l3kernel (5jun11)

git-svn-id: svn://tug.org/texlive/trunk@23282 c570f23f-e606-0410-a88d-b1316a301751

60 files changed, 51935 insertions, 0 deletions

diff --git a/Master/texmf-dist/doc/latex/l3kernel/README b/Master/texmf-dist/doc/latex/l3kernel/README
new file mode 100644
index 00000000000..9281d08f4c1
--- /dev/null
+++ b/Master/texmf-dist/doc/latex/l3kernel/README
@@ -0,0 +1,89 @@
+
+   An Experimental LaTeX3 Programming Convention
+   =============================================
+
+   2011/05/10
+
+
+WHERE TO GET IT
+---------------
+
+The files in this distribution represent a snapshot of selected files
+from the Subversion (SVN) repository of the LaTeX3 Project.
+
+To obtain current versions of the files, visit
+<http://www.latex-project.org/code.html> which contains further
+instructions.
+
+OVERVIEW
+--------
+
+The files of the expl3 bundle provide a low-level API for TeX
+programmers with special syntax conventions, completely separating it
+from document level syntax. Hence, the commands provided are not
+intended for use at the document level nor for use in describing
+design layouts in document class files.
+
+This API provides the foundation on which the LaTeX3 kernel and other
+advanced extensions are built. Special care has been taken so that
+they can be used within a LaTeX2e context as regular packages.
+
+While expl3 is still experimental, the bundle is now regarded as
+broadly stable. The syntax conventions and functions provided are now
+ready for wider use. There may still be changes to some functions, but
+these will be minor when compared to the scope of expl3.
+
+REQUIREMENTS
+------------
+
+The expl3 bundle requires the e-TeX extensions and the functionality
+of the \pdfstrcmp primitive. As a result, the bundle will only work
+with the following engines:
+
+ - pdfTeX v1.30 or later
+ - XeTeX v0.9994 or later
+ - LuaTeX v0.40 or later
+ 
+pdfTeX v1.30 was released in 2005, and so any recent TeX distribution
+will support expl3. Both XeTeX and LuaTeX have developed more
+actively over the past few years, and for this reason only recent 
+releases of these engines are supported.
+
+THE GUILTY PERSONS
+------------------
+
+   Frank Mittelbach, Denys Duchier, Johannes Braams, Michael Downes,
+   David Carlisle, Alan Jeffrey, Chris Rowley, Rainer Schoepf 
+   Javier Bezos, Morten Hoegholm, Thomas Lotze, Will Robertson,
+   Joseph Wright, Bruno Le Floch
+
+
+DISCUSSION
+----------
+
+Discussion concerning the approach, suggestions for improvements, 
+changes, additions, etc. should be addressed to the list LaTeX-L. 
+
+You can subscribe to this list by sending mail to
+
+  listserv@urz.uni-heidelberg.de
+
+with the body containing
+
+  subscribe LATEX-L  <Your-First-Name> <Your-Second-Name>
+
+
+BUGS
+----
+
+The issue tracker for LaTeX3 bugs is currently located at
+
+  https://github.com/latex3/svn-mirror/issues
+  
+Please report specific issues with LaTeX3 code there. More general
+discussion should be directed to the LaTeX-L lists.
+
+=====================================================================
+
+--- Copyright 1998-2011 The LaTeX3 Project. All rights reserved ---
+
diff --git a/Master/texmf-dist/doc/latex/l3kernel/expl3.pdf b/Master/texmf-dist/doc/latex/l3kernel/expl3.pdf
new file mode 100644
index 00000000000..875541e93ba
--- /dev/null
+++ b/Master/texmf-dist/doc/latex/l3kernel/expl3.pdf
diff --git a/Master/texmf-dist/doc/latex/l3kernel/interface3.tex b/Master/texmf-dist/doc/latex/l3kernel/interface3.tex
new file mode 100644
index 00000000000..f4c4fa4e2bf
--- /dev/null
+++ b/Master/texmf-dist/doc/latex/l3kernel/interface3.tex
@@ -0,0 +1,89 @@
+% \iffalse meta-comment
+%
+%% File: interfaces3.tex Copyright (C) 1990-2011 The LaTeX3 Project
+%%
+%% It may be distributed and/or modified under the conditions of the
+%% LaTeX Project Public License (LPPL), either version 1.3c of this
+%% license or (at your option) any later version.  The latest version
+%% of this license is in the file
+%%
+%%    http://www.latex-project.org/lppl.txt
+%%
+%% This file is part of the "expl3 bundle" (The Work in LPPL)
+%% and all files in that bundle must be distributed together.
+%%
+%% The released version of this bundle is available from CTAN.
+%%
+%% -----------------------------------------------------------------------
+%%
+%% The development version of the bundle can be found at
+%%
+%%    http://www.latex-project.org/svnroot/experimental/trunk/
+%%
+%% for those people who are interested.
+%%
+%%%%%%%%%%%
+%% NOTE: %%
+%%%%%%%%%%%
+%%
+%%   Snapshots taken from the repository represent work in progress and may
+%%   not work or may contain conflicting material!  We therefore ask
+%%   people _not_ to put them into distributions, archives, etc. without
+%%   prior consultation with the LaTeX3 Project.
+%%
+%% -----------------------------------------------------------------------
+%
+% \fi
+
+%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
+%
+% This document will typeset the LaTeX3 interface descriptions a single
+% document. This will produce quite a large file (more than 170 pages
+% currently).
+%
+% There is also a full version of the sources (source3.tex) which additionally
+% also typesets the command implementations.
+%
+% Do not forget to generate the index (as explained on the terminal output
+% near the end of the run)!
+%
+%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
+
+
+\documentclass{l3doc} 
+\listfiles
+
+\begin{document}
+
+\title{The \LaTeX3 Interfaces}
+\author{%
+ The \LaTeX3 Project\thanks
+   {%
+     E-mail:
+       \href{mailto:latex-team@latex-project.org}
+         {latex-team@latex-project.org}%
+   }%
+}
+
+\pagenumbering{roman}
+\maketitle
+
+%
+% First load all modules and typeset the documentation parts
+%
+
+\input{source3body}    % all the individual modules
+
+\clearpage
+
+\begingroup
+  \def\endash{--}
+  \catcode`\-\active
+  \def-{\futurelet\temp\indexdash}
+  \def\indexdash{\ifx\temp-\endash\fi}
+  \DelayPrintIndex
+\endgroup
+
+\end{document}
+
+
diff --git a/Master/texmf-dist/doc/latex/l3kernel/source3.pdf b/Master/texmf-dist/doc/latex/l3kernel/source3.pdf
new file mode 100644
index 00000000000..946105cd5c8
--- /dev/null
+++ b/Master/texmf-dist/doc/latex/l3kernel/source3.pdf
diff --git a/Master/texmf-dist/doc/latex/l3kernel/source3.tex b/Master/texmf-dist/doc/latex/l3kernel/source3.tex
new file mode 100644
index 00000000000..927e483cce9
--- /dev/null
+++ b/Master/texmf-dist/doc/latex/l3kernel/source3.tex
@@ -0,0 +1,106 @@
+% \iffalse meta-comment
+%
+%% File: source3.tex Copyright (C) 1990-2011 The LaTeX3 Project
+%%
+%% It may be distributed and/or modified under the conditions of the
+%% LaTeX Project Public License (LPPL), either version 1.3c of this
+%% license or (at your option) any later version.  The latest version
+%% of this license is in the file
+%%
+%%    http://www.latex-project.org/lppl.txt
+%%
+%% This file is part of the "expl3 bundle" (The Work in LPPL)
+%% and all files in that bundle must be distributed together.
+%%
+%% The released version of this bundle is available from CTAN.
+%%
+%% -----------------------------------------------------------------------
+%%
+%% The development version of the bundle can be found at
+%%
+%%    http://www.latex-project.org/svnroot/experimental/trunk/
+%%
+%% for those people who are interested.
+%%
+%%%%%%%%%%%
+%% NOTE: %%
+%%%%%%%%%%%
+%%
+%%   Snapshots taken from the repository represent work in progress and may
+%%   not work or may contain conflicting material!  We therefore ask
+%%   people _not_ to put them into distributions, archives, etc. without
+%%   prior consultation with the LaTeX3 Project.
+%%
+%% -----------------------------------------------------------------------
+%
+% \fi
+
+%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
+%
+% This document will typeset the LaTeX3 sources as a single document.
+% This will produce quite a large file (more than 400 pages).
+%
+% There is also a shorter version (interface3.tex) that only typesets the
+% command % interface descriptions.
+%
+% Do not forget to generate the index (as explained on the terminal output
+% near the end of the run)!
+%
+%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
+
+\documentclass{l3doc}
+\listfiles
+
+\begin{document}
+
+\title{The \LaTeX3 Sources}
+\author{%
+ The \LaTeX3 Project\thanks
+   {%
+     E-mail:
+       \href{mailto:latex-team@latex-project.org}
+         {latex-team@latex-project.org}%
+   }%
+}
+
+\pagenumbering{roman}
+\maketitle
+
+%
+% First load all modules and typeset the documentation parts
+%
+
+\input{source3body}    % all the individual modules
+
+%
+% Now reload all modules and typeset the implementation parts
+%
+
+\part{Implementation}
+
+\def\maketitle{}
+\EnableImplementation
+\DisableDocumentation
+\DocInputAgain
+
+\clearpage
+\pagestyle{headings}
+
+% Make TeX shut up.
+\hbadness=10000
+\newcount\hbadness
+\hfuzz=\maxdimen
+
+\PrintChanges
+\clearpage
+
+\begingroup
+  \def\endash{--}
+  \catcode`\-\active
+  \def-{\futurelet\temp\indexdash}
+  \def\indexdash{\ifx\temp-\endash\fi}
+
+  \DelayPrintIndex
+\endgroup
+
+\end{document}
\ No newline at end of file
diff --git a/Master/texmf-dist/doc/latex/l3kernel/source3body.tex b/Master/texmf-dist/doc/latex/l3kernel/source3body.tex
new file mode 100644
index 00000000000..fbffe013ae8
--- /dev/null
+++ b/Master/texmf-dist/doc/latex/l3kernel/source3body.tex
@@ -0,0 +1,376 @@
+% \iffalse meta-comment
+%
+%% File: source3body.tex Copyright (C) 1990-2011 The LaTeX3 Project
+%%
+%% It may be distributed and/or modified under the conditions of the
+%% LaTeX Project Public License (LPPL), either version 1.3c of this
+%% license or (at your option) any later version.  The latest version
+%% of this license is in the file
+%%
+%%    http://www.latex-project.org/lppl.txt
+%%
+%% This file is part of the "expl3 bundle" (The Work in LPPL)
+%% and all files in that bundle must be distributed together.
+%%
+%% The released version of this bundle is available from CTAN.
+%%
+%% -----------------------------------------------------------------------
+%%
+%% The development version of the bundle can be found at
+%%
+%%    http://www.latex-project.org/svnroot/experimental/trunk/
+%%
+%% for those people who are interested.
+%%
+%%%%%%%%%%%
+%% NOTE: %%
+%%%%%%%%%%%
+%%
+%%   Snapshots taken from the repository represent work in progress and may
+%%   not work or may contain conflicting material!  We therefore ask
+%%   people _not_ to put them into distributions, archives, etc. without
+%%   prior consultation with the LaTeX3 Project.
+%%
+%% -----------------------------------------------------------------------
+%
+% \fi
+
+%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
+%
+% This file is used by
+%
+%  source3.tex            % documentation including implementation
+%
+%  interface3.tex         % only interface documentation
+%
+%
+%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
+
+
+\begin{abstract}
+
+\setlength\parindent{0pt}
+\setlength\parskip{\baselineskip}
+
+\noindent
+This is the reference documentation for the \pkg{expl3} 
+programming environment. The \pkg{expl3} modules set up an experimental
+naming scheme for \LaTeX{} commands, which allow the \LaTeX{} programmer
+to systematically name functions and variables, and specify the argument
+types of functions.
+
+The \TeX{} and \eTeX{} primitives are all given a new name according to
+these conventions. However, in the main direct use of the primitives is
+not required or encouraged: the \pkg{expl3} modules define an
+independent low-level \LaTeX3 programming language.
+
+At present, the \pkg{expl3} modules are designed to be loaded on top of
+\LaTeXe{}. In time, a \LaTeX3 format will be produced based on this code.
+This allows the code to be used in \LaTeXe{} packages \emph{now} while a
+stand-alone \LaTeX3 is developed.
+
+\begin{bfseries}
+  While \pkg{expl3} is still experimental, the bundle is now regarded as
+  broadly stable. The syntax conventions and functions provided are now
+  ready for wider use. There may still be changes to some functions, but
+  these will be minor when compared to the scope of \pkg{expl3}.
+
+  New modules will be added to the distributed version of \pkg{expl3} as
+  they reach maturity.
+\end{bfseries}
+
+\end{abstract}
+
+\clearpage
+
+{%
+  \def\\{:}% fix "newlines" in the ToC
+  \tableofcontents
+}
+
+\clearpage
+\pagenumbering{arabic}
+
+%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
+%
+% Each of the following \DocInput lines includes a file with extension
+% .dtx. Each of these files may be typeset separately. For instance
+%   pdflatex l3box.dtx
+% will typeset the source of the LaTeX3 box commands. If you use the
+% Makefile, the index will be generated automatically; e.g.,
+%   make doc F=l3box
+%
+% If this file is processed, each of these separate dtx files will be
+% contained as a part of a single document.
+%
+%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
+
+\makeatletter
+\def\partname{Part}
+\def\maketitle{\part{\@title}}
+\let\thanks\@gobble
+\let\DelayPrintIndex\PrintIndex
+\let\PrintIndex\@empty
+\makeatother
+
+\part{Introduction to \pkg{expl3} and this document}
+
+This document is intended to act as a comprehensive reference manual 
+for the \pkg{expl3} language. A general guide to the \LaTeX3 
+programming language is found in \href{expl3.pdf}{expl3.pdf}.
+
+\section{Naming functions and variables} 
+
+\LaTeX3 does not use \texttt{@} as a \enquote{letter} for defining
+internal macros.  Instead, the symbols |_| and \texttt{:}
+are used in internal macro names to provide structure. The name of
+each \emph{function} is divided into logical units using \texttt{_}, 
+while \texttt{:} separates the \emph{name} of the function from the 
+\emph{argument specifier} (\enquote{arg-spec}). This describes the arguments 
+expected by the function. In most cases, each argument is represented 
+by a single letter. The complete list of arg-spec letters for a function
+is referred to as the \emph{signature} of the function.
+
+Each function name starts with the \emph{module} to which it belongs.
+Thus apart from a small number of very basic functions, all \pkg{expl3}
+function names contain at least one underscore to divide the module
+name from the descriptive name of the function. For example, all
+functions concerned with comma lists are in module \texttt{clist} and
+begin \cs{clist_}.
+
+Every function must include an argument specifier. For functions which
+take no arguments, this will be blank and the function name will end
+\texttt{:}. Most functions take one or more arguments, and use the
+following argument specifiers:
+\begin{description}
+  \item[\texttt{D}] The \texttt{D} specifier means \emph{do not use}. 
+    All of the \TeX{} primitives are initially \cs{let} to a \texttt{D}
+    name, and some are then given a second name.  Only the kernel 
+    team should use anything with a \texttt{D} specifier!
+  \item[\texttt{N} and \texttt{n}] These mean \emph{no manipulation},
+    of a single token for \texttt{N} and of a set of tokens given in 
+    braces for \texttt{n}. Both pass the argument though exactly as 
+    given. Usually, if you use a single token for an \texttt{n} argument, 
+    all will be well.
+  \item[\texttt{c}] This means \emph{csname}, and indicates that the 
+    argument will be turned into a csname before being used. So
+    So \cs{foo:c} |{ArgumentOne}| will act in the same way as \cs{foo:N}
+    \cs{ArgumentOne}.
+  \item[\texttt{V} and \texttt{v}] These mean \emph{value 
+    of variable}. The \texttt{V} and \texttt{v} specifiers are used to 
+    get the content of a variable without needing to worry about the
+    underlying \TeX{} structure containing the data. A \texttt{V} 
+    argument will be a single token (similar to \texttt{N}), for example
+    \cs{foo:V} \cs{MyVariable}; on the other hand, using \texttt{v} a 
+    csname is constructed first, and then the value is recovered, for 
+    example \cs{foo:v} |{MyVariable}|.
+  \item[\texttt{o}] This means \emph{expansion once}. In general, the
+    \texttt{V} and \texttt{v} specifiers are favoured over \texttt{o}
+    for recovering stored information. However, \texttt{o} is useful
+    for correctly processing information with delimited arguments.
+  \item[\texttt{x}] The \texttt{x} specifier stands for \emph{exhaustive 
+    expansion}: the plain \TeX{} \cs{edef}.
+  \item[\texttt{f}] The \texttt{f} specifier stands for \emph{full 
+    expansion}, and in contrast to \emph{x} stops at the first 
+    non-expandable item without trying to execute it.
+  \item[\texttt{T} and \texttt{F}] For logic tests, there are the branch
+    specifiers \texttt{T} (\emph{true}) and \texttt{F} (\emph{false}). 
+    Both specifiers treat the input in the same way as \texttt{n} (no 
+    change), but make the logic much easier to see.
+  \item[\texttt{p}] The letter \texttt{p} indicates \TeX{} 
+    \emph{parameters}. Normally this will be used for delimited 
+    functions as \pkg{expl3} provides better methods for creating simple
+    sequential arguments.    
+  \item[\texttt{w}] Finally, there is the \texttt{w} specifier for 
+    \emph{weird} arguments. This covers everything else, but mainly 
+    applies to delimited values (where the argument must be terminated 
+    by some arbitrary string).
+\end{description}
+Notice that the argument specifier describes how the argument is 
+processed prior to being passed to the underlying function. For example,
+\cs{foo:c} will take its argument, convert it to a control sequence and
+pass it to \cs{foo:N}.
+
+Variables are named in a similar manner to functions, but begin with
+a single letter to define the type of variable:
+\begin{description}
+  \item[\texttt{c}] Constant: global parameters whose value should not
+    be changed.
+  \item[\texttt{g}] Parameters whose value should only be set globally.
+  \item[\texttt{l}] Parameters whose value should only be set locally.
+\end{description}
+Each variable name is then build up in a similar way to that of a 
+function, typically starting with the module\footnote{The module names are
+  not used in case of generic scratch registers defined in the data
+  type modules, e.g., the
+  \texttt{int} module contains some scratch variables called \cs{l_tmpa_int},
+  \cs{l_tmpb_int}, and so on. In such a case adding the module name up front
+  to denote the module
+  and in the back to indicate the type, as in
+  \cs{l_int_tmpa_int} would be very unreadable.}  name
+and then a descriptive part.
+Variables end with a short identifier to show the variable type:
+\begin{description}
+  \item[\texttt{bool}] Either true or false.
+  \item[\texttt{box}] Box register.
+  \item[\texttt{clist}] Comma separated list.
+  \item[\texttt{dim}] \enquote{Rigid} lengths.
+  \item[\texttt{int}] Integer-valued count register.
+  \item[\texttt{prop}] Property list.
+  \item[\texttt{skip}] \enquote{Rubber} lengths.
+  \item[\texttt{seq}] \enquote{Sequence}: a data-type used to implement lists 
+    (with access at both ends) and stacks.
+  \item[\texttt{stream}] An input or output stream (for reading from or 
+    writing to, respectively).
+  \item[\texttt{tl}] Token list variables: placeholder for a token list.
+  \item[\texttt{toks}] Token register.
+\end{description}
+
+\subsection{Terminological inexactitude}
+
+A word of warning. In this document, and others referring to the \pkg{expl3}
+programming modules, we often refer to \enquote{variables} and 
+\enquote{functions} as if they
+were actual constructs from a real programming language.  In truth, \TeX{}
+is a macro processor, and functions are simply macros that may or may not take
+arguments and expand to their replacement text.  Many of the common variables
+are \emph{also} macros, and if placed into the input stream will simply expand
+to their definition as well~--- a \enquote{function} with no arguments and a 
+\enquote{token list variable} are in truth one and the same.  On the other
+hand, some \enquote{variables} are actually registers that must be
+initialised and their values set and retrieved with specific functions.
+
+The conventions of the \pkg{expl3} code are designed to clearly separate the
+ideas of \enquote{macros that contain data} and 
+\enquote{macros that contain code}, and a
+consistent wrapper is applied to all forms of \enquote{data} whether they be
+macros or
+actually registers.  This means that sometimes we will use phrases like 
+\enquote{the function returns a value}, when actually we just mean 
+\enquote{the macro expands to something}. Similarly, the term 
+\enquote{execute} might be used in place of \enquote{expand}
+or it might refer to the more specific case of
+\enquote{processing in \TeX's stomach}
+(if you are familiar with the \TeX{}book parlance).
+
+If in doubt, please ask; chances are we've been hasty in writing certain
+definitions and need to be told to tighten up our terminology.
+
+\section{Documentation conventions}
+
+This document is typeset with the experimental \pkg{l3doc} class; 
+several conventions are used to help describe the features of the code.
+A number of conventions are used here to make the documentation clearer.
+
+Each group of related functions is given in a box. For a function with
+a \enquote{user} name, this might read:
+\begin{function}{\ExplSyntaxOn, \ExplSyntaxOff}
+  \begin{syntax}
+    \cs{ExplSyntaxOn} \dots{} \cs{ExplSyntaxOff}
+  \end{syntax}
+  The textual description of how the function works would appear here. The 
+  syntax of the function is shown in mono-spaced text to the right of 
+  the box. In this example, the function takes no arguments and so the 
+  name of the function is simply reprinted.
+\end{function}
+
+For programming functions, which use \texttt{_} and \texttt{:} in their name
+there are a few additional conventions: If two related functions are given
+with identical names but different argument specifiers, these are termed
+\emph{variants} of each other, and the latter functions are printed in grey to
+show this more clearly. They will carry out the same function but will take
+different types of argument:
+\begin{function}{\seq_new:N, \seq_new:c}
+  \begin{syntax}
+    \cs{seq_new:N} \meta{sequence}
+  \end{syntax}
+  When a number of variants are described, the arguments are usually
+  illustrated only for the base function. Here, \meta{sequence} indicates 
+  that \cs{seq_new:N} expects the name of a sequence. From the argument
+  specifier, \cs{seq_new:c} also expects a sequence name, but as a 
+  name rather than as a control sequence. Each argument given in the
+  illustration should be described in the following text.
+\end{function}
+
+Some functions are fully expandable, which allows it to be used within 
+an \texttt{x}-type argument (in plain \TeX{} terms, inside an \cs{edef}).
+These fully expandable functions are indicated in the documentation by 
+a star:
+\begin{function}[EXP]{\cs_to_str:N}
+  \begin{syntax}
+    \cs{cs_to_str:N} \meta{cs}
+  \end{syntax}
+  As with other functions, some text should follow which explains how
+  the function works. Usually, only the star will indicate that the 
+  function is expandable. In this case, the function expects a \meta{cs}, 
+  shorthand for a \meta{control sequence}.
+\end{function}
+
+Conditional (\texttt{if}) functions are normally defined in three variants,
+with
+\texttt{T}, \texttt{F} and \texttt{TF} argument specifiers. This allows
+them to be used for different \enquote{true}/\enquote{false} branches,
+depending on 
+which outcome the conditional is being used to test. To indicate this 
+without repetition, this information is given in a shortened form:
+\begin{function}[EXP,pTF]{\xetex_if_engine:}
+  \begin{syntax}
+    \cs{xetex_if_engine:TF} \Arg{true code} \Arg{false code}
+  \end{syntax}
+  The underlining and italic of \texttt{TF} indicates that 
+  \cs{xetex_if_engine:T}, \cs{xetex_if_engine:F} and 
+  \cs{xetex_if_engine:TF} are all available. Usually, the illustration 
+  will use the \texttt{TF} variant, and so both \meta{true code}
+  and \meta{false code} will be shown. The two variant forms \texttt{T} and
+  \texttt{F} take only \meta{true code} and \meta{false code}, respectively.
+  Here, the star also shows that this function is expandable.
+  With some minor exceptions, \emph{all} conditional functions in the 
+  \pkg{expl3} modules should be defined in this way.
+\end{function}
+
+Variables, constants and so on are described in a similar manner:
+\begin{variable}{\l_tmpa_tl}
+  A short piece of text will describe the variable: there is no 
+  syntax illustration in this case.
+\end{variable}
+
+In some cases, the function is similar to one in \LaTeXe{} or plain \TeX{}.
+In these cases, the text will include an extra \enquote{\textbf{\TeX{}hackers 
+note}} section:
+\begin{function}[EXP]{\token_to_str:N}
+  \begin{syntax}
+    \cs{token_to_str:N} \meta{token}
+  \end{syntax}
+  The normal description text.
+  \begin{texnote}
+    Detail for the experienced \TeX{} or \LaTeXe\ programmer. In this 
+    case, it would point out that this function is the \TeX{} primitive
+    \cs{string}.
+  \end{texnote}
+\end{function}
+
+\DisableImplementation
+
+\DocInput{l3bootstrap.dtx}
+\DocInput{l3names.dtx}
+\DocInput{l3basics.dtx}
+\DocInput{l3expan.dtx}
+\DocInput{l3prg.dtx}
+\DocInput{l3quark.dtx}
+\DocInput{l3token.dtx}
+\DocInput{l3int.dtx}
+\DocInput{l3skip.dtx}
+\DocInput{l3tl.dtx}
+\DocInput{l3seq.dtx}
+\DocInput{l3clist.dtx}
+\DocInput{l3prop.dtx}
+\DocInput{l3box.dtx}
+\DocInput{l3io.dtx}
+\DocInput{l3msg.dtx}
+\DocInput{l3keyval.dtx}
+\DocInput{l3keys.dtx}
+\DocInput{l3file.dtx}
+\DocInput{l3fp.dtx}
+\DocInput{l3luatex.dtx}
+
+\endinput
diff --git a/Master/texmf-dist/makeindex/l3kernel/l3doc.ist b/Master/texmf-dist/makeindex/l3kernel/l3doc.ist
new file mode 100644
index 00000000000..e1496e34d73
--- /dev/null
+++ b/Master/texmf-dist/makeindex/l3kernel/l3doc.ist
@@ -0,0 +1,72 @@
+%%
+%% This is file `l3doc.ist',
+%% generated with the docstrip utility.
+%%
+%% The original source files were:
+%%
+%% l3doc.dtx  (with options: `docist')
+%% 
+%% 
+%% EXPERIMENTAL CODE
+%% 
+%% Do not distribute this file without also distributing the
+%% source files specified above.
+%% 
+%% Do not distribute a modified version of this file.
+%% 
+%% 
+%% File: l3doc.dtx Copyright (C) 1990-2011 The LaTeX3 project
+%%
+%% It may be distributed and/or modified under the conditions of the
+%% LaTeX Project Public License (LPPL), either version 1.3c of this
+%% license or (at your option) any later version.  The latest version
+%% of this license is in the file
+%%
+%%    http://www.latex-project.org/lppl.txt
+%%
+%% This file is part of the "expl3 bundle" (The Work in LPPL)
+%% and all files in that bundle must be distributed together.
+%%
+%% The released version of this bundle is available from CTAN.
+%%
+%% -----------------------------------------------------------------------
+%%
+%% The development version of the bundle can be found at
+%%
+%%    http://www.latex-project.org/svnroot/experimental/trunk/
+%%
+%% for those people who are interested.
+%%
+%%%%%%%%%%%
+%% NOTE: %%
+%%%%%%%%%%%
+%%
+%%   Snapshots taken from the repository represent work in progress and may
+%%   not work or may contain conflicting material!  We therefore ask
+%%   people _not_ to put them into distributions, archives, etc. without
+%%   prior consultation with the LaTeX3 Project.
+%%
+%% -----------------------------------------------------------------------
+actual '='
+quote '!'
+level '#'
+preamble
+"\n \\begin{theindex} \n \\makeatletter\\scan@allowedfalse\n"
+postamble
+"\n\n \\end{theindex}\n"
+item_x1   "\\efill \n \\subitem "
+item_x2   "\\efill \n \\subsubitem "
+delim_0   "\\pfill "
+delim_1   "\\pfill "
+delim_2   "\\pfill "
+lethead_prefix   "{\\bfseries\\hfil "
+lethead_suffix   "\\hfil}\\nopagebreak\n"
+lethead_flag       1
+heading_prefix   "{\\bfseries\\hfil "
+heading_suffix   "\\hfil}\\nopagebreak\n"
+headings_flag       1
+
+page_precedence "rnaA"
+%% 
+%%
+%% End of file `l3doc.ist'.
diff --git a/Master/texmf-dist/source/latex/l3kernel/expl3.dtx b/Master/texmf-dist/source/latex/l3kernel/expl3.dtx
new file mode 100644
index 00000000000..443c87b3916
--- /dev/null
+++ b/Master/texmf-dist/source/latex/l3kernel/expl3.dtx
@@ -0,0 +1,1099 @@
+% \iffalse meta-comment
+%
+%% File: expl3.dtx Copyright (C) 1990-2011 The LaTeX3 Project
+%%
+%% It may be distributed and/or modified under the conditions of the
+%% LaTeX Project Public License (LPPL), either version 1.3c of this
+%% license or (at your option) any later version.  The latest version
+%% of this license is in the file
+%%
+%%    http://www.latex-project.org/lppl.txt
+%%
+%% This file is part of the "expl3 bundle" (The Work in LPPL)
+%% and all files in that bundle must be distributed together.
+%%
+%% The released version of this bundle is available from CTAN.
+%%
+%% -----------------------------------------------------------------------
+%%
+%% The development version of the bundle can be found at
+%%
+%%    http://www.latex-project.org/svnroot/experimental/trunk/
+%%
+%% for those people who are interested.
+%%
+%%%%%%%%%%%
+%% NOTE: %%
+%%%%%%%%%%%
+%%
+%%   Snapshots taken from the repository represent work in progress and may
+%%   not work or may contain conflicting material!  We therefore ask
+%%   people _not_ to put them into distributions, archives, etc. without
+%%   prior consultation with the LaTeX3 Project.
+%%
+%% -----------------------------------------------------------------------
+%
+%<*driver|package>
+\RequirePackage{l3names}
+\GetIdInfo$Id: expl3.dtx 2478 2011-06-19 21:34:23Z joseph $
+  {L3 Experimental code bundle wrapper}
+%</driver|package>
+%<*driver>
+\documentclass[full]{l3doc}
+\begin{document}
+  \DocInput{\jobname.dtx}
+\end{document}
+%</driver>
+% \fi
+% 
+% \providecommand\acro[1]{\textsc{\MakeLowercase{#1}}}
+% \newenvironment{arg-description}{%
+%   \begin{itemize}\def\makelabel##1{\hss\llap{\bfseries##1}}}{\end{itemize}}
+% 
+% \title{^^A
+%   The \textsf{expl3} package and \LaTeX3 programming^^A
+%   \thanks{This file describes v\ExplFileVersion,
+%     last revised \ExplFileDate.}^^A
+% }
+%         
+% \author{^^A
+%  The \LaTeX3 Project\thanks
+%    {^^A
+%      E-mail:
+%        \href{mailto:latex-team@latex-project.org}
+%          {latex-team@latex-project.org}^^A
+%    }^^A
+% }
+%
+% \date{Released \ExplFileDate}
+%
+% \maketitle
+%
+% \begin{documentation}
+%
+% \begin{abstract}
+% 
+% This document gives an introduction to a new set of programming
+% conventions that have been designed to meet the requirements of
+% implementing large scale \TeX{} macro programming projects such as
+% \LaTeX{}. These programming conventions are the base layer of \LaTeX3.
+% 
+% The main features of the system described are:
+% \begin{itemize}
+%   \item classification of the macros (or, in \LaTeX{} terminology,
+%     commands) into \LaTeX{} functions and \LaTeX{} parameters, and also
+%     into modules containing related commands;
+%   \item  a systematic naming scheme based on these classifications;
+%   \item  a simple mechanism for controlling the expansion of a function's
+%     arguments.
+% \end{itemize}
+% This system is being used as the basis for \TeX{} programming within 
+% the \LaTeX3 project. Note that the language is not intended for either
+% document mark-up or style specification. Instead, it is intended that
+% such features will be built on top of the conventions described here.
+%
+% This document is an introduction to the ideas behind the \pkg{expl3}
+% programming interface. For the complete documentation of the programming
+% layer provided by the \LaTeX3 Project, see the accompanying \texttt{source3}
+% document.
+%
+% \end{abstract} 
+% 
+% \section{Introduction}
+% 
+% The first step to develop a \LaTeX{} kernel beyond \LaTeXe{} is to
+% address how the underlying system is programmed.  Rather than the 
+% current mix of \LaTeX{} and \TeX{} macros, the \LaTeX3 system provides 
+% its own consistent interface to all of the functions needed to 
+% control \TeX{}.  A key part of this work is to ensure that everything
+% is documented, so that \LaTeX{} programmers and users can work 
+% efficiently without needing to be familiar with the internal nature
+% of the kernel or with plain \TeX{}.
+% 
+% The \pkg{expl3} bundle provides this new programming interface for
+% \LaTeX. To make programming systematic, \LaTeX3 uses some very
+% different conventions to \LaTeXe{} or plain TeX{}. As a result,
+% programmers starting with \LaTeX3 will need to become familiar with
+% the syntax of the new language.
+% 
+% The next section shows where this language fits into a complete
+% \TeX{}-based document processing system.  We then describe the major
+% features of the syntactic structure of command names, including the
+% argument specification syntax used in function names.
+% 
+% The practical ideas behind this argument syntax will be explained,
+% together with the expansion control mechanism and the interface 
+% used to define variant forms of functions.  
+% 
+% As we shall demonstrate, the use of a structured naming scheme and of
+% variant forms for functions greatly improves the readability of the
+% code and hence also its reliability.  Moreover, experience has shown
+% that the longer command names which result from the new syntax do not
+% make the process of \emph{writing} code significantly harder.
+% 
+% \section{Languages and interfaces}
+% 
+% It is possible to identify several distinct languages related to the
+% various interfaces that are needed in a \TeX{}-based document processing
+% system.  This section looks at those we consider most important for
+% the \LaTeX3 system.
+% 
+% \begin{description}
+%   \item[Document mark-up] This comprises those commands (often called 
+%     tags) that are to embedded in the document (the |.tex| file).
+% 
+%     It is generally accepted that such mark-up should be essentially
+%     \emph{declarative}. It may be traditional \TeX{}-based mark-up such
+%      as \LaTeXe{}, as described in~\cite{A-W:LLa94} and~\cite{A-W:GMS94},
+%     or a mark-up language defined via \acro{HTML} or \acro{XML}.
+%   
+%     One problem with more traditional \TeX{} coding conventions (as
+%     described in~\cite{A-W:K-TB}) is that the names and syntax of \TeX{}'s
+%     primitive formatting commands are ingeniously designed to be
+%     \enquote{natural} when used directly by the author as document mark-up or in
+%     macros.  Ironically, the ubiquity (and widely recognised
+%     superiority) of logical mark-up has meant that such explicit
+%     formatting commands are almost never needed in documents or in
+%     author-defined macros.  Thus they are used almost exclusively by
+%     \TeX{} programmers to define higher-level commands, and their
+%     idiosyncratic syntax is not at all popular with this community.
+%     Moreover, many of them have names that could be very useful as
+%     document mark-up tags were they not pre-empted as primitives 
+%     (\emph{e.g.}~\cs{box} or \cs{special}).
+%   
+%   \item[Designer interface] This relates a (human) typographic
+%     designer's specification for a document to a program that 
+%     \enquote{formats
+%     the document}.  It should ideally use a declarative language that
+%     facilitates expression of the relationship and spacing rules 
+%     specified for the layout of the various document elements.
+% 
+%     This language is not embedded in document text and it will be very
+%     different in form to the document mark-up language.  For \LaTeX, 
+%     this level was almost completely missing  from \LaTeX2.09; \LaTeXe{} 
+%     made some improvements in this area but it is still the case that 
+%     implementing a design specification in  \LaTeX{} requires far more 
+%     \enquote{low-level} coding than is acceptable.
+%   
+%   \item[Programmer interface] 
+%     This language is the implementation language within which the 
+%     basic typesetting functionality is implemented, building upon the 
+%     primitives of \TeX{} (or a  successor program).  It may also be used
+%     to implement the previous two languages \enquote{within} \TeX{}, as in
+%     the
+%     current \LaTeX{} system.
+%   
+% \end{description}
+% 
+% The last layer is covered by the conventions described in this 
+% document, which describes a system aimed at providing a suitable 
+% basis for coding \LaTeX3. Its main distinguishing features are
+% summarised here: 
+% \begin{itemize}
+%   \item A consistent naming scheme for all commands, including \TeX{}
+%     primitives.
+%   \item The classification of commands as \LaTeX{} functions or \LaTeX{}
+%     parameters, and also their division into modules according to their
+%     functionality.
+%   \item A simple mechanism for controlling argument expansion.
+%   \item Provision of a set of core \LaTeX{} functions that is sufficient
+%     for handling programming constructs such as queues, sets, stacks, 
+%     property lists.
+%   \item A \TeX{} programming environment in which, for example, all
+%     white space is ignored.
+% \end{itemize}
+% 
+% \section{The naming scheme}
+% 
+% \LaTeX3 does not use |@| as a \enquote{letter} for defining
+% internal macros.  Instead, the symbols |_| and |:|
+% are used in internal macro names to provide structure. In
+% contrast to the plain \TeX{} format and the \LaTeXe{} kernel, these
+% extra letters are used only between parts of a macro name (no
+% strange vowel replacement).
+%
+% While \TeX{} is actually a macro processor, by
+% convention for the \pkg{expl3} programming language we distinguish between
+% \emph{functions} and \emph{variables}. Functions can have arguments and they
+% are either expanded or executed.  Variables can be assigned values and they
+% are used in arguments to functions; they are not used directly but are
+% manipulated by functions (including getting and setting functions).
+% Functions and variables with a related functionality (for example accessing
+% counters, or manipulating token lists, \emph{etc.})\ are collected together
+% into a
+% \emph{module}.
+%
+% \subsection{Examples}
+% 
+% Before giving the details of the naming scheme, here are a few typical
+% examples to indicate the flavour of the scheme; first some variable
+% names.
+% \begin{quote}
+%   \cs{l_tmpa_box} is a local variable (hence the~|l_| prefix)
+%     corresponding to a box register.\\
+%   \cs{g_tmpa_int} is a global variable (hence the~|g_| prefix)
+%     corresponding to an integer register (i.e.~a \TeX{} count 
+%     register).\\
+%   \cs{c_empty_tl} is the constant~(|c_|) token list variable
+%     that is always empty.
+% \end{quote}
+%
+% Now here is an example of a typical function name.
+% 
+% \cs{seq_push:Nn} is the function which puts the token list specified 
+% by its second argument onto the stack specified by its first argument.
+% The different natures of the two arguments are indicated by the~|:Nn|
+% suffix. The first argument must be a single token which \enquote{names}
+% the stack parameter: such single-token arguments are denoted~|N|.
+% The second argument is a normal \TeX{} \enquote{undelimited argument},
+% which
+% may either be a single token or a balanced, brace-delimited token
+% list (which we shall here call a \textit{braced token list}): the~|n|
+% denotes such a \enquote{normal} argument form. The name of the function
+% indicates it belongs to the |seq| module.
+% 
+% \subsection{Formal naming syntax}
+% 
+% We shall now look in more detail at the syntax of these names. A
+% function name in \LaTeX3 will have a name consisting of three parts:
+% \begin{quote}
+%   |\|\meta{module}|_|\meta{description}|:|\meta{arg-spec}
+% \end{quote}
+% while a variable will have (up to) four distinct parts to its name:
+% \begin{quote}
+%   |\|\meta{scope}|_|\meta{module}|_|\meta{description}|_|\meta{type}
+% \end{quote}
+%
+% The syntax of all names contains
+% \begin{quote}
+%   \meta{module} and \meta{description}
+% \end{quote}
+% these both give information about the command.
+% 
+% A \emph{module} is a collection of closely related functions and 
+% variables. Typical module names include~|int| for integer parameters
+% and related functions,~|seq| for sequences and~|box| for boxes.
+% 
+% Packages providing new programming functionality will add new modules
+% as needed; the programmer can choose any unused name, consisting
+% of letters only, for a module. In general, the module name and module
+% prefix should be related: for example, the kernel module containing
+% \texttt{box} functions is called \texttt{l3box}.
+% 
+% The \emph{description} gives more detailed information about the
+% function or parameter, and provides a unique name for it.  It should
+% consist of letters and, possibly,~|_|~characters. In general, the 
+% description should use |_| to divide up \enquote{words} or other easy to 
+% follow parts of the name.  For example, the \LaTeX3 kernel provides
+% \cs{if_cs_exist:N} which, as might be expected, tests if a command
+% name exists.
+% 
+% As a semi-formalized concept the letter |g| is sometimes used to
+% prefix certain parts of the \meta{description}
+% to mark the function as \enquote{globally acting}, \emph{e.g.},
+% \cs{int_set:Nn} is
+% a local operation while \cs{int_gset:Nn} is a global operation. This
+% of course goes hand in hand with when to use |l_| and |g_|
+% variable prefixes.
+%
+ %\subsubsection{Variables: scope and type}
+% 
+% The \meta{scope} part of the name describes how the variable can be
+% accessed.  Variables are classified as local, global or constant.  
+% This \emph{scope} type appears as a code at the beginning of the name;
+% the codes used are:
+% \begin{arg-description}
+%   \item[c] constants (global variables whose value should not be 
+%     changed);
+%   \item[g] variables whose value should only be set globally;
+%   \item[l] variables whose value should only be set locally.
+% \end{arg-description}
+%
+% Separate functions are provided to assign data to local and global
+% variables; for example, |\tl_set:Nn| and |\tl_gset:Nn| respectively
+% set the value of a local or global \enquote{token list} variable.
+% Note that it is a poor \TeX{} practise to intermix local and global
+% assignments to a variable; otherwise you risk exhausting the save stack.
+% \unskip\footnote{See \emph{The \TeX{}book}, p.\,301, for further
+% information.}
+% 
+% The \meta{type} will be in the list of available 
+% \emph{data-types};\footnote{Of course, if a totally new data type is
+% needed then this will not be the case. However, it is hoped that only
+% the kernel team will need to create new data types.} these include the
+% primitive \TeX{} data-types, such as the various registers, but to
+% these will be added data-types built within the \LaTeX{} programming
+% system.
+% 
+% The data types in \LaTeX3 are:
+% \begin{description}
+%   \item[bool]   either true or false (the \LaTeX3 implementation does
+%                 not use \cs{iftrue} or \cs{iffalse});
+%   \item[box]    box register;
+%   \item[clist]  comma separated list;
+%   \item[coffin] a \enquote{box with handles} --- a higher-level data
+%                 type for carrying out |box| alignment operations;
+%   \item[dim]    \enquote{rigid} lengths;
+%   \item[fp]     floating-point values;
+%   \item[int]    integer-valued count register;
+%   \item[prop]   property list;
+%   \item[seq]    sequence: a data-type used to implement lists (with 
+%                 access at both ends) and stacks;
+%   \item[skip]   \enquote{rubber} lengths;
+%   \item[stream] an input or output stream (for reading from or writing 
+%                 to, respectively);
+%   \item[tl]     \enquote{token list variables}: placeholders for token lists.
+% \end{description}
+% When the \meta{type} and \meta{module} are identical (as often happens in
+% the more basic modules) the \meta{module} part is often omitted for
+% aesthetic reasons.
+% 
+% The name \enquote{token list} may cause confusion, and so some background is 
+% useful.  \TeX{} works with tokens and lists of tokens, rather than 
+% characters. It provides two ways to store these token lists: within 
+% macros and as token registers (|toks|). The implementation in \LaTeX3 means
+% that
+% |toks| are not required, and that all operations for storing tokens can us
+% the |tl| variable type.
+% 
+% Experienced \TeX{} programmers will notice that some of the variable
+% types listed are native \TeX{} registers whilst others are not. In
+% general, the underlying \TeX{} implementation for a data structure may
+% vary but the \emph{documented interface} will be stable. For example,
+% the |prop| data type was originally implemented as a |toks|, but
+% is currently built on top of the |tl| data structure.
+% 
+% \subsubsection{Variables: guidance}
+% 
+% Although \TeX{} provides token registers for certain programming
+% tasks, \LaTeX3 is written so that token list variables can be used
+% for all stored lists of tokens. The \cs{toks_\ldots} functions are
+% therefore for internal use within \LaTeX3 and for supporting
+% \LaTeXe, and should not be needed in new \LaTeX3-based code.
+% 
+% Both comma lists and sequences both have similar characteristics.
+% They both use special delimiters to mark out one entry from the
+% next, and are both accessible at both ends. In general, it is
+% easier to create comma lists `by hand' as they can be typed
+% in directly. User input often takes the form of a comma separated
+% list and so there are many cases where this is the obvious
+% data type to use. On the other hand, sequences use special internal
+% tokens to separate entries. This means that they can be used to
+% contain material that comma lists cannot (such as items that may
+% themselves contain commas!). In general, comma lists should be
+% preferred for creating fixed lists inside programs and for
+% handling user input where commas will not occur. On the other
+% hand, sequences should be used to store arbitrary lists of
+% data.
+% 
+% \pkg{expl3} implements stacks using the sequence data structure.
+% Thus creating stacks involves first creating a sequence, and 
+% then using the sequence functions which work in a stack manner
+% (\cs{seq_push:Nn}, \emph{etc}.).
+% 
+%  Due to the nature of the underlying \TeX{} implementation, it is 
+%  possible to assign values to token list variables and comma lists 
+%  without first declaring them. However, this is \emph{not supported
+%  behaviour}. The \LaTeX3 coding convention is that all variables must
+%  be declared before use.
+%
+%  The \pkg{expl3} package can be loaded with the |check-declarations| option 
+%  to verify that all variables are declared before use. This has a 
+%  performance implication and is therefore intended for testing during 
+%  development and not for use in production documents. 
+% 
+% \subsubsection{Functions: argument specifications}
+% 
+% Function names end with an \meta{arg-spec} after a colon.  This
+% gives an indication of the types of argument that a function takes,
+% and provides a convenient method of naming similar functions that
+% differ only in their argument forms (see the next section for
+% examples).
+% 
+% The \meta{arg-spec} consists of a (possibly empty) list of letters,
+% each denoting one argument of the function. The letter, including
+% its case, conveys information about the type of argument required.
+%
+% All functions have a base form with arguments using one of the
+% following argument specifiers:
+% \begin{arg-description}
+%   \item[n]  Unexpanded token or braced token list.\\
+%     This is a standard \TeX{} undelimited macro argument.
+%   \item[N]  Single token (unlike~|n|, the argument must \emph{not} be
+%     surrounded by braces).\\
+%     A typical example of a command taking an~|N|
+%     argument is~|\cs_set|, in which the command being defined must be
+%     unbraced.
+%   \item[p]  Primitive \TeX{} parameter specification.\\
+%     This can be something simple like~|#1#2#3|, but may use arbitrary
+%     delimited argument syntax such as: |#1,#2\q_stop#3|. This is used
+%     when defining functions.
+%   \item[T,F]
+%     These are special cases of~|n| arguments, used for the
+%     true and false code in conditional commands.
+% \end{arg-description}
+% There are two other specifiers with more general meanings:
+% \begin{arg-description}
+%  \item[D] This means: \textbf{Do not use}. This special case is used
+%     for \TeX{} primitives and other commands that are provided for use
+%     only while bootstrapping the \LaTeX{} kernel.  Programmers outside
+%     the kernel team should not use these functions!
+%   \item[w] This means that the argument syntax is \enquote{weird} in that it
+%     does not follow any standard rule.  It is used for functions with
+%     arguments that take non standard forms: examples are \TeX-level
+%     delimited arguments and the boolean tests needed after certain
+%     primitive |\if|\ldots\ commands.
+% \end{arg-description}
+% 
+% In case of |n| arguments that consist of a single token the 
+% surrounding braces can be omitted in nearly all
+% situations---functions that force the use of braces even for single
+% token arguments are explicitly mentioned. However, programmers are
+% encouraged to always use braces around \texttt{n} arguments, as this 
+% makes the relationship between function and argument clearer.
+% 
+% Further argument specifiers are available as part of the expansion
+% control system.  These are discussed in the next section.
+% 
+% \section{Expansion control}
+%
+% Let's take a look at some typical operations one might want to
+% perform. Suppose we maintain a stack of open files and we use the
+% stack |\g_ior_file_name_seq| to keep track of them (|io| is the
+% file reading and writing module: it uses the module prefix
+% \texttt{ior} for reading functions and variables). The basic 
+% operation here is to push a name onto this stack which could be 
+% done by the operation 
+% \begin{quote}
+%   |\seq_gpush:Nn \g_ior_file_name_seq {#1}|
+% \end{quote}
+% where |#1| is the filename. In other words, this operation would
+% push the file name as is onto the stack.
+%
+% However, we might face a situation where the filename is stored in 
+% a variable of some sort, say |\l_ior_curr_file_tl|. In this case we 
+% want to retrieve the value of the variable. If we simply use
+% \begin{quote}
+%   |\seq_gpush:Nn \g_io_file_name_seq \l_io_curr_file_tl|
+% \end{quote}
+% we will not get the value of the variable pushed onto the stack,
+% only the variable name itself. Instead a suitable number of
+% |\exp_after:wN| would be necessary (together with extra braces) to
+% change the order of expansion,\footnote{\cs{exp_after:wN} is
+% the \LaTeX3 name for the \TeX{} \cs{expandafter} primitive.}, \emph{i.e.}
+% \begin{quote}
+%   |\exp_after:wN|                              \\
+%   |   \seq_gpush:Nn|                           \\
+%   |\exp_after:wN|                              \\
+%   |   \g_io_file_name_seq|                   \\
+%   |\exp_after:wN|                              \\
+%   |   { \l_io_curr_file_tl }|
+% \end{quote}
+%
+% The above example is probably the simplest case but already shows
+% how the code changes to something difficult to understand. 
+% Furthermore there is an assumption in this: that the storage bin 
+% reveals its contents after exactly one expansion. Relying on this 
+% means that you cannot do proper checking plus you have to know 
+% exactly how a storage bin acts in order to get the correct number
+% of expansions.  Therefore \LaTeX3 provides the programmer with a 
+% general scheme that keeps the code compact and easy to understand.
+%
+% To denote that some argument to a function needs special treatment one
+% just uses different letters in the arg-spec part of the function to
+% mark the desired behaviour. In the above example one would write
+% \begin{quote}
+%   |\seq_gpush:NV \g_io_file_name_seq \l_io_curr_file_tl|
+% \end{quote}
+% to achieve the desired effect. Here the |V| (the second argument) 
+% is for \enquote{retrieve the value of the variable} before passing it to
+% the base function.
+%
+% The following letters can be used to denote special treatment of
+% arguments before passing it to the base function:
+% \begin{description}
+%   \item[c] Character string used as a command name.\\ The argument (a
+%     token or braced token list) must, when fully expanded, produce a
+%     sequence of characters which is then used to construct a command
+%     name (\emph{via}~|\csname| \ldots |\endcsname|). This command name is the
+%     single token that is passed to the function as the argument. Hence
+%     \begin{quote}
+%       |\seq_gpush:cV { g_file_name_seq } \l_tmpa_tl|
+%     \end{quote}
+%     is equivalent to 
+%     \begin{quote}
+%       |\seq_gpush:NV \g_file_name_seq \l_tmpa_tl"|
+%     \end{quote}
+%     Remember that \texttt{c} arguments are \emph{fully expanded} by
+%     \TeX{} when creating csnames. This means that (a) the entire 
+%     argument must be expandable and (b) any variables will be 
+%     converted to their content. So the preceding examples are also
+%     equivalent to
+%     \begin{quote}
+%       |\tl_new:N \g_file_seq_name_tl| \\
+%       |\tl_gset:Nn \g_file_seq_name_tl { g_file_name_seq }| \\
+%       |\seq_gpush:cV { \g_file_seq_name_tl } \l_tmpa_tl|.
+%     \end{quote}
+%     (Token list variables are expandable and do not require an
+%     accessor function. Other variable types require the appropriate
+%     \cs{<var>_use:N} functions to be used in this context.)
+%   \item[V]  Value of a variable.\\
+%     This means that the contents of the register in question is used as the
+%     argument, be it an integer, a length-type register, a token list variable
+%     or similar. The value is passed to the function as a braced token list.
+%   \item[v] Value of a register, constructed from a character string
+%     used as a command name.\\
+%     This is a combination of |c| and |V| which first constructs a
+%     control sequence from the argument and then passes the value of the
+%     resulting register to the function.
+%   \item[x]  Fully-expanded token or braced token list.\\
+%     This means that the argument is expanded as in the replacement
+%     text of an~|\edef|, and the expansion is passed to the function as
+%     a braced token list. This means that expansion takes place until
+%     only unexpandable tokens are left.
+%  \item[o]  One-level-expanded token or braced token list.\\  
+%     This means that the argument is expanded one level, as by
+%     \cs{expandafter}, and the expansion is passed to the function as a
+%     braced token list.  Note that if the original argument is a braced
+%     token list then only the first token in that list is expanded.
+%     In general, using \texttt{V} should be preferred to using 
+%     \texttt{o} for simple variable retrieval.
+%   \item[f] Almost the same as the |x| type except here the token list
+%     is expanded fully until the first unexpandable token is found and
+%     the rest is left unchanged. Note that if this function finds a
+%     space at the beginning of the argument it will gobble it and not
+%     expand the next argument.
+% \end{description}
+%
+% \subsection{Simpler means better}
+% 
+% Anyone who programs in \TeX{} is frustratingly familiar with the
+% problem of arranging that arguments to functions are suitably expanded
+% before the function is called.  To illustrate how expansion control
+% can bring instant relief to this problem we shall consider two
+% examples copied from \texttt{latex.ltx}.
+% 
+% \begin{verbatim}
+%        \global\expandafter\let
+%              \csname\cf@encoding \string#1\expandafter\endcsname
+%              \csname ?\string#1\endcsname
+% \end{verbatim}
+% This first piece of code is in essence simply a global |\let| whose
+% two arguments firstly have to be constructed before |\let| is
+% executed. The |#1| is a control sequence name such as
+% |\textcurrency|. The token to be defined is obtained by
+% concatenating the characters of the current font encoding stored in
+% |\cf@encoding|, which has to be fully expanded, and the name of the
+% symbol. The second token is the same except it uses the default
+% encoding |?|. The result is a mess of interwoven |\expandafter|
+% and |\csname| beloved of all \TeX{} programmers, and the code is
+% essentially unreadable.
+% 
+% Using the conventions and functionality outlined here, the task would
+% be achieved with code such as this:
+% \begin{verbatim}
+%   \cs_gset_eq:cc 
+%     { \cf@encoding \token_to_str:N  #1 } { ? \token_to_str:N #1 }
+% \end{verbatim}
+% The command |\cs_gset_eq:cc| is a global~|\let| that generates
+% command names out of both of its arguments before making the
+% definition. This produces code that is far more readable and more
+% likely to be correct first time. (\cs{token_to_str:N} is the \LaTeX3
+% name for \cs{string}.)
+% 
+% Here is the second example.
+% \begin{verbatim}
+%   \expandafter
+%     \in@
+%   \csname sym#3%
+%     \expandafter
+%       \endcsname
+%     \expandafter
+%       {%
+%     \group@list}%
+% \end{verbatim}
+% This piece of code is part of the definition of another function. It
+% first produces two things: a token list, by expanding |\group@list| once;
+% and a token whose name comes from~`|sym#3|'.  Then the function~|\in@|
+% is called and this tests if its first argument occurs in the token list
+% of its second argument.
+% 
+% Again we can improve enormously on the code.  First we shall rename
+% the function~|\in@| according to our conventions.  A function such as
+% this but taking two normal \enquote{\texttt{n}} arguments might reasonably be
+% named |\seq_test_in:Nn|; thus the variant function we need will be
+% defined with the appropriate argument types and its name will be
+% |\seq_test_in:cV|.  Now this code fragment will be simply:
+% \begin{verbatim}
+%   \seq_test_in:cV { sym #3 } \l_group_seq
+% \end{verbatim}
+% Note that, in addition to the lack of |\expandafter|, the space after
+% the~|}| will be silently ignored since all white space is ignored in
+% this programming environment.
+% 
+% \subsection{New functions from old}
+% 
+% For many common functions the \LaTeX3 kernel will provide variants
+% with a range of argument forms, and similarly it is expected that
+% extension packages providing new functions will make them available in
+% the all the commonly needed forms.
+% 
+% However, there will be occasions where it is necessary to construct a
+% new such variant form; therefore the expansion module provides a
+% straightforward mechanism for the creation of functions with any
+% required argument type, starting from a function that takes \enquote{normal}
+% \TeX{} undelimited arguments.
+% 
+% To illustrate this let us suppose you have a \enquote{base function}
+% |\demo_cmd:Nnn| that takes three normal arguments, and that you need
+% to construct the variant |\demo_cmd:cnx|, for which the first argument
+% is used to construct the \emph{name} of a command, whilst the third
+% argument must be fully expanded before being passed to
+% |\demo_cmd:Nnn|.
+% To produce the variant form from the base form, simply use this:
+% \begin{verbatim}
+%   \cs_generate_variant:Nn \demo_cmd:Nnn { cnx } 
+% \end{verbatim}
+% This defines the variant form so that you can then write, for example:
+% \begin{verbatim}
+%   \demo_cmd:cnx { abc } { pq } { \rst \xyz }
+% \end{verbatim}
+% rather than \ldots\ well, something like this!
+% \begin{verbatim}
+%   \def \tempa {{pq}}%
+%   \edef \tempb {\rst \xyz}%
+%   \expandafter
+%     \demo@cmd:nnn 
+%   \csname abc%
+%     \expandafter
+%       \expandafter
+%     \expandafter
+%         \endcsname
+%     \expandafter
+%       \tempa
+%     \expandafter
+%       {%
+%     \tempb
+%       }%
+% \end{verbatim}
+% 
+% Another example: you may wish to declare a function
+% |\demo_cmd_b:xcxcx|, a variant of an existing function
+% |\demo_cmd_b:nnnnn|, that fully
+% expands arguments 1,~3 and~5, and produces commands to pass as
+% arguments 2 and~4 using~|\csname|.
+% The definition you need is simply
+% \begin{verbatim}
+%   \cs_generate_variant:Nn \demo_cmd_b:nnnnn { xcxcx }
+% \end{verbatim}
+% 
+% This extension mechanism is written so that if the same new form of 
+% some existing command is implemented by two extension packages then the
+% two definitions will be identical and thus no conflict will occur.
+% 
+% \section{The distribution}
+% 
+% At present, the \pkg{expl3} modules are designed to be loaded on top 
+% of \LaTeXe. In time, a \LaTeX3 format will be produced based on this
+% code. This allows the code to be used in \LaTeXe{} packages \emph{now}
+% while a stand-alone \LaTeX3 is developed.
+% 
+% \begin{bfseries}
+%   While \pkg{expl3} is still experimental, the bundle is now regarded 
+%   as broadly stable. The syntax conventions and functions provided 
+%   are now ready for wider use. There may still be changes to some 
+%   functions, but these will be minor when compared to the scope of 
+%   \pkg{expl3}.
+% \end{bfseries}
+%
+%  New modules will be added to the distributed version of \pkg{expl3}
+%  as they reach maturity. At present, the \pkg{expl3} bundle consists 
+%  of a number of modules, most of which are loaded by including the
+%  line:
+% \begin{verbatim}
+%   \RequirePackage{expl3}
+% \end{verbatim}  
+% in a \LaTeXe{} package, class or other file. The \pkg{expl3} modules
+% regarded as stable, and therefore suitable for basing real code on,
+% are as follows:
+% 
+% \begin{description}
+%   \providecommand\explpkg[2]{\item[#1]#2}
+%   \explpkg{l3basics}{
+%     This contains the basic definition modules used
+%     by the other packages.
+%   }
+%   \explpkg{l3box}{
+%     Primitives for dealing with boxes.
+%   }
+%   \explpkg{l3clist}{
+%     Methods for manipulating comma-separated token lists.
+%   }
+%   \explpkg{l3expan}{ 
+%     This is the argument expansion module discussed earlier in this
+%     document.
+%   }
+%   \explpkg{l3int}{
+%     This implements the integer data-type \texttt{int}.
+%   }
+%   \explpkg{l3io}{
+%     A module providing low level input and output functions.
+%   }
+%   \explpkg{l3keyval}{
+%     Low-level implementation for processing lists of the form
+%   \texttt{\{ key1=val1 , key2=val2 \}}.
+%   }
+%   \explpkg{l3keys}{
+%     Higher-level implementation of key--value methods: intended to work
+%     as a \LaTeX3 version of \pkg{xkeyval}/\pkg{kvoptions}, although
+%     with input syntax more like that of \pkg{pgfkeys}.
+%   }
+%   \explpkg{l3msg}{
+%     Communicating with the user: includes low-level hooks to allow
+%     messages to be filtered (higher-level interface for filtering
+%     to be written!).
+%   }
+%   \explpkg{l3names}{ 
+%     This sets up the basic naming scheme and renames all
+%     the \TeX{} primitives.  
+%   }
+%   \explpkg{l3prg}{
+%     Program control structures such as boolean data type |bool|, generic
+%     do-while loops, case-switches, sorting routines and stepwise loops. 
+%   }
+%   \explpkg{l3prop}{
+%     This implements the data-type for \enquote{property lists} that are used, in
+%     particular, for storing key/value pairs.
+%   }
+%   \explpkg{l3quark}{ 
+%     A \enquote{quark} is a command that is defined to expand to
+%     itself!  Therefore they must never be expanded as this will generate
+%     infinite recursion; they do however have many uses, \emph{e.g.}~as
+%     special markers and delimiters within code.
+%   }
+%   \explpkg{l3seq}{
+%     This implements data-types such as queues and stacks.
+%   }
+%   \explpkg{l3skip}{
+%     Implements the \enquote{rubber length} datatype \texttt{skip} and the
+%     \enquote{rigid length} datatype |dim|. 
+%   }
+%   \explpkg{l3tl}{
+%     This implements a basic data-type, called a \textit{token-list
+%     variable} (|tl var.|), used for storing named token lists: these are
+%     \TeX{} macros with no arguments.
+%   }
+%   \explpkg{l3token}{
+%     Analysing token lists and token streams, including peeking ahead to
+%     see what's coming next and inspecting tokens to detect which kind
+%     they are.
+%   }
+% \end{description}
+%
+% \section{Moving from \LaTeXe{} to \LaTeX3}
+%
+% To help programmers to use \LaTeX3 code in existing \LaTeXe{} package,
+% some short notes on making the change are probably desirable. 
+% Suggestions for inclusion here are welcome! Some of the following
+% is concerned with code, and some with coding style.
+% 
+% \begin{itemize}
+%   \item \pkg{expl3} is mainly focussed on programming. This means that
+%     some areas still require the use of \LaTeXe{} internal macros.
+%     For example, you may well need \cs{@ifpackageloaded}, as there
+%     is currently no native \LaTeX3 package loading module.
+%   \item User level macros should be generated using the mechanism 
+%     available in the \pkg{xparse} package, which is available from CTAN
+%     or the \LaTeX3 SVN repository.
+%   \item At an internal level, most functions should be generated 
+%     \cs{long} (using \cs{cs_new:Npn}) rather than \enquote{short} (using
+%     \cs{cs_new_nopar:Npn}). The exceptions are:
+%     \begin{itemize}
+%       \item Functions which take no arguments;
+%       \item Functions which are used with pre-set arguments which 
+%         therefore cannot be \cs{long}: this is mainly the case with
+%         auxiliary functions.
+%     \end{itemize}
+%   \item Where possible, declare all variables and functions (using 
+%     \cs{cs_new:Npn}, \cs{tl_new:N}, etc.) before use.
+%   \item Prefer \enquote{higher-level} functions over \enquote{lower-level},
+%   where
+%     possible. So for example use \cs{cs_if_exist:N(TF)} in preference 
+%     \cs{if_cs_exist:N}.
+%   \item Use space to make code readable. In general, we recommend
+%     a layout such as:
+%     \begin{verbatim}
+%       \cs_new:Npn \foo_bar:Nn #1#2
+%         {
+%           \cs_if_exist:NTF #1
+%             { \foo_bar_aux_i:n {#2} }
+%             { \foo_bar_aux_ii:nn {#2} { literal } }
+%         }
+%     \end{verbatim}
+%     where spaces are used around |{| and |}| except for isolated
+%     |#1|, |#2|, etc.
+%   \item Put different code items on separate lines: readability is
+%     much more useful than compactness.
+%   \item Use long, descriptive names for functions and variables, 
+%     and for auxiliary functions use the parent function name plus
+%     |aux|, |aux_i|, |aux_ii| and so on.
+%   \item If in doubt, ask the team via the LaTeX-L list: someone will
+%     soon get back to you!  
+% \end{itemize} 
+% 
+% \section{The \LaTeX3 Project}
+% 
+% Development of \LaTeX3 is carried out by The \LaTeX3 Project. Over time,
+% the membership of this team has naturally varied. Currently, the members
+% are
+% \begin{itemize}
+%   \item Johannes Braams
+%   \item David Carlisle
+%   \item Robin Fairbairns
+%   \item Morten H{\o}gholm
+%   \item Bruno Le Floch
+%   \item Thomas Lotze
+%   \item Frank Mittelbach
+%   \item Will Robertson
+%   \item Chris Rowley
+%   \item Rainer Sch{\"o}pf
+%   \item Joseph Wright
+% \end{itemize}
+% while former members are
+% \begin{itemize}
+%   \item Michael Downes
+%   \item Denys Duchier
+%   \item Alan Jeffrey
+%   \item Martin Schr{\"o}der
+% \end{itemize}
+% 
+% \begin{thebibliography}{1}
+% 
+%   \bibitem{A-W:K-TB}
+%     Donald E Knuth
+%     \newblock \emph{The \TeX{}book}.
+%     \newblock Addison-Wesley, Reading, Massachusetts, 1984.
+% 
+%   \bibitem{A-W:GMS94}
+%     Goossens, Mittelbach and Samarin.
+%     \newblock \emph{ The \LaTeX{} Companion}.
+%     \newblock Addison-Wesley, Reading, Massachusetts, 1994.
+% 
+%   \bibitem{A-W:LLa94}
+%     Leslie Lamport.
+%     \newblock \emph{\LaTeX{}: A Document Preparation System}.
+%     \newblock Addison-Wesley, Reading, Massachusetts, second edition, 1994.
+% 
+%   \bibitem{tub:MR97-1}
+%     Frank Mittelbach and Chris Rowley.
+%     \newblock \enquote{The \LaTeX3 Project}.
+%     \newblock \emph{TUGboat}, 
+%     Vol.\,18, No.\,3, pp.\,195--198, 1997.
+% 
+% \end{thebibliography}
+% 
+% \end{documentation}
+%
+% \begin{implementation}
+%
+% \section{\pkg{expl3} implementation}
+%
+%    \begin{macrocode}
+%<*package>
+%    \end{macrocode}
+%    
+% Load \pkg{etex} as otherwise we are likely to get into trouble
+% with registers. Some inserts are reserved also as these have to
+% be from the standard pool.
+%    \begin{macrocode}
+\RequirePackage{etex}
+\reserveinserts{32}
+%    \end{macrocode}
+%    
+% We want \pkg{calc} to allow \LaTeXe{} to do \eTeX{}-like setting (which will
+% be native in \LaTeX3).
+%    \begin{macrocode}
+\RequirePackage{calc}
+%    \end{macrocode}
+%
+% Do the package identification: this will turn on code syntax.
+%    \begin{macrocode}
+\ProvidesExplPackage
+  {\ExplFileName}{\ExplFileDate}{\ExplFileVersion}{\ExplFileDescription}
+%    \end{macrocode}
+%    
+% \begin{macro}{\@l@expl@check@declarations@bool}
+% \begin{macro}{\@l@expl@log@functions@bool}
+%   A couple of package options to be provided.
+%   These have to be done by hand as there is no \pkg{expl3} yet: the logging
+%   option
+%   is needed before loading \pkg{l3basics}!
+%    \begin{macrocode}
+\newcommand* \@l@expl@check@declarations@bool { 0 }
+\newcommand*\@l@expl@log@functions@bool { 0 }
+\DeclareOption { check-declarations }
+  { \def \@l@expl@check@declarations@bool { 1 } }
+\DeclareOption { log-functions }
+   {\def \@l@expl@log@functions@bool { 1 } }
+\ProcessOptions \relax
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}   
+%    
+% We already loaded \pkg{l3names} at the beginning of the \texttt{dtx} file.
+% We now load the base of \LaTeX3, stopping once token list variables are
+% defined.
+%    \begin{macrocode}
+\RequirePackage{ l3basics, l3expan, l3tl }
+%    \end{macrocode}
+%
+% \begin{macro}{\tl_check_exists:N}
+%   When used as a package, there is an option to be picky and to
+%   check definitions exist. This part of the process is done now, so that
+%   variable types based on |tl| (for example |clist|, |seq| and |prop|) will
+%   inherit
+%   the appropriate definitions.
+%    \begin{macrocode}
+\ifodd \@l@expl@check@declarations@bool \relax
+  \cs_new_protected:Npn \tl_check_exists:N #1 
+    { 
+      \cs_if_exist:NF #1 
+        { 
+          \msg_kernel_error:nnx { check } { non-declared-variable }
+            { \token_to_str:N #1 }
+        } 
+    }
+  \cs_set_protected:Npn \tl_set:Nn #1#2 
+    {
+      \tl_check_exists:N #1
+      \cs_set_nopar:Npx #1 { \exp_not:n {#2} }
+    }
+  \cs_set_protected:Npn \tl_set:Nx #1#2 
+    {
+      \tl_check_exists:N #1
+      \cs_set_nopar:Npx #1 {#2}
+    }
+  \cs_set_protected:Npn \tl_gset:Nn #1#2 
+    {
+      \tl_check_exists:N #1
+      \cs_gset_nopar:Npx #1 { \exp_not:n {#2} }
+    }
+  \cs_set_protected:Npn \tl_gset:Nx #1#2 
+    {
+      \tl_check_exists:N #1
+      \cs_gset_nopar:Npx #1 {#2}
+    }
+  \cs_set_protected:Npn \tl_set_eq:NN #1#2
+    {
+      \tl_check_exists:N #1
+      \tl_check_exists:N #2
+      \cs_set_eq:NN #1 #2
+    }
+  \cs_set_protected:Npn \tl_gset_eq:NN #1#2
+    {
+      \tl_check_exists:N #1
+      \tl_check_exists:N #2
+      \cs_gset_eq:NN #1 #2
+    }
+\fi
+%    \end{macrocode}
+%\end{macro}
+%    
+% The rest of the code can now be loaded.
+%    \begin{macrocode}
+\RequirePackage{
+  l3seq,
+  l3int,
+  l3quark,
+  l3prg,
+  l3clist,
+  l3token,
+  l3prop,
+  l3msg,
+  l3io,
+  l3file, 
+  l3skip,
+  l3box,
+  l3keyval,
+  l3keys,
+  l3fp,
+  l3luatex
+}
+%    \end{macrocode}
+% 
+% Depreciated, but loaded for the moment.
+%    \begin{macrocode}
+\RequirePackage{l3toks}
+%    \end{macrocode}
+% 
+% The error message for variables cannot be created before now: this means that
+% \pkg{expl3} itself may cause some odd errors.
+%    \begin{macrocode}
+\bool_if:nT { \@l@expl@check@declarations@bool }
+  {
+    \msg_kernel_new:nnnn { check } { non-declared-variable }
+      { The~variable~#1~has~not~been~declared~\msg_line_context:. }
+      { 
+        Checking~is~active,~and~you~have~tried~do~so~something~like: \\
+        \ \ \tl_set:Nn #1 ~
+          \iow_char:N \{ ~ ... ~ \iow_char:N \} \\
+        without~first~having: \\
+        \ \ \tl_new:N #1  \\ 
+        \\ 
+        LaTeX~will~create~the~variable~and~continue.
+      }  
+  }
+%    \end{macrocode}
+%
+% \begin{macro}{\ExplSyntaxOff, \ExplSyntaxOn}
+%   These are redefined here to reduce unnecessary work when switching.
+%   The definition for \cs{ExplSyntaxOff} copies the code from earlier,
+%   so that category codes are restored.
+%    \begin{macrocode}    
+\cs_set_protected_nopar:Npn \ExplSyntaxOn
+  {
+    \bool_if:NF \l_expl_status_bool
+      {
+        \cs_set_protected_nopar:Npx \ExplSyntaxOff
+          {
+            \char_set_catcode:nn { 9 }   { \char_value_catcode:n { 9 } }
+            \char_set_catcode:nn { 32 }  { \char_value_catcode:n { 32 } }
+            \char_set_catcode:nn { 34 }  { \char_value_catcode:n { 34 } }
+            \char_set_catcode:nn { 38 }  { \char_value_catcode:n { 38 } }
+            \char_set_catcode:nn { 58 }  { \char_value_catcode:n { 58 } }
+            \char_set_catcode:nn { 94 }  { \char_value_catcode:n { 94 } }
+            \char_set_catcode:nn { 95 }  { \char_value_catcode:n { 95 } }
+            \char_set_catcode:nn { 124 } { \char_value_catcode:n { 124 } }
+            \char_set_catcode:nn { 126 } { \char_value_catcode:n { 126 } }
+            \tex_endlinechar:D = \tex_the:D \tex_endlinechar:D \scan_stop:
+            \bool_set_false:N \l_expl_status_bool
+            \cs_set_protected_nopar:Npn \ExplSyntaxOff { }
+          }
+      }
+    \char_set_catcode_ignore:n           { 9 }   % tab
+    \char_set_catcode_ignore:n           { 32 }  % space
+    \char_set_catcode_other:n            { 34 }  % double quote
+    \char_set_catcode_alignment:n        { 38 }  % ampersand
+    \char_set_catcode_letter:n           { 58 }  % colon
+    \char_set_catcode_math_superscript:n { 94 }  % circumflex
+    \char_set_catcode_letter:n           { 95 }  % underscore
+    \char_set_catcode_other:n            { 124 } % pipe
+    \char_set_catcode_space:n            { 126 } % tilde  
+    \tex_endlinechar:D = 32 \scan_stop:
+    \bool_set_true:N \l_expl_status_bool
+  }
+\cs_set_protected_nopar:Npx \ExplSyntaxOff
+  {
+    \exp_not:o \ExplSyntaxOff
+    \cs_set_protected_nopar:Npn \ExplSyntaxOff { }
+  }
+%    \end{macrocode}
+% \end{macro}
+%    
+%    \begin{macrocode}
+%</package>
+%    \end{macrocode}
+%
+% \end{implementation}
+% 
+% \PrintIndex
diff --git a/Master/texmf-dist/source/latex/l3kernel/l3alloc.dtx b/Master/texmf-dist/source/latex/l3kernel/l3alloc.dtx
new file mode 100644
index 00000000000..8e175a2e74d
--- /dev/null
+++ b/Master/texmf-dist/source/latex/l3kernel/l3alloc.dtx
@@ -0,0 +1,216 @@
+% \iffalse meta-comment
+%
+%% File: l3alloc.dtx Copyright (C) 1990-2011 The LaTeX3 Project
+%%
+%% It may be distributed and/or modified under the conditions of the
+%% LaTeX Project Public License (LPPL), either version 1.3c of this
+%% license or (at your option) any later version.  The latest version
+%% of this license is in the file
+%%
+%%    http://www.latex-project.org/lppl.txt
+%%
+%% This file is part of the "expl3 bundle" (The Work in LPPL)
+%% and all files in that bundle must be distributed together.
+%%
+%% The released version of this bundle is available from CTAN.
+%%
+%% -----------------------------------------------------------------------
+%%
+%% The development version of the bundle can be found at
+%%
+%%    http://www.latex-project.org/svnroot/experimental/trunk/
+%%
+%% for those people who are interested.
+%%
+%%%%%%%%%%%
+%% NOTE: %%
+%%%%%%%%%%%
+%%
+%%   Snapshots taken from the repository represent work in progress and may
+%%   not work or may contain conflicting material!  We therefore ask
+%%   people _not_ to put them into distributions, archives, etc. without
+%%   prior consultation with the LaTeX3 Project.
+%%
+%% -----------------------------------------------------------------------
+%
+%<*driver|package>
+\RequirePackage{l3names}
+\GetIdInfo$Id: l3alloc.dtx 2478 2011-06-19 21:34:23Z joseph $
+  {L3 Experimental register allocation}
+%</driver|package>
+%<*driver>
+\documentclass[full]{l3doc}
+\begin{document}
+  \DocInput{\jobname.dtx}
+\end{document}
+%</driver>
+% \fi
+%
+% \title{^^A
+%   The \pkg{l3alloc} package\\ Register allocation^^A
+%   \thanks{This file describes v\ExplFileVersion,
+%      last revised \ExplFileDate.}^^A
+% }
+%
+% \author{^^A
+%  The \LaTeX3 Project\thanks
+%    {^^A
+%      E-mail:
+%        \href{mailto:latex-team@latex-project.org}
+%          {latex-team@latex-project.org}^^A
+%    }^^A
+% }
+%
+% \date{Released \ExplFileDate}
+%
+% \maketitle
+%
+% \begin{documentation}
+%
+%  Note that this module is only used for generating an pkg{expl3}-based
+%  format. Under \LaTeX{}, the \pkg{etex} package is used for allocation
+%  management.
+%
+%  This module provides the basic mechanism for allocating \TeX{}s
+%  registers. While designing this we have to take into account the
+%  following characteristics:
+%  \begin{itemize}
+%    \item |\box255| is reserved for use in the output routine, so it
+%      should not be allocated otherwise.
+%    \item \TeX{} can load up 256 hyphenation patterns (registers
+%      |\tex_language:D| 0--255),
+%    \item \TeX{} can load no more than 16 math families,
+%    \item \TeX{} supports no more than 16 I/O streams for reading
+%      (|\tex_read:D|) and 16 I/O streams for writing (|\tex_write:D|),
+%    \item \TeX{} supports no more than 256 inserts.
+%    \item The other registers (|\tex_count:D|, |\tex_dimen:D|,
+%      |\tex_skip:D|, |\tex_muskip:D|, |\tex_box:D|, and |\tex_toks:D|
+%      range from 0 to 32768, but registers numbered above 255 are
+%      accessed somewhat less efficiently.
+%    \item Registers could be allocated both globally and locally; the
+%      use of registers could also be globally or locally. Here we
+%      provide support for globally allocated registers for both
+%      global and local use.
+%    \end{itemize}
+%  We also need to allow for some bookkeeping: we need to know which
+%  register was allocated last and which registers can not be
+%  allocated by the standard mechanisms.
+%
+% \begin{function}{\alloc_new:nnnN}
+%   \begin{syntax}
+%     \cs{alloc_new:nnnN} \Arg{type} \Arg{min} \Arg{max} \meta{function}
+%   \end{syntax}
+%   Shorthand for allocating new registers. Defines \cs{<type>_new:N} as
+%   and allocator function of the specified \meta{type}, indexed up from
+%   \meta{min} to a \meta{max}, and with assignment carried out by
+%   the \meta{function}. This process will create two token lists,
+%   \cs{g_\meta{type}_allocation_tl} and \cs{c_\meta{type}_allocation_max_tl},
+%   to store the current and maximum allocation numbers, respectively.
+%   It will also create \cs{g_\meta{type}_allocation_seq} to store allocation
+%   numbers which should not be used.
+% \end{function}
+%
+% \section{Internal functions}
+%
+%  \begin{function}{\alloc_setup_type:nnn}
+%    \begin{syntax}
+%      \cs{alloc_setup_type:nnn} \Arg{type} \Arg{min} \Arg{max}
+%    \end{syntax}
+%    Sets up the storage needed for the administration of registers of
+%    \Arg{type}, which will start allocating at \meta{min} and will issue
+%    an error if there is an attempt to allocate past the \meta{max}.
+%  \end{function}
+%
+%  \begin{function}{\alloc_reg:nNN}
+%    \begin{syntax}
+%      \cs{alloc_reg:nNN} \Arg{type} \meta{function} \meta{register}
+%    \end{syntax}
+%    Preforms the allocation for a new \meta{register} of \meta{type},
+%    using the allocator \meta{function} (which will be either a 
+%    primitive \cs{tex_\ldots def:D} or \cs{tex_chardef:D}).
+%  \end{function}
+%
+% \end{documentation}
+%
+% \begin{implementation}
+%
+% \section{\pkg{l3alloc} implementation}
+%
+%    \begin{macrocode}
+%<*initex>
+%    \end{macrocode}
+%    
+% \begin{macro}{\alloc_new:nnnN}
+%   Shorthand for defining new register types and their allocators:
+%    \begin{macrocode}
+\cs_new:Npn \alloc_new:nnnN #1#2#3#4
+  {
+    \alloc_setup_type:nnn {#1} {#2} {#3}
+    \cs_new_nopar:cpn { #1 _new:N } ##1 { \alloc_reg:nNN {#1} #4 ##1 }
+  }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}[int]{\alloc_setup_type:nnn}
+%   For each type of register we need to \enquote{counters} that hold the
+%   last allocated global register, plus a constant for the maximum
+%   allocation. We also need a sequence to store the \enquote{exceptions}.
+%    \begin{macrocode}
+\cs_new_nopar:Npn \alloc_setup_type:nnn #1#2#3
+  {
+    \tl_new:c    { g_ #1 _allocation_tl }
+    \tl_gset:cx  { g_ #1 _allocation_tl }     { \int_eval:n {#2} }
+    \tl_const:cx { c_ #1 _allocation_max_tl } { \int_eval:n {#3} }
+    \seq_new:c   { g_ #1 _allocation_seq }
+  }
+%    \end{macrocode}
+% \end{macro}
+%
+%  \begin{macro}[int]{\alloc_reg:nNN}
+%  This internal macro performs the actual allocation.
+%    \begin{macrocode}
+\cs_new_nopar:Npn \alloc_reg:nNN #1#2#3
+  {
+    \chk_if_free_cs:N #3
+    \pref_global:D #2 #3 \tl_use:c { g_ #1 _allocation_tl } \scan_stop:
+    \iow_log:x
+      { 
+        \token_to_str:N #3 ~=~ #1 ~register~
+        \tl_use:c { g_ #1 _allocation_tl }
+      }
+    \alloc_next:n {#1}
+ }
+%    \end{macrocode}
+%  \end{macro}
+%
+% \begin{macro}[aux]{\alloc_next:n}
+%   Finding the next register to use is a question of doing an increment
+%   then a check: if we run out of registers then it is a fatal error, so
+%   there is no need to unwind the change. (The check could be done
+%   first but that needs an additional calculation.) There is a built-in
+%   loop to handle reserved allocation positions.
+%    \begin{macrocode}
+\cs_new_nopar:Npn \alloc_next:n #1
+  {
+    \tl_gset:cx { g_ #1 _allocation_tl }
+      { \int_eval:n { \tl_use:c { g_ #1 _allocation_tl } + 1 } }
+    \int_compare:nNnTF
+      { \tl_use:c { g_ #1 _allocation_tl } } 
+      > { \tl_use:c { c_ #1 _allocation_max_tl } }
+      { \msg_kernel_fatal:nnx { alloc } { out-of-registers } {#1} }
+      {
+        \seq_if_in:cxT { g_ #1 _allocation_seq }
+          { \tl_use:c { g_ #1 _allocation_tl } }
+          { \alloc_next:n {#1} }
+      }
+  }
+%    \end{macrocode}
+%  \end{macro}
+%
+%    \begin{macrocode}
+%</initex>
+%    \end{macrocode}
+%
+% \end{implementation}
+% 
+% \PrintIndex
diff --git a/Master/texmf-dist/source/latex/l3kernel/l3basics.dtx b/Master/texmf-dist/source/latex/l3kernel/l3basics.dtx
new file mode 100644
index 00000000000..52cc56c699a
--- /dev/null
+++ b/Master/texmf-dist/source/latex/l3kernel/l3basics.dtx
@@ -0,0 +1,3208 @@
+% \iffalse meta-comment
+%
+%% File: l3basics.dtx Copyright (C) 1990-2011 The LaTeX3 project
+%%
+%% It may be distributed and/or modified under the conditions of the
+%% LaTeX Project Public License (LPPL), either version 1.3c of this
+%% license or (at your option) any later version.  The latest version
+%% of this license is in the file
+%%
+%%    http://www.latex-project.org/lppl.txt
+%%
+%% This file is part of the "expl3 bundle" (The Work in LPPL)
+%% and all files in that bundle must be distributed together.
+%%
+%% The released version of this bundle is available from CTAN.
+%%
+%% -----------------------------------------------------------------------
+%%
+%% The development version of the bundle can be found at
+%%
+%%    http://www.latex-project.org/svnroot/experimental/trunk/
+%%
+%% for those people who are interested.
+%%
+%%%%%%%%%%%
+%% NOTE: %%
+%%%%%%%%%%%
+%%
+%%   Snapshots taken from the repository represent work in progress and may
+%%   not work or may contain conflicting material!  We therefore ask
+%%   people _not_ to put them into distributions, archives, etc. without
+%%   prior consultation with the LaTeX3 Project.
+%%
+%% -----------------------------------------------------------------------
+%
+%<*driver|package>
+\RequirePackage{l3names}
+\GetIdInfo$Id: l3basics.dtx 2493 2011-07-04 09:50:22Z bruno $
+  {L3 Experimental basic definitions}
+%</driver|package>
+%<*driver>
+\documentclass[full]{l3doc}
+\begin{document}
+  \DocInput{\jobname.dtx}
+\end{document}
+%</driver>
+% \fi
+%
+% \title{^^A
+%   The \pkg{l3basics} package\\ Basic definitions^^A
+%   \thanks{This file describes v\ExplFileVersion,
+%      last revised \ExplFileDate.}^^A
+% }
+%
+% \author{^^A
+%  The \LaTeX3 Project\thanks
+%    {^^A
+%      E-mail:
+%        \href{mailto:latex-team@latex-project.org}
+%          {latex-team@latex-project.org}^^A
+%    }^^A
+% }
+%
+% \date{Released \ExplFileDate}
+%
+% \maketitle
+%
+% \begin{documentation}
+%
+% As the name suggest this package holds some basic definitions which
+% are needed by most or all other packages in this set.
+%
+% Here we describe those functions that are used all over the place. With
+% that we mean functions dealing with the construction and testing of
+% control sequences. Furthermore the basic parts of conditional
+% processing are covered; conditional processing dealing with specific
+% data types is described in the modules specific for the respective
+% data types.
+% 
+% \section{No operation functions}
+%
+% \begin{function}[EXP]{\prg_do_nothing:}
+%   \begin{syntax}
+%     \cs{prg_do_nothing:}
+%   \end{syntax}
+%   An expandable function which does nothing at all: leaves nothing
+%   in the input stream after a single expansion.
+% \end{function}
+%
+% \begin{function}{\scan_stop:}
+%   \begin{syntax}
+%     \cs{scan_stop:}
+%   \end{syntax}
+%   A non-expandable function which does nothing. Does not vanish on
+%   expansion but produces no typeset output.
+% \end{function}
+%
+% \section{Grouping material}
+%
+% \begin{function}{\group_begin:, \group_end:}
+%   \begin{syntax}
+%     \cs{group_begin:}
+%     \cs{group_end:}
+%   \end{syntax}
+%   These functions begin and end a group for definition purposes.
+%   Assignments are local to groups unless carried out in a global
+%   manner. (A small number of exceptions to this rule will be noted
+%   as necessary elsewhere in this document.) Each \cs{group_begin:}
+%   must be matched by a \cs{group_end:}, although this does not have
+%   to occur within the same function. Indeed, it is often necessary
+%   to start a group within one function and finish it within another,
+%   for example when seeking to use non-standard category codes.
+% \end{function}
+%
+% \begin{function}{\group_insert_after:N}
+%   \begin{syntax}
+%     \cs{group_insert_after:N} \meta{token}
+%   \end{syntax}
+%   Adds \meta{token} to the list of \meta{tokens} to be inserted
+%   when the current group level ends. The list of \meta{tokens} to be
+%   inserted will be empty at the beginning of a group: multiple
+%   applications of \cs{group_insert_after:N} may be used to build
+%   the inserted list one \meta{token} at a time. The current group
+%   level may be closed by a \cs{group_end:} function or by a token
+%   with category code $2$ (close-group). The later will be a ^^A{
+%   |}| if standard category codes apply.
+% \end{function}
+% 
+% \section{Control sequences and functions}
+%
+% As \TeX{} is a macro language, creating new functions means
+% creating macros. At point of use, a function is replaced by
+% the replacement text (\enquote{code}) in which each parameter
+% in the code (|#1|, |#2|, \emph{etc.}) is replaced the appropriate
+% arguments absorbed by the function. In the following, \meta{code}
+% is therefore used as a shorthand for \enquote{replacement text}.
+% 
+% Functions which are not \enquote{protected} will be fully expanded
+% inside an \texttt{x} expansion. In contrast, \enquote{protected}
+% functions are not expanded within \texttt{x} expansions.
+% 
+% \subsection{Defining functions}
+%
+% Functions can be created with no requirement that they are declared
+% first (in contrast to variables, which must always be declared). 
+% Declaring a function before setting up the code means that the name
+% chosen will be checked and an error raised if it is already in use.
+% The name of a function can be checked at the point of definition using
+% the \cs{cs_new\ldots} functions: this is recommended for all
+% functions which are defined for the first time.
+% 
+% \subsection{Defining new functions using primitive parameter text}
+%
+% \begin{function}{\cs_new:Npn, \cs_new:cpn, \cs_new:Npx, \cs_new:cpx}
+%   \begin{syntax}
+%     \cs{cs_new:Npn} \meta{function} \meta{parameters} \Arg{code}
+%   \end{syntax}
+%   Creates \meta{function} to expand to \meta{code} as replacement text.
+%   Within the \meta{code}, the \meta{parameters} (|#1|, |#2|,
+%   \emph{etc.}) will be replaced by those absorbed by the function.
+%   The definition is global and an error will result if the
+%   \meta{function} is already defined.
+% \end{function}
+%
+% \begin{function}
+%   {
+%     \cs_new_nopar:Npn, \cs_new_nopar:cpn,
+%     \cs_new_nopar:Npx, \cs_new_nopar:cpx
+%   }
+%   \begin{syntax}
+%     \cs{cs_new_nopar:Npn} \meta{function} \meta{parameters} \Arg{code}
+%   \end{syntax}
+%   Creates \meta{function} to expand to \meta{code} as replacement text.
+%   Within the \meta{code}, the \meta{parameters} (|#1|, |#2|,
+%   \emph{etc.}) will be replaced by those absorbed by the function.
+%   When the \meta{function} is used the \meta{parameters} absorbed
+%   cannot contain \cs{par} tokens. The definition is global and
+%   an error will result if the \meta{function} is already defined.
+% \end{function}
+%
+% \begin{function}
+%   {
+%     \cs_new_protected:Npn, \cs_new_protected:cpn,
+%     \cs_new_protected:Npx, \cs_new_protected:cpx
+%   }
+%   \begin{syntax}
+%     \cs{cs_new_protected:Npn} \meta{function} \meta{parameters}
+%     ~~\Arg{code}
+%   \end{syntax}
+%   Creates \meta{function} to expand to \meta{code} as replacement text.
+%   Within the \meta{code}, the \meta{parameters} (|#1|, |#2|,
+%   \emph{etc.}) will be replaced by those absorbed by the function.
+%   The \meta{function} will not expand within an \texttt{x}-type
+%   argument. The definition is global and an error will result if the
+%   \meta{function} is already defined.
+% \end{function}
+%
+% \begin{function}
+%   {
+%     \cs_new_protected_nopar:Npn, \cs_new_protected_nopar:cpn ,
+%     \cs_new_protected_nopar:Npx, \cs_new_protected_nopar:cpx
+%   }
+%   \begin{syntax}
+%     \cs{cs_new_protected_nopar:Npn} \meta{function} \meta{parameters}
+%     ~~\Arg{code}
+%   \end{syntax}
+%   Creates \meta{function} to expand to \meta{code} as replacement text.
+%   Within the \meta{code}, the \meta{parameters} (|#1|, |#2|,
+%   \emph{etc.}) will be replaced by those absorbed by the function.
+%   When the \meta{function} is used the \meta{parameters} absorbed
+%   cannot contain \cs{par} tokens. The \meta{function} will not
+%   expand within an \texttt{x}-type argument. The definition is global
+%   and an error will result if the \meta{function} is already defined.
+% \end{function}
+% 
+% \begin{function}{\cs_set:Npn, \cs_set:cpn, \cs_set:Npx, \cs_set:cpx}
+%   \begin{syntax}
+%     \cs{cs_set:Npn} \meta{function} \meta{parameters} \Arg{code}
+%   \end{syntax}
+%   Sets \meta{function} to expand to \meta{code} as replacement text. 
+%   Within the \meta{code}, the \meta{parameters} (|#1|, |#2|,
+%   \emph{etc.}) will be replaced by those absorbed by the function.
+%   The assignment of a meaning to \meta{function} is restricted to
+%   the current \TeX{} group level.
+% \end{function}
+%
+% \begin{function}
+%   {
+%     \cs_set_nopar:Npn, \cs_set_nopar:cpn, 
+%     \cs_set_nopar:Npx, \cs_set_nopar:cpx
+%   }
+%   \begin{syntax}
+%     \cs{cs_set_nopar:Npn} \meta{function} \meta{parameters} \Arg{code}
+%   \end{syntax}
+%   Sets \meta{function} to expand to \meta{code} as replacement text. 
+%   Within the \meta{code}, the \meta{parameters} (|#1|, |#2|,
+%   \emph{etc.}) will be replaced by those absorbed by the function.
+%   When the \meta{function} is used the \meta{parameters} absorbed
+%   cannot contain \cs{par} tokens. The assignment of a meaning
+%   to \meta{function} is restricted to the current \TeX{} group
+%   level.
+% \end{function}
+%
+% \begin{function}
+%   {
+%     \cs_set_protected:Npn, \cs_set_protected:cpn,
+%     \cs_set_protected:Npx, \cs_set_protected:cpx
+%   }
+%   \begin{syntax}
+%     \cs{cs_set_protected:Npn} \meta{function} \meta{parameters}
+%     ~~\Arg{code}
+%   \end{syntax}
+%   Sets \meta{function} to expand to \meta{code} as replacement text. 
+%   Within the \meta{code}, the \meta{parameters} (|#1|, |#2|,
+%   \emph{etc.}) will be replaced by those absorbed by the function.
+%   The assignment of a meaning to \meta{function} is restricted to
+%   the current \TeX{} group level. The \meta{function} will
+%   not expand within an \texttt{x}-type argument.
+% \end{function}
+%
+% \begin{function}
+%   {
+%     \cs_set_protected_nopar:Npn, \cs_set_protected_nopar:cpn ,
+%     \cs_set_protected_nopar:Npx, \cs_set_protected_nopar:cpx ,
+%   }
+%   \begin{syntax}
+%     \cs{cs_set_protected_nopar:Npn} \meta{function}
+%     ~~\meta{parameters} \Arg{code}
+%   \end{syntax}
+%   Sets \meta{function} to expand to \meta{code} as replacement text. 
+%   Within the \meta{code}, the \meta{parameters} (|#1|, |#2|, 
+%   \emph{etc.}) will be replaced by those absorbed by the function.
+%   When the \meta{function} is used the \meta{parameters} absorbed
+%   cannot contain \cs{par} tokens. The assignment of a meaning
+%   to \meta{function} is restricted to the current \TeX{} group
+%   level. The \meta{function} will not expand within an
+%   \texttt{x}-type argument.
+% \end{function}
+%
+% \begin{function}{\cs_gset:Npn, \cs_gset:cpn, \cs_gset:Npx, \cs_gset:cpx}
+%   \begin{syntax}
+%     \cs{cs_gset:Npn} \meta{function} \meta{parameters} \Arg{code}
+%   \end{syntax}
+%   Globally sets \meta{function} to expand to \meta{code} as replacement
+%   text. Within the \meta{code}, the \meta{parameters} (|#1|, |#2|, 
+%   \emph{etc.}) will be replaced by those absorbed by the function.
+%   The assignment of a meaning to \meta{function} is \emph{not}
+%   restricted to the current \TeX{} group level: the assignment is
+%   global.
+% \end{function}
+%
+% \begin{function}
+%   {
+%     \cs_gset_nopar:Npn, \cs_gset_nopar:cpn, 
+%     \cs_gset_nopar:Npx , \cs_gset_nopar:cpx
+%   }
+%   \begin{syntax}
+%     \cs{cs_gset_nopar:Npn} \meta{function} \meta{parameters} \Arg{code}
+%   \end{syntax}
+%   Globally sets \meta{function} to expand to \meta{code} as replacement
+%   text. Within the \meta{code}, the \meta{parameters} (|#1|, |#2|, 
+%   \emph{etc.}) will be replaced by those absorbed by the function.
+%   When the \meta{function} is used the \meta{parameters} absorbed
+%   cannot contain \cs{par} tokens. The assignment of a meaning to
+%   \meta{function} is \emph{not} restricted to the current \TeX{}
+%   group level: the assignment is global.
+% \end{function}
+%
+% \begin{function}
+%   {
+%     \cs_gset_protected:Npn, \cs_gset_protected:cpn,
+%     \cs_gset_protected:Npx, \cs_gset_protected:cpx
+%   }
+%   \begin{syntax}
+%     \cs{cs_gset_protected:Npn} \meta{function} \meta{parameters}
+%     ~~\Arg{code}
+%   \end{syntax}
+%   Globally sets \meta{function} to expand to \meta{code} as replacement
+%   text. Within the \meta{code}, the \meta{parameters} (|#1|, |#2|, 
+%   \emph{etc.}) will be replaced by those absorbed by the function.
+%   The assignment of a meaning to \meta{function} is \emph{not}
+%   restricted to the current \TeX{} group level: the assignment is
+%   global. The \meta{function} will not expand within an
+%   \texttt{x}-type argument.
+% \end{function}
+%
+% \begin{function}
+%   {
+%     \cs_gset_protected_nopar:Npn, \cs_gset_protected_nopar:cpn,
+%     \cs_gset_protected_nopar:Npx, \cs_gset_protected_nopar:cpx
+%   }
+%   \begin{syntax}
+%     \cs{cs_gset_protected_nopar:Npn} \meta{function}
+%     ~~\meta{parameters} \Arg{code}
+%   \end{syntax}
+%   Globally sets \meta{function} to expand to \meta{code} as replacement
+%   text. Within the \meta{code}, the \meta{parameters} (|#1|, |#2|, 
+%   \emph{etc.}) will be replaced by those absorbed by the function.
+%   When the \meta{function} is used the \meta{parameters} absorbed
+%   cannot contain \cs{par} tokens. The assignment of a meaning to
+%   \meta{function} is \emph{not} restricted to the current \TeX{}
+%   group level: the assignment is global. The \meta{function} will
+%   not expand within an \texttt{x}-type argument.
+% \end{function}
+% 
+% \subsection{Defining new functions using the signature}
+% 
+% \begin{function}
+%   {
+%     \cs_new:Nn, \cs_new:cn,
+%     \cs_new:Nx, \cs_new:cx
+%   }
+%   \begin{syntax}
+%     \cs{cs_new:Nn} \meta{function} \Arg{code}
+%   \end{syntax}
+%   Creates \meta{function} to expand to \meta{code} as replacement text.
+%   Within the \meta{code}, the number of \meta{parameters} is detected
+%   automatically from the function signature. These \meta{parameters}
+%   (|#1|, |#2|, \emph{etc.}) will be replaced by those absorbed by the
+%   function. The definition is global and
+%   an error will result if the \meta{function} is already defined.
+% \end{function}
+% 
+% \begin{function}
+%   {
+%     \cs_new_nopar:Nn, \cs_new_nopar:cn,
+%     \cs_new_nopar:Nx, \cs_new_nopar:cx
+%   }
+%   \begin{syntax}
+%     \cs{cs_new_nopar:Nn} \meta{function} \Arg{code}
+%   \end{syntax}
+%   Creates \meta{function} to expand to \meta{code} as replacement text.
+%   Within the \meta{code}, the number of \meta{parameters} is detected
+%   automatically from the function signature. These \meta{parameters}
+%   (|#1|, |#2|, \emph{etc.}) will be replaced by those absorbed by the
+%   function.  When the \meta{function} is used the \meta{parameters}
+%   absorbed cannot contain \cs{par} tokens. The definition is global and
+%   an error will result if the \meta{function} is already defined.
+% \end{function}
+% 
+% \begin{function}
+%   {
+%     \cs_new_protected:Nn, \cs_new_protected:cn,
+%     \cs_new_protected:Nx, \cs_new_protected:cx
+%   }
+%   \begin{syntax}
+%     \cs{cs_new_protected:Nn} \meta{function} \Arg{code}
+%   \end{syntax}
+%   Creates \meta{function} to expand to \meta{code} as replacement text.
+%   Within the \meta{code}, the number of \meta{parameters} is detected
+%   automatically from the function signature. These \meta{parameters}
+%   (|#1|, |#2|, \emph{etc.}) will be replaced by those absorbed by the
+%   function. The \meta{function} will not expand within an \texttt{x}-type
+%   argument. The definition is global and
+%   an error will result if the \meta{function} is already defined.
+% \end{function}
+% 
+% \begin{function}
+%   {
+%     \cs_new_protected_nopar:Nn, \cs_new_protected_nopar:cn,
+%     \cs_new_protected_nopar:Nx, \cs_new_protected_nopar:cx
+%   }
+%   \begin{syntax}
+%     \cs{cs_new_protected_nopar:Nn} \meta{function} \Arg{code}
+%   \end{syntax}
+%   Creates \meta{function} to expand to \meta{code} as replacement text.
+%   Within the \meta{code}, the number of \meta{parameters} is detected
+%   automatically from the function signature. These \meta{parameters}
+%   (|#1|, |#2|, \emph{etc.}) will be replaced by those absorbed by the
+%   function.  When the \meta{function} is used the \meta{parameters}
+%   absorbed cannot contain \cs{par} tokens. The \meta{function} will not
+%   expand within an \texttt{x}-type argument. The definition is global and
+%   an error will result if the \meta{function} is already defined.
+% \end{function}
+% 
+% \begin{function}
+%   {
+%     \cs_set:Nn, \cs_set:cn,
+%     \cs_set:Nx, \cs_set:cx
+%   }
+%   \begin{syntax}
+%     \cs{cs_set:Nn} \meta{function} \Arg{code}
+%   \end{syntax}
+%   Sets \meta{function} to expand to \meta{code} as replacement text.
+%   Within the \meta{code}, the number of \meta{parameters} is detected
+%   automatically from the function signature. These \meta{parameters}
+%   (|#1|, |#2|, \emph{etc.}) will be replaced by those absorbed by the
+%   function. 
+%   The assignment of a meaning to \meta{function} is restricted to
+%   the current \TeX{} group level.
+% \end{function}
+% 
+% \begin{function}
+%   {
+%     \cs_set_nopar:Nn, \cs_set_nopar:cn,
+%     \cs_set_nopar:Nx, \cs_set_nopar:cx
+%   }
+%   \begin{syntax}
+%     \cs{cs_set_nopar:Nn} \meta{function} \Arg{code}
+%   \end{syntax}
+%   Sets \meta{function} to expand to \meta{code} as replacement text.
+%   Within the \meta{code}, the number of \meta{parameters} is detected
+%   automatically from the function signature. These \meta{parameters}
+%   (|#1|, |#2|, \emph{etc.}) will be replaced by those absorbed by the
+%   function.  When the \meta{function} is used the \meta{parameters}
+%   absorbed cannot contain \cs{par} tokens.
+%   The assignment of a meaning to \meta{function} is restricted to
+%   the current \TeX{} group level.
+% \end{function}
+% 
+% \begin{function}
+%   {
+%     \cs_set_protected:Nn, \cs_set_protected:cn,
+%     \cs_set_protected:Nx, \cs_set_protected:cx
+%   }
+%   \begin{syntax}
+%     \cs{cs_set_protected:Nn} \meta{function} \Arg{code}
+%   \end{syntax}
+%   Sets \meta{function} to expand to \meta{code} as replacement text.
+%   Within the \meta{code}, the number of \meta{parameters} is detected
+%   automatically from the function signature. These \meta{parameters}
+%   (|#1|, |#2|, \emph{etc.}) will be replaced by those absorbed by the
+%   function. The \meta{function} will not expand within an \texttt{x}-type
+%   argument.
+%   The assignment of a meaning to \meta{function} is restricted to
+%   the current \TeX{} group level.
+% \end{function}
+% 
+% \begin{function}
+%   {
+%     \cs_set_protected_nopar:Nn, \cs_set_protected_nopar:cn,
+%     \cs_set_protected_nopar:Nx, \cs_set_protected_nopar:cx
+%   }
+%   \begin{syntax}
+%     \cs{cs_set_protected_nopar:Nn} \meta{function} \Arg{code}
+%   \end{syntax}
+%   Sets \meta{function} to expand to \meta{code} as replacement text.
+%   Within the \meta{code}, the number of \meta{parameters} is detected
+%   automatically from the function signature. These \meta{parameters}
+%   (|#1|, |#2|, \emph{etc.}) will be replaced by those absorbed by the
+%   function.  When the \meta{function} is used the \meta{parameters}
+%   absorbed cannot contain \cs{par} tokens. The \meta{function} will not
+%   expand within an \texttt{x}-type argument.
+%   The assignment of a meaning to \meta{function} is restricted to
+%   the current \TeX{} group level.
+% \end{function}
+% 
+% \begin{function}
+%   {
+%     \cs_gset:Nn, \cs_gset:cn,
+%     \cs_gset:Nx, \cs_gset:cx
+%   }
+%   \begin{syntax}
+%     \cs{cs_gset:Nn} \meta{function} \Arg{code}
+%   \end{syntax}
+%   Sets \meta{function} to expand to \meta{code} as replacement text.
+%   Within the \meta{code}, the number of \meta{parameters} is detected
+%   automatically from the function signature. These \meta{parameters}
+%   (|#1|, |#2|, \emph{etc.}) will be replaced by those absorbed by the
+%   function. 
+%   The assignment of a meaning to \meta{function} is  global.
+% \end{function}
+% 
+% \begin{function}
+%   {
+%     \cs_gset_nopar:Nn, \cs_gset_nopar:cn,
+%     \cs_gset_nopar:Nx, \cs_gset_nopar:cx
+%   }
+%   \begin{syntax}
+%     \cs{cs_gset_nopar:Nn} \meta{function} \Arg{code}
+%   \end{syntax}
+%   Sets \meta{function} to expand to \meta{code} as replacement text.
+%   Within the \meta{code}, the number of \meta{parameters} is detected
+%   automatically from the function signature. These \meta{parameters}
+%   (|#1|, |#2|, \emph{etc.}) will be replaced by those absorbed by the
+%   function.  When the \meta{function} is used the \meta{parameters}
+%   absorbed cannot contain \cs{par} tokens.
+%   The assignment of a meaning to \meta{function} is global.
+% \end{function}
+% 
+% \begin{function}
+%   {
+%     \cs_gset_protected:Nn, \cs_gset_protected:cn,
+%     \cs_gset_protected:Nx, \cs_gset_protected:cx
+%   }
+%   \begin{syntax}
+%     \cs{cs_gset_protected:Nn} \meta{function} \Arg{code}
+%   \end{syntax}
+%   Sets \meta{function} to expand to \meta{code} as replacement text.
+%   Within the \meta{code}, the number of \meta{parameters} is detected
+%   automatically from the function signature. These \meta{parameters}
+%   (|#1|, |#2|, \emph{etc.}) will be replaced by those absorbed by the
+%   function. The \meta{function} will not expand within an \texttt{x}-type
+%   argument.
+%   The assignment of a meaning to \meta{function} is  global.
+% \end{function}
+% 
+% \begin{function}
+%   {
+%     \cs_gset_protected_nopar:Nn, \cs_gset_protected_nopar:cn,
+%     \cs_gset_protected_nopar:Nx, \cs_gset_protected_nopar:cx
+%   }
+%   \begin{syntax}
+%     \cs{cs_gset_protected_nopar:Nn} \meta{function} \Arg{code}
+%   \end{syntax}
+%   Sets \meta{function} to expand to \meta{code} as replacement text.
+%   Within the \meta{code}, the number of \meta{parameters} is detected
+%   automatically from the function signature. These \meta{parameters}
+%   (|#1|, |#2|, \emph{etc.}) will be replaced by those absorbed by the
+%   function.  When the \meta{function} is used the \meta{parameters}
+%   absorbed cannot contain \cs{par} tokens. The \meta{function} will not
+%   expand within an \texttt{x}-type argument.
+%   The assignment of a meaning to \meta{function} is global.
+% \end{function}
+% 
+% \begin{function}
+%   {\cs_generate_from_arg_count:NNnn, \cs_generate_from_arg_count:cNnn}
+%   \begin{syntax}
+%     \cs{cs_generate_from_arg_count:NNnn} \meta{function}
+%     ~~\meta{creator} \meta{number} \meta{code}
+%   \end{syntax}
+%   Uses the \meta{creator} function (which should have signature
+%   |Npn|, for example \cs{cs_new:Npn}) to define a \meta{function}
+%   which takes \meta{number} arguments and has \meta{code} as
+%   replacement text. The \meta{number} of arguments is an integer expression,
+%   evaluated as detailed for \cs{int_eval:n}.
+% \end{function}
+%
+% \subsection{Copying control sequences}
+%
+% Control sequences (not just functions as defined above) can
+% be set to have the same meaning using the functions described
+% here. Making two control sequences equivalent means that the
+% second control sequence is a \emph{copy} of the first (rather than
+% a pointer to it). Thus the old and new control sequence are not
+% tided together: changes to one are not reflected in the other.
+% 
+% In the following text \enquote{cs} is used as an abbreviation for
+% \enquote{control sequence}.
+% 
+% \begin{function}
+%   {\cs_new_eq:NN, \cs_new_eq:Nc, \cs_new_eq:cN, \cs_new_eq:cc}
+%   \begin{syntax}
+%     \cs{cs_new_eq:NN} \meta{cs 1} \meta{cs 2}
+%   \end{syntax}
+%   Creates \meta{control sequence 1} and sets it to have the same
+%   meaning as \meta{control sequence 2} at the point where
+%   \cs{cs_new_eq:NN} is executed. The two control sequences may
+%   subsequently be altered without affecting the copy. The assignment
+%   of a meaning to \meta{control sequence 1} is global.
+% \end{function}
+% 
+% \begin{function}
+%   {\cs_set_eq:NN, \cs_set_eq:Nc, \cs_set_eq:cN, \cs_set_eq:cc}
+%   \begin{syntax}
+%     \cs{cs_set_eq:NN} \meta{cs 1} \meta{cs 2}
+%   \end{syntax}
+%   Sets \meta{control sequence 1} to have the same meaning as
+%   \meta{control sequence 2} at the point where \cs{cs_set_eq:NN}
+%   is executed. The two control sequences may subsequently be
+%   altered without affecting the copy. The assignment of a meaning
+%   to \meta{control sequence 1} is restricted to the current
+%   \TeX{} group level.
+% \end{function}
+%
+% \begin{function}
+%   {\cs_gset_eq:NN, \cs_gset_eq:Nc, \cs_gset_eq:cN, \cs_gset_eq:cc}
+%   \begin{syntax}
+%     \cs{cs_gset_eq:NN} \meta{cs 1} \meta{cs 2}
+%   \end{syntax}
+%   Globally sets \meta{control sequence 1} to have the same meaning as
+%   \meta{control sequence 2} at the point where \cs{cs_gset_eq:NN}
+%   is executed. The two control sequences may subsequently be
+%   altered without affecting the copy. The assignment of a meaning to
+%   \meta{control sequence 1} is \emph{not} restricted to the current
+%   \TeX{} group level: the assignment is global.
+% \end{function}
+%
+% \subsection{Deleting control sequences}
+%
+% There are occasions where control sequences need to be deleted.
+% This is handled in a very simple manner.
+% 
+% \begin{function}{\cs_undefine:N, \cs_undefine:c}
+%   \begin{syntax}
+%     \cs{cs_undefine:N} \meta{control sequence}
+%   \end{syntax}
+%   Sets \meta{control sequence} to be globally undefined.
+% \end{function}
+%
+% \subsection{Showing control sequences}
+% 
+% \begin{function}[EXP]{\cs_meaning:N, \cs_meaning:c}
+%   \begin{syntax}
+%     \cs{cs_meaning:N} \meta{control sequence}
+%   \end{syntax}
+%   This function expands to the \emph{meaning} of the \meta{control sequence}
+%   control sequence. This will show the \meta{replacement text} for a
+%   macro.
+%   \begin{texnote}
+%     This is \TeX{}'s \cs{meaning} primitive.
+%   \end{texnote}
+% \end{function}
+%
+% \begin{function}{\cs_show:N, \cs_show:c}
+%   \begin{syntax}
+%     \cs{cs_show:N} \meta{control sequence}
+%   \end{syntax}
+%   Displays the definition of the \meta{control sequence} on the
+%   terminal.
+%   \begin{texnote}
+%     This is the \TeX{} primitive \cs{show}.
+%   \end{texnote}
+% \end{function}
+% 
+
+% \subsection{Converting to and from control sequences}
+%
+% \begin{function}[EXP]{\use:c}
+%   \begin{syntax}
+%     \cs{use:c} \Arg{control sequence name}
+%   \end{syntax}
+%   Converts the given \meta{control sequence name} into a single
+%   control sequence token. This process requires two expansions. 
+%   The content for \meta{control sequence name} may be literal
+%   material or from other expandable functions. The
+%   \meta{control sequence name} must, when fully expanded, consist
+%   of character tokens which are not active: typically, they will
+%   be of category code $10$ (space), $11$ (letter)
+%   or $12$ (other), or a mixture of these. As an example, both
+%   \begin{verbatim}
+%     \use:c { a b c }
+%   \end{verbatim}
+%   and
+%   \begin{verbatim}
+%     \tl_new:N  \l_my_tl
+%     \tl_set:Nn \l_my_tl { a b c }
+%     \use:c { \tl_use:N \l_my_tl }
+%   \end{verbatim}
+%   would be equivalent to
+%   \begin{verbatim}
+%     \abc
+%   \end{verbatim}
+%   after two expansions of \cs{use:c}.
+% \end{function}
+% 
+% \begin{function}[EXP]{\cs:w, \cs_end:}
+%   \begin{syntax}
+%     \cs{cs:w} \meta{control sequence name} \cs{cs_end:}
+%   \end{syntax}
+%   Converts the given \meta{control sequence name} into a single
+%   control sequence token. This process requires one expansion. 
+%   The content for \meta{control sequence name} may be literal
+%   material or from other expandable functions. The
+%   \meta{control sequence name} must, when fully expanded, consist
+%   of character tokens which are not active: typically, they will
+%   be of category code $10$ (space), $11$ (letter)
+%   or $12$ (other), or a mixture of these. As an example, both
+%   \begin{verbatim}
+%     \cs:w a b c \cs_end:
+%   \end{verbatim}
+%   and
+%   \begin{verbatim}
+%     \tl_new:N  \l_my_tl
+%     \tl_set:Nn \l_my_tl { a b c }
+%     \cs:w \tl_use:N \l_my_tl \cs_end:
+%   \end{verbatim}
+%   would be equivalent to
+%   \begin{verbatim}
+%     \abc
+%   \end{verbatim}
+%   after one expansion of \cs{cs:w}.
+%   \begin{texnote}
+%     These are the \TeX{} primitives \cs{csname} and \cs{endcsname}.
+%   \end{texnote}
+% \end{function}
+%
+% \begin{function}[EXP]{\cs_to_str:N}
+%   \begin{syntax}
+%     \cs{cs_to_str:N} \Arg{control sequence}
+%   \end{syntax}
+%   Converts the given \meta{control sequence} into a series of
+%   characters with category code $12$ (other), except spaces,
+%   of category code $10$. The sequence will \emph{not} include
+%   the current escape token, \emph{cf.}~\cs{token_to_str:N}.
+%   Full expansion of this function requires a variable number
+%   of expansion steps (either 3 or 4), and so an
+%   \texttt{f}- or \texttt{x}-type expansion will be required to
+%   convert the \meta{control sequence} to a sequence of characters
+%   in the input stream.
+% \end{function}
+%
+% \section{Using or removing tokens and arguments}
+%
+% Tokens in the input can be read and used or read and discarded.
+% If one or more tokens are wrapped in braces then in absorbing them
+% the outer set will be removed. At the same time, the category code
+% of each token is set when the token is read by a function (if it
+% is read more than once, the category code is determined by the
+% the situation in force when first function absorbs the token).
+% 
+% \begin{function}[EXP]{\use:n, \use:nn, \use:nnn, \use:nnnn}
+%   \begin{syntax}
+%     \cs{use:n}   \Arg{group1}
+%     \cs{use:nn}  \Arg{group1} \Arg{group2}
+%     \cs{use:nnn} \Arg{group1} \Arg{group2} \Arg{group3}
+%     \cs{use:nnn} \Arg{group1} \Arg{group2} \Arg{group3} \Arg{group4}
+%   \end{syntax}
+%   As illustrated, these functions will absorb between one and four
+%   arguments, as indicated by the argument specifier. The braces
+%   surrounding each argument will be removed leaving the remaining
+%   tokens in the input stream. The category code of these tokens will
+%   also be fixed by this process (if it has not already been by some
+%   other absorption). All of these functions require only a single
+%   expansion to operate, so that one expansion of
+%   \begin{verbatim}
+%     \use:nn { abc } { { def } }
+%   \end{verbatim}
+%   will result in the input stream containing
+%   \begin{verbatim}
+%     abc { def }
+%   \end{verbatim}
+%   \emph{i.e.} only the outer braces will be removed.
+% \end{function}
+%
+% \begin{function}[EXP]{\use_i:nn, \use_ii:nn}
+%    \begin{syntax}
+%     \cs{use_i:nn} \Arg{group1} \Arg{group2}
+%   \end{syntax}
+%   These functions will absorb two groups and leave only the
+%   first or the second in the input stream. The braces surrounding the
+%   arguments will be removed as part of this process. The category code
+%   of these tokens will also be fixed (if it has not already been by
+%   some other absorption). A single expansion is needed for the
+%   functions to take effect.
+% \end{function}
+%
+% \begin{function}[EXP]{\use_i:nnn, \use_ii:nnn, \use_iii:nnn}
+%   \begin{syntax}
+%     \cs{use_i:nnn} \Arg{group1} \Arg{group2} \Arg{group3}
+%   \end{syntax}
+%   These functions will absorb three groups and leave only of these
+%   in the input stream. The braces surrounding the arguments will be
+%   removed as part of this process. The category code of these tokens
+%   will also be fixed (if it has not already been by some other
+%   absorption). A single expansion is needed for the functions to take
+%   effect.
+% \end{function}
+%
+% \begin{function}[EXP]
+%   {\use_i:nnnn, \use_ii:nnnn, \use_iii:nnnn , \use_iv:nnnn}
+%   \begin{syntax}
+%     \cs{use_i:nnnn} \Arg{group1} \Arg{group2} \Arg{group3} \Arg{group4}
+%   \end{syntax}
+%   These functions will absorb four groups and leave only of these
+%   in the input stream. The braces surrounding the arguments will be
+%   removed as part of this process. The category code of these tokens
+%   will also be fixed (if it has not already been by some other
+%   absorption). A single expansion is needed for the functions to take
+%   effect.
+% \end{function}
+%
+% \begin{function}[EXP]{\use_i_ii:nnn}
+%   \begin{syntax}
+%     \cs{use_i_ii:nnn} \Arg{group1} \Arg{group2} \Arg{group3}
+%   \end{syntax}
+%   This functions will absorb three groups and leave the first and
+%   second in the input stream. The braces surrounding the arguments
+%   will be removed as part of this process. The category code of
+%   these tokens will also be fixed (if it has not already been by
+%   some other absorption). A single expansion is needed for the
+%   functions to take effect. An example:
+%   \begin{verbatim}
+%     \use_i_ii:nnn { abc } { { def } } { ghi }
+%   \end{verbatim}
+%   will result in the input stream containing
+%   \begin{verbatim}
+%     abc { def }
+%   \end{verbatim}
+%   \emph{i.e.} the outer braces will be removed and the third group
+%   will be removed.
+% \end{function}
+%
+% \begin{function}[EXP]
+%   {
+%     \use_none:n         ,
+%     \use_none:nn        ,
+%     \use_none:nnn       ,
+%     \use_none:nnnn      ,
+%     \use_none:nnnnn     ,
+%     \use_none:nnnnnn    ,
+%     \use_none:nnnnnnn   ,
+%     \use_none:nnnnnnnn  ,
+%     \use_none:nnnnnnnnn
+%   }
+%   \begin{syntax}
+%     \cs{use_none:n} \Arg{group1}
+%   \end{syntax}
+%   These functions absorb between one and nine groups from the
+%   input stream, leaving nothing on the resulting input stream.
+%   These functions work after a single expansion. One or more of the
+%   \texttt{n} arguments may be an unbraced single token
+%   (\emph{i.e.}~an \texttt{N} argument).
+% \end{function}
+%
+% \begin{function}{\use:x}
+%   \begin{syntax}
+%     \cs{use:x} \Arg{expandable tokens}
+%   \end{syntax}
+%   Fully expands the \meta{expandable tokens} and inserts the
+%   result into the input stream at the current location.
+% \end{function}
+% 
+% \subsection{Selecting tokens from delimited arguments}
+%
+% A different kind of function for selecting tokens from the token
+% stream are those that use delimited arguments.
+%
+% \begin{function}[EXP]
+%   {
+%     \use_none_delimit_by_q_nil:w,
+%     \use_none_delimit_by_q_stop:w,
+%     \use_none_delimit_by_q_recursion_stop:w
+%   }
+%   \begin{syntax}
+%    \cs{use_none_delimit_by_q_nil:w} \meta{balanced text} \cs{q_nil}
+%   \end{syntax}
+%   Absorb the \meta{balanced} text form the input stream delimited by
+%   the marker given in the function name, leaving nothing in the
+%   input stream.
+% \end{function}
+% 
+% \begin{function}[EXP]
+%   {
+%     \use_i_delimit_by_q_nil:nw,
+%     \use_i_delimit_by_q_stop:nw,
+%     \use_i_delimit_by_q_recursion_stop:nw
+%   }
+%   \begin{syntax}
+%    \cs{use_i_delimit_by_q_nil:nw} \Arg{inserted tokens}
+%    ~~\meta{balanced text} \cs{q_nil}
+%   \end{syntax}
+%   Absorb the \meta{balanced} text form the input stream delimited by
+%   the marker given in the function name, leaving \meta{inserted tokens}
+%   in the input stream for further processing.
+% \end{function}
+%
+%
+% \begin{function}[EXP]
+%   {
+%     \use_i_after_fi:nw,
+%     \use_i_after_else:nw,
+%     \use_i_after_or:nw,
+%     \use_i_after_orelse:nw
+%   }
+%   \begin{syntax}
+%      \cs{use_i_after_fi:nw}     \Arg{inserted tokens} \cs{fi:}
+%      \cs{use_i_after_else:nw}   \Arg{inserted tokens} \cs{else:}
+%      ~~\meta{balanced text} \cs{fi:}
+%      \cs{use_i_after_or:nw}     \Arg{inserted tokens} \cs{or:}
+%      ~~\meta{balanced text} \cs{fi:}
+%      \cs{use_i_after_orelse:nw} \Arg{inserted tokens} \cs{or:} or \cs{else:}
+%      ~~\meta{balanced text} \cs{fi:}
+%   \end{syntax}
+%   Absorb the \meta{balanced text}, if appropriate, delimited by
+%   the function name given. The \meta{inserted tokens} are then placed
+%   in the input stream after the delimiter. Thus for example
+%   \begin{verbatim}
+%     \use_i_after_fi:nw { some tokens } \fi:
+%   \end{verbatim}
+%   will leave
+%   \begin{verbatim}
+%     \fi:  some tokens
+%   \end{verbatim}
+%   in the input stream for further processing. See the discussion of the
+%   primitive \TeX{} conditionals for more detail on \cs{else:}, \cs{fi:}
+%   and \cs{or:}.
+% \end{function}
+%
+% \subsection{Decomposing control sequences} 
+%
+% \begin{function}[EXP]{\cs_get_arg_count_from_signature:N}
+%   \begin{syntax}
+%     \cs{cs_get_arg_count_from_signature:N} \meta{function}
+%   \end{syntax}
+%   Splits the \meta{function} into the name (\emph{i.e.}~the part
+%   before the colon) and the signature (\emph{i.e.}~after the colon).
+%   The \meta{number} of tokens in the \meta{signature} is then left in
+%   the input stream. If there was no \meta{signature} then the result is
+%   the marker value $-1$.
+% \end{function}
+%
+% \begin{function}[EXP]{\cs_get_function_name:N}
+%   \begin{syntax}
+%     \cs{cs_get_function_name:NN} \meta{function}
+%   \end{syntax}
+%   Splits the \meta{function} into the name (\emph{i.e.}~the part
+%   before the colon) and the signature (\emph{i.e.}~after the colon).
+%   The \meta{name} is then left in the input stream without the escape
+%   character present made up of tokens with category code $12$
+%   (other).
+% \end{function}
+%
+% \begin{function}{[EXP]\cs_get_function_signature:N}
+%   \begin{syntax}
+%     \cs{cs_get_function_signature:NN} \meta{function}
+%   \end{syntax}
+%   Splits the \meta{function} into the name (\emph{i.e.}~the part
+%   before the colon) and the signature (\emph{i.e.}~after the colon).
+%   The \meta{signature} is then left in the input stream made up of
+%   tokens with category code $12$ (other).
+% \end{function}
+% 
+% \begin{function}EXP]{\cs_split_function:NN}
+%   \begin{syntax}
+%     \cs{cs_split_function:NN} \meta{function} \meta{processor}
+%   \end{syntax}
+%   Splits the \meta{function} into the name (\emph{i.e.}~the part
+%   before the colon) and the signature (\emph{i.e.}~after the colon).
+%   This information is then placed in the input stream after the
+%   \meta{processor} function in three parts: the \meta{name}, the
+%   \meta{signature} and a logic token indicating if a colon was found
+%   (to differentiate variables from function names). The \meta{name}
+%   will not include the escape character, and both the \meta{name} and
+%   \meta{signature} are made up of tokens with category code $12$
+%   (other). The \meta{processor} should be a function with argument
+%   specification \texttt{:nnN} (plus any trailing arguments needed).
+% \end{function}
+%
+% \begin{function}[EXP]{\cs_to_str:N}
+%   \begin{syntax}
+%     \cs{cs_to_str:N} \Arg{control sequence}
+%   \end{syntax}
+%   Converts the given \meta{control sequence} into a series of
+%   characters with category code $12$ (other), except spaces,
+%   of category code $10$. The sequence will \emph{not} include
+%   the current escape token, \emph{cf.}~\cs{token_to_str:N}.
+%   Full expansion of this function requires a variable number
+%   of expansion steps (either 3 or 4), and so an
+%   \texttt{f}- or \texttt{x}-type expansion will be required to
+%   convert the \meta{control sequence} to a sequence of characters
+%   in the input stream.
+% \end{function}
+%
+% \section{Predicates and conditionals}
+% \label{sec:predicates}
+%
+% \LaTeX3 has three concepts for conditional flow processing:
+% \begin{description}
+%   \item[Branching conditionals]
+%     Functions that carry out a test and then execute, depending on its
+%     result, either the code supplied in the \meta{true arg} or the 
+%     \meta{false arg}. 
+%     These arguments are denoted with |T| and |F|, respectively. An
+%     example would be
+%     \begin{quote}
+%      |\cs_if_free:cTF{abc}| \Arg{true code} \Arg{false code}
+%     \end{quote}
+%     a function that will turn the first argument into a control sequence
+%     (since it's marked as |c|) then checks whether this control sequence
+%     is still free and then depending on the result carry out the code in
+%     the second argument (true case) or in the third argument (false
+%     case).
+%
+%     These type of functions are known as \enquote{conditionals}; 
+%     whenever a |TF| function is defined it will usually be accompanied by
+%     |T| and |F| functions as well. These are provided for convenience when
+%     the branch only needs to go a single way. Package writers are free to
+%     choose which types to define but the kernel definitions will always
+%     provide all three versions.
+%
+%     Important to note is that these branching conditionals with \meta{true
+%     code} and/or \meta{false code} are always defined in a way that the
+%     code of the chosen alternative can operate on following tokens in
+%     the input stream.
+%
+%     These conditional functions may or may not be fully expandable, but if
+%     they are expandable they will be accompanied by a \enquote{predicate}
+%     for the same test as described below.
+%
+%   \item[Predicates]
+%     \enquote{Predicates} are functions that return a special type of
+%     boolean value which can be tested by the function \cs{if_predicate:w}
+%     or in the boolean expression parser. All functions of this type
+%     are expandable and have names that end with |_p| in the
+%     description part.  For example,
+%     \begin{quote}
+%       |\cs_if_free_p:N|
+%     \end{quote}
+%     would be a predicate function for the same type of test as the
+%     conditional described above. It would return \enquote{true} if its
+%     argument (a single token denoted by |N|) is still free for definition.
+%     It would be used in constructions like
+%     \begin{quote}
+%       |\if_predicate:w \cs_if_free_p:N \l_tmpz_tl| \\
+%       |  |\meta{true code} \\
+%       |else:| \\
+%       |  |\meta{false code}\\
+%       |\fi:| \\
+%     \end{quote}
+%     or in expressions utilizing the boolean logic parser:
+%     \begin{quote}
+%       |\bool_if:nTF {| \\
+%       \verb"  \cs_if_free_p:N \l_tmpz_tl || \cs_if_free_p:N \g_tmpz_tl " \\
+%       |}|
+%       \Arg{true code} \Arg{false code} 
+%     \end{quote}
+% 
+%     Like their branching cousins, predicate functions ensure that all
+%     underlying primitive |\else:| or |\fi:| have been removed before
+%     returning the boolean true or false values.\footnote{If defined
+%     using the interface provided.}
+%
+%     For each predicate defined, a \enquote{predicate conditional} will
+%     also exist that behaves like a conditional described above.
+%
+%   \item[Primitive conditionals]
+%      There is a third variety of conditional, which is the original
+%      concept used in plain \TeX{} and \LaTeXe{}. Their use is discouraged
+%      in \pkg{expl3} (although still used in low-level definitions)
+%      because they are more fragile and in many cases require more
+%      expansion control (hence more code) than the two types of 
+%      conditionals described above.
+% \end{description}
+% 
+% \begin{variable}{\c_true_bool, \c_false_bool}
+%   Constants that represent |true| and |false|, respectively. Used to
+%   implement predicates.
+% \end{variable} 
+% 
+% \subsection{Tests on control sequences}
+%
+% \begin{function}[EXP,pTF]{\cs_if_eq:NN}
+%   \begin{syntax}
+%     \cs{cs_if_eq_p:NN} \Arg{cs1} \Arg{cs2}
+%     \cs{cs_if_eq:NNTF} \Arg{cs1} \Arg{cs2} \Arg{true code}
+%     ~~\Arg{false code}
+%   \end{syntax}
+%   Compares the definition of two \meta{control sequences} and
+%   is logically \texttt{true} if the two are the same. The branching
+%   versions then leave either \meta{true code} or \meta{false code}
+%   in the input stream, as appropriate to the truth of the test and the
+%   variant of the function chosen. The logical truth of the test is left
+%   in the input stream by the predicate version.
+% \end{function}
+%
+% \begin{function}[EXP,pTF]{\cs_if_exist:N, \cs_if_exist:c}
+%   \begin{syntax}
+%     \cs{cs_if_exist_p:N} \meta{control sequence}
+%     \cs{cs_if_exist:NTF} \meta{control sequence} \meta{true code}
+%     ~~\meta{false code}
+%   \end{syntax}
+%   Tests whether the \meta{control sequence} is currently defined
+%   (whether as a function or another control sequence type). Any
+%   valid definition of \meta{control sequence} will evaluate as
+%   \texttt{true}. The branching versions then leave either
+%   \meta{true code} or \meta{false code}  in the input stream, as
+%   appropriate to the truth of the test and the variant of the function
+%   chosen. The logical truth of the test is left in the input stream by
+%   the predicate version.
+% \end{function}
+%
+% \begin{function}[EXP,pTF]{\cs_if_free:N, \cs_if_free:c}
+%   \begin{syntax}
+%     \cs{cs_if_free_p:N} \meta{control sequence}
+%     \cs{cs_if_free:NTF} \meta{control sequence} \meta{true code}
+%     ~~\meta{false code}
+%   \end{syntax}
+%   Tests whether the \meta{control sequence} is currently free to
+%   be defined. This test will be \texttt{false} if the
+%   \meta{control sequence} currently exists (as defined by
+%   \cs{cs_if_exist:N}). The branching versions then leave either
+%   \meta{true code} or \meta{false code}  in the input stream, as
+%   appropriate to the truth of the test and the variant of the function
+%   chosen. The logical truth of the test is left in the input stream by
+%   the predicate version.
+% \end{function}
+% 
+% \subsection{Testing string equality}
+%
+% \begin{function}[EXP,pTF]
+%   { 
+%     \str_if_eq:nn, \str_if_eq:Vn, \str_if_eq:on, \str_if_eq:no,
+%     \str_if_eq:nV, \str_if_eq:VV, \str_if_eq:xx
+%   }
+%   \begin{syntax}
+%     \cs{str_if_eq_p:nn} \Arg{tl1} \Arg{tl2}
+%     \cs{str_if_eq:nnTF} \Arg{tl1} \Arg{tl2} \Arg{true code} \Arg{false code}
+%   \end{syntax}
+%   Compares the two \meta{token lists} on a character by character
+%   basis, and is \texttt{true} if the two lists contain the same
+%   characters in the same order. Thus for example
+%   \begin{verbatim}
+%     \str_if_eq_p:xx { abc } { \tl_to_str:n { abc } }
+%   \end{verbatim}
+%   is logically \texttt{true}. The branching versions then leave either
+%   \meta{true code} or \meta{false code} in the input stream, as
+%   appropriate to the truth of the test and the variant of the function
+%   chosen. The logical truth of the test is left in the input stream by
+%   the predicate version. All versions of these functions are fully
+%   expandable (including those involving an \texttt{x}-type
+%   expansion).
+% \end{function}
+% 
+% \subsection{Engine-specific conditionals}
+%
+% \begin{function}[EXP,TF]{\luatex_if_engine:}
+%   \begin{syntax}
+%     \cs{luatex_if_luatex:TF} \Arg{true code} \Arg{false code}
+%   \end{syntax}
+%   Detects is the document is being compiled using \LuaTeX{}. The
+%   branching versions then leave either \meta{true code} or
+%   \meta{false code} in the input stream, as appropriate to the truth
+%   of the test and the variant of the function chosen.
+% \end{function}
+%
+% \begin{function}[EXP,TF]{\pdftex_if_engine:}
+%   \begin{syntax}
+%     \cs{pdftex_if_engine:TF} \Arg{true code} \Arg{false code}
+%   \end{syntax}
+%   Detects is the document is being compiled using \pdfTeX{}. The
+%   branching versions then leave either \meta{true code} or
+%   \meta{false code} in the input stream, as appropriate to the truth
+%   of the test and the variant of the function chosen.
+% \end{function}
+%
+% \begin{function}[EXP,TF]{\xetex_if_engine:}
+%   \begin{syntax}
+%     \cs{xetex_if_engine:TF} \Arg{true code} \Arg{false code}
+%   \end{syntax}
+%   Detects is the document is being compiled using \XeTeX{}. The
+%   branching versions then leave either \meta{true code} or
+%   \meta{false code} in the input stream, as appropriate to the truth
+%   of the test and the variant of the function chosen.
+% \end{function}
+%
+% \begin{variable}
+%   {
+%     \c_luatex_is_engine_bool, \c_pdftex_is_engine_bool,
+%     \c_xetex_is_engine_bool
+%   }
+%   Boolean versions of the engine conditionals, for use in predicate
+%   tests.
+% \end{variable}
+%
+% \subsection{Primitive conditionals}
+%
+% The \eTeX{} engine itself provides many different conditionals. Some
+% expand whatever comes after them and others don't. Hence the names
+% for these underlying functions will often contain a |:w| part but
+% higher level functions are often available. See for instance
+% |\int_compare_p:nNn| which is a wrapper for |\if_num:w|.
+%
+% Certain conditionals deal with specific data types like boxes and
+% fonts and are described there. The ones described below are either
+% the universal conditionals or deal with control sequences. We will
+% prefix primitive conditionals with |\if_|.
+% 
+% \begin{function}[EXP]
+%   {\if_true:, \if_false:, \or:, \else:, \fi:, \reverse_if:N}
+% \begin{syntax}
+%   "\if_true:" <true code> "\else:" <false code> "\fi:" \\
+%   "\if_false:" <true code> "\else:" <false code> "\fi:" \\
+%   "\reverse_if:N" <primitive conditional>
+% \end{syntax}
+% "\if_true:" always executes <true code>, while "\if_false:" always
+% executes <false code>. "\reverse_if:N" reverses any two-way primitive
+% conditional. "\else:" and "\fi:" delimit the branches of the
+% conditional. "\or:" is used in case switches, see \pkg{l3intexpr}
+% for more.
+% \begin{texnote}
+% These are equivalent to their corresponding \TeX\ primitive 
+% conditionals; |\reverse_if:N| is \eTeX's |\unless|.
+% \end{texnote}
+% \end{function}
+% 
+% \begin{function}[EXP]{\if_meaning:w}
+% \begin{syntax}
+%   "\if_meaning:w" <arg1> <arg2> <true code> "\else:" <false code> "\fi:"
+% \end{syntax}
+% "\if_meaning:w" executes <true code> when <arg1> and <arg2> are the same, 
+% otherwise it executes <false code>. 
+% <arg1> and <arg2> could be functions, variables, tokens; in all cases the
+% \emph{unexpanded} definitions are compared.
+% \begin{texnote}
+% This is \TeX's |\ifx|.
+% \end{texnote}
+% \end{function}
+%
+% \begin{function}[EXP]{\if:w, \if_charcode:w,\if_catcode:w}
+% \begin{syntax}
+%   "\if:w" <token1> <token2> <true code> "\else:" <false code> "\fi:" \\
+%   "\if_catcode:w" <token1> <token2> <true code> "\else:" <false
+%   code> "\fi:"
+% \end{syntax}
+% These conditionals will expand any following tokens until two
+% unexpandable tokens are left. If you wish to prevent this expansion,
+% prefix the token in question with "\exp_not:N". "\if_catcode:w"
+% tests if the category codes of the two tokens are the same whereas
+% "\if:w" tests if the character codes are
+% identical. "\if_charcode:w" is an alternative name for "\if:w".
+% \end{function}
+%
+% \begin{function}[EXP]{\if_predicate:w}
+% \begin{syntax}
+%  "\if_predicate:w" <predicate> <true code> "\else:" <false code> "\fi:"
+% \end{syntax}
+% This function takes a predicate function and
+% branches according to the result.  (In practice this function would also
+% accept a single boolean variable in place of the <predicate> but to make the
+% coding clearer this should be done through "\if_bool:N".)
+% \end{function}
+%
+% \begin{function}[EXP]{\if_bool:N}
+%   \begin{syntax}
+%     "\if_bool:N" <boolean> <true code> "\else:" <false code> "\fi:"
+%   \end{syntax}
+%   This function takes a boolean variable and
+%   branches according to the result.
+% \end{function}
+%
+% \begin{function}[EXP]{\if_cs_exist:N, \if_cs_exist:w}
+%   \begin{syntax}
+%     "\if_cs_exist:N" <cs> <true code> "\else:" <false code> "\fi:" \\
+%     "\if_cs_exist:w" <tokens> "\cs_end:" <true code> "\else:" <false
+%     code> "\fi:"
+%   \end{syntax}
+%   Check if <cs> appears in the hash table or if the control sequence
+%   that can be formed from <tokens> appears in the hash table. The
+%   latter function does not turn the control sequence in question into
+%   "\scan_stop:"! This can be useful when dealing with control
+%   sequences which cannot be entered as a single token.
+% \end{function}
+%
+% \begin{function}[EXP]
+%   {
+%     \if_mode_horizontal:, \if_mode_vertical:,
+%     \if_mode_math:, \if_mode_inner:
+%   }
+%   \begin{syntax}
+%     "\if_mode_horizontal:" <true code> "\else:" <false code> "\fi:"
+%   \end{syntax}
+%   Execute <true code> if currently in horizontal mode, otherwise
+%   execute <false code>. Similar for the other functions.
+% \end{function}
+% 
+% \section{Internal kernel functions}
+% 
+% \begin{function}{\chk_if_exist_cs:N, \chk_if_exist_cs:c}
+%   \begin{syntax}
+%     \cs{chk_if_exist_cs:N} \meta{cs}
+%   \end{syntax}
+%   This function checks that \meta{cs} exists according to the
+%   criteria for \cs{cs_if_exist_p:N}, and if not raises a kernel-level
+%   error.
+% \end{function}
+%
+% \begin{function}{\chk_if_free_cs:N, \chk_if_free_cs:c}
+%   \begin{syntax}
+%     \cs{chk_if_free_cs:N} \meta{cs}
+%   \end{syntax}
+%   This function checks that \meta{cs} is free according to the
+%   criteria for \cs{cs_if_free_p:N}, and if not raises a kernel-level
+%   error.
+% \end{function}
+% 
+% \begin{function}{\pref_global:D, \pref_long:D, \pref_protected:D}
+%   \begin{syntax}
+%     \cs{pref_global:D} \cs{cs_set_nopar:Npn}
+%   \end{syntax}
+%   Prefix functions that can be used in front of some definition
+%   functions (namely \ldots). The result of prefixing a function
+%   definition with \cs{pref_global:D} makes the definition global,
+%   \cs{pref_long:D} change the argument scanning mechanism so that it
+%   allows \cs{par} tokens in the argument of the prefixed function,
+%   and \cs{pref_protected:D} makes the definition robust inside
+%   \texttt{x}-type expansions.
+%
+%   None of these internal functions should be used by a programmer since
+%   the necessary combinations are all available as separate function,
+%   \emph{e.g.} \cs{cs_set:Npn} is internally implemented as \cs{pref_long:D}
+%   \c{cs_set_nopar:Npn}.
+%   \begin{texnote}
+%     These prefixes are the primitives \tn{global}, \tn{long}, and
+%     \cs{protected}.
+%   \end{texnote}
+% \end{function}
+%
+% \end{documentation}
+%
+% \begin{implementation}
+%
+% \section{\pkg{l3basics} implementation}
+%
+%    \begin{macrocode}
+%<*initex|package>
+%    \end{macrocode}
+%
+%    \begin{macrocode}
+%<*package>
+\ProvidesExplPackage
+  {\ExplFileName}{\ExplFileDate}{\ExplFileVersion}{\ExplFileDescription}
+\package_check_loaded_expl:
+%</package>
+%    \end{macrocode}
+%
+% \subsection{Renaming some \TeX{} primitives (again)}
+% 
+% Having given all the \TeX{} primitives a consistent name, we need to
+% give sensible names to the ones we actually want to use.
+% These will be defined as needed in the appropriate modules, but
+% do a few now, just to get started.\footnote{This renaming gets expensive
+% in terms of csname usage, an alternative scheme would be to just use
+% the \cs{tex\ldots:D} name in the cases where no good alternative exists.}
+%
+% \begin{macro}{\cs_set_eq:NwN}
+%   A pretty basic requirement: \cs{let} one control sequence to another.
+%    \begin{macrocode}>
+\tex_let:D \cs_set_eq:NwN            \tex_let:D
+%    \end{macrocode}
+% \end{macro}
+% 
+% \begin{macro}{\if_true:}
+% \begin{macro}{\if_false:}
+% \begin{macro}{\or:}
+% \begin{macro}{\else:}
+% \begin{macro}{\fi:}
+% \begin{macro}{\reverse_if:N}
+% \begin{macro}{\if:w}
+% \begin{macro}{\if_bool:N}
+% \begin{macro}{\if_predicate:w}
+% \begin{macro}{\if_charcode:w}
+% \begin{macro}{\if_catcode:w}
+% Then some conditionals.
+%    \begin{macrocode}
+\cs_set_eq:NwN \if_true:           \tex_iftrue:D
+\cs_set_eq:NwN \if_false:          \tex_iffalse:D
+\cs_set_eq:NwN \or:                \tex_or:D
+\cs_set_eq:NwN \else:              \tex_else:D
+\cs_set_eq:NwN \fi:                \tex_fi:D
+\cs_set_eq:NwN \reverse_if:N       \etex_unless:D
+\cs_set_eq:NwN \if:w               \tex_if:D
+\cs_set_eq:NwN \if_bool:N          \tex_ifodd:D
+\cs_set_eq:NwN \if_predicate:w     \tex_ifodd:D
+\cs_set_eq:NwN \if_charcode:w      \tex_if:D
+\cs_set_eq:NwN \if_catcode:w       \tex_ifcat:D
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\if_meaning:w}
+%    \begin{macrocode}
+\cs_set_eq:NwN \if_meaning:w       \tex_ifx:D
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\if_mode_math:}
+% \begin{macro}{\if_mode_horizontal:}
+% \begin{macro}{\if_mode_vertical:}
+% \begin{macro}{\if_mode_inner:}
+%   \TeX{} lets us detect some if its modes.
+%    \begin{macrocode}
+\cs_set_eq:NwN \if_mode_math:       \tex_ifmmode:D
+\cs_set_eq:NwN \if_mode_horizontal: \tex_ifhmode:D
+\cs_set_eq:NwN \if_mode_vertical:   \tex_ifvmode:D
+\cs_set_eq:NwN \if_mode_inner:      \tex_ifinner:D
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% 
+% \begin{macro}{\if_cs_exist:N}
+% \begin{macro}{\if_cs_exist:w}
+%    \begin{macrocode}
+\cs_set_eq:NwN \if_cs_exist:N      \etex_ifdefined:D
+\cs_set_eq:NwN \if_cs_exist:w      \etex_ifcsname:D
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\exp_after:wN}
+% \begin{macro}{\exp_not:N}
+% \begin{macro}{\exp_not:n}
+%    The three |\exp_| functions are used in the \textsf{l3expan} module
+%    where they are described.
+%    \begin{macrocode}
+\cs_set_eq:NwN \exp_after:wN       \tex_expandafter:D
+\cs_set_eq:NwN \exp_not:N          \tex_noexpand:D
+\cs_set_eq:NwN \exp_not:n          \etex_unexpanded:D
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% 
+% \begin{macro}{\token_to_meaning:N}
+% \begin{macro}{\token_to_str:N}
+% \begin{macro}{\cs:w}
+% \begin{macro}{\cs_end:}
+% \begin{macro}{\cs_meaning:N}
+% \begin{macro}{\cs_show:N}
+%    \begin{macrocode}
+\cs_set_eq:NwN \token_to_meaning:N \tex_meaning:D
+\cs_set_eq:NwN \token_to_str:N     \tex_string:D
+\cs_set_eq:NwN \cs:w               \tex_csname:D
+\cs_set_eq:NwN \cs_end:            \tex_endcsname:D
+\cs_set_eq:NwN \cs_meaning:N       \tex_meaning:D
+\cs_set_eq:NwN \cs_show:N          \tex_show:D
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\scan_stop:}
+% \begin{macro}{\group_begin:}
+% \begin{macro}{\group_end:}
+%    The next three are basic functions for which there also exist
+%    versions that are safe inside alignments. These safe versions are
+%    defined in the \textsf{l3prg} module.
+%    \begin{macrocode}
+\cs_set_eq:NwN \scan_stop:         \tex_relax:D
+\cs_set_eq:NwN \group_begin:       \tex_begingroup:D
+\cs_set_eq:NwN \group_end:         \tex_endgroup:D
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% 
+% \begin{macro}{\if_int_compare:w}
+% \begin{macro}{\int_to_roman:w}
+%    \begin{macrocode}
+\cs_set_eq:NwN \if_int_compare:w   \tex_ifnum:D  
+\cs_set_eq:NwN \int_to_roman:w     \tex_romannumeral:D  
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% 
+% \begin{macro}{\group_insert_after:N}
+%    \begin{macrocode}
+\cs_set_eq:NwN \group_insert_after:N \tex_aftergroup:D      
+%    \end{macrocode}
+% \end{macro}
+% 
+% \begin{macro}{\pref_global:D}
+% \begin{macro}{\pref_long:D}
+% \begin{macro}{\pref_protected:D}
+%    \begin{macrocode}
+\cs_set_eq:NwN \pref_global:D      \tex_global:D
+\cs_set_eq:NwN \pref_long:D        \tex_long:D
+\cs_set_eq:NwN \pref_protected:D   \etex_protected:D
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% 
+% \begin{macro}{\exp_args:Nc}
+%   Discussed in \pkg{l3expan}, but needed much earlier.
+%    \begin{macrocode}
+\tex_long:D \tex_def:D \exp_args:Nc #1#2 { \exp_after:wN #1 \cs:w #2 \cs_end: }
+%    \end{macrocode}
+% \end{macro}
+% 
+% \begin{macro}{\token_to_str:c}
+% \begin{macro}{\cs_meaning:c}
+% \begin{macro}{\cs_show:c}
+%%  A small number of variants by hand.
+%    \begin{macrocode}
+\tex_def:D \cs_meaning:c   { \exp_args:Nc \cs_meaning:N }
+\tex_def:D \token_to_str:c { \exp_args:Nc \token_to_str:N }
+\tex_def:D \cs_show:c      { \exp_args:Nc \cs_show:N }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \subsection {Defining functions}
+%
+% We start by providing functions for the typical definition
+% functions. First the local ones.
+%
+% \begin{macro}{\cs_set_nopar:Npn}
+% \begin{macro}{\cs_set_nopar:Npx}
+% \begin{macro}{\cs_set:Npn}
+% \begin{macro}{\cs_set:Npx}
+% \begin{macro}{\cs_set_protected_nopar:Npn}
+% \begin{macro}{\cs_set_protected_nopar:Npx}
+% \begin{macro}{\cs_set_protected:Npn}
+% \begin{macro}{\cs_set_protected:Npx}
+%   All assignment functions in \LaTeX3 should be naturally robust;
+%   after all, the \TeX{} primitives for assignments are and it can be
+%   a cause of problems if others aren't.
+%    \begin{macrocode}
+\cs_set_eq:NwN \cs_set_nopar:Npn            \tex_def:D
+\cs_set_eq:NwN \cs_set_nopar:Npx            \tex_edef:D
+\pref_protected:D \cs_set_nopar:Npn \cs_set:Npn
+  { \pref_long:D \cs_set_nopar:Npn }
+\pref_protected:D \cs_set_nopar:Npn \cs_set:Npx
+  { \pref_long:D \cs_set_nopar:Npx }
+\pref_protected:D \cs_set_nopar:Npn \cs_set_protected_nopar:Npn
+  { \pref_protected:D \cs_set_nopar:Npn }
+\pref_protected:D \cs_set_nopar:Npn \cs_set_protected_nopar:Npx
+  { \pref_protected:D \cs_set_nopar:Npx }
+\cs_set_protected_nopar:Npn \cs_set_protected:Npn
+  { \pref_protected:D \pref_long:D \cs_set_nopar:Npn }
+\cs_set_protected_nopar:Npn \cs_set_protected:Npx
+  { \pref_protected:D \pref_long:D \cs_set_nopar:Npx }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\cs_gset_nopar:Npn}
+% \begin{macro}{\cs_gset_nopar:Npx}
+% \begin{macro}{\cs_gset:Npn}
+% \begin{macro}{\cs_gset:Npx}
+% \begin{macro}{\cs_gset_protected_nopar:Npn}
+% \begin{macro}{\cs_gset_protected_nopar:Npx}
+% \begin{macro}{\cs_gset_protected:Npn}
+% \begin{macro}{\cs_gset_protected:Npx}
+%   Global versions of the above functions.
+%    \begin{macrocode}
+\cs_set_eq:NwN \cs_gset_nopar:Npn           \tex_gdef:D
+\cs_set_eq:NwN \cs_gset_nopar:Npx           \tex_xdef:D
+\cs_set_protected_nopar:Npn \cs_gset:Npn
+  { \pref_long:D \cs_gset_nopar:Npn }
+\cs_set_protected_nopar:Npn \cs_gset:Npx
+  { \pref_long:D \cs_gset_nopar:Npx }
+\cs_set_protected_nopar:Npn \cs_gset_protected_nopar:Npn
+  { \pref_protected:D \cs_gset_nopar:Npn }
+\cs_set_protected_nopar:Npn \cs_gset_protected_nopar:Npx
+  { \pref_protected:D \cs_gset_nopar:Npx }
+\cs_set_protected_nopar:Npn \cs_gset_protected:Npn
+  { \pref_protected:D \pref_long:D \cs_gset_nopar:Npn }
+\cs_set_protected_nopar:Npn \cs_gset_protected:Npx
+  { \pref_protected:D \pref_long:D \cs_gset_nopar:Npx }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \subsection{Selecting tokens}
+%
+% \begin{macro}{\use:c}
+%    This macro grabs its argument and returns a csname from it.
+%    \begin{macrocode}
+\cs_set:Npn \use:c #1 { \cs:w #1 \cs_end: }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\use:x}
+% \begin{macro}[aux]{\cs_tmp:w}
+% Fully expands its argument and passes it to the input stream.
+% Uses |\cs_tmp:| as a scratch register but does not affect it.
+%    \begin{macrocode}
+\cs_set_protected:Npn \use:x #1
+  {
+    \group_begin:
+      \cs_set:Npx \cs_tmp:w {#1}
+      \exp_after:wN
+    \group_end:
+    \cs_tmp:w
+  }
+\cs_set:Npn \cs_tmp:w { }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\use:n}
+% \begin{macro}{\use:nn}
+% \begin{macro}{\use:nnn}
+% \begin{macro}{\use:nnnn}
+%    These macro grabs its arguments and returns it back to the input
+%    (with outer braces removed).
+%    \begin{macrocode}
+\cs_set:Npn \use:n    #1       {#1}
+\cs_set:Npn \use:nn   #1#2     {#1#2}
+\cs_set:Npn \use:nnn  #1#2#3   {#1#2#3}
+\cs_set:Npn \use:nnnn #1#2#3#4 {#1#2#3#4}
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\use_i:nn}
+% \begin{macro}{\use_ii:nn}
+%    The equivalent to \LaTeXe{}'s \cs{@firstoftwo} and \cs{@secondoftwo}.
+%    \begin{macrocode}
+\cs_set:Npn \use_i:nn  #1#2 {#1}
+\cs_set:Npn \use_ii:nn #1#2 {#2}
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\use_i:nnn}
+% \begin{macro}{\use_ii:nnn}
+% \begin{macro}{\use_iii:nnn}
+% \begin{macro}{\use_i_ii:nnn}
+% \begin{macro}{\use_i:nnnn}
+% \begin{macro}{\use_ii:nnnn}
+% \begin{macro}{\use_iii:nnnn}
+% \begin{macro}{\use_iv:nnnn}
+%    We also need something for picking up arguments from a longer
+%    list.
+%    \begin{macrocode}
+\cs_set:Npn \use_i:nnn    #1#2#3 {#1}
+\cs_set:Npn \use_ii:nnn   #1#2#3 {#2}
+\cs_set:Npn \use_iii:nnn  #1#2#3 {#3}
+\cs_set:Npn \use_i_ii:nnn #1#2#3 {#1#2}
+\cs_set:Npn \use_i:nnnn   #1#2#3#4 {#1}
+\cs_set:Npn \use_ii:nnnn  #1#2#3#4 {#2}
+\cs_set:Npn \use_iii:nnnn #1#2#3#4 {#3}
+\cs_set:Npn \use_iv:nnnn  #1#2#3#4 {#4}
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\use_none_delimit_by_q_nil:w}
+% \begin{macro}{\use_none_delimit_by_q_stop:w}
+% \begin{macro}{\use_none_delimit_by_q_recursion_stop:w}
+%   Functions that gobble everything until they see either \cs{q_nil} or
+%   \cs{q_stop}, respectively.
+%    \begin{macrocode}
+\cs_set:Npn \use_none_delimit_by_q_nil:w  #1 \q_nil  { }
+\cs_set:Npn \use_none_delimit_by_q_stop:w #1 \q_stop { }
+\cs_set:Npn \use_none_delimit_by_q_recursion_stop:w #1 \q_recursion_stop { }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\use_i_delimit_by_q_nil:nw}
+% \begin{macro}{\use_i_delimit_by_q_stop:nw}
+% \begin{macro}{\use_i_delimit_by_q_recursion_stop:nw}
+%   Same as above but execute first argument after gobbling. Very
+%   useful when you need to skip the rest of a mapping sequence but
+%   want an easy way to control what should be expanded next.
+%    \begin{macrocode}
+\cs_set:Npn \use_i_delimit_by_q_nil:nw  #1#2 \q_nil  {#1}
+\cs_set:Npn \use_i_delimit_by_q_stop:nw #1#2 \q_stop {#1}
+\cs_set:Npn \use_i_delimit_by_q_recursion_stop:nw #1#2 \q_recursion_stop {#1}
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\use_i_after_fi:nw}
+% \begin{macro}{\use_i_after_else:nw}
+% \begin{macro}{\use_i_after_or:nw}
+% \begin{macro}{\use_i_after_orelse:nw}
+%   Returns the first argument after ending the conditional.
+%    \begin{macrocode}
+\cs_set:Npn \use_i_after_fi:nw #1 \fi: { \fi: #1 }
+\cs_set:Npn \use_i_after_else:nw #1 \else: #2 \fi: { \fi: #1 }
+\cs_set:Npn \use_i_after_or:nw #1 \or: #2 \fi: { \fi: #1 }
+\cs_set:Npn \use_i_after_orelse:nw #1#2#3 \fi: { \fi: #1 }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \subsection{Gobbling tokens from input}
+%
+% \begin{macro}{\use_none:n}
+% \begin{macro}{\use_none:nn}
+% \begin{macro}{\use_none:nnn}
+% \begin{macro}{\use_none:nnnn}
+% \begin{macro}{\use_none:nnnnn}
+% \begin{macro}{\use_none:nnnnnn}
+% \begin{macro}{\use_none:nnnnnnn}
+% \begin{macro}{\use_none:nnnnnnnn}
+% \begin{macro}{\use_none:nnnnnnnnn}
+%   To gobble tokens from the input we use a standard naming
+%   convention: the number of tokens gobbled is given by the number of
+%   |n|'s following the |:| in the name. Although defining
+%   |\use_none:nnn| and above as separate calls of |\use_none:n| and
+%   |\use_none:nn| is slightly faster, this is very non-intuitive to
+%   the programmer who will assume that expanding such a function once
+%   will take care of gobbling all the tokens in one go.
+%    \begin{macrocode}
+\cs_set:Npn \use_none:n         #1                 { }
+\cs_set:Npn \use_none:nn        #1#2               { }
+\cs_set:Npn \use_none:nnn       #1#2#3             { }
+\cs_set:Npn \use_none:nnnn      #1#2#3#4           { }
+\cs_set:Npn \use_none:nnnnn     #1#2#3#4#5         { }
+\cs_set:Npn \use_none:nnnnnn    #1#2#3#4#5#6       { }
+\cs_set:Npn \use_none:nnnnnnn   #1#2#3#4#5#6#7     { }
+\cs_set:Npn \use_none:nnnnnnnn  #1#2#3#4#5#6#7#8   { }
+\cs_set:Npn \use_none:nnnnnnnnn #1#2#3#4#5#6#7#8#9 { }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \subsection{Conditional processing and definitions}
+%
+% Underneath any predicate function (|_p|) or other conditional forms
+% (|TF|, etc.) is a built-in logic saying that it after all of the
+% testing and processing must return the \meta{state} this leaves
+% \TeX{} in. Therefore, a simple user interface could be something like
+% \begin{verbatim}
+%   \if_meaning:w #1#2   \prg_return_true:  \else: 
+%      \if_meaning:w #1#3  \prg_return_true: \else:
+%      \prg_return_false: 
+%  \fi: \fi:
+% \end{verbatim}
+% Usually, a \TeX{} programmer would have to insert a number of
+% |\exp_after:wN|s to ensure the state value is returned at exactly
+% the point where the last conditional is finished.  However, that
+% obscures the code and forces the \TeX{} programmer to prove that
+% he/she knows the $2^{n}-1$ table.  We therefore provide the simpler
+% interface.
+%
+% \begin{macro}{\prg_return_true:}
+% \begin{macro}{\prg_return_false:}
+%   The idea here is that \cs{int_to_roman:w} will expand fully any
+%   \cs{else:} and the \cs{fi:} that are waiting to be discarded,
+%   before reaching the \cs{c_zero} which will leave the expansion null.
+%   The code can then leave either the first or second argument in the
+%   input stream. This means that all of the branching code has to contain
+%   at least two tokens: see how the logical tests are actually implemented
+%   to see this.
+%    \begin{macrocode}
+\cs_set_nopar:Npn \prg_return_true:
+  { \exp_after:wN \use_i:nn  \int_to_roman:w }
+\cs_set_nopar:Npn \prg_return_false:
+  { \exp_after:wN \use_ii:nn \int_to_roman:w}
+%    \end{macrocode}
+%   An extended state space could be implemented by including a more
+%   elaborate function in place of \cs{use_i:nn}/\cs{use_ii:nn}. Provided
+%   two arguments are absorbed then the code will work.
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\prg_set_conditional:Npnn,\prg_new_conditional:Npnn,
+%     \prg_set_protected_conditional:Npnn,\prg_new_protected_conditional:Npnn}
+%   The user functions for the types using parameter text from the
+%   programmer. Call aux function to grab parameters, split the base
+%   function into name and signature and then use, \emph{e.g.}, |\cs_set:Npn|
+%   to define it with.
+%    \begin{macrocode}
+\cs_set_protected:Npn \prg_set_conditional:Npnn #1
+  {
+    \prg_get_parm_aux:nw
+      {
+        \cs_split_function:NN #1 \prg_generate_conditional_aux:nnNNnnnn 
+        \cs_set:Npn { parm }
+      }
+  }
+\cs_set_protected:Npn \prg_new_conditional:Npnn #1
+  {
+    \prg_get_parm_aux:nw
+      {
+        \cs_split_function:NN #1 \prg_generate_conditional_aux:nnNNnnnn 
+        \cs_new:Npn { parm }
+      }
+  }
+\cs_set_protected:Npn \prg_set_protected_conditional:Npnn #1
+  {
+    \prg_get_parm_aux:nw{
+      \cs_split_function:NN #1 \prg_generate_conditional_aux:nnNNnnnn 
+      \cs_set_protected:Npn { parm }
+    }
+  }
+\cs_set_protected:Npn \prg_new_protected_conditional:Npnn #1
+  {
+    \prg_get_parm_aux:nw
+      {
+        \cs_split_function:NN #1 \prg_generate_conditional_aux:nnNNnnnn
+        \cs_new_protected:Npn { parm }
+     }
+  }
+%    \end{macrocode}
+% \end{macro} 
+%
+% \begin{macro}{\prg_set_conditional:Nnn,\prg_new_conditional:Nnn,
+%     \prg_set_protected_conditional:Nnn,\prg_new_protected_conditional:Nnn}
+%   The user functions for the types automatically inserting the
+%   correct parameter text based on the signature. Call aux function
+%   after calculating number of arguments, split the base function
+%   into name and signature and then use, \emph{e.g.}, |\cs_set:Npn| to
+%   define it with.
+%    \begin{macrocode}
+\cs_set_protected:Npn \prg_set_conditional:Nnn #1
+  {
+    \exp_args:Nnf \prg_get_count_aux:nn
+      {
+        \cs_split_function:NN #1 \prg_generate_conditional_aux:nnNNnnnn
+        \cs_set:Npn { count }
+      }
+      { \cs_get_arg_count_from_signature:N #1 }
+  }
+\cs_set_protected:Npn \prg_new_conditional:Nnn #1
+  {
+    \exp_args:Nnf \prg_get_count_aux:nn
+      {
+        \cs_split_function:NN #1 \prg_generate_conditional_aux:nnNNnnnn
+        \cs_new:Npn { count}
+      }
+      { \cs_get_arg_count_from_signature:N #1 }
+  }
+
+\cs_set_protected:Npn \prg_set_protected_conditional:Nnn #1{
+  \exp_args:Nnf \prg_get_count_aux:nn{
+    \cs_split_function:NN #1 \prg_generate_conditional_aux:nnNNnnnn
+    \cs_set_protected:Npn {count}
+  }{\cs_get_arg_count_from_signature:N #1}
+}
+
+\cs_set_protected:Npn \prg_new_protected_conditional:Nnn #1
+  {
+    \exp_args:Nnf \prg_get_count_aux:nn
+      {
+        \cs_split_function:NN #1 \prg_generate_conditional_aux:nnNNnnnn
+        \cs_new_protected:Npn {count}
+      }
+      { \cs_get_arg_count_from_signature:N #1 }
+  }
+%    \end{macrocode}
+% \end{macro}
+% 
+% \begin{macro}{\prg_set_eq_conditional:NNn,\prg_new_eq_conditional:NNn}
+%    The obvious setting-equal functions.
+%    \begin{macrocode}
+\cs_set_protected:Npn \prg_set_eq_conditional:NNn #1#2#3
+  { \prg_set_eq_conditional_aux:NNNn \cs_set_eq:cc #1#2 {#3} }
+\cs_set_protected:Npn \prg_new_eq_conditional:NNn #1#2#3
+  { \prg_set_eq_conditional_aux:NNNn \cs_new_eq:cc #1#2 {#3} }
+%    \end{macrocode}
+%    \end{macro}
+%
+% \begin{macro}[aux]{\prg_get_parm_aux:nw,\prg_get_count_aux:nn}
+%   For the |Npnn| type we must grab the parameter text before
+%   continuing. We make this a very generic function that takes one
+%   argument before reading everything up to a left brace. Something
+%   similar for the |Nnn| type.
+%    \begin{macrocode}
+\cs_set:Npn \prg_get_count_aux:nn #1#2 { #1 {#2} }
+\cs_set:Npn \prg_get_parm_aux:nw #1#2# { #1 {#2} }
+%    \end{macrocode}
+% \end{macro}
+% \begin{macro}[aux]{\prg_generate_conditional_parm_aux:nnNNnnnn,
+%     \prg_generate_conditional_parm_aux:nw}
+%   The workhorse here is going through a list of desired forms, \emph{i.e.},
+%   |p|, |TF|, |T| and |F|. The first three arguments come from splitting up
+%   the base form of the conditional, which gives the name, signature
+%   and a boolean to signal whether or not there was a colon in the
+%   name. For the time being, we do not use this piece of information
+%   but could well throw an error. The fourth argument is how to
+%   define this function, the fifth is the text |parm| or |count| for
+%   which version to use to define the functions, the sixth is the
+%   parameters to use (possibly empty) or number of arguments, the
+%   seventh is the list of forms to define, the eight is the
+%   replacement text which we will augment when defining the forms.
+%    \begin{macrocode}
+\cs_set_protected:Npn \prg_generate_conditional_aux:nnNNnnnn #1#2#3#4#5#6#7#8
+  {
+    \prg_generate_conditional_aux:nnw {#5}
+      {
+        #4 {#1} {#2} {#6} {#8}
+      }
+      #7 , ? , \q_recursion_stop
+  }
+%    \end{macrocode}
+%   Looping through the list of desired forms. First is the text |parm|
+%   or |count|, second is five arguments packed together and third is
+%   the form. Use text and form to call the correct type.
+%    \begin{macrocode}
+\cs_set_protected:Npn \prg_generate_conditional_aux:nnw #1#2#3 ,
+  {
+    \if:w ?#3
+      \exp_after:wN \use_none_delimit_by_q_recursion_stop:w
+    \fi:
+    \use:c { prg_generate_#3_form_#1:Nnnnn } #2
+    \prg_generate_conditional_aux:nnw {#1} {#2}
+  }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}[aux]
+%   {
+%     \prg_generate_p_form_parm:Nnnnn,
+%     \prg_generate_TF_form_parm:Nnnnn,
+%     \prg_generate_T_form_parm:Nnnnn,
+%     \prg_generate_F_form_parm:Nnnnn
+%   }
+%   How to generate the various forms. The |parm| types here takes the
+%   following arguments: 1: how to define (an N-type), 2: name, 3:
+%   signature, 4: parameter text (or empty), 5: replacement. Remember that
+%   the logic-returning functions expect two arguments to be present after
+%   \cs{c_zero}: notice the construction  of the different variants
+%   relies on this, and that the |TF| variant will be slightly faster
+%   than the |T| version.
+%    \begin{macrocode}
+\cs_set_protected:Npn \prg_generate_p_form_parm:Nnnnn #1#2#3#4#5
+  {
+    \exp_args:Nc #1 { #2 _p: #3 } #4
+      {
+        #5 \c_zero 
+        \c_true_bool \c_false_bool
+      }
+  }
+\cs_set_protected:Npn \prg_generate_T_form_parm:Nnnnn #1#2#3#4#5
+  {
+    \exp_args:Nc #1 { #2 : #3 T } #4
+      {
+        #5 \c_zero
+        \use:n \use_none:n
+      }
+  }
+\cs_set_protected:Npn \prg_generate_F_form_parm:Nnnnn #1#2#3#4#5
+  {
+    \exp_args:Nc #1 { #2 : #3 F } #4
+      {
+        #5 \c_zero
+        { }
+      }
+  }
+\cs_set_protected:Npn \prg_generate_TF_form_parm:Nnnnn #1#2#3#4#5
+  {
+    \exp_args:Nc #1 { #2 : #3 TF } #4
+      { #5 \c_zero }
+  }
+%    \end{macrocode}
+% \end{macro}
+% \begin{macro}[aux]
+%   {
+%     \prg_generate_p_form_count:Nnnnn,
+%     \prg_generate_TF_form_count:Nnnnn,
+%     \prg_generate_T_form_count:Nnnnn,
+%     \prg_generate_F_form_count:Nnnnn
+%   }
+%   The |count| form is similar, but of course requires a number rather
+%   than a primitive argument specification.
+%    \begin{macrocode}
+\cs_set_protected:Npn \prg_generate_p_form_count:Nnnnn #1#2#3#4#5
+  {
+    \cs_generate_from_arg_count:cNnn { #2 _p: #3 } #1 {#4}
+      {
+        #5 \c_zero 
+        \c_true_bool \c_false_bool
+      }
+  }
+\cs_set_protected:Npn \prg_generate_T_form_count:Nnnnn #1#2#3#4#5
+  {
+    \cs_generate_from_arg_count:cNnn { #2 : #3 T } #1 {#4}
+      {
+        #5 \c_zero
+        \use:n \use_none:n
+      }
+  }
+\cs_set_protected:Npn \prg_generate_F_form_count:Nnnnn #1#2#3#4#5
+  {
+    \cs_generate_from_arg_count:cNnn { #2 : #3 F } #1 {#4}
+      {
+        #5 \c_zero
+        { }
+      }
+  }
+\cs_set_protected:Npn \prg_generate_TF_form_count:Nnnnn #1#2#3#4#5
+  {
+    \cs_generate_from_arg_count:cNnn { #2 : #3 TF } #1 {#4}
+      { #5 \c_zero }
+  }
+%    \end{macrocode}
+% \end{macro}
+% 
+% \begin{macro}[aux]{\prg_set_eq_conditional_aux:NNNn,
+%                    \prg_set_eq_conditional_aux:NNNw}
+%    \begin{macrocode}
+\cs_set_protected:Npn \prg_set_eq_conditional_aux:NNNn #1#2#3#4
+ { \prg_set_eq_conditional_aux:NNNw #1#2#3#4 , ? , \q_recursion_stop }
+%    \end{macrocode}
+% Manual clist loop over argument |#4|.
+%    \begin{macrocode}
+\cs_set_protected:Npn \prg_set_eq_conditional_aux:NNNw #1#2#3#4 ,
+  {
+    \if:w ? #4 \scan_stop:
+      \exp_after:wN \use_none_delimit_by_q_recursion_stop:w
+    \fi:
+    #1
+    { \exp_args:NNc \cs_split_function:NN #2 { prg_conditional_form_#4:nnn } }
+    { \exp_args:NNc \cs_split_function:NN #3 { prg_conditional_form_#4:nnn } }
+    \prg_set_eq_conditional_aux:NNNw #1 {#2} {#3}
+  }
+%    \end{macrocode}
+%    \begin{macrocode}
+\cs_set:Npn \prg_conditional_form_p:nnn  #1#2#3 { #1 _p : #2 }
+\cs_set:Npn \prg_conditional_form_TF:nnn #1#2#3 { #1    : #2 TF }
+\cs_set:Npn \prg_conditional_form_T:nnn  #1#2#3 { #1    : #2 T }
+\cs_set:Npn \prg_conditional_form_F:nnn  #1#2#3 { #1    : #2 F }
+%    \end{macrocode}
+% \end{macro}
+%
+% All that is left is to define the canonical boolean true and false.
+% I think Michael originated the idea of expandable boolean tests.  At
+% first these were supposed to expand into either \texttt{TT} or
+% \texttt{TF} to be tested using |\if:w| but this was later changed to
+% |00| and |01|, so they could be used in logical
+% operations. Later again they were changed to being numerical
+% constants with values of $1$ for true and $0$ for false. We need
+% this from the get-go.
+%
+% \begin{macro}{\c_true_bool}
+% \begin{macro}{\c_false_bool}
+%    Here are the canonical boolean values.
+%    \begin{macrocode}
+\tex_chardef:D \c_true_bool  = 1~
+\tex_chardef:D \c_false_bool = 0~
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \subsection{Dissecting a control sequence}
+%
+% \begin{macro}{\cs_to_str:N}
+% \begin{macro}[aux]{\cs_to_str_aux:w}
+%   This converts a control sequence into the character string of its
+%   name, removing the leading escape character. This turns out to be
+%   a non-trivial matter as there a different cases:
+%   \begin{itemize}
+%     \item The usual case of a printable escape character;
+%     \item the case of a non-printable escape characters, e.g., when
+%     the value of |\tex_escapechar:D| is negative;
+%     \item when the escape character is a space.
+%   \end{itemize}
+%   One approach to solve this is to test how many tokens result from
+%   |\token_to_str:N \a|. If there are two tokens, then the escape
+%   character is printable, while if it is non-printable then only
+%   one is present.
+% 
+%   However, there is an additional complication: the control
+%   sequence itself may start with a space. Clearly that should \emph{not} be
+%   lost in the process of converting to a string. So the approach adopted is
+%   a little more intricate still. When the escape character is printable,
+%   \verb*|\token_to_str:N \ | yields the escape character itself and a space.
+%   The escape sequence will terminate the expansion started by
+%   \cs{int_to_roman:w}, which is a negative number and so will not
+%   gobble the escape character even if it's a number. The \cs{tex_if:D}
+%   test will then be \texttt{false}, and the na\"ive approach of gobbling
+%   the first character of the \cs{token_to_str:N} version of the control
+%   sequence will work, even if the first character is a space.
+%   The second case is that the escape character is itself a space. In this
+%   case, the escape character space is consumed terminating the first
+%   \cs{int_to_roman:w}, and \cs{cs_to_str_aux:w} is expanded. This
+%   inserts a space, making the \cs{if:w} test \texttt{true}. The
+%   second \cs{int_to_roman:w} will then execute the \cs{token_to_str:N},
+%   with the escape-character space being consumed by the
+%   \cs{int_to_roman:w}, and thus leaving the control sequence name in the
+%   input stream. The final case is where the escape character is not
+%   printable.
+%   The flow here starts with the \verb*|\token_to_str:N \ |  giving just a
+%   space, which terminates the first \cs{int_to_roman:w} but leaves no
+%   token for the \cs{if:w} test. This means that the \cs{int_to_roman:w}
+%   is executed before the test is finished. The result is that the
+%   \cs{fi:}, expanded before the \cs{tex_if:D} is finished, becomes
+%   \cs{scan_stop:} \cs{fi:}, and the \cs{scan_stop:} is then used in
+%   the \cs{if:w} test.
+%   In this case, \cs{token_to_str:N} is therefore used with no gobbling at
+%   all, which is exactly what is needed in this case.
+%    \begin{macrocode}
+\cs_set_nopar:Npn \cs_to_str:N
+  {
+    \if:w \int_to_roman:w - `0 \token_to_str:N \ %
+      \cs_to_str_aux:w
+    \fi:
+    \exp_after:wN \use_none:n \token_to_str:N
+ }
+\cs_set_nopar:Npn \cs_to_str_aux:w #1 \use_none:n
+  { ~ \int_to_roman:w - `0 \fi: }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\cs_split_function:NN}
+% \begin{macro}[aux]{\cs_split_function_aux:w}
+% \begin{macro}[aux]{\cs_split_function_auxii:w}
+%   This function takes a function name and splits it into name with
+%   the escape char removed and argument specification. In addition to
+%   this, a third argument, a boolean \meta{true} or \meta{false} is
+%   returned with \meta{true} for when there is a colon in the function
+%   and \meta{false} if there is not. Lastly, the second argument of
+%   |\cs_split_function:NN| is supposed to be a function
+%   taking three variables, one for name, one for signature, and one
+%   for the boolean.  For example,
+%   |\cs_split_function:NN\foo_bar:cnx\use_i:nnn| as input
+%   becomes |\use_i:nnn {foo_bar}{cnx}\c_true_bool|.
+%
+%   Can't use a literal |:| because it has the wrong catcode here, so
+%   it's transformed from |@| with |\tex_lowercase:D|.
+%    \begin{macrocode}
+\group_begin:
+  \tex_lccode:D  `\@ = `\: \scan_stop:
+  \tex_catcode:D `\@ = 12~
+\tex_lowercase:D
+  {
+    \group_end:
+%    \end{macrocode}
+%   First ensure that we actually get a properly evaluated str as we
+%   don't know how many expansions |\cs_to_str:N| requires.  Insert
+%   extra colon to catch the error cases.
+%    \begin{macrocode}
+    \cs_set:Npn \cs_split_function:NN #1#2
+      {
+        \exp_after:wN \cs_split_function_aux:w 
+          \int_to_roman:w - `\q \cs_to_str:N #1 @ a \q_stop #2 
+      }
+%    \end{macrocode}
+%   If no colon in the name, |#2| is |a| with catcode 11 and |#3| is
+%   empty.  If colon in the name, then either |#2| is a colon or the
+%   first letter of the signature.  The letters here have catcode 12.
+%   If a colon was given we need to a) split off the colon and quark at
+%   the end and b) ensure we return the name, signature and boolean true
+%   We can't use |\quark_if_no_value:NTF| yet but this is very safe
+%   anyway as all tokens have catcode~12.
+%    \begin{macrocode}
+    \cs_set:Npn \cs_split_function_aux:w #1 @ #2#3 \q_stop #4
+      {
+        \if_meaning:w a #2 
+          \exp_after:wN \use_i:nn 
+        \else:
+          \exp_after:wN\use_ii:nn
+        \fi: 
+        { #4 {#1} { } \c_false_bool }
+        { \cs_split_function_auxii:w #2#3 \q_stop #4 {#1} }
+      }
+    \cs_set:Npn \cs_split_function_auxii:w #1 @a \q_stop #2#3
+      { #2{#3}{#1}\c_true_bool }
+%    \end{macrocode}
+% End of lowercase
+%    \begin{macrocode}
+  }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\cs_get_function_name:N, \cs_get_function_signature:N}
+%     Now returning the name is trivial: just discard the last two
+%     arguments. Similar for signature.
+%    \begin{macrocode}
+\cs_set:Npn \cs_get_function_name:N #1
+  { \cs_split_function:NN #1 \use_i:nnn }
+\cs_set:Npn \cs_get_function_signature:N #1
+  { \cs_split_function:NN #1 \use_ii:nnn }
+%    \end{macrocode}
+% \end{macro}
+%
+% \subsection{Exist or free}
+%
+% A control sequence is said to \emph{exist} (to be used) if has an entry in
+% the hash table and its meaning is different from the primitive
+% |\tex_relax:D| token. A control sequence is said to be \emph{free}
+% (to be defined) if it does not already exist.
+%
+% \begin{macro}[pTF]{\cs_if_exist:N,\cs_if_exist:c}
+%   Two versions for checking existence. For the |N| form we firstly
+%   check for |\scan_stop:| and then if it is in the hash
+%   table. There is no problem when inputting something like |\else:|
+%   or |\fi:| as \TeX{} will only ever skip input in case the token
+%   tested against is |\scan_stop:|.
+%    \begin{macrocode}
+\prg_set_conditional:Npnn \cs_if_exist:N #1 { p , T , F , TF }
+  {
+    \if_meaning:w #1 \scan_stop:
+      \prg_return_false:
+    \else:
+      \if_cs_exist:N #1
+        \prg_return_true:
+      \else:
+        \prg_return_false:
+      \fi:
+    \fi:
+  }
+%    \end{macrocode}
+%   For the |c| form we firstly check if it is in the hash table and
+%   then for |\scan_stop:| so that we do not add it to the hash table
+%   unless it was already there. Here we have to be careful as the text
+%   to be skipped if the first test is false may contain tokens that
+%   disturb the scanner. Therefore, we ensure that the second test is
+%   performed after the first one has concluded completely.
+%    \begin{macrocode}
+\prg_set_conditional:Npnn \cs_if_exist:c #1 { p , T , F , TF }
+  {
+    \if_cs_exist:w #1 \cs_end: 
+      \exp_after:wN \use_i:nn
+    \else:
+      \exp_after:wN \use_ii:nn
+    \fi:
+    {
+      \exp_after:wN \if_meaning:w \cs:w #1 \cs_end: \scan_stop:
+        \prg_return_false:
+      \else:   
+        \prg_return_true:
+      \fi:
+    }
+    \prg_return_false:
+  }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}[pTF]{\cs_if_free:N,\cs_if_free:c}
+%   The logical reversal of the above.
+%    \begin{macrocode}
+\prg_set_conditional:Npnn \cs_if_free:N #1 { p , T , F , TF }
+  {
+    \if_meaning:w #1 \scan_stop:
+      \prg_return_true:
+    \else:
+      \if_cs_exist:N #1
+        \prg_return_false:
+      \else:
+        \prg_return_true:
+      \fi:
+    \fi:
+  }
+\prg_set_conditional:Npnn \cs_if_free:c #1 { p , T , F , TF }
+  {
+    \if_cs_exist:w #1 \cs_end: 
+      \exp_after:wN \use_i:nn
+    \else:
+      \exp_after:wN \use_ii:nn
+    \fi:
+      {
+        \exp_after:wN \if_meaning:w \cs:w #1 \cs_end: \scan_stop:
+          \prg_return_true:
+        \else:   
+          \prg_return_false:
+        \fi:
+      }
+      { \prg_return_true: }
+  }
+%    \end{macrocode}
+% \end{macro}
+%
+% \subsection{Defining and checking (new) functions}
+%
+% \begin{macro}{\c_minus_one, \c_zero, \c_sixteen}
+% \begin{macro}{\c_six, \c_seven, \c_twelve}
+%    We need the constants |\c_minus_one| and |\c_sixteen| now for
+%    writing information to the log and the terminal and |\c_zero|
+%    which is used by some functions in the \textsf{l3alloc} module. The
+%    rest are defined in the \textsf{l3int} module -- at least for the
+%    ones that can be defined with |\tex_chardef:D| or
+%    |\tex_mathchardef:D|. For other constants the \textsf{l3int} module is
+%    required but it can't be used until the allocation has been set
+%    up properly! The actual allocation mechanism is in
+%    \textsf{l3alloc} and as \TeX{} wants to reserve count registers
+%    0--9, the first available one is~10 so we use that for
+%    |\c_minus_one|.
+%    \begin{macrocode}
+%<*package>
+\cs_set_eq:NwN \c_minus_one \m@ne
+%</package>
+%<*initex>
+\tex_countdef:D \c_minus_one = 10 ~
+\c_minus_one = -1 ~
+%</initex>
+\tex_chardef:D \c_sixteen = 16~
+\tex_chardef:D \c_zero    = 0~
+\tex_chardef:D \c_six     = 6~
+\tex_chardef:D \c_seven   = 7~
+\tex_chardef:D \c_twelve  = 12~
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\c_max_register_int}
+%   This is here as this particular integer is needed both in package
+%   mode and to bootstrap \pkg{l3alloc}
+%    \begin{macrocode}
+\tex_mathchardef:D \c_max_register_int = 32 767 \scan_stop:
+%    \end{macrocode}
+% \end{macro}
+%
+%    We provide two kinds of functions that can be used to define
+%    control sequences. On the one hand we have functions that check
+%    if their argument doesn't already exist, they are called
+%    |\..._new|. The second type of defining functions doesn't check
+%    if the argument is already defined.
+%
+%    Before we can define them, we need some auxiliary macros that
+%    allow us to generate error messages. The definitions here are
+%    only temporary, they will be redefined later on.
+%
+% \begin{macro}{\iow_log:x}
+% \begin{macro}{\iow_term:x}
+%    We define a routine to write only to the log file. And a
+%    similar one for writing to both the log file and the terminal.
+%    These will be redefined later by \pkg{l3io}.
+%    \begin{macrocode}
+\cs_set_protected_nopar:Npn \iow_log:x
+  { \tex_immediate:D \tex_write:D \c_minus_one }
+\cs_set_protected_nopar:Npn \iow_term:x
+  { \tex_immediate:D \tex_write:D \c_sixteen }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\msg_kernel_error:nnxx}
+% \begin{macro}{\msg_kernel_error:nnx}
+% \begin{macro}{\msg_kernel_error:nn}
+%   If an internal error occurs before \LaTeX3 has loaded \pkg{l3msg} then
+%   the code should issue a usable if terse error message and halt. This
+%   can only happen if a coding error is made by the team, so this is
+%   a reasonable response.
+%    \begin{macrocode}
+\cs_set_protected_nopar:Npn \msg_kernel_error:nnxx #1#2#3#4
+  { 
+    \tex_errmessage:D
+      {
+        !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!~! ^^J
+        Argh,~internal~LaTeX3~error! ^^J ^^J
+        Module ~ #1 , ~ message~name~"#2": ^^J
+        Arguments~'#3'~and~'#4' ^^J ^^J
+        This~is~one~for~The~LaTeX3~Project:~bailing~out
+      }
+    \tex_end:D
+  }
+\cs_set_protected_nopar:Npn \msg_kernel_error:nnx #1#2#3
+  { \msg_kernel_error:nnxx {#1} {#2} {#3} { } }
+\cs_set_protected_nopar:Npn \msg_kernel_error:nn #1#2
+  { \msg_kernel_error:nnxx {#1} {#2} { } { } }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% 
+% \begin{macro}{\msg_line_context:}
+%   Another one from \pkg{l3msg} which will be altered later.
+%    \begin{macrocode}
+\cs_set_nopar:Npn \msg_line_context:
+  { on~line~\tex_the:D \tex_inputlineno:D }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\cs_record_meaning:N}
+%    This macro will be used later on for tracing purposes. But we
+%    need some more modules to define it, so we just give some dummy
+%    definition here.
+%    \begin{macrocode}
+\cs_set:Npn \cs_record_meaning:N #1 { }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\chk_if_free_cs:N, \chk_if_free_cs:c}
+%   This command is called by |\cs_new_nopar:Npn| and |\cs_new_eq:NN|
+%   \emph{etc.}\
+%   to make sure that the argument sequence is not already in use. If
+%   it is, an error is signalled.  It checks if \meta{csname} is
+%   undefined or |\scan_stop:|. Otherwise an error message is
+%   issued. We have to make sure we don't put the argument into the
+%   conditional processing since it may be an |\if...| type function!
+%    \begin {macrocode}
+\cs_set_protected_nopar:Npn \chk_if_free_cs:N #1
+  {
+    \cs_if_free:NF #1
+      {
+        \msg_kernel_error:nnxx { kernel } { command-already-defined }
+          { \token_to_str:N #1 } { \token_to_meaning:N #1 }
+      }
+}
+%<*package>
+\tex_ifodd:D \@l@expl@log@functions@bool
+  \cs_set_protected_nopar:Npn \chk_if_free_cs:N #1
+    {
+      \cs_if_free:NF #1
+        {
+          \msg_kernel_error:nnxx { kernel } { command-already-defined }
+            { \token_to_str:N #1 } { \token_to_meaning:N #1 }
+        }
+    \iow_log:x { Defining~\token_to_str:N #1~ \msg_line_context: }
+  }
+\fi:
+%</package>
+\cs_set_protected_nopar:Npn \chk_if_free_cs:c 
+  { \exp_args:Nc \chk_if_free_cs:N }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\chk_if_exist_cs:N, \chk_if_exist_cs:c}
+%    This function issues a warning message when the control sequence
+%    in its argument does not exist.
+%    \begin{macrocode}
+\cs_set_protected_nopar:Npn \chk_if_exist_cs:N #1
+  {
+    \cs_if_exist:NF #1
+      {
+        \msg_kernel_error:nnxx { kernel } { command-not-defined }
+          { \token_to_str:N #1 } { \token_to_meaning:N #1 }
+      }
+  }
+\cs_set_protected_nopar:Npn \chk_if_exist_cs:c
+  { \exp_args:Nc \chk_if_exist_cs:N }
+%    \end{macrocode}
+% \end{macro}
+%
+% \subsection{More new definitions}
+%
+% \begin{macro}{\cs_new_nopar:Npn}
+% \begin{macro}{\cs_new_nopar:Npx}
+% \begin{macro}{\cs_new:Npn}
+% \begin{macro}{\cs_new:Npx}
+% \begin{macro}{\cs_new_protected_nopar:Npn}
+% \begin{macro}{\cs_new_protected_nopar:Npx}
+% \begin{macro}{\cs_new_protected:Npn}
+% \begin{macro}{\cs_new_protected:Npx}
+%   Global versions of the above functions.
+%     \begin {macrocode}
+\cs_set:Npn \cs_tmp:w #1#2
+  {
+    \cs_set_protected_nopar:Npn #1 ##1 
+       { 
+         \chk_if_free_cs:N ##1
+         #2 ##1
+      }
+  }
+\cs_tmp:w \cs_new_nopar:Npn           \cs_gset_nopar:Npn
+\cs_tmp:w \cs_new_nopar:Npx           \cs_gset_nopar:Npx
+\cs_tmp:w \cs_new:Npn                 \cs_gset:Npn
+\cs_tmp:w \cs_new:Npx                 \cs_gset:Npx
+\cs_tmp:w \cs_new_protected_nopar:Npn \cs_gset_protected_nopar:Npn
+\cs_tmp:w \cs_new_protected_nopar:Npx \cs_gset_protected_nopar:Npx
+\cs_tmp:w \cs_new_protected:Npn       \cs_gset_protected:Npn
+\cs_tmp:w \cs_new_protected:Npx       \cs_gset_protected:Npx
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\cs_set_nopar:cpn}
+% \begin{macro}{\cs_set_nopar:cpx}
+% \begin{macro}{\cs_gset_nopar:cpn}
+% \begin{macro}{\cs_gset_nopar:cpx}
+% \begin{macro}{\cs_new_nopar:cpn}
+% \begin{macro}{\cs_new_nopar:cpx}
+%   Like |\cs_set_nopar:Npn| and |\cs_new_nopar:Npn|, except that the
+%   first argument consists of the sequence of characters that should
+%   be used to form the name of the desired control sequence (the |c|
+%   stands for csname argument, see the expansion module). Global
+%   versions are also provided.
+%
+%   |\cs_set_nopar:cpn|\meta{string}\meta{rep-text} will turn \meta{string}
+%   into a csname and then assign \m {rep-text} to it by using
+%   |\cs_set_nopar:Npn|.  This means that there might be a parameter
+%   string between the two arguments.
+%    \begin{macrocode}
+\cs_set:Npn \cs_tmp:w #1#2
+  { \cs_new_protected_nopar:Npn #1 { \exp_args:Nc #2 } }
+\cs_tmp:w \cs_set_nopar:cpn  \cs_set_nopar:Npn 
+\cs_tmp:w \cs_set_nopar:cpx  \cs_set_nopar:Npx 
+\cs_tmp:w \cs_gset_nopar:cpn \cs_gset_nopar:Npn 
+\cs_tmp:w \cs_gset_nopar:cpx \cs_gset_nopar:Npx 
+\cs_tmp:w \cs_new_nopar:cpn  \cs_new_nopar:Npn
+\cs_tmp:w \cs_new_nopar:cpx  \cs_new_nopar:Npx
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\cs_set:cpn}
+% \begin{macro}{\cs_set:cpx}
+% \begin{macro}{\cs_gset:cpn}
+% \begin{macro}{\cs_gset:cpx}
+% \begin{macro}{\cs_new:cpn}
+% \begin{macro}{\cs_new:cpx}
+%   Variants of the |\cs_set:Npn| versions which make a csname out
+%   of the first arguments. We may also do this globally.
+%    \begin{macrocode}
+\cs_tmp:w \cs_set:cpn  \cs_set:Npn 
+\cs_tmp:w \cs_set:cpx  \cs_set:Npx 
+\cs_tmp:w \cs_gset:cpn \cs_gset:Npn 
+\cs_tmp:w \cs_gset:cpx \cs_gset:Npx 
+\cs_tmp:w \cs_new:cpn  \cs_new:Npn 
+\cs_tmp:w \cs_new:cpx  \cs_new:Npx 
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\cs_set_protected_nopar:cpn}
+% \begin{macro}{\cs_set_protected_nopar:cpx}
+% \begin{macro}{\cs_gset_protected_nopar:cpn}
+% \begin{macro}{\cs_gset_protected_nopar:cpx}
+% \begin{macro}{\cs_new_protected_nopar:cpn}
+% \begin{macro}{\cs_new_protected_nopar:cpx}
+%   Variants of the |\cs_set_protected_nopar:Npn| versions which make a csname
+%   out of the first arguments. We may also do this globally.
+%    \begin{macrocode}
+\cs_tmp:w \cs_set_protected_nopar:cpn  \cs_set_protected_nopar:Npn 
+\cs_tmp:w \cs_set_protected_nopar:cpx  \cs_set_protected_nopar:Npx  
+\cs_tmp:w \cs_gset_protected_nopar:cpn \cs_gset_protected_nopar:Npn
+\cs_tmp:w \cs_gset_protected_nopar:cpx \cs_gset_protected_nopar:Npx
+\cs_tmp:w \cs_new_protected_nopar:cpn  \cs_new_protected_nopar:Npn
+\cs_tmp:w \cs_new_protected_nopar:cpx  \cs_new_protected_nopar:Npx 
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\cs_set_protected:cpn}
+% \begin{macro}{\cs_set_protected:cpx}
+% \begin{macro}{\cs_gset_protected:cpn}
+% \begin{macro}{\cs_gset_protected:cpx}
+% \begin{macro}{\cs_new_protected:cpn}
+% \begin{macro}{\cs_new_protected:cpx}
+%   Variants of the |\cs_set_protected:Npn| versions which make a csname
+%   out of the first arguments. We may also do this globally.
+%    \begin{macrocode}
+\cs_tmp:w \cs_set_protected:cpn  \cs_set_protected:Npn 
+\cs_tmp:w \cs_set_protected:cpx  \cs_set_protected:Npx
+\cs_tmp:w \cs_gset_protected:cpn \cs_gset_protected:Npn 
+\cs_tmp:w \cs_gset_protected:cpx \cs_gset_protected:Npx
+\cs_tmp:w \cs_new_protected:cpn  \cs_new_protected:Npn 
+\cs_tmp:w \cs_new_protected:cpx  \cs_new_protected:Npx 
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}[aux]{\use_0_parameter:}
+% \begin{macro}[aux]{\use_1_parameter:}
+% \begin{macro}[aux]{\use_2_parameter:}
+% \begin{macro}[aux]{\use_3_parameter:}
+% \begin{macro}[aux]{\use_4_parameter:}
+% \begin{macro}[aux]{\use_5_parameter:}
+% \begin{macro}[aux]{\use_6_parameter:}
+% \begin{macro}[aux]{\use_7_parameter:}
+% \begin{macro}[aux]{\use_8_parameter:}
+% \begin{macro}[aux]{\use_9_parameter:}
+%   For using parameters, \emph{i.e.}, when you need to define a function to
+%   process three parameters. See \textsf{xparse} for an application.
+%    \begin{macrocode}
+\cs_new_nopar:cpn { use_0_parameter: } { }
+\cs_new_nopar:cpn { use_1_parameter: } { {##1} }
+\cs_new_nopar:cpn { use_2_parameter: } { {##1} {##2} }
+\cs_new_nopar:cpn { use_3_parameter: } { {##1} {##2} {##3} }
+\cs_new_nopar:cpn { use_4_parameter: } { {##1} {##2} {##3} {##4} }
+\cs_new_nopar:cpn { use_5_parameter: } { {##1} {##2} {##3} {##4} {##5} }
+\cs_new_nopar:cpn { use_6_parameter: } { {##1} {##2} {##3} {##4} {##5} {##6} }
+\cs_new_nopar:cpn { use_7_parameter: }
+  { {##1} {##2} {##3} {##4} {##5}{##6} {##7} }
+\cs_new_nopar:cpn { use_8_parameter: }
+  { {##1} {##2} {##3} {##4} {##5} {##6} {##7} {##8} }
+\cs_new_nopar:cpn{ use_9_parameter: }
+  { {##1} {##2} {##3} {##4} {##5} {##6} {##7} {##8} {##9} }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \subsection{Copying definitions}
+%
+% \begin{macro}{\cs_set_eq:NN}
+% \begin{macro}{\cs_set_eq:cN}
+% \begin{macro}{\cs_set_eq:Nc}
+% \begin{macro}{\cs_set_eq:cc}
+%    These macros allow us to copy the definition of a control sequence
+%    to another control sequence.
+%
+%    The |=| sign allows us to define funny char tokens like |=|
+%    itself or \verb*| | with this function. For the definition of
+%    |\c_space_char{~}| to work we need the |~| after the |=|.
+%
+%    |\cs_set_eq:NN| is long to avoid problems with a literal argument
+%    of |\par|.  While |\cs_new_eq:NN| will probably never be correct
+%    with a first argument of |\par|, define it long in order to throw
+%    an \enquote{already defined} error rather than
+%    \enquote{runaway argument}.
+%    \begin{macrocode}
+\cs_new_protected:Npn \cs_set_eq:NN #1 { \cs_set_eq:NwN #1 =~ }
+\cs_new_protected_nopar:Npn \cs_set_eq:cN { \exp_args:Nc  \cs_set_eq:NN }
+\cs_new_protected_nopar:Npn \cs_set_eq:Nc { \exp_args:NNc \cs_set_eq:NN }
+\cs_new_protected_nopar:Npn \cs_set_eq:cc { \exp_args:Ncc \cs_set_eq:NN }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\cs_new_eq:NN}
+% \begin{macro}{\cs_new_eq:cN}
+% \begin{macro}{\cs_new_eq:Nc}
+% \begin{macro}{\cs_new_eq:cc}
+%    \begin{macrocode}
+\cs_new_protected:Npn \cs_new_eq:NN #1 
+  {
+    \chk_if_free_cs:N #1
+    \pref_global:D \cs_set_eq:NN #1
+  }
+\cs_new_protected_nopar:Npn \cs_new_eq:cN { \exp_args:Nc  \cs_new_eq:NN }
+\cs_new_protected_nopar:Npn \cs_new_eq:Nc { \exp_args:NNc \cs_new_eq:NN }
+\cs_new_protected_nopar:Npn \cs_new_eq:cc { \exp_args:Ncc \cs_new_eq:NN }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\cs_gset_eq:NN}
+% \begin{macro}{\cs_gset_eq:cN}
+% \begin{macro}{\cs_gset_eq:Nc}
+% \begin{macro}{\cs_gset_eq:cc}
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \cs_gset_eq:NN { \pref_global:D  \cs_set_eq:NN }
+\cs_new_protected_nopar:Npn \cs_gset_eq:Nc { \exp_args:NNc  \cs_gset_eq:NN }
+\cs_new_protected_nopar:Npn \cs_gset_eq:cN { \exp_args:Nc   \cs_gset_eq:NN }
+\cs_new_protected_nopar:Npn \cs_gset_eq:cc { \exp_args:Ncc  \cs_gset_eq:NN }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \subsection{Undefining functions}
+%
+% \begin{macro}{\cs_undefine:N , \cs_undefine:c}
+%    The following function is used to free the main memory from the
+%    definition of some function that isn't in use any longer.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \cs_undefine:N #1
+  { \cs_gset_eq:NN #1 \c_undefined:D }
+\cs_new_protected_nopar:Npn \cs_undefine:c #1
+  { \cs_gset_eq:cN {#1} \c_undefined:D }
+%    \end{macrocode}
+% \end{macro}
+%
+% \subsection{Defining functions from a given number of arguments}
+%
+% \begin{macro}{\cs_get_arg_count_from_signature:N}
+% \begin{macro}[aux]{\cs_get_arg_count_from_signature_aux:nnN}
+% \begin{macro}[aux]{\cs_get_arg_count_from_signature_auxii:w}
+%   Counting the number of tokens in the signature, i.e., the number
+%   of arguments the function should take.  If there is no signature,
+%   we return that there is $-1$ arguments to signal an error.
+%   Otherwise we insert the string |9876543210| after the signature.
+%   If the signature is empty, the number we want is $0$ so we remove
+%   the first nine tokens and return the tenth.  Similarly, if the
+%   signature is |nnn| we want to remove the nine tokens |nnn987654|
+%   and return $3$.  Therefore, we simply remove the first nine tokens
+%   and then return the tenth.
+%    \begin{macrocode}
+\cs_new:Npn \cs_get_arg_count_from_signature:N #1
+  { \cs_split_function:NN #1 \cs_get_arg_count_from_signature_aux:nnN }
+\cs_new:Npn \cs_get_arg_count_from_signature_aux:nnN #1#2#3
+  {
+    \if_predicate:w #3
+      \exp_after:wN \use_i:nn 
+    \else:
+      \exp_after:wN\use_ii:nn
+    \fi: 
+    {
+      \exp_after:wN \cs_get_arg_count_from_signature_auxii:w 
+        \use_none:nnnnnnnnn  #2 9876543210 \q_stop 
+    }
+    { -1 }
+}
+\cs_new:Npn \cs_get_arg_count_from_signature_auxii:w #1#2 \q_stop {#1}
+%    \end{macrocode}
+% A variant form we need right away.
+%    \begin{macrocode}
+\cs_new_nopar:Npn \cs_get_arg_count_from_signature:c
+  { \exp_args:Nc \cs_get_arg_count_from_signature:N }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}[aux]{\cs_generate_from_arg_count:NNnn}
+% \begin{macro}[aux]{\cs_generate_from_arg_count_error_msg:Nn}
+%   We provide a constructor function for defining functions with a
+%   given number of arguments.  For this we need to choose the correct
+%   parameter text and then use that when defining.  Since \TeX{}
+%   supports from zero to nine arguments, we use a simple switch to
+%   choose the correct parameter text, ensuring the result is returned
+%   after finishing the conditional.  If it is not between zero and
+%   nine, we throw an error.
+%
+%   1: function to define, 2: with what to define it, 3: the number of
+%   args it requires and 4: the replacement text
+%    \begin{macrocode}
+\cs_new_protected:Npn \cs_generate_from_arg_count:NNnn #1#2#3#4
+  {
+    \if_case:w \int_eval:w #3 \int_eval_end:
+      \use_i_after_orelse:nw {#2#1}
+    \or:
+      \use_i_after_orelse:nw {#2#1 ##1}
+    \or:
+      \use_i_after_orelse:nw {#2#1 ##1##2}
+    \or:
+      \use_i_after_orelse:nw {#2#1 ##1##2##3}
+    \or:
+      \use_i_after_orelse:nw {#2#1 ##1##2##3##4}
+    \or:
+      \use_i_after_orelse:nw {#2#1 ##1##2##3##4##5}
+    \or:
+      \use_i_after_orelse:nw {#2#1 ##1##2##3##4##5##6}
+    \or:
+      \use_i_after_orelse:nw {#2#1 ##1##2##3##4##5##6##7}
+    \or:
+      \use_i_after_orelse:nw {#2#1 ##1##2##3##4##5##6##7##8}
+    \or:
+      \use_i_after_orelse:nw {#2#1 ##1##2##3##4##5##6##7##8##9}
+    \else:
+      \use_i_after_fi:nw
+        {
+          \cs_generate_from_arg_count_error_msg:Nn #1 {#3}
+          \use_none:n
+        }
+    \fi:
+    {#4}
+  }
+%    \end{macrocode}
+% A variant form we need right away.
+%    \begin{macrocode}
+\cs_new_nopar:Npn \cs_generate_from_arg_count:cNnn
+  { \exp_args:Nc \cs_generate_from_arg_count:NNnn }
+%    \end{macrocode}
+% The error message. Elsewhere we use the value of $-1$ to signal a
+% missing colon in a function, so provide a hint for help on this.
+%    \begin{macrocode}
+\cs_new:Npn \cs_generate_from_arg_count_error_msg:Nn #1#2
+  {
+    \msg_kernel_error:nnxx { kernel } { bad-number-of-arguments }
+      { \token_to_str:N #1 } { \int_eval:n {#2} }
+  } 
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \subsection{Using the signature to define functions}
+%
+% We can now combine some of the tools we have to provide a simple
+% interface for defining functions.  We define some simpler functions
+% with user interface |\cs_set:Nn \foo_bar:nn {#1,#2}|, \emph{i.e.}, the
+% number of arguments is read from the signature.
+%
+% 
+% \begin{macro}{\cs_set:Nn}
+% \begin{macro}{\cs_set:Nx}
+% \begin{macro}{\cs_set_nopar:Nn}
+% \begin{macro}{\cs_set_nopar:Nx}
+% \begin{macro}{\cs_set_protected:Nn}
+% \begin{macro}{\cs_set_protected:Nx}
+% \begin{macro}{\cs_set_protected_nopar:Nn}
+% \begin{macro}{\cs_set_protected_nopar:Nx}
+% \begin{macro}{\cs_gset:Nn}
+% \begin{macro}{\cs_gset:Nx}
+% \begin{macro}{\cs_gset_nopar:Nn}
+% \begin{macro}{\cs_gset_nopar:Nx}
+% \begin{macro}{\cs_gset_protected:Nn}
+% \begin{macro}{\cs_gset_protected:Nx}
+% \begin{macro}{\cs_gset_protected_nopar:Nn}
+% \begin{macro}{\cs_gset_protected_nopar:Nx}
+% We want to define |\cs_set:Nn| as 
+% \begin{verbatim}
+%   \cs_set_protected:Npn \cs_set:Nn #1#2
+%     {
+%       \cs_generate_from_arg_count:NNnn #1 \cs_set:Npn 
+%         { \cs_get_arg_count_from_signature:N #1 } {#2}
+%     }
+% \end{verbatim}
+% In short, to define |\cs_set:Nn| we need just use |\cs_set:Npn|,
+% everything else is the same for each variant.  Therefore, we can
+% make it simpler by temporarily defining a function to do this for
+% us.
+%    \begin{macrocode}
+\cs_set:Npn \cs_tmp:w #1#2#3
+  {
+    \cs_set_protected:cpx { cs_ #1 : #2 } ##1##2
+      {
+        \exp_not:N \cs_generate_from_arg_count:NNnn ##1
+        \exp_after:wN \exp_not:N \cs:w cs_#1 : #3 \cs_end: 
+          { \exp_not:N\cs_get_arg_count_from_signature:N ##1 }{##2}
+      }
+  }
+%    \end{macrocode}
+% Then we define the 32 variants beginning with |N|. 
+%    \begin{macrocode}
+\cs_tmp:w { set }                  { Nn } { Npn }
+\cs_tmp:w { set }                  { Nx } { Npx }
+\cs_tmp:w { set_nopar }            { Nn } { Npn }
+\cs_tmp:w { set_nopar }            { Nx } { Npx }
+\cs_tmp:w { set_protected }        { Nn } { Npn }
+\cs_tmp:w { set_protected }        { Nx } { Npx }
+\cs_tmp:w { set_protected_nopar }  { Nn } { Npn }
+\cs_tmp:w { set_protected_nopar }  { Nx } { Npx }
+\cs_tmp:w { gset }                 { Nn } { Npn }
+\cs_tmp:w { gset }                 { Nx } { Npx }
+\cs_tmp:w { gset_nopar }           { Nn } { Npn }
+\cs_tmp:w { gset_nopar }           { Nx } { Npx }
+\cs_tmp:w { gset_protected }       { Nn } { Npn }
+\cs_tmp:w { gset_protected }       { Nx } { Npx }
+\cs_tmp:w { gset_protected_nopar } { Nn } { Npn }
+\cs_tmp:w { gset_protected_nopar } { Nx } { Npx }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\cs_new:Nn}
+% \begin{macro}{\cs_new:Nx}
+% \begin{macro}{\cs_new_nopar:Nn}
+% \begin{macro}{\cs_new_nopar:Nx}
+% \begin{macro}{\cs_new_protected:Nn}
+% \begin{macro}{\cs_new_protected:Nx}
+% \begin{macro}{\cs_new_protected_nopar:Nn}
+% \begin{macro}{\cs_new_protected_nopar:Nx}
+%    \begin{macrocode}
+\cs_tmp:w { new }                 { Nn } { Npn }
+\cs_tmp:w { new }                 { Nx } { Npx }
+\cs_tmp:w { new_nopar }           { Nn } { Npn }
+\cs_tmp:w { new_nopar }           { Nx } { Npx }
+\cs_tmp:w { new_protected }       { Nn } { Npn }
+\cs_tmp:w { new_protected }       { Nx } { Npx }
+\cs_tmp:w { new_protected_nopar } { Nn } { Npn }
+\cs_tmp:w { new_protected_nopar } { Nx } { Npx }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% Then something similar for the |c| variants. 
+% \begin{verbatim}
+%   \cs_set_protected:Npn \cs_set:cn #1#2
+%     {
+%       \cs_generate_from_arg_count:cNnn {#1} \cs_set:Npn 
+%         { \cs_get_arg_count_from_signature:c {#1} } {#2}
+%     }
+% \end{verbatim}
+%    \begin{macrocode}
+\cs_set:Npn \cs_tmp:w #1#2#3
+  {
+    \cs_set_protected:cpx {cs_#1:#2}##1##2{
+      \exp_not:N\cs_generate_from_arg_count:cNnn {##1}
+      \exp_after:wN \exp_not:N \cs:w cs_#1:#3 \cs_end: 
+        { \exp_not:N \cs_get_arg_count_from_signature:c {##1} } {##2}
+    }
+  }
+%    \end{macrocode}
+% \begin{macro}{\cs_set:cn}
+% \begin{macro}{\cs_set:cx}
+% \begin{macro}{\cs_set_nopar:cn}
+% \begin{macro}{\cs_set_nopar:cx}
+% \begin{macro}{\cs_set_protected:cn}
+% \begin{macro}{\cs_set_protected:cx}
+% \begin{macro}{\cs_set_protected_nopar:cn}
+% \begin{macro}{\cs_set_protected_nopar:cx}
+% \begin{macro}{\cs_gset:cn}
+% \begin{macro}{\cs_gset:cx}
+% \begin{macro}{\cs_gset_nopar:cn}
+% \begin{macro}{\cs_gset_nopar:cx}
+% \begin{macro}{\cs_gset_protected:cn}
+% \begin{macro}{\cs_gset_protected:cx}
+% \begin{macro}{\cs_gset_protected_nopar:cn}
+% \begin{macro}{\cs_gset_protected_nopar:cx}
+% The 32 |c| variants.
+%    \begin{macrocode}
+\cs_tmp:w { set }                  { cn } { Npn }
+\cs_tmp:w { set }                  { cx } { Npx }
+\cs_tmp:w { set_nopar }            { cn } { Npn }
+\cs_tmp:w { set_nopar }            { cx } { Npx }
+\cs_tmp:w { set_protected }        { cn } { Npn }
+\cs_tmp:w { set_protected }        { cx } { Npx }
+\cs_tmp:w { set_protected_nopar }  { cn } { Npn }
+\cs_tmp:w { set_protected_nopar }  { cx } { Npx }
+\cs_tmp:w { gset }                 { cn } { Npn }
+\cs_tmp:w { gset }                 { cx } { Npx }
+\cs_tmp:w { gset_nopar }           { cn } { Npn }
+\cs_tmp:w { gset_nopar }           { cx } { Npx }
+\cs_tmp:w { gset_protected }       { cn } { Npn }
+\cs_tmp:w { gset_protected }       { cx } { Npx }
+\cs_tmp:w { gset_protected_nopar } { cn } { Npn }
+\cs_tmp:w { gset_protected_nopar } { cx } { Npx }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\cs_new:cn}
+% \begin{macro}{\cs_new:cx}
+% \begin{macro}{\cs_new_nopar:cn}
+% \begin{macro}{\cs_new_nopar:cx}
+% \begin{macro}{\cs_new_protected:cn}
+% \begin{macro}{\cs_new_protected:cx}
+% \begin{macro}{\cs_new_protected_nopar:cn}
+% \begin{macro}{\cs_new_protected_nopar:cx}
+%    \begin{macrocode}
+\cs_tmp:w { new }                 { cn } { Npn }
+\cs_tmp:w { new }                 { cx } { Npx }
+\cs_tmp:w { new_nopar }           { cn } { Npn }
+\cs_tmp:w { new_nopar }           { cx } { Npx }
+\cs_tmp:w { new_protected }       { cn } { Npn }
+\cs_tmp:w { new_protected }       { cx } { Npx }
+\cs_tmp:w { new_protected_nopar } { cn } { Npn }
+\cs_tmp:w { new_protected_nopar } { cx } { Npx }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% 
+% \subsection{Checking control sequence equality}
+%
+% \begin{macro}[pTF]{\cs_if_eq:NN,\cs_if_eq:cN,\cs_if_eq:Nc,\cs_if_eq:cc}
+% Check if two control sequences are identical.
+%    \begin{macrocode}
+\prg_new_conditional:Npnn \cs_if_eq:NN #1#2 { p , T , F , TF }
+  {
+    \if_meaning:w #1#2
+      \prg_return_true: \else: \prg_return_false: \fi:
+  }
+\cs_new_nopar:Npn \cs_if_eq_p:cN { \exp_args:Nc  \cs_if_eq_p:NN }
+\cs_new_nopar:Npn \cs_if_eq:cNTF { \exp_args:Nc  \cs_if_eq:NNTF }
+\cs_new_nopar:Npn \cs_if_eq:cNT  { \exp_args:Nc  \cs_if_eq:NNT }
+\cs_new_nopar:Npn \cs_if_eq:cNF  { \exp_args:Nc  \cs_if_eq:NNF }
+\cs_new_nopar:Npn \cs_if_eq_p:Nc { \exp_args:NNc \cs_if_eq_p:NN }
+\cs_new_nopar:Npn \cs_if_eq:NcTF { \exp_args:NNc \cs_if_eq:NNTF }
+\cs_new_nopar:Npn \cs_if_eq:NcT  { \exp_args:NNc \cs_if_eq:NNT }
+\cs_new_nopar:Npn \cs_if_eq:NcF  { \exp_args:NNc \cs_if_eq:NNF }
+\cs_new_nopar:Npn \cs_if_eq_p:cc { \exp_args:Ncc \cs_if_eq_p:NN }
+\cs_new_nopar:Npn \cs_if_eq:ccTF { \exp_args:Ncc \cs_if_eq:NNTF }
+\cs_new_nopar:Npn \cs_if_eq:ccT  { \exp_args:Ncc \cs_if_eq:NNT }
+\cs_new_nopar:Npn \cs_if_eq:ccF  { \exp_args:Ncc \cs_if_eq:NNF }
+%    \end{macrocode}
+% \end{macro}
+% 
+% \subsection{Diagnostic wrapper functions}
+%
+% \begin{macro}{\kernel_register_show:N, \kernel_register_show:c}
+%    \begin{macrocode}
+\cs_new_nopar:Npn \kernel_register_show:N #1
+  {
+    \cs_if_exist:NTF #1
+     { \tex_showthe:D #1 }
+     {
+       \msg_kernel_error:nnx { kernel } { variable-not-defined }
+         { \token_to_str:N #1 }
+     }
+  }
+\cs_new_nopar:Npn \kernel_register_show:c { \exp_args:Nc \int_show:N }
+%    \end{macrocode}
+% \end{macro}
+%
+%  \subsection{Engine specific definitions}
+%
+% \begin{macro}
+%   {
+%     \c_pdftex_is_engine_bool, \c_luatex_is_engine_bool,
+%     \c_xetex_is_engine_bool
+%   }
+% \begin{macro}[TF]{\xetex_if_engine:, \luatex_if_engine:, \pdftex_if_engine:}
+%   In some cases it will be useful to know which engine we're running.
+%   Don't provide a |_p| predicate because the |_bool| is used for the
+%   same thing. This can all be hard-coded for speed.
+%    \begin{macrocode}
+\cs_new_eq:NN \luatex_if_engine:T  \use_none:n
+\cs_new_eq:NN \luatex_if_engine:F  \use:n
+\cs_new_eq:NN \luatex_if_engine:TF \use_ii:nn
+\cs_new_eq:NN \pdftex_if_engine:T  \use:n
+\cs_new_eq:NN \pdftex_if_engine:F  \use_none:n
+\cs_new_eq:NN \pdftex_if_engine:TF \use_i:nn
+\cs_new_eq:NN \xetex_if_engine:T   \use_none:n
+\cs_new_eq:NN \xetex_if_engine:F   \use:n
+\cs_new_eq:NN \xetex_if_engine:TF  \use_ii:nn
+\cs_new_eq:NN \c_luatex_is_engine_bool \c_false_bool
+\cs_new_eq:NN \c_pdftex_is_engine_bool \c_true_bool
+\cs_new_eq:NN \c_xetex_is_engine_bool  \c_false_bool    
+\cs_if_exist:NT \xetex_XeTeXversion:D
+  {
+    \cs_set_eq:NN \pdftex_if_engine:T  \use_none:n
+    \cs_set_eq:NN \pdftex_if_engine:F  \use:n
+    \cs_set_eq:NN \pdftex_if_engine:TF \use_ii:nn
+    \cs_set_eq:NN \xetex_if_engine:T   \use:n
+    \cs_set_eq:NN \xetex_if_engine:F   \use_none:n
+    \cs_set_eq:NN \xetex_if_engine:TF  \use_i:nn
+    \cs_set_eq:NN \c_pdftex_is_engine_bool \c_false_bool
+    \cs_set_eq:NN \c_xetex_is_engine_bool  \c_true_bool    
+  }
+\cs_if_exist:NT \luatex_directlua:D
+  {
+    \cs_set_eq:NN \luatex_if_engine:T  \use:n
+    \cs_set_eq:NN \luatex_if_engine:F  \use_none:n
+    \cs_set_eq:NN \luatex_if_engine:TF \use_i:nn
+    \cs_set_eq:NN \pdftex_if_engine:T  \use_none:n
+    \cs_set_eq:NN \pdftex_if_engine:F  \use:n
+    \cs_set_eq:NN \pdftex_if_engine:TF \use_ii:nn
+    \cs_set_eq:NN \c_luatex_is_engine_bool \c_true_bool  
+    \cs_set_eq:NN \c_pdftex_is_engine_bool \c_false_bool  
+  }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \subsection{Doing nothing functions}
+%
+% \begin{macro}{\prg_do_nothing:}
+%   This does not fit anywhere else!
+%    \begin{macrocode}
+\cs_new_nopar:Npn \prg_do_nothing: { }
+%    \end{macrocode}
+% \end{macro}
+% 
+% \subsection{String comparisons}
+%
+% \begin{macro}[pTF]{\str_if_eq:nn}
+% \begin{macro}[pTF]{\str_if_eq:xx}
+%   Modern engines provide a direct way of comparing two token lists,
+%   but returning a number. This set of conditionals therefore make life
+%   a bit clearer. The \texttt{nn} and \texttt{xx} versions are created
+%   directly as this is most efficient. These should eventually
+%   move somewhere else.
+%    \begin{macrocode}
+\prg_new_conditional:Npnn \str_if_eq:nn #1#2 { p , T , F , TF }
+  {
+    \if_int_compare:w \pdftex_strcmp:D { \exp_not:n {#1} } { \exp_not:n {#2} }
+      = \c_zero
+     \prg_return_true: \else: \prg_return_false: \fi: 
+  }
+\prg_new_conditional:Npnn \str_if_eq:xx #1#2 { p , T , F , TF }
+  {
+    \if_int_compare:w \pdftex_strcmp:D {#1} {#2} = \c_zero
+      \prg_return_true: \else: \prg_return_false: \fi: 
+  }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \subsection{Deprecated functions}
+% 
+% Deprecated on 2011-05-27, for removal by 2011-08-31.
+% 
+%    \begin{macrocode}
+\cs_new_eq:NN           \cs_gnew_nopar:Npn            \cs_new_nopar:Npn
+\cs_new_eq:NN                 \cs_gnew:Npn                  \cs_new:Npn
+\cs_new_eq:NN \cs_gnew_protected_nopar:Npn  \cs_new_protected_nopar:Npn
+\cs_new_eq:NN       \cs_gnew_protected:Npn        \cs_new_protected:Npn
+\cs_new_eq:NN           \cs_gnew_nopar:Npx            \cs_new_nopar:Npx
+\cs_new_eq:NN                 \cs_gnew:Npx                  \cs_new:Npx
+\cs_new_eq:NN \cs_gnew_protected_nopar:Npx  \cs_new_protected_nopar:Npx
+\cs_new_eq:NN       \cs_gnew_protected:Npx        \cs_new_protected:Npx
+\cs_new_eq:NN           \cs_gnew_nopar:cpn            \cs_new_nopar:cpn
+\cs_new_eq:NN                 \cs_gnew:cpn                  \cs_new:cpn
+\cs_new_eq:NN \cs_gnew_protected_nopar:cpn  \cs_new_protected_nopar:cpn
+\cs_new_eq:NN       \cs_gnew_protected:cpn        \cs_new_protected:cpn
+\cs_new_eq:NN           \cs_gnew_nopar:cpx            \cs_new_nopar:cpx
+\cs_new_eq:NN                 \cs_gnew:cpx                  \cs_new:cpx
+\cs_new_eq:NN \cs_gnew_protected_nopar:cpx  \cs_new_protected_nopar:cpx
+\cs_new_eq:NN       \cs_gnew_protected:cpx        \cs_new_protected:cpx
+%    \end{macrocode}
+%    
+%    \begin{macrocode}
+\cs_new_eq:NN \cs_gnew_eq:NN \cs_new_eq:NN
+\cs_new_eq:NN \cs_gnew_eq:cN \cs_new_eq:cN
+\cs_new_eq:NN \cs_gnew_eq:Nc \cs_new_eq:Nc
+\cs_new_eq:NN \cs_gnew_eq:cc \cs_new_eq:cc
+%    \end{macrocode}
+%    
+%    \begin{macrocode}
+\cs_new_eq:NN \cs_gundefine:N  \cs_undefine:N
+\cs_new_eq:NN \cs_gundefine:c  \cs_undefine:c
+%    \end{macrocode}
+%    
+%    \begin{macrocode}
+\cs_new_eq:NN \group_execute_after:N \group_insert_after:N
+%    \end{macrocode}
+%
+%    \begin{macrocode}
+%</initex|package>
+%    \end{macrocode}
+%
+% \end{implementation}
+%
+% \PrintIndex
diff --git a/Master/texmf-dist/source/latex/l3kernel/l3bootstrap.dtx b/Master/texmf-dist/source/latex/l3kernel/l3bootstrap.dtx
new file mode 100644
index 00000000000..69aaf46357b
--- /dev/null
+++ b/Master/texmf-dist/source/latex/l3kernel/l3bootstrap.dtx
@@ -0,0 +1,647 @@
+% \iffalse meta-comment
+%
+%% File: l3bootstrap.dtx Copyright (C) 2011 The LaTeX3 project
+%%
+%% It may be distributed and/or modified under the conditions of the
+%% LaTeX Project Public License (LPPL), either version 1.3c of this
+%% license or (at your option) any later version.  The latest version
+%% of this license is in the file
+%%
+%%    http://www.latex-project.org/lppl.txt
+%%
+%% This file is part of the "expl3 bundle" (The Work in LPPL)
+%% and all files in that bundle must be distributed together.
+%%
+%% The released version of this bundle is available from CTAN.
+%%
+%% -----------------------------------------------------------------------
+%%
+%% The development version of the bundle can be found at
+%%
+%%    http://www.latex-project.org/svnroot/experimental/trunk/
+%%
+%% for those people who are interested.
+%%
+%%%%%%%%%%%
+%% NOTE: %%
+%%%%%%%%%%%
+%%
+%%   Snapshots taken from the repository represent work in progress and may
+%%   not work or may contain conflicting material!  We therefore ask
+%%   people _not_ to put them into distributions, archives, etc. without
+%%   prior consultation with the LaTeX3 Project.
+%%
+%% -----------------------------------------------------------------------
+%
+%<*package>
+\def\GetIdInfo$%
+  Id: #1 #2 #3-#4-#5 #6 #7 $#8%
+  {%
+    \def\ExplFileDate{#3/#4/#5}%
+    \def\ExplFileVersion{#2}%
+  }
+\GetIdInfo$Id: l3bootstrap.dtx 2478 2011-06-19 21:34:23Z joseph $
+  {L3 experimental bootstrap code}
+%</package>
+%<*driver>
+\documentclass[full]{l3doc}
+\begin{document}
+  \DocInput{\jobname.dtx}
+\end{document}
+%</driver>
+% \fi
+%
+% \title{^^A
+%   The \pkg{l3bootstrap} package\\ Bootstrap code^^A
+%   \thanks{This file describes v\ExplFileVersion,
+%     last revised \ExplFileDate.}^^A
+% }
+%
+% \author{^^A
+%  The \LaTeX3 Project\thanks
+%    {^^A
+%      E-mail:
+%        \href{mailto:latex-team@latex-project.org}
+%          {latex-team@latex-project.org}^^A
+%    }^^A
+% }
+%
+% \date{Released \ExplFileDate}
+%
+% \maketitle
+%
+% \begin{documentation}
+%
+% \section{Using the \LaTeX3 modules}
+%
+% The modules documented in \file{source3} are designed to be used on top of
+% \LaTeXe{} and are loaded all as one with the usual |\usepackage{expl3}| or
+% |\RequirePackage{expl3}| instructions. These modules will also form the
+% basis of the \LaTeX3 format, but work in this area is incomplete and not
+% included in this documentation at present.
+%
+% As the modules use a coding syntax different from standard
+% \LaTeXe{} it provides a few functions for setting it up.
+%
+% \begin{function}{\ExplSyntaxOn, \ExplSyntaxOff}
+%   \begin{syntax}
+%     \cs{ExplSyntaxOn} \meta{code}  \cs{ExplSyntaxOff}
+%   \end{syntax}
+%   The \cs{ExplSyntaxOn} function switches to a category code
+%   regim{\'e} in which spaces are ignored and in which the colon (|:|)
+%   and underscore (|_|) are treated as \enquote{letters}, thus allowing
+%   access to the names of code functions and variables. Within this
+%   environment, |~| is used to input a space. The \cs{ExplSyntaxOff}
+%   reverts to the document category code regim{\'e}.
+% \end{function}
+%
+% \begin{function}{\ExplSyntaxNamesOn, \ExplSyntaxNamesOff}
+%   \begin{syntax}
+%     \cs{ExplSyntaxNamesOn} \meta{code}  \cs{ExplSyntaxNamesOff}
+%   \end{syntax}
+%   The \cs{ExplSyntaxOn} function switches to a category code
+%   regim{\'e} in which the colon (|:|) and underscore (|_|) are treated as
+%   \enquote{letters}, thus allowing access to the names of code functions
+%   and variables. In contrast to \cs{ExplSyntaxOn}, using
+%   \cs{ExplSyntaxNamesOn} does not cause spaces to be ignored.
+%   The \cs{ExplSyntaxNamesOff} reverts to the document category code
+%   regim{\'e}.
+% \end{function}
+%
+% \begin{function}{\ProvidesExplPackage, \ProvidesExplClass, \ProvidesExplFile}
+%   \begin{syntax}
+%     |\RequirePackage{expl3}| \\
+%     \cs{ProvidesExplPackage} \Arg{package} \Arg{date} \Arg{version}
+%     ~~\Arg{description}
+%   \end{syntax}
+%   These functions act broadly in the same way as the \LaTeXe{} kernel
+%   functions \cs{ProvidesPackage}, \cs{ProvidesClass} and
+%   \cs{ProvidesFile}. However, they also implicitly switch
+%   \cs{ExplSyntaxOn} for the remainder of the code with the file. At the
+%   end of the file, \cs{ExplSyntaxOff} will be called to reverse this.
+%   (This is the same concept as \LaTeXe{} provides in turning on
+%   \cs{makeatletter} within package and class code.)
+% \end{function}
+%
+% \begin{function}{\GetIdInfo}
+%   \begin{syntax}
+%     |\RequirePackage{l3names}|
+%     \cs{GetIdInfo} |$Id:| \meta{SVN info field} |$| \Arg{description}
+%   \end{syntax}
+%   Extracts all information from a SVN field. Spaces are not
+%   ignored in these fields. The information pieces are stored in
+%   separate control sequences with \cs{ExplFileName} for the part of the
+%   file name leading up to the period, \cs{ExplFileDate} for date,
+%   \cs{ExplFileVersion} for version and \cs{ExplFileDescription} for the
+%   description.
+% \end{function}
+%
+% To summarize: Every single package using this syntax should identify
+% itself using one of the above methods. Special care is taken so that
+% every package or class file loaded with \cs{RequirePackage} or alike
+% are loaded with usual \LaTeXe{} category codes and the \LaTeX3 category code
+% scheme is reloaded when needed afterwards. See implementation for
+% details. If you use the \cs{GetIdInfo} command you can use the
+% information when loading a package with
+% \begin{verbatim}
+%   \ProvidesExplPackage{\ExplFileName}{\ExplFileDate}{\ExplFileVersion}{\ExplFileDescription}
+% \end{verbatim}
+%
+% \end{documentation}
+%
+% \begin{implementation}
+%
+% \section{Bootstrap code}
+%
+%    \begin{macrocode}
+%<*initex|package>
+%    \end{macrocode}
+%
+% \subsection{Format-specific code}
+%
+% The very first thing to do is to bootstrap the \IniTeX{} system so
+% that everything else will actually work. \TeX{} does not start with
+% some pretty basic character codes set up.
+%    \begin{macrocode}
+%<*initex>
+\catcode `\{ = 1 \relax
+\catcode `\} = 2 \relax
+\catcode `\# = 6 \relax
+\catcode `\^ = 7 \relax
+%</initex>
+%    \end{macrocode}
+%
+% Tab characters should not show up in the code, but to be on the
+% safe side.
+%    \begin{macrocode}
+%<*initex>
+\catcode `\^^I = 10 \relax
+%</initex>
+%    \end{macrocode}
+%
+% For \LuaTeX{} the extra primitives need to be enabled before
+% they can be use. No \cs{ifdefined} yet, so do it the old-fashioned
+% way. The primitive \cs{strcmp} is simulated using some \Lua{}
+% code, which currently has to be applied to every job as the \Lua{}
+% code is not part of the format. Thanks to Taco Hoekwater for this
+% code. The odd \cs{csname} business is needed so that the later
+% deletion code will work.
+%    \begin{macrocode}
+%<*initex>
+\begingroup\expandafter\expandafter\expandafter\endgroup
+\expandafter\ifx\csname directlua\endcsname\relax
+\else
+  \directlua
+    {
+      tex.enableprimitives('',tex.extraprimitives ())
+      lua.bytecode[1] = function ()
+        function strcmp (A, B)
+          if A == B then
+            tex.write("0")
+          elseif A < B then
+            tex.write("-1")
+          else
+            tex.write("1")
+          end
+        end
+      end
+      lua.bytecode[1]()
+    }
+  \everyjob\expandafter
+    {\csname\detokenize{luatex_directlua:D}\endcsname{lua.bytecode[1]()}}
+  \long\edef\pdfstrcmp#1#2%
+     {%
+       \expandafter\noexpand\csname\detokenize{luatex_directlua:D}\endcsname
+         {%
+           strcmp%
+             (%
+               "\noexpand\luaescapestring{#1}",%
+               "\noexpand\luaescapestring{#2}"%
+             )%
+         }%
+     }
+\fi
+%</initex>
+%    \end{macrocode}
+%
+% \subsection{Package-specific code}
+%
+% The package starts by identifying itself: the information itself is
+% taken from the SVN |Id| string at the start of the source file.
+%    \begin{macrocode}
+%<*package>
+\ProvidesPackage{l3bootstrap}
+  [%
+    \ExplFileDate\space v\ExplFileVersion\space
+    L3 Experimental bootstrap code%
+  ]
+%</package>
+%    \end{macrocode}
+%
+% For \LuaTeX{} the functionality of the \cs{pdfstrcmp} primitive
+% needs to be provided: the \pkg{pdftexmcds} package is used to
+% do this if necessary. At present, there is also a need to deal
+% with some low-level allocation stuff that could usefully be added
+% to \texttt{lualatex.ini}. As it is currently not, load Heiko
+% Oberdiek's \pkg{luatex} package instead.
+%    \begin{macrocode}
+%<*package>
+\def\@tempa%
+  {%
+    \def\@tempa{}%
+    \RequirePackage{luatex}%
+    \RequirePackage{pdftexcmds}%
+    \let\pdfstrcmp\pdf@strcmp
+  }
+\begingroup\expandafter\expandafter\expandafter\endgroup
+\expandafter\ifx\csname directlua\endcsname\relax
+\else
+  \expandafter\@tempa
+\fi
+%</package>
+%    \end{macrocode}
+%
+% \begin{macro}{\ExplSyntaxOff, \ExplSyntaxOn}
+%   Experimental syntax switching is set up here for the package-loading
+%   process.  These are redefined in \pkg{expl3} for the package and in
+%   \pkg{l3final} for the format.
+%    \begin{macrocode}
+%<*package>
+\protected\edef\ExplSyntaxOff
+  {%
+    \catcode   9 = \the\catcode   9\relax
+    \catcode  32 = \the\catcode  32\relax
+    \catcode  34 = \the\catcode  34\relax
+    \catcode  38 = \the\catcode  38\relax
+    \catcode  58 = \the\catcode  58\relax
+    \catcode  94 = \the\catcode  94\relax
+    \catcode  95 = \the\catcode  95\relax
+    \catcode 124 = \the\catcode 124\relax
+    \catcode 126 = \the\catcode 126\relax
+    \endlinechar = \the\endlinechar\relax
+    \chardef\csname\detokenize{l_expl_status_bool}\endcsname = 0 \relax
+  }
+\protected\edef\ExplSyntaxOn
+  {
+    \catcode 9   = 9  \relax
+    \catcode 32  = 9  \relax
+    \catcode 34  = 12 \relax
+    \catcode 58  = 11 \relax
+    \catcode 94  = 7  \relax
+    \catcode 95  = 11 \relax
+    \catcode 124 = 12 \relax
+    \catcode 126 = 10 \relax
+    \endlinechar = 32 \relax
+    \chardef\csname\detokenize{l_expl_status_bool}\endcsname = 1 \relax
+  }
+%</package>
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\l_expl_status_bool}
+%   The status for experimental code syntax: this is off at present. This
+%   code is used by both the package and the format.
+%    \begin{macrocode}
+\expandafter\chardef\csname\detokenize{l_expl_status_bool}\endcsname = 0 \relax
+%    \end{macrocode}
+%\end{macro}
+%
+% \subsection{Dealing with package-mode meta-data}
+%
+% \begin{macro}{\GetIdInfo}
+% \begin{macro}[aux]
+%   {
+%     \GetIdInfoFull, \GetIdInfoAuxI, \GetIdInfoAuxII, \GetIdInfoAuxIII,
+%     \GetIdInfoAuxCVS, \GetIdInfoAuxSVN
+%   }
+%   Functions for collecting up meta-data from the SVN information
+%   used by the \LaTeX3 Project.
+%    \begin{macrocode}
+%<*package>
+\protected\def\GetIdInfo
+  {
+    \begingroup
+    \catcode 32 = 10 \relax
+    \GetIdInfoAuxI
+  }
+\protected\def\GetIdInfoAuxI$#1$#2%
+  {
+    \def\tempa{#1}%
+    \def\tempb{Id}%
+    \ifx\tempa\tempb
+      \def\tempa
+        {%
+          \endgroup
+          \def\ExplFileDate{0000/00/00}%
+          \def\ExplFileDescription{#2}%
+          \def\ExplFileName{[unknown~name]}%
+          \def\ExplFileVersion{000}%
+        }%
+    \else
+      \def\tempa
+        {%
+          \endgroup
+          \GetIdInfoAuxII$#1${#2}%
+        }%
+    \fi
+    \tempa
+  }
+\protected\def\GetIdInfoAuxII$#1 #2.#3 #4 #5 #6 #7 #8$#9%
+  {%
+    \def\ExplFileName{#2}%
+    \def\ExplFileVersion{#4}%
+    \def\ExplFileDescription{#9}%
+    \GetIdInfoAuxIII#5\relax#3\relax#5\relax#6\relax
+  }
+\protected\def\GetIdInfoAuxIII#1#2#3#4#5#6\relax
+  {%
+    \ifx#5/%
+      \expandafter\GetIdInfoAuxCVS
+    \else
+      \expandafter\GetIdInfoAuxSVN
+    \fi
+  }
+\protected\def\GetIdInfoAuxCVS#1,v\relax#2\relax#3\relax
+  {\def\ExplFileDate{#2}}
+\protected\def\GetIdInfoAuxSVN#1\relax#2-#3-#4\relax#5Z\relax
+  {\def\ExplFileDate{#2/#3/#4}}
+%</package>
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\ProvidesExplPackage, \ProvidesExplClass, \ProvidesExplFile}
+%   For other packages and classes building on this one it is convenient
+%   not to need \cs{ExplSyntaxOn} each time.
+%    \begin{macrocode}
+%<*package>
+\protected\def\ProvidesExplPackage#1#2#3#4%
+  {%
+    \ProvidesPackage{#1}[#2 v#3 #4]%
+    \ExplSyntaxOn
+  }
+\protected\def\ProvidesExplClass#1#2#3#4%
+  {%
+    \ProvidesClass{#1}[#2 v#3 #4]%
+    \ExplSyntaxOn
+  }
+\protected\def\ProvidesExplFile#1#2#3#4%
+  {%
+    \ProvidesFile{#1}[#2 v#3 #4]%
+    \ExplSyntaxOn
+  }
+%</package>
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\@pushfilename, \@popfilename}
+%   The idea here is to use \LaTeXe{}'s \cs{@pushfilename} and
+%   \cs{@popfilename} to track the current syntax status. This can be
+%   achieved by saving the current status flag at each push to a stack,
+%   then recovering it at the pop stage and checking if the code
+%   environment should still be active.
+%    \begin{macrocode}
+%<*package>
+\edef\@pushfilename
+  {%
+    \edef\expandafter\noexpand
+      \csname\detokenize{l_expl_status_stack_tl}\endcsname
+      {%
+        \noexpand\ifodd\expandafter\noexpand
+          \csname\detokenize{l_expl_status_bool}\endcsname
+          1%
+        \noexpand\else
+          0%
+        \noexpand\fi
+        \expandafter\noexpand
+          \csname\detokenize{l_expl_status_stack_tl}\endcsname
+      }%
+    \ExplSyntaxOff
+    \unexpanded\expandafter{\@pushfilename}%
+  }
+\edef\@popfilename
+  {%
+    \unexpanded\expandafter{\@popfilename}%
+    \noexpand\if a\expandafter\noexpand\csname
+      \detokenize{l_expl_status_stack_tl}\endcsname a%
+      \ExplSyntaxOff
+    \noexpand\else
+      \noexpand\expandafter
+        \expandafter\noexpand\csname
+          \detokenize{expl_status_pop:w}\endcsname
+          \expandafter\noexpand\csname
+            \detokenize{l_expl_status_stack_tl}\endcsname
+          \noexpand\@nil
+   \noexpand\fi
+  }
+%</package>
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\l_expl_status_stack_tl}
+%   As \pkg{expl3} itself cannot be loaded with the code environment
+%   already active, at the end of the package \cs{ExplSyntaxOff} can
+%   safely be called.
+%    \begin{macrocode}
+%<*package>
+\@namedef{\detokenize{l_expl_status_stack_tl}}{0}
+%</package>
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}[aux]{\expl_status_pop:w}
+%   The pop auxiliary function removes the first item from the stack,
+%   saves the rest of the stack and then does the test. As
+%   \cs{ExplSyntaxOff} is already defined as a protected macro, there
+%   is no need for \cs{noexpand} here.
+%    \begin{macrocode}
+%<*package>
+\expandafter\edef\csname\detokenize{expl_status_pop:w}\endcsname#1#2\@nil
+  {%
+    \def\expandafter\noexpand
+      \csname\detokenize{l_expl_status_stack_tl}\endcsname{#2}%
+    \noexpand\ifodd#1\space
+      \noexpand\expandafter\noexpand\ExplSyntaxOn
+    \noexpand\else
+      \noexpand\expandafter\ExplSyntaxOff
+    \noexpand\fi
+  }
+%</package>
+%    \end{macrocode}
+% \end{macro}
+%
+% We want the \pkg{expl3} bundle to be loaded \enquote{as one};
+% this command is
+% used to ensure that one of the |l3| packages isn't loaded on its own.
+%    \begin{macrocode}
+%<*package>
+\expandafter\protected\expandafter\def
+  \csname\detokenize{package_check_loaded_expl:}\endcsname
+  {%
+    \@ifpackageloaded{expl3}
+      {}
+      {%
+        \PackageError{expl3}
+          {Cannot load the expl3 modules separately}
+          {%
+            The expl3 modules cannot be loaded separately;\MessageBreak
+            please \string\usepackage\string{expl3\string} instead.
+          }%
+      }%
+  }
+%</package>
+%    \end{macrocode}
+%
+% \subsection{The \cs{pdfstrcmp} primitive in \XeTeX{}}
+%
+% Only \pdfTeX{} has a primitive called \cs{pdfstrcmp}. The \XeTeX{}
+% version is just \cs{strcmp}, so there is some shuffling to do.
+%    \begin{macrocode}
+\begingroup\expandafter\expandafter\expandafter\endgroup
+  \expandafter\ifx\csname pdfstrcmp\endcsname\relax
+  \let\pdfstrcmp\strcmp
+\fi
+%    \end{macrocode}
+%
+% \subsection{Engine requirements}
+%
+% The code currently requires functionality equivalent to \cs{pdfstrcmp}
+% in addition to \eTeX{}. The former is therefore used as a test for
+% a suitable engine.
+%    \begin{macrocode}
+\begingroup\expandafter\expandafter\expandafter\endgroup
+\expandafter\ifx\csname pdfstrcmp\endcsname\relax
+%<*package>
+  \PackageError{l3names}{Required primitive not found: \protect\pdfstrcmp}
+    {%
+      LaTeX3 requires the e-TeX primitives and
+        \string\pdfstrcmp.\MessageBreak
+      These are available in engine versions: \MessageBreak
+      - pdfTeX 1.30 \MessageBreak
+      - XeTeX 0.9994 \MessageBreak
+      - LuaTeX 0.60 \MessageBreak
+      or later. \MessageBreak
+      \MessageBreak
+      Loading of expl3 will abort!
+    }
+%</package>
+%<*initex>
+  \newlinechar`\^^J\relax
+  \errhelp{%
+      LaTeX3 requires the e-TeX primitives and
+        \string\pdfstrcmp. ^^J
+      These are available in engine versions:  ^^J
+      - pdfTeX 1.30  ^^J
+      - XeTeX 0.9994 ^^J
+      - LuaTeX 0.60  ^^J
+      or later. ^^J
+      For pdfTeX and XeTeX the '-etex' command-line switch is also
+      needed.^^J
+       ^^J
+      Format building will abort!
+  }
+%</initex>
+  \expandafter\endinput
+\fi
+%    \end{macrocode}
+%
+% \subsection{The \LaTeX3 code environment}
+%
+% \begin{macro}{\ExplSyntaxNamesOn, \ExplSyntaxNamesOff}
+%   These can be set up early, as they are not used anywhere in the
+%   package or format itself. Using an \cs{edef} here makes the
+%   definitions that bit clearer later.
+%    \begin{macrocode}
+\protected\edef\ExplSyntaxNamesOn
+  {%
+    \expandafter\noexpand
+      \csname\detokenize{char_set_catcode_letter:n}\endcsname{58}%
+    \expandafter\noexpand
+      \csname\detokenize{char_set_catcode_letter:n}\endcsname{95}%
+  }
+\protected\edef\ExplSyntaxNamesOff
+  {%
+    \expandafter\noexpand
+      \csname\detokenize{char_set_catcode_math_subscript:n}\endcsname{58}%
+    \expandafter\noexpand
+      \csname\detokenize{char_set_catcode_other:n}\endcsname{95}%
+  }
+%    \end{macrocode}
+% \end{macro}
+%
+% The code environment is now set up for the format: the package deals
+% with this using \cs{ProvidesExplPackage}.
+%    \begin{macrocode}
+%<*initex>
+\catcode 9   = 9  \relax
+\catcode 32  = 9  \relax
+\catcode 34  = 12 \relax
+\catcode 58  = 11 \relax
+\catcode 94  = 7  \relax
+\catcode 95  = 11 \relax
+\catcode 124 = 12 \relax
+\catcode 126 = 10 \relax
+\endlinechar = 32 \relax
+%</initex>
+%    \end{macrocode}
+%    
+% \begin{macro}{\ExplSyntaxOn, \ExplSyntaxOff}
+%  The idea here is that multiple \cs{ExplSyntaxOn} calls are not
+%  going to mess up category codes, and that multiple calls to
+%  \cs{ExplSyntaxOff} are also not wasting time.
+%    \begin{macrocode}
+%<*initex>
+\protected \def \ExplSyntaxOn
+  {
+    \bool_if:NF \l_expl_status_bool
+      {
+        \cs_set_protected_nopar:Npx \ExplSyntaxOff
+          {
+            \char_set_catcode:nn { 9 }   { \char_value_catcode:n { 9 } }
+            \char_set_catcode:nn { 32 }  { \char_value_catcode:n { 32 } }
+            \char_set_catcode:nn { 34 }  { \char_value_catcode:n { 34 } }
+            \char_set_catcode:nn { 38 }  { \char_value_catcode:n { 38 } }
+            \char_set_catcode:nn { 58 }  { \char_value_catcode:n { 58 } }
+            \char_set_catcode:nn { 94 }  { \char_value_catcode:n { 94 } }
+            \char_set_catcode:nn { 95 }  { \char_value_catcode:n { 95 } }
+            \char_set_catcode:nn { 124 } { \char_value_catcode:n { 124 } }
+            \char_set_catcode:nn { 126 } { \char_value_catcode:n { 126 } }
+            \tex_endlinechar:D = 
+              \tex_the:D \tex_endlinechar:D \scan_stop:
+            \bool_set_false:N \l_expl_status_bool
+            \cs_set_protected_nopar:Npn \ExplSyntaxOff { }
+          }
+      }
+    \char_set_catcode_ignore:n           { 9 }   % tab
+    \char_set_catcode_ignore:n           { 32 }  % space
+    \char_set_catcode_other:n            { 34 }  % double quote
+    \char_set_catcode_alignment:n        { 38 }  % ampersand
+    \char_set_catcode_letter:n           { 58 }  % colon
+    \char_set_catcode_math_superscript:n { 94 }  % circumflex
+    \char_set_catcode_letter:n           { 95 }  % underscore
+    \char_set_catcode_other:n            { 124 } % pipe
+    \char_set_catcode_space:n            { 126 } % tilde  
+    \tex_endlinechar:D = 32 \scan_stop:
+    \bool_set_true:N \l_expl_status_bool
+  }
+\protected \def \ExplSyntaxOff { }
+%</initex>
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{variable}{\l_expl_status_bool}
+%   A flag to show the current syntax status. 
+%    \begin{macrocode}
+%<*initex>
+\chardef \l_expl_status_bool = 0 ~
+%</initex>
+%    \end{macrocode}
+% \end{variable}
+%
+%    \begin{macrocode}
+%</initex|package>
+%    \end{macrocode}
+%
+% \end{implementation}
+%
+% \PrintIndex
\ No newline at end of file
diff --git a/Master/texmf-dist/source/latex/l3kernel/l3box.dtx b/Master/texmf-dist/source/latex/l3kernel/l3box.dtx
new file mode 100644
index 00000000000..1947f85b356
--- /dev/null
+++ b/Master/texmf-dist/source/latex/l3kernel/l3box.dtx
@@ -0,0 +1,1237 @@
+% \iffalse meta-comment
+%
+%% File: l3box.dtx Copyright (C) 2005-2011 The LaTeX3 Project
+%%
+%% It may be distributed and/or modified under the conditions of the
+%% LaTeX Project Public License (LPPL), either version 1.3c of this
+%% license or (at your option) any later version.  The latest version
+%% of this license is in the file
+%%
+%%    http://www.latex-project.org/lppl.txt
+%%
+%% This file is part of the "expl3 bundle" (The Work in LPPL)
+%% and all files in that bundle must be distributed together.
+%%
+%% The released version of this bundle is available from CTAN.
+%%
+%% -----------------------------------------------------------------------
+%%
+%% The development version of the bundle can be found at
+%%
+%%    http://www.latex-project.org/svnroot/experimental/trunk/
+%%
+%% for those people who are interested.
+%%
+%%%%%%%%%%%
+%% NOTE: %%
+%%%%%%%%%%%
+%%
+%%   Snapshots taken from the repository represent work in progress and may
+%%   not work or may contain conflicting material!  We therefore ask
+%%   people _not_ to put them into distributions, archives, etc. without
+%%   prior consultation with the LaTeX3 Project.
+%%
+%% -----------------------------------------------------------------------
+%
+%<*driver|package>
+\RequirePackage{l3names}
+\GetIdInfo$Id: l3box.dtx 2478 2011-06-19 21:34:23Z joseph $
+  {L3 Experimental boxes}
+%</driver|package>
+%<*driver>
+\documentclass[full]{l3doc}
+\begin{document}
+  \DocInput{\jobname.dtx}
+\end{document}
+%</driver>
+% \fi
+%
+% \title{^^A
+%   The \pkg{l3box} package\\ Boxes^^A
+%   \thanks{This file describes v\ExplFileVersion,
+%      last revised \ExplFileDate.}^^A
+% }
+%
+% \author{^^A
+%  The \LaTeX3 Project\thanks
+%    {^^A
+%      E-mail:
+%        \href{mailto:latex-team@latex-project.org}
+%          {latex-team@latex-project.org}^^A
+%    }^^A
+% }
+%
+% \date{Released \ExplFileDate}
+%
+% \maketitle
+%
+% \begin{documentation}
+%
+% There are three kinds of box operations: horizontal mode denoted
+% with prefix |\hbox_|, vertical mode with prefix |\vbox_|, and the
+% generic operations working in both modes with prefix |\box_|.
+%
+% \section{Creating and initialising boxes}
+%
+% \begin{function}{\box_new:N, \box_new:c}
+%   \begin{syntax}
+%     \cs{box_new:N} \meta{box}
+%   \end{syntax}
+%   Creates a new \meta{box} or raises an error if the name is
+%   already taken. The declaration is global. The \meta{box} will
+%   initially be void.
+% \end{function}
+%
+% \begin{function}{\box_clear:N, \box_clear:c}
+%   \begin{syntax}
+%     \cs{box_clear:N} \meta{box}
+%   \end{syntax}
+%   Clears the content of the \meta{box} by setting the box equal to
+%   \cs{c_void_box} within the current \TeX{} group level.
+% \end{function}
+%
+% \begin{function}{\box_gclear:N, \box_gclear:c}
+%   \begin{syntax}
+%     \cs{box_gclear:N} \meta{box}
+%   \end{syntax}
+%   Clears the content of the \meta{box} by setting the box equal to
+%   \cs{c_void_box} globally.
+% \end{function}
+%
+% \begin{function}{\box_clear_new:N, \box_clear_new:c}
+%   \begin{syntax}
+%     \cs{box_clear_new:N} \meta{box}
+%   \end{syntax}
+%   If the \meta{box} is not defined, globally creates it. If the \meta{box}
+%   is defined,
+%   clears the content of the \meta{box} by setting the box equal to
+%   \cs{c_void_box} within the current \TeX{} group level.
+% \end{function}
+%
+% \begin{function}{\box_gclear_new:N, \box_gclear_new:c}
+%   \begin{syntax}
+%     \cs{box_gclear_new:N} \meta{box}
+%   \end{syntax}
+%   If the \meta{box} is not defined, globally creates it. If the \meta{box}
+%   is defined,
+%   clears the content of the \meta{box} by setting the box equal to
+%   \cs{c_void_box} globally.
+% \end{function}
+%
+% \begin{function}
+%   {\box_set_eq:NN, \box_set_eq:cN, \box_set_eq:Nc, \box_set_eq:cc}
+%   \begin{syntax}
+%     \cs{box_set_eq:NN} \meta{box1} \meta{box2}
+%   \end{syntax}
+%   Sets the content of \meta{box1} equal to that of \meta{box2}. This
+%   assignment is restricted to the current \TeX{} group level.
+% \end{function}
+%
+% \begin{function}
+%   {\box_gset_eq:NN, \box_gset_eq:cN, \box_gset_eq:Nc, \box_gset_eq:cc}
+%   \begin{syntax}
+%     \cs{box_gset_eq:NN} \meta{box1} \meta{box2}
+%   \end{syntax}
+%   Sets the content of \meta{box1} equal to that of \meta{box2} globally.
+% \end{function}
+%
+% \begin{function}
+%   {
+%     \box_set_eq_clear:NN, \box_set_eq_clear:cN,
+%     \box_set_eq_clear:Nc, \box_set_eq_clear:cc
+%   }
+%   \begin{syntax}
+%     \cs{box_set_eq_clear:NN} \meta{box1} \meta{box2}
+%   \end{syntax}
+%   Sets the content of \meta{box1} within the current \TeX{} group
+%   equal to that of \meta{box2}, then clears \meta{box2} globally.
+% \end{function}
+%
+% \begin{function}
+%   {
+%     \box_gset_eq_clear:NN, \box_gset_eq_clear:cN,
+%     \box_gset_eq_clear:Nc, \box_gset_eq_clear:cc
+%   }
+%   \begin{syntax}
+%     \cs{box_gset_eq_clear:NN} \meta{box1} \meta{box2}
+%   \end{syntax}
+%   Sets the content of \meta{box1} equal to that of \meta{box2}, then
+%   clears \meta{box2}. These assignments are global.
+% \end{function}
+%
+% \section{Using boxes}
+%
+% \begin{function}{\box_use:N, \box_use:c}
+%   \begin{syntax}
+%     \cs{box_use:N} \meta{box}
+%   \end{syntax}
+%   Inserts the current content of the \meta{box} onto the current
+%   list for typesetting.
+%   \begin{texnote}
+%     This is the \TeX{} primitive \cs{copy}.
+%   \end{texnote}
+% \end{function}
+%
+% \begin{function}{\box_use_clear:N, \box_use_clear:c}
+%   \begin{syntax}
+%     \cs{box_use_clear:N} \meta{box}
+%   \end{syntax}
+%   Inserts the current content of the \meta{box} onto the current
+%   list for typesetting, then globally clears the content of the
+%   \meta{box}.
+%   \begin{texnote}
+%     This is the \TeX{} primitive \cs{box}.
+%   \end{texnote}
+% \end{function}
+%
+% \begin{function}{\box_move_right:nn, \box_move_left:nn}
+%   \begin{syntax}
+%     \cs{box_move_right:nn} \Arg{dimexpr} \Arg{box function}
+%   \end{syntax}
+%   This function operates in vertical mode, and inserts the
+%   material specified by the \meta{box function}
+%   such that its reference point is displaced horizontally by the given
+%   \meta{dimexpr} from the reference point for typesetting, to the right
+%   or left as appropriate. The \meta{box function} should be
+%   a box opeeration such as |\box_use:N \<box>| or a \enquote{raw}
+%   box specification such as |\vbox:n { xyz }|.
+% \end{function}
+%
+% \begin{function}{\box_move_up:nn, \box_move_down:nn}
+%   \begin{syntax}
+%     \cs{box_move_up:nn} \Arg{dimexpr} \Arg{box function}
+%   \end{syntax}
+%   This function operates in horizontal mode, and inserts the
+%   material specified by the \meta{box function}
+%   such that its reference point is displaced vertical by the given
+%   \meta{dimexpr} from the reference point for typesetting, up
+%   or down as appropriate. The \meta{box function} should be
+%   a box opeeration such as |\box_use:N \<box>| or a \enquote{raw}
+%   box specification such as |\vbox:n { xyz }|.
+% \end{function}
+%
+% \section{Measuring and setting box dimensions}
+%
+% \begin{function}{\box_dp:N, \box_dp:c}
+%   \begin{syntax}
+%     \cs{box_dp:N} \meta{box}
+%   \end{syntax}
+%   Calculates the depth (below the baseline) of the \meta{box}
+%   and leaves this in the input stream. The output of this function
+%   is suitable for use in a \meta{dimension expression} for
+%   calculations.
+%   \begin{texnote}
+%     This is the \TeX{} primitive \cs{dp}.
+%   \end{texnote}
+% \end{function}
+%
+% \begin{function}{\box_ht:N, \box_ht:c}
+%   \begin{syntax}
+%     \cs{box_ht:N} \meta{box}
+%   \end{syntax}
+%   Calculates the height (above the baseline) of the \meta{box}
+%   and leaves this in the input stream. The output of this function
+%   is suitable for use in a \meta{dimension expression} for
+%   calculations.
+%   \begin{texnote}
+%     This is the \TeX{} primitive \cs{ht}.
+%   \end{texnote}
+% \end{function}
+%
+% \begin{function}{\box_wd:N, \box_wd:c}
+%   \begin{syntax}
+%     \cs{box_wd:N} \meta{box}
+%   \end{syntax}
+%   Calculates the width of the \meta{box} and leaves this in the
+%   input stream. The output of this function is suitable for use in a
+%   \meta{dimension expression} for calculations.
+%   \begin{texnote}
+%     This is the \TeX{} primitive \cs{wd}.
+%   \end{texnote}
+% \end{function}
+%
+% \begin{function}{\box_set_dp:Nn, \box_set_dp:cn}
+%   \begin{syntax}
+%     \cs{box_set_dp:Nn} \meta{box} \Arg{dimension expression}
+%   \end{syntax}
+%   Set the depth (below the baseline) of the \meta{box} to the value of
+%   the \Arg{dimension expression}. This is a global assignment.
+% \end{function}
+%
+% \begin{function}{\box_set_ht:Nn, \box_set_ht:cn}
+%   \begin{syntax}
+%     \cs{box_set_ht:Nn} \meta{box} \Arg{dimension expression}
+%   \end{syntax}
+%   Set the height (above the baseline) of the \meta{box} to the value of
+%   the \Arg{dimension expression}. This is a global assignment.
+% \end{function}
+%
+% \begin{function}{\box_set_wd:Nn, \box_set_wd:cn}
+%   \begin{syntax}
+%     \cs{box_set_wd:Nn} \meta{box} \Arg{dimension expression}
+%   \end{syntax}
+%   Set the width of the \meta{box} to the value of the
+%   \Arg{dimension expression}. This is a global assignment.
+% \end{function}
+%
+% \section{Box conditionals}
+%
+% \begin{function}[EXP,pTF]{\box_if_empty:N, \box_if_empty:c}
+%   \begin{syntax}
+%     \cs{box_if_empty_p:N} \meta{box}
+%     \cs{box_if_empty:NTF} \meta{box} \Arg{true code} \Arg{false code}
+%   \end{syntax}
+%   Tests if \meta{box} is a empty (equal to \cs{c_empty_box}).
+%   The branching versions then leave either
+%   \meta{true code} or \meta{false code} in the input stream, as
+%   appropriate to the truth of the test and the variant of the function
+%   chosen. The logical truth of the test is left in the input stream by
+%   the predicate version.
+% \end{function}
+%
+% \begin{function}[EXP,pTF]{\box_if_horizontal:N, \box_if_horizontal:c}
+%   \begin{syntax}
+%     \cs{box_if_horizontal_p:N} \meta{box}
+%     \cs{box_if_horizontal:NTF} \meta{box} \Arg{true code} \Arg{false code}
+%   \end{syntax}
+%   Tests if \meta{box} is a horizontal box. The branching
+%   versions then leave either \meta{true code} or \meta{false code}
+%   in the input stream, as appropriate to the truth of the test and
+%   the variant of the function chosen. The logical truth of the test
+%   is left in the input stream by the predicate version.
+% \end{function}
+%
+% \begin{function}[EXP,pTF]{\box_if_vertical:N, \box_if_vertical:c}
+%   \begin{syntax}
+%     \cs{box_if_vertical_p:N} \meta{box}
+%     \cs{box_if_vertical:NTF} \meta{box} \Arg{true code} \Arg{false code}
+%   \end{syntax}
+%   Tests if \meta{box} is a vertical box. The branching
+%   versions then leave either \meta{true code} or \meta{false code}
+%   in the input stream, as appropriate to the truth of the test and
+%   the variant of the function chosen. The logical truth of the test
+%   is left in the input stream by the predicate version.
+% \end{function}
+%
+% \section{The last box inserted}
+%
+% \begin{variable}{\l_last_box}
+%   This is a box containing the last item added to the current partial
+%   list, except in the case of the main vertical list (main galley), in
+%   which case this box is always void. Notice that although this is
+%   not a constant, it is \emph{not} settable by the programmer but is
+%   instead varied by \TeX{}.
+%   \begin{texnote}
+%     This is the \TeX{} primitive \cs{lastbox} renamed.
+%   \end{texnote}
+% \end{variable}
+%
+% \section{Constant boxes}
+%
+%  \begin{variable}{\c_empty_box}
+%    This is a permanently empty box, which is neither set as horizontal
+%    nor vertical.
+%  \end{variable}
+%
+%  \section{Scratch boxes}
+%
+% \begin{variable}{\l_tmpa_tl, \l_tmpb_tl}
+%   Scratch boxes for local assignment. These are never used by
+%   the kernel code, and so are safe for use with any \LaTeX3-defined
+%   function. However, they may be overwritten by other non-kernel
+%   code and so should only be used for short-term storage.
+% \end{variable}
+%
+% \section{Viewing box contents}
+%
+% \begin{function}{\box_show:N, \box_show:c}
+%   \begin{syntax}
+%      \cs{box_show:N} \meta{box}
+%   \end{syntax}
+%   Writes the contents of \meta{box} to the log file.
+%   \begin{texnote}
+%     This is the \TeX{} primitive \cs{showbox}.
+%   \end{texnote}
+% \end{function}
+%
+% \section{Horizontal mode boxes}
+%
+% \begin{function}{\hbox:n}
+%   \begin{syntax}
+%     \cs{hbox:n} \Arg{contents}
+%   \end{syntax}
+%   Typesets the \meta{contents} into a horizontal box of natural
+%   width and then includes this box in the current list for typesetting.
+%   \begin{texnote}
+%     This is the \TeX{} primitive \cs{hbox}.
+%   \end{texnote}
+% \end{function}
+%
+% \begin{function}{\hbox_to_wd:nn}
+%   \begin{syntax}
+%     \cs{hbox_to_wd:nn} \Arg{dimexpr} \Arg{contents}
+%   \end{syntax}
+%   Typesets the \meta{contents} into a horizontal box of width
+%   \meta{dimexpr} and then includes this box in the current list for
+%   typesetting.
+% \end{function}
+%
+% \begin{function}{\hbox_to_zero:n}
+%   \begin{syntax}
+%     \cs{hbox_to_zero:n} \Arg{contents}
+%   \end{syntax}
+%   Typesets the \meta{contents} into a horizontal box of zero width
+%   and then includes this box in the current list for typesetting.
+% \end{function}
+%
+% \begin{function}{\hbox_set:Nn, \hbox_set:cn}
+%   \begin{syntax}
+%     \cs{hbox_set:Nn} \meta{box} \Arg{contents}
+%   \end{syntax}
+%   Typesets the \meta{contents} at natural width and then stores the
+%   result inside the \meta{box}. The assignment is local.
+% \end{function}
+%
+% \begin{function}{\hbox_gset:Nn, \hbox_gset:cn}
+%   \begin{syntax}
+%     \cs{hbox_gset:Nn} \meta{box} \Arg{contents}
+%   \end{syntax}
+%   Typesets the \meta{contents} at natural width and then stores the
+%   result inside the \meta{box}. The assignment is global.
+% \end{function}
+%
+% \begin{function}{\hbox_set_to_wd:Nnn, \hbox_set_to_wd:cnn}
+%   \begin{syntax}
+%     \cs{hbox_set_to_wd:Nnn} \meta{box} \Arg{dimexpr} \Arg{contents}
+%   \end{syntax}
+%   Typesets the \meta{contents} to the width given by the \meta{dimexpr}
+%   and then stores the result inside the \meta{box}. The assignment is
+%   local.
+% \end{function}
+%
+% \begin{function}{\hbox_gset_to_wd:Nnn, \hbox_gset_to_wd:cnn}
+%   \begin{syntax}
+%     \cs{hbox_gset_to_wd:Nnn} \meta{box} \Arg{dimexpr} \Arg{contents}
+%   \end{syntax}
+%   Typesets the \meta{contents} to the width given by the \meta{dimexpr}
+%   and then stores the result inside the \meta{box}. The assignment is
+%   global.
+% \end{function}
+%
+% \begin{function}{\hbox_overlap_right:n}
+%   \begin{syntax}
+%     \cs{hbox_overlap_right:n} \Arg{contents}
+%   \end{syntax}
+%   Typesets the \meta{contents} into a horizontal box of zero width
+%   such that material will protrude to the right of the insertion
+%   point.
+% \end{function}
+%
+% \begin{function}{\hbox_overlap_left:n}
+%   \begin{syntax}
+%     \cs{hbox_overlap_left:n} \Arg{contents}
+%   \end{syntax}
+%   Typesets the \meta{contents} into a horizontal box of zero width
+%   such that material will protrude to the left of the insertion
+%   point.
+% \end{function}
+%
+% \begin{function}
+%   {
+%     \hbox_set_inline_begin:N, \hbox_set_inline_begin:c,
+%     \hbox_set_inline_end:
+%   }
+%   \begin{syntax}
+%     \cs{hbox_set_inline_begin:N} \meta{box} \meta{contents}
+%     ~~\cs{hbox_set_inline_end:}
+%   \end{syntax}
+%   Typesets the \meta{contents} at natural width and then stores the
+%   result inside the \meta{box}. The assignment is local. In contrast
+%   to \cs{hbox_set:Nn} this function does not absorb the argument
+%   when finding the \meta{content}, and so can be used in circumstances
+%   where the \meta{content} may not be a simple argument.
+% \end{function}
+%
+% \begin{function}
+%   {
+%     \hbox_gset_inline_begin:N, \hbox_gset_inline_begin:c,
+%     \hbox_gset_inline_end:
+%   }
+%   \begin{syntax}
+%     \cs{hbox_gset_inline_begin:N} \meta{box} \meta{contents}
+%     ~~\cs{hbox_gset_inline_end:}
+%   \end{syntax}
+%   Typesets the \meta{contents} at natural width and then stores the
+%   result inside the \meta{box}. The assignment is global. In contrast
+%   to \cs{hbox_set:Nn} this function does not absorb the argument
+%   when finding the \meta{content}, and so can be used in circumstances
+%   where the \meta{content} may not be a simple argument.
+% \end{function}
+%
+% \begin{function}{\hbox_unpack:N, \hbox_unpack:c}
+%   \begin{syntax}
+%     \cs{hbox_unpack:N} \meta{box}
+%   \end{syntax}
+%   Unpacks the content of the horizontal \meta{box}, retaining any stretching
+%   or shrinking applied when the \meta{box} was set.
+%   \begin{texnote}
+%     This is the \TeX{} primitive \cs{unhcopy}.
+%   \end{texnote}
+% \end{function}
+%
+% \begin{function}{\hbox_unpack_clear:N, \hbox_unpack_clear:c}
+%   \begin{syntax}
+%     \cs{hbox_unpack_clear:N} \meta{box}
+%   \end{syntax}
+%   Unpacks the content of the horizontal \meta{box}, retaining any stretching
+%   or shrinking applied when the \meta{box} was set. The \meta{box} is
+%   then cleared globally.
+%   \begin{texnote}
+%     This is the \TeX{} primitive \cs{unhbox}.
+%   \end{texnote}
+% \end{function}
+%
+% \section{Vertical mode boxes}
+%
+% Vertical boxes inherit their baseline from their contents. The
+% standard case is that the baseline of the box is at the same position
+% as that of the last item added to the box. This means that the box
+% will have no depth unless the last item added to it had depth. As a
+% result most vertical boxes have a large height value and small or
+% zero depth. The exception are |_top| boxes, where the reference point
+% is that of the first item added. These tend to have a large depth and
+% small height, although the later will typically be non-zero.
+%
+% \begin{function}{\vbox:n}
+%   \begin{syntax}
+%     \cs{vbox:n} \Arg{contents}
+%   \end{syntax}
+%   Typesets the \meta{contents} into a vertical box of natural height
+%   and includes this box in the current list for typesetting.
+%   \begin{texnote}
+%     This is the \TeX{} primitive \cs{vbox}.
+%   \end{texnote}
+% \end{function}
+%
+% \begin{function}{\vbox_top:n}
+%   \begin{syntax}
+%     \cs{vbox_top:n} \Arg{contents}
+%   \end{syntax}
+%   Typesets the \meta{contents} into a vertical box of natural height
+%   and includes this box in the current list for typesetting. The
+%   baseline of the box will tbe equal to that of the \emph{first}
+%   item added to the box.
+%   \begin{texnote}
+%     This is the \TeX{} primitive \cs{vtop}.
+%   \end{texnote}
+% \end{function}
+%
+% \begin{function}{\vbox_to_ht:nn}
+%   \begin{syntax}
+%     \cs{vbox_to_ht:n} \Arg{contents} \Arg{dimexpr}
+%   \end{syntax}
+%   Typesets the \meta{contents} into a vertical box of height
+%   \meta{dimexpr} and then includes this box in the current list for
+%   typesetting.
+% \end{function}
+%
+% \begin{function}{\vbox_to_zero:n}
+%   \begin{syntax}
+%     \cs{vbox_to_zero:n} \Arg{contents}
+%   \end{syntax}
+%   Typesets the \meta{contents} into a vertical box of zero height
+%   and then includes this box in the current list for typesetting.
+% \end{function}
+%
+%  \begin{function}{\vbox_set:Nn, \vbox_set:cn}
+%   \begin{syntax}
+%     \cs{vbox_set:Nn} \meta{box} \Arg{contents}
+%   \end{syntax}
+%   Typesets the \meta{contents} at natural height and then stores the
+%   result inside the \meta{box}. The assignment is local.
+% \end{function}
+%
+%  \begin{function}{\vbox_gset:Nn, \vbox_gset:cn}
+%   \begin{syntax}
+%     \cs{vbox_gset:Nn} \meta{box} \Arg{contents}
+%   \end{syntax}
+%   Typesets the \meta{contents} at natural height and then stores the
+%   result inside the \meta{box}. The assignment is global.
+% \end{function}
+%
+% \begin{function}{\vbox_set_top:Nn, \vbox_set_top:cn}
+%   \begin{syntax}
+%     \cs{vbox_set_top:Nn} \meta{box} \Arg{contents}
+%   \end{syntax}
+%   Typesets the \meta{contents} at natural height and then stores the
+%   result inside the \meta{box}. The baseline of the box will tbe equal
+%   to that of the \emph{first} item added to the box. The assignment is
+%   local.
+% \end{function}
+%
+% \begin{function}{\vbox_gset_top:Nn, \vbox_gset_top:cn}
+%   \begin{syntax}
+%     \cs{vbox_gset_top:Nn} \meta{box} \Arg{contents}
+%   \end{syntax}
+%   Typesets the \meta{contents} at natural height and then stores the
+%   result inside the \meta{box}. The baseline of the box will tbe equal
+%   to that of the \emph{first} item added to the box. The assignment is
+%   global.
+% \end{function}
+%
+% \begin{function}{\vbox_set_to_ht:Nnn, \vbox_set_to_ht:cnn}
+%   \begin{syntax}
+%     \cs{vbox_set_to_ht:Nnn} \meta{box} \Arg{dimexpr} \Arg{contents}
+%   \end{syntax}
+%   Typesets the \meta{contents} to the height given by the
+%   \meta{dimexpr} and then stores the result inside the \meta{box}.
+%   The assignment is local.
+% \end{function}
+%
+% \begin{function}{\vbox_gset_to_ht:Nnn, \vbox_gset_to_ht:cnn}
+%   \begin{syntax}
+%     \cs{vbox_gset_to_ht:Nnn} \meta{box} \Arg{dimexpr} \Arg{contents}
+%   \end{syntax}
+%   Typesets the \meta{contents} to the height given by the
+%   \meta{dimexpr} and then stores the result inside the \meta{box}.
+%   The assignment is global.
+% \end{function}
+%
+% \begin{function}
+%  {
+%    \vbox_set_inline_begin:N, \vbox_set_inline_begin:c,
+%    \vbox_set_inline_end:
+%  }
+%    \begin{syntax}
+%      \cs{vbox_set_inline_begin:N} \meta{box} \meta{contents}
+%      ~~\cs{vbox_set_inline_end:}
+%    \end{syntax}
+%    Typesets the \meta{contents} at natural height and then stores the
+%    result inside the \meta{box}. The assignment is local. In contrast
+%    to \cs{vbox_set:Nn} this function does not absorb the argument
+%    when finding the \meta{content}, and so can be used in circumstances
+%    where the \meta{content} may not be a simple argument.
+%  \end{function}
+%
+% \begin{function}
+%  {
+%    \vbox_gset_inline_begin:N, \vbox_gset_inline_begin:c,
+%    \vbox_gset_inline_end:
+%  }
+%    \begin{syntax}
+%      \cs{vbox_gset_inline_begin:N} \meta{box} \meta{contents}
+%      ~~\cs{vbox_set_inline_end:}
+%    \end{syntax}
+%    Typesets the \meta{contents} at natural height and then stores the
+%    result inside the \meta{box}. The assignment is global. In contrast
+%    to \cs{vbox_set:Nn} this function does not absorb the argument
+%    when finding the \meta{content}, and so can be used in circumstances
+%    where the \meta{content} may not be a simple argument.
+%  \end{function}
+%
+% \begin{function}{\vbox_set_split_to_ht:NNn}
+%   \begin{syntax}
+%      \cs{vbox_set_split_to_ht:NNn} \meta{box1} \meta{box2} \Arg{dimexpr}
+%   \end{syntax}
+%   Sets \meta{box1} to contain material to the height given by the
+%   \meta{dimexpr} by removing content from the top of \meta{box2}
+%   (which must be a vertical box).
+%   \begin{texnote}
+%     This is the \TeX{} primitive \cs{vsplit}.
+%   \end{texnote}
+% \end{function}
+%
+% \begin{function}{\vbox_unpack:N, \vbox_unpack:c}
+%   \begin{syntax}
+%     \cs{vbox_unpack:N} \meta{box}
+%   \end{syntax}
+%   Unpacks the content of the vertical \meta{box}, retaining any stretching
+%   or shrinking applied when the \meta{box} was set.
+%   \begin{texnote}
+%     This is the \TeX{} primitive \cs{unvcopy}.
+%   \end{texnote}
+% \end{function}
+%
+% \begin{function}{\vbox_unpack_clear:N, \vbox_unpack_clear:c}
+%   \begin{syntax}
+%     \cs{vbox_unpack:N} \meta{box}
+%   \end{syntax}
+%   Unpacks the content of the vertical \meta{box}, retaining any stretching
+%   or shrinking applied when the \meta{box} was set. The \meta{box}
+%   is then cleared globally.
+%   \begin{texnote}
+%     This is the \TeX{} primitive \cs{unvbox}.
+%   \end{texnote}
+% \end{function}
+%
+% \section{Primitive box conditionals}
+%
+% \begin{function}[EXP]{\if_hbox:N}
+%   \begin{syntax}
+%     \cs{if_hbox:N} \meta{box}
+%     ~~\meta{true code}
+%     \cs{else:}
+%     ~~\meta{false code}
+%     \cs{fi:}
+%   \end{syntax}
+%   Tests is \meta{box} is a horizontal box.
+%   \begin{texnote}
+%     This is the \TeX{} primitive \cs{ifhbox}.
+%   \end{texnote}
+% \end{function}
+%
+% \begin{function}[EXP]{\if_vbox:N}
+%   \begin{syntax}
+%     \cs{if_vbox:N} \meta{box}
+%     ~~\meta{true code}
+%     \cs{else:}
+%     ~~\meta{false code}
+%     \cs{fi:}
+%   \end{syntax}
+%   Tests is \meta{box} is a vertical box.
+%   \begin{texnote}
+%     This is the \TeX{} primitive \cs{ifvbox}.
+%   \end{texnote}
+% \end{function}
+%
+% \begin{function}[EXP]{\if_box_empty:N}
+%   \begin{syntax}
+%     \cs{if_box_empty:N} \meta{box}
+%     ~~\meta{true code}
+%     \cs{else:}
+%     ~~\meta{false code}
+%     \cs{fi:}
+%   \end{syntax}
+%   Tests is \meta{box} is an empty (void) box.
+%   \begin{texnote}
+%     This is the \TeX{} primitive \cs{ifvoid}.
+%   \end{texnote}
+% \end{function}
+%
+% \end{documentation}
+%
+% \begin{implementation}
+%
+% \section{\pkg{l3box} implementation}
+%
+%    \begin{macrocode}
+%<*initex|package>
+%    \end{macrocode}
+%
+%    \begin{macrocode}
+%<*package>
+\ProvidesExplPackage
+  {\ExplFileName}{\ExplFileDate}{\ExplFileVersion}{\ExplFileDescription}
+\package_check_loaded_expl:
+%<*initex|package>
+%    \end{macrocode}
+%
+%  The code in this module is very straight forward so I'm not going to
+%  comment it very extensively.
+%
+% \subsection{Creating and initialising boxes}
+%
+% \TestFiles{m3box001.lvt}
+%
+% \begin{macro}{\box_new:N,\box_new:c}
+%   Defining a new \meta{box} register: remember that box $255$ is not
+%   generally available.
+%    \begin{macrocode}
+%<*initex>
+\alloc_new:nnnN { box } \c_zero \c_max_register_int \tex_mathchardef:D
+\seq_put_right:Nn \g_box_allocation_seq {255}
+%</initex>
+%<*package>
+\cs_new_protected:Npn \box_new:N #1
+  {
+    \chk_if_free_cs:N #1
+    \newbox #1
+  }
+%</package>
+\cs_generate_variant:Nn \box_new:N { c }
+%    \end{macrocode}
+%
+% \begin{macro}{\box_clear:N, \box_clear:c}
+% \begin{macro}{\box_gclear:N, \box_gclear:c}
+% \testfile*
+%   Clear a \meta{box} register.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \box_clear:N #1
+  { \box_set_eq:NN  #1 \c_empty_box }
+\cs_new_protected_nopar:Npn \box_gclear:N #1
+  { \box_gset_eq:NN #1 \c_empty_box }
+\cs_generate_variant:Nn \box_clear:N  { c }
+\cs_generate_variant:Nn \box_gclear:N { c }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\box_clear_new:N, \box_clear_new:c}
+% \begin{macro}{\box_gclear_new:N, \box_gclear_new:c}
+% \testfile*
+%   Clear or new.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \box_clear_new:N #1
+  {
+    \cs_if_exist:NTF #1
+      { \box_set_eq:NN #1 \c_empty_box }
+      { \box_new:N #1 }
+  }
+\cs_new_protected_nopar:Npn \box_gclear_new:N #1
+  {
+    \cs_if_exist:NTF #1
+      { \box_gset_eq:NN #1 \c_empty_box }
+      { \box_new:N #1 }
+  }
+\cs_generate_variant:Nn \box_clear_new:N  { c }
+\cs_generate_variant:Nn \box_gclear_new:N { c }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+%  \begin{macro}
+%    {\box_set_eq:NN, \box_set_eq:cN, \box_set_eq:Nc, \box_set_eq:cc}
+% \testfile*
+%  \begin{macro}
+%    {\box_gset_eq:NN, \box_gset_eq:cN, \box_gset_eq:Nc, \box_gset_eq:cc}
+% \testfile*
+%   Assigning the contents of a box to be another box.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \box_set_eq:NN #1#2
+  { \tex_setbox:D #1 \tex_copy:D #2 }
+\cs_new_protected_nopar:Npn \box_gset_eq:NN
+  { \pref_global:D \box_set_eq:NN }
+\cs_generate_variant:Nn \box_set_eq:NN  { cN , Nc , cc }
+\cs_generate_variant:Nn \box_gset_eq:NN { cN , Nc , cc }
+%    \end{macrocode}
+%  \end{macro}
+%  \end{macro}
+%
+% \begin{macro}
+%   {
+%     \box_set_eq_clear:NN, \box_set_eq_clear:cN,
+%     \box_set_eq_clear:Nc, \box_set_eq_clear:cc
+%   }
+% \testfile*
+% \begin{macro}
+%   {
+%     \box_gset_eq_clear:NN, \box_gset_eq_clear:cN,
+%     \box_gset_eq_clear:Nc, \box_gset_eq_clear:cc
+%   }
+% \testfile*
+%    Assigning the contents of a box to be another box.
+%    This clears the second box globally (that's how \TeX{} does it).
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \box_set_eq_clear:NN #1#2
+  { \tex_setbox:D #1 \tex_box:D #2 }
+\cs_new_protected_nopar:Npn \box_gset_eq_clear:NN
+  { \pref_global:D  \box_set_eq_clear:NN }
+\cs_generate_variant:Nn \box_set_eq_clear:NN  { cN , Nc , cc }
+\cs_generate_variant:Nn \box_gset_eq_clear:NN { cN , Nc , cc }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \subsection{Measuring and setting box dimensions}
+%
+% \begin{macro}{\box_ht:N,\box_ht:c}
+% \begin{macro}{\box_dp:N,\box_dp:c}
+% \begin{macro}{\box_wd:N,\box_wd:c}
+% \testfile*
+%    Accessing the height, depth, and width of a \meta{box} register.
+%    \begin{macrocode}
+\cs_new_eq:NN \box_ht:N \tex_ht:D
+\cs_new_eq:NN \box_dp:N \tex_dp:D
+\cs_new_eq:NN \box_wd:N \tex_wd:D
+\cs_generate_variant:Nn \box_ht:N { c }
+\cs_generate_variant:Nn \box_dp:N { c }
+\cs_generate_variant:Nn \box_wd:N { c }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\box_set_ht:Nn, \box_set_ht:cn}
+% \begin{macro}{\box_set_dp:Nn, \box_set_dp:cn}
+% \begin{macro}{\box_set_wd:Nn, \box_set_wd:cn}
+%   Measuring is easy: all primitive work. These primitives are not
+%   expandable, so the derived functions are not either.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \box_set_dp:Nn #1#2
+  { \box_dp:N #1 \dim_eval:w #2 \dim_eval_end: }
+\cs_new_protected_nopar:Npn \box_set_ht:Nn #1#2
+  { \box_ht:N #1 \dim_eval:w #2 \dim_eval_end: }
+\cs_new_protected_nopar:Npn \box_set_wd:Nn #1#2
+  { \box_wd:N #1 \dim_eval:w #2 \dim_eval_end: }
+\cs_generate_variant:Nn \box_set_ht:Nn { c }
+\cs_generate_variant:Nn \box_set_dp:Nn { c }
+\cs_generate_variant:Nn \box_set_wd:Nn { c }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \subsection{Using boxes}
+%
+% \begin{macro}{\box_use_clear:N, \box_use_clear:c}
+% \begin{macro}{\box_use:N, \box_use:c}
+%   Using a \meta{box}. These are just \TeX{} primitives with meaningful
+%   names.
+%    \begin{macrocode}
+\cs_new_eq:NN \box_use_clear:N \tex_box:D
+\cs_new_eq:NN \box_use:N \tex_copy:D
+\cs_generate_variant:Nn \box_use_clear:N { c }
+\cs_generate_variant:Nn \box_use:N { c }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\box_move_left:nn,\box_move_right:nn}
+% \begin{macro}{\box_move_up:nn,\box_move_down:nn}
+% \testfile*
+%   Move box material in different directions.
+%    \begin{macrocode}
+\cs_new_protected:Npn \box_move_left:nn #1#2
+  { \tex_moveleft:D \dim_eval:w #1 \dim_eval_end: #2 }
+\cs_new_protected:Npn \box_move_right:nn #1#2
+  { \tex_moveright:D \dim_eval:w #1 \dim_eval_end: #2 }
+\cs_new_protected:Npn \box_move_up:nn #1#2
+  { \tex_raise:D \dim_eval:w #1 \dim_eval_end: #2 }
+\cs_new_protected:Npn \box_move_down:nn #1#2
+  { \tex_lower:D \dim_eval:w #1 \dim_eval_end: #2 }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \subsection{Box conditionals}
+%
+% \begin{macro}{\if_hbox:N}
+% \begin{macro}{\if_vbox:N}
+% \begin{macro}{\if_box_empty:N}
+%  \testfile*
+%    The primitives for testing if a \meta{box} is empty/void or which
+%    type of box it is.
+%    \begin{macrocode}
+\cs_new_eq:NN \if_hbox:N      \tex_ifhbox:D
+\cs_new_eq:NN \if_vbox:N      \tex_ifvbox:D
+\cs_new_eq:NN \if_box_empty:N \tex_ifvoid:D
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}[pTF]{\box_if_horizontal:N,\box_if_horizontal:c}
+% \testfile*
+% \begin{macro}[pTF]{\box_if_vertical:N,\box_if_vertical:c}
+% \testfile*
+%    \begin{macrocode}
+\prg_new_conditional:Npnn \box_if_horizontal:N #1 { p , T , F , TF }
+  { \if_hbox:N #1 \prg_return_true: \else: \prg_return_false: \fi: }
+\prg_new_conditional:Npnn \box_if_vertical:N #1 { p , T , F , TF }
+  { \if_vbox:N #1 \prg_return_true: \else: \prg_return_false: \fi: }
+\cs_generate_variant:Nn \box_if_horizontal_p:N { c }
+\cs_generate_variant:Nn \box_if_horizontal:NT  { c }
+\cs_generate_variant:Nn \box_if_horizontal:NF  { c }
+\cs_generate_variant:Nn \box_if_horizontal:NTF { c }
+\cs_generate_variant:Nn \box_if_vertical_p:N { c }
+\cs_generate_variant:Nn \box_if_vertical:NT  { c }
+\cs_generate_variant:Nn \box_if_vertical:NF  { c }
+\cs_generate_variant:Nn \box_if_vertical:NTF { c }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}[pTF]{\box_if_empty:N, \box_if_empty:c}
+% \testfile*
+%   Testing if a \meta{box} is empty/void.
+%    \begin{macrocode}
+\prg_new_conditional:Npnn \box_if_empty:N #1 { p , T , F , TF }
+  { \if_box_empty:N #1 \prg_return_true: \else: \prg_return_false: \fi: }
+\cs_generate_variant:Nn \box_if_empty_p:N { c }
+\cs_generate_variant:Nn \box_if_empty:NT  { c }
+\cs_generate_variant:Nn \box_if_empty:NF  { c }
+\cs_generate_variant:Nn \box_if_empty:NTF { c }
+%    \end{macrocode}
+%  \end{macro}
+%  \end{macro}
+%
+% \subsection{The last box inserted}
+%
+% \begin{variable}{\l_last_box}
+%    A different name for this read-only primitive.
+%    \begin{macrocode}
+\cs_new_eq:NN \l_last_box \tex_lastbox:D
+%    \end{macrocode}
+% \end{variable}
+%
+% \begin{macro}{\box_set_to_last:N, \box_set_to_last:c}
+% \begin{macro}{\box_gset_to_last:N, \box_gset_to_last:c}
+% \testfile*
+%    Set a box to the previous box.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \box_set_to_last:N #1
+  { \tex_setbox:D #1 \l_last_box }
+\cs_new_protected_nopar:Npn \box_gset_to_last:N
+  { \pref_global:D \box_set_to_last:N }
+\cs_generate_variant:Nn \box_set_to_last:N  { c }
+\cs_generate_variant:Nn \box_gset_to_last:N { c }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \subsection{Constant boxes}
+%
+%  \begin{variable}{\c_empty_box}
+%    \begin{macrocode}
+%<*package>
+\cs_new_eq:NN \c_empty_box \voidb@x
+%</package>
+%<*initex>
+\box_new:N \c_empty_box
+%</initex>
+%    \end{macrocode}
+% \end{variable}
+%
+% \subsection{Scratch boxes}
+%
+%  \begin{variable}{\l_tmpa_box, \l_tmpb_box}
+%    \begin{macrocode}
+%<*package>
+\cs_new_eq:NN \l_tmpa_box \@tempboxa
+%</package>
+%<*initex>
+\box_new:N \l_tmpa_box
+%</initex>
+\box_new:N \l_tmpb_box
+%    \end{macrocode}
+% \end{variable}
+%
+% \subsection{Viewing box contents}
+%
+% \begin{macro}{\box_show:N,\box_show:c}
+% \testfile*
+%   Show the contents of a box and write it into the log file.
+%    \begin{macrocode}
+\cs_new_eq:NN \box_show:N \tex_showbox:D
+\cs_generate_variant:Nn \box_show:N { c }
+%    \end{macrocode}
+%  \end{macro}
+%
+%  \subsection{Horizontal mode boxes}
+%
+% \begin{macro}{\hbox:n}
+% \testfile{m3box002.lvt}
+%   Put a horizontal box directly into the input stream.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \hbox:n { \tex_hbox:D \scan_stop: }
+%    \end{macrocode}
+%  \end{macro}
+%
+% \begin{macro}{\hbox_set:Nn,\hbox_set:cn}
+% \begin{macro}{\hbox_gset:Nn,\hbox_gset:cn}
+% \testfile*
+%    \begin{macrocode}
+\cs_new_protected:Npn \hbox_set:Nn #1#2 { \tex_setbox:D #1 \tex_hbox:D {#2} }
+\cs_new_protected_nopar:Npn \hbox_gset:Nn { \pref_global:D \hbox_set:Nn }
+\cs_generate_variant:Nn \hbox_set:Nn { c }
+\cs_generate_variant:Nn \hbox_gset:Nn { c }
+%    \end{macrocode}
+%  \end{macro}
+%  \end{macro}
+%
+% \begin{macro}{\hbox_set_to_wd:Nnn,\hbox_set_to_wd:cnn}
+% \begin{macro}{\hbox_gset_to_wd:Nnn,\hbox_gset_to_wd:cnn}
+% \testfile*
+%   Storing material in a horizontal box with a specified width.
+%    \begin{macrocode}
+\cs_new_protected:Npn \hbox_set_to_wd:Nnn #1#2#3
+  { \tex_setbox:D #1 \tex_hbox:D to \dim_eval:w #2 \dim_eval_end: {#3} }
+\cs_new_protected_nopar:Npn \hbox_gset_to_wd:Nnn
+  { \pref_global:D \hbox_set_to_wd:Nnn }
+\cs_generate_variant:Nn \hbox_set_to_wd:Nnn { c }
+\cs_generate_variant:Nn \hbox_gset_to_wd:Nnn {cnn}
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\hbox_set_inline_begin:N, \hbox_set_inline_begin:c}
+% \begin{macro}{\hbox_gset_inline_begin:N, \hbox_gset_inline_begin:c}
+% \begin{macro}{\hbox_set_inline_end:}
+% \begin{macro}{\hbox_gset_inline_end:}
+% \testfile*
+%    Storing material in a horizontal box. This type is useful in
+%    environment definitions.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \hbox_set_inline_begin:N  #1
+  { \tex_setbox:D #1 \tex_hbox:D \c_group_begin_token }
+\cs_new_protected_nopar:Npn \hbox_gset_inline_begin:N
+  { \pref_global:D \hbox_set_inline_begin:N }
+\cs_generate_variant:Nn \hbox_set_inline_begin:N  { c }
+\cs_generate_variant:Nn \hbox_gset_inline_begin:N { c }
+\cs_new_eq:NN \hbox_set_inline_end:  \c_group_end_token
+\cs_new_eq:NN \hbox_gset_inline_end: \c_group_end_token
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+%  \begin{macro}{\hbox_to_wd:nn}
+%  \begin{macro}{\hbox_to_zero:n}
+%  \testfile*
+%   Put a horizontal box directly into the input stream.
+%    \begin{macrocode}
+\cs_new_protected:Npn \hbox_to_wd:nn #1#2
+   { \tex_hbox:D to \dim_eval:w #1 \dim_eval_end: {#2} }
+\cs_new_protected:Npn \hbox_to_zero:n #1 { \tex_hbox:D to \c_zero_skip {#1} }
+%    \end{macrocode}
+%  \end{macro}
+%  \end{macro}
+%
+% \begin{macro}{\hbox_overlap_left:n,}
+% \begin{macro}{\hbox_overlap_right:n}
+%   Put a zero-sized box with the contents pushed against one side (which
+%   makes it stick out on the other) directly into the input stream.
+%    \begin{macrocode}
+\cs_new_protected:Npn \hbox_overlap_left:n  #1
+  { \hbox_to_zero:n { \tex_hss:D #1 } }
+\cs_new_protected:Npn \hbox_overlap_right:n #1
+  { \hbox_to_zero:n { #1 \tex_hss:D } }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\hbox_unpack:N, \hbox_unpack:c}
+% \begin{macro}{\hbox_unpack_clear:N, \hbox_unpack_clear:c}
+% \testfile*
+%   Unpacking a box and if requested also clear it.
+%    \begin{macrocode}
+\cs_new_eq:NN \hbox_unpack:N \tex_unhcopy:D
+\cs_new_eq:NN \hbox_unpack_clear:N \tex_unhbox:D
+\cs_generate_variant:Nn \hbox_unpack:N { c }
+\cs_generate_variant:Nn \hbox_unpack_clear:N { c }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+%  \subsection{Vertical mode boxes}
+%
+% \begin{macro}{\vbox:n}
+% \TestFiles{m3box003.lvt}
+% \begin{macro}{\vbox:n\vbox_top:n}
+% \TestFiles{m3box003.lvt}
+%   Put a vertical box directly into the input stream.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \vbox:n { \tex_vbox:D \scan_stop: }
+\cs_new_protected_nopar:Npn \vbox_top:n { \tex_vtop:D \scan_stop: }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\vbox_to_ht:nn,\vbox_to_zero:n}
+% \begin{macro}{\vbox_to_ht:nn,\vbox_to_zero:n}
+% \testfile*
+%   Put a vertical box directly into the input stream.
+%    \begin{macrocode}
+\cs_new_protected:Npn \vbox_to_ht:nn #1#2
+  { \tex_vbox:D to \dim_eval:w #1 \dim_eval_end: {#2} }
+\cs_new_protected:Npn \vbox_to_zero:n #1 { \tex_vbox:D to \c_zero_dim {#1} }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\vbox_set:Nn, \vbox_set:cn}
+% \begin{macro}{\vbox_gset:Nn, \vbox_gset:cn}
+% \testfile*
+%   Storing material in a vertical box with a natural height.
+%    \begin{macrocode}
+\cs_new_protected:Npn \vbox_set:Nn #1#2 { \tex_setbox:D #1 \tex_vbox:D {#2} }
+\cs_new_protected_nopar:Npn \vbox_gset:Nn  { \pref_global:D \vbox_set:Nn }
+\cs_generate_variant:Nn \vbox_set:Nn  { c }
+\cs_generate_variant:Nn \vbox_gset:Nn { c }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\vbox_set_top:Nn, \vbox_set_top:cn}
+% \begin{macro}{\vbox_gset_top:Nn, \vbox_gset_top:cn}
+% \testfile*
+%   Storing material in a vertical box with a natural height and reference
+%   point at the baseline of the first object in the box.
+%    \begin{macrocode}
+\cs_new_protected:Npn \vbox_set_top:Nn #1#2
+  { \tex_setbox:D #1 \tex_vtop:D {#2} }
+\cs_new_protected_nopar:Npn \vbox_gset_top:Nn
+  { \pref_global:D \vbox_set_top:Nn }
+\cs_generate_variant:Nn \vbox_set_top:Nn { c }
+\cs_generate_variant:Nn \vbox_gset_top:Nn { c }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\vbox_set_to_ht:Nnn,\vbox_set_to_ht:cnn}
+% \begin{macro}{\vbox_gset_to_ht:Nnn,\vbox_gset_to_ht:cnn}
+%  \testfile*
+%  Storing material in a vertical box with a specified height.
+%    \begin{macrocode}
+\cs_new_protected:Npn \vbox_set_to_ht:Nnn #1#2#3
+  { \tex_setbox:D #1 \tex_vbox:D to \dim_eval:w #2 \dim_eval_end: {#3} }
+\cs_new_protected_nopar:Npn \vbox_gset_to_ht:Nnn
+  { \pref_global:D \vbox_set_to_ht:Nnn }
+\cs_generate_variant:Nn \vbox_set_to_ht:Nnn  { c }
+\cs_generate_variant:Nn \vbox_gset_to_ht:Nnn { c }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\vbox_set_inline_begin:N, \vbox_set_inline_begin:c}
+% \begin{macro}{\vbox_gset_inline_begin:N, \vbox_gset_inline_begin:c}
+% \begin{macro}{\vbox_set_inline_end:}
+% \begin{macro}{\vbox_gset_inline_end:}
+% \testfile*
+%   Storing material in a vertical box. This type is useful in
+%   environment definitions.
+%    \begin{macrocode}
+\cs_new_nopar:Npn \vbox_set_inline_begin:N #1
+  { \tex_setbox:D #1 \tex_vbox:D \c_group_begin_token }
+\cs_new_protected_nopar:Npn \vbox_gset_inline_begin:N
+  { \pref_global:D \vbox_set_inline_begin:N }
+\cs_generate_variant:Nn \vbox_set_inline_begin:N  { c }
+\cs_generate_variant:Nn \vbox_gset_inline_begin:N { c }
+\cs_new_eq:NN \vbox_set_inline_end:  \c_group_end_token
+\cs_new_eq:NN \vbox_gset_inline_end: \c_group_end_token
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\vbox_unpack:N, \vbox_unpack:c}
+% \begin{macro}{\vbox_unpack_clear:N, \vbox_unpack_clear:c}
+% \testfile*
+%   Unpacking a box and if requested also clear it.
+%    \begin{macrocode}
+\cs_new_eq:NN \vbox_unpack:N \tex_unvcopy:D
+\cs_new_eq:NN \vbox_unpack_clear:N \tex_unvbox:D
+\cs_generate_variant:Nn \vbox_unpack:N { c }
+\cs_generate_variant:Nn \vbox_unpack_clear:N { c }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\vbox_set_split_to_ht:NNn}
+% \testfile*
+%   Splitting a vertical box in two.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \vbox_set_split_to_ht:NNn #1#2#3
+  { \tex_setbox:D #1 \tex_vsplit:D #2 to \dim_eval:w #3 \dim_eval_end: }
+%    \end{macrocode}
+% \end{macro}
+%
+%    \begin{macrocode}
+%</initex|package>
+%    \end{macrocode}
+%
+% \end{implementation}
+%
+% \PrintIndex
diff --git a/Master/texmf-dist/source/latex/l3kernel/l3clist.dtx b/Master/texmf-dist/source/latex/l3kernel/l3clist.dtx
new file mode 100644
index 00000000000..d031b15f552
--- /dev/null
+++ b/Master/texmf-dist/source/latex/l3kernel/l3clist.dtx
@@ -0,0 +1,1437 @@
+% \iffalse meta-comment
+%
+%% File: l3clist.dtx Copyright (C) 2004-2011 Frank Mittelbach,
+%%                                 The LaTeX3 project
+%%
+%% It may be distributed and/or modified under the conditions of the
+%% LaTeX Project Public License (LPPL), either version 1.3c of this
+%% license or (at your option) any later version.  The latest version
+%% of this license is in the file
+%%
+%%    http://www.latex-project.org/lppl.txt
+%%
+%% This file is part of the "expl3 bundle" (The Work in LPPL)
+%% and all files in that bundle must be distributed together.
+%%
+%% The released version of this bundle is available from CTAN.
+%%
+%% -----------------------------------------------------------------------
+%%
+%% The development version of the bundle can be found at
+%%
+%%    http://www.latex-project.org/svnroot/experimental/trunk/
+%%
+%% for those people who are interested.
+%%
+%%%%%%%%%%%
+%% NOTE: %%
+%%%%%%%%%%%
+%%
+%%   Snapshots taken from the repository represent work in progress and may
+%%   not work or may contain conflicting material!  We therefore ask
+%%   people _not_ to put them into distributions, archives, etc. without
+%%   prior consultation with the LaTeX3 Project.
+%%
+%% -----------------------------------------------------------------------
+%
+%<*driver|package>
+\RequirePackage{l3names}
+\GetIdInfo$Id: l3clist.dtx 2498 2011-07-09 17:51:47Z joseph $
+  {L3 Experimental comma separated lists}
+%</driver|package>
+%<*driver>
+\documentclass[full]{l3doc}
+\begin{document}
+  \DocInput{\jobname.dtx}
+\end{document}
+%</driver>
+% \fi
+%
+% \title{^^A
+%   The \pkg{l3clist} package\\ Comma separated lists^^A
+%   \thanks{This file describes v\ExplFileVersion,
+%      last revised \ExplFileDate.}^^A
+% }
+%
+% \author{^^A
+%  The \LaTeX3 Project\thanks
+%    {^^A
+%      E-mail:
+%        \href{mailto:latex-team@latex-project.org}
+%          {latex-team@latex-project.org}^^A
+%    }^^A
+% }
+%
+% \date{Released \ExplFileDate}
+%
+% \maketitle
+%
+% \begin{documentation}
+%
+% Comma lists contain ordered data where items can be added to the left
+% or right end of the sequence. This gives an ordered list which can
+% then be utilised with the \cs{clist_map_function:NN} function. Comma
+% Comma lists cannot contain empty items, thus
+% \begin{verbatim}
+%   \clist_new:N \l_my_clist
+%   \clist_put_right:Nn \l_my_clist { }
+%   \clist_if_empty:NTF \l_my_clist { true } { false }
+% \end{verbatim}
+% will leave \texttt{true} in the input stream.
+%
+% \section{Creating and initialising comma lists}
+%
+% \begin{function}{\clist_new:N, \clist_new:c}
+%   \begin{syntax}
+%     \cs{clist_new:N} \meta{comma list}
+%   \end{syntax}
+%   Creates a new \meta{comma list} or raises an error if the name is
+%   already taken. The declaration is global. The \meta{comma list} will
+%   initially contain no items.
+% \end{function}
+%
+% \begin{function}{\clist_clear:N, \clist_clear:c}
+%   \begin{syntax}
+%     \cs{clist_clear:N} \meta{comma list}
+%   \end{syntax}
+%   Clears all items from the \meta{comma list} within the scope of
+%   the current \TeX{} group.
+% \end{function}
+%
+% \begin{function}{\clist_gclear:N, \clist_gclear:c}
+%   \begin{syntax}
+%     \cs{clist_gclear:N} \meta{comma list}
+%   \end{syntax}
+%   Clears all entries from the \meta{comma list} globally.
+% \end{function}
+%
+% \begin{function}{\clist_clear_new:N, \clist_clear_new:c}
+%   \begin{syntax}
+%     \cs{clist_clear_new:N} \meta{comma list}
+%   \end{syntax}
+%   If the \meta{comma list} already exists, clears it within the scope
+%   of the current \TeX{} group. If the \meta{comma list} is not defined,
+%   it will be created (using \cs{clist_new:N}). Thus the comma list is
+%   guaranteed to be available and clear within the current \TeX{}
+%   group. The \meta{comma list} will exist globally, but the content
+%   outside of the current \TeX{} group is not specified.
+% \end{function}
+%
+% \begin{function}{\clist_gclear_new:N, \clist_gclear_new:c}
+%   \begin{syntax}
+%     \cs{clist_gclear_new:N} \meta{comma list}
+%   \end{syntax}
+%   If the \meta{comma list} already exists, clears it globally. If the
+%   \meta{comma list} is not defined, it will be created (using
+%   \cs{clist_new:N}). Thus the comma list is guaranteed to be available
+%   and globally clear.
+% \end{function}
+%
+% \begin{function}
+%   {\clist_set_eq:NN, \clist_set_eq:cN, \clist_set_eq:Nc, \clist_set_eq:cc}
+%   \begin{syntax}
+%     \cs{clist_set_eq:NN} \meta{comma list1} \meta{comma list2}
+%   \end{syntax}
+%   Sets the content of \meta{comma list1} equal to that of
+%   \meta{comma list2}. This assignment is restricted to the current
+%   \TeX{} group level.
+% \end{function}
+%
+% \begin{function}
+%   {
+%     \clist_gset_eq:NN, \clist_gset_eq:cN,
+%     \clist_gset_eq:Nc, \clist_gset_eq:cc
+%   }
+%   \begin{syntax}
+%     \cs{clist_gset_eq:NN} \meta{comma list1} \meta{comma list2}
+%   \end{syntax}
+%   Sets the content of \meta{comma list1} equal to that of
+%   \meta{comma list2}. This assignment is global and so is not
+%   limited by the current \TeX{} group level.
+% \end{function}
+%
+% \begin{function}{\clist_concat:NNN, \clist_concat:ccc}
+%   \begin{syntax}
+%     \cs{clist_concat:NNN} \meta{comma list1} \meta{comma list2}
+%     ~~\meta{comma list3}
+%   \end{syntax}
+%   Concatenates the content of \meta{comma list2} and \meta{comma list3}
+%   together and saves the result in \meta{comma list1}. The items in
+%   \meta{comma list2} will be placed at the left side of the new comma list.
+%   This  operation is local to the current \TeX{} group and will
+%   remove any existing content in \meta{comma list1}.
+% \end{function}
+%
+% \begin{function}{\clist_gconcat:NNN, \clist_gconcat:ccc}
+%   \begin{syntax}
+%     \cs{clist_gconcat:NNN} \meta{comma list1} \meta{comma list2}
+%     ~~\meta{comma list3}
+%   \end{syntax}
+%   Concatenates the content of \meta{comma list2} and \meta{comma list3}
+%   together and saves the result in \meta{comma list1}. The items in
+%   \meta{comma list2}  will be placed at the left side of the new comma list.
+%   This operation is global and will remove any existing content in
+%   \meta{comma list1}.
+% \end{function}
+%
+% \section{Appending items to comma lists}
+%
+% \begin{function}{
+%   \clist_put_left:Nn, \clist_put_left:NV,
+%   \clist_put_left:No, \clist_put_left:Nx,
+%   \clist_put_left:cn, \clist_put_left:cV,
+%   \clist_put_left:co, \clist_put_left:cx
+% }
+%   \begin{syntax}
+%     \cs{clist_put_left:Nn} \meta{comma list} \Arg{item}
+%   \end{syntax}
+%   Appends the \meta{item} to the left of the \meta{comma list}.
+%   The assignment is restricted to the current \TeX{} group.
+% \end{function}
+%
+% \begin{function}{
+%   \clist_gput_left:Nn, \clist_gput_left:NV,
+%   \clist_gput_left:No, \clist_gput_left:Nx,
+%   \clist_gput_left:cn, \clist_gput_left:cV,
+%   \clist_gput_left:co, \clist_gput_left:cx
+% }
+%   \begin{syntax}
+%     \cs{clist_gput_left:Nn} \meta{comma list} \Arg{item}
+%   \end{syntax}
+%   Appends the \meta{item} to the left of the \meta{comma list}.
+%   The assignment is global.
+% \end{function}
+%
+% \begin{function}{
+%   \clist_put_right:Nn, \clist_put_right:NV,
+%   \clist_put_right:No, \clist_put_right:Nx,
+%   \clist_put_right:cn, \clist_put_right:cV,
+%   \clist_put_right:co, \clist_put_right:cx
+% }
+%   \begin{syntax}
+%     \cs{clist_put_right:Nn} \meta{comma list} \Arg{item}
+%   \end{syntax}
+%   Appends the \meta{item} to the right of the \meta{comma list}.
+%   The assignment is restricted to the current \TeX{} group.
+% \end{function}
+%
+% \begin{function}{
+%   \clist_gput_right:Nn, \clist_gput_right:NV,
+%   \clist_gput_right:No, \clist_gput_right:Nx,
+%   \clist_gput_right:cn, \clist_gput_right:cV,
+%   \clist_gput_right:co, \clist_gput_right:cx
+% }
+%   \begin{syntax}
+%     \cs{clist_gput_right:Nn} \meta{comma list} \Arg{item}
+%   \end{syntax}
+%   Appends the \meta{item} to the right of the \meta{comma list}.
+%   The assignment is global.
+% \end{function}
+%
+% \section{Comma lists as stacks}
+%
+% \begin{function}{\clist_get:NN, \clist_get:cN}
+%   \begin{syntax}
+%     \cs{clist_get:NN} \meta{comma list} \meta{token list variable}
+%   \end{syntax}
+%   Stores the left-most item from a \meta{comma list} in the
+%   \meta{token list variable} without removing it from the
+%   \meta{comma list}. The \meta{token list variable} is assigned locally.
+% \end{function}
+%
+% \begin{function}{\clist_get:NN, \clist_get:cN}
+%   \begin{syntax}
+%     \cs{clist_get:NN} \meta{comma list} \meta{token list variable}
+%   \end{syntax}
+%   Stores the right-most item from a \meta{comma list} in the
+%   \meta{token list variable} without removing it from the
+%   \meta{comma list}. The \meta{token list variable} is assigned locally.
+% \end{function}
+%
+% \begin{function}{\clist_pop:NN, \clist_pop:cN}
+%   \begin{syntax}
+%     \cs{clist_pop:NN} \meta{comma list} \meta{token list variable}
+%   \end{syntax}
+%   Pops the left-most item from a \meta{comma list} into the
+%   \meta{token list variable}, \emph{i.e.}~removes the item from the
+%   comma list and stores it in the \meta{token list variable}.
+%   Both of the variables are assigned locally.
+% \end{function}
+%
+% \begin{function}{\clist_gpop:NN, \clist_gpop:cN}
+%   \begin{syntax}
+%     \cs{clist_gpop:NN} \meta{comma list} \meta{token list variable}
+%   \end{syntax}
+%   Pops the left-most item from a \meta{comma list} into the
+%   \meta{token list variable}, \emph{i.e.}~removes the item from the
+%   comma list and stores it in the \meta{token list variable}.
+%   The \meta{comma list} is modified globally, while the assignment of
+%   the \meta{token list variable} is local.
+% \end{function}
+%
+% \begin{function}
+%   {
+%     \clist_push:Nn, \clist_push:NV, \clist_push:No, \clist_push:Nx,
+%     \clist_push:cn, \clist_push:cV, \clist_push:co, \clist_push:cx
+%   }
+%   \begin{syntax}
+%     \cs{clist_push:Nn} \meta{sequence} \Arg{item}
+%   \end{syntax}
+%   Adds the \Arg{item} to the top of the \meta{comma list}.
+%   The assignment is restricted to the current \TeX{} group.
+% \end{function}
+%
+% \begin{function}
+%   {
+%     \clist_gpush:Nn, \clist_gpush:NV, \clist_gpush:No, \clist_gpush:Nx,
+%     \clist_gpush:cn, \clist_gpush:cV, \clist_gpush:co, \clist_gpush:cx
+%   }
+%   \begin{syntax}
+%     \cs{clist_gpush:Nn} \meta{sequence} \Arg{item}
+%   \end{syntax}
+%   Pushes the \meta{item} onto the end of the top of the
+%   \meta{comma list}. The assignment is global.
+% \end{function}
+%
+% \section{Using comma lists}
+%
+% \begin{function}[EXP]{\clist_use:N, \clist_use:c}
+%   \begin{syntax}
+%     \cs{clist_use:N} \meta{comma list}
+%   \end{syntax}
+%   Places the \meta{comma list} directly into the input stream, thus
+%   treating it as a \meta{token list}.
+% \end{function}
+%
+% \section{Modifying comma lists}
+%
+%  While comma lists are normally used as ordered lists, it may be
+%  necessary to modify the content. The functions here may be used
+%  to update comma lists, while retaining the order of the unaffected
+%  entries.
+%
+% \begin{function}{\clist_remove_duplicates:N, \clist_remove_duplicates:c}
+%   \begin{syntax}
+%     \cs{clist_remove_duplicates:N} \meta{comma list}
+%   \end{syntax}
+%   Removes duplicate items from the \meta{comma list}, leaving the
+%   left most copy of each item in the \meta{comma list}.  The \meta{item}
+%   comparison takes place on a token basis, as for \cs{tl_if_eq:nn(TF)}.
+%   The removal is local to the current \TeX{} group.
+%   \begin{texnote}
+%     This function iterates through every item in the \meta{comma list} and
+%     does a comparison with the \meta{items} already checked. It is therefore
+%     relatively slow with large comma lists.
+%   \end{texnote}
+% \end{function}
+%
+% \begin{function}{\clist_gremove_duplicates:N, \clist_gremove_duplicates:c}
+%   \begin{syntax}
+%     \cs{clist_gremove_duplicates:N} \meta{comma list}
+%   \end{syntax}
+%   Removes duplicate items from the \meta{comma list}, leaving the
+%   left most copy of each item in the \meta{comma list}.  The \meta{item}
+%   comparison takes place on a token basis, as for \cs{tl_if_eq:nn(TF)}.
+%   The removal is applied globally.
+%   \begin{texnote}
+%     This function iterates through every item in the \meta{comma list} and
+%     does a comparison with the \meta{items} already checked. It is therefore
+%     relatively slow with large comma lists.
+%   \end{texnote}
+% \end{function}
+%
+% \begin{function}{\clist_remove_all:Nn, \clist_remove_all:cn}
+%   \begin{syntax}
+%     \cs{clist_remove_all:Nn} \meta{comma list} \Arg{item}
+%   \end{syntax}
+%   Removes every occurrence of \meta{item} from the \meta{comma list}.
+%   The \meta{item} comparison takes place on a token basis, as for
+%   \cs{tl_if_eq:nn(TF)}. The  removal is local to the current \TeX{} group.
+% \end{function}
+%
+% \begin{function}{\clist_gremove_all:Nn, \clist_gremove_all:cn}
+%   \begin{syntax}
+%     \cs{clist_gremove_all:Nn} \meta{comma list} \Arg{item}
+%   \end{syntax}
+%   Removes each occurrence of \meta{item} from the \meta{comma list}.
+%   The \meta{item} comparison takes place on a token basis, as for
+%   \cs{tl_if_eq:nn(TF)}. The removal is applied globally.
+% \end{function}
+% 
+% \begin{function}{\clist_trim_spaces:N, \clist_trim_spaces:c}
+%   \begin{syntax}
+%     \cs{clist_trim_spaces:N} \meta{comma list}
+%   \end{syntax}
+%   Removes leading and trailing spaces from each \meta{item} in the
+%   \meta{comma list} within the current \TeX{} group. This space-removal
+%   process takes place as described for \cs{tl_trim_spaces:n}.
+% \end{function}
+%
+% \begin{function}{\clist_gtrim_spaces:N, \clist_gtrim_spaces:c}
+%   \begin{syntax}
+%     \cs{clist_gtrim_spaces:N} \meta{comma list}
+%   \end{syntax}
+%   Removes leading and trailing spaces from each \meta{item} in the
+%   \meta{comma list} globally. This space-removal
+%   process takes place as described for \cs{tl_trim_spaces:n}.
+% \end{function}
+% 
+% \begin{function}[EXP]{\clist_trim_spaces:n}
+%   \begin{syntax}
+%     \cs{clist_trim_spaces:N} \meta{comma list}
+%   \end{syntax}
+%   Removes leading and trailing spaces from each \meta{item} in the
+%   \meta{comma list}, leaving the resulting modified list in the
+%   input stream.
+% \end{function}
+%
+% \section{Comma list conditionals}
+%
+% \begin{function}[EXP,pTF]{\clist_if_empty:N, \clist_if_empty:c}
+%   \begin{syntax}
+%     \cs{clist_if_empty_p:N} \meta{comma list}
+%     \cs{clist_if_empty:NTF} \meta{comma list}
+%     ~~\Arg{true code} \Arg{false code}
+%   \end{syntax}
+%   Tests if the \meta{comma list} is empty (containing no items). The
+%   branching versions then leave either \meta{true code} or
+%   \meta{false code} in the input stream, as appropriate to the truth of
+%   the test and the variant of the function chosen. The logical truth of
+%   the test is left in the input stream by the predicate version.
+% \end{function}
+%
+% \begin{function}[EXP,pTF]
+%   {\clist_if_eq:NN, \clist_if_eq:Nc, \clist_if_eq:cN, \clist_if_eq:cc}
+%   \begin{syntax}
+%     \cs{clist_if_eq_p:NN} \Arg{clist1} \Arg{clist2}
+%     \cs{clist_if_eq:NNTF} \Arg{clist1} \Arg{clist2} \Arg{true code}
+%     ~~\Arg{false code}
+%   \end{syntax}
+%   Compares the content of two \meta{comma lists} and
+%   is logically \texttt{true} if the two contain the same list of
+%   entries in the same order. The branching versions then leave either
+%   \meta{true code} or \meta{false code} in the input stream, as
+%   appropriate to the truth of the test and the variant of the function
+%   chosen. The logical truth of the test is left in the input stream by
+%   the predicate version.
+% \end{function}
+%
+% \begin{function}[TF]{
+%    \clist_if_in:Nn, \clist_if_in:NV, \clist_if_in:No,
+%    \clist_if_in:cn, \clist_if_in:cV, \clist_if_in:co,
+%    \clist_if_in:nn, \clist_if_in:nV, \clist_if_in:no
+% }
+%   \begin{syntax}
+%     \cs{clist_if_in:NnTF} \meta{comma list} \Arg{item}
+%     ~~\Arg{true code} \Arg{false code}
+%   \end{syntax}
+%   Tests if the \meta{item} is present in the \meta{comma list}.
+%   Either the \meta{true code} or \meta{false code} is left in the input
+%   stream, as appropriate to the truth of the test and the variant of the
+%   function
+%   chosen.
+% \end{function}
+%
+% \section{Mapping to comma lists}
+%
+% \begin{function}[EXP]
+%   {\clist_map_function:NN, \clist_map_function:cN, \clist_map_function:nN}
+%   \begin{syntax}
+%     \cs{clist_map_function:NN} \meta{comma list} \meta{function}
+%   \end{syntax}
+%   Applies \meta{function} to every \meta{item} stored in the
+%   \meta{comma list}. The \meta{function} will receive one argument for
+%   each iteration. The \meta{items} are returned from left to right.
+%   The function \cs{clist_map_inline:Nn} is in general more efficient
+%   than \cs{clist_map_function:NN}.
+%   One mapping may be nested inside another.
+% \end{function}
+%
+% \begin{function}
+%   {\clist_map_inline:Nn, \clist_map_inline:cn, \clist_map_inline:nn}
+%   \begin{syntax}
+%     \cs{clist_map_inline:Nn} \meta{comma list} \Arg{inline function}
+%   \end{syntax}
+%   Applies \meta{inline function} to every \meta{item} stored
+%   within the \meta{comma list}. The \meta{inline function} should
+%   consist of code which will receive the \meta{item} as |#1|.
+%   One in line mapping can be nested inside another. The \meta{items}
+%   are returned from left to right.
+% \end{function}
+%
+% \begin{function}
+%   {\clist_map_variable:NNn, \clist_map_variable:cNn, \clist_map_variable:nNn}
+%   \begin{syntax}
+%     \cs{clist_map_variable:NNn} \meta{comma list}
+%     ~~\meta{tl~var.} \Arg{function using tl~var.}
+%   \end{syntax}
+%   Stores each entry in the \meta{comma list} in turn in the
+%   \meta{tl~var.}\ and applies the \meta{function using tl~var.}
+%   The \meta{function} will usually consist of code making use of
+%   the \meta{tl~var.}, but this is not enforced.  One variable
+%   mapping can be nested inside another. The \meta{items}
+%   are returned from left to right.
+% \end{function}
+%
+% \begin{function}[EXP]{\clist_map_break:}
+%   \begin{syntax}
+%     \cs{clist_map_break:}
+%   \end{syntax}
+%   Used to terminate a \cs{clist_map_\ldots} function before all
+%   entries in the \meta{comma list} have been processed. This will
+%   normally take place within a conditional statement, for example
+%   \begin{verbatim}
+%     \clist_map_inline:Nn \l_my_clist
+%       {
+%         \str_if_eq:nnTF { #1 } { bingo }
+%           { \clist_map_break: }
+%           {
+%             % Do something useful
+%           }
+%       }
+%   \end{verbatim}
+%   Use outside of a \cs{clist_map_\ldots} scenario will lead to low
+%   level \TeX{} errors.
+%   \begin{texnote}
+%     When the mapping is broken, additional tokens may be inserted by the
+%     internal macro \cs{clist_break_point:n} before further items are taken
+%     from the input stream. This will depend on the design of the mapping
+%     function.
+%   \end{texnote}
+% \end{function}
+%
+% \begin{function}[EXP]{\clist_map_break:n}
+%   \begin{syntax}
+%     \cs{clist_map_break:n} \Arg{tokens}
+%   \end{syntax}
+%   Used to terminate a \cs{clist_map_\ldots} function before all
+%   entries in the \meta{comma list} have been processed, inserting
+%   the \meta{tokens} after the mapping has ended. This will
+%   normally take place within a conditional statement, for example
+%   \begin{verbatim}
+%     \clist_map_inline:Nn \l_my_clist
+%       {
+%         \str_if_eq:nnTF { #1 } { bingo }
+%           { \clist_map_break:n { <tokens> } }
+%           {
+%             % Do something useful
+%           }
+%       }
+%   \end{verbatim}
+%   Use outside of a \cs{clist_map_\ldots} scenario will lead to low
+%   level \TeX{} errors.
+%   \begin{texnote}
+%     When the mapping is broken, additional tokens may be inserted by the
+%     internal macro \cs{clist_break_point:n} before the \meta{tokens} are
+%     inserted into the input stream.
+%     This will depend on the design of the mapping function.
+%   \end{texnote}
+% \end{function}
+%
+% \section{Comma lists as stacks}
+%
+% Comma lists can be used as stacks, where data is pushed to and popped
+% from the top of the comma list. (The left of a comma list is the top, for
+% performance reasons.) The stack functions for comma lists are not
+% intended to be mixed with the general ordered data functions detailed
+% in the previous section: a comma list should either be used as an
+% ordered data type or as a stack, but not in both ways.
+%
+% \begin{function}{\clist_get:NN, \clist_get:cN}
+%   \begin{syntax}
+%     \cs{clist_get:NN} \meta{comma list} \meta{token list variable}
+%   \end{syntax}
+%   Reads the top item from a \meta{comma list} into the
+%   \meta{token list variable} without removing it from the
+%   \meta{comma list}. The \meta{token list variable} is assigned locally.
+%   If \meta{comma list} is empty an error will be raised.
+% \end{function}
+%
+% \begin{function}{\clist_pop:NN, \clist_pop:cN}
+%   \begin{syntax}
+%     \cs{clist_pop:NN} \meta{comma list} \meta{token list variable}
+%   \end{syntax}
+%   Pops the top item from a \meta{comma list} into the
+%   \meta{token list variable}. Both of the variables are assigned
+%   locally. If \meta{comma list} is empty an error will be raised.
+% \end{function}
+%
+% \begin{function}{\clist_gpop:NN, \clist_gpop:cN}
+%   \begin{syntax}
+%     \cs{clist_gpop:NN} \meta{comma list} \meta{token list variable}
+%   \end{syntax}
+%   Pops the top item from a \meta{comma list} into the
+%   \meta{token list variable}. The \meta{comma list} is modified globally,
+%   while the \meta{token list variable} is assigned locally. If
+%   \meta{comma list} is empty an error will be raised.
+% \end{function}
+%
+% \begin{function}{
+%   \clist_push:Nn, \clist_push:NV, \clist_push:No, \clist_push:Nx,
+%   \clist_push:cn, \clist_push:cV, \clist_push:co, \clist_push:cx
+% }
+%   \begin{syntax}
+%     \cs{clist_push:Nn} \meta{comma list} \Arg{item}
+%   \end{syntax}
+%   Adds the \Arg{item} to the top of the \meta{comma list}.
+%   The assignment is restricted to the current \TeX{}  group.
+% \end{function}
+%
+% \begin{function}{
+%   \clist_gpush:Nn, \clist_gpush:NV, \clist_gpush:No, \clist_gpush:Nx,
+%   \clist_gpush:cn, \clist_gpush:cV, \clist_gpush:co, \clist_gpush:cx
+% }
+%   \begin{syntax}
+%     \cs{clist_gpush:Nn} \meta{comma list} \Arg{item}
+%   \end{syntax}
+%   Pushes the \meta{item} onto the end of the top of the
+%   \meta{comma list}. The assignment is global.
+% \end{function}
+%
+% \section{Viewing comma lists}
+%
+% \begin{function}{\clist_show:N, \clist_show:c}
+%   \begin{syntax}
+%     \cs{clist_show:N} \meta{comma list}
+%   \end{syntax}
+%   Displays the entries in the \meta{comma list} in the terminal.
+% \end{function}
+%
+% \section{Experimental comma list functions}
+%
+% This section contains functions which may or may not be retained, depending
+% on how useful they are found to be.
+%
+% \begin{function}[EXP]{\clist_length:N, \clist_length:c,\clist_length:n}
+%   \begin{syntax}
+%     \cs{clist_length:N} \meta{comma list}
+%   \end{syntax}
+%   Leaves the number of items in the \meta{comma list} in the input
+%   stream as an \meta{integer denotation}. The total number of items
+%   in a \meta{comma list} will include those which are empty and duplicates,
+%   \emph{i.e.}~every item in a \meta{comma list} is unique.
+% \end{function}
+%
+% \begin{function}[EXP]{\clist_item:Nn, \clist_item:cn, \clist_item:nn}
+%   \begin{syntax}
+%     \cs{clist_item:Nn} \meta{comma list} \Arg{integer expression}
+%   \end{syntax}
+%   Indexing items in the \meta{comma list} from $0$ at the top (left), this
+%   function will evaluate the \meta{integer expression} and leave the
+%   appropriate item from the comma list in the input stream. If the
+%   \meta{integer expression} is negative, indexing occurs from the
+%   bottom (right) of the comma list. When the \meta{integer expression}
+%   is larger than the number of items in the \meta{comma list} (as
+%   calculated by \cs{clist_length:N}) then the function will expand to
+%   nothing.
+% \end{function}
+%
+% \begin{function}
+%   {
+%     \clist_set_from_seq:NN, \clist_set_from_seq:cN,
+%     \clist_set_from_seq:Nc, \clist_set_from_seq:cc
+%   }
+%   \begin{syntax}
+%     \cs{clist_set_from_seq:NN} \meta{comma list} \meta{sequence}
+%   \end{syntax}
+%   Sets the \meta{comma list} within the current \TeX{} group to be equal
+%   to the content of the \meta{sequence}.
+% \end{function}
+%
+% \begin{function}
+%   {
+%     \clist_gset_from_seq:NN, \clist_gset_from_seq:cN,
+%     \clist_gset_from_seq:Nc, \clist_gset_from_seq:cc
+%   }
+%   \begin{syntax}
+%     \cs{clist_gset_from_seq:NN} \meta{comma list} \meta{sequence}
+%   \end{syntax}
+%   Sets the \meta{comma list} globally to equal to the content of the
+%   \meta{sequence}.
+% \end{function}
+%
+% \end{documentation}
+%
+% \begin{implementation}
+%
+% \section{\pkg{l3clist} implementation}
+%
+% \TestFiles{m3clist002}
+%
+%    \begin{macrocode}
+%<*initex|package>
+%    \end{macrocode}
+%
+%    \begin{macrocode}
+%<*package>
+\ProvidesExplPackage
+  {\ExplFileName}{\ExplFileDate}{\ExplFileVersion}{\ExplFileDescription}
+\package_check_loaded_expl:
+%</package>
+%    \end{macrocode}
+%
+% \begin{variable}{\l_clist_tmpa_tl, \l_clist_tmpb_tl}
+%   Scratch space for various internal uses.
+%    \begin{macrocode}
+\tl_new:N \l_clist_tmpa_tl
+\tl_new:N \l_clist_tmpb_tl
+%    \end{macrocode}
+% \end{variable}
+%
+% \subsection{Allocation and initialisation}
+%
+% \begin{macro}{\clist_new:N,\clist_new:c}
+% \UnitTested
+%   Internally, comma lists are just token lists.
+%    \begin{macrocode}
+\cs_new_eq:NN \clist_new:N \tl_new:N
+\cs_new_eq:NN \clist_new:c \tl_new:c
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\clist_clear:N, \clist_clear:c}
+% \UnitTested
+% \begin{macro}{\clist_gclear:N, \clist_gclear:c}
+% \UnitTested
+%   Clearing comma lists is just the same as clearing token lists.
+%    \begin{macrocode}
+\cs_new_eq:NN \clist_clear:N  \tl_clear:N
+\cs_new_eq:NN \clist_clear:c  \tl_clear:c
+\cs_new_eq:NN \clist_gclear:N \tl_gclear:N
+\cs_new_eq:NN \clist_gclear:c \tl_gclear:c
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\clist_clear_new:N, \clist_clear_new:c}
+% \UnitTested
+% \begin{macro}{\clist_gclear_new:N, \clist_gclear_new:c}
+% \UnitTested
+%   Once again a copy from the token list functions.
+%    \begin{macrocode}
+\cs_new_eq:NN \clist_clear_new:N  \tl_clear_new:N
+\cs_new_eq:NN \clist_clear_new:c  \tl_clear_new:c
+\cs_new_eq:NN \clist_gclear_new:N \tl_gclear_new:N
+\cs_new_eq:NN \clist_gclear_new:c \tl_gclear_new:c
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}
+%   {\clist_set_eq:NN, \clist_set_eq:cN, \clist_set_eq:Nc, \clist_set_eq:cc}
+% \UnitTested
+% \begin{macro}
+%   {
+%     \clist_gset_eq:NN, \clist_gset_eq:cN,
+%     \clist_gset_eq:Nc, \clist_gset_eq:cc
+%   }
+% \UnitTested
+%   Once again, these are simple copies from the token list functions.
+%    \begin{macrocode}
+\cs_new_eq:NN \clist_set_eq:NN  \tl_set_eq:NN
+\cs_new_eq:NN \clist_set_eq:Nc  \tl_set_eq:Nc
+\cs_new_eq:NN \clist_set_eq:cN  \tl_set_eq:cN
+\cs_new_eq:NN \clist_set_eq:cc  \tl_set_eq:cc
+\cs_new_eq:NN \clist_gset_eq:NN \tl_gset_eq:NN
+\cs_new_eq:NN \clist_gset_eq:Nc \tl_gset_eq:Nc
+\cs_new_eq:NN \clist_gset_eq:cN \tl_gset_eq:cN
+\cs_new_eq:NN \clist_gset_eq:cc \tl_gset_eq:cc
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\clist_concat:NNN, \clist_concat:ccc}
+% \UnitTested
+% \begin{macro}{\clist_gconcat:NNN, \clist_gconcat:ccc}
+% \UnitTested
+% \begin{macro}[aux]{\clist_concat_aux:NNNN}
+%   Concatenating sequences is not quite as easy as it seems, as there is
+%   the danger that |#1| may be the same as either |#2| or |#3|. Also,
+%   there needs to be the correct addition of a comma to the output. So
+%   a little work to do.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \clist_concat:NNN
+  { \clist_concat_aux:NNNN \tl_set:Nx }
+\cs_new_protected_nopar:Npn \clist_gconcat:NNN
+  { \clist_concat_aux:NNNN \tl_gset:Nx }
+\cs_new_protected_nopar:Npn \clist_concat_aux:NNNN #1#2#3#4
+  {
+    #1 #2
+      {
+        \clist_if_empty:NTF #3
+          { \exp_not:o #4 }
+          {
+            \exp_not:o #3
+            \clist_if_empty:NF #4
+              {
+                ,
+                \exp_not:o #4
+              }
+          }
+      }
+  }
+\cs_generate_variant:Nn \clist_concat:NNN  { ccc }
+\cs_generate_variant:Nn \clist_gconcat:NNN { ccc }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \subsection{Appending items to comma lists}
+%
+% \begin{macro}
+%   {
+%     \clist_put_left:Nn, \clist_put_left:NV,
+%     \clist_put_left:No, \clist_put_left:Nx,
+%     \clist_put_left:cn, \clist_put_left:cV,
+%     \clist_put_left:co, \clist_put_left:cx
+%   }
+% \UnitTested
+% \begin{macro}
+%   {
+%     \clist_gput_left:Nn, \clist_gput_left:NV,
+%     \clist_gput_left:No, \clist_gput_left:Nx,
+%     \clist_gput_left:cn, \clist_gput_left:cV,
+%     \clist_gput_left:co, \clist_gput_left:cx
+%   }
+% \UnitTested
+%   Comma lists cannot hold empty values: there are therefore a couple
+%   of sanity checks to avoid accumulating commas.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \clist_put_left:Nn
+  { \clist_put_aux:NNnnNn \tl_set:Nn \tl_put_left:Nn { } , }
+\cs_new_protected_nopar:Npn \clist_gput_left:Nn
+  { \clist_put_aux:NNnnNn \tl_gset:Nn \tl_gput_left:Nn { } , }
+\cs_new_protected:Npn \clist_put_aux:NNnnNn #1#2#3#4#5#6
+  {
+    \clist_if_empty:NTF #5
+      { #1 #5 {#6} }
+      { \tl_if_empty:nF {#6} { #2 #5 { #3 #6 #4 } } }
+  }
+\cs_generate_variant:Nn \clist_put_left:Nn  {     NV , No , Nx }
+\cs_generate_variant:Nn \clist_put_left:Nn  { c , cV , co , cx }
+\cs_generate_variant:Nn \clist_gput_left:Nn {     NV , No , Nx }
+\cs_generate_variant:Nn \clist_gput_left:Nn { c , cV , co , cx }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}
+%   {
+%     \clist_put_right:Nn, \clist_put_right:NV,
+%     \clist_put_right:No, \clist_put_right:Nx,
+%     \clist_put_right:cn, \clist_put_right:cV,
+%     \clist_put_right:co, \clist_put_right:cx
+%   }
+% \begin{macro}
+%   {
+%     \clist_gput_right:Nn, \clist_gput_right:NV,
+%     \clist_gput_right:No, \clist_gput_right:Nx,
+%     \clist_gput_right:cn, \clist_gput_right:cV,
+%     \clist_gput_right:co, \clist_gput_right:cx
+%   }
+%   The same for the right side.
+%    \begin{macrocode}
+\cs_new_protected:Npn \clist_put_right:Nn
+  {  \clist_put_aux:NNnnNn \tl_set:Nn \tl_put_right:Nn , { } }
+\cs_new_protected_nopar:Npn \clist_gput_right:Nn
+  { \clist_put_aux:NNnnNn \tl_gset:Nn \tl_gput_right:Nn , { } }
+\cs_generate_variant:Nn \clist_put_right:Nn  {     NV , No , Nx }
+\cs_generate_variant:Nn \clist_put_right:Nn  { c , cV , co , cx }
+\cs_generate_variant:Nn \clist_gput_right:Nn {     NV , No , Nx }
+\cs_generate_variant:Nn \clist_gput_right:Nn { c , cV , co , cx }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \subsection{Comma lists as stacks}
+%
+% \begin{macro}{\clist_get:NN, \clist_get:cN}
+% \UnitTested
+% \begin{macro}[aux]{\clist_get_aux:wN}
+%   Getting an item from the left of a comma list is pretty easy: just
+%   trim off the first item using the comma.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \clist_get:NN #1#2
+  { \exp_after:wN \clist_get_aux:wN #1 , \q_stop #2 }
+\cs_new_protected:Npn \clist_get_aux:wN #1 , #2 \q_stop #3
+  { \tl_set:Nn #3 {#1} }
+\cs_generate_variant:Nn \clist_get:NN { c }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\clist_pop:NN, \clist_pop:cN}
+% \UnitTested
+% \begin{macro}{\clist_gpop:NN, \clist_gpop:cN}
+% \UnitTested
+% \begin{macro}[aux]{\clist_pop_aux:NNN}
+% \begin{macro}[aux]{\clist_pop_aux:wNNN}
+%   The aim here is to get the popped item as |#1| in the auxiliary, with
+%   |#2| containing either the remainder of the list \emph{or} \cs{q_nil}
+%   if there were insufficient items. That keeps the number of auxiliary
+%   functions down.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \clist_pop:NN
+  { \clist_pop_aux:NNN \tl_set:Nn }
+\cs_new_protected_nopar:Npn \clist_gpop:NN
+  { \clist_pop_aux:NNN \tl_gset:Nn }
+\cs_new_protected_nopar:Npn \clist_pop_aux:NNN #1#2#3
+  { \exp_after:wN \clist_pop_aux:wNNN #2 , \q_nil , \q_nil , \q_stop #1#2#3 }
+\cs_new_protected:Npn \clist_pop_aux:wNNN #1 , #2 , \q_nil , #3 \q_stop #4#5#6
+  {
+    \quark_if_nil:nTF {#2}
+      { #4 #5 { } }
+      { #4 #5 {#2} }
+    \tl_set:Nn #6 {#1}
+  }
+\cs_generate_variant:Nn \clist_pop:NN  { c }
+\cs_generate_variant:Nn \clist_gpop:NN { c }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{
+%   \clist_push:Nn, \clist_push:NV, \clist_push:No, \clist_push:Nx,
+%   \clist_push:cn, \clist_push:cV, \clist_push:co, \clist_push:cx
+% }
+% \UnitTested
+% \begin{macro}{
+%   \clist_gpush:Nn, \clist_gpush:NV, \clist_gpush:No, \clist_gpush:Nx,
+%   \clist_gpush:cn, \clist_gpush:cV, \clist_gpush:co, \clist_gpush:cx
+% }
+% \UnitTested
+%   Pushing to a sequence is the same as adding on the left.
+%    \begin{macrocode}
+\cs_new_eq:NN \clist_push:Nn  \clist_put_left:Nn
+\cs_new_eq:NN \clist_push:NV  \clist_put_left:NV
+\cs_new_eq:NN \clist_push:No  \clist_put_left:No
+\cs_new_eq:NN \clist_push:Nx  \clist_put_left:Nx
+\cs_new_eq:NN \clist_push:cn  \clist_put_left:cn
+\cs_new_eq:NN \clist_push:cV  \clist_put_left:cV
+\cs_new_eq:NN \clist_push:co  \clist_put_left:co
+\cs_new_eq:NN \clist_push:cx  \clist_put_left:cx
+\cs_new_eq:NN \clist_gpush:Nn \clist_gput_left:Nn
+\cs_new_eq:NN \clist_gpush:NV \clist_gput_left:NV
+\cs_new_eq:NN \clist_gpush:No \clist_gput_left:No
+\cs_new_eq:NN \clist_gpush:Nx \clist_gput_left:Nx
+\cs_new_eq:NN \clist_gpush:cn \clist_gput_left:cn
+\cs_new_eq:NN \clist_gpush:cV \clist_gput_left:cV
+\cs_new_eq:NN \clist_gpush:co \clist_gput_left:co
+\cs_new_eq:NN \clist_gpush:cx \clist_gput_left:cx
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \subsection{Using comma lists}
+%
+% \begin{macro}{\clist_use:N, \clist_use:c}
+% \UnitTested
+%   The approach is the same as for \cs{tl_use:N}.
+%    \begin{macrocode}
+\cs_new_eq:NN \clist_use:N \tl_use:N
+\cs_new_eq:NN \clist_use:c \tl_use:c
+%    \end{macrocode}
+% \end{macro}
+%
+% \subsection{Modifying comma lists}
+%
+% \begin{variable}{\l_clist_remove_clist}
+%   An internal comma list for the removal routines.
+%    \begin{macrocode}
+\clist_new:N \l_clist_remove_clist
+%    \end{macrocode}
+% \end{variable}
+%
+% \begin{macro}{\clist_remove_duplicates:N, \clist_remove_duplicates:c}
+% \UnitTested
+% \begin{macro}{\clist_gremove_duplicates:N, \clist_gremove_duplicates:c}
+% \UnitTested
+% \begin{macro}[aux]{\clist_remove_duplicates_aux:NN}
+%   Removing duplicates means making a new list then copying it.
+%    \begin{macrocode}
+\cs_new_protected:Npn \clist_remove_duplicates:N
+  { \clist_remove_duplicates_aux:NN \clist_set_eq:NN }
+\cs_new_protected:Npn \clist_gremove_duplicates:N
+  { \clist_remove_duplicates_aux:NN \clist_gset_eq:NN }
+\cs_new_protected:Npn \clist_remove_duplicates_aux:NN #1#2
+  {
+    \clist_clear:N \l_clist_remove_clist
+    \clist_map_inline:Nn #2
+      {
+        \clist_if_in:NnF \l_clist_remove_clist {##1}
+          { \clist_put_right:Nn \l_clist_remove_clist {##1} }
+      }
+    #1 #2 \l_clist_remove_clist
+  }
+\cs_generate_variant:Nn \clist_remove_duplicates:N  { c }
+\cs_generate_variant:Nn \clist_gremove_duplicates:N { c }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\clist_remove_all:Nn, \clist_remove_all:cn}
+% \UnitTested
+% \begin{macro}{\clist_gremove_all:Nn, \clist_gremove_all:cn}
+% \UnitTested
+% \begin{macro}[aux]{\clist_remove_all_aux:NNn}
+% \begin{macro}[aux]{\clist_remove_all_aux:w}
+%   Removing an item from a comma list is done without looping over
+%   the entire list, as the performance of that approach is very bad for
+%   long lists. Instead, a delimited function is needed. For this to work
+%   correctly, there is a need to add an additional comma at the start of
+%   the list, and to remove it again once the removal is complete. Of
+%   course, the list can end up empty, which is the reason for the test
+%   before copying back to the parent.
+%    \begin{macrocode}
+\cs_new_protected:Npn \clist_remove_all:Nn
+  { \clist_remove_all_aux:NNn \clist_set_eq:NN }
+\cs_new_protected:Npn \clist_gremove_all:Nn
+  { \clist_remove_all_aux:NNn \clist_gset_eq:NN }
+\cs_new_protected:Npn \clist_remove_all_aux:NNn #1#2#3
+  {
+    \clist_if_empty:NF #2
+      {
+        \clist_clear:N \l_clist_remove_clist
+        \cs_set_protected:Npn \clist_remove_all_aux:w
+          ##1 , #3 , ##2 \q_stop
+          {
+            \tl_put_right:Nn \l_clist_remove_clist {##1}
+            \quark_if_no_value:nF {##2}
+              { \clist_remove_all_aux:w , ##2 \q_stop }
+          }
+        \exp_after:wN \clist_remove_all_aux:w
+          \exp_after:wN , #2 , #3 , \q_no_value \q_stop
+        \tl_if_empty:NF \l_clist_remove_clist
+          {
+            \exp_after:wN \tl_set:No \exp_after:wN
+              \l_clist_remove_clist \exp_after:wN
+                { \exp_after:wN \use_none:n \l_clist_remove_clist }
+          }
+        #1 #2 \l_clist_remove_clist
+      }
+  }
+\cs_new_protected:Npn \clist_remove_all_aux:w { }
+\cs_generate_variant:Nn \clist_remove_all:Nn  { c }
+\cs_generate_variant:Nn \clist_gremove_all:Nn { c }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% 
+% \begin{macro}{\clist_trim_spaces:n}
+% \begin{macro}
+%   {
+%     \clist_trim_spaces:N, \clist_trim_spaces:c,
+%     \clist_gtrim_spaces:N, \clist_gtrim_spaces:c
+%   }
+% \begin{macro}{\clist_trim_spaces_aux_i:n}
+% \begin{macro}{\clist_trim_spaces_aux_ii:n}
+%   Here, the basic plan is to use \cs{tl_trim_spaces:n} to do the work:
+%   the only issue is to make sure that the number of commas at the end of
+%   the process is correct.
+%    \begin{macrocode}
+\cs_new:Npn \clist_trim_spaces:n #1
+  {
+    \exp_args:Nf \clist_trim_spaces_aux_i:n
+      { \clist_map_function:nN {#1} \clist_trim_spaces_aux_ii:n }
+  }
+\cs_new:Npn \clist_trim_spaces_aux_i:n #1 { \use_ii:nn #1 }
+\cs_new:Npn \clist_trim_spaces_aux_ii:n #1
+  { , \tl_trim_spaces:n {#1} }
+\cs_new_protected:Npn \clist_trim_spaces:N #1
+  { \tl_set:Nf #1 { \exp_args:No \clist_trim_spaces:n #1 } }
+\cs_new_protected:Npn \clist_gtrim_spaces:N #1
+  { \tl_gset:Nf #1 { \exp_args:No \clist_trim_spaces:n #1 } }
+\cs_generate_variant:Nn \clist_trim_spaces:N  { c }
+\cs_generate_variant:Nn \clist_gtrim_spaces:N { c }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \subsection{Comma list conditionals}
+%
+% \begin{macro}[aux]{\clist_tmp:w}
+%   A temporary function for comparison.
+%    \begin{macrocode}
+\cs_new_protected:Npn \clist_tmp:w { }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}[pTF]{\clist_if_empty:N, \clist_if_empty:c}
+% \UnitTested
+%   Simple copies from the token list variable material.
+%    \begin{macrocode}
+\prg_new_eq_conditional:NNn \clist_if_empty:N \tl_if_empty:N { p , T , F , TF }
+\prg_new_eq_conditional:NNn \clist_if_empty:c \tl_if_empty:c { p , T , F , TF }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}[pTF]
+%   {\clist_if_eq:NN, \clist_if_eq:Nc, \clist_if_eq:cN, \clist_if_eq:cc}
+% \UnitTested
+%   Simple copies from the token list variable material.
+%    \begin{macrocode}
+\prg_new_eq_conditional:NNn \clist_if_eq:NN \tl_if_eq:NN { p , T , F , TF }
+\prg_new_eq_conditional:NNn \clist_if_eq:Nc \tl_if_eq:Nc { p , T , F , TF }
+\prg_new_eq_conditional:NNn \clist_if_eq:cN \tl_if_eq:cN { p , T , F , TF }
+\prg_new_eq_conditional:NNn \clist_if_eq:cc \tl_if_eq:cc { p , T , F , TF }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}[TF]
+%   {
+%     \clist_if_in:Nn, \clist_if_in:NV, \clist_if_in:No,
+%     \clist_if_in:cn, \clist_if_in:cV, \clist_if_in:co
+%   }
+% \UnitTested
+%    \begin{macrocode}
+\prg_new_protected_conditional:Npnn \clist_if_in:Nn #1#2
+  { T , F , TF }
+  {
+    \cs_set_protected:Npn \clist_tmp:w ##1 , #2 , ##2##3 \q_stop
+      {
+        \if_meaning:w \q_no_value ##2
+          \prg_return_false:
+        \else:
+           \prg_return_true:
+        \fi:
+      }
+    \exp_last_unbraced:NNo \clist_tmp:w , #1 , #2 , \q_no_value \q_stop
+  }
+\prg_new_protected_conditional:Npnn \clist_if_in:nn #1#2
+  { T , F , TF }
+  {
+    \cs_set_protected:Npn \clist_tmp:w ##1 , #2 , ##2##3 \q_stop
+      {
+        \if_meaning:w \q_no_value ##2
+          \prg_return_false:
+        \else:
+           \prg_return_true:
+        \fi:
+      }
+    \clist_tmp:w , #1 , #2 , \q_no_value \q_stop
+  }
+\cs_generate_variant:Nn \clist_if_in:NnT  {     NV , No }
+\cs_generate_variant:Nn \clist_if_in:NnT  { c , cV , co }
+\cs_generate_variant:Nn \clist_if_in:NnF  {     NV , No }
+\cs_generate_variant:Nn \clist_if_in:NnF  { c , cV , co }
+\cs_generate_variant:Nn \clist_if_in:NnTF {     NV , No }
+\cs_generate_variant:Nn \clist_if_in:NnTF { c , cV , co }
+\cs_generate_variant:Nn \clist_if_in:nnT  {     nV , no }
+\cs_generate_variant:Nn \clist_if_in:nnF  {     nV , no }
+\cs_generate_variant:Nn \clist_if_in:nnTF {     nV , no }
+%    \end{macrocode}
+% \end{macro}
+%
+% \subsection{Mapping to comma lists}
+%
+% \begin{macro}{\clist_map_function:NN, \clist_map_function:cN}
+% \UnitTested
+% \begin{macro}{\clist_map_function:nN}
+% \UnitTested
+% \begin{macro}[aux]{\clist_map_function_aux:Nw}
+%   Mapping to comma lists is pretty simple, if not massively efficient.
+%    \begin{macrocode}
+\cs_new_nopar:Npn \clist_map_function:NN #1#2
+  {
+    \clist_if_empty:NF #1
+      {
+        \exp_last_unbraced:NNo \clist_map_function_aux:Nw #2 #1
+          , \q_recursion_tail , \q_recursion_stop
+      }
+  }
+\cs_new:Npn \clist_map_function:nN #1#2
+  {
+    \tl_if_empty:nF {#1}
+      {
+        \clist_map_function_aux:Nw #2 #1
+          , \q_recursion_tail , \q_recursion_stop
+      }
+  }
+\cs_new:Npn \clist_map_function_aux:Nw #1#2 ,
+  {
+    \quark_if_recursion_tail_stop:n {#2}
+    #1 {#2}
+    \clist_map_function_aux:Nw #1
+  }
+\cs_generate_variant:Nn \clist_map_function:NN { c }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{variable}{\g_clist_map_inline_int}
+%   For the nesting of mappings.
+%    \begin{macrocode}
+\int_new:N \g_clist_map_inline_int
+%    \end{macrocode}
+% \end{variable}
+%
+% \begin{macro}{\clist_map_inline:Nn, \clist_map_inline:cn}
+% \UnitTested
+% \begin{macro}{\clist_map_inline:nn}
+% \UnitTested
+%   Inline mapping is done by creating a suitable function
+%   \enquote{on the fly}:
+%   this is done globally to avoid any issues with \TeX{}'s groups.
+%    \begin{macrocode}
+\cs_new_protected:Npn \clist_map_inline:Nn #1#2
+  {
+    \int_gincr:N \g_clist_map_inline_int
+    \cs_gset:cpn { clist_map_inline_ \int_use:N \g_clist_map_inline_int :n }
+      ##1
+      {#2}
+    \exp_args:NNc \clist_map_function:NN #1
+      { clist_map_inline_ \int_use:N \g_clist_map_inline_int :n }
+    \int_gdecr:N \g_clist_map_inline_int
+  }
+\cs_new_protected:Npn \clist_map_inline:nn #1#2
+  {
+    \int_gincr:N \g_clist_map_inline_int
+    \cs_gset:cpn { clist_map_inline_ \int_use:N \g_clist_map_inline_int :n }
+      ##1
+      {#2}
+    \exp_args:Nnc \clist_map_function:nN {#1}
+      { clist_map_inline_ \int_use:N \g_clist_map_inline_int :n }
+    \int_gdecr:N \g_clist_map_inline_int
+  }
+\cs_generate_variant:Nn \clist_map_inline:Nn { c }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\clist_map_variable:NNn, \clist_map_variable:cNn}
+% \UnitTested
+%   This is just a dedicated version of the inline mapping.
+%    \begin{macrocode}
+\cs_new_protected:Npn \clist_map_variable:NNn #1#2#3
+  {
+    \clist_map_inline:Nn #1
+      {
+        \tl_set:Nn #2 {##1}
+        #3
+      }
+  }
+\cs_new_protected:Npn \clist_map_variable:nNn #1#2#3
+  {
+    \clist_map_inline:nn {#1}
+      {
+        \tl_set:Nn #2 {##1}
+        #3
+      }
+  }
+\cs_generate_variant:Nn \clist_map_variable:NNn { c }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\clist_map_break:}
+% \UnitTested
+% \begin{macro}{\clist_map_break:n}
+%   Both are simple renaming.
+%    \begin{macrocode}
+\cs_new_eq:NN \clist_map_break:  \use_none_delimit_by_q_recursion_stop:w
+\cs_new_eq:NN \clist_map_break:n \use_i_delimit_by_q_recursion_stop:nw
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \section{Viewing comma lists}
+%
+% \begin{macro}{\clist_show:N, \clist_show:c}
+% \UnitTested
+% \begin{macro}[aux]{\clist_show_aux:n}
+% \begin{macro}[aux]{\clist_show_aux:w}
+%   The aim of the mapping here is to create a token list containing the
+%   formatted comma list. The very first item needs the new line and \verb*|> |
+%   removing, which is achieved using a \texttt{w}-type auxiliary. To avoid
+%   a low-level \TeX{} error if there is an empty comma list, a simple test is
+%   used to keep the output \enquote{clean}.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \clist_show:N #1
+  {
+    \clist_if_empty:NTF #1
+      {
+        \iow_term:x { Comma~list~\token_to_str:N #1 \c_space_tl is~empty }
+        \tl_show:n { }
+      }
+      {
+        \iow_term:x
+          {
+            Comma~list~\token_to_str:N #1 \c_space_tl
+            contains~the~items~(without~outer~braces):
+          }
+        \tl_set:Nx \l_clist_show_tl
+          { \clist_map_function:NN #1 \clist_show_aux:n }
+          \etex_showtokens:D \exp_after:wN \exp_after:wN \exp_after:wN
+            { \exp_after:wN \clist_show_aux:w \l_clist_show_tl }
+      }
+  }
+\cs_new:Npn \clist_show_aux:n #1
+  {
+    \iow_newline: > \c_space_tl \c_space_tl
+    \iow_char:N \{ \exp_not:n {#1} \iow_char:N \}
+  }
+\cs_new:Npn \clist_show_aux:w #1 > ~ { }
+\cs_generate_variant:Nn \clist_show:N { c }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \subsection{Experimental functions}
+%
+% \begin{macro}{\clist_length:N, \clist_length:c}
+% \begin{macro}{\clist_length:n}
+% \begin{macro}[aux]{\clist_length_aux:n}
+%   Counting the items in a comma list is done using the same approach as for
+%   other length functions: turn each entry into a \texttt{+1} then use
+%   integer evaluation to actually do the mathematics.
+%    \begin{macrocode}
+\cs_new:Npn \clist_length:N #1
+  {
+    \int_eval:n
+      {
+        0
+        \clist_map_function:NN #1 \clist_length_aux:n
+      }
+  }
+\cs_new:Npn \clist_length:n #1
+  {
+    \int_eval:n
+      {
+        0
+        \clist_map_function:nN {#1} \clist_length_aux:n
+      }
+  }
+\cs_new:Npn \clist_length_aux:n #1 { +1 }
+\cs_generate_variant:Nn \clist_length:N { c }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\clist_item:Nn, \clist_item:cn}
+% \begin{macro}{\clist_item:nn}
+% \begin{macro}[aux]{\clist_item_aux:nnn}
+%   The idea here is to find the offset of the item from the left, then use
+%   a loop to grab the correct item. If the resulting offset is too large,
+%   then \cs{quark_if_recursion_stop:n} will be true,
+%   terminating the loop and returning nothing at all.
+%    \begin{macrocode}
+\cs_set_nopar:Npn \clist_item:Nn #1#2
+  { \exp_args:No \clist_item:nn #1 {#2} }
+\cs_set:Npn \clist_item:nn #1#2
+  {
+    \int_compare:nNnTF {#2} < \c_zero
+      {
+        \exp_args:Nf \clist_item_aux:nw
+          { \int_eval:n { \clist_length:n {#1} + #2 } }
+          #1 , \q_recursion_tail \q_recursion_stop
+      }
+      { \clist_item_aux:nw {#2} #1 , \q_recursion_tail \q_recursion_stop }
+  }
+\cs_set:Npn \clist_item_aux:nw #1#2 , #3
+  {
+    \int_compare:nNnTF {#1} = \c_zero
+      { \use_i_delimit_by_q_recursion_stop:nw {#2} }
+      {
+        \quark_if_recursion_tail_stop:n {#3}
+        \exp_args:Nf \clist_item_aux:nw
+          { \int_eval:n { #1 - 1 } }
+          #3
+      }
+  }
+\cs_generate_variant:Nn \clist_item:Nn { c }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}
+%   {
+%     \clist_set_from_seq:NN, \clist_set_from_seq:cN,
+%     \clist_set_from_seq:Nc, \clist_set_from_seq:cc
+%   }
+% \begin{macro}
+%   {
+%     \clist_gset_from_seq:NN, \clist_gset_from_seq:cN,
+%     \clist_gset_from_seq:Nc, \clist_gset_from_seq:cc
+%   }
+%   Setting a comma list from a comma-separated list is done using a simple
+%   mapping. We wrap each item with braces, \cs{exp_not:n}, and a comma. The
+%   first comma must be removed, except in the case of an empty comma-list.
+%    \begin{macrocode}
+\cs_new_protected:Npn \clist_set_from_seq:NN #1#2
+  {
+    \seq_if_empty:NTF #2
+      { \clist_clear:N #1 }
+      {
+        \seq_push_item_def:n { , \exp_not:n {{##1}} }
+        \tl_set:Nx #1
+          { \exp_after:wN \use_none:n \tex_romannumeral:D -`\0 #2 }
+        \seq_pop_item_def:
+      }
+  }
+\cs_new_protected:Npn \clist_gset_from_seq:NN #1#2
+  {
+    \seq_if_empty:NTF #2
+      { \clist_gclear:N #1 }
+      {
+        \seq_push_item_def:n { , \exp_not:n {##1} }
+        \tl_gset:Nx #1
+          { \exp_after:wN \use_none:n \tex_romannumeral:D -`\0 #2 }
+        \seq_pop_item_def:
+      }
+  }
+\cs_generate_variant:Nn \clist_set_from_seq:NN  {     Nc }
+\cs_generate_variant:Nn \clist_set_from_seq:NN  { c , cc }
+\cs_generate_variant:Nn \clist_gset_from_seq:NN {     Nc }
+\cs_generate_variant:Nn \clist_gset_from_seq:NN { c , cc }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \subsection{Deprecated interfaces}
+%
+% Deprecated on 2011-05-27, for removal by 2011-08-31.
+%
+% \begin{macro}{\clist_top:NN, \clist_top:cN}
+%   These are old stack functions.
+%    \begin{macrocode}
+\cs_new_eq:NN \clist_top:NN \clist_get:NN
+\cs_new_eq:NN \clist_top:cN \clist_get:cN
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\clist_remove_element:Nn}
+% \begin{macro}{\clist_gremove_element:Nn}
+%   An older name for \cs{clist_remove_all:Nn}.
+%    \begin{macrocode}
+\cs_new_eq:NN \clist_remove_element:Nn  \clist_remove_all:Nn
+\cs_new_eq:NN \clist_gremove_element:Nn \clist_gremove_all:Nn
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\clist_display:N, \clist_display:c}
+%   An older name for \cs{clist_show:N}.
+%    \begin{macrocode}
+\cs_new_eq:NN \clist_display:N \clist_show:N
+\cs_new_eq:NN \clist_display:c \clist_show:c
+%    \end{macrocode}
+% \end{macro}
+%
+%    \begin{macrocode}
+%</initex|package>
+%    \end{macrocode}
+%
+% \end{implementation}
+%
+% \PrintIndex
\ No newline at end of file
diff --git a/Master/texmf-dist/source/latex/l3kernel/l3doc.dtx b/Master/texmf-dist/source/latex/l3kernel/l3doc.dtx
new file mode 100644
index 00000000000..5039b8c0f62
--- /dev/null
+++ b/Master/texmf-dist/source/latex/l3kernel/l3doc.dtx
@@ -0,0 +1,2064 @@
+% \iffalse meta-comment
+%
+%% File: l3doc.dtx Copyright (C) 1990-2011 The LaTeX3 project
+%%
+%% It may be distributed and/or modified under the conditions of the
+%% LaTeX Project Public License (LPPL), either version 1.3c of this
+%% license or (at your option) any later version.  The latest version
+%% of this license is in the file
+%%
+%%    http://www.latex-project.org/lppl.txt
+%%
+%% This file is part of the "expl3 bundle" (The Work in LPPL)
+%% and all files in that bundle must be distributed together.
+%%
+%% The released version of this bundle is available from CTAN.
+%%
+%% -----------------------------------------------------------------------
+%%
+%% The development version of the bundle can be found at
+%%
+%%    http://www.latex-project.org/svnroot/experimental/trunk/
+%%
+%% for those people who are interested.
+%%
+%%%%%%%%%%%
+%% NOTE: %%
+%%%%%%%%%%%
+%%
+%%   Snapshots taken from the repository represent work in progress and may
+%%   not work or may contain conflicting material!  We therefore ask
+%%   people _not_ to put them into distributions, archives, etc. without
+%%   prior consultation with the LaTeX3 Project.
+%%
+%% -----------------------------------------------------------------------
+%
+%<*driver>
+\def\nameofplainTeX{plain}
+\ifx\fmtname\nameofplainTeX\else
+  \expandafter\begingroup
+\fi
+\input docstrip.tex
+\askforoverwritefalse
+\preamble
+
+
+EXPERIMENTAL CODE
+
+Do not distribute this file without also distributing the
+source files specified above.
+
+Do not distribute a modified version of this file.
+
+
+\endpreamble
+% stop docstrip adding \endinput
+\postamble
+\endpostamble
+\generate{\file{l3doc.cls}{\from{l3doc.dtx}{class}}}
+\generate{\file{l3doc.ist}{\from{l3doc.dtx}{docist}}}
+\ifx\fmtname\nameofplainTeX
+  \expandafter\endbatchfile
+\else
+  \expandafter\endgroup
+\fi
+%</driver>
+%
+%<*driver|class>
+\RequirePackage{l3names}
+%</driver|class>
+%
+% Need to protect the file metadata for any modules that load l3doc.
+% This is restored after "\ProvideExplClass" below.
+%    \begin{macrocode}
+%<class>\let        \filenameOld        \ExplFileName
+%<class>\let        \filedateOld        \ExplFileDate
+%<class>\let     \fileversionOld        \ExplFileVersion
+%<class>\let \filedescriptionOld        \ExplFileDescription
+%    \end{macrocode}
+%
+%<*driver|class>
+\GetIdInfo$Id: l3doc.dtx 2478 2011-06-19 21:34:23Z joseph $
+          {L3 Experimental documentation class}
+%</driver|class>
+%
+%<*driver>
+\ProvidesFile{\ExplFileName.dtx}
+  [\ExplFileDate\space v\ExplFileVersion\space\ExplFileDescription]
+\documentclass{l3doc}
+\usepackage{framed}
+\begin{document}
+  \DocInput{l3doc.dtx}
+\end{document}
+%</driver>
+%
+% This isn't included in the typeset documentation because it's a bit ugly:
+%<*class>
+\ProvidesExplClass
+  {\ExplFileName}{\ExplFileDate}{\ExplFileVersion}{\ExplFileDescription}
+\RequirePackage{expl3,xparse}
+\let        \ExplFileName        \filenameOld
+\let        \ExplFileDate        \filedateOld
+\let        \ExplFileVersion     \fileversionOld
+\let        \ExplFileDescription \filedescriptionOld
+%</class>
+% \fi
+%
+% \title{The \pkg{l3doc} class\thanks{This file
+%         has version number v\ExplFileVersion, last
+%         revised \ExplFileDate.}}
+% \author{\Team}
+% \date{\ExplFileDate}
+% \maketitle
+% \tableofcontents
+%
+% \begin{documentation}
+%
+% \section{Introduction}
+%
+% This is an ad-hoc class for documenting the \pkg{expl3} bundle,
+% a collection of modules or packages that make up \LaTeX3's programming
+% environment. Eventually it will replace the "ltxdoc" class for \LaTeX3,
+% but not before the good ideas in \pkg{hypdoc}, \cls{xdoc2}, \pkg{docmfp}, and
+% \cls{gmdoc} are incorporated.
+%
+% \textbf{It is even less stable than the main \pkg{expl3} packages. Use at own risk!}
+%
+% It is written as a `self-contained' docstrip file: executing
+% "latex l3doc.dtx"
+% will generate the "l3doc.cls" file and typeset this
+% documentation; execute "tex l3doc.dtx" to only generate the ".cls" file.
+%
+% \section{Features of other packages}
+%
+% This class builds on the \pkg{ltxdoc} class and the \pkg{doc} package, but
+% in the time since they were originally written some
+% improvements and replacements have appeared that we would like to use as
+% inspiration.
+%
+% These packages or classes are \pkg{hypdoc}, \pkg{docmfp}, \pkg{gmdoc},
+% and \pkg{xdoc}. I have summarised them below in order to work out what
+% sort of features we should aim at a minimum for \pkg{l3doc}.
+%
+% \subsection{The \pkg{hypdoc} package}
+%
+% This package provides hyperlink support for the \pkg{doc} package. I have
+% included it in this list to remind me that cross-referencing between
+% documentation and implementation of methods is not very good. (E.g., it
+% would be nice to be able to automatically hyperlink the documentation for
+% a function from its implementation and vice-versa.)
+%
+% \subsection{The \pkg{docmfp} package}
+%
+% \begin{itemize}
+% \item Provides "\DescribeRoutine" and the "routine" environment (etc.)
+%       for  MetaFont and MetaPost code.
+% \item Provides "\DescribeVariable" and the "variable" environment (etc.)
+%       for more general code.
+% \item Provides "\Describe" and the "Code" environment (etc.) as a
+%       generalisation of the above two instantiations.
+% \item Small tweaks to the DocStrip system to aid non-\LaTeX\ use.
+% \end{itemize}
+%
+% \subsection{The \pkg{xdoc2} package}
+%
+% \begin{itemize}
+% \item Two-sided printing.
+% \item "\NewMacroEnvironment", "\NewDescribeEnvironment"; similar idea
+%       to \pkg{docmfp} but more comprehensive.
+% \item Tons of small improvements.
+% \end{itemize}
+%
+% \subsection{The \pkg{gmdoc} package}
+%
+% Radical re-implementation of \pkg{doc} as a package or class.
+% \begin{itemize}
+% \item Requires no "\begin{macrocode}" blocks!
+% \item Automatically inserts "\begin{macro}" blocks!
+% \item And a whole bunch of other little things.
+% \end{itemize}
+%
+% \section{Problems \& Todo}
+%
+% Problems at the moment: (1)~not flexible in the types of things that can be documented; (2)~very nonstandard markup (e.g., the odd `"/ (...)"' tags; (3)~no obvious link between the "\begin{function}" environment for documenting things to the "\begin{macro}" function that's used analogously in the implementation.
+%
+% The "macro" should probably be renamed to "function" when it is used within
+% an implementation section. But they should have the same syntax before that happens!
+%
+% Furthermore, we need another `layer' of documentation commands to
+% account for `user-macro' as opposed to `code-functions'; the \pkg{expl3}
+% functions should be documented differently, probably, to the \pkg{xparse}
+% user macros (at least in terms of indexing).
+%
+% In no particular order, a list of things to do:
+% \begin{itemize}
+% \item Rename \env{function}/\env{macro} environments to better describe
+%       their use.
+% \item Generalise \env{function}/\env{macro} for documenting `other things',
+%       such as environment names, package options, even keyval options.
+% \item Use \pkg{xparse}.
+% \item New function like "\part" but for files (remove awkward `File' as "\partname").
+% \item Something better to replace "\StopEventually"; I'm thinking two
+%       environments \env{documentation} and \env{implementation} that can
+%       conditionally typeset/ignore their material.
+%       (This has been implemented but needs further consideration.)
+% \item Hyperlink documentation and implementation of macros (see
+%       the \textsc{dtx} file of \pkg{svn-multi} v2 as an example).
+% \end{itemize}
+%
+% \section{Bugs}
+%
+% \begin{itemize}
+% \item Spaces are ignored entirely within \env{function} and \env{macro}
+%       arguments. This is just waiting for a convenient space-trimming
+%       macro in \pkg{expl3}.
+% \end{itemize}
+%
+% \section{Documentation}
+%
+% \subsection{Configuration}
+%
+% Before class options are processed, \pkg{l3doc} loads a configuration
+% file "l3doc.cfg" if it exists, allowing you to customise the behaviour of
+% the class without having to change the documentation source files.
+%
+% For example, to produce documentation on letter-sized paper instead of the
+% default A4 size, create |l3doc.cfg| and include the line
+% \begin{verbatim}
+% \PassOptionsToClass{letterpaper}{l3doc}
+% \end{verbatim}
+%
+% By default, \pkg{l3doc} selects the |T1| font encoding and loads the
+% Latin Modern fonts.
+% To prevent this, use the class option |cm-default|.
+%
+% \subsection{Partitioning documentation and implementation}
+%
+% \pkg{doc} uses the "\OnlyDocumentation"/"\AlsoImplementation" macros
+% to guide the use of "\StopEventually{}", which is intended to be placed
+% to partition the documentation and implementation within a single DTX file.
+% 
+% This isn't very flexible, since it assumes that we \emph{always} want
+% to print the documentation. For the \pkg{expl3} sources, I wanted to be
+% be able to input DTX files in two modes: only displaying the documentation,
+% and only displaying the implementation. For example:
+%
+% \begin{verbatim}
+% \DisableImplementation
+% \DocInput{l3basics,l3prg,...}
+% \EnableImplementation
+% \DisableDocumentation
+% \DocInputAgain
+% \end{verbatim}
+%
+% The idea being that the entire \pkg{expl3} bundle can be documented,
+% with the implementation included at the back. Now, this isn't perfect,
+% but it's a start.
+%
+% Use "\begin{documentation}...\end{documentation}" around the documentation,
+% and "\begin{implementation}...\end{implementation}" around the implementation.
+% The "\EnableDocumentation"/"\EnableImplementation" will cause them to be typeset
+% when the DTX file is "\DocInput"; use "\DisableDocumentation"/"\DisableImplementation"
+% to omit the contents of those environments.
+%
+% Note that \cmd\DocInput\ now takes comma-separated arguments, and \cmd\DocInputAgain\
+% can be used to re-input all DTX files previously input in this way.
+%
+% \subsection{Describing functions in the documentation}
+%
+% \DescribeEnv{function}
+% \DescribeEnv{syntax}
+% Two heavily-used environments are defined to describe the syntax
+% of \textsf{expl3} functions and variables.
+% \begin{framed}
+% \vspace{-\baselineskip}
+% \begin{verbatim}
+% \begin{function}{ list_of , functions }
+%   \begin{syntax}
+%     "\foo_bar:" \Arg{meta} <test1>
+%   \end{syntax}
+% <description>
+% \end{function}
+% \end{verbatim}
+% \vspace{-2\baselineskip}
+% \hrulefill
+% \begin{function}{ list_of , functions }
+%   \begin{syntax}
+%     "\foo_bar:" \Arg{meta} <test1>
+%   \end{syntax}
+% <description>
+% \end{function}
+% \end{framed}
+%
+% Function environments take an optional argument to indicate whether the function(s) it
+% describes are expandable or defined in conditional forms. Use "EXP", "TF", or "pTF"
+% for this; note that "pTF" implies "EXP" since predicates must always be expandable.
+% As an example:
+% \begin{framed}
+% \vspace{-\baselineskip}
+% \begin{verbatim}
+% \begin{function}[pTF]{ \cs_if_exist:N }
+%   \begin{syntax}
+%     "\cs_if_exist_p:N" <cs>
+%   \end{syntax}
+% <description>
+% \end{function}
+% \end{verbatim}
+% \vspace{-2\baselineskip}
+% \hrulefill
+% \begin{function}[pTF]{ \cs_if_exist:N }
+%   \begin{syntax}
+%     "\cs_if_exist_p:N" <cs>
+%   \end{syntax}
+% <description>
+% \end{function}
+% \end{framed}
+%
+% Note that the list of functions used to use "|" as a separator instead,
+% and this syntax is still supported.
+%
+% In the old syntax, individual functions could be suffixed by an optional flag or
+% two to indicate the same information given in the optional argument.
+% This is still supported; use "/" to separate the function name
+% from the flag(s) and then add any of "(EXP)", "(TF)", or "(pTF)".
+%
+% \DescribeEnv{variable}
+% If you are documenting a variable instead of a function, use the "variable" environment instead; it behaves identically to the "function" environment above.
+%
+% \subsection{Describing functions in the implementation}
+%
+% \DescribeEnv{macro}
+% The well-used environment from \LaTeXe\ for marking up the implementation
+% of macros/functions remains the \env{macro} environment.
+% Some changes in \pkg{l3doc}: it now accepts comma-separated lists
+% of functions, to avoid a very large number of consecutive "\end{macro}"
+% statements.
+% \begin{verbatim}
+% % \begin{macro}{\foo:N,\foo:c}
+% %   \begin{macrocode}
+% ... code for \foo:N and \foo:c ...
+% %   \end{macrocode}
+% % \end{macro}
+% \end{verbatim}
+% If you are documenting an auxiliary macro, it's generally not necessary
+% to highlight it as much and you also don't need to check it for, say,
+% having a test function and having a documentation chunk earlier in a "function"
+% environment. In this case, write "\begin[aux]{macro}" and it will be
+% marked as such; its margin call-out will be printed in grey.
+%
+% Similarly, an internal package function still requires documentation
+% but usually will not be documented for users to see; these can be marked
+% as such with "\begin[internal]{macro}".
+%
+% For documenting \pkg{expl3}-type conditionals, you may also pass this
+% environment a "TF" option (and omit it from the function name) to denote that
+% the function is provided with "T", "F", and "TF" suffixes.
+% A similar "pTF" option will print both "TF" and "_p" predicate forms.
+%
+%
+% \DescribeMacro{\TestFiles}
+% \cs{TestFiles}\marg{list of files} is used to indicate which test files
+% are used for the current code; they will be printed in the documentation.
+%
+% \DescribeMacro{\UnitTested}
+% Within a "macro" environment, it is a good idea to mark whether a unit test has
+% been created for the commands it defines. This is indicated by writing \cs{UnitTested}
+% anywhere within "\begin{macro}" \dots "\end{macro}".
+%
+% If the class option "checktest" is enabled, then it is an \emph{error} to have a
+% "macro" environment without a \cs{testfile} file. This is intended for large packages such
+% as \pkg{expl3} that should have absolutely comprehensive tests suites and whose authors
+% may not always be as sharp at adding new tests with new code as they should be.
+%
+% \DescribeMacro{\TestMissing}
+% If a function is missing a test, this may be flagged by writing (as many times as needed)
+% \cs{TestMissing}\marg{explanation of test required}.
+% These missing tests will be summarised in the listing printed at the end of the
+% compilation run.
+%
+% \DescribeEnv{variable}
+% When documenting variable definitions, use the "variable" environment instead.
+% It will, here, behave identically to the "macro" environment, except that if the class
+% option "checktest" is enabled, variables will not be required to have a test file.
+%
+% \DescribeEnv{arguments}
+% Within a \env{macro} environment, you may use the \env{arguments} environment
+% to describe the arguments taken by the function(s). It behaves
+% like a modified enumerate environment.
+% \begin{verbatim}
+% % \begin{macro}{\foo:nn,\foo:VV}
+% % \begin{arguments}
+% %   \item Name of froozle to be frazzled
+% %   \item Name of muble to be jubled
+% % \end{arguments}
+% %   \begin{macrocode}
+% ... code for \foo:nn and \foo:VV ...
+% %   \end{macrocode}
+% % \end{macro}
+% \end{verbatim}
+%
+% \bigskip
+% \textbf{OPTIONS FOR THE FUTURE}\qquad Any improvements to the markup
+% for the \env{function} environment would be good to mirror in \env{macro}.
+%
+% Perhaps this would be a better syntax for describing arguments?
+% \begin{verbatim}
+% \begin{macro}{\foo:nn,foo:VV}
+% \dArg{Name of froozle to be frazzled}
+% \dArg{Name of mumble to be jumbled}
+% ...
+% \end{verbatim}
+% I.e., get rid of the environment and do things like in, say, \pkg{fontspec}.
+%
+% \subsection{Keeping things consistent}
+%
+% Whenever a function is either documented or defined with \env{function}
+% and \env{macro} respectively, its name is stored in a sequence for later
+% processing.
+%
+% At the end of the document (i.e., after the \textsc{dtx} file has finished
+% processing), the list of names is analysed to check whether all defined
+% functions have been documented and vice versa. The results are printed
+% in the console output.
+%
+% If you need to do more serious work with these lists of names, take a
+% look at the implementation for the data structures and methods used to
+% store and access them directly.
+%
+% \subsection{Documenting templates}
+%
+% The following macros are provided for documenting templates; might
+% end up being something completely different but who knows.
+% \begin{quote}\parskip=0pt\obeylines
+% "\begin{TemplateInterfaceDescription}" \Arg{template type name}
+% "  \TemplateArgument{none}{---}"
+% \textsc{or one or more of these:}
+% "  \TemplateArgument" \Arg{arg no} \Arg{meaning}
+% \textsc{and}
+% "\TemplateSemantics"
+% "  " \meta{text describing the template type semantics}
+% "\end{TemplateInterfaceDescription}"
+% \end{quote}
+%
+% \begin{quote}\parskip=0pt\obeylines
+% "\begin{TemplateDescription}" \Arg{template type name} \Arg{name}
+% \textsc{one or more of these:}
+% "  \TemplateKey" \marg{key name} \marg{type of key}
+% "    "\marg{textual description of meaning}
+% "    "\marg{default value if any}
+% \textsc{and}
+% "\TemplateSemantics"
+% "  " \meta{text describing special additional semantics of the template}
+% "\end{TemplateDescription}"
+% \end{quote}
+%
+% \begin{quote}\parskip=0pt\obeylines
+% "\begin{InstanceDescription}" \oarg{text to specify key column width (optional)}
+% \hfill\marg{template type name}\marg{instance name}\marg{template name}
+% \textsc{one or more of these:}
+% "  \InstanceKey" \marg{key name} \marg{value}
+% \textsc{and}
+% "\InstanceSemantics"
+% "  " \meta{text describing the result of this instance}
+% "\end{InstanceDescription}"
+% \end{quote}
+%
+% \end{documentation}
+%
+% \begin{implementation}
+%
+% \section{\pkg{l3doc} implementation}
+%
+%    \begin{macrocode}
+%<*class>
+%    \end{macrocode}
+%
+% The Guilty Parties.
+%    \begin{macrocode}
+\cs_new_nopar:Npn\Team{%
+  The~\LaTeX3~Project\thanks{%
+  Frank~Mittelbach,~Denys~Duchier,~Chris~Rowley,~
+  Rainer~Sch\"opf,~Johannes~Braams,~Michael~Downes,~
+  David~Carlisle,~Alan~Jeffrey,~Morten~H\o{}gholm,~Thomas~Lotze,~
+  Javier~Bezos,~Will~Robertson,~Joseph~Wright}}
+%    \end{macrocode}
+%
+% \subsection{Options and configuration}
+%
+%    \begin{macrocode}
+\DeclareOption{a5paper}{\@latexerr{Option not supported}{}}
+%    \end{macrocode}
+%
+%    \begin{macrocode}
+\bool_new:N \g_doc_full_bool
+\bool_new:N \g_doc_lmodern_bool
+\bool_new:N \g_doc_checkfunc_bool
+\bool_new:N \g_doc_checktest_bool
+%    \end{macrocode}
+%
+%    \begin{macrocode}
+\DeclareOption{full}{ \bool_set_true:N \g_doc_full_bool }
+\DeclareOption{onlydoc}{ \bool_set_false:N \g_doc_full_bool }
+%    \end{macrocode}
+%
+%    \begin{macrocode}
+\DeclareOption{check}{ \bool_set_true:N \g_doc_checkfunc_bool }
+\DeclareOption{nocheck}{ \bool_set_false:N \g_doc_checkfunc_bool }
+%    \end{macrocode}
+%
+%    \begin{macrocode}
+\DeclareOption{checktest}{ \bool_set_true:N \g_doc_checktest_bool }
+\DeclareOption{nochecktest}{ \bool_set_false:N \g_doc_checktest_bool }
+%    \end{macrocode}
+%
+%    \begin{macrocode}
+\DeclareOption{cm-default}{ \bool_set_false:N \g_doc_lmodern_bool }
+\DeclareOption{lm-default}{ \bool_set_true:N \g_doc_lmodern_bool }
+%    \end{macrocode}
+%
+%    \begin{macrocode}
+\DeclareOption*{\PassOptionsToClass{\CurrentOption}{article}}
+\ExecuteOptions{full,a4paper,nocheck,nochecktest,lm-default}
+%    \end{macrocode}
+%
+% Input a local configuration file, if it exists.
+%
+%    \begin{macrocode}
+\InputIfFileExists{l3doc.cfg}
+  {
+    \typeout{*************************************^^J
+             *~Local~config~file~l3doc.cfg~used   ^^J
+             *************************************}
+  }
+  { \@input{l3doc.ltx} }
+%    \end{macrocode}
+%
+%    \begin{macrocode}
+\ProcessOptions
+%    \end{macrocode}
+%
+%
+% \subsection{Class and package loading}
+%
+%    \begin{macrocode}
+\LoadClass{article}
+\RequirePackage{doc}
+\RequirePackage{array,alphalph,booktabs,color,fixltx2e,enumitem,textcomp,trace,underscore,csquotes}
+%    \end{macrocode}
+%
+%    \begin{macrocode}
+\bool_if:NT \g_doc_lmodern_bool {
+  \RequirePackage[T1]{fontenc}
+  \RequirePackage{lmodern}
+}
+%    \end{macrocode}
+%
+%    \begin{macrocode}
+\RequirePackage{hypdoc}
+%    \end{macrocode}
+% Just want the "comment" environment from \pkg{verbatim}:
+%    \begin{macrocode}
+\let\doc@verbatim\verbatim
+\let\enddoc@verbatim\endverbatim
+\let\doc@@verbatim\@verbatim
+\expandafter\let\csname doc@verbatim*\expandafter\endcsname
+  \csname verbatim*\endcsname
+\expandafter\let\csname enddoc@verbatim*\expandafter\endcsname
+  \csname endverbatim*\endcsname
+\expandafter\let\csname doc@@verbatim*\expandafter\endcsname
+  \csname @verbatim*\endcsname
+\RequirePackage{verbatim}
+\AtBeginDocument{%
+  \let\verbatim\doc@verbatim
+  \let\endverbatim\enddoc@verbatim
+  \let\@verbatim\doc@@verbatim
+  \expandafter\let\csname verbatim*\expandafter\endcsname
+    \csname doc@verbatim*\endcsname
+  \expandafter\let\csname endverbatim*\expandafter\endcsname
+    \csname enddoc@verbatim*\endcsname
+  \expandafter\let\csname @verbatim*\expandafter\endcsname
+    \csname doc@@verbatim*\endcsname
+}
+%    \end{macrocode}
+%
+% \subsection{Configuration}
+%
+% \begin{macro}{\MakePrivateLetters}
+%    \begin{macrocode}
+\cs_set_nopar:Npn \MakePrivateLetters {
+  \char_set_catcode_letter:N \@
+  \char_set_catcode_letter:N \_
+  \char_set_catcode_letter:N \:
+}
+%    \end{macrocode}
+% \end{macro}
+%
+%    \begin{macrocode}
+\setcounter{StandardModuleDepth}{1}
+\@addtoreset{CodelineNo}{part}
+\cs_set_nopar:Npn \theCodelineNo {
+  \textcolor[gray]{0.5}{ \sffamily\tiny\arabic{CodelineNo} }
+}
+%    \end{macrocode}
+%
+%
+%
+% \subsection{Design}
+%
+% Increase the text width slightly so that width the standard fonts
+% 72 columns of code may appear in a |macrocode| environment.
+% Increase the marginpar width slightly, for long command names.
+% And increase the left margin by a similar amount.
+%    \begin{macrocode}
+\setlength   \textwidth      { 385pt }
+\addtolength \marginparwidth {  30pt }
+\addtolength \oddsidemargin  {  20pt }
+\addtolength \evensidemargin {  20pt }
+%    \end{macrocode}
+% (These were introduced when "article" was the documentclass, but
+%  I've left them here for now to remind me to do something about them
+%  later; we still have the problem of \emph{very long} command names.)
+%
+% Customise lists:
+%    \begin{macrocode}
+\cs_set_eq:NN \@@oldlist\list
+\cs_set_nopar:Npn \list#1#2{\@@oldlist{#1}{#2\listparindent\z@}}
+\setlength \parindent  { 0pt }
+\setlength \itemindent { 0pt }
+\setlength \parskip    { \medskipamount }
+%    \end{macrocode}
+%
+% \subsection{Text markup}
+%
+%    Make "|" and |"| be `short verb' characters, but not in
+%    the document preamble, where an active character may interfere
+%    with packages that are loaded.
+%    \begin{macrocode}
+\AtBeginDocument {
+  \MakeShortVerb \"
+  \MakeShortVerb \|
+}
+%    \end{macrocode}
+%
+%    \begin{macrocode}
+\providecommand*\eTeX{
+  \if b\expandafter\@car\f@series\@nil\boldmath\fi
+  $\m@th\varepsilon$-\TeX
+}
+\providecommand*\IniTeX{Ini\TeX}
+\providecommand*\Lua{Lua}
+\providecommand*\LuaTeX{\Lua\TeX}
+\providecommand*\pdfTeX{pdf\TeX}
+\RequirePackage{graphicx}
+\cs_if_free:NT \XeTeX {
+  \cs_new_protected_nopar:Npn \XeTeX
+    {X\kern-.125em\lower.5ex\hbox{\reflectbox{E}}\kern-.1667em\TeX}
+}
+%    \end{macrocode}
+%
+% \begin{macro}{\cmd,\cs}
+% |\cmd{\foo}| Prints |\foo| verbatim. It may be used inside moving
+% arguments. |\cs{foo}| also prints |\foo|, for those who prefer that
+% syntax.
+%    \begin{macrocode}
+\cs_set_nopar:Npn \cmd #1 { \cs{\expandafter\cmd@to@cs\string#1} }
+\cs_set_nopar:Npn \cmd@to@cs #1#2 { \char\number`#2\relax }
+\DeclareRobustCommand \cs [1] { \texttt { \char`\\ #1 } }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\marg,\oarg,\parg}
+%    |\marg{text}| prints \marg{text}, `mandatory argument'.\\
+%    |\oarg{text}| prints \oarg{text}, `optional argument'.\\
+%    |\parg{te,xt}| prints \parg{te,xt}, `picture mode argument'.
+%    \begin{macrocode}
+\providecommand\marg[1]{ \texttt{\char`\{} \meta{#1} \texttt{\char`\}} }
+\providecommand\oarg[1]{ \texttt[ \meta{#1} \texttt] }
+\providecommand\parg[1]{ \texttt( \meta{#1} \texttt) }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\m,\file,\env,\pkg,\cls}
+% This list may change\dots this is just my preference for markup.
+%    \begin{macrocode}
+\cs_set_eq:NN \m    \meta
+\cs_set_eq:NN \file \nolinkurl
+\DeclareRobustCommand \env {\texttt}
+\DeclareRobustCommand \pkg {\textsf}
+\DeclareRobustCommand \cls {\textsf}
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{environment}{texnote}
+%    \begin{macrocode}
+\newenvironment{texnote}{
+  \endgraf
+  \vspace{3mm}
+  \small\textbf{\TeX~hackers~note:}
+}{
+  \vspace{3mm}
+}
+%    \end{macrocode}
+% \end{environment}
+%
+% \begin{macro}{\tn}
+% As \cs{cs}. Use this to mark up all \TeX\ and \LaTeXe\ commands; they 
+% then end up together in the index. TODO: hyperlinks in the index entries.
+%    \begin{macrocode}
+\newcommand\tn[1]{
+  \texttt{\bslash #1}
+  \index{TeX~and~LaTeX2e~commands\actualchar
+          \string\TeX{}~and~\string\LaTeXe{}~commands:\levelchar
+          #1\actualchar{\string\ttfamily\string\bslash{}#1}}}
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{environment}{documentation}
+% \begin{environment}{implementation}
+% \begin{macro}{\EnableDocumentation,\EnableImplementation}
+% \begin{macro}{\DisableDocumentation,\DisableImplementation}
+%    \begin{macrocode}
+\cs_new:Npn \doc_implementation: {
+  \cs_set:Npn \variable {\macro[var]}
+}
+\cs_new:Npn \doc_docu: {
+  \cs_set_eq:NN \variable \variabledoc
+}
+\AtEndOfPackage{\doc_docu:}
+\newenvironment{documentation}{\doc_docu:}{}
+\newenvironment{implementation}{\doc_implementation:}{}
+\newcommand\EnableDocumentation{%
+  \renewenvironment{documentation}{\doc_docu:}{}%
+}
+\newcommand\EnableImplementation{%
+  \renewenvironment{implementation}{\doc_implementation:}{}%
+}
+\newcommand\DisableDocumentation{%
+  \cs_set_eq:NN \documentation \comment
+  \cs_set_eq:NN \enddocumentation \endcomment
+}
+\newcommand\DisableImplementation{%
+  \cs_set_eq:NN \implementation \comment
+  \cs_set_eq:NN \endimplementation \endcomment
+}
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{environment}
+% \end{environment}
+%
+% \begin{environment}{arguments}
+% This environment is designed to be used within a \env{macro} environment
+% to describe the arguments of the macro/function.
+%    \begin{macrocode}
+\newenvironment{arguments}{
+  \enumerate[
+    nolistsep,
+    label=\texttt{\#\arabic*}~:,
+    labelsep=*,
+  ]
+}{
+  \endenumerate
+}
+%    \end{macrocode}
+% \end{environment}
+%
+% \begin{environment}{function}
+% \begin{environment}{variabledoc}
+% Environment for documenting function(s).
+% Stick the function names in a box. Use a "|" as delimiter and
+% allow |<...>| to be used as markup for |\meta{...}|.
+% Ignore spaces and sanitize with catcodes before reading argument.
+%    \begin{macrocode}
+\char_set_catcode_active:N \<
+\DeclareDocumentCommand \function { O{} } {
+
+  \phantomsection
+  
+  \bool_gset_false:N \l_doc_meta_TF_bool
+  \bool_gset_false:N \l_doc_meta_pTF_bool
+  \bool_gset_false:N \l_doc_meta_EXP_bool
+
+  \cs_set_nopar:Npn \KV_key_no_value_elt:n ##1 { \use:c {doc_meta_opt_##1:} }
+  \keyval_parse:NNn \KV_key_no_value_elt:n \use_none:nn {#1}
+
+  \char_set_catcode_active:N \<
+  \cs_set_eq:NN < \doc_open_meta:n
+  
+  \group_begin:
+    \MakePrivateLetters
+    \char_set_catcode_other:N \|
+    \char_set_catcode_other:N \\
+    \char_set_catcode_space:N \~
+    \char_set_catcode_ignore:N \ % space
+    \char_set_catcode_ignore:N \^^M
+    \char_set_catcode_ignore:N \^^I
+    \function_aux:n
+}
+\char_set_catcode_other:N \<
+\cs_set_eq:NN \variabledoc \function
+%    \end{macrocode}
+%
+%    \begin{macrocode}
+\cs_set:Nn \doc_meta_opt_TF: {
+  \bool_gset_true:N \l_doc_meta_TF_bool
+}
+\cs_set:Nn \doc_meta_opt_EXP: {
+  \bool_gset_true:N \l_doc_meta_EXP_bool
+}
+\cs_set:Nn \doc_meta_opt_pTF: {
+  \bool_gset_true:N \l_doc_meta_pTF_bool
+  \bool_gset_true:N \l_doc_meta_TF_bool
+  \bool_gset_true:N \l_doc_meta_EXP_bool
+}
+%    \end{macrocode}
+%
+% \begin{macro}{\function_aux:n}
+% \begin{arguments}
+% \item Vertical bar--separated list of functions with optional metadata;
+%       input has already been sanitised by catcode changes before reading
+%       the argument.
+% \end{arguments}
+% Because vertical bars are odd delimiters to choose, we also now iterate over commas!
+%    \begin{macrocode}
+\group_begin:
+\char_set_catcode_other:N \|
+\cs_gset_nopar:Npn \function_aux:n #1 {
+    \cs_set_nopar:Npn \nextnewline{\cs_gset_nopar:Npn\nextnewline{\\}}
+    \tl_gset_eq:NN \g_doc_macro_tl \c_empty_tl
+    \bigskip\endgraf\noindent\ttfamily
+    \tabular[b]{ | l @{} c | }
+      \hline
+      \clist_map_inline:nn {#1} {\doc_showmacro:w ##1 | \q_stop} \\
+      \hline
+    \endtabular
+  \group_end:
+}
+\group_end:
+%    \end{macrocode}
+% \end{macro}
+% \end{environment}
+% \end{environment}
+%
+% \begin{macro}{\doc_showmacro:w}
+% This function reads in a "|"-separated list, passing each item to
+% the auxiliary function "\doc_showmacro_aux:w".
+%    \begin{macrocode}
+\group_begin:
+\char_set_catcode_other:N \|
+\cs_gset_nopar:Npn \doc_showmacro:w #1 | {
+  \tl_if_blank:nTF {#1} {
+    \use_none:n
+  }{
+    \doc_showmacro_aux:w #1 / \q_stop
+    \peek_meaning:NTF \q_stop { \use_none:n } { \doc_showmacro:w }
+  }
+}
+\group_end:
+%    \end{macrocode}
+% \end{macro}
+%
+%    \begin{macrocode}
+\bool_new:N \l_doc_meta_TF_bool
+\bool_new:N \l_doc_meta_pTF_bool
+\bool_new:N \l_doc_meta_EXP_bool
+%    \end{macrocode}
+%
+% \begin{macro}{\doc_showmacro_aux:w}
+% This macro is passed one of:
+% \begin{quote}
+%   "\abc:cnx / (EXP) / \q_stop" \\
+%   "\abc:cnx / \q_stop" \\
+% \end{quote}
+% We also have some code here to print out every documented macro at the end
+% of the document.
+% \begin{arguments}
+% \item Function/macro/variable name \item Metadata tags (if any)
+% \end{arguments}
+%    \begin{macrocode}
+\cs_new_nopar:Npn \doc_showmacro_aux:w #1 / #2 \q_stop {
+
+  \tl_if_in:nnT {#2} { (TF)  } {
+    \bool_gset_true:N \l_doc_meta_TF_bool
+  }
+  
+  \tl_if_in:nnT {#2} { (EXP) } {
+    \bool_gset_true:N \l_doc_meta_EXP_bool
+  }
+  
+  \tl_if_in:nnT {#2} { (pTF)  } {
+    \bool_gset_true:N \l_doc_meta_TF_bool
+    \bool_gset_true:N \l_doc_meta_pTF_bool
+    \bool_gset_true:N \l_doc_meta_EXP_bool
+  }
+
+  \bool_if:NT \l_doc_meta_pTF_bool {
+    \tl_set:Nx \l_doc_pTF_name_tl { \doc_predicate_from_base:w #1 \q_nil }
+    \doc_special_main_index:o { \l_doc_pTF_name_tl }
+    \seq_gput_right:Nx \g_doc_functions_seq { \tl_to_str:N \l_doc_pTF_name_tl }
+  }
+
+  \bool_if:NTF \l_doc_meta_TF_bool {
+    \doc_special_main_index:o { #1 TF }
+    \seq_gput_right:Nx \g_doc_functions_seq { \tl_to_str:n { #1 TF } }
+    \seq_gput_right:Nx \g_doc_functions_seq { \tl_to_str:n { #1 T  } }
+    \seq_gput_right:Nx \g_doc_functions_seq { \tl_to_str:n { #1  F } }
+  }{
+    \doc_special_main_index:o { #1 }
+    \seq_gput_right:Nx \g_doc_functions_seq { \tl_to_str:n { #1    } }
+  }
+
+  \bool_if:NTF \l_doc_meta_pTF_bool {
+    \bool_gset_false:N \l_doc_meta_TF_bool
+    \exp_after:wN \doc_showmacro_aux_ii:w \l_doc_pTF_name_tl ::\q_stop
+    \bool_gset_true:N \l_doc_meta_TF_bool
+    \doc_showmacro_aux_ii:w #1::\q_stop
+  }{
+    \doc_showmacro_aux_ii:w #1::\q_stop
+  }
+}
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\doc_showmacro_aux_ii:w}
+% This macro is passed arguments in the following ways:
+% \begin{quote}
+%   "\doc_showmacro_aux_ii:w "^^A
+%    \makebox[\widthof{\texttt{123456}}]{\meta{name}}^^A
+%                        " ::\q_stop" \\
+%   "\doc_showmacro_aux_ii:w \foo   ::\q_stop" \\
+%   "\doc_showmacro_aux_ii:w \foo:  ::\q_stop" \\
+%   "\doc_showmacro_aux_ii:w \foo:Z ::\q_stop" \\
+% \end{quote}
+% Notice that for "\foo", "#2" and "#3" are empty,\\
+% for "\foo:", "#2" is empty, and "#3" is `":"'\\
+% for "\foo:Z", "#2" is `"Z"' and "#3" is `":"' .
+% \begin{arguments}
+% \item Function name \item Possible arg.\ spec. \item Possible colon
+% \end{arguments}
+%    \begin{macrocode}
+\cs_set_nopar:Npn \doc_showmacro_aux_ii:w #1:#2:#3 \q_stop {
+  \nextnewline
+  \str_if_eq:xxTF {#1} {\g_doc_macro_tl} {
+    \doc_typeset_aux:n
+  }{
+    \tl_gset:Nn \g_doc_macro_tl {#1}
+    \use:n
+  }
+  { \g_doc_macro_tl }
+  #3
+  #2
+  \bool_if:NT \l_doc_meta_TF_bool { \doc_typeset_TF: }
+  &
+  \bool_if:NT \l_doc_meta_EXP_bool {
+    \hspace{\tabcolsep}
+    $\star$
+  }
+  \tl_set:Nx \g_doc_macro_tl { \tl_to_str:N \g_doc_macro_tl }
+  \exp_args:NNf \tl_replace_all_in:Nnn \g_doc_macro_tl {\token_to_str:N _} {/}
+  \exp_args:NNf \tl_replace_all_in:Nnn \g_doc_macro_tl {\@backslashchar} {}
+  \bool_if:NT \g_doc_full_bool {
+    \exp_args:Nf\label{doc/function/\g_doc_macro_tl#3#2}
+  }
+}
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{environment}{syntax}
+% Syntax block placed next to the list of functions to illustrate their use.
+%    \begin{macrocode}
+\dim_new:N \g_doc_syntax_dim
+\dim_set:Nn \g_doc_syntax_dim {0.7\textwidth}
+\newenvironment{syntax}{
+  \minipage[b]{\g_doc_syntax_dim}
+    \cs_set_nopar:Npn \meta@font@select{\rmfamily\itshape} % (Will: I HATE italic cmtt!)
+    \small\ttfamily\raggedright
+    \obeyspaces\obeylines
+}{
+  \endminipage
+  \hfil\break
+  \global\@ignoretrue
+}
+%    \end{macrocode}
+% \end{environment}
+%
+% Perhaps these belong in \file{l3token}?
+%    \begin{macrocode}
+\tl_map_inline:nn {0123456789} { \cs_gset_eq:cN {char_other_#1} #1 }
+%    \end{macrocode}
+%
+% \begin{macro}{\doc_open_meta:n,\doc_close_meta:n}
+% This code turns all numbers within "<...>" markup to be set as subscripts.
+% You can use escaped numbers to get the real thing (e.g., "\1" = `1').
+%    \begin{macrocode}
+\group_begin:
+  \tl_map_inline:nn {0123456789} { \char_set_catcode_active:N #1 }
+  \cs_new:Npn \doc_open_meta:n {
+    \group_begin:
+      \tl_map_function:nN {0123456789} \doc_assign_num:n
+      \doc_close_meta:w
+  }
+  \cs_new:Npn \Arg {
+    \texttt{ \char`\{ }
+    \group_begin:
+      \tl_map_function:nN {0123456789} \doc_assign_num:n
+      \doc_close_Arg:n
+  }
+\group_end:
+\cs_new_nopar:Npn \doc_close_meta:w #1> { \meta {#1} \group_end: }
+\cs_new_nopar:Npn \doc_close_Arg:n #1 {
+  \meta {#1}
+  \group_end:
+  \texttt{ \char`\} }
+}
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\doc_assign_num:n}
+% This function takes a numeral (`0'), defines its escaped self to be equal
+% to itself ("\0" $\to$ `0'), makes it active, and turns itself into a subscript
+% instead (`0' $\to$ `${}_0$').
+%    \begin{macrocode}
+\cs_new_nopar:Npn \doc_assign_num:n #1 {
+  \cs_set_eq:cc { \string #1 } { char_other_\string #1 }
+  \char_set_catcode_active:N #1
+  \cs_set_nopar:Npn #1 { \unskip \, $ {} \sb { \use:c { char_other_\string #1 } } $ }
+}
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{environment}{macro}
+% We want to extend the old definition to allow comma-separated lists of
+% macros, rather than one at a time. keyval processing is very rudimentary;
+% awaiting a more robust solution.
+%    \begin{macrocode}
+\renewcommand \macro [1][] {
+
+  \int_compare:nNnTF \currentgrouplevel=2
+    { \int_gzero:N \g_doc_nested_macro_int }
+    { \int_incr:N  \g_doc_nested_macro_int }
+
+  \bool_set_false:N \l_doc_macro_aux_bool
+  \bool_set_false:N \l_doc_macro_internal_bool
+  \bool_set_false:N \l_doc_macro_TF_bool
+  \bool_set_false:N \l_doc_macro_pTF_bool
+  \bool_set_false:N \l_doc_macro_var_bool
+  \bool_set_false:N \l_doc_tested_bool
+  
+  \cs_set_eq:NN \doc_macroname_prefix:n \use:n
+  \cs_set_eq:NN \doc_macroname_suffix: \c_empty_tl
+  
+  \cs_set_nopar:Npn \KV_key_no_value_elt:n ##1 { \use:c {doc_macro_opt_##1:} }
+  \keyval_parse:NNn \KV_key_no_value_elt:n \use_none:nn {#1}
+  
+  \cs_set_eq:NN \testfile \doc_print_testfile:n
+
+  \group_begin:
+    \MakePrivateLetters
+    \char_set_catcode_letter:N \\
+    \char_set_catcode_ignore:N \ % space
+    \char_set_catcode_ignore:N \^^M
+    \char_set_catcode_ignore:N \^^I
+    \doc_macro_aux:n
+}
+%    \end{macrocode}
+% After changing the catcodes, parse the arguments:
+%    \begin{macrocode}
+\cs_new_nopar:Npn \doc_macro_aux:n #1 {
+  \group_end:
+  \cs_set:Npn \l_doc_macro_input_clist {#1}
+  \bool_if:NTF \l_doc_macro_pTF_bool
+  {
+    \clist_map_inline:nn {#1}
+      {
+        \tl_if_blank:nF {##1}
+          {
+            \exp_args:Nf \doc_macro_single
+              { \doc_predicate_from_base:w ##1 \q_nil }
+          }
+      }
+    \bool_set_true:N \l_doc_macro_TF_bool
+    \clist_map_inline:nn {#1}
+      { \tl_if_blank:nF {##1} { \doc_macro_single {##1} } }
+    \bool_set_false:N \l_doc_macro_TF_bool
+  }
+  {
+    \clist_map_inline:nn {#1} 
+      { \tl_if_blank:nF {##1} { \doc_macro_single {##1} } }
+  }
+}
+%    \end{macrocode}
+%
+%    \begin{macrocode}
+\bool_new:N \l_doc_macro_internal_bool
+\bool_new:N \l_doc_macro_aux_bool
+\bool_new:N \l_doc_macro_TF_bool
+\bool_new:N \l_doc_macro_pTF_bool
+\bool_new:N \l_doc_macro_var_bool
+\cs_set_nopar:Npn \doc_macro_opt_aux: { \bool_set_true:N \l_doc_macro_aux_bool }
+\cs_set_nopar:Npn \doc_macro_opt_internal: { \bool_set_true:N \l_doc_macro_internal_bool }
+\cs_set_nopar:Npn \doc_macro_opt_TF:  { \bool_set_true:N \l_doc_macro_TF_bool  }
+\cs_set_nopar:Npn \doc_macro_opt_pTF: { \bool_set_true:N \l_doc_macro_pTF_bool }
+\cs_set_nopar:Npn \doc_macro_opt_var: { \bool_set_true:N \l_doc_macro_var_bool }
+%    \end{macrocode}
+% \end{environment}
+%
+%    \begin{macrocode}
+\cs_set:Npn \doc_predicate_from_base:w #1:#2 \q_nil {#1_p:#2}
+%    \end{macrocode}
+%
+% \begin{environment}{doc_macro_single}
+% Let's start to mess around with "doc"'s "macro" environment. See \file{doc.dtx}
+% for a full explanation of the original environment. It's
+% rather \emph{enthusiastically} commented.
+% \begin{arguments}
+% \item Macro/function/whatever name; input has already been sanitised.
+% \end{arguments}
+%    \begin{macrocode}
+\int_new:N \l_doc_macro_int
+\cs_set_nopar:Npn \doc_macro_single #1 {
+  \int_incr:N \l_doc_macro_int
+  \tl_set:Nx \saved@macroname { \token_to_str:N #1 }
+  \topsep\MacroTopsep
+  \trivlist
+  \cs_set_nopar:Npn \makelabel ##1 { \llap{##1} }
+  \if@inlabel
+    \cs_set_eq:NN \@tempa \@empty
+    \count@ \macro@cnt
+    \loop \ifnum\count@>\z@
+      \cs_set_nopar:Npx \@tempa{\@tempa\hbox{\strut}}
+      \advance\count@\m@ne
+    \repeat
+    \cs_set_nopar:Npx \makelabel ##1 {
+      \llap{\vtop to\baselineskip {\@tempa\hbox{##1}\vss}}
+    }
+    \advance \macro@cnt \@ne
+  \else
+    \macro@cnt \@ne
+  \fi
+
+  \bool_if:NT \l_doc_macro_aux_bool {
+    \cs_set_eq:NN \doc_macroname_prefix:n \doc_typeset_aux:n
+  }
+  \bool_if:NT \l_doc_macro_TF_bool {
+    \cs_set_eq:NN \doc_macroname_suffix: \doc_typeset_TF:
+  }
+  
+  \bool_if:NF \l_doc_macro_aux_bool {
+    \tl_gset:Nx \l_doc_macro_tl { \tl_to_str:n {#1} }
+    \exp_args:NNf \tl_greplace_all_in:Nnn \l_doc_macro_tl {\token_to_str:N _} {/}
+    \exp_args:NNf \tl_greplace_all_in:Nnn \l_doc_macro_tl {\@backslashchar} {}
+  }
+
+  \use:x {
+    \exp_not:N \item [ \exp_not:N \doc_print_macroname:n {
+      \tl_to_str:n {#1}
+    }]
+  }
+  \global\advance \c@CodelineNo \@ne
+
+  \bool_if:NF \l_doc_macro_aux_bool {
+    \bool_if:NTF \l_doc_macro_TF_bool {
+      \seq_gput_right:Nx \g_doc_macros_seq { \tl_to_str:n { #1 TF } }
+      \seq_gput_right:Nx \g_doc_macros_seq { \tl_to_str:n { #1 T  } }
+      \seq_gput_right:Nx \g_doc_macros_seq { \tl_to_str:n { #1 F  } }
+    }{
+      \seq_gput_right:Nx \g_doc_macros_seq { \tl_to_str:n {#1} }
+    }
+  }
+  \bool_if:NTF \l_doc_macro_TF_bool {
+    \SpecialMainIndex{#1 TF}\nobreak
+    \DoNotIndex{#1 TF}
+  }{
+    \SpecialMainIndex{#1}\nobreak
+    \DoNotIndex{#1}
+  }
+  
+  \global\advance \c@CodelineNo \m@ne
+  \ignorespaces
+}
+%    \end{macrocode}
+%
+% \begin{macro}{\doc_print_macroname:n}
+%    \begin{macrocode}
+\tl_clear:N \l_doc_macro_tl
+\cs_set_nopar:Npn \doc_print_macroname:n #1 {
+  \strut
+  \int_compare:nTF { \tl_elt_count:n {#1} <= 28 }
+    { \MacroFont } { \MacroLongFont }
+
+  % INEFFICIENT: (!)
+  \exp_args:NNx \seq_if_in:NnTF \g_doc_functions_seq
+  { #1 \bool_if:NT \l_doc_macro_TF_bool { \tl_to_str:n {TF} } }
+  {
+    \hyperref [doc/function/\l_doc_macro_tl]
+  }
+  { \use:n }
+  {
+    \doc_macroname_prefix:n {#1} \doc_macroname_suffix: \ % space!
+  }
+}
+%    \end{macrocode}
+% \end{macro}
+% \end{environment}
+%
+% \begin{macro}{\MacroLongFont}
+%    \begin{macrocode}
+\providecommand \MacroLongFont {
+  \fontfamily{lmtt}\fontseries{lc}\small
+}
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\doc_typeset_TF:,\doc_typeset_aux:n}
+% Used by \cs{doc_macro_single} and \cs{doc_showmacro_aux_ii:w} to typeset
+% conditionals and auxiliary functions.
+%    \begin{macrocode}
+\cs_set_nopar:Npn \doc_typeset_TF: {
+  \itshape TF%
+  \makebox[0pt][r]{%
+    \color[gray]{0.5}
+    \underline { \phantom{\itshape TF} \kern-0.1em }
+  }
+}
+\cs_set_nopar:Npn \doc_typeset_aux:n #1 {
+  {\color[gray]{0.5} #1}
+}
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\doc_print_testfile:n}
+% Used to show that a macro has a test, somewhere.
+%    \begin{macrocode}
+\DeclareDocumentCommand \doc_print_testfile:n {m} {
+  \bool_set_true:N \l_doc_tested_bool
+  \tl_if_eq:nnF {#1} {*} {
+    \seq_if_in:NnF \g_doc_testfiles_seq {#1}
+    {
+      \par{\footnotesize(\textit{
+        The~ test~ suite~ for~ this~ command,~ and~ others~ in~ this~ file,~ is~ \textsf{#1}}.
+      )\par}
+      \seq_gput_right:Nn \g_doc_testfiles_seq {#1}
+    }
+  }
+}
+\seq_new:N \g_doc_testfiles_seq
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\TestFiles}
+%    \begin{macrocode}
+\DeclareDocumentCommand \TestFiles {m} {
+  \par
+  {\itshape
+    The~ following~ test~ files~ are~ used~ for~ this~ code:~ \textsf{#1}.
+  }
+  \par\ignorespaces
+}
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\UnitTested}
+%    \begin{macrocode}
+\DeclareDocumentCommand \UnitTested {} {
+  \testfile*
+}
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\TestMissing}
+%    \begin{macrocode}
+\cs_generate_variant:Nn \prop_gput:Nnn {NVx}
+\prop_new:N \g_doc_missing_tests_prop
+\DeclareDocumentCommand \TestMissing {m} {
+  \prop_if_in:NVTF \g_doc_missing_tests_prop \l_doc_macro_input_clist
+  {
+    \prop_get:NVN \g_doc_missing_tests_prop \l_doc_macro_input_clist \l_tmpa_tl
+    \prop_gput:NVx \g_doc_missing_tests_prop \l_doc_macro_input_clist
+    {
+      *~ \l_tmpa_tl
+      ^^J \exp_not:n {\space\space\space\space\space\space}
+      *~ #1
+    }
+  }
+  { \prop_gput:NVn \g_doc_missing_tests_prop \l_doc_macro_input_clist {#1} }
+}
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\endmacro}
+%    \begin{macrocode}
+\int_new:N \g_doc_nested_macro_int
+\cs_set:Nn \doc_texttt_comma:n {\,,~\texttt{#1}}
+\cs_set:Npn \endmacro {
+  \int_compare:nT {\g_doc_nested_macro_int<1}
+  {
+  \par\nobreak{\footnotesize(\emph{
+    End~ definition~ for~
+    \prg_case_int:nnn { \clist_length:N \l_doc_macro_input_clist }
+    {
+      {1} { \texttt{ \clist_use:N \l_doc_macro_input_clist }. }
+      {2}
+      {
+        \tl_set:Nx \l_clist_first_tl { \clist_item:Nn \l_doc_macro_input_clist {0} }
+        \tl_set:Nx \l_clist_last_tl { \clist_item:Nn \l_doc_macro_input_clist {1} }
+        \texttt{\l_clist_first_tl}\,~ and~ \texttt{\l_clist_last_tl}\,.
+      }
+      {3}
+      {
+        \tl_set:Nx \l_clist_first_tl { \clist_item:Nn \l_doc_macro_input_clist {0} }
+        \tl_set:Nx \l_clist_mid_tl   { \clist_item:Nn \l_doc_macro_input_clist {1} }
+        \tl_set:Nx \l_clist_last_tl  { \clist_item:Nn \l_doc_macro_input_clist {2} }
+        \texttt{\l_clist_first_tl}\,,~
+        \texttt{\l_clist_mid_tl}\,,~
+        and~ \texttt{\l_clist_last_tl}\,.
+      }
+    }
+    {
+      \tl_set:Nx \l_clist_first_tl { \clist_item:Nn \l_doc_macro_input_clist {0} }
+      \texttt{\l_clist_first_tl}\,~and~others.
+    }
+    \bool_if:nT {
+      !\l_doc_macro_aux_bool &&
+      !\l_doc_macro_internal_bool &&
+      \int_compare_p:n {\g_doc_nested_macro_int<1}
+    }
+    {
+        \int_compare:nNnTF \l_doc_macro_int=1 {~This~} {~These~}
+        \bool_if:NTF \l_doc_macro_var_bool{variable}{function}
+        \int_compare:nNnTF \l_doc_macro_int=1 {~is~}{s~are~}
+        documented~on~page~
+        \exp_args:Nx\pageref{doc/function/\l_doc_macro_tl}.
+    }
+  })\par}
+  }
+  \bool_if:nT
+  { \g_doc_checktest_bool &&
+    !( \l_doc_macro_aux_bool || \l_doc_macro_var_bool ) &&
+    !\l_doc_tested_bool
+  }
+  {
+    \seq_gput_right:Nx \g_doc_not_tested_seq
+    {
+      \l_doc_macro_input_clist
+      \bool_if:NT \l_doc_macro_pTF_bool {~(pTF)}
+      \bool_if:NT \l_doc_macro_TF_bool {~(TF)}
+    }
+  }
+}
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\DescribeOption}
+% For describing package options. Due to Joseph Wright.
+% Name/usage might change soon.
+%    \begin{macrocode}
+\newcommand*{\DescribeOption}{
+ \leavevmode
+ \@bsphack
+ \begingroup
+   \MakePrivateLetters
+   \Describe@Option
+}
+\newcommand*{\Describe@Option}[1]{
+ \endgroup
+ \marginpar{
+   \raggedleft
+   \PrintDescribeEnv{#1}
+ }
+ \SpecialOptionIndex{#1}
+ \@esphack
+ \ignorespaces
+}
+\newcommand*{\SpecialOptionIndex}[1]{
+ \@bsphack
+ \begingroup
+   \HD@target
+   \let\HDorg@encapchar\encapchar
+   \edef\encapchar usage{
+     \HDorg@encapchar hdclindex{\the\c@HD@hypercount}{usage}
+   }
+   \index{
+     #1\actualchar{\protect\ttfamily#1}~(option)
+     \encapchar usage
+   }
+   \index{
+     options:\levelchar#1\actualchar{\protect\ttfamily#1}
+     \encapchar usage
+   }
+ \endgroup
+ \@esphack
+}
+%    \end{macrocode}
+% \end{macro}
+%
+% Here are some definitions for additional markup that will help to
+% structure your documentation.
+%
+% \begin{environment}{danger}
+% \begin{environment}{ddanger}
+% \begin{syntax}
+% |\begin{[d]danger}|\\
+% dangerous code\\
+% |\end{[d]danger}|
+% \end{syntax}
+%
+% \begin{danger}
+%   Provides a danger bend, as known from the \TeX{}book.
+% \end{danger}
+% The actual character from the font |manfnt|:
+%    \begin{macrocode}
+\font\manual=manfnt
+\cs_set_nopar:Npn \dbend { {\manual\char127} }
+%    \end{macrocode}
+%
+% Defines the single danger bend. Use it whenever there is a feature in your
+% package that might be tricky to use.
+% FIXME: Has to be fixed when in combination with a macro-definition.
+%    \begin{macrocode}
+\newenvironment {danger} {
+  \begin{trivlist}\item[]\noindent
+  \begingroup\hangindent=2pc\hangafter=-2
+  \cs_set_nopar:Npn \par{\endgraf\endgroup}
+  \hbox to0pt{\hskip-\hangindent\dbend\hfill}\ignorespaces
+}{
+  \par\end{trivlist}
+}
+%    \end{macrocode}
+%
+% \begin{ddanger}
+%   Use the double danger bend if there is something which could cause serious
+%   problems when used in a wrong way. Better the normal user does not know
+%   about such things.
+% \end{ddanger}
+%    \begin{macrocode}
+\newenvironment {ddanger} {
+  \begin{trivlist}\item[]\noindent
+  \begingroup\hangindent=3.5pc\hangafter=-2
+  \cs_set_nopar:Npn \par{\endgraf\endgroup}
+  \hbox to0pt{\hskip-\hangindent\dbend\kern2pt\dbend\hfill}\ignorespaces
+}{
+  \par\end{trivlist}
+}
+%    \end{macrocode}
+% \end{environment}
+% \end{environment}
+%
+% \subsection{Documenting templates}
+%
+%    \begin{macrocode}
+\newenvironment{TemplateInterfaceDescription}[1]
+  {\subsection{The~object~type~`#1'}%
+   \begingroup
+   \@beginparpenalty\@M
+   \description
+   \def\TemplateArgument##1##2{\item[Arg:~##1]##2\par}%
+   \def\TemplateSemantics{\enddescription\endgroup
+       \subsubsection*{Semantics:}}%
+  }
+  {\par\bigskip}
+%    \end{macrocode}
+%
+%    \begin{macrocode}
+\newenvironment{TemplateDescription}[2]
+  {\subsection{The~template~`#2'~(object~type~#1)}%
+   \subsubsection*{Attributes:}%
+   \begingroup
+   \@beginparpenalty\@M
+   \description
+   \def\TemplateKey##1##2##3##4{\item[##1~(##2)]##3%
+     \ifx\TemplateKey##4\TemplateKey\else
+%         \hskip0ptplus3em\penalty-500\hskip 0pt plus 1filll Default:~##4%
+         \hfill\penalty500\hbox{}\hfill Default:~##4%
+         \nobreak\hskip-\parfillskip\hskip0pt\relax
+     \fi
+     \par}%
+   \def\TemplateSemantics{\enddescription\endgroup
+       \subsubsection*{Semantics~\&~Comments:}}%
+  }
+  {\par\bigskip}
+%    \end{macrocode}
+%
+%    \begin{macrocode}
+\newenvironment{InstanceDescription}[4][xxxxxxxxxxxxxxx]
+  {\subsubsection{The~instance~`#3'~(template~#2/#4)}%
+   \subsubsection*{Attribute~values:}%
+   \begingroup
+   \@beginparpenalty\@M
+   \def\InstanceKey##1##2{\>\textbf{##1}\>##2\\}%
+   \def\InstanceSemantics{\endtabbing\endgroup
+       \vskip-30pt\vskip0pt
+       \subsubsection*{Layout~description~\&~Comments:}}%
+   \tabbing
+   xxxx\=#1\=\kill
+  }
+  {\par\bigskip}
+%    \end{macrocode}
+%
+% \subsection{Inheriting doc}
+%
+% Code here is taken from \pkg{doc}, stripped of comments and translated
+% into \pkg{expl3} syntax. New features are added in various places.
+%
+% \begin{macro}{\StopEventually,\Finale,\AlsoImplementation,\OnlyDescription}
+%    \begin{macrocode}
+\bool_new:N \g_doc_implementation_bool
+\cs_set_nopar:Npn \AlsoImplementation {
+  \bool_set_true:N \g_doc_implementation_bool
+  \cs_set:Npn \StopEventually ##1 {
+    \@bsphack
+    \cs_gset_nopar:Npn \Finale { ##1 \check@checksum }
+    \init@checksum
+    \@esphack
+  }
+}
+\AlsoImplementation
+\cs_set_nopar:Npn \OnlyDescription {
+  \@bsphack
+  \bool_set_false:N \g_doc_implementation_bool
+  \cs_set:Npn \StopEventually ##1 { ##1 \endinput }
+  \@esphack
+}
+\cs_set_eq:NN \Finale \relax
+%    \end{macrocode}
+% \end{macro}
+%
+%    \begin{macrocode}
+\cs_set_nopar:Npn \partname{File}
+%    \end{macrocode}
+%
+% \begin{macro}{\DocInput}
+% From \pkg{doc}. Now accepts comma-list input (who has commas in filenames?).
+%    \begin{macrocode}
+\clist_new:N \g_docinput_clist
+\cs_set:Npn \DocInput #1 {
+  \clist_map_inline:nn {#1} {
+    \clist_put_right:Nn \g_docinput_clist {##1}
+    \MakePercentIgnore
+    \input{##1}
+    \MakePercentComment
+  }
+}
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\DocInputAgain}
+% Uses "\g_docinput_clist" to re-input whatever's already been "\DocInput"-ed until now.
+% May be used multiple times.
+%    \begin{macrocode}
+\cs_set:Npn \DocInputAgain {
+  \clist_map_inline:Nn \g_docinput_clist {
+    \MakePercentIgnore
+    \input{##1}
+    \MakePercentComment
+  }
+}
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\DocInclude}
+% More or less exactly the same as |\include|, but uses |\DocInput|
+% on a |dtx| file, not |\input| on a |tex| file.
+%    \begin{macrocode}
+\cs_set_nopar:Npn \partname{File}
+%    \end{macrocode}
+%
+%    \begin{macrocode}
+\newcommand*{\DocInclude}[1]{%
+  \relax\clearpage
+  \docincludeaux
+  \IfFileExists{#1.fdd}{
+    \cs_set_nopar:Npn \currentfile{#1.fdd}
+  }{
+    \cs_set_nopar:Npn \currentfile{#1.dtx}
+  }
+  \ifnum\@auxout=\@partaux
+    \@latexerr{\string\include\space cannot~be~nested}\@eha
+  \else
+    \@docinclude #1
+  \fi
+}
+%    \end{macrocode}
+%
+%    \begin{macrocode}
+\cs_set_nopar:Npn \@docinclude #1 {
+  \clearpage
+  \immediate\write\@mainaux{\string\@input{#1.aux}}
+  \@tempswatrue
+  \if@partsw
+    \@tempswafalse
+    \cs_set_nopar:Npx \@tempb{#1}
+    \@for\@tempa:=\@partlist\do{
+      \ifx\@tempa\@tempb\@tempswatrue\fi
+    }
+  \fi
+  \if@tempswa
+    \cs_set_eq:NN \@auxout\@partaux
+    \immediate\openout\@partaux #1.aux
+    \immediate\write\@partaux{\relax}
+    \cs_set_eq:NN \@ltxdoc@PrintIndex\PrintIndex
+    \cs_set_eq:NN \PrintIndex\relax
+    \cs_set_eq:NN \@ltxdoc@PrintChanges\PrintChanges
+    \cs_set_eq:NN \PrintChanges\relax
+    \cs_set_eq:NN \@ltxdoc@theglossary\theglossary
+    \cs_set_eq:NN \@ltxdoc@endtheglossary\endtheglossary
+    \part{\currentfile}
+    {
+      \cs_set_eq:NN \ttfamily\relax
+      \cs_gset_nopar:Npx \filekey{\filekey, \thepart={\ttfamily\currentfile}}
+    }
+    \DocInput{\currentfile}
+    \cs_set_eq:NN \PrintIndex\@ltxdoc@PrintIndex
+    \cs_set_eq:NN \PrintChanges\@ltxdoc@PrintChanges
+    \cs_set_eq:NN \theglossary\@ltxdoc@theglossary
+    \cs_set_eq:NN \endtheglossary\@ltxdoc@endtheglossary
+    \clearpage
+    \@writeckpt{#1}
+    \immediate\closeout\@partaux
+  \else
+    \@nameuse{cp@#1}
+  \fi
+  \cs_set_eq:NN \@auxout\@mainaux
+}
+%    \end{macrocode}
+%
+%    \begin{macrocode}
+\cs_gset_nopar:Npn \codeline@wrindex #1 {
+  \immediate\write\@indexfile {
+    \string\indexentry{#1}
+    {\filesep\number\c@CodelineNo}
+  }
+}
+%    \end{macrocode}
+% \end{macro}
+%
+%    \begin{macrocode}
+\cs_set_eq:NN \filesep \@empty
+%    \end{macrocode}
+%
+% \begin{macro}{\docincludeaux}
+%    \begin{macrocode}
+\cs_set_nopar:Npn \docincludeaux {
+  \cs_set_nopar:Npn \thepart {\alphalph{part}}
+  \cs_set_nopar:Npn \filesep {\thepart-}
+  \cs_set_eq:NN \filekey\@gobble
+  \g@addto@macro\index@prologue{
+    \cs_gset_nopar:Npn\@oddfoot{
+      \parbox{\textwidth}{
+        \strut\footnotesize
+        \raggedright{\bfseries File~Key:}~\filekey
+      }
+    }
+    \cs_set_eq:NN \@evenfoot\@oddfoot
+  }
+  \cs_gset_eq:NN \docincludeaux\relax
+  \cs_gset_nopar:Npn\@oddfoot{
+    \expandafter\ifx\csname ver@\currentfile\endcsname\relax
+      File~\thepart :~{\ttfamily\currentfile}~
+    \else
+      \GetFileInfo{\currentfile}
+      File~\thepart :~{\ttfamily\filename}~
+      Date:~\ExplFileDate\ % space
+      Version~\ExplFileVersion
+    \fi
+    \hfill\thepage
+  }
+  \cs_set_eq:NN \@evenfoot \@oddfoot
+}
+%    \end{macrocode}
+% \end{macro}
+%
+% \subsection{At end document}
+%
+% Print all defined and documented macros/functions.
+%
+%    \begin{macrocode}
+\seq_new:N \g_doc_functions_seq
+\seq_new:N \g_doc_macros_seq
+\seq_new:N \g_doc_not_tested_seq
+%    \end{macrocode}
+%
+%    \begin{macrocode}
+\iow_open:Nn \g_write_func_stream { \jobname.cmds }
+%    \end{macrocode}
+%
+%    \begin{macrocode}
+\cs_new_nopar:Npn \doc_show_functions_defined: {
+  \bool_if:nT { \g_doc_implementation_bool && \g_doc_checkfunc_bool } {
+    \typeout{ ======================================== ^^J }
+
+    \tl_clear:N \l_tmpa_tl
+    \seq_map_inline:Nn \g_doc_functions_seq {
+      \seq_if_in:NnT \g_doc_macros_seq {##1} {
+        \tl_put_right:Nn \l_tmpa_tl { ##1 ^^J }
+        \iow_now:Nn \g_write_func_stream { ##1 }
+      }
+    }
+    \iow_close:N \g_write_func_stream
+    \doc_functions_typeout:n {
+      Functions~both~documented~and~defined:^^J (In~order~of~being~documented)
+    }
+
+    \seq_map_inline:Nn \g_doc_functions_seq {
+      \seq_if_in:NnF \g_doc_macros_seq {##1} {
+        \tl_put_right:Nn \l_tmpa_tl { ##1 ^^J }
+      }
+    }
+    \doc_functions_typeout:n { Functions~documented~but~not~defined: }
+
+    \seq_map_inline:Nn \g_doc_macros_seq {
+      \seq_if_in:NnF \g_doc_functions_seq {##1} {
+        \tl_put_right:Nn \l_tmpa_tl { ##1 ^^J }
+      }
+    }
+    \doc_functions_typeout:n { Functions~defined~but~not~documented: }
+
+    \typeout{ ======================================== }
+  }
+}
+\AtEndDocument{ \doc_show_functions_defined: }
+%    \end{macrocode}
+%
+%    \begin{macrocode}
+\cs_set_nopar:Npn \doc_functions_typeout:n #1 {
+  \tl_if_empty:NF \l_tmpa_tl {
+    \typeout{
+      -------------------------------------- ^^J #1 ^^J
+      -------------------------------------- ^^J \l_tmpa_tl
+    }
+    \tl_clear:N \l_tmpa_tl
+  }
+}
+%    \end{macrocode}
+%
+%    \begin{macrocode}
+\cs_new:Npn \doc_show_not_tested: {
+  \bool_if:NT \g_doc_checktest_bool
+  {
+    \bool_if:nT { !(\seq_if_empty_p:N  \g_doc_not_tested_seq) ||
+                  !(\prop_if_empty_p:N \g_doc_missing_tests_prop) }
+    {
+      \tl_clear:N \l_tmpa_tl
+      \prop_if_empty:NF \g_doc_missing_tests_prop
+      {
+        \tl_put_right:Nn \l_tmpa_tl
+        {
+          ^^J^^JThe~ following~ macro(s)~ have~ incomplete~ tests:^^J
+        }
+        \prop_map_inline:Nn \g_doc_missing_tests_prop
+        {
+          \tl_put_right:Nn \l_tmpa_tl
+          {^^J\space\space\space\space ##1
+           ^^J\space\space\space\space\space\space ##2} 
+        }
+      }
+      \seq_if_empty:NF \g_doc_not_tested_seq
+      {
+        \tl_put_right:Nn \l_tmpa_tl
+        {
+          ^^J^^J
+          The~ following~ macro(s)~ do~ not~ have~ any~ tests:^^J
+        }
+        \seq_map_inline:Nn \g_doc_not_tested_seq
+        {
+          \clist_map_inline:nn {##1}
+          {
+            \tl_put_right:Nn \l_tmpa_tl {^^J\space\space\space\space ####1}
+          }
+        }
+        \int_set:Nn \l_tmpa_int {\etex_interactionmode:D}
+        \errorstopmode
+        \ClassError{l3doc}{\l_tmpa_tl}{}
+        \int_set:Nn \etex_interactionmode:D {\l_tmpa_int}
+      }
+    }
+  }
+}
+\AtEndDocument{ \doc_show_not_tested: }
+%    \end{macrocode}
+%
+% \subsection{Indexing}
+%
+% Fix index (for now):
+%    \begin{macrocode}
+\g@addto@macro\theindex{\MakePrivateLetters}
+\cs_set:Npn \verbatimchar {&}
+%    \end{macrocode}
+%
+%    \begin{macrocode}
+\setcounter{IndexColumns}{2}
+%    \end{macrocode}
+%
+% Set up the Index to use "\part"
+%    \begin{macrocode}
+\IndexPrologue{
+  \part*{Index}
+  \markboth{Index}{Index}
+  \addcontentsline{toc}{part}{Index}
+  The~italic~numbers~denote~the~pages~where~the~
+  corresponding~entry~is~described,~
+  numbers~underlined~point~to~the~definition,~
+  all~others~indicate~the~places~where~it~is~used.
+}
+%    \end{macrocode}
+%
+%
+% \begin{macro}{\doc_special_main_index:n,\doc_special_main_index:o,\hdpgindex}
+% Heiko's replacement to play nicely with |hypdoc|:
+%    \begin{macrocode}
+
+\cs_set_nopar:Npn \doc_special_main_index:n #1 {
+  \index{
+    \@gobble#1
+    \actualchar
+    \string\verb\quotechar*\verbatimchar#1\verbatimchar
+    \encapchar
+    hdpgindex{\thepage}{usage}
+  }
+}
+\cs_set_nopar:Npn \doc_special_main_index:o { \exp_args:No \doc_special_main_index:n }
+%    \end{macrocode}
+%    \begin{macrocode}
+\cs_set_nopar:Npn \hdpgindex #1#2#3 {
+  \csname\ifx\\#2\\relax\else#2\fi\endcsname{
+    \hyperlink{page.#1}{#3}
+  }
+}
+%    \end{macrocode}
+% \end{macro}
+%
+%    \begin{macrocode}
+\g@addto@macro \PrintIndex { \AtEndDocument{ \typeout{^^J
+  ========================================^^J
+  Generate~the~index~by~executing^^J
+  \c_space_tl \c_space_tl \c_space_tl \c_space_tl
+  makeindex~-s~l3doc.ist~-o~\jobname.ind~\jobname.idx^^J
+  ========================================^^J
+  }}
+}
+%    \end{macrocode}
+%
+% \subsection{Change history}
+%
+% Set the change history to use "\part".
+% Allow control names to be hyphenated in here...
+%    \begin{macrocode}
+\GlossaryPrologue{
+  \part*{Change~History}
+  {\GlossaryParms\ttfamily\hyphenchar\font=`\-}
+  \markboth{Change~History}{Change~History}
+  \addcontentsline{toc}{part}{Change~History}
+}
+%    \end{macrocode}
+%
+%    \begin{macrocode}
+\g@addto@macro \PrintChanges { \AtEndDocument{ \typeout{^^J
+  ========================================^^J
+  Generate~the~change~list~by~executing^^J
+  \c_space_tl \c_space_tl \c_space_tl \c_space_tl
+   makeindex~-s~gglo.ist~~-o~\jobname.gls~\jobname.glo^^J
+  ========================================^^J
+  }}
+}
+%    \end{macrocode}
+%
+%^^A The standard \changes command modified slightly to better cope with
+%^^A this multiple file document.
+%^^A\def\changes@#1#2#3{%
+%^^A  \let\protect\@unexpandable@protect
+%^^A  \edef\@tempa{\noexpand\glossary{#2\space\currentfile\space#1\levelchar
+%^^A                                 \ifx\saved@macroname\@empty
+%^^A                                   \space
+%^^A                                   \actualchar
+%^^A                                   \generalname
+%^^A                                 \else
+%^^A                                   \expandafter\@gobble
+%^^A                                   \saved@macroname
+%^^A                                   \actualchar
+%^^A                                   \string\verb\quotechar*%
+%^^A                                   \verbatimchar\saved@macroname
+%^^A                                   \verbatimchar
+%^^A                                 \fi
+%^^A                                 :\levelchar #3}}%
+%^^A  \@tempa\endgroup\@esphack}
+%
+% \subsection{cfg}
+%
+%    \begin{macrocode}
+\bool_if:NTF \g_doc_full_bool {
+  \RecordChanges
+  \CodelineIndex
+  \EnableCrossrefs
+  \AlsoImplementation
+}{
+  \CodelineNumbered
+  \DisableCrossrefs
+  \OnlyDescription
+}
+%    \end{macrocode}
+%
+%
+%    \begin{macrocode}
+%</class>
+%    \end{macrocode}
+%
+%
+% \subsection{Makeindex configuration}
+%
+% The makeindex style "l3doc.ist" is used in place of the usual
+% "gind.ist" to ensure that I is used in the sequence I J K
+% not I II II, which would be the default makeindex behaviour.
+%
+% Will: Do we need this?
+%
+%    \begin{macrocode}
+%<*docist>
+actual '='
+quote '!'
+level '#'
+preamble
+"\n \\begin{theindex} \n \\makeatletter\\scan@allowedfalse\n"
+postamble
+"\n\n \\end{theindex}\n"
+item_x1   "\\efill \n \\subitem "
+item_x2   "\\efill \n \\subsubitem "
+delim_0   "\\pfill "
+delim_1   "\\pfill "
+delim_2   "\\pfill "
+% The next lines will produce some warnings when
+% running Makeindex as they try to cover two different
+% versions of the program:
+lethead_prefix   "{\\bfseries\\hfil "
+lethead_suffix   "\\hfil}\\nopagebreak\n"
+lethead_flag       1
+heading_prefix   "{\\bfseries\\hfil "
+heading_suffix   "\\hfil}\\nopagebreak\n"
+headings_flag       1
+
+% and just for source3:
+% Remove R so I is treated in sequence I J K not I II III
+page_precedence "rnaA"
+%</docist>
+%    \end{macrocode}
+%
+% \section{Testing}
+%\ExplSyntaxOn 
+%\cs_set_eq:NN\g_saved_doc_functions_seq\g_doc_functions_seq
+%\cs_set_eq:NN\g_saved_doc_macros_seq\g_doc_macros_seq
+%\ExplSyntaxOff
+%
+% \begin{function}{\example_foo:N|\example_foo:c}
+% \begin{syntax}
+%   "\example_foo:N" <arg1>
+% \end{syntax}
+% <0123456789> <\0\1\2\3\4\5\6\7\8\9>  \Arg{arg1} ":("
+% \end{function}
+%
+% \begin{function}{ \foo | \foo: | \foo:x | \barrz: }
+% \begin{syntax}
+%   "\example_foo:N" <arg1>
+% \end{syntax}
+% <0123456789> <\0\1\2\3\4\5\6\7\8\9>
+% \end{function}
+%
+% \begin{function}{\foo:N / (TF) | \foo_if:c / (TF) (EXP)}
+% Test.
+% \end{function}
+%
+% \begin{function}[EXP]{\fffoo:N}
+% Test.
+% \end{function}
+% \begin{function}[TF]{\ffffoo:N}
+% Test.
+% \end{function}
+% \begin{function}[pTF]{\ffoo:N}
+% Test.
+% \end{function}
+%
+% \begin{function}{ \bar / (EXP) | \bar: / (EXP)  | \bar:x / (EXP)  | }
+% \begin{syntax}
+%   "\example_foo:N" <arg1>
+% \end{syntax}
+% <0123456789> <\0\1\2\3\4\5\6\7\8\9>
+% \end{function}
+%
+% \begin{macro}{ \foo , \foo: , \foo:x }
+% Testing.
+% \end{macro}
+%
+% \bigskip\bigskip
+%
+% \begin{macro}[aux]{ \foo_aux: }
+% Testing.
+% \end{macro}
+%
+% \bigskip\bigskip
+%
+% \begin{macro}[TF]{ \foo_if:c }
+% Testing.
+% \end{macro}
+%
+% \begin{macro}[internal]{ \foo_if:d }
+% Testing.
+% \end{macro}
+%
+% \bigskip\bigskip
+%
+% \begin{macro}{\aaaa_bbbb_cccc_dddd_eeee_ffff_gggg_hhhh}
+% Long macro names need to be printed in a shorter font.
+%    \begin{macrocode}
+%    \end{macrocode}
+% \end{macro}
+%
+%
+% \begin{macro}{\c_minus_one,
+% \c_zero,
+% \c_one,
+% \c_two,
+% \c_three,
+% \c_four,
+% \c_five,
+% \c_six,
+% \c_seven,
+% \c_eight,
+% \c_nine,
+% \c_ten,
+% \c_eleven,
+% \c_sixteen,
+% \c_thirty_two,
+% \c_hundred_one,
+% \c_twohundred_fifty_five,
+% \c_twohundred_fifty_six,
+% \c_thousand,
+% \c_ten_thousand,
+% \c_ten_thousand_one}
+% \begin{arguments}
+% \item name
+% \item parameters
+% \end{arguments}
+% Another test.
+% \end{macro}
+%
+%  
+%\ExplSyntaxOn 
+%\cs_set_eq:NN\g_doc_functions_seq\g_saved_doc_functions_seq
+%\cs_set_eq:NN\g_doc_macros_seq\g_saved_doc_macros_seq
+%\ExplSyntaxOff
+%
+%
+% \subsection{Macros}
+% \raggedright
+% \ExplSyntaxOn
+% \seq_map_inline:Nn \g_doc_macros_seq { `\texttt{#1}' \quad }
+% \ExplSyntaxOff
+%
+% \subsection{Functions}
+% \ExplSyntaxOn
+% \seq_map_inline:Nn \g_doc_functions_seq { `\texttt{#1}' \quad }
+% \ExplSyntaxOff
+%
+% \end{implementation}
+%
+% \PrintIndex
+%
+% \endinput
diff --git a/Master/texmf-dist/source/latex/l3kernel/l3expan.dtx b/Master/texmf-dist/source/latex/l3kernel/l3expan.dtx
new file mode 100644
index 00000000000..5b37fd49aaf
--- /dev/null
+++ b/Master/texmf-dist/source/latex/l3kernel/l3expan.dtx
@@ -0,0 +1,1416 @@
+% \iffalse meta-comment
+%
+%% File: l3expan.dtx Copyright (C) 1990-2011 The LaTeX3 project
+%%
+%% It may be distributed and/or modified under the conditions of the
+%% LaTeX Project Public License (LPPL), either version 1.3c of this
+%% license or (at your option) any later version.  The latest version
+%% of this license is in the file
+%%
+%%    http://www.latex-project.org/lppl.txt
+%%
+%% This file is part of the "expl3 bundle" (The Work in LPPL)
+%% and all files in that bundle must be distributed together.
+%%
+%% The released version of this bundle is available from CTAN.
+%%
+%% -----------------------------------------------------------------------
+%%
+%% The development version of the bundle can be found at
+%%
+%%    http://www.latex-project.org/svnroot/experimental/trunk/
+%%
+%% for those people who are interested.
+%%
+%%%%%%%%%%%
+%% NOTE: %%
+%%%%%%%%%%%
+%%
+%%   Snapshots taken from the repository represent work in progress and may
+%%   not work or may contain conflicting material!  We therefore ask
+%%   people _not_ to put them into distributions, archives, etc. without
+%%   prior consultation with the LaTeX3 Project.
+%%
+%% -----------------------------------------------------------------------
+%
+%<*driver|package>
+\RequirePackage{l3names}
+\GetIdInfo$Id: l3expan.dtx 2480 2011-06-20 17:25:23Z bruno $
+  {L3 Experimental argument expansion}
+%</driver|package>
+%<*driver>
+\documentclass[full]{l3doc}
+\begin{document}
+  \DocInput{\jobname.dtx}
+\end{document}
+%</driver>
+% \fi
+%
+% \title{^^A
+%   The \pkg{l3expan} package\\ Argument expansion^^A
+%   \thanks{This file describes v\ExplFileVersion,
+%      last revised \ExplFileDate.}^^A
+% }
+%
+% \author{^^A
+%  The \LaTeX3 Project\thanks
+%    {^^A
+%      E-mail:
+%        \href{mailto:latex-team@latex-project.org}
+%          {latex-team@latex-project.org}^^A
+%    }^^A
+% }
+%
+% \date{Released \ExplFileDate}
+%
+% \maketitle
+%
+% \begin{documentation}
+%
+% This module provides generic methods for expanding \TeX{} arguments in a
+% systematic manner. The functions in this module all have prefix |exp|.
+%
+% Not all possible variations are implemented for every base
+% function. Instead only those that are used within the \LaTeX3 kernel
+% or otherwise seem to be of general interest are implemented.
+% Consult the module description to find out which functions are
+% actually defined. The next section explains how to define missing
+% variants.
+%
+% \section{Defining new variants}
+%
+% The definition of variant forms for base functions may be necessary
+% when writing new functions or when applying a kernel function in a
+% situation that we haven't thought of before.
+%
+% Internally preprocessing of arguments is done with functions from the
+% |\exp_| module.  They all look alike, an example would be
+% |\exp_args:NNo|. This function has three arguments, the first and the
+% second are a single tokens  the third argument gets
+% expanded once. If |\seq_gpush:No| was not defined the example
+% above could be coded in the following way:
+% \begin{verbatim}
+%   \exp_args:NNo \seq_gpush:Nn
+%      \g_file_name_stack
+%      \l_tmpa_tl
+% \end{verbatim}
+% In other words, the first argument to |\exp_args:NNo| is the base
+% function and the other arguments are preprocessed and then passed to
+% this base function. In the example the first argument to the base
+% function should be a single token which is left unchanged while the
+% second argument is expanded once. From this example we can also see
+% how the variants are defined. They just expand into the appropriate
+% |\exp_| function followed by the desired base function, \emph{e.g.}
+% \begin{quote}
+%   |\cs_new_nopar:Npn\seq_gpush:No{\exp_args:NNo\seq_gpush:Nn}|
+% \end{quote}
+% Providing variants in this way in style files is uncritical as the
+% |\cs_new_nopar:Npn| function will silently accept definitions whenever the
+% new definition is identical to an already given one. Therefore adding
+% such definition to later releases of the kernel will not make such
+% style files obsolete.
+%
+% The steps above may be automated by using the function
+% |\cs_generate_variant:Nn|, described next.
+%
+% \section{Methods for defining variants}
+%
+% \begin{function}{\cs_generate_variant:Nn}
+%   \begin{syntax}
+%     \cs{cs_generate_variant:Nn} \meta{parent control sequence}
+%     ~~\Arg{variant argument specifiers}
+%   \end{syntax}
+%   This function is used to define argument-specifier variants of the
+%   \meta{parent control sequence} for \LaTeX3 code-level macros. The
+%   \meta{parent control sequence} is first separated into the
+%   \meta{base name} and \meta{original argument specifier}. The
+%   comma-separated list of \meta{variant argument specifiers} is
+%   then used to define variants of the
+%   \meta{original argument specifier} where these are not already
+%   defined. For each \meta{variant} given, a function is created
+%   which will expand its arguments as detailed and pass them
+%   to the \meta{parent control sequence}. So for example
+%   \begin{verbatim}
+%     \cs_set:Npn \foo:Nn #1#2 { code here }
+%     \cs_generate_variant:Nn \foo:Nn { c }
+%   \end{verbatim}
+%   will create a new function \cs{foo:cn} which will expand its first
+%   argument into a control sequence name and pass the result to
+%   \cs{foo:Nn}. Similarly
+%   \begin{verbatim}
+%     \cs_generate_variant:Nn \foo:Nn { NV , cV }
+%   \end{verbatim}
+%   would generate the functions \cs{foo:NV} and \cs{foo:cV} in the
+%   same way. The \cs{cs_generate_variant:Nn} function can only be
+%   applied if the \meta{parent control sequence} is already defined. If
+%   the \meta{parent control sequence} is protected then the new sequence
+%   will also be protected. The \meta{variant} is created globally, as
+%   is any \cs{exp_args:N\meta{variant}} function needed to carry out
+%  the expansion.
+% \end{function}
+%
+% \section{Introducing the variants}
+%
+% The available internal functions for argument expansion come in two
+% flavours, some of them are faster then others. Therefore it is usually
+% best to follow the following guidelines when defining new functions
+% that are supposed to come with variant forms:
+% \begin{itemize}
+%   \item
+%     Arguments that might need expansion should come first in the list of
+%     arguments to make processing faster.
+%   \item
+%     Arguments that should consist of single tokens should come first.
+%   \item
+%     Arguments that need full expansion (\emph{i.e.}, are denoted
+%     with |x|) should be avoided if possible as they can not be
+%     processed expandably, \emph{i.e.}, functions of this type will
+%     not work correctly in arguments that are itself subject to |x|
+%     expansion.
+%   \item
+%     In general, unless in the last position, multi-token arguments
+%     |n|, |f|, and |o| will need special processing which is not fast.
+%     Therefore it is best to use the optimized functions, namely
+%     those that contain only |N|, |c|, |V|, and |v|, and, in the last
+%     position, |o|, |f|, with possible trailing |N| or |n|, which are
+%     not expanded.
+% \end{itemize}
+%
+% The |V| type returns the value of a register, which can be one of
+% |tl|, |num|, |int|, |skip|, |dim|, |toks|, or built-in \TeX{}
+% registers. The |v| type is the same except it first creates a
+% control sequence out of its argument before returning the
+% value. This recent addition to the argument specifiers may shake
+% things up a bit as most places where |o| is used will be replaced by
+% |V|. The documentation you are currently reading will therefore
+% require a fair bit of re-writing.
+%
+% In general, the programmer should not need to be concerned with
+% expansion control. When simply using the content of a variable,
+% functions with a |V| specifier should be used. For those referred to by
+% (cs)name, the |v| specifier is available for the same purpose. Only when
+% specific expansion steps are needed, such as when using delimited
+% arguments, should the lower-level functions with |o| specifiers be employed.
+%
+% The |f| type is so special that it deserves an example.
+% Let's pretend we want to set |\aaa| equal to the control sequence
+% stemming from turning |b \l_tmpa_tl b| into a control
+% sequence. Furthermore we want to store the execution of it in a
+% \meta{tl~var}. In this example we assume |\l_tmpa_tl| contains
+% the text string |lur|. The straightforward approach is
+% \begin{quote}
+%   |\tl_set:No \l_tmpb_tl  {\cs_set_eq:Nc \aaa { b \l_tmpa_tl b } }|
+% \end{quote}
+% Unfortunately this only puts
+% |\exp_args:NNc \cs_set_eq:NN \aaa {b \l_tmpa_tl b}| into |\l_tmpb_tl|
+% and not |\cs_set_eq:NwN \aaa = \blurb| as we probably wanted. Using
+% |\tl_set:Nx| is not an option as that will die horribly. Instead
+% we can do a
+% \begin{quote}
+%   |\tl_set:Nf \l_tmpb_tl {\cs_set_eq:Nc \aaa { b \l_tmpa_tl b } }|
+% \end{quote}
+% which puts the desired result in |\l_tmpb_tl|. It requires
+% |\toks_set:Nf| to be defined as
+% \begin{quote}
+%   |\cs_set_nopar:Npn \tl_set:Nf { \exp_args:NNf \tl_set:Nn }|
+% \end{quote}
+% If you use this type of expansion in conditional processing then
+% you should stick to using |TF|  type functions only as it does not
+% try to finish any |\if... \fi:| itself!
+%
+% \section{Manipulating the first argument}
+%
+% These functions are described in detail: expansion of multiple tokens follows
+% the same rules but is described in a shorter fashion.
+%
+% \begin{function}[EXP]{\exp_args:No}
+%   \begin{syntax}
+%     \cs{exp_args:No} \meta{function} \Arg{tokens} \Arg{tokens2} ...
+%   \end{syntax}
+%   This function absorbs two arguments (the \meta{function} name and
+%   the \meta{tokens}). The \meta{tokens} are expanded once, and the result
+%   is inserted in braces into the input stream \emph{after} reinsertion
+%   of the \meta{function}. Thus the \meta{function} may take more than
+%   one argument: all others will be left unchanged.
+% \end{function}
+%
+% \begin{function}[EXP]{\exp_args:Nc, \exp_args:cc}
+%   \begin{syntax}
+%     \cs{exp_args:Nc} \meta{function} \Arg{tokens} \Arg{tokens2} ...
+%   \end{syntax}
+%   This function absorbs two arguments (the \meta{function} name and
+%   the \meta{tokens}). The \meta{tokens} are expanded until only characters
+%   remain, and are then turned into a control sequence. (An internal error
+%   will occur if such a conversion is not possible). The result
+%   is inserted into the input stream \emph{after} reinsertion
+%   of the \meta{function}. Thus the \meta{function} may take more than
+%   one argument: all others will be left unchanged.
+%
+%   The |:cc| variant constructs the \meta{function} name in the same
+%   manner as described for the \meta{tokens}.
+% \end{function}
+%
+% \begin{function}[EXP]{\exp_args:NV}
+%   \begin{syntax}
+%     \cs{exp_args:NV} \meta{function} \meta{variable} \Arg{tokens2} ...
+%   \end{syntax}
+%   This function absorbs two arguments (the names of the \meta{function} and
+%   the  the \meta{variable}). The content of the \meta{variable} are recovered
+%   and placed inside braces into the input stream \emph{after} reinsertion
+%   of the \meta{function}. Thus the \meta{function} may take more than
+%   one argument: all others will be left unchanged.
+% \end{function}
+%
+% \begin{function}[EXP]{\exp_args:Nv}
+%   \begin{syntax}
+%     \cs{exp_args:Nv} \meta{function} \Arg{tokens} \Arg{tokens2} ...
+%   \end{syntax}
+%   This function absorbs two arguments (the \meta{function} name and
+%   the \meta{tokens}). The \meta{tokens} are expanded until only characters
+%   remain, and are then turned into a control sequence. (An internal error
+%   will occur if such a conversion is not possible). This control sequence
+%   should
+%   be the name of a \meta{variable}.   The content of the \meta{variable} are
+%   recovered and placed inside braces into the input stream \emph{after}
+%   reinsertion of the \meta{function}. Thus the \meta{function} may take more
+%   than one argument: all others will be left unchanged.
+% \end{function}
+%
+% \begin{function}[EXP]{\exp_args:Nf}
+%   \begin{syntax}
+%     \cs{exp_args:Nf} \meta{function} \Arg{tokens} \Arg{tokens2} ...
+%   \end{syntax}
+%   This function absorbs two arguments (the \meta{function} name and
+%   the \meta{tokens}). The \meta{tokens} are fully expanded until the
+%   first non-expandable token or space is found, and the result
+%   is inserted in braces into the input stream \emph{after} reinsertion
+%   of the \meta{function}. Thus the \meta{function} may take more than
+%   one argument: all others will be left unchanged.
+% \end{function}
+%
+% \begin{function}{\exp_args:Nx}
+%   \begin{syntax}
+%     \cs{exp_args:Nx} \meta{function} \Arg{tokens} \Arg{tokens2} ...
+%   \end{syntax}
+%   This function absorbs two arguments (the \meta{function} name and
+%   the \meta{tokens}) and exhaustively expands the \meta{tokens}
+%   second. The result is inserted in braces into the input stream
+%   \emph{after} reinsertion of the \meta{function}.
+%   Thus the \meta{function} may take more
+%   than one argument: all others will be left unchanged.
+% \end{function}
+%
+% \section{Manipulating two arguments}
+%
+% \begin{function}[EXP]
+%   {
+%     \exp_args:NNo,
+%     \exp_args:NNc,
+%     \exp_args:NNv,
+%     \exp_args:NNV,
+%     \exp_args:NNf,
+%     \exp_args:Nco,
+%     \exp_args:Ncf,
+%     \exp_args:Ncc,
+%     \exp_args:NVV
+%   }
+%   \begin{syntax}
+%     \cs{exp_args:NNc} \meta{token1} \meta{token2} \Arg{tokens}
+%   \end{syntax}
+%   These optimized functions absorb three arguments and expand the second and
+%   third as detailed by their argument specifier. The first argument
+%   of the function is then the next item on the input stream, followed
+%   by the expansion of the second and third arguments.
+% \end{function}
+%
+% \begin{function}[EXP]
+%   {
+%     \exp_args:Nno,
+%     \exp_args:NnV,
+%     \exp_args:Nnf,
+%     \exp_args:Noo,
+%     \exp_args:Noc,
+%     \exp_args:Nff,
+%     \exp_args:Nfo,
+%     \exp_args:Nnc
+%   }
+%   \begin{syntax}
+%     \cs{exp_args:Noo} \meta{token} \Arg{tokens1} \Arg{tokens2}
+%   \end{syntax}
+%   These functions absorb three arguments and expand the second and
+%   third as detailed by their argument specifier. The first argument
+%   of the function is then the next item on the input stream, followed
+%   by the expansion of the second and third arguments.
+%   These functions need special (slower) processing.
+% \end{function}
+% 
+% \begin{function}
+%   {
+%     \exp_args:NNx,
+%     \exp_args:Nnx,
+%     \exp_args:Ncx,
+%     \exp_args:Nox,
+%     \exp_args:Nxo,
+%     \exp_args:Nxx
+%   }
+%   \begin{syntax}
+%     \cs{exp_args:NNx} \meta{token1} \meta{token2} \Arg{tokens}
+%   \end{syntax}
+%   These functions absorb three arguments and expand the second and
+%   third as detailed by their argument specifier. The first argument
+%   of the function is then the next item on the input stream, followed
+%   by the expansion of the second and third arguments. These functions
+%   are not expandable.
+% \end{function}
+%
+% \section{Manipulating three arguments}
+%
+% \begin{function}[EXP]
+%   {
+%     \exp_args:NNNo,
+%     \exp_args:NNNV,
+%     \exp_args:Nccc,
+%     \exp_args:NcNc,
+%     \exp_args:NcNo,
+%     \exp_args:Ncco
+%   }
+%   \begin{syntax}
+%     \cs{exp_args:NNNo} \meta{token1} \meta{token2} \meta{token3}
+%     ~~\Arg{tokens}
+%   \end{syntax}
+%   These optimized functions absorb four arguments and expand the second, third
+%   and fourth as detailed by their argument specifier. The first
+%   argument of the function is then the next item on the input stream,
+%   followed by the expansion of the second argument, \emph{etc}.
+% \end{function}
+%
+% \begin{function}[EXP]
+%   {
+%     \exp_args:NNoo,
+%     \exp_args:NNno,
+%     \exp_args:Nnno,
+%     \exp_args:Nnnc,
+%     \exp_args:Nooo,
+%   }
+%   \begin{syntax}
+%     \cs{exp_args:NNNo} \meta{token1} \meta{token2} \meta{token3}
+%     ~~\Arg{tokens}
+%   \end{syntax}
+%   These functions absorb four arguments and expand the second, third
+%   and fourth as detailed by their argument specifier. The first
+%   argument of the function is then the next item on the input stream,
+%   followed by the expansion of the second argument, \emph{etc}.
+%   These functions need special (slower) processing.
+% \end{function}
+%
+% \begin{function}
+%   {
+%     \exp_args:NNnx,
+%     \exp_args:NNox,
+%     \exp_args:Nnnx,
+%     \exp_args:Nnox,
+%     \exp_args:Noox,
+%     \exp_args:Ncnx,
+%     \exp_args:Nccx
+%   }
+%   \begin{syntax}
+%     \cs{exp_args:NNnx} \meta{token1} \meta{token2} \meta{tokens1}
+%     ~~\Arg{tokens2}
+%   \end{syntax}
+%   These functions absorb four arguments and expand the second, third
+%   and fourth as detailed by their argument specifier. The first
+%   argument of the function is then the next item on the input stream,
+%   followed by the expansion of the second argument, \emph{etc.}
+% \end{function}
+%
+% \section{Unbraced expansion}
+%
+% \begin{function}
+%   {
+%     \exp_last_unbraced:Nf,
+%     \exp_last_unbraced:NV,
+%     \exp_last_unbraced:No,
+%     \exp_last_unbraced:Nv,
+%     \exp_last_unbraced:NcV,
+%     \exp_last_unbraced:NNV,
+%     \exp_last_unbraced:NNo,
+%     \exp_last_unbraced:Nfo,
+%     \exp_last_unbraced:NNNV,
+%     \exp_last_unbraced:NNNo
+%     }
+%   \begin{syntax}
+%     \cs{exp_last_unbraced:Nno} \meta{token} \meta{tokens1} \meta{tokens2}
+%    \end{syntax}
+%   These functions absorb the number of arguments given by their
+%   specification, carry out the expansion
+%   indicated and leave the the results in the input stream, with the
+%   last argument not surrounded by the usual braces.
+%   Of these, \cs{exp_last_unbraced:Nfo} needs special (slower) processing.
+% \end{function}
+% 
+% \begin{function}[EXP]{\exp_last_two_unbraced:Noo}
+%   \begin{syntax}
+%     \cs{exp_last_two_unbraced:Noo} \meta{token} \meta{tokens1} \Arg{tokens2}
+%   \end{syntax}
+%   This function absorbs three arguments and expand the second and third
+%   once. The first argument of the function is then the next item on the
+%   input stream, followed by the expansion of the second and third arguments,
+%   which are not wrapped in braces.
+%   This function needs special (slower) processing.
+% \end{function}
+%
+% \begin{function}[EXP]{\exp_after:wN}
+%   \begin{syntax}
+%     \cs{exp_after:wN} \meta{token1} \meta{token2}
+%   \end{syntax}
+%   Carries out a single expansion of \meta{token2} prior to expansion
+%   of \meta{token1}. If \meta{token2} is a \TeX{} primitive, it will
+%   be executed rather than expanded, while if \meta{token2} has not
+%   expansion (for example, if it is a character) then it will be left
+%   unchanged. It is important to notice that \meta{token1} may be
+%   \emph{any} single token, including group-opening and -closing
+%   tokens (|{| or |}"| assuming normal \TeX{} category codes). Unless
+%    specifically required, expansion should be carried out using an
+%   appropriate argument specifier variant or the appropriate
+%   \cs{exp_arg:N} function.
+%   \begin{texnote}
+%     This is the \TeX{} primitive \cs{expandafter} renamed.
+%   \end{texnote}
+% \end{function}
+%
+% \section{Preventing expansion}
+%
+% \begin{function}{\exp_not:N}
+%   \begin{syntax}
+%     \cs{exp_not:N} \meta{token}
+%   \end{syntax}
+%   Prevents expansion of the \meta{token} in a context where it would otherwise
+%   be expanded, for example an |x|-type argument.
+%   \begin{texnote}
+%     This is the \TeX{} \cs{noexpand} primitive.
+%   \end{texnote}
+% \end{function}
+%
+% \begin{function}{\exp_not:c}
+%   \begin{syntax}
+%     \cs{exp_not:c} \Arg{tokens}
+%   \end{syntax}
+%   Expands the \meta{tokens} until only unexpandable content remains, and then
+%   converts this into a control sequence. Further expansion of this control
+%   sequence is then inhibited.
+% \end{function}
+%
+% \begin{function}{\exp_not:n}
+%   \begin{syntax}
+%     \cs{exp_not:n} \Arg{tokens}
+%   \end{syntax}
+%   Prevents expansion of the \meta{tokens} in a context where they would
+%   otherwise
+%   be expanded, for example an |x|-type argument.
+%   \begin{texnote}
+%     This is the \eTeX{} \cs{unexpanded} primitive.
+%   \end{texnote}
+% \end{function}
+%
+% \begin{function}{\exp_not:V}
+%   \begin{syntax}
+%     \cs{exp_not:V} \meta{variable}
+%   \end{syntax}
+%   Recovers the content of the \meta{variable}, then prevents expansion
+%   of the this material in a context where it would otherwise
+%   be expanded, for example an |x|-type argument.
+% \end{function}
+%
+% \begin{function}{\exp_not:v}
+%   \begin{syntax}
+%     \cs{exp_not:v} \Arg{tokens}
+%   \end{syntax}
+%   Expands the \meta{tokens} until only unexpandable content remains, and then
+%   converts this into a control sequence (which should be a \meta{variable}
+%   name). The content of the \meta{variable} is recovered, and further
+%   expansion is prevented in a context where it would otherwise
+%   be expanded, for example an |x|-type argument.
+% \end{function}
+%
+% \begin{function}{\exp_not:o}
+%   \begin{syntax}
+%     \cs{exp_not:o} \Arg{tokens}
+%   \end{syntax}
+%   Expands the \meta{tokens} once, then prevents any further expansion in a
+%   context where they would otherwise
+%   be expanded, for example an |x|-type argument.
+% \end{function}
+%
+% \begin{function}[EXP]{\exp_not:f}
+%   \begin{syntax}
+%     \cs{exp_not:f} \meta{tokens}
+%   \end{syntax}
+%   Expands \meta{tokens} fully until the first unexpandable token
+%   is found. Expansion then stops, and the result of the expansion
+%   (including any tokens which were not expanded) is protected from
+%   further expansion.
+% \end{function}
+%
+% \begin{function}[EXP]{\exp_stop_f:}
+%   \begin{syntax}
+%     \cs{function:f} \meta{tokens} \cs{exp_stop_f:} \meta{more tokens}
+%   \end{syntax}
+%   This function terminates an \texttt{f}-type expansion. Thus if
+%   a function \cs{function:f} starts an \texttt{f}-type expansion
+%   and all of \meta{tokens} are expandable \cs{exp_stop:f} will
+%   terminate the expansion of tokens even if \meta{more tokens}
+%   are also expandable. The function itself is an implicit space
+%   token. Inside an \texttt{x}-type expansion, it will retain its
+%   form, but when typeset it produces the underlying space (\verb*| |).
+% \end{function}
+%
+% \section{Internal functions and variables}
+%
+% \begin{variable}{\l_exp_tl}
+%   The |\exp_| module has its private variables to temporarily store
+%   results of the argument expansion. This is done to avoid interference
+%   with other functions using temporary variables.
+% \end{variable}
+%
+% \begin{function}[EXP]{\exp_eval_register:N, \exp_eval_register:c}
+%   \begin{syntax}
+%     \cs{exp_eval_register:N} \meta{variable}
+%   \end{syntax}
+%   These functions evaluates a \meta{variable} as part of a |V| or |v|
+%   expansion (respectively), preceeded by \cs{c_zero} which stops
+%   the expansion of a previous \cs{tex_romannumeral:D}.
+%   A \meta{variable} might exist as
+%   one of two things: a parameter-less non-long, non-protected macro
+%   or a built-in \TeX{} register such as |\count|.
+% \end{function}
+%
+% \begin{function}{\::n, \::N, \::c, \::o, \::f, \::x, \::v, \::V, \:::}
+%   \begin{syntax}
+%     |\cs_set_nopar:Npn \exp_args:Ncof { \::c \::o \::f \::: }|
+%   \end{syntax}
+%   Internal forms for the base expansion types. These names do \emph{not}
+%   conform to the general \LaTeX3 approach as this makes them more readily
+%   visible in the log and so forth.
+% \end{function}
+%
+% \begin{function}{\cs_generate_internal_variant:n}
+%   \begin{syntax}
+%     \cs{cs_generate_internal_variant:n} \meta{arg~spec}
+%   \end{syntax}
+%   Tests if the function |\exp_args:N|\meta{arg~spec} exists, and defines it
+%   if it does not. The \meta{arg~spec} should be a series of one or more
+%   of the letters |N|, |c|, |n|, |o|, |V|, |v|, |f| and |x|.
+% \end{function}
+%
+% \end{documentation}
+%
+% \begin{implementation}
+%
+% \section{\pkg{l3expan} implementation}
+%
+%    \begin{macrocode}
+%<*initex|package>
+%    \end{macrocode}
+%
+%    We start by ensuring that the required packages are loaded.
+%    \begin{macrocode}
+%<*package>
+\ProvidesExplPackage
+  {\ExplFileName}{\ExplFileDate}{\ExplFileVersion}{\ExplFileDescription}
+\package_check_loaded_expl:
+%</package>
+%    \end{macrocode}
+%
+% \begin{macro}{\exp_after:wN}
+% \begin{macro}{\exp_not:N}
+% \begin{macro}{\exp_not:n}
+%   These are defined in \pkg{l3basics}.
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \subsection{General expansion}
+%
+% In this section a general mechanism for defining functions to handle
+% argument handling is defined.  These general expansion functions are
+% expandable unless |x| is used.  (Any version of |x| is going to have
+% to use one of the \LaTeX3 names for |\cs_set_nopar:Npx| at some point, and
+% so is never going to be expandable.\footnote{However, some
+% primitives have certain characteristics that means that their
+% arguments undergo an \texttt{x} type expansion but the primitive
+% is in fact still expandable. We shall make it very clear when such
+% a function is expandable.})
+%
+% The definition of expansion functions with this technique happens
+% in section~\ref{sec:gendef}.
+% In section~\ref{sec:handtune} some common cases are coded by a more direct
+% method for efficiency, typically using calls to |\exp_after:wN|.
+%
+% \begin{variable}{\l_exp_tl}
+%   We need a scratch token list variable.
+%   We don't use |tl| methods so that \pkg{l3expan} can be loaded earlier.
+%    \begin{macrocode}
+\cs_new_nopar:Npn \l_exp_tl { }
+%    \end{macrocode}
+% \end{variable}
+%
+% This code uses internal functions with names that start with |\::|
+% to perform the expansions. All macros are |long| as this turned out
+% to be desirable since the tokens undergoing expansion may be
+% arbitrary user input.
+%
+% An argument manipulator |\::|\meta{Z} always has signature |#1\:::#2#3|
+% where |#1| holds the remaining argument manipulations to be performed,
+% |\:::| serves as an end marker for the list of manipulations, |#2|
+% is the carried over result of the previous expansion steps and |#3| is
+% the argument about to be processed.
+%
+% \begin{macro}[aux]{\exp_arg_next:nnn}
+% \begin{macro}[aux]{\exp_arg_next_nobrace:nnn}
+%   |#1| is the result of an expansion step, |#2| is the remaining
+%   argument manipulations and |#3| is the current result of the
+%   expansion chain.  This auxiliary function moves |#1| back after
+%   |#3| in the input stream and checks if any expansion is left to
+%   be done by calling |#2|. In by far the most cases we will require
+%   to add a set of braces to the result of an argument manipulation
+%   so it is more effective to do it directly here. Actually, so far
+%   only the |c| of the final argument manipulation variants does not
+%   require a set of braces.
+%    \begin{macrocode}
+\cs_new:Npn \exp_arg_next:nnn #1#2#3 { #2 \::: { #3 {#1} } }
+\cs_new:Npn \exp_arg_next_nobrace:nnn #1#2#3 { #2 \::: { #3 #1 } }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}[int]{\:::}
+%   The end marker is just another name for the identity function.
+%    \begin{macrocode}
+\cs_new:Npn \::: #1 {#1}
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}[int]{\::n}
+%   This function is used to skip an argument that doesn't need to
+%   be expanded.
+%    \begin{macrocode}
+\cs_new:Npn \::n #1 \::: #2#3 { #1 \::: { #2 {#3} } }
+%    \end{macrocode}
+% \end{macro}
+%
+%  \begin{macro}[int]{\::N}
+%   This function is used to skip an argument that consists of a
+%   single token and doesn't need to be expanded.
+%    \begin{macrocode}
+\cs_new:Npn \::N #1 \::: #2#3 { #1 \::: {#2#3} }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}[int]{\::c}
+%   This function is used to skip an argument that is turned into
+%   as control sequence without expansion.
+%    \begin{macrocode}
+\cs_new:Npn \::c #1 \::: #2#3
+  { \exp_after:wN \exp_arg_next_nobrace:nnn \cs:w #3 \cs_end: {#1} {#2} }
+%    \end{macrocode}
+% \end{macro}
+%
+%  \begin{macro}[int]{\::o}
+%   This function is used to expand an argument once.
+%    \begin{macrocode}
+\cs_new:Npn \::o #1 \::: #2#3
+  { \exp_after:wN \exp_arg_next:nnn \exp_after:wN {#3} {#1} {#2} }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\::f}
+% \begin{macro}{\exp_stop_f:}
+%   This function is used to expand a token list until the first
+%   unexpandable token is found. The underlying |\tex_romannumeral:D -`0|
+%   expands everything in its way to find something terminating the
+%   number and thereby expands the function in front of it. This
+%   scanning procedure is terminated once the expansion hits
+%   something non-expandable or a space. We introduce |\exp_stop_f:|
+%   to mark such an end of expansion marker; in case the scanner hits
+%   a number, this number also terminates the scanning and is left
+%   untouched. In the example shown earlier the scanning was stopped
+%   once \TeX{} had fully expanded |\cs_set_eq:Nc \aaa { b \l_tmpa_tl b }|
+%   into |\cs_set_eq:NwN \aaa = \blurb| which then turned out to contain
+%   the non-expandable token |\cs_set_eq:NwN|.  Since the expansion of
+%   |\tex_romannumeral:D -`0| is \meta{null}, we wind up with a fully
+%   expanded list, only \TeX{} has not tried to execute any of the
+%   non-expandable tokens. This is what differentiates this function
+%   from the |x| argument type.
+%    \begin{macrocode}
+\cs_new:Npn \::f #1 \::: #2#3
+  {
+    \exp_after:wN \exp_arg_next:nnn
+      \exp_after:wN { \tex_romannumeral:D -`0 #3 }
+      {#1} {#2}
+  }
+\use:nn { \cs_new_eq:NN \exp_stop_f: } { ~ }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}[int]{\::x}
+%   This function is used to expand an argument fully.
+%    \begin{macrocode}
+\cs_new_protected:Npn \::x #1 \::: #2#3
+  {
+    \cs_set_nopar:Npx \l_exp_tl { {#3} }
+    \exp_after:wN \exp_arg_next:nnn \l_exp_tl {#1} {#2}
+  }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}[int]{\::v}
+% \begin{macro}[int]{\::V}
+%   These functions return the value of a register, i.e., one of
+%   |tl|, |num|, |int|, |skip|, |dim| and |muskip|. The |V| version
+%   expects a single token whereas |v| like |c| creates a csname from
+%   its argument given in braces and then evaluates it as if it was a
+%   |V|. The primitive \cs{tex_romannumeral:D} sets off an expansion
+%   similar to an |f| type expansion, which we will terminate using
+%   \cs{c_zero}. The argument is returned in braces.
+%    \begin{macrocode}
+\cs_new:Npn \::V #1 \::: #2#3
+  {
+    \exp_after:wN \exp_arg_next:nnn
+      \exp_after:wN { \tex_romannumeral:D \exp_eval_register:N #3 }
+      {#1} {#2}
+}
+\cs_new:Npn \::v # 1\::: #2#3
+  {
+    \exp_after:wN \exp_arg_next:nnn
+      \exp_after:wN { \tex_romannumeral:D \exp_eval_register:c {#3} }
+      {#1} {#2}
+  }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+%  \begin{macro}[int]{\exp_eval_register:N, \exp_eval_register:c}
+%  \begin{macro}[aux]{\exp_eval_error_msg:w}
+%    This function evaluates a register. Now a register might exist as
+%    one of two things: A parameter-less macro or a built-in \TeX{}
+%    register such as |\count|. For the \TeX{} registers we have to
+%    utilize a |\tex_the:D| whereas for the macros we merely have to
+%    expand them once. The trick is to find out when to use
+%    |\tex_the:D| and when not to. What we do here is try to find out
+%    whether the token will expand to something else when hit with
+%    |\exp_after:wN|. The technique is to compare the meaning of the
+%    register in question when it has been prefixed with |\exp_not:N|
+%    and the register itself. If it is a macro, the prefixed
+%    |\exp_not:N| will temporarily turn it into the primitive
+%    |\scan_stop:|.
+%    \begin{macrocode}
+\cs_new_nopar:Npn \exp_eval_register:N #1
+  {
+    \exp_after:wN \if_meaning:w \exp_not:N #1 #1
+%    \end{macrocode}
+%   If the token was not a macro it may be a malformed variable from a
+%   |c| expansion in which case it is equal to the primitive
+%   |\scan_stop:|. In that case we throw an error. We could let \TeX{}
+%   do it for us but that would result in the rather obscure
+%   \begin{quote}
+%     |! You can't use `\relax' after \the.|
+%   \end{quote}
+%   which while quite true doesn't give many hints as to what actually
+%   went wrong. We provide something more sensible.
+%    \begin{macrocode}
+      \if_meaning:w \scan_stop: #1
+        \exp_eval_error_msg:w
+      \fi:
+%    \end{macrocode}
+%   The next bit requires some explanation. The function must be
+%   initiated by the primitive \cs{tex_romannumeral:D} and we want to
+%   terminate this expansion chain by inserting the \cs{c_zero} integer
+%   constant. However, we have to expand the register |#1| before we do
+%   that. If it is a \TeX{} register, we need to execute the sequence
+%   |\exp_after:wN \c_zero \tex_the:D #1| and if it is a macro we
+%   need to execute |\exp_after:wN \c_zero #1|. We therefore issue
+%   the longer of the two sequences and if the register is a macro, we
+%   remove the |\tex_the:D|.
+%    \begin{macrocode}
+    \else:
+      \exp_after:wN \use_i_ii:nnn
+    \fi:
+    \exp_after:wN \c_zero \tex_the:D #1
+  }
+\cs_new_nopar:Npn \exp_eval_register:c #1
+  { \exp_after:wN \exp_eval_register:N \cs:w #1 \cs_end: }
+%    \end{macrocode}
+%   Clean up nicely, then call the undefined control sequence. The
+%   result is an error message looking like this:
+%   \begin{verbatim}
+%     ! Undefined control sequence.
+%     \exp_eval_error_msg:w ...erroneous variable used!
+%
+%     l.55 \tl_set:Nv \l_tmpa_tl {undefined_tl}
+%   \end{verbatim}
+%    \begin{macrocode}
+\group_begin:%
+\tex_catcode:D`\!=11\scan_stop:%
+\tex_catcode:D`\ =11\scan_stop:%
+\cs_new:Npn\exp_eval_error_msg:w#1\tex_the:D#2{%
+\fi:\fi:\exp_after:wN\c_zero\erroneous variable used!}%
+\group_end:%
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \subsection{Hand-tuned definitions}
+% \label{sec:handtune}
+%
+% One of the most important features of these functions is that they
+% are fully expandable and therefore allow to prefix them with
+% |\pref_global:D| for example.
+%
+% \begin{macro}{\exp_args:No}
+% \begin{macro}{\exp_args:NNo}
+% \begin{macro}{\exp_args:NNNo}
+%   Those lovely runs of expansion!
+%    \begin{macrocode}
+\cs_new:Npn \exp_args:No #1#2 { \exp_after:wN #1 \exp_after:wN {#2} }
+\cs_new:Npn \exp_args:NNo #1#2#3
+  { \exp_after:wN #1 \exp_after:wN #2 \exp_after:wN {#3} }
+\cs_new:Npn \exp_args:NNNo #1#2#3#4
+  { \exp_after:wN #1 \exp_after:wN#2 \exp_after:wN #3 \exp_after:wN {#4} }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\exp_args:Nc}
+% In \pkg{l3basics}
+%\end{macro}
+%
+% \begin{macro}{\exp_args:cc, \exp_args:NNc, \exp_args:Ncc, \exp_args:Nccc}
+%    Here are the functions that turn their argument into csnames but
+%    are  expandable.
+%    \begin{macrocode}
+\cs_new:Npn \exp_args:cc #1#2
+  { \cs:w #1 \exp_after:wN \cs_end: \cs:w #2 \cs_end: }
+\cs_new:Npn \exp_args:NNc #1#2#3
+  { \exp_after:wN #1 \exp_after:wN #2 \cs:w # 3\cs_end: }
+\cs_new:Npn \exp_args:Ncc #1#2#3
+  { \exp_after:wN #1 \cs:w #2 \exp_after:wN \cs_end: \cs:w #3 \cs_end: }
+\cs_new:Npn \exp_args:Nccc #1#2#3#4
+  {
+    \exp_after:wN #1
+      \cs:w #2 \exp_after:wN \cs_end:
+      \cs:w #3 \exp_after:wN \cs_end:
+      \cs:w #4 \cs_end:
+  }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\exp_args:Nf, \exp_args:NV, \exp_args:Nv, \exp_args:Nx}
+%    \begin{macrocode}
+\cs_new:Npn \exp_args:Nf #1#2
+  { \exp_after:wN #1 \exp_after:wN { \tex_romannumeral:D -`0 #2 } }
+\cs_new:Npn \exp_args:Nv #1#2
+  {
+    \exp_after:wN #1 \exp_after:wN
+      { \tex_romannumeral:D \exp_eval_register:c {#2} }
+  }
+\cs_new:Npn \exp_args:NV #1#2
+  {
+    \exp_after:wN #1 \exp_after:wN
+      { \tex_romannumeral:D \exp_eval_register:N #2 }
+  }
+%    \end{macrocode}
+% \end{macro}
+%
+%  \begin{macro}{\exp_args:NNV,\exp_args:NNv,\exp_args:NNf,
+%                \exp_args:NVV,
+%                \exp_args:Ncf,\exp_args:Nco}
+%    Some more hand-tuned function with three arguments.
+%    If we force that an |o| argument always has braces,
+%    we could implement \cs{exp_args:Nco} with less tokens
+%    and only two arguments.
+%    \begin{macrocode}
+\cs_new:Npn \exp_args:NNf #1#2#3
+  {
+    \exp_after:wN #1
+    \exp_after:wN #2
+    \exp_after:wN { \tex_romannumeral:D -`0 #3 }
+  }
+\cs_new:Npn \exp_args:NNv #1#2#3
+  {
+    \exp_after:wN #1
+    \exp_after:wN #2
+    \exp_after:wN { \tex_romannumeral:D \exp_eval_register:c {#3} }
+  }
+\cs_new:Npn \exp_args:NNV #1#2#3
+  {
+    \exp_after:wN #1
+    \exp_after:wN #2
+    \exp_after:wN { \tex_romannumeral:D \exp_eval_register:N #3 }
+  }
+\cs_new:Npn \exp_args:Nco #1#2#3
+  {
+    \exp_after:wN #1
+    \cs:w #2 \exp_after:wN \cs_end:
+    \exp_after:wN {#3}
+  }
+\cs_new:Npn \exp_args:Ncf #1#2#3
+  {
+    \exp_after:wN #1
+    \cs:w #2 \exp_after:wN \cs_end:
+    \exp_after:wN { \tex_romannumeral:D -`0 #3 }
+  }
+\cs_new_nopar:Npn \exp_args:NVV #1#2#3
+  {
+    \exp_after:wN #1
+    \exp_after:wN { \tex_romannumeral:D \exp_after:wN
+      \exp_eval_register:N \exp_after:wN #2 \exp_after:wN }
+    \exp_after:wN { \tex_romannumeral:D \exp_eval_register:N #3 }
+  }
+%    \end{macrocode}
+%  \end{macro}
+%
+% \begin{macro}{\exp_args:Ncco,
+%               \exp_args:NcNc,
+%               \exp_args:NcNo,
+%               \exp_args:NNNV}
+%   A few more that we can hand-tune.
+%    \begin{macrocode}
+\cs_new:Npn \exp_args:NNNV #1#2#3#4
+  {
+    \exp_after:wN #1
+    \exp_after:wN #2
+    \exp_after:wN #3
+    \exp_after:wN { \tex_romannumeral:D \exp_eval_register:N #4 }
+  }
+\cs_new:Npn \exp_args:NcNc #1#2#3#4
+  {
+    \exp_after:wN #1
+    \cs:w #2 \exp_after:wN \cs_end:
+    \exp_after:wN #3
+    \cs:w #4 \cs_end:
+  }
+\cs_new:Npn \exp_args:NcNo #1#2#3#4
+  {
+    \exp_after:wN #1
+    \cs:w #2 \exp_after:wN \cs_end:
+    \exp_after:wN #3
+    \exp_after:wN {#4}
+  }
+\cs_new:Npn \exp_args:Ncco #1#2#3#4
+  {
+    \exp_after:wN #1
+    \cs:w #2 \exp_after:wN \cs_end:
+    \cs:w #3 \exp_after:wN \cs_end:
+    \exp_after:wN {#4}
+  }
+%    \end{macrocode}
+%  \end{macro}
+%  
+% \subsection{Definitions with the automated technique}
+% \label{sec:gendef}
+%
+% Some of these could be done more efficiently, but the complexity of
+% coding then becomes an issue. Notice that the auto-generated functions
+% are all not long: they don't actually take any arguments themselves.
+%
+%  \begin{macro}{\exp_args:Nx}
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \exp_args:Nx { \::x \::: }
+%    \end{macrocode}
+%  \end{macro}
+%
+%  \begin{macro}{\exp_args:NNx,
+%                \exp_args:Nnc,
+%                \exp_args:Ncx,
+%                \exp_args:Nfo,\exp_args:Nff,
+%                \exp_args:Nnf,\exp_args:Nno,\exp_args:NnV,\exp_args:Nnx,
+%                \exp_args:Noo,\exp_args:Noc,\exp_args:Nox,
+%                \exp_args:Nxo,\exp_args:Nxx}
+%    Here are the actual function definitions, using the helper functions
+%    above.
+%    \begin{macrocode}
+\cs_new_nopar:Npn \exp_args:Nnc { \::n \::c \::: }
+\cs_new_nopar:Npn \exp_args:Nfo { \::f \::o \::: }
+\cs_new_nopar:Npn \exp_args:Nff { \::f \::f \::: }
+\cs_new_nopar:Npn \exp_args:Nnf { \::n \::f \::: }
+\cs_new_nopar:Npn \exp_args:Nno { \::n \::o \::: }
+\cs_new_nopar:Npn \exp_args:NnV { \::n \::V \::: }
+\cs_new_nopar:Npn \exp_args:Noc { \::o \::c \::: }
+\cs_new_nopar:Npn \exp_args:Noo { \::o \::o \::: }
+\cs_new_protected_nopar:Npn \exp_args:NNx { \::N \::x \::: }
+\cs_new_protected_nopar:Npn \exp_args:Ncx { \::c \::x \::: }
+\cs_new_protected_nopar:Npn \exp_args:Nnx { \::n \::x \::: }
+\cs_new_protected_nopar:Npn \exp_args:Nox { \::o \::x \::: }
+\cs_new_protected_nopar:Npn \exp_args:Nxo { \::x \::o \::: }
+\cs_new_protected_nopar:Npn \exp_args:Nxx { \::x \::x \::: }
+%    \end{macrocode}
+%  \end{macro}
+%
+% \begin{macro}{\exp_args:Nccx,
+%               \exp_args:Ncnx,
+%               \exp_args:NNno,
+%               \exp_args:Nnno,
+%               \exp_args:Nnnx,
+%               \exp_args:Nnox,
+%               \exp_args:Nooo,
+%               \exp_args:Noox,
+%               \exp_args:Nnnc,
+%               \exp_args:NNnx,
+%               \exp_args:NNoo,
+%               \exp_args:NNox}
+%    \begin{macrocode}
+\cs_new_nopar:Npn \exp_args:NNno { \::N \::n \::o \::: }
+\cs_new_nopar:Npn \exp_args:NNoo { \::N \::o \::o \::: }
+\cs_new_nopar:Npn \exp_args:Nnnc { \::n \::n \::c \::: }
+\cs_new_nopar:Npn \exp_args:Nnno { \::n \::n \::o \::: }
+\cs_new_nopar:Npn \exp_args:Nooo { \::o \::o \::o \::: }
+\cs_new_protected_nopar:Npn \exp_args:NNnx { \::N \::n \::x \::: }
+\cs_new_protected_nopar:Npn \exp_args:NNox { \::N \::o \::x \::: }
+\cs_new_protected_nopar:Npn \exp_args:Nnnx { \::n \::n \::x \::: }
+\cs_new_protected_nopar:Npn \exp_args:Nnox { \::n \::o \::x \::: }
+\cs_new_protected_nopar:Npn \exp_args:Nccx { \::c \::c \::x \::: }
+\cs_new_protected_nopar:Npn \exp_args:Ncnx { \::c \::n \::x \::: }
+\cs_new_protected_nopar:Npn \exp_args:Noox { \::o \::o \::x \::: }
+%    \end{macrocode}
+%  \end{macro}
+%
+% \subsection{Last-unbraced versions}
+%
+% \begin{macro}[aux]{\exp_arg_last_unbraced:nn}
+% \begin{macro}[aux]{\::f_unbraced}
+% \begin{macro}[aux]{\::o_unbraced}
+% \begin{macro}[aux]{\::V_unbraced}
+% \begin{macro}[aux]{\::v_unbraced}
+%   There are a few places where the last argument needs to be available
+%   unbraced. First some helper macros.
+%    \begin{macrocode}
+\cs_new:Npn \exp_arg_last_unbraced:nn #1#2 { #2#1 }
+\cs_new:Npn \::f_unbraced \::: #1#2
+  {
+    \exp_after:wN \exp_arg_last_unbraced:nn
+      \exp_after:wN { \tex_romannumeral:D -`0 #2 } {#1}
+  }
+\cs_new:Npn \::o_unbraced \::: #1#2
+  { \exp_after:wN \exp_arg_last_unbraced:nn \exp_after:wN {#2} {#1} }
+\cs_new:Npn \::V_unbraced \::: #1#2
+  {
+    \exp_after:wN \exp_arg_last_unbraced:nn
+      \exp_after:wN { \tex_romannumeral:D \exp_eval_register:N #2 } {#1}
+  }
+\cs_new:Npn \::v_unbraced \::: #1#2
+  {
+    \exp_after:wN \exp_arg_last_unbraced:nn
+      \exp_after:wN { \tex_romannumeral:D \exp_eval_register:c {#2} } {#1}
+  }
+%    \end{macrocode}
+%\end{macro}
+%\end{macro}
+%\end{macro}
+%\end{macro}
+%\end{macro}
+%
+% \begin{macro}{\exp_last_unbraced:NV}
+% \begin{macro}{\exp_last_unbraced:Nv}
+% \begin{macro}{\exp_last_unbraced:Nf}
+% \begin{macro}{\exp_last_unbraced:No}
+% \begin{macro}{\exp_last_unbraced:NcV}
+% \begin{macro}{\exp_last_unbraced:NNV}
+% \begin{macro}{\exp_last_unbraced:NNo}
+% \begin{macro}{\exp_last_unbraced:Noo}
+% \begin{macro}{\exp_last_unbraced:Nfo}
+% \begin{macro}{\exp_last_unbraced:NNNV}
+% \begin{macro}{\exp_last_unbraced:NNNo}
+%   Now the business end: most of these are hand-tuned for speed, but the
+%   general system is in place.
+%    \begin{macrocode}
+\cs_new:Npn \exp_last_unbraced:NV #1#2
+  { \exp_after:wN #1 \tex_romannumeral:D \exp_eval_register:N #2 }
+\cs_new:Npn \exp_last_unbraced:Nv #1#2
+  { \exp_after:wN #1 \tex_romannumeral:D \exp_eval_register:c {#2} }
+\cs_new:Npn \exp_last_unbraced:No #1#2 { \exp_after:wN #1 #2 }
+\cs_new:Npn \exp_last_unbraced:Nf #1#2
+  { \exp_after:wN #1 \tex_romannumeral:D -`0 #2 }
+\cs_new:Npn \exp_last_unbraced:NcV #1#2#3
+  {
+    \exp_after:wN #1
+    \cs:w #2 \exp_after:wN \cs_end:
+    \tex_romannumeral:D \exp_eval_register:N #3
+  }
+\cs_new:Npn \exp_last_unbraced:NNV #1#2#3
+  {
+    \exp_after:wN #1
+    \exp_after:wN #2
+    \tex_romannumeral:D \exp_eval_register:N #3
+  }
+\cs_new:Npn \exp_last_unbraced:NNo #1#2#3
+  { \exp_after:wN #1 \exp_after:wN #2 #3 }
+\cs_new_nopar:Npn \exp_last_unbraced:Noo { \::o \::o_unbraced \::: }
+\cs_new_nopar:Npn \exp_last_unbraced:Nfo { \::f \::o_unbraced \::: }
+\cs_new:Npn \exp_last_unbraced:NNNV #1#2#3#4
+  {
+    \exp_after:wN #1
+    \exp_after:wN #2
+    \exp_after:wN #3
+    \tex_romannumeral:D \exp_eval_register:N #4
+  }
+\cs_new:Npn \exp_last_unbraced:NNNo #1#2#3#4
+  { \exp_after:wN #1 \exp_after:wN #2 \exp_after:wN #3 #4 }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% 
+% \begin{macro}{\exp_last_two_unbraced:Noo}
+%   If |#2| is a single token then this can be implemented as
+%   \begin{verbatim}
+%     \cs_new:Npn \exp_last_two_unbraced:Noo #1 #2 #3
+%      { \exp_after:wN \exp_after:wN \exp_after:wN #1 \exp_after:wN #2 #3 }
+%   \end{verbatim}
+%   However, for robustness this is not suitable. Instead, a bit of a
+%   shuffle is used to ensure that |#2| can be multiple tokens.
+%    \begin{macrocode}
+\cs_new:Npn \exp_last_two_unbraced:Noo #1#2#3
+  { \exp_after:wN \exp_last_two_unbraced_aux:nnN \exp_after:wN {#3} {#2} #1 }
+\cs_new:Npn \exp_last_two_unbraced_aux:nnN #1#2#3
+   { \exp_after:wN #3 #2 #1 }
+%    \end{macrocode}
+% \end{macro}
+%
+% \subsection{Preventing expansion}
+%
+%  \begin{macro}{\exp_not:o}
+%  \begin{macro}{\exp_not:f}
+%  \begin{macro}{\exp_not:V}
+%  \begin{macro}{\exp_not:v}
+%    \begin{macrocode}
+\cs_new:Npn \exp_not:o #1 { \etex_unexpanded:D \exp_after:wN {#1} }
+\cs_new:Npn \exp_not:f #1
+  { \etex_unexpanded:D \exp_after:wN { \tex_romannumeral:D -`0 #1 } }
+\cs_new:Npn \exp_not:V #1
+  {
+    \etex_unexpanded:D \exp_after:wN
+      { \tex_romannumeral:D \exp_eval_register:N #1 }
+  }
+\cs_new:Npn \exp_not:v #1
+  {
+    \etex_unexpanded:D \exp_after:wN
+      { \tex_romannumeral:D \exp_eval_register:c {#1} }
+  }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\exp_not:c}
+%    A helper function.
+%    \begin{macrocode}
+\cs_new:Npn \exp_not:c #1 { \exp_after:wN \exp_not:N \cs:w #1 \cs_end: }
+%    \end{macrocode}
+%  \end{macro}
+%
+% \subsection{Defining function variants}
+%
+% \begin{macro}{\cs_generate_variant:Nn}
+% \begin{macro}[aux]{\cs_generate_variant_aux:nnNNn}
+% \begin{macro}[aux]{\cs_generate_variant_aux:Nnnw}
+% \begin{macro}[aux]{\cs_generate_variant_aux:NNn}
+% \begin{macro}[aux]{\cs_generate_variant_aux:N}
+%   \begin{arguments}
+%     \item Base form of a function; \emph{e.g.},~|\tl_set:Nn|
+%     \item One or more variant argument specifiers; e.g., |{Nx,c,cx}|
+%   \end{arguments}
+%   Split up the original base function to grab its name and signature
+%   consisting of $k$ letters. Then we wish to iterate through the list
+%   of variant argument specifiers, and for each one construct a new
+%   function name using the original base name, the variant signature
+%   consisting of $l$ letters and the last $k-l$ letters of the base
+%   signature. For example, for a base function |\tl_set:Nn| which
+%   needs a |c| variant form, we want the new signature to be |cn|.
+%    \begin{macrocode}
+\cs_new_protected:Npn \cs_generate_variant:Nn #1
+  {
+    \chk_if_exist_cs:N #1
+    \cs_split_function:NN #1 \cs_generate_variant_aux:nnNNn
+    #1
+  }
+%    \end{macrocode}
+%   We discard the boolean |#3| and then set off a loop through the desired
+%   variant forms. The original function is retained as |#4| for efficiency.
+%    \begin{macrocode}
+\cs_new:Npn \cs_generate_variant_aux:nnNNn #1#2#3#4#5
+  { \cs_generate_variant_aux:Nnnw #4 {#1}{#2} #5 , ? , \q_recursion_stop }
+%    \end{macrocode}
+%   Next is the real work to be done. We now have 1: original function,
+%   2: base name, 3: base signature, 4: beginning of variant signature.
+%   To construct the new
+%   csname and the |\exp_args:Ncc| form, we need the variant signature.
+%   In our example, we wanted to discard the first two letters of the
+%   base signature because the variant form started with |cc|. This is
+%   the same as putting first |cc| in the signature and then
+%   |\use_none:nn| followed by the base signature |NNn|. We therefore
+%   call a small loop that outputs an |n| for each letter in the variant
+%   signature and use this to call the correct |\use_none:| variant.
+%
+%   Firstly though, we check whether to terminate the loop.
+%   Then build the variant function once, to avoid repeating this
+%   relatively expensive operation. Then recurse.
+%    \begin{macrocode}
+\cs_new:Npn \cs_generate_variant_aux:Nnnw #1#2#3#4 ,
+  {
+    \if:w ? #4
+      \exp_after:wN \use_none_delimit_by_q_recursion_stop:w
+    \fi:
+    \exp_args:NNc \cs_generate_variant_aux:NNn
+      #1
+      { #2 : #4 \use:c { use_none: \cs_generate_variant_aux:N #4 ? } #3 }
+      {#4}
+    \cs_generate_variant_aux:Nnnw #1 {#2} {#3}
+  }
+%    \end{macrocode}
+%   Check if the variant form has already been defined.
+%   If not, then define it and then additionally check if
+%   the |\exp_args:N| form needed is defined.
+%   Otherwise tell that it was already defined.
+%    \begin{macrocode}
+\cs_new:Npn \cs_generate_variant_aux:NNn #1 #2 #3
+  {
+    \cs_if_free:NTF #2
+      {
+        \cs_generate_variant_aux:NNpx #1 #2
+          { \exp_not:c { exp_args:N #3 } \exp_not:N #1 }
+        \cs_generate_internal_variant:n {#3}
+      }
+      {
+        \iow_log:x
+          {
+            Variant~\token_to_str:N #2~%
+            already~defined;~ not~ changing~ it~on~line~%
+            \tex_the:D \tex_inputlineno:D
+          }
+      }
+  }
+%    \end{macrocode}
+%   The small loop for defining the required number of |n|s. Break when
+%   seeing a |?|.
+%    \begin{macrocode}
+\cs_new:Npn \cs_generate_variant_aux:N #1
+  {
+    \if:w ? #1
+      \exp_after:wN \use_none:nn
+    \fi:
+    n
+    \cs_generate_variant_aux:N
+  }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}[aux]{\cs_generate_variant_aux:NNpx}
+% \begin{macro}[aux]{\cs_generate_variant_aux:w}
+%   The idea here is to pick up protected parent functions, using the
+%   nature of the meaning string that they generate. The test here is
+%   almost the same as \cs{tl_if_empty:nTF}, but has to be hard-coded as
+%   that function is not yet available and because it has to match both
+%   long and short macros.
+%    \begin{macrocode}
+\group_begin:
+  \tex_lccode:D `\Z = `\d \scan_stop:
+  \tex_lccode:D `\? =`\\ \scan_stop:
+  \tex_catcode:D `\P = 12 \scan_stop:
+  \tex_catcode:D `\R = 12 \scan_stop:
+  \tex_catcode:D `\O = 12 \scan_stop:
+  \tex_catcode:D `\T = 12 \scan_stop:
+  \tex_catcode:D `\E = 12 \scan_stop:
+  \tex_catcode:D `\C = 12 \scan_stop:
+  \tex_catcode:D `\Z = 12 \scan_stop:
+\tex_lowercase:D
+  {
+    \group_end:
+    \cs_new_nopar:Npn \cs_generate_variant_aux:NNpx #1
+      {
+        \exp_after:wN \cs_generate_variant_aux:w
+          \token_to_meaning:N #1 ? PROTECTEZ \q_stop
+      }
+    \cs_new:Npn \cs_generate_variant_aux:w #1 ? PROTECTEZ #2 \q_stop
+      {
+        \if_catcode:w a \etex_detokenize:D \exp_after:wN {#1} a
+          \exp_after:wN \cs_new_protected_nopar:Npx
+        \else:
+          \exp_after:wN \cs_new_nopar:Npx
+        \fi:
+      }
+  }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}[int]{\cs_generate_internal_variant:n}
+% \begin{macro}[aux]{\cs_generate_internal_variant_aux:N}
+%   Test if |exp_args:N #1| is already defined
+%   and if not define it via the
+%   |\::| commands using the chars in |#1|
+%    \begin{macrocode}
+\cs_new_protected:Npn \cs_generate_internal_variant:n #1
+  {
+    \cs_if_free:cT { exp_args:N #1 }
+      {
+        \cs_new:cpx { exp_args:N #1 }
+          { \cs_generate_internal_variant_aux:N #1 : }
+     }
+  }
+%    \end{macrocode}
+%   This command grabs char by char outputting |\::#1| (not expanded
+%   further) until we see a |:|. That colon is in fact also turned into
+%   |\:::| so that the required structure for |\exp_args...| commands
+%   is correctly terminated.
+%    \begin{macrocode}
+\cs_new:Npn \cs_generate_internal_variant_aux:N #1
+  {
+    \exp_not:c { :: #1 }
+    \if_meaning:w #1 :
+      \exp_after:wN \use_none:n
+    \fi:
+    \cs_generate_internal_variant_aux:N
+  }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+%\subsection{Variants which cannot be created earlier}
+%
+% \begin{macro}[pTF]
+%   {\str_if_eq:Vn, \str_if_eq:on, \str_if_eq:nV, \str_if_eq:no, \str_if_eq:VV}
+%   These cannot come earlier as they need \cs{cs_generate_variant:Nn}.
+%    \begin{macrocode}
+\cs_generate_variant:Nn \str_if_eq_p:nn {  V ,  o }
+\cs_generate_variant:Nn \str_if_eq_p:nn { nV , no , VV }
+\cs_generate_variant:Nn \str_if_eq:nnT  {  V ,  o }
+\cs_generate_variant:Nn \str_if_eq:nnT  { nV , no , VV }
+\cs_generate_variant:Nn \str_if_eq:nnF  {  V ,  o }
+\cs_generate_variant:Nn \str_if_eq:nnF  { nV , no , VV }
+\cs_generate_variant:Nn \str_if_eq:nnTF {  V ,  o }
+\cs_generate_variant:Nn \str_if_eq:nnTF { nV , no , VV }
+%    \end{macrocode}
+% \end{macro}
+%
+%    \begin{macrocode}
+%</initex|package>
+%    \end{macrocode}
+%
+% \end{implementation}
+%
+% \PrintIndex
diff --git a/Master/texmf-dist/source/latex/l3kernel/l3file.dtx b/Master/texmf-dist/source/latex/l3kernel/l3file.dtx
new file mode 100644
index 00000000000..7ec9de82c5e
--- /dev/null
+++ b/Master/texmf-dist/source/latex/l3kernel/l3file.dtx
@@ -0,0 +1,418 @@
+% \iffalse meta-comment
+%
+%% File: l3file.dtx Copyright (C) 2009-2011 The LaTeX3 Project
+%%
+%% It may be distributed and/or modified under the conditions of the
+%% LaTeX Project Public License (LPPL), either version 1.3c of this
+%% license or (at your option) any later version.  The latest version
+%% of this license is in the file
+%%
+%%    http://www.latex-project.org/lppl.txt
+%%
+%% This file is part of the "expl3 bundle" (The Work in LPPL)
+%% and all files in that bundle must be distributed together.
+%%
+%% The released version of this bundle is available from CTAN.
+%%
+%% -----------------------------------------------------------------------
+%%
+%% The development version of the bundle can be found at
+%%
+%%    http://www.latex-project.org/svnroot/experimental/trunk/
+%%
+%% for those people who are interested.
+%%
+%%%%%%%%%%%
+%% NOTE: %%
+%%%%%%%%%%%
+%%
+%%   Snapshots taken from the repository represent work in progress and may
+%%   not work or may contain conflicting material!  We therefore ask
+%%   people _not_ to put them into distributions, archives, etc. without
+%%   prior consultation with the LaTeX3 Project.
+%%
+%% -----------------------------------------------------------------------
+%
+%<*driver|package>
+\RequirePackage{l3names}
+\GetIdInfo$Id: l3file.dtx 2478 2011-06-19 21:34:23Z joseph $
+  {L3 Experimental file operations}
+%</driver|package>
+%<*driver>
+\documentclass[full]{l3doc}
+\begin{document}
+  \DocInput{\jobname.dtx}
+\end{document}
+%</driver>
+% \fi
+%
+% \title{^^A
+%   The \pkg{l3file} package\\ File operations^^A
+%   \thanks{This file describes v\ExplFileVersion,
+%      last revised \ExplFileDate.}^^A
+% }
+%
+% \author{^^A
+%  The \LaTeX3 Project\thanks
+%    {^^A
+%      E-mail:
+%        \href{mailto:latex-team@latex-project.org}
+%          {latex-team@latex-project.org}^^A
+%    }^^A
+% }
+%
+% \date{Released \ExplFileDate}
+%
+% \maketitle
+%
+% \begin{documentation}
+%
+% In contrast to the \pkg{l3io} module, which deals with the lowest
+% level of file management, the \pkg{l3file} module provides a higher
+% level interface for handling file contents. This involves providing
+% convenient wrappers around many of the functions in \pkg{l3io}
+% to make them more generally accessible.
+%
+% It is important to remember that \TeX{} will attempt to locate files using
+% both the operating system path and entries in the \TeX{} file database (most
+% \TeX{} systems use such a database). Thus the \enquote{current path} for
+% \TeX{} is somewhat broader than that for other programs.
+%
+% \section{File operation functions}
+%
+% \begin{variable}{\g_file_current_name_tl}
+%   Contains the name of the current \LaTeX{} file. This variable
+%   should not be modified: it is intended for information only. It
+%   will be equal to \cs{c_job_name_tl} at the start of a \LaTeX{}
+%   run and will be modified each time a file is read using
+%   \cs{file_input:n}.
+% \end{variable}
+%
+% \begin{function}[TF]{\file_if_exist:n}
+%   \begin{syntax}
+%     \cs{file_if_exist:nTF} \Arg{file name} \Arg{true code} \Arg{false code}
+%   \end{syntax}
+%   Searches for \meta{file name} using the current \TeX{} search
+%   path and the additional paths controlled by
+%   \cs{file_path_include:n}). The branching versions then leave
+%   either \meta{true code} or \meta{false code} in the input stream,
+%   as appropriate to the truth of the test and the variant of the
+%   function chosen.
+%   \begin{texnote}
+%     The \meta{file name} may contain both literal items and expandable
+%     content, which should on full expansion be the desired file name.
+%     The expansion occurs when \TeX{} searches for the file.
+%   \end{texnote}
+% \end{function}
+%
+% \begin{function}{\file_add_path:nN}
+%   \begin{syntax}
+%     \cs{file_add_path:nN} \Arg{file name} \meta{tl var}
+%   \end{syntax}
+%   Searches for \meta{file name} in the path as detailed for
+%   \cs{file_if_exist:nTF}, and if found sets the \meta{tl var} the
+%   fully-qualified name of the file, \emph{i.e}.~the path and file name.
+%   If the file is not found then the \meta{tl var} will be empty.
+%   \begin{texnote}
+%     The \meta{file name} may contain both literal items and expandable
+%     content, which should on full expansion be the desired file name.
+%     The expansion occurs when \TeX{} searches for the file.
+%   \end{texnote}
+% \end{function}
+%
+% \begin{function}{\file_input:n}
+%   \begin{syntax}
+%     \cs{file_input:n} \Arg{file name}
+%   \end{syntax}
+%   Searches for \meta{file name} in the path as detailed for
+%   \cs{file_if_exist:nTF}, and if found reads in the file as
+%   additional \LaTeX{} source. All files read are recorded
+%   for information and the file name stack is updated by this
+%   function.
+%   \begin{texnote}
+%     The \meta{file name} may contain both literal items and expandable
+%     content, which should on full expansion be the desired file name.
+%     The expansion occurs when \TeX{} searches for the file.
+%   \end{texnote}
+% \end{function}
+%
+% \begin{function}{\file_path_include:n}
+%   \begin{syntax}
+%     \cs{file_path_include:n} \Arg{path}
+%   \end{syntax}
+%   Adds \meta{path} to the list of those used to search for files
+%   by the \cs{file_input:n} and \cs{file_if_exist:n} function. The
+%   assignment is local.
+% \end{function}
+%
+% \begin{function}{\file_path_remove:n}
+%   \begin{syntax}
+%     \cs{file_path_remove:n} \Arg{path}
+%   \end{syntax}
+%   Removes \meta{path} from the list of those used to search for files
+%   by the \cs{file_input:n} and \cs{file_if_exist:n} function. The
+%   assignment is local.
+% \end{function}
+%
+% \begin{function}{\file_list:}
+%   \begin{syntax}
+%     \cs{file_list:}
+%   \end{syntax}
+%   This function will list all files loaded using \cs{file_input:n}
+%   in the log file.
+% \end{function}
+%
+% \section{Internal file functions}
+%
+% \begin{variable}{\g_file_stack_seq}
+%   Stores the stack of nested files loaded using \cs{file_input:n}. This
+%   is needed to restore the appropriate file name to
+%   \cs{g_file_current_name_tl} at the end of each file.
+% \end{variable}
+%
+% \begin{variable}{\g_file_record_seq}
+%   Stores the name of every file loaded using \cs{file_input:n}. In
+%   contrast to \cs{g_file_stack_seq}, no items are ever removed from this
+%   sequence.
+% \end{variable}
+%
+% \begin{variable}{\l_file_name_tl}
+%   Used to return the full name of a file for internal use.
+% \end{variable}
+%
+% \begin{variable}{\l_file_search_path_seq}
+%   The sequence of file paths to search when loading a file.
+% \end{variable}
+%
+% \begin{variable}{\l_file_search_path_saved_seq}
+%   When loaded on top of \LaTeXe{}, there is a need to save the search
+%   path so that \cs{input@path} can be used as appropriate.
+% \end{variable}
+%
+% \end{documentation}
+%
+% \begin{implementation}
+%
+% \section{\pkg{l3file} implementation}
+%
+% \TestFiles{m3file001}
+%
+%    \begin{macrocode}
+%<*initex|package>
+%    \end{macrocode}
+%
+%    \begin{macrocode}
+%<*package>
+\ProvidesExplPackage
+  {\ExplFileName}{\ExplFileDate}{\ExplFileVersion}{\ExplFileDescription}
+\package_check_loaded_expl:
+%</package>
+%    \end{macrocode}
+%
+% \begin{variable}{\g_file_current_name_tl}
+%   The name of the current file should be available at all times.
+%    \begin{macrocode}
+\tl_new:N \g_file_current_name_tl
+%    \end{macrocode}
+%   For the format the file name needs to be picked up at the start of the
+%   file. In package mode the current file name is collected from \LaTeXe{}.
+%    \begin{macrocode}
+%<*initex>
+\tex_everyjob:D \exp_after:wN
+  {
+    \tex_the:D \tex_everyjob:D
+    \tl_gset:Nx \g_file_current_name_tl { \tex_jobname:D }
+  }
+%</initex>
+%<*package>
+\tl_gset_eq:NN \g_file_current_name_tl \@currname
+%</package>
+%    \end{macrocode}
+% \end{variable}
+%
+% \begin{variable}{\g_file_stack_seq}
+%   The input list of files is stored as a sequence stack.
+%    \begin{macrocode}
+\seq_new:N \g_file_stack_seq
+%    \end{macrocode}
+% \end{variable}
+%
+% \begin{variable}{\g_file_record_seq}
+%   The total list of files used is recorded separately from the current file
+%   stack, as nothing is ever popped from this list.
+%    \begin{macrocode}
+\seq_new:N \g_file_record_seq
+%    \end{macrocode}
+% The current file name should be included in the file list!
+%    \begin{macrocode}
+%<*initex>
+\tex_everyjob:D \exp_after:wN
+  {
+    \tex_the:D \tex_everyjob:D
+    \seq_gput_right:NV \g_file_record_seq \g_file_current_name_tl
+  }
+%</initex>
+%    \end{macrocode}
+% \end{variable}
+%
+% \begin{variable}{\l_file_name_tl}
+%   Used to return the fully-qualified name of a file.
+%    \begin{macrocode}
+\tl_new:N \l_file_name_tl
+%    \end{macrocode}
+% \end{variable}
+%
+% \begin{variable}{\l_file_search_path_seq}
+%   The current search path.
+%    \begin{macrocode}
+\seq_new:N \l_file_search_path_seq
+%    \end{macrocode}
+% \end{variable}
+%
+% \begin{variable}{\l_file_search_path_saved_seq}
+%   The current search path has to be saved for package use.
+%    \begin{macrocode}
+%<*package>
+\seq_new:N \l_file_search_path_saved_seq
+%</package>
+%    \end{macrocode}
+% \end{variable}
+%
+% \begin{macro}{\file_add_path:nN}
+% \begin{variable}{\g_file_test_stream}
+% \begin{macro}[aux]{\file_add_path_search:nN}
+%   The way to test if a file exists is to try to open it: if it does not
+%   exist then \TeX{} will report end-of-file. For files which are in the
+%   current directory, this is straight-forward.  For other locations, a
+%   search has to be made looking at each potential path in turn. The first
+%   location is of course treated as the correct one. If nothing is found,
+%   |#2| is returned empty.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \file_add_path:nN #1#2
+  {
+    \ior_open:Nn \g_file_test_stream {#1}
+    \ior_if_eof:NTF \g_file_test_stream
+      { \file_add_path_search:nN {#1} #2 }
+      {
+        \ior_close:N \g_file_test_stream
+        \tl_set:Nx #2 {#1}
+      }
+  }
+\cs_new_protected_nopar:Npn \file_add_path_search:nN #1#2
+  {
+    \tl_clear:N #2
+%<*package>
+    \cs_if_exist:NT \input@path
+      {
+        \seq_set_eq:NN \l_file_search_path_saved_seq \l_file_search_path_seq
+        \clist_map_inline:Nn \input@path
+          { \seq_put_right:Nn \l_file_search_path_seq {##1} }
+      }
+%</package>
+    \seq_map_inline:Nn \l_file_search_path_seq
+      {
+        \ior_open:Nn \g_file_test_stream { ##1 #1 }
+        \ior_if_eof:NF \g_file_test_stream
+          {
+            \tl_set:Nx #2 { ##1 #1 }
+            \seq_map_break:
+          }
+      }
+%<*package>
+    \cs_if_exist:NT \input@path
+      { \seq_set_eq:NN \l_file_search_path_seq \l_file_search_path_saved_seq }
+%</package>
+    \ior_close:N \g_file_test_stream
+  }
+%    \end{macrocode}
+% \end{macro}
+% \end{variable}
+% \end{macro}
+%
+% \begin{macro}[TF]{\file_if_exist:n}
+%   The test for the existence of a file is a wrapper around the function to
+%   add a path to a file. If the file was found, the path will contain
+%   something, whereas if the file was not located then the return value
+%   will be empty.
+%    \begin{macrocode}
+\prg_new_protected_conditional:Nnn \file_if_exist:n { T , F , TF }
+  {
+    \file_add_path:nN {#1} \l_file_name_tl
+    \tl_if_empty:NTF \l_file_name_tl
+      { \prg_return_false: }
+      { \prg_return_true: }
+  }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\file_input:n}
+%   Loading a file is done in a safe way, checking first that the file
+%   exists and loading only if it does.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \file_input:n #1
+  {
+    \file_add_path:nN {#1} \l_file_name_tl
+    \tl_if_empty:NF \l_file_name_tl
+      {
+%<*initex>
+        \seq_gput_right:Nx \g_file_record_seq {#1}
+%</initex>
+%<*package>
+        \@addtofilelist {#1}
+%</package>
+        \seq_gpush:NV \g_file_stack_seq \g_file_current_name_tl
+        \tl_gset:Nn \g_file_current_name_tl {#1}
+        \exp_after:wN \tex_input:D \l_file_name_tl ~
+        \seq_gpop:NN \g_file_stack_seq \g_file_current_name_tl
+      }
+  }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\file_path_include:n}
+% \begin{macro}{\file_path_remove:n}
+%   Wrapper functions to manage the search path.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \file_path_include:n #1
+  {
+    \seq_if_in:NnF  \l_file_search_path_seq {#1}
+      { \seq_put_right:Nn \l_file_search_path_seq {#1} }
+  }
+\cs_new_protected_nopar:Npn \file_path_remove:n #1
+  { \seq_remove_all:Nn \l_file_search_path_seq {#1} }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\file_list:}
+%   A function to list all files used to the log.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \file_list:
+  {
+    \seq_remove_duplicates:N \g_file_record_seq
+    \iow_log:n { *~File~List~* }
+    \seq_map_inline:Nn \g_file_record_seq { \iow_log:n {##1} }
+    \iow_log:n { ************* }
+  }
+%    \end{macrocode}
+% \end{macro}
+%
+%   When used as a package, there is a need to hold onto the standard
+%   file list as well as the new one here.
+%    \begin{macrocode}
+%<*package>
+\AtBeginDocument
+  {
+    \clist_map_inline:Nn \@filelist
+      { \seq_put_right:Nn \g_file_record_seq {#1} }
+  }
+%</package>
+%    \end{macrocode}
+%
+%    \begin{macrocode}
+%</initex|package>
+%    \end{macrocode}
+%
+% \end{implementation}
+%
+% \PrintIndex
\ No newline at end of file
diff --git a/Master/texmf-dist/source/latex/l3kernel/l3final.dtx b/Master/texmf-dist/source/latex/l3kernel/l3final.dtx
new file mode 100644
index 00000000000..bc37e2aaee4
--- /dev/null
+++ b/Master/texmf-dist/source/latex/l3kernel/l3final.dtx
@@ -0,0 +1,112 @@
+% \iffalse meta-comment
+%
+%% File: l3final.dtx Copyright (C) 1990-2011 The LaTeX3 Project
+%%
+%% It may be distributed and/or modified under the conditions of the
+%% LaTeX Project Public License (LPPL), either version 1.3c of this
+%% license or (at your option) any later version.  The latest version
+%% of this license is in the file
+%%
+%%    http://www.latex-project.org/lppl.txt
+%%
+%% This file is part of the "expl3 bundle" (The Work in LPPL)
+%% and all files in that bundle must be distributed together.
+%%
+%% The released version of this bundle is available from CTAN.
+%%
+%% -----------------------------------------------------------------------
+%%
+%% The development version of the bundle can be found at
+%%
+%%    http://www.latex-project.org/svnroot/experimental/trunk/
+%%
+%% for those people who are interested.
+%%
+%%%%%%%%%%%
+%% NOTE: %%
+%%%%%%%%%%%
+%%
+%%   Snapshots taken from the repository represent work in progress and may
+%%   not work or may contain conflicting material!  We therefore ask
+%%   people _not_ to put them into distributions, archives, etc. without
+%%   prior consultation with the LaTeX3 Project.
+%%
+%% -----------------------------------------------------------------------
+%
+%<*driver|package>
+\RequirePackage{l3names}
+\GetIdInfo$Id: l3final.dtx 2478 2011-06-19 21:34:23Z joseph $
+  {L3 Experimental format finalisation}
+%</driver|package>
+%<*driver>
+\documentclass[full]{l3doc}
+\begin{document}
+  \DocInput{\jobname.dtx}
+\end{document}
+%</driver>
+% \fi
+%
+% \title{^^A
+%   The \pkg{l3final} package\\ Format finalisation^^A
+%   \thanks{This file describes v\ExplFileVersion,
+%      last revised \ExplFileDate.}^^A
+% }
+%
+% \author{^^A
+%  The \LaTeX3 Project\thanks
+%    {^^A
+%      E-mail:
+%        \href{mailto:latex-team@latex-project.org}
+%          {latex-team@latex-project.org}^^A
+%    }^^A
+% }
+%
+% \date{Released \ExplFileDate}
+%
+% \maketitle
+%
+% \begin{documentation}
+%
+% This module is the end of the \LaTeX3 format file. Currently, there
+% is not a lot happening here.
+% 
+% \end{documentation}
+% 
+% \begin{implementation}
+%
+% \section{\pkg{l3final} Implementation}
+%
+%    \begin{macrocode}
+%<*initex>
+%    \end{macrocode}
+%
+% \begin{macro}{\par}
+%   \TeX{} has a nasty habit of inserting a command with the name \cs{par}
+%   so we had better make sure that that command at least has a definition.
+%    \begin{macrocode}
+\cs_set_eq:NN \par \tex_par:D
+%    \end{macrocode}
+% \end{macro}
+%
+% The very last job is to dump the format, taking care to first leave
+% the code environment.
+%    \begin{macrocode}
+\use:n
+  {
+    \char_set_catcode_space:n          { 9 }
+    \char_set_catcode_space:n          { 32 }
+    \char_set_catcode_math_subscript:n { 94 }
+    \char_set_catcode_other:n          { 58 }
+    \char_set_catcode_active:n         { 126 }
+    \tex_endlinechar:D = 13 \scan_stop:
+    \tex_dump:D
+  }
+%    \end{macrocode}
+%
+%    \begin{macrocode}
+%</initex>
+%    \end{macrocode}
+%
+% \end{implementation}
+%
+%\PrintIndex
diff --git a/Master/texmf-dist/source/latex/l3kernel/l3fp.dtx b/Master/texmf-dist/source/latex/l3kernel/l3fp.dtx
new file mode 100644
index 00000000000..36ac323debd
--- /dev/null
+++ b/Master/texmf-dist/source/latex/l3kernel/l3fp.dtx
@@ -0,0 +1,5676 @@
+% \iffalse meta-comment
+%
+%% File: l3fp.dtx Copyright (C) 2010,2011 The LaTeX3 Project
+%%
+%% It may be distributed and/or modified under the conditions of the
+%% LaTeX Project Public License (LPPL), either version 1.3c of this
+%% license or (at your option) any later version.  The latest version
+%% of this license is in the file
+%%
+%%    http://www.latex-project.org/lppl.txt
+%%
+%% This file is part of the "expl3 bundle" (The Work in LPPL)
+%% and all files in that bundle must be distributed together.
+%%
+%% The released version of this bundle is available from CTAN.
+%%
+%% -----------------------------------------------------------------------
+%%
+%% The development version of the bundle can be found at
+%%
+%%    http://www.latex-project.org/svnroot/experimental/trunk/
+%%
+%% for those people who are interested.
+%%
+%%%%%%%%%%%
+%% NOTE: %%
+%%%%%%%%%%%
+%%
+%%   Snapshots taken from the repository represent work in progress and may
+%%   not work or may contain conflicting material!  We therefore ask
+%%   people _not_ to put them into distributions, archives, etc. without
+%%   prior consultation with the LaTeX3 Project.
+%%
+%% -----------------------------------------------------------------------
+%
+%<*driver|package>
+\RequirePackage{l3names}
+\GetIdInfo$Id: l3fp.dtx 2478 2011-06-19 21:34:23Z joseph $
+  {L3 Experimental floating-point operations}
+%</driver|package>
+%<*driver>
+\documentclass[full]{l3doc}
+\begin{document}
+  \DocInput{\jobname.dtx}
+\end{document}
+%</driver>
+% \fi
+%
+% \title{^^A
+%   The \pkg{l3fp} package\\ Floating-point operations^^A
+%   \thanks{This file describes v\ExplFileVersion,
+%      last revised \ExplFileDate.}^^A
+% }
+%
+% \author{^^A
+%  The \LaTeX3 Project\thanks
+%    {^^A
+%      E-mail:
+%        \href{mailto:latex-team@latex-project.org}
+%          {latex-team@latex-project.org}^^A
+%    }^^A
+% }
+%
+% \date{Released \ExplFileDate}
+%
+% \maketitle
+%
+% \begin{documentation}
+%
+% A floating point number is one which is stored as a mantissa and
+% a separate exponent. This module implements arithmetic using radix
+% $10$ floating point numbers. This means that the mantissa should
+% be a real number in the range $1 \le \expandafter\mathopen\string|
+% x \expandafter\mathclose\string| < 10$, with the
+% exponent given as an integer between $-99$ and $99$. In the
+% input, the exponent part is represented starting with an \texttt{e}.
+% As this is a low-level module, error-checking is minimal. Numbers
+% which are too large for the floating point unit to handle will result
+% in errors, either from \TeX{} or from \LaTeX{}. The \LaTeX{} code does not
+% check that the input will not overflow, hence the possibility of a
+% \TeX{} error. On the other hand, numbers which are too small will be
+% dropped, which will mean that extra decimal digits will simply be
+% lost.
+%
+% When parsing numbers, any missing parts will be interpreted as
+% zero. So for example
+%\begin{verbatim}
+%  \fp_set:Nn \l_my_fp { }
+%  \fp_set:Nn \l_my_fp { . }
+%  \fp_set:Nn \l_my_fp { - }
+% \end{verbatim}
+% will all be interpreted as zero values without raising an error.
+%
+% Operations which give an undefined result (such as division by
+% $0$) will not lead to errors. Instead special marker values are
+% returned, which can be tested for using fr example
+% \cs{fp_if_undefined:N(TF)}. In this way it is possible to work with
+% asymptotic functions without first checking the input. If these
+% special values are carried forward in calculations they will be
+% treated as $0$.
+%
+% Floating point numbers are stored in the \texttt{fp} floating point
+% variable type. This has a standard range of functions for
+% variable management.
+%
+% \section{Floating-point variables}
+%
+% \begin{function}{\fp_new:N, \fp_new:c}
+%   \begin{syntax}
+%     \cs{fp_new:N} \meta{floating point variable}
+%   \end{syntax}
+%   Creates a new \meta{floating point variable} or raises an error if
+%   the name is already taken. The declaration global. The
+%   \meta{floating point} will initially be set to  |+0.000000000e0|
+%   (the zero floating point).
+% \end{function}
+%
+% \begin{function}{\fp_const:Nn, \fp_const:cn}
+%   \begin{syntax}
+%     \cs{fp_const:Nn} \meta{floating point variable} \Arg{value}
+%   \end{syntax}
+%   Creates a new constant \meta{floating point variable} or raises an
+%   error  if the name is already taken. The value of the
+%   \meta{floating point variable} will be set globally to the
+%   \meta{value}.
+% \end{function}
+%
+% \begin{function}{\fp_set_eq:NN, \fp_set_eq:cN, \fp_set_eq:Nc, \fp_set_eq:cc}
+%   \begin{syntax}
+%     \cs{fp_set_eq:NN} \meta{fp var1} \meta{fp var2}
+%   \end{syntax}
+%   Sets the value of \meta{floating point variable1} equal to that of
+%   \meta{floating point variable2}. This assignment is restricted to the
+%   current \TeX{} group level.
+% \end{function}
+%
+% \begin{function}
+%   {\fp_gset_eq:NN, \fp_gset_eq:cN, \fp_gset_eq:Nc, \fp_gset_eq:cc}
+%   \begin{syntax}
+%     \cs{fp_gset_eq:NN} \meta{fp var1} \meta{fp var2}
+%   \end{syntax}
+%   Sets the value of \meta{floating point variable1} equal to that of
+%   \meta{floating point variable2}. This assignment is global and so is
+%   not limited by the current \TeX{} group level.
+% \end{function}
+%
+% \begin{function}{\fp_zero:N, \fp_zero:c}
+%   \begin{syntax}
+%     \cs{fp_zero:N} \meta{floating point variable}
+%   \end{syntax}
+%   Sets the \meta{floating point variable} to |+0.000000000e0| within
+%   the current scope.
+% \end{function}
+%
+% \begin{function}{\fp_gzero:N, \fp_gzero:c}
+%   \begin{syntax}
+%     \cs{fp_gzero:N} \meta{floating point variable}
+%   \end{syntax}
+%   Sets the \meta{floating point variable} to |+0.000000000e0| globally.
+% \end{function}
+%
+% \begin{function}{\fp_set:Nn, \fp_set:cn}
+%   \begin{syntax}
+%     \cs{fp_set:Nn} \meta{floating point variable} \Arg{value}
+%   \end{syntax}
+%   Sets the \meta{floating point variable} variable to \meta{value}
+%   within the scope of the current \TeX{} group.
+% \end{function}
+%
+% \begin{function}{\fp_gset:Nn, \fp_gset:cn}
+%   \begin{syntax}
+%     \cs{fp_gset:Nn} \meta{floating point variable} \Arg{value}
+%   \end{syntax}
+%   Sets the \meta{floating point variable} variable to \meta{value}
+%   globally.
+% \end{function}
+%
+% \begin{function}{\fp_set_from_dim:Nn, \fp_set_from_dim:cn}
+%   \begin{syntax}
+%     \cs{fp_set_from_dim:Nn} \meta{floating point variable} \Arg{dimexpr}
+%   \end{syntax}
+%   Sets the \meta{floating point variable} to the distance represented
+%   by the \meta{dimension expression} in the units points. This means
+%   that distances given in other units are first converted to points
+%   before being assigned to the \meta{floating point variable}. The
+%   assignment is local.
+% \end{function}
+%
+% \begin{function}{\fp_gset_from_dim:Nn, \fp_gset_from_dim:cn}
+%   \begin{syntax}
+%     \cs{fp_gset_from_dim:Nn} \meta{floating point variable} \Arg{dimexpr}
+%   \end{syntax}
+%   Sets the \meta{floating point variable} to the distance represented
+%   by the \meta{dimension expression} in the units points. This means
+%   that distances given in other units are first converted to points
+%   before being assigned to the \meta{floating point variable}. The
+%   assignment is global.
+% \end{function}
+%
+% \begin{function}[EXP]{\fp_use:N, \fp_use:c}
+%   \begin{syntax}
+%     \cs{fp_use:N} \meta{floating point variable}
+%   \end{syntax}
+%   Inserts the value of the \meta{floating point variable} into the
+%   input stream. The value will be given as a real number without any
+%   exponent part, and will always include a decimal point. For example,
+%   \begin{verbatim}
+%     \fp_new:Nn \test
+%     \fp_set:Nn \test { 1.234 e 5 }
+%     \fp_use:N \test
+%   \end{verbatim}
+%   will insert |12345.00000| into the input stream.
+%   As illustrated, a floating point will always be inserted with ten
+%   significant digits given. Very large and very small values will
+%   include additional zeros for place value.
+% \end{function}
+%
+% \begin{function}{\fp_show:N, \fp_show:c}
+%   \begin{syntax}
+%     \cs{fp_show:N} \meta{floating point variable}
+%   \end{syntax}
+%   Displays the content of the \meta{floating point variable} on the
+%   terminal.
+% \end{function}
+%
+% \section{Conversion of floating point values to other formats}
+%
+% It is useful to be able to convert floating point variables to
+% other forms. These functions are expandable, so that the material
+% can be used in a variety of contexts. The \cs{fp_use:N} function
+% should also be consulted in this context, as it will insert the
+% value of the floating point variable as a real number.
+%
+% \begin{function}[EXP]{\fp_to_dim:N, \fp_to_dim:c}
+%   \begin{syntax}
+%     \cs{fp_to_dim:N} \meta{floating point variable}
+%   \end{syntax}
+%   Inserts the value of the \meta{floating point variable}
+%   into the input stream converted into a dimension in points.
+% \end{function}
+%
+% \begin{function}[EXP]{\fp_to_int:N, \fp_to_int:c}
+%   \begin{syntax}
+%     \cs{fp_to_int:N} \meta{floating point variable}
+%   \end{syntax}
+%   Inserts the integer value of the \meta{floating point variable}
+%   into the input stream. The decimal part of the number will not be
+%   included, but will be used to round the integer.
+% \end{function}
+%
+% \begin{function}[EXP]{\fp_to_tl:N, \fp_to_tl:c}
+%   \begin{syntax}
+%     \cs{fp_to_tl:N} \meta{floating point variable}
+%   \end{syntax}
+%   Inserts a representation of the \meta{floating point variable} into
+%   the input stream as a token list. The representation follows the
+%   conventions of a pocket calculator:
+%   \begin{center}
+%     \ttfamily
+%     \begin{tabular}{r@{.}lr@{.}l}
+%       \toprule
+%         \multicolumn{2}{l}{\rmfamily{Floating point value}} &
+%         \multicolumn{2}{l}{\rmfamily{Representation}} \\
+%       \midrule
+%          1 & 234000000000e0  &  1 & 234    \\
+%         -1 & 234000000000e0  & -1 & 234    \\
+%          1 & 234000000000e3  &  \multicolumn{2}{l}{1234} \\
+%          1 & 234000000000e13 &  \multicolumn{2}{l}{1234e13} \\
+%          1 & 234000000000e-1 & 0 & 1234   \\
+%          1 & 234000000000e-2 & 0 & 01234  \\
+%          1 & 234000000000e-3 & 1 & 234e-3 \\
+%       \bottomrule
+%     \end{tabular}
+%   \end{center}
+%   Notice that trailing zeros are removed in this process, and that
+%   numbers which do not require a decimal part do \emph{not} include
+%   a decimal marker.
+% \end{function}
+%
+% \section{Rounding floating point values}
+%
+% The module can round floating point values to either decimal places
+% or significant figures using the usual method in which exact halves
+% are rounded up.
+%
+% \begin{function}{\fp_round_figures:Nn, \fp_round_figures:cn}
+%   \begin{syntax}
+%     \cs{fp_round_figures:Nn} \meta{floating point variable} \Arg{target}
+%   \end{syntax}
+%   Rounds the \meta{floating point variable} to the \meta{target} number
+%   of significant figures (an integer expression). The rounding is
+%   carried out locally.
+% \end{function}
+%
+% \begin{function}{\fp_ground_figures:Nn, \fp_ground_figures:cn}
+%   \begin{syntax}
+%     \cs{fp_ground_figures:Nn} \meta{floating point variable} \Arg{target}
+%   \end{syntax}
+%   Rounds the \meta{floating point variable} to the \meta{target} number
+%   of significant figures (an integer expression). The rounding is
+%   carried out globally.
+% \end{function}
+%
+% \begin{function}{\fp_round_places:Nn, \fp_round_places:cn}
+%   \begin{syntax}
+%     \cs{fp_round_places:Nn} \meta{floating point variable} \Arg{target}
+%   \end{syntax}
+%   Rounds the \meta{floating point variable} to the \meta{target} number
+%   of decimal places (an integer expression). The rounding is
+%   carried out locally.
+% \end{function}
+%
+% \begin{function}{\fp_ground_places:Nn, \fp_ground_places:cn}
+%   \begin{syntax}
+%     \cs{fp_ground_places:Nn} \meta{floating point variable} \Arg{target}
+%   \end{syntax}
+%   Rounds the \meta{floating point variable} to the \meta{target} number
+%   of decimal places (an integer expression). The rounding is
+%   carried out globally.
+% \end{function}
+%
+% \section{Floating-point conditionals}
+%
+% \begin{function}[EXP,pTF]{\fp_if_undefined:N}
+%   \begin{syntax}
+%     \cs{fp_if_undefined_p:N} \meta{fixed-point}
+%     \cs{fp_if_undefined:NTF} \meta{fixed-point}
+%     ~~\Arg{true code} \Arg{false code}
+%   \end{syntax}
+%   Tests if \meta{floating point} is undefined (\emph{i.e.}~equal to the
+%   special \cs{c_undefined_fp} variable). The branching versions then
+%   leave either \meta{true code} or \meta{false code} in the input
+%   stream, as appropriate to the truth of the test and the variant of
+%   the function chosen. The logical truth of the test is left in the
+%   input stream by the predicate version.
+% \end{function}
+%
+% \begin{function}[EXP]{\fp_if_zero:N}
+%   \begin{syntax}
+%     \cs{fp_if_zero_p:N} \meta{fixed-point}
+%     \cs{fp_if_zero:NTF} \meta{fixed-point} \Arg{true code} \Arg{false code}
+%   \end{syntax}
+%   Tests if \meta{floating point} is equal to zero (\emph{i.e.}~equal to
+%   the special \cs{c_zero_fp} variable). The branching versions then
+%   leave either \meta{true code} or \meta{false code} in the input
+%   stream, as appropriate to the truth of the test and the variant of
+%   the function chosen. The logical truth of the test is left in the
+%   input stream by the predicate version.
+% \end{function}
+%
+% \begin{function}[TF]{\fp_compare:nNn}
+%   \begin{syntax}
+%     \cs{fp_compare:nNnTF}
+%     ~~\Arg{floating point1} \meta{relation} \Arg{floating point2}
+%     ~~\Arg{true code} \Arg{false code}
+%   \end{syntax}
+%   This function compared the two \meta{floating point} values, which
+%   may be stored as \texttt{fp} variables, using the \meta{relation}:
+%   \begin{center}
+%     \begin{tabular}{ll}
+%       Equal                 & |=| \\
+%       Greater than          & |>| \\
+%       Less than             & |<| \\
+%     \end{tabular}
+%   \end{center}
+%   Either \meta{true code} or \meta{false code} is then left in the
+%   input stream, as appropriate to the truth of the test and the variant
+%   of the function chosen. The tests treat undefined floating points as
+%   zero as the comparison is intended for real numbers only.
+% \end{function}
+%
+% \begin{function}[TF]{\fp_compare:n}
+%   \begin{syntax}
+%     \cs{fp_compare:nTF}
+%     ~~\{ \meta{floating point1} \meta{relation} \meta{floating point2} \}
+%     ~~\Arg{true code} \Arg{false code}
+%   \end{syntax}
+%   This function compared the two \meta{floating point} values, which
+%   may be stored as \texttt{fp} variables, using the \meta{relation}:
+%   \begin{center}
+%     \begin{tabular}{ll}
+%       Equal                 & |=| or |==| \\
+%       Greater than          & |>|  \\
+%       Greater than or equal & |>=| \\
+%       Less than             & |<|  \\
+%       Less than or equal    & |<=| \\
+%       Not equal             & |!=| \\
+%     \end{tabular}
+%   \end{center}
+%   Either \meta{true code} or \meta{false code} is then left in the
+%   input stream, as appropriate to the truth of the test and the variant
+%   of the function chosen. The tests treat undefined floating points as
+%   zero as the comparison is intended for real numbers only.
+% \end{function}
+%
+% \section{Unary floating-point operations}
+%
+% The unary operations alter the value stored within an \texttt{fp}
+% variable.
+%
+% \begin{function}{\fp_abs:N, \fp_abs:c}
+%   \begin{syntax}
+%     \cs{fp_abs:N} \meta{floating point variable}
+%   \end{syntax}
+%   Converts the \meta{floating point variable} to its absolute value,
+%   assigning the result within the current \TeX\ group.
+% \end{function}
+%
+% \begin{function}{\fp_gabs:N, \fp_gabs:c}
+%   \begin{syntax}
+%     \cs{fp_gabs:N} \meta{floating point variable}
+%   \end{syntax}
+%   Converts the \meta{floating point variable} to its absolute value,
+%   assigning the result globally.
+% \end{function}
+%
+% \begin{function}{\fp_neg:N, \fp_neg:c}
+%   \begin{syntax}
+%     \cs{fp_neg:N} \meta{floating point variable}
+%   \end{syntax}
+%   Reverse the sign of the \meta{floating point variable}, assigning the
+%   result within the current \TeX\ group.
+% \end{function}
+%
+% \begin{function}{\fp_gneg:N, \fp_gneg:c}
+%   \begin{syntax}
+%     \cs{fp_gneg:N} \meta{floating point variable}
+%   \end{syntax}
+%   Reverse the sign of the \meta{floating point variable}, assigning the
+%   result globally.
+% \end{function}
+%
+% \section{Floating-point arithmetic}
+%
+% Binary arithmetic operations act on the value stored in an
+% \texttt{fp}, so for example
+% \begin{verbatim}
+%   \fp_set:Nn \l_my_fp { 1.234 }
+%   \fp_sub:Nn \l_my_fp { 5.678 }
+% \end{verbatim}
+%  sets \cs{l_my_fp} to the result of $1.234 - 5.678$
+%  (\emph{i.e.}~$-4.444$).
+%
+% \begin{function}{\fp_add:Nn, \fp_add:cn}
+%   \begin{syntax}
+%     \cs{fp_add:Nn} \meta{floating point} \Arg{value}
+%   \end{syntax}
+%   Adds the \meta{value} to the \meta{floating point}, making the
+%   assignment within the current \TeX{} group level.
+% \end{function}
+%
+% \begin{function}{\fp_gadd:Nn, \fp_gadd:cn}
+%   \begin{syntax}
+%     \cs{fp_gadd:Nn} \meta{floating point} \Arg{value}
+%   \end{syntax}
+%   Adds the \meta{value} to the \meta{floating point}, making the
+%   assignment globally.
+% \end{function}
+%
+% \begin{function}{\fp_sub:Nn, \fp_sub:cn}
+%   \begin{syntax}
+%     \cs{fp_sub:Nn} \meta{floating point} \Arg{value}
+%   \end{syntax}
+%   Subtracts the \meta{value} from the \meta{floating point}, making the
+%   assignment within the current \TeX{} group level.
+% \end{function}
+%
+% \begin{function}{\fp_gsub:Nn, \fp_gsub:cn}
+%   \begin{syntax}
+%     \cs{fp_gsub:Nn} \meta{floating point} \Arg{value}
+%   \end{syntax}
+%   Subtracts the \meta{value} from the \meta{floating point}, making the
+%   assignment globally.
+% \end{function}
+%
+% \begin{function}{\fp_mul:Nn, \fp_mul:cn}
+%   \begin{syntax}
+%     \cs{fp_mul:Nn} \meta{floating point} \Arg{value}
+%   \end{syntax}
+%   Multiples the \meta{floating point} by the \meta{value}, making the
+%   assignment within the current \TeX{} group level.
+% \end{function}
+%
+% \begin{function}{\fp_gmul:Nn, \fp_gmul:cn}
+%   \begin{syntax}
+%     \cs{fp_gmul:Nn} \meta{floating point} \Arg{value}
+%   \end{syntax}
+%   Multiples the \meta{floating point} by the \meta{value}, making the
+%   assignment globally.
+% \end{function}
+%
+% \begin{function}{\fp_div:Nn, \fp_div:cn}
+%   \begin{syntax}
+%     \cs{fp_div:Nn} \meta{floating point} \Arg{value}
+%   \end{syntax}
+%   Divides the \meta{floating point} by the \meta{value}, making the
+%   assignment within the current \TeX{} group level. If the \meta{value}
+%   is zero, the \meta{floating point} will be set to
+%   \cs{c_undefined_fp}. The assignment is local.
+% \end{function}
+%
+% \begin{function}{\fp_gdiv:Nn, \fp_gdiv:cn}
+%   \begin{syntax}
+%     \cs{fp_gdiv:Nn} \meta{floating point} \Arg{value}
+%   \end{syntax}
+%   Divides the \meta{floating point} by the \meta{value}, making the
+%   assignment globally. If the \meta{value} is zero, the
+%   \meta{floating point} will be set to \cs{c_undefined_fp}.
+%   The assignment is global.
+% \end{function}
+%
+% \section{Floating-point power operations}
+%
+% \begin{function}{\fp_pow:Nn, \fp_pow:cn}
+%   \begin{syntax}
+%     \cs{fp_pow:Nn} \meta{floating point} \Arg{value}
+%   \end{syntax}
+%   Raises the \meta{floating point} to the given \meta{value}. If the
+%   \meta{floating point} is negative, then the \meta{value} should be
+%   either a positive real number or a negative integer. If the
+%   \meta{floating point} is positive, then the \meta{value} may be any
+%   real value. Mathematically invalid operations such as $0^{0}$
+%   will give set the \meta{floating point} to  to \cs{c_undefined_fp}.
+%   The assignment is local.
+% \end{function}
+%
+% \begin{function}{\fp_gpow:Nn, \fp_gpow:cn}
+%   \begin{syntax}
+%     \cs{fp_gpow:Nn} \meta{floating point} \Arg{value}
+%   \end{syntax}
+%   Raises the \meta{floating point} to the given \meta{value}. If the
+%   \meta{floating point} is negative, then the \meta{value} should be
+%   either a positive real number or a negative integer. If the
+%   \meta{floating point} is positive, then the \meta{value} may be any
+%   real value. Mathematically invalid operations such as $0^{0}$
+%   will give set the \meta{floating point} to  to \cs{c_undefined_fp}.
+%   The assignment is global.
+% \end{function}
+%
+% \section{Exponential and logarithm functions}
+%
+% \begin{function}{\fp_exp:Nn, \fp_exp:cn}
+%   \begin{syntax}
+%     \cs{fp_exp:Nn} \meta{floating point} \Arg{value}
+%   \end{syntax}
+%   Calculates the exponential of the \meta{value} and assigns this
+%   to the \meta{floating point}. The assignment is local.
+% \end{function}
+%
+% \begin{function}{\fp_gexp:Nn, \fp_gexp:cn}
+%   \begin{syntax}
+%     \cs{fp_gexp:Nn} \meta{floating point} \Arg{value}
+%   \end{syntax}
+%   Calculates the exponential of the \meta{value} and assigns this
+%   to the \meta{floating point}. The assignment is global.
+% \end{function}
+%
+% \begin{function}{\fp_ln:Nn, \fp_ln:cn}
+%   \begin{syntax}
+%     \cs{fp_ln:Nn} \meta{floating point} \Arg{value}
+%   \end{syntax}
+%   Calculates the natural logarithm of the \meta{value} and assigns
+%   this to the \meta{floating point}. The assignment is local.
+% \end{function}
+%
+% \begin{function}{\fp_gln:Nn, \fp_gln:cn}
+%   \begin{syntax}
+%     \cs{fp_gln:Nn} \meta{floating point} \Arg{value}
+%   \end{syntax}
+%   Calculates the natural logarithm of the \meta{value} and assigns
+%   this to the \meta{floating point}. The assignment is global.
+% \end{function}
+%
+% \section{Trigonometric functions}
+%
+% The trigonometric functions all work in radians. They accept a maximum
+% input value of $100\,000\,000$, as there are issues with range
+% reduction and very large input values.
+%
+% \begin{function}{\fp_sin:Nn, \fp_sin:cn}
+%   \begin{syntax}
+%     \cs{fp_sin:Nn} \meta{floating point} \Arg{value}
+%   \end{syntax}
+%   Assigns the sine of the \meta{value} to the \meta{floating point}.
+%   The \meta{value} should be given in radians. The assignment is
+%   local.
+% \end{function}
+%
+% \begin{function}{\fp_gsin:Nn, \fp_gsin:cn}
+%   \begin{syntax}
+%     \cs{fp_gsin:Nn} \meta{floating point} \Arg{value}
+%   \end{syntax}
+%   Assigns the sine of the \meta{value} to the \meta{floating point}.
+%   The \meta{value} should be given in radians. The assignment is
+%   global.
+% \end{function}
+%
+% \begin{function}{\fp_cos:Nn, \fp_cos:cn}
+%   \begin{syntax}
+%     \cs{fp_cos:Nn} \meta{floating point} \Arg{value}
+%   \end{syntax}
+%   Assigns the cosine of the \meta{value} to the \meta{floating point}.
+%   The \meta{value} should be given in radians. The assignment is
+%   local.
+% \end{function}
+%
+% \begin{function}{\fp_gcos:Nn, \fp_gcos:cn}
+%   \begin{syntax}
+%     \cs{fp_gcos:Nn} \meta{floating point} \Arg{value}
+%   \end{syntax}
+%   Assigns the cosine of the \meta{value} to the \meta{floating point}.
+%   The \meta{value} should be given in radians. The assignment is
+%   global.
+% \end{function}
+%
+% \begin{function}{\fp_tan:Nn, \fp_tan:cn}
+%   \begin{syntax}
+%     \cs{fp_tan:Nn} \meta{floating point} \Arg{value}
+%   \end{syntax}
+%   Assigns the tangent of the \meta{value} to the \meta{floating point}.
+%   The \meta{value} should be given in radians. The assignment is
+%   local.
+% \end{function}
+%
+% \begin{function}{\fp_gtan:Nn, \fp_gtan:cn}
+%   \begin{syntax}
+%     \cs{fp_gtan:Nn} \meta{floating point} \Arg{value}
+%   \end{syntax}
+%   Assigns the tangent of the \meta{value} to the \meta{floating point}.
+%   The \meta{value} should be given in radians. The assignment is
+%   global.
+% \end{function}
+%
+% \section{Constant floating point values}
+%
+% \begin{variable}{\c_e_fp}
+%   The value of the base of natural numbers, $\mathrm{e}$.
+% \end{variable}
+%
+% \begin{variable}{\c_one_fp}
+%   A floating point variable with permanent value $1$: used for
+%   speeding up some comparisons.
+% \end{variable}
+%
+% \begin{variable}{\c_pi_fp}
+%   The value of $\pi$.
+% \end{variable}
+%
+% \begin{variable}{\c_undefined_fp}
+%   A special marker floating point variable representing the result of
+%   an operation which does not give a defined result (such as division
+%   by $0$).
+% \end{variable}
+%
+% \begin{variable}{\c_zero_fp}
+%   A permanently zero floating point variable.
+% \end{variable}
+%
+% \section{Notes on the floating point unit}
+%
+% As calculation of the elemental transcendental functions is
+% computationally expensive compared to storage of results, after
+% calculating a trigonometric function, exponent, \emph{etc.}~the module
+% stored the result for reuse. Thus the performance of the module for
+% repeated operations, most probably trigonometric functions, should be
+% much higher than if the values were re-calculated every time they
+% were needed.
+%
+% Anyone with experience of programming floating point calculations will
+% know that this is a complex area. The aim of the unit is to be
+% accurate enough for the likely applications in a typesetting context.
+% The arithmetic operations are therefore intended to provide ten digit
+% accuracy with the last digit accurate to $\pm 1$. The elemental
+% transcendental functions may not provide such high accuracy in every
+% case, although the design aim has been to provide $10$ digit
+% accuracy for cases likely to be relevant in typesetting situations.
+% A good overview of the challenges in this area can be found in
+% J.-M.~Muller, \emph{Elementary functions: algorithms and
+% implementation}, 2nd edition, Birkh{\"a}uer Boston, New York, USA,
+% 2006.
+%
+% The internal representation of numbers is tuned to the needs of the
+% underlying \TeX{} system. This means that the format is somewhat
+% different from that used in, for example, computer floating point
+% units. Programming in \TeX{} makes it most convenient to use a
+% radix $10$ system, using \TeX{} \texttt{count} registers for
+% storage and taking advantage where possible of delimited arguments.
+%
+% \end{documentation}
+%
+% \begin{implementation}
+%
+% \section{\pkg{l3fp} Implementation}
+%
+% \TestFiles{m3fp003.lvt}
+%
+%    \begin{macrocode}
+%<*initex|package>
+%    \end{macrocode}
+%
+%    \begin{macrocode}
+%<*package>
+\ProvidesExplPackage
+  {\ExplFileName}{\ExplFileDate}{\ExplFileVersion}{\ExplFileDescription}
+\package_check_loaded_expl:
+%</package>
+%    \end{macrocode}
+%
+% \subsection{Constants}
+%
+% \begin{variable}{\c_forty_four}
+% \begin{variable}{\c_one_million}
+% \begin{variable}{\c_one_hundred_million}
+% \begin{variable}{\c_five_hundred_million}
+% \begin{variable}{\c_one_thousand_million}
+%   There is some speed to gain by moving numbers into fixed positions.
+%    \begin{macrocode}
+\int_const:Nn \c_forty_four { 44 }
+\int_const:Nn \c_one_million { 1 000 000 }
+\int_const:Nn \c_one_hundred_million { 100 000 000 }
+\int_const:Nn \c_five_hundred_million { 500 000 000 }
+\int_const:Nn \c_one_thousand_million { 1 000 000 000 }
+%    \end{macrocode}
+% \end{variable}
+% \end{variable}
+% \end{variable}
+% \end{variable}
+% \end{variable}
+%
+% \begin{variable}{\c_fp_pi_by_four_decimal_int}
+% \begin{variable}{\c_fp_pi_by_four_extended_int}
+% \begin{variable}{\c_fp_pi_decimal_int}
+% \begin{variable}{\c_fp_pi_extended_int}
+% \begin{variable}{\c_fp_two_pi_decimal_int}
+% \begin{variable}{\c_fp_two_pi_extended_int}
+%   Parts of $\pi$ for trigonometric range reduction, implemented
+%   as \texttt{int} variables for speed.
+%    \begin{macrocode}
+\int_new:N  \c_fp_pi_by_four_decimal_int
+\int_set:Nn \c_fp_pi_by_four_decimal_int { 785 398 158 }
+\int_new:N  \c_fp_pi_by_four_extended_int
+\int_set:Nn \c_fp_pi_by_four_extended_int { 897 448 310 }
+\int_new:N  \c_fp_pi_decimal_int
+\int_set:Nn \c_fp_pi_decimal_int { 141 592 653 }
+\int_new:N  \c_fp_pi_extended_int
+\int_set:Nn \c_fp_pi_extended_int { 589 793 238 }
+\int_new:N  \c_fp_two_pi_decimal_int
+\int_set:Nn \c_fp_two_pi_decimal_int { 283 185 307 }
+\int_new:N  \c_fp_two_pi_extended_int
+\int_set:Nn \c_fp_two_pi_extended_int { 179 586 477 }
+%    \end{macrocode}
+% \end{variable}
+% \end{variable}
+% \end{variable}
+% \end{variable}
+% \end{variable}
+% \end{variable}
+%
+% \begin{variable}{\c_e_fp}
+%   The value $\mathrm{e}$ as a \enquote{machine number}.
+%    \begin{macrocode}
+\tl_const:Nn \c_e_fp { + 2.718281828 e 0 }
+%    \end{macrocode}
+% \end{variable}
+%
+% \begin{variable}{\c_one_fp}
+%   The constant value $1$: used for fast comparisons.
+%    \begin{macrocode}
+\tl_const:Nn \c_one_fp { + 1.000000000 e 0 }
+%    \end{macrocode}
+% \end{variable}
+%
+% \begin{variable}{\c_pi_fp}
+%   The value $\pi$ as a \enquote{machine number}.
+%    \begin{macrocode}
+\tl_const:Nn \c_pi_fp { + 3.141592654 e 0 }
+%    \end{macrocode}
+% \end{variable}
+%
+% \begin{variable}{\c_undefined_fp}
+%   A marker for undefined values.
+%    \begin{macrocode}
+\tl_const:Nn \c_undefined_fp { X 0.000000000 e 0 }
+%    \end{macrocode}
+% \end{variable}
+%
+% \begin{variable}{\c_zero_fp}
+%   The constant zero value.
+%    \begin{macrocode}
+\tl_const:Nn \c_zero_fp { + 0.000000000 e 0 }
+%    \end{macrocode}
+% \end{variable}
+%
+% \subsection{Variables}
+%
+% \begin{variable}{\l_fp_arg_tl}
+%   A token list to store the formalised representation of the input
+%   for transcendental functions.
+%    \begin{macrocode}
+\tl_new:N \l_fp_arg_tl
+%    \end{macrocode}
+% \end{variable}
+%
+% \begin{variable}{\l_fp_count_int}
+%   A counter for things like the number of divisions possible.
+%    \begin{macrocode}
+\int_new:N \l_fp_count_int
+%    \end{macrocode}
+% \end{variable}
+%
+% \begin{variable}{\l_fp_div_offset_int}
+%   When carrying out division, an offset is used for the results to
+%   get the decimal part correct.
+%    \begin{macrocode}
+\int_new:N \l_fp_div_offset_int
+%    \end{macrocode}
+% \end{variable}
+%
+% \begin{variable}{\l_fp_exp_integer_int}
+% \begin{variable}{\l_fp_exp_decimal_int}
+% \begin{variable}{\l_fp_exp_extended_int}
+% \begin{variable}{\l_fp_exp_exponent_int}
+%   Used for the calculation of exponent values.
+%    \begin{macrocode}
+\int_new:N \l_fp_exp_integer_int
+\int_new:N \l_fp_exp_decimal_int
+\int_new:N \l_fp_exp_extended_int
+\int_new:N \l_fp_exp_exponent_int
+%    \end{macrocode}
+% \end{variable}
+% \end{variable}
+% \end{variable}
+% \end{variable}
+%
+% \begin{variable}{\l_fp_input_a_sign_int}
+% \begin{variable}{\l_fp_input_a_integer_int}
+% \begin{variable}{\l_fp_input_a_decimal_int}
+% \begin{variable}{\l_fp_input_a_exponent_int}
+% \begin{variable}{\l_fp_input_b_sign_int}
+% \begin{variable}{\l_fp_input_b_integer_int}
+% \begin{variable}{\l_fp_input_b_decimal_int}
+% \begin{variable}{\l_fp_input_b_exponent_int}
+%   Storage for the input: two storage areas as there are at most two
+%   inputs.
+%    \begin{macrocode}
+\int_new:N \l_fp_input_a_sign_int
+\int_new:N \l_fp_input_a_integer_int
+\int_new:N \l_fp_input_a_decimal_int
+\int_new:N \l_fp_input_a_exponent_int
+\int_new:N \l_fp_input_b_sign_int
+\int_new:N \l_fp_input_b_integer_int
+\int_new:N \l_fp_input_b_decimal_int
+\int_new:N \l_fp_input_b_exponent_int
+%    \end{macrocode}
+% \end{variable}
+% \end{variable}
+% \end{variable}
+% \end{variable}
+% \end{variable}
+% \end{variable}
+% \end{variable}
+% \end{variable}
+%
+% \begin{variable}{\l_fp_input_a_extended_int}
+% \begin{variable}{\l_fp_input_b_extended_int}
+%   For internal use, \enquote{extended} floating point numbers are
+%   needed.
+%    \begin{macrocode}
+\int_new:N \l_fp_input_a_extended_int
+\int_new:N \l_fp_input_b_extended_int
+%    \end{macrocode}
+% \end{variable}
+% \end{variable}
+%
+% \begin{variable}{\l_fp_mul_a_i_int}
+% \begin{variable}{\l_fp_mul_a_ii_int}
+% \begin{variable}{\l_fp_mul_a_iii_int}
+% \begin{variable}{\l_fp_mul_a_iv_int}
+% \begin{variable}{\l_fp_mul_a_v_int}
+% \begin{variable}{\l_fp_mul_a_vi_int}
+% \begin{variable}{\l_fp_mul_b_i_int}
+% \begin{variable}{\l_fp_mul_b_ii_int}
+% \begin{variable}{\l_fp_mul_b_iii_int}
+% \begin{variable}{\l_fp_mul_b_iv_int}
+% \begin{variable}{\l_fp_mul_b_v_int}
+% \begin{variable}{\l_fp_mul_b_vi_int}
+%   Multiplication requires that the decimal part is split into parts
+%   so that there are no overflows.
+%    \begin{macrocode}
+\int_new:N \l_fp_mul_a_i_int
+\int_new:N \l_fp_mul_a_ii_int
+\int_new:N \l_fp_mul_a_iii_int
+\int_new:N \l_fp_mul_a_iv_int
+\int_new:N \l_fp_mul_a_v_int
+\int_new:N \l_fp_mul_a_vi_int
+\int_new:N \l_fp_mul_b_i_int
+\int_new:N \l_fp_mul_b_ii_int
+\int_new:N \l_fp_mul_b_iii_int
+\int_new:N \l_fp_mul_b_iv_int
+\int_new:N \l_fp_mul_b_v_int
+\int_new:N \l_fp_mul_b_vi_int
+%    \end{macrocode}
+% \end{variable}
+% \end{variable}
+% \end{variable}
+% \end{variable}
+% \end{variable}
+% \end{variable}
+% \end{variable}
+% \end{variable}
+% \end{variable}
+% \end{variable}
+% \end{variable}
+% \end{variable}
+%
+% \begin{variable}{\l_fp_mul_output_int}
+% \begin{variable}{\l_fp_mul_output_tl}
+%   Space for multiplication results.
+%    \begin{macrocode}
+\int_new:N \l_fp_mul_output_int
+\tl_new:N  \l_fp_mul_output_tl
+%    \end{macrocode}
+% \end{variable}
+% \end{variable}
+%
+% \begin{variable}{\l_fp_output_sign_int}
+% \begin{variable}{\l_fp_output_integer_int}
+% \begin{variable}{\l_fp_output_decimal_int}
+% \begin{variable}{\l_fp_output_exponent_int}
+%   Output is stored in the same way as input.
+%    \begin{macrocode}
+\int_new:N \l_fp_output_sign_int
+\int_new:N \l_fp_output_integer_int
+\int_new:N \l_fp_output_decimal_int
+\int_new:N \l_fp_output_exponent_int
+%    \end{macrocode}
+% \end{variable}
+% \end{variable}
+% \end{variable}
+% \end{variable}
+%
+% \begin{variable}{\l_fp_output_extended_int}
+%   Again, for calculations an extended part.
+%    \begin{macrocode}
+\int_new:N \l_fp_output_extended_int
+%    \end{macrocode}
+% \end{variable}
+%
+% \begin{variable}{\l_fp_round_carry_bool}
+%   To indicate that a digit needs to be carried forward.
+%    \begin{macrocode}
+\bool_new:N \l_fp_round_carry_bool
+%    \end{macrocode}
+% \end{variable}
+%
+% \begin{variable}{\l_fp_round_decimal_tl}
+%   A temporary store when rounding, to build up the decimal part without
+%   needing to do any maths.
+%    \begin{macrocode}
+\tl_new:N \l_fp_round_decimal_tl
+%    \end{macrocode}
+% \end{variable}
+%
+% \begin{variable}{\l_fp_round_position_int}
+% \begin{variable}{\l_fp_round_target_int}
+%   Used to check the position for rounding.
+%    \begin{macrocode}
+\int_new:N \l_fp_round_position_int
+\int_new:N \l_fp_round_target_int
+%    \end{macrocode}
+% \end{variable}
+% \end{variable}
+%
+% \begin{variable}{\l_fp_sign_tl}
+%   There are places where the sign needs to be set up \enquote{early},
+%   so that the registers can be re-used.
+%    \begin{macrocode}
+\tl_new:N \l_fp_sign_tl
+%    \end{macrocode}
+% \end{variable}
+%
+% \begin{variable}{\l_fp_split_sign_int}
+%   When splitting the input it is fastest to use a fixed name for the
+%   sign part, and to transfer it after the split is complete.
+%    \begin{macrocode}
+\int_new:N \l_fp_split_sign_int
+%    \end{macrocode}
+% \end{variable}
+%
+% \begin{variable}{\l_fp_tmp_int}
+%   A scratch \texttt{int}: used only where the value is not carried
+%   forward.
+%    \begin{macrocode}
+\int_new:N \l_fp_tmp_int
+%    \end{macrocode}
+% \end{variable}
+%
+% \begin{variable}{\l_fp_tmp_tl}
+%   A scratch token list variable for expanding material.
+%    \begin{macrocode}
+\tl_new:N \l_fp_tmp_tl
+%    \end{macrocode}
+% \end{variable}
+%
+% \begin{variable}{\l_fp_trig_octant_int}
+%   To track which octant the trigonometric input is in.
+%    \begin{macrocode}
+\int_new:N \l_fp_trig_octant_int
+%    \end{macrocode}
+% \end{variable}
+%
+% \begin{variable}{\l_fp_trig_sign_int}
+% \begin{variable}{\l_fp_trig_decimal_int}
+% \begin{variable}{\l_fp_trig_extended_int}
+%   Used for the calculation of trigonometric values.
+%    \begin{macrocode}
+\int_new:N \l_fp_trig_sign_int
+\int_new:N \l_fp_trig_decimal_int
+\int_new:N \l_fp_trig_extended_int
+%    \end{macrocode}
+% \end{variable}
+% \end{variable}
+% \end{variable}
+%
+% \subsection{Parsing numbers}
+%
+% \begin{macro}{\fp_read:N}
+% \begin{macro}[aux]{\fp_read_aux:w}
+%   Reading a stored value is made easier as the format is designed to
+%   match the delimited function. This is always used to read the first
+%   value (register |a|).
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \fp_read:N #1
+  { \exp_after:wN \fp_read_aux:w #1 \q_stop }
+\cs_new_protected_nopar:Npn \fp_read_aux:w #1#2 . #3 e #4 \q_stop
+  {
+    \if:w #1 -
+      \l_fp_input_a_sign_int \c_minus_one
+    \else:
+      \l_fp_input_a_sign_int \c_one
+    \fi:
+    \l_fp_input_a_integer_int  #2 \scan_stop:
+    \l_fp_input_a_decimal_int  #3 \scan_stop:
+    \l_fp_input_a_exponent_int #4 \scan_stop:
+  }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\fp_split:Nn}
+% \begin{macro}[aux]{\fp_split_sign:}
+% \begin{macro}[aux]{\fp_split_exponent:}
+% \begin{macro}[aux]{\fp_split_aux_i:w}
+% \begin{macro}[aux]{\fp_split_aux_ii:w}
+% \begin{macro}[aux]{\fp_split_aux_iii:w}
+% \begin{macro}[aux]{\fp_split_decimal:w}
+% \begin{macro}[aux]{\fp_split_decimal_aux:w}
+%   The aim here is to use as much of \TeX{}'s mechanism as possible to pick
+%   up the numerical input without any mistakes. In particular, negative
+%   numbers have to be filtered out first in case the integer part is
+%   $0$ (in which case \TeX{} would drop the |-| sign). That process
+%   has to be done in a loop for cases where the sign is repeated.
+%   Finding an exponent is relatively easy, after which the next phase is
+%   to find the integer part, which will terminate with a |.|, and trigger
+%   the decimal-finding code. The later will allow the decimal to be too
+%   long, truncating the result.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \fp_split:Nn #1#2
+  {
+    \tl_set:Nx \l_fp_tmp_tl {#2}
+    \tl_set_rescan:Nno \l_fp_tmp_tl { \char_set_catcode_ignore:n { 32 } }
+      { \l_fp_tmp_tl }
+    \l_fp_split_sign_int \c_one
+    \fp_split_sign:
+    \use:c { l_fp_input_ #1 _sign_int } \l_fp_split_sign_int
+    \exp_after:wN \fp_split_exponent:w \l_fp_tmp_tl e e \q_stop #1
+  }
+\cs_new_protected_nopar:Npn \fp_split_sign:
+  {
+    \if_int_compare:w \pdftex_strcmp:D
+      { \exp_after:wN \tl_head:w \l_fp_tmp_tl ? \q_stop } { - }
+        = \c_zero
+      \tl_set:Nx \l_fp_tmp_tl
+        {
+          \exp_after:wN
+            \tl_tail:w \l_fp_tmp_tl \prg_do_nothing: \q_stop
+        }
+      \l_fp_split_sign_int -\l_fp_split_sign_int
+      \exp_after:wN \fp_split_sign:
+    \else:
+      \if_int_compare:w \pdftex_strcmp:D
+        { \exp_after:wN \tl_head:w \l_fp_tmp_tl ? \q_stop } { + }
+          = \c_zero
+        \tl_set:Nx \l_fp_tmp_tl
+          {
+            \exp_after:wN
+              \tl_tail:w \l_fp_tmp_tl \prg_do_nothing: \q_stop
+          }
+        \exp_after:wN \exp_after:wN \exp_after:wN \fp_split_sign:
+       \fi:
+    \fi:
+  }
+\cs_new_protected_nopar:Npn \fp_split_exponent:w #1 e #2 e #3 \q_stop #4
+  {
+    \use:c { l_fp_input_ #4 _exponent_int }
+      \int_eval:w 0 #2 \scan_stop:
+    \tex_afterassignment:D \fp_split_aux_i:w
+    \use:c { l_fp_input_ #4 _integer_int }
+      \int_eval:w 0 #1 . . \q_stop #4
+  }
+\cs_new_protected_nopar:Npn \fp_split_aux_i:w #1 . #2 . #3 \q_stop
+  { \fp_split_aux_ii:w #2 000000000 \q_stop }
+\cs_new_protected_nopar:Npn \fp_split_aux_ii:w #1#2#3#4#5#6#7#8#9
+  { \fp_split_aux_iii:w {#1#2#3#4#5#6#7#8#9} }
+\cs_new_protected_nopar:Npn \fp_split_aux_iii:w #1#2 \q_stop
+  {
+    \l_fp_tmp_int 1 #1 \scan_stop:
+    \exp_after:wN \fp_split_decimal:w
+      \int_use:N \l_fp_tmp_int 000000000 \q_stop
+  }
+\cs_new_protected_nopar:Npn \fp_split_decimal:w #1#2#3#4#5#6#7#8#9
+  { \fp_split_decimal_aux:w {#2#3#4#5#6#7#8#9} }
+\cs_new_protected_nopar:Npn \fp_split_decimal_aux:w #1#2#3 \q_stop #4
+  {
+    \use:c { l_fp_input_ #4 _decimal_int } #1#2 \scan_stop:
+    \if_int_compare:w
+      \int_eval:w
+        \use:c { l_fp_input_ #4 _integer_int } +
+        \use:c { l_fp_input_ #4 _decimal_int }
+      \scan_stop:
+        = \c_zero
+      \use:c { l_fp_input_ #4 _sign_int } \c_one
+    \fi:
+    \if_int_compare:w
+      \use:c { l_fp_input_ #4 _integer_int } < \c_one_thousand_million
+    \else:
+      \exp_after:wN \fp_overflow_msg:
+    \fi:
+  }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\fp_standardise:NNNN}
+% \begin{macro}[aux]{\fp_standardise_aux:NNNN}
+% \begin{macro}[aux]{\fp_standardise_aux:}
+% \begin{macro}[aux]{\fp_standardise_aux:w}
+%   The idea here is to shift the input into a known exponent range. This
+%   is done using \TeX{} tokens where possible, as this is faster than
+%   arithmetic.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \fp_standardise:NNNN #1#2#3#4
+  {
+    \if_int_compare:w
+      \int_eval:w #2 + #3 = \c_zero
+      #1 \c_one
+      #4 \c_zero
+      \exp_after:wN \use_none:nnnn
+    \else:
+      \exp_after:wN \fp_standardise_aux:NNNN
+    \fi:
+    #1#2#3#4
+  }
+\cs_new_protected_nopar:Npn \fp_standardise_aux:NNNN #1#2#3#4
+  {
+    \cs_set_protected_nopar:Npn \fp_standardise_aux:
+      {
+        \if_int_compare:w #2 = \c_zero
+          \tex_advance:D #3 \c_one_thousand_million
+          \exp_after:wN \fp_standardise_aux:w
+            \int_use:N #3 \q_stop
+           \exp_after:wN \fp_standardise_aux:
+         \fi:
+      }
+    \cs_set_protected_nopar:Npn
+      \fp_standardise_aux:w ##1##2##3##4##5##6##7##8##9 \q_stop
+      {
+        #2 ##2 \scan_stop:
+        #3 ##3##4##5##6##7##8##9 0 \scan_stop:
+        \tex_advance:D #4 \c_minus_one
+      }
+    \fp_standardise_aux:
+    \cs_set_protected_nopar:Npn \fp_standardise_aux:
+      {
+        \if_int_compare:w #2 > \c_nine
+          \tex_advance:D #2 \c_one_thousand_million
+          \exp_after:wN \use_i:nn \exp_after:wN
+            \fp_standardise_aux:w \int_use:N #2
+           \exp_after:wN \fp_standardise_aux:
+         \fi:
+      }
+    \cs_set_protected_nopar:Npn
+      \fp_standardise_aux:w ##1##2##3##4##5##6##7##8##9
+      {
+        #2 ##1##2##3##4##5##6##7##8 \scan_stop:
+        \tex_advance:D #3 \c_one_thousand_million
+        \tex_divide:D #3 \c_ten
+        \tl_set:Nx \l_fp_tmp_tl
+          {
+            ##9
+            \exp_after:wN \use_none:n \int_use:N #3
+          }
+        #3 \l_fp_tmp_tl \scan_stop:
+        \tex_advance:D #4 \c_one
+      }
+    \fp_standardise_aux:
+    \if_int_compare:w #4 < \c_one_hundred
+      \if_int_compare:w #4 > -\c_one_hundred
+      \else:
+        #1 \c_one
+        #2 \c_zero
+        #3 \c_zero
+        #4 \c_zero
+      \fi:
+    \else:
+      \exp_after:wN \fp_overflow_msg:
+    \fi:
+  }
+\cs_new_protected_nopar:Npn \fp_standardise_aux: { }
+\cs_new_protected_nopar:Npn \fp_standardise_aux:w { }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \subsection{Internal utilities}
+%
+% \begin{macro}{\fp_level_input_exponents:}
+% \begin{macro}[aux]{\fp_level_input_exponents_a:}
+% \begin{macro}[aux]{\fp_level_input_exponents_a:NNNNNNNNN}
+% \begin{macro}[aux]{\fp_level_input_exponents_b:}
+% \begin{macro}[aux]{\fp_level_input_exponents_b:NNNNNNNNN}
+%   The routines here are similar to those used to standardise the
+%   exponent. However, the aim here is different: the two exponents need
+%   to end up the same.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \fp_level_input_exponents:
+  {
+    \if_int_compare:w \l_fp_input_a_exponent_int > \l_fp_input_b_exponent_int
+      \exp_after:wN \fp_level_input_exponents_a:
+    \else:
+      \exp_after:wN \fp_level_input_exponents_b:
+    \fi:
+  }
+\cs_new_protected_nopar:Npn \fp_level_input_exponents_a:
+  {
+    \if_int_compare:w \l_fp_input_a_exponent_int > \l_fp_input_b_exponent_int
+      \tex_advance:D \l_fp_input_b_integer_int \c_one_thousand_million
+      \exp_after:wN \use_i:nn \exp_after:wN
+        \fp_level_input_exponents_a:NNNNNNNNN
+          \int_use:N \l_fp_input_b_integer_int
+      \exp_after:wN \fp_level_input_exponents_a:
+    \fi:
+  }
+\cs_new_protected_nopar:Npn \fp_level_input_exponents_a:NNNNNNNNN
+  #1#2#3#4#5#6#7#8#9
+  {
+    \l_fp_input_b_integer_int #1#2#3#4#5#6#7#8 \scan_stop:
+    \tex_advance:D \l_fp_input_b_decimal_int \c_one_thousand_million
+    \tex_divide:D \l_fp_input_b_decimal_int \c_ten
+    \tl_set:Nx \l_fp_tmp_tl
+      {
+        #9
+        \exp_after:wN \use_none:n
+          \int_use:N \l_fp_input_b_decimal_int
+      }
+    \l_fp_input_b_decimal_int \l_fp_tmp_tl \scan_stop:
+    \tex_advance:D \l_fp_input_b_exponent_int \c_one
+  }
+\cs_new_protected_nopar:Npn \fp_level_input_exponents_b:
+  {
+    \if_int_compare:w \l_fp_input_b_exponent_int > \l_fp_input_a_exponent_int
+      \tex_advance:D \l_fp_input_a_integer_int \c_one_thousand_million
+      \exp_after:wN \use_i:nn \exp_after:wN
+        \fp_level_input_exponents_b:NNNNNNNNN
+          \int_use:N \l_fp_input_a_integer_int
+      \exp_after:wN \fp_level_input_exponents_b:
+    \fi:
+  }
+\cs_new_protected_nopar:Npn \fp_level_input_exponents_b:NNNNNNNNN
+  #1#2#3#4#5#6#7#8#9
+  {
+    \l_fp_input_a_integer_int #1#2#3#4#5#6#7#8 \scan_stop:
+    \tex_advance:D \l_fp_input_a_decimal_int \c_one_thousand_million
+    \tex_divide:D \l_fp_input_a_decimal_int \c_ten
+    \tl_set:Nx \l_fp_tmp_tl
+      {
+        #9
+        \exp_after:wN \use_none:n
+          \int_use:N \l_fp_input_a_decimal_int
+      }
+    \l_fp_input_a_decimal_int \l_fp_tmp_tl \scan_stop:
+    \tex_advance:D \l_fp_input_a_exponent_int \c_one
+  }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}[aux]{\fp_tmp:w}
+%   Used for output of results, cutting down on \cs{exp_after:wN}.
+%   This is just a place holder definition.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \fp_tmp:w #1#2 { }
+%    \end{macrocode}
+% \end{macro}
+%
+% \subsection{Operations for \texttt{fp} variables}
+%
+% The format of \texttt{fp} variables is tightly defined, so that
+% they can be read quickly by the internal code. The format is a single
+% sign token, a single number, the decimal point, nine decimal numbers,
+% an |e| and finally the exponent. This final part may vary in length.
+% When stored, floating points will always be stored with a value in
+% the integer position unless the number is zero.
+%
+% \begin{macro}{\fp_new:N, \fp_new:c}
+% \UnitTested
+%   Fixed-points always have a value, and of course this has to be
+%   initialised globally.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \fp_new:N #1
+  {
+    \tl_new:N #1
+    \tl_gset_eq:NN #1 \c_zero_fp
+  }
+\cs_generate_variant:Nn \fp_new:N { c }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\fp_const:Nn, \fp_const:cn}
+%   A simple wrapper.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \fp_const:Nn #1#2
+  {
+    \fp_new:N #1
+    \fp_gset:Nn #1 {#2}
+  }
+\cs_generate_variant:Nn \fp_const:Nn { c }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\fp_zero:N, \fp_zero:c}
+% \UnitTested
+% \begin{macro}{\fp_gzero:N, \fp_gzero:c}
+% \UnitTested
+%   Zeroing fixed-points is pretty obvious.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \fp_zero:N #1
+  { \tl_set_eq:NN #1 \c_zero_fp }
+\cs_new_protected_nopar:Npn \fp_gzero:N #1
+  { \tl_gset_eq:NN #1 \c_zero_fp }
+\cs_generate_variant:Nn \fp_zero:N { c }
+\cs_generate_variant:Nn \fp_gzero:N { c }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\fp_set:Nn, \fp_set:cn}
+% \UnitTested
+% \begin{macro}{\fp_gset:Nn, \fp_gset:cn}
+% \UnitTested
+% \begin{macro}[aux]{\fp_set_aux:NNn}
+%   To trap any input errors, a very simple version of the parser is run
+%   here. This will pick up any invalid characters at this stage, saving
+%   issues later. The splitting approach is the same as the more
+%   advanced function later.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \fp_set:Nn  { \fp_set_aux:NNn \tl_set:Nn }
+\cs_new_protected_nopar:Npn \fp_gset:Nn { \fp_set_aux:NNn \tl_gset:Nn }
+\cs_new_protected_nopar:Npn \fp_set_aux:NNn #1#2#3
+  {
+    \group_begin:
+      \fp_split:Nn a {#3}
+      \fp_standardise:NNNN
+        \l_fp_input_a_sign_int
+        \l_fp_input_a_integer_int
+        \l_fp_input_a_decimal_int
+        \l_fp_input_a_exponent_int
+      \tex_advance:D \l_fp_input_a_decimal_int \c_one_thousand_million
+      \cs_set_protected_nopar:Npx \fp_tmp:w
+        {
+          \group_end:
+          #1 \exp_not:N #2
+            {
+              \if_int_compare:w \l_fp_input_a_sign_int < \c_zero
+                -
+              \else:
+                +
+              \fi:
+              \int_use:N \l_fp_input_a_integer_int
+              .
+              \exp_after:wN \use_none:n
+                \int_use:N \l_fp_input_a_decimal_int
+              e
+              \int_use:N \l_fp_input_a_exponent_int
+            }
+        }
+    \fp_tmp:w
+  }
+\cs_generate_variant:Nn \fp_set:Nn  { c }
+\cs_generate_variant:Nn \fp_gset:Nn { c }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+%
+%
+% \begin{macro}{\fp_set_from_dim:Nn,  \fp_set_from_dim:cn}
+% \UnitTested
+% \begin{macro}{\fp_gset_from_dim:Nn, \fp_gset_from_dim:cn}
+% \UnitTested
+% \begin{macro}[aux]{\fp_set_from_dim_aux:NNn}
+% \begin{macro}[aux]{\fp_set_from_dim_aux:w}
+% \begin{variable}{\l_fp_tmp_dim}
+% \begin{variable}{\l_fp_tmp_skip}
+%   Here, dimensions are converted to fixed-points \emph{via} a
+%   temporary variable. This ensures that they always convert as points.
+%   The code is then essentially the same as for \cs{fp_set:Nn}, but with
+%   the dimension passed so that it will be striped of the |pt| on the
+%   way through. The passage through a skip is used to remove any rubber
+%   part.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \fp_set_from_dim:Nn
+  { \fp_set_from_dim_aux:NNn \tl_set:Nx }
+\cs_new_protected_nopar:Npn \fp_gset_from_dim:Nn
+  { \fp_set_from_dim_aux:NNn \tl_gset:Nx }
+\cs_new_protected_nopar:Npn \fp_set_from_dim_aux:NNn #1#2#3
+  {
+    \group_begin:
+      \l_fp_tmp_skip \etex_glueexpr:D #3 \scan_stop:
+      \l_fp_tmp_dim \l_fp_tmp_skip
+      \fp_split:Nn a
+        {
+          \exp_after:wN \fp_set_from_dim_aux:w
+            \dim_use:N \l_fp_tmp_dim
+        }
+      \fp_standardise:NNNN
+        \l_fp_input_a_sign_int
+        \l_fp_input_a_integer_int
+        \l_fp_input_a_decimal_int
+        \l_fp_input_a_exponent_int
+      \tex_advance:D \l_fp_input_a_decimal_int \c_one_thousand_million
+      \cs_set_protected_nopar:Npx \fp_tmp:w
+        {
+          \group_end:
+          #1 \exp_not:N #2
+            {
+              \if_int_compare:w \l_fp_input_a_sign_int < \c_zero
+                -
+              \else:
+                +
+              \fi:
+              \int_use:N \l_fp_input_a_integer_int
+              .
+              \exp_after:wN \use_none:n
+                \int_use:N \l_fp_input_a_decimal_int
+              e
+              \int_use:N \l_fp_input_a_exponent_int
+            }
+        }
+    \fp_tmp:w
+  }
+\cs_set_protected_nopar:Npx \fp_set_from_dim_aux:w
+  {
+    \cs_set_nopar:Npn \exp_not:N \fp_set_from_dim_aux:w
+      ##1 \tl_to_str:n { pt } {##1}
+  }
+\fp_set_from_dim_aux:w
+\cs_generate_variant:Nn \fp_set_from_dim:Nn  { c }
+\cs_generate_variant:Nn \fp_gset_from_dim:Nn { c }
+\dim_new:N \l_fp_tmp_dim
+\skip_new:N \l_fp_tmp_skip
+%    \end{macrocode}
+% \end{variable}
+% \end{variable}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\fp_set_eq:NN, \fp_set_eq:cN, \fp_set_eq:Nc, \fp_set_eq:cc}
+% \UnitTested
+% \begin{macro}{\fp_gset_eq:NN, \fp_gset_eq:cN, \fp_gset_eq:Nc, \fp_gset_eq:cc}
+% \UnitTested
+%   Pretty simple, really.
+%    \begin{macrocode}
+\cs_new_eq:NN \fp_set_eq:NN  \tl_set_eq:NN
+\cs_new_eq:NN \fp_set_eq:cN  \tl_set_eq:cN
+\cs_new_eq:NN \fp_set_eq:Nc  \tl_set_eq:Nc
+\cs_new_eq:NN \fp_set_eq:cc  \tl_set_eq:cc
+\cs_new_eq:NN \fp_gset_eq:NN \tl_gset_eq:NN
+\cs_new_eq:NN \fp_gset_eq:cN \tl_gset_eq:cN
+\cs_new_eq:NN \fp_gset_eq:Nc \tl_gset_eq:Nc
+\cs_new_eq:NN \fp_gset_eq:cc \tl_gset_eq:cc
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\fp_show:N, \fp_show:c}
+% \UnitTested
+%   Simple showing of the underlying variable.
+%    \begin{macrocode}
+\cs_new_eq:NN \fp_show:N \tl_show:N
+\cs_new_eq:NN \fp_show:c \tl_show:c
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\fp_use:N, \fp_use:c}
+% \UnitTested
+% \begin{macro}[aux]{\fp_use_aux:w}
+% \begin{macro}[aux]{\fp_use_none:w}
+% \begin{macro}[aux]{\fp_use_small:w}
+% \begin{macro}[aux]{\fp_use_large:w}
+% \begin{macro}[aux]{\fp_use_large_aux_i:w}
+% \begin{macro}[aux]{\fp_use_large_aux_1:w}
+% \begin{macro}[aux]{\fp_use_large_aux_2:w}
+% \begin{macro}[aux]{\fp_use_large_aux_3:w}
+% \begin{macro}[aux]{\fp_use_large_aux_4:w}
+% \begin{macro}[aux]{\fp_use_large_aux_5:w}
+% \begin{macro}[aux]{\fp_use_large_aux_6:w}
+% \begin{macro}[aux]{\fp_use_large_aux_7:w}
+% \begin{macro}[aux]{\fp_use_large_aux_8:w}
+% \begin{macro}[aux]{\fp_use_large_aux_i:w}
+% \begin{macro}[aux]{\fp_use_large_aux_ii:w}
+%   The idea of the \cs{fp_use:N} function to convert the stored
+%   value into something suitable for \TeX{} to use as a number in an
+%   expandable manner. The first step is to deal with the sign, then
+%   work out how big the input is.
+%    \begin{macrocode}
+\cs_new_nopar:Npn \fp_use:N #1
+  { \exp_after:wN \fp_use_aux:w #1 \q_stop }
+\cs_generate_variant:Nn \fp_use:N { c }
+\cs_new_nopar:Npn \fp_use_aux:w #1#2 e #3 \q_stop
+  {
+    \if:w #1 -
+      -
+    \fi:
+    \if_int_compare:w #3 > \c_zero
+      \exp_after:wN \fp_use_large:w
+    \else:
+      \if_int_compare:w #3 < \c_zero
+        \exp_after:wN \exp_after:wN \exp_after:wN
+          \fp_use_small:w
+      \else:
+        \exp_after:wN \exp_after:wN \exp_after:wN \fp_use_none:w
+      \fi:
+    \fi:
+    #2 e #3 \q_stop
+  }
+%    \end{macrocode}
+%   When the exponent is zero, the input is simply returned as output.
+%    \begin{macrocode}
+\cs_new_nopar:Npn \fp_use_none:w #1 e #2 \q_stop {#1}
+%    \end{macrocode}
+%   For small numbers (less than $1$) the correct number of zeros
+%   have to be inserted, but the decimal point is easy.
+%    \begin{macrocode}
+\cs_new_nopar:Npn \fp_use_small:w #1 . #2 e #3 \q_stop
+  {
+    0 .
+    \prg_replicate:nn { -#3 - 1 } { 0 }
+    #1#2
+  }
+%    \end{macrocode}
+%   Life is more complex for large numbers. The decimal point needs to
+%   be shuffled, with potentially some zero-filling for very large values.
+%    \begin{macrocode}
+\cs_new_nopar:Npn \fp_use_large:w #1 . #2 e #3 \q_stop
+  {
+    \if_int_compare:w #3 < \c_ten
+      \exp_after:wN \fp_use_large_aux_i:w
+    \else:
+      \exp_after:wN \fp_use_large_aux_ii:w
+    \fi:
+    #1#2 e #3 \q_stop
+  }
+\cs_new_nopar:Npn \fp_use_large_aux_i:w #1#2 e #3 \q_stop
+  {
+    #1
+    \use:c { fp_use_large_aux_ #3 :w } #2 \q_stop
+  }
+\cs_new_nopar:cpn { fp_use_large_aux_1:w } #1#2 \q_stop { #1 . #2 }
+\cs_new_nopar:cpn { fp_use_large_aux_2:w } #1#2#3 \q_stop
+  { #1#2 . #3 }
+\cs_new_nopar:cpn { fp_use_large_aux_3:w } #1#2#3#4 \q_stop
+  { #1#2#3 . #4 }
+\cs_new_nopar:cpn { fp_use_large_aux_4:w } #1#2#3#4#5 \q_stop
+  { #1#2#3#4 . #5 }
+\cs_new_nopar:cpn { fp_use_large_aux_5:w } #1#2#3#4#5#6 \q_stop
+  { #1#2#3#4#5 . #6 }
+\cs_new_nopar:cpn { fp_use_large_aux_6:w } #1#2#3#4#5#6#7 \q_stop
+  { #1#2#3#4#5#6 . #7 }
+\cs_new_nopar:cpn { fp_use_large_aux_7:w } #1#2#3#4#5#6#7#8 \q_stop
+  { #1#2#3#4#6#7 . #8 }
+\cs_new_nopar:cpn { fp_use_large_aux_8:w } #1#2#3#4#5#6#7#8#9 \q_stop
+  { #1#2#3#4#5#6#7#8 . #9 }
+\cs_new_nopar:cpn { fp_use_large_aux_9:w } #1 \q_stop { #1 . }
+\cs_new_nopar:Npn \fp_use_large_aux_ii:w #1 e #2 \q_stop
+  {
+    #1
+    \prg_replicate:nn { #2 - 9 } { 0 }
+    .
+  }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \subsection{Transferring to other types}
+%
+% The \cs{fp_use:N} function converts a floating point variable to
+% a form that can be used by \TeX{}. Here, the functions are slightly
+% different, as some information may be discarded.
+%
+% \begin{macro}{\fp_to_dim:N, \fp_to_dim:c}
+%   A very simple wrapper.
+%    \begin{macrocode}
+\cs_new_nopar:Npn \fp_to_dim:N #1 { \fp_use:N #1 pt }
+\cs_generate_variant:Nn \fp_to_dim:N { c }
+%    \end{macrocode}
+% \end{macro}
+%
+%
+% \begin{macro}{\fp_to_int:N, \fp_to_int:c}
+% \UnitTested
+% \begin{macro}[aux]{\fp_to_int_aux:w}
+% \begin{macro}[aux]{\fp_to_int_none:w}
+% \begin{macro}[aux]{\fp_to_int_small:w}
+% \begin{macro}[aux]{\fp_to_int_large:w}
+% \begin{macro}[aux]{\fp_to_int_large_aux_i:w}
+% \begin{macro}[aux]{\fp_to_int_large_aux_1:w}
+% \begin{macro}[aux]{\fp_to_int_large_aux_2:w}
+% \begin{macro}[aux]{\fp_to_int_large_aux_3:w}
+% \begin{macro}[aux]{\fp_to_int_large_aux_4:w}
+% \begin{macro}[aux]{\fp_to_int_large_aux_5:w}
+% \begin{macro}[aux]{\fp_to_int_large_aux_6:w}
+% \begin{macro}[aux]{\fp_to_int_large_aux_7:w}
+% \begin{macro}[aux]{\fp_to_int_large_aux_8:w}
+% \begin{macro}[aux]{\fp_to_int_large_aux_i:w}
+% \begin{macro}[aux]{\fp_to_int_large_aux:nnn}
+% \begin{macro}[aux]{\fp_to_int_large_aux_ii:w}
+%   Converting to integers in an expandable manner is very similar to
+%   simply using floating point variables, particularly in the lead-off.
+%    \begin{macrocode}
+\cs_new_nopar:Npn \fp_to_int:N #1
+  { \exp_after:wN \fp_to_int_aux:w #1 \q_stop }
+\cs_generate_variant:Nn \fp_to_int:N { c }
+\cs_new_nopar:Npn \fp_to_int_aux:w #1#2 e #3 \q_stop
+  {
+    \if:w #1 -
+      -
+    \fi:
+    \if_int_compare:w #3 < \c_zero
+      \exp_after:wN \fp_to_int_small:w
+    \else:
+      \exp_after:wN \fp_to_int_large:w
+    \fi:
+    #2 e #3 \q_stop
+  }
+%    \end{macrocode}
+%   For small numbers, if the decimal part is greater than a half then
+%   there is rounding up to do.
+%    \begin{macrocode}
+\cs_new_nopar:Npn \fp_to_int_small:w #1 . #2 e #3 \q_stop
+  {
+    \if_int_compare:w #3 > \c_one
+    \else:
+      \if_int_compare:w #1 < \c_five
+        0
+      \else:
+        1
+      \fi:
+    \fi:
+  }
+%    \end{macrocode}
+% For large numbers, the idea is to split off the part for rounding,
+% do the rounding and fill if needed.
+%    \begin{macrocode}
+\cs_new_nopar:Npn \fp_to_int_large:w #1 . #2 e #3 \q_stop
+  {
+    \if_int_compare:w #3 < \c_ten
+      \exp_after:wN \fp_to_int_large_aux_i:w
+    \else:
+      \exp_after:wN \fp_to_int_large_aux_ii:w
+    \fi:
+    #1#2 e #3 \q_stop
+  }
+\cs_new_nopar:Npn \fp_to_int_large_aux_i:w #1#2 e #3 \q_stop
+  { \use:c { fp_to_int_large_aux_ #3 :w } #2 \q_stop {#1} }
+\cs_new_nopar:cpn { fp_to_int_large_aux_1:w } #1#2 \q_stop
+  { \fp_to_int_large_aux:nnn { #2 0 } {#1} }
+\cs_new_nopar:cpn { fp_to_int_large_aux_2:w } #1#2#3 \q_stop
+  { \fp_to_int_large_aux:nnn { #3 00 } {#1#2} }
+\cs_new_nopar:cpn { fp_to_int_large_aux_3:w } #1#2#3#4 \q_stop
+  { \fp_to_int_large_aux:nnn { #4 000 } {#1#2#3} }
+\cs_new_nopar:cpn { fp_to_int_large_aux_4:w } #1#2#3#4#5 \q_stop
+  { \fp_to_int_large_aux:nnn { #5 0000 } {#1#2#3#4} }
+\cs_new_nopar:cpn { fp_to_int_large_aux_5:w } #1#2#3#4#5#6 \q_stop
+  { \fp_to_int_large_aux:nnn { #6 00000 } {#1#2#3#4#5} }
+\cs_new_nopar:cpn { fp_to_int_large_aux_6:w } #1#2#3#4#5#6#7 \q_stop
+  { \fp_to_int_large_aux:nnn { #7 000000 } {#1#2#3#4#5#6} }
+\cs_new_nopar:cpn { fp_to_int_large_aux_7:w } #1#2#3#4#5#6#7#8 \q_stop
+  { \fp_to_int_large_aux:nnn { #8 0000000 } {#1#2#3#4#5#6#7} }
+\cs_new_nopar:cpn { fp_to_int_large_aux_8:w } #1#2#3#4#5#6#7#8#9 \q_stop
+  { \fp_to_int_large_aux:nnn { #9 00000000 } {#1#2#3#4#5#6#7#8} }
+\cs_new_nopar:cpn { fp_to_int_large_aux_9:w } #1 \q_stop {#1}
+\cs_new_nopar:Npn \fp_to_int_large_aux:nnn #1#2#3
+  {
+    \if_int_compare:w #1 < \c_five_hundred_million
+      #3#2
+    \else:
+      \int_value:w \int_eval:w #3#2 + 1 \int_eval_end:
+    \fi:
+  }
+\cs_new_nopar:Npn \fp_to_int_large_aux_ii:w #1 e #2 \q_stop
+  {
+    #1
+    \prg_replicate:nn { #2 - 9 } { 0 }
+  }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\fp_to_tl:N, \fp_to_tl:c}
+% \UnitTested
+% \begin{macro}[aux]{\fp_to_tl_aux:w}
+% \begin{macro}[aux]{\fp_to_tl_large:w}
+% \begin{macro}[aux]{\fp_to_tl_large_aux_i:w}
+% \begin{macro}[aux]{\fp_to_tl_large_aux_ii:w}
+% \begin{macro}[aux]{\fp_to_tl_large_0:w}
+% \begin{macro}[aux]{\fp_to_tl_large_1:w}
+% \begin{macro}[aux]{\fp_to_tl_large_2:w}
+% \begin{macro}[aux]{\fp_to_tl_large_3:w}
+% \begin{macro}[aux]{\fp_to_tl_large_4:w}
+% \begin{macro}[aux]{\fp_to_tl_large_5:w}
+% \begin{macro}[aux]{\fp_to_tl_large_6:w}
+% \begin{macro}[aux]{\fp_to_tl_large_7:w}
+% \begin{macro}[aux]{\fp_to_tl_large_8:w}
+% \begin{macro}[aux]{\fp_to_tl_large_8_aux:w}
+% \begin{macro}[aux]{\fp_to_tl_large_9:w}
+% \begin{macro}[aux]{\fp_to_tl_small:w}
+% \begin{macro}[aux]{\fp_to_tl_small_one:w}
+% \begin{macro}[aux]{\fp_to_tl_small_two:w}
+% \begin{macro}[aux]{\fp_to_tl_small_aux:w}
+% \begin{macro}[aux]{\fp_to_tl_large_zeros:NNNNNNNNN}
+% \begin{macro}[aux]{\fp_to_tl_small_zeros:NNNNNNNNN}
+% \begin{macro}[aux]{\fp_use_iix_ix:NNNNNNNNN}
+% \begin{macro}[aux]{\fp_use_ix:NNNNNNNNN}
+% \begin{macro}[aux]{\fp_use_i_to_vii:NNNNNNNNN}
+% \begin{macro}[aux]{\fp_use_i_to_iix:NNNNNNNNN}
+%   Converting to integers in an expandable manner is very similar to
+%   simply using floating point variables, particularly in the lead-off.
+%    \begin{macrocode}
+\cs_new_nopar:Npn \fp_to_tl:N #1
+  { \exp_after:wN \fp_to_tl_aux:w #1 \q_stop }
+\cs_generate_variant:Nn \fp_to_tl:N { c }
+\cs_new_nopar:Npn \fp_to_tl_aux:w #1#2 e #3 \q_stop
+  {
+    \if:w #1 -
+      -
+    \fi:
+    \if_int_compare:w #3 < \c_zero
+      \exp_after:wN \fp_to_tl_small:w
+    \else:
+      \exp_after:wN \fp_to_tl_large:w
+    \fi:
+    #2 e #3 \q_stop
+  }
+%    \end{macrocode}
+%   For \enquote{large} numbers (exponent $\ge 0$) there are two
+%   cases. For very large exponents ($ \ge 10 $) life is easy: apart
+%   from dropping extra zeros there is no work to do. On the other hand,
+%   for intermediate exponent values the decimal needs to be moved, then
+%   zeros can be dropped.
+%    \begin{macrocode}
+\cs_new_nopar:Npn \fp_to_tl_large:w #1 e #2 \q_stop
+  {
+    \if_int_compare:w #2 < \c_ten
+      \exp_after:wN \fp_to_tl_large_aux_i:w
+    \else:
+      \exp_after:wN \fp_to_tl_large_aux_ii:w
+    \fi:
+    #1 e #2 \q_stop
+  }
+\cs_new_nopar:Npn \fp_to_tl_large_aux_i:w #1 e #2 \q_stop
+  { \use:c { fp_to_tl_large_ #2 :w } #1 \q_stop }
+\cs_new_nopar:Npn \fp_to_tl_large_aux_ii:w #1 . #2 e #3 \q_stop
+  {
+    #1
+    \fp_to_tl_large_zeros:NNNNNNNNN #2
+    e #3
+  }
+\cs_new_nopar:cpn { fp_to_tl_large_0:w } #1 . #2 \q_stop
+  {
+    #1
+    \fp_to_tl_large_zeros:NNNNNNNNN #2
+  }
+\cs_new_nopar:cpn { fp_to_tl_large_1:w } #1 . #2#3 \q_stop
+  {
+    #1#2
+    \fp_to_tl_large_zeros:NNNNNNNNN #3 0
+  }
+\cs_new_nopar:cpn { fp_to_tl_large_2:w } #1 . #2#3#4 \q_stop
+  {
+    #1#2#3
+    \fp_to_tl_large_zeros:NNNNNNNNN #4 00
+  }
+\cs_new_nopar:cpn { fp_to_tl_large_3:w } #1 . #2#3#4#5 \q_stop
+  {
+    #1#2#3#4
+    \fp_to_tl_large_zeros:NNNNNNNNN #5 000
+  }
+\cs_new_nopar:cpn { fp_to_tl_large_4:w } #1 . #2#3#4#5#6 \q_stop
+  {
+    #1#2#3#4#5
+    \fp_to_tl_large_zeros:NNNNNNNNN #6 0000
+  }
+\cs_new_nopar:cpn { fp_to_tl_large_5:w } #1 . #2#3#4#5#6#7 \q_stop
+  {
+    #1#2#3#4#5#6
+    \fp_to_tl_large_zeros:NNNNNNNNN #7 00000
+  }
+\cs_new_nopar:cpn { fp_to_tl_large_6:w } #1 . #2#3#4#5#6#7#8 \q_stop
+  {
+    #1#2#3#4#5#6#7
+    \fp_to_tl_large_zeros:NNNNNNNNN #8 000000
+  }
+\cs_new_nopar:cpn { fp_to_tl_large_7:w } #1 . #2#3#4#5#6#7#8#9 \q_stop
+  {
+    #1#2#3#4#5#6#7#8
+    \fp_to_tl_large_zeros:NNNNNNNNN #9 0000000
+  }
+\cs_new_nopar:cpn { fp_to_tl_large_8:w } #1 .
+  {
+    #1
+    \use:c { fp_to_tl_large_8_aux:w }
+  }
+\cs_new_nopar:cpn { fp_to_tl_large_8_aux:w } #1#2#3#4#5#6#7#8#9 \q_stop
+  {
+    #1#2#3#4#5#6#7#8
+    \fp_to_tl_large_zeros:NNNNNNNNN #9 00000000
+  }
+\cs_new_nopar:cpn { fp_to_tl_large_9:w } #1 . #2 \q_stop {#1#2}
+%    \end{macrocode}
+%   Dealing with small numbers is a bit more complex as there has to be
+%   rounding. This makes life rather awkward, as there need to be a series
+%   of tests and calculations, as things cannot be stored in an
+%   expandable system.
+%    \begin{macrocode}
+\cs_new_nopar:Npn \fp_to_tl_small:w #1 e #2 \q_stop
+  {
+    \if_int_compare:w #2 = \c_minus_one
+      \exp_after:wN \fp_to_tl_small_one:w
+    \else:
+      \if_int_compare:w #2 = -\c_two
+        \exp_after:wN \exp_after:wN \exp_after:wN \fp_to_tl_small_two:w
+      \else:
+        \exp_after:wN \exp_after:wN \exp_after:wN \fp_to_tl_small_aux:w
+      \fi:
+    \fi:
+    #1 e #2 \q_stop
+  }
+\cs_new_nopar:Npn \fp_to_tl_small_one:w #1 . #2 e #3 \q_stop
+  {
+    \if_int_compare:w \fp_use_ix:NNNNNNNNN #2 > \c_four
+      \if_int_compare:w
+        \int_eval:w #1 \fp_use_i_to_iix:NNNNNNNNN #2 + 1
+          < \c_one_thousand_million
+        0.
+        \exp_after:wN \fp_to_tl_small_zeros:NNNNNNNNN
+          \int_value:w \int_eval:w
+              #1 \fp_use_i_to_iix:NNNNNNNNN #2 + 1
+            \int_eval_end:
+      \else:
+        1
+      \fi:
+    \else:
+      0. #1
+      \fp_to_tl_small_zeros:NNNNNNNNN #2
+    \fi:
+  }
+\cs_new_nopar:Npn \fp_to_tl_small_two:w #1 . #2 e #3 \q_stop
+  {
+    \if_int_compare:w \fp_use_iix_ix:NNNNNNNNN #2 > \c_forty_four
+      \if_int_compare:w
+        \int_eval:w #1 \fp_use_i_to_vii:NNNNNNNNN #2 0 + \c_ten
+          < \c_one_thousand_million
+        0.0
+        \exp_after:wN \fp_to_tl_small_zeros:NNNNNNNNN
+          \int_value:w \int_eval:w
+              #1 \fp_use_i_to_vii:NNNNNNNNN #2 0 + \c_ten
+            \int_eval_end:
+      \else:
+        0.1
+      \fi:
+    \else:
+      0.0
+      #1
+      \fp_to_tl_small_zeros:NNNNNNNNN #2
+    \fi:
+  }
+\cs_new_nopar:Npn \fp_to_tl_small_aux:w #1 . #2 e #3 \q_stop
+  {
+    #1
+    \fp_to_tl_large_zeros:NNNNNNNNN #2
+    e #3
+  }
+%    \end{macrocode}
+%   Rather than a complex recursion, the tests for finding trailing zeros
+%   are written out long-hand.  The difference between the two is only the
+%   need for a decimal marker.
+%    \begin{macrocode}
+\cs_new_nopar:Npn \fp_to_tl_large_zeros:NNNNNNNNN #1#2#3#4#5#6#7#8#9
+  {
+    \if_int_compare:w #9 = \c_zero
+      \if_int_compare:w #8 = \c_zero
+        \if_int_compare:w #7 = \c_zero
+          \if_int_compare:w #6 = \c_zero
+            \if_int_compare:w #5 = \c_zero
+              \if_int_compare:w #4 = \c_zero
+                \if_int_compare:w #3 = \c_zero
+                  \if_int_compare:w #2 = \c_zero
+                    \if_int_compare:w #1 = \c_zero
+                    \else:
+                      . #1
+                    \fi:
+                  \else:
+                    . #1#2
+                  \fi:
+                \else:
+                  . #1#2#3
+                \fi:
+              \else:
+                . #1#2#3#4
+              \fi:
+            \else:
+              . #1#2#3#4#5
+            \fi:
+          \else:
+            . #1#2#3#4#5#6
+          \fi:
+        \else:
+          . #1#2#3#4#5#6#7
+        \fi:
+      \else:
+         . #1#2#3#4#5#6#7#8
+      \fi:
+    \else:
+      . #1#2#3#4#5#6#7#8#9
+    \fi:
+  }
+\cs_new_nopar:Npn \fp_to_tl_small_zeros:NNNNNNNNN #1#2#3#4#5#6#7#8#9
+  {
+    \if_int_compare:w #9 = \c_zero
+      \if_int_compare:w #8 = \c_zero
+        \if_int_compare:w #7 = \c_zero
+          \if_int_compare:w #6 = \c_zero
+            \if_int_compare:w #5 = \c_zero
+              \if_int_compare:w #4 = \c_zero
+                \if_int_compare:w #3 = \c_zero
+                  \if_int_compare:w #2 = \c_zero
+                    \if_int_compare:w #1 = \c_zero
+                    \else:
+                      #1
+                    \fi:
+                  \else:
+                    #1#2
+                  \fi:
+                \else:
+                  #1#2#3
+                \fi:
+              \else:
+                #1#2#3#4
+              \fi:
+            \else:
+              #1#2#3#4#5
+            \fi:
+          \else:
+            #1#2#3#4#5#6
+          \fi:
+        \else:
+          #1#2#3#4#5#6#7
+        \fi:
+      \else:
+         #1#2#3#4#5#6#7#8
+      \fi:
+    \else:
+      #1#2#3#4#5#6#7#8#9
+    \fi:
+  }
+%    \end{macrocode}
+%   Some quick \enquote{return a few} functions.
+%    \begin{macrocode}
+\cs_new_nopar:Npn \fp_use_iix_ix:NNNNNNNNN #1#2#3#4#5#6#7#8#9 {#8#9}
+\cs_new_nopar:Npn \fp_use_ix:NNNNNNNNN #1#2#3#4#5#6#7#8#9 {#9}
+\cs_new_nopar:Npn \fp_use_i_to_vii:NNNNNNNNN #1#2#3#4#5#6#7#8#9
+  {#1#2#3#4#5#6#7}
+\cs_new_nopar:Npn \fp_use_i_to_iix:NNNNNNNNN #1#2#3#4#5#6#7#8#9
+  {#1#2#3#4#5#6#7#8}
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \subsection{Rounding numbers}
+%
+% The results may well need to be rounded. A couple of related functions
+% to do this for a stored value.
+%
+% \begin{macro}{\fp_round_figures:Nn,  \fp_round_figures:cn}
+% \UnitTested
+% \begin{macro}{\fp_ground_figures:Nn, \fp_ground_figures:cn}
+% \UnitTested
+% \begin{macro}[aux]{\fp_round_figures_aux:NNn}
+%   Rounding to figures needs only an adjustment to the target by one
+%   (as the target is in decimal places).
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \fp_round_figures:Nn
+  { \fp_round_figures_aux:NNn \tl_set:Nn }
+\cs_generate_variant:Nn \fp_round_figures:Nn { c }
+\cs_new_protected_nopar:Npn \fp_ground_figures:Nn
+  { \fp_round_figures_aux:NNn \tl_gset:Nn }
+\cs_generate_variant:Nn \fp_ground_figures:Nn { c }
+\cs_new_protected_nopar:Npn \fp_round_figures_aux:NNn #1#2#3
+  {
+    \group_begin:
+      \fp_read:N #2
+      \int_set:Nn \l_fp_round_target_int { #3 - 1 }
+      \if_int_compare:w \l_fp_round_target_int < \c_ten
+        \exp_after:wN \fp_round:
+      \fi:
+      \tex_advance:D \l_fp_input_a_decimal_int \c_one_thousand_million
+      \cs_set_protected_nopar:Npx \fp_tmp:w
+        {
+          \group_end:
+          #1 \exp_not:N #2
+            {
+              \if_int_compare:w \l_fp_input_a_sign_int < \c_zero
+                -
+              \else:
+                +
+              \fi:
+              \int_use:N \l_fp_input_a_integer_int
+              .
+              \exp_after:wN \use_none:n
+                \int_use:N \l_fp_input_a_decimal_int
+              e
+              \int_use:N \l_fp_input_a_exponent_int
+            }
+        }
+    \fp_tmp:w
+  }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\fp_round_places:Nn,  \fp_round_places:cn}
+% \UnitTested
+% \begin{macro}{\fp_ground_places:Nn, \fp_ground_places:cn}
+% \UnitTested
+% \begin{macro}[aux]{\fp_round_places_aux:NNn}
+%   Rounding to places needs an adjustment for the exponent value, which
+%   will mean that everything should be correct.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \fp_round_places:Nn
+  { \fp_round_places_aux:NNn \tl_set:Nn }
+\cs_generate_variant:Nn \fp_round_places:Nn { c }
+\cs_new_protected_nopar:Npn \fp_ground_places:Nn
+  { \fp_round_places_aux:NNn \tl_gset:Nn }
+\cs_generate_variant:Nn \fp_ground_places:Nn { c }
+\cs_new_protected_nopar:Npn \fp_round_places_aux:NNn #1#2#3
+  {
+    \group_begin:
+      \fp_read:N #2
+      \int_set:Nn \l_fp_round_target_int
+        { #3 + \l_fp_input_a_exponent_int }
+      \if_int_compare:w \l_fp_round_target_int < \c_ten
+        \exp_after:wN \fp_round:
+      \fi:
+      \tex_advance:D \l_fp_input_a_decimal_int \c_one_thousand_million
+      \cs_set_protected_nopar:Npx \fp_tmp:w
+        {
+          \group_end:
+          #1 \exp_not:N #2
+            {
+              \if_int_compare:w \l_fp_input_a_sign_int < \c_zero
+                -
+              \else:
+                +
+              \fi:
+              \int_use:N \l_fp_input_a_integer_int
+              .
+              \exp_after:wN \use_none:n
+                \int_use:N \l_fp_input_a_decimal_int
+              e
+              \int_use:N \l_fp_input_a_exponent_int
+            }
+        }
+    \fp_tmp:w
+  }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\fp_round:}
+% \begin{macro}[aux]{\fp_round_aux:NNNNNNNNN}
+% \begin{macro}{\fp_round_loop:N}
+%   The rounding approach is the same for decimal places and significant
+%   figures. There are always nine decimal digits to round, so the code
+%   can be written to account for this. The basic logic is simply to
+%   find the rounding, track any carry digit and move along. At the end
+%   of the loop there is a possible shuffle if the integer part has
+%   become $10$.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \fp_round:
+  {
+    \bool_set_false:N \l_fp_round_carry_bool
+    \l_fp_round_position_int \c_eight
+    \tl_clear:N \l_fp_round_decimal_tl
+    \tex_advance:D \l_fp_input_a_decimal_int \c_one_thousand_million
+    \exp_after:wN \use_i:nn \exp_after:wN
+      \fp_round_aux:NNNNNNNNN \int_use:N \l_fp_input_a_decimal_int
+  }
+\cs_new_protected_nopar:Npn \fp_round_aux:NNNNNNNNN #1#2#3#4#5#6#7#8#9
+  {
+    \fp_round_loop:N #9#8#7#6#5#4#3#2#1
+    \bool_if:NT \l_fp_round_carry_bool
+      { \tex_advance:D \l_fp_input_a_integer_int \c_one }
+    \l_fp_input_a_decimal_int \l_fp_round_decimal_tl \scan_stop:
+    \if_int_compare:w \l_fp_input_a_integer_int < \c_ten
+    \else:
+      \l_fp_input_a_integer_int \c_one
+      \tex_divide:D \l_fp_input_a_decimal_int \c_ten
+      \tex_advance:D \l_fp_input_a_exponent_int \c_one
+    \fi:
+  }
+\cs_new_protected_nopar:Npn \fp_round_loop:N #1
+  {
+    \if_int_compare:w \l_fp_round_position_int < \l_fp_round_target_int
+      \bool_if:NTF \l_fp_round_carry_bool
+        { \l_fp_tmp_int \int_eval:w #1 + \c_one \scan_stop: }
+        { \l_fp_tmp_int \int_eval:w #1 \scan_stop: }
+      \if_int_compare:w \l_fp_tmp_int = \c_ten
+        \l_fp_tmp_int \c_zero
+      \else:
+        \bool_set_false:N \l_fp_round_carry_bool
+      \fi:
+      \tl_set:Nx \l_fp_round_decimal_tl
+        { \int_use:N \l_fp_tmp_int \l_fp_round_decimal_tl }
+    \else:
+      \tl_set:Nx \l_fp_round_decimal_tl { 0 \l_fp_round_decimal_tl }
+      \if_int_compare:w \l_fp_round_position_int = \l_fp_round_target_int
+        \if_int_compare:w #1 > \c_four
+          \bool_set_true:N \l_fp_round_carry_bool
+        \fi:
+      \fi:
+    \fi:
+    \tex_advance:D \l_fp_round_position_int \c_minus_one
+    \if_int_compare:w \l_fp_round_position_int > \c_minus_one
+      \exp_after:wN \fp_round_loop:N
+    \fi:
+  }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \subsection{Unary functions}
+%
+% \begin{macro}{\fp_abs:N,  \fp_abs:c}
+% \UnitTested
+% \begin{macro}{\fp_gabs:N, \fp_gabs:c}
+% \UnitTested
+% \begin{macro}[aux]{\fp_abs_aux:NN}
+%   Setting the absolute value is easy: read the value, ignore the sign,
+%   return the result.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \fp_abs:N  { \fp_abs_aux:NN \tl_set:Nn }
+\cs_new_protected_nopar:Npn \fp_gabs:N { \fp_abs_aux:NN \tl_gset:Nn }
+\cs_generate_variant:Nn \fp_abs:N  { c }
+\cs_generate_variant:Nn \fp_gabs:N { c }
+\cs_new_protected_nopar:Npn \fp_abs_aux:NN #1#2
+  {
+    \group_begin:
+      \fp_read:N #2
+      \tex_advance:D \l_fp_input_a_decimal_int \c_one_thousand_million
+      \cs_set_protected_nopar:Npx \fp_tmp:w
+        {
+          \group_end:
+          #1 \exp_not:N #2
+            {
+              +
+              \int_use:N \l_fp_input_a_integer_int
+              .
+              \exp_after:wN \use_none:n
+                \int_use:N \l_fp_input_a_decimal_int
+              e
+              \int_use:N \l_fp_input_a_exponent_int
+            }
+        }
+    \fp_tmp:w
+  }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\fp_neg:N,  \fp_neg:c}
+% \UnitTested
+% \begin{macro}{\fp_gneg:N, \fp_gneg:c}
+% \UnitTested
+% \begin{macro}[aux]{\fp_neg:NN}
+% Just a bit more complex: read the input, reverse the sign and
+% output the result.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \fp_neg:N  { \fp_neg_aux:NN \tl_set:Nn }
+\cs_new_protected_nopar:Npn \fp_gneg:N { \fp_neg_aux:NN \tl_gset:Nn }
+\cs_generate_variant:Nn \fp_neg:N  { c }
+\cs_generate_variant:Nn \fp_gneg:N { c }
+\cs_new_protected_nopar:Npn \fp_neg_aux:NN #1#2
+  {
+    \group_begin:
+      \fp_read:N #2
+      \tex_advance:D \l_fp_input_a_decimal_int \c_one_thousand_million
+      \tl_set:Nx \l_fp_tmp_tl
+        {
+          \if_int_compare:w \l_fp_input_a_sign_int < \c_zero
+            +
+          \else:
+            -
+          \fi:
+          \int_use:N \l_fp_input_a_integer_int
+          .
+          \exp_after:wN \use_none:n
+            \int_use:N \l_fp_input_a_decimal_int
+          e
+          \int_use:N \l_fp_input_a_exponent_int
+        }
+    \exp_after:wN \group_end: \exp_after:wN
+    #1 \exp_after:wN #2 \exp_after:wN { \l_fp_tmp_tl }
+  }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \subsection{Basic arithmetic}
+%
+% \begin{macro}{\fp_add:Nn, \fp_add:cn}
+% \UnitTested
+% \begin{macro}{\fp_gadd:Nn,\fp_gadd:cn}
+% \UnitTested
+% \begin{macro}[aux]{\fp_add_aux:NNn}
+% \begin{macro}[aux]{\fp_add_core:}
+% \begin{macro}[aux]{\fp_add_sum:}
+% \begin{macro}[aux]{\fp_add_difference:}
+%   The various addition functions are simply different ways to call the
+%   single master function below. This pattern is repeated for the
+%   other arithmetic functions.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \fp_add:Nn  { \fp_add_aux:NNn \tl_set:Nn }
+\cs_new_protected_nopar:Npn \fp_gadd:Nn { \fp_add_aux:NNn \tl_gset:Nn }
+\cs_generate_variant:Nn \fp_add:Nn   { c }
+\cs_generate_variant:Nn \fp_gadd:Nn  { c }
+%    \end{macrocode}
+% Addition takes place using one of two paths. If the signs of the
+% two parts are the same, they are simply combined. On the other
+% hand, if the signs are different the calculation finds this
+% difference.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \fp_add_aux:NNn #1#2#3
+  {
+    \group_begin:
+      \fp_read:N #2
+      \fp_split:Nn b {#3}
+      \fp_standardise:NNNN
+        \l_fp_input_b_sign_int
+        \l_fp_input_b_integer_int
+        \l_fp_input_b_decimal_int
+        \l_fp_input_b_exponent_int
+      \fp_add_core:
+    \fp_tmp:w #1#2
+  }
+\cs_new_protected_nopar:Npn \fp_add_core:
+  {
+    \fp_level_input_exponents:
+    \if_int_compare:w
+      \int_eval:w
+        \l_fp_input_a_sign_int * \l_fp_input_b_sign_int
+        > \c_zero
+      \exp_after:wN \fp_add_sum:
+    \else:
+      \exp_after:wN \fp_add_difference:
+    \fi:
+    \l_fp_output_exponent_int \l_fp_input_a_exponent_int
+    \fp_standardise:NNNN
+      \l_fp_output_sign_int
+      \l_fp_output_integer_int
+      \l_fp_output_decimal_int
+      \l_fp_output_exponent_int
+    \cs_set_protected_nopar:Npx \fp_tmp:w ##1##2
+      {
+        \group_end:
+        ##1 ##2
+          {
+             \if_int_compare:w \l_fp_output_sign_int < \c_zero
+              -
+            \else:
+              +
+            \fi:
+            \int_use:N \l_fp_output_integer_int
+            .
+            \exp_after:wN \use_none:n
+              \int_value:w \int_eval:w
+                 \l_fp_output_decimal_int + \c_one_thousand_million
+            e
+            \int_use:N \l_fp_output_exponent_int
+          }
+      }
+  }
+%    \end{macrocode}
+%   Finding the sum of two numbers is trivially easy.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \fp_add_sum:
+  {
+    \l_fp_output_sign_int \l_fp_input_a_sign_int
+    \l_fp_output_integer_int
+      \int_eval:w
+        \l_fp_input_a_integer_int + \l_fp_input_b_integer_int
+      \scan_stop:
+    \l_fp_output_decimal_int
+      \int_eval:w
+        \l_fp_input_a_decimal_int + \l_fp_input_b_decimal_int
+      \scan_stop:
+    \if_int_compare:w \l_fp_output_decimal_int < \c_one_thousand_million
+    \else:
+      \tex_advance:D \l_fp_output_integer_int \c_one
+      \tex_advance:D \l_fp_output_decimal_int -\c_one_thousand_million
+    \fi:
+  }
+%    \end{macrocode}
+%   When the signs of the two parts of the input are different, the
+%   absolute difference is worked out first. There is then a calculation
+%   to see which way around everything has worked out, so that the final
+%   sign is correct. The difference might also give a zero result with
+%   a negative sign, which is reversed as zero is regarded as positive.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \fp_add_difference:
+  {
+    \l_fp_output_integer_int
+      \int_eval:w
+        \l_fp_input_a_integer_int - \l_fp_input_b_integer_int
+      \scan_stop:
+    \l_fp_output_decimal_int
+      \int_eval:w
+        \l_fp_input_a_decimal_int - \l_fp_input_b_decimal_int
+      \scan_stop:
+    \if_int_compare:w \l_fp_output_decimal_int < \c_zero
+      \tex_advance:D \l_fp_output_integer_int \c_minus_one
+      \tex_advance:D \l_fp_output_decimal_int \c_one_thousand_million
+    \fi:
+    \if_int_compare:w \l_fp_output_integer_int < \c_zero
+      \l_fp_output_sign_int \l_fp_input_b_sign_int
+      \if_int_compare:w \l_fp_output_decimal_int = \c_zero
+        \l_fp_output_integer_int -\l_fp_output_integer_int
+      \else:
+        \l_fp_output_decimal_int
+          \int_eval:w
+            \c_one_thousand_million - \l_fp_output_decimal_int
+          \scan_stop:
+        \l_fp_output_integer_int
+           \int_eval:w
+             - \l_fp_output_integer_int - \c_one
+           \scan_stop:
+      \fi:
+    \else:
+      \l_fp_output_sign_int \l_fp_input_a_sign_int
+    \fi:
+  }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\fp_sub:Nn, \fp_sub:cn}
+% \UnitTested
+% \begin{macro}{\fp_gsub:Nn,\fp_gsub:cn}
+% \UnitTested
+% \begin{macro}[aux]{\fp_sub_aux:NNn}
+%   Subtraction is essentially the same as addition, but with the sign
+%   of the second component reversed. Thus the core of the two function
+%   groups is the same, with just a little set up here.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \fp_sub:Nn  { \fp_sub_aux:NNn \tl_set:Nn }
+\cs_new_protected_nopar:Npn \fp_gsub:Nn { \fp_sub_aux:NNn \tl_gset:Nn }
+\cs_generate_variant:Nn \fp_sub:Nn   { c }
+\cs_generate_variant:Nn \fp_gsub:Nn  { c }
+\cs_new_protected_nopar:Npn \fp_sub_aux:NNn #1#2#3
+  {
+    \group_begin:
+      \fp_read:N #2
+      \fp_split:Nn b {#3}
+      \fp_standardise:NNNN
+        \l_fp_input_b_sign_int
+        \l_fp_input_b_integer_int
+        \l_fp_input_b_decimal_int
+        \l_fp_input_b_exponent_int
+      \tex_multiply:D \l_fp_input_b_sign_int \c_minus_one
+      \fp_add_core:
+    \fp_tmp:w #1#2
+  }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\fp_mul:Nn, \fp_mul:cn}
+% \UnitTested
+% \begin{macro}{\fp_gmul:Nn,\fp_gmul:cn}
+% \UnitTested
+% \begin{macro}[aux]{\fp_mul_aux:NNn}
+% \begin{macro}[aux]{\fp_mul_internal:}
+% \begin{macro}[aux]{\fp_mul_split:NNNN}
+% \begin{macro}[aux]{\fp_mul_split:w}
+% \begin{macro}[aux]{\fp_mul_end_level:}
+% \begin{macro}[aux]{\fp_mul_end_level:NNNNNNNNN}
+%   The pattern is much the same for multiplication.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \fp_mul:Nn  { \fp_mul_aux:NNn \tl_set:Nn }
+\cs_new_protected_nopar:Npn \fp_gmul:Nn { \fp_mul_aux:NNn \tl_gset:Nn }
+\cs_generate_variant:Nn \fp_mul:Nn  { c }
+\cs_generate_variant:Nn \fp_gmul:Nn { c }
+%    \end{macrocode}
+%   The approach to multiplication is as follows. First, the two numbers
+%   are split into blocks of three digits. These are then multiplied
+%   together to find products for each group of three output digits. This
+%   is al written out in full for speed reasons. Between each block of
+%   three digits in the output, there is a carry step. The very lowest
+%   digits are not calculated, while
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \fp_mul_aux:NNn #1#2#3
+  {
+    \group_begin:
+      \fp_read:N #2
+      \fp_split:Nn b {#3}
+      \fp_standardise:NNNN
+        \l_fp_input_b_sign_int
+        \l_fp_input_b_integer_int
+        \l_fp_input_b_decimal_int
+        \l_fp_input_b_exponent_int
+      \fp_mul_internal:
+      \l_fp_output_exponent_int
+        \int_eval:w
+          \l_fp_input_a_exponent_int + \l_fp_input_b_exponent_int
+        \scan_stop:
+      \fp_standardise:NNNN
+        \l_fp_output_sign_int
+        \l_fp_output_integer_int
+        \l_fp_output_decimal_int
+        \l_fp_output_exponent_int
+      \cs_set_protected_nopar:Npx \fp_tmp:w
+        {
+          \group_end:
+          #1 \exp_not:N #2
+            {
+              \if_int_compare:w
+                \int_eval:w
+                  \l_fp_input_a_sign_int * \l_fp_input_b_sign_int
+                  < \c_zero
+                \if_int_compare:w
+                  \int_eval:w
+                    \l_fp_output_integer_int + \l_fp_output_decimal_int
+                    = \c_zero
+                  +
+                \else:
+                  -
+                \fi:
+              \else:
+                +
+             \fi:
+              \int_use:N \l_fp_output_integer_int
+              .
+              \exp_after:wN \use_none:n
+                \int_value:w \int_eval:w
+                   \l_fp_output_decimal_int + \c_one_thousand_million
+              e
+              \int_use:N \l_fp_output_exponent_int
+            }
+        }
+    \fp_tmp:w
+  }
+%    \end{macrocode}
+%   Done separately so that the internal use is a bit easier.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \fp_mul_internal:
+  {
+    \fp_mul_split:NNNN \l_fp_input_a_decimal_int
+      \l_fp_mul_a_i_int \l_fp_mul_a_ii_int \l_fp_mul_a_iii_int
+    \fp_mul_split:NNNN \l_fp_input_b_decimal_int
+      \l_fp_mul_b_i_int \l_fp_mul_b_ii_int \l_fp_mul_b_iii_int
+    \l_fp_mul_output_int \c_zero
+    \tl_clear:N \l_fp_mul_output_tl
+    \fp_mul_product:NN \l_fp_mul_a_i_int         \l_fp_mul_b_iii_int
+    \fp_mul_product:NN \l_fp_mul_a_ii_int        \l_fp_mul_b_ii_int
+    \fp_mul_product:NN \l_fp_mul_a_iii_int       \l_fp_mul_b_i_int
+    \tex_divide:D \l_fp_mul_output_int \c_one_thousand
+    \fp_mul_product:NN \l_fp_input_a_integer_int \l_fp_mul_b_iii_int
+    \fp_mul_product:NN \l_fp_mul_a_i_int         \l_fp_mul_b_ii_int
+    \fp_mul_product:NN \l_fp_mul_a_ii_int        \l_fp_mul_b_i_int
+    \fp_mul_product:NN \l_fp_mul_a_iii_int       \l_fp_input_b_integer_int
+    \fp_mul_end_level:
+    \fp_mul_product:NN \l_fp_input_a_integer_int \l_fp_mul_b_ii_int
+    \fp_mul_product:NN \l_fp_mul_a_i_int         \l_fp_mul_b_i_int
+    \fp_mul_product:NN \l_fp_mul_a_ii_int        \l_fp_input_b_integer_int
+    \fp_mul_end_level:
+    \fp_mul_product:NN \l_fp_input_a_integer_int \l_fp_mul_b_i_int
+    \fp_mul_product:NN \l_fp_mul_a_i_int         \l_fp_input_b_integer_int
+    \fp_mul_end_level:
+    \l_fp_output_decimal_int 0 \l_fp_mul_output_tl \scan_stop:
+    \tl_clear:N \l_fp_mul_output_tl
+    \fp_mul_product:NN \l_fp_input_a_integer_int \l_fp_input_b_integer_int
+    \fp_mul_end_level:
+    \l_fp_output_integer_int 0 \l_fp_mul_output_tl \scan_stop:
+  }
+%    \end{macrocode}
+%   The split works by making a $10$ digit number, from which
+%   the first digit can then be dropped using a delimited argument. The
+%   groups of three digits are then assigned to the various parts of
+%   the input: notice that |##9| contains the last two digits of the
+%   smallest part of the input.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \fp_mul_split:NNNN #1#2#3#4
+  {
+    \tex_advance:D #1 \c_one_thousand_million
+    \cs_set_protected_nopar:Npn \fp_mul_split_aux:w
+       ##1##2##3##4##5##6##7##8##9 \q_stop {
+        #2 ##2##3##4 \scan_stop:
+        #3 ##5##6##7 \scan_stop:
+        #4 ##8##9    \scan_stop:
+      }
+    \exp_after:wN \fp_mul_split_aux:w \int_use:N #1 \q_stop
+    \tex_advance:D #1 -\c_one_thousand_million
+  }
+\cs_new_protected_nopar:Npn \fp_mul_product:NN #1#2
+  {
+    \l_fp_mul_output_int
+      \int_eval:w \l_fp_mul_output_int + #1 * #2 \scan_stop:
+  }
+%    \end{macrocode}
+%   At the end of each output group of three, there is a transfer of
+%   information so that there is no danger of an overflow. This is done by
+%   expansion to keep the number of calculations down.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \fp_mul_end_level:
+  {
+    \tex_advance:D \l_fp_mul_output_int \c_one_thousand_million
+    \exp_after:wN \use_i:nn \exp_after:wN
+      \fp_mul_end_level:NNNNNNNNN \int_use:N \l_fp_mul_output_int
+  }
+\cs_new_protected_nopar:Npn \fp_mul_end_level:NNNNNNNNN #1#2#3#4#5#6#7#8#9
+  {
+    \tl_set:Nx \l_fp_mul_output_tl { #7#8#9 \l_fp_mul_output_tl }
+    \l_fp_mul_output_int #1#2#3#4#5#6 \scan_stop:
+  }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\fp_div:Nn, \fp_div:cn}
+% \UnitTested
+% \begin{macro}{\fp_gdiv:Nn,\fp_gdiv:cn}
+% \UnitTested
+% \begin{macro}[aux]{\fp_div_aux:NNn}
+% \begin{macro}{\fp_div_internal:}
+% \begin{macro}[aux]{\fp_div_loop:}
+% \begin{macro}[aux]{\fp_div_divide:}
+% \begin{macro}[aux]{\fp_div_divide_aux:}
+% \begin{macro}[aux]{\fp_div_store:}
+% \begin{macro}[aux]{\fp_div_store_integer:}
+% \begin{macro}[aux]{\fp_div_store_decimal:}
+%   The pattern is much the same for multiplication.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \fp_div:Nn  { \fp_div_aux:NNn \tl_set:Nn }
+\cs_new_protected_nopar:Npn \fp_gdiv:Nn { \fp_div_aux:NNn \tl_gset:Nn }
+\cs_generate_variant:Nn \fp_div:Nn  { c }
+\cs_generate_variant:Nn \fp_gdiv:Nn { c }
+%    \end{macrocode}
+%   Division proper starts with a couple of tests. If the denominator is
+%   zero then a error is issued. On the other hand, if the numerator is
+%   zero then the result must be $0.0$ and can be given with no
+%   further work.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \fp_div_aux:NNn #1#2#3
+  {
+    \group_begin:
+      \fp_read:N #2
+      \fp_split:Nn b {#3}
+      \fp_standardise:NNNN
+        \l_fp_input_b_sign_int
+        \l_fp_input_b_integer_int
+        \l_fp_input_b_decimal_int
+        \l_fp_input_b_exponent_int
+      \if_int_compare:w
+        \int_eval:w
+          \l_fp_input_b_integer_int + \l_fp_input_b_decimal_int
+          = \c_zero
+        \cs_set_protected_nopar:Npx \fp_tmp:w ##1##2
+          {
+            \group_end:
+            #1 \exp_not:N #2 { \c_undefined_fp }
+          }
+      \else:
+        \if_int_compare:w
+          \int_eval:w
+            \l_fp_input_a_integer_int + \l_fp_input_a_decimal_int
+            = \c_zero
+          \cs_set_protected_nopar:Npx \fp_tmp:w ##1##2
+            {
+              \group_end:
+              #1 \exp_not:N #2 { \c_zero_fp }
+            }
+        \else:
+          \exp_after:wN \exp_after:wN \exp_after:wN \fp_div_internal:
+        \fi:
+      \fi:
+    \fp_tmp:w #1#2
+  }
+%    \end{macrocode}
+%   The main division algorithm works by finding how many times |b| can
+%   be removed from |a|, storing the result and doing the subtraction.
+%   Input |a| is then multiplied by $10$, and the process is repeated.
+%   The looping ends either when there is nothing left of |a|
+%   (\emph{i.e.}~an exact result) or when the code reaches the ninth
+%   decimal place. Most of the process takes place in the loop function
+%   below.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \fp_div_internal: {
+  \l_fp_output_integer_int \c_zero
+  \l_fp_output_decimal_int \c_zero
+  \cs_set_eq:NN \fp_div_store: \fp_div_store_integer:
+  \l_fp_div_offset_int \c_one_hundred_million
+  \fp_div_loop:
+  \l_fp_output_exponent_int
+    \int_eval:w
+      \l_fp_input_a_exponent_int - \l_fp_input_b_exponent_int
+    \scan_stop:
+  \fp_standardise:NNNN
+    \l_fp_output_sign_int
+    \l_fp_output_integer_int
+    \l_fp_output_decimal_int
+    \l_fp_output_exponent_int
+  \cs_set_protected_nopar:Npx \fp_tmp:w ##1##2
+    {
+      \group_end:
+      ##1 ##2
+        {
+          \if_int_compare:w
+            \int_eval:w
+              \l_fp_input_a_sign_int * \l_fp_input_b_sign_int
+              < \c_zero
+            \if_int_compare:w
+              \int_eval:w
+                \l_fp_output_integer_int + \l_fp_output_decimal_int
+                = \c_zero
+              +
+            \else:
+              -
+            \fi:
+          \else:
+            +
+          \fi:
+          \int_use:N \l_fp_output_integer_int
+          .
+          \exp_after:wN \use_none:n
+            \int_value:w \int_eval:w
+               \l_fp_output_decimal_int + \c_one_thousand_million
+             \int_eval_end:
+          e
+          \int_use:N \l_fp_output_exponent_int
+        }
+    }
+}
+%    \end{macrocode}
+%   The main loop implements the approach described above. The storing
+%   function is done as a function so that the integer and decimal parts
+%   can be done separately but rapidly.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \fp_div_loop:
+  {
+    \l_fp_count_int \c_zero
+    \fp_div_divide:
+    \fp_div_store:
+    \tex_multiply:D \l_fp_input_a_integer_int \c_ten
+    \tex_advance:D \l_fp_input_a_decimal_int \c_one_thousand_million
+    \exp_after:wN \fp_div_loop_step:w
+      \int_use:N \l_fp_input_a_decimal_int \q_stop
+    \if_int_compare:w
+      \int_eval:w \l_fp_input_a_integer_int + \l_fp_input_a_decimal_int
+        > \c_zero
+        \if_int_compare:w \l_fp_div_offset_int > \c_zero
+          \exp_after:wN \exp_after:wN \exp_after:wN
+            \fp_div_loop:
+        \fi:
+    \fi:
+  }
+%    \end{macrocode}
+%   Checking to see if the numerator can be divides needs quite an
+%   involved check. Either the integer part has to be bigger for the
+%   numerator or, if it is not smaller then the decimal part of the
+%   numerator must not be smaller than that of the denominator. Once
+%   the test is right the rest is much as elsewhere.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \fp_div_divide:
+  {
+    \if_int_compare:w \l_fp_input_a_integer_int > \l_fp_input_b_integer_int
+      \exp_after:wN \fp_div_divide_aux:
+    \else:
+      \if_int_compare:w \l_fp_input_a_integer_int < \l_fp_input_b_integer_int
+      \else:
+        \if_int_compare:w
+          \l_fp_input_a_decimal_int < \l_fp_input_b_decimal_int
+        \else:
+          \exp_after:wN \exp_after:wN \exp_after:wN
+            \exp_after:wN \exp_after:wN \exp_after:wN
+            \exp_after:wN \fp_div_divide_aux:
+        \fi:
+      \fi:
+    \fi:
+  }
+\cs_new_protected_nopar:Npn \fp_div_divide_aux:
+  {
+    \tex_advance:D \l_fp_count_int \c_one
+    \tex_advance:D \l_fp_input_a_integer_int -\l_fp_input_b_integer_int
+    \tex_advance:D \l_fp_input_a_decimal_int -\l_fp_input_b_decimal_int
+    \if_int_compare:w \l_fp_input_a_decimal_int < \c_zero
+      \tex_advance:D \l_fp_input_a_integer_int \c_minus_one
+      \tex_advance:D \l_fp_input_a_decimal_int \c_one_thousand_million
+    \fi:
+    \fp_div_divide:
+  }
+%    \end{macrocode}
+%   Storing the number of each division is done differently for the
+%   integer and decimal. The integer is easy and a one-off, while the
+%   decimal also needs to account for the position of the digit to store.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \fp_div_store: { }
+\cs_new_protected_nopar:Npn \fp_div_store_integer:
+  {
+    \l_fp_output_integer_int \l_fp_count_int
+    \cs_set_eq:NN \fp_div_store: \fp_div_store_decimal:
+  }
+\cs_new_protected_nopar:Npn \fp_div_store_decimal:
+  {
+    \l_fp_output_decimal_int
+      \int_eval:w
+        \l_fp_output_decimal_int +
+        \l_fp_count_int * \l_fp_div_offset_int
+      \int_eval_end:
+    \tex_divide:D \l_fp_div_offset_int \c_ten
+  }
+\cs_new_protected_nopar:Npn \fp_div_loop_step:w #1#2#3#4#5#6#7#8#9 \q_stop
+  {
+    \l_fp_input_a_integer_int
+      \int_eval:w #2 + \l_fp_input_a_integer_int  \int_eval_end:
+    \l_fp_input_a_decimal_int #3#4#5#6#7#8#9 0 \scan_stop:
+  }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \subsection{Arithmetic for internal use}
+%
+% For the more complex functions, it is only possible to deliver
+% reliable $10$ digit accuracy if the internal calculations are
+% carried out to a higher degree of precision. This is done using a
+% second set of functions so that the `user' versions are not
+% slowed down. These versions are also focussed on the needs of internal
+% calculations. No error checking, sign checking or exponent levelling
+% is done. For addition and subtraction, the arguments are:
+% \begin{itemize}
+%   \item Integer part of input |a|.
+%   \item Decimal part of input |a|.
+%   \item Additional decimal part of input |a|.
+%   \item Integer part of input |b|.
+%   \item Decimal part of input |b|.
+%   \item Additional decimal part of input |b|.
+%   \item Integer part of output.
+%   \item Decimal part of output.
+%   \item Additional decimal part of output.
+% \end{itemize}
+% The situation for multiplication and division is a little different as
+% they only deal with the decimal part.
+%
+% \begin{macro}{\fp_add:NNNNNNNNN}
+%   The internal sum is always exactly that: it is always a sum and there
+%   is no sign check.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \fp_add:NNNNNNNNN #1#2#3#4#5#6#7#8#9
+  {
+    #7 \int_eval:w #1 + #4 \int_eval_end:
+    #8 \int_eval:w #2 + #5 \int_eval_end:
+    #9 \int_eval:w #3 + #6 \int_eval_end:
+    \if_int_compare:w #9 < \c_one_thousand_million
+    \else:
+      \tex_advance:D #8 \c_one
+      \tex_advance:D #9 -\c_one_thousand_million
+    \fi:
+    \if_int_compare:w #8 < \c_one_thousand_million
+    \else:
+      \tex_advance:D #7 \c_one
+      \tex_advance:D #8 -\c_one_thousand_million
+    \fi:
+  }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\fp_sub:NNNNNNNNN}
+%   Internal subtraction is needed only when the first number is bigger
+%   than the second, so there is no need to worry about the sign. This is
+%   a good job as there are no arguments left. The flipping flag is
+%   used in the rare case where a sign change is possible.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \fp_sub:NNNNNNNNN #1#2#3#4#5#6#7#8#9
+  {
+    #7 \int_eval:w #1 - #4 \int_eval_end:
+    #8 \int_eval:w #2 - #5 \int_eval_end:
+    #9 \int_eval:w #3 - #6 \int_eval_end:
+    \if_int_compare:w #9 < \c_zero
+      \tex_advance:D #8 \c_minus_one
+      \tex_advance:D #9 \c_one_thousand_million
+    \fi:
+    \if_int_compare:w #8 < \c_zero
+      \tex_advance:D #7 \c_minus_one
+      \tex_advance:D #8 \c_one_thousand_million
+    \fi:
+    \if_int_compare:w #7 < \c_zero
+      \if_int_compare:w \int_eval:w #8 + #9 = \c_zero
+        #7 -#7
+      \else:
+        \tex_advance:D #7 \c_one
+        #8 \int_eval:w \c_one_thousand_million - #8 \int_eval_end:
+        #9 \int_eval:w \c_one_thousand_million - #9 \int_eval_end:
+      \fi:
+    \fi:
+  }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\fp_mul:NNNNNN}
+% Decimal-part only multiplication but with higher accuracy than the
+% user version.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \fp_mul:NNNNNN #1#2#3#4#5#6
+  {
+    \fp_mul_split:NNNN #1
+      \l_fp_mul_a_i_int \l_fp_mul_a_ii_int \l_fp_mul_a_iii_int
+    \fp_mul_split:NNNN #2
+      \l_fp_mul_a_iv_int \l_fp_mul_a_v_int \l_fp_mul_a_vi_int
+    \fp_mul_split:NNNN #3
+      \l_fp_mul_b_i_int \l_fp_mul_b_ii_int \l_fp_mul_b_iii_int
+    \fp_mul_split:NNNN #4
+      \l_fp_mul_b_iv_int \l_fp_mul_b_v_int \l_fp_mul_b_vi_int
+    \l_fp_mul_output_int \c_zero
+    \tl_clear:N \l_fp_mul_output_tl
+    \fp_mul_product:NN \l_fp_mul_a_i_int         \l_fp_mul_b_vi_int
+    \fp_mul_product:NN \l_fp_mul_a_ii_int        \l_fp_mul_b_v_int
+    \fp_mul_product:NN \l_fp_mul_a_iii_int       \l_fp_mul_b_iv_int
+    \fp_mul_product:NN \l_fp_mul_a_iv_int        \l_fp_mul_b_iii_int
+    \fp_mul_product:NN \l_fp_mul_a_v_int         \l_fp_mul_b_ii_int
+    \fp_mul_product:NN \l_fp_mul_a_vi_int        \l_fp_mul_b_i_int
+    \tex_divide:D \l_fp_mul_output_int \c_one_thousand
+    \fp_mul_product:NN \l_fp_mul_a_i_int        \l_fp_mul_b_v_int
+    \fp_mul_product:NN \l_fp_mul_a_ii_int       \l_fp_mul_b_iv_int
+    \fp_mul_product:NN \l_fp_mul_a_iii_int      \l_fp_mul_b_iii_int
+    \fp_mul_product:NN \l_fp_mul_a_iv_int       \l_fp_mul_b_ii_int
+    \fp_mul_product:NN \l_fp_mul_a_v_int        \l_fp_mul_b_i_int
+    \fp_mul_end_level:
+    \fp_mul_product:NN \l_fp_mul_a_i_int        \l_fp_mul_b_iv_int
+    \fp_mul_product:NN \l_fp_mul_a_ii_int       \l_fp_mul_b_iii_int
+    \fp_mul_product:NN \l_fp_mul_a_iii_int      \l_fp_mul_b_ii_int
+    \fp_mul_product:NN \l_fp_mul_a_iv_int       \l_fp_mul_b_i_int
+    \fp_mul_end_level:
+    \fp_mul_product:NN \l_fp_mul_a_i_int        \l_fp_mul_b_iii_int
+    \fp_mul_product:NN \l_fp_mul_a_ii_int       \l_fp_mul_b_ii_int
+    \fp_mul_product:NN \l_fp_mul_a_iii_int      \l_fp_mul_b_i_int
+    \fp_mul_end_level:
+    #6 0 \l_fp_mul_output_tl \scan_stop:
+    \tl_clear:N \l_fp_mul_output_tl
+    \fp_mul_product:NN \l_fp_mul_a_i_int        \l_fp_mul_b_ii_int
+    \fp_mul_product:NN \l_fp_mul_a_ii_int       \l_fp_mul_b_i_int
+    \fp_mul_end_level:
+    \fp_mul_product:NN \l_fp_mul_a_i_int        \l_fp_mul_b_i_int
+    \fp_mul_end_level:
+    \fp_mul_end_level:
+    #5 0 \l_fp_mul_output_tl \scan_stop:
+  }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\fp_mul:NNNNNNNNN}
+%   For internal multiplication where the integer does need to be
+%   retained. This means of course that this code is quite slow, and so
+%   is only used when necessary.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \fp_mul:NNNNNNNNN #1#2#3#4#5#6#7#8#9
+  {
+    \fp_mul_split:NNNN #2
+      \l_fp_mul_a_i_int \l_fp_mul_a_ii_int \l_fp_mul_a_iii_int
+    \fp_mul_split:NNNN #3
+      \l_fp_mul_a_iv_int \l_fp_mul_a_v_int \l_fp_mul_a_vi_int
+    \fp_mul_split:NNNN #5
+      \l_fp_mul_b_i_int \l_fp_mul_b_ii_int \l_fp_mul_b_iii_int
+    \fp_mul_split:NNNN #6
+      \l_fp_mul_b_iv_int \l_fp_mul_b_v_int \l_fp_mul_b_vi_int
+    \l_fp_mul_output_int \c_zero
+    \tl_clear:N \l_fp_mul_output_tl
+    \fp_mul_product:NN \l_fp_mul_a_i_int         \l_fp_mul_b_vi_int
+    \fp_mul_product:NN \l_fp_mul_a_ii_int        \l_fp_mul_b_v_int
+    \fp_mul_product:NN \l_fp_mul_a_iii_int       \l_fp_mul_b_iv_int
+    \fp_mul_product:NN \l_fp_mul_a_iv_int        \l_fp_mul_b_iii_int
+    \fp_mul_product:NN \l_fp_mul_a_v_int         \l_fp_mul_b_ii_int
+    \fp_mul_product:NN \l_fp_mul_a_vi_int        \l_fp_mul_b_i_int
+    \tex_divide:D \l_fp_mul_output_int \c_one_thousand
+    \fp_mul_product:NN #1                       \l_fp_mul_b_vi_int
+    \fp_mul_product:NN \l_fp_mul_a_i_int        \l_fp_mul_b_v_int
+    \fp_mul_product:NN \l_fp_mul_a_ii_int       \l_fp_mul_b_iv_int
+    \fp_mul_product:NN \l_fp_mul_a_iii_int      \l_fp_mul_b_iii_int
+    \fp_mul_product:NN \l_fp_mul_a_iv_int       \l_fp_mul_b_ii_int
+    \fp_mul_product:NN \l_fp_mul_a_v_int        \l_fp_mul_b_i_int
+    \fp_mul_product:NN \l_fp_mul_a_vi_int       #4
+    \fp_mul_end_level:
+    \fp_mul_product:NN #1                       \l_fp_mul_b_v_int
+    \fp_mul_product:NN \l_fp_mul_a_i_int        \l_fp_mul_b_iv_int
+    \fp_mul_product:NN \l_fp_mul_a_ii_int       \l_fp_mul_b_iii_int
+    \fp_mul_product:NN \l_fp_mul_a_iii_int      \l_fp_mul_b_ii_int
+    \fp_mul_product:NN \l_fp_mul_a_iv_int       \l_fp_mul_b_i_int
+    \fp_mul_product:NN \l_fp_mul_a_v_int        #4
+    \fp_mul_end_level:
+    \fp_mul_product:NN #1                       \l_fp_mul_b_iv_int
+    \fp_mul_product:NN \l_fp_mul_a_i_int        \l_fp_mul_b_iii_int
+    \fp_mul_product:NN \l_fp_mul_a_ii_int       \l_fp_mul_b_ii_int
+    \fp_mul_product:NN \l_fp_mul_a_iii_int      \l_fp_mul_b_i_int
+    \fp_mul_product:NN \l_fp_mul_a_iv_int       #4
+    \fp_mul_end_level:
+    #9 0 \l_fp_mul_output_tl \scan_stop:
+    \tl_clear:N \l_fp_mul_output_tl
+    \fp_mul_product:NN #1                       \l_fp_mul_b_iii_int
+    \fp_mul_product:NN \l_fp_mul_a_i_int        \l_fp_mul_b_ii_int
+    \fp_mul_product:NN \l_fp_mul_a_ii_int       \l_fp_mul_b_i_int
+    \fp_mul_product:NN \l_fp_mul_a_iii_int      #4
+    \fp_mul_end_level:
+    \fp_mul_product:NN #1                       \l_fp_mul_b_ii_int
+    \fp_mul_product:NN \l_fp_mul_a_i_int        \l_fp_mul_b_i_int
+    \fp_mul_product:NN \l_fp_mul_a_ii_int       #4
+    \fp_mul_end_level:
+    \fp_mul_product:NN #1                       \l_fp_mul_b_i_int
+    \fp_mul_product:NN \l_fp_mul_a_i_int        #4
+    \fp_mul_end_level:
+    #8 0 \l_fp_mul_output_tl \scan_stop:
+    \tl_clear:N \l_fp_mul_output_tl
+    \fp_mul_product:NN #1 #4
+    \fp_mul_end_level:
+    #7 0 \l_fp_mul_output_tl \scan_stop:
+  }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\fp_div_integer:NNNNN}
+%   Here, division is always by an integer, and so it is possible to
+%   use \TeX{}'s native calculations rather than doing it in macros.
+%   The idea here is to divide the decimal part, find any remainder,
+%   then do the real division of the two parts before adding in what
+%   is needed for the remainder.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \fp_div_integer:NNNNN #1#2#3#4#5
+  {
+    \l_fp_tmp_int #1
+    \tex_divide:D \l_fp_tmp_int #3
+    \l_fp_tmp_int \int_eval:w #1 - \l_fp_tmp_int * #3 \int_eval_end:
+    #4 #1
+    \tex_divide:D #4 #3
+    #5 #2
+    \tex_divide:D #5 #3
+    \tex_multiply:D \l_fp_tmp_int \c_one_thousand
+    \tex_divide:D \l_fp_tmp_int #3
+    #5 \int_eval:w #5 + \l_fp_tmp_int * \c_one_million \int_eval_end:
+    \if_int_compare:w #5 > \c_one_thousand_million
+      \tex_advance:D #4 \c_one
+      \tex_advance:D #5 -\c_one_thousand_million
+    \fi:
+  }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\fp_extended_normalise:}
+% \begin{macro}[aux]{\fp_extended_normalise_aux_i:}
+% \begin{macro}[aux]{\fp_extended_normalise_aux_i:w}
+% \begin{macro}[aux]{\fp_extended_normalise_aux_ii:w}
+% \begin{macro}[aux]{\fp_extended_normalise_aux_ii:}
+% \begin{macro}[aux]{\fp_extended_normalise_aux:NNNNNNNNN}
+%   The \enquote{extended} integers for internal use are mainly used in
+%   fixed-point mode. This comes up in a few places, so a generalised
+%   utility is made available to carry out the change. This function
+%   simply calls the two loops to shift the input to the point of
+%   having a zero exponent.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \fp_extended_normalise:
+  {
+    \fp_extended_normalise_aux_i:
+    \fp_extended_normalise_aux_ii:
+  }
+\cs_new_protected_nopar:Npn \fp_extended_normalise_aux_i:
+  {
+    \if_int_compare:w \l_fp_input_a_exponent_int > \c_zero
+      \tex_multiply:D \l_fp_input_a_integer_int \c_ten
+      \tex_advance:D \l_fp_input_a_decimal_int \c_one_thousand_million
+      \exp_after:wN \fp_extended_normalise_aux_i:w
+        \int_use:N \l_fp_input_a_decimal_int \q_stop
+       \exp_after:wN \fp_extended_normalise_aux_i:
+     \fi:
+  }
+\cs_new_protected_nopar:Npn \fp_extended_normalise_aux_i:w
+  #1#2#3#4#5#6#7#8#9 \q_stop
+  {
+    \l_fp_input_a_integer_int
+      \int_eval:w \l_fp_input_a_integer_int + #2 \scan_stop:
+    \l_fp_input_a_decimal_int #3#4#5#6#7#8#9 0 \scan_stop:
+    \tex_advance:D \l_fp_input_a_extended_int \c_one_thousand_million
+    \exp_after:wN \fp_extended_normalise_aux_ii:w
+      \int_use:N \l_fp_input_a_extended_int \q_stop
+  }
+\cs_new_protected_nopar:Npn \fp_extended_normalise_aux_ii:w
+  #1#2#3#4#5#6#7#8#9 \q_stop
+  {
+    \l_fp_input_a_decimal_int
+      \int_eval:w \l_fp_input_a_decimal_int + #2 \scan_stop:
+    \l_fp_input_a_extended_int #3#4#5#6#7#8#9 0 \scan_stop:
+    \tex_advance:D \l_fp_input_a_exponent_int \c_minus_one
+  }
+\cs_new_protected_nopar:Npn \fp_extended_normalise_aux_ii:
+  {
+    \if_int_compare:w \l_fp_input_a_exponent_int < \c_zero
+      \tex_advance:D \l_fp_input_a_decimal_int \c_one_thousand_million
+      \exp_after:wN \use_i:nn \exp_after:wN
+        \fp_extended_normalise_ii_aux:NNNNNNNNN
+        \int_use:N \l_fp_input_a_decimal_int
+       \exp_after:wN \fp_extended_normalise_aux_ii:
+     \fi:
+  }
+\cs_new_protected_nopar:Npn \fp_extended_normalise_ii_aux:NNNNNNNNN
+  #1#2#3#4#5#6#7#8#9
+  {
+    \if_int_compare:w \l_fp_input_a_integer_int = \c_zero
+      \l_fp_input_a_decimal_int #1#2#3#4#5#6#7#8 \scan_stop:
+    \else:
+      \tl_set:Nx \l_fp_tmp_tl
+        {
+          \int_use:N \l_fp_input_a_integer_int
+          #1#2#3#4#5#6#7#8
+        }
+      \l_fp_input_a_integer_int \c_zero
+      \l_fp_input_a_decimal_int \l_fp_tmp_tl \scan_stop:
+    \fi:
+    \tex_divide:D \l_fp_input_a_extended_int \c_ten
+    \tl_set:Nx \l_fp_tmp_tl
+      {
+        #9
+        \int_use:N \l_fp_input_a_extended_int
+      }
+    \l_fp_input_a_extended_int \l_fp_tmp_tl \scan_stop:
+    \tex_advance:D \l_fp_input_a_exponent_int \c_one
+  }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\fp_extended_normalise_output:}
+% \begin{macro}[aux]{\fp_extended_normalise_output_aux_i:NNNNNNNNN}
+% \begin{macro}[aux]{\fp_extended_normalise_output_aux_ii:NNNNNNNNN}
+% \begin{macro}[aux]{\fp_extended_normalise_output_aux:N}
+%   At some stages in working out extended output, it is possible for the
+%   value to need shifting to keep the integer part in range. This only
+%   ever happens such that the integer needs to be made smaller.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \fp_extended_normalise_output:
+  {
+    \if_int_compare:w \l_fp_output_integer_int > \c_nine
+      \tex_advance:D \l_fp_output_integer_int \c_one_thousand_million
+      \exp_after:wN \use_i:nn \exp_after:wN
+        \fp_extended_normalise_output_aux_i:NNNNNNNNN
+        \int_use:N \l_fp_output_integer_int
+      \exp_after:wN \fp_extended_normalise_output:
+    \fi:
+  }
+\cs_new_protected_nopar:Npn \fp_extended_normalise_output_aux_i:NNNNNNNNN
+  #1#2#3#4#5#6#7#8#9
+  {
+    \l_fp_output_integer_int #1#2#3#4#5#6#7#8 \scan_stop:
+    \tex_advance:D \l_fp_output_decimal_int \c_one_thousand_million
+    \tl_set:Nx \l_fp_tmp_tl
+      {
+        #9
+        \exp_after:wN \use_none:n
+        \int_use:N \l_fp_output_decimal_int
+      }
+    \exp_after:wN \fp_extended_normalise_output_aux_ii:NNNNNNNNN
+      \l_fp_tmp_tl
+  }
+\cs_new_protected_nopar:Npn \fp_extended_normalise_output_aux_ii:NNNNNNNNN
+  #1#2#3#4#5#6#7#8#9
+  {
+    \l_fp_output_decimal_int #1#2#3#4#5#6#7#8#9 \scan_stop:
+    \fp_extended_normalise_output_aux:N
+  }
+\cs_new_protected_nopar:Npn \fp_extended_normalise_output_aux:N #1
+  {
+    \tex_advance:D \l_fp_output_extended_int \c_one_thousand_million
+    \tex_divide:D \l_fp_output_extended_int \c_ten
+    \tl_set:Nx \l_fp_tmp_tl
+      {
+        #1
+        \exp_after:wN \use_none:n
+          \int_use:N \l_fp_output_extended_int
+      }
+    \l_fp_output_extended_int \l_fp_tmp_tl \scan_stop:
+    \tex_advance:D \l_fp_output_exponent_int \c_one
+  }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \subsection{Trigonometric functions}
+%
+% \begin{macro}{\fp_trig_normalise:}
+% \begin{macro}[aux]{\fp_trig_normalise_aux:}
+% \begin{macro}[aux]{\fp_trig_sub:NNN}
+%   For normalisation, the code essentially switches to fixed-point
+%   arithmetic. There is a shift of the exponent, then repeated
+%   subtractions. The end result is a number in the range
+%   $ -\pi < x \le \pi $.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \fp_trig_normalise:
+  {
+    \if_int_compare:w \l_fp_input_a_exponent_int < \c_ten
+      \l_fp_input_a_extended_int \c_zero
+      \fp_extended_normalise:
+      \fp_trig_normalise_aux:
+      \if_int_compare:w \l_fp_input_a_integer_int < \c_zero
+        \l_fp_input_a_sign_int -\l_fp_input_a_sign_int
+        \l_fp_input_a_integer_int -\l_fp_input_a_integer_int
+      \fi:
+       \exp_after:wN \fp_trig_octant:
+    \else:
+      \l_fp_input_a_sign_int    \c_one
+      \l_fp_output_integer_int  \c_zero
+      \l_fp_output_decimal_int  \c_zero
+      \l_fp_output_exponent_int \c_zero
+      \exp_after:wN \fp_trig_overflow_msg:
+    \fi:
+  }
+\cs_new_protected_nopar:Npn \fp_trig_normalise_aux:
+  {
+    \if_int_compare:w \l_fp_input_a_integer_int > \c_three
+      \fp_trig_sub:NNN
+        \c_six \c_fp_two_pi_decimal_int \c_fp_two_pi_extended_int
+      \exp_after:wN \fp_trig_normalise_aux:
+    \else:
+      \if_int_compare:w \l_fp_input_a_integer_int > \c_two
+        \if_int_compare:w \l_fp_input_a_decimal_int > \c_fp_pi_decimal_int
+          \fp_trig_sub:NNN
+            \c_six \c_fp_two_pi_decimal_int \c_fp_two_pi_extended_int
+          \exp_after:wN \exp_after:wN \exp_after:wN
+          \exp_after:wN \exp_after:wN \exp_after:wN
+            \exp_after:wN \fp_trig_normalise_aux:
+          \fi:
+      \fi:
+    \fi:
+  }
+%    \end{macrocode}
+%   Here, there may be a sign change but there will never be any
+%   variation in the input. So a dedicated function can be used.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \fp_trig_sub:NNN #1#2#3
+  {
+    \l_fp_input_a_integer_int
+      \int_eval:w \l_fp_input_a_integer_int - #1 \int_eval_end:
+    \l_fp_input_a_decimal_int
+      \int_eval:w \l_fp_input_a_decimal_int - #2 \int_eval_end:
+    \l_fp_input_a_extended_int
+      \int_eval:w \l_fp_input_a_extended_int - #3 \int_eval_end:
+    \if_int_compare:w \l_fp_input_a_extended_int < \c_zero
+      \tex_advance:D \l_fp_input_a_decimal_int \c_minus_one
+      \tex_advance:D \l_fp_input_a_extended_int \c_one_thousand_million
+    \fi:
+    \if_int_compare:w \l_fp_input_a_decimal_int < \c_zero
+      \tex_advance:D \l_fp_input_a_integer_int \c_minus_one
+      \tex_advance:D \l_fp_input_a_decimal_int \c_one_thousand_million
+    \fi:
+    \if_int_compare:w \l_fp_input_a_integer_int < \c_zero
+      \l_fp_input_a_sign_int -\l_fp_input_a_sign_int
+      \if_int_compare:w
+        \int_eval:w
+          \l_fp_input_a_decimal_int + \l_fp_input_a_extended_int
+        = \c_zero
+        \l_fp_input_a_integer_int -\l_fp_input_a_integer_int
+      \else:
+        \l_fp_input_a_integer_int
+           \int_eval:w
+             - \l_fp_input_a_integer_int - \c_one
+           \int_eval_end:
+        \l_fp_input_a_decimal_int
+          \int_eval:w
+            \c_one_thousand_million - \l_fp_input_a_decimal_int
+          \int_eval_end:
+        \l_fp_input_a_extended_int
+          \int_eval:w
+            \c_one_thousand_million - \l_fp_input_a_extended_int
+          \int_eval_end:
+      \fi:
+    \fi:
+  }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\fp_trig_octant:}
+% \begin{macro}[aux]{\fp_trig_octant_aux:}
+%   Here, the input is further reduced into the range
+%   $ 0 \le x < \pi / 4 $. This is pretty simple: check if
+%   $ \pi / 4 $ can be taken off and if it can do it and loop. The
+%   check at the end is to \enquote{mop up} values which are so close to
+%   $ \pi / 4 $ that they should be treated as such.  The test for
+%   an even octant is needed as the `remainder' needed is from
+%   the nearest $ \pi / 2 $.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \fp_trig_octant:
+  {
+    \l_fp_trig_octant_int \c_one
+    \fp_trig_octant_aux:
+    \if_int_compare:w \l_fp_input_a_decimal_int < \c_ten
+      \l_fp_input_a_decimal_int  \c_zero
+      \l_fp_input_a_extended_int \c_zero
+    \fi:
+    \if_int_odd:w \l_fp_trig_octant_int
+    \else:
+      \fp_sub:NNNNNNNNN
+        \c_zero \c_fp_pi_by_four_decimal_int \c_fp_pi_by_four_extended_int
+        \l_fp_input_a_integer_int \l_fp_input_a_decimal_int
+          \l_fp_input_a_extended_int
+        \l_fp_input_a_integer_int \l_fp_input_a_decimal_int
+          \l_fp_input_a_extended_int
+    \fi:
+  }
+\cs_new_protected_nopar:Npn \fp_trig_octant_aux:
+  {
+    \if_int_compare:w \l_fp_input_a_integer_int > \c_zero
+      \fp_sub:NNNNNNNNN
+        \l_fp_input_a_integer_int \l_fp_input_a_decimal_int
+          \l_fp_input_a_extended_int
+        \c_zero \c_fp_pi_by_four_decimal_int \c_fp_pi_by_four_extended_int
+        \l_fp_input_a_integer_int \l_fp_input_a_decimal_int
+          \l_fp_input_a_extended_int
+      \tex_advance:D \l_fp_trig_octant_int \c_one
+      \exp_after:wN \fp_trig_octant_aux:
+    \else:
+      \if_int_compare:w
+        \l_fp_input_a_decimal_int > \c_fp_pi_by_four_decimal_int
+        \fp_sub:NNNNNNNNN
+          \l_fp_input_a_integer_int \l_fp_input_a_decimal_int
+            \l_fp_input_a_extended_int
+          \c_zero \c_fp_pi_by_four_decimal_int
+            \c_fp_pi_by_four_extended_int
+          \l_fp_input_a_integer_int \l_fp_input_a_decimal_int
+            \l_fp_input_a_extended_int
+        \tex_advance:D \l_fp_trig_octant_int \c_one
+        \exp_after:wN \exp_after:wN \exp_after:wN
+          \fp_trig_octant_aux:
+      \fi:
+    \fi:
+  }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\fp_sin:Nn, \fp_sin:cn}
+% \UnitTested
+% \begin{macro}{\fp_gsin:Nn,\fp_gsin:cn}
+% \UnitTested
+% \begin{macro}[aux]{\fp_sin_aux:NNn}
+% \begin{macro}[aux]{\fp_sin_aux_i:}
+% \begin{macro}[aux]{\fp_sin_aux_ii:}
+%   Calculating the sine starts off in the usual way. There is a check
+%   to see if the value has already been worked out before proceeding
+%   further.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \fp_sin:Nn  { \fp_sin_aux:NNn \tl_set:Nn }
+\cs_new_protected_nopar:Npn \fp_gsin:Nn { \fp_sin_aux:NNn \tl_gset:Nn }
+\cs_generate_variant:Nn \fp_sin:Nn   { c }
+\cs_generate_variant:Nn \fp_gsin:Nn  { c }
+%    \end{macrocode}
+%   The internal routine for sines does a check to see if the value is
+%   already known. This saves a lot of repetition when doing rotations.
+%   For very small values it is best to simply return the input as the
+%   sine: the cut-off is $ 1 \times 10^{-5} $.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \fp_sin_aux:NNn #1#2#3
+  {
+    \group_begin:
+      \fp_split:Nn a {#3}
+      \fp_standardise:NNNN
+        \l_fp_input_a_sign_int
+        \l_fp_input_a_integer_int
+        \l_fp_input_a_decimal_int
+        \l_fp_input_a_exponent_int
+      \tl_set:Nx \l_fp_arg_tl
+        {
+          \if_int_compare:w \l_fp_input_a_sign_int < \c_zero
+            -
+          \else:
+            +
+          \fi:
+          \int_use:N \l_fp_input_a_integer_int
+          .
+          \exp_after:wN \use_none:n
+            \int_value:w \int_eval:w
+              \l_fp_input_a_decimal_int + \c_one_thousand_million
+          e
+          \int_use:N \l_fp_input_a_exponent_int
+        }
+      \if_int_compare:w \l_fp_input_a_exponent_int < -\c_five
+        \cs_set_protected_nopar:Npx \fp_tmp:w
+        {
+          \group_end:
+          #1 \exp_not:N #2 { \l_fp_arg_tl }
+        }
+      \else:
+        \if_cs_exist:w
+          c_fp_sin ( \l_fp_arg_tl ) _fp
+        \cs_end:
+        \else:
+          \exp_after:wN \exp_after:wN \exp_after:wN
+            \fp_sin_aux_i:
+        \fi:
+        \cs_set_protected_nopar:Npx \fp_tmp:w
+          {
+            \group_end:
+            #1 \exp_not:N #2
+              { \use:c { c_fp_sin ( \l_fp_arg_tl ) _fp } }
+          }
+      \fi:
+    \fp_tmp:w
+  }
+%    \end{macrocode}
+%   The internals for sine first normalise the input into an octant, then
+%   choose the correct set up for the Taylor series. The sign for the sine
+%   function is easy, so there is no worry about it. So the only thing to
+%   do is to get the output standardised.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \fp_sin_aux_i:
+  {
+    \fp_trig_normalise:
+    \fp_sin_aux_ii:
+    \if_int_compare:w \l_fp_output_integer_int = \c_one
+      \l_fp_output_exponent_int \c_zero
+    \else:
+      \l_fp_output_integer_int \l_fp_output_decimal_int
+      \l_fp_output_decimal_int \l_fp_output_extended_int
+      \l_fp_output_exponent_int -\c_nine
+    \fi:
+    \fp_standardise:NNNN
+      \l_fp_input_a_sign_int
+      \l_fp_output_integer_int
+      \l_fp_output_decimal_int
+      \l_fp_output_exponent_int
+    \tl_new:c { c_fp_sin ( \l_fp_arg_tl ) _fp }
+    \tl_gset:cx { c_fp_sin ( \l_fp_arg_tl ) _fp }
+      {
+        \if_int_compare:w \l_fp_input_a_sign_int > \c_zero
+          +
+        \else:
+          -
+        \fi:
+        \int_use:N \l_fp_output_integer_int
+        .
+        \exp_after:wN \use_none:n
+          \int_value:w \int_eval:w
+             \l_fp_output_decimal_int + \c_one_thousand_million
+        e
+        \int_use:N \l_fp_output_exponent_int
+      }
+  }
+\cs_new_protected_nopar:Npn \fp_sin_aux_ii:
+  {
+    \if_case:w \l_fp_trig_octant_int
+    \or:
+      \exp_after:wN \fp_trig_calc_sin:
+    \or:
+      \exp_after:wN \fp_trig_calc_cos:
+    \or:
+      \exp_after:wN \fp_trig_calc_cos:
+    \or:
+      \exp_after:wN \fp_trig_calc_sin:
+    \fi:
+  }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\fp_cos:Nn, \fp_cos:cn}
+% \UnitTested
+% \begin{macro}{\fp_gcos:Nn,\fp_gcos:cn}
+% \UnitTested
+% \begin{macro}[aux]{\fp_cos_aux:NNn}
+% \begin{macro}[aux]{\fp_cos_aux_i:}
+% \begin{macro}[aux]{\fp_cos_aux_ii:}
+%   Cosine is almost identical, but there is no short cut code here.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \fp_cos:Nn  { \fp_cos_aux:NNn \tl_set:Nn }
+\cs_new_protected_nopar:Npn \fp_gcos:Nn { \fp_cos_aux:NNn \tl_gset:Nn }
+\cs_generate_variant:Nn \fp_cos:Nn   { c }
+\cs_generate_variant:Nn \fp_gcos:Nn  { c }
+\cs_new_protected_nopar:Npn \fp_cos_aux:NNn #1#2#3
+  {
+    \group_begin:
+      \fp_split:Nn a {#3}
+      \fp_standardise:NNNN
+        \l_fp_input_a_sign_int
+        \l_fp_input_a_integer_int
+        \l_fp_input_a_decimal_int
+        \l_fp_input_a_exponent_int
+      \tl_set:Nx \l_fp_arg_tl
+        {
+          \if_int_compare:w \l_fp_input_a_sign_int < \c_zero
+            -
+          \else:
+            +
+          \fi:
+          \int_use:N \l_fp_input_a_integer_int
+          .
+        \exp_after:wN \use_none:n
+          \int_value:w \int_eval:w
+             \l_fp_input_a_decimal_int + \c_one_thousand_million
+          e
+          \int_use:N \l_fp_input_a_exponent_int
+        }
+      \if_cs_exist:w c_fp_cos ( \l_fp_arg_tl ) _fp \cs_end:
+      \else:
+        \exp_after:wN \fp_cos_aux_i:
+      \fi:
+      \cs_set_protected_nopar:Npx \fp_tmp:w
+        {
+          \group_end:
+          #1 \exp_not:N #2
+            { \use:c { c_fp_cos ( \l_fp_arg_tl ) _fp } }
+        }
+    \fp_tmp:w
+  }
+%    \end{macrocode}
+% Almost the same as for sine: just a bit of correction for the sign
+% of the output.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \fp_cos_aux_i:
+  {
+    \fp_trig_normalise:
+    \fp_cos_aux_ii:
+    \if_int_compare:w \l_fp_output_integer_int = \c_one
+      \l_fp_output_exponent_int \c_zero
+    \else:
+      \l_fp_output_integer_int \l_fp_output_decimal_int
+      \l_fp_output_decimal_int \l_fp_output_extended_int
+      \l_fp_output_exponent_int -\c_nine
+    \fi:
+    \fp_standardise:NNNN
+      \l_fp_input_a_sign_int
+      \l_fp_output_integer_int
+      \l_fp_output_decimal_int
+      \l_fp_output_exponent_int
+    \tl_new:c { c_fp_cos ( \l_fp_arg_tl ) _fp }
+    \tl_gset:cx { c_fp_cos ( \l_fp_arg_tl ) _fp }
+      {
+        \if_int_compare:w \l_fp_input_a_sign_int > \c_zero
+          +
+        \else:
+          -
+        \fi:
+        \int_use:N \l_fp_output_integer_int
+        .
+        \exp_after:wN \use_none:n
+          \int_value:w \int_eval:w
+             \l_fp_output_decimal_int + \c_one_thousand_million
+        e
+        \int_use:N \l_fp_output_exponent_int
+      }
+  }
+\cs_new_protected_nopar:Npn \fp_cos_aux_ii:
+  {
+    \if_case:w \l_fp_trig_octant_int
+    \or:
+      \exp_after:wN \fp_trig_calc_cos:
+    \or:
+      \exp_after:wN \fp_trig_calc_sin:
+    \or:
+      \exp_after:wN \fp_trig_calc_sin:
+    \or:
+      \exp_after:wN \fp_trig_calc_cos:
+    \fi:
+    \if_int_compare:w \l_fp_input_a_sign_int > \c_zero
+      \if_int_compare:w \l_fp_trig_octant_int > \c_two
+        \l_fp_input_a_sign_int \c_minus_one
+      \fi:
+    \else:
+      \if_int_compare:w \l_fp_trig_octant_int > \c_two
+      \else:
+        \l_fp_input_a_sign_int \c_one
+      \fi:
+    \fi:
+  }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\fp_trig_calc_cos:}
+% \begin{macro}{\fp_trig_calc_sin:}
+% \begin{macro}[aux]{\fp_trig_calc_Taylor:}
+%   These functions actually do the calculation for sine and cosine.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \fp_trig_calc_cos:
+  {
+    \if_int_compare:w \l_fp_input_a_decimal_int = \c_zero
+      \l_fp_output_integer_int \c_one
+      \l_fp_output_decimal_int \c_zero
+    \else:
+      \l_fp_trig_sign_int \c_minus_one
+      \fp_mul:NNNNNN
+        \l_fp_input_a_decimal_int \l_fp_input_a_extended_int
+        \l_fp_input_a_decimal_int \l_fp_input_a_extended_int
+        \l_fp_trig_decimal_int \l_fp_trig_extended_int
+      \fp_div_integer:NNNNN
+        \l_fp_trig_decimal_int \l_fp_trig_extended_int
+        \c_two
+        \l_fp_trig_decimal_int \l_fp_trig_extended_int
+      \l_fp_count_int \c_three
+      \if_int_compare:w \l_fp_trig_extended_int = \c_zero
+        \if_int_compare:w \l_fp_trig_decimal_int = \c_zero
+          \l_fp_output_integer_int \c_one
+          \l_fp_output_decimal_int \c_zero
+          \l_fp_output_extended_int \c_zero
+        \else:
+          \l_fp_output_integer_int \c_zero
+          \l_fp_output_decimal_int \c_one_thousand_million
+          \l_fp_output_extended_int \c_zero
+        \fi:
+      \else:
+        \l_fp_output_integer_int \c_zero
+        \l_fp_output_decimal_int 999999999 \scan_stop:
+        \l_fp_output_extended_int \c_one_thousand_million
+      \fi:
+      \tex_advance:D \l_fp_output_extended_int -\l_fp_trig_extended_int
+      \tex_advance:D \l_fp_output_decimal_int -\l_fp_trig_decimal_int
+      \exp_after:wN \fp_trig_calc_Taylor:
+    \fi:
+  }
+\cs_new_protected_nopar:Npn \fp_trig_calc_sin:
+  {
+    \l_fp_output_integer_int \c_zero
+    \if_int_compare:w \l_fp_input_a_decimal_int = \c_zero
+      \l_fp_output_decimal_int \c_zero
+    \else:
+      \l_fp_output_decimal_int \l_fp_input_a_decimal_int
+      \l_fp_output_extended_int \l_fp_input_a_extended_int
+      \l_fp_trig_sign_int \c_one
+      \l_fp_trig_decimal_int \l_fp_input_a_decimal_int
+      \l_fp_trig_extended_int \l_fp_input_a_extended_int
+      \l_fp_count_int \c_two
+      \exp_after:wN \fp_trig_calc_Taylor:
+    \fi:
+  }
+%    \end{macrocode}
+%   This implements a Taylor series calculation for the trigonometric
+%   functions. Lots of shuffling about as \TeX\ is not exactly a natural
+%   choice for this sort of thing.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \fp_trig_calc_Taylor:
+  {
+    \l_fp_trig_sign_int -\l_fp_trig_sign_int
+    \fp_mul:NNNNNN
+      \l_fp_trig_decimal_int \l_fp_trig_extended_int
+      \l_fp_input_a_decimal_int \l_fp_input_a_extended_int
+      \l_fp_trig_decimal_int \l_fp_trig_extended_int
+    \fp_mul:NNNNNN
+      \l_fp_trig_decimal_int \l_fp_trig_extended_int
+      \l_fp_input_a_decimal_int \l_fp_input_a_extended_int
+      \l_fp_trig_decimal_int \l_fp_trig_extended_int
+    \fp_div_integer:NNNNN
+      \l_fp_trig_decimal_int \l_fp_trig_extended_int
+      \l_fp_count_int
+      \l_fp_trig_decimal_int \l_fp_trig_extended_int
+    \tex_advance:D \l_fp_count_int \c_one
+    \fp_div_integer:NNNNN
+      \l_fp_trig_decimal_int \l_fp_trig_extended_int
+      \l_fp_count_int
+      \l_fp_trig_decimal_int \l_fp_trig_extended_int
+    \tex_advance:D \l_fp_count_int \c_one
+    \if_int_compare:w \l_fp_trig_decimal_int > \c_zero
+      \if_int_compare:w \l_fp_trig_sign_int > \c_zero
+        \tex_advance:D \l_fp_output_decimal_int \l_fp_trig_decimal_int
+        \tex_advance:D \l_fp_output_extended_int
+          \l_fp_trig_extended_int
+        \if_int_compare:w \l_fp_output_extended_int < \c_one_thousand_million
+        \else:
+          \tex_advance:D \l_fp_output_decimal_int \c_one
+          \tex_advance:D \l_fp_output_extended_int
+            -\c_one_thousand_million
+        \fi:
+        \if_int_compare:w \l_fp_output_decimal_int < \c_one_thousand_million
+        \else:
+          \tex_advance:D \l_fp_output_integer_int \c_one
+          \tex_advance:D \l_fp_output_decimal_int
+            -\c_one_thousand_million
+        \fi:
+      \else:
+        \tex_advance:D \l_fp_output_decimal_int -\l_fp_trig_decimal_int
+        \tex_advance:D \l_fp_output_extended_int
+          -\l_fp_input_a_extended_int
+        \if_int_compare:w \l_fp_output_extended_int < \c_zero
+          \tex_advance:D \l_fp_output_decimal_int \c_minus_one
+          \tex_advance:D \l_fp_output_extended_int \c_one_thousand_million
+        \fi:
+        \if_int_compare:w \l_fp_output_decimal_int < \c_zero
+          \tex_advance:D \l_fp_output_integer_int \c_minus_one
+          \tex_advance:D \l_fp_output_decimal_int \c_one_thousand_million
+        \fi:
+      \fi:
+      \exp_after:wN \fp_trig_calc_Taylor:
+    \fi:
+  }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\fp_tan:Nn, \fp_tan:cn}
+% \UnitTested
+% \begin{macro}{\fp_gtan:Nn,\fp_gtan:cn}
+% \UnitTested
+% \begin{macro}[aux]{\fp_tan_aux:NNn}
+% \begin{macro}[aux]{\fp_tan_aux_i:}
+% \begin{macro}[aux]{\fp_tan_aux_ii:}
+% \begin{macro}[aux]{\fp_tan_aux_iii:}
+% \begin{macro}[aux]{\fp_tan_aux_iv:}
+%   As might be expected, tangents are calculated from the sine and cosine
+%   by division. So there is a bit of set up, the two subsidiary pieces
+%   of work are done and then a division takes place. For small numbers,
+%   the same approach is used as for sines, with the input value simply
+%   returned as is.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \fp_tan:Nn  { \fp_tan_aux:NNn \tl_set:Nn }
+\cs_new_protected_nopar:Npn \fp_gtan:Nn { \fp_tan_aux:NNn \tl_gset:Nn }
+\cs_generate_variant:Nn \fp_tan:Nn   { c }
+\cs_generate_variant:Nn \fp_gtan:Nn  { c }
+\cs_new_protected_nopar:Npn \fp_tan_aux:NNn #1#2#3
+  {
+    \group_begin:
+      \fp_split:Nn a {#3}
+      \fp_standardise:NNNN
+        \l_fp_input_a_sign_int
+        \l_fp_input_a_integer_int
+        \l_fp_input_a_decimal_int
+        \l_fp_input_a_exponent_int
+      \tl_set:Nx \l_fp_arg_tl
+        {
+          \if_int_compare:w \l_fp_input_a_sign_int < \c_zero
+            -
+          \else:
+            +
+          \fi:
+          \int_use:N \l_fp_input_a_integer_int
+          .
+          \exp_after:wN \use_none:n
+            \int_value:w \int_eval:w
+              \l_fp_input_a_decimal_int + \c_one_thousand_million
+          e
+          \int_use:N \l_fp_input_a_exponent_int
+        }
+      \if_int_compare:w \l_fp_input_a_exponent_int < -\c_five
+        \cs_set_protected_nopar:Npx \fp_tmp:w
+        {
+          \group_end:
+          #1 \exp_not:N #2 { \l_fp_arg_tl }
+        }
+      \else:
+        \if_cs_exist:w
+          c_fp_tan ( \l_fp_arg_tl ) _fp
+        \cs_end:
+        \else:
+          \exp_after:wN \exp_after:wN \exp_after:wN
+            \fp_tan_aux_i:
+        \fi:
+        \cs_set_protected_nopar:Npx \fp_tmp:w
+          {
+            \group_end:
+            #1 \exp_not:N #2
+              { \use:c { c_fp_tan ( \l_fp_arg_tl ) _fp } }
+          }
+      \fi:
+    \fp_tmp:w
+  }
+%    \end{macrocode}
+%   The business of the calculation does not check for stored sines or
+%   cosines as there would then be an overhead to reading them back in.
+%   There is also no need to worry about \enquote{small} sine values as
+%   these will have been dealt with earlier. There is a two-step lead off
+%   so that undefined division is not even attempted.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \fp_tan_aux_i:
+  {
+    \if_int_compare:w \l_fp_input_a_exponent_int < \c_ten
+      \exp_after:wN \fp_tan_aux_ii:
+    \else:
+      \cs_new_eq:cN { c_fp_tan ( \l_fp_arg_tl ) _fp }
+        \c_zero_fp
+      \exp_after:wN \fp_trig_overflow_msg:
+    \fi:
+  }
+\cs_new_protected_nopar:Npn \fp_tan_aux_ii:
+  {
+    \fp_trig_normalise:
+    \if_int_compare:w \l_fp_input_a_sign_int > \c_zero
+      \if_int_compare:w \l_fp_trig_octant_int > \c_two
+        \l_fp_output_sign_int \c_minus_one
+      \else:
+        \l_fp_output_sign_int \c_one
+      \fi:
+    \else:
+      \if_int_compare:w \l_fp_trig_octant_int > \c_two
+        \l_fp_output_sign_int \c_one
+      \else:
+        \l_fp_output_sign_int \c_minus_one
+      \fi:
+    \fi:
+    \fp_cos_aux_ii:
+    \if_int_compare:w \l_fp_input_a_decimal_int = \c_zero
+      \if_int_compare:w \l_fp_input_a_integer_int = \c_zero
+        \cs_new_eq:cN { c_fp_tan ( \l_fp_arg_tl ) _fp }
+          \c_undefined_fp
+      \else:
+        \exp_after:wN \exp_after:wN \exp_after:wN
+          \fp_tan_aux_iii:
+      \fi:
+    \else:
+      \exp_after:wN \fp_tan_aux_iii:
+    \fi:
+  }
+%    \end{macrocode}
+%   The division is done here using the same code as the standard division
+%   unit, shifting the digits in the calculated sine and cosine to
+%   maintain accuracy.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \fp_tan_aux_iii:
+  {
+    \l_fp_input_b_integer_int \l_fp_output_decimal_int
+    \l_fp_input_b_decimal_int \l_fp_output_extended_int
+    \l_fp_input_b_exponent_int -\c_nine
+    \fp_standardise:NNNN
+      \l_fp_input_b_sign_int
+      \l_fp_input_b_integer_int
+      \l_fp_input_b_decimal_int
+      \l_fp_input_b_exponent_int
+    \fp_sin_aux_ii:
+    \l_fp_input_a_integer_int \l_fp_output_decimal_int
+    \l_fp_input_a_decimal_int \l_fp_output_extended_int
+    \l_fp_input_a_exponent_int -\c_nine
+    \fp_standardise:NNNN
+      \l_fp_input_a_sign_int
+      \l_fp_input_a_integer_int
+      \l_fp_input_a_decimal_int
+      \l_fp_input_a_exponent_int
+    \if_int_compare:w \l_fp_input_a_decimal_int = \c_zero
+      \if_int_compare:w \l_fp_input_a_integer_int = \c_zero
+        \cs_new_eq:cN { c_fp_tan ( \l_fp_arg_tl ) _fp }
+          \c_zero_fp
+      \else:
+        \exp_after:wN \exp_after:wN \exp_after:wN \fp_tan_aux_iv:
+      \fi:
+    \else:
+      \exp_after:wN \fp_tan_aux_iv:
+    \fi:
+  }
+\cs_new_protected_nopar:Npn \fp_tan_aux_iv:
+  {
+    \l_fp_output_integer_int \c_zero
+    \l_fp_output_decimal_int \c_zero
+    \cs_set_eq:NN \fp_div_store: \fp_div_store_integer:
+    \l_fp_div_offset_int \c_one_hundred_million
+    \fp_div_loop:
+    \l_fp_output_exponent_int
+      \int_eval:w
+        \l_fp_input_a_exponent_int - \l_fp_input_b_exponent_int
+      \int_eval_end:
+    \fp_standardise:NNNN
+      \l_fp_output_sign_int
+      \l_fp_output_integer_int
+      \l_fp_output_decimal_int
+      \l_fp_output_exponent_int
+    \tl_new:c { c_fp_tan ( \l_fp_arg_tl ) _fp }
+    \tl_gset:cx { c_fp_tan ( \l_fp_arg_tl ) _fp }
+      {
+        \if_int_compare:w \l_fp_output_sign_int > \c_zero
+          +
+        \else:
+          -
+        \fi:
+        \int_use:N \l_fp_output_integer_int
+        .
+        \exp_after:wN \use_none:n
+          \int_value:w \int_eval:w
+             \l_fp_output_decimal_int + \c_one_thousand_million
+        e
+        \int_use:N \l_fp_output_exponent_int
+      }
+  }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \subsection{Exponent and logarithm functions}
+%
+% \begin{variable}{\c_fp_exp_1_tl}
+% \begin{variable}{\c_fp_exp_2_tl}
+% \begin{variable}{\c_fp_exp_3_tl}
+% \begin{variable}{\c_fp_exp_4_tl}
+% \begin{variable}{\c_fp_exp_5_tl}
+% \begin{variable}{\c_fp_exp_6_tl}
+% \begin{variable}{\c_fp_exp_7_tl}
+% \begin{variable}{\c_fp_exp_8_tl}
+% \begin{variable}{\c_fp_exp_9_tl}
+% \begin{variable}{\c_fp_exp_10_tl}
+% \begin{variable}{\c_fp_exp_20_tl}
+% \begin{variable}{\c_fp_exp_30_tl}
+% \begin{variable}{\c_fp_exp_40_tl}
+% \begin{variable}{\c_fp_exp_50_tl}
+% \begin{variable}{\c_fp_exp_60_tl}
+% \begin{variable}{\c_fp_exp_70_tl}
+% \begin{variable}{\c_fp_exp_80_tl}
+% \begin{variable}{\c_fp_exp_90_tl}
+% \begin{variable}{\c_fp_exp_100_tl}
+% \begin{variable}{\c_fp_exp_200_tl}
+%   Calculation of exponentials requires a number of precomputed values:
+%   first the positive integers.
+%    \begin{macrocode}
+\tl_const:cn { c_fp_exp_1_tl }   { { 2 } { 718281828 } { 459045235 } { 0 } }
+\tl_const:cn { c_fp_exp_2_tl }   { { 7 } { 389056098 } { 930650227 } { 0 } }
+\tl_const:cn { c_fp_exp_3_tl }   { { 2 } { 008553692 } { 318766774 } { 1 } }
+\tl_const:cn { c_fp_exp_4_tl }   { { 5 } { 459815003 } { 314423908 } { 1 } }
+\tl_const:cn { c_fp_exp_5_tl }   { { 1 } { 484131591 } { 025766034 } { 2 } }
+\tl_const:cn { c_fp_exp_6_tl }   { { 4 } { 034287934 } { 927351226 } { 2 } }
+\tl_const:cn { c_fp_exp_7_tl }   { { 1 } { 096633158 } { 428458599 } { 3 } }
+\tl_const:cn { c_fp_exp_8_tl }   { { 2 } { 980957987 } { 041728275 } { 3 } }
+\tl_const:cn { c_fp_exp_9_tl }   { { 8 } { 103083927 } { 575384008 } { 3 } }
+\tl_const:cn { c_fp_exp_10_tl }  { { 2 } { 202646579 } { 480671652 } { 4 } }
+\tl_const:cn { c_fp_exp_20_tl }  { { 4 } { 851651954 } { 097902280 } { 8 } }
+\tl_const:cn { c_fp_exp_30_tl }  { { 1 } { 068647458 } { 152446215 } { 13 } }
+\tl_const:cn { c_fp_exp_40_tl }  { { 2 } { 353852668 } { 370199854 } { 17 } }
+\tl_const:cn { c_fp_exp_50_tl }  { { 5 } { 184705528 } { 587072464 } { 21 } }
+\tl_const:cn { c_fp_exp_60_tl }  { { 1 } { 142007389 } { 815684284 } { 26 } }
+\tl_const:cn { c_fp_exp_70_tl }  { { 2 } { 515438670 } { 919167006 } { 30 } }
+\tl_const:cn { c_fp_exp_80_tl }  { { 5 } { 540622384 } { 393510053 } { 34 } }
+\tl_const:cn { c_fp_exp_90_tl }  { { 1 } { 220403294 } { 317840802 } { 39 } }
+\tl_const:cn { c_fp_exp_100_tl } { { 2 } { 688117141 } { 816135448 } { 43 } }
+\tl_const:cn { c_fp_exp_200_tl } { { 7 } { 225973768 } { 125749258 } { 86 } }
+%    \end{macrocode}
+% \end{variable}
+% \end{variable}
+% \end{variable}
+% \end{variable}
+% \end{variable}
+% \end{variable}
+% \end{variable}
+% \end{variable}
+% \end{variable}
+% \end{variable}
+% \end{variable}
+% \end{variable}
+% \end{variable}
+% \end{variable}
+% \end{variable}
+% \end{variable}
+% \end{variable}
+% \end{variable}
+% \end{variable}
+% \end{variable}
+%
+% \begin{variable}{\c_fp_exp_-1_tl}
+% \begin{variable}{\c_fp_exp_-2_tl}
+% \begin{variable}{\c_fp_exp_-3_tl}
+% \begin{variable}{\c_fp_exp_-4_tl}
+% \begin{variable}{\c_fp_exp_-5_tl}
+% \begin{variable}{\c_fp_exp_-6_tl}
+% \begin{variable}{\c_fp_exp_-7_tl}
+% \begin{variable}{\c_fp_exp_-8_tl}
+% \begin{variable}{\c_fp_exp_-9_tl}
+% \begin{variable}{\c_fp_exp_-10_tl}
+% \begin{variable}{\c_fp_exp_-20_tl}
+% \begin{variable}{\c_fp_exp_-30_tl}
+% \begin{variable}{\c_fp_exp_-40_tl}
+% \begin{variable}{\c_fp_exp_-50_tl}
+% \begin{variable}{\c_fp_exp_-60_tl}
+% \begin{variable}{\c_fp_exp_-70_tl}
+% \begin{variable}{\c_fp_exp_-80_tl}
+% \begin{variable}{\c_fp_exp_-90_tl}
+% \begin{variable}{\c_fp_exp_-100_tl}
+% \begin{variable}{\c_fp_exp_-200_tl}
+% Now the negative integers.
+%    \begin{macrocode}
+\tl_const:cn { c_fp_exp_-1_tl }   { { 3 } { 678794411 } { 71442322 }  { -1 } }
+\tl_const:cn { c_fp_exp_-2_tl }   { { 1 } { 353352832 } { 366132692 } { -1 } }
+\tl_const:cn { c_fp_exp_-3_tl }   { { 4 } { 978706836 } { 786394298 } { -2 } }
+\tl_const:cn { c_fp_exp_-4_tl }   { { 1 } { 831563888 } { 873418029 } { -2 } }
+\tl_const:cn { c_fp_exp_-5_tl }   { { 6 } { 737946999 } { 085467097 } { -3 } }
+\tl_const:cn { c_fp_exp_-6_tl }   { { 2 } { 478752176 } { 666358423 } { -3 } }
+\tl_const:cn { c_fp_exp_-7_tl }   { { 9 } { 118819655 } { 545162080 } { -4 } }
+\tl_const:cn { c_fp_exp_-8_tl }   { { 3 } { 354626279 } { 025118388 } { -4 } }
+\tl_const:cn { c_fp_exp_-9_tl }   { { 1 } { 234098040 } { 866795495 } { -4 } }
+\tl_const:cn { c_fp_exp_-10_tl }  { { 4 } { 539992976 } { 248451536 } { -5 } }
+\tl_const:cn { c_fp_exp_-20_tl }  { { 2 } { 061153622 } { 438557828 } { -9 } }
+\tl_const:cn { c_fp_exp_-30_tl }  { { 9 } { 357622968 } { 840174605 } { -14 } }
+\tl_const:cn { c_fp_exp_-40_tl }  { { 4 } { 248354255 } { 291588995 } { -18 } }
+\tl_const:cn { c_fp_exp_-50_tl }  { { 1 } { 928749847 } { 963917783 } { -22 } }
+\tl_const:cn { c_fp_exp_-60_tl }  { { 8 } { 756510762 } { 696520338 } { -27 } }
+\tl_const:cn { c_fp_exp_-70_tl }  { { 3 } { 975449735 } { 908646808 } { -31 } }
+\tl_const:cn { c_fp_exp_-80_tl }  { { 1 } { 804851387 } { 845415172 } { -35 } }
+\tl_const:cn { c_fp_exp_-90_tl }  { { 8 } { 194012623 } { 990515430 } { -40 } }
+\tl_const:cn { c_fp_exp_-100_tl } { { 3 } { 720075976 } { 020835963 } { -44 } }
+\tl_const:cn { c_fp_exp_-200_tl } { { 1 } { 383896526 } { 736737530 } { -87 } }
+%    \end{macrocode}
+% \end{variable}
+% \end{variable}
+% \end{variable}
+% \end{variable}
+% \end{variable}
+% \end{variable}
+% \end{variable}
+% \end{variable}
+% \end{variable}
+% \end{variable}
+% \end{variable}
+% \end{variable}
+% \end{variable}
+% \end{variable}
+% \end{variable}
+% \end{variable}
+% \end{variable}
+% \end{variable}
+% \end{variable}
+% \end{variable}
+%
+% \begin{macro}{\fp_exp:Nn, \fp_exp:cn}
+% \UnitTested
+% \begin{macro}{\fp_gexp:Nn,\fp_gexp:cn}
+% \UnitTested
+% \begin{macro}[aux]{\fp_exp_aux:NNn}
+% \begin{macro}[aux]{\fp_exp_internal:}
+% \begin{macro}[aux]{\fp_exp_aux:}
+% \begin{macro}[aux]{\fp_exp_integer:}
+% \begin{macro}[aux]{\fp_exp_integer_tens:}
+% \begin{macro}[aux]{\fp_exp_integer_units:}
+% \begin{macro}[aux]{\fp_exp_integer_const:n}
+% \begin{macro}[aux]{\fp_exp_integer_const:nnnn}
+% \begin{macro}[aux]{\fp_exp_decimal:}
+% \begin{macro}[aux]{\fp_exp_Taylor:}
+% \begin{macro}[aux]{\fp_exp_const:Nx}
+% \begin{macro}[aux]{\fp_exp_const:cx}
+%   The calculation of an exponent starts off starts in much the same
+%   way as the trigonometric functions: normalise the input, look for
+%   a pre-defined value and if one is not found hand off to the real
+%   workhorse function. The test for a definition of the result is used
+%   so that overflows do not result in any outcome being defined.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \fp_exp:Nn  { \fp_exp_aux:NNn \tl_set:Nn }
+\cs_new_protected_nopar:Npn \fp_gexp:Nn { \fp_exp_aux:NNn \tl_gset:Nn }
+\cs_generate_variant:Nn \fp_exp:Nn  { c }
+\cs_generate_variant:Nn \fp_gexp:Nn { c }
+\cs_new_protected_nopar:Npn \fp_exp_aux:NNn #1#2#3
+  {
+    \group_begin:
+      \fp_split:Nn a {#3}
+      \fp_standardise:NNNN
+        \l_fp_input_a_sign_int
+        \l_fp_input_a_integer_int
+        \l_fp_input_a_decimal_int
+        \l_fp_input_a_exponent_int
+      \l_fp_input_a_extended_int \c_zero
+      \tl_set:Nx \l_fp_arg_tl
+        {
+          \if_int_compare:w \l_fp_input_a_sign_int < \c_zero
+            -
+          \else:
+            +
+          \fi:
+          \int_use:N \l_fp_input_a_integer_int
+          .
+          \exp_after:wN \use_none:n
+            \int_value:w \int_eval:w
+               \l_fp_input_a_decimal_int + \c_one_thousand_million
+          e
+          \int_use:N \l_fp_input_a_exponent_int
+        }
+      \if_cs_exist:w c_fp_exp ( \l_fp_arg_tl ) _fp \cs_end:
+      \else:
+        \exp_after:wN \fp_exp_internal:
+      \fi:
+      \cs_set_protected_nopar:Npx \fp_tmp:w
+        {
+          \group_end:
+          #1 \exp_not:N #2
+            {
+              \if_cs_exist:w c_fp_exp ( \l_fp_arg_tl ) _fp
+                \cs_end:
+                \use:c { c_fp_exp ( \l_fp_arg_tl ) _fp }
+              \else:
+                \c_zero_fp
+              \fi:
+            }
+        }
+    \fp_tmp:w
+  }
+%    \end{macrocode}
+%   The first real step is to convert the input into a fixed-point
+%   representation for further calculation: anything which is dropped
+%   here as too small would not influence the output in any case. There
+%   are a couple of overflow tests: the maximum
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \fp_exp_internal:
+  {
+    \if_int_compare:w \l_fp_input_a_exponent_int < \c_three
+      \fp_extended_normalise:
+      \if_int_compare:w \l_fp_input_a_sign_int > \c_zero
+        \if_int_compare:w \l_fp_input_a_integer_int < 230 \scan_stop:
+          \exp_after:wN \exp_after:wN \exp_after:wN
+          \exp_after:wN \exp_after:wN \exp_after:wN
+            \exp_after:wN \fp_exp_aux:
+        \else:
+          \exp_after:wN \exp_after:wN \exp_after:wN
+          \exp_after:wN \exp_after:wN \exp_after:wN
+            \exp_after:wN \fp_exp_overflow_msg:
+          \fi:
+      \else:
+        \if_int_compare:w \l_fp_input_a_integer_int < 230 \scan_stop:
+          \exp_after:wN \exp_after:wN \exp_after:wN
+          \exp_after:wN \exp_after:wN \exp_after:wN
+            \exp_after:wN  \fp_exp_aux:
+        \else:
+          \fp_exp_const:cx { c_fp_exp ( \l_fp_arg_tl ) _fp }
+            { \c_zero_fp }
+        \fi:
+      \fi:
+    \else:
+      \exp_after:wN \fp_exp_overflow_msg:
+    \fi:
+  }
+%    \end{macrocode}
+%   The main algorithm makes use of the fact that
+%   \[
+%     \mathrm{e}^{nmp.q} =
+%       \mathrm{e}^{n}
+%       \mathrm{e}^{m}
+%       \mathrm{e}^{p}
+%       \mathrm{e}^{0.q}
+%   \]
+%   and that there is a Taylor series that can be used to calculate
+%   $ \mathrm{e}^{0.q} $. Thus the approach needed is in three parts.
+%   First, the exponent of the integer part of the input is found
+%   using the pre-calculated constants. Second, the Taylor series is
+%   used to find the exponent for the decimal part of the input. Finally,
+%   the two parts are multiplied together to give the result. As the
+%   normalisation code will already have dealt with any overflowing
+%   values, there are no further checks needed.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \fp_exp_aux:
+  {
+    \if_int_compare:w \l_fp_input_a_integer_int > \c_zero
+      \exp_after:wN \fp_exp_integer:
+    \else:
+      \l_fp_output_integer_int  \c_one
+      \l_fp_output_decimal_int  \c_zero
+      \l_fp_output_extended_int \c_zero
+      \l_fp_output_exponent_int \c_zero
+      \exp_after:wN \fp_exp_decimal:
+    \fi:
+  }
+%    \end{macrocode}
+%   The integer part calculation starts with the hundreds. This is
+%   set up such that very large negative numbers can short-cut the entire
+%   procedure and simply return zero. In other cases, the code either
+%   recovers the exponent of the hundreds value or sets the appropriate
+%   storage to one (so that multiplication works correctly).
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \fp_exp_integer:
+  {
+    \if_int_compare:w \l_fp_input_a_integer_int < \c_one_hundred
+      \l_fp_exp_integer_int  \c_one
+      \l_fp_exp_decimal_int  \c_zero
+      \l_fp_exp_extended_int \c_zero
+      \l_fp_exp_exponent_int \c_zero
+      \exp_after:wN \fp_exp_integer_tens:
+    \else:
+      \tl_set:Nx \l_fp_tmp_tl
+        {
+          \exp_after:wN \use_i:nnn
+            \int_use:N \l_fp_input_a_integer_int
+        }
+      \l_fp_input_a_integer_int
+        \int_eval:w
+          \l_fp_input_a_integer_int - \l_fp_tmp_tl 00
+        \int_eval_end:
+      \if_int_compare:w \l_fp_input_a_sign_int < \c_zero
+        \if_int_compare:w \l_fp_output_integer_int > 200 \scan_stop:
+          \fp_exp_const:cx { c_fp_exp ( \l_fp_arg_tl ) _fp }
+            { \c_zero_fp }
+        \else:
+          \fp_exp_integer_const:n { - \l_fp_tmp_tl 00 }
+          \exp_after:wN \exp_after:wN \exp_after:wN
+            \exp_after:wN \exp_after:wN \exp_after:wN
+            \exp_after:wN \fp_exp_integer_tens:
+        \fi:
+      \else:
+        \fp_exp_integer_const:n { \l_fp_tmp_tl 00 }
+        \exp_after:wN \exp_after:wN \exp_after:wN
+          \exp_after:wN \fp_exp_integer_tens:
+      \fi:
+    \fi:
+  }
+%    \end{macrocode}
+%   The tens and units parts are handled in a similar way, with a
+%   multiplication step to build up the final value. That also includes a
+%   correction step to avoid an overflow of the integer part.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \fp_exp_integer_tens:
+  {
+    \l_fp_output_integer_int  \l_fp_exp_integer_int
+    \l_fp_output_decimal_int  \l_fp_exp_decimal_int
+    \l_fp_output_extended_int \l_fp_exp_extended_int
+    \l_fp_output_exponent_int \l_fp_exp_exponent_int
+    \if_int_compare:w \l_fp_input_a_integer_int > \c_nine
+      \tl_set:Nx \l_fp_tmp_tl
+        {
+          \exp_after:wN \use_i:nn
+            \int_use:N \l_fp_input_a_integer_int
+        }
+      \l_fp_input_a_integer_int
+        \int_eval:w
+          \l_fp_input_a_integer_int - \l_fp_tmp_tl 0
+        \int_eval_end:
+      \if_int_compare:w \l_fp_input_a_sign_int > \c_zero
+        \fp_exp_integer_const:n { \l_fp_tmp_tl 0 }
+      \else:
+        \fp_exp_integer_const:n { - \l_fp_tmp_tl 0 }
+      \fi:
+      \fp_mul:NNNNNNNNN
+        \l_fp_exp_integer_int \l_fp_exp_decimal_int \l_fp_exp_extended_int
+        \l_fp_output_integer_int \l_fp_output_decimal_int
+          \l_fp_output_extended_int
+        \l_fp_output_integer_int \l_fp_output_decimal_int
+          \l_fp_output_extended_int
+      \tex_advance:D \l_fp_output_exponent_int \l_fp_exp_exponent_int
+      \fp_extended_normalise_output:
+    \fi:
+    \fp_exp_integer_units:
+  }
+\cs_new_protected_nopar:Npn \fp_exp_integer_units:
+  {
+    \if_int_compare:w \l_fp_input_a_integer_int > \c_zero
+      \if_int_compare:w \l_fp_input_a_sign_int > \c_zero
+        \fp_exp_integer_const:n { \int_use:N \l_fp_input_a_integer_int }
+      \else:
+        \fp_exp_integer_const:n
+          { - \int_use:N \l_fp_input_a_integer_int }
+      \fi:
+      \fp_mul:NNNNNNNNN
+        \l_fp_exp_integer_int \l_fp_exp_decimal_int \l_fp_exp_extended_int
+        \l_fp_output_integer_int \l_fp_output_decimal_int
+          \l_fp_output_extended_int
+        \l_fp_output_integer_int \l_fp_output_decimal_int
+          \l_fp_output_extended_int
+      \tex_advance:D \l_fp_output_exponent_int \l_fp_exp_exponent_int
+      \fp_extended_normalise_output:
+    \fi:
+    \fp_exp_decimal:
+  }
+%    \end{macrocode}
+%   Recovery of the stored constant values into the separate registers
+%   is done with a simple expansion then assignment.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \fp_exp_integer_const:n #1
+  {
+    \exp_after:wN \exp_after:wN \exp_after:wN
+      \fp_exp_integer_const:nnnn
+      \cs:w c_fp_exp_ #1 _tl \cs_end:
+  }
+\cs_new_protected_nopar:Npn \fp_exp_integer_const:nnnn #1#2#3#4
+  {
+    \l_fp_exp_integer_int  #1 \scan_stop:
+    \l_fp_exp_decimal_int  #2 \scan_stop:
+    \l_fp_exp_extended_int #3 \scan_stop:
+    \l_fp_exp_exponent_int #4 \scan_stop:
+  }
+%    \end{macrocode}
+%   Finding the exponential for the decimal part of the number requires
+%   a Taylor series calculation. The set up is done here with the loop
+%   itself a separate function. Once the decimal part is available this
+%   is multiplied by the integer part already worked out to give
+%   the final result.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \fp_exp_decimal:
+  {
+    \if_int_compare:w \l_fp_input_a_decimal_int > \c_zero
+      \if_int_compare:w \l_fp_input_a_sign_int > \c_zero
+        \l_fp_exp_integer_int  \c_one
+        \l_fp_exp_decimal_int  \l_fp_input_a_decimal_int
+        \l_fp_exp_extended_int \l_fp_input_a_extended_int
+      \else:
+        \l_fp_exp_integer_int \c_zero
+        \if_int_compare:w \l_fp_exp_extended_int = \c_zero
+          \l_fp_exp_decimal_int
+            \int_eval:w
+              \c_one_thousand_million - \l_fp_input_a_decimal_int
+            \int_eval_end:
+          \l_fp_exp_extended_int \c_zero
+        \else:
+          \l_fp_exp_decimal_int
+            \int_eval:w
+              999999999 - \l_fp_input_a_decimal_int
+            \scan_stop:
+          \l_fp_exp_extended_int
+            \int_eval:w
+              \c_one_thousand_million - \l_fp_input_a_extended_int
+            \int_eval_end:
+        \fi:
+      \fi:
+      \l_fp_input_b_sign_int     \l_fp_input_a_sign_int
+      \l_fp_input_b_decimal_int  \l_fp_input_a_decimal_int
+      \l_fp_input_b_extended_int \l_fp_input_a_extended_int
+      \l_fp_count_int \c_one
+      \fp_exp_Taylor:
+      \fp_mul:NNNNNNNNN
+        \l_fp_exp_integer_int \l_fp_exp_decimal_int \l_fp_exp_extended_int
+        \l_fp_output_integer_int \l_fp_output_decimal_int
+          \l_fp_output_extended_int
+        \l_fp_output_integer_int \l_fp_output_decimal_int
+          \l_fp_output_extended_int
+    \fi:
+    \if_int_compare:w \l_fp_output_extended_int < \c_five_hundred_million
+    \else:
+      \tex_advance:D \l_fp_output_decimal_int \c_one
+      \if_int_compare:w \l_fp_output_decimal_int < \c_one_thousand_million
+      \else:
+        \l_fp_output_decimal_int \c_zero
+        \tex_advance:D \l_fp_output_integer_int \c_one
+      \fi:
+    \fi:
+    \fp_standardise:NNNN
+      \l_fp_output_sign_int
+      \l_fp_output_integer_int
+      \l_fp_output_decimal_int
+      \l_fp_output_exponent_int
+    \fp_exp_const:cx { c_fp_exp ( \l_fp_arg_tl ) _fp }
+      {
+        +
+        \int_use:N \l_fp_output_integer_int
+        .
+        \exp_after:wN \use_none:n
+          \int_value:w \int_eval:w
+             \l_fp_output_decimal_int + \c_one_thousand_million
+        e
+        \int_use:N \l_fp_output_exponent_int
+      }
+  }
+%    \end{macrocode}
+%   The Taylor series for $ \exp(x) $ is
+%   \[
+%     1 + x + \frac{x^2}{2!} + \frac{x^3}{3!} + \frac{x^4}{4!} + \cdots
+%   \]
+%   which converges for $ -1 < x < 1 $. The code above sets up
+%   the $ x $ part, leaving the loop to multiply the running
+%   value by $ x / n $ and add it onto the sum. The way that this is
+%   done is that the running total is stored in the \texttt{exp} set of
+%   registers, while the current item is stored as \texttt{input_b}.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \fp_exp_Taylor:
+  {
+    \tex_advance:D \l_fp_count_int \c_one
+    \tex_multiply:D \l_fp_input_b_sign_int \l_fp_input_a_sign_int
+    \fp_mul:NNNNNN
+      \l_fp_input_a_decimal_int \l_fp_input_a_extended_int
+      \l_fp_input_b_decimal_int \l_fp_input_b_extended_int
+      \l_fp_input_b_decimal_int \l_fp_input_b_extended_int
+    \fp_div_integer:NNNNN
+      \l_fp_input_b_decimal_int \l_fp_input_b_extended_int
+      \l_fp_count_int
+      \l_fp_input_b_decimal_int \l_fp_input_b_extended_int
+    \if_int_compare:w
+      \int_eval:w
+        \l_fp_input_b_decimal_int + \l_fp_input_b_extended_int
+        > \c_zero
+      \if_int_compare:w \l_fp_input_b_sign_int > \c_zero
+        \tex_advance:D \l_fp_exp_decimal_int \l_fp_input_b_decimal_int
+        \tex_advance:D \l_fp_exp_extended_int
+          \l_fp_input_b_extended_int
+        \if_int_compare:w \l_fp_exp_extended_int < \c_one_thousand_million
+      \else:
+          \tex_advance:D \l_fp_exp_decimal_int \c_one
+          \tex_advance:D \l_fp_exp_extended_int
+            -\c_one_thousand_million
+        \fi:
+        \if_int_compare:w \l_fp_exp_decimal_int < \c_one_thousand_million
+        \else:
+          \tex_advance:D \l_fp_exp_integer_int \c_one
+          \tex_advance:D \l_fp_exp_decimal_int
+            -\c_one_thousand_million
+        \fi:
+      \else:
+        \tex_advance:D \l_fp_exp_decimal_int -\l_fp_input_b_decimal_int
+        \tex_advance:D \l_fp_exp_extended_int
+          -\l_fp_input_a_extended_int
+        \if_int_compare:w \l_fp_exp_extended_int < \c_zero
+          \tex_advance:D \l_fp_exp_decimal_int \c_minus_one
+          \tex_advance:D \l_fp_exp_extended_int \c_one_thousand_million
+        \fi:
+        \if_int_compare:w \l_fp_exp_decimal_int < \c_zero
+          \tex_advance:D \l_fp_exp_integer_int \c_minus_one
+          \tex_advance:D \l_fp_exp_decimal_int \c_one_thousand_million
+        \fi:
+      \fi:
+      \exp_after:wN \fp_exp_Taylor:
+    \fi:
+  }
+%    \end{macrocode}
+%   This is set up as a function so that the power code can redirect
+%   the effect.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \fp_exp_const:Nx #1#2
+  {
+    \tl_new:N #1
+    \tl_gset:Nx #1 {#2}
+  }
+\cs_generate_variant:Nn \fp_exp_const:Nx { c }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{variable}{\c_fp_ln_10_1_tl}
+% \begin{variable}{\c_fp_ln_10_2_tl}
+% \begin{variable}{\c_fp_ln_10_3_tl}
+% \begin{variable}{\c_fp_ln_10_4_tl}
+% \begin{variable}{\c_fp_ln_10_5_tl}
+% \begin{variable}{\c_fp_ln_10_6_tl}
+% \begin{variable}{\c_fp_ln_10_7_tl}
+% \begin{variable}{\c_fp_ln_10_8_tl}
+% \begin{variable}{\c_fp_ln_10_9_tl}
+%   Constants for working out logarithms: first those for the powers of
+%   ten.
+%    \begin{macrocode}
+\tl_const:cn { c_fp_ln_10_1_tl } { { 2 } { 302585092 } { 994045684 } { 0 } }
+\tl_const:cn { c_fp_ln_10_2_tl } { { 4 } { 605170185 } { 988091368 } { 0 } }
+\tl_const:cn { c_fp_ln_10_3_tl } { { 6 } { 907755278 } { 982137052 } { 0 } }
+\tl_const:cn { c_fp_ln_10_4_tl } { { 9 } { 210340371 } { 976182736 } { 0 } }
+\tl_const:cn { c_fp_ln_10_5_tl } { { 1 } { 151292546 } { 497022842 } { 1 } }
+\tl_const:cn { c_fp_ln_10_6_tl } { { 1 } { 381551055 } { 796427410 } { 1 } }
+\tl_const:cn { c_fp_ln_10_7_tl } { { 1 } { 611809565 } { 095831979 } { 1 } }
+\tl_const:cn { c_fp_ln_10_8_tl } { { 1 } { 842068074 } { 395226547 } { 1 } }
+\tl_const:cn { c_fp_ln_10_9_tl } { { 2 } { 072326583 } { 694641116 } { 1 } }
+%    \end{macrocode}
+% \end{variable}
+% \end{variable}
+% \end{variable}
+% \end{variable}
+% \end{variable}
+% \end{variable}
+% \end{variable}
+% \end{variable}
+% \end{variable}
+%^^A 0.69...309 4.
+% \begin{variable}{\c_fp_ln_2_1_tl }
+% \begin{variable}{\c_fp_ln_2_2_tl }
+% \begin{variable}{\c_fp_ln_2_3_tl }
+% The smaller set for powers of two.
+%    \begin{macrocode}
+\tl_const:cn { c_fp_ln_2_1_tl } { { 0 } { 693147180 } { 559945309 } { 0 } }
+\tl_const:cn { c_fp_ln_2_2_tl } { { 1 } { 386294361 } { 119890618 } { 0 } }
+\tl_const:cn { c_fp_ln_2_3_tl } { { 2 } { 079441541 } { 679835928 } { 0 } }
+%    \end{macrocode}
+% \end{variable}
+% \end{variable}
+% \end{variable}
+%
+% \begin{macro}{\fp_ln:Nn, \fp_ln:cn}
+% \UnitTested
+% \begin{macro}{\fp_gln:Nn,\fp_gln:cn}
+% \UnitTested
+% \begin{macro}[aux]{\fp_ln_aux:NNn}
+% \begin{macro}[aux]{\fp_ln_aux:}
+% \begin{macro}[aux]{\fp_ln_exponent:}
+% \begin{macro}[aux]{\fp_ln_internal:}
+% \begin{macro}[aux]{\fp_ln_exponent_tens:}
+% \begin{macro}[aux]{\fp_ln_exponent_units:}
+% \begin{macro}[aux]{\fp_ln_normalise:}
+% \begin{macro}[aux]{\fp_ln_nornalise_aux:NNNNNNNNN}
+% \begin{macro}[aux]{\fp_ln_mantissa:}
+% \begin{macro}[aux]{\fp_ln_mantissa_aux:}
+% \begin{macro}[aux]{\fp_ln_mantissa_divide_two:}
+% \begin{macro}[aux]{\fp_ln_integer_const:nn}
+% \begin{macro}[aux]{\fp_ln_Taylor:}
+% \begin{macro}[aux]{\fp_ln_fixed:}
+% \begin{macro}[aux]{\fp_ln_fixed_aux:NNNNNNNNN}
+% \begin{macro}[aux]{\fp_ln_Taylor_aux:}
+%   The approach for logarithms is again based on a mix of tables and
+%   Taylor series. Here, the initial validation is a bit easier and so it
+%   is set up earlier, meaning less need to escape later on.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \fp_ln:Nn  { \fp_ln_aux:NNn \tl_set:Nn  }
+\cs_new_protected_nopar:Npn \fp_gln:Nn { \fp_ln_aux:NNn \tl_gset:Nn }
+\cs_generate_variant:Nn \fp_ln:Nn  { c }
+\cs_generate_variant:Nn \fp_gln:Nn { c }
+\cs_new_protected_nopar:Npn \fp_ln_aux:NNn #1#2#3
+  {
+    \group_begin:
+      \fp_split:Nn a {#3}
+      \fp_standardise:NNNN
+        \l_fp_input_a_sign_int
+        \l_fp_input_a_integer_int
+        \l_fp_input_a_decimal_int
+        \l_fp_input_a_exponent_int
+      \if_int_compare:w \l_fp_input_a_sign_int > \c_zero
+        \if_int_compare:w
+          \int_eval:w
+            \l_fp_input_a_integer_int + \l_fp_input_a_decimal_int
+            > \c_zero
+          \exp_after:wN \exp_after:wN \exp_after:wN \fp_ln_aux:
+        \else:
+          \cs_set_protected_nopar:Npx \fp_tmp:w ##1##2
+            {
+              \group_end:
+              ##1 \exp_not:N ##2 { \c_zero_fp }
+            }
+          \exp_after:wN \exp_after:wN \exp_after:wN \fp_ln_error_msg:
+        \fi:
+      \else:
+        \cs_set_protected_nopar:Npx \fp_tmp:w ##1##2
+          {
+            \group_end:
+            ##1 \exp_not:N ##2 { \c_zero_fp }
+          }
+        \exp_after:wN \fp_ln_error_msg:
+      \fi:
+    \fp_tmp:w #1 #2
+  }
+%    \end{macrocode}
+%   As the input at this stage meets the validity criteria above, the
+%   argument can now be saved for further processing. There is no need
+%   to look at the sign of the input as it must be positive. The function
+%   here simply sets up to either do the full calculation or recover
+%   the stored value, as appropriate.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \fp_ln_aux:
+  {
+    \tl_set:Nx \l_fp_arg_tl
+      {
+        +
+        \int_use:N \l_fp_input_a_integer_int
+        .
+        \exp_after:wN \use_none:n
+          \int_value:w \int_eval:w
+             \l_fp_input_a_decimal_int + \c_one_thousand_million
+        e
+        \int_use:N \l_fp_input_a_exponent_int
+      }
+    \if_cs_exist:w c_fp_ln ( \l_fp_arg_tl ) _fp \cs_end:
+    \else:
+      \exp_after:wN \fp_ln_exponent:
+    \fi:
+    \cs_set_protected_nopar:Npx \fp_tmp:w ##1##2
+      {
+        \group_end:
+        ##1 \exp_not:N ##2
+          { \use:c { c_fp_ln ( \l_fp_arg_tl ) _fp } }
+      }
+  }
+%    \end{macrocode}
+%   The main algorithm here uses the fact the logarithm can be divided
+%   up, first taking out the powers of ten, then powers of two and finally
+%   using a Taylor series for the remainder.
+%   \[
+%     \ln ( 10^{n} \times 2^{m} \times x )
+%       = \ln ( 10^{n} ) + \ln ( 2^{m} ) + \ln ( x )
+%   \]
+%   The second point to remember is that
+%   \[
+%     \ln ( x^{-1} ) = - \ln ( x )
+%   \]
+%   which means that for the powers of $ 10 $ and $ 2 $ constants
+%   are only needed for positive powers.
+%
+%   The first step is to set up the sign for the output functions and
+%   work out the powers of ten in the exponent. First the larger powers
+%   are sorted out. The values for the constants are the same as those
+%   for the smaller ones, just with a shift in the exponent.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \fp_ln_exponent:
+  {
+    \fp_ln_internal:
+    \if_int_compare:w \l_fp_output_extended_int < \c_five_hundred_million
+    \else:
+      \tex_advance:D \l_fp_output_decimal_int \c_one
+      \if_int_compare:w \l_fp_output_decimal_int < \c_one_thousand_million
+      \else:
+        \l_fp_output_decimal_int \c_zero
+        \tex_advance:D \l_fp_output_integer_int \c_one
+      \fi:
+    \fi:
+    \fp_standardise:NNNN
+      \l_fp_output_sign_int
+      \l_fp_output_integer_int
+      \l_fp_output_decimal_int
+      \l_fp_output_exponent_int
+    \tl_const:cx { c_fp_ln ( \l_fp_arg_tl ) _fp }
+      {
+        \if_int_compare:w \l_fp_output_sign_int > \c_zero
+          +
+        \else:
+          -
+        \fi:
+        \int_use:N \l_fp_output_integer_int
+        .
+        \exp_after:wN \use_none:n
+          \int_value:w \int_eval:w
+             \l_fp_output_decimal_int + \c_one_thousand_million
+           \scan_stop:
+        e
+        \int_use:N \l_fp_output_exponent_int
+      }
+  }
+\cs_new_protected_nopar:Npn \fp_ln_internal:
+  {
+    \if_int_compare:w \l_fp_input_a_exponent_int < \c_zero
+      \l_fp_input_a_exponent_int -\l_fp_input_a_exponent_int
+      \l_fp_output_sign_int \c_minus_one
+    \else:
+      \l_fp_output_sign_int \c_one
+    \fi:
+    \if_int_compare:w \l_fp_input_a_exponent_int > \c_nine
+      \exp_after:wN \fp_ln_exponent_tens:NN
+        \int_use:N \l_fp_input_a_exponent_int
+    \else:
+      \l_fp_output_integer_int  \c_zero
+      \l_fp_output_decimal_int  \c_zero
+      \l_fp_output_extended_int \c_zero
+      \l_fp_output_exponent_int \c_zero
+    \fi:
+    \fp_ln_exponent_units:
+  }
+\cs_new_protected_nopar:Npn \fp_ln_exponent_tens:NN #1 #2
+  {
+    \l_fp_input_a_exponent_int #2 \scan_stop:
+    \fp_ln_const:nn { 10 } { #1 }
+    \tex_advance:D \l_fp_exp_exponent_int \c_one
+    \l_fp_output_integer_int  \l_fp_exp_integer_int
+    \l_fp_output_decimal_int  \l_fp_exp_decimal_int
+    \l_fp_output_extended_int \l_fp_exp_extended_int
+    \l_fp_output_exponent_int \l_fp_exp_exponent_int
+  }
+%    \end{macrocode}
+%   Next the smaller powers of ten, which will need to be combined
+%   with the above: always an additive process.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \fp_ln_exponent_units:
+  {
+    \if_int_compare:w \l_fp_input_a_exponent_int > \c_zero
+      \fp_ln_const:nn { 10 } { \int_use:N \l_fp_input_a_exponent_int }
+      \fp_ln_normalise:
+      \fp_add:NNNNNNNNN
+        \l_fp_exp_integer_int \l_fp_exp_decimal_int \l_fp_exp_extended_int
+        \l_fp_output_integer_int \l_fp_output_decimal_int
+          \l_fp_output_extended_int
+        \l_fp_output_integer_int \l_fp_output_decimal_int
+          \l_fp_output_extended_int
+    \fi:
+    \fp_ln_mantissa:
+  }
+%    \end{macrocode}
+%   The smaller table-based parts may need to be exponent shifted so that
+%   they stay in line with the larger parts. This is similar to the
+%   approach in other places, but here there is a need to watch the
+%   extended part of the number. The only case where the new exponent is
+%   larger than the old is if there was no previous part. Then simply set
+%   the exponent.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \fp_ln_normalise:
+  {
+    \if_int_compare:w \l_fp_exp_exponent_int < \l_fp_output_exponent_int
+      \tex_advance:D \l_fp_exp_decimal_int \c_one_thousand_million
+      \exp_after:wN \use_i:nn \exp_after:wN
+        \fp_ln_normalise_aux:NNNNNNNNN
+        \int_use:N \l_fp_exp_decimal_int
+       \exp_after:wN \fp_ln_normalise:
+    \else:
+      \l_fp_output_exponent_int \l_fp_exp_exponent_int
+    \fi:
+  }
+\cs_new_protected_nopar:Npn \fp_ln_normalise_aux:NNNNNNNNN #1#2#3#4#5#6#7#8#9
+  {
+    \if_int_compare:w \l_fp_exp_integer_int = \c_zero
+      \l_fp_exp_decimal_int #1#2#3#4#5#6#7#8 \scan_stop:
+    \else:
+      \tl_set:Nx \l_fp_tmp_tl
+        {
+          \int_use:N \l_fp_exp_integer_int
+          #1#2#3#4#5#6#7#8
+        }
+      \l_fp_exp_integer_int \c_zero
+      \l_fp_exp_decimal_int \l_fp_tmp_tl \scan_stop:
+    \fi:
+    \tex_divide:D \l_fp_exp_extended_int \c_ten
+    \tl_set:Nx \l_fp_tmp_tl
+      {
+        #9
+        \int_use:N \l_fp_exp_extended_int
+      }
+    \l_fp_exp_extended_int \l_fp_tmp_tl \scan_stop:
+    \tex_advance:D \l_fp_exp_exponent_int \c_one
+  }
+%    \end{macrocode}
+%   The next phase is to decompose the mantissa by division by two to
+%   leave a value which is in the range $ 1 \le x < 2 $. The sum of the
+%   two powers needs to take account of the sign of the output: if it
+%   is negative then the result gets \emph{smaller} as the mantissa gets
+%   \emph{bigger}.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \fp_ln_mantissa:
+  {
+    \l_fp_count_int \c_zero
+    \l_fp_input_a_extended_int \c_zero
+    \fp_ln_mantissa_aux:
+    \if_int_compare:w \l_fp_count_int > \c_zero
+      \fp_ln_const:nn { 2 } { \int_use:N \l_fp_count_int }
+      \fp_ln_normalise:
+      \if_int_compare:w \l_fp_output_sign_int > \c_zero
+        \exp_after:wN \fp_add:NNNNNNNNN
+      \else:
+        \exp_after:wN \fp_sub:NNNNNNNNN
+      \fi:
+      \l_fp_output_integer_int \l_fp_output_decimal_int
+        \l_fp_output_extended_int
+      \l_fp_exp_integer_int \l_fp_exp_decimal_int \l_fp_exp_extended_int
+      \l_fp_output_integer_int \l_fp_output_decimal_int
+        \l_fp_output_extended_int
+    \fi:
+    \if_int_compare:w
+      \int_eval:w
+        \l_fp_input_a_integer_int + \l_fp_input_a_decimal_int > \c_one
+      \exp_after:wN \fp_ln_Taylor:
+    \fi:
+  }
+\cs_new_protected_nopar:Npn \fp_ln_mantissa_aux:
+  {
+    \if_int_compare:w \l_fp_input_a_integer_int > \c_one
+      \tex_advance:D \l_fp_count_int \c_one
+      \fp_ln_mantissa_divide_two:
+      \exp_after:wN \fp_ln_mantissa_aux:
+    \fi:
+  }
+%    \end{macrocode}
+%   A fast one-shot division by two.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \fp_ln_mantissa_divide_two:
+  {
+    \if_int_odd:w \l_fp_input_a_decimal_int
+      \tex_advance:D \l_fp_input_a_extended_int \c_one_thousand_million
+    \fi:
+    \if_int_odd:w \l_fp_input_a_integer_int
+      \tex_advance:D \l_fp_input_a_decimal_int \c_one_thousand_million
+    \fi:
+    \tex_divide:D \l_fp_input_a_integer_int  \c_two
+    \tex_divide:D \l_fp_input_a_decimal_int  \c_two
+    \tex_divide:D \l_fp_input_a_extended_int \c_two
+  }
+%    \end{macrocode}
+%   Recovering constants makes use of the same auxiliary code as for
+%   exponents.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \fp_ln_const:nn #1#2
+  {
+    \exp_after:wN \exp_after:wN \exp_after:wN
+      \fp_exp_integer_const:nnnn
+      \cs:w c_fp_ln_ #1 _ #2 _tl \cs_end:
+  }
+%    \end{macrocode}
+%   The Taylor series for the logarithm function is best implemented using
+%   the identity
+%   \[
+%     \ln(x) = \ln\left( \frac{y + 1}{y - 1} \right)
+%   \]
+%   with
+%   \[
+%     y = \frac{x - 1}{x + 1}
+%   \]
+%   This leads to the series
+%   \[
+%     \ln(x)
+%       = 2y
+%         \left(
+%           1 + y^{2}
+%             \left(
+%               \frac{1}{3} + y^{2}
+%                 \left(
+%                   \frac{1}{5} + y^{2}
+%                     \left(
+%                       \frac{1}{7} + y^{2}
+%                         \left(
+%                           \frac{1}{9} + \cdots
+%                         \right)
+%                     \right)
+%                 \right)
+%             \right)
+%         \right)
+%   \]
+%   This expansion has the advantage that a lot of the work can be
+%   loaded up early by finding $ y^{2} $ before the loop itself starts.
+%   (In practice, the implementation does the multiplication by two at the
+%   end of the loop, and expands out the brackets as this is an overall
+%   more efficient approach.)
+%
+%   At the implementation level, the code starts by calculating $ y $
+%   and storing that in input \texttt{a} (which is no longer needed
+%   for other purposes). That is done using the full division system
+%   avoiding the parsing step. The value is then switched to a fixed-point
+%   representation. There is then some shuffling to get all of the working
+%   space set up. At this stage, a lot of registers are in use and so
+%   the Taylor series is calculated within a group so that the
+%   \texttt{output} variables can be used to hold the result. The value
+%   of $ y^{2} $ is held in input \texttt{b} (there are a few
+%   assignments saved by choosing this over \texttt{a}), while input
+%   \texttt{a} is used for the \enquote{loop value}.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \fp_ln_Taylor:
+  {
+    \group_begin:
+      \l_fp_input_a_integer_int \c_zero
+      \l_fp_input_a_exponent_int \c_zero
+      \l_fp_input_b_integer_int \c_two
+      \l_fp_input_b_decimal_int \l_fp_input_a_decimal_int
+      \l_fp_input_b_exponent_int \c_zero
+      \fp_div_internal:
+      \fp_ln_fixed:
+      \l_fp_input_a_integer_int  \l_fp_output_integer_int
+      \l_fp_input_a_decimal_int  \l_fp_output_decimal_int
+      \l_fp_input_a_extended_int \c_zero
+      \l_fp_input_a_exponent_int \l_fp_output_exponent_int
+      \l_fp_output_decimal_int \c_zero %^^A Bug?
+      \l_fp_output_decimal_int  \l_fp_input_a_decimal_int
+      \l_fp_output_extended_int \l_fp_input_a_extended_int
+      \fp_mul:NNNNNN
+        \l_fp_input_a_decimal_int \l_fp_input_a_extended_int
+        \l_fp_input_a_decimal_int \l_fp_input_a_extended_int
+        \l_fp_input_b_decimal_int \l_fp_input_b_extended_int
+      \l_fp_count_int \c_one
+      \fp_ln_Taylor_aux:
+      \cs_set_protected_nopar:Npx \fp_tmp:w
+        {
+          \group_end:
+          \l_fp_exp_integer_int \c_zero
+          \exp_not:N \l_fp_exp_decimal_int
+            \int_use:N \l_fp_output_decimal_int \scan_stop:
+          \exp_not:N \l_fp_exp_extended_int
+            \int_use:N \l_fp_output_extended_int \scan_stop:
+          \exp_not:N \l_fp_exp_exponent_int
+            \int_use:N \l_fp_output_exponent_int \scan_stop:
+        }
+    \fp_tmp:w
+%    \end{macrocode}
+%   After the loop part of the Taylor series, the factor of $ 2 $ needs
+%   to be included. The total for the result can then be constructed.
+%    \begin{macrocode}
+    \tex_advance:D \l_fp_exp_decimal_int \l_fp_exp_decimal_int
+    \if_int_compare:w \l_fp_exp_extended_int < \c_five_hundred_million
+    \else:
+      \tex_advance:D \l_fp_exp_extended_int -\c_five_hundred_million
+      \tex_advance:D \l_fp_exp_decimal_int \c_one
+    \fi:
+    \tex_advance:D \l_fp_exp_extended_int \l_fp_exp_extended_int
+    \fp_ln_normalise:
+    \if_int_compare:w \l_fp_output_sign_int > \c_zero
+      \exp_after:wN \fp_add:NNNNNNNNN
+    \else:
+      \exp_after:wN \fp_sub:NNNNNNNNN
+    \fi:
+    \l_fp_output_integer_int \l_fp_output_decimal_int
+      \l_fp_output_extended_int
+    \c_zero \l_fp_exp_decimal_int \l_fp_exp_extended_int
+    \l_fp_output_integer_int \l_fp_output_decimal_int
+      \l_fp_output_extended_int
+  }
+%    \end{macrocode}
+% The usual shifts to move to fixed-point working. This is done using
+% the \texttt{output} registers as this saves a reassignment here.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \fp_ln_fixed:
+  {
+    \if_int_compare:w \l_fp_output_exponent_int < \c_zero
+      \tex_advance:D \l_fp_output_decimal_int \c_one_thousand_million
+      \exp_after:wN \use_i:nn \exp_after:wN
+        \fp_ln_fixed_aux:NNNNNNNNN
+        \int_use:N \l_fp_output_decimal_int
+       \exp_after:wN \fp_ln_fixed:
+     \fi:
+  }
+\cs_new_protected_nopar:Npn \fp_ln_fixed_aux:NNNNNNNNN #1#2#3#4#5#6#7#8#9
+  {
+    \if_int_compare:w \l_fp_output_integer_int = \c_zero
+      \l_fp_output_decimal_int #1#2#3#4#5#6#7#8 \scan_stop:
+    \else:
+      \tl_set:Nx \l_fp_tmp_tl
+        {
+          \int_use:N \l_fp_output_integer_int
+          #1#2#3#4#5#6#7#8
+        }
+      \l_fp_output_integer_int \c_zero
+      \l_fp_output_decimal_int \l_fp_tmp_tl \scan_stop:
+    \fi:
+    \tex_advance:D \l_fp_output_exponent_int \c_one
+  }
+%    \end{macrocode}
+%   The main loop for the Taylor series: unlike some of the other similar
+%   functions, the result here is not the final value and is therefore
+%   subject to further manipulation outside of the loop.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \fp_ln_Taylor_aux:
+  {
+    \tex_advance:D \l_fp_count_int \c_two
+    \fp_mul:NNNNNN
+      \l_fp_input_a_decimal_int \l_fp_input_a_extended_int
+      \l_fp_input_b_decimal_int \l_fp_input_b_extended_int
+      \l_fp_input_a_decimal_int \l_fp_input_a_extended_int
+    \if_int_compare:w
+      \int_eval:w
+        \l_fp_input_a_decimal_int + \l_fp_input_a_extended_int
+        > \c_zero
+      \fp_div_integer:NNNNN
+        \l_fp_input_a_decimal_int \l_fp_input_a_extended_int
+        \l_fp_count_int
+        \l_fp_exp_decimal_int \l_fp_exp_extended_int
+        \tex_advance:D \l_fp_output_decimal_int \l_fp_exp_decimal_int
+        \tex_advance:D \l_fp_output_extended_int \l_fp_exp_extended_int
+        \if_int_compare:w \l_fp_output_extended_int < \c_one_thousand_million
+        \else:
+          \tex_advance:D \l_fp_output_decimal_int \c_one
+          \tex_advance:D \l_fp_output_extended_int
+            -\c_one_thousand_million
+        \fi:
+        \if_int_compare:w \l_fp_output_decimal_int < \c_one_thousand_million
+        \else:
+          \tex_advance:D \l_fp_output_integer_int \c_one
+          \tex_advance:D \l_fp_output_decimal_int
+            -\c_one_thousand_million
+        \fi:
+      \exp_after:wN \fp_ln_Taylor_aux:
+    \fi:
+  }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\fp_pow:Nn, \fp_pow:cn}
+% \UnitTested
+% \begin{macro}{\fp_gpow:Nn,\fp_gpow:cn}
+% \UnitTested
+% \begin{macro}[aux]{\fp_pow_aux:NNn}
+% \begin{macro}[aux]{\fp_pow_aux_i:}
+% \begin{macro}[aux]{\fp_pow_positive:}
+% \begin{macro}[aux]{\fp_pow_negative:}
+% \begin{macro}[aux]{\fp_pow_aux_ii:}
+% \begin{macro}[aux]{\fp_pow_aux_iii:}
+% \begin{macro}[aux]{\fp_pow_aux_iv:}
+%   The approach used for working out powers is to first filter out the
+%   various special cases and then do most of the work using the
+%   logarithm and exponent functions. The two storage areas are used
+%   in the reverse of the `natural' logic as this avoids some
+%   re-assignment in the sanity checking code.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \fp_pow:Nn  { \fp_pow_aux:NNn \tl_set:Nn }
+\cs_new_protected_nopar:Npn \fp_gpow:Nn { \fp_pow_aux:NNn \tl_gset:Nn }
+\cs_generate_variant:Nn \fp_pow:Nn  { c }
+\cs_generate_variant:Nn \fp_gpow:Nn { c }
+\cs_new_protected_nopar:Npn \fp_pow_aux:NNn #1#2#3
+  {
+    \group_begin:
+      \fp_read:N #2
+      \l_fp_input_b_sign_int     \l_fp_input_a_sign_int
+      \l_fp_input_b_integer_int  \l_fp_input_a_integer_int
+      \l_fp_input_b_decimal_int  \l_fp_input_a_decimal_int
+      \l_fp_input_b_exponent_int \l_fp_input_a_exponent_int
+      \fp_split:Nn a {#3}
+      \fp_standardise:NNNN
+        \l_fp_input_a_sign_int
+        \l_fp_input_a_integer_int
+        \l_fp_input_a_decimal_int
+        \l_fp_input_a_exponent_int
+      \if_int_compare:w
+        \int_eval:w
+          \l_fp_input_b_integer_int + \l_fp_input_b_decimal_int
+          = \c_zero
+         \if_int_compare:w
+           \int_eval:w
+             \l_fp_input_a_integer_int + \l_fp_input_a_decimal_int
+             = \c_zero
+             \cs_set_protected_nopar:Npx \fp_tmp:w ##1##2
+               {
+                 \group_end:
+                 ##1 ##2 { \c_undefined_fp }
+               }
+           \else:
+             \cs_set_protected_nopar:Npx \fp_tmp:w ##1##2
+               {
+                 \group_end:
+                 ##1 ##2 { \c_zero_fp }
+               }
+          \fi:
+       \else:
+         \if_int_compare:w
+           \int_eval:w
+             \l_fp_input_a_integer_int + \l_fp_input_a_decimal_int
+             = \c_zero
+             \cs_set_protected_nopar:Npx \fp_tmp:w ##1##2
+               {
+                 \group_end:
+                 ##1 ##2 { \c_one_fp }
+               }
+           \else:
+             \exp_after:wN \exp_after:wN \exp_after:wN
+               \fp_pow_aux_i:
+          \fi:
+       \fi:
+    \fp_tmp:w #1 #2
+}
+%    \end{macrocode}
+%   Simply using the logarithm function directly will fail when negative
+%   numbers are raised to integer powers, which is a mathematically valid
+%   operation. So there are some more tests to make, after forcing the
+%   power into an integer and decimal parts, if necessary.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \fp_pow_aux_i:
+  {
+    \if_int_compare:w \l_fp_input_b_sign_int > \c_zero
+      \tl_set:Nn \l_fp_sign_tl { + }
+      \exp_after:wN \fp_pow_aux_ii:
+    \else:
+      \l_fp_input_a_extended_int \c_zero
+      \if_int_compare:w \l_fp_input_a_exponent_int < \c_ten
+        \group_begin:
+        \fp_extended_normalise:
+        \if_int_compare:w
+          \int_eval:w
+            \l_fp_input_a_decimal_int + \l_fp_input_a_extended_int
+            = \c_zero
+           \group_end:
+          \tl_set:Nn \l_fp_sign_tl { - }
+          \exp_after:wN \exp_after:wN \exp_after:wN
+          \exp_after:wN \exp_after:wN \exp_after:wN
+          \exp_after:wN \fp_pow_aux_ii:
+        \else:
+          \group_end:
+          \cs_set_protected_nopar:Npx \fp_tmp:w ##1##2
+            {
+              \group_end:
+              ##1 ##2 { \c_undefined_fp }
+           }
+        \fi:
+      \else:
+        \cs_set_protected_nopar:Npx \fp_tmp:w ##1##2
+          {
+            \group_end:
+            ##1 ##2 { \c_undefined_fp }
+         }
+      \fi:
+    \fi:
+  }
+%    \end{macrocode}
+%   The approach used here for powers works well in most cases but gives
+%   poorer results for negative integer powers, which often have exact
+%   values.  So there is some filtering to do. For negative powers where
+%   the power is small, an alternative approach is used in which the
+%   positive value is worked out and the reciprocal is then taken. The
+%   filtering is unfortunately rather long.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \fp_pow_aux_ii:
+  {
+    \if_int_compare:w \l_fp_input_a_sign_int > \c_zero
+      \exp_after:wN \fp_pow_aux_iv:
+    \else:
+      \if_int_compare:w \l_fp_input_a_exponent_int < \c_ten
+        \group_begin:
+        \l_fp_input_a_extended_int \c_zero
+        \fp_extended_normalise:
+        \if_int_compare:w \l_fp_input_a_decimal_int = \c_zero
+          \if_int_compare:w \l_fp_input_a_integer_int > \c_ten
+            \group_end:
+            \exp_after:wN \exp_after:wN \exp_after:wN
+            \exp_after:wN \exp_after:wN \exp_after:wN
+            \exp_after:wN \exp_after:wN \exp_after:wN
+            \exp_after:wN \exp_after:wN \exp_after:wN
+            \exp_after:wN \exp_after:wN \exp_after:wN
+              \fp_pow_aux_iv:
+          \else:
+            \group_end:
+            \exp_after:wN \exp_after:wN \exp_after:wN
+            \exp_after:wN \exp_after:wN \exp_after:wN
+            \exp_after:wN \exp_after:wN \exp_after:wN
+            \exp_after:wN \exp_after:wN \exp_after:wN
+            \exp_after:wN \exp_after:wN \exp_after:wN
+              \exp_after:wN \fp_pow_aux_iii:
+          \fi:
+        \else:
+          \group_end:
+            \exp_after:wN \exp_after:wN \exp_after:wN
+            \exp_after:wN \exp_after:wN \exp_after:wN
+              \exp_after:wN \fp_pow_aux_iv:
+        \fi:
+      \else:
+        \exp_after:wN \exp_after:wN \exp_after:wN
+          \fp_pow_aux_iv:
+      \fi:
+    \fi:
+    \cs_set_protected_nopar:Npx \fp_tmp:w ##1##2
+      {
+        \group_end:
+        ##1 ##2
+          {
+            \l_fp_sign_tl
+            \int_use:N \l_fp_output_integer_int
+            .
+            \exp_after:wN \use_none:n
+              \int_value:w \int_eval:w
+                 \l_fp_output_decimal_int + \c_one_thousand_million
+            e
+            \int_use:N \l_fp_output_exponent_int
+          }
+      }
+  }
+%    \end{macrocode}
+%   For the small negative integer powers, the calculation is done for
+%   the positive power and the reciprocal is then taken.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \fp_pow_aux_iii:
+  {
+    \l_fp_input_a_sign_int \c_one
+    \fp_pow_aux_iv:
+    \l_fp_input_a_integer_int  \c_one
+    \l_fp_input_a_decimal_int  \c_zero
+    \l_fp_input_a_exponent_int \c_zero
+    \l_fp_input_b_integer_int  \l_fp_output_integer_int
+    \l_fp_input_b_decimal_int  \l_fp_output_decimal_int
+    \l_fp_input_b_exponent_int \l_fp_output_exponent_int
+    \fp_div_internal:
+  }
+%    \end{macrocode}
+%   The business end of the code starts by finding the logarithm of the
+%   given base. There is a bit of a shuffle so that this does not have
+%   to be re-parsed and so that the output ends up in the correct place.
+%   There is also a need to enable using the short-cut for a
+%   pre-calculated result. The internal part of the multiplication
+%   function can then be used to do the second part of the calculation
+%   directly. There is some more set up before doing the exponential:
+%   the idea here is to deactivate some internals so that everything works
+%   smoothly.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \fp_pow_aux_iv:
+  {
+    \group_begin:
+      \l_fp_input_a_integer_int  \l_fp_input_b_integer_int
+      \l_fp_input_a_decimal_int  \l_fp_input_b_decimal_int
+      \l_fp_input_a_exponent_int \l_fp_input_b_exponent_int
+      \fp_ln_internal:
+      \cs_set_protected_nopar:Npx \fp_tmp:w
+        {
+          \group_end:
+          \exp_not:N \l_fp_input_b_sign_int
+            \int_use:N \l_fp_output_sign_int \scan_stop:
+          \exp_not:N \l_fp_input_b_integer_int
+            \int_use:N \l_fp_output_integer_int \scan_stop:
+          \exp_not:N \l_fp_input_b_decimal_int
+            \int_use:N \l_fp_output_decimal_int \scan_stop:
+          \exp_not:N \l_fp_input_b_extended_int
+            \int_use:N \l_fp_output_extended_int \scan_stop:
+          \exp_not:N \l_fp_input_b_exponent_int
+            \int_use:N \l_fp_output_exponent_int \scan_stop:
+        }
+    \fp_tmp:w
+    \l_fp_input_a_extended_int  \c_zero
+    \fp_mul:NNNNNNNNN
+      \l_fp_input_a_integer_int \l_fp_input_a_decimal_int
+        \l_fp_input_a_extended_int
+      \l_fp_input_b_integer_int \l_fp_input_b_decimal_int
+        \l_fp_input_b_extended_int
+      \l_fp_output_integer_int \l_fp_output_decimal_int
+        \l_fp_output_extended_int
+    \l_fp_output_exponent_int
+      \int_eval:w
+        \l_fp_input_a_exponent_int + \l_fp_input_b_exponent_int
+      \scan_stop:
+    \fp_extended_normalise_output:
+    \tex_multiply:D \l_fp_input_a_sign_int \l_fp_input_b_sign_int
+    \l_fp_input_a_integer_int  \l_fp_output_integer_int
+    \l_fp_input_a_decimal_int  \l_fp_output_decimal_int
+    \l_fp_input_a_extended_int \l_fp_output_extended_int
+    \l_fp_input_a_exponent_int \l_fp_output_exponent_int
+    \l_fp_output_integer_int  \c_zero
+    \l_fp_output_decimal_int  \c_zero
+    \l_fp_output_extended_int \c_zero
+    \l_fp_output_exponent_int \c_zero
+    \cs_set_eq:NN \fp_exp_const:Nx \use_none:nn
+    \fp_exp_internal:
+  }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \subsection{Tests for special values}
+%
+% \begin{macro}[pTF]{\fp_if_undefined:N}
+% \UnitTested
+%   Testing for an undefined value is easy.
+%    \begin{macrocode}
+\prg_new_conditional:Npnn \fp_if_undefined:N #1 { p , T , F , TF }
+  {
+    \if_meaning:w #1 \c_undefined_fp
+      \prg_return_true:
+    \else:
+      \prg_return_false:
+    \fi:
+  }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}[pTF]{\fp_if_zero:N}
+% \UnitTested
+%   Testing for a zero fixed-point is also easy.
+%    \begin{macrocode}
+\prg_new_conditional:Npnn \fp_if_zero:N #1 { p , T , F , TF }
+  {
+    \if_meaning:w #1 \c_zero_fp
+      \prg_return_true:
+    \else:
+      \prg_return_false:
+    \fi:
+  }
+%    \end{macrocode}
+% \end{macro}
+%
+% \subsection{Floating-point conditionals}
+%
+% \begin{macro}[TF]{\fp_compare:nNn}
+% \begin{macro}[TF]{\fp_compare:NNN}
+% \UnitTested
+% \begin{macro}[aux]{\fp_compare_aux:N}
+% \begin{macro}[aux]{\fp_compare_=:}
+% \begin{macro}[aux]{\fp_compare_<:}
+% \begin{macro}[aux]{\fp_compare_<_aux:}
+% \begin{macro}[aux]{\fp_compare_absolute_a>b:}
+% \begin{macro}[aux]{\fp_compare_absolute_a<b:}
+% \begin{macro}[aux]{\fp_compare_>:}
+%   The idea for the comparisons is to provide two versions: slower and
+%   faster. The lead off for both is the same: get the two numbers
+%   read and then look for a function to handle the comparison.
+%    \begin{macrocode}
+\prg_new_protected_conditional:Npnn \fp_compare:nNn #1#2#3 { T , F , TF }
+  {
+    \group_begin:
+      \fp_split:Nn a {#1}
+      \fp_standardise:NNNN
+        \l_fp_input_a_sign_int
+        \l_fp_input_a_integer_int
+        \l_fp_input_a_decimal_int
+        \l_fp_input_a_exponent_int
+      \fp_split:Nn b {#3}
+      \fp_standardise:NNNN
+        \l_fp_input_b_sign_int
+        \l_fp_input_b_integer_int
+        \l_fp_input_b_decimal_int
+        \l_fp_input_b_exponent_int
+      \fp_compare_aux:N #2
+  }
+\prg_new_protected_conditional:Npnn \fp_compare:NNN #1#2#3 { T , F , TF }
+  {
+    \group_begin:
+      \fp_read:N #3
+      \l_fp_input_b_sign_int     \l_fp_input_a_sign_int
+      \l_fp_input_b_integer_int  \l_fp_input_a_integer_int
+      \l_fp_input_b_decimal_int  \l_fp_input_a_decimal_int
+      \l_fp_input_b_exponent_int \l_fp_input_a_exponent_int
+      \fp_read:N #1
+      \fp_compare_aux:N #2
+  }
+\cs_new_protected_nopar:Npn \fp_compare_aux:N #1
+  {
+    \cs_if_exist:cTF { fp_compare_#1: }
+      { \use:c { fp_compare_#1: } }
+      {
+        \group_end:
+        \prg_return_false:
+      }
+  }
+%    \end{macrocode}
+%   For equality, the test is pretty easy as things are either equal or
+%   they are not.
+%    \begin{macrocode}
+\cs_new_protected_nopar:cpn { fp_compare_=: }
+  {
+    \if_int_compare:w \l_fp_input_a_sign_int = \l_fp_input_b_sign_int
+      \if_int_compare:w \l_fp_input_a_integer_int = \l_fp_input_b_integer_int
+        \if_int_compare:w \l_fp_input_a_decimal_int = \l_fp_input_b_decimal_int
+          \if_int_compare:w
+            \l_fp_input_a_exponent_int = \l_fp_input_b_exponent_int
+            \group_end:
+            \prg_return_true:
+          \else:
+            \group_end:
+            \prg_return_false:
+          \fi:
+        \else:
+          \group_end:
+          \prg_return_false:
+        \fi:
+      \else:
+        \group_end:
+        \prg_return_false:
+      \fi:
+    \else:
+      \group_end:
+      \prg_return_false:
+    \fi:
+  }
+%    \end{macrocode}
+%   Comparing two values is quite complex. First, there is a filter step
+%   to check if one or other of the given values is zero. If it is then
+%   the result is relatively easy to determine.
+%    \begin{macrocode}
+\cs_new_protected_nopar:cpn { fp_compare_>: }
+  {
+    \if_int_compare:w \int_eval:w
+      \l_fp_input_a_integer_int + \l_fp_input_a_decimal_int
+      = \c_zero
+      \if_int_compare:w \int_eval:w
+        \l_fp_input_b_integer_int + \l_fp_input_b_decimal_int
+        = \c_zero
+        \group_end:
+        \prg_return_false:
+      \else:
+        \if_int_compare:w \l_fp_input_b_sign_int > \c_zero
+          \group_end:
+          \prg_return_false:
+        \else:
+          \group_end:
+          \prg_return_true:
+        \fi:
+      \fi:
+    \else:
+      \if_int_compare:w \int_eval:w
+        \l_fp_input_b_integer_int + \l_fp_input_b_decimal_int
+        = \c_zero
+        \if_int_compare:w \l_fp_input_a_sign_int > \c_zero
+          \group_end:
+          \prg_return_true:
+        \else:
+          \group_end:
+          \prg_return_false:
+        \fi:
+      \else:
+        \use:c { fp_compare_>_aux: }
+      \fi:
+    \fi:
+  }
+%    \end{macrocode}
+%   Next, check the sign of the input: this again may give an obvious
+%   result. If both signs are the same, then hand off to comparing the
+%   absolute values.
+%    \begin{macrocode}
+\cs_new_protected_nopar:cpn { fp_compare_>_aux: }
+  {
+    \if_int_compare:w \l_fp_input_a_sign_int > \l_fp_input_b_sign_int
+      \group_end:
+      \prg_return_true:
+    \else:
+      \if_int_compare:w \l_fp_input_a_sign_int < \l_fp_input_b_sign_int
+        \group_end:
+        \prg_return_false:
+      \else:
+        \if_int_compare:w \l_fp_input_a_sign_int > \c_zero
+          \use:c { fp_compare_absolute_a>b: }
+        \else:
+          \use:c { fp_compare_absolute_a<b: }
+        \fi:
+      \fi:
+    \fi:
+  }
+%    \end{macrocode}
+%   Rather long runs of checks, as there is the need to go through each
+%   layer of the input and do the comparison. There is also the need to
+%   avoid messing up with equal inputs at each stage.
+%    \begin{macrocode}
+\cs_new_protected_nopar:cpn { fp_compare_absolute_a>b: }
+  {
+    \if_int_compare:w \l_fp_input_a_exponent_int > \l_fp_input_b_exponent_int
+      \group_end:
+      \prg_return_true:
+    \else:
+      \if_int_compare:w \l_fp_input_a_exponent_int < \l_fp_input_b_exponent_int
+        \group_end:
+        \prg_return_false:
+      \else:
+        \if_int_compare:w \l_fp_input_a_integer_int > \l_fp_input_b_integer_int
+          \group_end:
+          \prg_return_true:
+        \else:
+          \if_int_compare:w
+            \l_fp_input_a_integer_int < \l_fp_input_b_integer_int
+            \group_end:
+            \prg_return_false:
+          \else:
+            \if_int_compare:w
+              \l_fp_input_a_decimal_int > \l_fp_input_b_decimal_int
+              \group_end:
+              \prg_return_true:
+            \else:
+              \group_end:
+              \prg_return_false:
+            \fi:
+          \fi:
+        \fi:
+      \fi:
+    \fi:
+  }
+\cs_new_protected_nopar:cpn { fp_compare_absolute_a<b: }
+  {
+    \if_int_compare:w \l_fp_input_b_exponent_int > \l_fp_input_a_exponent_int
+      \group_end:
+      \prg_return_true:
+    \else:
+      \if_int_compare:w \l_fp_input_b_exponent_int < \l_fp_input_a_exponent_int
+        \group_end:
+        \prg_return_false:
+      \else:
+        \if_int_compare:w \l_fp_input_b_integer_int > \l_fp_input_a_integer_int
+          \group_end:
+          \prg_return_true:
+        \else:
+          \if_int_compare:w
+            \l_fp_input_b_integer_int < \l_fp_input_a_integer_int
+            \group_end:
+            \prg_return_false:
+          \else:
+            \if_int_compare:w
+              \l_fp_input_b_decimal_int > \l_fp_input_a_decimal_int
+              \group_end:
+              \prg_return_true:
+            \else:
+              \group_end:
+              \prg_return_false:
+            \fi:
+          \fi:
+        \fi:
+      \fi:
+    \fi:
+  }
+%    \end{macrocode}
+%   This is just a case of reversing the two input values and then
+%   running the tests already defined.
+%    \begin{macrocode}
+\cs_new_protected_nopar:cpn { fp_compare_<: }
+  {
+    \tl_set:Nx \l_fp_tmp_tl
+      {
+        \int_set:Nn \exp_not:N \l_fp_input_a_sign_int
+          { \int_use:N \l_fp_input_b_sign_int }
+        \int_set:Nn \exp_not:N \l_fp_input_a_integer_int
+          { \int_use:N \l_fp_input_b_integer_int }
+        \int_set:Nn \exp_not:N \l_fp_input_a_decimal_int
+          { \int_use:N \l_fp_input_b_decimal_int }
+        \int_set:Nn \exp_not:N \l_fp_input_a_exponent_int
+          { \int_use:N \l_fp_input_b_exponent_int }
+        \int_set:Nn \exp_not:N \l_fp_input_b_sign_int
+          { \int_use:N \l_fp_input_a_sign_int }
+        \int_set:Nn \exp_not:N \l_fp_input_b_integer_int
+          { \int_use:N \l_fp_input_a_integer_int }
+        \int_set:Nn \exp_not:N \l_fp_input_b_decimal_int
+          { \int_use:N \l_fp_input_a_decimal_int }
+        \int_set:Nn \exp_not:N \l_fp_input_b_exponent_int
+          { \int_use:N \l_fp_input_a_exponent_int }
+      }
+    \l_fp_tmp_tl
+    \use:c { fp_compare_>: }
+  }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}[TF]{\fp_compare:n}
+% \begin{macro}[aux]
+%   {
+%     \fp_compare_aux_i:w,  \fp_compare_aux_ii:w, \fp_compare_aux_iii:w,
+%     \fp_compare_aux_iv:w, \fp_compare_aux_v:w,  \fp_compare_aux_vi:w,
+%     \fp_compare_aux_vii:w
+%   }
+%   As \TeX{} cannot help out here, a daisy-chain of delimited functions
+%   are used. This is very much a first-generation approach: revision will
+%   be needed if these functions are really useful.
+%    \begin{macrocode}
+\prg_new_protected_conditional:Npnn \fp_compare:n #1 { T , F , TF }
+  {
+    \group_begin:
+      \tl_set:Nx \l_fp_tmp_tl
+        {
+          \group_end:
+          \fp_compare_aux_i:w #1 \exp_not:n { == \q_nil == \q_stop }
+        }
+     \l_fp_tmp_tl
+  }
+\cs_new_protected_nopar:Npn \fp_compare_aux_i:w #1 == #2 == #3 \q_stop
+  {
+    \quark_if_nil:nTF {#2}
+      { \fp_compare_aux_ii:w #1 != \q_nil != \q_stop }
+      { \fp_compare:nNnTF {#1} = {#2} \prg_return_true: \prg_return_false: }
+  }
+\cs_new_protected_nopar:Npn \fp_compare_aux_ii:w #1 != #2 != #3 \q_stop
+  {
+    \quark_if_nil:nTF {#2}
+      { \fp_compare_aux_iii:w #1 <= \q_nil <= \q_stop }
+      { \fp_compare:nNnTF {#1} = {#2} \prg_return_false: \prg_return_true: }
+  }
+\cs_new_protected_nopar:Npn \fp_compare_aux_iii:w #1 <= #2 <= #3 \q_stop
+  {
+    \quark_if_nil:nTF {#2}
+      { \fp_compare_aux_iv:w #1 >= \q_nil >= \q_stop }
+      { \fp_compare:nNnTF {#1} > {#2} \prg_return_false: \prg_return_true: }
+  }
+\cs_new_protected_nopar:Npn \fp_compare_aux_iv:w #1 >= #2 >= #3 \q_stop
+  {
+    \quark_if_nil:nTF {#2}
+      { \fp_compare_aux_v:w #1 = \q_nil  \q_stop }
+      { \fp_compare:nNnTF {#1} < {#2} \prg_return_false: \prg_return_true: }
+  }
+\cs_new_protected_nopar:Npn \fp_compare_aux_v:w #1 = #2 = #3 \q_stop
+  {
+    \quark_if_nil:nTF {#2}
+      { \fp_compare_aux_vi:w #1 < \q_nil < \q_stop }
+      { \fp_compare:nNnTF {#1} = {#2} \prg_return_true: \prg_return_false: }
+  }
+\cs_new_protected_nopar:Npn \fp_compare_aux_vi:w #1 < #2 < #3 \q_stop
+  {
+    \quark_if_nil:nTF {#2}
+      { \fp_compare_aux_vii:w #1 > \q_nil > \q_stop }
+      { \fp_compare:nNnTF {#1} < {#2} \prg_return_true: \prg_return_false: }
+  }
+\cs_new_protected_nopar:Npn \fp_compare_aux_vii:w #1 > #2 > #3 \q_stop
+  {
+    \quark_if_nil:nTF {#2}
+      { \prg_return_false: }
+      { \fp_compare:nNnTF {#1} > {#2} \prg_return_true: \prg_return_false: }
+  }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \subsection{Messages}
+%
+% \begin{macro}{\fp_overflow_msg:}
+% A generic overflow message, used whenever there is a possible
+% overflow.
+%    \begin{macrocode}
+\msg_kernel_new:nnnn { fpu } { overflow }
+  { Number~too~big. }
+  {
+    The~input~given~is~too~big~for~the~LaTeX~floating~point~unit. \\
+    Further~errors~may~well~occur!
+  }
+\cs_new_protected_nopar:Npn \fp_overflow_msg:
+  { \msg_kernel_error:nn { fpu } { overflow } }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\fp_exp_overflow_msg:}
+% A slightly more helpful message for exponent overflows.
+%    \begin{macrocode}
+\msg_kernel_new:nnnn { fpu } { exponent-overflow }
+  { Number~too~big~for~exponent~unit. }
+  {
+    The~exponent~of~the~input~given~is~too~big~for~the~floating~point~
+    unit:~the~maximum~input~value~for~an~exponent~is~230.
+  }
+\cs_new_protected_nopar:Npn \fp_exp_overflow_msg:
+  { \msg_kernel_error:nn { fpu } { exponent-overflow } }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\fp_ln_error_msg:}
+% Logarithms are only valid for positive number
+%    \begin{macrocode}
+\msg_kernel_new:nnnn { fpu } { logarithm-input-error }
+  { Invalid~input~to~ln~function. }
+  { Logarithms~can~only~be~calculated~for~positive~numbers. }
+\cs_new_protected_nopar:Npn \fp_ln_error_msg: {
+  \msg_kernel_error:nn { fpu } { logarithm-input-error }
+}
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\fp_trig_overflow_msg:}
+% A slightly more helpful message for trigonometric overflows.
+%    \begin{macrocode}
+\msg_kernel_new:nnnn { fpu } { trigonometric-overflow }
+  { Number~too~big~for~trigonometry~unit. }
+  {
+    The~trigonometry~code~can~only~work~with~numbers~smaller~
+    than~1000000000.
+  }
+\cs_new_protected_nopar:Npn \fp_trig_overflow_msg:
+  { \msg_kernel_error:nn { fpu } { trigonometric-overflow } }
+%    \end{macrocode}
+% \end{macro}
+%
+%    \begin{macrocode}
+%</initex|package>
+%    \end{macrocode}
+%
+% \end{implementation}
+%
+%\PrintIndex
diff --git a/Master/texmf-dist/source/latex/l3kernel/l3int.dtx b/Master/texmf-dist/source/latex/l3kernel/l3int.dtx
new file mode 100644
index 00000000000..70264286d27
--- /dev/null
+++ b/Master/texmf-dist/source/latex/l3kernel/l3int.dtx
@@ -0,0 +1,2197 @@
+% \iffalse meta-comment
+%
+%% File: l3int.dtx Copyright (C) 1990-2011 The LaTeX3 Project
+%%
+%% It may be distributed and/or modified under the conditions of the
+%% LaTeX Project Public License (LPPL), either version 1.3c of this
+%% license or (at your option) any later version.  The latest version
+%% of this license is in the file
+%%
+%%    http://www.latex-project.org/lppl.txt
+%%
+%% This file is part of the "expl3 bundle" (The Work in LPPL)
+%% and all files in that bundle must be distributed together.
+%%
+%% The released version of this bundle is available from CTAN.
+%%
+%% -----------------------------------------------------------------------
+%%
+%% The development version of the bundle can be found at
+%%
+%%    http://www.latex-project.org/svnroot/experimental/trunk/
+%%
+%% for those people who are interested.
+%%
+%%%%%%%%%%%
+%% NOTE: %%
+%%%%%%%%%%%
+%%
+%%   Snapshots taken from the repository represent work in progress and may
+%%   not work or may contain conflicting material!  We therefore ask
+%%   people _not_ to put them into distributions, archives, etc. without
+%%   prior consultation with the LaTeX3 Project.
+%%
+%% -----------------------------------------------------------------------
+%
+%<*driver|package>
+\RequirePackage{l3names}
+\GetIdInfo$Id: l3int.dtx 2478 2011-06-19 21:34:23Z joseph $
+  {L3 Experimental integers}
+%</driver|package>
+%<*driver>
+\documentclass[full]{l3doc}
+\begin{document}
+  \DocInput{\jobname.dtx}
+\end{document}
+%</driver>
+% \fi
+%
+% \title{^^A
+%   The \pkg{l3int} package\\ Integers^^A
+%   \thanks{This file describes v\ExplFileVersion,
+%      last revised \ExplFileDate.}^^A
+% }
+%
+% \author{^^A
+%  The \LaTeX3 Project\thanks
+%    {^^A
+%      E-mail:
+%        \href{mailto:latex-team@latex-project.org}
+%          {latex-team@latex-project.org}^^A
+%    }^^A
+% }
+%
+% \date{Released \ExplFileDate}
+%
+% \maketitle
+%
+% \begin{documentation}
+%
+% Calculation and comparison of integer values can be carried out
+% using literal numbers, \texttt{int} registers, constants and
+% integers stored in token list variables. The standard operators
+% \texttt{+}, \texttt{-}, \texttt{/} and \texttt{*} and
+% parentheses can be used within such expressions to carry
+% arithmetic operations. This module carries out these functions
+% on \emph{integer expressions} (\enquote{\texttt{int expr}}).
+%
+% \section{Integer expressions}
+%
+% \begin{function}[EXP]{\int_eval:n}
+%   \begin{syntax}
+%     \cs{int_eval:n} \Arg{integer expression}
+%   \end{syntax}
+%   Evaluates the \meta{integer expression}, expanding any
+%   integer and token list variables within the \meta{expression}
+%   to their content (without requiring \cs{int_use:N}/\cs{tl_use:N})
+%   and applying the standard mathematical rules. For example both
+%   \begin{verbatim}
+%     \int_eval:n { 5 +  4 * 3 - ( 3 + 4 * 5 ) }
+%   \end{verbatim}
+%   and
+%   \begin{verbatim}
+%     \tl_new:N  \l_my_tl
+%     \tl_set:Nn \l_my_tl { 5 }
+%     \int_new:N  \l_my_int
+%     \int\set:Nn \l_my_int { 4 }
+%    \int_eval:n { \l_my_tl +  \l_my_int * 3 - ( 3 + 4 * 5 ) }
+%   \end{verbatim}
+%   both evaluate to \( -6 \). The  \Arg{integer expression} may
+%   contain the operators \texttt{+}, \texttt{-}, \texttt{*} and
+%   \texttt{/}, along with parenthesis \texttt{(} and \texttt{)}.
+%   After two expansions, \cs{int_eval:n} yields a
+%   \meta{integer denotation} which is left in the input stream. This is
+%   \emph{not} an \meta{internal integer}, and therefore requires
+%   suitable termination if used in a \TeX{}-style integer assignment.
+% \end{function}
+%
+% \begin{function}[EXP]{\int_abs:n}
+%   \begin{syntax}
+%     \cs{int_abs:n} \Arg{integer expression}
+%   \end{syntax}
+%   Evaluates the \meta{integer expression} as described for
+%   \cs{int_eval:n} and leaves the absolute value of the result in
+%   the input stream as an \meta{integer denotation} after two
+%   expansions.
+% \end{function}
+%
+% \begin{function}[EXP]{\int_div_round:nn}
+%   \begin{syntax}
+%     \cs{int_div_round:nn} \Arg{intexpr1} \Arg{intexpr2}
+%   \end{syntax}
+%   Evaluates the two \meta{integer expressions} as described earlier,
+%   then calculates the result of dividing the first value by the
+%    second, round any remainder. Note that this is identical to using
+%   |/| directly in an \meta{integer expression}. The result is left in
+%   the input stream as a \meta{integer denotation} after two expansions.
+% \end{function}
+%
+% \begin{function}[EXP]{\int_div_truncate:nn}
+%   \begin{syntax}
+%     \cs{int_div_truncate:nn} \Arg{intexpr1} \Arg{intexpr2}
+%   \end{syntax}
+%   Evaluates the two \meta{integer expressions} as described earlier,
+%   then calculates the result of dividing the first value by the
+%   second, truncating any remainder. Note that division using |/|
+%   rounds the result. The result is left in the input stream as a
+%   \meta{integer denotation} after two expansions.
+% \end{function}
+%
+% \begin{function}[EXP]{\int_max:nn, \int_min:nn}
+%   \begin{syntax}
+%     \cs{int_max:nn} \Arg{intexpr1} \Arg{intexpr2}
+%     \cs{int_min:nn} \Arg{intexpr1} \Arg{intexpr2}
+%   \end{syntax}
+%   Evaluates the \meta{integer expressions} as described for
+%   \cs{int_eval:n} and leaves either the larger or smaller value
+%   in the input stream as an \meta{integer denotation} after two
+%   expansions.
+% \end{function}
+%
+% \begin{function}[EXP]{\int_mod:nn}
+%   \begin{syntax}
+%     \cs{int_mod:nn} \Arg{intexpr1} \Arg{intexpr2}
+%   \end{syntax}
+%   Evaluates the two \meta{integer expressions} as described earlier,
+%   then calculates the integer remainder of dividing the first
+%   expression by the second. This is left in the input stream as an
+%   \meta{integer denotation} after two expansions.
+% \end{function}
+%
+% \section{Creating and initialising integers}
+%
+% \begin{function}{\int_new:N, \int_new:c}
+%   \begin{syntax}
+%     \cs{int_new:N} \meta{integer}
+%   \end{syntax}
+%   Creates a new \meta{integer} or raises an error if the name is
+%   already taken. The declaration is global. The \meta{integer} will
+%   initially be equal to $0$.
+% \end{function}
+%
+% \begin{function}{\int_const:Nn, \int_const:cn}
+%   \begin{syntax}
+%     \cs{int_const:Nn} \meta{integer} \Arg{integer expression}
+%   \end{syntax}
+%   Creates a new constant \meta{integer} or raises an error if the name
+%   is already taken. The value of the \meta{integer} will be set
+%   globally to the \meta{integer expression}.
+% \end{function}
+%
+% \begin{function}{\int_zero:N, \int_zero:c}
+%   \begin{syntax}
+%     \cs{int_zero:N} \meta{integer}
+%   \end{syntax}
+%   Sets \meta{integer} to $0$ within the scope of the current \TeX{} group.
+% \end{function}
+%
+% \begin{function}{\int_gzero:N, \int_gzero:c}
+%   \begin{syntax}
+%     \cs{int_gzero:N} \meta{integer}
+%   \end{syntax}
+%   Sets \meta{integer} to $0$ globally, \emph{i.e.}~not
+%   restricted by the current \TeX{} group level.
+% \end{function}
+%
+% \begin{function}
+%   {\int_set_eq:NN, \int_set_eq:cN, \int_set_eq:Nc, \int_set_eq:cc}
+%   \begin{syntax}
+%     \cs{int_set_eq:NN} \meta{integer1} \meta{integer2}
+%   \end{syntax}
+%   Sets the content of \meta{integer1} equal to that of
+%   \meta{integer2}. This assignment is restricted to the current
+%   \TeX{} group level.
+% \end{function}
+%
+% \begin{function}
+%   {\int_gset_eq:NN, \int_gset_eq:cN, \int_gset_eq:Nc, \int_gset_eq:cc}
+%   \begin{syntax}
+%     \cs{int_gset_eq:NN} \meta{integer1} \meta{integer2}
+%   \end{syntax}
+%   Sets the content of \meta{integer1} equal to that of \meta{integer2}.
+%   This assignment is global and so is not limited by the current
+%   \TeX{} group level.
+% \end{function}
+%
+% \section{Setting and incrementing integers}
+%
+% \begin{function}{\int_add:Nn, \int_add:cn}
+%   \begin{syntax}
+%     \cs{int_add:Nn} \meta{integer} \Arg{integer expression}
+%   \end{syntax}
+%   Adds the result of the \meta{integer expression} to the current
+%   content of the \meta{integer}. This assignment is local.
+% \end{function}
+%
+% \begin{function}{\int_gadd:Nn, \int_gadd:cn}
+%   \begin{syntax}
+%     \cs{int_gadd:Nn} \meta{integer} \Arg{integer expression}
+%   \end{syntax}
+%   Adds the result of the \meta{integer expression} to the current
+%   content of the \meta{integer}. This assignment is global.
+% \end{function}
+%
+% \begin{function}{\int_decr:N, \int_decr:c}
+%   \begin{syntax}
+%     \cs{int_decr:N} \meta{integer}
+%   \end{syntax}
+%   Decreases the value stored in \meta{integer} by $1$ within
+%   the scope of the current \TeX{} group.
+% \end{function}
+%
+% \begin{function}{\int_gdecr:N, \int_gdecr:c}
+%   \begin{syntax}
+%     \cs{int_incr:N} \meta{integer}
+%   \end{syntax}
+%   Decreases the value stored in \meta{integer} by $1$ globally
+%   (\emph{i.e.}~not limited by the current group level).
+% \end{function}
+%
+% \begin{function}{\int_incr:N, \int_incr:c}
+%   \begin{syntax}
+%     \cs{int_incr:N} \meta{integer}
+%   \end{syntax}
+%   Increases the value stored in \meta{integer} by $1$ within
+%   the scope of the current \TeX{} group.
+% \end{function}
+%
+% \begin{function}{\int_gincr:N, \int_gincr:c}
+%   \begin{syntax}
+%     \cs{int_incr:N} \meta{integer}
+%   \end{syntax}
+%   Increases the value stored in \meta{integer} by $1$ globally
+%   (\emph{i.e.}~not limited by the current group level).
+% \end{function}
+%
+% \begin{function}{\int_set:Nn, \int_set:cn}
+%   \begin{syntax}
+%     \cs{int_set:Nn} \meta{integer} \Arg{integer expression}
+%   \end{syntax}
+%   Sets \meta{integer} to the value of \meta{integer expression},
+%   which must evaluate to an integer (as described for
+%   \cs{int_eval:n}).  This assignment is restricted to the
+%   current \TeX{} group.
+% \end{function}
+%
+% \begin{function}{\int_gset:Nn, \int_gset:cn}
+%   \begin{syntax}
+%     \cs{int_gset:Nn} \meta{integer} \Arg{integer expression}
+%   \end{syntax}
+%   Sets \meta{integer} to the value of \meta{integer expression},
+%   which must evaluate to an integer (as described for
+%   \cs{int_eval:n}). This assignment is global and is not limited
+%   to the current \TeX{} group level.
+% \end{function}
+%
+% \begin{function}{\int_sub:Nn, \int_sub:cn}
+%   \begin{syntax}
+%     \cs{int_sub:Nn} \meta{integer} \Arg{integer expression}
+%   \end{syntax}
+%   Subtracts the result of the \meta{integer expression} to the
+%   current content of the \meta{integer}. This assignment is local.
+% \end{function}
+%
+% \begin{function}{\int_gsub:Nn, \int_gsub:cn}
+%   \begin{syntax}
+%     \cs{int_gsub:Nn} \meta{integer} \Arg{integer expression}
+%   \end{syntax}
+%   Subtracts the result of the \meta{integer expression} to the
+%   current content of the \meta{integer}. This assignment is global.
+% \end{function}
+%
+% \section{Using integers}
+%
+% \begin{function}[EXP]{\int_use:N, \int_use:c}
+%   \begin{syntax}
+%     \cs{int_use:N} \meta{integer}
+%   \end{syntax}
+%   Recovers the content of a \meta{integer} and places it directly
+%   in the input stream. An error will be raised if the variable does
+%   not exist or if it is invalid. Can be omitted in places where a
+%   \meta{integer} is required (such as in the first and third arguments
+%   of \cs{int_compare:nNnTF}).
+%   \begin{texnote}
+%     \cs{int_use:N} is the \TeX{} primitive \cs{the}: this is one of
+%     several \LaTeX3 names for this primitive.
+%   \end{texnote}
+% \end{function}
+%
+% \section{Integer expression conditionals}
+%
+% \begin{function}[EXP,pTF]{\int_compare:nNn}
+%   \begin{syntax}
+%     \cs{int_compare_p:nNn}
+%     ~~\Arg{intexpr1} \meta{relation} \Arg{intexpr2}
+%     \cs{int_compare:nNnTF}
+%     ~~\Arg{intexpr1} \meta{relation} \Arg{intexpr2}
+%     ~~\Arg{true code} \Arg{false code}
+%   \end{syntax}
+%   This function first evaluates each of the \meta{integer expressions}
+%   as described for \cs{int_eval:n}. The two results are then
+%   compared using the \meta{relation}:
+%   \begin{center}
+%     \begin{tabular}{ll}
+%       Equal                 & |=| \\
+%       Greater than          & |>| \\
+%       Less than             & |<| \\
+%     \end{tabular}
+%   \end{center}
+%   The branching versions then leave either \meta{true code} or
+%   \meta{false code} in the input stream, as appropriate to the truth
+%   of the test and the variant of the function chosen. The logical
+%   truth of the test is left in the input stream by the predicate
+%   version.
+% \end{function}
+%
+% \begin{function}[EXP,pTF]{\int_compare:n}
+%   \begin{syntax}
+%     \cs{int_compare_p:n}
+%     ~~\{ \meta{intexpr1} \meta{relation} \meta{intexpr2} \}
+%     \cs{int_compare:nTF}
+%     ~~\{ \meta{intexpr1} \meta{relation} \meta{intexpr2} \}
+%     ~~\Arg{true code} \Arg{false code}
+%   \end{syntax}
+%   This function first evaluates each of the \meta{integer expressions}
+%   as described for \cs{int_eval:n}. The two results are then
+%   compared using the \meta{relation}:
+%   \begin{center}
+%     \begin{tabular}{ll}
+%       Equal                    & |=| or |==| \\
+%       Greater than or equal to & |=>|        \\
+%       Greater than             & |>|         \\
+%       Less than or equal to    & |=<|        \\
+%       Less than                & |<|         \\
+%       Not equal                & |!=|        \\
+%     \end{tabular}
+%   \end{center}
+%   The branching versions then leave either \meta{true code} or
+%   \meta{false code} in the input stream, as appropriate to the truth
+%   of the test and the variant of the function chosen. The logical
+%   truth of the test is left in the input stream by the predicate
+%   version.
+% \end{function}
+%
+% \begin{function}[EXP,pTF]{\int_if_even:n, \int_if_odd:n}
+%   \begin{syntax}
+%     \cs{int_if_odd_p:n} \Arg{integer expression}
+%     \cs{int_if_odd:nTF} \Arg{integer expression}
+%     ~~\Arg{true code} \Arg{false code}
+%   \end{syntax}
+%   This function first evaluates the \meta{integer expression}
+%   as described for \cs{int_eval:n}. It then evaluates if this
+%   is odd or even, as appropriate. The branching versions then leave
+%   either \meta{true code} or \meta{false code} in the input stream,
+%   as appropriate to the truth of the test and the variant of the
+%   function chosen. The logical truth of the test is left in the input
+%   stream by the predicate version.
+% \end{function}
+%
+% \section{Integer expression loops}
+%
+% \begin{function}[EXP]{\int_do_while:nNnn}
+%   \begin{syntax}
+%      \cs{int_do_while:nNnn}
+%      ~~\Arg{intexpr1} \meta{relation} \Arg{intexpr2} \Arg{code}
+%   \end{syntax}
+%   Evaluates the relationship between the two \meta{integer expressions}
+%   as described for \cs{int_compare:nNnTF}, and then places the
+%   \meta{code} in the input stream if the \meta{relation} is
+%   \texttt{true}. After the \meta{code} has been processed by \TeX{} the
+%   test will be repeated, and a loop will occur until the test is
+%   \texttt{false}.
+% \end{function}
+%
+% \begin{function}[EXP]{\int_do_until:nNnn}
+%   \begin{syntax}
+%      \cs{int_do_until:nNnn}
+%      ~~\Arg{intexpr1} \meta{relation} \Arg{intexpr2} \Arg{code}
+%   \end{syntax}
+%   Evaluates the relationship between the two \meta{integer expressions}
+%   as described for \cs{int_compare:nNnTF}, and then places the
+%   \meta{code} in the input stream if the \meta{relation} is
+%   \texttt{false}. After the \meta{code} has been processed by \TeX{} the
+%   test will be repeated, and a loop will occur until the test is
+%   \texttt{true}.
+% \end{function}
+%
+% \begin{function}[EXP]{\int_until_do:nNnn}
+%   \begin{syntax}
+%      \cs{int_until_do:nNnn}
+%      ~~\Arg{intexpr1} \meta{relation} \Arg{intexpr2} \Arg{code}
+%   \end{syntax}
+%   Places the \meta{code} in the input stream for \TeX{} to process, and
+%   then evaluates the relationship between the two
+%   \meta{integer expressions} as described for \cs{int_compare:nNnTF}.
+%   If the test is \texttt{false} then the \meta{code} will be inserted
+%   into the input stream again and a loop will occur until the
+%   \meta{relation} is \texttt{true}.
+% \end{function}
+%
+% \begin{function}[EXP]{\int_while_do:nNnn}
+%   \begin{syntax}
+%      \cs{int_while_do:nNnn} \
+%      ~~\Arg{intexpr1} \meta{relation} \Arg{intexpr2} \Arg{code}
+%   \end{syntax}
+%   Places the \meta{code} in the input stream for \TeX{} to process, and
+%   then evaluates the relationship between the two
+%   \meta{integer expressions} as described for \cs{int_compare:nNnTF}.
+%   If the test is \texttt{true} then the \meta{code} will be inserted
+%   into the input stream again and a loop will occur until the
+%   \meta{relation} is \texttt{false}.
+% \end{function}
+%
+% \begin{function}[EXP]{\int_do_while:nn}
+%   \begin{syntax}
+%      \cs{int_do_while:nNnn}
+%      ~~\{ \meta{intexpr1} \meta{relation} \meta{intexpr2} \} \Arg{code}
+%   \end{syntax}
+%   Evaluates the relationship between the two \meta{integer expressions}
+%   as described for \cs{int_compare:nTF}, and then places the
+%   \meta{code} in the input stream if the \meta{relation} is
+%   \texttt{true}. After the \meta{code} has been processed by \TeX{} the
+%   test will be repeated, and a loop will occur until the test is
+%   \texttt{false}.
+% \end{function}
+%
+% \begin{function}[EXP]{\int_do_until:nn}
+%   \begin{syntax}
+%      \cs{int_do_until:nn}
+%      ~~\{ \meta{intexpr1} \meta{relation} \meta{intexpr2} \} \Arg{code}
+%   \end{syntax}
+%   Evaluates the relationship between the two \meta{integer expressions}
+%   as described for \cs{int_compare:nTF}, and then places the
+%   \meta{code} in the input stream if the \meta{relation} is
+%   \texttt{false}. After the \meta{code} has been processed by \TeX{} the
+%   test will be repeated, and a loop will occur until the test is
+%   \texttt{true}.
+% \end{function}
+%
+% \begin{function}[EXP]{\int_until_do:nn}
+%   \begin{syntax}
+%      \cs{int_until_do:nn}
+%      ~~\{ \meta{intexpr1} \meta{relation} \meta{intexpr2} \} \Arg{code}
+%   \end{syntax}
+%   Places the \meta{code} in the input stream for \TeX\ to process, and
+%   then evaluates the relationship between the two
+%   \meta{integer expressions} as described for \cs{int_compare:nTF}.
+%   If the test is \texttt{false} then the \meta{code} will be inserted
+%   into the input stream again and a loop will occur until the
+%   \meta{relation} is \texttt{true}.
+% \end{function}
+%
+% \begin{function}[EXP]{\int_while_do:nn}
+%   \begin{syntax}
+%      \cs{int_while_do:nn} \
+%      ~~\{ \meta{intexpr1} \meta{relation} \meta{intexpr2} \} \Arg{code}
+%   \end{syntax}
+%   Places the \meta{code} in the input stream for \TeX{} to process, and
+%   then evaluates the relationship between the two
+%   \meta{integer expressions} as described for \cs{int_compare:nTF}.
+%   If the test is \texttt{true} then the \meta{code} will be inserted
+%   into the input stream again and a loop will occur until the
+%   \meta{relation} is \texttt{false}.
+% \end{function}
+%
+% \section{Formatting integers}
+%
+% Integers can be placed into the output stream with formatting. These
+% conversions apply to any integer expressions.
+%
+% \begin{function}[EXP]{\int_to_arabic:n}
+%   \begin{syntax}
+%     \cs{int_to_arabic:n} \Arg{integer expression}
+%   \end{syntax}
+%   Places the value of the \meta{integer expression} in the input
+%   stream as digits, with category code $12$ (other).
+% \end{function}
+%
+% \begin{function}[EXP]{\int_to_alph:n, \int_to_Alph:n}
+%   \begin{syntax}
+%     \cs{int_to_alph:n} \Arg{integer expression}
+%   \end{syntax}
+%   Evaluates the \meta{integer expression} and converts the result
+%   into a series of letters, which are then left in the input stream.
+%   The conversion rule uses the $26$ letters of the English
+%   alphabet, in order. Thus
+%   \begin{verbatim}
+%     \int_to_alph:n { 1 }
+%   \end{verbatim}
+%   places |a| in the input stream,
+%   \begin{verbatim}
+%     \int_to_alph:n { 26 }
+%   \end{verbatim}
+%   is represented as |z| and
+%   \begin{verbatim}
+%     \int_to_alph:n { 27 }
+%   \end{verbatim}
+%   is converted to |aa|. For conversions using other alphabets, use
+%   \cs{int_convert_to_symbols:nnn} to define an alphabet-specific
+%   function. The basic \cs{int_to_alph:n} and \cs{int_to_Alph:n}
+%   functions should not be modified.
+% \end{function}
+%
+% \begin{function}[EXP]{\int_to_symbols:nnn}
+%   \begin{syntax}
+%     \cs{int_to_symbols:nnn}
+%     ~~\Arg{integer expression} \Arg{total symbols}
+%     ~~\meta{value to symbol mapping}
+%   \end{syntax}
+%   This is the low-level function for conversion of an
+%   \meta{integer expression} into a symbolic form (which will often
+%   be letters). The \meta{total symbols} available should be given
+%   as an integer expression. Values are actually converted to symbols
+%   according to the \meta{value to symbol mapping}. This should be given
+%   as \meta{total symbols} pairs of entries, a number and the
+%   appropriate symbol. Thus the \cs{int_to_alph:n} function is defined
+%   as
+%   \begin{verbatim}
+%     \cs_new:Npn \int_to_alph:n #1
+%       {
+%         \int_convert_to_sybols:nnn {#1} { 26 }
+%           {
+%             {  1 } { a }
+%             {  2 } { b }
+%             {  3 } { c }
+%             {  4 } { d }
+%             {  5 } { e }
+%             {  6 } { f }
+%             {  7 } { g }
+%             {  8 } { h }
+%             {  9 } { i }
+%             { 10 } { j }
+%             { 11 } { k }
+%             { 12 } { l }
+%             { 13 } { m }
+%             { 14 } { n }
+%             { 15 } { o }
+%             { 16 } { p }
+%             { 17 } { q }
+%             { 18 } { r }
+%             { 19 } { s }
+%             { 20 } { t }
+%             { 21 } { u }
+%             { 22 } { v }
+%             { 23 } { w }
+%             { 24 } { x }
+%             { 25 } { y }
+%             { 26 } { z }
+%           }
+%       }
+%   \end{verbatim}
+% \end{function}
+%
+% \begin{function}[EXP]{\int_to_binary:n}
+%   \begin{syntax}
+%     \cs{int_to_binary:n} \Arg{integer expression}
+%   \end{syntax}
+%   Calculates the value of the \meta{integer expression} and places
+%   the binary representation of the result in the input stream.
+% \end{function}
+%
+% \begin{function}[EXP]{\int_to_hexadecimal:n}
+%   \begin{syntax}
+%     \cs{int_to_binary:n} \Arg{integer expression}
+%   \end{syntax}
+%   Calculates the value of the \meta{integer expression} and places
+%   the hexadecimal (base~$16$) representation of the result in the
+%   input stream. Upper case letters are used for digits beyond $9$.
+% \end{function}
+%
+% \begin{function}[EXP]{\int_to_octal:n}
+%   \begin{syntax}
+%     \cs{int_to_octal:n} \Arg{integer expression}
+%   \end{syntax}
+%   Calculates the value of the \meta{integer expression} and places
+%   the octal (base~$8$) representation of the result in the input
+%   stream.
+% \end{function}
+%
+% \begin{function}[EXP]{\int_to_base:nn}
+%   \begin{syntax}
+%     \cs{int_to_base:nn} \Arg{integer expression} \Arg{base}
+%   \end{syntax}
+%   Calculates the value of the \meta{integer expression} and
+%   converts it into the appropriate representation in the \meta{base};
+%   the later may be given as an integer expression. For bases greater
+%   than $10$ the higher \enquote{digits} are represented by the upper case
+%   letters from the English alphabet.
+%   The maximum \meta{base} value is $36$.
+%   \begin{texnote}
+%     This is a generic version of \cs{int_to_binary:n}, \emph{etc.}
+%   \end{texnote}
+% \end{function}
+%
+% \begin{function}[EXP]{\int_to_roman:n, \int_to_Roman:n}
+%   \begin{syntax}
+%     \cs{int_to_roman:n} \Arg{integer expression}
+%   \end{syntax}
+%   Places the value of the \meta{integer expression} in the input
+%   stream as Roman numerals, either lower case (\cs{int_to_roman:n})
+%   or upper case (\cs{int_to_Roman:n}). The Roman numerals are letters
+%   with category code $11$ (letter).
+% \end{function}
+%
+% \begin{function}[EXP]{\int_to_symbol:n}
+%   \begin{syntax}
+%     \cs{int_to_symbol:n} \Arg{integer expression}
+%   \end{syntax}
+%   Calculates the value of the \meta{integer expression} and places
+%   the symbol representation of the result in the input stream. The
+%   list of symbols used is equivalent to \LaTeXe{}'s \cs{@fnsymbol}
+%   set.
+% \end{function}
+%
+% \section{Converting from other formats to integers}
+%
+% \begin{function}[EXP]{\int_from_alph:n}
+%   \begin{syntax}
+%     \cs{int_from_alpa:n} \Arg{letters}
+%   \end{syntax}
+%   Converts the \meta{letters} into the integer (base~$10$)
+%   representation and leaves this in the input stream. The
+%   \meta{letters} are treated using the English alphabet only, with
+%   \enquote{a} equal to $1$ through to \enquote{z} equal to $26$. Either lower
+%   or upper case letters may be used. This is the inverse function of
+%   \cs{int_to_alph:n}.
+% \end{function}
+%
+% \begin{function}[EXP]{\int_from_binary:n}
+%   \begin{syntax}
+%     \cs{int_from_binary:n} \Arg{binary number}
+%   \end{syntax}
+%   Converts the \meta{binary number} into the integer (base~$10$)
+%   representation and leaves this in the input stream.
+% \end{function}
+%
+% \begin{function}[EXP]{\int_from_hexadecimal:n}
+%   \begin{syntax}
+%     \cs{int_from_binary:n} \Arg{hexadecimal number}
+%   \end{syntax}
+%   Converts the \meta{hexadecimal number} into the integer
+%   (base~$10$) representation and leaves this in the input stream.
+%   Digits greater than $9$ may be represented in the
+%   \meta{hexadecimal number} by upper or lower case letters.
+% \end{function}
+%
+% \begin{function}[EXP]{\int_from_octal:n}
+%   \begin{syntax}
+%     \cs{int_from_octal:n} \Arg{octal number}
+%   \end{syntax}
+%   Converts the \meta{octal number} into the integer (base~$10$)
+%   representation and leaves this in the input stream.
+% \end{function}
+%
+% \begin{function}[EXP]{\int_from_roman:n}
+%   \begin{syntax}
+%     \cs{int_from_roman:n} \Arg{roman numeral}
+%   \end{syntax}
+%   Converts the \meta{roman numeral} into the integer (base~$10$)
+%   representation and leaves this in the input stream. The
+%   \meta{roman numeral} may be in upper or lower case; if the numeral
+%   is not valid then the resulting value will be $-1$.
+% \end{function}
+%
+% \begin{function}[EXP]{\int_from_base:nn}
+%   \begin{syntax}
+%     \cs{int_from_base:nn} \Arg{number}
+%     ~~\Arg{base}
+%   \end{syntax}
+%   Converts the \meta{number} in \meta{base} into the appropriate
+%   value in base $10$. The \meta{number} should consist of
+%   digits and letters (either lower or upper case), plus optionally
+%   a leading sign. The maximum \meta{base} value is $36$.
+% \end{function}
+%
+% \section{Viewing integers}
+%
+% \begin{function}{\int_show:N, \int_show:c}
+%   \begin{syntax}
+%     \cs{int_show:N} \meta{integer}
+%   \end{syntax}
+%   Displays the value of the \meta{integer} on the terminal.
+% \end{function}
+%
+% \section{Constant integers}
+%
+% \begin{variable}
+%   {
+%     \c_minus_one              ,
+%     \c_zero                   ,
+%     \c_one                    ,
+%     \c_two                    ,
+%     \c_three                  ,
+%     \c_four                   ,
+%     \c_five                   ,
+%     \c_six                    ,
+%     \c_seven                  ,
+%     \c_eight                  ,
+%     \c_nine                   ,
+%     \c_ten                    ,
+%     \c_eleven                 ,
+%     \c_twelve                 ,
+%     \c_thirteen               ,
+%     \c_fourteen               ,
+%     \c_fifteen                ,
+%     \c_sixteen                ,
+%     \c_thirty_two             ,
+%     \c_one_hundred            ,
+%     \c_two_hundred_fifty_five ,
+%     \c_two_hundred_fifty_six  ,
+%     \c_one_thousand           ,
+%     \c_ten_thousand
+% }
+%   Integer values used with primitive tests and assignments:
+%   self-terminating nature makes these more convenient and faster than
+%   literal numbers.
+% \end{variable}
+%
+% \begin{variable}{\c_max_int}
+%   The maximum value that can be stored as an integer.
+% \end{variable}
+%
+% \begin{variable}{\c_max_register_int}
+%   Maximum number of registers.
+% \end{variable}
+%
+% \section{Scratch integers}
+%
+% \begin{variable}{\l_tmpa_int, \l_tmpb_int, \l_tmpc_int}
+%   Scratch integer for local assignment. These are never used by
+%   the kernel code, and so are safe for use with any \LaTeX3-defined
+%   function. However, they may be overwritten by other non-kernel
+%   code and so should only be used for short-term storage.
+% \end{variable}
+%
+% \begin{variable}{\g_tmpa_int, \g_tmpb_int}
+%   Scratch integer for global assignment. These are never used by
+%   the kernel code, and so are safe for use with any \LaTeX3-defined
+%   function. However, they may be overwritten by other non-kernel
+%   code and so should only be used for short-term storage.
+% \end{variable}
+%
+% \section{Internal functions}
+%
+% \begin{function}[EXP]{int_get_digits:n}
+%   \begin{syntax}
+%     \cs{int_get_digits:n} \meta{value}
+%   \end{syntax}
+%     Parses the \meta{value} to leave the absolute \meta{value}
+%     in the input stream. This may therefore be used
+%     to remove multiple sign tokens from the \meta{value}
+%     (which may be symbolic).
+% \end{function}
+%
+% \begin{function}[EXP]{int_get_sign:n}
+%   \begin{syntax}
+%     \cs{int_get_sign:n} \meta{value}
+%   \end{syntax}
+%     Parses the \meta{value} to leave a single sign token
+%     (either |+| or |-|) in the input stream. This may therefore be used
+%     to sanitise sign tokens from the \meta{value} (which may be symbolic).
+% \end{function}
+%
+% \begin{function}[EXP]{int_to_letter:n}
+%   \begin{syntax}
+%     \cs{int_to_letter:n} \meta{integer value}
+%   \end{syntax}
+%   For \meta{integer values} from $0$ to $9$, leaves the \meta{value}
+%   in the input stream unchanged. For \meta{integer values} from
+%   $10$ to $35$, leaves the appropriate upper case letter (from the
+%   standard English alphabet) in the input stream: for example,
+%   $10$ is converted to |A|, $11$ to |B|, \emph{etc.}
+% \end{function}
+%
+% \begin{function}[EXP]{\int_to_roman:w}
+%   \begin{syntax}
+%     \cs{int_to_roman:w} \meta{integer}
+%     ~~\meta{space} \textit{or} \meta{non-expandable token}
+%   \end{syntax}
+%   Converts \meta{integer} to it lower case Roman representation. Expansion
+%   ends when a space or non-expandable token is found.
+%   Note that this function produces a string of letters with category code
+%   $12$ and that protected functions \emph{are} expanded by this
+%   process. Negative \meta{integer} values result in no output, although
+%   the function does not terminate expansion until a suitable endpoint
+%   is found in the same way as for positive numbers.
+%   \begin{texnote}
+%     This is the \TeX{} primitive \tn{romannumeral} renamed.
+%   \end{texnote}
+% \end{function}
+%
+% \begin{function}[EXP]{\if_num:w, \if_int_compare:w}
+%   \begin{syntax}
+%     \cs{if_num:w} \meta{integer1} \meta{relation} \meta{integer2}
+%     ~~\meta{true code}
+%     \cs{else:}
+%     ~~\meta{false code}
+%     \cs{fi:}
+%   \end{syntax}
+%   Compare two integers using \meta{relation}, which must be one of
+%   |=|, |<| or |>| with category code $12$.
+%   The \cs{else:} branch is optional.
+%   \begin{texnote}
+%     These are both names for the \TeX{} primitive \cs{ifnum}.
+%   \end{texnote}
+% \end{function}
+%
+% \begin{function}[EXP]{\if_case:w, \or:}
+%   \begin{syntax}
+%     \cs{if_case:w} \meta{integer} \meta{case0}
+%     ~~\cs{or:} \meta{case1}
+%     ~~\cs{or:} |...|
+%     ~~\cs{else:} \meta{default}
+%     \cs{fi:}
+%   \end{syntax}
+%   Selects a case to execute based on the value of the \meta{integer}. The
+%   first case (\meta{case0}) is executed if \meta{integer} is $0$, the second
+%   (\meta{case1}) if the \meta{integer} is $1$, \emph{etc.} The
+%   \meta{integer} may be a literal, a constant or an integer
+%   expression (\emph{e.g.}~using \cs{int_eval:n}).
+%   \begin{texnote}
+%     These are the \TeX{} primitives \cs{ifcase}  and \cs{or}.
+%   \end{texnote}
+% \end{function}
+%
+% \begin{function}[EXP]{\int_value:w}
+%   \begin{syntax}
+%     \cs{int_value:w} \meta{integer}
+%     \cs{int_value:w} \meta{tokens} \meta{optional space}
+%   \end{syntax}
+%   Expands \meta{tokens} until an \meta{integer} is formed. One space may be
+%   gobbled in the process.
+%   \begin{texnote}
+%     This is the \TeX{} primitive \cs{number}.
+%   \end{texnote}
+% \end{function}
+%
+% \begin{function}[EXP]{\int_eval:w, \int_eval_end:}
+%   \begin{syntax}
+%     \cs{int_eval:w} \meta{intexpr} \cs{int_eval_end:}
+%   \end{syntax}
+%   Evaluates \meta{integer expression} as described for \cs{int_eval:n}.
+%   The evaluation stops when an unexpandable token which is not a valid
+%   part of an integer is read or when \cs{int_eval_end:} is
+%   reached. The latter is gobbled by the scanner mechanism:
+%   \cs{int_eval_end:} itself is unexpandable but used correctly
+%   the entire construct is expandable.
+%   \begin{texnote}
+%     This is the \eTeX{} primitive \cs{numexpr}.
+%   \end{texnote}
+% \end{function}
+%
+% \begin{function}[EXP]{\if_int_odd:w}
+%   \begin{syntax}
+%     \cs{if_int_odd:w} \meta{tokens}  \meta{optional space}
+%     ~~\meta{true code}
+%     \cs{else:}
+%     ~~\meta{true code}
+%     \cs{fi:}
+%   \end{syntax}
+%   Expands \meta{tokens} until a non-numeric token or a space is found, and
+%   tests whether the resulting \meta{integer} is odd. If so, \meta{true code}
+%   is executed. The \cs{else:} branch is optional.
+%   \begin{texnote}
+%     This is the \TeX{} primitive \cs{ifodd}.
+%   \end{texnote}
+% \end{function}
+%
+% \end{documentation}
+%
+% \begin{implementation}
+%
+% \section{\pkg{l3int} implementation}
+%
+%    \begin{macrocode}
+%<*initex|package>
+%    \end{macrocode}
+%
+% \TestFiles{m3int001,m3int002,m3int03}
+%
+%    \begin{macrocode}
+%<*package>
+\ProvidesExplPackage
+  {\ExplFileName}{\ExplFileDate}{\ExplFileVersion}{\ExplFileDescription}
+\package_check_loaded_expl:
+%</package>
+%    \end{macrocode}
+%    
+% \begin{macro}{\int_to_roman:w}
+% \begin{macro}{\if_int_compare:w}
+%   Done in \pkg{l3basics}.
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\int_value:w}
+% \begin{macro}{\int_eval:w}
+% \begin{macro}{\int_eval_end:}
+% \begin{macro}{\if_num:w}
+% \begin{macro}{\if_int_odd:w}
+% \begin{macro}{\if_case:w}
+%   Here are the remaining primitives for number comparisons and
+%   expressions.
+%    \begin{macrocode}
+\cs_set_eq:NN \int_value:w      \tex_number:D
+\cs_set_eq:NN \int_eval:w       \etex_numexpr:D
+\cs_new_eq:NN \int_eval_end:    \tex_relax:D
+\cs_new_eq:NN \if_num:w         \tex_ifnum:D
+\cs_set_eq:NN \if_int_odd:w     \tex_ifodd:D
+\cs_new_eq:NN \if_case:w        \tex_ifcase:D
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \subsection{Integer expressions}
+%
+% \begin{macro}{\int_eval:n}
+%   Wrapper for \cs{int_eval:w}. Can be used in an integer expression
+%   or directly in the input stream.
+%    \begin{macrocode}
+\cs_new:Npn \int_eval:n #1
+  { \int_value:w \int_eval:w #1 \int_eval_end: }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\int_max:nn}
+% \UnitTested
+% \begin{macro}{\int_min:nn}
+% \UnitTested
+% \begin{macro}{\int_abs:n}
+% \UnitTested
+% Functions for $\min$, $\max$, and absolute value.
+%    \begin{macrocode}
+\cs_new:Npn \int_abs:n #1
+  {
+    \int_value:w
+      \if_int_compare:w \int_eval:w #1 < \c_zero
+        -
+      \fi:
+    \int_eval:w #1 \int_eval_end:
+  }
+\cs_new:Npn \int_max:nn #1#2
+  {
+    \int_value:w \int_eval:w
+      \if_int_compare:w
+        \int_eval:w #1 > \int_eval:w #2 \int_eval_end:
+        #1
+      \else:
+        #2
+      \fi:
+    \int_eval_end:
+  }
+\cs_new:Npn \int_min:nn #1#2
+  {
+    \int_value:w \int_eval:w
+      \if_int_compare:w
+        \int_eval:w #1 < \int_eval:w #2 \int_eval_end:
+        #1
+      \else:
+        #2
+      \fi:
+    \int_eval_end:
+  }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\int_div_truncate:nn}
+% \UnitTested
+% \begin{macro}{\int_div_round:nn}
+% \UnitTested
+% \begin{macro}{\int_mod:nn}
+% \UnitTested
+%   As \cs{int_eval:w} rounds the result of a division we also
+%   provide a version that truncates the result. This version is thanks
+%   to Heiko Oberdiek: getting things right in all cases is not so easy.
+%    \begin{macrocode}
+\cs_new:Npn \int_div_truncate:nn #1#2
+  {
+    \int_value:w \int_eval:w
+      \if_int_compare:w \int_eval:w #1 = \c_zero
+        0
+      \else:
+        ( #1 % )
+        \if_int_compare:w \int_eval:w #1 < \c_zero
+          \if_int_compare:w \int_eval:w #2 < \c_zero
+            - ( #2 + % )
+          \else:
+            + ( #2 - % )
+          \fi:
+        \else:
+          \if_int_compare:w \int_eval:w #2 < \c_zero
+            + ( #2 + % )
+          \else:
+            - ( #2 - % )
+          \fi:
+        \fi: % ( (
+        1 ) / 2 )
+      \fi:
+      / ( #2 )
+    \int_eval_end:
+  }
+%    \end{macrocode}
+%   For the sake of completeness:
+%    \begin{macrocode}
+\cs_new:Npn \int_div_round:nn #1#2 { \int_eval:n { ( #1 ) / ( #2 ) } }
+%    \end{macrocode}
+%   Finally there's the modulus operation.
+%    \begin{macrocode}
+\cs_new:Npn \int_mod:nn #1#2
+  {
+    \int_value:w \int_eval:w
+      #1 - \int_div_truncate:nn {#1} {#2} * ( #2 )
+    \int_eval_end:
+  }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \subsection{Creating and initialising integers}
+%
+% \begin{macro}{\int_new:N, \int_new:c}
+% \UnitTested
+%   Two ways to do this: one for the format and one for the \LaTeXe{}
+%   package.
+%    \begin{macrocode}
+%<*initex>
+\alloc_new:nnnN { int } { 11 } { \c_max_register_int } \tex_countdef:D
+%</initex>
+%<*package>
+\cs_new_protected_nopar:Npn \int_new:N #1
+  {
+    \chk_if_free_cs:N #1
+    \newcount #1
+  }
+%</package>
+\cs_generate_variant:Nn \int_new:N { c }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\int_const:Nn, \int_const:cn}
+% \UnitTested
+%   As stated, most constants can be defined as \cs{tex_chardef:D} or
+%   \cs{tex_mathchardef:D} but that's engine dependent.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \int_const:Nn #1#2
+  {
+    \int_compare:nNnTF {#2} > \c_minus_one
+      {
+        \int_compare:nNnTF {#2} > \c_max_register_int
+          {
+            \int_new:N #1
+            \int_gset:Nn #1 {#2}
+          }
+          {
+            \chk_if_free_cs:N #1
+            \pref_global:D \tex_mathchardef:D #1 =
+              \int_eval:w #2 \int_eval_end:
+          }
+      }
+      {
+        \int_new:N #1
+        \int_gset:Nn #1 {#2}
+      }
+  }
+\cs_generate_variant:Nn \int_const:Nn { c }
+%    \end{macrocode}
+%  \end{macro}
+%
+% \begin{macro}{\int_zero:N, \int_zero:c}
+% \UnitTested
+% \begin{macro}{\int_gzero:N, \int_gzero:c}
+% \UnitTested
+%   Functions that reset an \meta{integer} register to zero.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \int_zero:N  #1 { #1 = \c_zero }
+\cs_new_protected_nopar:Npn \int_gzero:N #1 { \pref_global:D #1 = \c_zero }
+\cs_generate_variant:Nn \int_zero:N  { c }
+\cs_generate_variant:Nn \int_gzero:N { c }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\int_set_eq:NN,\int_set_eq:cN, \int_set_eq:Nc,\int_set_eq:cc}
+% \UnitTested
+% \begin{macro}
+%   {\int_gset_eq:NN,\int_gset_eq:cN, \int_gset_eq:Nc,\int_gset_eq:cc}
+% \UnitTested
+%   Setting equal means using one integer inside the set function of
+%   another.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \int_set_eq:NN #1#2 { #1 = #2 }
+\cs_generate_variant:Nn \int_set_eq:NN {       c }
+\cs_generate_variant:Nn \int_set_eq:NN { Nc , cc }
+\cs_new_protected_nopar:Npn \int_gset_eq:NN #1#2 { \pref_global:D #1 = #2 }
+\cs_generate_variant:Nn \int_gset_eq:NN {       c }
+\cs_generate_variant:Nn \int_gset_eq:NN { Nc , cc }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \subsection{Setting and incrementing integers}
+%
+% \begin{macro}{\int_add:Nn, \int_add:cn}
+% \UnitTested
+% \begin{macro}{\int_gadd:Nn, \int_gadd:cn}
+% \UnitTested
+% \begin{macro}{\int_sub:Nn, \int_sub:cn}
+% \UnitTested
+% \begin{macro}{\int_gsub:Nn, \int_gsub:cn}
+% \UnitTested
+%    Adding and subtracting to and from a counter \ldots
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \int_add:Nn #1#2
+  { \tex_advance:D #1 by \int_eval:w #2 \int_eval_end: }
+\cs_new_nopar:Npn \int_sub:Nn #1#2
+  { \tex_advance:D #1 by - \int_eval:w #2 \int_eval_end: }
+\cs_new_protected_nopar:Npn \int_gadd:Nn
+  { \pref_global:D \int_add:Nn }
+\cs_new_protected_nopar:Npn \int_gsub:Nn
+  { \pref_global:D \int_sub:Nn }
+\cs_generate_variant:Nn \int_add:Nn  { c }
+\cs_generate_variant:Nn \int_gadd:Nn { c }
+\cs_generate_variant:Nn \int_sub:Nn  { c }
+\cs_generate_variant:Nn \int_gsub:Nn { c }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\int_incr:N, \int_incr:c}
+% \UnitTested
+% \begin{macro}{\int_gincr:N, \int_gincr:c}
+% \UnitTested
+% \begin{macro}{\int_decr:N, \int_decr:c}
+% \UnitTested
+% \begin{macro}{\int_gdecr:N, \int_gdecr:c}
+% \UnitTested
+%   Incrementing and decrementing of integer registers is done with
+%   the following functions.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \int_incr:N #1
+  { \tex_advance:D #1 \c_one }
+\cs_new_protected_nopar:Npn \int_decr:N #1
+  { \tex_advance:D #1 \c_minus_one }
+\cs_new_protected_nopar:Npn \int_gincr:N
+  { \pref_global:D \int_incr:N }
+\cs_new_protected_nopar:Npn \int_gdecr:N
+  { \pref_global:D \int_decr:N }
+\cs_generate_variant:Nn \int_incr:N  { c }
+\cs_generate_variant:Nn \int_decr:N  { c }
+\cs_generate_variant:Nn \int_gincr:N { c }
+\cs_generate_variant:Nn \int_gdecr:N { c }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\int_set:Nn, \int_set:cn}
+% \UnitTested
+% \begin{macro}{\int_gset:Nn, \int_gset:cn}
+% \UnitTested
+%   As integers are register-based \TeX{} will issue an error
+%   if they are not defined. Thus there is no need for the checking
+%   code seen with token list variables.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \int_set:Nn #1#2
+  { #1 ~ \int_eval:w #2\int_eval_end: }
+\cs_new_protected_nopar:Npn \int_gset:Nn { \pref_global:D \int_set:Nn }
+\cs_generate_variant:Nn \int_set:Nn  { c }
+\cs_generate_variant:Nn \int_gset:Nn { c }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \subsection{Using integers}
+%
+% \begin{macro}{\int_use:N, \int_use:c}
+% \UnitTested
+%    Here is how counters are accessed:
+%    \begin{macrocode}
+\cs_new_eq:NN \int_use:N \tex_the:D
+\cs_new_nopar:Npn \int_use:c #1 { \int_use:N \cs:w #1 \cs_end: }
+%    \end{macrocode}
+%  \end{macro}
+%
+% \subsection{Integer expression conditionals}
+%
+% \begin{macro}[pTF]{\int_compare:n}
+% \begin{macro}[aux]{\int_compare_aux:nw}
+% \begin{macro}[aux]{\int_compare_aux:Nw}
+% \begin{macro}[aux]{int_compare_=:w}
+% \begin{macro}[aux]{int_compare_==:w}
+% \begin{macro}[aux]{int_compare_!=:w}
+% \begin{macro}[aux]{int_compare_<:w}
+% \begin{macro}[aux]{int_compare_>:w}
+% \begin{macro}[aux]{int_compare_<=:w}
+% \begin{macro}[aux]{int_compare_>=:w}
+%   Comparison tests using a simple syntax where only one set of braces
+%   is required and additional operators such as |!=| and |>=| are
+%   supported. First some notes on the idea behind this. We wish to
+%   support writing code like
+%   \begin{verbatim}
+%     \int_compare_p:n { 5 + \l_tmpa_int != 4 - \l_tmpb_int }
+%   \end{verbatim}
+%   In other words, we want to somehow add the missing \cs{int_eval:w}
+%   where required.  We can start evaluating from the left using
+%   \cs{int_eval:w}, and we know that since the relation symbols |<|, |>|,
+%   |=| and |!| are not allowed in such expressions, they will terminate
+%   the expression. Therefore, we first let \TeX{} evaluate this left
+%   hand side of the (in)equality.
+%    \begin{macrocode}
+\prg_new_conditional:Npnn \int_compare:n #1 { p , T , F , TF }
+  { \exp_after:wN \int_compare_aux:nw \int_value:w \int_eval:w #1 \q_stop }
+%    \end{macrocode}
+%   Then the next step is to figure out which relation we should use, so
+%   we have to somehow get rid of the first evaluation so that we can
+%   see what stopped it. \cs{int_to_roman:w} is handy here since its
+%   expansion given a non-positive number is \meta{null}. We therefore
+%   simply check if the first token of the left hand side evaluation is
+%   a minus. If not, we insert it and issue \cs{int_to_roman:w},
+%   thereby ridding us of the left hand side evaluation. We do however
+%   save it for later.
+%    \begin{macrocode}
+\cs_new:Npn \int_compare_aux:nw #1#2 \q_stop
+  {
+    \exp_after:wN \int_compare_aux:Nw
+      \int_to_roman:w
+        \if:w #1 -
+        \else:
+          -
+        \fi:
+          #1#2 \q_mark #1#2 \q_stop
+  }
+%    \end{macrocode}
+%   This leaves the first relation symbol in front and assuming the
+%   right hand side has been input, at least one other token as well. We
+%   support the following forms: |=|, |<|, |>| and the extended |!=|,
+%   |==|, |<=| and |>=|. All the extended forms have an extra |=| so we
+%   check if that is present as well. Then use specific function to
+%   perform the test.
+%    \begin{macrocode}
+\cs_new:Npn \int_compare_aux:Nw #1#2#3 \q_mark
+  { \use:c { int_compare_  #1  \if_meaning:w = #2 =  \fi: :w } }
+%    \end{macrocode}
+%   The actual comparisons are then simple function calls, using the
+%   relation as delimiter for a delimited argument.
+%   Equality is easy:
+%    \begin{macrocode}
+\cs_new:cpn { int_compare_=:w } #1 = #2 \q_stop
+  {
+    \if_int_compare:w #1 = \int_eval:w #2 \int_eval_end:
+      \prg_return_true:
+    \else:
+      \prg_return_false:
+    \fi:
+  }
+%    \end{macrocode}
+%   So is the one using |==| we just have to use |==| in the
+%   parameter text.
+%    \begin{macrocode}
+\cs_new:cpn { int_compare_==:w } #1 == #2 \q_stop
+  {
+    \if_int_compare:w #1 = \int_eval:w #2 \int_eval_end:
+      \prg_return_true:
+    \else:
+      \prg_return_false:
+    \fi:
+  }
+%    \end{macrocode}
+%   Not equal is just about reversing the truth value.
+%    \begin{macrocode}
+\cs_new:cpn { int_compare_!=:w } #1 != #2 \q_stop
+  {
+    \if_int_compare:w #1 = \int_eval:w #2 \int_eval_end:
+      \prg_return_false:
+    \else:
+      \prg_return_true:
+    \fi:
+  }
+%    \end{macrocode}
+%   Less than and greater than are also straight forward.
+%    \begin{macrocode}
+\cs_new:cpn { int_compare_<:w } #1 < #2 \q_stop
+  {
+    \if_int_compare:w #1 < \int_eval:w #2 \int_eval_end:
+      \prg_return_true:
+    \else:
+      \prg_return_false:
+    \fi:
+  }
+\cs_new:cpn { int_compare_>:w } #1 > #2 \q_stop
+  {
+    \if_int_compare:w #1 > \int_eval:w #2 \int_eval_end:
+      \prg_return_true:
+    \else:
+      \prg_return_false:
+    \fi:
+  }
+%    \end{macrocode}
+%   The less than or equal operation is just the opposite of the greater
+%   than operation. \emph{Vice versa} for less than or equal.
+%    \begin{macrocode}
+\cs_new:cpn { int_compare_<=:w } #1 <= #2 \q_stop
+  {
+    \if_int_compare:w #1 > \int_eval:w #2 \int_eval_end:
+      \prg_return_false:
+    \else:
+      \prg_return_true:
+    \fi:
+  }
+\cs_new:cpn { int_compare_>=:w } #1 >= #2 \q_stop
+  {
+    \if_int_compare:w #1 < \int_eval:w #2 \int_eval_end:
+      \prg_return_false:
+    \else:
+      \prg_return_true:
+    \fi:
+  }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}[pTF]{\int_compare:nNn}
+% \UnitTested
+% More efficient but less natural in typing.
+%    \begin{macrocode}
+\prg_new_conditional:Npnn \int_compare:nNn #1#2#3 { p , T , F , TF}
+  {
+    \if_int_compare:w \int_eval:w #1 #2 \int_eval:w #3 \int_eval_end:
+      \prg_return_true:
+    \else:
+      \prg_return_false:
+    \fi:
+  }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}[pTF]{\int_if_odd:n}
+% \UnitTested
+% \begin{macro}[pTF]{\int_if_even:n}
+% \UnitTested
+%   A predicate function.
+%    \begin{macrocode}
+\prg_new_conditional:Npnn \int_if_odd:n #1 { p , T , F , TF}
+  {
+    \if_int_odd:w \int_eval:w #1 \int_eval_end:
+      \prg_return_true:
+    \else:
+      \prg_return_false:
+    \fi:
+  }
+\prg_new_conditional:Npnn \int_if_even:n #1 { p , T , F , TF}
+  {
+    \if_int_odd:w \int_eval:w #1 \int_eval_end:
+      \prg_return_false:
+    \else:
+      \prg_return_true:
+    \fi:
+  }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \subsection{Integer expression loops}
+%
+% \begin{macro}{\int_while_do:nn}
+% \UnitTested
+% \TestMissing{Boundary cases}
+% \begin{macro}{\int_until_do:nn}
+% \UnitTested
+% \TestMissing{Boundary cases}
+% \begin{macro}{\int_do_while:nn}
+% \UnitTested
+% \TestMissing{Boundary cases}
+% \begin{macro}{\int_do_until:nn}
+% \UnitTested
+% \TestMissing{Boundary cases}
+%   These are quite easy given the above functions. The |while| versions
+%   test first and then execute the body. The |do_while| does it the
+%   other way round.
+%    \begin{macrocode}
+\cs_new:Npn \int_while_do:nn #1#2
+  {
+    \int_compare:nT {#1}
+      {
+        #2
+        \int_while_do:nn {#1} {#2}
+      }
+  }
+\cs_new:Npn \int_until_do:nn #1#2
+  {
+    \int_compare:nF {#1}
+      {
+        #2
+        \int_until_do:nn {#1} {#2}
+      }
+  }
+\cs_new:Npn \int_do_while:nn #1#2
+  {
+    #2
+    \int_compare:nT {#1}
+      { \int_do_while:nNnn {#1} {#2} }
+  }
+\cs_new:Npn \int_do_until:nn #1#2
+  {
+    #2
+    \int_compare:nF {#1}
+      { \int_do_until:nn {#1} {#2} }
+  }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\int_while_do:nNnn}
+% \begin{macro}{\int_until_do:nNnn}
+% \begin{macro}{\int_do_while:nNnn}
+% \begin{macro}{\int_do_until:nNnn}
+%    As above but not using the more natural syntax.
+%    \begin{macrocode}
+\cs_new:Npn \int_while_do:nNnn #1#2#3#4
+  {
+    \int_compare:nNnT {#1} #2 {#3}
+      {
+        #4
+        \int_while_do:nNnn {#1} #2 {#3} {#4}
+      }
+  }
+\cs_new:Npn \int_until_do:nNnn #1#2#3#4
+  {
+  \int_compare:nNnF {#1} #2 {#3}
+    {
+      #4
+      \int_until_do:nNnn {#1} #2 {#3} {#4}
+    }
+  }
+\cs_new:Npn \int_do_while:nNnn #1#2#3#4
+  {
+    #4
+    \int_compare:nNnT {#1} #2 {#3}
+      { \int_do_while:nNnn {#1} #2 {#3} {#4} }
+  }
+\cs_new:Npn \int_do_until:nNnn #1#2#3#4
+  {
+    #4
+    \int_compare:nNnF {#1} #2 {#3}
+      { \int_do_until:nNnn {#1} #2 {#3} {#4} }
+  }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \subsection{Formatting integers}
+%
+% \begin{macro}{\int_to_arabic:n}
+% \UnitTested
+%   Nothing exciting here.
+%    \begin{macrocode}
+\cs_new_nopar:Npn \int_to_arabic:n #1 { \int_eval:n {#1} }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\int_to_symbols:nnn}
+% \UnitTested
+%   For conversion of integers to arbitrary symbols the method is in
+%   general as follows. The input number (|#1|) is compared to the total
+%   number of symbols available at each place (|#2|). If the input is larger
+%   than the total number of symbols available then the modulus is needed,
+%   with one added so that the positions don't have to number from
+%   zero. Using an \texttt{f}-type expansion, this is done so that the system
+%   is recursive. The actual conversion function therefore gets a `nice'
+%   number at each stage. Of course, if the initial input was small enough
+%   then there is no problem and everything is easy. This is more or less the
+%   same as \cs{int_convert_number_with_rule:nnN} but \enquote{pre-packaged}.
+%    \begin{macrocode}
+\cs_new_nopar:Npn \int_to_symbols:nnn #1#2#3
+  {
+    \int_compare:nNnTF {#1} > {#2}
+      {
+        \exp_args:Nf \int_to_symbols:nnn
+          { \int_div_truncate:nn { #1 - 1 } {#2} } {#2} {#3}
+        \exp_args:Nf \prg_case_int:nnn
+          { \int_eval:n { 1 + \int_mod:nn { #1 - 1 } {#2} } }
+          {#3} { }
+      }
+      { \exp_args:Nf \prg_case_int:nnn { \int_eval:n {#1} } {#3} { } }
+  }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\int_to_alph:n,\int_to_Alph:n}
+% \UnitTested
+%   These both use the above function with input functions that make sense
+%   for the alphabet in English.
+%    \begin{macrocode}
+\cs_new:Npn \int_to_alph:n #1
+  {
+    \int_to_symbols:nnn {#1} { 26 }
+      {
+        {  1 } { a }
+        {  2 } { b }
+        {  3 } { c }
+        {  4 } { d }
+        {  5 } { e }
+        {  6 } { f }
+        {  7 } { g }
+        {  8 } { h }
+        {  9 } { i }
+        { 10 } { j }
+        { 11 } { k }
+        { 12 } { l }
+        { 13 } { m }
+        { 14 } { n }
+        { 15 } { o }
+        { 16 } { p }
+        { 17 } { q }
+        { 18 } { r }
+        { 19 } { s }
+        { 20 } { t }
+        { 21 } { u }
+        { 22 } { v }
+        { 23 } { w }
+        { 24 } { x }
+        { 25 } { y }
+        { 26 } { z }
+      }
+  }
+\cs_new:Npn \int_to_Alph:n #1
+  {
+    \int_to_symbols:nnn {#1} { 26 }
+      {
+        {  1 } { A }
+        {  2 } { B }
+        {  3 } { C }
+        {  4 } { D }
+        {  5 } { E }
+        {  6 } { F }
+        {  7 } { G }
+        {  8 } { H }
+        {  9 } { I }
+        { 10 } { J }
+        { 11 } { K }
+        { 12 } { L }
+        { 13 } { M }
+        { 14 } { N }
+        { 15 } { O }
+        { 16 } { P }
+        { 17 } { Q }
+        { 18 } { R }
+        { 19 } { S }
+        { 20 } { T }
+        { 21 } { U }
+        { 22 } { V }
+        { 23 } { W }
+        { 24 } { X }
+        { 25 } { Y }
+        { 26 } { Z }
+      }
+  }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\int_to_base:nn}
+% \UnitTested
+% \begin{macro}[aux]{\int_to_base_aux:nnn}
+% \begin{macro}[int]{\int_to_letter:n}
+% \UnitTested
+%   Converting from base ten (|#1|) to a second base (|#2|) starts with
+%   a simple sign check. As the input is base $10$ \TeX{} can then
+%   do the actual work with the sign itself.
+%    \begin{macrocode}
+\cs_new:Npn \int_to_base:nn #1#2
+  {
+    \int_compare:nNnTF {#1} < \c_zero
+      {
+        -
+        \exp_args:Nnf \int_to_base_aux:nnn
+          { } { \int_eval:n { 0 - ( #1 ) } } {#2}
+      }
+      {
+      \exp_args:Nnf \int_to_base_aux:nnn
+        { } { \int_eval:n {#1} } {#2}
+    }
+}
+%    \end{macrocode}
+%   Here, the idea is to provide a recursive system to deal with the
+%   input. The output is build up as argument |#1|, which is why it
+%   starts off empty in the above. At each pass, the value in |#2| is
+%   checked to see if it is less than the new base (|#3|). If it is
+%   the it is converted directly and the rest of the output is added in.
+%   On the other hand, if the value to convert is greater than or equal
+%   to the new base then the modulus and remainder values are found. The
+%   modulus is converted to a symbol and the remainder is carried forward
+%   to the next round.
+%    \begin{macrocode}
+\cs_new:Npn \int_to_base_aux:nnn #1#2#3 {
+  \int_compare:nNnTF {#2} < {#3}
+    {
+      \int_to_letter:n {#2}
+      #1
+    }
+    {
+      \exp_args:Nff \int_to_base_aux:nnn
+        {
+          \int_to_letter:n { \int_mod:nn {#2} {#3} }
+          #1
+        }
+        { \int_div_truncate:nn {#2} {#3} }
+        {#3}
+    }
+  }
+%    \end{macrocode}
+%   Convert to a letter only if necessary, otherwise simply return the
+%   value unchanged.
+%    \begin{macrocode}
+\cs_new:Npn \int_to_letter:n #1
+  {
+    \prg_case_int:nnn { #1 - 9 }
+      {
+        {  1 } { A }
+        {  2 } { B }
+        {  3 } { C }
+        {  4 } { D }
+        {  5 } { E }
+        {  6 } { F }
+        {  7 } { G }
+        {  8 } { H }
+        {  9 } { I }
+        { 10 } { J }
+        { 11 } { K }
+        { 12 } { L }
+        { 13 } { M }
+        { 14 } { N }
+        { 15 } { O }
+        { 16 } { P }
+        { 17 } { Q }
+        { 18 } { R }
+        { 19 } { S }
+        { 20 } { T }
+        { 21 } { U }
+        { 22 } { V }
+        { 23 } { W }
+        { 24 } { X }
+        { 25 } { Y }
+        { 26 } { Z }
+      }
+      {#1}
+  }
+%    \end{macrocode}
+%\end{macro}
+%\end{macro}
+%\end{macro}
+%
+% \begin{macro}{\int_to_binary:n, \int_to_hexadecimal:n, \int_to_octal:n}
+% \UnitTested
+%   Wrappers around the generic function.
+%    \begin{macrocode}
+\cs_new:Npn \int_to_binary:n #1
+  { \int_to_base:nn {#1} { 2 } }
+\cs_new:Npn \int_to_hexadecimal:n #1
+  { \int_to_base:nn {#1} { 16 } }
+\cs_new:Npn \int_to_octal:n #1
+  { \int_to_base:nn {#1} { 8 } }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\int_to_roman:n, \int_to_Roman:n}
+% \UnitTested
+% \begin{macro}[aux]{\int_to_roman_aux:N, \int_to_roman_aux:N}
+% \begin{macro}[aux]
+%   {
+%     \int_to_roman_i:w, \int_to_roman_v:w, \int_to_roman_x:w,
+%     \int_to_roman_l:w, \int_to_roman_c:w, \int_to_roman_d:w,
+%     \int_to_roman_m:w, \int_to_roman_Q:w,
+%     \int_to_Roman_i:w, \int_to_Roman_v:w, \int_to_Roman_x:w,
+%     \int_to_Roman_l:w, \int_to_Roman_c:w, \int_to_Roman_d:w,
+%     \int_to_Roman_m:w, \int_to_Roman_Q:w,
+%   }
+% The \cs{int_to_roman:w} primitive creates tokens of category
+% code $12$ (other). Usually, what is actually wanted is letters.
+% The approach here is to convert the output of the primitive into
+% letters using appropriate control sequence names. That keeps
+% everything expandable. The loop will be terminated by the conversion
+% of the |Q|.
+%    \begin{macrocode}
+\cs_new_nopar:Npn \int_to_roman:n #1
+  {
+    \exp_after:wN \int_to_roman_aux:N
+      \int_to_roman:w \int_eval:n {#1} Q
+  }
+\cs_new_nopar:Npn \int_to_roman_aux:N #1
+  {
+    \use:c { int_to_roman_ #1 :w }
+    \int_to_roman_aux:N
+  }
+\cs_new_nopar:Npn \int_to_Roman:n #1
+  {
+    \exp_after:wN \int_to_Roman_aux:N
+      \int_to_roman:w \int_eval:n {#1} Q
+  }
+\cs_new_nopar:Npn \int_to_Roman_aux:N #1
+  {
+    \use:c { int_to_Roman_ #1 :w }
+    \int_to_Roman_aux:N
+  }
+\cs_new_nopar:Npn \int_to_roman_i:w { i }
+\cs_new_nopar:Npn \int_to_roman_v:w { v }
+\cs_new_nopar:Npn \int_to_roman_x:w { x }
+\cs_new_nopar:Npn \int_to_roman_l:w { l }
+\cs_new_nopar:Npn \int_to_roman_c:w { c }
+\cs_new_nopar:Npn \int_to_roman_d:w { d }
+\cs_new_nopar:Npn \int_to_roman_m:w { m }
+\cs_new_nopar:Npn \int_to_roman_Q:w #1 { }
+\cs_new_nopar:Npn \int_to_Roman_i:w { I }
+\cs_new_nopar:Npn \int_to_Roman_v:w { V }
+\cs_new_nopar:Npn \int_to_Roman_x:w { X }
+\cs_new_nopar:Npn \int_to_Roman_l:w { L }
+\cs_new_nopar:Npn \int_to_Roman_c:w { C }
+\cs_new_nopar:Npn \int_to_Roman_d:w { D }
+\cs_new_nopar:Npn \int_to_Roman_m:w { M }
+\cs_new_nopar:Npn \int_to_Roman_Q:w #1 { }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \subsection{Converting from other formats to integers}
+%
+% \begin{macro}[int]{\int_get_sign:n}
+% \begin{macro}[int]{\int_get_digits:n}
+% \begin{macro}[aux]
+%   {\int_get_sign_and_digits_aux:nNNN, \int_get_sign_and_digits_aux:oNNN}
+%   Finding a number and its sign requires dealing with an arbitrary
+%   list of |+| and |-| symbols. This is done by working through token
+%   by token until there is something else at the start of the input.
+%   The sign of the input is tracked by the first Boolean used by the
+%   auxiliary function.
+%    \begin{macrocode}
+\cs_new:Npn \int_get_sign:n #1
+  {
+    \int_get_sign_and_digits_aux:nNNN {#1}
+      \c_true_bool \c_true_bool \c_false_bool
+  }
+\cs_new:Npn \int_get_digits:n #1
+  {
+    \int_get_sign_and_digits_aux:nNNN {#1}
+      \c_true_bool \c_false_bool \c_true_bool
+  }
+%    \end{macrocode}
+%   The auxiliary loops through, finding sign tokens and removing them.
+%   The sign itself is carried through as a flag.
+%    \begin{macrocode}
+\cs_new:Npn \int_get_sign_and_digits_aux:nNNN #1#2#3#4
+  {
+    \tl_if_head_eq_charcode:fNTF {#1} -
+      {
+        \bool_if:NTF #2
+          {
+            \int_get_sign_and_digits_aux:oNNN
+              { \use_none:n #1 } \c_false_bool #3#4
+          }
+          {
+            \int_get_sign_and_digits_aux:oNNN
+              { \use_none:n #1 } \c_true_bool #3#4
+          }
+      }
+      {
+        \tl_if_head_eq_charcode:fNTF {#1} +
+          { \int_get_sign_and_digits_aux:oNNN { \use_none:n #1 } #2#3#4 }
+          {
+            \bool_if:NT #3 { \bool_if:NF #2 - }
+            \bool_if:NT #4 {#1}
+          }
+      }
+  }
+\cs_generate_variant:Nn \int_get_sign_and_digits_aux:nNNN { o }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\int_from_alph:n}
+% \UnitTested
+% \begin{macro}[aux]{\int_from_alph_aux:n}
+% \begin{macro}[aux]{\int_from_alph_aux:nN}
+% \begin{macro}[aux]{\int_from_alph_aux:N}
+%   The aim here is to iterate through the input, converting one letter at
+%   a time to a number. The same approach is also used for base
+%   conversion, but this needs a different final auxiliary.
+%    \begin{macrocode}
+\cs_new:Npn \int_from_alph:n #1
+  {
+    \int_eval:n
+      {
+        \int_get_sign:n {#1}
+        \exp_args:Nf \int_from_alph_aux:n { \int_get_digits:n {#1} }
+      }
+  }
+\cs_new:Npn \int_from_alph_aux:n #1
+  { \int_from_alph_aux:nN { 0 } #1 \q_nil }
+\cs_new:Npn \int_from_alph_aux:nN #1#2
+  {
+    \quark_if_nil:NTF #2
+      {#1}
+      {
+        \exp_args:Nf \int_from_alph_aux:nN
+          { \int_eval:n { #1 * 26 + \int_from_alph_aux:N #2 } }
+      }
+  }
+\cs_new:Npn \int_from_alph_aux:N #1
+  { \int_eval:n { `#1 - \int_compare:nNnTF { `#1 } < { 91 } { 64 } { 96 } } }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\int_from_base:nn}
+% \UnitTested
+% \begin{macro}[aux]{\int_from_base_aux:nn}
+% \begin{macro}[aux]{\int_from_base_aux:nnN}
+% \begin{macro}[aux]{\int_from_base_aux:N}
+%   Conversion to base ten means stripping off the sign then iterating
+%   through the input one token at a time. The total number is then added
+%   up as the code loops.
+%    \begin{macrocode}
+\cs_new:Npn \int_from_base:nn #1#2
+  {
+    \int_eval:n
+      {
+        \int_get_sign:n {#1}
+        \exp_args:Nf \int_from_base_aux:nn
+          { \int_get_digits:n {#1} } {#2}
+      }
+  }
+\cs_new:Npn \int_from_base_aux:nn #1#2
+  { \int_from_base_aux:nnN { 0 } { #2 } #1 \q_nil }
+\cs_new:Npn \int_from_base_aux:nnN #1#2#3
+  {
+    \quark_if_nil:NTF #3
+      {#1}
+      {
+        \exp_args:Nf \int_from_base_aux:nnN
+          { \int_eval:n { #1 * #2 + \int_from_base_aux:N #3 } }
+          {#2}
+      }
+  }
+%    \end{macrocode}
+%   The conversion here will take lower or upper case letters and turn
+%   them into the appropriate number, hence the two-part nature of the
+%   function.
+%    \begin{macrocode}
+\cs_new:Npn \int_from_base_aux:N #1
+  {
+    \int_compare:nNnTF { `#1 } < { 58 }
+      {#1}
+      {
+        \int_eval:n
+          { `#1 - \int_compare:nNnTF { `#1 } < { 91 } { 55 } { 87 } }
+      }
+  }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\int_from_binary:n, \int_from_hexadecimal:n, \int_from_octal:n}
+% \UnitTested
+%   Wrappers around the generic function.
+%    \begin{macrocode}
+\cs_new:Npn \int_from_binary:n #1
+  { \int_from_base:nn {#1} \c_two }
+\cs_new:Npn \int_from_hexadecimal:n #1
+  { \int_from_base:nn {#1} \c_sixteen }
+\cs_new:Npn \int_from_octal:n #1
+  { \int_from_base:nn {#1} \c_eight }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{variable}[aux]
+%   {
+%     \c_int_from_roman_i_int, \c_int_from_roman_v_int,
+%     \c_int_from_roman_x_int, \c_int_from_roman_l_int,
+%     \c_int_from_roman_c_int, \c_int_from_roman_d_int,
+%     \c_int_from_roman_m_int,
+%     \c_int_from_roman_I_int, \c_int_from_roman_V_int,
+%     \c_int_from_roman_X_int, \c_int_from_roman_L_int,
+%     \c_int_from_roman_C_int, \c_int_from_roman_D_int,
+%     \c_int_from_roman_M_int
+%  }
+% Constants used to convert from Roman numerals to integers.
+%    \begin{macrocode}
+\int_const:cn { c_int_from_roman_i_int } { 1 }
+\int_const:cn { c_int_from_roman_v_int } { 5 }
+\int_const:cn { c_int_from_roman_x_int } { 10 }
+\int_const:cn { c_int_from_roman_l_int } { 50 }
+\int_const:cn { c_int_from_roman_c_int } { 100 }
+\int_const:cn { c_int_from_roman_d_int } { 500 }
+\int_const:cn { c_int_from_roman_m_int } { 1000 }
+\int_const:cn { c_int_from_roman_I_int } { 1 }
+\int_const:cn { c_int_from_roman_V_int } { 5 }
+\int_const:cn { c_int_from_roman_X_int } { 10 }
+\int_const:cn { c_int_from_roman_L_int } { 50 }
+\int_const:cn { c_int_from_roman_C_int } { 100 }
+\int_const:cn { c_int_from_roman_D_int } { 500 }
+\int_const:cn { c_int_from_roman_M_int } { 1000 }
+%    \end{macrocode}
+% \end{variable}
+%
+% \begin{macro}{\int_from_roman:n}
+% \UnitTested
+% \TestMissing{boundary cases / wrong input?}
+% \begin{macro}[aux]{\int_from_roman_aux:NN}
+% \begin{macro}[aux]{\int_from_roman_end:w}
+% \begin{macro}[aux]{\int_from_roman_clean_up:w}
+%   The method here is to iterate through the input, finding the
+%   appropriate value for each letter and building up a sum. This is
+%   then evaluated by \TeX{}.
+%    \begin{macrocode}
+\cs_new_nopar:Npn \int_from_roman:n #1
+  {
+    \tl_if_blank:nF {#1}
+      {
+        \exp_after:wN \int_from_roman_end:w
+          \int_value:w \int_eval:w
+            \int_from_roman_aux:NN #1 Q \q_stop
+      }
+  }
+\cs_new_nopar:Npn \int_from_roman_aux:NN #1#2
+  {
+    \str_if_eq:nnTF {#1} { Q }
+      {#1#2}
+      {
+        \str_if_eq:nnTF {#2} { Q }
+          {
+            \cs_if_exist:cF { c_int_from_roman_ #1 _int }
+              { \int_from_roman_clean_up:w }
+            +
+            \use:c { c_int_from_roman_ #1 _int }
+            #2
+          }
+          {
+            \cs_if_exist:cF { c_int_from_roman_ #1 _int }
+              { \int_from_roman_clean_up:w }
+            \cs_if_exist:cF { c_int_from_roman_ #2 _int }
+              { \int_from_roman_clean_up:w }
+            \int_compare:nNnTF
+              { \use:c { c_int_from_roman_ #1 _int } }
+              <
+              { \use:c { c_int_from_roman_ #2 _int  } }
+              {
+                + \use:c { c_int_from_roman_ #2 _int }
+                - \use:c { c_int_from_roman_ #1 _int }
+                \int_from_roman_aux:NN
+              }
+              {
+                + \use:c { c_int_from_roman_ #1 _int }
+                \int_from_roman_aux:NN #2
+              }
+          }
+      }
+  }
+\cs_new_nopar:Npn \int_from_roman_end:w #1 Q #2 \q_stop
+  { \tl_if_empty:nTF {#2} {#1} {#2} }
+\cs_new_nopar:Npn \int_from_roman_clean_up:w #1 Q { + 0 Q -1 }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \subsection{Viewing integer}
+%
+% \begin{macro}{\int_show:N,\int_show:c}
+% \UnitTested
+%    \begin{macrocode}
+\cs_new_eq:NN \int_show:N \kernel_register_show:N
+\cs_new_eq:NN \int_show:c \kernel_register_show:c
+%    \end{macrocode}
+% \end{macro}
+%
+% \subsection{Constant integers}
+%
+% \begin{variable}{\c_minus_one}
+%   This is needed early, and so is in \pkg{l3basics}
+% \end{variable}
+%
+% \begin{variable}{\c_zero}
+%   Again, one in \pkg{l3basics} for obvious reasons.
+% \end{variable}
+%
+% \begin{variable}{\c_six, \c_seven}
+% \begin{variable}{\c_twelve}
+% \begin{variable}{\c_sixteen}
+%   Once again, in \pkg{l3basics}.
+% \end{variable}
+% \end{variable}
+% \end{variable}
+%
+% \begin{variable}
+%   {
+%     \c_one, \c_two, \c_three, \c_four, \c_five, \c_eight, \c_nine, \c_ten,
+%     \c_eleven, \c_thirteen, \c_fourteen, \c_fifteen,
+%   }
+% Low-number values not previously defined.
+%    \begin{macrocode}
+\int_const:Nn \c_one      {  1 }
+\int_const:Nn \c_two      {  2 }
+\int_const:Nn \c_three    {  3 }
+\int_const:Nn \c_four     {  4 }
+\int_const:Nn \c_five     {  5 }
+\int_const:Nn \c_eight    {  8 }
+\int_const:Nn \c_nine     {  9 }
+\int_const:Nn \c_ten      { 10 }
+\int_const:Nn \c_eleven   { 11 }
+\int_const:Nn \c_thirteen { 13 }
+\int_const:Nn \c_fourteen { 14 }
+\int_const:Nn \c_fifteen  { 15 }
+%    \end{macrocode}
+% \end{variable}
+%
+% \begin{variable}{\c_thirty_two}
+%   One middling value.
+%    \begin{macrocode}
+\int_const:Nn \c_thirty_two { 32 }
+%    \end{macrocode}
+% \end{variable}
+%
+% \begin{variable}{\c_two_hundred_fifty_five, \c_two_hundred_fifty_six}
+%   Two classic mid-range integer constants.
+%    \begin{macrocode}
+\int_const:Nn \c_two_hundred_fifty_five { 255 }
+\int_const:Nn \c_two_hundred_fifty_six  { 256 }
+%    \end{macrocode}
+%\end{variable}
+%
+% \begin{variable}
+%   {\c_one_hundred, \c_one_thousand, \c_ten_thousand}
+%   Simple runs of powers of ten.
+%    \begin{macrocode}
+\int_const:Nn \c_one_hundred  {   100 }
+\int_const:Nn \c_one_thousand {  1000 }
+\int_const:Nn \c_ten_thousand { 10000 }
+%    \end{macrocode}
+% \end{variable}
+%
+% \begin{variable}{\c_max_int}
+%   The largest number allowed is $2^{31}-1$
+%    \begin{macrocode}
+\int_const:Nn \c_max_int { 2 147 483 647 }
+%    \end{macrocode}
+% \end{variable}
+%
+% \subsection{Scratch integers}
+%
+% \begin{variable}{\l_tmpa_int, \l_tmpb_int, \l_tmpc_int}
+% \begin{variable}{\g_tmpa_int, \g_tmpb_int}
+%    We provide four local and two global scratch counters, maybe we
+%    need more or less.
+%    \begin{macrocode}
+\int_new:N \l_tmpa_int
+\int_new:N \l_tmpb_int
+\int_new:N \l_tmpc_int
+\int_new:N \g_tmpa_int
+\int_new:N \g_tmpb_int
+%    \end{macrocode}
+% \end{variable}
+% \end{variable}
+%
+% \subsection{Registers for earlier modules}
+%
+% Needed from other modules:
+%    \begin{macrocode}
+\int_new:N \g_seq_nesting_depth_int
+\int_new:N \g_tl_inline_level_int
+%    \end{macrocode}
+%
+% \subsection{Deprecated functions}
+%
+% Deprecated on 2011-05-27, for removal by 2011-08-31.
+%
+% \begin{macro}{\int_convert_from_base_ten:nn}
+% \begin{macro}{\int_convert_to_symbols:nnn}
+% \begin{macro}{\int_convert_to_base_ten:nn}
+% Some simple renames.
+%    \begin{macrocode}
+\cs_new_eq:NN \int_convert_from_base_ten:nn \int_to_base:nn
+\cs_new_eq:NN \int_convert_to_symbols:nnn   \int_to_symbols:nnn
+\cs_new_eq:NN \int_convert_to_base_ten:nn   \int_from_base:nn
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\int_to_symbol:n}
+% \begin{macro}[aux]{\int_to_symbol_math:n}
+% \begin{macro}[aux]{\int_to_symbol_text:n}
+% \UnitTested
+%   This is rather too tied to \LaTeXe{}.
+%    \begin{macrocode}
+\cs_new_nopar:Npn \int_to_symbol:n
+  {
+    \mode_if_math:TF
+      { \int_to_symbol_math:n }
+      { \int_to_symbol_text:n }
+  }
+\cs_new:Npn \int_to_symbol_math:n #1
+  {
+    \int_to_symbols:nnn {#1} { 9 }
+      {
+        { 1 } {                 * }
+        { 2 } {           \dagger }
+        { 3 } {          \ddagger }
+        { 4 } {      \mathsection }
+        { 5 } {    \mathparagraph }
+        { 6 } {                \| }
+        { 7 } {                ** }
+        { 8 } {   \dagger \dagger }
+        { 9 } { \ddagger \ddagger }
+      }
+  }
+\cs_new:Npn \int_to_symbol_text:n #1
+  {
+    \int_to_symbols:nnn {#1} { 9 }
+      {
+        { 1 } {                       \textasteriskcentered }
+        { 2 } {                                 \textdagger }
+        { 3 } {                              \textdaggerdbl }
+        { 4 } {                                \textsection }
+        { 5 } {                              \textparagraph }
+        { 6 } {                                 \textbardbl }
+        { 7 } { \textasteriskcentered \textasteriskcentered }
+        { 8 } {                     \textdagger \textdagger }
+        { 9 } {               \textdaggerdbl \textdaggerdbl }
+      }
+  }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+%    \begin{macrocode}
+%</initex|package>
+%    \end{macrocode}
+%
+% \end{implementation}
+%
+% \PrintIndex
diff --git a/Master/texmf-dist/source/latex/l3kernel/l3io.dtx b/Master/texmf-dist/source/latex/l3kernel/l3io.dtx
new file mode 100644
index 00000000000..bf71c3df878
--- /dev/null
+++ b/Master/texmf-dist/source/latex/l3kernel/l3io.dtx
@@ -0,0 +1,1185 @@
+% \iffalse meta-comment
+%
+%% File: l3io.dtx Copyright (C) 1990-2011 The LaTeX3 Project
+%%
+%% It may be distributed and/or modified under the conditions of the
+%% LaTeX Project Public License (LPPL), either version 1.3c of this
+%% license or (at your option) any later version.  The latest version
+%% of this license is in the file
+%%
+%%    http://www.latex-project.org/lppl.txt
+%%
+%% This file is part of the "expl3 bundle" (The Work in LPPL)
+%% and all files in that bundle must be distributed together.
+%%
+%% The released version of this bundle is available from CTAN.
+%%
+%% -----------------------------------------------------------------------
+%%
+%% The development version of the bundle can be found at
+%%
+%%    http://www.latex-project.org/svnroot/experimental/trunk/
+%%
+%% for those people who are interested.
+%%
+%%%%%%%%%%%
+%% NOTE: %%
+%%%%%%%%%%%
+%%
+%%   Snapshots taken from the repository represent work in progress and may
+%%   not work or may contain conflicting material!  We therefore ask
+%%   people _not_ to put them into distributions, archives, etc. without
+%%   prior consultation with the LaTeX3 Project.
+%%
+%% -----------------------------------------------------------------------
+%
+%<*driver|package>
+\RequirePackage{l3names}
+\GetIdInfo$Id: l3io.dtx 2478 2011-06-19 21:34:23Z joseph $
+  {L3 Experimental input-output operations}
+%</driver|package>
+%<*driver>
+\documentclass[full]{l3doc}
+\begin{document}
+  \DocInput{\jobname.dtx}
+\end{document}
+%</driver>
+% \fi
+%
+% \title{^^A
+%   The \pkg{l3io} package\\ Input--output operations^^A
+%   \thanks{This file describes v\ExplFileVersion,
+%      last revised \ExplFileDate.}^^A
+% }
+%
+% \author{^^A
+%  The \LaTeX3 Project\thanks
+%    {^^A
+%      E-mail:
+%        \href{mailto:latex-team@latex-project.org}
+%          {latex-team@latex-project.org}^^A
+%    }^^A
+% }
+%
+% \date{Released \ExplFileDate}
+%
+% \maketitle
+%
+% \begin{documentation}
+%
+% Reading and writing from file streams is handled in \LaTeX3 using
+% functions with prefixes \cs{iow_\ldots} (file reading) and
+% \cs{ior_\ldots} (file writing). Many of the basic functions are
+% very similar, with reading and writing using the same syntax and
+% function concepts. As a result, the reading and writing functions
+% are documented together where this makes sense.
+%
+% As \TeX{} is limited to $16$ input streams and $16$ output
+% streams, direct use of the streams by the programmer is not
+% supported in \LaTeX3. Instead, an internal pool of streams is
+% maintained, and these are allocated and deallocated as needed
+% by other modules. As a result, the programmer should close streams
+% when they are no longer needed, to release them for other processes.
+%
+% Reading from or writing to a file requires a \meta{stream} to
+% be used. This is a csname which refers to the file being processed,
+% and is independent of the name of the file (except of course that
+% the file name is needed when the file is opened).
+%
+% \section{Opening and closing streams}
+%
+% \begin{function}{\ior_open:Nn, \ior_open:cn}
+%   \begin{syntax}
+%     \cs{ior_open:Nn} \meta{stream} \Arg{file name}
+%   \end{syntax}
+%   Opens \meta{file name} for reading using \meta{stream} as the
+%   control sequence for file access. If the \meta{stream} was already
+%   open it is closed before the new operation begins. The
+%   \meta{stream} is available for access immediately and will remain
+%   allocated to \meta{file name} until a \cs{ior_close:N} instruction
+%   is given or the file ends.
+% \end{function}
+%
+% \begin{function}{\iow_open:Nn, \iow_open:cn}
+%   \begin{syntax}
+%     \cs{iow_open:Nn} \meta{stream} \Arg{file name}
+%   \end{syntax}
+%   Opens \meta{file name} for writing using \meta{stream} as the
+%   control sequence for file access. If the \meta{stream} was already
+%   open it is closed before the new operation begins. The
+%   \meta{stream} is available for access immediately and will remain
+%   allocated to \meta{file name} until a \cs{iow_close:N} instruction
+%   is given or the file ends. Opening a file for writing will clear
+%   any existing content in the file (\emph{i.e.}~writing is \emph{not}
+%   additive).
+% \end{function}
+%
+% \begin{function}{\ior_close:N, \ior_close:c}
+%   \begin{syntax}
+%     \cs{ior_close:N} \meta{stream}
+%   \end{syntax}
+%   Closes the \meta{stream}. Streams should always be closed when
+%   they are finished with as this ensures that they remain available
+%   to other programmer. The name of the \meta{stream} will be freed at
+%   this stage, to ensure that any further attempts to read from it results
+%   in an error.
+% \end{function}
+%
+% \begin{function}{\iow_close:N, \iow_close:c}
+%   \begin{syntax}
+%     \cs{iow_close:N} \meta{stream}
+%   \end{syntax}
+%   Closes the \meta{stream}. Streams should always be closed when
+%   they are finished with as this ensures that they remain available
+%   to other programmer. The name of the \meta{stream} will be freed at
+%   this stage, to ensure that any further attempts to write to it results
+%   in an error.
+% \end{function}
+%
+% \begin{function}{\ior_list_streams:, \iow_list_streams:}
+%   \begin{syntax}
+%     \cs{ior_list_streams:} \\
+%     \cs{iow_list_streams:}
+%   \end{syntax}
+%   Displays a list of the file names associated with each open
+%   stream: intended for tracking down problems.
+% \end{function}
+%
+% \section{Writing to files}
+%
+% \begin{function}{\iow_now:Nn, \iow_now:Nx}
+%   \begin{syntax}
+%     \cs{iow_now:Nn} \meta{stream} \Arg{tokens}
+%   \end{syntax}
+%   This functions writes \meta{tokens} to the specified
+%   \meta{stream} immediately (\emph{i.e.}~the write operation is called
+%   on expansion of \cs{iow_now:Nn}).
+%   \begin{texnote}
+%     \cs{iow_now:Nx} is a protected macro which expands to
+%     the two \TeX{} primitives \cs{immediate}\cs{write}.
+%   \end{texnote}
+% \end{function}
+%
+% \begin{function}{\iow_log:n, \iow_log:x}
+%   \begin{syntax}
+%     \cs{iow_log:n} \Arg{tokens}
+%   \end{syntax}
+%   This function writes the given \meta{tokens} to the log (transcript)
+%   file immediately: it is a dedicated version of \cs{iow_now:Nn}.
+% \end{function}
+%
+% \begin{function}{\iow_term:n, \iow_term:x}
+%   \begin{syntax}
+%     \cs{iow_term:n} \Arg{tokens}
+%   \end{syntax}
+%   This function writes the given \meta{tokens} to the terminal
+%   file immediately: it is a dedicated version of \cs{iow_now:Nn}.
+% \end{function}
+%
+% \begin{function}{\iow_now_when_avail:Nn, \iow_now_when_avail:Nx}
+%   \begin{syntax}
+%     \cs{iow_now_when_avail:Nn} \meta{stream} \Arg{tokens}
+%   \end{syntax}
+%   If \meta{stream} is open, writes the \meta{tokens} to the \meta{stream}
+%   in the  same manner as \cs{iow_now:Nn}. If the \meta{stream} is not open,
+%   the \meta{tokens}are simply thrown away.
+% \end{function}
+%
+% \begin{function}{\iow_shipout:Nn, \iow_shipout:Nx}
+%   \begin{syntax}
+%     \cs{iow_shipout:Nn} \meta{stream} \Arg{tokens}
+%   \end{syntax}
+%   This functions writes \meta{tokens} to the specified
+%   \meta{stream} when the current page is finalised (\emph{i.e.}~at
+%   shipout). The \texttt{x}-type variants expand the \meta{tokens}
+%   at the point where the function is used but \emph{not} when the
+%   resulting tokens are written to the \meta{stream}
+%   (\emph{cf.}~\cs{iow_shipout_x:Nn}).
+% \end{function}
+%
+% \begin{function}{\iow_shipout_x:Nn, \iow_shipout_x:Nx}
+%   \begin{syntax}
+%     \cs{iow_shipout_x:Nn} \meta{stream} \Arg{tokens}
+%   \end{syntax}
+%   This functions writes \meta{tokens} to the specified
+%   \meta{stream} when the current page is finalised (\emph{i.e}.~at
+%   shipout). The \meta{tokens} are expanded at the time of writing
+%   in addition to any expansion when the function is used. This makes
+%   these functions suitable for including material finalised during
+%   the page building process (such as the page number integer).
+%   \begin{texnote}
+%     \cs{iow_shipout_x:Nn} is the \TeX{} primitive \cs{write} renamed.
+%   \end{texnote}
+% \end{function}
+%
+% \begin{function}[EXP]{\iow_char:N}
+%   \begin{syntax}
+%     \cs{iow_char:N} \meta{token}
+%   \end{syntax}
+%   Inserts \meta{token} into the output stream. Useful when trying to
+%   write difficult characters such as "%", "{", "}",
+%   \emph{etc}.~in messages, for example:
+%   \begin{verbatim}
+%     \iow_now:Nx \g_my_stream { \iow_char:N \{ text \iow_char:N \} }
+%   \end{verbatim}
+%   The function has no effect if writing is taking place without
+%   expansion (\emph{e.g.}~in the second argument of \cs{iow_now:Nn}).
+% \end{function}
+%
+% \begin{function}[EXP]{\iow_newline:}
+%   \begin{syntax}
+%     \cs{iow_newline:}
+%   \end{syntax}
+%   Function to add a new line within the \meta{tokens} written to a
+%   file. The function has no effect if writing is taking place without
+%   expansion (\emph{e.g.}~in the second argument of \cs{iow_now:Nn}).
+% \end{function}
+%
+% \section{Wrapping lines in output}
+%
+% \begin{function}{\iow_wrap:xnnnN}
+%   \begin{syntax}
+%     \cs{iow_wrap:xnnnN} \Arg{text} \Arg{run-on text} \Arg{run-on length}
+%     ~~\Arg{set up} \meta{function}
+%   \end{syntax}
+%   This function will wrap the \meta{text} to a fixed number of
+%   characters per line. At the start of each line which is wrapped,
+%   the \meta{run-on text} will be inserted.  The line length targeted
+%   will be the value of \cs{l_iow_line_length_int} minus the
+%   \meta{run-on length}. The later value should be the number of
+%   characters in the \meta{run-on text}. Additional functions may be
+%   added to the wrapping by using the \meta{set up}, which is executed
+%   before the wrapping takes place. The result of the wrapping operation
+%   is passed as a braced argument to the \meta{function}, which will
+%   typically be a wrapper around a writing operation. Within the
+%   \meta{text}, |\\| may be used to force a new line and \verb|\ | may be
+%   used to represent a forced space (for example after a control
+%   sequence). Both the wrapping process and the subsequent write operation
+%   will perform \texttt{x}-type expansion. For this reason, material which
+%   is to be written \enquote{as is} should be given as the argument to
+%   \cs{token_to_str:N} or \cs{tl_to_str:n} (as appropriate) within
+%   the \meta{text}. The output of \cs{iow_wrap:xnnnN} (\emph{i.e.}~the
+%   argument passed to the \meta{function}) will consist of characters of
+%   category code $12$ (other) and $10$ (space) only. This means that the
+%   output will \emph{not} expand further when written to a file.
+% \end{function}
+%
+% \begin{variable}{\l_iow_line_length_int}
+%   The maximum length of a line to be written by the \cs{iow_wrap:xxnnN}
+%   function. This value depends on the \TeX{} system in use: the standard
+%   value is $78$, which is typically correct for unmodified \TeX{}live
+%   and MiK\TeX{} systems.
+% \end{variable}
+%
+% \section{Reading from files}
+%
+% \begin{function}{\ior_to:NN}
+%   \begin{syntax}
+%     \cs{ior_to:NN} \meta{stream} \meta{token list variable}
+%   \end{syntax}
+%   Functions that reads one or more lines (until an equal number of left
+%   and right braces are found) from the input \meta{stream} and stores
+%   the result locally in the \meta{token list} variable. If the
+%   \meta{stream} is not open, input is requested from the terminal.
+%   The material read from the \meta{stream} will be tokenized by
+%   \TeX{} according to the category codes in force when the function
+%   is used.
+%   \begin{texnote}
+%     The is protected macro which expands to the \TeX{} primitive \cs{read}
+%     along with the |to| keyword.
+%   \end{texnote}
+% \end{function}
+%
+% \begin{function}{\ior_gto:NN}
+%   \begin{syntax}
+%     \cs{ior_gto:NN} \meta{stream} \meta{token list variable}
+%   \end{syntax}
+%   Functions that reads one or more lines (until an equal number of left
+%   and right braces are found) from the input \meta{stream} and stores
+%   the result globally in the \meta{token list} variable. If the
+%   \meta{stream} is not open, input is requested from the terminal.
+%   The material read from the \meta{stream} will be tokenized by
+%   \TeX{} according to the category codes in force when the function
+%   is used.
+%   \begin{texnote}
+%     The is protected macro which expands to the \TeX{} primitives
+%     \cs{global} \cs{read} along with the |to| keyword.
+%   \end{texnote}
+% \end{function}
+%
+% \begin{function}{\ior_str_to:NN}
+%   \begin{syntax}
+%     \cs{ior_str_to:NN} \meta{stream} \meta{token list variable}
+%   \end{syntax}
+%   Functions that reads one or more lines (until an equal number of left
+%   and right braces are found) from the input \meta{stream} and stores
+%   the result locally in the \meta{token list} variable. If the
+%   \meta{stream} is not open, input is requested from the terminal.
+%   The material read from the \meta{stream} as a series of tokens with
+%   category code $12$ (other), with the exception of space
+%   characters which are given category code $10$ (space).
+%   \begin{texnote}
+%     The is protected macro which expands to the \eTeX{} primitive
+%     \cs{readline} along with the |to| keyword.
+%   \end{texnote}
+% \end{function}
+%
+% \begin{function}{\ior_str_gto:NN}
+%   \begin{syntax}
+%     \cs{ior_str_gto:NN} \meta{stream} \meta{token list variable}
+%   \end{syntax}
+%   Functions that reads one or more lines (until an equal number of left
+%   and right braces are found) from the input \meta{stream} and stores
+%   the result globally in the \meta{token list} variable. If the
+%   \meta{stream} is not open, input is requested from the terminal.
+%   The material read from the \meta{stream} as a series of tokens with
+%   category code $12$ (other), with the exception of space
+%   characters which are given category code $10$ (space).
+%   \begin{texnote}
+%     The is protected macro which expands to the primitives
+%     \cs{global} \cs{readline} along with the |to| keyword.
+%   \end{texnote}
+% \end{function}
+%
+%\begin{function}[EXP,pTF]{\ior_if_eof:N}
+%  \begin{syntax}
+%    \cs{ior_if_eof_p:N} \meta{stream}
+%    \cs{ior_if_eof:NTF} \meta{stream} \Arg{true code} \Arg{false code}
+%  \end{syntax}
+%  Tests if the end of a \meta{stream} has been reached during a reading
+%  operation. The test will also return a \texttt{true} value if
+%  the \meta{stream} is not open or the \meta{file name} associated with
+%  a \meta{stream} does not exist at all. The branching versions then
+%  leave either \meta{true code} or \meta{false code} in the input
+%  stream, as appropriate to the truth of the test and the variant of
+%  the function chosen. The logical truth of the test is left in the
+%  input stream by the predicate version.
+%\end{function}
+%
+% \section{Internal input--output functions}
+% 
+% \begin{function}[EXP]{\if_eof:w}
+%   \begin{syntax}
+%     \cs{if_eof:w} \meta{stream}
+%     ~~\meta{true code}
+%     \cs{else:}
+%     ~~\meta{false code}
+%     \cs{fi:}
+%   \end{syntax}
+%   Tests if the \meta{stream} returns \enquote{end of file}, which is true
+%   for non-existent files. The \cs{else:} branch is optional.
+%   \begin{texnote}
+%     This is the \TeX{} primitive \cs{ifeof}.
+%   \end{texnote}
+% \end{function}
+%
+% \begin{function}{\ior_raw_new:N, \ior_raw_new:c}
+%   \begin{syntax}
+%     \cs{ior_raw_new:N} \meta{stream}
+%   \end{syntax}
+%   Directly allocates a new stream for reading, bypassing the stack
+%   system. This is to be used only when a new stream is required at a
+%   \TeX{} level, when a new stream is requested by the stack itself.
+% \end{function}
+%
+% \begin{function}{\iow_raw_new:N, \iow_raw_new:c}
+%   \begin{syntax}
+%     \cs{iow_raw_new:N} \meta{stream}
+%   \end{syntax}
+%   Directly allocates a new stream for writing, bypassing the stack
+%   system. This is to be used only when a new stream is required at a
+%   \TeX{} level, when a new stream is requested by the stack itself.
+% \end{function}
+%
+% \end{documentation}
+%
+% \begin{implementation}
+%
+% \section{\pkg{l3io} implementation}
+%
+%    \begin{macrocode}
+%<*initex|package>
+%    \end{macrocode}
+%
+%    \begin{macrocode}
+%<*package>
+\ProvidesExplPackage
+  {\ExplFileName}{\ExplFileDate}{\ExplFileVersion}{\ExplFileDescription}
+\package_check_loaded_expl:
+%</package>
+%    \end{macrocode}
+%    
+% \subsection{Primitives}
+%
+% \begin{macro}{\if_eof:w}
+%   The primitive conditional
+%    \begin{macrocode}
+\cs_new_eq:NN \if_eof:w \tex_ifeof:D
+%    \end{macrocode}
+% \end{macro}
+%
+% \subsection{Variables and constants}
+%
+% \begin{variable}{\c_iow_term_stream, \c_ior_term_stream}
+% \begin{variable}{\c_iow_log_stream, \c_ior_log_stream}
+%   Here we allocate two output streams for writing to the transcript
+%   file only (\cs{c_iow_log_stream}) and to both the terminal and
+%   transcript file (\cs{c_iow_term_stream}). Both can be used to read
+%   from and have equivalent \cs{c_ior} versions.
+%    \begin{macrocode}
+\cs_new_eq:NN \c_iow_term_stream \c_sixteen
+\cs_new_eq:NN \c_ior_term_stream \c_sixteen
+\cs_new_eq:NN \c_iow_log_stream  \c_minus_one
+\cs_new_eq:NN \c_ior_log_stream  \c_minus_one
+%    \end{macrocode}
+% \end{variable}
+% \end{variable}
+%
+% \begin{variable}{\c_iow_streams_tl, \c_ior_streams_tl}
+%   The list of streams available, by number.
+%    \begin{macrocode}
+\tl_const:Nn \c_iow_streams_tl
+  {
+    \c_zero
+    \c_one
+    \c_two
+    \c_three
+    \c_four
+    \c_five
+    \c_six
+    \c_seven
+    \c_eight
+    \c_nine
+    \c_ten
+    \c_eleven
+    \c_twelve
+    \c_thirteen
+    \c_fourteen
+    \c_fifteen
+  }
+\cs_new_eq:NN \c_ior_streams_tl \c_iow_streams_tl
+%    \end{macrocode}
+% \end{variable}
+%
+% \begin{variable}{\g_iow_streams_prop, \g_ior_streams_prop}
+%   The allocations for streams are stored in property lists, which
+%   are set up to have a \enquote{full} set of allocations from the start.
+%   In package mode, a few slots are always taken, so these are
+%   blocked off from use.
+%    \begin{macrocode}
+\prop_new:N \g_iow_streams_prop
+\prop_new:N \g_ior_streams_prop
+%<*package>
+\prop_put:Nnn \g_iow_streams_prop { 0 } { LaTeX2e~reserved }
+\prop_put:Nnn \g_iow_streams_prop { 1 } { LaTeX2e~reserved }
+\prop_put:Nnn \g_iow_streams_prop { 2 } { LaTeX2e~reserved }
+\prop_put:Nnn \g_ior_streams_prop { 0 } { LaTeX2e~reserved }
+%</package>
+%    \end{macrocode}
+% \end{variable}
+%
+% \begin{variable}{\l_iow_stream_int, \l_ior_stream_int}
+%   Used to track the number allocated to the stream being created:
+%   this is taken from the property list but does alter.
+%    \begin{macrocode}
+\int_new:N \l_iow_stream_int
+\cs_new_eq:NN \l_ior_stream_int \l_iow_stream_int
+%    \end{macrocode}
+% \end{variable}
+%
+% \subsection{Stream management}
+%
+% \begin{macro}[int]{\ior_raw_new:N, \ior_raw_new:c}
+% \begin{macro}[int]{\iow_raw_new:N, \iow_raw_new:c}
+% The lowest level for stream management is actually creating raw \TeX{}
+% streams. As these are very limited (even with \eTeX{}), this should not
+% be addressed directly.
+%    \begin{macrocode}
+%<*initex>
+\alloc_setup_type:nnn { ior } \c_zero \c_sixteen
+\cs_new_protected_nopar:Npn \ior_raw_new:N #1
+  { \alloc_reg:NnNN g { ior } \tex_chardef:D #1 }
+\alloc_setup_type:nnn { iow } \c_zero \c_sixteen
+\cs_new_protected_nopar:Npn \iow_raw_new:N #1
+  { \alloc_reg:NnNN g { iow } \tex_chardef:D #1 }
+%</initex>
+%<*package>
+\cs_set_eq:NN \iow_raw_new:N \newwrite
+\cs_set_eq:NN \ior_raw_new:N \newread
+%</package>
+\cs_generate_variant:Nn \ior_raw_new:N { c }
+\cs_generate_variant:Nn \iow_raw_new:N { c }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\ior_open:Nn, \ior_open:cn}
+% \begin{macro}{\iow_open:Nn, \iow_open:cn}
+%   In both cases, opening a stream starts with a call to the closing
+%   function: this is safest. There is then a loop through the
+%   allocation number list to find the first free stream number.
+%   When one is found the allocation can take place, the information
+%   can be stored and finally the file can actually be opened.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \ior_open:Nn #1#2
+  {
+    \ior_close:N #1
+    \int_set:Nn \l_ior_stream_int \c_sixteen
+    \tl_map_function:NN \c_ior_streams_tl \ior_alloc_read:n
+    \int_compare:nNnTF \l_ior_stream_int = \c_sixteen
+      { \msg_kernel_error:nn { ior } { streams-exhausted } }
+      {
+        \ior_stream_alloc:N #1
+        \prop_gput:NVn \g_ior_streams_prop \l_ior_stream_int {#2}
+        \tex_openin:D #1#2 \scan_stop:
+      }
+  }
+\cs_new_protected_nopar:Npn \iow_open:Nn #1#2
+  {
+    \iow_close:N #1
+    \int_set:Nn \l_iow_stream_int \c_sixteen
+    \tl_map_function:NN \c_iow_streams_tl \iow_alloc_write:n
+    \int_compare:nNnTF \l_iow_stream_int = \c_sixteen
+      { \msg_kernel_error:nn { iow } { streams-exhausted } }
+      {
+        \iow_stream_alloc:N #1
+        \prop_gput:NVn \g_iow_streams_prop \l_iow_stream_int {#2}
+        \tex_immediate:D \tex_openout:D #1#2 \scan_stop:
+      }
+  }
+\cs_generate_variant:Nn \ior_open:Nn { c }
+\cs_generate_variant:Nn \iow_open:Nn { c }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}[aux]{\ior_alloc_read:n}
+% \begin{macro}[aux]{\iow_alloc_write:n}
+% These functions are used to see if a particular stream is available.
+% The property list contains file names for streams in use, so
+% any unused ones are for the taking.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \iow_alloc_write:n #1
+  {
+    \prop_if_in:NnF \g_iow_streams_prop {#1}
+      {
+        \int_set:Nn \l_iow_stream_int {#1}
+        \tl_map_break:
+      }
+  }
+\cs_new_protected_nopar:Npn \ior_alloc_read:n #1
+  {
+    \prop_if_in:NnF \g_iow_streams_prop {#1}
+      {
+        \int_set:Nn \l_ior_stream_int {#1}
+        \tl_map_break:
+      }
+  }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}[aux]{\iow_stream_alloc:N}
+% \begin{macro}[aux]{\ior_stream_alloc:N}
+% \begin{macro}[aux]{\iow_stream_alloc_aux:}
+% \begin{macro}[aux]{\ior_stream_alloc_aux:}
+% \begin{variable}{\g_iow_tmp_stream}
+% \begin{variable}{\g_ior_tmp_stream}
+%   Allocating a raw stream is much easier in \IniTeX{} mode than for
+%   the package. For the format, all streams will be allocated by
+%   \pkg{l3io} and so there is a simple check to see if a raw
+%   stream is actually available. On the other hand, for the
+%   package there will be non-managed streams. So if the managed
+%   one is not open, a check is made to see if some other managed
+%   stream is available before deciding to open a new one. If a new
+%   one is needed, we get the number allocated by \LaTeXe{} to get
+%   \enquote{back on track} with allocation.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \iow_stream_alloc:N #1
+  {
+    \cs_if_exist:cTF { g_iow_ \int_use:N \l_iow_stream_int _stream }
+      { \cs_gset_eq:Nc #1 { g_iow_ \int_use:N \l_iow_stream_int _stream } }
+      {
+%<*package>
+        \iow_stream_alloc_aux:
+        \int_compare:nNnT \l_iow_stream_int = \c_sixteen
+          {
+            \iow_raw_new:N \g_iow_tmp_stream
+            \int_set:Nn \l_iow_stream_int { \g_iow_tmp_stream }
+            \cs_gset_eq:cN
+              { g_iow_ \int_use:N \l_iow_stream_int _stream }
+              \g_iow_tmp_stream
+          }
+%</package>
+%<*initex>
+        \iow_raw_new:c { g_iow_ \int_use:N \l_iow_stream_int _stream }
+%</initex>
+        \cs_gset_eq:Nc #1 { g_iow_ \int_use:N \l_iow_stream_int _stream }
+      }
+  }
+%<*package>
+\cs_new_protected_nopar:Npn \iow_stream_alloc_aux:
+  {
+    \int_incr:N \l_iow_stream_int
+    \int_compare:nNnT \l_iow_stream_int < \c_sixteen
+      {
+         \cs_if_exist:cTF { g_iow_ \int_use:N \l_iow_stream_int _stream }
+           {
+             \prop_if_in:NVT \g_iow_streams_prop \l_iow_stream_int
+               { \iow_stream_alloc_aux: }
+           }
+           { \iow_stream_alloc_aux: }
+      }
+  }
+%</package>
+\cs_new_protected_nopar:Npn \ior_stream_alloc:N #1
+  {
+    \cs_if_exist:cTF { g_ior_ \int_use:N \l_ior_stream_int _stream }
+      { \cs_gset_eq:Nc #1 { g_ior_ \int_use:N \l_ior_stream_int _stream } }
+      {
+%<*package>
+        \ior_stream_alloc_aux:
+        \int_compare:nNnT \l_ior_stream_int = \c_sixteen
+          {
+            \ior_raw_new:N \g_ior_tmp_stream
+            \int_set:Nn \l_ior_stream_int { \g_ior_tmp_stream }
+            \cs_gset_eq:cN
+              { g_ior_ \int_use:N \l_iow_stream_int _stream }
+              \g_ior_tmp_stream
+          }
+%</package>
+%<*initex>
+        \ior_raw_new:c { g_ior_ \int_use:N \l_ior_stream_int _stream }
+%</initex>
+        \cs_gset_eq:Nc #1 { g_ior_ \int_use:N \l_ior_stream_int _stream }
+      }
+  }
+%<*package>
+\cs_new_protected_nopar:Npn \ior_stream_alloc_aux:
+  {
+    \int_incr:N \l_ior_stream_int
+    \int_compare:nNnT \l_ior_stream_int < \c_sixteen
+      {
+         \cs_if_exist:cTF { g_ior_ \int_use:N \l_ior_stream_int _stream }
+           {
+             \prop_if_in:NVT \g_ior_streams_prop \l_ior_stream_int
+               { \ior_stream_alloc_aux: }
+           }
+           { \ior_stream_alloc_aux: }
+      }
+  }
+%</package>
+%    \end{macrocode}
+% \end{variable}
+% \end{variable}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\iow_close:N, \iow_close:c}
+% \begin{macro}{\iow_close:N, \iow_close:c}
+%   Closing a stream is not quite the reverse of opening one. First,
+%   the close operation is easier than the open one, and second as the
+%   stream is actually a number we can use it directly to show that the
+%   slot has been freed up.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \ior_close:N #1
+  {
+    \cs_if_exist:NT #1
+      {
+        \int_compare:nNnF #1 = \c_minus_one
+          {
+            \tex_closein:D #1
+            \prop_gdel:NV \g_ior_streams_prop #1
+            \cs_undefine:N #1
+          }
+      }
+  }
+\cs_new_protected_nopar:Npn \iow_close:N #1
+  {
+    \cs_if_exist:NT #1
+      {
+        \int_compare:nNnF #1 = \c_minus_one
+          {
+            \tex_immediate:D \tex_closeout:D #1
+            \prop_gdel:NV \g_iow_streams_prop #1
+            \cs_undefine:N #1
+          }
+      }
+  }
+\cs_generate_variant:Nn \ior_close:N { c }
+\cs_generate_variant:Nn \iow_close:N { c }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\ior_list_streams:}
+% \begin{macro}{\iow_list_streams:}
+% \begin{macro}[aux]{\iow_show_aux:nn}
+% \begin{macro}[aux]{\ior_show_aux:nn}
+% Show the property lists, but with some \enquote{pretty printing}.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \ior_list_streams:
+  {
+    \prop_if_empty:NTF \g_ior_streams_prop
+      {
+        \iow_term:x { No~input~streams~are~open }
+        \tl_show:n { }
+      }
+      {
+        \iow_term:x { The~following~input~streams~are~in~use: }
+        \tl_set:Nx \l_prop_show_tl
+          { \prop_map_function:NN \g_ior_streams_prop \ior_show_aux:nn }
+          \etex_showtokens:D \exp_after:wN \exp_after:wN \exp_after:wN
+            { \exp_after:wN \prop_show_aux:w \l_prop_show_tl }
+      }
+  }
+\cs_new:Npn \ior_show_aux:nn #1#2
+  {
+    \iow_newline: > \c_space_tl \c_space_tl
+    #1 \iow_char:N
+    \c_space_tl \c_space_tl => \c_space_tl \c_space_tl
+    \exp_not:n {#2}
+  }
+\cs_new_protected_nopar:Npn \iow_list_streams:
+  {
+    \prop_if_empty:NTF \g_iow_streams_prop
+      {
+        \iow_term:x { No~output~streams~are~open }
+        \tl_show:n { }
+      }
+      {
+        \iow_term:x { The~following~output~streams~are~in~use: }
+        \tl_set:Nx \l_prop_show_tl
+          { \prop_map_function:NN \g_iow_streams_prop \iow_show_aux:nn }
+          \etex_showtokens:D \exp_after:wN \exp_after:wN \exp_after:wN
+            { \exp_after:wN \prop_show_aux:w \l_prop_show_tl }
+      }
+  }
+\cs_new_eq:NN \iow_show_aux:nn \ior_show_aux:nn
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% Text for the error messages.
+%    \begin{macrocode}
+\msg_kernel_new:nnnn { iow } { streams-exhausted }
+  { Output~streams~exhausted }
+  {
+    TeX~can~only~open~up~to~16~output~streams~at~one~time.\\
+    All~16 are currently~in~use,~and~something~wanted~to~open
+    another~one.
+  }
+\msg_kernel_new:nnnn { ior } { streams-exhausted }
+  { Input~streams~exhausted }
+  {
+    TeX~can~only~open~up~to~16~input~streams~at~one~time.\\
+    All~16 are currently~in~use,~and~something~wanted~to~open
+    another~one.
+  }
+%    \end{macrocode}
+%
+% \subsection{Deferred writing}
+%
+% \begin{macro}{\iow_shipout_x:Nn, \iow_shipout_x:Nx}
+%   First the easy part, this is the primitive.
+%    \begin{macrocode}
+\cs_new_eq:NN \iow_shipout_x:Nn \tex_write:D
+\cs_generate_variant:Nn \iow_shipout_x:Nn { Nx }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\iow_shipout:Nn, \iow_shipout:Nx}
+%   With \eTeX{} available deferred writing is easy.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \iow_shipout:Nn #1#2
+  { \iow_shipout_x:Nn #1 { \exp_not:n {#2} } }
+\cs_generate_variant:Nn \iow_shipout:Nn { Nx }
+%    \end{macrocode}
+% \end{macro}
+%
+% \subsection{Immediate writing}
+%
+% \begin{macro}{\iow_now:Nx}
+%   An abbreviation for an often used operation, which immediately
+%   writes its second argument expanded to the output stream.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \iow_now:Nx { \tex_immediate:D \iow_shipout_x:Nn }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\iow_now:Nn}
+%   This routine writes the second argument onto the output stream without
+%   expansion. If this stream isn't open, the output goes to the terminal
+%   instead. If the first argument is no output stream at all, we get an
+%   internal error.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \iow_now:Nn #1#2
+  { \iow_now:Nx #1 { \exp_not:n {#2} } }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\iow_log:n, \iow_log:x}
+% \begin{macro}{\iow_term:n, \iow_term:x}
+% Writing to the log and the terminal directly are relatively easy.
+%    \begin{macrocode}
+\cs_set_protected_nopar:Npn \iow_log:x  { \iow_now:Nx \c_iow_log_stream  }
+\cs_new_protected_nopar:Npn \iow_log:n  { \iow_now:Nn \c_iow_log_stream  }
+\cs_set_protected_nopar:Npn \iow_term:x { \iow_now:Nx \c_iow_term_stream }
+\cs_new_protected_nopar:Npn \iow_term:n { \iow_now:Nn \c_iow_term_stream }
+%    \end{macrocode}
+%\end{macro}
+%\end{macro}
+%
+% \begin{macro}{\iow_now_when_avail:Nn, \iow_now_when_avail:Nx}
+% For writing only if the stream requested is open at all.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \iow_now_when_avail:Nn #1
+  { \cs_if_free:NTF #1 { \use_none:n } { \iow_now:Nn #1 } }
+\cs_new_protected_nopar:Npn \iow_now_when_avail:Nx #1
+  { \cs_if_free:NTF #1 { \use_none:n } { \iow_now:Nx #1 } }
+%    \end{macrocode}
+% \end{macro}
+%
+% \subsection{Hard-wrapping lines based on length}
+%
+% The code here implements a generic hard-wrapping function. This is
+% used by the messaging system, but is designed such that it is
+% available for other uses.
+%
+% \begin{macro}{\l_iow_line_length_int}
+%   The is the \enquote{raw} length of a line which can be written to
+%   file. The standard value is the line length typically used by
+%   \TeX{}Live and Mik\TeX{}.
+%    \begin{macrocode}
+\int_new:N  \l_iow_line_length_int
+\int_set:Nn \l_iow_line_length_int { 78 }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}[aux]{\l_iow_target_length_int}
+%   This stores the target line length: the full length minus any
+%   part for a leader at the start of each line.
+%    \begin{macrocode}
+\int_new:N \l_iow_target_length_int
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}[aux]{\l_iow_current_line_int, \l_iow_current_word_int}
+%   These store the number of characters in the line and word currently
+%   being constructed, respectively.
+%    \begin{macrocode}
+\int_new:N \l_iow_current_line_int
+\int_new:N \l_iow_current_word_int
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}[aux]{\l_iow_current_line_tl, \l_iow_current_word_tl}
+%   These hold the current line of text and current word, respectively.
+%    \begin{macrocode}
+\tl_new:N \l_iow_current_line_tl
+\tl_new:N \l_iow_current_word_tl
+%    \end{macrocode}
+%\end{macro}
+%
+% \begin{macro}[aux]{\l_iow_wrap_tl}
+%   Used for the expansion step before detokenizing.
+%    \begin{macrocode}
+\tl_new:N \l_iow_wrap_tl
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}[aux]{\l_iow_wrapped_tl}
+%   The output from wrapping text: fully expanded and with lines
+%   which are not overly long.
+%    \begin{macrocode}
+\tl_new:N \l_iow_wrapped_tl
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\q_iow_stop}
+%   A quark which will not appear elsewhere.
+%    \begin{macrocode}
+\quark_new:N \q_iow_stop
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\l_iow_line_start_bool}
+%   Boolean to avoid adding a space at the beginning of lines.
+%    \begin{macrocode}
+\bool_new:N \l_iow_line_start_bool
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\iow_wrap:xnnnN}
+% \begin{macro}[aux]{\iow_wrap_loop:w}
+% \begin{macro}[aux]{\iow_wrap_word:}
+% \begin{macro}[aux]{\iow_wrap_word_fits:}
+% \begin{macro}[aux]{\iow_wrap_word_newline:}
+% \begin{macro}[aux]{\iow_wrap_newline:}
+% \begin{macro}[aux]{\iow_wrap_end:}
+%   The main wrapping function works as follows. The target number of
+%   characters in a line is calculated, before fully-expanding the input
+%   such that |\\| and \verb*|\ | are converted into the appropriate
+%   values. There is then a loop over each word in the input, which
+%   will do the actual wrapping. After the loop, the resulting text is
+%   passed on to the function which has been given as a post-processor.
+%   The argument |#4| is available for additional set up steps for
+%   the output. 
+%    \begin{macrocode}
+\cs_new_protected:Npn \iow_wrap:xnnnN #1#2#3#4#5
+  {
+    \group_begin:
+      \int_set:Nn \l_iow_target_length_int { \l_iow_line_length_int - ( #3 ) }
+      \int_zero:N \l_iow_current_line_int
+      \tl_clear:N \l_iow_current_line_tl
+      \tl_clear:N \l_iow_wrap_tl
+      \bool_set_true:N \l_iow_line_start_bool
+      \cs_set:Npx \\ { \c_space_tl \iow_newline: \c_space_tl }
+      \cs_set_eq:NN \  \c_space_tl
+      #4
+%<*initex>
+      \tl_set:Nx \l_iow_wrap_tl {#1}
+%</initex>
+%<*package>
+      \protected@edef \l_iow_wrap_tl {#1}
+%</package>
+      \cs_set:Npn \\ { \iow_newline: #2 }
+      \use:x
+        {
+          \exp_not:N \iow_wrap_loop:w
+          \tl_to_str:N \l_iow_wrap_tl \c_space_tl
+          \exp_not:N \q_iow_stop \c_space_tl
+        }
+    \exp_args:NNo \group_end:
+    #5 \l_iow_wrapped_tl
+  }
+%    \end{macrocode}
+% The loop grabs one word in the input, and checks whether it is
+% the end, or a forced new line, or a normal word.
+%    \begin{macrocode}
+\cs_new_protected:Npn \iow_wrap_loop:w #1 ~ %
+  {
+    \tl_set:Nn \l_iow_current_word_tl {#1}
+    \tl_if_eq:NNTF \l_iow_current_word_tl \iow_newline:
+      { \iow_wrap_newline: }
+      {
+        \tl_if_eq:NNTF \l_iow_current_word_tl \q_iow_stop
+          { \iow_wrap_end: }
+          { \iow_wrap_word: }
+      }
+  }
+%    \end{macrocode}
+% For a normal word, update the line length, then test if the current
+% word would fit in the current line, and call the appropriate function.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \iow_wrap_word:
+  {
+    \int_set:Nn \l_iow_current_word_int
+      { \str_length_skip_spaces:N \l_iow_current_word_tl }
+    \int_add:Nn \l_iow_current_line_int { \l_iow_current_word_int }
+    \int_compare:nNnTF \l_iow_current_line_int
+                     < \l_iow_target_length_int
+      { \iow_wrap_word_fits: }
+      { \iow_wrap_word_newline: }
+    \iow_wrap_loop:w
+  }
+%    \end{macrocode}
+% If the word fits in the current line, add it to the line, preceded by
+% a space unless it is the first word of the line.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \iow_wrap_word_fits:
+  {
+    \bool_if:NTF \l_iow_line_start_bool
+      {
+        \bool_set_false:N \l_iow_line_start_bool
+        \tl_set_eq:NN \l_iow_current_line_tl \l_iow_current_word_tl
+      }
+      {
+        \tl_put_right:Nx \l_iow_current_line_tl
+          { ~ \l_iow_current_word_tl }
+        \int_incr:N \l_iow_current_line_int
+      }
+  }
+%    \end{macrocode}
+% Otherwise, the current line is added to the result, with the run-on text.
+% The current word (and its length) are then put in the new line.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \iow_wrap_word_newline:
+  {
+    \tl_put_right:Nx \l_iow_wrapped_tl
+      { \l_iow_current_line_tl \\ }
+    \int_set_eq:NN \l_iow_current_line_int \l_iow_current_word_int
+    \tl_set_eq:NN  \l_iow_current_line_tl  \l_iow_current_word_tl
+  }
+%    \end{macrocode}
+% Forced newlines are almost identical to those caused by overflow,
+% except that here the word is empty. And remember to continue the loop!
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \iow_wrap_newline:
+  {
+    \tl_put_right:Nx \l_iow_wrapped_tl
+      { \l_iow_current_line_tl \\ }
+    \int_zero:N \l_iow_current_line_int
+    \tl_clear:N \l_iow_current_line_tl
+    \bool_set_true:N \l_iow_line_start_bool
+    \iow_wrap_loop:w
+  }
+%    \end{macrocode}
+% At the end, we simply save the last line (without the run-on text).
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \iow_wrap_end:
+  {
+    \tl_put_right:Nx \l_iow_wrapped_tl
+      { \l_iow_current_line_tl }
+  }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\str_length_skip_spaces:N}
+% \begin{macro}{\str_length_skip_spaces:n}
+% \begin{macro}{\str_length_loop:NNNNNNNNN}
+%   The wrapping code requires to measure the number of character
+%   in each word. This could be done with \cs{tl_length:n}, but
+%   it is ten times faster (literally) to use the code below.
+%    \begin{macrocode}
+\cs_new_nopar:Npn \str_length_skip_spaces:N
+  { \exp_args:No \str_length_skip_spaces:n }
+\cs_new:Npn \str_length_skip_spaces:n #1
+  {
+    \int_value:w \int_eval:w
+      \exp_after:wN \str_length_loop:NNNNNNNNN \tl_to_str:n {#1}
+      {X8}{X7}{X6}{X5}{X4}{X3}{X2}{X1}{X0} \q_stop
+    \int_eval_end:
+  }
+\cs_new:Npn \str_length_loop:NNNNNNNNN #1#2#3#4#5#6#7#8#9
+  {
+    \if_catcode:w X #9
+      \exp_after:wN \use_none_delimit_by_q_stop:w
+    \else:
+      9 + 
+      \exp_after:wN \str_length_loop:NNNNNNNNN
+    \fi:
+  }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \subsection{Special characters for writing}
+%
+% \begin{macro}{\iow_newline:}
+%   Global variable holding the character that forces a new line when
+%   something is written to an output stream
+%    \begin{macrocode}
+\cs_new_nopar:Npn \iow_newline: { ^^J }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\iow_char:N}
+%   Function to write any escaped char to an output stream.
+%    \begin{macrocode}
+\cs_new_eq:NN \iow_char:N \cs_to_str:N
+%    \end{macrocode}
+% \end{macro}
+%
+% \subsection{Reading input}
+%
+% \begin{macro}[pTF]{\ior_if_eof_p:N}
+% To test if some particular input stream is exhausted the following
+% conditional is provided. As the pool model means that closed
+% streams are undefined control sequences, the test has two parts.
+%    \begin{macrocode}
+\prg_new_conditional:Nnn \ior_if_eof:N { p , T , F , TF }
+  {
+    \cs_if_exist:NTF #1
+      {
+        \if_eof:w #1
+          \prg_return_true:
+        \else:
+          \prg_return_false:
+        \fi:
+      }
+      { \prg_return_true: }
+  }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\ior_to:NN}
+% \begin{macro}{\ior_gto:NN}
+%  And here we read from files.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \ior_to:NN #1#2
+  { \tex_read:D #1 to #2 }
+\cs_new_protected_nopar:Npn \ior_gto:NN #1#2
+  { \pref_global:D \tex_read:D #1 to #2 }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\ior_str_to:NN}
+% \begin{macro}{\ior_str_gto:NN}
+%  Reading as strings is also a primitive wrapper.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \ior_str_to:NN #1#2
+  { \etex_readline:D #1 to #2 }
+\cs_new_protected_nopar:Npn \ior_str_gto:NN #1#2
+  { \pref_global:D \etex_readline:D #1 to #2 }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \subsection{Deprecated functions}
+%
+% Deprecated on 2011-05-27, for removal by 2011-08-31.
+%
+% \begin{macro}{\iow_now_buffer_safe:Nn, \iow_now_buffer_safe:Nx}
+%   This is much more easily done using the wrapping system: there is
+%   an expansion there, so a bit of a hack is needed.
+%    \begin{macrocode}
+\cs_new_protected:Npn \iow_now_buffer_safe:Nn #1#2
+  { \iow_wrap:xnnnN { \exp_not:n {#2} } { } \c_zero { } \iow_now:Nn #1 }
+\cs_new_protected:Npn \iow_now_buffer_safe:Nx #1#2
+  { \iow_wrap:xnnnN {#2} { } \c_zero { } \iow_now:Nn #1 }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\ior_new:N, \ior_new:c}
+% \begin{macro}{\iow_new:N, \iow_new:c}
+%   As input--output operations are done using a stack, |new| operations seem
+%   out-of-place. They are therefore set up just to gobble the input.
+%    \begin{macrocode}
+\cs_new_eq:NN \ior_new:N \use_none:n
+\cs_new_eq:NN \ior_new:c \use_none:n
+\cs_new_eq:NN \iow_new:N \use_none:n
+\cs_new_eq:NN \iow_new:c \use_none:n
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\ior_open_streams:}
+% \begin{macro}{\iow_open_streams:}
+%   Slightly misleading names.
+%    \begin{macrocode}
+\cs_new_eq:NN \ior_open_streams: \ior_list_streams:
+\cs_new_eq:NN \iow_open_streams: \iow_list_streams:
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+%    \begin{macrocode}
+%</initex|package>
+%    \end{macrocode}
+%
+% \end{implementation}
+%
+% \PrintIndex
diff --git a/Master/texmf-dist/source/latex/l3kernel/l3keys.dtx b/Master/texmf-dist/source/latex/l3kernel/l3keys.dtx
new file mode 100644
index 00000000000..cbc1fb62a7d
--- /dev/null
+++ b/Master/texmf-dist/source/latex/l3kernel/l3keys.dtx
@@ -0,0 +1,1365 @@
+% \iffalse meta-comment
+%
+%% File: l3keys.dtx Copyright (C) 2009-2011 The LaTeX3 Project
+%%
+%% It may be distributed and/or modified under the conditions of the
+%% LaTeX Project Public License (LPPL), either version 1.3c of this
+%% license or (at your option) any later version.  The latest version
+%% of this license is in the file
+%%
+%%    http://www.latex-project.org/lppl.txt
+%%
+%% This file is part of the "expl3 bundle" (The Work in LPPL)
+%% and all files in that bundle must be distributed together.
+%%
+%% The released version of this bundle is available from CTAN.
+%%
+%% -----------------------------------------------------------------------
+%%
+%% The development version of the bundle can be found at
+%%
+%%    http://www.latex-project.org/svnroot/experimental/trunk/
+%%
+%% for those people who are interested.
+%%
+%%%%%%%%%%%
+%% NOTE: %%
+%%%%%%%%%%%
+%%
+%%   Snapshots taken from the repository represent work in progress and may
+%%   not work or may contain conflicting material!  We therefore ask
+%%   people _not_ to put them into distributions, archives, etc. without
+%%   prior consultation with the LaTeX3 Project.
+%%
+%% -----------------------------------------------------------------------
+%
+%<*driver|package>
+\RequirePackage{l3names}
+\GetIdInfo$Id: l3keys.dtx 2478 2011-06-19 21:34:23Z joseph $
+  {L3 Experimental key-value interfaces}
+%</driver|package>
+%<*driver>
+\documentclass[full]{l3doc}
+\begin{document}
+  \DocInput{\jobname.dtx}
+\end{document}
+%</driver>
+% \fi
+%
+% \title{^^A
+%   The \pkg{l3keys} package\\ Key--value interfaces^^A
+%   \thanks{This file describes v\ExplFileVersion,
+%      last revised \ExplFileDate.}^^A
+% }
+%
+% \author{^^A
+%  The \LaTeX3 Project\thanks
+%    {^^A
+%      E-mail:
+%        \href{mailto:latex-team@latex-project.org}
+%          {latex-team@latex-project.org}^^A
+%    }^^A
+% }
+%
+% \date{Released \ExplFileDate}
+%
+% \maketitle
+%
+% \begin{documentation}
+%
+% The key--value method is a popular system for creating large numbers
+% of settings for controlling function or package behaviour.  For the
+% user, the system normally results in input of the form
+% \begin{verbatim}
+%   \PackageControlMacro{
+%     key-one = value one,
+%     key-two = value two
+%   }
+% \end{verbatim}
+% or
+% \begin{verbatim}
+%   \PackageMacro[
+%     key-one = value one,
+%     key-two = value two
+%   ]{argument}.
+% \end{verbatim}
+%
+% The high level functions here are intended as a method to create
+% key--value controls. Keys are themselves created using a key--value
+% interface, minimising the number of functions and arguments
+% required. Each key is created by setting one or more \emph{properties}
+% of the key:
+% \begin{verbatim}
+%   \keys_define:nn { module }
+%     {
+%       key-one .code:n   = code including parameter #1,
+%       key-two .tl_set:N = \l_module_store_tl
+%     }
+% \end{verbatim}
+% These values can then be set as with other key--value approaches:
+% \begin{verbatim}
+%   \keys_set:nn { module }
+%     {
+%       key-one = value one,
+%       key-two = value two
+%     }
+% \end{verbatim}
+%
+% At a document level, \cs{keys_set:nn} will be used within a
+% document function, for example
+% \begin{verbatim}
+%   \DeclareDocumentCommand \SomePackageSetup { m }
+%     { \keys_set:nn { module } { #1 }  }
+%  \DeclareDocumentCommand \SomePackageMacro { o m }
+%    {
+%      \group_begin:
+%        \keys_set:nn { module } { #1 }
+%        % Main code for \SomePackageMacro
+%      \group_end:
+%    }
+% \end{verbatim}
+%
+% Key names may contain any tokens, as they are handled internally
+% using \cs{tl_to_str:n}. As will be discussed in
+% section~\ref{sec:subdivision}, it is suggested that the character
+% |/| is reserved for sub-division of keys into logical
+% groups. Functions and variables are \emph{not} expanded when creating
+% key names, and so
+% \begin{verbatim}
+%   \tl_set:Nn \l_module_tmp_tl { key }
+%   \keys_define:nn { module }
+%     {
+%       \l_module_tmp_tl .code:n = code
+%     }
+% \end{verbatim}
+% will create a key called \cs{l_module_tmp_tl}, and not one called
+% \texttt{key}.
+%
+% \section{Creating keys}
+%
+% \begin{function}{\keys_define:nn}
+%   \begin{syntax}
+%     \cs{keys_define:nn} \Arg{module} \Arg{keyval list}
+%   \end{syntax}
+%   Parses the \meta{keyval list} and defines the keys listed there for
+%   \meta{module}. The \meta{module} name should be a text value, but
+%   there are no restrictions on the nature of the text. In practice the
+%   \meta{module} should be chosen to be unique to the module in question
+%   (unless deliberately adding keys to an existing module).
+%
+%   The \meta{keyval list} should consist of one or more key names along
+%   with an associated key \emph{property}. The properties of a key
+%   determine how it acts. The individual properties are described
+%   in the following text; a typical use of \cs{keys_define:nn} might
+%   read
+%   \begin{verbatim}
+%     \keys_define:nn { mymodule }
+%       {
+%         keyname .code:n = Some~code~using~#1,
+%         keyname .value_required:
+%       }
+%   \end{verbatim}
+%   where the properties of the key begin from the |.| after the key
+%   name.
+% \end{function}
+%
+% The various properties available take either no arguments at
+% all, or require exactly one argument. This is indicated in the
+% name of the property using an argument specification. In the following
+% discussion, each property is illustrated attached to an
+% arbitrary \meta{key}, which when used may be supplied with a
+% \meta{value}. All key \emph{definitions} are local.
+%
+% \begin{function}{.bool_set:N}
+%   \begin{syntax}
+%     \meta{key} .bool_set:N = \meta{boolean}
+%   \end{syntax}
+%   Defines \meta{key} to set \meta{boolean} to \meta{value} (which
+%   must be either \texttt{true} or \texttt{false}).  If the variable
+%   does not exist, it will be created at the point that the key is
+%   set up. The \meta{boolean} will be assigned locally.
+% \end{function}
+%
+% \begin{function}{.bool_gset:N}
+%   \begin{syntax}
+%     \meta{key} .bool_gset:N = \meta{boolean}
+%   \end{syntax}
+%   Defines \meta{key} to set \meta{boolean} to \meta{value} (which
+%   must be either \texttt{true} or \texttt{false}).  If the variable
+%   does not exist, it will be created at the point that the key is
+%   set up. The \meta{boolean} will be assigned globally.
+% \end{function}
+%
+% \begin{function}{.choice:}
+%   \begin{syntax}
+%     \meta{key} .choice:
+%   \end{syntax}
+%   Sets \meta{key} to act as a multiple choice key. Each valid choice
+%   for \meta{key} must then be created, as discussed in
+%   section~\ref{sec:choice}.
+% \end{function}
+%
+% \begin{function}{.choice_code:n, .choice_code:x}
+%   \begin{syntax}
+%     \meta{key} .choice_code:n = \meta{code}
+%   \end{syntax}
+%   Stores \meta{code} for use when \texttt{.generate_choices:n} creates
+%   one or more choice sub-keys of the current key. Inside \meta{code},
+%   \cs{l_keys_choice_tl} will expand to the name of the choice made, and
+%   \cs{l_keys_choice_int} will be the position of the choice in the list
+%   given to \texttt{.generate_choices:n}. Choices are discussed in
+%   detail in section~\ref{sec:choice}.
+% \end{function}
+%
+% \begin{function}{.code:n, .code:x}
+%   \begin{syntax}
+%     \meta{key} .code:n = \meta{code}
+%   \end{syntax}
+%   Stores the \meta{code} for execution when \meta{key} is used. The
+%   The \meta{code} can include one parameter ("#1"), which will be the
+%   \meta{value} given for the \meta{key}. The \texttt{x}-type variant
+%   will expand \meta{code} at the point  where the \meta{key} is
+%   created.
+% \end{function}
+%
+% \begin{function}{.default:n, .default:V}
+%   \begin{syntax}
+%     \meta{key} .default:n = \meta{default}
+%   \end{syntax}
+%   Creates a \meta{default} value for \meta{key}, which is used if no
+%   value is given. This will be used if only the key name is given,
+%   but not if a blank \meta{value} is given:
+%   \begin{verbatim}
+%     \keys_define:nn { module }
+%       {
+%         key .code:n    = Hello~#1,
+%         key .default:n = World
+%       }
+%     \keys_set:nn { module }
+%       {
+%         key = Fred, % Prints 'Hello Fred'
+%         key,        % Prints 'Hello World'
+%         key = ,     % Prints 'Hello '
+%       }
+%   \end{verbatim}
+% \end{function}
+%
+% \begin{function}{.dim_set:N, .dim_set:c}
+%   \begin{syntax}
+%     \meta{key} .dim_set:N = \meta{dimension}
+%   \end{syntax}
+%   Defines \meta{key} to set \meta{dimension} to \meta{value} (which
+%   must a dimension expression).  If the variable does not exist, it
+%   will be created at the point that the key is set up. The
+%   \meta{dimension} will be assigned locally.
+% \end{function}
+%
+% \begin{function}{.dim_gset:N, .dim_gset:c}
+%   \begin{syntax}
+%     \meta{key} .dim_gset:N = \meta{dimension}
+%   \end{syntax}
+%   Defines \meta{key} to set \meta{dimension} to \meta{value} (which
+%   must a dimension expression).  If the variable does not exist, it
+%   will be created at the point that the key is set up. The
+%   \meta{dimension} will be assigned globally.
+% \end{function}
+%
+% \begin{function}{.fp_set:N, .fp_set:c}
+%   \begin{syntax}
+%     \meta{key} .fp_set:N = \meta{floating point}
+%   \end{syntax}
+%   Defines \meta{key} to set \meta{floating point} to \meta{value}
+%   (which must a integer expression).  If the variable does not exist,
+%   it will be created at the point that the key is set up. The
+%   \meta{integer} will be assigned locally.
+% \end{function}
+%
+% \begin{function}{.fp_gset:N, .fp_gset:c}
+%   \begin{syntax}
+%     \meta{key} .fp_gset:N = \meta{floating point}
+%   \end{syntax}
+%   Defines \meta{key} to set \meta{floating-point} to \meta{value}
+%   (which must a integer expression).  If the variable does not exist,
+%   it will be created at the point that the key is set up. The
+%   \meta{integer} will be assigned globally.
+% \end{function}
+%
+% \begin{function}{.generate_choices:n}
+%   \begin{syntax}
+%     \meta{key} .generate_choices:n = \Arg{list}
+%   \end{syntax}
+%   This property will mark \meta{key} as a multiple choice key,
+%   and will use the \meta{list} to define the choices. The \meta{list}
+%   should consist of a comma-separated list of choice names. Each
+%   choice will be set up to execute \meta{code} as set using
+%   \texttt{.choice_code:n} (or \texttt{.choice_code:x}). Choices are
+%   discussed in detail in section~\ref{sec:choice}.
+% \end{function}
+%
+% \begin{function}{.int_set:N, .int_set:c}
+%   \begin{syntax}
+%     \meta{key} .int_set:N = \meta{integer}
+%   \end{syntax}
+%   Defines \meta{key} to set \meta{integer} to \meta{value} (which
+%   must a integer expression).  If the variable does not exist, it
+%   will be created at the point that the key is set up. The
+%   \meta{integer} will be assigned locally.
+% \end{function}
+%
+% \begin{function}{.int_gset:N, .int_gset:c}
+%   \begin{syntax}
+%     \meta{key} .int_gset:N = \meta{integer}
+%   \end{syntax}
+%   Defines \meta{key} to set \meta{integer} to \meta{value} (which
+%   must a integer expression).  If the variable does not exist, it
+%   will be created at the point that the key is set up. The
+%   \meta{integer} will be assigned globally.
+% \end{function}
+%
+% \begin{function}{.meta:n, .meta:x}
+%   \begin{syntax}
+%     \meta{key} .meta:n = \Arg{keyval list}
+%   \end{syntax}
+%   Makes \meta{key} a meta-key, which will set \meta{keyval list} in
+%   one go.  If \meta{key} is given with a value at the time the key
+%   is used, then the value will be passed through to the subsidiary
+%   \meta{keys} for processing (as |#1|).
+% \end{function}
+%
+% \begin{function}{.skip_set:N, .skip_set:c}
+%   \begin{syntax}
+%     \meta{key} .skip_set:N = \meta{skip}
+%   \end{syntax}
+%   Defines \meta{key} to set \meta{skip} to \meta{value} (which
+%   must a skip expression). If the variable does not exist, it
+%   will be created at the point that the key is set up. The
+%   \meta{skip} will be assigned locally.
+% \end{function}
+%
+% \begin{function}{.skip_gset:N, .skip_gset:c}
+%   \begin{syntax}
+%     \meta{key} .skip_gset:N = \meta{skip}
+%   \end{syntax}
+%   Defines \meta{key} to set \meta{skip} to \meta{value} (which
+%   must a skip expression). If the variable does not exist, it
+%   will be created at the point that the key is set up. The
+%   \meta{skip} will be assigned globally.
+% \end{function}
+%
+% \begin{function}{.tl_set:N, .tl_set:c}
+%   \begin{syntax}
+%     \meta{key} .tl_set:N = \meta{token list variable}
+%   \end{syntax}
+%   Defines \meta{key} to set \meta{token list variable} to \meta{value}.
+%   If the variable does not exist, it will be created at the point that
+%   the key is set up. The \meta{token list variable} will be assigned
+%   locally.
+% \end{function}
+%
+% \begin{function}{.tl_gset:N, .tl_gset:c}
+%   \begin{syntax}
+%     \meta{key} .tl_gset:N = \meta{token list variable}
+%   \end{syntax}
+%   Defines \meta{key} to set \meta{token list variable} to \meta{value}.
+%   If the variable does not exist, it will be created at the point that
+%   the key is set up. The \meta{token list variable} will be assigned
+%   globally.
+% \end{function}
+%
+% \begin{function}{.tl_set_x:N, .tl_set_x:c}
+%   \begin{syntax}
+%     \meta{key} .tl_set_x:N = \meta{token list variable}
+%   \end{syntax}
+%   Defines \meta{key} to set \meta{token list variable} to \meta{value},
+%   which will be subjected to an \texttt{x}-type expansion
+%   (\emph{i.e.}~using \cs{tl_set:Nx}). If the variable does not exist,
+%   it will be created at the point that the key is set up. The
+%   \meta{token list variable} will be assigned locally.
+% \end{function}
+%
+% \begin{function}{.tl_gset_x:N, .tl_gset_x:c}
+%   \begin{syntax}
+%     \meta{key} .tl_gset_x:N = \meta{token list variable}
+%   \end{syntax}
+%   Defines \meta{key} to set \meta{token list variable} to \meta{value},
+%   which will be subjected to an \texttt{x}-type expansion
+%   (\emph{i.e.}~using \cs{tl_set:Nx}). If the variable does not exist,
+%   it will be created at the point that the key is set up. The
+%   \meta{token list variable} will be assigned globally.
+% \end{function}
+%
+% \begin{function}{.value_forbidden:}
+%   \begin{syntax}
+%     \meta{key} .value_forbidden:
+%   \end{syntax}
+%   Specifies that \meta{key} cannot receive a \meta{value} when used.
+%   If a \meta{value} is given then an error will be issued.
+% \end{function}
+%
+% \begin{function}{.value_required:}
+%   \begin{syntax}
+%      \meta{key} .value_required:
+%   \end{syntax}
+%   Specifies that \meta{key} must receive a \meta{value} when used.
+%   If a \meta{value} is not given then an error will be issued.
+% \end{function}
+%
+% \section{Sub-dividing keys}
+% \label{sec:subdivision}
+%
+% When creating large numbers of keys, it may be desirable to divide
+% them into several sub-groups for a given module. This can be achieved
+% either by adding a sub-division to the module name:
+% \begin{verbatim}
+%   \keys_define:nn { module / subgroup }
+%     { key .code:n = code }
+% \end{verbatim}
+% or to the key name:
+% \begin{verbatim}
+%   \keys_define:nn { module }
+%     { subgroup / key .code:n = code }
+% \end{verbatim}
+% As illustrated, the best choice of token for sub-dividing keys in
+% this way is |/|. This is because of the method that is
+% used to represent keys internally. Both of the above code fragments
+% set the same key, which has full name \texttt{module/subgroup/key}.
+%
+% As will be illustrated in the next section, this subdivision is
+% particularly relevant to making multiple choices.
+%
+% \section{Multiple choice keys}
+% \label{sec:choice}
+%
+% Multiple choices are created by setting the \texttt{.choice:}
+% property:
+% \begin{verbatim}
+%   \keys_define:nn { module }
+%     { key .choice: }
+% \end{verbatim}
+% For keys which are set up as choices, the valid choices are generated
+% by creating sub-keys of the choice key. This can be carried out in
+% two ways.
+%
+% In many cases, choices execute similar code which is dependant only
+% on the name of the choice or the position of the choice in the
+% list of choices. Here, the keys can share the same code, and can
+% be rapidly created using the  \texttt{.choice_code:n} and
+% \texttt{.generate_choices:n} properties:
+% \begin{verbatim}
+%   \keys_define:nn { module }
+%     {
+%       key .choice_code:n =
+%         {
+%           You~gave~choice~'\int_use:N \l_keys_choice_tl',~
+%           which~is~in~position~
+%           \int_use:N \l_keys_choice_int \c_space_tl
+%           in~the~list.
+%         },
+%       key .generate_choices:n =
+%         { choice-a, choice-b, choice-c }
+%     }
+% \end{verbatim}
+% Following common computing practice, \cs{l_keys_choice_int} is
+% indexed from  $0$ (as an offset), so that the value of
+% \cs{l_keys_choice_int} for the first choice in a list will be
+% zero.
+%
+% \begin{variable}{\l_keys_choice_int, \l_keys_choice_tl}
+%   Inside the code block for a choice generated using
+%   \texttt{.generate_choice:}, the variables \cs{l_keys_choice_tl} and
+%   \cs{l_keys_choice_int} are available to indicate the name of the
+%   current choice, and its position in the comma list.  The position
+%   is indexed from $0$.
+% \end{variable}
+%
+% On the other hand, it is sometimes useful to create choices which
+% use entirely different code from one another. This can be achieved
+% by setting the \texttt{.choice:} property of a key, then manually
+% defining sub-keys.
+% \begin{verbatim}
+%   \keys_define:nn { module }
+%     {
+%       key .choice:n,
+%       key / choice-a .code:n = code-a,
+%       key / choice-b .code:n = code-b,
+%       key / choice-c .code:n = code-c,
+%     }
+%\end{verbatim}
+%
+% It is possible to mix the two methods, but manually-created choices
+% should \emph{not} use \cs{l_keys_choice_tl} or \cs{l_keys_choice_int}.
+% These variables do not have defined behaviour when used outside of
+% code created using \texttt{.generate_choices:n}
+% (\emph{i.e.}~anything might happen).
+%
+% \section{Setting keys}
+%
+% \begin{function}{\keys_set:nn, \keys_set:nV, \keys_set:nv, \keys_set:no}
+%   \begin{syntax}
+%     \cs{keys_set:nn} \Arg{module} \Arg{keyval list}
+%   \end{syntax}
+%   Parses the \meta{keyval list}, and sets those keys which are defined
+%   for \meta{module}. The behaviour on finding an unknown key can be set
+%   by defining a special \texttt{unknown} key: this will be illustrated
+%   later. In contrast to \cs{keys_define:nn}, this function does check
+%   category codes and ignore spaces, and is therefore suitable for user
+%   input.
+% \end{function}
+%
+% If a key is not known, \cs{keys_set:nn} will look for a special
+% \texttt{unknown} key for the same module. This mechanism can be
+% used to create new keys from user input.
+% \begin{verbatim}
+%   \keys_define:nn { module }
+%     {
+%       unknown .code:n =
+%         You~tried~to~set~key~'\l_keys_key_tl'~to~'#1'.
+%     }
+%\end{verbatim}
+%
+% \begin{variable}{\l_keys_key_tl}
+%   When processing an unknown key, the name of the key is available
+%   as \cs{l_keys_key_tl}. Note that this will have been processed
+%   using \cs{tl_to_str:n}. The value passed to the key (if any) is
+%   available as the macro parameter |#1|.
+% \end{variable}
+%
+% \begin{variable}{\l_keys_path_tl}
+%   When processing an unknown key, the path of the key used is available
+%   as \cs{l_keys_path_tl}. Note that this will have been processed
+%   using \cs{tl_to_str:n}.
+% \end{variable}
+%
+% \section{Utility functions for keys}
+%
+% \begin{function}[EXP,pTF]{\keys_if_exist:nn}
+%   \begin{syntax}
+%     \cs{keys_if_exist:nn} \meta{module} \meta{key}
+%     \cs{keys_if_exist:nn} \meta{module} \meta{key}
+%     ~~\Arg{true code} \Arg{false code}
+%   \end{syntax}
+%   Tests if the \meta{key} exists for \meta{module}, \emph{i.e.}~if any code
+%   has been defined for \meta{key}.
+%   The branching versions then leave either \meta{true code} or
+%   \meta{false code} in the input stream, as appropriate to the truth
+%   of the test and the variant of the function chosen. The logical
+%   truth of the test is left in the input stream by the predicate
+%   version.
+% \end{function}
+%
+% \begin{function}{\keys_show:nn}
+%   \begin{syntax}
+%     \cs{keys_show:nn} \Arg{module} \Arg{key}
+%   \end{syntax}
+%   Shows the function which is used to actually implement a
+%   \meta{key} for a \meta{module}.
+% \end{function}
+%
+% \end{documentation}
+%
+% \begin{implementation}
+%
+% \section{\pkg{l3keys} Implementation}
+%
+%    \begin{macrocode}
+%<*initex|package>
+%    \end{macrocode}
+%
+%    \begin{macrocode}
+%<*package>
+\ProvidesExplPackage
+  {\ExplFileName}{\ExplFileDate}{\ExplFileVersion}{\ExplFileDescription}
+\package_check_loaded_expl:
+%</package>
+%    \end{macrocode}
+%
+% \subsection{Constants and variables}
+%
+% \begin{variable}{\c_keys_code_root_tl, \c_keys_vars_root_tl}
+%   The prefixes for the code and variables of the keys themselves.
+%    \begin{macrocode}
+\tl_const:Nn \c_keys_code_root_tl { key~code~>~ }
+\tl_const:Nn \c_keys_vars_root_tl { key~var~>~ }
+%    \end{macrocode}
+% \end{variable}
+%
+% \begin{variable}{\c_keys_props_root_tl}
+%   The prefix for storing properties.
+%    \begin{macrocode}
+\tl_const:Nn \c_keys_props_root_tl { key~prop~>~ }
+%    \end{macrocode}
+% \end{variable}
+%
+% \begin{variable}{\c_keys_value_forbidden_tl, \c_keys_value_required_tl}
+%   Two marker token lists.
+%    \begin{macrocode}
+\tl_const:Nn \c_keys_value_forbidden_tl { forbidden }
+\tl_const:Nn \c_keys_value_required_tl  { required }
+%    \end{macrocode}
+% \end{variable}
+%
+% \begin{variable}{\l_keys_choice_int, \l_keys_choices_tl}
+%   Publicly accessible data on which choice is being used when several
+%   are generated as a set.
+%    \begin{macrocode}
+\int_new:N \l_keys_choice_int
+\tl_new:N \l_keys_choices_tl
+%    \end{macrocode}
+% \end{variable}
+%
+% \begin{variable}{\l_keys_key_tl}
+%   The name of a key itself: needed when setting keys.
+%    \begin{macrocode}
+\tl_new:N \l_keys_key_tl
+%    \end{macrocode}
+% \end{variable}
+%
+% \begin{variable}{\l_keys_module_tl}
+%   The module for an entire set of keys.
+%    \begin{macrocode}
+\tl_new:N \l_keys_module_tl
+%    \end{macrocode}
+% \end{variable}
+%
+% \begin{variable}{\l_keys_no_value_bool}
+%   A marker is needed internally to show if only a key or a key plus a
+%   value was seen: this is recorded here.
+%    \begin{macrocode}
+\bool_new:N \l_keys_no_value_bool
+%    \end{macrocode}
+% \end{variable}
+%
+% \begin{variable}{\l_keys_path_tl}
+%   The \enquote{path} of the current key is stored here: this is
+%   available to the programmer and so is public.
+%    \begin{macrocode}
+\tl_new:N \l_keys_path_tl
+%    \end{macrocode}
+% \end{variable}
+%
+% \begin{variable}{\l_keys_property_tl}
+%   The \enquote{property} begin set for a key at definition time is
+%   stored here.
+%    \begin{macrocode}
+\tl_new:N \l_keys_property_tl
+%    \end{macrocode}
+% \end{variable}
+%
+% \begin{variable}{\l_keys_value_tl}
+%   The value given for a key: may be empty if no value was given.
+%    \begin{macrocode}
+\tl_new:N \l_keys_value_tl
+%    \end{macrocode}
+% \end{variable}
+%
+% \subsection{The key defining mechanism}
+%
+% \begin{macro}{\keys_define:nn}
+% \begin{macro}[aux]{\keys_define_aux:nnn, \keys_define_aux:onn}
+%   The public function for definitions is just a wrapper for the lower
+%   level mechanism, more or less. The outer function is designed to
+%   keep a track of the current module, to allow safe nesting.
+%    \begin{macrocode}
+\cs_new_protected:Npn \keys_define:nn
+  { \keys_define_aux:onn \l_keys_module_tl }
+\cs_new_protected:Npn \keys_define_aux:nnn #1#2#3
+  {
+    \tl_set:Nx \l_keys_module_tl { \tl_to_str:n {#2} }
+    \keyval_parse:NNn \keys_define_elt:n \keys_define_elt:nn {#3}
+    \tl_set:Nn \l_keys_module_tl {#1}
+  }
+\cs_generate_variant:Nn \keys_define_aux:nnn { o }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}[int]{\keys_define_elt:n}
+% \begin{macro}[int]{\keys_define_elt:nn}
+% \begin{macro}[aux]{\keys_define_elt_aux:nn}
+%   The outer functions here record whether a value was given and then
+%   converge on a common internal mechanism. There is first a search for
+%   a property in the current key name, then a check to make sure it is
+%   known before the code hands off to the next step.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \keys_define_elt:n #1
+  {
+    \bool_set_true:N \l_keys_no_value_bool
+    \keys_define_elt_aux:nn {#1} { }
+  }
+\cs_new_protected:Npn \keys_define_elt:nn #1#2
+  {
+    \bool_set_false:N \l_keys_no_value_bool
+    \keys_define_elt_aux:nn {#1} {#2}
+  }
+\cs_new_protected:Npn \keys_define_elt_aux:nn #1#2 {
+  \keys_property_find:n {#1}
+  \cs_if_exist:cTF { \c_keys_props_root_tl \l_keys_property_tl }
+    { \keys_define_key:n {#2} }
+    {
+      \msg_kernel_error:nnxx { keys } { property-unknown }
+        { \l_keys_property_tl } { \l_keys_path_tl }
+    }
+}
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}[int]{\keys_property_find:n}
+% \begin{macro}[aux]{\keys_property_find_aux:w}
+%   Searching for a property means finding the last |.| in the input,
+%   and storing the text before and after it. Everything is turned into
+%   strings, so there is no problem using an \texttt{x}-type expansion.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \keys_property_find:n #1
+  {
+    \tl_set:Nx \l_keys_path_tl { \l_keys_module_tl / }
+    \tl_if_in:nnTF {#1} { . }
+      { \keys_property_find_aux:w #1 \q_stop }
+      { \msg_kernel_error:nnx { keys } { key-no-property } {#1} }
+  }
+\cs_new_protected_nopar:Npn \keys_property_find_aux:w #1 . #2 \q_stop
+  {
+    \tl_set:Nx \l_keys_path_tl { \l_keys_path_tl \tl_to_str:n {#1} }
+    \tl_if_in:nnTF {#2} { . }
+      {
+        \tl_set:Nx \l_keys_path_tl { \l_keys_path_tl . }
+        \keys_property_find_aux:w #2 \q_stop
+      }
+      { \tl_set:Nn \l_keys_property_tl { . #2 } }
+  }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}[int]{\keys_define_key:n}
+% \begin{macro}[aux]{\keys_define_key_aux:w}
+%   Two possible cases. If there is a value for the key, then just use
+%   the function. If not, then a check to make sure there is no need for
+%   a value with the property. If there should be one then complain,
+%   otherwise execute it. There is no need to check for a |:| as if it
+%   is missing the earlier tests will have failed.
+%    \begin{macrocode}
+\cs_new_protected:Npn \keys_define_key:n #1
+  {
+    \bool_if:NTF \l_keys_no_value_bool
+      {
+        \exp_after:wN \keys_define_key_aux:w
+          \l_keys_property_tl \q_stop
+          { \use:c { \c_keys_props_root_tl \l_keys_property_tl } }
+          {
+            \msg_kernel_error:nnxx { keys }
+              { property-requires-value } { \l_keys_property_tl }
+              { \l_keys_path_tl }
+            }
+      }
+      { \use:c { \c_keys_props_root_tl \l_keys_property_tl } {#1} }
+  }
+\cs_new_protected:Npn \keys_define_key_aux:w #1 : #2 \q_stop
+  { \tl_if_empty:nTF {#2} }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \subsection{Turning properties into actions}
+%
+% \begin{macro}[int]{\keys_bool_set:NN}
+%   Boolean keys are really just choices, but all done by hand. The
+%   second argument here is the scope: either empty or \texttt{g} for
+%   global.
+%    \begin{macrocode}
+\cs_new_nopar:Npn \keys_bool_set:NN #1#2
+  {
+    \cs_if_exist:NF #1 { \bool_new:N #1 }
+    \keys_choice_make:
+    \keys_cmd_set:nx { \l_keys_path_tl / true }
+      { \exp_not:c { bool_ #2 set_true:N } \exp_not:N #1 }
+    \keys_cmd_set:nx { \l_keys_path_tl / false }
+      { \exp_not:c { bool_ #2 set_false:N } \exp_not:N #1 }
+    \keys_cmd_set:nn { \l_keys_path_tl / unknown }
+      {
+        \msg_kernel_error:nnx { keys } { boolean-values-only }
+          { \l_keys_key_tl }
+      }
+    \keys_default_set:n { true }
+  }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}[int]{\keys_choice_make:}
+%   To make a choice from a key, two steps: set the code, and set the
+%   unknown key.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \keys_choice_make:
+  {
+    \keys_cmd_set:nn { \l_keys_path_tl }
+      { \keys_choice_find:n {##1} }
+    \keys_cmd_set:nn { \l_keys_path_tl / unknown }
+      {
+        \msg_kernel_error:nnxx { keys } { choice-unknown }
+          { \l_keys_path_tl } {##1}
+      }
+  }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}[int]{\keys_choices_generate:n}
+% \begin{macro}[aux]{\keys_choices_generate_aux:n}
+%   Creating multiple-choices means setting up the \enquote{indicator}
+%   code, then applying whatever the user wanted.
+%    \begin{macrocode}
+\cs_new_protected:Npn \keys_choices_generate:n #1
+  {
+    \cs_if_exist:cTF
+      { \c_keys_vars_root_tl \l_keys_path_tl .choice~code }
+      {
+        \keys_choice_make:
+        \int_zero:N \l_keys_choice_int
+        \clist_map_function:nN {#1} \keys_choices_generate_aux:n
+      }
+      {
+        \msg_kernel_error:nnx { keys }
+          { generate-choices-before-code } { \l_keys_path_tl }
+      }
+  }
+\cs_new_protected_nopar:Npn \keys_choices_generate_aux:n #1
+  {
+    \keys_cmd_set:nx { \l_keys_path_tl / #1 }
+      {
+        \exp_not:n { \tl_set:Nn \l_keys_choice_tl } {#1}
+        \exp_not:n { \int_set:Nn \l_keys_choice_int }
+          { \int_use:N \l_keys_choice_int }
+        \exp_not:v
+          { \c_keys_vars_root_tl \l_keys_path_tl .choice~code }
+      }
+    \int_incr:N \l_keys_choice_int
+  }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}[int]{\keys_choice_code_store:x}
+%   The code for making multiple choices is stored in a token list.
+%    \begin{macrocode}
+\cs_new_protected:Npn \keys_choice_code_store:x #1
+  {
+    \cs_if_exist:cF
+      { \c_keys_vars_root_tl \l_keys_path_tl .choice~code }
+      {
+        \tl_new:c
+          { \c_keys_vars_root_tl \l_keys_path_tl .choice~code }
+      }
+    \tl_set:cx { \c_keys_vars_root_tl \l_keys_path_tl .choice~code }
+      {#1}
+  }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}[int]{\keys_cmd_set:nn, \keys_cmd_set:nx}
+% \begin{macro}[aux]{\keys_cmd_set_aux:n}
+%   Creating a new command means tidying up the properties and then making
+%   the internal function which actually does the work.
+%    \begin{macrocode}
+\cs_new_protected:Npn \keys_cmd_set:nn #1#2
+  {
+    \keys_cmd_set_aux:n {#1}
+    \cs_set:cpn { \c_keys_code_root_tl #1 } ##1 {#2}
+  }
+\cs_new_protected:Npn \keys_cmd_set:nx #1#2
+  {
+    \keys_cmd_set_aux:n {#1}
+    \cs_set:cpx { \c_keys_code_root_tl #1 } ##1 {#2}
+  }
+\cs_new_protected_nopar:Npn \keys_cmd_set_aux:n #1
+  {
+    \tl_clear_new:c { \c_keys_vars_root_tl #1 .default }
+    \tl_set:cn { \c_keys_vars_root_tl #1 .default } { \q_no_value }
+    \tl_clear_new:c { \c_keys_vars_root_tl #1 .req }
+  }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}[int]{\keys_default_set:n, \keys_default_set:V}
+%   Setting a default value is easy.
+%    \begin{macrocode}
+\cs_new_protected:Npn \keys_default_set:n #1
+  { \tl_set:cn { \c_keys_vars_root_tl \l_keys_path_tl .default } {#1} }
+\cs_generate_variant:Nn \keys_default_set:n { V }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\keys_meta_make:n, \keys_meta_make:x}
+%   To create a meta-key, simply set up to pass data through.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \keys_meta_make:n #1
+  {
+    \exp_args:NNo \keys_cmd_set:nn \l_keys_path_tl
+      { \exp_after:wN \keys_set:nn  \exp_after:wN { \l_keys_module_tl } {#1} }
+  }
+\cs_new_protected_nopar:Npn \keys_meta_make:x #1
+  {
+    \keys_cmd_set:nx { \l_keys_path_tl }
+      { \exp_not:N \keys_set:nn { \l_keys_module_tl } {#1} }
+  }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}[int]{\keys_value_requirement:n}
+%   Values can be required or forbidden by having the appropriate marker
+%   set.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \keys_value_requirement:n #1
+  {
+    \tl_set_eq:cc
+      { \c_keys_vars_root_tl \l_keys_path_tl .req }
+      { c_keys_value_ #1 _tl }
+  }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}[int]{\keys_variable_set:NnNN, \keys_variable_set:cnNN}
+% \begin{macro}[int]{\keys_variable_set:NnN, \keys_variable_set:cnN}
+%   Setting a variable takes the type and scope separately so that
+%   it is easy to make a new variable if needed. The three-argument
+%   version is set up so that the use of |{ }| as an \texttt{N}-type
+%   variable is only done once!
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \keys_variable_set:NnNN #1#2#3#4
+  {
+    \cs_if_exist:NF #1 { \use:c { #2 _new:N } #1 }
+    \keys_cmd_set:nx { \l_keys_path_tl }
+      { \exp_not:c { #2 _ #3 set:N #4 } \exp_not:N #1 {##1} }
+  }
+\cs_new_protected_nopar:Npn \keys_variable_set:NnN #1#2#3
+  { \keys_variable_set:NnNN #1 {#2} { } #3 }
+\cs_generate_variant:Nn \keys_variable_set:NnNN { c }
+\cs_generate_variant:Nn \keys_variable_set:NnN  { c }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \subsection{Creating key properties}
+%
+% The key property functions are all wrappers for internal functions,
+% meaning that things stay readable and can also be altered later on.
+%
+% \begin{macro}{.bool_set:N}
+% \begin{macro}{.bool_gset:N}
+%   One function for this.
+%    \begin{macrocode}
+\cs_new_protected_nopar:cpn { \c_keys_props_root_tl .bool_set:N } #1
+  { \keys_bool_set:NN #1 { } }
+\cs_new_protected_nopar:cpn { \c_keys_props_root_tl .bool_gset:N } #1
+  { \keys_bool_set:NN #1 g }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{.choice:}
+%   Making a choice is handled internally, as it is also needed by
+% \texttt{.generate_choices:n}.
+%    \begin{macrocode}
+\cs_new_protected_nopar:cpn { \c_keys_props_root_tl .choice: }
+  { \keys_choice_make: }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{.code:n, .code:x}
+%   Creating code is simply a case of passing through to the underlying
+%   \texttt{set} function.
+%    \begin{macrocode}
+\cs_new_protected:cpn { \c_keys_props_root_tl .code:n } #1
+  { \keys_cmd_set:nn { \l_keys_path_tl } {#1} }
+\cs_new_protected:cpn { \c_keys_props_root_tl .code:x } #1
+  { \keys_cmd_set:nx { \l_keys_path_tl } {#1} }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{.choice_code:n, .choice_code:x}
+%   Storing the code for choices, using \cs{exp_not:n} to avoid needing
+%   two internal functions.
+%    \begin{macrocode}
+\cs_new_protected:cpn { \c_keys_props_root_tl .choice_code:n } #1
+  { \keys_choice_code_store:x { \exp_not:n {#1} } }
+\cs_new_protected:cpn { \c_keys_props_root_tl .choice_code:x } #1
+  { \keys_choice_code_store:x {#1} }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{.default:n, .default:V}
+%   Expansion is left to the internal functions.
+%    \begin{macrocode}
+\cs_new_protected:cpn { \c_keys_props_root_tl .default:n } #1
+  { \keys_default_set:n {#1} }
+\cs_new_protected:cpn { \c_keys_props_root_tl .default:V } #1
+  { \keys_default_set:V #1 }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{.dim_set:N, .dim_set:c}
+% \begin{macro}{.dim_gset:N, .dim_gset:c}
+% Setting a variable is very easy: just pass the data along.
+%    \begin{macrocode}
+\cs_new_protected_nopar:cpn { \c_keys_props_root_tl .dim_set:N } #1
+  { \keys_variable_set:NnN #1 { dim } n }
+\cs_new_protected_nopar:cpn { \c_keys_props_root_tl .dim_set:c } #1
+  { \keys_variable_set:cnN {#1} { dim } n }
+\cs_new_protected_nopar:cpn { \c_keys_props_root_tl .dim_gset:N } #1
+  { \keys_variable_set:NnNN #1 { dim } g n }
+\cs_new_protected_nopar:cpn { \c_keys_props_root_tl .dim_gset:c } #1
+  { \keys_variable_set:cnNN {#1} { dim } g n }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{.fp_set:N, .fp_set:c}
+% \begin{macro}{.fp_gset:N, .fp_gset:c}
+%   Setting a variable is very easy: just pass the data along.
+%    \begin{macrocode}
+\cs_new_protected_nopar:cpn { \c_keys_props_root_tl .fp_set:N } #1
+  { \keys_variable_set:NnN #1 { fp } n }
+\cs_new_protected_nopar:cpn { \c_keys_props_root_tl .fp_set:c } #1
+  { \keys_variable_set:cnN {#1} { fp } n }
+\cs_new_protected_nopar:cpn { \c_keys_props_root_tl .fp_gset:N } #1
+  { \keys_variable_set:NnNN #1 { fp } g n }
+\cs_new_protected_nopar:cpn { \c_keys_props_root_tl .fp_gset:c } #1
+  { \keys_variable_set:cnNN {#1} { fp } g n }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{.generate_choices:n}
+%   Making choices is easy.
+%    \begin{macrocode}
+\cs_new_protected:cpn { \c_keys_props_root_tl .generate_choices:n } #1
+  { \keys_choices_generate:n {#1} }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{.int_set:N, .int_set:c}
+% \begin{macro}{.int_gset:N, .int_gset:c}
+%   Setting a variable is very easy: just pass the data along.
+%    \begin{macrocode}
+\cs_new_protected_nopar:cpn { \c_keys_props_root_tl .int_set:N } #1
+  { \keys_variable_set:NnN #1 { int } n }
+\cs_new_protected_nopar:cpn { \c_keys_props_root_tl .int_set:c } #1
+  { \keys_variable_set:cnN {#1} { int } n }
+\cs_new_protected_nopar:cpn { \c_keys_props_root_tl .int_gset:N } #1
+  { \keys_variable_set:NnNN #1 { int } g n }
+\cs_new_protected_nopar:cpn { \c_keys_props_root_tl .int_gset:c } #1
+  { \keys_variable_set:cnNN {#1} { int } g n }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+%\begin{macro}{.meta:n, .meta:x}
+% Making a meta is handled internally.
+%    \begin{macrocode}
+\cs_new_protected:cpn { \c_keys_props_root_tl .meta:n } #1
+  { \keys_meta_make:n {#1} }
+\cs_new_protected:cpn { \c_keys_props_root_tl .meta:x } #1
+  { \keys_meta_make:x {#1} }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{.skip_set:N, .skip_set:c}
+% \begin{macro}{.skip_gset:N, .skip_gset:c}
+%   Setting a variable is very easy: just pass the data along.
+%    \begin{macrocode}
+\cs_new_protected_nopar:cpn { \c_keys_props_root_tl .skip_set:N } #1
+  { \keys_variable_set:NnN #1 { skip } n }
+\cs_new_protected_nopar:cpn { \c_keys_props_root_tl .skip_set:c } #1
+  { \keys_variable_set:cnN {#1} { skip } n }
+\cs_new_protected_nopar:cpn { \c_keys_props_root_tl .skip_gset:N } #1
+  { \keys_variable_set:NnNN #1 { skip } g n }
+\cs_new_protected_nopar:cpn { \c_keys_props_root_tl .skip_gset:c } #1
+  { \keys_variable_set:cnNN {#1} { skip } g n }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+%
+% \begin{macro}{.tl_set:N, .tl_set:c}
+% \begin{macro}{.tl_gset:N, .tl_gset:c}
+% \begin{macro}{.tl_set_x:N, .tl_set_x:c}
+% \begin{macro}{.tl_gset_x:N, .tl_gset_x:c}
+%   Setting a variable is very easy: just pass the data along.
+%    \begin{macrocode}
+\cs_new_protected_nopar:cpn { \c_keys_props_root_tl .tl_set:N } #1
+   { \keys_variable_set:NnN #1 { tl } n }
+\cs_new_protected_nopar:cpn { \c_keys_props_root_tl .tl_set:c } #1
+  { \keys_variable_set:cnN {#1} { tl } n }
+\cs_new_protected_nopar:cpn { \c_keys_props_root_tl .tl_set_x:N } #1
+  { \keys_variable_set:NnN #1 { tl } x }
+\cs_new_protected_nopar:cpn { \c_keys_props_root_tl .tl_set_x:c } #1
+  { \keys_variable_set:cnN {#1} { tl } x }
+\cs_new_protected_nopar:cpn { \c_keys_props_root_tl .tl_gset:N } #1
+  { \keys_variable_set:NnNN #1 { tl } g n }
+\cs_new_protected_nopar:cpn { \c_keys_props_root_tl .tl_gset:c } #1
+  { \keys_variable_set:cnNN {#1} { tl } g n }
+\cs_new_protected_nopar:cpn { \c_keys_props_root_tl .tl_gset_x:N } #1
+  { \keys_variable_set:NnNN #1 { tl } g x }
+\cs_new_protected_nopar:cpn { \c_keys_props_root_tl .tl_gset_x:c } #1
+  { \keys_variable_set:cnNN {#1} { tl } g x }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{.value_forbidden:}
+% \begin{macro}{.value_required:}
+%   These are very similar, so both call the same function.
+%    \begin{macrocode}
+\cs_new_protected_nopar:cpn { \c_keys_props_root_tl .value_forbidden: }
+  { \keys_value_requirement:n { forbidden } }
+\cs_new_protected_nopar:cpn { \c_keys_props_root_tl .value_required: }
+  { \keys_value_requirement:n { required } }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \subsection{Setting keys}
+%
+% \begin{macro}{\keys_set:nn, \keys_set:nV, \keys_set:nv, \keys_set:no}
+% \begin{macro}[aux]{\keys_set_aux:nnn, \keys_set_aux:onn}
+%   A simple wrapper again.
+%    \begin{macrocode}
+\cs_new_protected:Npn \keys_set:nn
+  { \keys_set_aux:onn { \l_keys_module_tl } }
+\cs_new_protected:Npn \keys_set_aux:nnn #1#2#3
+  {
+    \tl_set:Nn \l_keys_module_tl {#2}
+    \keyval_parse:NNn \keys_set_elt:n \keys_set_elt:nn {#3}
+    \tl_set:Nn \l_keys_module_tl {#1}
+  }
+\cs_generate_variant:Nn \keys_set:nn { nV , nv , no }
+\cs_generate_variant:Nn \keys_set_aux:nnn { o }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}[int]{\keys_set_elt:n, \keys_set_elt:nn}
+% \begin{macro}[aux]{\keys_set_elt_aux:nn}
+%   A shared system once again. First, set the current path and add a
+%   default if needed. There are then checks to see if the a value is
+%   required or forbidden. If everything passes, move on to execute the
+%   code.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \keys_set_elt:n #1
+  {
+    \bool_set_true:N \l_keys_no_value_bool
+    \keys_set_elt_aux:nn {#1} { }
+  }
+\cs_new_protected:Npn \keys_set_elt:nn #1#2
+  {
+    \bool_set_false:N \l_keys_no_value_bool
+    \keys_set_elt_aux:nn {#1} {#2}
+  }
+\cs_new_protected:Npn \keys_set_elt_aux:nn #1#2
+  {
+    \tl_set:Nx \l_keys_key_tl { \tl_to_str:n {#1} }
+    \tl_set:Nx \l_keys_path_tl { \l_keys_module_tl / \l_keys_key_tl }
+    \keys_value_or_default:n {#2}
+    \bool_if:nTF
+      {
+        \keys_if_value_p:n { required } &&
+        \l_keys_no_value_bool
+      }
+      {
+        \msg_kernel_error:nnx { keys } { value-required }
+          { \l_keys_path_tl }
+      }
+      {
+        \bool_if:nTF
+          {
+              \keys_if_value_p:n { forbidden } &&
+            ! \l_keys_no_value_bool
+          }
+          {
+            \msg_kernel_error:nnxx { keys } { value-forbidden }
+              { \l_keys_path_tl } { \l_keys_value_tl }
+          }
+          { \keys_execute: }
+      }
+  }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}[int]{\keys_value_or_default:n}
+%   If a value is given, return it as |#1|, otherwise send a default if
+%   available.
+%    \begin{macrocode}
+\cs_new_protected:Npn \keys_value_or_default:n #1
+  {
+    \tl_set:Nn \l_keys_value_tl {#1}
+    \bool_if:NT \l_keys_no_value_bool
+      {
+        \quark_if_no_value:cF { \c_keys_vars_root_tl \l_keys_path_tl .default }
+          {
+            \tl_set_eq:Nc \l_keys_value_tl
+              { \c_keys_vars_root_tl \l_keys_path_tl .default }
+          }
+      }
+  }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}[int]{\keys_if_value_p:n}
+%   To test if a value is required or forbidden. A simple check for
+%   the existence of the appropriate marker.
+%    \begin{macrocode}
+\prg_new_conditional:Npnn \keys_if_value:n #1 { p }
+  {
+    \tl_if_eq:ccTF { c_keys_value_ #1 _tl }
+      { \c_keys_vars_root_tl \l_keys_path_tl .req }
+      { \prg_return_true: }
+      { \prg_return_false: }
+  }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}[int]{\keys_execute:}
+% \begin{macro}[aux]{\keys_execute_unknown:}
+% \begin{macro}[aux]{\keys_execute:nn}
+%   Actually executing a key is done in two parts. First, look for the
+%   key itself, then look for the \texttt{unknown} key with the same
+%   path. If both of these fail, complain.
+%    \begin{macrocode}
+\cs_new_nopar:Npn \keys_execute:
+  { \keys_execute:nn { \l_keys_path_tl } { \keys_execute_unknown: } }
+\cs_new_nopar:Npn \keys_execute_unknown:
+  {
+    \keys_execute:nn { \l_keys_module_tl / unknown }
+      {
+        \msg_kernel_error:nnxx { keys } { key-unknown }
+          { \l_keys_path_tl } { \l_keys_module_tl }
+      }
+  }
+\cs_new_nopar:Npn \keys_execute:nn #1#2
+  {
+    \cs_if_exist:cTF { \c_keys_code_root_tl #1 }
+      {
+        \exp_args:Nno \use:c { \c_keys_code_root_tl #1 }
+          \l_keys_value_tl
+      }
+      {#2}
+  }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\keys_choice_find:n}
+%   Executing a choice has two parts. First, try the choice given, then
+%   if that fails call the unknown key. That will exist, as it is created
+%   when a choice is first made. So there is no need for any escape code.
+%    \begin{macrocode}
+\cs_new_nopar:Npn \keys_choice_find:n #1
+  {
+    \keys_execute:nn { \l_keys_path_tl / \tl_to_str:n {#1} }
+      { \keys_execute:nn { \l_keys_path_tl / unknown } { } }
+  }
+%    \end{macrocode}
+% \end{macro}
+%
+% \subsection{Utilities}
+%
+% \begin{macro}[pTF]{\keys_if_exist:nn}
+% A utility for others to see if a key exists.
+%    \begin{macrocode}
+\prg_new_conditional:Npnn \keys_if_exist:nn #1#2 { p , T , F , TF }
+  {
+    \cs_if_exist:cTF { \c_keys_code_root_tl #1 / #2 }
+      { \prg_return_true:}
+      { \prg_return_false:}
+  }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\keys_show:nn}
+%   Showing a key is just a question of using the correct name.
+%    \begin{macrocode}
+\cs_new_nopar:Npn \keys_show:nn #1#2
+  { \cs_show:c { \c_keys_code_root_tl #1 / \tl_to_str:n {#2} } }
+%    \end{macrocode}
+% \end{macro}
+%
+% \subsection{Messages}
+%
+% For when there is a need to complain.
+%    \begin{macrocode}
+\msg_kernel_new:nnnn { keys } { boolean-values-only }
+  { Key~'#1'~accepts~boolean~values~only. }
+  { The~key~'#1'~only~accepts~the~values~'true'~and~'false'. }
+\msg_kernel_new:nnnn { keys } { choice-unknown }
+  { Choice~'#2'~unknown~for~key~'#1'. }
+  {
+    The~key~'#1'~takes~a~limited~number~of~values.\\
+    The~input~given,~'#2',~is~not~on~the~list~accepted.
+  }
+\msg_kernel_new:nnnn { keys } { generate-choices-before-code }
+  { No~code~available~to~generate~choices~for~key~'#1'. }
+  {
+    \c_msg_coding_error_text_tl
+    Before~using~.generate_choices:n~the~code~should~be~defined~
+    with~'.choice_code:n'~or~'.choice_code:x'.
+  }
+\msg_kernel_new:nnnn { keys } { key-no-property }
+  { No~property~given~in~definition~of~key~'#1'. }
+  {
+    \c_msg_coding_error_text_tl
+    Inside~\keys_define:nn each~key~name
+    needs~a~property:  \\
+    ~ ~ #1 .<property> \\
+    LaTeX~did~not~find~a~'.'~to~indicate~the~start~of~a~property.
+  }
+\msg_kernel_new:nnnn { keys } { key-unknown }
+  { The~key~'#1'~is~unknown~and~is~being~ignored. }
+  {
+    The~module~'#2'~does~not~have~a~key~called~#1'.\\
+    Check~that~you~have~spelled~the~key~name~correctly.
+  }
+\msg_kernel_new:nnnn { keys } { option-unknown }
+  { Unknown~option~'#1'~for~package~#2. }
+  {
+    LaTeX~has~been~asked~to~set~an~option~called~'#1'~
+    but~the~#2~package~has~not~created~an~option~with~this~name.
+  }
+\msg_kernel_new:nnnn { keys } { property-requires-value }
+  { The~property~'#1'~requires~a~value. }
+  {
+    \c_msg_coding_error_text_tl
+    LaTeX~was~asked~to~set~property~'#2'~for~key~'#1'.\\
+    No~value~was~given~for~the~property,~and~one~is~required.
+  }
+\msg_kernel_new:nnnn { keys } { property-unknown }
+  { The~key~property~'#1'~is~unknown. }
+  {
+    \c_msg_coding_error_text_tl
+    LaTeX~has~been~asked~to~set~the~property~'#1'~for~key~'#2':~
+    this~property~is~not~defined.
+  }
+\msg_kernel_new:nnnn { keys } { value-forbidden }
+  { The~key~'#1'~does~not~taken~a~value. }
+  {
+    The~key~'#1'~should~be~given~without~a~value.\\
+    LaTeX~will~ignore~the~given~value~'#2'.
+  }
+\msg_kernel_new:nnnn { keys } { value-required }
+  { The~key~'#1'~requires~a~value. }
+  {
+    The~key~'#1'~must~have~a~value.\\
+    No~value~was~present:~the~key~will~be~ignored.
+  }
+%    \end{macrocode}
+%
+%    \begin{macrocode}
+%</initex|package>
+%    \end{macrocode}
+%
+%\end{implementation}
+%
+%\PrintIndex
\ No newline at end of file
diff --git a/Master/texmf-dist/source/latex/l3kernel/l3keyval.dtx b/Master/texmf-dist/source/latex/l3kernel/l3keyval.dtx
new file mode 100644
index 00000000000..999ae7e5b08
--- /dev/null
+++ b/Master/texmf-dist/source/latex/l3kernel/l3keyval.dtx
@@ -0,0 +1,405 @@
+% \iffalse meta-comment
+%
+%% File: l3keyval.dtx Copyright (C) 2006-2011 The LaTeX3 Project
+%%
+%% It may be distributed and/or modified under the conditions of the
+%% LaTeX Project Public License (LPPL), either version 1.3c of this
+%% license or (at your option) any later version.  The latest version
+%% of this license is in the file
+%%
+%%    http://www.latex-project.org/lppl.txt
+%%
+%% This file is part of the "expl3 bundle" (The Work in LPPL)
+%% and all files in that bundle must be distributed together.
+%%
+%% The released version of this bundle is available from CTAN.
+%%
+%% -----------------------------------------------------------------------
+%%
+%% The development version of the bundle can be found at
+%%
+%%    http://www.latex-project.org/svnroot/experimental/trunk/
+%%
+%% for those people who are interested.
+%%
+%%%%%%%%%%%
+%% NOTE: %%
+%%%%%%%%%%%
+%%
+%%   Snapshots taken from the repository represent work in progress and may
+%%   not work or may contain conflicting material!  We therefore ask
+%%   people _not_ to put them into distributions, archives, etc. without
+%%   prior consultation with the LaTeX3 Project.
+%%
+%% -----------------------------------------------------------------------
+%
+%<*driver|package>
+\RequirePackage{l3names}
+\GetIdInfo$Id: l3keyval.dtx 2478 2011-06-19 21:34:23Z joseph $
+  {L3 Experimental key-value parsing}
+%</driver|package>
+%<*driver>
+\documentclass[full]{l3doc}
+\begin{document}
+  \DocInput{\jobname.dtx}
+\end{document}
+%</driver>
+% \fi
+%
+% \title{^^A
+%   The \pkg{l3keyval} package\\ Key--value parsing^^A
+%   \thanks{This file describes v\ExplFileVersion,
+%      last revised \ExplFileDate.}^^A
+% }
+%
+% \author{^^A
+%  The \LaTeX3 Project\thanks
+%    {^^A
+%      E-mail:
+%        \href{mailto:latex-team@latex-project.org}
+%          {latex-team@latex-project.org}^^A
+%    }^^A
+% }
+%
+% \date{Released \ExplFileDate}
+%
+% \maketitle
+%
+% \begin{documentation}
+% 
+% A key--value list is input of the form
+% \begin{verbatim}
+%   KeyOne = ValueOne ,
+%   KeyTwo = ValueTwo ,
+%   KeyThree   
+% \end{verbatim} 
+% where each key--value pair is separated by a comma from the rest of
+% the list, and each key--value pair does not necessarily contain an
+% equals sign or a value! Processing this type of input correctly 
+% requires a number of careful steps, to correctly account for
+% braces, spaces and the category codes of separators.
+% 
+% This module provides the low-level machinery for processing arbitrary
+% key--value lists. The \pkg{l3keys} module provides a higher-level
+% interface for managing run-time settings using key--value input,
+% while other parts of \LaTeX3 also use key--value input based on
+% \pkg{l3keyval} (for example the \pkg{xtemplate} module).
+% 
+ %\section{Parsing key--value lists}
+%
+% The low-level parsing system converts a \meta{key--value list}
+% into \meta{keys} and associated \meta{values}. After the parsing phase
+% is completed, the resulting keys and values (or keys alone) are
+% available for further processing. This is not carried out by the
+% low-level parser, and so the parser requires the names of
+% two functions along with the key--value list. One function is
+% needed to process key--value pairs (\emph{i.e}~two arguments), 
+% and a second function if required for keys given without arguments
+% (\emph{i.e.}~a single argument).
+% 
+% The parser does not double |#| tokens or expand any input. The tokens
+% |=| and |,| are corrected so that the parser does not \enquote{miss}
+% any due to category code changes. Spaces are removed from the ends
+% of the keys and values. Values which are given in braces will have
+% exactly one set removed, thus
+% \begin{verbatim}
+%    key = {value here},
+% \end{verbatim}
+% and
+% \begin{verbatim}
+%   key = value here,
+% \end{verbatim}
+% are treated identically.
+% 
+ %\begin{function}{\keyval_parse:NNn}
+%   \begin{syntax}
+%     \cs{keyval_parse:NNn} \meta{function1} \meta{function2}
+%     ~~\Arg{key--value list}
+%   \end{syntax}
+%   Parses the \meta{key--value list} into a series of \meta{keys} and
+%   associated \meta{values}, or keys alone (if no \meta{value} was
+%   given).  \meta{function1} should take one argument, while
+%   \meta{function2} should absorb two arguments. After
+%   \cs{keyval_parse:NNn} has parsed the \meta{key--value list}, 
+%   \meta{function1} will be used to process keys given with no value
+%   and \meta{function2} will be used to process keys given with a
+%   value. The order of the \meta{keys} in the \meta{key--value list}
+%   will be preserved. Thus
+%   \begin{verbatim}
+%     \keyval_parse:NNn \function:n \function:nn
+%       { key1 = value1 , key2 = value2, key3 = , key4 }
+%   \end{verbatim}
+%   will be converted into an input stream
+%   \begin{verbatim}
+%     \function:nn { key1 } { value1 }
+%     \function:nn { key2 } { value2 }
+%     \function:nn { key3 } { }
+%     \function:n  { key4 }
+%   \end{verbatim}
+%   Note that there is a difference between an empty value (an equals
+%   sign followed by nothing) and a missing value (no equals sign at
+%   all). 
+% \end{function}
+%
+% \end{documentation}
+%
+% \begin{implementation}
+%
+% \section{\pkg{l3keyval} implementation}
+%
+%    \begin{macrocode}
+%<*initex|package>
+%    \end{macrocode}
+%    
+%    \begin{macrocode}
+%<*package>
+\ProvidesExplPackage
+  {\ExplFileName}{\ExplFileDate}{\ExplFileVersion}{\ExplFileDescription}
+\package_check_loaded_expl:
+%</package>
+%    \end{macrocode}
+%    
+% \begin{variable}{\g_keyval_level_int}
+%   For nesting purposes an integer is needed for the current level.
+%    \begin{macrocode}
+\int_new:N \g_keyval_level_int
+%    \end{macrocode}
+% \end{variable}
+%
+% \begin{variable}{\l_keyval_key_tl, \l_keyval_value_tl}
+%   The current key name and value.
+%    \begin{macrocode}
+\tl_new:N \l_keyval_key_tl
+\tl_new:N \l_keyval_value_tl
+%    \end{macrocode}
+% \end{variable}
+%
+% \begin{variable}{\l_keyval_sanitise_tl}
+% \begin{variable}{\l_keyval_parse_tl}
+%   Token list variables for dealing with awkward category codes in the
+%   input.
+%    \begin{macrocode}
+\tl_new:N \l_keyval_sanitise_tl
+\tl_new:N \l_keyval_parse_tl
+%    \end{macrocode}
+%\end{variable}
+%\end{variable}
+%
+% \begin{macro}{\keyval_parse:n}
+%   The parsing function first deals with the category codes for
+%   |=| and |,|, so that there are no odd events. The input is then
+%   handed off to the element by element system. 
+%    \begin{macrocode}
+\group_begin:
+  \char_set_catcode_active:n { `\= }
+  \char_set_catcode_active:n { `\, }
+  \char_set_lccode:nn { `\8 } { `\= }
+  \char_set_lccode:nn { `\9 } { `\, }
+\tl_to_lowercase:n
+  {
+    \group_end:
+    \cs_new_protected:Npn \keyval_parse:n #1
+      {
+        \group_begin:
+          \tl_clear:N \l_keyval_sanitise_tl
+          \tl_set:Nn \l_keyval_sanitise_tl {#1}
+          \tl_replace_all_in:Nnn \l_keyval_sanitise_tl { = } { 8 }
+          \tl_replace_all_in:Nnn \l_keyval_sanitise_tl { , } { 9 }
+          \tl_clear:N \l_keyval_parse_tl
+          \exp_after:wN \keyval_parse_elt:w \exp_after:wN
+            \q_no_value \l_keyval_sanitise_tl 9 \q_nil 9
+        \exp_after:wN \group_end:
+        \l_keyval_parse_tl
+      }
+  }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\keyval_parse_elt:w}
+%   Each item to be parsed will have \cs{q_no_value} added to the front.
+%   Hence the blank test here can always be used to find a totally
+%   empty argument. If this is the case, the system loops round. If there
+%   is something to parse, there is a check for the \cs{q_nil} marker
+%   and if not a hand-off.
+%    \begin{macrocode}
+\cs_new_protected:Npn \keyval_parse_elt:w #1 ,
+  {
+    \tl_if_blank:oTF { \use_none:n #1 }
+      { \keyval_parse_elt:w \q_no_value }
+      {
+        \quark_if_nil:oF { \use_ii:nn #1 }
+          {
+            \keyval_split_key_value:w #1 = = \q_stop
+            \keyval_parse_elt:w \q_no_value
+          }
+      }
+  }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\keyval_split_key_value:w}
+% \begin{macro}[aux]{\keyval_split_key_value_aux:wTF}
+%   The key and value are handled separately. First the key is grabbed and
+%   saved as \cs{l_keyval_key_tl}. Then a check is need to see if there is
+%   a value at all: if not then the key name is simply added to the output.
+%   If there is a value then there is a check to ensure that there was
+%   only one |=| in the input (remembering some extra ones are around at
+%   the moment to prevent errors). All being well, there is an
+%   hand-off to find the value: the \cs{q_nil} is there to prevent loss
+%   of braces.
+%    \begin{macrocode}
+\cs_new_protected:Npn \keyval_split_key_value:w #1 = #2 \q_stop
+  {
+    \keyval_split_key:w #1 \q_stop
+    \str_if_eq:nnTF {#2} { = }
+      {
+        \tl_put_right:Nx \l_keyval_parse_tl
+          {
+            \exp_not:c { keyval_key_no_value_elt_ \int_use:N \g_keyval_level_int :n }
+              { \exp_not:o \l_keyval_key_tl }
+          }
+      }
+      {
+        \keyval_split_key_value_aux:wTF #2 \q_no_value \q_stop
+          { \keyval_split_value:w \q_nil #2 }
+          { \msg_kernel_error:nn { keyval } { misplaced-equals-sign } }
+      }
+  }
+\cs_new:Npn \keyval_split_key_value_aux:wTF #1 = #2#3 \q_stop
+  { \tl_if_head_eq_meaning:nNTF {#3} \q_no_value }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\keyval_split_key:w}
+% \begin{macro}{\keyval_remove_spaces:w}
+% \begin{macro}[aux]{\keyval_split_key_aux:w}
+% \begin{macro}[aux]{\keyval_remove_spaces_aux:w}
+%   The aim here is to remove spaces and also exactly one set of braces.
+%   The spaces are trimmed off from each end using a \enquote{funny}
+%   |Q|, which will never turn up in normal use. The idea is that
+%   the \texttt{f}-type expansion will stop if it finds an unexpandable
+%   token or a space, and will gobble the space. To avoid expanding
+%   anything else, the \cs{exp_not:N} works by ensuring that the first
+%   non-space token in the setting will stop the \texttt{f}-type
+%   expansion. The \cs{use_none:n} is needed to remove the leading
+%   quark, while the second setting of \cs{l_keyval_key_tl}
+%   removes exactly one set of braces.
+%    \begin{macrocode}
+\group_begin:
+  \char_set_catcode_math_toggle:n { `\Q }
+  \cs_new_protected:Npn \keyval_split_key:w #1 \q_stop
+    {
+      \exp_args:NNf \tl_set:Nn \l_keyval_key_tl
+        {
+          \exp_after:wN \keyval_remove_spaces:w \exp_after:wN
+            \exp_not:N \use_none:n #1 Q ~ Q
+        }
+      \tl_set:Nx \l_keyval_key_tl
+        { \exp_after:wN \keyval_split_key_aux:w \l_keyval_key_tl \q_stop }
+    }
+  \cs_gset:Npn \keyval_split_key_aux:w #1 \q_stop { \exp_not:n {#1} }
+  \cs_gset:Npn \keyval_remove_spaces:w #1 ~ Q { \keyval_remove_spaces_aux:w #1 Q }
+  \cs_gset:Npn \keyval_remove_spaces_aux:w #1 Q #2 {#1}
+\group_end:
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\keyval_split_value:w}
+%   Here the value has to be separated from the equals signs and the
+%   leading \cs{q_nil} added in to keep the brace levels. Fist the
+%   processing function can be added to the output list. If there is no
+%   value, setting \cs{l_keyval_value_tl} with three groups removed will
+%   leave nothing at all, and so an empty group can be added to the
+%   parsed list. On the other hand, if the value is entirely contained
+%   within a set of braces then \cs{l_keyval_value_tl} will contain
+%   \cs{q_nil} only. In that case, strip off the leading quark using
+%   \cs{use_ii:nnn}, which also deals with any spaces.
+%    \begin{macrocode}
+\cs_new_protected:Npn \keyval_split_value:w #1 = =
+  {
+    \tl_put_right:Nx \l_keyval_parse_tl
+      {
+        \exp_not:c { keyval_key_value_elt_ \int_use:N \g_keyval_level_int :nn }
+          { \exp_not:o \l_keyval_key_tl }
+      }
+    \tl_set:Nx \l_keyval_value_tl { \exp_not:o { \use_none:nnn #1 \q_nil \q_nil } }
+    \tl_if_empty:NTF \l_keyval_value_tl
+      { \tl_put_right:Nn \l_keyval_parse_tl { { } } }
+      {
+        \quark_if_nil:NTF \l_keyval_value_tl
+          {
+            \tl_put_right:Nx \l_keyval_parse_tl
+              { { \exp_not:o { \use_ii:nnn #1 \q_nil } } }
+          }
+          { \keyval_split_value_aux:w #1 \q_stop }
+      }
+  }
+%    \end{macrocode}
+%   A similar idea to the key code: remove the spaces from each end and
+%   deal with one set of braces.
+%    \begin{macrocode}
+\group_begin:
+  \char_set_catcode_math_toggle:n { `\Q }
+  \cs_new_protected:Npn \keyval_split_value_aux:w \q_nil #1 \q_stop
+    {
+      \exp_args:NNf \tl_set:Nn \l_keyval_value_tl
+        { \keyval_remove_spaces:w \exp_not:N #1 Q ~ Q }
+      \tl_put_right:Nx \l_keyval_parse_tl { { \exp_not:o \l_keyval_value_tl } }
+    }
+\group_end:
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\keyval_parse:NNn}
+%   The outer parsing routine just sets up the processing functions and
+%   hands off.
+%    \begin{macrocode}
+\cs_new_protected:Npn \keyval_parse:NNn #1#2#3
+  {
+    \int_gincr:N \g_keyval_level_int
+    \cs_gset_eq:cN { keyval_key_no_value_elt_ \int_use:N \g_keyval_level_int :n } #1
+    \cs_gset_eq:cN { keyval_key_value_elt_ \int_use:N \g_keyval_level_int :nn }   #2
+    \keyval_parse:n {#3}
+    \int_gdecr:N \g_keyval_level_int
+  }
+%    \end{macrocode}
+% \end{macro}
+%
+% One message for the low level parsing system.
+%    \begin{macrocode}
+\msg_kernel_new:nnnn { keyval } { misplaced-equals-sign }
+  { Misplaced~equals~sign~in~key-value~input~\msg_line_number: }
+  {
+    LaTeX~is~attempting~to~parse~some~key-value~input~but~found~
+    two~equals~signs~not~separated~by~a~comma.
+  }
+%    \end{macrocode}
+%    
+% \subsection{Deprecated functions}
+% 
+% Deprecated on 2011-05-27, for removal by 2011-08-31.
+%
+% \begin{macro}{\KV_process_space_removal_sanitize:NNn}
+% \begin{macro}{\KV_process_space_removal_no_sanitize:NNn}
+% \begin{macro}{\KV_process_no_space_removal_no_sanitize:NNn}
+% There is just one function for this now.
+%    \begin{macrocode}
+\cs_new_eq:NN \KV_process_space_removal_sanitize:NNn       \keyval_parse:NNn
+\cs_new_eq:NN \KV_process_space_removal_no_sanitize:NNn    \keyval_parse:NNn
+\cs_new_eq:NN \KV_process_no_space_removal_no_sanitize:NNn \keyval_parse:NNn
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+%    \begin{macrocode}
+%</initex|package>
+%    \end{macrocode}
+% 
+% \end{implementation}
+% 
+% \PrintIndex
diff --git a/Master/texmf-dist/source/latex/l3kernel/l3luatex.dtx b/Master/texmf-dist/source/latex/l3kernel/l3luatex.dtx
new file mode 100644
index 00000000000..70259184340
--- /dev/null
+++ b/Master/texmf-dist/source/latex/l3kernel/l3luatex.dtx
@@ -0,0 +1,459 @@
+% \iffalse meta-comment
+%
+%% File: l3luatex.dtx Copyright (C) 2010,2011 The LaTeX3 Project
+%%
+%% It may be distributed and/or modified under the conditions of the
+%% LaTeX Project Public License (LPPL), either version 1.3c of this
+%% license or (at your option) any later version.  The latest version
+%% of this license is in the file
+%%
+%%    http://www.latex-project.org/lppl.txt
+%%
+%% This file is part of the "expl3 bundle" (The Work in LPPL)
+%% and all files in that bundle must be distributed together.
+%%
+%% The released version of this bundle is available from CTAN.
+%%
+%% -----------------------------------------------------------------------
+%%
+%% The development version of the bundle can be found at
+%%
+%%    http://www.latex-project.org/svnroot/experimental/trunk/
+%%
+%% for those people who are interested.
+%%
+%%%%%%%%%%%
+%% NOTE: %%
+%%%%%%%%%%%
+%%
+%%   Snapshots taken from the repository represent work in progress and may
+%%   not work or may contain conflicting material!  We therefore ask
+%%   people _not_ to put them into distributions, archives, etc. without
+%%   prior consultation with the LaTeX3 Project.
+%%
+%% -----------------------------------------------------------------------
+%
+%<*driver|package>
+\RequirePackage{l3names}
+\GetIdInfo$Id: l3luatex.dtx 2478 2011-06-19 21:34:23Z joseph $
+  {L3 Experimental LuaTeX-specific functions}
+%</driver|package>
+%<*driver>
+\documentclass[full]{l3doc}
+\begin{document}
+  \DocInput{\jobname.dtx}
+\end{document}
+%</driver>
+% \fi
+%
+% \title{^^A
+%   The \pkg{l3luatex} package\\ LuaTeX-specific functions^^A
+%   \thanks{This file describes v\ExplFileVersion,
+%      last revised \ExplFileDate.}^^A
+% }
+%
+% \author{^^A
+%  The \LaTeX3 Project\thanks
+%    {^^A
+%      E-mail:
+%        \href{mailto:latex-team@latex-project.org}
+%          {latex-team@latex-project.org}^^A
+%    }^^A
+% }
+%
+% \date{Released \ExplFileDate}
+%
+% \maketitle
+%
+% \begin{documentation}
+%
+% \section{Breaking out to \Lua{}}
+%
+% The \LuaTeX{} engine provides access to the \Lua{} programming language, 
+% and with it access to the \enquote{internals} of \TeX{}. In order to use 
+% this within the framework provided here, a family of functions is
+% available. When used with \pdfTeX{} or \XeTeX{} these will raise an
+% error: use \cs{luatex_if_engine:T} to avoid this. Details of coding
+% the \LuaTeX{} engine are detailed in the \LuaTeX{} manual.
+% 
+% \begin{function}[EXP]{\lua_now:n, \lua_now:x}
+%   \begin{syntax}
+%     \cs{lua_now:n} \Arg{token list}
+%   \end{syntax}
+%   The \meta{token list} is first tokenized by \TeX{}, which will include
+%   converting line ends to spaces in the usual \TeX{} manner and which
+%   respects currently-applicable \TeX\ category codes. The resulting 
+%   \meta{\Lua{} input} is passed to the \Lua{} interpreter for processing.
+%   Each \cs{lua_now:n} block is treated by \Lua{} as a separate chunk.
+%   The \Lua{} interpreter will execute the \meta{\Lua{} input} immediately, 
+%   and in an expandable manner.
+%   \begin{texnote}
+%     \cs{lua_now:x} is the \LuaTeX{} primitive \cs{directlua} renamed.
+%   \end{texnote}
+% \end{function}
+%
+% \begin{function}{\lua_shipout:n, \lua_shipout:x}
+%   \begin{syntax}
+%     \cs{lua_shipout:x} \Arg{token list}
+%   \end{syntax}
+%   The \meta{token list} is first tokenized by \TeX{}, which will include
+%   converting line ends to spaces in the usual \TeX{} manner and which
+%   respects currently-applicable \TeX{} category codes.  The resulting 
+%   \meta{\Lua{} input} is passed to the \Lua\ interpreter when the 
+%   current page is finalised (\emph{i.e.}~at shipout).  Each
+%   \cs{lua_shipout:n} block is treated by \Lua{} as a separate chunk.
+%   The \Lua{} interpreter will execute the \meta{\Lua{} input} during the
+%   page-building routine: no \TeX{} expansion of the \meta{\Lua{} input}
+%   will occur at this stage.
+%   \begin{texnote}
+%     At a \TeX{} level, the \meta{\Lua{} input} is stored as a 
+%     \enquote{whatsit}.
+%   \end{texnote}
+% \end{function}
+%
+% \begin{function}{\lua_shipout_x:n, \lua_shipout_x:x}
+%   \begin{syntax}
+%     \cs{lua_shipout:n} \Arg{token list}
+%   \end{syntax}
+%   The \meta{token list} is first tokenized by \TeX{}, which will include
+%   converting line ends to spaces in the usual \TeX{} manner and which
+%   respects currently-applicable \TeX{} category codes.  The resulting 
+%   \meta{\Lua{} input} is passed to the \Lua{} interpreter when the 
+%   current page is finalised (\emph{i.e.}~at shipout).  Each
+%   \cs{lua_shipout:n} block is treated by \Lua{} as a separate chunk.
+%   The \Lua{} interpreter will execute the \meta{\Lua{} input} during the
+%   page-building routine: the \meta{\Lua{} input} is expanded during this
+%   process in addition to any expansion when the argument was read. This
+%   makes these functions suitable for including material finalised
+%   during the page building process (such as the page number).
+%   \begin{texnote}
+%     \cs{lua_shipout_x:n} is the \LuaTeX{} primitive \cs{latelua}
+%     named using the \LaTeX3 scheme.
+%  
+%     At a \TeX{} level, the \meta{\Lua{} input} is stored as a 
+%     \enquote{whatsit}.
+%   \end{texnote}
+% \end{function}
+%
+% \section{Category code tables}
+%
+% As well as providing methods to break out into \Lua{}, there are
+% places where additional \LaTeX3 functions are provided by the 
+% \LuaTeX{} engine. In particular, \LuaTeX{} provides category code 
+% tables. These can be used to ensure that a set of category codes
+% are in force in a more robust way than is possible with other 
+% engines. These are therefore used by \cs{ExplSyntaxOn} and
+% \pkg{ExplSyntaxOff} when using the \LuaTeX{} engine.
+% 
+% \begin{function}{\cctab_new:N}
+%   \begin{syntax}
+%     \cs{cctab_new:N} \meta{category code table}
+%   \end{syntax}
+%   Creates a new category code table, initially with the codes as
+%   used by \IniTeX{}.
+% \end{function}
+%
+% \begin{function}{\cctab_gset:Nn}
+%   \begin{syntax}
+%     \cs{cctab_gset:Nn} \meta{category code table}
+%       \Arg{category code set up}
+%   \end{syntax}
+%   Sets the \meta{category code table} to apply the category codes
+%   which apply when the prevailing regime is modified by the
+%   \meta{category code set up}. Thus within a standard code block
+%   the starting point will be the code applied by \cs{c_code_cctab}.
+%   The assignment of the table is global: the underlying primitive does
+%   not respect grouping.
+% \end{function}
+% 
+% \begin{function}{\cctab_begin:N}
+%   \begin{syntax}
+%     \cs{cctab_begin:N} \meta{category code table}
+%   \end{syntax}
+%   Switches the category codes in force to those stored in the
+%   \meta{category code table}.  The prevailing codes before the 
+%   function is called are added to a stack, for use with
+%   \cs{cctab_end:}.
+% \end{function}
+%
+% \begin{function}{\cctab_end:}
+%   \begin{syntax}
+%     \cs{cctab_end:}
+%   \end{syntax}
+%   Ends the scope of a \meta{category code table} started using
+%   \cs{cctab_begin:N}, retuning the codes to those in force before the
+%   matching \cs{cctab_begin:N} was used.
+% \end{function}
+%
+% \begin{variable}{\c_code_cctab}
+%   Category code table for the code environment. This does not include
+%   setting the behaviour of the line-end character, which is only
+%   altered by \cs{ExplSyntaxOn}.
+% \end{variable}
+%
+% \begin{variable}{\c_document_cctab}
+%   Category code table for a standard \LaTeX{} document. This does not 
+%   include setting the behaviour of the line-end character, which is 
+%   only altered by \cs{ExplSyntaxOff}.
+% \end{variable}
+%
+% \begin{variable}{\c_initex_cctab}
+%   Category code table as set up by \IniTeX{}.
+% \end{variable}
+%
+% \begin{variable}{\c_other_cctab}
+%   Category code table where all characters have category code $12$
+%   (other).
+% \end{variable}
+%
+% \begin{variable}{\c_string_cctab}
+%   Category code table where all characters have category code $12$
+%   (other) with the exception of spaces, which have category code
+%   $10$ (space).
+% \end{variable}
+%
+% \end{documentation}
+%
+% \begin{implementation}
+%
+% \section{\pkg{l3luatex} implementation}
+% 
+%    \begin{macrocode}
+%<*initex|package>
+%    \end{macrocode}
+% 
+%    Announce and ensure that the required packages are loaded.
+%    \begin{macrocode}
+%<*package>
+\ProvidesExplPackage
+  {\ExplFileName}{\ExplFileDate}{\ExplFileVersion}{\ExplFileDescription}
+\package_check_loaded_expl:
+%</package>
+%<*initex|package>
+%    \end{macrocode}
+%    
+% \begin{macro}{\lua_now:n, \lua_now:x}
+% \begin{macro}{\lua_shipout_x:n, \lua_shipout_x:x}
+% \begin{macro}{\lua_shipout:n, \lua_shipout:x}
+% \begin{macro}[aux]{\lua_wrong_engine:}
+%   When \LuaTeX{} is in use, this is all a question of primitives with new
+%   names. On the other hand, for \pdfTeX{} and \XeTeX{} the argument should
+%   be removed from the input stream before issuing an error. This needs 
+%   to be expandable, so the same idea is used as for \texttt{V}-type
+%   expansion, with an appropriately-named but undefined function.
+%    \begin{macrocode}
+\luatex_if_engine:TF
+  {
+    \cs_new_eq:NN \lua_now:x       \luatex_directlua:D
+    \cs_new_eq:NN \lua_shipout_x:n \luatex_latelua:D
+  }
+  {
+    \cs_new:Npn \lua_now:x #1 { \lua_wrong_engine: }
+    \cs_new_protected:Npn \lua_shipout_x:n #1 { \lua_wrong_engine: }
+  }
+\cs_new:Npn \lua_now:n #1
+  { \lua_now:x { \exp_not:n {#1} } }
+\cs_generate_variant:Nn \lua_shipout_x:n { x }
+\cs_new_protected:Npn \lua_shipout:n #1
+  { \lua_shipout_x:n { \exp_not:n {#1} } }
+\cs_generate_variant:Nn \lua_shipout:n { x }
+\group_begin:
+\char_set_catcode_letter:N\!
+\char_set_catcode_letter:N\ %
+\cs_gset:Npn\lua_wrong_engine:{%
+\LuaTeX engine not in use!%
+}%
+\group_end:%
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \subsection{Category code tables}
+%
+% \begin{variable}{\g_cctab_allocate_int}
+% \begin{variable}{\g_cctab_stack_int}
+% \begin{variable}{\g_cctab_stack_seq}
+%   To allocate category code tables, both the read-only and stack 
+%   tables need to be followed. There is also a sequence stack for the
+%   dynamic tables themselves.
+%    \begin{macrocode}
+\int_new:N  \g_cctab_allocate_int
+\int_set:Nn \g_cctab_allocate_int { -1 }
+\int_new:N \g_cctab_stack_int
+\seq_new:N \g_cctab_stack_seq
+%    \end{macrocode}
+% \end{variable}
+% \end{variable} 
+% \end{variable}  
+%
+% \begin{macro}{\cctab_new:N}
+%   Creating a new category code table is done slightly differently
+%   from other registers. Low-numbered tables are more efficiently-stored
+%   than high-numbered ones. There is also a need to have a stack of 
+%   flexible tables as well as the set of read-only ones. To satisfy both
+%   of these requirements, odd numbered tables are used for read-only
+%   tables, and even ones for the stack. Here, therefore, the odd numbers
+%   are allocated.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \cctab_new:N #1
+  {
+    \cs_if_free:NTF #1
+      { 
+        \int_gadd:Nn \g_cctab_allocate_int { 2 }
+         \int_compare:nNnTF 
+           { \g_cctab_allocate_int } <  { \c_max_register_int + 1 }
+           { 
+             \pref_global:D \tex_mathchardef:D #1 \g_cctab_allocate_int 
+             \luatex_initcatcodetable:D #1
+           }
+           { \msg_kernel_fatal:nnx { alloc } { out-of-registers } { cctab } }  
+       }
+       {
+         \msg_kernel_error:nnx { code } { variable-already-defined } 
+           { \token_to_str:N #1 } 
+       }
+  } 
+\luatex_if_engine:F
+  { \cs_set_protected_nopar:Npn \cctab_new:N #1 { \lua_wrong_engine: } }
+%<*package>
+\luatex_if_engine:T
+  {
+    \cs_set_protected_nopar:Npn \cctab_new:N #1 
+      { 
+        \newcatcodetable #1
+        \luatex_initcatcodetable:D #1
+      }
+  }
+%</package>
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\cctab_begin:N}
+% \begin{macro}{\cctab_end:}
+% \begin{variable}{\l_cctab_tmp_tl}
+%   The aim here is to ensure that the saved tables are read-only. This is
+%   done by using a stack of tables which are not read only, and actually
+%   having them as \enquote{in use} copies.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \cctab_begin:N #1
+  {
+    \seq_gpush:Nx \g_cctab_stack_seq { \tex_the:D \luatex_catcodetable:D }
+    \luatex_catcodetable:D #1
+    \int_gadd:Nn \g_cctab_stack_int { 2 }
+    \int_compare:nNnT { \g_cctab_stack_int } > { 268 435 453 }
+      { \msg_kernel_error:nn { code } { cctab-stack-full } }
+    \luatex_savecatcodetable:D \g_cctab_stack_int 
+    \luatex_catcodetable:D \g_cctab_stack_int 
+  }
+\cs_new_protected_nopar:Npn \cctab_end:
+  {
+    \int_gsub:Nn \g_cctab_stack_int { 2 }
+    \seq_gpop:NN \g_cctab_stack_seq \l_cctab_tmp_tl
+    \quark_if_no_value:NT \l_cctab_tmp_tl
+      { \tl_set:Nn \l_cctab_tmp_tl { 0 } }
+    \luatex_catcodetable:D \l_cctab_tmp_tl \scan_stop:
+  }
+\luatex_if_engine:F
+  {
+    \cs_set_protected_nopar:Npn \cctab_begin:N #1 { \lua_wrong_engine: }
+    \cs_set_protected_nopar:Npn \cctab_end: { \lua_wrong_engine: }
+  }
+%<*package>
+\luatex_if_engine:T
+  {
+    \cs_set_protected_nopar:Npn \cctab_begin:N #1 { \BeginCatcodeRegime #1 }
+    \cs_set_protected_nopar:Npn \cctab_end: { \EndCatcodeRegime }
+  }
+%</package>
+\tl_new:N \l_cctab_tmp_tl
+%    \end{macrocode}
+% \end{variable}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\cctab_gset:Nn}
+%   Category code tables are always global, so only one version is needed.
+%   The set up here is simple, and means that at the point of use there is
+%   no need to worry about escaping category codes.
+%    \begin{macrocode}
+\cs_new_protected:Npn \cctab_gset:Nn #1#2
+  {
+    \group_begin:
+      #2
+      \luatex_savecatcodetable:D #1
+    \group_end:
+  }
+\luatex_if_engine:F
+  { \cs_set_protected_nopar:Npn \cctab_gset:Nn #1#2 { \lua_wrong_engine: } }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{variable}{\c_code_cctab}
+% \begin{variable}{\c_document_cctab}
+% \begin{variable}{\c_initex_cctab}
+% \begin{variable}{\c_other_cctab}
+% \begin{variable}{\c_string_cctab}
+%   Creating category code tables is easy using the function above.
+%   The \texttt{other} and \texttt{string} ones are done by completely
+%   ignoring the existing codes as this makes life a lot less complex. The
+%   table for \pkg{expl3} category codes is always needed, whereas when in
+%   package mode the rest can be copied from the existing \LaTeXe{} package 
+%   \pkg{luatex}.
+%    \begin{macrocode}
+\luatex_if_engine:T
+  {
+    \cctab_new:N \c_code_cctab
+    \cctab_gset:Nn \c_code_cctab { }
+  }
+%<*package>  
+\luatex_if_engine:T
+  {
+    \cs_new_eq:NN \c_document_cctab \CatcodeTableLaTeX
+    \cs_new_eq:NN \c_initex_cctab   \CatcodeTableIniTeX
+    \cs_new_eq:NN \c_other_cctab    \CatcodeTableOther
+    \cs_new_eq:NN \c_string_cctab   \CatcodeTableString
+  }
+%</package>   
+%<*initex>    
+\luatex_if_engine:T
+  {
+    \cctab_new:N \c_document_cctab
+    \cctab_new:N \c_other_cctab
+    \cctab_new:N \c_string_cctab
+    \cctab_gset:Nn \c_document_cctab
+      {
+        \char_set_catcode_space:n          { 9 }
+        \char_set_catcode_space:n          { 32 }
+        \char_set_catcode_other:n          { 58 }
+        \char_set_catcode_math_subscript:n { 95 }
+        \char_set_catcode_active:n         { 126 }
+      }
+    \cctab_gset:Nn \c_other_cctab
+      {
+        \prg_stepwise_inline:nnnn { 0 } { 1 } { 127 }
+          { \char_set_catcode_other:n {#1} }
+      } 
+    \cctab_gset:Nn \c_string_cctab
+      {
+        \prg_stepwise_inline:nnnn { 0 } { 1 } { 127 }
+          { \char_set_catcode_other:n {#1} }
+        \char_set_catcode_space:n { 32 }  
+      }  
+  }
+%</initex>
+%    \end{macrocode}
+% \end{variable}
+% \end{variable}
+% \end{variable}
+% \end{variable}
+% \end{variable}
+% 
+%    \begin{macrocode}
+%</initex|package>
+%    \end{macrocode}
+%
+%\end{implementation}
+%
+%\PrintIndex
\ No newline at end of file
diff --git a/Master/texmf-dist/source/latex/l3kernel/l3msg.dtx b/Master/texmf-dist/source/latex/l3kernel/l3msg.dtx
new file mode 100644
index 00000000000..eceadf3c1a8
--- /dev/null
+++ b/Master/texmf-dist/source/latex/l3kernel/l3msg.dtx
@@ -0,0 +1,1609 @@
+% \iffalse meta-comment
+%
+%% File: l3msg.dtx Copyright (C) 2009-2011 The LaTeX3 Project
+%%
+%% It may be distributed and/or modified under the conditions of the
+%% LaTeX Project Public License (LPPL), either version 1.3c of this
+%% license or (at your option) any later version.  The latest version
+%% of this license is in the file
+%%
+%%    http://www.latex-project.org/lppl.txt
+%%
+%% This file is part of the "expl3 bundle" (The Work in LPPL)
+%% and all files in that bundle must be distributed together.
+%%
+%% The released version of this bundle is available from CTAN.
+%%
+%% -----------------------------------------------------------------------
+%%
+%% The development version of the bundle can be found at
+%%
+%%    http://www.latex-project.org/svnroot/experimental/trunk/
+%%
+%% for those people who are interested.
+%%
+%%%%%%%%%%%
+%% NOTE: %%
+%%%%%%%%%%%
+%%
+%%   Snapshots taken from the repository represent work in progress and may
+%%   not work or may contain conflicting material!  We therefore ask
+%%   people _not_ to put them into distributions, archives, etc. without
+%%   prior consultation with the LaTeX3 Project.
+%%
+%% -----------------------------------------------------------------------
+%
+%<*driver|package>
+\RequirePackage{l3names}
+\GetIdInfo$Id: l3msg.dtx 2478 2011-06-19 21:34:23Z joseph $
+  {L3 Experimental messages}
+%</driver|package>
+%<*driver>
+\documentclass[full]{l3doc}
+\begin{document}
+  \DocInput{\jobname.dtx}
+\end{document}
+%</driver>
+% \fi
+%
+% \title{^^A
+%   The \pkg{l3msg} package\\ Messages^^A
+%   \thanks{This file describes v\ExplFileVersion,
+%      last revised \ExplFileDate.}^^A
+% }
+%
+% \author{^^A
+%  The \LaTeX3 Project\thanks
+%    {^^A
+%      E-mail:
+%        \href{mailto:latex-team@latex-project.org}
+%          {latex-team@latex-project.org}^^A
+%    }^^A
+% }
+%
+% \date{Released \ExplFileDate}
+%
+% \maketitle
+%
+% \begin{documentation}
+%
+% Messages need to be passed to the user by modules, either when errors
+% occur or to indicate how the code is proceeding. The \pkg{l3msg}
+% module provides a consistent method for doing this (as opposed to
+% writing directly to the terminal or log).
+%
+% The system used by \pkg{l3msg} to create messages divides the process
+% into two distinct parts. Named messages are created in the first part
+% of the process; at this stage, no decision is made about the type
+% output that the message will produce. The second part of the process
+% is actually producing a message. At this stage a choice of message
+% \emph{class} has to be made, for example \texttt{error},
+% \texttt{warning} or \texttt{info}.
+%
+% By separating out the creation and use of messages, several benefits
+% are available. First, the messages can be altered later without
+% needing details of where they are used in the code. This makes it
+% possible to alter the language used, the detail level and so on.
+% Secondly, the output which results from a given message can be
+% altered. This can be done on a message class, module or message name
+% basis. In this way, message behaviour can be altered and messages can
+% be entirely suppressed.
+%
+% \section{Creating new messages}
+%
+% All messages have to be created before they can be used. All message
+% setting is local, with the general assumption that messages will
+% be managed as part of module set up outside of any \TeX{} grouping.
+%
+% The text of messages will automatically by wrapped to the length
+% available in the console. As a result, formatting is only needed
+% where it will help to show meaning. In particular, |\\| may be
+% used to force a new line and \verb*|\ | forces an explicit space.
+%
+% \begin{function}{\msg_new:nnnn, \msg_new:nnn}
+%   \begin{syntax}
+%     \cs{msg_new:nnnn} \Arg{module} \Arg{message} \Arg{text}
+%     ~~\Arg{more text}
+%   \end{syntax}
+%   Creates a \meta{message} for a given \meta{module}.
+%   The message will be defined to first give \meta{text} and then
+%   \meta{more text} if the user requests it. If no \meta{more text} is
+%   available then a standard text is given instead. Within \meta{text}
+%   and \meta{more text} four parameters (|#1| to |#4|) can be used:
+%   these will be supplied at the time the message is used. The
+%   parameters will be expanded when the message is used. Within the
+%   \meta{text} and \meta{more text} |\\| can be used to start a new
+%   line. An error will be raised if the \meta{message} already exists.
+% \end{function}
+%
+% \begin{function}{\msg_set:nnnn, \msg_set:nnn}
+%   \begin{syntax}
+%     \cs{msg_set:nnnn} \Arg{module} \Arg{message} \Arg{text}
+%     ~~\Arg{more text}
+%   \end{syntax}
+%   Sets up the text for a \meta{message} for a given \meta{module}.
+%   The message will be defined to first give \meta{text} and then
+%   \meta{more text} if the user requests it. If no \meta{more text} is
+%   available then a standard text is given instead. Within \meta{text}
+%   and \meta{more text} four parameters (|#1| to |#4|) can be used:
+%   these will be supplied at the time the message is used. The
+%   parameters will be expanded when the message is used. Within the
+%   \meta{text} and \meta{more text} |\\| can be used to start a new
+%   line.
+% \end{function}
+%
+% \section{Contextual information for messages}
+%
+% \begin{function}[EXP]{\msg_line_context:}
+%   \begin{syntax}
+%     \cs{msg_line_context:}
+%   \end{syntax}
+%   Prints the current line number when a message is given, and
+%   thus suitable for giving context to messages. The number itself
+%   is proceeded by the text |on line|.
+% \end{function}
+%
+% \begin{function}[EXP]{\msg_line_number:}
+%   \begin{syntax}
+%     \cs{msg_line_number:}
+%   \end{syntax}
+%   Prints the current line number when a message is given.
+% \end{function}
+%
+% \begin{variable}{\c_msg_return_text_tl}
+%  Standard text to indicate that the user should try pressing
+%  \meta{return} to continue. The standard definition reads:
+%  \begin{verbatim}
+%    Try typing <return> to proceed.
+%
+%    If that doesn't work, type X <return> to quit.
+%  \end{verbatim}
+% \end{variable}
+%
+% \begin{variable}{\c_msg_trouble_text_tl}
+%  Standard text to indicate that the more errors are likely and
+%  that aborting the run is advised. The standard definition reads:
+%  \begin{verbatim}
+%    More errors will almost certainly follow:
+%    the LaTeX run should be aborted.
+%  \end{verbatim}
+% \end{variable}
+%
+% \begin{function}[EXP]{\msg_fatal_text:n}
+%   \begin{syntax}
+%     \cs{msg_fatal_text:n} \Arg{module}
+%   \end{syntax}
+%   Produces the standard text:
+%   \begin{verbatim}
+%     Fatal <module> error
+%   \end{verbatim}
+%   This function can be redefined to alter the language in which the
+%   message is give, using |#1| as the name of the \meta{module} to
+%   be included.
+% \end{function}
+%
+% \begin{function}[EXP]{\msg_critical_text:n}
+%   \begin{syntax}
+%     \cs{msg_critical_text:n} \Arg{module}
+%   \end{syntax}
+%   Produces the standard text:
+%   \begin{verbatim}
+%     Critical <module> error
+%   \end{verbatim}
+%   This function can be redefined to alter the language in which the
+%   message is give, using |#1| as the name of the \meta{module} to
+%   be included.
+% \end{function}
+%
+% \begin{function}[EXP]{\msg_error_text:n}
+%   \begin{syntax}
+%     \cs{msg_error_text:n} \Arg{module}
+%   \end{syntax}
+%   Produces the standard text:
+%   \begin{verbatim}
+%     <module> error
+%   \end{verbatim}
+%   This function can be redefined to alter the language in which the
+%   message is give, using |#1| as the name of the \meta{module} to
+%   be included.
+% \end{function}
+%
+% \begin{function}[EXP]{\msg_warning_text:n}
+%   \begin{syntax}
+%     \cs{msg_warning_text:n} \Arg{module}
+%   \end{syntax}
+%   Produces the standard text:
+%   \begin{verbatim}
+%     <module> warning
+%   \end{verbatim}
+%   This function can be redefined to alter the language in which the
+%   message is give, using |#1| as the name of the \meta{module} to
+%   be included.
+% \end{function}
+%
+% \begin{function}[EXP]{\msg_info_text:n}
+%   \begin{syntax}
+%     \cs{msg_info_text:n} \Arg{module}
+%   \end{syntax}
+%   Produces the standard text:
+%   \begin{verbatim}
+%     <module> info
+%   \end{verbatim}
+%   This function can be redefined to alter the language in which the
+%   message is give, using |#1| as the name of the \meta{module} to
+%   be included.
+% \end{function}
+%
+% \section{Issuing messages}
+%
+% Messages behave differently depending on the message class. A number
+% of standard message classes are supplied, but more can be
+% created.
+% 
+% When issuing messages, any arguments passed should use
+% \cs{tl_to_str:n} or \cs{token_to_str:N} to prevent unwanted expansion
+% of the material.
+%
+% \begin{function}{\msg_class_set:nn}
+%   \begin{syntax}
+%     \cs{msg_class_set:nn} \Arg{class} \Arg{code}
+%   \end{syntax}
+%   Sets a \meta{class} to output a message, using \meta{code}
+%   to process the message text. The \meta{class} should be a text
+%   value, while the \meta{code} may be any arbitrary material.
+%   The \meta{code} will receive $6$ arguments: the module
+%   name (|#1|), the message name (|#2|) and the four arguments
+%   taken by the message text (|#3| to |#6|).
+% \end{function}
+%
+% The kernel defines several common message classes. The following
+% describes the standard behaviour of each class if no redirection of
+% the class or message is active. In all cases, the message may be
+% issued supplying $0$ to $4$ arguments. The code will
+% ensure that there an no errors if the number of arguments supplied
+% here does not match the number in the definition of the message
+% (although of course the sense of the message may be impaired).
+%
+% \begin{function}
+%   {
+%     \msg_fatal:nnxxxx ,
+%     \msg_fatal:nnxxx  ,
+%     \msg_fatal:nnxx   ,
+%     \msg_fatal:nnx    ,
+%     \msg_fatal:nn
+%   }
+%   \begin{syntax}
+%     \cs{msg_fatal:nnxxxx} \Arg{module} \Arg{message} \Arg{arg one}
+%       ~~\Arg{arg two} \Arg{arg three} \Arg{arg four}
+%   \end{syntax}
+%   Issues \meta{module} error \meta{message}, passing \meta{arg one} to
+%   \meta{arg four} to the text-creating functions. After issuing a
+%   fatal error the \TeX{} run will halt.
+% \end{function}
+%
+% \begin{function}
+%   {
+%    \msg_critical:nnxxxx ,
+%    \msg_critical:nnxxx  ,
+%    \msg_critical:nnxx   ,
+%    \msg_critical:nnx    ,
+%    \msg_critical:nn
+%  }
+%   \begin{syntax}
+%     \cs{msg_critical:nnxxxx} \Arg{module} \Arg{message} \Arg{arg one}
+%      ~~\Arg{arg two} \Arg{arg three} \Arg{arg four}
+%   \end{syntax}
+%   Issues \meta{module} error \meta{message}, passing \meta{arg one} to
+%   \meta{arg four} to the text-creating functions. After issuing the
+%   message reading the current input file will stop. This may halt
+%   the \TeX{} run (if the current file is the main file) or
+%   may abort reading a sub-file.
+% \end{function}
+%
+% \begin{function}
+%   {
+%     \msg_error:nnxxxx ,
+%     \msg_error:nnxxx  ,
+%     \msg_error:nnxx   ,
+%     \msg_error:nnx    ,
+%     \msg_error:nn
+%   }
+%   \begin{syntax}
+%     \cs{msg_error:nnxxxx} \Arg{module} \Arg{message} \Arg{arg one}
+%      ~~\Arg{arg two} \Arg{arg three} \Arg{arg four}
+%   \end{syntax}
+%   Issues \meta{module} error \meta{message}, passing \meta{arg one} to
+%   \meta{arg four} to the text-creating functions. The error will
+%   stop processing and issue the text at the terminal. After user input,
+%   the run will continue.
+% \end{function}
+%
+%  \begin{function}
+%    {
+%      \msg_warning:nnxxxx ,
+%      \msg_warning:nnxxx  ,
+%      \msg_warning:nnxx   ,
+%      \msg_warning:nnx    ,
+%      \msg_warning:nn
+%    }
+%    \begin{syntax}
+%      \cs{msg_warning:nnxxxx} \Arg{module} \Arg{message} \Arg{arg one}
+%       ~~\Arg{arg two} \Arg{arg three} \Arg{arg four}
+%    \end{syntax}
+%    Issues \meta{module} warning \meta{message}, passing \meta{arg one} to
+%    \meta{arg four} to the text-creating functions. The warning text
+%    will be added to the log file, but the \TeX{} run will not be
+%    interrupted.
+%  \end{function}
+%
+% \begin{function}
+%   {
+%     \msg_info:nnxxxx ,
+%     \msg_info:nnxxx  ,
+%     \msg_info:nnxx   ,
+%     \msg_info:nnx    ,
+%     \msg_info:nn
+%   }
+%   \begin{syntax}
+%     \cs{msg_info:nnxxxx} \Arg{module} \Arg{message} \Arg{arg one}
+%      ~~\Arg{arg two} \Arg{arg three} \Arg{arg four}
+%   \end{syntax}
+%   Issues \meta{module} information \meta{message}, passing
+%   \meta{arg one} to \meta{arg four} to the text-creating functions.
+%   The information text will be added to the log file.
+% \end{function}
+%
+% \begin{function}
+%   {
+%     \msg_log:nnxxxx ,
+%     \msg_log:nnxxx  ,
+%     \msg_log:nnxx   ,
+%     \msg_log:nnx    ,
+%     \msg_log:nn
+%   }
+%   \begin{syntax}
+%     \cs{msg_log:nnxxxx} \Arg{module} \Arg{message} \Arg{arg one}
+%      ~~\Arg{arg two} \Arg{arg three} \Arg{arg four}
+%   \end{syntax}
+%   Issues \meta{module} information \meta{message}, passing
+%   \meta{arg one} to \meta{arg four} to the text-creating functions.
+%   The information text will be added to the log file: the output
+%   is briefer than \cs{msg_info:nnxxxx}.
+% \end{function}
+%
+% \begin{function}
+%   {
+%     \msg_none:nnxxxx ,
+%     \msg_none:nnxxx  ,
+%     \msg_none:nnxx   ,
+%     \msg_none:nnx    ,
+%     \msg_none:nn
+%   }
+%   \begin{syntax}
+%     \cs{msg_none:nnxxxx} \Arg{module} \Arg{message} \Arg{arg one}
+%      ~~\Arg{arg two} \Arg{arg three} \Arg{arg four}
+%   \end{syntax}
+%   Does nothing: used as a message class to prevent any output at
+%   all (see the discussion of message redirection).
+% \end{function}
+%
+% \section{Redirecting messages}
+%
+% \begin{function}{\msg_redirect_class:nn}
+%   \begin{syntax}
+%     \cs{msg_redirect_class:nn} \Arg{class one} \Arg{class two}
+%   \end{syntax}
+%   Changes the behaviour of messages of \meta{class one} so that they
+%   are processed using the code for those of \meta{class two}. Multiple
+%   redirections are possible. Redirection to a missing class or
+%   infinite loops will raise errors when the messages are used,
+%   rather than at the point of redirection.
+% \end{function}
+%
+% \begin{function}{\msg_redirect_module:nnn}
+%   \begin{syntax}
+%     \cs{msg_redirect_module:nnn} \Arg{module} \Arg{class one}
+%     ~~\Arg{class two}
+%   \end{syntax}
+%   Redirects message of \meta{class one} for \meta{module} to act as
+%   though they were from \meta{class two}. Messages of \meta{class one}
+%   from sources other than \meta{module} are not affected by this
+%   redirection. This function can be used to make some messages
+%   \enquote{silent} by default. For example, all of the
+%   \texttt{trace} messages of \meta{module} could be turned off with:
+%   \begin{verbatim}
+%     \msg_redirect_module:nnn { module } { trace } { none }
+%   \end{verbatim}
+% \end{function}
+%
+% \begin{function}{\msg_redirect_name:nnn}
+%   \begin{syntax}
+%     \cs{msg_redirect_name:nn} \Arg{module} \Arg{message} \Arg{class}
+%   \end{syntax}
+%   Redirects a specific \meta{message} from a specific \meta{module}
+%   to act as a member of \meta{class} of messages. This function can
+%   be used to make a selected message \enquote{silent} without
+%   changing global parameters:
+%   \begin{verbatim}
+%     \msg_redirect_name:nnn { module } { annoying-message } { none }
+%   \end{verbatim}
+% \end{function}
+%
+% \section{Low-level message functions}
+%
+% The lower-level message functions should usually be accessed from the
+% higher-level system. However, there are occasions where direct
+% access to these functions is desirable.
+%
+% \begin{function}[EXP]{\msg_newline:, \msg_two_newlines:}
+%   \begin{syntax}
+%     \cs{msg_newline:}
+%   \end{syntax}
+%   Forces a new line in a message. This is a low-level function, which
+%   will not include any additional printing information in the message:
+%   contrast with |\\| in messages. The |two| version adds two lines.
+% \end{function}
+% 
+% \begin{function}{\msg_interrupt:xxx}
+%   \begin{syntax}
+%     \cs{msg_interrupt:xxx} \Arg{first line} \Arg{text} \Arg{extra text}
+%   \end{syntax}
+%   Interrupts the \TeX{} run, issuing a formatted message comprising
+%   \meta{first line} and \meta{text} laid out in the format
+%   \begin{verbatim}
+%     !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
+%     !
+%     ! <first line>
+%     ! 
+%     ! <text>
+%     !............................................... 
+%   \end{verbatim}
+%   where the \meta{text} will be wrapped to fit within the current
+%   line length. The user may then request more information, at which 
+%   stage the \meta{extra text} will be shown in the terminal in the
+%   format
+%   \begin{verbatim}
+%     |'''''''''''''''''''''''''''''''''''''''''''''''
+%     |  <extra text>
+%     |...............................................
+%   \end{verbatim}
+%   where the \meta{extra text} will be wrapped to fit within the current
+%   line length.
+% \end{function}
+%
+% \begin{function}{\msg_log:x}
+%   \begin{syntax}
+%     \cs{msg_log:x} \Arg{text}
+%   \end{syntax}
+%   Writes to the log file with the \meta{text} laid out in the format
+%   \begin{verbatim}
+%     .................................................
+%     . <text>
+%     .................................................
+%   \end{verbatim}
+%   where the \meta{text} will be wrapped to fit within the current
+%   line length. 
+% \end{function}
+%
+% \begin{function}{\msg_term:x}
+%   \begin{syntax}
+%     \cs{msg_term:x} \Arg{text}
+%   \end{syntax}
+%   Writes to the terminal and log file with the \meta{text} laid out in the
+%   format
+%   \begin{verbatim}
+%     *************************************************
+%     * <text>
+%     *************************************************
+%   \end{verbatim}
+%   where the \meta{text} will be wrapped to fit within the current
+%   line length. 
+% \end{function}
+%
+% \section{Kernel-specific functions}
+%
+% Messages from \LaTeX3 itself are handled by the general message system,
+% but have their own functions. This allows some text to be pre-defined,
+% and also ensures that serious errors can be handled properly.
+%
+% \begin{function}{\msg_kernel_new:nnnn, \msg_kernel_new:nnn}
+%   \begin{syntax}
+%     \cs{msg_kernel_new:nnnn} \Arg{module} \Arg{message} \Arg{text}
+%     ~~\Arg{more text}
+%   \end{syntax}
+%   Creates a kernel \meta{message} for a given \meta{module}.
+%   The message will be defined to first give \meta{text} and then
+%   \meta{more text} if the user requests it. If no \meta{more text} is
+%   available then a standard text is given instead. Within \meta{text}
+%   and \meta{more text} four parameters (|#1| to |#4|) can be used:
+%   these will be supplied at the time the message is used. The
+%   parameters will be expanded when the message is used. Within the
+%   \meta{text} and \meta{more text} |\\| can be used to start a new
+%   line. An error will be raised if the \meta{message} already exists.
+% \end{function}
+%
+% \begin{function}{\msg_kernel_set:nnnn, \msg_kernel_set:nnn}
+%   \begin{syntax}
+%     \cs{msg_kernel_set:nnnn} \Arg{module} \Arg{message} \Arg{text}
+%     ~~\Arg{more text}
+%   \end{syntax}
+%   Sets up the text for a kernel \meta{message} for a given \meta{module}.
+%   The message will be defined to first give \meta{text} and then
+%   \meta{more text} if the user requests it. If no \meta{more text} is
+%   available then a standard text is given instead. Within \meta{text}
+%   and \meta{more text} four parameters (|#1| to |#4|) can be used:
+%   these will be supplied at the time the message is used. The
+%   parameters will be expanded when the message is used. Within the
+%   \meta{text} and \meta{more text} |\\| can be used to start a new
+%   line.
+% \end{function}
+%
+% \begin{function}
+%   {
+%     \msg_kernel_fatal:nnxxxx ,
+%     \msg_kernel_fatal:nnxxx  ,
+%     \msg_kernel_fatal:nnxx   ,
+%     \msg_kernel_fatal:nnx    ,
+%     \msg_kernel_fatal:nn
+%   }
+%   \begin{syntax}
+%     \cs{msg_kernel_fatal:nnxxxx} \Arg{module} \Arg{message} \Arg{arg one}
+%       ~~\Arg{arg two} \Arg{arg three} \Arg{arg four}
+%   \end{syntax}
+%   Issues kernel \meta{module} error \meta{message}, passing \meta{arg one}
+%   to \meta{arg four} to the text-creating functions. After issuing a
+%   fatal error the \TeX{} run will halt. Cannot be redirected.
+% \end{function}
+%
+% \begin{function}
+%   {
+%     \msg_kernel_error:nnxxxx ,
+%     \msg_kernel_error:nnxxx  ,
+%     \msg_kernel_error:nnxx   ,
+%     \msg_kernel_error:nnx    ,
+%     \msg_kernel_error:nn
+%   }
+%   \begin{syntax}
+%     \cs{msg_kernel_error:nnxxxx} \Arg{module} \Arg{message} \Arg{arg one}
+%      ~~\Arg{arg two} \Arg{arg three} \Arg{arg four}
+%   \end{syntax}
+%   Issues kernel \meta{module} error \meta{message}, passing \meta{arg one}
+%   to
+%   \meta{arg four} to the text-creating functions. The error will
+%   stop processing and issue the text at the terminal. After user input,
+%   the run will continue. Cannot be redirected.
+% \end{function}
+%
+%  \begin{function}
+%    {
+%      \msg_kernel_warning:nnxxxx ,
+%      \msg_kernel_warning:nnxxx  ,
+%      \msg_kernel_warning:nnxx   ,
+%      \msg_kernel_warning:nnx    ,
+%      \msg_kernel_warning:nn
+%    }
+%    \begin{syntax}
+%      \cs{msg_kernel_warning:nnxxxx} \Arg{module} \Arg{message} \Arg{arg one}
+%       ~~\Arg{arg two} \Arg{arg three} \Arg{arg four}
+%    \end{syntax}
+%    Issues kernel \meta{module} warning \meta{message}, passing
+%    \meta{arg one} to
+%    \meta{arg four} to the text-creating functions. The warning text
+%    will be added to the log file, but the \TeX{} run will not be
+%    interrupted.
+%  \end{function}
+%
+% \begin{function}
+%   {
+%     \msg_kernel_info:nnxxxx ,
+%     \msg_kernel_info:nnxxx  ,
+%     \msg_kernel_info:nnxx   ,
+%     \msg_kernel_info:nnx    ,
+%     \msg_kernel_info:nn
+%   }
+%   \begin{syntax}
+%     \cs{msg_kernel_info:nnxxxx} \Arg{module} \Arg{message} \Arg{arg one}
+%      ~~\Arg{arg two} \Arg{arg three} \Arg{arg four}
+%   \end{syntax}
+%   Issues kernel \meta{module} information \meta{message}, passing
+%   \meta{arg one} to \meta{arg four} to the text-creating functions.
+%   The information text will be added to the log file.
+% \end{function}
+%
+% \end{documentation}
+%
+% \begin{implementation}
+%
+% \section{\pkg{l3msg} implementation}
+%
+%    \begin{macrocode}
+%<*initex|package>
+%    \end{macrocode}
+%
+%    \begin{macrocode}
+%<*package>
+\ProvidesExplPackage
+  {\ExplFileName}{\ExplFileDate}{\ExplFileVersion}{\ExplFileDescription}
+\package_check_loaded_expl:
+%</package>
+%    \end{macrocode}
+%    
+% \begin{variable}{\l_msg_tmp_tl}
+%   A general scratch for the module.
+%    \begin{macrocode}
+\tl_new:N \l_msg_tmp_tl
+%    \end{macrocode}
+% \end{variable}
+%
+%\section{Creating messages}
+%
+% Messages are created and used separately, so there two parts to
+% the code here. First, a mechanism for creating message text.
+% This is pretty simple, as there is not actually a lot to do.
+%
+%\begin{variable}{\c_msg_text_prefix_tl, \c_msg_more_text_prefix_tl}
+%   Locations for the text of messages.
+%    \begin{macrocode}
+\tl_const:Nn \c_msg_text_prefix_tl      { msg~text~>~ }
+\tl_const:Nn \c_msg_more_text_prefix_tl { msg~extra~text~>~ }
+%    \end{macrocode}
+% \end{variable}
+%
+% \begin{macro}{\msg_new:nnnn, \msg_new:nnn}
+% \begin{macro}{\msg_set:nnnn, \msg_set:nnn}
+%   Setting a message simply means saving the appropriate text
+%   into two functions. A sanity check first.
+%    \begin{macrocode}
+\cs_new_protected:Npn \msg_new:nnnn #1#2
+  {
+    \cs_if_exist:cT { \c_msg_text_prefix_tl #1 / #2 }
+      {
+        \msg_kernel_error:nn { msg } { message-already-defined }
+          {#1} {#2}
+      }
+    \msg_set:nnnn {#1} {#2}
+  }
+\cs_new_protected:Npn \msg_new:nnn #1#2#3
+  { \msg_new:nnnn {#1} {#2} {#3} { } }
+\cs_new_protected:Npn \msg_set:nnnn #1#2#3#4
+  {
+    \cs_set:cpn { \c_msg_text_prefix_tl #1 / #2 }
+      ##1##2##3##4 {#3}
+    \cs_set:cpn { \c_msg_more_text_prefix_tl #1 / #2 }
+      ##1##2##3##4 {#4}
+  }
+\cs_new_protected:Npn \msg_set:nnn #1#2#3
+  { \msg_set:nnnn {#1} {#2} {#3} { } }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \subsection{Messages: support functions and text}
+%
+% \begin{variable}
+%   {
+%     \c_msg_coding_error_text_tl ,
+%     \c_msg_continue_text_tl     ,
+%     \c_msg_critical_text_tl     ,
+%     \c_msg_fatal_text_tl        ,
+%     \c_msg_help_text_tl         ,
+%     \c_msg_no_info_text_tl      ,
+%     \c_msg_on_line_tl           ,
+%     \c_msg_return_text_tl       ,
+%     \c_msg_trouble_text_tl
+%   }
+% Simple pieces of text for messages.
+%    \begin{macrocode}
+\tl_const:Nn \c_msg_coding_error_text_tl
+  {
+    This~is~a~coding~error.
+    \\ \\
+  }
+\tl_const:Nn \c_msg_continue_text_tl
+  { Type~<return>~to~continue }
+\tl_const:Nn \c_msg_critical_text_tl
+  { Reading~the~current~file~will~stop }
+\tl_const:Nn \c_msg_fatal_text_tl
+  { This~is~a~fatal~error:~LaTeX~will~abort }
+\tl_const:Nn \c_msg_help_text_tl
+  { For~immediate~help~type~H~<return> }
+\tl_const:Nn \c_msg_no_info_text_tl
+  {
+    LaTeX~does~not~know~anything~more~about~this~error,~sorry.
+    \c_msg_return_text_tl
+  }
+\tl_const:Nn \c_msg_on_line_text_tl { on~line }
+\tl_const:Nn \c_msg_return_text_tl
+  {
+    \\ \\
+    Try~typing~<return>~to~proceed.
+    \\
+    If~that~doesn't~work,~type~X~<return>~to~quit.
+  }
+\tl_const:Nn \c_msg_trouble_text_tl
+  {
+    \\ \\
+    More~errors~will~almost~certainly~follow: \\
+    the~LaTeX~run~should~be~aborted.
+  }
+%    \end{macrocode}
+% \end{variable}
+%
+% \begin{macro}{\msg_newline:, \msg_two_newlines:}
+%   New lines are printed in the same way as for low-level file writing.
+%    \begin{macrocode}
+\cs_new_nopar:Npn \msg_newline: { ^^J }
+\cs_new_nopar:Npn \msg_two_newlines: { ^^J ^^J }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\msg_line_number:}
+% \begin{macro}{\msg_line_context:}
+%   For writing the line number nicely.
+%    \begin{macrocode}
+\cs_new_nopar:Npn \msg_line_number: { \int_use:N \tex_inputlineno:D }
+\cs_set_nopar:Npn \msg_line_context:
+  {
+    \c_msg_on_line_text_tl
+    \c_space_tl
+    \msg_line_number:
+  }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \subsection{Showing messages: low level mechanism}
+%
+% \begin{variable}{\c_msg_hide_tl}
+% \begin{variable}[aux]{\c_msg_hide_tl<dots>}
+%   An empty variable with a number of (category code 11) periods at the
+%   end of its name. This is used to push the \TeX{} part of an error
+%   message \enquote{off the screen}. Using two variables here means that
+%   later life is a little easier.
+%    \begin{macrocode}
+\char_set_catcode_letter:N \.
+\tl_new:N
+  \c_msg_hide_tl..........................................................
+\tl_const:Nn \c_msg_hide_tl
+ { \c_msg_hide_tl.......................................................... }
+\char_set_catcode_other:N \.
+%    \end{macrocode}
+% \end{variable}
+% \end{variable}
+%
+% \begin{macro}{\msg_interrupt:xxx}
+% \begin{macro}[aux]{\msg_interrupt_no_details:xx}
+% \begin{macro}[aux]{\msg_interrupt_details:xxx}
+% \begin{macro}[aux]{\msg_interrupt_text:n}
+% \begin{macro}[aux]{\msg_interrupt_more_text:n}
+% \begin{macro}[aux]{\msg_interrupt_aux:}
+%   The low-level interruption macro is rather opaque, unfortunately. The
+%   idea here is to create a a message which hides all of \TeX{}'s own
+%   information by filling the output up with dots. To achieve this,
+%   dots have to be letters. The odd
+%   \cs{c_msg_hide_tl<dots>} actually does the hiding: it is the
+%   large run of dots in the name that is important here. The meaning
+%   of |\\| is altered so that the explanation text is a simple run
+%   whilst the initial error has line-continuation shown.
+%    \begin{macrocode}
+\cs_new_protected:Npn \msg_interrupt:xxx #1#2#3
+  {
+    \group_begin:
+      \tl_if_empty:nTF {#3}
+        { \msg_interrupt_no_details:xx {#1} {#2} }
+        { \msg_interrupt_details:xxx {#1} {#2} {#3} }
+      \msg_interrupt_aux:
+    \group_end:
+  }
+%    \end{macrocode}
+%    \begin{macrocode}
+% Depending on the availability of more information there is a choice of
+% how to set up the further help. The extra help text has to be set
+% before the message itself can be issued. Everything is done using
+% \texttt{x}-type expansion as the new line markers are different for
+% the two type of text and need to be correctly set up.
+%    \begin{macrocode}
+\cs_new_protected:Npn \msg_interrupt_no_details:xx #1#2
+  {
+    \iow_wrap:xnnnN
+      { \\ \c_msg_no_info_text_tl }
+      { |~ } { 2 } { } \msg_interrupt_more_text:n
+    \iow_wrap:xnnnN { #1 \\ \\ #2 \\ \\ \c_msg_continue_text_tl }
+      { ! ~ } { 2 } {} \msg_interrupt_text:n
+  }
+\cs_new_protected:Npn \msg_interrupt_details:xxx #1#2#3
+  {
+    \iow_wrap:xnnnN
+      { \\ #3 }
+      { |~ } { 2 } { } \msg_interrupt_more_text:n
+    \iow_wrap:xnnnN { #1 \\ \\ #2 \\ \\ \c_msg_help_text_tl }
+      { ! ~ } { 2 } { } \msg_interrupt_text:n
+  }
+\cs_new_protected:Npn \msg_interrupt_text:n #1
+  { \tl_set:Nn \l_msg_text_tl {#1} }
+\cs_new_protected:Npn \msg_interrupt_more_text:n #1
+  {
+%<*initex>
+    \tl_set:Nx \l_msg_tmp_tl
+%</initex>
+%<*package>
+    \protected@edef \l_msg_tmp_tl
+%</package>
+    {
+      |'''''''''''''''''''''''''''''''''''''''''''''''
+      #1
+      \msg_newline:
+      |...............................................
+    }
+    \tex_errhelp:D \exp_after:wN { \l_msg_tmp_tl }
+  }
+%    \end{macrocode}
+%   The business end of the process starts by producing some visual
+%   separation of the message from the main part of the log. It then
+%   adds the hiding text to the message to print. The error message needs
+%   to be printed with everything made \enquote{invisible}: this is where
+%   the strange business with |&| comes in: this is made into another
+%   |!|. There is also a closing brace that will show up in the output,
+%   which is turned into a blank space.
+%    \begin{macrocode}
+\group_begin: % {
+  \char_set_lccode:w `\} = `\  \scan_stop:
+  \char_set_lccode:w `\& = `\! \scan_stop:
+  \char_set_catcode_active:N \&
+\tl_to_lowercase:n
+  {
+    \group_end:
+    \cs_new_protected:Npn \msg_interrupt_aux:
+      {
+        \iow_term:x
+          {
+            \iow_newline:
+            !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
+            \iow_newline:
+            !
+          }
+        \tl_put_right:No \l_msg_text_tl { \c_msg_hide_tl }
+        \cs_set_protected_nopar:Npx &
+          { \tex_errmessage:D { \exp_not:o { \l_msg_text_tl } } }
+        &
+      }
+  }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\msg_log:x}
+% \begin{macro}{\msg_term:x}
+%   Printing to the log or terminal without a stop is rather easier.
+%   A bit of simple visual work sets things off nicely.
+%    \begin{macrocode}
+\cs_new_protected:Npn \msg_log:x #1
+  {
+    \iow_log:x { ................................................. }
+    \iow_wrap:xnnnN { . ~ #1} { . ~ } { 2 } { }
+      \iow_log:x
+    \iow_log:x { ................................................. }
+  }
+\cs_new_protected:Npn \msg_term:x #1
+  {
+    \iow_term:x { ************************************************* }
+    \iow_wrap:xnnnN { * ~ #1} { * ~ } { 2 } { }
+      \iow_term:x
+    \iow_term:x { ************************************************* }
+  }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \subsection{Displaying messages}
+%
+% \LaTeX{} is handling error messages and so the \TeX{} ones are disabled.
+%    \begin{macrocode}
+\int_set:Nn \tex_errorcontextlines:D { -1 }
+%    \end{macrocode}
+%
+% \begin{macro}
+%   {
+%     \msg_fatal_text:n    ,
+%     \msg_critical_text:n ,
+%     \msg_error_text:n    ,
+%     \msg_warning_text:n  ,
+%     \msg_info_text:n
+%   }
+%   A function for issuing messages: both the text and order could
+%   in principal vary.
+%    \begin{macrocode}
+\cs_new_nopar:Npn \msg_fatal_text:n #1 { Fatal~#1~error }
+\cs_new_nopar:Npn \msg_critical_text:n #1 { Critical~#1~error }
+\cs_new_nopar:Npn \msg_error_text:n #1 { #1~error }
+\cs_new_nopar:Npn \msg_warning_text:n #1 { #1~warning }
+\cs_new_nopar:Npn \msg_info_text:n #1 { #1~info }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\msg_see_documentation_text:n}
+%   Contextual footer information.
+%    \begin{macrocode}
+\cs_new_nopar:Npn \msg_see_documentation_text:n #1
+  { \\ \\ See~the~#1~documentation~for~further~information. }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{variable}{\l_msg_redirect_classes_prop, \l_msg_redirect_names_prop}
+%   For filtering messages, a list of all messages and of those which have
+%   to be modified is required.
+%    \begin{macrocode}
+\prop_new:N \l_msg_redirect_classes_prop
+\prop_new:N \l_msg_redirect_names_prop
+%    \end{macrocode}
+% \end{variable}
+%
+% \begin{macro}{\msg_class_set:nn}
+%  Setting up a message class does two tasks. Any existing redirection
+%  is cleared, and the various message functions are created
+%  to simply use the code stored for the message.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \msg_class_set:nn #1#2
+  {
+    \prop_clear_new:c { l_msg_redirect_ #1 _prop }
+    \cs_set_protected:cpn { msg_ #1 :nnxxxx } ##1##2##3##4##5##6
+      { \msg_use:nnnnxxxx {#1} {#2} {##1} {##2} {##3} {##4} {##5} {##6} }
+    \cs_set_protected:cpx { msg_ #1 :nnxxx } ##1##2##3##4##5
+      { \exp_not:c { msg_ #1 :nnxxxx } {##1} {##2} {##3} {##4} {##5} { } }
+    \cs_set_protected:cpx { msg_ #1 :nnxx } ##1##2##3##4
+      { \exp_not:c { msg_ #1 :nnxxxx } {##1} {##2} {##3} {##4} { } { } }
+    \cs_set_protected:cpx { msg_ #1 :nnx } ##1##2##3
+      { \exp_not:c { msg_ #1 :nnxxxx } {##1} {##2} {##3} { } { } { } }
+    \cs_set_protected:cpx { msg_ #1 :nn } ##1##2
+      { \exp_not:c { msg_ #1 :nnxxxx } {##1} {##2} { } { } { } { } }
+  }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}[pTF]{\msg_if_more_text:N, \msg_if_more_text:c}
+% \begin{macro}[aux]{\msg_no_more_text:xxxx}
+%   A test to see if any more text is available, using a
+%   permanently-empty text function.
+%    \begin{macrocode}
+\prg_set_conditional:Npnn \msg_if_more_text:N #1 { p , T , F , TF }
+  {
+    \cs_if_eq:NNTF #1 \msg_no_more_text:xxxx
+      { \prg_return_false:  }
+      { \prg_return_true: }
+  }
+\cs_new:Npn \msg_no_more_text:xxxx #1#2#3#4 { }
+\cs_generate_variant:Nn \msg_if_more_text_p:N { c }
+\cs_generate_variant:Nn \msg_if_more_text:NT  { c }
+\cs_generate_variant:Nn \msg_if_more_text:NF  { c }
+\cs_generate_variant:Nn \msg_if_more_text:NTF { c }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+%\begin{macro}
+%  {
+%    \msg_fatal:nnxxxx ,
+%    \msg_fatal:nnxxx  ,
+%    \msg_fatal:nnxx   ,
+%    \msg_fatal:nnx    ,
+%    \msg_fatal:nn
+%  }
+%   For fatal errors, after the error message \TeX{} bails out.
+%    \begin{macrocode}
+\msg_class_set:nn { fatal }
+  {
+    \msg_interrupt:xxx
+      { \msg_fatal_text:n {#1} : ~ "#2" }
+      {
+        \use:c { \c_msg_text_prefix_tl #1 / #2 } {#3} {#4} {#5} {#6}
+        \msg_see_documentation_text:n {#1}
+      }
+      { \c_msg_fatal_text_tl }
+    \tex_end:D
+  }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}
+%   {
+%     \msg_critical:nnxxxx ,
+%     \msg_critical:nnxxx  ,
+%     \msg_critical:nnxx   ,
+%     \msg_critical:nnx    ,
+%     \msg_critical:nn
+%  }
+%   Not quite so bad: just end the current file.
+%    \begin{macrocode}
+\msg_class_set:nn { critical }
+  {
+    \msg_interrupt:xxx
+      { \msg_critical_text:n {#1} : ~ "#2" }
+      {
+        \use:c { \c_msg_text_prefix_tl #1 / #2 } {#3} {#4} {#5} {#6}
+        \msg_see_documentation_text:n {#1}
+      }
+      { \c_msg_critical_text_tl }
+    \tex_endinput:D
+  }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}
+%   {
+%     \msg_error:nnxxxx ,
+%     \msg_error:nnxxx  ,
+%     \msg_error:nnxx   ,
+%     \msg_error:nnx    ,
+%     \msg_error:nn
+%   }
+%   For an error, the interrupt routine is called, then any recovery code
+%   is tried.
+%    \begin{macrocode}
+\msg_class_set:nn { error }
+  {
+    \msg_if_more_text:cTF { \c_msg_more_text_prefix_tl #1 / #2 }
+      {
+        \msg_interrupt:xxx
+          { \msg_error_text:n {#1} : ~ "#2" }
+          {
+            \use:c { \c_msg_text_prefix_tl #1 / #2 } {#3} {#4} {#5} {#6}
+            \msg_see_documentation_text:n {#1}
+          }
+          { \use:c { \c_msg_more_text_prefix_tl #1 / #2 } {#3} {#4} {#5} {#6} }
+     }
+     {
+        \msg_interrupt:xxx
+          { \msg_error_text:n {#1} : ~ "#2" }
+          {
+            \use:c { \c_msg_text_prefix_tl #1 / #2 } {#3} {#4} {#5} {#6}
+            \msg_see_documentation_text:n {#1}
+          }
+          { }
+     }
+  }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}
+%   {
+%     \msg_warning:nnxxxx ,
+%     \msg_warning:nnxxx  ,
+%     \msg_warning:nnxx   ,
+%     \msg_warning:nnx    ,
+%     \msg_warning:nn
+%   }
+%   Warnings are printed to the terminal.
+%    \begin{macrocode}
+\msg_class_set:nn { warning }
+  {
+    \msg_term:x
+      {
+        \msg_warning_text:n {#1} : ~ "#2" \\ \\
+        \use:c { \c_msg_text_prefix_tl #1 / #2 } {#3} {#4} {#5} {#6}
+      }
+  }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}
+%   {
+%     \msg_info:nnxxxx ,
+%     \msg_info:nnxxx  ,
+%     \msg_info:nnxx   ,
+%     \msg_info:nnx    ,
+%     \msg_info:nn
+%   }
+%   Information only goes into the log.
+%    \begin{macrocode}
+\msg_class_set:nn { info }
+  {
+    \msg_log:x
+      {
+        \msg_info_text:n {#1} : ~ "#2" \\ \\
+        \use:c { \c_msg_text_prefix_tl #1 / #2 } {#3} {#4} {#5} {#6}
+      }
+  }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}
+%   {
+%     \msg_log:nnxxxx ,
+%     \msg_log:nnxxx  ,
+%     \msg_log:nnxx   ,
+%     \msg_log:nnx    ,
+%     \msg_log:nn
+%   }
+%   \enquote{Log} data is very similar to information, but with no extras
+%   added.
+%    \begin{macrocode}
+\msg_class_set:nn { log }
+  {
+    \msg_log:x
+      { \use:c { \c_msg_text_prefix_tl #1 / #2 } {#3} {#4} {#5} {#6} }
+  }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}
+%   {
+%     \msg_none:nnxxxx ,
+%     \msg_none:nnxxx  ,
+%     \msg_none:nnxx   ,
+%     \msg_none:nnx    ,
+%     \msg_none:nn
+%   }
+%   The \texttt{none} message type is needed so that input can be gobbled.
+%    \begin{macrocode}
+\msg_class_set:nn { none } { }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{variable}
+%   {
+%     \l_msg_redirect_classes_seq ,
+%     \l_msg_class_tl             ,
+%     \l_msg_current_class_tl     ,
+%     \l_msg_current_module_tl
+% }
+%   Support variables needed for the redirection system.
+%    \begin{macrocode}
+\seq_new:N \l_msg_redirect_classes_seq
+\tl_new:N \l_msg_class_tl
+\tl_new:N \l_msg_current_class_tl
+\tl_new:N \l_msg_current_module_tl
+%    \end{macrocode}
+%\end{variable}
+%
+% \begin{macro}[int]{\msg_use:nnnnxxxx}
+% \begin{macro}[aux]{\msg_use_aux:nnn}
+% \begin{macro}[aux]{\msg_use_aux:nn}
+% \begin{macro}[aux]{\msg_use_loop_check:nn}
+% \begin{macro}[aux]{\msg_use_code:}
+% \begin{macro}[aux]{\msg_use_loop:n, \msg_use_loop:o}
+%   The main message-using macro creates two auxiliary functions: one
+%   containing the code for the message, and the second a loop function.
+%   There is then a hand-off to the system for checking if redirection is
+%   needed.
+%    \begin{macrocode}
+\cs_new_protected:Npn \msg_use:nnnnxxxx #1#2#3#4#5#6#7#8
+  {
+    \cs_set_protected_nopar:Npx \msg_use_code:
+      {
+        \seq_clear:N \exp_not:N \l_msg_redirect_classes_seq
+        \exp_not:n {#2}
+      }
+    \cs_set_protected:Npx \msg_use_loop:n ##1
+      {
+        \seq_if_in:NnTF \exp_not:n \l_msg_redirect_classes_seq {#1}
+          { \msg_kernel_error:nn { msg }  { message-loop } {#1} }
+          {
+            \seq_put_right:Nn \exp_not:N \l_msg_redirect_classes_seq {#1}
+            \exp_not:N \cs_if_exist:cTF { msg_ ##1 :nnxxxx }
+              {
+                \exp_not:N \use:c { msg_ ##1 :nnxxxx }
+                  \exp_not:n { {#3} {#4} {#5} {#6} {#7} {#8} }
+              }
+              {
+                \msg_kernel_error:nnx { msg } { message-class-unknown } {##1}
+              }
+          }
+      }
+    \cs_if_exist:cTF { \c_msg_text_prefix_tl #3 / #4 }
+      { \msg_use_aux:nnn {#1} {#3} {#4} }
+      { \msg_kernel_error:nnxx { msg } { message-unknown } {#3} {#4} }
+  }
+%    \end{macrocode}
+%   The first auxiliary macro looks for a match by name: the most
+%   restrictive check.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \msg_use_aux:nnn #1#2#3
+  {
+    \tl_set:Nn \l_msg_current_class_tl  {#1}
+    \tl_set:Nn \l_msg_current_module_tl {#2}
+    \prop_if_in:NnTF \l_msg_redirect_names_prop { // #2 / #3 / }
+      { \msg_use_loop_check:nn { names } { // #2 / #3 / } }
+      { \msg_use_aux:nn {#1} {#2} }
+  }
+%    \end{macrocode}
+%   The second function checks for general matches by module or for
+%   all modules.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \msg_use_aux:nn #1#2
+  {
+    \prop_if_in:cnTF { l_msg_redirect_ #1 _prop } {#2}
+      { \msg_use_loop_check:nn {#1} {#2} }
+      {
+        \prop_if_in:cnTF { l_msg_redirect_ #1 _prop } { * }
+          { \msg_use_loop_check:nn {#1} { * } }
+          { \msg_use_code: }
+      }
+  }
+%    \end{macrocode}
+%   When checking whether to loop, the same code is needed in a few
+%   places.
+%    \begin{macrocode}
+\cs_new_protected:Npn \msg_use_loop_check:nn #1#2
+  {
+    \prop_get:cnN { l_msg_redirect_ #1 _prop } {#2} \l_msg_class_tl
+    \tl_if_eq:NNTF \l_msg_current_class_tl \l_msg_class_tl
+      {
+        { \msg_use_code: }
+        { \msg_use_loop:o \l_msg_class_tl }
+      }
+  }
+\cs_new_protected_nopar:Npn \msg_use_code: { }
+\cs_new_protected:Npn \msg_use_loop:n #1 { }
+\cs_generate_variant:Nn \msg_use_loop:n { o }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\msg_redirect_class:nn}
+%   Converts class one into class two.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \msg_redirect_class:nn #1#2
+  { \prop_put:cnn { l_msg_redirect_ #1 _prop } { * } {#2} }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\msg_redirect_module:nnn}
+%   For when all messages of a class should be altered for a given module.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \msg_redirect_module:nnn #1#2#3
+  { \prop_put:cnn { l_msg_redirect_ #2 _prop } {#1} {#3} }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\msg_redirect_name:nnn}
+%   Named message will always use the given class.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \msg_redirect_name:nnn #1#2#3
+  { \prop_put:Nnn \l_msg_redirect_names_prop { // #1 / #2 / } {#3} }
+%    \end{macrocode}
+%\end{macro}
+%
+% \subsection{Kernel-specific functions}
+%
+% \begin{macro}{\msg_kernel_new:nnnn}
+% \begin{macro}{\msg_kernel_new:nnn}
+% \begin{macro}{\msg_kernel_set:nnnn}
+% \begin{macro}{\msg_kernel_set:nnn}
+%   The kernel needs some messages of its own. These are created using
+%   pre-built functions. Two functions are provided: one more general
+%   and one which only has the short text part.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \msg_kernel_new:nnnn #1#2
+  { \msg_new:nnnn { LaTeX } { #1 / #2 } }
+\cs_new_protected_nopar:Npn \msg_kernel_new:nnn #1#2
+  { \msg_new:nnn { LaTeX } { #1 / #2 } }
+\cs_new_protected_nopar:Npn \msg_kernel_set:nnnn #1#2
+  { \msg_set:nnnn { LaTeX } { #1 / #2 } }
+\cs_new_protected_nopar:Npn \msg_kernel_set:nnn #1#2
+  { \msg_set:nnn { LaTeX } { #1 / #2 } }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\msg_kernel_fatal:nnxxxx}
+% \begin{macro}{\msg_kernel_fatal:nnxxx}
+% \begin{macro}{\msg_kernel_fatal:nnxx}
+% \begin{macro}{\msg_kernel_fatal:nnx}
+% \begin{macro}{\msg_kernel_fatal:nn}
+%   Fatal kernel errors cannot be re-defined.
+%    \begin{macrocode}
+\cs_new_protected:Npn \msg_kernel_fatal:nnxxxx #1#2#3#4#5#6
+  {
+    \msg_interrupt:xxx
+      { \msg_fatal_text:n { LaTeX } : ~ "#1 / #2" }
+      {
+        \use:c { \c_msg_text_prefix_tl LaTeX / #1 / #2 }
+          {#3} {#4} {#5} {#6}
+        \msg_see_documentation_text:n { LaTeX3 }
+      }
+      { \c_msg_fatal_text_tl }
+    \tex_end:D
+  }
+\cs_new_protected:Npn \msg_kernel_fatal:nnxxx #1#2#3#4#5
+  {\msg_kernel_fatal:nnxxxx {#1} {#2} {#3} {#4} {#5} { } }
+\cs_new_protected:Npn \msg_kernel_fatal:nnxx #1#2#3#4
+  { \msg_kernel_fatal:nnxxxx {#1} {#2} {#3} {#4} { } { } }
+\cs_new_protected:Npn \msg_kernel_fatal:nnx #1#2#3
+  { \msg_kernel_fatal:nnxxxx {#1} {#2} {#3} { } { } { } }
+\cs_new_protected:Npn \msg_kernel_fatal:nn #1#2
+  { \msg_kernel_fatal:nnxxxx {#1} {#2} { } { } { } { } }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\msg_kernel_error:nnxxxx}
+% \begin{macro}{\msg_kernel_error:nnxxx}
+% \begin{macro}{\msg_kernel_error:nnxx}
+% \begin{macro}{\msg_kernel_error:nnx}
+% \begin{macro}{\msg_kernel_error:nn}
+% Neither can kernel errors.
+%    \begin{macrocode}
+\cs_new_protected:Npn \msg_kernel_error:nnxxxx #1#2#3#4#5#6
+  {
+    \msg_if_more_text:cTF { \c_msg_more_text_prefix_tl LaTeX / #1 / #2 }
+      {
+        \msg_interrupt:xxx
+          { \msg_error_text:n { LaTeX } : ~ " #1 / #2 " }
+          {
+            \use:c { \c_msg_text_prefix_tl LaTeX / #1 / #2 }
+              {#3} {#4} {#5} {#6}
+            \msg_see_documentation_text:n { LaTeX3 }
+          }
+          {
+            \use:c { \c_msg_more_text_prefix_tl LaTeX / #1 / #2 }
+              {#3} {#4} {#5} {#6}
+          }
+      }
+      {
+        \msg_interrupt:xxx
+          { \msg_error_text:n { LaTeX } : ~ " #1 / #2 " }
+          {
+            \use:c { \c_msg_text_prefix_tl LaTeX / #1 / #2 }
+              {#3} {#4} {#5} {#6}
+            \msg_see_documentation_text:n { LaTeX3 }
+          }
+          { }
+      }
+  }
+\cs_new_protected:Npn \msg_kernel_error:nnxxx #1#2#3#4#5
+  {\msg_kernel_error:nnxxxx {#1} {#2} {#3} {#4} {#5} { } }
+\cs_set_protected:Npn \msg_kernel_error:nnxx #1#2#3#4
+  { \msg_kernel_error:nnxxxx {#1} {#2} {#3} {#4} { } { } }
+\cs_set_protected:Npn \msg_kernel_error:nnx #1#2#3
+  { \msg_kernel_error:nnxxxx {#1} {#2} {#3} { } { } { } }
+\cs_set_protected:Npn \msg_kernel_error:nn #1#2
+  { \msg_kernel_error:nnxxxx {#1} {#2} { } { } { } { } }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\msg_kernel_warning:nnxxxx}
+% \begin{macro}{\msg_kernel_warning:nnxxx}
+% \begin{macro}{\msg_kernel_warning:nnxx}
+% \begin{macro}{\msg_kernel_warning:nnx}
+% \begin{macro}{\msg_kernel_warning:nn}
+% \begin{macro}{\msg_kernel_info:nnxxxx}
+% \begin{macro}{\msg_kernel_info:nnxxx}
+% \begin{macro}{\msg_kernel_info:nnxx}
+% \begin{macro}{\msg_kernel_info:nnx}
+% \begin{macro}{\msg_kernel_info:nn}
+%   Kernel messages which can be redirected.
+%    \begin{macrocode}
+\prop_new:N \l_msg_redirect_kernel_warning_prop
+\cs_new_protected:Npn \msg_kernel_warning:nnxxxx #1#2#3#4#5#6
+  {
+    \msg_use:nnnnxxxx { warning }
+      {
+        \msg_term:x
+          {
+            \msg_warning_text:n { LaTeX } : ~ " #1 / #2 " \\ \\
+            \use:c { \c_msg_text_prefix_tl LaTeX / #1 / #2 }
+              {#3} {#4} {#5} {#6}
+          }
+      }
+      { LaTeX } { #1 / #2 } {#3} {#4} {#5} {#6}
+  }
+\cs_new_protected:Npn \msg_kernel_warning:nnxxx #1#2#3#4#5
+  { \msg_kernel_warning:nnxxxx {#1} {#2} {#3} {#4} {#5} { } }
+\cs_new_protected:Npn \msg_kernel_warning:nnxx #1#2#3#4
+  { \msg_kernel_warning:nnxxxx {#1} {#2} {#3} {#4} { } { } }
+\cs_new_protected:Npn \msg_kernel_warning:nnx #1#2#3
+  { \msg_kernel_warning:nnxxxx {#1} {#2} {#3} { } { } { } }
+\cs_new_protected:Npn \msg_kernel_warning:nn #1#2
+  { \msg_kernel_warning:nnxxxx {#1} {#2} { } { } { } { } }
+\prop_new:N \l_msg_redirect_kernel_info_prop
+\cs_new_protected:Npn \msg_kernel_info:nnxxxx #1#2#3#4#5#6
+  {
+    \msg_use:nnnnxxxx { info }
+      {
+        \msg_log:x
+          {
+            \msg_info_text:n { LaTeX } : ~ " #1 / #2 " \\ \\
+            \use:c { \c_msg_text_prefix_tl LaTeX / #1 / #2 }
+              {#3} {#4} {#5} {#6}
+          }
+      }
+      { LaTeX } { #1 / #2 } {#3} {#4} {#5} {#6}
+  }
+\cs_new_protected:Npn \msg_kernel_info:nnxxx #1#2#3#4#5
+  { \msg_kernel_info:nnxxxx {#1} {#2} {#3} {#4} {#5} { } }
+\cs_new_protected:Npn \msg_kernel_info:nnxx #1#2#3#4
+  { \msg_kernel_info:nnxxxx {#1} {#2} {#3} {#4} { } { } }
+\cs_new_protected:Npn \msg_kernel_info:nnx #1#2#3
+  { \msg_kernel_info:nnxxxx {#1} {#2} {#3} { } { } { } }
+\cs_new_protected:Npn \msg_kernel_info:nn #1#2
+  { \msg_kernel_info:nnxxxx {#1} {#2} { } { } { } { } }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% Error messages needed to actually implement the message system
+% itself.
+%    \begin{macrocode}
+\msg_kernel_new:nnnn { msg } { message-already-defined }
+  { Message~'#2'~for~module~'#1'~already~defined. }
+  {
+    \c_msg_coding_error_text_tl
+    LaTeX~was~asked~to~define~a~new~message~called~'#2'
+    by~the~module~'#1'~module:\\
+    this~message~already~exists.
+    \c_msg_return_text_tl
+  }
+\msg_kernel_new:nnnn { msg } { message-unknown }
+  { Unknown~message~'#2'~for~module~'#1'. }
+  {
+    \c_msg_coding_error_text_tl
+    LaTeX~was~asked~to~display~a~message~called~'#2'\\
+    by~the~module~'#1'~module:~this~message~does~not~exist.
+    \c_msg_return_text_tl
+  }
+\msg_kernel_new:nnnn { msg } { message-class-unknown }
+  { Unknown~message~class~'#1'. }
+  {
+    LaTeX~has~been~asked~to~redirect~messages~to~a~class~'#1':\\
+    this~was~never~defined.
+    \c_msg_return_text_tl
+  }
+\msg_kernel_new:nnnn { msg } { redirect-loop }
+  { Message~redirection~loop~for~message~class~'#1'. }
+  {
+    LaTeX~has~been~asked~to~redirect~messages~in~an~infinite~loop.\\
+    The~original~message~here~has~been~lost.
+    \c_msg_return_text_tl
+  }
+%    \end{macrocode}
+% 
+% Messages for earlier kernel modules.
+%    \begin{macrocode}
+\msg_kernel_new:nnnn { kernel } { bad-number-of-arguments }
+  { Function~'#1'~cannot~be~defined~with~#2~arguments. }
+  {
+    \c_msg_coding_error_text_tl
+    LaTeX~has~been~asked~to~define~a~function~'#1'~with~
+    #2~arguments. \\
+    TeX~allows~between~0~and~9~arguments~for~a~single~function.
+  }
+\msg_kernel_new:nnnn { kernel } { command-already-defined }
+  { Control~sequence~#1~already~defined. }
+  {
+    \c_msg_coding_error_text_tl
+    LaTeX~has~been~asked~to~create~a~new~control~sequence~'#1'~
+    but~this~name~has~already~been~used~elsewhere. \\ \\
+    The~current~meaning~is:\\
+    \ \ #2
+  }
+\msg_kernel_new:nnnn { kernel } { command-not-defined }
+  { Control~sequence~#1~undefined. }
+  {
+    \c_msg_coding_error_text_tl
+    LaTeX~has~been~asked~to~use~a~command~#1,~but~this~has~not~
+    been~defined~yet.
+  }
+\msg_kernel_new:nnnn { kernel } { variable-not-defined }
+  { Variable~#1~undefined. }
+  {
+    \c_msg_coding_error_text_tl
+    LaTeX~has~been~asked~to~show~a~variable~#1,~but~this~has~not~
+    been~defined~yet.
+  }
+\msg_kernel_new:nnnn { seq } { empty-sequence }
+  { Empty~sequence~#1. }
+  {
+    \c_msg_coding_error_text_tl
+    LaTeX~has~been~asked~to~recover~an~entry~from~a~sequence~that~
+    has~no~content:~that~cannot~happen!
+  }
+%    \end{macrocode}
+%
+% \begin{macro}{\msg_kernel_bug:x}
+% \begin{variable}{\c_msg_kernel_bug_text_tl, \c_msg_kernel_bug_more_text_tl}
+% The \LaTeX{} coding bug error gets re-visited here.
+%    \begin{macrocode}
+\cs_set_protected:Npn \msg_kernel_bug:x #1
+  {
+    \msg_interrupt:xxx { \c_msg_kernel_bug_text_tl }
+      {
+        #1
+       \msg_see_documentation_text:n { LaTeX3 }
+      }
+      { \c_msg_kernel_bug_more_text_tl }
+  }
+\tl_const:Nn \c_msg_kernel_bug_text_tl
+  { This~is~a~LaTeX~bug:~check~coding! }
+\tl_const:Nn \c_msg_kernel_bug_more_text_tl
+  {
+    There~is~a~coding~bug~somewhere~around~here. \\
+    This~probably~needs~examining~by~an~expert.
+    \c_msg_return_text_tl
+  }
+%    \end{macrocode}
+% \end{variable}
+% \end{macro}
+%
+% \subsection{Deprecated functions}
+%
+% Deprecated on 2011-05-27, for removal by 2011-08-31.
+%
+% \begin{macro}{\msg_class_new:nn}
+% This is only ever used in a |set| fashion.
+%    \begin{macrocode}
+\cs_new_eq:NN \msg_class_new:nn \msg_class_set:nn
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}
+%   {
+%     \msg_trace:nnxxxx, \msg_trace:nnxxx, \msg_trace:nnxx,
+%     \msg_trace:nnx,    \msg_trace:nn
+%   }
+%   The performance here is never going to be good enough for tracing
+%   code, so let's be realistic.
+%    \begin{macrocode}
+\cs_new_eq:NN \msg_trace:nnxxxx \msg_log:nnxxxx
+\cs_new_eq:NN \msg_trace:nnxxx  \msg_log:nnxxx
+\cs_new_eq:NN \msg_trace:nnxx   \msg_log:nnxx
+\cs_new_eq:NN \msg_trace:nnx    \msg_log:nnx
+\cs_new_eq:NN \msg_trace:nn     \msg_log:nn
+%    \end{macrocode}
+%\end{macro}
+%
+% \begin{macro}{\msg_generic_new:nnn}
+% \begin{macro}{\msg_generic_new:nn}
+% \begin{macro}{\msg_generic_set:nnn}
+% \begin{macro}{\msg_generic_set:nn}
+% \begin{macro}{\msg_direct_interrupt:xxxxx}
+% \begin{macro}{\msg_direct_log:xx}
+% \begin{macro}{\msg_direct_term:xx}
+%  These were all too low-level.
+%    \begin{macrocode}
+\cs_new_protected:Npn \msg_generic_new:nnn #1#2#3 { \deprecated }
+\cs_new_protected:Npn \msg_generic_new:nn  #1#2   { \deprecated }
+\cs_new_protected:Npn \msg_generic_set:nnn #1#2#3 { \deprecated }
+\cs_new_protected:Npn \msg_generic_set:nn  #1#2   { \deprecated }
+\cs_new_protected:Npn \msg_direct_interrupt:xxxxx #1#2#3#4#5 { \deprecated }
+\cs_new_protected:Npn \msg_direct_log:xx #1#2  { \deprecated }
+\cs_new_protected:Npn \msg_direct_term:xx #1#2 { \deprecated }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+%    \begin{macrocode}
+%</initex|package>
+%    \end{macrocode}
+%
+% \end{implementation}
+%
+% \PrintIndex
diff --git a/Master/texmf-dist/source/latex/l3kernel/l3names.dtx b/Master/texmf-dist/source/latex/l3kernel/l3names.dtx
new file mode 100644
index 00000000000..d8a33196962
--- /dev/null
+++ b/Master/texmf-dist/source/latex/l3kernel/l3names.dtx
@@ -0,0 +1,629 @@
+% \iffalse meta-comment
+%
+%% File: l3names.dtx Copyright (C) 1990-2011 The LaTeX3 project
+%%
+%% It may be distributed and/or modified under the conditions of the
+%% LaTeX Project Public License (LPPL), either version 1.3c of this
+%% license or (at your option) any later version.  The latest version
+%% of this license is in the file
+%%
+%%    http://www.latex-project.org/lppl.txt
+%%
+%% This file is part of the "expl3 bundle" (The Work in LPPL)
+%% and all files in that bundle must be distributed together.
+%%
+%% The released version of this bundle is available from CTAN.
+%%
+%% -----------------------------------------------------------------------
+%%
+%% The development version of the bundle can be found at
+%%
+%%    http://www.latex-project.org/svnroot/experimental/trunk/
+%%
+%% for those people who are interested.
+%%
+%%%%%%%%%%%
+%% NOTE: %%
+%%%%%%%%%%%
+%%
+%%   Snapshots taken from the repository represent work in progress and may
+%%   not work or may contain conflicting material!  We therefore ask
+%%   people _not_ to put them into distributions, archives, etc. without
+%%   prior consultation with the LaTeX3 Project.
+%%
+%% -----------------------------------------------------------------------
+%
+%<*package>
+\RequirePackage{l3bootstrap}
+\GetIdInfo$Id: l3names.dtx 2478 2011-06-19 21:34:23Z joseph $
+  {L3 Experimental namespace for primitives}
+%</package>
+%<*driver>
+\documentclass[full]{l3doc}
+\begin{document}
+  \DocInput{\jobname.dtx}
+\end{document}
+%</driver>
+% \fi
+%
+% \title{^^A
+%   The \pkg{l3names} package\\ Namespace for primitives^^A
+%   \thanks{This file describes v\ExplFileVersion,
+%      last revised \ExplFileDate.}^^A
+% }
+%
+% \author{^^A
+%  The \LaTeX3 Project\thanks
+%    {^^A
+%      E-mail:
+%        \href{mailto:latex-team@latex-project.org}
+%          {latex-team@latex-project.org}^^A
+%    }^^A
+% }
+%
+% \date{Released \ExplFileDate}
+%
+% \maketitle
+%
+% \begin{documentation}
+%
+% \section{Setting up the \LaTeX3 programming language}
+%
+% This module is at the core of the \LaTeX3 programming language. It
+% performs the following tasks:
+% \begin{itemize}
+%   \item defines new names for all \TeX{} primitives;
+%   \item switches to the category code regime for programming;
+%   \item provides support settings for building the code as a \TeX{} format.
+% \end{itemize}
+%
+% This module is entirely dedicated to primitives, which should not
+% be used directly within \LaTeX3 code (outside of \enquote{kernel-level}
+% code). As such, the primitives are not documented here:
+% \emph{The \TeX{}book}, \emph{\TeX{} by Topic} and the manuals for
+% \pdfTeX{}, \XeTeX{} and \LuaTeX{} should be consulted for details of
+% the primitives. These are named based on the engine which first introduced
+% them:
+% \begin{itemize}
+%   \item[\cs{tex_\ldots}] Introduced by \TeX{} itself;
+%   \item[\cs{etex_\ldots}] Introduced by the \eTeX{} extensions;
+%   \item[\cs{pdftex_\ldots}] Introduced by \pdfTeX{};
+%   \item[\cs{xetex_\ldots}] Introduced by \XeTeX{};
+%   \item[\cs{luatex_\ldots}] Introduced by \LuaTeX{}.
+% \end{itemize}
+%
+% \end{documentation}
+%
+% \begin{implementation}
+%
+% \section{\pkg{l3names} implementation}
+%
+%    \begin{macrocode}
+%<*initex|package>
+%    \end{macrocode}
+%
+%    \begin{macrocode}
+%<*package>
+\ProvidesExplPackage
+  {\ExplFileName}{\ExplFileDate}{\ExplFileVersion}{\ExplFileDescription}
+%</package>
+%    \end{macrocode}
+%
+% The code here simply renames all of the primitives to new, internal,
+% names. In format mode, it also deletes all of the existing names (although
+% some od come back later).
+%
+% \begin{macro}{\tex_undefined:D}
+% This function does not exist at all, but is the name used by the
+% plain \TeX{} format for an undefined function. So it should
+% be marked here as \enquote{taken}.
+% \end{macro}
+%
+% The \cs{let} primitive is renamed by hand first as it is essential for
+% the entire process to follow. This also uses \cs{global}, as that way
+% we avoid leaving an unneeded csname in the hash table.
+%    \begin{macrocode}
+\let \tex_global:D \global
+\let \tex_let:D    \let
+%    \end{macrocode}
+%
+% Everything is inside a (rather long) group, which kees \cs{name_primitve:NN}
+% trapped.
+%    \begin{macrocode}
+\begingroup
+%    \end{macrocode}
+%
+% \begin{macro}[aux]{\name_primitive:NN}
+%   A temporary function to actually do the renaming. This also allows the
+%   original names to be removed in format mode.
+%    \begin{macrocode}
+  \long \def \name_primitive:NN #1#2
+    {
+      \tex_global:D \tex_let:D #2 #1
+%<*initex>
+      \tex_global:D \tex_let:D #1 \tex_undefined:D
+%</initex>
+    }
+%    \end{macrocode}
+%\end{macro}
+%
+% In the current incarnation of this package, all \TeX{} primitives
+% are given a new name of the form |\tex_|\emph{oldname}|:D|.
+% But first three special cases which have symbolic original names.
+% These are given modified new names, so that they may be entered
+% without catcode tricks.
+%    \begin{macrocode}
+  \name_primitive:NN \                      \tex_space:D
+  \name_primitive:NN \/                     \tex_italiccor:D
+  \name_primitive:NN \-                     \tex_hyphen:D
+%    \end{macrocode}
+%
+% Now all the other primitives.
+%    \begin{macrocode}
+  \name_primitive:NN \let                   \tex_let:D
+  \name_primitive:NN \def                   \tex_def:D
+  \name_primitive:NN \edef                  \tex_edef:D
+  \name_primitive:NN \gdef                  \tex_gdef:D
+  \name_primitive:NN \xdef                  \tex_xdef:D
+  \name_primitive:NN \chardef               \tex_chardef:D
+  \name_primitive:NN \countdef              \tex_countdef:D
+  \name_primitive:NN \dimendef              \tex_dimendef:D
+  \name_primitive:NN \skipdef               \tex_skipdef:D
+  \name_primitive:NN \muskipdef             \tex_muskipdef:D
+  \name_primitive:NN \mathchardef           \tex_mathchardef:D
+  \name_primitive:NN \toksdef               \tex_toksdef:D
+  \name_primitive:NN \futurelet             \tex_futurelet:D
+  \name_primitive:NN \advance               \tex_advance:D
+  \name_primitive:NN \divide                \tex_divide:D
+  \name_primitive:NN \multiply              \tex_multiply:D
+  \name_primitive:NN \font                  \tex_font:D
+  \name_primitive:NN \fam                   \tex_fam:D
+  \name_primitive:NN \global                \tex_global:D
+  \name_primitive:NN \long                  \tex_long:D
+  \name_primitive:NN \outer                 \tex_outer:D
+  \name_primitive:NN \setlanguage           \tex_setlanguage:D
+  \name_primitive:NN \globaldefs            \tex_globaldefs:D
+  \name_primitive:NN \afterassignment       \tex_afterassignment:D
+  \name_primitive:NN \aftergroup            \tex_aftergroup:D
+  \name_primitive:NN \expandafter           \tex_expandafter:D
+  \name_primitive:NN \noexpand              \tex_noexpand:D
+  \name_primitive:NN \begingroup            \tex_begingroup:D
+  \name_primitive:NN \endgroup              \tex_endgroup:D
+  \name_primitive:NN \halign                \tex_halign:D
+  \name_primitive:NN \valign                \tex_valign:D
+  \name_primitive:NN \cr                    \tex_cr:D
+  \name_primitive:NN \crcr                  \tex_crcr:D
+  \name_primitive:NN \noalign               \tex_noalign:D
+  \name_primitive:NN \omit                  \tex_omit:D
+  \name_primitive:NN \span                  \tex_span:D
+  \name_primitive:NN \tabskip               \tex_tabskip:D
+  \name_primitive:NN \everycr               \tex_everycr:D
+  \name_primitive:NN \if                    \tex_if:D
+  \name_primitive:NN \ifcase                \tex_ifcase:D
+  \name_primitive:NN \ifcat                 \tex_ifcat:D
+  \name_primitive:NN \ifnum                 \tex_ifnum:D
+  \name_primitive:NN \ifodd                 \tex_ifodd:D
+  \name_primitive:NN \ifdim                 \tex_ifdim:D
+  \name_primitive:NN \ifeof                 \tex_ifeof:D
+  \name_primitive:NN \ifhbox                \tex_ifhbox:D
+  \name_primitive:NN \ifvbox                \tex_ifvbox:D
+  \name_primitive:NN \ifvoid                \tex_ifvoid:D
+  \name_primitive:NN \ifx                   \tex_ifx:D
+  \name_primitive:NN \iffalse               \tex_iffalse:D
+  \name_primitive:NN \iftrue                \tex_iftrue:D
+  \name_primitive:NN \ifhmode               \tex_ifhmode:D
+  \name_primitive:NN \ifmmode               \tex_ifmmode:D
+  \name_primitive:NN \ifvmode               \tex_ifvmode:D
+  \name_primitive:NN \ifinner               \tex_ifinner:D
+  \name_primitive:NN \else                  \tex_else:D
+  \name_primitive:NN \fi                    \tex_fi:D
+  \name_primitive:NN \or                    \tex_or:D
+  \name_primitive:NN \immediate             \tex_immediate:D
+  \name_primitive:NN \closeout              \tex_closeout:D
+  \name_primitive:NN \openin                \tex_openin:D
+  \name_primitive:NN \openout               \tex_openout:D
+  \name_primitive:NN \read                  \tex_read:D
+  \name_primitive:NN \write                 \tex_write:D
+  \name_primitive:NN \closein               \tex_closein:D
+  \name_primitive:NN \newlinechar           \tex_newlinechar:D
+  \name_primitive:NN \input                 \tex_input:D
+  \name_primitive:NN \endinput              \tex_endinput:D
+  \name_primitive:NN \inputlineno           \tex_inputlineno:D
+  \name_primitive:NN \errmessage            \tex_errmessage:D
+  \name_primitive:NN \message               \tex_message:D
+  \name_primitive:NN \show                  \tex_show:D
+  \name_primitive:NN \showthe               \tex_showthe:D
+  \name_primitive:NN \showbox               \tex_showbox:D
+  \name_primitive:NN \showlists             \tex_showlists:D
+  \name_primitive:NN \errhelp               \tex_errhelp:D
+  \name_primitive:NN \errorcontextlines     \tex_errorcontextlines:D
+  \name_primitive:NN \tracingcommands       \tex_tracingcommands:D
+  \name_primitive:NN \tracinglostchars      \tex_tracinglostchars:D
+  \name_primitive:NN \tracingmacros         \tex_tracingmacros:D
+  \name_primitive:NN \tracingonline         \tex_tracingonline:D
+  \name_primitive:NN \tracingoutput         \tex_tracingoutput:D
+  \name_primitive:NN \tracingpages          \tex_tracingpages:D
+  \name_primitive:NN \tracingparagraphs     \tex_tracingparagraphs:D
+  \name_primitive:NN \tracingrestores       \tex_tracingrestores:D
+  \name_primitive:NN \tracingstats          \tex_tracingstats:D
+  \name_primitive:NN \pausing               \tex_pausing:D
+  \name_primitive:NN \showboxbreadth        \tex_showboxbreadth:D
+  \name_primitive:NN \showboxdepth          \tex_showboxdepth:D
+  \name_primitive:NN \batchmode             \tex_batchmode:D
+  \name_primitive:NN \errorstopmode         \tex_errorstopmode:D
+  \name_primitive:NN \nonstopmode           \tex_nonstopmode:D
+  \name_primitive:NN \scrollmode            \tex_scrollmode:D
+  \name_primitive:NN \end                   \tex_end:D
+  \name_primitive:NN \csname                \tex_csname:D
+  \name_primitive:NN \endcsname             \tex_endcsname:D
+  \name_primitive:NN \ignorespaces          \tex_ignorespaces:D
+  \name_primitive:NN \relax                 \tex_relax:D
+  \name_primitive:NN \the                   \tex_the:D
+  \name_primitive:NN \mag                   \tex_mag:D
+  \name_primitive:NN \language              \tex_language:D
+  \name_primitive:NN \mark                  \tex_mark:D
+  \name_primitive:NN \topmark               \tex_topmark:D
+  \name_primitive:NN \firstmark             \tex_firstmark:D
+  \name_primitive:NN \botmark               \tex_botmark:D
+  \name_primitive:NN \splitfirstmark        \tex_splitfirstmark:D
+  \name_primitive:NN \splitbotmark          \tex_splitbotmark:D
+  \name_primitive:NN \fontname              \tex_fontname:D
+  \name_primitive:NN \escapechar            \tex_escapechar:D
+  \name_primitive:NN \endlinechar           \tex_endlinechar:D
+  \name_primitive:NN \mathchoice            \tex_mathchoice:D
+  \name_primitive:NN \delimiter             \tex_delimiter:D
+  \name_primitive:NN \mathaccent            \tex_mathaccent:D
+  \name_primitive:NN \mathchar              \tex_mathchar:D
+  \name_primitive:NN \mskip                 \tex_mskip:D
+  \name_primitive:NN \radical               \tex_radical:D
+  \name_primitive:NN \vcenter               \tex_vcenter:D
+  \name_primitive:NN \mkern                 \tex_mkern:D
+  \name_primitive:NN \above                 \tex_above:D
+  \name_primitive:NN \abovewithdelims       \tex_abovewithdelims:D
+  \name_primitive:NN \atop                  \tex_atop:D
+  \name_primitive:NN \atopwithdelims        \tex_atopwithdelims:D
+  \name_primitive:NN \over                  \tex_over:D
+  \name_primitive:NN \overwithdelims        \tex_overwithdelims:D
+  \name_primitive:NN \displaystyle          \tex_displaystyle:D
+  \name_primitive:NN \textstyle             \tex_textstyle:D
+  \name_primitive:NN \scriptstyle           \tex_scriptstyle:D
+  \name_primitive:NN \scriptscriptstyle     \tex_scriptscriptstyle:D
+  \name_primitive:NN \nonscript             \tex_nonscript:D
+  \name_primitive:NN \eqno                  \tex_eqno:D
+  \name_primitive:NN \leqno                 \tex_leqno:D
+  \name_primitive:NN \abovedisplayshortskip \tex_abovedisplayshortskip:D
+  \name_primitive:NN \abovedisplayskip      \tex_abovedisplayskip:D
+  \name_primitive:NN \belowdisplayshortskip \tex_belowdisplayshortskip:D
+  \name_primitive:NN \belowdisplayskip      \tex_belowdisplayskip:D
+  \name_primitive:NN \displaywidowpenalty   \tex_displaywidowpenalty:D
+  \name_primitive:NN \displayindent         \tex_displayindent:D
+  \name_primitive:NN \displaywidth          \tex_displaywidth:D
+  \name_primitive:NN \everydisplay          \tex_everydisplay:D
+  \name_primitive:NN \predisplaysize        \tex_predisplaysize:D
+  \name_primitive:NN \predisplaypenalty     \tex_predisplaypenalty:D
+  \name_primitive:NN \postdisplaypenalty    \tex_postdisplaypenalty:D
+  \name_primitive:NN \mathbin               \tex_mathbin:D
+  \name_primitive:NN \mathclose             \tex_mathclose:D
+  \name_primitive:NN \mathinner             \tex_mathinner:D
+  \name_primitive:NN \mathop                \tex_mathop:D
+  \name_primitive:NN \displaylimits         \tex_displaylimits:D
+  \name_primitive:NN \limits                \tex_limits:D
+  \name_primitive:NN \nolimits              \tex_nolimits:D
+  \name_primitive:NN \mathopen              \tex_mathopen:D
+  \name_primitive:NN \mathord               \tex_mathord:D
+  \name_primitive:NN \mathpunct             \tex_mathpunct:D
+  \name_primitive:NN \mathrel               \tex_mathrel:D
+  \name_primitive:NN \overline              \tex_overline:D
+  \name_primitive:NN \underline             \tex_underline:D
+  \name_primitive:NN \left                  \tex_left:D
+  \name_primitive:NN \right                 \tex_right:D
+  \name_primitive:NN \binoppenalty          \tex_binoppenalty:D
+  \name_primitive:NN \relpenalty            \tex_relpenalty:D
+  \name_primitive:NN \delimitershortfall    \tex_delimitershortfall:D
+  \name_primitive:NN \delimiterfactor       \tex_delimiterfactor:D
+  \name_primitive:NN \nulldelimiterspace    \tex_nulldelimiterspace:D
+  \name_primitive:NN \everymath             \tex_everymath:D
+  \name_primitive:NN \mathsurround          \tex_mathsurround:D
+  \name_primitive:NN \medmuskip             \tex_medmuskip:D
+  \name_primitive:NN \thinmuskip            \tex_thinmuskip:D
+  \name_primitive:NN \thickmuskip           \tex_thickmuskip:D
+  \name_primitive:NN \scriptspace           \tex_scriptspace:D
+  \name_primitive:NN \noboundary            \tex_noboundary:D
+  \name_primitive:NN \accent                \tex_accent:D
+  \name_primitive:NN \char                  \tex_char:D
+  \name_primitive:NN \discretionary         \tex_discretionary:D
+  \name_primitive:NN \hfil                  \tex_hfil:D
+  \name_primitive:NN \hfilneg               \tex_hfilneg:D
+  \name_primitive:NN \hfill                 \tex_hfill:D
+  \name_primitive:NN \hskip                 \tex_hskip:D
+  \name_primitive:NN \hss                   \tex_hss:D
+  \name_primitive:NN \vfil                  \tex_vfil:D
+  \name_primitive:NN \vfilneg               \tex_vfilneg:D
+  \name_primitive:NN \vfill                 \tex_vfill:D
+  \name_primitive:NN \vskip                 \tex_vskip:D
+  \name_primitive:NN \vss                   \tex_vss:D
+  \name_primitive:NN \unskip                \tex_unskip:D
+  \name_primitive:NN \kern                  \tex_kern:D
+  \name_primitive:NN \unkern                \tex_unkern:D
+  \name_primitive:NN \hrule                 \tex_hrule:D
+  \name_primitive:NN \vrule                 \tex_vrule:D
+  \name_primitive:NN \leaders               \tex_leaders:D
+  \name_primitive:NN \cleaders              \tex_cleaders:D
+  \name_primitive:NN \xleaders              \tex_xleaders:D
+  \name_primitive:NN \lastkern              \tex_lastkern:D
+  \name_primitive:NN \lastskip              \tex_lastskip:D
+  \name_primitive:NN \indent                \tex_indent:D
+  \name_primitive:NN \par                   \tex_par:D
+  \name_primitive:NN \noindent              \tex_noindent:D
+  \name_primitive:NN \vadjust               \tex_vadjust:D
+  \name_primitive:NN \baselineskip          \tex_baselineskip:D
+  \name_primitive:NN \lineskip              \tex_lineskip:D
+  \name_primitive:NN \lineskiplimit         \tex_lineskiplimit:D
+  \name_primitive:NN \clubpenalty           \tex_clubpenalty:D
+  \name_primitive:NN \widowpenalty          \tex_widowpenalty:D
+  \name_primitive:NN \exhyphenpenalty       \tex_exhyphenpenalty:D
+  \name_primitive:NN \hyphenpenalty         \tex_hyphenpenalty:D
+  \name_primitive:NN \linepenalty           \tex_linepenalty:D
+  \name_primitive:NN \doublehyphendemerits  \tex_doublehyphendemerits:D
+  \name_primitive:NN \finalhyphendemerits   \tex_finalhyphendemerits:D
+  \name_primitive:NN \adjdemerits           \tex_adjdemerits:D
+  \name_primitive:NN \hangafter             \tex_hangafter:D
+  \name_primitive:NN \hangindent            \tex_hangindent:D
+  \name_primitive:NN \parshape              \tex_parshape:D
+  \name_primitive:NN \hsize                 \tex_hsize:D
+  \name_primitive:NN \lefthyphenmin         \tex_lefthyphenmin:D
+  \name_primitive:NN \righthyphenmin        \tex_righthyphenmin:D
+  \name_primitive:NN \leftskip              \tex_leftskip:D
+  \name_primitive:NN \rightskip             \tex_rightskip:D
+  \name_primitive:NN \looseness             \tex_looseness:D
+  \name_primitive:NN \parskip               \tex_parskip:D
+  \name_primitive:NN \parindent             \tex_parindent:D
+  \name_primitive:NN \uchyph                \tex_uchyph:D
+  \name_primitive:NN \emergencystretch      \tex_emergencystretch:D
+  \name_primitive:NN \pretolerance          \tex_pretolerance:D
+  \name_primitive:NN \tolerance             \tex_tolerance:D
+  \name_primitive:NN \spaceskip             \tex_spaceskip:D
+  \name_primitive:NN \xspaceskip            \tex_xspaceskip:D
+  \name_primitive:NN \parfillskip           \tex_parfillskip:D
+  \name_primitive:NN \everypar              \tex_everypar:D
+  \name_primitive:NN \prevgraf              \tex_prevgraf:D
+  \name_primitive:NN \spacefactor           \tex_spacefactor:D
+  \name_primitive:NN \shipout               \tex_shipout:D
+  \name_primitive:NN \vsize                 \tex_vsize:D
+  \name_primitive:NN \interlinepenalty      \tex_interlinepenalty:D
+  \name_primitive:NN \brokenpenalty         \tex_brokenpenalty:D
+  \name_primitive:NN \topskip               \tex_topskip:D
+  \name_primitive:NN \maxdeadcycles         \tex_maxdeadcycles:D
+  \name_primitive:NN \maxdepth              \tex_maxdepth:D
+  \name_primitive:NN \output                \tex_output:D
+  \name_primitive:NN \deadcycles            \tex_deadcycles:D
+  \name_primitive:NN \pagedepth             \tex_pagedepth:D
+  \name_primitive:NN \pagestretch           \tex_pagestretch:D
+  \name_primitive:NN \pagefilstretch        \tex_pagefilstretch:D
+  \name_primitive:NN \pagefillstretch       \tex_pagefillstretch:D
+  \name_primitive:NN \pagefilllstretch      \tex_pagefilllstretch:D
+  \name_primitive:NN \pageshrink            \tex_pageshrink:D
+  \name_primitive:NN \pagegoal              \tex_pagegoal:D
+  \name_primitive:NN \pagetotal             \tex_pagetotal:D
+  \name_primitive:NN \outputpenalty         \tex_outputpenalty:D
+  \name_primitive:NN \hoffset               \tex_hoffset:D
+  \name_primitive:NN \voffset               \tex_voffset:D
+  \name_primitive:NN \insert                \tex_insert:D
+  \name_primitive:NN \holdinginserts        \tex_holdinginserts:D
+  \name_primitive:NN \floatingpenalty       \tex_floatingpenalty:D
+  \name_primitive:NN \insertpenalties       \tex_insertpenalties:D
+  \name_primitive:NN \lower                 \tex_lower:D
+  \name_primitive:NN \moveleft              \tex_moveleft:D
+  \name_primitive:NN \moveright             \tex_moveright:D
+  \name_primitive:NN \raise                 \tex_raise:D
+  \name_primitive:NN \copy                  \tex_copy:D
+  \name_primitive:NN \lastbox               \tex_lastbox:D
+  \name_primitive:NN \vsplit                \tex_vsplit:D
+  \name_primitive:NN \unhbox                \tex_unhbox:D
+  \name_primitive:NN \unhcopy               \tex_unhcopy:D
+  \name_primitive:NN \unvbox                \tex_unvbox:D
+  \name_primitive:NN \unvcopy               \tex_unvcopy:D
+  \name_primitive:NN \setbox                \tex_setbox:D
+  \name_primitive:NN \hbox                  \tex_hbox:D
+  \name_primitive:NN \vbox                  \tex_vbox:D
+  \name_primitive:NN \vtop                  \tex_vtop:D
+  \name_primitive:NN \prevdepth             \tex_prevdepth:D
+  \name_primitive:NN \badness               \tex_badness:D
+  \name_primitive:NN \hbadness              \tex_hbadness:D
+  \name_primitive:NN \vbadness              \tex_vbadness:D
+  \name_primitive:NN \hfuzz                 \tex_hfuzz:D
+  \name_primitive:NN \vfuzz                 \tex_vfuzz:D
+  \name_primitive:NN \overfullrule          \tex_overfullrule:D
+  \name_primitive:NN \boxmaxdepth           \tex_boxmaxdepth:D
+  \name_primitive:NN \splitmaxdepth         \tex_splitmaxdepth:D
+  \name_primitive:NN \splittopskip          \tex_splittopskip:D
+  \name_primitive:NN \everyhbox             \tex_everyhbox:D
+  \name_primitive:NN \everyvbox             \tex_everyvbox:D
+  \name_primitive:NN \nullfont              \tex_nullfont:D
+  \name_primitive:NN \textfont              \tex_textfont:D
+  \name_primitive:NN \scriptfont            \tex_scriptfont:D
+  \name_primitive:NN \scriptscriptfont      \tex_scriptscriptfont:D
+  \name_primitive:NN \fontdimen             \tex_fontdimen:D
+  \name_primitive:NN \hyphenchar            \tex_hyphenchar:D
+  \name_primitive:NN \skewchar              \tex_skewchar:D
+  \name_primitive:NN \defaulthyphenchar     \tex_defaulthyphenchar:D
+  \name_primitive:NN \defaultskewchar       \tex_defaultskewchar:D
+  \name_primitive:NN \number                \tex_number:D
+  \name_primitive:NN \romannumeral          \tex_romannumeral:D
+  \name_primitive:NN \string                \tex_string:D
+  \name_primitive:NN \lowercase             \tex_lowercase:D
+  \name_primitive:NN \uppercase             \tex_uppercase:D
+  \name_primitive:NN \meaning               \tex_meaning:D
+  \name_primitive:NN \penalty               \tex_penalty:D
+  \name_primitive:NN \unpenalty             \tex_unpenalty:D
+  \name_primitive:NN \lastpenalty           \tex_lastpenalty:D
+  \name_primitive:NN \special               \tex_special:D
+  \name_primitive:NN \dump                  \tex_dump:D
+  \name_primitive:NN \patterns              \tex_patterns:D
+  \name_primitive:NN \hyphenation           \tex_hyphenation:D
+  \name_primitive:NN \time                  \tex_time:D
+  \name_primitive:NN \day                   \tex_day:D
+  \name_primitive:NN \month                 \tex_month:D
+  \name_primitive:NN \year                  \tex_year:D
+  \name_primitive:NN \jobname               \tex_jobname:D
+  \name_primitive:NN \everyjob              \tex_everyjob:D
+  \name_primitive:NN \count                 \tex_count:D
+  \name_primitive:NN \dimen                 \tex_dimen:D
+  \name_primitive:NN \skip                  \tex_skip:D
+  \name_primitive:NN \toks                  \tex_toks:D
+  \name_primitive:NN \muskip                \tex_muskip:D
+  \name_primitive:NN \box                   \tex_box:D
+  \name_primitive:NN \wd                    \tex_wd:D
+  \name_primitive:NN \ht                    \tex_ht:D
+  \name_primitive:NN \dp                    \tex_dp:D
+  \name_primitive:NN \catcode               \tex_catcode:D
+  \name_primitive:NN \delcode               \tex_delcode:D
+  \name_primitive:NN \sfcode                \tex_sfcode:D
+  \name_primitive:NN \lccode                \tex_lccode:D
+  \name_primitive:NN \uccode                \tex_uccode:D
+  \name_primitive:NN \mathcode              \tex_mathcode:D
+%    \end{macrocode}
+% Since \LaTeX3 requires at least the \eTeX{} extensions,
+% we also rename the additional primitives. These are all
+% given the prefix |\etex_|.
+%    \begin{macrocode}
+  \name_primitive:NN \ifdefined             \etex_ifdefined:D
+  \name_primitive:NN \ifcsname              \etex_ifcsname:D
+  \name_primitive:NN \unless                \etex_unless:D
+  \name_primitive:NN \eTeXversion           \etex_eTeXversion:D
+  \name_primitive:NN \eTeXrevision          \etex_eTeXrevision:D
+  \name_primitive:NN \marks                 \etex_marks:D
+  \name_primitive:NN \topmarks              \etex_topmarks:D
+  \name_primitive:NN \firstmarks            \etex_firstmarks:D
+  \name_primitive:NN \botmarks              \etex_botmarks:D
+  \name_primitive:NN \splitfirstmarks       \etex_splitfirstmarks:D
+  \name_primitive:NN \splitbotmarks         \etex_splitbotmarks:D
+  \name_primitive:NN \unexpanded            \etex_unexpanded:D
+  \name_primitive:NN \detokenize            \etex_detokenize:D
+  \name_primitive:NN \scantokens            \etex_scantokens:D
+  \name_primitive:NN \showtokens            \etex_showtokens:D
+  \name_primitive:NN \readline              \etex_readline:D
+  \name_primitive:NN \tracingassigns        \etex_tracingassigns:D
+  \name_primitive:NN \tracingscantokens     \etex_tracingscantokens:D
+  \name_primitive:NN \tracingnesting        \etex_tracingnesting:D
+  \name_primitive:NN \tracingifs            \etex_tracingifs:D
+  \name_primitive:NN \currentiflevel        \etex_currentiflevel:D
+  \name_primitive:NN \currentifbranch       \etex_currentifbranch:D
+  \name_primitive:NN \currentiftype         \etex_currentiftype:D
+  \name_primitive:NN \tracinggroups         \etex_tracinggroups:D
+  \name_primitive:NN \currentgrouplevel     \etex_currentgrouplevel:D
+  \name_primitive:NN \currentgrouptype      \etex_currentgrouptype:D
+  \name_primitive:NN \showgroups            \etex_showgroups:D
+  \name_primitive:NN \showifs               \etex_showifs:D
+  \name_primitive:NN \interactionmode       \etex_interactionmode:D
+  \name_primitive:NN \lastnodetype          \etex_lastnodetype:D
+  \name_primitive:NN \iffontchar            \etex_iffontchar:D
+  \name_primitive:NN \fontcharht            \etex_fontcharht:D
+  \name_primitive:NN \fontchardp            \etex_fontchardp:D
+  \name_primitive:NN \fontcharwd            \etex_fontcharwd:D
+  \name_primitive:NN \fontcharic            \etex_fontcharic:D
+  \name_primitive:NN \parshapeindent        \etex_parshapeindent:D
+  \name_primitive:NN \parshapelength        \etex_parshapelength:D
+  \name_primitive:NN \parshapedimen         \etex_parshapedimen:D
+  \name_primitive:NN \numexpr               \etex_numexpr:D
+  \name_primitive:NN \dimexpr               \etex_dimexpr:D
+  \name_primitive:NN \glueexpr              \etex_glueexpr:D
+  \name_primitive:NN \muexpr                \etex_muexpr:D
+  \name_primitive:NN \gluestretch           \etex_gluestretch:D
+  \name_primitive:NN \glueshrink            \etex_glueshrink:D
+  \name_primitive:NN \gluestretchorder      \etex_gluestretchorder:D
+  \name_primitive:NN \glueshrinkorder       \etex_glueshrinkorder:D
+  \name_primitive:NN \gluetomu              \etex_gluetomu:D
+  \name_primitive:NN \mutoglue              \etex_mutoglue:D
+  \name_primitive:NN \lastlinefit           \etex_lastlinefit:D
+  \name_primitive:NN \interlinepenalties    \etex_interlinepenalties:D
+  \name_primitive:NN \clubpenalties         \etex_clubpenalties:D
+  \name_primitive:NN \widowpenalties        \etex_widowpenalties:D
+  \name_primitive:NN \displaywidowpenalties \etex_displaywidowpenalties:D
+  \name_primitive:NN \middle                \etex_middle:D
+  \name_primitive:NN \savinghyphcodes       \etex_savinghyphcodes:D
+  \name_primitive:NN \savingvdiscards       \etex_savingvdiscards:D
+  \name_primitive:NN \pagediscards          \etex_pagediscards:D
+  \name_primitive:NN \splitdiscards         \etex_splitdiscards:D
+  \name_primitive:NN \TeXXETstate           \etex_TeXXETstate:D
+  \name_primitive:NN \beginL                \etex_beginL:D
+  \name_primitive:NN \endL                  \etex_endL:D
+  \name_primitive:NN \beginR                \etex_beginR:D
+  \name_primitive:NN \endR                  \etex_endR:D
+  \name_primitive:NN \predisplaydirection   \etex_predisplaydirection:D
+  \name_primitive:NN \everyeof              \etex_everyeof:D
+  \name_primitive:NN \protected             \etex_protected:D
+%    \end{macrocode}
+% The newer primitives are more complex: there are an awful lot of them,
+% and we don't use them all at the moment. So the following is selective.
+% In the case of the \pdfTeX{} primitives, we retain |pdf| at the start
+% of the names \emph{only} for directly PDF-related primitives, as
+% there are a lot of \pdfTeX{} primitives that start \cs{pdf\ldots} but
+% are not related to PDF output. These ones re
+%    \begin{macrocode}
+  \name_primitive:NN \pdfcreationdate       \pdftex_pdfcreationdate:D
+  \name_primitive:NN \pdfcolorstack         \pdftex_pdfcolorstack:D
+  \name_primitive:NN \pdfcompresslevel      \pdftex_pdfcompresslevel:D
+  \name_primitive:NN \pdfdecimaldigits      \pdftex_pdfdecimaldigits:D
+  \name_primitive:NN \pdfhorigin            \pdftex_pdfhorigin:D
+  \name_primitive:NN \pdfinfo               \pdftex_pdfinfo:D
+  \name_primitive:NN \pdfliteral            \pdftex_pdfliteral:D
+  \name_primitive:NN \pdfminorversion       \pdftex_pdfminorversion:D
+  \name_primitive:NN \pdfobjcompresslevel   \pdftex_pdfobjcompresslevel:D
+  \name_primitive:NN \pdfoutput             \pdftex_pdfoutput:D
+  \name_primitive:NN \pdfrestore            \pdftex_pdfrestore:D
+  \name_primitive:NN \pdfsave               \pdftex_pdfsave:D
+  \name_primitive:NN \pdfsetmatrix          \pdftex_pdfsetmatrix:D
+  \name_primitive:NN \pdfpkresolution       \pdftex_pdfpkresolution:D
+  \name_primitive:NN \pdftexrevision        \pdftex_pdftextrevision:D
+  \name_primitive:NN \pdfvorigin            \pdftex_pdfvorigin:D
+%    \end{macrocode}
+% While these are not.
+%    \begin{macrocode}
+  \name_primitive:NN \pdfstrcmp             \pdftex_strcmp:D
+%    \end{macrocode}
+% \XeTeX{}-specific primitives.
+%    \begin{macrocode}
+  \name_primitive:NN \XeTeXversion          \xetex_XeTeXversion:D
+%    \end{macrocode}
+% Primitives from \LuaTeX.
+%    \begin{macrocode}
+  \name_primitive:NN \catcodetable          \luatex_catcodetable:D
+  \name_primitive:NN \directlua             \luatex_directlua:D
+  \name_primitive:NN \initcatcodetable      \luatex_initcatcodetable:D
+  \name_primitive:NN \latelua               \luatex_latelua:D
+  \name_primitive:NN \luatexversion         \luatex_luatexversion:D
+  \name_primitive:NN \savecatcodetable      \luatex_savecatcodetable:D
+%    \end{macrocode}
+% The job is done: close the group (using the primitive renamed!).
+%    \begin{macrocode}
+\tex_endgroup:D
+%    \end{macrocode}
+%
+% \LaTeXe{} will have moved a few primitives, so these are sorted out.
+%    \begin{macrocode}
+%<*package>
+\tex_let:D \tex_end:D               \@@end
+\tex_let:D \tex_everydisplay:D      \frozen@everydisplay
+\tex_let:D \tex_everymath:D         \frozen@everymath
+\tex_let:D \tex_hyphen:D            \@@hyph
+\tex_let:D \tex_input:D             \@@input
+\tex_let:D \tex_italic_correction:D \@@italiccorr
+\tex_let:D \tex_underline:D         \@@underline
+%    \end{macrocode}
+% That is also true for the \pkg{luatex} package for \LaTeXe{}.
+%    \begin{macrocode}
+\tex_let:D \luatex_catcodetable:D      \luatexcatcodetable
+\tex_let:D \luatex_initcatcodetable:D  \luatexinitcatcodetable
+\tex_let:D \luatex_latelua:D           \luatexlatelua
+\tex_let:D \luatex_savecatcodetable:D  \luatexsavecatcodetable
+%</package>
+%    \end{macrocode}
+%
+%    \begin{macrocode}
+%</initex|package>
+%    \end{macrocode}
+%
+% \end{implementation}
+%
+% \PrintIndex
\ No newline at end of file
diff --git a/Master/texmf-dist/source/latex/l3kernel/l3prg.dtx b/Master/texmf-dist/source/latex/l3kernel/l3prg.dtx
new file mode 100644
index 00000000000..b4cf9368f2a
--- /dev/null
+++ b/Master/texmf-dist/source/latex/l3kernel/l3prg.dtx
@@ -0,0 +1,2046 @@
+% \iffalse meta-comment
+%
+%% File: l3prg.dtx Copyright (C) 2005-2011 The LaTeX3 Project
+%%
+%% It may be distributed and/or modified under the conditions of the
+%% LaTeX Project Public License (LPPL), either version 1.3c of this
+%% license or (at your option) any later version.  The latest version
+%% of this license is in the file
+%%
+%%    http://www.latex-project.org/lppl.txt
+%%
+%% This file is part of the "expl3 bundle" (The Work in LPPL)
+%% and all files in that bundle must be distributed together.
+%%
+%% The released version of this bundle is available from CTAN.
+%%
+%% -----------------------------------------------------------------------
+%%
+%% The development version of the bundle can be found at
+%%
+%%    http://www.latex-project.org/svnroot/experimental/trunk/
+%%
+%% for those people who are interested.
+%%
+%%%%%%%%%%%
+%% NOTE: %%
+%%%%%%%%%%%
+%%
+%%   Snapshots taken from the repository represent work in progress and may
+%%   not work or may contain conflicting material!  We therefore ask
+%%   people _not_ to put them into distributions, archives, etc. without
+%%   prior consultation with the LaTeX3 Project.
+%%
+%% -----------------------------------------------------------------------
+%
+%<*driver|package>
+\RequirePackage{l3names}
+\GetIdInfo$Id: l3prg.dtx 2495 2011-07-06 16:57:08Z bruno $
+  {L3 Experimental control structures}
+%</driver|package>
+%<*driver>
+\documentclass[full]{l3doc}
+\begin{document}
+  \DocInput{\jobname.dtx}
+\end{document}
+%</driver>
+% \fi
+%
+% \title{^^A
+%   The \pkg{l3prg} package\\ Control structures^^A
+%   \thanks{This file describes v\ExplFileVersion,
+%      last revised \ExplFileDate.}^^A
+% }
+%
+% \author{^^A
+%  The \LaTeX3 Project\thanks
+%    {^^A
+%      E-mail:
+%        \href{mailto:latex-team@latex-project.org}
+%          {latex-team@latex-project.org}^^A
+%    }^^A
+% }
+%
+% \date{Released \ExplFileDate}
+%
+% \maketitle
+%
+% \begin{documentation}
+%
+% Conditional processing in \LaTeX3 is defined as something that
+% performs a series of tests, possibly involving assignments and
+% calling other functions that do not read further ahead in the input
+% stream. After processing the input, a \emph{state} is returned. The
+% typical states returned are \meta{true} and \meta{false} but other
+% states are possible, say an \meta{error} state for erroneous
+% input, \emph{e.g.}, text as input in a function comparing integers.
+%
+% \LaTeX3 has two primary forms of conditional flow processing based
+% on these states. One type is predicate functions that turn the
+% returned state into a boolean \meta{true} or \meta{false}. For
+% example, the function |\cs_if_free_p:N| checks whether the control
+% sequence given as its argument is free and then returns the boolean
+% \meta{true} or \meta{false} values to be used in testing with
+% |\if_predicate:w| or in functions to be described below.  The other type
+% is the kind of functions choosing a particular argument from the
+% input stream based on the result of the testing as in
+% |\cs_if_free:NTF| which also takes one argument (the |N|) and then
+% executes either \meta{true} or \meta{false} depending on the
+% result. Important to note here is that the arguments are executed
+% after exiting the underlying |\if...\fi:| structure
+%
+% \section{Defining a set of conditional functions}
+%
+% \begin{function}
+%   {
+%     \prg_new_conditional:Npnn, \prg_set_conditional:Npnn,
+%     \prg_new_conditional:Nnn,  \prg_set_conditional:Nnn
+%    }
+%   \begin{syntax}
+%     \cs{prg_set_conditional:Npnn} \cs{\meta{name}:\meta{arg spec}}
+%     ~~\meta{parameters} \Arg{conditions} \Arg{code}
+%     \cs{prg_set_conditional:Nnn} \cs{\meta{name}:\meta{arg spec}}
+%     ~~\Arg{conditions} \Arg{code}
+%   \end{syntax}
+%   These functions creates a family of conditionals using the same
+%   \Arg{code} to perform the test created. The \texttt{new} version will
+%   check for existing definitions (\emph{cf.}~\cs{cs_new:Npn}) whereas
+%   the \texttt{set} version will not (\emph{cf.}~\cs{cs_set:Npn}). The
+%   conditionals created are depended on the comma-separated list of
+%   \meta{conditions}, which should be one or more of \texttt{p},
+%   \texttt{T}, \texttt{F} and \texttt{TF}. The conditionals are then
+%   defined in the obvious way as:
+%   \begin{itemize}
+%     \item \cs{\meta{name}_p:\meta{arg spec}}, a predicate function
+%       which will supply either a logical \texttt{true} or
+%       logical \texttt{false}. This function is intended for use
+%       in cases where one or more logical tests are combined to
+%       lead to a final outcome.
+%     \item \cs{\meta{name}:\meta{arg spec}T}, a function with one
+%       more argument than the original \meta{arg spec} demands. The
+%       \meta{true branch} code in this additional argument will be
+%       left on the input stream only if the test is \texttt{true}.
+%     \item \cs{\meta{name}:\meta{arg spec}F}, a function with one
+%       more argument than the original \meta{arg spec} demands. The
+%       \meta{false branch} code in this additional argument will be
+%       left on the input stream only if the test is \texttt{false}.
+%     \item \cs{\meta{name}:\meta{arg spec}TF} , a function with two
+%       more argument than the original \meta{arg spec} demands. The
+%       \meta{true branch} code in the first additional argument will
+%       be left on the input stream if the test is \texttt{true}, while
+%       the \meta{false branch} code in the second argument will be
+%       left on the input stream if the test is \texttt{false}.
+%   \end{itemize}
+%   The \meta{code} of the test may use \meta{parameters} as specified
+%   by the second argument to \cs{prg_set_conditional:Npnn}: this should
+%   match the \meta{argument specification} but this is not enforced.
+%   The |Nnn| versions infer the number of arguments from the argument
+%   specification given (\emph{cf.}~\cs{cs_new:Nn}, \emph{etc.}).
+%   Within the \meta{code}, the functions \cs{prg_return_true:} and
+%   \cs{prg_return_false:} are used to indicate the logical outcomes of
+%   the test. If \meta{code} is expandable then
+%   \cs{prg_set_conditional:Npnn} will generate a family of conditionals
+%   which are also expandable. All of the functions are created globally.
+%
+%   An example can easily clarify matters here:
+%   \begin{verbatim}
+%     \prg_set_conditional:Nnn \foo_if_bar:NN { p , T , TF }
+%       {
+%         \if_meaning:w \l_tmpa_tl #1
+%           \prg_return_true:
+%         \else:
+%           \if_meaning:w \l_tmpa_tl #2
+%             \prg_return_true:
+%           \else:
+%             \prg_return_false:
+%          \fi:
+%        \fi:
+%      }
+%   \end{verbatim}
+%   This defines the function |\foo_if_bar_p:NN|, |\foo_if_bar:NNTF|,
+%   |\foo_if_bar:NNT| but not |\foo_if_bar:NNF| (because |F| is missing from
+%   the \meta{conds} list). The return statements
+%   take care of resolving the remaining |\else:| and |\fi:| before
+%   returning the state. There must be a return statement for each
+%   branch, failing to do so will result in an error if that branch is
+%   executed.
+% \end{function}
+%
+% \begin{function}
+%   {
+%     \prg_new_protected_conditional:Npnn, \prg_set_protected_conditional:Npnn,
+%     \prg_new_protected_conditional:Nnn,  \prg_set_protected_conditional:Nnn
+%   }
+%   \begin{syntax}
+%     \cs{prg_set_protected_conditional:Npnn}
+%     ~~\cs{\meta{name}:\meta{arg spec}} \meta{parameters}
+%     ~~\meta{conditions} \Arg{code}
+%     \cs{prg_set_protected_conditional:Nnn}
+%     ~~\cs{\meta{name}:\meta{arg spec}} \meta{conditions} \Arg{code}
+%   \end{syntax}
+%   These functions creates a family of conditionals using the same
+%   \Arg{code} to perform the test created. The \texttt{new} version will
+%   check for existing definitions (\emph{cf.}~\cs{cs_new:Npn}) whereas
+%   the \texttt{set} version will not (\emph{cf.}~\cs{cs_set:Npn}). The
+%   conditionals created are depended on the comma-separated list of
+%   \meta{conditions}, which should be one or more of \texttt{T},
+%   \texttt{F} and \texttt{TF}. The conditionals are then defined in the
+%   obvious way as:
+%   \begin{itemize}
+%     \item \cs{\meta{name}:\meta{arg spec}T}, a function with one
+%       more argument than the original \meta{arg spec} demands. The
+%       \meta{true branch} code in this additional argument will be
+%       left on the input stream only if the test is \texttt{true}.
+%     \item \cs{\meta{name}:\meta{arg spec}F}, a function with one
+%       more argument than the original \meta{arg spec} demands. The
+%       \meta{false branch} code in this additional argument will be
+%       left on the input stream only if the test is \texttt{false}.
+%     \item \cs{\meta{name}:\meta{arg spec}TF} , a function with two
+%       more argument than the original \meta{arg spec} demands. The
+%       \meta{true branch} code in the first additional argument will
+%       be left on the input stream if the test is \texttt{true}, while
+%       the \meta{false branch} code in the second argument will be
+%       left on the input stream if the test is \texttt{false}.
+%   \end{itemize}
+%   The \meta{code} of the test may use \meta{parameters} as specified
+%   by the second argument to \cs{prg_set_conditional:Npn}: this should
+%   match the \meta{argument specification} but this is not enforced.
+%   The |Nnn| versions infer the number of arguments from the argument
+%   specification given (\emph{cf.}~\cs{cs_new:Nn}, \emph{etc.}).
+%   Within the \meta{code}, the functions \cs{prg_return_true:} and
+%   \cs{prg_return_false:} are used to indicate the logical outcomes of
+%   the test. \cs{prg_set_protected_conditional:Npn} will generate
+%   a family of protected conditional functions, and so \meta{code}
+%   does not need to be expandable. All of the functions are created
+%   globally.
+%\end{function}
+%
+% \begin{function}{\prg_new_eq_conditional:NN, \prg_set_eq_conditional:NN}
+%   \begin{syntax}
+%     \cs{prg_new_eq_conditional:NN}
+%     ~~\cs{\meta{name1}:\meta{arg spec1}} \cs{\meta{name2}:\meta{arg spec2}}
+%   \end{syntax}
+%   These will set the definitions of the functions
+%   \begin{itemize}
+%     \item \cs{\meta{name1}_p:\meta{arg spec1}}
+%     \item \cs{\meta{name1}:\meta{arg spec1}T}
+%     \item \cs{\meta{name1}:\meta{arg spec1}F}
+%     \item \cs{\meta{name1}:\meta{arg spec1}TF}
+%   \end{itemize}
+%   equal to those for
+%   \begin{itemize}
+%     \item \cs{\meta{name2}_p:\meta{arg spec2}}
+%     \item \cs{\meta{name2}:\meta{arg spec2}T}
+%     \item \cs{\meta{name2}:\meta{arg spec2}F}
+%     \item \cs{\meta{name2}:\meta{arg spec2}TF}
+%   \end{itemize}
+%   In most cases, the two \meta{arg specs} will be identical, although
+%   this is not enforced. In the case of the \texttt{new} function, a
+%   check is made for any existing definitions for \meta{name1}. The
+%   functions are set globally.
+% \end{function}
+%
+% \begin{function}[EXP]{\prg_return_true:, \prg_return_false:}
+%   \begin{syntax}
+%     \cs{prg_return_true:}
+%     \cs{prg_return_false:}
+%   \end{syntax}
+%   These functions define the logical state at the end of a conditional.
+%   As such, they should appear within the code for a conditional
+%   statement generated by \cs{prg_set_conditional:Npnn}, \emph{etc}.
+% \end{function}
+%
+% \section{The boolean data type}
+%
+% This section describes a boolean data type which is closely
+% connected to conditional processing as sometimes you want to
+% execute some code depending on the value of a switch
+% (\emph{e.g.},~draft/final) and other times you perhaps want to use it as a
+% predicate function in an |\if_predicate:w| test. The problem of the
+% primitive |\if_false:| and |\if_true:| tokens is that it is not
+% always safe to pass them around as they may interfere with scanning
+% for termination of primitive conditional processing. Therefore, we
+% employ two canonical booleans: |\c_true_bool| or
+% |\c_false_bool|. Besides preventing problems as described above, it
+% also allows us to implement a simple boolean parser supporting the
+% logical operations And, Or, Not, \emph{etc.}\ which can then be used on
+% both the boolean type and predicate functions.
+%
+% All conditional |\bool_| functions are expandable and expect the
+% input to also be fully expandable (which will generally mean being
+% constructed from predicate functions, possibly nested).
+%
+% \begin{function}{\bool_new:N, \bool_new:c}
+%   \begin{syntax}
+%     \cs{bool_new:N} \meta{boolean}
+%   \end{syntax}
+%   Creates a new \meta{boolean} or raises an error if the
+%   name is already taken. The declaration is global. The
+%   \meta{boolean} will initially be \texttt{false}.
+% \end{function}
+%
+% \begin{function}{\bool_set_false:N, \bool_set_false:c}
+%   \begin{syntax}
+%     \cs{bool_set_false:N} \meta{boolean}
+%   \end{syntax}
+%   Sets \meta{boolean} logically \texttt{false} within the current
+%   \TeX{} group.
+% \end{function}
+%
+% \begin{function}{\bool_gset_false:N, \bool_gset_false:c}
+%   \begin{syntax}
+%     \cs{bool_sget_false:N} \meta{boolean}
+%   \end{syntax}
+%   Sets \meta{boolean} logically \texttt{false} globally.
+% \end{function}
+%
+% \begin{function}{\bool_set_true:N, \bool_set_true:c}
+%   \begin{syntax}
+%     \cs{bool_set_true:N} \meta{boolean}
+%   \end{syntax}
+%   Sets \meta{boolean} logically \texttt{true} within the current
+%   \TeX{} group.
+% \end{function}
+%
+% \begin{function}{\bool_gset_true:N, \bool_gset_true:c}
+%   \begin{syntax}
+%     \cs{bool_gset_true:N} \meta{boolean}
+%   \end{syntax}
+%   Sets \meta{boolean} logically \texttt{true} globally.
+% \end{function}
+%
+% \begin{function}
+%   {\bool_set_eq:NN, \bool_set_eq:cN, \bool_set_eq:Nc, \bool_set_eq:cc}
+%   \begin{syntax}
+%     \cs{bool_set_eq:NN} \meta{boolean1} \meta{boolean2}
+%   \end{syntax}
+%   Sets the content of \meta{boolean1} equal to that of \meta{boolean2}.
+%   This assignment is restricted to the current \TeX{} group level.
+% \end{function}
+%
+% \begin{function}
+%   {\bool_gset_eq:NN, \bool_gset_eq:cN, \bool_gset_eq:Nc, \bool_gset_eq:cc}
+%   \begin{syntax}
+%     \cs{bool_gset_eq:NN} \meta{boolean1} \meta{boolean2}
+%   \end{syntax}
+%   Sets the content of \meta{boolean1} equal to that of \meta{boolean2}.
+%   This assignment is global and so is not limited by the current
+%   \TeX{} group level.
+% \end{function}
+%
+% \begin{function}{\bool_set:Nn, \bool_set:cn}
+%   \begin{syntax}
+%     \cs{bool_set:Nn} \meta{boolean} \Arg{boolexpr}
+%   \end{syntax}
+%   Evaluates the \meta{boolean expression} as described for
+%   \cs{bool_if:n(TF)}, and sets the \meta{boolean} variable to
+%   the logical truth of this evaluation. This assignment is local.
+% \end{function}
+%
+% \begin{function}{\bool_gset:Nn, \bool_gset:cn}
+%   \begin{syntax}
+%     \cs{bool_gset:Nn} \meta{boolean} \Arg{boolexpr}
+%   \end{syntax}
+%   Evaluates the \meta{boolean expression} as described for
+%   \cs{bool_if:n(TF)}, and sets the \meta{boolean} variable to
+%   the logical truth of this evaluation. This assignment is global.
+% \end{function}
+%
+% \begin{function}[EXP,pTF]{\bool_if:N, \bool_if:c}
+%   \begin{syntax}
+%     \cs{bool_if_p:N} \Arg{boolean}
+%     \cs{bool_if:NTF} \Arg{boolean} \Arg{true code} \Arg{false code}
+%   \end{syntax}
+%   Tests the current truth of \meta{boolean}, and continues expansion
+%   based on this result. The branching versions then leave either
+%   \meta{true code} or \meta{false code} in the input stream, as
+%   appropriate to the truth of the test and the variant of the function
+%   chosen. The logical truth of the test is left in the input stream by
+%   the predicate version.
+% \end{function}
+%
+% \begin{variable}{\l_tmpa_bool}
+%   A scratch boolean for local assignment. It is never used by
+%   the kernel code, and so is safe for use with any \LaTeX3-defined
+%   function. However, it may be overwritten by other non-kernel
+%   code and so should only be used for short-term storage.
+% \end{variable}
+%
+% \begin{variable}{\g_tmpa_bool}
+%   A scratch boolean for global assignment. It is never used by
+%   the kernel code, and so is safe for use with any \LaTeX3-defined
+%   function. However, it may be overwritten by other non-kernel
+%   code and so should only be used for short-term storage.
+% \end{variable}
+%
+% \section{Boolean expressions}
+%
+% As we have a boolean datatype and predicate functions returning
+% boolean \meta{true} or \meta{false} values, it seems only fitting
+% that we also provide a parser for \meta{boolean expressions}.
+%
+% A boolean expression is an expression which given input in the form
+% of predicate functions and boolean variables, return boolean
+% \meta{true} or \meta{false}. It supports the logical operations And,
+% Or and Not as the well-known infix operators |&&|, \verb"||" and |!|. In
+% addition to this, parentheses can be used to isolate
+% sub-expressions. For example,
+% \begin{verbatim}
+%     \int_compare_p:n { 1 = 1 } &&
+%       (
+%         \int_compare_p:n { 2 = 3 } ||
+%         \int_compare_p:n { 4 = 4 } ||
+%         \int_compare_p:n { 1 = \error } % is skipped
+%       ) &&
+%     ! ( \int_compare_p:n { 2 = 4 } )
+% \end{verbatim}
+% is a valid boolean expression. Note that minimal evaluation is
+% carried out whenever possible so that whenever a truth value cannot
+% be changed any more, the remaining tests within the current group
+% are skipped.
+%
+% \begin{function}[EXP,pTF]{\bool_if:n}
+%   \begin{syntax}
+%     \cs{bool_if_p:n} \Arg{boolean expression}
+%     \cs{bool_if:nTF} \Arg{boolean expression} \Arg{true code}
+%     ~~\Arg{false code}
+%   \end{syntax}
+%   Tests the current truth of \meta{boolean expression}, and
+%   continues expansion based on this result. The
+%   \meta{boolean expression} should consist of a series of predicates
+%   or boolean variables with the logical relationship between these
+%   defined using |&&| (\enquote{And}), \verb"||" (\enquote{Or}),
+%   |!| (\enquote{Not}) and parentheses. Minimal evaluation is used
+%   in the processing, so that once a result is defined there is
+%   not further expansion of the tests. For example
+%  \begin{verbatim}
+%     \bool_if_p:n
+%       {
+%        \int_compare_p:nNn { 1 } = { 1 }
+%        &&
+%        (
+%          \int_compare_p:nNn { 2 } = { 3 } ||
+%          \int_compare_p:nNn { 4 } = { 4 } ||
+%          \int_compare_p:nNn { 1 } = { \error } % is skipped
+%        )
+%        &&
+%        ! ( \int_compare_p:nNn { 2 } = { 4 } )
+%     }
+%   \end{verbatim}
+%   will be \texttt{true} and will not evaluate
+%   |\int_compare_p:nNn { 1 } = { \error }|. The logical Not applies to
+%   the next single predicate or group. As shown above, this means that
+%   any predicates requiring an argument have to be given within
+%   parentheses. The branching versions then leave either
+%   \meta{true code} or \meta{false code} in the input stream, as
+%   appropriate to the truth of the test and the variant of the function
+%   chosen. The logical truth of the test is left in the input stream by
+%   the predicate version.
+% \end{function}
+%
+% \begin{function}[EXP]{\bool_not_p:n}
+%   \begin{syntax}
+%      \cs{bool_not_p:n} \Arg{boolean expression}
+%   \end{syntax}
+%   Function version of |!(|\meta{boolean expression}|)| within a boolean
+%   expression.
+% \end{function}
+%
+% \begin{function}[EXP]{\bool_xor_p:nn}
+%   \begin{syntax}
+%     \cs{bool_xor_p:nn} \Arg{boolexpr1} \Arg{boolexpr1}
+%   \end{syntax}
+%   Implements an \enquote{exclusive or} operation between two boolean
+%   expressions. There is no infix operation for this logical
+%   operator.
+% \end{function}
+%
+% \section{Logical loops}
+%
+%  Loops using either boolean expressions or stored boolean values.
+%
+% \begin{function}[EXP]{\bool_until_do:Nn, \bool_until_do:cn}
+%   \begin{syntax}
+%     \cs{bool_until_do:Nn} \Arg{boolean} \Arg{code}
+%   \end{syntax}
+%   This function firsts checks the logical value of the \meta{boolean}.
+%   If it is \texttt{false} the \meta{code} is placed in the input stream
+%   and expanded. After the completion of the \meta{code} the truth
+%   of the \meta{boolean} is re-evaluated. The process will then loop
+%   until the \meta{boolean} is \texttt{true}.
+% \end{function}
+%
+% \begin{function}[EXP]{\bool_while_do:Nn, \bool_while_do:cn}
+%   \begin{syntax}
+%     \cs{bool_while_do:Nn} \Arg{boolean} \Arg{code}
+%   \end{syntax}
+%   This function firsts checks the logical value of the \meta{boolean}.
+%   If it is \texttt{true} the \meta{code} is placed in the input stream
+%   and expanded. After the completion of the \meta{code} the truth
+%   of the \meta{boolean} is re-evaluated. The process will then loop
+%   until the \meta{boolean} is \texttt{false}.
+% \end{function}
+%
+% \begin{function}[EXP]{\bool_until_do:nn}
+%   \begin{syntax}
+%     \cs{bool_until_do:nn} \Arg{boolean expression} \Arg{code}
+%   \end{syntax}
+%   This function firsts checks the logical value of the
+%   \meta{boolean expression} (as described for \cs{bool_if:nTF}).
+%   If it is \texttt{false} the \meta{code} is placed in the input stream
+%   and expanded. After the completion of the \meta{code} the truth
+%   of the \meta{boolean expression} is re-evaluated. The process will
+%   then loop until the \meta{boolean expression} is \texttt{true}.
+% \end{function}
+%
+% \begin{function}[EXP]{\bool_while_do:nn}
+%   \begin{syntax}
+%     \cs{bool_while_do:nn} \Arg{boolean expression} \Arg{code}
+%   \end{syntax}
+%   This function firsts checks the logical value of the
+%   \meta{boolean expression} (as described for \cs{bool_if:nTF}).
+%   If it is \texttt{true} the \meta{code} is placed in the input stream
+%   and expanded. After the completion of the \meta{code} the truth
+%   of the \meta{boolean expression} is re-evaluated. The process will
+%   then loop until the \meta{boolean expression} is \texttt{false}.
+% \end{function}
+%
+% \section{Switching by case}
+%
+% For cases where a number of cases need to be considered a family of
+% case-selecting functions are available.
+%
+% \begin{function}[EXP]{\prg_case_int:nnn}
+%   \begin{syntax}
+%     \cs{prg_case_int:nnn}
+%     ~~\Arg{test integer expression}
+%     ~~|{|
+%     ~~~~\Arg{intexpr case1} \Arg{code case1}
+%     ~~~~\Arg{intexpr case2} \Arg{code case2}
+%     ~~~~\ldots
+%     ~~~~\Arg{intexpr case$_n$} \Arg{code case$_n$}
+%     ~~|}|
+%     ~~\Arg{else case}
+%   \end{syntax}
+%   This function evaluates the \meta{test integer expression} and
+%   compares this in turn to each of the
+%   \meta{integer expression cases}. If the two are equal then the
+%   associated \meta{code} is left in the input stream. If none of
+%   the tests are \texttt{true} then the \texttt{else code} will be
+%   left in the input stream. For example
+%   \begin{verbatim}
+%     \prg_case_int:nnn
+%       { 2 * 5 }
+%       {
+%         { 5 }       { Small }
+%         { 4 + 6 }   { Medium }
+%         { -2 * 10 } { Negative }
+%       }
+%       { No idea! }
+%    \end{verbatim}
+%   will leave \enquote{\texttt{Medium}} in the input stream.
+% \end{function}
+%
+% \begin{function}[EXP]{\prg_case_dim:nnn}
+%   \begin{syntax}
+%     \cs{prg_case_dim:nnn}
+%     ~~\Arg{test dimension expression}
+%     ~~|{|
+%     ~~~~\Arg{dimexpr case1} \Arg{code case1}
+%     ~~~~\Arg{dimexpr case2} \Arg{code case2}
+%     ~~~~\ldots
+%     ~~~~\Arg{dimexpr case$_n$} \Arg{code case$_n$}
+%     ~~|}|
+%     ~~\Arg{else case}
+%   \end{syntax}
+%   This function evaluates the \meta{test dimension expression} and
+%   compares this in turn to each of the
+%   \meta{dimension expression cases}. If the two are equal then the
+%   associated \meta{code} is left in the input stream. If none of
+%   the tests are \texttt{true} then the \texttt{else code} will be
+%   left in the input stream.
+% \end{function}
+%
+% \begin{function}[EXP]
+%   {\prg_case_str:nnn, \prg_case_str:onn, \prg_case_str:xxn}
+%   \begin{syntax}
+%     \cs{prg_case_str:nnn}
+%     ~~\Arg{test string}
+%     ~~|{|
+%     ~~~~\Arg{string case1} \Arg{code case1}
+%     ~~~~\Arg{string case2} \Arg{code case2}
+%     ~~~~\ldots
+%     ~~~~\Arg{string case$_n$} \Arg{code case$_n$}
+%     ~~|}|
+%     ~~\Arg{else case}
+%   \end{syntax}
+%   This function compares the \meta{test string} in turn with each
+%   of the \meta{string cases}. If the two are equal (as described for
+%   \cs{str_if_eq:nnTF} then the
+%   associated \meta{code} is left in the input stream. If none of
+%   the tests are \texttt{true} then the \texttt{else code} will be
+%   left in the input stream. The |xx| variant is fully expandable,
+%   in the same way as the underlying \cs{str_if_eq:xxTF} test.
+% \end{function}
+%
+% \begin{function}[EXP]{\prg_case_tl:Nnn, \prg_case_tl:cnn}
+%   \begin{syntax}
+%     \cs{prg_case_tl:Nnn}
+%     ~~\meta{test token list variable}
+%     ~~"{"
+%     ~~~~\meta{token list variable case1} \Arg{code case1}
+%     ~~~~\meta{token list variable case2} \Arg{code case2}
+%     ~~~~\ldots
+%     ~~~~\meta{token list variable case$_n$} \Arg{code case$_n$}
+%     ~~"}"
+%     ~~\Arg{else case}
+%   \end{syntax}
+%   This function compares the \meta{test token list variable} in turn
+%   with each of the \meta{token list variable cases}. If the two
+%   are equal (as described for
+%   \cs{tl_if_eq:nnTF}
+%   then the associated \meta{code} is left in the input
+%   stream. If none of the tests are \texttt{true} then the
+%   \texttt{else code} will be left in the input stream.
+% \end{function}
+%
+% \section{Producing $n$ copies}
+%
+% \begin{function}[EXP]{\prg_replicate:nn}
+%   \begin{syntax}
+%     \cs{prg_replicate:nn} \Arg{integer expression} \Arg{tokens}
+%   \end{syntax}
+%   Evaluates the \meta{integer expression} (which should be
+%   zero or positive) and creates the resulting number of copies
+%   of the \meta{tokens}. The function is both expandable and safe for
+%   nesting. It yields its result after two expansion steps.
+% \end{function}
+%
+% \begin{function}[EXP]{\prg_stepwise_function:nnnN}
+%   \begin{syntax}
+%     \cs{prg_stepwise_function:nnnN} \Arg{initial value} \Arg{step}
+%     ~~\Arg{final value} \meta{function}
+%   \end{syntax}
+%   This function first evaluates the \meta{initial value}, \meta{step}
+%   and \meta{final value}, all of which should be integer expressions.
+%   The \meta{function} is then placed in front of each \meta{value}
+%   from the \meta{initial value} to the \meta{final value} in turn
+%   (using \meta{step} between each \meta{value}). Thus \meta{function}
+%   should absorb one numerical argument. For example
+%   \begin{verbatim}
+%     \cs_set_nopar:Npn \my_func:n #1 { I~saw~#1 \\ }
+%     \prg_stepwise_function:nnnN { 1 } { 5 } { 1 } \my_func:n
+%   \end{verbatim}
+%   would print
+%   \begin{quote}
+%     I saw 1 \\
+%     I saw 2 \\
+%     I saw 3 \\
+%     I saw 4 \\
+%     I saw 5 \\
+%   \end{quote}
+% \end{function}
+%
+% \begin{function}{\prg_stepwise_inline:nnnn}
+%   \begin{syntax}
+%     \cs{prg_stepwise_inline:nnnn} \Arg{initial value} \Arg{step}
+%     ~~\Arg{final value} \Arg{code}
+%   \end{syntax}
+%   This function first evaluates the \meta{initial value}, \meta{step}
+%   and \meta{final value}, all of which should be integer expressions.
+%   The \meta{code} is then placed in front of each \meta{value}
+%   from the \meta{initial value} to the \meta{final value} in turn
+%   (using \meta{step} between each \meta{value}). Thus the \meta{code}
+%   should define a function of one argument (|#1|).
+% \end{function}
+%
+% \begin{function}{\prg_stepwise_variable:nnnn}
+%   \begin{syntax}
+%     \cs{prg_stepwise_inline:nnnn} \Arg{initial value} \Arg{step}
+%     ~~\Arg{final value} \meta{tl~var} \Arg{code}
+%   \end{syntax}
+%   This function first evaluates the \meta{initial value}, \meta{step}
+%   and \meta{final value}, all of which should be integer expressions.
+%   The \meta{code} is inserted into the input stream, with the
+%   \meta{tl~var} defined as the current \meta{value}. Thus the
+%   \meta{code} should make use of the \meta{tl~var}.
+% \end{function}
+%
+% \section{Detecting \TeX{}'s mode}
+%
+% \begin{function}[EXP,pTF]{\mode_if_horizontal:}
+%   \begin{syntax}
+%     \cs{mode_if_horizontal_p:}
+%     \cs{mode_if_horizontal:TF} \Arg{true code} \Arg{false code}
+%   \end{syntax}
+%   Detects if \TeX{} is currently in horizontal mode. The branching
+%   versions then leave either \meta{true code} or \meta{false code}
+%   in the input stream, as appropriate to the truth of the test and
+%   the variant of the function chosen. The logical truth of the test
+%   is left in the input stream by the predicate version.
+% \end{function}
+%
+% \begin{function}[EXP,pTF]{\mode_if_inner:}
+%   \begin{syntax}
+%     \cs{mode_if_inner_p:}
+%     \cs{mode_if_inner:TF} \Arg{true code} \Arg{false code}
+%   \end{syntax}
+%   Detects if \TeX{} is currently in inner mode. The branching
+%   versions then leave either \meta{true code} or \meta{false code}
+%   in the input stream, as appropriate to the truth of the test and
+%   the variant of the function chosen. The logical truth of the test
+%   is left in the input stream by the predicate version.
+% \end{function}
+%
+% \begin{function}[EXP,TF]{\mode_if_math:}
+%   \begin{syntax}
+%     \cs{mode_if_math:TF} \Arg{true code} \Arg{false code}
+%   \end{syntax}
+%   Detects if \TeX{} is currently in maths mode. The branching
+%   versions then leave either \meta{true code} or \meta{false code}
+%   in the input stream, as appropriate to the truth of the test and
+%   the variant of the function chosen.
+% \end{function}
+%
+% \begin{function}[EXP,pTF]{\mode_if_vertical:}
+%   \begin{syntax}
+%     \cs{mode_if_vertical_p:}
+%     \cs{mode_if_vertical:TF} \Arg{true code} \Arg{false code}
+%   \end{syntax}
+%   Detects if \TeX{} is currently in vertical mode. The branching
+%   versions then leave either \meta{true code} or \meta{false code}
+%   in the input stream, as appropriate to the truth of the test and
+%   the variant of the function chosen. The logical truth of the test
+%   is left in the input stream by the predicate version.
+% \end{function}
+%
+% \section{Internal programming functions}
+%
+% \begin{function}[EXP]{\group_align_safe_begin:, \group_align_safe_end:}
+%   \begin{syntax}
+%     \cs{group_align_safe_begin:}
+%     \ldots
+%     \cs{group_align_safe_end:}
+%   \end{syntax}
+%   These functions are used to enclose material in a \TeX{} alignment
+%   environment within a specially-constructed group. This group is
+%   designed in such a way that it does not add brace groups to the
+%   output but does act as a group for the |&| token inside
+%   \cs{tex_halign:D}. This is necessary to allow grabbing of tokens
+%   for testing purposes, as \TeX{} uses group level to determine the
+%   effect of alignment tokens. Without the special grouping, the use of
+%   a function such as \cs{peek_after:Nw} will result in a forbidden
+%   comparison of the internal \cs{endtemplate} token, yielding a
+%    fatal error. Each \cs{group_align_safe_begin:} must be matched by a
+%   \cs{group_align_safe_end:}, although this does not have to occur
+%   within the same function.
+% \end{function}
+%
+% \begin{function}{\scan_align_safe_stop:}
+%   \begin{syntax}
+%     \cs{scan_align_safe_stop:}
+%   \end{syntax}
+%   This function gets \TeX{} on the right track inside an alignment
+%   cell but without destroying any kerning.
+% \end{function}
+%
+% \begin{function}[EXP]{\prg_variable_get_scope:N}
+%   \begin{syntax}
+%     \cs{prg_variable_get_scope:N} \meta{variable}
+%   \end{syntax}
+%   Returns the scope (\texttt{g} for global, blank otherwise) for the
+%   \meta{variable}.
+% \end{function}
+%
+% \begin{function}[EXP]{\prg_variable_get_type:N}
+%   \begin{syntax}
+%     \cs{prg_variable_get_type:N} \meta{variable}
+%   \end{syntax}
+%   Returns the type of \meta{variable} (\texttt{tl}, \texttt{int},
+%   \emph{etc.})
+% \end{function}
+% 
+% \section{Experimental programmings functions}
+%
+% \begin{function}{\prg_quicksort:n}
+%   \begin{syntax}
+%    \cs{prg_quicksort:n} |{| \Arg{item~1} \Arg{item~2} \dots \Arg{item~n} |}|
+%   \end{syntax}
+%   Performs a quicksort on the token list. The comparisons are
+%   performed by the function \cs{prg_quicksort_compare:nnTF} which is up
+%   to the programmer to define. When the sorting process is over, all
+%   items are given as argument to the function
+%   \cs{prg_quicksort_function:n} which the programmer also controls.
+% \end{function}
+%
+%  \begin{function}{
+%                   \prg_quicksort_function:n |
+%                   \prg_quicksort_compare:nnTF
+%  }
+%  \begin{syntax}
+%    "\prg_quicksort_function:n" \Arg{element} \\
+%    "\prg_quicksort_compare:nnTF" \Arg{element 1} \Arg{element 2}\\
+%  \end{syntax}
+%  The two functions the programmer must define before calling
+%  |\prg_quicksort:n|. As an example we could define
+% \begin{quote}
+% |\cs_set_nopar:Npn\prg_quicksort_function:n #1{{#1}}|\\
+% |\cs_set_nopar:Npn\prg_quicksort_compare:nnTF #1#2#3#4 {\int_compare:nNnTF{#1}>{#2}}|
+% \end{quote}
+% Then the function call
+% \begin{quote}
+% |\prg_quicksort:n {876234520}|
+% \end{quote}
+% would return |{0}{2}{2}{3}{4}{5}{6}{7}{8}|. An alternative example
+% where one sorts a list of words, |\prg_quicksort_compare:nnTF| could
+% be defined as
+% \begin{quote}
+% |\cs_set_nopar:Npn\prg_quicksort_compare:nnTF #1#2 {|\\
+% |  \int_compare:nNnTF{\tl_compare:nn{#1}{#2}}>\c_zero }|
+% \end{quote}
+%
+%  \end{function}
+%
+% \end{documentation}
+%
+% \begin{implementation}
+%
+% \section{\pkg{l3prg} implementation}
+%
+% \TestFiles{m3prg001.lvt,m3prg002.lvt,m3prg003.lvt}
+%%
+%    \begin{macrocode}
+%<*initex|package>
+%    \end{macrocode}
+%
+%    \begin{macrocode}
+%<*package>
+\ProvidesExplPackage
+  {\ExplFileName}{\ExplFileDate}{\ExplFileVersion}{\ExplFileDescription}
+\package_check_loaded_expl:
+%</package>
+%    \end{macrocode}
+%
+% \subsection{Defining a set of conditional functions}
+%
+% \begin{macro}
+%   {
+%     \prg_set_conditional:Npnn,
+%     \prg_new_conditional:Npnn,
+%     \prg_set_protected_conditional:Npnn,
+%     \prg_new_protected_conditional:Npnn
+%   }
+% \begin{macro}
+%   {
+%     \prg_set_conditional:Nnn,
+%     \prg_new_conditional:Nnn,
+%     \prg_set_protected_conditional:Nnn,
+%     \prg_new_protected_conditional:Nnn
+%   }
+% \begin{macro}{\prg_set_eq_conditional:NNn, \prg_new_eq_conditional:NNn}
+% \begin{macro}{\prg_return_true:}
+% \TestMissing
+%   {This function is implicitly tested with all other conditionals!}
+% \begin{macro}{\prg_return_false:}
+% \TestMissing
+%   {This function is also implicitly tested with all other conditionals!}
+%   These are all defined in \pkg{l3basics}, as they are needed
+%   \enquote{early}. This is just a reminder that that is the case!
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \subsection{The boolean data type}
+%
+% \begin{macro}{\bool_new:N, \bool_new:c}
+% \UnitTested
+%   Boolean variables have to be initiated when they are created. Other
+%   than that there is not much to say here.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \bool_new:N #1 { \cs_new_eq:NN #1 \c_false_bool }
+\cs_generate_variant:Nn \bool_new:N { c }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}
+%   {
+%     \bool_set_true:N,   \bool_set_true:c,
+%     \bool_gset_true:N,  \bool_gset_true:c,
+%     \bool_set_false:N,  \bool_set_false:c,
+%     \bool_gset_false:N, \bool_gset_false:c
+%   }
+% \UnitTested
+%   Setting is already pretty easy.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \bool_set_true:N #1
+  { \cs_set_eq:NN #1 \c_true_bool }
+\cs_new_protected_nopar:Npn \bool_set_false:N #1
+  { \cs_set_eq:NN #1 \c_false_bool }
+\cs_new_protected_nopar:Npn \bool_gset_true:N #1
+  { \cs_gset_eq:NN #1 \c_true_bool }
+\cs_new_protected_nopar:Npn \bool_gset_false:N #1
+  { \cs_gset_eq:NN #1 \c_false_bool }
+\cs_generate_variant:Nn \bool_set_true:N   { c }
+\cs_generate_variant:Nn \bool_set_false:N  { c }
+\cs_generate_variant:Nn \bool_gset_true:N  { c }
+\cs_generate_variant:Nn \bool_gset_false:N { c }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}
+%   {
+%     \bool_set_eq:NN,  \bool_set_eq:cN,
+%     \bool_set_eq:Nc,  \bool_set_eq:cc,
+%     \bool_gset_eq:NN, \bool_gset_eq:cN,
+%     \bool_gset_eq:Nc, \bool_gset_eq:cc
+%   }
+% \UnitTested
+%   The usual copy code.
+%    \begin{macrocode}
+\cs_new_eq:NN \bool_set_eq:NN  \cs_set_eq:NN
+\cs_new_eq:NN \bool_set_eq:Nc  \cs_set_eq:Nc
+\cs_new_eq:NN \bool_set_eq:cN  \cs_set_eq:cN
+\cs_new_eq:NN \bool_set_eq:cc  \cs_set_eq:cc
+\cs_new_eq:NN \bool_gset_eq:NN \cs_gset_eq:NN
+\cs_new_eq:NN \bool_gset_eq:Nc \cs_gset_eq:Nc
+\cs_new_eq:NN \bool_gset_eq:cN \cs_gset_eq:cN
+\cs_new_eq:NN \bool_gset_eq:cc \cs_gset_eq:cc
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\bool_set:Nn,\bool_set:cn}
+% \begin{macro}{\bool_gset:Nn,\bool_gset:cn}
+%   This function evaluates a boolean expression and assigns the first
+%   argument the meaning |\c_true_bool| or |\c_false_bool|.
+%    \begin{macrocode}
+\cs_new:Npn \bool_set:Nn #1#2
+  { \tex_chardef:D #1 = \bool_if_p:n {#2} }
+\cs_new:Npn \bool_gset:Nn #1#2
+  { \tex_global:D \tex_chardef:D #1 = \bool_if_p:n {#2} }
+\cs_generate_variant:Nn \bool_set:Nn  { c }
+\cs_generate_variant:Nn \bool_gset:Nn { c }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}[pTF]{\bool_if:N, \bool_if:c}
+% \UnitTested
+%  Straight forward here. We could optimize here if we wanted to as
+%  the boolean can just be input directly.
+%    \begin{macrocode}
+\prg_new_conditional:Npnn \bool_if:N #1 { p , T , F , TF }
+  {
+    \if_bool:N #1
+      \prg_return_true:
+    \else:
+      \prg_return_false:
+    \fi:
+  }
+\cs_generate_variant:Nn \bool_if_p:N { c }
+\cs_generate_variant:Nn \bool_if:NT  { c }
+\cs_generate_variant:Nn \bool_if:NF  { c }
+\cs_generate_variant:Nn \bool_if:NTF { c }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{variable}{\l_tmpa_bool, \g_tmpa_bool}
+%    A few booleans just if you need them.
+%    \begin{macrocode}
+\bool_new:N \l_tmpa_bool
+\bool_new:N \g_tmpa_bool
+%    \end{macrocode}
+% \end{variable}
+%
+% \subsection{Boolean expressions}
+%
+% \begin{macro}[pTF]{\bool_if:n}
+% \UnitTested
+% \begin{macro}[aux]{\bool_get_next:N}
+% \begin{macro}[aux]{\bool_cleanup:N}
+% \begin{macro}[aux]{\bool_choose:NN}
+% \begin{macro}[aux]
+%   {
+%     bool_!:w,
+%     \bool_Not:w,
+%     \bool_Not:w,
+%     \bool_(:w,
+%     \bool_p:w,
+%     \bool_8_1:w,
+%     \bool_I_1:w,
+%     \bool_8_0:w,
+%     \bool_I_0:w,
+%     \bool_)_0:w,
+%     \bool_)_1:w,
+%     \bool_S_0:w,
+%     \bool_S_1:w
+% }
+% \begin{macro}[aux]
+%   {
+%     \bool_eval_skip_to_end:Nw, \bool_eval_skip_to_end_aux:Nw,
+%     \bool_eval_skip_to_end_aux_ii:Nw
+%  }
+%   Evaluating the truth value of a list of predicates is done using
+%   an input syntax somewhat similar to the one found in other
+%   programming languages with |(| and |)| for grouping, |!| for
+%   logical \enquote{Not}, |&&| for logical \enquote{And} and \verb"||"
+%   for logical \enquote{Or}. We shall use the terms Not, And, Or, Open and
+%   Close for  these operations.
+%
+%   Any expression is terminated by a Close operation. Evaluation
+%   happens from left to right in the following manner using a GetNext
+%   function:
+%   \begin{itemize}
+%     \item If an Open is seen, start evaluating a new expression using
+%       the Eval function and call GetNext again.
+%     \item If a Not is seen, insert a negating function (if-even in
+%       this case) and call GetNext.
+%     \item If none of the above, start evaluating a new expression by
+%       reinserting the token found (this is supposed to be a predicate
+%       function) in front of Eval.
+%   \end{itemize}
+%   The Eval function then contains a post-processing operation which
+%   grabs the instruction following the predicate. This is either And,
+%   Or or Close. In each case the truth value is used to determine
+%   where to go next. The following situations can arise:
+%   \begin{description}
+%     \item[\meta{true}And] Current truth value is true, logical And
+%       seen, continue with GetNext to examine truth value of next
+%       boolean (sub-)expression.
+%     \item[\meta{false}And] Current truth value is false, logical And
+%       seen, stop evaluating the predicates within this sub-expression
+%       and break to the nearest Close. Then return \meta{false}.
+%     \item[\meta{true}Or] Current truth value is true, logical Or
+%       seen, stop evaluating the predicates within this sub-expression
+%       and break to the nearest Close. Then return \meta{true}.
+%     \item[\meta{false}Or] Current truth value is false, logical Or
+%       seen, continue with GetNext to examine truth value of next
+%       boolean (sub-)expression.
+%     \item[\meta{true}Close] Current truth value is true, Close
+%       seen, return \meta{true}.
+%     \item[\meta{false}Close] Current truth value is false, Close
+%       seen, return \meta{false}.
+%   \end{description}
+%   We introduce an additional Stop operation with the following
+%   semantics:
+%   \begin{description}
+%     \item[\meta{true}Stop] Current truth value is true, return
+%       \meta{true}.
+%     \item[\meta{false}Stop] Current truth value is false, return
+%       \meta{false}.
+%   \end{description}
+%   The reasons for this follow below.
+%
+%   Now for how these works in practice. The canonical true and false
+%   values have numerical values $1$ and $0$ respectively. We evaluate
+%   this using the primitive |\int_value:w:D| operation. First we
+%   issue a |\group_align_safe_begin:| as we are using |&&| as syntax
+%   shorthand for the And operation and we need to hide it for \TeX{}.
+%   We also need to finish this special group before finally
+%   returning a |\c_true_bool| or |\c_false_bool| as there might
+%   otherwise be something left in front in the input stream. For
+%   this we call the Stop operation, denoted simply by a |S|
+%   following the last Close operation.
+%    \begin{macrocode}
+\prg_new_conditional:Npnn \bool_if:n #1 { T , F , TF }
+  {
+    \if_predicate:w \bool_if_p:n {#1}
+      \prg_return_true:
+    \else:
+      \prg_return_false:
+    \fi:
+  }
+\cs_new:Npn \bool_if_p:n #1
+  {
+    \group_align_safe_begin:
+    \bool_get_next:N ( #1 ) S
+  }
+%    \end{macrocode}
+%   The GetNext operation. We make it a switch: If not a |!| or |(|, we
+%   assume it is a predicate.
+%    \begin{macrocode}
+\cs_new:Npn \bool_get_next:N #1
+  {
+    \use:c
+      {
+        bool_
+        \if_meaning:w !#1 ! \else: \if_meaning:w (#1 ( \else: p \fi: \fi:
+        :w
+      }
+      #1
+  }
+%    \end{macrocode}
+%   This variant gets called when a Not has just been entered.
+%   It (eventually) results in a reversal of the logic of the directly
+%   following material.
+%    \begin{macrocode}
+\cs_new:Npn \bool_get_not_next:N #1
+  {
+    \use:c
+      {
+      bool_not_
+      \if_meaning:w !#1 ! \else: \if_meaning:w (#1 ( \else: p \fi: \fi:
+      :w
+      }
+        #1
+  }
+%    \end{macrocode}
+%   We need these later on to nullify the unity operation |!!|.
+%    \begin{macrocode}
+\cs_new:Npn \bool_get_next:NN #1#2 { \bool_get_next:N #2 }
+\cs_new:Npn \bool_get_not_next:NN #1#2 { \bool_get_not_next:N #2 }
+%    \end{macrocode}
+%   The Not operation. Discard the token read and reverse the truth
+%   value of the next expression if there
+%   are brackets; otherwise
+%   if we're coming up to a |!| then we don't need to reverse anything
+%   (but we then want to continue scanning ahead in case some fool has written
+%   |!!(...)|);
+%   otherwise we have a boolean that we can reverse here and now.
+%    \begin{macrocode}
+\cs_new:cpn { bool_!:w } #1#2
+  {
+    \if_meaning:w ( #2
+      \exp_after:wN \bool_Not:w
+    \else:
+      \if_meaning:w ! #2
+        \exp_after:wN \exp_after:wN \exp_after:wN \bool_get_next:NN
+      \else:
+        \exp_after:wN \exp_after:wN \exp_after:wN \bool_Not:N
+      \fi:
+    \fi:
+    #2
+  }
+%    \end{macrocode}
+%   Variant called when already inside a Not.
+%   Essentially the opposite of the above.
+%    \begin{macrocode}
+\cs_new:cpn { bool_not_!:w } #1#2
+  {
+    \if_meaning:w ( #2
+      \exp_after:wN \bool_not_Not:w
+    \else:
+      \if_meaning:w ! #2
+        \exp_after:wN \exp_after:wN \exp_after:wN \bool_get_not_next:NN
+      \else:
+        \exp_after:wN \exp_after:wN \exp_after:wN \bool_not_Not:N
+      \fi:
+    \fi:
+    #2
+  }
+%    \end{macrocode}
+%   These occur when processing |!(...)|. The idea is to use a variant
+%   of |\bool_get_next:N| that finishes its parsing with a logic reversal.
+%   Of course, the double logic reversal gets us back to where we started.
+%    \begin{macrocode}
+\cs_new:Npn \bool_Not:w { \exp_after:wN \int_value:w \bool_get_not_next:N }
+\cs_new:Npn \bool_not_Not:w { \exp_after:wN \int_value:w \bool_get_next:N }
+%    \end{macrocode}
+%   These occur when processing |!<bool>| and can be evaluated directly.
+%    \begin{macrocode}
+\cs_new:Npn \bool_Not:N #1
+  {
+    \exp_after:wN \bool_p:w
+    \if_meaning:w #1 \c_true_bool
+      \c_false_bool
+    \else:
+      \c_true_bool
+    \fi:
+  }
+\cs_new:Npn \bool_not_Not:N #1
+  {
+    \exp_after:wN \bool_p:w
+    \if_meaning:w #1 \c_true_bool
+      \c_true_bool
+    \else:
+      \c_false_bool
+    \fi:
+  }
+%    \end{macrocode}
+%   The Open operation. Discard the token read and start a
+%   sub-expression.
+%   |\bool_get_next:N| continues building up the logical expressions as usual;
+%   |\bool_not_cleanup:N| is what reverses the logic if we're inside |!(...)|.
+%    \begin{macrocode}
+\cs_new:cpn { bool_(:w } #1
+  { \exp_after:wN  \bool_cleanup:N \int_value:w \bool_get_next:N }
+\cs_new:cpn { bool_not_(:w } #1
+  { \exp_after:wN  \bool_not_cleanup:N \int_value:w \bool_get_next:N }
+%    \end{macrocode}
+%   Otherwise just evaluate the predicate and look for And, Or or Close
+%   afterwards.
+%    \begin{macrocode}
+\cs_new:cpn { bool_p:w } { \exp_after:wN \bool_cleanup:N \int_value:w  }
+\cs_new:cpn { bool_not_p:w } {\exp_after:wN \bool_not_cleanup:N \int_value:w }
+%    \end{macrocode}
+%   This cleanup function can be omitted once predicates return their
+%   true/false booleans outside the conditionals.
+%    \begin{macrocode}
+\cs_new:Npn \bool_cleanup:N #1
+  {
+    \exp_after:wN \bool_choose:NN \exp_after:wN #1
+    \int_to_roman:w - `\q
+  }
+\cs_new:Npn \bool_not_cleanup:N #1
+  {
+    \exp_after:wN \bool_not_choose:NN \exp_after:wN #1
+    \int_to_roman:w - `\q
+  }
+%    \end{macrocode}
+%   Branching the six way switch.
+%   Reversals should be reasonably straightforward.
+%    \begin{macrocode}
+\cs_new_nopar:Npn \bool_choose:NN #1#2 { \use:c { bool_ #2 _ #1 :w } }
+\cs_new_nopar:Npn \bool_not_choose:NN #1#2 { \use:c { bool_not_ #2 _ #1 :w } }
+%    \end{macrocode}
+%   Continues scanning. Must remove the second "&" or \verb"|".
+%    \begin{macrocode}
+\cs_new_nopar:cpn { bool_&_1:w } & { \bool_get_next:N }
+\cs_new_nopar:cpn { bool_|_0:w } | { \bool_get_next:N }
+\cs_new_nopar:cpn { bool_not_&_0:w } & { \bool_get_next:N }
+\cs_new_nopar:cpn { bool_not_|_1:w } | { \bool_get_next:N }
+%    \end{macrocode}
+%   Closing a group is just about returning the result. The Stop
+%   operation is similar except it closes the special alignment group
+%   before returning the boolean.
+%    \begin{macrocode}
+\cs_new_nopar:cpn { bool_)_0:w } { \c_false_bool }
+\cs_new_nopar:cpn { bool_)_1:w } { \c_true_bool }
+\cs_new_nopar:cpn { bool_not_)_0:w } { \c_true_bool }
+\cs_new_nopar:cpn { bool_not_)_1:w } { \c_false_bool }
+\cs_new_nopar:cpn { bool_S_0:w } { \group_align_safe_end: \c_false_bool }
+\cs_new_nopar:cpn { bool_S_1:w } { \group_align_safe_end: \c_true_bool }
+%    \end{macrocode}
+%   When the truth value has already been decided, we have to throw away
+%   the remainder of the current group as we are doing minimal
+%   evaluation. This is slightly tricky as there are no braces so we
+%   have to play match the |()| manually.
+%    \begin{macrocode}
+\cs_new_nopar:cpn { bool_&_0:w } & { \bool_eval_skip_to_end:Nw \c_false_bool }
+\cs_new_nopar:cpn { bool_|_1:w } | { \bool_eval_skip_to_end:Nw \c_true_bool }
+\cs_new_nopar:cpn { bool_not_&_1:w } &
+  { \bool_eval_skip_to_end:Nw \c_false_bool }
+\cs_new_nopar:cpn { bool_not_|_0:w } |
+  { \bool_eval_skip_to_end:Nw \c_true_bool }
+%    \end{macrocode}
+%   There is always at least one |)| waiting, namely the outer
+%   one. However, we are facing the problem that there may be more than
+%   one that need to be finished off and we have to detect the correct
+%   number of them. Here is a complicated example showing how this is
+%   done. After evaluating the following, we realize we must skip
+%   everything after the first And. Note the extra Close at the end.
+%   \begin{quote}
+%     |\c_false_bool  && ((abc) && xyz) && ((xyz) && (def)))|
+%   \end{quote}
+%   First read up to the first Close. This gives us the list we first
+%   read up until the first right parenthesis so we are looking at the
+%   token list
+%   \begin{quote}
+%     |((abc|
+%   \end{quote}
+%   This contains two Open markers so we must remove two groups. Since
+%   no evaluation of the contents is to be carried out, it doesn't
+%   matter how we remove the groups as long as we wind up with the
+%   correct result. We therefore first remove a |()| pair and what
+%   preceded the Open -- but leave the contents as it may contain Open
+%   tokens itself -- leaving
+%   \begin{quote}
+%     |(abc && xyz) && ((xyz) && (def)))|
+%   \end{quote}
+%   Another round of this gives us
+%   \begin{quote}
+%     |(abc && xyz|
+%   \end{quote}
+%   which still contains an Open so we remove another |()| pair, giving us
+%   \begin{quote}
+%     |abc && xyz && ((xyz) && (def)))|
+%   \end{quote}
+%   Again we read up to a Close and again find Open tokens:
+%   \begin{quote}
+%     |abc && xyz && ((xyz|
+%   \end{quote}
+%   Further reduction gives us
+%   \begin{quote}
+%     |(xyz && (def)))|
+%   \end{quote}
+%   and then
+%   \begin{quote}
+%     |(xyz && (def|
+%   \end{quote}
+%   with reduction to
+%   \begin{quote}
+%     |xyz && (def))|
+%   \end{quote}
+%   and ultimately we arrive at no Open tokens being skipped and we can
+%   finally close the group nicely.
+%    \begin{macrocode}
+%% (
+\cs_new:Npn \bool_eval_skip_to_end:Nw #1#2 )
+  {
+    \bool_eval_skip_to_end_aux:Nw #1#2 ( % )
+    \q_no_value \q_stop
+    {#2}
+  }
+%    \end{macrocode}
+%   If no right parenthesis, then |#3| is no_value and we are done, return
+%   the boolean |#1|.  If there is, we need to grab a |()| pair and then
+%   recurse
+%    \begin{macrocode}
+\cs_new:Npn \bool_eval_skip_to_end_aux:Nw #1#2 ( #3#4 \q_stop #5 % )
+  {
+    \quark_if_no_value:NTF #3
+    {#1}
+    { \bool_eval_skip_to_end_aux_ii:Nw #1 #5 }
+  }
+%    \end{macrocode}
+%   Keep the boolean, throw away anything up to the |(| as it is
+%   irrelevant, remove a |()| pair but remember to reinsert |#3| as it may
+%   contain |(| tokens!
+%    \begin{macrocode}
+\cs_new:Npn \bool_eval_skip_to_end_aux_ii:Nw #1#2 ( #3 )
+  { % (
+    \bool_eval_skip_to_end:Nw #1#3 )
+  }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\bool_not_p:n}
+% \UnitTested
+%   The Not variant just reverses the outcome of |\bool_if_p:n|. Can
+%   be optimized but this is nice and simple and according to the
+%   implementation plan. Not even particularly useful to have it when
+%   the infix notation is easier to use.
+%    \begin{macrocode}
+\cs_new:Npn \bool_not_p:n #1 { \bool_if_p:n { ! ( #1 ) } }
+%    \end{macrocode}
+%  \end{macro}
+%
+% \begin{macro}{\bool_xor_p:nn}
+% \UnitTested
+%    Exclusive or. If the boolean expressions have same truth value,
+%    return false, otherwise return true.
+%    \begin{macrocode}
+\cs_new:Npn \bool_xor_p:nn #1#2
+  {
+    \int_compare:nNnTF { \bool_if_p:n {#1} } = { \bool_if_p:n {#2} }
+      \c_false_bool
+      \c_true_bool
+  }
+%    \end{macrocode}
+% \end{macro}
+%
+% \subsection{Logical loops}
+%
+% \begin{macro}{\bool_while_do:Nn,\bool_while_do:cn}
+% \UnitTested
+% \begin{macro}{\bool_until_do:Nn,\bool_until_do:cn}
+% \UnitTested
+%   A |while| loop where the boolean is tested before executing the
+%   statement. The \enquote{while} version executes the code as long as the
+%   boolean is true; the \enquote{until} version executes the code as
+%   long as the boolean is false.
+%    \begin{macrocode}
+\cs_new:Npn \bool_while_do:Nn #1#2
+  { \bool_if:NT #1 { #2 \bool_while_do:Nn #1 {#2} } }
+\cs_new:Npn \bool_until_do:Nn #1#2
+  { \bool_if:NF #1 { #2 \bool_until_do:Nn #1 {#2} } }
+\cs_generate_variant:Nn \bool_while_do:Nn { c }
+\cs_generate_variant:Nn \bool_until_do:Nn { c }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\bool_do_while:Nn,\bool_do_while:cn}
+% \UnitTested
+% \begin{macro}{\bool_do_until:Nn,\bool_do_until:cn}
+% \UnitTested
+%   A |do-while| loop where the body is performed at least once and the
+%   boolean is tested after executing the body. Otherwise identical to
+%   the above functions.
+%    \begin{macrocode}
+\cs_new:Npn \bool_do_while:Nn #1#2
+  { #2 \bool_if:NT #1 { \bool_do_while:Nn #1 {#2} } }
+\cs_new:Npn \bool_do_until:Nn #1#2
+  { #2 \bool_if:NF #1 { \bool_do_until:Nn #1 {#2} } }
+\cs_generate_variant:Nn \bool_do_while:Nn { c }
+\cs_generate_variant:Nn \bool_do_until:Nn { c }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}
+%   {
+%     \bool_while_do:nn, \bool_do_while:nn ,
+%     \bool_until_do:nn, \bool_do_until:nn
+%   }
+%  \UnitTested
+%   Loop functions with the test either before or after the first body
+%   expansion.
+%    \begin{macrocode}
+\cs_new:Npn \bool_while_do:nn #1#2
+  {
+    \bool_if:nT {#1}
+      {
+        #2
+        \bool_while_do:nn {#1} {#2}
+      }
+  }
+\cs_new:Npn \bool_do_while:nn #1#2
+  {
+    #2
+    \bool_if:nT {#1} { \bool_do_while:nn {#1} {#2} }
+  }
+\cs_new:Npn \bool_until_do:nn #1#2
+  {
+    \bool_if:nF {#1}
+      {
+        #2
+        \bool_until_do:nn {#1} {#2}
+      }
+  }
+\cs_new:Npn \bool_do_until:nn #1#2
+  {
+    #2
+    \bool_if:nF {#1} { \bool_do_until:nn {#1} {#2}  }
+  }
+%    \end{macrocode}
+% \end{macro}
+%
+% \subsection{Switching by case}
+%
+% A family of functions to select one case of a number: the same ideas
+% are used for a number of different situations.
+%
+% \begin{macro}[aux]{\prg_case_end:nw}
+%   In all cases the end statement is the same. Here, |#1| will be the
+%   code needed, |#2| the other cases to throw away, including the
+% \enquote{else} case.
+%    \begin{macrocode}
+\cs_new_eq:NN \prg_case_end:nw \use_i_delimit_by_q_recursion_stop:nw
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\prg_case_int:nnn}
+% \UnitTested
+% \begin{macro}[aux]{\prg_case_int_aux:nnn,\prg_case_int_aux:nw}
+%   For integer cases, the first task to fully expand the check
+%   condition. After that, a loop is started to compare each possible
+%   value and stop if the test is true. The tested value is put at the
+%   end to ensure that there is necessarily a match, which will fire the
+% \enquote{else} pathway.
+%    \begin{macrocode}
+\cs_new:Npn \prg_case_int:nnn #1
+  { \exp_args:Nf \prg_case_int_aux:nnn { \int_eval:n {#1} } }
+\cs_new:Npn \prg_case_int_aux:nnn #1 #2 #3
+  { \prg_case_int_aux:nw {#1} #2 {#1} {#3} \q_recursion_stop }
+\cs_new:Npn \prg_case_int_aux:nw #1#2#3
+  {
+    \int_compare:nNnTF {#1} = {#2}
+      { \prg_case_end:nw {#3} }
+      { \prg_case_int_aux:nw {#1} }
+  }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\prg_case_dim:nnn}
+% \UnitTested
+% \begin{macro}[aux]{\prg_case_dim_aux:nnn,\prg_case_dim_aux:nw}
+%   The dimension function is the same, just a change of calculation
+%   method.
+%    \begin{macrocode}
+\cs_new:Npn \prg_case_dim:nnn #1
+  { \exp_args:Nf \prg_case_dim_aux:nnn { \dim_eval:n {#1} } }
+\cs_new:Npn \prg_case_dim_aux:nnn #1 #2 #3
+  { \prg_case_dim_aux:nw {#1} #2 {#1} {#3} \q_recursion_stop }
+\cs_new:Npn \prg_case_dim_aux:nw #1#2#3
+  {
+    \dim_compare:nNnTF {#1} = {#2}
+      { \prg_case_end:nw {#3} }
+      { \prg_case_dim_aux:nw {#1} }
+  }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\prg_case_str:nnn, \prg_case_str:onn, \prg_case_str:xxn}
+% \UnitTested
+% \begin{macro}[aux]{\prg_case_str_aux:nw, \prg_case_str_x_aux:nw}
+%   No calculations for strings, otherwise no surprises.
+%    \begin{macrocode}
+\cs_new:Npn \prg_case_str:nnn #1#2#3
+  { \prg_case_str_aux:nw {#1} #2 {#1} {#3} \q_recursion_stop }
+\cs_new:Npn \prg_case_str_aux:nw #1#2#3
+  {
+    \str_if_eq:nnTF {#1} {#2}
+      { \prg_case_end:nw {#3} }
+      { \prg_case_str_aux:nw {#1} }
+  }
+\cs_generate_variant:Nn \prg_case_str:nnn { o }
+\cs_new:Npn \prg_case_str:xxn #1#2#3
+  { \prg_case_str_x_aux:nw {#1} #2 {#1} {#3} \q_recursion_stop }
+\cs_new:Npn \prg_case_str_x_aux:nw #1#2#3
+  {
+    \str_if_eq:xxTF {#1} {#2}
+      { \prg_case_end:nw {#3} }
+      { \prg_case_str_aux:nw {#1} }
+  }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\prg_case_tl:Nnn, \prg_case_tl:cnn}
+% \UnitTested
+% \begin{macro}[aux]{\prg_case_tl_aux:Nw}
+%   Similar again, but this time with some variants.
+%    \begin{macrocode}
+\cs_new:Npn \prg_case_tl:Nnn #1#2#3
+  { \prg_case_tl_aux:Nw #1 #2 #1 {#3} \q_recursion_stop }
+\cs_new:Npn \prg_case_tl_aux:Nw #1#2#3
+  {
+    \tl_if_eq:NNTF #1 #2
+      { \prg_case_end:nw {#3} }
+      { \prg_case_tl_aux:Nw #1 }
+  }
+\cs_generate_variant:Nn \prg_case_tl:Nnn { c }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \subsection{Producing $n$ copies}
+%
+% \begin{macro}{\prg_replicate:nn}
+% \UnitTested
+% \begin{macro}[aux]{\prg_replicate_aux:N, \prg_replicate_first_aux:N}
+% \begin{macro}[aux]{\prg_replicate_}
+% \begin{macro}[aux]
+%   {
+%     \prg_replicate_0:n,
+%     \prg_replicate_1:n,
+%     \prg_replicate_2:n,
+%     \prg_replicate_3:n,
+%     \prg_replicate_4:n,
+%     \prg_replicate_5:n,
+%     \prg_replicate_6:n,
+%     \prg_replicate_7:n,
+%     \prg_replicate_8:n,
+%     \prg_replicate_9:n
+%   }
+% \begin{macro}[aux]
+%   {
+%     \prg_replicate_first_-:n,
+%     \prg_replicate_first_0:n,
+%     \prg_replicate_first_1:n,
+%     \prg_replicate_first_2:n,
+%     \prg_replicate_first_3:n,
+%     \prg_replicate_first_4:n,
+%     \prg_replicate_first_5:n,
+%     \prg_replicate_first_6:n,
+%     \prg_replicate_first_7:n,
+%     \prg_replicate_first_8:n,
+%     \prg_replicate_first_9:n
+%   }
+%   This function uses a cascading csname technique by David Kastrup
+%   (who else :-)
+%
+%   The idea is to make the input |25| result in first adding five, and
+%   then 20 copies of the code to be replicated. The technique uses
+%   cascading csnames which means that we start building several csnames
+%   so we end up with a list of functions to be called in reverse
+%   order. This is important here (and other places) because it means
+%   that we can for instance make the function that inserts five copies
+%   of something to also hand down ten to the next function in
+%   line. This is exactly what happens here: in the example with |25|
+%   then the next function is the one that inserts two copies but it
+%   sees the ten copies handed down by the previous function. In order
+%   to avoid the last function to insert say, $100$ copies of the original
+%   argument just to gobble them again we define separate functions to
+%   be inserted first. These functions also close the expansion of
+%   \cs{int_to_roman:w}, which ensures that \cs{prg_replicate:nn} only
+%   requires two steps of expansion.
+%
+%   This function has one flaw though: Since it constantly passes down
+%   ten copies of its previous argument it will severely affect the main
+%   memory once you start demanding hundreds of thousands of copies. Now
+%   I don't think this is a real limitation for any ordinary use, and if
+%   necessary, it is possible to write
+%   |\prg_replicate:nn{1000}{\prg_replicate:nn{1000}{|\meta{code}|}}|. An
+%   alternative approach is to create a string of |m|'s with
+%   \cs{int_to_roman:w} which can be done with just four macros but that
+%   method has its own problems since it can exhaust the string
+%   pool. Also, it is considerably slower than what we use here so the
+%   few extra csnames are well spent I would say.
+%    \begin{macrocode}
+\cs_new_nopar:Npn \prg_replicate:nn #1
+  {
+    \int_to_roman:w
+      \exp_after:wN \prg_replicate_first_aux:N
+        \int_value:w \int_eval:w #1 \int_eval_end:
+      \cs_end:
+  }
+\cs_new_nopar:Npn \prg_replicate_aux:N #1
+  { \cs:w prg_replicate_#1 :n \prg_replicate_aux:N }
+\cs_new_nopar:Npn \prg_replicate_first_aux:N #1
+  { \cs:w prg_replicate_first_ #1 :n \prg_replicate_aux:N }
+%    \end{macrocode}
+% \end{macro}
+% Then comes all the functions that do the hard work of inserting all
+% the copies.
+%    \begin{macrocode}
+\cs_new_nopar:Npn \prg_replicate_ :n #1 { \cs_end: }
+\cs_new:cpn { prg_replicate_0:n } #1 { \cs_end: {#1#1#1#1#1#1#1#1#1#1} }
+\cs_new:cpn { prg_replicate_1:n } #1 { \cs_end: {#1#1#1#1#1#1#1#1#1#1} #1 }
+\cs_new:cpn { prg_replicate_2:n } #1 { \cs_end: {#1#1#1#1#1#1#1#1#1#1} #1#1 }
+\cs_new:cpn { prg_replicate_3:n } #1
+  { \cs_end: {#1#1#1#1#1#1#1#1#1#1} #1#1#1 }
+\cs_new:cpn { prg_replicate_4:n } #1
+  { \cs_end: {#1#1#1#1#1#1#1#1#1#1} #1#1#1#1 }
+\cs_new:cpn { prg_replicate_5:n } #1
+  { \cs_end: {#1#1#1#1#1#1#1#1#1#1} #1#1#1#1#1 }
+\cs_new:cpn { prg_replicate_6:n } #1
+  { \cs_end: {#1#1#1#1#1#1#1#1#1#1} #1#1#1#1#1#1 }
+\cs_new:cpn { prg_replicate_7:n } #1
+  { \cs_end: {#1#1#1#1#1#1#1#1#1#1} #1#1#1#1#1#1#1 }
+\cs_new:cpn { prg_replicate_8:n } #1
+  { \cs_end: {#1#1#1#1#1#1#1#1#1#1} #1#1#1#1#1#1#1#1 }
+\cs_new:cpn { prg_replicate_9:n } #1
+  { \cs_end: {#1#1#1#1#1#1#1#1#1#1} #1#1#1#1#1#1#1#1#1 }
+%    \end{macrocode}
+%    Users shouldn't ask for something to be replicated once or even
+%    not at all but\dots
+%    \begin{macrocode}
+\cs_new:cpn { prg_replicate_first_-:n } #1 { \c_zero \negative_replication }
+\cs_new:cpn { prg_replicate_first_0:n } #1 { \c_zero }
+\cs_new:cpn { prg_replicate_first_1:n } #1 { \c_zero #1 }
+\cs_new:cpn { prg_replicate_first_2:n } #1 { \c_zero #1#1 }
+\cs_new:cpn { prg_replicate_first_3:n } #1 { \c_zero #1#1#1 }
+\cs_new:cpn { prg_replicate_first_4:n } #1 { \c_zero #1#1#1#1 }
+\cs_new:cpn { prg_replicate_first_5:n } #1 { \c_zero #1#1#1#1#1 }
+\cs_new:cpn { prg_replicate_first_6:n } #1 { \c_zero #1#1#1#1#1#1 }
+\cs_new:cpn { prg_replicate_first_7:n } #1 { \c_zero #1#1#1#1#1#1#1 }
+\cs_new:cpn { prg_replicate_first_8:n } #1 { \c_zero #1#1#1#1#1#1#1#1 }
+\cs_new:cpn { prg_replicate_first_9:n } #1 { \c_zero #1#1#1#1#1#1#1#1#1 }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\prg_stepwise_function:nnnN}
+% \begin{macro}[aux]
+%  {\prg_stepwise_function_incr:nnnN, \prg_stepwise_function_decr:nnnN}
+%   Repeating a function by steps fist needs a check on the direction
+%   of the steps. After that, do the function for the start value
+%   then step and loop around.
+%    \begin{macrocode}
+\cs_new:Npn \prg_stepwise_function:nnnN #1#2
+  {
+    \int_compare:nNnTF {#2} > { 0 }
+      { \exp_args:Nf \prg_stepwise_function_incr:nnnN }
+      { \exp_args:Nf \prg_stepwise_function_decr:nnnN }
+        { \int_eval:n {#1} } {#2}
+  }
+\cs_new:Npn \prg_stepwise_function_incr:nnnN #1#2#3#4
+  {
+    \int_compare:nNnF {#1} > {#3}
+      {
+        #4 {#1}
+        \exp_args:Nf \prg_stepwise_function_incr:nnnN
+          { \int_eval:n { #1 + #2 } } {#2} {#3} #4
+      }
+  }
+\cs_new:Npn \prg_stepwise_function_decr:nnnN #1#2#3#4
+  {
+    \int_compare:nNnF {#1}  < {#3}
+      {
+        #4 {#1}
+        \exp_args:Nf \prg_stepwise_function_decr:nnnN
+          { \int_eval:n { #1 + #2 } } {#2} {#3} #4
+      }
+  }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+%\begin{macro}[aux]{\g_prg_stepwise_level_int}
+% For nesting, the usual approach of using a counter.
+%    \begin{macrocode}
+\int_new:N \g_prg_stepwise_level_int
+%    \end{macrocode}
+%\end{macro}
+%
+%\begin{macro}{\prg_stepwise_inline:nnnn}
+%\begin{macro}[aux]
+%  {\prg_stepwise_inline_incr:Nnnn, \prg_stepwise_inline_decr:Nnnn}
+%   The approach here is similar but with a global integer required
+%   to make the nesting safe (as seen in other in line functions).
+%    \begin{macrocode}
+\cs_new_protected:Npn \prg_stepwise_inline:nnnn #1#2#3#4
+  {
+    \int_gincr:N \g_prg_stepwise_level_int
+    \cs_gset_nopar:cpn
+      { g_prg_stepwise_ \int_use:N \g_prg_stepwise_level_int :n }
+        ##1 {#4}
+    \int_compare:nNnTF {#2} > { 0 }
+      { \exp_args:Ncf \prg_stepwise_inline_incr:Nnnn }
+      { \exp_args:Ncf \prg_stepwise_inline_decr:Nnnn }
+        { g_prg_stepwise_ \int_use:N \g_prg_stepwise_level_int :n }
+        { \int_eval:n {#1} } {#2} {#3}
+    \int_gdecr:N \g_prg_stepwise_level_int
+  }
+\cs_new_protected:Npn \prg_stepwise_inline_incr:Nnnn #1#2#3#4
+  {
+    \int_compare:nNnF {#2} > {#4}
+      {
+        #1 {#2}
+        \exp_args:NNf \prg_stepwise_inline_incr:Nnnn #1
+          { \int_eval:n { #2 + #3 } } {#3} {#4}
+      }
+  }
+\cs_new_protected:Npn \prg_stepwise_inline_decr:Nnnn #1#2#3#4
+  {
+    \int_compare:nNnF {#2} < {#4}
+      {
+        #1 {#2}
+        \exp_args:NNf \prg_stepwise_inline_decr:Nnnn #1
+          { \int_eval:n { #2 + #3 } } {#3} {#4}
+      }
+  }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\prg_stepwise_variable:nnnNn}
+% \UnitTested
+%   A wrapper for the above.
+%    \begin{macrocode}
+\cs_new_protected:Npn \prg_stepwise_variable:nnnNn #1#2#3#4#5
+  {
+    \prg_stepwise_inline:nnnn {#1} {#2} {#3}
+      {
+        \tl_set:Nn #4 {##1}
+        #5
+      }
+  }
+%    \end{macrocode}
+% \end{macro}
+%
+% \subsection{Detecting \TeX{}'s mode}
+%
+% \begin{macro}[pTF]{\mode_if_vertical:}
+% \UnitTested
+%   For testing vertical mode. Strikes me here on the bus with David,
+%   that as long as we are just talking about returning true and
+%   false states, we can just use the primitive conditionals for this
+%   and gobbling the |\c_zero| in the input stream. However this
+%   requires knowledge of the implementation so we keep things nice
+%   and clean and use the return statements.
+%    \begin{macrocode}
+\prg_new_conditional:Npnn \mode_if_vertical: { p , T , F , TF }
+  { \if_mode_vertical: \prg_return_true: \else: \prg_return_false: \fi: }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}[pTF]{\mode_if_horizontal:}
+% \UnitTested
+%   For testing horizontal mode.
+%    \begin{macrocode}
+\prg_new_conditional:Npnn \mode_if_horizontal: { p , T , F , TF }
+  { \if_mode_horizontal: \prg_return_true: \else: \prg_return_false: \fi: }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}[pTF]{\mode_if_inner:}
+% \UnitTested
+%   For testing inner mode.
+%    \begin{macrocode}
+\prg_new_conditional:Npnn \mode_if_inner: { p , T , F , TF }
+  { \if_mode_inner: \prg_return_true: \else: \prg_return_false: \fi: }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}[pTF]{\mode_if_math:}
+% \UnitTested
+%   For testing math mode: without \cs{} things go wrong in alignments.
+%    \begin{macrocode}
+\prg_new_conditional:Npnn \mode_if_math: { p , T , F , TF }
+  {
+    \scan_align_safe_stop:
+    \if_mode_math: \prg_return_true: \else: \prg_return_false: \fi:
+  }
+%    \end{macrocode}
+% \end{macro}
+%
+% \subsection{Internal programming functions}
+%
+% \begin{macro}[int]{\group_align_safe_begin:, \group_align_safe_end:}
+%   \TeX{}'s alignment structures present many problems. As Knuth says
+%   himself in \emph{\TeX : The Program}: \enquote{It's sort of a miracle
+%   whenever \cs{halign} or \cs{valign} work, [\ldots]} One problem relates
+%   to commands that internally issues a |\cr| but also peek ahead for
+%   the next character for use in, say, an optional argument. If the
+%   next token happens to be a |&| with category code~4 we will get some
+%   sort of weird error message because the underlying
+%   |\tex_futurelet:D| will store the token at the end of the alignment
+%   template. This could be a |&|$_4$ giving a message like
+%   |! Misplaced \cr.| or even worse: it could be the |\endtemplate|
+%   token causing even more trouble! To solve this we have to open a
+%   special group so that \TeX{} still thinks it's on safe ground but at
+%   the same time we don't want to introduce any brace group that may
+%   find its way to the output. The following functions help with this
+%   by using code documented only in Appendix~D of
+%   \emph{The \TeX{}book}\dots
+%    \begin{macrocode}
+\cs_new_nopar:Npn \group_align_safe_begin:
+  { \if_false: { \fi: \if_int_compare:w `} = \c_zero \fi: }
+\cs_new_nopar:Npn \group_align_safe_end:
+  { \if_int_compare:w `{ = \c_zero } \fi: }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}[int]{\scan_align_safe_stop:}
+%   When \TeX{} is in the beginning of an align cell (right after the
+%   |\cr|) it is in a somewhat strange mode as it is looking ahead to
+%   find an |\tex_omit:D| or |\tex_noalign:D| and hasn't looked at the
+%   preamble yet. Thus an |\tex_ifmmode:D| test will always fail unless
+%   we insert |\scan_stop:| to stop \TeX{}'s scanning ahead. On the other
+%   hand we don't want to insert a |\scan_stop:| every time as that will
+%   destroy kerning between letters\footnote{Unless we enforce an extra
+%   pass with an appropriate value of \cs{pretolerance}.}
+%   Unfortunately there is no way to detect if we're in the beginning of
+%   an alignment cell as they have different characteristics depending
+%   on column number, \emph{etc.} However we \emph{can} detect if we're in an
+%   alignment cell by checking the current group type and we can also
+%   check if the previous node was a character or ligature. What is done
+%   here is that |\scan_stop:| is only inserted if an only
+%   if a)~we're in the
+%   outer part of an alignment cell and b)~the last node \emph{wasn't} a
+%   char node or a ligature node.
+%    \begin{macrocode}
+\cs_new_nopar:Npn \scan_align_safe_stop:
+  {
+    \int_compare:nNnT \etex_currentgrouptype:D = \c_six
+      {
+        \int_compare:nNnF \etex_lastnodetype:D = \c_zero
+          { \int_compare:nNnF \etex_lastnodetype:D = \c_seven { \scan_stop: } }
+      }
+  }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}[int]{\prg_variable_get_scope:N}
+% \begin{macro}[aux]{\prg_variable_get_scope_aux:w}
+% \begin{macro}[int]{\prg_variable_get_type:N}
+% \begin{macro}[aux]{\prg_variable_get_type:w}
+%   Expandable functions to find the type of a variable, and to
+%   return \texttt{g} if the variable is global. The trick for
+%   \cs{prg_variable_get_scope:N} is the same as that in
+%   \cs{cs_split_function:NN}, but it can be simplified as the
+%   requirements here are less complex.
+%    \begin{macrocode}
+\group_begin:
+  \tex_lccode:D `\& = `\g \scan_stop:
+  \tex_catcode:D `\& = \c_twelve
+\tl_to_lowercase:n
+  {
+    \group_end:
+    \cs_new_nopar:Npn \prg_variable_get_scope:N #1
+      {
+        \exp_last_unbraced:Nf \prg_variable_get_scope_aux:w
+          { \cs_to_str:N #1 \exp_stop_f: \q_stop }
+      }
+    \cs_new_nopar:Npn \prg_variable_get_scope_aux:w #1#2 \q_stop
+      { \token_if_eq_meaning:NNT & #1 { g } }
+  }
+\group_begin:
+  \tex_lccode:D `\& = `\_ \scan_stop:
+  \tex_catcode:D `\& = \c_twelve
+\tl_to_lowercase:n
+  {
+    \group_end:
+    \cs_new_nopar:Npn \prg_variable_get_type:N #1
+      {
+        \exp_after:wN \prg_variable_get_type_aux:w
+          \token_to_str:N #1 & a \q_stop
+      }
+    \cs_new_nopar:Npn \prg_variable_get_type_aux:w #1 & #2#3 \q_stop
+      {
+        \token_if_eq_meaning:NNTF a #2
+          {#1}
+          { \prg_variable_get_type_aux:w #2#3 \q_stop }
+      }
+  }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% 
+% \subsection{Experimental programmings functions}
+%
+%
+% \begin{macro}[aux]{\prg_define_quicksort:nnn}
+%   |#1| is the name, |#2| and |#3| are the tokens enclosing the
+%   argument. For the somewhat strange \meta{clist} type which doesn't
+%   enclose the items but uses a separator we define it by hand
+%   afterwards. When doing the first pass, the algorithm wraps all
+%   elements in braces and then uses a generic quicksort which works
+%   on token lists.
+%
+%   As an example
+%   \begin{quote}
+%   |\prg_define_quicksort:nnn{seq}{\seq_elt:w}{\seq_elt_end:w}|
+%   \end{quote}
+%   defines the user function |\seq_quicksort:n| and furthermore
+%   expects to use the two functions |\seq_quicksort_compare:nnTF|
+%   which compares the items and |\seq_quicksort_function:n| which is
+%   placed before each sorted item. It is up to the programmer to
+%   define these functions when needed. For the |seq| type a sequence
+%   is a token list variable, so one additionally has to define
+%   \begin{quote}
+%     |\cs_set_nopar:Npn \seq_quicksort:N{\exp_args:No\seq_quicksort:n}|
+%   \end{quote}
+%
+%
+%   For details on the implementation see \enquote{Sorting in \TeX{}'s Mouth}
+%   by Bernd Raichle. Firstly we define the function for parsing the
+%   initial list and then the braced list afterwards.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \prg_define_quicksort:nnn #1#2#3 {
+  \cs_set:cpx{#1_quicksort:n}##1{
+    \exp_not:c{#1_quicksort_start_partition:w} ##1
+    \exp_not:n{#2\q_nil#3\q_stop}
+  }
+  \cs_set:cpx{#1_quicksort_braced:n}##1{
+    \exp_not:c{#1_quicksort_start_partition_braced:n} ##1
+    \exp_not:N\q_nil\exp_not:N\q_stop
+  }
+  \cs_set:cpx {#1_quicksort_start_partition:w} #2 ##1 #3{
+    \exp_not:N \quark_if_nil:nT {##1}\exp_not:N \use_none_delimit_by_q_stop:w
+    \exp_not:c{#1_quicksort_do_partition_i:nnnw} {##1}{}{}
+  }
+  \cs_set:cpx {#1_quicksort_start_partition_braced:n} ##1 {
+    \exp_not:N \quark_if_nil:nT {##1}\exp_not:N \use_none_delimit_by_q_stop:w
+    \exp_not:c{#1_quicksort_do_partition_i_braced:nnnn} {##1}{}{}
+  }
+%    \end{macrocode}
+% Now for doing the partitions.
+%    \begin{macrocode}
+  \cs_set:cpx {#1_quicksort_do_partition_i:nnnw} ##1##2##3 #2 ##4 #3 {
+    \exp_not:N \quark_if_nil:nTF {##4} \exp_not:c {#1_do_quicksort_braced:nnnnw}
+    {
+      \exp_not:c{#1_quicksort_compare:nnTF}{##1}{##4}
+      \exp_not:c{#1_quicksort_partition_greater_ii:nnnn}
+      \exp_not:c{#1_quicksort_partition_less_ii:nnnn}
+    }
+    {##1}{##2}{##3}{##4}
+  }
+  \cs_set:cpx {#1_quicksort_do_partition_i_braced:nnnn} ##1##2##3##4 {
+    \exp_not:N \quark_if_nil:nTF {##4} \exp_not:c {#1_do_quicksort_braced:nnnnw}
+    {
+      \exp_not:c{#1_quicksort_compare:nnTF}{##1}{##4}
+      \exp_not:c{#1_quicksort_partition_greater_ii_braced:nnnn}
+      \exp_not:c{#1_quicksort_partition_less_ii_braced:nnnn}
+    }
+    {##1}{##2}{##3}{##4}
+  }
+  \cs_set:cpx {#1_quicksort_do_partition_ii:nnnw} ##1##2##3 #2 ##4 #3 {
+    \exp_not:N \quark_if_nil:nTF {##4} \exp_not:c {#1_do_quicksort_braced:nnnnw}
+    {
+      \exp_not:c{#1_quicksort_compare:nnTF}{##4}{##1}
+      \exp_not:c{#1_quicksort_partition_less_i:nnnn}
+      \exp_not:c{#1_quicksort_partition_greater_i:nnnn}
+    }
+    {##1}{##2}{##3}{##4}
+  }
+  \cs_set:cpx {#1_quicksort_do_partition_ii_braced:nnnn} ##1##2##3##4 {
+    \exp_not:N \quark_if_nil:nTF {##4} \exp_not:c {#1_do_quicksort_braced:nnnnw}
+    {
+      \exp_not:c{#1_quicksort_compare:nnTF}{##4}{##1}
+      \exp_not:c{#1_quicksort_partition_less_i_braced:nnnn}
+      \exp_not:c{#1_quicksort_partition_greater_i_braced:nnnn}
+    }
+    {##1}{##2}{##3}{##4}
+  }
+%    \end{macrocode}
+% This part of the code handles the two branches in each
+% sorting. Again we will also have to do it braced.
+%    \begin{macrocode}
+  \cs_set:cpx {#1_quicksort_partition_less_i:nnnn} ##1##2##3##4{
+    \exp_not:c{#1_quicksort_do_partition_i:nnnw}{##1}{##2}{{##4}##3}}
+  \cs_set:cpx {#1_quicksort_partition_less_ii:nnnn} ##1##2##3##4{
+    \exp_not:c{#1_quicksort_do_partition_ii:nnnw}{##1}{##2}{##3{##4}}}
+  \cs_set:cpx {#1_quicksort_partition_greater_i:nnnn} ##1##2##3##4{
+    \exp_not:c{#1_quicksort_do_partition_i:nnnw}{##1}{{##4}##2}{##3}}
+  \cs_set:cpx {#1_quicksort_partition_greater_ii:nnnn} ##1##2##3##4{
+    \exp_not:c{#1_quicksort_do_partition_ii:nnnw}{##1}{##2{##4}}{##3}}
+  \cs_set:cpx {#1_quicksort_partition_less_i_braced:nnnn} ##1##2##3##4{
+    \exp_not:c{#1_quicksort_do_partition_i_braced:nnnn}{##1}{##2}{{##4}##3}}
+  \cs_set:cpx {#1_quicksort_partition_less_ii_braced:nnnn} ##1##2##3##4{
+    \exp_not:c{#1_quicksort_do_partition_ii_braced:nnnn}{##1}{##2}{##3{##4}}}
+  \cs_set:cpx {#1_quicksort_partition_greater_i_braced:nnnn} ##1##2##3##4{
+    \exp_not:c{#1_quicksort_do_partition_i_braced:nnnn}{##1}{{##4}##2}{##3}}
+  \cs_set:cpx {#1_quicksort_partition_greater_ii_braced:nnnn} ##1##2##3##4{
+    \exp_not:c{#1_quicksort_do_partition_ii_braced:nnnn}{##1}{##2{##4}}{##3}}
+%    \end{macrocode}
+% Finally, the big kahuna! This is where the sub-lists are sorted.
+%    \begin{macrocode}
+  \cs_set:cpx {#1_do_quicksort_braced:nnnnw} ##1##2##3##4\q_stop {
+    \exp_not:c{#1_quicksort_braced:n}{##2}
+    \exp_not:c{#1_quicksort_function:n}{##1}
+    \exp_not:c{#1_quicksort_braced:n}{##3}
+  }
+}
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\prg_quicksort:n}
+% \UnitTested
+%   A simple version. Sorts a list of tokens, uses the function
+%   |\prg_quicksort_compare:nnTF| to compare items, and places the
+%   function |\prg_quicksort_function:n| in front of each of them.
+%    \begin{macrocode}
+\prg_define_quicksort:nnn {prg}{}{}
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\prg_quicksort_function:n}
+% \UnitTested
+% \begin{macro}{\prg_quicksort_compare:nnTF}
+% \UnitTested
+%    \begin{macrocode}
+\cs_set:Npn \prg_quicksort_function:n {\ERROR}
+\cs_set:Npn \prg_quicksort_compare:nnTF {\ERROR}
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \subsection{Deprecated functions}
+%
+% These were depreciated on 2011-05-27 and will be removed entirely by
+% 2011-08-31.
+%
+% \begin{macro}{\prg_new_map_functions:Nn}
+% \begin{macro}{\prg_set_map_functions:Nn}
+%   As we have restructured the structured variables, these are no
+%   longer needed.
+%    \begin{macrocode}
+\cs_new_protected:Npn \prg_new_map_functions:Nn #1#2 { \deprectiated }
+\cs_new_protected:Npn \prg_set_map_functions:Nn #1#2 { \deprectiated }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+%    \begin{macrocode}
+%</initex|package>
+%    \end{macrocode}
+%
+% \end{implementation}
+%
+% \PrintIndex
\ No newline at end of file
diff --git a/Master/texmf-dist/source/latex/l3kernel/l3prop.dtx b/Master/texmf-dist/source/latex/l3kernel/l3prop.dtx
new file mode 100644
index 00000000000..744cea629fc
--- /dev/null
+++ b/Master/texmf-dist/source/latex/l3kernel/l3prop.dtx
@@ -0,0 +1,1206 @@
+% \iffalse meta-comment
+%
+%% File: l3prop.dtx Copyright (C) 1990-2011 The LaTeX3 Project
+%%
+%% It may be distributed and/or modified under the conditions of the
+%% LaTeX Project Public License (LPPL), either version 1.3c of this
+%% license or (at your option) any later version.  The latest version
+%% of this license is in the file
+%%
+%%    http://www.latex-project.org/lppl.txt
+%%
+%% This file is part of the "expl3 bundle" (The Work in LPPL)
+%% and all files in that bundle must be distributed together.
+%%
+%% The released version of this bundle is available from CTAN.
+%%
+%% -----------------------------------------------------------------------
+%%
+%% The development version of the bundle can be found at
+%%
+%%    http://www.latex-project.org/svnroot/experimental/trunk/
+%%
+%% for those people who are interested.
+%%
+%%%%%%%%%%%
+%% NOTE: %%
+%%%%%%%%%%%
+%%
+%%   Snapshots taken from the repository represent work in progress and may
+%%   not work or may contain conflicting material!  We therefore ask
+%%   people _not_ to put them into distributions, archives, etc. without
+%%   prior consultation with the LaTeX3 Project.
+%%
+%% -----------------------------------------------------------------------
+%
+%<*driver|package>
+\RequirePackage{l3names}
+\GetIdInfo$Id: l3prop.dtx 2478 2011-06-19 21:34:23Z joseph $
+  {L3 Experimental property lists}
+%</driver|package>
+%<*driver>
+\documentclass[full]{l3doc}
+\begin{document}
+  \DocInput{\jobname.dtx}
+\end{document}
+%</driver>
+% \fi
+%
+% \title{^^A
+%   The \pkg{l3prop} package\\ Property lists^^A
+%   \thanks{This file describes v\ExplFileVersion,
+%      last revised \ExplFileDate.}^^A
+% }
+%
+% \author{^^A
+%  The \LaTeX3 Project\thanks
+%    {^^A
+%      E-mail:
+%        \href{mailto:latex-team@latex-project.org}
+%          {latex-team@latex-project.org}^^A
+%    }^^A
+% }
+%
+% \date{Released \ExplFileDate}
+%
+% \maketitle
+%
+% \begin{documentation}
+%
+% \LaTeX3 implements a \enquote{property list} data type, which contain
+% an unordered list of entries each of which consists of a \meta{key} and
+% an associated \meta{value}. The \meta{key} and \meta{value} may both be
+% any \meta{balanced text}. It is possible to map functions to property lists
+% such that the function is applied to every key--value pair within
+% the list.
+%
+% Each entry in a property list must have a unique \meta{key}: if an entry is
+% added to a property list which already contains the \meta{key} then the new
+% entry will overwrite the existing one. The \meta{keys} are compared on a
+% string basis, using the same method as \cs{str_if_eq:nn}.
+%
+% \section{Creating and initialising property lists}
+%
+% \begin{function}{\prop_new:N, \prop_new:c}
+%   \begin{syntax}
+%     \cs{prop_new:N} \meta{property list}
+%   \end{syntax}
+%   Creates a new \meta{property list} or raises an error if the name is
+%   already taken. The declaration is global. The \meta{property lists} will
+%   initially contain no entries.
+% \end{function}
+%
+% \begin{function}{\prop_clear:N, \prop_clear:c}
+%   \begin{syntax}
+%     \cs{prop_clear:N} \meta{property list}
+%   \end{syntax}
+%   Clears all entries from the \meta{property list} within the scope of
+%   the current \TeX{} group.
+% \end{function}
+%
+% \begin{function}{\prop_gclear:N, \prop_gclear:c}
+%   \begin{syntax}
+%     \cs{prop_gclear:N} \meta{property list}
+%   \end{syntax}
+%   Clears all entries from the \meta{property list} globally.
+% \end{function}
+%
+% \begin{function}{\prop_clear_new:N, \prop_clear_new:c}
+%   \begin{syntax}
+%     \cs{prop_clear_new:N} \meta{property list}
+%   \end{syntax}
+%   If the \meta{property list} already exists, clears it within the scope
+%   of the current \TeX{} group. If the \meta{property list} is not defined,
+%   it will be created (using \cs{prop_new:N}). Thus the property list is
+%   guaranteed to be available and clear within the current \TeX{}
+%   group. The \meta{property list} will exist globally, but the content
+%   outside of the current \TeX{} group is not specified.
+% \end{function}
+%
+% \begin{function}{\prop_gclear_new:N, \prop_gclear_new:c}
+%   \begin{syntax}
+%     \cs{prop_gclear_new:N} \meta{property list}
+%   \end{syntax}
+%   If the \meta{property list} already exists, clears it globally. If the
+%   \meta{property list} is not defined, it will be created (using
+%   \cs{prop_new:N}). Thus the property list is guaranteed to be available
+%   and globally clear.
+% \end{function}
+%
+% \begin{function}
+%   {\prop_set_eq:NN, \prop_set_eq:cN, \prop_set_eq:Nc, \prop_set_eq:cc}
+%   \begin{syntax}
+%     \cs{prop_set_eq:NN} \meta{property list1} \meta{property list2}
+%   \end{syntax}
+%   Sets the content of \meta{property list1} equal to that of
+%   \meta{property list2}. This assignment is restricted to the current
+%   \TeX{} group level.
+% \end{function}
+%
+% \begin{function}
+%   {\prop_gset_eq:NN, \prop_gset_eq:cN, \prop_gset_eq:Nc, \prop_gset_eq:cc}
+%   \begin{syntax}
+%     \cs{prop_gset_eq:NN} \meta{property list1} \meta{property list2}
+%   \end{syntax}
+%   Sets the content of \meta{property list1} equal to that of
+%   \meta{property list2}. This assignment is global and so is not
+%   limited by the current \TeX{} group level.
+% \end{function}
+%
+% \section{Adding entries to property lists}
+%
+% \begin{function}
+%   {
+%     \prop_put:Nnn, \prop_put:NnV, \prop_put:Nno, \prop_put:Nnx,
+%     \prop_put:NVn, \prop_put:NVV, \prop_put:Non, \prop_put:Noo,
+%     \prop_put:cnn, \prop_put:cnV, \prop_put:cno, \prop_put:cnx,
+%     \prop_put:cVn, \prop_put:cVV, \prop_put:con, \prop_put:coo
+%   }
+%   \begin{syntax}
+%     \cs{prop_put:Nnn} \meta{property list} \Arg{key} \Arg{value}
+%   \end{syntax}
+%   Adds an entry to the \meta{property list} which may be accessed
+%   using the \meta{key} and which has \meta{value}. Both the \meta{key}
+%   and \meta{value} may contain any \meta{balanced text}. The \meta{key}
+%   is stored after processing with \cs{tl_to_str:n}, meaning that
+%   category codes are ignored. If the \meta{key} is already present
+%   in the \meta{property list}, the existing entry is overwritten
+%   by the new \meta{value}. The assignment is restricted to the current
+%   \TeX{} group.
+% \end{function}
+%
+% \begin{function}
+%   {
+%     \prop_gput:Nnn, \prop_gput:NnV, \prop_gput:Nno, \prop_gput:Nnx,
+%     \prop_gput:NVn, \prop_gput:NVV, \prop_gput:Non, \prop_gput:Noo,
+%     \prop_gput:cnn, \prop_gput:cnV, \prop_gput:cno, \prop_gput:cnx,
+%     \prop_gput:cVn, \prop_gput:cVV, \prop_gput:con, \prop_gput:coo
+%   }
+%   \begin{syntax}
+%     \cs{prop_gput:Nnn} \meta{property list} \Arg{key} \Arg{value}
+%   \end{syntax}
+%   Adds an entry to the \meta{property list} which may be accessed
+%   using the \meta{key} and which has \meta{value}. Both the \meta{key}
+%   and \meta{value} may contain any \meta{balanced text}. The \meta{key}
+%   is stored after processing with \cs{tl_to_str:n}, meaning that
+%   category codes are ignored. If the \meta{key} is already present
+%   in the \meta{property list}, the existing entry is overwritten
+%   by the new \meta{value}. The assignment is global.
+% \end{function}
+%
+% \begin{function}{\prop_put_if_new:Nnn, \prop_put_if_new:cnn}
+%   \begin{syntax}
+%     \cs{prop_put_if_new:Nnn} \meta{property list} \Arg{key} \Arg{value}
+%   \end{syntax}
+%   If the \meta{key} is present in the \meta{property list} then
+%   no action is taken. If the \meta{key} is not present in the
+%   \meta{property list} then a new entry is added. Both the \meta{key}
+%   and \meta{value} may contain any \meta{balanced text}. The \meta{key}
+%   is stored after processing with \cs{tl_to_str:n}, meaning that
+%   category codes are ignored. The assignment is restricted to the current
+%   \TeX{} group.
+% \end{function}
+%
+% \begin{function}{\prop_gput_if_new:Nnn, \prop_gput_if_new:cnn}
+%   \begin{syntax}
+%     \cs{prop_gput_if_new:Nnn} \meta{property list} \Arg{key} \Arg{value}
+%   \end{syntax}
+%   If the \meta{key} is present in the \meta{property list} then
+%   no action is taken. If the \meta{key} is not present in the
+%   \meta{property list} then a new entry is added. Both the \meta{key}
+%   and \meta{value} may contain any \meta{balanced text}. The \meta{key}
+%   is stored after processing with \cs{tl_to_str:n}, meaning that
+%   category codes are ignored. The assignment is global.
+% \end{function}
+%
+% \section{Recovering values from property lists}
+%
+% \begin{function}
+%   {
+%     \prop_get:NnN, \prop_get:NVN, \prop_get:NoN,
+%     \prop_get:cnN, \prop_get:cVN, \prop_get:coN,
+%   }
+%   \begin{syntax}
+%     \cs{prop_get:NnN} \meta{property list} \Arg{key} \meta{tl var}
+%   \end{syntax}
+%   Recovers the \meta{value} stored with \meta{key} from the
+%   \meta{property list}, and places this in the \meta{token list
+%   variable}. If the \meta{key} is not found in the
+%   \meta{property list} then the \meta{token list variable} will
+%   contain the special marker \cs{q_no_value}. The \meta{token list
+%   variable} is set within the current \TeX{} group.
+% \end{function}
+%
+% \begin{function}{\prop_pop:NnN, \prop_pop:NoN, \prop_pop:cnN, \prop_pop:coN}
+%   \begin{syntax}
+%     \cs{prop_pop:NnN} \meta{property list} \Arg{key} \meta{tl var}
+%   \end{syntax}
+%   Recovers the \meta{value} stored with \meta{key} from the
+%   \meta{property list}, and places this in the \meta{token list
+%   variable}. If the \meta{key} is not found in the
+%   \meta{property list} then the \meta{token list variable} will
+%   contain the special marker \cs{q_no_value}. The \meta{key} and
+%   \meta{value} are then deleted from the property list. Both
+%   assignments are local.
+% \end{function}
+%
+% \begin{function}
+%   {\prop_gpop:NnN, \prop_gpop:NoN, \prop_gpop:cnN, \prop_gpop:coN}
+%   \begin{syntax}
+%     \cs{prop_gpop:NnN} \meta{property list} \Arg{key} \meta{tl var}
+%   \end{syntax}
+%   Recovers the \meta{value} stored with \meta{key} from the
+%   \meta{property list}, and places this in the \meta{token list
+%   variable}. If the \meta{key} is not found in the
+%   \meta{property list} then the \meta{token list variable} will
+%   contain the special marker \cs{q_no_value}. The \meta{key} and
+%   \meta{value} are then deleted from the property list.
+%   The \meta{property list} is modified globally, while the assignment of
+%   the \meta{token list variable} is local.
+% \end{function}
+%
+% \section{Modifying property lists}
+%
+% \begin{function}{\prop_del:Nn, \prop_del:NV, \prop_del:cn, \prop_del:cV}
+%   \begin{syntax}
+%     \cs{prop_del:Nn} \meta{property list} \Arg{key}
+%   \end{syntax}
+%   Deletes the entry listed under \meta{key} from the
+%   \meta{property list} which may be accessed. If the \meta{key} is
+%   not found in the \meta{property list} no change occurs,
+%   \emph{i.e}~there is no need to test for the existence of a key before
+%   deleting it. The deletion is restricted to the current \TeX{} group.
+% \end{function}
+%
+% \begin{function}{\prop_gdel:Nn, \prop_gdel:NV, \prop_gdel:cn, \prop_gdel:cV}
+%   \begin{syntax}
+%     \cs{prop_gdel:Nn} \meta{property list} \Arg{key}
+%   \end{syntax}
+%   Deletes the entry listed under \meta{key} from the
+%   \meta{property list} which may be accessed. If the \meta{key} is
+%    not found in the \meta{property list} no change occurs,
+%   \emph{i.e}~there is no need to test for the existence of a key before
+%   deleting it. The deletion is not restricted to the current \TeX{}
+%   group: it is global.
+% \end{function}
+%
+% \section{Property list conditionals}
+%
+% \begin{function}[EXP,pTF]{\prop_if_empty:N, \prop_if_empty:c}
+%   \begin{syntax}
+%     \cs{prop_if_empty_p:N} \meta{property list}
+%     \cs{prop_if_empty:NTF} \meta{property list}
+%     ~~\Arg{true code} \Arg{false code}
+%   \end{syntax}
+%   Tests if the \meta{property list} is empty (containing no entries). The
+%   branching versions then leave either \meta{true code} or
+%   \meta{false code} in the input stream, as appropriate to the truth of
+%   the test and the variant of the function chosen. The logical truth of
+%   the test is left in the input stream by the predicate version.
+% \end{function}
+%
+% \begin{function}[EXP,pTF]{
+%    \prop_if_in:Nn, \prop_if_in:NV, \prop_if_in:No,
+%    \prop_if_in:cn, \prop_if_in:cV, \prop_if_in:co
+% }
+%   \begin{syntax}
+%     \cs{prop_if_in:NnTF} \meta{property list} \Arg{key}
+%     ~~\Arg{true code} \Arg{false code}
+%   \end{syntax}
+%   Tests if the \meta{key} is present in the \meta{property list},
+%   making the comparison using the method described by \cs{str_if_eq:nnTF}.
+%   Either the \meta{true code} or \meta{false code} in the input stream, as
+%   appropriate to the truth of the test and the variant of the function
+%   chosen.
+%   \begin{texnote}
+%     This function iterates through every key--value pair in the
+%     \meta{property list} and is therefore slower than using the
+%     non-expandable \cs{prop_get:NnNTF}.
+%   \end{texnote}
+% \end{function}
+%
+% \section{Mapping to property lists}
+%
+% \begin{function}[EXP]{\prop_map_function:NN, \prop_map_function:cN}
+%   \begin{syntax}
+%     \cs{prop_map_function:NN} \meta{property list} \meta{function}
+%   \end{syntax}
+%   Applies \meta{function} to every \meta{entry} stored in the
+%   \meta{property list}. The \meta{function} will receive two argument for
+%   each iteration.: the \meta{key} and associated \meta{value}.
+%   The order in which \meta{entries} are returned is not defined and
+%   should not be relied upon.
+% \end{function}
+%
+% \begin{function}{\prop_map_inline:Nn, \prop_map_inline:cn}
+%   \begin{syntax}
+%     \cs{prop_map_inline:Nn} \meta{property list} \Arg{inline function}
+%   \end{syntax}
+%   Applies \meta{inline function} to every \meta{entry} stored
+%   within the \meta{property list}. The \meta{inline function} should
+%   consist of code which will receive the \meta{key} as |#1| and the
+%   \meta{value} as |#2|.
+%   The order in which \meta{entries} are returned is not defined and
+%   should not be relied upon.
+% \end{function}
+%
+% \begin{function}[EXP]{\prop_map_break:}
+%   \begin{syntax}
+%     \cs{prop_map_break:}
+%   \end{syntax}
+%   Used to terminate a \cs{prop_map_\ldots} function before all
+%   entries in the \meta{property list} have been processed. This will
+%   normally take place within a conditional statement, for example
+%   \begin{verbatim}
+%     \prop_map_inline:Nn \l_my_prop
+%       {
+%         \str_if_eq:nnTF { #1 } { bingo }
+%           { \prop_map_break: }
+%           {
+%             % Do something useful
+%           }
+%       }
+%   \end{verbatim}
+%   Use outside of a \cs{prop_map_\ldots} scenario will lead low
+%   level \TeX{} errors.
+% \end{function}
+%
+% \begin{function}[EXP]{\prop_map_break:n}
+%   \begin{syntax}
+%     \cs{prop_map_break:n} \Arg{tokens}
+%   \end{syntax}
+%   Used to terminate a \cs{prop_map_\ldots} function before all
+%   entries in the \meta{property list} have been processed, inserting
+%   the \meta{tokens} after the mapping has ended. This will
+%   normally take place within a conditional statement, for example
+%   \begin{verbatim}
+%     \prop_map_inline:Nn \l_my_prop
+%       {
+%         \str_if_eq:nnTF { #1 } { bingo }
+%           { \prop_map_break:n { <tokens> } }
+%           {
+%             % Do something useful
+%           }
+%       }
+%   \end{verbatim}
+%   Use outside of a \cs{prop_map_\ldots} scenario will lead low
+%   level \TeX{} errors.
+% \end{function}
+%
+% \section{Viewing property lists}
+%
+% \begin{function}{\prop_show:N, \prop_show:c}
+%   \begin{syntax}
+%     \cs{prop_show:N} \meta{property list}
+%   \end{syntax}
+%   Displays the entries in the \meta{property list} in the terminal.
+% \end{function}
+%
+% \section{Experimental property list functions}
+%
+% This section contains functions which may or may not be retained, depending
+% on how useful they are found to be.
+%
+% \begin{function}[TF]{\prop_get:NnN}
+%   \begin{syntax}
+%     \cs{prop_get:NnNTF} \meta{property list} \Arg{key}
+%     ~~\meta{token list variable} \Arg{true code} \Arg{false code}
+%   \end{syntax}
+%   If the \meta{key} is not present in the \meta{property list}, leaves
+%   the \meta{false code} in the input stream and leaves the
+%   \meta{token list variable} unchanged. If the \meta{key} is present in
+%   the \meta{property list}, stores the corresponding \meta{value}
+%   in the \meta{token list variable} without removing it from the
+%   \meta{property list}. The \meta{true code} is then left in the input
+%   stream. The \meta{token list variable} is assigned locally.
+% \end{function}
+%
+% \begin{function}[TF]{\prop_pop:NnN}
+%   \begin{syntax}
+%     \cs{prop_pop:NnNTF} \meta{property list} \Arg{key}
+%     ~~\meta{token list variable} \Arg{true code} \Arg{false code}
+%   \end{syntax}
+%   If the \meta{key} is not present in the \meta{property list}, leaves
+%   the \meta{false code} in the input stream and leaves the
+%   \meta{token list variable} unchanged. If the \meta{key} is present in
+%   the \meta{property list}, pops the corresponding \meta{value}
+%   in the \meta{token list variable}, \emph{i.e.}~removes the item from
+%   the \meta{property list}. The \meta{true code} is then left in the input
+%   stream. Both the \meta{property list} and the \meta{token list variable}
+%   are assigned locally.
+% \end{function}
+%
+% \begin{function}[TF]{\prop_gpop:NnN}
+%   \begin{syntax}
+%     \cs{prop_gpop:NnNTF} \meta{property list} \Arg{key}
+%     ~~\meta{token list variable} \Arg{true code} \Arg{false code}
+%   \end{syntax}
+%   If the \meta{key} is not present in the \meta{property list}, leaves
+%   the \meta{false code} in the input stream and leaves the
+%   \meta{token list variable} unchanged. If the \meta{key} is present in
+%   the \meta{property list}, pops the corresponding \meta{value}
+%   in the \meta{token list variable}, \emph{i.e.}~removes the item from
+%   the \meta{property list}. The \meta{true code} is then left in the input
+%   stream. The \meta{property list} is modified globally, while the
+%   \meta{token list variable} is assigned locally.
+% \end{function}
+%
+% \begin{function}[EXP]{\prop_map_tokens:Nn}
+%   \begin{syntax}
+%     \cs{prop_map_tokens:Nn} \meta{property list} \Arg{code}
+%   \end{syntax}
+%   Analogue of \cs{prop_map_function:NN} which maps several tokens instead
+%   of a single function. Useful in particular when mapping through a
+%   property list while keeping track of a given key.
+% \end{function}
+%
+% \begin{function}[EXP]{\prop_get:Nn}
+%   \begin{syntax}
+%     \cs{prop_get:Nn} \meta{property list} \Arg{key}
+%   \end{syntax}
+%   Expands to the \meta{value} corresponding to the \meta{key} in
+%   the \meta{property list}. If the \meta{key} is missing, this has
+%   an empty expansion.
+%   \begin{texnote}
+%     This function is slower than the non-expandable analogue
+%     \cs{prop_get:NnN}.
+%   \end{texnote}
+% \end{function}
+%
+% \section{Internal property list functions}
+%
+% \begin{function}{\q_prop}
+%   The internal token used to separate out property list entries, separating
+%   both the \meta{key} from the \meta{value} and also one entry from another.
+% \end{function}
+%
+% \begin{function}{\c_empty_prop}
+%   A permanently-empty property list used for internal comparisons.
+% \end{function}
+%
+% \begin{function}{\prop_split:Nnn}
+%   \begin{syntax}
+%     \cs{prop_spilt:Nnn} \meta{property list} \Arg{key} \Arg{code}
+%   \end{syntax}
+%   Splits the \meta{property list} at the \meta{key}, giving three
+%   groups: the \meta{extract} of \meta{property list} before the
+%   \meta{key}, the \meta{value} associated with the \meta{key} and the
+%   \meta{extract} of the \meta{property list} after the \meta{value}.
+%   The first \meta{extract} retains the internal structure of a property
+%   list. The second is only missing the leading separator \cs{q_prop}.
+%   This ensures that the concatenation of the two \meta{extracts} is a
+%   property list.
+%   If the \meta{key} is not present in the \meta{property list} then
+%   the second group will contain the marker \cs{q_no_value} and the third
+%   is empty.
+%   Once the split has occurred, the \meta{code} is inserted followed by the
+%   three groups: thus the \meta{code} should properly absorb three
+%   arguments.
+%   The \meta{key} comparison takes place as described for \cs{str_if_eq:nn}.
+% \end{function}
+%
+% \begin{function}{\prop_split:NnTF}
+%   \begin{syntax}
+%     \cs{prop_spilt:NnTF} \meta{property list} \Arg{key}
+%     ~~\Arg{true code} \Arg{false code}
+%   \end{syntax}
+%   Splits the \meta{property list} at the \meta{key}, giving three
+%   groups: the \meta{extract} of \meta{property list} before the
+%   \meta{key}, the \meta{value} associated with the \meta{key} and the
+%   \meta{extract} of the \meta{property list} after the \meta{value}.
+%   The first \meta{extract} retains the internal structure of a property
+%   list. The second is only missing the leading separator \cs{q_prop}.
+%   This ensures that the concatenation of the two \meta{extracts} is a
+%   property list.
+%   If the \meta{key} is present in the \meta{property list} then the
+%   \meta{true code} is left in the input stream, followed by the three
+%   groups: thus the \meta{true code} should properly absorb three arguments.
+%   If the \meta{key} is not present in the \meta{property list} then
+%   the \meta{false code} is left in the input stream, with no trailing
+%   material.
+%   The \meta{key} comparison takes place as described for \cs{str_if_eq:nn}.
+% \end{function}
+%
+% \end{documentation}
+%
+% \begin{implementation}
+%
+% \section{\pkg{l3prop} implementation}
+%
+% \TestFiles{m3prop001}
+%
+%    \begin{macrocode}
+%<*initex|package>
+%    \end{macrocode}
+%
+%    \begin{macrocode}
+%<*package>
+\ProvidesExplPackage
+  {\ExplFileName}{\ExplFileDate}{\ExplFileVersion}{\ExplFileDescription}
+\package_check_loaded_expl:
+%</package>
+%    \end{macrocode}
+%
+% A property list is a macro whose top-level expansion is for the form
+% \enquote{\cs{q_prop} \meta{key$_0$} \cs{q_prop} \Arg{value$_0$} \cs{q_prop}
+% \ldots \cs{q_prop} \meta{key$_{n-1}$} \cs{q_prop} \Arg{value$_{n-1}$}
+% \cs{q_prop}}.
+% The trailing \cs{q_prop} is always present for performance reasons: this
+% means that empty property lists are not actually empty.
+%
+% \begin{macro}[int]{\q_prop}
+%   A private quark is used as a marker between entries.
+%    \begin{macrocode}
+\quark_new:N \q_prop
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{variable}{\c_empty_prop}
+%   An empty prop contains exactly one \cs{q_prop}.
+%    \begin{macrocode}
+\tl_const:Nn \c_empty_prop { \q_prop }
+%    \end{macrocode}
+% \end{variable}
+%
+% \subsection{Allocation and initialisation}
+%
+% \begin{macro}{\prop_new:N,\prop_new:c}
+%   Internally, property lists are token lists, but an empty prop
+%   is not an empty tl, so we need to do things by hand.
+%    \begin{macrocode}
+\cs_new_protected:Npn \prop_new:N #1 { \cs_new_eq:NN  #1  \c_empty_prop }
+\cs_new_protected:Npn \prop_new:c #1 { \cs_new_eq:cN {#1} \c_empty_prop }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\prop_clear:N, \prop_clear:c}
+% \begin{macro}{\prop_gclear:N, \prop_gclear:c}
+%   The same idea for  clearing
+%    \begin{macrocode}
+\cs_new_protected:Npn \prop_clear:N #1  { \cs_set_eq:NN   #1  \c_empty_prop }
+\cs_new_protected:Npn \prop_clear:c #1  { \cs_set_eq:cN  {#1} \c_empty_prop }
+\cs_new_protected:Npn \prop_gclear:N #1 { \cs_gset_eq:NN  #1  \c_empty_prop }
+\cs_new_protected:Npn \prop_gclear:c #1 { \cs_gset_eq:cN {#1} \c_empty_prop }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\prop_clear_new:N, \prop_clear_new:c}
+% \begin{macro}{\prop_gclear_new:N, \prop_gclear_new:c}
+%   Once again a simple copy from the token list functions.
+%    \begin{macrocode}
+\cs_new_protected:Npn \prop_clear_new:N #1
+  { \cs_if_exist:NTF #1 { \prop_clear:N #1 } { \prop_new:N #1 } }
+\cs_generate_variant:Nn \prop_clear_new:N {c}
+\cs_new_protected:Npn \prop_gclear_new:N #1
+  { \cs_if_exist:NTF #1 { \prop_gclear:N #1 } { \prop_new:N #1 } }
+\cs_new_eq:NN \prop_gclear_new:c \prop_gclear:c
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}
+%   {\prop_set_eq:NN, \prop_set_eq:cN, \prop_set_eq:Nc, \prop_set_eq:cc}
+% \begin{macro}
+%   {\prop_gset_eq:NN, \prop_gset_eq:cN, \prop_gset_eq:Nc, \prop_gset_eq:cc}
+%   Once again, these are simply copies from the token list functions.
+%    \begin{macrocode}
+\cs_new_eq:NN \prop_set_eq:NN  \tl_set_eq:NN
+\cs_new_eq:NN \prop_set_eq:Nc  \tl_set_eq:Nc
+\cs_new_eq:NN \prop_set_eq:cN  \tl_set_eq:cN
+\cs_new_eq:NN \prop_set_eq:cc  \tl_set_eq:cc
+\cs_new_eq:NN \prop_gset_eq:NN \tl_gset_eq:NN
+\cs_new_eq:NN \prop_gset_eq:Nc \tl_gset_eq:Nc
+\cs_new_eq:NN \prop_gset_eq:cN \tl_gset_eq:cN
+\cs_new_eq:NN \prop_gset_eq:cc \tl_gset_eq:cc
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \subsection{Accessing data in property lists}
+%
+% \begin{macro}[int]{\prop_split:NnTF}
+% \begin{macro}[aux]{\prop_split_aux:NnTF}
+% \begin{macro}[aux]{\prop_split_aux:nnnn}
+% \begin{macro}[aux]{\prop_split_aux:w}
+%   This function is used by most of the module, and hence must be fast.
+%   The aim here is to split a property list at a given key into the part
+%   before the key--value pair, the value associated with the key and the part
+%   after the key--value pair. To do this, the key is first detokenized (to
+%   avoid repeatedly doing this), then a delimited function is constructed to
+%   match the key. It will match \cs{q_prop} \meta{detokenized key} \cs{q_prop}
+%   \marg{value} \meta{extra argument}, effectively separating an
+%   \meta{extract1} before the key in the property list and an \meta{extract2}
+%   after the key.
+%
+%   If the key is present in the property list, then \meta{extra argument}
+%   is simply \cs{q_prop}, and \cs{prop_splot_aux:nnnn} will gobble this
+%   and the false branch (|#4|), leaving the correct code on the input
+%   stream. More precisely, it leaves the user code (true branch), followed
+%   by three groups, \Arg{extract1} \Arg{value} \Arg{extract2}.
+%   In order for \meta{extract1}\meta{extract2} to be a well-formed
+%   property list, \meta{extract1} has a leading and trailing \cs{q_prop},
+%   retaining exactly the structure of a property list, while \meta{extract2}
+%   omits the leading \cs{q_prop}.
+%
+%   If the key is not there, then \meta{extra argument} is
+%   |? \use_ii:nn { }|, and |\prop_split_aux:nnnn ? \use_ii:nn { }| removes
+%   the three brace groups that just follow. Then \cs{use_ii:nn} removes
+%   the true branch, leaving the false branch, with no trailing material.
+%    \begin{macrocode}
+\cs_set_protected:Npn \prop_split:NnTF #1#2
+  { \exp_args:NNo \prop_split_aux:NnTF #1 { \tl_to_str:n {#2} } }
+\cs_new_protected:Npn \prop_split_aux:NnTF #1#2
+  {
+    \cs_set_protected:Npn \prop_split_aux:w
+      ##1 \q_prop #2 \q_prop ##2 ##3 ##4 \q_mark ##5 \q_stop
+      { \prop_split_aux:nnnn ##3 { {##1 \q_prop } {##2} {##4} } }
+    \exp_after:wN \prop_split_aux:w #1 \q_mark
+         \q_prop #2 \q_prop { } { ? \use_ii:nn { } } \q_mark \q_stop
+  }
+\cs_new:Npn \prop_split_aux:nnnn #1#2#3#4 { #3 #2 }
+\cs_new_protected:Npn \prop_split_aux:w { }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}[int]{\prop_split:Nnn}
+%   The goal here is to provide a common interface for both true and false
+%   branches of \cs{prop_split:NnTF}. In both cases, the code given by the
+%   user will be placed in front of three brace groups, \Arg{extract1}
+%   \Arg{value} \Arg{extract2}.
+%   If the key was missing from the property list, then \meta{extract1}
+%   is the full property list, \meta{value} is \cs{q_no_value}, and
+%   \meta{extract2} is empty.
+%   Otherwise, \meta{extract1} is the part of the property list before the
+%   \meta{key}, and has the structure of a property list, \meta{value} is
+%   the value corresponding to the \meta{key}, and \meta{extract2} (the part
+%   after the \meta{key}) is missing the leading \cs{q_prop}.
+%    \begin{macrocode}
+\cs_set_protected:Npn \prop_split:Nnn #1#2#3
+  {
+    \prop_split:NnTF #1 {#2}
+    {#3}
+    { \exp_args:Nno \use:n {#3} {#1} { \q_no_value } { } }
+  }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\prop_del:Nn, \prop_del:NV, \prop_del:cn, \prop_del:cV}
+% \begin{macro}{\prop_gdel:Nn, \prop_gdel:NV, \prop_gdel:cn, \prop_gdel:cV}
+% \begin{macro}[aux]{\prop_del_aux:NNnnn}
+%   Deleting from a property starts by splitting the list.
+%   If the key is present in the property list, the returned value is ignored.
+%   If the key is missing, nothing happens.
+%    \begin{macrocode}
+\cs_new_protected:Npn \prop_del:Nn #1#2
+  { \prop_split:NnTF #1 {#2} { \prop_del_aux:NNnnn \tl_set:Nn #1 } { } }
+\cs_new_protected:Npn \prop_gdel:Nn #1#2
+  { \prop_split:NnTF #1 {#2} { \prop_del_aux:NNnnn \tl_gset:Nn #1 } { } }
+\cs_new_protected:Npn \prop_del_aux:NNnnn #1#2#3#4#5
+  { #1 #2 { #3 #5 } }
+\cs_generate_variant:Nn \prop_del:Nn  {     NV }
+\cs_generate_variant:Nn \prop_del:Nn  { c , cV }
+\cs_generate_variant:Nn \prop_gdel:Nn {     NV }
+\cs_generate_variant:Nn \prop_gdel:Nn { c , cV }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}
+%   {
+%     \prop_get:NnN, \prop_get:NVN, \prop_get:NoN,
+%     \prop_get:cnN, \prop_get:cVN, \prop_get:NoN
+%   }
+% \begin{macro}[aux]{\prop_get_aux:Nnnn}
+%   Getting an item from a list is very easy: after splitting,
+%   if the key is in the property list, just set the token list variable
+%   to the return value, otherwise to \cs{q_no_value}.
+%    \begin{macrocode}
+\cs_new_protected:Npn \prop_get:NnN #1#2#3
+  {
+    \prop_split:NnTF #1 {#2}
+      { \prop_get_aux:Nnnn #3 }
+      { \tl_set:Nn #3 { \q_no_value } }
+  }
+\cs_new_protected:Npn \prop_get_aux:Nnnn #1#2#3#4
+  { \tl_set:Nn #1 {#3} }
+\cs_generate_variant:Nn \prop_get:NnN {     NV , No }
+\cs_generate_variant:Nn \prop_get:NnN { c , cV , co }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\prop_pop:NnN, \prop_pop:NoN, \prop_pop:cnN, \prop_pop:coN}
+% \begin{macro}{\prop_gpop:NnN, \prop_gpop:NoN, \prop_gpop:cnN, \prop_gpop:coN}
+% \begin{macro}[aux]{\prop_pop_aux:NNNnnn}
+%   Popping a value also starts by doing the split.
+%   If the key is present, save the value in the token list and update the
+%   property list as when deleting.
+%   If the key is missing, save \cs{q_no_value} in the token list.
+%    \begin{macrocode}
+\cs_new_protected:Npn \prop_pop:NnN #1#2#3
+  {
+    \prop_split:NnTF #1 {#2}
+      { \prop_pop_aux:NNNnnn \tl_set:Nn #1 #3 }
+      { \tl_set:Nn #3 { \q_no_value } }
+  }
+\cs_new_protected:Npn \prop_gpop:NnN #1#2#3
+  {
+    \prop_split:NnTF #1 {#2}
+      { \prop_pop_aux:NNNnnn \tl_gset:Nn #1 #3 }
+      { \tl_set:Nn #3 { \q_no_value } }
+  }
+\cs_new_protected:Npn \prop_pop_aux:NNNnnn #1#2#3#4#5#6
+  {
+    \tl_set:Nn #3 {#5}
+    #1 #2 { #4 #6 }
+  }
+\cs_generate_variant:Nn \prop_pop:NnN  {     No }
+\cs_generate_variant:Nn \prop_pop:NnN  { c , co }
+\cs_generate_variant:Nn \prop_gpop:NnN {     No }
+\cs_generate_variant:Nn \prop_gpop:NnN { c , co }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}
+%   {
+%     \prop_put:Nnn, \prop_put:NnV, \prop_put:Nno, \prop_put:Nnx,
+%     \prop_put:NVn, \prop_put:NVV, \prop_put:Non, \prop_put:Noo,
+%     \prop_put:cnn, \prop_put:cnV, \prop_put:cno, \prop_put:cnx,
+%     \prop_put:cVn, \prop_put:cVV, \prop_put:con, \prop_put:coo
+%   }
+% \begin{macro}
+%   {
+%     \prop_gput:Nnn, \prop_gput:NnV, \prop_gput:Nno, \prop_gput:Nnx,
+%     \prop_gput:NVn, \prop_gput:NVV, \prop_gput:Non, \prop_gput:Noo,
+%     \prop_gput:cnn, \prop_gput:cnV, \prop_gput:cno, \prop_gput:cnx,
+%     \prop_gput:cVn, \prop_gput:cVV, \prop_gput:con, \prop_gput:coo
+%   }
+% \begin{macro}[aux]{\prop_put_aux:NNnnnnn}
+%   Putting a key--value pair in a property list starts by splitting
+%   to remove any existing value. The property list is then
+%   reconstructed with the two remaining parts |#5| and |#7| first,
+%   followed by the new or updated entry.
+%    \begin{macrocode}
+\cs_new_protected:Npn \prop_put:Nnn { \prop_put_aux:NNnn \tl_set:Nx }
+\cs_new_protected:Npn \prop_gput:Nnn { \prop_put_aux:NNnn \tl_gset:Nx }
+\cs_new_protected:Npn \prop_put_aux:NNnn #1#2#3#4
+  {
+    \prop_split:Nnn #2 {#3} { \prop_put_aux:NNnnnnn #1 #2 {#3} {#4} }
+  }
+\cs_new_protected:Npn \prop_put_aux:NNnnnnn #1#2#3#4#5#6#7
+  {
+    #1 #2
+      {
+        \exp_not:n { #5 #7 }
+        \tl_to_str:n {#3} \exp_not:n { \q_prop {#4} \q_prop }
+      }
+  }
+\cs_generate_variant:Nn \prop_put:Nnn
+  {     NnV , Nno , Nnx , NV , NVV , No , Noo }
+\cs_generate_variant:Nn \prop_put:Nnn
+  { c , cnV , cno , cnx , cV , cVV , co , coo }
+\cs_generate_variant:Nn \prop_gput:Nnn
+  {     NnV , Nno , Nnx , NV , NVV , No , Noo }
+\cs_generate_variant:Nn \prop_gput:Nnn
+  { c , cnV , cno , cnx , cV , cVV , co , coo }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\prop_put_if_new:Nnn, \prop_put_if_new:cnn}
+% \begin{macro}{\prop_gput_if_new:Nnn, \prop_gput_if_new:cnn}
+%   Adding conditionally also splits. If the key is already present,
+%   the three brace groups given by \cs{prop_split:NnTF} are removed.
+%   If the key is new, then the value is added, being careful to
+%   convert the key to a string using \cs{tl_to_str:n}.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \prop_put_if_new:Nnn
+  { \prop_put_if_new_aux:NNnn \tl_put_right:Nx }
+\cs_new_protected_nopar:Npn \prop_gput_if_new:Nnn
+  { \prop_put_if_new_aux:NNnn \tl_gput_right:Nx }
+\cs_new_protected:Npn \prop_put_if_new_aux:NNnn #1#2#3#4
+  {
+    \prop_split:NnTF #2 {#3}
+      { \use_none:nnn }
+      {
+        #1 #2
+          { \tl_to_str:n {#3} \exp_not:n { \q_prop {#4} \q_prop } }
+      }
+  }
+\cs_generate_variant:Nn \prop_put_if_new:Nnn  { c }
+\cs_generate_variant:Nn \prop_gput_if_new:Nnn { c }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \subsection{Property list conditionals}
+%
+% \begin{macro}[pTF]{\prop_if_empty:N, \prop_if_empty:c}
+%   The test here uses \cs{c_empty_prop} as it is not really empty!
+%    \begin{macrocode}
+\prg_new_conditional:Npnn \prop_if_empty:N #1 { p, T , F , TF }
+  {
+    \if_meaning:w #1 \c_empty_prop
+      \prg_return_true:
+    \else:
+      \prg_return_false:
+    \fi:
+  }
+\cs_generate_variant:Nn \prop_if_empty_p:N {c}
+\cs_generate_variant:Nn \prop_if_empty:NTF {c}
+\cs_generate_variant:Nn \prop_if_empty:NT  {c}
+\cs_generate_variant:Nn \prop_if_empty:NF  {c}
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}[pTF]
+%   {
+%     \prop_if_in:Nn, \prop_if_in:NV, \prop_if_in:No,
+%     \prop_if_in:cn, \prop_if_in:cV, \prop_if_in:co
+%   }
+% \begin{macro}[aux]{\prop_if_in_aux:w}
+%   Testing expandably if a key is in a property list
+%   requires to go through the key--value pairs one by one.
+%   This is rather slow, and a faster test would be
+%   \begin{verbatim}
+%     \prg_new_protected_conditional:Npnn \prop_if_in:Nn #1 #2
+%       {
+%         \prop_split:NnTF #1 {#2}
+%           {
+%             \prg_return_true:
+%             \use_none:nnn
+%           }
+%         { \prg_return_false: }
+%       }
+%   \end{verbatim}
+%   but \cs{prop_split:NnTF} is non-expandable.
+%
+%   Instead, the key is compared to each key in turn using \cs{str_if_eq:nn},
+%   which is expandable. The mapping is stopped using |A|, which cannot appear
+%   within a key of the property list, since keys are strings.
+%   Here, \cs{prop_map_function:NN} is not sufficient for the mapping,
+%   since it can only map a single token, and cannot carry the key that
+%   is searched for.
+%    \begin{macrocode}
+\prg_new_conditional:Npnn \prop_if_in:Nn #1#2 { p , T , F , TF }
+  {
+    \exp_last_unbraced:Noo \prop_if_in_aux:nwn
+      { \tl_to_str:n {#2} } #1
+    A \q_prop { } \q_stop
+  }
+\cs_new:Npn \prop_if_in_aux:nwn #1 \q_prop #2 \q_prop #3
+  {
+    \if_catcode:w A #2
+       \prg_return_false:
+       \exp_after:wN \use_none_delimit_by_q_stop:w
+    \fi:
+    \str_if_eq:nnT {#1} {#2}
+      {
+        \prg_return_true:
+        \use_none_delimit_by_q_stop:w
+      }
+    \prop_if_in_aux:nwn {#1}
+  }
+\cs_generate_variant:Nn \prop_if_in_p:Nn {     NV , No }
+\cs_generate_variant:Nn \prop_if_in_p:Nn { c , cV , co }
+\cs_generate_variant:Nn \prop_if_in:NnT  {     NV , No }
+\cs_generate_variant:Nn \prop_if_in:NnT  { c , cV , co }
+\cs_generate_variant:Nn \prop_if_in:NnF  {     NV , No }
+\cs_generate_variant:Nn \prop_if_in:NnF  { c , cV , co }
+\cs_generate_variant:Nn \prop_if_in:NnTF {     NV , No }
+\cs_generate_variant:Nn \prop_if_in:NnTF { c , cV , co }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \subsection{Mapping to property lists}
+%
+% \begin{macro}
+%   {
+%     \prop_map_function:NN, \prop_map_function:Nc,
+%     \prop_map_function:cN, \prop_map_function:cc
+%   }
+% \begin{macro}[aux]{\prop_map_function_aux:Nwn}
+%   The fastest way to do a recursion here is to use an
+%   \cs{if_catcode:w} test: the keys are strings, and thus
+%   cannot match the marker |A| (which has catcode \enquote{letter}).
+%    \begin{macrocode}
+\cs_new_nopar:Npn \prop_map_function:NN #1#2
+  {
+    \exp_last_unbraced:NNo \prop_map_function_aux:Nwn #2
+      #1 A \q_prop { } \q_recursion_stop
+  }
+\cs_new:Npn \prop_map_function_aux:Nwn #1 \q_prop #2 \q_prop #3
+  {
+    \if_catcode:w A #2
+      \exp_after:wN \prop_map_break:
+    \fi:
+    #1 {#2} {#3}
+    \prop_map_function_aux:Nwn #1
+  }
+\cs_generate_variant:Nn \prop_map_function:NN {     Nc }
+\cs_generate_variant:Nn \prop_map_function:NN { c , cc }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{variable}{\g_prop_map_inline_int}
+%   A nesting counter for mapping.
+%    \begin{macrocode}
+\int_new:N \g_prop_map_inline_int
+%    \end{macrocode}
+% \end{variable}
+%
+% \begin{macro}{\prop_map_inline:Nn, \prop_map_inline:cn}
+%   Mapping in line requires a nesting level counter.
+%    \begin{macrocode}
+\cs_new_protected:Npn \prop_map_inline:Nn #1#2
+  {
+    \int_gincr:N \g_prop_map_inline_int
+    \cs_gset:cpn { prop_map_inline_ \int_use:N \g_prop_map_inline_int :nn }
+      ##1##2 {#2}
+    \prop_map_function:Nc #1
+      { prop_map_inline_ \int_use:N \g_prop_map_inline_int :nn }
+    \int_gdecr:N \g_prop_map_inline_int
+  }
+\cs_generate_variant:Nn \prop_map_inline:Nn { c }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\prop_map_break:}
+%   Breaking the map function simply means removing everything up to
+%   the \cs{q_stop} marker.
+%    \begin{macrocode}
+\cs_new_eq:NN \prop_map_break: \use_none_delimit_by_q_recursion_stop:w
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\prop_map_break:n}
+%   The same idea for using one set of tokens.
+%    \begin{macrocode}
+\cs_new_eq:NN \prop_map_break:n \use_i_delimit_by_q_recursion_stop:nw
+%    \end{macrocode}
+% \end{macro}
+%
+% \subsection{Viewing property lists}
+%
+% \begin{variable}{\l_prop_show_tl}
+%   Used to store the material for display.
+%    \begin{macrocode}
+\tl_new:N \l_prop_show_tl
+%    \end{macrocode}
+% \end{variable}
+%
+% \begin{macro}{\prop_show:N, \prop_show:c}
+% \begin{macro}[aux]{\prop_show_aux:n}
+% \begin{macro}[aux]{\prop_show_aux:w}
+%   The aim of the mapping here is to create a token list containing the
+%   formatted property list. The very first item needs the new line and
+%   \verb*|> | removing, which is achieved using a \texttt{w}-type auxiliary.
+%   To avoid a low-level \TeX{} error if there is an empty property list, a
+%   simple test is used to keep the output \enquote{clean}.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \prop_show:N #1
+  {
+    \prop_if_empty:NTF #1
+      {
+        \iow_term:x { Property~list~\token_to_str:N #1 \c_space_tl is~empty }
+        \tl_show:n { }
+      }
+      {
+        \iow_term:x
+          {
+            Property~list~\token_to_str:N #1 \c_space_tl
+            contains~the~pairs~(without~outer~braces):
+          }
+        \tl_set:Nx \l_prop_show_tl
+          { \prop_map_function:NN #1 \prop_show_aux:nn }
+        \tl_show:n \exp_after:wN \exp_after:wN \exp_after:wN
+          { \exp_after:wN \prop_show_aux:w \l_prop_show_tl }
+      }
+  }
+\cs_new:Npn \prop_show_aux:nn #1#2
+  {
+    \iow_newline: > \c_space_tl \c_space_tl
+    \iow_char:N \{ #1 \iow_char:N \}
+    \c_space_tl \c_space_tl => \c_space_tl \c_space_tl
+    \iow_char:N \{ \exp_not:n {#2} \iow_char:N \}
+  }
+\cs_new:Npn \prop_show_aux:w #1 > ~ { }
+\cs_generate_variant:Nn \prop_show:N { c }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \subsection{Experimental functions}
+%
+% \begin{macro}[TF]{\prop_get:NnN}
+% \begin{macro}[aux]{\prop_get_aux_true:Nnnn}
+%   Getting the value corresponding to a key, keeping track of whether
+%   the key was present or not, is implemented as a conditional (with
+%   side effects). If the key was absent, the token list is not altered.
+%    \begin{macrocode}
+\prg_new_protected_conditional:Npnn \prop_get:NnN #1#2#3 { T , F , TF }
+  {
+    \prop_split:NnTF #1 {#2}
+      { \prop_get_aux_true:Nnnn #3 }
+      { \prg_return_false: }
+  }
+\cs_new_protected:Npn \prop_get_aux_true:Nnnn #1#2#3#4
+  {
+    \tl_set:Nn #1 {#3}
+    \prg_return_true:
+  }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}[TF]{\prop_pop:NnN,\prop_gpop:NnN}
+% \begin{macro}[aux]{\prop_pop_aux_true:NNNnnn}
+%   Popping an item from a property list, keeping track of whether
+%   the key was present or not, is implemented as a conditional.
+%   If the key was missing, neither the property list, nor the token
+%   list are altered. Otherwise, \cs{prg_return_true:} is used after
+%   the assignments.
+%    \begin{macrocode}
+\prg_new_protected_conditional:Npnn \prop_pop:NnN #1#2#3 {T,F,TF}
+  {
+    \prop_split:NnTF #1 {#2}
+      { \prop_pop_aux_true:NNNnnn \tl_set:Nn #1 #3 }
+      { \prg_return_false: }
+  }
+\prg_new_protected_conditional:Npnn \prop_gpop:NnN #1#2#3 {T,F,TF}
+  {
+    \prop_split:NnTF #1 {#2}
+      { \prop_pop_aux_true:NNNnnn \tl_gset:Nn #1 #3 }
+      { \prg_return_false: }
+  }
+\cs_new_protected:Npn \prop_pop_aux_true:NNNnnn #1#2#3#4#5#6
+  {
+    \tl_set:Nn #3 {#5}
+    #1 #2 { #4 #6 }
+    \prg_return_true:
+  }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}[EXP]{\prop_map_tokens:Nn}
+% \begin{macro}[aux]{\prop_map_tokens_aux:nwn}
+%   The mapping grabs one key--value pair at a time, and stops when
+%   reaching the marker key |A|, with catcode \enquote{letter}, which
+%   cannot appear in normal keys since those are strings. The odd
+%   construction |\use:n {#1}| allows |#1| to contain any token.
+%    \begin{macrocode}
+\cs_new:Npn \prop_map_tokens:Nn #1#2
+  {
+    \exp_last_unbraced:Nno \prop_map_tokens_aux:nwn {#2} #1
+    A \q_prop { } \q_recursion_stop
+  }
+\cs_new:Npn \prop_map_tokens_aux:nwn #1 \q_prop #2 \q_prop #3
+  {
+    \if_catcode:w A #2
+       \exp_after:wN \prop_map_break:
+    \fi:
+    \use:n {#1} {#2} {#3}
+    \prop_map_tokens_aux:nwn {#1}
+  }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}[EXP]{\prop_get:Nn}
+% \begin{macro}[aux]{\prop_get_aux:nnn}
+%   Getting expandably the value corresponding to a key in a property list
+%   is a simple instance of mapping some tokens.
+%   Map the function \cs{prop_get_aux:nnn} which takes as its three
+%   arguments the \meta{key} that we are looking for, the current
+%   \meta{key} and the current \meta{value}. If the \meta{keys} match,
+%   the \meta{value} is returned. If none of the keys match, this expands
+%   to nothing.
+%    \begin{macrocode}
+\cs_new:Npn \prop_get:Nn #1 #2
+  { \prop_map_tokens:Nn #1 { \prop_get_aux:nnn {#2} } }
+\cs_new:Npn \prop_get_aux:nnn #1 #2 #3
+  { \str_if_eq:nnT {#1} {#2} { \prop_map_break:n {#3} } }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \subsection{Deprecated interfaces}
+%
+% Deprecated on 2011-05-27, for removal by 2011-08-31.
+%
+% \begin{macro}{\prop_display:N, \prop_display:c}
+%   An older name for \cs{prop_show:N}.
+%    \begin{macrocode}
+\cs_new_eq:NN \prop_display:N \prop_show:N
+\cs_new_eq:NN \prop_display:c \prop_show:c
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\prop_gget:NnN, \prop_gget:NVN, \prop_gget:cnN, \prop_gget:cVN}
+% \begin{macro}[aux]{\prop_gget_aux:Nnnn}
+%   Getting globally is no longer supported: this is a conceptual change, so
+%   the necessary code for the transition is provided directly.
+%    \begin{macrocode}
+\cs_new_protected:Npn \prop_gget:NnN #1#2#3
+  { \prop_split:Nnn #1 {#2} { \prop_gget_aux:Nnnn #3 } }
+\cs_new_protected:Npn \prop_gget_aux:Nnnn #1#2#3#4
+  { \tl_gset:Nn #1 {#3} }
+\cs_generate_variant:Nn \prop_gget:NnN {     NV }
+\cs_generate_variant:Nn \prop_gget:NnN { c , cV }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\prop_get_gdel:NnN}
+%   This name seems very odd.
+%    \begin{macrocode}
+\cs_new_eq:NN \prop_get_gdel:NnN \prop_gpop:NnN
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}[TF]{\prop_if_in:cc}
+%   A hang-over from an ancient implementation
+%    \begin{macrocode}
+\cs_generate_variant:Nn \prop_if_in:NnT  { cc }
+\cs_generate_variant:Nn \prop_if_in:NnF  { cc }
+\cs_generate_variant:Nn \prop_if_in:NnTF { cc }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\prop_gput:ccx}
+%   Another one.
+%    \begin{macrocode}
+\cs_generate_variant:Nn \prop_gput:Nnn { ccx }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}[pTF]
+%   {\prop_if_eq:NN, \prop_if_eq:Nc, \prop_if_eq:cN, \prop_if_eq:cc}
+%   These ones do no even make sense!
+%    \begin{macrocode}
+\prg_new_eq_conditional:NNn \prop_if_eq:NN \tl_if_eq:NN { p , T , F , TF }
+\prg_new_eq_conditional:NNn \prop_if_eq:cN \tl_if_eq:cN { p , T , F , TF }
+\prg_new_eq_conditional:NNn \prop_if_eq:Nc \tl_if_eq:Nc { p , T , F , TF }
+\prg_new_eq_conditional:NNn \prop_if_eq:cc \tl_if_eq:cc { p , T , F , TF }
+%    \end{macrocode}
+% \end{macro}
+%
+%    \begin{macrocode}
+%</initex|package>
+%    \end{macrocode}
+%
+% \end{implementation}
+%
+% \PrintIndex
\ No newline at end of file
diff --git a/Master/texmf-dist/source/latex/l3kernel/l3quark.dtx b/Master/texmf-dist/source/latex/l3kernel/l3quark.dtx
new file mode 100644
index 00000000000..2326fd4519d
--- /dev/null
+++ b/Master/texmf-dist/source/latex/l3kernel/l3quark.dtx
@@ -0,0 +1,486 @@
+% \iffalse meta-comment
+%
+%% File: l3quark.dtx Copyright (C) 1990-2011 The LaTeX3 Project
+%%
+%% It may be distributed and/or modified under the conditions of the
+%% LaTeX Project Public License (LPPL), either version 1.3c of this
+%% license or (at your option) any later version.  The latest version
+%% of this license is in the file
+%%
+%%    http://www.latex-project.org/lppl.txt
+%%
+%% This file is part of the "expl3 bundle" (The Work in LPPL)
+%% and all files in that bundle must be distributed together.
+%%
+%% The released version of this bundle is available from CTAN.
+%%
+%% -----------------------------------------------------------------------
+%%
+%% The development version of the bundle can be found at
+%%
+%%    http://www.latex-project.org/svnroot/experimental/trunk/
+%%
+%% for those people who are interested.
+%%
+%%%%%%%%%%%
+%% NOTE: %%
+%%%%%%%%%%%
+%%
+%%   Snapshots taken from the repository represent work in progress and may
+%%   not work or may contain conflicting material!  We therefore ask
+%%   people _not_ to put them into distributions, archives, etc. without
+%%   prior consultation with the LaTeX3 Project.
+%%
+%% -----------------------------------------------------------------------
+%
+%<*driver|package>
+\RequirePackage{l3names}
+\GetIdInfo$Id: l3quark.dtx 2478 2011-06-19 21:34:23Z joseph $
+  {L3 Experimental quarks}
+%</driver|package>
+%<*driver>
+\documentclass[full]{l3doc}
+\begin{document}
+  \DocInput{\jobname.dtx}
+\end{document}
+%</driver>
+% \fi
+%
+% \title{^^A
+%   The \pkg{l3quark} package\\ Quarks^^A
+%   \thanks{This file describes v\ExplFileVersion,
+%      last revised \ExplFileDate.}^^A
+% }
+%
+% \author{^^A
+%  The \LaTeX3 Project\thanks
+%    {^^A
+%      E-mail:
+%        \href{mailto:latex-team@latex-project.org}
+%          {latex-team@latex-project.org}^^A
+%    }^^A
+% }
+%
+% \date{Released \ExplFileDate}
+%
+% \maketitle
+%
+% \begin{documentation}
+%
+% A special type of constants in \LaTeX3 are \enquote{quarks}. These are
+% control
+% sequences that expand to themselves and should therefore \emph{never} be
+% executed directly in the code. This would result in an endless loop!
+%
+% They are meant to be used as delimiter is weird functions (for
+% example as the stop token (\emph{i.e.}~\cs{q_stop}). They also permit the
+% following ingenious trick: when you pick up a token in a temporary,
+% and you want to know whether you have picked up a particular quark,
+% all you have to do is compare the temporary to the quark using
+% \cs{if_meaning:w}. A set of special quark testing functions is set up
+% below. All the quark testing functions are expandable although the
+% ones testing only single tokens are much faster.
+%
+% By convention all constants of type quark start out with |\q_|.
+%
+% \section{Defining quarks}
+%
+% \begin{function}{\quark_new:N}
+%   \begin{syntax}
+%     \cs{quark_new:N} \meta{quark}
+%   \end{syntax}
+%   Creates a new \meta{quark} which expands only to \meta{quark}.
+%   The \meta{quark} will be defined globally, and an error message
+%   will be raised if the name was already taken.
+% \end{function}
+%
+% \begin{variable}{\q_stop}
+%   Used as a marker for delimited arguments, such as
+%   \begin{verbatim}
+%     \cs_set:Npn \tmp:w #1#2 \q_stop {#1}
+%   \end{verbatim}
+% \end{variable}
+%
+% \begin{variable}{\q_mark}
+%   Used as a marker for delimited arguments when \cs{q_stop} is
+%   already in use.
+% \end{variable}
+%
+% \begin{variable}{\q_nil}
+%   Quark to mark a null value in structured variables or functions. Used
+%   as an end delimiter when this may itself may need to be tested
+%   (in contrast to \cs{q_stop}, which is only ever used as a delimiter).
+% \end{variable}
+%
+% \begin{variable}{\q_no_value}
+%   A canonical value for a missing value, when one is requested from
+%   a data structure. This is therefore used as a \enquote{return} value
+%   by functions such as \cs{prop_get:NnN} if there is no data to
+%   return.
+% \end{variable}
+%
+% \section{Quark tests}
+%
+% The method used to define quarks means that the single token (\texttt{N})
+% tests are faster than the multi-token (\texttt{n}) tests. The later
+% should therefore only be used when the argument can definitely take
+% more than a single token.
+%
+% \begin{function}[EXP,pTF]{\quark_if_nil:N}
+%   \begin{syntax}
+%     \cs{quark_if_nil_p:N} \meta{token}
+%     \cs{quark_if_nil:NTF} \meta{token} \Arg{true code} \Arg{false code}
+%   \end{syntax}
+%   Tests if the \meta{token} is equal to \cs{q_nil}. The branching
+%   versions then leave either \meta{true code} or \meta{false code} in
+%   the input stream, as appropriate to the truth of the test and the
+%   variant of the function chosen. The logical truth of the test is
+%   left in the input stream by the predicate version.
+% \end{function}
+%
+% \begin{function}[EXP,pTF]{\quark_if_nil:n, \quark_if_nil:o, \quark_if_nil:V}
+%   \begin{syntax}
+%     \cs{quark_if_nil_p:n} \Arg{token list}
+%     \cs{quark_if_nil:nTF} \Arg{token list} \Arg{true code} \Arg{false code}
+%   \end{syntax}
+%   Tests if the \meta{token list} contains only \cs{q_nil} (distinct
+%   from \meta{token list} being empty or containing \cs{q_nil} plus one
+%   or more other tokens). The branching versions then leave either
+%   \meta{true code} or \meta{false code} in the input stream, as
+%   appropriate to the truth of the test and the variant of the
+%   function chosen. The logical truth of the test is left in the input
+%   stream by the predicate version.
+% \end{function}
+%
+% \begin{function}[EXP,pTF]{\quark_if_no_value:N, \quark_if_no_value:c}
+%   \begin{syntax}
+%     \cs{quark_if_no_value_p:N} \meta{token}
+%     \cs{quark_if_no_value:NTF} \meta{token} \Arg{true code}
+%     ~~\Arg{false code}
+%   \end{syntax}
+%   Tests if the \meta{token} is equal to \cs{q_no_value}. The branching
+%   versions then leave either \meta{true code} or \meta{false code} in
+%   the input stream, as appropriate to the truth of the test and the
+%   variant of the function chosen. The logical truth of the test is
+%   left in the input stream by the predicate version.
+% \end{function}
+%
+% \begin{function}[EXP,pTF]{\quark_if_no_value:n}
+%   \begin{syntax}
+%      \cs{quark_if_no_value_p:n} \Arg{token list}
+%     \cs{quark_if_no_value:nTF} \Arg{token list} \Arg{true code}
+%     ~~\Arg{false code}
+%   \end{syntax}
+%   Tests if the \meta{token list} contains only \cs{q_no_value}
+%   (distinct from \meta{token list} being empty or containing
+%   \cs{q_no_value} plus one or more other tokens). The branching
+%   versions then leave either \meta{true code} or \meta{false code} in
+%   the input stream, as appropriate to the truth of the test and the
+%   variant of the function chosen. The logical truth of the test is
+%   left in the input stream by the predicate version.
+% \end{function}
+%
+% \section{Recursion}
+%
+% This module provides a uniform interface to intercepting and
+% terminating loops as when one is doing tail recursion. The building
+% blocks follow below.
+%
+% \begin{variable}{\q_recursion_tail}
+%   This quark is appended to the data structure in question and
+%   appears as a real element there. This means it gets any list
+%   separators around it.
+% \end{variable}
+%
+% \begin{variable}{\q_recursion_stop}
+%   This quark is added \emph{after} the data structure. Its purpose
+%   is to make it possible to terminate the recursion at any point
+%   easily.
+% \end{variable}
+%
+% \begin{function}{\quark_if_recursion_tail_stop:N}
+%   \begin{syntax}
+%     \cs{quark_if_recursion_tail_stop:N} \Arg{token}
+%   \end{syntax}
+%   Tests if \meta{token} contains only the marker
+%   \cs{q_recursion_tail}, and if so terminates the recursion this is
+%   part of using \cs{use_none_delimit_by_q_recursion_stop:w}. The
+%   recursion input must include the marker tokens \cs{q_recursion_tail}
+%   and \cs{q_recursion_stop} as the last two items.
+% \end{function}
+%
+% \begin{function}
+%   {\quark_if_recursion_tail_stop:n, \quark_if_recursion_tail_stop:o}
+%   \begin{syntax}
+%     \cs{quark_if_recursion_tail_stop:n} \Arg{tokens}
+%   \end{syntax}
+%   Tests if \meta{tokens} consists of the single token
+%   \cs{q_recursion_tail}, and if so terminates the recursion this is
+%   part of using \cs{use_none_delimit_by_q_recursion_stop:w}. The
+%   recursion input must include the marker tokens \cs{q_recursion_tail}
+%   and \cs{q_recursion_stop} as the last two items.
+% \end{function}
+%
+% \begin{function}{\quark_if_recursion_tail_stop_do:Nn}
+%   \begin{syntax}
+%     \cs{quark_if_recursion_tail_stop_do:nn} \Arg{token} \Arg{insertion}
+%   \end{syntax}
+%   Tests if \meta{token} contains only the marker
+%   \cs{q_recursion_tail}, and if so terminates the recursion this is
+%   part of using \cs{use_none_delimit_by_q_recursion_stop:w}. The
+%   recursion input must include the marker tokens \cs{q_recursion_tail}
+%   and \cs{q_recursion_stop} as the last two items. The \meta{insertion}
+%   code is then added to the input stream after the recursion has
+%   ended.
+% \end{function}
+%
+% \begin{function}
+%   {\quark_if_recursion_tail_stop_do:nn, \quark_if_recursion_tail_stop_do:on}
+%   \begin{syntax}
+%     \cs{quark_if_recursion_tail_stop_do:nn} \Arg{tokens} \Arg{insertion}
+%   \end{syntax}
+%   Tests if \meta{tokens} consists of the single token
+%   \cs{q_recursion_tail}, and if so terminates the recursion this is
+%   part of using \cs{use_none_delimit_by_q_recursion_stop:w}. The
+%   recursion input must include the marker tokens \cs{q_recursion_tail}
+%   and \cs{q_recursion_stop} as the last two items. The \meta{insertion}
+%   code is then added to the input stream after the recursion has
+%   ended.
+% \end{function}
+%
+% \section{Internal quark functions}
+%
+% \begin{function}{\use_none_delimit_by_q_recursion_stop:w}
+%   \begin{syntax}
+%     \cs{use_none_delimit_by_q_recursion_stop:w}
+%     ~~\meta{tokens} \cs{q_recursion_stop}
+%   \end{syntax}
+%   Used to prematurely terminate a recursion using \cs{q_recursion_stop}
+%   as the end marker, removing any remaining \meta{tokens} from the
+%   input stream.
+% \end{function}
+%
+% \begin{function}{\use_i_delimit_by_q_recursion_stop:nw}
+%   \begin{syntax}
+%     \cs{use_i_delimit_by_q_recursion_stop:nw} \Arg{insertion}
+%     ~~\meta{tokens} \cs{q_recursion_stop}
+%   \end{syntax}
+%   Used to prematurely terminate a recursion using \cs{q_recursion_stop}
+%   as the end marker, removing any remaining \meta{tokens} from the
+%   input stream. The \meta{insertion} is then made into the input
+%   stream after the end of the recursion.
+% \end{function}
+%
+% \end{documentation}
+%
+% \begin{implementation}
+%
+% \section{\pkg{l3quark} implementation}
+%
+% \TestFiles{m3quark001.lvt}
+%
+%    \begin{macrocode}
+%<*initex|package>
+%    \end{macrocode}
+%
+%    \begin{macrocode}
+%<*package>
+\ProvidesExplPackage
+  {\ExplFileName}{\ExplFileDate}{\ExplFileVersion}{\ExplFileDescription}
+\package_check_loaded_expl:
+%</package>
+%    \end{macrocode}
+%
+% \begin{macro}{\quark_new:N}
+% \UnitTested
+%    Allocate a new quark.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \quark_new:N #1 { \tl_const:Nn #1 {#1} }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{variable}{\q_nil, \q_mark, \q_no_value, \q_stop}
+%   Some \enquote{public} quarks. \cs{q_stop} is an \enquote{end of
+%   argument} marker, \cs{q_nil} is a empty value and \cs{q_no_value}
+%   marks an empty argument.
+%    \begin{macrocode}
+\quark_new:N \q_nil
+\quark_new:N \q_mark
+\quark_new:N \q_no_value
+\quark_new:N \q_stop
+%    \end{macrocode}
+% \end{variable}
+%
+% \begin{variable}{\q_recursion_tail,\q_recursion_stop}
+%   Quarks for ending recursions. Only ever used there!
+%   \cs{q_recursion_tail} is appended to whatever list structure we are
+%   doing recursion on, meaning it is added as a proper list item with
+%   whatever list separator is in use.  \cs{q_recursion_stop} is placed
+%   directly after the list.
+%    \begin{macrocode}
+\quark_new:N \q_recursion_tail
+\quark_new:N \q_recursion_stop
+%    \end{macrocode}
+% \end{variable}
+%
+% \begin{macro}{\quark_if_recursion_tail_stop:N}
+% \UnitTested
+% \begin{macro}{\quark_if_recursion_tail_stop_do:Nn}
+% \UnitTested
+%   When doing recursions, it is easy to spend a lot of time testing if the
+%   end marker has been found. To avoid this, a dedicated end marker is used
+%   each time a recursion is set up. Thus if the marker is found everything
+%   can be wrapper up and finished off. The simple case is when the test
+%   can guarantee that only a single token is being tested. In this case,
+%   there is just a dedicated copy of the standard quark test. Both a gobbling
+%   version and one inserting end code are provided.
+%    \begin{macrocode}
+\cs_new:Npn \quark_if_recursion_tail_stop:N #1
+  {
+    \if_meaning:w #1 \q_recursion_tail
+      \exp_after:wN \use_none_delimit_by_q_recursion_stop:w
+    \fi:
+  }
+\cs_new:Npn \quark_if_recursion_tail_stop_do:Nn #1#2
+  {
+    \if_meaning:w #1 \q_recursion_tail
+      \exp_after:wN \use_i_delimit_by_q_recursion_stop:nw
+    \else:
+      \exp_after:wN \use_none:n
+    \fi:
+      {#2}
+  }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}
+%   {\quark_if_recursion_tail_stop:n,\quark_if_recursion_tail_stop:o}
+% \UnitTested
+% \begin{macro}
+%   {\quark_if_recursion_tail_stop_do:nn,\quark_if_recursion_tail_stop_do:on}
+% \UnitTested
+% \begin{macro}[aux]{\quark_if_recursion_tail_aux:w}
+%   The same idea applies when testing multiple tokens, but here a little more
+%   care is needed. It is possible that |#1| might be something like
+%   |{{{a}}}| or |{ab\iffalse}\fi|, which will therefore need to be tested
+%   in a detokenized manner. The way that this is done is using
+%   \cs{if_catcode:w}, with the idea being that this test will be \texttt{true}
+%   provided the auxiliary function returns nothing at all. If the auxiliary
+%   returns anything, it will be detokenized and so the test will be both
+%   \texttt{false} and safe.
+%    \begin{macrocode}
+\cs_new:Npn \quark_if_recursion_tail_stop:n #1
+  {
+    \if_catcode:w
+      A
+      \etex_detokenize:D \exp_after:wN
+        {
+          \quark_if_recursion_tail_aux:w #1 \q_recursion_stop
+            \q_recursion_tail \q_recursion_stop \q_stop
+        }
+      A
+      \exp_after:wN \use_none_delimit_by_q_recursion_stop:w
+    \fi:
+  }
+\cs_new:Npn \quark_if_recursion_tail_stop_do:nn #1#2
+  {
+    \if_catcode:w
+      A
+      \etex_detokenize:D \exp_after:wN
+        {
+          \quark_if_recursion_tail_aux:w #1 \q_recursion_stop
+            \q_recursion_tail \q_recursion_stop \q_stop
+        }
+        A
+      \exp_after:wN \use_i_delimit_by_q_recursion_stop:nw
+    \else:
+      \exp_after:wN \use_none:n
+    \fi:
+      {#2}
+  }
+\cs_new:Npn \quark_if_recursion_tail_aux:w
+  #1 \q_recursion_tail #2 \q_recursion_stop #3 \q_stop
+  { #1 #2 }
+\cs_generate_variant:Nn \quark_if_recursion_tail_stop:n { o }
+\cs_generate_variant:Nn \quark_if_recursion_tail_stop_do:nn { o }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}[pTF]{\quark_if_nil:N}
+% \UnitTested
+% \begin{macro}[pTF]{\quark_if_no_value:N., \quark_if_no_value:c}
+% \UnitTested
+%   Here we test if we found a special quark as the first argument.
+%   We better start with \cs{q_no_value} as the first argument since
+%   the whole thing may otherwise loop if |#1| is wrongly given
+%   a string like |aabc| instead of a single token.\footnote{It may
+%   still loop in special circumstances however!}
+%    \begin{macrocode}
+\prg_new_conditional:Nnn \quark_if_nil:N { p, T , F , TF }
+  {
+    \if_meaning:w \q_nil #1
+      \prg_return_true:
+    \else:
+      \prg_return_false:
+    \fi:
+  }
+\prg_new_conditional:Nnn \quark_if_no_value:N { p, T , F , TF }
+  {
+    \if_meaning:w \q_no_value #1
+      \prg_return_true:
+    \else:
+      \prg_return_false:
+    \fi:
+  }
+\cs_generate_variant:Nn \quark_if_no_value_p:N { c }
+\cs_generate_variant:Nn \quark_if_no_value:NT  { c }
+\cs_generate_variant:Nn \quark_if_no_value:NF  { c }
+\cs_generate_variant:Nn \quark_if_no_value:NTF { c }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}[pTF]{\quark_if_nil:n, \quark_if_nil:V, \quark_if_nil:o}
+% \UnitTested
+% \begin{macro}[pTF]{\quark_if_no_value:n}
+% \UnitTested
+%   These are essentially \cs{str_if_eq:nn} tests but done directly.
+%    \begin{macrocode}
+\prg_new_conditional:Nnn \quark_if_nil:n { p, T , F , TF }
+  {
+    \if_int_compare:w \pdftex_strcmp:D
+      { \exp_not:N \q_nil } { \exp_not:n {#1} } = \c_zero
+      \prg_return_true:
+    \else:
+      \prg_return_false:
+    \fi:
+  }
+\prg_new_conditional:Nnn \quark_if_no_value:n { p, T , F , TF }
+  {
+    \if_int_compare:w \pdftex_strcmp:D
+      { \exp_not:N \q_no_value } { \exp_not:n {#1} } = \c_zero
+      \prg_return_true:
+    \else:
+      \prg_return_false:
+    \fi:
+  }
+\cs_generate_variant:Nn \quark_if_nil_p:n { V , o }
+\cs_generate_variant:Nn \quark_if_nil:nTF { V , o }
+\cs_generate_variant:Nn \quark_if_nil:nT  { V , o }
+\cs_generate_variant:Nn \quark_if_nil:nF  { V , o }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+%    \begin{macrocode}
+%</initex|package>
+%    \end{macrocode}
+%
+% \end{implementation}
+%
+% \PrintIndex
+%
+% \endinput
diff --git a/Master/texmf-dist/source/latex/l3kernel/l3seq.dtx b/Master/texmf-dist/source/latex/l3kernel/l3seq.dtx
new file mode 100644
index 00000000000..088165a02fc
--- /dev/null
+++ b/Master/texmf-dist/source/latex/l3kernel/l3seq.dtx
@@ -0,0 +1,1812 @@
+% \iffalse meta-comment
+%
+%% File: l3seq.dtx Copyright (C) 1990-2011 The LaTeX3 Project
+%%
+%% It may be distributed and/or modified under the conditions of the
+%% LaTeX Project Public License (LPPL), either version 1.3c of this
+%% license or (at your option) any later version.  The latest version
+%% of this license is in the file
+%%
+%%    http://www.latex-project.org/lppl.txt
+%%
+%% This file is part of the "expl3 bundle" (The Work in LPPL)
+%% and all files in that bundle must be distributed together.
+%%
+%% The released version of this bundle is available from CTAN.
+%%
+%% -----------------------------------------------------------------------
+%%
+%% The development version of the bundle can be found at
+%%
+%%    http://www.latex-project.org/svnroot/experimental/trunk/
+%%
+%% for those people who are interested.
+%%
+%%%%%%%%%%%
+%% NOTE: %%
+%%%%%%%%%%%
+%%
+%%   Snapshots taken from the repository represent work in progress and may
+%%   not work or may contain conflicting material!  We therefore ask
+%%   people _not_ to put them into distributions, archives, etc. without
+%%   prior consultation with the LaTeX3 Project.
+%%
+%% -----------------------------------------------------------------------
+%
+%<*driver|package>
+\RequirePackage{l3names}
+\GetIdInfo$Id: l3seq.dtx 2478 2011-06-19 21:34:23Z joseph $
+  {L3 Experimental sequences and stacks}
+%</driver|package>
+%<*driver>
+\documentclass[full]{l3doc}
+\begin{document}
+  \DocInput{\jobname.dtx}
+\end{document}
+%</driver>
+% \fi
+%
+% \title{^^A
+%   The \pkg{l3seq} package\\ Sequences and stacks^^A
+%   \thanks{This file describes v\ExplFileVersion,
+%      last revised \ExplFileDate.}^^A
+% }
+%
+% \author{^^A
+%  The \LaTeX3 Project\thanks
+%    {^^A
+%      E-mail:
+%        \href{mailto:latex-team@latex-project.org}
+%          {latex-team@latex-project.org}^^A
+%    }^^A
+% }
+%
+% \date{Released \ExplFileDate}
+%
+% \maketitle
+%
+% \begin{documentation}
+%
+% \LaTeX3 implements a \enquote{sequence} data type, which contain
+% an ordered list of entries which may contain any \meta{balanced text}.
+% It is possible to map functions to sequences such that the function
+% is applied to every item in the sequence.
+%
+% Sequences are also used to implement stack functions in \LaTeX3. This
+% is achieved using a number of dedicated stack functions.
+%
+% \section{Creating and initialising sequences}
+%
+% \begin{function}{\seq_new:N, \seq_new:c}
+%   \begin{syntax}
+%     \cs{seq_new:N} \meta{sequence}
+%   \end{syntax}
+%   Creates a new \meta{sequence} or raises an error if the name is
+%   already taken. The declaration is global. The \meta{sequence} will
+%   initially contain no items.
+% \end{function}
+%
+% \begin{function}{\seq_clear:N, \seq_clear:c}
+%   \begin{syntax}
+%     \cs{seq_clear:N} \meta{sequence}
+%   \end{syntax}
+%   Clears all items from the \meta{sequence} within the scope of
+%   the current \TeX{} group.
+% \end{function}
+%
+% \begin{function}{\seq_gclear:N, \seq_gclear:c}
+%   \begin{syntax}
+%     \cs{seq_gclear:N} \meta{sequence}
+%   \end{syntax}
+%   Clears all entries from the \meta{sequence} globally.
+% \end{function}
+%
+% \begin{function}{\seq_clear_new:N, \seq_clear_new:c}
+%   \begin{syntax}
+%     \cs{seq_clear_new:N} \meta{sequence}
+%   \end{syntax}
+%   If the \meta{sequence} already exists, clears it within the scope
+%   of the current \TeX{} group. If the \meta{sequence} is not defined,
+%   it will be created (using \cs{seq_new:N}). Thus the sequence is
+%   guaranteed to be available and clear within the current \TeX{}
+%   group. The \meta{sequence} will exist globally, but the content
+%   outside of the current \TeX{} group is not specified.
+% \end{function}
+%
+% \begin{function}{\seq_gclear_new:N, \seq_gclear_new:c}
+%   \begin{syntax}
+%     \cs{seq_gclear_new:N} \meta{sequence}
+%   \end{syntax}
+%   If the \meta{sequence} already exists, clears it globally. If the
+%   \meta{sequence} is not defined, it will be created (using
+%   \cs{seq_new:N}). Thus the sequence is guaranteed to be available
+%   and globally clear.
+% \end{function}
+%
+% \begin{function}
+%   {\seq_set_eq:NN, \seq_set_eq:cN, \seq_set_eq:Nc, \seq_set_eq:cc}
+%   \begin{syntax}
+%     \cs{seq_set_eq:NN} \meta{sequence1} \meta{sequence2}
+%   \end{syntax}
+%   Sets the content of \meta{sequence1} equal to that of
+%   \meta{sequence2}. This assignment is restricted to the current
+%   \TeX{} group level.
+% \end{function}
+%
+% \begin{function}
+%   {\seq_gset_eq:NN, \seq_gset_eq:cN, \seq_gset_eq:Nc, \seq_gset_eq:cc}
+%   \begin{syntax}
+%     \cs{seq_gset_eq:NN} \meta{sequence1} \meta{sequence2}
+%   \end{syntax}
+%   Sets the content of \meta{sequence1} equal to that of
+%   \meta{sequence2}. This assignment is global and so is not
+%   limited by the current \TeX{} group level.
+% \end{function}
+%
+% \begin{function}{\seq_concat:NNN, \seq_concat:ccc}
+%   \begin{syntax}
+%     \cs{seq_concat:NNN} \meta{sequence1} \meta{sequence2} \meta{sequence3}
+%   \end{syntax}
+%   Concatenates the content of \meta{sequence2} and \meta{sequence3}
+%   together and saves the result in \meta{sequence1}. The items in
+%   \meta{sequence2} will be placed at the left side of the new sequence.
+%   This  operation is local to the current \TeX{} group and will
+%   remove any existing content in \meta{sequence1}.
+% \end{function}
+%
+% \begin{function}{\seq_gconcat:NNN, \seq_gconcat:ccc}
+%   \begin{syntax}
+%     \cs{seq_gconcat:NNN} \meta{sequence1} \meta{sequence2} \meta{sequence3}
+%   \end{syntax}
+%   Concatenates the content of \meta{sequence2} and \meta{sequence3}
+%   together and saves the result in \meta{sequence1}. The items in
+%   \meta{sequence2}  will be placed at the left side of the new sequence.
+%   This operation is global and will remove any existing content in
+%   \meta{sequence1}.
+% \end{function}
+%
+% \section{Appending data to sequences}
+%
+% \begin{function}{
+%   \seq_put_left:Nn, \seq_put_left:NV, \seq_put_left:Nv,
+%   \seq_put_left:No, \seq_put_left:Nx,
+%   \seq_put_left:cn, \seq_put_left:cV, \seq_put_left:cv,
+%   \seq_put_left:co, \seq_put_left:cx
+% }
+%   \begin{syntax}
+%     \cs{seq_put_left:Nn} \meta{sequence} \Arg{item}
+%   \end{syntax}
+%   Appends the \meta{item} to the left of the \meta{sequence}.
+%   The assignment is restricted to the current \TeX{} group.
+% \end{function}
+%
+% \begin{function}{
+%   \seq_gput_left:Nn, \seq_gput_left:NV, \seq_gput_left:Nv,
+%   \seq_gput_left:No, \seq_gput_left:Nx,
+%   \seq_gput_left:cn, \seq_gput_left:cV, \seq_gput_left:cv,
+%   \seq_gput_left:co, \seq_gput_left:cx
+% }
+%   \begin{syntax}
+%     \cs{seq_gput_left:Nn} \meta{sequence} \Arg{item}
+%   \end{syntax}
+%   Appends the \meta{item} to the left of the \meta{sequence}.
+%   The assignment is global.
+% \end{function}
+%
+% \begin{function}{
+%   \seq_put_right:Nn, \seq_put_right:NV, \seq_put_right:Nv,
+%   \seq_put_right:No, \seq_put_right:Nx,
+%   \seq_put_right:cn, \seq_put_right:cV, \seq_put_right:cv,
+%   \seq_put_right:co, \seq_put_right:cx
+% }
+%   \begin{syntax}
+%     \cs{seq_put_right:Nn} \meta{sequence} \Arg{item}
+%   \end{syntax}
+%   Appends the \meta{item} to the right of the \meta{sequence}.
+%   The assignment is restricted to the current \TeX{} group.
+% \end{function}
+%
+% \begin{function}{
+%   \seq_gput_right:Nn, \seq_gput_right:NV, \seq_gput_right:Nv,
+%   \seq_gput_right:No, \seq_gput_right:Nx,
+%   \seq_gput_right:cn, \seq_gput_right:cV, \seq_gput_right:cv,
+%   \seq_gput_right:co, \seq_gput_right:cx
+% }
+%   \begin{syntax}
+%     \cs{seq_gput_right:Nn} \meta{sequence} \Arg{item}
+%   \end{syntax}
+%   Appends the \meta{item} to the right of the \meta{sequence}.
+%   The assignment is global.
+% \end{function}
+%
+% \section{Recovering items from sequences}
+%
+% Items can be recovered from either the left or the right of sequences.
+% For implementation reasons, the actions at the left of the sequence are
+% faster than those acting on the right. These functions all assign the
+% recovered material locally, \emph{i.e.}~setting the
+% \meta{token list variable} used with \cs{tl_set:Nn} and \emph{never}
+% \cs{tl_gset:Nn}.
+%
+% \begin{function}{\seq_get_left:NN, \seq_get_left:cN}
+%   \begin{syntax}
+%     \cs{seq_get_left:NN} \meta{sequence} \meta{token list variable}
+%   \end{syntax}
+%   Stores the left-most item from a \meta{sequence} in the
+%   \meta{token list variable} without removing it from the
+%   \meta{sequence}. The \meta{token list variable} is assigned locally.
+%   If \meta{sequence} is empty an error will be raised.
+% \end{function}
+%
+% \begin{function}{\seq_get_right:NN, \seq_get_right:cN}
+%   \begin{syntax}
+%     \cs{seq_get_right:NN} \meta{sequence} \meta{token list variable}
+%   \end{syntax}
+%   Stores the right-most item from a \meta{sequence} in the
+%   \meta{token list variable} without removing it from the
+%   \meta{sequence}. The \meta{token list variable} is assigned locally.
+%   If \meta{sequence} is empty an error will be raised.
+% \end{function}
+%
+% \begin{function}{\seq_pop_left:NN, \seq_pop_left:cN}
+%   \begin{syntax}
+%     \cs{seq_pop_left:NN} \meta{sequence} \meta{token list variable}
+%   \end{syntax}
+%   Pops the left-most item from a \meta{sequence} into the
+%   \meta{token list variable}, \emph{i.e.}~removes the item from the
+%   sequence and stores it in the \meta{token list variable}.
+%   Both of the variables are assigned locally. If \meta{sequence} is
+%   empty an error will be raised.
+% \end{function}
+%
+% \begin{function}{\seq_gpop_left:NN, \seq_gpop_left:cN}
+%   \begin{syntax}
+%     \cs{seq_gpop_left:NN} \meta{sequence} \meta{token list variable}
+%   \end{syntax}
+%   Pops the left-most item from a \meta{sequence} into the
+%   \meta{token list variable}, \emph{i.e.}~removes the item from the
+%   sequence and stores it in the \meta{token list variable}.
+%   The \meta{sequence} is modified globally, while the assignment of
+%   the \meta{token list variable} is local.
+%   If \meta{sequence} is  empty an error will be raised.
+% \end{function}
+%
+% \begin{function}{\seq_pop_right:NN, \seq_pop_right:cN}
+%   \begin{syntax}
+%     \cs{seq_pop_right:NN} \meta{sequence} \meta{token list variable}
+%   \end{syntax}
+%   Pops the right-most item from a \meta{sequence} into the
+%   \meta{token list variable}, \emph{i.e.}~removes the item from the
+%   sequence and stores it in in the \meta{token list variable}.
+%   Both of the variables are assigned locally. If \meta{sequence} is
+%   empty an error will be raised.
+% \end{function}
+%
+% \begin{function}{\seq_gpop_right:NN, \seq_gpop_right:cN}
+%   \begin{syntax}
+%     \cs{seq_gpop_right:NN} \meta{sequence} \meta{token list variable}
+%   \end{syntax}
+%   Pops the right-most item from a \meta{sequence} into the
+%   \meta{token list variable}, \emph{i.e.}~removes the item from the
+%   sequence and stores it in the \meta{token list variable}.
+%   The \meta{sequence} is modified globally, while the assignment of
+%   the \meta{token list variable} is local.
+%   If \meta{sequence} is  empty an error will be raised.
+% \end{function}
+%
+% \section{Modifying sequences}
+%
+%  While sequences are normally used as ordered lists, it may be
+%  necessary to modify the content. The functions here may be used
+%  to update sequences, while retaining the order of the unaffected
+%  entries.
+%
+% \begin{function}{\seq_remove_duplicates:N, \seq_remove_duplicates:c}
+%   \begin{syntax}
+%     \cs{seq_remove_duplicates:N} \meta{sequence}
+%   \end{syntax}
+%   Removes duplicate items from the \meta{sequence}, leaving the
+%   left most copy of each item in the \meta{sequence}.  The \meta{item}
+%   comparison takes place on a token basis, as for \cs{tl_if_eq:nn(TF)}.
+%   The removal is local to the current \TeX{} group.
+%   \begin{texnote}
+%     This function iterates through every item in the \meta{sequence} and
+%     does a comparison with the \meta{items} already checked. It is therefore
+%     relatively slow with large sequences.
+%   \end{texnote}
+% \end{function}
+%
+% \begin{function}{\seq_gremove_duplicates:N, \seq_gremove_duplicates:c}
+%   \begin{syntax}
+%     \cs{seq_gremove_duplicates:N} \meta{sequence}
+%   \end{syntax}
+%   Removes duplicate items from the \meta{sequence}, leaving the
+%   left most copy of each item in the \meta{sequence}.  The \meta{item}
+%   comparison takes place on a token basis, as for \cs{tl_if_eq:nn(TF)}.
+%   The removal is applied globally.
+%   \begin{texnote}
+%     This function iterates through every item in the \meta{sequence} and
+%     does a comparison with the \meta{items} already checked. It is therefore
+%     relatively slow with large sequences.
+%   \end{texnote}
+% \end{function}
+%
+% \begin{function}{\seq_remove_all:Nn, \seq_remove_all:cn}
+%   \begin{syntax}
+%     \cs{seq_remove_all:Nn} \meta{sequence} \Arg{item}
+%   \end{syntax}
+%   Removes every occurrence of \meta{item} from the \meta{sequence}.
+%   The \meta{item} comparison takes place on a token basis, as for
+%   \cs{tl_if_eq:nn(TF)}. The  removal is local to the current \TeX{} group.
+% \end{function}
+%
+% \begin{function}{\seq_gremove_all:Nn, \seq_gremove_all:cn}
+%   \begin{syntax}
+%     \cs{seq_gremove_all:Nn} \meta{sequence} \Arg{item}
+%   \end{syntax}
+%   Removes each occurrence of \meta{item} from the \meta{sequence}.
+%   The \meta{item} comparison takes place on a token basis, as for
+%   \cs{tl_if_eq:nn(TF)}. The removal is applied globally.
+% \end{function}
+%
+% \section{Sequence conditionals}
+%
+% \begin{function}[EXP,pTF]{\seq_if_empty:N, \seq_if_empty:c}
+%   \begin{syntax}
+%     \cs{seq_if_empty_p:N} \meta{sequence}
+%     \cs{seq_if_empty:NTF} \meta{sequence} \Arg{true code} \Arg{false code}
+%   \end{syntax}
+%   Tests if the \meta{sequence} is empty (containing no items). The
+%   branching versions then leave either \meta{true code} or
+%   \meta{false code} in the input stream, as appropriate to the truth of
+%   the test and the variant of the function chosen. The logical truth of
+%   the test is left in the input stream by the predicate version.
+% \end{function}
+%
+% \begin{function}[TF]{
+%    \seq_if_in:Nn, \seq_if_in:NV, \seq_if_in:Nv, \seq_if_in:No, \seq_if_in:Nx,
+%    \seq_if_in:cn, \seq_if_in:cV, \seq_if_in:cv, \seq_if_in:co, \seq_if_in:cx
+% }
+%   \begin{syntax}
+%     \cs{seq_if_in:NnTF} \meta{sequence} \Arg{item}
+%     ~~\Arg{true code} \Arg{false code}
+%   \end{syntax}
+%   Tests if the \meta{item} is present in the \meta{sequence}.
+%   Either the \meta{true code} or \meta{false code} is left in the input
+%   stream, as appropriate to the truth of the test and the variant of the
+%   function
+%   chosen.
+% \end{function}
+%
+% \section{Mapping to sequences}
+%
+% \begin{function}[EXP]{\seq_map_function:NN, \seq_map_function:cN}
+%   \begin{syntax}
+%     \cs{seq_map_function:NN} \meta{sequence} \meta{function}
+%   \end{syntax}
+%   Applies \meta{function} to every \meta{item} stored in the
+%   \meta{sequence}. The \meta{function} will receive one argument for
+%   each iteration. The \meta{items} are returned from left to right.
+%   The function \cs{seq_map_inline:Nn} is in general more efficient
+%   than \cs{seq_map_function:NN}.
+%   One mapping may be nested inside another.
+% \end{function}
+%
+% \begin{function}{\seq_map_inline:Nn, \seq_map_inline:cn}
+%   \begin{syntax}
+%     \cs{seq_map_inline:Nn} \meta{sequence} \Arg{inline function}
+%   \end{syntax}
+%   Applies \meta{inline function} to every \meta{item} stored
+%   within the \meta{sequence}. The \meta{inline function} should
+%   consist of code which will receive the \meta{item} as |#1|.
+%   One in line mapping can be nested inside another. The \meta{items}
+%   are returned from left to right.
+% \end{function}
+%
+% \begin{function}{
+%   \seq_map_variable:NNn, \seq_map_variable:Ncn,
+%   \seq_map_variable:cNn, \seq_map_variable:ccn
+% }
+%   \begin{syntax}
+%     \cs{seq_map_variable:NNn} \meta{sequence}
+%     ~~\meta{tl~var.} \Arg{function using tl~var.}
+%   \end{syntax}
+%   Stores each entry in the \meta{sequence} in turn in the
+%   \meta{tl~var.}\ and applies the \meta{function using tl~var.}
+%   The \meta{function} will usually consist of code making use of
+%   the \meta{tl~var.}, but this is not enforced.  One variable
+%   mapping can be nested inside another. The \meta{items}
+%   are returned from left to right.
+% \end{function}
+%
+% \begin{function}[EXP]{\seq_map_break:}
+%   \begin{syntax}
+%     \cs{seq_map_break:}
+%   \end{syntax}
+%   Used to terminate a \cs{seq_map_\ldots} function before all
+%   entries in the \meta{sequence} have been processed. This will
+%   normally take place within a conditional statement, for example
+%   \begin{verbatim}
+%     \seq_map_inline:Nn \l_my_seq
+%       {
+%         \str_if_eq:nnTF { #1 } { bingo }
+%           { \seq_map_break: }
+%           {
+%             % Do something useful
+%           }
+%       }
+%   \end{verbatim}
+%   Use outside of a \cs{seq_map_\ldots} scenario will lead to low
+%   level \TeX{} errors.
+%   \begin{texnote}
+%     When the mapping is broken, additional tokens may be inserted by the
+%     internal macro \cs{seq_break_point:n} before further items are taken
+%     from the input stream. This will depend on the design of the mapping
+%     function.
+%   \end{texnote}
+% \end{function}
+%
+% \begin{function}[EXP]{\seq_map_break:n}
+%   \begin{syntax}
+%     \cs{seq_map_break:n} \Arg{tokens}
+%   \end{syntax}
+%   Used to terminate a \cs{seq_map_\ldots} function before all
+%   entries in the \meta{sequence} have been processed, inserting
+%   the \meta{tokens} after the mapping has ended. This will
+%   normally take place within a conditional statement, for example
+%   \begin{verbatim}
+%     \seq_map_inline:Nn \l_my_seq
+%       {
+%         \str_if_eq:nnTF { #1 } { bingo }
+%           { \seq_map_break:n { <tokens> } }
+%           {
+%             % Do something useful
+%           }
+%       }
+%   \end{verbatim}
+%   Use outside of a \cs{seq_map_\ldots} scenario will lead to low
+%   level \TeX{} errors.
+%   \begin{texnote}
+%     When the mapping is broken, additional tokens may be inserted by the
+%     internal macro \cs{seq_break_point:n} before the \meta{tokens} are
+%     inserted into the input stream.
+%     This will depend on the design of the mapping function.
+%   \end{texnote}
+% \end{function}
+%
+% \section{Sequences as stacks}
+%
+% Sequences can be used as stacks, where data is pushed to and popped
+% from the top of the sequence. (The left of a sequence is the top, for
+% performance reasons.) The stack functions for sequences are not
+% intended to be mixed with the general ordered data functions detailed
+% in the previous section: a sequence should either be used as an
+% ordered data type or as a stack, but not in both ways.
+%
+% \begin{function}{\seq_get:NN, \seq_get:cN}
+%   \begin{syntax}
+%     \cs{seq_get:NN} \meta{sequence} \meta{token list variable}
+%   \end{syntax}
+%   Reads the top item from a \meta{sequence} into the
+%   \meta{token list variable} without removing it from the
+%   \meta{sequence}. The \meta{token list variable} is assigned locally.
+%   If \meta{sequence} is empty an error will be raised.
+% \end{function}
+%
+% \begin{function}{\seq_pop:NN, \seq_pop:cN}
+%   \begin{syntax}
+%     \cs{seq_pop:NN} \meta{sequence} \meta{token list variable}
+%   \end{syntax}
+%   Pops the top item from a \meta{sequence} into the
+%   \meta{token list variable}. Both of the variables are assigned
+%   locally. If \meta{sequence} is empty an error will be raised.
+% \end{function}
+%
+% \begin{function}{\seq_gpop:NN, \seq_gpop:cN}
+%   \begin{syntax}
+%     \cs{seq_gpop:NN} \meta{sequence} \meta{token list variable}
+%   \end{syntax}
+%   Pops the top item from a \meta{sequence} into the
+%   \meta{token list variable}. The \meta{sequence} is modified globally,
+%   while the \meta{token list variable} is assigned locally. If
+%   \meta{sequence} is empty an error will be raised.
+% \end{function}
+%
+% \begin{function}
+%   {
+%     \seq_push:Nn, \seq_push:NV, \seq_push:Nv, \seq_push:No, \seq_push:Nx,
+%      seq_push:cn, \seq_push:cV, \seq_push:cv, \seq_push:co, \seq_push:cx
+%   }
+%   \begin{syntax}
+%     \cs{seq_push:Nn} \meta{sequence} \Arg{item}
+%   \end{syntax}
+%   Adds the \Arg{item} to the top of the \meta{sequence}.
+%   The assignment is restricted to the current \TeX{}  group.
+% \end{function}
+%
+% \begin{function}
+%   {
+%     \seq_gpush:Nn, \seq_gpush:NV, \seq_gpush:Nv,
+%     \seq_gpush:No, \seq_gpush:Nx,
+%     \seq_gpush:cn, \seq_gpush:cV, \seq_gpush:cv,
+%     \seq_gpush:co, \seq_gpush:cx
+%   }
+%   \begin{syntax}
+%     \cs{seq_gpush:Nn} \meta{sequence} \Arg{item}
+%   \end{syntax}
+%   Pushes the \meta{item} onto the end of the top of the
+%   \meta{sequence}. The assignment is global.
+% \end{function}
+%
+% \section{Viewing sequences}
+%
+% \begin{function}{\seq_show:N, \seq_show:c}
+%   \begin{syntax}
+%     \cs{seq_show:N} \meta{sequence}
+%   \end{syntax}
+%   Displays the entries in the \meta{sequence} in the terminal.
+% \end{function}
+%
+% \section{Experimental sequence functions}
+%
+% This section contains functions which may or may not be retained, depending
+% on how useful they are found to be.
+%
+% \begin{function}[TF]{\seq_get_left:NN, \seq_get_left:cN}
+%   \begin{syntax}
+%     \cs{seq_get_left:NNTF} \meta{sequence} \meta{token list variable}
+%     ~~\Arg{true code} \Arg{false code}
+%   \end{syntax}
+%   If the \meta{sequence} is empty, leaves the \meta{false code} in the
+%   input stream and leaves the \meta{token list variable} unchanged. If the
+%   \meta{sequence} is non-empty, stores the left-most item from a \meta{sequence}
+%   in the \meta{token list variable} without removing it from a
+%   \meta{sequence}. The \meta{true code} is then left in the input stream.
+%   The \meta{token list variable} is assigned locally.
+% \end{function}
+%
+% \begin{function}[TF]{\seq_get_right:NN, \seq_get_right:cN}
+%   \begin{syntax}
+%     \cs{seq_get_right:NNTF} \meta{sequence} \meta{token list variable}
+%     ~~\Arg{true code} \Arg{false code}
+%   \end{syntax}
+%   If the \meta{sequence} is empty, leaves the \meta{false code} in the
+%   input stream and leaves the \meta{token list variable} unchanged. If the
+%   \meta{sequence} is non-empty, stores the right-most item from a \meta{sequence}
+%   in the \meta{token list variable} without removing it from a
+%   \meta{sequence}. The \meta{true code} is then left in the input stream.
+%   The \meta{token list variable} is assigned locally.
+% \end{function}
+%
+% \begin{function}[TF]{\seq_pop_left:NN, \seq_pop_left:cN}
+%   \begin{syntax}
+%     \cs{seq_pop_left:NNTF} \meta{sequence} \meta{token list variable}
+%     ~~\Arg{true code} \Arg{false code}
+%   \end{syntax}
+%   If the \meta{sequence} is empty, leaves the \meta{false code} in the
+%   input stream and leaves the \meta{token list variable} unchanged. If the
+%   \meta{sequence} is non-empty, pops the left-most item from a \meta{sequence}
+%   in the \meta{token list variable}, \emph{i.e.}~removes the item from a
+%   \meta{sequence}. The \meta{true code} is then left in the input stream.
+%   Both the \meta{sequence} and the \meta{token list variable} are assigned
+%   locally.
+% \end{function}
+%
+% \begin{function}[TF]{\seq_gpop_left:NN, \seq_gpop_left:cN}
+%   \begin{syntax}
+%     \cs{seq_gpop_left:NNTF} \meta{sequence} \meta{token list variable}
+%     ~~\Arg{true code} \Arg{false code}
+%   \end{syntax}
+%   If the \meta{sequence} is empty, leaves the \meta{false code} in the
+%   input stream and leaves the \meta{token list variable} unchanged. If the
+%   \meta{sequence} is non-empty, pops the left-most item from a \meta{sequence}
+%   in the \meta{token list variable}, \emph{i.e.}~removes the item from a
+%   \meta{sequence}. The \meta{true code} is then left in the input stream.
+%   The \meta{sequence} is modified globally, while the \meta{token list variable}
+%   is assigned locally.
+% \end{function}
+%
+% \begin{function}[TF]{\seq_pop_right:NN, \seq_pop_right:cN}
+%   \begin{syntax}
+%     \cs{seq_pop_right:NNTF} \meta{sequence} \meta{token list variable}
+%     ~~\Arg{true code} \Arg{false code}
+%   \end{syntax}
+%   If the \meta{sequence} is empty, leaves the \meta{false code} in the
+%   input stream and leaves the \meta{token list variable} unchanged. If the
+%   \meta{sequence} is non-empty, pops the right-most item from a \meta{sequence}
+%   in the \meta{token list variable}, \emph{i.e.}~removes the item from a
+%   \meta{sequence}. The \meta{true code} is then left in the input stream.
+%   Both the \meta{sequence} and the \meta{token list variable} are assigned
+%   locally.
+% \end{function}
+%
+% \begin{function}[TF]{\seq_gpop_right:NN, \seq_gpop_right:cN}
+%   \begin{syntax}
+%     \cs{seq_gpop_right:NNTF} \meta{sequence} \meta{token list variable}
+%     ~~\Arg{true code} \Arg{false code}
+%   \end{syntax}
+%   If the \meta{sequence} is empty, leaves the \meta{false code} in the
+%   input stream and leaves the \meta{token list variable} unchanged. If the
+%   \meta{sequence} is non-empty, pops the right-most item from a \meta{sequence}
+%   in the \meta{token list variable}, \emph{i.e.}~removes the item from a
+%   \meta{sequence}. The \meta{true code} is then left in the input stream.
+%   The \meta{sequence} is modified globally, while the \meta{token list variable}
+%   is assigned locally.
+% \end{function}
+%
+% \begin{function}[EXP]{\seq_length:N, \seq_length:c}
+%   \begin{syntax}
+%     \cs{seq_length:N} \meta{sequence}
+%   \end{syntax}
+%   Leaves the number of items in the \meta{sequence} in the input
+%   stream as an \meta{integer denotation}. The total number of items
+%   in a \meta{sequence} will include those which are empty and duplicates,
+%   \emph{i.e.}~every item in a \meta{sequence} is unique.
+% \end{function}
+%
+% \begin{function}[EXP]{\seq_item:Nn, \seq_item:cn}
+%   \begin{syntax}
+%     \cs{seq_item:Nn} \meta{sequence} \Arg{integer expression}
+%   \end{syntax}
+%   Indexing items in the \meta{sequence} from $0$ at the top (left), this
+%   function will evaluate the \meta{integer expression} and leave the
+%   appropriate item from the sequence in the input stream. If the
+%   \meta{integer expression} is negative, indexing occurs from the
+%   bottom (right) of the sequence. When the \meta{integer expression}
+%   is larger than the number of items in the \meta{sequence} (as
+%   calculated by \cs{seq_length:N}) then the function will expand to
+%   nothing.
+% \end{function}
+%
+% \begin{function}[EXP]{\seq_use:N, \seq_use:c}
+%   \begin{syntax}
+%     \cs{seq_use:N} \meta{sequence}
+%   \end{syntax}
+%   Places each \meta{item} in the \meta{sequence} in turn in the input stream.
+%   This occurs in an expandable fashion, and is implemented as a mapping.
+%   This means that the process may be prematurely terminated using
+%   \cs{seq_map_break:} or \cs{seq_map_break:n}. The \meta{items} in the
+%   \meta{sequence} will be used from left (top) to right (bottom).
+% \end{function}
+%
+% \begin{function}[EXP]
+%   {
+%     \seq_mapthread_function:NNN, \seq_mapthread_function:NcN,
+%     \seq_mapthread_function:cNN, \seq_mapthread_function:ccN
+%   }
+%   \begin{syntax}
+%     \cs{seq_mapthread_function:NNN} \meta{seq1} \meta{seq2} \meta{function}
+%   \end{syntax}
+%   Applies \meta{function} to every pair of items
+%   \meta{seq1-item}--\meta{seq2-item} from the two sequences, returning
+%   items from both sequences from left to right.   The \meta{function} will
+%   receive two \texttt{n}-type arguments for each iteration. The  mapping
+%   will terminate when
+%   the end of either sequence is reached (\emph{i.e.}~whichever sequence has
+%   fewer items determines how many iterations
+%   occur).
+% \end{function}
+%
+% \begin{function}
+%   {
+%     \seq_set_from_clist:NN, \seq_set_from_clist:cN,
+%     \seq_set_from_clist:Nc, \seq_set_from_clist:cc,
+%     \seq_set_from_clist:Nn, \seq_set_from_clist:cn
+%   }
+%   \begin{syntax}
+%     \cs{seq_set_from_clist:NN} \meta{sequence} \meta{comma-list}
+%   \end{syntax}
+%   Sets the \meta{sequence} within the current \TeX{} group to be equal
+%   to the content of the \meta{comma-list}.
+% \end{function}
+%
+% \begin{function}
+%   {
+%     \seq_gset_from_clist:NN, \seq_gset_from_clist:cN,
+%     \seq_gset_from_clist:Nc, \seq_gset_from_clist:cc,
+%     \seq_gset_from_clist:Nn, \seq_gset_from_clist:cn
+%   }
+%   \begin{syntax}
+%     \cs{seq_gset_from_clist:NN} \meta{sequence} \meta{comma-list}
+%   \end{syntax}
+%   Sets the \meta{sequence} globally to equal to the content of the
+%   \meta{comma-list}.
+% \end{function}
+%
+% \section{Internal sequence functions}
+%
+% \begin{function}{\seq_if_empty_err_break:N}
+%   \begin{syntax}
+%     \cs{seq_if_empty_err_break:N} \meta{sequence}
+%   \end{syntax}
+%   Tests if the \meta{sequence} is empty, and if so issues an error
+%   message before skipping over any tokens up to \cs{seq_break_point:n}.
+%   This function is used to avoid more serious errors which would
+%   otherwise occur if some internal functions were applied to an
+%   empty \meta{sequence}.
+% \end{function}
+%
+% \begin{function}[EXP]{\seq_item:n}
+%   \begin{syntax}
+%     \cs{seq_item:n} \meta{item}
+%   \end{syntax}
+%   The internal token used to begin each sequence entry. If expanded
+%   outside of a mapping or manipulation function, an error will be
+%   raised. The definition should always be set globally.
+% \end{function}
+%
+% \begin{function}{\seq_push_item_def:n, \seq_push_item_def:x}
+%   \begin{syntax}
+%     \cs{seq_push_item_def:n} \Arg{code}
+%   \end{syntax}
+%   Saves the definition of \cs{seq_item:n} and redefines it to
+%   accept one parameter and expand to \meta{code}. This function
+%   should always be balanced by use of \cs{seq_pop_item_def:}.
+% \end{function}
+%
+% \begin{function}{\seq_pop_item_def:}
+%   \begin{syntax}
+%     \cs{seq_pop_item_def:}
+%   \end{syntax}
+%   Restores the definition of \cs{seq_item:n} most recently saved by
+%   \cs{seq_push_item_def:n}. This function should always be used in
+%   a balanced pair with \cs{seq_push_item_def:n}.
+% \end{function}
+%
+% \begin{function}[EXP]{\seq_break:}
+%   \begin{syntax}
+%     \cs{seq_break:}
+%   \end{syntax}
+%   Used to terminate sequence functions by gobbling all tokens
+%   up to \cs{seq_break_point:n}. This function is a copy of
+%   \cs{seq_map_break:}, but is used in situations which are
+%   not mappings.
+% \end{function}
+%
+% \begin{function}[EXP]{\seq_break:n}
+%   \begin{syntax}
+%     \cs{seq_break:n} \Arg{tokens}
+%   \end{syntax}
+%   Used to terminate sequence functions by gobbling all tokens
+%   up to \cs{seq_break_point:n}, then inserting the \meta{tokens}
+%   before continuing reading the input stream. This function is a copy
+%   of \cs{seq_map_break:n}, but is used in situations which are
+%   not mappings.
+% \end{function}
+%
+% \begin{function}[EXP]{\seq_break_point:n}
+%   \begin{syntax}
+%     \cs{seq_break_point:n} \meta{tokens}
+%   \end{syntax}
+%   Used to mark the end of a recursion or mapping: the functions
+%   \cs{seq_map_break:} and \cs{seq_map_break:n} use this to break out
+%   of the loop. After the loop ends, the \meta{tokens} are inserted into
+%   the input stream. This occurs even if the the break functions are
+%   \emph{not} applied: \cs{seq_break_point:n} is functionally-equivalent
+%   in these cases to \cs{use:n}.
+% \end{function}
+%
+% \end{documentation}
+%
+% \begin{implementation}
+%
+% \section{\pkg{l3seq} implementation}
+%
+% \TestFiles{m3seq002,m3seq003}
+%
+%    \begin{macrocode}
+%<*initex|package>
+%    \end{macrocode}
+%
+%    \begin{macrocode}
+%<*package>
+\ProvidesExplPackage
+  {\ExplFileName}{\ExplFileDate}{\ExplFileVersion}{\ExplFileDescription}
+\package_check_loaded_expl:
+%</package>
+%    \end{macrocode}
+%
+% A sequence is a control sequence whose top-level expansion is of
+% the form \enquote{\cs{seq_item:n} \marg{item$_0$}
+% \ldots \cs{seq_item:n} \marg{item$_{n-1}$}}. An earlier implementation
+% used the structure \enquote{\cs{seq_elt:w} \meta{item$_1$}
+% \cs{seq_elt_end:} \ldots \cs{seq_elt:w} \meta{item$_n$}
+% \cs{seq_elt_end:}}. This allows rapid searching using a delimited
+% function, but is not suitable for items containing |{|, |}| and |#|
+% tokens, and also leads to the loss of surrounding braces
+% around items.
+%
+% \begin{macro}[int]{\seq_item:n}
+%   The delimiter is always defined, but when used incorrectly simply
+%   removes its argument and hits an undefined control sequence to
+%   raise an error.
+%    \begin{macrocode}
+\cs_new:Npn \seq_item:n
+  {
+    \seq_use_error:
+    \use_none:n
+  }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{variable}{\l_seq_tmpa_tl, \l_seq_tmpb_tl}
+%   Scratch space for various internal uses.
+%    \begin{macrocode}
+\tl_new:N \l_seq_tmpa_tl
+\tl_new:N \l_seq_tmpb_tl
+%    \end{macrocode}
+% \end{variable}
+%
+% \subsection{Allocation and initialisation}
+%
+% \begin{macro}{\seq_new:N,\seq_new:c}
+% \UnitTested
+%   Internally, sequences are just token lists.
+%    \begin{macrocode}
+\cs_new_eq:NN \seq_new:N \tl_new:N
+\cs_new_eq:NN \seq_new:c \tl_new:c
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\seq_clear:N, \seq_clear:c}
+% \UnitTested
+% \begin{macro}{\seq_gclear:N, \seq_gclear:c}
+% \UnitTested
+%   Clearing sequences is just the same as clearing token lists.
+%    \begin{macrocode}
+\cs_new_eq:NN \seq_clear:N  \tl_clear:N
+\cs_new_eq:NN \seq_clear:c  \tl_clear:c
+\cs_new_eq:NN \seq_gclear:N \tl_gclear:N
+\cs_new_eq:NN \seq_gclear:c \tl_gclear:c
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\seq_clear_new:N, \seq_clear_new:c}
+% \UnitTested
+% \begin{macro}{\seq_gclear_new:N, \seq_gclear_new:c}
+% \UnitTested
+%   Once again a copy from the token list functions.
+%    \begin{macrocode}
+\cs_new_eq:NN \seq_clear_new:N  \tl_clear_new:N
+\cs_new_eq:NN \seq_clear_new:c  \tl_clear_new:c
+\cs_new_eq:NN \seq_gclear_new:N \tl_gclear_new:N
+\cs_new_eq:NN \seq_gclear_new:c \tl_gclear_new:c
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\seq_set_eq:NN, \seq_set_eq:cN, \seq_set_eq:Nc, \seq_set_eq:cc}
+% \UnitTested
+% \begin{macro}
+%   {\seq_gset_eq:NN, \seq_gset_eq:cN, \seq_gset_eq:Nc, \seq_gset_eq:cc}
+% \UnitTested
+%   Once again, these are simple copies from the token list functions.
+%    \begin{macrocode}
+\cs_new_eq:NN \seq_set_eq:NN  \tl_set_eq:NN
+\cs_new_eq:NN \seq_set_eq:Nc  \tl_set_eq:Nc
+\cs_new_eq:NN \seq_set_eq:cN  \tl_set_eq:cN
+\cs_new_eq:NN \seq_set_eq:cc  \tl_set_eq:cc
+\cs_new_eq:NN \seq_gset_eq:NN \tl_gset_eq:NN
+\cs_new_eq:NN \seq_gset_eq:Nc \tl_gset_eq:Nc
+\cs_new_eq:NN \seq_gset_eq:cN \tl_gset_eq:cN
+\cs_new_eq:NN \seq_gset_eq:cc \tl_gset_eq:cc
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\seq_concat:NNN, \seq_concat:ccc}
+% \UnitTested
+% \begin{macro}{\seq_gconcat:NNN, \seq_gconcat:ccc}
+% \UnitTested
+%   Concatenating sequences is easy.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \seq_concat:NNN #1#2#3
+  { \tl_set:Nx #1 { \exp_not:o {#2} \exp_not:o {#3} } }
+\cs_new_protected_nopar:Npn \seq_gconcat:NNN #1#2#3
+  { \tl_gset:Nx #1 { \exp_not:o {#2} \exp_not:o {#3} } }
+\cs_generate_variant:Nn \seq_concat:NNN  { ccc }
+\cs_generate_variant:Nn \seq_gconcat:NNN { ccc }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \subsection{Appending data to either end}
+%
+% \begin{macro}{
+%   \seq_put_left:Nn, \seq_put_left:NV, \seq_put_left:Nv,
+%   \seq_put_left:No, \seq_put_left:Nx,
+%   \seq_put_left:cn, \seq_put_left:cV, \seq_put_left:cv,
+%   \seq_put_left:co, \seq_put_left:cx
+% }
+% \UnitTested
+% \begin{macro}{
+%   \seq_put_right:Nn, \seq_put_right:NV, \seq_put_right:Nv,
+%   \seq_put_right:No, \seq_put_right:Nx,
+%   \seq_put_right:cn, \seq_put_right:cV, \seq_put_right:cv,
+%   \seq_put_right:co, \seq_put_right:cx
+% }
+% \UnitTested
+%   The code here is just a wrapper for adding to token lists.
+%    \begin{macrocode}
+\cs_new_protected:Npn \seq_put_left:Nn #1#2
+  { \tl_put_left:Nn #1 { \seq_item:n {#2} } }
+\cs_new_protected:Npn \seq_put_right:Nn #1#2
+  { \tl_put_right:Nn #1 { \seq_item:n {#2} } }
+\cs_generate_variant:Nn \seq_put_left:Nn  {     NV , Nv , No , Nx }
+\cs_generate_variant:Nn \seq_put_left:Nn  { c , cV , cv , co , cx }
+\cs_generate_variant:Nn \seq_put_right:Nn {     NV , Nv , No , Nx }
+\cs_generate_variant:Nn \seq_put_right:Nn { c , cV , cv , co , cx }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{
+%   \seq_gput_left:Nn, \seq_gput_left:NV, \seq_gput_left:Nv,
+%   \seq_gput_left:No, \seq_gput_left:Nx,
+%   \seq_gput_left:cn, \seq_gput_left:cV, \seq_gput_left:cv,
+%   \seq_gput_left:co, \seq_gput_left:cx
+% }
+% \begin{macro}{
+%   \seq_gput_right:Nn, \seq_gput_right:NV, \seq_gput_right:Nv,
+%   \seq_gput_right:No, \seq_gput_right:Nx,
+%   \seq_gput_right:cn, \seq_gput_right:cV,\seq_gput_right:cv,
+%   \seq_gput_right:co, \seq_gput_right:cx
+% }
+%   The same for global addition.
+%    \begin{macrocode}
+\cs_new_protected:Npn \seq_gput_left:Nn #1#2
+  { \tl_gput_left:Nn #1 { \seq_item:n {#2} } }
+\cs_new_protected:Npn \seq_gput_right:Nn #1#2
+  { \tl_gput_right:Nn #1 { \seq_item:n {#2} } }
+\cs_generate_variant:Nn \seq_gput_left:Nn  {     NV , Nv , No , Nx }
+\cs_generate_variant:Nn \seq_gput_left:Nn  { c , cV , cv , co , cx }
+\cs_generate_variant:Nn \seq_gput_right:Nn {     NV , Nv , No , Nx }
+\cs_generate_variant:Nn \seq_gput_right:Nn { c , cV , cv , co , cx }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \subsection{Modifying sequences}
+%
+% \begin{variable}{\l_seq_remove_seq}
+%   An internal sequence for the removal routines.
+%    \begin{macrocode}
+\seq_new:N \l_seq_remove_seq
+%    \end{macrocode}
+% \end{variable}
+%
+% \begin{macro}{\seq_remove_duplicates:N, \seq_remove_duplicates:c}
+% \UnitTested
+% \begin{macro}{\seq_gremove_duplicates:N, \seq_gremove_duplicates:c}
+% \UnitTested
+% \begin{macro}[aux]{\seq_remove_duplicates_aux:NN}
+%   Removing duplicates means making a new list then copying it.
+%    \begin{macrocode}
+\cs_new_protected:Npn \seq_remove_duplicates:N
+  { \seq_remove_duplicates_aux:NN \seq_set_eq:NN }
+\cs_new_protected:Npn \seq_gremove_duplicates:N
+  { \seq_remove_duplicates_aux:NN \seq_gset_eq:NN }
+\cs_new_protected:Npn \seq_remove_duplicates_aux:NN #1#2
+  {
+    \seq_clear:N \l_seq_remove_seq
+    \seq_map_inline:Nn #2
+      {
+        \seq_if_in:NnF \l_seq_remove_seq {##1}
+          { \seq_put_right:Nn \l_seq_remove_seq {##1} }
+      }
+    #1 #2 \l_seq_remove_seq
+  }
+\cs_generate_variant:Nn \seq_remove_duplicates:N  { c }
+\cs_generate_variant:Nn \seq_gremove_duplicates:N { c }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\seq_remove_all:Nn, \seq_remove_all:cn}
+% \UnitTested
+% \begin{macro}{\seq_gremove_all:Nn, \seq_gremove_all:cn}
+% \UnitTested
+% \begin{macro}[aux]{\seq_remove_all_aux:NNn}
+%   The idea of the code here is to avoid a relatively expensive addition of
+%   items one at a time to an intermediate sequence.
+%   The approach taken is therefore similar to
+%   that in \cs{seq_pop_right_aux_ii:NNN}, using a \enquote{flexible}
+%   \texttt{x}-type expansion to do most of the work. As \cs{tl_if_eq:nnT}
+%   is not expandable, a two-part strategy is needed. First, the
+%   \texttt{x}-type expansion uses \cs{str_if_eq:nnT} to find potential
+%   matches. If one is found, the expansion is halted and the necessary
+%   set up takes place to use the \cs{tl_if_eq:NNT} test. The \texttt{x}-type
+%   is started again, including all of the items copied already. This will
+%   happen repeatedly until the entire sequence has been scanned. The code
+%   is set up to avoid needing and intermediate scratch list: the lead-off
+%   \texttt{x}-type expansion (|#1 #2 {#2}|) will ensure that nothing is
+%   lost.
+%    \begin{macrocode}
+\cs_new_protected:Npn \seq_remove_all:Nn
+  { \seq_remove_all_aux:NNn \tl_set:Nx }
+\cs_new_protected:Npn \seq_gremove_all:Nn
+  { \seq_remove_all_aux:NNn \tl_gset:Nx }
+\cs_new_protected:Npn \seq_remove_all_aux:NNn #1#2#3
+  {
+    \seq_push_item_def:n
+      {
+        \str_if_eq:nnT {##1} {#3}
+          {
+            \if_false: { \fi: }
+            \tl_set:Nn \l_seq_tmpb_tl {##1}
+            #1 #2
+               { \if_false: } \fi:
+                 \exp_not:o {#2}
+                 \tl_if_eq:NNT \l_seq_tmpa_tl \l_seq_tmpb_tl
+                   { \use_none:nn }
+          }
+        \exp_not:n { \seq_item:n {##1} }
+      }
+    \tl_set:Nn \l_seq_tmpa_tl {#3}
+    #1 #2 {#2}
+    \seq_pop_item_def:
+  }
+\cs_generate_variant:Nn \seq_remove_all:Nn  { c }
+\cs_generate_variant:Nn \seq_gremove_all:Nn { c }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \subsection{Sequence conditionals}
+%
+% \begin{macro}[pTF]{\seq_if_empty:N, \seq_if_empty:c}
+% \UnitTested
+%   Simple copies from the token list variable material.
+%    \begin{macrocode}
+\prg_new_eq_conditional:NNn \seq_if_empty:N \tl_if_empty:N
+  { p , T , F , TF }
+\prg_new_eq_conditional:NNn \seq_if_empty:c \tl_if_empty:c
+  { p , T , F , TF }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}[TF]{
+%   \seq_if_in:Nn, \seq_if_in:NV, \seq_if_in:Nv, \seq_if_in:No, \seq_if_in:Nx,
+%   \seq_if_in:cn, \seq_if_in:cV, \seq_if_in:cv, \seq_if_in:co, \seq_if_in:cx
+% }
+% \UnitTested
+% \begin{macro}[aux]{\seq_if_in_aux:}
+%   The approach here is to define \cs{seq_item:n} to compare its
+%   argument with the test sequence. If the two items are equal, the
+%   mapping is terminated and \cs{prg_return_true:} is inserted. On the
+%   other hand, if there is no match then the loop will break returning
+%   \cs{prg_return_false:}. In either case, \cs{seq_break_point:n}
+%   ensures that the group ends before the logical value is returned.
+%   Everything is inside a group so that \cs{seq_item:n} is preserved
+%   in nested situations.
+%    \begin{macrocode}
+\prg_new_protected_conditional:Npnn \seq_if_in:Nn #1#2
+  { T , F , TF }
+  {
+    \group_begin:
+      \tl_set:Nn \l_seq_tmpa_tl {#2}
+      \cs_set_protected:Npn \seq_item:n ##1
+        {
+          \tl_set:Nn \l_seq_tmpb_tl {##1}
+          \if_meaning:w \l_seq_tmpa_tl \l_seq_tmpb_tl
+            \exp_after:wN \seq_if_in_aux:
+          \fi:
+        }
+      #1
+      \seq_break:n { \prg_return_false: }
+    \seq_break_point:n { \group_end: }
+  }
+\cs_new_nopar:Npn \seq_if_in_aux: { \seq_break:n { \prg_return_true: } }
+\cs_generate_variant:Nn \seq_if_in:NnT  {     NV , Nv , No , Nx }
+\cs_generate_variant:Nn \seq_if_in:NnT  { c , cV , cv , co , cx }
+\cs_generate_variant:Nn \seq_if_in:NnF  {     NV , Nv , No , Nx }
+\cs_generate_variant:Nn \seq_if_in:NnF  { c , cV , cv , co , cx }
+\cs_generate_variant:Nn \seq_if_in:NnTF {     NV , Nv , No , Nx }
+\cs_generate_variant:Nn \seq_if_in:NnTF { c , cV , cv , co , cx }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \subsection{Recovering data from sequences}
+%
+% \begin{macro}{\seq_get_left:NN, \seq_get_left:cN}
+% \UnitTested
+% \begin{macro}[aux]{\seq_get_left_aux:NnwN}
+%   Getting an item from the left of a sequence is pretty easy: just
+%   trim off the first item after removing the \cs{seq_item:n} at
+%   the start.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \seq_get_left:NN #1#2
+  {
+    \seq_if_empty_err_break:N #1
+    \exp_after:wN \seq_get_left_aux:NnwN #1 \q_stop #2
+    \seq_break_point:n { }
+  }
+\cs_new_protected:Npn \seq_get_left_aux:NnwN \seq_item:n #1#2 \q_stop #3
+  { \tl_set:Nn #3 {#1} }
+\cs_generate_variant:Nn \seq_get_left:NN { c }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\seq_pop_left:NN, \seq_pop_left:cN}
+% \UnitTested
+% \begin{macro}{\seq_gpop_left:NN, \seq_gpop_left:cN}
+% \UnitTested
+% \begin{macro}[aux]{\seq_pop_left_aux:NNN}
+% \begin{macro}[aux]{\seq_pop_left_aux:NnwNNN}
+%   The approach to popping an item is pretty similar to that to get
+%   an item, with the only difference being that the sequence itself has
+%   to be redefined. This makes it more sensible to use an auxiliary
+%   function for the local and global cases.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \seq_pop_left:NN
+  { \seq_pop_left_aux:NNN \tl_set:Nn }
+\cs_new_protected_nopar:Npn \seq_gpop_left:NN
+  { \seq_pop_left_aux:NNN \tl_gset:Nn }
+\cs_new_protected_nopar:Npn \seq_pop_left_aux:NNN #1#2#3
+  {
+    \seq_if_empty_err_break:N #2
+    \exp_after:wN \seq_pop_left_aux:NnwNNN #2 \q_stop #1#2#3
+    \seq_break_point:n { }
+  }
+\cs_new_protected:Npn \seq_pop_left_aux:NnwNNN \seq_item:n #1#2 \q_stop #3#4#5
+  {
+    #3 #4 {#2}
+    \tl_set:Nn #5 {#1}
+  }
+\cs_generate_variant:Nn \seq_pop_left:NN  { c }
+\cs_generate_variant:Nn \seq_gpop_left:NN { c }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\seq_get_right:NN, \seq_get_right:cN}
+% \UnitTested
+% \begin{macro}[aux]{\seq_get_right_aux:NN}
+% \begin{macro}[aux]{\seq_get_right_loop:nn}
+%   The idea here is to remove the very first \cs{seq_item:n} from the
+%   sequence, leaving a token list starting with the first braced entry.
+%   Two arguments at a time are then grabbed: apart from the right-hand end of
+%   the sequence, this will be a brace group followed by \cs{seq_item:n}. The
+%   set up code means that these all disappear. At the end of the sequence,
+%   the assignment is placed in front of the very last entry in the sequence,
+%   before a tidying-up step takes place to remove the loop and reset the
+%   meaning of \cs{seq_item:n}.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \seq_get_right:NN #1#2
+  {
+    \seq_if_empty_err_break:N #1
+    \seq_get_right_aux:NN #1#2
+    \seq_break_point:n { }
+  }
+\cs_new_protected_nopar:Npn \seq_get_right_aux:NN #1#2
+  {
+    \seq_push_item_def:n { }
+    \exp_after:wN \exp_after:wN \exp_after:wN \seq_get_right_loop:nn
+      \exp_after:wN \use_none:n #1
+      { \tl_set:Nn #2 }
+      { }
+      {
+        \seq_pop_item_def:
+        \seq_break:
+      }
+}
+\cs_new:Npn \seq_get_right_loop:nn #1#2
+  {
+    #2 {#1}
+    \seq_get_right_loop:nn
+  }
+\cs_generate_variant:Nn \seq_get_right:NN { c }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\seq_pop_right:NN, \seq_pop_right:cN}
+% \UnitTested
+% \begin{macro}{\seq_gpop_right:NN, \seq_gpop_right:cN}
+% \UnitTested
+% \begin{macro}[aux]{\seq_pop_right_aux:NNN, \seq_pop_right_aux_ii:NNN}
+%   The approach to popping from the right is a bit more involved, but does
+%   use some of the same ideas as getting from the right. What is needed is a
+%   \enquote{flexible length} way to set a token list variable. This is
+%   supplied by the |{ \if_false:} \fi:| \ldots
+%   |\if_false: { \fi: }| construct. Using an \texttt{x}-type
+%   expansion and a \enquote{non-expanding} definition for \cs{seq_item:n},
+%   the left-most $n - 1$ entries in a sequence of $n$ items will be stored
+%   back in the sequence. That needs a loop of unknown length, hence using the
+%   strange \cs{if_false:} way of including brackets. When the last item
+%   of the sequence is reached, the closing bracket for the assignment is
+%   inserted, and |\tl_set:Nn #3| is inserted in front of the final entry.
+%   This therefore does the pop assignment, then a final loop clears up the
+%   code.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \seq_pop_right:NN
+  { \seq_pop_right_aux:NNN \tl_set:Nx }
+\cs_new_protected_nopar:Npn \seq_gpop_right:NN
+  { \seq_pop_right_aux:NNN \tl_gset:Nx }
+\cs_new_protected_nopar:Npn \seq_pop_right_aux:NNN #1#2#3
+  {
+    \seq_if_empty_err_break:N #2
+    \seq_pop_right_aux_ii:NNN #1 #2 #3
+    \seq_break_point:n { }
+  }
+\cs_new_protected_nopar:Npn \seq_pop_right_aux_ii:NNN #1#2#3
+  {
+    \seq_push_item_def:n { \exp_not:n { \seq_item:n {##1} } }
+    #1 #2 { \if_false: } \fi:
+      \exp_after:wN \exp_after:wN \exp_after:wN \seq_get_right_loop:nn
+         \exp_after:wN \use_none:n #2
+        {
+          \if_false: { \fi: }
+          \tl_set:Nn #3
+        }
+        { }
+        {
+          \seq_pop_item_def:
+          \seq_break:
+        }
+  }
+\cs_generate_variant:Nn \seq_pop_right:NN  { c }
+\cs_generate_variant:Nn \seq_gpop_right:NN { c }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \subsection{Mapping to sequences}
+%
+% \begin{macro}[int]{\seq_break:}
+% \begin{macro}[int]{\seq_break:n}
+%   To break a function, the special token \cs{seq_break_point:n} is
+%   used to find the end of the code. Any ending code is then inserted
+%   before the return value of \cs{seq_map_break:n} is inserted.
+%    \begin{macrocode}
+\cs_new:Npn \seq_break:  #1   \seq_break_point:n #2 {#2}
+\cs_new:Npn \seq_break:n #1#2 \seq_break_point:n #3 { #3 #1 }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\seq_map_break:}
+% \UnitTested
+% \begin{macro}{\seq_map_break:n}
+% \UnitTested
+%   Semantically-logical copies of the break functions for use inside
+%   mappings.
+%    \begin{macrocode}
+\cs_new_eq:NN \seq_map_break:  \seq_break:
+\cs_new_eq:NN \seq_map_break:n \seq_break:n
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}[int]{\seq_break_point:n}
+%   Normally, the marker token will not be executed, but if it is then
+%   the end code is simply inserted.
+%    \begin{macrocode}
+\cs_new_eq:NN \seq_break_point:n \use:n
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}[int]{\seq_if_empty_err_break:N}
+%   A function to check that sequences really have some content. This
+%   is optimised for speed, hence the direct primitive use.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \seq_if_empty_err_break:N #1
+  {
+    \if_meaning:w #1 \c_empty_tl
+      \msg_kernel_error:nnx { seq } { empty-sequence } { \token_to_str:N #1 }
+      \exp_after:wN \seq_break:
+    \fi:
+  }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\seq_map_function:NN, \seq_map_function:cN}
+% \UnitTested
+% \begin{macro}[aux]{\seq_map_function_aux:NNn}
+%   The idea here is to apply the code of |#2| to each item in the
+%   sequence without altering the definition of \cs{seq_item:n}. This
+%   is done as by noting that every odd token in the sequence must be
+%   \cs{seq_item:n}, which can be gobbled by \cs{use_none:n}. At the end of
+%   the loop, |#2| is instead |? \seq_map_break:|, which therefore breaks the
+%   loop without needing to do a (relatively-expensive) quark test.
+%    \begin{macrocode}
+\cs_new:Npn \seq_map_function:NN #1#2
+  {
+    \exp_after:wN \seq_map_function_aux:NNn \exp_after:wN #2 #1
+      { ? \seq_map_break: } { }
+    \seq_break_point:n { }
+  }
+\cs_new:Npn \seq_map_function_aux:NNn #1#2#3
+  {
+    \use_none:n #2
+    #1 {#3}
+    \seq_map_function_aux:NNn #1
+  }
+\cs_generate_variant:Nn \seq_map_function:NN { c }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{variable}{\g_seq_nesting_depth_int}
+% A counter to keep track of nested functions: defined in \pkg{l3int}.
+% \end{variable}
+%
+% \begin{macro}[int]{\seq_push_item_def:n, \seq_push_item_def:x}
+% \begin{macro}[aux]{\seq_push_item_def_aux:}
+% \begin{macro}[int]{\seq_pop_item_def:}
+%   The definition of \cs{seq_item:n} needs to be saved and restored at
+%   various points within the mapping and manipulation code. That is handled
+%   here: as always, this approach uses global assignments.
+%    \begin{macrocode}
+\cs_new_protected:Npn \seq_push_item_def:n
+  {
+    \seq_push_item_def_aux:
+    \cs_gset:Npn \seq_item:n ##1
+  }
+\cs_new_protected:Npn \seq_push_item_def:x
+  {
+    \seq_push_item_def_aux:
+    \cs_gset:Npx \seq_item:n ##1
+  }
+\cs_new_protected:Npn \seq_push_item_def_aux:
+  {
+    \cs_gset_eq:cN { seq_item_ \int_use:N \g_seq_nesting_depth_int :n }
+      \seq_item:n
+    \int_gincr:N \g_seq_nesting_depth_int
+  }
+\cs_new_protected_nopar:Npn \seq_pop_item_def:
+  {
+    \int_gdecr:N \g_seq_nesting_depth_int
+    \cs_gset_eq:Nc \seq_item:n
+      { seq_item_ \int_use:N \g_seq_nesting_depth_int :n }
+  }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\seq_map_inline:Nn, \seq_map_inline:cn}
+% \UnitTested
+%   The idea here is that \cs{seq_item:n} is already \enquote{applied} to
+%   each item in a sequence, and so an in-line mapping is just a case of
+%   redefining \cs{seq_item:n}.
+%    \begin{macrocode}
+\cs_new_protected:Npn \seq_map_inline:Nn #1#2
+  {
+    \seq_push_item_def:n {#2}
+    #1
+    \seq_break_point:n { \seq_pop_item_def: }
+  }
+\cs_generate_variant:Nn \seq_map_inline:Nn { c }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}
+%   {
+%     \seq_map_variable:NNn,\seq_map_variable:Ncn,
+%     \seq_map_variable:cNn,\seq_map_variable:ccn
+%   }
+% \UnitTested
+%   This is just a specialised version of the in-line mapping function,
+%   using an \texttt{x}-type expansion for the code set up so that the
+%   number of |#| tokens required is as expected.
+%    \begin{macrocode}
+\cs_new_protected:Npn \seq_map_variable:NNn #1#2#3
+  {
+    \seq_push_item_def:x
+      {
+        \tl_set:Nn \exp_not:N #2 {##1}
+        \exp_not:n {#3}
+      }
+    #1
+    \seq_break_point:n { \seq_pop_item_def: }
+  }
+\cs_generate_variant:Nn \seq_map_variable:NNn {     Nc }
+\cs_generate_variant:Nn \seq_map_variable:NNn { c , cc }
+%    \end{macrocode}
+% \end{macro}
+%
+% \subsection{Sequence stacks}
+%
+% The same functions as for sequences, but with the correct naming.
+%
+% \begin{macro}{
+%   \seq_push:Nn, \seq_push:NV, \seq_push:Nv, \seq_push:No, \seq_push:Nx,
+%   \seq_push:cn, \seq_push:cV, \seq_push:cV, \seq_push:co, \seq_push:cx
+% }
+% \UnitTested
+% \begin{macro}{
+%   \seq_gpush:Nn, \seq_gpush:NV, \seq_gpush:Nv, \seq_gpush:No, \seq_gpush:Nx,
+%   \seq_gpush:cn, \seq_gpush:cV, \seq_gpush:cv, \seq_gpush:co, \seq_gpush:cx
+% }
+% \UnitTested
+%   Pushing to a sequence is the same as adding on the left.
+%    \begin{macrocode}
+\cs_new_eq:NN \seq_push:Nn  \seq_put_left:Nn
+\cs_new_eq:NN \seq_push:NV  \seq_put_left:NV
+\cs_new_eq:NN \seq_push:Nv  \seq_put_left:Nv
+\cs_new_eq:NN \seq_push:No  \seq_put_left:No
+\cs_new_eq:NN \seq_push:Nx  \seq_put_left:Nx
+\cs_new_eq:NN \seq_push:cn  \seq_put_left:cn
+\cs_new_eq:NN \seq_push:cV  \seq_put_left:cV
+\cs_new_eq:NN \seq_push:cv  \seq_put_left:cv
+\cs_new_eq:NN \seq_push:co  \seq_put_left:co
+\cs_new_eq:NN \seq_push:cx  \seq_put_left:cx
+\cs_new_eq:NN \seq_gpush:Nn \seq_gput_left:Nn
+\cs_new_eq:NN \seq_gpush:NV \seq_gput_left:NV
+\cs_new_eq:NN \seq_gpush:Nv \seq_gput_left:Nv
+\cs_new_eq:NN \seq_gpush:No \seq_gput_left:No
+\cs_new_eq:NN \seq_gpush:Nx \seq_gput_left:Nx
+\cs_new_eq:NN \seq_gpush:cn \seq_gput_left:cn
+\cs_new_eq:NN \seq_gpush:cV \seq_gput_left:cV
+\cs_new_eq:NN \seq_gpush:cv \seq_gput_left:cv
+\cs_new_eq:NN \seq_gpush:co \seq_gput_left:co
+\cs_new_eq:NN \seq_gpush:cx \seq_gput_left:cx
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\seq_get:NN, \seq_get:cN}
+% \UnitTested
+% \begin{macro}{\seq_pop:NN, \seq_pop:cN}
+% \UnitTested
+% \begin{macro}{\seq_gpop:NN, \seq_gpop:cN}
+% \UnitTested
+%   In most cases, getting items from the stack does not need to specify
+%   that this is from the left. So alias are provided.
+%    \begin{macrocode}
+\cs_new_eq:NN \seq_get:NN \seq_get_left:NN
+\cs_new_eq:NN \seq_get:cN \seq_get_left:cN
+\cs_new_eq:NN \seq_pop:NN \seq_pop_left:NN
+\cs_new_eq:NN \seq_pop:cN \seq_pop_left:cN
+\cs_new_eq:NN \seq_gpop:NN \seq_gpop_left:NN
+\cs_new_eq:NN \seq_gpop:cN \seq_gpop_left:cN
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \subsection{Viewing sequences}
+%
+% \begin{variable}{\l_seq_show_tl}
+%   Used to store the material for display.
+%    \begin{macrocode}
+\tl_new:N \l_seq_show_tl
+%    \end{macrocode}
+% \end{variable}
+%
+% \begin{macro}{\seq_show:N, \seq_show:c}
+% \UnitTested
+% \begin{macro}[aux]{\seq_show_aux:n}
+% \begin{macro}[aux]{\seq_show_aux:w}
+%   The aim of the mapping here is to create a token list containing the
+%   formatted sequence. The very first item needs the new line and \verb*|> |
+%   removing, which is achieved using a \texttt{w}-type auxiliary. To avoid
+%   a low-level \TeX{} error if there is an empty sequence, a simple test is
+%   used to keep the output \enquote{clean}.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \seq_show:N #1
+  {
+    \seq_if_empty:NTF #1
+      {
+        \iow_term:x { Sequence~\token_to_str:N #1 \c_space_tl is~empty }
+        \tl_show:n { }
+      }
+      {
+        \iow_term:x
+          {
+            Sequence~\token_to_str:N #1 \c_space_tl
+            contains~the~items~(without~outer~braces):
+          }
+        \tl_set:Nx \l_seq_show_tl
+          { \seq_map_function:NN #1 \seq_show_aux:n }
+          \etex_showtokens:D \exp_after:wN \exp_after:wN \exp_after:wN
+            { \exp_after:wN \seq_show_aux:w \l_seq_show_tl }
+      }
+  }
+\cs_new:Npn \seq_show_aux:n #1
+  {
+    \iow_newline: > \c_space_tl \c_space_tl
+    \iow_char:N \{ \exp_not:n {#1} \iow_char:N \}
+  }
+\cs_new:Npn \seq_show_aux:w #1 > ~ { }
+\cs_generate_variant:Nn \seq_show:N { c }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \subsection{Experimental functions}
+%
+% \begin{macro}[aux]{\seq_if_empty_break_return_false:N}
+% The name says it all: of the sequence is empty, returns logical
+% \texttt{false}.
+%    \begin{macrocode}
+\cs_new_nopar:Npn \seq_if_empty_break_return_false:N #1
+  {
+    \if_meaning:w #1 \c_empty_tl
+      \prg_return_false:
+      \exp_after:wN \seq_break:
+    \fi:
+  }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}[TF]{\seq_get_left:NN, \seq_get_left:cN}
+% \begin{macro}[TF]{\seq_get_right:NN, \seq_get_right:cN}
+% Getting from the left or right with a check on the results.
+%    \begin{macrocode}
+\prg_new_protected_conditional:Npnn \seq_get_left:NN #1 #2 { T , F , TF }
+  {
+    \seq_if_empty_break_return_false:N #1
+    \exp_after:wN \seq_get_left_aux:Nw #1 \q_stop #2
+    \prg_return_true:
+    \seq_break:
+    \seq_break_point:n { }
+  }
+\prg_new_protected_conditional:Npnn \seq_get_right:NN #1#2 { T , F , TF }
+  {
+    \seq_if_empty_break_return_false:N #1
+    \seq_get_right_aux:NN #1#2
+    \prg_return_true: \seq_break:
+    \seq_break_point:n { }
+  }
+\cs_generate_variant:Nn \seq_get_left:NNT   { c }
+\cs_generate_variant:Nn \seq_get_left:NNF   { c }
+\cs_generate_variant:Nn \seq_get_left:NNTF  { c }
+\cs_generate_variant:Nn \seq_get_right:NNT  { c }
+\cs_generate_variant:Nn \seq_get_right:NNF  { c }
+\cs_generate_variant:Nn \seq_get_right:NNTF { c }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}[TF]{\seq_pop_left:NN, \seq_pop_left:cN}
+% \begin{macro}[TF]{\seq_gpop_left:NN, \seq_gpop_left:cN}
+% \begin{macro}[TF]{\seq_pop_right:NN, \seq_pop_right:cN}
+% \begin{macro}[TF]{\seq_gpop_right:NN, \seq_gpop_right:cN}
+% More or less the same for popping.
+%    \begin{macrocode}
+\prg_new_protected_conditional:Npnn \seq_pop_left:NN #1#2 { T , F , TF }
+  {
+    \seq_if_empty_break_return_false:N #1
+    \exp_after:wN \seq_pop_left_aux:NnwNNN #1 \q_stop \tl_set:Nn #1#2
+    \prg_return_true: \seq_break:
+    \seq_break_point:n { }
+  }
+\prg_new_protected_conditional:Npnn \seq_gpop_left:NN #1#2 { T , F , TF }
+  {
+    \seq_if_empty_break_return_false:N #1
+    \exp_after:wN \seq_pop_left_aux:NnwNNN #1 \q_stop \tl_gset:Nn #1#2
+    \prg_return_true: \seq_break:
+    \seq_break_point:n { }
+  }
+\prg_new_protected_conditional:Npnn \seq_pop_right:NN #1#2 { T , F , TF }
+  {
+    \seq_if_empty_break_return_false:N #1
+    \seq_pop_right_aux_ii:NNN \tl_set:Nx #1 #2
+    \prg_return_true: \seq_break:
+    \seq_break_point:n { }
+  }
+\prg_new_protected_conditional:Npnn \seq_gpop_right:NN #1#2 { T , F , TF }
+  {
+    \seq_if_empty_break_return_false:N #1
+    \seq_pop_right_aux_ii:NNN \tl_gset:Nx #1 #2
+    \prg_return_true: \seq_break:
+    \seq_break_point:n { }
+  }
+\cs_generate_variant:Nn \seq_pop_left:NNT    { c }
+\cs_generate_variant:Nn \seq_pop_left:NNF    { c }
+\cs_generate_variant:Nn \seq_pop_left:NNTF   { c }
+\cs_generate_variant:Nn \seq_gpop_left:NNT   { c }
+\cs_generate_variant:Nn \seq_gpop_left:NNF   { c }
+\cs_generate_variant:Nn \seq_gpop_left:NNTF  { c }
+\cs_generate_variant:Nn \seq_pop_right:NNT   { c }
+\cs_generate_variant:Nn \seq_pop_right:NNF   { c }
+\cs_generate_variant:Nn \seq_pop_right:NNTF  { c }
+\cs_generate_variant:Nn \seq_gpop_right:NNT  { c }
+\cs_generate_variant:Nn \seq_gpop_right:NNF  { c }
+\cs_generate_variant:Nn \seq_gpop_right:NNTF { c }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\seq_length:N, \seq_length:c}
+% \begin{macro}[aux]{\seq_length_aux:n}
+%   Counting the items in a sequence is done using the same approach as for
+%   other length functions: turn each entry into a \texttt{+1} then use
+%   integer evaluation to actually do the mathematics.
+%    \begin{macrocode}
+\cs_new:Npn \seq_length:N #1
+  {
+    \int_eval:n
+      {
+        0
+        \seq_map_function:NN #1 \seq_length_aux:n
+      }
+  }
+\cs_new:Npn \seq_length_aux:n #1 { +1 }
+\cs_generate_variant:Nn \seq_length:N { c }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\seq_item:Nn, \seq_item:cn}
+% \begin{macro}[aux]{\seq_item_aux:nnn}
+%   The idea here is to find the offset of the item from the left, then use
+%   a loop to grab the correct item. If the resulting offset is too large,
+%   then the stop code |{ ? \seq_break } { }| will be used by the auxiliary,
+%   terminating the loop and returning nothing at all.
+%    \begin{macrocode}
+\cs_new_nopar:Npn \seq_item:Nn #1#2
+  {
+    \exp_last_unbraced:Nfo \seq_item_aux:nnn
+      {
+        \int_eval:n
+          {
+            \int_compare:nNnT {#2} < \c_zero
+              { \seq_length:N #1 + }
+            #2
+          }
+      }
+    #1
+    { ? \seq_break: }
+    { }
+    \seq_break_point:n { }
+  }
+\cs_new_nopar:Npn \seq_item_aux:nnn #1#2#3
+  {
+    \use_none:n #2
+    \int_compare:nNnTF {#1} = \c_zero
+      { \seq_break:n {#3} }
+      { \exp_args:Nf \seq_item_aux:nnn { #1 - 1 } }
+  }
+\cs_generate_variant:Nn \seq_item:Nn { c }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\seq_use:N, \seq_use:c}
+%   A simple short cut for a mapping.
+%    \begin{macrocode}
+\cs_new_nopar:Npn \seq_use:N #1 { \seq_map_function:NN #1 \use:n }
+\cs_generate_variant:Nn \seq_use:N { c }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}
+%   {
+%     \seq_mapthread_function:NNN, \seq_mapthread_function:NcN,
+%     \seq_mapthread_function:cNN, \seq_mapthread_function:ccN
+%   }
+% \begin{macro}[aux]{\seq_mapthread_function_aux:NN}
+% \begin{macro}[aux]{\seq_mapthread_function_aux:Nnnwnn}
+%   The idea here is to first expand both of the sequences, adding the usual
+%   |{ ? \seq_break: } { }| to the end of each on. This is most conveniently
+%   done in two steps using an auxiliary function. The mapping then throws
+%   away the first token of |#2| and |#5|, which for items in the sequences
+%   will both be \cs{seq_item:n}. The function to be mapped will then be
+%   applied to the two entries. When the code hits the end of one of the
+%   sequences, the break material will stop the entire loop and tidy up. This
+%   avoids needing to find the length of the two sequences, or worrying about
+%   which is longer.
+%    \begin{macrocode}
+\cs_new_nopar:Npn \seq_mapthread_function:NNN #1#2#3
+  {
+    \exp_after:wN \seq_mapthread_function_aux:NN
+      \exp_after:wN #3
+      \exp_after:wN #1
+      #2
+      { ? \seq_break: } { }
+    \seq_break_point:n { }
+  }
+\cs_new_nopar:Npn \seq_mapthread_function_aux:NN #1#2
+  {
+    \exp_after:wN \seq_mapthread_function_aux:Nnnwnn
+      \exp_after:wN #1
+      #2
+      { ? \seq_break: } { }
+      \q_stop
+  }
+\cs_new:Npn \seq_mapthread_function_aux:Nnnwnn #1#2#3#4 \q_stop #5#6
+  {
+    \use_none:n #2
+    \use_none:n #5
+    #1 {#3} {#6}
+    \seq_mapthread_function_aux:Nnnwnn #1 #4 \q_stop
+  }
+\cs_generate_variant:Nn \seq_mapthread_function:NNN {     Nc }
+\cs_generate_variant:Nn \seq_mapthread_function:NNN { c , cc }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}
+%   {
+%     \seq_set_from_clist:NN, \seq_set_from_clist:cN,
+%     \seq_set_from_clist:Nc, \seq_set_from_clist:cc,
+%     \seq_set_from_clist:Nn, \seq_set_from_clist:cn
+%   }
+% \begin{macro}
+%   {
+%     \seq_gset_from_clist:NN, \seq_gset_from_clist:cN,
+%     \seq_gset_from_clist:Nc, \seq_gset_from_clist:cc,
+%     \seq_gset_from_clist:Nn, \seq_gset_from_clist:cn
+%   }
+% \begin{macro}[aux]{\seq_wrap_item:n}
+%   Setting a sequence from a comma-separated list is done using a simple
+%   mapping.
+%    \begin{macrocode}
+\cs_new_protected:Npn \seq_set_from_clist:NN #1#2
+  {
+    \tl_set:Nx #1
+      { \clist_map_function:NN #2 \seq_wrap_item:n }
+  }
+\cs_new_protected:Npn \seq_set_from_clist:Nn #1#2
+  {
+    \tl_set:Nx #1
+      { \clist_map_function:nN {#2} \seq_wrap_item:n }
+  }
+\cs_new_protected:Npn \seq_gset_from_clist:NN #1#2
+  {
+    \tl_gset:Nx #1
+      { \clist_map_function:NN #2 \seq_wrap_item:n }
+  }
+\cs_new_protected:Npn \seq_gset_from_clist:Nn #1#2
+  {
+    \tl_gset:Nx #1
+      { \clist_map_function:nN {#2} \seq_wrap_item:n }
+  }
+\cs_new:Npn \seq_wrap_item:n #1 { \exp_not:n { \seq_item:n {#1} } }
+\cs_generate_variant:Nn \seq_set_from_clist:NN  {     Nc }
+\cs_generate_variant:Nn \seq_set_from_clist:NN  { c , cc }
+\cs_generate_variant:Nn \seq_set_from_clist:Nn  { c      }
+\cs_generate_variant:Nn \seq_gset_from_clist:NN {     Nc }
+\cs_generate_variant:Nn \seq_gset_from_clist:NN { c , cc }
+\cs_generate_variant:Nn \seq_gset_from_clist:Nn { c      }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \subsection{Deprecated interfaces}
+%
+%   A few functions which are no longer documented: these were moved
+%   here on or before 2011-04-20, and will be removed entirely by
+%   2011-07-20.
+%
+% \begin{macro}{\seq_top:NN, \seq_top:cN}
+%   These are old stack functions.
+%    \begin{macrocode}
+\cs_new_eq:NN \seq_top:NN \seq_get_left:NN
+\cs_new_eq:NN \seq_top:cN \seq_get_left:cN
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\seq_display:N, \seq_display:c}
+%   An older name for \cs{seq_show:N}.
+%    \begin{macrocode}
+\cs_new_eq:NN \seq_display:N \seq_show:N
+\cs_new_eq:NN \seq_display:c \seq_show:c
+%    \end{macrocode}
+% \end{macro}
+%
+%    \begin{macrocode}
+%</initex|package>
+%    \end{macrocode}
+%
+% \end{implementation}
+%
+% \PrintIndex
\ No newline at end of file
diff --git a/Master/texmf-dist/source/latex/l3kernel/l3skip.dtx b/Master/texmf-dist/source/latex/l3kernel/l3skip.dtx
new file mode 100644
index 00000000000..ecd3ca2e397
--- /dev/null
+++ b/Master/texmf-dist/source/latex/l3kernel/l3skip.dtx
@@ -0,0 +1,1675 @@
+% \iffalse meta-comment
+%
+%% File: l3skip.dtx Copyright (C) 2004-2011 Frank Mittelbach, The LaTeX3 Project
+%%
+%% It may be distributed and/or modified under the conditions of the
+%% LaTeX Project Public License (LPPL), either version 1.3c of this
+%% license or (at your option) any later version.  The latest version
+%% of this license is in the file
+%%
+%%    http://www.latex-project.org/lppl.txt
+%%
+%% This file is part of the "expl3 bundle" (The Work in LPPL)
+%% and all files in that bundle must be distributed together.
+%%
+%% The released version of this bundle is available from CTAN.
+%%
+%% -----------------------------------------------------------------------
+%%
+%% The development version of the bundle can be found at
+%%
+%%    http://www.latex-project.org/svnroot/experimental/trunk/
+%%
+%% for those people who are interested.
+%%
+%%%%%%%%%%%
+%% NOTE: %%
+%%%%%%%%%%%
+%%
+%%   Snapshots taken from the repository represent work in progress and may
+%%   not work or may contain conflicting material!  We therefore ask
+%%   people _not_ to put them into distributions, archives, etc. without
+%%   prior consultation with the LaTeX3 Project.
+%%
+%% -----------------------------------------------------------------------
+%
+%<*driver|package>
+\RequirePackage{l3names}
+\GetIdInfo$Id: l3skip.dtx 2478 2011-06-19 21:34:23Z joseph $
+  {L3 Experimental dimensions and skips}
+%</driver|package>
+%<*driver>
+\documentclass[full]{l3doc}
+\begin{document}
+  \DocInput{\jobname.dtx}
+\end{document}
+%</driver>
+% \fi
+%
+% \title{^^A
+%   The \pkg{l3skip} package\\ Dimensions and skips^^A
+%   \thanks{This file describes v\ExplFileVersion,
+%      last revised \ExplFileDate.}^^A
+% }
+%
+% \author{^^A
+%  The \LaTeX3 Project\thanks
+%    {^^A
+%      E-mail:
+%        \href{mailto:latex-team@latex-project.org}
+%          {latex-team@latex-project.org}^^A
+%    }^^A
+% }
+%
+% \date{Released \ExplFileDate}
+%
+% \maketitle
+%
+% \begin{documentation}
+%
+% \LaTeX3 provides two general length variables: \texttt{dim} and
+% \texttt{skip}. Lengths stored as \texttt{dim} variables have a fixed
+% length, whereas \texttt{skip} lengths have a rubber (stretch/shrink)
+% component. In addition, the \texttt{muskip} type is available for
+% use in math mode: this is a special form of \texttt{skip} where the
+% lengths involved are determined by the current math font (in
+% \texttt{mu)}. There are common features in the creation and setting of
+% length variables, but for clarity the functions are grouped by variable
+% type.
+%
+% \section{Creating and initialising \texttt{dim} variables}
+%
+% \begin{function}{\dim_new:N, \dim_new:c}
+%   \begin{syntax}
+%     \cs{dim_new:N} \meta{dimension}
+%   \end{syntax}
+%   Creates a new \meta{dimension} or raises an error if the name is
+%   already taken. The declaration is global. The \meta{dimension}
+%   will initially be equal to $0$\,pt.
+% \end{function}
+%
+% \begin{function}{\dim_zero:N, \dim_zero:c}
+%   \begin{syntax}
+%     \cs{dim_zero:N} \meta{dimension}
+%   \end{syntax}
+%   Sets \meta{dimension} to $0$\,pt within the scope of the current
+%   \TeX{} group.
+% \end{function}
+%
+% \begin{function}{\dim_gzero:N, \dim_gzero:c}
+%   \begin{syntax}
+%     \cs{dim_gzero:N} \meta{dimension}
+%   \end{syntax}
+%   Sets \meta{dimension} to $0$\,pt globally, \emph{i.e.}~not
+%   restricted by the current \TeX{} group level.
+% \end{function}
+%
+% \section{Setting \texttt{dim} variables}
+%
+% \begin{function}{\dim_add:Nn, \dim_add:cn}
+%   \begin{syntax}
+%     \cs{dim_add:Nn} \meta{dimension} \Arg{dimension expression}
+%   \end{syntax}
+%   Adds the result of the \meta{dimension expression} to the current
+%   content of the \meta{dimension}. This assignment is local.
+% \end{function}
+%
+% \begin{function}{\dim_gadd:Nn, \dim_gadd:cn}
+%   \begin{syntax}
+%     \cs{dim_gadd:Nn} \meta{dimension} \Arg{dimension expression}
+%   \end{syntax}
+%   Adds the result of the \meta{dimension expression} to the
+%   current content of the \meta{dimension}. This assignment is global.
+% \end{function}
+%
+% \begin{function}{\dim_set:Nn, \dim_set:cn}
+%   \begin{syntax}
+%     \cs{dim_set:Nn} \meta{dimension} \Arg{dimension expression}
+%   \end{syntax}
+%   Sets \meta{dimension} to the value of \meta{dimension expression}, which
+%   must evaluate to a length with units.  This assignment is
+%   restricted to the current \TeX{} group.
+% \end{function}
+%
+% \begin{function}{\dim_gset:Nn, \dim_gset:cn}
+%   \begin{syntax}
+%     \cs{dim_gset:Nn} \meta{dimension} \Arg{dimension expression}
+%   \end{syntax}
+%   Sets \meta{dimension} to the value of \meta{dimension expression}, which
+%   must evaluate to a length with units and may include a rubber
+%   component (for example |1 cm plus 0.5 cm|.  This assignment is
+%   global and is not limited to the current \TeX{} group level.
+% \end{function}
+% 
+% \begin{function}
+%   {\dim_set_eq:NN, \dim_set_eq:cN, \dim_set_eq:Nc, \dim_set_eq:cc}
+%   \begin{syntax}
+%     \cs{dim_set_eq:NN} \meta{dimension1} \meta{dimension2}
+%   \end{syntax}
+%   Sets the content of \meta{dimension1} equal to that of
+%   \meta{dimension2}. This assignment is restricted to the current
+%   \TeX{} group level.
+% \end{function}
+%
+% \begin{function}
+%   {\dim_gset_eq:NN, \dim_gset_eq:cN, \dim_gset_eq:Nc, \dim_gset_eq:cc}
+%   \begin{syntax}
+%     \cs{dim_gset_eq:NN} \meta{dimension1} \meta{dimension2}
+%   \end{syntax}
+%   Sets the content of \meta{dimension1} equal to that of \meta{dimension2}.
+%   This assignment is global and so is not limited by the current
+%   \TeX{} group level.
+% \end{function}
+%
+% \begin{function}{\dim_set_max:Nn, \dim_set_max:cn}
+%   \begin{syntax}
+%     \cs{dim_set_max:Nn} \meta{dimension} \Arg{dimension expression}
+%   \end{syntax}
+%   Compares the current value of the \meta{dimension} with that of the
+%   \meta{dimension expression}, and sets the \meta{dimension} to the
+%   larger of these two value. This assignment is local to the current
+%   \TeX{} group.
+% \end{function}
+%
+% \begin{function}{\dim_gset_max:Nn, \dim_gset_max:cn}
+%   \begin{syntax}
+%     \cs{dim_gset_max:Nn} \meta{dimension} \Arg{dimension expression}
+%   \end{syntax}
+%   Compares the current value of the \meta{dimension} with that of the
+%   \meta{dimension expression}, and sets the \meta{dimension} to the
+%   larger of these two value. This assignment is global.
+% \end{function}
+%
+% \begin{function}{\dim_set_min:Nn, \dim_set_min:cn}
+%   \begin{syntax}
+%     \cs{dim_set_min:Nn} \meta{dimension} \Arg{dimension expression}
+%   \end{syntax}
+%   Compares the current value of the \meta{dimension} with that of the
+%   \meta{dimension expression}, and sets the \meta{dimension} to the
+%   smaller of these two value. This assignment is local to the current
+%   \TeX{} group.
+% \end{function}
+%
+% \begin{function}{\dim_gset_min:Nn, \dim_gset_min:cn}
+%   \begin{syntax}
+%     \cs{dim_gset_min:Nn} \meta{dimension} \Arg{dimension expression}
+%   \end{syntax}
+%   Compares the current value of the \meta{dimension} with that of the
+%   \meta{dimension expression}, and sets the \meta{dimension} to the
+%   smaller of these two value. This assignment is global.
+% \end{function}
+%
+% \begin{function}{\dim_sub:Nn, \dim_sub:cn}
+%   \begin{syntax}
+%     \cs{dim_sub:Nn} \meta{dimension} \Arg{dimension expression}
+%   \end{syntax}
+%   Subtracts the result of the \meta{dimension expression} to the
+%   current content of the \meta{dimension}. This assignment is local.
+% \end{function}
+%
+% \begin{function}{\dim_gsub:Nn, \dim_gsub:cn}
+%   \begin{syntax}
+%     \cs{dim_gsub:Nn} \meta{dimension} \Arg{dimension expression}
+%   \end{syntax}
+%   Subtracts the result of the \meta{dimension expression} to the
+%   current content of the \meta{dimension}. This assignment is global.
+% \end{function}
+% 
+% \section{Utilities for dimension calculations}
+% 
+% \begin{function}[EXP]{\dim_ratio:nn}
+%   \begin{syntax}
+%     \cs{dim_ratio:nn} \Arg{dimexpr1} \Arg{dimexpr2}
+%   \end{syntax}
+%   Parses the two \meta{dimension expressions} and converts the ratio of
+%   the two to a form suitable for use inside a \meta{dimension expression}.
+%   This ratio is then left in the input stream, allowing syntax such as
+%   \begin{verbatim}
+%     \dim_set:Nn \l_my_dim
+%       { 10 pt * \dim_ratio:nn { 5 pt } { 10 pt } }
+%   \end{verbatim}
+%   The output of \cs{dim_ratio:nn} on full expansion is a ration expression
+%   between two integers, with all distances converted to scaled points.
+%   Thus 
+%   \begin{verbatim}
+%     \tl_set:Nx \l_my_tl { \dim_ratio:nn { 5 pt } { 10 pt } }
+%     \tl_show:N \l_my_tl
+%   \end{verbatim}
+%   will display |327680/655360| on the terminal.
+% \end{function}
+%
+% \section{Dimension expression conditionals}
+%
+% \begin{function}[EXP,pTF]{\dim_compare:nNn}
+%   \begin{syntax}
+%     \cs{dim_compare_p:nNn}
+%     ~~\Arg{dimexpr1} \meta{relation} \Arg{dimexpr2}
+%     \cs{dim_compare:nNnTF}
+%     ~~\Arg{dimexpr1} \meta{relation} \Arg{dimexpr2}
+%     ~~\Arg{true code} \Arg{false code}
+%   \end{syntax}
+%   This function first evaluates each of the \meta{dimension expressions}
+%   as described for \cs{dim_eval:n}. The two results are then
+%   compared using the \meta{relation}:
+%   \begin{center}
+%     \begin{tabular}{ll}
+%       Equal                 & |=| \\
+%       Greater than          & |>| \\
+%       Less than             & |<| \\
+%     \end{tabular}
+%   \end{center}
+%   The branching versions then leave either \meta{true code} or
+%   \meta{false code} in the input stream, as appropriate to the truth
+%   of the test and the variant of the function chosen. The logical
+%   truth of the test is left in the input stream by the predicate
+%   version.
+% \end{function}
+%
+% \begin{function}[EXP,pTF]{\dim_compare:n}
+%   \begin{syntax}
+%     \cs{dim_compare_p:n}
+%     ~~\{ \meta{dimexpr1} \meta{relation} \meta{dimexpr2} \}
+%     \cs{dim_compare:nTF}
+%     ~~\{ \meta{dimexpr1} \meta{relation} \meta{dimexpr2} \}
+%     ~~\Arg{true code} \Arg{false code}
+%   \end{syntax}
+%   This function first evaluates each of the \meta{dimension expressions}
+%   as described for \cs{dim_eval:n}. The two results are then
+%   compared using the \meta{relation}:
+%   \begin{center}
+%     \begin{tabular}{ll}
+%       Equal                    & |=| or |==| \\
+%       Greater than or equal to & |=>|        \\
+%       Greater than             & |>|         \\
+%       Less than or equal to    & |=<|        \\
+%       Less than                & |<|         \\
+%       Not equal                & |!=|        \\
+%     \end{tabular}
+%   \end{center}
+%   The branching versions then leave either \meta{true code} or
+%   \meta{false code} in the input stream, as appropriate to the truth
+%   of the test and the variant of the function chosen. The logical
+%   truth of the test is left in the input stream by the predicate
+%   version.
+% \end{function}
+%
+% \section{Dimension expression loops}
+%
+% \begin{function}[EXP]{\dim_do_while:nNnn}
+%   \begin{syntax}
+%      \cs{dim_do_while:nNnn}
+%      ~~\Arg{dimexpr1} \meta{relation} \Arg{dimexpr2} \Arg{code}
+%   \end{syntax}
+%   Evaluates the relationship between the two \meta{dimension expressions}
+%   as described for \cs{dim_compare:nNnTF}, and then places the
+%   \meta{code} in the input stream if the \meta{relation} is
+%   \texttt{true}. After the \meta{code} has been processed by \TeX{} the
+%   test will be repeated, and a loop will occur until the test is
+%   \texttt{false}.
+% \end{function}
+%
+% \begin{function}[EXP]{\dim_do_until:nNnn}
+%   \begin{syntax}
+%      \cs{dim_do_until:nNnn}
+%      ~~\Arg{dimexpr1} \meta{relation} \Arg{dimexpr2} \Arg{code}
+%   \end{syntax}
+%   Evaluates the relationship between the two \meta{dimension expressions}
+%   as described for \cs{dim_compare:nNnTF}, and then places the
+%   \meta{code} in the input stream if the \meta{relation} is
+%   \texttt{false}. After the \meta{code} has been processed by \TeX{} the
+%   test will be repeated, and a loop will occur until the test is
+%   \texttt{true}.
+% \end{function}
+%
+% \begin{function}[EXP]{\dim_until_do:nNnn}
+%   \begin{syntax}
+%      \cs{dim_until_do:nNnn}
+%      ~~\Arg{dimexpr1} \meta{relation} \Arg{dimexpr2} \Arg{code}
+%   \end{syntax}
+%   Places the \meta{code} in the input stream for \TeX{} to process, and
+%   then evaluates the relationship between the two
+%   \meta{dimension expressions} as described for \cs{dim_compare:nNnTF}.
+%   If the test is \texttt{false} then the \meta{code} will be inserted
+%   into the input stream again and a loop will occur until the
+%   \meta{relation} is \texttt{true}.
+% \end{function}
+%
+% \begin{function}[EXP]{\dim_while_do:nNnn}
+%   \begin{syntax}
+%      \cs{dim_while_do:nNnn}
+%      ~~\Arg{dimexpr1} \meta{relation} \Arg{dimexpr2} \Arg{code}
+%   \end{syntax}
+%   Places the \meta{code} in the input stream for \TeX{} to process, and
+%   then evaluates the relationship between the two
+%   \meta{dimension expressions} as described for \cs{dim_compare:nNnTF}.
+%   If the test is \texttt{true} then the \meta{code} will be inserted
+%   into the input stream again and a loop will occur until the
+%   \meta{relation} is \texttt{false}.
+% \end{function}
+%
+% \begin{function}[EXP]{\dim_do_while:nn}
+%   \begin{syntax}
+%      \cs{dim_do_while:nNnn}
+%      ~~\{ \meta{dimexpr1} \meta{relation} \meta{dimexpr2} \} \Arg{code}
+%   \end{syntax}
+%   Evaluates the relationship between the two \meta{dimension expressions}
+%   as described for \cs{dim_compare:nTF}, and then places the
+%   \meta{code} in the input stream if the \meta{relation} is
+%   \texttt{true}. After the \meta{code} has been processed by \TeX{} the
+%   test will be repeated, and a loop will occur until the test is
+%   \texttt{false}.
+% \end{function}
+%
+% \begin{function}[EXP]{\dim_do_until:nn}
+%   \begin{syntax}
+%      \cs{dim_do_until:nn}
+%      ~~\{ \meta{dimexpr1} \meta{relation} \meta{dimexpr2} \} \Arg{code}
+%   \end{syntax}
+%   Evaluates the relationship between the two \meta{dimension expressions}
+%   as described for \cs{dim_compare:nTF}, and then places the
+%   \meta{code} in the input stream if the \meta{relation} is
+%   \texttt{false}. After the \meta{code} has been processed by \TeX{} the
+%   test will be repeated, and a loop will occur until the test is
+%   \texttt{true}.
+% \end{function}
+%
+% \begin{function}[EXP]{\dim_until_do:nn}
+%   \begin{syntax}
+%      \cs{dim_until_do:nn}
+%      ~~\{ \meta{dimexpr1} \meta{relation} \meta{dimexpr2} \} \Arg{code}
+%   \end{syntax}
+%   Places the \meta{code} in the input stream for \TeX{} to process, and
+%   then evaluates the relationship between the two
+%   \meta{dimension expressions} as described for \cs{dim_compare:nTF}.
+%   If the test is \texttt{false} then the \meta{code} will be inserted
+%   into the input stream again and a loop will occur until the
+%   \meta{relation} is \texttt{true}.
+% \end{function}
+%
+% \begin{function}[EXP]{\dim_while_do:nn}
+%   \begin{syntax}
+%      \cs{dim_while_do:nn}
+%      ~~\{ \meta{dimexpr1} \meta{relation} \meta{dimexpr2} \} \Arg{code}
+%   \end{syntax}
+%   Places the \meta{code} in the input stream for \TeX{} to process, and
+%   then evaluates the relationship between the two
+%   \meta{dimension expressions} as described for \cs{dim_compare:nTF}.
+%   If the test is \texttt{true} then the \meta{code} will be inserted
+%   into the input stream again and a loop will occur until the
+%   \meta{relation} is \texttt{false}.
+% \end{function}
+%
+% \section{Using \texttt{dim} expressions and variables}
+%
+% \begin{function}[EXP]{\dim_eval:n}
+%   \begin{syntax}
+%     \cs{dim_eval:n} \Arg{dimension expression}
+%   \end{syntax}
+%   Evaluates the \meta{dimension expression}, expanding any
+%   dimensions and token list variables within the \meta{expression}
+%   to their content (without requiring \cs{dim_use:N}/\cs{tl_use:N})
+%   and applying the standard mathematical rules. The result of the
+%   calculation is left in the input stream as a
+%   \meta{dimension denotation} after two expansions. This will be
+%   expressed in points (\texttt{pt}), and will require suitable
+%   termination if used in a \TeX{}-style assignment as it is \emph{not}
+%   an \meta{internal dimension}.
+% \end{function}
+%
+% \begin{function}[EXP]{\dim_use:N, \dim_use:c}
+%   \begin{syntax}
+%     \cs{dim_use:N} \meta{dimension}
+%   \end{syntax}
+%   Recovers the content of a \meta{dimension} and places it directly
+%   in the input stream. An error will be raised if the variable does
+%   not exist or if it is invalid. Can be omitted in places where a
+%   \meta{dimension} is required (such as in the argument of
+%   \cs{dim_eval:n}).
+%   \begin{texnote}
+%     \cs{dim_use:N} is the \TeX{} primitive \cs{the}: this is one of
+%     several \LaTeX3 names for this primitive.
+%   \end{texnote}
+% \end{function}
+%
+% \section{Viewing \texttt{dim} variables}
+%
+% \begin{function}{\dim_show:N, \dim_show:c}
+%   \begin{syntax}
+%     \cs{dim_show:N} \meta{dimension}
+%    \end{syntax}
+%   Displays the value of the \meta{dimension} on the terminal.
+% \end{function}
+%
+% \section{Constant dimensions}
+%
+% \begin{variable}{\c_max_dim}
+%   The maximum value that can be stored as a dimension or skip (these
+%   are equivalent).
+% \end{variable}
+%
+% \begin{variable}{\c_zero_dim}
+%   A zero length as a dimension or a skip (these are equivalent).
+% \end{variable}
+%
+% \section{Scratch dimensions}
+%
+% \begin{variable}{\l_tmpa_dim, \l_tmpb_dim, \l_tmpc_dim}
+%   Scratch dimension for local assignment. These are never used by
+%   the kernel code, and so are safe for use with any \LaTeX3-defined
+%   function. However, they may be overwritten by other non-kernel
+%   code and so should only be used for short-term storage.
+% \end{variable}
+%
+% \begin{variable}{\g_tmpa_dim, \g_tmpb_dim}
+%   Scratch dimension for global assignment. These are never used by
+%   the kernel code, and so are safe for use with any \LaTeX3-defined
+%   function. However, they may be overwritten by other non-kernel
+%   code and so should only be used for short-term storage.
+% \end{variable}
+%
+% \section{Creating and initialising \texttt{skip} variables}
+%
+% \begin{function}{\skip_new:N, \skip_new:c}
+%   \begin{syntax}
+%     \cs{skip_new:N} \meta{skip}
+%   \end{syntax}
+%   Creates a new \meta{skip} or raises an error if the name is
+%   already taken. The declaration is global. The \meta{skip}
+%   will initially be equal to $0$\,pt.
+% \end{function}
+%
+% \begin{function}{\skip_zero:N, \skip_zero:c}
+%   \begin{syntax}
+%     \cs{skip_zero:N} \meta{skip}
+%   \end{syntax}
+%   Sets \meta{skip} to $0$\,pt within the scope of the current
+%   \TeX{} group.
+% \end{function}
+%
+% \begin{function}{\skip_gzero:N, \skip_gzero:c}
+%   \begin{syntax}
+%     \cs{skip_gzero:N} \meta{skip}
+%   \end{syntax}
+%   Sets \meta{skip} to $0$\,pt globally, \emph{i.e.}~not
+%   restricted by the current \TeX{} group level.
+% \end{function}
+%
+% \section{Setting \texttt{skip} variables}
+%
+% \begin{function}{\skip_add:Nn, \skip_add:cn}
+%   \begin{syntax}
+%     \cs{skip_add:Nn} \meta{skip} \Arg{skip expression}
+%   \end{syntax}
+%   Adds the result of the \meta{skip expression} to the current
+%   content of the \meta{skip}. This assignment is local.
+% \end{function}
+%
+% \begin{function}{\skip_gadd:Nn, \skip_gadd:cn}
+%   \begin{syntax}
+%     \cs{skip_gadd:Nn} \meta{skip} \Arg{skip expression}
+%   \end{syntax}
+%   Adds the result of the \meta{skip expression} to the
+%   current content of the \meta{skip}. This assignment is global.
+% \end{function}
+%
+% \begin{function}{\skip_set:Nn, \skip_set:cn}
+%   \begin{syntax}
+%     \cs{skip_set:Nn} \meta{skip} \Arg{skip expression}
+%   \end{syntax}
+%   Sets \meta{skip} to the value of \meta{skip expression}, which
+%   must evaluate to a length with units and may include a rubber
+%   component (for example |1 cm plus 0.5 cm|.  This assignment is
+%   restricted to the current \TeX{} group.
+% \end{function}
+%
+% \begin{function}
+%   {\skip_gset_eq:NN, \skip_gset_eq:cN, \skip_gset_eq:Nc, \skip_gset_eq:cc}
+%   \begin{syntax}
+%     \cs{skip_gset_eq:NN} \meta{skip1} \meta{skip2}
+%   \end{syntax}
+%   Sets the content of \meta{skip1} equal to that of \meta{skip2}.
+%   This assignment is global and so is not limited by the current
+%   \TeX{} group level.
+% \end{function}
+%
+% \begin{function}{\skip_gset:Nn, \skip_gset:cn}
+%   \begin{syntax}
+%     \cs{skip_gset:Nn} \meta{skip} \Arg{skip expression}
+%   \end{syntax}
+%   Sets \meta{skip} to the value of \meta{skip expression}, which
+%   must evaluate to a length with units and may include a rubber
+%   component (for example |1 cm plus 0.5 cm|.  This assignment is
+%   global and is not limited to the current \TeX{} group level.
+% \end{function}
+% 
+% \begin{function}
+%   {\skip_set_eq:NN, \skip_set_eq:cN, \skip_set_eq:Nc, \skip_set_eq:cc}
+%   \begin{syntax}
+%     \cs{skip_set_eq:NN} \meta{skip1} \meta{skip2}
+%   \end{syntax}
+%   Sets the content of \meta{skip1} equal to that of
+%   \meta{skip2}. This assignment is restricted to the current
+%   \TeX{} group level.
+% \end{function}
+%
+% \begin{function}{\skip_sub:Nn, \skip_sub:cn}
+%   \begin{syntax}
+%     \cs{skip_sub:Nn} \meta{skip} \Arg{skip expression}
+%   \end{syntax}
+%   Subtracts the result of the \meta{skip expression} to the
+%   current content of the \meta{skip}. This assignment is local.
+% \end{function}
+%
+% \begin{function}{\skip_gsub:Nn, \skip_gsub:cn}
+%   \begin{syntax}
+%     \cs{skip_gsub:Nn} \meta{skip} \Arg{skip expression}
+%   \end{syntax}
+%   Subtracts the result of the \meta{skip expression} to the
+%   current content of the \meta{skip}. This assignment is global.
+% \end{function}
+%
+% \section{Skip expression conditionals}
+%
+%  \begin{function}[EXP,pTF]{\skip_if_eq:nn}
+%    \begin{syntax}
+%      \cs{skip_if_eq_p:nn}
+%      ~~\Arg{skipexpr1} \Arg{skipexpr2}
+%      \cs{dim_compare:nTF}
+%      ~~\Arg{skipexpr1} \Arg{skipexpr2}
+%      ~~\Arg{true code} \Arg{false code}
+%    \end{syntax}
+%    This function first evaluates each of the
+%    \meta{skip expressions} as described for \cs{skip_eval:n}.
+%    The two results are then compared for exact equality,
+%    \emph{i.e.}~both the fixed and rubber components must be the same
+%    for the test to be true.  The branching versions then leave either
+%    \meta{true code} or \meta{false code} in the input stream, as
+%    appropriate to the truth of the test and the variant of the function
+%    chosen. The logical truth of the test is left in the input stream by
+%    the predicate version.
+%  \end{function}
+%
+% \begin{function}[EXP,pTF]{\skip_if_infinite_glue:n}
+%   \begin{syntax}
+%     \cs{skip_if_infinite_glue_p:n} \Arg{skipexpr}
+%     \cs{skip_if_infinite_glue:nTF} \Arg{skipexpr}
+%     ~~\Arg{true code} \Arg{false code}
+%   \end{syntax}
+%   Evaluates the \meta{skip expression} as described for \cs{skip_eval:n},
+%   and then tests if this contains an infinite stretch or shrink
+%   component (or both).
+%   The branching versions then leave either \meta{true code} or
+%   \meta{false code} in the input stream, as appropriate to the truth
+%   of the test and the variant of the function chosen. The logical
+%   truth of the test is left in the input stream by the predicate
+% \end{function}
+%
+% \section{Using \texttt{skip} expressions and variables}
+%
+% \begin{function}[EXP]{\skip_eval:n}
+%   \begin{syntax}
+%     \cs{skip_eval:n} \Arg{skip expression}
+%   \end{syntax}
+%   Evaluates the \meta{skip expression}, expanding any skips
+%   and token list variables within the \meta{expression}
+%   to their content (without requiring \cs{skip_use:N}/\cs{tl_use:N})
+%   and applying the standard mathematical rules. The result of the
+%   calculation is left in the input stream as a \meta{glue denotation}
+%   after two expansions. This will be expressed in points (\texttt{pt}),
+%   and will require suitable termination if used in a \TeX{}-style
+%   assignment as it is \emph{not} an \meta{internal glue}.
+% \end{function}
+%
+% \begin{function}[EXP]{\skip_use:N, \skip_use:c}
+%   \begin{syntax}
+%     \cs{skip_use:N} \meta{skip}
+%   \end{syntax}
+%   Recovers the content of a \meta{skip} and places it directly
+%   in the input stream. An error will be raised if the variable does
+%   not exist or if it is invalid. Can be omitted in places where a
+%   \meta{dimension} is required (such as in the argument of
+%   \cs{skip_eval:n}).
+%   \begin{texnote}
+%     \cs{skip_use:N} is the \TeX{} primitive \cs{the}: this is one of
+%     several \LaTeX3 names for this primitive.
+%   \end{texnote}
+% \end{function}
+%
+% \section{Viewing \texttt{skip} variables}
+%
+% \begin{function}{\skip_show:N, \skip_show:c}
+%   \begin{syntax}
+%     \cs{skip_show:N} \meta{skip}
+%    \end{syntax}
+%   Displays the value of the \meta{skip} on the terminal.
+% \end{function}
+%
+% \section{Constant skips}
+%
+% \begin{variable}{\c_max_skip}
+%   The maximum value that can be stored as a dimension or skip (these
+%   are equivalent).
+% \end{variable}
+%
+% \begin{variable}{\c_zero_skip}
+%   A zero length as a dimension or a skip (these are equivalent).
+% \end{variable}
+%
+% \section{Scratch skips}
+%
+% \begin{variable}{\l_tmpa_skip, \l_tmpb_skip, \l_tmpc_skip}
+%   Scratch skip for local assignment. These are never used by
+%   the kernel code, and so are safe for use with any \LaTeX3-defined
+%   function. However, they may be overwritten by other non-kernel
+%   code and so should only be used for short-term storage.
+% \end{variable}
+%
+% \begin{variable}{\g_tmpa_skip, \g_tmpb_skip}
+%   Scratch skip for global assignment. These are never used by
+%   the kernel code, and so are safe for use with any \LaTeX3-defined
+%   function. However, they may be overwritten by other non-kernel
+%   code and so should only be used for short-term storage.
+% \end{variable}
+%
+% \section{Creating and initialising \texttt{muskip} variables}
+%
+% \begin{function}{\muskip_new:N, \muskip_new:c}
+%   \begin{syntax}
+%     \cs{muskip_new:N} \meta{muskip}
+%   \end{syntax}
+%   Creates a new \meta{muskip} or raises an error if the name is
+%   already taken. The declaration is global. The \meta{muskip}
+%   will initially be equal to $0$\,mu.
+% \end{function}
+%
+% \begin{function}{\muskip_zero:N, \muskip_zero:c}
+%   \begin{syntax}
+%     \cs{skip_zero:N} \meta{muskip}
+%   \end{syntax}
+%   Sets \meta{muskip} to $0$\,mu within the scope of the current
+%   \TeX{} group.
+% \end{function}
+%
+% \begin{function}{\muskip_gzero:N, \muskip_gzero:c}
+%   \begin{syntax}
+%     \cs{muskip_gzero:N} \meta{muskip}
+%   \end{syntax}
+%   Sets \meta{muskip} to $0$\,mu globally, \emph{i.e.}~not
+%   restricted by the current \TeX{} group level.
+% \end{function}
+%
+% \section{Setting \texttt{muskip} variables}
+%
+% \begin{function}{\muskip_add:Nn, \muskip_add:cn}
+%   \begin{syntax}
+%     \cs{muskip_add:Nn} \meta{muskip} \Arg{muskip expression}
+%   \end{syntax}
+%   Adds the result of the \meta{muskip expression} to the current
+%   content of the \meta{muskip}. This assignment is local.
+% \end{function}
+%
+% \begin{function}{\muskip_gadd:Nn, \muskip_gadd:cn}
+%   \begin{syntax}
+%     \cs{muskip_gadd:Nn} \meta{muskip} \Arg{muskip expression}
+%   \end{syntax}
+%   Adds the result of the \meta{muskip expression} to the
+%   current content of the \meta{muskip}. This assignment is global.
+% \end{function}
+%
+% \begin{function}{\muskip_set:Nn, \muskip_set:cn}
+%   \begin{syntax}
+%     \cs{muskip_set:Nn} \meta{muskip} \Arg{muskip expression}
+%   \end{syntax}
+%   Sets \meta{muskip} to the value of \meta{muskip expression}, which
+%   must evaluate to a math length with units and may include a rubber
+%   component (for example |1 mu plus 0.5 mu|.  This assignment is
+%   restricted to the current \TeX{} group.
+% \end{function}
+%
+% \begin{function}{\muskip_gset:Nn, \muskip_gset:cn}
+%   \begin{syntax}
+%     \cs{muskip_gset:Nn} \meta{muskip} \Arg{muskip expression}
+%   \end{syntax}
+%   Sets \meta{muskip} to the value of \meta{muskip expression}, which
+%   must evaluate to a math length with units and may include a rubber
+%   component (for example |1 mu plus 0.5 mu|.  This assignment is
+%   global and is not limited to the current \TeX{} group level.
+% \end{function}
+% 
+% \begin{function}
+%   {
+%     \muskip_set_eq:NN, \muskip_set_eq:cN,
+%     \muskip_set_eq:Nc, \muskip_set_eq:cc
+%   }
+%   \begin{syntax}
+%     \cs{muskip_set_eq:NN} \meta{muskip1} \meta{muskip2}
+%   \end{syntax}
+%   Sets the content of \meta{muskip1} equal to that of
+%   \meta{muskip2}. This assignment is restricted to the current
+%   \TeX{} group level.
+% \end{function}
+%
+% \begin{function}
+%   {
+%     \muskip_gset_eq:NN, \muskip_gset_eq:cN,
+%     \muskip_gset_eq:Nc, \muskip_gset_eq:cc
+%    }
+%   \begin{syntax}
+%     \cs{muskip_gset_eq:NN} \meta{muskip1} \meta{muskip2}
+%   \end{syntax}
+%   Sets the content of \meta{muskip1} equal to that of \meta{muskip2}.
+%   This assignment is global and so is not limited by the current
+%   \TeX{} group level.
+% \end{function}
+%
+% \begin{function}{\muskip_sub:Nn, \muskip_sub:cn}
+%   \begin{syntax}
+%     \cs{muskip_sub:Nn} \meta{muskip} \Arg{muskip expression}
+%   \end{syntax}
+%   Subtracts the result of the \meta{muskip expression} to the
+%   current content of the \meta{skip}. This assignment is local.
+% \end{function}
+%
+% \begin{function}{\muskip_gsub:Nn, \muskip_gsub:cn}
+%   \begin{syntax}
+%     \cs{muskip_gsub:Nn} \meta{muskip} \Arg{muskip expression}
+%   \end{syntax}
+%   Subtracts the result of the \meta{muskip expression} to the
+%   current content of the \meta{muskip}. This assignment is global.
+% \end{function}
+%
+% \section{Using \texttt{muskip} expressions and variables}
+%
+% \begin{function}[EXP]{\muskip_eval:n}
+%   \begin{syntax}
+%     \cs{muskip_eval:n} \Arg{muskip expression}
+%   \end{syntax}
+%   Evaluates the \meta{muskip expression}, expanding any skips
+%   and token list variables within the \meta{expression}
+%   to their content (without requiring \cs{muskip_use:N}/\cs{tl_use:N})
+%   and applying the standard mathematical rules. The result of the
+%   calculation is left in the input stream as a \meta{muglue denotation}
+%   after two expansions. This will be expressed in \texttt{mu},
+%   and will require suitable termination if used in a \TeX{}-style
+%   assignment as it is \emph{not} an \meta{internal muglue}.
+% \end{function}
+%
+% \begin{function}[EXP]{\muskip_use:N, \muskip_use:c}
+%   \begin{syntax}
+%     \cs{muskip_use:N} \meta{muskip}
+%   \end{syntax}
+%   Recovers the content of a \meta{skip} and places it directly
+%   in the input stream. An error will be raised if the variable does
+%   not exist or if it is invalid. Can be omitted in places where a
+%   \meta{dimension} is required (such as in the argument of
+%   \cs{muskip_eval:n}).
+%   \begin{texnote}
+%     \cs{muskip_use:N} is the \TeX{} primitive \cs{the}: this is one of
+%     several \LaTeX3 names for this primitive.
+%   \end{texnote}
+% \end{function}
+%
+% \section{Inserting skips into the output}
+%
+% \begin{function}{\skip_horizontal:N, \skip_horizontal:c, \skip_horizontal:n}
+%   \begin{syntax}
+%     \cs{skip_horizontal:N} \meta{skip}
+%     \cs{skip_horizontal:n} \Arg{skipexpr}
+%   \end{syntax}
+%   Inserts a horizontal \meta{skip} into the current list.
+%   \begin{texnote}
+%     \cs{skip_horizontal:N} is the \TeX{} primitive \cs{hskip} renamed.
+%   \end{texnote}
+% \end{function}
+%
+% \begin{function}{\skip_vertical:N, \skip_vertical:c, \skip_vertical:n}
+%   \begin{syntax}
+%     \cs{skip_vertical:N} \meta{skip}
+%     \cs{skip_vertical:n} \Arg{skipexpr}
+%   \end{syntax}
+%   Inserts a vertical \meta{skip} into the current list.
+%   \begin{texnote}
+%     \cs{skip_vertical:N} is the \TeX{} primitive \cs{vskip} renamed.
+%   \end{texnote}
+% \end{function}
+%
+% \section{Viewing \texttt{muskip} variables}
+%
+% \begin{function}{\muskip_show:N, \muskip_show:c}
+%   \begin{syntax}
+%     \cs{muskip_show:N} \meta{muskip}
+%    \end{syntax}
+%   Displays the value of the \meta{muskip} on the terminal.
+% \end{function}
+%
+% \section{Internal functions}
+%
+% \begin{function}{\if_dim:w}
+%   \begin{syntax}
+%     \cs{if_dim:w} \meta{dimen1} \meta{relation} \meta{dimen1}
+%     ~~\meta{true code}
+%     \cs{else:}
+%     ~~\meta{false}
+%     \cs{fi:}
+%   \end{syntax}
+%   Compare two dimensions. The \meta{relation} is one of
+%   |<|, |=| or |>| with category code $12$.
+%   \begin{texnote}
+%     This is the \TeX{} primitive \cs{ifdim}.
+%   \end{texnote}
+% \end{function}
+% 
+% \begin{function}[EXP]{\dim_eval:w, \dim_eval_end:}
+%   \begin{syntax}
+%     \cs{dim_eval:w} \meta{dimexpr} \cs{dim_eval_end:}
+%   \end{syntax}
+%   Evaluates \meta{dimension expression} as described for \cs{dim_eval:n}.
+%   The evaluation stops when an unexpandable token which is not a valid
+%   part of a dimension is read or when \cs{dim_eval_end:} is
+%   reached. The latter is gobbled by the scanner mechanism:
+%   \cs{dim_eval_end:} itself is unexpandable but used correctly
+%   the entire construct is expandable.
+%   \begin{texnote}
+%     This is the \eTeX{} primitive \cs{dimexpr}.
+%   \end{texnote}
+% \end{function}
+%
+% \section{Experimental skip functions}
+%
+% \begin{function}{\skip_split_finite_else_action:nnNN}
+%   \begin{syntax}
+%     \cs{skip_split_finite_else_action:nnNN} \Arg{skipexpr} \Arg{action}
+%     ~~\meta{dimen1} \meta{dimen2}
+%   \end{syntax}
+%   Checks if the \meta{skipexpr} contains finite glue. If it does then it
+%   assigns
+%   \meta{dimen1} the stretch component and \meta{dimen2} the shrink
+%   component. If
+%   it contains infinite glue set \meta{dimen1} and \meta{dimen2} to $0$\,pt
+%   and place |#2| into the input stream: this is usually an error or
+%   warning message of some sort.
+% \end{function}
+%
+% \end{documentation}
+%
+% \begin{implementation}
+%
+% \section{\pkg{l3skip} implementation}
+%
+%    \begin{macrocode}
+%<*initex|package>
+%    \end{macrocode}
+%
+%    \begin{macrocode}
+%<*package>
+\ProvidesExplPackage
+  {\ExplFileName}{\ExplFileDate}{\ExplFileVersion}{\ExplFileDescription}
+\package_check_loaded_expl:
+%</package>
+%    \end{macrocode}
+%
+% \subsection{Length primitives renamed}
+%
+% \begin{macro}{\if_dim:w}
+% \begin{macro}{\dim_eval:w}
+% \begin{macro}{\dim_eval_end:}
+%   Primitives renamed.
+%    \begin{macrocode}
+\cs_new_eq:NN \if_dim:w      \tex_ifdim:D
+\cs_new_eq:NN \dim_eval:w    \etex_dimexpr:D
+\cs_new_eq:NN \dim_eval_end: \tex_relax:D
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \subsection{Creating and initialising \texttt{dim} variables}
+%
+% \begin{macro}{\dim_new:N,\dim_new:c}
+%    Allocating \meta{dim} registers \ldots
+%    \begin{macrocode}
+%<*initex>
+\alloc_new:nnnN { dim } \c_zero \c_max_register_int \tex_dimendef:D
+%</initex>
+%<*package>
+\cs_new_protected_nopar:Npn \dim_new:N #1
+  {
+    \chk_if_free_cs:N #1
+    \newdimen #1
+  }
+%</package>
+\cs_generate_variant:Nn \dim_new:N { c }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\dim_zero:N, \dim_zero:c}
+% \begin{macro}{\dim_gzero:N, \dim_gzero:c}
+%   Reset the register to zero.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \dim_zero:N #1 { #1 \c_zero_dim }
+\cs_new_protected_nopar:Npn \dim_gzero:N { \pref_global:D \dim_zero:N }
+\cs_generate_variant:Nn \dim_zero:N  { c }
+\cs_generate_variant:Nn \dim_gzero:N { c }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \subsection{Setting \texttt{dim} variables}
+%
+% \begin{macro}{\dim_set:Nn, \dim_set:cn}
+% \begin{macro}{\dim_gset:Nn, \dim_gset:cn}
+%   Setting dimensions is easy enough.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \dim_set:Nn #1#2
+  { #1 ~ \dim_eval:w #2 \dim_eval_end: }
+\cs_new_protected_nopar:Npn \dim_gset:Nn { \pref_global:D \dim_set:Nn }
+\cs_generate_variant:Nn \dim_set:Nn  { c }
+\cs_generate_variant:Nn \dim_gset:Nn { c }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% 
+% \begin{macro}{\dim_set_eq:NN,\dim_set_eq:cN, \dim_set_eq:Nc,\dim_set_eq:cc}
+% \begin{macro}
+%   {\dim_gset_eq:NN,\dim_gset_eq:cN, \dim_gset_eq:Nc,\dim_gset_eq:cc}
+%   All straightforward.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \dim_set_eq:NN #1#2 { #1 = #2 }
+\cs_generate_variant:Nn \dim_set_eq:NN {       c }
+\cs_generate_variant:Nn \dim_set_eq:NN { Nc , cc }
+\cs_new_protected_nopar:Npn \dim_gset_eq:NN #1#2 { \pref_global:D #1 = #2 }
+\cs_generate_variant:Nn \dim_gset_eq:NN {       c }
+\cs_generate_variant:Nn \dim_gset_eq:NN { Nc , cc }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\dim_set_max:Nn, \dim_set_max:cn}
+% \begin{macro}{\dim_set_min:Nn, \dim_set_min:cn}
+% \begin{macro}{\dim_gset_max:Nn, \dim_gset_max:cn}
+% \begin{macro}{\dim_gset_min:Nn, \dim_gset_min:cn}
+%   Setting maximum and minimum values is simply a case of so build-in
+%   comparison. This only applies to dimensions as skips are not ordered.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \dim_set_max:Nn #1#2
+  { \dim_compare:nNnT {#1} < {#2} { \dim_set:Nn #1 {#2} } }
+\cs_new_protected_nopar:Npn \dim_gset_max:Nn #1#2
+  { \dim_compare:nNnT {#1} < {#2} { \dim_gset:Nn #1 {#2} } }
+\cs_new_protected_nopar:Npn \dim_set_min:Nn #1#2
+  { \dim_compare:nNnT {#1} > {#2} { \dim_set:Nn #1 {#2} } }
+\cs_new_protected_nopar:Npn \dim_gset_min:Nn #1#2
+  { \dim_compare:nNnT {#1} > {#2} { \dim_gset:Nn #1 {#2} } }
+\cs_generate_variant:Nn \dim_set_max:Nn  { c }
+\cs_generate_variant:Nn \dim_gset_max:Nn { c }
+\cs_generate_variant:Nn \dim_set_min:Nn  { c }
+\cs_generate_variant:Nn \dim_gset_min:Nn { c }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\dim_add:Nn, \dim_add:cn}
+% \begin{macro}{\dim_gadd:Nn, \dim_gadd:cn}
+% \begin{macro}{\dim_sub:Nn, \dim_sub:cn}
+% \begin{macro}{\dim_gsub:Nn, \dim_gsub:cn}
+%   Using |by| here deals with the (incorrect) case |\dimen123|.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \dim_add:Nn #1#2
+  { \tex_advance:D #1 by \dim_eval:w #2 \dim_eval_end: }
+\cs_new_protected_nopar:Npn \dim_gadd:Nn { \pref_global:D \dim_add:Nn }
+\cs_generate_variant:Nn \dim_add:Nn  { c }
+\cs_generate_variant:Nn \dim_gadd:Nn { c }
+\cs_new_protected_nopar:Npn \dim_sub:Nn #1#2
+  { \tex_advance:D #1 by - \dim_eval:w #2 \dim_eval_end: }
+\cs_new_protected_nopar:Npn \dim_gsub:Nn { \pref_global:D \dim_sub:Nn }
+\cs_generate_variant:Nn \dim_sub:Nn  { c }
+\cs_generate_variant:Nn \dim_gsub:Nn { c }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% 
+% \subsection{Utilities for dimension calculations}
+%
+% \begin{macro}{\dim_ratio:nn}
+% \begin{macro}[aux]{\dim_ratio_aux:n}
+%   With dimension expressions, something like |10 pt * ( 5 pt / 10 pt )| will
+%   not work. Instead, the ratio part needs to be converted to an integer
+%   expression. Using \cs{int_value:w} forces everything into |sp|, avoiding
+%   any decimal parts.
+%    \begin{macrocode}
+\cs_new_nopar:Npn \dim_ratio:nn #1#2
+  { \dim_ratio_aux:n {#1} / \dim_ratio_aux:n {#2} }
+\cs_new_nopar:Npn \dim_ratio_aux:n #1
+  { \exp_after:wN \int_value:w \dim_eval:w #1 \dim_eval_end: }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \subsection{Dimension expression conditionals}
+%
+% \begin{macro}{\dim_compare_p:nNn}
+% \begin{macro}[TF]{\dim_compare:nNn}
+%    \begin{macrocode}
+\prg_new_conditional:Npnn \dim_compare:nNn #1#2#3 { p , T , F , TF }
+  {
+    \if_dim:w \dim_eval:w #1 #2 \dim_eval:w #3 \dim_eval_end:
+      \prg_return_true: \else: \prg_return_false: \fi:
+  }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}[pTF]{\dim_compare:n}
+% \begin{macro}[aux]{\dim_compare_aux:wNN}
+% \begin{macro}[aux]{\dim_compare_<:nw}
+% \begin{macro}[aux]{\dim_compare_=:nw}
+% \begin{macro}[aux]{\dim_compare_>:nw}
+% \begin{macro}[aux]{\dim_compare_==:nw}
+% \begin{macro}[aux]{\dim_compare_<=:nw}
+% \begin{macro}[aux]{\dim_compare_!=:nw}
+% \begin{macro}[aux]{\dim_compare_>=:nw}
+%   [This code plus comments are adapted from the \cs{int_compare:nTF}
+%   function.]
+%   Comparison tests using a simple syntax where only one set of braces
+%   is required and additional operators such as |!=| and |>=| are
+%   supported. First some notes on the idea behind this. We wish to
+%   support writing code like
+%   \begin{verbatim}
+%     \dim_compare_p:n { 5mm + \l_tmpa_dim >= 4pt - \l_tmpb_dim }
+%   \end{verbatim}
+%   In other words, we want to somehow add the missing |\dim_eval:w|
+%   where required.  We can start evaluating from the left using
+%   |\dim_use:N \dim_eval:w|, and we know that since the relation
+%   symbols |<|, |>|, |=| and |!| are not allowed in such expressions,
+%   they will terminate the expression. Therefore, we first let \TeX{}
+%   evaluate this left hand side of the (in)equality.
+%
+%   Eventually, we will convert the relation symbol to the appropriate
+%   version of \cs{if_dim:w}, and add \cs{dim_eval:w} after it. We
+%   optimize by placing the end-code already here: this avoids needless
+%   grabbing of arguments later.
+%    \begin{macrocode}
+\prg_new_conditional:Npnn \dim_compare:n #1 { p , T , F , TF }
+  {
+    \exp_after:wN \dim_compare_aux:wNN \dim_use:N \dim_eval:w #1
+        \dim_eval_end:
+      \prg_return_true:
+    \else:
+      \prg_return_false:
+    \fi:
+  }
+%    \end{macrocode}
+%
+%   Contrarily to the case of integers, where we have to remove the
+%   result in order to access the relation, \cs{dim_use:N} nicely
+%   produces a result which ends in |pt|. We can thus use a delimited
+%   argument to find the relation. \cs{tl_to_str:n} is needed to
+%   convert |pt| to \enquote{other} characters.
+%
+%   The relation might be one character, |#2|, or two characters |#2#3|.
+%   We support the following forms: |=|, |<|, |>| and the extended |!=|,
+%   |==|, |<=| and |>=|. All the extended forms have an extra |=| so we
+%   check if that is present as well. Then use specific function to
+%   perform the (unbalanced) test.
+%    \begin{macrocode}
+\exp_args:Nno \use:nn
+  { \cs_new:Npn \dim_compare_aux:wNN #1 }
+  { \tl_to_str:n { pt } }
+  #2 #3
+    {
+      \use:c
+        {
+          dim_compare_ #2
+            \if_meaning:w = #3 = \fi:
+          :nw
+        }
+      { #1 pt } #3
+    }
+%    \end{macrocode}
+%   Here, \cs{dim_eval:w} will begin the right hand side of a dimension
+%   comparison (with \cs{if_dim:w}), closed cleanly by the trailing tokens
+%   we put in the definition of \cs{dim_compare:n}.
+%
+%   The actual comparisons take as a first argument the left-hand side
+%   of the comparison (a length). In the case of normal comparisons,
+%   just place the relevant \cs{if_dim:w}, with a trailing \cs{dim_eval:w}
+%   to evaluate the right hand side. For extended comparisons, remove
+%   the trailing |=| that we left, before evaluating with \cs{dim_eval:w}.
+%   In both cases, the expansion of \cs{dim_eval:w} is stopped properly,
+%   and the conditional ended correctly by the tokens we put in the
+%   definition of \cs{dim_compare:n}.
+%
+%   Equal, less than and greater than are straighforward.
+%    \begin{macrocode}
+\cs_new:cpn { dim_compare_<:nw } #1 { \if_dim:w #1 < \dim_eval:w }
+\cs_new:cpn { dim_compare_=:nw } #1 { \if_dim:w #1 = \dim_eval:w }
+\cs_new:cpn { dim_compare_>:nw } #1 { \if_dim:w #1 > \dim_eval:w }
+%    \end{macrocode}
+%   For the extended syntax |==|, we remove |#2|, trailing |=| sign,
+%   and otherwise act as for |=|.
+%    \begin{macrocode}
+\cs_new:cpn {dim_compare_==:nw} #1#2 { \if_dim:w #1 = \dim_eval:w }
+%    \end{macrocode}
+%
+%   Not equal, greater than or equal, less than or equal follow the
+%   same scheme as the extended equality syntax, with an additional
+%   \cs{reverse_if:N} to get the opposite of their \enquote{simple}
+%   analog.
+%    \begin{macrocode}
+\cs_new:cpn {dim_compare_<=:nw} #1#2 {\reverse_if:N \if_dim:w #1 > \dim_eval:w}
+\cs_new:cpn {dim_compare_!=:nw} #1#2 {\reverse_if:N \if_dim:w #1 = \dim_eval:w}
+\cs_new:cpn {dim_compare_>=:nw} #1#2 {\reverse_if:N \if_dim:w #1 < \dim_eval:w}
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+%
+%
+% \subsection{Dimension expression loops}
+%
+% \begin{macro}{\dim_while_do:nn}
+% \begin{macro}{\dim_until_do:nn}
+% \begin{macro}{\dim_do_while:nn}
+% \begin{macro}{\dim_do_until:nn}
+%   |while_do| and |do_while| functions for dimensions. Same as for the
+%   |int| type only the names have changed.
+%    \begin{macrocode}
+\cs_set:Npn \dim_while_do:nn #1#2
+  {
+    \dim_compare:nT {#1}
+      {
+        #2
+        \dim_while_do:nn {#1} {#2}
+      }
+  }
+\cs_set:Npn \dim_until_do:nn #1#2
+  {
+    \dim_compare:nF {#1}
+      {
+        #2
+        \dim_until_do:nn {#1} {#2}
+      }
+  }
+\cs_set:Npn \dim_do_while:nn #1#2
+  {
+    #2
+    \dim_compare:nT {#1}
+      { \dim_do_while:nNnn {#1} {#2} }
+  }
+\cs_set:Npn \dim_do_until:nn #1#2
+  {
+    #2
+    \dim_compare:nF {#1}
+      { \dim_do_until:nn {#1} {#2} }
+  }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\dim_while_do:nNnn}
+% \begin{macro}{\dim_until_do:nNnn}
+% \begin{macro}{\dim_do_while:nNnn}
+% \begin{macro}{\dim_do_until:nNnn}
+%   |while_do| and |do_while| functions for dimensions. Same as for the
+%   |int| type only the names have changed.
+%    \begin{macrocode}
+\cs_set:Npn \dim_while_do:nNnn #1#2#3#4
+  {
+    \dim_compare:nNnT {#1} #2 {#3}
+      {
+        #4
+        \dim_while_do:nNnn {#1} #2 {#3} {#4}
+      }
+  }
+\cs_set:Npn \dim_until_do:nNnn #1#2#3#4
+  {
+  \dim_compare:nNnF {#1} #2 {#3}
+    {
+      #4
+      \dim_until_do:nNnn {#1} #2 {#3} {#4}
+    }
+  }
+\cs_set:Npn \dim_do_while:nNnn #1#2#3#4
+  {
+    #4
+    \dim_compare:nNnT {#1} #2 {#3}
+      { \dim_do_while:nNnn {#1} #2 {#3} {#4} }
+  }
+\cs_set:Npn \dim_do_until:nNnn #1#2#3#4
+  {
+    #4
+    \dim_compare:nNnF {#1} #2 {#3}
+      { \dim_do_until:nNnn {#1} #2 {#3} {#4} }
+  }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \subsection{Using \texttt{dim} expressions and variables}
+%
+% \begin{macro}{\dim_eval:n}
+%   Evaluating a dimension expression expandably.
+%    \begin{macrocode}
+\cs_new_nopar:Npn \dim_eval:n #1
+  { \dim_use:N \dim_eval:w #1 \dim_eval_end: }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\dim_use:N, \dim_use:c}
+%   Accessing a \meta{dim}.
+%    \begin{macrocode}
+\cs_new_eq:NN \dim_use:N \tex_the:D
+\cs_generate_variant:Nn \dim_use:N { c }
+%    \end{macrocode}
+% \end{macro}
+%
+% \subsection{Viewing \texttt{dim} variables}
+%
+% \begin{macro}{\dim_show:N, \dim_show:c}
+%   Diagnostics.
+%    \begin{macrocode}
+\cs_new_eq:NN  \dim_show:N \kernel_register_show:N
+\cs_generate_variant:Nn \dim_show:N { c }
+%    \end{macrocode}
+%  \end{macro}
+%
+% \subsection{Constant dimensions}
+%
+% \begin{variable}{\c_zero_dim}
+% \begin{variable}{\c_max_dim}
+%   The source for these depends on whether we are in package mode.
+%    \begin{macrocode}
+%<*initex>
+\dim_new:N \c_zero_dim 
+\dim_new:N \c_max_dim
+\dim_set:Nn \c_max_dim { 16383.99999 pt }
+%</initex>
+%<*package>
+\cs_new_eq:NN \c_zero_dim \z@
+\cs_new_eq:NN \c_max_dim  \maxdimen
+%</package>
+%    \end{macrocode}
+% \end{variable}
+% \end{variable}
+%
+% \subsection{Scratch dimensions}
+%
+% \begin{variable}{\l_tmpa_dim, \l_tmpb_dim, \l_tmpc_dim}
+% \begin{variable}{\g_tmpa_dim, \g_tmpb_dim}
+%    We provide three local and two global scratch registers, maybe we
+%    need more or less.
+%    \begin{macrocode}
+\dim_new:N \l_tmpa_dim
+\dim_new:N \l_tmpb_dim
+\dim_new:N \l_tmpc_dim
+\dim_new:N \g_tmpa_dim
+\dim_new:N \g_tmpb_dim
+%    \end{macrocode}
+% \end{variable}
+% \end{variable}
+%
+% \subsection{Creating and initialising \texttt{skip} variables}
+%
+% \begin{macro}{\skip_new:N,\skip_new:c}
+%    Allocation of a new internal registers.
+%    \begin{macrocode}
+%<*initex>
+\alloc_new:nnnN { skip } \c_zero \c_max_register_int \tex_skipdef:D
+%</initex>
+%<*package>
+\cs_new_protected_nopar:Npn \skip_new:N #1
+  {
+    \chk_if_free_cs:N #1
+    \newskip #1
+  }
+%</package>
+\cs_generate_variant:Nn \skip_new:N { c }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\skip_zero:N, \skip_zero:c}
+% \begin{macro}{\skip_gzero:N, \skip_gzero:c}
+%   Reset the register to zero.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \skip_zero:N #1 { #1 \c_zero_skip }
+\cs_new_protected_nopar:Npn \skip_gzero:N { \pref_global:D \skip_zero:N }
+\cs_generate_variant:Nn \skip_zero:N  { c }
+\cs_generate_variant:Nn \skip_gzero:N { c }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \subsection{Setting \texttt{skip} variables}
+%
+% \begin{macro}{\skip_set:Nn, \skip_set:cn}
+% \begin{macro}{\skip_gset:Nn, \skip_gset:cn}
+%   Much the same as for dimensions.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \skip_set:Nn #1#2
+  { #1 ~ \etex_glueexpr:D #2 \scan_stop: }
+\cs_new_protected_nopar:Npn \skip_gset:Nn { \pref_global:D \skip_set:Nn }
+\cs_generate_variant:Nn \skip_set:Nn  { c }
+\cs_generate_variant:Nn \skip_gset:Nn { c }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% 
+% \begin{macro}
+%   {\skip_set_eq:NN,\skip_set_eq:cN, \skip_set_eq:Nc,\skip_set_eq:cc}
+% \begin{macro}
+%   {\skip_gset_eq:NN,\skip_gset_eq:cN, \skip_gset_eq:Nc,\skip_gset_eq:cc}
+%   All straightforward.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \skip_set_eq:NN #1#2 { #1 = #2 }
+\cs_generate_variant:Nn \skip_set_eq:NN {       c }
+\cs_generate_variant:Nn \skip_set_eq:NN { Nc , cc }
+\cs_new_protected_nopar:Npn \skip_gset_eq:NN #1#2 { \pref_global:D #1 = #2 }
+\cs_generate_variant:Nn \skip_gset_eq:NN {       c }
+\cs_generate_variant:Nn \skip_gset_eq:NN { Nc , cc }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\skip_add:Nn, \skip_add:cn}
+% \begin{macro}{\skip_gadd:Nn, \skip_gadd:cn}
+% \begin{macro}{\skip_sub:Nn, \skip_sub:cn}
+% \begin{macro}{\skip_gsub:Nn, \skip_gsub:cn}
+%   Using |by| here deals with the (incorrect) case |\skip123|.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \skip_add:Nn #1#2
+  { \tex_advance:D #1 by \etex_glueexpr:D #2 \scan_stop: }
+\cs_new_protected_nopar:Npn \skip_gadd:Nn { \pref_global:D \skip_add:Nn }
+\cs_generate_variant:Nn \skip_add:Nn  { c }
+\cs_generate_variant:Nn \skip_gadd:Nn { c }
+\cs_new_protected_nopar:Npn \skip_sub:Nn #1#2
+  { \tex_advance:D #1 by - \etex_glueexpr:D #2 \scan_stop: }
+\cs_new_protected_nopar:Npn \skip_gsub:Nn { \pref_global:D \skip_sub:Nn }
+\cs_generate_variant:Nn \skip_sub:Nn  { c }
+\cs_generate_variant:Nn \skip_gsub:Nn { c }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \subsection{Skip expression conditionals}
+%
+% \begin{macro}[pTF]{\skip_if_eq:nn}
+%   Comparing skips means doing two expansions to make strings, and then
+%   testing them. As a result, only equality is tested.
+%    \begin{macrocode}
+\prg_new_conditional:Npnn \skip_if_eq:nn #1#2 { p , T , F , TF }
+  {
+    \if_int_compare:w
+      \pdftex_strcmp:D { \skip_eval:n { #1 } } { \skip_eval:n { #2 } }
+      = \c_zero
+        \prg_return_true:
+    \else:
+        \prg_return_false:
+    \fi:
+  }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}[pTF]{\skip_if_infinite_glue:n}
+%   With \eTeX{} we all of a sudden get access to a lot information we
+%   should otherwise consider ourselves lucky to get. One is
+%   the stretch and shrink components of a skip register and the order
+%   or those components. cs{skip_if_infinite_glue:nTF} tests it directly by
+%   looking at the stretch and shrink order. If either of the predicate
+%   functions return \meta{true}, |\bool_if:nTF| will return \meta{true}
+%   and the logic test will take the true branch.
+%    \begin{macrocode}
+\prg_new_conditional:Npnn \skip_if_infinite_glue:n #1 { p , T , F , TF }
+  {
+    \bool_if:nTF
+      {
+        \int_compare_p:nNn { \etex_gluestretchorder:D #1 } > \c_zero ||
+        \int_compare_p:nNn { \etex_glueshrinkorder:D  #1 } > \c_zero
+      }
+      { \prg_return_true: }
+      { \prg_return_false: }
+ }
+%    \end{macrocode}
+%  \end{macro}
+%
+% \subsection{Using \texttt{skip} expressions and variables}
+%
+% \begin{macro}{\skip_eval:n}
+%   Evaluating a skip expression expandably.
+%    \begin{macrocode}
+\cs_new_nopar:Npn \skip_eval:n #1
+  { \skip_use:N \etex_glueexpr:D #1 \scan_stop: }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\skip_use:N, \skip_use:c}
+%   Accessing a \meta{skip}.
+%    \begin{macrocode}
+\cs_new_eq:NN \skip_use:N \tex_the:D
+\cs_generate_variant:Nn \skip_use:N { c }
+%    \end{macrocode}
+% \end{macro}
+%
+% \subsection{Inserting skips into the output}
+%
+% \begin{macro}{\skip_horizontal:N, \skip_horizontal:c, \skip_horizontal:n}
+% \begin{macro}{\skip_vertical:N, \skip_vertical:c, \skip_vertical:n}
+%    Inserting skips.
+%    \begin{macrocode}
+\cs_new_eq:NN  \skip_horizontal:N \tex_hskip:D
+\cs_new_nopar:Npn \skip_horizontal:n #1
+  { \skip_horizontal:N \etex_glueexpr:D #1 \scan_stop: }
+\cs_new_eq:NN  \skip_vertical:N \tex_vskip:D
+\cs_new_nopar:Npn \skip_vertical:n #1
+  { \skip_vertical:N \etex_glueexpr:D #1 \scan_stop: }
+\cs_generate_variant:Nn \skip_horizontal:N { c }
+\cs_generate_variant:Nn \skip_vertical:N { c }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \subsection{Viewing \texttt{skip} variables}
+%
+% \begin{macro}{\skip_show:N, \skip_show:c}
+%   Diagnostics.
+%    \begin{macrocode}
+\cs_new_eq:NN  \skip_show:N \kernel_register_show:N
+\cs_generate_variant:Nn \skip_show:N { c }
+%    \end{macrocode}
+%  \end{macro}
+%
+% \subsection{Constant skips}
+%
+% \begin{macro}{\c_zero_skip}
+% \begin{macro}{\c_max_skip}
+%   Skips with no rubber component are just dimensions
+%    \begin{macrocode}
+\cs_new_eq:NN \c_zero_skip \c_zero_dim
+\cs_new_eq:NN \c_max_skip  \c_max_dim
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \subsection{Scratch skips}
+%
+% \begin{variable}{\l_tmpa_skip, \l_tmpb_skip, \l_tmpc_skip}
+% \begin{variable}{\g_tmpa_skip, \g_tmpb_skip}
+%    We provide three local and two global scratch registers, maybe we
+%    need more or less.
+%    \begin{macrocode}
+\skip_new:N \l_tmpa_skip
+\skip_new:N \l_tmpb_skip
+\skip_new:N \l_tmpc_skip
+\skip_new:N \g_tmpa_skip
+\skip_new:N \g_tmpb_skip
+%    \end{macrocode}
+% \end{variable}
+% \end{variable}
+%
+% \subsection{Creating and initialising \texttt{muskip} variables}
+%
+% \begin{macro}{\muskip_new:N, \muskip_new:c}
+% And then we add muskips.
+%    \begin{macrocode}
+%<*initex>
+\alloc_new:nnnN { muskip } \c_zero \c_max_register_int \tex_muskipdef:D
+%</initex>
+%<*package>
+\cs_new_protected_nopar:Npn \muskip_new:N #1
+  {
+    \chk_if_free_cs:N #1
+    \newmuskip #1
+  }
+%</package>
+\cs_generate_variant:Nn \muskip_new:N { c }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\muskip_zero:N, \muskip_zero:c}
+% \begin{macro}{\muskip_gzero:N, \muskip_gzero:c}
+%   Reset the register to zero.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \muskip_zero:N #1
+  { #1 \c_zero_muskip }
+\cs_new_protected_nopar:Npn \muskip_gzero:N { \pref_global:D \muskip_zero:N }
+\cs_generate_variant:Nn \muskip_zero:N  { c }
+\cs_generate_variant:Nn \muskip_gzero:N { c }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \subsection{Setting \texttt{muskip} variables}
+%
+% \begin{macro}{\muskip_set:Nn, \muskip_set:cn}
+% \begin{macro}{\muskip_gset:Nn, \muskip_gset:cn}
+%   This should be pretty familiar.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \muskip_set:Nn #1#2
+  { #1 ~ \etex_muexpr:D #2 \scan_stop: }
+\cs_new_protected_nopar:Npn \muskip_gset:Nn { \pref_global:D \muskip_set:Nn }
+\cs_generate_variant:Nn \muskip_set:Nn  { c }
+\cs_generate_variant:Nn \muskip_gset:Nn { c }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% 
+% \begin{macro}
+%   {
+%     \muskip_set_eq:NN, \muskip_set_eq:cN,
+%     \muskip_set_eq:Nc, \muskip_set_eq:cc
+%   }
+% \begin{macro}
+%   {
+%     \muskip_gset_eq:NN,\muskip_gset_eq:cN,
+%     \muskip_gset_eq:Nc,\muskip_gset_eq:cc
+%   }
+%   All straightforward.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \muskip_set_eq:NN #1#2 { #1 = #2 }
+\cs_generate_variant:Nn \muskip_set_eq:NN {       c }
+\cs_generate_variant:Nn \muskip_set_eq:NN { Nc , cc }
+\cs_new_protected_nopar:Npn \muskip_gset_eq:NN #1#2 { \pref_global:D #1 = #2 }
+\cs_generate_variant:Nn \muskip_gset_eq:NN {       c }
+\cs_generate_variant:Nn \muskip_gset_eq:NN { Nc , cc }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\muskip_add:Nn, \muskip_add:cn}
+% \begin{macro}{\muskip_gadd:Nn, \muskip_gadd:cn}
+% \begin{macro}{\muskip_sub:Nn, \muskip_sub:cn}
+% \begin{macro}{\muskip_gsub:Nn, \muskip_gsub:cn}
+%   Using |by| here deals with the (incorrect) case |\muskip123|.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \muskip_add:Nn #1#2
+  { \tex_advance:D #1 by \etex_muexpr:D #2 \scan_stop: }
+\cs_new_protected_nopar:Npn \muskip_gadd:Nn { \pref_global:D \muskip_add:Nn }
+\cs_generate_variant:Nn \muskip_add:Nn  { c }
+\cs_generate_variant:Nn \muskip_gadd:Nn { c }
+\cs_new_protected_nopar:Npn \muskip_sub:Nn #1#2
+  { \tex_advance:D #1 by - \etex_muexpr:D #2 \scan_stop: }
+\cs_new_protected_nopar:Npn \muskip_gsub:Nn { \pref_global:D \muskip_sub:Nn }
+\cs_generate_variant:Nn \muskip_sub:Nn  { c }
+\cs_generate_variant:Nn \muskip_gsub:Nn { c }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \subsection{Using \texttt{muskip} expressions and variables}
+%
+% \begin{macro}{\muskip_eval:n}
+%   Evaluating a muskip expression expandably.
+%    \begin{macrocode}
+\cs_new_nopar:Npn \muskip_eval:n #1
+  { \muskip_use:N \etex_muexpr:D #1 \scan_stop: }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\muskip_use:N, \muskip_use:c}
+%   Accessing a \meta{muskip}.
+%    \begin{macrocode}
+\cs_new_eq:NN \muskip_use:N \tex_the:D
+\cs_generate_variant:Nn \muskip_use:N { c }
+%    \end{macrocode}
+% \end{macro}
+%
+% \subsection{Viewing \texttt{muskip} variables}
+%
+% \begin{macro}{\muskip_show:N, \muskip_show:c}
+%   Diagnostics.
+%    \begin{macrocode}
+\cs_new_eq:NN  \muskip_show:N \kernel_register_show:N
+\cs_generate_variant:Nn \muskip_show:N { c }
+%    \end{macrocode}
+%  \end{macro}
+%
+% \subsection{Experimental skip functions}
+%
+% \begin{macro}{\skip_split_finite_else_action:nnNN}
+%   This macro is useful when performing error checking in certain
+%   circumstances. If the \m{skip} register holds finite glue it sets
+%   |#3| and |#4| to the stretch and shrink component, resp. If it holds
+%   infinite glue set |#3| and |#4| to zero and issue the special action
+%   |#2| which is probably an error message.
+%   Assignments are global.
+%    \begin{macrocode}
+\cs_new_nopar:Npn \skip_split_finite_else_action:nnNN #1#2#3#4
+  {
+    \skip_if_infinite_glue:nTF {#1}
+    {
+      #3 = \c_zero_skip
+      #4 = \c_zero_skip
+      #2
+    }
+    {
+      #3 = \etex_gluestretch:D #1 \scan_stop:
+      #4 = \etex_glueshrink:D  #1 \scan_stop:
+    }
+  }
+%    \end{macrocode}
+%  \end{macro}
+%
+%    \begin{macrocode}
+%</initex|package>
+%    \end{macrocode}
+%
+% \end{implementation}
+%
+% \PrintIndex
diff --git a/Master/texmf-dist/source/latex/l3kernel/l3tl.dtx b/Master/texmf-dist/source/latex/l3kernel/l3tl.dtx
new file mode 100644
index 00000000000..a7f292d14c7
--- /dev/null
+++ b/Master/texmf-dist/source/latex/l3kernel/l3tl.dtx
@@ -0,0 +1,2268 @@
+% \iffalse meta-comment
+%
+%% File: l3tl.dtx Copyright (C) 1990-2011 The LaTeX3 Project
+%%
+%% It may be distributed and/or modified under the conditions of the
+%% LaTeX Project Public License (LPPL), either version 1.3c of this
+%% license or (at your option) any later version.  The latest version
+%% of this license is in the file
+%%
+%%    http://www.latex-project.org/lppl.txt
+%%
+%% This file is part of the "expl3 bundle" (The Work in LPPL)
+%% and all files in that bundle must be distributed together.
+%%
+%% The released version of this bundle is available from CTAN.
+%%
+%% -----------------------------------------------------------------------
+%%
+%% The development version of the bundle can be found at
+%%
+%%    http://www.latex-project.org/svnroot/experimental/trunk/
+%%
+%% for those people who are interested.
+%%
+%%%%%%%%%%%
+%% NOTE: %%
+%%%%%%%%%%%
+%%
+%%   Snapshots taken from the repository represent work in progress and may
+%%   not work or may contain conflicting material!  We therefore ask
+%%   people _not_ to put them into distributions, archives, etc. without
+%%   prior consultation with the LaTeX3 Project.
+%%
+%% -----------------------------------------------------------------------
+%
+%<*driver|package>
+\RequirePackage{l3names}
+\GetIdInfo$Id: l3tl.dtx 2498 2011-07-09 17:51:47Z joseph $
+  {L3 Experimental token lists}
+%</driver|package>
+%<*driver>
+\documentclass[full]{l3doc}
+\begin{document}
+  \DocInput{\jobname.dtx}
+\end{document}
+%</driver>
+% \fi
+%
+% \title{^^A
+%   The \pkg{l3tl} package\\ Token lists^^A
+%   \thanks{This file describes v\ExplFileVersion,
+%      last revised \ExplFileDate.}^^A
+% }
+%
+% \author{^^A
+%  The \LaTeX3 Project\thanks
+%    {^^A
+%      E-mail:
+%        \href{mailto:latex-team@latex-project.org}
+%          {latex-team@latex-project.org}^^A
+%    }^^A
+% }
+%
+% \date{Released \ExplFileDate}
+%
+% \maketitle
+%
+% \begin{documentation}
+%
+% \LaTeX3 stores lists of token in variables also called \enquote{token lists}.
+% Variables of this type get the suffix |tl| and functions of this type
+% have the prefix |tl|. To use a token list variable you simply call
+% the corresponding variable.
+%
+% Often you find yourself with not a token list variable but an
+% arbitrary token list which has to undergo certain tests. We will \emph{also}
+% prefix these functions with |tl|. While token list variables are
+% always single tokens, token lists are always surrounded by
+% braces. Many of the functions for token lists and token list variables
+% are very similar, and so are grouped together here.
+%
+% \section{Creating and initialising token list variables}
+%
+% \begin{function}{\tl_new:N, \tl_new:c}
+%   \begin{syntax}
+%     \cs{tl_new:N} \meta{tl~var}
+%   \end{syntax}
+%   Creates a new \meta{tl~var} or raises an error if the
+%   name is already taken. The declaration is global. The
+%   \meta{tl~var} will initially be empty.
+% \end{function}
+%
+% \begin{function}{\tl_const:Nn, \tl_const:Nx, \tl_const:cn, \tl_const:cx}
+%   \begin{syntax}
+%     \cs{tl_const:Nn} \meta{tl~var} \Arg{token list}
+%   \end{syntax}
+%   Creates a new constant \meta{tl~var} or raises an error
+%   if the name is already taken. The value of the
+%   \meta{tl~var} will be set globally to the
+%   \meta{token list}.
+% \end{function}
+%
+% \begin{function}{\tl_clear:N, \tl_clear:c}
+%   \begin{syntax}
+%     \cs{tl_clear:N} \meta{tl~var}
+%   \end{syntax}
+%   Clears all entries from the \meta{tl~var} within the scope of
+%   the current \TeX{} group.
+% \end{function}
+%
+% \begin{function}{\tl_gclear:N, \tl_gclear:c}
+%   \begin{syntax}
+%     \cs{tl_gclear:N} \meta{tl~var}
+%   \end{syntax}
+%   Clears all entries from the \meta{tl~var} globally.
+% \end{function}
+%
+% \begin{function}{\tl_clear_new:N, \tl_clear_new:c}
+%   \begin{syntax}
+%     \cs{tl_clear_new:N} \meta{tl~var}
+%   \end{syntax}
+%   If the \meta{tl~var} already exists, clears it within the scope
+%   of the current \TeX{} group. If the \meta{tl~var} is not defined,
+%   it will be created (using \cs{tl_new:N}). Thus the sequence is
+%   guaranteed to be available and clear within the current \TeX{}
+%   group. The \meta{tl~var} will exist globally, but the content
+%   outside of the current \TeX{} group is not specified.
+% \end{function}
+%
+% \begin{function}{\tl_gclear_new:N, \tl_gclear_new:c}
+%   \begin{syntax}
+%     \cs{tl_gclear_new:N} \meta{tl~var}
+%   \end{syntax}
+%   If the \meta{tl~var} already exists, clears it globally. If the
+%   \meta{tl~var} is not defined, it will be created (using
+%   \cs{tl_new:N}). Thus the sequence is guaranteed to be available
+%   and globally clear.
+% \end{function}
+%
+% \begin{function}{\tl_set_eq:NN, \tl_set_eq:cN, \tl_set_eq:Nc, \tl_set_eq:cc}
+%   \begin{syntax}
+%     \cs{tl_set_eq:NN} \meta{tl~var1} \meta{tl~var2}
+%   \end{syntax}
+%   Sets the content of \meta{tl~var1} equal to that of
+%   \meta{tl~var2}. This assignment is restricted to the current
+%   \TeX{} group level.
+% \end{function}
+%
+% \begin{function}
+%   {\tl_gset_eq:NN, \tl_gset_eq:cN, \tl_gset_eq:Nc, \tl_gset_eq:cc}
+%   \begin{syntax}
+%     \cs{tl_gset_eq:NN} \meta{tl~var1} \meta{tl~var2}
+%   \end{syntax}
+%   Sets the content of \meta{tl~var1} equal to that of
+%   \meta{tl~var2}. This assignment is global and so is not
+%   limited by the current \TeX{} group level.
+% \end{function}
+%
+% \section{Adding data to token list variables}
+%
+% \begin{function}
+%   {
+%     \tl_set:Nn, \tl_set:NV, \tl_set:Nv, \tl_set:No, \tl_set:Nf, \tl_set:Nx,
+%     \tl_set:cn, \tl_set:NV, \tl_set:Nv, \tl_set:co, \tl_set:cf, \tl_set:cx
+%   }
+%   \begin{syntax}
+%     \cs{tl_set:Nn} \meta{tl~var} \Arg{tokens}
+%   \end{syntax}
+%   Sets \meta{tl~var} to contain \meta{tokens},
+%   removing any previous content from the variable. This assignment
+%   is restricted to the current \TeX{} group.
+% \end{function}
+%
+% \begin{function}
+%   {
+%     \tl_gset:Nn, \tl_gset:NV, \tl_gset:Nv,
+%     \tl_gset:No, \tl_gset:Nf, \tl_gset:Nx,
+%     \tl_gset:cn, \tl_gset:cV, \tl_gset:cv,
+%     \tl_gset:co, \tl_gset:cf, \tl_gset:cx
+%   }
+%  \begin{syntax}
+%     \cs{tl_gset:Nn} \meta{tl~var} \Arg{tokens}
+%   \end{syntax}
+%   Sets \meta{tl~var} to contain \meta{tokens},
+%   removing any previous content from the variable. This assignment
+%   is global and is not limited to the current \TeX{} group level.
+% \end{function}
+%
+% \begin{function}
+%   {
+%     \tl_put_left:Nn, \tl_put_left:NV, \tl_put_left:No, \tl_put_left:Nx,
+%     \tl_put_left:cn, \tl_put_left:cV, \tl_put_left:co, \tl_put_left:cx
+%   }
+%   \begin{syntax}
+%     \cs{tl_put_left:Nn} \meta{tl~var} \Arg{tokens}
+%   \end{syntax}
+%   Appends \meta{tokens} to the left side of the current content of
+%   \meta{tl~var}. This modification is restricted to the
+%   current \TeX{} group level.
+% \end{function}
+%
+% \begin{function}
+%   {
+%     \tl_gput_left:Nn, \tl_gput_left:NV, \tl_gput_left:No, \tl_gput_left:Nx,
+%     \tl_gput_left:cn, \tl_gput_left:cV, \tl_gput_left:co, \tl_gput_left:cx
+%   }
+%   \begin{syntax}
+%     \cs{tl_gput_left:Nn} \meta{tl~var} \Arg{tokens}
+%   \end{syntax}
+%   Globally appends \meta{tokens} to the left side of the current
+%   content of \meta{tl~var}. This modification is not
+%   limited by \TeX{} grouping.
+% \end{function}
+%
+% \begin{function}
+%   {
+%     \tl_put_right:Nn, \tl_put_right:NV, \tl_put_right:No, \tl_put_right:Nx,
+%     \tl_put_right:cn, \tl_put_right:cV, \tl_put_right:co, \tl_put_right:cx
+%   }
+%   \begin{syntax}
+%     \cs{tl_put_right:Nn} \meta{tl~var} \Arg{tokens}
+%   \end{syntax}
+%   Appends \meta{tokens} to the right side of the current content of
+%   \meta{tl~var}. This modification is restricted to the
+%   current \TeX{} group level.
+% \end{function}
+%
+% \begin{function}
+%   {
+%     \tl_gput_right:Nn, \tl_gput_right:NV, \tl_gput_right:No,
+%     \tl_gput_right:Nx,
+%     \tl_gput_right:cn, \tl_gput_right:cV, \tl_gput_right:co,
+%     \tl_gput_right:cx
+%   }
+%   \begin{syntax}
+%     \cs{tl_gput_right:Nn} \meta{tl~var} \Arg{tokens}
+%   \end{syntax}
+%   Globally appends \meta{tokens} to the right side of the current
+%   content of \meta{tl~var}. This modification is not
+%   limited by \TeX{} grouping.
+% \end{function}
+%
+% \section{Modifying token list variables}
+%
+% \begin{function}{\tl_replace_once:Nnn, \tl_replace_once:cnn}
+%   \begin{syntax}
+%     \cs{tl_replace_once:Nnn} \meta{tl~var} \Arg{old tokens}
+%     ~~\Arg{new tokens}
+%   \end{syntax}
+%   Replaces the first (leftmost) occurrence of \meta{old tokens} in the
+%   \meta{tl~var} with \meta{new tokens}. \meta{Old tokens}
+%   cannot contain |{|, |}| or |#|
+%   (assuming normal \TeX{} category codes). The assignment is
+%   restricted to the current \TeX{} group.
+% \end{function}
+%
+% \begin{function}{\tl_greplace_once:Nnn, \tl_greplace_once:cnn}
+%   \begin{syntax}
+%     \cs{tl_greplace_once:Nnn} \meta{tl~var} \Arg{old tokens}
+%     ~~\Arg{new tokens}
+%   \end{syntax}
+%   Replaces the first (leftmost) occurrence of \meta{old tokens} in the
+%   \meta{tl~var} with \meta{new tokens}. \meta{Old tokens}
+%   cannot contain |{|, |}| or |#|
+%   (assuming normal \TeX{} category codes). The assignment is
+%   applied globally.
+% \end{function}
+%
+% \begin{function}{\tl_replace_all:Nnn, \tl_replace_all:cnn}
+%   \begin{syntax}
+%     \cs{tl_replace_all:Nnn} \meta{tl~var} \Arg{old tokens}
+%     ~~\Arg{new tokens}
+%   \end{syntax}
+%   Replaces all occurrences of \meta{old tokens} in the
+%   \meta{tl~var} with \meta{new tokens}. \meta{Old tokens}
+%   cannot contain |{|, |}| or |#|
+%   (assuming normal \TeX{} category codes). The assignment is
+%   restricted to the current \TeX{} group.
+% \end{function}
+%
+% \begin{function}{\tl_greplace_all:Nnn, \tl_greplace_all:cnn}
+%   \begin{syntax}
+%     \cs{tl_greplace_all:Nnn} \meta{tl~var} \Arg{old tokens}
+%     ~~\Arg{new tokens}
+%   \end{syntax}
+%   Replaces all occurrences of \meta{old tokens} in the
+%   \meta{tl~var} with \meta{new tokens}. \meta{Old tokens}
+%   cannot contain |{|, |}| or |#|
+%   (assuming normal \TeX{} category codes). The assignment is
+%   applied globally.
+% \end{function}
+%
+% \begin{function}{\tl_remove_once:Nn, \tl_remove_once:cn}
+%   \begin{syntax}
+%     \cs{tl_remove_once:Nn} \meta{tl~var} \Arg{tokens}
+%   \end{syntax}
+%   Removes the first (leftmost) occurrence of \meta{tokens} from the
+%   \meta{tl~var}. \meta{Tokens} cannot contain |{|, |}| or
+%   |#| (assuming normal \TeX{} category codes). The assignment is
+%   restricted to the current \TeX{} group.
+% \end{function}
+%
+% \begin{function}{\tl_gremove_once:Nn, \tl_gremove_once:cn}
+%   \begin{syntax}
+%     \cs{tl_gremove_once:Nn} \meta{tl~var} \Arg{tokens}
+%   \end{syntax}
+%   Removes the first (leftmost) occurrence of \meta{tokens} from the
+%   \meta{tl~var}. \meta{Tokens} cannot contain |{|, |}| or
+%   |#| (assuming normal \TeX{} category codes). The assignment is
+%   applied globally.
+% \end{function}
+%
+% \begin{function}{\tl_remove_all:Nn, \tl_remove_all:cn}
+%   \begin{syntax}
+%     \cs{tl_remove_all:Nn} \meta{tl~var} \Arg{tokens}
+%   \end{syntax}
+%   Removes all occurrences of \meta{tokens} from the
+%   \meta{tl~var}. \meta{Tokens} cannot contain |{|, |}| or
+%   |#| (assuming normal \TeX{} category codes). The assignment is
+%   restricted to the current \TeX{} group.
+% \end{function}
+%
+% \begin{function}{\tl_gremove_all:Nn, \tl_gremove_all:cn}
+%   \begin{syntax}
+%     \cs{tl_gremove_all:Nn} \meta{tl~var} \Arg{tokens}
+%   \end{syntax}
+%   Removes all occurrences of \meta{tokens} from the
+%   \meta{tl~var}. \meta{Tokens} cannot contain |{|, |}| or
+%   |#| (assuming normal \TeX{} category codes). The assignment is
+%   applied globally.
+% \end{function}
+%
+% \section{Reassigning token list category codes}
+%
+% \begin{function}
+%   {
+%     \tl_set_rescan:Nnn, \tl_set_rescan:Nno, \tl_set_rescan:Nnx,
+%     \tl_set_rescan:cnn, \tl_set_rescan:cno, \tl_set_rescan:cnx,
+%   }
+%   \begin{syntax}
+%     \cs{tl_set_rescan:Nnn} \meta{tl~var} \Arg{setup}
+%     ~~\Arg{tokens}
+%   \end{syntax}
+%   Sets \meta{tl~var} to contain \meta{tokens},
+%   applying the category code regim{\'e} specified in the
+%   \meta{setup} before carrying out the assignment. This allows the
+%   \meta{tl~var} to contain material with category codes
+%   other than those that apply when \meta{tokens} are absorbed. The
+%   assignment is local to the current \TeX{} group. See also
+%   \cs{tl_rescan:nn}.
+% \end{function}
+%
+% \begin{function}
+%   {
+%     \tl_gset_rescan:Nnn, \tl_gset_rescan:Nno, \tl_gset_rescan:Nnx,
+%     \tl_gset_rescan:cnn, \tl_gset_rescan:cno, \tl_gset_rescan:cnx,
+%   }
+%   \begin{syntax}
+%     \cs{tl_gset_rescan:Nnn} \meta{tl~var} \Arg{setup}
+%     ~~\Arg{tokens}
+%   \end{syntax}
+%   Sets \meta{tl~var} to contain \meta{tokens},
+%   applying the category code regim{\'e} specified in the
+%   \meta{setup} before carrying out the assignment. This allows the
+%   \meta{tl~var} to contain material with category codes
+%   other than those that apply when \meta{tokens} are absorbed. The
+%   assignment is global. See also \cs{tl_rescan:nn}.
+% \end{function}
+%
+% \begin{function}{\tl_rescan:nn}
+%   \begin{syntax}
+%     \cs{tl_rescan:nn} \Arg{setup} \Arg{tokens}
+%   \end{syntax}
+%   Rescans \meta{tokens} applying the category code regim{\'e} specified
+%   in the \meta{setup}, and leaves the resulting tokens in the input
+%   stream. See also \cs{tl_set_rescan:Nnn}.
+% \end{function}
+%
+% \section{Reassigning token list character codes}
+%
+% \begin{function}{\tl_to_lowercase:n}
+%   \begin{syntax}
+%     \cs{tl_to_lowercase:n} \Arg{tokens}
+%   \end{syntax}
+%   Works through all of the \meta{tokens}, replacing each character
+%   with the lower case equivalent as defined by \cs{char_set_lccode:nn}.
+%   Characters with no defined lower case character code are left
+%   unchanged. This process does not alter the category code assigned
+%   to the \meta{tokens}.
+%   \begin{texnote}
+%     This is the \TeX{} primitive \cs{lowercase} renamed.
+%     As a result, this function takes place on execution and
+%     not on expansion.
+%   \end{texnote}
+% \end{function}
+%
+% \begin{function}{\tl_to_uppercase:n}
+%   \begin{syntax}
+%     \cs{tl_to_uppercase:n} \Arg{tokens}
+%   \end{syntax}
+%   Works through all of the \meta{tokens}, replacing each character
+%   with the upper case equivalent as defined by \cs{char_set_uccode:nn}.
+%   Characters with no defined lower case character code are left
+%   unchanged. This process does not alter the category code assigned
+%   to the \meta{tokens}.
+%   \begin{texnote}
+%     This is the \TeX{} primitive \cs{uppercase} renamed..
+%     As a result, this function takes place on execution and
+%     not on expansion.
+%   \end{texnote}
+% \end{function}
+%
+% \section{Token list conditionals}
+%
+% \begin{function}[EXP,pTF]{\tl_if_blank:n, \tl_if_blank:V, \tl_if_blank:o}
+%   \begin{syntax}
+%     \cs{tl_if_blank_p:n} \Arg{token list}
+%     \cs{tl_if_blank:nTF} \Arg{token list} \Arg{true code} \Arg{false code}
+%   \end{syntax}
+%   Tests if the \meta{token list} consists only of blank spaces. The
+%   test is \texttt{true} if \meta{token list} is empty or consists
+%   entirely of explicit tokens of character code $32$
+%   and category code $10$, and is \texttt{false}
+%   otherwise. The branching versions then leave either \meta{true code}
+%   or \meta{false code}  in the input stream, as appropriate to the
+%   truth of the test and the variant of the function chosen. The logical
+%   truth of the test is left in the input stream by the predicate
+%   version.
+% \end{function}
+%
+% \begin{function}[EXP,pTF]{\tl_if_empty:N, \tl_if_empty:c}
+%   \begin{syntax}
+%     \cs{tl_if_empty_p:N} \meta{tl var}
+%     \cs{tl_if_empty:NTF} \meta{tl var} \Arg{true code} \Arg{false code}
+%   \end{syntax}
+%   Tests if the \meta{token list variable} is entirely empty
+%   (\emph{i.e.}~contains no tokens at all). The branching versions then
+%   leave either \meta{true code} or \meta{false code} in the input
+%   stream, as appropriate to the truth of the test and the variant of
+%   the function chosen. The logical truth of the test is left in the
+%   input stream by the predicate version.
+% \end{function}
+%
+% \begin{function}[EXP,pTF]{\tl_if_empty:n, \tl_if_empty:V, \tl_if_empty:o}
+%   \begin{syntax}
+%     \cs{tl_if_empty_p:n} \Arg{token list}
+%     \cs{tl_if_empty:nTF} \Arg{token list} \Arg{true code} \Arg{false code}
+%   \end{syntax}
+%   Tests if the \meta{token list} is entirely empty
+%   (\emph{i.e.}~contains no tokens at all). The branching versions then
+%   leave either \meta{true code} or \meta{false code} in the input
+%   stream, as appropriate to the truth of the test and the variant of
+%   the function chosen. The logical truth of the test is left in the
+%   input stream by the predicate version.
+% \end{function}
+%
+% \begin{function}[EXP,pTF]
+%   {\tl_if_eq:NN, \tl_if_eq:Nc, \tl_if_eq:cN, \tl_if_eq:cc}
+%   \begin{syntax}
+%     \cs{tl_if_eq_p:NN} \Arg{tl var1} \Arg{tl var2}
+%     \cs{tl_if_eq:NNTF} \Arg{tl var1} \Arg{tl var2} \Arg{true code}
+%     ~~\Arg{false code}
+%   \end{syntax}
+%   Compares the content of two \meta{token list variables} and
+%   is logically \texttt{true} if the two contain the same list of
+%   tokens (\emph{i.e.}~identical in both the list of characters they
+%   contain and the category codes of those characters). Thus for example
+%   \begin{verbatim}
+%     \tl_set:Nn \l_tmpa_tl { abc }
+%     \tl_set:Nx \l_tmpb_tl { \tl_to_str:n { abc } }
+%     \tl_if_eq_p:NN \l_tmpa_tl \l_tmpb_tl
+%   \end{verbatim}
+%   is logically \texttt{false}. The branching versions then leave either
+%   \meta{true code} or \meta{false code} in the input stream, as
+%   appropriate to the truth of the test and the variant of the function
+%   chosen. The logical truth of the test is left in the input stream by
+%   the predicate version.
+% \end{function}
+%
+% \begin{function}[TF]{\tl_if_eq:nn}
+%   \begin{syntax}
+%     \cs{tl_if_eq:nnTF} \meta{token list1} \Arg{token list2} \Arg{true code}
+%     ~~\Arg{false code}
+%   \end{syntax}
+%   Tests if \meta{token list1} and \meta{token list2} are equal, both in
+%   respect of character codes and category codes. Either the
+%   \meta{true code} or \meta{false code} in the input stream, as
+%   appropriate to the truth of the test and the variant of the function
+%   chosen.
+% \end{function}
+%
+% \begin{function}[TF]{\tl_if_in:Nn, \tl_if_in:cn}
+%   \begin{syntax}
+%     \cs{tl_if_in:NnTF} \meta{tl~var} \Arg{token list} \Arg{true code}
+%     ~~\Arg{false code}
+%   \end{syntax}
+%   Tests if the \meta{token list} is found in the content of the
+%   \meta{token list variable}. The \meta{token list} cannot contain
+%   the tokens |{|, |}| or |#| (assuming the usual \TeX{} category
+%   codes apply). Either the \meta{true code} or \meta{false code}
+%   is left in the
+%   input stream, as appropriate to the truth of the test and the variant
+%   of the function chosen.
+% \end{function}
+%
+% \begin{function}[TF]
+%   {\tl_if_in:nn, \tl_if_in:Vn, \tl_if_in:on, \tl_if_in:on}
+%   \begin{syntax}
+%     \cs{tl_if_in:nnTF} \meta{token list1} \Arg{token list2} \Arg{true code}
+%     ~~\Arg{false code}
+%   \end{syntax}
+%   Tests if the \meta{token list1} is found inside \meta{token list2}.
+%   The \meta{token list} cannot contain the tokens |{|, |}| or |#|
+%   (assuming the usual \TeX{} category codes apply). Either the
+%   \meta{true code} or \meta{false code} is left in the input stream, as
+%   appropriate to the truth of the test and the variant of the function
+%   chosen.
+% \end{function}
+%
+% \begin{function}[EXP,pTF]{\tl_if_single:N, \tl_if_single:c}
+%   \begin{syntax}
+%     \cs{tl_if_single_p:N} \Arg{tl~var}
+%     \cs{tl_if_single:nTF} \Arg{tl~var} \Arg{true code} \Arg{false code}
+%   \end{syntax}
+%   Tests if the content of the \meta{tl~var} consists of a
+%   single token or token group. The test is \texttt{true} if
+%   \meta{token list} contains exactly one token, if it consists
+%   entirely of explicit tokens of character code $32$ and category
+%   code $10$, or if it contains one braced token group (optionally
+%   preceeded by spaces), and it is \texttt{false} otherwise. The branching
+%   versions then leave either \meta{true code} or \meta{false code}
+%   in the input stream, as appropriate to the truth of the test and the
+%   variant of the function chosen. The logical truth of the test is left
+%   in the input stream by the predicate version.
+% \end{function}
+%
+% \begin{function}[EXP,pTF]{\tl_if_single:n}
+%   \begin{syntax}
+%     \cs{tl_if_single_p:n} \Arg{token list}
+%     \cs{tl_if_single:nTF} \Arg{token list} \Arg{true code} \Arg{false code}
+%   \end{syntax}
+%   Tests if the \meta{token list} consists of a single token
+%   or token group. The test is \texttt{true} if \meta{token list}
+%   contains exactly one token, or if it consists entirely of
+%   explicit tokens of character code $32$ and category code $10$,
+%   or if it contains a braced token group (optionally preceeded by blank
+%   spaces), and it is
+%   \texttt{false} otherwise. The branching versions then leave
+%   either \meta{true code} or \meta{false code}  in the input stream, as
+%   appropriate to the truth of the test and the variant of the function
+%   chosen. The logical truth of the test is left in the input stream by
+%   the predicate version.
+% \end{function}
+%
+% \section{Mapping to token lists}
+%
+% \begin{function}[EXP]{\tl_map_function:NN, \tl_map_function:cN}
+%   \begin{syntax}
+%     \cs{tl_map_function:NN} \meta{tl~var} \meta{function}
+%   \end{syntax}
+%   Applies \meta{function} to every \meta{token group} stored in the
+%   \meta{tl~var}. The \meta{function}
+%   will receive one argument for each iteration. This may be a number
+%   of tokens if the \meta{token group} was an item stored within
+%   braces. Hence the \meta{function} should anticipate receiving
+%   \texttt{n}-type arguments. See also \cs{tl_map_function:nN}.
+% \end{function}
+%
+% \begin{function}[EXP]{\tl_map_function:nN}
+%   \begin{syntax}
+%     \cs{tl_map_function:nN} \meta{token list} \meta{function}
+%   \end{syntax}
+%   Applies \meta{function} to every \meta{token group} in the
+%   \meta{token list}. The \meta{function}
+%   will receive one argument for each iteration. This may be a number
+%   of tokens if the \meta{token group} was an item stored within
+%   braces. Hence the \meta{function} should anticipate receiving
+%   \texttt{n}-type arguments. See also \cs{tl_map_function:NN}.
+% \end{function}
+%
+% \begin{function}{\tl_map_inline:Nn, \tl_map_inline:cn}
+%   \begin{syntax}
+%     \cs{tl_map_inline:Nn} \meta{tl~var} \Arg{inline function}
+%   \end{syntax}
+%   Applies the \meta{inline function} to every \meta{token group} stored
+%   within the \meta{tl~var}. The \meta{inline function}
+%   should consist of code which will receive the \meta{token group} as
+%   |#1|. One in line mapping can be nested inside another. See also
+%   \cs{tl_map_function:Nn}.
+% \end{function}
+%
+% \begin{function}{\tl_map_inline:nn}
+%   \begin{syntax}
+%     \cs{tl_map_inline:nn} \meta{token list} \Arg{inline function}
+%   \end{syntax}
+%   Applies the \meta{inline function} to every \meta{token group} stored
+%   within the \meta{token list}. The \meta{inline function}  should
+%   consist of code which will receive the \meta{token group} as
+%   |#1|. One in line mapping can be nested inside another. See also
+%   \cs{tl_map_function:nn}.
+% \end{function}
+%
+% \begin{function}{\tl_map_variable:NNn, \tl_map_variable:cNn}
+%   \begin{syntax}
+%     \cs{tl_map_variable:NNn} \meta{tl~var} \meta{variable} \Arg{function}
+%   \end{syntax}
+%   Applies the \meta{function} to every \meta{token group} stored
+%   within the \meta{tl~var}. The \meta{function} should consist of code
+%   which will receive the \meta{token group} stored in the \meta{variable}.
+%   One variable mapping can be nested inside another. See also
+%   \cs{tl_map_inline:Nn}.
+% \end{function}
+%
+% \begin{function}{\tl_map_variable:nNn}
+%   \begin{syntax}
+%     \cs{tl_map_variable:nNn} \meta{token list} \meta{variable} \Arg{function}
+%   \end{syntax}
+%   Applies the \meta{function} to every \meta{token group} stored
+%   within the \meta{token list}. The \meta{function} should consist of code
+%   which will receive the \meta{token group} stored in the \meta{variable}.
+%   One variable mapping can be nested inside another. See also
+%   \cs{tl_map_inline:nn}.
+% \end{function}
+%
+% \begin{function}[EXP]{\tl_map_break:}
+%   \begin{syntax}
+%     \cs{tl_map_break:}
+%   \end{syntax}
+%   Used to terminate a \cs{tl_map_\ldots} function before all
+%   entries in the \meta{token list variable} have been processed. This
+%   will normally take place within a conditional statement, for example
+%   \begin{verbatim}
+%     \tl_map_inline:Nn \l_my_tl
+%       {
+%         \str_if_eq:nnTF { #1 } { bingo }
+%           { \tl_map_break: }
+%           {
+%             % Do something useful
+%           }
+%       }
+%   \end{verbatim}
+%   Use outside of a \cs{tl_map_\ldots} scenario will lead low
+%   level \TeX{} errors.
+% \end{function}
+%
+% \section{Using token lists}
+%
+% \begin{function}[EXP]{\tl_to_str:N, \tl_to_str:c}
+%   \begin{syntax}
+%     \cs{tl_to_str:N} \meta{tl~var}
+%   \end{syntax}
+%   Converts the content of the  \meta{tl~var} into a series of characters
+%   with category code $12$ (other) with the exception of spaces, which
+%   retain category code $10$ (space). This \meta{string} is then left
+%   in the input stream.
+% \end{function}
+%
+% \begin{function}[EXP]{\tl_to_str:n}
+%   \begin{syntax}
+%     \cs{tl_to_str:n} \Arg{tokens}
+%   \end{syntax}
+%   Converts the given \meta{tokens} into a series of characters with
+%   category code $12$ (other) with the exception of spaces, which
+%   retain category code $10$ (space). This \meta{string} is then left
+%   in the input stream. Note that this function requires only a single
+%   expansion.
+%   \begin{texnote}
+%     This is the \eTeX{} primitive \cs{detokenize}.
+%   \end{texnote}
+% \end{function}
+%
+% \begin{function}[EXP]{\tl_use:N, \tl_use:c}
+%   \begin{syntax}
+%     \cs{tl_use:N} \meta{tl~var}
+%   \end{syntax}
+%   Recovers the content of a \meta{tl~var} and places it
+%   directly in the input stream. An error will be raised if the variable
+%   does not exist or if it is invalid. Note that it is possible to use
+%   a \meta{tl~var} directly without an accessor function.
+% \end{function}
+%
+% \section{Working with the content of token lists}
+%
+% \begin{function}[EXP]{\tl_length:n, \tl_length:V, \tl_length:o}
+%   \begin{syntax}
+%     \cs{tl_length:n} \Arg{tokens}
+%   \end{syntax}
+%   Counts the number of token groups in \meta{tokens} and leaves this
+%   information in the input stream. Unbraced tokens count as one
+%   element as do each token group (|{|\ldots|}|). This process will
+%   ignore any unprotected spaces within \meta{tokens}. See also
+%   \cs{tl_length:N}. This function requires three expansions,
+%   giving an \meta{integer denotation}.
+% \end{function}
+%
+% \begin{function}[EXP]{\tl_length:N, \tl_length:c}
+%   \begin{syntax}
+%     \cs{tl_length:N} \Arg{tl~var}
+%   \end{syntax}
+%   Counts the number of token groups in the \meta{tl~var}
+%   and leaves this information in the input stream. Unbraced tokens
+%   count as one element as do each token group (|{|\ldots|}|). This
+%   process will ignore any unprotected spaces within \meta{tokens}.
+%   See also \cs{tl_length:n}. This function requires three expansions,
+%   giving an \meta{integer denotation}.
+% \end{function}
+%
+% \begin{function}[EXP]{\tl_reverse:n, \tl_reverse:V, \tl_reverse:o}
+%   \begin{syntax}
+%     \cs{tl_reverse:n} \Arg{tokens}
+%   \end{syntax}
+%   Reverses the order of the \meta{tokens}, so that
+%   \meta{token1}\meta{token2}\meta{token3} \ldots \meta{token$_n$}
+%   becomes \meta{token$_n$}\ldots
+%   \meta{token3}\meta{token2}\meta{token1}. This process will remove
+%   any unprotected space within the \meta{tokens}. Tokens are not reversed
+%   within braced token groups, which lose their outer set of braces.
+%   See also \cs{tl_reverse:N}.
+% \end{function}
+%
+% \begin{function}{\tl_reverse:N, \tl_reverse:c}
+%   \begin{syntax}
+%     \cs{tl_reverse:N} \Arg{tl~var}
+%   \end{syntax}
+%   Reverses the order of the \meta{tokens} stored in
+%   \meta{tl~var}, so that
+%   \meta{token1}\meta{token2}\meta{token3} \ldots \meta{token$_n$}
+%   becomes \meta{token$_n$}\ldots
+%   \meta{token3}\meta{token2}\meta{token1}. This process will
+%   remove any unprotected spaces within the
+%   \meta{token list variable}. Braced token groups are copied without
+%   reversing the order of tokens, and the outer set of braces is removed.
+%   The reversal is local to the current
+%   \TeX{} group. See also \cs{tl_reverse:n}.
+% \end{function}
+%
+% \begin{function}[EXP]{\tl_trim_spaces:n}
+%   \begin{syntax}
+%     \cs{tl_trim_spaces:n} \meta{token list}
+%   \end{syntax}
+%   Removes any leading and trailing spaces from the \meta{token list}
+%   and leaves the result in the input stream. This process requires
+%   two expansions.
+% \end{function}
+%
+% \begin{function}{\tl_trim_spaces:N, \tl_trim_spaces:c}
+%   \begin{syntax}
+%     \cs{tl_trim_spaces:N} \meta{tl~var}
+%   \end{syntax}
+%   Removes any leading and trailing spaces from the content of the
+%   \meta{tl~var} within the current \TeX{} group.
+% \end{function}
+% 
+% \begin{function}{\tl_gtrim_spaces:N, \tl_gtrim_spaces:c}
+%   \begin{syntax}
+%     \cs{tl_gtrim_spaces:N} \meta{tl~var}
+%   \end{syntax}
+%   Removes any leading and trailing spaces from the content of the
+%   \meta{tl~var} globally.
+% \end{function}
+%
+% \section{The first token from a token list}
+%
+% Functions which deal with either onlt the very first token of a
+% token list or everything except the first token.
+%
+% \begin{function}[EXP]{\tl_head:n, \tl_head:V, \tl_head:v, \tl_head:f}
+%   \begin{syntax}
+%     \cs{tl_head:n} \Arg{tokens}
+%   \end{syntax}
+%   Leaves only the first \meta{token} in \meta{tokens} in the input
+%   stream, discarding the remainder.
+% \end{function}
+%
+% \begin{function}[EXP]{\tl_head:w}
+%   \begin{syntax}
+%     \cs{tl_head:w} \Arg{tokens} \cs{q_stop}
+%   \end{syntax}
+%   Leaves only the first \meta{token} in \meta{tokens} in the input
+%   stream, discarding the remainder. This function requires only a
+%   single expansion, and so is suitable for use inside an
+%   \texttt{o}-type expansion. In general \cs{tl_head:n} should be
+%   preferred.
+% \end{function}
+%
+% \begin{function}[EXP]{\tl_tail:n, \tl_tail:V, \tl_tail:v, \tl_tail:f}
+%   \begin{syntax}
+%     \cs{tl_tail:n} \Arg{tokens}
+%   \end{syntax}
+%   Discards the first \meta{token} of the \meta{tokens} and leaves the
+%   remainder in the input stream.
+% \end{function}
+%
+% \begin{function}[EXP]{\tl_tail:w}
+%   \begin{syntax}
+%     \cs{tl_tail:w} \Arg{tokens} \cs{q_stop}
+%   \end{syntax}
+%   Discards the first \meta{token} of the \meta{tokens} and leaves the
+%   remainder in the input stream. This function requires only a
+%   single expansion, and so is suitable for use inside an
+%   \texttt{o}-type expansion. In general \cs{tl_tail:n} should be
+%   preferred.
+% \end{function}
+%
+% \begin{function}[EXP,pTF]{\tl_if_head_eq_catcode:nN}
+%   \begin{syntax}
+%     \cs{tl_if_head_eq_catcode_p:n} \Arg{token list} \meta{test token}
+%     \cs{tl_if_head_eq_catcode:nTF} \Arg{token list} \meta{test token}
+%     ~~\Arg{true code} \Arg{false code}
+%   \end{syntax}
+%   Tests if the first \meta{token} in the \meta{token list} has the same
+%   category code as the \meta{test token}. The branching versions then
+%   leave either \meta{true code} or \meta{false code}  in the input
+%   stream, as appropriate to the truth of the test and the variant of
+%   the function chosen. The logical truth of the test is left in the
+%   input stream by the predicate
+%   version.
+% \end{function}
+%
+% \begin{function}[EXP,pTF]
+%   {\tl_if_head_eq_charcode:nN, \tl_if_head_eq_charcode:fN}
+%   \begin{syntax}
+%     \cs{tl_if_head_eq_charcode_p:n} \Arg{token list} \meta{test token}
+%     \cs{tl_if_head_eq_charcode:nTF} \Arg{token list} \meta{test token}
+%     ~~\Arg{true code} \Arg{false code}
+%   \end{syntax}
+%   Tests if the first \meta{token} in the \meta{token list} has the same
+%   character code as the \meta{test token}. The branching versions then
+%   leave either \meta{true code} or \meta{false code}  in the input
+%   stream, as appropriate to the truth of the test and the variant of
+%   the function chosen. The logical truth of the test is left in the
+%   input stream by the predicate
+%   version.
+% \end{function}
+%
+% \begin{function}[EXP,pTF]{\tl_if_head_eq_meaning:nN}
+%   \begin{syntax}
+%     \cs{tl_if_head_eq_meaning_p:n} \Arg{token list} \meta{test token}
+%     \cs{tl_if_head_eq_meaning:nTF} \Arg{token list} \meta{test token}
+%     ~~\Arg{true code} \Arg{false code}
+%   \end{syntax}
+%   Tests if the first \meta{token} in the \meta{token list} has the same
+%   meaning as the \meta{test token}. The branching versions then
+%   leave either \meta{true code} or \meta{false code}  in the input
+%   stream, as appropriate to the truth of the test and the variant of
+%   the function chosen. The logical truth of the test is left in the
+%   input stream by the predicate
+%   version.
+% \end{function}
+%
+% \section{Viewing token lists}
+%
+% \begin{function}{\tl_show:N, \tl_show:c}
+%   \begin{syntax}
+%     \cs{tl_show:N} \meta{tl~var}
+%   \end{syntax}
+%   Displays the content of the \meta{tl~var} on the terminal.
+%   \begin{texnote}
+%     \cs{tl_show:N} is the \TeX{} primitive \cs{show}.
+%   \end{texnote}
+% \end{function}
+%
+% \begin{function}{\tl_show:n}
+%   \begin{syntax}
+%     \cs{tl_show:n} \meta{token list}
+%   \end{syntax}
+%   Displays the \meta{token list} on the terminal.
+%   \begin{texnote}
+%     \cs{tl_show:n} is the \eTeX{} primitive \cs{showtokens}.
+%   \end{texnote}
+% \end{function}
+%
+% \section{Constant token lists}
+%
+% \begin{variable}{\c_job_name_tl}
+%   Constant that gets the \enquote{job name} assigned when \TeX{} starts.
+%   \begin{texnote}
+%     This is the new name for the primitive \cs{jobname}. It is a constant
+%     that is set by \TeX{} and should not be overwritten by the package.
+%   \end{texnote}
+% \end{variable}
+%
+% \begin{variable}{\c_empty_tl}
+%   Constant that is always empty.
+% \end{variable}
+%
+% \begin{variable}{\c_space_tl}
+%   A space token contained in a token list (compare this with
+%   \cs{c_space_token}). For use where an explicit space is required.
+% \end{variable}
+%
+% \section{Scratch token lists}
+%
+% \begin{variable}{\l_tmpa_tl, \l_tmpb_tl}
+%   Scratch token lists for local assignment. These are never used by
+%   the kernel code, and so are safe for use with any \LaTeX3-defined
+%   function. However, they may be overwritten by other non-kernel
+%   code and so should only be used for short-term storage.
+% \end{variable}
+%
+% \begin{variable}{\g_tmpa_tl, \g_tmpb_tl}
+%   Scratch token lists for global assignment. These are never used by
+%   the kernel code, and so are safe for use with any \LaTeX3-defined
+%   function. However, they may be overwritten by other non-kernel
+%   code and so should only be used for short-term storage.
+% \end{variable}
+%
+% \section{Experimental token list functions}
+%
+% \begin{function}[EXP,pTF]{\tl_if_single_item:n}
+%   \begin{syntax}
+%     \cs{tl_if_single_item_p:n} \Arg{token list}
+%     \cs{tl_if_single_item:nTF} \Arg{token list}
+%     ~~\Arg{true code} \Arg{false code}
+%   \end{syntax}
+%   Tests if the token list has exactly one item, i.e., is either
+%   a single normal token or a single brace group, surrounded by
+%   optional spaces on both sides. In other words, such a token list
+%   has length $1$ according to \cs{tl_length:n}. The branching
+%   versions then leave either \meta{true code} or \meta{false code}
+%   in the input stream, as appropriate to the truth of the test and
+%   the variant of the function chosen. The logical truth of the test
+%   is left in the input stream by the predicate.
+% \end{function}
+%
+% \begin{function}[EXP,pTF]{\tl_if_head_begin_group:n}
+%   \begin{syntax}
+%     \cs{tl_if_head_begin_group_p:n} \Arg{token list}
+%     \cs{tl_if_head_begin_group:nTF} \Arg{token list}
+%     ~~\Arg{true code} \Arg{false code}
+%   \end{syntax}
+%   Tests if the first \meta{token} in the \meta{token list}
+%   has catcode~1, i.e., is a begin-group character. This conditional
+%   is mainly intended to be used in combination with
+%   \cs{tl_if_head_eq_space:n} to check if grabbing an undelimited
+%   argument from the token list is safe.  The branching
+%   versions leave either \meta{true code} or \meta{false code}
+%   in the input stream, as appropriate to the truth of the test and
+%   the variant of the function chosen. The logical truth of the test
+%   is left in the input stream by the predicate.
+% \end{function}
+%
+% \begin{function}[EXP,pTF]{\tl_if_head_eq_space:n}
+%   \begin{syntax}
+%     \cs{tl_if_head_eq_space_p:n} \Arg{token list}
+%     \cs{tl_if_head_eq_space:nTF} \Arg{token list}
+%     ~~\Arg{true code} \Arg{false code}
+%   \end{syntax}
+%^^A Needs rewriting. We need to add a discussion of
+%^^A explicit vs implicit tokens somewhere in the doc.
+%   Tests if the first token of the token list is an explicit space,
+%   i.e., a token which matches \verb*+ + both in character code and
+%   in category code. In the case of an implicit space token, such as
+%   \cs{c_space_token}, the test will return \meta{false}. 
+%   This conditional is mainly intended to be used in combination with
+%   \cs{tl_if_head_begin_group:n} to test if grabbing an undelimited
+%   argument is safe (see \cs{tl_if_single_token:n} for an example.
+%   The branching versions leave either \meta{true code} or
+%   \meta{false code} in the input stream, as appropriate to the truth
+%   of the test and the variant of the function chosen. The logical
+%   truth of the test is left in the input stream by the predicate.
+%   \begin{texnote}
+%     When \TeX{}'s reads a character of category code $10$ for the
+%     first time, it is converted to an explicit space token, with
+%     character code $32$, regardless of the initial character code.
+%     ``Funny'' spaces with a different category code, can be produced
+%     using \cs{tex_lowercase:D}. Explicit spaces are also produced
+%     as a result of \cs{token_to_str:N}, \cs{tl_to_str:n}, etc.
+%   \end{texnote}
+% \end{function}
+%
+% \begin{function}[EXP,pTF]{\tl_if_single_token:n}
+%   \begin{syntax}
+%   \cs{tl_if_single_token_p:n} \Arg{token list}
+%   \cs{tl_if_single_token:nTF} \Arg{token list}
+%   ~~\Arg{true code} \Arg{false code}
+%   \end{syntax}
+%   Tests if the token list consists of exactly one token, i.e.,
+%   is either a single space character or a single ``normal'' token.
+%   Token groups are not single tokens. The branching
+%   versions leaves either \meta{true code} or \meta{false code}
+%   in the input stream, as appropriate to the truth of the test and
+%   the variant of the function chosen. The logical truth of the test
+%   is left in the input stream by the predicate.
+% \end{function}
+%
+% \end{documentation}
+%
+% \begin{implementation}
+%
+% \section{\pkg{l3tl} implementation}
+%
+%    \begin{macrocode}
+%<*initex|package>
+%    \end{macrocode}
+%
+%    \begin{macrocode}
+%<*package>
+\ProvidesExplPackage
+  {\ExplFileName}{\ExplFileDate}{\ExplFileVersion}{\ExplFileDescription}
+\package_check_loaded_expl:
+%</package>
+%    \end{macrocode}
+%
+% A token list variable is a \TeX{} macro that holds tokens. By using the
+% \eTeX{} primitive \cs{unexpanded} inside a \TeX{} \cs{edef} it is
+% possible to store any tokens, including |#|, in this way.
+%
+% \subsection{Functions}
+%
+% \begin{macro}{\tl_new:N, \tl_new:c}
+%   Creating new token list variables is a case of checking for an
+%   existing definition and if free doing the definition.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \tl_new:N #1
+  {
+    \chk_if_free_cs:N #1
+    \cs_gset_eq:NN #1 \c_empty_tl
+  }
+\cs_generate_variant:Nn \tl_new:N { c }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\tl_const:Nn, \tl_const:Nx, \tl_const:cn, \tl_const:cx}
+%   Constants are also easy to generate.
+%    \begin{macrocode}
+\cs_new_protected:Npn \tl_const:Nn #1#2
+  {
+    \chk_if_free_cs:N #1
+    \cs_gset_nopar:Npx #1 { \exp_not:n {#2} }
+  }
+\cs_new_protected:Npn \tl_const:Nx #1#2
+  {
+    \chk_if_free_cs:N #1
+    \cs_gset_nopar:Npx #1 {#2}
+  }
+\cs_generate_variant:Nn \tl_const:Nn { c }
+\cs_generate_variant:Nn \tl_const:Nx { c }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\tl_clear:N, \tl_clear:c}
+% \begin{macro}{\tl_gclear:N, \tl_gclear:c}
+%   Clearing a token list variable means setting it to an empty value.
+%   Error checking will be sorted out by the parent function.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \tl_clear:N  #1
+  { \tl_set_eq:NN #1 \c_empty_tl }
+\cs_new_protected_nopar:Npn \tl_gclear:N #1
+  { \tl_gset_eq:NN #1 \c_empty_tl }
+\cs_generate_variant:Nn \tl_clear:N  { c }
+\cs_generate_variant:Nn \tl_gclear:N { c }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\tl_clear_new:N, \tl_clear_new:c}
+% \begin{macro}{\tl_gclear_new:N, \tl_gclear_new:c}
+%   Clearing a token list variable means setting it to an empty value.
+%   Error checking will be sorted out by the parent function.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \tl_clear_new:N  #1
+  { \cs_if_exist:NTF #1 { \tl_clear:N #1 } { \tl_new:N #1 } }
+\cs_new_protected_nopar:Npn \tl_gclear_new:N #1
+  { \cs_if_exist:NTF #1 { \tl_gclear:N #1 } { \tl_new:N #1 } }
+\cs_generate_variant:Nn \tl_clear_new:N  { c }
+\cs_generate_variant:Nn \tl_gclear_new:N { c }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\tl_set_eq:NN, \tl_set_eq:Nc, \tl_set_eq:cN, \tl_set_eq:cc}
+% \begin{macro}{\tl_gset_eq:NN, \tl_gset_eq:Nc, \tl_gset_eq:cN, \tl_gset_eq:cc}
+%   For setting token list variables equal to each other.
+%    \begin{macrocode}
+\cs_new_eq:NN \tl_set_eq:NN  \cs_set_eq:NN
+\cs_new_eq:NN \tl_set_eq:cN  \cs_set_eq:cN
+\cs_new_eq:NN \tl_set_eq:Nc  \cs_set_eq:Nc
+\cs_new_eq:NN \tl_set_eq:cc  \cs_set_eq:cc
+\cs_new_eq:NN \tl_gset_eq:NN \cs_gset_eq:NN
+\cs_new_eq:NN \tl_gset_eq:cN \cs_gset_eq:cN
+\cs_new_eq:NN \tl_gset_eq:Nc \cs_gset_eq:Nc
+\cs_new_eq:NN \tl_gset_eq:cc \cs_gset_eq:cc
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \subsection{Adding to token list variables}
+%
+% \begin{macro}
+%   {
+%     \tl_set:Nn, \tl_set:NV, \tl_set:Nv, \tl_set:No, \tl_set:Nf, \tl_set:Nx,
+%     \tl_set:cn, \tl_set:NV, \tl_set:Nv, \tl_set:co, \tl_set:cf, \tl_set:cx
+%   }
+% \begin{macro}
+%   {
+%     \tl_gset:Nn, \tl_gset:NV, \tl_gset:Nv,
+%     \tl_gset:No, \tl_gset:Nf, \tl_gset:Nx,
+%     \tl_gset:cn, \tl_gset:NV, \tl_gset:Nv,
+%     \tl_gset:co, \tl_gset:cf, \tl_gset:cx
+%   }
+%   By using \cs{exp_not:n} token list variables can contain |#| tokens,
+%   which makes the token list registers provided by \TeX{}
+%   more or less redundant. The \cs{tl_set:No} version is done
+%   \enquote{by hand} as it is used quite a lot.
+%    \begin{macrocode}
+\cs_new_protected:Npn \tl_set:Nn #1#2
+  { \cs_set_nopar:Npx #1 { \exp_not:n {#2} } }
+\cs_new_protected:Npn \tl_set:No #1#2
+  { \cs_set_nopar:Npx #1 { \exp_not:o {#2} } }
+\cs_new_protected:Npn \tl_set:Nx #1#2
+  { \cs_set_nopar:Npx #1 {#2} }
+\cs_new_protected:Npn \tl_gset:Nn #1#2
+  { \cs_gset_nopar:Npx #1 { \exp_not:n {#2} } }
+\cs_new_protected:Npn \tl_gset:No #1#2
+  { \cs_gset_nopar:Npx #1 { \exp_not:o {#2} } }
+\cs_new_protected:Npn \tl_gset:Nx #1#2
+  { \cs_gset_nopar:Npx #1 {#2} }
+\cs_generate_variant:Nn \tl_set:Nn  {         NV , Nv , Nf }
+\cs_generate_variant:Nn \tl_set:Nx  { c }
+\cs_generate_variant:Nn \tl_set:Nn  { c, co , cV , cv , cf }
+\cs_generate_variant:Nn \tl_gset:Nn {         NV , Nv , Nf }
+\cs_generate_variant:Nn \tl_gset:Nx { c }
+\cs_generate_variant:Nn \tl_gset:Nn { c, co , cV , cv , cf }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}
+%   {
+%     \tl_put_left:Nn, \tl_put_left:NV, \tl_put_left:No, \tl_put_left:Nx,
+%     \tl_put_left:cn, \tl_put_left:cV, \tl_put_left:co, \tl_put_left:cx
+%   }
+% \begin{macro}
+%   {
+%     \tl_gput_left:Nn, \tl_gput_left:NV, \tl_gput_left:No, \tl_gput_left:Nx,
+%     \tl_gput_left:cn, \tl_gput_left:cV, \tl_gput_left:co, \tl_gput_left:cx
+%   }
+% Adding to the left is done directly to gain a little performance.
+%    \begin{macrocode}
+\cs_new_protected:Npn \tl_put_left:Nn #1#2
+  { \cs_set_nopar:Npx #1 { \exp_not:n {#2} \exp_not:o #1 } }
+\cs_new_protected:Npn \tl_put_left:NV #1#2
+  { \cs_set_nopar:Npx #1 { \exp_not:V #2 \exp_not:o #1 } }
+\cs_new_protected:Npn \tl_put_left:No #1#2
+  { \cs_set_nopar:Npx #1 { \exp_not:o {#2} \exp_not:o #1 } }
+\cs_new_protected:Npn \tl_put_left:Nx #1#2
+  { \cs_set_nopar:Npx #1 { #2 \exp_not:o #1 } }
+\cs_new_protected:Npn \tl_gput_left:Nn #1#2
+  { \cs_gset_nopar:Npx #1 { \exp_not:n {#2} \exp_not:o #1 } }
+\cs_new_protected:Npn \tl_gput_left:NV #1#2
+  { \cs_gset_nopar:Npx #1 { \exp_not:V #2 \exp_not:o #1 } }
+\cs_new_protected:Npn \tl_gput_left:No #1#2
+  { \cs_gset_nopar:Npx #1 { \exp_not:o {#2} \exp_not:o #1 } }
+\cs_new_protected:Npn \tl_gput_left:Nx #1#2
+  { \cs_gset_nopar:Npx #1 { #2 \exp_not:o {#1} } }
+\cs_generate_variant:Nn \tl_put_left:Nn  { c }
+\cs_generate_variant:Nn \tl_put_left:NV  { c }
+\cs_generate_variant:Nn \tl_put_left:No  { c }
+\cs_generate_variant:Nn \tl_put_left:Nx  { c }
+\cs_generate_variant:Nn \tl_gput_left:Nn { c }
+\cs_generate_variant:Nn \tl_gput_left:NV { c }
+\cs_generate_variant:Nn \tl_gput_left:No { c }
+\cs_generate_variant:Nn \tl_gput_left:Nx { c }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}
+%   {
+%     \tl_put_right:Nn, \tl_put_right:NV, \tl_put_right:No, \tl_put_right:Nx,
+%     \tl_put_right:cn, \tl_put_right:cV, \tl_put_right:co, \tl_put_right:cx
+%   }
+% \begin{macro}
+%   {
+%     \tl_gput_right:Nn, \tl_gput_right:NV, \tl_gput_right:No,
+%     \tl_gput_right:Nx,
+%     \tl_gput_right:cn, \tl_gput_right:cV, \tl_gput_right:co,
+%     \tl_gput_right:cx
+%   }
+% The same on the right.
+%    \begin{macrocode}
+\cs_new_protected:Npn \tl_put_right:Nn #1#2
+  { \cs_set_nopar:Npx #1 { \exp_not:o #1 \exp_not:n {#2} } }
+\cs_new_protected:Npn \tl_put_right:NV #1#2
+  { \cs_set_nopar:Npx #1 { \exp_not:o #1 \exp_not:V #2 } }
+\cs_new_protected:Npn \tl_put_right:No #1#2
+  { \cs_set_nopar:Npx #1 { \exp_not:o #1 \exp_not:o {#2} } }
+\cs_new_protected:Npn \tl_put_right:Nx #1#2
+  { \cs_set_nopar:Npx #1 { \exp_not:o #1 #2 } }
+\cs_new_protected:Npn \tl_gput_right:Nn #1#2
+  { \cs_gset_nopar:Npx #1 { \exp_not:o #1 \exp_not:n {#2} } }
+\cs_new_protected:Npn \tl_gput_right:NV #1#2
+  { \cs_gset_nopar:Npx #1 { \exp_not:o #1 \exp_not:V #2 } }
+\cs_new_protected:Npn \tl_gput_right:No #1#2
+  { \cs_gset_nopar:Npx #1 { \exp_not:o #1 \exp_not:o {#2} } }
+\cs_new_protected:Npn \tl_gput_right:Nx #1#2
+  { \cs_gset_nopar:Npx #1 { \exp_not:o {#1} #2 } }
+\cs_generate_variant:Nn \tl_put_right:Nn  { c }
+\cs_generate_variant:Nn \tl_put_right:NV  { c }
+\cs_generate_variant:Nn \tl_put_right:No  { c }
+\cs_generate_variant:Nn \tl_put_right:Nx  { c }
+\cs_generate_variant:Nn \tl_gput_right:Nn { c }
+\cs_generate_variant:Nn \tl_gput_right:NV { c }
+\cs_generate_variant:Nn \tl_gput_right:No { c }
+\cs_generate_variant:Nn \tl_gput_right:Nx { c }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \subsection{Reassigning token list category codes}
+%
+% \begin{variable}{\c_tl_rescan_marker_tl}
+%   The rescanning code needs a special token list containing the same
+%   character with two different category codes. This is set up here,
+%   while the detail is described below.
+%    \begin{macrocode}
+\group_begin:
+  \tex_lccode:D `\A = `\@ \scan_stop:
+  \tex_lccode:D `\B = `\@ \scan_stop:
+  \tex_catcode:D `\A = 8 \scan_stop:
+  \tex_catcode:D `\B = 3 \scan_stop:
+\tex_lowercase:D
+  {
+    \group_end:
+    \tl_const:Nn \c_tl_rescan_marker_tl { A B }
+  }
+%    \end{macrocode}
+% \end{variable}
+%
+% \begin{variable}{\l_tl_rescan_tl}
+%   A token list variable to actually store the material being processed.
+%    \begin{macrocode}
+\tl_new:N \l_tl_rescan_tl
+%    \end{macrocode}
+% \end{variable}
+%
+% \begin{macro}
+%   {
+%     \tl_set_rescan:Nnn, \tl_set_rescan:Nno,
+%     \tl_set_rescan:cnn, \tl_set_rescan:cno
+%   }
+% \begin{macro}
+%   {
+%     \tl_gset_rescan:Nnn, \tl_gset_rescan:Nno,
+%     \tl_gset_rescan:cnn, \tl_gset_rescan:cno
+%   }
+% \begin{macro}[aux]{\tl_set_rescan_aux:NNnn}
+% \begin{macro}[aux]{\tl_rescan_aux:w}
+%   The idea here is to deal cleanly with the problem that
+%   \cs{tex_scantokens:D} treats the argument as a file, and without
+%   the correct settings a \TeX{} error occurs:
+%   \begin{verbatim}
+%    ! File ended while scanning definition of ...
+%   \end{verbatim}
+%   When expanding a token list this can be handled using \cs{exp_not:N}
+%   but this fails if the token list is not being expanded. So instead
+%   a delimited argument is used with an end marker which cannot appear
+%   within the token list which is scanned: two |@| symbols with different
+%   category codes. The rescanned token list cannot contain the end marker,
+%   because all |@| present in the token list are read with the same category
+%   code. The setting of \cs{tex_endlinechar:D}
+%   is needed to avoid introducing an extraneous space in the result.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \tl_set_rescan:Nnn
+  { \tl_set_rescan_aux:NNnn \tl_set:Nn }
+\cs_new_protected_nopar:Npn \tl_gset_rescan:Nnn
+  { \tl_set_rescan_aux:NNnn \tl_gset:Nn }
+\cs_new_protected:Npn \tl_set_rescan_aux:NNnn #1#2#3#4
+  {
+    \group_begin:
+      \exp_args:No \etex_everyeof:D { \c_tl_rescan_marker_tl }
+      \tex_endlinechar:D \c_minus_one
+      #3
+      \tl_clear:N \l_tl_rescan_tl
+      \exp_after:wN \tl_rescan_aux:w \etex_scantokens:D {#4}
+    \exp_args:NNNo \group_end:
+    #1 #2 \l_tl_rescan_tl
+  }
+\cs_new_nopar:Npx \tl_rescan_aux:w
+  {
+    \cs_set_protected:Npn \exp_not:N \tl_rescan_aux:w ##1
+      \c_tl_rescan_marker_tl
+      { \tl_set:Nn \exp_not:N \l_tl_rescan_tl {##1} }
+  }
+\tl_rescan_aux:w
+\cs_generate_variant:Nn \tl_set_rescan:Nnn  {     Nno }
+\cs_generate_variant:Nn \tl_set_rescan:Nnn  { c , cno }
+\cs_generate_variant:Nn \tl_gset_rescan:Nnn {     Nno }
+\cs_generate_variant:Nn \tl_gset_rescan:Nnn { c , cno }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\tl_set_rescan:Nnx, \tl_set_rescan:cnx}
+% \begin{macro}{\tl_gset_rescan:Nnx, \tl_gset_rescan:cnx}
+% \begin{macro}[aux]{\tl_set_rescan_aux:NNnx}
+%   With \texttt{x}-type expansion the \cs{tex_everyoef:D} method
+%   does apply and the code is simple.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \tl_set_rescan:Nnx
+  { \tl_set_rescan_aux:NNnx \tl_set:Nn }
+\cs_new_protected_nopar:Npn \tl_gset_rescan:Nnx
+  { \tl_set_rescan_aux:NNnx \tl_gset:Nn }
+\cs_new_protected_nopar:Npn \tl_set_rescan_aux:NNnx #1#2#3#4
+  {
+    \group_begin:
+      \etex_everyeof:D { \exp_not:N }
+      \tex_endlinechar:D \c_minus_one
+      #3
+      \tl_set:Nx \l_tl_rescan_tl { \etex_scantokens:D {#4} }
+    \exp_args:NNNo \group_end:
+    #1 #2 \l_tl_rescan_tl
+  }
+\cs_generate_variant:Nn \tl_set_rescan:Nnx  { c }
+\cs_generate_variant:Nn \tl_gset_rescan:Nnx { c }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\tl_rescan:nn}
+%   The same idea is also applied to in line token lists.
+%    \begin{macrocode}
+\cs_new_protected:Npn \tl_rescan:nn #1#2
+  {
+    \group_begin:
+      \exp_args:No \etex_everyeof:D { \c_tl_rescan_marker_tl }
+      \tex_endlinechar:D \c_minus_one
+      #1
+      \exp_after:wN \tl_rescan_aux:w \etex_scantokens:D {#2}
+    \exp_args:No \group_end:
+    \l_tl_rescan_tl
+  }
+%    \end{macrocode}
+% \end{macro}
+%
+% \subsection{Reassigning token list character codes}
+%
+% \begin{macro}{\tl_to_lowercase:n}
+% \begin{macro}{\tl_to_uppercase:n}
+%   Just some names for a few primitives.
+%    \begin{macrocode}
+\cs_new_eq:NN \tl_to_lowercase:n \tex_lowercase:D
+\cs_new_eq:NN \tl_to_uppercase:n \tex_uppercase:D
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \subsection{Modifying token list variables}
+%
+% \begin{variable}{\l_tl_replace_tl}
+%   A scratch variable for doing token replacement.
+%    \begin{macrocode}
+\tl_new:N \l_tl_replace_tl
+%    \end{macrocode}
+% \end{variable}
+%
+% \begin{macro}{\tl_replace_once:Nnn, \tl_replace_once:cnn}
+% \begin{macro}{\tl_greplace_once:Nnn, \tl_greplace_once:cnn}
+% \begin{macro}[aux]{\tl_replace_once_aux:NNnn}
+%   The concept here is that only the first occurrence should be
+%   replaced. The first step is to define an auxiliary which will
+%   match the appropriate item, with a trailing marker. If the last token
+%   is the marker there is nothing to do, otherwise replace the token
+%   and clean up (hence the second use of \cs{tl_tmp:w}). To prevent
+%   loosing braces or spaces there are a couple of empty groups and
+%   the strange-looking \cs{use:n}. There is a \cs{q_nil} between the
+%   original and the search input, to prevent cases where
+%   the end of the original and the start of the search run together
+%   to give an erroneous test result.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \tl_replace_once:Nnn
+  { \tl_replace_once_aux:NNnn \tl_set_eq:NN }
+\cs_new_protected_nopar:Npn \tl_greplace_once:Nnn
+  { \tl_replace_once_aux:NNnn \tl_gset_eq:NN }
+\cs_new_protected:Npn \tl_replace_once_aux:NNnn #1#2#3#4
+  {
+    \cs_set_protected:Npx \tl_tmp:w ##1 #3 ##2 \q_stop
+      {
+        \exp_not:N \quark_if_no_value:nF {##2}
+          {
+            \tl_set:No \exp_not:N \l_tl_replace_tl { ##1 \exp_not:n{#4} }
+            \exp_not:n
+              {
+                \cs_set_protected:Npn \tl_tmp:w ##1 \q_nil #3 \q_no_value
+                  { \tl_put_right:No \l_tl_replace_tl {##1} }
+              }
+            \exp_not:n { \tl_tmp:w \prg_do_nothing: } ##2
+            \exp_not:n { #1 #2 \l_tl_replace_tl }
+          }
+      }
+    \exp_after:wN \tl_tmp:w \exp_after:wN \prg_do_nothing:
+    #2 \q_nil #3 \q_no_value \q_stop
+  }
+\cs_generate_variant:Nn \tl_replace_once:Nnn  { c }
+\cs_generate_variant:Nn \tl_greplace_once:Nnn { c }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\tl_replace_all:Nnn, \tl_replace_all:cnn}
+% \begin{macro}{\tl_greplace_all:Nnn, \tl_greplace_all:cnn}
+% \begin{macro}[aux]{\tl_replace_all_aux:NNnn}
+%   A similar approach here but with a loop built in.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \tl_replace_all:Nnn
+  { \tl_replace_all_aux:NNnn \tl_set_eq:NN }
+\cs_new_protected_nopar:Npn \tl_greplace_all:Nnn
+  { \tl_replace_all_aux:NNnn \tl_gset_eq:NN }
+\cs_new_protected:Npn \tl_replace_all_aux:NNnn #1#2#3#4
+   {
+     \tl_clear:N \l_tl_replace_tl
+     \cs_set_protected:Npx \tl_tmp:w ##1 #3 ##2 \q_stop
+       {
+         \exp_not:N \quark_if_no_value:nTF {##2}
+           {
+             \exp_not:n
+               {
+                 \cs_set_protected:Npn \tl_tmp:w ##1 \q_nil ##2 \q_stop
+                   { \tl_put_right:No \l_tl_replace_tl {##1} }
+               }
+             \exp_not:N \tl_tmp:w ##1 \exp_not:N \q_stop
+           }
+           {
+             \exp_not:n { \tl_put_right:No \l_tl_replace_tl }
+               { ##1 \exp_not:n{#4} }
+             \exp_not:n { \tl_tmp:w \prg_do_nothing: } ##2 \exp_not:N \q_stop
+           }
+       }
+     \exp_after:wN \tl_tmp:w \exp_after:wN \prg_do_nothing:
+     #2 \q_nil #3 \q_no_value \q_stop
+     #1 #2 \l_tl_replace_tl
+   }
+\cs_generate_variant:Nn \tl_replace_all:Nnn  { c }
+\cs_generate_variant:Nn \tl_greplace_all:Nnn { c }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\tl_remove_once:Nn, \tl_remove_once:cn}
+% \begin{macro}{\tl_gremove_once:Nn, \tl_gremove_once:cn}
+%   Removal is just a special case of replacement.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \tl_remove_once:Nn #1#2
+  { \tl_replace_once:Nnn #1 {#2} { } }
+\cs_new_protected_nopar:Npn \tl_gremove_once:Nn #1#2
+  { \tl_greplace_once:Nnn #1 {#2} { } }
+\cs_generate_variant:Nn \tl_remove_once:Nn  { c }
+\cs_generate_variant:Nn \tl_gremove_once:Nn { c }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\tl_remove_all:Nn, \tl_remove_all:cn}
+% \begin{macro}{\tl_gremove_all:Nn, \tl_gremove_all:cn}
+%   Removal is just a special case of replacement.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \tl_remove_all:Nn #1#2
+  { \tl_replace_all:Nnn #1 {#2} { } }
+\cs_new_protected_nopar:Npn \tl_gremove_all:Nn #1#2
+  { \tl_greplace_all:Nnn #1 {#2} { } }
+\cs_generate_variant:Nn \tl_remove_all:Nn  { c }
+\cs_generate_variant:Nn \tl_gremove_all:Nn { c }
+%    \end{macrocode}
+% \end{macro}
+%
+% \subsection{Token list conditionals}
+%
+% \begin{macro}[pTF]{\tl_if_blank:n,\tl_if_blank:V,\tl_if_blank:o}
+% \begin{macro}[aux]{\tl_if_blank_p_aux:NNw}
+%   \TeX{} skips spaces when reading a non-delimited arguments. Thus,
+%   a \meta{token list} is blank if and only if \cs{use_none:n}
+%   \meta{token list} |?| is empty. For performance reasons, we hard-code
+%   the emptyness test done in \cs{tl_if_empty:n(TF)}: convert to harmless
+%   characters with \cs{tl_to_str:n}, and then use
+%   \cs{if_meaning:w} \cs{q_nil} |...| \cs{q_nil}.
+%   Note that converting to a string is done after reading the delimited
+%   argument for \cs{use_none:n}. The similar construction
+%   \cs{exp_after:wN} \cs{use_none:n} \cs{tl_to_str:n} \Arg{token list} |?|
+%   would fail if the token list contains the control sequence \cs{ },
+%   while \cs{tex_escapechar:D} is a space or is unprintable.
+%    \begin{macrocode}
+\prg_new_conditional:Npnn \tl_if_blank:n #1 { p , T , F , TF }
+  { \tl_if_empty_return:o { \use_none:n #1 ? } }
+\cs_generate_variant:Nn \tl_if_blank_p:n { V }
+\cs_generate_variant:Nn \tl_if_blank:nT  { V }
+\cs_generate_variant:Nn \tl_if_blank:nF  { V }
+\cs_generate_variant:Nn \tl_if_blank:nTF { V }
+\cs_generate_variant:Nn \tl_if_blank_p:n { o }
+\cs_generate_variant:Nn \tl_if_blank:nT  { o }
+\cs_generate_variant:Nn \tl_if_blank:nF  { o }
+\cs_generate_variant:Nn \tl_if_blank:nTF { o }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}[pTF]{\tl_if_empty:N,\tl_if_empty:c}
+%    These functions check whether the token list in the argument is
+%    empty and execute the proper code from their argument(s).
+%    \begin{macrocode}
+\prg_set_conditional:Npnn \tl_if_empty:N #1 { p , T , F , TF }
+  {
+    \if_meaning:w #1 \c_empty_tl
+      \prg_return_true:
+    \else:
+      \prg_return_false:
+    \fi:
+  }
+\cs_generate_variant:Nn \tl_if_empty_p:N { c }
+\cs_generate_variant:Nn \tl_if_empty:NT  { c }
+\cs_generate_variant:Nn \tl_if_empty:NF  { c }
+\cs_generate_variant:Nn \tl_if_empty:NTF { c }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}[pTF]{\tl_if_empty:n,\tl_if_empty:V}
+%   It would be tempting to just use |\if_meaning:w \q_nil #1 \q_nil| as
+%   a test since this works really well. However, it fails on a token
+%   list starting with |\q_nil| of course but more troubling is the
+%   case where argument is a complete conditional such as |\if_true:|
+%   a |\else:| b |\fi:| because then |\if_true:| is used by
+%   |\if_meaning:w|, the test turns out false, the |\else:| executes
+%   the false branch, the |\fi:| ends it and the |\q_nil| at the end
+%   starts executing\dots{} A safer route is to convert the entire
+%   token list into harmless characters first and then compare
+%   that. This way the test will even accept |\q_nil| as the first
+%   token.
+%    \begin{macrocode}
+\prg_new_conditional:Npnn \tl_if_empty:n #1 { p,TF,T,F} {
+  \exp_after:wN \if_meaning:w \exp_after:wN \q_nil \tl_to_str:n {#1} \q_nil
+    \prg_return_true:
+  \else:
+    \prg_return_false:
+  \fi:
+}
+\cs_generate_variant:Nn \tl_if_empty_p:n { V }
+\cs_generate_variant:Nn \tl_if_empty:nTF { V }
+\cs_generate_variant:Nn \tl_if_empty:nT  { V }
+\cs_generate_variant:Nn \tl_if_empty:nF  { V }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}[pTF]{\tl_if_empty:o}
+% \begin{macro}[EXP,aux]{\tl_if_empty_return:o}
+%   The auxiliary function \cs{tl_if_empty_return:o} is for use
+%   in conditionals on token lists, which mostly reduce to testing
+%   if a given token list is empty after applying a simple function
+%   to it.
+%   The test for emptyness is based on \cs{tl_if_empty:n(TF)}, but
+%   the expansion is hard-coded for efficiency, as this auxiliary
+%   function is used in many places.
+%   Note that this works because \cs{tl_to_str:n} expands tokens
+%   that follow until reading a catcode $1$ (begin-group) token.
+%    \begin{macrocode}
+\cs_new:Npn \tl_if_empty_return:o #1 {
+  \exp_after:wN \if_meaning:w \exp_after:wN \q_nil
+          \tl_to_str:n \exp_after:wN {#1} \q_nil
+    \prg_return_true:
+  \else:
+    \prg_return_false:
+  \fi:
+}
+\prg_new_conditional:Npnn \tl_if_empty:o #1 { p,TF,T,F}
+  { \tl_if_empty_return:o {#1} }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}[pTF]{\tl_if_eq:NN,\tl_if_eq:Nc,\tl_if_eq:cN,\tl_if_eq:cc}
+%   Returns \cs{c_true_bool} if and only if the two token list variables are
+%   equal.
+%    \begin{macrocode}
+\prg_new_conditional:Npnn \tl_if_eq:NN #1#2 { p , T , F , TF }
+  {
+    \if_meaning:w #1 #2
+      \prg_return_true:
+    \else:
+      \prg_return_false:
+    \fi:
+  }
+\cs_generate_variant:Nn \tl_if_eq_p:NN { Nc , c , cc }
+\cs_generate_variant:Nn \tl_if_eq:NNTF { Nc , c , cc }
+\cs_generate_variant:Nn \tl_if_eq:NNT  { Nc , c , cc }
+\cs_generate_variant:Nn \tl_if_eq:NNF  { Nc , c , cc }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}[TF]{\tl_if_eq:nn}
+% \begin{variable}{\l_tl_tmpa_tl, \l_tl_tmpb_tl}
+%   A simple store and compare routine.
+%    \begin{macrocode}
+\prg_new_protected_conditional:Npnn \tl_if_eq:nn #1#2 { T , F ,  TF }
+  {
+    \group_begin:
+      \tl_set:Nn \l_tl_tmpa_tl {#1}
+      \tl_set:Nn \l_tl_tmpb_tl {#2}
+      \if_meaning:w \l_tl_tmpa_tl \l_tl_tmpb_tl
+        \group_end:
+        \prg_return_true:
+      \else:
+        \group_end:
+        \prg_return_false:
+      \fi:
+  }
+\tl_new:N \l_tl_tmpa_tl
+\tl_new:N \l_tl_tmpb_tl
+%    \end{macrocode}
+% \end{variable}
+% \end{macro}
+%
+% \begin{macro}[TF]{\tl_if_in:Nn, \tl_if_in:cn}
+%   See \cs{tl_if_in:nn(TF)} for further comments. Here we simply
+%   expand the token list variable and pass it to \cs{tl_if_in:nn(TF)}.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \tl_if_in:NnT  { \exp_args:No \tl_if_in:nnT  }
+\cs_new_protected_nopar:Npn \tl_if_in:NnF  { \exp_args:No \tl_if_in:nnF  }
+\cs_new_protected_nopar:Npn \tl_if_in:NnTF { \exp_args:No \tl_if_in:nnTF }
+\cs_generate_variant:Nn \tl_if_in:NnT { c }
+\cs_generate_variant:Nn \tl_if_in:NnF  { c }
+\cs_generate_variant:Nn \tl_if_in:NnTF { c }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}[TF]{\tl_if_in:nn, \tl_if_in:Vn, \tl_if_in:on, \tl_if_in:no}
+%   Once more, the test relies on \cs{tl_to_str:n} for robustness.
+%   The function \cs{tl_tmp:w} removes tokens until the first occurrence
+%   of |#2|. If this does not appear in |#1|, then the final |#2| is removed,
+%   leaving an empty token list. Otherwise some tokens remain, and the
+%   test is false. See \cs{tl_if_empty:n(TF)} for details on
+%   the emptyness test.
+%
+%   Special care is needed to treat correctly cases like
+%   |\tl_if_in:nnTF {a state}{states}|, where |#1#2| contains |#2| before
+%   the end. To cater for this case, we insert |{}{}| between the two token
+%   lists. This marker may not appear in |#2| because of \TeX{} limitations
+%   on what can delimit a parameter, hence we are safe. Using two brace
+%   groups makes the test work also for empty arguments.
+%    \begin{macrocode}
+\prg_new_protected_conditional:Npnn \tl_if_in:nn #1#2 { T  , F , TF }
+  {
+    \cs_set:Npn \tl_tmp:w ##1 #2 { }
+    \tl_if_empty:oTF { \tl_tmp:w #1 {} {} #2 }
+      { \prg_return_false: } { \prg_return_true: }
+  }
+\cs_generate_variant:Nn \tl_if_in:nnT  { V }
+\cs_generate_variant:Nn \tl_if_in:nnF  { V }
+\cs_generate_variant:Nn \tl_if_in:nnTF { V }
+\cs_generate_variant:Nn \tl_if_in:nnT  { o }
+\cs_generate_variant:Nn \tl_if_in:nnF  { o }
+\cs_generate_variant:Nn \tl_if_in:nnTF { o }
+\cs_generate_variant:Nn \tl_if_in:nnT  { no }
+\cs_generate_variant:Nn \tl_if_in:nnF  { no }
+\cs_generate_variant:Nn \tl_if_in:nnTF { no }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}[pTF]{\tl_if_single:n}
+% \begin{macro}[aux]{\tl_if_single_aux:n}
+%   If the argument is a single token, or a single brace group
+%   preceeded by optional (explicit) spaces, or a non-zero number
+%   of spaces.
+%    \begin{macrocode}
+\prg_new_conditional:Npnn \tl_if_single:n #1 { p , T , F , TF }
+  {
+    \tl_if_blank:nTF {#1}
+      { \tl_if_empty:nTF {#1} { \prg_return_false: } { \prg_return_true: } }
+      { \tl_if_single_aux:n {#1} }
+  }
+\prg_new_conditional:Npnn \tl_if_single:N #1 { p , T , F , TF }
+  {
+    \tl_if_blank:oTF #1
+      { \tl_if_empty:NTF #1 { \prg_return_false: } { \prg_return_true: } }
+      { \exp_args:No \tl_if_single_aux:n {#1} }
+  }
+\cs_new:Npn \tl_if_single_aux:n #1
+  { \tl_if_empty_return:o { \use_none:n #1 } }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \subsection{Mapping to token lists}
+%
+% \begin{macro}{\tl_map_function:nN}
+% \begin{macro}{\tl_map_function:NN, \tl_map_function:cN}
+% \begin{macro}[aux]{\tl_map_function_aux:NN}
+%   Expandable loop macro for token lists. These have the advantage of not
+%   needing to test if the argument is empty, because if it is, the stop
+%   marker will be read immediately and the loop terminated.
+%    \begin{macrocode}
+\cs_new:Npn \tl_map_function:nN #1#2
+  { \tl_map_function_aux:Nn #2 #1 \q_recursion_tail \q_recursion_stop }
+\cs_new_nopar:Npn \tl_map_function:NN #1#2
+  {
+    \exp_after:wN \tl_map_function_aux:Nn
+    \exp_after:wN #2 #1 \q_recursion_tail \q_recursion_stop
+  }
+\cs_new:Npn \tl_map_function_aux:Nn #1#2
+  {
+    \quark_if_recursion_tail_stop:n {#2}
+    #1 {#2} \tl_map_function_aux:Nn #1
+  }
+\cs_generate_variant:Nn \tl_map_function:NN { c }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\tl_map_inline:nn}
+% \begin{macro}{\tl_map_inline:Nn, \tl_map_inline:cn}
+% \begin{macro}[aux]{\tl_map_inline_aux:n}
+% \begin{variable}{\g_tl_inline_level_int}
+%   The inline functions are straight forward by now. We use a little
+%   trick with the counter \cs{g_tl_inline_level_int} to make
+%   them nestable. We can also make use of \cs{tl_map_function:Nn}
+%   from before. (\cs{g_tl_inline_level_int} is defined in \pkg{l3int}
+%   for order-of-loading reasons.)
+%    \begin{macrocode}
+\cs_new_protected:Npn \tl_map_inline:nn #1#2
+  {
+    \int_gincr:N \g_tl_inline_level_int
+    \cs_gset:cpn { tl_map_inline_ \int_use:N \g_tl_inline_level_int :n }
+      ##1 {#2}
+    \exp_args:Nc \tl_map_function_aux:Nn
+      { tl_map_inline_ \int_use:N \g_tl_inline_level_int :n }
+      #1 \q_recursion_tail \q_recursion_stop
+    \int_gdecr:N \g_tl_inline_level_int
+  }
+\cs_new_protected:Npn \tl_map_inline:Nn #1#2
+  {
+    \int_gincr:N \g_tl_inline_level_int
+    \cs_gset:cpn { tl_map_inline_ \int_use:N \g_tl_inline_level_int :n }
+      ##1 {#2}
+    \exp_last_unbraced:NcV \tl_map_function_aux:Nn
+      { tl_map_inline_ \int_use:N \g_tl_inline_level_int :n }
+      #1 \q_recursion_tail\q_recursion_stop
+    \int_gdecr:N \g_tl_inline_level_int
+  }
+\cs_generate_variant:Nn \tl_map_inline:Nn { c }
+%    \end{macrocode}
+% \end{variable}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\tl_map_variable:nNn}
+% \begin{macro}{\tl_map_variable:NNn, \tl_map_variable:cNn}
+% \begin{macro}[aux]{\tl_map_variable_aux:NnN}
+%   \cs{tl_map_variable:nNn} \meta{token list} \meta{temp} \meta{action}
+%   assigns
+%   \meta{temp} to each element and executes \meta{action}.
+%    \begin{macrocode}
+\cs_new_protected:Npn \tl_map_variable:nNn #1#2#3
+  { \tl_map_variable_aux:Nnn #2 {#3} #1 \q_recursion_tail \q_recursion_stop }
+\cs_new_protected_nopar:Npn \tl_map_variable:NNn
+  { \exp_args:No \tl_map_variable:nNn }
+\cs_new_protected:Npn \tl_map_variable_aux:Nnn #1#2#3
+  {
+    \tl_set:Nn #1 {#3}
+    \quark_if_recursion_tail_stop:N #1
+    #2 \tl_map_variable_aux:Nnn #1 {#2}
+  }
+\cs_generate_variant:Nn \tl_map_variable:NNn { c }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\tl_map_break:}
+%   The break statement.
+%    \begin{macrocode}
+\cs_new_eq:NN \tl_map_break: \use_none_delimit_by_q_recursion_stop:w
+%    \end{macrocode}
+% \end{macro}
+%
+% \subsection{Using token lists}
+%
+% \begin{macro}{\tl_to_str:n}
+%   Another name for a primitive.
+%    \begin{macrocode}
+\cs_new_eq:NN \tl_to_str:n \etex_detokenize:D
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\tl_to_str:N, \tl_to_str:c}
+%    These functions return the replacement text of a token list as a
+%    string.
+%    \begin{macrocode}
+\cs_new_nopar:Npn \tl_to_str:N #1 { \etex_detokenize:D \exp_after:wN {#1} }
+\cs_generate_variant:Nn \tl_to_str:N { c }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\tl_use:N, \tl_use:c}
+% \begin{macro}[aux]{\tl_error_message:}
+% Token lists which are simply not defined will give a clear \TeX{}
+% error here. No such luck for ones equal to \cs{scan_stop:} so
+% instead a test is made and if there is an issue an error is forced.
+%    \begin{macrocode}
+\cs_new_eq:NN \tl_use:N \prg_do_nothing:
+\cs_new_nopar:Npn \tl_use:c #1
+  {
+    \if_cs_exist:w #1 \cs_end:
+      \cs:w #1 \exp_after:wN \cs_end:
+    \else:
+      \exp_after:wN \tl_error_message:
+    \fi:
+  }
+\group_begin:
+\tex_catcode:D `\! = 11 \scan_stop:
+\tex_catcode:D `\ = 11\scan_stop:%
+\cs_gset_nopar:Npn\tl_error_message:{\undefined variable name!}%
+\group_end:%
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \subsection{Working with the contents of token lists}
+%
+% \begin{macro}{\tl_length:n, \tl_length:V, \tl_length:o}
+% \begin{macro}{\tl_length:N, \tl_length:c}
+% \begin{macro}[aux]{\tl_length_aux:n}
+%   Count number of elements within a token list or token list
+%   variable. Brace groups within the list are read as a single
+%   element. Spaces are ignored.
+%   \cs{tl_length_aux:n} grabs the element and replaces it by |+1|.
+%   The |0| to ensure it works on an empty list.
+%    \begin{macrocode}
+\cs_new:Npn \tl_length:n #1
+  {
+    \int_eval:n
+      { 0 \tl_map_function:nN {#1} \tl_length_aux:n }
+  }
+\cs_new_nopar:Npn \tl_length:N #1
+  {
+    \int_eval:n
+      { 0 \tl_map_function:NN #1 \tl_length_aux:n }
+  }
+\cs_new:Npn \tl_length_aux:n #1 { + 1 }
+\cs_generate_variant:Nn \tl_length:n { V , o }
+\cs_generate_variant:Nn \tl_length:N { c }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\tl_reverse:n, \tl_reverse:V, \tl_reverse:o}
+% \begin{macro}[aux]{\tl_reverse_aux:nN}
+%    Reversal of a token list is done by taking one token at a time
+%    and putting it in front of the ones before it.
+%    \begin{macrocode}
+\cs_new:Npn \tl_reverse:n #1
+  { \tl_reverse_aux:nN { } #1 \q_recursion_tail \q_recursion_stop }
+\cs_new:Npn \tl_reverse_aux:nN #1#2
+  {
+    \quark_if_recursion_tail_stop_do:nn {#2} {#1}
+    \tl_reverse_aux:nN { #2 #1 }
+  }
+\cs_generate_variant:Nn \tl_reverse:n {V,o}
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+%  \begin{macro}{\tl_reverse:N, \tl_reverse:c}
+% This reverses the list, leaving |\exp_stop_f:| in front, which in turn
+% is removed by the |f| expansion which comes to a halt.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \tl_reverse:N #1
+  { \tl_set:Nf #1 { \tl_reverse:o { #1 \exp_stop_f: } } }
+\cs_generate_variant:Nn \tl_reverse:N { c }
+%    \end{macrocode}
+%  \end{macro}
+%  
+% \begin{macro}{\tl_trim_spaces:n}
+% \begin{macro}
+%   {
+%     \tl_trim_spaces:N, \tl_trim_spaces:c,
+%     \tl_gtrim_spaces:N, \tl_gtrim_spaces:c
+%   }
+% \begin{macro}[aux]{\tl_trim_spaces_exp:n}
+% \begin{macro}[aux]{\tl_trim_spaces_aux_i:w}
+% \begin{macro}[aux]{\tl_trim_spaces_aux_ii:w}
+%   Trimming spaces from around the input uses the idea of a sufficiently
+%   odd token to allow for a delimited argument to do this. Here, the
+%   standard approach [a |Q| with category code 3 (math toggle)] is used.
+%   The \cs{etex_unexpanded:D} here is used so that space trimming will
+%   behave correctly within an \texttt{x}-type expansion.
+%    \begin{macrocode}
+\cs_new:Npn \tl_trim_spaces:n #1
+  {
+    \etex_unexpanded:D \exp_after:wN \exp_after:wN \exp_after:wN
+      { \tl_trim_spaces_exp:n {#1} }
+  }
+\cs_new_protected:Npn \tl_trim_spaces:N #1
+  { \tl_set:Nx #1 { \exp_after:wN \tl_trim_spaces:n \exp_after:wN {#1} } }
+\cs_new_protected:Npn \tl_gtrim_spaces:N #1
+  { \tl_gset:Nx #1 { \exp_after:wN \tl_trim_spaces:n \exp_after:wN {#1} } }
+\group_begin:
+  \tex_catcode:D `\Q = 3 \scan_stop:
+  \cs_new:Npn \tl_trim_spaces_exp:n #1
+    { \tex_romannumeral:D - `0 \tl_trim_spaces_aux_i:w \exp_not:N #1 Q ~ Q }
+  \cs_new:Npn \tl_trim_spaces_aux_i:w #1 ~ Q { \tl_trim_spaces_aux_ii:w #1 Q }
+  \cs_new:Npn \tl_trim_spaces_aux_ii:w #1 Q #2 {#1}
+\group_end:
+\cs_generate_variant:Nn \tl_trim_spaces:N  { c }
+\cs_generate_variant:Nn \tl_gtrim_spaces:N { c }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \subsection{The first token from a token list}
+%
+% \begin{macro}{\tl_head:n, \tl_head:V, \tl_head:v, \tl_head:f}
+% \begin{macro}{\tl_head:w}
+% \begin{macro}{\tl_tail:n, \tl_tail:V, \tl_tail:v, \tl_tail:f}
+% \begin{macro}{\tl_tail:w}
+%  These functions pick up either the head or the tail of a list.
+%    \begin{macrocode}
+\cs_new:Npn \tl_head:n #1 { \tl_head:w #1 \q_stop }
+\cs_new:Npn \tl_tail:n #1 { \tl_tail:w #1 \q_stop }
+\cs_new:Npn \tl_head:w #1#2 \q_stop {#1}
+\cs_new:Npn \tl_tail:w #1#2 \q_stop {#2}
+\cs_generate_variant:Nn \tl_head:n { V , v , f }
+\cs_generate_variant:Nn \tl_tail:n { V , v , f }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}[pTF]{\tl_if_head_eq_meaning:nN}
+% \begin{macro}[pTF]{\tl_if_head_eq_charcode:nN}
+% \begin{macro}[pTF]{\tl_if_head_eq_charcode:fN}
+% \begin{macro}[pTF]{\tl_if_head_eq_catcode:nN}
+%   When we want to check if the first token of a list equals something
+%   specific it is usually either to see if it is a control sequence or
+%   a character. Hence we make two different functions as the internal
+%   test is different.
+%   \cs{tl_if_head_meaning_eq:nNTF} uses \cs{if_meaning:w} and will
+%   consider the tokens |b|$\sb{11}$ and |b|$\sb{12}$ different.
+%   \cs{tl_if_head_char_eq:nNTF} on the other hand only compares
+%   character codes so would regard |b|$\sb{11}$ and |b|$\sb{12}$ as
+%   equal but would also regard two primitives as equal.
+%    \begin{macrocode}
+\prg_new_conditional:Npnn \tl_if_head_eq_meaning:nN #1#2 { p , T , F , TF }
+  {
+    \exp_after:wN \if_meaning:w \tl_head:w #1 \q_stop #2
+      \prg_return_true:
+    \else:
+      \prg_return_false:
+    \fi:
+  }
+%    \end{macrocode}
+%   For the charcode and catcode versions we insert |\exp_not:N| in
+%   front of both tokens. If you need them to expand fully as \TeX{}
+%   does itself with these you can use an |f| type expansion.
+%    \begin{macrocode}
+\prg_new_conditional:Npnn \tl_if_head_eq_charcode:nN #1#2 { p , T , F , TF }
+  {
+    \exp_after:wN \if:w
+      \exp_after:wN \exp_not:N \tl_head:w #1 \q_stop \exp_not:N #2
+      \prg_return_true:
+    \else:
+      \prg_return_false:
+    \fi:
+  }
+\cs_generate_variant:Nn \tl_if_head_eq_charcode_p:nN { f }
+\cs_generate_variant:Nn \tl_if_head_eq_charcode:nNTF { f }
+\cs_generate_variant:Nn \tl_if_head_eq_charcode:nNT  { f }
+\cs_generate_variant:Nn \tl_if_head_eq_charcode:nNF  { f }
+%    \end{macrocode}
+% And now catcodes:
+%    \begin{macrocode}
+\prg_new_conditional:Npnn \tl_if_head_eq_catcode:nN #1#2 { p , T , F , TF }
+  {
+    \exp_after:wN \if_catcode:w \exp_after:wN \exp_not:N
+        \tl_head:w #1 \q_stop \exp_not:N #2
+      \prg_return_true: \else: \prg_return_false: \fi:
+  }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \subsection{Viewing token lists}
+%
+% \begin{macro}{\tl_show:N, \tl_show:c}
+%   Showing token list variables is done directly: at the moment do not
+%   worry if they are defined.
+%    \begin{macrocode}
+\cs_new_protected:Npn \tl_show:N #1 { \cs_show:N #1 }
+\cs_generate_variant:Nn \tl_show:N { c }
+%    \end{macrocode}
+%\end{macro}
+%
+% \begin{macro}{\tl_show:n}
+%   For literal token lists, life is easy.
+%    \begin{macrocode}
+\cs_new_eq:NN \tl_show:n \etex_showtokens:D
+%    \end{macrocode}
+%\end{macro}
+%
+% \subsection{Constant token lists}
+%
+% \begin{variable}{\c_job_name_tl}
+%   Inherited from the \LaTeX3 name for the primitive: this needs to
+%   actually contain the text of the job name rather than the name of
+%   the primitive, of course.
+%    \begin{macrocode}
+%<*package>
+\tl_const:Nx \c_job_name_tl { \tex_jobname:D }
+%</package>
+%<*initex>
+\tex_everyjob:D \exp_after:wN
+  {
+    \tex_the:D \tex_everyjob:D
+    \luatex_if_engine:TF
+      {
+        \tl_if_in:onTF { \tex_jobname:D } { ~ }
+          { \tl_const:Nx \c_job_name_tl { " \tex_jobname:D " } }
+          { \tl_const:Nx \c_job_name_tl { \tex_jobname:D } }
+      }
+      { \tl_const:Nx \c_job_name_tl { \tex_jobname:D } }
+  }
+%</initex>
+%    \end{macrocode}
+% \end{variable}
+%
+% \begin{variable}{\c_empty_tl}
+%   Never full.
+%    \begin{macrocode}
+\tl_const:Nn \c_empty_tl { }
+%    \end{macrocode}
+% \end{variable}
+%
+% \begin{variable}{\c_space_tl}
+%   A space as a token list (as opposed to as a character).
+%    \begin{macrocode}
+\tl_const:Nn \c_space_tl { ~ }
+%    \end{macrocode}
+% \end{variable}
+%
+% \subsection{Scratch token lists}
+%
+% \begin{variable}{\g_tmpa_tl, \g_tmpb_tl}
+%    Global temporary token list variables.
+%    They are supposed to be set and used immediately,
+%    with no delay between the definition and the use because you
+%    can't count on other macros not to redefine them from under you.
+%    \begin{macrocode}
+\tl_new:N \g_tmpa_tl
+\tl_new:N \g_tmpb_tl
+%    \end{macrocode}
+% \end{variable}
+%
+% \begin{variable}{\l_tmpa_tl, \l_tmpb_tl}
+%    These are local temporary token list variables. Be sure not to assume
+%    that the value you put into them will survive for
+%    long---see discussion above.
+%    \begin{macrocode}
+\tl_new:N \l_tmpa_tl
+\tl_new:N \l_tmpb_tl
+%    \end{macrocode}
+% \end{variable}
+%
+% \subsection{Experimental functions}
+%
+% \begin{macro}[EXP,pTF]{\tl_if_single_item:n}
+%   A token list has exactly one item if it is either a single
+%   token surrounded by optional explicit spaces, or a single brace
+%   group surrounded by optional explicit spaces. The naive
+%   version of this test would do \cs{use_none:n} |#1|, and
+%   test if the result is empty. However, this will fail when
+%   the token list is empty. Furthermore, it does not allow optional
+%   trailing spaces.
+%    \begin{macrocode}
+\cs_new:Npn \tl_if_single_item:nTF #1
+  { \str_if_eq:onTF { \use_ii:nn #1 {?} ? } {??} }
+\cs_new:Npn \tl_if_single_item:nT #1
+  { \str_if_eq:onT { \use_ii:nn #1 {?} ? } {??} }
+\cs_new:Npn \tl_if_single_item:nF #1
+  { \str_if_eq:onF { \use_ii:nn #1 {?} ? } {??} }
+\cs_new:Npn \tl_if_single_item_p:n #1
+  { \str_if_eq_p:on { \use_ii:nn #1 {?} ? } {??} }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}[EXP,pTF]{\tl_if_head_begin_group:n}
+%    \begin{macrocode}
+\prg_new_conditional:Npnn \tl_if_head_begin_group:n #1 { p , T , F , TF }
+  {
+    \exp_after:wN \use_none:n
+      \exp_after:wN {
+        \exp_after:wN {
+          \token_to_str:N #1 .
+        }
+        \prg_return_true: \exp_after:wN \use_none:nn \token_to_str:N
+      }
+      \prg_return_false:
+  }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}[EXP,pTF]{\tl_if_head_eq_space:n}
+% \begin{macro}[EXP,aux]{\tl_if_head_eq_space_aux:w}
+%   If the first token of the token list is an explicit space, i.e.,
+%   a character token with character code $32$ and category code $10$,
+%   then this test will be \meta{true}. It is \meta{false} if the token
+%   list is empty, if the first token is an implicit space token,
+%   such as \cs{c_space_token}, or any token other than an explicit space.
+%    \begin{macrocode}
+\prg_new_conditional:Npnn \tl_if_head_eq_space:n #1 { p , T , F , TF }
+  {
+    \if_false: { \fi:
+      \tl_if_head_eq_space_aux:w \prg_do_nothing: #1 ? ~ }
+  }
+\cs_new:Npn \tl_if_head_eq_space_aux:w #1 ~ %
+  {
+    \tl_if_empty_return:o { #1 }
+    \exp_after:wN \use_none:n \exp_after:wN { \if_false: } \fi:
+  }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% 
+% \begin{macro}[EXP,pTF]{\tl_if_single_token:n}
+%   If the token list starts with a space, then it is
+%   a single token if and only if its string representation
+%   is exactly one space character. Otherwise, we check if removing
+%   the first token group makes the token list empty. If it doesn't,
+%   the token list is not a single token. If it does, then the token
+%   list is either a single token or a single brace group. This is
+%   tested for using \cs{tl_if_head_begin_group:n}.
+%    \begin{macrocode}
+\prg_new_conditional:Npnn \tl_if_single_token:n #1 { p , T , F , TF }
+  {
+    \tl_if_head_eq_space:nTF {#1}
+      {
+        \str_if_eq:nnTF {#1} {~}
+          \prg_return_true: \prg_return_false:
+      }
+      {
+        \str_if_eq:onTF { \use_none:n #1 ? } {?}
+          {
+            \tl_if_head_begin_group:nTF {#1}
+              \prg_return_false: \prg_return_true:
+          }
+          \prg_return_false:
+      }
+  }
+%    \end{macrocode}
+% \end{macro}
+%
+% \subsection{Deprecated functions}
+%
+% \begin{macro}{\tl_new:Nn, \tl_new:cn, \tl_new:Nx}
+%   Use either \cs{tl_const:Nn} or \cs{tl_new:N}.
+%    \begin{macrocode}
+\cs_new_protected:Npn \tl_new:Nn #1#2
+  {
+    \tl_new:N #1
+    \tl_gset:Nn #1 {#2}
+  }
+\cs_generate_variant:Nn \tl_new:Nn { c }
+\cs_generate_variant:Nn \tl_new:Nn { Nx }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\tl_gset:Nc}
+% \begin{macro}{\tl_set:Nc}
+%   This was useful once, but nowadays does not make much sense.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \tl_gset:Nc
+  { \pref_global:D \tl_set:Nc }
+\cs_new_protected_nopar:Npn \tl_set:Nc #1#2
+  { \tl_set:No #1 { \cs:w #2 \cs_end: } }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\tl_replace_in:Nnn, \tl_replace_in:cnn}
+% \begin{macro}{\tl_greplace_in:Nnn, \tl_greplace_in:cnn}
+% \begin{macro}{\tl_replace_all_in:Nnn, \tl_replace_all_in:cnn}
+% \begin{macro}{\tl_greplace_all_in:Nnn, \tl_greplace_all_in:cnn}
+%   These are renamed.
+%    \begin{macrocode}
+\cs_new_eq:NN \tl_replace_in:Nnn  \tl_replace_once:Nnn
+\cs_new_eq:NN \tl_replace_in:cnn  \tl_replace_once:cnn
+\cs_new_eq:NN \tl_greplace_in:Nnn \tl_greplace_once:Nnn
+\cs_new_eq:NN \tl_greplace_in:cnn \tl_greplace_once:cnn
+\cs_new_eq:NN \tl_replace_all_in:Nnn  \tl_replace_all:Nnn
+\cs_new_eq:NN \tl_replace_all_in:cnn  \tl_replace_all:cnn
+\cs_new_eq:NN \tl_greplace_all_in:Nnn \tl_greplace_all:Nnn
+\cs_new_eq:NN \tl_greplace_all_in:cnn \tl_greplace_all:cnn
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\tl_remove_in:Nn, \tl_remove_in:cn}
+% \begin{macro}{\tl_gremove_in:Nn, \tl_gremove_in:cn}
+% \begin{macro}{\tl_remove_all_in:Nn, \tl_remove_all_in:cn}
+% \begin{macro}{\tl_gremove_all_in:Nn, \tl_gremove_all_in:cn}
+%   Also renamed.
+%    \begin{macrocode}
+\cs_new_eq:NN \tl_remove_in:Nn  \tl_remove_once:Nn
+\cs_new_eq:NN \tl_remove_in:cn  \tl_remove_once:cn
+\cs_new_eq:NN \tl_gremove_in:Nn \tl_gremove_once:Nn
+\cs_new_eq:NN \tl_gremove_in:cn \tl_gremove_once:cn
+\cs_new_eq:NN \tl_remove_all_in:Nn  \tl_remove_all:Nn
+\cs_new_eq:NN \tl_remove_all_in:cn  \tl_remove_all:cn
+\cs_new_eq:NN \tl_gremove_all_in:Nn \tl_gremove_all:Nn
+\cs_new_eq:NN \tl_gremove_all_in:cn \tl_gremove_all:cn
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\tl_elt_count:n, \tl_elt_count:V, \tl_elt_count:o}
+% \begin{macro}{\tl_elt_count:N, \tl_elt_count:c}
+% Another renaming job.
+%    \begin{macrocode}
+\cs_new_eq:NN \tl_elt_count:n \tl_length:n
+\cs_new_eq:NN \tl_elt_count:V \tl_length:V
+\cs_new_eq:NN \tl_elt_count:o \tl_length:o
+\cs_new_eq:NN \tl_elt_count:N \tl_length:N
+\cs_new_eq:NN \tl_elt_count:c \tl_length:c
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\tl_head_i:n}
+% \begin{macro}{\tl_head_i:w}
+% \begin{macro}{\tl_head_iii:n}
+% \begin{macro}{\tl_head_iii:f}
+% \begin{macro}{\tl_head_iii:w}
+%  Two renames, and a few that are rather too specialised.
+%    \begin{macrocode}
+\cs_new_eq:NN \tl_head_i:n \tl_head:n
+\cs_new_eq:NN \tl_head_i:w \tl_head:w
+\cs_new:Npn \tl_head_iii:n #1 { \tl_head_iii:w #1 \q_stop }
+\cs_generate_variant:Nn \tl_head_iii:n { f }
+\cs_new:Npn \tl_head_iii:w #1#2#3#4 \q_stop {#1#2#3}
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+%    \begin{macrocode}
+%</initex|package>
+%    \end{macrocode}
+%
+% \end{implementation}
+%
+% \PrintIndex
diff --git a/Master/texmf-dist/source/latex/l3kernel/l3token.dtx b/Master/texmf-dist/source/latex/l3kernel/l3token.dtx
new file mode 100644
index 00000000000..153ca92f7e0
--- /dev/null
+++ b/Master/texmf-dist/source/latex/l3kernel/l3token.dtx
@@ -0,0 +1,2495 @@
+% \iffalse meta-comment
+%
+%% File: l3token.dtx Copyright (C) 2005-2011 The LaTeX3 Project
+%%
+%% It may be distributed and/or modified under the conditions of the
+%% LaTeX Project Public License (LPPL), either version 1.3c of this
+%% license or (at your option) any later version.  The latest version
+%% of this license is in the file
+%%
+%%    http://www.latex-project.org/lppl.txt
+%%
+%% This file is part of the "expl3 bundle" (The Work in LPPL)
+%% and all files in that bundle must be distributed together.
+%%
+%% The released version of this bundle is available from CTAN.
+%%
+%% -----------------------------------------------------------------------
+%%
+%% The development version of the bundle can be found at
+%%
+%%    http://www.latex-project.org/svnroot/experimental/trunk/
+%%
+%% for those people who are interested.
+%%
+%%%%%%%%%%%
+%% NOTE: %%
+%%%%%%%%%%%
+%%
+%%   Snapshots taken from the repository represent work in progress and may
+%%   not work or may contain conflicting material!  We therefore ask
+%%   people _not_ to put them into distributions, archives, etc. without
+%%   prior consultation with the LaTeX3 Project.
+%%
+%% -----------------------------------------------------------------------
+%
+%<*driver|package>
+\RequirePackage{l3names}
+\GetIdInfo$Id: l3token.dtx 2491 2011-07-02 11:49:37Z joseph $
+  {L3 Experimental token manipulation}
+%</driver|package>
+%<*driver>
+\documentclass[full]{l3doc}
+\begin{document}
+  \DocInput{\jobname.dtx}
+\end{document}
+%</driver>
+% \fi
+%
+% \title{^^A
+%   The \pkg{l3token} package\\ Token manipulation^^A
+%   \thanks{This file describes v\ExplFileVersion,
+%      last revised \ExplFileDate.}^^A
+% }
+%
+% \author{^^A
+%  The \LaTeX3 Project\thanks
+%    {^^A
+%      E-mail:
+%        \href{mailto:latex-team@latex-project.org}
+%          {latex-team@latex-project.org}^^A
+%    }^^A
+% }
+%
+% \date{Released \ExplFileDate}
+%
+% \maketitle
+%
+% \begin{documentation}
+%
+% This module deals with tokens. Now this is perhaps not the most
+% precise description so let's try with a better description: When
+% programming in \TeX{}, it is often desirable to know just what a
+% certain token is: is it a control sequence or something
+% else. Similarly one often needs to know if a control sequence is
+% expandable or not, a macro or a primitive, how many arguments it
+% takes etc. Another thing of great importance (especially when it
+% comes to document commands) is looking ahead in the token stream to
+% see if a certain character is present and maybe even remove it or
+% disregard other tokens while scanning. This module provides
+% functions for both and as such will have two primary function
+% categories: |\token| for anything that deals with tokens and
+% |\peek| for looking ahead in the token stream.
+%
+% Most of the time we will be using the term \enquote{token} but most of the
+% time the function we're describing can equally well by used on a
+% control sequence as such one is one token as well.
+%
+% We shall refer to list of tokens as |tlist|s and such lists
+% represented by a single control sequence is a \enquote{token list variable}
+% |tl var|. Functions for these two types are found in the \textsf{l3tl}
+% module.
+%
+% \section{All possible tokens}
+%
+% Let us start by reviewing every case that a given token can fall into.
+% It is very important to distinguish two aspects of a token: its meaning,
+% and what it looks like.
+%
+% For instance, \cs{if:w}, \cs{if_charcode:w}, and \cs{tex_if:D} are
+% three for the same internal operation of \TeX{}, namely the primitive
+% testing the next two characters for equality of their character code.
+% They behave identically in many situations. However, \TeX{}
+% distinguishes them when searching for a delimited argument. Namely, the
+% example function \cs{show_until_if:w} defined below will take everything
+% until \cs{if:w} as an argument, despite the presence of other copies of
+% \cs{if:w} under different names.
+% \begin{verbatim}
+% \cs_new:Npn \show_until_if:w #1 \if:w { \tl_show:n {#1} }
+% \show_until_if:w \tex_if:D \if_charcode:w \if:w
+% \end{verbatim}
+%
+% It is easier to start by considering the possibilities for what a
+% token looks like, in other words, how \TeX{} sees it from the point
+% of view of delimited arguments. Two cases: the token is either a
+% control sequence or a single character. 
+% \begin{center}
+%   \begin{tabular}{p{4cm}p{4cm}}
+%     \toprule
+%     Control sequence & Character \\
+%     \midrule
+%     Control word: an escape character followed by some letters. &
+%   \end{tabular}
+% \end{center}
+% The token can be a control sequence,
+% in other words, an escape character followed by some letters, or by
+% a single non-letter character. Examples of this include \cs{begin},
+% \cs{;}, \cs{emph}\ldots{} The other case 
+% There are two cases. either the token
+% is a single character, in which case it is associated with a category
+% code:
+% 
+%
+% \section{Character tokens}
+%
+% \begin{function}
+%   {
+%     \char_set_catcode_escape:N           ,
+%     \char_set_catcode_group_begin:N      ,
+%     \char_set_catcode_group_end:N        ,
+%     \char_set_catcode_math_toggle:N      ,
+%     \char_set_catcode_alignment:N        ,
+%     \char_set_catcode_end_line:N         ,
+%     \char_set_catcode_parameter:N        ,
+%     \char_set_catcode_math_superscript:N ,
+%     \char_set_catcode_math_subscript:N   ,
+%     \char_set_catcode_ignore:N           ,
+%     \char_set_catcode_space:N            ,
+%     \char_set_catcode_letter:N           ,
+%     \char_set_catcode_other:N            ,
+%     \char_set_catcode_active:N           ,
+%     \char_set_catcode_comment:N          ,
+%     \char_set_catcode_invalid:N
+%   }
+%   \begin{syntax}
+%     \cs{char_make_letter:N} \meta{character}
+%   \end{syntax}
+%   Sets the category code of the \meta{character} to that indicated in
+%   the function name. Depending on the current category code of the
+%   \meta{token} the escape token may also be needed:
+%   \begin{verbatim}
+%     \char_set_catcode_other:N \%
+%   \end{verbatim}
+%   The assignment is local.
+% \end{function}
+%
+% \begin{function}
+%   {
+%     \char_set_catcode_escape:n           ,
+%     \char_set_catcode_group_begin:n      ,
+%     \char_set_catcode_group_end:n        ,
+%     \char_set_catcode_math_toggle:n       ,
+%     \char_set_catcode_alignment:n        ,
+%     \char_set_catcode_end_line:n         ,
+%     \char_set_catcode_parameter:n        ,
+%     \char_set_catcode_math_superscript:n ,
+%     \char_set_catcode_math_subscript:n   ,
+%     \char_set_catcode_ignore:n           ,
+%     \char_set_catcode_space:n            ,
+%     \char_set_catcode_letter:n           ,
+%     \char_set_catcode_other:n            ,
+%     \char_set_catcode_active:n           ,
+%     \char_set_catcode_comment:n          ,
+%     \char_set_catcode_invalid:n
+%   }
+%   \begin{syntax}
+%     \cs{char_make_letter:n} \Arg{integer expression}
+%   \end{syntax}
+%   Sets the category code of the \meta{character} which has character
+%   code as given by the \meta{integer expression}. This version can be
+%   used to set up characters which cannot otherwise be given
+%   (\emph{cf.}~the \texttt{N}-type variants). The assignment is local.
+% \end{function}
+%
+% \begin{function}{\char_set_catcode:nn}
+%   \begin{syntax}
+%     \cs{char_set_catcode:nn} \Arg{intexpr1} \Arg{intexpr2}
+%   \end{syntax}
+%   These functions set the category code of the \meta{character} which
+%   has character code as given by the \meta{integer expression}.
+%   The first \meta{integer expression}
+%   is the character code and the second is the category code to apply.
+%   The setting applies within the current \TeX{} group. In general, the
+%   symbolic functions \cs{char_make_\meta{type}} should be preferred,
+%   but there are cases where these lower-level functions may be useful.
+% \end{function}
+%
+% \begin{function}[EXP]{\char_value_catcode:n}
+%   \begin{syntax}
+%     \cs{char_value_catcode:n} \Arg{integer expression}
+%   \end{syntax}
+%   Expands to the current category code of the \meta{character} with
+%   character code given by the
+%   \meta{integer expression}.
+% \end{function}
+%
+% \begin{function}{\char_show_value_catcode:n}
+%   \begin{syntax}
+%     \cs{char_show_value_catcode:n} \Arg{integer expression}
+%   \end{syntax}
+%   Displays the current category code of the \meta{character} with
+%   character code given by the  \meta{integer expression} on the
+%   terminal.
+% \end{function}
+%
+% \begin{function}{\char_set_lccode:nn}
+%   \begin{syntax}
+%     \cs{char_set_lcode:nn} \Arg{intexpr1} \Arg{intexpr2}
+%   \end{syntax}
+%   This function set up the behaviour of \meta{character} when
+%   found inside \cs{tl_to_lowercase:n}, such that \meta{character1}
+%   will be converted into \meta{character2}. The two \meta{characters}
+%   may be specified using an \meta{integer expression} for the character code
+%   concerned. This may include the \TeX{} |`|\meta{character}
+%   method for converting a single character into its character
+%   code:
+%   \begin{verbatim}
+%     \char_set_lccode:nn { `\A } { `\a } % Standard behaviour
+%     \char_set_lccode:nn { `\A } { `\A + 32 }
+%     \char_set_lccode:nn { 50 } { 60 }
+%   \end{verbatim}
+%   The setting applies within the current \TeX{} group.
+% \end{function}
+%
+% \begin{function}[EXP]{\char_value_lccode:n}
+%   \begin{syntax}
+%     \cs{char_value_lccode:n} \Arg{integer expression}
+%   \end{syntax}
+%   Expands to the current lower case code of the \meta{character} with
+%   character code given by the
+%   \meta{integer expression}.
+% \end{function}
+%
+% \begin{function}{\char_show_value_lccode:n}
+%   \begin{syntax}
+%     \cs{char_show_value_lccode:n} \Arg{integer expression}
+%   \end{syntax}
+%   Displays the current lower case code of the \meta{character} with
+%   character code given by the  \meta{integer expression} on the
+%   terminal.
+% \end{function}
+%
+% \begin{function}{\char_set_uccode:nn}
+%   \begin{syntax}
+%     \cs{char_set_uccode:nn} \Arg{intexpr1} \Arg{intexpr2}
+%   \end{syntax}
+%   This function set up the behaviour of \meta{character} when
+%   found inside \cs{tl_to_uppeercase:n}, such that \meta{character1}
+%   will be converted into \meta{character2}. The two \meta{characters}
+%   may be specified using an \meta{integer expression} for the character code
+%   concerned. This may include the \TeX{} |`|\meta{character}
+%   method for converting a single character into its character
+%   code:
+%   \begin{verbatim}
+%     \char_set_uccode:nn { `\a } { `\A } % Standard behaviour
+%     \char_set_uccode:nn { `\A } { `\A - 32 }
+%     \char_set_uccode:nn { 60 } { 50 }
+%   \end{verbatim}
+%   The setting applies within the current \TeX{} group.
+% \end{function}
+%
+% \begin{function}[EXP]{\char_value_uccode:n}
+%   \begin{syntax}
+%     \cs{char_value_uccode:n} \Arg{integer expression}
+%   \end{syntax}
+%   Expands to the current upper case code of the \meta{character} with
+%   character code given by the
+%   \meta{integer expression}.
+% \end{function}
+%
+% \begin{function}{\char_show_value_uccode:n}
+%   \begin{syntax}
+%     \cs{char_show_value_uccode:n} \Arg{integer expression}
+%   \end{syntax}
+%   Displays the current upper case code of the \meta{character} with
+%   character code given by the  \meta{integer expression} on the
+%   terminal.
+% \end{function}
+%
+% \begin{function}{\char_set_mathcode:nn}
+%   \begin{syntax}
+%     \cs{char_set_mathcode:nn} \Arg{intexpr1} \Arg{intexpr2}
+%   \end{syntax}
+%   This function sets up the math code of \meta{character}.
+%   The \meta{character} is specified as
+%   an \meta{integer expression} which will be used as the character
+%   code of the relevant character. The setting applies within the
+%   current \TeX{} group.
+% \end{function}
+%
+% \begin{function}[EXP]{\char_value_mathcode:n}
+%   \begin{syntax}
+%     \cs{char_value_mathcode:n} \Arg{integer expression}
+%   \end{syntax}
+%   Expands to the current math code of the \meta{character} with
+%   character code given by the
+%   \meta{integer expression}.
+% \end{function}
+%
+% \begin{function}{\char_show_value_mathcode:n}
+%   \begin{syntax}
+%     \cs{char_show_value_mathcode:n} \Arg{integer expression}
+%   \end{syntax}
+%   Displays the current math code of the \meta{character} with
+%   character code given by the  \meta{integer expression} on the
+%   terminal.
+% \end{function}
+%
+% \begin{function}{\char_set_sfcode:nn}
+%   \begin{syntax}
+%     \cs{char_set_sfcode:nn} \Arg{intexpr1} \Arg{intexpr2}
+%   \end{syntax}
+%   This function sets up the space factor for the \meta{character}.
+%   The \meta{character} is specified as
+%   an \meta{integer expression} which will be used as the character
+%   code of the relevant character. The setting applies within the
+%   current \TeX{} group.
+% \end{function}
+%
+% \begin{function}[EXP]{\char_value_sfcode:n}
+%   \begin{syntax}
+%     \cs{char_value_sfcode:n} \Arg{integer expression}
+%   \end{syntax}
+%   Expands to the current space factor for the \meta{character} with
+%   character code given by the
+%   \meta{integer expression}.
+% \end{function}
+%
+% \begin{function}{\char_show_value_sfcode:n}
+%   \begin{syntax}
+%     \cs{char_show_value_sfcode:n} \Arg{integer expression}
+%   \end{syntax}
+%   Displays the current space factor for the \meta{character} with
+%   character code given by the  \meta{integer expression} on the
+%   terminal.
+% \end{function}
+%
+% \section{Generic tokens}
+%
+% \begin{function}{\token_new:Nn}
+%   \begin{syntax}
+%     \cs{token_new:Nn} \meta{token1} \Arg{token2}
+%   \end{syntax}
+%   Defines \meta{token1} to globally be a snapshot of \meta{token2}.
+%   This will be an implicit representation of \meta{token2}.
+% \end{function}
+%
+% \begin{variable}
+%   {
+%     \c_group_begin_token,
+%     \c_group_end_token,
+%     \c_math_toggle_token,
+%     \c_alignment_token,
+%     \c_parameter_token,
+%     \c_math_superscript_token,
+%     \c_math_subscript_token,
+%     \c_space_token
+%   }
+%   These are implicit tokens which have the category code described
+%   by their name. They are used internally for test purposes but
+%   are also available to the programmer for other uses.
+% \end{variable}
+% 
+% \begin{variable}
+%   {
+%     \c_catcode_letter_token,
+%     \c_catcode_other_token
+%   }
+%   These are implicit tokens which have the category code described
+%   by their name. They are used internally for test purposes and should
+%   not be used other than for category code tests.
+% \end{variable}
+% 
+% \begin{variable}{\c_catcode_active_tl}
+%   A token list containing an active token. This is used internally
+%   for test purposes and should not be used other than in
+%   appropriately-constructed category code tests.
+% \end{variable}
+% 
+% \section{Converting tokens}
+% 
+% \begin{function}[EXP]{\token_to_meaning:N}
+%   \begin{syntax}
+%     \cs{token_to_meaning:N} \meta{token}
+%   \end{syntax}
+%   Inserts the current meaning of the \meta{token} into the input
+%   stream as a series of characters of category code $12$ (other).
+%   This will be the primitive \TeX{} description of the \meta{token},
+%   thus for example both functions defined by \cs{cs_set_nopar:Npn}
+%   and token list variables defined using \cs{tl_new:N} will be described
+%   as |macro|s.
+%   \begin{texnote}
+%     This is the \TeX{} primitive \tn{meaning}.
+%   \end{texnote}
+% \end{function}
+% 
+% \begin{function}[EXP]{\token_to_str:N, \token_to_str:c}
+%   \begin{syntax}
+%     \cs{token_to_str:N} \meta{token}
+%   \end{syntax}
+%   Converts the given \meta{token} into a series of characters with
+%   category code $12$ (other). The current escape character will be
+%   the first character in the sequence, although this will also have
+%   category code $12$ (the escape character is part of the
+%   \meta{token}). This function requires only a single expansion.
+%   \begin{texnote}
+%     \cs{token_to_str:N} is the \TeX{} primitive \cs{string} renamed.
+%   \end{texnote}
+% \end{function}
+%
+% \section{Token conditionals}
+%
+% \begin{function}[EXP,pTF]{\token_if_group_begin:N}
+%   \begin{syntax}
+%     \cs{token_if_group_begin_p:N} \meta{token}
+%     \cs{token_if_group_begin:NTF} \meta{token} \Arg{true code}
+%     ~~\Arg{false code}
+%   \end{syntax}
+%   Tests if \meta{token} has the category code of a begin group token
+%   (|{| when normal \TeX{} category codes are in ^^A }
+%   force). The branching versions then leave either \meta{true code} or
+%   \meta{false code} in the input stream, as appropriate to the truth of
+%   the test and the variant of the function chosen. The logical truth of
+%   the test is left in the input stream by the predicate version.
+%   Note that an explicit begin group token cannot be tested in this way,
+%   as it is not a valid \texttt{N}-type argument.
+% \end{function}
+%
+% \begin{function}[EXP,pTF]{\token_if_group_end:N}
+%   \begin{syntax}
+%     \cs{token_if_group_end_p:N} \meta{token}
+%     \cs{token_if_group_end:NTF} \meta{token} \Arg{true code}
+%     ~~\Arg{false code}
+%    \end{syntax}
+%   Tests if \meta{token} has the category code of an end group token
+%   (^^A {
+%   |}| when normal \TeX{} category codes are in force). The branching
+%   versions then leave either \meta{true code} or \meta{false code} in the
+%   input stream, as appropriate to the truth of the test and the variant
+%   of the function chosen. The logical truth of the test is left in the
+%   input stream by the predicate version.
+%   Note that an explicit end group token cannot be tested in this way,
+%   as it is not a valid \texttt{N}-type argument.
+% \end{function}
+%
+% \begin{function}[EXP,pTF]{\token_if_math_toggle:N}
+%   \begin{syntax}
+%     \cs{token_if_math_toggle_p:N} \meta{token}
+%     \cs{token_if_math_toggle:NTF} \meta{token} \Arg{true code}
+%     ~~\Arg{false code}
+%   \end{syntax}
+%   Tests if \meta{token} has the category code of a math shift token
+%   (|$| when normal \TeX{} category codes are in force). The
+%   branching versions then leave either \meta{true code} or
+%   \meta{false code} in the input stream, as appropriate to the truth of
+%   the test and the variant of the function chosen. The logical truth of
+%   the test is left in the input stream by the predicate version.
+% \end{function}
+%
+% \begin{function}[EXP,pTF]{\token_if_alignment:N}
+%   \begin{syntax}
+%     \cs{token_if_alignment_p:N} \meta{token}
+%     \cs{token_if_alignment:NTF} \meta{token} \Arg{true code}
+%     ~~\Arg{false code}
+%   \end{syntax}
+%   Tests if \meta{token} has the category code of an alignment token
+%   (|&| when normal \TeX{} category codes are in force). The
+%   branching versions then leave either \meta{true code} or
+%   \meta{false code} in the input stream, as appropriate to the truth of
+%   the test and the variant of the function chosen. The logical truth of
+%   the test is left in the input stream by the predicate version.
+% \end{function}
+%
+% \begin{function}[EXP,pTF]{\token_if_parameter:N}
+%   \begin{syntax}
+%     \cs{token_if_parameter_p:N} \meta{token}
+%     \cs{token_if_alignment:NTF} \meta{token} \Arg{true code}
+%     ~~\Arg{false code}
+%   \end{syntax}
+%   Tests if \meta{token} has the category code of a macro parameter token
+%   (|#| when normal \TeX{} category codes are in force). The
+%   branching versions then leave either \meta{true code} or
+%   \meta{false code} in the input stream, as appropriate to the truth of
+%   the test and the variant of the function chosen. The logical truth of
+%   the test is left in the input stream by the predicate version.
+% \end{function}
+%
+% \begin{function}[EXP,pTF]{\token_if_math_superscript:N}
+%   \begin{syntax}
+%     \cs{token_if_math_superscript_p:N} \meta{token}
+%     \cs{token_if_math_superscript:NTF} \meta{token} \Arg{true code}
+%     ~~\Arg{false code}
+%   \end{syntax}
+%   Tests if \meta{token} has the category code of a superscript token
+%   (|^| when normal \TeX{} category codes are in force). The
+%   branching versions then leave either \meta{true code} or
+%   \meta{false code} in the input stream, as appropriate to the truth of
+%   the test and the variant of the function chosen. The logical truth of
+%   the test is left in the input stream by the predicate version.
+% \end{function}
+%
+% \begin{function}[EXP,pTF]{\token_if_math_subscript:N}
+%   \begin{syntax}
+%     \cs{token_if_math_subscript_p:N} \meta{token}
+%     \cs{token_if_math_subscript:NTF} \meta{token} \Arg{true code}
+%     ~~\Arg{false code}
+%   \end{syntax}
+%   Tests if \meta{token} has the category code of a subscript token
+%   (|_| when normal \TeX{} category codes are in force). The
+%   branching versions then leave either \meta{true code} or
+%   \meta{false code} in the input stream, as appropriate to the truth of
+%   the test and the variant of the function chosen. The logical truth of
+%   the test is left in the input stream by the predicate version.
+% \end{function}
+%
+% \begin{function}[EXP,pTF]{\token_if_space:N}
+%   \begin{syntax}
+%     \cs{token_if_space_p:N} \meta{token}
+%     \cs{token_if_space:NTF} \meta{token} \Arg{true code}
+%     ~~\Arg{false code}
+%   \end{syntax}
+%   Tests if \meta{token} has the category code of a space token. The
+%   branching versions then leave either \meta{true code} or
+%   \meta{false code} in the input stream, as appropriate to the truth of
+%   the test and the variant of the function chosen. The logical truth of
+%   the test is left in the input stream by the predicate version.
+%   Note that an explicit space token with character code $32$ cannot
+%   be tested in this way, as it is not a valid \texttt{N}-type argument.
+% \end{function}
+%
+% \begin{function}[EXP,pTF]{\token_if_letter:N}
+%   \begin{syntax}
+%     \cs{token_if_letter_p:N} \meta{token}
+%     \cs{token_if_letter:NTF} \meta{token} \Arg{true code}
+%     ~~\Arg{false code}
+%   \end{syntax}
+%   Tests if \meta{token} has the category code of a letter token. The
+%   branching versions then leave either \meta{true code} or
+%   \meta{false code} in the input stream, as appropriate to the truth of
+%   the test and the variant of the function chosen. The logical truth of
+%   the test is left in the input stream by the predicate version.
+% \end{function}
+%
+% \begin{function}[EXP,pTF]{\token_if_other:N}
+%   \begin{syntax}
+%     \cs{token_if_other_p:N} \meta{token}
+%     \cs{token_if_other:NTF} \meta{token} \Arg{true code}
+%     ~~\Arg{false code}
+%   \end{syntax}
+%   Tests if \meta{token} has the category code of an \enquote{other}
+%   token. The branching versions then leave either \meta{true code} or
+%   \meta{false code} in the input stream, as appropriate to the truth of
+%   the test and the variant of the function chosen. The logical truth of
+%   the test is left in the input stream by the predicate version.
+% \end{function}
+%
+% \begin{function}[EXP,pTF]{\token_if_active:N}
+%   \begin{syntax}
+%     \cs{token_if_active_p:N} \meta{token}
+%     \cs{token_if_active:NTF} \meta{token} \Arg{true code}
+%     ~~\Arg{false code}
+%   \end{syntax}
+%   Tests if \meta{token} has the category code of an active character.
+%   The branching versions then leave either \meta{true code} or
+%   \meta{false code} in the input stream, as appropriate to the truth of
+%   the test and the variant of the function chosen. The logical truth of
+%   the test is left in the input stream by the predicate version.
+% \end{function}
+%
+% \begin{function}[EXP,pTF]{\token_if_eq_catcode:NN}
+%   \begin{syntax}
+%     \cs{token_if_eq_catcode_p:NN} \meta{token1} \meta{token2}
+%     \cs{token_if_eq_catcode:NNTF} \meta{token1} \meta{token2}
+%     ~~\Arg{true code} \Arg{false code}
+%   \end{syntax}
+%   Tests if the two \meta{tokens} have the same category code. The
+%   branching versions then leave either \meta{true code} or
+%   \meta{false code} in the input stream, as appropriate to the truth of
+%   the test and the variant of the function chosen. The logical truth of
+%   the test is left in the input stream by the predicate version.
+% \end{function}
+%
+% \begin{function}[EXP,pTF]{\token_if_eq_charcode:NN}
+%   \begin{syntax}
+%     \cs{token_if_eq_charcode_p:NN} \meta{token1} \meta{token2}
+%     \cs{token_if_eq_charcode:NNTF} \meta{token1} \meta{token2}
+%     ~~\Arg{true code} \Arg{false code}
+%   \end{syntax}
+%   Tests if the two \meta{tokens} have the same character code. The
+%   branching versions then leave either \meta{true code} or
+%   \meta{false code} in the input stream, as appropriate to the truth of
+%   the test and the variant of the function chosen. The logical truth of
+%   the test is left in the input stream by the predicate version.
+% \end{function}
+%
+% \begin{function}[EXP,pTF]{\token_if_eq_meaning:NN}
+%   \begin{syntax}
+%     \cs{token_if_eq_meaning_p:NN} \meta{token1} \meta{token2}
+%     \cs{token_if_eq_meaning:NNTF} \meta{token1} \meta{token2}
+%     ~~\Arg{true code} \Arg{false code}
+%   \end{syntax}
+%   Tests if the two \meta{tokens} have the same meaning when expanded.
+%   The branching versions then leave either \meta{true code} or
+%   \meta{false code} in the input stream, as appropriate to the truth of
+%   the test and the variant of the function chosen. The logical truth of
+%   the test is left in the input stream by the predicate version.
+% \end{function}
+%
+% \begin{function}[EXP,pTF]{\token_if_macro:N}
+%   \begin{syntax}
+%     \cs{token_if_macro_p:N} \meta{token}
+%     \cs{token_if_macro:NTF} \meta{token} \Arg{true code} \Arg{false code}
+%   \end{syntax}
+%   Tests if the \meta{token} is a \TeX{} macro.
+%   The branching versions then leave either \meta{true code} or
+%   \meta{false code} in the input stream, as appropriate to the truth of
+%   the test and the variant of the function chosen. The logical truth of
+%   the test is left in the input stream by the predicate version.
+% \end{function}
+%
+% \begin{function}[EXP,pTF]{\token_if_cs:N}
+%   \begin{syntax}
+%     \cs{token_if_cs_p:N} \meta{token}
+%     \cs{token_if_cs:NTF} \meta{token} \Arg{true code} \Arg{false code}
+%   \end{syntax}
+%   Tests if the \meta{token} is a control sequence.
+%   The branching versions then leave either \meta{true code} or
+%   \meta{false code} in the input stream, as appropriate to the truth of
+%   the test and the variant of the function chosen. The logical truth of
+%   the test is left in the input stream by the predicate version.
+% \end{function}
+%
+% \begin{function}[EXP,pTF]{\token_if_expandable:N}
+%   \begin{syntax}
+%     \cs{token_if_expandable_p:N} \meta{token}
+%     \cs{token_if_expandable:NTF} \meta{token}
+%     ~~\Arg{true code} \Arg{false code}
+%   \end{syntax}
+%   Tests if the \meta{token} is expandable. This test returns \meta{false}
+%   for an undefined token.
+%   The branching versions then leave either \meta{true code} or
+%   \meta{false code} in the input stream, as appropriate to the truth of
+%   the test and the variant of the function chosen. The logical truth of
+%   the test is left in the input stream by the predicate version.
+% \end{function}
+%
+% \begin{function}[EXP,pTF]{\token_if_long_macro:N}
+%   \begin{syntax}
+%     \cs{token_if_long_macro_p:N} \meta{token}
+%     \cs{token_if_long_macro:NTF} \meta{token}
+%     ~~\Arg{true code} \Arg{false code}
+%   \end{syntax}
+%   Tests if the \meta{token} is a long macro.
+%   The branching versions then leave either \meta{true code} or
+%   \meta{false code} in the input stream, as appropriate to the truth of
+%   the test and the variant of the function chosen. The logical truth of
+%   the test is left in the input stream by the predicate version.
+% \end{function}
+%
+% \begin{function}[EXP,pTF]{\token_if_protected_macro:N}
+%   \begin{syntax}
+%     \cs{token_if_protected_macro_p:N} \meta{token}
+%     \cs{token_if_protected_macro:NTF} \meta{token}
+%     ~~\Arg{true code} \Arg{false code}
+%   \end{syntax}
+%   Tests if the \meta{token} is a protected macro: a macro which
+%   is both protected and long will return logical false.
+%   The branching versions then leave either \meta{true code} or
+%   \meta{false code} in the input stream, as appropriate to the truth of
+%   the test and the variant of the function chosen. The logical truth of
+%   the test is left in the input stream by the predicate version.
+% \end{function}
+%
+%% \begin{function}[EXP,pTF]{\token_if_protected_long_macro:N}
+%   \begin{syntax}
+%     \cs{token_if_protected_long_macro_p:N} \meta{token}
+%     \cs{token_if_protected_long_macro:NTF} \meta{token}
+%     ~~\Arg{true code} \Arg{false code}
+%   \end{syntax}
+%   Tests if the \meta{token} is a protected long macro.
+%   The branching versions then leave either \meta{true code} or
+%   \meta{false code} in the input stream, as appropriate to the truth of
+%   the test and the variant of the function chosen. The logical truth of
+%   the test is left in the input stream by the predicate version.
+% \end{function}
+%
+% \begin{function}[EXP,pTF]{\token_if_chardef:N}
+%   \begin{syntax}
+%     \cs{token_if_chardef_p:N} \meta{token}
+%     \cs{token_if_chardef:NTF} \meta{token}
+%     ~~\Arg{true code} \Arg{false code}
+%   \end{syntax}
+%   Tests if the \meta{token} is defined to be a chardef.
+%   The branching versions then leave either \meta{true code} or
+%   \meta{false code} in the input stream, as appropriate to the truth of
+%   the test and the variant of the function chosen. The logical truth of
+%   the test is left in the input stream by the predicate version.
+% \end{function}
+%
+% \begin{function}[EXP,pTF]{\token_if_mathchardef:N}
+%   \begin{syntax}
+%     \cs{token_if_mathchardef_p:N} \meta{token}
+%     \cs{token_if_mathchardef:NTF} \meta{token}
+%     ~~\Arg{true code} \Arg{false code}
+%   \end{syntax}
+%   Tests if the \meta{token} is defined to be a mathchardef.
+%   The branching versions then leave either \meta{true code} or
+%   \meta{false code} in the input stream, as appropriate to the truth of
+%   the test and the variant of the function chosen. The logical truth of
+%   the test is left in the input stream by the predicate version.
+% \end{function}
+%
+% \begin{function}[EXP,pTF]{\token_if_dim_register:N}
+%   \begin{syntax}
+%     \cs{token_if_dim_register_p:N} \meta{token}
+%     \cs{token_if_dim_register:NTF} \meta{token}
+%     ~~\Arg{true code} \Arg{false code}
+%   \end{syntax}
+%   Tests if the \meta{token} is defined to be a dimension register.
+%   The branching versions then leave either \meta{true code} or
+%   \meta{false code} in the input stream, as appropriate to the truth of
+%   the test and the variant of the function chosen. The logical truth of
+%   the test is left in the input stream by the predicate version.
+% \end{function}
+%
+% \begin{function}[EXP,pTF]{\token_if_int_register:N}
+%   \begin{syntax}
+%     \cs{token_if_int_register_p:N} \meta{token}
+%     \cs{token_if_int_register:NTF} \meta{token}
+%     ~~\Arg{true code} \Arg{false code}
+%   \end{syntax}
+%   Tests if the \meta{token} is defined to be a integer register.
+%   The branching versions then leave either \meta{true code} or
+%   \meta{false code} in the input stream, as appropriate to the truth of
+%   the test and the variant of the function chosen. The logical truth of
+%   the test is left in the input stream by the predicate version.
+% \end{function}
+%
+% \begin{function}[EXP,pTF]{\token_if_skip_register:N}
+%   \begin{syntax}
+%     \cs{token_if_skip_register_p:N} \meta{token}
+%     \cs{token_if_skip_register:NTF} \meta{token}
+%     ~~\Arg{true code} \Arg{false code}
+%   \end{syntax}
+%   Tests if the \meta{token} is defined to be a skip register.
+%   The branching versions then leave either \meta{true code} or
+%   \meta{false code} in the input stream, as appropriate to the truth of
+%   the test and the variant of the function chosen. The logical truth of
+%   the test is left in the input stream by the predicate version.
+% \end{function}
+%
+% \begin{function}[EXP,pTF]{\token_if_toks_register:N}
+%   \begin{syntax}
+%     \cs{token_if_toks_register_p:N} \meta{token}
+%     \cs{token_if_toks_register:NTF} \meta{token}
+%     ~~\Arg{true code} \Arg{false code}
+%   \end{syntax}
+%   Tests if the \meta{token} is defined to be a toks register
+%   (not used by\LaTeX3).
+%   The branching versions then leave either \meta{true code} or
+%   \meta{false code} in the input stream, as appropriate to the truth of
+%   the test and the variant of the function chosen. The logical truth of
+%   the test is left in the input stream by the predicate version.
+% \end{function}
+%
+% \begin{function}[EXP,pTF]{\token_if_primitive:N}
+%   \begin{syntax}
+%     \cs{token_if_primitive_p:N} \meta{token}
+%     \cs{token_if_primitive:NTF} \meta{token}
+%     ~~\Arg{true code} \Arg{false code}
+%   \end{syntax}
+%   Tests if the \meta{token} is an engine primitive.
+%   The branching versions then leave either \meta{true code} or
+%   \meta{false code} in the input stream, as appropriate to the truth of
+%   the test and the variant of the function chosen. The logical truth of
+%   the test is left in the input stream by the predicate version.
+% \end{function}
+%
+% \section{Peeking ahead at the next token}
+%
+% There is often a need to look ahead at the next token in the input
+% stream while leaving it in place. This is handled using the
+% \enquote{peek} functions. The generic \cs{peek_after:Nw} is
+% provided along with a family of predefined tests for common cases.
+% As peeking ahead does \emph{not} skip spaces the predefined tests
+% include both a space-respecting and space-skipping version.
+%
+% \begin{function}{\peek_after:Nw}
+%   \begin{syntax}
+%     \cs{peek_after:Nw} \meta{function} \meta{token}
+%   \end{syntax}
+%   Locally sets the test variable \cs{l_peek_token} equal to \meta{token}
+%   (as an implicit token, \emph{not} as a token list), and then
+%   expands the \meta{function}. The \meta{token} will remain in
+%   the input stream as the next item after the \meta{function}.
+%   The \meta{token} here may be \verb*| |, |{| or |}| (assuming
+%   normal \TeX{} category codes), \emph{i.e.}~it is not necessarily the
+%   next argument which would be grabbed by a normal function.
+% \end{function}
+%
+% \begin{function}{\peek_gafter:Nw}
+%   \begin{syntax}
+%     \cs{peek_gafter:Nw} \meta{function} \meta{token}
+%   \end{syntax}
+%   Globally sets the test variable \cs{g_peek_token} equal to \meta{token}
+%   (as an implicit token, \emph{not} as a token list), and then
+%   expands the \meta{function}. The \meta{token} will remain in
+%   the input stream as the next item after the \meta{function}.
+%   The \meta{token} here may be \verb*| |, |{| or |}| (assuming
+%   normal \TeX{} category codes), \emph{i.e.}~it is not necessarily the
+%   next argument which would be grabbed by a normal function.
+% \end{function}
+%
+% \begin{variable}{\l_peek_token}
+%  Token set by \cs{peek_after:Nw} and available for testing
+%  as described above.
+% \end{variable}
+%
+% \begin{variable}{\g_peek_token}
+%  Token set by \cs{peek_gafter:Nw} and available for testing
+%  as described above.
+% \end{variable}
+%
+% \begin{function}[TF]{\peek_catcode:N}
+%   \begin{syntax}
+%     \cs{peek_catcode:NTF} \meta{test token} \Arg{true code} \Arg{false code}
+%   \end{syntax}
+%   Tests if the next \meta{token} in the input stream has the same
+%   category code as the \meta{test token} (as defined by the test
+%   \cs{token_if_eq_catcode:NNTF}). Spaces are respected by the test
+%   and the \meta{token} will be left in the input stream after
+%   the \meta{true code} or \meta{false code} (as appropriate to the
+%   result of the test).
+% \end{function}
+%
+% \begin{function}[TF]{\peek_catcode_ignore_spaces:N}
+%   \begin{syntax}
+%     \cs{peek_catcode_ignore_spaces:NTF} \meta{test token}
+%     ~~\Arg{true code} \Arg{false code}
+%   \end{syntax}
+%   Tests if the next \meta{token} in the input stream has the same
+%   category code as the \meta{test token} (as defined by the test
+%   \cs{token_if_eq_catcode:NNTF}). Spaces are ignored by the test
+%   and the \meta{token} will be left in the input stream after
+%   the \meta{true code} or \meta{false code} (as appropriate to the
+%   result of the test).
+% \end{function}
+%
+% \begin{function}[TF]{\peek_catcode_remove:N}
+%   \begin{syntax}
+%     \cs{peek_catcode_remove:NTF} \meta{test token}
+%     ~~\Arg{true code} \Arg{false code}
+%   \end{syntax}
+%   Tests if the next \meta{token} in the input stream has the same
+%   category code as the \meta{test token} (as defined by the test
+%   \cs{token_if_eq_catcode:NNTF}). Spaces are respected by the test
+%   and the \meta{token} will be removed from the input stream if the
+%   test is true. The function will then place either the
+%   \meta{true code} or \meta{false code} in the input stream (as
+%   appropriate to the result of the test).
+% \end{function}
+%
+% \begin{function}[TF]{\peek_catcode_remove_ignore_spaces:N}
+%   \begin{syntax}
+%     \cs{peek_catcode_remove_ignore_spaces:NTF} \meta{test token}
+%     ~~\Arg{true code} \Arg{false code}
+%   \end{syntax}
+%   Tests if the next \meta{token} in the input stream has the same
+%   category code as the \meta{test token} (as defined by the test
+%   \cs{token_if_eq_catcode:NNTF}). Spaces are ignored by the test
+%   and the \meta{token} will be removed from the input stream if the
+%   test is true. The function will then place either the
+%   \meta{true code} or \meta{false code} in the input stream (as
+%   appropriate to the result of the test).
+% \end{function}
+%
+% \begin{function}[TF]{\peek_charcode:N}
+%   \begin{syntax}
+%     \cs{peek_charcode:NTF} \meta{test token}
+%     ~~\Arg{true code} \Arg{false code}
+%   \end{syntax}
+%   Tests if the next \meta{token} in the input stream has the same
+%   character code as the \meta{test token} (as defined by the test
+%   \cs{token_if_eq_charcode:NNTF}). Spaces are respected by the test
+%   and the \meta{token} will be left in the input stream after
+%   the \meta{true code} or \meta{false code} (as appropriate to the
+%   result of the test).
+% \end{function}
+%
+% \begin{function}[TF]{\peek_charcode_ignore_spaces:N}
+%   \begin{syntax}
+%     \cs{peek_charcode_ignore_spaces:NTF} \meta{test token}
+%     ~~\Arg{true code} \Arg{false code}
+%   \end{syntax}
+%   Tests if the next \meta{token} in the input stream has the same
+%   character code as the \meta{test token} (as defined by the test
+%   \cs{token_if_eq_charcode:NNTF}). Spaces are ignored by the test
+%   and the \meta{token} will be left in the input stream after
+%   the \meta{true code} or \meta{false code} (as appropriate to the
+%   result of the test).
+% \end{function}
+%
+% \begin{function}[TF]{\peek_charcode_remove:N}
+%   \begin{syntax}
+%     \cs{peek_charcode_remove:NTF} \meta{test token}
+%     ~~\Arg{true code} \Arg{false code}
+%   \end{syntax}
+%   Tests if the next \meta{token} in the input stream has the same
+%   character code as the \meta{test token} (as defined by the test
+%   \cs{token_if_eq_charcode:NNTF}). Spaces are respected by the test
+%   and the \meta{token} will be removed from the input stream if the
+%   test is true. The function will then place either the
+%   \meta{true code} or \meta{false code} in the input stream (as
+%   appropriate to the result of the test).
+% \end{function}
+%
+% \begin{function}[TF]{\peek_charcode_remove_ignore_spaces:N}
+%   \begin{syntax}
+%     \cs{peek_charcode_remove_ignore_spaces:NTF} \meta{test token}
+%     ~~\Arg{true code} \Arg{false code}
+%   \end{syntax}
+%    Tests if the next \meta{token} in the input stream has the same
+%   character code as the \meta{test token} (as defined by the test
+%   \cs{token_if_eq_charcode:NNTF}). Spaces are ignored by the test
+%   and the \meta{token} will be removed from the input stream if the
+%   test is true. The function will then place either the
+%   \meta{true code} or \meta{false code} in the input stream (as
+%   appropriate to the result of the test).
+% \end{function}
+%
+% \begin{function}[TF]{\peek_meaning:N}
+%   \begin{syntax}
+%     \cs{peek_meaning:NTF} \meta{test token} \Arg{true code} \Arg{false code}
+%   \end{syntax}
+%   Tests if the next \meta{token} in the input stream has the same
+%   meaning as the \meta{test token} (as defined by the test
+%   \cs{token_if_eq_meaning:NNTF}). Spaces are respected by the test
+%   and the \meta{token} will be left in the input stream after
+%   the \meta{true code} or \meta{false code} (as appropriate to the
+%   result of the test).
+% \end{function}
+%
+% \begin{function}[TF]{\peek_meaning_ignore_spaces:N}
+%   \begin{syntax}
+%     \cs{peek_meaning_ignore_spaces:NTF} \meta{test token}
+%     ~~\Arg{true code} \Arg{false code}
+%   \end{syntax}
+%   Tests if the next \meta{token} in the input stream has the same
+%   meaning as the \meta{test token} (as defined by the test
+%   \cs{token_if_eq_meaning:NNTF}). Spaces are ignored by the test
+%   and the \meta{token} will be left in the input stream after
+%   the \meta{true code} or \meta{false code} (as appropriate to the
+%   result of the test).
+% \end{function}
+%
+% \begin{function}[TF]{\peek_meaning_remove:N}
+%   \begin{syntax}
+%     \cs{peek_meaning_remove:NTF} \meta{test token}
+%     ~~\Arg{true code} \Arg{false code}
+%   \end{syntax}
+%   Tests if the next \meta{token} in the input stream has the same
+%   meaning as the \meta{test token} (as defined by the test
+%   \cs{token_if_eq_meaning:NNTF}). Spaces are respected by the test
+%   and the \meta{token} will be removed from the input stream if the
+%   test is true. The function will then place either the
+%   \meta{true code} or \meta{false code} in the input stream (as
+%   appropriate to the result of the test).
+% \end{function}
+%
+% \begin{function}[TF]{\peek_meaning_remove_ignore_spaces:N}
+%   \begin{syntax}
+%     \cs{peek_meaning_remove_ignore_spaces:NTF} \meta{test token}
+%     ~~\Arg{true code} \Arg{false code}
+%   \end{syntax}
+%   Tests if the next \meta{token} in the input stream has the same
+%   meaning as the \meta{test token} (as defined by the test
+%   \cs{token_if_eq_meaning:NNTF}). Spaces are ignored by the test
+%   and the \meta{token} will be removed from the input stream if the
+%   test is true. The function will then place either the
+%   \meta{true code} or \meta{false code} in the input stream (as
+%   appropriate to the result of the test).
+% \end{function}
+%
+% \section{Decomposing a macro definition}
+%
+% These functions decompose \TeX{} macros into their constituent
+% parts: if the \meta{token} passed is not a macro then no decomposition
+% can occur. In the later case, all three functions leave \cs{scan_stop:}
+% in the input stream.
+%
+% \begin{function}[EXP]{\token_get_arg_spec:N}
+%   \begin{syntax}
+%     \cs{token_get_arg_spec:N} \meta{token}
+%   \end{syntax}
+%   If the \meta{token} is a macro, this function will leave
+%   the primitive \TeX{} argument specification in input stream as
+%   a string of tokens of category code $12$ (with spaces having category
+%   code $10$). Thus for example for a token \cs{next} defined by
+%   \begin{verbatim}
+%     \cs_set:Npn \next #1#2 { x #1 y #2 }
+%   \end{verbatim}
+%   will leave |#1#2| in the input stream. If the \meta{token} is
+%   not a macro then \cs{scan_stop:} will be left in the input stream
+%   \begin{texnote}
+%     If the arg~spec. contains the string |->|, then the |spec| function
+%     will produce incorrect results.
+%   \end{texnote}
+% \end{function}
+%
+% \begin{function}[EXP]{\token_get_replacement_text:N}
+%   \begin{syntax}
+%     \cs{token_get_replacement_text:N} \meta{token}
+%   \end{syntax}
+%   If the \meta{token} is a macro, this function will leave
+%   the replacement text in input stream as
+%   a string of tokens of category code $12$ (with spaces having category
+%   code $10$). Thus for example for a token \cs{next} defined by
+%   \begin{verbatim}
+%     \cs_set:Npn \next #1#2 { x #1~y #2 }
+%   \end{verbatim}
+%   will leave \verb|x#1 y#2| in the input stream. If the \meta{token} is
+%   not a macro then \cs{scan_stop:} will be left in the input stream
+% \end{function}
+%
+% \begin{function}[EXP]{\token_get_prefix_spec:N}
+%   \begin{syntax}
+%     \cs{token_get_prefix_spec:N} \meta{token}
+%   \end{syntax}
+%   If the \meta{token} is a macro, this function will leave
+%   the \TeX{} prefixes applicable in input stream as
+%   a string of tokens of category code $12$ (with spaces having category
+%   code $10$). Thus for example for a token \cs{next} defined by
+%   \begin{verbatim}
+%     \cs_set:Npn \next #1#2 { x #1~y #2 }
+%   \end{verbatim}
+%   will leave |\long| in the input stream. If the \meta{token} is
+%   not a macro then \cs{scan_stop:} will be left in the input stream
+% \end{function}
+% 
+% \section{Experimental token functions}
+%
+% \begin{function}{\char_active_set:Npn, \char_active_set:Npx}
+%   \begin{syntax}
+%      \cs{char_active_set:Npn} \meta{char} \meta{parameters} \Arg{code}
+%   \end{syntax}
+%   Makes \meta{char} an active character to expand to \meta{code} as
+%   replacement text.
+%   Within the \meta{code}, the \meta{parameters} (|#1|, |#2|,
+%   \emph{etc.}) will be replaced by those absorbed
+%   This definition is local to the current \TeX{} group.
+% \end{function}
+% 
+% \begin{function}{\char_active_gset:Npn, \char_active_gset:Npx}
+%   \begin{syntax}
+%      \cs{char_active_gset:Npn} \meta{char} \meta{parameters} \Arg{code}
+%   \end{syntax}
+%   Makes \meta{char} an active character to expand to \meta{code} as
+%   replacement text.
+%   Within the \meta{code}, the \meta{parameters} (|#1|, |#2|,
+%   \emph{etc.}) will be replaced by those absorbed
+%   This definition is global.
+% \end{function}
+%
+% \begin{function}{\char_active_set_eq:NN}
+%   \begin{syntax}
+%      \cs{char_active_set_eq:NN} \meta{char} \meta{function}
+%   \end{syntax}
+%   Makes \meta{char} an active character equivalent in meaning to the
+%   \meta{fiucntion} (which may itself be an active character). This
+%   definition is local to the current \TeX{} group.
+% \end{function}
+% 
+% \begin{function}{\char_active_gset_eq:NN}
+%   \begin{syntax}
+%      \cs{char_active_gset_eq:NN} \meta{char} \meta{function}
+%   \end{syntax}
+%   Makes \meta{char} an active character equivalent in meaning to the
+%   \meta{fiucntion} (which may itself be an active character). This
+%   definition is global.
+% \end{function}
+%
+% \end{documentation}
+%
+% \begin{implementation}
+%
+% \section{\pkg{l3token} implementation}
+%
+%    \begin{macrocode}
+%<*initex|package>
+%    \end{macrocode}
+%
+%    \begin{macrocode}
+%<*package>
+\ProvidesExplPackage
+  {\ExplFileName}{\ExplFileDate}{\ExplFileVersion}{\ExplFileDescription}
+\package_check_loaded_expl:
+%</package>
+%    \end{macrocode}
+%
+% \subsection{Character tokens}
+%
+% \begin{macro}{\char_set_catcode:nn}
+% \begin{macro}{\char_value_catcode:n}
+% \begin{macro}{\char_show_value_catcode:n}
+%   Category code changes.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \char_set_catcode:nn #1#2
+  { \tex_catcode:D #1 = \int_eval:w #2 \int_eval_end: }
+\cs_new_nopar:Npn \char_value_catcode:n #1
+  { \tex_the:D \tex_catcode:D \int_eval:w #1\int_eval_end: }
+\cs_new_nopar:Npn \char_show_value_catcode:n #1
+  { \tex_showthe:D \tex_catcode:D \int_eval:w #1 \int_eval_end: }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}
+%   {
+%     \char_set_catcode_escape:N           ,
+%     \char_set_catcode_group_begin:N      ,
+%     \char_set_catcode_group_end:N        ,
+%     \char_set_catcode_math_toggle:N      ,
+%     \char_set_catcode_alignment:N        , 
+%     \char_set_catcode_end_line:N         ,
+%     \char_set_catcode_parameter:N        ,
+%     \char_set_catcode_math_superscript:N ,
+%     \char_set_catcode_math_subscript:N   , 
+%     \char_set_catcode_ignore:N           ,
+%     \char_set_catcode_space:N            , 
+%     \char_set_catcode_letter:N           ,
+%     \char_set_catcode_other:N            , 
+%     \char_set_catcode_active:N           ,
+%     \char_set_catcode_comment:N          , 
+%     \char_set_catcode_invalid:N          
+%   }
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \char_set_catcode_escape:N #1
+  { \char_set_catcode:nn { `#1 } \c_zero }
+\cs_new_protected_nopar:Npn \char_set_catcode_group_begin:N #1
+  { \char_set_catcode:nn { `#1 } \c_one }
+\cs_new_protected_nopar:Npn \char_set_catcode_group_end:N #1
+  { \char_set_catcode:nn { `#1 } \c_two }
+\cs_new_protected_nopar:Npn \char_set_catcode_math_toggle:N #1
+  { \char_set_catcode:nn { `#1 } \c_three }
+\cs_new_protected_nopar:Npn \char_set_catcode_alignment:N #1
+  { \char_set_catcode:nn { `#1 } \c_four }
+\cs_new_protected_nopar:Npn \char_set_catcode_end_line:N #1
+  { \char_set_catcode:nn { `#1 } \c_five }
+\cs_new_protected_nopar:Npn \char_set_catcode_parameter:N #1
+  { \char_set_catcode:nn { `#1 } \c_six }
+\cs_new_protected_nopar:Npn \char_set_catcode_math_superscript:N #1
+  { \char_set_catcode:nn { `#1 } \c_seven }
+\cs_new_protected_nopar:Npn \char_set_catcode_math_subscript:N #1
+  { \char_set_catcode:nn { `#1 } \c_eight }
+\cs_new_protected_nopar:Npn \char_set_catcode_ignore:N #1
+  { \char_set_catcode:nn { `#1 } \c_nine }
+\cs_new_protected_nopar:Npn \char_set_catcode_space:N #1
+  { \char_set_catcode:nn { `#1 } \c_ten }
+\cs_new_protected_nopar:Npn \char_set_catcode_letter:N #1
+  { \char_set_catcode:nn { `#1 } \c_eleven }
+\cs_new_protected_nopar:Npn \char_set_catcode_other:N #1
+  { \char_set_catcode:nn { `#1 } \c_twelve }
+\cs_new_protected_nopar:Npn \char_set_catcode_active:N #1
+  { \char_set_catcode:nn { `#1 } \c_thirteen }
+\cs_new_protected_nopar:Npn \char_set_catcode_comment:N #1
+  { \char_set_catcode:nn { `#1 } \c_fourteen }
+\cs_new_protected_nopar:Npn \char_set_catcode_invalid:N #1
+  { \char_set_catcode:nn { `#1 } \c_fifteen }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}
+%   {
+%     \char_set_catcode_escape:n           ,
+%     \char_set_catcode_group_begin:n      ,
+%     \char_set_catcode_group_end:n        ,
+%     \char_set_catcode_math_toggle:n      ,
+%     \char_set_catcode_alignment:n        , 
+%     \char_set_catcode_end_line:n         ,
+%     \char_set_catcode_parameter:n        ,
+%     \char_set_catcode_math_superscript:n ,
+%     \char_set_catcode_math_subscript:n   , 
+%     \char_set_catcode_ignore:n           ,
+%     \char_set_catcode_space:n            , 
+%     \char_set_catcode_letter:n           ,
+%     \char_set_catcode_other:n            , 
+%     \char_set_catcode_active:n           ,
+%     \char_set_catcode_comment:n          , 
+%     \char_set_catcode_invalid:n          
+%   }
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \char_set_catcode_escape:n #1
+  { \char_set_catcode:nn {#1} \c_zero }
+\cs_new_protected_nopar:Npn \char_set_catcode_group_begin:n #1
+  { \char_set_catcode:nn {#1} \c_one }
+\cs_new_protected_nopar:Npn \char_set_catcode_group_end:n #1
+  { \char_set_catcode:nn {#1} \c_two }
+\cs_new_protected_nopar:Npn \char_set_catcode_math_toggle:n #1
+  { \char_set_catcode:nn {#1} \c_three }
+\cs_new_protected_nopar:Npn \char_set_catcode_alignment:n #1
+  { \char_set_catcode:nn {#1} \c_four }
+\cs_new_protected_nopar:Npn \char_set_catcode_end_line:n #1
+  { \char_set_catcode:nn {#1} \c_five }
+\cs_new_protected_nopar:Npn \char_set_catcode_parameter:n #1
+  { \char_set_catcode:nn {#1} \c_six }
+\cs_new_protected_nopar:Npn \char_set_catcode_math_superscript:n #1
+  { \char_set_catcode:nn {#1} \c_seven }
+\cs_new_protected_nopar:Npn \char_set_catcode_math_subscript:n #1
+  { \char_set_catcode:nn {#1} \c_eight }
+\cs_new_protected_nopar:Npn \char_set_catcode_ignore:n #1
+  { \char_set_catcode:nn {#1} \c_nine }
+\cs_new_protected_nopar:Npn \char_set_catcode_space:n #1
+  { \char_set_catcode:nn {#1} \c_ten }
+\cs_new_protected_nopar:Npn \char_set_catcode_letter:n #1
+  { \char_set_catcode:nn {#1} \c_eleven }
+\cs_new_protected_nopar:Npn \char_set_catcode_other:n #1
+  { \char_set_catcode:nn {#1} \c_twelve }
+\cs_new_protected_nopar:Npn \char_set_catcode_active:n #1
+  { \char_set_catcode:nn {#1} \c_thirteen }
+\cs_new_protected_nopar:Npn \char_set_catcode_comment:n #1
+  { \char_set_catcode:nn {#1} \c_fourteen }
+\cs_new_protected_nopar:Npn \char_set_catcode_invalid:n #1
+  { \char_set_catcode:nn {#1} \c_fifteen }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\char_set_mathcode:nn}
+% \begin{macro}{\char_value_mathcode:n}
+% \begin{macro}{\char_show_value_mathcode:n}
+% \begin{macro}{\char_set_lccode:nn}
+% \begin{macro}{\char_value_lccode:n}
+% \begin{macro}{\char_show_value_lccode:n}
+% \begin{macro}{\char_set_uccode:nn}
+% \begin{macro}{\char_value_uccode:n}
+% \begin{macro}{\char_show_value_uccode:n}
+% \begin{macro}{\char_set_sfcode:nn}
+% \begin{macro}{\char_value_sfcode:n}
+% \begin{macro}{\char_show_value_sfcode:n}
+%   Pretty repetitive, but necessary!
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \char_set_mathcode:nn #1#2
+  { \tex_mathcode:D #1 = \int_eval:w #2 \int_eval_end: }
+\cs_new_nopar:Npn \char_value_mathcode:n #1
+  { \tex_the:D \tex_mathcode:D \int_eval:w #1\int_eval_end: }
+\cs_new_nopar:Npn \char_show_value_mathcode:n #1
+  { \tex_showthe:D \tex_mathcode:D \int_eval:w #1 \int_eval_end: }
+\cs_new_protected_nopar:Npn \char_set_lccode:nn #1#2
+  { \tex_lccode:D #1 = \int_eval:w #2 \int_eval_end: }
+\cs_new_nopar:Npn \char_value_lccode:n #1
+  { \tex_the:D \tex_lccode:D \int_eval:w #1\int_eval_end: }
+\cs_new_nopar:Npn \char_show_value_lccode:n #1
+  { \tex_showthe:D \tex_lccode:D \int_eval:w #1 \int_eval_end: }
+\cs_new_protected_nopar:Npn \char_set_uccode:nn #1#2
+  { \tex_uccode:D #1 = \int_eval:w #2 \int_eval_end: }
+\cs_new_nopar:Npn \char_value_uccode:n #1
+  { \tex_the:D \tex_uccode:D \int_eval:w #1\int_eval_end: }
+\cs_new_nopar:Npn \char_show_value_uccode:n #1
+  { \tex_showthe:D \tex_uccode:D \int_eval:w #1 \int_eval_end: }
+\cs_new_protected_nopar:Npn \char_set_sfcode:nn #1#2
+  { \tex_sfcode:D #1 = \int_eval:w #2 \int_eval_end: }
+\cs_new_nopar:Npn \char_value_sfcode:n #1
+  { \tex_the:D \tex_sfcode:D \int_eval:w #1\int_eval_end: }
+\cs_new_nopar:Npn \char_show_value_sfcode:n #1
+  { \tex_showthe:D \tex_sfcode:D \int_eval:w #1 \int_eval_end: }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \subsection{Generic tokens}
+%
+% \begin{macro}{\token_new:Nn}
+%  Creates a new token.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \token_new:Nn #1#2 { \cs_new_eq:NN #1 #2 }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}
+%   {
+%     \c_group_begin_token,
+%     \c_group_end_token,
+%     \c_math_toggle_token,
+%     \c_alignment_token,
+%     \c_parameter_token,
+%     \c_math_superscript_token,
+%     \c_math_subscript_token,
+%     \c_space_token,
+%     \c_catcode_letter_token,
+%     \c_catcode_other_token
+%   }
+%    We define these useful tokens. We have to do it by hand with the
+%    brace tokens for obvious reasons.
+%    \begin{macrocode}
+\cs_new_eq:NN \c_group_begin_token {
+\cs_new_eq:NN \c_group_end_token }
+\group_begin:
+  \char_set_catcode_math_toggle:N \*
+  \token_new:Nn \c_math_toggle_token { * }
+  \char_set_catcode_alignment:N \*
+  \token_new:Nn \c_alignment_token { * }
+  \token_new:Nn \c_parameter_token { # }
+  \token_new:Nn \c_math_superscript_token { ^ }
+  \char_set_catcode_math_subscript:N \*
+  \token_new:Nn \c_math_subscript_token { * }
+  \token_new:Nn \c_space_token { ~ }
+  \token_new:Nn \c_catcode_letter_token { a }
+  \token_new:Nn \c_catcode_other_token { 1 }
+\group_end:
+%    \end{macrocode}
+% \end{macro}
+% 
+% \begin{variable}{\c_catcode_active_tl}
+%   Not an implicit token!
+%    \begin{macrocode}
+\group_begin:
+  \char_set_catcode_active:N \*
+  \cs_new_nopar:Npn \c_catcode_active_tl { \exp_not:N * }
+\group_end:
+%    \end{macrocode}
+% \end{variable}
+%
+% \subsection{Token conditionals}
+%
+% \begin{macro}[pTF]{\token_if_group_begin:N}
+%   Check if token is a begin group token. We use the constant
+%   |\c_group_begin_token| for this.
+%    \begin{macrocode}
+\prg_new_conditional:Npnn \token_if_group_begin:N #1 { p , T ,  F , TF }
+  {
+    \if_catcode:w \exp_not:N #1 \c_group_begin_token
+      \prg_return_true: \else: \prg_return_false: \fi:
+  }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}[pTF]{\token_if_group_end:N}
+%   Check if token is a end group token. We use the constant
+%   |\c_group_end_token| for this.
+%    \begin{macrocode}
+\prg_new_conditional:Npnn \token_if_group_end:N #1 { p , T ,  F , TF }
+  {
+    \if_catcode:w \exp_not:N #1 \c_group_end_token
+      \prg_return_true: \else: \prg_return_false: \fi:
+  }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}[pTF]{\token_if_math_toggle:N}
+%   Check if token is a math shift token. We use the constant
+%   |\c_math_toggle_token| for this.
+%    \begin{macrocode}
+\prg_new_conditional:Npnn \token_if_math_toggle:N #1 { p , T ,  F , TF }
+  {
+    \if_catcode:w \exp_not:N #1 \c_math_toggle_token
+      \prg_return_true: \else: \prg_return_false: \fi:
+  }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}[pTF]{\token_if_alignment:N}
+%   Check if token is an alignment tab token. We use the constant
+%   |\c_alignment_tab_token| for this.
+%    \begin{macrocode}
+\prg_new_conditional:Npnn \token_if_alignment:N #1 { p , T ,  F , TF }
+  {
+    \if_catcode:w \exp_not:N #1 \c_alignment_token
+      \prg_return_true: \else: \prg_return_false: \fi:
+  }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}[pTF]{\token_if_parameter:N}
+%   Check if token is a parameter token. We use the constant
+%   |\c_parameter_token| for this. We have to trick \TeX{} a bit to
+%   avoid an error message: within a group we prevent
+%   \cs{c_parameter_token} from behaving like a macro parameter character.
+%   The definitions of \cs{prg_new_conditional:Npnn} are global, so they
+%   will remain after the group.
+%    \begin{macrocode}
+\group_begin:
+\cs_set_eq:NN \c_parameter_token \scan_stop:
+\prg_new_conditional:Npnn \token_if_parameter:N #1 { p , T ,  F , TF }
+  {
+    \if_catcode:w \exp_not:N #1 \c_parameter_token 
+      \prg_return_true: \else: \prg_return_false: \fi:
+  }
+\group_end:
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}[pTF]{\token_if_math_superscript:N}
+%   Check if token is a math superscript token. We use the constant
+%   |\c_superscript_token| for this.
+%    \begin{macrocode}
+\prg_new_conditional:Npnn \token_if_math_superscript:N #1 { p , T ,  F , TF }
+  {
+    \if_catcode:w \exp_not:N #1 \c_math_superscript_token
+      \prg_return_true: \else: \prg_return_false: \fi:
+  }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}[pTF]{\token_if_math_subscript:N}
+%   Check if token is a math subscript token. We use the constant
+%   |\c_subscript_token| for this.
+%    \begin{macrocode}
+\prg_new_conditional:Npnn \token_if_math_subscript:N #1 { p , T ,  F , TF }
+  {
+    \if_catcode:w \exp_not:N #1 \c_math_subscript_token
+      \prg_return_true: \else: \prg_return_false: \fi:
+  }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}[pTF]{\token_if_space:N}
+%   Check if token is a space token. We use the constant
+%   |\c_space_token| for this.
+%    \begin{macrocode}
+\prg_new_conditional:Npnn \token_if_space:N #1 { p , T ,  F , TF }
+  {
+    \if_catcode:w \exp_not:N #1 \c_space_token
+      \prg_return_true: \else: \prg_return_false: \fi:
+  }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}[pTF]{\token_if_letter:N}
+%   Check if token is a letter token. We use the constant
+%   |\c_letter_token| for this.
+%    \begin{macrocode}
+\prg_new_conditional:Npnn \token_if_letter:N #1 { p , T ,  F , TF }
+  {
+    \if_catcode:w \exp_not:N #1 \c_catcode_letter_token
+      \prg_return_true: \else: \prg_return_false: \fi:
+  }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}[pTF]{\token_if_other:N}
+%   Check if token is an other char token. We use the constant
+%   |\c_other_char_token| for this.
+%    \begin{macrocode}
+\prg_new_conditional:Npnn \token_if_other:N #1 { p , T ,  F , TF }
+  {
+    \if_catcode:w \exp_not:N #1 \c_catcode_other_token
+      \prg_return_true: \else: \prg_return_false: \fi:
+  }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}[pTF]{\token_if_active:N}
+%   Check if token is an active char token. We use the constant
+%   |\c_active_char_tl| for this. A technical point is that
+%   \cs{c_active_char_tl} is in fact a macro expanding to
+%   |\exp_not:N *|, where |*| is active.
+%    \begin{macrocode}
+\prg_new_conditional:Npnn \token_if_active:N #1 { p , T ,  F , TF }
+  {
+    \if_catcode:w \exp_not:N #1 \c_catcode_active_tl
+      \prg_return_true: \else: \prg_return_false: \fi:
+  }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}[pTF]{\token_if_eq_meaning:NN}
+%   Check if the tokens |#1| and |#2| have same meaning.
+%    \begin{macrocode}
+\prg_new_conditional:Npnn \token_if_eq_meaning:NN #1#2 { p , T ,  F , TF }
+  {
+    \if_meaning:w  #1  #2
+      \prg_return_true: \else: \prg_return_false: \fi:
+  }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}[pTF]{\token_if_eq_catcode:NN}
+%  Check if the tokens |#1| and |#2| have same category code.
+%    \begin{macrocode}
+\prg_new_conditional:Npnn \token_if_eq_catcode:NN #1#2 { p , T ,  F , TF }
+  {
+    \if_catcode:w \exp_not:N #1 \exp_not:N #2
+      \prg_return_true: \else: \prg_return_false: \fi:
+  }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}[pTF]{\token_if_eq_charcode:NN}
+%  Check if the tokens |#1| and |#2| have same character code.
+%    \begin{macrocode}
+\prg_new_conditional:Npnn \token_if_eq_charcode:NN #1#2 { p , T ,  F , TF }
+  {
+    \if_charcode:w \exp_not:N #1 \exp_not:N #2
+      \prg_return_true: \else: \prg_return_false: \fi:
+  }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}[pTF]{\token_if_macro:N}
+% \begin{macro}[aux]{\token_if_macro_p_aux:w}
+%   When a token is a macro, |\token_to_meaning:N| will always output
+%   something like |\long macro:#1->#1| so we could naively check to
+%   see if the meaning contains |->|. However, this can fail the five
+%   \cs{tex_...mark:D} primitives, whose meaning has the form
+%   |...mark:|\meta{user material}. The problem is that the
+%   \meta{user material} can contain |->|.
+%
+%   However, only characters, macros, and marks can contain the colon
+%   character. The idea is thus to grab until the first |:|, and analyse
+%   what is left. However, macros can have any combination of |\long|,
+%   |\protected| or |\outer| (not used in \LaTeX3) before the string
+%   |macro:|. We thus only select the part of the meaning between
+%   the first |ma| and the first following |:|. If this string is
+%   |cro|, then we have a macro. If the string is |rk|, then we have
+%   a mark. The string can also be |cro parameter character | for a
+%   colon with a weird category code (namely the usual category code
+%   of |#|). Otherwise, it is empty.
+%
+%   This relies on the fact that |\long|, |\protected|, |\outer|
+%   cannot contain |ma|, regardless of the escape character, even if
+%   the escape character is |m|\ldots{}
+%
+%   Both |ma| and |:| must be of category code $12$ (other), and we
+%   achieve using the standard lowercasing technique.
+%
+%    \begin{macrocode}
+\group_begin:
+\char_set_catcode_other:N \M
+\char_set_catcode_other:N \A
+\char_set_lccode:nn { `\; } { `\: }
+\char_set_lccode:nn { `\T } { `\T }
+\char_set_lccode:nn { `\F } { `\F }
+\tl_to_lowercase:n
+  {
+    \group_end:
+    \prg_new_conditional:Npnn \token_if_macro:N #1 { p , T ,  F , TF }
+      {
+        \exp_after:wN \token_if_macro_p_aux:w
+        \token_to_meaning:N #1 MA; \q_stop
+      }
+    \cs_new_nopar:Npn \token_if_macro_p_aux:w #1 MA #2 ; #3 \q_stop
+      {
+        \if_int_compare:w \pdftex_strcmp:D { #2 } { cro } = \c_zero
+            \prg_return_true:
+        \else:
+            \prg_return_false:
+        \fi:
+      }
+  }
+%    \end{macrocode}
+%  \end{macro}
+%  \end{macro}
+%
+% \begin{macro}[pTF]{\token_if_cs:N}
+%   Check if token has same catcode as a control sequence. This
+%   follows the same pattern as for \cs{token_if_letter:N} \emph{etc.}
+%   We use |\scan_stop:| for this.
+%    \begin{macrocode}
+\prg_new_conditional:Npnn \token_if_cs:N #1 { p , T ,  F , TF }
+  {
+    \if_catcode:w \exp_not:N #1 \scan_stop:
+      \prg_return_true: \else: \prg_return_false: \fi:
+  }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}[pTF]{\token_if_expandable:N}
+%   Check if token is expandable. We use the fact that \TeX{} will
+%   temporarily convert |\exp_not:N| \meta{token} into |\scan_stop:| if
+%   \meta{token} is expandable.
+%    \begin{macrocode}
+\prg_new_conditional:Npnn \token_if_expandable:N #1 { p , T ,  F , TF }
+  {
+    \cs_if_exist:NTF #1
+      {
+        \exp_after:wN \if_meaning:w \exp_not:N #1 #1
+          \prg_return_false: \else: \prg_return_true: \fi:
+      }
+      { \prg_return_false: }
+  }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}[pTF]{\token_if_chardef:N,\token_if_mathchardef:N,
+%   \token_if_long_macro:N,             \token_if_protected_macro:N,
+%   \token_if_protected_long_macro:N,   \token_if_dim_register:N,
+%   \token_if_skip_register:N,          \token_if_int_register:N,
+%   \token_if_toks_register:N}
+% \begin{macro}[aux]{
+%               \token_if_chardef_p_aux:w,
+%               \token_if_mathchardef_p_aux:w,
+%               \token_if_int_register_p_aux:w,
+%               \token_if_skip_register_p_aux:w,
+%               \token_if_dim_register_p_aux:w,
+%               \token_if_toks_register_p_aux:w,
+%               \token_if_protected_macro_p_aux:w,
+%               \token_if_long_macro_p_aux:w,
+%               \token_if_protected_long_macro_p_aux:w}
+%   Most of these functions have to check the meaning of the token in
+%   question so we need to do some checkups on which characters are
+%   output by |\token_to_meaning:N|. As usual, these characters have
+%   catcode 12 so we must do some serious substitutions in the code
+%   below\dots
+%    \begin{macrocode}
+\group_begin:
+  \char_set_lccode:nn { `\T } { `\T }
+  \char_set_lccode:nn { `\F } { `\F }
+  \char_set_lccode:nn { `\X } { `\n }
+  \char_set_lccode:nn { `\Y } { `\t }
+  \char_set_lccode:nn { `\Z } { `\d }
+  \char_set_lccode:nn { `\? } { `\\ }
+  \tl_map_inline:nn { \X \Y \Z \M \C \H \A \R \O \U \S \K \I \P \L \G \P \E }
+    { \char_set_catcode:nn { `#1 } \c_twelve }
+%    \end{macrocode}
+%   We convert the token list to lower case and restore the catcode and
+%   lowercase code changes.
+%    \begin{macrocode}
+\tl_to_lowercase:n
+  {
+    \group_end:
+%    \end{macrocode}
+%   First up is checking if something has been defined with
+%   |\tex_chardef:D| or |\tex_mathchardef:D|. This is easy since \TeX{}
+%   thinks of such tokens as hexadecimal so it stores them as
+%   |\char"|\meta{hex~number} or |\mathchar"|\meta{hex~number}.
+%    \begin{macrocode}
+    \prg_new_conditional:Npnn \token_if_chardef:N #1 { p , T ,  F , TF }
+      {
+        \exp_after:wN \token_if_chardef_aux:w
+          \token_to_meaning:N #1 ?CHAR" \q_stop
+      }
+    \cs_new_nopar:Npn \token_if_chardef_aux:w #1 ?CHAR" #2 \q_stop
+      { \tl_if_empty:nTF {#1} { \prg_return_true: } { \prg_return_false: } }
+%    \end{macrocode}
+%
+%    \begin{macrocode}
+    \prg_new_conditional:Npnn \token_if_mathchardef:N #1 { p , T ,  F , TF }
+      {
+        \exp_after:wN \token_if_mathchardef_aux:w
+          \token_to_meaning:N #1 ?MAYHCHAR" \q_stop
+      }
+    \cs_new_nopar:Npn \token_if_mathchardef_aux:w #1 ?MAYHCHAR" #2 \q_stop
+      { \tl_if_empty:nTF {#1} { \prg_return_true: } { \prg_return_false: } }
+%    \end{macrocode}
+%   Integer registers are a little more difficult since they expand to
+%   |\count|\meta{number} and there is also a primitive |\countdef|. So
+%   we have to check for that primitive as well.
+%    \begin{macrocode}
+    \prg_new_conditional:Npnn \token_if_int_register:N #1 { p , T ,  F , TF }
+      {
+        \if_meaning:w \tex_countdef:D #1
+          \prg_return_false:
+        \else:
+          \exp_after:wN \token_if_int_register_aux:w
+            \token_to_meaning:N #1 ?COUXY \q_stop
+        \fi:
+      }
+    \cs_new_nopar:Npn \token_if_int_register_aux:w #1 ?COUXY #2 \q_stop
+      { \tl_if_empty:nTF {#1} { \prg_return_true: } { \prg_return_false: } }
+%    \end{macrocode}
+%   Skip registers are done the same way as the integer registers.
+%    \begin{macrocode}
+    \prg_new_conditional:Npnn \token_if_skip_register:N #1 { p , T ,  F , TF }
+      {
+        \if_meaning:w \tex_skipdef:D #1
+          \prg_return_false:
+        \else:
+          \exp_after:wN \token_if_skip_register_aux:w
+            \token_to_meaning:N #1?SKIP\q_stop
+        \fi:
+      }
+    \cs_new_nopar:Npn \token_if_skip_register_aux:w #1 ?SKIP #2 \q_stop
+      { \tl_if_empty:nTF {#1} { \prg_return_true: } { \prg_return_false: } }
+%    \end{macrocode}
+%   Dim registers. No news here
+%    \begin{macrocode}
+    \prg_new_conditional:Npnn \token_if_dim_register:N #1 { p , T ,  F , TF }
+      {
+        \if_meaning:w \tex_dimendef:D #1
+          \c_false_bool
+        \else:
+          \exp_after:wN \token_if_dim_register_aux:w
+            \token_to_meaning:N #1 ?ZIMEX \q_stop
+        \fi:
+      }
+    \cs_new_nopar:Npn \token_if_dim_register_aux:w #1 ?ZIMEX #2 \q_stop
+      { \tl_if_empty:nTF {#1} { \prg_return_true: } { \prg_return_false: } }
+%    \end{macrocode}
+%   Toks registers.
+%    \begin{macrocode}
+    \prg_new_conditional:Npnn \token_if_toks_register:N #1 { p , T ,  F , TF }
+      {
+        \if_meaning:w \tex_toksdef:D #1
+          \prg_return_false:
+        \else:
+          \exp_after:wN \token_if_toks_register_aux:w
+            \token_to_meaning:N #1 ?YOKS \q_stop
+        \fi:
+      }
+     \cs_new_nopar:Npn \token_if_toks_register_aux:w #1 ?YOKS #2 \q_stop
+       { \tl_if_empty:nTF {#1} { \prg_return_true: } { \prg_return_false: } }
+%    \end{macrocode}
+%   Protected macros.
+%    \begin{macrocode}
+    \prg_new_conditional:Npnn \token_if_protected_macro:N #1
+      { p , T ,  F , TF }
+      {
+        \exp_after:wN \token_if_protected_macro_aux:w
+          \token_to_meaning:N #1 ?PROYECYEZ~MACRO \q_stop
+      }
+    \cs_new_nopar:Npn \token_if_protected_macro_aux:w
+      #1 ?PROYECYEZ~MACRO #2 \q_stop
+      { \tl_if_empty:nTF {#1} { \prg_return_true: } { \prg_return_false: } }
+%    \end{macrocode}
+%   Long macros.
+%    \begin{macrocode}
+    \prg_new_conditional:Npnn \token_if_long_macro:N #1 { p , T ,  F , TF }
+      {
+        \exp_after:wN \token_if_long_macro_aux:w
+          \token_to_meaning:N #1 ?LOXG~MACRO \q_stop
+      }
+    \cs_new_nopar:Npn \token_if_long_macro_aux:w #1 ?LOXG~MACRO #2 \q_stop
+      { \tl_if_empty:nTF {#1} { \prg_return_true: } { \prg_return_false: } }
+%    \end{macrocode}
+%   Finally protected long macros where we for once don't have to add an
+%   extra test since there is no primitive for the combined prefixes.
+%    \begin{macrocode}
+    \prg_new_conditional:Npnn \token_if_protected_long_macro:N #1
+      { p , T ,  F , TF }
+      {
+        \exp_after:wN \token_if_protected_long_macro_aux:w
+          \token_to_meaning:N #1 ?PROYECYEZ?LOXG~MACRO \q_stop
+      }
+    \cs_new_nopar:Npn \token_if_protected_long_macro_aux:w
+      #1 ?PROYECYEZ?LOXG~MACRO #2 \q_stop
+      { \tl_if_empty:nTF {#1} { \prg_return_true: } { \prg_return_false: } }
+%    \end{macrocode}
+% Finally the |\tl_to_lowercase:n| ends!
+%    \begin{macrocode}
+  }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}[pTF]{\token_if_primitive:N}
+% \begin{macro}[aux]{\token_if_primitive_aux:NNw,
+%     \token_if_primitive_aux_space:w,
+%     \token_if_primitive_aux_nullfont:N,
+%     \token_if_primitive_aux_loop:N,
+%     \token_if_primitive_auxii:Nw,
+%     \token_if_primitive_aux_undefined:N}
+%^^A See http://groups.google.com/group/comp.text.tex/browse_thread/thread/0a72666873f8753d#
+%
+%   We filter out macros first, because they cause endless trouble later
+%   otherwise.
+%
+%   Primitives are almost distinguished by the fact that the result
+%   of \cs{token_to_meaning:N} is formed from letters only. Every other
+%   token has either a space (e.g., |the letter A|), a digit
+%   (e.g., |\count123|) or a double quote (e.g., |\char"A|).
+%
+%   Ten exceptions: on the one hand, \cs{c_undefined:D} is not a
+%   primitive, but its meaning is |undefined|, only letters;
+%   on the other hand, \cs{tex_space:D}, \cs{tex_italiccorr:D},
+%   \cs{tex_hyphen:D}, \cs{tex_firstmark:D}, \cs{tex_topmark:D},
+%   \cs{tex_botmark:D}, \cs{tex_splitfirstmark:D}, \cs{tex_splitbotmark:D},
+%   and \cs{tex_nullfont:D} are primitives, but have non-letters
+%   in their meaning.
+%
+%   We start by removing the two first (non-space) characters from
+%   the meaning. This removes the escape character (which may be
+%   inexistent depending on \cs{tex_endlinechar:D}), and takes care
+%   of three of the exceptions: \cs{tex_space:D}, \cs{tex_italiccorr:D}
+%   and \cs{tex_hyphen:D}, whose meaning is at most two characters.
+%   This leaves a string terminated by some |:|, and \cs{q_stop}. 
+%
+%   The meaning of each one of the five \cs{tex_...mark:D} primitives
+%   has the form \meta{letters}|:|\meta{user material}. In other words,
+%   the first non-letter is a colon. We remove everything after the first
+%   colon.
+%
+%   We are now left with a string, which we must analyze. For primitives,
+%   it contains only letters. For non-primitives, it contains either
+%   |"|, or a space, or a digit. Two exceptions remain: \cs{c_undefined:D},
+%   which is not a primitive, and \cs{tex_nullfont:D}, which is a primitive.
+%
+%   Spaces cannot be grabbed in an undelimited way, so we check them
+%   separately. If there is a space, we test for \cs{tex_nullfont:D}.
+%   Otherwise, we go through characters one by one, and stop at the
+%   first character less than |`A| (this is not quite a test for
+%   \enquote{only letters}, but is close enough to work in this context).
+%   If this first character is |:| then we have a primitive, or
+%   \cs{c_undefined:D}, and if it is |"| or a digit, then the token
+%   is not a primitive.
+%
+%    \begin{macrocode}
+\tex_chardef:D \c_token_A_int = `A ~ %
+\group_begin:
+\char_set_catcode_other:N \;
+\char_set_lccode:nn { `\; } { `\: }
+\char_set_lccode:nn { `\T } { `\T }
+\char_set_lccode:nn { `\F } { `\F }
+\tl_to_lowercase:n {
+  \group_end:
+  \prg_new_conditional:Npnn \token_if_primitive:N #1 { p , T , F , TF }
+    {
+      \token_if_macro:NTF #1
+        \prg_return_false:
+        {
+          \exp_after:wN \token_if_primitive_aux:NNw
+          \token_to_meaning:N #1 ; ; ; \q_stop #1
+        }
+    }
+  \cs_new_nopar:Npn \token_if_primitive_aux:NNw #1#2 #3 ; #4 \q_stop
+    {
+      \tl_if_empty:oTF { \token_if_primitive_aux_space:w #3 ~ }
+        { \token_if_primitive_aux_loop:N #3 ; \q_stop }
+        { \token_if_primitive_aux_nullfont:N }
+    }
+}
+\cs_new_nopar:Npn \token_if_primitive_aux_space:w #1 ~ { }
+\cs_new:Npn \token_if_primitive_aux_nullfont:N #1
+  {
+    \if_meaning:w \tex_nullfont:D #1
+      \prg_return_true:
+    \else:
+      \prg_return_false:
+    \fi:
+  }
+\cs_new_nopar:Npn \token_if_primitive_aux_loop:N #1
+  {
+    \if_num:w `#1 < \c_token_A_int %
+      \exp_after:wN \token_if_primitive_auxii:Nw
+      \exp_after:wN #1
+    \else:
+      \exp_after:wN \token_if_primitive_aux_loop:N
+    \fi:
+  }
+\cs_new_nopar:Npn \token_if_primitive_auxii:Nw #1 #2 \q_stop
+  {
+    \if:w : #1
+      \exp_after:wN \token_if_primitive_aux_undefined:N
+    \else:
+      \prg_return_false:
+      \exp_after:wN \use_none:n
+    \fi:
+  }
+\cs_new:Npn \token_if_primitive_aux_undefined:N #1
+  {
+    \if_cs_exist:N #1
+      \prg_return_true:
+    \else:
+      \prg_return_false:
+    \fi:
+  }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \subsection{Peeking ahead at the next token}
+%
+% Peeking ahead is implemented using a two part mechanism. The
+% outer level provides a defined interface to the lower level material.
+% This allows a large amount of code to be shared. There are four
+% cases:
+% \begin{enumerate}
+%   \item peek at the next token;
+%   \item peek at the next non-space token;
+%   \item peek at the next token and remove it;
+%   \item peek at the next non-space token and remove it.
+% \end{enumerate}
+%
+% \begin{variable}{\l_peek_token}
+% \begin{variable}{\g_peek_token}
+%   Storage tokens which are publicly documented: the token peeked.
+%    \begin{macrocode}
+\cs_new_eq:NN \l_peek_token ?
+\cs_new_eq:NN \g_peek_token ?
+%    \end{macrocode}
+% \end{variable}
+% \end{variable}
+%
+% \begin{variable}{\l_peek_search_token}
+%   The token to search for as an implicit token:
+%   \emph{cf.}~\cs{l_peek_search_tl}.
+%    \begin{macrocode}
+\cs_new_eq:NN \l_peek_search_token ?
+%    \end{macrocode}
+% \end{variable}
+%
+% \begin{variable}{\l_peek_search_tl}
+%   The token to search for as an explicit token:
+%   \emph{cf.}~\cs{l_peek_search_token}.
+%    \begin{macrocode}
+\cs_new_nopar:Npn \l_peek_search_tl { }
+%    \end{macrocode}
+% \end{variable}
+%
+% \begin{macro}[aux]
+%   {\peek_true:w, \peek_true_aux:w, \peek_false:w, \peek_tmp:w}
+%   Functions used by the branching and space-stripping code.
+%    \begin{macrocode}
+\cs_new_nopar:Npn \peek_true:w  { }
+\cs_new_nopar:Npn \peek_true_aux:w  { }
+\cs_new_nopar:Npn \peek_false:w { }
+\cs_new:Npn \peek_tmp:w { }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\peek_after:Nw}
+% \begin{macro}{\peek_after:Nw}
+%   Simple wrappers for \cs{tex_futurelet:D}: no arguments absorbed
+%   here.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \peek_after:Nw
+  { \tex_futurelet:D \l_peek_token }
+\cs_new_protected_nopar:Npn \peek_gafter:Nw
+  { \pref_global:D \tex_futurelet:D \g_peek_token }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}[aux]{\peek_true_remove:w}
+%   A function to remove the next token and then regain control.
+%    \begin{macrocode}
+\cs_new_protected:Npn \peek_true_remove:w
+  {
+    \group_align_safe_end:
+    \tex_afterassignment:D \peek_true_aux:w
+    \cs_set_eq:NN \peek_tmp:w
+  }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}[TF]{\peek_token_generic:NN}
+%   The generic function stores the test token in both implicit and
+%   explicit modes, and the \texttt{true} and \texttt{false} code as
+%   token lists, more or less. The two branches have to be absorbed here
+%   as the input stream needs to be cleared for the peek function itself.
+%    \begin{macrocode}
+\cs_new_protected:Npn \peek_token_generic:NNTF #1#2#3#4
+  {
+    \cs_set_eq:NN \l_peek_search_token #2
+    \tl_set:Nn \l_peek_search_tl {#2}
+    \cs_set_nopar:Npx \peek_true:w
+      {
+        \exp_not:N \group_align_safe_end:
+        \exp_not:n {#3}
+      }
+    \cs_set_nopar:Npx \peek_false:w
+      {
+        \exp_not:N \group_align_safe_end:
+        \exp_not:n {#4}
+      }
+    \group_align_safe_begin:
+      \peek_after:Nw #1
+  }
+\cs_new_protected:Npn \peek_token_generic:NNT #1#2#3
+  { \peek_token_generic:NNTF #1 #2 {#3} { } }
+\cs_new_protected:Npn \peek_token_generic:NNF #1#2#3
+  { \peek_token_generic:NNTF #1 #2 { } {#3} }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}[TF]{\peek_token_remove_generic:NN}
+%   For token removal there needs to be a call to the auxiliary
+%   function which does the work.
+%    \begin{macrocode}
+\cs_new_protected:Npn \peek_token_remove_generic:NNTF #1#2#3#4
+  {
+    \cs_set_eq:NN \l_peek_search_token #2
+    \tl_set:Nn \l_peek_search_tl {#2}
+    \cs_set_eq:NN \peek_true:w \peek_true_remove:w
+    \cs_set_nopar:Npx \peek_true_aux:w { \exp_not:n {#3} }
+    \cs_set_nopar:Npx \peek_false:w
+      {
+        \exp_not:N \group_align_safe_end:
+        \exp_not:n {#4}
+      }
+    \group_align_safe_begin:
+      \peek_after:Nw #1
+  }
+\cs_new_protected:Npn \peek_token_remove_generic:NNT #1#2#3
+  { \peek_token_remove_generic:NNTF #1 #2 {#3} { } }
+\cs_new_protected:Npn \peek_token_remove_generic:NNF #1#2#3
+  { \peek_token_remove_generic:NNTF #1 #2 { } {#3} }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}
+%   {\peek_execute_branches_catcode:, \peek_execute_branches_meaning:}
+%   The category code and meaning tests are straight forward.
+%    \begin{macrocode}
+\cs_new_nopar:Npn \peek_execute_branches_catcode:
+  {
+    \if_catcode:w
+      \exp_not:N \l_peek_token \exp_not:N \l_peek_search_token
+      \exp_after:wN \peek_true:w
+    \else:
+      \exp_after:wN \peek_false:w
+    \fi:
+  }
+\cs_new_nopar:Npn \peek_execute_branches_meaning:
+  {
+    \if_meaning:w \l_peek_token \l_peek_search_token
+      \exp_after:wN \peek_true:w
+    \else:
+      \exp_after:wN \peek_false:w
+    \fi:
+  }
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\peek_execute_branches_charcode:}
+% \begin{macro}[aux]{\peek_execute_branches_charcode:NN}
+%   First the character code test there is a need to worry about \TeX{}
+%   grabbing brace group or skipping spaces. These are all tested for
+%   using a category code check before grabbing what must be a real
+%   single token and doing the comparison.
+%    \begin{macrocode}
+\cs_new_nopar:Npn \peek_execute_branches_charcode:
+  {
+    \bool_if:nTF
+      {
+           \token_if_eq_catcode_p:NN \l_peek_token \c_group_begin_token
+        || \token_if_eq_catcode_p:NN \l_peek_token \c_group_end_token
+        || \token_if_eq_meaning_p:NN \l_peek_token \c_space_token
+      }
+      { \peek_false:w }
+      {
+        \exp_after:wN \peek_execute_branches_charcode_aux:NN
+          \l_peek_search_tl
+      }
+  }
+\cs_new:Npn \peek_execute_branches_charcode_aux:NN #1#2
+  {
+    \if:w \exp_not:N #1 \exp_not:N #2
+      \exp_after:wN \peek_true:w
+    \else:
+      \exp_after:wN \peek_false:w
+    \fi:
+    #2
+  }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\peek_ignore_spaces_execute_branches:}
+% \begin{macro}[aux]{\peek_ignore_spaces_execute_branches_aux:}
+%   This function removes one token at a time with a mechanism that can
+%   be applied to things other than spaces.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \peek_ignore_spaces_execute_branches:
+  {
+    \token_if_eq_meaning:NNTF \l_peek_token \c_space_token
+      {
+        \tex_afterassignment:D \peek_ignore_spaces_execute_branches_aux:
+        \cs_set_eq:NN \peek_tmp:w
+      }
+      { \peek_execute_branches: }
+  }
+\cs_new_protected_nopar:Npn \peek_ignore_spaces_execute_branches_aux:
+  { \peek_after:Nw \peek_ignore_spaces_execute_branches: }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}[aux]{\peek_def:nnnn}
+% \begin{macro}[aux]{\peek_def_aux:nnnnn}
+%   The public functions themselves cannot be defined using
+%   \cs{prg_set_conditional:Npnn} and so a couple of auxiliary functions
+%   are used. As a result, everything is done inside a group. As a result
+%   things are a bit complicated.
+%    \begin{macrocode}
+\group_begin:
+  \cs_set_nopar:Npn \peek_def:nnnn #1#2#3#4
+    {
+      \peek_def_aux:nnnnn {#1} {#2} {#3} {#4} { TF }
+      \peek_def_aux:nnnnn {#1} {#2} {#3} {#4} { T }
+      \peek_def_aux:nnnnn {#1} {#2} {#3} {#4} { F }
+    }
+  \cs_set_nopar:Npn \peek_def_aux:nnnnn #1#2#3#4#5
+    {
+      \cs_gset_nopar:cpx { #1 #5 }
+        {
+          \tl_if_empty:nF {#2}
+            { \exp_not:n { \cs_set_eq:NN \peek_execute_branches: #2 } }
+          \exp_not:c { #3 #5 }
+          \exp_not:n {#4}
+        }
+    }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \begin{macro}[TF]
+%   {
+%     \peek_catcode:N, \peek_catcode_ignore_spaces:N,
+%     \peek_catcode_remove:N, \peek_catcode_remove_ignore_spaces:N
+%   }
+%   With everything in place the definitions can take place. First for
+%   category codes.
+%    \begin{macrocode}
+  \peek_def:nnnn { peek_catcode:N }
+    { }
+    { peek_token_generic:NN }
+    { \peek_execute_branches_catcode: }
+  \peek_def:nnnn { peek_catcode_ignore_spaces:N }
+    { \peek_execute_branches_catcode: }
+    { peek_token_generic:NN }
+    { \peek_ignore_spaces_execute_branches: }
+  \peek_def:nnnn { peek_catcode_remove:N }
+    { }
+    { peek_token_remove_generic:NN }
+    { \peek_execute_branches_catcode: }
+  \peek_def:nnnn { peek_catcode_remove_ignore_spaces:N }
+    { \peek_execute_branches_catcode: }
+    { peek_token_remove_generic:NN }
+    { \peek_ignore_spaces_execute_branches: }
+%    \end{macrocode}
+% \end{macro}
+% \begin{macro}[TF]
+%   {
+%     \peek_charcode:N, \peek_charcode_ignore_spaces:N,
+%     \peek_charcode_remove:N, \peek_charcode_remove_ignore_spaces:N
+%   }
+%   Then for character codes.
+%    \begin{macrocode}
+  \peek_def:nnnn { peek_charcode:N }
+    { }
+    { peek_token_generic:NN }
+    { \peek_execute_branches_charcode: }
+  \peek_def:nnnn { peek_charcode_ignore_spaces:N }
+    { \peek_execute_branches_charcode: }
+    { peek_token_generic:NN }
+    { \peek_ignore_spaces_execute_branches: }
+  \peek_def:nnnn { peek_charcode_remove:N }
+    { }
+    { peek_token_remove_generic:NN }
+    { \peek_execute_branches_charcode: }
+  \peek_def:nnnn { peek_charcode_remove_ignore_spaces:N }
+    { \peek_execute_branches_charcode: }
+    { peek_token_remove_generic:NN }
+    { \peek_ignore_spaces_execute_branches: }
+%    \end{macrocode}
+% \end{macro}
+% \begin{macro}[TF]
+%   {
+%     \peek_meaning:N, \peek_meaning_ignore_spaces:N,
+%     \peek_meaning_remove:N, \peek_meaning_remove_ignore_spaces:N
+%   }
+%   Finally for meaning, with the group closed to remove the temporary
+%   definition functions.
+%    \begin{macrocode}
+  \peek_def:nnnn { peek_meaning:N }
+    { }
+    { peek_token_generic:NN }
+    { \peek_execute_branches_meaning: }
+  \peek_def:nnnn { peek_meaning_ignore_spaces:N }
+    { \peek_execute_branches_meaning: }
+    { peek_token_generic:NN }
+    { \peek_ignore_spaces_execute_branches: }
+  \peek_def:nnnn { peek_meaning_remove:N }
+    { }
+    { peek_token_remove_generic:NN }
+    { \peek_execute_branches_meaning: }
+  \peek_def:nnnn { peek_meaning_remove_ignore_spaces:N }
+    { \peek_execute_branches_meaning: }
+    { peek_token_remove_generic:NN }
+    { \peek_ignore_spaces_execute_branches: }
+\group_end:
+%    \end{macrocode}
+% \end{macro}
+%
+% \subsection{Decomposing a macro definition}
+%
+% \begin{macro}{\token_get_prefix_spec:N}
+% \begin{macro}{\token_get_arg_spec:N}
+% \begin{macro}{\token_get_replacement_spec:N}
+% \begin{macro}[aux]{\token_get_prefix_arg_replacement_aux:wN}
+%   We sometimes want to test if a
+%   control sequence can be expanded to reveal a hidden
+%   value. However, we cannot just expand the macro blindly as it may
+%   have arguments and none might be present. Therefore we define
+%   these functions to pick either the prefix(es), the argument
+%   specification, or the replacement text from a macro. All of this
+%   information is returned as characters with catcode~$12$. If the
+%   token in question isn't a macro, the token |\scan_stop:| is
+%   returned instead.
+%    \begin{macrocode}
+\exp_args:Nno \use:nn
+  { \cs_new_nopar:Npn \token_get_prefix_arg_replacement_aux:wN #1 }
+  { \tl_to_str:n { macro : } #2 -> #3 \q_stop #4 }
+  { #4 {#1} {#2} {#3} }
+\cs_new:Npn \token_get_prefix_spec:N #1
+  {
+    \token_if_macro:NTF #1
+      {
+        \exp_after:wN \token_get_prefix_arg_replacement_aux:wN
+          \token_to_meaning:N #1 \q_stop \use_i:nnn
+      }
+      { \scan_stop: }
+  }
+\cs_new:Npn \token_get_arg_spec:N #1
+  {
+    \token_if_macro:NTF #1
+      {
+        \exp_after:wN \token_get_prefix_arg_replacement_aux:wN
+          \token_to_meaning:N #1 \q_stop \use_ii:nnn
+      }
+      { \scan_stop: }
+  }
+\cs_new:Npn \token_get_replacement_spec:N #1
+  {
+    \token_if_macro:NTF #1
+      {
+        \exp_after:wN \token_get_prefix_arg_replacement_aux:wN
+          \token_to_meaning:N #1 \q_stop \use_iii:nnn
+      }
+      { \scan_stop: }
+  }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% 
+% \subsection{Experimental token functions}
+%
+% \begin{macro}{\char_active_set:Npn,\char_active_set:Npx}
+% \begin{macro}{\char_active_set:Npn,\char_active_set:Npx}
+% \begin{macro}{\char_active_set_eq:NN,\char_active_gset_eq:NN}
+%    \begin{macrocode}
+\group_begin:
+  \char_set_catcode_active:N \^^@
+  \cs_set:Npn \char_tmp:NN #1#2 
+    {
+      \cs_new:Npn #1 ##1
+        {
+          \char_set_catcode_active:n { `##1 }
+          \group_begin:
+          \char_set_lccode:nn { `\^^@ } { `##1 }
+          \tl_to_lowercase:n { \group_end: #2 ^^@ }
+        }
+    }
+  \char_tmp:NN \char_active_set:Npn    \cs_set:Npn
+  \char_tmp:NN \char_active_set:Npx    \cs_set:Npx
+  \char_tmp:NN \char_active_gset:Npn   \cs_gset:Npn
+  \char_tmp:NN \char_active_gset:Npx   \cs_gset:Npx
+  \char_tmp:NN \char_active_set_eq:NN  \cs_set_eq:NN
+  \char_tmp:NN \char_active_gset_eq:NN \cs_gset_eq:NN
+\group_end:
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \subsection{Deprecated functions}
+%
+% Deprecated on 2011-05-27, for removal by 2011-08-31.
+%
+% \begin{macro}{\char_set_catcode:w}
+% \begin{macro}{\char_set_mathcode:w}
+% \begin{macro}{\char_set_lccode:w}
+% \begin{macro}{\char_set_uccode:w}
+% \begin{macro}{\char_set_sfcode:w}
+%   Primitives renamed.
+%    \begin{macrocode}
+\cs_new_eq:NN \char_set_catcode:w  \tex_catcode:D
+\cs_new_eq:NN \char_set_mathcode:w \tex_mathcode:D
+\cs_new_eq:NN \char_set_lccode:w   \tex_lccode:D
+\cs_new_eq:NN \char_set_uccode:w   \tex_uccode:D
+\cs_new_eq:NN \char_set_sfcode:w   \tex_sfcode:D
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\char_value_catcode:w}
+% \begin{macro}{\char_show_value_catcode:w}
+% \begin{macro}{\char_value_mathcode:w}
+% \begin{macro}{\char_show_value_mathcode:w}
+% \begin{macro}{\char_value_lccode:w}
+% \begin{macro}{\char_show_value_lccode:w}
+% \begin{macro}{\char_value_uccode:w}
+% \begin{macro}{\char_show_value_uccode:w}
+% \begin{macro}{\char_value_sfcode:w}
+% \begin{macro}{\char_show_value_sfcode:w}
+%   More |w| functions we should not have.
+%    \begin{macrocode}
+\cs_new_nopar:Npn \char_value_catcode:w { \tex_the:D \char_set_catcode:w }
+\cs_new_nopar:Npn \char_show_value_catcode:w
+  { \tex_showthe:D \char_set_catcode:w }
+\cs_new_nopar:Npn \char_value_mathcode:w { \tex_the:D \char_set_mathcode:w }
+\cs_new_nopar:Npn \char_show_value_mathcode:w
+  { \tex_showthe:D \char_set_mathcode:w }
+\cs_new_nopar:Npn \char_value_lccode:w { \tex_the:D \char_set_lccode:w }
+\cs_new_nopar:Npn \char_show_value_lccode:w
+  { \tex_showthe:D \char_set_lccode:w }
+\cs_new_nopar:Npn \char_value_uccode:w { \tex_the:D \char_set_uccode:w }
+\cs_new_nopar:Npn \char_show_value_uccode:w
+  { \tex_showthe:D \char_set_uccode:w }
+\cs_new_nopar:Npn \char_value_sfcode:w { \tex_the:D \char_set_sfcode:w }
+\cs_new_nopar:Npn \char_show_value_sfcode:w
+  { \tex_showthe:D \char_set_sfcode:w }
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\peek_after:NN}
+% \begin{macro}{\peek_gafter:NN}
+%   The second argument here must be |w|.
+%    \begin{macrocode}
+\cs_new_eq:NN \peek_after:NN  \peek_after:Nw
+\cs_new_eq:NN \peek_gafter:NN \peek_gafter:Nw
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% 
+% Functions deprecated 2011-05-28 for removal by 2011-08-31.
+% 
+% \begin{macro}{\c_alignment_tab_token}
+% \begin{macro}{\c_math_shift_token}
+% \begin{macro}{\c_letter_token}
+% \begin{macro}{\c_other_char_token}
+%    \begin{macrocode}
+\cs_new_eq:NN \c_alignment_tab_token \c_alignment_token
+\cs_new_eq:NN \c_math_shift_token    \c_math_toggle_token
+\cs_new_eq:NN \c_letter_token        \c_catcode_letter_token
+\cs_new_eq:NN \c_other_char_token    \c_catcode_other_token
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% 
+% \begin{macro}{\c_active_char_token}
+%   An odd one: this was never a |token|!
+%    \begin{macrocode}
+\cs_new_eq:NN \c_active_char_token \c_catcode_active_tl
+%    \end{macrocode}
+% \end{macro}
+% 
+% \begin{macro}
+%   {
+%     \char_make_escape:N           ,
+%     \char_make_group_begin:N      ,
+%     \char_make_group_end:N        ,
+%     \char_make_math_toggle:N      ,
+%     \char_make_alignment:N        , 
+%     \char_make_end_line:N         ,
+%     \char_make_parameter:N        ,
+%     \char_make_math_superscript:N ,
+%     \char_make_math_subscript:N   , 
+%     \char_make_ignore:N           ,
+%     \char_make_space:N            , 
+%     \char_make_letter:N           ,
+%     \char_make_other:N            , 
+%     \char_make_active:N           ,
+%     \char_make_comment:N          , 
+%     \char_make_invalid:N          
+%   }
+% \begin{macro}
+%   {
+%     \char_make_escape:n           ,
+%     \char_make_group_begin:n      ,
+%     \char_make_group_end:n        ,
+%     \char_make_math_toggle:n      ,
+%     \char_make_alignment:n        , 
+%     \char_make_end_line:n         ,
+%     \char_make_parameter:n        ,
+%     \char_make_math_superscript:n ,
+%     \char_make_math_subscript:n   , 
+%     \char_make_ignore:n           ,
+%     \char_make_space:n            , 
+%     \char_make_letter:n           ,
+%     \char_make_other:n            , 
+%     \char_make_active:n           ,
+%     \char_make_comment:n          , 
+%     \char_make_invalid:n          
+%   }
+%  Two renames in one block! 
+%    \begin{macrocode}
+\cs_new_eq:NN \char_make_escape:N           \char_set_catcode_escape:N
+\cs_new_eq:NN \char_make_begin_group:N      \char_set_catcode_group_begin:N
+\cs_new_eq:NN \char_make_end_group:N        \char_set_catcode_group_end:N
+\cs_new_eq:NN \char_make_math_shift:N       \char_set_catcode_math_toggle:N
+\cs_new_eq:NN \char_make_alignment_tab:N    \char_set_catcode_alignment:N
+\cs_new_eq:NN \char_make_end_line:N         \char_set_catcode_end_line:N
+\cs_new_eq:NN \char_make_parameter:N        \char_set_catcode_parameter:N
+\cs_new_eq:NN \char_make_math_superscript:N
+  \char_set_catcode_math_superscript:N
+\cs_new_eq:NN \char_make_math_subscript:N
+  \char_set_catcode_math_subscript:N
+\cs_new_eq:NN \char_make_ignore:N           \char_set_catcode_ignore:N
+\cs_new_eq:NN \char_make_space:N            \char_set_catcode_space:N
+\cs_new_eq:NN \char_make_letter:N           \char_set_catcode_letter:N
+\cs_new_eq:NN \char_make_other:N            \char_set_catcode_other:N
+\cs_new_eq:NN \char_make_active:N           \char_set_catcode_active:N
+\cs_new_eq:NN \char_make_comment:N          \char_set_catcode_comment:N
+\cs_new_eq:NN \char_make_invalid:N          \char_set_catcode_invalid:N
+\cs_new_eq:NN \char_make_escape:n           \char_set_catcode_escape:n
+\cs_new_eq:NN \char_make_begin_group:n      \char_set_catcode_group_begin:n
+\cs_new_eq:NN \char_make_end_group:n        \char_set_catcode_group_end:n
+\cs_new_eq:NN \char_make_math_shift:n       \char_set_catcode_math_toggle:n
+\cs_new_eq:NN \char_make_alignment_tab:n    \char_set_catcode_alignment:n
+\cs_new_eq:NN \char_make_end_line:n         \char_set_catcode_end_line:n
+\cs_new_eq:NN \char_make_parameter:n        \char_set_catcode_parameter:n
+\cs_new_eq:NN \char_make_math_superscript:n
+  \char_set_catcode_math_superscript:n
+\cs_new_eq:NN \char_make_math_subscript:n
+  \char_set_catcode_math_subscript:n
+\cs_new_eq:NN \char_make_ignore:n           \char_set_catcode_ignore:n
+\cs_new_eq:NN \char_make_space:n           \char_set_catcode_space:n
+\cs_new_eq:NN \char_make_letter:n           \char_set_catcode_letter:n
+\cs_new_eq:NN \char_make_other:n            \char_set_catcode_other:n
+\cs_new_eq:NN \char_make_active:n           \char_set_catcode_active:n
+\cs_new_eq:NN \char_make_comment:n          \char_set_catcode_comment:n
+\cs_new_eq:NN \char_make_invalid:n          \char_set_catcode_invalid:n
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% 
+% \begin{macro}[pTF]{\token_if_alignment_tab:N}
+% \begin{macro}[pTF]{\token_if_math_shift:N}
+% \begin{macro}[pTF]{\token_if_other_char:N}
+% \begin{macro}[pTF]{\token_if_active_char:N}
+%    \begin{macrocode}
+\cs_new_eq:NN \token_if_alignment_tab_p:N \token_if_alignment_p:N
+\cs_new_eq:NN \token_if_alignment_tab:NT  \token_if_alignment:NT
+\cs_new_eq:NN \token_if_alignment_tab:NF  \token_if_alignment:NF
+\cs_new_eq:NN \token_if_alignment_tab:NTF \token_if_alignment:NTF
+\cs_new_eq:NN \token_if_math_shift_p:N \token_if_math_toggle_p:N
+\cs_new_eq:NN \token_if_math_shift:NT  \token_if_math_toggle:NT
+\cs_new_eq:NN \token_if_math_shift:NF  \token_if_math_toggle:NF
+\cs_new_eq:NN \token_if_math_shift:NTF \token_if_math_toggle:NTF
+\cs_new_eq:NN \token_if_other_char_p:N \token_if_other_p:N
+\cs_new_eq:NN \token_if_other_char:NT  \token_if_other:NT
+\cs_new_eq:NN \token_if_other_char:NF  \token_if_other:NF
+\cs_new_eq:NN \token_if_other_char:NTF \token_if_other:NTF
+\cs_new_eq:NN \token_if_active_char_p:N \token_if_active_p:N
+\cs_new_eq:NN \token_if_active_char:NT  \token_if_active:NT
+\cs_new_eq:NN \token_if_active_char:NF  \token_if_active:NF
+\cs_new_eq:NN \token_if_active_char:NTF \token_if_active:NTF
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+%    \begin{macrocode}
+%</initex|package>
+%    \end{macrocode}
+%
+% \end{implementation}
+%
+% \PrintIndex
diff --git a/Master/texmf-dist/source/latex/l3kernel/l3toks.dtx b/Master/texmf-dist/source/latex/l3kernel/l3toks.dtx
new file mode 100644
index 00000000000..62824848afc
--- /dev/null
+++ b/Master/texmf-dist/source/latex/l3kernel/l3toks.dtx
@@ -0,0 +1,548 @@
+% \iffalse meta-comment
+%
+%% File: l3toks.dtx Copyright (C) 1990-2011 The LaTeX3 project
+%%
+%% It may be distributed and/or modified under the conditions of the
+%% LaTeX Project Public License (LPPL), either version 1.3c of this
+%% license or (at your option) any later version.  The latest version
+%% of this license is in the file
+%%
+%%    http://www.latex-project.org/lppl.txt
+%%
+%% This file is part of the "expl3 bundle" (The Work in LPPL)
+%% and all files in that bundle must be distributed together.
+%%
+%% The released version of this bundle is available from CTAN.
+%%
+%% -----------------------------------------------------------------------
+%%
+%% The development version of the bundle can be found at
+%%
+%%    http://www.latex-project.org/svnroot/experimental/trunk/
+%%
+%% for those people who are interested.
+%%
+%%%%%%%%%%%
+%% NOTE: %%
+%%%%%%%%%%%
+%%
+%%   Snapshots taken from the repository represent work in progress and may
+%%   not work or may contain conflicting material!  We therefore ask
+%%   people _not_ to put them into distributions, archives, etc. without
+%%   prior consultation with the LaTeX3 Project.
+%%
+%% -----------------------------------------------------------------------
+%
+%<*driver|package>
+\RequirePackage{l3names}
+\GetIdInfo$Id: l3toks.dtx 2478 2011-06-19 21:34:23Z joseph $
+  {L3 Experimental token registers}
+%</driver|package>
+%<*driver>
+\documentclass[full]{l3doc}
+\begin{document}
+  \DocInput{\jobname.dtx}
+\end{document}
+%</driver>
+% \fi
+%
+% \title{^^A
+%   The \pkg{l3toks} package\\ Token registers^^A
+%   \thanks{This file describes v\ExplFileVersion,
+%     last revised \ExplFileDate.}^^A
+% }
+%
+% \author{^^A
+%  The \LaTeX3 Project\thanks
+%    {^^A
+%      E-mail:
+%        \href{mailto:latex-team@latex-project.org}
+%          {latex-team@latex-project.org}^^A
+%    }^^A
+% }
+%
+% \date{Released \ExplFileDate}
+%
+% \maketitle
+%
+% \begin{documentation}
+%
+% \textbf{THIS ENTIRE MODULE IS DEPRECIATED: DO NOT USE \cs{TOKS_\ldots}
+% FUNCTIONS}
+%
+% \end{documentation}
+%
+% \begin{implementation}
+%
+%    \section{\pkg{l3toks} implementation}
+%
+%    We start by ensuring that the required packages are loaded.
+%    \begin{macrocode}
+%<*package>
+\ProvidesExplPackage
+  {\ExplFileName}{\ExplFileDate}{\ExplFileVersion}{\ExplFileDescription}
+\package_check_loaded_expl:
+%</package>
+%<*initex|package>
+%    \end{macrocode}
+%
+% \subsection{Allocation and use}
+%
+% \begin{macro}{\toks_new:N,\toks_new:c}
+%    Allocates a new token register.
+%    \begin{macrocode}
+%<*initex>
+\alloc_new:nnnN {toks} \c_zero \c_max_register_int \tex_toksdef:D
+%</initex>
+%    \end{macrocode}
+%
+%    \begin{macrocode}
+%<*package>
+\cs_new_protected_nopar:Npn \toks_new:N #1 {
+  \chk_if_free_cs:N #1
+  \newtoks #1
+}
+%</package>
+%    \end{macrocode}
+%
+%    \begin{macrocode}
+\cs_generate_variant:Nn \toks_new:N {c}
+%    \end{macrocode}
+% \end{macro}
+%
+%
+% \begin{macro}{\toks_use:N}
+% \begin{macro}{\toks_use:c}
+%    This function returns the contents of a token register.
+%    \begin{macrocode}
+\cs_new_eq:NN  \toks_use:N \tex_the:D
+\cs_generate_variant:Nn \toks_use:N {c}
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\toks_set:Nn}
+% \begin{macro}{\toks_set:NV}
+% \begin{macro}{\toks_set:Nv}
+% \begin{macro}{\toks_set:No}
+% \begin{macro}{\toks_set:Nx}
+% \begin{macro}{\toks_set:Nf}
+% \begin{macro}{\toks_set:cn}
+% \begin{macro}{\toks_set:co}
+% \begin{macro}{\toks_set:cV}
+% \begin{macro}{\toks_set:cv}
+% \begin{macro}{\toks_set:cx}
+% \begin{macro}{\toks_set:cf}
+%    |\toks_set:Nn|\m{toks}\m{stuff} stores \m{stuff} without expansion
+%    in \m{toks}. |\toks_set:No| and |\toks_set:Nx| expand \m{stuff} once
+%    and fully.
+%    \begin{macrocode}
+%<*check>
+\cs_new_protected_nopar:Npn \toks_set:Nn #1 { \chk_local:N #1 #1 }
+\cs_generate_variant:Nn \toks_set:Nn {No,Nf}
+%</check>
+%    \end{macrocode}
+%    If we don't check if \m{toks} is a local register then the
+%    |\toks_set:Nn| function has nothing to do.
+%    We implement |\toks_set:No|/|d|/|f| by hand when not checking because this
+%    is going to be used \emph{extensively} in keyval processing!
+%    TODO: (Will) Can we get some numbers published on how necessary this is?
+%    On the other hand I'm happy to believe Morten |:)|
+%    \begin{macrocode}
+%<*!check>
+\cs_new_eq:NN       \toks_set:Nn \prg_do_nothing:
+\cs_new_protected:Npn \toks_set:NV #1#2 { 
+  #1 \exp_after:wN { \int_to_roman:w -`0 \exp_eval_register:N #2 }
+}
+\cs_new_protected:Npn \toks_set:Nv #1#2 { 
+  #1 \exp_after:wN { \int_to_roman:w -`0 \exp_eval_register:c {#2} }
+}
+\cs_new_protected:Npn \toks_set:No #1#2 { #1 \exp_after:wN {#2} }
+\cs_new_protected:Npn \toks_set:Nf #1#2 {
+  #1 \exp_after:wN { \int_to_roman:w -`0#2 }
+}
+%</!check>
+%    \end{macrocode}
+% 
+%    \begin{macrocode}
+\cs_generate_variant:Nn \toks_set:Nn {Nx,cn,cV,cv,co,cx,cf}
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\toks_gset:Nn}
+% \begin{macro}{\toks_gset:NV}
+% \begin{macro}{\toks_gset:No}
+% \begin{macro}{\toks_gset:Nx}
+% \begin{macro}{\toks_gset:cn}
+% \begin{macro}{\toks_gset:cV}
+% \begin{macro}{\toks_gset:co}
+% \begin{macro}{\toks_gset:cx}
+%    These functions are the global variants of the above.
+%    \begin{macrocode}
+%<check>\cs_new_protected_nopar:Npn \toks_gset:Nn #1 { \chk_global:N #1 \pref_global:D #1 }
+%<!check>\cs_new_eq:NN \toks_gset:Nn \pref_global:D
+\cs_generate_variant:Nn \toks_gset:Nn {NV,No,Nx,cn,cV,co,cx}
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+%  \begin{macro}{\toks_set_eq:NN}
+%  \begin{macro}{\toks_set_eq:Nc}
+%  \begin{macro}{\toks_set_eq:cN}
+%  \begin{macro}{\toks_set_eq:cc}
+%  \begin{macro}{\toks_gset_eq:NN}
+%  \begin{macro}{\toks_gset_eq:Nc}
+%  \begin{macro}{\toks_gset_eq:cN}
+%  \begin{macro}{\toks_gset_eq:cc}
+%    |\toks_set_eq:NN|\m{toks1}\m{toks2} copies the contents of \m{toks2}
+%    in \m{toks1}.
+%    \begin{macrocode}
+%<*check>
+\cs_new_protected_nopar:Npn\toks_set_eq:NN #1#2 {
+  \chk_local:N #1
+  \chk_var_or_const:N #2
+  #1 #2
+}
+\cs_new_protected_nopar:Npn\toks_gset_eq:NN #1#2 {
+  \chk_global:N #1
+  \chk_var_or_const:N #2
+  \pref_global:D #1 #2
+}
+%</check>
+%<*!check>
+\cs_new_eq:NN \toks_set_eq:NN  \prg_do_nothing:
+\cs_new_eq:NN \toks_gset_eq:NN \pref_global:D
+%</!check>
+\cs_generate_variant:Nn \toks_set_eq:NN  {Nc,cN,cc}
+\cs_generate_variant:Nn \toks_gset_eq:NN  {Nc,cN,cc}
+%    \end{macrocode}
+%  \end{macro}
+%  \end{macro}
+%  \end{macro}
+%  \end{macro}
+%  \end{macro}
+%  \end{macro}
+%  \end{macro}
+%  \end{macro}
+%
+% \begin{macro}{\toks_clear:N,\toks_gclear:N,\toks_clear:c,\toks_gclear:c}
+%    These functions clear a token register, either locally or globally.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \toks_clear:N #1 {
+  #1\c_empty_toks
+%<check>\chk_local_or_pref_global:N #1
+}
+%    \end{macrocode}
+% 
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \toks_gclear:N {
+%<check>  \pref_global_chk:
+%<!check>  \pref_global:D
+  \toks_clear:N
+}
+%    \end{macrocode}
+% 
+%    \begin{macrocode}
+\cs_generate_variant:Nn \toks_clear:N  {c}
+\cs_generate_variant:Nn \toks_gclear:N {c}
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\toks_use_clear:N}
+% \begin{macro}{\toks_use_clear:c}
+% \begin{macro}{\toks_use_gclear:N}
+% \begin{macro}{\toks_use_gclear:c}
+%    These functions clear a token register (locally or globally) after
+%    returning the contents.
+%
+%    They make sure that clearing the register does not
+%    interfere with following tokens. In other words, the contents of
+%    the register might operate on what follows in the input stream.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \toks_use_clear:N #1 {
+  \exp_last_unbraced:NNV \toks_clear:N #1 #1
+}
+%    \end{macrocode}
+%
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \toks_use_gclear:N {
+%<check>  \pref_global_chk:
+%<!check>  \pref_global:D
+  \toks_use_clear:N
+}
+%    \end{macrocode}
+%
+%    \begin{macrocode}
+\cs_generate_variant:Nn \toks_use_clear:N  {c}
+\cs_generate_variant:Nn \toks_use_gclear:N {c}
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\toks_show:N}
+% \begin{macro}{\toks_show:c}
+%    This function shows the contents of a token register on the terminal.
+%    \begin{macrocode}
+\cs_new_eq:NN           \toks_show:N \kernel_register_show:N
+\cs_generate_variant:Nn \toks_show:N {c}
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \subsection{Adding to token registers' contents}
+%
+% \begin{macro}{\toks_put_left:Nn}
+% \begin{macro}{\toks_put_left:NV}
+% \begin{macro}{\toks_put_left:No}
+% \begin{macro}{\toks_put_left:Nx}
+% \begin{macro}{\toks_put_left:cn}
+% \begin{macro}{\toks_put_left:cV}
+% \begin{macro}{\toks_put_left:co}
+% \begin{macro}{\toks_gput_left:Nn}
+% \begin{macro}{\toks_gput_left:NV}
+% \begin{macro}{\toks_gput_left:No}
+% \begin{macro}{\toks_gput_left:Nx}
+% \begin{macro}{\toks_gput_left:cn}
+% \begin{macro}{\toks_gput_left:cV}
+% \begin{macro}{\toks_gput_left:co}
+% \begin{macro}[aux]{\toks_put_left_aux:w}
+%   |\toks_put_left:Nn |\meta{toks}\meta{stuff\/} adds the tokens of
+%   \textsl{stuff} on the `left-side' of the token register
+%   \m{toks}. |\toks_put_left:No| does the same, but expands the
+%   tokens once. We need to look out for brace stripping so we add a
+%   token, which is then later removed.
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \toks_put_left:Nn #1 {
+  \exp_after:wN \toks_put_left_aux:w \exp_after:wN \q_nil
+    \toks_use:N #1 \q_stop #1
+}
+%    \end{macrocode}
+% 
+%    \begin{macrocode}
+\cs_generate_variant:Nn \toks_put_left:Nn  {NV,No,Nx,cn,co,cV}
+%    \end{macrocode}
+% 
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \toks_gput_left:Nn {
+%<check>  \pref_global_chk:
+%<!check>  \pref_global:D
+  \toks_put_left:Nn
+}
+%    \end{macrocode}
+% 
+%    \begin{macrocode}
+\cs_generate_variant:Nn \toks_gput_left:Nn  {NV,No,Nx,cn,cV,co}
+%    \end{macrocode}
+% A helper function for |\toks_put_left:Nn|. Its arguments are
+% subsequently the tokens of \meta{stuff\/}, the token register
+% \meta{toks} and the current contents of \meta{toks}. We make sure to
+% remove the token we inserted earlier.
+%    \begin{macrocode}
+\cs_new:Npn \toks_put_left_aux:w #1\q_stop #2#3 {
+  #2 \exp_after:wN { \use_i:nn {#3} #1 }
+%<check>  \chk_local_or_pref_global:N #2
+}
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\toks_put_right:Nn}
+% \begin{macro}{\toks_put_right:NV}
+% \begin{macro}{\toks_put_right:No}
+% \begin{macro}{\toks_put_right:Nx}
+% \begin{macro}{\toks_put_right:cn}
+% \begin{macro}{\toks_put_right:cV}
+% \begin{macro}{\toks_put_right:co}
+% \begin{macro}{\toks_gput_right:Nn}
+% \begin{macro}{\toks_gput_right:NV}
+% \begin{macro}{\toks_gput_right:No}
+% \begin{macro}{\toks_gput_right:Nx}
+% \begin{macro}{\toks_gput_right:cn}
+% \begin{macro}{\toks_gput_right:cV}
+% \begin{macro}{\toks_gput_right:co}
+%    These macros add a list of tokens to the right of a token register.
+%    \begin{macrocode}
+\cs_new_protected:Npn \toks_put_right:Nn #1#2 {
+  #1 \exp_after:wN { \toks_use:N #1 #2 }
+%<check>  \chk_local_or_pref_global:N #1
+}
+%    \end{macrocode}
+% 
+%    \begin{macrocode}
+\cs_new_protected_nopar:Npn \toks_gput_right:Nn {
+%<check>  \pref_global_chk:
+%<!check>  \pref_global:D
+  \toks_put_right:Nn
+}
+%    \end{macrocode}
+%  A couple done by hand for speed.
+%    \begin{macrocode}
+%<check>\cs_generate_variant:Nn \toks_put_right:Nn {No}
+%<*!check>
+\cs_new_protected:Npn \toks_put_right:NV #1#2 {
+  #1 \exp_after:wN \exp_after:wN \exp_after:wN {
+     \exp_after:wN \toks_use:N \exp_after:wN #1 
+     \int_to_roman:w -`0 \exp_eval_register:N #2
+  }
+}
+\cs_new_protected:Npn \toks_put_right:No #1#2 {
+  #1 \exp_after:wN \exp_after:wN \exp_after:wN { 
+    \exp_after:wN \toks_use:N \exp_after:wN #1 #2 
+  }
+}
+%</!check>
+\cs_generate_variant:Nn \toks_put_right:Nn {Nx,cn,cV,co}
+\cs_generate_variant:Nn \toks_gput_right:Nn {NV,No,Nx,cn,cV,co}
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\toks_put_right:Nf}
+% We implement |\toks_put_right:Nf| by hand because I think I might
+% use it in the \textsf{l3keyval} module in which case it is going to
+% be used a lot.
+%    \begin{macrocode}
+%<check>\cs_generate_variant:Nn \toks_put_right:Nn {Nf}
+%<*!check>
+\cs_new_protected:Npn \toks_put_right:Nf #1#2 {
+  #1 \exp_after:wN \exp_after:wN \exp_after:wN {
+     \exp_after:wN \toks_use:N \exp_after:wN #1 \int_to_roman:w -`0#2
+  }
+}
+%</!check>
+%    \end{macrocode}
+% \end{macro}
+%
+% \subsection{Predicates and conditionals}
+%
+% \begin{macro}{\toks_if_empty_p:N,\toks_if_empty_p:c}
+% \begin{macro}[TF]{\toks_if_empty:N,\toks_if_empty:c}
+%   |\toks_if_empty:NTF|\m{toks}\m{true code}\m{false code} tests if a
+%   token register is empty and executes either \m{true code} or
+%   \m{false code}. This test had the advantage of being
+%   expandable. Otherwise one has to do an |x| type expansion in order
+%   to prevent problems with parameter tokens.
+%    \begin{macrocode}
+\prg_new_conditional:Nnn \toks_if_empty:N {p,TF,T,F} {
+  \tl_if_empty:VTF #1 {\prg_return_true:} {\prg_return_false:}
+}
+%    \end{macrocode}
+%    \begin{macrocode}
+\cs_generate_variant:Nn \toks_if_empty_p:N {c}
+\cs_generate_variant:Nn \toks_if_empty:NTF {c}
+\cs_generate_variant:Nn \toks_if_empty:NT  {c}
+\cs_generate_variant:Nn \toks_if_empty:NF  {c}
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\toks_if_eq_p:NN,\toks_if_eq_p:cN,\toks_if_eq_p:Nc,\toks_if_eq_p:cc}
+% \begin{macro}[TF]{\toks_if_eq:NN,\toks_if_eq:Nc,\toks_if_eq:cN,\toks_if_eq:cc}
+%    This function test whether two token registers have the same contents.
+%    \begin{macrocode}
+\prg_new_conditional:Nnn \toks_if_eq:NN {p,TF,T,F} {
+  \str_if_eq:xxTF {\toks_use:N #1} {\toks_use:N #2}
+      {\prg_return_true:} {\prg_return_false:}
+}
+\cs_generate_variant:Nn \toks_if_eq_p:NN {Nc,c,cc}
+\cs_generate_variant:Nn \toks_if_eq:NNTF {Nc,c,cc}
+\cs_generate_variant:Nn \toks_if_eq:NNT  {Nc,c,cc}
+\cs_generate_variant:Nn \toks_if_eq:NNF  {Nc,c,cc}
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \subsection{Variables and constants}
+%
+% \begin{macro}{\l_tmpa_toks}
+% \begin{macro}{\l_tmpb_toks}
+% \begin{macro}{\l_tmpc_toks}
+% \begin{macro}{\g_tmpa_toks}
+% \begin{macro}{\g_tmpb_toks}
+% \begin{macro}{\g_tmpc_toks}
+%    Some scratch registers \ldots
+%    \begin{macrocode}
+\tex_toksdef:D \l_tmpa_toks = 255\scan_stop:
+%<initex>\seq_put_right:Nn \g_toks_allocation_seq {255}
+\toks_new:N \l_tmpb_toks
+\toks_new:N \l_tmpc_toks
+\toks_new:N \g_tmpa_toks
+\toks_new:N \g_tmpb_toks
+\toks_new:N \g_tmpc_toks
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\c_empty_toks}
+%    And here is a constant, which is a (permanently) empty token register.
+%    \begin{macrocode}
+\toks_new:N \c_empty_toks
+%    \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\l_tl_replace_toks}
+%    And here is one for tl vars. Can't define it there as the allocation
+%    isn't set up at that point.
+%    \begin{macrocode}
+\toks_new:N \l_tl_replace_toks
+%    \end{macrocode}
+% \end{macro}
+%
+%    \begin{macrocode}
+%</initex|package>
+%    \end{macrocode}
+%
+%
+% \end{implementation}
+% \PrintIndex
+%
+% \endinput
+
diff --git a/Master/texmf-dist/tex/latex/l3kernel/expl3.sty b/Master/texmf-dist/tex/latex/l3kernel/expl3.sty
new file mode 100644
index 00000000000..d0ac38f18e1
--- /dev/null
+++ b/Master/texmf-dist/tex/latex/l3kernel/expl3.sty
@@ -0,0 +1,179 @@
+%%
+%% This is file `expl3.sty',
+%% generated with the docstrip utility.
+%%
+%% The original source files were:
+%%
+%% expl3.dtx  (with options: `package')
+%% 
+%% EXPERIMENTAL CODE
+%% 
+%% Do not distribute this file without also distributing the
+%% source files specified above.
+%% 
+%% Do not distribute a modified version of this file.
+%% 
+%% File: expl3.dtx Copyright (C) 1990-2011 The LaTeX3 Project
+%%
+%% It may be distributed and/or modified under the conditions of the
+%% LaTeX Project Public License (LPPL), either version 1.3c of this
+%% license or (at your option) any later version.  The latest version
+%% of this license is in the file
+%%
+%%    http://www.latex-project.org/lppl.txt
+%%
+%% This file is part of the "expl3 bundle" (The Work in LPPL)
+%% and all files in that bundle must be distributed together.
+%%
+%% The released version of this bundle is available from CTAN.
+%%
+%% -----------------------------------------------------------------------
+%%
+%% The development version of the bundle can be found at
+%%
+%%    http://www.latex-project.org/svnroot/experimental/trunk/
+%%
+%% for those people who are interested.
+%%
+%%%%%%%%%%%
+%% NOTE: %%
+%%%%%%%%%%%
+%%
+%%   Snapshots taken from the repository represent work in progress and may
+%%   not work or may contain conflicting material!  We therefore ask
+%%   people _not_ to put them into distributions, archives, etc. without
+%%   prior consultation with the LaTeX3 Project.
+%%
+%% -----------------------------------------------------------------------
+\RequirePackage{l3names}
+\GetIdInfo$Id: expl3.dtx 2478 2011-06-19 21:34:23Z joseph $
+  {L3 Experimental code bundle wrapper}
+ %\subsubsection{Variables: scope and type}
+\RequirePackage{etex}
+\reserveinserts{32}
+\RequirePackage{calc}
+\ProvidesExplPackage
+  {\ExplFileName}{\ExplFileDate}{\ExplFileVersion}{\ExplFileDescription}
+\newcommand* \@l@expl@check@declarations@bool { 0 }
+\newcommand*\@l@expl@log@functions@bool { 0 }
+\DeclareOption { check-declarations }
+  { \def \@l@expl@check@declarations@bool { 1 } }
+\DeclareOption { log-functions }
+   {\def \@l@expl@log@functions@bool { 1 } }
+\ProcessOptions \relax
+\RequirePackage{ l3basics, l3expan, l3tl }
+\ifodd \@l@expl@check@declarations@bool \relax
+  \cs_new_protected:Npn \tl_check_exists:N #1
+    {
+      \cs_if_exist:NF #1
+        {
+          \msg_kernel_error:nnx { check } { non-declared-variable }
+            { \token_to_str:N #1 }
+        }
+    }
+  \cs_set_protected:Npn \tl_set:Nn #1#2
+    {
+      \tl_check_exists:N #1
+      \cs_set_nopar:Npx #1 { \exp_not:n {#2} }
+    }
+  \cs_set_protected:Npn \tl_set:Nx #1#2
+    {
+      \tl_check_exists:N #1
+      \cs_set_nopar:Npx #1 {#2}
+    }
+  \cs_set_protected:Npn \tl_gset:Nn #1#2
+    {
+      \tl_check_exists:N #1
+      \cs_gset_nopar:Npx #1 { \exp_not:n {#2} }
+    }
+  \cs_set_protected:Npn \tl_gset:Nx #1#2
+    {
+      \tl_check_exists:N #1
+      \cs_gset_nopar:Npx #1 {#2}
+    }
+  \cs_set_protected:Npn \tl_set_eq:NN #1#2
+    {
+      \tl_check_exists:N #1
+      \tl_check_exists:N #2
+      \cs_set_eq:NN #1 #2
+    }
+  \cs_set_protected:Npn \tl_gset_eq:NN #1#2
+    {
+      \tl_check_exists:N #1
+      \tl_check_exists:N #2
+      \cs_gset_eq:NN #1 #2
+    }
+\fi
+\RequirePackage{
+  l3seq,
+  l3int,
+  l3quark,
+  l3prg,
+  l3clist,
+  l3token,
+  l3prop,
+  l3msg,
+  l3io,
+  l3file,
+  l3skip,
+  l3box,
+  l3keyval,
+  l3keys,
+  l3fp,
+  l3luatex
+}
+\RequirePackage{l3toks}
+\bool_if:nT { \@l@expl@check@declarations@bool }
+  {
+    \msg_kernel_new:nnnn { check } { non-declared-variable }
+      { The~variable~#1~has~not~been~declared~\msg_line_context:. }
+      {
+        Checking~is~active,~and~you~have~tried~do~so~something~like: \\
+        \ \ \tl_set:Nn #1 ~
+          \iow_char:N \{ ~ ... ~ \iow_char:N \} \\
+        without~first~having: \\
+        \ \ \tl_new:N #1  \\
+        \\
+        LaTeX~will~create~the~variable~and~continue.
+      }
+  }
+\cs_set_protected_nopar:Npn \ExplSyntaxOn
+  {
+    \bool_if:NF \l_expl_status_bool
+      {
+        \cs_set_protected_nopar:Npx \ExplSyntaxOff
+          {
+            \char_set_catcode:nn { 9 }   { \char_value_catcode:n { 9 } }
+            \char_set_catcode:nn { 32 }  { \char_value_catcode:n { 32 } }
+            \char_set_catcode:nn { 34 }  { \char_value_catcode:n { 34 } }
+            \char_set_catcode:nn { 38 }  { \char_value_catcode:n { 38 } }
+            \char_set_catcode:nn { 58 }  { \char_value_catcode:n { 58 } }
+            \char_set_catcode:nn { 94 }  { \char_value_catcode:n { 94 } }
+            \char_set_catcode:nn { 95 }  { \char_value_catcode:n { 95 } }
+            \char_set_catcode:nn { 124 } { \char_value_catcode:n { 124 } }
+            \char_set_catcode:nn { 126 } { \char_value_catcode:n { 126 } }
+            \tex_endlinechar:D = \tex_the:D \tex_endlinechar:D \scan_stop:
+            \bool_set_false:N \l_expl_status_bool
+            \cs_set_protected_nopar:Npn \ExplSyntaxOff { }
+          }
+      }
+    \char_set_catcode_ignore:n           { 9 }   % tab
+    \char_set_catcode_ignore:n           { 32 }  % space
+    \char_set_catcode_other:n            { 34 }  % double quote
+    \char_set_catcode_alignment:n        { 38 }  % ampersand
+    \char_set_catcode_letter:n           { 58 }  % colon
+    \char_set_catcode_math_superscript:n { 94 }  % circumflex
+    \char_set_catcode_letter:n           { 95 }  % underscore
+    \char_set_catcode_other:n            { 124 } % pipe
+    \char_set_catcode_space:n            { 126 } % tilde
+    \tex_endlinechar:D = 32 \scan_stop:
+    \bool_set_true:N \l_expl_status_bool
+  }
+\cs_set_protected_nopar:Npx \ExplSyntaxOff
+  {
+    \exp_not:o \ExplSyntaxOff
+    \cs_set_protected_nopar:Npn \ExplSyntaxOff { }
+  }
+%% 
+%%
+%% End of file `expl3.sty'.
diff --git a/Master/texmf-dist/tex/latex/l3kernel/l3basics.sty b/Master/texmf-dist/tex/latex/l3kernel/l3basics.sty
new file mode 100644
index 00000000000..edecb52c1c7
--- /dev/null
+++ b/Master/texmf-dist/tex/latex/l3kernel/l3basics.sty
@@ -0,0 +1,784 @@
+%%
+%% This is file `l3basics.sty',
+%% generated with the docstrip utility.
+%%
+%% The original source files were:
+%%
+%% l3basics.dtx  (with options: `package')
+%% 
+%% EXPERIMENTAL CODE
+%% 
+%% Do not distribute this file without also distributing the
+%% source files specified above.
+%% 
+%% Do not distribute a modified version of this file.
+%% 
+%% File: l3basics.dtx Copyright (C) 1990-2011 The LaTeX3 project
+%%
+%% It may be distributed and/or modified under the conditions of the
+%% LaTeX Project Public License (LPPL), either version 1.3c of this
+%% license or (at your option) any later version.  The latest version
+%% of this license is in the file
+%%
+%%    http://www.latex-project.org/lppl.txt
+%%
+%% This file is part of the "expl3 bundle" (The Work in LPPL)
+%% and all files in that bundle must be distributed together.
+%%
+%% The released version of this bundle is available from CTAN.
+%%
+%% -----------------------------------------------------------------------
+%%
+%% The development version of the bundle can be found at
+%%
+%%    http://www.latex-project.org/svnroot/experimental/trunk/
+%%
+%% for those people who are interested.
+%%
+%%%%%%%%%%%
+%% NOTE: %%
+%%%%%%%%%%%
+%%
+%%   Snapshots taken from the repository represent work in progress and may
+%%   not work or may contain conflicting material!  We therefore ask
+%%   people _not_ to put them into distributions, archives, etc. without
+%%   prior consultation with the LaTeX3 Project.
+%%
+%% -----------------------------------------------------------------------
+\RequirePackage{l3names}
+\GetIdInfo$Id: l3basics.dtx 2493 2011-07-04 09:50:22Z bruno $
+  {L3 Experimental basic definitions}
+
+\ProvidesExplPackage
+  {\ExplFileName}{\ExplFileDate}{\ExplFileVersion}{\ExplFileDescription}
+\package_check_loaded_expl:
+\tex_let:D \cs_set_eq:NwN            \tex_let:D
+\cs_set_eq:NwN \if_true:           \tex_iftrue:D
+\cs_set_eq:NwN \if_false:          \tex_iffalse:D
+\cs_set_eq:NwN \or:                \tex_or:D
+\cs_set_eq:NwN \else:              \tex_else:D
+\cs_set_eq:NwN \fi:                \tex_fi:D
+\cs_set_eq:NwN \reverse_if:N       \etex_unless:D
+\cs_set_eq:NwN \if:w               \tex_if:D
+\cs_set_eq:NwN \if_bool:N          \tex_ifodd:D
+\cs_set_eq:NwN \if_predicate:w     \tex_ifodd:D
+\cs_set_eq:NwN \if_charcode:w      \tex_if:D
+\cs_set_eq:NwN \if_catcode:w       \tex_ifcat:D
+\cs_set_eq:NwN \if_meaning:w       \tex_ifx:D
+\cs_set_eq:NwN \if_mode_math:       \tex_ifmmode:D
+\cs_set_eq:NwN \if_mode_horizontal: \tex_ifhmode:D
+\cs_set_eq:NwN \if_mode_vertical:   \tex_ifvmode:D
+\cs_set_eq:NwN \if_mode_inner:      \tex_ifinner:D
+\cs_set_eq:NwN \if_cs_exist:N      \etex_ifdefined:D
+\cs_set_eq:NwN \if_cs_exist:w      \etex_ifcsname:D
+\cs_set_eq:NwN \exp_after:wN       \tex_expandafter:D
+\cs_set_eq:NwN \exp_not:N          \tex_noexpand:D
+\cs_set_eq:NwN \exp_not:n          \etex_unexpanded:D
+\cs_set_eq:NwN \token_to_meaning:N \tex_meaning:D
+\cs_set_eq:NwN \token_to_str:N     \tex_string:D
+\cs_set_eq:NwN \cs:w               \tex_csname:D
+\cs_set_eq:NwN \cs_end:            \tex_endcsname:D
+\cs_set_eq:NwN \cs_meaning:N       \tex_meaning:D
+\cs_set_eq:NwN \cs_show:N          \tex_show:D
+\cs_set_eq:NwN \scan_stop:         \tex_relax:D
+\cs_set_eq:NwN \group_begin:       \tex_begingroup:D
+\cs_set_eq:NwN \group_end:         \tex_endgroup:D
+\cs_set_eq:NwN \if_int_compare:w   \tex_ifnum:D
+\cs_set_eq:NwN \int_to_roman:w     \tex_romannumeral:D
+\cs_set_eq:NwN \group_insert_after:N \tex_aftergroup:D
+\cs_set_eq:NwN \pref_global:D      \tex_global:D
+\cs_set_eq:NwN \pref_long:D        \tex_long:D
+\cs_set_eq:NwN \pref_protected:D   \etex_protected:D
+\tex_long:D \tex_def:D \exp_args:Nc #1#2 { \exp_after:wN #1 \cs:w #2 \cs_end: }
+%%  A small number of variants by hand.
+\tex_def:D \cs_meaning:c   { \exp_args:Nc \cs_meaning:N }
+\tex_def:D \token_to_str:c { \exp_args:Nc \token_to_str:N }
+\tex_def:D \cs_show:c      { \exp_args:Nc \cs_show:N }
+\cs_set_eq:NwN \cs_set_nopar:Npn            \tex_def:D
+\cs_set_eq:NwN \cs_set_nopar:Npx            \tex_edef:D
+\pref_protected:D \cs_set_nopar:Npn \cs_set:Npn
+  { \pref_long:D \cs_set_nopar:Npn }
+\pref_protected:D \cs_set_nopar:Npn \cs_set:Npx
+  { \pref_long:D \cs_set_nopar:Npx }
+\pref_protected:D \cs_set_nopar:Npn \cs_set_protected_nopar:Npn
+  { \pref_protected:D \cs_set_nopar:Npn }
+\pref_protected:D \cs_set_nopar:Npn \cs_set_protected_nopar:Npx
+  { \pref_protected:D \cs_set_nopar:Npx }
+\cs_set_protected_nopar:Npn \cs_set_protected:Npn
+  { \pref_protected:D \pref_long:D \cs_set_nopar:Npn }
+\cs_set_protected_nopar:Npn \cs_set_protected:Npx
+  { \pref_protected:D \pref_long:D \cs_set_nopar:Npx }
+\cs_set_eq:NwN \cs_gset_nopar:Npn           \tex_gdef:D
+\cs_set_eq:NwN \cs_gset_nopar:Npx           \tex_xdef:D
+\cs_set_protected_nopar:Npn \cs_gset:Npn
+  { \pref_long:D \cs_gset_nopar:Npn }
+\cs_set_protected_nopar:Npn \cs_gset:Npx
+  { \pref_long:D \cs_gset_nopar:Npx }
+\cs_set_protected_nopar:Npn \cs_gset_protected_nopar:Npn
+  { \pref_protected:D \cs_gset_nopar:Npn }
+\cs_set_protected_nopar:Npn \cs_gset_protected_nopar:Npx
+  { \pref_protected:D \cs_gset_nopar:Npx }
+\cs_set_protected_nopar:Npn \cs_gset_protected:Npn
+  { \pref_protected:D \pref_long:D \cs_gset_nopar:Npn }
+\cs_set_protected_nopar:Npn \cs_gset_protected:Npx
+  { \pref_protected:D \pref_long:D \cs_gset_nopar:Npx }
+\cs_set:Npn \use:c #1 { \cs:w #1 \cs_end: }
+\cs_set_protected:Npn \use:x #1
+  {
+    \group_begin:
+      \cs_set:Npx \cs_tmp:w {#1}
+      \exp_after:wN
+    \group_end:
+    \cs_tmp:w
+  }
+\cs_set:Npn \cs_tmp:w { }
+\cs_set:Npn \use:n    #1       {#1}
+\cs_set:Npn \use:nn   #1#2     {#1#2}
+\cs_set:Npn \use:nnn  #1#2#3   {#1#2#3}
+\cs_set:Npn \use:nnnn #1#2#3#4 {#1#2#3#4}
+\cs_set:Npn \use_i:nn  #1#2 {#1}
+\cs_set:Npn \use_ii:nn #1#2 {#2}
+\cs_set:Npn \use_i:nnn    #1#2#3 {#1}
+\cs_set:Npn \use_ii:nnn   #1#2#3 {#2}
+\cs_set:Npn \use_iii:nnn  #1#2#3 {#3}
+\cs_set:Npn \use_i_ii:nnn #1#2#3 {#1#2}
+\cs_set:Npn \use_i:nnnn   #1#2#3#4 {#1}
+\cs_set:Npn \use_ii:nnnn  #1#2#3#4 {#2}
+\cs_set:Npn \use_iii:nnnn #1#2#3#4 {#3}
+\cs_set:Npn \use_iv:nnnn  #1#2#3#4 {#4}
+\cs_set:Npn \use_none_delimit_by_q_nil:w  #1 \q_nil  { }
+\cs_set:Npn \use_none_delimit_by_q_stop:w #1 \q_stop { }
+\cs_set:Npn \use_none_delimit_by_q_recursion_stop:w #1 \q_recursion_stop { }
+\cs_set:Npn \use_i_delimit_by_q_nil:nw  #1#2 \q_nil  {#1}
+\cs_set:Npn \use_i_delimit_by_q_stop:nw #1#2 \q_stop {#1}
+\cs_set:Npn \use_i_delimit_by_q_recursion_stop:nw #1#2 \q_recursion_stop {#1}
+\cs_set:Npn \use_i_after_fi:nw #1 \fi: { \fi: #1 }
+\cs_set:Npn \use_i_after_else:nw #1 \else: #2 \fi: { \fi: #1 }
+\cs_set:Npn \use_i_after_or:nw #1 \or: #2 \fi: { \fi: #1 }
+\cs_set:Npn \use_i_after_orelse:nw #1#2#3 \fi: { \fi: #1 }
+\cs_set:Npn \use_none:n         #1                 { }
+\cs_set:Npn \use_none:nn        #1#2               { }
+\cs_set:Npn \use_none:nnn       #1#2#3             { }
+\cs_set:Npn \use_none:nnnn      #1#2#3#4           { }
+\cs_set:Npn \use_none:nnnnn     #1#2#3#4#5         { }
+\cs_set:Npn \use_none:nnnnnn    #1#2#3#4#5#6       { }
+\cs_set:Npn \use_none:nnnnnnn   #1#2#3#4#5#6#7     { }
+\cs_set:Npn \use_none:nnnnnnnn  #1#2#3#4#5#6#7#8   { }
+\cs_set:Npn \use_none:nnnnnnnnn #1#2#3#4#5#6#7#8#9 { }
+\cs_set_nopar:Npn \prg_return_true:
+  { \exp_after:wN \use_i:nn  \int_to_roman:w }
+\cs_set_nopar:Npn \prg_return_false:
+  { \exp_after:wN \use_ii:nn \int_to_roman:w}
+\cs_set_protected:Npn \prg_set_conditional:Npnn #1
+  {
+    \prg_get_parm_aux:nw
+      {
+        \cs_split_function:NN #1 \prg_generate_conditional_aux:nnNNnnnn
+        \cs_set:Npn { parm }
+      }
+  }
+\cs_set_protected:Npn \prg_new_conditional:Npnn #1
+  {
+    \prg_get_parm_aux:nw
+      {
+        \cs_split_function:NN #1 \prg_generate_conditional_aux:nnNNnnnn
+        \cs_new:Npn { parm }
+      }
+  }
+\cs_set_protected:Npn \prg_set_protected_conditional:Npnn #1
+  {
+    \prg_get_parm_aux:nw{
+      \cs_split_function:NN #1 \prg_generate_conditional_aux:nnNNnnnn
+      \cs_set_protected:Npn { parm }
+    }
+  }
+\cs_set_protected:Npn \prg_new_protected_conditional:Npnn #1
+  {
+    \prg_get_parm_aux:nw
+      {
+        \cs_split_function:NN #1 \prg_generate_conditional_aux:nnNNnnnn
+        \cs_new_protected:Npn { parm }
+     }
+  }
+\cs_set_protected:Npn \prg_set_conditional:Nnn #1
+  {
+    \exp_args:Nnf \prg_get_count_aux:nn
+      {
+        \cs_split_function:NN #1 \prg_generate_conditional_aux:nnNNnnnn
+        \cs_set:Npn { count }
+      }
+      { \cs_get_arg_count_from_signature:N #1 }
+  }
+\cs_set_protected:Npn \prg_new_conditional:Nnn #1
+  {
+    \exp_args:Nnf \prg_get_count_aux:nn
+      {
+        \cs_split_function:NN #1 \prg_generate_conditional_aux:nnNNnnnn
+        \cs_new:Npn { count}
+      }
+      { \cs_get_arg_count_from_signature:N #1 }
+  }
+
+\cs_set_protected:Npn \prg_set_protected_conditional:Nnn #1{
+  \exp_args:Nnf \prg_get_count_aux:nn{
+    \cs_split_function:NN #1 \prg_generate_conditional_aux:nnNNnnnn
+    \cs_set_protected:Npn {count}
+  }{\cs_get_arg_count_from_signature:N #1}
+}
+
+\cs_set_protected:Npn \prg_new_protected_conditional:Nnn #1
+  {
+    \exp_args:Nnf \prg_get_count_aux:nn
+      {
+        \cs_split_function:NN #1 \prg_generate_conditional_aux:nnNNnnnn
+        \cs_new_protected:Npn {count}
+      }
+      { \cs_get_arg_count_from_signature:N #1 }
+  }
+\cs_set_protected:Npn \prg_set_eq_conditional:NNn #1#2#3
+  { \prg_set_eq_conditional_aux:NNNn \cs_set_eq:cc #1#2 {#3} }
+\cs_set_protected:Npn \prg_new_eq_conditional:NNn #1#2#3
+  { \prg_set_eq_conditional_aux:NNNn \cs_new_eq:cc #1#2 {#3} }
+\cs_set:Npn \prg_get_count_aux:nn #1#2 { #1 {#2} }
+\cs_set:Npn \prg_get_parm_aux:nw #1#2# { #1 {#2} }
+\cs_set_protected:Npn \prg_generate_conditional_aux:nnNNnnnn #1#2#3#4#5#6#7#8
+  {
+    \prg_generate_conditional_aux:nnw {#5}
+      {
+        #4 {#1} {#2} {#6} {#8}
+      }
+      #7 , ? , \q_recursion_stop
+  }
+\cs_set_protected:Npn \prg_generate_conditional_aux:nnw #1#2#3 ,
+  {
+    \if:w ?#3
+      \exp_after:wN \use_none_delimit_by_q_recursion_stop:w
+    \fi:
+    \use:c { prg_generate_#3_form_#1:Nnnnn } #2
+    \prg_generate_conditional_aux:nnw {#1} {#2}
+  }
+\cs_set_protected:Npn \prg_generate_p_form_parm:Nnnnn #1#2#3#4#5
+  {
+    \exp_args:Nc #1 { #2 _p: #3 } #4
+      {
+        #5 \c_zero
+        \c_true_bool \c_false_bool
+      }
+  }
+\cs_set_protected:Npn \prg_generate_T_form_parm:Nnnnn #1#2#3#4#5
+  {
+    \exp_args:Nc #1 { #2 : #3 T } #4
+      {
+        #5 \c_zero
+        \use:n \use_none:n
+      }
+  }
+\cs_set_protected:Npn \prg_generate_F_form_parm:Nnnnn #1#2#3#4#5
+  {
+    \exp_args:Nc #1 { #2 : #3 F } #4
+      {
+        #5 \c_zero
+        { }
+      }
+  }
+\cs_set_protected:Npn \prg_generate_TF_form_parm:Nnnnn #1#2#3#4#5
+  {
+    \exp_args:Nc #1 { #2 : #3 TF } #4
+      { #5 \c_zero }
+  }
+\cs_set_protected:Npn \prg_generate_p_form_count:Nnnnn #1#2#3#4#5
+  {
+    \cs_generate_from_arg_count:cNnn { #2 _p: #3 } #1 {#4}
+      {
+        #5 \c_zero
+        \c_true_bool \c_false_bool
+      }
+  }
+\cs_set_protected:Npn \prg_generate_T_form_count:Nnnnn #1#2#3#4#5
+  {
+    \cs_generate_from_arg_count:cNnn { #2 : #3 T } #1 {#4}
+      {
+        #5 \c_zero
+        \use:n \use_none:n
+      }
+  }
+\cs_set_protected:Npn \prg_generate_F_form_count:Nnnnn #1#2#3#4#5
+  {
+    \cs_generate_from_arg_count:cNnn { #2 : #3 F } #1 {#4}
+      {
+        #5 \c_zero
+        { }
+      }
+  }
+\cs_set_protected:Npn \prg_generate_TF_form_count:Nnnnn #1#2#3#4#5
+  {
+    \cs_generate_from_arg_count:cNnn { #2 : #3 TF } #1 {#4}
+      { #5 \c_zero }
+  }
+\cs_set_protected:Npn \prg_set_eq_conditional_aux:NNNn #1#2#3#4
+ { \prg_set_eq_conditional_aux:NNNw #1#2#3#4 , ? , \q_recursion_stop }
+\cs_set_protected:Npn \prg_set_eq_conditional_aux:NNNw #1#2#3#4 ,
+  {
+    \if:w ? #4 \scan_stop:
+      \exp_after:wN \use_none_delimit_by_q_recursion_stop:w
+    \fi:
+    #1
+    { \exp_args:NNc \cs_split_function:NN #2 { prg_conditional_form_#4:nnn } }
+    { \exp_args:NNc \cs_split_function:NN #3 { prg_conditional_form_#4:nnn } }
+    \prg_set_eq_conditional_aux:NNNw #1 {#2} {#3}
+  }
+\cs_set:Npn \prg_conditional_form_p:nnn  #1#2#3 { #1 _p : #2 }
+\cs_set:Npn \prg_conditional_form_TF:nnn #1#2#3 { #1    : #2 TF }
+\cs_set:Npn \prg_conditional_form_T:nnn  #1#2#3 { #1    : #2 T }
+\cs_set:Npn \prg_conditional_form_F:nnn  #1#2#3 { #1    : #2 F }
+\tex_chardef:D \c_true_bool  = 1~
+\tex_chardef:D \c_false_bool = 0~
+\cs_set_nopar:Npn \cs_to_str:N
+  {
+    \if:w \int_to_roman:w - `0 \token_to_str:N \ %
+      \cs_to_str_aux:w
+    \fi:
+    \exp_after:wN \use_none:n \token_to_str:N
+ }
+\cs_set_nopar:Npn \cs_to_str_aux:w #1 \use_none:n
+  { ~ \int_to_roman:w - `0 \fi: }
+\group_begin:
+  \tex_lccode:D  `\@ = `\: \scan_stop:
+  \tex_catcode:D `\@ = 12~
+\tex_lowercase:D
+  {
+    \group_end:
+    \cs_set:Npn \cs_split_function:NN #1#2
+      {
+        \exp_after:wN \cs_split_function_aux:w
+          \int_to_roman:w - `\q \cs_to_str:N #1 @ a \q_stop #2
+      }
+    \cs_set:Npn \cs_split_function_aux:w #1 @ #2#3 \q_stop #4
+      {
+        \if_meaning:w a #2
+          \exp_after:wN \use_i:nn
+        \else:
+          \exp_after:wN\use_ii:nn
+        \fi:
+        { #4 {#1} { } \c_false_bool }
+        { \cs_split_function_auxii:w #2#3 \q_stop #4 {#1} }
+      }
+    \cs_set:Npn \cs_split_function_auxii:w #1 @a \q_stop #2#3
+      { #2{#3}{#1}\c_true_bool }
+  }
+\cs_set:Npn \cs_get_function_name:N #1
+  { \cs_split_function:NN #1 \use_i:nnn }
+\cs_set:Npn \cs_get_function_signature:N #1
+  { \cs_split_function:NN #1 \use_ii:nnn }
+\prg_set_conditional:Npnn \cs_if_exist:N #1 { p , T , F , TF }
+  {
+    \if_meaning:w #1 \scan_stop:
+      \prg_return_false:
+    \else:
+      \if_cs_exist:N #1
+        \prg_return_true:
+      \else:
+        \prg_return_false:
+      \fi:
+    \fi:
+  }
+\prg_set_conditional:Npnn \cs_if_exist:c #1 { p , T , F , TF }
+  {
+    \if_cs_exist:w #1 \cs_end:
+      \exp_after:wN \use_i:nn
+    \else:
+      \exp_after:wN \use_ii:nn
+    \fi:
+    {
+      \exp_after:wN \if_meaning:w \cs:w #1 \cs_end: \scan_stop:
+        \prg_return_false:
+      \else:
+        \prg_return_true:
+      \fi:
+    }
+    \prg_return_false:
+  }
+\prg_set_conditional:Npnn \cs_if_free:N #1 { p , T , F , TF }
+  {
+    \if_meaning:w #1 \scan_stop:
+      \prg_return_true:
+    \else:
+      \if_cs_exist:N #1
+        \prg_return_false:
+      \else:
+        \prg_return_true:
+      \fi:
+    \fi:
+  }
+\prg_set_conditional:Npnn \cs_if_free:c #1 { p , T , F , TF }
+  {
+    \if_cs_exist:w #1 \cs_end:
+      \exp_after:wN \use_i:nn
+    \else:
+      \exp_after:wN \use_ii:nn
+    \fi:
+      {
+        \exp_after:wN \if_meaning:w \cs:w #1 \cs_end: \scan_stop:
+          \prg_return_true:
+        \else:
+          \prg_return_false:
+        \fi:
+      }
+      { \prg_return_true: }
+  }
+\cs_set_eq:NwN \c_minus_one \m@ne
+\tex_chardef:D \c_sixteen = 16~
+\tex_chardef:D \c_zero    = 0~
+\tex_chardef:D \c_six     = 6~
+\tex_chardef:D \c_seven   = 7~
+\tex_chardef:D \c_twelve  = 12~
+\tex_mathchardef:D \c_max_register_int = 32 767 \scan_stop:
+\cs_set_protected_nopar:Npn \iow_log:x
+  { \tex_immediate:D \tex_write:D \c_minus_one }
+\cs_set_protected_nopar:Npn \iow_term:x
+  { \tex_immediate:D \tex_write:D \c_sixteen }
+\cs_set_protected_nopar:Npn \msg_kernel_error:nnxx #1#2#3#4
+  {
+    \tex_errmessage:D
+      {
+        !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!~! ^^J
+        Argh,~internal~LaTeX3~error! ^^J ^^J
+        Module ~ #1 , ~ message~name~"#2": ^^J
+        Arguments~'#3'~and~'#4' ^^J ^^J
+        This~is~one~for~The~LaTeX3~Project:~bailing~out
+      }
+    \tex_end:D
+  }
+\cs_set_protected_nopar:Npn \msg_kernel_error:nnx #1#2#3
+  { \msg_kernel_error:nnxx {#1} {#2} {#3} { } }
+\cs_set_protected_nopar:Npn \msg_kernel_error:nn #1#2
+  { \msg_kernel_error:nnxx {#1} {#2} { } { } }
+\cs_set_nopar:Npn \msg_line_context:
+  { on~line~\tex_the:D \tex_inputlineno:D }
+\cs_set:Npn \cs_record_meaning:N #1 { }
+\cs_set_protected_nopar:Npn \chk_if_free_cs:N #1
+  {
+    \cs_if_free:NF #1
+      {
+        \msg_kernel_error:nnxx { kernel } { command-already-defined }
+          { \token_to_str:N #1 } { \token_to_meaning:N #1 }
+      }
+}
+\tex_ifodd:D \@l@expl@log@functions@bool
+  \cs_set_protected_nopar:Npn \chk_if_free_cs:N #1
+    {
+      \cs_if_free:NF #1
+        {
+          \msg_kernel_error:nnxx { kernel } { command-already-defined }
+            { \token_to_str:N #1 } { \token_to_meaning:N #1 }
+        }
+    \iow_log:x { Defining~\token_to_str:N #1~ \msg_line_context: }
+  }
+\fi:
+\cs_set_protected_nopar:Npn \chk_if_free_cs:c
+  { \exp_args:Nc \chk_if_free_cs:N }
+\cs_set_protected_nopar:Npn \chk_if_exist_cs:N #1
+  {
+    \cs_if_exist:NF #1
+      {
+        \msg_kernel_error:nnxx { kernel } { command-not-defined }
+          { \token_to_str:N #1 } { \token_to_meaning:N #1 }
+      }
+  }
+\cs_set_protected_nopar:Npn \chk_if_exist_cs:c
+  { \exp_args:Nc \chk_if_exist_cs:N }
+\cs_set:Npn \cs_tmp:w #1#2
+  {
+    \cs_set_protected_nopar:Npn #1 ##1
+       {
+         \chk_if_free_cs:N ##1
+         #2 ##1
+      }
+  }
+\cs_tmp:w \cs_new_nopar:Npn           \cs_gset_nopar:Npn
+\cs_tmp:w \cs_new_nopar:Npx           \cs_gset_nopar:Npx
+\cs_tmp:w \cs_new:Npn                 \cs_gset:Npn
+\cs_tmp:w \cs_new:Npx                 \cs_gset:Npx
+\cs_tmp:w \cs_new_protected_nopar:Npn \cs_gset_protected_nopar:Npn
+\cs_tmp:w \cs_new_protected_nopar:Npx \cs_gset_protected_nopar:Npx
+\cs_tmp:w \cs_new_protected:Npn       \cs_gset_protected:Npn
+\cs_tmp:w \cs_new_protected:Npx       \cs_gset_protected:Npx
+\cs_set:Npn \cs_tmp:w #1#2
+  { \cs_new_protected_nopar:Npn #1 { \exp_args:Nc #2 } }
+\cs_tmp:w \cs_set_nopar:cpn  \cs_set_nopar:Npn
+\cs_tmp:w \cs_set_nopar:cpx  \cs_set_nopar:Npx
+\cs_tmp:w \cs_gset_nopar:cpn \cs_gset_nopar:Npn
+\cs_tmp:w \cs_gset_nopar:cpx \cs_gset_nopar:Npx
+\cs_tmp:w \cs_new_nopar:cpn  \cs_new_nopar:Npn
+\cs_tmp:w \cs_new_nopar:cpx  \cs_new_nopar:Npx
+\cs_tmp:w \cs_set:cpn  \cs_set:Npn
+\cs_tmp:w \cs_set:cpx  \cs_set:Npx
+\cs_tmp:w \cs_gset:cpn \cs_gset:Npn
+\cs_tmp:w \cs_gset:cpx \cs_gset:Npx
+\cs_tmp:w \cs_new:cpn  \cs_new:Npn
+\cs_tmp:w \cs_new:cpx  \cs_new:Npx
+\cs_tmp:w \cs_set_protected_nopar:cpn  \cs_set_protected_nopar:Npn
+\cs_tmp:w \cs_set_protected_nopar:cpx  \cs_set_protected_nopar:Npx
+\cs_tmp:w \cs_gset_protected_nopar:cpn \cs_gset_protected_nopar:Npn
+\cs_tmp:w \cs_gset_protected_nopar:cpx \cs_gset_protected_nopar:Npx
+\cs_tmp:w \cs_new_protected_nopar:cpn  \cs_new_protected_nopar:Npn
+\cs_tmp:w \cs_new_protected_nopar:cpx  \cs_new_protected_nopar:Npx
+\cs_tmp:w \cs_set_protected:cpn  \cs_set_protected:Npn
+\cs_tmp:w \cs_set_protected:cpx  \cs_set_protected:Npx
+\cs_tmp:w \cs_gset_protected:cpn \cs_gset_protected:Npn
+\cs_tmp:w \cs_gset_protected:cpx \cs_gset_protected:Npx
+\cs_tmp:w \cs_new_protected:cpn  \cs_new_protected:Npn
+\cs_tmp:w \cs_new_protected:cpx  \cs_new_protected:Npx
+\cs_new_nopar:cpn { use_0_parameter: } { }
+\cs_new_nopar:cpn { use_1_parameter: } { {##1} }
+\cs_new_nopar:cpn { use_2_parameter: } { {##1} {##2} }
+\cs_new_nopar:cpn { use_3_parameter: } { {##1} {##2} {##3} }
+\cs_new_nopar:cpn { use_4_parameter: } { {##1} {##2} {##3} {##4} }
+\cs_new_nopar:cpn { use_5_parameter: } { {##1} {##2} {##3} {##4} {##5} }
+\cs_new_nopar:cpn { use_6_parameter: } { {##1} {##2} {##3} {##4} {##5} {##6} }
+\cs_new_nopar:cpn { use_7_parameter: }
+  { {##1} {##2} {##3} {##4} {##5}{##6} {##7} }
+\cs_new_nopar:cpn { use_8_parameter: }
+  { {##1} {##2} {##3} {##4} {##5} {##6} {##7} {##8} }
+\cs_new_nopar:cpn{ use_9_parameter: }
+  { {##1} {##2} {##3} {##4} {##5} {##6} {##7} {##8} {##9} }
+\cs_new_protected:Npn \cs_set_eq:NN #1 { \cs_set_eq:NwN #1 =~ }
+\cs_new_protected_nopar:Npn \cs_set_eq:cN { \exp_args:Nc  \cs_set_eq:NN }
+\cs_new_protected_nopar:Npn \cs_set_eq:Nc { \exp_args:NNc \cs_set_eq:NN }
+\cs_new_protected_nopar:Npn \cs_set_eq:cc { \exp_args:Ncc \cs_set_eq:NN }
+\cs_new_protected:Npn \cs_new_eq:NN #1
+  {
+    \chk_if_free_cs:N #1
+    \pref_global:D \cs_set_eq:NN #1
+  }
+\cs_new_protected_nopar:Npn \cs_new_eq:cN { \exp_args:Nc  \cs_new_eq:NN }
+\cs_new_protected_nopar:Npn \cs_new_eq:Nc { \exp_args:NNc \cs_new_eq:NN }
+\cs_new_protected_nopar:Npn \cs_new_eq:cc { \exp_args:Ncc \cs_new_eq:NN }
+\cs_new_protected_nopar:Npn \cs_gset_eq:NN { \pref_global:D  \cs_set_eq:NN }
+\cs_new_protected_nopar:Npn \cs_gset_eq:Nc { \exp_args:NNc  \cs_gset_eq:NN }
+\cs_new_protected_nopar:Npn \cs_gset_eq:cN { \exp_args:Nc   \cs_gset_eq:NN }
+\cs_new_protected_nopar:Npn \cs_gset_eq:cc { \exp_args:Ncc  \cs_gset_eq:NN }
+\cs_new_protected_nopar:Npn \cs_undefine:N #1
+  { \cs_gset_eq:NN #1 \c_undefined:D }
+\cs_new_protected_nopar:Npn \cs_undefine:c #1
+  { \cs_gset_eq:cN {#1} \c_undefined:D }
+\cs_new:Npn \cs_get_arg_count_from_signature:N #1
+  { \cs_split_function:NN #1 \cs_get_arg_count_from_signature_aux:nnN }
+\cs_new:Npn \cs_get_arg_count_from_signature_aux:nnN #1#2#3
+  {
+    \if_predicate:w #3
+      \exp_after:wN \use_i:nn
+    \else:
+      \exp_after:wN\use_ii:nn
+    \fi:
+    {
+      \exp_after:wN \cs_get_arg_count_from_signature_auxii:w
+        \use_none:nnnnnnnnn  #2 9876543210 \q_stop
+    }
+    { -1 }
+}
+\cs_new:Npn \cs_get_arg_count_from_signature_auxii:w #1#2 \q_stop {#1}
+\cs_new_nopar:Npn \cs_get_arg_count_from_signature:c
+  { \exp_args:Nc \cs_get_arg_count_from_signature:N }
+\cs_new_protected:Npn \cs_generate_from_arg_count:NNnn #1#2#3#4
+  {
+    \if_case:w \int_eval:w #3 \int_eval_end:
+      \use_i_after_orelse:nw {#2#1}
+    \or:
+      \use_i_after_orelse:nw {#2#1 ##1}
+    \or:
+      \use_i_after_orelse:nw {#2#1 ##1##2}
+    \or:
+      \use_i_after_orelse:nw {#2#1 ##1##2##3}
+    \or:
+      \use_i_after_orelse:nw {#2#1 ##1##2##3##4}
+    \or:
+      \use_i_after_orelse:nw {#2#1 ##1##2##3##4##5}
+    \or:
+      \use_i_after_orelse:nw {#2#1 ##1##2##3##4##5##6}
+    \or:
+      \use_i_after_orelse:nw {#2#1 ##1##2##3##4##5##6##7}
+    \or:
+      \use_i_after_orelse:nw {#2#1 ##1##2##3##4##5##6##7##8}
+    \or:
+      \use_i_after_orelse:nw {#2#1 ##1##2##3##4##5##6##7##8##9}
+    \else:
+      \use_i_after_fi:nw
+        {
+          \cs_generate_from_arg_count_error_msg:Nn #1 {#3}
+          \use_none:n
+        }
+    \fi:
+    {#4}
+  }
+\cs_new_nopar:Npn \cs_generate_from_arg_count:cNnn
+  { \exp_args:Nc \cs_generate_from_arg_count:NNnn }
+\cs_new:Npn \cs_generate_from_arg_count_error_msg:Nn #1#2
+  {
+    \msg_kernel_error:nnxx { kernel } { bad-number-of-arguments }
+      { \token_to_str:N #1 } { \int_eval:n {#2} }
+  }
+\cs_set:Npn \cs_tmp:w #1#2#3
+  {
+    \cs_set_protected:cpx { cs_ #1 : #2 } ##1##2
+      {
+        \exp_not:N \cs_generate_from_arg_count:NNnn ##1
+        \exp_after:wN \exp_not:N \cs:w cs_#1 : #3 \cs_end:
+          { \exp_not:N\cs_get_arg_count_from_signature:N ##1 }{##2}
+      }
+  }
+\cs_tmp:w { set }                  { Nn } { Npn }
+\cs_tmp:w { set }                  { Nx } { Npx }
+\cs_tmp:w { set_nopar }            { Nn } { Npn }
+\cs_tmp:w { set_nopar }            { Nx } { Npx }
+\cs_tmp:w { set_protected }        { Nn } { Npn }
+\cs_tmp:w { set_protected }        { Nx } { Npx }
+\cs_tmp:w { set_protected_nopar }  { Nn } { Npn }
+\cs_tmp:w { set_protected_nopar }  { Nx } { Npx }
+\cs_tmp:w { gset }                 { Nn } { Npn }
+\cs_tmp:w { gset }                 { Nx } { Npx }
+\cs_tmp:w { gset_nopar }           { Nn } { Npn }
+\cs_tmp:w { gset_nopar }           { Nx } { Npx }
+\cs_tmp:w { gset_protected }       { Nn } { Npn }
+\cs_tmp:w { gset_protected }       { Nx } { Npx }
+\cs_tmp:w { gset_protected_nopar } { Nn } { Npn }
+\cs_tmp:w { gset_protected_nopar } { Nx } { Npx }
+\cs_tmp:w { new }                 { Nn } { Npn }
+\cs_tmp:w { new }                 { Nx } { Npx }
+\cs_tmp:w { new_nopar }           { Nn } { Npn }
+\cs_tmp:w { new_nopar }           { Nx } { Npx }
+\cs_tmp:w { new_protected }       { Nn } { Npn }
+\cs_tmp:w { new_protected }       { Nx } { Npx }
+\cs_tmp:w { new_protected_nopar } { Nn } { Npn }
+\cs_tmp:w { new_protected_nopar } { Nx } { Npx }
+\cs_set:Npn \cs_tmp:w #1#2#3
+  {
+    \cs_set_protected:cpx {cs_#1:#2}##1##2{
+      \exp_not:N\cs_generate_from_arg_count:cNnn {##1}
+      \exp_after:wN \exp_not:N \cs:w cs_#1:#3 \cs_end:
+        { \exp_not:N \cs_get_arg_count_from_signature:c {##1} } {##2}
+    }
+  }
+\cs_tmp:w { set }                  { cn } { Npn }
+\cs_tmp:w { set }                  { cx } { Npx }
+\cs_tmp:w { set_nopar }            { cn } { Npn }
+\cs_tmp:w { set_nopar }            { cx } { Npx }
+\cs_tmp:w { set_protected }        { cn } { Npn }
+\cs_tmp:w { set_protected }        { cx } { Npx }
+\cs_tmp:w { set_protected_nopar }  { cn } { Npn }
+\cs_tmp:w { set_protected_nopar }  { cx } { Npx }
+\cs_tmp:w { gset }                 { cn } { Npn }
+\cs_tmp:w { gset }                 { cx } { Npx }
+\cs_tmp:w { gset_nopar }           { cn } { Npn }
+\cs_tmp:w { gset_nopar }           { cx } { Npx }
+\cs_tmp:w { gset_protected }       { cn } { Npn }
+\cs_tmp:w { gset_protected }       { cx } { Npx }
+\cs_tmp:w { gset_protected_nopar } { cn } { Npn }
+\cs_tmp:w { gset_protected_nopar } { cx } { Npx }
+\cs_tmp:w { new }                 { cn } { Npn }
+\cs_tmp:w { new }                 { cx } { Npx }
+\cs_tmp:w { new_nopar }           { cn } { Npn }
+\cs_tmp:w { new_nopar }           { cx } { Npx }
+\cs_tmp:w { new_protected }       { cn } { Npn }
+\cs_tmp:w { new_protected }       { cx } { Npx }
+\cs_tmp:w { new_protected_nopar } { cn } { Npn }
+\cs_tmp:w { new_protected_nopar } { cx } { Npx }
+\prg_new_conditional:Npnn \cs_if_eq:NN #1#2 { p , T , F , TF }
+  {
+    \if_meaning:w #1#2
+      \prg_return_true: \else: \prg_return_false: \fi:
+  }
+\cs_new_nopar:Npn \cs_if_eq_p:cN { \exp_args:Nc  \cs_if_eq_p:NN }
+\cs_new_nopar:Npn \cs_if_eq:cNTF { \exp_args:Nc  \cs_if_eq:NNTF }
+\cs_new_nopar:Npn \cs_if_eq:cNT  { \exp_args:Nc  \cs_if_eq:NNT }
+\cs_new_nopar:Npn \cs_if_eq:cNF  { \exp_args:Nc  \cs_if_eq:NNF }
+\cs_new_nopar:Npn \cs_if_eq_p:Nc { \exp_args:NNc \cs_if_eq_p:NN }
+\cs_new_nopar:Npn \cs_if_eq:NcTF { \exp_args:NNc \cs_if_eq:NNTF }
+\cs_new_nopar:Npn \cs_if_eq:NcT  { \exp_args:NNc \cs_if_eq:NNT }
+\cs_new_nopar:Npn \cs_if_eq:NcF  { \exp_args:NNc \cs_if_eq:NNF }
+\cs_new_nopar:Npn \cs_if_eq_p:cc { \exp_args:Ncc \cs_if_eq_p:NN }
+\cs_new_nopar:Npn \cs_if_eq:ccTF { \exp_args:Ncc \cs_if_eq:NNTF }
+\cs_new_nopar:Npn \cs_if_eq:ccT  { \exp_args:Ncc \cs_if_eq:NNT }
+\cs_new_nopar:Npn \cs_if_eq:ccF  { \exp_args:Ncc \cs_if_eq:NNF }
+\cs_new_nopar:Npn \kernel_register_show:N #1
+  {
+    \cs_if_exist:NTF #1
+     { \tex_showthe:D #1 }
+     {
+       \msg_kernel_error:nnx { kernel } { variable-not-defined }
+         { \token_to_str:N #1 }
+     }
+  }
+\cs_new_nopar:Npn \kernel_register_show:c { \exp_args:Nc \int_show:N }
+\cs_new_eq:NN \luatex_if_engine:T  \use_none:n
+\cs_new_eq:NN \luatex_if_engine:F  \use:n
+\cs_new_eq:NN \luatex_if_engine:TF \use_ii:nn
+\cs_new_eq:NN \pdftex_if_engine:T  \use:n
+\cs_new_eq:NN \pdftex_if_engine:F  \use_none:n
+\cs_new_eq:NN \pdftex_if_engine:TF \use_i:nn
+\cs_new_eq:NN \xetex_if_engine:T   \use_none:n
+\cs_new_eq:NN \xetex_if_engine:F   \use:n
+\cs_new_eq:NN \xetex_if_engine:TF  \use_ii:nn
+\cs_new_eq:NN \c_luatex_is_engine_bool \c_false_bool
+\cs_new_eq:NN \c_pdftex_is_engine_bool \c_true_bool
+\cs_new_eq:NN \c_xetex_is_engine_bool  \c_false_bool
+\cs_if_exist:NT \xetex_XeTeXversion:D
+  {
+    \cs_set_eq:NN \pdftex_if_engine:T  \use_none:n
+    \cs_set_eq:NN \pdftex_if_engine:F  \use:n
+    \cs_set_eq:NN \pdftex_if_engine:TF \use_ii:nn
+    \cs_set_eq:NN \xetex_if_engine:T   \use:n
+    \cs_set_eq:NN \xetex_if_engine:F   \use_none:n
+    \cs_set_eq:NN \xetex_if_engine:TF  \use_i:nn
+    \cs_set_eq:NN \c_pdftex_is_engine_bool \c_false_bool
+    \cs_set_eq:NN \c_xetex_is_engine_bool  \c_true_bool
+  }
+\cs_if_exist:NT \luatex_directlua:D
+  {
+    \cs_set_eq:NN \luatex_if_engine:T  \use:n
+    \cs_set_eq:NN \luatex_if_engine:F  \use_none:n
+    \cs_set_eq:NN \luatex_if_engine:TF \use_i:nn
+    \cs_set_eq:NN \pdftex_if_engine:T  \use_none:n
+    \cs_set_eq:NN \pdftex_if_engine:F  \use:n
+    \cs_set_eq:NN \pdftex_if_engine:TF \use_ii:nn
+    \cs_set_eq:NN \c_luatex_is_engine_bool \c_true_bool
+    \cs_set_eq:NN \c_pdftex_is_engine_bool \c_false_bool
+  }
+\cs_new_nopar:Npn \prg_do_nothing: { }
+\prg_new_conditional:Npnn \str_if_eq:nn #1#2 { p , T , F , TF }
+  {
+    \if_int_compare:w \pdftex_strcmp:D { \exp_not:n {#1} } { \exp_not:n {#2} }
+      = \c_zero
+     \prg_return_true: \else: \prg_return_false: \fi:
+  }
+\prg_new_conditional:Npnn \str_if_eq:xx #1#2 { p , T , F , TF }
+  {
+    \if_int_compare:w \pdftex_strcmp:D {#1} {#2} = \c_zero
+      \prg_return_true: \else: \prg_return_false: \fi:
+  }
+\cs_new_eq:NN           \cs_gnew_nopar:Npn            \cs_new_nopar:Npn
+\cs_new_eq:NN                 \cs_gnew:Npn                  \cs_new:Npn
+\cs_new_eq:NN \cs_gnew_protected_nopar:Npn  \cs_new_protected_nopar:Npn
+\cs_new_eq:NN       \cs_gnew_protected:Npn        \cs_new_protected:Npn
+\cs_new_eq:NN           \cs_gnew_nopar:Npx            \cs_new_nopar:Npx
+\cs_new_eq:NN                 \cs_gnew:Npx                  \cs_new:Npx
+\cs_new_eq:NN \cs_gnew_protected_nopar:Npx  \cs_new_protected_nopar:Npx
+\cs_new_eq:NN       \cs_gnew_protected:Npx        \cs_new_protected:Npx
+\cs_new_eq:NN           \cs_gnew_nopar:cpn            \cs_new_nopar:cpn
+\cs_new_eq:NN                 \cs_gnew:cpn                  \cs_new:cpn
+\cs_new_eq:NN \cs_gnew_protected_nopar:cpn  \cs_new_protected_nopar:cpn
+\cs_new_eq:NN       \cs_gnew_protected:cpn        \cs_new_protected:cpn
+\cs_new_eq:NN           \cs_gnew_nopar:cpx            \cs_new_nopar:cpx
+\cs_new_eq:NN                 \cs_gnew:cpx                  \cs_new:cpx
+\cs_new_eq:NN \cs_gnew_protected_nopar:cpx  \cs_new_protected_nopar:cpx
+\cs_new_eq:NN       \cs_gnew_protected:cpx        \cs_new_protected:cpx
+\cs_new_eq:NN \cs_gnew_eq:NN \cs_new_eq:NN
+\cs_new_eq:NN \cs_gnew_eq:cN \cs_new_eq:cN
+\cs_new_eq:NN \cs_gnew_eq:Nc \cs_new_eq:Nc
+\cs_new_eq:NN \cs_gnew_eq:cc \cs_new_eq:cc
+\cs_new_eq:NN \cs_gundefine:N  \cs_undefine:N
+\cs_new_eq:NN \cs_gundefine:c  \cs_undefine:c
+\cs_new_eq:NN \group_execute_after:N \group_insert_after:N
+%% 
+%%
+%% End of file `l3basics.sty'.
diff --git a/Master/texmf-dist/tex/latex/l3kernel/l3bootstrap.sty b/Master/texmf-dist/tex/latex/l3kernel/l3bootstrap.sty
new file mode 100644
index 00000000000..54b4a0cf7c4
--- /dev/null
+++ b/Master/texmf-dist/tex/latex/l3kernel/l3bootstrap.sty
@@ -0,0 +1,256 @@
+%%
+%% This is file `l3bootstrap.sty',
+%% generated with the docstrip utility.
+%%
+%% The original source files were:
+%%
+%% l3bootstrap.dtx  (with options: `package')
+%% 
+%% EXPERIMENTAL CODE
+%% 
+%% Do not distribute this file without also distributing the
+%% source files specified above.
+%% 
+%% Do not distribute a modified version of this file.
+%% 
+%% File: l3bootstrap.dtx Copyright (C) 2011 The LaTeX3 project
+%%
+%% It may be distributed and/or modified under the conditions of the
+%% LaTeX Project Public License (LPPL), either version 1.3c of this
+%% license or (at your option) any later version.  The latest version
+%% of this license is in the file
+%%
+%%    http://www.latex-project.org/lppl.txt
+%%
+%% This file is part of the "expl3 bundle" (The Work in LPPL)
+%% and all files in that bundle must be distributed together.
+%%
+%% The released version of this bundle is available from CTAN.
+%%
+%% -----------------------------------------------------------------------
+%%
+%% The development version of the bundle can be found at
+%%
+%%    http://www.latex-project.org/svnroot/experimental/trunk/
+%%
+%% for those people who are interested.
+%%
+%%%%%%%%%%%
+%% NOTE: %%
+%%%%%%%%%%%
+%%
+%%   Snapshots taken from the repository represent work in progress and may
+%%   not work or may contain conflicting material!  We therefore ask
+%%   people _not_ to put them into distributions, archives, etc. without
+%%   prior consultation with the LaTeX3 Project.
+%%
+%% -----------------------------------------------------------------------
+\def\GetIdInfo$%
+  Id: #1 #2 #3-#4-#5 #6 #7 $#8%
+  {%
+    \def\ExplFileDate{#3/#4/#5}%
+    \def\ExplFileVersion{#2}%
+  }
+\GetIdInfo$Id: l3bootstrap.dtx 2478 2011-06-19 21:34:23Z joseph $
+  {L3 experimental bootstrap code}
+\ProvidesPackage{l3bootstrap}
+  [%
+    \ExplFileDate\space v\ExplFileVersion\space
+    L3 Experimental bootstrap code%
+  ]
+\def\@tempa%
+  {%
+    \def\@tempa{}%
+    \RequirePackage{luatex}%
+    \RequirePackage{pdftexcmds}%
+    \let\pdfstrcmp\pdf@strcmp
+  }
+\begingroup\expandafter\expandafter\expandafter\endgroup
+\expandafter\ifx\csname directlua\endcsname\relax
+\else
+  \expandafter\@tempa
+\fi
+\protected\edef\ExplSyntaxOff
+  {%
+    \catcode   9 = \the\catcode   9\relax
+    \catcode  32 = \the\catcode  32\relax
+    \catcode  34 = \the\catcode  34\relax
+    \catcode  38 = \the\catcode  38\relax
+    \catcode  58 = \the\catcode  58\relax
+    \catcode  94 = \the\catcode  94\relax
+    \catcode  95 = \the\catcode  95\relax
+    \catcode 124 = \the\catcode 124\relax
+    \catcode 126 = \the\catcode 126\relax
+    \endlinechar = \the\endlinechar\relax
+    \chardef\csname\detokenize{l_expl_status_bool}\endcsname = 0 \relax
+  }
+\protected\edef\ExplSyntaxOn
+  {
+    \catcode 9   = 9  \relax
+    \catcode 32  = 9  \relax
+    \catcode 34  = 12 \relax
+    \catcode 58  = 11 \relax
+    \catcode 94  = 7  \relax
+    \catcode 95  = 11 \relax
+    \catcode 124 = 12 \relax
+    \catcode 126 = 10 \relax
+    \endlinechar = 32 \relax
+    \chardef\csname\detokenize{l_expl_status_bool}\endcsname = 1 \relax
+  }
+\expandafter\chardef\csname\detokenize{l_expl_status_bool}\endcsname = 0 \relax
+\protected\def\GetIdInfo
+  {
+    \begingroup
+    \catcode 32 = 10 \relax
+    \GetIdInfoAuxI
+  }
+\protected\def\GetIdInfoAuxI$#1$#2%
+  {
+    \def\tempa{#1}%
+    \def\tempb{Id}%
+    \ifx\tempa\tempb
+      \def\tempa
+        {%
+          \endgroup
+          \def\ExplFileDate{0000/00/00}%
+          \def\ExplFileDescription{#2}%
+          \def\ExplFileName{[unknown~name]}%
+          \def\ExplFileVersion{000}%
+        }%
+    \else
+      \def\tempa
+        {%
+          \endgroup
+          \GetIdInfoAuxII$#1${#2}%
+        }%
+    \fi
+    \tempa
+  }
+\protected\def\GetIdInfoAuxII$#1 #2.#3 #4 #5 #6 #7 #8$#9%
+  {%
+    \def\ExplFileName{#2}%
+    \def\ExplFileVersion{#4}%
+    \def\ExplFileDescription{#9}%
+    \GetIdInfoAuxIII#5\relax#3\relax#5\relax#6\relax
+  }
+\protected\def\GetIdInfoAuxIII#1#2#3#4#5#6\relax
+  {%
+    \ifx#5/%
+      \expandafter\GetIdInfoAuxCVS
+    \else
+      \expandafter\GetIdInfoAuxSVN
+    \fi
+  }
+\protected\def\GetIdInfoAuxCVS#1,v\relax#2\relax#3\relax
+  {\def\ExplFileDate{#2}}
+\protected\def\GetIdInfoAuxSVN#1\relax#2-#3-#4\relax#5Z\relax
+  {\def\ExplFileDate{#2/#3/#4}}
+\protected\def\ProvidesExplPackage#1#2#3#4%
+  {%
+    \ProvidesPackage{#1}[#2 v#3 #4]%
+    \ExplSyntaxOn
+  }
+\protected\def\ProvidesExplClass#1#2#3#4%
+  {%
+    \ProvidesClass{#1}[#2 v#3 #4]%
+    \ExplSyntaxOn
+  }
+\protected\def\ProvidesExplFile#1#2#3#4%
+  {%
+    \ProvidesFile{#1}[#2 v#3 #4]%
+    \ExplSyntaxOn
+  }
+\edef\@pushfilename
+  {%
+    \edef\expandafter\noexpand
+      \csname\detokenize{l_expl_status_stack_tl}\endcsname
+      {%
+        \noexpand\ifodd\expandafter\noexpand
+          \csname\detokenize{l_expl_status_bool}\endcsname
+          1%
+        \noexpand\else
+          0%
+        \noexpand\fi
+        \expandafter\noexpand
+          \csname\detokenize{l_expl_status_stack_tl}\endcsname
+      }%
+    \ExplSyntaxOff
+    \unexpanded\expandafter{\@pushfilename}%
+  }
+\edef\@popfilename
+  {%
+    \unexpanded\expandafter{\@popfilename}%
+    \noexpand\if a\expandafter\noexpand\csname
+      \detokenize{l_expl_status_stack_tl}\endcsname a%
+      \ExplSyntaxOff
+    \noexpand\else
+      \noexpand\expandafter
+        \expandafter\noexpand\csname
+          \detokenize{expl_status_pop:w}\endcsname
+          \expandafter\noexpand\csname
+            \detokenize{l_expl_status_stack_tl}\endcsname
+          \noexpand\@nil
+   \noexpand\fi
+  }
+\@namedef{\detokenize{l_expl_status_stack_tl}}{0}
+\expandafter\edef\csname\detokenize{expl_status_pop:w}\endcsname#1#2\@nil
+  {%
+    \def\expandafter\noexpand
+      \csname\detokenize{l_expl_status_stack_tl}\endcsname{#2}%
+    \noexpand\ifodd#1\space
+      \noexpand\expandafter\noexpand\ExplSyntaxOn
+    \noexpand\else
+      \noexpand\expandafter\ExplSyntaxOff
+    \noexpand\fi
+  }
+\expandafter\protected\expandafter\def
+  \csname\detokenize{package_check_loaded_expl:}\endcsname
+  {%
+    \@ifpackageloaded{expl3}
+      {}
+      {%
+        \PackageError{expl3}
+          {Cannot load the expl3 modules separately}
+          {%
+            The expl3 modules cannot be loaded separately;\MessageBreak
+            please \string\usepackage\string{expl3\string} instead.
+          }%
+      }%
+  }
+\begingroup\expandafter\expandafter\expandafter\endgroup
+  \expandafter\ifx\csname pdfstrcmp\endcsname\relax
+  \let\pdfstrcmp\strcmp
+\fi
+\begingroup\expandafter\expandafter\expandafter\endgroup
+\expandafter\ifx\csname pdfstrcmp\endcsname\relax
+  \PackageError{l3names}{Required primitive not found: \protect\pdfstrcmp}
+    {%
+      LaTeX3 requires the e-TeX primitives and
+        \string\pdfstrcmp.\MessageBreak
+      These are available in engine versions: \MessageBreak
+      - pdfTeX 1.30 \MessageBreak
+      - XeTeX 0.9994 \MessageBreak
+      - LuaTeX 0.60 \MessageBreak
+      or later. \MessageBreak
+      \MessageBreak
+      Loading of expl3 will abort!
+    }
+  \expandafter\endinput
+\fi
+\protected\edef\ExplSyntaxNamesOn
+  {%
+    \expandafter\noexpand
+      \csname\detokenize{char_set_catcode_letter:n}\endcsname{58}%
+    \expandafter\noexpand
+      \csname\detokenize{char_set_catcode_letter:n}\endcsname{95}%
+  }
+\protected\edef\ExplSyntaxNamesOff
+  {%
+    \expandafter\noexpand
+      \csname\detokenize{char_set_catcode_math_subscript:n}\endcsname{58}%
+    \expandafter\noexpand
+      \csname\detokenize{char_set_catcode_other:n}\endcsname{95}%
+  }
+%% 
+%%
+%% End of file `l3bootstrap.sty'.
diff --git a/Master/texmf-dist/tex/latex/l3kernel/l3box.sty b/Master/texmf-dist/tex/latex/l3kernel/l3box.sty
new file mode 100644
index 00000000000..68db799fdf3
--- /dev/null
+++ b/Master/texmf-dist/tex/latex/l3kernel/l3box.sty
@@ -0,0 +1,219 @@
+%%
+%% This is file `l3box.sty',
+%% generated with the docstrip utility.
+%%
+%% The original source files were:
+%%
+%% l3box.dtx  (with options: `package')
+%% 
+%% EXPERIMENTAL CODE
+%% 
+%% Do not distribute this file without also distributing the
+%% source files specified above.
+%% 
+%% Do not distribute a modified version of this file.
+%% 
+%% File: l3box.dtx Copyright (C) 2005-2011 The LaTeX3 Project
+%%
+%% It may be distributed and/or modified under the conditions of the
+%% LaTeX Project Public License (LPPL), either version 1.3c of this
+%% license or (at your option) any later version.  The latest version
+%% of this license is in the file
+%%
+%%    http://www.latex-project.org/lppl.txt
+%%
+%% This file is part of the "expl3 bundle" (The Work in LPPL)
+%% and all files in that bundle must be distributed together.
+%%
+%% The released version of this bundle is available from CTAN.
+%%
+%% -----------------------------------------------------------------------
+%%
+%% The development version of the bundle can be found at
+%%
+%%    http://www.latex-project.org/svnroot/experimental/trunk/
+%%
+%% for those people who are interested.
+%%
+%%%%%%%%%%%
+%% NOTE: %%
+%%%%%%%%%%%
+%%
+%%   Snapshots taken from the repository represent work in progress and may
+%%   not work or may contain conflicting material!  We therefore ask
+%%   people _not_ to put them into distributions, archives, etc. without
+%%   prior consultation with the LaTeX3 Project.
+%%
+%% -----------------------------------------------------------------------
+\RequirePackage{l3names}
+\GetIdInfo$Id: l3box.dtx 2478 2011-06-19 21:34:23Z joseph $
+  {L3 Experimental boxes}
+\ProvidesExplPackage
+  {\ExplFileName}{\ExplFileDate}{\ExplFileVersion}{\ExplFileDescription}
+\package_check_loaded_expl:
+\cs_new_protected:Npn \box_new:N #1
+  {
+    \chk_if_free_cs:N #1
+    \newbox #1
+  }
+\cs_generate_variant:Nn \box_new:N { c }
+\cs_new_protected_nopar:Npn \box_clear:N #1
+  { \box_set_eq:NN  #1 \c_empty_box }
+\cs_new_protected_nopar:Npn \box_gclear:N #1
+  { \box_gset_eq:NN #1 \c_empty_box }
+\cs_generate_variant:Nn \box_clear:N  { c }
+\cs_generate_variant:Nn \box_gclear:N { c }
+\cs_new_protected_nopar:Npn \box_clear_new:N #1
+  {
+    \cs_if_exist:NTF #1
+      { \box_set_eq:NN #1 \c_empty_box }
+      { \box_new:N #1 }
+  }
+\cs_new_protected_nopar:Npn \box_gclear_new:N #1
+  {
+    \cs_if_exist:NTF #1
+      { \box_gset_eq:NN #1 \c_empty_box }
+      { \box_new:N #1 }
+  }
+\cs_generate_variant:Nn \box_clear_new:N  { c }
+\cs_generate_variant:Nn \box_gclear_new:N { c }
+\cs_new_protected_nopar:Npn \box_set_eq:NN #1#2
+  { \tex_setbox:D #1 \tex_copy:D #2 }
+\cs_new_protected_nopar:Npn \box_gset_eq:NN
+  { \pref_global:D \box_set_eq:NN }
+\cs_generate_variant:Nn \box_set_eq:NN  { cN , Nc , cc }
+\cs_generate_variant:Nn \box_gset_eq:NN { cN , Nc , cc }
+\cs_new_protected_nopar:Npn \box_set_eq_clear:NN #1#2
+  { \tex_setbox:D #1 \tex_box:D #2 }
+\cs_new_protected_nopar:Npn \box_gset_eq_clear:NN
+  { \pref_global:D  \box_set_eq_clear:NN }
+\cs_generate_variant:Nn \box_set_eq_clear:NN  { cN , Nc , cc }
+\cs_generate_variant:Nn \box_gset_eq_clear:NN { cN , Nc , cc }
+\cs_new_eq:NN \box_ht:N \tex_ht:D
+\cs_new_eq:NN \box_dp:N \tex_dp:D
+\cs_new_eq:NN \box_wd:N \tex_wd:D
+\cs_generate_variant:Nn \box_ht:N { c }
+\cs_generate_variant:Nn \box_dp:N { c }
+\cs_generate_variant:Nn \box_wd:N { c }
+\cs_new_protected_nopar:Npn \box_set_dp:Nn #1#2
+  { \box_dp:N #1 \dim_eval:w #2 \dim_eval_end: }
+\cs_new_protected_nopar:Npn \box_set_ht:Nn #1#2
+  { \box_ht:N #1 \dim_eval:w #2 \dim_eval_end: }
+\cs_new_protected_nopar:Npn \box_set_wd:Nn #1#2
+  { \box_wd:N #1 \dim_eval:w #2 \dim_eval_end: }
+\cs_generate_variant:Nn \box_set_ht:Nn { c }
+\cs_generate_variant:Nn \box_set_dp:Nn { c }
+\cs_generate_variant:Nn \box_set_wd:Nn { c }
+\cs_new_eq:NN \box_use_clear:N \tex_box:D
+\cs_new_eq:NN \box_use:N \tex_copy:D
+\cs_generate_variant:Nn \box_use_clear:N { c }
+\cs_generate_variant:Nn \box_use:N { c }
+\cs_new_protected:Npn \box_move_left:nn #1#2
+  { \tex_moveleft:D \dim_eval:w #1 \dim_eval_end: #2 }
+\cs_new_protected:Npn \box_move_right:nn #1#2
+  { \tex_moveright:D \dim_eval:w #1 \dim_eval_end: #2 }
+\cs_new_protected:Npn \box_move_up:nn #1#2
+  { \tex_raise:D \dim_eval:w #1 \dim_eval_end: #2 }
+\cs_new_protected:Npn \box_move_down:nn #1#2
+  { \tex_lower:D \dim_eval:w #1 \dim_eval_end: #2 }
+\cs_new_eq:NN \if_hbox:N      \tex_ifhbox:D
+\cs_new_eq:NN \if_vbox:N      \tex_ifvbox:D
+\cs_new_eq:NN \if_box_empty:N \tex_ifvoid:D
+\prg_new_conditional:Npnn \box_if_horizontal:N #1 { p , T , F , TF }
+  { \if_hbox:N #1 \prg_return_true: \else: \prg_return_false: \fi: }
+\prg_new_conditional:Npnn \box_if_vertical:N #1 { p , T , F , TF }
+  { \if_vbox:N #1 \prg_return_true: \else: \prg_return_false: \fi: }
+\cs_generate_variant:Nn \box_if_horizontal_p:N { c }
+\cs_generate_variant:Nn \box_if_horizontal:NT  { c }
+\cs_generate_variant:Nn \box_if_horizontal:NF  { c }
+\cs_generate_variant:Nn \box_if_horizontal:NTF { c }
+\cs_generate_variant:Nn \box_if_vertical_p:N { c }
+\cs_generate_variant:Nn \box_if_vertical:NT  { c }
+\cs_generate_variant:Nn \box_if_vertical:NF  { c }
+\cs_generate_variant:Nn \box_if_vertical:NTF { c }
+\prg_new_conditional:Npnn \box_if_empty:N #1 { p , T , F , TF }
+  { \if_box_empty:N #1 \prg_return_true: \else: \prg_return_false: \fi: }
+\cs_generate_variant:Nn \box_if_empty_p:N { c }
+\cs_generate_variant:Nn \box_if_empty:NT  { c }
+\cs_generate_variant:Nn \box_if_empty:NF  { c }
+\cs_generate_variant:Nn \box_if_empty:NTF { c }
+\cs_new_eq:NN \l_last_box \tex_lastbox:D
+\cs_new_protected_nopar:Npn \box_set_to_last:N #1
+  { \tex_setbox:D #1 \l_last_box }
+\cs_new_protected_nopar:Npn \box_gset_to_last:N
+  { \pref_global:D \box_set_to_last:N }
+\cs_generate_variant:Nn \box_set_to_last:N  { c }
+\cs_generate_variant:Nn \box_gset_to_last:N { c }
+\cs_new_eq:NN \c_empty_box \voidb@x
+\cs_new_eq:NN \l_tmpa_box \@tempboxa
+\box_new:N \l_tmpb_box
+\cs_new_eq:NN \box_show:N \tex_showbox:D
+\cs_generate_variant:Nn \box_show:N { c }
+\cs_new_protected_nopar:Npn \hbox:n { \tex_hbox:D \scan_stop: }
+\cs_new_protected:Npn \hbox_set:Nn #1#2 { \tex_setbox:D #1 \tex_hbox:D {#2} }
+\cs_new_protected_nopar:Npn \hbox_gset:Nn { \pref_global:D \hbox_set:Nn }
+\cs_generate_variant:Nn \hbox_set:Nn { c }
+\cs_generate_variant:Nn \hbox_gset:Nn { c }
+\cs_new_protected:Npn \hbox_set_to_wd:Nnn #1#2#3
+  { \tex_setbox:D #1 \tex_hbox:D to \dim_eval:w #2 \dim_eval_end: {#3} }
+\cs_new_protected_nopar:Npn \hbox_gset_to_wd:Nnn
+  { \pref_global:D \hbox_set_to_wd:Nnn }
+\cs_generate_variant:Nn \hbox_set_to_wd:Nnn { c }
+\cs_generate_variant:Nn \hbox_gset_to_wd:Nnn {cnn}
+\cs_new_protected_nopar:Npn \hbox_set_inline_begin:N  #1
+  { \tex_setbox:D #1 \tex_hbox:D \c_group_begin_token }
+\cs_new_protected_nopar:Npn \hbox_gset_inline_begin:N
+  { \pref_global:D \hbox_set_inline_begin:N }
+\cs_generate_variant:Nn \hbox_set_inline_begin:N  { c }
+\cs_generate_variant:Nn \hbox_gset_inline_begin:N { c }
+\cs_new_eq:NN \hbox_set_inline_end:  \c_group_end_token
+\cs_new_eq:NN \hbox_gset_inline_end: \c_group_end_token
+\cs_new_protected:Npn \hbox_to_wd:nn #1#2
+   { \tex_hbox:D to \dim_eval:w #1 \dim_eval_end: {#2} }
+\cs_new_protected:Npn \hbox_to_zero:n #1 { \tex_hbox:D to \c_zero_skip {#1} }
+\cs_new_protected:Npn \hbox_overlap_left:n  #1
+  { \hbox_to_zero:n { \tex_hss:D #1 } }
+\cs_new_protected:Npn \hbox_overlap_right:n #1
+  { \hbox_to_zero:n { #1 \tex_hss:D } }
+\cs_new_eq:NN \hbox_unpack:N \tex_unhcopy:D
+\cs_new_eq:NN \hbox_unpack_clear:N \tex_unhbox:D
+\cs_generate_variant:Nn \hbox_unpack:N { c }
+\cs_generate_variant:Nn \hbox_unpack_clear:N { c }
+\cs_new_protected_nopar:Npn \vbox:n { \tex_vbox:D \scan_stop: }
+\cs_new_protected_nopar:Npn \vbox_top:n { \tex_vtop:D \scan_stop: }
+\cs_new_protected:Npn \vbox_to_ht:nn #1#2
+  { \tex_vbox:D to \dim_eval:w #1 \dim_eval_end: {#2} }
+\cs_new_protected:Npn \vbox_to_zero:n #1 { \tex_vbox:D to \c_zero_dim {#1} }
+\cs_new_protected:Npn \vbox_set:Nn #1#2 { \tex_setbox:D #1 \tex_vbox:D {#2} }
+\cs_new_protected_nopar:Npn \vbox_gset:Nn  { \pref_global:D \vbox_set:Nn }
+\cs_generate_variant:Nn \vbox_set:Nn  { c }
+\cs_generate_variant:Nn \vbox_gset:Nn { c }
+\cs_new_protected:Npn \vbox_set_top:Nn #1#2
+  { \tex_setbox:D #1 \tex_vtop:D {#2} }
+\cs_new_protected_nopar:Npn \vbox_gset_top:Nn
+  { \pref_global:D \vbox_set_top:Nn }
+\cs_generate_variant:Nn \vbox_set_top:Nn { c }
+\cs_generate_variant:Nn \vbox_gset_top:Nn { c }
+\cs_new_protected:Npn \vbox_set_to_ht:Nnn #1#2#3
+  { \tex_setbox:D #1 \tex_vbox:D to \dim_eval:w #2 \dim_eval_end: {#3} }
+\cs_new_protected_nopar:Npn \vbox_gset_to_ht:Nnn
+  { \pref_global:D \vbox_set_to_ht:Nnn }
+\cs_generate_variant:Nn \vbox_set_to_ht:Nnn  { c }
+\cs_generate_variant:Nn \vbox_gset_to_ht:Nnn { c }
+\cs_new_nopar:Npn \vbox_set_inline_begin:N #1
+  { \tex_setbox:D #1 \tex_vbox:D \c_group_begin_token }
+\cs_new_protected_nopar:Npn \vbox_gset_inline_begin:N
+  { \pref_global:D \vbox_set_inline_begin:N }
+\cs_generate_variant:Nn \vbox_set_inline_begin:N  { c }
+\cs_generate_variant:Nn \vbox_gset_inline_begin:N { c }
+\cs_new_eq:NN \vbox_set_inline_end:  \c_group_end_token
+\cs_new_eq:NN \vbox_gset_inline_end: \c_group_end_token
+\cs_new_eq:NN \vbox_unpack:N \tex_unvcopy:D
+\cs_new_eq:NN \vbox_unpack_clear:N \tex_unvbox:D
+\cs_generate_variant:Nn \vbox_unpack:N { c }
+\cs_generate_variant:Nn \vbox_unpack_clear:N { c }
+\cs_new_protected_nopar:Npn \vbox_set_split_to_ht:NNn #1#2#3
+  { \tex_setbox:D #1 \tex_vsplit:D #2 to \dim_eval:w #3 \dim_eval_end: }
+%% 
+%%
+%% End of file `l3box.sty'.
diff --git a/Master/texmf-dist/tex/latex/l3kernel/l3clist.sty b/Master/texmf-dist/tex/latex/l3kernel/l3clist.sty
new file mode 100644
index 00000000000..ab92e649563
--- /dev/null
+++ b/Master/texmf-dist/tex/latex/l3kernel/l3clist.sty
@@ -0,0 +1,426 @@
+%%
+%% This is file `l3clist.sty',
+%% generated with the docstrip utility.
+%%
+%% The original source files were:
+%%
+%% l3clist.dtx  (with options: `package')
+%% 
+%% EXPERIMENTAL CODE
+%% 
+%% Do not distribute this file without also distributing the
+%% source files specified above.
+%% 
+%% Do not distribute a modified version of this file.
+%% 
+%% File: l3clist.dtx Copyright (C) 2004-2011 Frank Mittelbach,
+%%                                 The LaTeX3 project
+%%
+%% It may be distributed and/or modified under the conditions of the
+%% LaTeX Project Public License (LPPL), either version 1.3c of this
+%% license or (at your option) any later version.  The latest version
+%% of this license is in the file
+%%
+%%    http://www.latex-project.org/lppl.txt
+%%
+%% This file is part of the "expl3 bundle" (The Work in LPPL)
+%% and all files in that bundle must be distributed together.
+%%
+%% The released version of this bundle is available from CTAN.
+%%
+%% -----------------------------------------------------------------------
+%%
+%% The development version of the bundle can be found at
+%%
+%%    http://www.latex-project.org/svnroot/experimental/trunk/
+%%
+%% for those people who are interested.
+%%
+%%%%%%%%%%%
+%% NOTE: %%
+%%%%%%%%%%%
+%%
+%%   Snapshots taken from the repository represent work in progress and may
+%%   not work or may contain conflicting material!  We therefore ask
+%%   people _not_ to put them into distributions, archives, etc. without
+%%   prior consultation with the LaTeX3 Project.
+%%
+%% -----------------------------------------------------------------------
+\RequirePackage{l3names}
+\GetIdInfo$Id: l3clist.dtx 2498 2011-07-09 17:51:47Z joseph $
+  {L3 Experimental comma separated lists}
+\ProvidesExplPackage
+  {\ExplFileName}{\ExplFileDate}{\ExplFileVersion}{\ExplFileDescription}
+\package_check_loaded_expl:
+\tl_new:N \l_clist_tmpa_tl
+\tl_new:N \l_clist_tmpb_tl
+\cs_new_eq:NN \clist_new:N \tl_new:N
+\cs_new_eq:NN \clist_new:c \tl_new:c
+\cs_new_eq:NN \clist_clear:N  \tl_clear:N
+\cs_new_eq:NN \clist_clear:c  \tl_clear:c
+\cs_new_eq:NN \clist_gclear:N \tl_gclear:N
+\cs_new_eq:NN \clist_gclear:c \tl_gclear:c
+\cs_new_eq:NN \clist_clear_new:N  \tl_clear_new:N
+\cs_new_eq:NN \clist_clear_new:c  \tl_clear_new:c
+\cs_new_eq:NN \clist_gclear_new:N \tl_gclear_new:N
+\cs_new_eq:NN \clist_gclear_new:c \tl_gclear_new:c
+\cs_new_eq:NN \clist_set_eq:NN  \tl_set_eq:NN
+\cs_new_eq:NN \clist_set_eq:Nc  \tl_set_eq:Nc
+\cs_new_eq:NN \clist_set_eq:cN  \tl_set_eq:cN
+\cs_new_eq:NN \clist_set_eq:cc  \tl_set_eq:cc
+\cs_new_eq:NN \clist_gset_eq:NN \tl_gset_eq:NN
+\cs_new_eq:NN \clist_gset_eq:Nc \tl_gset_eq:Nc
+\cs_new_eq:NN \clist_gset_eq:cN \tl_gset_eq:cN
+\cs_new_eq:NN \clist_gset_eq:cc \tl_gset_eq:cc
+\cs_new_protected_nopar:Npn \clist_concat:NNN
+  { \clist_concat_aux:NNNN \tl_set:Nx }
+\cs_new_protected_nopar:Npn \clist_gconcat:NNN
+  { \clist_concat_aux:NNNN \tl_gset:Nx }
+\cs_new_protected_nopar:Npn \clist_concat_aux:NNNN #1#2#3#4
+  {
+    #1 #2
+      {
+        \clist_if_empty:NTF #3
+          { \exp_not:o #4 }
+          {
+            \exp_not:o #3
+            \clist_if_empty:NF #4
+              {
+                ,
+                \exp_not:o #4
+              }
+          }
+      }
+  }
+\cs_generate_variant:Nn \clist_concat:NNN  { ccc }
+\cs_generate_variant:Nn \clist_gconcat:NNN { ccc }
+\cs_new_protected_nopar:Npn \clist_put_left:Nn
+  { \clist_put_aux:NNnnNn \tl_set:Nn \tl_put_left:Nn { } , }
+\cs_new_protected_nopar:Npn \clist_gput_left:Nn
+  { \clist_put_aux:NNnnNn \tl_gset:Nn \tl_gput_left:Nn { } , }
+\cs_new_protected:Npn \clist_put_aux:NNnnNn #1#2#3#4#5#6
+  {
+    \clist_if_empty:NTF #5
+      { #1 #5 {#6} }
+      { \tl_if_empty:nF {#6} { #2 #5 { #3 #6 #4 } } }
+  }
+\cs_generate_variant:Nn \clist_put_left:Nn  {     NV , No , Nx }
+\cs_generate_variant:Nn \clist_put_left:Nn  { c , cV , co , cx }
+\cs_generate_variant:Nn \clist_gput_left:Nn {     NV , No , Nx }
+\cs_generate_variant:Nn \clist_gput_left:Nn { c , cV , co , cx }
+\cs_new_protected:Npn \clist_put_right:Nn
+  {  \clist_put_aux:NNnnNn \tl_set:Nn \tl_put_right:Nn , { } }
+\cs_new_protected_nopar:Npn \clist_gput_right:Nn
+  { \clist_put_aux:NNnnNn \tl_gset:Nn \tl_gput_right:Nn , { } }
+\cs_generate_variant:Nn \clist_put_right:Nn  {     NV , No , Nx }
+\cs_generate_variant:Nn \clist_put_right:Nn  { c , cV , co , cx }
+\cs_generate_variant:Nn \clist_gput_right:Nn {     NV , No , Nx }
+\cs_generate_variant:Nn \clist_gput_right:Nn { c , cV , co , cx }
+\cs_new_protected_nopar:Npn \clist_get:NN #1#2
+  { \exp_after:wN \clist_get_aux:wN #1 , \q_stop #2 }
+\cs_new_protected:Npn \clist_get_aux:wN #1 , #2 \q_stop #3
+  { \tl_set:Nn #3 {#1} }
+\cs_generate_variant:Nn \clist_get:NN { c }
+\cs_new_protected_nopar:Npn \clist_pop:NN
+  { \clist_pop_aux:NNN \tl_set:Nn }
+\cs_new_protected_nopar:Npn \clist_gpop:NN
+  { \clist_pop_aux:NNN \tl_gset:Nn }
+\cs_new_protected_nopar:Npn \clist_pop_aux:NNN #1#2#3
+  { \exp_after:wN \clist_pop_aux:wNNN #2 , \q_nil , \q_nil , \q_stop #1#2#3 }
+\cs_new_protected:Npn \clist_pop_aux:wNNN #1 , #2 , \q_nil , #3 \q_stop #4#5#6
+  {
+    \quark_if_nil:nTF {#2}
+      { #4 #5 { } }
+      { #4 #5 {#2} }
+    \tl_set:Nn #6 {#1}
+  }
+\cs_generate_variant:Nn \clist_pop:NN  { c }
+\cs_generate_variant:Nn \clist_gpop:NN { c }
+\cs_new_eq:NN \clist_push:Nn  \clist_put_left:Nn
+\cs_new_eq:NN \clist_push:NV  \clist_put_left:NV
+\cs_new_eq:NN \clist_push:No  \clist_put_left:No
+\cs_new_eq:NN \clist_push:Nx  \clist_put_left:Nx
+\cs_new_eq:NN \clist_push:cn  \clist_put_left:cn
+\cs_new_eq:NN \clist_push:cV  \clist_put_left:cV
+\cs_new_eq:NN \clist_push:co  \clist_put_left:co
+\cs_new_eq:NN \clist_push:cx  \clist_put_left:cx
+\cs_new_eq:NN \clist_gpush:Nn \clist_gput_left:Nn
+\cs_new_eq:NN \clist_gpush:NV \clist_gput_left:NV
+\cs_new_eq:NN \clist_gpush:No \clist_gput_left:No
+\cs_new_eq:NN \clist_gpush:Nx \clist_gput_left:Nx
+\cs_new_eq:NN \clist_gpush:cn \clist_gput_left:cn
+\cs_new_eq:NN \clist_gpush:cV \clist_gput_left:cV
+\cs_new_eq:NN \clist_gpush:co \clist_gput_left:co
+\cs_new_eq:NN \clist_gpush:cx \clist_gput_left:cx
+\cs_new_eq:NN \clist_use:N \tl_use:N
+\cs_new_eq:NN \clist_use:c \tl_use:c
+\clist_new:N \l_clist_remove_clist
+\cs_new_protected:Npn \clist_remove_duplicates:N
+  { \clist_remove_duplicates_aux:NN \clist_set_eq:NN }
+\cs_new_protected:Npn \clist_gremove_duplicates:N
+  { \clist_remove_duplicates_aux:NN \clist_gset_eq:NN }
+\cs_new_protected:Npn \clist_remove_duplicates_aux:NN #1#2
+  {
+    \clist_clear:N \l_clist_remove_clist
+    \clist_map_inline:Nn #2
+      {
+        \clist_if_in:NnF \l_clist_remove_clist {##1}
+          { \clist_put_right:Nn \l_clist_remove_clist {##1} }
+      }
+    #1 #2 \l_clist_remove_clist
+  }
+\cs_generate_variant:Nn \clist_remove_duplicates:N  { c }
+\cs_generate_variant:Nn \clist_gremove_duplicates:N { c }
+\cs_new_protected:Npn \clist_remove_all:Nn
+  { \clist_remove_all_aux:NNn \clist_set_eq:NN }
+\cs_new_protected:Npn \clist_gremove_all:Nn
+  { \clist_remove_all_aux:NNn \clist_gset_eq:NN }
+\cs_new_protected:Npn \clist_remove_all_aux:NNn #1#2#3
+  {
+    \clist_if_empty:NF #2
+      {
+        \clist_clear:N \l_clist_remove_clist
+        \cs_set_protected:Npn \clist_remove_all_aux:w
+          ##1 , #3 , ##2 \q_stop
+          {
+            \tl_put_right:Nn \l_clist_remove_clist {##1}
+            \quark_if_no_value:nF {##2}
+              { \clist_remove_all_aux:w , ##2 \q_stop }
+          }
+        \exp_after:wN \clist_remove_all_aux:w
+          \exp_after:wN , #2 , #3 , \q_no_value \q_stop
+        \tl_if_empty:NF \l_clist_remove_clist
+          {
+            \exp_after:wN \tl_set:No \exp_after:wN
+              \l_clist_remove_clist \exp_after:wN
+                { \exp_after:wN \use_none:n \l_clist_remove_clist }
+          }
+        #1 #2 \l_clist_remove_clist
+      }
+  }
+\cs_new_protected:Npn \clist_remove_all_aux:w { }
+\cs_generate_variant:Nn \clist_remove_all:Nn  { c }
+\cs_generate_variant:Nn \clist_gremove_all:Nn { c }
+\cs_new:Npn \clist_trim_spaces:n #1
+  {
+    \exp_args:Nf \clist_trim_spaces_aux_i:n
+      { \clist_map_function:nN {#1} \clist_trim_spaces_aux_ii:n }
+  }
+\cs_new:Npn \clist_trim_spaces_aux_i:n #1 { \use_ii:nn #1 }
+\cs_new:Npn \clist_trim_spaces_aux_ii:n #1
+  { , \tl_trim_spaces:n {#1} }
+\cs_new_protected:Npn \clist_trim_spaces:N #1
+  { \tl_set:Nf #1 { \exp_args:No \clist_trim_spaces:n #1 } }
+\cs_new_protected:Npn \clist_gtrim_spaces:N #1
+  { \tl_gset:Nf #1 { \exp_args:No \clist_trim_spaces:n #1 } }
+\cs_generate_variant:Nn \clist_trim_spaces:N  { c }
+\cs_generate_variant:Nn \clist_gtrim_spaces:N { c }
+\cs_new_protected:Npn \clist_tmp:w { }
+\prg_new_eq_conditional:NNn \clist_if_empty:N \tl_if_empty:N { p , T , F , TF }
+\prg_new_eq_conditional:NNn \clist_if_empty:c \tl_if_empty:c { p , T , F , TF }
+\prg_new_eq_conditional:NNn \clist_if_eq:NN \tl_if_eq:NN { p , T , F , TF }
+\prg_new_eq_conditional:NNn \clist_if_eq:Nc \tl_if_eq:Nc { p , T , F , TF }
+\prg_new_eq_conditional:NNn \clist_if_eq:cN \tl_if_eq:cN { p , T , F , TF }
+\prg_new_eq_conditional:NNn \clist_if_eq:cc \tl_if_eq:cc { p , T , F , TF }
+\prg_new_protected_conditional:Npnn \clist_if_in:Nn #1#2
+  { T , F , TF }
+  {
+    \cs_set_protected:Npn \clist_tmp:w ##1 , #2 , ##2##3 \q_stop
+      {
+        \if_meaning:w \q_no_value ##2
+          \prg_return_false:
+        \else:
+           \prg_return_true:
+        \fi:
+      }
+    \exp_last_unbraced:NNo \clist_tmp:w , #1 , #2 , \q_no_value \q_stop
+  }
+\prg_new_protected_conditional:Npnn \clist_if_in:nn #1#2
+  { T , F , TF }
+  {
+    \cs_set_protected:Npn \clist_tmp:w ##1 , #2 , ##2##3 \q_stop
+      {
+        \if_meaning:w \q_no_value ##2
+          \prg_return_false:
+        \else:
+           \prg_return_true:
+        \fi:
+      }
+    \clist_tmp:w , #1 , #2 , \q_no_value \q_stop
+  }
+\cs_generate_variant:Nn \clist_if_in:NnT  {     NV , No }
+\cs_generate_variant:Nn \clist_if_in:NnT  { c , cV , co }
+\cs_generate_variant:Nn \clist_if_in:NnF  {     NV , No }
+\cs_generate_variant:Nn \clist_if_in:NnF  { c , cV , co }
+\cs_generate_variant:Nn \clist_if_in:NnTF {     NV , No }
+\cs_generate_variant:Nn \clist_if_in:NnTF { c , cV , co }
+\cs_generate_variant:Nn \clist_if_in:nnT  {     nV , no }
+\cs_generate_variant:Nn \clist_if_in:nnF  {     nV , no }
+\cs_generate_variant:Nn \clist_if_in:nnTF {     nV , no }
+\cs_new_nopar:Npn \clist_map_function:NN #1#2
+  {
+    \clist_if_empty:NF #1
+      {
+        \exp_last_unbraced:NNo \clist_map_function_aux:Nw #2 #1
+          , \q_recursion_tail , \q_recursion_stop
+      }
+  }
+\cs_new:Npn \clist_map_function:nN #1#2
+  {
+    \tl_if_empty:nF {#1}
+      {
+        \clist_map_function_aux:Nw #2 #1
+          , \q_recursion_tail , \q_recursion_stop
+      }
+  }
+\cs_new:Npn \clist_map_function_aux:Nw #1#2 ,
+  {
+    \quark_if_recursion_tail_stop:n {#2}
+    #1 {#2}
+    \clist_map_function_aux:Nw #1
+  }
+\cs_generate_variant:Nn \clist_map_function:NN { c }
+\int_new:N \g_clist_map_inline_int
+\cs_new_protected:Npn \clist_map_inline:Nn #1#2
+  {
+    \int_gincr:N \g_clist_map_inline_int
+    \cs_gset:cpn { clist_map_inline_ \int_use:N \g_clist_map_inline_int :n }
+      ##1
+      {#2}
+    \exp_args:NNc \clist_map_function:NN #1
+      { clist_map_inline_ \int_use:N \g_clist_map_inline_int :n }
+    \int_gdecr:N \g_clist_map_inline_int
+  }
+\cs_new_protected:Npn \clist_map_inline:nn #1#2
+  {
+    \int_gincr:N \g_clist_map_inline_int
+    \cs_gset:cpn { clist_map_inline_ \int_use:N \g_clist_map_inline_int :n }
+      ##1
+      {#2}
+    \exp_args:Nnc \clist_map_function:nN {#1}
+      { clist_map_inline_ \int_use:N \g_clist_map_inline_int :n }
+    \int_gdecr:N \g_clist_map_inline_int
+  }
+\cs_generate_variant:Nn \clist_map_inline:Nn { c }
+\cs_new_protected:Npn \clist_map_variable:NNn #1#2#3
+  {
+    \clist_map_inline:Nn #1
+      {
+        \tl_set:Nn #2 {##1}
+        #3
+      }
+  }
+\cs_new_protected:Npn \clist_map_variable:nNn #1#2#3
+  {
+    \clist_map_inline:nn {#1}
+      {
+        \tl_set:Nn #2 {##1}
+        #3
+      }
+  }
+\cs_generate_variant:Nn \clist_map_variable:NNn { c }
+\cs_new_eq:NN \clist_map_break:  \use_none_delimit_by_q_recursion_stop:w
+\cs_new_eq:NN \clist_map_break:n \use_i_delimit_by_q_recursion_stop:nw
+\cs_new_protected_nopar:Npn \clist_show:N #1
+  {
+    \clist_if_empty:NTF #1
+      {
+        \iow_term:x { Comma~list~\token_to_str:N #1 \c_space_tl is~empty }
+        \tl_show:n { }
+      }
+      {
+        \iow_term:x
+          {
+            Comma~list~\token_to_str:N #1 \c_space_tl
+            contains~the~items~(without~outer~braces):
+          }
+        \tl_set:Nx \l_clist_show_tl
+          { \clist_map_function:NN #1 \clist_show_aux:n }
+          \etex_showtokens:D \exp_after:wN \exp_after:wN \exp_after:wN
+            { \exp_after:wN \clist_show_aux:w \l_clist_show_tl }
+      }
+  }
+\cs_new:Npn \clist_show_aux:n #1
+  {
+    \iow_newline: > \c_space_tl \c_space_tl
+    \iow_char:N \{ \exp_not:n {#1} \iow_char:N \}
+  }
+\cs_new:Npn \clist_show_aux:w #1 > ~ { }
+\cs_generate_variant:Nn \clist_show:N { c }
+\cs_new:Npn \clist_length:N #1
+  {
+    \int_eval:n
+      {
+        0
+        \clist_map_function:NN #1 \clist_length_aux:n
+      }
+  }
+\cs_new:Npn \clist_length:n #1
+  {
+    \int_eval:n
+      {
+        0
+        \clist_map_function:nN {#1} \clist_length_aux:n
+      }
+  }
+\cs_new:Npn \clist_length_aux:n #1 { +1 }
+\cs_generate_variant:Nn \clist_length:N { c }
+\cs_set_nopar:Npn \clist_item:Nn #1#2
+  { \exp_args:No \clist_item:nn #1 {#2} }
+\cs_set:Npn \clist_item:nn #1#2
+  {
+    \int_compare:nNnTF {#2} < \c_zero
+      {
+        \exp_args:Nf \clist_item_aux:nw
+          { \int_eval:n { \clist_length:n {#1} + #2 } }
+          #1 , \q_recursion_tail \q_recursion_stop
+      }
+      { \clist_item_aux:nw {#2} #1 , \q_recursion_tail \q_recursion_stop }
+  }
+\cs_set:Npn \clist_item_aux:nw #1#2 , #3
+  {
+    \int_compare:nNnTF {#1} = \c_zero
+      { \use_i_delimit_by_q_recursion_stop:nw {#2} }
+      {
+        \quark_if_recursion_tail_stop:n {#3}
+        \exp_args:Nf \clist_item_aux:nw
+          { \int_eval:n { #1 - 1 } }
+          #3
+      }
+  }
+\cs_generate_variant:Nn \clist_item:Nn { c }
+\cs_new_protected:Npn \clist_set_from_seq:NN #1#2
+  {
+    \seq_if_empty:NTF #2
+      { \clist_clear:N #1 }
+      {
+        \seq_push_item_def:n { , \exp_not:n {{##1}} }
+        \tl_set:Nx #1
+          { \exp_after:wN \use_none:n \tex_romannumeral:D -`\0 #2 }
+        \seq_pop_item_def:
+      }
+  }
+\cs_new_protected:Npn \clist_gset_from_seq:NN #1#2
+  {
+    \seq_if_empty:NTF #2
+      { \clist_gclear:N #1 }
+      {
+        \seq_push_item_def:n { , \exp_not:n {##1} }
+        \tl_gset:Nx #1
+          { \exp_after:wN \use_none:n \tex_romannumeral:D -`\0 #2 }
+        \seq_pop_item_def:
+      }
+  }
+\cs_generate_variant:Nn \clist_set_from_seq:NN  {     Nc }
+\cs_generate_variant:Nn \clist_set_from_seq:NN  { c , cc }
+\cs_generate_variant:Nn \clist_gset_from_seq:NN {     Nc }
+\cs_generate_variant:Nn \clist_gset_from_seq:NN { c , cc }
+\cs_new_eq:NN \clist_top:NN \clist_get:NN
+\cs_new_eq:NN \clist_top:cN \clist_get:cN
+\cs_new_eq:NN \clist_remove_element:Nn  \clist_remove_all:Nn
+\cs_new_eq:NN \clist_gremove_element:Nn \clist_gremove_all:Nn
+\cs_new_eq:NN \clist_display:N \clist_show:N
+\cs_new_eq:NN \clist_display:c \clist_show:c
+%% 
+%%
+%% End of file `l3clist.sty'.
diff --git a/Master/texmf-dist/tex/latex/l3kernel/l3doc.cls b/Master/texmf-dist/tex/latex/l3kernel/l3doc.cls
new file mode 100644
index 00000000000..258a8475eb9
--- /dev/null
+++ b/Master/texmf-dist/tex/latex/l3kernel/l3doc.cls
@@ -0,0 +1,1001 @@
+%%
+%% This is file `l3doc.cls',
+%% generated with the docstrip utility.
+%%
+%% The original source files were:
+%%
+%% l3doc.dtx  (with options: `class')
+%% 
+%% 
+%% EXPERIMENTAL CODE
+%% 
+%% Do not distribute this file without also distributing the
+%% source files specified above.
+%% 
+%% Do not distribute a modified version of this file.
+%% 
+%% 
+%% File: l3doc.dtx Copyright (C) 1990-2011 The LaTeX3 project
+%%
+%% It may be distributed and/or modified under the conditions of the
+%% LaTeX Project Public License (LPPL), either version 1.3c of this
+%% license or (at your option) any later version.  The latest version
+%% of this license is in the file
+%%
+%%    http://www.latex-project.org/lppl.txt
+%%
+%% This file is part of the "expl3 bundle" (The Work in LPPL)
+%% and all files in that bundle must be distributed together.
+%%
+%% The released version of this bundle is available from CTAN.
+%%
+%% -----------------------------------------------------------------------
+%%
+%% The development version of the bundle can be found at
+%%
+%%    http://www.latex-project.org/svnroot/experimental/trunk/
+%%
+%% for those people who are interested.
+%%
+%%%%%%%%%%%
+%% NOTE: %%
+%%%%%%%%%%%
+%%
+%%   Snapshots taken from the repository represent work in progress and may
+%%   not work or may contain conflicting material!  We therefore ask
+%%   people _not_ to put them into distributions, archives, etc. without
+%%   prior consultation with the LaTeX3 Project.
+%%
+%% -----------------------------------------------------------------------
+\RequirePackage{l3names}
+\let        \filenameOld        \ExplFileName
+\let        \filedateOld        \ExplFileDate
+\let     \fileversionOld        \ExplFileVersion
+\let \filedescriptionOld        \ExplFileDescription
+\GetIdInfo$Id: l3doc.dtx 2478 2011-06-19 21:34:23Z joseph $
+          {L3 Experimental documentation class}
+\ProvidesExplClass
+  {\ExplFileName}{\ExplFileDate}{\ExplFileVersion}{\ExplFileDescription}
+\RequirePackage{expl3,xparse}
+\let        \ExplFileName        \filenameOld
+\let        \ExplFileDate        \filedateOld
+\let        \ExplFileVersion     \fileversionOld
+\let        \ExplFileDescription \filedescriptionOld
+\cs_new_nopar:Npn\Team{%
+  The~\LaTeX3~Project\thanks{%
+  Frank~Mittelbach,~Denys~Duchier,~Chris~Rowley,~
+  Rainer~Sch\"opf,~Johannes~Braams,~Michael~Downes,~
+  David~Carlisle,~Alan~Jeffrey,~Morten~H\o{}gholm,~Thomas~Lotze,~
+  Javier~Bezos,~Will~Robertson,~Joseph~Wright}}
+\DeclareOption{a5paper}{\@latexerr{Option not supported}{}}
+\bool_new:N \g_doc_full_bool
+\bool_new:N \g_doc_lmodern_bool
+\bool_new:N \g_doc_checkfunc_bool
+\bool_new:N \g_doc_checktest_bool
+\DeclareOption{full}{ \bool_set_true:N \g_doc_full_bool }
+\DeclareOption{onlydoc}{ \bool_set_false:N \g_doc_full_bool }
+\DeclareOption{check}{ \bool_set_true:N \g_doc_checkfunc_bool }
+\DeclareOption{nocheck}{ \bool_set_false:N \g_doc_checkfunc_bool }
+\DeclareOption{checktest}{ \bool_set_true:N \g_doc_checktest_bool }
+\DeclareOption{nochecktest}{ \bool_set_false:N \g_doc_checktest_bool }
+\DeclareOption{cm-default}{ \bool_set_false:N \g_doc_lmodern_bool }
+\DeclareOption{lm-default}{ \bool_set_true:N \g_doc_lmodern_bool }
+\DeclareOption*{\PassOptionsToClass{\CurrentOption}{article}}
+\ExecuteOptions{full,a4paper,nocheck,nochecktest,lm-default}
+\InputIfFileExists{l3doc.cfg}
+  {
+    \typeout{*************************************^^J
+             *~Local~config~file~l3doc.cfg~used   ^^J
+             *************************************}
+  }
+  { \@input{l3doc.ltx} }
+\ProcessOptions
+\LoadClass{article}
+\RequirePackage{doc}
+\RequirePackage{array,alphalph,booktabs,color,fixltx2e,enumitem,textcomp,trace,underscore,csquotes}
+\bool_if:NT \g_doc_lmodern_bool {
+  \RequirePackage[T1]{fontenc}
+  \RequirePackage{lmodern}
+}
+\RequirePackage{hypdoc}
+\let\doc@verbatim\verbatim
+\let\enddoc@verbatim\endverbatim
+\let\doc@@verbatim\@verbatim
+\expandafter\let\csname doc@verbatim*\expandafter\endcsname
+  \csname verbatim*\endcsname
+\expandafter\let\csname enddoc@verbatim*\expandafter\endcsname
+  \csname endverbatim*\endcsname
+\expandafter\let\csname doc@@verbatim*\expandafter\endcsname
+  \csname @verbatim*\endcsname
+\RequirePackage{verbatim}
+\AtBeginDocument{%
+  \let\verbatim\doc@verbatim
+  \let\endverbatim\enddoc@verbatim
+  \let\@verbatim\doc@@verbatim
+  \expandafter\let\csname verbatim*\expandafter\endcsname
+    \csname doc@verbatim*\endcsname
+  \expandafter\let\csname endverbatim*\expandafter\endcsname
+    \csname enddoc@verbatim*\endcsname
+  \expandafter\let\csname @verbatim*\expandafter\endcsname
+    \csname doc@@verbatim*\endcsname
+}
+\cs_set_nopar:Npn \MakePrivateLetters {
+  \char_set_catcode_letter:N \@
+  \char_set_catcode_letter:N \_
+  \char_set_catcode_letter:N \:
+}
+\setcounter{StandardModuleDepth}{1}
+\@addtoreset{CodelineNo}{part}
+\cs_set_nopar:Npn \theCodelineNo {
+  \textcolor[gray]{0.5}{ \sffamily\tiny\arabic{CodelineNo} }
+}
+\setlength   \textwidth      { 385pt }
+\addtolength \marginparwidth {  30pt }
+\addtolength \oddsidemargin  {  20pt }
+\addtolength \evensidemargin {  20pt }
+\cs_set_eq:NN \@@oldlist\list
+\cs_set_nopar:Npn \list#1#2{\@@oldlist{#1}{#2\listparindent\z@}}
+\setlength \parindent  { 0pt }
+\setlength \itemindent { 0pt }
+\setlength \parskip    { \medskipamount }
+\AtBeginDocument {
+  \MakeShortVerb \"
+  \MakeShortVerb \|
+}
+\providecommand*\eTeX{
+  \if b\expandafter\@car\f@series\@nil\boldmath\fi
+  $\m@th\varepsilon$-\TeX
+}
+\providecommand*\IniTeX{Ini\TeX}
+\providecommand*\Lua{Lua}
+\providecommand*\LuaTeX{\Lua\TeX}
+\providecommand*\pdfTeX{pdf\TeX}
+\RequirePackage{graphicx}
+\cs_if_free:NT \XeTeX {
+  \cs_new_protected_nopar:Npn \XeTeX
+    {X\kern-.125em\lower.5ex\hbox{\reflectbox{E}}\kern-.1667em\TeX}
+}
+\cs_set_nopar:Npn \cmd #1 { \cs{\expandafter\cmd@to@cs\string#1} }
+\cs_set_nopar:Npn \cmd@to@cs #1#2 { \char\number`#2\relax }
+\DeclareRobustCommand \cs [1] { \texttt { \char`\\ #1 } }
+\providecommand\marg[1]{ \texttt{\char`\{} \meta{#1} \texttt{\char`\}} }
+\providecommand\oarg[1]{ \texttt[ \meta{#1} \texttt] }
+\providecommand\parg[1]{ \texttt( \meta{#1} \texttt) }
+\cs_set_eq:NN \m    \meta
+\cs_set_eq:NN \file \nolinkurl
+\DeclareRobustCommand \env {\texttt}
+\DeclareRobustCommand \pkg {\textsf}
+\DeclareRobustCommand \cls {\textsf}
+\newenvironment{texnote}{
+  \endgraf
+  \vspace{3mm}
+  \small\textbf{\TeX~hackers~note:}
+}{
+  \vspace{3mm}
+}
+\newcommand\tn[1]{
+  \texttt{\bslash #1}
+  \index{TeX~and~LaTeX2e~commands\actualchar
+          \string\TeX{}~and~\string\LaTeXe{}~commands:\levelchar
+          #1\actualchar{\string\ttfamily\string\bslash{}#1}}}
+\cs_new:Npn \doc_implementation: {
+  \cs_set:Npn \variable {\macro[var]}
+}
+\cs_new:Npn \doc_docu: {
+  \cs_set_eq:NN \variable \variabledoc
+}
+\AtEndOfPackage{\doc_docu:}
+\newenvironment{documentation}{\doc_docu:}{}
+\newenvironment{implementation}{\doc_implementation:}{}
+\newcommand\EnableDocumentation{%
+  \renewenvironment{documentation}{\doc_docu:}{}%
+}
+\newcommand\EnableImplementation{%
+  \renewenvironment{implementation}{\doc_implementation:}{}%
+}
+\newcommand\DisableDocumentation{%
+  \cs_set_eq:NN \documentation \comment
+  \cs_set_eq:NN \enddocumentation \endcomment
+}
+\newcommand\DisableImplementation{%
+  \cs_set_eq:NN \implementation \comment
+  \cs_set_eq:NN \endimplementation \endcomment
+}
+\newenvironment{arguments}{
+  \enumerate[
+    nolistsep,
+    label=\texttt{\#\arabic*}~:,
+    labelsep=*,
+  ]
+}{
+  \endenumerate
+}
+\char_set_catcode_active:N \<
+\DeclareDocumentCommand \function { O{} } {
+
+  \phantomsection
+
+  \bool_gset_false:N \l_doc_meta_TF_bool
+  \bool_gset_false:N \l_doc_meta_pTF_bool
+  \bool_gset_false:N \l_doc_meta_EXP_bool
+
+  \cs_set_nopar:Npn \KV_key_no_value_elt:n ##1 { \use:c {doc_meta_opt_##1:} }
+  \keyval_parse:NNn \KV_key_no_value_elt:n \use_none:nn {#1}
+
+  \char_set_catcode_active:N \<
+  \cs_set_eq:NN < \doc_open_meta:n
+
+  \group_begin:
+    \MakePrivateLetters
+    \char_set_catcode_other:N \|
+    \char_set_catcode_other:N \\
+    \char_set_catcode_space:N \~
+    \char_set_catcode_ignore:N \ % space
+    \char_set_catcode_ignore:N \^^M
+    \char_set_catcode_ignore:N \^^I
+    \function_aux:n
+}
+\char_set_catcode_other:N \<
+\cs_set_eq:NN \variabledoc \function
+\cs_set:Nn \doc_meta_opt_TF: {
+  \bool_gset_true:N \l_doc_meta_TF_bool
+}
+\cs_set:Nn \doc_meta_opt_EXP: {
+  \bool_gset_true:N \l_doc_meta_EXP_bool
+}
+\cs_set:Nn \doc_meta_opt_pTF: {
+  \bool_gset_true:N \l_doc_meta_pTF_bool
+  \bool_gset_true:N \l_doc_meta_TF_bool
+  \bool_gset_true:N \l_doc_meta_EXP_bool
+}
+\group_begin:
+\char_set_catcode_other:N \|
+\cs_gset_nopar:Npn \function_aux:n #1 {
+    \cs_set_nopar:Npn \nextnewline{\cs_gset_nopar:Npn\nextnewline{\\}}
+    \tl_gset_eq:NN \g_doc_macro_tl \c_empty_tl
+    \bigskip\endgraf\noindent\ttfamily
+    \tabular[b]{ | l @{} c | }
+      \hline
+      \clist_map_inline:nn {#1} {\doc_showmacro:w ##1 | \q_stop} \\
+      \hline
+    \endtabular
+  \group_end:
+}
+\group_end:
+\group_begin:
+\char_set_catcode_other:N \|
+\cs_gset_nopar:Npn \doc_showmacro:w #1 | {
+  \tl_if_blank:nTF {#1} {
+    \use_none:n
+  }{
+    \doc_showmacro_aux:w #1 / \q_stop
+    \peek_meaning:NTF \q_stop { \use_none:n } { \doc_showmacro:w }
+  }
+}
+\group_end:
+\bool_new:N \l_doc_meta_TF_bool
+\bool_new:N \l_doc_meta_pTF_bool
+\bool_new:N \l_doc_meta_EXP_bool
+\cs_new_nopar:Npn \doc_showmacro_aux:w #1 / #2 \q_stop {
+
+  \tl_if_in:nnT {#2} { (TF)  } {
+    \bool_gset_true:N \l_doc_meta_TF_bool
+  }
+
+  \tl_if_in:nnT {#2} { (EXP) } {
+    \bool_gset_true:N \l_doc_meta_EXP_bool
+  }
+
+  \tl_if_in:nnT {#2} { (pTF)  } {
+    \bool_gset_true:N \l_doc_meta_TF_bool
+    \bool_gset_true:N \l_doc_meta_pTF_bool
+    \bool_gset_true:N \l_doc_meta_EXP_bool
+  }
+
+  \bool_if:NT \l_doc_meta_pTF_bool {
+    \tl_set:Nx \l_doc_pTF_name_tl { \doc_predicate_from_base:w #1 \q_nil }
+    \doc_special_main_index:o { \l_doc_pTF_name_tl }
+    \seq_gput_right:Nx \g_doc_functions_seq { \tl_to_str:N \l_doc_pTF_name_tl }
+  }
+
+  \bool_if:NTF \l_doc_meta_TF_bool {
+    \doc_special_main_index:o { #1 TF }
+    \seq_gput_right:Nx \g_doc_functions_seq { \tl_to_str:n { #1 TF } }
+    \seq_gput_right:Nx \g_doc_functions_seq { \tl_to_str:n { #1 T  } }
+    \seq_gput_right:Nx \g_doc_functions_seq { \tl_to_str:n { #1  F } }
+  }{
+    \doc_special_main_index:o { #1 }
+    \seq_gput_right:Nx \g_doc_functions_seq { \tl_to_str:n { #1    } }
+  }
+
+  \bool_if:NTF \l_doc_meta_pTF_bool {
+    \bool_gset_false:N \l_doc_meta_TF_bool
+    \exp_after:wN \doc_showmacro_aux_ii:w \l_doc_pTF_name_tl ::\q_stop
+    \bool_gset_true:N \l_doc_meta_TF_bool
+    \doc_showmacro_aux_ii:w #1::\q_stop
+  }{
+    \doc_showmacro_aux_ii:w #1::\q_stop
+  }
+}
+\cs_set_nopar:Npn \doc_showmacro_aux_ii:w #1:#2:#3 \q_stop {
+  \nextnewline
+  \str_if_eq:xxTF {#1} {\g_doc_macro_tl} {
+    \doc_typeset_aux:n
+  }{
+    \tl_gset:Nn \g_doc_macro_tl {#1}
+    \use:n
+  }
+  { \g_doc_macro_tl }
+  #3
+  #2
+  \bool_if:NT \l_doc_meta_TF_bool { \doc_typeset_TF: }
+  &
+  \bool_if:NT \l_doc_meta_EXP_bool {
+    \hspace{\tabcolsep}
+    $\star$
+  }
+  \tl_set:Nx \g_doc_macro_tl { \tl_to_str:N \g_doc_macro_tl }
+  \exp_args:NNf \tl_replace_all_in:Nnn \g_doc_macro_tl {\token_to_str:N _} {/}
+  \exp_args:NNf \tl_replace_all_in:Nnn \g_doc_macro_tl {\@backslashchar} {}
+  \bool_if:NT \g_doc_full_bool {
+    \exp_args:Nf\label{doc/function/\g_doc_macro_tl#3#2}
+  }
+}
+\dim_new:N \g_doc_syntax_dim
+\dim_set:Nn \g_doc_syntax_dim {0.7\textwidth}
+\newenvironment{syntax}{
+  \minipage[b]{\g_doc_syntax_dim}
+    \cs_set_nopar:Npn \meta@font@select{\rmfamily\itshape} % (Will: I HATE italic cmtt!)
+    \small\ttfamily\raggedright
+    \obeyspaces\obeylines
+}{
+  \endminipage
+  \hfil\break
+  \global\@ignoretrue
+}
+\tl_map_inline:nn {0123456789} { \cs_gset_eq:cN {char_other_#1} #1 }
+\group_begin:
+  \tl_map_inline:nn {0123456789} { \char_set_catcode_active:N #1 }
+  \cs_new:Npn \doc_open_meta:n {
+    \group_begin:
+      \tl_map_function:nN {0123456789} \doc_assign_num:n
+      \doc_close_meta:w
+  }
+  \cs_new:Npn \Arg {
+    \texttt{ \char`\{ }
+    \group_begin:
+      \tl_map_function:nN {0123456789} \doc_assign_num:n
+      \doc_close_Arg:n
+  }
+\group_end:
+\cs_new_nopar:Npn \doc_close_meta:w #1> { \meta {#1} \group_end: }
+\cs_new_nopar:Npn \doc_close_Arg:n #1 {
+  \meta {#1}
+  \group_end:
+  \texttt{ \char`\} }
+}
+\cs_new_nopar:Npn \doc_assign_num:n #1 {
+  \cs_set_eq:cc { \string #1 } { char_other_\string #1 }
+  \char_set_catcode_active:N #1
+  \cs_set_nopar:Npn #1 { \unskip \, $ {} \sb { \use:c { char_other_\string #1 } } $ }
+}
+\renewcommand \macro [1][] {
+
+  \int_compare:nNnTF \currentgrouplevel=2
+    { \int_gzero:N \g_doc_nested_macro_int }
+    { \int_incr:N  \g_doc_nested_macro_int }
+
+  \bool_set_false:N \l_doc_macro_aux_bool
+  \bool_set_false:N \l_doc_macro_internal_bool
+  \bool_set_false:N \l_doc_macro_TF_bool
+  \bool_set_false:N \l_doc_macro_pTF_bool
+  \bool_set_false:N \l_doc_macro_var_bool
+  \bool_set_false:N \l_doc_tested_bool
+
+  \cs_set_eq:NN \doc_macroname_prefix:n \use:n
+  \cs_set_eq:NN \doc_macroname_suffix: \c_empty_tl
+
+  \cs_set_nopar:Npn \KV_key_no_value_elt:n ##1 { \use:c {doc_macro_opt_##1:} }
+  \keyval_parse:NNn \KV_key_no_value_elt:n \use_none:nn {#1}
+
+  \cs_set_eq:NN \testfile \doc_print_testfile:n
+
+  \group_begin:
+    \MakePrivateLetters
+    \char_set_catcode_letter:N \\
+    \char_set_catcode_ignore:N \ % space
+    \char_set_catcode_ignore:N \^^M
+    \char_set_catcode_ignore:N \^^I
+    \doc_macro_aux:n
+}
+\cs_new_nopar:Npn \doc_macro_aux:n #1 {
+  \group_end:
+  \cs_set:Npn \l_doc_macro_input_clist {#1}
+  \bool_if:NTF \l_doc_macro_pTF_bool
+  {
+    \clist_map_inline:nn {#1}
+      {
+        \tl_if_blank:nF {##1}
+          {
+            \exp_args:Nf \doc_macro_single
+              { \doc_predicate_from_base:w ##1 \q_nil }
+          }
+      }
+    \bool_set_true:N \l_doc_macro_TF_bool
+    \clist_map_inline:nn {#1}
+      { \tl_if_blank:nF {##1} { \doc_macro_single {##1} } }
+    \bool_set_false:N \l_doc_macro_TF_bool
+  }
+  {
+    \clist_map_inline:nn {#1}
+      { \tl_if_blank:nF {##1} { \doc_macro_single {##1} } }
+  }
+}
+\bool_new:N \l_doc_macro_internal_bool
+\bool_new:N \l_doc_macro_aux_bool
+\bool_new:N \l_doc_macro_TF_bool
+\bool_new:N \l_doc_macro_pTF_bool
+\bool_new:N \l_doc_macro_var_bool
+\cs_set_nopar:Npn \doc_macro_opt_aux: { \bool_set_true:N \l_doc_macro_aux_bool }
+\cs_set_nopar:Npn \doc_macro_opt_internal: { \bool_set_true:N \l_doc_macro_internal_bool }
+\cs_set_nopar:Npn \doc_macro_opt_TF:  { \bool_set_true:N \l_doc_macro_TF_bool  }
+\cs_set_nopar:Npn \doc_macro_opt_pTF: { \bool_set_true:N \l_doc_macro_pTF_bool }
+\cs_set_nopar:Npn \doc_macro_opt_var: { \bool_set_true:N \l_doc_macro_var_bool }
+\cs_set:Npn \doc_predicate_from_base:w #1:#2 \q_nil {#1_p:#2}
+\int_new:N \l_doc_macro_int
+\cs_set_nopar:Npn \doc_macro_single #1 {
+  \int_incr:N \l_doc_macro_int
+  \tl_set:Nx \saved@macroname { \token_to_str:N #1 }
+  \topsep\MacroTopsep
+  \trivlist
+  \cs_set_nopar:Npn \makelabel ##1 { \llap{##1} }
+  \if@inlabel
+    \cs_set_eq:NN \@tempa \@empty
+    \count@ \macro@cnt
+    \loop \ifnum\count@>\z@
+      \cs_set_nopar:Npx \@tempa{\@tempa\hbox{\strut}}
+      \advance\count@\m@ne
+    \repeat
+    \cs_set_nopar:Npx \makelabel ##1 {
+      \llap{\vtop to\baselineskip {\@tempa\hbox{##1}\vss}}
+    }
+    \advance \macro@cnt \@ne
+  \else
+    \macro@cnt \@ne
+  \fi
+
+  \bool_if:NT \l_doc_macro_aux_bool {
+    \cs_set_eq:NN \doc_macroname_prefix:n \doc_typeset_aux:n
+  }
+  \bool_if:NT \l_doc_macro_TF_bool {
+    \cs_set_eq:NN \doc_macroname_suffix: \doc_typeset_TF:
+  }
+
+  \bool_if:NF \l_doc_macro_aux_bool {
+    \tl_gset:Nx \l_doc_macro_tl { \tl_to_str:n {#1} }
+    \exp_args:NNf \tl_greplace_all_in:Nnn \l_doc_macro_tl {\token_to_str:N _} {/}
+    \exp_args:NNf \tl_greplace_all_in:Nnn \l_doc_macro_tl {\@backslashchar} {}
+  }
+
+  \use:x {
+    \exp_not:N \item [ \exp_not:N \doc_print_macroname:n {
+      \tl_to_str:n {#1}
+    }]
+  }
+  \global\advance \c@CodelineNo \@ne
+
+  \bool_if:NF \l_doc_macro_aux_bool {
+    \bool_if:NTF \l_doc_macro_TF_bool {
+      \seq_gput_right:Nx \g_doc_macros_seq { \tl_to_str:n { #1 TF } }
+      \seq_gput_right:Nx \g_doc_macros_seq { \tl_to_str:n { #1 T  } }
+      \seq_gput_right:Nx \g_doc_macros_seq { \tl_to_str:n { #1 F  } }
+    }{
+      \seq_gput_right:Nx \g_doc_macros_seq { \tl_to_str:n {#1} }
+    }
+  }
+  \bool_if:NTF \l_doc_macro_TF_bool {
+    \SpecialMainIndex{#1 TF}\nobreak
+    \DoNotIndex{#1 TF}
+  }{
+    \SpecialMainIndex{#1}\nobreak
+    \DoNotIndex{#1}
+  }
+
+  \global\advance \c@CodelineNo \m@ne
+  \ignorespaces
+}
+\tl_clear:N \l_doc_macro_tl
+\cs_set_nopar:Npn \doc_print_macroname:n #1 {
+  \strut
+  \int_compare:nTF { \tl_elt_count:n {#1} <= 28 }
+    { \MacroFont } { \MacroLongFont }
+
+  % INEFFICIENT: (!)
+  \exp_args:NNx \seq_if_in:NnTF \g_doc_functions_seq
+  { #1 \bool_if:NT \l_doc_macro_TF_bool { \tl_to_str:n {TF} } }
+  {
+    \hyperref [doc/function/\l_doc_macro_tl]
+  }
+  { \use:n }
+  {
+    \doc_macroname_prefix:n {#1} \doc_macroname_suffix: \ % space!
+  }
+}
+\providecommand \MacroLongFont {
+  \fontfamily{lmtt}\fontseries{lc}\small
+}
+\cs_set_nopar:Npn \doc_typeset_TF: {
+  \itshape TF%
+  \makebox[0pt][r]{%
+    \color[gray]{0.5}
+    \underline { \phantom{\itshape TF} \kern-0.1em }
+  }
+}
+\cs_set_nopar:Npn \doc_typeset_aux:n #1 {
+  {\color[gray]{0.5} #1}
+}
+\DeclareDocumentCommand \doc_print_testfile:n {m} {
+  \bool_set_true:N \l_doc_tested_bool
+  \tl_if_eq:nnF {#1} {*} {
+    \seq_if_in:NnF \g_doc_testfiles_seq {#1}
+    {
+      \par{\footnotesize(\textit{
+        The~ test~ suite~ for~ this~ command,~ and~ others~ in~ this~ file,~ is~ \textsf{#1}}.
+      )\par}
+      \seq_gput_right:Nn \g_doc_testfiles_seq {#1}
+    }
+  }
+}
+\seq_new:N \g_doc_testfiles_seq
+\DeclareDocumentCommand \TestFiles {m} {
+  \par
+  {\itshape
+    The~ following~ test~ files~ are~ used~ for~ this~ code:~ \textsf{#1}.
+  }
+  \par\ignorespaces
+}
+\DeclareDocumentCommand \UnitTested {} {
+  \testfile*
+}
+\cs_generate_variant:Nn \prop_gput:Nnn {NVx}
+\prop_new:N \g_doc_missing_tests_prop
+\DeclareDocumentCommand \TestMissing {m} {
+  \prop_if_in:NVTF \g_doc_missing_tests_prop \l_doc_macro_input_clist
+  {
+    \prop_get:NVN \g_doc_missing_tests_prop \l_doc_macro_input_clist \l_tmpa_tl
+    \prop_gput:NVx \g_doc_missing_tests_prop \l_doc_macro_input_clist
+    {
+      *~ \l_tmpa_tl
+      ^^J \exp_not:n {\space\space\space\space\space\space}
+      *~ #1
+    }
+  }
+  { \prop_gput:NVn \g_doc_missing_tests_prop \l_doc_macro_input_clist {#1} }
+}
+\int_new:N \g_doc_nested_macro_int
+\cs_set:Nn \doc_texttt_comma:n {\,,~\texttt{#1}}
+\cs_set:Npn \endmacro {
+  \int_compare:nT {\g_doc_nested_macro_int<1}
+  {
+  \par\nobreak{\footnotesize(\emph{
+    End~ definition~ for~
+    \prg_case_int:nnn { \clist_length:N \l_doc_macro_input_clist }
+    {
+      {1} { \texttt{ \clist_use:N \l_doc_macro_input_clist }. }
+      {2}
+      {
+        \tl_set:Nx \l_clist_first_tl { \clist_item:Nn \l_doc_macro_input_clist {0} }
+        \tl_set:Nx \l_clist_last_tl { \clist_item:Nn \l_doc_macro_input_clist {1} }
+        \texttt{\l_clist_first_tl}\,~ and~ \texttt{\l_clist_last_tl}\,.
+      }
+      {3}
+      {
+        \tl_set:Nx \l_clist_first_tl { \clist_item:Nn \l_doc_macro_input_clist {0} }
+        \tl_set:Nx \l_clist_mid_tl   { \clist_item:Nn \l_doc_macro_input_clist {1} }
+        \tl_set:Nx \l_clist_last_tl  { \clist_item:Nn \l_doc_macro_input_clist {2} }
+        \texttt{\l_clist_first_tl}\,,~
+        \texttt{\l_clist_mid_tl}\,,~
+        and~ \texttt{\l_clist_last_tl}\,.
+      }
+    }
+    {
+      \tl_set:Nx \l_clist_first_tl { \clist_item:Nn \l_doc_macro_input_clist {0} }
+      \texttt{\l_clist_first_tl}\,~and~others.
+    }
+    \bool_if:nT {
+      !\l_doc_macro_aux_bool &&
+      !\l_doc_macro_internal_bool &&
+      \int_compare_p:n {\g_doc_nested_macro_int<1}
+    }
+    {
+        \int_compare:nNnTF \l_doc_macro_int=1 {~This~} {~These~}
+        \bool_if:NTF \l_doc_macro_var_bool{variable}{function}
+        \int_compare:nNnTF \l_doc_macro_int=1 {~is~}{s~are~}
+        documented~on~page~
+        \exp_args:Nx\pageref{doc/function/\l_doc_macro_tl}.
+    }
+  })\par}
+  }
+  \bool_if:nT
+  { \g_doc_checktest_bool &&
+    !( \l_doc_macro_aux_bool || \l_doc_macro_var_bool ) &&
+    !\l_doc_tested_bool
+  }
+  {
+    \seq_gput_right:Nx \g_doc_not_tested_seq
+    {
+      \l_doc_macro_input_clist
+      \bool_if:NT \l_doc_macro_pTF_bool {~(pTF)}
+      \bool_if:NT \l_doc_macro_TF_bool {~(TF)}
+    }
+  }
+}
+\newcommand*{\DescribeOption}{
+ \leavevmode
+ \@bsphack
+ \begingroup
+   \MakePrivateLetters
+   \Describe@Option
+}
+\newcommand*{\Describe@Option}[1]{
+ \endgroup
+ \marginpar{
+   \raggedleft
+   \PrintDescribeEnv{#1}
+ }
+ \SpecialOptionIndex{#1}
+ \@esphack
+ \ignorespaces
+}
+\newcommand*{\SpecialOptionIndex}[1]{
+ \@bsphack
+ \begingroup
+   \HD@target
+   \let\HDorg@encapchar\encapchar
+   \edef\encapchar usage{
+     \HDorg@encapchar hdclindex{\the\c@HD@hypercount}{usage}
+   }
+   \index{
+     #1\actualchar{\protect\ttfamily#1}~(option)
+     \encapchar usage
+   }
+   \index{
+     options:\levelchar#1\actualchar{\protect\ttfamily#1}
+     \encapchar usage
+   }
+ \endgroup
+ \@esphack
+}
+\font\manual=manfnt
+\cs_set_nopar:Npn \dbend { {\manual\char127} }
+\newenvironment {danger} {
+  \begin{trivlist}\item[]\noindent
+  \begingroup\hangindent=2pc\hangafter=-2
+  \cs_set_nopar:Npn \par{\endgraf\endgroup}
+  \hbox to0pt{\hskip-\hangindent\dbend\hfill}\ignorespaces
+}{
+  \par\end{trivlist}
+}
+\newenvironment {ddanger} {
+  \begin{trivlist}\item[]\noindent
+  \begingroup\hangindent=3.5pc\hangafter=-2
+  \cs_set_nopar:Npn \par{\endgraf\endgroup}
+  \hbox to0pt{\hskip-\hangindent\dbend\kern2pt\dbend\hfill}\ignorespaces
+}{
+  \par\end{trivlist}
+}
+\newenvironment{TemplateInterfaceDescription}[1]
+  {\subsection{The~object~type~`#1'}%
+   \begingroup
+   \@beginparpenalty\@M
+   \description
+   \def\TemplateArgument##1##2{\item[Arg:~##1]##2\par}%
+   \def\TemplateSemantics{\enddescription\endgroup
+       \subsubsection*{Semantics:}}%
+  }
+  {\par\bigskip}
+\newenvironment{TemplateDescription}[2]
+  {\subsection{The~template~`#2'~(object~type~#1)}%
+   \subsubsection*{Attributes:}%
+   \begingroup
+   \@beginparpenalty\@M
+   \description
+   \def\TemplateKey##1##2##3##4{\item[##1~(##2)]##3%
+     \ifx\TemplateKey##4\TemplateKey\else
+         \hfill\penalty500\hbox{}\hfill Default:~##4%
+         \nobreak\hskip-\parfillskip\hskip0pt\relax
+     \fi
+     \par}%
+   \def\TemplateSemantics{\enddescription\endgroup
+       \subsubsection*{Semantics~\&~Comments:}}%
+  }
+  {\par\bigskip}
+\newenvironment{InstanceDescription}[4][xxxxxxxxxxxxxxx]
+  {\subsubsection{The~instance~`#3'~(template~#2/#4)}%
+   \subsubsection*{Attribute~values:}%
+   \begingroup
+   \@beginparpenalty\@M
+   \def\InstanceKey##1##2{\>\textbf{##1}\>##2\\}%
+   \def\InstanceSemantics{\endtabbing\endgroup
+       \vskip-30pt\vskip0pt
+       \subsubsection*{Layout~description~\&~Comments:}}%
+   \tabbing
+   xxxx\=#1\=\kill
+  }
+  {\par\bigskip}
+\bool_new:N \g_doc_implementation_bool
+\cs_set_nopar:Npn \AlsoImplementation {
+  \bool_set_true:N \g_doc_implementation_bool
+  \cs_set:Npn \StopEventually ##1 {
+    \@bsphack
+    \cs_gset_nopar:Npn \Finale { ##1 \check@checksum }
+    \init@checksum
+    \@esphack
+  }
+}
+\AlsoImplementation
+\cs_set_nopar:Npn \OnlyDescription {
+  \@bsphack
+  \bool_set_false:N \g_doc_implementation_bool
+  \cs_set:Npn \StopEventually ##1 { ##1 \endinput }
+  \@esphack
+}
+\cs_set_eq:NN \Finale \relax
+\cs_set_nopar:Npn \partname{File}
+\clist_new:N \g_docinput_clist
+\cs_set:Npn \DocInput #1 {
+  \clist_map_inline:nn {#1} {
+    \clist_put_right:Nn \g_docinput_clist {##1}
+    \MakePercentIgnore
+    \input{##1}
+    \MakePercentComment
+  }
+}
+\cs_set:Npn \DocInputAgain {
+  \clist_map_inline:Nn \g_docinput_clist {
+    \MakePercentIgnore
+    \input{##1}
+    \MakePercentComment
+  }
+}
+\cs_set_nopar:Npn \partname{File}
+\newcommand*{\DocInclude}[1]{%
+  \relax\clearpage
+  \docincludeaux
+  \IfFileExists{#1.fdd}{
+    \cs_set_nopar:Npn \currentfile{#1.fdd}
+  }{
+    \cs_set_nopar:Npn \currentfile{#1.dtx}
+  }
+  \ifnum\@auxout=\@partaux
+    \@latexerr{\string\include\space cannot~be~nested}\@eha
+  \else
+    \@docinclude #1
+  \fi
+}
+\cs_set_nopar:Npn \@docinclude #1 {
+  \clearpage
+  \immediate\write\@mainaux{\string\@input{#1.aux}}
+  \@tempswatrue
+  \if@partsw
+    \@tempswafalse
+    \cs_set_nopar:Npx \@tempb{#1}
+    \@for\@tempa:=\@partlist\do{
+      \ifx\@tempa\@tempb\@tempswatrue\fi
+    }
+  \fi
+  \if@tempswa
+    \cs_set_eq:NN \@auxout\@partaux
+    \immediate\openout\@partaux #1.aux
+    \immediate\write\@partaux{\relax}
+    \cs_set_eq:NN \@ltxdoc@PrintIndex\PrintIndex
+    \cs_set_eq:NN \PrintIndex\relax
+    \cs_set_eq:NN \@ltxdoc@PrintChanges\PrintChanges
+    \cs_set_eq:NN \PrintChanges\relax
+    \cs_set_eq:NN \@ltxdoc@theglossary\theglossary
+    \cs_set_eq:NN \@ltxdoc@endtheglossary\endtheglossary
+    \part{\currentfile}
+    {
+      \cs_set_eq:NN \ttfamily\relax
+      \cs_gset_nopar:Npx \filekey{\filekey, \thepart={\ttfamily\currentfile}}
+    }
+    \DocInput{\currentfile}
+    \cs_set_eq:NN \PrintIndex\@ltxdoc@PrintIndex
+    \cs_set_eq:NN \PrintChanges\@ltxdoc@PrintChanges
+    \cs_set_eq:NN \theglossary\@ltxdoc@theglossary
+    \cs_set_eq:NN \endtheglossary\@ltxdoc@endtheglossary
+    \clearpage
+    \@writeckpt{#1}
+    \immediate\closeout\@partaux
+  \else
+    \@nameuse{cp@#1}
+  \fi
+  \cs_set_eq:NN \@auxout\@mainaux
+}
+\cs_gset_nopar:Npn \codeline@wrindex #1 {
+  \immediate\write\@indexfile {
+    \string\indexentry{#1}
+    {\filesep\number\c@CodelineNo}
+  }
+}
+\cs_set_eq:NN \filesep \@empty
+\cs_set_nopar:Npn \docincludeaux {
+  \cs_set_nopar:Npn \thepart {\alphalph{part}}
+  \cs_set_nopar:Npn \filesep {\thepart-}
+  \cs_set_eq:NN \filekey\@gobble
+  \g@addto@macro\index@prologue{
+    \cs_gset_nopar:Npn\@oddfoot{
+      \parbox{\textwidth}{
+        \strut\footnotesize
+        \raggedright{\bfseries File~Key:}~\filekey
+      }
+    }
+    \cs_set_eq:NN \@evenfoot\@oddfoot
+  }
+  \cs_gset_eq:NN \docincludeaux\relax
+  \cs_gset_nopar:Npn\@oddfoot{
+    \expandafter\ifx\csname ver@\currentfile\endcsname\relax
+      File~\thepart :~{\ttfamily\currentfile}~
+    \else
+      \GetFileInfo{\currentfile}
+      File~\thepart :~{\ttfamily\filename}~
+      Date:~\ExplFileDate\ % space
+      Version~\ExplFileVersion
+    \fi
+    \hfill\thepage
+  }
+  \cs_set_eq:NN \@evenfoot \@oddfoot
+}
+\seq_new:N \g_doc_functions_seq
+\seq_new:N \g_doc_macros_seq
+\seq_new:N \g_doc_not_tested_seq
+\iow_open:Nn \g_write_func_stream { \jobname.cmds }
+\cs_new_nopar:Npn \doc_show_functions_defined: {
+  \bool_if:nT { \g_doc_implementation_bool && \g_doc_checkfunc_bool } {
+    \typeout{ ======================================== ^^J }
+
+    \tl_clear:N \l_tmpa_tl
+    \seq_map_inline:Nn \g_doc_functions_seq {
+      \seq_if_in:NnT \g_doc_macros_seq {##1} {
+        \tl_put_right:Nn \l_tmpa_tl { ##1 ^^J }
+        \iow_now:Nn \g_write_func_stream { ##1 }
+      }
+    }
+    \iow_close:N \g_write_func_stream
+    \doc_functions_typeout:n {
+      Functions~both~documented~and~defined:^^J (In~order~of~being~documented)
+    }
+
+    \seq_map_inline:Nn \g_doc_functions_seq {
+      \seq_if_in:NnF \g_doc_macros_seq {##1} {
+        \tl_put_right:Nn \l_tmpa_tl { ##1 ^^J }
+      }
+    }
+    \doc_functions_typeout:n { Functions~documented~but~not~defined: }
+
+    \seq_map_inline:Nn \g_doc_macros_seq {
+      \seq_if_in:NnF \g_doc_functions_seq {##1} {
+        \tl_put_right:Nn \l_tmpa_tl { ##1 ^^J }
+      }
+    }
+    \doc_functions_typeout:n { Functions~defined~but~not~documented: }
+
+    \typeout{ ======================================== }
+  }
+}
+\AtEndDocument{ \doc_show_functions_defined: }
+\cs_set_nopar:Npn \doc_functions_typeout:n #1 {
+  \tl_if_empty:NF \l_tmpa_tl {
+    \typeout{
+      -------------------------------------- ^^J #1 ^^J
+      -------------------------------------- ^^J \l_tmpa_tl
+    }
+    \tl_clear:N \l_tmpa_tl
+  }
+}
+\cs_new:Npn \doc_show_not_tested: {
+  \bool_if:NT \g_doc_checktest_bool
+  {
+    \bool_if:nT { !(\seq_if_empty_p:N  \g_doc_not_tested_seq) ||
+                  !(\prop_if_empty_p:N \g_doc_missing_tests_prop) }
+    {
+      \tl_clear:N \l_tmpa_tl
+      \prop_if_empty:NF \g_doc_missing_tests_prop
+      {
+        \tl_put_right:Nn \l_tmpa_tl
+        {
+          ^^J^^JThe~ following~ macro(s)~ have~ incomplete~ tests:^^J
+        }
+        \prop_map_inline:Nn \g_doc_missing_tests_prop
+        {
+          \tl_put_right:Nn \l_tmpa_tl
+          {^^J\space\space\space\space ##1
+           ^^J\space\space\space\space\space\space ##2}
+        }
+      }
+      \seq_if_empty:NF \g_doc_not_tested_seq
+      {
+        \tl_put_right:Nn \l_tmpa_tl
+        {
+          ^^J^^J
+          The~ following~ macro(s)~ do~ not~ have~ any~ tests:^^J
+        }
+        \seq_map_inline:Nn \g_doc_not_tested_seq
+        {
+          \clist_map_inline:nn {##1}
+          {
+            \tl_put_right:Nn \l_tmpa_tl {^^J\space\space\space\space ####1}
+          }
+        }
+        \int_set:Nn \l_tmpa_int {\etex_interactionmode:D}
+        \errorstopmode
+        \ClassError{l3doc}{\l_tmpa_tl}{}
+        \int_set:Nn \etex_interactionmode:D {\l_tmpa_int}
+      }
+    }
+  }
+}
+\AtEndDocument{ \doc_show_not_tested: }
+\g@addto@macro\theindex{\MakePrivateLetters}
+\cs_set:Npn \verbatimchar {&}
+\setcounter{IndexColumns}{2}
+\IndexPrologue{
+  \part*{Index}
+  \markboth{Index}{Index}
+  \addcontentsline{toc}{part}{Index}
+  The~italic~numbers~denote~the~pages~where~the~
+  corresponding~entry~is~described,~
+  numbers~underlined~point~to~the~definition,~
+  all~others~indicate~the~places~where~it~is~used.
+}
+
+\cs_set_nopar:Npn \doc_special_main_index:n #1 {
+  \index{
+    \@gobble#1
+    \actualchar
+    \string\verb\quotechar*\verbatimchar#1\verbatimchar
+    \encapchar
+    hdpgindex{\thepage}{usage}
+  }
+}
+\cs_set_nopar:Npn \doc_special_main_index:o { \exp_args:No \doc_special_main_index:n }
+\cs_set_nopar:Npn \hdpgindex #1#2#3 {
+  \csname\ifx\\#2\\relax\else#2\fi\endcsname{
+    \hyperlink{page.#1}{#3}
+  }
+}
+\g@addto@macro \PrintIndex { \AtEndDocument{ \typeout{^^J
+  ========================================^^J
+  Generate~the~index~by~executing^^J
+  \c_space_tl \c_space_tl \c_space_tl \c_space_tl
+  makeindex~-s~l3doc.ist~-o~\jobname.ind~\jobname.idx^^J
+  ========================================^^J
+  }}
+}
+\GlossaryPrologue{
+  \part*{Change~History}
+  {\GlossaryParms\ttfamily\hyphenchar\font=`\-}
+  \markboth{Change~History}{Change~History}
+  \addcontentsline{toc}{part}{Change~History}
+}
+\g@addto@macro \PrintChanges { \AtEndDocument{ \typeout{^^J
+  ========================================^^J
+  Generate~the~change~list~by~executing^^J
+  \c_space_tl \c_space_tl \c_space_tl \c_space_tl
+   makeindex~-s~gglo.ist~~-o~\jobname.gls~\jobname.glo^^J
+  ========================================^^J
+  }}
+}
+\bool_if:NTF \g_doc_full_bool {
+  \RecordChanges
+  \CodelineIndex
+  \EnableCrossrefs
+  \AlsoImplementation
+}{
+  \CodelineNumbered
+  \DisableCrossrefs
+  \OnlyDescription
+}
+%% 
+%%
+%% End of file `l3doc.cls'.
diff --git a/Master/texmf-dist/tex/latex/l3kernel/l3expan.sty b/Master/texmf-dist/tex/latex/l3kernel/l3expan.sty
new file mode 100644
index 00000000000..028e9cd3bc0
--- /dev/null
+++ b/Master/texmf-dist/tex/latex/l3kernel/l3expan.sty
@@ -0,0 +1,394 @@
+%%
+%% This is file `l3expan.sty',
+%% generated with the docstrip utility.
+%%
+%% The original source files were:
+%%
+%% l3expan.dtx  (with options: `package')
+%% 
+%% EXPERIMENTAL CODE
+%% 
+%% Do not distribute this file without also distributing the
+%% source files specified above.
+%% 
+%% Do not distribute a modified version of this file.
+%% 
+%% File: l3expan.dtx Copyright (C) 1990-2011 The LaTeX3 project
+%%
+%% It may be distributed and/or modified under the conditions of the
+%% LaTeX Project Public License (LPPL), either version 1.3c of this
+%% license or (at your option) any later version.  The latest version
+%% of this license is in the file
+%%
+%%    http://www.latex-project.org/lppl.txt
+%%
+%% This file is part of the "expl3 bundle" (The Work in LPPL)
+%% and all files in that bundle must be distributed together.
+%%
+%% The released version of this bundle is available from CTAN.
+%%
+%% -----------------------------------------------------------------------
+%%
+%% The development version of the bundle can be found at
+%%
+%%    http://www.latex-project.org/svnroot/experimental/trunk/
+%%
+%% for those people who are interested.
+%%
+%%%%%%%%%%%
+%% NOTE: %%
+%%%%%%%%%%%
+%%
+%%   Snapshots taken from the repository represent work in progress and may
+%%   not work or may contain conflicting material!  We therefore ask
+%%   people _not_ to put them into distributions, archives, etc. without
+%%   prior consultation with the LaTeX3 Project.
+%%
+%% -----------------------------------------------------------------------
+\RequirePackage{l3names}
+\GetIdInfo$Id: l3expan.dtx 2480 2011-06-20 17:25:23Z bruno $
+  {L3 Experimental argument expansion}
+\ProvidesExplPackage
+  {\ExplFileName}{\ExplFileDate}{\ExplFileVersion}{\ExplFileDescription}
+\package_check_loaded_expl:
+\cs_new_nopar:Npn \l_exp_tl { }
+\cs_new:Npn \exp_arg_next:nnn #1#2#3 { #2 \::: { #3 {#1} } }
+\cs_new:Npn \exp_arg_next_nobrace:nnn #1#2#3 { #2 \::: { #3 #1 } }
+\cs_new:Npn \::: #1 {#1}
+\cs_new:Npn \::n #1 \::: #2#3 { #1 \::: { #2 {#3} } }
+\cs_new:Npn \::N #1 \::: #2#3 { #1 \::: {#2#3} }
+\cs_new:Npn \::c #1 \::: #2#3
+  { \exp_after:wN \exp_arg_next_nobrace:nnn \cs:w #3 \cs_end: {#1} {#2} }
+\cs_new:Npn \::o #1 \::: #2#3
+  { \exp_after:wN \exp_arg_next:nnn \exp_after:wN {#3} {#1} {#2} }
+\cs_new:Npn \::f #1 \::: #2#3
+  {
+    \exp_after:wN \exp_arg_next:nnn
+      \exp_after:wN { \tex_romannumeral:D -`0 #3 }
+      {#1} {#2}
+  }
+\use:nn { \cs_new_eq:NN \exp_stop_f: } { ~ }
+\cs_new_protected:Npn \::x #1 \::: #2#3
+  {
+    \cs_set_nopar:Npx \l_exp_tl { {#3} }
+    \exp_after:wN \exp_arg_next:nnn \l_exp_tl {#1} {#2}
+  }
+\cs_new:Npn \::V #1 \::: #2#3
+  {
+    \exp_after:wN \exp_arg_next:nnn
+      \exp_after:wN { \tex_romannumeral:D \exp_eval_register:N #3 }
+      {#1} {#2}
+}
+\cs_new:Npn \::v # 1\::: #2#3
+  {
+    \exp_after:wN \exp_arg_next:nnn
+      \exp_after:wN { \tex_romannumeral:D \exp_eval_register:c {#3} }
+      {#1} {#2}
+  }
+\cs_new_nopar:Npn \exp_eval_register:N #1
+  {
+    \exp_after:wN \if_meaning:w \exp_not:N #1 #1
+      \if_meaning:w \scan_stop: #1
+        \exp_eval_error_msg:w
+      \fi:
+    \else:
+      \exp_after:wN \use_i_ii:nnn
+    \fi:
+    \exp_after:wN \c_zero \tex_the:D #1
+  }
+\cs_new_nopar:Npn \exp_eval_register:c #1
+  { \exp_after:wN \exp_eval_register:N \cs:w #1 \cs_end: }
+\group_begin:%
+\tex_catcode:D`\!=11\scan_stop:%
+\tex_catcode:D`\ =11\scan_stop:%
+\cs_new:Npn\exp_eval_error_msg:w#1\tex_the:D#2{%
+\fi:\fi:\exp_after:wN\c_zero\erroneous variable used!}%
+\group_end:%
+\cs_new:Npn \exp_args:No #1#2 { \exp_after:wN #1 \exp_after:wN {#2} }
+\cs_new:Npn \exp_args:NNo #1#2#3
+  { \exp_after:wN #1 \exp_after:wN #2 \exp_after:wN {#3} }
+\cs_new:Npn \exp_args:NNNo #1#2#3#4
+  { \exp_after:wN #1 \exp_after:wN#2 \exp_after:wN #3 \exp_after:wN {#4} }
+\cs_new:Npn \exp_args:cc #1#2
+  { \cs:w #1 \exp_after:wN \cs_end: \cs:w #2 \cs_end: }
+\cs_new:Npn \exp_args:NNc #1#2#3
+  { \exp_after:wN #1 \exp_after:wN #2 \cs:w # 3\cs_end: }
+\cs_new:Npn \exp_args:Ncc #1#2#3
+  { \exp_after:wN #1 \cs:w #2 \exp_after:wN \cs_end: \cs:w #3 \cs_end: }
+\cs_new:Npn \exp_args:Nccc #1#2#3#4
+  {
+    \exp_after:wN #1
+      \cs:w #2 \exp_after:wN \cs_end:
+      \cs:w #3 \exp_after:wN \cs_end:
+      \cs:w #4 \cs_end:
+  }
+\cs_new:Npn \exp_args:Nf #1#2
+  { \exp_after:wN #1 \exp_after:wN { \tex_romannumeral:D -`0 #2 } }
+\cs_new:Npn \exp_args:Nv #1#2
+  {
+    \exp_after:wN #1 \exp_after:wN
+      { \tex_romannumeral:D \exp_eval_register:c {#2} }
+  }
+\cs_new:Npn \exp_args:NV #1#2
+  {
+    \exp_after:wN #1 \exp_after:wN
+      { \tex_romannumeral:D \exp_eval_register:N #2 }
+  }
+\cs_new:Npn \exp_args:NNf #1#2#3
+  {
+    \exp_after:wN #1
+    \exp_after:wN #2
+    \exp_after:wN { \tex_romannumeral:D -`0 #3 }
+  }
+\cs_new:Npn \exp_args:NNv #1#2#3
+  {
+    \exp_after:wN #1
+    \exp_after:wN #2
+    \exp_after:wN { \tex_romannumeral:D \exp_eval_register:c {#3} }
+  }
+\cs_new:Npn \exp_args:NNV #1#2#3
+  {
+    \exp_after:wN #1
+    \exp_after:wN #2
+    \exp_after:wN { \tex_romannumeral:D \exp_eval_register:N #3 }
+  }
+\cs_new:Npn \exp_args:Nco #1#2#3
+  {
+    \exp_after:wN #1
+    \cs:w #2 \exp_after:wN \cs_end:
+    \exp_after:wN {#3}
+  }
+\cs_new:Npn \exp_args:Ncf #1#2#3
+  {
+    \exp_after:wN #1
+    \cs:w #2 \exp_after:wN \cs_end:
+    \exp_after:wN { \tex_romannumeral:D -`0 #3 }
+  }
+\cs_new_nopar:Npn \exp_args:NVV #1#2#3
+  {
+    \exp_after:wN #1
+    \exp_after:wN { \tex_romannumeral:D \exp_after:wN
+      \exp_eval_register:N \exp_after:wN #2 \exp_after:wN }
+    \exp_after:wN { \tex_romannumeral:D \exp_eval_register:N #3 }
+  }
+\cs_new:Npn \exp_args:NNNV #1#2#3#4
+  {
+    \exp_after:wN #1
+    \exp_after:wN #2
+    \exp_after:wN #3
+    \exp_after:wN { \tex_romannumeral:D \exp_eval_register:N #4 }
+  }
+\cs_new:Npn \exp_args:NcNc #1#2#3#4
+  {
+    \exp_after:wN #1
+    \cs:w #2 \exp_after:wN \cs_end:
+    \exp_after:wN #3
+    \cs:w #4 \cs_end:
+  }
+\cs_new:Npn \exp_args:NcNo #1#2#3#4
+  {
+    \exp_after:wN #1
+    \cs:w #2 \exp_after:wN \cs_end:
+    \exp_after:wN #3
+    \exp_after:wN {#4}
+  }
+\cs_new:Npn \exp_args:Ncco #1#2#3#4
+  {
+    \exp_after:wN #1
+    \cs:w #2 \exp_after:wN \cs_end:
+    \cs:w #3 \exp_after:wN \cs_end:
+    \exp_after:wN {#4}
+  }
+\cs_new_protected_nopar:Npn \exp_args:Nx { \::x \::: }
+\cs_new_nopar:Npn \exp_args:Nnc { \::n \::c \::: }
+\cs_new_nopar:Npn \exp_args:Nfo { \::f \::o \::: }
+\cs_new_nopar:Npn \exp_args:Nff { \::f \::f \::: }
+\cs_new_nopar:Npn \exp_args:Nnf { \::n \::f \::: }
+\cs_new_nopar:Npn \exp_args:Nno { \::n \::o \::: }
+\cs_new_nopar:Npn \exp_args:NnV { \::n \::V \::: }
+\cs_new_nopar:Npn \exp_args:Noc { \::o \::c \::: }
+\cs_new_nopar:Npn \exp_args:Noo { \::o \::o \::: }
+\cs_new_protected_nopar:Npn \exp_args:NNx { \::N \::x \::: }
+\cs_new_protected_nopar:Npn \exp_args:Ncx { \::c \::x \::: }
+\cs_new_protected_nopar:Npn \exp_args:Nnx { \::n \::x \::: }
+\cs_new_protected_nopar:Npn \exp_args:Nox { \::o \::x \::: }
+\cs_new_protected_nopar:Npn \exp_args:Nxo { \::x \::o \::: }
+\cs_new_protected_nopar:Npn \exp_args:Nxx { \::x \::x \::: }
+\cs_new_nopar:Npn \exp_args:NNno { \::N \::n \::o \::: }
+\cs_new_nopar:Npn \exp_args:NNoo { \::N \::o \::o \::: }
+\cs_new_nopar:Npn \exp_args:Nnnc { \::n \::n \::c \::: }
+\cs_new_nopar:Npn \exp_args:Nnno { \::n \::n \::o \::: }
+\cs_new_nopar:Npn \exp_args:Nooo { \::o \::o \::o \::: }
+\cs_new_protected_nopar:Npn \exp_args:NNnx { \::N \::n \::x \::: }
+\cs_new_protected_nopar:Npn \exp_args:NNox { \::N \::o \::x \::: }
+\cs_new_protected_nopar:Npn \exp_args:Nnnx { \::n \::n \::x \::: }
+\cs_new_protected_nopar:Npn \exp_args:Nnox { \::n \::o \::x \::: }
+\cs_new_protected_nopar:Npn \exp_args:Nccx { \::c \::c \::x \::: }
+\cs_new_protected_nopar:Npn \exp_args:Ncnx { \::c \::n \::x \::: }
+\cs_new_protected_nopar:Npn \exp_args:Noox { \::o \::o \::x \::: }
+\cs_new:Npn \exp_arg_last_unbraced:nn #1#2 { #2#1 }
+\cs_new:Npn \::f_unbraced \::: #1#2
+  {
+    \exp_after:wN \exp_arg_last_unbraced:nn
+      \exp_after:wN { \tex_romannumeral:D -`0 #2 } {#1}
+  }
+\cs_new:Npn \::o_unbraced \::: #1#2
+  { \exp_after:wN \exp_arg_last_unbraced:nn \exp_after:wN {#2} {#1} }
+\cs_new:Npn \::V_unbraced \::: #1#2
+  {
+    \exp_after:wN \exp_arg_last_unbraced:nn
+      \exp_after:wN { \tex_romannumeral:D \exp_eval_register:N #2 } {#1}
+  }
+\cs_new:Npn \::v_unbraced \::: #1#2
+  {
+    \exp_after:wN \exp_arg_last_unbraced:nn
+      \exp_after:wN { \tex_romannumeral:D \exp_eval_register:c {#2} } {#1}
+  }
+\cs_new:Npn \exp_last_unbraced:NV #1#2
+  { \exp_after:wN #1 \tex_romannumeral:D \exp_eval_register:N #2 }
+\cs_new:Npn \exp_last_unbraced:Nv #1#2
+  { \exp_after:wN #1 \tex_romannumeral:D \exp_eval_register:c {#2} }
+\cs_new:Npn \exp_last_unbraced:No #1#2 { \exp_after:wN #1 #2 }
+\cs_new:Npn \exp_last_unbraced:Nf #1#2
+  { \exp_after:wN #1 \tex_romannumeral:D -`0 #2 }
+\cs_new:Npn \exp_last_unbraced:NcV #1#2#3
+  {
+    \exp_after:wN #1
+    \cs:w #2 \exp_after:wN \cs_end:
+    \tex_romannumeral:D \exp_eval_register:N #3
+  }
+\cs_new:Npn \exp_last_unbraced:NNV #1#2#3
+  {
+    \exp_after:wN #1
+    \exp_after:wN #2
+    \tex_romannumeral:D \exp_eval_register:N #3
+  }
+\cs_new:Npn \exp_last_unbraced:NNo #1#2#3
+  { \exp_after:wN #1 \exp_after:wN #2 #3 }
+\cs_new_nopar:Npn \exp_last_unbraced:Noo { \::o \::o_unbraced \::: }
+\cs_new_nopar:Npn \exp_last_unbraced:Nfo { \::f \::o_unbraced \::: }
+\cs_new:Npn \exp_last_unbraced:NNNV #1#2#3#4
+  {
+    \exp_after:wN #1
+    \exp_after:wN #2
+    \exp_after:wN #3
+    \tex_romannumeral:D \exp_eval_register:N #4
+  }
+\cs_new:Npn \exp_last_unbraced:NNNo #1#2#3#4
+  { \exp_after:wN #1 \exp_after:wN #2 \exp_after:wN #3 #4 }
+\cs_new:Npn \exp_last_two_unbraced:Noo #1#2#3
+  { \exp_after:wN \exp_last_two_unbraced_aux:nnN \exp_after:wN {#3} {#2} #1 }
+\cs_new:Npn \exp_last_two_unbraced_aux:nnN #1#2#3
+   { \exp_after:wN #3 #2 #1 }
+\cs_new:Npn \exp_not:o #1 { \etex_unexpanded:D \exp_after:wN {#1} }
+\cs_new:Npn \exp_not:f #1
+  { \etex_unexpanded:D \exp_after:wN { \tex_romannumeral:D -`0 #1 } }
+\cs_new:Npn \exp_not:V #1
+  {
+    \etex_unexpanded:D \exp_after:wN
+      { \tex_romannumeral:D \exp_eval_register:N #1 }
+  }
+\cs_new:Npn \exp_not:v #1
+  {
+    \etex_unexpanded:D \exp_after:wN
+      { \tex_romannumeral:D \exp_eval_register:c {#1} }
+  }
+\cs_new:Npn \exp_not:c #1 { \exp_after:wN \exp_not:N \cs:w #1 \cs_end: }
+\cs_new_protected:Npn \cs_generate_variant:Nn #1
+  {
+    \chk_if_exist_cs:N #1
+    \cs_split_function:NN #1 \cs_generate_variant_aux:nnNNn
+    #1
+  }
+\cs_new:Npn \cs_generate_variant_aux:nnNNn #1#2#3#4#5
+  { \cs_generate_variant_aux:Nnnw #4 {#1}{#2} #5 , ? , \q_recursion_stop }
+\cs_new:Npn \cs_generate_variant_aux:Nnnw #1#2#3#4 ,
+  {
+    \if:w ? #4
+      \exp_after:wN \use_none_delimit_by_q_recursion_stop:w
+    \fi:
+    \exp_args:NNc \cs_generate_variant_aux:NNn
+      #1
+      { #2 : #4 \use:c { use_none: \cs_generate_variant_aux:N #4 ? } #3 }
+      {#4}
+    \cs_generate_variant_aux:Nnnw #1 {#2} {#3}
+  }
+\cs_new:Npn \cs_generate_variant_aux:NNn #1 #2 #3
+  {
+    \cs_if_free:NTF #2
+      {
+        \cs_generate_variant_aux:NNpx #1 #2
+          { \exp_not:c { exp_args:N #3 } \exp_not:N #1 }
+        \cs_generate_internal_variant:n {#3}
+      }
+      {
+        \iow_log:x
+          {
+            Variant~\token_to_str:N #2~%
+            already~defined;~ not~ changing~ it~on~line~%
+            \tex_the:D \tex_inputlineno:D
+          }
+      }
+  }
+\cs_new:Npn \cs_generate_variant_aux:N #1
+  {
+    \if:w ? #1
+      \exp_after:wN \use_none:nn
+    \fi:
+    n
+    \cs_generate_variant_aux:N
+  }
+\group_begin:
+  \tex_lccode:D `\Z = `\d \scan_stop:
+  \tex_lccode:D `\? =`\\ \scan_stop:
+  \tex_catcode:D `\P = 12 \scan_stop:
+  \tex_catcode:D `\R = 12 \scan_stop:
+  \tex_catcode:D `\O = 12 \scan_stop:
+  \tex_catcode:D `\T = 12 \scan_stop:
+  \tex_catcode:D `\E = 12 \scan_stop:
+  \tex_catcode:D `\C = 12 \scan_stop:
+  \tex_catcode:D `\Z = 12 \scan_stop:
+\tex_lowercase:D
+  {
+    \group_end:
+    \cs_new_nopar:Npn \cs_generate_variant_aux:NNpx #1
+      {
+        \exp_after:wN \cs_generate_variant_aux:w
+          \token_to_meaning:N #1 ? PROTECTEZ \q_stop
+      }
+    \cs_new:Npn \cs_generate_variant_aux:w #1 ? PROTECTEZ #2 \q_stop
+      {
+        \if_catcode:w a \etex_detokenize:D \exp_after:wN {#1} a
+          \exp_after:wN \cs_new_protected_nopar:Npx
+        \else:
+          \exp_after:wN \cs_new_nopar:Npx
+        \fi:
+      }
+  }
+\cs_new_protected:Npn \cs_generate_internal_variant:n #1
+  {
+    \cs_if_free:cT { exp_args:N #1 }
+      {
+        \cs_new:cpx { exp_args:N #1 }
+          { \cs_generate_internal_variant_aux:N #1 : }
+     }
+  }
+\cs_new:Npn \cs_generate_internal_variant_aux:N #1
+  {
+    \exp_not:c { :: #1 }
+    \if_meaning:w #1 :
+      \exp_after:wN \use_none:n
+    \fi:
+    \cs_generate_internal_variant_aux:N
+  }
+\cs_generate_variant:Nn \str_if_eq_p:nn {  V ,  o }
+\cs_generate_variant:Nn \str_if_eq_p:nn { nV , no , VV }
+\cs_generate_variant:Nn \str_if_eq:nnT  {  V ,  o }
+\cs_generate_variant:Nn \str_if_eq:nnT  { nV , no , VV }
+\cs_generate_variant:Nn \str_if_eq:nnF  {  V ,  o }
+\cs_generate_variant:Nn \str_if_eq:nnF  { nV , no , VV }
+\cs_generate_variant:Nn \str_if_eq:nnTF {  V ,  o }
+\cs_generate_variant:Nn \str_if_eq:nnTF { nV , no , VV }
+%% 
+%%
+%% End of file `l3expan.sty'.
diff --git a/Master/texmf-dist/tex/latex/l3kernel/l3file.sty b/Master/texmf-dist/tex/latex/l3kernel/l3file.sty
new file mode 100644
index 00000000000..e8272cde35f
--- /dev/null
+++ b/Master/texmf-dist/tex/latex/l3kernel/l3file.sty
@@ -0,0 +1,133 @@
+%%
+%% This is file `l3file.sty',
+%% generated with the docstrip utility.
+%%
+%% The original source files were:
+%%
+%% l3file.dtx  (with options: `package')
+%% 
+%% EXPERIMENTAL CODE
+%% 
+%% Do not distribute this file without also distributing the
+%% source files specified above.
+%% 
+%% Do not distribute a modified version of this file.
+%% 
+%% File: l3file.dtx Copyright (C) 2009-2011 The LaTeX3 Project
+%%
+%% It may be distributed and/or modified under the conditions of the
+%% LaTeX Project Public License (LPPL), either version 1.3c of this
+%% license or (at your option) any later version.  The latest version
+%% of this license is in the file
+%%
+%%    http://www.latex-project.org/lppl.txt
+%%
+%% This file is part of the "expl3 bundle" (The Work in LPPL)
+%% and all files in that bundle must be distributed together.
+%%
+%% The released version of this bundle is available from CTAN.
+%%
+%% -----------------------------------------------------------------------
+%%
+%% The development version of the bundle can be found at
+%%
+%%    http://www.latex-project.org/svnroot/experimental/trunk/
+%%
+%% for those people who are interested.
+%%
+%%%%%%%%%%%
+%% NOTE: %%
+%%%%%%%%%%%
+%%
+%%   Snapshots taken from the repository represent work in progress and may
+%%   not work or may contain conflicting material!  We therefore ask
+%%   people _not_ to put them into distributions, archives, etc. without
+%%   prior consultation with the LaTeX3 Project.
+%%
+%% -----------------------------------------------------------------------
+\RequirePackage{l3names}
+\GetIdInfo$Id: l3file.dtx 2478 2011-06-19 21:34:23Z joseph $
+  {L3 Experimental file operations}
+\ProvidesExplPackage
+  {\ExplFileName}{\ExplFileDate}{\ExplFileVersion}{\ExplFileDescription}
+\package_check_loaded_expl:
+\tl_new:N \g_file_current_name_tl
+\tl_gset_eq:NN \g_file_current_name_tl \@currname
+\seq_new:N \g_file_stack_seq
+\seq_new:N \g_file_record_seq
+\tl_new:N \l_file_name_tl
+\seq_new:N \l_file_search_path_seq
+\seq_new:N \l_file_search_path_saved_seq
+\cs_new_protected_nopar:Npn \file_add_path:nN #1#2
+  {
+    \ior_open:Nn \g_file_test_stream {#1}
+    \ior_if_eof:NTF \g_file_test_stream
+      { \file_add_path_search:nN {#1} #2 }
+      {
+        \ior_close:N \g_file_test_stream
+        \tl_set:Nx #2 {#1}
+      }
+  }
+\cs_new_protected_nopar:Npn \file_add_path_search:nN #1#2
+  {
+    \tl_clear:N #2
+    \cs_if_exist:NT \input@path
+      {
+        \seq_set_eq:NN \l_file_search_path_saved_seq \l_file_search_path_seq
+        \clist_map_inline:Nn \input@path
+          { \seq_put_right:Nn \l_file_search_path_seq {##1} }
+      }
+    \seq_map_inline:Nn \l_file_search_path_seq
+      {
+        \ior_open:Nn \g_file_test_stream { ##1 #1 }
+        \ior_if_eof:NF \g_file_test_stream
+          {
+            \tl_set:Nx #2 { ##1 #1 }
+            \seq_map_break:
+          }
+      }
+    \cs_if_exist:NT \input@path
+      { \seq_set_eq:NN \l_file_search_path_seq \l_file_search_path_saved_seq }
+    \ior_close:N \g_file_test_stream
+  }
+\prg_new_protected_conditional:Nnn \file_if_exist:n { T , F , TF }
+  {
+    \file_add_path:nN {#1} \l_file_name_tl
+    \tl_if_empty:NTF \l_file_name_tl
+      { \prg_return_false: }
+      { \prg_return_true: }
+  }
+\cs_new_protected_nopar:Npn \file_input:n #1
+  {
+    \file_add_path:nN {#1} \l_file_name_tl
+    \tl_if_empty:NF \l_file_name_tl
+      {
+        \@addtofilelist {#1}
+        \seq_gpush:NV \g_file_stack_seq \g_file_current_name_tl
+        \tl_gset:Nn \g_file_current_name_tl {#1}
+        \exp_after:wN \tex_input:D \l_file_name_tl ~
+        \seq_gpop:NN \g_file_stack_seq \g_file_current_name_tl
+      }
+  }
+\cs_new_protected_nopar:Npn \file_path_include:n #1
+  {
+    \seq_if_in:NnF  \l_file_search_path_seq {#1}
+      { \seq_put_right:Nn \l_file_search_path_seq {#1} }
+  }
+\cs_new_protected_nopar:Npn \file_path_remove:n #1
+  { \seq_remove_all:Nn \l_file_search_path_seq {#1} }
+\cs_new_protected_nopar:Npn \file_list:
+  {
+    \seq_remove_duplicates:N \g_file_record_seq
+    \iow_log:n { *~File~List~* }
+    \seq_map_inline:Nn \g_file_record_seq { \iow_log:n {##1} }
+    \iow_log:n { ************* }
+  }
+\AtBeginDocument
+  {
+    \clist_map_inline:Nn \@filelist
+      { \seq_put_right:Nn \g_file_record_seq {#1} }
+  }
+%% 
+%%
+%% End of file `l3file.sty'.
diff --git a/Master/texmf-dist/tex/latex/l3kernel/l3fp.sty b/Master/texmf-dist/tex/latex/l3kernel/l3fp.sty
new file mode 100644
index 00000000000..d3f98d4353a
--- /dev/null
+++ b/Master/texmf-dist/tex/latex/l3kernel/l3fp.sty
@@ -0,0 +1,3403 @@
+%%
+%% This is file `l3fp.sty',
+%% generated with the docstrip utility.
+%%
+%% The original source files were:
+%%
+%% l3fp.dtx  (with options: `package')
+%% 
+%% EXPERIMENTAL CODE
+%% 
+%% Do not distribute this file without also distributing the
+%% source files specified above.
+%% 
+%% Do not distribute a modified version of this file.
+%% 
+%% File: l3fp.dtx Copyright (C) 2010,2011 The LaTeX3 Project
+%%
+%% It may be distributed and/or modified under the conditions of the
+%% LaTeX Project Public License (LPPL), either version 1.3c of this
+%% license or (at your option) any later version.  The latest version
+%% of this license is in the file
+%%
+%%    http://www.latex-project.org/lppl.txt
+%%
+%% This file is part of the "expl3 bundle" (The Work in LPPL)
+%% and all files in that bundle must be distributed together.
+%%
+%% The released version of this bundle is available from CTAN.
+%%
+%% -----------------------------------------------------------------------
+%%
+%% The development version of the bundle can be found at
+%%
+%%    http://www.latex-project.org/svnroot/experimental/trunk/
+%%
+%% for those people who are interested.
+%%
+%%%%%%%%%%%
+%% NOTE: %%
+%%%%%%%%%%%
+%%
+%%   Snapshots taken from the repository represent work in progress and may
+%%   not work or may contain conflicting material!  We therefore ask
+%%   people _not_ to put them into distributions, archives, etc. without
+%%   prior consultation with the LaTeX3 Project.
+%%
+%% -----------------------------------------------------------------------
+\RequirePackage{l3names}
+\GetIdInfo$Id: l3fp.dtx 2478 2011-06-19 21:34:23Z joseph $
+  {L3 Experimental floating-point operations}
+\ProvidesExplPackage
+  {\ExplFileName}{\ExplFileDate}{\ExplFileVersion}{\ExplFileDescription}
+\package_check_loaded_expl:
+\int_const:Nn \c_forty_four { 44 }
+\int_const:Nn \c_one_million { 1 000 000 }
+\int_const:Nn \c_one_hundred_million { 100 000 000 }
+\int_const:Nn \c_five_hundred_million { 500 000 000 }
+\int_const:Nn \c_one_thousand_million { 1 000 000 000 }
+\int_new:N  \c_fp_pi_by_four_decimal_int
+\int_set:Nn \c_fp_pi_by_four_decimal_int { 785 398 158 }
+\int_new:N  \c_fp_pi_by_four_extended_int
+\int_set:Nn \c_fp_pi_by_four_extended_int { 897 448 310 }
+\int_new:N  \c_fp_pi_decimal_int
+\int_set:Nn \c_fp_pi_decimal_int { 141 592 653 }
+\int_new:N  \c_fp_pi_extended_int
+\int_set:Nn \c_fp_pi_extended_int { 589 793 238 }
+\int_new:N  \c_fp_two_pi_decimal_int
+\int_set:Nn \c_fp_two_pi_decimal_int { 283 185 307 }
+\int_new:N  \c_fp_two_pi_extended_int
+\int_set:Nn \c_fp_two_pi_extended_int { 179 586 477 }
+\tl_const:Nn \c_e_fp { + 2.718281828 e 0 }
+\tl_const:Nn \c_one_fp { + 1.000000000 e 0 }
+\tl_const:Nn \c_pi_fp { + 3.141592654 e 0 }
+\tl_const:Nn \c_undefined_fp { X 0.000000000 e 0 }
+\tl_const:Nn \c_zero_fp { + 0.000000000 e 0 }
+\tl_new:N \l_fp_arg_tl
+\int_new:N \l_fp_count_int
+\int_new:N \l_fp_div_offset_int
+\int_new:N \l_fp_exp_integer_int
+\int_new:N \l_fp_exp_decimal_int
+\int_new:N \l_fp_exp_extended_int
+\int_new:N \l_fp_exp_exponent_int
+\int_new:N \l_fp_input_a_sign_int
+\int_new:N \l_fp_input_a_integer_int
+\int_new:N \l_fp_input_a_decimal_int
+\int_new:N \l_fp_input_a_exponent_int
+\int_new:N \l_fp_input_b_sign_int
+\int_new:N \l_fp_input_b_integer_int
+\int_new:N \l_fp_input_b_decimal_int
+\int_new:N \l_fp_input_b_exponent_int
+\int_new:N \l_fp_input_a_extended_int
+\int_new:N \l_fp_input_b_extended_int
+\int_new:N \l_fp_mul_a_i_int
+\int_new:N \l_fp_mul_a_ii_int
+\int_new:N \l_fp_mul_a_iii_int
+\int_new:N \l_fp_mul_a_iv_int
+\int_new:N \l_fp_mul_a_v_int
+\int_new:N \l_fp_mul_a_vi_int
+\int_new:N \l_fp_mul_b_i_int
+\int_new:N \l_fp_mul_b_ii_int
+\int_new:N \l_fp_mul_b_iii_int
+\int_new:N \l_fp_mul_b_iv_int
+\int_new:N \l_fp_mul_b_v_int
+\int_new:N \l_fp_mul_b_vi_int
+\int_new:N \l_fp_mul_output_int
+\tl_new:N  \l_fp_mul_output_tl
+\int_new:N \l_fp_output_sign_int
+\int_new:N \l_fp_output_integer_int
+\int_new:N \l_fp_output_decimal_int
+\int_new:N \l_fp_output_exponent_int
+\int_new:N \l_fp_output_extended_int
+\bool_new:N \l_fp_round_carry_bool
+\tl_new:N \l_fp_round_decimal_tl
+\int_new:N \l_fp_round_position_int
+\int_new:N \l_fp_round_target_int
+\tl_new:N \l_fp_sign_tl
+\int_new:N \l_fp_split_sign_int
+\int_new:N \l_fp_tmp_int
+\tl_new:N \l_fp_tmp_tl
+\int_new:N \l_fp_trig_octant_int
+\int_new:N \l_fp_trig_sign_int
+\int_new:N \l_fp_trig_decimal_int
+\int_new:N \l_fp_trig_extended_int
+\cs_new_protected_nopar:Npn \fp_read:N #1
+  { \exp_after:wN \fp_read_aux:w #1 \q_stop }
+\cs_new_protected_nopar:Npn \fp_read_aux:w #1#2 . #3 e #4 \q_stop
+  {
+    \if:w #1 -
+      \l_fp_input_a_sign_int \c_minus_one
+    \else:
+      \l_fp_input_a_sign_int \c_one
+    \fi:
+    \l_fp_input_a_integer_int  #2 \scan_stop:
+    \l_fp_input_a_decimal_int  #3 \scan_stop:
+    \l_fp_input_a_exponent_int #4 \scan_stop:
+  }
+\cs_new_protected_nopar:Npn \fp_split:Nn #1#2
+  {
+    \tl_set:Nx \l_fp_tmp_tl {#2}
+    \tl_set_rescan:Nno \l_fp_tmp_tl { \char_set_catcode_ignore:n { 32 } }
+      { \l_fp_tmp_tl }
+    \l_fp_split_sign_int \c_one
+    \fp_split_sign:
+    \use:c { l_fp_input_ #1 _sign_int } \l_fp_split_sign_int
+    \exp_after:wN \fp_split_exponent:w \l_fp_tmp_tl e e \q_stop #1
+  }
+\cs_new_protected_nopar:Npn \fp_split_sign:
+  {
+    \if_int_compare:w \pdftex_strcmp:D
+      { \exp_after:wN \tl_head:w \l_fp_tmp_tl ? \q_stop } { - }
+        = \c_zero
+      \tl_set:Nx \l_fp_tmp_tl
+        {
+          \exp_after:wN
+            \tl_tail:w \l_fp_tmp_tl \prg_do_nothing: \q_stop
+        }
+      \l_fp_split_sign_int -\l_fp_split_sign_int
+      \exp_after:wN \fp_split_sign:
+    \else:
+      \if_int_compare:w \pdftex_strcmp:D
+        { \exp_after:wN \tl_head:w \l_fp_tmp_tl ? \q_stop } { + }
+          = \c_zero
+        \tl_set:Nx \l_fp_tmp_tl
+          {
+            \exp_after:wN
+              \tl_tail:w \l_fp_tmp_tl \prg_do_nothing: \q_stop
+          }
+        \exp_after:wN \exp_after:wN \exp_after:wN \fp_split_sign:
+       \fi:
+    \fi:
+  }
+\cs_new_protected_nopar:Npn \fp_split_exponent:w #1 e #2 e #3 \q_stop #4
+  {
+    \use:c { l_fp_input_ #4 _exponent_int }
+      \int_eval:w 0 #2 \scan_stop:
+    \tex_afterassignment:D \fp_split_aux_i:w
+    \use:c { l_fp_input_ #4 _integer_int }
+      \int_eval:w 0 #1 . . \q_stop #4
+  }
+\cs_new_protected_nopar:Npn \fp_split_aux_i:w #1 . #2 . #3 \q_stop
+  { \fp_split_aux_ii:w #2 000000000 \q_stop }
+\cs_new_protected_nopar:Npn \fp_split_aux_ii:w #1#2#3#4#5#6#7#8#9
+  { \fp_split_aux_iii:w {#1#2#3#4#5#6#7#8#9} }
+\cs_new_protected_nopar:Npn \fp_split_aux_iii:w #1#2 \q_stop
+  {
+    \l_fp_tmp_int 1 #1 \scan_stop:
+    \exp_after:wN \fp_split_decimal:w
+      \int_use:N \l_fp_tmp_int 000000000 \q_stop
+  }
+\cs_new_protected_nopar:Npn \fp_split_decimal:w #1#2#3#4#5#6#7#8#9
+  { \fp_split_decimal_aux:w {#2#3#4#5#6#7#8#9} }
+\cs_new_protected_nopar:Npn \fp_split_decimal_aux:w #1#2#3 \q_stop #4
+  {
+    \use:c { l_fp_input_ #4 _decimal_int } #1#2 \scan_stop:
+    \if_int_compare:w
+      \int_eval:w
+        \use:c { l_fp_input_ #4 _integer_int } +
+        \use:c { l_fp_input_ #4 _decimal_int }
+      \scan_stop:
+        = \c_zero
+      \use:c { l_fp_input_ #4 _sign_int } \c_one
+    \fi:
+    \if_int_compare:w
+      \use:c { l_fp_input_ #4 _integer_int } < \c_one_thousand_million
+    \else:
+      \exp_after:wN \fp_overflow_msg:
+    \fi:
+  }
+\cs_new_protected_nopar:Npn \fp_standardise:NNNN #1#2#3#4
+  {
+    \if_int_compare:w
+      \int_eval:w #2 + #3 = \c_zero
+      #1 \c_one
+      #4 \c_zero
+      \exp_after:wN \use_none:nnnn
+    \else:
+      \exp_after:wN \fp_standardise_aux:NNNN
+    \fi:
+    #1#2#3#4
+  }
+\cs_new_protected_nopar:Npn \fp_standardise_aux:NNNN #1#2#3#4
+  {
+    \cs_set_protected_nopar:Npn \fp_standardise_aux:
+      {
+        \if_int_compare:w #2 = \c_zero
+          \tex_advance:D #3 \c_one_thousand_million
+          \exp_after:wN \fp_standardise_aux:w
+            \int_use:N #3 \q_stop
+           \exp_after:wN \fp_standardise_aux:
+         \fi:
+      }
+    \cs_set_protected_nopar:Npn
+      \fp_standardise_aux:w ##1##2##3##4##5##6##7##8##9 \q_stop
+      {
+        #2 ##2 \scan_stop:
+        #3 ##3##4##5##6##7##8##9 0 \scan_stop:
+        \tex_advance:D #4 \c_minus_one
+      }
+    \fp_standardise_aux:
+    \cs_set_protected_nopar:Npn \fp_standardise_aux:
+      {
+        \if_int_compare:w #2 > \c_nine
+          \tex_advance:D #2 \c_one_thousand_million
+          \exp_after:wN \use_i:nn \exp_after:wN
+            \fp_standardise_aux:w \int_use:N #2
+           \exp_after:wN \fp_standardise_aux:
+         \fi:
+      }
+    \cs_set_protected_nopar:Npn
+      \fp_standardise_aux:w ##1##2##3##4##5##6##7##8##9
+      {
+        #2 ##1##2##3##4##5##6##7##8 \scan_stop:
+        \tex_advance:D #3 \c_one_thousand_million
+        \tex_divide:D #3 \c_ten
+        \tl_set:Nx \l_fp_tmp_tl
+          {
+            ##9
+            \exp_after:wN \use_none:n \int_use:N #3
+          }
+        #3 \l_fp_tmp_tl \scan_stop:
+        \tex_advance:D #4 \c_one
+      }
+    \fp_standardise_aux:
+    \if_int_compare:w #4 < \c_one_hundred
+      \if_int_compare:w #4 > -\c_one_hundred
+      \else:
+        #1 \c_one
+        #2 \c_zero
+        #3 \c_zero
+        #4 \c_zero
+      \fi:
+    \else:
+      \exp_after:wN \fp_overflow_msg:
+    \fi:
+  }
+\cs_new_protected_nopar:Npn \fp_standardise_aux: { }
+\cs_new_protected_nopar:Npn \fp_standardise_aux:w { }
+\cs_new_protected_nopar:Npn \fp_level_input_exponents:
+  {
+    \if_int_compare:w \l_fp_input_a_exponent_int > \l_fp_input_b_exponent_int
+      \exp_after:wN \fp_level_input_exponents_a:
+    \else:
+      \exp_after:wN \fp_level_input_exponents_b:
+    \fi:
+  }
+\cs_new_protected_nopar:Npn \fp_level_input_exponents_a:
+  {
+    \if_int_compare:w \l_fp_input_a_exponent_int > \l_fp_input_b_exponent_int
+      \tex_advance:D \l_fp_input_b_integer_int \c_one_thousand_million
+      \exp_after:wN \use_i:nn \exp_after:wN
+        \fp_level_input_exponents_a:NNNNNNNNN
+          \int_use:N \l_fp_input_b_integer_int
+      \exp_after:wN \fp_level_input_exponents_a:
+    \fi:
+  }
+\cs_new_protected_nopar:Npn \fp_level_input_exponents_a:NNNNNNNNN
+  #1#2#3#4#5#6#7#8#9
+  {
+    \l_fp_input_b_integer_int #1#2#3#4#5#6#7#8 \scan_stop:
+    \tex_advance:D \l_fp_input_b_decimal_int \c_one_thousand_million
+    \tex_divide:D \l_fp_input_b_decimal_int \c_ten
+    \tl_set:Nx \l_fp_tmp_tl
+      {
+        #9
+        \exp_after:wN \use_none:n
+          \int_use:N \l_fp_input_b_decimal_int
+      }
+    \l_fp_input_b_decimal_int \l_fp_tmp_tl \scan_stop:
+    \tex_advance:D \l_fp_input_b_exponent_int \c_one
+  }
+\cs_new_protected_nopar:Npn \fp_level_input_exponents_b:
+  {
+    \if_int_compare:w \l_fp_input_b_exponent_int > \l_fp_input_a_exponent_int
+      \tex_advance:D \l_fp_input_a_integer_int \c_one_thousand_million
+      \exp_after:wN \use_i:nn \exp_after:wN
+        \fp_level_input_exponents_b:NNNNNNNNN
+          \int_use:N \l_fp_input_a_integer_int
+      \exp_after:wN \fp_level_input_exponents_b:
+    \fi:
+  }
+\cs_new_protected_nopar:Npn \fp_level_input_exponents_b:NNNNNNNNN
+  #1#2#3#4#5#6#7#8#9
+  {
+    \l_fp_input_a_integer_int #1#2#3#4#5#6#7#8 \scan_stop:
+    \tex_advance:D \l_fp_input_a_decimal_int \c_one_thousand_million
+    \tex_divide:D \l_fp_input_a_decimal_int \c_ten
+    \tl_set:Nx \l_fp_tmp_tl
+      {
+        #9
+        \exp_after:wN \use_none:n
+          \int_use:N \l_fp_input_a_decimal_int
+      }
+    \l_fp_input_a_decimal_int \l_fp_tmp_tl \scan_stop:
+    \tex_advance:D \l_fp_input_a_exponent_int \c_one
+  }
+\cs_new_protected_nopar:Npn \fp_tmp:w #1#2 { }
+\cs_new_protected_nopar:Npn \fp_new:N #1
+  {
+    \tl_new:N #1
+    \tl_gset_eq:NN #1 \c_zero_fp
+  }
+\cs_generate_variant:Nn \fp_new:N { c }
+\cs_new_protected_nopar:Npn \fp_const:Nn #1#2
+  {
+    \fp_new:N #1
+    \fp_gset:Nn #1 {#2}
+  }
+\cs_generate_variant:Nn \fp_const:Nn { c }
+\cs_new_protected_nopar:Npn \fp_zero:N #1
+  { \tl_set_eq:NN #1 \c_zero_fp }
+\cs_new_protected_nopar:Npn \fp_gzero:N #1
+  { \tl_gset_eq:NN #1 \c_zero_fp }
+\cs_generate_variant:Nn \fp_zero:N { c }
+\cs_generate_variant:Nn \fp_gzero:N { c }
+\cs_new_protected_nopar:Npn \fp_set:Nn  { \fp_set_aux:NNn \tl_set:Nn }
+\cs_new_protected_nopar:Npn \fp_gset:Nn { \fp_set_aux:NNn \tl_gset:Nn }
+\cs_new_protected_nopar:Npn \fp_set_aux:NNn #1#2#3
+  {
+    \group_begin:
+      \fp_split:Nn a {#3}
+      \fp_standardise:NNNN
+        \l_fp_input_a_sign_int
+        \l_fp_input_a_integer_int
+        \l_fp_input_a_decimal_int
+        \l_fp_input_a_exponent_int
+      \tex_advance:D \l_fp_input_a_decimal_int \c_one_thousand_million
+      \cs_set_protected_nopar:Npx \fp_tmp:w
+        {
+          \group_end:
+          #1 \exp_not:N #2
+            {
+              \if_int_compare:w \l_fp_input_a_sign_int < \c_zero
+                -
+              \else:
+                +
+              \fi:
+              \int_use:N \l_fp_input_a_integer_int
+              .
+              \exp_after:wN \use_none:n
+                \int_use:N \l_fp_input_a_decimal_int
+              e
+              \int_use:N \l_fp_input_a_exponent_int
+            }
+        }
+    \fp_tmp:w
+  }
+\cs_generate_variant:Nn \fp_set:Nn  { c }
+\cs_generate_variant:Nn \fp_gset:Nn { c }
+\cs_new_protected_nopar:Npn \fp_set_from_dim:Nn
+  { \fp_set_from_dim_aux:NNn \tl_set:Nx }
+\cs_new_protected_nopar:Npn \fp_gset_from_dim:Nn
+  { \fp_set_from_dim_aux:NNn \tl_gset:Nx }
+\cs_new_protected_nopar:Npn \fp_set_from_dim_aux:NNn #1#2#3
+  {
+    \group_begin:
+      \l_fp_tmp_skip \etex_glueexpr:D #3 \scan_stop:
+      \l_fp_tmp_dim \l_fp_tmp_skip
+      \fp_split:Nn a
+        {
+          \exp_after:wN \fp_set_from_dim_aux:w
+            \dim_use:N \l_fp_tmp_dim
+        }
+      \fp_standardise:NNNN
+        \l_fp_input_a_sign_int
+        \l_fp_input_a_integer_int
+        \l_fp_input_a_decimal_int
+        \l_fp_input_a_exponent_int
+      \tex_advance:D \l_fp_input_a_decimal_int \c_one_thousand_million
+      \cs_set_protected_nopar:Npx \fp_tmp:w
+        {
+          \group_end:
+          #1 \exp_not:N #2
+            {
+              \if_int_compare:w \l_fp_input_a_sign_int < \c_zero
+                -
+              \else:
+                +
+              \fi:
+              \int_use:N \l_fp_input_a_integer_int
+              .
+              \exp_after:wN \use_none:n
+                \int_use:N \l_fp_input_a_decimal_int
+              e
+              \int_use:N \l_fp_input_a_exponent_int
+            }
+        }
+    \fp_tmp:w
+  }
+\cs_set_protected_nopar:Npx \fp_set_from_dim_aux:w
+  {
+    \cs_set_nopar:Npn \exp_not:N \fp_set_from_dim_aux:w
+      ##1 \tl_to_str:n { pt } {##1}
+  }
+\fp_set_from_dim_aux:w
+\cs_generate_variant:Nn \fp_set_from_dim:Nn  { c }
+\cs_generate_variant:Nn \fp_gset_from_dim:Nn { c }
+\dim_new:N \l_fp_tmp_dim
+\skip_new:N \l_fp_tmp_skip
+\cs_new_eq:NN \fp_set_eq:NN  \tl_set_eq:NN
+\cs_new_eq:NN \fp_set_eq:cN  \tl_set_eq:cN
+\cs_new_eq:NN \fp_set_eq:Nc  \tl_set_eq:Nc
+\cs_new_eq:NN \fp_set_eq:cc  \tl_set_eq:cc
+\cs_new_eq:NN \fp_gset_eq:NN \tl_gset_eq:NN
+\cs_new_eq:NN \fp_gset_eq:cN \tl_gset_eq:cN
+\cs_new_eq:NN \fp_gset_eq:Nc \tl_gset_eq:Nc
+\cs_new_eq:NN \fp_gset_eq:cc \tl_gset_eq:cc
+\cs_new_eq:NN \fp_show:N \tl_show:N
+\cs_new_eq:NN \fp_show:c \tl_show:c
+\cs_new_nopar:Npn \fp_use:N #1
+  { \exp_after:wN \fp_use_aux:w #1 \q_stop }
+\cs_generate_variant:Nn \fp_use:N { c }
+\cs_new_nopar:Npn \fp_use_aux:w #1#2 e #3 \q_stop
+  {
+    \if:w #1 -
+      -
+    \fi:
+    \if_int_compare:w #3 > \c_zero
+      \exp_after:wN \fp_use_large:w
+    \else:
+      \if_int_compare:w #3 < \c_zero
+        \exp_after:wN \exp_after:wN \exp_after:wN
+          \fp_use_small:w
+      \else:
+        \exp_after:wN \exp_after:wN \exp_after:wN \fp_use_none:w
+      \fi:
+    \fi:
+    #2 e #3 \q_stop
+  }
+\cs_new_nopar:Npn \fp_use_none:w #1 e #2 \q_stop {#1}
+\cs_new_nopar:Npn \fp_use_small:w #1 . #2 e #3 \q_stop
+  {
+    0 .
+    \prg_replicate:nn { -#3 - 1 } { 0 }
+    #1#2
+  }
+\cs_new_nopar:Npn \fp_use_large:w #1 . #2 e #3 \q_stop
+  {
+    \if_int_compare:w #3 < \c_ten
+      \exp_after:wN \fp_use_large_aux_i:w
+    \else:
+      \exp_after:wN \fp_use_large_aux_ii:w
+    \fi:
+    #1#2 e #3 \q_stop
+  }
+\cs_new_nopar:Npn \fp_use_large_aux_i:w #1#2 e #3 \q_stop
+  {
+    #1
+    \use:c { fp_use_large_aux_ #3 :w } #2 \q_stop
+  }
+\cs_new_nopar:cpn { fp_use_large_aux_1:w } #1#2 \q_stop { #1 . #2 }
+\cs_new_nopar:cpn { fp_use_large_aux_2:w } #1#2#3 \q_stop
+  { #1#2 . #3 }
+\cs_new_nopar:cpn { fp_use_large_aux_3:w } #1#2#3#4 \q_stop
+  { #1#2#3 . #4 }
+\cs_new_nopar:cpn { fp_use_large_aux_4:w } #1#2#3#4#5 \q_stop
+  { #1#2#3#4 . #5 }
+\cs_new_nopar:cpn { fp_use_large_aux_5:w } #1#2#3#4#5#6 \q_stop
+  { #1#2#3#4#5 . #6 }
+\cs_new_nopar:cpn { fp_use_large_aux_6:w } #1#2#3#4#5#6#7 \q_stop
+  { #1#2#3#4#5#6 . #7 }
+\cs_new_nopar:cpn { fp_use_large_aux_7:w } #1#2#3#4#5#6#7#8 \q_stop
+  { #1#2#3#4#6#7 . #8 }
+\cs_new_nopar:cpn { fp_use_large_aux_8:w } #1#2#3#4#5#6#7#8#9 \q_stop
+  { #1#2#3#4#5#6#7#8 . #9 }
+\cs_new_nopar:cpn { fp_use_large_aux_9:w } #1 \q_stop { #1 . }
+\cs_new_nopar:Npn \fp_use_large_aux_ii:w #1 e #2 \q_stop
+  {
+    #1
+    \prg_replicate:nn { #2 - 9 } { 0 }
+    .
+  }
+\cs_new_nopar:Npn \fp_to_dim:N #1 { \fp_use:N #1 pt }
+\cs_generate_variant:Nn \fp_to_dim:N { c }
+\cs_new_nopar:Npn \fp_to_int:N #1
+  { \exp_after:wN \fp_to_int_aux:w #1 \q_stop }
+\cs_generate_variant:Nn \fp_to_int:N { c }
+\cs_new_nopar:Npn \fp_to_int_aux:w #1#2 e #3 \q_stop
+  {
+    \if:w #1 -
+      -
+    \fi:
+    \if_int_compare:w #3 < \c_zero
+      \exp_after:wN \fp_to_int_small:w
+    \else:
+      \exp_after:wN \fp_to_int_large:w
+    \fi:
+    #2 e #3 \q_stop
+  }
+\cs_new_nopar:Npn \fp_to_int_small:w #1 . #2 e #3 \q_stop
+  {
+    \if_int_compare:w #3 > \c_one
+    \else:
+      \if_int_compare:w #1 < \c_five
+        0
+      \else:
+        1
+      \fi:
+    \fi:
+  }
+\cs_new_nopar:Npn \fp_to_int_large:w #1 . #2 e #3 \q_stop
+  {
+    \if_int_compare:w #3 < \c_ten
+      \exp_after:wN \fp_to_int_large_aux_i:w
+    \else:
+      \exp_after:wN \fp_to_int_large_aux_ii:w
+    \fi:
+    #1#2 e #3 \q_stop
+  }
+\cs_new_nopar:Npn \fp_to_int_large_aux_i:w #1#2 e #3 \q_stop
+  { \use:c { fp_to_int_large_aux_ #3 :w } #2 \q_stop {#1} }
+\cs_new_nopar:cpn { fp_to_int_large_aux_1:w } #1#2 \q_stop
+  { \fp_to_int_large_aux:nnn { #2 0 } {#1} }
+\cs_new_nopar:cpn { fp_to_int_large_aux_2:w } #1#2#3 \q_stop
+  { \fp_to_int_large_aux:nnn { #3 00 } {#1#2} }
+\cs_new_nopar:cpn { fp_to_int_large_aux_3:w } #1#2#3#4 \q_stop
+  { \fp_to_int_large_aux:nnn { #4 000 } {#1#2#3} }
+\cs_new_nopar:cpn { fp_to_int_large_aux_4:w } #1#2#3#4#5 \q_stop
+  { \fp_to_int_large_aux:nnn { #5 0000 } {#1#2#3#4} }
+\cs_new_nopar:cpn { fp_to_int_large_aux_5:w } #1#2#3#4#5#6 \q_stop
+  { \fp_to_int_large_aux:nnn { #6 00000 } {#1#2#3#4#5} }
+\cs_new_nopar:cpn { fp_to_int_large_aux_6:w } #1#2#3#4#5#6#7 \q_stop
+  { \fp_to_int_large_aux:nnn { #7 000000 } {#1#2#3#4#5#6} }
+\cs_new_nopar:cpn { fp_to_int_large_aux_7:w } #1#2#3#4#5#6#7#8 \q_stop
+  { \fp_to_int_large_aux:nnn { #8 0000000 } {#1#2#3#4#5#6#7} }
+\cs_new_nopar:cpn { fp_to_int_large_aux_8:w } #1#2#3#4#5#6#7#8#9 \q_stop
+  { \fp_to_int_large_aux:nnn { #9 00000000 } {#1#2#3#4#5#6#7#8} }
+\cs_new_nopar:cpn { fp_to_int_large_aux_9:w } #1 \q_stop {#1}
+\cs_new_nopar:Npn \fp_to_int_large_aux:nnn #1#2#3
+  {
+    \if_int_compare:w #1 < \c_five_hundred_million
+      #3#2
+    \else:
+      \int_value:w \int_eval:w #3#2 + 1 \int_eval_end:
+    \fi:
+  }
+\cs_new_nopar:Npn \fp_to_int_large_aux_ii:w #1 e #2 \q_stop
+  {
+    #1
+    \prg_replicate:nn { #2 - 9 } { 0 }
+  }
+\cs_new_nopar:Npn \fp_to_tl:N #1
+  { \exp_after:wN \fp_to_tl_aux:w #1 \q_stop }
+\cs_generate_variant:Nn \fp_to_tl:N { c }
+\cs_new_nopar:Npn \fp_to_tl_aux:w #1#2 e #3 \q_stop
+  {
+    \if:w #1 -
+      -
+    \fi:
+    \if_int_compare:w #3 < \c_zero
+      \exp_after:wN \fp_to_tl_small:w
+    \else:
+      \exp_after:wN \fp_to_tl_large:w
+    \fi:
+    #2 e #3 \q_stop
+  }
+\cs_new_nopar:Npn \fp_to_tl_large:w #1 e #2 \q_stop
+  {
+    \if_int_compare:w #2 < \c_ten
+      \exp_after:wN \fp_to_tl_large_aux_i:w
+    \else:
+      \exp_after:wN \fp_to_tl_large_aux_ii:w
+    \fi:
+    #1 e #2 \q_stop
+  }
+\cs_new_nopar:Npn \fp_to_tl_large_aux_i:w #1 e #2 \q_stop
+  { \use:c { fp_to_tl_large_ #2 :w } #1 \q_stop }
+\cs_new_nopar:Npn \fp_to_tl_large_aux_ii:w #1 . #2 e #3 \q_stop
+  {
+    #1
+    \fp_to_tl_large_zeros:NNNNNNNNN #2
+    e #3
+  }
+\cs_new_nopar:cpn { fp_to_tl_large_0:w } #1 . #2 \q_stop
+  {
+    #1
+    \fp_to_tl_large_zeros:NNNNNNNNN #2
+  }
+\cs_new_nopar:cpn { fp_to_tl_large_1:w } #1 . #2#3 \q_stop
+  {
+    #1#2
+    \fp_to_tl_large_zeros:NNNNNNNNN #3 0
+  }
+\cs_new_nopar:cpn { fp_to_tl_large_2:w } #1 . #2#3#4 \q_stop
+  {
+    #1#2#3
+    \fp_to_tl_large_zeros:NNNNNNNNN #4 00
+  }
+\cs_new_nopar:cpn { fp_to_tl_large_3:w } #1 . #2#3#4#5 \q_stop
+  {
+    #1#2#3#4
+    \fp_to_tl_large_zeros:NNNNNNNNN #5 000
+  }
+\cs_new_nopar:cpn { fp_to_tl_large_4:w } #1 . #2#3#4#5#6 \q_stop
+  {
+    #1#2#3#4#5
+    \fp_to_tl_large_zeros:NNNNNNNNN #6 0000
+  }
+\cs_new_nopar:cpn { fp_to_tl_large_5:w } #1 . #2#3#4#5#6#7 \q_stop
+  {
+    #1#2#3#4#5#6
+    \fp_to_tl_large_zeros:NNNNNNNNN #7 00000
+  }
+\cs_new_nopar:cpn { fp_to_tl_large_6:w } #1 . #2#3#4#5#6#7#8 \q_stop
+  {
+    #1#2#3#4#5#6#7
+    \fp_to_tl_large_zeros:NNNNNNNNN #8 000000
+  }
+\cs_new_nopar:cpn { fp_to_tl_large_7:w } #1 . #2#3#4#5#6#7#8#9 \q_stop
+  {
+    #1#2#3#4#5#6#7#8
+    \fp_to_tl_large_zeros:NNNNNNNNN #9 0000000
+  }
+\cs_new_nopar:cpn { fp_to_tl_large_8:w } #1 .
+  {
+    #1
+    \use:c { fp_to_tl_large_8_aux:w }
+  }
+\cs_new_nopar:cpn { fp_to_tl_large_8_aux:w } #1#2#3#4#5#6#7#8#9 \q_stop
+  {
+    #1#2#3#4#5#6#7#8
+    \fp_to_tl_large_zeros:NNNNNNNNN #9 00000000
+  }
+\cs_new_nopar:cpn { fp_to_tl_large_9:w } #1 . #2 \q_stop {#1#2}
+\cs_new_nopar:Npn \fp_to_tl_small:w #1 e #2 \q_stop
+  {
+    \if_int_compare:w #2 = \c_minus_one
+      \exp_after:wN \fp_to_tl_small_one:w
+    \else:
+      \if_int_compare:w #2 = -\c_two
+        \exp_after:wN \exp_after:wN \exp_after:wN \fp_to_tl_small_two:w
+      \else:
+        \exp_after:wN \exp_after:wN \exp_after:wN \fp_to_tl_small_aux:w
+      \fi:
+    \fi:
+    #1 e #2 \q_stop
+  }
+\cs_new_nopar:Npn \fp_to_tl_small_one:w #1 . #2 e #3 \q_stop
+  {
+    \if_int_compare:w \fp_use_ix:NNNNNNNNN #2 > \c_four
+      \if_int_compare:w
+        \int_eval:w #1 \fp_use_i_to_iix:NNNNNNNNN #2 + 1
+          < \c_one_thousand_million
+        0.
+        \exp_after:wN \fp_to_tl_small_zeros:NNNNNNNNN
+          \int_value:w \int_eval:w
+              #1 \fp_use_i_to_iix:NNNNNNNNN #2 + 1
+            \int_eval_end:
+      \else:
+        1
+      \fi:
+    \else:
+      0. #1
+      \fp_to_tl_small_zeros:NNNNNNNNN #2
+    \fi:
+  }
+\cs_new_nopar:Npn \fp_to_tl_small_two:w #1 . #2 e #3 \q_stop
+  {
+    \if_int_compare:w \fp_use_iix_ix:NNNNNNNNN #2 > \c_forty_four
+      \if_int_compare:w
+        \int_eval:w #1 \fp_use_i_to_vii:NNNNNNNNN #2 0 + \c_ten
+          < \c_one_thousand_million
+        0.0
+        \exp_after:wN \fp_to_tl_small_zeros:NNNNNNNNN
+          \int_value:w \int_eval:w
+              #1 \fp_use_i_to_vii:NNNNNNNNN #2 0 + \c_ten
+            \int_eval_end:
+      \else:
+        0.1
+      \fi:
+    \else:
+      0.0
+      #1
+      \fp_to_tl_small_zeros:NNNNNNNNN #2
+    \fi:
+  }
+\cs_new_nopar:Npn \fp_to_tl_small_aux:w #1 . #2 e #3 \q_stop
+  {
+    #1
+    \fp_to_tl_large_zeros:NNNNNNNNN #2
+    e #3
+  }
+\cs_new_nopar:Npn \fp_to_tl_large_zeros:NNNNNNNNN #1#2#3#4#5#6#7#8#9
+  {
+    \if_int_compare:w #9 = \c_zero
+      \if_int_compare:w #8 = \c_zero
+        \if_int_compare:w #7 = \c_zero
+          \if_int_compare:w #6 = \c_zero
+            \if_int_compare:w #5 = \c_zero
+              \if_int_compare:w #4 = \c_zero
+                \if_int_compare:w #3 = \c_zero
+                  \if_int_compare:w #2 = \c_zero
+                    \if_int_compare:w #1 = \c_zero
+                    \else:
+                      . #1
+                    \fi:
+                  \else:
+                    . #1#2
+                  \fi:
+                \else:
+                  . #1#2#3
+                \fi:
+              \else:
+                . #1#2#3#4
+              \fi:
+            \else:
+              . #1#2#3#4#5
+            \fi:
+          \else:
+            . #1#2#3#4#5#6
+          \fi:
+        \else:
+          . #1#2#3#4#5#6#7
+        \fi:
+      \else:
+         . #1#2#3#4#5#6#7#8
+      \fi:
+    \else:
+      . #1#2#3#4#5#6#7#8#9
+    \fi:
+  }
+\cs_new_nopar:Npn \fp_to_tl_small_zeros:NNNNNNNNN #1#2#3#4#5#6#7#8#9
+  {
+    \if_int_compare:w #9 = \c_zero
+      \if_int_compare:w #8 = \c_zero
+        \if_int_compare:w #7 = \c_zero
+          \if_int_compare:w #6 = \c_zero
+            \if_int_compare:w #5 = \c_zero
+              \if_int_compare:w #4 = \c_zero
+                \if_int_compare:w #3 = \c_zero
+                  \if_int_compare:w #2 = \c_zero
+                    \if_int_compare:w #1 = \c_zero
+                    \else:
+                      #1
+                    \fi:
+                  \else:
+                    #1#2
+                  \fi:
+                \else:
+                  #1#2#3
+                \fi:
+              \else:
+                #1#2#3#4
+              \fi:
+            \else:
+              #1#2#3#4#5
+            \fi:
+          \else:
+            #1#2#3#4#5#6
+          \fi:
+        \else:
+          #1#2#3#4#5#6#7
+        \fi:
+      \else:
+         #1#2#3#4#5#6#7#8
+      \fi:
+    \else:
+      #1#2#3#4#5#6#7#8#9
+    \fi:
+  }
+\cs_new_nopar:Npn \fp_use_iix_ix:NNNNNNNNN #1#2#3#4#5#6#7#8#9 {#8#9}
+\cs_new_nopar:Npn \fp_use_ix:NNNNNNNNN #1#2#3#4#5#6#7#8#9 {#9}
+\cs_new_nopar:Npn \fp_use_i_to_vii:NNNNNNNNN #1#2#3#4#5#6#7#8#9
+  {#1#2#3#4#5#6#7}
+\cs_new_nopar:Npn \fp_use_i_to_iix:NNNNNNNNN #1#2#3#4#5#6#7#8#9
+  {#1#2#3#4#5#6#7#8}
+\cs_new_protected_nopar:Npn \fp_round_figures:Nn
+  { \fp_round_figures_aux:NNn \tl_set:Nn }
+\cs_generate_variant:Nn \fp_round_figures:Nn { c }
+\cs_new_protected_nopar:Npn \fp_ground_figures:Nn
+  { \fp_round_figures_aux:NNn \tl_gset:Nn }
+\cs_generate_variant:Nn \fp_ground_figures:Nn { c }
+\cs_new_protected_nopar:Npn \fp_round_figures_aux:NNn #1#2#3
+  {
+    \group_begin:
+      \fp_read:N #2
+      \int_set:Nn \l_fp_round_target_int { #3 - 1 }
+      \if_int_compare:w \l_fp_round_target_int < \c_ten
+        \exp_after:wN \fp_round:
+      \fi:
+      \tex_advance:D \l_fp_input_a_decimal_int \c_one_thousand_million
+      \cs_set_protected_nopar:Npx \fp_tmp:w
+        {
+          \group_end:
+          #1 \exp_not:N #2
+            {
+              \if_int_compare:w \l_fp_input_a_sign_int < \c_zero
+                -
+              \else:
+                +
+              \fi:
+              \int_use:N \l_fp_input_a_integer_int
+              .
+              \exp_after:wN \use_none:n
+                \int_use:N \l_fp_input_a_decimal_int
+              e
+              \int_use:N \l_fp_input_a_exponent_int
+            }
+        }
+    \fp_tmp:w
+  }
+\cs_new_protected_nopar:Npn \fp_round_places:Nn
+  { \fp_round_places_aux:NNn \tl_set:Nn }
+\cs_generate_variant:Nn \fp_round_places:Nn { c }
+\cs_new_protected_nopar:Npn \fp_ground_places:Nn
+  { \fp_round_places_aux:NNn \tl_gset:Nn }
+\cs_generate_variant:Nn \fp_ground_places:Nn { c }
+\cs_new_protected_nopar:Npn \fp_round_places_aux:NNn #1#2#3
+  {
+    \group_begin:
+      \fp_read:N #2
+      \int_set:Nn \l_fp_round_target_int
+        { #3 + \l_fp_input_a_exponent_int }
+      \if_int_compare:w \l_fp_round_target_int < \c_ten
+        \exp_after:wN \fp_round:
+      \fi:
+      \tex_advance:D \l_fp_input_a_decimal_int \c_one_thousand_million
+      \cs_set_protected_nopar:Npx \fp_tmp:w
+        {
+          \group_end:
+          #1 \exp_not:N #2
+            {
+              \if_int_compare:w \l_fp_input_a_sign_int < \c_zero
+                -
+              \else:
+                +
+              \fi:
+              \int_use:N \l_fp_input_a_integer_int
+              .
+              \exp_after:wN \use_none:n
+                \int_use:N \l_fp_input_a_decimal_int
+              e
+              \int_use:N \l_fp_input_a_exponent_int
+            }
+        }
+    \fp_tmp:w
+  }
+\cs_new_protected_nopar:Npn \fp_round:
+  {
+    \bool_set_false:N \l_fp_round_carry_bool
+    \l_fp_round_position_int \c_eight
+    \tl_clear:N \l_fp_round_decimal_tl
+    \tex_advance:D \l_fp_input_a_decimal_int \c_one_thousand_million
+    \exp_after:wN \use_i:nn \exp_after:wN
+      \fp_round_aux:NNNNNNNNN \int_use:N \l_fp_input_a_decimal_int
+  }
+\cs_new_protected_nopar:Npn \fp_round_aux:NNNNNNNNN #1#2#3#4#5#6#7#8#9
+  {
+    \fp_round_loop:N #9#8#7#6#5#4#3#2#1
+    \bool_if:NT \l_fp_round_carry_bool
+      { \tex_advance:D \l_fp_input_a_integer_int \c_one }
+    \l_fp_input_a_decimal_int \l_fp_round_decimal_tl \scan_stop:
+    \if_int_compare:w \l_fp_input_a_integer_int < \c_ten
+    \else:
+      \l_fp_input_a_integer_int \c_one
+      \tex_divide:D \l_fp_input_a_decimal_int \c_ten
+      \tex_advance:D \l_fp_input_a_exponent_int \c_one
+    \fi:
+  }
+\cs_new_protected_nopar:Npn \fp_round_loop:N #1
+  {
+    \if_int_compare:w \l_fp_round_position_int < \l_fp_round_target_int
+      \bool_if:NTF \l_fp_round_carry_bool
+        { \l_fp_tmp_int \int_eval:w #1 + \c_one \scan_stop: }
+        { \l_fp_tmp_int \int_eval:w #1 \scan_stop: }
+      \if_int_compare:w \l_fp_tmp_int = \c_ten
+        \l_fp_tmp_int \c_zero
+      \else:
+        \bool_set_false:N \l_fp_round_carry_bool
+      \fi:
+      \tl_set:Nx \l_fp_round_decimal_tl
+        { \int_use:N \l_fp_tmp_int \l_fp_round_decimal_tl }
+    \else:
+      \tl_set:Nx \l_fp_round_decimal_tl { 0 \l_fp_round_decimal_tl }
+      \if_int_compare:w \l_fp_round_position_int = \l_fp_round_target_int
+        \if_int_compare:w #1 > \c_four
+          \bool_set_true:N \l_fp_round_carry_bool
+        \fi:
+      \fi:
+    \fi:
+    \tex_advance:D \l_fp_round_position_int \c_minus_one
+    \if_int_compare:w \l_fp_round_position_int > \c_minus_one
+      \exp_after:wN \fp_round_loop:N
+    \fi:
+  }
+\cs_new_protected_nopar:Npn \fp_abs:N  { \fp_abs_aux:NN \tl_set:Nn }
+\cs_new_protected_nopar:Npn \fp_gabs:N { \fp_abs_aux:NN \tl_gset:Nn }
+\cs_generate_variant:Nn \fp_abs:N  { c }
+\cs_generate_variant:Nn \fp_gabs:N { c }
+\cs_new_protected_nopar:Npn \fp_abs_aux:NN #1#2
+  {
+    \group_begin:
+      \fp_read:N #2
+      \tex_advance:D \l_fp_input_a_decimal_int \c_one_thousand_million
+      \cs_set_protected_nopar:Npx \fp_tmp:w
+        {
+          \group_end:
+          #1 \exp_not:N #2
+            {
+              +
+              \int_use:N \l_fp_input_a_integer_int
+              .
+              \exp_after:wN \use_none:n
+                \int_use:N \l_fp_input_a_decimal_int
+              e
+              \int_use:N \l_fp_input_a_exponent_int
+            }
+        }
+    \fp_tmp:w
+  }
+\cs_new_protected_nopar:Npn \fp_neg:N  { \fp_neg_aux:NN \tl_set:Nn }
+\cs_new_protected_nopar:Npn \fp_gneg:N { \fp_neg_aux:NN \tl_gset:Nn }
+\cs_generate_variant:Nn \fp_neg:N  { c }
+\cs_generate_variant:Nn \fp_gneg:N { c }
+\cs_new_protected_nopar:Npn \fp_neg_aux:NN #1#2
+  {
+    \group_begin:
+      \fp_read:N #2
+      \tex_advance:D \l_fp_input_a_decimal_int \c_one_thousand_million
+      \tl_set:Nx \l_fp_tmp_tl
+        {
+          \if_int_compare:w \l_fp_input_a_sign_int < \c_zero
+            +
+          \else:
+            -
+          \fi:
+          \int_use:N \l_fp_input_a_integer_int
+          .
+          \exp_after:wN \use_none:n
+            \int_use:N \l_fp_input_a_decimal_int
+          e
+          \int_use:N \l_fp_input_a_exponent_int
+        }
+    \exp_after:wN \group_end: \exp_after:wN
+    #1 \exp_after:wN #2 \exp_after:wN { \l_fp_tmp_tl }
+  }
+\cs_new_protected_nopar:Npn \fp_add:Nn  { \fp_add_aux:NNn \tl_set:Nn }
+\cs_new_protected_nopar:Npn \fp_gadd:Nn { \fp_add_aux:NNn \tl_gset:Nn }
+\cs_generate_variant:Nn \fp_add:Nn   { c }
+\cs_generate_variant:Nn \fp_gadd:Nn  { c }
+\cs_new_protected_nopar:Npn \fp_add_aux:NNn #1#2#3
+  {
+    \group_begin:
+      \fp_read:N #2
+      \fp_split:Nn b {#3}
+      \fp_standardise:NNNN
+        \l_fp_input_b_sign_int
+        \l_fp_input_b_integer_int
+        \l_fp_input_b_decimal_int
+        \l_fp_input_b_exponent_int
+      \fp_add_core:
+    \fp_tmp:w #1#2
+  }
+\cs_new_protected_nopar:Npn \fp_add_core:
+  {
+    \fp_level_input_exponents:
+    \if_int_compare:w
+      \int_eval:w
+        \l_fp_input_a_sign_int * \l_fp_input_b_sign_int
+        > \c_zero
+      \exp_after:wN \fp_add_sum:
+    \else:
+      \exp_after:wN \fp_add_difference:
+    \fi:
+    \l_fp_output_exponent_int \l_fp_input_a_exponent_int
+    \fp_standardise:NNNN
+      \l_fp_output_sign_int
+      \l_fp_output_integer_int
+      \l_fp_output_decimal_int
+      \l_fp_output_exponent_int
+    \cs_set_protected_nopar:Npx \fp_tmp:w ##1##2
+      {
+        \group_end:
+        ##1 ##2
+          {
+             \if_int_compare:w \l_fp_output_sign_int < \c_zero
+              -
+            \else:
+              +
+            \fi:
+            \int_use:N \l_fp_output_integer_int
+            .
+            \exp_after:wN \use_none:n
+              \int_value:w \int_eval:w
+                 \l_fp_output_decimal_int + \c_one_thousand_million
+            e
+            \int_use:N \l_fp_output_exponent_int
+          }
+      }
+  }
+\cs_new_protected_nopar:Npn \fp_add_sum:
+  {
+    \l_fp_output_sign_int \l_fp_input_a_sign_int
+    \l_fp_output_integer_int
+      \int_eval:w
+        \l_fp_input_a_integer_int + \l_fp_input_b_integer_int
+      \scan_stop:
+    \l_fp_output_decimal_int
+      \int_eval:w
+        \l_fp_input_a_decimal_int + \l_fp_input_b_decimal_int
+      \scan_stop:
+    \if_int_compare:w \l_fp_output_decimal_int < \c_one_thousand_million
+    \else:
+      \tex_advance:D \l_fp_output_integer_int \c_one
+      \tex_advance:D \l_fp_output_decimal_int -\c_one_thousand_million
+    \fi:
+  }
+\cs_new_protected_nopar:Npn \fp_add_difference:
+  {
+    \l_fp_output_integer_int
+      \int_eval:w
+        \l_fp_input_a_integer_int - \l_fp_input_b_integer_int
+      \scan_stop:
+    \l_fp_output_decimal_int
+      \int_eval:w
+        \l_fp_input_a_decimal_int - \l_fp_input_b_decimal_int
+      \scan_stop:
+    \if_int_compare:w \l_fp_output_decimal_int < \c_zero
+      \tex_advance:D \l_fp_output_integer_int \c_minus_one
+      \tex_advance:D \l_fp_output_decimal_int \c_one_thousand_million
+    \fi:
+    \if_int_compare:w \l_fp_output_integer_int < \c_zero
+      \l_fp_output_sign_int \l_fp_input_b_sign_int
+      \if_int_compare:w \l_fp_output_decimal_int = \c_zero
+        \l_fp_output_integer_int -\l_fp_output_integer_int
+      \else:
+        \l_fp_output_decimal_int
+          \int_eval:w
+            \c_one_thousand_million - \l_fp_output_decimal_int
+          \scan_stop:
+        \l_fp_output_integer_int
+           \int_eval:w
+             - \l_fp_output_integer_int - \c_one
+           \scan_stop:
+      \fi:
+    \else:
+      \l_fp_output_sign_int \l_fp_input_a_sign_int
+    \fi:
+  }
+\cs_new_protected_nopar:Npn \fp_sub:Nn  { \fp_sub_aux:NNn \tl_set:Nn }
+\cs_new_protected_nopar:Npn \fp_gsub:Nn { \fp_sub_aux:NNn \tl_gset:Nn }
+\cs_generate_variant:Nn \fp_sub:Nn   { c }
+\cs_generate_variant:Nn \fp_gsub:Nn  { c }
+\cs_new_protected_nopar:Npn \fp_sub_aux:NNn #1#2#3
+  {
+    \group_begin:
+      \fp_read:N #2
+      \fp_split:Nn b {#3}
+      \fp_standardise:NNNN
+        \l_fp_input_b_sign_int
+        \l_fp_input_b_integer_int
+        \l_fp_input_b_decimal_int
+        \l_fp_input_b_exponent_int
+      \tex_multiply:D \l_fp_input_b_sign_int \c_minus_one
+      \fp_add_core:
+    \fp_tmp:w #1#2
+  }
+\cs_new_protected_nopar:Npn \fp_mul:Nn  { \fp_mul_aux:NNn \tl_set:Nn }
+\cs_new_protected_nopar:Npn \fp_gmul:Nn { \fp_mul_aux:NNn \tl_gset:Nn }
+\cs_generate_variant:Nn \fp_mul:Nn  { c }
+\cs_generate_variant:Nn \fp_gmul:Nn { c }
+\cs_new_protected_nopar:Npn \fp_mul_aux:NNn #1#2#3
+  {
+    \group_begin:
+      \fp_read:N #2
+      \fp_split:Nn b {#3}
+      \fp_standardise:NNNN
+        \l_fp_input_b_sign_int
+        \l_fp_input_b_integer_int
+        \l_fp_input_b_decimal_int
+        \l_fp_input_b_exponent_int
+      \fp_mul_internal:
+      \l_fp_output_exponent_int
+        \int_eval:w
+          \l_fp_input_a_exponent_int + \l_fp_input_b_exponent_int
+        \scan_stop:
+      \fp_standardise:NNNN
+        \l_fp_output_sign_int
+        \l_fp_output_integer_int
+        \l_fp_output_decimal_int
+        \l_fp_output_exponent_int
+      \cs_set_protected_nopar:Npx \fp_tmp:w
+        {
+          \group_end:
+          #1 \exp_not:N #2
+            {
+              \if_int_compare:w
+                \int_eval:w
+                  \l_fp_input_a_sign_int * \l_fp_input_b_sign_int
+                  < \c_zero
+                \if_int_compare:w
+                  \int_eval:w
+                    \l_fp_output_integer_int + \l_fp_output_decimal_int
+                    = \c_zero
+                  +
+                \else:
+                  -
+                \fi:
+              \else:
+                +
+             \fi:
+              \int_use:N \l_fp_output_integer_int
+              .
+              \exp_after:wN \use_none:n
+                \int_value:w \int_eval:w
+                   \l_fp_output_decimal_int + \c_one_thousand_million
+              e
+              \int_use:N \l_fp_output_exponent_int
+            }
+        }
+    \fp_tmp:w
+  }
+\cs_new_protected_nopar:Npn \fp_mul_internal:
+  {
+    \fp_mul_split:NNNN \l_fp_input_a_decimal_int
+      \l_fp_mul_a_i_int \l_fp_mul_a_ii_int \l_fp_mul_a_iii_int
+    \fp_mul_split:NNNN \l_fp_input_b_decimal_int
+      \l_fp_mul_b_i_int \l_fp_mul_b_ii_int \l_fp_mul_b_iii_int
+    \l_fp_mul_output_int \c_zero
+    \tl_clear:N \l_fp_mul_output_tl
+    \fp_mul_product:NN \l_fp_mul_a_i_int         \l_fp_mul_b_iii_int
+    \fp_mul_product:NN \l_fp_mul_a_ii_int        \l_fp_mul_b_ii_int
+    \fp_mul_product:NN \l_fp_mul_a_iii_int       \l_fp_mul_b_i_int
+    \tex_divide:D \l_fp_mul_output_int \c_one_thousand
+    \fp_mul_product:NN \l_fp_input_a_integer_int \l_fp_mul_b_iii_int
+    \fp_mul_product:NN \l_fp_mul_a_i_int         \l_fp_mul_b_ii_int
+    \fp_mul_product:NN \l_fp_mul_a_ii_int        \l_fp_mul_b_i_int
+    \fp_mul_product:NN \l_fp_mul_a_iii_int       \l_fp_input_b_integer_int
+    \fp_mul_end_level:
+    \fp_mul_product:NN \l_fp_input_a_integer_int \l_fp_mul_b_ii_int
+    \fp_mul_product:NN \l_fp_mul_a_i_int         \l_fp_mul_b_i_int
+    \fp_mul_product:NN \l_fp_mul_a_ii_int        \l_fp_input_b_integer_int
+    \fp_mul_end_level:
+    \fp_mul_product:NN \l_fp_input_a_integer_int \l_fp_mul_b_i_int
+    \fp_mul_product:NN \l_fp_mul_a_i_int         \l_fp_input_b_integer_int
+    \fp_mul_end_level:
+    \l_fp_output_decimal_int 0 \l_fp_mul_output_tl \scan_stop:
+    \tl_clear:N \l_fp_mul_output_tl
+    \fp_mul_product:NN \l_fp_input_a_integer_int \l_fp_input_b_integer_int
+    \fp_mul_end_level:
+    \l_fp_output_integer_int 0 \l_fp_mul_output_tl \scan_stop:
+  }
+\cs_new_protected_nopar:Npn \fp_mul_split:NNNN #1#2#3#4
+  {
+    \tex_advance:D #1 \c_one_thousand_million
+    \cs_set_protected_nopar:Npn \fp_mul_split_aux:w
+       ##1##2##3##4##5##6##7##8##9 \q_stop {
+        #2 ##2##3##4 \scan_stop:
+        #3 ##5##6##7 \scan_stop:
+        #4 ##8##9    \scan_stop:
+      }
+    \exp_after:wN \fp_mul_split_aux:w \int_use:N #1 \q_stop
+    \tex_advance:D #1 -\c_one_thousand_million
+  }
+\cs_new_protected_nopar:Npn \fp_mul_product:NN #1#2
+  {
+    \l_fp_mul_output_int
+      \int_eval:w \l_fp_mul_output_int + #1 * #2 \scan_stop:
+  }
+\cs_new_protected_nopar:Npn \fp_mul_end_level:
+  {
+    \tex_advance:D \l_fp_mul_output_int \c_one_thousand_million
+    \exp_after:wN \use_i:nn \exp_after:wN
+      \fp_mul_end_level:NNNNNNNNN \int_use:N \l_fp_mul_output_int
+  }
+\cs_new_protected_nopar:Npn \fp_mul_end_level:NNNNNNNNN #1#2#3#4#5#6#7#8#9
+  {
+    \tl_set:Nx \l_fp_mul_output_tl { #7#8#9 \l_fp_mul_output_tl }
+    \l_fp_mul_output_int #1#2#3#4#5#6 \scan_stop:
+  }
+\cs_new_protected_nopar:Npn \fp_div:Nn  { \fp_div_aux:NNn \tl_set:Nn }
+\cs_new_protected_nopar:Npn \fp_gdiv:Nn { \fp_div_aux:NNn \tl_gset:Nn }
+\cs_generate_variant:Nn \fp_div:Nn  { c }
+\cs_generate_variant:Nn \fp_gdiv:Nn { c }
+\cs_new_protected_nopar:Npn \fp_div_aux:NNn #1#2#3
+  {
+    \group_begin:
+      \fp_read:N #2
+      \fp_split:Nn b {#3}
+      \fp_standardise:NNNN
+        \l_fp_input_b_sign_int
+        \l_fp_input_b_integer_int
+        \l_fp_input_b_decimal_int
+        \l_fp_input_b_exponent_int
+      \if_int_compare:w
+        \int_eval:w
+          \l_fp_input_b_integer_int + \l_fp_input_b_decimal_int
+          = \c_zero
+        \cs_set_protected_nopar:Npx \fp_tmp:w ##1##2
+          {
+            \group_end:
+            #1 \exp_not:N #2 { \c_undefined_fp }
+          }
+      \else:
+        \if_int_compare:w
+          \int_eval:w
+            \l_fp_input_a_integer_int + \l_fp_input_a_decimal_int
+            = \c_zero
+          \cs_set_protected_nopar:Npx \fp_tmp:w ##1##2
+            {
+              \group_end:
+              #1 \exp_not:N #2 { \c_zero_fp }
+            }
+        \else:
+          \exp_after:wN \exp_after:wN \exp_after:wN \fp_div_internal:
+        \fi:
+      \fi:
+    \fp_tmp:w #1#2
+  }
+\cs_new_protected_nopar:Npn \fp_div_internal: {
+  \l_fp_output_integer_int \c_zero
+  \l_fp_output_decimal_int \c_zero
+  \cs_set_eq:NN \fp_div_store: \fp_div_store_integer:
+  \l_fp_div_offset_int \c_one_hundred_million
+  \fp_div_loop:
+  \l_fp_output_exponent_int
+    \int_eval:w
+      \l_fp_input_a_exponent_int - \l_fp_input_b_exponent_int
+    \scan_stop:
+  \fp_standardise:NNNN
+    \l_fp_output_sign_int
+    \l_fp_output_integer_int
+    \l_fp_output_decimal_int
+    \l_fp_output_exponent_int
+  \cs_set_protected_nopar:Npx \fp_tmp:w ##1##2
+    {
+      \group_end:
+      ##1 ##2
+        {
+          \if_int_compare:w
+            \int_eval:w
+              \l_fp_input_a_sign_int * \l_fp_input_b_sign_int
+              < \c_zero
+            \if_int_compare:w
+              \int_eval:w
+                \l_fp_output_integer_int + \l_fp_output_decimal_int
+                = \c_zero
+              +
+            \else:
+              -
+            \fi:
+          \else:
+            +
+          \fi:
+          \int_use:N \l_fp_output_integer_int
+          .
+          \exp_after:wN \use_none:n
+            \int_value:w \int_eval:w
+               \l_fp_output_decimal_int + \c_one_thousand_million
+             \int_eval_end:
+          e
+          \int_use:N \l_fp_output_exponent_int
+        }
+    }
+}
+\cs_new_protected_nopar:Npn \fp_div_loop:
+  {
+    \l_fp_count_int \c_zero
+    \fp_div_divide:
+    \fp_div_store:
+    \tex_multiply:D \l_fp_input_a_integer_int \c_ten
+    \tex_advance:D \l_fp_input_a_decimal_int \c_one_thousand_million
+    \exp_after:wN \fp_div_loop_step:w
+      \int_use:N \l_fp_input_a_decimal_int \q_stop
+    \if_int_compare:w
+      \int_eval:w \l_fp_input_a_integer_int + \l_fp_input_a_decimal_int
+        > \c_zero
+        \if_int_compare:w \l_fp_div_offset_int > \c_zero
+          \exp_after:wN \exp_after:wN \exp_after:wN
+            \fp_div_loop:
+        \fi:
+    \fi:
+  }
+\cs_new_protected_nopar:Npn \fp_div_divide:
+  {
+    \if_int_compare:w \l_fp_input_a_integer_int > \l_fp_input_b_integer_int
+      \exp_after:wN \fp_div_divide_aux:
+    \else:
+      \if_int_compare:w \l_fp_input_a_integer_int < \l_fp_input_b_integer_int
+      \else:
+        \if_int_compare:w
+          \l_fp_input_a_decimal_int < \l_fp_input_b_decimal_int
+        \else:
+          \exp_after:wN \exp_after:wN \exp_after:wN
+            \exp_after:wN \exp_after:wN \exp_after:wN
+            \exp_after:wN \fp_div_divide_aux:
+        \fi:
+      \fi:
+    \fi:
+  }
+\cs_new_protected_nopar:Npn \fp_div_divide_aux:
+  {
+    \tex_advance:D \l_fp_count_int \c_one
+    \tex_advance:D \l_fp_input_a_integer_int -\l_fp_input_b_integer_int
+    \tex_advance:D \l_fp_input_a_decimal_int -\l_fp_input_b_decimal_int
+    \if_int_compare:w \l_fp_input_a_decimal_int < \c_zero
+      \tex_advance:D \l_fp_input_a_integer_int \c_minus_one
+      \tex_advance:D \l_fp_input_a_decimal_int \c_one_thousand_million
+    \fi:
+    \fp_div_divide:
+  }
+\cs_new_protected_nopar:Npn \fp_div_store: { }
+\cs_new_protected_nopar:Npn \fp_div_store_integer:
+  {
+    \l_fp_output_integer_int \l_fp_count_int
+    \cs_set_eq:NN \fp_div_store: \fp_div_store_decimal:
+  }
+\cs_new_protected_nopar:Npn \fp_div_store_decimal:
+  {
+    \l_fp_output_decimal_int
+      \int_eval:w
+        \l_fp_output_decimal_int +
+        \l_fp_count_int * \l_fp_div_offset_int
+      \int_eval_end:
+    \tex_divide:D \l_fp_div_offset_int \c_ten
+  }
+\cs_new_protected_nopar:Npn \fp_div_loop_step:w #1#2#3#4#5#6#7#8#9 \q_stop
+  {
+    \l_fp_input_a_integer_int
+      \int_eval:w #2 + \l_fp_input_a_integer_int  \int_eval_end:
+    \l_fp_input_a_decimal_int #3#4#5#6#7#8#9 0 \scan_stop:
+  }
+\cs_new_protected_nopar:Npn \fp_add:NNNNNNNNN #1#2#3#4#5#6#7#8#9
+  {
+    #7 \int_eval:w #1 + #4 \int_eval_end:
+    #8 \int_eval:w #2 + #5 \int_eval_end:
+    #9 \int_eval:w #3 + #6 \int_eval_end:
+    \if_int_compare:w #9 < \c_one_thousand_million
+    \else:
+      \tex_advance:D #8 \c_one
+      \tex_advance:D #9 -\c_one_thousand_million
+    \fi:
+    \if_int_compare:w #8 < \c_one_thousand_million
+    \else:
+      \tex_advance:D #7 \c_one
+      \tex_advance:D #8 -\c_one_thousand_million
+    \fi:
+  }
+\cs_new_protected_nopar:Npn \fp_sub:NNNNNNNNN #1#2#3#4#5#6#7#8#9
+  {
+    #7 \int_eval:w #1 - #4 \int_eval_end:
+    #8 \int_eval:w #2 - #5 \int_eval_end:
+    #9 \int_eval:w #3 - #6 \int_eval_end:
+    \if_int_compare:w #9 < \c_zero
+      \tex_advance:D #8 \c_minus_one
+      \tex_advance:D #9 \c_one_thousand_million
+    \fi:
+    \if_int_compare:w #8 < \c_zero
+      \tex_advance:D #7 \c_minus_one
+      \tex_advance:D #8 \c_one_thousand_million
+    \fi:
+    \if_int_compare:w #7 < \c_zero
+      \if_int_compare:w \int_eval:w #8 + #9 = \c_zero
+        #7 -#7
+      \else:
+        \tex_advance:D #7 \c_one
+        #8 \int_eval:w \c_one_thousand_million - #8 \int_eval_end:
+        #9 \int_eval:w \c_one_thousand_million - #9 \int_eval_end:
+      \fi:
+    \fi:
+  }
+\cs_new_protected_nopar:Npn \fp_mul:NNNNNN #1#2#3#4#5#6
+  {
+    \fp_mul_split:NNNN #1
+      \l_fp_mul_a_i_int \l_fp_mul_a_ii_int \l_fp_mul_a_iii_int
+    \fp_mul_split:NNNN #2
+      \l_fp_mul_a_iv_int \l_fp_mul_a_v_int \l_fp_mul_a_vi_int
+    \fp_mul_split:NNNN #3
+      \l_fp_mul_b_i_int \l_fp_mul_b_ii_int \l_fp_mul_b_iii_int
+    \fp_mul_split:NNNN #4
+      \l_fp_mul_b_iv_int \l_fp_mul_b_v_int \l_fp_mul_b_vi_int
+    \l_fp_mul_output_int \c_zero
+    \tl_clear:N \l_fp_mul_output_tl
+    \fp_mul_product:NN \l_fp_mul_a_i_int         \l_fp_mul_b_vi_int
+    \fp_mul_product:NN \l_fp_mul_a_ii_int        \l_fp_mul_b_v_int
+    \fp_mul_product:NN \l_fp_mul_a_iii_int       \l_fp_mul_b_iv_int
+    \fp_mul_product:NN \l_fp_mul_a_iv_int        \l_fp_mul_b_iii_int
+    \fp_mul_product:NN \l_fp_mul_a_v_int         \l_fp_mul_b_ii_int
+    \fp_mul_product:NN \l_fp_mul_a_vi_int        \l_fp_mul_b_i_int
+    \tex_divide:D \l_fp_mul_output_int \c_one_thousand
+    \fp_mul_product:NN \l_fp_mul_a_i_int        \l_fp_mul_b_v_int
+    \fp_mul_product:NN \l_fp_mul_a_ii_int       \l_fp_mul_b_iv_int
+    \fp_mul_product:NN \l_fp_mul_a_iii_int      \l_fp_mul_b_iii_int
+    \fp_mul_product:NN \l_fp_mul_a_iv_int       \l_fp_mul_b_ii_int
+    \fp_mul_product:NN \l_fp_mul_a_v_int        \l_fp_mul_b_i_int
+    \fp_mul_end_level:
+    \fp_mul_product:NN \l_fp_mul_a_i_int        \l_fp_mul_b_iv_int
+    \fp_mul_product:NN \l_fp_mul_a_ii_int       \l_fp_mul_b_iii_int
+    \fp_mul_product:NN \l_fp_mul_a_iii_int      \l_fp_mul_b_ii_int
+    \fp_mul_product:NN \l_fp_mul_a_iv_int       \l_fp_mul_b_i_int
+    \fp_mul_end_level:
+    \fp_mul_product:NN \l_fp_mul_a_i_int        \l_fp_mul_b_iii_int
+    \fp_mul_product:NN \l_fp_mul_a_ii_int       \l_fp_mul_b_ii_int
+    \fp_mul_product:NN \l_fp_mul_a_iii_int      \l_fp_mul_b_i_int
+    \fp_mul_end_level:
+    #6 0 \l_fp_mul_output_tl \scan_stop:
+    \tl_clear:N \l_fp_mul_output_tl
+    \fp_mul_product:NN \l_fp_mul_a_i_int        \l_fp_mul_b_ii_int
+    \fp_mul_product:NN \l_fp_mul_a_ii_int       \l_fp_mul_b_i_int
+    \fp_mul_end_level:
+    \fp_mul_product:NN \l_fp_mul_a_i_int        \l_fp_mul_b_i_int
+    \fp_mul_end_level:
+    \fp_mul_end_level:
+    #5 0 \l_fp_mul_output_tl \scan_stop:
+  }
+\cs_new_protected_nopar:Npn \fp_mul:NNNNNNNNN #1#2#3#4#5#6#7#8#9
+  {
+    \fp_mul_split:NNNN #2
+      \l_fp_mul_a_i_int \l_fp_mul_a_ii_int \l_fp_mul_a_iii_int
+    \fp_mul_split:NNNN #3
+      \l_fp_mul_a_iv_int \l_fp_mul_a_v_int \l_fp_mul_a_vi_int
+    \fp_mul_split:NNNN #5
+      \l_fp_mul_b_i_int \l_fp_mul_b_ii_int \l_fp_mul_b_iii_int
+    \fp_mul_split:NNNN #6
+      \l_fp_mul_b_iv_int \l_fp_mul_b_v_int \l_fp_mul_b_vi_int
+    \l_fp_mul_output_int \c_zero
+    \tl_clear:N \l_fp_mul_output_tl
+    \fp_mul_product:NN \l_fp_mul_a_i_int         \l_fp_mul_b_vi_int
+    \fp_mul_product:NN \l_fp_mul_a_ii_int        \l_fp_mul_b_v_int
+    \fp_mul_product:NN \l_fp_mul_a_iii_int       \l_fp_mul_b_iv_int
+    \fp_mul_product:NN \l_fp_mul_a_iv_int        \l_fp_mul_b_iii_int
+    \fp_mul_product:NN \l_fp_mul_a_v_int         \l_fp_mul_b_ii_int
+    \fp_mul_product:NN \l_fp_mul_a_vi_int        \l_fp_mul_b_i_int
+    \tex_divide:D \l_fp_mul_output_int \c_one_thousand
+    \fp_mul_product:NN #1                       \l_fp_mul_b_vi_int
+    \fp_mul_product:NN \l_fp_mul_a_i_int        \l_fp_mul_b_v_int
+    \fp_mul_product:NN \l_fp_mul_a_ii_int       \l_fp_mul_b_iv_int
+    \fp_mul_product:NN \l_fp_mul_a_iii_int      \l_fp_mul_b_iii_int
+    \fp_mul_product:NN \l_fp_mul_a_iv_int       \l_fp_mul_b_ii_int
+    \fp_mul_product:NN \l_fp_mul_a_v_int        \l_fp_mul_b_i_int
+    \fp_mul_product:NN \l_fp_mul_a_vi_int       #4
+    \fp_mul_end_level:
+    \fp_mul_product:NN #1                       \l_fp_mul_b_v_int
+    \fp_mul_product:NN \l_fp_mul_a_i_int        \l_fp_mul_b_iv_int
+    \fp_mul_product:NN \l_fp_mul_a_ii_int       \l_fp_mul_b_iii_int
+    \fp_mul_product:NN \l_fp_mul_a_iii_int      \l_fp_mul_b_ii_int
+    \fp_mul_product:NN \l_fp_mul_a_iv_int       \l_fp_mul_b_i_int
+    \fp_mul_product:NN \l_fp_mul_a_v_int        #4
+    \fp_mul_end_level:
+    \fp_mul_product:NN #1                       \l_fp_mul_b_iv_int
+    \fp_mul_product:NN \l_fp_mul_a_i_int        \l_fp_mul_b_iii_int
+    \fp_mul_product:NN \l_fp_mul_a_ii_int       \l_fp_mul_b_ii_int
+    \fp_mul_product:NN \l_fp_mul_a_iii_int      \l_fp_mul_b_i_int
+    \fp_mul_product:NN \l_fp_mul_a_iv_int       #4
+    \fp_mul_end_level:
+    #9 0 \l_fp_mul_output_tl \scan_stop:
+    \tl_clear:N \l_fp_mul_output_tl
+    \fp_mul_product:NN #1                       \l_fp_mul_b_iii_int
+    \fp_mul_product:NN \l_fp_mul_a_i_int        \l_fp_mul_b_ii_int
+    \fp_mul_product:NN \l_fp_mul_a_ii_int       \l_fp_mul_b_i_int
+    \fp_mul_product:NN \l_fp_mul_a_iii_int      #4
+    \fp_mul_end_level:
+    \fp_mul_product:NN #1                       \l_fp_mul_b_ii_int
+    \fp_mul_product:NN \l_fp_mul_a_i_int        \l_fp_mul_b_i_int
+    \fp_mul_product:NN \l_fp_mul_a_ii_int       #4
+    \fp_mul_end_level:
+    \fp_mul_product:NN #1                       \l_fp_mul_b_i_int
+    \fp_mul_product:NN \l_fp_mul_a_i_int        #4
+    \fp_mul_end_level:
+    #8 0 \l_fp_mul_output_tl \scan_stop:
+    \tl_clear:N \l_fp_mul_output_tl
+    \fp_mul_product:NN #1 #4
+    \fp_mul_end_level:
+    #7 0 \l_fp_mul_output_tl \scan_stop:
+  }
+\cs_new_protected_nopar:Npn \fp_div_integer:NNNNN #1#2#3#4#5
+  {
+    \l_fp_tmp_int #1
+    \tex_divide:D \l_fp_tmp_int #3
+    \l_fp_tmp_int \int_eval:w #1 - \l_fp_tmp_int * #3 \int_eval_end:
+    #4 #1
+    \tex_divide:D #4 #3
+    #5 #2
+    \tex_divide:D #5 #3
+    \tex_multiply:D \l_fp_tmp_int \c_one_thousand
+    \tex_divide:D \l_fp_tmp_int #3
+    #5 \int_eval:w #5 + \l_fp_tmp_int * \c_one_million \int_eval_end:
+    \if_int_compare:w #5 > \c_one_thousand_million
+      \tex_advance:D #4 \c_one
+      \tex_advance:D #5 -\c_one_thousand_million
+    \fi:
+  }
+\cs_new_protected_nopar:Npn \fp_extended_normalise:
+  {
+    \fp_extended_normalise_aux_i:
+    \fp_extended_normalise_aux_ii:
+  }
+\cs_new_protected_nopar:Npn \fp_extended_normalise_aux_i:
+  {
+    \if_int_compare:w \l_fp_input_a_exponent_int > \c_zero
+      \tex_multiply:D \l_fp_input_a_integer_int \c_ten
+      \tex_advance:D \l_fp_input_a_decimal_int \c_one_thousand_million
+      \exp_after:wN \fp_extended_normalise_aux_i:w
+        \int_use:N \l_fp_input_a_decimal_int \q_stop
+       \exp_after:wN \fp_extended_normalise_aux_i:
+     \fi:
+  }
+\cs_new_protected_nopar:Npn \fp_extended_normalise_aux_i:w
+  #1#2#3#4#5#6#7#8#9 \q_stop
+  {
+    \l_fp_input_a_integer_int
+      \int_eval:w \l_fp_input_a_integer_int + #2 \scan_stop:
+    \l_fp_input_a_decimal_int #3#4#5#6#7#8#9 0 \scan_stop:
+    \tex_advance:D \l_fp_input_a_extended_int \c_one_thousand_million
+    \exp_after:wN \fp_extended_normalise_aux_ii:w
+      \int_use:N \l_fp_input_a_extended_int \q_stop
+  }
+\cs_new_protected_nopar:Npn \fp_extended_normalise_aux_ii:w
+  #1#2#3#4#5#6#7#8#9 \q_stop
+  {
+    \l_fp_input_a_decimal_int
+      \int_eval:w \l_fp_input_a_decimal_int + #2 \scan_stop:
+    \l_fp_input_a_extended_int #3#4#5#6#7#8#9 0 \scan_stop:
+    \tex_advance:D \l_fp_input_a_exponent_int \c_minus_one
+  }
+\cs_new_protected_nopar:Npn \fp_extended_normalise_aux_ii:
+  {
+    \if_int_compare:w \l_fp_input_a_exponent_int < \c_zero
+      \tex_advance:D \l_fp_input_a_decimal_int \c_one_thousand_million
+      \exp_after:wN \use_i:nn \exp_after:wN
+        \fp_extended_normalise_ii_aux:NNNNNNNNN
+        \int_use:N \l_fp_input_a_decimal_int
+       \exp_after:wN \fp_extended_normalise_aux_ii:
+     \fi:
+  }
+\cs_new_protected_nopar:Npn \fp_extended_normalise_ii_aux:NNNNNNNNN
+  #1#2#3#4#5#6#7#8#9
+  {
+    \if_int_compare:w \l_fp_input_a_integer_int = \c_zero
+      \l_fp_input_a_decimal_int #1#2#3#4#5#6#7#8 \scan_stop:
+    \else:
+      \tl_set:Nx \l_fp_tmp_tl
+        {
+          \int_use:N \l_fp_input_a_integer_int
+          #1#2#3#4#5#6#7#8
+        }
+      \l_fp_input_a_integer_int \c_zero
+      \l_fp_input_a_decimal_int \l_fp_tmp_tl \scan_stop:
+    \fi:
+    \tex_divide:D \l_fp_input_a_extended_int \c_ten
+    \tl_set:Nx \l_fp_tmp_tl
+      {
+        #9
+        \int_use:N \l_fp_input_a_extended_int
+      }
+    \l_fp_input_a_extended_int \l_fp_tmp_tl \scan_stop:
+    \tex_advance:D \l_fp_input_a_exponent_int \c_one
+  }
+\cs_new_protected_nopar:Npn \fp_extended_normalise_output:
+  {
+    \if_int_compare:w \l_fp_output_integer_int > \c_nine
+      \tex_advance:D \l_fp_output_integer_int \c_one_thousand_million
+      \exp_after:wN \use_i:nn \exp_after:wN
+        \fp_extended_normalise_output_aux_i:NNNNNNNNN
+        \int_use:N \l_fp_output_integer_int
+      \exp_after:wN \fp_extended_normalise_output:
+    \fi:
+  }
+\cs_new_protected_nopar:Npn \fp_extended_normalise_output_aux_i:NNNNNNNNN
+  #1#2#3#4#5#6#7#8#9
+  {
+    \l_fp_output_integer_int #1#2#3#4#5#6#7#8 \scan_stop:
+    \tex_advance:D \l_fp_output_decimal_int \c_one_thousand_million
+    \tl_set:Nx \l_fp_tmp_tl
+      {
+        #9
+        \exp_after:wN \use_none:n
+        \int_use:N \l_fp_output_decimal_int
+      }
+    \exp_after:wN \fp_extended_normalise_output_aux_ii:NNNNNNNNN
+      \l_fp_tmp_tl
+  }
+\cs_new_protected_nopar:Npn \fp_extended_normalise_output_aux_ii:NNNNNNNNN
+  #1#2#3#4#5#6#7#8#9
+  {
+    \l_fp_output_decimal_int #1#2#3#4#5#6#7#8#9 \scan_stop:
+    \fp_extended_normalise_output_aux:N
+  }
+\cs_new_protected_nopar:Npn \fp_extended_normalise_output_aux:N #1
+  {
+    \tex_advance:D \l_fp_output_extended_int \c_one_thousand_million
+    \tex_divide:D \l_fp_output_extended_int \c_ten
+    \tl_set:Nx \l_fp_tmp_tl
+      {
+        #1
+        \exp_after:wN \use_none:n
+          \int_use:N \l_fp_output_extended_int
+      }
+    \l_fp_output_extended_int \l_fp_tmp_tl \scan_stop:
+    \tex_advance:D \l_fp_output_exponent_int \c_one
+  }
+\cs_new_protected_nopar:Npn \fp_trig_normalise:
+  {
+    \if_int_compare:w \l_fp_input_a_exponent_int < \c_ten
+      \l_fp_input_a_extended_int \c_zero
+      \fp_extended_normalise:
+      \fp_trig_normalise_aux:
+      \if_int_compare:w \l_fp_input_a_integer_int < \c_zero
+        \l_fp_input_a_sign_int -\l_fp_input_a_sign_int
+        \l_fp_input_a_integer_int -\l_fp_input_a_integer_int
+      \fi:
+       \exp_after:wN \fp_trig_octant:
+    \else:
+      \l_fp_input_a_sign_int    \c_one
+      \l_fp_output_integer_int  \c_zero
+      \l_fp_output_decimal_int  \c_zero
+      \l_fp_output_exponent_int \c_zero
+      \exp_after:wN \fp_trig_overflow_msg:
+    \fi:
+  }
+\cs_new_protected_nopar:Npn \fp_trig_normalise_aux:
+  {
+    \if_int_compare:w \l_fp_input_a_integer_int > \c_three
+      \fp_trig_sub:NNN
+        \c_six \c_fp_two_pi_decimal_int \c_fp_two_pi_extended_int
+      \exp_after:wN \fp_trig_normalise_aux:
+    \else:
+      \if_int_compare:w \l_fp_input_a_integer_int > \c_two
+        \if_int_compare:w \l_fp_input_a_decimal_int > \c_fp_pi_decimal_int
+          \fp_trig_sub:NNN
+            \c_six \c_fp_two_pi_decimal_int \c_fp_two_pi_extended_int
+          \exp_after:wN \exp_after:wN \exp_after:wN
+          \exp_after:wN \exp_after:wN \exp_after:wN
+            \exp_after:wN \fp_trig_normalise_aux:
+          \fi:
+      \fi:
+    \fi:
+  }
+\cs_new_protected_nopar:Npn \fp_trig_sub:NNN #1#2#3
+  {
+    \l_fp_input_a_integer_int
+      \int_eval:w \l_fp_input_a_integer_int - #1 \int_eval_end:
+    \l_fp_input_a_decimal_int
+      \int_eval:w \l_fp_input_a_decimal_int - #2 \int_eval_end:
+    \l_fp_input_a_extended_int
+      \int_eval:w \l_fp_input_a_extended_int - #3 \int_eval_end:
+    \if_int_compare:w \l_fp_input_a_extended_int < \c_zero
+      \tex_advance:D \l_fp_input_a_decimal_int \c_minus_one
+      \tex_advance:D \l_fp_input_a_extended_int \c_one_thousand_million
+    \fi:
+    \if_int_compare:w \l_fp_input_a_decimal_int < \c_zero
+      \tex_advance:D \l_fp_input_a_integer_int \c_minus_one
+      \tex_advance:D \l_fp_input_a_decimal_int \c_one_thousand_million
+    \fi:
+    \if_int_compare:w \l_fp_input_a_integer_int < \c_zero
+      \l_fp_input_a_sign_int -\l_fp_input_a_sign_int
+      \if_int_compare:w
+        \int_eval:w
+          \l_fp_input_a_decimal_int + \l_fp_input_a_extended_int
+        = \c_zero
+        \l_fp_input_a_integer_int -\l_fp_input_a_integer_int
+      \else:
+        \l_fp_input_a_integer_int
+           \int_eval:w
+             - \l_fp_input_a_integer_int - \c_one
+           \int_eval_end:
+        \l_fp_input_a_decimal_int
+          \int_eval:w
+            \c_one_thousand_million - \l_fp_input_a_decimal_int
+          \int_eval_end:
+        \l_fp_input_a_extended_int
+          \int_eval:w
+            \c_one_thousand_million - \l_fp_input_a_extended_int
+          \int_eval_end:
+      \fi:
+    \fi:
+  }
+\cs_new_protected_nopar:Npn \fp_trig_octant:
+  {
+    \l_fp_trig_octant_int \c_one
+    \fp_trig_octant_aux:
+    \if_int_compare:w \l_fp_input_a_decimal_int < \c_ten
+      \l_fp_input_a_decimal_int  \c_zero
+      \l_fp_input_a_extended_int \c_zero
+    \fi:
+    \if_int_odd:w \l_fp_trig_octant_int
+    \else:
+      \fp_sub:NNNNNNNNN
+        \c_zero \c_fp_pi_by_four_decimal_int \c_fp_pi_by_four_extended_int
+        \l_fp_input_a_integer_int \l_fp_input_a_decimal_int
+          \l_fp_input_a_extended_int
+        \l_fp_input_a_integer_int \l_fp_input_a_decimal_int
+          \l_fp_input_a_extended_int
+    \fi:
+  }
+\cs_new_protected_nopar:Npn \fp_trig_octant_aux:
+  {
+    \if_int_compare:w \l_fp_input_a_integer_int > \c_zero
+      \fp_sub:NNNNNNNNN
+        \l_fp_input_a_integer_int \l_fp_input_a_decimal_int
+          \l_fp_input_a_extended_int
+        \c_zero \c_fp_pi_by_four_decimal_int \c_fp_pi_by_four_extended_int
+        \l_fp_input_a_integer_int \l_fp_input_a_decimal_int
+          \l_fp_input_a_extended_int
+      \tex_advance:D \l_fp_trig_octant_int \c_one
+      \exp_after:wN \fp_trig_octant_aux:
+    \else:
+      \if_int_compare:w
+        \l_fp_input_a_decimal_int > \c_fp_pi_by_four_decimal_int
+        \fp_sub:NNNNNNNNN
+          \l_fp_input_a_integer_int \l_fp_input_a_decimal_int
+            \l_fp_input_a_extended_int
+          \c_zero \c_fp_pi_by_four_decimal_int
+            \c_fp_pi_by_four_extended_int
+          \l_fp_input_a_integer_int \l_fp_input_a_decimal_int
+            \l_fp_input_a_extended_int
+        \tex_advance:D \l_fp_trig_octant_int \c_one
+        \exp_after:wN \exp_after:wN \exp_after:wN
+          \fp_trig_octant_aux:
+      \fi:
+    \fi:
+  }
+\cs_new_protected_nopar:Npn \fp_sin:Nn  { \fp_sin_aux:NNn \tl_set:Nn }
+\cs_new_protected_nopar:Npn \fp_gsin:Nn { \fp_sin_aux:NNn \tl_gset:Nn }
+\cs_generate_variant:Nn \fp_sin:Nn   { c }
+\cs_generate_variant:Nn \fp_gsin:Nn  { c }
+\cs_new_protected_nopar:Npn \fp_sin_aux:NNn #1#2#3
+  {
+    \group_begin:
+      \fp_split:Nn a {#3}
+      \fp_standardise:NNNN
+        \l_fp_input_a_sign_int
+        \l_fp_input_a_integer_int
+        \l_fp_input_a_decimal_int
+        \l_fp_input_a_exponent_int
+      \tl_set:Nx \l_fp_arg_tl
+        {
+          \if_int_compare:w \l_fp_input_a_sign_int < \c_zero
+            -
+          \else:
+            +
+          \fi:
+          \int_use:N \l_fp_input_a_integer_int
+          .
+          \exp_after:wN \use_none:n
+            \int_value:w \int_eval:w
+              \l_fp_input_a_decimal_int + \c_one_thousand_million
+          e
+          \int_use:N \l_fp_input_a_exponent_int
+        }
+      \if_int_compare:w \l_fp_input_a_exponent_int < -\c_five
+        \cs_set_protected_nopar:Npx \fp_tmp:w
+        {
+          \group_end:
+          #1 \exp_not:N #2 { \l_fp_arg_tl }
+        }
+      \else:
+        \if_cs_exist:w
+          c_fp_sin ( \l_fp_arg_tl ) _fp
+        \cs_end:
+        \else:
+          \exp_after:wN \exp_after:wN \exp_after:wN
+            \fp_sin_aux_i:
+        \fi:
+        \cs_set_protected_nopar:Npx \fp_tmp:w
+          {
+            \group_end:
+            #1 \exp_not:N #2
+              { \use:c { c_fp_sin ( \l_fp_arg_tl ) _fp } }
+          }
+      \fi:
+    \fp_tmp:w
+  }
+\cs_new_protected_nopar:Npn \fp_sin_aux_i:
+  {
+    \fp_trig_normalise:
+    \fp_sin_aux_ii:
+    \if_int_compare:w \l_fp_output_integer_int = \c_one
+      \l_fp_output_exponent_int \c_zero
+    \else:
+      \l_fp_output_integer_int \l_fp_output_decimal_int
+      \l_fp_output_decimal_int \l_fp_output_extended_int
+      \l_fp_output_exponent_int -\c_nine
+    \fi:
+    \fp_standardise:NNNN
+      \l_fp_input_a_sign_int
+      \l_fp_output_integer_int
+      \l_fp_output_decimal_int
+      \l_fp_output_exponent_int
+    \tl_new:c { c_fp_sin ( \l_fp_arg_tl ) _fp }
+    \tl_gset:cx { c_fp_sin ( \l_fp_arg_tl ) _fp }
+      {
+        \if_int_compare:w \l_fp_input_a_sign_int > \c_zero
+          +
+        \else:
+          -
+        \fi:
+        \int_use:N \l_fp_output_integer_int
+        .
+        \exp_after:wN \use_none:n
+          \int_value:w \int_eval:w
+             \l_fp_output_decimal_int + \c_one_thousand_million
+        e
+        \int_use:N \l_fp_output_exponent_int
+      }
+  }
+\cs_new_protected_nopar:Npn \fp_sin_aux_ii:
+  {
+    \if_case:w \l_fp_trig_octant_int
+    \or:
+      \exp_after:wN \fp_trig_calc_sin:
+    \or:
+      \exp_after:wN \fp_trig_calc_cos:
+    \or:
+      \exp_after:wN \fp_trig_calc_cos:
+    \or:
+      \exp_after:wN \fp_trig_calc_sin:
+    \fi:
+  }
+\cs_new_protected_nopar:Npn \fp_cos:Nn  { \fp_cos_aux:NNn \tl_set:Nn }
+\cs_new_protected_nopar:Npn \fp_gcos:Nn { \fp_cos_aux:NNn \tl_gset:Nn }
+\cs_generate_variant:Nn \fp_cos:Nn   { c }
+\cs_generate_variant:Nn \fp_gcos:Nn  { c }
+\cs_new_protected_nopar:Npn \fp_cos_aux:NNn #1#2#3
+  {
+    \group_begin:
+      \fp_split:Nn a {#3}
+      \fp_standardise:NNNN
+        \l_fp_input_a_sign_int
+        \l_fp_input_a_integer_int
+        \l_fp_input_a_decimal_int
+        \l_fp_input_a_exponent_int
+      \tl_set:Nx \l_fp_arg_tl
+        {
+          \if_int_compare:w \l_fp_input_a_sign_int < \c_zero
+            -
+          \else:
+            +
+          \fi:
+          \int_use:N \l_fp_input_a_integer_int
+          .
+        \exp_after:wN \use_none:n
+          \int_value:w \int_eval:w
+             \l_fp_input_a_decimal_int + \c_one_thousand_million
+          e
+          \int_use:N \l_fp_input_a_exponent_int
+        }
+      \if_cs_exist:w c_fp_cos ( \l_fp_arg_tl ) _fp \cs_end:
+      \else:
+        \exp_after:wN \fp_cos_aux_i:
+      \fi:
+      \cs_set_protected_nopar:Npx \fp_tmp:w
+        {
+          \group_end:
+          #1 \exp_not:N #2
+            { \use:c { c_fp_cos ( \l_fp_arg_tl ) _fp } }
+        }
+    \fp_tmp:w
+  }
+\cs_new_protected_nopar:Npn \fp_cos_aux_i:
+  {
+    \fp_trig_normalise:
+    \fp_cos_aux_ii:
+    \if_int_compare:w \l_fp_output_integer_int = \c_one
+      \l_fp_output_exponent_int \c_zero
+    \else:
+      \l_fp_output_integer_int \l_fp_output_decimal_int
+      \l_fp_output_decimal_int \l_fp_output_extended_int
+      \l_fp_output_exponent_int -\c_nine
+    \fi:
+    \fp_standardise:NNNN
+      \l_fp_input_a_sign_int
+      \l_fp_output_integer_int
+      \l_fp_output_decimal_int
+      \l_fp_output_exponent_int
+    \tl_new:c { c_fp_cos ( \l_fp_arg_tl ) _fp }
+    \tl_gset:cx { c_fp_cos ( \l_fp_arg_tl ) _fp }
+      {
+        \if_int_compare:w \l_fp_input_a_sign_int > \c_zero
+          +
+        \else:
+          -
+        \fi:
+        \int_use:N \l_fp_output_integer_int
+        .
+        \exp_after:wN \use_none:n
+          \int_value:w \int_eval:w
+             \l_fp_output_decimal_int + \c_one_thousand_million
+        e
+        \int_use:N \l_fp_output_exponent_int
+      }
+  }
+\cs_new_protected_nopar:Npn \fp_cos_aux_ii:
+  {
+    \if_case:w \l_fp_trig_octant_int
+    \or:
+      \exp_after:wN \fp_trig_calc_cos:
+    \or:
+      \exp_after:wN \fp_trig_calc_sin:
+    \or:
+      \exp_after:wN \fp_trig_calc_sin:
+    \or:
+      \exp_after:wN \fp_trig_calc_cos:
+    \fi:
+    \if_int_compare:w \l_fp_input_a_sign_int > \c_zero
+      \if_int_compare:w \l_fp_trig_octant_int > \c_two
+        \l_fp_input_a_sign_int \c_minus_one
+      \fi:
+    \else:
+      \if_int_compare:w \l_fp_trig_octant_int > \c_two
+      \else:
+        \l_fp_input_a_sign_int \c_one
+      \fi:
+    \fi:
+  }
+\cs_new_protected_nopar:Npn \fp_trig_calc_cos:
+  {
+    \if_int_compare:w \l_fp_input_a_decimal_int = \c_zero
+      \l_fp_output_integer_int \c_one
+      \l_fp_output_decimal_int \c_zero
+    \else:
+      \l_fp_trig_sign_int \c_minus_one
+      \fp_mul:NNNNNN
+        \l_fp_input_a_decimal_int \l_fp_input_a_extended_int
+        \l_fp_input_a_decimal_int \l_fp_input_a_extended_int
+        \l_fp_trig_decimal_int \l_fp_trig_extended_int
+      \fp_div_integer:NNNNN
+        \l_fp_trig_decimal_int \l_fp_trig_extended_int
+        \c_two
+        \l_fp_trig_decimal_int \l_fp_trig_extended_int
+      \l_fp_count_int \c_three
+      \if_int_compare:w \l_fp_trig_extended_int = \c_zero
+        \if_int_compare:w \l_fp_trig_decimal_int = \c_zero
+          \l_fp_output_integer_int \c_one
+          \l_fp_output_decimal_int \c_zero
+          \l_fp_output_extended_int \c_zero
+        \else:
+          \l_fp_output_integer_int \c_zero
+          \l_fp_output_decimal_int \c_one_thousand_million
+          \l_fp_output_extended_int \c_zero
+        \fi:
+      \else:
+        \l_fp_output_integer_int \c_zero
+        \l_fp_output_decimal_int 999999999 \scan_stop:
+        \l_fp_output_extended_int \c_one_thousand_million
+      \fi:
+      \tex_advance:D \l_fp_output_extended_int -\l_fp_trig_extended_int
+      \tex_advance:D \l_fp_output_decimal_int -\l_fp_trig_decimal_int
+      \exp_after:wN \fp_trig_calc_Taylor:
+    \fi:
+  }
+\cs_new_protected_nopar:Npn \fp_trig_calc_sin:
+  {
+    \l_fp_output_integer_int \c_zero
+    \if_int_compare:w \l_fp_input_a_decimal_int = \c_zero
+      \l_fp_output_decimal_int \c_zero
+    \else:
+      \l_fp_output_decimal_int \l_fp_input_a_decimal_int
+      \l_fp_output_extended_int \l_fp_input_a_extended_int
+      \l_fp_trig_sign_int \c_one
+      \l_fp_trig_decimal_int \l_fp_input_a_decimal_int
+      \l_fp_trig_extended_int \l_fp_input_a_extended_int
+      \l_fp_count_int \c_two
+      \exp_after:wN \fp_trig_calc_Taylor:
+    \fi:
+  }
+\cs_new_protected_nopar:Npn \fp_trig_calc_Taylor:
+  {
+    \l_fp_trig_sign_int -\l_fp_trig_sign_int
+    \fp_mul:NNNNNN
+      \l_fp_trig_decimal_int \l_fp_trig_extended_int
+      \l_fp_input_a_decimal_int \l_fp_input_a_extended_int
+      \l_fp_trig_decimal_int \l_fp_trig_extended_int
+    \fp_mul:NNNNNN
+      \l_fp_trig_decimal_int \l_fp_trig_extended_int
+      \l_fp_input_a_decimal_int \l_fp_input_a_extended_int
+      \l_fp_trig_decimal_int \l_fp_trig_extended_int
+    \fp_div_integer:NNNNN
+      \l_fp_trig_decimal_int \l_fp_trig_extended_int
+      \l_fp_count_int
+      \l_fp_trig_decimal_int \l_fp_trig_extended_int
+    \tex_advance:D \l_fp_count_int \c_one
+    \fp_div_integer:NNNNN
+      \l_fp_trig_decimal_int \l_fp_trig_extended_int
+      \l_fp_count_int
+      \l_fp_trig_decimal_int \l_fp_trig_extended_int
+    \tex_advance:D \l_fp_count_int \c_one
+    \if_int_compare:w \l_fp_trig_decimal_int > \c_zero
+      \if_int_compare:w \l_fp_trig_sign_int > \c_zero
+        \tex_advance:D \l_fp_output_decimal_int \l_fp_trig_decimal_int
+        \tex_advance:D \l_fp_output_extended_int
+          \l_fp_trig_extended_int
+        \if_int_compare:w \l_fp_output_extended_int < \c_one_thousand_million
+        \else:
+          \tex_advance:D \l_fp_output_decimal_int \c_one
+          \tex_advance:D \l_fp_output_extended_int
+            -\c_one_thousand_million
+        \fi:
+        \if_int_compare:w \l_fp_output_decimal_int < \c_one_thousand_million
+        \else:
+          \tex_advance:D \l_fp_output_integer_int \c_one
+          \tex_advance:D \l_fp_output_decimal_int
+            -\c_one_thousand_million
+        \fi:
+      \else:
+        \tex_advance:D \l_fp_output_decimal_int -\l_fp_trig_decimal_int
+        \tex_advance:D \l_fp_output_extended_int
+          -\l_fp_input_a_extended_int
+        \if_int_compare:w \l_fp_output_extended_int < \c_zero
+          \tex_advance:D \l_fp_output_decimal_int \c_minus_one
+          \tex_advance:D \l_fp_output_extended_int \c_one_thousand_million
+        \fi:
+        \if_int_compare:w \l_fp_output_decimal_int < \c_zero
+          \tex_advance:D \l_fp_output_integer_int \c_minus_one
+          \tex_advance:D \l_fp_output_decimal_int \c_one_thousand_million
+        \fi:
+      \fi:
+      \exp_after:wN \fp_trig_calc_Taylor:
+    \fi:
+  }
+\cs_new_protected_nopar:Npn \fp_tan:Nn  { \fp_tan_aux:NNn \tl_set:Nn }
+\cs_new_protected_nopar:Npn \fp_gtan:Nn { \fp_tan_aux:NNn \tl_gset:Nn }
+\cs_generate_variant:Nn \fp_tan:Nn   { c }
+\cs_generate_variant:Nn \fp_gtan:Nn  { c }
+\cs_new_protected_nopar:Npn \fp_tan_aux:NNn #1#2#3
+  {
+    \group_begin:
+      \fp_split:Nn a {#3}
+      \fp_standardise:NNNN
+        \l_fp_input_a_sign_int
+        \l_fp_input_a_integer_int
+        \l_fp_input_a_decimal_int
+        \l_fp_input_a_exponent_int
+      \tl_set:Nx \l_fp_arg_tl
+        {
+          \if_int_compare:w \l_fp_input_a_sign_int < \c_zero
+            -
+          \else:
+            +
+          \fi:
+          \int_use:N \l_fp_input_a_integer_int
+          .
+          \exp_after:wN \use_none:n
+            \int_value:w \int_eval:w
+              \l_fp_input_a_decimal_int + \c_one_thousand_million
+          e
+          \int_use:N \l_fp_input_a_exponent_int
+        }
+      \if_int_compare:w \l_fp_input_a_exponent_int < -\c_five
+        \cs_set_protected_nopar:Npx \fp_tmp:w
+        {
+          \group_end:
+          #1 \exp_not:N #2 { \l_fp_arg_tl }
+        }
+      \else:
+        \if_cs_exist:w
+          c_fp_tan ( \l_fp_arg_tl ) _fp
+        \cs_end:
+        \else:
+          \exp_after:wN \exp_after:wN \exp_after:wN
+            \fp_tan_aux_i:
+        \fi:
+        \cs_set_protected_nopar:Npx \fp_tmp:w
+          {
+            \group_end:
+            #1 \exp_not:N #2
+              { \use:c { c_fp_tan ( \l_fp_arg_tl ) _fp } }
+          }
+      \fi:
+    \fp_tmp:w
+  }
+\cs_new_protected_nopar:Npn \fp_tan_aux_i:
+  {
+    \if_int_compare:w \l_fp_input_a_exponent_int < \c_ten
+      \exp_after:wN \fp_tan_aux_ii:
+    \else:
+      \cs_new_eq:cN { c_fp_tan ( \l_fp_arg_tl ) _fp }
+        \c_zero_fp
+      \exp_after:wN \fp_trig_overflow_msg:
+    \fi:
+  }
+\cs_new_protected_nopar:Npn \fp_tan_aux_ii:
+  {
+    \fp_trig_normalise:
+    \if_int_compare:w \l_fp_input_a_sign_int > \c_zero
+      \if_int_compare:w \l_fp_trig_octant_int > \c_two
+        \l_fp_output_sign_int \c_minus_one
+      \else:
+        \l_fp_output_sign_int \c_one
+      \fi:
+    \else:
+      \if_int_compare:w \l_fp_trig_octant_int > \c_two
+        \l_fp_output_sign_int \c_one
+      \else:
+        \l_fp_output_sign_int \c_minus_one
+      \fi:
+    \fi:
+    \fp_cos_aux_ii:
+    \if_int_compare:w \l_fp_input_a_decimal_int = \c_zero
+      \if_int_compare:w \l_fp_input_a_integer_int = \c_zero
+        \cs_new_eq:cN { c_fp_tan ( \l_fp_arg_tl ) _fp }
+          \c_undefined_fp
+      \else:
+        \exp_after:wN \exp_after:wN \exp_after:wN
+          \fp_tan_aux_iii:
+      \fi:
+    \else:
+      \exp_after:wN \fp_tan_aux_iii:
+    \fi:
+  }
+\cs_new_protected_nopar:Npn \fp_tan_aux_iii:
+  {
+    \l_fp_input_b_integer_int \l_fp_output_decimal_int
+    \l_fp_input_b_decimal_int \l_fp_output_extended_int
+    \l_fp_input_b_exponent_int -\c_nine
+    \fp_standardise:NNNN
+      \l_fp_input_b_sign_int
+      \l_fp_input_b_integer_int
+      \l_fp_input_b_decimal_int
+      \l_fp_input_b_exponent_int
+    \fp_sin_aux_ii:
+    \l_fp_input_a_integer_int \l_fp_output_decimal_int
+    \l_fp_input_a_decimal_int \l_fp_output_extended_int
+    \l_fp_input_a_exponent_int -\c_nine
+    \fp_standardise:NNNN
+      \l_fp_input_a_sign_int
+      \l_fp_input_a_integer_int
+      \l_fp_input_a_decimal_int
+      \l_fp_input_a_exponent_int
+    \if_int_compare:w \l_fp_input_a_decimal_int = \c_zero
+      \if_int_compare:w \l_fp_input_a_integer_int = \c_zero
+        \cs_new_eq:cN { c_fp_tan ( \l_fp_arg_tl ) _fp }
+          \c_zero_fp
+      \else:
+        \exp_after:wN \exp_after:wN \exp_after:wN \fp_tan_aux_iv:
+      \fi:
+    \else:
+      \exp_after:wN \fp_tan_aux_iv:
+    \fi:
+  }
+\cs_new_protected_nopar:Npn \fp_tan_aux_iv:
+  {
+    \l_fp_output_integer_int \c_zero
+    \l_fp_output_decimal_int \c_zero
+    \cs_set_eq:NN \fp_div_store: \fp_div_store_integer:
+    \l_fp_div_offset_int \c_one_hundred_million
+    \fp_div_loop:
+    \l_fp_output_exponent_int
+      \int_eval:w
+        \l_fp_input_a_exponent_int - \l_fp_input_b_exponent_int
+      \int_eval_end:
+    \fp_standardise:NNNN
+      \l_fp_output_sign_int
+      \l_fp_output_integer_int
+      \l_fp_output_decimal_int
+      \l_fp_output_exponent_int
+    \tl_new:c { c_fp_tan ( \l_fp_arg_tl ) _fp }
+    \tl_gset:cx { c_fp_tan ( \l_fp_arg_tl ) _fp }
+      {
+        \if_int_compare:w \l_fp_output_sign_int > \c_zero
+          +
+        \else:
+          -
+        \fi:
+        \int_use:N \l_fp_output_integer_int
+        .
+        \exp_after:wN \use_none:n
+          \int_value:w \int_eval:w
+             \l_fp_output_decimal_int + \c_one_thousand_million
+        e
+        \int_use:N \l_fp_output_exponent_int
+      }
+  }
+\tl_const:cn { c_fp_exp_1_tl }   { { 2 } { 718281828 } { 459045235 } { 0 } }
+\tl_const:cn { c_fp_exp_2_tl }   { { 7 } { 389056098 } { 930650227 } { 0 } }
+\tl_const:cn { c_fp_exp_3_tl }   { { 2 } { 008553692 } { 318766774 } { 1 } }
+\tl_const:cn { c_fp_exp_4_tl }   { { 5 } { 459815003 } { 314423908 } { 1 } }
+\tl_const:cn { c_fp_exp_5_tl }   { { 1 } { 484131591 } { 025766034 } { 2 } }
+\tl_const:cn { c_fp_exp_6_tl }   { { 4 } { 034287934 } { 927351226 } { 2 } }
+\tl_const:cn { c_fp_exp_7_tl }   { { 1 } { 096633158 } { 428458599 } { 3 } }
+\tl_const:cn { c_fp_exp_8_tl }   { { 2 } { 980957987 } { 041728275 } { 3 } }
+\tl_const:cn { c_fp_exp_9_tl }   { { 8 } { 103083927 } { 575384008 } { 3 } }
+\tl_const:cn { c_fp_exp_10_tl }  { { 2 } { 202646579 } { 480671652 } { 4 } }
+\tl_const:cn { c_fp_exp_20_tl }  { { 4 } { 851651954 } { 097902280 } { 8 } }
+\tl_const:cn { c_fp_exp_30_tl }  { { 1 } { 068647458 } { 152446215 } { 13 } }
+\tl_const:cn { c_fp_exp_40_tl }  { { 2 } { 353852668 } { 370199854 } { 17 } }
+\tl_const:cn { c_fp_exp_50_tl }  { { 5 } { 184705528 } { 587072464 } { 21 } }
+\tl_const:cn { c_fp_exp_60_tl }  { { 1 } { 142007389 } { 815684284 } { 26 } }
+\tl_const:cn { c_fp_exp_70_tl }  { { 2 } { 515438670 } { 919167006 } { 30 } }
+\tl_const:cn { c_fp_exp_80_tl }  { { 5 } { 540622384 } { 393510053 } { 34 } }
+\tl_const:cn { c_fp_exp_90_tl }  { { 1 } { 220403294 } { 317840802 } { 39 } }
+\tl_const:cn { c_fp_exp_100_tl } { { 2 } { 688117141 } { 816135448 } { 43 } }
+\tl_const:cn { c_fp_exp_200_tl } { { 7 } { 225973768 } { 125749258 } { 86 } }
+\tl_const:cn { c_fp_exp_-1_tl }   { { 3 } { 678794411 } { 71442322 }  { -1 } }
+\tl_const:cn { c_fp_exp_-2_tl }   { { 1 } { 353352832 } { 366132692 } { -1 } }
+\tl_const:cn { c_fp_exp_-3_tl }   { { 4 } { 978706836 } { 786394298 } { -2 } }
+\tl_const:cn { c_fp_exp_-4_tl }   { { 1 } { 831563888 } { 873418029 } { -2 } }
+\tl_const:cn { c_fp_exp_-5_tl }   { { 6 } { 737946999 } { 085467097 } { -3 } }
+\tl_const:cn { c_fp_exp_-6_tl }   { { 2 } { 478752176 } { 666358423 } { -3 } }
+\tl_const:cn { c_fp_exp_-7_tl }   { { 9 } { 118819655 } { 545162080 } { -4 } }
+\tl_const:cn { c_fp_exp_-8_tl }   { { 3 } { 354626279 } { 025118388 } { -4 } }
+\tl_const:cn { c_fp_exp_-9_tl }   { { 1 } { 234098040 } { 866795495 } { -4 } }
+\tl_const:cn { c_fp_exp_-10_tl }  { { 4 } { 539992976 } { 248451536 } { -5 } }
+\tl_const:cn { c_fp_exp_-20_tl }  { { 2 } { 061153622 } { 438557828 } { -9 } }
+\tl_const:cn { c_fp_exp_-30_tl }  { { 9 } { 357622968 } { 840174605 } { -14 } }
+\tl_const:cn { c_fp_exp_-40_tl }  { { 4 } { 248354255 } { 291588995 } { -18 } }
+\tl_const:cn { c_fp_exp_-50_tl }  { { 1 } { 928749847 } { 963917783 } { -22 } }
+\tl_const:cn { c_fp_exp_-60_tl }  { { 8 } { 756510762 } { 696520338 } { -27 } }
+\tl_const:cn { c_fp_exp_-70_tl }  { { 3 } { 975449735 } { 908646808 } { -31 } }
+\tl_const:cn { c_fp_exp_-80_tl }  { { 1 } { 804851387 } { 845415172 } { -35 } }
+\tl_const:cn { c_fp_exp_-90_tl }  { { 8 } { 194012623 } { 990515430 } { -40 } }
+\tl_const:cn { c_fp_exp_-100_tl } { { 3 } { 720075976 } { 020835963 } { -44 } }
+\tl_const:cn { c_fp_exp_-200_tl } { { 1 } { 383896526 } { 736737530 } { -87 } }
+\cs_new_protected_nopar:Npn \fp_exp:Nn  { \fp_exp_aux:NNn \tl_set:Nn }
+\cs_new_protected_nopar:Npn \fp_gexp:Nn { \fp_exp_aux:NNn \tl_gset:Nn }
+\cs_generate_variant:Nn \fp_exp:Nn  { c }
+\cs_generate_variant:Nn \fp_gexp:Nn { c }
+\cs_new_protected_nopar:Npn \fp_exp_aux:NNn #1#2#3
+  {
+    \group_begin:
+      \fp_split:Nn a {#3}
+      \fp_standardise:NNNN
+        \l_fp_input_a_sign_int
+        \l_fp_input_a_integer_int
+        \l_fp_input_a_decimal_int
+        \l_fp_input_a_exponent_int
+      \l_fp_input_a_extended_int \c_zero
+      \tl_set:Nx \l_fp_arg_tl
+        {
+          \if_int_compare:w \l_fp_input_a_sign_int < \c_zero
+            -
+          \else:
+            +
+          \fi:
+          \int_use:N \l_fp_input_a_integer_int
+          .
+          \exp_after:wN \use_none:n
+            \int_value:w \int_eval:w
+               \l_fp_input_a_decimal_int + \c_one_thousand_million
+          e
+          \int_use:N \l_fp_input_a_exponent_int
+        }
+      \if_cs_exist:w c_fp_exp ( \l_fp_arg_tl ) _fp \cs_end:
+      \else:
+        \exp_after:wN \fp_exp_internal:
+      \fi:
+      \cs_set_protected_nopar:Npx \fp_tmp:w
+        {
+          \group_end:
+          #1 \exp_not:N #2
+            {
+              \if_cs_exist:w c_fp_exp ( \l_fp_arg_tl ) _fp
+                \cs_end:
+                \use:c { c_fp_exp ( \l_fp_arg_tl ) _fp }
+              \else:
+                \c_zero_fp
+              \fi:
+            }
+        }
+    \fp_tmp:w
+  }
+\cs_new_protected_nopar:Npn \fp_exp_internal:
+  {
+    \if_int_compare:w \l_fp_input_a_exponent_int < \c_three
+      \fp_extended_normalise:
+      \if_int_compare:w \l_fp_input_a_sign_int > \c_zero
+        \if_int_compare:w \l_fp_input_a_integer_int < 230 \scan_stop:
+          \exp_after:wN \exp_after:wN \exp_after:wN
+          \exp_after:wN \exp_after:wN \exp_after:wN
+            \exp_after:wN \fp_exp_aux:
+        \else:
+          \exp_after:wN \exp_after:wN \exp_after:wN
+          \exp_after:wN \exp_after:wN \exp_after:wN
+            \exp_after:wN \fp_exp_overflow_msg:
+          \fi:
+      \else:
+        \if_int_compare:w \l_fp_input_a_integer_int < 230 \scan_stop:
+          \exp_after:wN \exp_after:wN \exp_after:wN
+          \exp_after:wN \exp_after:wN \exp_after:wN
+            \exp_after:wN  \fp_exp_aux:
+        \else:
+          \fp_exp_const:cx { c_fp_exp ( \l_fp_arg_tl ) _fp }
+            { \c_zero_fp }
+        \fi:
+      \fi:
+    \else:
+      \exp_after:wN \fp_exp_overflow_msg:
+    \fi:
+  }
+\cs_new_protected_nopar:Npn \fp_exp_aux:
+  {
+    \if_int_compare:w \l_fp_input_a_integer_int > \c_zero
+      \exp_after:wN \fp_exp_integer:
+    \else:
+      \l_fp_output_integer_int  \c_one
+      \l_fp_output_decimal_int  \c_zero
+      \l_fp_output_extended_int \c_zero
+      \l_fp_output_exponent_int \c_zero
+      \exp_after:wN \fp_exp_decimal:
+    \fi:
+  }
+\cs_new_protected_nopar:Npn \fp_exp_integer:
+  {
+    \if_int_compare:w \l_fp_input_a_integer_int < \c_one_hundred
+      \l_fp_exp_integer_int  \c_one
+      \l_fp_exp_decimal_int  \c_zero
+      \l_fp_exp_extended_int \c_zero
+      \l_fp_exp_exponent_int \c_zero
+      \exp_after:wN \fp_exp_integer_tens:
+    \else:
+      \tl_set:Nx \l_fp_tmp_tl
+        {
+          \exp_after:wN \use_i:nnn
+            \int_use:N \l_fp_input_a_integer_int
+        }
+      \l_fp_input_a_integer_int
+        \int_eval:w
+          \l_fp_input_a_integer_int - \l_fp_tmp_tl 00
+        \int_eval_end:
+      \if_int_compare:w \l_fp_input_a_sign_int < \c_zero
+        \if_int_compare:w \l_fp_output_integer_int > 200 \scan_stop:
+          \fp_exp_const:cx { c_fp_exp ( \l_fp_arg_tl ) _fp }
+            { \c_zero_fp }
+        \else:
+          \fp_exp_integer_const:n { - \l_fp_tmp_tl 00 }
+          \exp_after:wN \exp_after:wN \exp_after:wN
+            \exp_after:wN \exp_after:wN \exp_after:wN
+            \exp_after:wN \fp_exp_integer_tens:
+        \fi:
+      \else:
+        \fp_exp_integer_const:n { \l_fp_tmp_tl 00 }
+        \exp_after:wN \exp_after:wN \exp_after:wN
+          \exp_after:wN \fp_exp_integer_tens:
+      \fi:
+    \fi:
+  }
+\cs_new_protected_nopar:Npn \fp_exp_integer_tens:
+  {
+    \l_fp_output_integer_int  \l_fp_exp_integer_int
+    \l_fp_output_decimal_int  \l_fp_exp_decimal_int
+    \l_fp_output_extended_int \l_fp_exp_extended_int
+    \l_fp_output_exponent_int \l_fp_exp_exponent_int
+    \if_int_compare:w \l_fp_input_a_integer_int > \c_nine
+      \tl_set:Nx \l_fp_tmp_tl
+        {
+          \exp_after:wN \use_i:nn
+            \int_use:N \l_fp_input_a_integer_int
+        }
+      \l_fp_input_a_integer_int
+        \int_eval:w
+          \l_fp_input_a_integer_int - \l_fp_tmp_tl 0
+        \int_eval_end:
+      \if_int_compare:w \l_fp_input_a_sign_int > \c_zero
+        \fp_exp_integer_const:n { \l_fp_tmp_tl 0 }
+      \else:
+        \fp_exp_integer_const:n { - \l_fp_tmp_tl 0 }
+      \fi:
+      \fp_mul:NNNNNNNNN
+        \l_fp_exp_integer_int \l_fp_exp_decimal_int \l_fp_exp_extended_int
+        \l_fp_output_integer_int \l_fp_output_decimal_int
+          \l_fp_output_extended_int
+        \l_fp_output_integer_int \l_fp_output_decimal_int
+          \l_fp_output_extended_int
+      \tex_advance:D \l_fp_output_exponent_int \l_fp_exp_exponent_int
+      \fp_extended_normalise_output:
+    \fi:
+    \fp_exp_integer_units:
+  }
+\cs_new_protected_nopar:Npn \fp_exp_integer_units:
+  {
+    \if_int_compare:w \l_fp_input_a_integer_int > \c_zero
+      \if_int_compare:w \l_fp_input_a_sign_int > \c_zero
+        \fp_exp_integer_const:n { \int_use:N \l_fp_input_a_integer_int }
+      \else:
+        \fp_exp_integer_const:n
+          { - \int_use:N \l_fp_input_a_integer_int }
+      \fi:
+      \fp_mul:NNNNNNNNN
+        \l_fp_exp_integer_int \l_fp_exp_decimal_int \l_fp_exp_extended_int
+        \l_fp_output_integer_int \l_fp_output_decimal_int
+          \l_fp_output_extended_int
+        \l_fp_output_integer_int \l_fp_output_decimal_int
+          \l_fp_output_extended_int
+      \tex_advance:D \l_fp_output_exponent_int \l_fp_exp_exponent_int
+      \fp_extended_normalise_output:
+    \fi:
+    \fp_exp_decimal:
+  }
+\cs_new_protected_nopar:Npn \fp_exp_integer_const:n #1
+  {
+    \exp_after:wN \exp_after:wN \exp_after:wN
+      \fp_exp_integer_const:nnnn
+      \cs:w c_fp_exp_ #1 _tl \cs_end:
+  }
+\cs_new_protected_nopar:Npn \fp_exp_integer_const:nnnn #1#2#3#4
+  {
+    \l_fp_exp_integer_int  #1 \scan_stop:
+    \l_fp_exp_decimal_int  #2 \scan_stop:
+    \l_fp_exp_extended_int #3 \scan_stop:
+    \l_fp_exp_exponent_int #4 \scan_stop:
+  }
+\cs_new_protected_nopar:Npn \fp_exp_decimal:
+  {
+    \if_int_compare:w \l_fp_input_a_decimal_int > \c_zero
+      \if_int_compare:w \l_fp_input_a_sign_int > \c_zero
+        \l_fp_exp_integer_int  \c_one
+        \l_fp_exp_decimal_int  \l_fp_input_a_decimal_int
+        \l_fp_exp_extended_int \l_fp_input_a_extended_int
+      \else:
+        \l_fp_exp_integer_int \c_zero
+        \if_int_compare:w \l_fp_exp_extended_int = \c_zero
+          \l_fp_exp_decimal_int
+            \int_eval:w
+              \c_one_thousand_million - \l_fp_input_a_decimal_int
+            \int_eval_end:
+          \l_fp_exp_extended_int \c_zero
+        \else:
+          \l_fp_exp_decimal_int
+            \int_eval:w
+              999999999 - \l_fp_input_a_decimal_int
+            \scan_stop:
+          \l_fp_exp_extended_int
+            \int_eval:w
+              \c_one_thousand_million - \l_fp_input_a_extended_int
+            \int_eval_end:
+        \fi:
+      \fi:
+      \l_fp_input_b_sign_int     \l_fp_input_a_sign_int
+      \l_fp_input_b_decimal_int  \l_fp_input_a_decimal_int
+      \l_fp_input_b_extended_int \l_fp_input_a_extended_int
+      \l_fp_count_int \c_one
+      \fp_exp_Taylor:
+      \fp_mul:NNNNNNNNN
+        \l_fp_exp_integer_int \l_fp_exp_decimal_int \l_fp_exp_extended_int
+        \l_fp_output_integer_int \l_fp_output_decimal_int
+          \l_fp_output_extended_int
+        \l_fp_output_integer_int \l_fp_output_decimal_int
+          \l_fp_output_extended_int
+    \fi:
+    \if_int_compare:w \l_fp_output_extended_int < \c_five_hundred_million
+    \else:
+      \tex_advance:D \l_fp_output_decimal_int \c_one
+      \if_int_compare:w \l_fp_output_decimal_int < \c_one_thousand_million
+      \else:
+        \l_fp_output_decimal_int \c_zero
+        \tex_advance:D \l_fp_output_integer_int \c_one
+      \fi:
+    \fi:
+    \fp_standardise:NNNN
+      \l_fp_output_sign_int
+      \l_fp_output_integer_int
+      \l_fp_output_decimal_int
+      \l_fp_output_exponent_int
+    \fp_exp_const:cx { c_fp_exp ( \l_fp_arg_tl ) _fp }
+      {
+        +
+        \int_use:N \l_fp_output_integer_int
+        .
+        \exp_after:wN \use_none:n
+          \int_value:w \int_eval:w
+             \l_fp_output_decimal_int + \c_one_thousand_million
+        e
+        \int_use:N \l_fp_output_exponent_int
+      }
+  }
+\cs_new_protected_nopar:Npn \fp_exp_Taylor:
+  {
+    \tex_advance:D \l_fp_count_int \c_one
+    \tex_multiply:D \l_fp_input_b_sign_int \l_fp_input_a_sign_int
+    \fp_mul:NNNNNN
+      \l_fp_input_a_decimal_int \l_fp_input_a_extended_int
+      \l_fp_input_b_decimal_int \l_fp_input_b_extended_int
+      \l_fp_input_b_decimal_int \l_fp_input_b_extended_int
+    \fp_div_integer:NNNNN
+      \l_fp_input_b_decimal_int \l_fp_input_b_extended_int
+      \l_fp_count_int
+      \l_fp_input_b_decimal_int \l_fp_input_b_extended_int
+    \if_int_compare:w
+      \int_eval:w
+        \l_fp_input_b_decimal_int + \l_fp_input_b_extended_int
+        > \c_zero
+      \if_int_compare:w \l_fp_input_b_sign_int > \c_zero
+        \tex_advance:D \l_fp_exp_decimal_int \l_fp_input_b_decimal_int
+        \tex_advance:D \l_fp_exp_extended_int
+          \l_fp_input_b_extended_int
+        \if_int_compare:w \l_fp_exp_extended_int < \c_one_thousand_million
+      \else:
+          \tex_advance:D \l_fp_exp_decimal_int \c_one
+          \tex_advance:D \l_fp_exp_extended_int
+            -\c_one_thousand_million
+        \fi:
+        \if_int_compare:w \l_fp_exp_decimal_int < \c_one_thousand_million
+        \else:
+          \tex_advance:D \l_fp_exp_integer_int \c_one
+          \tex_advance:D \l_fp_exp_decimal_int
+            -\c_one_thousand_million
+        \fi:
+      \else:
+        \tex_advance:D \l_fp_exp_decimal_int -\l_fp_input_b_decimal_int
+        \tex_advance:D \l_fp_exp_extended_int
+          -\l_fp_input_a_extended_int
+        \if_int_compare:w \l_fp_exp_extended_int < \c_zero
+          \tex_advance:D \l_fp_exp_decimal_int \c_minus_one
+          \tex_advance:D \l_fp_exp_extended_int \c_one_thousand_million
+        \fi:
+        \if_int_compare:w \l_fp_exp_decimal_int < \c_zero
+          \tex_advance:D \l_fp_exp_integer_int \c_minus_one
+          \tex_advance:D \l_fp_exp_decimal_int \c_one_thousand_million
+        \fi:
+      \fi:
+      \exp_after:wN \fp_exp_Taylor:
+    \fi:
+  }
+\cs_new_protected_nopar:Npn \fp_exp_const:Nx #1#2
+  {
+    \tl_new:N #1
+    \tl_gset:Nx #1 {#2}
+  }
+\cs_generate_variant:Nn \fp_exp_const:Nx { c }
+\tl_const:cn { c_fp_ln_10_1_tl } { { 2 } { 302585092 } { 994045684 } { 0 } }
+\tl_const:cn { c_fp_ln_10_2_tl } { { 4 } { 605170185 } { 988091368 } { 0 } }
+\tl_const:cn { c_fp_ln_10_3_tl } { { 6 } { 907755278 } { 982137052 } { 0 } }
+\tl_const:cn { c_fp_ln_10_4_tl } { { 9 } { 210340371 } { 976182736 } { 0 } }
+\tl_const:cn { c_fp_ln_10_5_tl } { { 1 } { 151292546 } { 497022842 } { 1 } }
+\tl_const:cn { c_fp_ln_10_6_tl } { { 1 } { 381551055 } { 796427410 } { 1 } }
+\tl_const:cn { c_fp_ln_10_7_tl } { { 1 } { 611809565 } { 095831979 } { 1 } }
+\tl_const:cn { c_fp_ln_10_8_tl } { { 1 } { 842068074 } { 395226547 } { 1 } }
+\tl_const:cn { c_fp_ln_10_9_tl } { { 2 } { 072326583 } { 694641116 } { 1 } }
+\tl_const:cn { c_fp_ln_2_1_tl } { { 0 } { 693147180 } { 559945309 } { 0 } }
+\tl_const:cn { c_fp_ln_2_2_tl } { { 1 } { 386294361 } { 119890618 } { 0 } }
+\tl_const:cn { c_fp_ln_2_3_tl } { { 2 } { 079441541 } { 679835928 } { 0 } }
+\cs_new_protected_nopar:Npn \fp_ln:Nn  { \fp_ln_aux:NNn \tl_set:Nn  }
+\cs_new_protected_nopar:Npn \fp_gln:Nn { \fp_ln_aux:NNn \tl_gset:Nn }
+\cs_generate_variant:Nn \fp_ln:Nn  { c }
+\cs_generate_variant:Nn \fp_gln:Nn { c }
+\cs_new_protected_nopar:Npn \fp_ln_aux:NNn #1#2#3
+  {
+    \group_begin:
+      \fp_split:Nn a {#3}
+      \fp_standardise:NNNN
+        \l_fp_input_a_sign_int
+        \l_fp_input_a_integer_int
+        \l_fp_input_a_decimal_int
+        \l_fp_input_a_exponent_int
+      \if_int_compare:w \l_fp_input_a_sign_int > \c_zero
+        \if_int_compare:w
+          \int_eval:w
+            \l_fp_input_a_integer_int + \l_fp_input_a_decimal_int
+            > \c_zero
+          \exp_after:wN \exp_after:wN \exp_after:wN \fp_ln_aux:
+        \else:
+          \cs_set_protected_nopar:Npx \fp_tmp:w ##1##2
+            {
+              \group_end:
+              ##1 \exp_not:N ##2 { \c_zero_fp }
+            }
+          \exp_after:wN \exp_after:wN \exp_after:wN \fp_ln_error_msg:
+        \fi:
+      \else:
+        \cs_set_protected_nopar:Npx \fp_tmp:w ##1##2
+          {
+            \group_end:
+            ##1 \exp_not:N ##2 { \c_zero_fp }
+          }
+        \exp_after:wN \fp_ln_error_msg:
+      \fi:
+    \fp_tmp:w #1 #2
+  }
+\cs_new_protected_nopar:Npn \fp_ln_aux:
+  {
+    \tl_set:Nx \l_fp_arg_tl
+      {
+        +
+        \int_use:N \l_fp_input_a_integer_int
+        .
+        \exp_after:wN \use_none:n
+          \int_value:w \int_eval:w
+             \l_fp_input_a_decimal_int + \c_one_thousand_million
+        e
+        \int_use:N \l_fp_input_a_exponent_int
+      }
+    \if_cs_exist:w c_fp_ln ( \l_fp_arg_tl ) _fp \cs_end:
+    \else:
+      \exp_after:wN \fp_ln_exponent:
+    \fi:
+    \cs_set_protected_nopar:Npx \fp_tmp:w ##1##2
+      {
+        \group_end:
+        ##1 \exp_not:N ##2
+          { \use:c { c_fp_ln ( \l_fp_arg_tl ) _fp } }
+      }
+  }
+\cs_new_protected_nopar:Npn \fp_ln_exponent:
+  {
+    \fp_ln_internal:
+    \if_int_compare:w \l_fp_output_extended_int < \c_five_hundred_million
+    \else:
+      \tex_advance:D \l_fp_output_decimal_int \c_one
+      \if_int_compare:w \l_fp_output_decimal_int < \c_one_thousand_million
+      \else:
+        \l_fp_output_decimal_int \c_zero
+        \tex_advance:D \l_fp_output_integer_int \c_one
+      \fi:
+    \fi:
+    \fp_standardise:NNNN
+      \l_fp_output_sign_int
+      \l_fp_output_integer_int
+      \l_fp_output_decimal_int
+      \l_fp_output_exponent_int
+    \tl_const:cx { c_fp_ln ( \l_fp_arg_tl ) _fp }
+      {
+        \if_int_compare:w \l_fp_output_sign_int > \c_zero
+          +
+        \else:
+          -
+        \fi:
+        \int_use:N \l_fp_output_integer_int
+        .
+        \exp_after:wN \use_none:n
+          \int_value:w \int_eval:w
+             \l_fp_output_decimal_int + \c_one_thousand_million
+           \scan_stop:
+        e
+        \int_use:N \l_fp_output_exponent_int
+      }
+  }
+\cs_new_protected_nopar:Npn \fp_ln_internal:
+  {
+    \if_int_compare:w \l_fp_input_a_exponent_int < \c_zero
+      \l_fp_input_a_exponent_int -\l_fp_input_a_exponent_int
+      \l_fp_output_sign_int \c_minus_one
+    \else:
+      \l_fp_output_sign_int \c_one
+    \fi:
+    \if_int_compare:w \l_fp_input_a_exponent_int > \c_nine
+      \exp_after:wN \fp_ln_exponent_tens:NN
+        \int_use:N \l_fp_input_a_exponent_int
+    \else:
+      \l_fp_output_integer_int  \c_zero
+      \l_fp_output_decimal_int  \c_zero
+      \l_fp_output_extended_int \c_zero
+      \l_fp_output_exponent_int \c_zero
+    \fi:
+    \fp_ln_exponent_units:
+  }
+\cs_new_protected_nopar:Npn \fp_ln_exponent_tens:NN #1 #2
+  {
+    \l_fp_input_a_exponent_int #2 \scan_stop:
+    \fp_ln_const:nn { 10 } { #1 }
+    \tex_advance:D \l_fp_exp_exponent_int \c_one
+    \l_fp_output_integer_int  \l_fp_exp_integer_int
+    \l_fp_output_decimal_int  \l_fp_exp_decimal_int
+    \l_fp_output_extended_int \l_fp_exp_extended_int
+    \l_fp_output_exponent_int \l_fp_exp_exponent_int
+  }
+\cs_new_protected_nopar:Npn \fp_ln_exponent_units:
+  {
+    \if_int_compare:w \l_fp_input_a_exponent_int > \c_zero
+      \fp_ln_const:nn { 10 } { \int_use:N \l_fp_input_a_exponent_int }
+      \fp_ln_normalise:
+      \fp_add:NNNNNNNNN
+        \l_fp_exp_integer_int \l_fp_exp_decimal_int \l_fp_exp_extended_int
+        \l_fp_output_integer_int \l_fp_output_decimal_int
+          \l_fp_output_extended_int
+        \l_fp_output_integer_int \l_fp_output_decimal_int
+          \l_fp_output_extended_int
+    \fi:
+    \fp_ln_mantissa:
+  }
+\cs_new_protected_nopar:Npn \fp_ln_normalise:
+  {
+    \if_int_compare:w \l_fp_exp_exponent_int < \l_fp_output_exponent_int
+      \tex_advance:D \l_fp_exp_decimal_int \c_one_thousand_million
+      \exp_after:wN \use_i:nn \exp_after:wN
+        \fp_ln_normalise_aux:NNNNNNNNN
+        \int_use:N \l_fp_exp_decimal_int
+       \exp_after:wN \fp_ln_normalise:
+    \else:
+      \l_fp_output_exponent_int \l_fp_exp_exponent_int
+    \fi:
+  }
+\cs_new_protected_nopar:Npn \fp_ln_normalise_aux:NNNNNNNNN #1#2#3#4#5#6#7#8#9
+  {
+    \if_int_compare:w \l_fp_exp_integer_int = \c_zero
+      \l_fp_exp_decimal_int #1#2#3#4#5#6#7#8 \scan_stop:
+    \else:
+      \tl_set:Nx \l_fp_tmp_tl
+        {
+          \int_use:N \l_fp_exp_integer_int
+          #1#2#3#4#5#6#7#8
+        }
+      \l_fp_exp_integer_int \c_zero
+      \l_fp_exp_decimal_int \l_fp_tmp_tl \scan_stop:
+    \fi:
+    \tex_divide:D \l_fp_exp_extended_int \c_ten
+    \tl_set:Nx \l_fp_tmp_tl
+      {
+        #9
+        \int_use:N \l_fp_exp_extended_int
+      }
+    \l_fp_exp_extended_int \l_fp_tmp_tl \scan_stop:
+    \tex_advance:D \l_fp_exp_exponent_int \c_one
+  }
+\cs_new_protected_nopar:Npn \fp_ln_mantissa:
+  {
+    \l_fp_count_int \c_zero
+    \l_fp_input_a_extended_int \c_zero
+    \fp_ln_mantissa_aux:
+    \if_int_compare:w \l_fp_count_int > \c_zero
+      \fp_ln_const:nn { 2 } { \int_use:N \l_fp_count_int }
+      \fp_ln_normalise:
+      \if_int_compare:w \l_fp_output_sign_int > \c_zero
+        \exp_after:wN \fp_add:NNNNNNNNN
+      \else:
+        \exp_after:wN \fp_sub:NNNNNNNNN
+      \fi:
+      \l_fp_output_integer_int \l_fp_output_decimal_int
+        \l_fp_output_extended_int
+      \l_fp_exp_integer_int \l_fp_exp_decimal_int \l_fp_exp_extended_int
+      \l_fp_output_integer_int \l_fp_output_decimal_int
+        \l_fp_output_extended_int
+    \fi:
+    \if_int_compare:w
+      \int_eval:w
+        \l_fp_input_a_integer_int + \l_fp_input_a_decimal_int > \c_one
+      \exp_after:wN \fp_ln_Taylor:
+    \fi:
+  }
+\cs_new_protected_nopar:Npn \fp_ln_mantissa_aux:
+  {
+    \if_int_compare:w \l_fp_input_a_integer_int > \c_one
+      \tex_advance:D \l_fp_count_int \c_one
+      \fp_ln_mantissa_divide_two:
+      \exp_after:wN \fp_ln_mantissa_aux:
+    \fi:
+  }
+\cs_new_protected_nopar:Npn \fp_ln_mantissa_divide_two:
+  {
+    \if_int_odd:w \l_fp_input_a_decimal_int
+      \tex_advance:D \l_fp_input_a_extended_int \c_one_thousand_million
+    \fi:
+    \if_int_odd:w \l_fp_input_a_integer_int
+      \tex_advance:D \l_fp_input_a_decimal_int \c_one_thousand_million
+    \fi:
+    \tex_divide:D \l_fp_input_a_integer_int  \c_two
+    \tex_divide:D \l_fp_input_a_decimal_int  \c_two
+    \tex_divide:D \l_fp_input_a_extended_int \c_two
+  }
+\cs_new_protected_nopar:Npn \fp_ln_const:nn #1#2
+  {
+    \exp_after:wN \exp_after:wN \exp_after:wN
+      \fp_exp_integer_const:nnnn
+      \cs:w c_fp_ln_ #1 _ #2 _tl \cs_end:
+  }
+\cs_new_protected_nopar:Npn \fp_ln_Taylor:
+  {
+    \group_begin:
+      \l_fp_input_a_integer_int \c_zero
+      \l_fp_input_a_exponent_int \c_zero
+      \l_fp_input_b_integer_int \c_two
+      \l_fp_input_b_decimal_int \l_fp_input_a_decimal_int
+      \l_fp_input_b_exponent_int \c_zero
+      \fp_div_internal:
+      \fp_ln_fixed:
+      \l_fp_input_a_integer_int  \l_fp_output_integer_int
+      \l_fp_input_a_decimal_int  \l_fp_output_decimal_int
+      \l_fp_input_a_extended_int \c_zero
+      \l_fp_input_a_exponent_int \l_fp_output_exponent_int
+      \l_fp_output_decimal_int \c_zero %^^A Bug?
+      \l_fp_output_decimal_int  \l_fp_input_a_decimal_int
+      \l_fp_output_extended_int \l_fp_input_a_extended_int
+      \fp_mul:NNNNNN
+        \l_fp_input_a_decimal_int \l_fp_input_a_extended_int
+        \l_fp_input_a_decimal_int \l_fp_input_a_extended_int
+        \l_fp_input_b_decimal_int \l_fp_input_b_extended_int
+      \l_fp_count_int \c_one
+      \fp_ln_Taylor_aux:
+      \cs_set_protected_nopar:Npx \fp_tmp:w
+        {
+          \group_end:
+          \l_fp_exp_integer_int \c_zero
+          \exp_not:N \l_fp_exp_decimal_int
+            \int_use:N \l_fp_output_decimal_int \scan_stop:
+          \exp_not:N \l_fp_exp_extended_int
+            \int_use:N \l_fp_output_extended_int \scan_stop:
+          \exp_not:N \l_fp_exp_exponent_int
+            \int_use:N \l_fp_output_exponent_int \scan_stop:
+        }
+    \fp_tmp:w
+    \tex_advance:D \l_fp_exp_decimal_int \l_fp_exp_decimal_int
+    \if_int_compare:w \l_fp_exp_extended_int < \c_five_hundred_million
+    \else:
+      \tex_advance:D \l_fp_exp_extended_int -\c_five_hundred_million
+      \tex_advance:D \l_fp_exp_decimal_int \c_one
+    \fi:
+    \tex_advance:D \l_fp_exp_extended_int \l_fp_exp_extended_int
+    \fp_ln_normalise:
+    \if_int_compare:w \l_fp_output_sign_int > \c_zero
+      \exp_after:wN \fp_add:NNNNNNNNN
+    \else:
+      \exp_after:wN \fp_sub:NNNNNNNNN
+    \fi:
+    \l_fp_output_integer_int \l_fp_output_decimal_int
+      \l_fp_output_extended_int
+    \c_zero \l_fp_exp_decimal_int \l_fp_exp_extended_int
+    \l_fp_output_integer_int \l_fp_output_decimal_int
+      \l_fp_output_extended_int
+  }
+\cs_new_protected_nopar:Npn \fp_ln_fixed:
+  {
+    \if_int_compare:w \l_fp_output_exponent_int < \c_zero
+      \tex_advance:D \l_fp_output_decimal_int \c_one_thousand_million
+      \exp_after:wN \use_i:nn \exp_after:wN
+        \fp_ln_fixed_aux:NNNNNNNNN
+        \int_use:N \l_fp_output_decimal_int
+       \exp_after:wN \fp_ln_fixed:
+     \fi:
+  }
+\cs_new_protected_nopar:Npn \fp_ln_fixed_aux:NNNNNNNNN #1#2#3#4#5#6#7#8#9
+  {
+    \if_int_compare:w \l_fp_output_integer_int = \c_zero
+      \l_fp_output_decimal_int #1#2#3#4#5#6#7#8 \scan_stop:
+    \else:
+      \tl_set:Nx \l_fp_tmp_tl
+        {
+          \int_use:N \l_fp_output_integer_int
+          #1#2#3#4#5#6#7#8
+        }
+      \l_fp_output_integer_int \c_zero
+      \l_fp_output_decimal_int \l_fp_tmp_tl \scan_stop:
+    \fi:
+    \tex_advance:D \l_fp_output_exponent_int \c_one
+  }
+\cs_new_protected_nopar:Npn \fp_ln_Taylor_aux:
+  {
+    \tex_advance:D \l_fp_count_int \c_two
+    \fp_mul:NNNNNN
+      \l_fp_input_a_decimal_int \l_fp_input_a_extended_int
+      \l_fp_input_b_decimal_int \l_fp_input_b_extended_int
+      \l_fp_input_a_decimal_int \l_fp_input_a_extended_int
+    \if_int_compare:w
+      \int_eval:w
+        \l_fp_input_a_decimal_int + \l_fp_input_a_extended_int
+        > \c_zero
+      \fp_div_integer:NNNNN
+        \l_fp_input_a_decimal_int \l_fp_input_a_extended_int
+        \l_fp_count_int
+        \l_fp_exp_decimal_int \l_fp_exp_extended_int
+        \tex_advance:D \l_fp_output_decimal_int \l_fp_exp_decimal_int
+        \tex_advance:D \l_fp_output_extended_int \l_fp_exp_extended_int
+        \if_int_compare:w \l_fp_output_extended_int < \c_one_thousand_million
+        \else:
+          \tex_advance:D \l_fp_output_decimal_int \c_one
+          \tex_advance:D \l_fp_output_extended_int
+            -\c_one_thousand_million
+        \fi:
+        \if_int_compare:w \l_fp_output_decimal_int < \c_one_thousand_million
+        \else:
+          \tex_advance:D \l_fp_output_integer_int \c_one
+          \tex_advance:D \l_fp_output_decimal_int
+            -\c_one_thousand_million
+        \fi:
+      \exp_after:wN \fp_ln_Taylor_aux:
+    \fi:
+  }
+\cs_new_protected_nopar:Npn \fp_pow:Nn  { \fp_pow_aux:NNn \tl_set:Nn }
+\cs_new_protected_nopar:Npn \fp_gpow:Nn { \fp_pow_aux:NNn \tl_gset:Nn }
+\cs_generate_variant:Nn \fp_pow:Nn  { c }
+\cs_generate_variant:Nn \fp_gpow:Nn { c }
+\cs_new_protected_nopar:Npn \fp_pow_aux:NNn #1#2#3
+  {
+    \group_begin:
+      \fp_read:N #2
+      \l_fp_input_b_sign_int     \l_fp_input_a_sign_int
+      \l_fp_input_b_integer_int  \l_fp_input_a_integer_int
+      \l_fp_input_b_decimal_int  \l_fp_input_a_decimal_int
+      \l_fp_input_b_exponent_int \l_fp_input_a_exponent_int
+      \fp_split:Nn a {#3}
+      \fp_standardise:NNNN
+        \l_fp_input_a_sign_int
+        \l_fp_input_a_integer_int
+        \l_fp_input_a_decimal_int
+        \l_fp_input_a_exponent_int
+      \if_int_compare:w
+        \int_eval:w
+          \l_fp_input_b_integer_int + \l_fp_input_b_decimal_int
+          = \c_zero
+         \if_int_compare:w
+           \int_eval:w
+             \l_fp_input_a_integer_int + \l_fp_input_a_decimal_int
+             = \c_zero
+             \cs_set_protected_nopar:Npx \fp_tmp:w ##1##2
+               {
+                 \group_end:
+                 ##1 ##2 { \c_undefined_fp }
+               }
+           \else:
+             \cs_set_protected_nopar:Npx \fp_tmp:w ##1##2
+               {
+                 \group_end:
+                 ##1 ##2 { \c_zero_fp }
+               }
+          \fi:
+       \else:
+         \if_int_compare:w
+           \int_eval:w
+             \l_fp_input_a_integer_int + \l_fp_input_a_decimal_int
+             = \c_zero
+             \cs_set_protected_nopar:Npx \fp_tmp:w ##1##2
+               {
+                 \group_end:
+                 ##1 ##2 { \c_one_fp }
+               }
+           \else:
+             \exp_after:wN \exp_after:wN \exp_after:wN
+               \fp_pow_aux_i:
+          \fi:
+       \fi:
+    \fp_tmp:w #1 #2
+}
+\cs_new_protected_nopar:Npn \fp_pow_aux_i:
+  {
+    \if_int_compare:w \l_fp_input_b_sign_int > \c_zero
+      \tl_set:Nn \l_fp_sign_tl { + }
+      \exp_after:wN \fp_pow_aux_ii:
+    \else:
+      \l_fp_input_a_extended_int \c_zero
+      \if_int_compare:w \l_fp_input_a_exponent_int < \c_ten
+        \group_begin:
+        \fp_extended_normalise:
+        \if_int_compare:w
+          \int_eval:w
+            \l_fp_input_a_decimal_int + \l_fp_input_a_extended_int
+            = \c_zero
+           \group_end:
+          \tl_set:Nn \l_fp_sign_tl { - }
+          \exp_after:wN \exp_after:wN \exp_after:wN
+          \exp_after:wN \exp_after:wN \exp_after:wN
+          \exp_after:wN \fp_pow_aux_ii:
+        \else:
+          \group_end:
+          \cs_set_protected_nopar:Npx \fp_tmp:w ##1##2
+            {
+              \group_end:
+              ##1 ##2 { \c_undefined_fp }
+           }
+        \fi:
+      \else:
+        \cs_set_protected_nopar:Npx \fp_tmp:w ##1##2
+          {
+            \group_end:
+            ##1 ##2 { \c_undefined_fp }
+         }
+      \fi:
+    \fi:
+  }
+\cs_new_protected_nopar:Npn \fp_pow_aux_ii:
+  {
+    \if_int_compare:w \l_fp_input_a_sign_int > \c_zero
+      \exp_after:wN \fp_pow_aux_iv:
+    \else:
+      \if_int_compare:w \l_fp_input_a_exponent_int < \c_ten
+        \group_begin:
+        \l_fp_input_a_extended_int \c_zero
+        \fp_extended_normalise:
+        \if_int_compare:w \l_fp_input_a_decimal_int = \c_zero
+          \if_int_compare:w \l_fp_input_a_integer_int > \c_ten
+            \group_end:
+            \exp_after:wN \exp_after:wN \exp_after:wN
+            \exp_after:wN \exp_after:wN \exp_after:wN
+            \exp_after:wN \exp_after:wN \exp_after:wN
+            \exp_after:wN \exp_after:wN \exp_after:wN
+            \exp_after:wN \exp_after:wN \exp_after:wN
+              \fp_pow_aux_iv:
+          \else:
+            \group_end:
+            \exp_after:wN \exp_after:wN \exp_after:wN
+            \exp_after:wN \exp_after:wN \exp_after:wN
+            \exp_after:wN \exp_after:wN \exp_after:wN
+            \exp_after:wN \exp_after:wN \exp_after:wN
+            \exp_after:wN \exp_after:wN \exp_after:wN
+              \exp_after:wN \fp_pow_aux_iii:
+          \fi:
+        \else:
+          \group_end:
+            \exp_after:wN \exp_after:wN \exp_after:wN
+            \exp_after:wN \exp_after:wN \exp_after:wN
+              \exp_after:wN \fp_pow_aux_iv:
+        \fi:
+      \else:
+        \exp_after:wN \exp_after:wN \exp_after:wN
+          \fp_pow_aux_iv:
+      \fi:
+    \fi:
+    \cs_set_protected_nopar:Npx \fp_tmp:w ##1##2
+      {
+        \group_end:
+        ##1 ##2
+          {
+            \l_fp_sign_tl
+            \int_use:N \l_fp_output_integer_int
+            .
+            \exp_after:wN \use_none:n
+              \int_value:w \int_eval:w
+                 \l_fp_output_decimal_int + \c_one_thousand_million
+            e
+            \int_use:N \l_fp_output_exponent_int
+          }
+      }
+  }
+\cs_new_protected_nopar:Npn \fp_pow_aux_iii:
+  {
+    \l_fp_input_a_sign_int \c_one
+    \fp_pow_aux_iv:
+    \l_fp_input_a_integer_int  \c_one
+    \l_fp_input_a_decimal_int  \c_zero
+    \l_fp_input_a_exponent_int \c_zero
+    \l_fp_input_b_integer_int  \l_fp_output_integer_int
+    \l_fp_input_b_decimal_int  \l_fp_output_decimal_int
+    \l_fp_input_b_exponent_int \l_fp_output_exponent_int
+    \fp_div_internal:
+  }
+\cs_new_protected_nopar:Npn \fp_pow_aux_iv:
+  {
+    \group_begin:
+      \l_fp_input_a_integer_int  \l_fp_input_b_integer_int
+      \l_fp_input_a_decimal_int  \l_fp_input_b_decimal_int
+      \l_fp_input_a_exponent_int \l_fp_input_b_exponent_int
+      \fp_ln_internal:
+      \cs_set_protected_nopar:Npx \fp_tmp:w
+        {
+          \group_end:
+          \exp_not:N \l_fp_input_b_sign_int
+            \int_use:N \l_fp_output_sign_int \scan_stop:
+          \exp_not:N \l_fp_input_b_integer_int
+            \int_use:N \l_fp_output_integer_int \scan_stop:
+          \exp_not:N \l_fp_input_b_decimal_int
+            \int_use:N \l_fp_output_decimal_int \scan_stop:
+          \exp_not:N \l_fp_input_b_extended_int
+            \int_use:N \l_fp_output_extended_int \scan_stop:
+          \exp_not:N \l_fp_input_b_exponent_int
+            \int_use:N \l_fp_output_exponent_int \scan_stop:
+        }
+    \fp_tmp:w
+    \l_fp_input_a_extended_int  \c_zero
+    \fp_mul:NNNNNNNNN
+      \l_fp_input_a_integer_int \l_fp_input_a_decimal_int
+        \l_fp_input_a_extended_int
+      \l_fp_input_b_integer_int \l_fp_input_b_decimal_int
+        \l_fp_input_b_extended_int
+      \l_fp_output_integer_int \l_fp_output_decimal_int
+        \l_fp_output_extended_int
+    \l_fp_output_exponent_int
+      \int_eval:w
+        \l_fp_input_a_exponent_int + \l_fp_input_b_exponent_int
+      \scan_stop:
+    \fp_extended_normalise_output:
+    \tex_multiply:D \l_fp_input_a_sign_int \l_fp_input_b_sign_int
+    \l_fp_input_a_integer_int  \l_fp_output_integer_int
+    \l_fp_input_a_decimal_int  \l_fp_output_decimal_int
+    \l_fp_input_a_extended_int \l_fp_output_extended_int
+    \l_fp_input_a_exponent_int \l_fp_output_exponent_int
+    \l_fp_output_integer_int  \c_zero
+    \l_fp_output_decimal_int  \c_zero
+    \l_fp_output_extended_int \c_zero
+    \l_fp_output_exponent_int \c_zero
+    \cs_set_eq:NN \fp_exp_const:Nx \use_none:nn
+    \fp_exp_internal:
+  }
+\prg_new_conditional:Npnn \fp_if_undefined:N #1 { p , T , F , TF }
+  {
+    \if_meaning:w #1 \c_undefined_fp
+      \prg_return_true:
+    \else:
+      \prg_return_false:
+    \fi:
+  }
+\prg_new_conditional:Npnn \fp_if_zero:N #1 { p , T , F , TF }
+  {
+    \if_meaning:w #1 \c_zero_fp
+      \prg_return_true:
+    \else:
+      \prg_return_false:
+    \fi:
+  }
+\prg_new_protected_conditional:Npnn \fp_compare:nNn #1#2#3 { T , F , TF }
+  {
+    \group_begin:
+      \fp_split:Nn a {#1}
+      \fp_standardise:NNNN
+        \l_fp_input_a_sign_int
+        \l_fp_input_a_integer_int
+        \l_fp_input_a_decimal_int
+        \l_fp_input_a_exponent_int
+      \fp_split:Nn b {#3}
+      \fp_standardise:NNNN
+        \l_fp_input_b_sign_int
+        \l_fp_input_b_integer_int
+        \l_fp_input_b_decimal_int
+        \l_fp_input_b_exponent_int
+      \fp_compare_aux:N #2
+  }
+\prg_new_protected_conditional:Npnn \fp_compare:NNN #1#2#3 { T , F , TF }
+  {
+    \group_begin:
+      \fp_read:N #3
+      \l_fp_input_b_sign_int     \l_fp_input_a_sign_int
+      \l_fp_input_b_integer_int  \l_fp_input_a_integer_int
+      \l_fp_input_b_decimal_int  \l_fp_input_a_decimal_int
+      \l_fp_input_b_exponent_int \l_fp_input_a_exponent_int
+      \fp_read:N #1
+      \fp_compare_aux:N #2
+  }
+\cs_new_protected_nopar:Npn \fp_compare_aux:N #1
+  {
+    \cs_if_exist:cTF { fp_compare_#1: }
+      { \use:c { fp_compare_#1: } }
+      {
+        \group_end:
+        \prg_return_false:
+      }
+  }
+\cs_new_protected_nopar:cpn { fp_compare_=: }
+  {
+    \if_int_compare:w \l_fp_input_a_sign_int = \l_fp_input_b_sign_int
+      \if_int_compare:w \l_fp_input_a_integer_int = \l_fp_input_b_integer_int
+        \if_int_compare:w \l_fp_input_a_decimal_int = \l_fp_input_b_decimal_int
+          \if_int_compare:w
+            \l_fp_input_a_exponent_int = \l_fp_input_b_exponent_int
+            \group_end:
+            \prg_return_true:
+          \else:
+            \group_end:
+            \prg_return_false:
+          \fi:
+        \else:
+          \group_end:
+          \prg_return_false:
+        \fi:
+      \else:
+        \group_end:
+        \prg_return_false:
+      \fi:
+    \else:
+      \group_end:
+      \prg_return_false:
+    \fi:
+  }
+\cs_new_protected_nopar:cpn { fp_compare_>: }
+  {
+    \if_int_compare:w \int_eval:w
+      \l_fp_input_a_integer_int + \l_fp_input_a_decimal_int
+      = \c_zero
+      \if_int_compare:w \int_eval:w
+        \l_fp_input_b_integer_int + \l_fp_input_b_decimal_int
+        = \c_zero
+        \group_end:
+        \prg_return_false:
+      \else:
+        \if_int_compare:w \l_fp_input_b_sign_int > \c_zero
+          \group_end:
+          \prg_return_false:
+        \else:
+          \group_end:
+          \prg_return_true:
+        \fi:
+      \fi:
+    \else:
+      \if_int_compare:w \int_eval:w
+        \l_fp_input_b_integer_int + \l_fp_input_b_decimal_int
+        = \c_zero
+        \if_int_compare:w \l_fp_input_a_sign_int > \c_zero
+          \group_end:
+          \prg_return_true:
+        \else:
+          \group_end:
+          \prg_return_false:
+        \fi:
+      \else:
+        \use:c { fp_compare_>_aux: }
+      \fi:
+    \fi:
+  }
+\cs_new_protected_nopar:cpn { fp_compare_>_aux: }
+  {
+    \if_int_compare:w \l_fp_input_a_sign_int > \l_fp_input_b_sign_int
+      \group_end:
+      \prg_return_true:
+    \else:
+      \if_int_compare:w \l_fp_input_a_sign_int < \l_fp_input_b_sign_int
+        \group_end:
+        \prg_return_false:
+      \else:
+        \if_int_compare:w \l_fp_input_a_sign_int > \c_zero
+          \use:c { fp_compare_absolute_a>b: }
+        \else:
+          \use:c { fp_compare_absolute_a<b: }
+        \fi:
+      \fi:
+    \fi:
+  }
+\cs_new_protected_nopar:cpn { fp_compare_absolute_a>b: }
+  {
+    \if_int_compare:w \l_fp_input_a_exponent_int > \l_fp_input_b_exponent_int
+      \group_end:
+      \prg_return_true:
+    \else:
+      \if_int_compare:w \l_fp_input_a_exponent_int < \l_fp_input_b_exponent_int
+        \group_end:
+        \prg_return_false:
+      \else:
+        \if_int_compare:w \l_fp_input_a_integer_int > \l_fp_input_b_integer_int
+          \group_end:
+          \prg_return_true:
+        \else:
+          \if_int_compare:w
+            \l_fp_input_a_integer_int < \l_fp_input_b_integer_int
+            \group_end:
+            \prg_return_false:
+          \else:
+            \if_int_compare:w
+              \l_fp_input_a_decimal_int > \l_fp_input_b_decimal_int
+              \group_end:
+              \prg_return_true:
+            \else:
+              \group_end:
+              \prg_return_false:
+            \fi:
+          \fi:
+        \fi:
+      \fi:
+    \fi:
+  }
+\cs_new_protected_nopar:cpn { fp_compare_absolute_a<b: }
+  {
+    \if_int_compare:w \l_fp_input_b_exponent_int > \l_fp_input_a_exponent_int
+      \group_end:
+      \prg_return_true:
+    \else:
+      \if_int_compare:w \l_fp_input_b_exponent_int < \l_fp_input_a_exponent_int
+        \group_end:
+        \prg_return_false:
+      \else:
+        \if_int_compare:w \l_fp_input_b_integer_int > \l_fp_input_a_integer_int
+          \group_end:
+          \prg_return_true:
+        \else:
+          \if_int_compare:w
+            \l_fp_input_b_integer_int < \l_fp_input_a_integer_int
+            \group_end:
+            \prg_return_false:
+          \else:
+            \if_int_compare:w
+              \l_fp_input_b_decimal_int > \l_fp_input_a_decimal_int
+              \group_end:
+              \prg_return_true:
+            \else:
+              \group_end:
+              \prg_return_false:
+            \fi:
+          \fi:
+        \fi:
+      \fi:
+    \fi:
+  }
+\cs_new_protected_nopar:cpn { fp_compare_<: }
+  {
+    \tl_set:Nx \l_fp_tmp_tl
+      {
+        \int_set:Nn \exp_not:N \l_fp_input_a_sign_int
+          { \int_use:N \l_fp_input_b_sign_int }
+        \int_set:Nn \exp_not:N \l_fp_input_a_integer_int
+          { \int_use:N \l_fp_input_b_integer_int }
+        \int_set:Nn \exp_not:N \l_fp_input_a_decimal_int
+          { \int_use:N \l_fp_input_b_decimal_int }
+        \int_set:Nn \exp_not:N \l_fp_input_a_exponent_int
+          { \int_use:N \l_fp_input_b_exponent_int }
+        \int_set:Nn \exp_not:N \l_fp_input_b_sign_int
+          { \int_use:N \l_fp_input_a_sign_int }
+        \int_set:Nn \exp_not:N \l_fp_input_b_integer_int
+          { \int_use:N \l_fp_input_a_integer_int }
+        \int_set:Nn \exp_not:N \l_fp_input_b_decimal_int
+          { \int_use:N \l_fp_input_a_decimal_int }
+        \int_set:Nn \exp_not:N \l_fp_input_b_exponent_int
+          { \int_use:N \l_fp_input_a_exponent_int }
+      }
+    \l_fp_tmp_tl
+    \use:c { fp_compare_>: }
+  }
+\prg_new_protected_conditional:Npnn \fp_compare:n #1 { T , F , TF }
+  {
+    \group_begin:
+      \tl_set:Nx \l_fp_tmp_tl
+        {
+          \group_end:
+          \fp_compare_aux_i:w #1 \exp_not:n { == \q_nil == \q_stop }
+        }
+     \l_fp_tmp_tl
+  }
+\cs_new_protected_nopar:Npn \fp_compare_aux_i:w #1 == #2 == #3 \q_stop
+  {
+    \quark_if_nil:nTF {#2}
+      { \fp_compare_aux_ii:w #1 != \q_nil != \q_stop }
+      { \fp_compare:nNnTF {#1} = {#2} \prg_return_true: \prg_return_false: }
+  }
+\cs_new_protected_nopar:Npn \fp_compare_aux_ii:w #1 != #2 != #3 \q_stop
+  {
+    \quark_if_nil:nTF {#2}
+      { \fp_compare_aux_iii:w #1 <= \q_nil <= \q_stop }
+      { \fp_compare:nNnTF {#1} = {#2} \prg_return_false: \prg_return_true: }
+  }
+\cs_new_protected_nopar:Npn \fp_compare_aux_iii:w #1 <= #2 <= #3 \q_stop
+  {
+    \quark_if_nil:nTF {#2}
+      { \fp_compare_aux_iv:w #1 >= \q_nil >= \q_stop }
+      { \fp_compare:nNnTF {#1} > {#2} \prg_return_false: \prg_return_true: }
+  }
+\cs_new_protected_nopar:Npn \fp_compare_aux_iv:w #1 >= #2 >= #3 \q_stop
+  {
+    \quark_if_nil:nTF {#2}
+      { \fp_compare_aux_v:w #1 = \q_nil  \q_stop }
+      { \fp_compare:nNnTF {#1} < {#2} \prg_return_false: \prg_return_true: }
+  }
+\cs_new_protected_nopar:Npn \fp_compare_aux_v:w #1 = #2 = #3 \q_stop
+  {
+    \quark_if_nil:nTF {#2}
+      { \fp_compare_aux_vi:w #1 < \q_nil < \q_stop }
+      { \fp_compare:nNnTF {#1} = {#2} \prg_return_true: \prg_return_false: }
+  }
+\cs_new_protected_nopar:Npn \fp_compare_aux_vi:w #1 < #2 < #3 \q_stop
+  {
+    \quark_if_nil:nTF {#2}
+      { \fp_compare_aux_vii:w #1 > \q_nil > \q_stop }
+      { \fp_compare:nNnTF {#1} < {#2} \prg_return_true: \prg_return_false: }
+  }
+\cs_new_protected_nopar:Npn \fp_compare_aux_vii:w #1 > #2 > #3 \q_stop
+  {
+    \quark_if_nil:nTF {#2}
+      { \prg_return_false: }
+      { \fp_compare:nNnTF {#1} > {#2} \prg_return_true: \prg_return_false: }
+  }
+\msg_kernel_new:nnnn { fpu } { overflow }
+  { Number~too~big. }
+  {
+    The~input~given~is~too~big~for~the~LaTeX~floating~point~unit. \\
+    Further~errors~may~well~occur!
+  }
+\cs_new_protected_nopar:Npn \fp_overflow_msg:
+  { \msg_kernel_error:nn { fpu } { overflow } }
+\msg_kernel_new:nnnn { fpu } { exponent-overflow }
+  { Number~too~big~for~exponent~unit. }
+  {
+    The~exponent~of~the~input~given~is~too~big~for~the~floating~point~
+    unit:~the~maximum~input~value~for~an~exponent~is~230.
+  }
+\cs_new_protected_nopar:Npn \fp_exp_overflow_msg:
+  { \msg_kernel_error:nn { fpu } { exponent-overflow } }
+\msg_kernel_new:nnnn { fpu } { logarithm-input-error }
+  { Invalid~input~to~ln~function. }
+  { Logarithms~can~only~be~calculated~for~positive~numbers. }
+\cs_new_protected_nopar:Npn \fp_ln_error_msg: {
+  \msg_kernel_error:nn { fpu } { logarithm-input-error }
+}
+\msg_kernel_new:nnnn { fpu } { trigonometric-overflow }
+  { Number~too~big~for~trigonometry~unit. }
+  {
+    The~trigonometry~code~can~only~work~with~numbers~smaller~
+    than~1000000000.
+  }
+\cs_new_protected_nopar:Npn \fp_trig_overflow_msg:
+  { \msg_kernel_error:nn { fpu } { trigonometric-overflow } }
+%% 
+%%
+%% End of file `l3fp.sty'.
diff --git a/Master/texmf-dist/tex/latex/l3kernel/l3int.sty b/Master/texmf-dist/tex/latex/l3kernel/l3int.sty
new file mode 100644
index 00000000000..7f009832fcf
--- /dev/null
+++ b/Master/texmf-dist/tex/latex/l3kernel/l3int.sty
@@ -0,0 +1,743 @@
+%%
+%% This is file `l3int.sty',
+%% generated with the docstrip utility.
+%%
+%% The original source files were:
+%%
+%% l3int.dtx  (with options: `package')
+%% 
+%% EXPERIMENTAL CODE
+%% 
+%% Do not distribute this file without also distributing the
+%% source files specified above.
+%% 
+%% Do not distribute a modified version of this file.
+%% 
+%% File: l3int.dtx Copyright (C) 1990-2011 The LaTeX3 Project
+%%
+%% It may be distributed and/or modified under the conditions of the
+%% LaTeX Project Public License (LPPL), either version 1.3c of this
+%% license or (at your option) any later version.  The latest version
+%% of this license is in the file
+%%
+%%    http://www.latex-project.org/lppl.txt
+%%
+%% This file is part of the "expl3 bundle" (The Work in LPPL)
+%% and all files in that bundle must be distributed together.
+%%
+%% The released version of this bundle is available from CTAN.
+%%
+%% -----------------------------------------------------------------------
+%%
+%% The development version of the bundle can be found at
+%%
+%%    http://www.latex-project.org/svnroot/experimental/trunk/
+%%
+%% for those people who are interested.
+%%
+%%%%%%%%%%%
+%% NOTE: %%
+%%%%%%%%%%%
+%%
+%%   Snapshots taken from the repository represent work in progress and may
+%%   not work or may contain conflicting material!  We therefore ask
+%%   people _not_ to put them into distributions, archives, etc. without
+%%   prior consultation with the LaTeX3 Project.
+%%
+%% -----------------------------------------------------------------------
+\RequirePackage{l3names}
+\GetIdInfo$Id: l3int.dtx 2478 2011-06-19 21:34:23Z joseph $
+  {L3 Experimental integers}
+\ProvidesExplPackage
+  {\ExplFileName}{\ExplFileDate}{\ExplFileVersion}{\ExplFileDescription}
+\package_check_loaded_expl:
+\cs_set_eq:NN \int_value:w      \tex_number:D
+\cs_set_eq:NN \int_eval:w       \etex_numexpr:D
+\cs_new_eq:NN \int_eval_end:    \tex_relax:D
+\cs_new_eq:NN \if_num:w         \tex_ifnum:D
+\cs_set_eq:NN \if_int_odd:w     \tex_ifodd:D
+\cs_new_eq:NN \if_case:w        \tex_ifcase:D
+\cs_new:Npn \int_eval:n #1
+  { \int_value:w \int_eval:w #1 \int_eval_end: }
+\cs_new:Npn \int_abs:n #1
+  {
+    \int_value:w
+      \if_int_compare:w \int_eval:w #1 < \c_zero
+        -
+      \fi:
+    \int_eval:w #1 \int_eval_end:
+  }
+\cs_new:Npn \int_max:nn #1#2
+  {
+    \int_value:w \int_eval:w
+      \if_int_compare:w
+        \int_eval:w #1 > \int_eval:w #2 \int_eval_end:
+        #1
+      \else:
+        #2
+      \fi:
+    \int_eval_end:
+  }
+\cs_new:Npn \int_min:nn #1#2
+  {
+    \int_value:w \int_eval:w
+      \if_int_compare:w
+        \int_eval:w #1 < \int_eval:w #2 \int_eval_end:
+        #1
+      \else:
+        #2
+      \fi:
+    \int_eval_end:
+  }
+\cs_new:Npn \int_div_truncate:nn #1#2
+  {
+    \int_value:w \int_eval:w
+      \if_int_compare:w \int_eval:w #1 = \c_zero
+        0
+      \else:
+        ( #1 % )
+        \if_int_compare:w \int_eval:w #1 < \c_zero
+          \if_int_compare:w \int_eval:w #2 < \c_zero
+            - ( #2 + % )
+          \else:
+            + ( #2 - % )
+          \fi:
+        \else:
+          \if_int_compare:w \int_eval:w #2 < \c_zero
+            + ( #2 + % )
+          \else:
+            - ( #2 - % )
+          \fi:
+        \fi: % ( (
+        1 ) / 2 )
+      \fi:
+      / ( #2 )
+    \int_eval_end:
+  }
+\cs_new:Npn \int_div_round:nn #1#2 { \int_eval:n { ( #1 ) / ( #2 ) } }
+\cs_new:Npn \int_mod:nn #1#2
+  {
+    \int_value:w \int_eval:w
+      #1 - \int_div_truncate:nn {#1} {#2} * ( #2 )
+    \int_eval_end:
+  }
+\cs_new_protected_nopar:Npn \int_new:N #1
+  {
+    \chk_if_free_cs:N #1
+    \newcount #1
+  }
+\cs_generate_variant:Nn \int_new:N { c }
+\cs_new_protected_nopar:Npn \int_const:Nn #1#2
+  {
+    \int_compare:nNnTF {#2} > \c_minus_one
+      {
+        \int_compare:nNnTF {#2} > \c_max_register_int
+          {
+            \int_new:N #1
+            \int_gset:Nn #1 {#2}
+          }
+          {
+            \chk_if_free_cs:N #1
+            \pref_global:D \tex_mathchardef:D #1 =
+              \int_eval:w #2 \int_eval_end:
+          }
+      }
+      {
+        \int_new:N #1
+        \int_gset:Nn #1 {#2}
+      }
+  }
+\cs_generate_variant:Nn \int_const:Nn { c }
+\cs_new_protected_nopar:Npn \int_zero:N  #1 { #1 = \c_zero }
+\cs_new_protected_nopar:Npn \int_gzero:N #1 { \pref_global:D #1 = \c_zero }
+\cs_generate_variant:Nn \int_zero:N  { c }
+\cs_generate_variant:Nn \int_gzero:N { c }
+\cs_new_protected_nopar:Npn \int_set_eq:NN #1#2 { #1 = #2 }
+\cs_generate_variant:Nn \int_set_eq:NN {       c }
+\cs_generate_variant:Nn \int_set_eq:NN { Nc , cc }
+\cs_new_protected_nopar:Npn \int_gset_eq:NN #1#2 { \pref_global:D #1 = #2 }
+\cs_generate_variant:Nn \int_gset_eq:NN {       c }
+\cs_generate_variant:Nn \int_gset_eq:NN { Nc , cc }
+\cs_new_protected_nopar:Npn \int_add:Nn #1#2
+  { \tex_advance:D #1 by \int_eval:w #2 \int_eval_end: }
+\cs_new_nopar:Npn \int_sub:Nn #1#2
+  { \tex_advance:D #1 by - \int_eval:w #2 \int_eval_end: }
+\cs_new_protected_nopar:Npn \int_gadd:Nn
+  { \pref_global:D \int_add:Nn }
+\cs_new_protected_nopar:Npn \int_gsub:Nn
+  { \pref_global:D \int_sub:Nn }
+\cs_generate_variant:Nn \int_add:Nn  { c }
+\cs_generate_variant:Nn \int_gadd:Nn { c }
+\cs_generate_variant:Nn \int_sub:Nn  { c }
+\cs_generate_variant:Nn \int_gsub:Nn { c }
+\cs_new_protected_nopar:Npn \int_incr:N #1
+  { \tex_advance:D #1 \c_one }
+\cs_new_protected_nopar:Npn \int_decr:N #1
+  { \tex_advance:D #1 \c_minus_one }
+\cs_new_protected_nopar:Npn \int_gincr:N
+  { \pref_global:D \int_incr:N }
+\cs_new_protected_nopar:Npn \int_gdecr:N
+  { \pref_global:D \int_decr:N }
+\cs_generate_variant:Nn \int_incr:N  { c }
+\cs_generate_variant:Nn \int_decr:N  { c }
+\cs_generate_variant:Nn \int_gincr:N { c }
+\cs_generate_variant:Nn \int_gdecr:N { c }
+\cs_new_protected_nopar:Npn \int_set:Nn #1#2
+  { #1 ~ \int_eval:w #2\int_eval_end: }
+\cs_new_protected_nopar:Npn \int_gset:Nn { \pref_global:D \int_set:Nn }
+\cs_generate_variant:Nn \int_set:Nn  { c }
+\cs_generate_variant:Nn \int_gset:Nn { c }
+\cs_new_eq:NN \int_use:N \tex_the:D
+\cs_new_nopar:Npn \int_use:c #1 { \int_use:N \cs:w #1 \cs_end: }
+\prg_new_conditional:Npnn \int_compare:n #1 { p , T , F , TF }
+  { \exp_after:wN \int_compare_aux:nw \int_value:w \int_eval:w #1 \q_stop }
+\cs_new:Npn \int_compare_aux:nw #1#2 \q_stop
+  {
+    \exp_after:wN \int_compare_aux:Nw
+      \int_to_roman:w
+        \if:w #1 -
+        \else:
+          -
+        \fi:
+          #1#2 \q_mark #1#2 \q_stop
+  }
+\cs_new:Npn \int_compare_aux:Nw #1#2#3 \q_mark
+  { \use:c { int_compare_  #1  \if_meaning:w = #2 =  \fi: :w } }
+\cs_new:cpn { int_compare_=:w } #1 = #2 \q_stop
+  {
+    \if_int_compare:w #1 = \int_eval:w #2 \int_eval_end:
+      \prg_return_true:
+    \else:
+      \prg_return_false:
+    \fi:
+  }
+\cs_new:cpn { int_compare_==:w } #1 == #2 \q_stop
+  {
+    \if_int_compare:w #1 = \int_eval:w #2 \int_eval_end:
+      \prg_return_true:
+    \else:
+      \prg_return_false:
+    \fi:
+  }
+\cs_new:cpn { int_compare_!=:w } #1 != #2 \q_stop
+  {
+    \if_int_compare:w #1 = \int_eval:w #2 \int_eval_end:
+      \prg_return_false:
+    \else:
+      \prg_return_true:
+    \fi:
+  }
+\cs_new:cpn { int_compare_<:w } #1 < #2 \q_stop
+  {
+    \if_int_compare:w #1 < \int_eval:w #2 \int_eval_end:
+      \prg_return_true:
+    \else:
+      \prg_return_false:
+    \fi:
+  }
+\cs_new:cpn { int_compare_>:w } #1 > #2 \q_stop
+  {
+    \if_int_compare:w #1 > \int_eval:w #2 \int_eval_end:
+      \prg_return_true:
+    \else:
+      \prg_return_false:
+    \fi:
+  }
+\cs_new:cpn { int_compare_<=:w } #1 <= #2 \q_stop
+  {
+    \if_int_compare:w #1 > \int_eval:w #2 \int_eval_end:
+      \prg_return_false:
+    \else:
+      \prg_return_true:
+    \fi:
+  }
+\cs_new:cpn { int_compare_>=:w } #1 >= #2 \q_stop
+  {
+    \if_int_compare:w #1 < \int_eval:w #2 \int_eval_end:
+      \prg_return_false:
+    \else:
+      \prg_return_true:
+    \fi:
+  }
+\prg_new_conditional:Npnn \int_compare:nNn #1#2#3 { p , T , F , TF}
+  {
+    \if_int_compare:w \int_eval:w #1 #2 \int_eval:w #3 \int_eval_end:
+      \prg_return_true:
+    \else:
+      \prg_return_false:
+    \fi:
+  }
+\prg_new_conditional:Npnn \int_if_odd:n #1 { p , T , F , TF}
+  {
+    \if_int_odd:w \int_eval:w #1 \int_eval_end:
+      \prg_return_true:
+    \else:
+      \prg_return_false:
+    \fi:
+  }
+\prg_new_conditional:Npnn \int_if_even:n #1 { p , T , F , TF}
+  {
+    \if_int_odd:w \int_eval:w #1 \int_eval_end:
+      \prg_return_false:
+    \else:
+      \prg_return_true:
+    \fi:
+  }
+\cs_new:Npn \int_while_do:nn #1#2
+  {
+    \int_compare:nT {#1}
+      {
+        #2
+        \int_while_do:nn {#1} {#2}
+      }
+  }
+\cs_new:Npn \int_until_do:nn #1#2
+  {
+    \int_compare:nF {#1}
+      {
+        #2
+        \int_until_do:nn {#1} {#2}
+      }
+  }
+\cs_new:Npn \int_do_while:nn #1#2
+  {
+    #2
+    \int_compare:nT {#1}
+      { \int_do_while:nNnn {#1} {#2} }
+  }
+\cs_new:Npn \int_do_until:nn #1#2
+  {
+    #2
+    \int_compare:nF {#1}
+      { \int_do_until:nn {#1} {#2} }
+  }
+\cs_new:Npn \int_while_do:nNnn #1#2#3#4
+  {
+    \int_compare:nNnT {#1} #2 {#3}
+      {
+        #4
+        \int_while_do:nNnn {#1} #2 {#3} {#4}
+      }
+  }
+\cs_new:Npn \int_until_do:nNnn #1#2#3#4
+  {
+  \int_compare:nNnF {#1} #2 {#3}
+    {
+      #4
+      \int_until_do:nNnn {#1} #2 {#3} {#4}
+    }
+  }
+\cs_new:Npn \int_do_while:nNnn #1#2#3#4
+  {
+    #4
+    \int_compare:nNnT {#1} #2 {#3}
+      { \int_do_while:nNnn {#1} #2 {#3} {#4} }
+  }
+\cs_new:Npn \int_do_until:nNnn #1#2#3#4
+  {
+    #4
+    \int_compare:nNnF {#1} #2 {#3}
+      { \int_do_until:nNnn {#1} #2 {#3} {#4} }
+  }
+\cs_new_nopar:Npn \int_to_arabic:n #1 { \int_eval:n {#1} }
+\cs_new_nopar:Npn \int_to_symbols:nnn #1#2#3
+  {
+    \int_compare:nNnTF {#1} > {#2}
+      {
+        \exp_args:Nf \int_to_symbols:nnn
+          { \int_div_truncate:nn { #1 - 1 } {#2} } {#2} {#3}
+        \exp_args:Nf \prg_case_int:nnn
+          { \int_eval:n { 1 + \int_mod:nn { #1 - 1 } {#2} } }
+          {#3} { }
+      }
+      { \exp_args:Nf \prg_case_int:nnn { \int_eval:n {#1} } {#3} { } }
+  }
+\cs_new:Npn \int_to_alph:n #1
+  {
+    \int_to_symbols:nnn {#1} { 26 }
+      {
+        {  1 } { a }
+        {  2 } { b }
+        {  3 } { c }
+        {  4 } { d }
+        {  5 } { e }
+        {  6 } { f }
+        {  7 } { g }
+        {  8 } { h }
+        {  9 } { i }
+        { 10 } { j }
+        { 11 } { k }
+        { 12 } { l }
+        { 13 } { m }
+        { 14 } { n }
+        { 15 } { o }
+        { 16 } { p }
+        { 17 } { q }
+        { 18 } { r }
+        { 19 } { s }
+        { 20 } { t }
+        { 21 } { u }
+        { 22 } { v }
+        { 23 } { w }
+        { 24 } { x }
+        { 25 } { y }
+        { 26 } { z }
+      }
+  }
+\cs_new:Npn \int_to_Alph:n #1
+  {
+    \int_to_symbols:nnn {#1} { 26 }
+      {
+        {  1 } { A }
+        {  2 } { B }
+        {  3 } { C }
+        {  4 } { D }
+        {  5 } { E }
+        {  6 } { F }
+        {  7 } { G }
+        {  8 } { H }
+        {  9 } { I }
+        { 10 } { J }
+        { 11 } { K }
+        { 12 } { L }
+        { 13 } { M }
+        { 14 } { N }
+        { 15 } { O }
+        { 16 } { P }
+        { 17 } { Q }
+        { 18 } { R }
+        { 19 } { S }
+        { 20 } { T }
+        { 21 } { U }
+        { 22 } { V }
+        { 23 } { W }
+        { 24 } { X }
+        { 25 } { Y }
+        { 26 } { Z }
+      }
+  }
+\cs_new:Npn \int_to_base:nn #1#2
+  {
+    \int_compare:nNnTF {#1} < \c_zero
+      {
+        -
+        \exp_args:Nnf \int_to_base_aux:nnn
+          { } { \int_eval:n { 0 - ( #1 ) } } {#2}
+      }
+      {
+      \exp_args:Nnf \int_to_base_aux:nnn
+        { } { \int_eval:n {#1} } {#2}
+    }
+}
+\cs_new:Npn \int_to_base_aux:nnn #1#2#3 {
+  \int_compare:nNnTF {#2} < {#3}
+    {
+      \int_to_letter:n {#2}
+      #1
+    }
+    {
+      \exp_args:Nff \int_to_base_aux:nnn
+        {
+          \int_to_letter:n { \int_mod:nn {#2} {#3} }
+          #1
+        }
+        { \int_div_truncate:nn {#2} {#3} }
+        {#3}
+    }
+  }
+\cs_new:Npn \int_to_letter:n #1
+  {
+    \prg_case_int:nnn { #1 - 9 }
+      {
+        {  1 } { A }
+        {  2 } { B }
+        {  3 } { C }
+        {  4 } { D }
+        {  5 } { E }
+        {  6 } { F }
+        {  7 } { G }
+        {  8 } { H }
+        {  9 } { I }
+        { 10 } { J }
+        { 11 } { K }
+        { 12 } { L }
+        { 13 } { M }
+        { 14 } { N }
+        { 15 } { O }
+        { 16 } { P }
+        { 17 } { Q }
+        { 18 } { R }
+        { 19 } { S }
+        { 20 } { T }
+        { 21 } { U }
+        { 22 } { V }
+        { 23 } { W }
+        { 24 } { X }
+        { 25 } { Y }
+        { 26 } { Z }
+      }
+      {#1}
+  }
+\cs_new:Npn \int_to_binary:n #1
+  { \int_to_base:nn {#1} { 2 } }
+\cs_new:Npn \int_to_hexadecimal:n #1
+  { \int_to_base:nn {#1} { 16 } }
+\cs_new:Npn \int_to_octal:n #1
+  { \int_to_base:nn {#1} { 8 } }
+\cs_new_nopar:Npn \int_to_roman:n #1
+  {
+    \exp_after:wN \int_to_roman_aux:N
+      \int_to_roman:w \int_eval:n {#1} Q
+  }
+\cs_new_nopar:Npn \int_to_roman_aux:N #1
+  {
+    \use:c { int_to_roman_ #1 :w }
+    \int_to_roman_aux:N
+  }
+\cs_new_nopar:Npn \int_to_Roman:n #1
+  {
+    \exp_after:wN \int_to_Roman_aux:N
+      \int_to_roman:w \int_eval:n {#1} Q
+  }
+\cs_new_nopar:Npn \int_to_Roman_aux:N #1
+  {
+    \use:c { int_to_Roman_ #1 :w }
+    \int_to_Roman_aux:N
+  }
+\cs_new_nopar:Npn \int_to_roman_i:w { i }
+\cs_new_nopar:Npn \int_to_roman_v:w { v }
+\cs_new_nopar:Npn \int_to_roman_x:w { x }
+\cs_new_nopar:Npn \int_to_roman_l:w { l }
+\cs_new_nopar:Npn \int_to_roman_c:w { c }
+\cs_new_nopar:Npn \int_to_roman_d:w { d }
+\cs_new_nopar:Npn \int_to_roman_m:w { m }
+\cs_new_nopar:Npn \int_to_roman_Q:w #1 { }
+\cs_new_nopar:Npn \int_to_Roman_i:w { I }
+\cs_new_nopar:Npn \int_to_Roman_v:w { V }
+\cs_new_nopar:Npn \int_to_Roman_x:w { X }
+\cs_new_nopar:Npn \int_to_Roman_l:w { L }
+\cs_new_nopar:Npn \int_to_Roman_c:w { C }
+\cs_new_nopar:Npn \int_to_Roman_d:w { D }
+\cs_new_nopar:Npn \int_to_Roman_m:w { M }
+\cs_new_nopar:Npn \int_to_Roman_Q:w #1 { }
+\cs_new:Npn \int_get_sign:n #1
+  {
+    \int_get_sign_and_digits_aux:nNNN {#1}
+      \c_true_bool \c_true_bool \c_false_bool
+  }
+\cs_new:Npn \int_get_digits:n #1
+  {
+    \int_get_sign_and_digits_aux:nNNN {#1}
+      \c_true_bool \c_false_bool \c_true_bool
+  }
+\cs_new:Npn \int_get_sign_and_digits_aux:nNNN #1#2#3#4
+  {
+    \tl_if_head_eq_charcode:fNTF {#1} -
+      {
+        \bool_if:NTF #2
+          {
+            \int_get_sign_and_digits_aux:oNNN
+              { \use_none:n #1 } \c_false_bool #3#4
+          }
+          {
+            \int_get_sign_and_digits_aux:oNNN
+              { \use_none:n #1 } \c_true_bool #3#4
+          }
+      }
+      {
+        \tl_if_head_eq_charcode:fNTF {#1} +
+          { \int_get_sign_and_digits_aux:oNNN { \use_none:n #1 } #2#3#4 }
+          {
+            \bool_if:NT #3 { \bool_if:NF #2 - }
+            \bool_if:NT #4 {#1}
+          }
+      }
+  }
+\cs_generate_variant:Nn \int_get_sign_and_digits_aux:nNNN { o }
+\cs_new:Npn \int_from_alph:n #1
+  {
+    \int_eval:n
+      {
+        \int_get_sign:n {#1}
+        \exp_args:Nf \int_from_alph_aux:n { \int_get_digits:n {#1} }
+      }
+  }
+\cs_new:Npn \int_from_alph_aux:n #1
+  { \int_from_alph_aux:nN { 0 } #1 \q_nil }
+\cs_new:Npn \int_from_alph_aux:nN #1#2
+  {
+    \quark_if_nil:NTF #2
+      {#1}
+      {
+        \exp_args:Nf \int_from_alph_aux:nN
+          { \int_eval:n { #1 * 26 + \int_from_alph_aux:N #2 } }
+      }
+  }
+\cs_new:Npn \int_from_alph_aux:N #1
+  { \int_eval:n { `#1 - \int_compare:nNnTF { `#1 } < { 91 } { 64 } { 96 } } }
+\cs_new:Npn \int_from_base:nn #1#2
+  {
+    \int_eval:n
+      {
+        \int_get_sign:n {#1}
+        \exp_args:Nf \int_from_base_aux:nn
+          { \int_get_digits:n {#1} } {#2}
+      }
+  }
+\cs_new:Npn \int_from_base_aux:nn #1#2
+  { \int_from_base_aux:nnN { 0 } { #2 } #1 \q_nil }
+\cs_new:Npn \int_from_base_aux:nnN #1#2#3
+  {
+    \quark_if_nil:NTF #3
+      {#1}
+      {
+        \exp_args:Nf \int_from_base_aux:nnN
+          { \int_eval:n { #1 * #2 + \int_from_base_aux:N #3 } }
+          {#2}
+      }
+  }
+\cs_new:Npn \int_from_base_aux:N #1
+  {
+    \int_compare:nNnTF { `#1 } < { 58 }
+      {#1}
+      {
+        \int_eval:n
+          { `#1 - \int_compare:nNnTF { `#1 } < { 91 } { 55 } { 87 } }
+      }
+  }
+\cs_new:Npn \int_from_binary:n #1
+  { \int_from_base:nn {#1} \c_two }
+\cs_new:Npn \int_from_hexadecimal:n #1
+  { \int_from_base:nn {#1} \c_sixteen }
+\cs_new:Npn \int_from_octal:n #1
+  { \int_from_base:nn {#1} \c_eight }
+\int_const:cn { c_int_from_roman_i_int } { 1 }
+\int_const:cn { c_int_from_roman_v_int } { 5 }
+\int_const:cn { c_int_from_roman_x_int } { 10 }
+\int_const:cn { c_int_from_roman_l_int } { 50 }
+\int_const:cn { c_int_from_roman_c_int } { 100 }
+\int_const:cn { c_int_from_roman_d_int } { 500 }
+\int_const:cn { c_int_from_roman_m_int } { 1000 }
+\int_const:cn { c_int_from_roman_I_int } { 1 }
+\int_const:cn { c_int_from_roman_V_int } { 5 }
+\int_const:cn { c_int_from_roman_X_int } { 10 }
+\int_const:cn { c_int_from_roman_L_int } { 50 }
+\int_const:cn { c_int_from_roman_C_int } { 100 }
+\int_const:cn { c_int_from_roman_D_int } { 500 }
+\int_const:cn { c_int_from_roman_M_int } { 1000 }
+\cs_new_nopar:Npn \int_from_roman:n #1
+  {
+    \tl_if_blank:nF {#1}
+      {
+        \exp_after:wN \int_from_roman_end:w
+          \int_value:w \int_eval:w
+            \int_from_roman_aux:NN #1 Q \q_stop
+      }
+  }
+\cs_new_nopar:Npn \int_from_roman_aux:NN #1#2
+  {
+    \str_if_eq:nnTF {#1} { Q }
+      {#1#2}
+      {
+        \str_if_eq:nnTF {#2} { Q }
+          {
+            \cs_if_exist:cF { c_int_from_roman_ #1 _int }
+              { \int_from_roman_clean_up:w }
+            +
+            \use:c { c_int_from_roman_ #1 _int }
+            #2
+          }
+          {
+            \cs_if_exist:cF { c_int_from_roman_ #1 _int }
+              { \int_from_roman_clean_up:w }
+            \cs_if_exist:cF { c_int_from_roman_ #2 _int }
+              { \int_from_roman_clean_up:w }
+            \int_compare:nNnTF
+              { \use:c { c_int_from_roman_ #1 _int } }
+              <
+              { \use:c { c_int_from_roman_ #2 _int  } }
+              {
+                + \use:c { c_int_from_roman_ #2 _int }
+                - \use:c { c_int_from_roman_ #1 _int }
+                \int_from_roman_aux:NN
+              }
+              {
+                + \use:c { c_int_from_roman_ #1 _int }
+                \int_from_roman_aux:NN #2
+              }
+          }
+      }
+  }
+\cs_new_nopar:Npn \int_from_roman_end:w #1 Q #2 \q_stop
+  { \tl_if_empty:nTF {#2} {#1} {#2} }
+\cs_new_nopar:Npn \int_from_roman_clean_up:w #1 Q { + 0 Q -1 }
+\cs_new_eq:NN \int_show:N \kernel_register_show:N
+\cs_new_eq:NN \int_show:c \kernel_register_show:c
+\int_const:Nn \c_one      {  1 }
+\int_const:Nn \c_two      {  2 }
+\int_const:Nn \c_three    {  3 }
+\int_const:Nn \c_four     {  4 }
+\int_const:Nn \c_five     {  5 }
+\int_const:Nn \c_eight    {  8 }
+\int_const:Nn \c_nine     {  9 }
+\int_const:Nn \c_ten      { 10 }
+\int_const:Nn \c_eleven   { 11 }
+\int_const:Nn \c_thirteen { 13 }
+\int_const:Nn \c_fourteen { 14 }
+\int_const:Nn \c_fifteen  { 15 }
+\int_const:Nn \c_thirty_two { 32 }
+\int_const:Nn \c_two_hundred_fifty_five { 255 }
+\int_const:Nn \c_two_hundred_fifty_six  { 256 }
+\int_const:Nn \c_one_hundred  {   100 }
+\int_const:Nn \c_one_thousand {  1000 }
+\int_const:Nn \c_ten_thousand { 10000 }
+\int_const:Nn \c_max_int { 2 147 483 647 }
+\int_new:N \l_tmpa_int
+\int_new:N \l_tmpb_int
+\int_new:N \l_tmpc_int
+\int_new:N \g_tmpa_int
+\int_new:N \g_tmpb_int
+\int_new:N \g_seq_nesting_depth_int
+\int_new:N \g_tl_inline_level_int
+\cs_new_eq:NN \int_convert_from_base_ten:nn \int_to_base:nn
+\cs_new_eq:NN \int_convert_to_symbols:nnn   \int_to_symbols:nnn
+\cs_new_eq:NN \int_convert_to_base_ten:nn   \int_from_base:nn
+\cs_new_nopar:Npn \int_to_symbol:n
+  {
+    \mode_if_math:TF
+      { \int_to_symbol_math:n }
+      { \int_to_symbol_text:n }
+  }
+\cs_new:Npn \int_to_symbol_math:n #1
+  {
+    \int_to_symbols:nnn {#1} { 9 }
+      {
+        { 1 } {                 * }
+        { 2 } {           \dagger }
+        { 3 } {          \ddagger }
+        { 4 } {      \mathsection }
+        { 5 } {    \mathparagraph }
+        { 6 } {                \| }
+        { 7 } {                ** }
+        { 8 } {   \dagger \dagger }
+        { 9 } { \ddagger \ddagger }
+      }
+  }
+\cs_new:Npn \int_to_symbol_text:n #1
+  {
+    \int_to_symbols:nnn {#1} { 9 }
+      {
+        { 1 } {                       \textasteriskcentered }
+        { 2 } {                                 \textdagger }
+        { 3 } {                              \textdaggerdbl }
+        { 4 } {                                \textsection }
+        { 5 } {                              \textparagraph }
+        { 6 } {                                 \textbardbl }
+        { 7 } { \textasteriskcentered \textasteriskcentered }
+        { 8 } {                     \textdagger \textdagger }
+        { 9 } {               \textdaggerdbl \textdaggerdbl }
+      }
+  }
+%% 
+%%
+%% End of file `l3int.sty'.
diff --git a/Master/texmf-dist/tex/latex/l3kernel/l3io.sty b/Master/texmf-dist/tex/latex/l3kernel/l3io.sty
new file mode 100644
index 00000000000..7239fbae288
--- /dev/null
+++ b/Master/texmf-dist/tex/latex/l3kernel/l3io.sty
@@ -0,0 +1,430 @@
+%%
+%% This is file `l3io.sty',
+%% generated with the docstrip utility.
+%%
+%% The original source files were:
+%%
+%% l3io.dtx  (with options: `package')
+%% 
+%% EXPERIMENTAL CODE
+%% 
+%% Do not distribute this file without also distributing the
+%% source files specified above.
+%% 
+%% Do not distribute a modified version of this file.
+%% 
+%% File: l3io.dtx Copyright (C) 1990-2011 The LaTeX3 Project
+%%
+%% It may be distributed and/or modified under the conditions of the
+%% LaTeX Project Public License (LPPL), either version 1.3c of this
+%% license or (at your option) any later version.  The latest version
+%% of this license is in the file
+%%
+%%    http://www.latex-project.org/lppl.txt
+%%
+%% This file is part of the "expl3 bundle" (The Work in LPPL)
+%% and all files in that bundle must be distributed together.
+%%
+%% The released version of this bundle is available from CTAN.
+%%
+%% -----------------------------------------------------------------------
+%%
+%% The development version of the bundle can be found at
+%%
+%%    http://www.latex-project.org/svnroot/experimental/trunk/
+%%
+%% for those people who are interested.
+%%
+%%%%%%%%%%%
+%% NOTE: %%
+%%%%%%%%%%%
+%%
+%%   Snapshots taken from the repository represent work in progress and may
+%%   not work or may contain conflicting material!  We therefore ask
+%%   people _not_ to put them into distributions, archives, etc. without
+%%   prior consultation with the LaTeX3 Project.
+%%
+%% -----------------------------------------------------------------------
+\RequirePackage{l3names}
+\GetIdInfo$Id: l3io.dtx 2478 2011-06-19 21:34:23Z joseph $
+  {L3 Experimental input-output operations}
+\ProvidesExplPackage
+  {\ExplFileName}{\ExplFileDate}{\ExplFileVersion}{\ExplFileDescription}
+\package_check_loaded_expl:
+\cs_new_eq:NN \if_eof:w \tex_ifeof:D
+\cs_new_eq:NN \c_iow_term_stream \c_sixteen
+\cs_new_eq:NN \c_ior_term_stream \c_sixteen
+\cs_new_eq:NN \c_iow_log_stream  \c_minus_one
+\cs_new_eq:NN \c_ior_log_stream  \c_minus_one
+\tl_const:Nn \c_iow_streams_tl
+  {
+    \c_zero
+    \c_one
+    \c_two
+    \c_three
+    \c_four
+    \c_five
+    \c_six
+    \c_seven
+    \c_eight
+    \c_nine
+    \c_ten
+    \c_eleven
+    \c_twelve
+    \c_thirteen
+    \c_fourteen
+    \c_fifteen
+  }
+\cs_new_eq:NN \c_ior_streams_tl \c_iow_streams_tl
+\prop_new:N \g_iow_streams_prop
+\prop_new:N \g_ior_streams_prop
+\prop_put:Nnn \g_iow_streams_prop { 0 } { LaTeX2e~reserved }
+\prop_put:Nnn \g_iow_streams_prop { 1 } { LaTeX2e~reserved }
+\prop_put:Nnn \g_iow_streams_prop { 2 } { LaTeX2e~reserved }
+\prop_put:Nnn \g_ior_streams_prop { 0 } { LaTeX2e~reserved }
+\int_new:N \l_iow_stream_int
+\cs_new_eq:NN \l_ior_stream_int \l_iow_stream_int
+\cs_set_eq:NN \iow_raw_new:N \newwrite
+\cs_set_eq:NN \ior_raw_new:N \newread
+\cs_generate_variant:Nn \ior_raw_new:N { c }
+\cs_generate_variant:Nn \iow_raw_new:N { c }
+\cs_new_protected_nopar:Npn \ior_open:Nn #1#2
+  {
+    \ior_close:N #1
+    \int_set:Nn \l_ior_stream_int \c_sixteen
+    \tl_map_function:NN \c_ior_streams_tl \ior_alloc_read:n
+    \int_compare:nNnTF \l_ior_stream_int = \c_sixteen
+      { \msg_kernel_error:nn { ior } { streams-exhausted } }
+      {
+        \ior_stream_alloc:N #1
+        \prop_gput:NVn \g_ior_streams_prop \l_ior_stream_int {#2}
+        \tex_openin:D #1#2 \scan_stop:
+      }
+  }
+\cs_new_protected_nopar:Npn \iow_open:Nn #1#2
+  {
+    \iow_close:N #1
+    \int_set:Nn \l_iow_stream_int \c_sixteen
+    \tl_map_function:NN \c_iow_streams_tl \iow_alloc_write:n
+    \int_compare:nNnTF \l_iow_stream_int = \c_sixteen
+      { \msg_kernel_error:nn { iow } { streams-exhausted } }
+      {
+        \iow_stream_alloc:N #1
+        \prop_gput:NVn \g_iow_streams_prop \l_iow_stream_int {#2}
+        \tex_immediate:D \tex_openout:D #1#2 \scan_stop:
+      }
+  }
+\cs_generate_variant:Nn \ior_open:Nn { c }
+\cs_generate_variant:Nn \iow_open:Nn { c }
+\cs_new_protected_nopar:Npn \iow_alloc_write:n #1
+  {
+    \prop_if_in:NnF \g_iow_streams_prop {#1}
+      {
+        \int_set:Nn \l_iow_stream_int {#1}
+        \tl_map_break:
+      }
+  }
+\cs_new_protected_nopar:Npn \ior_alloc_read:n #1
+  {
+    \prop_if_in:NnF \g_iow_streams_prop {#1}
+      {
+        \int_set:Nn \l_ior_stream_int {#1}
+        \tl_map_break:
+      }
+  }
+\cs_new_protected_nopar:Npn \iow_stream_alloc:N #1
+  {
+    \cs_if_exist:cTF { g_iow_ \int_use:N \l_iow_stream_int _stream }
+      { \cs_gset_eq:Nc #1 { g_iow_ \int_use:N \l_iow_stream_int _stream } }
+      {
+        \iow_stream_alloc_aux:
+        \int_compare:nNnT \l_iow_stream_int = \c_sixteen
+          {
+            \iow_raw_new:N \g_iow_tmp_stream
+            \int_set:Nn \l_iow_stream_int { \g_iow_tmp_stream }
+            \cs_gset_eq:cN
+              { g_iow_ \int_use:N \l_iow_stream_int _stream }
+              \g_iow_tmp_stream
+          }
+        \cs_gset_eq:Nc #1 { g_iow_ \int_use:N \l_iow_stream_int _stream }
+      }
+  }
+\cs_new_protected_nopar:Npn \iow_stream_alloc_aux:
+  {
+    \int_incr:N \l_iow_stream_int
+    \int_compare:nNnT \l_iow_stream_int < \c_sixteen
+      {
+         \cs_if_exist:cTF { g_iow_ \int_use:N \l_iow_stream_int _stream }
+           {
+             \prop_if_in:NVT \g_iow_streams_prop \l_iow_stream_int
+               { \iow_stream_alloc_aux: }
+           }
+           { \iow_stream_alloc_aux: }
+      }
+  }
+\cs_new_protected_nopar:Npn \ior_stream_alloc:N #1
+  {
+    \cs_if_exist:cTF { g_ior_ \int_use:N \l_ior_stream_int _stream }
+      { \cs_gset_eq:Nc #1 { g_ior_ \int_use:N \l_ior_stream_int _stream } }
+      {
+        \ior_stream_alloc_aux:
+        \int_compare:nNnT \l_ior_stream_int = \c_sixteen
+          {
+            \ior_raw_new:N \g_ior_tmp_stream
+            \int_set:Nn \l_ior_stream_int { \g_ior_tmp_stream }
+            \cs_gset_eq:cN
+              { g_ior_ \int_use:N \l_iow_stream_int _stream }
+              \g_ior_tmp_stream
+          }
+        \cs_gset_eq:Nc #1 { g_ior_ \int_use:N \l_ior_stream_int _stream }
+      }
+  }
+\cs_new_protected_nopar:Npn \ior_stream_alloc_aux:
+  {
+    \int_incr:N \l_ior_stream_int
+    \int_compare:nNnT \l_ior_stream_int < \c_sixteen
+      {
+         \cs_if_exist:cTF { g_ior_ \int_use:N \l_ior_stream_int _stream }
+           {
+             \prop_if_in:NVT \g_ior_streams_prop \l_ior_stream_int
+               { \ior_stream_alloc_aux: }
+           }
+           { \ior_stream_alloc_aux: }
+      }
+  }
+\cs_new_protected_nopar:Npn \ior_close:N #1
+  {
+    \cs_if_exist:NT #1
+      {
+        \int_compare:nNnF #1 = \c_minus_one
+          {
+            \tex_closein:D #1
+            \prop_gdel:NV \g_ior_streams_prop #1
+            \cs_undefine:N #1
+          }
+      }
+  }
+\cs_new_protected_nopar:Npn \iow_close:N #1
+  {
+    \cs_if_exist:NT #1
+      {
+        \int_compare:nNnF #1 = \c_minus_one
+          {
+            \tex_immediate:D \tex_closeout:D #1
+            \prop_gdel:NV \g_iow_streams_prop #1
+            \cs_undefine:N #1
+          }
+      }
+  }
+\cs_generate_variant:Nn \ior_close:N { c }
+\cs_generate_variant:Nn \iow_close:N { c }
+\cs_new_protected_nopar:Npn \ior_list_streams:
+  {
+    \prop_if_empty:NTF \g_ior_streams_prop
+      {
+        \iow_term:x { No~input~streams~are~open }
+        \tl_show:n { }
+      }
+      {
+        \iow_term:x { The~following~input~streams~are~in~use: }
+        \tl_set:Nx \l_prop_show_tl
+          { \prop_map_function:NN \g_ior_streams_prop \ior_show_aux:nn }
+          \etex_showtokens:D \exp_after:wN \exp_after:wN \exp_after:wN
+            { \exp_after:wN \prop_show_aux:w \l_prop_show_tl }
+      }
+  }
+\cs_new:Npn \ior_show_aux:nn #1#2
+  {
+    \iow_newline: > \c_space_tl \c_space_tl
+    #1 \iow_char:N
+    \c_space_tl \c_space_tl => \c_space_tl \c_space_tl
+    \exp_not:n {#2}
+  }
+\cs_new_protected_nopar:Npn \iow_list_streams:
+  {
+    \prop_if_empty:NTF \g_iow_streams_prop
+      {
+        \iow_term:x { No~output~streams~are~open }
+        \tl_show:n { }
+      }
+      {
+        \iow_term:x { The~following~output~streams~are~in~use: }
+        \tl_set:Nx \l_prop_show_tl
+          { \prop_map_function:NN \g_iow_streams_prop \iow_show_aux:nn }
+          \etex_showtokens:D \exp_after:wN \exp_after:wN \exp_after:wN
+            { \exp_after:wN \prop_show_aux:w \l_prop_show_tl }
+      }
+  }
+\cs_new_eq:NN \iow_show_aux:nn \ior_show_aux:nn
+\msg_kernel_new:nnnn { iow } { streams-exhausted }
+  { Output~streams~exhausted }
+  {
+    TeX~can~only~open~up~to~16~output~streams~at~one~time.\\
+    All~16 are currently~in~use,~and~something~wanted~to~open
+    another~one.
+  }
+\msg_kernel_new:nnnn { ior } { streams-exhausted }
+  { Input~streams~exhausted }
+  {
+    TeX~can~only~open~up~to~16~input~streams~at~one~time.\\
+    All~16 are currently~in~use,~and~something~wanted~to~open
+    another~one.
+  }
+\cs_new_eq:NN \iow_shipout_x:Nn \tex_write:D
+\cs_generate_variant:Nn \iow_shipout_x:Nn { Nx }
+\cs_new_protected_nopar:Npn \iow_shipout:Nn #1#2
+  { \iow_shipout_x:Nn #1 { \exp_not:n {#2} } }
+\cs_generate_variant:Nn \iow_shipout:Nn { Nx }
+\cs_new_protected_nopar:Npn \iow_now:Nx { \tex_immediate:D \iow_shipout_x:Nn }
+\cs_new_protected_nopar:Npn \iow_now:Nn #1#2
+  { \iow_now:Nx #1 { \exp_not:n {#2} } }
+\cs_set_protected_nopar:Npn \iow_log:x  { \iow_now:Nx \c_iow_log_stream  }
+\cs_new_protected_nopar:Npn \iow_log:n  { \iow_now:Nn \c_iow_log_stream  }
+\cs_set_protected_nopar:Npn \iow_term:x { \iow_now:Nx \c_iow_term_stream }
+\cs_new_protected_nopar:Npn \iow_term:n { \iow_now:Nn \c_iow_term_stream }
+\cs_new_protected_nopar:Npn \iow_now_when_avail:Nn #1
+  { \cs_if_free:NTF #1 { \use_none:n } { \iow_now:Nn #1 } }
+\cs_new_protected_nopar:Npn \iow_now_when_avail:Nx #1
+  { \cs_if_free:NTF #1 { \use_none:n } { \iow_now:Nx #1 } }
+\int_new:N  \l_iow_line_length_int
+\int_set:Nn \l_iow_line_length_int { 78 }
+\int_new:N \l_iow_target_length_int
+\int_new:N \l_iow_current_line_int
+\int_new:N \l_iow_current_word_int
+\tl_new:N \l_iow_current_line_tl
+\tl_new:N \l_iow_current_word_tl
+\tl_new:N \l_iow_wrap_tl
+\tl_new:N \l_iow_wrapped_tl
+\quark_new:N \q_iow_stop
+\bool_new:N \l_iow_line_start_bool
+\cs_new_protected:Npn \iow_wrap:xnnnN #1#2#3#4#5
+  {
+    \group_begin:
+      \int_set:Nn \l_iow_target_length_int { \l_iow_line_length_int - ( #3 ) }
+      \int_zero:N \l_iow_current_line_int
+      \tl_clear:N \l_iow_current_line_tl
+      \tl_clear:N \l_iow_wrap_tl
+      \bool_set_true:N \l_iow_line_start_bool
+      \cs_set:Npx \\ { \c_space_tl \iow_newline: \c_space_tl }
+      \cs_set_eq:NN \  \c_space_tl
+      #4
+      \protected@edef \l_iow_wrap_tl {#1}
+      \cs_set:Npn \\ { \iow_newline: #2 }
+      \use:x
+        {
+          \exp_not:N \iow_wrap_loop:w
+          \tl_to_str:N \l_iow_wrap_tl \c_space_tl
+          \exp_not:N \q_iow_stop \c_space_tl
+        }
+    \exp_args:NNo \group_end:
+    #5 \l_iow_wrapped_tl
+  }
+\cs_new_protected:Npn \iow_wrap_loop:w #1 ~ %
+  {
+    \tl_set:Nn \l_iow_current_word_tl {#1}
+    \tl_if_eq:NNTF \l_iow_current_word_tl \iow_newline:
+      { \iow_wrap_newline: }
+      {
+        \tl_if_eq:NNTF \l_iow_current_word_tl \q_iow_stop
+          { \iow_wrap_end: }
+          { \iow_wrap_word: }
+      }
+  }
+\cs_new_protected_nopar:Npn \iow_wrap_word:
+  {
+    \int_set:Nn \l_iow_current_word_int
+      { \str_length_skip_spaces:N \l_iow_current_word_tl }
+    \int_add:Nn \l_iow_current_line_int { \l_iow_current_word_int }
+    \int_compare:nNnTF \l_iow_current_line_int
+                     < \l_iow_target_length_int
+      { \iow_wrap_word_fits: }
+      { \iow_wrap_word_newline: }
+    \iow_wrap_loop:w
+  }
+\cs_new_protected_nopar:Npn \iow_wrap_word_fits:
+  {
+    \bool_if:NTF \l_iow_line_start_bool
+      {
+        \bool_set_false:N \l_iow_line_start_bool
+        \tl_set_eq:NN \l_iow_current_line_tl \l_iow_current_word_tl
+      }
+      {
+        \tl_put_right:Nx \l_iow_current_line_tl
+          { ~ \l_iow_current_word_tl }
+        \int_incr:N \l_iow_current_line_int
+      }
+  }
+\cs_new_protected_nopar:Npn \iow_wrap_word_newline:
+  {
+    \tl_put_right:Nx \l_iow_wrapped_tl
+      { \l_iow_current_line_tl \\ }
+    \int_set_eq:NN \l_iow_current_line_int \l_iow_current_word_int
+    \tl_set_eq:NN  \l_iow_current_line_tl  \l_iow_current_word_tl
+  }
+\cs_new_protected_nopar:Npn \iow_wrap_newline:
+  {
+    \tl_put_right:Nx \l_iow_wrapped_tl
+      { \l_iow_current_line_tl \\ }
+    \int_zero:N \l_iow_current_line_int
+    \tl_clear:N \l_iow_current_line_tl
+    \bool_set_true:N \l_iow_line_start_bool
+    \iow_wrap_loop:w
+  }
+\cs_new_protected_nopar:Npn \iow_wrap_end:
+  {
+    \tl_put_right:Nx \l_iow_wrapped_tl
+      { \l_iow_current_line_tl }
+  }
+\cs_new_nopar:Npn \str_length_skip_spaces:N
+  { \exp_args:No \str_length_skip_spaces:n }
+\cs_new:Npn \str_length_skip_spaces:n #1
+  {
+    \int_value:w \int_eval:w
+      \exp_after:wN \str_length_loop:NNNNNNNNN \tl_to_str:n {#1}
+      {X8}{X7}{X6}{X5}{X4}{X3}{X2}{X1}{X0} \q_stop
+    \int_eval_end:
+  }
+\cs_new:Npn \str_length_loop:NNNNNNNNN #1#2#3#4#5#6#7#8#9
+  {
+    \if_catcode:w X #9
+      \exp_after:wN \use_none_delimit_by_q_stop:w
+    \else:
+      9 +
+      \exp_after:wN \str_length_loop:NNNNNNNNN
+    \fi:
+  }
+\cs_new_nopar:Npn \iow_newline: { ^^J }
+\cs_new_eq:NN \iow_char:N \cs_to_str:N
+\prg_new_conditional:Nnn \ior_if_eof:N { p , T , F , TF }
+  {
+    \cs_if_exist:NTF #1
+      {
+        \if_eof:w #1
+          \prg_return_true:
+        \else:
+          \prg_return_false:
+        \fi:
+      }
+      { \prg_return_true: }
+  }
+\cs_new_protected_nopar:Npn \ior_to:NN #1#2
+  { \tex_read:D #1 to #2 }
+\cs_new_protected_nopar:Npn \ior_gto:NN #1#2
+  { \pref_global:D \tex_read:D #1 to #2 }
+\cs_new_protected_nopar:Npn \ior_str_to:NN #1#2
+  { \etex_readline:D #1 to #2 }
+\cs_new_protected_nopar:Npn \ior_str_gto:NN #1#2
+  { \pref_global:D \etex_readline:D #1 to #2 }
+\cs_new_protected:Npn \iow_now_buffer_safe:Nn #1#2
+  { \iow_wrap:xnnnN { \exp_not:n {#2} } { } \c_zero { } \iow_now:Nn #1 }
+\cs_new_protected:Npn \iow_now_buffer_safe:Nx #1#2
+  { \iow_wrap:xnnnN {#2} { } \c_zero { } \iow_now:Nn #1 }
+\cs_new_eq:NN \ior_new:N \use_none:n
+\cs_new_eq:NN \ior_new:c \use_none:n
+\cs_new_eq:NN \iow_new:N \use_none:n
+\cs_new_eq:NN \iow_new:c \use_none:n
+\cs_new_eq:NN \ior_open_streams: \ior_list_streams:
+\cs_new_eq:NN \iow_open_streams: \iow_list_streams:
+%% 
+%%
+%% End of file `l3io.sty'.
diff --git a/Master/texmf-dist/tex/latex/l3kernel/l3keys.sty b/Master/texmf-dist/tex/latex/l3kernel/l3keys.sty
new file mode 100644
index 00000000000..b12add10c9c
--- /dev/null
+++ b/Master/texmf-dist/tex/latex/l3kernel/l3keys.sty
@@ -0,0 +1,475 @@
+%%
+%% This is file `l3keys.sty',
+%% generated with the docstrip utility.
+%%
+%% The original source files were:
+%%
+%% l3keys.dtx  (with options: `package')
+%% 
+%% EXPERIMENTAL CODE
+%% 
+%% Do not distribute this file without also distributing the
+%% source files specified above.
+%% 
+%% Do not distribute a modified version of this file.
+%% 
+%% File: l3keys.dtx Copyright (C) 2009-2011 The LaTeX3 Project
+%%
+%% It may be distributed and/or modified under the conditions of the
+%% LaTeX Project Public License (LPPL), either version 1.3c of this
+%% license or (at your option) any later version.  The latest version
+%% of this license is in the file
+%%
+%%    http://www.latex-project.org/lppl.txt
+%%
+%% This file is part of the "expl3 bundle" (The Work in LPPL)
+%% and all files in that bundle must be distributed together.
+%%
+%% The released version of this bundle is available from CTAN.
+%%
+%% -----------------------------------------------------------------------
+%%
+%% The development version of the bundle can be found at
+%%
+%%    http://www.latex-project.org/svnroot/experimental/trunk/
+%%
+%% for those people who are interested.
+%%
+%%%%%%%%%%%
+%% NOTE: %%
+%%%%%%%%%%%
+%%
+%%   Snapshots taken from the repository represent work in progress and may
+%%   not work or may contain conflicting material!  We therefore ask
+%%   people _not_ to put them into distributions, archives, etc. without
+%%   prior consultation with the LaTeX3 Project.
+%%
+%% -----------------------------------------------------------------------
+\RequirePackage{l3names}
+\GetIdInfo$Id: l3keys.dtx 2478 2011-06-19 21:34:23Z joseph $
+  {L3 Experimental key-value interfaces}
+\ProvidesExplPackage
+  {\ExplFileName}{\ExplFileDate}{\ExplFileVersion}{\ExplFileDescription}
+\package_check_loaded_expl:
+\tl_const:Nn \c_keys_code_root_tl { key~code~>~ }
+\tl_const:Nn \c_keys_vars_root_tl { key~var~>~ }
+\tl_const:Nn \c_keys_props_root_tl { key~prop~>~ }
+\tl_const:Nn \c_keys_value_forbidden_tl { forbidden }
+\tl_const:Nn \c_keys_value_required_tl  { required }
+\int_new:N \l_keys_choice_int
+\tl_new:N \l_keys_choices_tl
+\tl_new:N \l_keys_key_tl
+\tl_new:N \l_keys_module_tl
+\bool_new:N \l_keys_no_value_bool
+\tl_new:N \l_keys_path_tl
+\tl_new:N \l_keys_property_tl
+\tl_new:N \l_keys_value_tl
+\cs_new_protected:Npn \keys_define:nn
+  { \keys_define_aux:onn \l_keys_module_tl }
+\cs_new_protected:Npn \keys_define_aux:nnn #1#2#3
+  {
+    \tl_set:Nx \l_keys_module_tl { \tl_to_str:n {#2} }
+    \keyval_parse:NNn \keys_define_elt:n \keys_define_elt:nn {#3}
+    \tl_set:Nn \l_keys_module_tl {#1}
+  }
+\cs_generate_variant:Nn \keys_define_aux:nnn { o }
+\cs_new_protected_nopar:Npn \keys_define_elt:n #1
+  {
+    \bool_set_true:N \l_keys_no_value_bool
+    \keys_define_elt_aux:nn {#1} { }
+  }
+\cs_new_protected:Npn \keys_define_elt:nn #1#2
+  {
+    \bool_set_false:N \l_keys_no_value_bool
+    \keys_define_elt_aux:nn {#1} {#2}
+  }
+\cs_new_protected:Npn \keys_define_elt_aux:nn #1#2 {
+  \keys_property_find:n {#1}
+  \cs_if_exist:cTF { \c_keys_props_root_tl \l_keys_property_tl }
+    { \keys_define_key:n {#2} }
+    {
+      \msg_kernel_error:nnxx { keys } { property-unknown }
+        { \l_keys_property_tl } { \l_keys_path_tl }
+    }
+}
+\cs_new_protected_nopar:Npn \keys_property_find:n #1
+  {
+    \tl_set:Nx \l_keys_path_tl { \l_keys_module_tl / }
+    \tl_if_in:nnTF {#1} { . }
+      { \keys_property_find_aux:w #1 \q_stop }
+      { \msg_kernel_error:nnx { keys } { key-no-property } {#1} }
+  }
+\cs_new_protected_nopar:Npn \keys_property_find_aux:w #1 . #2 \q_stop
+  {
+    \tl_set:Nx \l_keys_path_tl { \l_keys_path_tl \tl_to_str:n {#1} }
+    \tl_if_in:nnTF {#2} { . }
+      {
+        \tl_set:Nx \l_keys_path_tl { \l_keys_path_tl . }
+        \keys_property_find_aux:w #2 \q_stop
+      }
+      { \tl_set:Nn \l_keys_property_tl { . #2 } }
+  }
+\cs_new_protected:Npn \keys_define_key:n #1
+  {
+    \bool_if:NTF \l_keys_no_value_bool
+      {
+        \exp_after:wN \keys_define_key_aux:w
+          \l_keys_property_tl \q_stop
+          { \use:c { \c_keys_props_root_tl \l_keys_property_tl } }
+          {
+            \msg_kernel_error:nnxx { keys }
+              { property-requires-value } { \l_keys_property_tl }
+              { \l_keys_path_tl }
+            }
+      }
+      { \use:c { \c_keys_props_root_tl \l_keys_property_tl } {#1} }
+  }
+\cs_new_protected:Npn \keys_define_key_aux:w #1 : #2 \q_stop
+  { \tl_if_empty:nTF {#2} }
+\cs_new_nopar:Npn \keys_bool_set:NN #1#2
+  {
+    \cs_if_exist:NF #1 { \bool_new:N #1 }
+    \keys_choice_make:
+    \keys_cmd_set:nx { \l_keys_path_tl / true }
+      { \exp_not:c { bool_ #2 set_true:N } \exp_not:N #1 }
+    \keys_cmd_set:nx { \l_keys_path_tl / false }
+      { \exp_not:c { bool_ #2 set_false:N } \exp_not:N #1 }
+    \keys_cmd_set:nn { \l_keys_path_tl / unknown }
+      {
+        \msg_kernel_error:nnx { keys } { boolean-values-only }
+          { \l_keys_key_tl }
+      }
+    \keys_default_set:n { true }
+  }
+\cs_new_protected_nopar:Npn \keys_choice_make:
+  {
+    \keys_cmd_set:nn { \l_keys_path_tl }
+      { \keys_choice_find:n {##1} }
+    \keys_cmd_set:nn { \l_keys_path_tl / unknown }
+      {
+        \msg_kernel_error:nnxx { keys } { choice-unknown }
+          { \l_keys_path_tl } {##1}
+      }
+  }
+\cs_new_protected:Npn \keys_choices_generate:n #1
+  {
+    \cs_if_exist:cTF
+      { \c_keys_vars_root_tl \l_keys_path_tl .choice~code }
+      {
+        \keys_choice_make:
+        \int_zero:N \l_keys_choice_int
+        \clist_map_function:nN {#1} \keys_choices_generate_aux:n
+      }
+      {
+        \msg_kernel_error:nnx { keys }
+          { generate-choices-before-code } { \l_keys_path_tl }
+      }
+  }
+\cs_new_protected_nopar:Npn \keys_choices_generate_aux:n #1
+  {
+    \keys_cmd_set:nx { \l_keys_path_tl / #1 }
+      {
+        \exp_not:n { \tl_set:Nn \l_keys_choice_tl } {#1}
+        \exp_not:n { \int_set:Nn \l_keys_choice_int }
+          { \int_use:N \l_keys_choice_int }
+        \exp_not:v
+          { \c_keys_vars_root_tl \l_keys_path_tl .choice~code }
+      }
+    \int_incr:N \l_keys_choice_int
+  }
+\cs_new_protected:Npn \keys_choice_code_store:x #1
+  {
+    \cs_if_exist:cF
+      { \c_keys_vars_root_tl \l_keys_path_tl .choice~code }
+      {
+        \tl_new:c
+          { \c_keys_vars_root_tl \l_keys_path_tl .choice~code }
+      }
+    \tl_set:cx { \c_keys_vars_root_tl \l_keys_path_tl .choice~code }
+      {#1}
+  }
+\cs_new_protected:Npn \keys_cmd_set:nn #1#2
+  {
+    \keys_cmd_set_aux:n {#1}
+    \cs_set:cpn { \c_keys_code_root_tl #1 } ##1 {#2}
+  }
+\cs_new_protected:Npn \keys_cmd_set:nx #1#2
+  {
+    \keys_cmd_set_aux:n {#1}
+    \cs_set:cpx { \c_keys_code_root_tl #1 } ##1 {#2}
+  }
+\cs_new_protected_nopar:Npn \keys_cmd_set_aux:n #1
+  {
+    \tl_clear_new:c { \c_keys_vars_root_tl #1 .default }
+    \tl_set:cn { \c_keys_vars_root_tl #1 .default } { \q_no_value }
+    \tl_clear_new:c { \c_keys_vars_root_tl #1 .req }
+  }
+\cs_new_protected:Npn \keys_default_set:n #1
+  { \tl_set:cn { \c_keys_vars_root_tl \l_keys_path_tl .default } {#1} }
+\cs_generate_variant:Nn \keys_default_set:n { V }
+\cs_new_protected_nopar:Npn \keys_meta_make:n #1
+  {
+    \exp_args:NNo \keys_cmd_set:nn \l_keys_path_tl
+      { \exp_after:wN \keys_set:nn  \exp_after:wN { \l_keys_module_tl } {#1} }
+  }
+\cs_new_protected_nopar:Npn \keys_meta_make:x #1
+  {
+    \keys_cmd_set:nx { \l_keys_path_tl }
+      { \exp_not:N \keys_set:nn { \l_keys_module_tl } {#1} }
+  }
+\cs_new_protected_nopar:Npn \keys_value_requirement:n #1
+  {
+    \tl_set_eq:cc
+      { \c_keys_vars_root_tl \l_keys_path_tl .req }
+      { c_keys_value_ #1 _tl }
+  }
+\cs_new_protected_nopar:Npn \keys_variable_set:NnNN #1#2#3#4
+  {
+    \cs_if_exist:NF #1 { \use:c { #2 _new:N } #1 }
+    \keys_cmd_set:nx { \l_keys_path_tl }
+      { \exp_not:c { #2 _ #3 set:N #4 } \exp_not:N #1 {##1} }
+  }
+\cs_new_protected_nopar:Npn \keys_variable_set:NnN #1#2#3
+  { \keys_variable_set:NnNN #1 {#2} { } #3 }
+\cs_generate_variant:Nn \keys_variable_set:NnNN { c }
+\cs_generate_variant:Nn \keys_variable_set:NnN  { c }
+\cs_new_protected_nopar:cpn { \c_keys_props_root_tl .bool_set:N } #1
+  { \keys_bool_set:NN #1 { } }
+\cs_new_protected_nopar:cpn { \c_keys_props_root_tl .bool_gset:N } #1
+  { \keys_bool_set:NN #1 g }
+\cs_new_protected_nopar:cpn { \c_keys_props_root_tl .choice: }
+  { \keys_choice_make: }
+\cs_new_protected:cpn { \c_keys_props_root_tl .code:n } #1
+  { \keys_cmd_set:nn { \l_keys_path_tl } {#1} }
+\cs_new_protected:cpn { \c_keys_props_root_tl .code:x } #1
+  { \keys_cmd_set:nx { \l_keys_path_tl } {#1} }
+\cs_new_protected:cpn { \c_keys_props_root_tl .choice_code:n } #1
+  { \keys_choice_code_store:x { \exp_not:n {#1} } }
+\cs_new_protected:cpn { \c_keys_props_root_tl .choice_code:x } #1
+  { \keys_choice_code_store:x {#1} }
+\cs_new_protected:cpn { \c_keys_props_root_tl .default:n } #1
+  { \keys_default_set:n {#1} }
+\cs_new_protected:cpn { \c_keys_props_root_tl .default:V } #1
+  { \keys_default_set:V #1 }
+\cs_new_protected_nopar:cpn { \c_keys_props_root_tl .dim_set:N } #1
+  { \keys_variable_set:NnN #1 { dim } n }
+\cs_new_protected_nopar:cpn { \c_keys_props_root_tl .dim_set:c } #1
+  { \keys_variable_set:cnN {#1} { dim } n }
+\cs_new_protected_nopar:cpn { \c_keys_props_root_tl .dim_gset:N } #1
+  { \keys_variable_set:NnNN #1 { dim } g n }
+\cs_new_protected_nopar:cpn { \c_keys_props_root_tl .dim_gset:c } #1
+  { \keys_variable_set:cnNN {#1} { dim } g n }
+\cs_new_protected_nopar:cpn { \c_keys_props_root_tl .fp_set:N } #1
+  { \keys_variable_set:NnN #1 { fp } n }
+\cs_new_protected_nopar:cpn { \c_keys_props_root_tl .fp_set:c } #1
+  { \keys_variable_set:cnN {#1} { fp } n }
+\cs_new_protected_nopar:cpn { \c_keys_props_root_tl .fp_gset:N } #1
+  { \keys_variable_set:NnNN #1 { fp } g n }
+\cs_new_protected_nopar:cpn { \c_keys_props_root_tl .fp_gset:c } #1
+  { \keys_variable_set:cnNN {#1} { fp } g n }
+\cs_new_protected:cpn { \c_keys_props_root_tl .generate_choices:n } #1
+  { \keys_choices_generate:n {#1} }
+\cs_new_protected_nopar:cpn { \c_keys_props_root_tl .int_set:N } #1
+  { \keys_variable_set:NnN #1 { int } n }
+\cs_new_protected_nopar:cpn { \c_keys_props_root_tl .int_set:c } #1
+  { \keys_variable_set:cnN {#1} { int } n }
+\cs_new_protected_nopar:cpn { \c_keys_props_root_tl .int_gset:N } #1
+  { \keys_variable_set:NnNN #1 { int } g n }
+\cs_new_protected_nopar:cpn { \c_keys_props_root_tl .int_gset:c } #1
+  { \keys_variable_set:cnNN {#1} { int } g n }
+\cs_new_protected:cpn { \c_keys_props_root_tl .meta:n } #1
+  { \keys_meta_make:n {#1} }
+\cs_new_protected:cpn { \c_keys_props_root_tl .meta:x } #1
+  { \keys_meta_make:x {#1} }
+\cs_new_protected_nopar:cpn { \c_keys_props_root_tl .skip_set:N } #1
+  { \keys_variable_set:NnN #1 { skip } n }
+\cs_new_protected_nopar:cpn { \c_keys_props_root_tl .skip_set:c } #1
+  { \keys_variable_set:cnN {#1} { skip } n }
+\cs_new_protected_nopar:cpn { \c_keys_props_root_tl .skip_gset:N } #1
+  { \keys_variable_set:NnNN #1 { skip } g n }
+\cs_new_protected_nopar:cpn { \c_keys_props_root_tl .skip_gset:c } #1
+  { \keys_variable_set:cnNN {#1} { skip } g n }
+\cs_new_protected_nopar:cpn { \c_keys_props_root_tl .tl_set:N } #1
+   { \keys_variable_set:NnN #1 { tl } n }
+\cs_new_protected_nopar:cpn { \c_keys_props_root_tl .tl_set:c } #1
+  { \keys_variable_set:cnN {#1} { tl } n }
+\cs_new_protected_nopar:cpn { \c_keys_props_root_tl .tl_set_x:N } #1
+  { \keys_variable_set:NnN #1 { tl } x }
+\cs_new_protected_nopar:cpn { \c_keys_props_root_tl .tl_set_x:c } #1
+  { \keys_variable_set:cnN {#1} { tl } x }
+\cs_new_protected_nopar:cpn { \c_keys_props_root_tl .tl_gset:N } #1
+  { \keys_variable_set:NnNN #1 { tl } g n }
+\cs_new_protected_nopar:cpn { \c_keys_props_root_tl .tl_gset:c } #1
+  { \keys_variable_set:cnNN {#1} { tl } g n }
+\cs_new_protected_nopar:cpn { \c_keys_props_root_tl .tl_gset_x:N } #1
+  { \keys_variable_set:NnNN #1 { tl } g x }
+\cs_new_protected_nopar:cpn { \c_keys_props_root_tl .tl_gset_x:c } #1
+  { \keys_variable_set:cnNN {#1} { tl } g x }
+\cs_new_protected_nopar:cpn { \c_keys_props_root_tl .value_forbidden: }
+  { \keys_value_requirement:n { forbidden } }
+\cs_new_protected_nopar:cpn { \c_keys_props_root_tl .value_required: }
+  { \keys_value_requirement:n { required } }
+\cs_new_protected:Npn \keys_set:nn
+  { \keys_set_aux:onn { \l_keys_module_tl } }
+\cs_new_protected:Npn \keys_set_aux:nnn #1#2#3
+  {
+    \tl_set:Nn \l_keys_module_tl {#2}
+    \keyval_parse:NNn \keys_set_elt:n \keys_set_elt:nn {#3}
+    \tl_set:Nn \l_keys_module_tl {#1}
+  }
+\cs_generate_variant:Nn \keys_set:nn { nV , nv , no }
+\cs_generate_variant:Nn \keys_set_aux:nnn { o }
+\cs_new_protected_nopar:Npn \keys_set_elt:n #1
+  {
+    \bool_set_true:N \l_keys_no_value_bool
+    \keys_set_elt_aux:nn {#1} { }
+  }
+\cs_new_protected:Npn \keys_set_elt:nn #1#2
+  {
+    \bool_set_false:N \l_keys_no_value_bool
+    \keys_set_elt_aux:nn {#1} {#2}
+  }
+\cs_new_protected:Npn \keys_set_elt_aux:nn #1#2
+  {
+    \tl_set:Nx \l_keys_key_tl { \tl_to_str:n {#1} }
+    \tl_set:Nx \l_keys_path_tl { \l_keys_module_tl / \l_keys_key_tl }
+    \keys_value_or_default:n {#2}
+    \bool_if:nTF
+      {
+        \keys_if_value_p:n { required } &&
+        \l_keys_no_value_bool
+      }
+      {
+        \msg_kernel_error:nnx { keys } { value-required }
+          { \l_keys_path_tl }
+      }
+      {
+        \bool_if:nTF
+          {
+              \keys_if_value_p:n { forbidden } &&
+            ! \l_keys_no_value_bool
+          }
+          {
+            \msg_kernel_error:nnxx { keys } { value-forbidden }
+              { \l_keys_path_tl } { \l_keys_value_tl }
+          }
+          { \keys_execute: }
+      }
+  }
+\cs_new_protected:Npn \keys_value_or_default:n #1
+  {
+    \tl_set:Nn \l_keys_value_tl {#1}
+    \bool_if:NT \l_keys_no_value_bool
+      {
+        \quark_if_no_value:cF { \c_keys_vars_root_tl \l_keys_path_tl .default }
+          {
+            \tl_set_eq:Nc \l_keys_value_tl
+              { \c_keys_vars_root_tl \l_keys_path_tl .default }
+          }
+      }
+  }
+\prg_new_conditional:Npnn \keys_if_value:n #1 { p }
+  {
+    \tl_if_eq:ccTF { c_keys_value_ #1 _tl }
+      { \c_keys_vars_root_tl \l_keys_path_tl .req }
+      { \prg_return_true: }
+      { \prg_return_false: }
+  }
+\cs_new_nopar:Npn \keys_execute:
+  { \keys_execute:nn { \l_keys_path_tl } { \keys_execute_unknown: } }
+\cs_new_nopar:Npn \keys_execute_unknown:
+  {
+    \keys_execute:nn { \l_keys_module_tl / unknown }
+      {
+        \msg_kernel_error:nnxx { keys } { key-unknown }
+          { \l_keys_path_tl } { \l_keys_module_tl }
+      }
+  }
+\cs_new_nopar:Npn \keys_execute:nn #1#2
+  {
+    \cs_if_exist:cTF { \c_keys_code_root_tl #1 }
+      {
+        \exp_args:Nno \use:c { \c_keys_code_root_tl #1 }
+          \l_keys_value_tl
+      }
+      {#2}
+  }
+\cs_new_nopar:Npn \keys_choice_find:n #1
+  {
+    \keys_execute:nn { \l_keys_path_tl / \tl_to_str:n {#1} }
+      { \keys_execute:nn { \l_keys_path_tl / unknown } { } }
+  }
+\prg_new_conditional:Npnn \keys_if_exist:nn #1#2 { p , T , F , TF }
+  {
+    \cs_if_exist:cTF { \c_keys_code_root_tl #1 / #2 }
+      { \prg_return_true:}
+      { \prg_return_false:}
+  }
+\cs_new_nopar:Npn \keys_show:nn #1#2
+  { \cs_show:c { \c_keys_code_root_tl #1 / \tl_to_str:n {#2} } }
+\msg_kernel_new:nnnn { keys } { boolean-values-only }
+  { Key~'#1'~accepts~boolean~values~only. }
+  { The~key~'#1'~only~accepts~the~values~'true'~and~'false'. }
+\msg_kernel_new:nnnn { keys } { choice-unknown }
+  { Choice~'#2'~unknown~for~key~'#1'. }
+  {
+    The~key~'#1'~takes~a~limited~number~of~values.\\
+    The~input~given,~'#2',~is~not~on~the~list~accepted.
+  }
+\msg_kernel_new:nnnn { keys } { generate-choices-before-code }
+  { No~code~available~to~generate~choices~for~key~'#1'. }
+  {
+    \c_msg_coding_error_text_tl
+    Before~using~.generate_choices:n~the~code~should~be~defined~
+    with~'.choice_code:n'~or~'.choice_code:x'.
+  }
+\msg_kernel_new:nnnn { keys } { key-no-property }
+  { No~property~given~in~definition~of~key~'#1'. }
+  {
+    \c_msg_coding_error_text_tl
+    Inside~\keys_define:nn each~key~name
+    needs~a~property:  \\
+    ~ ~ #1 .<property> \\
+    LaTeX~did~not~find~a~'.'~to~indicate~the~start~of~a~property.
+  }
+\msg_kernel_new:nnnn { keys } { key-unknown }
+  { The~key~'#1'~is~unknown~and~is~being~ignored. }
+  {
+    The~module~'#2'~does~not~have~a~key~called~#1'.\\
+    Check~that~you~have~spelled~the~key~name~correctly.
+  }
+\msg_kernel_new:nnnn { keys } { option-unknown }
+  { Unknown~option~'#1'~for~package~#2. }
+  {
+    LaTeX~has~been~asked~to~set~an~option~called~'#1'~
+    but~the~#2~package~has~not~created~an~option~with~this~name.
+  }
+\msg_kernel_new:nnnn { keys } { property-requires-value }
+  { The~property~'#1'~requires~a~value. }
+  {
+    \c_msg_coding_error_text_tl
+    LaTeX~was~asked~to~set~property~'#2'~for~key~'#1'.\\
+    No~value~was~given~for~the~property,~and~one~is~required.
+  }
+\msg_kernel_new:nnnn { keys } { property-unknown }
+  { The~key~property~'#1'~is~unknown. }
+  {
+    \c_msg_coding_error_text_tl
+    LaTeX~has~been~asked~to~set~the~property~'#1'~for~key~'#2':~
+    this~property~is~not~defined.
+  }
+\msg_kernel_new:nnnn { keys } { value-forbidden }
+  { The~key~'#1'~does~not~taken~a~value. }
+  {
+    The~key~'#1'~should~be~given~without~a~value.\\
+    LaTeX~will~ignore~the~given~value~'#2'.
+  }
+\msg_kernel_new:nnnn { keys } { value-required }
+  { The~key~'#1'~requires~a~value. }
+  {
+    The~key~'#1'~must~have~a~value.\\
+    No~value~was~present:~the~key~will~be~ignored.
+  }
+%% 
+%%
+%% End of file `l3keys.sty'.
diff --git a/Master/texmf-dist/tex/latex/l3kernel/l3keyval.sty b/Master/texmf-dist/tex/latex/l3kernel/l3keyval.sty
new file mode 100644
index 00000000000..c42867b5b48
--- /dev/null
+++ b/Master/texmf-dist/tex/latex/l3kernel/l3keyval.sty
@@ -0,0 +1,177 @@
+%%
+%% This is file `l3keyval.sty',
+%% generated with the docstrip utility.
+%%
+%% The original source files were:
+%%
+%% l3keyval.dtx  (with options: `package')
+%% 
+%% EXPERIMENTAL CODE
+%% 
+%% Do not distribute this file without also distributing the
+%% source files specified above.
+%% 
+%% Do not distribute a modified version of this file.
+%% 
+%% File: l3keyval.dtx Copyright (C) 2006-2011 The LaTeX3 Project
+%%
+%% It may be distributed and/or modified under the conditions of the
+%% LaTeX Project Public License (LPPL), either version 1.3c of this
+%% license or (at your option) any later version.  The latest version
+%% of this license is in the file
+%%
+%%    http://www.latex-project.org/lppl.txt
+%%
+%% This file is part of the "expl3 bundle" (The Work in LPPL)
+%% and all files in that bundle must be distributed together.
+%%
+%% The released version of this bundle is available from CTAN.
+%%
+%% -----------------------------------------------------------------------
+%%
+%% The development version of the bundle can be found at
+%%
+%%    http://www.latex-project.org/svnroot/experimental/trunk/
+%%
+%% for those people who are interested.
+%%
+%%%%%%%%%%%
+%% NOTE: %%
+%%%%%%%%%%%
+%%
+%%   Snapshots taken from the repository represent work in progress and may
+%%   not work or may contain conflicting material!  We therefore ask
+%%   people _not_ to put them into distributions, archives, etc. without
+%%   prior consultation with the LaTeX3 Project.
+%%
+%% -----------------------------------------------------------------------
+\RequirePackage{l3names}
+\GetIdInfo$Id: l3keyval.dtx 2478 2011-06-19 21:34:23Z joseph $
+  {L3 Experimental key-value parsing}
+ %\section{Parsing key--value lists}
+ %\begin{function}{\keyval_parse:NNn}
+\ProvidesExplPackage
+  {\ExplFileName}{\ExplFileDate}{\ExplFileVersion}{\ExplFileDescription}
+\package_check_loaded_expl:
+\int_new:N \g_keyval_level_int
+\tl_new:N \l_keyval_key_tl
+\tl_new:N \l_keyval_value_tl
+\tl_new:N \l_keyval_sanitise_tl
+\tl_new:N \l_keyval_parse_tl
+\group_begin:
+  \char_set_catcode_active:n { `\= }
+  \char_set_catcode_active:n { `\, }
+  \char_set_lccode:nn { `\8 } { `\= }
+  \char_set_lccode:nn { `\9 } { `\, }
+\tl_to_lowercase:n
+  {
+    \group_end:
+    \cs_new_protected:Npn \keyval_parse:n #1
+      {
+        \group_begin:
+          \tl_clear:N \l_keyval_sanitise_tl
+          \tl_set:Nn \l_keyval_sanitise_tl {#1}
+          \tl_replace_all_in:Nnn \l_keyval_sanitise_tl { = } { 8 }
+          \tl_replace_all_in:Nnn \l_keyval_sanitise_tl { , } { 9 }
+          \tl_clear:N \l_keyval_parse_tl
+          \exp_after:wN \keyval_parse_elt:w \exp_after:wN
+            \q_no_value \l_keyval_sanitise_tl 9 \q_nil 9
+        \exp_after:wN \group_end:
+        \l_keyval_parse_tl
+      }
+  }
+\cs_new_protected:Npn \keyval_parse_elt:w #1 ,
+  {
+    \tl_if_blank:oTF { \use_none:n #1 }
+      { \keyval_parse_elt:w \q_no_value }
+      {
+        \quark_if_nil:oF { \use_ii:nn #1 }
+          {
+            \keyval_split_key_value:w #1 = = \q_stop
+            \keyval_parse_elt:w \q_no_value
+          }
+      }
+  }
+\cs_new_protected:Npn \keyval_split_key_value:w #1 = #2 \q_stop
+  {
+    \keyval_split_key:w #1 \q_stop
+    \str_if_eq:nnTF {#2} { = }
+      {
+        \tl_put_right:Nx \l_keyval_parse_tl
+          {
+            \exp_not:c { keyval_key_no_value_elt_ \int_use:N \g_keyval_level_int :n }
+              { \exp_not:o \l_keyval_key_tl }
+          }
+      }
+      {
+        \keyval_split_key_value_aux:wTF #2 \q_no_value \q_stop
+          { \keyval_split_value:w \q_nil #2 }
+          { \msg_kernel_error:nn { keyval } { misplaced-equals-sign } }
+      }
+  }
+\cs_new:Npn \keyval_split_key_value_aux:wTF #1 = #2#3 \q_stop
+  { \tl_if_head_eq_meaning:nNTF {#3} \q_no_value }
+\group_begin:
+  \char_set_catcode_math_toggle:n { `\Q }
+  \cs_new_protected:Npn \keyval_split_key:w #1 \q_stop
+    {
+      \exp_args:NNf \tl_set:Nn \l_keyval_key_tl
+        {
+          \exp_after:wN \keyval_remove_spaces:w \exp_after:wN
+            \exp_not:N \use_none:n #1 Q ~ Q
+        }
+      \tl_set:Nx \l_keyval_key_tl
+        { \exp_after:wN \keyval_split_key_aux:w \l_keyval_key_tl \q_stop }
+    }
+  \cs_gset:Npn \keyval_split_key_aux:w #1 \q_stop { \exp_not:n {#1} }
+  \cs_gset:Npn \keyval_remove_spaces:w #1 ~ Q { \keyval_remove_spaces_aux:w #1 Q }
+  \cs_gset:Npn \keyval_remove_spaces_aux:w #1 Q #2 {#1}
+\group_end:
+\cs_new_protected:Npn \keyval_split_value:w #1 = =
+  {
+    \tl_put_right:Nx \l_keyval_parse_tl
+      {
+        \exp_not:c { keyval_key_value_elt_ \int_use:N \g_keyval_level_int :nn }
+          { \exp_not:o \l_keyval_key_tl }
+      }
+    \tl_set:Nx \l_keyval_value_tl { \exp_not:o { \use_none:nnn #1 \q_nil \q_nil } }
+    \tl_if_empty:NTF \l_keyval_value_tl
+      { \tl_put_right:Nn \l_keyval_parse_tl { { } } }
+      {
+        \quark_if_nil:NTF \l_keyval_value_tl
+          {
+            \tl_put_right:Nx \l_keyval_parse_tl
+              { { \exp_not:o { \use_ii:nnn #1 \q_nil } } }
+          }
+          { \keyval_split_value_aux:w #1 \q_stop }
+      }
+  }
+\group_begin:
+  \char_set_catcode_math_toggle:n { `\Q }
+  \cs_new_protected:Npn \keyval_split_value_aux:w \q_nil #1 \q_stop
+    {
+      \exp_args:NNf \tl_set:Nn \l_keyval_value_tl
+        { \keyval_remove_spaces:w \exp_not:N #1 Q ~ Q }
+      \tl_put_right:Nx \l_keyval_parse_tl { { \exp_not:o \l_keyval_value_tl } }
+    }
+\group_end:
+\cs_new_protected:Npn \keyval_parse:NNn #1#2#3
+  {
+    \int_gincr:N \g_keyval_level_int
+    \cs_gset_eq:cN { keyval_key_no_value_elt_ \int_use:N \g_keyval_level_int :n } #1
+    \cs_gset_eq:cN { keyval_key_value_elt_ \int_use:N \g_keyval_level_int :nn }   #2
+    \keyval_parse:n {#3}
+    \int_gdecr:N \g_keyval_level_int
+  }
+\msg_kernel_new:nnnn { keyval } { misplaced-equals-sign }
+  { Misplaced~equals~sign~in~key-value~input~\msg_line_number: }
+  {
+    LaTeX~is~attempting~to~parse~some~key-value~input~but~found~
+    two~equals~signs~not~separated~by~a~comma.
+  }
+\cs_new_eq:NN \KV_process_space_removal_sanitize:NNn       \keyval_parse:NNn
+\cs_new_eq:NN \KV_process_space_removal_no_sanitize:NNn    \keyval_parse:NNn
+\cs_new_eq:NN \KV_process_no_space_removal_no_sanitize:NNn \keyval_parse:NNn
+%% 
+%%
+%% End of file `l3keyval.sty'.
diff --git a/Master/texmf-dist/tex/latex/l3kernel/l3luatex.sty b/Master/texmf-dist/tex/latex/l3kernel/l3luatex.sty
new file mode 100644
index 00000000000..b7e63c89882
--- /dev/null
+++ b/Master/texmf-dist/tex/latex/l3kernel/l3luatex.sty
@@ -0,0 +1,160 @@
+%%
+%% This is file `l3luatex.sty',
+%% generated with the docstrip utility.
+%%
+%% The original source files were:
+%%
+%% l3luatex.dtx  (with options: `package')
+%% 
+%% EXPERIMENTAL CODE
+%% 
+%% Do not distribute this file without also distributing the
+%% source files specified above.
+%% 
+%% Do not distribute a modified version of this file.
+%% 
+%% File: l3luatex.dtx Copyright (C) 2010,2011 The LaTeX3 Project
+%%
+%% It may be distributed and/or modified under the conditions of the
+%% LaTeX Project Public License (LPPL), either version 1.3c of this
+%% license or (at your option) any later version.  The latest version
+%% of this license is in the file
+%%
+%%    http://www.latex-project.org/lppl.txt
+%%
+%% This file is part of the "expl3 bundle" (The Work in LPPL)
+%% and all files in that bundle must be distributed together.
+%%
+%% The released version of this bundle is available from CTAN.
+%%
+%% -----------------------------------------------------------------------
+%%
+%% The development version of the bundle can be found at
+%%
+%%    http://www.latex-project.org/svnroot/experimental/trunk/
+%%
+%% for those people who are interested.
+%%
+%%%%%%%%%%%
+%% NOTE: %%
+%%%%%%%%%%%
+%%
+%%   Snapshots taken from the repository represent work in progress and may
+%%   not work or may contain conflicting material!  We therefore ask
+%%   people _not_ to put them into distributions, archives, etc. without
+%%   prior consultation with the LaTeX3 Project.
+%%
+%% -----------------------------------------------------------------------
+\RequirePackage{l3names}
+\GetIdInfo$Id: l3luatex.dtx 2478 2011-06-19 21:34:23Z joseph $
+  {L3 Experimental LuaTeX-specific functions}
+\ProvidesExplPackage
+  {\ExplFileName}{\ExplFileDate}{\ExplFileVersion}{\ExplFileDescription}
+\package_check_loaded_expl:
+\luatex_if_engine:TF
+  {
+    \cs_new_eq:NN \lua_now:x       \luatex_directlua:D
+    \cs_new_eq:NN \lua_shipout_x:n \luatex_latelua:D
+  }
+  {
+    \cs_new:Npn \lua_now:x #1 { \lua_wrong_engine: }
+    \cs_new_protected:Npn \lua_shipout_x:n #1 { \lua_wrong_engine: }
+  }
+\cs_new:Npn \lua_now:n #1
+  { \lua_now:x { \exp_not:n {#1} } }
+\cs_generate_variant:Nn \lua_shipout_x:n { x }
+\cs_new_protected:Npn \lua_shipout:n #1
+  { \lua_shipout_x:n { \exp_not:n {#1} } }
+\cs_generate_variant:Nn \lua_shipout:n { x }
+\group_begin:
+\char_set_catcode_letter:N\!
+\char_set_catcode_letter:N\ %
+\cs_gset:Npn\lua_wrong_engine:{%
+\LuaTeX engine not in use!%
+}%
+\group_end:%
+\int_new:N  \g_cctab_allocate_int
+\int_set:Nn \g_cctab_allocate_int { -1 }
+\int_new:N \g_cctab_stack_int
+\seq_new:N \g_cctab_stack_seq
+\cs_new_protected_nopar:Npn \cctab_new:N #1
+  {
+    \cs_if_free:NTF #1
+      {
+        \int_gadd:Nn \g_cctab_allocate_int { 2 }
+         \int_compare:nNnTF
+           { \g_cctab_allocate_int } <  { \c_max_register_int + 1 }
+           {
+             \pref_global:D \tex_mathchardef:D #1 \g_cctab_allocate_int
+             \luatex_initcatcodetable:D #1
+           }
+           { \msg_kernel_fatal:nnx { alloc } { out-of-registers } { cctab } }
+       }
+       {
+         \msg_kernel_error:nnx { code } { variable-already-defined }
+           { \token_to_str:N #1 }
+       }
+  }
+\luatex_if_engine:F
+  { \cs_set_protected_nopar:Npn \cctab_new:N #1 { \lua_wrong_engine: } }
+\luatex_if_engine:T
+  {
+    \cs_set_protected_nopar:Npn \cctab_new:N #1
+      {
+        \newcatcodetable #1
+        \luatex_initcatcodetable:D #1
+      }
+  }
+\cs_new_protected_nopar:Npn \cctab_begin:N #1
+  {
+    \seq_gpush:Nx \g_cctab_stack_seq { \tex_the:D \luatex_catcodetable:D }
+    \luatex_catcodetable:D #1
+    \int_gadd:Nn \g_cctab_stack_int { 2 }
+    \int_compare:nNnT { \g_cctab_stack_int } > { 268 435 453 }
+      { \msg_kernel_error:nn { code } { cctab-stack-full } }
+    \luatex_savecatcodetable:D \g_cctab_stack_int
+    \luatex_catcodetable:D \g_cctab_stack_int
+  }
+\cs_new_protected_nopar:Npn \cctab_end:
+  {
+    \int_gsub:Nn \g_cctab_stack_int { 2 }
+    \seq_gpop:NN \g_cctab_stack_seq \l_cctab_tmp_tl
+    \quark_if_no_value:NT \l_cctab_tmp_tl
+      { \tl_set:Nn \l_cctab_tmp_tl { 0 } }
+    \luatex_catcodetable:D \l_cctab_tmp_tl \scan_stop:
+  }
+\luatex_if_engine:F
+  {
+    \cs_set_protected_nopar:Npn \cctab_begin:N #1 { \lua_wrong_engine: }
+    \cs_set_protected_nopar:Npn \cctab_end: { \lua_wrong_engine: }
+  }
+\luatex_if_engine:T
+  {
+    \cs_set_protected_nopar:Npn \cctab_begin:N #1 { \BeginCatcodeRegime #1 }
+    \cs_set_protected_nopar:Npn \cctab_end: { \EndCatcodeRegime }
+  }
+\tl_new:N \l_cctab_tmp_tl
+\cs_new_protected:Npn \cctab_gset:Nn #1#2
+  {
+    \group_begin:
+      #2
+      \luatex_savecatcodetable:D #1
+    \group_end:
+  }
+\luatex_if_engine:F
+  { \cs_set_protected_nopar:Npn \cctab_gset:Nn #1#2 { \lua_wrong_engine: } }
+\luatex_if_engine:T
+  {
+    \cctab_new:N \c_code_cctab
+    \cctab_gset:Nn \c_code_cctab { }
+  }
+\luatex_if_engine:T
+  {
+    \cs_new_eq:NN \c_document_cctab \CatcodeTableLaTeX
+    \cs_new_eq:NN \c_initex_cctab   \CatcodeTableIniTeX
+    \cs_new_eq:NN \c_other_cctab    \CatcodeTableOther
+    \cs_new_eq:NN \c_string_cctab   \CatcodeTableString
+  }
+%% 
+%%
+%% End of file `l3luatex.sty'.
diff --git a/Master/texmf-dist/tex/latex/l3kernel/l3msg.sty b/Master/texmf-dist/tex/latex/l3kernel/l3msg.sty
new file mode 100644
index 00000000000..bf1d732514f
--- /dev/null
+++ b/Master/texmf-dist/tex/latex/l3kernel/l3msg.sty
@@ -0,0 +1,574 @@
+%%
+%% This is file `l3msg.sty',
+%% generated with the docstrip utility.
+%%
+%% The original source files were:
+%%
+%% l3msg.dtx  (with options: `package')
+%% 
+%% EXPERIMENTAL CODE
+%% 
+%% Do not distribute this file without also distributing the
+%% source files specified above.
+%% 
+%% Do not distribute a modified version of this file.
+%% 
+%% File: l3msg.dtx Copyright (C) 2009-2011 The LaTeX3 Project
+%%
+%% It may be distributed and/or modified under the conditions of the
+%% LaTeX Project Public License (LPPL), either version 1.3c of this
+%% license or (at your option) any later version.  The latest version
+%% of this license is in the file
+%%
+%%    http://www.latex-project.org/lppl.txt
+%%
+%% This file is part of the "expl3 bundle" (The Work in LPPL)
+%% and all files in that bundle must be distributed together.
+%%
+%% The released version of this bundle is available from CTAN.
+%%
+%% -----------------------------------------------------------------------
+%%
+%% The development version of the bundle can be found at
+%%
+%%    http://www.latex-project.org/svnroot/experimental/trunk/
+%%
+%% for those people who are interested.
+%%
+%%%%%%%%%%%
+%% NOTE: %%
+%%%%%%%%%%%
+%%
+%%   Snapshots taken from the repository represent work in progress and may
+%%   not work or may contain conflicting material!  We therefore ask
+%%   people _not_ to put them into distributions, archives, etc. without
+%%   prior consultation with the LaTeX3 Project.
+%%
+%% -----------------------------------------------------------------------
+\RequirePackage{l3names}
+\GetIdInfo$Id: l3msg.dtx 2478 2011-06-19 21:34:23Z joseph $
+  {L3 Experimental messages}
+\ProvidesExplPackage
+  {\ExplFileName}{\ExplFileDate}{\ExplFileVersion}{\ExplFileDescription}
+\package_check_loaded_expl:
+\tl_new:N \l_msg_tmp_tl
+\tl_const:Nn \c_msg_text_prefix_tl      { msg~text~>~ }
+\tl_const:Nn \c_msg_more_text_prefix_tl { msg~extra~text~>~ }
+\cs_new_protected:Npn \msg_new:nnnn #1#2
+  {
+    \cs_if_exist:cT { \c_msg_text_prefix_tl #1 / #2 }
+      {
+        \msg_kernel_error:nn { msg } { message-already-defined }
+          {#1} {#2}
+      }
+    \msg_set:nnnn {#1} {#2}
+  }
+\cs_new_protected:Npn \msg_new:nnn #1#2#3
+  { \msg_new:nnnn {#1} {#2} {#3} { } }
+\cs_new_protected:Npn \msg_set:nnnn #1#2#3#4
+  {
+    \cs_set:cpn { \c_msg_text_prefix_tl #1 / #2 }
+      ##1##2##3##4 {#3}
+    \cs_set:cpn { \c_msg_more_text_prefix_tl #1 / #2 }
+      ##1##2##3##4 {#4}
+  }
+\cs_new_protected:Npn \msg_set:nnn #1#2#3
+  { \msg_set:nnnn {#1} {#2} {#3} { } }
+\tl_const:Nn \c_msg_coding_error_text_tl
+  {
+    This~is~a~coding~error.
+    \\ \\
+  }
+\tl_const:Nn \c_msg_continue_text_tl
+  { Type~<return>~to~continue }
+\tl_const:Nn \c_msg_critical_text_tl
+  { Reading~the~current~file~will~stop }
+\tl_const:Nn \c_msg_fatal_text_tl
+  { This~is~a~fatal~error:~LaTeX~will~abort }
+\tl_const:Nn \c_msg_help_text_tl
+  { For~immediate~help~type~H~<return> }
+\tl_const:Nn \c_msg_no_info_text_tl
+  {
+    LaTeX~does~not~know~anything~more~about~this~error,~sorry.
+    \c_msg_return_text_tl
+  }
+\tl_const:Nn \c_msg_on_line_text_tl { on~line }
+\tl_const:Nn \c_msg_return_text_tl
+  {
+    \\ \\
+    Try~typing~<return>~to~proceed.
+    \\
+    If~that~doesn't~work,~type~X~<return>~to~quit.
+  }
+\tl_const:Nn \c_msg_trouble_text_tl
+  {
+    \\ \\
+    More~errors~will~almost~certainly~follow: \\
+    the~LaTeX~run~should~be~aborted.
+  }
+\cs_new_nopar:Npn \msg_newline: { ^^J }
+\cs_new_nopar:Npn \msg_two_newlines: { ^^J ^^J }
+\cs_new_nopar:Npn \msg_line_number: { \int_use:N \tex_inputlineno:D }
+\cs_set_nopar:Npn \msg_line_context:
+  {
+    \c_msg_on_line_text_tl
+    \c_space_tl
+    \msg_line_number:
+  }
+\char_set_catcode_letter:N \.
+\tl_new:N
+  \c_msg_hide_tl..........................................................
+\tl_const:Nn \c_msg_hide_tl
+ { \c_msg_hide_tl.......................................................... }
+\char_set_catcode_other:N \.
+\cs_new_protected:Npn \msg_interrupt:xxx #1#2#3
+  {
+    \group_begin:
+      \tl_if_empty:nTF {#3}
+        { \msg_interrupt_no_details:xx {#1} {#2} }
+        { \msg_interrupt_details:xxx {#1} {#2} {#3} }
+      \msg_interrupt_aux:
+    \group_end:
+  }
+\cs_new_protected:Npn \msg_interrupt_no_details:xx #1#2
+  {
+    \iow_wrap:xnnnN
+      { \\ \c_msg_no_info_text_tl }
+      { |~ } { 2 } { } \msg_interrupt_more_text:n
+    \iow_wrap:xnnnN { #1 \\ \\ #2 \\ \\ \c_msg_continue_text_tl }
+      { ! ~ } { 2 } {} \msg_interrupt_text:n
+  }
+\cs_new_protected:Npn \msg_interrupt_details:xxx #1#2#3
+  {
+    \iow_wrap:xnnnN
+      { \\ #3 }
+      { |~ } { 2 } { } \msg_interrupt_more_text:n
+    \iow_wrap:xnnnN { #1 \\ \\ #2 \\ \\ \c_msg_help_text_tl }
+      { ! ~ } { 2 } { } \msg_interrupt_text:n
+  }
+\cs_new_protected:Npn \msg_interrupt_text:n #1
+  { \tl_set:Nn \l_msg_text_tl {#1} }
+\cs_new_protected:Npn \msg_interrupt_more_text:n #1
+  {
+    \protected@edef \l_msg_tmp_tl
+    {
+      |'''''''''''''''''''''''''''''''''''''''''''''''
+      #1
+      \msg_newline:
+      |...............................................
+    }
+    \tex_errhelp:D \exp_after:wN { \l_msg_tmp_tl }
+  }
+\group_begin: % {
+  \char_set_lccode:w `\} = `\  \scan_stop:
+  \char_set_lccode:w `\& = `\! \scan_stop:
+  \char_set_catcode_active:N \&
+\tl_to_lowercase:n
+  {
+    \group_end:
+    \cs_new_protected:Npn \msg_interrupt_aux:
+      {
+        \iow_term:x
+          {
+            \iow_newline:
+            !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
+            \iow_newline:
+            !
+          }
+        \tl_put_right:No \l_msg_text_tl { \c_msg_hide_tl }
+        \cs_set_protected_nopar:Npx &
+          { \tex_errmessage:D { \exp_not:o { \l_msg_text_tl } } }
+        &
+      }
+  }
+\cs_new_protected:Npn \msg_log:x #1
+  {
+    \iow_log:x { ................................................. }
+    \iow_wrap:xnnnN { . ~ #1} { . ~ } { 2 } { }
+      \iow_log:x
+    \iow_log:x { ................................................. }
+  }
+\cs_new_protected:Npn \msg_term:x #1
+  {
+    \iow_term:x { ************************************************* }
+    \iow_wrap:xnnnN { * ~ #1} { * ~ } { 2 } { }
+      \iow_term:x
+    \iow_term:x { ************************************************* }
+  }
+\int_set:Nn \tex_errorcontextlines:D { -1 }
+\cs_new_nopar:Npn \msg_fatal_text:n #1 { Fatal~#1~error }
+\cs_new_nopar:Npn \msg_critical_text:n #1 { Critical~#1~error }
+\cs_new_nopar:Npn \msg_error_text:n #1 { #1~error }
+\cs_new_nopar:Npn \msg_warning_text:n #1 { #1~warning }
+\cs_new_nopar:Npn \msg_info_text:n #1 { #1~info }
+\cs_new_nopar:Npn \msg_see_documentation_text:n #1
+  { \\ \\ See~the~#1~documentation~for~further~information. }
+\prop_new:N \l_msg_redirect_classes_prop
+\prop_new:N \l_msg_redirect_names_prop
+\cs_new_protected_nopar:Npn \msg_class_set:nn #1#2
+  {
+    \prop_clear_new:c { l_msg_redirect_ #1 _prop }
+    \cs_set_protected:cpn { msg_ #1 :nnxxxx } ##1##2##3##4##5##6
+      { \msg_use:nnnnxxxx {#1} {#2} {##1} {##2} {##3} {##4} {##5} {##6} }
+    \cs_set_protected:cpx { msg_ #1 :nnxxx } ##1##2##3##4##5
+      { \exp_not:c { msg_ #1 :nnxxxx } {##1} {##2} {##3} {##4} {##5} { } }
+    \cs_set_protected:cpx { msg_ #1 :nnxx } ##1##2##3##4
+      { \exp_not:c { msg_ #1 :nnxxxx } {##1} {##2} {##3} {##4} { } { } }
+    \cs_set_protected:cpx { msg_ #1 :nnx } ##1##2##3
+      { \exp_not:c { msg_ #1 :nnxxxx } {##1} {##2} {##3} { } { } { } }
+    \cs_set_protected:cpx { msg_ #1 :nn } ##1##2
+      { \exp_not:c { msg_ #1 :nnxxxx } {##1} {##2} { } { } { } { } }
+  }
+\prg_set_conditional:Npnn \msg_if_more_text:N #1 { p , T , F , TF }
+  {
+    \cs_if_eq:NNTF #1 \msg_no_more_text:xxxx
+      { \prg_return_false:  }
+      { \prg_return_true: }
+  }
+\cs_new:Npn \msg_no_more_text:xxxx #1#2#3#4 { }
+\cs_generate_variant:Nn \msg_if_more_text_p:N { c }
+\cs_generate_variant:Nn \msg_if_more_text:NT  { c }
+\cs_generate_variant:Nn \msg_if_more_text:NF  { c }
+\cs_generate_variant:Nn \msg_if_more_text:NTF { c }
+\msg_class_set:nn { fatal }
+  {
+    \msg_interrupt:xxx
+      { \msg_fatal_text:n {#1} : ~ "#2" }
+      {
+        \use:c { \c_msg_text_prefix_tl #1 / #2 } {#3} {#4} {#5} {#6}
+        \msg_see_documentation_text:n {#1}
+      }
+      { \c_msg_fatal_text_tl }
+    \tex_end:D
+  }
+\msg_class_set:nn { critical }
+  {
+    \msg_interrupt:xxx
+      { \msg_critical_text:n {#1} : ~ "#2" }
+      {
+        \use:c { \c_msg_text_prefix_tl #1 / #2 } {#3} {#4} {#5} {#6}
+        \msg_see_documentation_text:n {#1}
+      }
+      { \c_msg_critical_text_tl }
+    \tex_endinput:D
+  }
+\msg_class_set:nn { error }
+  {
+    \msg_if_more_text:cTF { \c_msg_more_text_prefix_tl #1 / #2 }
+      {
+        \msg_interrupt:xxx
+          { \msg_error_text:n {#1} : ~ "#2" }
+          {
+            \use:c { \c_msg_text_prefix_tl #1 / #2 } {#3} {#4} {#5} {#6}
+            \msg_see_documentation_text:n {#1}
+          }
+          { \use:c { \c_msg_more_text_prefix_tl #1 / #2 } {#3} {#4} {#5} {#6} }
+     }
+     {
+        \msg_interrupt:xxx
+          { \msg_error_text:n {#1} : ~ "#2" }
+          {
+            \use:c { \c_msg_text_prefix_tl #1 / #2 } {#3} {#4} {#5} {#6}
+            \msg_see_documentation_text:n {#1}
+          }
+          { }
+     }
+  }
+\msg_class_set:nn { warning }
+  {
+    \msg_term:x
+      {
+        \msg_warning_text:n {#1} : ~ "#2" \\ \\
+        \use:c { \c_msg_text_prefix_tl #1 / #2 } {#3} {#4} {#5} {#6}
+      }
+  }
+\msg_class_set:nn { info }
+  {
+    \msg_log:x
+      {
+        \msg_info_text:n {#1} : ~ "#2" \\ \\
+        \use:c { \c_msg_text_prefix_tl #1 / #2 } {#3} {#4} {#5} {#6}
+      }
+  }
+\msg_class_set:nn { log }
+  {
+    \msg_log:x
+      { \use:c { \c_msg_text_prefix_tl #1 / #2 } {#3} {#4} {#5} {#6} }
+  }
+\msg_class_set:nn { none } { }
+\seq_new:N \l_msg_redirect_classes_seq
+\tl_new:N \l_msg_class_tl
+\tl_new:N \l_msg_current_class_tl
+\tl_new:N \l_msg_current_module_tl
+\cs_new_protected:Npn \msg_use:nnnnxxxx #1#2#3#4#5#6#7#8
+  {
+    \cs_set_protected_nopar:Npx \msg_use_code:
+      {
+        \seq_clear:N \exp_not:N \l_msg_redirect_classes_seq
+        \exp_not:n {#2}
+      }
+    \cs_set_protected:Npx \msg_use_loop:n ##1
+      {
+        \seq_if_in:NnTF \exp_not:n \l_msg_redirect_classes_seq {#1}
+          { \msg_kernel_error:nn { msg }  { message-loop } {#1} }
+          {
+            \seq_put_right:Nn \exp_not:N \l_msg_redirect_classes_seq {#1}
+            \exp_not:N \cs_if_exist:cTF { msg_ ##1 :nnxxxx }
+              {
+                \exp_not:N \use:c { msg_ ##1 :nnxxxx }
+                  \exp_not:n { {#3} {#4} {#5} {#6} {#7} {#8} }
+              }
+              {
+                \msg_kernel_error:nnx { msg } { message-class-unknown } {##1}
+              }
+          }
+      }
+    \cs_if_exist:cTF { \c_msg_text_prefix_tl #3 / #4 }
+      { \msg_use_aux:nnn {#1} {#3} {#4} }
+      { \msg_kernel_error:nnxx { msg } { message-unknown } {#3} {#4} }
+  }
+\cs_new_protected_nopar:Npn \msg_use_aux:nnn #1#2#3
+  {
+    \tl_set:Nn \l_msg_current_class_tl  {#1}
+    \tl_set:Nn \l_msg_current_module_tl {#2}
+    \prop_if_in:NnTF \l_msg_redirect_names_prop { // #2 / #3 / }
+      { \msg_use_loop_check:nn { names } { // #2 / #3 / } }
+      { \msg_use_aux:nn {#1} {#2} }
+  }
+\cs_new_protected_nopar:Npn \msg_use_aux:nn #1#2
+  {
+    \prop_if_in:cnTF { l_msg_redirect_ #1 _prop } {#2}
+      { \msg_use_loop_check:nn {#1} {#2} }
+      {
+        \prop_if_in:cnTF { l_msg_redirect_ #1 _prop } { * }
+          { \msg_use_loop_check:nn {#1} { * } }
+          { \msg_use_code: }
+      }
+  }
+\cs_new_protected:Npn \msg_use_loop_check:nn #1#2
+  {
+    \prop_get:cnN { l_msg_redirect_ #1 _prop } {#2} \l_msg_class_tl
+    \tl_if_eq:NNTF \l_msg_current_class_tl \l_msg_class_tl
+      {
+        { \msg_use_code: }
+        { \msg_use_loop:o \l_msg_class_tl }
+      }
+  }
+\cs_new_protected_nopar:Npn \msg_use_code: { }
+\cs_new_protected:Npn \msg_use_loop:n #1 { }
+\cs_generate_variant:Nn \msg_use_loop:n { o }
+\cs_new_protected_nopar:Npn \msg_redirect_class:nn #1#2
+  { \prop_put:cnn { l_msg_redirect_ #1 _prop } { * } {#2} }
+\cs_new_protected_nopar:Npn \msg_redirect_module:nnn #1#2#3
+  { \prop_put:cnn { l_msg_redirect_ #2 _prop } {#1} {#3} }
+\cs_new_protected_nopar:Npn \msg_redirect_name:nnn #1#2#3
+  { \prop_put:Nnn \l_msg_redirect_names_prop { // #1 / #2 / } {#3} }
+\cs_new_protected_nopar:Npn \msg_kernel_new:nnnn #1#2
+  { \msg_new:nnnn { LaTeX } { #1 / #2 } }
+\cs_new_protected_nopar:Npn \msg_kernel_new:nnn #1#2
+  { \msg_new:nnn { LaTeX } { #1 / #2 } }
+\cs_new_protected_nopar:Npn \msg_kernel_set:nnnn #1#2
+  { \msg_set:nnnn { LaTeX } { #1 / #2 } }
+\cs_new_protected_nopar:Npn \msg_kernel_set:nnn #1#2
+  { \msg_set:nnn { LaTeX } { #1 / #2 } }
+\cs_new_protected:Npn \msg_kernel_fatal:nnxxxx #1#2#3#4#5#6
+  {
+    \msg_interrupt:xxx
+      { \msg_fatal_text:n { LaTeX } : ~ "#1 / #2" }
+      {
+        \use:c { \c_msg_text_prefix_tl LaTeX / #1 / #2 }
+          {#3} {#4} {#5} {#6}
+        \msg_see_documentation_text:n { LaTeX3 }
+      }
+      { \c_msg_fatal_text_tl }
+    \tex_end:D
+  }
+\cs_new_protected:Npn \msg_kernel_fatal:nnxxx #1#2#3#4#5
+  {\msg_kernel_fatal:nnxxxx {#1} {#2} {#3} {#4} {#5} { } }
+\cs_new_protected:Npn \msg_kernel_fatal:nnxx #1#2#3#4
+  { \msg_kernel_fatal:nnxxxx {#1} {#2} {#3} {#4} { } { } }
+\cs_new_protected:Npn \msg_kernel_fatal:nnx #1#2#3
+  { \msg_kernel_fatal:nnxxxx {#1} {#2} {#3} { } { } { } }
+\cs_new_protected:Npn \msg_kernel_fatal:nn #1#2
+  { \msg_kernel_fatal:nnxxxx {#1} {#2} { } { } { } { } }
+\cs_new_protected:Npn \msg_kernel_error:nnxxxx #1#2#3#4#5#6
+  {
+    \msg_if_more_text:cTF { \c_msg_more_text_prefix_tl LaTeX / #1 / #2 }
+      {
+        \msg_interrupt:xxx
+          { \msg_error_text:n { LaTeX } : ~ " #1 / #2 " }
+          {
+            \use:c { \c_msg_text_prefix_tl LaTeX / #1 / #2 }
+              {#3} {#4} {#5} {#6}
+            \msg_see_documentation_text:n { LaTeX3 }
+          }
+          {
+            \use:c { \c_msg_more_text_prefix_tl LaTeX / #1 / #2 }
+              {#3} {#4} {#5} {#6}
+          }
+      }
+      {
+        \msg_interrupt:xxx
+          { \msg_error_text:n { LaTeX } : ~ " #1 / #2 " }
+          {
+            \use:c { \c_msg_text_prefix_tl LaTeX / #1 / #2 }
+              {#3} {#4} {#5} {#6}
+            \msg_see_documentation_text:n { LaTeX3 }
+          }
+          { }
+      }
+  }
+\cs_new_protected:Npn \msg_kernel_error:nnxxx #1#2#3#4#5
+  {\msg_kernel_error:nnxxxx {#1} {#2} {#3} {#4} {#5} { } }
+\cs_set_protected:Npn \msg_kernel_error:nnxx #1#2#3#4
+  { \msg_kernel_error:nnxxxx {#1} {#2} {#3} {#4} { } { } }
+\cs_set_protected:Npn \msg_kernel_error:nnx #1#2#3
+  { \msg_kernel_error:nnxxxx {#1} {#2} {#3} { } { } { } }
+\cs_set_protected:Npn \msg_kernel_error:nn #1#2
+  { \msg_kernel_error:nnxxxx {#1} {#2} { } { } { } { } }
+\prop_new:N \l_msg_redirect_kernel_warning_prop
+\cs_new_protected:Npn \msg_kernel_warning:nnxxxx #1#2#3#4#5#6
+  {
+    \msg_use:nnnnxxxx { warning }
+      {
+        \msg_term:x
+          {
+            \msg_warning_text:n { LaTeX } : ~ " #1 / #2 " \\ \\
+            \use:c { \c_msg_text_prefix_tl LaTeX / #1 / #2 }
+              {#3} {#4} {#5} {#6}
+          }
+      }
+      { LaTeX } { #1 / #2 } {#3} {#4} {#5} {#6}
+  }
+\cs_new_protected:Npn \msg_kernel_warning:nnxxx #1#2#3#4#5
+  { \msg_kernel_warning:nnxxxx {#1} {#2} {#3} {#4} {#5} { } }
+\cs_new_protected:Npn \msg_kernel_warning:nnxx #1#2#3#4
+  { \msg_kernel_warning:nnxxxx {#1} {#2} {#3} {#4} { } { } }
+\cs_new_protected:Npn \msg_kernel_warning:nnx #1#2#3
+  { \msg_kernel_warning:nnxxxx {#1} {#2} {#3} { } { } { } }
+\cs_new_protected:Npn \msg_kernel_warning:nn #1#2
+  { \msg_kernel_warning:nnxxxx {#1} {#2} { } { } { } { } }
+\prop_new:N \l_msg_redirect_kernel_info_prop
+\cs_new_protected:Npn \msg_kernel_info:nnxxxx #1#2#3#4#5#6
+  {
+    \msg_use:nnnnxxxx { info }
+      {
+        \msg_log:x
+          {
+            \msg_info_text:n { LaTeX } : ~ " #1 / #2 " \\ \\
+            \use:c { \c_msg_text_prefix_tl LaTeX / #1 / #2 }
+              {#3} {#4} {#5} {#6}
+          }
+      }
+      { LaTeX } { #1 / #2 } {#3} {#4} {#5} {#6}
+  }
+\cs_new_protected:Npn \msg_kernel_info:nnxxx #1#2#3#4#5
+  { \msg_kernel_info:nnxxxx {#1} {#2} {#3} {#4} {#5} { } }
+\cs_new_protected:Npn \msg_kernel_info:nnxx #1#2#3#4
+  { \msg_kernel_info:nnxxxx {#1} {#2} {#3} {#4} { } { } }
+\cs_new_protected:Npn \msg_kernel_info:nnx #1#2#3
+  { \msg_kernel_info:nnxxxx {#1} {#2} {#3} { } { } { } }
+\cs_new_protected:Npn \msg_kernel_info:nn #1#2
+  { \msg_kernel_info:nnxxxx {#1} {#2} { } { } { } { } }
+\msg_kernel_new:nnnn { msg } { message-already-defined }
+  { Message~'#2'~for~module~'#1'~already~defined. }
+  {
+    \c_msg_coding_error_text_tl
+    LaTeX~was~asked~to~define~a~new~message~called~'#2'
+    by~the~module~'#1'~module:\\
+    this~message~already~exists.
+    \c_msg_return_text_tl
+  }
+\msg_kernel_new:nnnn { msg } { message-unknown }
+  { Unknown~message~'#2'~for~module~'#1'. }
+  {
+    \c_msg_coding_error_text_tl
+    LaTeX~was~asked~to~display~a~message~called~'#2'\\
+    by~the~module~'#1'~module:~this~message~does~not~exist.
+    \c_msg_return_text_tl
+  }
+\msg_kernel_new:nnnn { msg } { message-class-unknown }
+  { Unknown~message~class~'#1'. }
+  {
+    LaTeX~has~been~asked~to~redirect~messages~to~a~class~'#1':\\
+    this~was~never~defined.
+    \c_msg_return_text_tl
+  }
+\msg_kernel_new:nnnn { msg } { redirect-loop }
+  { Message~redirection~loop~for~message~class~'#1'. }
+  {
+    LaTeX~has~been~asked~to~redirect~messages~in~an~infinite~loop.\\
+    The~original~message~here~has~been~lost.
+    \c_msg_return_text_tl
+  }
+\msg_kernel_new:nnnn { kernel } { bad-number-of-arguments }
+  { Function~'#1'~cannot~be~defined~with~#2~arguments. }
+  {
+    \c_msg_coding_error_text_tl
+    LaTeX~has~been~asked~to~define~a~function~'#1'~with~
+    #2~arguments. \\
+    TeX~allows~between~0~and~9~arguments~for~a~single~function.
+  }
+\msg_kernel_new:nnnn { kernel } { command-already-defined }
+  { Control~sequence~#1~already~defined. }
+  {
+    \c_msg_coding_error_text_tl
+    LaTeX~has~been~asked~to~create~a~new~control~sequence~'#1'~
+    but~this~name~has~already~been~used~elsewhere. \\ \\
+    The~current~meaning~is:\\
+    \ \ #2
+  }
+\msg_kernel_new:nnnn { kernel } { command-not-defined }
+  { Control~sequence~#1~undefined. }
+  {
+    \c_msg_coding_error_text_tl
+    LaTeX~has~been~asked~to~use~a~command~#1,~but~this~has~not~
+    been~defined~yet.
+  }
+\msg_kernel_new:nnnn { kernel } { variable-not-defined }
+  { Variable~#1~undefined. }
+  {
+    \c_msg_coding_error_text_tl
+    LaTeX~has~been~asked~to~show~a~variable~#1,~but~this~has~not~
+    been~defined~yet.
+  }
+\msg_kernel_new:nnnn { seq } { empty-sequence }
+  { Empty~sequence~#1. }
+  {
+    \c_msg_coding_error_text_tl
+    LaTeX~has~been~asked~to~recover~an~entry~from~a~sequence~that~
+    has~no~content:~that~cannot~happen!
+  }
+\cs_set_protected:Npn \msg_kernel_bug:x #1
+  {
+    \msg_interrupt:xxx { \c_msg_kernel_bug_text_tl }
+      {
+        #1
+       \msg_see_documentation_text:n { LaTeX3 }
+      }
+      { \c_msg_kernel_bug_more_text_tl }
+  }
+\tl_const:Nn \c_msg_kernel_bug_text_tl
+  { This~is~a~LaTeX~bug:~check~coding! }
+\tl_const:Nn \c_msg_kernel_bug_more_text_tl
+  {
+    There~is~a~coding~bug~somewhere~around~here. \\
+    This~probably~needs~examining~by~an~expert.
+    \c_msg_return_text_tl
+  }
+\cs_new_eq:NN \msg_class_new:nn \msg_class_set:nn
+\cs_new_eq:NN \msg_trace:nnxxxx \msg_log:nnxxxx
+\cs_new_eq:NN \msg_trace:nnxxx  \msg_log:nnxxx
+\cs_new_eq:NN \msg_trace:nnxx   \msg_log:nnxx
+\cs_new_eq:NN \msg_trace:nnx    \msg_log:nnx
+\cs_new_eq:NN \msg_trace:nn     \msg_log:nn
+\cs_new_protected:Npn \msg_generic_new:nnn #1#2#3 { \deprecated }
+\cs_new_protected:Npn \msg_generic_new:nn  #1#2   { \deprecated }
+\cs_new_protected:Npn \msg_generic_set:nnn #1#2#3 { \deprecated }
+\cs_new_protected:Npn \msg_generic_set:nn  #1#2   { \deprecated }
+\cs_new_protected:Npn \msg_direct_interrupt:xxxxx #1#2#3#4#5 { \deprecated }
+\cs_new_protected:Npn \msg_direct_log:xx #1#2  { \deprecated }
+\cs_new_protected:Npn \msg_direct_term:xx #1#2 { \deprecated }
+%% 
+%%
+%% End of file `l3msg.sty'.
diff --git a/Master/texmf-dist/tex/latex/l3kernel/l3names.sty b/Master/texmf-dist/tex/latex/l3kernel/l3names.sty
new file mode 100644
index 00000000000..a504049d1ed
--- /dev/null
+++ b/Master/texmf-dist/tex/latex/l3kernel/l3names.sty
@@ -0,0 +1,489 @@
+%%
+%% This is file `l3names.sty',
+%% generated with the docstrip utility.
+%%
+%% The original source files were:
+%%
+%% l3names.dtx  (with options: `package')
+%% 
+%% EXPERIMENTAL CODE
+%% 
+%% Do not distribute this file without also distributing the
+%% source files specified above.
+%% 
+%% Do not distribute a modified version of this file.
+%% 
+%% File: l3names.dtx Copyright (C) 1990-2011 The LaTeX3 project
+%%
+%% It may be distributed and/or modified under the conditions of the
+%% LaTeX Project Public License (LPPL), either version 1.3c of this
+%% license or (at your option) any later version.  The latest version
+%% of this license is in the file
+%%
+%%    http://www.latex-project.org/lppl.txt
+%%
+%% This file is part of the "expl3 bundle" (The Work in LPPL)
+%% and all files in that bundle must be distributed together.
+%%
+%% The released version of this bundle is available from CTAN.
+%%
+%% -----------------------------------------------------------------------
+%%
+%% The development version of the bundle can be found at
+%%
+%%    http://www.latex-project.org/svnroot/experimental/trunk/
+%%
+%% for those people who are interested.
+%%
+%%%%%%%%%%%
+%% NOTE: %%
+%%%%%%%%%%%
+%%
+%%   Snapshots taken from the repository represent work in progress and may
+%%   not work or may contain conflicting material!  We therefore ask
+%%   people _not_ to put them into distributions, archives, etc. without
+%%   prior consultation with the LaTeX3 Project.
+%%
+%% -----------------------------------------------------------------------
+\RequirePackage{l3bootstrap}
+\GetIdInfo$Id: l3names.dtx 2478 2011-06-19 21:34:23Z joseph $
+  {L3 Experimental namespace for primitives}
+\ProvidesExplPackage
+  {\ExplFileName}{\ExplFileDate}{\ExplFileVersion}{\ExplFileDescription}
+\let \tex_global:D \global
+\let \tex_let:D    \let
+\begingroup
+  \long \def \name_primitive:NN #1#2
+    {
+      \tex_global:D \tex_let:D #2 #1
+    }
+  \name_primitive:NN \                      \tex_space:D
+  \name_primitive:NN \/                     \tex_italiccor:D
+  \name_primitive:NN \-                     \tex_hyphen:D
+  \name_primitive:NN \let                   \tex_let:D
+  \name_primitive:NN \def                   \tex_def:D
+  \name_primitive:NN \edef                  \tex_edef:D
+  \name_primitive:NN \gdef                  \tex_gdef:D
+  \name_primitive:NN \xdef                  \tex_xdef:D
+  \name_primitive:NN \chardef               \tex_chardef:D
+  \name_primitive:NN \countdef              \tex_countdef:D
+  \name_primitive:NN \dimendef              \tex_dimendef:D
+  \name_primitive:NN \skipdef               \tex_skipdef:D
+  \name_primitive:NN \muskipdef             \tex_muskipdef:D
+  \name_primitive:NN \mathchardef           \tex_mathchardef:D
+  \name_primitive:NN \toksdef               \tex_toksdef:D
+  \name_primitive:NN \futurelet             \tex_futurelet:D
+  \name_primitive:NN \advance               \tex_advance:D
+  \name_primitive:NN \divide                \tex_divide:D
+  \name_primitive:NN \multiply              \tex_multiply:D
+  \name_primitive:NN \font                  \tex_font:D
+  \name_primitive:NN \fam                   \tex_fam:D
+  \name_primitive:NN \global                \tex_global:D
+  \name_primitive:NN \long                  \tex_long:D
+  \name_primitive:NN \outer                 \tex_outer:D
+  \name_primitive:NN \setlanguage           \tex_setlanguage:D
+  \name_primitive:NN \globaldefs            \tex_globaldefs:D
+  \name_primitive:NN \afterassignment       \tex_afterassignment:D
+  \name_primitive:NN \aftergroup            \tex_aftergroup:D
+  \name_primitive:NN \expandafter           \tex_expandafter:D
+  \name_primitive:NN \noexpand              \tex_noexpand:D
+  \name_primitive:NN \begingroup            \tex_begingroup:D
+  \name_primitive:NN \endgroup              \tex_endgroup:D
+  \name_primitive:NN \halign                \tex_halign:D
+  \name_primitive:NN \valign                \tex_valign:D
+  \name_primitive:NN \cr                    \tex_cr:D
+  \name_primitive:NN \crcr                  \tex_crcr:D
+  \name_primitive:NN \noalign               \tex_noalign:D
+  \name_primitive:NN \omit                  \tex_omit:D
+  \name_primitive:NN \span                  \tex_span:D
+  \name_primitive:NN \tabskip               \tex_tabskip:D
+  \name_primitive:NN \everycr               \tex_everycr:D
+  \name_primitive:NN \if                    \tex_if:D
+  \name_primitive:NN \ifcase                \tex_ifcase:D
+  \name_primitive:NN \ifcat                 \tex_ifcat:D
+  \name_primitive:NN \ifnum                 \tex_ifnum:D
+  \name_primitive:NN \ifodd                 \tex_ifodd:D
+  \name_primitive:NN \ifdim                 \tex_ifdim:D
+  \name_primitive:NN \ifeof                 \tex_ifeof:D
+  \name_primitive:NN \ifhbox                \tex_ifhbox:D
+  \name_primitive:NN \ifvbox                \tex_ifvbox:D
+  \name_primitive:NN \ifvoid                \tex_ifvoid:D
+  \name_primitive:NN \ifx                   \tex_ifx:D
+  \name_primitive:NN \iffalse               \tex_iffalse:D
+  \name_primitive:NN \iftrue                \tex_iftrue:D
+  \name_primitive:NN \ifhmode               \tex_ifhmode:D
+  \name_primitive:NN \ifmmode               \tex_ifmmode:D
+  \name_primitive:NN \ifvmode               \tex_ifvmode:D
+  \name_primitive:NN \ifinner               \tex_ifinner:D
+  \name_primitive:NN \else                  \tex_else:D
+  \name_primitive:NN \fi                    \tex_fi:D
+  \name_primitive:NN \or                    \tex_or:D
+  \name_primitive:NN \immediate             \tex_immediate:D
+  \name_primitive:NN \closeout              \tex_closeout:D
+  \name_primitive:NN \openin                \tex_openin:D
+  \name_primitive:NN \openout               \tex_openout:D
+  \name_primitive:NN \read                  \tex_read:D
+  \name_primitive:NN \write                 \tex_write:D
+  \name_primitive:NN \closein               \tex_closein:D
+  \name_primitive:NN \newlinechar           \tex_newlinechar:D
+  \name_primitive:NN \input                 \tex_input:D
+  \name_primitive:NN \endinput              \tex_endinput:D
+  \name_primitive:NN \inputlineno           \tex_inputlineno:D
+  \name_primitive:NN \errmessage            \tex_errmessage:D
+  \name_primitive:NN \message               \tex_message:D
+  \name_primitive:NN \show                  \tex_show:D
+  \name_primitive:NN \showthe               \tex_showthe:D
+  \name_primitive:NN \showbox               \tex_showbox:D
+  \name_primitive:NN \showlists             \tex_showlists:D
+  \name_primitive:NN \errhelp               \tex_errhelp:D
+  \name_primitive:NN \errorcontextlines     \tex_errorcontextlines:D
+  \name_primitive:NN \tracingcommands       \tex_tracingcommands:D
+  \name_primitive:NN \tracinglostchars      \tex_tracinglostchars:D
+  \name_primitive:NN \tracingmacros         \tex_tracingmacros:D
+  \name_primitive:NN \tracingonline         \tex_tracingonline:D
+  \name_primitive:NN \tracingoutput         \tex_tracingoutput:D
+  \name_primitive:NN \tracingpages          \tex_tracingpages:D
+  \name_primitive:NN \tracingparagraphs     \tex_tracingparagraphs:D
+  \name_primitive:NN \tracingrestores       \tex_tracingrestores:D
+  \name_primitive:NN \tracingstats          \tex_tracingstats:D
+  \name_primitive:NN \pausing               \tex_pausing:D
+  \name_primitive:NN \showboxbreadth        \tex_showboxbreadth:D
+  \name_primitive:NN \showboxdepth          \tex_showboxdepth:D
+  \name_primitive:NN \batchmode             \tex_batchmode:D
+  \name_primitive:NN \errorstopmode         \tex_errorstopmode:D
+  \name_primitive:NN \nonstopmode           \tex_nonstopmode:D
+  \name_primitive:NN \scrollmode            \tex_scrollmode:D
+  \name_primitive:NN \end                   \tex_end:D
+  \name_primitive:NN \csname                \tex_csname:D
+  \name_primitive:NN \endcsname             \tex_endcsname:D
+  \name_primitive:NN \ignorespaces          \tex_ignorespaces:D
+  \name_primitive:NN \relax                 \tex_relax:D
+  \name_primitive:NN \the                   \tex_the:D
+  \name_primitive:NN \mag                   \tex_mag:D
+  \name_primitive:NN \language              \tex_language:D
+  \name_primitive:NN \mark                  \tex_mark:D
+  \name_primitive:NN \topmark               \tex_topmark:D
+  \name_primitive:NN \firstmark             \tex_firstmark:D
+  \name_primitive:NN \botmark               \tex_botmark:D
+  \name_primitive:NN \splitfirstmark        \tex_splitfirstmark:D
+  \name_primitive:NN \splitbotmark          \tex_splitbotmark:D
+  \name_primitive:NN \fontname              \tex_fontname:D
+  \name_primitive:NN \escapechar            \tex_escapechar:D
+  \name_primitive:NN \endlinechar           \tex_endlinechar:D
+  \name_primitive:NN \mathchoice            \tex_mathchoice:D
+  \name_primitive:NN \delimiter             \tex_delimiter:D
+  \name_primitive:NN \mathaccent            \tex_mathaccent:D
+  \name_primitive:NN \mathchar              \tex_mathchar:D
+  \name_primitive:NN \mskip                 \tex_mskip:D
+  \name_primitive:NN \radical               \tex_radical:D
+  \name_primitive:NN \vcenter               \tex_vcenter:D
+  \name_primitive:NN \mkern                 \tex_mkern:D
+  \name_primitive:NN \above                 \tex_above:D
+  \name_primitive:NN \abovewithdelims       \tex_abovewithdelims:D
+  \name_primitive:NN \atop                  \tex_atop:D
+  \name_primitive:NN \atopwithdelims        \tex_atopwithdelims:D
+  \name_primitive:NN \over                  \tex_over:D
+  \name_primitive:NN \overwithdelims        \tex_overwithdelims:D
+  \name_primitive:NN \displaystyle          \tex_displaystyle:D
+  \name_primitive:NN \textstyle             \tex_textstyle:D
+  \name_primitive:NN \scriptstyle           \tex_scriptstyle:D
+  \name_primitive:NN \scriptscriptstyle     \tex_scriptscriptstyle:D
+  \name_primitive:NN \nonscript             \tex_nonscript:D
+  \name_primitive:NN \eqno                  \tex_eqno:D
+  \name_primitive:NN \leqno                 \tex_leqno:D
+  \name_primitive:NN \abovedisplayshortskip \tex_abovedisplayshortskip:D
+  \name_primitive:NN \abovedisplayskip      \tex_abovedisplayskip:D
+  \name_primitive:NN \belowdisplayshortskip \tex_belowdisplayshortskip:D
+  \name_primitive:NN \belowdisplayskip      \tex_belowdisplayskip:D
+  \name_primitive:NN \displaywidowpenalty   \tex_displaywidowpenalty:D
+  \name_primitive:NN \displayindent         \tex_displayindent:D
+  \name_primitive:NN \displaywidth          \tex_displaywidth:D
+  \name_primitive:NN \everydisplay          \tex_everydisplay:D
+  \name_primitive:NN \predisplaysize        \tex_predisplaysize:D
+  \name_primitive:NN \predisplaypenalty     \tex_predisplaypenalty:D
+  \name_primitive:NN \postdisplaypenalty    \tex_postdisplaypenalty:D
+  \name_primitive:NN \mathbin               \tex_mathbin:D
+  \name_primitive:NN \mathclose             \tex_mathclose:D
+  \name_primitive:NN \mathinner             \tex_mathinner:D
+  \name_primitive:NN \mathop                \tex_mathop:D
+  \name_primitive:NN \displaylimits         \tex_displaylimits:D
+  \name_primitive:NN \limits                \tex_limits:D
+  \name_primitive:NN \nolimits              \tex_nolimits:D
+  \name_primitive:NN \mathopen              \tex_mathopen:D
+  \name_primitive:NN \mathord               \tex_mathord:D
+  \name_primitive:NN \mathpunct             \tex_mathpunct:D
+  \name_primitive:NN \mathrel               \tex_mathrel:D
+  \name_primitive:NN \overline              \tex_overline:D
+  \name_primitive:NN \underline             \tex_underline:D
+  \name_primitive:NN \left                  \tex_left:D
+  \name_primitive:NN \right                 \tex_right:D
+  \name_primitive:NN \binoppenalty          \tex_binoppenalty:D
+  \name_primitive:NN \relpenalty            \tex_relpenalty:D
+  \name_primitive:NN \delimitershortfall    \tex_delimitershortfall:D
+  \name_primitive:NN \delimiterfactor       \tex_delimiterfactor:D
+  \name_primitive:NN \nulldelimiterspace    \tex_nulldelimiterspace:D
+  \name_primitive:NN \everymath             \tex_everymath:D
+  \name_primitive:NN \mathsurround          \tex_mathsurround:D
+  \name_primitive:NN \medmuskip             \tex_medmuskip:D
+  \name_primitive:NN \thinmuskip            \tex_thinmuskip:D
+  \name_primitive:NN \thickmuskip           \tex_thickmuskip:D
+  \name_primitive:NN \scriptspace           \tex_scriptspace:D
+  \name_primitive:NN \noboundary            \tex_noboundary:D
+  \name_primitive:NN \accent                \tex_accent:D
+  \name_primitive:NN \char                  \tex_char:D
+  \name_primitive:NN \discretionary         \tex_discretionary:D
+  \name_primitive:NN \hfil                  \tex_hfil:D
+  \name_primitive:NN \hfilneg               \tex_hfilneg:D
+  \name_primitive:NN \hfill                 \tex_hfill:D
+  \name_primitive:NN \hskip                 \tex_hskip:D
+  \name_primitive:NN \hss                   \tex_hss:D
+  \name_primitive:NN \vfil                  \tex_vfil:D
+  \name_primitive:NN \vfilneg               \tex_vfilneg:D
+  \name_primitive:NN \vfill                 \tex_vfill:D
+  \name_primitive:NN \vskip                 \tex_vskip:D
+  \name_primitive:NN \vss                   \tex_vss:D
+  \name_primitive:NN \unskip                \tex_unskip:D
+  \name_primitive:NN \kern                  \tex_kern:D
+  \name_primitive:NN \unkern                \tex_unkern:D
+  \name_primitive:NN \hrule                 \tex_hrule:D
+  \name_primitive:NN \vrule                 \tex_vrule:D
+  \name_primitive:NN \leaders               \tex_leaders:D
+  \name_primitive:NN \cleaders              \tex_cleaders:D
+  \name_primitive:NN \xleaders              \tex_xleaders:D
+  \name_primitive:NN \lastkern              \tex_lastkern:D
+  \name_primitive:NN \lastskip              \tex_lastskip:D
+  \name_primitive:NN \indent                \tex_indent:D
+  \name_primitive:NN \par                   \tex_par:D
+  \name_primitive:NN \noindent              \tex_noindent:D
+  \name_primitive:NN \vadjust               \tex_vadjust:D
+  \name_primitive:NN \baselineskip          \tex_baselineskip:D
+  \name_primitive:NN \lineskip              \tex_lineskip:D
+  \name_primitive:NN \lineskiplimit         \tex_lineskiplimit:D
+  \name_primitive:NN \clubpenalty           \tex_clubpenalty:D
+  \name_primitive:NN \widowpenalty          \tex_widowpenalty:D
+  \name_primitive:NN \exhyphenpenalty       \tex_exhyphenpenalty:D
+  \name_primitive:NN \hyphenpenalty         \tex_hyphenpenalty:D
+  \name_primitive:NN \linepenalty           \tex_linepenalty:D
+  \name_primitive:NN \doublehyphendemerits  \tex_doublehyphendemerits:D
+  \name_primitive:NN \finalhyphendemerits   \tex_finalhyphendemerits:D
+  \name_primitive:NN \adjdemerits           \tex_adjdemerits:D
+  \name_primitive:NN \hangafter             \tex_hangafter:D
+  \name_primitive:NN \hangindent            \tex_hangindent:D
+  \name_primitive:NN \parshape              \tex_parshape:D
+  \name_primitive:NN \hsize                 \tex_hsize:D
+  \name_primitive:NN \lefthyphenmin         \tex_lefthyphenmin:D
+  \name_primitive:NN \righthyphenmin        \tex_righthyphenmin:D
+  \name_primitive:NN \leftskip              \tex_leftskip:D
+  \name_primitive:NN \rightskip             \tex_rightskip:D
+  \name_primitive:NN \looseness             \tex_looseness:D
+  \name_primitive:NN \parskip               \tex_parskip:D
+  \name_primitive:NN \parindent             \tex_parindent:D
+  \name_primitive:NN \uchyph                \tex_uchyph:D
+  \name_primitive:NN \emergencystretch      \tex_emergencystretch:D
+  \name_primitive:NN \pretolerance          \tex_pretolerance:D
+  \name_primitive:NN \tolerance             \tex_tolerance:D
+  \name_primitive:NN \spaceskip             \tex_spaceskip:D
+  \name_primitive:NN \xspaceskip            \tex_xspaceskip:D
+  \name_primitive:NN \parfillskip           \tex_parfillskip:D
+  \name_primitive:NN \everypar              \tex_everypar:D
+  \name_primitive:NN \prevgraf              \tex_prevgraf:D
+  \name_primitive:NN \spacefactor           \tex_spacefactor:D
+  \name_primitive:NN \shipout               \tex_shipout:D
+  \name_primitive:NN \vsize                 \tex_vsize:D
+  \name_primitive:NN \interlinepenalty      \tex_interlinepenalty:D
+  \name_primitive:NN \brokenpenalty         \tex_brokenpenalty:D
+  \name_primitive:NN \topskip               \tex_topskip:D
+  \name_primitive:NN \maxdeadcycles         \tex_maxdeadcycles:D
+  \name_primitive:NN \maxdepth              \tex_maxdepth:D
+  \name_primitive:NN \output                \tex_output:D
+  \name_primitive:NN \deadcycles            \tex_deadcycles:D
+  \name_primitive:NN \pagedepth             \tex_pagedepth:D
+  \name_primitive:NN \pagestretch           \tex_pagestretch:D
+  \name_primitive:NN \pagefilstretch        \tex_pagefilstretch:D
+  \name_primitive:NN \pagefillstretch       \tex_pagefillstretch:D
+  \name_primitive:NN \pagefilllstretch      \tex_pagefilllstretch:D
+  \name_primitive:NN \pageshrink            \tex_pageshrink:D
+  \name_primitive:NN \pagegoal              \tex_pagegoal:D
+  \name_primitive:NN \pagetotal             \tex_pagetotal:D
+  \name_primitive:NN \outputpenalty         \tex_outputpenalty:D
+  \name_primitive:NN \hoffset               \tex_hoffset:D
+  \name_primitive:NN \voffset               \tex_voffset:D
+  \name_primitive:NN \insert                \tex_insert:D
+  \name_primitive:NN \holdinginserts        \tex_holdinginserts:D
+  \name_primitive:NN \floatingpenalty       \tex_floatingpenalty:D
+  \name_primitive:NN \insertpenalties       \tex_insertpenalties:D
+  \name_primitive:NN \lower                 \tex_lower:D
+  \name_primitive:NN \moveleft              \tex_moveleft:D
+  \name_primitive:NN \moveright             \tex_moveright:D
+  \name_primitive:NN \raise                 \tex_raise:D
+  \name_primitive:NN \copy                  \tex_copy:D
+  \name_primitive:NN \lastbox               \tex_lastbox:D
+  \name_primitive:NN \vsplit                \tex_vsplit:D
+  \name_primitive:NN \unhbox                \tex_unhbox:D
+  \name_primitive:NN \unhcopy               \tex_unhcopy:D
+  \name_primitive:NN \unvbox                \tex_unvbox:D
+  \name_primitive:NN \unvcopy               \tex_unvcopy:D
+  \name_primitive:NN \setbox                \tex_setbox:D
+  \name_primitive:NN \hbox                  \tex_hbox:D
+  \name_primitive:NN \vbox                  \tex_vbox:D
+  \name_primitive:NN \vtop                  \tex_vtop:D
+  \name_primitive:NN \prevdepth             \tex_prevdepth:D
+  \name_primitive:NN \badness               \tex_badness:D
+  \name_primitive:NN \hbadness              \tex_hbadness:D
+  \name_primitive:NN \vbadness              \tex_vbadness:D
+  \name_primitive:NN \hfuzz                 \tex_hfuzz:D
+  \name_primitive:NN \vfuzz                 \tex_vfuzz:D
+  \name_primitive:NN \overfullrule          \tex_overfullrule:D
+  \name_primitive:NN \boxmaxdepth           \tex_boxmaxdepth:D
+  \name_primitive:NN \splitmaxdepth         \tex_splitmaxdepth:D
+  \name_primitive:NN \splittopskip          \tex_splittopskip:D
+  \name_primitive:NN \everyhbox             \tex_everyhbox:D
+  \name_primitive:NN \everyvbox             \tex_everyvbox:D
+  \name_primitive:NN \nullfont              \tex_nullfont:D
+  \name_primitive:NN \textfont              \tex_textfont:D
+  \name_primitive:NN \scriptfont            \tex_scriptfont:D
+  \name_primitive:NN \scriptscriptfont      \tex_scriptscriptfont:D
+  \name_primitive:NN \fontdimen             \tex_fontdimen:D
+  \name_primitive:NN \hyphenchar            \tex_hyphenchar:D
+  \name_primitive:NN \skewchar              \tex_skewchar:D
+  \name_primitive:NN \defaulthyphenchar     \tex_defaulthyphenchar:D
+  \name_primitive:NN \defaultskewchar       \tex_defaultskewchar:D
+  \name_primitive:NN \number                \tex_number:D
+  \name_primitive:NN \romannumeral          \tex_romannumeral:D
+  \name_primitive:NN \string                \tex_string:D
+  \name_primitive:NN \lowercase             \tex_lowercase:D
+  \name_primitive:NN \uppercase             \tex_uppercase:D
+  \name_primitive:NN \meaning               \tex_meaning:D
+  \name_primitive:NN \penalty               \tex_penalty:D
+  \name_primitive:NN \unpenalty             \tex_unpenalty:D
+  \name_primitive:NN \lastpenalty           \tex_lastpenalty:D
+  \name_primitive:NN \special               \tex_special:D
+  \name_primitive:NN \dump                  \tex_dump:D
+  \name_primitive:NN \patterns              \tex_patterns:D
+  \name_primitive:NN \hyphenation           \tex_hyphenation:D
+  \name_primitive:NN \time                  \tex_time:D
+  \name_primitive:NN \day                   \tex_day:D
+  \name_primitive:NN \month                 \tex_month:D
+  \name_primitive:NN \year                  \tex_year:D
+  \name_primitive:NN \jobname               \tex_jobname:D
+  \name_primitive:NN \everyjob              \tex_everyjob:D
+  \name_primitive:NN \count                 \tex_count:D
+  \name_primitive:NN \dimen                 \tex_dimen:D
+  \name_primitive:NN \skip                  \tex_skip:D
+  \name_primitive:NN \toks                  \tex_toks:D
+  \name_primitive:NN \muskip                \tex_muskip:D
+  \name_primitive:NN \box                   \tex_box:D
+  \name_primitive:NN \wd                    \tex_wd:D
+  \name_primitive:NN \ht                    \tex_ht:D
+  \name_primitive:NN \dp                    \tex_dp:D
+  \name_primitive:NN \catcode               \tex_catcode:D
+  \name_primitive:NN \delcode               \tex_delcode:D
+  \name_primitive:NN \sfcode                \tex_sfcode:D
+  \name_primitive:NN \lccode                \tex_lccode:D
+  \name_primitive:NN \uccode                \tex_uccode:D
+  \name_primitive:NN \mathcode              \tex_mathcode:D
+  \name_primitive:NN \ifdefined             \etex_ifdefined:D
+  \name_primitive:NN \ifcsname              \etex_ifcsname:D
+  \name_primitive:NN \unless                \etex_unless:D
+  \name_primitive:NN \eTeXversion           \etex_eTeXversion:D
+  \name_primitive:NN \eTeXrevision          \etex_eTeXrevision:D
+  \name_primitive:NN \marks                 \etex_marks:D
+  \name_primitive:NN \topmarks              \etex_topmarks:D
+  \name_primitive:NN \firstmarks            \etex_firstmarks:D
+  \name_primitive:NN \botmarks              \etex_botmarks:D
+  \name_primitive:NN \splitfirstmarks       \etex_splitfirstmarks:D
+  \name_primitive:NN \splitbotmarks         \etex_splitbotmarks:D
+  \name_primitive:NN \unexpanded            \etex_unexpanded:D
+  \name_primitive:NN \detokenize            \etex_detokenize:D
+  \name_primitive:NN \scantokens            \etex_scantokens:D
+  \name_primitive:NN \showtokens            \etex_showtokens:D
+  \name_primitive:NN \readline              \etex_readline:D
+  \name_primitive:NN \tracingassigns        \etex_tracingassigns:D
+  \name_primitive:NN \tracingscantokens     \etex_tracingscantokens:D
+  \name_primitive:NN \tracingnesting        \etex_tracingnesting:D
+  \name_primitive:NN \tracingifs            \etex_tracingifs:D
+  \name_primitive:NN \currentiflevel        \etex_currentiflevel:D
+  \name_primitive:NN \currentifbranch       \etex_currentifbranch:D
+  \name_primitive:NN \currentiftype         \etex_currentiftype:D
+  \name_primitive:NN \tracinggroups         \etex_tracinggroups:D
+  \name_primitive:NN \currentgrouplevel     \etex_currentgrouplevel:D
+  \name_primitive:NN \currentgrouptype      \etex_currentgrouptype:D
+  \name_primitive:NN \showgroups            \etex_showgroups:D
+  \name_primitive:NN \showifs               \etex_showifs:D
+  \name_primitive:NN \interactionmode       \etex_interactionmode:D
+  \name_primitive:NN \lastnodetype          \etex_lastnodetype:D
+  \name_primitive:NN \iffontchar            \etex_iffontchar:D
+  \name_primitive:NN \fontcharht            \etex_fontcharht:D
+  \name_primitive:NN \fontchardp            \etex_fontchardp:D
+  \name_primitive:NN \fontcharwd            \etex_fontcharwd:D
+  \name_primitive:NN \fontcharic            \etex_fontcharic:D
+  \name_primitive:NN \parshapeindent        \etex_parshapeindent:D
+  \name_primitive:NN \parshapelength        \etex_parshapelength:D
+  \name_primitive:NN \parshapedimen         \etex_parshapedimen:D
+  \name_primitive:NN \numexpr               \etex_numexpr:D
+  \name_primitive:NN \dimexpr               \etex_dimexpr:D
+  \name_primitive:NN \glueexpr              \etex_glueexpr:D
+  \name_primitive:NN \muexpr                \etex_muexpr:D
+  \name_primitive:NN \gluestretch           \etex_gluestretch:D
+  \name_primitive:NN \glueshrink            \etex_glueshrink:D
+  \name_primitive:NN \gluestretchorder      \etex_gluestretchorder:D
+  \name_primitive:NN \glueshrinkorder       \etex_glueshrinkorder:D
+  \name_primitive:NN \gluetomu              \etex_gluetomu:D
+  \name_primitive:NN \mutoglue              \etex_mutoglue:D
+  \name_primitive:NN \lastlinefit           \etex_lastlinefit:D
+  \name_primitive:NN \interlinepenalties    \etex_interlinepenalties:D
+  \name_primitive:NN \clubpenalties         \etex_clubpenalties:D
+  \name_primitive:NN \widowpenalties        \etex_widowpenalties:D
+  \name_primitive:NN \displaywidowpenalties \etex_displaywidowpenalties:D
+  \name_primitive:NN \middle                \etex_middle:D
+  \name_primitive:NN \savinghyphcodes       \etex_savinghyphcodes:D
+  \name_primitive:NN \savingvdiscards       \etex_savingvdiscards:D
+  \name_primitive:NN \pagediscards          \etex_pagediscards:D
+  \name_primitive:NN \splitdiscards         \etex_splitdiscards:D
+  \name_primitive:NN \TeXXETstate           \etex_TeXXETstate:D
+  \name_primitive:NN \beginL                \etex_beginL:D
+  \name_primitive:NN \endL                  \etex_endL:D
+  \name_primitive:NN \beginR                \etex_beginR:D
+  \name_primitive:NN \endR                  \etex_endR:D
+  \name_primitive:NN \predisplaydirection   \etex_predisplaydirection:D
+  \name_primitive:NN \everyeof              \etex_everyeof:D
+  \name_primitive:NN \protected             \etex_protected:D
+  \name_primitive:NN \pdfcreationdate       \pdftex_pdfcreationdate:D
+  \name_primitive:NN \pdfcolorstack         \pdftex_pdfcolorstack:D
+  \name_primitive:NN \pdfcompresslevel      \pdftex_pdfcompresslevel:D
+  \name_primitive:NN \pdfdecimaldigits      \pdftex_pdfdecimaldigits:D
+  \name_primitive:NN \pdfhorigin            \pdftex_pdfhorigin:D
+  \name_primitive:NN \pdfinfo               \pdftex_pdfinfo:D
+  \name_primitive:NN \pdfliteral            \pdftex_pdfliteral:D
+  \name_primitive:NN \pdfminorversion       \pdftex_pdfminorversion:D
+  \name_primitive:NN \pdfobjcompresslevel   \pdftex_pdfobjcompresslevel:D
+  \name_primitive:NN \pdfoutput             \pdftex_pdfoutput:D
+  \name_primitive:NN \pdfrestore            \pdftex_pdfrestore:D
+  \name_primitive:NN \pdfsave               \pdftex_pdfsave:D
+  \name_primitive:NN \pdfsetmatrix          \pdftex_pdfsetmatrix:D
+  \name_primitive:NN \pdfpkresolution       \pdftex_pdfpkresolution:D
+  \name_primitive:NN \pdftexrevision        \pdftex_pdftextrevision:D
+  \name_primitive:NN \pdfvorigin            \pdftex_pdfvorigin:D
+  \name_primitive:NN \pdfstrcmp             \pdftex_strcmp:D
+  \name_primitive:NN \XeTeXversion          \xetex_XeTeXversion:D
+  \name_primitive:NN \catcodetable          \luatex_catcodetable:D
+  \name_primitive:NN \directlua             \luatex_directlua:D
+  \name_primitive:NN \initcatcodetable      \luatex_initcatcodetable:D
+  \name_primitive:NN \latelua               \luatex_latelua:D
+  \name_primitive:NN \luatexversion         \luatex_luatexversion:D
+  \name_primitive:NN \savecatcodetable      \luatex_savecatcodetable:D
+\tex_endgroup:D
+\tex_let:D \tex_end:D               \@@end
+\tex_let:D \tex_everydisplay:D      \frozen@everydisplay
+\tex_let:D \tex_everymath:D         \frozen@everymath
+\tex_let:D \tex_hyphen:D            \@@hyph
+\tex_let:D \tex_input:D             \@@input
+\tex_let:D \tex_italic_correction:D \@@italiccorr
+\tex_let:D \tex_underline:D         \@@underline
+\tex_let:D \luatex_catcodetable:D      \luatexcatcodetable
+\tex_let:D \luatex_initcatcodetable:D  \luatexinitcatcodetable
+\tex_let:D \luatex_latelua:D           \luatexlatelua
+\tex_let:D \luatex_savecatcodetable:D  \luatexsavecatcodetable
+%% 
+%%
+%% End of file `l3names.sty'.
diff --git a/Master/texmf-dist/tex/latex/l3kernel/l3prg.sty b/Master/texmf-dist/tex/latex/l3kernel/l3prg.sty
new file mode 100644
index 00000000000..4a20c040547
--- /dev/null
+++ b/Master/texmf-dist/tex/latex/l3kernel/l3prg.sty
@@ -0,0 +1,563 @@
+%%
+%% This is file `l3prg.sty',
+%% generated with the docstrip utility.
+%%
+%% The original source files were:
+%%
+%% l3prg.dtx  (with options: `package')
+%% 
+%% EXPERIMENTAL CODE
+%% 
+%% Do not distribute this file without also distributing the
+%% source files specified above.
+%% 
+%% Do not distribute a modified version of this file.
+%% 
+%% File: l3prg.dtx Copyright (C) 2005-2011 The LaTeX3 Project
+%%
+%% It may be distributed and/or modified under the conditions of the
+%% LaTeX Project Public License (LPPL), either version 1.3c of this
+%% license or (at your option) any later version.  The latest version
+%% of this license is in the file
+%%
+%%    http://www.latex-project.org/lppl.txt
+%%
+%% This file is part of the "expl3 bundle" (The Work in LPPL)
+%% and all files in that bundle must be distributed together.
+%%
+%% The released version of this bundle is available from CTAN.
+%%
+%% -----------------------------------------------------------------------
+%%
+%% The development version of the bundle can be found at
+%%
+%%    http://www.latex-project.org/svnroot/experimental/trunk/
+%%
+%% for those people who are interested.
+%%
+%%%%%%%%%%%
+%% NOTE: %%
+%%%%%%%%%%%
+%%
+%%   Snapshots taken from the repository represent work in progress and may
+%%   not work or may contain conflicting material!  We therefore ask
+%%   people _not_ to put them into distributions, archives, etc. without
+%%   prior consultation with the LaTeX3 Project.
+%%
+%% -----------------------------------------------------------------------
+\RequirePackage{l3names}
+\GetIdInfo$Id: l3prg.dtx 2495 2011-07-06 16:57:08Z bruno $
+  {L3 Experimental control structures}
+%%
+\ProvidesExplPackage
+  {\ExplFileName}{\ExplFileDate}{\ExplFileVersion}{\ExplFileDescription}
+\package_check_loaded_expl:
+\cs_new_protected_nopar:Npn \bool_new:N #1 { \cs_new_eq:NN #1 \c_false_bool }
+\cs_generate_variant:Nn \bool_new:N { c }
+\cs_new_protected_nopar:Npn \bool_set_true:N #1
+  { \cs_set_eq:NN #1 \c_true_bool }
+\cs_new_protected_nopar:Npn \bool_set_false:N #1
+  { \cs_set_eq:NN #1 \c_false_bool }
+\cs_new_protected_nopar:Npn \bool_gset_true:N #1
+  { \cs_gset_eq:NN #1 \c_true_bool }
+\cs_new_protected_nopar:Npn \bool_gset_false:N #1
+  { \cs_gset_eq:NN #1 \c_false_bool }
+\cs_generate_variant:Nn \bool_set_true:N   { c }
+\cs_generate_variant:Nn \bool_set_false:N  { c }
+\cs_generate_variant:Nn \bool_gset_true:N  { c }
+\cs_generate_variant:Nn \bool_gset_false:N { c }
+\cs_new_eq:NN \bool_set_eq:NN  \cs_set_eq:NN
+\cs_new_eq:NN \bool_set_eq:Nc  \cs_set_eq:Nc
+\cs_new_eq:NN \bool_set_eq:cN  \cs_set_eq:cN
+\cs_new_eq:NN \bool_set_eq:cc  \cs_set_eq:cc
+\cs_new_eq:NN \bool_gset_eq:NN \cs_gset_eq:NN
+\cs_new_eq:NN \bool_gset_eq:Nc \cs_gset_eq:Nc
+\cs_new_eq:NN \bool_gset_eq:cN \cs_gset_eq:cN
+\cs_new_eq:NN \bool_gset_eq:cc \cs_gset_eq:cc
+\cs_new:Npn \bool_set:Nn #1#2
+  { \tex_chardef:D #1 = \bool_if_p:n {#2} }
+\cs_new:Npn \bool_gset:Nn #1#2
+  { \tex_global:D \tex_chardef:D #1 = \bool_if_p:n {#2} }
+\cs_generate_variant:Nn \bool_set:Nn  { c }
+\cs_generate_variant:Nn \bool_gset:Nn { c }
+\prg_new_conditional:Npnn \bool_if:N #1 { p , T , F , TF }
+  {
+    \if_bool:N #1
+      \prg_return_true:
+    \else:
+      \prg_return_false:
+    \fi:
+  }
+\cs_generate_variant:Nn \bool_if_p:N { c }
+\cs_generate_variant:Nn \bool_if:NT  { c }
+\cs_generate_variant:Nn \bool_if:NF  { c }
+\cs_generate_variant:Nn \bool_if:NTF { c }
+\bool_new:N \l_tmpa_bool
+\bool_new:N \g_tmpa_bool
+\prg_new_conditional:Npnn \bool_if:n #1 { T , F , TF }
+  {
+    \if_predicate:w \bool_if_p:n {#1}
+      \prg_return_true:
+    \else:
+      \prg_return_false:
+    \fi:
+  }
+\cs_new:Npn \bool_if_p:n #1
+  {
+    \group_align_safe_begin:
+    \bool_get_next:N ( #1 ) S
+  }
+\cs_new:Npn \bool_get_next:N #1
+  {
+    \use:c
+      {
+        bool_
+        \if_meaning:w !#1 ! \else: \if_meaning:w (#1 ( \else: p \fi: \fi:
+        :w
+      }
+      #1
+  }
+\cs_new:Npn \bool_get_not_next:N #1
+  {
+    \use:c
+      {
+      bool_not_
+      \if_meaning:w !#1 ! \else: \if_meaning:w (#1 ( \else: p \fi: \fi:
+      :w
+      }
+        #1
+  }
+\cs_new:Npn \bool_get_next:NN #1#2 { \bool_get_next:N #2 }
+\cs_new:Npn \bool_get_not_next:NN #1#2 { \bool_get_not_next:N #2 }
+\cs_new:cpn { bool_!:w } #1#2
+  {
+    \if_meaning:w ( #2
+      \exp_after:wN \bool_Not:w
+    \else:
+      \if_meaning:w ! #2
+        \exp_after:wN \exp_after:wN \exp_after:wN \bool_get_next:NN
+      \else:
+        \exp_after:wN \exp_after:wN \exp_after:wN \bool_Not:N
+      \fi:
+    \fi:
+    #2
+  }
+\cs_new:cpn { bool_not_!:w } #1#2
+  {
+    \if_meaning:w ( #2
+      \exp_after:wN \bool_not_Not:w
+    \else:
+      \if_meaning:w ! #2
+        \exp_after:wN \exp_after:wN \exp_after:wN \bool_get_not_next:NN
+      \else:
+        \exp_after:wN \exp_after:wN \exp_after:wN \bool_not_Not:N
+      \fi:
+    \fi:
+    #2
+  }
+\cs_new:Npn \bool_Not:w { \exp_after:wN \int_value:w \bool_get_not_next:N }
+\cs_new:Npn \bool_not_Not:w { \exp_after:wN \int_value:w \bool_get_next:N }
+\cs_new:Npn \bool_Not:N #1
+  {
+    \exp_after:wN \bool_p:w
+    \if_meaning:w #1 \c_true_bool
+      \c_false_bool
+    \else:
+      \c_true_bool
+    \fi:
+  }
+\cs_new:Npn \bool_not_Not:N #1
+  {
+    \exp_after:wN \bool_p:w
+    \if_meaning:w #1 \c_true_bool
+      \c_true_bool
+    \else:
+      \c_false_bool
+    \fi:
+  }
+\cs_new:cpn { bool_(:w } #1
+  { \exp_after:wN  \bool_cleanup:N \int_value:w \bool_get_next:N }
+\cs_new:cpn { bool_not_(:w } #1
+  { \exp_after:wN  \bool_not_cleanup:N \int_value:w \bool_get_next:N }
+\cs_new:cpn { bool_p:w } { \exp_after:wN \bool_cleanup:N \int_value:w  }
+\cs_new:cpn { bool_not_p:w } {\exp_after:wN \bool_not_cleanup:N \int_value:w }
+\cs_new:Npn \bool_cleanup:N #1
+  {
+    \exp_after:wN \bool_choose:NN \exp_after:wN #1
+    \int_to_roman:w - `\q
+  }
+\cs_new:Npn \bool_not_cleanup:N #1
+  {
+    \exp_after:wN \bool_not_choose:NN \exp_after:wN #1
+    \int_to_roman:w - `\q
+  }
+\cs_new_nopar:Npn \bool_choose:NN #1#2 { \use:c { bool_ #2 _ #1 :w } }
+\cs_new_nopar:Npn \bool_not_choose:NN #1#2 { \use:c { bool_not_ #2 _ #1 :w } }
+\cs_new_nopar:cpn { bool_&_1:w } & { \bool_get_next:N }
+\cs_new_nopar:cpn { bool_|_0:w } | { \bool_get_next:N }
+\cs_new_nopar:cpn { bool_not_&_0:w } & { \bool_get_next:N }
+\cs_new_nopar:cpn { bool_not_|_1:w } | { \bool_get_next:N }
+\cs_new_nopar:cpn { bool_)_0:w } { \c_false_bool }
+\cs_new_nopar:cpn { bool_)_1:w } { \c_true_bool }
+\cs_new_nopar:cpn { bool_not_)_0:w } { \c_true_bool }
+\cs_new_nopar:cpn { bool_not_)_1:w } { \c_false_bool }
+\cs_new_nopar:cpn { bool_S_0:w } { \group_align_safe_end: \c_false_bool }
+\cs_new_nopar:cpn { bool_S_1:w } { \group_align_safe_end: \c_true_bool }
+\cs_new_nopar:cpn { bool_&_0:w } & { \bool_eval_skip_to_end:Nw \c_false_bool }
+\cs_new_nopar:cpn { bool_|_1:w } | { \bool_eval_skip_to_end:Nw \c_true_bool }
+\cs_new_nopar:cpn { bool_not_&_1:w } &
+  { \bool_eval_skip_to_end:Nw \c_false_bool }
+\cs_new_nopar:cpn { bool_not_|_0:w } |
+  { \bool_eval_skip_to_end:Nw \c_true_bool }
+%% (
+\cs_new:Npn \bool_eval_skip_to_end:Nw #1#2 )
+  {
+    \bool_eval_skip_to_end_aux:Nw #1#2 ( % )
+    \q_no_value \q_stop
+    {#2}
+  }
+\cs_new:Npn \bool_eval_skip_to_end_aux:Nw #1#2 ( #3#4 \q_stop #5 % )
+  {
+    \quark_if_no_value:NTF #3
+    {#1}
+    { \bool_eval_skip_to_end_aux_ii:Nw #1 #5 }
+  }
+\cs_new:Npn \bool_eval_skip_to_end_aux_ii:Nw #1#2 ( #3 )
+  { % (
+    \bool_eval_skip_to_end:Nw #1#3 )
+  }
+\cs_new:Npn \bool_not_p:n #1 { \bool_if_p:n { ! ( #1 ) } }
+\cs_new:Npn \bool_xor_p:nn #1#2
+  {
+    \int_compare:nNnTF { \bool_if_p:n {#1} } = { \bool_if_p:n {#2} }
+      \c_false_bool
+      \c_true_bool
+  }
+\cs_new:Npn \bool_while_do:Nn #1#2
+  { \bool_if:NT #1 { #2 \bool_while_do:Nn #1 {#2} } }
+\cs_new:Npn \bool_until_do:Nn #1#2
+  { \bool_if:NF #1 { #2 \bool_until_do:Nn #1 {#2} } }
+\cs_generate_variant:Nn \bool_while_do:Nn { c }
+\cs_generate_variant:Nn \bool_until_do:Nn { c }
+\cs_new:Npn \bool_do_while:Nn #1#2
+  { #2 \bool_if:NT #1 { \bool_do_while:Nn #1 {#2} } }
+\cs_new:Npn \bool_do_until:Nn #1#2
+  { #2 \bool_if:NF #1 { \bool_do_until:Nn #1 {#2} } }
+\cs_generate_variant:Nn \bool_do_while:Nn { c }
+\cs_generate_variant:Nn \bool_do_until:Nn { c }
+\cs_new:Npn \bool_while_do:nn #1#2
+  {
+    \bool_if:nT {#1}
+      {
+        #2
+        \bool_while_do:nn {#1} {#2}
+      }
+  }
+\cs_new:Npn \bool_do_while:nn #1#2
+  {
+    #2
+    \bool_if:nT {#1} { \bool_do_while:nn {#1} {#2} }
+  }
+\cs_new:Npn \bool_until_do:nn #1#2
+  {
+    \bool_if:nF {#1}
+      {
+        #2
+        \bool_until_do:nn {#1} {#2}
+      }
+  }
+\cs_new:Npn \bool_do_until:nn #1#2
+  {
+    #2
+    \bool_if:nF {#1} { \bool_do_until:nn {#1} {#2}  }
+  }
+\cs_new_eq:NN \prg_case_end:nw \use_i_delimit_by_q_recursion_stop:nw
+\cs_new:Npn \prg_case_int:nnn #1
+  { \exp_args:Nf \prg_case_int_aux:nnn { \int_eval:n {#1} } }
+\cs_new:Npn \prg_case_int_aux:nnn #1 #2 #3
+  { \prg_case_int_aux:nw {#1} #2 {#1} {#3} \q_recursion_stop }
+\cs_new:Npn \prg_case_int_aux:nw #1#2#3
+  {
+    \int_compare:nNnTF {#1} = {#2}
+      { \prg_case_end:nw {#3} }
+      { \prg_case_int_aux:nw {#1} }
+  }
+\cs_new:Npn \prg_case_dim:nnn #1
+  { \exp_args:Nf \prg_case_dim_aux:nnn { \dim_eval:n {#1} } }
+\cs_new:Npn \prg_case_dim_aux:nnn #1 #2 #3
+  { \prg_case_dim_aux:nw {#1} #2 {#1} {#3} \q_recursion_stop }
+\cs_new:Npn \prg_case_dim_aux:nw #1#2#3
+  {
+    \dim_compare:nNnTF {#1} = {#2}
+      { \prg_case_end:nw {#3} }
+      { \prg_case_dim_aux:nw {#1} }
+  }
+\cs_new:Npn \prg_case_str:nnn #1#2#3
+  { \prg_case_str_aux:nw {#1} #2 {#1} {#3} \q_recursion_stop }
+\cs_new:Npn \prg_case_str_aux:nw #1#2#3
+  {
+    \str_if_eq:nnTF {#1} {#2}
+      { \prg_case_end:nw {#3} }
+      { \prg_case_str_aux:nw {#1} }
+  }
+\cs_generate_variant:Nn \prg_case_str:nnn { o }
+\cs_new:Npn \prg_case_str:xxn #1#2#3
+  { \prg_case_str_x_aux:nw {#1} #2 {#1} {#3} \q_recursion_stop }
+\cs_new:Npn \prg_case_str_x_aux:nw #1#2#3
+  {
+    \str_if_eq:xxTF {#1} {#2}
+      { \prg_case_end:nw {#3} }
+      { \prg_case_str_aux:nw {#1} }
+  }
+\cs_new:Npn \prg_case_tl:Nnn #1#2#3
+  { \prg_case_tl_aux:Nw #1 #2 #1 {#3} \q_recursion_stop }
+\cs_new:Npn \prg_case_tl_aux:Nw #1#2#3
+  {
+    \tl_if_eq:NNTF #1 #2
+      { \prg_case_end:nw {#3} }
+      { \prg_case_tl_aux:Nw #1 }
+  }
+\cs_generate_variant:Nn \prg_case_tl:Nnn { c }
+\cs_new_nopar:Npn \prg_replicate:nn #1
+  {
+    \int_to_roman:w
+      \exp_after:wN \prg_replicate_first_aux:N
+        \int_value:w \int_eval:w #1 \int_eval_end:
+      \cs_end:
+  }
+\cs_new_nopar:Npn \prg_replicate_aux:N #1
+  { \cs:w prg_replicate_#1 :n \prg_replicate_aux:N }
+\cs_new_nopar:Npn \prg_replicate_first_aux:N #1
+  { \cs:w prg_replicate_first_ #1 :n \prg_replicate_aux:N }
+\cs_new_nopar:Npn \prg_replicate_ :n #1 { \cs_end: }
+\cs_new:cpn { prg_replicate_0:n } #1 { \cs_end: {#1#1#1#1#1#1#1#1#1#1} }
+\cs_new:cpn { prg_replicate_1:n } #1 { \cs_end: {#1#1#1#1#1#1#1#1#1#1} #1 }
+\cs_new:cpn { prg_replicate_2:n } #1 { \cs_end: {#1#1#1#1#1#1#1#1#1#1} #1#1 }
+\cs_new:cpn { prg_replicate_3:n } #1
+  { \cs_end: {#1#1#1#1#1#1#1#1#1#1} #1#1#1 }
+\cs_new:cpn { prg_replicate_4:n } #1
+  { \cs_end: {#1#1#1#1#1#1#1#1#1#1} #1#1#1#1 }
+\cs_new:cpn { prg_replicate_5:n } #1
+  { \cs_end: {#1#1#1#1#1#1#1#1#1#1} #1#1#1#1#1 }
+\cs_new:cpn { prg_replicate_6:n } #1
+  { \cs_end: {#1#1#1#1#1#1#1#1#1#1} #1#1#1#1#1#1 }
+\cs_new:cpn { prg_replicate_7:n } #1
+  { \cs_end: {#1#1#1#1#1#1#1#1#1#1} #1#1#1#1#1#1#1 }
+\cs_new:cpn { prg_replicate_8:n } #1
+  { \cs_end: {#1#1#1#1#1#1#1#1#1#1} #1#1#1#1#1#1#1#1 }
+\cs_new:cpn { prg_replicate_9:n } #1
+  { \cs_end: {#1#1#1#1#1#1#1#1#1#1} #1#1#1#1#1#1#1#1#1 }
+\cs_new:cpn { prg_replicate_first_-:n } #1 { \c_zero \negative_replication }
+\cs_new:cpn { prg_replicate_first_0:n } #1 { \c_zero }
+\cs_new:cpn { prg_replicate_first_1:n } #1 { \c_zero #1 }
+\cs_new:cpn { prg_replicate_first_2:n } #1 { \c_zero #1#1 }
+\cs_new:cpn { prg_replicate_first_3:n } #1 { \c_zero #1#1#1 }
+\cs_new:cpn { prg_replicate_first_4:n } #1 { \c_zero #1#1#1#1 }
+\cs_new:cpn { prg_replicate_first_5:n } #1 { \c_zero #1#1#1#1#1 }
+\cs_new:cpn { prg_replicate_first_6:n } #1 { \c_zero #1#1#1#1#1#1 }
+\cs_new:cpn { prg_replicate_first_7:n } #1 { \c_zero #1#1#1#1#1#1#1 }
+\cs_new:cpn { prg_replicate_first_8:n } #1 { \c_zero #1#1#1#1#1#1#1#1 }
+\cs_new:cpn { prg_replicate_first_9:n } #1 { \c_zero #1#1#1#1#1#1#1#1#1 }
+\cs_new:Npn \prg_stepwise_function:nnnN #1#2
+  {
+    \int_compare:nNnTF {#2} > { 0 }
+      { \exp_args:Nf \prg_stepwise_function_incr:nnnN }
+      { \exp_args:Nf \prg_stepwise_function_decr:nnnN }
+        { \int_eval:n {#1} } {#2}
+  }
+\cs_new:Npn \prg_stepwise_function_incr:nnnN #1#2#3#4
+  {
+    \int_compare:nNnF {#1} > {#3}
+      {
+        #4 {#1}
+        \exp_args:Nf \prg_stepwise_function_incr:nnnN
+          { \int_eval:n { #1 + #2 } } {#2} {#3} #4
+      }
+  }
+\cs_new:Npn \prg_stepwise_function_decr:nnnN #1#2#3#4
+  {
+    \int_compare:nNnF {#1}  < {#3}
+      {
+        #4 {#1}
+        \exp_args:Nf \prg_stepwise_function_decr:nnnN
+          { \int_eval:n { #1 + #2 } } {#2} {#3} #4
+      }
+  }
+\int_new:N \g_prg_stepwise_level_int
+\cs_new_protected:Npn \prg_stepwise_inline:nnnn #1#2#3#4
+  {
+    \int_gincr:N \g_prg_stepwise_level_int
+    \cs_gset_nopar:cpn
+      { g_prg_stepwise_ \int_use:N \g_prg_stepwise_level_int :n }
+        ##1 {#4}
+    \int_compare:nNnTF {#2} > { 0 }
+      { \exp_args:Ncf \prg_stepwise_inline_incr:Nnnn }
+      { \exp_args:Ncf \prg_stepwise_inline_decr:Nnnn }
+        { g_prg_stepwise_ \int_use:N \g_prg_stepwise_level_int :n }
+        { \int_eval:n {#1} } {#2} {#3}
+    \int_gdecr:N \g_prg_stepwise_level_int
+  }
+\cs_new_protected:Npn \prg_stepwise_inline_incr:Nnnn #1#2#3#4
+  {
+    \int_compare:nNnF {#2} > {#4}
+      {
+        #1 {#2}
+        \exp_args:NNf \prg_stepwise_inline_incr:Nnnn #1
+          { \int_eval:n { #2 + #3 } } {#3} {#4}
+      }
+  }
+\cs_new_protected:Npn \prg_stepwise_inline_decr:Nnnn #1#2#3#4
+  {
+    \int_compare:nNnF {#2} < {#4}
+      {
+        #1 {#2}
+        \exp_args:NNf \prg_stepwise_inline_decr:Nnnn #1
+          { \int_eval:n { #2 + #3 } } {#3} {#4}
+      }
+  }
+\cs_new_protected:Npn \prg_stepwise_variable:nnnNn #1#2#3#4#5
+  {
+    \prg_stepwise_inline:nnnn {#1} {#2} {#3}
+      {
+        \tl_set:Nn #4 {##1}
+        #5
+      }
+  }
+\prg_new_conditional:Npnn \mode_if_vertical: { p , T , F , TF }
+  { \if_mode_vertical: \prg_return_true: \else: \prg_return_false: \fi: }
+\prg_new_conditional:Npnn \mode_if_horizontal: { p , T , F , TF }
+  { \if_mode_horizontal: \prg_return_true: \else: \prg_return_false: \fi: }
+\prg_new_conditional:Npnn \mode_if_inner: { p , T , F , TF }
+  { \if_mode_inner: \prg_return_true: \else: \prg_return_false: \fi: }
+\prg_new_conditional:Npnn \mode_if_math: { p , T , F , TF }
+  {
+    \scan_align_safe_stop:
+    \if_mode_math: \prg_return_true: \else: \prg_return_false: \fi:
+  }
+\cs_new_nopar:Npn \group_align_safe_begin:
+  { \if_false: { \fi: \if_int_compare:w `} = \c_zero \fi: }
+\cs_new_nopar:Npn \group_align_safe_end:
+  { \if_int_compare:w `{ = \c_zero } \fi: }
+\cs_new_nopar:Npn \scan_align_safe_stop:
+  {
+    \int_compare:nNnT \etex_currentgrouptype:D = \c_six
+      {
+        \int_compare:nNnF \etex_lastnodetype:D = \c_zero
+          { \int_compare:nNnF \etex_lastnodetype:D = \c_seven { \scan_stop: } }
+      }
+  }
+\group_begin:
+  \tex_lccode:D `\& = `\g \scan_stop:
+  \tex_catcode:D `\& = \c_twelve
+\tl_to_lowercase:n
+  {
+    \group_end:
+    \cs_new_nopar:Npn \prg_variable_get_scope:N #1
+      {
+        \exp_last_unbraced:Nf \prg_variable_get_scope_aux:w
+          { \cs_to_str:N #1 \exp_stop_f: \q_stop }
+      }
+    \cs_new_nopar:Npn \prg_variable_get_scope_aux:w #1#2 \q_stop
+      { \token_if_eq_meaning:NNT & #1 { g } }
+  }
+\group_begin:
+  \tex_lccode:D `\& = `\_ \scan_stop:
+  \tex_catcode:D `\& = \c_twelve
+\tl_to_lowercase:n
+  {
+    \group_end:
+    \cs_new_nopar:Npn \prg_variable_get_type:N #1
+      {
+        \exp_after:wN \prg_variable_get_type_aux:w
+          \token_to_str:N #1 & a \q_stop
+      }
+    \cs_new_nopar:Npn \prg_variable_get_type_aux:w #1 & #2#3 \q_stop
+      {
+        \token_if_eq_meaning:NNTF a #2
+          {#1}
+          { \prg_variable_get_type_aux:w #2#3 \q_stop }
+      }
+  }
+\cs_new_protected_nopar:Npn \prg_define_quicksort:nnn #1#2#3 {
+  \cs_set:cpx{#1_quicksort:n}##1{
+    \exp_not:c{#1_quicksort_start_partition:w} ##1
+    \exp_not:n{#2\q_nil#3\q_stop}
+  }
+  \cs_set:cpx{#1_quicksort_braced:n}##1{
+    \exp_not:c{#1_quicksort_start_partition_braced:n} ##1
+    \exp_not:N\q_nil\exp_not:N\q_stop
+  }
+  \cs_set:cpx {#1_quicksort_start_partition:w} #2 ##1 #3{
+    \exp_not:N \quark_if_nil:nT {##1}\exp_not:N \use_none_delimit_by_q_stop:w
+    \exp_not:c{#1_quicksort_do_partition_i:nnnw} {##1}{}{}
+  }
+  \cs_set:cpx {#1_quicksort_start_partition_braced:n} ##1 {
+    \exp_not:N \quark_if_nil:nT {##1}\exp_not:N \use_none_delimit_by_q_stop:w
+    \exp_not:c{#1_quicksort_do_partition_i_braced:nnnn} {##1}{}{}
+  }
+  \cs_set:cpx {#1_quicksort_do_partition_i:nnnw} ##1##2##3 #2 ##4 #3 {
+    \exp_not:N \quark_if_nil:nTF {##4} \exp_not:c {#1_do_quicksort_braced:nnnnw}
+    {
+      \exp_not:c{#1_quicksort_compare:nnTF}{##1}{##4}
+      \exp_not:c{#1_quicksort_partition_greater_ii:nnnn}
+      \exp_not:c{#1_quicksort_partition_less_ii:nnnn}
+    }
+    {##1}{##2}{##3}{##4}
+  }
+  \cs_set:cpx {#1_quicksort_do_partition_i_braced:nnnn} ##1##2##3##4 {
+    \exp_not:N \quark_if_nil:nTF {##4} \exp_not:c {#1_do_quicksort_braced:nnnnw}
+    {
+      \exp_not:c{#1_quicksort_compare:nnTF}{##1}{##4}
+      \exp_not:c{#1_quicksort_partition_greater_ii_braced:nnnn}
+      \exp_not:c{#1_quicksort_partition_less_ii_braced:nnnn}
+    }
+    {##1}{##2}{##3}{##4}
+  }
+  \cs_set:cpx {#1_quicksort_do_partition_ii:nnnw} ##1##2##3 #2 ##4 #3 {
+    \exp_not:N \quark_if_nil:nTF {##4} \exp_not:c {#1_do_quicksort_braced:nnnnw}
+    {
+      \exp_not:c{#1_quicksort_compare:nnTF}{##4}{##1}
+      \exp_not:c{#1_quicksort_partition_less_i:nnnn}
+      \exp_not:c{#1_quicksort_partition_greater_i:nnnn}
+    }
+    {##1}{##2}{##3}{##4}
+  }
+  \cs_set:cpx {#1_quicksort_do_partition_ii_braced:nnnn} ##1##2##3##4 {
+    \exp_not:N \quark_if_nil:nTF {##4} \exp_not:c {#1_do_quicksort_braced:nnnnw}
+    {
+      \exp_not:c{#1_quicksort_compare:nnTF}{##4}{##1}
+      \exp_not:c{#1_quicksort_partition_less_i_braced:nnnn}
+      \exp_not:c{#1_quicksort_partition_greater_i_braced:nnnn}
+    }
+    {##1}{##2}{##3}{##4}
+  }
+  \cs_set:cpx {#1_quicksort_partition_less_i:nnnn} ##1##2##3##4{
+    \exp_not:c{#1_quicksort_do_partition_i:nnnw}{##1}{##2}{{##4}##3}}
+  \cs_set:cpx {#1_quicksort_partition_less_ii:nnnn} ##1##2##3##4{
+    \exp_not:c{#1_quicksort_do_partition_ii:nnnw}{##1}{##2}{##3{##4}}}
+  \cs_set:cpx {#1_quicksort_partition_greater_i:nnnn} ##1##2##3##4{
+    \exp_not:c{#1_quicksort_do_partition_i:nnnw}{##1}{{##4}##2}{##3}}
+  \cs_set:cpx {#1_quicksort_partition_greater_ii:nnnn} ##1##2##3##4{
+    \exp_not:c{#1_quicksort_do_partition_ii:nnnw}{##1}{##2{##4}}{##3}}
+  \cs_set:cpx {#1_quicksort_partition_less_i_braced:nnnn} ##1##2##3##4{
+    \exp_not:c{#1_quicksort_do_partition_i_braced:nnnn}{##1}{##2}{{##4}##3}}
+  \cs_set:cpx {#1_quicksort_partition_less_ii_braced:nnnn} ##1##2##3##4{
+    \exp_not:c{#1_quicksort_do_partition_ii_braced:nnnn}{##1}{##2}{##3{##4}}}
+  \cs_set:cpx {#1_quicksort_partition_greater_i_braced:nnnn} ##1##2##3##4{
+    \exp_not:c{#1_quicksort_do_partition_i_braced:nnnn}{##1}{{##4}##2}{##3}}
+  \cs_set:cpx {#1_quicksort_partition_greater_ii_braced:nnnn} ##1##2##3##4{
+    \exp_not:c{#1_quicksort_do_partition_ii_braced:nnnn}{##1}{##2{##4}}{##3}}
+  \cs_set:cpx {#1_do_quicksort_braced:nnnnw} ##1##2##3##4\q_stop {
+    \exp_not:c{#1_quicksort_braced:n}{##2}
+    \exp_not:c{#1_quicksort_function:n}{##1}
+    \exp_not:c{#1_quicksort_braced:n}{##3}
+  }
+}
+\prg_define_quicksort:nnn {prg}{}{}
+\cs_set:Npn \prg_quicksort_function:n {\ERROR}
+\cs_set:Npn \prg_quicksort_compare:nnTF {\ERROR}
+\cs_new_protected:Npn \prg_new_map_functions:Nn #1#2 { \deprectiated }
+\cs_new_protected:Npn \prg_set_map_functions:Nn #1#2 { \deprectiated }
+%% 
+%%
+%% End of file `l3prg.sty'.
diff --git a/Master/texmf-dist/tex/latex/l3kernel/l3prop.sty b/Master/texmf-dist/tex/latex/l3kernel/l3prop.sty
new file mode 100644
index 00000000000..5d2a1ff69d0
--- /dev/null
+++ b/Master/texmf-dist/tex/latex/l3kernel/l3prop.sty
@@ -0,0 +1,333 @@
+%%
+%% This is file `l3prop.sty',
+%% generated with the docstrip utility.
+%%
+%% The original source files were:
+%%
+%% l3prop.dtx  (with options: `package')
+%% 
+%% EXPERIMENTAL CODE
+%% 
+%% Do not distribute this file without also distributing the
+%% source files specified above.
+%% 
+%% Do not distribute a modified version of this file.
+%% 
+%% File: l3prop.dtx Copyright (C) 1990-2011 The LaTeX3 Project
+%%
+%% It may be distributed and/or modified under the conditions of the
+%% LaTeX Project Public License (LPPL), either version 1.3c of this
+%% license or (at your option) any later version.  The latest version
+%% of this license is in the file
+%%
+%%    http://www.latex-project.org/lppl.txt
+%%
+%% This file is part of the "expl3 bundle" (The Work in LPPL)
+%% and all files in that bundle must be distributed together.
+%%
+%% The released version of this bundle is available from CTAN.
+%%
+%% -----------------------------------------------------------------------
+%%
+%% The development version of the bundle can be found at
+%%
+%%    http://www.latex-project.org/svnroot/experimental/trunk/
+%%
+%% for those people who are interested.
+%%
+%%%%%%%%%%%
+%% NOTE: %%
+%%%%%%%%%%%
+%%
+%%   Snapshots taken from the repository represent work in progress and may
+%%   not work or may contain conflicting material!  We therefore ask
+%%   people _not_ to put them into distributions, archives, etc. without
+%%   prior consultation with the LaTeX3 Project.
+%%
+%% -----------------------------------------------------------------------
+\RequirePackage{l3names}
+\GetIdInfo$Id: l3prop.dtx 2478 2011-06-19 21:34:23Z joseph $
+  {L3 Experimental property lists}
+\ProvidesExplPackage
+  {\ExplFileName}{\ExplFileDate}{\ExplFileVersion}{\ExplFileDescription}
+\package_check_loaded_expl:
+\quark_new:N \q_prop
+\tl_const:Nn \c_empty_prop { \q_prop }
+\cs_new_protected:Npn \prop_new:N #1 { \cs_new_eq:NN  #1  \c_empty_prop }
+\cs_new_protected:Npn \prop_new:c #1 { \cs_new_eq:cN {#1} \c_empty_prop }
+\cs_new_protected:Npn \prop_clear:N #1  { \cs_set_eq:NN   #1  \c_empty_prop }
+\cs_new_protected:Npn \prop_clear:c #1  { \cs_set_eq:cN  {#1} \c_empty_prop }
+\cs_new_protected:Npn \prop_gclear:N #1 { \cs_gset_eq:NN  #1  \c_empty_prop }
+\cs_new_protected:Npn \prop_gclear:c #1 { \cs_gset_eq:cN {#1} \c_empty_prop }
+\cs_new_protected:Npn \prop_clear_new:N #1
+  { \cs_if_exist:NTF #1 { \prop_clear:N #1 } { \prop_new:N #1 } }
+\cs_generate_variant:Nn \prop_clear_new:N {c}
+\cs_new_protected:Npn \prop_gclear_new:N #1
+  { \cs_if_exist:NTF #1 { \prop_gclear:N #1 } { \prop_new:N #1 } }
+\cs_new_eq:NN \prop_gclear_new:c \prop_gclear:c
+\cs_new_eq:NN \prop_set_eq:NN  \tl_set_eq:NN
+\cs_new_eq:NN \prop_set_eq:Nc  \tl_set_eq:Nc
+\cs_new_eq:NN \prop_set_eq:cN  \tl_set_eq:cN
+\cs_new_eq:NN \prop_set_eq:cc  \tl_set_eq:cc
+\cs_new_eq:NN \prop_gset_eq:NN \tl_gset_eq:NN
+\cs_new_eq:NN \prop_gset_eq:Nc \tl_gset_eq:Nc
+\cs_new_eq:NN \prop_gset_eq:cN \tl_gset_eq:cN
+\cs_new_eq:NN \prop_gset_eq:cc \tl_gset_eq:cc
+\cs_set_protected:Npn \prop_split:NnTF #1#2
+  { \exp_args:NNo \prop_split_aux:NnTF #1 { \tl_to_str:n {#2} } }
+\cs_new_protected:Npn \prop_split_aux:NnTF #1#2
+  {
+    \cs_set_protected:Npn \prop_split_aux:w
+      ##1 \q_prop #2 \q_prop ##2 ##3 ##4 \q_mark ##5 \q_stop
+      { \prop_split_aux:nnnn ##3 { {##1 \q_prop } {##2} {##4} } }
+    \exp_after:wN \prop_split_aux:w #1 \q_mark
+         \q_prop #2 \q_prop { } { ? \use_ii:nn { } } \q_mark \q_stop
+  }
+\cs_new:Npn \prop_split_aux:nnnn #1#2#3#4 { #3 #2 }
+\cs_new_protected:Npn \prop_split_aux:w { }
+\cs_set_protected:Npn \prop_split:Nnn #1#2#3
+  {
+    \prop_split:NnTF #1 {#2}
+    {#3}
+    { \exp_args:Nno \use:n {#3} {#1} { \q_no_value } { } }
+  }
+\cs_new_protected:Npn \prop_del:Nn #1#2
+  { \prop_split:NnTF #1 {#2} { \prop_del_aux:NNnnn \tl_set:Nn #1 } { } }
+\cs_new_protected:Npn \prop_gdel:Nn #1#2
+  { \prop_split:NnTF #1 {#2} { \prop_del_aux:NNnnn \tl_gset:Nn #1 } { } }
+\cs_new_protected:Npn \prop_del_aux:NNnnn #1#2#3#4#5
+  { #1 #2 { #3 #5 } }
+\cs_generate_variant:Nn \prop_del:Nn  {     NV }
+\cs_generate_variant:Nn \prop_del:Nn  { c , cV }
+\cs_generate_variant:Nn \prop_gdel:Nn {     NV }
+\cs_generate_variant:Nn \prop_gdel:Nn { c , cV }
+\cs_new_protected:Npn \prop_get:NnN #1#2#3
+  {
+    \prop_split:NnTF #1 {#2}
+      { \prop_get_aux:Nnnn #3 }
+      { \tl_set:Nn #3 { \q_no_value } }
+  }
+\cs_new_protected:Npn \prop_get_aux:Nnnn #1#2#3#4
+  { \tl_set:Nn #1 {#3} }
+\cs_generate_variant:Nn \prop_get:NnN {     NV , No }
+\cs_generate_variant:Nn \prop_get:NnN { c , cV , co }
+\cs_new_protected:Npn \prop_pop:NnN #1#2#3
+  {
+    \prop_split:NnTF #1 {#2}
+      { \prop_pop_aux:NNNnnn \tl_set:Nn #1 #3 }
+      { \tl_set:Nn #3 { \q_no_value } }
+  }
+\cs_new_protected:Npn \prop_gpop:NnN #1#2#3
+  {
+    \prop_split:NnTF #1 {#2}
+      { \prop_pop_aux:NNNnnn \tl_gset:Nn #1 #3 }
+      { \tl_set:Nn #3 { \q_no_value } }
+  }
+\cs_new_protected:Npn \prop_pop_aux:NNNnnn #1#2#3#4#5#6
+  {
+    \tl_set:Nn #3 {#5}
+    #1 #2 { #4 #6 }
+  }
+\cs_generate_variant:Nn \prop_pop:NnN  {     No }
+\cs_generate_variant:Nn \prop_pop:NnN  { c , co }
+\cs_generate_variant:Nn \prop_gpop:NnN {     No }
+\cs_generate_variant:Nn \prop_gpop:NnN { c , co }
+\cs_new_protected:Npn \prop_put:Nnn { \prop_put_aux:NNnn \tl_set:Nx }
+\cs_new_protected:Npn \prop_gput:Nnn { \prop_put_aux:NNnn \tl_gset:Nx }
+\cs_new_protected:Npn \prop_put_aux:NNnn #1#2#3#4
+  {
+    \prop_split:Nnn #2 {#3} { \prop_put_aux:NNnnnnn #1 #2 {#3} {#4} }
+  }
+\cs_new_protected:Npn \prop_put_aux:NNnnnnn #1#2#3#4#5#6#7
+  {
+    #1 #2
+      {
+        \exp_not:n { #5 #7 }
+        \tl_to_str:n {#3} \exp_not:n { \q_prop {#4} \q_prop }
+      }
+  }
+\cs_generate_variant:Nn \prop_put:Nnn
+  {     NnV , Nno , Nnx , NV , NVV , No , Noo }
+\cs_generate_variant:Nn \prop_put:Nnn
+  { c , cnV , cno , cnx , cV , cVV , co , coo }
+\cs_generate_variant:Nn \prop_gput:Nnn
+  {     NnV , Nno , Nnx , NV , NVV , No , Noo }
+\cs_generate_variant:Nn \prop_gput:Nnn
+  { c , cnV , cno , cnx , cV , cVV , co , coo }
+\cs_new_protected_nopar:Npn \prop_put_if_new:Nnn
+  { \prop_put_if_new_aux:NNnn \tl_put_right:Nx }
+\cs_new_protected_nopar:Npn \prop_gput_if_new:Nnn
+  { \prop_put_if_new_aux:NNnn \tl_gput_right:Nx }
+\cs_new_protected:Npn \prop_put_if_new_aux:NNnn #1#2#3#4
+  {
+    \prop_split:NnTF #2 {#3}
+      { \use_none:nnn }
+      {
+        #1 #2
+          { \tl_to_str:n {#3} \exp_not:n { \q_prop {#4} \q_prop } }
+      }
+  }
+\cs_generate_variant:Nn \prop_put_if_new:Nnn  { c }
+\cs_generate_variant:Nn \prop_gput_if_new:Nnn { c }
+\prg_new_conditional:Npnn \prop_if_empty:N #1 { p, T , F , TF }
+  {
+    \if_meaning:w #1 \c_empty_prop
+      \prg_return_true:
+    \else:
+      \prg_return_false:
+    \fi:
+  }
+\cs_generate_variant:Nn \prop_if_empty_p:N {c}
+\cs_generate_variant:Nn \prop_if_empty:NTF {c}
+\cs_generate_variant:Nn \prop_if_empty:NT  {c}
+\cs_generate_variant:Nn \prop_if_empty:NF  {c}
+\prg_new_conditional:Npnn \prop_if_in:Nn #1#2 { p , T , F , TF }
+  {
+    \exp_last_unbraced:Noo \prop_if_in_aux:nwn
+      { \tl_to_str:n {#2} } #1
+    A \q_prop { } \q_stop
+  }
+\cs_new:Npn \prop_if_in_aux:nwn #1 \q_prop #2 \q_prop #3
+  {
+    \if_catcode:w A #2
+       \prg_return_false:
+       \exp_after:wN \use_none_delimit_by_q_stop:w
+    \fi:
+    \str_if_eq:nnT {#1} {#2}
+      {
+        \prg_return_true:
+        \use_none_delimit_by_q_stop:w
+      }
+    \prop_if_in_aux:nwn {#1}
+  }
+\cs_generate_variant:Nn \prop_if_in_p:Nn {     NV , No }
+\cs_generate_variant:Nn \prop_if_in_p:Nn { c , cV , co }
+\cs_generate_variant:Nn \prop_if_in:NnT  {     NV , No }
+\cs_generate_variant:Nn \prop_if_in:NnT  { c , cV , co }
+\cs_generate_variant:Nn \prop_if_in:NnF  {     NV , No }
+\cs_generate_variant:Nn \prop_if_in:NnF  { c , cV , co }
+\cs_generate_variant:Nn \prop_if_in:NnTF {     NV , No }
+\cs_generate_variant:Nn \prop_if_in:NnTF { c , cV , co }
+\cs_new_nopar:Npn \prop_map_function:NN #1#2
+  {
+    \exp_last_unbraced:NNo \prop_map_function_aux:Nwn #2
+      #1 A \q_prop { } \q_recursion_stop
+  }
+\cs_new:Npn \prop_map_function_aux:Nwn #1 \q_prop #2 \q_prop #3
+  {
+    \if_catcode:w A #2
+      \exp_after:wN \prop_map_break:
+    \fi:
+    #1 {#2} {#3}
+    \prop_map_function_aux:Nwn #1
+  }
+\cs_generate_variant:Nn \prop_map_function:NN {     Nc }
+\cs_generate_variant:Nn \prop_map_function:NN { c , cc }
+\int_new:N \g_prop_map_inline_int
+\cs_new_protected:Npn \prop_map_inline:Nn #1#2
+  {
+    \int_gincr:N \g_prop_map_inline_int
+    \cs_gset:cpn { prop_map_inline_ \int_use:N \g_prop_map_inline_int :nn }
+      ##1##2 {#2}
+    \prop_map_function:Nc #1
+      { prop_map_inline_ \int_use:N \g_prop_map_inline_int :nn }
+    \int_gdecr:N \g_prop_map_inline_int
+  }
+\cs_generate_variant:Nn \prop_map_inline:Nn { c }
+\cs_new_eq:NN \prop_map_break: \use_none_delimit_by_q_recursion_stop:w
+\cs_new_eq:NN \prop_map_break:n \use_i_delimit_by_q_recursion_stop:nw
+\tl_new:N \l_prop_show_tl
+\cs_new_protected_nopar:Npn \prop_show:N #1
+  {
+    \prop_if_empty:NTF #1
+      {
+        \iow_term:x { Property~list~\token_to_str:N #1 \c_space_tl is~empty }
+        \tl_show:n { }
+      }
+      {
+        \iow_term:x
+          {
+            Property~list~\token_to_str:N #1 \c_space_tl
+            contains~the~pairs~(without~outer~braces):
+          }
+        \tl_set:Nx \l_prop_show_tl
+          { \prop_map_function:NN #1 \prop_show_aux:nn }
+        \tl_show:n \exp_after:wN \exp_after:wN \exp_after:wN
+          { \exp_after:wN \prop_show_aux:w \l_prop_show_tl }
+      }
+  }
+\cs_new:Npn \prop_show_aux:nn #1#2
+  {
+    \iow_newline: > \c_space_tl \c_space_tl
+    \iow_char:N \{ #1 \iow_char:N \}
+    \c_space_tl \c_space_tl => \c_space_tl \c_space_tl
+    \iow_char:N \{ \exp_not:n {#2} \iow_char:N \}
+  }
+\cs_new:Npn \prop_show_aux:w #1 > ~ { }
+\cs_generate_variant:Nn \prop_show:N { c }
+\prg_new_protected_conditional:Npnn \prop_get:NnN #1#2#3 { T , F , TF }
+  {
+    \prop_split:NnTF #1 {#2}
+      { \prop_get_aux_true:Nnnn #3 }
+      { \prg_return_false: }
+  }
+\cs_new_protected:Npn \prop_get_aux_true:Nnnn #1#2#3#4
+  {
+    \tl_set:Nn #1 {#3}
+    \prg_return_true:
+  }
+\prg_new_protected_conditional:Npnn \prop_pop:NnN #1#2#3 {T,F,TF}
+  {
+    \prop_split:NnTF #1 {#2}
+      { \prop_pop_aux_true:NNNnnn \tl_set:Nn #1 #3 }
+      { \prg_return_false: }
+  }
+\prg_new_protected_conditional:Npnn \prop_gpop:NnN #1#2#3 {T,F,TF}
+  {
+    \prop_split:NnTF #1 {#2}
+      { \prop_pop_aux_true:NNNnnn \tl_gset:Nn #1 #3 }
+      { \prg_return_false: }
+  }
+\cs_new_protected:Npn \prop_pop_aux_true:NNNnnn #1#2#3#4#5#6
+  {
+    \tl_set:Nn #3 {#5}
+    #1 #2 { #4 #6 }
+    \prg_return_true:
+  }
+\cs_new:Npn \prop_map_tokens:Nn #1#2
+  {
+    \exp_last_unbraced:Nno \prop_map_tokens_aux:nwn {#2} #1
+    A \q_prop { } \q_recursion_stop
+  }
+\cs_new:Npn \prop_map_tokens_aux:nwn #1 \q_prop #2 \q_prop #3
+  {
+    \if_catcode:w A #2
+       \exp_after:wN \prop_map_break:
+    \fi:
+    \use:n {#1} {#2} {#3}
+    \prop_map_tokens_aux:nwn {#1}
+  }
+\cs_new:Npn \prop_get:Nn #1 #2
+  { \prop_map_tokens:Nn #1 { \prop_get_aux:nnn {#2} } }
+\cs_new:Npn \prop_get_aux:nnn #1 #2 #3
+  { \str_if_eq:nnT {#1} {#2} { \prop_map_break:n {#3} } }
+\cs_new_eq:NN \prop_display:N \prop_show:N
+\cs_new_eq:NN \prop_display:c \prop_show:c
+\cs_new_protected:Npn \prop_gget:NnN #1#2#3
+  { \prop_split:Nnn #1 {#2} { \prop_gget_aux:Nnnn #3 } }
+\cs_new_protected:Npn \prop_gget_aux:Nnnn #1#2#3#4
+  { \tl_gset:Nn #1 {#3} }
+\cs_generate_variant:Nn \prop_gget:NnN {     NV }
+\cs_generate_variant:Nn \prop_gget:NnN { c , cV }
+\cs_new_eq:NN \prop_get_gdel:NnN \prop_gpop:NnN
+\cs_generate_variant:Nn \prop_if_in:NnT  { cc }
+\cs_generate_variant:Nn \prop_if_in:NnF  { cc }
+\cs_generate_variant:Nn \prop_if_in:NnTF { cc }
+\cs_generate_variant:Nn \prop_gput:Nnn { ccx }
+\prg_new_eq_conditional:NNn \prop_if_eq:NN \tl_if_eq:NN { p , T , F , TF }
+\prg_new_eq_conditional:NNn \prop_if_eq:cN \tl_if_eq:cN { p , T , F , TF }
+\prg_new_eq_conditional:NNn \prop_if_eq:Nc \tl_if_eq:Nc { p , T , F , TF }
+\prg_new_eq_conditional:NNn \prop_if_eq:cc \tl_if_eq:cc { p , T , F , TF }
+%% 
+%%
+%% End of file `l3prop.sty'.
diff --git a/Master/texmf-dist/tex/latex/l3kernel/l3quark.sty b/Master/texmf-dist/tex/latex/l3kernel/l3quark.sty
new file mode 100644
index 00000000000..35a8ca8f8e7
--- /dev/null
+++ b/Master/texmf-dist/tex/latex/l3kernel/l3quark.sty
@@ -0,0 +1,154 @@
+%%
+%% This is file `l3quark.sty',
+%% generated with the docstrip utility.
+%%
+%% The original source files were:
+%%
+%% l3quark.dtx  (with options: `package')
+%% 
+%% EXPERIMENTAL CODE
+%% 
+%% Do not distribute this file without also distributing the
+%% source files specified above.
+%% 
+%% Do not distribute a modified version of this file.
+%% 
+%% File: l3quark.dtx Copyright (C) 1990-2011 The LaTeX3 Project
+%%
+%% It may be distributed and/or modified under the conditions of the
+%% LaTeX Project Public License (LPPL), either version 1.3c of this
+%% license or (at your option) any later version.  The latest version
+%% of this license is in the file
+%%
+%%    http://www.latex-project.org/lppl.txt
+%%
+%% This file is part of the "expl3 bundle" (The Work in LPPL)
+%% and all files in that bundle must be distributed together.
+%%
+%% The released version of this bundle is available from CTAN.
+%%
+%% -----------------------------------------------------------------------
+%%
+%% The development version of the bundle can be found at
+%%
+%%    http://www.latex-project.org/svnroot/experimental/trunk/
+%%
+%% for those people who are interested.
+%%
+%%%%%%%%%%%
+%% NOTE: %%
+%%%%%%%%%%%
+%%
+%%   Snapshots taken from the repository represent work in progress and may
+%%   not work or may contain conflicting material!  We therefore ask
+%%   people _not_ to put them into distributions, archives, etc. without
+%%   prior consultation with the LaTeX3 Project.
+%%
+%% -----------------------------------------------------------------------
+\RequirePackage{l3names}
+\GetIdInfo$Id: l3quark.dtx 2478 2011-06-19 21:34:23Z joseph $
+  {L3 Experimental quarks}
+\ProvidesExplPackage
+  {\ExplFileName}{\ExplFileDate}{\ExplFileVersion}{\ExplFileDescription}
+\package_check_loaded_expl:
+\cs_new_protected_nopar:Npn \quark_new:N #1 { \tl_const:Nn #1 {#1} }
+\quark_new:N \q_nil
+\quark_new:N \q_mark
+\quark_new:N \q_no_value
+\quark_new:N \q_stop
+\quark_new:N \q_recursion_tail
+\quark_new:N \q_recursion_stop
+\cs_new:Npn \quark_if_recursion_tail_stop:N #1
+  {
+    \if_meaning:w #1 \q_recursion_tail
+      \exp_after:wN \use_none_delimit_by_q_recursion_stop:w
+    \fi:
+  }
+\cs_new:Npn \quark_if_recursion_tail_stop_do:Nn #1#2
+  {
+    \if_meaning:w #1 \q_recursion_tail
+      \exp_after:wN \use_i_delimit_by_q_recursion_stop:nw
+    \else:
+      \exp_after:wN \use_none:n
+    \fi:
+      {#2}
+  }
+\cs_new:Npn \quark_if_recursion_tail_stop:n #1
+  {
+    \if_catcode:w
+      A
+      \etex_detokenize:D \exp_after:wN
+        {
+          \quark_if_recursion_tail_aux:w #1 \q_recursion_stop
+            \q_recursion_tail \q_recursion_stop \q_stop
+        }
+      A
+      \exp_after:wN \use_none_delimit_by_q_recursion_stop:w
+    \fi:
+  }
+\cs_new:Npn \quark_if_recursion_tail_stop_do:nn #1#2
+  {
+    \if_catcode:w
+      A
+      \etex_detokenize:D \exp_after:wN
+        {
+          \quark_if_recursion_tail_aux:w #1 \q_recursion_stop
+            \q_recursion_tail \q_recursion_stop \q_stop
+        }
+        A
+      \exp_after:wN \use_i_delimit_by_q_recursion_stop:nw
+    \else:
+      \exp_after:wN \use_none:n
+    \fi:
+      {#2}
+  }
+\cs_new:Npn \quark_if_recursion_tail_aux:w
+  #1 \q_recursion_tail #2 \q_recursion_stop #3 \q_stop
+  { #1 #2 }
+\cs_generate_variant:Nn \quark_if_recursion_tail_stop:n { o }
+\cs_generate_variant:Nn \quark_if_recursion_tail_stop_do:nn { o }
+\prg_new_conditional:Nnn \quark_if_nil:N { p, T , F , TF }
+  {
+    \if_meaning:w \q_nil #1
+      \prg_return_true:
+    \else:
+      \prg_return_false:
+    \fi:
+  }
+\prg_new_conditional:Nnn \quark_if_no_value:N { p, T , F , TF }
+  {
+    \if_meaning:w \q_no_value #1
+      \prg_return_true:
+    \else:
+      \prg_return_false:
+    \fi:
+  }
+\cs_generate_variant:Nn \quark_if_no_value_p:N { c }
+\cs_generate_variant:Nn \quark_if_no_value:NT  { c }
+\cs_generate_variant:Nn \quark_if_no_value:NF  { c }
+\cs_generate_variant:Nn \quark_if_no_value:NTF { c }
+\prg_new_conditional:Nnn \quark_if_nil:n { p, T , F , TF }
+  {
+    \if_int_compare:w \pdftex_strcmp:D
+      { \exp_not:N \q_nil } { \exp_not:n {#1} } = \c_zero
+      \prg_return_true:
+    \else:
+      \prg_return_false:
+    \fi:
+  }
+\prg_new_conditional:Nnn \quark_if_no_value:n { p, T , F , TF }
+  {
+    \if_int_compare:w \pdftex_strcmp:D
+      { \exp_not:N \q_no_value } { \exp_not:n {#1} } = \c_zero
+      \prg_return_true:
+    \else:
+      \prg_return_false:
+    \fi:
+  }
+\cs_generate_variant:Nn \quark_if_nil_p:n { V , o }
+\cs_generate_variant:Nn \quark_if_nil:nTF { V , o }
+\cs_generate_variant:Nn \quark_if_nil:nT  { V , o }
+\cs_generate_variant:Nn \quark_if_nil:nF  { V , o }
+%% 
+%%
+%% End of file `l3quark.sty'.
diff --git a/Master/texmf-dist/tex/latex/l3kernel/l3seq.sty b/Master/texmf-dist/tex/latex/l3kernel/l3seq.sty
new file mode 100644
index 00000000000..13136da4828
--- /dev/null
+++ b/Master/texmf-dist/tex/latex/l3kernel/l3seq.sty
@@ -0,0 +1,531 @@
+%%
+%% This is file `l3seq.sty',
+%% generated with the docstrip utility.
+%%
+%% The original source files were:
+%%
+%% l3seq.dtx  (with options: `package')
+%% 
+%% EXPERIMENTAL CODE
+%% 
+%% Do not distribute this file without also distributing the
+%% source files specified above.
+%% 
+%% Do not distribute a modified version of this file.
+%% 
+%% File: l3seq.dtx Copyright (C) 1990-2011 The LaTeX3 Project
+%%
+%% It may be distributed and/or modified under the conditions of the
+%% LaTeX Project Public License (LPPL), either version 1.3c of this
+%% license or (at your option) any later version.  The latest version
+%% of this license is in the file
+%%
+%%    http://www.latex-project.org/lppl.txt
+%%
+%% This file is part of the "expl3 bundle" (The Work in LPPL)
+%% and all files in that bundle must be distributed together.
+%%
+%% The released version of this bundle is available from CTAN.
+%%
+%% -----------------------------------------------------------------------
+%%
+%% The development version of the bundle can be found at
+%%
+%%    http://www.latex-project.org/svnroot/experimental/trunk/
+%%
+%% for those people who are interested.
+%%
+%%%%%%%%%%%
+%% NOTE: %%
+%%%%%%%%%%%
+%%
+%%   Snapshots taken from the repository represent work in progress and may
+%%   not work or may contain conflicting material!  We therefore ask
+%%   people _not_ to put them into distributions, archives, etc. without
+%%   prior consultation with the LaTeX3 Project.
+%%
+%% -----------------------------------------------------------------------
+\RequirePackage{l3names}
+\GetIdInfo$Id: l3seq.dtx 2478 2011-06-19 21:34:23Z joseph $
+  {L3 Experimental sequences and stacks}
+\ProvidesExplPackage
+  {\ExplFileName}{\ExplFileDate}{\ExplFileVersion}{\ExplFileDescription}
+\package_check_loaded_expl:
+\cs_new:Npn \seq_item:n
+  {
+    \seq_use_error:
+    \use_none:n
+  }
+\tl_new:N \l_seq_tmpa_tl
+\tl_new:N \l_seq_tmpb_tl
+\cs_new_eq:NN \seq_new:N \tl_new:N
+\cs_new_eq:NN \seq_new:c \tl_new:c
+\cs_new_eq:NN \seq_clear:N  \tl_clear:N
+\cs_new_eq:NN \seq_clear:c  \tl_clear:c
+\cs_new_eq:NN \seq_gclear:N \tl_gclear:N
+\cs_new_eq:NN \seq_gclear:c \tl_gclear:c
+\cs_new_eq:NN \seq_clear_new:N  \tl_clear_new:N
+\cs_new_eq:NN \seq_clear_new:c  \tl_clear_new:c
+\cs_new_eq:NN \seq_gclear_new:N \tl_gclear_new:N
+\cs_new_eq:NN \seq_gclear_new:c \tl_gclear_new:c
+\cs_new_eq:NN \seq_set_eq:NN  \tl_set_eq:NN
+\cs_new_eq:NN \seq_set_eq:Nc  \tl_set_eq:Nc
+\cs_new_eq:NN \seq_set_eq:cN  \tl_set_eq:cN
+\cs_new_eq:NN \seq_set_eq:cc  \tl_set_eq:cc
+\cs_new_eq:NN \seq_gset_eq:NN \tl_gset_eq:NN
+\cs_new_eq:NN \seq_gset_eq:Nc \tl_gset_eq:Nc
+\cs_new_eq:NN \seq_gset_eq:cN \tl_gset_eq:cN
+\cs_new_eq:NN \seq_gset_eq:cc \tl_gset_eq:cc
+\cs_new_protected_nopar:Npn \seq_concat:NNN #1#2#3
+  { \tl_set:Nx #1 { \exp_not:o {#2} \exp_not:o {#3} } }
+\cs_new_protected_nopar:Npn \seq_gconcat:NNN #1#2#3
+  { \tl_gset:Nx #1 { \exp_not:o {#2} \exp_not:o {#3} } }
+\cs_generate_variant:Nn \seq_concat:NNN  { ccc }
+\cs_generate_variant:Nn \seq_gconcat:NNN { ccc }
+\cs_new_protected:Npn \seq_put_left:Nn #1#2
+  { \tl_put_left:Nn #1 { \seq_item:n {#2} } }
+\cs_new_protected:Npn \seq_put_right:Nn #1#2
+  { \tl_put_right:Nn #1 { \seq_item:n {#2} } }
+\cs_generate_variant:Nn \seq_put_left:Nn  {     NV , Nv , No , Nx }
+\cs_generate_variant:Nn \seq_put_left:Nn  { c , cV , cv , co , cx }
+\cs_generate_variant:Nn \seq_put_right:Nn {     NV , Nv , No , Nx }
+\cs_generate_variant:Nn \seq_put_right:Nn { c , cV , cv , co , cx }
+\cs_new_protected:Npn \seq_gput_left:Nn #1#2
+  { \tl_gput_left:Nn #1 { \seq_item:n {#2} } }
+\cs_new_protected:Npn \seq_gput_right:Nn #1#2
+  { \tl_gput_right:Nn #1 { \seq_item:n {#2} } }
+\cs_generate_variant:Nn \seq_gput_left:Nn  {     NV , Nv , No , Nx }
+\cs_generate_variant:Nn \seq_gput_left:Nn  { c , cV , cv , co , cx }
+\cs_generate_variant:Nn \seq_gput_right:Nn {     NV , Nv , No , Nx }
+\cs_generate_variant:Nn \seq_gput_right:Nn { c , cV , cv , co , cx }
+\seq_new:N \l_seq_remove_seq
+\cs_new_protected:Npn \seq_remove_duplicates:N
+  { \seq_remove_duplicates_aux:NN \seq_set_eq:NN }
+\cs_new_protected:Npn \seq_gremove_duplicates:N
+  { \seq_remove_duplicates_aux:NN \seq_gset_eq:NN }
+\cs_new_protected:Npn \seq_remove_duplicates_aux:NN #1#2
+  {
+    \seq_clear:N \l_seq_remove_seq
+    \seq_map_inline:Nn #2
+      {
+        \seq_if_in:NnF \l_seq_remove_seq {##1}
+          { \seq_put_right:Nn \l_seq_remove_seq {##1} }
+      }
+    #1 #2 \l_seq_remove_seq
+  }
+\cs_generate_variant:Nn \seq_remove_duplicates:N  { c }
+\cs_generate_variant:Nn \seq_gremove_duplicates:N { c }
+\cs_new_protected:Npn \seq_remove_all:Nn
+  { \seq_remove_all_aux:NNn \tl_set:Nx }
+\cs_new_protected:Npn \seq_gremove_all:Nn
+  { \seq_remove_all_aux:NNn \tl_gset:Nx }
+\cs_new_protected:Npn \seq_remove_all_aux:NNn #1#2#3
+  {
+    \seq_push_item_def:n
+      {
+        \str_if_eq:nnT {##1} {#3}
+          {
+            \if_false: { \fi: }
+            \tl_set:Nn \l_seq_tmpb_tl {##1}
+            #1 #2
+               { \if_false: } \fi:
+                 \exp_not:o {#2}
+                 \tl_if_eq:NNT \l_seq_tmpa_tl \l_seq_tmpb_tl
+                   { \use_none:nn }
+          }
+        \exp_not:n { \seq_item:n {##1} }
+      }
+    \tl_set:Nn \l_seq_tmpa_tl {#3}
+    #1 #2 {#2}
+    \seq_pop_item_def:
+  }
+\cs_generate_variant:Nn \seq_remove_all:Nn  { c }
+\cs_generate_variant:Nn \seq_gremove_all:Nn { c }
+\prg_new_eq_conditional:NNn \seq_if_empty:N \tl_if_empty:N
+  { p , T , F , TF }
+\prg_new_eq_conditional:NNn \seq_if_empty:c \tl_if_empty:c
+  { p , T , F , TF }
+\prg_new_protected_conditional:Npnn \seq_if_in:Nn #1#2
+  { T , F , TF }
+  {
+    \group_begin:
+      \tl_set:Nn \l_seq_tmpa_tl {#2}
+      \cs_set_protected:Npn \seq_item:n ##1
+        {
+          \tl_set:Nn \l_seq_tmpb_tl {##1}
+          \if_meaning:w \l_seq_tmpa_tl \l_seq_tmpb_tl
+            \exp_after:wN \seq_if_in_aux:
+          \fi:
+        }
+      #1
+      \seq_break:n { \prg_return_false: }
+    \seq_break_point:n { \group_end: }
+  }
+\cs_new_nopar:Npn \seq_if_in_aux: { \seq_break:n { \prg_return_true: } }
+\cs_generate_variant:Nn \seq_if_in:NnT  {     NV , Nv , No , Nx }
+\cs_generate_variant:Nn \seq_if_in:NnT  { c , cV , cv , co , cx }
+\cs_generate_variant:Nn \seq_if_in:NnF  {     NV , Nv , No , Nx }
+\cs_generate_variant:Nn \seq_if_in:NnF  { c , cV , cv , co , cx }
+\cs_generate_variant:Nn \seq_if_in:NnTF {     NV , Nv , No , Nx }
+\cs_generate_variant:Nn \seq_if_in:NnTF { c , cV , cv , co , cx }
+\cs_new_protected_nopar:Npn \seq_get_left:NN #1#2
+  {
+    \seq_if_empty_err_break:N #1
+    \exp_after:wN \seq_get_left_aux:NnwN #1 \q_stop #2
+    \seq_break_point:n { }
+  }
+\cs_new_protected:Npn \seq_get_left_aux:NnwN \seq_item:n #1#2 \q_stop #3
+  { \tl_set:Nn #3 {#1} }
+\cs_generate_variant:Nn \seq_get_left:NN { c }
+\cs_new_protected_nopar:Npn \seq_pop_left:NN
+  { \seq_pop_left_aux:NNN \tl_set:Nn }
+\cs_new_protected_nopar:Npn \seq_gpop_left:NN
+  { \seq_pop_left_aux:NNN \tl_gset:Nn }
+\cs_new_protected_nopar:Npn \seq_pop_left_aux:NNN #1#2#3
+  {
+    \seq_if_empty_err_break:N #2
+    \exp_after:wN \seq_pop_left_aux:NnwNNN #2 \q_stop #1#2#3
+    \seq_break_point:n { }
+  }
+\cs_new_protected:Npn \seq_pop_left_aux:NnwNNN \seq_item:n #1#2 \q_stop #3#4#5
+  {
+    #3 #4 {#2}
+    \tl_set:Nn #5 {#1}
+  }
+\cs_generate_variant:Nn \seq_pop_left:NN  { c }
+\cs_generate_variant:Nn \seq_gpop_left:NN { c }
+\cs_new_protected_nopar:Npn \seq_get_right:NN #1#2
+  {
+    \seq_if_empty_err_break:N #1
+    \seq_get_right_aux:NN #1#2
+    \seq_break_point:n { }
+  }
+\cs_new_protected_nopar:Npn \seq_get_right_aux:NN #1#2
+  {
+    \seq_push_item_def:n { }
+    \exp_after:wN \exp_after:wN \exp_after:wN \seq_get_right_loop:nn
+      \exp_after:wN \use_none:n #1
+      { \tl_set:Nn #2 }
+      { }
+      {
+        \seq_pop_item_def:
+        \seq_break:
+      }
+}
+\cs_new:Npn \seq_get_right_loop:nn #1#2
+  {
+    #2 {#1}
+    \seq_get_right_loop:nn
+  }
+\cs_generate_variant:Nn \seq_get_right:NN { c }
+\cs_new_protected_nopar:Npn \seq_pop_right:NN
+  { \seq_pop_right_aux:NNN \tl_set:Nx }
+\cs_new_protected_nopar:Npn \seq_gpop_right:NN
+  { \seq_pop_right_aux:NNN \tl_gset:Nx }
+\cs_new_protected_nopar:Npn \seq_pop_right_aux:NNN #1#2#3
+  {
+    \seq_if_empty_err_break:N #2
+    \seq_pop_right_aux_ii:NNN #1 #2 #3
+    \seq_break_point:n { }
+  }
+\cs_new_protected_nopar:Npn \seq_pop_right_aux_ii:NNN #1#2#3
+  {
+    \seq_push_item_def:n { \exp_not:n { \seq_item:n {##1} } }
+    #1 #2 { \if_false: } \fi:
+      \exp_after:wN \exp_after:wN \exp_after:wN \seq_get_right_loop:nn
+         \exp_after:wN \use_none:n #2
+        {
+          \if_false: { \fi: }
+          \tl_set:Nn #3
+        }
+        { }
+        {
+          \seq_pop_item_def:
+          \seq_break:
+        }
+  }
+\cs_generate_variant:Nn \seq_pop_right:NN  { c }
+\cs_generate_variant:Nn \seq_gpop_right:NN { c }
+\cs_new:Npn \seq_break:  #1   \seq_break_point:n #2 {#2}
+\cs_new:Npn \seq_break:n #1#2 \seq_break_point:n #3 { #3 #1 }
+\cs_new_eq:NN \seq_map_break:  \seq_break:
+\cs_new_eq:NN \seq_map_break:n \seq_break:n
+\cs_new_eq:NN \seq_break_point:n \use:n
+\cs_new_protected_nopar:Npn \seq_if_empty_err_break:N #1
+  {
+    \if_meaning:w #1 \c_empty_tl
+      \msg_kernel_error:nnx { seq } { empty-sequence } { \token_to_str:N #1 }
+      \exp_after:wN \seq_break:
+    \fi:
+  }
+\cs_new:Npn \seq_map_function:NN #1#2
+  {
+    \exp_after:wN \seq_map_function_aux:NNn \exp_after:wN #2 #1
+      { ? \seq_map_break: } { }
+    \seq_break_point:n { }
+  }
+\cs_new:Npn \seq_map_function_aux:NNn #1#2#3
+  {
+    \use_none:n #2
+    #1 {#3}
+    \seq_map_function_aux:NNn #1
+  }
+\cs_generate_variant:Nn \seq_map_function:NN { c }
+\cs_new_protected:Npn \seq_push_item_def:n
+  {
+    \seq_push_item_def_aux:
+    \cs_gset:Npn \seq_item:n ##1
+  }
+\cs_new_protected:Npn \seq_push_item_def:x
+  {
+    \seq_push_item_def_aux:
+    \cs_gset:Npx \seq_item:n ##1
+  }
+\cs_new_protected:Npn \seq_push_item_def_aux:
+  {
+    \cs_gset_eq:cN { seq_item_ \int_use:N \g_seq_nesting_depth_int :n }
+      \seq_item:n
+    \int_gincr:N \g_seq_nesting_depth_int
+  }
+\cs_new_protected_nopar:Npn \seq_pop_item_def:
+  {
+    \int_gdecr:N \g_seq_nesting_depth_int
+    \cs_gset_eq:Nc \seq_item:n
+      { seq_item_ \int_use:N \g_seq_nesting_depth_int :n }
+  }
+\cs_new_protected:Npn \seq_map_inline:Nn #1#2
+  {
+    \seq_push_item_def:n {#2}
+    #1
+    \seq_break_point:n { \seq_pop_item_def: }
+  }
+\cs_generate_variant:Nn \seq_map_inline:Nn { c }
+\cs_new_protected:Npn \seq_map_variable:NNn #1#2#3
+  {
+    \seq_push_item_def:x
+      {
+        \tl_set:Nn \exp_not:N #2 {##1}
+        \exp_not:n {#3}
+      }
+    #1
+    \seq_break_point:n { \seq_pop_item_def: }
+  }
+\cs_generate_variant:Nn \seq_map_variable:NNn {     Nc }
+\cs_generate_variant:Nn \seq_map_variable:NNn { c , cc }
+\cs_new_eq:NN \seq_push:Nn  \seq_put_left:Nn
+\cs_new_eq:NN \seq_push:NV  \seq_put_left:NV
+\cs_new_eq:NN \seq_push:Nv  \seq_put_left:Nv
+\cs_new_eq:NN \seq_push:No  \seq_put_left:No
+\cs_new_eq:NN \seq_push:Nx  \seq_put_left:Nx
+\cs_new_eq:NN \seq_push:cn  \seq_put_left:cn
+\cs_new_eq:NN \seq_push:cV  \seq_put_left:cV
+\cs_new_eq:NN \seq_push:cv  \seq_put_left:cv
+\cs_new_eq:NN \seq_push:co  \seq_put_left:co
+\cs_new_eq:NN \seq_push:cx  \seq_put_left:cx
+\cs_new_eq:NN \seq_gpush:Nn \seq_gput_left:Nn
+\cs_new_eq:NN \seq_gpush:NV \seq_gput_left:NV
+\cs_new_eq:NN \seq_gpush:Nv \seq_gput_left:Nv
+\cs_new_eq:NN \seq_gpush:No \seq_gput_left:No
+\cs_new_eq:NN \seq_gpush:Nx \seq_gput_left:Nx
+\cs_new_eq:NN \seq_gpush:cn \seq_gput_left:cn
+\cs_new_eq:NN \seq_gpush:cV \seq_gput_left:cV
+\cs_new_eq:NN \seq_gpush:cv \seq_gput_left:cv
+\cs_new_eq:NN \seq_gpush:co \seq_gput_left:co
+\cs_new_eq:NN \seq_gpush:cx \seq_gput_left:cx
+\cs_new_eq:NN \seq_get:NN \seq_get_left:NN
+\cs_new_eq:NN \seq_get:cN \seq_get_left:cN
+\cs_new_eq:NN \seq_pop:NN \seq_pop_left:NN
+\cs_new_eq:NN \seq_pop:cN \seq_pop_left:cN
+\cs_new_eq:NN \seq_gpop:NN \seq_gpop_left:NN
+\cs_new_eq:NN \seq_gpop:cN \seq_gpop_left:cN
+\tl_new:N \l_seq_show_tl
+\cs_new_protected_nopar:Npn \seq_show:N #1
+  {
+    \seq_if_empty:NTF #1
+      {
+        \iow_term:x { Sequence~\token_to_str:N #1 \c_space_tl is~empty }
+        \tl_show:n { }
+      }
+      {
+        \iow_term:x
+          {
+            Sequence~\token_to_str:N #1 \c_space_tl
+            contains~the~items~(without~outer~braces):
+          }
+        \tl_set:Nx \l_seq_show_tl
+          { \seq_map_function:NN #1 \seq_show_aux:n }
+          \etex_showtokens:D \exp_after:wN \exp_after:wN \exp_after:wN
+            { \exp_after:wN \seq_show_aux:w \l_seq_show_tl }
+      }
+  }
+\cs_new:Npn \seq_show_aux:n #1
+  {
+    \iow_newline: > \c_space_tl \c_space_tl
+    \iow_char:N \{ \exp_not:n {#1} \iow_char:N \}
+  }
+\cs_new:Npn \seq_show_aux:w #1 > ~ { }
+\cs_generate_variant:Nn \seq_show:N { c }
+\cs_new_nopar:Npn \seq_if_empty_break_return_false:N #1
+  {
+    \if_meaning:w #1 \c_empty_tl
+      \prg_return_false:
+      \exp_after:wN \seq_break:
+    \fi:
+  }
+\prg_new_protected_conditional:Npnn \seq_get_left:NN #1 #2 { T , F , TF }
+  {
+    \seq_if_empty_break_return_false:N #1
+    \exp_after:wN \seq_get_left_aux:Nw #1 \q_stop #2
+    \prg_return_true:
+    \seq_break:
+    \seq_break_point:n { }
+  }
+\prg_new_protected_conditional:Npnn \seq_get_right:NN #1#2 { T , F , TF }
+  {
+    \seq_if_empty_break_return_false:N #1
+    \seq_get_right_aux:NN #1#2
+    \prg_return_true: \seq_break:
+    \seq_break_point:n { }
+  }
+\cs_generate_variant:Nn \seq_get_left:NNT   { c }
+\cs_generate_variant:Nn \seq_get_left:NNF   { c }
+\cs_generate_variant:Nn \seq_get_left:NNTF  { c }
+\cs_generate_variant:Nn \seq_get_right:NNT  { c }
+\cs_generate_variant:Nn \seq_get_right:NNF  { c }
+\cs_generate_variant:Nn \seq_get_right:NNTF { c }
+\prg_new_protected_conditional:Npnn \seq_pop_left:NN #1#2 { T , F , TF }
+  {
+    \seq_if_empty_break_return_false:N #1
+    \exp_after:wN \seq_pop_left_aux:NnwNNN #1 \q_stop \tl_set:Nn #1#2
+    \prg_return_true: \seq_break:
+    \seq_break_point:n { }
+  }
+\prg_new_protected_conditional:Npnn \seq_gpop_left:NN #1#2 { T , F , TF }
+  {
+    \seq_if_empty_break_return_false:N #1
+    \exp_after:wN \seq_pop_left_aux:NnwNNN #1 \q_stop \tl_gset:Nn #1#2
+    \prg_return_true: \seq_break:
+    \seq_break_point:n { }
+  }
+\prg_new_protected_conditional:Npnn \seq_pop_right:NN #1#2 { T , F , TF }
+  {
+    \seq_if_empty_break_return_false:N #1
+    \seq_pop_right_aux_ii:NNN \tl_set:Nx #1 #2
+    \prg_return_true: \seq_break:
+    \seq_break_point:n { }
+  }
+\prg_new_protected_conditional:Npnn \seq_gpop_right:NN #1#2 { T , F , TF }
+  {
+    \seq_if_empty_break_return_false:N #1
+    \seq_pop_right_aux_ii:NNN \tl_gset:Nx #1 #2
+    \prg_return_true: \seq_break:
+    \seq_break_point:n { }
+  }
+\cs_generate_variant:Nn \seq_pop_left:NNT    { c }
+\cs_generate_variant:Nn \seq_pop_left:NNF    { c }
+\cs_generate_variant:Nn \seq_pop_left:NNTF   { c }
+\cs_generate_variant:Nn \seq_gpop_left:NNT   { c }
+\cs_generate_variant:Nn \seq_gpop_left:NNF   { c }
+\cs_generate_variant:Nn \seq_gpop_left:NNTF  { c }
+\cs_generate_variant:Nn \seq_pop_right:NNT   { c }
+\cs_generate_variant:Nn \seq_pop_right:NNF   { c }
+\cs_generate_variant:Nn \seq_pop_right:NNTF  { c }
+\cs_generate_variant:Nn \seq_gpop_right:NNT  { c }
+\cs_generate_variant:Nn \seq_gpop_right:NNF  { c }
+\cs_generate_variant:Nn \seq_gpop_right:NNTF { c }
+\cs_new:Npn \seq_length:N #1
+  {
+    \int_eval:n
+      {
+        0
+        \seq_map_function:NN #1 \seq_length_aux:n
+      }
+  }
+\cs_new:Npn \seq_length_aux:n #1 { +1 }
+\cs_generate_variant:Nn \seq_length:N { c }
+\cs_new_nopar:Npn \seq_item:Nn #1#2
+  {
+    \exp_last_unbraced:Nfo \seq_item_aux:nnn
+      {
+        \int_eval:n
+          {
+            \int_compare:nNnT {#2} < \c_zero
+              { \seq_length:N #1 + }
+            #2
+          }
+      }
+    #1
+    { ? \seq_break: }
+    { }
+    \seq_break_point:n { }
+  }
+\cs_new_nopar:Npn \seq_item_aux:nnn #1#2#3
+  {
+    \use_none:n #2
+    \int_compare:nNnTF {#1} = \c_zero
+      { \seq_break:n {#3} }
+      { \exp_args:Nf \seq_item_aux:nnn { #1 - 1 } }
+  }
+\cs_generate_variant:Nn \seq_item:Nn { c }
+\cs_new_nopar:Npn \seq_use:N #1 { \seq_map_function:NN #1 \use:n }
+\cs_generate_variant:Nn \seq_use:N { c }
+\cs_new_nopar:Npn \seq_mapthread_function:NNN #1#2#3
+  {
+    \exp_after:wN \seq_mapthread_function_aux:NN
+      \exp_after:wN #3
+      \exp_after:wN #1
+      #2
+      { ? \seq_break: } { }
+    \seq_break_point:n { }
+  }
+\cs_new_nopar:Npn \seq_mapthread_function_aux:NN #1#2
+  {
+    \exp_after:wN \seq_mapthread_function_aux:Nnnwnn
+      \exp_after:wN #1
+      #2
+      { ? \seq_break: } { }
+      \q_stop
+  }
+\cs_new:Npn \seq_mapthread_function_aux:Nnnwnn #1#2#3#4 \q_stop #5#6
+  {
+    \use_none:n #2
+    \use_none:n #5
+    #1 {#3} {#6}
+    \seq_mapthread_function_aux:Nnnwnn #1 #4 \q_stop
+  }
+\cs_generate_variant:Nn \seq_mapthread_function:NNN {     Nc }
+\cs_generate_variant:Nn \seq_mapthread_function:NNN { c , cc }
+\cs_new_protected:Npn \seq_set_from_clist:NN #1#2
+  {
+    \tl_set:Nx #1
+      { \clist_map_function:NN #2 \seq_wrap_item:n }
+  }
+\cs_new_protected:Npn \seq_set_from_clist:Nn #1#2
+  {
+    \tl_set:Nx #1
+      { \clist_map_function:nN {#2} \seq_wrap_item:n }
+  }
+\cs_new_protected:Npn \seq_gset_from_clist:NN #1#2
+  {
+    \tl_gset:Nx #1
+      { \clist_map_function:NN #2 \seq_wrap_item:n }
+  }
+\cs_new_protected:Npn \seq_gset_from_clist:Nn #1#2
+  {
+    \tl_gset:Nx #1
+      { \clist_map_function:nN {#2} \seq_wrap_item:n }
+  }
+\cs_new:Npn \seq_wrap_item:n #1 { \exp_not:n { \seq_item:n {#1} } }
+\cs_generate_variant:Nn \seq_set_from_clist:NN  {     Nc }
+\cs_generate_variant:Nn \seq_set_from_clist:NN  { c , cc }
+\cs_generate_variant:Nn \seq_set_from_clist:Nn  { c      }
+\cs_generate_variant:Nn \seq_gset_from_clist:NN {     Nc }
+\cs_generate_variant:Nn \seq_gset_from_clist:NN { c , cc }
+\cs_generate_variant:Nn \seq_gset_from_clist:Nn { c      }
+\cs_new_eq:NN \seq_top:NN \seq_get_left:NN
+\cs_new_eq:NN \seq_top:cN \seq_get_left:cN
+\cs_new_eq:NN \seq_display:N \seq_show:N
+\cs_new_eq:NN \seq_display:c \seq_show:c
+%% 
+%%
+%% End of file `l3seq.sty'.
diff --git a/Master/texmf-dist/tex/latex/l3kernel/l3skip.sty b/Master/texmf-dist/tex/latex/l3kernel/l3skip.sty
new file mode 100644
index 00000000000..ae4a46fc805
--- /dev/null
+++ b/Master/texmf-dist/tex/latex/l3kernel/l3skip.sty
@@ -0,0 +1,332 @@
+%%
+%% This is file `l3skip.sty',
+%% generated with the docstrip utility.
+%%
+%% The original source files were:
+%%
+%% l3skip.dtx  (with options: `package')
+%% 
+%% EXPERIMENTAL CODE
+%% 
+%% Do not distribute this file without also distributing the
+%% source files specified above.
+%% 
+%% Do not distribute a modified version of this file.
+%% 
+%% File: l3skip.dtx Copyright (C) 2004-2011 Frank Mittelbach, The LaTeX3 Project
+%%
+%% It may be distributed and/or modified under the conditions of the
+%% LaTeX Project Public License (LPPL), either version 1.3c of this
+%% license or (at your option) any later version.  The latest version
+%% of this license is in the file
+%%
+%%    http://www.latex-project.org/lppl.txt
+%%
+%% This file is part of the "expl3 bundle" (The Work in LPPL)
+%% and all files in that bundle must be distributed together.
+%%
+%% The released version of this bundle is available from CTAN.
+%%
+%% -----------------------------------------------------------------------
+%%
+%% The development version of the bundle can be found at
+%%
+%%    http://www.latex-project.org/svnroot/experimental/trunk/
+%%
+%% for those people who are interested.
+%%
+%%%%%%%%%%%
+%% NOTE: %%
+%%%%%%%%%%%
+%%
+%%   Snapshots taken from the repository represent work in progress and may
+%%   not work or may contain conflicting material!  We therefore ask
+%%   people _not_ to put them into distributions, archives, etc. without
+%%   prior consultation with the LaTeX3 Project.
+%%
+%% -----------------------------------------------------------------------
+\RequirePackage{l3names}
+\GetIdInfo$Id: l3skip.dtx 2478 2011-06-19 21:34:23Z joseph $
+  {L3 Experimental dimensions and skips}
+\ProvidesExplPackage
+  {\ExplFileName}{\ExplFileDate}{\ExplFileVersion}{\ExplFileDescription}
+\package_check_loaded_expl:
+\cs_new_eq:NN \if_dim:w      \tex_ifdim:D
+\cs_new_eq:NN \dim_eval:w    \etex_dimexpr:D
+\cs_new_eq:NN \dim_eval_end: \tex_relax:D
+\cs_new_protected_nopar:Npn \dim_new:N #1
+  {
+    \chk_if_free_cs:N #1
+    \newdimen #1
+  }
+\cs_generate_variant:Nn \dim_new:N { c }
+\cs_new_protected_nopar:Npn \dim_zero:N #1 { #1 \c_zero_dim }
+\cs_new_protected_nopar:Npn \dim_gzero:N { \pref_global:D \dim_zero:N }
+\cs_generate_variant:Nn \dim_zero:N  { c }
+\cs_generate_variant:Nn \dim_gzero:N { c }
+\cs_new_protected_nopar:Npn \dim_set:Nn #1#2
+  { #1 ~ \dim_eval:w #2 \dim_eval_end: }
+\cs_new_protected_nopar:Npn \dim_gset:Nn { \pref_global:D \dim_set:Nn }
+\cs_generate_variant:Nn \dim_set:Nn  { c }
+\cs_generate_variant:Nn \dim_gset:Nn { c }
+\cs_new_protected_nopar:Npn \dim_set_eq:NN #1#2 { #1 = #2 }
+\cs_generate_variant:Nn \dim_set_eq:NN {       c }
+\cs_generate_variant:Nn \dim_set_eq:NN { Nc , cc }
+\cs_new_protected_nopar:Npn \dim_gset_eq:NN #1#2 { \pref_global:D #1 = #2 }
+\cs_generate_variant:Nn \dim_gset_eq:NN {       c }
+\cs_generate_variant:Nn \dim_gset_eq:NN { Nc , cc }
+\cs_new_protected_nopar:Npn \dim_set_max:Nn #1#2
+  { \dim_compare:nNnT {#1} < {#2} { \dim_set:Nn #1 {#2} } }
+\cs_new_protected_nopar:Npn \dim_gset_max:Nn #1#2
+  { \dim_compare:nNnT {#1} < {#2} { \dim_gset:Nn #1 {#2} } }
+\cs_new_protected_nopar:Npn \dim_set_min:Nn #1#2
+  { \dim_compare:nNnT {#1} > {#2} { \dim_set:Nn #1 {#2} } }
+\cs_new_protected_nopar:Npn \dim_gset_min:Nn #1#2
+  { \dim_compare:nNnT {#1} > {#2} { \dim_gset:Nn #1 {#2} } }
+\cs_generate_variant:Nn \dim_set_max:Nn  { c }
+\cs_generate_variant:Nn \dim_gset_max:Nn { c }
+\cs_generate_variant:Nn \dim_set_min:Nn  { c }
+\cs_generate_variant:Nn \dim_gset_min:Nn { c }
+\cs_new_protected_nopar:Npn \dim_add:Nn #1#2
+  { \tex_advance:D #1 by \dim_eval:w #2 \dim_eval_end: }
+\cs_new_protected_nopar:Npn \dim_gadd:Nn { \pref_global:D \dim_add:Nn }
+\cs_generate_variant:Nn \dim_add:Nn  { c }
+\cs_generate_variant:Nn \dim_gadd:Nn { c }
+\cs_new_protected_nopar:Npn \dim_sub:Nn #1#2
+  { \tex_advance:D #1 by - \dim_eval:w #2 \dim_eval_end: }
+\cs_new_protected_nopar:Npn \dim_gsub:Nn { \pref_global:D \dim_sub:Nn }
+\cs_generate_variant:Nn \dim_sub:Nn  { c }
+\cs_generate_variant:Nn \dim_gsub:Nn { c }
+\cs_new_nopar:Npn \dim_ratio:nn #1#2
+  { \dim_ratio_aux:n {#1} / \dim_ratio_aux:n {#2} }
+\cs_new_nopar:Npn \dim_ratio_aux:n #1
+  { \exp_after:wN \int_value:w \dim_eval:w #1 \dim_eval_end: }
+\prg_new_conditional:Npnn \dim_compare:nNn #1#2#3 { p , T , F , TF }
+  {
+    \if_dim:w \dim_eval:w #1 #2 \dim_eval:w #3 \dim_eval_end:
+      \prg_return_true: \else: \prg_return_false: \fi:
+  }
+\prg_new_conditional:Npnn \dim_compare:n #1 { p , T , F , TF }
+  {
+    \exp_after:wN \dim_compare_aux:wNN \dim_use:N \dim_eval:w #1
+        \dim_eval_end:
+      \prg_return_true:
+    \else:
+      \prg_return_false:
+    \fi:
+  }
+\exp_args:Nno \use:nn
+  { \cs_new:Npn \dim_compare_aux:wNN #1 }
+  { \tl_to_str:n { pt } }
+  #2 #3
+    {
+      \use:c
+        {
+          dim_compare_ #2
+            \if_meaning:w = #3 = \fi:
+          :nw
+        }
+      { #1 pt } #3
+    }
+\cs_new:cpn { dim_compare_<:nw } #1 { \if_dim:w #1 < \dim_eval:w }
+\cs_new:cpn { dim_compare_=:nw } #1 { \if_dim:w #1 = \dim_eval:w }
+\cs_new:cpn { dim_compare_>:nw } #1 { \if_dim:w #1 > \dim_eval:w }
+\cs_new:cpn {dim_compare_==:nw} #1#2 { \if_dim:w #1 = \dim_eval:w }
+\cs_new:cpn {dim_compare_<=:nw} #1#2 {\reverse_if:N \if_dim:w #1 > \dim_eval:w}
+\cs_new:cpn {dim_compare_!=:nw} #1#2 {\reverse_if:N \if_dim:w #1 = \dim_eval:w}
+\cs_new:cpn {dim_compare_>=:nw} #1#2 {\reverse_if:N \if_dim:w #1 < \dim_eval:w}
+\cs_set:Npn \dim_while_do:nn #1#2
+  {
+    \dim_compare:nT {#1}
+      {
+        #2
+        \dim_while_do:nn {#1} {#2}
+      }
+  }
+\cs_set:Npn \dim_until_do:nn #1#2
+  {
+    \dim_compare:nF {#1}
+      {
+        #2
+        \dim_until_do:nn {#1} {#2}
+      }
+  }
+\cs_set:Npn \dim_do_while:nn #1#2
+  {
+    #2
+    \dim_compare:nT {#1}
+      { \dim_do_while:nNnn {#1} {#2} }
+  }
+\cs_set:Npn \dim_do_until:nn #1#2
+  {
+    #2
+    \dim_compare:nF {#1}
+      { \dim_do_until:nn {#1} {#2} }
+  }
+\cs_set:Npn \dim_while_do:nNnn #1#2#3#4
+  {
+    \dim_compare:nNnT {#1} #2 {#3}
+      {
+        #4
+        \dim_while_do:nNnn {#1} #2 {#3} {#4}
+      }
+  }
+\cs_set:Npn \dim_until_do:nNnn #1#2#3#4
+  {
+  \dim_compare:nNnF {#1} #2 {#3}
+    {
+      #4
+      \dim_until_do:nNnn {#1} #2 {#3} {#4}
+    }
+  }
+\cs_set:Npn \dim_do_while:nNnn #1#2#3#4
+  {
+    #4
+    \dim_compare:nNnT {#1} #2 {#3}
+      { \dim_do_while:nNnn {#1} #2 {#3} {#4} }
+  }
+\cs_set:Npn \dim_do_until:nNnn #1#2#3#4
+  {
+    #4
+    \dim_compare:nNnF {#1} #2 {#3}
+      { \dim_do_until:nNnn {#1} #2 {#3} {#4} }
+  }
+\cs_new_nopar:Npn \dim_eval:n #1
+  { \dim_use:N \dim_eval:w #1 \dim_eval_end: }
+\cs_new_eq:NN \dim_use:N \tex_the:D
+\cs_generate_variant:Nn \dim_use:N { c }
+\cs_new_eq:NN  \dim_show:N \kernel_register_show:N
+\cs_generate_variant:Nn \dim_show:N { c }
+\cs_new_eq:NN \c_zero_dim \z@
+\cs_new_eq:NN \c_max_dim  \maxdimen
+\dim_new:N \l_tmpa_dim
+\dim_new:N \l_tmpb_dim
+\dim_new:N \l_tmpc_dim
+\dim_new:N \g_tmpa_dim
+\dim_new:N \g_tmpb_dim
+\cs_new_protected_nopar:Npn \skip_new:N #1
+  {
+    \chk_if_free_cs:N #1
+    \newskip #1
+  }
+\cs_generate_variant:Nn \skip_new:N { c }
+\cs_new_protected_nopar:Npn \skip_zero:N #1 { #1 \c_zero_skip }
+\cs_new_protected_nopar:Npn \skip_gzero:N { \pref_global:D \skip_zero:N }
+\cs_generate_variant:Nn \skip_zero:N  { c }
+\cs_generate_variant:Nn \skip_gzero:N { c }
+\cs_new_protected_nopar:Npn \skip_set:Nn #1#2
+  { #1 ~ \etex_glueexpr:D #2 \scan_stop: }
+\cs_new_protected_nopar:Npn \skip_gset:Nn { \pref_global:D \skip_set:Nn }
+\cs_generate_variant:Nn \skip_set:Nn  { c }
+\cs_generate_variant:Nn \skip_gset:Nn { c }
+\cs_new_protected_nopar:Npn \skip_set_eq:NN #1#2 { #1 = #2 }
+\cs_generate_variant:Nn \skip_set_eq:NN {       c }
+\cs_generate_variant:Nn \skip_set_eq:NN { Nc , cc }
+\cs_new_protected_nopar:Npn \skip_gset_eq:NN #1#2 { \pref_global:D #1 = #2 }
+\cs_generate_variant:Nn \skip_gset_eq:NN {       c }
+\cs_generate_variant:Nn \skip_gset_eq:NN { Nc , cc }
+\cs_new_protected_nopar:Npn \skip_add:Nn #1#2
+  { \tex_advance:D #1 by \etex_glueexpr:D #2 \scan_stop: }
+\cs_new_protected_nopar:Npn \skip_gadd:Nn { \pref_global:D \skip_add:Nn }
+\cs_generate_variant:Nn \skip_add:Nn  { c }
+\cs_generate_variant:Nn \skip_gadd:Nn { c }
+\cs_new_protected_nopar:Npn \skip_sub:Nn #1#2
+  { \tex_advance:D #1 by - \etex_glueexpr:D #2 \scan_stop: }
+\cs_new_protected_nopar:Npn \skip_gsub:Nn { \pref_global:D \skip_sub:Nn }
+\cs_generate_variant:Nn \skip_sub:Nn  { c }
+\cs_generate_variant:Nn \skip_gsub:Nn { c }
+\prg_new_conditional:Npnn \skip_if_eq:nn #1#2 { p , T , F , TF }
+  {
+    \if_int_compare:w
+      \pdftex_strcmp:D { \skip_eval:n { #1 } } { \skip_eval:n { #2 } }
+      = \c_zero
+        \prg_return_true:
+    \else:
+        \prg_return_false:
+    \fi:
+  }
+\prg_new_conditional:Npnn \skip_if_infinite_glue:n #1 { p , T , F , TF }
+  {
+    \bool_if:nTF
+      {
+        \int_compare_p:nNn { \etex_gluestretchorder:D #1 } > \c_zero ||
+        \int_compare_p:nNn { \etex_glueshrinkorder:D  #1 } > \c_zero
+      }
+      { \prg_return_true: }
+      { \prg_return_false: }
+ }
+\cs_new_nopar:Npn \skip_eval:n #1
+  { \skip_use:N \etex_glueexpr:D #1 \scan_stop: }
+\cs_new_eq:NN \skip_use:N \tex_the:D
+\cs_generate_variant:Nn \skip_use:N { c }
+\cs_new_eq:NN  \skip_horizontal:N \tex_hskip:D
+\cs_new_nopar:Npn \skip_horizontal:n #1
+  { \skip_horizontal:N \etex_glueexpr:D #1 \scan_stop: }
+\cs_new_eq:NN  \skip_vertical:N \tex_vskip:D
+\cs_new_nopar:Npn \skip_vertical:n #1
+  { \skip_vertical:N \etex_glueexpr:D #1 \scan_stop: }
+\cs_generate_variant:Nn \skip_horizontal:N { c }
+\cs_generate_variant:Nn \skip_vertical:N { c }
+\cs_new_eq:NN  \skip_show:N \kernel_register_show:N
+\cs_generate_variant:Nn \skip_show:N { c }
+\cs_new_eq:NN \c_zero_skip \c_zero_dim
+\cs_new_eq:NN \c_max_skip  \c_max_dim
+\skip_new:N \l_tmpa_skip
+\skip_new:N \l_tmpb_skip
+\skip_new:N \l_tmpc_skip
+\skip_new:N \g_tmpa_skip
+\skip_new:N \g_tmpb_skip
+\cs_new_protected_nopar:Npn \muskip_new:N #1
+  {
+    \chk_if_free_cs:N #1
+    \newmuskip #1
+  }
+\cs_generate_variant:Nn \muskip_new:N { c }
+\cs_new_protected_nopar:Npn \muskip_zero:N #1
+  { #1 \c_zero_muskip }
+\cs_new_protected_nopar:Npn \muskip_gzero:N { \pref_global:D \muskip_zero:N }
+\cs_generate_variant:Nn \muskip_zero:N  { c }
+\cs_generate_variant:Nn \muskip_gzero:N { c }
+\cs_new_protected_nopar:Npn \muskip_set:Nn #1#2
+  { #1 ~ \etex_muexpr:D #2 \scan_stop: }
+\cs_new_protected_nopar:Npn \muskip_gset:Nn { \pref_global:D \muskip_set:Nn }
+\cs_generate_variant:Nn \muskip_set:Nn  { c }
+\cs_generate_variant:Nn \muskip_gset:Nn { c }
+\cs_new_protected_nopar:Npn \muskip_set_eq:NN #1#2 { #1 = #2 }
+\cs_generate_variant:Nn \muskip_set_eq:NN {       c }
+\cs_generate_variant:Nn \muskip_set_eq:NN { Nc , cc }
+\cs_new_protected_nopar:Npn \muskip_gset_eq:NN #1#2 { \pref_global:D #1 = #2 }
+\cs_generate_variant:Nn \muskip_gset_eq:NN {       c }
+\cs_generate_variant:Nn \muskip_gset_eq:NN { Nc , cc }
+\cs_new_protected_nopar:Npn \muskip_add:Nn #1#2
+  { \tex_advance:D #1 by \etex_muexpr:D #2 \scan_stop: }
+\cs_new_protected_nopar:Npn \muskip_gadd:Nn { \pref_global:D \muskip_add:Nn }
+\cs_generate_variant:Nn \muskip_add:Nn  { c }
+\cs_generate_variant:Nn \muskip_gadd:Nn { c }
+\cs_new_protected_nopar:Npn \muskip_sub:Nn #1#2
+  { \tex_advance:D #1 by - \etex_muexpr:D #2 \scan_stop: }
+\cs_new_protected_nopar:Npn \muskip_gsub:Nn { \pref_global:D \muskip_sub:Nn }
+\cs_generate_variant:Nn \muskip_sub:Nn  { c }
+\cs_generate_variant:Nn \muskip_gsub:Nn { c }
+\cs_new_nopar:Npn \muskip_eval:n #1
+  { \muskip_use:N \etex_muexpr:D #1 \scan_stop: }
+\cs_new_eq:NN \muskip_use:N \tex_the:D
+\cs_generate_variant:Nn \muskip_use:N { c }
+\cs_new_eq:NN  \muskip_show:N \kernel_register_show:N
+\cs_generate_variant:Nn \muskip_show:N { c }
+\cs_new_nopar:Npn \skip_split_finite_else_action:nnNN #1#2#3#4
+  {
+    \skip_if_infinite_glue:nTF {#1}
+    {
+      #3 = \c_zero_skip
+      #4 = \c_zero_skip
+      #2
+    }
+    {
+      #3 = \etex_gluestretch:D #1 \scan_stop:
+      #4 = \etex_glueshrink:D  #1 \scan_stop:
+    }
+  }
+%% 
+%%
+%% End of file `l3skip.sty'.
diff --git a/Master/texmf-dist/tex/latex/l3kernel/l3tl.sty b/Master/texmf-dist/tex/latex/l3kernel/l3tl.sty
new file mode 100644
index 00000000000..34238f643f0
--- /dev/null
+++ b/Master/texmf-dist/tex/latex/l3kernel/l3tl.sty
@@ -0,0 +1,628 @@
+%%
+%% This is file `l3tl.sty',
+%% generated with the docstrip utility.
+%%
+%% The original source files were:
+%%
+%% l3tl.dtx  (with options: `package')
+%% 
+%% EXPERIMENTAL CODE
+%% 
+%% Do not distribute this file without also distributing the
+%% source files specified above.
+%% 
+%% Do not distribute a modified version of this file.
+%% 
+%% File: l3tl.dtx Copyright (C) 1990-2011 The LaTeX3 Project
+%%
+%% It may be distributed and/or modified under the conditions of the
+%% LaTeX Project Public License (LPPL), either version 1.3c of this
+%% license or (at your option) any later version.  The latest version
+%% of this license is in the file
+%%
+%%    http://www.latex-project.org/lppl.txt
+%%
+%% This file is part of the "expl3 bundle" (The Work in LPPL)
+%% and all files in that bundle must be distributed together.
+%%
+%% The released version of this bundle is available from CTAN.
+%%
+%% -----------------------------------------------------------------------
+%%
+%% The development version of the bundle can be found at
+%%
+%%    http://www.latex-project.org/svnroot/experimental/trunk/
+%%
+%% for those people who are interested.
+%%
+%%%%%%%%%%%
+%% NOTE: %%
+%%%%%%%%%%%
+%%
+%%   Snapshots taken from the repository represent work in progress and may
+%%   not work or may contain conflicting material!  We therefore ask
+%%   people _not_ to put them into distributions, archives, etc. without
+%%   prior consultation with the LaTeX3 Project.
+%%
+%% -----------------------------------------------------------------------
+\RequirePackage{l3names}
+\GetIdInfo$Id: l3tl.dtx 2498 2011-07-09 17:51:47Z joseph $
+  {L3 Experimental token lists}
+\ProvidesExplPackage
+  {\ExplFileName}{\ExplFileDate}{\ExplFileVersion}{\ExplFileDescription}
+\package_check_loaded_expl:
+\cs_new_protected_nopar:Npn \tl_new:N #1
+  {
+    \chk_if_free_cs:N #1
+    \cs_gset_eq:NN #1 \c_empty_tl
+  }
+\cs_generate_variant:Nn \tl_new:N { c }
+\cs_new_protected:Npn \tl_const:Nn #1#2
+  {
+    \chk_if_free_cs:N #1
+    \cs_gset_nopar:Npx #1 { \exp_not:n {#2} }
+  }
+\cs_new_protected:Npn \tl_const:Nx #1#2
+  {
+    \chk_if_free_cs:N #1
+    \cs_gset_nopar:Npx #1 {#2}
+  }
+\cs_generate_variant:Nn \tl_const:Nn { c }
+\cs_generate_variant:Nn \tl_const:Nx { c }
+\cs_new_protected_nopar:Npn \tl_clear:N  #1
+  { \tl_set_eq:NN #1 \c_empty_tl }
+\cs_new_protected_nopar:Npn \tl_gclear:N #1
+  { \tl_gset_eq:NN #1 \c_empty_tl }
+\cs_generate_variant:Nn \tl_clear:N  { c }
+\cs_generate_variant:Nn \tl_gclear:N { c }
+\cs_new_protected_nopar:Npn \tl_clear_new:N  #1
+  { \cs_if_exist:NTF #1 { \tl_clear:N #1 } { \tl_new:N #1 } }
+\cs_new_protected_nopar:Npn \tl_gclear_new:N #1
+  { \cs_if_exist:NTF #1 { \tl_gclear:N #1 } { \tl_new:N #1 } }
+\cs_generate_variant:Nn \tl_clear_new:N  { c }
+\cs_generate_variant:Nn \tl_gclear_new:N { c }
+\cs_new_eq:NN \tl_set_eq:NN  \cs_set_eq:NN
+\cs_new_eq:NN \tl_set_eq:cN  \cs_set_eq:cN
+\cs_new_eq:NN \tl_set_eq:Nc  \cs_set_eq:Nc
+\cs_new_eq:NN \tl_set_eq:cc  \cs_set_eq:cc
+\cs_new_eq:NN \tl_gset_eq:NN \cs_gset_eq:NN
+\cs_new_eq:NN \tl_gset_eq:cN \cs_gset_eq:cN
+\cs_new_eq:NN \tl_gset_eq:Nc \cs_gset_eq:Nc
+\cs_new_eq:NN \tl_gset_eq:cc \cs_gset_eq:cc
+\cs_new_protected:Npn \tl_set:Nn #1#2
+  { \cs_set_nopar:Npx #1 { \exp_not:n {#2} } }
+\cs_new_protected:Npn \tl_set:No #1#2
+  { \cs_set_nopar:Npx #1 { \exp_not:o {#2} } }
+\cs_new_protected:Npn \tl_set:Nx #1#2
+  { \cs_set_nopar:Npx #1 {#2} }
+\cs_new_protected:Npn \tl_gset:Nn #1#2
+  { \cs_gset_nopar:Npx #1 { \exp_not:n {#2} } }
+\cs_new_protected:Npn \tl_gset:No #1#2
+  { \cs_gset_nopar:Npx #1 { \exp_not:o {#2} } }
+\cs_new_protected:Npn \tl_gset:Nx #1#2
+  { \cs_gset_nopar:Npx #1 {#2} }
+\cs_generate_variant:Nn \tl_set:Nn  {         NV , Nv , Nf }
+\cs_generate_variant:Nn \tl_set:Nx  { c }
+\cs_generate_variant:Nn \tl_set:Nn  { c, co , cV , cv , cf }
+\cs_generate_variant:Nn \tl_gset:Nn {         NV , Nv , Nf }
+\cs_generate_variant:Nn \tl_gset:Nx { c }
+\cs_generate_variant:Nn \tl_gset:Nn { c, co , cV , cv , cf }
+\cs_new_protected:Npn \tl_put_left:Nn #1#2
+  { \cs_set_nopar:Npx #1 { \exp_not:n {#2} \exp_not:o #1 } }
+\cs_new_protected:Npn \tl_put_left:NV #1#2
+  { \cs_set_nopar:Npx #1 { \exp_not:V #2 \exp_not:o #1 } }
+\cs_new_protected:Npn \tl_put_left:No #1#2
+  { \cs_set_nopar:Npx #1 { \exp_not:o {#2} \exp_not:o #1 } }
+\cs_new_protected:Npn \tl_put_left:Nx #1#2
+  { \cs_set_nopar:Npx #1 { #2 \exp_not:o #1 } }
+\cs_new_protected:Npn \tl_gput_left:Nn #1#2
+  { \cs_gset_nopar:Npx #1 { \exp_not:n {#2} \exp_not:o #1 } }
+\cs_new_protected:Npn \tl_gput_left:NV #1#2
+  { \cs_gset_nopar:Npx #1 { \exp_not:V #2 \exp_not:o #1 } }
+\cs_new_protected:Npn \tl_gput_left:No #1#2
+  { \cs_gset_nopar:Npx #1 { \exp_not:o {#2} \exp_not:o #1 } }
+\cs_new_protected:Npn \tl_gput_left:Nx #1#2
+  { \cs_gset_nopar:Npx #1 { #2 \exp_not:o {#1} } }
+\cs_generate_variant:Nn \tl_put_left:Nn  { c }
+\cs_generate_variant:Nn \tl_put_left:NV  { c }
+\cs_generate_variant:Nn \tl_put_left:No  { c }
+\cs_generate_variant:Nn \tl_put_left:Nx  { c }
+\cs_generate_variant:Nn \tl_gput_left:Nn { c }
+\cs_generate_variant:Nn \tl_gput_left:NV { c }
+\cs_generate_variant:Nn \tl_gput_left:No { c }
+\cs_generate_variant:Nn \tl_gput_left:Nx { c }
+\cs_new_protected:Npn \tl_put_right:Nn #1#2
+  { \cs_set_nopar:Npx #1 { \exp_not:o #1 \exp_not:n {#2} } }
+\cs_new_protected:Npn \tl_put_right:NV #1#2
+  { \cs_set_nopar:Npx #1 { \exp_not:o #1 \exp_not:V #2 } }
+\cs_new_protected:Npn \tl_put_right:No #1#2
+  { \cs_set_nopar:Npx #1 { \exp_not:o #1 \exp_not:o {#2} } }
+\cs_new_protected:Npn \tl_put_right:Nx #1#2
+  { \cs_set_nopar:Npx #1 { \exp_not:o #1 #2 } }
+\cs_new_protected:Npn \tl_gput_right:Nn #1#2
+  { \cs_gset_nopar:Npx #1 { \exp_not:o #1 \exp_not:n {#2} } }
+\cs_new_protected:Npn \tl_gput_right:NV #1#2
+  { \cs_gset_nopar:Npx #1 { \exp_not:o #1 \exp_not:V #2 } }
+\cs_new_protected:Npn \tl_gput_right:No #1#2
+  { \cs_gset_nopar:Npx #1 { \exp_not:o #1 \exp_not:o {#2} } }
+\cs_new_protected:Npn \tl_gput_right:Nx #1#2
+  { \cs_gset_nopar:Npx #1 { \exp_not:o {#1} #2 } }
+\cs_generate_variant:Nn \tl_put_right:Nn  { c }
+\cs_generate_variant:Nn \tl_put_right:NV  { c }
+\cs_generate_variant:Nn \tl_put_right:No  { c }
+\cs_generate_variant:Nn \tl_put_right:Nx  { c }
+\cs_generate_variant:Nn \tl_gput_right:Nn { c }
+\cs_generate_variant:Nn \tl_gput_right:NV { c }
+\cs_generate_variant:Nn \tl_gput_right:No { c }
+\cs_generate_variant:Nn \tl_gput_right:Nx { c }
+\group_begin:
+  \tex_lccode:D `\A = `\@ \scan_stop:
+  \tex_lccode:D `\B = `\@ \scan_stop:
+  \tex_catcode:D `\A = 8 \scan_stop:
+  \tex_catcode:D `\B = 3 \scan_stop:
+\tex_lowercase:D
+  {
+    \group_end:
+    \tl_const:Nn \c_tl_rescan_marker_tl { A B }
+  }
+\tl_new:N \l_tl_rescan_tl
+\cs_new_protected_nopar:Npn \tl_set_rescan:Nnn
+  { \tl_set_rescan_aux:NNnn \tl_set:Nn }
+\cs_new_protected_nopar:Npn \tl_gset_rescan:Nnn
+  { \tl_set_rescan_aux:NNnn \tl_gset:Nn }
+\cs_new_protected:Npn \tl_set_rescan_aux:NNnn #1#2#3#4
+  {
+    \group_begin:
+      \exp_args:No \etex_everyeof:D { \c_tl_rescan_marker_tl }
+      \tex_endlinechar:D \c_minus_one
+      #3
+      \tl_clear:N \l_tl_rescan_tl
+      \exp_after:wN \tl_rescan_aux:w \etex_scantokens:D {#4}
+    \exp_args:NNNo \group_end:
+    #1 #2 \l_tl_rescan_tl
+  }
+\cs_new_nopar:Npx \tl_rescan_aux:w
+  {
+    \cs_set_protected:Npn \exp_not:N \tl_rescan_aux:w ##1
+      \c_tl_rescan_marker_tl
+      { \tl_set:Nn \exp_not:N \l_tl_rescan_tl {##1} }
+  }
+\tl_rescan_aux:w
+\cs_generate_variant:Nn \tl_set_rescan:Nnn  {     Nno }
+\cs_generate_variant:Nn \tl_set_rescan:Nnn  { c , cno }
+\cs_generate_variant:Nn \tl_gset_rescan:Nnn {     Nno }
+\cs_generate_variant:Nn \tl_gset_rescan:Nnn { c , cno }
+\cs_new_protected_nopar:Npn \tl_set_rescan:Nnx
+  { \tl_set_rescan_aux:NNnx \tl_set:Nn }
+\cs_new_protected_nopar:Npn \tl_gset_rescan:Nnx
+  { \tl_set_rescan_aux:NNnx \tl_gset:Nn }
+\cs_new_protected_nopar:Npn \tl_set_rescan_aux:NNnx #1#2#3#4
+  {
+    \group_begin:
+      \etex_everyeof:D { \exp_not:N }
+      \tex_endlinechar:D \c_minus_one
+      #3
+      \tl_set:Nx \l_tl_rescan_tl { \etex_scantokens:D {#4} }
+    \exp_args:NNNo \group_end:
+    #1 #2 \l_tl_rescan_tl
+  }
+\cs_generate_variant:Nn \tl_set_rescan:Nnx  { c }
+\cs_generate_variant:Nn \tl_gset_rescan:Nnx { c }
+\cs_new_protected:Npn \tl_rescan:nn #1#2
+  {
+    \group_begin:
+      \exp_args:No \etex_everyeof:D { \c_tl_rescan_marker_tl }
+      \tex_endlinechar:D \c_minus_one
+      #1
+      \exp_after:wN \tl_rescan_aux:w \etex_scantokens:D {#2}
+    \exp_args:No \group_end:
+    \l_tl_rescan_tl
+  }
+\cs_new_eq:NN \tl_to_lowercase:n \tex_lowercase:D
+\cs_new_eq:NN \tl_to_uppercase:n \tex_uppercase:D
+\tl_new:N \l_tl_replace_tl
+\cs_new_protected_nopar:Npn \tl_replace_once:Nnn
+  { \tl_replace_once_aux:NNnn \tl_set_eq:NN }
+\cs_new_protected_nopar:Npn \tl_greplace_once:Nnn
+  { \tl_replace_once_aux:NNnn \tl_gset_eq:NN }
+\cs_new_protected:Npn \tl_replace_once_aux:NNnn #1#2#3#4
+  {
+    \cs_set_protected:Npx \tl_tmp:w ##1 #3 ##2 \q_stop
+      {
+        \exp_not:N \quark_if_no_value:nF {##2}
+          {
+            \tl_set:No \exp_not:N \l_tl_replace_tl { ##1 \exp_not:n{#4} }
+            \exp_not:n
+              {
+                \cs_set_protected:Npn \tl_tmp:w ##1 \q_nil #3 \q_no_value
+                  { \tl_put_right:No \l_tl_replace_tl {##1} }
+              }
+            \exp_not:n { \tl_tmp:w \prg_do_nothing: } ##2
+            \exp_not:n { #1 #2 \l_tl_replace_tl }
+          }
+      }
+    \exp_after:wN \tl_tmp:w \exp_after:wN \prg_do_nothing:
+    #2 \q_nil #3 \q_no_value \q_stop
+  }
+\cs_generate_variant:Nn \tl_replace_once:Nnn  { c }
+\cs_generate_variant:Nn \tl_greplace_once:Nnn { c }
+\cs_new_protected_nopar:Npn \tl_replace_all:Nnn
+  { \tl_replace_all_aux:NNnn \tl_set_eq:NN }
+\cs_new_protected_nopar:Npn \tl_greplace_all:Nnn
+  { \tl_replace_all_aux:NNnn \tl_gset_eq:NN }
+\cs_new_protected:Npn \tl_replace_all_aux:NNnn #1#2#3#4
+   {
+     \tl_clear:N \l_tl_replace_tl
+     \cs_set_protected:Npx \tl_tmp:w ##1 #3 ##2 \q_stop
+       {
+         \exp_not:N \quark_if_no_value:nTF {##2}
+           {
+             \exp_not:n
+               {
+                 \cs_set_protected:Npn \tl_tmp:w ##1 \q_nil ##2 \q_stop
+                   { \tl_put_right:No \l_tl_replace_tl {##1} }
+               }
+             \exp_not:N \tl_tmp:w ##1 \exp_not:N \q_stop
+           }
+           {
+             \exp_not:n { \tl_put_right:No \l_tl_replace_tl }
+               { ##1 \exp_not:n{#4} }
+             \exp_not:n { \tl_tmp:w \prg_do_nothing: } ##2 \exp_not:N \q_stop
+           }
+       }
+     \exp_after:wN \tl_tmp:w \exp_after:wN \prg_do_nothing:
+     #2 \q_nil #3 \q_no_value \q_stop
+     #1 #2 \l_tl_replace_tl
+   }
+\cs_generate_variant:Nn \tl_replace_all:Nnn  { c }
+\cs_generate_variant:Nn \tl_greplace_all:Nnn { c }
+\cs_new_protected_nopar:Npn \tl_remove_once:Nn #1#2
+  { \tl_replace_once:Nnn #1 {#2} { } }
+\cs_new_protected_nopar:Npn \tl_gremove_once:Nn #1#2
+  { \tl_greplace_once:Nnn #1 {#2} { } }
+\cs_generate_variant:Nn \tl_remove_once:Nn  { c }
+\cs_generate_variant:Nn \tl_gremove_once:Nn { c }
+\cs_new_protected_nopar:Npn \tl_remove_all:Nn #1#2
+  { \tl_replace_all:Nnn #1 {#2} { } }
+\cs_new_protected_nopar:Npn \tl_gremove_all:Nn #1#2
+  { \tl_greplace_all:Nnn #1 {#2} { } }
+\cs_generate_variant:Nn \tl_remove_all:Nn  { c }
+\cs_generate_variant:Nn \tl_gremove_all:Nn { c }
+\prg_new_conditional:Npnn \tl_if_blank:n #1 { p , T , F , TF }
+  { \tl_if_empty_return:o { \use_none:n #1 ? } }
+\cs_generate_variant:Nn \tl_if_blank_p:n { V }
+\cs_generate_variant:Nn \tl_if_blank:nT  { V }
+\cs_generate_variant:Nn \tl_if_blank:nF  { V }
+\cs_generate_variant:Nn \tl_if_blank:nTF { V }
+\cs_generate_variant:Nn \tl_if_blank_p:n { o }
+\cs_generate_variant:Nn \tl_if_blank:nT  { o }
+\cs_generate_variant:Nn \tl_if_blank:nF  { o }
+\cs_generate_variant:Nn \tl_if_blank:nTF { o }
+\prg_set_conditional:Npnn \tl_if_empty:N #1 { p , T , F , TF }
+  {
+    \if_meaning:w #1 \c_empty_tl
+      \prg_return_true:
+    \else:
+      \prg_return_false:
+    \fi:
+  }
+\cs_generate_variant:Nn \tl_if_empty_p:N { c }
+\cs_generate_variant:Nn \tl_if_empty:NT  { c }
+\cs_generate_variant:Nn \tl_if_empty:NF  { c }
+\cs_generate_variant:Nn \tl_if_empty:NTF { c }
+\prg_new_conditional:Npnn \tl_if_empty:n #1 { p,TF,T,F} {
+  \exp_after:wN \if_meaning:w \exp_after:wN \q_nil \tl_to_str:n {#1} \q_nil
+    \prg_return_true:
+  \else:
+    \prg_return_false:
+  \fi:
+}
+\cs_generate_variant:Nn \tl_if_empty_p:n { V }
+\cs_generate_variant:Nn \tl_if_empty:nTF { V }
+\cs_generate_variant:Nn \tl_if_empty:nT  { V }
+\cs_generate_variant:Nn \tl_if_empty:nF  { V }
+\cs_new:Npn \tl_if_empty_return:o #1 {
+  \exp_after:wN \if_meaning:w \exp_after:wN \q_nil
+          \tl_to_str:n \exp_after:wN {#1} \q_nil
+    \prg_return_true:
+  \else:
+    \prg_return_false:
+  \fi:
+}
+\prg_new_conditional:Npnn \tl_if_empty:o #1 { p,TF,T,F}
+  { \tl_if_empty_return:o {#1} }
+\prg_new_conditional:Npnn \tl_if_eq:NN #1#2 { p , T , F , TF }
+  {
+    \if_meaning:w #1 #2
+      \prg_return_true:
+    \else:
+      \prg_return_false:
+    \fi:
+  }
+\cs_generate_variant:Nn \tl_if_eq_p:NN { Nc , c , cc }
+\cs_generate_variant:Nn \tl_if_eq:NNTF { Nc , c , cc }
+\cs_generate_variant:Nn \tl_if_eq:NNT  { Nc , c , cc }
+\cs_generate_variant:Nn \tl_if_eq:NNF  { Nc , c , cc }
+\prg_new_protected_conditional:Npnn \tl_if_eq:nn #1#2 { T , F ,  TF }
+  {
+    \group_begin:
+      \tl_set:Nn \l_tl_tmpa_tl {#1}
+      \tl_set:Nn \l_tl_tmpb_tl {#2}
+      \if_meaning:w \l_tl_tmpa_tl \l_tl_tmpb_tl
+        \group_end:
+        \prg_return_true:
+      \else:
+        \group_end:
+        \prg_return_false:
+      \fi:
+  }
+\tl_new:N \l_tl_tmpa_tl
+\tl_new:N \l_tl_tmpb_tl
+\cs_new_protected_nopar:Npn \tl_if_in:NnT  { \exp_args:No \tl_if_in:nnT  }
+\cs_new_protected_nopar:Npn \tl_if_in:NnF  { \exp_args:No \tl_if_in:nnF  }
+\cs_new_protected_nopar:Npn \tl_if_in:NnTF { \exp_args:No \tl_if_in:nnTF }
+\cs_generate_variant:Nn \tl_if_in:NnT { c }
+\cs_generate_variant:Nn \tl_if_in:NnF  { c }
+\cs_generate_variant:Nn \tl_if_in:NnTF { c }
+\prg_new_protected_conditional:Npnn \tl_if_in:nn #1#2 { T  , F , TF }
+  {
+    \cs_set:Npn \tl_tmp:w ##1 #2 { }
+    \tl_if_empty:oTF { \tl_tmp:w #1 {} {} #2 }
+      { \prg_return_false: } { \prg_return_true: }
+  }
+\cs_generate_variant:Nn \tl_if_in:nnT  { V }
+\cs_generate_variant:Nn \tl_if_in:nnF  { V }
+\cs_generate_variant:Nn \tl_if_in:nnTF { V }
+\cs_generate_variant:Nn \tl_if_in:nnT  { o }
+\cs_generate_variant:Nn \tl_if_in:nnF  { o }
+\cs_generate_variant:Nn \tl_if_in:nnTF { o }
+\cs_generate_variant:Nn \tl_if_in:nnT  { no }
+\cs_generate_variant:Nn \tl_if_in:nnF  { no }
+\cs_generate_variant:Nn \tl_if_in:nnTF { no }
+\prg_new_conditional:Npnn \tl_if_single:n #1 { p , T , F , TF }
+  {
+    \tl_if_blank:nTF {#1}
+      { \tl_if_empty:nTF {#1} { \prg_return_false: } { \prg_return_true: } }
+      { \tl_if_single_aux:n {#1} }
+  }
+\prg_new_conditional:Npnn \tl_if_single:N #1 { p , T , F , TF }
+  {
+    \tl_if_blank:oTF #1
+      { \tl_if_empty:NTF #1 { \prg_return_false: } { \prg_return_true: } }
+      { \exp_args:No \tl_if_single_aux:n {#1} }
+  }
+\cs_new:Npn \tl_if_single_aux:n #1
+  { \tl_if_empty_return:o { \use_none:n #1 } }
+\cs_new:Npn \tl_map_function:nN #1#2
+  { \tl_map_function_aux:Nn #2 #1 \q_recursion_tail \q_recursion_stop }
+\cs_new_nopar:Npn \tl_map_function:NN #1#2
+  {
+    \exp_after:wN \tl_map_function_aux:Nn
+    \exp_after:wN #2 #1 \q_recursion_tail \q_recursion_stop
+  }
+\cs_new:Npn \tl_map_function_aux:Nn #1#2
+  {
+    \quark_if_recursion_tail_stop:n {#2}
+    #1 {#2} \tl_map_function_aux:Nn #1
+  }
+\cs_generate_variant:Nn \tl_map_function:NN { c }
+\cs_new_protected:Npn \tl_map_inline:nn #1#2
+  {
+    \int_gincr:N \g_tl_inline_level_int
+    \cs_gset:cpn { tl_map_inline_ \int_use:N \g_tl_inline_level_int :n }
+      ##1 {#2}
+    \exp_args:Nc \tl_map_function_aux:Nn
+      { tl_map_inline_ \int_use:N \g_tl_inline_level_int :n }
+      #1 \q_recursion_tail \q_recursion_stop
+    \int_gdecr:N \g_tl_inline_level_int
+  }
+\cs_new_protected:Npn \tl_map_inline:Nn #1#2
+  {
+    \int_gincr:N \g_tl_inline_level_int
+    \cs_gset:cpn { tl_map_inline_ \int_use:N \g_tl_inline_level_int :n }
+      ##1 {#2}
+    \exp_last_unbraced:NcV \tl_map_function_aux:Nn
+      { tl_map_inline_ \int_use:N \g_tl_inline_level_int :n }
+      #1 \q_recursion_tail\q_recursion_stop
+    \int_gdecr:N \g_tl_inline_level_int
+  }
+\cs_generate_variant:Nn \tl_map_inline:Nn { c }
+\cs_new_protected:Npn \tl_map_variable:nNn #1#2#3
+  { \tl_map_variable_aux:Nnn #2 {#3} #1 \q_recursion_tail \q_recursion_stop }
+\cs_new_protected_nopar:Npn \tl_map_variable:NNn
+  { \exp_args:No \tl_map_variable:nNn }
+\cs_new_protected:Npn \tl_map_variable_aux:Nnn #1#2#3
+  {
+    \tl_set:Nn #1 {#3}
+    \quark_if_recursion_tail_stop:N #1
+    #2 \tl_map_variable_aux:Nnn #1 {#2}
+  }
+\cs_generate_variant:Nn \tl_map_variable:NNn { c }
+\cs_new_eq:NN \tl_map_break: \use_none_delimit_by_q_recursion_stop:w
+\cs_new_eq:NN \tl_to_str:n \etex_detokenize:D
+\cs_new_nopar:Npn \tl_to_str:N #1 { \etex_detokenize:D \exp_after:wN {#1} }
+\cs_generate_variant:Nn \tl_to_str:N { c }
+\cs_new_eq:NN \tl_use:N \prg_do_nothing:
+\cs_new_nopar:Npn \tl_use:c #1
+  {
+    \if_cs_exist:w #1 \cs_end:
+      \cs:w #1 \exp_after:wN \cs_end:
+    \else:
+      \exp_after:wN \tl_error_message:
+    \fi:
+  }
+\group_begin:
+\tex_catcode:D `\! = 11 \scan_stop:
+\tex_catcode:D `\ = 11\scan_stop:%
+\cs_gset_nopar:Npn\tl_error_message:{\undefined variable name!}%
+\group_end:%
+\cs_new:Npn \tl_length:n #1
+  {
+    \int_eval:n
+      { 0 \tl_map_function:nN {#1} \tl_length_aux:n }
+  }
+\cs_new_nopar:Npn \tl_length:N #1
+  {
+    \int_eval:n
+      { 0 \tl_map_function:NN #1 \tl_length_aux:n }
+  }
+\cs_new:Npn \tl_length_aux:n #1 { + 1 }
+\cs_generate_variant:Nn \tl_length:n { V , o }
+\cs_generate_variant:Nn \tl_length:N { c }
+\cs_new:Npn \tl_reverse:n #1
+  { \tl_reverse_aux:nN { } #1 \q_recursion_tail \q_recursion_stop }
+\cs_new:Npn \tl_reverse_aux:nN #1#2
+  {
+    \quark_if_recursion_tail_stop_do:nn {#2} {#1}
+    \tl_reverse_aux:nN { #2 #1 }
+  }
+\cs_generate_variant:Nn \tl_reverse:n {V,o}
+\cs_new_protected_nopar:Npn \tl_reverse:N #1
+  { \tl_set:Nf #1 { \tl_reverse:o { #1 \exp_stop_f: } } }
+\cs_generate_variant:Nn \tl_reverse:N { c }
+\cs_new:Npn \tl_trim_spaces:n #1
+  {
+    \etex_unexpanded:D \exp_after:wN \exp_after:wN \exp_after:wN
+      { \tl_trim_spaces_exp:n {#1} }
+  }
+\cs_new_protected:Npn \tl_trim_spaces:N #1
+  { \tl_set:Nx #1 { \exp_after:wN \tl_trim_spaces:n \exp_after:wN {#1} } }
+\cs_new_protected:Npn \tl_gtrim_spaces:N #1
+  { \tl_gset:Nx #1 { \exp_after:wN \tl_trim_spaces:n \exp_after:wN {#1} } }
+\group_begin:
+  \tex_catcode:D `\Q = 3 \scan_stop:
+  \cs_new:Npn \tl_trim_spaces_exp:n #1
+    { \tex_romannumeral:D - `0 \tl_trim_spaces_aux_i:w \exp_not:N #1 Q ~ Q }
+  \cs_new:Npn \tl_trim_spaces_aux_i:w #1 ~ Q { \tl_trim_spaces_aux_ii:w #1 Q }
+  \cs_new:Npn \tl_trim_spaces_aux_ii:w #1 Q #2 {#1}
+\group_end:
+\cs_generate_variant:Nn \tl_trim_spaces:N  { c }
+\cs_generate_variant:Nn \tl_gtrim_spaces:N { c }
+\cs_new:Npn \tl_head:n #1 { \tl_head:w #1 \q_stop }
+\cs_new:Npn \tl_tail:n #1 { \tl_tail:w #1 \q_stop }
+\cs_new:Npn \tl_head:w #1#2 \q_stop {#1}
+\cs_new:Npn \tl_tail:w #1#2 \q_stop {#2}
+\cs_generate_variant:Nn \tl_head:n { V , v , f }
+\cs_generate_variant:Nn \tl_tail:n { V , v , f }
+\prg_new_conditional:Npnn \tl_if_head_eq_meaning:nN #1#2 { p , T , F , TF }
+  {
+    \exp_after:wN \if_meaning:w \tl_head:w #1 \q_stop #2
+      \prg_return_true:
+    \else:
+      \prg_return_false:
+    \fi:
+  }
+\prg_new_conditional:Npnn \tl_if_head_eq_charcode:nN #1#2 { p , T , F , TF }
+  {
+    \exp_after:wN \if:w
+      \exp_after:wN \exp_not:N \tl_head:w #1 \q_stop \exp_not:N #2
+      \prg_return_true:
+    \else:
+      \prg_return_false:
+    \fi:
+  }
+\cs_generate_variant:Nn \tl_if_head_eq_charcode_p:nN { f }
+\cs_generate_variant:Nn \tl_if_head_eq_charcode:nNTF { f }
+\cs_generate_variant:Nn \tl_if_head_eq_charcode:nNT  { f }
+\cs_generate_variant:Nn \tl_if_head_eq_charcode:nNF  { f }
+\prg_new_conditional:Npnn \tl_if_head_eq_catcode:nN #1#2 { p , T , F , TF }
+  {
+    \exp_after:wN \if_catcode:w \exp_after:wN \exp_not:N
+        \tl_head:w #1 \q_stop \exp_not:N #2
+      \prg_return_true: \else: \prg_return_false: \fi:
+  }
+\cs_new_protected:Npn \tl_show:N #1 { \cs_show:N #1 }
+\cs_generate_variant:Nn \tl_show:N { c }
+\cs_new_eq:NN \tl_show:n \etex_showtokens:D
+\tl_const:Nx \c_job_name_tl { \tex_jobname:D }
+\tl_const:Nn \c_empty_tl { }
+\tl_const:Nn \c_space_tl { ~ }
+\tl_new:N \g_tmpa_tl
+\tl_new:N \g_tmpb_tl
+\tl_new:N \l_tmpa_tl
+\tl_new:N \l_tmpb_tl
+\cs_new:Npn \tl_if_single_item:nTF #1
+  { \str_if_eq:onTF { \use_ii:nn #1 {?} ? } {??} }
+\cs_new:Npn \tl_if_single_item:nT #1
+  { \str_if_eq:onT { \use_ii:nn #1 {?} ? } {??} }
+\cs_new:Npn \tl_if_single_item:nF #1
+  { \str_if_eq:onF { \use_ii:nn #1 {?} ? } {??} }
+\cs_new:Npn \tl_if_single_item_p:n #1
+  { \str_if_eq_p:on { \use_ii:nn #1 {?} ? } {??} }
+\prg_new_conditional:Npnn \tl_if_head_begin_group:n #1 { p , T , F , TF }
+  {
+    \exp_after:wN \use_none:n
+      \exp_after:wN {
+        \exp_after:wN {
+          \token_to_str:N #1 .
+        }
+        \prg_return_true: \exp_after:wN \use_none:nn \token_to_str:N
+      }
+      \prg_return_false:
+  }
+\prg_new_conditional:Npnn \tl_if_head_eq_space:n #1 { p , T , F , TF }
+  {
+    \if_false: { \fi:
+      \tl_if_head_eq_space_aux:w \prg_do_nothing: #1 ? ~ }
+  }
+\cs_new:Npn \tl_if_head_eq_space_aux:w #1 ~ %
+  {
+    \tl_if_empty_return:o { #1 }
+    \exp_after:wN \use_none:n \exp_after:wN { \if_false: } \fi:
+  }
+\prg_new_conditional:Npnn \tl_if_single_token:n #1 { p , T , F , TF }
+  {
+    \tl_if_head_eq_space:nTF {#1}
+      {
+        \str_if_eq:nnTF {#1} {~}
+          \prg_return_true: \prg_return_false:
+      }
+      {
+        \str_if_eq:onTF { \use_none:n #1 ? } {?}
+          {
+            \tl_if_head_begin_group:nTF {#1}
+              \prg_return_false: \prg_return_true:
+          }
+          \prg_return_false:
+      }
+  }
+\cs_new_protected:Npn \tl_new:Nn #1#2
+  {
+    \tl_new:N #1
+    \tl_gset:Nn #1 {#2}
+  }
+\cs_generate_variant:Nn \tl_new:Nn { c }
+\cs_generate_variant:Nn \tl_new:Nn { Nx }
+\cs_new_protected_nopar:Npn \tl_gset:Nc
+  { \pref_global:D \tl_set:Nc }
+\cs_new_protected_nopar:Npn \tl_set:Nc #1#2
+  { \tl_set:No #1 { \cs:w #2 \cs_end: } }
+\cs_new_eq:NN \tl_replace_in:Nnn  \tl_replace_once:Nnn
+\cs_new_eq:NN \tl_replace_in:cnn  \tl_replace_once:cnn
+\cs_new_eq:NN \tl_greplace_in:Nnn \tl_greplace_once:Nnn
+\cs_new_eq:NN \tl_greplace_in:cnn \tl_greplace_once:cnn
+\cs_new_eq:NN \tl_replace_all_in:Nnn  \tl_replace_all:Nnn
+\cs_new_eq:NN \tl_replace_all_in:cnn  \tl_replace_all:cnn
+\cs_new_eq:NN \tl_greplace_all_in:Nnn \tl_greplace_all:Nnn
+\cs_new_eq:NN \tl_greplace_all_in:cnn \tl_greplace_all:cnn
+\cs_new_eq:NN \tl_remove_in:Nn  \tl_remove_once:Nn
+\cs_new_eq:NN \tl_remove_in:cn  \tl_remove_once:cn
+\cs_new_eq:NN \tl_gremove_in:Nn \tl_gremove_once:Nn
+\cs_new_eq:NN \tl_gremove_in:cn \tl_gremove_once:cn
+\cs_new_eq:NN \tl_remove_all_in:Nn  \tl_remove_all:Nn
+\cs_new_eq:NN \tl_remove_all_in:cn  \tl_remove_all:cn
+\cs_new_eq:NN \tl_gremove_all_in:Nn \tl_gremove_all:Nn
+\cs_new_eq:NN \tl_gremove_all_in:cn \tl_gremove_all:cn
+\cs_new_eq:NN \tl_elt_count:n \tl_length:n
+\cs_new_eq:NN \tl_elt_count:V \tl_length:V
+\cs_new_eq:NN \tl_elt_count:o \tl_length:o
+\cs_new_eq:NN \tl_elt_count:N \tl_length:N
+\cs_new_eq:NN \tl_elt_count:c \tl_length:c
+\cs_new_eq:NN \tl_head_i:n \tl_head:n
+\cs_new_eq:NN \tl_head_i:w \tl_head:w
+\cs_new:Npn \tl_head_iii:n #1 { \tl_head_iii:w #1 \q_stop }
+\cs_generate_variant:Nn \tl_head_iii:n { f }
+\cs_new:Npn \tl_head_iii:w #1#2#3#4 \q_stop {#1#2#3}
+%% 
+%%
+%% End of file `l3tl.sty'.
diff --git a/Master/texmf-dist/tex/latex/l3kernel/l3token.sty b/Master/texmf-dist/tex/latex/l3kernel/l3token.sty
new file mode 100644
index 00000000000..c4ca0549b2f
--- /dev/null
+++ b/Master/texmf-dist/tex/latex/l3kernel/l3token.sty
@@ -0,0 +1,739 @@
+%%
+%% This is file `l3token.sty',
+%% generated with the docstrip utility.
+%%
+%% The original source files were:
+%%
+%% l3token.dtx  (with options: `package')
+%% 
+%% EXPERIMENTAL CODE
+%% 
+%% Do not distribute this file without also distributing the
+%% source files specified above.
+%% 
+%% Do not distribute a modified version of this file.
+%% 
+%% File: l3token.dtx Copyright (C) 2005-2011 The LaTeX3 Project
+%%
+%% It may be distributed and/or modified under the conditions of the
+%% LaTeX Project Public License (LPPL), either version 1.3c of this
+%% license or (at your option) any later version.  The latest version
+%% of this license is in the file
+%%
+%%    http://www.latex-project.org/lppl.txt
+%%
+%% This file is part of the "expl3 bundle" (The Work in LPPL)
+%% and all files in that bundle must be distributed together.
+%%
+%% The released version of this bundle is available from CTAN.
+%%
+%% -----------------------------------------------------------------------
+%%
+%% The development version of the bundle can be found at
+%%
+%%    http://www.latex-project.org/svnroot/experimental/trunk/
+%%
+%% for those people who are interested.
+%%
+%%%%%%%%%%%
+%% NOTE: %%
+%%%%%%%%%%%
+%%
+%%   Snapshots taken from the repository represent work in progress and may
+%%   not work or may contain conflicting material!  We therefore ask
+%%   people _not_ to put them into distributions, archives, etc. without
+%%   prior consultation with the LaTeX3 Project.
+%%
+%% -----------------------------------------------------------------------
+\RequirePackage{l3names}
+\GetIdInfo$Id: l3token.dtx 2491 2011-07-02 11:49:37Z joseph $
+  {L3 Experimental token manipulation}
+%% \begin{function}[EXP,pTF]{\token_if_protected_long_macro:N}
+\ProvidesExplPackage
+  {\ExplFileName}{\ExplFileDate}{\ExplFileVersion}{\ExplFileDescription}
+\package_check_loaded_expl:
+\cs_new_protected_nopar:Npn \char_set_catcode:nn #1#2
+  { \tex_catcode:D #1 = \int_eval:w #2 \int_eval_end: }
+\cs_new_nopar:Npn \char_value_catcode:n #1
+  { \tex_the:D \tex_catcode:D \int_eval:w #1\int_eval_end: }
+\cs_new_nopar:Npn \char_show_value_catcode:n #1
+  { \tex_showthe:D \tex_catcode:D \int_eval:w #1 \int_eval_end: }
+\cs_new_protected_nopar:Npn \char_set_catcode_escape:N #1
+  { \char_set_catcode:nn { `#1 } \c_zero }
+\cs_new_protected_nopar:Npn \char_set_catcode_group_begin:N #1
+  { \char_set_catcode:nn { `#1 } \c_one }
+\cs_new_protected_nopar:Npn \char_set_catcode_group_end:N #1
+  { \char_set_catcode:nn { `#1 } \c_two }
+\cs_new_protected_nopar:Npn \char_set_catcode_math_toggle:N #1
+  { \char_set_catcode:nn { `#1 } \c_three }
+\cs_new_protected_nopar:Npn \char_set_catcode_alignment:N #1
+  { \char_set_catcode:nn { `#1 } \c_four }
+\cs_new_protected_nopar:Npn \char_set_catcode_end_line:N #1
+  { \char_set_catcode:nn { `#1 } \c_five }
+\cs_new_protected_nopar:Npn \char_set_catcode_parameter:N #1
+  { \char_set_catcode:nn { `#1 } \c_six }
+\cs_new_protected_nopar:Npn \char_set_catcode_math_superscript:N #1
+  { \char_set_catcode:nn { `#1 } \c_seven }
+\cs_new_protected_nopar:Npn \char_set_catcode_math_subscript:N #1
+  { \char_set_catcode:nn { `#1 } \c_eight }
+\cs_new_protected_nopar:Npn \char_set_catcode_ignore:N #1
+  { \char_set_catcode:nn { `#1 } \c_nine }
+\cs_new_protected_nopar:Npn \char_set_catcode_space:N #1
+  { \char_set_catcode:nn { `#1 } \c_ten }
+\cs_new_protected_nopar:Npn \char_set_catcode_letter:N #1
+  { \char_set_catcode:nn { `#1 } \c_eleven }
+\cs_new_protected_nopar:Npn \char_set_catcode_other:N #1
+  { \char_set_catcode:nn { `#1 } \c_twelve }
+\cs_new_protected_nopar:Npn \char_set_catcode_active:N #1
+  { \char_set_catcode:nn { `#1 } \c_thirteen }
+\cs_new_protected_nopar:Npn \char_set_catcode_comment:N #1
+  { \char_set_catcode:nn { `#1 } \c_fourteen }
+\cs_new_protected_nopar:Npn \char_set_catcode_invalid:N #1
+  { \char_set_catcode:nn { `#1 } \c_fifteen }
+\cs_new_protected_nopar:Npn \char_set_catcode_escape:n #1
+  { \char_set_catcode:nn {#1} \c_zero }
+\cs_new_protected_nopar:Npn \char_set_catcode_group_begin:n #1
+  { \char_set_catcode:nn {#1} \c_one }
+\cs_new_protected_nopar:Npn \char_set_catcode_group_end:n #1
+  { \char_set_catcode:nn {#1} \c_two }
+\cs_new_protected_nopar:Npn \char_set_catcode_math_toggle:n #1
+  { \char_set_catcode:nn {#1} \c_three }
+\cs_new_protected_nopar:Npn \char_set_catcode_alignment:n #1
+  { \char_set_catcode:nn {#1} \c_four }
+\cs_new_protected_nopar:Npn \char_set_catcode_end_line:n #1
+  { \char_set_catcode:nn {#1} \c_five }
+\cs_new_protected_nopar:Npn \char_set_catcode_parameter:n #1
+  { \char_set_catcode:nn {#1} \c_six }
+\cs_new_protected_nopar:Npn \char_set_catcode_math_superscript:n #1
+  { \char_set_catcode:nn {#1} \c_seven }
+\cs_new_protected_nopar:Npn \char_set_catcode_math_subscript:n #1
+  { \char_set_catcode:nn {#1} \c_eight }
+\cs_new_protected_nopar:Npn \char_set_catcode_ignore:n #1
+  { \char_set_catcode:nn {#1} \c_nine }
+\cs_new_protected_nopar:Npn \char_set_catcode_space:n #1
+  { \char_set_catcode:nn {#1} \c_ten }
+\cs_new_protected_nopar:Npn \char_set_catcode_letter:n #1
+  { \char_set_catcode:nn {#1} \c_eleven }
+\cs_new_protected_nopar:Npn \char_set_catcode_other:n #1
+  { \char_set_catcode:nn {#1} \c_twelve }
+\cs_new_protected_nopar:Npn \char_set_catcode_active:n #1
+  { \char_set_catcode:nn {#1} \c_thirteen }
+\cs_new_protected_nopar:Npn \char_set_catcode_comment:n #1
+  { \char_set_catcode:nn {#1} \c_fourteen }
+\cs_new_protected_nopar:Npn \char_set_catcode_invalid:n #1
+  { \char_set_catcode:nn {#1} \c_fifteen }
+\cs_new_protected_nopar:Npn \char_set_mathcode:nn #1#2
+  { \tex_mathcode:D #1 = \int_eval:w #2 \int_eval_end: }
+\cs_new_nopar:Npn \char_value_mathcode:n #1
+  { \tex_the:D \tex_mathcode:D \int_eval:w #1\int_eval_end: }
+\cs_new_nopar:Npn \char_show_value_mathcode:n #1
+  { \tex_showthe:D \tex_mathcode:D \int_eval:w #1 \int_eval_end: }
+\cs_new_protected_nopar:Npn \char_set_lccode:nn #1#2
+  { \tex_lccode:D #1 = \int_eval:w #2 \int_eval_end: }
+\cs_new_nopar:Npn \char_value_lccode:n #1
+  { \tex_the:D \tex_lccode:D \int_eval:w #1\int_eval_end: }
+\cs_new_nopar:Npn \char_show_value_lccode:n #1
+  { \tex_showthe:D \tex_lccode:D \int_eval:w #1 \int_eval_end: }
+\cs_new_protected_nopar:Npn \char_set_uccode:nn #1#2
+  { \tex_uccode:D #1 = \int_eval:w #2 \int_eval_end: }
+\cs_new_nopar:Npn \char_value_uccode:n #1
+  { \tex_the:D \tex_uccode:D \int_eval:w #1\int_eval_end: }
+\cs_new_nopar:Npn \char_show_value_uccode:n #1
+  { \tex_showthe:D \tex_uccode:D \int_eval:w #1 \int_eval_end: }
+\cs_new_protected_nopar:Npn \char_set_sfcode:nn #1#2
+  { \tex_sfcode:D #1 = \int_eval:w #2 \int_eval_end: }
+\cs_new_nopar:Npn \char_value_sfcode:n #1
+  { \tex_the:D \tex_sfcode:D \int_eval:w #1\int_eval_end: }
+\cs_new_nopar:Npn \char_show_value_sfcode:n #1
+  { \tex_showthe:D \tex_sfcode:D \int_eval:w #1 \int_eval_end: }
+\cs_new_protected_nopar:Npn \token_new:Nn #1#2 { \cs_new_eq:NN #1 #2 }
+\cs_new_eq:NN \c_group_begin_token {
+\cs_new_eq:NN \c_group_end_token }
+\group_begin:
+  \char_set_catcode_math_toggle:N \*
+  \token_new:Nn \c_math_toggle_token { * }
+  \char_set_catcode_alignment:N \*
+  \token_new:Nn \c_alignment_token { * }
+  \token_new:Nn \c_parameter_token { # }
+  \token_new:Nn \c_math_superscript_token { ^ }
+  \char_set_catcode_math_subscript:N \*
+  \token_new:Nn \c_math_subscript_token { * }
+  \token_new:Nn \c_space_token { ~ }
+  \token_new:Nn \c_catcode_letter_token { a }
+  \token_new:Nn \c_catcode_other_token { 1 }
+\group_end:
+\group_begin:
+  \char_set_catcode_active:N \*
+  \cs_new_nopar:Npn \c_catcode_active_tl { \exp_not:N * }
+\group_end:
+\prg_new_conditional:Npnn \token_if_group_begin:N #1 { p , T ,  F , TF }
+  {
+    \if_catcode:w \exp_not:N #1 \c_group_begin_token
+      \prg_return_true: \else: \prg_return_false: \fi:
+  }
+\prg_new_conditional:Npnn \token_if_group_end:N #1 { p , T ,  F , TF }
+  {
+    \if_catcode:w \exp_not:N #1 \c_group_end_token
+      \prg_return_true: \else: \prg_return_false: \fi:
+  }
+\prg_new_conditional:Npnn \token_if_math_toggle:N #1 { p , T ,  F , TF }
+  {
+    \if_catcode:w \exp_not:N #1 \c_math_toggle_token
+      \prg_return_true: \else: \prg_return_false: \fi:
+  }
+\prg_new_conditional:Npnn \token_if_alignment:N #1 { p , T ,  F , TF }
+  {
+    \if_catcode:w \exp_not:N #1 \c_alignment_token
+      \prg_return_true: \else: \prg_return_false: \fi:
+  }
+\group_begin:
+\cs_set_eq:NN \c_parameter_token \scan_stop:
+\prg_new_conditional:Npnn \token_if_parameter:N #1 { p , T ,  F , TF }
+  {
+    \if_catcode:w \exp_not:N #1 \c_parameter_token
+      \prg_return_true: \else: \prg_return_false: \fi:
+  }
+\group_end:
+\prg_new_conditional:Npnn \token_if_math_superscript:N #1 { p , T ,  F , TF }
+  {
+    \if_catcode:w \exp_not:N #1 \c_math_superscript_token
+      \prg_return_true: \else: \prg_return_false: \fi:
+  }
+\prg_new_conditional:Npnn \token_if_math_subscript:N #1 { p , T ,  F , TF }
+  {
+    \if_catcode:w \exp_not:N #1 \c_math_subscript_token
+      \prg_return_true: \else: \prg_return_false: \fi:
+  }
+\prg_new_conditional:Npnn \token_if_space:N #1 { p , T ,  F , TF }
+  {
+    \if_catcode:w \exp_not:N #1 \c_space_token
+      \prg_return_true: \else: \prg_return_false: \fi:
+  }
+\prg_new_conditional:Npnn \token_if_letter:N #1 { p , T ,  F , TF }
+  {
+    \if_catcode:w \exp_not:N #1 \c_catcode_letter_token
+      \prg_return_true: \else: \prg_return_false: \fi:
+  }
+\prg_new_conditional:Npnn \token_if_other:N #1 { p , T ,  F , TF }
+  {
+    \if_catcode:w \exp_not:N #1 \c_catcode_other_token
+      \prg_return_true: \else: \prg_return_false: \fi:
+  }
+\prg_new_conditional:Npnn \token_if_active:N #1 { p , T ,  F , TF }
+  {
+    \if_catcode:w \exp_not:N #1 \c_catcode_active_tl
+      \prg_return_true: \else: \prg_return_false: \fi:
+  }
+\prg_new_conditional:Npnn \token_if_eq_meaning:NN #1#2 { p , T ,  F , TF }
+  {
+    \if_meaning:w  #1  #2
+      \prg_return_true: \else: \prg_return_false: \fi:
+  }
+\prg_new_conditional:Npnn \token_if_eq_catcode:NN #1#2 { p , T ,  F , TF }
+  {
+    \if_catcode:w \exp_not:N #1 \exp_not:N #2
+      \prg_return_true: \else: \prg_return_false: \fi:
+  }
+\prg_new_conditional:Npnn \token_if_eq_charcode:NN #1#2 { p , T ,  F , TF }
+  {
+    \if_charcode:w \exp_not:N #1 \exp_not:N #2
+      \prg_return_true: \else: \prg_return_false: \fi:
+  }
+\group_begin:
+\char_set_catcode_other:N \M
+\char_set_catcode_other:N \A
+\char_set_lccode:nn { `\; } { `\: }
+\char_set_lccode:nn { `\T } { `\T }
+\char_set_lccode:nn { `\F } { `\F }
+\tl_to_lowercase:n
+  {
+    \group_end:
+    \prg_new_conditional:Npnn \token_if_macro:N #1 { p , T ,  F , TF }
+      {
+        \exp_after:wN \token_if_macro_p_aux:w
+        \token_to_meaning:N #1 MA; \q_stop
+      }
+    \cs_new_nopar:Npn \token_if_macro_p_aux:w #1 MA #2 ; #3 \q_stop
+      {
+        \if_int_compare:w \pdftex_strcmp:D { #2 } { cro } = \c_zero
+            \prg_return_true:
+        \else:
+            \prg_return_false:
+        \fi:
+      }
+  }
+\prg_new_conditional:Npnn \token_if_cs:N #1 { p , T ,  F , TF }
+  {
+    \if_catcode:w \exp_not:N #1 \scan_stop:
+      \prg_return_true: \else: \prg_return_false: \fi:
+  }
+\prg_new_conditional:Npnn \token_if_expandable:N #1 { p , T ,  F , TF }
+  {
+    \cs_if_exist:NTF #1
+      {
+        \exp_after:wN \if_meaning:w \exp_not:N #1 #1
+          \prg_return_false: \else: \prg_return_true: \fi:
+      }
+      { \prg_return_false: }
+  }
+\group_begin:
+  \char_set_lccode:nn { `\T } { `\T }
+  \char_set_lccode:nn { `\F } { `\F }
+  \char_set_lccode:nn { `\X } { `\n }
+  \char_set_lccode:nn { `\Y } { `\t }
+  \char_set_lccode:nn { `\Z } { `\d }
+  \char_set_lccode:nn { `\? } { `\\ }
+  \tl_map_inline:nn { \X \Y \Z \M \C \H \A \R \O \U \S \K \I \P \L \G \P \E }
+    { \char_set_catcode:nn { `#1 } \c_twelve }
+\tl_to_lowercase:n
+  {
+    \group_end:
+    \prg_new_conditional:Npnn \token_if_chardef:N #1 { p , T ,  F , TF }
+      {
+        \exp_after:wN \token_if_chardef_aux:w
+          \token_to_meaning:N #1 ?CHAR" \q_stop
+      }
+    \cs_new_nopar:Npn \token_if_chardef_aux:w #1 ?CHAR" #2 \q_stop
+      { \tl_if_empty:nTF {#1} { \prg_return_true: } { \prg_return_false: } }
+    \prg_new_conditional:Npnn \token_if_mathchardef:N #1 { p , T ,  F , TF }
+      {
+        \exp_after:wN \token_if_mathchardef_aux:w
+          \token_to_meaning:N #1 ?MAYHCHAR" \q_stop
+      }
+    \cs_new_nopar:Npn \token_if_mathchardef_aux:w #1 ?MAYHCHAR" #2 \q_stop
+      { \tl_if_empty:nTF {#1} { \prg_return_true: } { \prg_return_false: } }
+    \prg_new_conditional:Npnn \token_if_int_register:N #1 { p , T ,  F , TF }
+      {
+        \if_meaning:w \tex_countdef:D #1
+          \prg_return_false:
+        \else:
+          \exp_after:wN \token_if_int_register_aux:w
+            \token_to_meaning:N #1 ?COUXY \q_stop
+        \fi:
+      }
+    \cs_new_nopar:Npn \token_if_int_register_aux:w #1 ?COUXY #2 \q_stop
+      { \tl_if_empty:nTF {#1} { \prg_return_true: } { \prg_return_false: } }
+    \prg_new_conditional:Npnn \token_if_skip_register:N #1 { p , T ,  F , TF }
+      {
+        \if_meaning:w \tex_skipdef:D #1
+          \prg_return_false:
+        \else:
+          \exp_after:wN \token_if_skip_register_aux:w
+            \token_to_meaning:N #1?SKIP\q_stop
+        \fi:
+      }
+    \cs_new_nopar:Npn \token_if_skip_register_aux:w #1 ?SKIP #2 \q_stop
+      { \tl_if_empty:nTF {#1} { \prg_return_true: } { \prg_return_false: } }
+    \prg_new_conditional:Npnn \token_if_dim_register:N #1 { p , T ,  F , TF }
+      {
+        \if_meaning:w \tex_dimendef:D #1
+          \c_false_bool
+        \else:
+          \exp_after:wN \token_if_dim_register_aux:w
+            \token_to_meaning:N #1 ?ZIMEX \q_stop
+        \fi:
+      }
+    \cs_new_nopar:Npn \token_if_dim_register_aux:w #1 ?ZIMEX #2 \q_stop
+      { \tl_if_empty:nTF {#1} { \prg_return_true: } { \prg_return_false: } }
+    \prg_new_conditional:Npnn \token_if_toks_register:N #1 { p , T ,  F , TF }
+      {
+        \if_meaning:w \tex_toksdef:D #1
+          \prg_return_false:
+        \else:
+          \exp_after:wN \token_if_toks_register_aux:w
+            \token_to_meaning:N #1 ?YOKS \q_stop
+        \fi:
+      }
+     \cs_new_nopar:Npn \token_if_toks_register_aux:w #1 ?YOKS #2 \q_stop
+       { \tl_if_empty:nTF {#1} { \prg_return_true: } { \prg_return_false: } }
+    \prg_new_conditional:Npnn \token_if_protected_macro:N #1
+      { p , T ,  F , TF }
+      {
+        \exp_after:wN \token_if_protected_macro_aux:w
+          \token_to_meaning:N #1 ?PROYECYEZ~MACRO \q_stop
+      }
+    \cs_new_nopar:Npn \token_if_protected_macro_aux:w
+      #1 ?PROYECYEZ~MACRO #2 \q_stop
+      { \tl_if_empty:nTF {#1} { \prg_return_true: } { \prg_return_false: } }
+    \prg_new_conditional:Npnn \token_if_long_macro:N #1 { p , T ,  F , TF }
+      {
+        \exp_after:wN \token_if_long_macro_aux:w
+          \token_to_meaning:N #1 ?LOXG~MACRO \q_stop
+      }
+    \cs_new_nopar:Npn \token_if_long_macro_aux:w #1 ?LOXG~MACRO #2 \q_stop
+      { \tl_if_empty:nTF {#1} { \prg_return_true: } { \prg_return_false: } }
+    \prg_new_conditional:Npnn \token_if_protected_long_macro:N #1
+      { p , T ,  F , TF }
+      {
+        \exp_after:wN \token_if_protected_long_macro_aux:w
+          \token_to_meaning:N #1 ?PROYECYEZ?LOXG~MACRO \q_stop
+      }
+    \cs_new_nopar:Npn \token_if_protected_long_macro_aux:w
+      #1 ?PROYECYEZ?LOXG~MACRO #2 \q_stop
+      { \tl_if_empty:nTF {#1} { \prg_return_true: } { \prg_return_false: } }
+  }
+\tex_chardef:D \c_token_A_int = `A ~ %
+\group_begin:
+\char_set_catcode_other:N \;
+\char_set_lccode:nn { `\; } { `\: }
+\char_set_lccode:nn { `\T } { `\T }
+\char_set_lccode:nn { `\F } { `\F }
+\tl_to_lowercase:n {
+  \group_end:
+  \prg_new_conditional:Npnn \token_if_primitive:N #1 { p , T , F , TF }
+    {
+      \token_if_macro:NTF #1
+        \prg_return_false:
+        {
+          \exp_after:wN \token_if_primitive_aux:NNw
+          \token_to_meaning:N #1 ; ; ; \q_stop #1
+        }
+    }
+  \cs_new_nopar:Npn \token_if_primitive_aux:NNw #1#2 #3 ; #4 \q_stop
+    {
+      \tl_if_empty:oTF { \token_if_primitive_aux_space:w #3 ~ }
+        { \token_if_primitive_aux_loop:N #3 ; \q_stop }
+        { \token_if_primitive_aux_nullfont:N }
+    }
+}
+\cs_new_nopar:Npn \token_if_primitive_aux_space:w #1 ~ { }
+\cs_new:Npn \token_if_primitive_aux_nullfont:N #1
+  {
+    \if_meaning:w \tex_nullfont:D #1
+      \prg_return_true:
+    \else:
+      \prg_return_false:
+    \fi:
+  }
+\cs_new_nopar:Npn \token_if_primitive_aux_loop:N #1
+  {
+    \if_num:w `#1 < \c_token_A_int %
+      \exp_after:wN \token_if_primitive_auxii:Nw
+      \exp_after:wN #1
+    \else:
+      \exp_after:wN \token_if_primitive_aux_loop:N
+    \fi:
+  }
+\cs_new_nopar:Npn \token_if_primitive_auxii:Nw #1 #2 \q_stop
+  {
+    \if:w : #1
+      \exp_after:wN \token_if_primitive_aux_undefined:N
+    \else:
+      \prg_return_false:
+      \exp_after:wN \use_none:n
+    \fi:
+  }
+\cs_new:Npn \token_if_primitive_aux_undefined:N #1
+  {
+    \if_cs_exist:N #1
+      \prg_return_true:
+    \else:
+      \prg_return_false:
+    \fi:
+  }
+\cs_new_eq:NN \l_peek_token ?
+\cs_new_eq:NN \g_peek_token ?
+\cs_new_eq:NN \l_peek_search_token ?
+\cs_new_nopar:Npn \l_peek_search_tl { }
+\cs_new_nopar:Npn \peek_true:w  { }
+\cs_new_nopar:Npn \peek_true_aux:w  { }
+\cs_new_nopar:Npn \peek_false:w { }
+\cs_new:Npn \peek_tmp:w { }
+\cs_new_protected_nopar:Npn \peek_after:Nw
+  { \tex_futurelet:D \l_peek_token }
+\cs_new_protected_nopar:Npn \peek_gafter:Nw
+  { \pref_global:D \tex_futurelet:D \g_peek_token }
+\cs_new_protected:Npn \peek_true_remove:w
+  {
+    \group_align_safe_end:
+    \tex_afterassignment:D \peek_true_aux:w
+    \cs_set_eq:NN \peek_tmp:w
+  }
+\cs_new_protected:Npn \peek_token_generic:NNTF #1#2#3#4
+  {
+    \cs_set_eq:NN \l_peek_search_token #2
+    \tl_set:Nn \l_peek_search_tl {#2}
+    \cs_set_nopar:Npx \peek_true:w
+      {
+        \exp_not:N \group_align_safe_end:
+        \exp_not:n {#3}
+      }
+    \cs_set_nopar:Npx \peek_false:w
+      {
+        \exp_not:N \group_align_safe_end:
+        \exp_not:n {#4}
+      }
+    \group_align_safe_begin:
+      \peek_after:Nw #1
+  }
+\cs_new_protected:Npn \peek_token_generic:NNT #1#2#3
+  { \peek_token_generic:NNTF #1 #2 {#3} { } }
+\cs_new_protected:Npn \peek_token_generic:NNF #1#2#3
+  { \peek_token_generic:NNTF #1 #2 { } {#3} }
+\cs_new_protected:Npn \peek_token_remove_generic:NNTF #1#2#3#4
+  {
+    \cs_set_eq:NN \l_peek_search_token #2
+    \tl_set:Nn \l_peek_search_tl {#2}
+    \cs_set_eq:NN \peek_true:w \peek_true_remove:w
+    \cs_set_nopar:Npx \peek_true_aux:w { \exp_not:n {#3} }
+    \cs_set_nopar:Npx \peek_false:w
+      {
+        \exp_not:N \group_align_safe_end:
+        \exp_not:n {#4}
+      }
+    \group_align_safe_begin:
+      \peek_after:Nw #1
+  }
+\cs_new_protected:Npn \peek_token_remove_generic:NNT #1#2#3
+  { \peek_token_remove_generic:NNTF #1 #2 {#3} { } }
+\cs_new_protected:Npn \peek_token_remove_generic:NNF #1#2#3
+  { \peek_token_remove_generic:NNTF #1 #2 { } {#3} }
+\cs_new_nopar:Npn \peek_execute_branches_catcode:
+  {
+    \if_catcode:w
+      \exp_not:N \l_peek_token \exp_not:N \l_peek_search_token
+      \exp_after:wN \peek_true:w
+    \else:
+      \exp_after:wN \peek_false:w
+    \fi:
+  }
+\cs_new_nopar:Npn \peek_execute_branches_meaning:
+  {
+    \if_meaning:w \l_peek_token \l_peek_search_token
+      \exp_after:wN \peek_true:w
+    \else:
+      \exp_after:wN \peek_false:w
+    \fi:
+  }
+\cs_new_nopar:Npn \peek_execute_branches_charcode:
+  {
+    \bool_if:nTF
+      {
+           \token_if_eq_catcode_p:NN \l_peek_token \c_group_begin_token
+        || \token_if_eq_catcode_p:NN \l_peek_token \c_group_end_token
+        || \token_if_eq_meaning_p:NN \l_peek_token \c_space_token
+      }
+      { \peek_false:w }
+      {
+        \exp_after:wN \peek_execute_branches_charcode_aux:NN
+          \l_peek_search_tl
+      }
+  }
+\cs_new:Npn \peek_execute_branches_charcode_aux:NN #1#2
+  {
+    \if:w \exp_not:N #1 \exp_not:N #2
+      \exp_after:wN \peek_true:w
+    \else:
+      \exp_after:wN \peek_false:w
+    \fi:
+    #2
+  }
+\cs_new_protected_nopar:Npn \peek_ignore_spaces_execute_branches:
+  {
+    \token_if_eq_meaning:NNTF \l_peek_token \c_space_token
+      {
+        \tex_afterassignment:D \peek_ignore_spaces_execute_branches_aux:
+        \cs_set_eq:NN \peek_tmp:w
+      }
+      { \peek_execute_branches: }
+  }
+\cs_new_protected_nopar:Npn \peek_ignore_spaces_execute_branches_aux:
+  { \peek_after:Nw \peek_ignore_spaces_execute_branches: }
+\group_begin:
+  \cs_set_nopar:Npn \peek_def:nnnn #1#2#3#4
+    {
+      \peek_def_aux:nnnnn {#1} {#2} {#3} {#4} { TF }
+      \peek_def_aux:nnnnn {#1} {#2} {#3} {#4} { T }
+      \peek_def_aux:nnnnn {#1} {#2} {#3} {#4} { F }
+    }
+  \cs_set_nopar:Npn \peek_def_aux:nnnnn #1#2#3#4#5
+    {
+      \cs_gset_nopar:cpx { #1 #5 }
+        {
+          \tl_if_empty:nF {#2}
+            { \exp_not:n { \cs_set_eq:NN \peek_execute_branches: #2 } }
+          \exp_not:c { #3 #5 }
+          \exp_not:n {#4}
+        }
+    }
+  \peek_def:nnnn { peek_catcode:N }
+    { }
+    { peek_token_generic:NN }
+    { \peek_execute_branches_catcode: }
+  \peek_def:nnnn { peek_catcode_ignore_spaces:N }
+    { \peek_execute_branches_catcode: }
+    { peek_token_generic:NN }
+    { \peek_ignore_spaces_execute_branches: }
+  \peek_def:nnnn { peek_catcode_remove:N }
+    { }
+    { peek_token_remove_generic:NN }
+    { \peek_execute_branches_catcode: }
+  \peek_def:nnnn { peek_catcode_remove_ignore_spaces:N }
+    { \peek_execute_branches_catcode: }
+    { peek_token_remove_generic:NN }
+    { \peek_ignore_spaces_execute_branches: }
+  \peek_def:nnnn { peek_charcode:N }
+    { }
+    { peek_token_generic:NN }
+    { \peek_execute_branches_charcode: }
+  \peek_def:nnnn { peek_charcode_ignore_spaces:N }
+    { \peek_execute_branches_charcode: }
+    { peek_token_generic:NN }
+    { \peek_ignore_spaces_execute_branches: }
+  \peek_def:nnnn { peek_charcode_remove:N }
+    { }
+    { peek_token_remove_generic:NN }
+    { \peek_execute_branches_charcode: }
+  \peek_def:nnnn { peek_charcode_remove_ignore_spaces:N }
+    { \peek_execute_branches_charcode: }
+    { peek_token_remove_generic:NN }
+    { \peek_ignore_spaces_execute_branches: }
+  \peek_def:nnnn { peek_meaning:N }
+    { }
+    { peek_token_generic:NN }
+    { \peek_execute_branches_meaning: }
+  \peek_def:nnnn { peek_meaning_ignore_spaces:N }
+    { \peek_execute_branches_meaning: }
+    { peek_token_generic:NN }
+    { \peek_ignore_spaces_execute_branches: }
+  \peek_def:nnnn { peek_meaning_remove:N }
+    { }
+    { peek_token_remove_generic:NN }
+    { \peek_execute_branches_meaning: }
+  \peek_def:nnnn { peek_meaning_remove_ignore_spaces:N }
+    { \peek_execute_branches_meaning: }
+    { peek_token_remove_generic:NN }
+    { \peek_ignore_spaces_execute_branches: }
+\group_end:
+\exp_args:Nno \use:nn
+  { \cs_new_nopar:Npn \token_get_prefix_arg_replacement_aux:wN #1 }
+  { \tl_to_str:n { macro : } #2 -> #3 \q_stop #4 }
+  { #4 {#1} {#2} {#3} }
+\cs_new:Npn \token_get_prefix_spec:N #1
+  {
+    \token_if_macro:NTF #1
+      {
+        \exp_after:wN \token_get_prefix_arg_replacement_aux:wN
+          \token_to_meaning:N #1 \q_stop \use_i:nnn
+      }
+      { \scan_stop: }
+  }
+\cs_new:Npn \token_get_arg_spec:N #1
+  {
+    \token_if_macro:NTF #1
+      {
+        \exp_after:wN \token_get_prefix_arg_replacement_aux:wN
+          \token_to_meaning:N #1 \q_stop \use_ii:nnn
+      }
+      { \scan_stop: }
+  }
+\cs_new:Npn \token_get_replacement_spec:N #1
+  {
+    \token_if_macro:NTF #1
+      {
+        \exp_after:wN \token_get_prefix_arg_replacement_aux:wN
+          \token_to_meaning:N #1 \q_stop \use_iii:nnn
+      }
+      { \scan_stop: }
+  }
+\group_begin:
+  \char_set_catcode_active:N \^^@
+  \cs_set:Npn \char_tmp:NN #1#2
+    {
+      \cs_new:Npn #1 ##1
+        {
+          \char_set_catcode_active:n { `##1 }
+          \group_begin:
+          \char_set_lccode:nn { `\^^@ } { `##1 }
+          \tl_to_lowercase:n { \group_end: #2 ^^@ }
+        }
+    }
+  \char_tmp:NN \char_active_set:Npn    \cs_set:Npn
+  \char_tmp:NN \char_active_set:Npx    \cs_set:Npx
+  \char_tmp:NN \char_active_gset:Npn   \cs_gset:Npn
+  \char_tmp:NN \char_active_gset:Npx   \cs_gset:Npx
+  \char_tmp:NN \char_active_set_eq:NN  \cs_set_eq:NN
+  \char_tmp:NN \char_active_gset_eq:NN \cs_gset_eq:NN
+\group_end:
+\cs_new_eq:NN \char_set_catcode:w  \tex_catcode:D
+\cs_new_eq:NN \char_set_mathcode:w \tex_mathcode:D
+\cs_new_eq:NN \char_set_lccode:w   \tex_lccode:D
+\cs_new_eq:NN \char_set_uccode:w   \tex_uccode:D
+\cs_new_eq:NN \char_set_sfcode:w   \tex_sfcode:D
+\cs_new_nopar:Npn \char_value_catcode:w { \tex_the:D \char_set_catcode:w }
+\cs_new_nopar:Npn \char_show_value_catcode:w
+  { \tex_showthe:D \char_set_catcode:w }
+\cs_new_nopar:Npn \char_value_mathcode:w { \tex_the:D \char_set_mathcode:w }
+\cs_new_nopar:Npn \char_show_value_mathcode:w
+  { \tex_showthe:D \char_set_mathcode:w }
+\cs_new_nopar:Npn \char_value_lccode:w { \tex_the:D \char_set_lccode:w }
+\cs_new_nopar:Npn \char_show_value_lccode:w
+  { \tex_showthe:D \char_set_lccode:w }
+\cs_new_nopar:Npn \char_value_uccode:w { \tex_the:D \char_set_uccode:w }
+\cs_new_nopar:Npn \char_show_value_uccode:w
+  { \tex_showthe:D \char_set_uccode:w }
+\cs_new_nopar:Npn \char_value_sfcode:w { \tex_the:D \char_set_sfcode:w }
+\cs_new_nopar:Npn \char_show_value_sfcode:w
+  { \tex_showthe:D \char_set_sfcode:w }
+\cs_new_eq:NN \peek_after:NN  \peek_after:Nw
+\cs_new_eq:NN \peek_gafter:NN \peek_gafter:Nw
+\cs_new_eq:NN \c_alignment_tab_token \c_alignment_token
+\cs_new_eq:NN \c_math_shift_token    \c_math_toggle_token
+\cs_new_eq:NN \c_letter_token        \c_catcode_letter_token
+\cs_new_eq:NN \c_other_char_token    \c_catcode_other_token
+\cs_new_eq:NN \c_active_char_token \c_catcode_active_tl
+\cs_new_eq:NN \char_make_escape:N           \char_set_catcode_escape:N
+\cs_new_eq:NN \char_make_begin_group:N      \char_set_catcode_group_begin:N
+\cs_new_eq:NN \char_make_end_group:N        \char_set_catcode_group_end:N
+\cs_new_eq:NN \char_make_math_shift:N       \char_set_catcode_math_toggle:N
+\cs_new_eq:NN \char_make_alignment_tab:N    \char_set_catcode_alignment:N
+\cs_new_eq:NN \char_make_end_line:N         \char_set_catcode_end_line:N
+\cs_new_eq:NN \char_make_parameter:N        \char_set_catcode_parameter:N
+\cs_new_eq:NN \char_make_math_superscript:N
+  \char_set_catcode_math_superscript:N
+\cs_new_eq:NN \char_make_math_subscript:N
+  \char_set_catcode_math_subscript:N
+\cs_new_eq:NN \char_make_ignore:N           \char_set_catcode_ignore:N
+\cs_new_eq:NN \char_make_space:N            \char_set_catcode_space:N
+\cs_new_eq:NN \char_make_letter:N           \char_set_catcode_letter:N
+\cs_new_eq:NN \char_make_other:N            \char_set_catcode_other:N
+\cs_new_eq:NN \char_make_active:N           \char_set_catcode_active:N
+\cs_new_eq:NN \char_make_comment:N          \char_set_catcode_comment:N
+\cs_new_eq:NN \char_make_invalid:N          \char_set_catcode_invalid:N
+\cs_new_eq:NN \char_make_escape:n           \char_set_catcode_escape:n
+\cs_new_eq:NN \char_make_begin_group:n      \char_set_catcode_group_begin:n
+\cs_new_eq:NN \char_make_end_group:n        \char_set_catcode_group_end:n
+\cs_new_eq:NN \char_make_math_shift:n       \char_set_catcode_math_toggle:n
+\cs_new_eq:NN \char_make_alignment_tab:n    \char_set_catcode_alignment:n
+\cs_new_eq:NN \char_make_end_line:n         \char_set_catcode_end_line:n
+\cs_new_eq:NN \char_make_parameter:n        \char_set_catcode_parameter:n
+\cs_new_eq:NN \char_make_math_superscript:n
+  \char_set_catcode_math_superscript:n
+\cs_new_eq:NN \char_make_math_subscript:n
+  \char_set_catcode_math_subscript:n
+\cs_new_eq:NN \char_make_ignore:n           \char_set_catcode_ignore:n
+\cs_new_eq:NN \char_make_space:n           \char_set_catcode_space:n
+\cs_new_eq:NN \char_make_letter:n           \char_set_catcode_letter:n
+\cs_new_eq:NN \char_make_other:n            \char_set_catcode_other:n
+\cs_new_eq:NN \char_make_active:n           \char_set_catcode_active:n
+\cs_new_eq:NN \char_make_comment:n          \char_set_catcode_comment:n
+\cs_new_eq:NN \char_make_invalid:n          \char_set_catcode_invalid:n
+\cs_new_eq:NN \token_if_alignment_tab_p:N \token_if_alignment_p:N
+\cs_new_eq:NN \token_if_alignment_tab:NT  \token_if_alignment:NT
+\cs_new_eq:NN \token_if_alignment_tab:NF  \token_if_alignment:NF
+\cs_new_eq:NN \token_if_alignment_tab:NTF \token_if_alignment:NTF
+\cs_new_eq:NN \token_if_math_shift_p:N \token_if_math_toggle_p:N
+\cs_new_eq:NN \token_if_math_shift:NT  \token_if_math_toggle:NT
+\cs_new_eq:NN \token_if_math_shift:NF  \token_if_math_toggle:NF
+\cs_new_eq:NN \token_if_math_shift:NTF \token_if_math_toggle:NTF
+\cs_new_eq:NN \token_if_other_char_p:N \token_if_other_p:N
+\cs_new_eq:NN \token_if_other_char:NT  \token_if_other:NT
+\cs_new_eq:NN \token_if_other_char:NF  \token_if_other:NF
+\cs_new_eq:NN \token_if_other_char:NTF \token_if_other:NTF
+\cs_new_eq:NN \token_if_active_char_p:N \token_if_active_p:N
+\cs_new_eq:NN \token_if_active_char:NT  \token_if_active:NT
+\cs_new_eq:NN \token_if_active_char:NF  \token_if_active:NF
+\cs_new_eq:NN \token_if_active_char:NTF \token_if_active:NTF
+%% 
+%%
+%% End of file `l3token.sty'.
diff --git a/Master/texmf-dist/tex/latex/l3kernel/l3toks.sty b/Master/texmf-dist/tex/latex/l3kernel/l3toks.sty
new file mode 100644
index 00000000000..1b1ab880687
--- /dev/null
+++ b/Master/texmf-dist/tex/latex/l3kernel/l3toks.sty
@@ -0,0 +1,163 @@
+%%
+%% This is file `l3toks.sty',
+%% generated with the docstrip utility.
+%%
+%% The original source files were:
+%%
+%% l3toks.dtx  (with options: `package')
+%% 
+%% EXPERIMENTAL CODE
+%% 
+%% Do not distribute this file without also distributing the
+%% source files specified above.
+%% 
+%% Do not distribute a modified version of this file.
+%% 
+%% File: l3toks.dtx Copyright (C) 1990-2011 The LaTeX3 project
+%%
+%% It may be distributed and/or modified under the conditions of the
+%% LaTeX Project Public License (LPPL), either version 1.3c of this
+%% license or (at your option) any later version.  The latest version
+%% of this license is in the file
+%%
+%%    http://www.latex-project.org/lppl.txt
+%%
+%% This file is part of the "expl3 bundle" (The Work in LPPL)
+%% and all files in that bundle must be distributed together.
+%%
+%% The released version of this bundle is available from CTAN.
+%%
+%% -----------------------------------------------------------------------
+%%
+%% The development version of the bundle can be found at
+%%
+%%    http://www.latex-project.org/svnroot/experimental/trunk/
+%%
+%% for those people who are interested.
+%%
+%%%%%%%%%%%
+%% NOTE: %%
+%%%%%%%%%%%
+%%
+%%   Snapshots taken from the repository represent work in progress and may
+%%   not work or may contain conflicting material!  We therefore ask
+%%   people _not_ to put them into distributions, archives, etc. without
+%%   prior consultation with the LaTeX3 Project.
+%%
+%% -----------------------------------------------------------------------
+\RequirePackage{l3names}
+\GetIdInfo$Id: l3toks.dtx 2478 2011-06-19 21:34:23Z joseph $
+  {L3 Experimental token registers}
+\ProvidesExplPackage
+  {\ExplFileName}{\ExplFileDate}{\ExplFileVersion}{\ExplFileDescription}
+\package_check_loaded_expl:
+\cs_new_protected_nopar:Npn \toks_new:N #1 {
+  \chk_if_free_cs:N #1
+  \newtoks #1
+}
+\cs_generate_variant:Nn \toks_new:N {c}
+\cs_new_eq:NN  \toks_use:N \tex_the:D
+\cs_generate_variant:Nn \toks_use:N {c}
+\cs_new_eq:NN       \toks_set:Nn \prg_do_nothing:
+\cs_new_protected:Npn \toks_set:NV #1#2 {
+  #1 \exp_after:wN { \int_to_roman:w -`0 \exp_eval_register:N #2 }
+}
+\cs_new_protected:Npn \toks_set:Nv #1#2 {
+  #1 \exp_after:wN { \int_to_roman:w -`0 \exp_eval_register:c {#2} }
+}
+\cs_new_protected:Npn \toks_set:No #1#2 { #1 \exp_after:wN {#2} }
+\cs_new_protected:Npn \toks_set:Nf #1#2 {
+  #1 \exp_after:wN { \int_to_roman:w -`0#2 }
+}
+\cs_generate_variant:Nn \toks_set:Nn {Nx,cn,cV,cv,co,cx,cf}
+\cs_new_eq:NN \toks_gset:Nn \pref_global:D
+\cs_generate_variant:Nn \toks_gset:Nn {NV,No,Nx,cn,cV,co,cx}
+\cs_new_eq:NN \toks_set_eq:NN  \prg_do_nothing:
+\cs_new_eq:NN \toks_gset_eq:NN \pref_global:D
+\cs_generate_variant:Nn \toks_set_eq:NN  {Nc,cN,cc}
+\cs_generate_variant:Nn \toks_gset_eq:NN  {Nc,cN,cc}
+\cs_new_protected_nopar:Npn \toks_clear:N #1 {
+  #1\c_empty_toks
+}
+\cs_new_protected_nopar:Npn \toks_gclear:N {
+  \pref_global:D
+  \toks_clear:N
+}
+\cs_generate_variant:Nn \toks_clear:N  {c}
+\cs_generate_variant:Nn \toks_gclear:N {c}
+\cs_new_protected_nopar:Npn \toks_use_clear:N #1 {
+  \exp_last_unbraced:NNV \toks_clear:N #1 #1
+}
+\cs_new_protected_nopar:Npn \toks_use_gclear:N {
+  \pref_global:D
+  \toks_use_clear:N
+}
+\cs_generate_variant:Nn \toks_use_clear:N  {c}
+\cs_generate_variant:Nn \toks_use_gclear:N {c}
+\cs_new_eq:NN           \toks_show:N \kernel_register_show:N
+\cs_generate_variant:Nn \toks_show:N {c}
+\cs_new_protected_nopar:Npn \toks_put_left:Nn #1 {
+  \exp_after:wN \toks_put_left_aux:w \exp_after:wN \q_nil
+    \toks_use:N #1 \q_stop #1
+}
+\cs_generate_variant:Nn \toks_put_left:Nn  {NV,No,Nx,cn,co,cV}
+\cs_new_protected_nopar:Npn \toks_gput_left:Nn {
+  \pref_global:D
+  \toks_put_left:Nn
+}
+\cs_generate_variant:Nn \toks_gput_left:Nn  {NV,No,Nx,cn,cV,co}
+\cs_new:Npn \toks_put_left_aux:w #1\q_stop #2#3 {
+  #2 \exp_after:wN { \use_i:nn {#3} #1 }
+}
+\cs_new_protected:Npn \toks_put_right:Nn #1#2 {
+  #1 \exp_after:wN { \toks_use:N #1 #2 }
+}
+\cs_new_protected_nopar:Npn \toks_gput_right:Nn {
+  \pref_global:D
+  \toks_put_right:Nn
+}
+\cs_new_protected:Npn \toks_put_right:NV #1#2 {
+  #1 \exp_after:wN \exp_after:wN \exp_after:wN {
+     \exp_after:wN \toks_use:N \exp_after:wN #1
+     \int_to_roman:w -`0 \exp_eval_register:N #2
+  }
+}
+\cs_new_protected:Npn \toks_put_right:No #1#2 {
+  #1 \exp_after:wN \exp_after:wN \exp_after:wN {
+    \exp_after:wN \toks_use:N \exp_after:wN #1 #2
+  }
+}
+\cs_generate_variant:Nn \toks_put_right:Nn {Nx,cn,cV,co}
+\cs_generate_variant:Nn \toks_gput_right:Nn {NV,No,Nx,cn,cV,co}
+\cs_new_protected:Npn \toks_put_right:Nf #1#2 {
+  #1 \exp_after:wN \exp_after:wN \exp_after:wN {
+     \exp_after:wN \toks_use:N \exp_after:wN #1 \int_to_roman:w -`0#2
+  }
+}
+\prg_new_conditional:Nnn \toks_if_empty:N {p,TF,T,F} {
+  \tl_if_empty:VTF #1 {\prg_return_true:} {\prg_return_false:}
+}
+\cs_generate_variant:Nn \toks_if_empty_p:N {c}
+\cs_generate_variant:Nn \toks_if_empty:NTF {c}
+\cs_generate_variant:Nn \toks_if_empty:NT  {c}
+\cs_generate_variant:Nn \toks_if_empty:NF  {c}
+\prg_new_conditional:Nnn \toks_if_eq:NN {p,TF,T,F} {
+  \str_if_eq:xxTF {\toks_use:N #1} {\toks_use:N #2}
+      {\prg_return_true:} {\prg_return_false:}
+}
+\cs_generate_variant:Nn \toks_if_eq_p:NN {Nc,c,cc}
+\cs_generate_variant:Nn \toks_if_eq:NNTF {Nc,c,cc}
+\cs_generate_variant:Nn \toks_if_eq:NNT  {Nc,c,cc}
+\cs_generate_variant:Nn \toks_if_eq:NNF  {Nc,c,cc}
+\tex_toksdef:D \l_tmpa_toks = 255\scan_stop:
+\toks_new:N \l_tmpb_toks
+\toks_new:N \l_tmpc_toks
+\toks_new:N \g_tmpa_toks
+\toks_new:N \g_tmpb_toks
+\toks_new:N \g_tmpc_toks
+\toks_new:N \c_empty_toks
+\toks_new:N \l_tl_replace_toks
+
+%% 
+%%
+%% End of file `l3toks.sty'.
diff --git a/Master/tlpkg/bin/tlpkg-ctan-check b/Master/tlpkg/bin/tlpkg-ctan-check
index 322b2078dff..27c3fad9b79 100755
--- a/Master/tlpkg/bin/tlpkg-ctan-check
+++ b/Master/tlpkg/bin/tlpkg-ctan-check
@@ -198,6 +198,7 @@ my @TLP_working = qw(
     knitting knittingpattern knuth
     koma-moderncvclassic koma-script kpfonts kurier
   l2picfaq l2tabu l2tabu-english l2tabu-french l2tabu-it l2tabu-spanish
+    l3kernel
     labbook labelcas labels lastpage
     latex latex-bib-ex latex-course latex-doc-ptr latex-notes-zh-cn
     latex-referenz latex-tabellen
diff --git a/Master/tlpkg/tlpsrc/collection-latexrecommended.tlpsrc b/Master/tlpkg/tlpsrc/collection-latexrecommended.tlpsrc
index ac0e739d737..8ad7bb8bdde 100644
--- a/Master/tlpkg/tlpsrc/collection-latexrecommended.tlpsrc
+++ b/Master/tlpkg/tlpsrc/collection-latexrecommended.tlpsrc
@@ -25,6 +25,7 @@ depend fp
 depend index
 depend jknapltx
 depend koma-script
+depend l3kernel
 depend listings
 depend mdwtools
 depend memoir
diff --git a/Master/tlpkg/tlpsrc/l3kernel.tlpsrc b/Master/tlpkg/tlpsrc/l3kernel.tlpsrc
new file mode 100644
index 00000000000..e69de29bb2d
--- /dev/null
+++ b/Master/tlpkg/tlpsrc/l3kernel.tlpsrc