summaryrefslogtreecommitdiff
path: root/Master/texmf-dist/source/latex/revtex4
diff options
context:
space:
mode:
authorNorbert Preining <preining@logic.at>2009-12-23 18:57:23 +0000
committerNorbert Preining <preining@logic.at>2009-12-23 18:57:23 +0000
commit954a951fc174a8b1b5e1ce335a3d0fa8ae77a25b (patch)
tree8d87babe00944a8357cb3309bb2e00fd5e00e0ba /Master/texmf-dist/source/latex/revtex4
parent8f0d30f2d461ba69f47843ce84fd1a0732cc7066 (diff)
add revtex4 package that ships the old revtex version
taken from TL 2007 (the only place where I could find it for now) git-svn-id: svn://tug.org/texlive/trunk@16488 c570f23f-e606-0410-a88d-b1316a301751
Diffstat (limited to 'Master/texmf-dist/source/latex/revtex4')
-rw-r--r--Master/texmf-dist/source/latex/revtex4/ltxdocext.dtx1208
-rw-r--r--Master/texmf-dist/source/latex/revtex4/ltxgrid.dtx4815
-rw-r--r--Master/texmf-dist/source/latex/revtex4/ltxutil.dtx3534
-rw-r--r--Master/texmf-dist/source/latex/revtex4/revtex4.dtx8728
-rw-r--r--Master/texmf-dist/source/latex/revtex4/revtex4.ins36
-rw-r--r--Master/texmf-dist/source/latex/revtex4/textcase.dtx404
6 files changed, 18725 insertions, 0 deletions
diff --git a/Master/texmf-dist/source/latex/revtex4/ltxdocext.dtx b/Master/texmf-dist/source/latex/revtex4/ltxdocext.dtx
new file mode 100644
index 00000000000..ee9d71c9c04
--- /dev/null
+++ b/Master/texmf-dist/source/latex/revtex4/ltxdocext.dtx
@@ -0,0 +1,1208 @@
+% \iffalse &pdflatex ltxdoc klootch
+% ltxdocext.dtx: package to extend the ltxdoc class of standard LaTeX
+% Copyright (c) 1999 Arthur Ogawa
+%
+% Disclaimer
+% This file is distributed WITHOUT ANY WARRANTY;
+% without even the implied warranty of
+% MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
+% ReadMe
+% For the documentation and more detailed instructions for
+% installation, typeset this document with \LaTeX.
+% \fi
+% \CheckSum{938}
+%% \CharacterTable
+%% {Upper-case \A\B\C\D\E\F\G\H\I\J\K\L\M\N\O\P\Q\R\S\T\U\V\W\X\Y\Z
+%% Lower-case \a\b\c\d\e\f\g\h\i\j\k\l\m\n\o\p\q\r\s\t\u\v\w\x\y\z
+%% Digits \0\1\2\3\4\5\6\7\8\9
+%% Exclamation \! Double quote \" Hash (number) \#
+%% Dollar \$ Percent \% Ampersand \&
+%% Acute accent \' Left paren \( Right paren \)
+%% Asterisk \* Plus \+ Comma \,
+%% Minus \- Point \. Solidus \/
+%% Colon \: Semicolon \; Less than \<
+%% Equals \= Greater than \> Question mark \?
+%% Commercial at \@ Left bracket \[ Backslash \\
+%% Right bracket \] Circumflex \^ Underscore \_
+%% Grave accent \` Left brace \{ Vertical bar \|
+%% Right brace \} Tilde \~}%
+%
+% \iffalse ltxdoc klootch
+%%% @LaTeX-file{
+%%% filename = "ltxdocext.dtx",
+%%% version = "0.0a0",
+%%% date = "1999/06/04",
+%%% time = "11:23:00 GMT",
+%%% checksum = "929",
+%%% author = "Arthur Ogawa (mailto:ogawa@teleport.com)",
+%%% copyright = "Copyright (C) 1999 Arthur Ogawa,
+%%% all rights reserved. Copying of this file is
+%%% authorized only if either:
+%%% (1) you make absolutely no changes to your copy,
+%%% including name; OR
+%%% (2) if you do make changes, you first rename it
+%%% to some other name.",
+%%% address = "Arthur Ogawa,
+%%% USA",
+%%% telephone = "",
+%%% FAX = "",
+%%% email = "ogawa@teleport.com",
+%%% codetable = "ISO/ASCII",
+%%% keywords = "latex, ltxdoc",
+%%% supported = "yes",
+%%% abstract = "extensions to the ltxdoc class",
+%%% docstring = "The checksum field above generated by ltxdoc",
+%%% }
+% \fi
+%
+% \iffalse ltxdoc klootch
+% The following references the \file{00readme.txt} file,
+% which contains basic information about this package.
+% The contents of this file are generated when
+% you typeset the programmer's documentation.
+% Search on "{filecontents*}{00readme.txt}" to locate it.
+% \fi\input{00readme.txt}%
+%
+% \subsection{Bill of Materials}
+%
+% Following is a list of the files in this distribution arranged
+% according to provenance.
+%
+% \subsubsection{Primary Source}%
+% One single file generates all.
+%\begin{verbatim}
+%ltxdocext.dtx
+%\end{verbatim}
+%
+% \subsubsection{Generated by \texttt{latex ltxdocext.dtx}}%
+% Typesetting the source file under \LaTeX\
+% generates the readme and the installer.
+%\begin{verbatim}
+%00readme.txt ltxdocext.ins
+%\end{verbatim}
+%
+% \subsubsection{Generated by \texttt{tex ltxdocext.ins}}%
+% Typesetting the installer generates
+% the package files.
+%\begin{verbatim}
+%ltxdocext.sty acrofont.sty
+%\end{verbatim}
+%
+% \subsubsection{Documentation}%
+% The following are the online documentation:
+% \begin{verbatim}
+%ltxdocext.pdf
+% \end{verbatim}
+%
+% \subsubsection{Auxiliary}%
+% The following are auxiliary files generated
+% in the course of running \LaTeX:
+% \begin{verbatim}
+%ltxdocext.aux ltxdocext.idx ltxdocext.ind ltxdocext.log ltxdocext.toc
+% \end{verbatim}
+%
+% \section{Code common to all modules}%
+%
+% The following may look a bit klootchy, but we
+% want to require only one place in this file
+% where the version number is stated,
+% and we also want to ensure that the version
+% number is embedded into every generated file.
+%
+% Now we declare that
+% these files can only be used with \LaTeXe.
+% An appropriate message is displayed if
+% a different \TeX{} format is used.
+% \begin{macrocode}
+%<*doc|extensions|fonts>
+\NeedsTeXFormat{LaTeX2e}[1995/12/01]%
+%</doc|extensions|fonts>
+% \end{macrocode}
+% As desired, the following modules all
+% take common version information:
+% \begin{macrocode}
+%<extensions>\ProvidesFile{ltxdocext.sty}%
+%<fonts>\ProvidesFile{acrofont.sty}%
+%<*doc>
+\expandafter\ProvidesFile\expandafter{\jobname.dtx}%
+%</doc>
+% \end{macrocode}
+%
+% The following line contains, for once and for all,
+% the version and date information.
+% By various means, this information is reproduced
+% consistently in all generated files and in the
+% typeset documentation.
+% \begin{macrocode}
+ [1999/06/04 0.0a0 ltxdoc extensions package]% \fileversion
+% \end{macrocode}
+%
+%
+% \section{The driver module \texttt{doc}}
+%
+% This module, consisting of the present section,
+% typesets the programmer's documentation,
+% generating the \file{.ins} installer and \file{00readme.txt} as required.
+%
+% Because the only uncommented-out lines of code at the beginning of
+% this file constitute the \file{doc} module itself,
+% we can simply typeset the \file{.dtx} file directly,
+% and there is thus rarely any need to
+% generate the ``doc'' {\sc docstrip} module.
+% Module delimiters are nonetheless required so that
+% this code does not find its way into the other modules.
+%
+% The \enve{document} command concludes the typesetting run.
+%
+% \begin{macrocode}
+%<*doc>
+% \end{macrocode}
+%
+% \subsection{Stubs for \file{ltxdocext.sty} and \file{acrofont.sty}}
+%
+% This \env{filecontents} environment generates stub files
+% \file{ltxdocext.sty} and \file{acrofont.sty}.
+% Stubs are needed because the documentation for this package
+% requires them and they might not yet be present.
+%
+% A real version of these files will be generated
+% when you typeset \file{ltxdocext.ins}, at which point
+% the stub file will be overwritten (you will have to give {\sc docstrip}
+% permission to do this).
+% \begin{macrocode}
+\begin{filecontents}{ltxdocext.sty}
+% Stub version of ltxdocext.sty
+\end{filecontents}
+\begin{filecontents}{acrofont.sty}
+% Stub version of acrofont.sty
+\end{filecontents}
+% \end{macrocode}
+%
+% \subsection{The Preamble}
+% The programmers documentation is formatted
+% with the \classname{ltxdoc} class with local customizations,
+% and with the usual code line indexing.
+% \begin{macrocode}
+\documentclass[draft]{ltxdoc}
+\RequirePackage{ltxdocext}%
+\RequirePackage{acrofont}%
+\CodelineIndex\EnableCrossrefs
+% \end{macrocode}
+%
+% \subsubsection{Docstrip and info directives}
+% We use so many {\sc docstrip} modules that we set the
+% \texttt{StandardModuleDepth} counter to 1.
+% \begin{macrocode}
+\setcounter{StandardModuleDepth}{1}
+% \end{macrocode}
+% The following command retrieves the date and version information
+% from this file.
+% \begin{macrocode}
+\expandafter\GetFileInfo\expandafter{\jobname.dtx}%
+% \end{macrocode}
+%
+%
+% \subsection{The installer file}
+%
+% The installer \file{ltxdocext.ins} appears here.
+% If you have retrieved the standard distribution of this package,
+% the installer file is already on your filesystem.
+% If you are bootstrapping,
+% the first typesetting of the \file{.dtx} file
+% will cause the installer to be generated.
+%
+% The following modules are used to direct
+% {\sc docstrip} in generating the external files:
+% \begin{center}
+% \begin{tabular}{lll}
+% \textbf{Module}&\textbf{File}&\textbf{Description}\\
+% doc &\file{ltxdocext.drv}&driver for programmer's documantation\\
+% extensions&\file{ltxdocext.sty}&ltxdoc extensions package\\
+% fonts &\file{acrofont.sty} &package to use only Acrobat fonts
+% \end{tabular}
+% \end{center}
+%
+% \begin{macrocode}
+\begin{filecontents}{ltxdocext.ins}
+%% This file will generate documentation and runtime files
+%% from ltxdocext.dtx when run through LaTeX or TeX.
+\input docstrip
+\preamble
+
+This is a generated file;
+altering it directly is inadvisable;
+instead, modify the original source file.
+See the URL in the file 00readme.txt.
+
+Copyright notice.
+
+ These files are distributed
+ WITHOUT ANY WARRANTY; without even the implied warranty of
+ MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
+
+\endpreamble
+\keepsilent
+ \generate{%
+ \file{ltxdocext.drv}{\from{ltxdocext.dtx}{doc}}%
+ \file{ltxdocext.sty}{\from{ltxdocext.dtx}{extensions}}%
+ \file{acrofont.sty}{\from{ltxdocext.dtx}{fonts}}%
+ }%
+\ifToplevel{
+\Msg{***********************************************************}
+\Msg{*}
+\Msg{* To finish the installation, please move}
+\Msg{* ltxdocext.sty}
+\Msg{* into a directory searched by TeX.}
+\Msg{*}
+\Msg{* To produce the user documentation,
+ run ltxdocext.tex through LaTeX.}
+\Msg{*}
+\Msg{* To produce the programmer documentation,
+ run ltxdocext.dtx through LaTeX.}
+\Msg{*}
+\Msg{* Happy TeXing}
+\Msg{***********************************************************}
+}
+\endbatchfile
+\end{filecontents}
+% \end{macrocode}
+% Note that, because all of the files generated by the installer
+% are part of the standard distribution, it will
+% be necessary to run the installer only when bootstrapping
+% (or, of course, during development).
+% Note, too, that it is rare to generate the \file{doc}
+% module because it suffices to simply typeset the \file{.dtx} file itself.
+%
+% \subsection{The ``Read Me'' File}
+% As promised above, here is the contents of the
+% ``Read Me'' file. That file serves a double purpose,
+% since it also constitutes the beginining of the
+% programmer's documentation. What better thing, after
+% all, to have appear at the beginning of the
+% typeset documentation?
+%
+% A good discussion of how to write a ReadMe file can be found in
+% Engst, Tonya, ``Writing a ReadMe File? Read This''
+% \emph{MacTech} October 1998, p. 58.
+%
+% Note the appearance of the
+% \cmd\StopEventually\ command, which marks the
+% dividing line between the user documentation
+% and the programmer documentation.
+%
+% The usual user will not be asked to
+% do a full build, not to speak
+% of the bootstrap.
+% Instructions for carrying these processes
+% begin the programmer's manual.
+%
+% \begin{macrocode}
+\begin{filecontents*}{00readme.txt}
+\title{%
+ Extensions to the \classname{ltxdoc} class%
+ \thanks{%
+ This file has version number \fileversion,
+ last revised \filedate.%
+ % For version number and date, search on "\fileversion" in the .dtx file,
+ % or see the end of the 00readme.txt file.
+ }%
+}%
+
+\author{%
+Arthur Ogawa (\texttt{mailto:ogawa@teleport.com}),
+\fileversion\\Copyright (C) 1999 Arthur Ogawa
+}%
+\maketitle
+
+This file embodies the \classname{ltxdocext} package,
+the implementation and its user documentation.
+
+The distribution point for this work is
+\url{ftp://ftp.teleport.com/users/ogawa/macros/latex/contrib/supported/ltxdocext...},
+which contains fully unpacked, prebuilt runtime files and documentation.
+
+To use this document class, you must have a working
+\TeX\ installation equipped with \LaTeXe\
+and possibly pdftex and Adobe Acrobat Reader or equivalent.
+
+To install, retrieve the distribution,
+unpack it into a directory on the target computer,
+and move the files \file{ltxdocext.sty} and \file{acrofont.sty}
+into a location in your filesystem where they will be found by \LaTeX.
+
+If you will be using the \classname{acrofont} package, you must
+also install the virtual fonts
+\file{zpsynocmrv}, \file{zptmnocmr},
+\file{zptmnocmrm}, and \file{zpzcnocmry}.
+The corresponding \file{.tfm}, \file{.vf}, and \file{.vpl}
+files are part of this distribution.
+
+To use, read the user documentation \file{ltxdocext.pdf}.
+The \file{.dtx} file, \file{ltxdocext.dtx}, constitutes
+in itself an instance of use of the \classname{ltxdocext}
+package and the \classname{acrofont} package.
+
+\tableofcontents
+
+\section{Processing Instructions}
+
+The package files \file{ltxdocext.sty} and \file{acrofont.sty}
+are generated from this file, \file{ltxdocext.dtx},
+via the {\sc docstrip} facility of \LaTeX
+via |tex ltxdocext.ins|.
+The typeset documentation that you are now reading is generated from
+the same file by typesetting it with \LaTeX\ or pdftex
+via |latex ltxdocext.dtx| or |pdflatex ltxdocext.dtx|.
+
+\subsection{Build Instructions}
+
+You may bootstrap this suite of files solely from \file{ltxdocext.dtx}.
+Prepare by installing \LaTeXe\ (and either tex or pdftex) on your computer,
+then carry out the following steps:
+\begin{enumerate}
+\item
+Within an otherwise empty directory,
+typeset \file{ltxdocext.dtx} with \LaTeX\ or pdflatex;
+you will obtain the typeset documentation you are now reading,
+along with
+the installer \file{ltxdocext.ins},
+and the file \file{00readme.txt}.
+\item
+Now typeset \file{ltxdocext.ins},
+thereby generating the package file \file{ltxdocext.sty},
+and the package file \file{acrofont.sty}.
+Make sure that {\sc docstrip} receives permission
+to overwrite existing versions of these packages.
+\item
+Install \classname{ltxdocext.sty} and \classname{acrofont.sty}
+by moving them to a location
+in your filesystem where they will be found by \LaTeX.
+\item
+Now complete the typesetting of the documentation by
+retypesetting \file{ltxdocext.dtx}.
+Note: you will have to run \LaTeX\ twice, then \file{makeindex}, then
+\LaTeX\ again in order to obtain a valid index and table of contents.
+\end{enumerate}
+\end{filecontents*}
+% \end{macrocode}
+%
+% \subsection{The Document Body}
+%
+% Here is the document body, containing only a
+% \cmd\DocInput\ directive---referring to this very file.
+% This very cute self-reference is a common \classname{ltxdoc} idiom.
+% \begin{macrocode}
+\begin{document}%
+\expandafter\DocInput\expandafter{\jobname.dtx}%
+% ^^A\PrintChanges
+\end{document}
+% \end{macrocode}
+%
+% \begin{macrocode}
+%</doc>
+% \end{macrocode}
+%
+% \section{Using the \classname{ltxdoc} and \classname{acrofont} packages}%
+% These packages are an adjunct to the
+% standard \LaTeX\ \classname{ltxdoc} class and may be
+% simply invoked as follows:
+%\begin{verbatim}
+%\documentclass[draft]{ltxdoc}
+%\RequirePackage{ltxdocext}%
+%\RequirePackage{acrofont}%
+%\CodelineIndex\EnableCrossrefs
+%\end{verbatim}
+%
+% Your document should simply cleave to the standards of the
+% \classname{ltxdoc} class, with extensions and alterations as noted.
+%
+% \subsection{Extensions to the \classname{ltxdoc} class}%
+%
+% \subsubsection{Extensions to the \env{verbatim} environment and \cs{verb} command}%
+%
+% The delimiters \verb$<<$ and \verb$>>$ within the scope of the verbatim environment
+% or within the argument of the \cmd\verb\ command produce
+% italics surrounded by angle brackets.
+% This typographic convention usually indicates
+% \emph{metalanguage}, i.e., a placeholder.
+%
+% To obtain the angle bracket character per se,
+% double the character, viz., ``\verb$the delimiter \verb+<<<<+$''.
+%
+% \subsubsection{The -\texttt{matter} Commands Work}
+% The sectioning commands \cmd\frontmatter, \cmd\mainmatter, and \cmd\backmatter
+% of the standard \LaTeX\ \classname{book} class are operative in the
+% \classname{ltxdoc} class.
+%
+% \subsubsection{The \cs{GetFileInfo} command}\label{sec:GetFileInfo}
+% You can use the \cmd\GetFileInfo\ command to extract
+% the date, version, and file info of a file which has registered itself
+% via the \cmd\ProvidesFile\ or \cmd\ProvidesClass\ command
+% (employing the optional argument thereto).
+%
+% For instance, if your document contains the following:
+%\begin{verbatim}
+%\RequirePackage{ltxdocext}%
+%\GetFileInfo{ltxdocext.sty}%
+%\end{verbatim}
+% then the following control sequence names will have
+% a value corresponding to that package's \cmd\ProvidesFile\
+% command:
+% \cmd\filedate: the file's date,
+% \cmd\fileversion: the file's version, and
+% \cmd\fileinfo: the file's info.
+%
+% \subsubsection{Self-Indexing Commands}%
+% Certain commands automatically produce an index entry
+% (or several related entries) according to the meaning.
+%
+% \begin{unnumtable}
+% \begin{tabular}{ll}
+% meta-text &\cmd\m\arg{text}\\
+% command &\cmd\cmd\cmd\csname\\
+% environment name &\cmd\env\arg{name}\\
+% \cmd\begin\verb`{foo}`&\cmd\envb\arg{foo}\\
+% \cmd\end\verb`{foo}` &\cmd\enve\arg{foo}\\
+% argument &\cmd\arg\arg{name}\\
+% optional &\cmd\oarg\arg{name}\\
+% filename &\cmd\file\arg{name}\\
+% url &\cmd\url\arg{name}\\
+% document class &\cmd\classname\arg{name}\\
+% document substyle &\cmd\substyle\arg{name}\\
+% class option &\cmd\classoption\arg{name}
+% \end{tabular}
+% \end{unnumtable}
+%
+% \subsubsection{Unnumbered Tables}%
+%
+% When your documentation requires the use of an unnumbered table,
+% use the \env{unnumtable} environment:
+% \begin{verbatim}
+%\begin{unnumtable}
+%\begin{tabular}{ll}
+%<table rows>
+%\end{tabular}
+%\end{unnumtable}
+% \end{verbatim}
+%
+% \subsubsection{Structuring Tables}%
+% The commands \cmd\toprule, \cmd\colrule, and \cmd\botrule
+% allow you to mark the beginning of the column heads
+% the beginning of the table body, and the end of
+% the table body, respectively.
+% In context,
+% \begin{verbatim}
+%\begin{tabular}{ll}
+%\toprule
+%<table head rows>
+%\colrule
+%<table rows>
+%\botrule
+%\end{tabular}
+% \end{verbatim}
+%
+% \subsubsection{A Sectioning Command Below \cs{subsection}}%
+% The \cmd\subsubsection\ command is defined.
+%
+%
+% \subsection{Alterations to the \classname{ltxdoc} class}%
+% The following involve no new markup, but they
+% do change the appearance of your formatted documentation:
+% \begin{enumerate}
+% \item
+% Using the \classname{acrofont} package causes your
+% document to be formatted using the standard
+% Acrobat fonts to the greatest extent possible.
+% This means that for most documents, Computer Modern
+% is not used at all.
+% Math that unavoidable must use CM still exists, however.
+% \item
+% An index will be produced at the end of the document
+% without your needing to explicitly mark it up, and
+% it will have an entry in the TOC.
+% \item
+% The \env{quote} environment has a slightly smaller left margin.
+% \item
+% Array columns are set tight by default.
+% \item
+% A host of \cmd\DoNotIndex\ directives are invoked.
+% I intend this list to grow to encompass
+% even more commands. Send me your suggestions.
+% \end{enumerate}
+%
+%
+%
+%\StopEventually{}
+%
+% \section{Extensions to the ltxdoc class}
+% The \file{extensions} {\sc docstrip} module comprises the
+% package \classname{ltxdocext.sty}, which provides extensions
+% to the standard \LaTeX\ \classname{ltxdoc} class.
+%
+% \subsection{Beginning of the \file{extensions} {\sc docstrip} module}
+% \begin{macrocode}
+%<*extensions>
+\def\class@name{ltxdocext}%
+\expandafter\PackageInfo\expandafter{\class@name}{%
+ An extension to the \protect\LaTeXe\space ltxdoc class
+ by A. Ogawa (ogawa@teleport.com)%
+}%
+% \end{macrocode}
+%
+% \subsection{Incorporate \classname{ltxguide.cls} extensions}
+%
+% Code extracted from \classname{ltxguide.cls}, by Alan Jeffrey.
+% ``This code stolen from \classname{ltxguide.cls}:
+% Some hacks with verbatim... NB: this would be better done with the
+% verbatim package, but this document has to run on any \LaTeX
+% installation.''
+% \begin{macrocode}
+\RequirePackage{verbatim}%
+\let\o@verbatim\verbatim
+\def\verbatim{%
+ \ifhmode\unskip\par\fi
+% \nopagebreak % Overridden by list penalty
+ \ifx\@currsize\normalsize
+ \small
+ \fi
+ \o@verbatim
+}%
+% \end{macrocode}
+%
+% Here we extend the font-setting command to include making \texttt{<>} active
+% (i.e., adjusting the input encoding).
+% \begin{macrocode}
+\renewcommand \verbatim@font {%
+ \normalfont \ttfamily
+ \catcode`\<=\active
+ \catcode`\>=\active
+}%
+% \end{macrocode}
+%
+% Make \verb$|...|$ a synonym for \cmd\verb\verb$|...|$.
+% \begin{macrocode}
+\RequirePackage{shortvrb}
+\AtBeginDocument{%
+ \MakeShortVerb{\|}%
+}%
+% \end{macrocode}
+%
+% Make active bracket characters produce italics surrounded by angle brackets
+% (used in \env{verbatim} and \cmd\verb).
+% \verb|<<| produces a less-than, and \verb|>>| produces a greater-than.
+% \begin{macrocode}
+\begingroup
+ \catcode`\<=\active
+ \catcode`\>=\active
+ \gdef<{\@ifnextchar<\@lt\@meta}
+ \gdef>{\@ifnextchar>\@gt\@gtr@err}
+ \gdef\@meta#1>{\m{#1}}
+ \gdef\@lt<{\char`\<}
+ \gdef\@gt>{\char`\>}
+\endgroup
+\def\@gtr@err{%
+ \ClassError{ltxguide}{%
+ Isolated \protect>%
+ }{%
+ In this document class, \protect<...\protect>
+ is used to indicate a parameter.\MessageBreak
+ I've just found a \protect> on its own.
+ Perhaps you meant to type \protect>\protect>?
+ }%
+}
+\def\verbatim@nolig@list{\do\`\do\,\do\'\do\-}
+% \end{macrocode}
+% End of code stolen from \file{ltxguide.cls}. Thanks, Alan.
+%
+% Add functionality from doc.dtx:
+% (code stolen from doc.dtx):
+% \begin{macrocode}
+\def\GetFileInfo#1{%
+ \def\filename{#1}%
+ \def\@tempb##1 ##2 ##3\relax##4\relax{%
+ \def\filedate{##1}%
+ \def\fileversion{##2}%
+ \def\fileinfo{##3}}%
+ \edef\@tempa{\csname ver@#1\endcsname}%
+ \expandafter\@tempb\@tempa\relax? ? \relax\relax}
+% \end{macrocode}
+% (end of code stolen from doc.dtx. Thanks FMi.)
+%
+% Various forms of self-indexing commands:
+% \begin{macrocode}
+\DeclareRobustCommand{\m}[1]{%
+ \meta{#1}%
+ \index{#1=\string\meta{#1} placeholder}\index{placeholder>#1=\string\meta{#1}}%
+}%
+\DeclareRobustCommand\meta[1]{%
+ \mbox{\LANGLE\itshape#1\/\RANGLE}%
+}%
+\def\LANGLE{$\langle$}%
+\def\RANGLE{$\rangle$}%
+\DeclareRobustCommand{\arg}[1]{%
+ {\ttfamily\string{}\meta{#1}{\ttfamily\string}}%
+ \index{#1=\string\ttt{#1}, argument}\index{argument>#1=\string\ttt{#1}}%
+}%
+\let\oarg\undefined
+\DeclareRobustCommand{\oarg}[1]{%
+ {\ttfamily[%]
+ }\meta{#1}{\ttfamily%[
+ ]}%
+ \index{#1=\string\ttt{#1}, optional argument}%
+ \index{argument, optional>#1=\string\ttt{#1}}%
+}%
+\DeclareRobustCommand\cmd{\begingroup\makeatletter\@cmd}%
+\def\@cmd#1{%
+ \endgroup
+ \cs{\expandafter\cmd@to@cs\string#1}%
+ \expandafter\cmd@to@index\string#1\@nil
+}%
+\def\cmd@to@cs#1#2{\char\number`#2\relax}%
+\def\cmd@to@index#1#2\@nil{%
+ \index{#2=\string\cmd#1#2}%\index{command>#2=\string\cmd#1#2}%
+}%
+\DeclareRobustCommand\cs[1]{{\ttfamily\char`\\#1}}%
+\def\scmd#1{%
+ \cs{\expandafter\cmd@to@cs\string#1}%
+ \expandafter\scmd@to@index\string#1\@nil
+}%
+\def\scmd@to@index#1#2\@nil#3{%
+ \index{\string$#3=\string\cmd#1#2---#3}%
+%\index{command>\string$#3=\string\cmd#1#2---#3}%
+}%
+\DeclareRobustCommand\env{\name@idx{environment}}%
+\DeclareRobustCommand\envb[1]{%
+ {\ttfamily\string\begin\string{}\env{#1}{\ttfamily\string}}%
+}%
+\DeclareRobustCommand\enve[1]{{\ttfamily\string\end\string{}\env{#1}{\ttfamily\string}}}%
+\DeclareRobustCommand*{\file}[1]{%
+ {\ttfamily#1}%
+ \index{#1=\string\ttt{#1}}\index{file>#1=\string\ttt{#1}}%
+}%
+\DeclareRobustCommand\substyle{\name@idx{document substyle}}%
+\DeclareRobustCommand\classoption{\name@idx{document class option}}%
+\DeclareRobustCommand\classname{\name@idx{document class}}%
+\def\name@idx#1#2{%
+ {\ttfamily#2}%
+ \index{#2\space#1=\string\ttt{#2}\space#1}\index{#1>#2=\string\ttt{#2}}%
+}%
+\DeclareRobustCommand\url@ltxdocext{\begingroup\catcode`\/\active\catcode`\.\active\catcode`\:\active\@url}%
+\AtBeginDocument{%
+ \ifx\url\undefined\let\url\url@ltxdocext\fi
+}%
+\def\@url#1{%
+ \url@break{\ttfamily#1}%
+ \url@char\edef\@tempa{#1=\string\url{#1}}%
+ \expandafter\index\expandafter{\@tempa}%
+ \expandafter\index\expandafter{\expandafter u\expandafter r\expandafter l\expandafter >\@tempa}%
+ \endgroup
+}%
+{\catcode`\:\active\aftergroup\def\aftergroup:}{\active@colon}%
+\def\colon@break{\colon@char\allowbreak}%
+\def\colon@char{:}%
+{\catcode`\/\active\aftergroup\def\aftergroup/}{\active@slash}%
+\def\slash@break{\slash@char\allowbreak}%
+\def\slash@char{/}%
+{\catcode`\.\active\aftergroup\def\aftergroup.}{\active@dot}%
+\def\dot@break{\dot@char\allowbreak}%
+\def\dot@char{.}%
+\def\url@break{\let\active@slash\slash@break\let\active@dot\dot@break\let\active@colon\colon@break}%
+\def\url@char{\let\active@slash\slash@char\let\active@dot\dot@char\let\active@colon\colon@char}%
+% \end{macrocode}
+%
+% \subsection{Changes to the base class of the ltxdoc class}
+% Modify \env{theindex} environment so that it produces a TOC entry
+% \begin{macrocode}
+\renewenvironment{theindex}
+ {\if@twocolumn
+ \@restonecolfalse
+ \else
+ \@restonecoltrue
+ \fi
+ \columnseprule \z@
+ \columnsep 35\p@
+\def\see##1##2{\textit{See} ##1}%
+\def\seealso##1##2{\textit{See also} ##1}%
+\def\cmd##1{\cs{\expandafter\cmd@to@cs\string##1}}%
+\def\@url##1{\url@break\ttt{##1}\endgroup}%
+\def\ttt##1{{\ttfamily##1}}%
+\mathchardef\save@secnumdepth\c@secnumdepth
+\c@secnumdepth\m@ne
+ \twocolumn[\section{\indexname}]%
+% \@mkboth{\MakeUppercase\indexname}%
+% {\MakeUppercase\indexname}%
+\c@secnumdepth\save@secnumdepth
+ \thispagestyle{plain}\parindent\z@
+ \parskip\z@ \@plus .3\p@\relax
+ \let\item\@idxitem}
+ {\if@restonecol\onecolumn\else\clearpage\fi}
+\renewenvironment{quote}
+ {\list{}{%
+ \leftmargin1em\relax
+ \rightmargin\leftmargin
+ }%
+ \item\relax}
+ {\endlist}
+% \end{macrocode}
+%
+% \subsection{Extensions to the base class of \classname{ltxdoc.cls}}
+%
+% Matter commands from \classname{book.cls}
+% \begin{macrocode}
+\newif\if@mainmatter
+\newif\if@openright
+\@openrighttrue
+\DeclareRobustCommand\frontmatter{%
+ \cleartorecto
+ \@mainmatterfalse
+ \pagenumbering{roman}%
+}%
+\DeclareRobustCommand\mainmatter{%
+ \cleartorecto
+ \@mainmattertrue
+ \pagenumbering{arabic}%
+}%
+\DeclareRobustCommand\backmatter{%
+ \if@openright
+ \cleartorecto
+ \else
+ \clearpage
+ \fi
+ \@mainmatterfalse
+}%
+\ifx\undefined\cleartorecto
+ \def\cleartorecto{\cleardoublepage}%
+\fi
+% \end{macrocode}
+%
+% Unnumbered tables
+%
+% \begin{environment}{unnumtable}
+% An unnumbered table does not float.
+% \begin{macrocode}
+\def\@to{to}%
+\newenvironment{unnumtable}{%
+ \par
+ \addpenalty\predisplaypenalty
+ \addvspace\abovedisplayskip
+ \hbox\@to\hsize\bgroup\hfil\ignorespaces
+ \let\@Hline\@empty
+}{%
+ \unskip\hfil\egroup
+ \penalty\postdisplaypenalty
+ \vskip\belowdisplayskip
+ \aftergroup\ignorespaces
+ \@endpetrue
+}%
+% \end{macrocode}
+% \end{environment}
+%
+% Emulate \cmd\toprule\ and friends
+% \begin{macrocode}
+\newcommand\toprule{\hline\hline}%
+\newcommand\colrule{\\\hline}%
+\newcommand\botrule{\\\hline\hline}%
+% \end{macrocode}
+%
+% Define sectioning command below \cmd\subsubsection.
+% \begin{macrocode}
+\DeclareRobustCommand\subsubsubsection{%
+ \@startsection{subsubsection}{4}%
+ {\z@}{-15\p@\@plus-5\p@\@minus-2\p@}%
+ {5\p@}{\normalfont\normalsize\itshape}%
+}%
+% \end{macrocode}
+%
+% \subsection{In lieu of \file{ltxdoc.cfg}}
+% We don't want everything to appear in the index
+% \begin{macrocode}
+\DoNotIndex{\',\.,\@M,\@@input,\@Alph,\@alph,\@addtoreset,\@arabic}
+\DoNotIndex{\@badmath,\@centercr,\@cite}
+\DoNotIndex{\@dotsep,\@empty,\@float,\@gobble,\@gobbletwo,\@ignoretrue}
+\DoNotIndex{\@input,\@ixpt,\@m,\@minus,\@mkboth}
+\DoNotIndex{\@ne,\@nil,\@nomath,\@plus,\roman,\@set@topoint}
+\DoNotIndex{\@tempboxa,\@tempcnta,\@tempdima,\@tempdimb}
+\DoNotIndex{\@tempswafalse,\@tempswatrue,\@viipt,\@viiipt,\@vipt}
+\DoNotIndex{\@vpt,\@warning,\@xiipt,\@xipt,\@xivpt,\@xpt,\@xviipt}
+\DoNotIndex{\@xxpt,\@xxvpt,\\,\ ,\addpenalty,\addtolength,\addvspace}
+\DoNotIndex{\advance,\ast,\begin,\begingroup,\bfseries,\bgroup,\box}
+\DoNotIndex{\bullet}
+\DoNotIndex{\cdot,\cite,\CodelineIndex,\cr,\day,\DeclareOption}
+\DoNotIndex{\def,\DisableCrossrefs,\divide,\DocInput,\documentclass}
+\DoNotIndex{\DoNotIndex,\egroup,\ifdim,\else,\fi,\em,\endtrivlist}
+\DoNotIndex{\EnableCrossrefs,\end,\end@dblfloat,\end@float,\endgroup}
+\DoNotIndex{\endlist,\everycr,\everypar,\ExecuteOptions,\expandafter}
+\DoNotIndex{\fbox}
+\DoNotIndex{\filedate,\filename,\fileversion,\fontsize,\framebox,\gdef}
+\DoNotIndex{\global,\halign,\hangindent,\hbox,\hfil,\hfill,\hrule}
+\DoNotIndex{\hsize,\hskip,\hspace,\hss,\if@tempswa,\ifcase,\or,\fi,\fi}
+\DoNotIndex{\ifhmode,\ifvmode,\ifnum,\iftrue,\ifx,\fi,\fi,\fi,\fi,\fi}
+\DoNotIndex{\input}
+\DoNotIndex{\jobname,\kern,\leavevmode,\let,\leftmark}
+\DoNotIndex{\list,\llap,\long,\m@ne,\m@th,\mark,\markboth,\markright}
+\DoNotIndex{\month,\newcommand,\newcounter,\newenvironment}
+\DoNotIndex{\NeedsTeXFormat,\newdimen}
+\DoNotIndex{\newlength,\newpage,\nobreak,\noindent,\null,\number}
+\DoNotIndex{\numberline,\OldMakeindex,\OnlyDescription,\p@}
+\DoNotIndex{\pagestyle,\par,\paragraph,\paragraphmark,\parfillskip}
+\DoNotIndex{\penalty,\PrintChanges,\PrintIndex,\ProcessOptions}
+\DoNotIndex{\protect,\ProvidesClass,\raggedbottom,\raggedright}
+\DoNotIndex{\refstepcounter,\relax,\renewcommand}
+\DoNotIndex{\rightmargin,\rightmark,\rightskip,\rlap,\rmfamily}
+\DoNotIndex{\secdef,\selectfont,\setbox,\setcounter,\setlength}
+\DoNotIndex{\settowidth,\sfcode,\skip,\sloppy,\slshape,\space}
+\DoNotIndex{\symbol,\the,\trivlist,\typeout,\tw@,\undefined,\uppercase}
+\DoNotIndex{\usecounter,\usefont,\usepackage,\vfil,\vfill,\viiipt}
+\DoNotIndex{\viipt,\vipt,\vskip,\vspace}
+\DoNotIndex{\wd,\xiipt,\year,\z@}
+\DoNotIndex{\next}
+% \end{macrocode}
+%
+% Direct \classname{ltxdoc} to produce an index.
+% \begin{macrocode}
+\AtEndDocument{\PrintIndex}%
+% \end{macrocode}
+%
+% \subsection{Extension to \LaTeX's \env{filecontents} Environment}
+% We want to
+% coax the version number into \env{filecontents}-generated files.
+% Note that we expect \cmd\fileversion\ and \cmd\filedate\ to
+% hold the needed information. For this to be the case,
+% your document should execute the \cmd\GetFileInfo\ command
+% (as documented in section~\ref{sec:GetFileInfo}) before
+% any instances of \env{filecontents}.
+% \begin{macrocode}
+\makeatletter
+\def\endfilecontents{%
+ \immediate\write\reserved@c{%
+ \string\iffalse\space ltxdoc klootch^^J%
+ \ifx\undefined\fileversion\else
+ \ifx\undefined\filedate\else
+ This file has version number \fileversion, last revised \filedate.%
+ \fi\fi
+ \string\fi
+ }%
+ \immediate\closeout\reserved@c
+ \def\T##1##2##3{%
+ \ifx##1\@undefined\else
+ \@latex@warning@no@line{##2 has been converted to Blank ##3e}%
+ \fi
+ }%
+ \T\L{Form Feed}{Lin}%
+ \T\I{Tab}{Spac}%
+ \immediate\write\@unused{}%
+}%
+\expandafter\let\csname endfilecontents*\endcsname\endfilecontents
+\makeatother
+% \end{macrocode}
+%
+% Alter formatting in arrays; set them tight.
+% \begin{macrocode}
+\setlength\arraycolsep{0pt}%
+% \end{macrocode}
+%
+% \subsection{End of the \file{extensions} {\sc docstrip} module}
+% Here ends the module.
+% \begin{macrocode}
+%</extensions>
+% \end{macrocode}
+%
+%
+% \section{Font Package for Acrobat Compatability}%
+% The package \classname{acrofont}
+% substitutes Acrobat-standard fonts for Computer Modern where possible,
+% even in math mode.
+% Documents typeset with this package in effect will require as little
+% downloaded font data as possible, but will not be exemplars of
+% fine math typesetting.
+%
+% \subsection{Beginning of the \file{fonts} {\sc docstrip} module}
+% The document class module comprises this and the next
+% four sections.
+% \begin{macro}{\class@base}
+% We define in exactly one spot the base class.
+% Typically that class will be one of \classname{book},
+% \classname{article}, or \classname{report}.
+% The base class effectively defines the use and the markup scheme
+% of the class of documents to be handled by this class.
+%
+% This class is a variant of the standard \LaTeX\ book class:
+% \url{ftp://ctan.tug.org/tex-archive/macros/latex/unpacked}.
+% \begin{macrocode}
+%<*fonts>
+\def\class@name{ltxdocext}%
+\expandafter\ClassInfo\expandafter{\class@name}{%
+ Written for \protect\LaTeXe\space
+ by A. Ogawa (ogawa@teleport.com)%
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \subsection{Variants on psfonts packages}%
+% The following uses \file{times.sty} from \url{/packages/psnfss/psfonts.dtx}
+% \begin{macrocode}
+\RequirePackage{times}%
+% \end{macrocode}
+%
+% The following uses \file{mathptm.sty} from \url{/packages/psnfss/psfonts.dtx}
+% \begin{macrocode}
+\RequirePackage{mathptm}%
+% \end{macrocode}
+%
+% The following is a customization of \file{ot1ptmcm.fd}.
+% The virtual font referred to here \file{zptmnocmr} is
+% a variant of Sebastian Rahtz's \file{zptmcmr}, but with
+% even more glyphs moved from CM to Acrobat-standard fonts.
+% \begin{macrocode}
+\DeclareFontFamily{OT1}{ptmcm}{}
+\DeclareFontShape{OT1}{ptmcm}{m}{n}{
+ <-> zptmnocmr
+}{}
+\DeclareFontShape{OT1}{ptmcm}{l}{n}{<->ssub * ptmnocm/m/n}{}
+% \end{macrocode}
+%
+% The following is a customization of \file{omlptmcm.fd}
+% The virtual font referred to here \file{zptmnocmrm} is
+% a variant of Sebastian Rahtz's \file{zptmcmrm}, but with
+% even more glyphs moved from CM to Acrobat-standard fonts.
+% \begin{macrocode}
+\DeclareFontFamily{OML}{ptmcm}{\skewchar \font =127}
+\DeclareFontShape{OML}{ptmcm}{m}{it}{
+ <-> zptmnocmrm
+}{}
+\DeclareFontShape{OML}{ptmcm}{l}{it}{<->ssub * ptmcm/m/it}{}
+\DeclareFontShape{OML}{ptmcm}{m}{sl}{<->ssub * ptmcm/m/it}{}
+\DeclareFontShape{OML}{ptmcm}{l}{sl}{<->ssub * ptmcm/m/sl}{}
+% \end{macrocode}
+%
+% The following is a customization of \file{omspzccm.fd}
+% The virtual font referred to here \file{zpzcnocmry} is
+% a variant of Sebastian Rahtz's \file{zpzccmry}, but with
+% even more glyphs moved from CM to Acrobat-standard fonts.
+% \begin{macrocode}
+\DeclareFontFamily{OMS}{pzccm}{}
+\DeclareFontShape{OMS}{pzccm}{m}{n}{
+ <-> zpzcnocmry
+}{}% cmsy10 Symbol Zapf Chancery Medium-Italic Times-Roman
+\DeclareFontShape{OMS}{pzccm}{l}{n}{<->ssub * pzccm/m/n}{}
+% \end{macrocode}
+%
+% The following is a customization of \file{omxpsycm.fd}
+% The virtual font referred to here \file{zpsynocmrv} is
+% a variant of Sebastian Rahtz's \file{zpsycmrv}, but with
+% even more glyphs moved from CM to Acrobat-standard fonts.
+% \begin{macrocode}
+\DeclareFontFamily{OMX}{psycm}{}
+\DeclareFontShape{OMX}{psycm}{m}{n}{
+ <-> zpsynocmrv
+}{}
+\DeclareFontShape{OMX}{psycm}{l}{n}{<->ssub * psycm/m/n}{}
+%
+\DeclareFontEncoding{8r}{}{}% from file: 8renc.def
+\DeclareFontFamily{8r}{cmr}{\hyphenchar\font45 }% from file: 8rcmr.fd
+\DeclareFontShape{8r}{cmr}{m}{n}{
+ <-> cmr10
+}{}
+% \end{macrocode}
+%
+% \subsection{Font definition files}%
+%
+% The following forces \LaTeX\ to do now what it would do anyway:
+% load the `font definition' information for the fonts that we
+% use. In this way, we prepare for faster processing through
+% the \cmd\dump\ of a preformatted macro package that will not
+% need to read in any packages or font definitions from disk.
+% \begin{macrocode}
+\input{8rphv.fd}%
+\input{8rptm.fd}%
+\input{ot1phv.fd}%
+\input{ot1ptm.fd}%
+\input{t1ptm.fd}%
+% \end{macrocode}
+%
+% \subsection{More math substitutions}%
+%
+% The following definitions arrange to get certain glyphs from the
+% text font instead of out of math pi fonts.
+% In particular, the copyright and registered symbols
+% are single glyphs instead of composites involving
+% the big circle from the \file{cmsy} font.
+% \begin{macrocode}
+ \def\eightRChar#1{{\def\encodingdefault{8r}\fontencoding\encodingdefault\selectfont\char"#1}}%
+ \def\LANGLE{$<$}%{\eightRChar{8B}}%
+ \def\RANGLE{$>$}%{\eightRChar{9B}}%
+%\def\ASTER{\eightRChar{2A}}%
+%\def\DAGGER{\eightRChar{86}}%
+%\def\DDAGGER{\eightRChar{87}}%
+ \def\BULLET{\eightRChar{95}}%
+%\def\SECTION{\eightRChar{A7}}%
+%\def\PARAGRAPH{\eightRChar{B6}}%
+ \def\VERTBAR{\eightRChar{7C}}%
+ \def\COPYRIGHT{\eightRChar{A9}}%
+ \def\REGISTERED{\eightRChar{AE}}%
+% \end{macrocode}
+%
+% \begin{macrocode}
+ \def\textbar{\VERTBAR}%
+ \def\textbullet{\BULLET}%
+ \def\textcopyright{\COPYRIGHT}%
+ \def\textregistered{\REGISTERED}%
+% \end{macrocode}
+%
+% I have removed \cmd\ensuremath\ from the following definition, and
+% all commands like \cmd\mathsection have been converted to
+% e.g., \cmd\textsection.
+% \begin{macrocode}
+\def\@makefnmark{\@thefnmark}%
+\def\@fnsymbol#1{{\ifcase#1\or *\or \dagger\or \ddagger\or
+ \textsection\or \textparagraph\or \|\or **\or \dagger\dagger
+ \or \ddagger\ddagger \else\@ctrerr\fi}}
+% \end{macrocode}
+%
+% \subsection{End of the \file{fonts} {\sc docstrip} module}
+% Here ends the module.
+% \begin{macrocode}
+%</fonts>
+% \end{macrocode}
+%
+%
+% \section{Programming Conventions}%
+% In writing the above code, I cleave to certain conventions, noted
+% here.
+% My goal in explaining them is to encourage others maintaining this
+% body of code to consider following them as well.
+% The benefits are twofold:
+% Some of the coding conventions aim to avoid programming pitfalls;
+% following them reduces maintenance costs.
+% Others make the code easier to read; following these eases the
+% process of understanding how the code works.
+%
+% And, for my part, I prefer to read code of this type.
+%
+% \subsection{Whitespace Conventions}%
+% Exactly where code lines break and indent, and where additional
+% whitespace is inserted is explained here.
+% \begin{itemize}
+% \item
+% Each new macro definition or register assignment begins a new line.
+% Therefore, \cmd\def, \cmd\newcommand, and their ilk will start in column
+% 1.
+% \item
+% Code is indented one space for each level of nesting within braces
+% \verb|{}|.
+% \item
+% Likewise, if possible, for \cmd\if\dots and matching \cmd\fi.
+% \item
+% However, the closing brace or \cmd\fi\ is outdented by one so that it
+% falls
+% at the same level of indentation as its matching brace or \cmd\if,
+% and it appears alone on its line.
+% \item
+% Use of the |tab| character is deprecated
+% (tabs are not standardized across all applications and operating
+% systems).
+% \item
+% Lines of code are limited to 72 characters.
+% I follow this convention mostly to ease the transmission of files
+% via email (a deprecated practice) and to accomodate people with
+% small monitors.
+% But \classname{ltxdoc} output looks better with the shorter lines, too.
+% \item
+% Extraneous whitespace is avoided by using the comment character |%|.
+% In most cases, if falling at the end of a line of code,
+% a brace will be immediately followed by a comment character,
+% as will the macro parameter |#1| and any one-letter control sequence,
+% like |\\|.
+% \end{itemize}
+% These conventions taken together are illustrated by the following:
+% \begin{verbatim}
+%\def\prepdef#1#2{%
+% \@ifxundefined#1{\toks@{}}{\toks@\expandafter{#1}}%
+% \toks@ii{#2}%
+% \edef#1{\the\toks@ii\the\toks@}%
+%}%
+% \end{verbatim}
+% In the above, the definition of \cmd\prepdef\ would not fit on a single
+% line,
+% and required breaking. The first and last lines have matching braces,
+% and are a the same level of indentation, with the last line containing
+% a single brace.
+%
+% Each of the three intervening lines has balanced braces and is
+% indented by one space. Each line that would otherwise end in a single
+% brace character is terminated by a comment character.
+%
+% Some coders rely on the fact that a space character seen by \TeX's
+% scanner while in vertical mode is a no-op.
+% Be that as it may, I eliminate them unless actually intentional.
+%
+% \subsection{Conventions For Procedures}%
+% Here are some of my preferences when writing procedures:
+% \begin{itemize}
+%
+% \item
+% I dislike defining a macro within another macro, especially when the
+% pattern part is non-nil.
+% You know, you are not saving much space in |mem| when you do this,
+% right?
+% You do save space in the hash table and the string pool, though.
+% On the other hand, we are not dealing with small \TeX\ engines
+% anymore; Team \LaTeX\ has made sure of this.
+%
+% \item
+% If two or more macros have very similar replacement parts, consider
+% layering.
+%
+% \item
+% Macros may perform parsing, may maintain tokens or registers, or may
+% set type (produce marks). I try to avoid combining the three functions
+% in a single macro.
+%
+% \item
+% When a procedure both does assignments and sets type, I try to have a
+% clean separations between the two activities. Try to avoid:
+% \begin{verbatim}
+% \vskip10pt
+% \parindent=0pt
+% \leavevmode
+% \end{verbatim}
+%
+% \item
+% The Boolean calculus (cf. \cmd\@ifx)
+% is very useful and produces code that executes nicely.
+% Using it also helps avoid your having to debug problems where
+% \cmd\if\dots\ and \cmd\fi\ are not properly balanced
+% (a nightmare in case you didn't already experience it).
+%
+% \end{itemize}
+%
+% \subsection{Conventions For \LaTeX}%
+% Team \LaTeX\ make certain recommendations in \file{clsguide.tex}.
+% Ones that I particularly pay attention to are:
+% \begin{itemize}
+%
+% \item
+% For the sake of ``color safety'',
+% use \cmd\sbox\ rather than \cmd\setbox, \cmd\mbox\ rather than \cmd\hbox, and
+% \cmd\parbox\ or \env{minipage} rather than \cmd\vbox.
+%
+% \item
+% Use \cmd\newcommand\ and \cmd\newenvironment\ to declare user-level commands
+% and environments. Avoid the idiom \cmd\def\cmd\foo, \cmd\def\cmd\endfoo\ to define
+% an environment.
+% Ideally, all user-level markup could be extracted from the
+% document class by grepping on \cmd\newcommand\ and \cmd\newenvironment.
+%
+% \item
+% Prefer to use \cmd\setlength\ to assign registers.
+%
+% \end{itemize}
+% I cannot help but notice that much of the code of \LaTeX\ itself fails
+% to comply with many of the coding recommendations of Team \LaTeX.
+%
+%
+% \Finale
+% %Here ends the programmer's documentation.
+% \endinput
+%
+\endinput
diff --git a/Master/texmf-dist/source/latex/revtex4/ltxgrid.dtx b/Master/texmf-dist/source/latex/revtex4/ltxgrid.dtx
new file mode 100644
index 00000000000..ab0209ed098
--- /dev/null
+++ b/Master/texmf-dist/source/latex/revtex4/ltxgrid.dtx
@@ -0,0 +1,4815 @@
+% \iffalse ltxdoc klootch
+% ltxgrid.dtx: package to change page grid, MVL.
+% Copyright (c) 1999 Arthur Ogawa
+%
+% Disclaimer
+% This file is distributed WITHOUT ANY WARRANTY;
+% without even the implied warranty of
+% MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
+% ReadMe
+% For the documentation and more detailed instructions for
+% installation, typeset this document with \LaTeX.
+% \fi
+% \GetFileInfo{ltxgrid.dtx}\CheckSum{4234}
+%
+% \iffalse ltxdoc klootch
+%<*ltxgrid>
+%%% @LaTeX-file{
+%%% filename = "ltxgrid.dtx",
+%%% version = "1.0rc5",
+%%% date = "2001/07/26",
+%%% time = "12:23:00 GMT+8",
+%%% checksum = "4234",
+%%% author = "Arthur Ogawa (mailto:ogawa@teleport.com),
+%%% commissioned by the American Physical Society.
+%%% ",
+%%% copyright = "Copyright (C) 1999, 2000 Arthur Ogawa,
+%%% distributed under the terms of the
+%%% LaTeX Project Public License, see
+%%% ftp://ctan.tug.org/macros/latex/base/lppl.txt
+%%% ",
+%%% address = "Arthur Ogawa,
+%%% USA",
+%%% telephone = "",
+%%% FAX = "",
+%%% email = "ogawa@teleport.com",
+%%% codetable = "ISO/ASCII",
+%%% keywords = "latex, page grid, main vertical list",
+%%% supported = "yes",
+%%% abstract = "package to change page grid, MVL",
+%%% docstring = "The checksum field above generated by ltxdoc",
+%%% }
+%</ltxgrid>
+% \fi
+%
+% \iffalse ltxdoc klootch
+% The following references the \file{00readme.txt} file,
+% which contains basic information about this package.
+% The contents of this file are generated when
+% you typeset the programmer's documentation.
+% Search on "{filecontents*}{00readme.txt}" to locate it.
+% \fi\input{00readme.txt}%
+%
+% \subsection{Bill of Materials}
+%
+% Following is a list of the files in this distribution arranged
+% according to provenance.
+%
+% \subsubsection{Primary Source}%
+% One single file generates all.
+%\begin{verbatim}
+%ltxgrid.dtx
+%\end{verbatim}
+%
+% \subsubsection{Generated by \texttt{latex ltxgrid.dtx}}%
+% Typesetting the source file under \LaTeX\
+% generates the readme and the installer.
+%\begin{verbatim}
+%00readme.txt ltxgrid.ins
+%\end{verbatim}
+%
+% \subsubsection{Generated by \texttt{tex ltxgrid.ins}}%
+% Typesetting the installer generates
+% the package files.
+%\begin{verbatim}
+%ltxgrid.sty
+%\end{verbatim}
+%
+% \subsubsection{Documentation}%
+% The following are the online documentation:
+% \begin{verbatim}
+%ltxgrid.pdf
+% \end{verbatim}
+%
+% \subsubsection{Auxiliary}%
+% The following are auxiliary files generated
+% in the course of running \LaTeX:
+% \begin{verbatim}
+%ltxgrid.aux ltxgrid.idx ltxgrid.ind ltxgrid.log ltxgrid.toc
+% \end{verbatim}
+%
+% \section{Code common to all modules}%
+%
+% The following may look a bit klootchy, but we
+% want to require only one place in this file
+% where the version number is stated,
+% and we also want to ensure that the version
+% number is embedded into every generated file.
+%
+% Now we declare that
+% these files can only be used with \LaTeXe.
+% An appropriate message is displayed if
+% a different \TeX{} format is used.
+% \begin{macrocode}
+%<*doc|ltxgrid>
+\NeedsTeXFormat{LaTeX2e}[1995/12/01]%
+%</doc|ltxgrid>
+% \end{macrocode}
+% As desired, the following modules all
+% take common version information:
+% \begin{macrocode}
+%<ltxgrid>\ProvidesFile{ltxgrid.sty}%
+%<*doc>
+\expandafter\ProvidesFile\expandafter{\jobname.dtx}%
+%</doc>
+% \end{macrocode}
+%
+% The following line contains, for once and for all,
+% the version and date information.
+% By various means, this information is reproduced
+% consistently in all generated files and in the
+% typeset documentation.
+% \begin{macrocode}
+%<*doc|ltxgrid>
+ [2001/07/26 1.0rc5 page grid package]% \fileversion
+%</doc|ltxgrid>
+% \end{macrocode}
+%
+%
+% \section{The driver module \texttt{doc}}
+%
+% This module, consisting of the present section,
+% typesets the programmer's documentation,
+% generating the \file{.ins} installer and \file{00readme.txt} as required.
+%
+% Because the only uncommented-out lines of code at the beginning of
+% this file constitute the \file{doc} module itself,
+% we can simply typeset the \file{.dtx} file directly,
+% and there is thus rarely any need to
+% generate the ``doc'' {\sc docstrip} module.
+% Module delimiters are nonetheless required so that
+% this code does not find its way into the other modules.
+%
+% The \enve{document} command concludes the typesetting run.
+%
+% \begin{macrocode}
+%<*doc>
+% \end{macrocode}
+%
+% \subsection{The Preamble}
+% The programmers documentation is formatted
+% with the \classname{ltxdoc} class with local customizations,
+% and with the usual code line indexing.
+% \begin{macrocode}
+\documentclass{ltxdoc}
+\RequirePackage{ltxdocext}%
+\RequirePackage[colorlinks=true,linkcolor=blue]{hyperref}%
+\ifx\package@font\@undefined\else
+ \expandafter\expandafter
+ \expandafter\RequirePackage
+ \expandafter\expandafter
+ \expandafter{%
+ \csname package@font\endcsname
+ }%
+\fi
+\CodelineIndex\EnableCrossrefs
+% \end{macrocode}
+%
+% \subsubsection{Docstrip and info directives}
+% We use so many {\sc docstrip} modules that we set the
+% \texttt{StandardModuleDepth} counter to 1.
+% \begin{macrocode}
+\setcounter{StandardModuleDepth}{1}
+% \end{macrocode}
+% The following command retrieves the date and version information
+% from this file.
+% \begin{macrocode}
+\expandafter\GetFileInfo\expandafter{\jobname.dtx}%
+% \end{macrocode}
+%
+%
+% \subsection{The installer file}
+%
+% The installer \file{ltxgrid.ins} appears here.
+% If you have retrieved the standard distribution of this package,
+% the installer file is already on your filesystem.
+% If you are bootstrapping,
+% the first typesetting of the \file{.dtx} file
+% will cause the installer to be generated.
+%
+% The following modules are used to direct
+% {\sc docstrip} in generating the external files:
+% \begin{center}
+% \begin{tabular}{lll}
+% \textbf{Module}&\textbf{File}&\textbf{Description}\\
+% doc &\file{ltxgrid.drv}&driver for programmer's documentation\\
+% ltxgrid,ltxgrid-krn &\file{ltxgrid.sty}&this package\\
+% ltxgrid-krn& &the portion of this package suitable for inclusion within another package
+% \end{tabular}
+% \end{center}
+%
+% \begin{macrocode}
+\begin{filecontents}{ltxgrid.ins}
+%% This file will generate documentation and runtime files
+%% from ltxgrid.dtx when run through LaTeX or TeX.
+\input docstrip
+\preamble
+
+This is a generated file;
+altering it directly is inadvisable;
+instead, modify the original source file.
+See the URL in the file 00readme.txt.
+
+Copyright notice.
+
+ These files are distributed
+ WITHOUT ANY WARRANTY; without even the implied warranty of
+ MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
+
+\endpreamble
+\keepsilent
+ \generate{%
+ \file{ltxgrid.drv}{\from{ltxgrid.dtx}{doc}}%
+ \file{ltxgrid.sty}{%
+ \from{ltxgrid.dtx}{ltxgrid,ltxgrid-krn}%
+ }%
+ }%
+\ifToplevel{
+\Msg{***********************************************************}
+\Msg{*}
+\Msg{* To finish the installation, please move}
+\Msg{* ltxgrid.sty}
+\Msg{* into a directory searched by TeX;}
+\Msg{* in a TDS-compliant installation:}
+\Msg{* texmf/tex/macros/latex/ao/.}
+\Msg{*}
+\Msg{* To produce the documentation,
+ run ltxgrid.dtx through LaTeX.}
+\Msg{*}
+\Msg{* Happy TeXing}
+\Msg{***********************************************************}
+}
+\endbatchfile
+\end{filecontents}
+% \end{macrocode}
+% Note that, because all of the files generated by the installer
+% are part of the standard distribution, it will
+% be necessary to run the installer only when bootstrapping
+% (or, of course, during development).
+% Note, too, that it is rare to generate the \file{doc}
+% module because it suffices to simply typeset the \file{.dtx} file itself.
+%
+% \subsection{The ``Read Me'' File}
+% As promised above, here is the contents of the
+% ``Read Me'' file. That file serves a double purpose,
+% since it also constitutes the beginining of the
+% programmer's documentation. What better thing, after
+% all, to have appear at the beginning of the
+% typeset documentation?
+%
+% A good discussion of how to write a ReadMe file can be found in
+% Engst, Tonya, ``Writing a ReadMe File? Read This''
+% \emph{MacTech} October 1998, p. 58.
+%
+% Note the appearance of the
+% \cmd\StopEventually\ command, which marks the
+% dividing line between the user documentation
+% and the programmer documentation.
+%
+% The usual user will not be asked to
+% do a full build, not to speak
+% of the bootstrap.
+% Instructions for carrying these processes
+% begin the programmer's manual.
+%
+% \begin{macrocode}
+\begin{filecontents*}{00readme.txt}
+\title{%
+ A \LaTeX\ Package for changing the page grid and MVL%
+ \thanks{%
+ This file has version number \fileversion,
+ last revised \filedate.%
+ % For version number and date,
+ % search on "\fileversion" in the .dtx file,
+ % or see the end of the 00readme.txt file.
+ }%
+}%
+
+\author{%
+Arthur Ogawa (\texttt{mailto:ogawa@teleport.com}),
+\fileversion\\Copyright (C) 1999, 2000 Arthur Ogawa
+}%
+\maketitle
+
+This file embodies the \classname{ltxgrid} package,
+the implementation and its user documentation.
+
+The distribution point for this work is
+\url{ftp://ftp.teleport.com/users/ogawa/macros/latex/contrib/supported/ltxgrid...},
+which contains fully unpacked, prebuilt runtime files and documentation.
+
+The \classname{ltxgrid} package was commissioned by the American Physical Society
+and is distributed under the terms of the \LaTeX\ Project Public License,
+the same license under which all the portions of \LaTeX\ itself is distributed.
+Please see \url{http://ctan.tug.org/macros/latex/base/lppl.txt} for details.
+
+To use this document class, you must have a working
+\TeX\ installation equipped with \LaTeXe\
+and possibly pdftex and Adobe Acrobat Reader or equivalent.
+
+To install, retrieve the distribution,
+unpack it into a directory on the target computer,
+into a location in your filesystem where it will be found by \LaTeX;
+in a TDS-compliant installation this would be:
+\file{texmf/tex/macros/latex/ao/.}
+
+To use, read the user documentation \file{ltxgrid.pdf}.
+
+\tableofcontents
+
+\section{Processing Instructions}
+
+The package file \file{ltxgrid.sty}
+is generated from this file, \file{ltxgrid.dtx},
+using the {\sc docstrip} facility of \LaTeX
+via |tex ltxgrid.ins|.
+The typeset documentation that you are now reading is generated from
+the same file by typesetting it with \LaTeX\ or pdftex
+via |latex ltxgrid.dtx| or |pdflatex ltxgrid.dtx|.
+
+\subsection{Build Instructions}
+
+You may bootstrap this suite of files solely from \file{ltxgrid.dtx}.
+Prepare by installing \LaTeXe\ (and either tex or pdftex) on your computer,
+then carry out the following steps:
+\begin{enumerate}
+\item
+Within an otherwise empty directory,
+typeset \file{ltxgrid.dtx} with \LaTeX\ or pdflatex;
+you will obtain the typeset documentation you are now reading,
+along with
+the installer \file{ltxgrid.ins},
+and the file \file{00readme.txt}.
+
+Note: you will have to run \LaTeX\ twice, then \file{makeindex}, then
+\LaTeX\ again in order to obtain a valid index and table of contents.
+\item
+Now typeset \file{ltxgrid.ins},
+thereby generating the package file \file{ltxgrid.sty}.
+\item
+Install \classname{ltxgrid.sty}
+by moving it to a location
+in your filesystem where they will be found by \LaTeX.
+\end{enumerate}
+\end{filecontents*}
+% \end{macrocode}
+%
+% \subsection{The Document Body}
+%
+% Here is the document body, containing only a
+% \cmd\DocInput\ directive---referring to this very file.
+% This very cute self-reference is a common \classname{ltxdoc} idiom.
+% \begin{macrocode}
+\begin{document}%
+\expandafter\DocInput\expandafter{\jobname.dtx}%
+% ^^A\PrintChanges
+\end{document}
+% \end{macrocode}
+%
+% \begin{macrocode}
+%</doc>
+% \end{macrocode}
+%
+% \section{Using this package}
+% Once this package is installed on your filesystem, you can employ it in
+% adding functionality to \LaTeX\ by invoking it in your document or document class.
+%
+% \subsection{Invoking the package}
+% In your document, you can simply call it up in your preamble:
+% \begin{verbatim}
+%\documentclass{book}%
+%\usepackage{ltxgrid}%
+%\begin{document}
+%<your document here>
+%\end{document}
+% \end{verbatim}
+% However, the preferred way is to invoke this package from within your
+% customized document class:
+% \begin{verbatim}
+%\NeedsTeXFormat{LaTeX2e}[1995/12/01]%
+%\ProvidesClass{myclass}%
+%\LoadClass{book}%
+%\RequirePackage{ltxgrid}%
+%<class customization commands>
+%\endinput
+% \end{verbatim}
+%
+% Note that this package requires the features of the \classname{ltxutil} package,
+% available at
+% \url{ftp://ftp.teleport.com/users/ogawa/macros/latex/contrib/supported/ltxutil/}.
+%
+% Once loaded, the package gives you acccess to certain procedures,
+% usually to be invoked by a \LaTeX\ command or environment, but not at the document level.
+%
+% \subsection{Changing the page grid}%
+% This package provides two procedures, \cmd\onecolumngrid, \cmd\twocolumngrid,
+% that change the page grid (it can be extended to more columns and to other page grids).
+%
+% They differ from standard \LaTeX's \cmd\onecolumn\ and \cmd\twocolumn\ commands in that
+% they do not force a page break. Also, upon leaving a multiple-column grid, the columns are balanced.
+% In other respects they work same.
+%
+% They differ from the grid-changing commands of Frank Mittelbach's \classname{multicol} package
+% in that they allow floats of all types (single- and double column floats, that is) and
+% preserve compatability with the \classname{longtable} package.
+%
+% These commands must be issued in vertical mode (conceivably via a \cmd\vadjust) such that
+% they are ultimately present in the MVL, where they can do their work.
+% Because they do not work in \LaTeX's left-right mode, they are unsuitable at the
+% document level.
+% Furthermore, packaging a grid command in a \cmd\vadjust, although possible, will probably
+% not acheive satisfactory page layout.
+%
+% Page grid commands are not intended to be issued unnecessarily: only the first of
+% two successive \cmd\onecolumngrid\ commands is effective; the second will be silently ignored.
+%
+% \DescribeMacro\onecolumngrid
+% You command \LaTeX\ to return to the one-column grid with the
+% \cmd\onecolumngrid\ command. If you are already in the one-column grid, this
+% is a no-op. The one-column grid is considered special of all page grids, in that
+% no portion of the page is held back (in \cmd\pagesofar); all items that might go
+% on the current page (with the exception of floats and footnotes) are on the MVL.
+%
+% \DescribeMacro\twocolumngrid
+% You command \LaTeX\ to return to the two-column grid with the
+% \cmd\twocolumngrid\ command. If you are already in the two-column grid, this
+% is a no-op.
+%
+% These two commands should be issued by a macro procedure that can ensure that
+% \TeX\ is in outer vertical mode.
+%
+% \subsection{Changing the MVL}%
+%
+% This package also provides commands to modify the main vertical list (MVL) in a safe way.
+% The scheme here is to structure, insofar possible, \TeX's MVL as follows:
+%\begin{quotation}
+%box or boxes\\
+%penalty\\
+%glue
+%\end{quotation}
+% This should be a familiar sequence. It is the prototype sequence for a vertical list,
+% and is followed when \TeX\ breaks paragraphs into lines, and when \TeX\ generates
+% a display math equation.
+%
+% If you (as a macro programmer) wish to modify the value of the penalty or glue item,
+% you can use one of the MVL-altering commands to do so. Certain operations are implemented
+% here; you can make up your own.
+%
+% Note that these commands must be issued in vertical mode, perhaps via a \cmd\vadjust\ or a \cmd\noalign.
+% They can work directly if you are in inner mode (say within a parbox or a minipage).
+%
+% \DescribeMacro\removestuff
+% You instruct \LaTeX\ to remove both the penalty and the glue item with this command.
+%
+% \DescribeMacro\addstuff
+% You issue the \cmd\addstuff\arg{penalty}\arg{glue} command to add a penalty, glue, or both.
+% If you do not wish to add one or the other, the corresponding argument should be nil.
+% Note that the effect of \cmd\addstuff\ is to stack the penalties and glue items.
+% Therefore, the lesser of the two penalties takes effect,
+% and the two glue items add together.
+%
+% \cmd\addstuff\ is limited because once applied, it cannot be applied again with correct results.
+%
+% \DescribeMacro\replacestuff
+% The \cmd\replacestuff\ command is syntactically the same as \cmd\addstuff, but works
+% differently: the existing penalty and glue are replaced or modified.
+%
+% The specified penalty is not inserted if the existing penalty is greater than 10000
+% (that is, in case of a \cmd\nobreak), otherwise, the lower (non-zero) of the two penalties is inserted.
+%
+% If the specified glue has a larger natural component than the existing glue, we replace the glue.
+% However, if the specified glue's natural component is negative, then the existing glue's natural component is
+% changed by that amount.
+%
+% \cmd\replacestuff\ can be applied mutiple times bceause it retains the list structure in the canonical form.
+%
+% Note that we treat two penalties specially (as does \TeX): a penalty of 10000 is considered
+% a garbage value, to be replaced if found. This is the signal value that \TeX\ inserts on
+% the MVL replacing the penalty that caused the page break (if the page break occurred at a penalty).
+% Also, a penalty of zero is indistinguishable from no penalty at all, so it will always
+% be replaced by the given value.
+%
+% Therefore, it is highly recommended to never set any of \TeX's penalty parameters to
+% zero (a value of, say, 1, is practically the same), nor should a skip parameter be set to zero
+% (instead, use, say, 1sp). Also, to prevent a pagebreak, do not use a penalty of 10000, use, say
+% 10001 instead.
+%
+% You can define your own construct that modifies the MVL:
+% Define a command, say, \cmd\myadjust, as follows:
+%\begin{verbatim}
+%\def\myadjust#1{\noexpand\do@main@vlist{\noexpand\@myadjust{#1}}\@tempa}%
+%\end{verbatim}
+% that is, \cmd\myadjust\ invokes \cmd\do@main@vlist, passing it the procedure name
+% \cmd\@myadjust\ along with the arguments thereof pre-expanded.
+% Next, define the procedure \cmd\@myadjust:
+%\begin{verbatim}
+%\def\@myadjust#1{<meddle with the MVL>}%
+%\end{verbatim}
+% when \cmd\@myadjust\ executes, you will be in the output routine (in inner vertical mode)
+% and the MVL will be that very vertical list.
+%
+%
+%
+%
+% \section{Compatability with \LaTeX's Required Packages}
+% Certain packages, usually ones written by members of the
+% \LaTeX\ Project itself, have been designated ``required'' and
+% are distributed as part of standard \LaTeX.
+% These packages have been placed in a priviledged position
+% vis \'a vis the \LaTeX\ kernel in that they override the definitions of certain kernel macros.
+%
+% Compatability between \classname{ltxgrid} and these packages is complicated
+% by a number of factors. First is that \classname{ltxgrid} alters the meaning of some of the same
+% kernel macros as certain of the ``required'' packages.
+% Second is that fact that certain of the ``required'' packages of \LaTeX\ are incompatible with
+% each other.
+%
+% Examples of the first kind are the \classname{ftnright}, \classname{multicol}, and \classname{longtable}
+% packages.
+% The \classname{ltxgrid} package is not compatible with \classname{multicol},
+% but if you are using \classname{ltxgrid}, you do not need to use \classname{ftnright} or \classname{multicol}
+% anyway. The \classname{ltxgrid} package does however attempt to be compatible with \classname{longtable}.
+%
+% Among the ``required'' packages that are mutually incompatible are \classname{multicol} and \classname{longtable},
+% the incompatibility arising because both packages replace \LaTeX's output routine:
+% if one package is active, the other must not be so.
+% This state of affairs has remained essentially unchanged since the introduction of the two as \LaTeX2.09 packages in the late 1980s.
+%
+% The reason that \classname{ltxgrid} can remain compatible with \classname{longtable} is due to the
+% introduction of a more modern architecture, the ``output routine dispatcher'', which allows all macro packages access to the
+% safe processing environment of the output routine, on an equal footing.
+% The relevant portions of the \classname{longtable} package are reimplemented in \classname{ltxgrid}
+% to take advantage of this mechanism.
+%
+% Timing is critical:
+% the \classname{ltxgrid} package will be incompatible with any package that
+% redefines any of the kernel macros that \classname{ltxgrid} patches---if that
+% package is loaded \emph{after} \classname{ltxgrid}.
+%
+% Hereinafter follows some notes on specific \LaTeX\ packages.
+%
+% \subsection{ftnright}
+% Frank Mittelbach's \classname{ftnright} package effects a change to \LaTeX's
+% \cmd\twocolumn\ mode such that footnotes are set at the bottom of the right-hand
+% column instead of at the foot of each of the two columns.
+%
+% Note that it overwrites three \LaTeX\ kernel macros: \cmd\@outputdblcol, \cmd\@startcolumn, and \cmd\@makecol.
+% Fortunately none of the three are patched by \classname{ltxgrid}, so that compatability
+% is not excluded on this basis.
+%
+% At the same time, it changes the meaning of \cmd\footnotesize, the macro that is automatically
+% invoked when setting a document's footnote into type.
+% One might well argue that it is an error for the meaning of \cmd\footnotesize\ to be determined by
+% a package such as \classname{ftnright}, that indeed such a choice should be made in the
+% document class, or in a file such as \file{bk10.clo}.
+%
+% To avoid being tripped up by this misfeature in \classname{ftnright}, it is only necessary to
+% reassert our meaning for \cmd\footnotesize\ later on, after \classname{ftnright} has been loaded.
+%
+% Note that \classname{ftnright} inserts code that demands that \LaTeX's flag \cmd\if@twocolumn\
+% is true, that is, it will complain if deployed in a \cmd\onecolumn\ document.
+% It is therefore necessary for any other multicolumn package to assert that flag in order to
+% avoid this package's complaint. It is an interesting question exactly why this package has
+% this limitation. After all, a one-column page grid is just a degenerate case of the
+% two column.
+%
+% \subsection{longtable}
+% David Carlisle's \classname{longtable} package sets tables that can be so long as to break over pages.
+% According to its author, it uses the same override of \LaTeX's output routine as
+% Frank Mittelbach's \classname{multicol} package. By implication, then, it has a hard
+% incompatability with the latter.
+%
+% The \classname{longtable} package also performs a check of whether the document is in
+% \cmd\twocolumn\ mode, and declines to work if this is the case. It is not clear, however,
+% that there is any true incompatability present if so. It's just that David did not see any reason
+% anyone would want to set such long tables in a multicolumn document, hence the check.
+%
+% There does not appear to be any indication that \classname{longtable} would work less
+% well under \classname{ltxgrid} than under standard \LaTeX's \cmd\twocolumn\ mode.
+% Therefore, this \classname{ltxgrid} patches \classname{longtable} (if loaded) so as to provide
+% compatability. In the course of which, \classname{longtable} becomes more robust
+% (\classname{longtable} has mumerous bugs and incompatabilities of long standing,
+% some of which are repaired by \classname{ltxgrid}).
+%
+% One problem remains, namely that, if a \env{longtable} environment breaks over columns
+% and thereby inserts its special headers and footers at that break, and those columms are then
+% balanced (due to a return to the one-column page grid), then those inserted rows
+% will remain, and may no longer fall at the column break. This will, of course look
+% wrong.
+%
+% The only way to fix this problem is to avoid doing column balancing in the way
+% I have implemented here; such an enhancement to this package is possible.
+%
+% \subsection{multicol}
+% Frank Mittelbach's \classname{multicol} package provides a page grid with many columns,
+% albeit denies the placement of floats in individual columns.
+%
+% It esablishes its own \cmd\output\ routine, which is the reason it runs afoul of
+% the \classname{longtable} package. On the other hand, \classname{ltxgrid} specifically
+% allows for the case where a package installs its own \cmd\output\ routine, so
+% there is no incompatability on that basis.
+%
+% Still, it is pointless to use \classname{multicol} if you are using \classname{ltxgrid},
+% since both packages provide multicolumn page layouts.
+% Therefore, \classname{multicol} is not supported by \classname{ltxgrid}.
+%
+% \subsection{ltxgrid}
+% It has been pointed out that one of the disadvantages of adopting the \classname{ltxgrid} package is that
+% it does alter the \LaTeX\ kernel.
+% Any package that itself alters the \LaTeX\ kernel may be incompatible with \classname{ltxgrid}, and new packages
+% (destined perhaps to become part of the successor to \LaTeXe) may break \classname{ltxgrid}.
+%
+% The consequence is that packages introduced in future, and future changes to \LaTeX\ may be incompatible
+% with \classname{ltxgrid}.
+% This is, of course, true.
+% The development plan for \classname{ltxgrid} is that when such packages and \LaTeX\ kernel changes come about,
+% the burden will be on \classname{ltxgrid} to change in a way that provides for continued compatability with
+% those packages and \LaTeX\ kernel changes.
+%
+%
+% \section{How \classname{ltxgrid} places footnotes}
+%
+% In conventional multicolumn layouts, a footnote will appear at the bottom of the column in which it is called out.
+% The \classname{ltxgrid} package implements this conventional layout choice by default.
+% However, other choices are possible (a la \classname{ftnright}, whose compatability with \classname{ltxgrid} has not been tested).
+%
+% One unusual feature of \classname{ltxgrid}'s default implementation must be mentioned, though,
+% namely the case in a two-column page grid, where a footnote is followed by a temporary change to the one-column page grid
+% (e.g., for a wide equation).
+% In such a case, the material above the wide material is split into two columns, and a footnote whose callout
+% appears in the right-hand column will nonetheless be set at the base of the left column.
+%
+% This arrangement was chosen because it ensures that the footnotes at the bottom of any page will appear in
+% numerical order. It can be argued that this choice is ``incorrect'', but be that as it may,
+% the \classname{ltxgrid} package does not foreclose on other arrangements for the footnotes.
+% The package can be adapted to accomodate any page design desired.
+%
+% \section{Limitations in \classname{ltxgrid}'s default column balancing method}%
+%
+% In a multicolumn page grid, when encountering a page that is not completely full,
+% it is customary to set the material in balanced columns (typically with the last column no longer than any of the others).
+% Such a case also crops up when temporarily interrupting the multicolumn grid to set material on the full width of the page:
+% the material on the page above the break is customarily set in balanced columns.
+%
+% An awkward case arises when we have already set one or more complete columns of type before encountering the need to
+% balance columns. In this subset of cases, the default in \classname{ltxgrid} is to
+% do an operation I call ``re-balancing'':
+% the material on the page so far is pasted back together into a single column, and new, balanced column breaks are
+% calculated.
+%
+% This scheme typically works fine, but it has a significant vulnerability:
+% any discardable items trimmed at the original column break is lost, never to be retrieved.
+% Consequently, after re-balancing, an element like, say, a section head can fail to have the correct amount of whitespace above.
+%
+% This problem is due to an unfortunate optimization in \TeX, wherein a certain class of nodes is trimmed from the
+% top of main vertical list upon returning from the output routine:
+% any penalty, glue, or leader node falls in to this class of discardable nodes,
+% and trimming proceeds until a non-discardable node (such as a box, or rule) is encountered.
+% It gets better: a third class of nodes is transparent to this trimming process;
+% they are neither discarded nor do they halt the process of trimming:
+% mark nodes and all whatsits fall into this class of transparent nodes;
+% they are quietly passed over during trimming.
+%
+% An alternative approach for \TeX\ to take would have been,
+% rather than discarding the node entirely, to simply mark it as discarded.
+% (Implementors of NTS, please note!)
+% Then, upon shipping out, such nodes would not make it into the DVI.
+% \TeX's optimization, driven by the small computer architectures current when it was developed,
+% does save mem, but at the cost of revisiting page breaks in a reliable way.
+%
+% FIXME: how to fix a column break in the above case? Widetext?
+%
+%\StopEventually{}
+%
+% \section{Implementation of package}
+%
+% Special acknowledgment: this package uses concepts pioneered
+% and first realized by William Baxter (mailto:web@superscript.com)
+% in his SuperScript line of commercial typesetting tools, and
+% which are used here with his permission. His thorough understanding
+% of \TeX's output routine underpins the entire \classname{ltxgrid}
+% package.
+%
+% \subsection{Beginning of the \file{ltxgrid} {\sc docstrip} module}
+% Requires the underpinnings of the \classname{ltxkrnext} package.
+% \begin{macrocode}
+%<*ltxgrid>
+\def\package@name{ltxgrid}%
+\expandafter\PackageInfo\expandafter{\package@name}{%
+ Page grid for \protect\LaTeXe,
+ by A. Ogawa (ogawa@teleport.com)%
+}%
+\RequirePackage{ltxutil}%
+%</ltxgrid>
+% \end{macrocode}
+%
+% \subsection{Banner}%
+% Credit where due.
+% \begin{macrocode}
+%<*ltxgrid-krn>
+\typeout{%
+ ltxgrid: portions licensed from W. E. Baxter (web@superscript.com)%
+}%
+% \end{macrocode}
+%
+% \subsection{Sundry}%
+% Here are assorted macro definitions.
+% \begin{macro}{\lineloop}
+% The document-level command \cmd\lineloop\ sets numbered lines until the
+% specified count is reached.
+% This command is mainly used to construct test documents.
+% \begin{macrocode}
+\newcounter{linecount}
+\def\lineloop#1{%
+ \loop
+ \ifnum\c@linecount<#1\relax
+ \global\advance\c@linecount\@ne
+ \par
+ \hb@xt@\hsize{%
+ \ifnum\c@linecount<100 0\fi\ifnum\c@linecount<10 0\fi\number\c@linecount
+ \vrule depth2.5\p@
+ \leaders\hrule\hfil
+ }%
+ \penalty\interlinepenalty
+ \repeat
+}%
+% \end{macrocode}
+% \end{macro}
+%
+%
+% \subsection{Mark Components}%
+%
+% Override LaTeX's mark macros to allow more components.
+%
+% We remain bound by the weakness of LaTeX's scheme in that
+% one cannot emulate the action of \TeX\ whereby
+% material with marks can be inserted in the middle of
+% a vertical list such that the marks are reliably calculated.
+% If we did that, \cmd\@themark\ would no longer be utilized.
+%
+% A more robust scheme involves placing all marks (component and value)
+% into a list (using global scoping, i.e., \cmd\gdef),
+% and using \cmd\@@mark to place an index on that list into the MVL.
+% Then, e.g., \cmd\@@botmark signifies the place where that list is to be cut,
+% and the \cmd\botmark\ of any component is
+% the value of the last element of the cut
+% list having the given component. The \cmd\firstmark\ and \cmd\topmark\
+% can likewise be defined relative to \cmd\@@firstmark\ and \cmd\@@topmark,
+% except in the latter case, we want the first following the cut instead of the last
+% preceding the cut.
+%
+% The limitation of this scheme is its demands upon \TeX's mem.
+% The list of marks would need to be trimmed back to, effectively,
+% \cmd\topmark\ at the beginning of every page.
+%
+% This approach is not yet part of the extended LaTeX kernel.
+%
+% \begin{macro}{\@@mark}
+% \begin{macro}{\@@topmark}
+% \begin{macro}{\@@firstmark}
+% \begin{macro}{\@@botmark}
+% \begin{macro}{\@@splitfirstmark}
+% \begin{macro}{\@@splitbotmark}
+% Remember primitives under a new set of names.
+% \begin{macrocode}
+\let\@@mark\mark
+\let\@@topmark\topmark
+\let\@@firstmark\firstmark
+\let\@@botmark\botmark
+\let\@@splitfirstmark\splitfirstmark
+\let\@@splitbotmark\splitbotmark
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \subsubsection{Procedures that expose the component data structure}%
+% This portion of the code exposes the internal representation of
+% the mark components. If we wish to add more components, we will have to revise
+% these macro definitions:
+% \cmd\@themark,
+% \cmd\nul@mark,
+% \cmd\set@mark@netw@,
+% \cmd\set@marktw@,
+% \cmd\set@markthr@@,
+% \cmd\get@mark@@ne,
+% \cmd\get@mark@tw@,
+% \cmd\get@mark@thr@@,
+% \cmd\get@mark@f@ur.
+%
+% \begin{macro}{\@themark}
+%FIXME: is it safer to eliminate \cmd\@themark\ in favor of a message that evaluates \cmd\@@botmark?
+%
+% Note: these definitions expose the data structure of mark components.
+% \begin{macrocode}
+\def\@themark{{}{}{}{}}%
+\def\nul@mark{{}{}{}{}\@@nul}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\set@mark@netw@}
+% \begin{macro}{\set@marktw@}
+% \begin{macro}{\set@markthr@@}
+% These procedures insert the new value of a particular mark component into the given argument.
+% They expose the data structure of mark components.
+%
+% \begin{macrocode}
+\def\set@mark@netw@#1#2#3#4#5#6#7{\gdef#1{{#6}{#7}{#4}{#5}}\do@mark}%
+\def\set@marktw@#1#2#3#4#5#6{\gdef#1{{#2}{#6}{#4}{#5}}\do@mark}%
+\def\set@markthr@@#1#2#3#4#5#6{\gdef#1{{#2}{#3}{#6}{#5}}\do@mark}%
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\get@mark@@ne}
+% \begin{macro}{\get@mark@tw@}
+% \begin{macro}{\get@mark@thr@@}
+% \begin{macro}{\get@mark@f@ur}
+% These procedures retreive the value of a particular mark component.
+% They expose the data structure of mark components.
+% \begin{macrocode}
+\def\get@mark@@ne#1#2#3#4#5\@@nul{#1}%
+\def\get@mark@tw@#1#2#3#4#5\@@nul{#2}%
+\def\get@mark@thr@@#1#2#3#4#5\@@nul{#3}%
+\def\get@mark@f@ur#1#2#3#4#5\@@nul{#4}%
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+%
+% \subsubsection{Procedures that do not expose the component data structure}%
+%
+% \begin{macro}{\mark@netw@}
+% \begin{macro}{\marktw@}
+% \begin{macro}{\markthr@@}
+% These procedures insert the new value of a particular mark component into \cmd\@themark,
+% then execute \cmd\do@mark.
+% They constitute the implementation layer for mark components one, two, and three.
+% An analogous procedure for component four could be defined; call it \cmd\markf@ur.
+%
+% \begin{macrocode}
+\def\mark@netw@{\expandafter\set@mark@netw@\expandafter\@themark\@themark}%
+\def\marktw@{\expandafter\set@marktw@\expandafter\@themark\@themark}%
+\def\markthr@@{\expandafter\set@markthr@@\expandafter\@themark\@themark}%
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\do@mark}
+% \begin{macro}{\do@@mark}
+% Access procedures \cmd\mark (AKA \cmd\@@mark).
+% The \cmd\do@mark\ procedure is used when a mark is being put down into the MVL;
+% \cmd\do@@mark\ when this happens in the output routine.
+% \begin{macrocode}
+\def\do@mark{\do@@mark\@themark\nobreak@mark}%
+\def\do@@mark#1{%
+ \begingroup
+ \let@mark
+ \@@mark{#1}%
+ \endgroup
+}%
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\let@mark}
+% \begin{macro}{\nobreak@mark}%
+% The procedure that makes \cmd\csname s robust within a mark.
+% Use \cmd\appdef\ and \cmd\robust@\ to extend the list.
+% \begin{macrocode}
+\def\let@mark{%
+ \let\protect\@unexpandable@protect
+ \let\label\relax
+ \let\index\relax
+ \let\glossary\relax
+}%
+\def\nobreak@mark{%
+ \@if@sw\if@nobreak\fi{\@ifvmode{\nobreak}{}}{}%
+}%
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+%
+% \subsubsection{Using mark components}%
+%
+% These procedures use the component mark mechanism to implement
+% a mark component that remembers the current environment (used in page makeup)
+% and the the two mark components left over from the original \LaTeX.
+% The fourth component is presently unused.
+%
+% \begin{macro}{\mark@envir}
+% The third mark component's access procedures.
+% The \cmd\mark@envir\ and \cmd\bot@envir\ commands are a good model of how to write
+% access procedures for a new mark component.
+% \begin{macrocode}
+\def\mark@envir{\markthr@@}%
+\def\bot@envir{%
+ \expandafter\expandafter
+ \expandafter\get@mark@thr@@
+ \expandafter\@@botmark
+ \nul@mark
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\markboth}
+% \begin{macro}{\markright}
+% \begin{macro}{\leftmark}
+% \begin{macro}{\rightmark}
+% Set procedures for legacy components.
+% \begin{macrocode}
+\def\markboth{\mark@netw@}%
+\def\markright{\marktw@}%
+% \end{macrocode}
+%
+% Retrieval procedures for legacy mark components.
+% The procedure for retrieving the first component from \cmd\botmark\
+% and the second component from \cmd\firstmark have names in \LaTeX;
+% they are called, respectively, \cmd\leftmark\ and \cmd\rightmark.
+%
+% It is possible to retrieve the components of \cmd\topmark\
+% as well: use \cmd\saved@@topmark.
+% \begin{macrocode}
+\def\leftmark{%
+ \expandafter\expandafter
+ \expandafter\get@mark@@ne
+ \expandafter\saved@@botmark
+ \nul@mark
+}%
+\def\rightmark{%
+ \expandafter\expandafter
+ \expandafter\get@mark@tw@
+ \expandafter\saved@@firstmark
+ \nul@mark
+}%
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+%
+%
+% \subsection{Output Super-routine}%
+%
+% We want to change \LaTeX's output routine, but do not wish to remain vulnerable
+% to interference from such ``required'' packages as
+% \classname{multicol} (authored by Frank Mittelbach)
+% and \classname{longtable} (authored by David P. Carlisle), which
+% swap in their own output routines when the respective package is active.
+%
+% The better mechanism, used here, is due to William Baxter (web@superscript.com),
+% who has allowed his several ideas to be used in this package.
+%
+% In what follows, we effectively wrap up the old \LaTeX\ output routine inside
+% a new, more flexible ``super routine''. When the output routine is called,
+% the ``super routine'' acts as a dispatcher. If the old routine is needed, it is called.
+%
+% If a package attempts to substitute in their own output routine, they will effectively
+% be modifying a token register by the name of \cmd\output.
+% The primitive \cmd\output\ is now known by a different name, which should no longer be
+% necessary to use.
+%
+% Usage note: to make a visit to the output routine employing the dispatcher, enter
+% with a value of \cmd\outputpenalty\ that corresponds to a macro. Defining as follows:
+%\begin{verbatim}
+%\@namedef{output@10000}{<your code here>}%
+%\end{verbatim}
+% by convention, your output routine should void out \cmd\box\cmd\@cclv.
+%
+% In rewriting \LaTeX's output dispatcher
+% in a much simpler form, we also avoid the sin of multiple \cmd\shipout s
+% within a single visit to the output routine.
+%
+% Conceptually, we divide visits to the output routine into two classes.
+% The first involves natural page breaks
+% (at a \cmd\newpage\ or when \cmd\pagetotal $>$ \cmd\pagegoal)
+% and usually resulting in \cmd\box\cmd\@cclv\ either being shipped out or
+% salted away (e.g., each column in a multicolumn layout).
+% We might call this class the ``natural output routines''; the \cmd\outputpenalty\
+% will never be less than $-10000$.
+% Furthermore, we ensure that \cmd\holdinginserts\ is cleared when
+% calling such routines.
+%
+% The other class involves a forced visit to the output routine
+% via a large negative penalty ($< -10000$). They do not generally
+% result in a \cmd\shipout\ of \cmd\box\cmd\@cclv: they may be dead cycles.
+% We provide a mechanism (call it a ``one-off'' output routine) that allows
+% us to specify certain processing to be done when \TeX\ reaches
+% the current position on the page.
+%
+% One-off output routines themselves fall into two divisions, ones
+% that process \cmd\box\cmd\@cclv, and ones that work on the main vertical list (MVL).
+% The former are typified by changes to the page grid, perhaps
+% even column balancing.
+% The latter involve the insertion of penalties or glue and the processing of floats.
+%
+% The natural output routine is a single procedure. We have not introduced multiple
+% natural output routines based on the \cmd\outputpenalty\ because \TeX\ does not
+% support such a thing: \TeX\ sometimes lays down a penalty whose value is the sum
+% of other penalties. Because of this, we cannot depend on the value of \cmd\outputpenalty\
+% in such areas.
+%
+% We do introduce flexibility in the form of a mechanism for patching into the
+% natural output routine. Three hooks are offered, allowing a procedure
+% to prepare for the upcoming visit to the output routine,
+% access to \cmd\box\cmd\@cclv, and after shipping out (or otherwise
+% committing the material to the page).
+%
+% Environments, commands, and even packages can install their
+% own procedures into these hooks.
+% For instance, if the longtable package is loaded, it will install
+% its procedures, but those procedures will punt if the page break
+% being processed does not actually fall within a longtable environment.
+%
+% \begin{macro}{\primitive@output}
+% Here we remember the \TeX\ primitive \cmd\output\ and its value,
+% and then proceed to take over the \cmd\csname\ of \cmd\output,
+% making it a \cmd\toks\ register and installing the old value of
+% the output routine.
+% \begin{macrocode}
+\let\primitive@output\output
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\output}
+% Grab the tokens in \cmd\the\cmd\output\ (but without the extra set of braces).
+% The value of \cmd\toks@\ must remain untouched until loaded into the appropriate token
+% register; this is done a few lines below.
+% \begin{macrocode}
+\long\def\@tempa#1\@@nil{#1}%
+\toks@
+\expandafter\expandafter
+\expandafter{%
+\expandafter \@tempa
+ \the\output
+ \@@nil
+ }%
+\newtoks\output
+\output\expandafter{\the\toks@}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\dispatch@output}
+% We now install our own output routine in place of the
+% old one, which is still available as \cmd\the\cmd\output.
+%
+% The output routine is simply the procedure \cmd\dispatch@output.
+% It either dispatches to a procedure based on a particular value of
+% \cmd\outputpenalty\ or it executes \cmd\the\cmd\output\ tokens.
+% \begin{macrocode}
+\primitive@output{\dispatch@output}%
+\def\dispatch@output{%
+ \let\par\@@par
+ \expandafter\let\expandafter\@tempa\csname output@\the\outputpenalty\endcsname
+ \outputdebug@sw{%
+ \saythe\badness
+ \saythe\outputpenalty
+ \saythe\holdinginserts
+ \say\thepagegrid
+ \saythe\pagegrid@col
+ \saythe\pagegrid@cur
+ %\say\bot@envir
+ \saythe\insertpenalties
+ %\say\@@topmark
+ %\say\saved@@topmark
+ %\say\@@firstmark
+ %\say\saved@@firstmark
+ \say\@@botmark
+ %\say\saved@@botmark
+ \saythe\pagegoal
+ \saythe\pagetotal
+ \saythe{\badness\@cclv}%
+ \expandafter\@ifx\expandafter{\csname output@-\the\execute@message@pen\endcsname\@tempa}{%
+ \say\@message@saved
+ }{%
+ \expandafter\say\csname output@\the\outputpenalty\endcsname
+ }%
+ \say\@toplist
+ \say\@botlist
+ \say\@dbltoplist
+ \say\@deferlist
+ {\tracingall\scrollmode
+ \showbox\@cclv
+ \showbox\@cclv@saved
+ \showbox\pagesofar
+ \showbox\footbox
+ \showbox\footins@saved
+ \showbox\footins
+ \showlists
+ }%
+ }{}%
+ \@ifnotrelax\@tempa{\@tempa}{\the\output}%
+}%
+\@ifxundefined{\outputdebug@sw}{%
+ \@booleanfalse\outputdebug@sw
+}{}%
+% \end{macrocode}
+% \end{macro}
+%
+%
+% \subsection{Further thoughts about inserts}
+%
+% The only safe way to deal with inserts is to either set \cmd\holdininserts\ or
+% to commit to using whatever insert comes your way: you cannot change your mind
+% once you see a non-void \cmd\box\cmd\footins, say.
+%
+% Therefore all output routine processing must proceed with \cmd\holdinginserts\ set
+% until you are sure of the material to be committed to the page. At that point, you
+% can clear \cmd\holdinginserts, spew \cmd\box\cmd\@cclv, put down the appropriate penalty,
+% and exit, with the knowledge that \TeX\ will re-find the same pagebreak, this time
+% visiting the output routine with everything, including inserts, in their proper
+% place.
+% This technique applies to split elements (screens, longtable, index) as well as to
+% manufactured pages (float pages and clearpage pages).
+%
+% Therefore, the output routine must not make assumptions about whether \cmd\holdinginserts\
+% should be cleared; instead this must be left to the one-off output routines or the natural output routine.
+%
+% If we are manufacturing pages (``float page processing''), and if \cmd\pagegoal\ is not equal to
+% \cmd\vsize, then inserts are at hand, and our criterion should take into account the insert
+% material, even though we cannot measure its height based on the size of \cmd\box\cmd\footins\
+% (because \cmd\holdinginserts\ is set, you see).
+%
+% It would be better to take the complement of \cmd\floatpagefraction\ and use that
+% as a standard for the looseness of the page. Since \cmd\pagegoal\ reflects the inserted material,
+% the criterion becomes the difference of the aggregate height of the floats and the \cmd\pagegoal\
+% versus this "page looseness" standard.
+%
+% As a check, consider what happens if we bail out: \cmd\@deferlist\ has never been touched, so it
+% requires no attention. Also, \cmd\holdinginserts\ has never been cleared, so inserts require
+% no attention. So we only have to ensure that marks are preserved, which is already taken
+% care of by the message handler mechanism.
+%
+% If we are doing ordinary page cutting, then the scheme would be to detect whether we are within
+% a screen (or longtable as may be), do the adjustment to the page height, and return, but this time
+% with \cmd\holdinginserts\ cleared. Upon reentering the output routine, we may or may not be within
+% the screen environment, but we are now sure to have a final page break, and we can commit this
+% material (by shipping out or by saving it out as a full column).
+%
+% In the above, the first of the two visits to the output routine is a dead cycle and requires
+% propagation of marks, but nothing else.
+%
+% The natural output routine
+%
+% Here is the portion of the output routine that fields cases not handled by
+% the dispatcher.
+%
+% The default is to ship out a page and then look around for more material
+% that might constitute a ``float page''. However, because \cmd\holdinginserts\
+% is normally set, this output routine must first have a dead cycle and
+% come back again with \cmd\holdinginserts\ cleared.
+% Then, after shipping out, it puts down a message that
+% will manufacture zero or more float pages, finally terminating
+% with a procedure that commits floats to a new unfinished page.
+%
+% To accomodate special processing, we execute hooks whose name is based
+% on the value of the "envir" mark component. The default is "document",
+% ensured by an initial mark of that value; the associated procedures
+% are all nil. Any unknown envir value will "\cmd\relax\ out".
+%
+% The code \cmd\move@insert@sw\ tells whether we are on our first visit to
+% the output routine (with \cmd\holdinginserts\ still set), or our second
+% (with \cmd\holdinginserts\ cleared). The output routine will toggle the
+% setting.
+%
+% The commands \cmd\hold@insertions\ and \cmd\move@insertions\ respectively
+% clear and set the state of \cmd\move@insert@sw, so this procedure effectively
+% clears \cmd\holdinginserts\ just long enough to pick up the insertions.
+% Important: any output routine that clears \cmd\holdinginserts\
+% must guarentee that it is restored on the subsequent visit to the output routine.
+% Or, to put it another way, if an output routine detects that \cmd\holdinginserts\
+% is cleared, it should take it upon itself to restore it before exiting.
+%
+% The branch with \cmd\holdinginserts\ set is executed first; the other
+% branch follows on practically immediately thereafter. In the first branch,
+% we simply execute the appropriate hook and then execute a dead cycle.
+%
+% In the branch with \cmd\holdinginserts\ cleared, the procedure
+% builds up the current column, which is now complete, with \cmd\@makecol, then
+% dispatches to the shipout routine associated with the current page grid, \cmd\output@column@.
+% At the end, it triggers the execution of an output routine to prepare the next column (or page).
+%
+% \subsection{Natural output routine}%
+%
+% \begin{macro}{\output}
+% Here is what has become of the output routine of \LaTeX.
+% It is of necessity divided into phases, \cmd\output@holding\ is executed upon first encountering the natural page-breaking point, while inserts are being held.
+% The second phase, \cmd\output@moving, is set in motion by the first: here the same material (in most cases) will be processed with \cmd\holdinginserts\ cleared.
+%
+% \begin{macrocode}
+\output={\toggle@insert\output@holding\output@moving}%
+% \end{macrocode}
+%
+% The procedure \cmd\output@holding\
+% is our first cycle through the output routine; \cmd\holdinginserts\ is still set.
+% We give the current environment a heads up
+% (it is through this means that \classname{longtable} sets its running header and footer),
+% then we execute a dead cycle, which should propagate marks.
+%
+% One corner case that can crop up is the presence of a single unbreakable chunk whose size is larger
+% than \cmd\vsize.
+% Doing a dead cycle under such circumstances will not find the same breakpoint as this time
+% (remember we threw in a \cmd\mark\ node).
+% Instead, we attempt to remove the excess height of the material, so we can continue to propagate marks.
+%
+% The corner case is at hand if the natural size of \cmd\box\cmd\@cclv\ exceeds \cmd\pagegoal\ and
+% the contents cannot be shrunk to fit.
+%
+% \begin{macrocode}
+\def\output@holding{%
+ \csname output@init@\bot@envir\endcsname
+%\vbadness\@M
+%\vfuzz\maxdimen
+ \@if@exceed@pagegoal{\unvcopy\@cclv}{%
+ \setbox\z@\vbox{\unvcopy\@cclv}%
+ \outputdebug@sw{{\tracingall\scrollmode\showbox\z@}}{}%
+ \dimen@\ht\@cclv\advance\dimen@-\ht\z@
+ \dead@cycle@repair\dimen@
+ }{%
+ \dead@cycle
+ }%
+}%
+\def\@if@exceed@pagegoal#1{%
+ \begingroup
+ \setbox\z@\vbox{#1}%
+ \dimen@\ht\z@\advance\dimen@\dp\z@
+ \outputdebug@sw{\saythe\dimen@}{}%
+ \@ifdim{\dimen@>\pagegoal}{%
+ \setbox\z@\vbox{\@@mark{}\unvbox\z@}%
+ \splittopskip\topskip
+ \splitmaxdepth\maxdepth
+ \vbadness\@M
+ \vfuzz\maxdimen
+ \setbox\tw@\vsplit\z@ to\pagegoal
+ \outputdebug@sw{{\tracingall\scrollmode\showbox\tw@\showbox\z@}}{}%
+ \setbox\tw@\vbox{\unvbox\tw@}%
+ \@ifdim{\ht\tw@=\z@}{%
+ \ltxgrid@info{Found overly large chunk while preparing to move insertions. Attempting repairs}%
+ \aftergroup\true@sw
+ }{%
+ \aftergroup\false@sw
+ }%
+ }{%
+ \aftergroup\false@sw
+ }%
+ \endgroup
+}%
+% \end{macrocode}
+%
+% The procedure \cmd\output@moving\
+% is our second cycle through the output routine; \cmd\holdinginserts\ is now cleared,
+% and \cmd\insert s will have been split off into their respective box registers, like \cmd\footins.
+%
+% \begin{enumerate}
+% \item
+% Set the values of \cmd\topmark\ and \cmd\firstmark.
+% \item
+% If we got here because of a \cmd\clearpage\ command, remove the protection box that this mechanism has left on the MVL.
+% \item
+% If the contents of \cmd\box\cmd\@cclv\ are non-trivial, commit it to the current page or ship it out as the case may call for.
+% \item
+% If not, discard it (we are at the end of \cmd\clearpage\ processing).
+% \item
+% Set various values, including the available space for setting type on the next column (\cmd\@colroom).
+% \end{enumerate}
+%
+% The processing for a non-trivial \cmd\box\cmd\@cclv\ are:
+% \begin{enumerate}
+% \item
+% Execute the head procedure for the current environment.
+% \item
+% Make up a column and ship it out (or commit it to the current page) via a procedure keyed to the current page grid.
+% \item
+% Put down an interrupt for \cmd\do@startcolumn@pen: this will force a visit to the output routine for the
+% purpose of committing floats to the next column.
+% \item
+% Possibly put down an interrupt to continue \cmd\clearpage\ proccessing.
+% \item
+% Execute the tail procedure for the current environment.
+% \end{enumerate}
+%
+%
+% The processing for a trivial \cmd\box\cmd\@cclv\ are:
+% \begin{enumerate}
+% \item
+% Void out \cmd\box\cmd\@cclv\ and give appropriate warning messages and diagnostics.
+%% \item
+%% Put down the same interrupts as for the non-trivial case above.
+% \end{enumerate}
+%
+% \begin{macrocode}
+\def\output@moving{%
+ \set@top@firstmark
+ \@ifnum{\outputpenalty=\do@newpage@pen}{%
+ \setbox\@cclv\vbox{%
+ \unvbox\@cclv
+ \setbox\z@\lastbox
+ \@ifdim{\ht\z@=\ht\@protection@box}{\box\lastbox}{\unskip}%
+ }%
+ }{}%
+ \@cclv@nontrivial@sw{%
+ \csname output@prep@\bot@envir \endcsname
+ \@makecol\csname output@column@\thepagegrid\endcsname
+ \protect@penalty\do@startcolumn@pen
+ \clearpage@sw{%
+ \protect@penalty\do@endpage@pen
+ }{}%
+ \csname output@post@\bot@envir \endcsname
+ }{%
+ {\setbox\z@\box\@cclv}%
+ }%
+ \set@colroom
+ \global\@mparbottom\z@
+ \global\@textfloatsheight\z@ %FIXME: this legacy LaTeX variable is set, but never queried!
+}%
+% \end{macrocode}
+%
+% The procedure \cmd\@cclv@nontrivial@sw\ determines if this visit to \cmd\output@moving\
+% is a trivial one, which happens at the end of \cmd\clearpage\ processing and under some pathological circumstances.
+% It emits a Boolean, so it is syntactically like \cmd\true@sw, albeit does not execute solely via expansion.
+%
+% Note: the case where \cmd\box\cmd\@cclv\ is void comes up at the very beginning of the job, when
+% typesetting a (full-page-width) title block in a two-column layout.
+%
+% Note: the code that removes the last box and skip from the output is intended to detect the case
+% where the output has whatit nodes followed by topskip and a protection box.
+% This is what happens under normal circumstances at the end of \cmd\clearpage\ processing.
+% \begin{macrocode}
+\def\@cclv@nontrivial@sw{%
+ \@ifx@empty\@toplist{%
+ \@ifx@empty\@botlist{%
+ \@ifvoid\footins{%
+ \@ifvoid\@cclv{%
+ \false@sw
+ }{%
+ \setbox\z@\vbox{\unvcopy\@cclv}%
+ \@ifdim{\ht\z@=\topskip}{%
+ \setbox\z@\vbox{%
+ \unvbox\z@
+ \setbox\z@\lastbox\dimen@\lastskip\unskip
+ \@ifdim{\ht\z@=\ht\@protection@box}{%
+ \advance\dimen@\ht\z@
+ \@ifdim{\dimen@=\topskip}{%
+ \aftergroup\true@sw
+ }{%
+ \aftergroup\false@sw
+ }%
+ }{%
+ \aftergroup\false@sw
+ }%
+ }%
+ {%
+ \false@sw % Normal for \clearpage
+ }{%
+ \true@sw
+ }%
+ }{%
+ \@ifdim{\ht\z@=\z@}{%
+ \ltxgrid@info{Found trivial column. Discarding it}%
+ \outputdebug@sw{{\tracingall\scrollmode\showbox\@cclv}}{}%
+ \false@sw
+ }{%
+ \true@sw
+ }%
+ }%
+ }%
+ }{%
+ \true@sw
+ }%
+ }{%
+ \true@sw
+ }%
+ }{%
+ \true@sw
+ }%
+}%
+% \end{macrocode}
+%
+%
+% \end{macro}
+%
+% \begin{macro}{\protect@penalty}
+% The procedure \cmd\protect@penalty\ is the utility procedure for invoking a
+% one-off output routine. Such a routine can expect to find the protection box
+% above it in \cmd\box\cmd\@cclv: it should remove that box.
+%
+% Note that \cmd\execute@message\ does the same thing as \cmd\protect@penalty, but
+% in a slightly different way.
+%
+% We create a specially formulated box that will be universally used when a protection box is needed.
+% In this way, we can always recognize when \cmd\box\cmd\@cclv\ is trivial:
+% it will consist of whatsits followed by \cmd\topskip\ glue and the \cmd\@protection@box.
+% \begin{macrocode}
+\def\protect@penalty#1{\protection@box\penalty-#1\relax}%
+\newbox\@protection@box
+\setbox\@protection@box\vbox to1986sp{\vfil}%
+\def\protection@box{\nointerlineskip\copy\@protection@box}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\dead@cycle}
+% \begin{macro}{\dead@cycle@repair}
+% The procedure \cmd\dead@cycle\ is defined separately as a utility which can be used by
+% any output processing routine to emulate what takes place in the standard output routine.
+%
+% Here, we have entered the output routine with \cmd\holdinginserts\ enabled, which means that we
+% are not yet ready to ship out material, because the \cmd\insert\ registers are being held.
+% We want to clear \cmd\holdinginserts\ and come back here with the same page break as before, whereupon
+% we may properly proceed with page makeup.
+%
+% To do this, we
+% propagate marks, then spew the contents of \cmd\box\cmd\@cclv\ followed by the
+% original output penalty that landed us here (but only if it is not 10000,
+% the flag value for a pagebreak not at a penalty).
+%
+% However, the natural output routine should do this only if \cmd\box\cmd\@cclv\ is nontrivial.
+% A pathological case exists wherein a box of height greater than \cmd\textheight\ would cause an infinite loop involving the output routine.
+% The procedure \cmd\dead@cycle@repair, attempts to catch this case and avoid the loop.
+%
+% The test of the height of \cmd\box\cmd\@cclv\ is not the correct one, because this test will run afoul in
+% the case where \cmd\box\cmd\@cclv\ contains nothing but an \cmd\insert\ node. What to do?
+%
+% It is possible that the pathological case can be detected by looking at \cmd\pagetotal. If that quantity is
+% zero, then \cmd\box\cmd\@cclv\ really is trivial.
+%
+% In the procedure \cmd\dead@cycle@repair, if \cmd\box\cmd\@cclv\ is nontrivial, we execute \cmd\dead@cycle,
+% otherwise it contains nothing but a mark, so we dispense with propagating marks
+% and we simply spew out \cmd\box\cmd\@cclv\ without an accompanying mark.
+% This has the effect of failing to propagate marks, but this problem is preferrable to the infinite loop,
+% which in principle could crash even a robust operating system by filling up the file system.
+%
+% If a document has such a large chunk, it should be fixed, so we give a message in the log.
+%
+% You ask, ``In what way does this infinite loop come about?'' Good question!
+%
+% The setup is a chunk in the MVL that is taller than \cmd\textheight.
+% (Yes, it's that simple.)
+% As soon as the previous page ships out, the MVL will contain a mark (propagated from the previous page) followed
+% by that large chunk (call it the `big bad box', albeit does not need to be a single box).
+% The next visit to the output routine will be a natural page break, but
+% \TeX\ will select the juncture between the mark and the big bad box as the least-cost page break.
+% Unless the test in \cmd\dead@cycle\ is done, the cycle is perpetuated when the macro
+% reinserts the mark.
+%
+% The crux matter is achieving, in a robust way, the goal of going from a \cmd\holdinginserts\ state to one
+% where the insertions are moving.
+%
+% \begin{macrocode}
+\def\dead@cycle@repair#1{%
+ \expandafter\do@@mark
+ \expandafter{%
+ \@@botmark
+ }%
+ \unvbox\@cclv
+ \nointerlineskip
+ \vbox to#1{\vss}%
+ \@ifnum{\outputpenalty<\@M}{\penalty\outputpenalty}{}%
+}%
+\def\dead@cycle@repair@protected#1{%
+ \expandafter\do@@mark
+ \expandafter{%
+ \@@botmark
+ }%
+ \begingroup
+ \unvbox\@cclv
+ \setbox\z@\lastbox % Remove protection box
+ \nointerlineskip
+ \advance#1-\ht\@protection@box
+ \vbox to#1{\vss}%
+ \protection@box % Reinsert protection box
+ \@ifnum{\outputpenalty<\@M}{\penalty\outputpenalty}{}%
+ \endgroup
+}%
+\def\dead@cycle{%
+ \expandafter\do@@mark
+ \expandafter{%
+ \@@botmark
+ }%
+ \unvbox\@cclv
+ \@ifnum{\outputpenalty<\@M}{\penalty\outputpenalty}{}%
+}%
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\output@init@document}
+% \begin{macro}{\output@prep@document}
+% \begin{macro}{\output@post@document}
+% The default processing simply provides for insertion of held-over footnotes.
+% At a natural page break, we are either at the bottom of a column or at the bottom
+% of a page. In either case, the \cmd\output@init@\ processing adjusts for the height
+% of the held-over footnotes and bails out.
+% Upon our return, at \cmd\output@prep@\ time, the page break will accomodate the material;
+% it is now actually inserted by concatenating it with the contents of \cmd\footins.
+% The default processing for \cmd\output@post@\ is nil.
+% \begin{macrocode}
+\def\output@init@document{%
+ \@ifvoid\footbox{}{%
+ \global\advance\vsize-\ht\footbox
+ \global\advance\vsize-\dp\footbox
+ }%
+}%
+\def\output@prep@document{%
+ \@ifvoid\footbox{}{%
+% {\tracingall\scrollmode\showbox\footbox\showbox\footins}%
+ \setbox\footins\vbox{\unvbox\footbox\unvbox\footins}%
+ }%
+}%
+\def\output@post@document{}%
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\@opcol}
+% The standard \LaTeX\ procedure \cmd\@opcol\ is now completely obsoleted.
+% \begin{macrocode}
+\let\@opcol\@undefined
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\@makecol}
+% The procedure \cmd\@makecol\ packages up a page along with all its insertions and floats.
+% Therefore it is essential that it be executed with \cmd\holdininserts\ cleared.
+%
+% Note that there is a corner case when in a multi-column grid, where the change back to
+% one-column grid occurs just after a complete page ships out. We want to detect when
+% \cmd\@cclv\ contains nothing but a \cmd\mark, but this is a \TeX\ impossibility.
+%
+% Note on \cmd\@kludgeins: we have removed this mechanism from \LaTeX, because the implementation
+% of \cmd\enlargethispage\ no longer requires it.
+% Here, for consistency sake, we remove \cmd\@makespecialcolbox.
+% \begin{macrocode}
+\def\@makecol{%
+ \setbox\@outputbox\vbox{%
+ \boxmaxdepth\@maxdepth
+ \@tempdima\dp\@cclv
+ \unvbox\@cclv
+ \vskip-\@tempdima
+ }%
+ \xdef\@freelist{\@freelist\@midlist}\global\let\@midlist\@empty
+ \@combinefloats
+ \@combineinserts\@outputbox\footins
+%\@ifvbox\@kludgeins{%
+% \@makespecialcolbox
+%}{%
+ \set@adj@colht\dimen@
+ \count@\vbadness
+ \vbadness\@M
+ \setbox\@outputbox\vbox to\dimen@{%
+ \@texttop
+ \dimen@\dp\@outputbox
+ \unvbox\@outputbox
+ \vskip-\dimen@
+ \@textbottom
+ }%
+ \vbadness\count@
+%}%
+ \global\maxdepth\@maxdepth
+}%
+\let\@makespecialcolbox\@undefined
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\@combineinserts}
+% We split out the procedure to add the \cmd\footins\ insertions to the packaged-up page.
+% Any other non-trivial insertions should also be dealt with at this time.
+% \begin{macrocode}
+\def\@combineinserts#1#2{%
+ \setbox#1\vbox{%
+ \unvbox#1%
+% {\tracingall\scrollmode\showbox#2}%
+ \vbox{%
+ \@ifvoid#2{}{%
+ \vskip\skip\footins
+ \color@begingroup
+ \normalcolor
+ \footnoterule
+ \nointerlineskip
+ \box#2%
+ \color@endgroup
+ }{}%
+ }%
+ }%
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\@floatplacement}
+% In standard \LaTeX, someone (DPC?) makes the assumption that \cmd\@fpmin\ can be assigned
+% locally. This is no longer true now that we ship no more than one page per visit to the output routine.
+% We apply a bandaid.
+% \begin{macrocode}
+\appdef\@floatplacement{%
+ \global\@fpmin\@fpmin
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\pagebreak@pen}
+% While we are in the way of registering certain penalty values,
+% let us register the smallest one that will force a visit to the output routine.
+% However, this penalty will not have an assciated macro: we wish to execute the
+% natural output routine instead.
+%
+% Note that this penalty is invoked by \cmd\clearpage\ and \cmd\newpage.
+% \begin{macrocode}
+\mathchardef\pagebreak@pen=\@M
+\expandafter\let\csname output@-\the\pagebreak@pen\endcsname\relax
+% \end{macrocode}
+% \end{macro}
+%
+%
+% \subsection{Float placement}%
+%
+% \begin{macro}{\do@startcolumn@pen}
+% The procedure \cmd\do@startcolumn@pen\ is executed as a one-off output routine
+% just after a page is shipped out (or, in a multicolumn page grid, a column is salted away).
+%
+% Its job is to either generate a ``float page'' (in reality a column) for shipping out,
+% or to commit deferred floats to the fresh column, concluding with a dead cycle.
+% In the former case, we accomodate split footnotes and other insertions (by comparing \cmd\vsize\ and \cmd\pagegoal):
+% the floats are spewed onto the page, whereupon \LaTeX's output routine will place the footnotes and ship out,
+% iterating the process once again.
+%
+% Note that when this procedure is invoked, \cmd\box\cmd\@cclv\ still has within it the protection box, so we
+% start by removing it. Note also that if there was a split insertion held over from the previous page, the
+% insert node will be present in \cmd\box\cmd\@cclv, \emph{prior to} the protection box. For this reason, we cannot
+% just throw away that box, as we might be tempted to do.
+%
+% FIXME: where else do we possibly inappropriately discard \cmd\box\cmd\@cclv?
+%
+% Note that, because a column or page page had previously just been completed,
+% we can assume that there is nothing of importance on the page,
+% and because no message is being passed, we can preserve marks in a simple way.
+%
+% A Note on terminology:
+% In a single-column page grid, you might expect that we would execute the procedure \cmd\do@startpage.
+% But this is not so.
+% \LaTeX\ has a confustion of long standing,
+% in which the procedures that handle full-page width floats in a two-column page grid all have in their names
+% the string `dbl', which erroneously suggests having something to do with ``double''. It does not:
+% when you see `dbl', think ``full page width''.
+% \begin{macrocode}
+\mathchardef\do@startcolumn@pen=10005
+\@namedef{output@-\the\do@startcolumn@pen}{\do@startcolumn}%
+\def\do@startcolumn{%
+ \setbox\@cclv\vbox{\unvbox\@cclv\setbox\z@\lastbox\unskip}%
+ \clearpage@sw{\@clearfloatplacement}{\@floatplacement}%
+ \set@colroom
+ \@booleanfalse\pfloat@avail@sw
+ \begingroup
+ \@colht\@colroom
+ \@booleanfalse\float@avail@sw
+ \@tryfcolumn\test@colfloat
+ \float@avail@sw{\aftergroup\@booleantrue\aftergroup\pfloat@avail@sw}{}%
+ \endgroup
+ \fcolmade@sw{%
+ \setbox\@cclv\vbox{\unvbox\@outputbox\unvbox\@cclv}%
+% \csname float@column@\thepagegrid\endcsname
+% \csname output@column@\thepagegrid\endcsname
+ \outputpenalty-\pagebreak@pen % ask for a return visit, this time with insertions and all.
+ \dead@cycle
+ }{%
+ \begingroup
+ \let\@elt\@scolelt
+ \let\reserved@b\@deferlist\global\let\@deferlist\@empty\reserved@b
+ \endgroup
+ \clearpage@sw{%
+ \outputpenalty\@M
+ }{%
+ \outputpenalty\do@newpage@pen
+ }%
+ \dead@cycle
+ }%
+ \check@deferlist@stuck\do@startcolumn
+ \set@vsize
+}%
+\def\@scolelt#1{\def\@currbox{#1}\@addtonextcol}%
+\def\test@colfloat#1{%
+ \csname @floatselect@sw@\thepagegrid\endcsname#1{}{\@testtrue}%
+ \@if@sw\if@test\fi{}{\aftergroup\@booleantrue\aftergroup\float@avail@sw}%
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\@addtonextcol}
+% We must adjust \cmd\@addtonextcol\ to take held-over inserts into account.
+% Now that all deferred floats are queued up together (in order), we must have a way of
+% differentiating them; this is done by the page grid-dependent procedure \cmd\@floatselect@sw@.
+% \begin{macrocode}
+\def\@addtonextcol{%
+ \begingroup
+ \@insertfalse
+ \@setfloattypecounts
+ \csname @floatselect@sw@\thepagegrid\endcsname\@currbox{%
+ \@ifnum{\@fpstype=8 }{}{%
+ \@ifnum{\@fpstype=24 }{}{%
+ \@flsettextmin
+ \@reqcolroom \ht\@currbox
+ \advance \@reqcolroom \@textmin
+ \advance \@reqcolroom \vsize % take into account split insertions
+ \advance \@reqcolroom -\pagegoal
+ \@ifdim{\@colroom>\@reqcolroom}{%
+ \@flsetnum \@colnum
+ \@ifnum{\@colnum>\z@}{%
+ \@bitor\@currtype\@deferlist
+ \@if@sw\if@test\fi{}{%
+ \@addtotoporbot
+ }%
+ }{}%
+ }{}%
+ }%
+ }%
+ }{}%
+ \@if@sw\if@insert\fi{}{%
+ \@cons\@deferlist\@currbox
+ }%
+ \endgroup
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\do@startpage@pen}
+% \begin{macro}{\forcefloats@sw}
+% \begin{macro}{\@sdblcolelt}
+% \begin{macro}{\test@dblfloat}
+% \begin{macro}{\@if@notdblfloat}
+% Similar to \cmd\do@startcolumn,
+% the procedure \cmd\do@startpage\ starts up a new page (not column) in a multi-column page grid.
+% It is invoked after a page is shipped out in a multi-column page grid, and
+% it commits full-page-width floats to the fresh page, possibly resulting in a float page.
+% In implementation, it is similar to \cmd\do@startcolumn, except that
+% it commits effectively via \cmd\@addtodblcol\ instead of \cmd\@addtonextcol.
+% Note that this procedure will inevitably be followed by \cmd\do@startcolumn.
+%
+% Some details of the procedure:
+%
+% We begin by removing the protection box from \cmd\box\cmd\@cclv, then setting the values of the
+% float placement parameters appropriately, and resetting \cmd\@colht, \cmd\@colroom, and \cmd\vsize\ to base values.
+%
+% Next we attempt to compose a float page, a page consisting entirely of floats. If successful,
+% we ship out the float page and lay down an interrupt that will send us back here for another try.
+%
+% If no float page is formed, we attempt to commit full-page-width floats to the text page, and return with a dead cycle.
+% We are now ready to compose columns of text.
+%
+% Note that all floats (both column floats and full-page-width floats) move through a single queue.
+% To differentiate between the two, the width of the float is compared to \cmd\textwidth.
+% This comparison is encapsulated in the macro \cmd\@if@notdblfloat, which should be used whenever
+% such a determination must be made. This procedure returns a Boolean.
+% \begin{macrocode}
+\mathchardef\do@startpage@pen=10006
+\@namedef{output@-\the\do@startpage@pen}{\do@startpage}%
+\def\do@startpage{%
+ \setbox\@cclv\vbox{\unvbox\@cclv\setbox\z@\lastbox\unskip}%
+ \clearpage@sw{\@clearfloatplacement}{\@dblfloatplacement}%
+ \set@colht
+ \@booleanfalse\pfloat@avail@sw
+ \begingroup
+ \@booleanfalse\float@avail@sw
+ \@tryfcolumn\test@dblfloat
+ \float@avail@sw{\aftergroup\@booleantrue\aftergroup\pfloat@avail@sw}{}%
+ \endgroup
+ \fcolmade@sw{%
+ \global\setbox\pagesofar\vbox{\unvbox\pagesofar\unvbox\@outputbox}%
+ \@combinepage
+ \@combinedblfloats
+ \@outputpage
+ \global\pagegrid@cur\@ne
+ \protect@penalty\do@startpage@pen
+ }{%
+ \begingroup
+ \@booleanfalse\float@avail@sw
+ \let\@elt\@sdblcolelt
+ \let\reserved@b\@deferlist\global\let\@deferlist\@empty\reserved@b
+ \endgroup
+ \@ifdim{\@colht=\textheight}{% No luck...
+ \pfloat@avail@sw{% ...but a float *was* available!
+ \forcefloats@sw{%
+ \ltxgrid@warn{Forced dequeueing of floats stalled}%
+ }{%
+ \ltxgrid@warn{Dequeueing of floats stalled}%
+ }%
+ }{}%
+ }{}%
+ \outputpenalty\@M
+ \dead@cycle
+ }%
+ \check@deferlist@stuck\do@startpage
+ \set@colht
+%\set@colroom
+}%
+\def\@sdblcolelt#1{\def\@currbox{#1}\@addtodblcol}%
+\def\test@dblfloat#1{%
+ \@if@notdblfloat{#1}{\@testtrue}{}%
+ \@if@sw\if@test\fi{}{\aftergroup\@booleantrue\aftergroup\float@avail@sw}%
+}%
+\def\@if@notdblfloat#1{\@ifdim{\wd#1<\textwidth}}%
+\@booleanfalse\forcefloats@sw
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\@addtodblcol}
+%
+% The procedure \cmd\@addtodblcol\ is called into play at the beginning of each fresh page
+% and operates on each deferred float, in the hopes of placing one or more such floats
+% at the top of the current page.
+%
+% We alter the procedure of standard \LaTeX\ by putting failed floats into
+% \cmd\@deferlist\ instead of \cmd\@dbldeferlist. Having done so, we must have a means
+% of differentiating full-page-width floats from column-width floats.
+% We assume that the latter will always be narrower than \cmd\textwidth.
+%
+% In aid of detecting a stalled float flushing process, we set a Boolean if we encounter
+% a qualified full-page-width float here. Any that qualify but fail the rest of the tests
+% might still pass when reconsidered on an otherwise blank page.
+% \begin{macrocode}
+\def\@addtodblcol{%
+ \begingroup
+ \@if@notdblfloat{\@currbox}{%
+ \false@sw
+ }{%
+ \@setfloattypecounts
+ \@getfpsbit \tw@
+ \@bitor \@currtype \@deferlist
+ \@if@sw\if@test\fi{%
+ \false@sw
+ }{%
+ \@ifodd\@tempcnta{%
+ \aftergroup\@booleantrue\aftergroup\float@avail@sw
+ \@flsetnum \@dbltopnum
+ \@ifnum{\@dbltopnum>\z@}{%
+ \@ifdim{\@dbltoproom>\ht\@currbox}{%
+ \true@sw
+ }{%
+ \@ifnum{\@fpstype<\sixt@@n}{%
+ \begingroup
+ \advance \@dbltoproom \@textmin
+ \@ifdim{\@dbltoproom>\ht\@currbox}{%
+ \endgroup\true@sw
+ }{%
+ \endgroup\false@sw
+ }%
+ }{%
+ \false@sw
+ }%
+ }%
+ }{%
+ \false@sw
+ }%
+ }{%
+ \false@sw
+ }%
+ }%
+ }%
+ {%
+ \@tempdima -\ht\@currbox
+ \advance\@tempdima
+ -\@ifx{\@dbltoplist\@empty}{\dbltextfloatsep}{\dblfloatsep}%
+ \global \advance \@dbltoproom \@tempdima
+ \global \advance \@colht \@tempdima
+ \global \advance \@dbltopnum \m@ne
+ \@cons \@dbltoplist \@currbox
+ }{%
+ \@cons \@deferlist \@currbox
+ }%
+ \endgroup
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\@tryfcolumn}
+% \begin{macro}{\@wtryfc}
+% \begin{macro}{\@xtryfc}
+% \begin{macro}{\@ztryfc}
+% Whenever a page is shipped out, \LaTeX\ automatically tries out a float column:
+% a page containing nothing but floats (and, as we have added here, split footnotes).
+%
+% The following four procedures employ certain macros to communicate between each other:
+%
+% \cmd\fcolmade@sw, a boolean, says whether we were successful in making a float column.
+%
+% \cmd\if@test, a \cmd\newif\ switch, says a float has failed some test.
+%
+% \cmd\@deferlist, is the input to the process, a list, of deferred floats.
+%
+% \cmd\@trylist, a list, stores the deferred floats to be tried out on the float column.
+%
+% \cmd\@failedlist, a list of floats that have failed the selection for the float column.
+%
+% \cmd\@flfail, a list of floats that have failed the second selection for the float column.
+%
+% \cmd\@flsucceed, a list, the floats that have been successfully placed on the float column.
+%
+% \cmd\@freelist, a list, receives any freed floats.
+%
+% \cmd\@colht, a dimen, the available space for the column, including column floats and insertions (footnotes).
+%
+% \cmd\@fpmin, a dimen, the required minimum height for the float column.
+%
+% \cmd\@outputbox, a box, the output of the process.
+%
+% \cmd\@fptop, \cmd\@fpsep, \cmd\@fpbot, glue, placed above, between, and below floats on the float column.
+%
+% \cmd\@currtype, a count, used temporarily for the float's bits.
+%
+% \cmd\@tempcnta, a count, used temporarily for the float's bits.
+%
+% In \cmd\@tryfcolumn, we alter the criterion for a float page, because if footnotes are present at this point
+% (presumably due to a split insertion) then \cmd\@fpmin is no longer the right threshold to apply.
+%
+% Note that we have changed \cmd\@tryfcolumn, \cmd\@xtryfc, and \cmd\@ztryfc\ syntactically so that the procedure
+% to test for the float's being a column float versus a full-page-width float is passed in as an
+% argument.
+%
+% \begin{macrocode}
+\def\@tryfcolumn#1{%
+ \global\@booleanfalse\fcolmade@sw
+ \@ifx@empty\@deferlist{}{%
+ \global\let\@trylist\@deferlist
+ \global\let\@failedlist\@empty
+ \begingroup
+ \dimen@\vsize\advance\dimen@-\pagegoal\@ifdim{\dimen@>\z@}{%
+ \advance\@fpmin-\dimen@
+ }{}%
+ \def\@elt{\@xtryfc#1}\@trylist
+ \endgroup
+ \fcolmade@sw{%
+ \global\setbox\@outputbox\vbox{\vskip \@fptop}%
+ \let \@elt \@wtryfc \@flsucceed
+ \global\setbox\@outputbox\vbox{\unvbox\@outputbox
+ \unskip \vskip \@fpbot
+ }%
+ \let \@elt \relax
+ \xdef\@deferlist{\@failedlist\@flfail}%
+ \xdef\@freelist{\@freelist\@flsucceed}%
+ }{}%
+ }%
+}%
+\def\@wtryfc #1{%
+ \global\setbox\@outputbox\vbox{\unvbox\@outputbox
+ \box #1\vskip\@fpsep
+ }%
+}%
+\def\@xtryfc#1#2{%
+ \@next\reserved@a\@trylist{}{}% trim \@trylist. Ugly!
+ \@currtype \count #2%
+ \divide\@currtype\@xxxii\multiply\@currtype\@xxxii
+ \@bitor \@currtype \@failedlist
+ \@testfp #2%
+ #1#2%
+ \@ifdim{\ht #2>\@colht }{\@testtrue}{}%
+ \@if@sw\if@test\fi{%
+ \@cons\@failedlist #2%
+ }{%
+ \begingroup
+ \gdef\@flsucceed{\@elt #2}%
+ \global\let\@flfail\@empty
+ \@tempdima\ht #2%
+ \def \@elt {\@ztryfc#1}\@trylist
+ \@ifdim{\@tempdima >\@fpmin}{%
+ \global\@booleantrue\fcolmade@sw
+ }{%
+ \@cons\@failedlist #2%
+ }%
+ \endgroup
+ \fcolmade@sw{%
+ \let \@elt \@gobble
+ }{}%
+ }%
+}%
+\def\@ztryfc #1#2{%
+ \@tempcnta \count#2%
+ \divide\@tempcnta\@xxxii\multiply\@tempcnta\@xxxii
+ \@bitor \@tempcnta {\@failedlist \@flfail}%
+ \@testfp #2%
+ #1#2%
+ \@tempdimb\@tempdima
+ \advance\@tempdimb \ht#2\advance\@tempdimb\@fpsep
+ \@ifdim{\@tempdimb >\@colht}{%
+ \@testtrue
+ }{}%
+ \@if@sw\if@test\fi{%
+ \@cons\@flfail #2%
+ }{%
+ \@cons\@flsucceed #2%
+ \@tempdima\@tempdimb
+ }%
+}%
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+%
+% \subsection{Clearing pages}%
+%
+% Clearing the page is an elaboration of ending the page: it entails flushing all floats.
+%
+% This package might make number of float flushing algorithms available,
+% a very simple one that does not try to produce excellent pages,
+% another that tries to make the best use of space,
+% and a more complex one that tries to balance columns.
+%
+% At the beginning of the page-clearing process, by definition all of the paragraph text involved is on the MVL and all floats have been encountered.
+% There may be material in \cmd\pagesofar, and (in a multi-column page grid) any number of columns of the page have been composed.
+% Also, there might be footnote material saved up in \cmd\footbox.
+%
+% Because we did not want to perform multiple \cmd\shipout s per visit to the output routine,
+% our multi-column page makeup will not compose multiple columns per visit.
+% This implementation detail may not require alteration, but it is not a limitation that is truly necessary:
+% it is only multiple \cmd\shipout s per visit that must be avoided.
+%
+% The crux matter is how to continue with flushing floats even after the material in the MVL is exhausted.
+% At that point, we must, upon completion of the output routine,
+% insert into the MVL an interrupt that triggers the next step in the processing.
+%
+% Therefore, after processing a \cmd\do@startcolumn\ interrupt, we must somehow force the completion of that column.
+% This could be done by inserting a \cmd\do@newpage@pen\ interrupt.
+%
+% And after processing a \cmd\do@startpage@pen\ interrupt, that results in \cmd\@dbltopinsert s,
+% we must ensure that the multiple columns on the page get completed, so that the page itself finally gets shipped out.
+% This part will proceed automatically given that \cmd\do@startcolumn\ processing completes successfully.
+%
+% The process will not be complete until all deferred floats have been placed and shipped out, and all saved-up footnotes have been inserted.
+%
+% Full-page-width floats can get out of order of column floats. This problem can be remedied by holding them all in the same list.
+% We therefore stop using \cmd\@dbldeferlist\ entirely, and all of the procedures that formerly used it have been rewritten to
+% use \cmd\@deferlist\ instead. When traversing the list, we apply a selector on the given box that determines whether it is a column-width or page-width float.
+% This selector is different depending on the page grid.
+%
+% When the \cmd\@deferlist\ is processed (by any means), we have to take care of the case where a float of one category is passed over but we are looking for a float of the other category.
+% Here, we must terminate processing, to avoid disordering the floats. This we do by the usual means.
+%
+% The system has a Boolean that says we are clearing pages: \cmd\clearpage@sw; if it is true,
+% then at the tail of \cmd\do@startcolumn\ processing, we should put down a (\cmd\vfil?) \cmd\do@newpage@pen\ interrupt.
+% This is because the MVL is now empty, so we have to force the columns to complete.
+%
+% One potential very pathological case would be where there is one or more deferred floats that never successfully get placed:
+% placing floats has stalled, and we will ship out blank pages indefinitely. How to detect this case?
+%
+% First, \cmd\do@startpage\ will evidently be stalled if the following are all true:
+% a) \cmd\@tryfcolumn\ and \cmd\@sdblcolelt\ both fail,
+% b) there are deferred floats available for page placement, and
+% c) the \cmd\@colht=\cmd\textheight, that is, the full page height is available for placement of column floats.
+%
+% Second, \cmd\do@startcolumn\ will evidently be stalled if the following are all true:
+% a) tryfcolumn fails,
+% b) there are deferred floats available for column placement, and
+% a) the \cmd\@colroom=\cmd\textheight, that is, the full page height is available for placement of column floats.
+%
+%
+% \begin{macro}{\cleardoublepage}
+% \begin{macro}{\clearpage}
+% \begin{macro}{\newpage}
+% \begin{macro}{\newpage@prep}
+% The function of \cmd\clearpage\ is to end the current page with \cmd\newpage\ and then
+% ship out additional pages until (\footins) inserts and (deferred) floats are exhausted.
+%
+% The method involves setting the float placement parameters to completely permissive values
+% and kicking out the current page (using a non-discardable penalty).
+% A possibly short page will be shipped out, followed
+% by any number of float pages. However these float pages, because using permissive float placement,
+% will exhaust all inserts and deferred floats.
+%
+% Bug Note: in the code for \cmd\clearpage, the first penalty we output is an unprotected \cmd\pagebreak@pen.
+% I tried using a protected \cmd\do@newpage@pen, but that gave rise to a corner case where a blank page
+% was output.
+%
+% At present, the \cmd\clearpage\ procedure does the same as \cmd\newpage,
+% except that \cmd\clearpage@sw\ is turned on,
+% and the (discardable) \cmd\newpage\ is inevitably followed by the same procedures
+% that are executed if a page is shipped out.
+%
+% FIXME: it seems that better than \cmd\pagebreak@pen\ would be an unprotected penalty of a special value that would
+% entail output routine processing consisting of the following steps:
+% 3) \cmd\unvbox\cmd\@cclv,
+% 1) set \cmd\clearpage@sw\ to \cmd\true@sw,
+% 2) put down a protected \cmd\do@startcolumn@pen,
+% 4) take a dead cycle.
+%
+% The effect would be to liberalize float placement options for the current column as well as further columns that may be output as part of \cmd\clearpage\ processing.
+% Of course, it would still be necessary to set \cmd\clearpage@sw\ again via an interrupt.
+%
+% An optimization might be to clear \cmd\clearpage@sw\ as part of the same interrupt,
+% but that would actually not work properly, because it is necessary for \cmd\do@endpage\ to
+% possibly invoke furhter visits to the output routine before clearpage processing ceases.
+% \begin{macrocode}
+\def\newpage@prep{%
+ \if@noskipsec
+ \ifx \@nodocument\relax
+ \leavevmode
+ \global \@noskipsecfalse
+ \fi
+ \fi
+ \if@inlabel
+ \leavevmode
+ \global \@inlabelfalse
+ \fi
+ \if@nobreak \@nobreakfalse \everypar{}\fi
+ \par
+}%
+\def \newpage {%
+ \newpage@prep
+ \do@output@MVL{%
+ \vfil
+ \penalty-\pagebreak@pen
+ }%
+}%
+\def\clearpage{%
+ \newpage@prep
+ \do@output@MVL{%
+ \vfil
+ \penalty-\pagebreak@pen
+ \global\@booleantrue\clearpage@sw
+ \protect@penalty\do@startcolumn@pen
+ \protect@penalty\do@endpage@pen
+ }%
+ \do@output@MVL{%
+ \global\@booleanfalse\clearpage@sw
+ }%
+}%
+\def\cleardoublepage{%
+ \clearpage
+ \@if@sw\if@twoside\fi{%
+ \@ifodd\c@page{}{%
+ \null\clearpage
+ }%
+ }{}%
+}%
+\@booleanfalse\clearpage@sw
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\do@endpage@pen}
+% The penalty \cmd\do@endpage@pen\ simply dispatches to the page grid procedure that forces an end page.
+% That procedure should test whether there is anything to ship out (say committed floats), then act accordingly.
+% Note that as part of this work, it should \cmd\unvbox\cmd\@cclv, which has been left boxed up so it can be
+% measured.
+% \begin{macrocode}
+\mathchardef\do@endpage@pen=10007
+\@namedef{output@-\the\do@endpage@pen}{%
+ \csname end@column@\thepagegrid\endcsname
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\do@newpage@pen}
+% The penalty \cmd\do@newpage@pen\ allows a ``non-discardable \cmd\newpage'' command:
+% a \cmd\newpage\ command that will not disappear at a pagebreak.
+% This visit to the output routine will not be dispatched to an interrupt,
+% rather the natural output routine will be executed, where it
+% will remove the protection box.
+%
+% Call this routine by executing \cmd\protect@penalty\cmd\do@newpage@pen.
+% \begin{macrocode}
+\mathchardef\do@newpage@pen=10001
+\expandafter\let\csname output@-\the\do@newpage@pen\endcsname\relax
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\@clearfloatplacement}
+% The procedure \cmd\@clearfloatplacement\ sets all of the float placement parameters
+% to completely permissive values. The standard values appear as comments.
+% \begin{macrocode}
+\def\@clearfloatplacement{%
+ \global\@topnum \maxdimen % \c@topnumber
+ \global\@toproom \maxdimen % \topfraction\@colht
+ \global\@botnum \maxdimen % \c@bottomnumber
+ \global\@botroom \maxdimen % \bottomfraction\@colht
+ \global\@colnum \maxdimen % \c@totalnumber
+%\global\@fpmin \z@ % \floatpagefraction\@colht
+ \global\@dbltopnum \maxdimen % \c@dbltopnumber
+ \global\@dbltoproom \maxdimen % \dbltopfraction\@colht
+ \global\@textmin \z@ % \@colht\advance \@textmin -\@dbltoproom
+ \global\@fpmin \z@ % \dblfloatpagefraction\textheight
+ \let\@testfp\@gobble
+ \appdef\@setfloattypecounts{\@fpstype16\advance\@fpstype\m@ne}%
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\@doclearpage}
+% The \cmd\@doclearpage\ procedure is now obsoleted, as is \cmd\@makefcolumn, which it invoked.
+% \begin{macrocode}
+\let\@doclearpage\@undefined
+\let\@makefcolumn\@undefined
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\clr@top@firstmark}
+% \begin{macro}{\set@top@firstmark}
+% \begin{macro}{\@outputpage}
+% We want accurate values of \cmd\topmark\ and \cmd\firstmark, but we must deal with
+% the fact that there are many different ways of contributing material
+% to the page. Only upon the first contribution to the page is the value of \cmd\topmark\
+% accurate. However, with \cmd\firstmark\ we must potentially examine each contribution
+% because the first mark on the page may happen to fall in the last piece of material contributed.
+%
+% To begin, we define the procedure that initializes the macros to appropriate flag values.
+% \begin{macrocode}
+\def\clr@top@firstmark{%
+ \global\let\saved@@topmark\@undefined
+ \global\let\saved@@firstmark\@empty
+ \global\let\saved@@botmark\@empty
+}%
+\clr@top@firstmark
+% \end{macrocode}
+%
+% Note that the flag value for \cmd\saved@@topmark\ is \cmd\@undefined, just as one would
+% expect. But that for \cmd\saved@@firstmark\ and \cmd\saved@@botmark\ is \cmd\@empty.
+%
+% Next, we define procedure \cmd\set@top@firstmark; it will be exercised everywhere material is contributed,
+% capturing the mark values if appropriate.
+% \begin{macrocode}
+\def\set@top@firstmark{%
+ \@ifxundefined\saved@@topmark{\expandafter\gdef\expandafter\saved@@topmark\expandafter{\@@topmark}}{}%
+ \@if@empty\saved@@firstmark{\expandafter\gdef\expandafter\saved@@firstmark\expandafter{\@@firstmark}}{}%
+ \@if@empty\@@botmark{}{\expandafter\gdef\expandafter\saved@@botmark\expandafter{\@@botmark}}%
+}%
+% \end{macrocode}
+% When should \cmd\set@top@firstmark\ be called?
+% A good candidate for a universal procedure for handling contributed material is
+% the natural output routine; are any other calls needed?
+%
+% Yes, in \cmd\save@column\ we must execute \cmd\set@top@firstmark\ because we are about to
+% save away \cmd\box\cmd\@cclv, and we will never see its marks again (unless it is unboxed into the MVL),
+% because \TeX\ lets one access a box's marks only within an output routine that has put that box into \cmd\box\cmd\@cclv.
+%
+% As soon as a page is shipped out, we initialize the two macros that
+% hold the values of \cmd\topmark\ and \cmd\firstmark, respectively.
+% \LaTeX\ has exactly one procedure \cmd\@outputpage\ that does \cmd\shipout,
+% which is as it should be: we tailpatch it, and the job is done.
+% \begin{macrocode}
+\appdef\@outputpage{%
+ \clr@top@firstmark
+}%
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+%
+% \subsection{Other interfaces to \LaTeX}%
+%
+% \begin{macro}{\@float}
+% \begin{macro}{\@dblfloat}
+% \begin{macro}{\@yfloat}
+% \begin{macro}{\fps@}
+% \begin{macro}{\fpsd@}
+% The \LaTeX\ kernel procedures \cmd\@float\ and \cmd\@dblfloat\ are treated on an equal footing.
+% Each now takes environment-specific float placement defaults.
+% If none are defined for the calling environment, we apply a default.
+%
+% A parameter is passed that will set the width of text within the float, normally \cmd\columnwidth,
+% and in the "dbl" version, \cmd\textwidth. However, an environment such as \env{turnpage}
+% may change the meanings of these macros to allow turnpage floats.
+%
+% \begin{macrocode}
+\def\@float#1{%
+ \@ifnextchar[{%}]{%Brace-matching klootch
+ \@yfloat\width@float{#1}%
+ }{%
+ \@ifxundefined@cs{fps@#1}{%
+ \edef\reserved@a{\noexpand\@yfloat\noexpand\width@float{#1}[\csname fps@\endcsname]}\reserved@a
+ }{%
+ \edef\reserved@a{\noexpand\@yfloat\noexpand\width@float{#1}[\csname fps@#1\endcsname]}\reserved@a
+ }%
+ }%
+}%
+\def\@dblfloat#1{%
+ \@ifnum{\pagegrid@col=\@ne}{%
+ \@float{#1}%
+ }{%
+ \@ifnextchar[{%}]{%Brace-matching klootch
+ \@yfloat\widthd@float{#1}%
+ }{%
+ \@ifxundefined@cs{fpsd@#1}{%
+ \edef\reserved@a{\noexpand\@yfloat\noexpand\widthd@float{#1}[\csname fpsd@\endcsname]}\reserved@a
+ }{%
+ \edef\reserved@a{\noexpand\@yfloat\noexpand\widthd@float{#1}[\csname fpsd@#1\endcsname]}\reserved@a
+ }%
+ }%
+ }%
+}%
+\def\@yfloat#1#2[#3]{%
+ \@xfloat{#2}[#3]%
+ \hsize#1\linewidth\hsize
+ \minipagefootnote@init
+}%
+\def\fps@{tbp}%
+\def\fpsd@{tp}%
+\def\width@float{\columnwidth}%
+\def\widthd@float{\textwidth}%
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\end@float}
+% \begin{macro}{\end@dblfloat}
+% \begin{macro}{\end@@float}
+% \begin{macro}{\check@currbox@count}
+% \begin{macro}{\minipagefootnote@init}
+% \begin{macro}{\minipagefootnote@here}
+% \LaTeX\ kernel procedures \cmd\end@float\ and \cmd\end@dblfloat\
+% have been changed to work alike; in particular, floats of both classes
+% are deferred into the same queue.
+% This measure ensures that they will be placed in their original order,
+% an aspect in which \LaTeX\ is broken.
+%
+% Note: when retrieving floats from the queues, we can differentiate those of the two categories
+% by the width of the box.
+%
+% Floats are processed via an output routine message, and are checked for
+% sanity in re the float placement options. In the case of full-page-width floats,
+% we ensure that the h and b float placement options are never asserted, because they
+% make no sense.
+%
+% Note that if we get to the end of the float box and still have pending
+% footnotes, we put then out.
+%
+% LaTeX Bug note: if a user types \cmd\begin{table*}[h], the float will never succeed in being placed!
+% we try to catch such cases.
+%
+% Note that the macro \cmd\check@currbox@count\ trys to catch cases where the float placement options
+% are such that the float can never be placed.
+% \begin{macrocode}
+\def\end@float{%
+ \end@@float{%
+ \check@currbox@count
+ }%
+}%
+\def\end@dblfloat{%
+ \@ifnum{\pagegrid@col=\@ne}{%
+ \end@float
+ }{%
+ \end@@float{%
+ \@boxfpsbit\@currbox{1}\@ifodd\@tempcnta{\global\advance\count\@currbox\m@ne}{}%
+ \@boxfpsbit\@currbox{4}\@ifodd\@tempcnta{\global\advance\count\@currbox-4\relax}{}%
+ \global\wd\@currbox\textwidth % Klootch
+ \check@currbox@count
+ }%
+ }%
+}%
+\def\end@@float#1{%
+ \minipagefootnote@here
+%\minipagefootnotes
+ \@endfloatbox
+ #1%
+ \@ifnum{\@floatpenalty <\z@}{%
+ \@largefloatcheck
+ \@cons\@currlist\@currbox
+ \@ifnum{\@floatpenalty <-\@Mii}{%
+ \do@output@cclv{\@add@float}%
+ }{%
+ \vadjust{\do@output@cclv{\@add@float}}%
+ \@Esphack
+ }%
+ }{}%
+}%
+\def\check@currbox@count{%
+ \@ifnum{\count\@currbox>\z@}{%
+ \count@\count\@currbox\divide\count@\sixt@@n\multiply\count@\sixt@@n
+ \@tempcnta\count\@currbox\advance\@tempcnta-\count@
+ \@ifnum{\@tempcnta=\z@}{%
+ \ltxgrid@warn{Float cannot be placed}%
+ }{}%
+ }{%
+ % Is a \marginpar
+ }%
+}%
+\providecommand\minipagefootnote@init{}%
+\providecommand\minipagefootnote@here{}%
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\@specialoutput}
+% The \cmd\@add@float\ procedure used to reside in standard \LaTeX's \cmd\@specialoutput,
+% which is no more.
+%
+% Historical Note: \cmd\@specialoutput\ and Lamport's method of an output routine dispatcher
+% is the genesis of our more powerful and refined way of using \TeX's output routine to
+% safely accomplish page makeup tasks. To it and to him we owe acknowledgement and thanks.
+% \begin{macrocode}
+\let\@specialoutput\@undefined
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\@add@float}
+% In the following, we do not need to execute \cmd\@reinserts, which was wrong anyway, as you cannot
+% reliably recover insertions when they split (unless you have a way of reinserting the captured insertion
+% ahead of the split-off part).
+%
+% Now that full-page-width floats are being processed the same as column floats, we
+% have to nip in here and cause them always to be deferred.
+%
+% At the very end, the \cmd\vsize\ is adjusted for any newly committed float.
+% \begin{macrocode}
+\def\@add@float{%
+ \@pageht\ht\@cclv\@pagedp\dp\@cclv
+ \unvbox\@cclv
+ \@next\@currbox\@currlist{%
+ \csname @floatselect@sw@\thepagegrid\endcsname\@currbox{%
+ \@ifnum{\count\@currbox>\z@}{%
+ \advance \@pageht \@pagedp
+ \advance \@pageht \vsize \advance \@pageht -\pagegoal % do not assume \holdinginserts is cleared!
+% \@ifvbox\@kludgeins{%
+% \@ifdim{\wd\@kludgeins=\z@}{%
+% \advance \@pageht \ht\@kludgeins
+% }{}%
+% }{}%
+% \@reinserts
+ \@addtocurcol % Commit an h float
+ }{%
+% \@reinserts
+ \@addmarginpar
+ }%
+ }{%
+ \@resethfps
+ \@cons\@deferlist\@currbox
+ }%
+ }{\@latexbug}%
+ \@ifnum{\outputpenalty<\z@}{%
+ \@if@sw\if@nobreak\fi{%
+ \nobreak
+ }{%
+ \addpenalty \interlinepenalty
+ }%
+ }{}%
+ \set@vsize
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\@reinserts}
+% The \cmd\@reinserts\ procedure of standard \LaTeX\ is now obsoleted.
+% \begin{macrocode}
+\let\@reinserts\@undefined
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\@addtocurcol}
+% We modify the \cmd\@addtocurcol\ procedure of standard \LaTeX\
+% so that a float placed ``here'' may break over pages.
+% \begin{macrocode}
+\def \@addtocurcol {%
+ \@insertfalse
+ \@setfloattypecounts
+ \ifnum \@fpstype=8
+ \else
+ \ifnum \@fpstype=24
+ \else
+ \@flsettextmin
+ \advance \@textmin \@textfloatsheight
+ \@reqcolroom \@pageht
+ \ifdim \@textmin>\@reqcolroom
+ \@reqcolroom \@textmin
+ \fi
+ \advance \@reqcolroom \ht\@currbox
+ \ifdim \@colroom>\@reqcolroom
+ \@flsetnum \@colnum
+ \ifnum \@colnum>\z@
+ \@bitor\@currtype\@deferlist
+ \if@test
+ \else
+ \@bitor\@currtype\@botlist
+ \if@test
+ \@addtobot
+ \else
+ \ifodd \count\@currbox
+ \advance \@reqcolroom \intextsep
+ \ifdim \@colroom>\@reqcolroom
+ \global \advance \@colnum \m@ne
+ \global \advance \@textfloatsheight \ht\@currbox
+ \global \advance \@textfloatsheight 2\intextsep
+ \@cons \@midlist \@currbox
+ \if@nobreak
+ \nobreak
+ \@nobreakfalse
+ \everypar{}%
+ \else
+ \addpenalty \interlinepenalty
+ \fi
+ \vskip \intextsep
+ \unvbox\@currbox %AO
+ \penalty\interlinepenalty
+ \vskip\intextsep
+ \ifnum\outputpenalty <-\@Mii \vskip -\parskip\fi
+ \outputpenalty \z@
+ \@inserttrue
+ \fi
+ \fi
+ \if@insert
+ \else
+ \@addtotoporbot
+ \fi
+ \fi
+ \fi
+ \fi
+ \fi
+ \fi
+ \fi
+ \if@insert
+ \else
+ \@resethfps
+ \@cons\@deferlist\@currbox
+ \fi
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\if@twocolumn}
+% The \cmd\newif\ switch \cmd\if@twocolumn\ is entirely unused. However its access words are invoked by
+% \LaTeX's \cmd\document\ procedure, so we de-fang it.
+% \begin{macrocode}
+\@twocolumnfalse
+\let\@twocolumntrue\@twocolumnfalse
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\@addmarginpar}
+% The procedure \cmd\@addmarginpar\ used to access \cmd\if@twocolumn, but that switch is not reliable;
+% the better way is to use \cmd\thepagegrid. We establish a convention for a page-grid-oriented
+% procedure, e.g., \cmd\@addmarginpar@one, that emits a boolean, telling this procedure
+% whether to set the marginpar on the left or right.
+% \begin{macrocode}
+\def\@addmarginpar{%
+ \@next\@marbox\@currlist{%
+ \@cons\@freelist\@marbox\@cons\@freelist\@currbox
+ }\@latexbug
+ \setbox\@marbox\hb@xt@\columnwidth{%
+ \csname @addmarginpar@\thepagegrid\endcsname{%
+ \hskip-\marginparsep\hskip-\marginparwidth
+ \box\@currbox
+ }{%
+ \hskip\columnwidth\hskip\marginparsep
+ \box\@marbox
+ }%
+ \hss
+ }%
+ \setbox\z@\box\@currbox
+ \@tempdima\@mparbottom
+ \advance\@tempdima -\@pageht
+ \advance\@tempdima\ht\@marbox
+ \@ifdim{\@tempdima >\z@}{%
+ \@latex@warning@no@line {Marginpar on page \thepage\space moved}%
+ }{%
+ \@tempdima\z@
+ }%
+ \global\@mparbottom\@pageht
+ \global\advance\@mparbottom\@tempdima
+ \global\advance\@mparbottom\dp\@marbox
+ \global\advance\@mparbottom\marginparpush
+ \advance\@tempdima -\ht\@marbox
+ \global\setbox \@marbox
+ \vbox {\vskip \@tempdima
+ \box \@marbox}%
+ \global \ht\@marbox \z@
+ \global \dp\@marbox \z@
+ \kern -\@pagedp
+ \nointerlineskip
+ \box\@marbox
+ \nointerlineskip
+ \hbox{\vrule \@height\z@ \@width\z@ \@depth\@pagedp}%
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{environment}{turnpage}
+% Any float (viz., \env{figure} or \env{table}) within the scope of this environment
+% will be a turnpage float: It will be assumed to occupy an entire page (constitute a float page),
+% the width will be \cmd\textheight, the height \cmd\textwidth, and the entire float will be presented
+% rotated 90 degrees.
+%
+% The implementation requires the services of the \cmd\rotatebox\ command, so we supply a dummy definition
+% that explains things to the user.
+% \begin{macrocode}
+\newenvironment{turnpage}{%
+ \def\width@float{\textheight}%
+ \def\widthd@float{\textheight}%
+ \appdef\@endfloatbox{%
+ \@ifxundefined\@currbox{%
+ \ltxgrid@warn{Cannot rotate! Not a float}%
+ }{%
+ \setbox\@currbox\vbox to\textwidth{\vfil\unvbox\@currbox\vfil}%
+ \global\setbox\@currbox\vbox{\rotatebox{90}{\box\@currbox}}%
+ }%
+ }%
+}{%
+}%
+\def\rotatebox@dummy#1#2{%
+ \ltxgrid@warn{You must load the graphics or graphicx package in order to use the turnpage environment}%
+ #2%
+}%
+\AtBeginDocument{%
+ \@ifxundefined\rotatebox{\let\rotatebox\rotatebox@dummy}{}%
+}%
+% \end{macrocode}
+% \end{environment}
+%
+%
+% \subsection{One-off output routines}
+%
+% These procedures are executed in lieu of \cmd\the\cmd\output\ when the output penalty has the associated flag value.
+%
+% \begin{macro}{output@-1073741824}
+% The first one-off output routine handles the end of the job, wherein
+% \LaTeX\ executes \cmd\@@end, and breaks to the output with a penalty of
+% $"40000000 = 2^{32}/4$. We simply discard \cmd\box\cmd\@cclv\ and leave.
+% This means that \LaTeX\ is obligated to do \cmd\clearpage\ as part of
+% its \enve{document} processing, otherwise material will be lost.
+% \begin{macrocode}
+\@namedef{output@-1073741824}{%"40000000
+ \deadcycles\z@
+%\showbox\@cclv
+ \setbox\z@\box\@cclv
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\save@column@pen}
+% The one-off output routine associated with \cmd\penalty\cmd\save@column@pen\
+% will be called within a sequence of three such routines by \cmd\execute@message
+% or its companion routine \cmd\execute@message@insert.
+% This procedure must save away any the current page and preserve marks.
+% \begin{macrocode}
+\mathchardef\save@column@pen=10016
+\@namedef{output@-\the\save@column@pen}{\save@column}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\@cclv@saved}
+% We take over the \cmd\@holdpg\ box register. Hereafter,
+% we no longer use the \cmd\@holdpg\ box register, so let the world know.
+% This should decisively break packages that assume standard \LaTeX.
+% Breaking decisively is preferred to quietly proceeding erroneously.
+% \begin{macrocode}
+\let \@cclv@saved \@holdpg
+\let \@holdpg \@undefined
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\save@column}
+% The procedure \cmd\save@column\ does the actual work of saving away the material
+% on the page. It is invoked both by \cmd\save@column@pen\ and by \cmd\save@column@insert@pen.
+% We save \cmd\box\cmd\@cclv\ and the primitive \cmd\@@topmark.
+% \begin{macrocode}
+\def\save@column{%
+ \@ifvoid\@cclv@saved{%
+ \set@top@firstmark
+ \global\@topmark@saved\expandafter{\@@topmark}%
+ }{}%
+ \global\setbox\@cclv@saved\vbox{%
+ \@ifvoid\@cclv@saved{}{%
+ \unvbox\@cclv@saved
+ \marry@baselines
+ }%
+ \unvbox\@cclv
+ \lose@breaks
+ \setbox\z@\lastbox
+ }%
+}%
+\newtoks\@topmark@saved
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\prep@cclv}
+% The procedure \cmd\prep@cclv\ is used by message handlers to set up their environment
+% to ape that of the usual output routine, with the boxed-up page in \cmd\box\cmd\@cclv.
+% Here, we retrieve the material from \cmd\@cclv@saved, where it was saved away by
+% the one-off output routine associated with \cmd\save@column@pen.
+% \begin{macrocode}
+\def\prep@cclv{%
+ \setbox\z@\box\@cclv
+ \setbox\@cclv\box\@cclv@saved
+ \vbadness\@M
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\save@column@insert@pen}
+% The one-off output routine associated with \cmd\penalty\cmd\save@column@insert@pen\
+% is similar to that of \cmd\save@column@pen\ augmented with the processing of insertions.
+% It is called by \cmd\execute@message@insert\ (i.e., at a grid change)
+% and saves away the current page and preserves marks.
+% In addition, it saves away any insertions that fall on the current page.
+% As with the regular output routine, it executes in two phases,
+% first with \cmd\holdinginserts\ set, then with it cleared.
+% \begin{macrocode}
+\mathchardef\save@column@insert@pen=10017
+\@namedef{output@-\the\save@column@insert@pen}{\toggle@insert\savecolumn@holding\savecolumn@moving}%
+% \end{macrocode}
+% The procedure \cmd\savecolumn@holding\ is the first phase of saving a column with its inserts.
+% This phase must detect and remedy the one circumstance that will confound our efforts to propagate marks.
+% It is similar to \cmd\output@holding, except that we have to deal with the protection box, which must
+% remain, because the messaging mechanism is being used.
+%
+% If it appears that we have the pathological ``Big Bad Box'' case at hand, we use the \cmd\dead@cycle@repair@protected\
+% procedure instead of \cmd\dead@cycle\ to do our dead cycle.
+% \begin{macrocode}
+\def\savecolumn@holding{%
+ \@if@exceed@pagegoal{\unvcopy\@cclv\setbox\z@\lastbox}{%
+ \setbox\z@\vbox{\unvcopy\@cclv\setbox\z@\lastbox}%
+ \outputdebug@sw{{\tracingall\scrollmode\showbox\z@}}{}%
+ \dimen@\ht\@cclv\advance\dimen@-\ht\z@
+ \dead@cycle@repair@protected\dimen@
+ }{%
+ \dead@cycle
+ }%
+}%
+% \end{macrocode}
+% The procedure \cmd\save@column@moving\ is the second phase of saving a column with its inserts.
+% Now that \cmd\holdinginserts\ is cleared, we can look in the various \cmd\insert\ registers for
+% our inserts (at present there is only one, \cmd\footins).
+% if anything is there, we save it away and ask for another cycle (because it may have split).
+%
+% Note that the message that is about to be executed had better deal properly with the contents of
+% the \cmd\footins@saved\ box.
+% \begin{macrocode}
+\def\savecolumn@moving{%
+ \@cclv@nontrivial@sw{%
+ \save@column
+ }{%
+ {\setbox\z@\box\@cclv}%
+ }%
+ \@ifvoid\footins{}{%
+ \outputdebug@sw{{\tracingall\scrollmode\showbox\footins}}{}%
+ \global\setbox\footins@saved\vbox{\unvbox\footins@saved\marry@baselines\unvbox\footins}%
+ \protect@penalty\save@column@insert@pen
+ }%
+}%
+\newbox\footins@saved
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\save@message@pen}
+% The one-off output routine associated with \cmd\penalty\cmd\save@message@pen\
+% saves away the message that has been passed.
+% This procedure is penultimate in a sequence of one-off output routine calls;
+% earlier ones have saved away the MVL and preserved marks, the last executes the message.
+%
+% Note that we are passing tokens to \TeX's primitive \cmd\mark\ mechanism, so we must ensure
+% that they are not inappropriately expanded. We use the same mechanism for all such cases,
+% namely \cmd\let@mark.
+%
+% Note: we expect that \cmd\box\cmd\@cclv's contents are well known:
+% \cmd\topskip, protection box, and a \cmd\mark, the latter havin the
+% message.
+% But if we came here via \cmd\penalty 10017, there might be an \cmd\insert\ node here as well, because
+% a footnote split.
+% Because this procedure simply voids out \cmd\box\cmd\@cclv, such material would be lost.
+% Perhaps we can repair things by manipulating the \cmd\insert \ mechanism temporarily.
+% \begin{macrocode}
+\mathchardef\save@message@pen=10018
+\@namedef{output@-\the\save@message@pen}{\save@message}%
+\def\save@message{%
+ \setbox\z@\box\@cclv %FIXME: what if \box\@cclv is not empty?
+ \toks@\expandafter{\@@firstmark}%
+ \expandafter\gdef\expandafter\@message@saved\expandafter{\the\toks@}%
+ \expandafter\do@@mark\expandafter{\the\@topmark@saved}%
+}%
+\gdef\@message@saved{}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\execute@message@pen}
+% The one-off output routine associated with \cmd\execute@message@pen\
+% simply executes the given message. It is last in a sequence of one-off output routine calls;
+% earlier ones have saved all that require saving.
+% \begin{macrocode}
+\mathchardef\execute@message@pen=10019
+\@namedef{output@-\the\execute@message@pen}{\@message@saved}%
+% \end{macrocode}
+% \end{macro}
+%
+%
+% \subsection{Output messages}
+%
+% Message handlers are procedures that execute output messages, tokens that are passed to
+% the output routine for execution in an environment appropriate to page makeup.
+%
+% How it works. We put down three large negative penalties, each of which will be handled by
+% the output dispatcher (not the output routine), each penalty being protected by
+% a removable, non-discardable item (i.e., a box).
+% Either three or four invocations of one-off output routines are involved per message.
+%
+% We make the last of the three protection boxes have a depth equal to the value of \cmd\prevdepth\
+% that was current when the procedure is called. This effectively restores \cmd\prevdepth.
+%
+% In each case, the one-off output routine will remove the extraneous box we have inserted.
+% And the second and third one-off routines will simply void \cmd\box\cmd\@cclv, because its contents
+% are entirely artificial.
+%
+% FIXME: not so! If \cmd\holdinginserts\ is cleared, that box may have an insert node; it must be preserved, too.
+%
+% The first routine saves away the current column contents and remembers the \cmd\topmark\ for later
+% use. There is a variant routine that first clears \cmd\holdinginserts, so that the message can
+% handle any inserts present in the boxed-up page; this of course entails yet another
+% visit to the output routine.
+%
+% The penultimate routine saves away the tokens transmitted in via the \cmd\@@mark: the argument of
+% the macro. These tokens are of course the very thing we wish to execute within the safety
+% of the output routine. It also puts down a mark containing the \cmd\topmark\ tokens saved
+% by the first routine. By this means, the mark, which we have clobbered, is restored.
+%
+% The last routine simply executes the given tokens.
+% In the course of doing this, it must take care of \cmd\box\cmd\@cclv, either by shipping it out,
+% or by \cmd\unvbox ing it onto the MVL.
+%
+% \begin{macro}{\execute@message}
+% The procedure \cmd\execute@message\ simply calls the utility procedure \cmd\@execute@message\
+% with a penalty value for the standard treatment.
+% \begin{macrocode}
+\def\execute@message{%
+ \@execute@message\save@column@pen %Implicit #2
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\execute@message@insert}
+% The procedure \cmd\execute@message@insert\ is like \cmd\execute@message\ in all respects
+% except that the penalty value is \cmd\save@column@insert@pen, which arranges for the
+% message handler involved to deal with the page's insertions.
+% At the same time, we prepare the \cmd\footins\ box so that these insertions can be dealt with.
+%
+% Note:
+% If more insertions are added to \LaTeX\ (presumably via \cmd\newinsert), then they
+% must be dealt with in a way entirely analogous to \cmd\footins.
+% \begin{macrocode}
+\def\execute@message@insert#1{%
+ \@execute@message\save@column@insert@pen{\setbox\footins\box\footins@saved#1}%
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\@execute@message}
+% The utility procedure \cmd\@execute@message\ is called by \cmd\execute@message\ and \cmd\execute@message@insert.
+% We prepare by creating a \cmd\vbox\ containing all the needed nodes and proceed by simply
+% \cmd\unvbox ing that box onto the MVL.
+% We ensure that \cmd\box\cmd\@cclv\ is properly set up for the output message handler
+% by always inserting \cmd\prep@cclv\ in advance of the argument.
+%
+% Note that each one-off output routine is invoked effectively the same as
+% \cmd\protect@penalty, except that the second invocation involves an additional
+% \cmd\mark\ node, and the third a specially prepared protection box.
+%
+% Note also that \TeX's primitive \cmd\mark\ is called here without any expansion protection.
+% This is the only place where it is called that way, but it's OK because those tokens
+% have have been pre-expanded by procedures that call \cmd\execute@message.
+% FIXME: all procedures calling \cmd\execute@message\ must pre-expand their tokens!
+% \begin{macrocode}
+\long\def\@execute@message#1#2{%
+ \begingroup
+ \dimen@\prevdepth\@ifdim{\dimen@<\z@}{\dimen@\z@}{}%
+ \setbox\z@\vbox{%
+ \protect@penalty#1%
+ \protection@box
+ \toks@{\prep@cclv#2}%
+ \@@mark{\the\toks@}%
+ \penalty-\save@message@pen
+% \hbox{\vrule\@height\z@\@width\z@\@depth\dimen@}%
+ \setbox\z@\null\dp\z@\dimen@\ht\z@-\dimen@
+ \nointerlineskip\box\z@
+ \penalty-\execute@message@pen
+ }\unvbox\z@
+ \endgroup
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\do@output@cclv}
+% The procedure \cmd\do@output@cclv\ provides access to message handlers at their simplest.
+% The message will execute in the usual environment of the output routine, with
+% the boxed-up page in \cmd\box\cmd\@cclv, and we assume that \cmd\holdinginserts\ remains set.
+% This procedure must be invoked within main vertical mode;
+% it is the obligation of the macro writer to ensure that this is the case.
+% \begin{macrocode}
+\def\do@output@cclv{\execute@message}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\do@output@MVL}
+% The procedure \cmd\do@output@MVL, like \cmd\do@output@cclv, is an interface for messages,
+% but provides two additional services: the command may also be invoked in horizontal mode, and
+% the message handler will execute with the MVL unboxed.
+% \begin{macrocode}
+\def\do@output@MVL#1{%
+ \@ifvmode{%
+ \begingroup\execute@message{\unvbox\@cclv#1}\endgroup
+ }{%
+ \@ifhmode{%
+ \vadjust{\execute@message{\unvbox\@cclv#1}}%
+ }{%
+ \@latexerr{\string\do@output@MVL\space cannot be executed in this mode!}\@eha
+ }%
+ }%
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\lose@breaks}
+% The purpose of this procedure is to get rid of all the extraneous
+% \cmd\penalty\cmd\@M\ nodes that tend to build up in the MVL.
+% \begin{macrocode}
+\def\lose@breaks{%
+ \loopwhile{%
+ \count@\lastpenalty
+ \@ifnum{\count@=\@M}{% 10000 is a TeX magic number!
+ \unpenalty\true@sw
+ }{%
+ \false@sw
+ }%
+ }%
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\removestuff}
+% \cmd\removestuff\ is a document-level command that removes the bottom skip glue item
+% from the MVL.
+% \begin{macrocode}
+\def\removestuff{\do@output@MVL{\unskip\unpenalty}}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\removephantombox}
+% The procedure \cmd\removephantombox\ is a special-purpose message handler exclusively for
+% preventing incorrect spacing above display math. It must be issued in
+% horizontal mode within the phantom paragraph generated when display math starts up in
+% vertical mode.
+% \begin{macrocode}
+\def\removephantombox{%
+ \vadjust{%
+ \execute@message{%
+ \unvbox\@cclv
+ \setbox\z@\lastbox
+ \unskip
+ \unskip
+ \unpenalty
+ \penalty\predisplaypenalty
+ \vskip\abovedisplayskip
+ }%
+ }%
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\addstuff}
+% \cmd\addstuff\ is a document-level command that adds penalty, glue, or both to the
+% MVL. The penalty and glue items are rearranged so that all penalties nodes precede all the glue nodes,
+% which is the canonical arrangement.
+% \begin{macrocode}
+\def\addstuff#1#2{\edef\@tempa{\noexpand\do@output@MVL{\noexpand\@addstuff{#1}{#2}}}\@tempa}%
+\def\@addstuff#1#2{%
+ \skip@\lastskip\unskip
+ \count@\lastpenalty\unpenalty
+ \@if@empty{#1}{}{\penalty#1\relax}%
+ \@ifnum{\count@=\z@}{}{\penalty\count@}%
+ \vskip\skip@
+ \@if@empty{#2}{}{\vskip#2\relax}%
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\replacestuff}
+% \cmd\replacestuff\ is a document-level command similar to \cmd\addstuff;
+% but it replaces penalty, glue, or both in the MVL.
+% The penalty and glue items are rearranged so that all penalties nodes precede all the glue nodes,
+% which is the canonical arrangement.
+% \begin{macrocode}
+\def\replacestuff#1#2{\edef\@tempa{\noexpand\do@output@MVL{\noexpand\@replacestuff{#1}{#2}}}\@tempa}%
+\def\@replacestuff#1#2{%
+ \skip@\lastskip\unskip
+ \count@\lastpenalty\unpenalty
+ \@if@empty{#1}{}{%
+ \@ifnum{\count@>\@M}{}{%
+ \@ifnum{\count@=\z@}{\count@=#1\relax}{%
+ \@ifnum{\count@<#1\relax}{}{%
+ \count@=#1\relax
+ }%
+ }%
+ }%
+ }%
+ \@ifnum{\count@=\z@}{}{\penalty\count@}%
+ \@if@empty{#2}{}{%
+ \@tempskipa#2\relax
+ \@ifdim{\z@>\@tempskipa}{%
+ \advance\skip@-\@tempskipa
+ }{%
+ \@ifdim{\skip@>\@tempskipa}{}{%
+ \skip@\@tempskipa
+ }%
+ }%
+ }%
+ \vskip\skip@
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\move@insertions}
+% \begin{macro}{\hold@insertions}
+% \begin{macro}{\move@insert@sw}
+% In order to avoid bolluxing up \cmd\insert\ registers by our one-off
+% output routines, we set \cmd\holdinginserts\ to zero by default and only
+% clear it (briefly) while we handle cases where we want inserts to show up.
+% \begin{macrocode}
+\def\move@insertions{\global\holdinginserts\z@}%
+\def\hold@insertions{\global\holdinginserts\@ne}%
+\hold@insertions
+\def\move@insert@sw{\@ifnum{\holdinginserts=\z@}}%
+\def\toggle@insert#1#2{%
+ \@ifnum{\holdinginserts=\z@}{\hold@insertions#2}{\move@insertions#1}%
+}%
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+%
+% \subsection{Messages to alter the page grid}%
+%
+% Here is the implementation of the grid-switching procedures.
+% We perform two checks when changing the page grid; first to ensure that
+% the target page grid is known (defensive programming), second to ensure that
+% the switch is a non-trivial one. The latter check must be performed within
+% the safety of the output routine, so requires using an output message.
+% Thus, a grid change requires two messages, for a total of six visits to the output routine.
+%
+% \begin{macro}{\do@columngrid}
+% Utility procedure \cmd\do@columngrid\ changes the page grid.
+% Note that this command forces an end to the current paragraph. This is necessary, because
+% a page grid change makes no sense unless we can alter the \cmd\hsize\ before commencing to
+% typeset the following paragraph. So the command should never be executed in horizontal mode
+% anyway.
+% \begin{macrocode}
+\def\do@columngrid#1#2{%
+ \par
+ \expandafter\let\expandafter\@tempa\csname open@column@#1\endcsname
+ \@ifx{\relax\@tempa}{%
+ \ltxgrid@warn{Unknown page grid #1. No action taken}%
+ }{%
+ \do@output@MVL{\start@column{#1}{#2}}%
+ }%
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\start@column}
+% Procedure \cmd\start@column\ lays down the interrupts to switch the page grid.
+% If the change to the page grid would have been trivial, it bails out.
+% It seems a reasonable tradeoff of processing versus security: once we commit
+% to changing the page grid, we clear \cmd\holdinginserts, so there is no turning back.
+%
+% Note that the second argument to the macro allows us to pass an argument to the
+% page grid that is starting up. This can be handy, because a single procedure can
+% handle multiple page grids, differing only by the value of a parameter.
+%
+% FIXME: this means that you cannot switch between mlt page grids in a single step.
+% But do we want to do this, at all, at all?
+% \begin{macrocode}
+\def\start@column#1#2{%
+ \def\@tempa{#1}\@ifx{\@tempa\thepagegrid}{%
+ \ltxgrid@info{Already in page grid \thepagegrid. No action taken}%
+ }{%
+ \expandafter\execute@message@insert
+ \expandafter{%
+ \csname shut@column@\thepagegrid\expandafter\endcsname
+ \csname open@column@#1\endcsname{#2}%
+ \set@vsize
+ }%
+ }%
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\thepagegrid}
+% The macro \cmd\thepagegrid\ tracks what kind of page grid we are in.
+%
+% Note: Access \cmd\thepagegrid\ only within the safety of the output routine.
+%
+% Warning: The page grid should be changed only within the safety of the output routine.
+% People who write multicol page grid mechanisms appear not to understand the matter, so they
+% should particularly heed this warning. Think about it: obviously Lamport did so, which is why
+% his \cmd\twocolumn\ command forced a pagebreak, which is limiting, but safe.
+% \begin{macrocode}
+\def\thepagegrid{one}%
+% \end{macrocode}
+% \end{macro}
+%
+%
+% \subsection{Application Note: implementing a page grid}
+%
+% If you want to create a new page grid for \LaTeX,
+% you must define five procedures with specific names:
+% \cmd\open@column@ name, \cmd\shut@column@ name, \cmd\end@column@ name,
+% \cmd\output@column@ name, and \cmd\@addmarginpar@ name, where ``name'' is the
+% name of your page grid.
+%
+% The procedure \cmd\open@column@ name starts the new page grid. It should define \cmd\thepagegrid,
+% deal with \cmd\box\cmd\pagesofar\ and \cmd\box\cmd\footbox\ (perhaps by leaving them alone),
+% and it should set the values of \LaTeX's page layout parameters for the column size and height.
+%
+% The procedure \cmd\shut@column@ name should expect to be called with \cmd\holdinginserts\
+% cleared (it can assume that \cmd\holdinginserts\ will automatically be restored).
+% It should properly deal with insertions (like footnotes); calling \cmd\@makecol\
+% is a good way to do this. It should know that the page grid is being terminated
+% in the middle of a page, so it should make arrangements to carry the footnotes down to the
+% bottom of the column or page, and it should possibly salt away the material for later
+% incorporation into the page. The box registers \cmd\footbox\ and \cmd\pagesofar\ are customarily used
+% for this purpose.
+%
+% The procedure \cmd\end@column@ name should kick out a possibly short page containing all the
+% floats committed to the page. It will be invoked during \cmd\clearpage\ processing.
+% After that, it should \cmd\unvbox\cmd\@cclv.
+%
+% The procedure \cmd\output@column@ name should ship out or commit the current \cmd\@outputbox.
+% In a one-column layout, you ship out; in a multicolumn layout, you commit the box as the contents
+% of a particular column, and if that column is the last, you ship out.
+%
+% The procedure \cmd\@addmarginpar@ name should return a boolean (either \cmd\true@sw\ or \cmd\false@sw\ or an equivalent)
+% to tell the marginpar mechanism to place the marginal material to the right or left, respectively.
+%
+% You can use the existing page grids ``one'' and ``mlt'' as a point of departure for creating others.
+% The former can be the basis for, say, a single-column page grid with a side column.
+%
+% \begin{macro}{\pagesofar}
+% \begin{macro}{\footbox}
+% \begin{macrocode}
+\newbox\pagesofar
+\newbox\footbox
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+%
+% \subsubsection{One-column page grid}
+%
+% \begin{macro}{\onecolumngrid}
+% \begin{macro}{\open@column@one}
+% \begin{macro}{\shut@column@one}
+% \begin{macro}{\float@column@one}
+% \begin{macro}{\end@column@one}
+% \begin{macro}{\output@column@one}
+% \begin{macro}{\@addmarginpar@one}
+% Here are all the procedures necessary for the standard page grid named ``one'':
+% a single column layout. It is, of course, \LaTeX's familiar \cmd\onecolumn\ layout.
+% We begin with the procedure exposed to the style writer.
+% This is, however, not a \LaTeX\ command; users should not change the page grid.
+% \begin{macrocode}
+\newcommand\onecolumngrid{\do@columngrid{one}{\@ne}}%
+% \end{macrocode}
+%
+% Note that a document class that issues the command \cmd\onecolumn\ will break. This includes
+% \LaTeX's standard classes.dtx-based classes: if your class descends from one of these, you
+% must expunge it of all such commands.
+% \begin{macrocode}
+\let\onecolumn\@undefined
+% \end{macrocode}
+%
+% The procedure \cmd\open@column@one\ takes advantage of the special nature of the one-column
+% page grid to deal with \cmd\box\cmd\pagesofar, therefore it must also reset \cmd\@colroom.
+% \begin{macrocode}
+\def\open@column@one#1{%
+ \unvbox\pagesofar
+ \gdef\thepagegrid{one}%
+ \global\pagegrid@col#1%
+ \global\pagegrid@cur\@ne
+ \set@colht
+%\set@colroom
+ \set@column@hsize\pagegrid@col
+}%
+% \end{macrocode}
+%
+% The procedure \cmd\shut@column@one\ saves away the one-column material into the box register \cmd\pagesofar.
+% Because it is called from a message handler, we are assured that marks are properly taken care of.
+% \begin{macrocode}
+\def\shut@column@one{%
+ \@makecol
+ \global\setbox\pagesofar\vbox{\unvbox\@outputbox\recover@footins}%
+ \set@colht
+%\set@colroom
+}%
+% \end{macrocode}
+%
+% The procedure \cmd\float@column@one\ takes care of a float column that has been built by \cmd\@tryfcolumn,
+% in the single-column page grid.
+% \begin{macrocode}
+\def\float@column@one{%
+ \@makecol
+ \@outputpage
+}%
+% \end{macrocode}
+%
+% The procedure \cmd\end@column@one\ is executed at the end of \cmd\clearpage\ processing,
+% if we were in a one-column page grid, once all permissive float pages have been shipped out.
+% At this point, one could perhaps
+% assume that nothing more need be done, but let us anyway test for committed floats and force a shipout.
+%
+% FIXME: this procedure does the same as \cmd\end@column@mlt
+% (except for the test of \cmd\@ifx@empty\cmd\@dbltoplist):
+% the two could almost be the same procedure.
+%
+% I have changed this procedure to avoid the testing it once did: it simply puts down interrupts,
+% upon which it relies to correctly do what \cmd\clearpage\ requires.
+% \begin{macrocode}
+\def\end@column@one{%
+ \unvbox\@cclv\setbox\z@\lastbox
+ \protect@penalty\do@newpage@pen
+}%
+% \end{macrocode}
+%
+% The procedure \cmd\output@column@one\ is dispatched from the output routine when
+% we have completed a page (that is, a column in a one-column page grid).
+% It ships out the page using the \cmd\@outputpage\ of standard \LaTeX,
+% which has been retained (it is needed also in \cmd\output@column@mlt,
+% and in any case should remain as the sole procedure in \LaTeX\ where \cmd\shipout\ is performed).
+% It will be followed up with an output routine message to prepare a new column.
+% \begin{macrocode}
+\def\output@column@one{%
+ \@outputpage
+}%
+% \end{macrocode}
+%
+% The following procedure determines which side of the page a marginpar will appear.
+% It reproduces the behavior of standard \LaTeX.
+% \begin{macrocode}
+\def\@addmarginpar@one{%
+ \@if@sw\if@mparswitch\fi{%
+ \@ifodd\c@page{\false@sw}{\true@sw}%
+ }{\false@sw}{%
+ \@if@sw\if@reversemargin\fi{\false@sw}{\true@sw}%
+ }{%
+ \@if@sw\if@reversemargin\fi{\true@sw}{\false@sw}%
+ }%
+}%
+% \end{macrocode}
+%
+% The following procedure yields a Boolean value; it determines whether a float in the deferred queue
+% is appropriate for placing. In the one-column grid, all floats are so.
+% \begin{macrocode}
+\def\@floatselect@sw@one#1{\true@sw}%
+% \end{macrocode}
+%
+% \begin{macrocode}
+\def\onecolumngrid@push{%
+ \do@output@MVL{%
+ \@ifnum{\pagegrid@col=\@ne}{%
+ \global\let\restorecolumngrid\@empty
+ }{%
+ \xdef\restorecolumngrid{%
+ \noexpand\start@column{\thepagegrid}{\the\pagegrid@col}%
+ }%
+ \start@column{one}{\@ne}%
+ }%
+ }%
+}%
+\def\onecolumngrid@pop{%
+ \do@output@MVL{\restorecolumngrid}%
+}%
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+%
+% \subsubsection{Two-column page grid}
+%
+% \begin{macro}{\twocolumngrid}
+% \begin{macro}{\open@column@mlt}
+% \begin{macro}{\shut@column@mlt}
+% \begin{macro}{\end@column@mlt}
+% \begin{macro}{\output@column@mlt}
+% \begin{macro}{\@addmarginpar@mlt}
+% Here are all the procedures necessary for the standard page grid named ``mlt'':
+% the multi-column page grid. With an argument of "2", it is,
+% of course, \LaTeX's familiar \cmd\twocolumn\ layout.
+%
+% We start with the procedure to switch to the two-column page grid.
+% \begin{macrocode}
+\newcommand\twocolumngrid{\do@columngrid{mlt}{\tw@}}%
+% \end{macrocode}
+%
+% The corresponding command of \LaTeX\ is obsolete.
+% \begin{macrocode}
+\let\twocolumn\@undefined
+% \end{macrocode}
+%
+% Of course, \cmd\@topnewpage\ is also obsolete. Just do
+%\begin{quote}
+% \cmd\clearpage\cmd\onecolumngrid<vertical mode material>\cmd\twocolumngrid.
+%\end{quote}
+% \begin{macrocode}
+\let\@topnewpage\@undefined
+% \end{macrocode}
+%
+% If your document class descends from one of \LaTeX's standard classes.dtx-derived
+% classes, it will break. You must expunge from it all such commands.
+%
+% \begin{macrocode}
+\def\open@column@mlt#1{%
+ \gdef\thepagegrid{mlt}%
+ \global\pagegrid@col#1%
+ \global\pagegrid@cur\@ne
+ \set@column@hsize\pagegrid@col
+ \set@colht
+%\set@colroom
+}%
+% \end{macrocode}
+%
+% The procedure \cmd\shut@column@mlt\ ends the current column, balances the columns, and
+% salts away all in \cmd\pagesofar. Because it is called in a message handler,
+% we are assured that marks are handled properly.
+% Attention: because this procedure balances columns, all footnotes are
+% held aside in \cmd\footbox\ for placement at the bottom of the page.
+%
+% Bug note: the last macro executed by this procedure is \cmd\set@colht, but had been erroneously \cmd\set@colroom.
+% I now believe that the latter should be changed pretty much everywhere to the former.
+% \begin{macrocode}
+\def\shut@column@mlt{%
+ \@cclv@nontrivial@sw{%
+ \@makecol
+ \@ifnum{\pagegrid@cur<\pagegrid@col}{%
+ \expandafter\global\expandafter\setbox\csname col@\the\pagegrid@cur\endcsname\box\@outputbox
+ \global\advance\pagegrid@cur\@ne
+ }{}%
+ }{%
+ {\setbox\z@\box\@cclv}%
+ }%
+ \@ifnum{\pagegrid@cur>\@ne}{%
+ \csname balance@\the\pagegrid@col\endcsname
+ \grid@column{}%
+ \@combinepage
+ \@combinedblfloats
+ \global\setbox\pagesofar\box\@outputbox
+ }{}%
+ \set@colht
+}%
+% \end{macrocode}
+%
+% The procedure \cmd\float@column@mlt\ takes care of a float page that has been built by \cmd\@tryfcolumn,
+% in the multi-column page grid.
+% \begin{macrocode}
+\def\float@column@mlt{%
+ \@combinepage
+ \@combinedblfloats
+ \@outputpage
+ \global\pagegrid@cur\@ne
+ \protect@penalty\do@startpage@pen
+}%
+% \end{macrocode}
+%
+% The procedure \cmd\end@column@mlt\ is executed at the end of \cmd\clearpage\ processing,
+% if we were in a multi-column page grid, once all permissive float pages have been shipped out.
+% If no floats are committed and if no columns are yet filled, we have nothing to do.
+% Otherwise, we kick out a column and try again.
+%
+% Note that in our code to kick out a column, we must deal properly with the case where the column
+% is trivial: it will have nothing but \cmd\topskip\ glue plus a protection box. We substitute an ordinary
+% \cmd\null\ for the protection box.
+%
+% \begin{macrocode}
+\def\end@column@mlt{%
+ \@ifx@empty\@toplist{%
+ \@ifx@empty\@botlist{%
+ \@ifx@empty\@dbltoplist{%
+ \@ifx@empty\@deferlist{%
+ \@ifnum{\pagegrid@cur=\@ne}{%
+ \false@sw
+ }{%
+ \true@sw
+ }%
+ }{%
+ \true@sw
+ }%
+ }{%
+ \true@sw
+ }%
+ }{%
+ \true@sw
+ }%
+ }{%
+ \true@sw
+ }%
+ % true = kick out a column and try again
+ {%
+ \@cclv@nontrivial@sw{%
+ \unvbox\@cclv\setbox\z@\lastbox
+ }{%
+ \unvbox\@cclv\setbox\z@\lastbox\unskip\null
+ }%
+ \protect@penalty\do@newpage@pen
+ \protect@penalty\do@endpage@pen
+ }{%
+ \unvbox\@cclv\setbox\z@\lastbox
+ }%
+}%
+% \end{macrocode}
+% The procedure \cmd\output@column@mlt (cf. \cmd\output@column@one)
+% is dispatched from the output routine when
+% we have completed a column in a multi-column page grid).
+% (It replaces the \cmd\@outputdblcol\ of standard \LaTeX.)
+% If a complete set of columns is at hand, it ships out the page and
+% lays down an interrupt for \cmd\do@startpage@pen, which will commit the
+% full-page-width floats to the next page.
+% Like \cmd\output@column@mlt, this is followed by
+% an output routine message to prepare a new column.
+% \begin{macrocode}
+\def\output@column@mlt{%
+ \@ifnum{\pagegrid@cur<\pagegrid@col}{%
+ \expandafter\global\expandafter\setbox\csname col@\the\pagegrid@cur\endcsname\box\@outputbox
+ \global\advance\pagegrid@cur\@ne
+ }{%
+ \set@adj@colht\dimen@
+% \advance\dimen@-\topskip
+ \grid@column{}%{\dimen@}%
+ \@combinepage
+ \@combinedblfloats
+ \@outputpage
+ \global\pagegrid@cur\@ne
+ \protect@penalty\do@startpage@pen
+ }%
+}%
+% \end{macrocode}
+% The procedure \cmd\output@column@mlt\ obsoletes \LaTeX's \cmd\@outputdblcol
+% \begin{macrocode}
+\let\@outputdblcol\@undefined
+% \end{macrocode}
+%
+% The following procedure yields a Boolean value; it determines whether a float in the deferred queue
+% is appropriate for placement in the column. In the multi-column grid, only those narrower than \cmd\textwidth\ are so.
+% \begin{macrocode}
+\def\@floatselect@sw@mlt#1{\@if@notdblfloat{#1}}%
+% \end{macrocode}
+%
+% The following procedure determines which side of the page a marginpar will appear.
+% It reproduces the behavior of standard \LaTeX.
+% \begin{macrocode}
+\def\@addmarginpar@mlt{% emits a boolean
+ \@ifnum{\pagegrid@cur=\@ne}%
+}%
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+%
+% \subsubsection{Page grid utility procedures}
+%
+% \begin{macro}{\pagegrid@cur}
+% \begin{macro}{\pagegrid@col}
+% \begin{macro}{\col@}
+% \begin{macro}{\pagegrid@init}
+% We take over \LaTeX's \cmd\col@number\ and \cmd\@leftcolumn, which are obsolete.
+% We create two counters to hold the columns in the page grid and the current column within.
+% We also create the first of a set of box registers to hold the committted columns.
+% \begin{macrocode}
+\let\pagegrid@cur\col@number
+\let\col@number\@undefined
+\newcount\pagegrid@col
+\pagegrid@cur\@ne
+\expandafter\let\csname col@\the\pagegrid@cur\endcsname\@leftcolumn
+\let\@leftcolumn\@undefined
+% \end{macrocode}
+%
+% The default is for maximum two columns.
+% If your class will require more columns, assign that number to \cmd\pagegrid@col\
+% before \envb{document} time.
+% \begin{macrocode}
+\pagegrid@col\tw@
+% \end{macrocode}
+%
+% The procedure \cmd\pagegrid@init\ exercises \cmd\newbox\ sufficiently to create the
+% boxes for holding the columns in the page grid.
+% \begin{macrocode}
+\def\pagegrid@init{%
+ \advance\pagegrid@cur\@ne
+ \@ifnum{\pagegrid@cur<\pagegrid@col}{%
+ \csname newbox\expandafter\endcsname\csname col@\the\pagegrid@cur\endcsname
+ \pagegrid@init
+ }{%
+ }%
+}%
+\appdef\class@documenthook{%
+ \pagegrid@init
+}%
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\grid@column}
+% The procedure \cmd\grid@column\ knows how to lay up the columns in a multi-column page grid.
+% It uses utility procedures \cmd\append@column\ and \cmd\box@column.
+%
+% \begin{macrocode}
+\def\grid@column#1{%
+ \global\setbox\@outputbox\vbox{%
+ \hb@xt@\textwidth{%
+ \vrule\@height\z@\@width\z@\@if@empty{#1}{}{\@depth#1}%
+ \pagegrid@cur\@ne
+ \append@column
+ \box@column\@outputbox
+ }%
+ \vskip\z@skip % FIXME: page depth!
+ }%
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\append@column}
+% \begin{macro}{\box@column}
+% \begin{macro}{\marry@baselines}
+% The procedure \cmd\append@column\ appends columns for \cmd\grid@column,
+% \cmd\box@column\ builds the columns for \cmd\append@column,
+% and \cmd\marry@baselines\ pastes vertical things back together.
+% \changes{v4.0rc1}{2001/06/18}
+% {Introoduce \cs{marry@height} }
+%
+% Note that \cmd\box@column\ makes an attempt to prevent excessive \cmd\topskip\
+% or \cmd\baselineskip\ glue
+% from being applied by \TeX\ when \cmd\@outputbox\ is contributed to the MVL.
+% If this is not done, it is possible to get into an infinite loop in the corner case,
+% wherein the page grid is changed to one column and the balanced-up columns are
+% already sufficient to fill the page.
+%
+% Note (AO 0920): I have changed the dimension involved with \cmd\box@column\ from
+% \cmd\vsize\ to \cmd\textheight, because the former is certainly not the correct value
+% to use: it will change if floats have been placed in the last column of the page.
+% I believe \cmd\textheight\ is the correct parameter to use here.
+%
+% A REVTeX4 beta user, Sergey Strelkov (strelkov@maik.rssi.ru), wants the option
+% of ragged-bottom columns. Implementing this feature properly means reboxing the
+% columns to their natural height only if \cmd\raggedcolumn@sw\ is true.
+% Otherwise, they get reboxed to their common height (\cmd\@colht?).
+%
+% Note that the default has hereby changed from ragged to flush.
+% It's not clear that anyone but Sergey will notice.
+%
+% The macro \cmd\marry@skip\ addresses (in a limited way)
+% the fact that neither the value of \cmd\baselineskip\ nor that of \cmd\topskip\
+% can be relied upon for the purpose of marrying the baselines of two split columns.
+% (Because there might have been a local change to their values at the point where
+% the output routine got triggered.)
+%
+% For best results, your document class should call for grid changes only when in basal text
+% settings. The \cmd\marry@baselines\ procedure will use the values appropriate to that point
+% when attempting to put the columns back together.
+%
+% In any case, we are not attempting to solve the more general problem of how to marry baselines
+% where the leading can change arbitrarily within the galley or where glue could have been trimmed
+% at a page top.
+% \begin{macrocode}
+\def\append@column{%
+ \@ifnum{\pagegrid@cur<\pagegrid@col}{%
+ \expandafter\box@column\csname col@\the\pagegrid@cur\endcsname
+ \hfil
+ \vrule \@width\columnseprule
+ \hfil
+ \advance\pagegrid@cur\@ne
+ \append@column
+ }{%
+ }%
+}%
+\def\box@column#1{%
+ \raise\topskip
+ \hb@xt@\columnwidth{%
+ \dimen@\ht#1\@ifdim{\dimen@>\@colht}{\dimen@\@colht}{}%
+% \advance\dimen@-\topskip
+ \count@\vbadness\vbadness\@M
+ \dimen@ii\vfuzz\vfuzz\maxdimen
+ \outputdebug@sw{\saythe\@colht\saythe\dimen@}{}%
+ \vtop to\dimen@
+% \@ifdim{\ht#1>\textheight}{to\textheight}{}%
+ {\hrule\@height\z@
+ \unvbox#1%
+ \raggedcolumn@skip
+ }%
+ \vfuzz\dimen@ii
+ \vbadness\count@
+ \hss
+ }%
+}%
+\def\marry@baselines{%
+%{\tracingall\scrollmode\showlists}%
+%\skip@\baselineskip\advance\skip@-\topskip %FIXME: cannot assume \baselineskip nor \topskip
+ \vskip\marry@skip\relax
+}%
+\gdef\marry@skip{\z@skip}%
+\def\set@marry@skip{%
+ \begingroup
+ \skip@\baselineskip\advance\skip@-\topskip
+ \@ifdim{\skip@>\z@}{%
+ \xdef\marry@skip{\the\skip@}%
+ }{}%
+ \endgroup
+}%
+\AtBeginDocument{%
+ \@ifxundefined\raggedcolumn@sw{\@booleanfalse\raggedcolumn@sw}{}%
+}%
+\def\raggedcolumn@skip{%
+ \vskip\z@\raggedcolumn@sw{\@plus.0001fil\@minus.0001fil}{}\relax
+}%
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\@combinepage}
+% The procedure \cmd\@combinepage\ prepends the stored page to \cmd\@outputbox.
+% \begin{macrocode}
+\def\@combinepage{%
+ \@ifvoid\pagesofar{}{%
+ \setbox\@outputbox\vbox{%
+ \unvbox\pagesofar
+ \marry@baselines
+ \unvbox\@outputbox
+ }%
+ }%
+ \@ifvoid\footbox{}{%
+ \setbox\@outputbox\vbox{%
+ \unvbox\@outputbox
+ \marry@baselines
+ \unvbox\footbox
+ }%
+ }%
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\@combinedblfloats}
+% We modify \LaTeX's \cmd\@combinedblfloats\ to be more appropriate for incremental page building:
+% we \cmd\unvbox\ the \cmd\@outputbox.
+% \begin{macrocode}
+\def\@combinedblfloats{%
+ \@ifx@empty\@dbltoplist{}{%
+ \setbox\@tempboxa\vbox{}%
+ \let\@elt\@comdblflelt\@dbltoplist
+ \let\@elt\relax\xdef\@freelist{\@freelist\@dbltoplist}%
+ \global\let\@dbltoplist\@empty
+ \setbox\@outputbox\vbox{%
+ %\boxmaxdepth\maxdepth %% probably not needed, CAR
+ \unvbox\@tempboxa\unskip
+ \@ifnum{\@dbltopnum>\m@ne}{\dblfigrule}{}%FIXME: how is \@dbltopnum maintained?
+ \vskip\dbltextfloatsep
+ \unvbox\@outputbox
+ }%
+ }%
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\set@column@hsize}%
+% The procedure \cmd\set@column@hsize\ takes care of setting up the horizontal dimensions
+% for the current page grid. The present routine will certainly not be adequate for more
+% complex page layouts (e.g., with a side column), but works for the common ones.
+% \changes{v4.0rc1}{2001/06/18}
+% {Introoduce \cs{set@marry@height} }
+% \begin{macrocode}
+\def\set@column@hsize#1{%
+ \pagegrid@col#1%
+ \global\columnwidth\textwidth
+ \global\advance\columnwidth\columnsep
+ \global\divide\columnwidth\pagegrid@col
+ \global\advance\columnwidth-\columnsep
+ \global\hsize\columnwidth
+ \global\linewidth\columnwidth
+ \skip@\baselineskip\advance\skip@-\topskip
+ \@ifnum{\pagegrid@col>\@ne}{\set@marry@skip}{}%
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\set@colht}%
+% \begin{macro}{\set@colroom}%
+% \begin{macro}{\set@vsize}%
+% \begin{macro}{\set@adj@colht}%
+% The story of \cmd\textheight, \cmd\@colht, \cmd\@colroom, and \cmd\vsize.
+%
+% \cmd\textheight---height of the text column. Not a running parameter, however, each time a page is
+% shipped out, the \cmd\textheight\ could in principle be altered. This must be done before
+%
+% \cmd\@colht---\cmd\textheight\ minus the height of any full-page-width floats. The latter are committed
+% only just after shipping out, and only if we are in a multicolumn page grid.
+% Therefore, \cmd\@colht\ should be set after a \cmd\shipout\ (by \cmd\@outputpage) and
+% will be adjusted when full-page-width floats are committed to the fresh page by \cmd\do@startpage.
+%
+% \cmd\@colroom---\cmd\@colht\ (adjusted by \cmd\pagesofar) minus the height of any column-width floats.
+% The latter are committed anywhere on the page, at which point \cmd\@colroom\ must be adjusted.
+% Therefore, \cmd\@colroom\ should be set (by \cmd\set@colroom) whenever a column is prepared (by ).
+%FIXME: committed (by \cmd\output@column@) and
+% will be adjusted (by \cmd\@add@float\ or \cmd\do@startcolumn) whenever a float is committted to the column.
+%
+% \cmd\vsize---\cmd\@colroom.
+% Therefore, \cmd\vsize\ should be set (by \cmd\set@vsize) whenever
+% the \cmd\@colroom\ is set (by \cmd\set@colroom) or adjusted (by \cmd\@add@float\ or \cmd\do@startcolumn)
+%FIXME: or when the \cmd\pagesofar\ box is changed (after invoking \cmd\open@column@).
+%
+% Question: what if there are committed floats? Footnotes?
+% Answer: full-page-width floats are only committed at top, and they are already reckoned with in \cmd\@colht.
+% Column-width committed floats are incorporated by \cmd\@makecol; footnotes need help.
+%
+% Note: FIXME: adjusting for \cmd\pagesofar\ is done at not quite the right time. I need to reexamine \cmd\set@colht,
+% because \cmd\@dbltoplist\ and \cmd\pagesofar\ really should be on the same footing.
+% Perhaps \cmd\@colht\ and \cmd\@colroom\ should both deal with their respective ``lists'' in the same way?
+%
+% These concerns will be particularly germane if we ever extend this package to deal with full-page-width floats
+% placed at the bottom of the page, or committed on the same page as called out.
+%
+% It occurs to me that we should ditch \cmd\set@colroom\ and only ever execute \cmd\set@colht, which sets \cmd\@colroom\ as a side effect.
+% If so, we can make \cmd\@colht\ take \cmd\pagesofar\ into account, as it should. Then \cmd\@colht\ will return to its
+% original significance as the value that \cmd\@colroom\ is set to after a column is committed.
+%
+% On the other hand, why not simply forget all this caching and (re-)calculate \cmd\vsize\ as late as possible?
+% Paticularly, \cmd\@colht\ is an artifact of the old way of doing things, where once it was set, it would never change.
+%
+% \begin{macrocode}
+\def\set@colht{%
+ \set@adj@textheight\@colht
+ \global\let\enlarge@colroom\@empty
+ \set@colroom
+}%
+\def\set@adj@textheight#1{%
+ #1\textheight
+ \def\@elt{\adj@page#1}%
+ \@booleantrue\firsttime@sw\@dbltoplist
+ \let\@elt\relax
+%\@ifvoid\pagesofar{}{%
+% \advance#1-\ht\pagesofar\advance#1-\dp\pagesofar
+%}%
+ \global#1#1\relax
+ \outputdebug@sw{\saythe#1}{}%
+}%
+\def\set@colroom{%
+ \set@adj@colht\@colroom
+ \@if@empty\enlarge@colroom{}{%
+ \global\advance\@colroom\enlarge@colroom\relax
+ }%
+ \outputdebug@sw{\saythe\@colroom}{}%
+ \@ifdim{\@colroom>\topskip}{}{%
+ \ltxgrid@info{Not enough room: \string\@colroom=\the\@colroom; increasing to \the\topskip}%
+ \@colroom\topskip
+ }%
+ \global\@colroom\@colroom
+ \set@vsize
+}%
+%
+\def\set@vsize{%
+ \global\vsize\@colroom
+ \outputdebug@sw{\saythe\vsize}{}%
+}%
+%
+\def\set@adj@colht#1{%
+ #1\@colht
+ \@ifvoid\pagesofar{}{%
+ \advance#1-\ht\pagesofar\advance#1-\dp\pagesofar
+ }%
+ \@ifvoid\footbox{}{%
+ \advance#1-\ht\footbox\advance#1-\dp\footbox
+ }%
+ \def\@elt{\adj@column#1}%
+ \@booleantrue\firsttime@sw\@toplist
+ \@booleantrue\firsttime@sw\@botlist
+ \let\@elt\relax
+ \outputdebug@sw{\saythe#1}{}%
+}%
+\def\adj@column#1#2{%
+ \advance#1-\ht#2%
+ \advance#1-\firsttime@sw{\textfloatsep\@booleanfalse\firsttime@sw}{\floatsep}%
+}%
+\def\adj@page#1#2{%
+ \advance#1-\ht#2%
+ \advance#1-\firsttime@sw{\dbltextfloatsep\@booleanfalse\firsttime@sw}{\dblfloatsep}%
+}%
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\@outputpage}%
+% At the tail of \cmd\@outputpage, we set \cmd\@colht\ and the float placement parameters
+% (this is the one point where it is appropriate to set \cmd\@colht).
+% At \cmd\do@startpage\ time, we adjust \cmd\@colht's value to reflect committed
+% full-page-width floats.
+%
+% Note: with a correctly written output routine, a call to \cmd\@outputpage\ will inevitably be
+% followed by a call to \cmd\do@startpage, so these procedure calls would be unneeded.
+% \begin{macrocode}
+\appdef\@outputpage{%
+ \set@colht % FIXME: needed?
+ \@floatplacement % FIXME: needed?
+ \@dblfloatplacement % FIXME: needed?
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{balance@2}
+% We define procedures for balancing columns in a multicolumn layout.
+% For now, we define only one: a procedure for the two-column grid.
+% All others will simply \cmd\relax\ out.
+% \begin{macrocode}
+\@namedef{balance@2}{%
+ \expandafter\balance@two\csname col@1\endcsname\@outputbox
+ % Avoid a bug by preventing a restore when leaving this group
+ \global\setbox\csname col@1\endcsname\box\csname col@1\endcsname
+ \@ifvoid\footbox{}{%
+ \global\setbox\footbox\vbox{%
+ \setbox\z@\box\@tempboxa
+ \let\recover@footins\relax
+ \balance@two\footbox\@tempboxa
+ \hb@xt@\textwidth{\box\footbox\hfil\box\@tempboxa}%
+ }%
+ }%
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\balance@two}
+% The procedure \cmd\balance@two\ takes two columns and balances them; in the process it removes
+% any footnotes that may be present to a place of safety, for later placement at the foot of the
+% shipped-out page.
+% The box register \cmd\box\cmd\@ne\ is the aggregate of all columns.
+% The box register \cmd\box\cmd \z@\ is the last column.
+% The box register \cmd\box\cmd\tw@\ is the first column.
+% The \cmd\dimen\ register \cmd\dimen@\ is the trial value to balance to,
+% initially half the height of \cmd\box\cmd\@ne.
+% The \cmd\dimen\ register \cmd\dimen@i\ is the increment for the next trial;
+% its initial value is equal to the initial value of \cmd\dimen@.
+% The \cmd\dimen\ register \cmd\dimen@ii\ is the difference of the heights of the two columns.
+%
+% The procedure uses a binary search for that value of \cmd\dimen@\ which is stable to within .5\cmd\p@\ and which
+% makes the last column be shorter than the others.
+%
+% This procedure can be extended to multiple columns simply by changing it to execute \cmd\vsplit\ multiple
+% times (one less than the total number of columns in the page layout) and to calculating \cmd\dimen@ii\
+% to be the difference of the heights of last column and the \cmd\dimen@.
+% Upon termination of the search, one would excute the \cmd\vsplit s once again, this time
+% using the actual \cmd\col@\ box registers to store the
+% balanced columns, thereby clobbering their former contents.
+%
+% Bug Note:
+% as originally written, this macro had a bug, which is well worth avoiding under similar circumstances anywhere.
+% So, learn from the mistakes of others, as they say.
+% In trying to remove the depth of the boxes created via \cmd\vsplit\ within the \cmd\loopwhile\ control,
+% I originally coded
+% \cmd\unvbox
+% \cmd\z@\
+% \cmd\setbox
+% \cmd\z@
+% \cmd\lastbox\
+% \cmd\dimen@
+% \cmd\dp
+% \cmd\z@\
+% \cmd\box
+% \cmd\z@\
+% \cmd\vskip-%
+% \cmd\dimen@.
+% The error here is that the shift of the last box in the vertical list will be lost in the process.
+% Simply put, \cmd\setbox\cmd\z@\cmd\lastbox\ fails to retain the shift of the box node in the vertical list,
+% and when it is put down again via \cmd\box\cmd\z@, it will no longer have the correct shift.
+%
+% This bug affected things placed in the MVL with \cmd\moveleft, \cmd\moveright, \cmd\parshape, and
+% \cmd\hangindent, as well as things shifted by \TeX's primitive mechanisms.
+%
+% A superior strategy for removing the depth of the last line of the list is more expensive, but safer:
+% make a separate copy of the list, measure the depth of the last box as above, but then discard
+% the list, retaining only the value of the dimension.
+%
+% Note that this procedure will not work if the material within is excessively chunky.
+% A particular failure mode exists where none of the material is allocated to the last (right) column.
+% We detect this case and revert to unbalanced columns.
+%
+% Another failure mode is where a large chunk occurs at the beginning of the composite box.
+% In this case, the left column may fill up even when \cmd\dimen@\ is very small.
+% If this configuration leaves the left column longer than the right, then we are done,
+% but \cmd\dimen@\ by no means represents the height of either finished box.
+%
+% Therefore the last step in the process is to rebox the two columns to a common height determined
+% independently of the balancing process.
+%
+% The dimension involved is checked against the current \cmd\@colroom\ to guard against the case where
+% excessive material happens to fall in either column.
+% \begin{macrocode}
+\def\balance@two#1#2{%
+ \outputdebug@sw{{\tracingall\scrollmode\showbox#1\showbox#2}}{}%
+ \setbox\@ne\vbox{%
+ \@ifvoid#1{}{%
+ \unvcopy#1\recover@footins
+ \@ifvoid#2{}{\marry@baselines}%
+ }%
+ \@ifvoid#2{}{%
+ \unvcopy#2\recover@footins
+ }%
+ }%
+ \dimen@\ht\@ne\divide\dimen@\tw@
+ \dimen@i\dimen@
+ \vbadness\@M
+ \vfuzz\maxdimen
+ \loopwhile{%
+ \dimen@i=.5\dimen@i
+ \outputdebug@sw{\saythe\dimen@\saythe\dimen@i\saythe\dimen@ii}{}%
+ \setbox\z@\copy\@ne\setbox\tw@\vsplit\z@ to\dimen@
+ \setbox\z@ \vbox{%
+ \unvcopy\z@
+ \setbox\z@\vbox{\unvbox\z@ \setbox\z@\lastbox\aftergroup\vskip\aftergroup-\expandafter}\the\dp\z@\relax
+ }%
+ \setbox\tw@\vbox{%
+ \unvcopy\tw@
+ \setbox\z@\vbox{\unvbox\tw@\setbox\z@\lastbox\aftergroup\vskip\aftergroup-\expandafter}\the\dp\z@\relax
+ }%
+ \dimen@ii\ht\tw@\advance\dimen@ii-\ht\z@
+ \@ifdim{\dimen@i>.5\p@}{%
+ \advance\dimen@\@ifdim{\dimen@ii<\z@}{}{-}\dimen@i
+ \true@sw
+ }{%
+ \@ifdim{\dimen@ii<\z@}{%
+ \advance\dimen@\tw@\dimen@i
+ \true@sw
+ }{%
+ \false@sw
+ }%
+ }%
+ }%
+ \outputdebug@sw{\saythe\dimen@\saythe\dimen@i\saythe\dimen@ii}{}%
+ \@ifdim{\ht\z@=\z@}{%
+ \@ifdim{\ht\tw@=\z@}{%
+ \true@sw
+ }{%
+ \false@sw
+ }%
+ }{%
+ \true@sw
+ }%
+ {%
+ }{%
+ \ltxgrid@info{Unsatifactorily balanced columns: giving up}%
+ \setbox\tw@\box#1%
+ \setbox\z@ \box#2%
+ }%
+ \setbox\tw@\vbox{\unvbox\tw@\vskip\z@skip}%
+ \setbox\z@ \vbox{\unvbox\z@ \vskip\z@skip}%
+ \set@colroom
+ \dimen@\ht\z@\@ifdim{\dimen@<\ht\tw@}{\dimen@\ht\tw@}{}%
+ \@ifdim{\dimen@>\@colroom}{\dimen@\@colroom}{}%
+ \outputdebug@sw{\saythe{\ht\z@}\saythe{\ht\tw@}\saythe\@colroom\saythe\dimen@}{}%
+ \setbox#1\vbox to\dimen@{\unvbox\tw@\unskip\raggedcolumn@skip}%
+ \setbox#2\vbox to\dimen@{\unvbox\z@ \unskip\raggedcolumn@skip}%
+ \outputdebug@sw{{\tracingall\scrollmode\showbox#1\showbox#2}}{}%
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\recover@footins}
+% The procedure \cmd\recover@footins\ is the utility procedure for recovering the footnotes
+% from the bottom of a column. It is used when the page grid is changed, so that footnotes can be
+% set at the bottom of the shipped out page.
+% \begin{macrocode}
+\def\recover@footins{%
+ \skip\z@ \lastskip\unskip
+ \skip\@ne\lastskip\unskip
+ \setbox\z@\lastbox
+ \@ifvbox\z@{%
+ \setbox\z@\vbox{%
+ \unvbox\z@
+ \setbox\z@\lastbox
+% \outputdebug@sw{{\tracingall\showbox\lastbox}}{}%
+ \@ifvoid\z@{}{%
+ \global\setbox\footbox\vbox{%
+ \unvbox\footbox
+ \@ifvbox\z@{%
+ \unvbox\z@
+ }{%
+ \box\z@
+ }%
+ }%
+ }%
+ }%
+ }{}%
+ \outputdebug@sw{{\tracingall\scrollmode\showbox\footbox}}{}%
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\@begindocumenthook}
+% Initialization:
+% we initialize to the page grid named ``one''.
+% If the class decides to initially set type in a different grid, it
+% should execute these same commands, but changing the first to the appropriate procedure.
+%
+% Note that the point where this sequence is executed would be an excellent place to arrange for
+% floats to be committed to the first page of a document.
+% That is, we execute \cmd\do@startpage, which triggers \cmd\do@startcolumn.
+%
+% FIXME: it should be the job of the page grid to determine the procedure to execute at
+% the start of the job. Make this a hook.
+% \begin{macrocode}
+\prepdef\@begindocumenthook{%
+ \open@column@one\@ne
+ \set@colht
+ \@floatplacement
+ \@dblfloatplacement
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% Comment: our technique of balancing columns is severely limited, because it cannot properly work
+% with \env{longtable}, which places material at the bottom and top of the column break.
+%
+% The proper way to handle a grid change in the middle of the page is to accumulate all the material for
+% an entire article (or chapter) and then assemble finished pages therefrom. This approach is fundamentally
+% superior for complex layouts: it corresponds to real-world workflows.
+% Such a scheme is an excellent subject for another \LaTeX\ package.
+%
+%
+% \subsection{Patches for the longtable package}%
+%
+% \LaTeX's ``required'' package \classname{longtable} (written by David P. Carlilsle),
+% which is part of /latex/required/tools, is incmpatible with both
+% \LaTeX's ``required'' package \classname{multicol} and with
+% \LaTeX's native \cmd\twocolumn\ capability. There is no essential reason
+% for this incompatability, aside from implementation details, and the
+% \classname{ltxgrid} package gives us the ability to lift them.
+%
+% Only four of \classname{longtable}'s procedures require rewriting:
+% \cmd\longtable,
+% \cmd\endlongtable,
+% \cmd\LT@start, and
+% \cmd\LT@end@hd@ft.
+% The procedure \cmd\switch@longtable\ checks against their expected meanings
+% and, if all is as expected, applies the patches.
+% In the process, we simplify things considerably and also make them more
+% secure.
+%
+% Why does \classname{longtable} need to access the output routine, anyway?
+% What it comes down to, is what happens when a pagebreak falls within a
+% long table. If this happens, we would like to append a row at the bottom of
+% the broken table and add a row at the top of the next page.
+%
+% These things can be accomodated easily by the \classname{ltxgrid} output
+% routine hooks.
+%
+%
+% \begin{macro}{\longtable}
+% \begin{macrocode}
+\def\longtable@longtable{%
+ \par
+ \ifx\multicols\@undefined\else\ifnum\col@number>\@ne\@twocolumntrue\fi\fi
+ \if@twocolumn\LT@err{longtable not in 1-column mode}\@ehc\fi
+ \begingroup
+ \@ifnextchar[\LT@array{\LT@array[x]}%
+}%
+\def\longtable@new{%
+ \par
+ \@ifnextchar[\LT@array{\LT@array[x]}%
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\endlongtable}
+% \begin{macrocode}
+\def\endlongtable@longtable{%
+ \crcr
+ \noalign{%
+ \let\LT@entry\LT@entry@chop
+ \xdef\LT@save@row{\LT@save@row}}%
+ \LT@echunk
+ \LT@start
+ \unvbox\z@
+ \LT@get@widths
+ \if@filesw
+ {\let\LT@entry\LT@entry@write\immediate\write\@auxout{%
+ \gdef\expandafter\noexpand
+ \csname LT@\romannumeral\c@LT@tables\endcsname
+ {\LT@save@row}}}%
+ \fi
+ \ifx\LT@save@row\LT@@save@row
+ \else
+ \LT@warn{Column \@width s have changed\MessageBreak
+ in table \thetable}%
+ \LT@final@warn
+ \fi
+ \endgraf\penalty -\LT@end@pen
+ \endgroup
+ \global\@mparbottom\z@
+ \pagegoal\vsize
+ \endgraf\penalty\z@\addvspace\LTpost
+ \ifvoid\footins\else\insert\footins{}\fi
+}%
+\def\endlongtable@new{%
+ \crcr
+ \noalign{%
+ \let\LT@entry\LT@entry@chop
+ \xdef\LT@save@row{\LT@save@row}%
+ }%
+ \LT@echunk
+ \LT@start
+ \unvbox\z@
+ \LT@get@widths
+ \@if@sw\if@filesw\fi{%
+ {%
+ \let\LT@entry\LT@entry@write
+ \immediate\write\@auxout{%
+ \gdef\expandafter\noexpand\csname LT@\romannumeral\c@LT@tables\endcsname
+ {\LT@save@row}%
+ }%
+ }%
+ }{}%
+ \@ifx\LT@save@row\LT@@save@row{}{%
+ \LT@warn{%
+ Column \@width s have changed\MessageBreak in table \thetable
+ }\LT@final@warn
+ }%
+ \endgraf
+ \nobreak
+ \box\@ifvoid\LT@lastfoot{\LT@foot}{\LT@lastfoot}%
+ \global\@mparbottom\z@
+ \endgraf
+ \LT@post
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\LT@start}
+% \begin{macrocode}
+\def\LT@start@longtable{%
+ \let\LT@start\endgraf
+ \endgraf
+ \penalty\z@
+ \vskip\LTpre
+ \dimen@\pagetotal
+ \advance\dimen@ \ht\ifvoid\LT@firsthead\LT@head\else\LT@firsthead\fi
+ \advance\dimen@ \dp\ifvoid\LT@firsthead\LT@head\else\LT@firsthead\fi
+ \advance\dimen@ \ht\LT@foot
+ \dimen@ii\vfuzz\vfuzz\maxdimen
+ \setbox\tw@\copy\z@
+ \setbox\tw@\vsplit\tw@ to \ht\@arstrutbox
+ \setbox\tw@\vbox{\unvbox\tw@}%
+ \vfuzz\dimen@ii
+ \advance\dimen@ \ht
+ \ifdim\ht\@arstrutbox>\ht\tw@\@arstrutbox\else\tw@\fi
+ \advance\dimen@\dp
+ \ifdim\dp\@arstrutbox>\dp\tw@\@arstrutbox\else\tw@\fi
+ \advance\dimen@ -\pagegoal
+ \ifdim \dimen@>\z@\vfil\break\fi
+ \global\@colroom\@colht
+ \ifvoid\LT@foot\else
+ \advance\vsize-\ht\LT@foot
+ \global\advance\@colroom-\ht\LT@foot
+ \dimen@\pagegoal\advance\dimen@-\ht\LT@foot\pagegoal\dimen@
+ \maxdepth\z@
+ \fi
+ \ifvoid\LT@firsthead\copy\LT@head\else\box\LT@firsthead\fi
+ \output{\LT@output}%
+}%
+\def\LT@start@new{%
+ \let\LT@start\endgraf
+ \endgraf
+ \markthr@@{}%
+ \LT@pre
+ \@ifvoid\LT@firsthead{\LT@top}{\box\LT@firsthead\nobreak}%
+ \mark@envir{longtable}%
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\LT@end}
+% \begin{macrocode}
+\def\LT@end@hd@ft@longtable#1{%
+ \LT@echunk
+ \ifx\LT@start\endgraf
+ \LT@err{Longtable head or foot not at start of table}{Increase LTchunksize}%
+ \fi
+ \setbox#1\box\z@
+ \LT@get@widths\LT@bchunk
+}%
+\def\LT@end@hd@ft@new#1{%
+ \LT@echunk
+ \@ifx{\LT@start\endgraf}{%
+ \LT@err{Longtable head or foot not at start of table}{Increase LTchunksize}%
+ }%
+ \global\setbox#1\box\z@
+ \LT@get@widths
+ \LT@bchunk
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\LT@array}
+% \begin{macrocode}
+\def\LT@array@longtable[#1]#2{%
+ \refstepcounter{table}\stepcounter{LT@tables}%
+ \if l#1%
+ \LTleft\z@ \LTright\fill
+ \else\if r#1%
+ \LTleft\fill \LTright\z@
+ \else\if c#1%
+ \LTleft\fill \LTright\fill
+ \fi\fi\fi
+ \let\LT@mcol\multicolumn
+ \let\LT@@tabarray\@tabarray
+ \let\LT@@hl\hline
+ \def\@tabarray{%
+ \let\hline\LT@@hl
+ \LT@@tabarray}%
+ \let\\\LT@tabularcr\let\tabularnewline\\%
+ \def\newpage{\noalign{\break}}%
+ \def\pagebreak{\noalign{\ifnum`}=0\fi\@testopt{\LT@no@pgbk-}4}%
+ \def\nopagebreak{\noalign{\ifnum`}=0\fi\@testopt\LT@no@pgbk4}%
+ \let\hline\LT@hline \let\kill\LT@kill\let\caption\LT@caption
+ \@tempdima\ht\strutbox
+ \let\@endpbox\LT@endpbox
+ \ifx\extrarowheight\@undefined
+ \let\@acol\@tabacol
+ \let\@classz\@tabclassz \let\@classiv\@tabclassiv
+ \def\@startpbox{\vtop\LT@startpbox}%
+ \let\@@startpbox\@startpbox
+ \let\@@endpbox\@endpbox
+ \let\LT@LL@FM@cr\@tabularcr
+ \else
+ \advance\@tempdima\extrarowheight
+ \col@sep\tabcolsep
+ \let\@startpbox\LT@startpbox\let\LT@LL@FM@cr\@arraycr
+ \fi
+ \setbox\@arstrutbox\hbox{\vrule
+ \@height \arraystretch \@tempdima
+ \@depth \arraystretch \dp \strutbox
+ \@width \z@}%
+ \let\@sharp##\let\protect\relax
+ \begingroup
+ \@mkpream{#2}%
+ \xdef\LT@bchunk{%
+ \global\advance\c@LT@chunks\@ne
+ \global\LT@rows\z@\setbox\z@\vbox\bgroup
+ \LT@setprevdepth
+ \tabskip\LTleft\halign to\hsize\bgroup
+ \tabskip\z@ \@arstrut \@preamble \tabskip\LTright \cr}%
+ \endgroup
+ \expandafter\LT@nofcols\LT@bchunk&\LT@nofcols
+ \LT@make@row
+ \m@th\let\par\@empty
+ \everycr{}\lineskip\z@\baselineskip\z@
+ \LT@bchunk
+}%
+\def\LT@LR@l{\LTleft\z@ \LTright\fill}%
+\def\LT@LR@r{\LTleft\fill \LTright\z@ }%
+\def\LT@LR@c{\LTleft\fill \LTright\fill}%
+\def\LT@array@new[#1]#2{%
+ \refstepcounter{table}\stepcounter{LT@tables}%
+ \table@hook
+ \LTleft\fill \LTright\fill
+ \csname LT@LR@#1\endcsname
+ \let\LT@mcol\multicolumn
+ \let\LT@@hl\hline
+ \prepdef\@tabarray{\let\hline\LT@@hl}%
+ \let\\\LT@tabularcr
+ \let\tabularnewline\\%
+ \def\newpage{\noalign{\break}}%
+ \def\pagebreak{\noalign{\ifnum`}=0\fi\@testopt{\LT@no@pgbk-}4}%
+ \def\nopagebreak{\noalign{\ifnum`}=0\fi\@testopt\LT@no@pgbk4}%
+ \let\hline\LT@hline
+ \let\kill\LT@kill
+ \let\caption\LT@caption
+ \@tempdima\ht\strutbox
+ \let\@endpbox\LT@endpbox
+ \@ifxundefined\extrarowheight{%
+ \let\@acol\@tabacol
+ \let\@classz\@tabclassz
+ \let\@classiv\@tabclassiv
+ \def\@startpbox{\vtop\LT@startpbox}%
+ \let\@@startpbox\@startpbox
+ \let\@@endpbox\@endpbox
+ \let\LT@LL@FM@cr\@tabularcr
+ }{%
+ \advance\@tempdima\extrarowheight
+ \col@sep\tabcolsep
+ \let\@startpbox\LT@startpbox
+ \let\LT@LL@FM@cr\@arraycr
+ }%
+%
+ \let\@acoll\@tabacoll
+ \let\@acolr\@tabacolr
+ \let\@acol\@tabacol
+%
+ \setbox\@arstrutbox\hbox{%
+ \vrule
+ \@height \arraystretch \@tempdima
+ \@depth \arraystretch \dp \strutbox
+ \@width \z@
+ }%
+ \let\@sharp##%
+ \let\protect\relax
+ \begingroup
+ \@mkpream{#2}%
+ \@mkpream@relax
+ \edef\@preamble{\@preamble}%
+ \prepdef\@preamble{%
+ \global\advance\c@LT@chunks\@ne
+ \global\LT@rows\z@
+ \setbox\z@\vbox\bgroup
+ \LT@setprevdepth
+ \tabskip\LTleft
+ \halign to\hsize\bgroup
+ \tabskip\z@
+ \@arstrut
+ }%
+ \appdef\@preamble{%
+ \tabskip\LTright
+ \cr
+ }%
+ \global\let\LT@bchunk\@preamble
+ \endgroup
+ \expandafter\LT@nofcols\LT@bchunk&\LT@nofcols
+ \LT@make@row
+ \m@th
+ \let\par\@empty
+ \everycr{}%
+ \lineskip\z@
+ \baselineskip\z@
+ \LT@bchunk
+}%
+\appdef\table@hook{}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\switch@longtable}
+%
+% Here is the switch from standard \classname{longtable} to the new, \classname{ltxgrid}-compatible values.
+%
+% At this point, we extend \env{longtable} with a \env{longtable*} form, which signifies that we want to
+% use the full page width for setting the table.
+% You can think this way: \env{longtable*} is to \env{longtable} as \env{table*} is to \env{table}.
+%
+%FIXME: the following is no longer true:
+%% Note that it is not enough to define the environment itself; we also have to create the corresponding
+%% \cmd\output\ routine procedures, which provide for continued footers and headers
+%% (the very feature of \env{longtable} requiring support in the output routine).
+%
+%% This same consideration would arise in defining any syntactic extension to \env{longtable}, because
+%% the environment name itself is exposed in the output routine.
+%
+% \begin{macrocode}
+\def\switch@longtable{%
+ \@ifpackageloaded{longtable}{%
+ \@ifx{\longtable\longtable@longtable}{%
+ \@ifx{\endlongtable\endlongtable@longtable}{%
+ \@ifx{\LT@start\LT@start@longtable}{%
+ \@ifx{\LT@end@hd@ft\LT@end@hd@ft@longtable}{%
+ \@ifx{\LT@array\LT@array@longtable}{%
+ \true@sw
+ }{\false@sw}%
+ }{\false@sw}%
+ }{\false@sw}%
+ }{\false@sw}%
+ }{\false@sw}%
+ {%
+ \class@info{Patching longtable package}%
+ }{%
+ \class@info{Patching unrecognized longtable package. (Proceeding with fingers crossed)}%
+ }%
+ \let\longtable\longtable@new
+ \let\endlongtable\endlongtable@new
+ \let\LT@start\LT@start@new
+ \let\LT@end@hd@ft\LT@end@hd@ft@new
+ \let\LT@array\LT@array@new
+ \newenvironment{longtable*}{%
+ \onecolumngrid@push
+ \longtable
+ }{%
+ \endlongtable
+ \onecolumngrid@pop
+ }%
+% \expandafter\let\csname output@init@longtable*\endcsname\output@init@longtable
+% \expandafter\let\csname output@prep@longtable*\endcsname\output@prep@longtable
+% \expandafter\let\csname output@post@longtable*\endcsname\output@post@longtable
+ }{}%
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\LT@pre}
+% \begin{macro}{\LT@bot}
+% \begin{macro}{\LT@top}
+% \begin{macro}{\LT@post}
+% \begin{macro}{\LT@adj}
+% Note that at the end of the longtable environment, we reestablish the \cmd\mark@envir\ of the
+% containing environment. We have left \cmd\curr@envir\ alone, so this will work.
+% \begin{macrocode}
+\def\LT@pre{\penalty\z@\vskip\LTpre}%
+\def\LT@bot{\nobreak\copy\LT@foot\vfil}%
+\def\LT@top{\copy\LT@head\nobreak}%
+\def\LT@post{\penalty\z@\addvspace\LTpost\mark@envir{\curr@envir}}%
+\def\LT@adj{%
+ \setbox\z@\vbox{\null}\dimen@-\ht\z@
+ \setbox\z@\vbox{\unvbox\z@\LT@bot}\advance\dimen@\ht\z@
+ \global\advance\vsize-\dimen@
+}%
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{output@init}
+% \begin{macro}{output@prep}
+% \begin{macro}{output@post}
+% \begin{macrocode}
+\def\output@init@longtable{\LT@adj}%
+\def\output@prep@longtable{\setbox\@cclv\vbox{\unvbox\@cclv\LT@bot}}%
+\def\output@post@longtable{\LT@top}%
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+%
+% \subsection{Patches for index processing}%
+%
+% Another feature that uses the output routine hooks occurs within
+% an index, where one wishes to apply a ``continue head'' when a
+% column breaks within a primary index entry.
+% Some book designs call for the continue head to only be applied
+% at a turnpage break.
+%
+% In any case, it is easy enough for \cmd\output@post@theindex\
+% to do this in conjunction with component marks.
+% Only the bare outlines are shown here.
+%
+% \begin{macro}{\output@init}
+% \begin{macro}{\output@prep}
+% \begin{macro}{\output@post}
+% \begin{macrocode}
+\let\output@init@theindex\@empty
+\let\output@prep@theindex\@empty
+\def\output@post@theindex{%
+ \@ifodd\c@page{}{%
+ \@ifnum{\pagegrid@cur=\@ne}{% we have the leftmost column of a verso page
+ % insert the current top-level continued head
+ }%
+ }%
+}%
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+%
+% \subsection{Checking the auxiliary file}%
+%
+% We relegate the checking of the auxiliary file to the output routine.
+% This task must wait until the last page is shipped out, because otherwise
+% the stream might get closed before the last page is shipped out.
+% Obviously, we must use \cmd\do@output@MVL\ for the job.
+%
+% \begin{macro}{\check@aux}
+% \begin{macrocode}
+\def\check@aux{\do@output@MVL{\do@check@aux}}%
+% \end{macrocode}
+% \end{macro}
+%
+%
+%
+% \subsection{Dealing with stuck floats and stalled float dequeueing}%
+%
+% \LaTeX's float placement mechanism is fundamentally flawed, as evidenced by
+% its warning message ``too many unprocessed floats'', which users understandably find frustrating.
+% The \classname{ltxgrid} package provides tools for ameliorating the situation somewhat.
+%
+% Two cases require detection and rectification:
+% \begin{enumerate}
+% \item
+% A float is ``stuck'' in the \cmd\@deferlist: for whatever reason, the float fails to be committed,
+% even at the start of a fresh page.
+% Once this condition prevails, following floats can never be committed, subsequently all of \LaTeX's
+% float registers are used up.
+%
+% If this condition is detected, we reconsider float dequeueing under permissive (\cmd\clearpage-style) processing.
+%
+% \item
+% The \cmd\@freelist\ is exhausted:
+% a large concentration of floats, say, uses up all of \LaTeX's float registers all at once.
+% This condition commonly occurs when the user collects floats at the end of the document, for some reason.
+%
+% When a float is encountered, \LaTeX\ uses a float register (allocated from a pool of free registers) to contain it until it can be placed.
+% However, no further action is taken until the pagebuilder is visited, so floats can accumulate.
+% Also, even after the pagebuilder is visited, deferred floats can accumulate, and these are not committed
+% until a column (or page) of text is completed.
+%
+% Once the last free float register is used, action should be taken that will commit some of the deferred floats,
+% even if this might require ending the page right where we are (resulting in a short page).
+%
+% Perhaps, committed floats should be stored using some mechanism other than a list, as is currently done.
+% A feasible alternative storage method would be to use a \cmd\box\ register in place of
+% \cmd\@toplist,
+% \cmd\@botlist, and
+% \cmd\@dbltoplist.
+% This is probably just fine, since such committed floats are not reconsidered (I think).
+%
+% \end{enumerate}
+%
+% The emergency processing implemented here immediately ends the current page and begins to output float pages under (\cmd\clearpage-style) rules.
+% It proceeds until all deferred floats have been flushed.
+%
+% Users should expect non-optimal page makeup under these circumstances.
+%
+% Note that there is a weakness in our approach that we have not attempted to repair: if floats are being
+% added as part of a paragraph, we will not be able to take these remedial steps until the paragraph ends.
+% This means that the approach implemented here cannot fix all \LaTeX\ documents. Users can still construct
+% documents that exhaust \LaTeX's pool of float registers!
+%
+% \begin{macro}{\check@deferlist@stuck}
+% \begin{macro}{\@outputpage}
+% We detect the case where, at the start of a fresh page, there are deferred floats, but none are
+% committed. We memorize the \cmd\@deferlist\ at \cmd\shipout\ time, then examine it at the point where
+% our efforts to commit floats to the new page are complete.
+% If it has not changed, the first float must be stuck, and we
+% attempt to fix things via \cmd\force@deferlist@stuck.
+%
+% This simple approach is comp[letely effective in for typical documents.
+%
+% Note that we try to avoid an infinite loop by examining the value of \cmd\clearpage@sw:
+% if we come here with that boolean true, we are in a loop.
+% \begin{macrocode}
+\def\check@deferlist@stuck#1{%
+ \@ifx{\@deferlist@postshipout\@empty}{}{%
+ \@ifx{\@deferlist@postshipout\@deferlist}{%
+ \@fltstk
+ \clearpage@sw{%
+ \ltxgrid@warn{Deferred float stuck during \string\clearpage\space processing}%
+ }{%
+ \force@deferlist@stuck#1%
+ }%
+ }{%
+ %Successfully committed float(s)
+ }%
+ \global\let\@deferlist@postshipout\@empty
+ }%
+}%
+\def\@fltstk{%
+ \@latex@warning{A float is stuck (cannot be placed without \string\clearpage)}%
+}%
+\appdef\@outputpage{%
+ \global\let\@deferlist@postshipout\@deferlist
+}%
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\@next}
+% \begin{macro}{\@xnext}
+% We rewrite the \LaTeX\ kernel macros that dequeue float registers from, e.g., \cmd\@deferlist,
+% providing a test for the condition where the pool of free registers is about to underflow.
+%
+% In this case, we attempt to fix things via \cmd\force@deferlist@empty.
+% \begin{macrocode}
+\def\@next#1#2{%
+ \@ifx{#2\@empty}{\false@sw}{%
+ \expandafter\@xnext#2\@@#1#2%
+ \true@sw
+ }%
+}%
+\def\@xnext\@elt#1#2\@@#3#4{%
+ \def#3{#1}%
+ \gdef#4{#2}%
+ \def\@tempa{#4}\def\@tempb{\@freelist}%
+ \@ifx{\@tempa\@tempb}{%
+ \@ifx{#4\@empty}{%
+ \force@deferlist@empty%{Float register pool exhausted}%
+ }{}%
+ }{}%
+}%
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\force@deferlist@stuck}
+% \begin{macro}{\force@deferlist@empty}
+% \begin{macro}{\force@deferlist@sw}
+% \begin{macro}{\do@forcecolumn@pen}
+% \begin{macro}{\do@forcecolumn}
+% The procedure \cmd\force@deferlist@empty\ is an attempt to rectify a situation where \LaTeX's float placement mechanism
+% may fail (``too many unprocessed floats'').
+%
+% We put down interrupts that call for the float placement to be redone, but under permissive conditions,
+% just the same as if \cmd\clearpage\ had been invoked.
+%
+% Note that the attempt to rectify the error is contingent on the setting of \cmd\force@deferlist@sw,
+% default false. A document class using this package that wishes to enable this error recovery mechanism should
+% set this boolean to true.
+%
+% The interrupt \cmd\do@forcecolumn@pen, which invokes the procedure \cmd\do@forcecolumn,
+% does the same as \cmd\do@startcolumn, except under permissive conditions:
+% we are trying to empty out the float registers completely.
+%
+% In order to properly with the case where there is material in \cmd\box\cmd\@cclv,
+% \cmd\@toplist, \cmd\@botlist, \cmd\@dbltoplist, etc,
+% we do what amounts to \cmd\newpage\ to get things rolling.
+%
+% In \cmd\force@deferlist@stuck, we take advantage of already being in the output routine:
+% simply reinvoke \cmd\do@startcolumn\ under permissive conditions.
+%
+% \begin{macrocode}
+\def\force@deferlist@stuck#1{%
+ \force@deferlist@sw{%
+ \@booleantrue\clearpage@sw
+ \@booleantrue\forcefloats@sw
+ #1%
+ }{%
+ }%
+}%
+\def\force@deferlist@empty{%
+ \force@deferlist@sw{%
+% \ltxgrid@info{#1, attempting rectification}%
+ \penalty-\pagebreak@pen
+ \protect@penalty\do@forcecolumn@pen
+ }{%
+% \ltxgrid@info{#1}%
+ }%
+}%
+\@booleanfalse\force@deferlist@sw
+\mathchardef\do@forcecolumn@pen=10009
+\@namedef{output@-\the\do@forcecolumn@pen}{\do@forcecolumn}%
+\def\do@forcecolumn{%
+ \@booleantrue\clearpage@sw
+ \@booleantrue\forcefloats@sw
+%\unvbox\@cclv
+%\vfil
+%\penalty-\pagebreak@pen
+ \do@startcolumn
+}%
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% A more thorough revision of \LaTeX's float placement mechanism would involve substituting a single \cmd\box\
+% register for the \cmd\@deferlist. This way, \LaTeX's ability to have latent floats would be limited by
+% box memory alone.
+%
+% Because only the \cmd\box\ and \cmd\count\ components of the float box register are actually used by \LaTeX,
+% our scheme can be accomplished if we can find a way to encode the information held in the \cmd\count\ component.
+%
+% A first-in, first-out mechanism exists, wherein a box-penalty pair is dequeued by \cmd\lastbox\cmd\lastpenalty\cmd\unpenalty\ and enqueued
+% by \cmd\setbox\cmd\foo=\cmd\hbox\cmd\bgroup\cmd\penalty\cmd\floatpenalty\cmd\box\cmd\floatbox\cmd\unhbox\cmd\foo\cmd\egroup.
+%
+% Note that this scheme is made possible by our change to \LaTeX's float placement mechanism,
+% wherein we consolidated the two \cmd\@deferlist s into one.
+%
+% \section{Support for legacy \LaTeX\ commands}
+%
+% We provide support for the \cmd\enlargethispage\ command.
+%
+% Note: using a command of this sort is questionable.
+% Instead, people should enlarge the entire spread.
+%
+% Timing Note: In a multicolumn page grid, the user should issue the \cmd\enlargethispage\ command
+% while the first column of the page is being typeset.
+% We provide a helpful message if the timing is wrong.
+%
+% This code can serve as a model for introducing commands that need to execute within the safety of the output routine.
+% We ensure that the arguments are fully expanded, then execute \cmd\do@output@MVL\ to cause an output procedure,
+% \cmd\@@enlargethispage, to execute. When it does execute, the MVL will be exposed.
+%
+% The \cmd\@@enlargethispage\ procedure simply adjusts the vertical dimensions of the page.
+% The adjustment will persist until the column is committed, at which point the page dimension
+% will revert to its standard value.
+% \begin{macrocode}
+\def\enlargethispage{%
+ \@ifstar{%
+ \@enlargethispage{}%
+ }{%
+ \@enlargethispage{}%
+ }%
+}%
+\def\@enlargethispage#1#2{%
+ \begingroup
+ \dimen@#2\relax
+ \edef\@tempa{#1}%
+ \edef\@tempa{\noexpand\@@enlargethispage{\@tempa}{\the\dimen@}}%
+ \expandafter\do@output@MVL\expandafter{\@tempa}%
+ \endgroup
+}%
+\def\@@enlargethispage#1#2{%
+ \def\@tempa{one}%
+ \@ifx{\thepagegrid\@tempa}{%
+ \true@sw
+ }{%
+ \def\@tempa{mlt}%
+ \@ifx{\thepagegrid\@tempa}{%
+ \@ifnum{\pagegrid@cur=\@ne}{% OK to adjust this page
+ \gdef\enlarge@colroom{#2}%
+ \true@sw
+ }{% Can only adjust this column; give up
+ \ltxgrid@warn{Too late to enlarge this page; move the command to the first column.}%
+ \false@sw
+ }%
+ }{% Unknown page grid
+ \ltxgrid@warn{Unable to enlarge a page of this kind.}%
+ \false@sw
+ }%
+ }%
+ {%
+ \class@info{Enlarging page \thepage\space by #2}%
+ \global\advance\@colroom#2\relax
+ \set@vsize
+ }{%
+ % Could not adjust this page
+ }%
+}%
+\let\enlarge@colroom\@empty
+% \end{macrocode}
+% The \cmd\@kludgeins\ insert register is now unneeded.
+% Ensure that packages using this mechanism break (preferrable to subtle bugs).
+% \begin{macrocode}
+\let\@kludgeins\@undefined
+% \end{macrocode}
+%
+% \subsubsection{Building the page for shipout}
+%
+% \begin{macro}{\@outputpage}
+% The procedures that build \cmd\@outputbox\ just before a page is shipped out by \cmd\@outputpage\ are:
+% \cmd\@makecol,
+% \cmd\@combinepage, and
+% \cmd\@combinedblfloats.
+% We headpatch \cmd\@outputpage\ to make the \cmd\@outputbox\ be of fixed height.
+% \begin{macrocode}
+\@booleantrue\textheight@sw
+\prepdef\@outputpage{%
+ \textheight@sw{%
+ \count@\vbadness\vbadness\@M
+ \dimen@\vfuzz\vfuzz\maxdimen
+ \setbox\@outputbox\vbox to\textheight{\unvbox\@outputbox}%
+ \vfuzz\dimen@
+ \vbadness\count@
+ }{}%
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \subsubsection{Warning message}
+%
+% \begin{macro}{\ltxgrid@info}
+% \begin{macro}{\ltxgrid@warn}
+% Something has happened that the user might be interested in.
+% Print a message to the log, but only if the user selected the verbose option.
+% \begin{macrocode}
+\def\ltxgrid@info{%
+ \ltxgrid@info@sw{\class@info}{\@gobble}%
+}%
+\@booleanfalse\ltxgrid@info@sw
+\def\ltxgrid@warn{%
+ \ltxgrid@warn@sw{\class@warn}{\@gobble}%
+}%
+\@booleantrue\ltxgrid@warn@sw
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \section{End of the \file{ltxgrid} {\sc docstrip} module}
+% Here ends the module.
+% \begin{macrocode}
+%</ltxgrid-krn>
+% \end{macrocode}
+%
+% \Finale
+% %Here ends the programmer's documentation.
+% \endinput
+%
+\endinput
+%%EOF
diff --git a/Master/texmf-dist/source/latex/revtex4/ltxutil.dtx b/Master/texmf-dist/source/latex/revtex4/ltxutil.dtx
new file mode 100644
index 00000000000..5553e3a863e
--- /dev/null
+++ b/Master/texmf-dist/source/latex/revtex4/ltxutil.dtx
@@ -0,0 +1,3534 @@
+% \iffalse ltxdoc klootch
+% ltxutil.dtx: package to add utilties to LaTeX
+% Copyright (c) 2000 Arthur Ogawa
+%
+% Disclaimer
+% This file is distributed WITHOUT ANY WARRANTY;
+% without even the implied warranty of
+% MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
+% ReadMe
+% For the documentation and more detailed instructions for
+% installation, typeset this document with \LaTeX.
+% \fi
+% \GetFileInfo{ltxutil.dtx}\CheckSum{3641}
+%
+% \iffalse ltxdoc klootch
+%<*ltxutil>
+%%% @LaTeX-file{
+%%% filename = "ltxutil.dtx",
+%%% version = "1.0rc5b",
+%%% date = "2001/07/31",
+%%% time = "12:23:00 GMT+8",
+%%% checksum = "3641",
+%%% author = "Arthur Ogawa (mailto:ogawa@teleport.com),
+%%% commissioned by the American Physical Society.
+%%% ",
+%%% copyright = "Copyright (C) 1999 Arthur Ogawa,
+%%% distributed under the terms of the
+%%% LaTeX Project Public License, see
+%%% ftp://ctan.tug.org/macros/latex/base/lppl.txt
+%%% ",
+%%% address = "Arthur Ogawa,
+%%% USA",
+%%% telephone = "",
+%%% FAX = "",
+%%% email = "ogawa@teleport.com",
+%%% codetable = "ISO/ASCII",
+%%% keywords = "latex, utility, kernel",
+%%% supported = "yes",
+%%% abstract = "package to add utilties to LaTeX",
+%%% docstring = "The checksum field above generated by ltxdoc",
+%%% }
+%</ltxutil>
+% \fi
+%
+% \iffalse ltxdoc klootch
+% The following references the \file{00readme.txt} file,
+% which contains basic information about this package.
+% The contents of this file are generated when
+% you typeset the programmer's documentation.
+% Search on "{filecontents*}{00readme.txt}" to locate it.
+% \fi\input{00readme.txt}%
+%
+% \subsection{Bill of Materials}
+%
+% Following is a list of the files in this distribution arranged
+% according to provenance.
+%
+% \subsubsection{Primary Source}%
+% One single file generates all.
+%\begin{verbatim}
+%ltxutil.dtx
+%\end{verbatim}
+%
+% \subsubsection{Generated by \texttt{latex ltxutil.dtx}}%
+% Typesetting the source file under \LaTeX\
+% generates the readme and the installer.
+%\begin{verbatim}
+%00readme.txt ltxutil.ins
+%\end{verbatim}
+%
+% \subsubsection{Generated by \texttt{tex ltxutil.ins}}%
+% Typesetting the installer generates
+% the package files.
+%\begin{verbatim}
+%ltxutil.sty
+%\end{verbatim}
+%
+% \subsubsection{Documentation}%
+% The following are the online documentation:
+% \begin{verbatim}
+%ltxutil.pdf
+% \end{verbatim}
+%
+% \subsubsection{Auxiliary}%
+% The following are auxiliary files generated
+% in the course of running \LaTeX:
+% \begin{verbatim}
+%ltxutil.aux ltxutil.idx ltxutil.ind ltxutil.log ltxutil.toc
+% \end{verbatim}
+%
+% \section{Code common to all modules}%
+%
+% The following may look a bit klootchy, but we
+% want to require only one place in this file
+% where the version number is stated,
+% and we also want to ensure that the version
+% number is embedded into every generated file.
+%
+% Now we declare that
+% these files can only be used with \LaTeXe.
+% An appropriate message is displayed if
+% a different \TeX{} format is used.
+% \begin{macrocode}
+%<*doc|ltxutil>
+\NeedsTeXFormat{LaTeX2e}[1995/12/01]%
+%</doc|ltxutil>
+% \end{macrocode}
+% As desired, the following modules all
+% take common version information:
+% \begin{macrocode}
+%<ltxutil>\ProvidesFile{ltxutil.sty}%
+%<*doc>
+\expandafter\ProvidesFile\expandafter{\jobname.dtx}%
+%</doc>
+% \end{macrocode}
+%
+% The following line contains, for once and for all,
+% the version and date information.
+% By various means, this information is reproduced
+% consistently in all generated files and in the
+% typeset documentation.
+% \begin{macrocode}
+%<*doc|ltxutil>
+ [2001/07/31 1.0rc5b utilities package]% \fileversion
+%</doc|ltxutil>
+% \end{macrocode}
+%
+%
+% \section{The driver module \texttt{doc}}
+%
+% This module, consisting of the present section,
+% typesets the programmer's documentation,
+% generating the \file{.ins} installer and \file{00readme.txt} as required.
+%
+% Because the only uncommented-out lines of code at the beginning of
+% this file constitute the \file{doc} module itself,
+% we can simply typeset the \file{.dtx} file directly,
+% and there is thus rarely any need to
+% generate the ``doc'' {\sc docstrip} module.
+% Module delimiters are nonetheless required so that
+% this code does not find its way into the other modules.
+%
+% The \enve{document} command concludes the typesetting run.
+%
+% \begin{macrocode}
+%<*doc>
+% \end{macrocode}
+%
+% \subsection{The Preamble}
+% The programmers documentation is formatted
+% with the \classname{ltxdoc} class with local customizations,
+% and with the usual code line indexing.
+% \begin{macrocode}
+\documentclass{ltxdoc}
+\RequirePackage{ltxdocext}%
+\let\url\undefined
+\RequirePackage[colorlinks=true,linkcolor=blue]{hyperref}%
+\expandafter\ifx\csname package@font\endcsname\@undefined\else
+ \expandafter\RequirePackage\expandafter{\csname package@font\endcsname}%
+\fi
+\CodelineIndex\EnableCrossrefs
+% \end{macrocode}
+%
+% \subsubsection{Docstrip and info directives}
+% We use so many {\sc docstrip} modules that we set the
+% \texttt{StandardModuleDepth} counter to 1.
+% \begin{macrocode}
+\setcounter{StandardModuleDepth}{1}
+% \end{macrocode}
+% The following command retrieves the date and version information
+% from this file.
+% \begin{macrocode}
+\expandafter\GetFileInfo\expandafter{\jobname.dtx}%
+% \end{macrocode}
+%
+%
+% \subsection{The installer file}
+%
+% The installer \file{ltxutil.ins} appears here.
+% If you have retrieved the standard distribution of this package,
+% the installer file is already on your filesystem.
+% If you are bootstrapping,
+% the first typesetting of the \file{.dtx} file
+% will cause the installer to be generated.
+%
+% The following modules are used to direct
+% {\sc docstrip} in generating the external files:
+% \begin{center}
+% \begin{tabular}{lll}
+% \textbf{Module}&\textbf{File}&\textbf{Description}\\
+% doc &\file{ltxutil.drv}&driver for programmer's documentation\\
+% ltxutil,ltxutil-krn &\file{ltxutil.sty}&this package\\
+% ltxutil-krn& &the portion of this package suitable for inclusion within another package
+% \end{tabular}
+% \end{center}
+%
+% \begin{macrocode}
+\begin{filecontents}{ltxutil.ins}
+%% This file will generate documentation and runtime files
+%% from ltxutil.dtx when run through LaTeX or TeX.
+\input docstrip
+\preamble
+
+This is a generated file;
+altering it directly is inadvisable;
+instead, modify the original source file.
+See the URL in the file 00readme.txt.
+
+Copyright notice.
+
+ These files are distributed
+ WITHOUT ANY WARRANTY; without even the implied warranty of
+ MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
+
+\endpreamble
+\keepsilent
+ \generate{%
+ \file{ltxutil.drv}{\from{ltxutil.dtx}{doc}}%
+ \file{ltxutil.sty}{%
+ \from{ltxutil.dtx}{ltxutil,ltxutil-krn}%
+ }%
+ }%
+\ifToplevel{
+\Msg{***********************************************************}
+\Msg{*}
+\Msg{* To finish the installation, please move}
+\Msg{* ltxutil.sty}
+\Msg{* into a directory searched by TeX.}
+\Msg{*}
+\Msg{* To produce the documentation,
+ run ltxutil.dtx through LaTeX.}
+\Msg{*}
+\Msg{* Happy TeXing}
+\Msg{***********************************************************}
+}
+\endbatchfile
+\end{filecontents}
+% \end{macrocode}
+% Note that, because all of the files generated by the installer
+% are part of the standard distribution, it will
+% be necessary to run the installer only when bootstrapping
+% (or, of course, during development).
+% Note, too, that it is rare to generate the \file{doc}
+% module because it suffices to simply typeset the \file{.dtx} file itself.
+%
+% \subsection{The ``Read Me'' File}
+% As promised above, here is the contents of the
+% ``Read Me'' file. That file serves a double purpose,
+% since it also constitutes the beginining of the
+% programmer's documentation. What better thing, after
+% all, to have appear at the beginning of the
+% typeset documentation?
+%
+% A good discussion of how to write a ReadMe file can be found in
+% Engst, Tonya, ``Writing a ReadMe File? Read This''
+% \emph{MacTech} October 1998, p. 58.
+%
+% Note the appearance of the
+% \cmd\StopEventually\ command, which marks the
+% dividing line between the user documentation
+% and the programmer documentation.
+%
+% The usual user will not be asked to
+% do a full build, not to speak
+% of the bootstrap.
+% Instructions for carrying these processes
+% begin the programmer's manual.
+%
+% \begin{macrocode}
+\begin{filecontents*}{00readme.txt}
+\title{%
+ A \LaTeX\ Package of utility macros%
+ \thanks{%
+ This file has version number \fileversion,
+ last revised \filedate.%
+ % For version number and date,
+ % search on "\fileversion" in the .dtx file,
+ % or see the end of the 00readme.txt file.
+ }%
+}%
+
+\author{%
+Arthur Ogawa (\texttt{mailto:ogawa@teleport.com}),
+\fileversion\\Copyright (C) 1999 Arthur Ogawa
+}%
+\maketitle
+
+This file embodies the \classname{ltxutil} package,
+the implementation and its user documentation.
+
+The distribution point for this work is
+\url{ftp://ftp.teleport.com/users/ogawa/macros/latex/contrib/supported/ltxutil...},
+which contains fully unpacked, prebuilt runtime files and documentation.
+
+The \classname{ltxutil} package was commissioned by the American Physical Society
+and is distributed under the terms of the \LaTeX\ Project Public License,
+the same license under which all the portions of \LaTeX\ itself is distributed.
+Please see \url{http://ctan.tug.org/macros/latex/base/lppl.txt} for details.
+
+To use this document class, you must have a working
+\TeX\ installation equipped with \LaTeXe\
+and possibly pdftex and Adobe Acrobat Reader or equivalent.
+
+To install, retrieve the distribution,
+unpack it into a directory on the target computer,
+and move the file \file{ltxutil.sty}
+into a location in your filesystem where it will be found by \LaTeX.
+
+To use, read the user documentation \file{ltxutil.pdf}.
+
+\tableofcontents
+
+\section{Processing Instructions}
+
+The package file \file{ltxutil.sty}
+is generated from this file, \file{ltxutil.dtx},
+using the {\sc docstrip} facility of \LaTeX
+via |tex ltxutil.ins|.
+The typeset documentation that you are now reading is generated from
+the same file by typesetting it with \LaTeX\ or pdftex
+via |latex ltxutil.dtx| or |pdflatex ltxutil.dtx|.
+
+\subsection{Build Instructions}
+
+You may bootstrap this suite of files solely from \file{ltxutil.dtx}.
+Prepare by installing \LaTeXe\ (and either tex or pdftex) on your computer,
+then carry out the following steps:
+\begin{enumerate}
+\item
+Within an otherwise empty directory,
+typeset \file{ltxutil.dtx} with \LaTeX\ or pdflatex;
+you will obtain the typeset documentation you are now reading,
+along with
+the installer \file{ltxutil.ins},
+and the file \file{00readme.txt}.
+
+Note: you will have to run \LaTeX\ twice, then \file{makeindex}, then
+\LaTeX\ again in order to obtain a valid index and table of contents.
+\item
+Now typeset \file{ltxutil.ins},
+thereby generating the package file \file{ltxutil.sty}.
+\item
+Install \classname{ltxutil.sty}
+by moving it to a location
+in your filesystem where they will be found by \LaTeX.
+\end{enumerate}
+\end{filecontents*}
+% \end{macrocode}
+%
+% \subsection{The Document Body}
+%
+% Here is the document body, containing only a
+% \cmd\DocInput\ directive---referring to this very file.
+% This very cute self-reference is a common \classname{ltxdoc} idiom.
+% \begin{macrocode}
+\begin{document}%
+\expandafter\DocInput\expandafter{\jobname.dtx}%
+% ^^A\PrintChanges
+\end{document}
+% \end{macrocode}
+%
+% \begin{macrocode}
+%</doc>
+% \end{macrocode}
+%
+% \section{Using this package}
+% Once this package is installed on your filesystem, you can employ it in
+% adding functionality to \LaTeX\ by invoking it in your document or document class.
+%
+% \subsection{Invoking the package}
+% In your document, you can simply call it up in your preamble:
+% \begin{verbatim}
+%\documentclass{book}%
+%\usepackage{ltxutil}%
+%\begin{document}
+%<your document here>
+%\end{document}
+% \end{verbatim}
+% However, the preferred way is to invoke this package from within your
+% customized document class:
+% \begin{verbatim}
+%\NeedsTeXFormat{LaTeX2e}[1995/12/01]%
+%\ProvidesClass{myclass}%
+%\RequirePackage{ltxutil}%
+%\LoadClass{book}%
+%<class customization commands>
+%\endinput
+% \end{verbatim}
+%
+% Once loaded, the package gives you acccess to certain procedures,
+% usually to be invoked by a \LaTeX\ command or environment, but not at the document level.
+%
+%
+% \section{Compatability with \LaTeX's Required Packages}
+% Certain packages, usually ones written by members of the
+% \LaTeX\ Project itself, have been designated ``required'' and
+% are distributed as part of standard \LaTeX.
+% These packages have been placed in a priviledged position
+% vis \'a vis the \LaTeX\ kernel in that they override the definitions of certain kernel macros.
+%
+% The \classname{ltxutil} package will be incompatible with any package that
+% redefines any of the kernel macros that \classname{ltxutil} patches---if that
+% package is loaded \emph{after} \classname{ltxutil}. This means that for
+% greatest compatability, \classname{ltxutil} should be loaded \emph{after},
+% say, \classname{ftnright}, which overwrites \LaTeX's kernel
+% procedures \cmd\@outputdblcol, \cmd\@startcolumn, and \cmd\@makecol.
+%
+% Hereinafter follows some notes on specific \LaTeX\ packages.
+%
+% \subsection{array}
+% This package alters the way tabular environments are done,
+% therefore it could run afoul of the \LaTeX\ ``required'' package \classname{array} or any
+% package that calls for it to be loaded.
+% However, this package has provisions for remaining compatible with \classname{array}.
+% So long as the version of \classname{array} that is used with this package has the appropriate
+% meanings for the procedures it overwrites, all should be well.
+%
+% \subsection{longtable}
+% David Carlisle's \classname{longtable} package modifies both the \LaTeX\ kernel and the
+% \classname{array} package. This package must therefore alter \cmd\LT@array.
+% For now, that job is handled by \classname{ltxgrid}.
+%
+%
+%\StopEventually{}
+%
+% \section{Implementation of package}
+%
+% Special acknowledgment: this package uses concepts pioneered
+% and first realized by William Baxter (mailto:web@superscript.com)
+% in his SuperScript line of commercial typesetting tools, and
+% which are used here with his permission.
+%
+% \subsection{Beginning of the \file{ltxutil} {\sc docstrip} module}
+% \begin{macrocode}
+%<*ltxutil>
+\def\package@name{ltxutil}%
+\expandafter\PackageInfo\expandafter{\package@name}{%
+ Utility macros for \protect\LaTeXe,
+ by A. Ogawa (ogawa@teleport.com)%
+}%
+%</ltxutil>
+% \end{macrocode}
+%
+% \subsection{Banner}%
+% Credit where due.
+% \begin{macrocode}
+%<*ltxutil-krn>
+\typeout{%
+ ltxutil: portions licensed from W. E. Baxter (web@superscript.com)%
+}%
+% \end{macrocode}
+%
+% \subsection{Errors and warnings}
+%
+% \begin{macro}{\class@err}
+% \begin{macro}{\class@warn}
+% \begin{macro}{\class@info}
+% A few shorthands for Class messages.
+% Your document class should define \cmd\class@name.
+% \begin{macrocode}
+\def\class@err#1{\ClassError{\class@name}{#1}\@eha}%
+\def\class@warn#1{\ClassWarningNoLine{\class@name}{#1}}%
+\def\class@info#1{\ClassInfo{\class@name}{#1}}%
+\def\obsolete@command#1{%
+ \class@warn@end{Command \string#1\space is obsolete.^^JPlease remove from your document}%
+ \global\let#1\@empty
+ #1%
+}%
+\def\replace@command#1#2{%
+ \class@warn@end{Command \string#1\space is obsolete;^^JUse \string#2\space instead}%
+ \global\let#1#2%
+ #1%
+}%
+\def\replace@environment#1#2{%
+ \class@warn@end{Environment #1 is obsolete;^^JUse #2 instead}%
+ \glet@environment{#1}{#2}%
+ \@nameuse{#1}%
+}%
+\def\incompatible@package#1{%
+ \@ifpackageloaded{#1}{%
+ \def\@tempa{I cannot continue. You must remove the \string\usepackage\ statement that caused that package to be loaded.}%
+ \ClassError{\class@name}{The #1 package cannot be used with \class@name}%
+ \@tempa\stop
+ }{%
+ \class@info{#1 was not loaded (OK!)}%
+ }%
+}%
+\def\class@warn@end#1{%
+ \gappdef\class@enddocumenthook{\class@warn{#1}}%
+}%
+\AtEndOfClass{%
+ \@ifxundefined\class@name{\def\class@name{Generic Class}}{}%
+}%
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \subsection{New Tools}%
+%
+% \begin{macro}{\t@}
+% \begin{macrocode}
+\def\t@{to}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\dimen@iii}
+% \begin{macrocode}
+\dimendef\dimen@iii\thr@@
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\halignt@}
+% \begin{macrocode}
+\def\halignt@{\halign\t@}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\f@ur}
+% Analogous to \cmd\@ne, \cmd\tw@, and \cmd\thr@@.
+% \begin{macrocode}
+\chardef\f@ur=4\relax
+\chardef\cat@letter=11\relax
+\chardef\other=12\relax
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\let@environment}
+% \begin{macro}{\glet@environment}
+% The directive \cmd\let@environment\ takes care of a common programming
+% idiom whereby one environment is made a synonym for another.
+% \begin{macrocode}
+\def\let@environment#1#2{%
+ \expandafter\let
+ \csname#1\expandafter\endcsname\csname#2\endcsname
+ \expandafter\let
+ \csname end#1\expandafter\endcsname\csname end#2\endcsname
+}%
+\def\glet@environment#1#2{%
+ \global\expandafter\let
+ \csname#1\expandafter\endcsname\csname#2\endcsname
+ \global\expandafter\let
+ \csname end#1\expandafter\endcsname\csname end#2\endcsname
+}%
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\tracingplain}
+% The command \cmd\tracingplain\ causes \TeX's tracing parameters to
+% return to the values set by default. This command is sometimes
+% useful when you have said \cmd\tracingall\ somewhere and want to
+% restore.
+% The \cmd\traceoutput\ command causes \cmd\tracingoutput\ diagnostics
+% upon \cmd\shipout.
+% \begin{macrocode}
+\newcommand\tracingplain{%
+ \tracingonline\z@\tracingcommands\z@\tracingstats\z@
+ \tracingpages\z@\tracingoutput\z@\tracinglostchars\@ne
+ \tracingmacros\z@\tracingparagraphs\z@\tracingrestores\z@
+ \showboxbreadth5\showboxdepth3\relax %\errorstopmode
+ }%
+\newcommand\traceoutput{%
+ \appdef\@resetactivechars{\showoutput}%
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\say}
+% \begin{macro}{\saythe}
+% The commands \cmd\say\ and \cmd\saythe\ cause diagnostic messages in the
+% \TeX\ log that give the value of a control sequence name or a register
+% respectively.
+% \begin{macrocode}
+\newcommand\say[1]{\typeout{<\noexpand#1=\meaning#1>}}%
+\newcommand\saythe[1]{\typeout{<\noexpand#1=\the#1>}}%
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\fullinterlineskip}
+% Resets the \cmd\prevdepth\ so that the full amount of \cmd\baselineskip\ glue will be inserted by
+% the \cmd\baselinesklip\ mechanism.
+% Can be invoked just after a \cmd\hrule\ to undo its default suppression of base line skip.
+% \begin{macrocode}
+\def\fullinterlineskip{\prevdepth\z@}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\count@i}
+% \begin{macro}{\count@ii}
+%
+% \begin{macrocode}
+\countdef\count@i\@ne
+\countdef\count@ii\tw@
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+%
+% \subsection{Boolean Control}%
+% We introduce just enough of the Boolean calculus for \TeX.
+% Alan Jeffrey was the pioneer here, with an article in TUGboat
+% (Vol. 11, No. 2, page 237).
+% This implementation owes a debt to
+% William Baxter (web@superscript.com).
+% See articles by Baxter and Ogawa in the proceedings of the
+% 1994 TUG meeting, TUGboat Vol.~15, No.~3.
+%
+% \begin{macro}{\prepdef}
+% \begin{macro}{\appdef}
+% \begin{macro}{\gappdef}
+%
+% Provide the capability of performing head- and tail patches.
+% The procedure \cmd\prepdef\ prepends to the given macro
+% the tokens specified in its second argument.
+% Likewise for \cmd\appdef, except that it appends.
+% Note that the first 10 toks registers are utility registers,
+% and we simply make a control sequence name, \cmd\toks@ii, for one of
+% them.
+% \begin{macrocode}
+\long\def\prepdef#1#2{%
+ \@ifxundefined#1{\toks@{}}{\toks@\expandafter{#1}}%
+ \toks@ii{#2}%
+ \edef#1{\the\toks@ii\the\toks@}%
+}%
+\long\def\appdef#1#2{%
+ \@ifxundefined#1{\toks@{}}{\toks@\expandafter{#1}}%
+ \toks@ii{#2}%
+ \edef#1{\the\toks@\the\toks@ii}%
+}%
+\long\def\gappdef#1#2{%
+ \@ifxundefined#1{\toks@{}}{\toks@\expandafter{#1}}%
+ \toks@ii{#2}%
+ \global\edef#1{\the\toks@\the\toks@ii}%
+}%
+\long\def\appdef@val#1#2{%
+ \appdef#1{{#2}}%
+}%
+\long\def\appdef@e#1#2{%
+ \expandafter\appdef
+ \expandafter#1%
+ \expandafter{#2}%
+}%
+\long\def\appdef@eval#1#2{%
+ \expandafter\appdef@val
+ \expandafter#1%
+ \expandafter{#2}%
+}%
+\toksdef\toks@ii=\tw@
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\@ifxundefined}
+% \begin{macro}{\@ifnotrelax}
+% \begin{macro}{\@argswap}
+% \begin{macro}{\@argswap@val}
+%
+% Certain utility procedures use \cmd\@ifxundefined,
+% which is defined here in terms of \cmd\@ifx.
+% Others use \cmd\@ifnotrelax, namely when
+% the control sequence name is manufactured by
+% the use of \cmd\csname.
+%
+% The procedures \cmd\@argswap and \cmd\@argswap@val
+% are used to facilitate control of expansion.
+%
+% \begin{macrocode}
+\long\def\@ifxundefined#1{\@ifx{\undefined#1}}%
+\long\def\@ifnotrelax#1#2#3{\@ifx{\relax#1}{#3}{#2}}%
+\long\def\@argswap#1#2{#2#1}%
+\long\def\@argswap@val#1#2{#2{#1}}%
+\def\@ifxundefined@cs#1{\expandafter\@ifx\expandafter{\csname#1\endcsname\relax}}%
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\@boolean}
+% \begin{macro}{\@boole@def}
+% In order to define \cmd\@ifx, we first must create the
+% ``defining word'' (term taken form our Forth vocabulary)
+% \cmd\@boole@def, which employs \cmd\@boolean\ to do its job.
+% \begin{macrocode}
+\def\@boolean#1#2{%
+ \long\def#1{%
+ #2% \if<something>
+ \expandafter\true@sw
+ \else
+ \expandafter\false@sw
+ \fi
+ }%
+}%
+\def\@boole@def#1#{\@boolean{#1}}% Implicit #2
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\@booleantrue}
+% \begin{macro}{\@booleanfalse}
+% The procedures \cmd\@booleantrue\ and
+% \cmd\@booleanfalse\ are assignment operators
+% for Boolean flags.
+% \begin{macrocode}
+\def\@booleantrue#1{\let#1\true@sw}%
+\def\@booleanfalse#1{\let#1\false@sw}%
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\@ifx}
+% \begin{macro}{\@ifx@empty}
+% \begin{macro}{\@if@empty}
+% \begin{macro}{\@ifcat}%
+% \begin{macro}{\@ifdim}%
+% \begin{macro}{\@ifeof}%
+% \begin{macro}{\@ifhbox}%
+% \begin{macro}{\@ifhmode}%
+% \begin{macro}{\@ifinner}%
+% \begin{macro}{\@ifmmode}%
+% \begin{macro}{\@ifnum}%
+% \begin{macro}{\@ifodd}%
+% \begin{macro}{\@ifvbox}%
+% \begin{macro}{\@ifvmode}%
+% \begin{macro}{\@ifvoid}%
+% We can now invoke the defining word to create
+% the procedures \cmd\@ifx\ and friends.
+%
+% Compatability Note: earlier versions of this package
+% defined a procedure \cmd\@ifempty. However, for compatability with AMS\LaTeX,
+% we must avoid the following three names:
+% \cmd\@ifempty, \cmd\@xifempty, and \cmd\@ifnotempty.
+%
+% \begin{macrocode}
+\@boole@def\@ifx#1{\ifx#1}%
+\@boole@def\@ifx@empty#1{\ifx\@empty#1}%
+\@boole@def\@if@empty#1{\if!#1!}%
+%\@boole@def\@if@sw#1{\csname if#1\endcsname}%
+\def\@if@sw#1#2{#1\expandafter\true@sw\else\expandafter\false@sw#2}%
+\@boole@def\@ifdim#1{\ifdim#1}%
+\@boole@def\@ifeof#1{\ifeof#1}%
+\@boole@def\@ifhbox#1{\ifhbox#1}%
+\@boole@def\@ifhmode{\ifhmode}%
+\@boole@def\@ifinner{\ifinner}%
+\@boole@def\@ifmmode{\ifmmode}%
+\@boole@def\@ifnum#1{\ifnum#1}%
+\@boole@def\@ifodd#1{\ifodd#1}%
+\@boole@def\@ifvbox#1{\ifvbox#1}%
+\@boole@def\@ifvmode{\ifvmode}%
+\@boole@def\@ifvoid#1{\ifvoid#1}%
+% \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}{\true@sw}
+% \begin{macro}{\false@sw}
+%
+% Note that when a Boolean operator expands, it
+% employs two macros that act as selectors, defined here.
+%
+% \begin{macrocode}
+\long\def\true@sw#1#2{#1}%
+\long\def\false@sw#1#2{#2}%
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\loopuntil}
+% \begin{macro}{\loopwhile}
+%
+% Loop control using the Boolean idiom.
+% Superior to \cmd\loop\dots\cmd\repeat\ because these can be nested.
+% The tail of the argument must have a Boolean predicate.
+%
+% \begin{macrocode}
+\long\def\loopuntil#1{#1{}{\loopuntil{#1}}}%
+\long\def\loopwhile#1{#1{\loopwhile{#1}}{}}%
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\@provide}
+%
+% A defining word that refuses to clobber a prior meaning.
+%
+% \begin{macrocode}
+\def\@provide#1{%
+ \@ifx{\undefined#1}{\true@sw}{\@ifx{\relax#1}{\true@sw}{\false@sw}}%
+ {\def#1}{\def\j@nk}%
+}%
+% \end{macrocode}
+% \end{macro}
+%
+%
+% \subsection{Begin Document Structure}
+% The standard \LaTeX\ mechanism \cmd\AtBeginDocument\
+% is inadequate because the \cmd\vsize\ is bound much too early.
+% We supply here a mechanism whereby decisions about the
+% page layout can be deferred until \cmd\AtBeginDocument\ time.
+%
+% The problem we are working around is that the \cmd\AtBeginDocument\
+% hook in \cmd\document\ appears long after the calculation of
+% \cmd\vsize\ and \cmd\hsize, that is, \LaTeX\ provides no mechanism
+% for deferring the decision about the page grid until \cmd\AtBeginDocument\ time.
+% We fix things by prepending a hook at the very beginning of \cmd\document.
+%
+% The price we pay for this facility is to depend on the stability of this part
+% of \LaTeX's kernel code (the first token of \cmd\document),
+% which could change, you see. But considering that \LaTeX\ is
+% at this point essentially stagnant once more, we risk it.
+%
+% \begin{macro}{\document}
+% We begin by installing hooks into \cmd\document\ that
+% we will manage ourselves.
+% First, we do as \cmd\document\ does: end the group
+% begun by \cmd\begin. Last, we conclude our shenanigans by absorbing
+% the first token of the expansion of \cmd\document, which
+% we assume to be \cmd\endgroup.
+% \begin{macrocode}
+\prepdef\document{%
+ \endgroup
+ \init@documenthook
+ \set@typesize@hook
+ \normalsize
+ \set@pica@hook
+ \true@sw{}%
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\class@documenthook}
+% \begin{macro}{\class@enddocumenthook}
+% We install the first \cmd\AtBeginDocument\ hook, namely the
+% procedure \cmd\class@documenthook. Within the document class,
+% we will use this hook exclusively, so as to avoid interference from other packages.
+% Similarly with \cmd\class@enddocumenthook, installed via \cmd\AtEndDocument.
+%
+% A document class using this package should do as this package does and
+% just say, \cmd\appdef\cmd\class@documenthook\ and \cmd\appdef\cmd\class@enddocumenthook\
+% instead of \cmd\AtBeginDocument\ and \cmd\AtEndDocument.
+% \begin{macrocode}
+\def\init@documenthook{}%
+\AtBeginDocument{%
+ \class@documenthook
+}%
+\AtEndDocument{%
+ \class@enddocumenthook
+}%
+\def\class@documenthook{}%
+\def\class@enddocumenthook{}%
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\set@typesize@hook}
+% \begin{macro}{\set@pica@hook}
+% The macros \cmd\set@typesize@hook\ and \cmd\set@pica@hook\ provide everything we need.
+% To use, simply \cmd\appdef\ your tokens to the appropriate hook.
+% \begin{macrocode}
+\def\set@typesize@hook{}%
+\def\set@pica@hook{}%
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\enddocument}
+% \begin{macro}{\check@aux}
+% \begin{macro}{\do@check@aux}
+% The standard \LaTeX\ \enve{document} processing is a potential problem,
+% particularly when the output routine has been changed by \classname{ltxgrid}.
+% We separate out the procedure that checks the auxiliary file at the end of
+% the job so that later it can be called from the safety of the output
+% routine.
+% We will do this to ensure that the \cmd\@mainaux\ stream is not closed until
+% the last page of the job is shipped out, and that can only be done by coordinating
+% with the output routine.
+% \begin{macrocode}
+\def\enddocument{%
+ \@enddocumenthook
+ \@checkend{document}%
+ \clear@document
+ \check@aux
+ \deadcycles\z@
+ \@@end
+}%
+\def\clear@document{\clearpage}%
+\def\check@aux{\do@check@aux}%
+\def\do@check@aux{%
+ \@if@sw\if@filesw\fi{%
+ \immediate\closeout\@mainaux
+ \let\@setckpt\@gobbletwo
+ \let\@newl@bel\@testdef
+ \@tempswafalse
+ \makeatletter
+ \input\jobname.aux\relax
+ }{}%
+ \@dofilelist
+ \@ifdim{\font@submax >\fontsubfuzz\relax}{%
+ \@font@warning{%
+ Size substitutions with differences\MessageBreak
+ up to \font@submax\space have occured.\@gobbletwo
+ }%
+ }{}%
+ \@defaultsubs
+ \@refundefined
+ \@if@sw\if@filesw\fi{%
+ \@ifx{\@multiplelabels\relax}{%
+ \@if@sw\if@tempswa\fi{%
+ \@latex@warning@no@line{%
+ Label(s) may have changed.
+ Rerun to get cross-references right
+ }%
+ }{}%
+ }{%
+ \@multiplelabels
+ }%
+ }{}%
+}%
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+%
+% \subsection{Type Tools}%
+%
+% \begin{macro}{\flushing}
+% Undoes \cmd\centering. Should also undo \cmd\raggedleft\ and \cmd\raggedright.
+% \begin{macrocode}
+\def\flushing{%
+ \let\\\@normalcr
+ \leftskip\z@skip
+ \rightskip\z@skip
+ \@rightskip\z@skip
+ \parfillskip\@flushglue
+}%
+% \end{macrocode}
+% \end{macro}
+%
+%
+% \subsection{Display Math}%
+%
+% \begin{macro}{\eqnarray@LaTeX}
+% \begin{macro}{\eqnarray@fleqn@fixed}
+% Team \LaTeX\ has stated they will never repair Leslie's broken definition of \env{eqnarray}.
+% Let us be bold\dots.
+%
+% Note on \classname{hyperref} package compatability: that package overrides
+% \cmd\eqnarray\ by wrapping it up in a larger procedure, so its changes
+% are compatible with this package's changes.
+%
+% \begin{macrocode}
+\def\eqnarray@LaTeX{%
+ \stepcounter{equation}%
+ \def\@currentlabel{\p@equation\theequation}%
+ \global\@eqnswtrue
+ \m@th
+ \global\@eqcnt\z@
+ \tabskip\@centering
+ \let\\\@eqncr
+ $$\everycr{}\halign\t@\displaywidth\bgroup
+ \hskip\@centering$\displaystyle\tabskip\z@skip{##}$\@eqnsel
+ &\global\@eqcnt\@ne\hskip \tw@\arraycolsep \hfil${##}$\hfil
+ &\global\@eqcnt\tw@ \hskip \tw@\arraycolsep
+ $\displaystyle{##}$\hfil\tabskip\@centering
+ &\global\@eqcnt\thr@@ \hb@xt@\z@\bgroup\hss##\egroup
+ \tabskip\z@skip
+ \cr
+}
+\long\def\eqnarray@fleqn@fixed{%
+ \stepcounter{equation}\def\@currentlabel{\p@equation\theequation}%
+ \global\@eqnswtrue\m@th\global\@eqcnt\z@
+ \tabskip\mathindent
+ \let\\=\@eqncr
+ \setlength\abovedisplayskip{\topsep}%
+ \ifvmode\addtolength\abovedisplayskip{\partopsep}\fi
+ \addtolength\abovedisplayskip{\parskip}%
+ \setlength\belowdisplayskip{\abovedisplayskip}%
+ \setlength\belowdisplayshortskip{\abovedisplayskip}%
+ \setlength\abovedisplayshortskip{\abovedisplayskip}%
+ $$%
+ \everycr{}%
+ \halignt@\linewidth\bgroup
+ \hskip\@centering$\displaystyle\tabskip\z@skip{##}$\@eqnsel
+ &\global\@eqcnt\@ne
+ \hskip\tw@\eqncolsep
+ \hfil${{}##{}}$\hfil
+ &\global\@eqcnt\tw@
+ \hskip\tw@\eqncolsep
+ $\displaystyle{##}$\hfil\tabskip\@centering
+ &\global\@eqcnt\thr@@\hb@xt@\z@\bgroup\hss##\egroup
+ \tabskip\z@skip
+ \cr
+}%
+\@ifx{\eqnarray\eqnarray@LaTeX}{%
+ \class@info{Repairing broken LaTeX eqnarray}%
+ \let\eqnarray\eqnarray@fleqn@fixed
+ \newlength\eqncolsep
+ \setlength\eqncolsep\z@
+ \let\eqnarray@LaTeX\relax
+ \let\eqnarray@fleqn@fixed\relax
+}{}%
+\def\mathindent{\@centering}%
+\def\set@eqnarray@skips{}%
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+%
+% \subsection{Footnotes}
+%
+% \changes{v4.0beta 4}{2000/04/10}
+% {New kernel fix. For bug 174}
+%
+% \begin{macro}{\footnote}
+% \begin{macro}{\footnotemark}
+% \begin{macro}{\@xfootnote}
+% \begin{macro}{\@xfootnotemark}
+% \begin{macro}{\@yfootnote}
+% We repair an error in the \LaTeX\ kernel (see \file{ltfloat.dtx}) involving footnotes.
+% The symptom is that the \cmd\footnotemark\ command does not work properly within a \env{minipage} environment.
+% The source of the problem is in the way the \cmd\footnotemark\ and \cmd\@xfootnotemark\ procedures are defined:
+% they do not share the method used by the \cmd\footnote\ and other procedures that allows a context switch to
+% change the way footnotes behave within a minipage environment.
+% This is a \LaTeX\ bug of long standing; this fix dates to 1987.
+%
+% While we are at it, we rewrite both the \cmd\footnote\ and \cmd\footnotemark\ procedures,
+% achieving a slightly cleaner separation of syntax and semantics.
+% Note that the \cmd\@footnotemark\ and \cmd\@footnotetext\ procedures are not altered here;
+% they continue as the methods of formatting the footnote mark and footnote text, respectively.
+%
+% A note about the context switch mentioned above:
+% the \env{minipage} environment executes the following in order to alter the way footnotes
+% behave:
+%\begin{verbatim}
+%\def\@mpfn{mpfootnote}%
+%\def\thempfn{\thempfootnote}%
+%\let\@footnotetext\@mpfootnotetext
+%\c@mpfootnote\z@
+%\end{verbatim}
+% This code changes the counter used in autonumbered footnotes, the choice of footnote marker,
+% and the procedure used on the footnote text. Changing the counter is needed because minipage
+% footnotes are in their own sequence, and the footnote marker is customarily different within
+% a minipage. The procedure that works on the footnote text must be different because the footnotes
+% are placed at the bottom of the minipage, not the bottom of the text column.
+%
+% Any procedure that establishes a minipage-like context (e.g., floats) can do the same.
+% \begin{macrocode}
+\def\footnote{%
+ \@ifnextchar[\@xfootnote{\@yfootnote\@footnotetext}%
+}%
+\def\footnotemark{%
+ \@ifnextchar[\@xfootnotemark{\@yfootnote}%
+}%
+\def\@xfootnote[#1]{%
+ \@xfootnotemark[#1]%
+ \@footnotetext
+}%
+\def\@xfootnotemark@ltx[#1]{%
+ \begingroup
+ \csname c@\@mpfn\endcsname #1\relax
+ \unrestored@protected@xdef\@thefnmark{\thempfn}%
+ \endgroup
+ \H@@footnotemark
+}%
+\def\@yfootnote{%
+ \stepcounter\@mpfn
+ \protected@xdef\@thefnmark{\thempfn}%
+ \H@@footnotemark
+}%
+% \end{macrocode}
+%
+% Note on \classname{hyperref} compatability:
+% In its ``Automated \LaTeX\ hypertext cross-references'',
+% the \classname{hyperref} package alters footnote processing,
+% thereby imperiling these fixes and necessiating defensive measures.
+%
+% The main thing \classname{hyperref} does is to take over the \cmd\@mpfootnotetext\ and
+% \cmd\@footnotetext\ procedures, feeding its own arguments to these macros.
+% It also rewrites \cmd\@footnotemark, making it a hyperlink.
+%
+% But at the same time, it attempts to turn off these changes during
+% \cmd\maketitle\ processing, necessitating rewriting \cmd\@xfootnotemark.
+% At this point it is on the slippery slope.
+%
+% We make ourself \classname{hyperref} friendly:
+% we give \classname{hyperref} what it needs, but avoid its change to
+% \cmd\@xfootnotemark.
+%
+% Any other package that rewrites \LaTeX's footnote macros will be incompatible
+% with this package.
+% \begin{macrocode}
+\appdef\class@documenthook{%
+ \@ifxundefined\H@@footnotemark{%
+ \let\H@@footnotemark\@footnotemark
+ }{}%
+ \let\@xfootnotemark\@xfootnotemark@ltx
+}%
+% \end{macrocode}
+%
+% Two thoughts about \classname{hyperref}: what for does it define \cmd\realfootnote?
+% Also: a document class that desires high hypertext capabilities might
+% well wish to reimplement \cmd\maketitle\ so that footnotes called out from there
+% are hypertext links: the \classname{hyperref} package's
+% ``Automated \LaTeX\ hypertext cross-references''
+% does not do any of this:
+%\begin{quotation}
+% But the special footnotes
+% in |\maketitle| are much too hard to deal with
+% properly. Let them revert to plain behaviour.
+%\end{quotation}
+% Note that the document class, in reimplementing \cmd\maketitle, must ensure
+% that the \classname{hyperref} package does not clobber its own definition!
+%
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\@footnotetext}
+% \begin{macro}{\@mpfootnotetext}
+% \begin{macro}{\@tpfootnotetext}
+% \begin{macro}{\make@footnotetext}
+% \begin{macro}{\set@footnotewidth}
+% The two procedures \cmd\@footnotetext\ and \cmd\@mpfootnotetext\ share code.
+% We make that explicit here.
+%
+% Note that the procedure calling \cmd\make@footnotetext\ will open a group
+% with \cmd\bgroup\ which is then closed by \cmd\minipagefootnote@drop.
+%
+% Difference from \LaTeX: here we do not set \cmd\floatingpenalty\ to infinity.
+% Doing this must date back to a time when \LaTeX\ could not accomodate split insertions (footnotes).
+% I cannot think of any other reason to do have done this.
+% At any rate, with the \classname{ltxgrid} package, split insertions are specifically properly
+% taken care of, so we allow it.
+%
+% We provide the hook \cmd\set@footnotewidth\
+% that sets the footnote on a particular measure.
+% Some page grids are such as to set a footnote in a context where \cmd\columnwidth
+% is not the right parameter to use for the set width of a footnote.
+% In such a case, for the applicable scope, you should define
+% \cmd\set@footnotewidth\ to perform this job correctly.
+%
+% A procedure, \cmd\set@footnotewidth@ii, illustrates how to do this when in a two-column page grid.
+% In general, remember that footnotes, like all insertions (including floats),
+% are a step outside of the galley context, and all aspects of insertions need to be
+% properly handled, including the set width.
+% \begin{macrocode}
+\long\def\@footnotetext{%
+ \insert\footins\bgroup
+ \make@footnotetext
+}%
+\long\def\@mpfootnotetext{%
+ \minipagefootnote@pick
+ \make@footnotetext
+}%
+\def\make@footnotetext#1{%
+ \reset@font\footnotesize
+ \interlinepenalty\interfootnotelinepenalty
+ \splittopskip\footnotesep
+ \splitmaxdepth\dp\strutbox
+% \floatingpenalty\@MM
+ \set@footnotewidth
+ \@parboxrestore
+ \protected@edef\@currentlabel{%
+ \csname p@footnote\endcsname\@thefnmark
+ }%
+ \color@begingroup
+ \@makefntext{%
+ \rule\z@\footnotesep\ignorespaces#1\@finalstrut\strutbox
+ }%
+ \color@endgroup
+ \minipagefootnote@drop
+}%
+\def\set@footnotewidth{%
+ \hsize\columnwidth
+ \linewidth\hsize
+}%
+\def\set@footnotewidth@ii{%
+ \hsize\textwidth
+ \advance\hsize\columnsep
+ \divide\hsize\tw@
+ \advance\hsize-\columnsep
+ \linewidth\hsize
+}%
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+%
+% \subsection{Floats}
+%
+% \subsubsection{Usage notes}%
+% We extend the \LaTeX\ kernel for three purposes:
+% \begin{enumerate}
+%
+% \item
+% When the \cmd\footnote\ command is used within the
+% scope of a float, we do as \env{minipage} does.
+%
+% \item
+% We provide a mechanism to write floats out to an external
+% stream for temporary storage (deferred floats).
+%
+% \item
+% We provide mechanism for placing a float \texttt{here}
+% invariably, that is, floats are unfloated.
+% This mechanism is used to read the external stream mentioned above.
+%
+% \end{enumerate}
+%
+% To use these mechanisms, the document class should
+% define a float, say, \env{figure} as per usual, and in addition:
+% \begin{enumerate}
+%
+% \item
+% Optionally define an alternative, say \env{figure@write} as follows:
+% \begin{verbatim}
+%\newenvironment{figure@write}{%
+% \write@float{figure}%
+%}{%
+% \endwrite@float
+%}%
+% \end{verbatim}
+% That is, the alternative environment executes \cmd\write@float\
+% instead of \cmd\@float.
+% Note that this step is not needed if the float environment
+% is defined in the simple way of \file{classes.dtx}.
+% However, an environment like \env{longtable} will require it.
+%
+% \item
+% Install into \cmd\AtBeginDocument\ a call to \cmd\do@if@floats,
+% with the float name and an appropriate file extension as its arguments.
+% \begin{verbatim}
+%\AtBeginDocument{\do@if@floats{figure}{.fgx}}%
+% \end{verbatim}
+%
+% \item
+% Optionally define a text entity \cmd\figuresname\ that will
+% be the text of the head that is set over the
+% deferred floats.
+% If not defined, there will be no head.
+%
+% \item
+% Optionally define a user-level command to allow
+% the document to determine where the figures are printed out
+% (default is to print at end of document). E.g.,
+% \begin{verbatim}
+%\newcommand\printfigures{\print@float{figure}}%
+% \end{verbatim}
+% \item
+% Install into \cmd\appdef\cmd\class@enddocumenthook\ a call to \cmd\printfigures,
+% or, if the latter is not defined, as follows:
+% \begin{verbatim}
+%\appdef\class@enddocumenthook{\print@float{figure}}%
+% \end{verbatim}
+% Note that installing this command into \cmd\AtBeginDocument
+% is best done earlier than calls that assume the last page of
+% the document is at hand.
+%
+% \end{enumerate}
+%
+% \subsubsection{Robustifying fragile commands}%
+% Certain of \LaTeX's commands cannot be written out to a file or appear within a \cmd\mark\ command argument
+% because they do calculations during expansion.
+% We provide for a little help, but without changing the meanings of these commands.
+%
+% \begin{macro}{\addtocontents}
+%
+% \begin{macrocode}
+\def\addtocontents#1#2{%
+ \protected@write\@auxout{%
+ \let \label \@gobble \let \index \@gobble \let \glossary \@gobble
+ \def\({\string\(}%
+ \def\){\string\)}%
+ \def\\{\string\\}%
+ }{\string \@writefile {#1}{#2}}%
+}%
+% \end{macrocode}
+%
+% \end{macro}
+%
+% \subsubsection{Preparing for the \classname{hyperref package}}%
+%
+% \begin{macro}{\addcontentsline}
+% \begin{macro}{\contentsline}
+% The \classname{hyperref} package assumes that the \cmd\contentsline\ command will be given four arguments.
+% Therefore it cannot successfully process a \filename{.toc} file that had been written by standard \LaTeX.
+% We fix things up by always writing that fourth argument and by supplying a \cmd\contentsline\ command that
+% can read them.
+%
+% We also give the \cmd\newlabel\ command's second argument five tokens.
+%
+% This means that a document class that uses this package will itself have trouble taking over a \filename{.toc} file that was written by standard \LaTeX. Sigh.
+% \begin{macrocode}
+\def\addcontentsline#1#2#3{%
+ \addtocontents{#1}{%
+ \protect\contentsline{#2}{#3}{\thepage}{}%
+ }%
+}%
+\def\contentsline#1#2#3#4{%
+ \csname l@#1\endcsname{#2}{#3}%
+}%
+\def\label#1{%
+ \@bsphack
+ \protected@write\@auxout{}{%
+ \string\newlabel{#1}{{\@currentlabel}{\thepage}{}{}{}}%
+ }%
+ \@esphack
+}%
+% \end{macrocode}
+%
+% \end{macro}
+% \end{macro}
+%
+% \subsubsection{Footnotes within floats, unfloating floats, float font}%
+%
+% \begin{macro}{\caption}
+% DPC: Er a bit of a hack, but seems best way of supporting normal
+% \LaTeX\ syntax at this point: If a caption is used below a table,
+% then put out the footnotes before the caption.
+% \changes{v4.0beta 2}{1999/06/20}
+% {Support the hack with \cs{prepdef}, and delay until \cs{AtBeginDocument} time,
+% since \classname{hyperref} clobbers \cs{caption}.}
+% \begin{macrocode}
+\appdef\class@documenthook{%
+ \prepdef\caption{\minipagefootnote@here}%
+}%
+% \end{macrocode}
+%
+% Note on \classname{hyperref} compatability:
+% this change to the \cmd\caption\ command is compatible with the
+% ``Automated \LaTeX\ hypertext cross-references'' patches of
+% that package.
+%
+% All the same, I think Sebastian's changes to \cmd\caption\ and \cmd\@caption\
+% could bear with some improvement.
+% The following implementation requires knowing only the pattern part of the
+% \cmd\@caption\ macro:
+%\begin{verbatim}
+%\def\caption{%
+% \H@refstepcounter\@captype
+% \hyper@makecurrent{\@captype}%
+% \@dblarg{\H@caption\@captype}%
+%}%
+%\def\H@caption#1[#2]#3{%
+% \@caption{#1}[#2]{%
+% \ifHy@nesting
+% \hyper@@anchor{\@currentHref}{#3}%
+% \else
+% \hyper@@anchor{\@currentHref}{\relax}#3%
+% \fi
+% }%
+%}%
+%
+%\end{verbatim}
+%
+% \end{macro}
+%
+% \begin{macro}{\minipagefootnote@init}
+% \begin{macro}{\minipagefootnote@here}
+% \begin{macro}{\minipagefootnote@foot}
+% \begin{macro}{\minipagefootnote@pick}
+% \begin{macro}{\minipagefootnote@drop}
+% Procedure to deal with footnotes accumulated within a minipage environment.
+% These procedures encapsulate all uses of the \cmd\@mpfootins\ box.
+%
+%
+% Note: \cmd\minipagefootnote@here\ must \emph{not} be executed within the MVL!
+% \begin{macrocode}
+\def\minipagefootnote@init{%
+ \setbox\@mpfootins\box\voidb@x
+}%
+\def\minipagefootnote@pick{%
+ \global\setbox\@mpfootins\vbox\bgroup
+ \unvbox\@mpfootins
+}%
+\def\minipagefootnote@drop{%
+ \egroup
+}%
+\def\minipagefootnote@here{%
+ \par
+ \@ifvoid\@mpfootins{}{%
+ \vskip\skip\@mpfootins
+ \fullinterlineskip
+ \@ifinner{%
+ \vtop{\unvcopy\@mpfootins}%
+ {\setbox\z@\lastbox}%
+ }{}%
+ \unvbox\@mpfootins
+ }%
+}%
+\def\minipagefootnote@foot{%
+ \@ifvoid\@mpfootins{}{%
+ \insert\footins\bgroup\unvbox\@mpfootins\egroup
+ }%
+}%
+\def\endminipage{%
+ \par
+ \unskip
+ \minipagefootnote@here
+ \@minipagefalse %% added 24 May 89
+ \color@endgroup
+ \egroup
+ \expandafter\@iiiparbox\@mpargs{\unvbox\@tempboxa}%
+}%
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\floats@sw}
+% The Boolean \cmd\floats@sw\ signifies that floats are to be floated;
+% if false, that floats are to be deferred to the end of the document.
+% Note that the state of this Boolean is to be changed by
+% the document class in response to user-selected options.
+% Here we display model code that assigns a default value
+% at \cmd\AtBeginDocument\ time.
+% \begin{verbatim}
+%\AtBeginDocument{%
+% \@ifxundefined\floats@sw{\@booleantrue\floats@sw}{}%
+%}%
+% \end{verbatim}
+% \end{macro}
+%
+% \begin{macro}{\@xfloat}
+% \begin{macro}{\@mpmakefntext}
+% The float start-code is redefined to set up footnotes in the style of minipage.
+% Also, the \cmd\floats@sw\ Boolean informs us that floats are to be
+% all placed \texttt{here}.
+% Note that, to protect against the Boolean being undefined
+% at this late hour, we default it globally to true.
+%
+% \changes{v4.0beta 2}{1999/06/20}
+% {AO: Removed superfluous \cs{def}s, changed to using \cs{floats@sw} as the flag.
+% Also stopped using DPC's \cs{if@twocolumn} flag: using \cs{floats@sw} instead.
+% Also added \cs{par}\cs{vskip}\cs{z@skip} after the \cs{minipagefootnotes}
+% so that the float box would have zero depth like the kernel one.
+% }
+% \changes{v4.0beta 3}{1999/11/13}
+% {bug fix 110. Install hooks for endfloats processing}
+% \changes{v4.0beta 4}{2000/04/10}
+% {bug fix 127. Floats placed [h] to allow page breaks}
+% \changes{v4.0beta 4}{2000/05/19}
+% {bug fix 224. Hyperref compatability.}
+% \changes{v4.0beta 5}{2000/11/16}
+% {bug fix 221. Remove samepage command from @xfloat@prep: If the float can break over pages, we want better control.}
+%FIXME: why does hyperref override \cmd\@xfootnotenext?
+%
+% \begin{macrocode}
+\let\@xfloat@LaTeX\@xfloat
+\def\@xfloat#1[#2]{%
+ \@xfloat@prep
+ \@nameuse{fp@proc@#2}%
+ \@ifxundefined\floats@sw{\global\@booleantrue\floats@sw}{}%
+ \floats@sw{\@xfloat@LaTeX{#1}[#2]}{\@xfloat@anchored{#1}[]}%
+}%
+\def\@xfloat@prep{%
+ \let\footnote\footnote@latex
+ \def\@mpfn{mpfootnote}%
+ \def\thempfn{\thempfootnote}%
+% \def\thefootnote{\thempfootnote}%
+ \c@mpfootnote\z@
+ \let\@footnotetext\@mpfootnotetext
+ \let\H@@footnotetext\@mpfootnotetext
+ \let\@makefntext\@mpmakefntext
+% \samepage
+}%
+\appdef\class@documenthook{%
+ \let\footnote@latex\footnote
+}%
+%\def\fp@proc@h{\@booleanfalse\floats@sw}%
+%\def\fp@proc@H{\@booleanfalse\floats@sw}%
+\def\@xfloat@anchored#1[#2]{%
+ \def\@captype{#1}%
+ \begin@float@pagebreak
+%\vbox\bgroup
+ \let\end@float\end@float@anchored
+ \let\end@dblfloat\end@float@anchored
+% do as \@xfloat does:
+ \hsize\columnwidth
+ \@parboxrestore
+ \@floatboxreset
+ \minipagefootnote@init
+% \pagegrid@col\@ne % Klootch to avoid processing as a float
+}%
+\def\end@float@anchored{%
+ \minipagefootnote@here
+ \par\vskip\z@skip %% \par\vskip\z@ added 15 Dec 87
+%\egroup
+ \par
+ \end@float@pagebreak
+}%
+\def\begin@float@pagebreak{\par\addvspace\intextsep}%
+\def\end@float@pagebreak{\par\addvspace\intextsep}%
+\def\@mpmakefntext#1{%
+ \parindent=1em
+ \noindent
+ \hb@xt@1em{\hss\@makefnmark}%
+ #1%
+}%
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+%
+% \subsubsection{Writing floats out to a file}%
+%
+% \begin{macro}{\do@if@floats}
+% The procedure \cmd\do@if@floats\ should be executed at
+% \cmd\AtBeginDocument\ time, and arranges to write out
+% the floats of the given class to a temporary file, to be
+% read back later (deferred floats),
+% given that \cmd\floats@sw\ is false.
+% Note that, to protect against the Boolean being undefined
+% at this late hour, we default it globally to true.
+% \begin{macrocode}
+\def\do@if@floats#1#2{%
+ \@ifxundefined\floats@sw{\global\@booleantrue\floats@sw}{}%
+ \floats@sw{}{%
+% \end{macrocode}
+% Open the stream to save out the document's floats of this class.
+% \begin{macrocode}
+ \expandafter\newwrite
+ \csname#1write\endcsname
+ \expandafter\def
+ \csname#1@stream\endcsname{\jobname#2}%
+ \expandafter\immediate
+ \expandafter\openout
+ \csname#1write\endcsname
+ \csname#1@stream\endcsname\relax
+% \end{macrocode}
+% Swap environments.
+% If the class writer has defined, e.g., \env{figure@write},
+% then we use this as the procedure to execute for writing
+% the float out to the external stream.
+% Otherwise, the replacement of \cmd\@float\ by \cmd\write@float\
+% should do the right thing for float environments defined
+% in the simple way of \classname{classes.dtx}.
+% \begin{macrocode}
+ \@ifxundefined\@float@LaTeX{%
+ \let\@float@LaTeX\@float
+ \let\@dblfloat@LaTeX\@dblfloat
+ \let\@float\write@float
+ \let\@dblfloat\write@floats
+ }{}%
+ \let@environment{#1@float}{#1}%
+ \let@environment{#1@floats}{#1*}%
+ \@ifxundefined@cs{#1@write}{}{%
+ \let@environment{#1}{#1@write}%
+ }%
+ }%
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\print@float}
+% The procedure \cmd\print@float\ prints out the
+% deferred floats.
+% \changes{v4.0beta 2}{1999/06/20}
+% {only execute if there really were floats of the given type}
+% \changes{v4.0beta 3}{1999/11/13}
+% {*-form mandates pagebreak at each float;
+% only print section head if there is something there.}
+% \changes{v4.0beta 4}{2000/05/23}
+% {Allow things to break over pages by setting array@default.}
+%
+% \begin{macrocode}
+\def\triggerpar{\leavevmode\@@par}%
+\def\oneapage{\def\begin@float@pagebreak{\newpage}\def\end@float@pagebreak{\newpage}}%
+\def\print@float#1#2{%
+ \@ifxundefined@cs{#1write}{}{%
+ \begingroup
+ \@booleanfalse\floats@sw
+ #2%
+ \raggedbottom
+ \def\array@default{v}% floats must
+ \let\@float\@float@LaTeX
+ \let\@dblfloat\@dblfloat@LaTeX
+ \let\trigger@float@par\triggerpar
+ \let@environment{#1}{#1@float}%
+ \let@environment{#1*}{#1@floats}%
+ \expandafter\prepdef\csname#1\endcsname{\trigger@float@par}%
+ \expandafter\prepdef\csname#1*\endcsname{\trigger@float@par}%
+ \@namedef{fps@#1}{h!}%
+ \expandafter\immediate
+ \expandafter\closeout
+ \csname#1write\endcsname
+ \everypar{%
+ \global\let\trigger@float@par\relax
+ \global\everypar{}\setbox\z@\lastbox
+ \@ifxundefined@cs{#1sname}{}{%
+ \begin@float@pagebreak
+ \expandafter\section
+ \expandafter*%
+ \expandafter{%
+ \csname#1sname\endcsname
+ }%
+ }%
+ }%
+ \input{\csname#1@stream\endcsname}%
+ \endgroup
+ \global\expandafter\let\csname#1write\endcsname\relax
+ }%
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\write@float}
+% \begin{macro}{\write@floats}
+% \begin{macro}{\write@@float}
+% Handles the case where the name of the float is the same as
+% that of the stream. Note that
+% \env{longtable} does \emph{not} fit this case.
+% Note also: \cmd\write@float\ is \emph{not} a user-level environment
+% therefore it is properly not defined with \cmd\newenvironment.
+% \begin{macrocode}
+\def\write@float#1{\write@@float{#1}{#1}}%
+\def\endwrite@float{\@Esphack}%
+\def\write@floats#1{\write@@float{#1*}{#1}}%
+\def\endwrite@floats{\@Esphack}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\write@@float}
+% \changes{v4.0beta 2}{1999/06/20}
+% {AO: Fixed spurious \texttt{CR} and (return) characters in output file.
+% Also, if the document did not have the \cs{end}\texttt{figure} on a line of its own,
+% the macro wouldn't work. Fixed.}
+% \begin{macrocode}
+\def\write@@float#1#2{%
+ \ifhmode
+ \@bsphack
+ \fi
+ \chardef\@tempc\csname#2write\endcsname
+ \toks@{\begin{#1}}%
+ \def\@tempb{#1}%
+ \expandafter\let\csname end#1\endcsname\endwrite@float
+ \catcode`\^^M\active
+ \@makeother\{\@makeother\}\@makeother\%
+ \write@floatline
+}%
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\write@floatline}
+% \begin{macro}{\@write@floatline}
+% \begin{macro}{\float@end@tag}
+% The procedure \cmd\write@floatline\ only parses, and passes
+% its result to \cmd\@write@floatline, which
+% writes the line to output, then tests the line
+% for the \cmd\end\arg{float} tokens with
+% aid of the \cmd\float@end@tag\ procedure.
+% \begin{macrocode}
+\begingroup
+ \catcode`\[\the\catcode`\{\catcode`\]\the\catcode`\}\@makeother\{\@makeother\}%
+ \gdef\float@end@tag#1\end{#2}#3\@nul[%
+ \def\@tempa[#2]%
+ \@ifx[\@tempa\@tempb][\end[#2]][\write@floatline]%
+ ]%
+ \obeylines%
+ \gdef\write@floatline#1^^M[%
+ \begingroup%
+ \newlinechar`\^^M%
+ \toks@\expandafter[\the\toks@#1]\immediate\write\@tempc[\the\toks@]%
+ \endgroup%
+ \toks@[]%
+ \float@end@tag#1\end{}\@nul%
+ ]%
+\endgroup
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+%
+%
+% \subsection{Counters}
+% The following definitions override those of the \LaTeX\ kernel,
+% providing for a greater range of inputs.
+% \begin{macrocode}
+\def\@alph#1{\ifcase#1\or a\or b\or c\or d\else\@ialph{#1}\fi}
+% \end{macrocode}
+%
+% \begin{macrocode}
+\def\@ialph#1{\ifcase#1\or \or \or \or \or e\or f\or g\or h\or i\or j\or
+ k\or l\or m\or n\or o\or p\or q\or r\or s\or t\or u\or v\or w\or x\or
+ y\or z\or aa\or bb\or cc\or dd\or ee\or ff\or gg\or hh\or ii\or jj\or
+ kk\or ll\or mm\or nn\or oo\or pp\or qq\or rr\or ss\or tt\or uu\or
+ vv\or ww\or xx\or yy\or zz\else\@ctrerr\fi}
+% \end{macrocode}
+%
+%
+% \subsection{Customization of Sections}%
+%
+% Patch the standard \LaTeX\ sectioning procedure to:
+%\begin{itemize}
+%\item
+% Allow a sectioning command to trigger the title page, or more generally
+% to recognize that it is the first object in the document,
+% so we headpatch \cmd\@startsection.
+%
+%\item
+% Allow a tail command in |#6| to uppercase the title, so we retain
+% DPC's braces.
+%
+%\item
+% Allow each type of sectioning command to format its number differently,
+% so we generalize \cmd\@seccntformat.
+%
+%\item
+% Allow each type of sectioning command to format its argument differently,
+% so we generalize \cmd\@hangfrom.
+%
+%\item
+% Allow the starred form of the command to
+% mark (the running head) and
+% make an entry in the TOC,
+% so we put \cmd\@ssect\ on the same footing as \cmd\@sect.
+%
+% Note that the tokens passed to the TOC now are \emph{not}
+% the optional argument of the command, but the required.
+% This means that the user can no longer use the former
+% to put variant content in to the TOC as the Manual says.
+%
+% Instead, the optional argument is used to put an alternative
+% title into the running headers, a better choice.
+%
+%\end{itemize}
+%
+% \begin{macro}{\@startsection}
+% Patch a head hook into the basic sectioning command.
+% Treat \cmd\@sect\ and \cmd\@ssect\ on an equal footing:
+% now their pattern parts are identical.
+% \begin{macrocode}
+\def\@startsection#1#2#3#4#5#6{%
+ \@startsection@hook
+ \if@noskipsec \leavevmode \fi
+ \par
+ \@tempskipa #4\relax
+ \@afterindenttrue
+ \ifdim \@tempskipa <\z@
+ \@tempskipa -\@tempskipa \@afterindentfalse
+ \fi
+ \if@nobreak
+ \everypar{}%
+ \else
+ \addpenalty\@secpenalty\addvspace\@tempskipa
+ \fi
+ \@ifstar
+ {\@dblarg{\@ssect@ltx{#1}{#2}{#3}{#4}{#5}{#6}}}%
+ {\@dblarg{\@sect@ltx {#1}{#2}{#3}{#4}{#5}{#6}}}%
+}%
+\def\@startsection@hook{}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\@sect}
+% When defining \cmd\@svsec, do not expand \cmd\@seccntformat.
+% Put brace characters back where they were before David Carlisle got at them
+% (i.e., as if \cmd\@hangfrom\ had two arguments).
+% Protect the mark mechanism from an undefined meaning.
+% Pass |#8| to the TOC instead of |#7|.
+% Remove \cmd\relax\ from the replacement part of \cmd\@svsec.
+%
+% The procedure \cmd\@hangfrom\ and \cmd\@runin@to\ can
+% be used to process the argument of the head.
+% The head can define, e.g., \cmd\@hangfrom@section, to
+% do its own processing.
+%
+% In using \cmd\H@refstepcounter\ in place of \cmd\refstepcounter\ we rely on
+% either loading before any package that patches the latter, or
+% the convention that the former is the original \LaTeX\ procedure.
+%
+% \begin{macrocode}
+\class@info
+ {Repairing broken LateX \string\@sect}%
+\def\@sect@ltx#1#2#3#4#5#6[#7]#8{%
+ \@ifnum{#2>\c@secnumdepth}{%
+ \def\H@svsec{\phantomsection}%
+ \let\@svsec\@empty
+ }{%
+ \H@refstepcounter{#1}%
+ \def\H@svsec{%
+ \phantomsection
+ }%
+ \protected@edef\@svsec{{#1}}%
+ \@ifundefined{@#1cntformat}{%
+ \prepdef\@svsec\@seccntformat
+ }{%
+ \expandafter\prepdef
+ \expandafter\@svsec
+ \csname @#1cntformat\endcsname
+ }%
+ }%
+ \@tempskipa #5\relax
+ \@ifdim{\@tempskipa>\z@}{%
+ \begingroup
+ \interlinepenalty \@M
+ #6{%
+ \@ifundefined{@hangfrom@#1}{\@hang@from}{\csname @hangfrom@#1\endcsname}%
+ {\hskip#3\relax\H@svsec}{\@svsec}{#8}%
+ }%
+ \@@par
+ \endgroup
+ \@ifundefined{#1mark}{\@gobble}{\csname #1mark\endcsname}{#7}%
+ \addcontentsline{toc}{#1}{%
+ \@ifnum{#2>\c@secnumdepth}{%
+ \protect\numberline{}%
+ }{%
+ \protect\numberline{\csname the#1\endcsname}%
+ }%
+ #8}%
+ }{%
+ \def\@svsechd{%
+ #6{%
+ \@ifundefined{@runin@to@#1}{\@runin@to}{\csname @runin@to@#1\endcsname}%
+ {\hskip#3\relax\H@svsec}{\@svsec}{#8}%
+ }%
+ \@ifundefined{#1mark}{\@gobble}{\csname #1mark\endcsname}{#7}%
+ \addcontentsline{toc}{#1}{%
+ \@ifnum{#2>\c@secnumdepth}{%
+ \protect\numberline{}%
+ }{%
+ \protect\numberline{\csname the#1\endcsname}%
+ }%
+ #8}%
+ }%
+ }%
+ \@xsect{#5}%
+}%
+\def\@hang@from#1#2#3{\@hangfrom{#1#2}#3}%
+\def\@runin@to #1#2#3{#1#2#3}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\@ssect}
+% Put brace characters back where they were before David Carlisle got at them
+% (as if \cmd\@hangfrom\ has two arguments).
+% Possibly set a mark.
+% Make a TOC entry.
+%
+% Note that, for compatability with the \classname{hyperref} package, we
+% need to provide the interface required by that package
+% (actually required by \file{pdfmark.def} and \file{nameref.sty}),
+% namely
+% the definition of \cmd\@currentlabelname\ (but now removed),
+% the insertion of the procedure \cmd\Sectionformat\ (but why is this needed?), and
+% the call to \cmd\phantomsection\ (which must precede the call to \cmd\addcontentsline).
+% We also have to sidestep the patch to \cmd\@ssect\ in that same file, therefore
+% we use a different control sequence name in the call from \cmd\@startsection.
+% \changes{v4.0beta 3}{1999/11/13}
+% {Bug 116: Hyperref compatability}
+% \changes{v4.0rc3b}{2001/07/13}
+% {Bug 404: Hyperref compatability}
+% \begin{macrocode}
+\def\@ssect@ltx#1#2#3#4#5#6[#7]#8{%
+% \def\@currentlabelname{#8}%
+ \def\H@svsec{\phantomsection}%
+ \@tempskipa #5\relax
+ \@ifdim{\@tempskipa>\z@}{%
+ \begingroup
+ \interlinepenalty \@M
+ #6{%
+ \@ifundefined{@hangfroms@#1}{\@hang@froms}{\csname @hangfroms@#1\endcsname}%
+% {\hskip#3\relax\H@svsec}{\Sectionformat{#8}{#1}}%
+ {\hskip#3\relax\H@svsec}{#8}%
+ }%
+ \@@par
+ \endgroup
+ \@ifundefined{#1smark}{\@gobble}{\csname #1smark\endcsname}{#7}%
+ \addcontentsline{toc}{#1}{\protect\numberline{}#8}%
+ }{%
+ \def\@svsechd{%
+ #6{%
+ \@ifundefined{@runin@tos@#1}{\@runin@tos}{\csname @runin@tos@#1\endcsname}%
+% {\hskip#3\relax\H@svsec}{\Sectionformat{#8}{#1}}%
+ {\hskip#3\relax\H@svsec}{#8}%
+ }%
+ \@ifundefined{#1smark}{\@gobble}{\csname #1smark\endcsname}{#7}%
+ \addcontentsline{toc}{#1}{\protect\numberline{}#8}%
+ }%
+ }%
+ \@xsect{#5}%
+}%
+\def\@hang@froms#1#2{#1#2}%
+\def\@runin@tos #1#2{#1#2}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\init@documenthook}
+% Document classes that incorporate this package will be \classname{hyperref}-savvy.
+% (To accomplish this, we ensure that \cmd\hyperanchor\ and \cmd\hyper@last\ are both defined.)
+% Being \classname{hyperref}-savvy levels some requirements on us, but the benefits are many.
+%
+% One is that the TOC will not get amnesia and require a full set of three typesetting runs before its formatting is stable.
+% Instead, only two runs are required: the first updates the auxiliary file, the second the TOC.
+% However, the formatting of the document does not change.
+%
+% Another aspect of being \classname{hyperref}-savvy is that the syntax of commands in the \filename{.aux} file will now change
+% if \classname{hyperref} is turned on or off.
+%
+% Note that \cmd\hyper@anchorstart\ and \cmd\hyper@anchorend\ constitute the programming interface
+% for a hypertext anchor (the target of a hypertext link); \cmd\hyper@linkstart\ and \cmd\hyper@linkend\
+% are the interface for a hypertext link.
+% \begin{macrocode}
+\appdef\init@documenthook{%
+ \providecommand\phantomsection{}%
+%\@ifx{\Sectionformat\@undefined}{\let\Sectionformat\@firstoftwo}{}%
+ \providecommand\hyper@anchor[1]{}%
+ \providecommand\hyper@last{}%
+ \providecommand\Hy@raisedlink[1]{#1}%
+ \providecommand\hyper@anchorstart[1]{}%
+ \providecommand\hyper@anchorend{}%
+ \providecommand\hyper@linkstart[2]{}%
+ \providecommand\hyper@linkend{}%
+}%
+\let\H@refstepcounter\refstepcounter
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\sec@upcase}
+% Upper case for sections (optional upper case items). These are
+% created so that some headings can be toggled between mixed case and
+% upper case readily.
+% Headings that might be changed can be wrapped in the style file in
+% \cmd\sec@upcase\arg{text} constructs;
+% the expansion of \cmd\sec@upcase\ is
+% controlled here. It is \cmd\relax\ by default (mixed case heads), and
+% can easily be changed to \cmd\uppercase\ if desired.
+% If mixed-case headings are wanted by the editor, authors {\em must}
+% supply mixed case text, although this is what authors should be doing
+% anyway.
+% (Mixed can be converted to upper,
+% but the reverse transformation cannot be automated.)
+%
+% The following setting gives the \LaTeX\ default.
+% \begin{macrocode}
+\def\sec@upcase#1{\relax{#1}}%
+% \end{macrocode}
+% \end{macro}
+%
+%
+% \subsection{Patch the \env{tabular} and \env{array} Environments}
+%
+% \begin{macro}{\endtabular}
+% \begin{macro}{\endarray}
+% We headpatch the begin processing and tailpatch the end processing
+% of the \env{tabular} and \env{array} environments.
+% A document class can define these hooks as needed.
+%
+% We proceed with care to make further patches to
+% support tabulars that break over pages.
+% Our patches will not necessarily be effective for
+% other packages that replace the \LaTeX\ \env{array} and \env{tabular}
+% environments. I know of none that do so.
+% \begin{macrocode}
+\appdef\class@documenthook{%
+ \@ifpackageloaded{array}{\switch@array}{\switch@tabular}%
+ \prepdef\endtabular{\endtabular@hook}%
+ \@provide\endtabular@hook{}%
+ \prepdef\endarray{\endarray@hook}%
+ \@provide\endarray@hook{}%
+ \providecommand\array@hook{}%
+% \end{macrocode}
+% Install, effectively, a head patch to \cmd\tabular.
+% In order to avoid interference from, e.g., the \classname{array} package,
+% we must perform this patch only \emph{after} packages load.
+% \changes{v4.0beta 3}{1999/11/13}
+% {Bug 130. Interference from array package}
+% \begin{macrocode}
+ \prepdef\@tabular{\tabular@hook}%
+ \@provide\tabular@hook{}%
+}%
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\switch@tabular}
+% \begin{macro}{\switch@array}
+% The two procedures \cmd\switch@tabular\ and \cmd\switch@array\
+% apply needed patches to the various tabular procedures,
+% the former applying to the \LaTeX\ kernel, the latter to the
+% required \classname{array} package (and to the number of other
+% required packages that load it).
+%
+% \begin{macrocode}
+\def\switch@tabular{%
+ \let\@array@sw\@array@sw@array
+ \@ifx{\@array\@array@LaTeX}{%
+ \@ifx{\multicolumn\multicolumn@LaTeX}{%
+ \@ifx{\@tabular\@tabular@LaTeX}{%
+ \@ifx{\@tabarray\@tabarray@LaTeX}{%
+ \@ifx{\array\array@LaTeX}{%
+ \@ifx{\endarray\endarray@LaTeX}{%
+ \@ifx{\endtabular\endtabular@LaTeX}{%
+ \@ifx{\@mkpream\@mkpream@LaTeX}{%
+ \@ifx{\@addamp\@addamp@LaTeX}{%
+ \@ifx{\@arrayacol\@arrayacol@LaTeX}{%
+ \@ifx{\@tabacol\@tabacol@LaTeX}{%
+ \@ifx{\@arrayclassz\@arrayclassz@LaTeX}{%
+ \@ifx{\@tabclassiv\@tabclassiv@LaTeX}{%
+ \@ifx{\@arrayclassiv\@arrayclassiv@LaTeX}{%
+ \@ifx{\@tabclassz\@tabclassz@LaTeX}{%
+ \@ifx{\@classv\@classv@LaTeX}{%
+ \@ifx{\hline\hline@LaTeX}{%
+ \@ifx{\@tabularcr\@tabularcr@LaTeX}{%
+ \@ifx{\@xtabularcr\@xtabularcr@LaTeX}{%
+ \@ifx{\@xargarraycr\@xargarraycr@LaTeX}{%
+ \@ifx{\@yargarraycr\@yargarraycr@LaTeX}{%
+ \true@sw
+ }{%
+ \false@sw
+ }%
+ }{%
+ \false@sw
+ }%
+ }{%
+ \false@sw
+ }%
+ }{%
+ \false@sw
+ }%
+ }{%
+ \false@sw
+ }%
+ }{%
+ \false@sw
+ }%
+ }{%
+ \false@sw
+ }%
+ }{%
+ \false@sw
+ }%
+ }{%
+ \false@sw
+ }%
+ }{%
+ \false@sw
+ }%
+ }{%
+ \false@sw
+ }%
+ }{%
+ \false@sw
+ }%
+ }{%
+ \false@sw
+ }%
+ }{%
+ \false@sw
+ }%
+ }{%
+ \false@sw
+ }%
+ }{%
+ \false@sw
+ }%
+ }{%
+ \false@sw
+ }%
+ }{%
+ \false@sw
+ }%
+ }{%
+ \false@sw
+ }%
+ }{%
+ \false@sw
+ }%
+ }{%
+ \false@sw
+ }%
+ {%
+ \class@info{Patching LaTeX tabular.}%
+ }{%
+ \class@info{Unrecognized LaTeX tabular. Please update this document class! (Proceeding with fingers crossed.)}%
+ }%
+ \let\@array\@array@ltx
+ \let\multicolumn\multicolumn@ltx
+ \let\@tabular\@tabular@ltx
+ \let\@tabarray\@tabarray@ltx
+ \let\array\array@ltx
+ \let\endarray\endarray@ltx
+ \let\endtabular\endtabular@ltx
+ \let\@mkpream\@mkpream@ltx
+ \let\@addamp\@addamp@ltx
+ \let\@arrayacol\@arrayacol@ltx
+ \let\@tabacol\@tabacol@ltx
+ \let\@arrayclassz\@arrayclassz@ltx
+ \let\@tabclassiv\@tabclassiv@ltx
+ \let\@arrayclassiv\@arrayclassiv@ltx
+ \let\@tabclassz\@tabclassz@ltx
+ \let\@classv\@classv@ltx
+ \let\hline\hline@ltx
+ \let\@tabularcr\@tabularcr@ltx
+ \let\@xtabularcr\@xtabularcr@ltx
+ \let\@xargarraycr\@xargarraycr@ltx
+ \let\@yargarraycr\@yargarraycr@ltx
+}%
+\def\switch@array{%
+ \let\@array@sw\@array@sw@LaTeX
+ \@ifx{\@array\@array@array}{%
+ \@ifx{\@tabular\@tabular@array}{%
+ \@ifx{\@tabarray\@tabarray@array}{%
+ \@ifx{\array\array@array}{%
+ \@ifx{\endarray\endarray@array}{%
+ \@ifx{\endtabular\endtabular@array}{%
+ \@ifx{\@mkpream\@mkpream@array}{%
+ \@ifx{\@classx\@classx@array}{%
+ \@ifx{\insert@column\insert@column@array}{%
+ \@ifx{\@arraycr\@arraycr@array}{%
+ \@ifx{\@xarraycr\@xarraycr@array}{%
+ \@ifx{\@xargarraycr\@xargarraycr@array}{%
+ \@ifx{\@yargarraycr\@yargarraycr@array}{%
+ \true@sw
+ }{%
+ \false@sw
+ }%
+ }{%
+ \false@sw
+ }%
+ }{%
+ \false@sw
+ }%
+ }{%
+ \false@sw
+ }%
+ }{%
+ \false@sw
+ }%
+ }{%
+ \false@sw
+ }%
+ }{%
+ \false@sw
+ }%
+ }{%
+ \false@sw
+ }%
+ }{%
+ \false@sw
+ }%
+ }{%
+ \false@sw
+ }%
+ }{%
+ \false@sw
+ }%
+ }{%
+ \false@sw
+ }%
+ }{%
+ \false@sw
+ }{%
+ \class@info{Patching array package.}%
+ }{%
+ \class@info{Unrecognized array package. Please update this document class! (Proceeding with fingers crossed.)}%
+ }%
+ \let\@array \@array@array@new
+ \let\@@array \@array % Cosi fan tutti
+ \let\@tabular \@tabular@array@new
+ \let\@tabarray \@tabarray@array@new
+ \let\array \array@array@new
+ \let\endarray \endarray@array@new
+ \let\endtabular\endtabular@array@new
+ \let\@mkpream \@mkpream@array@new
+ \let\@classx \@classx@array@new
+ \let\@arrayacol\@arrayacol@ltx
+ \let\@tabacol \@tabacol@ltx
+ \let\insert@column\insert@column@array@new
+ \expandafter\let\csname endtabular*\endcsname\endtabular % Cosi fan tutti
+ \let\@arraycr \@arraycr@new
+ \let\@xarraycr \@xarraycr@new
+ \let\@xargarraycr\@xargarraycr@new
+ \let\@yargarraycr\@yargarraycr@new
+}%
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\@array@sw}
+% The Boolean \cmd\@array@sw\ must be different depending on
+% whether the \classname{array} package is loaded.
+% \begin{macrocode}
+\def\@array@sw@LaTeX{\@ifx{\\\@tabularcr}}%
+\def\@array@sw@array{\@ifx{\d@llarbegin\begingroup}}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\@tabular}
+% We provide the old versions of \cmd\@tabular\ along with the respective new versions.
+% The change here is to avoid committing to LR mode. That will be done later (as late as possible, naturally).
+% \begin{macrocode}
+\def\@tabular@LaTeX{%
+ \leavevmode
+ \hbox\bgroup$%
+ \let\@acol\@tabacol
+ \let\@classz\@tabclassz
+ \let\@classiv\@tabclassiv
+ \let\\\@tabularcr
+ \@tabarray
+}%
+\def\@tabular@ltx{%
+ \let\@acoll\@tabacoll
+ \let\@acolr\@tabacolr
+ \let\@acol\@tabacol
+ \let\@classz\@tabclassz
+ \let\@classiv\@tabclassiv
+ \let\\\@tabularcr
+ \@tabarray
+}%
+\def\@tabular@array{%
+ \leavevmode
+ \hbox\bgroup$%
+ \col@sep\tabcolsep
+ \let\d@llarbegin\begingroup
+ \let\d@llarend\endgroup
+ \@tabarray
+}%
+\def\@tabular@array@new{%
+ \let\@acoll\@tabacoll
+ \let\@acolr\@tabacolr
+ \let\@acol\@tabacol
+ \let\col@sep\@undefined
+ \let\d@llarbegin\begingroup
+ \let\d@llarend\endgroup
+ \@tabarray
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\@tabarray}
+% Here we provide old and new versions of the \cmd\@tabarray\ procedure.
+% The change here is to parametrize the default vertical alignment,
+% which is 'c' in standard \LaTeX.
+% Under some circumstances, we want to change this to, say, 'v'.
+%
+% FIXME: must decouple \env{array} and \env{tabular}.
+% \begin{macrocode}
+\def\@tabarray@LaTeX{%
+ \m@th\@ifnextchar[\@array{\@array[c]}%
+}%
+\def\@tabarray@ltx{%
+ \m@th\@ifnextchar[\@array{\expandafter\@array\expandafter[\array@default]}%
+}%
+\def\@tabarray@array{%
+ \@ifnextchar[{\@@array}{\@@array[c]}%
+}%
+\def\@tabarray@array@new{%
+ \@ifnextchar[{\@@array}{\expandafter\@@array\expandafter[\array@default]}%
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\@tabularcr}
+% \begin{macro}{\@tbpen}
+% \begin{macro}{\@tabularcr}
+% \begin{macro}{\@xtabularcr}
+% \begin{macro}{\@xargarraycr}
+% \begin{macro}{\@yargarraycr}
+% \begin{macro}{\@arraycr}
+% \begin{macro}{\@xarraycr}
+% We provide for the \cmd\\ command within \env{tabular} to provide control over page breaking, just the same as
+% that of \env{eqnarray}.
+%
+% The count register \cmd\intertabularlinepenalty\ is similar to \cmd\interdisplaylinepenalty: it is the penalty
+% associated with each row of a tabular. When it is set to \cmd\@M, the tabular will cleave together.
+%
+% The count register \cmd\@tbpen\ is similar to \cmd\@eqpen: it memorizes the penalty to use after the current tabular row.
+% If the \cmd\\ command is in its star form, then \cmd\@eqpen\ is set to \cmd\@M.
+%
+% We append code to \cmd\samepage\ so that a tabular within its scope will cleave together.
+%
+% We keep the standard definition of \cmd\@tabularcr\ in \cmd\@tabularcr@LaTeX\ for reference,
+% and provide a new definition that works like \cmd\@eqncr: it sets \cmd\@tbpen\ to \cmd\@M\ if the star was given.
+%
+% We also provide new versions of \cmd\@xtabularcr, \cmd\@xargarraycr, and \cmd\@yargarraycr, all of which invoke \cmd\@tbpen.
+%
+% The \cmd\switch@tabular\ procedure switches in the new definitions.
+% \begin{macrocode}
+\newcount\intertabularlinepenalty
+\intertabularlinepenalty=100
+\newcount\@tbpen
+\appdef\samepage{\intertabularlinepenalty\@M}%
+\def\@tabularcr@LaTeX{{\ifnum 0=`}\fi \@ifstar \@xtabularcr \@xtabularcr}%
+\def\@tabularcr@ltx{{\ifnum 0=`}\fi \@ifstar {\global \@tbpen \@M \@xtabularcr }{\global \@tbpen \intertabularlinepenalty \@xtabularcr }}%
+\def\@xtabularcr@LaTeX{\@ifnextchar [\@argtabularcr {\ifnum 0=`{\fi }\cr }}%
+\def\@xtabularcr@ltx{\@ifnextchar [\@argtabularcr {\ifnum 0=`{\fi }\cr \noalign {\penalty \@tbpen }}}%
+\def\@xargarraycr@LaTeX#1{\@tempdima #1\advance \@tempdima \dp \@arstrutbox \vrule \@height \z@ \@depth \@tempdima \@width \z@ \cr}%
+\def\@xargarraycr@ltx#1{\@tempdima #1\advance \@tempdima \dp \@arstrutbox \vrule \@height \z@ \@depth \@tempdima \@width \z@ \cr \noalign {\penalty \@tbpen }}%
+\def\@yargarraycr@LaTeX#1{\cr \noalign {\vskip #1}}%
+\def\@yargarraycr@ltx#1{\cr \noalign {\penalty \@tbpen \vskip #1}}%
+% \end{macrocode}
+%
+% If the \classname{array} package has been loaded, we must alter the meanings of
+% \cmd\@arraycr, \cmd\@xarraycr, \cmd\@xargarraycr, and \cmd\@yargarraycr.
+% In this case, it is \cmd\switch@array\ that switches in the new definitions.
+% \begin{macrocode}
+\def\@arraycr@array{%
+ \relax
+ \iffalse{\fi\ifnum 0=`}\fi
+ \@ifstar \@xarraycr \@xarraycr
+}%
+\def\@arraycr@new{%
+ \relax
+ \iffalse{\fi\ifnum 0=`}\fi
+ \@ifstar {\global \@tbpen \@M \@xarraycr }{\global \@tbpen \intertabularlinepenalty \@xarraycr }%
+}%
+\def\@xarraycr@array{%
+ \@ifnextchar [%]
+ \@argarraycr {\ifnum 0=`{}\fi\cr}%
+}%
+\def\@xarraycr@new{%
+ \@ifnextchar [%]
+ \@argarraycr {\ifnum 0=`{}\fi\cr \noalign {\penalty \@tbpen }}%
+}%
+\def\@xargarraycr@array#1{%
+ \unskip
+ \@tempdima #1\advance\@tempdima \dp\@arstrutbox
+ \vrule \@depth\@tempdima \@width\z@
+ \cr
+}%
+\def\@xargarraycr@new#1{%
+ \unskip
+ \@tempdima #1\advance\@tempdima \dp\@arstrutbox
+ \vrule \@depth\@tempdima \@width\z@
+ \cr
+ \noalign {\penalty \@tbpen }%
+}%
+\def\@yargarraycr@array#1{%
+ \cr
+ \noalign{\vskip #1}%
+}%
+\def\@yargarraycr@new#1{%
+ \cr
+ \noalign{\penalty \@tbpen \vskip #1}%
+}%
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\array}
+% We provide old and new versions of the \cmd\array\ procedure for both \LaTeX\ and the \classname{array} package.
+% The change here is to accomodate the new procedures that will be called for the array boundaries, even
+% though at present they are not special.
+% A thought: here is where matrices can be readily accomodated.
+% \begin{macrocode}
+\def\array@LaTeX{%
+ \let\@acol\@arrayacol
+ \let\@classz\@arrayclassz
+ \let\@classiv\@arrayclassiv
+ \let\\\@arraycr
+ \let\@halignto\@empty
+ \@tabarray
+}%
+\def\array@ltx{%
+ \@ifmmode{}{\@badmath$}%
+ \let\@acoll\@arrayacol
+ \let\@acolr\@arrayacol
+ \let\@acol\@arrayacol
+ \let\@classz\@arrayclassz
+ \let\@classiv\@arrayclassiv
+ \let\\\@arraycr
+ \let\@halignto\@empty
+ \@tabarray
+}%
+\def\array@array{%
+ \col@sep\arraycolsep
+ \def\d@llarbegin{$}\let\d@llarend\d@llarbegin\gdef\@halignto{}%
+ \@tabarray
+}
+\def\array@array@new{%
+ \@ifmmode{}{\@badmath$}%
+ \let\@acoll\@arrayacol
+ \let\@acolr\@arrayacol
+ \let\@acol\@arrayacol
+\let\col@sep\@undefined
+ \def\d@llarbegin{$}%
+ \let\d@llarend\d@llarbegin
+ \gdef\@halignto{}%
+ \@tabarray
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\@array}
+% Here we provide old and new versions of \cmd\@array.
+% The change here is to provide a convenient, flexible, and extensible
+% mechanism for new vertical alignment options.
+%
+% Instead of testing the optional argument with \cmd\if, we
+% use a dispatcher based on \cmd\csname.
+%
+% We also refrain from using \cmd\ialign, which would set
+% the \cmd\tabskip\ to the wrong value.
+%
+% Finally, the procedure to set the \cmd\@arstrutbox\
+% is broken out so that it can be patched.
+% \begin{macrocode}
+\def\@array@LaTeX[#1]#2{%
+ \if #1t\vtop \else \if#1b\vbox \else \vcenter \fi\fi
+ \bgroup
+ \setbox\@arstrutbox\hbox{%
+ \vrule \@height\arraystretch\ht\strutbox
+ \@depth\arraystretch \dp\strutbox
+ \@width\z@}%
+ \@mkpream{#2}%
+ \edef\@preamble{%
+ \ialign \noexpand\@halignto
+ \bgroup \@arstrut \@preamble \tabskip\z@skip \cr}%
+ \let\@startpbox\@@startpbox \let\@endpbox\@@endpbox
+ \let\tabularnewline\\%
+ \let\par\@empty
+ \let\@sharp##%
+ \set@typeset@protect
+ \lineskip\z@skip\baselineskip\z@skip
+ \ifhmode \@preamerr\z@ \@@par\fi
+ \@preamble
+}%
+\def\@array@ltx[#1]#2{%
+ \@nameuse{@array@align@#1}%
+ \set@arstrutbox
+ \@mkpream{#2}%
+ \prepdef\@preamble{%
+ \tabskip\tabmid@skip
+ \@arstrut
+ }%
+ \appdef\@preamble{%
+ \tabskip\tabright@skip
+ \cr
+ \array@row@pre
+ }%
+% \let\@startpbox\@@startpbox
+% \let\@endpbox\@@endpbox
+ \let\tabularnewline\\%
+ \let\par\@empty
+ \let\@sharp##%
+ \set@typeset@protect
+ \lineskip\z@skip\baselineskip\z@skip
+ \tabskip\tableft@skip\relax
+ \ifhmode \@preamerr\z@ \@@par\fi
+ \everycr{}%
+ \expandafter\halign\expandafter\@halignto\expandafter\bgroup\@preamble
+}%
+%
+\def\set@arstrutbox{%
+ \setbox\@arstrutbox\hbox{%
+ \vrule \@height\arraystretch\ht\strutbox
+ \@depth\arraystretch \dp\strutbox
+ \@width\z@
+ }%
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\@array@array}
+%
+% \begin{macrocode}
+\def\@array@array[#1]#2{%
+ \@tempdima \ht \strutbox
+ \advance \@tempdima by\extrarowheight
+ \setbox \@arstrutbox \hbox{\vrule
+ \@height \arraystretch \@tempdima
+ \@depth \arraystretch \dp \strutbox
+ \@width \z@}%
+ \begingroup
+ \@mkpream{#2}%
+ \xdef\@preamble{\noexpand \ialign \@halignto
+ \bgroup \@arstrut \@preamble
+ \tabskip \z@ \cr}%
+ \endgroup
+ \@arrayleft
+ \if #1t\vtop \else \if#1b\vbox \else \vcenter \fi \fi
+ \bgroup
+ \let \@sharp ##\let \protect \relax
+ \lineskip \z@
+ \baselineskip \z@
+ \m@th
+ \let\\\@arraycr \let\tabularnewline\\\let\par\@empty \@preamble
+}%
+\def\@array@array@new[#1]#2{%
+ \@tempdima\ht\strutbox
+ \advance\@tempdima by\extrarowheight
+ \setbox\@arstrutbox\hbox{%
+ \vrule \@height\arraystretch\@tempdima
+ \@depth \arraystretch\dp\strutbox
+ \@width \z@
+ }%
+ \begingroup
+ \@mkpream{#2}%
+ \xdef\@preamble{\@preamble}%
+ \endgroup
+ \prepdef\@preamble{%
+ \tabskip\tabmid@skip
+ \@arstrut
+ }%
+ \appdef\@preamble{%
+ \tabskip\tabright@skip
+ \cr
+ \array@row@pre
+ }%
+ \@arrayleft
+ \@nameuse{@array@align@#1}%
+ \m@th
+ \let\\\@arraycr
+ \let\tabularnewline\\%
+ \let\par\@empty
+ \let\@sharp##%
+ \set@typeset@protect
+ \lineskip\z@\baselineskip\z@
+ \tabskip\tableft@skip
+ \everycr{}%
+ \expandafter\halign\expandafter\@halignto\expandafter\bgroup\@preamble
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\endarray}
+% Here we provide old and new versions of \cmd\endarray.
+% The change here is to use a single procedure to close
+% out any array-like structure, namely \cmd\endarray@ltx.
+% It merely closes out the \cmd\halign.
+% \begin{macrocode}
+\def\endarray@LaTeX{%
+ \crcr\egroup\egroup
+}%
+\def\endarray@ltx{%
+ \crcr\array@row@pst\egroup\egroup
+}%
+\def\endarray@array{%
+ \crcr \egroup \egroup \@arrayright \gdef\@preamble{}%
+}%
+\def\endarray@array@new{%
+ \crcr\array@row@pst\egroup\egroup % Same as \endarray@ltx
+ \@arrayright
+ \global\let\@preamble\@empty
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\endtabular}
+%
+% \begin{macrocode}
+\def\endtabular@LaTeX{%
+ \crcr\egroup\egroup $\egroup
+}%
+\def\endtabular@ltx{%
+ \endarray
+}%
+\def\endtabular@array{%
+ \endarray $\egroup
+}%
+\def\endtabular@array@new{%
+ \endarray
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{endtabular*}
+% Here we provide a proper definition for the star-form of \enve{endtabular}.
+% It is one of the enduring curiosities that the \LaTeX\ kernel continues to use
+% dangerously and inappropriately ``optimized'' definitions for such commands.
+% \begin{macrocode}
+\@namedef{endtabular*}{\endtabular}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\multicolumn}
+%
+% \begin{macrocode}
+\long\def\multicolumn@LaTeX#1#2#3{%
+ \multispan{#1}\begingroup
+ \@mkpream{#2}%
+ \def\@sharp{#3}\set@typeset@protect
+ \let\@startpbox\@@startpbox\let\@endpbox\@@endpbox
+ \@arstrut \@preamble\hbox{}\endgroup\ignorespaces
+}%
+\long\def\multicolumn@ltx#1#2#3{%
+ \multispan{#1}%
+ \begingroup
+ \@mkpream{#2}%
+ \def\@sharp{#3}%
+ \set@typeset@protect
+ %\let\@startpbox\@@startpbox\let\@endpbox\@@endpbox
+ \@arstrut
+ \@preamble
+ \hbox{}%
+ \endgroup
+ \ignorespaces
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\@array@align@}
+% \begin{macro}{\array@default}
+% Here are the various procedures for the vertical alignment options.
+% The change from standard \LaTeX\ is that we do not go into math mode
+% in every case: only when required by \cmd\vcenter.
+% Also, we use \cmd\aftergroup\ to close out the boxes and modes we have started.
+% It requires only that each procedure issue exactly one unmatched \cmd\bgroup.
+%
+% We establish here the default vertical alignment.
+% \begin{macrocode}
+\def\@array@align@t{\leavevmode\vtop\bgroup}%
+\def\@array@align@b{\leavevmode\vbox\bgroup}%
+\def\@array@align@c{\leavevmode\@ifmmode{\vcenter\bgroup}{$\vcenter\bgroup\aftergroup$\aftergroup\relax}}%
+\def\@array@align@v{%
+ \@ifmmode{%
+ \@badmath
+ \vcenter\bgroup
+ }{%
+ \@ifinner{%
+ $\vcenter\bgroup\aftergroup$
+ }{%
+ \@@par\bgroup
+ }%
+ }%
+}%
+\def\array@default{c}%
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\array@row@pre}
+% \begin{macro}{\array@row@pst}
+% \begin{macro}{\array@row@rst}
+% The procedure \cmd\array@row@rst\ reestablishes a default context for
+% an alignment, so that they can be nested.
+% Any environment or procedure that alters the way alignments are formatted
+% must patch this procedure to restore from that alteration.
+% To start things off, we equate \cmd\@array@align@v\ to \cmd\@array@align@c,
+% because it does not make sense to do the former in any context other
+% than the MVL or in a list that will be unboxed onto the MVL.
+% \begin{macrocode}
+\def\array@row@rst{%
+ \let\@array@align@v\@array@align@c
+}%
+\def\array@row@pre{}%
+\def\array@row@pst{}%
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\toprule}
+% \begin{macro}{\colrule}
+% \begin{macro}{\botrule}
+% Default definitions for \cmd\toprule, \cmd\colrule, \cmd\botrule
+% \begin{macrocode}
+\newcommand\toprule{\tab@rule{\column@font}{\column@fil}{\frstrut}}%
+\newcommand\colrule{\unskip\lrstrut\\\tab@rule{\body@font}{}{\frstrut}}%
+\newcommand\botrule{\unskip\lrstrut\\\noalign{\hline@rule}{}}%
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\hline}
+% \begin{macrocode}
+\def\hline@LaTeX{%
+ \noalign{\ifnum0=`}\fi\hrule \@height \arrayrulewidth \futurelet
+ \reserved@a\@xhline
+}%
+\def\hline@ltx{%
+ \noalign{%
+ \ifnum0=`}\fi
+ \hline@rule
+ \futurelet\reserved@a\@xhline
+ % \noalign ended in \@xhline
+}%
+\def\@xhline@unneeded{%
+ \say\reserved@a
+ \ifx\reserved@a\hline
+ \vskip\doublerulesep
+ \vskip-\arrayrulewidth
+ \fi
+ \ifnum0=`{\fi}%
+}%
+\def\tab@rule#1#2#3{%
+ \crcr
+ \noalign{%
+ \hline@rule
+ \gdef\@arstrut@hook{%
+ \global\let\@arstrut@hook\@empty
+ #3%
+ }%
+ \gdef\cell@font{#1}%
+ \gdef\cell@fil{#2}%
+ }%
+}%
+\def\column@font{}%
+\def\column@fil{}%
+\def\body@font{}%
+\def\cell@font{}%
+\def\frstrut{}%
+\def\lrstrut{}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\@arstrut@hline}
+% \begin{macro}{\@arstrut@org}
+% \begin{macro}{\@arstrut@hook}
+% \begin{macro}{\@arstrutbox@hline}
+% \begin{macro}{\set@arstrutbox}
+% \begin{macro}{\hline@rule}
+% The procedure \cmd\@arstrut@hline\ is substantially the same as
+% \cmd\@arstrut, except the strut copied in is \cmd\@arstrutbox@hline
+% instead of \cmd\@arstrutbox.
+%
+% The procedure \cmd\@arstrut@hook\ is redefined in \cmd\tab@rule!
+%
+% The register \cmd\@arstrutbox@hline.
+%
+% We append to \cmd\set@arstrutbox\ the code necessary to set a strut following an \cmd\hline.
+%
+% The procedure \cmd\hline@rule\ lays down a rule, and changes the meaning of \cmd\@arstrut\
+% so that the next line will be correctly strutted.
+%
+% The \cmd\@arstrut@hline@clnc\ is a klootch, a magic number.
+% \begin{macrocode}
+\def\@arstrut@hline{%
+ \relax
+ \@ifmmode{\copy}{\unhcopy}\@arstrutbox@hline
+ \@arstrut@hook
+}%
+%
+\let\@arstrut@org\@arstrut
+\def\@arstrut@hook{%
+ \global\let\@arstrut\@arstrut@org
+}%
+%
+\newbox\@arstrutbox@hline
+\appdef\set@arstrutbox{%
+ \setbox\@arstrutbox@hline\hbox{%
+ \setbox\z@\hbox{$0^{0}_{}$}%
+ \dimen@\ht\z@\advance\dimen@\@arstrut@hline@clnc
+ \@ifdim{\dimen@<\arraystretch\ht\strutbox}{\dimen@=\arraystretch\ht\strutbox}{}%
+ \vrule \@height\dimen@
+ \@depth\arraystretch \dp\strutbox
+ \@width\z@
+ }%
+}%
+%
+\def\hline@rule{%
+ \hrule \@height \arrayrulewidth
+ \global\let\@arstrut\@arstrut@hline
+}%
+\def\@arstrut@hline@clnc{2\p@}% % Klootch: magic number
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\tableft@skip}
+% \begin{macrocode}
+\def\tableft@skip{\z@skip}%
+\def\tabmid@skip{\z@skip}%\@flushglue
+\def\tabright@skip{\z@skip}%
+\def\tableftsep{\tabcolsep}%
+\def\tabmidsep{\tabcolsep}%
+\def\tabrightsep{\tabcolsep}%
+\def\cell@fil{}%
+\def\pbox@hook{}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\@arstrut}
+% \begin{macrocode}
+\appdef\@arstrut{\@arstrut@hook}%
+\let\@arstrut@hook\@empty
+\def\@addtopreamble{\appdef\@preamble}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\@mkpream}
+% \begin{macrocode}
+\def\@mkpream@LaTeX#1{%
+ \@firstamptrue\@lastchclass6
+ \let\@preamble\@empty
+ \let\protect\@unexpandable@protect
+ \let\@sharp\relax
+ \let\@startpbox\relax\let\@endpbox\relax
+ \@expast{#1}%
+ \expandafter\@tfor \expandafter
+ \@nextchar \expandafter:\expandafter=\reserved@a\do
+ {\@testpach\@nextchar
+ \ifcase \@chclass \@classz \or \@classi \or \@classii \or \@classiii
+ \or \@classiv \or\@classv \fi\@lastchclass\@chclass}%
+ \ifcase \@lastchclass \@acol
+ \or \or \@preamerr \@ne\or \@preamerr \tw@\or \or \@acol \fi
+}%
+\def\@mkpream@ltx#1{%
+ \@firstamptrue
+ \@lastchclass6
+ \let\@preamble\@empty
+ \let\protect\@unexpandable@protect
+ \let\@sharp\relax
+%\let\@startpbox\relax\let\@endpbox\relax
+ \@expast{#1}%
+ \expandafter\@tfor\expandafter\@nextchar\expandafter:\expandafter=\reserved@a
+ \do{%
+ \expandafter\@testpach\expandafter{\@nextchar}%
+ \ifcase\@chclass
+ \@classz
+ \or
+ \@classi
+ \or
+ \@classii
+ \or
+ \@classiii
+ \or
+ \@classiv
+ \or
+ \@classv
+ \fi
+ \@lastchclass\@chclass
+ }%
+ \ifcase\@lastchclass
+ \@acolr % right-hand column
+ \or
+ \or
+ \@preamerr\@ne
+ \or
+ \@preamerr\tw@
+ \or
+ \or
+ \@acolr % right-hand column
+ \fi
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\insert@column}
+% \begin{macrocode}
+\def\insert@column@array{%
+ \the@toks \the \@tempcnta
+ \ignorespaces \@sharp \unskip
+ \the@toks \the \count@ \relax
+}%
+\def\insert@column@array@new{%
+ \the@toks\the\@tempcnta
+ \array@row@rst\cell@font
+ \ignorespaces\@sharp\unskip
+ \the@toks\the\count@
+ \relax
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\@mkpream@relax}
+% The procedure \cmd\@mkpream@relax\ participates in a strange and wonderful
+% method of binding the alignment procedure---but only certain parts thereof.
+%
+% Here is how it works: in \LaTeX, the \classname{array} package, and in the
+% \classname{longtable} package alike, there is a need to create an alignment
+% preamble (using \cmd\@mkpream) for use by the upcoming \cmd\halign.
+% Then, in both \classname{array} and \classname{longtable}, \TeX's \cmd\edef\
+% is used to `compile in place' that alignment preamble.
+%
+% In the case of \classname{array}, the operation is done in order to
+% pre-expand the use of \texttt{*}, in \classname{longtable}, it is to
+% set the widths of the columns.
+%
+% Now, during this \cmd\edef, certain control sequence names must \emph{not}
+% be expanded, and those are robustified by \cmd\@mkpream@relax.
+%
+% \begin{macrocode}
+\def\@mkpream@relax{%
+ \let\tableftsep\relax
+ \let\tabmidsep\relax
+ \let\tabrightsep\relax
+ \let\array@row@rst\relax
+ \let\cell@font\relax
+ \let\@startpbox\relax
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\@mkpream}
+% \begin{macrocode}
+\def\@mkpream@array#1{%
+ \gdef\@preamble{}\@lastchclass 4 \@firstamptrue
+ \let\@sharp\relax \let\@startpbox\relax \let\@endpbox\relax
+ \@temptokena{#1}\@tempswatrue
+ \@whilesw\if@tempswa\fi{\@tempswafalse\the\NC@list}%
+ \count@\m@ne
+ \let\the@toks\relax
+ \prepnext@tok
+ \expandafter \@tfor \expandafter \@nextchar
+ \expandafter :\expandafter =\the\@temptokena \do
+ {\@testpach
+ \ifcase \@chclass \@classz \or \@classi \or \@classii
+ \or \save@decl \or \or \@classv \or \@classvi
+ \or \@classvii \or \@classviii
+ \or \@classx
+ \or \@classx \fi
+ \@lastchclass\@chclass}%
+ \ifcase\@lastchclass
+ \@acol \or
+ \or
+ \@acol \or
+ \@preamerr \thr@@ \or
+ \@preamerr \tw@ \@addtopreamble\@sharp \or
+ \or
+ \else \@preamerr \@ne \fi
+ \def\the@toks{\the\toks}%
+}%
+\def\@mkpream@array@new#1{%
+ \gdef\@preamble{}%
+ \@lastchclass\f@ur
+ \@firstamptrue
+ \let\@sharp\relax
+ \@mkpream@relax
+%\let\@startpbox\relax\let\@endpbox\relax
+ \@temptokena{#1}\@tempswatrue
+ \@whilesw\if@tempswa\fi{\@tempswafalse\the\NC@list}%
+ \count@\m@ne
+ \let\the@toks\relax
+ \prepnext@tok
+ \expandafter\@tfor\expandafter\@nextchar\expandafter:\expandafter=\the\@temptokena
+ \do{%
+ \@testpach
+ \ifcase\@chclass
+ \@classz
+ \or
+ \@classi
+ \or
+ \@classii
+ \or
+ \save@decl
+ \or
+ \or
+ \@classv
+ \or
+ \@classvi
+ \or
+ \@classvii
+ \or
+ \@classviii
+ \or
+ \@classx
+ \or
+ \@classx
+ \fi
+ \@lastchclass\@chclass
+ }%
+ \ifcase\@lastchclass
+ \@acolr % right-hand column
+ \or
+ \or
+ \@acolr % right-hand column
+ \or
+ \@preamerr\thr@@
+ \or
+ \@preamerr\tw@\@addtopreamble\@sharp
+ \or
+ \or
+ \else
+ \@preamerr\@ne
+ \fi
+ \def\the@toks{\the\toks}%
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\@addamp}
+% \begin{macrocode}
+\def\@addamp@LaTeX{%
+ \if@firstamp\@firstampfalse\else\edef\@preamble{\@preamble &}\fi
+}%
+\def\@addamp@ltx{%
+ \if@firstamp\@firstampfalse\else\@addtopreamble{&}\fi
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\@arrayacol}
+% \begin{macrocode}
+\def\@arrayacol@LaTeX{%
+ \edef\@preamble{\@preamble \hskip \arraycolsep}%
+}%
+\def\@arrayacol@ltx{%
+ \@addtopreamble{\hskip\arraycolsep}%
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\@tabacol}
+% \begin{macrocode}
+\def\@tabacoll{%
+ \@addtopreamble{\hskip\tableftsep\relax}%
+}%
+\def\@tabacol@LaTeX{%
+ \edef\@preamble{\@preamble \hskip \tabcolsep}%
+}%
+\def\@tabacol@ltx{%
+ \@addtopreamble{\hskip\tabmidsep\relax}%
+}%
+\def\@tabacolr{%
+ \@addtopreamble{\hskip\tabrightsep\relax}%
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\@arrayclassz}
+% \begin{macrocode}
+\def\@arrayclassz@LaTeX{%
+ \ifcase \@lastchclass \@acolampacol \or \@ampacol \or
+ \or \or \@addamp \or
+ \@acolampacol \or \@firstampfalse \@acol \fi
+ \edef\@preamble{\@preamble
+ \ifcase \@chnum
+ \hfil$\relax\@sharp$\hfil \or $\relax\@sharp$\hfil
+ \or \hfil$\relax\@sharp$\fi}%
+}%
+\def\@arrayclassz@ltx{%
+ \ifcase\@lastchclass
+ \@acolampacol
+ \or
+ \@ampacol
+ \or
+ \or
+ \or
+ \@addamp
+ \or
+ \@acolampacol
+ \or
+ \@firstampfalse\@acoll
+ \fi
+ \ifcase\@chnum
+ \@addtopreamble{%
+ \hfil\array@row@rst$\relax\@sharp$\hfil
+ }%
+ \or
+ \@addtopreamble{%
+ \array@row@rst$\relax\@sharp$\hfil
+ }%
+ \or
+ \@addtopreamble{%
+ \hfil\array@row@rst$\relax\@sharp$%
+ }%
+ \fi
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\@tabclassz}
+% \begin{macrocode}
+\def\@tabclassz@LaTeX{%
+ \ifcase\@lastchclass
+ \@acolampacol
+ \or
+ \@ampacol
+ \or
+ \or
+ \or
+ \@addamp
+ \or
+ \@acolampacol
+ \or
+ \@firstampfalse\@acol
+ \fi
+ \edef\@preamble{%
+ \@preamble{%
+ \ifcase\@chnum
+ \hfil\ignorespaces\@sharp\unskip\hfil
+ \or
+ \hskip1sp\ignorespaces\@sharp\unskip\hfil
+ \or
+ \hfil\hskip1sp\ignorespaces\@sharp\unskip
+ \fi}}%
+}%
+\def\@tabclassz@ltx{%
+ \ifcase\@lastchclass
+ \@acolampacol
+ \or
+ \@ampacol
+ \or
+ \or
+ \or
+ \@addamp
+ \or
+ \@acolampacol
+ \or
+ \@firstampfalse\@acoll
+ \fi
+ \ifcase\@chnum
+ \@addtopreamble{%
+ {\hfil\array@row@rst\cell@font\ignorespaces\@sharp\unskip\hfil}%
+ }%
+ \or
+ \@addtopreamble{%
+ {\cell@fil\hskip1sp\array@row@rst\cell@font\ignorespaces\@sharp\unskip\hfil}%
+ }%
+ \or
+ \@addtopreamble{%
+ {\hfil\hskip1sp\array@row@rst\cell@font\ignorespaces\@sharp\unskip\cell@fil}%
+ }%
+ \fi
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\@tabclassiv}
+% \begin{macrocode}
+\def\@tabclassiv@LaTeX{%
+ \@addtopreamble\@nextchar
+}%
+\def\@tabclassiv@ltx{%
+ \expandafter\@addtopreamble\expandafter{\@nextchar}%
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\@arrayclassiv}
+% \begin{macrocode}
+\def\@arrayclassiv@LaTeX{%
+ \@addtopreamble{$\@nextchar$}%
+}%
+\def\@arrayclassiv@ltx{%
+ \expandafter\@addtopreamble\expandafter{\expandafter$\@nextchar$}%
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\@classv}
+% \begin{macrocode}
+\def\@classv@LaTeX{%
+ \@addtopreamble{\@startpbox{\@nextchar}\ignorespaces
+ \@sharp\@endpbox}%
+}%
+\def\@classv@ltx{%
+ \expandafter\@addtopreamble
+ \expandafter{%
+ \expandafter \@startpbox
+ \expandafter {\@nextchar}%
+ \pbox@hook\array@row@rst\cell@font\ignorespaces\@sharp\@endpbox
+ }%
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\@classx}
+% \begin{macrocode}
+\def\@classx@array{%
+ \ifcase \@lastchclass
+ \@acolampacol \or
+ \@addamp \@acol \or
+ \@acolampacol \or
+ \or
+ \@acol \@firstampfalse \or
+ \@addamp
+ \fi
+}%
+\def\@classx@array@new{%
+ \ifcase \@lastchclass
+ \@acolampacol
+ \or
+ \@addamp \@acol
+ \or
+ \@acolampacol
+ \or
+ \or
+ \@firstampfalse\@acoll
+ \or
+ \@addamp
+ \fi
+}%
+% \end{macrocode}
+% \end{macro}
+%
+%
+% \subsection{Repair other broken parts of \LaTeX}
+%
+% \begin{macro}{\@xbitor}
+% Expansion part has extraneous space token. Removed.
+% \begin{macrocode}
+\def\@xbitor@LaTeX #1{\@tempcntb \count#1
+ \ifnum \@tempcnta =\z@
+ \else
+ \divide\@tempcntb\@tempcnta
+ \ifodd\@tempcntb \@testtrue\fi
+ \fi}%
+\def\@xbitor@ltx#1{%
+ \@tempcntb\count#1%
+ \@ifnum{\@tempcnta=\z@}{}{%
+ \divide\@tempcntb\@tempcnta
+ \@ifodd\@tempcntb{\@testtrue}{}%
+ }%
+}%
+\@ifx{\@xbitor\@xbitor@LaTeX}{%
+ \class@info{Repairing broken LaTeX \string\@xbitor}%
+}{%
+ \class@info{Unrecognized LaTeX \string\@xbitor. Please update this document class! (Proceeding with fingers crossed.)}%
+}%
+\let\@xbitor\@xbitor@ltx
+% \end{macrocode}
+% \end{macro}
+%
+%
+% \subsection{Syntax}
+% \begin{macro}{\@gobble@opt@one}
+% The \cmd\@gobble@opt@one\ command eats up an optional argument
+% and one required argument.
+% \begin{macrocode}
+\newcommand*\@gobble@opt@one[2][]{}%
+% \end{macrocode}
+% \end{macro}
+%
+% \subsection{Auto-indented Contents}
+% Facility to automatically determine the proper indentation of
+% the TOC entries.
+%
+% Note on \classname{hyperref} compatibility:
+% We must respect that
+% \cmd\contentsline now has a 4th argument.
+% So, instead of trying to override the meaning of \cmd\contentsline,
+% we use the aux file to remember max values from one run to the next.
+%
+% In this respect, this package retains compatability with
+% \classname{hyperref}.
+%
+% \begin{macro}{\@starttoc}
+% Install hooks at beginning and end of the TOC processing.
+% \begin{macrocode}
+\def\@starttoc#1{%
+ \begingroup
+ \toc@pre
+ \makeatletter
+ \@input{\jobname.#1}%
+ \if@filesw
+ \expandafter\newwrite\csname tf@#1\endcsname
+ \immediate\openout \csname tf@#1\endcsname \jobname.#1\relax
+ \fi
+ \@nobreakfalse
+ \toc@post
+ \endgroup
+}%
+\def\toc@pre{}%
+\def\toc@post{}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\toc@@font}
+% Interface for setting the formatting characteristics of this part
+% of the TOC.
+%
+% Note: \cmd\toc@@font\ is the common font for all auto-sizing toc commands,
+% although this, too, could become a dispatcher.
+% \begin{macrocode}
+\def\toc@@font{}%{\footnotesize\rmfamily}%
+\def\@dotsep{\z@}%{5.5pt}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\l@section}
+% Interface for determining which TOC elements are automatically indented.
+%
+% All of the \cmd\l@\dots\ commands simply go through the
+% bottleproc \cmd\l@@sections. The calling convention is
+% to pass the name of self and the name of parent.
+% If you want to exclude any of these from the indentation
+% scheme, simply leave the \cmd\l@\dots\ command undefined.
+%
+% Note that the parent of ``section'' is nil, so we have to
+% define a stub.
+% \begin{verbatim}
+%\def\l@section{%
+% \l@@sections{}{section}% Implicit #3#4
+%}%
+%\def\tocleft@{\z@}%
+%\def\l@subsection{%
+% \l@@sections{section}{subsection}% Implicit #3#4
+%}%
+%\def\l@subsubsection{%
+% \l@@sections{subsection}{subsubsection}% Implicit #3#4
+%}%
+%\def\l@paragraph{%
+% \l@@sections{subsubsection}{paragraph}% Implicit #3#4
+%}%
+%\def\l@subparagraph#1#2{%
+% \l@@sections{paragraph}{subparagraph}% Implicit #3#4
+%}%
+% \end{verbatim}
+% \end{macro}
+%
+% Glom some \cmd\dimen\ registers.
+% \begin{macrocode}
+\let\tocdim@section \leftmargini
+\let\tocdim@subsection \leftmarginii
+\let\tocdim@subsubsection \leftmarginiii
+\let\tocdim@paragraph \leftmarginiv
+\let\tocdim@appendix \leftmarginv
+\let\tocdim@pagenum \leftmarginvi
+% \end{macrocode}
+%
+% \begin{macro}{\toc@pre@auto}
+% \begin{macro}{\toc@post@auto}
+% We patch \cmd\@starttoc\ to:
+% 1) before TOC processing,
+% initialize the max registers and
+% set the needed dimensions from
+% the values stored in the auxiliary file, and
+% 2) after TOC processing,
+% store out those max register values into the auxiliary file.
+%
+% Note that the font is set here: all other TOC entries must
+% override these font settings.
+%
+% To activate this override of the standard \LaTeX\ processing,
+% the substyle does: \cmd\let\cmd\toc@pre\cmd\toc@pre@auto\
+% and \cmd\let\cmd\toc@post\cmd\toc@post@auto.
+% \begin{macrocode}
+\def\toc@pre@auto{%
+ \toc@@font
+ \@tempdima\z@
+ \toc@setindent\@tempdima{section}%
+ \toc@setindent\@tempdima{subsection}%
+ \toc@setindent\@tempdima{subsubsection}%
+ \toc@setindent\@tempdima{paragraph}%
+ \toc@letdimen{appendix}%
+ \toc@letdimen{pagenum}%
+}%
+\def\toc@post@auto{%
+ \if@filesw
+ \begingroup
+ \toc@writedimen{section}%
+ \toc@writedimen{subsection}%
+ \toc@writedimen{subsubsection}%
+ \toc@writedimen{paragraph}%
+ \toc@writedimen{appendix}%
+ \toc@writedimen{pagenum}%
+ \endgroup
+ \fi
+}%
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\toc@setindent}
+% \begin{macrocode}
+\def\toc@setindent#1#2{%
+ \csname tocdim@#2\endcsname\tocdim@min\relax
+ \@ifundefined{tocmax@#2}{\@namedef{tocmax@#2}{\z@}}{}%
+ \advance#1\@nameuse{tocmax@#2}\relax
+ \expandafter\edef\csname tocleft@#2\endcsname{\the#1}%
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\toc@letdimen}
+% \begin{macrocode}
+\def\toc@letdimen#1{%
+ \csname tocdim@#1\endcsname\tocdim@min\relax
+ \@ifundefined{tocmax@#1}{\@namedef{tocmax@#1}{\z@}}{}%
+ \expandafter\let\csname tocleft@#1\expandafter\endcsname\csname tocmax@#1\endcsname
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\toc@writedimen}
+% \begin{macrocode}
+\def\toc@writedimen#1{%
+ \immediate\write\@auxout{%
+ \gdef\expandafter\string\csname tocmax@#1\endcsname{%
+ \expandafter\the\csname tocdim@#1\endcsname
+ }%
+ }%
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\l@@sections}
+% The procedure for formatting the indented TOC entries.
+% We use control sequence names such as \cmd\tocmax@section\ and
+% \cmd\tocleft@section, the former being written to the auxiliary file
+% and the latter only defined for the duration of the TOC processing.
+%
+% Note that the assignment of \cmd\box\cmd\z@\ must endure
+% over the invocation of |#3|.
+% \begin{macrocode}
+\def\l@@sections#1#2#3#4{%
+ % #1 - superior section
+ % #2 - this section
+ % #3 - content, including possible \numberline
+ % #4 - page number
+ \begingroup
+ \everypar{}%
+ \set@tocdim@pagenum{#4}%
+ \global\@tempdima\csname tocdim@#2\endcsname
+ \leftskip\csname tocleft@#2\endcsname\relax
+ \dimen@\csname tocleft@#1\endcsname\relax
+ \parindent-\leftskip\advance\parindent\dimen@
+ \rightskip\tocleft@pagenum plus 1fil\relax
+ \skip@\parfillskip\parfillskip\z@
+ \let\numberline\numberline@@sections
+ \@nameuse{l@f@#2}%
+ \ignorespaces#3\unskip\nobreak\hskip\skip@
+ \hb@xt@\rightskip{\hfil\unhbox\@tempboxa}\hskip-\rightskip\hskip\z@skip
+ \par
+ \expandafter\aftergroup\csname tocdim@#2\endcsname\expandafter
+ \endgroup\the\@tempdima\relax
+}%
+\def\set@tocdim@pagenum#1{%
+ \setbox\@tempboxa\hbox{\ignorespaces#1}%
+ \@ifdim{\tocdim@pagenum<\wd\z@}{\global\tocdim@pagenum\wd\z@}{}%
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\numberline@@sections}
+% The bottleproc for all \cmd\numberline\ processing in indented TOC entries.
+% The first argument is self.
+%
+% We use \cmd\@tempdima\ to pass a value around (via global assignment) because
+% \cmd\numberline\ executes inside a group if the
+% \classname{hyperref} package is loaded.
+% Would that it were not so!
+% \begin{macrocode}
+\def\numberline@@sections#1{%
+ \leavevmode\hb@xt@-\parindent{%
+ \hfil
+ \@if@empty{#1}{}{%
+ \setbox\z@\hbox{#1.\kern\@dotsep}%
+ \@ifdim{\@tempdima<\wd\z@}{\global\@tempdima\wd\z@}{}%
+ \unhbox\z@
+ }%
+ }%
+ \ignorespaces
+}%
+\def\tocdim@min{\z@}%
+% \end{macrocode}
+% \end{macro}
+%
+%
+% \subsection{Lists}
+% \begin{macro}{\list}
+% Using \cmd\parshape\ to implement lists was always suspect
+% (can you get behind \cmd\parshape\cmd\@ne?) and we now see that
+% it was a mistake all along. Why? Because \cmd\parshape, like
+% \cmd\hangindent, achieves its effect via ``shifting'' the \cmd\hbox es
+% in a paragraph
+% instead of using \cmd\leftskip\ and \cmd\parindent, which is
+% robust during column balancing.
+%
+% We introduce the alternative method with a hook into
+% the \LaTeX\ kernel procedure \cmd\list, which is
+% the implementation of all lists.
+%
+% \begin{macrocode}
+\def\list#1#2{%
+ \ifnum \@listdepth >5\relax
+ \@toodeep
+ \else
+ \global\advance\@listdepth\@ne
+ \fi
+ \rightmargin\z@
+ \listparindent\z@
+ \itemindent\z@
+ \csname @list\romannumeral\the\@listdepth\endcsname
+ \def\@itemlabel{#1}%
+ \let\makelabel\@mklab
+ \@nmbrlistfalse
+ #2\relax
+ \@trivlist
+ \parskip\parsep
+ \set@listindent
+ \ignorespaces
+}%
+\def\set@listindent@parshape{%
+ \parindent\listparindent
+ \advance\@totalleftmargin\leftmargin
+ \advance\linewidth-\rightmargin
+ \advance\linewidth-\leftmargin
+ \parshape\@ne\@totalleftmargin\linewidth
+}%
+\def\set@listindent@{%
+ \parindent\listparindent
+ \advance\@totalleftmargin\leftmargin
+ \advance\rightskip\rightmargin
+ \advance\leftskip\@totalleftmargin
+}%
+\let\set@listindent\set@listindent@parshape
+% \end{macrocode}
+% \end{macro}
+%
+% \subsection{End of the \file{ltxutil} {\sc docstrip} module}
+% Here ends the module.
+% \begin{macrocode}
+%</ltxutil-krn>
+% \end{macrocode}
+%
+%
+% \Finale
+% %Here ends the programmer's documentation.
+% \endinput
+%
+\endinput
+%%EOF
diff --git a/Master/texmf-dist/source/latex/revtex4/revtex4.dtx b/Master/texmf-dist/source/latex/revtex4/revtex4.dtx
new file mode 100644
index 00000000000..693747ef47f
--- /dev/null
+++ b/Master/texmf-dist/source/latex/revtex4/revtex4.dtx
@@ -0,0 +1,8728 @@
+% \iffalse
+% ****** Start of file revtex4.dtx ****** %
+%
+% This file is part of the APS files in the REVTeX 4 distribution.
+% For version number, search on the string "%FileInfo"
+%
+% First revision by David Carlisle,
+% later versions by Arthur Ogawa.
+%
+% Copyright (c) 2000 The American Physical Society.
+% http://publish.aps.org/revtex4/
+% mailto:revtex@aps.org
+%
+%
+% See the REVTeX 4 README file for restrictions and more information.
+% \fi
+% \iffalse
+%<*driver>
+ \ProvidesFile{revtex4.dtx}
+%</driver>
+%<class-pre|aps|rmp>\NeedsTeXFormat{LaTeX2e}[1996/12/01]
+%<class-pre>\ProvidesClass{revtex4}
+%<aps>\ProvidesFile{aps.rtx}
+%<rmp>\ProvidesFile{rmp.rtx}
+%<10pt>\ProvidesFile{10pt.rtx}
+%<11pt>\ProvidesFile{11pt.rtx}
+%<12pt>\ProvidesFile{12pt.rtx}
+%<revsymb&!class-pre>\ProvidesPackage{revsymb}
+%<driver>\ProvidesFile{revtex.drv}
+%<*!class-pst>
+ [2001/08/02 v4.0 (http://publish.aps.org/revtex4/ for documentation)]%FileInfo
+%</!class-pst>
+%<*driver>
+\listfiles
+\documentclass{ltxdoc}
+\RequirePackage{ltxdocext}%
+\let\url\undefined
+\RequirePackage[colorlinks=true,linkcolor=blue]{hyperref}%
+\expandafter\ifx\csname package@font\endcsname\relax\else
+ \expandafter\expandafter
+ \expandafter\RequirePackage
+ \expandafter\expandafter
+ \expandafter{\csname package@font\endcsname}%
+\fi
+\CodelineIndex\EnableCrossrefs
+\begin{document}
+\gdef\lquote{`}\gdef\rquote{'}%
+%{\tracingall\showlists}
+\DocInput{revtex4.dtx}
+\end{document}
+%</driver>
+% \fi
+%
+% \GetFileInfo{revtex4.dtx}\CheckSum{7177}
+%
+% \title{The \classname{revtex4} document class of the American Physical Society}
+% \author{Arthur Ogawa%
+% \thanks{First revision by David Carlisle}%
+% }%
+% \date{Version \fileversion, dated \filedate}%
+% \newcommand\revtex{REV\TeX}
+%
+% \maketitle
+%
+% This file embodies the implementation of the APS \revtex\ 4 document class
+% for electronic submissions to journals.
+%
+% The distribution point for this work is
+% \url{http://publish.aps.org/revtex4/},
+% which contains fully unpacked, prebuilt runtime files and documentation.
+%
+% \tableofcontents
+%
+% \section{Using \protect\revtex}
+%
+% The file \file{README} has retrieval and installation information.
+%
+% User documentation is presented separately in \file{revguide.tex}.
+%
+% The file \file{template.aps} is a boilerplate file.
+%
+% \StopEventually{}
+%
+% \changes{v4.a}{1998/01/16}
+% {Initial version}
+% \changes{v4.0beta 2}{1999/06/20}
+% {First modifications by Arthur Ogawa (mailto:ogawa@teleport.com)}
+% \changes{v4.0beta 3}{1999/11/13}
+% {New ltxgrid-based code, other bug fixes}
+% \changes{v4.0beta 4}{2000/05/10}
+% {More features and bug fixes: compatability with longtable and array packages.
+% Now certainly incompatible with multicol.}
+% \changes{v4.0beta 5}{2000/11/14}
+% {Bug fixes and minor new features: title block affiliations can have ancillary data, just like authors;
+% clearpage processing revamped, with floats staying in order; widetext ornaments.}
+% \changes{v4.0beta 6}{2001/02/13}
+% {Last bug fixes before release.}
+%
+% \section{Overview}%
+% \revtex\ is a \LaTeXe\ document class, aomewhat like a hybrid of
+% the standard \LaTeX\ \classname{book} and \classname{article} classes.
+%
+% Certain packages are (should be) loaded by this class in any case:
+% \classname{amssymb}, \classname{amsmath}, \classname{bm}, \classname{natbib}.
+%
+% Certain packages are automatically loaded by this class
+% when a corresponding class option has been invoked:
+%
+% \begin{tabular}{ll}
+% \textbf{\revtex\ option}&\textbf{package}\\
+% \classoption{amsfonts}&\classname{amsfonts}\\
+% \classoption{amssymb}&\classname{amssymb}\\
+% \classoption{aps}&\classname{overcite}\\
+% \end{tabular}
+%
+% Certain other packages are to be loaded by the
+% document through explicit use of \cmd\usepackage.
+% Some mentioned in the user documentation are
+% \classname{graphicx}, \classname{longtable},
+% and \classname{bm}.
+%
+% Certain commonly used packages are known to be incompatible with
+% \revtex, among them \classname{multicol} and \classname{cite}.
+% If such a package is found to be loaded, \revtex\ issues an error message
+% and halts the job.
+% Halting might be considered severe punishment for loading an incompatible
+% package, but if we were to proceed, an even weirder error might be encountered
+% further down the road.
+%
+% This document class implements the substyle: a set of mutually
+% exclusive class options that, in this case, allow the
+% document class to address multiple societies.
+% It also implements a sub-substyle, giving the
+% journal of the given society.
+% The only society currently implemented is \classoption{aps},
+% however, code is in place for another society, the \classoption{osa}.
+%
+% FIXME: should always load the \classname{graphicx} package.
+%
+% QUERY: since \classname{amsfonts} and \classname{amssymb} extend
+% syntax, why not load them in any case?
+%
+% QUERY: if you say \cmd\documentclass|[prb,osa]{revtex4}|, then
+% the value of \cmd\@society\ will be wrong, or? Fixed.
+%
+% Certain events occur at \cmd\AtEndOfClass\ time:
+% \begin{enumerate}
+% \item
+% Optionally read in packages
+% \classname{amsfonts}, \classname{amssymb}.
+% \item
+% Read in society substyle \file{.rtx} file.
+% \item
+% Read in type size \file{.rtx} file.
+% \item
+% Read in the job's style file, the \file{.rty} file.
+% \item
+% Define things based on \cmd\secnumarabic@sw.
+% \item
+% Install procedures to execute at the very end of the class's
+% \cmd\AtBeginDocument\ processing, such as
+% \begin{enumerate}
+% \item
+% closing out the page grid
+% \item
+% putting out the LastPage label.
+% \item
+% issuing a \cmd\bibliographystyle\ command, based on the value set
+% by the society substyle.
+% \item
+% setting default values for parameters used in the document.
+%FIXME: differentiate between class's parameters that can wait until
+% they are used in the document, and parameters that are used at \cmd\AtBeginDocument\ time.
+% \end{enumerate}
+% \item
+% Install procedures to execute the very last at \cmd\AtEndDocument\ time,
+% such as the \cmd\clearpage\ processing.
+% \end{enumerate}
+%
+% Certain events are optionally scheduled for \cmd\AtBeginDocument\ time:
+% \begin{enumerate}
+% \item
+% Setting default values for the Booleans and for other procedures used in formatting.
+% \item
+% In response to class options options, adjusting parameters and procedures used in formatting.
+% \item
+% Implementing the \classoption{eqsecnum} option, if required.
+% \item
+% Setting the state engine for data commands.
+% \item
+% Memorizing procedures for later use.
+% \item
+% Setting type size and area, for use by later calculations.
+% \end{enumerate}
+%
+% Certain events are scheduled for \cmd\class@enddocumenthook\ time:
+% \begin{enumerate}
+% \item
+% Print out the migrated floats or the end notes, if needed.
+% \item
+% Close out the page grid.
+% \item
+% Label the last page of the document
+% \item
+% (\classname{natbib}) prepare to read in the \file{.aux} file.
+% \end{enumerate}
+%
+%
+% \section{The installer file}
+% We include here a reference copy of the installer file, \file{revtex4.ins}.
+% It is not intended that one actually generate that file using \file{docstrip}
+% via the revtex4 installer file, because that would involve overwriting the
+% file that \TeX\ is typesetting. Instead, the installer file is provided
+% pre-generated.
+%
+% For the most part, the installer is simple: it generates \file{revtex4.cls} and
+% a collection of \file{.rtx} files from \file{revtex4.dtx}.
+%
+% However, the process of generating \file{revtex4.cls} involves the incorporation
+% of portions of two of the \file{ltx} series of files, \file{ltxutil.dtx},
+% and \file{ltxgrid.dtx}. The former is a collection of macro definitions that
+% serve as the lowest level of extensions to \LaTeX, the latter implements
+% an alternative output routine and page gridding mechanism that provides for
+% multicolumn page layouts, but without the sorry limitations of the
+% \LaTeX\ required package, \classname{multicol}, and is also compatible with
+% the \classname{longtable} package (which \classname{multicol} is not).
+% \begin{macrocode}
+%<*ins>
+% This file requires docstrip version 2.4 or higher available from
+% ftp://ctan.tug.org/tex-archive/macros/latex/unpacked/docstrip.tex
+\input docstrip
+\preamble
+
+This file is part of the APS files in the REVTeX 4 distribution.
+For the version number, search on the string %FileInfo
+
+Copyright (c) 2000 The American Physical Society.
+http://publish.aps.org/revtex4/
+mailto:revtex@aps.org
+
+Maintained by Arthur Ogawa (mailto:ogawa@teleport.com)
+under contract to the American Physical Society.
+
+See the REVTeX 4 README file for restrictions and more information.
+
+\endpreamble
+\keepsilent
+\askforoverwritefalse
+\generate{%
+ \file{revtex4.cls}{%
+ \from{revtex4.dtx}{class-pre}%
+ \from{ltxutil.dtx}{ltxutil-krn}%
+ \from{ltxgrid.dtx}{ltxgrid-krn}%
+ \from{textcase.dtx}{package}%
+ \from{revtex4.dtx}{class-pst}%
+ }%
+ \file{aps.rtx}{\from{revtex4.dtx}{aps}}%
+ \file{rmp.rtx}{\from{revtex4.dtx}{rmp}}%
+ \file{10pt.rtx}{\from{revtex4.dtx}{10pt}}%
+ \file{11pt.rtx}{\from{revtex4.dtx}{11pt}}%
+ \file{12pt.rtx}{\from{revtex4.dtx}{12pt}}%
+ \file{revsymb.sty}{\from{revtex4.dtx}{revsymb}}%
+}%
+\endbatchfile
+%</ins>
+% \end{macrocode}
+%
+%
+% \section{Writing journal-specifc extensions to \revtex}
+%
+% With this version of \revtex, we introduce a somewhat different
+% scheme for adapting \revtex\ to the needs of a specific journal.
+%
+% To create a journal substyle, you create new class options in
+% \revtex\ for the society, say \classoption{osa}, and any of that society's journals,
+% one of which is, say, \classoption{josaa},
+% using the code for the APS as a guide.
+% In particular, each of your new options should separately define \cmd\@society\
+% and \cmd\@journal. That for the former will be the same for
+% all options relating to a particular society.
+%
+% Then, for the society, you create a corresponding \file{.rtx}
+% file, in our case \file{osa.rtx}.
+% Within that file, you override procedures and parameter assignments as you see fit.
+% Ideally they will be generally applicable to all of that society's
+% journals (see the file \file{aps.rtx} for a realization of this scheme).
+% Also within that file, you include a
+% section of code for each journal, that for josaa looks like:
+% \begin{verbatim}
+% \@ifx{\@journal\journal@josaa}{%
+% <code specific to the josaa>
+% }{}%
+% \end{verbatim}
+%
+% Thus far, the scheme is similar to that used in \revtex\ 3.1.
+% However, the new scheme does differ from the old in that
+% the \file{.rtx} file should define no syntactical extensions
+% to \revtex.
+%
+% \section{The \classname{revtex4} Document Class}
+%
+% \begin{macrocode}
+%<*class-pre>
+\def\class@name{revtex4}%
+% \end{macrocode}
+%
+% Print a banner in the log:
+% \begin{macrocode}
+\GenericInfo{}{\space
+ Copyright (c) 2001 The American Physical Society.^^J
+ mailto:revtex@aps.org^^J
+ Licensed under the LPPL:^^Jhttp://www.ctan.org/tex-archive/macros/latex/base/lppl.txt^^J
+ Arthur Ogawa <ogawa@teleport.com>^^J
+ Based on work by David Carlisle <david@dcarlisle.demon.co.uk>%
+ \@gobble
+}%
+% \end{macrocode}
+%
+% \subsection{Compatability Processing}
+%
+% If the document has \cmd\documentstyle|{revtex4}|,
+% then, instead of attempting to run in compatability mode,
+% just complain and exit.
+% \changes{v4.0beta 3}{1999/11/13}
+% {Lose compatability mode.}
+% \begin{macrocode}
+\if@compatibility
+ \edef\reserved@a{\errhelp{%
+ Change your \string\documentstyle\space statement to
+ \string\documentclass\space and rerun.
+ }}\reserved@a
+ \errmessage{You cannot run \class@name\space in compatability mode}%
+ \expandafter\@@end
+\fi
+% \end{macrocode}
+%
+%
+% \section{Extensions to the \LaTeX\ Kernel}
+% I am introducing
+% some portions of the \classname{ltxkrnext} package
+% (\url{ftp://ftp.teleport.com/users/ogawa/macros/latex/contrib/supported/ltxkrnext/}).
+%
+% First, de-fang unwanted commands and turn \cmd\ProvidesPackage\ into an informative message.
+% \begin{macrocode}
+\let\DeclareOption@latex\DeclareOption
+\let\DeclareOption\@gobbletwo
+\let\ProcessOptions@latex\ProcessOptions
+\let\ProcessOptions\@empty
+\let\ProvidesPackage@latex\ProvidesPackage
+\def\ProvidesPackage#1[#2]{%
+ \class@info{Incorporating package #1 [#2].}%
+}%
+%</class-pre>
+% \end{macrocode}
+% Now load the utility and page grid packages.
+% The \classname{ltxutil} and \classname{ltxgrid} packages are distributed with
+% \revtex.
+% \begin{macrocode}
+%\RequirePackage{ltxutil}%
+%\RequirePackage{ltxgrid}%
+% \end{macrocode}
+%
+% The \classname{textcase} package is available from
+% \url{CTAN:/macros/latex/contrib/supported/carlisle/textcase.dtx} and is
+% used at the suggestion, and by permission, of David~P. Carlisle.
+% \begin{quotation}
+% \cmd\ProvidesFile{textcase.dtx}
+% [1998/11/12 v0.06 Text only upper/lower case changing (DPC)]
+% \end{quotation}
+%
+% David, in turn, suggests that Donald Arseneau be credited for some of the ideas in
+% the \classname{textcase} package, so let it hereby be so understood.
+%
+% Note that David's package has a bug, which is fixed as follows:
+% Change:
+%\begin{verbatim}
+% \def\NoCaseChange##1{\noexpand\NoCaseChange{\noexpand##1}}
+%\end{verbatim}
+% To:
+%\begin{verbatim}
+% \def\NoCaseChange##1{\noexpand\NoCaseChange{\noexpand##1}}%
+%\end{verbatim}
+%
+% \begin{macrocode}
+%\RequirePackage{textcase}%
+% \end{macrocode}
+%
+% Next, restore those cammands we de-fanged:
+% \begin{macrocode}
+%<*class-pst>
+\let\ProvidesPackage\ProvidesPackage@latex
+\let\ProcessOptions\ProcessOptions@latex
+\let\DeclareOption\DeclareOption@latex
+% \end{macrocode}
+% And let the \classname{textcase} commands override those of \LaTeX:
+% \begin{macrocode}
+\expandafter
+\let\csname MakeUppercase \expandafter\endcsname
+ \csname MakeTextUppercase \endcsname
+\expandafter
+\let\csname MakeLowercase \expandafter\endcsname
+ \csname MakeTextLowercase \endcsname
+% \end{macrocode}
+%
+% Try to patch \classname{longtable}.
+% \begin{macrocode}
+\appdef\class@documenthook{%
+ \switch@longtable
+}%
+% \end{macrocode}
+%
+% Try to provide a more robust version of \cmd\boldmath.
+% If Team \LaTeX\ ever gets around to fixing this problem, we will bow out.
+% (This relates to bug \#394.)
+% \begin{macrocode}
+\appdef\class@documenthook{%
+ \expandafter\@ifnotrelax\csname boldmath \endcsname{}{%
+ \expandafter\let\csname boldmath \endcsname\boldmath
+ \expandafter\def\expandafter\boldmath\expandafter{\expandafter\protect\csname boldmath \endcsname}%
+ }%
+}%
+% \end{macrocode}
+%
+%
+%
+% \section{Options}\label{sec:options}%
+%
+% \subsection{Define Booleans Used in Options}%
+% The following Booleans are used within the document class to
+% allow the document or the substyle to make selections
+% of formatting.
+% Because an explicit document class option always overrides a substyle,
+% any substyle wishing to assign a value to a flag must first test the flag to
+% confirm that it is still undefined.
+% This further means that the default value of the flag must be
+% assigned late, at the end of \cmd\AtBeginDocument\ time, and that
+% the flag may be queried no earlier than that time.
+%
+% \begin{macro}{\class@inithook}
+% The hook \cmd\class@inithook\ is installed into \cmd\class@documenthook,
+% and thereafter serves as the focal point for all code that
+% gives default values to the class's parameters.
+%
+% A typical code might be:
+% \begin{verbatim}
+%\appdef\class@inithook{%
+% \@ifxundefined\twocolumn@sw{\@booleanfalse\twocolumn@sw}{}%
+%}%
+% \end{verbatim}
+% which ensures that the switch \cmd\twocolumn@sw\ is never left undefined.
+%
+% \begin{macrocode}
+\appdef\class@documenthook{%
+ \class@inithook
+}%
+\def\class@inithook{}%
+%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\twocolumn@sw}
+% The boolean \cmd\twocolumn@sw\ signifies that we are to use a two-column grid.
+%
+% \begin{macrocode}
+\appdef\class@inithook{%
+ \@ifxundefined\twocolumn@sw{\@booleanfalse\twocolumn@sw}{}%
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\footinbib@sw}
+% The boolean \cmd\footinbib@sw\ signifies that
+% footnotes are to be set in the bibliography, as endnotes.
+% \begin{macrocode}
+\appdef\class@inithook{%
+ \@ifxundefined\footinbib@sw{\@booleanfalse\footinbib@sw}{}%
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\preprintsty@sw}
+% The boolean \cmd\preprintsty@sw\ signifies that
+% the document is to be formatted in preprint style.
+% \begin{macrocode}
+\appdef\class@inithook{%
+ \@ifxundefined\preprintsty@sw{\@booleanfalse\preprintsty@sw}{}%
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\preprint@sw}
+% The boolean \cmd\preprint@sw\ signifies that
+% the preprints (cf. \cmd\preprint) are to be formatted (usually on the title page).
+% The default is to do so; the option \classoption{nopreprintnumbers}
+% declares not to.
+% \begin{macrocode}
+\appdef\class@inithook{%
+ \@ifxundefined\preprint@sw{\@booleanfalse\preprint@sw}{}%
+ \preprintsty@sw{\@booleantrue\preprint@sw}{}%
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\galley@sw}
+% The boolean \cmd\galley@sw\ signifies that
+% the document is to be formatted in galley style.
+%
+% Asserting both \cmd\galley@sw\ and \cmd\preprintsty@sw\
+% may produce strange formatting results, but it is not illegal.
+% However, it \emph{is} illegal to assert galley and any
+% twocolumn option.
+% \begin{macrocode}
+\appdef\class@inithook{%
+ \@ifxundefined\galley@sw{\@booleanfalse\galley@sw}{}%
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\groupauthors@sw}
+% The flag \cmd\groupauthors@sw\ signifies that authors are to be grouped.
+% This affects the meaning of \cmd\@author@present\ and \cmd\@tempaffil.
+%
+% FIXME: apparently \cmd\@tempaffil\ and its aliases are never used.
+% \end{macro}
+%
+% \begin{macro}{\runinaddress@sw}
+% The flag \cmd\runinaddress@sw\ signifies that author addresses are to be
+% run in.
+% \end{macro}
+%
+% \begin{macro}{\@affils@sw}
+% If \cmd\@affils@sw\ is false, an address is never recognised as ``new''
+% and is therefore always entered into the affiliation list,
+% stopping groups of authors at the same address being amalgamated
+% into the same list:
+% the address will be printed the same number of times it is entered.
+% \end{macro}
+%
+% \begin{macro}{\showPACS@sw}
+% \begin{macro}{\showKEYS@sw}
+% If \cmd\showPACS@sw\ is true, print the PACS information in the title block,
+% otherwise not.
+% Similarly for \cmd\showKEYS@sw\ and the keywords.
+% \begin{macrocode}
+\appdef\class@inithook{%
+ \@ifxundefined\showPACS@sw{\@booleanfalse\showPACS@sw}{}%
+ \@ifxundefined\showKEYS@sw{\@booleanfalse\showKEYS@sw}{}%
+}%
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\floats@sw}
+% The Boolean \cmd\floats@sw\ signifies that floats are to be floated;
+% if false, that floats are to be deferred to the end of the document.
+% By default, the former.
+% Note that the state of this Boolean is to be changed by
+% the document class in response to user-selected options.
+%
+% This boolean and the assignment of its default value is done by
+% the \classname{ltxutil} package.
+% \end{macro}
+%
+% \begin{macro}{\floatp@sw}
+% The Boolean \cmd\floatp@sw\ signifies that endfloats are to be set one per page;
+% if false, that endfloats are to be set with multiple floats per page permitted.
+% By default, the latter.
+% Note that the state of this Boolean is to be changed by
+% the document class in response to user-selected options.
+% The default is established here.
+% \end{macro}
+%
+% \begin{macro}{\twoside@sw}
+% The flag \cmd\twoside@sw\ signifies that the document is to be
+% formatted for duplex printing.
+% At \cmd\AtBeginDocument\ time, we must align the value
+% of the kernel \cmd\newif\ switch \cmd\if@twoside\
+% to that of \cmd\twoside@sw.
+% \begin{macrocode}
+\appdef\class@inithook{%
+ \@ifxundefined\twoside@sw{\@booleanfalse\twoside@sw}{}%
+ \twoside@sw{\@twosidetrue}{\@twosidefalse}%
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\draft@sw}
+% The flag \cmd\draft@sw\ signifies that the document is to be
+% formatted in draft mode.
+% Certain packages may pay attention to the
+% class option \classoption{draft} that sets this Boolean.
+% \begin{macrocode}
+\appdef\class@inithook{%
+ \@ifxundefined\draft@sw{\@booleanfalse\draft@sw}{}%
+ \draft@sw{\overfullrule 5\p@}{\overfullrule\z@}%
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\tightenlines@sw}
+% The boolean \cmd\tightenlines@sw\ signifies that
+% the leading is to be made standard amount.
+% If false, it means that the leading is to be set extra open.
+% Has no effect on 10pt size option.
+% \begin{macrocode}
+\appdef\class@inithook{%
+ \@ifxundefined\tightenlines@sw{\@booleanfalse\tightenlines@sw}{}%
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\lengthcheck@sw}
+% The flag \cmd\lengthcheck@sw\ signifies that
+% the length checking is in effect.
+% It is up to the individual journal substyle
+% to alter its formatting accordingly.
+% \begin{macrocode}
+\appdef\class@inithook{%
+ \@ifxundefined\lengthcheck@sw{\@booleanfalse\lengthcheck@sw}{}%
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\eqsecnum@sw}
+% The flag \cmd\eqsecnum@sw\ signifies that equations are to be
+% numbered with the section, e.g., ``Eq. (2.13)''.
+% This flag is neither set nor queried.
+% \begin{macrocode}
+\appdef\class@inithook{%
+ \@ifxundefined\eqsecnum@sw{\@booleanfalse\eqsecnum@sw}{}%
+ \eqsecnum@sw{%
+ \@addtoreset{equation}{section}%
+ \def\theequation@prefix{\arabic{section}.}%
+ }{}%
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\byrevtex@sw}
+% The flag \cmd\byrevtex@sw\ signifies that the document should
+% bear an imprint to the effect that it was formatted by this document class.
+% \begin{macrocode}
+\appdef\class@inithook{%
+ \@ifxundefined\byrevtex@sw{\@booleanfalse\byrevtex@sw}{}%
+}%
+% \end{macrocode}
+% \end{macro}
+%
+%
+% \subsection{Declare Options}%
+%
+% \subsubsection{Checkin: for Editorial Use}%
+% A document class option declaring that the document is being processed by the
+% editorial staff.
+%
+% This option should:
+% \begin{itemize}
+% \item
+% put date in footer along with folio
+% \item
+% Have the effect of selecting the \classoption{preprint} class option.
+% \item
+% Have the effect of selecting the \classoption{showpacs} class option.
+% \item
+% specify that when a float is placed h or H, it will be allowed to break over pages.
+% (Note: be sure that if the enclosed \env{tabular} has an optional argument, you change it to [v], or remove it entirely.)
+% \end{itemize}
+% \changes{v4.0beta 3}{1999/11/13}
+% {New option ``checkin''}
+% \begin{macrocode}
+\DeclareOption{checkin}{%
+ \@booleantrue\dateinRH@sw
+ \@booleantrue\preprintsty@sw
+ \def\@pointsize{12}%
+ \@booleantrue\showPACS@sw
+ \@booleantrue\showKEYS@sw
+ \def\fp@proc@h{\allow@breaking@tables}%
+ \def\fp@proc@H{\allow@breaking@tables}%
+}%
+\@booleanfalse\dateinRH@sw
+\def\checkindate{\dateinRH@sw{{\tiny(\today)}}{}}%
+\def\allow@breaking@tables{%
+ \def\array@default{v}% tabular can break over pages
+ \@booleanfalse\floats@sw % table can break over pages
+}%
+% \end{macrocode}
+%
+% \subsubsection{Preprint Style}%
+% \changes{v4.d}{1998/01/31}
+% {Rearrange the ordering so numerical ones come first. AO: David, what does this mean?}
+%
+% \changes{v4.g}{1998/06/10}
+% {comma not space between email and homepage}
+%
+% \changes{v4.0beta 2}{1999/06/20}
+% {Move this ``complex'' option to the front, where it can be overridden by ``simple'' options.}
+% \begin{macrocode}
+\DeclareOption{preprint}{%
+ \@booleantrue\preprintsty@sw
+ \def\@pointsize{12}%
+}%
+\DeclareOption{manuscript}{%
+ \class@warn{Document class option manuscript is obsolete; use preprint instead}%
+ \ExecuteOptions{preprint}%
+}%
+% \end{macrocode}
+%
+% \subsubsection{Showing PACS and keywords}%
+% \changes{v4.0beta 2}{1999/06/20}
+% {New option}
+% \changes{v4.0beta 5}{2000/09/20}
+% {New option showkeys}
+% \begin{macrocode}
+\DeclareOption{showpacs}{%
+ \@booleantrue\showPACS@sw
+}%
+\DeclareOption{noshowpacs}{%
+ \@booleanfalse\showPACS@sw
+}%
+\DeclareOption{showkeys}{%
+ \@booleantrue\showKEYS@sw
+}%
+\DeclareOption{noshowkeys}{%
+ \@booleanfalse\showKEYS@sw
+}%
+% \end{macrocode}
+%
+% \subsubsection{Balance the last page when in two-column page grid}%
+% If we are in a two-column page grid, we may wish to balance the columns
+% of the last page. This will be done automatically if the
+% \classoption{twocolumn} document class option is chosen.
+% This action will be turned off by the \classoption{nobalancelastpage}.
+% A complementary class option, \classoption{balancelastpage} is also provided.
+% \begin{macrocode}
+\DeclareOption{balancelastpage}{%
+ \@booleantrue\balancelastpage@sw
+}%
+\DeclareOption{nobalancelastpage}{%
+ \@booleanfalse\balancelastpage@sw
+}%
+\appdef\class@inithook{%
+ \@ifxundefined\balancelastpage@sw{%
+ \@booleantrue\balancelastpage@sw
+ }{}%
+}%
+% \end{macrocode}
+%
+% \subsubsection{Showing preprint numbers}%
+% \changes{v4.0beta 4}{2000/04/10}
+% {New option}
+% \begin{macrocode}
+\DeclareOption{nopreprintnumbers}{\@booleanfalse\preprint@sw}%
+\DeclareOption{preprintnumbers}{\@booleantrue\preprint@sw}%
+% \end{macrocode}
+%
+% \subsubsection{Hypertext Option}%
+%
+% \changes{v4.e}{1998/01/31}
+% {protect against hyperref revtex kludges which are not needed now}
+% \changes{v4.0beta 2}{1999/06/20}
+% {Move this ``complex'' option to the front, where it can be overridden by ``simple'' options.}
+% \changes{v4.0rc4}{2001/07/23}
+% {hyperref is no longer loaded via class option: use a usepackage statement instead}
+% \begin{macrocode}
+\DeclareOption{hyperref}{%
+%\AtEndOfClass{%
+% \begingroup
+% \edef\@tempa{%
+% \let
+% \noexpand\@clsextension
+% \noexpand\@empty
+% \noexpand\RequirePackage{hyperref}%
+% \def\noexpand
+% \@clsextension{\@clsextension}%
+% }%
+% \expandafter
+% \endgroup
+% \@tempa
+%}%
+% \end{macrocode}
+%
+% If you have a hyper-foo enabled browser you may prefer this format
+% which does not print the URL for the home page, but just makes the
+% name a link, but by default print it so it works on paper.
+% \changes{v4.e}{1998/01/31}
+% {print homepage}
+%
+% \begin{macrocode}
+% \def\@pointsize{10}%
+ \class@warn{Class option "hyperref" is no longer supported.^^JEmploy \string\usepackage{hyperref} instead}%
+}%
+% \end{macrocode}
+%
+% \subsubsection{Type Size}%
+% Use \cmd\@pointsize|=10| rather than \cmd\@ptsize|=0| to allow easy extensions
+% to 9pt or whatever.
+% Note: the three alternatives are mutually exclusive. Furthermore
+% if \cmd\@pointsize\ is still undefined at \cmd\AtEndOfClass\ time,
+% then the substyle can conclude that the user accepts the default.
+% \begin{macrocode}
+\DeclareOption{10pt}{\def\@pointsize{10}}
+\DeclareOption{11pt}{\def\@pointsize{11}}
+\DeclareOption{12pt}{\def\@pointsize{12}}
+% \end{macrocode}
+%
+% \subsubsection{Media Size}%
+% \changes{v4.0beta 2}{1999/06/20}
+% {Restore all media size class option of \protect\file{classes.dtx}}
+% \begin{macrocode}
+\DeclareOption{a4paper}
+ {\setlength\paperheight {297mm}%
+ \setlength\paperwidth {210mm}}
+\DeclareOption{a5paper}
+ {\setlength\paperheight {210mm}%
+ \setlength\paperwidth {148mm}}
+\DeclareOption{b5paper}
+ {\setlength\paperheight {250mm}%
+ \setlength\paperwidth {176mm}}
+\DeclareOption{letterpaper}
+ {\setlength\paperheight {11in}%
+ \setlength\paperwidth {8.5in}}
+\DeclareOption{legalpaper}
+ {\setlength\paperheight {14in}%
+ \setlength\paperwidth {8.5in}}
+\DeclareOption{executivepaper}
+ {\setlength\paperheight {10.5in}%
+ \setlength\paperwidth {7.25in}}
+\DeclareOption{landscape}
+ {\setlength\@tempdima {\paperheight}%
+ \setlength\paperheight {\paperwidth}%
+ \setlength\paperwidth {\@tempdima}}
+% \end{macrocode}
+%
+% \subsubsection{Bibnotes}%
+% Frontmatter footnotes result from frontmatter commands like
+% \cmd\email, \cmd\homepage, \cmd\altaffiliation, and \cmd\thanks.
+% The default for \cmd\frontmatter@footnote@produce\ is
+% \cmd\frontmatter@footnote@produce@footnote, which formats the frontmatter footnotes
+% at the foot of the title page. The \classoption{bibnotes} class option
+% defers them to the bibliography.
+% \begin{macrocode}
+\DeclareOption{bibnotes}{\@booleantrue\bibnotes@sw}%
+\DeclareOption{nobibnotes}{\@booleanfalse\bibnotes@sw}%
+\appdef\class@enddocumenthook{\auto@bib}%
+\appdef\class@inithook{%
+ \@ifxundefined\bibnotes@sw{\@booleanfalse\bibnotes@sw}{}%
+ \bibnotes@sw{%
+ \let\frontmatter@footnote@produce\frontmatter@footnote@produce@endnote
+ }{%
+ \let\frontmatter@footnote@produce\frontmatter@footnote@produce@footnote
+ }%
+}%
+% \end{macrocode}
+%
+% \subsubsection{Footinbib}%
+% \begin{macrocode}
+\DeclareOption{footinbib}{\@booleantrue\footinbib@sw}
+\DeclareOption{nofootinbib}{\@booleanfalse\footinbib@sw}
+% \end{macrocode}
+%
+% \subsubsection{altaffilletter}%
+%
+% \begin{macrocode}
+\DeclareOption{altaffilletter}{\@booleantrue\altaffilletter@sw}%
+\DeclareOption{altaffilsymbol}{\@booleanfalse\altaffilletter@sw}%
+\appdef\class@inithook{%
+ \@ifxundefined\altaffilletter@sw{\@booleanfalse\altaffilletter@sw}{}%
+}%
+% \end{macrocode}
+%
+% \subsubsection{superbib}%
+% The document may invoke superscript bibliography style.
+% If not, the journal substyle may make a selection.
+% If neither, the class itself selects a default.
+% This is really a quite deep nesting of defaults.
+% \begin{macrocode}
+\DeclareOption{superbib}{%
+ \def\place@bibnumber{\place@bibnumber@sup}%
+}%
+% \end{macrocode}
+%
+% \subsubsection{citeautoscript}%
+% This class option allows you to automatically accomodate a change from non-superscripted, numbered references
+% to superscripted, bunmbered references.
+%
+% Note: you should always mark up your document with the assumption that references are \emph{not} going to be superscripted.
+% Otherwise this option has no hope of working properly.
+%
+% \begin{macrocode}
+\DeclareOption{citeautoscript}{%
+ \@booleantrue\citeautoscript@sw
+}%
+\@booleanfalse\citeautoscript@sw
+% \end{macrocode}
+%
+% \subsubsection{Simplex/Duplex Pages}%
+% \changes{v4.e}{1998/01/31}
+% {4.d had twoside option setting twoside switch to false}
+% The complementary options \classoption{twoside} and \classoption{oneside}
+% assert formatting for duplex or simplex printing, respectively.
+% At the same time, we arrange for the selection of the page grid
+% with respect to the marginal column:
+% Because \cmd\if@reversemargin\ remains default (false),
+% if duplex printing, this column will always be on the (right),
+% if simplex printing, it will always be on the (outside).
+% QUERY: correct choice?
+% \begin{macrocode}
+\DeclareOption{twoside}{\@booleantrue\twoside@sw \@mparswitchfalse}
+\DeclareOption{oneside}{\@booleanfalse\twoside@sw\@mparswitchtrue}
+% \end{macrocode}
+%
+% \subsubsection{Two-Column Page Grid}%
+%
+% \begin{macro}{\twocolumn@sw}
+% The flag \cmd\twocolumn@sw\ signifies that the document is to be
+% formatted in the two-column page grid.
+%
+% If no options relating to page grid are invoked by
+% \cmd\AtBeginDcoument\ time, we set default values.
+% Up to that point, the class can check if \cmd\twocolumn@sw\
+% is \cmd\undefined\ to see if any related options have been invoked.
+%
+% At \cmd\AtBeginDocument\ time, we must align the value
+% of the kernel \cmd\newif\ switch \cmd\if@twocolumn\
+% to that of \cmd\twocolumn@sw.
+%
+% We install into \cmd\class@enddocumenthook\
+% all tasks that we know must be completed at \cmd\AtEndDocument\ time.
+% The first thing it does is end the current paragraph.
+% The last is
+% to return us to one-column page grid.
+% (This task is installed as late as possible,
+% i.e., after items like \cmd\printtables.)
+% \begin{macrocode}
+\AtEndOfClass{%
+ \def\clear@document{%
+ \close@column
+ }%
+ \appdef\class@inithook{%
+ \@ifxundefined\close@column{%
+ \let\close@column\close@column@default
+ }{}%
+ }%
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% The \classoption{twocolumn} class option employs the standard mechanism for
+% changing the column grid: the \classname{ltxgrid} package.
+% \begin{macrocode}
+\DeclareOption{twocolumn}{%
+ \@booleantrue\twocolumn@sw
+}%
+% \end{macrocode}
+%
+% \classoption{onecolumn} specifies one-column page grid.
+% \begin{macrocode}
+\DeclareOption{onecolumn}{%
+ \@booleanfalse\twocolumn@sw
+}%
+% \end{macrocode}
+%
+% \classoption{galley} emulates setting the galleys of
+% a two-column journal.
+% FIXME: this option should effectively set \cmd\preprintsty@sw\ false.
+% NOTE: it makes no sense to assert both galley and twocolumn.
+%
+% \begin{macrocode}
+\DeclareOption{galley}{%
+ \ExecuteOptions{onecolumn}%
+ \@booleantrue\galley@sw
+ \appdef\set@pica@hook{%
+ \advance\textwidth-\columnsep
+ \textwidth.5\textwidth
+ }%
+}%
+% \end{macrocode}
+%
+%
+% \subsubsection{raggedbottom or flushbottom}%
+% The class options \classoption{raggedbottom} and \classoption{flushbottom} determine whether the columns (page)
+% are ragged bottom or flush bottom.
+% Note that we do not select a default here; that is done by the journal substyle.
+% \begin{macrocode}
+\DeclareOption{raggedbottom}{\@booleantrue\raggedcolumn@sw}
+\DeclareOption{flushbottom}{\@booleanfalse\raggedcolumn@sw}
+% \end{macrocode}
+%
+% \subsubsection{tightenlines}%
+% This class option specifies that standard leading
+% is to be used to set the type.
+% If lacking, the leading will be loose.
+%
+% \begin{macrocode}
+\DeclareOption{tightenlines}{\@booleantrue\tightenlines@sw}
+% \end{macrocode}
+%
+% \subsubsection{lengthcheck}%
+% This class option specifies that the formatted document
+% should approach as closely as possible the
+% formatting of an actual journal article
+% to facilitate the author's performance of a length check.
+%
+% \begin{macrocode}
+\DeclareOption{lengthcheck}{%
+ \@booleantrue\lengthcheck@sw
+ \@booleantrue\twoside@sw
+ \@booleantrue\twocolumn@sw
+}%
+% \end{macrocode}
+%
+% \subsubsection{Draft and Final}%
+% \begin{macrocode}
+\DeclareOption{draft}{%
+ \@booleantrue\draft@sw
+}%
+\DeclareOption{final}{%
+ \@booleanfalse\draft@sw
+}%
+% \end{macrocode}
+%
+% \subsubsection{eqsecnum}%
+% The \classoption{eqsecnum} class option signifies that equations are
+% to be numbered within sections.
+% \begin{macrocode}
+\DeclareOption{eqsecnum}{\@booleantrue\eqsecnum@sw}
+% \end{macrocode}
+%
+% \subsubsection{secnumarabic}%
+% The \classoption{secnumarabic} class option signifies that sectioning commands are
+% to be numbered arabic.
+% \begin{macrocode}
+\DeclareOption{secnumarabic}{\@booleantrue\secnumarabic@sw}
+% \end{macrocode}
+% The code that polls the Boolean \cmd\secnumarabic@sw\ appears
+% in Section~\ref{sec:secnumarabic}.
+%
+% FIXME: model fleqn after amsfonts. I no longer understand why I said this.
+% \begin{macrocode}
+\DeclareOption{fleqn}{\input{fleqn.clo}}
+% \end{macrocode}
+%
+% \subsubsection{floats/endfloats}%
+% These options control, via the Boolean \cmd\floats@sw,
+% whether floats are to be migrated to the end of the document.
+% \begin{macrocode}
+\DeclareOption{floats}{\@booleantrue\floats@sw\@booleanfalse\floatp@sw}
+\DeclareOption{endfloats}{\@booleanfalse\floats@sw\@booleanfalse\floatp@sw}
+\DeclareOption{endfloats*}{\@booleanfalse\floats@sw\@booleantrue\floatp@sw}
+\appdef\class@inithook{%
+ \@ifxundefined\floats@sw{\@booleantrue\floats@sw}{}%
+ \@ifxundefined\floatp@sw{\@booleantrue\floatp@sw}{}%
+}%
+% \end{macrocode}
+%
+% \subsubsection{titlepage/notitlepage}%
+% These options control, via \cmd\titlepage@sw,
+% whether the title block is to be set on a separate page.
+% \begin{macro}{\titlepage@sw}
+% The flag \cmd\titlepage@sw\ signifies that a forced
+% page break is to follow the title page: the article
+% title appears on a page by itself.
+%
+% The Boolean \cmd\titlepage@sw\ is in an odd situation:
+% its value is set by one or more substyles, contingent
+% upon the value of another Boolean.
+% Therefore that code must be invoked at \cmd\AtBeginDocument\ time.
+% Therefore we must assign the default value at the very end
+% of \cmd\AtBeginDocument\ processing: we use \cmd\defaults@hook.
+% Luckily, this Boolean does not need to be queried at
+% \cmd\AtBeginDocument\ time, or we'd be in dire straights.
+% \begin{macrocode}
+\DeclareOption{titlepage}{\@booleantrue\titlepage@sw}
+\DeclareOption{notitlepage}{\@booleanfalse\titlepage@sw}
+\appdef\defaults@hook{%
+ \@ifxundefined\titlepage@sw{\@booleanfalse\titlepage@sw}{}%
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \subsubsection{Substyle and Sub-substyle}%
+%
+% If the society has already been assigned,
+% notify user that \cmd\@society\ is being overridden.
+%
+% \begin{macrocode}
+\def\change@society#1{%
+ \def\@tempa{#1}%
+ \@ifxundefined\@society{}{%
+ \@ifx{\@tempa\@society}{}{%
+ \class@warn{Conflicting society \@tempa<>\@society.}%
+ }%
+ }%
+ \let\@society\@tempa
+}%
+% \end{macrocode}
+% Here are the class options relating to the APS
+% \begin{macrocode}
+\DeclareOption{aps}{\change@society{aps}\def\@journal{}}%
+\DeclareOption{pra}{\change@society{aps}\def\@journal{pra}}%
+\DeclareOption{prb}{\change@society{aps}\def\@journal{prb}}%
+\DeclareOption{prc}{\change@society{aps}\def\@journal{prc}}%
+\DeclareOption{prd}{\change@society{aps}\def\@journal{prd}}%
+\DeclareOption{pre}{\change@society{aps}\def\@journal{pre}}%
+\DeclareOption{prl}{\change@society{aps}\def\@journal{prl}}%
+\DeclareOption{prstab}{\change@society{aps}\def\@journal{prstab}}%
+\DeclareOption{rmp}{\change@society{aps}\def\@journal{rmp}}%
+% \end{macrocode}
+% Here are the class options relating to the Optical Society of America
+% \begin{macrocode}
+\DeclareOption{osa}{\change@society{osa}\def\@journal{}}%
+\DeclareOption{osameet}{\change@society{osa}\def\@journal{osameet}}%
+\DeclareOption{opex}{\change@society{osa}\def\@journal{opex}}%
+\DeclareOption{tops}{\change@society{osa}\def\@journal{tops}}%
+\DeclareOption{josa}{\change@society{osa}\def\@journal{josa}}%
+% \end{macrocode}
+%
+% NOTE: \file{revguide.tex} does not mention any \cmd\@journal\ options
+% besides \classoption{prb}.
+%
+% \begin{macro}{amsfonts}
+% The class option \classoption{amsfonts} has the
+% same effect as if the document preamble contained
+% a \cmd\usepackage\verb${amsfonts}$ statement.
+% \begin{macrocode}
+\DeclareOption{amsfonts}{%
+ \def\class@amsfonts{\RequirePackage{amsfonts}}%
+}%
+\DeclareOption{noamsfonts}{%
+ \let\class@amsfonts\@empty
+}%
+\AtEndOfClass{%
+ \@ifxundefined\class@amsfonts{}{\class@amsfonts}%
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{amssymb}
+% The class option \classoption{amssymb} has the
+% same effect as if the document preamble contained
+% a \cmd\usepackage\verb${amssymb}$ statement.
+% \begin{macrocode}
+\DeclareOption{amssymb}{%
+ \def\class@amssymb{\RequirePackage{amssymb}}%
+}%
+\DeclareOption{noamssymb}{%
+ \let\class@amssymb\@empty
+}%
+\AtEndOfClass{%
+ \@ifxundefined\class@amssymb{}{\class@amssymb}%
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{amsmath}
+% The class option \classoption{amsmath} has the
+% same effect as if the document preamble contained
+% a \cmd\usepackage\verb${amsmath}$ statement.
+%
+% We require version 1.2 (datestamped 1997/03/20) or later.
+% The \expandafter\cmd\csname ver@amsmath.sty\endcsname, will
+%
+% \LaTeX\ note:
+% Certain \LaTeX\ procedures have an arbitrary and pointless restriction that they may be used only within the preamble.
+% We get around this by preserving the procedures in private \cmd\csname s.
+%
+% FIXME note: it is difficult to ensure that an error summary will be printed on the console at the very end,
+% but \classname{ltxgrid} allows accomplishing this via an interrupt, put down at \cmd\AtEndDocument\ time.
+%
+% \begin{macrocode}
+\DeclareOption{amsmath}{%
+ \def\class@amsmath{\RequirePackage{amsmath}[\ver@amsmath@prefer]}%
+}%
+\DeclareOption{noamsmath}{%
+ \let\class@amsmath\@empty
+}%
+\AtEndOfClass{%
+ \preserve@LaTeX
+ \@ifxundefined\class@amsmath{}{\class@amsmath}%
+ \appdef\class@enddocumenthook{\test@amsmath@ver}%
+}%
+\appdef\preserve@LaTeX{%
+ \let\@ifl@aded@LaTeX\@ifl@aded
+ \let\@ifpackageloaded@LaTeX\@ifpackageloaded
+ \let\@pkgextension@LaTeX\@pkgextension
+ \let\@ifpackagelater@LaTeX\@ifpackagelater
+ \let\@ifl@ter@LaTeX\@ifl@ter
+ \let\@ifl@t@r@LaTeX\@ifl@t@r
+ \let\@parse@version@LaTeX\@parse@version
+}%
+\appdef\restore@LaTeX{%
+ \let\@ifl@aded\@ifl@aded@LaTeX
+ \let\@ifpackageloaded\@ifpackageloaded@LaTeX
+ \let\@pkgextension\@pkgextension@LaTeX
+ \let\@ifpackagelater\@ifpackagelater@LaTeX
+ \let\@ifl@ter\@ifl@ter@LaTeX
+ \let\@ifl@t@r\@ifl@t@r@LaTeX
+ \let\@parse@version\@parse@version@LaTeX
+}%
+\def\test@amsmath@ver{%
+ \begingroup
+ \restore@LaTeX
+ \@ifpackageloaded{amsmath}{%
+ \@ifpackagelater{amsmath}{\ver@amsmath@prefer}{}{%
+ \class@warn{%
+ You have loaded amsmath, version "\csname ver@amsmath.sty\endcsname",\MessageBreak
+ but this class requires version "\ver@amsmath@prefer", or later.\MessageBreak
+ Please update your LaTeX installation.
+ }%
+ }%
+ }{%
+ }%
+ \endgroup
+}%
+\def\ver@amsmath@prefer{2000/01/15 v2.05 AMS math features}%
+% \end{macrocode}
+% \end{macro}
+%
+% \subsubsection{Presenting Authors and Their Affiliations}%
+% \changes{v4.0beta 2}{1999/06/20}
+% {These options are now mutually exclusive. But is this correct?}
+%
+% The \classoption{runinaddress} option is the only one that sets \cmd\runinaddress@sw\ to true.
+%
+% The \classoption{unsortedaddress} option is the only one that sets \cmd\@affils@sw\ to false.
+%
+% If we take the \classoption{groupedaddress} option as the default, then we can look upon
+% the \classoption{superscriptaddress} option as simply turning \cmd\groupauthors@sw\ to false.
+%
+% \classoption{groupedaddress}, the default, groups authors above a common address.
+% In order to assert the default and still provide for proper precedence
+% for document-supplied options, we carry out our action at \cmd\AtBeginDocument\ time,
+% and only if none of the related class options were invoked.
+% \begin{macrocode}
+\DeclareOption{groupedaddress}{\clo@groupedaddress}%
+\def\clo@groupedaddress{%
+ \@booleantrue\groupauthors@sw
+ \@booleantrue\@affils@sw
+ \@booleanfalse\runinaddress@sw
+}%
+\appdef\class@inithook{%
+ \@ifxundefined\groupauthors@sw{\clo@groupedaddress}{}%
+}%
+% \end{macrocode}
+%
+% \classoption{unsortedaddress} is similar to \classoption{groupedaddress},
+% but turns off \cmd\@affils@sw,
+% with the result that each address that is entered will be printed.
+% \begin{macrocode}
+\DeclareOption{unsortedaddress}{\clo@unsortedaddress}%
+\def\clo@unsortedaddress{%
+ \@booleantrue\groupauthors@sw
+ \@booleanfalse\@affils@sw
+ \@booleanfalse\runinaddress@sw
+}%
+% \end{macrocode}
+%
+% \classoption{runinaddress} is similar to \classoption{groupedaddress},
+% with the addition of the Boolean \cmd\runinaddress@sw,
+% which causes the authors associated with a particular address group
+% to be formatted in a paragraph instead of on separate lines.
+% \begin{macrocode}
+\DeclareOption{runinaddress}{\clo@runinaddress}%
+\def\clo@runinaddress{%
+ \@booleantrue\groupauthors@sw
+ \@booleantrue\@affils@sw
+ \@booleantrue\runinaddress@sw
+}%
+% \end{macrocode}
+%
+% \classoption{superscriptaddress} presents author affiliations
+% as superscripts. Authors with like affiliations share the same
+% superscript.
+%
+% FIXME: if \cmd\groupauthors@sw\ is false, would not \cmd\@affils@sw\ be a don't care?
+% \begin{macrocode}
+\DeclareOption{superscriptaddress}{\clo@superscriptaddress}%
+\def\clo@superscriptaddress{%
+ \@booleanfalse\groupauthors@sw
+ \@booleantrue\@affils@sw
+ \@booleanfalse\runinaddress@sw
+}%
+% \end{macrocode}
+%
+%
+% \subsubsection{Typeset by \revtex}%
+%
+% The class option \classoption{byrevtex} signifies that you want the
+% ``Typeset by \revtex'' byline to appear on your formatted output.
+% By default, no such byline appears.
+% \begin{macrocode}
+\DeclareOption{byrevtex}{\@booleantrue\byrevtex@sw}%
+% \end{macrocode}
+%
+%
+% \subsection{Attempt to fix float placement failure}%
+%
+% \revtex\ uses the \classname{ltxgrid} package, which provides the ability to attempt repairs when
+% \LaTeX's float placement mechanism is about to fail, but that facility is turned off by default.
+% Users should invoke the \classoption{floatfix} document class option to enable this \LaTeX\ extension.
+% If not, a helpful message is printed in the log, indicating how to work around the difficulty.
+% \begin{macrocode}
+\DeclareOption{floatfix}{\@booleantrue\force@deferlist@sw}%
+\DeclareOption{nofloatfix}{\@booleanfalse\force@deferlist@sw}%
+\appdef\class@inithook{%
+ \@ifxundefined\force@deferlist@sw{\@booleanfalse\force@deferlist@sw}{}%
+}%
+% \end{macrocode}
+%
+% \begin{macro}{\@fltovf}
+% \begin{macro}{\@fltstk}
+% The \LaTeX\ kernel error message \cmd\@fltovf\ may now be a bit more helpful to the user;
+% likewise for the \cmd\@fltstk\ message of \classname{ltxgrid}.
+% \begin{macrocode}
+\gdef\@fltovf{%
+ \@latex@error{%
+ Too many unprocessed floats%
+ \force@deferlist@sw{}{; try class option [floatfix]}%
+ }\@ehb
+}%
+\def\@fltstk{%
+ \@latex@warning{%
+ A float is stuck (cannot be placed)%
+ \force@deferlist@sw{}{; try class option [floatfix]}%
+ }%
+}%
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\ltxgrid@info@sw}
+% \begin{macro}{\outputdebug@sw}
+% The two options \classoption{ltxgridinfo} and \classoption{outputdebug} turn on informative diagnostics within the
+% package \classname{ltxgrid}. Only people who really want to see this output will select these class options.
+% Consult documentation for the \classname{ltxgrid} package to see what output the related switches enable.
+% \begin{macrocode}
+\DeclareOption{ltxgridinfo}{\@booleantrue\ltxgrid@info@sw}%
+\DeclareOption{outputdebug}{\@booleantrue\outputdebug@sw}%
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \subsection{Option to relax page height}%
+%
+% The \classname{ltxgrid} package can set text pages to their natural height or force them to the full
+% text height; the latter is the default.
+% If setting the pages with a variable length,
+% the running foot will move up or down with the natural length of the text column.
+% While I recommend against doing so, this option will turn that switch to the latter setting.
+% \begin{macrocode}
+\DeclareOption{raggedfooter}{\@booleanfalse\textheight@sw}%
+\DeclareOption{noraggedfooter}{\@booleantrue\textheight@sw}%
+% \end{macrocode}
+%
+% \subsection{Selecting procedure for processing abstract}%
+%
+% There are two different procedures for processing an abstract; the options
+% \classoption{newabstract} and \classoption{oldabstract} determine which one the user
+% prefers. At present, the default is to use the old, but this may change.
+% \begin{macrocode}
+\DeclareOption{newabstract}{\@booleanfalse\oldabstract@sw}%
+\DeclareOption{oldabstract}{\@booleantrue \oldabstract@sw}%
+\appdef\class@inithook{%
+ \@ifxundefined\oldabstract@sw{\@booleanfalse\oldabstract@sw}{}%
+ \oldabstract@sw{%
+ \let@environment{abstract}{abstract-old}%
+ }{}%
+}%
+% \end{macrocode}
+%
+%
+%
+%
+% \subsection{Default Option}%
+%
+% The default handling for a document class option is to
+% treat it as the name of a society and
+% hunt for a corresponding \file{.rtx} file to read in.
+% This behavior is much like the \LaTeX2.09 handling, where
+% one looked for a \file{.sty} file, except that
+% in this case, we must catch the case where multiple
+% unknown options are declared.
+%
+% \begin{macrocode}
+\DeclareOption*{%
+ \IfFileExists{\CurrentOption\substyle@ext}{%
+ \expandafter\change@society\expandafter{\CurrentOption}%
+ }{%
+ \OptionNotUsed
+ }%
+}%
+\def\substyle@ext{.rtx}%
+\AtEndOfClass{%
+ \@ifxundefined\@society{%
+ \class@warn{No Society specified, using generic APS style}%
+ \def\@society{aps}\def\@journal{}%
+ }{}%
+ \expandafter\input\expandafter{\@society\substyle@ext}%
+}%
+% \end{macrocode}
+%
+% \subsection{Class-Asserted Options}%
+% \changes{v4.0beta 2}{1999/06/20}
+% {%
+% Removed invocation of nonexistent class option \protect\classoption{groupauthors}
+% and all other class options that should only be invoked by the document.
+% (Otherwise precedence of class options does not work.)
+% }
+% Here we establish the default document class options.
+% Those of the document itself will override these.
+% \begin{macrocode}
+\ExecuteOptions{10pt}
+% \end{macrocode}
+%
+% \subsection{Execute Options}%
+% The document class options are processed in the order
+% they are declared in this document class file.
+%
+% Precedence is established in the following way:
+%
+% First, the single \file{.rtx} file is read in.
+% It may set values to any of the Booleans that represent class options,
+% and it may elect to do so only if that Boolean is undefined, signifying
+% that its corresponding document class options were absent from
+% the document's option list.
+% In this way correct precedence of class options is maintained.
+%
+% Second, the type size option is acted upon. Type size code
+% should not require altering the values of any of the Booleans carrying
+% class options.
+%
+% Next, at \cmd\AtEndOfClass\ time, certain packages are read in.
+%
+% Finally, at \cmd\AtBeginDocument\ time, the state of the Booleans determines
+% how the formatting will go.
+%
+% \begin{macrocode}
+\ProcessOptions\relax
+% \end{macrocode}
+%
+%
+% \section{Procedures Dependent Upon Options}
+%
+% Here we introduce \classname{classes.dtx} definitions for
+% the page styles that people will expect to be able to use.
+% \begin{macro}{\ps@headings}
+% \begin{macro}{\ps@myheadings}
+% \begin{macrocode}
+ \def\ps@headings{%
+ \let\@oddfoot\@empty\let\@evenfoot\@empty
+ \def\@evenhead{\thepage\hfil\slshape\leftmark}%
+ \def\@oddhead{{\slshape\rightmark}\hfil\thepage}%
+ \let\@mkboth\markboth
+ \def\sectionmark##1{%
+ \markboth {\MakeUppercase{%
+ \ifnum \c@secnumdepth >\z@
+ \thesection\quad
+ \fi
+ ##1}}{}}%
+ \def\subsectionmark##1{%
+ \markright {%
+ \ifnum \c@secnumdepth >\@ne
+ \thesubsection\quad
+ \fi
+ ##1}}}%
+\def\ps@myheadings{%
+ \let\@oddfoot\@empty\let\@evenfoot\@empty
+ \def\@evenhead{\thepage\hfil\slshape\leftmark}%
+ \def\@oddhead{{\slshape\rightmark}\hfil\thepage}%
+ \let\@mkboth\@gobbletwo
+ \let\sectionmark\@gobble
+ \let\subsectionmark\@gobble
+ }%
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\ps@article@twoside}
+% \begin{macro}{\ps@article@oneside}
+% \begin{macro}{\ps@preprint}
+% \changes{v4.e}{1998/01/31}
+% {Move after process options, so the following test works}
+% \changes{v4.0beta 2}{1999/06/20}
+% {Defer decision until \cs{AtBeginDocument} time}
+% \changes{v4.0beta 3}{1999/11/13}
+% {Insert procedure \cs{checkindate}}
+% \changes{v4.0rc1}{2001/06/17}
+% {Running headers always as if two-sided}
+% \begin{macrocode}
+\def\ps@article{%
+ \@provide\@evenhead{\let\\\heading@cr\thepage\quad\checkindate\hfil{\leftmark}}%
+ \@provide\@oddhead{\let\\\heading@cr{\rightmark}\hfil\checkindate\quad\thepage}%
+ \@provide\@oddfoot{}%
+ \@provide\@evenfoot{}%
+ \let\@mkboth\markboth
+ \let\sectionmark\@gobble
+ \let\subsectionmark\@gobble
+}%
+\def\ps@article@final{%
+ \@provide\@evenhead{\let\\\heading@cr\thepage\quad\checkindate\hfil{\leftmark}}%
+ \@provide\@oddhead{\let\\\heading@cr{\rightmark}\hfil\checkindate\quad\thepage}%
+ \@provide\@oddfoot{}%
+ \@provide\@evenfoot{}%
+ \let\@mkboth\markboth
+ \def\sectionmark##1{%
+ \markboth{%
+ \MakeTextUppercase{%
+ \@ifnum{\c@secnumdepth >\z@}{\thesection\hskip 1em\relax}{}%
+ ##1%
+ }%
+ }{}%
+ }%
+ \def\subsectionmark##1{%
+ \markright {%
+ \@ifnum{\c@secnumdepth >\@ne}{\thesubsection\hskip 1em\relax}{}%
+ ##1%
+ }%
+ }%
+}%
+\appdef\class@inithook{%
+%\twoside@sw{%
+%}{%
+%}%
+}%
+% \end{macrocode}
+%
+% \begin{macrocode}
+\def\heading@cr{\unskip\space\ignorespaces}%
+% \end{macrocode}
+%
+% \changes{v4.0beta 3}{1999/11/13}
+% {Do not put by REVTeX in every page foot}
+% \changes{v4.0beta 3}{1999/11/13}
+% {Insert procedure \cs{checkindate}}
+% \begin{macrocode}
+\def\ps@preprint{%
+ \@provide\@oddfoot{\hfil\thepage\quad\checkindate\hfil}%
+ \@provide\@evenfoot{\hfil\thepage\quad\checkindate\hfil}%
+ \@provide\@oddhead{}%
+ \@provide\@evenhead{}%
+ \let\@mkboth\@gobbletwo
+ \let\sectionmark\@gobble
+ \let\subsectionmark\@gobble
+}%
+% \end{macrocode}
+%
+% \begin{macrocode}
+\let\@oddhead\@undefined
+\let\@evenhead\@undefined
+\let\@oddfoot\@undefined
+\let\@evenfoot\@undefined
+\AtEndOfClass{%
+ \appdef\class@documenthook{%
+ \@provide\@oddhead{}%
+ \@provide\@evenhead{}%
+ \@provide\@oddfoot{}%
+ \@provide\@evenfoot{}%
+ }%
+}%
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\lastpage@putlabel}
+% Support the default meaning of \cmd\@endpage.
+% \changes{v4.d}{1998/01/31}
+% {Move after process options, so \cs{clearpage} not in scope of
+% twocolumn}
+% Name of this macro (and the \cmd\label\ key) taken from
+% \url{CTAN:/macros/latex/contrib/other/lastpage}%
+% with code optimised slightly.
+% \begin{macrocode}
+\def\lastpage@putlabel{%
+ \if@filesw
+ \begingroup
+ \advance\c@page\m@ne
+ \immediate\write\@auxout{\string\newlabel{LastPage}{{}{\thepage}}}%
+ \endgroup
+ \fi
+}%
+% \end{macrocode}
+% Install a late hook into \cmd\class@enddocumenthook\ that
+% labels the last page of the document.
+% This is done just before the \file{.aux} file
+% is closed, and does not require a \cmd\shipout, because it writes
+% directly to the \file{.aux} file.
+% Note that we invoke \cmd\clearpage: this means that no more typesetting
+% should be done past this point.
+% CHECKME: are there any more things installed via
+% \cmd\AtEndOfClass\allowbreak\cmd\class@enddocumenthook?
+% \begin{macrocode}
+\AtEndOfClass{%
+ \appdef\clear@document{%
+ \clearpage
+ \lastpage@putlabel
+ }%
+}%
+% \end{macrocode}
+% \end{macro}
+%
+%
+% \section{Required Packages}
+% \url{CTAN:macros/latex/contrib/other/misc/url.sty}%
+% \begin{macrocode}
+\RequirePackage{url}%
+% \end{macrocode}
+%
+% \section{Incompatible Packages}
+% Known to be incompatible are \classname{multicol} and \classname{cite}.
+% The former is incompatible with \classname{ltxgrid}, which replaces it.
+% The latter is incompatible with \classname{natbib}, which replaces its functionality.
+% \begin{macrocode}
+\appdef\class@inithook{%
+ \incompatible@package{cite}%
+ \incompatible@package{multicol}%
+}%
+% \end{macrocode}
+%
+% \section{Society- and Journal-Specific Code}
+%
+% \begin{macro}{\@journal}
+% Journal test helper, used as
+% \begin{verbatim}
+%\@ifx{\@journal\journal@pra}{%
+% <journal-specific setup>
+%\fi
+% \end{verbatim}
+% Journal code might like to further specify (if as yet undefined) or distinguish on
+% the following Booleans.
+%
+% Note: the journal substyle code should only alter the value of one of these Booleans if
+% the Boolean is \cmd\undefined. This convention is what makes the document's
+% options take precedence over the values set by the journal.
+%
+% FIXME: make this table an exhaustive listing of all the parameters set by
+% the class options.
+%
+% \begin{tabular}{ll}
+% \cmd\@pointsize &(10|11|12), depending on the type size\\
+% \cmd\footinbib@sw &true if footnotes are to be formatted in the bibliography\\
+% \cmd\preprintsty@sw &true for preprint and hyperpreprint\\
+% \cmd\eqsecnum@sw &true means that equations are numbered within sections\\
+% \cmd\secnumarabic@sw &true means the sections are to be numbered arabic\\
+% \cmd\groupauthors@sw &true means authors listed separately for each address\\
+% \cmd\preprint@sw &true means to produce the preprint numbers as part of the title block\\
+% \cmd\showPACS@sw &true means to produce the PACS as part of the title block\\
+% \cmd\showKEYS@sw &true means to produce the keywords as part of the title block\\
+% \cmd\@affils@sw &true means each affiliation is printed, for each author\\
+% \cmd\runinaddress@sw &true means author addresses are printed run-in\\
+% \cmd\draft@sw &true implies that PACS will be printed\\
+% \cmd\tightenlines@sw &true if preprint single spaced\\
+% \cmd\lengthcheck@sw &true if length checking is in effect\\
+% \cmd\byrevtex@sw &true means to announce ``typeset by \revtex''\\
+% \cmd\titlepage@sw &true for title is to be set on a separate page\\
+% \cmd\twocolumn@sw &true if two-column page grid\\
+% \cmd\twocolumn@sw &true if we are to automatically balance the columns of the last page\\
+% \cmd\twoside@sw &true means to format pages for duplex printing\\
+% \cmd\floats@sw &false means floats are migrated to end of document\\
+% \cmd\floatp@sw &true means endfloats are set one to a page\\
+% \cmd\class@amsfonts &if \cmd\@empty, means that \classname{amsfonts} will \emph{not} be loaded\\
+% \cmd\class@amssymb &if \cmd\@empty, means that \classname{amssymb} will \emph{not} be loaded\\
+% \cmd\frontmatter@footnote &if \cmd\undefined, means that the default (\cmd\footnote) will be used\\
+% \cmd\place@bibnumber &if \cmd\undefined, means that the default (inline) will be used\\
+% \end{tabular}
+%
+% Note: if \cmd\twocolumn@sw\ and \cmd\preprintsty@sw\ are both false,
+% then `galley' style is in effect. The \classoption{galley} option invokes
+% \classoption{onecolumn}, but does not affect the \cmd\preprintsty@sw.
+%
+% Note: \cmd\paperwidth\ and \cmd\paperheight\ are not integrated into this scheme,
+% and should be selected by the document alone.
+% \end{macro}
+%
+%
+% \section{Front Matter}
+%
+% \subsection{The title command}%
+%
+% \begin{macro}{\title}
+% \begin{macro}{\do@title}
+% The author uses \cmd\title\ to specify the title. As in the AMS
+% classes, \cmd\title\ has an optional argument specifying a short form
+% for use in running heads.
+%
+% Once the \cmd\title\ is given, you can specify any combination of
+% \cmd\thanks, \cmd\email, \cmd\homepage, and \cmd\altaffiliation\ commands,
+% applying to the title of the document.
+% \begin{macrocode}
+\renewcommand\title[2][]{%
+ \def\@title{#2}%
+ \def\@shorttitle{#1}%
+ \let\@AF@join\@title@join
+}%
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\@title}
+% \begin{macro}{\@shorttitle}
+% The arguments to \cmd\title\ are saved in these internal
+% macros, which are set up to produce a warning if \cmd\title\ has not
+% been used before \cmd\maketitle.
+% \begin{macrocode}
+\appdef\class@inithook{%
+ \def\@title{\class@warn{No title}}%
+ \let\@shorttitle\@empty
+ \let\@title@aux\@title@aux@cleared
+}%
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\@title@join}
+% \begin{macro}{\@title@join@}
+% \begin{macro}{\@title@aux@cleared}
+%
+% \begin{macrocode}
+\def\@title@join{\expandafter\@title@join@\@title@aux}%
+\def\@title@join@#1#2{%
+ \def\@title@aux{{\@join{\@separator}{#1}{#2}}}%
+}%
+\def\@title@aux@cleared{{}}%
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+%
+% \subsection{The author, affiliation, and related commands}%
+%
+% \begin{macro}{\c@affil}
+% A counter to store the affiliation numbers used for the
+% superscript marks. This command defines \cmd\theaffil.
+% \begin{macrocode}
+\newcounter{affil}
+% \end{macrocode}
+% \end{macro}
+%
+% \subsubsection{Commands for author or collaboration}%
+%
+% \begin{macro}{\author}
+% Each author is given in a \emph{separate} \cmd\author\ command.
+% This is similar to the AMS classes, but the AMS also try to support
+% the \classname{article} class's \cmd\and\ command for putting multiple
+% authors in one \cmd\author\ command.
+% \revtex, however, does not support the \cmd\and\ command.
+%
+% When the \cmd\author\ command is encountered, the previous author, author group, affiliation, and affiliation group are at an end.
+% This procedure moves the previous author to the author group (\cmd\move@AU),
+% moves the previous affiliation to the affiliation group (\cmd\move@AF),
+% and terminates the previous author/affiliation group (\cmd\move@AUAF) as the case may require.
+% \begin{macrocode}
+\renewcommand\author{% implicit #1
+ \@author@def{}% implicit #2
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\collaboration}
+% \begin{macro}{\@collaboration@}
+% Specify the collaboration (given after a group of authors).
+%
+% This procedure does exactly what \cmd\author\ does, and sets
+% a flag signifying that the \cmd\collaboration\ command was given.
+%
+% Note: earlier versions attempted to ignore the collaboration command if it was considered inappropriate.
+% We must not do this, however, because it throws the state machine out of kilter.
+% \begin{macrocode}
+\def\collaboration{% implicit #1
+ \groupauthors@sw{%
+ \class@warn{%
+ \string\collaboration\space is only allowed with superscriptaddress option.
+ Use superscriptaddress in place of groupedaddress, unsortedaddress, or runinaddress%
+ }%
+ }{}%
+ \@author@def{\@booleantrue\collaboration@sw}% implicit #2
+}%
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\@author@init}
+% \begin{macro}{\@author@cleared}
+% \begin{macro}{\@authorclear@sw}
+% \begin{macro}{\@author@def}
+% \begin{macro}{\@author@join@}
+% \begin{macro}{\@author@join}
+% \changes{v4.0beta 2}{1999/06/20}
+% {Change handling of \cs{@author} to prevent parsing problems.
+% (Helps in error recovery when syntax errors exist)}
+%
+% Here are all the access procedures for the author data structure.
+%
+% We define a flag value for \cmd\@author (private),
+% a procedure for setting \cmd\@author\ to the flag value (\cmd\@author@init),
+% a Boolean to test against the flag value (\cmd\@authorclear@sw),
+% and a procedure to add new ancillary data to the author.
+% \begin{macrocode}
+\def\@author@cleared{{}{}}%
+\def\@author@init{%
+ \let\@author\@author@cleared
+}%
+\def\@authorclear@sw{\@ifx{\@author\@author@cleared}}%
+\appdef\class@inithook{%
+ \@author@init
+}%
+\def\@author@def#1#2{%
+ \move@AU\move@AF\move@AUAF
+ \let\@AF@join\@author@join
+ #1%
+ \def\@author{{#2}{}}%
+}%
+\def\@author@join@#1#2#3{%
+ \def\@author{{#1}{\@join{\@separator}{#2}{#3}}}%
+}%
+\def\@author@join{\expandafter\@author@join@\@author}%
+\def\@author@gobble#1#2{}%
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\move@AU}
+% Move the most recent author to the list of current authors.
+% If we recently did a \cmd\@author@init, this is a no-op.
+% If we recently encountered a \cmd\collaboration,
+% we preface the author with the \cmd\CO@opr \ intercollated operator,
+% otherwise we apply \cmd\AU@opr.
+% \begin{macrocode}
+\def\move@AU{%
+ \@authorclear@sw{}{%
+ \collaboration@sw{%
+ \expandafter\appdef
+ \expandafter\AU@grp
+ \expandafter{%
+ \expandafter \CO@opr
+ \@author
+ }%
+ }{%
+ \expandafter\appdef
+ \expandafter\AU@grp
+ \expandafter{%
+ \expandafter \AU@opr
+ \@author
+ }%
+ }%
+ }%
+ \@author@init
+ \@booleanfalse\collaboration@sw
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\@author@finish}
+% We define a procedure to terminate author/affiliation parsing just before the title block is formatted.
+% We detect the case where an author group is not followed by an affiliation group.
+% \changes{v4.0beta 6}{2001/02/13}
+% {Actually append AU@grp to @AAClist.}
+%
+% We first finish off any author- or affiliation groups in progress.
+% Next, we detect the case where there is an author group in progress, but no affiliation group.
+% If so, we emit a diagnositic message and act as if a \cmd\noaffiliation\ command had been given.
+% In any case, we finish off any author/affiliation pair that may remain.
+% Formatting of the title block may now proceed.
+% \begin{macrocode}
+\def\@author@finish{%
+ \move@AU\move@AF
+ \@ifx{\AU@grp\@empty}{}{%
+ \@ifx{\AF@grp\@empty}{%
+ \begingroup
+ \let\href\@secondoftwo
+ \let\AU@opr\@firstoftwo
+ \let\CO@opr\@firstoftwo
+ \class@warn{Assuming \string\noaffiliation\space for \AU@grp}%
+ \endgroup
+ \begingroup\@affiliation{\blankaffiliation}%\endgroup in \@affiliation
+ \move@AF
+%
+% \appdef \@AAC@list{\AF@opr}%
+% \appdef@eval\@AAC@list{\AF@grp}%
+% \appdef@e \@AAC@list{\AU@grp}%
+% \let\AU@grp\@empty
+% \let\AF@grp\@empty
+ }{}%
+ }%
+ \move@AUAF
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\@join}
+% The procedure \cmd\@join\ inserts
+% a separator between two tokens, or, if the first token is
+% nil, elides both that token and the separator.
+% \begin{macrocode}
+\def\@join#1#2#3{%
+ \@if@empty{#2}{#3}{#2#1#3}%
+}%
+\def\@separator{;\space}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\surname}
+% \begin{macro}{\firstname}
+% No-op to allow better post-processing of the file.
+% \begin{macrocode}
+\let\surname\@firstofone
+\let\firstname\@firstofone
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\AU@grp}
+% The control sequence name \cmd\AU@grp\ accretes
+% consecutive \cmd\author\ entries (with \cmd\AU@opr\ as the intercollated operator)
+% and \cmd\collaboration\ entries (with \cmd\CO@opr\ as the intercollated operator).
+% Ultimately, its contents are appended to the author/affiliation list \cmd\@AAC@list.
+% It must be initialized to \cmd\@empty.
+%
+% Note on \cmd\AU@opr\ and \cmd\CO@opr: it is essential to treat these two operators to the greatest extent possible on an equal footing.
+% Therefore we invariably assign values to the pair of them within the same procedure.
+%
+% \begin{macrocode}
+\let\AU@grp\@empty
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\and}
+% The original \LaTeX\ idea of using one \cmd\author\ command to capture
+% all authors and their address just doesn't work with multiple authors
+% possibly sharing addresses, so in this class disable \cmd\and.
+% \begin{macrocode}
+\renewcommand\and{\class@err{\protect\and\space is not supported}}
+% \end{macrocode}
+% \end{macro}
+%
+%
+% \subsubsection{Commands for affiliation}%
+%
+% \begin{macro}{\affiliation}
+% After each group of authors with the same address,
+% give that address in \cmd\affiliation.
+% If later in the list you have an author
+% with the address of an earlier author, repeat the \cmd\affiliation\
+% command: the system will detect the similarity and (if using the \classoption{superscriptaddress} option) only print the
+% affiliation once, reusing the superscript marker.
+%
+% When the \cmd\affiliation\ command is encountered, the current author and author group (if any) are at an end.
+% Also, the current affiliation is at an end.
+%
+% FIXME: changes to catcode required elsewhere now?
+% \begin{macrocode}
+\def\cat@comma@active{\catcode`\,\active}%
+{\cat@comma@active\gdef,{\active@comma}}%
+\def\active@comma{,\penalty-300\relax}%
+\newcommand\affiliation{%
+ \move@AU\move@AF
+ \begingroup
+ \cat@comma@active
+ \@affiliation
+}%
+\def\@affiliation#1{%
+ \endgroup
+ \let\@AF@join\@affil@join
+ \@affil@def{#1}%
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\noaffiliation}
+% User-level command signifying that the current group of authors has \emph{no} affiliation.
+%
+% This is implemented by acting as if the \cmd\affiliation\ command has been given, and using a
+% flag value of \cmd\blankaffiliation, which \cmd\@affil@group\ can recognize as a no-op.
+% \begin{macrocode}
+\newcommand\noaffiliation{%
+ \move@AU\move@AF
+% \edef\AF@grp{\AF@grp}%
+ \begingroup\@affiliation{\blankaffiliation}%
+% \move@AF
+}%
+\def\blankaffiliation{}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\@affil@cleared}
+% \begin{macro}{\@affil@init}
+% \begin{macro}{\@affilclear@sw}
+% \begin{macro}{\@affil@def}
+% \begin{macro}{\@affil@join@}
+% \begin{macro}{\@affil@join}
+%
+% Here are all the access procedures for the affiliation data structure.
+% Note the similarity to those of the \cmd\@author\ data structure.
+%
+% We define a flag value for \cmd\@affil\ (private),
+% a procedure for setting \cmd\@affil\ to the flag value (\cmd\@affil@init),
+% a Boolean to test against the flag value (\cmd\@affilclear@sw),
+% and a procedure to add new ancillary data to the affiliation.
+% \begin{macrocode}
+\def\@affil@cleared{{}{}}%
+\def\@affil@init{%
+ \let\@affil\@affil@cleared
+}%
+\def\@affilclear@sw{\@ifx{\@affil\@affil@cleared}}%
+\appdef\class@inithook{%
+ \@affil@init
+}%
+\def\@affil@def#1{\def\@affil{{#1}{}}}%
+\def\@affil@join@#1#2#3{%
+ \def\@affil{{#1}{\@join{\@separator}{#2}{#3}}}%
+}%
+\def\@affil@join{\expandafter\@affil@join@\@affil}%
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\move@AF}
+% Move the most recent affiliation to the affiliation group.
+% If we recently did a \cmd\@affil@init, this is a no-op.
+% \begin{macrocode}
+\def\move@AF{%
+ \@affilclear@sw{}{%
+% \end{macrocode}
+% Set the \cmd\temp@sw\ to false, then execute the affiliation list (\cmd\@AFF@list)
+% with the intercollated operator set to \cmd\@affil@match\
+% and with \cmd\@tempa\ holding the first component of \cmd\@affil, the current affiliation, which is to be matched.
+% If the current affiliation has been seen before, \cmd\temp@sw\ will be set to true,
+% \cmd\@tempc\ will be set to the existing affiliation number,
+% and \cmd\@tempd\ will be set to the affiliation's ancillary data.
+% The Boolean \cmd\@affils@sw\ being false prevents this test from ever returning a positive result.
+%
+% This procedure uses \cmd\@tempa\ and sets \cmd\@tempc, \cmd\@tempd, and \cmd\temp@sw.
+% \begin{macrocode}
+ \@booleanfalse\temp@sw
+ \let\@tempd\@empty
+ \@affils@sw{%
+ \expandafter\@firstoftwodef\expandafter\@tempa\@affil
+ \def\AFF@opr{\@affil@match\@tempa}%
+ \@AFF@list
+ }{}\temp@sw
+% \end{macrocode}
+% True clause: This affiliation has been seen before.
+% If ancillary data for the affiliation have been given but are not identical to those seen before, give a warning.
+% \begin{macrocode}
+ {%
+ \expandafter\@secondoftwodef\expandafter\@tempb\@affil
+ \@ifx{\@tempb\@empty}{}{%
+ \@ifx{\@tempb\@tempd}{}{%
+ \class@warn{%
+ Ancillary information for \@tempa\space must not be different!
+ Please put it on the first instance%
+ }%
+ }%
+ }%
+ \appdef@eval\AF@grp\@tempc
+ }%
+% \end{macrocode}
+% False clause: This is a new address:
+% increment the counter to get a unique ID for the new affiliation
+% and append the new number, address, and footnote to the list \cmd\@AFF@list.
+% \begin{macrocode}
+ {%
+ \@if@empty{\expandafter\@firstoftwo\@affil}{%
+ \edef\@tempc{\z@}%
+ }{%
+ \stepcounter{affil}%
+ \edef\@tempc{\theaffil}%
+ }%
+ \appdef \@AFF@list{\AFF@opr}%
+ \appdef@eval\@AFF@list\@tempc
+ \appdef@e \@AFF@list{\@affil}%
+ \appdef@eval\AF@grp\@tempc
+ }%
+% \end{macrocode}
+% Append the ID of the affiliation to the affiliation group of the current author group.
+% \begin{macrocode}
+ \@affil@init
+ }%
+}%
+\def\@firstoftwodef#1#2#3{%
+ \def#1{#2}%
+}%
+\def\@secondoftwodef#1#2#3{%
+ \def#1{#3}%
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\move@AUAF}
+% \changes{v4.h}{1998/06/25}
+% {Changes to support groups of affiliations}
+% Append \cmd\AF@grp\ and \cmd\AU@grp\ to \cmd\@AAC@list\ %
+% and merge any new \cmd\AF@grp\ to \cmd\@AFG@list.
+% \begin{macrocode}
+\def\move@AUAF{%
+% \end{macrocode}
+% Entire procedure is predicated on something non-trivial to move.
+%
+% If both author and affiliation are nil, bail out.
+%
+% No, try that again:
+%
+% If we have seen any affiliations, then it is time to move the current
+% affiliation group and author group to \cmd\@AAC@list.
+% If not, we are picking up authors into an author group, and we
+% should bail out.
+% \begin{macrocode}
+%\@ifx{\AU@grp\@empty}{\@ifx{\AF@grp\@empty}{\true@sw}{\false@sw}}{\false@sw}%
+ \@ifx{\AF@grp\@empty}%
+ {}{%
+% \end{macrocode}
+% First, append current affiliation group (expansion of \cmd\AF@grp)
+% to the author/affiliation list (\cmd\@AAC@list),
+% using \cmd\AF@opr\ as the intercollated operator;
+% also append the current authors (\cmd\AU@grp).
+% \begin{macrocode}
+ \appdef \@AAC@list{\AF@opr}%
+ \appdef@eval\@AAC@list{\AF@grp}%
+ \appdef@e \@AAC@list{\AU@grp}%
+% \end{macrocode}
+% Next, append the current affiliation group to \cmd\@AFG@list,
+% (if it is not already there)
+% using \cmd\AFS@opr\ as the intercollated operator.
+%
+% Note that \cmd\@AFG@list\ is a list of \emph{unique} affiliation groups.
+% \begin{macrocode}
+ \@ifx{\@empty\AU@grp}{}{%
+ \@ifx{\@empty\AF@grp}{}{%
+ \@booleanfalse\temp@sw
+ \def\AFS@opr{\x@match\AF@grp}%
+ \@AFG@list
+ \temp@sw{}{%
+ \appdef \@AFG@list{\AFS@opr}%
+ \appdef@eval\@AFG@list{\AF@grp}%
+ }%
+ }%
+ }%
+% \end{macrocode}
+% Finally, null out the macros holding the author group and affiliation group, respectively.
+% \begin{macrocode}
+ \let\AU@grp\@empty
+ \let\AF@grp\@empty
+ }%
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\@AAC@list}
+% The \cmd\@AAC@list\ macro accretes
+% authors (using \cmd\AU@opr\ as the intercollated operator),
+% collaborations (using \cmd\CO@opr\ as the intercollated operator),
+% and
+% affiliations (using \cmd\AF@opr\ as the intercollated operator).
+% \begin{macrocode}
+\appdef\class@inithook{%
+ \let\@AAC@list\@empty
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\@AFG@list}
+% The macro \cmd\@AFG@list\ accretes \cmd\AF@grp,
+% with \cmd\AFS@opr\ as the intercollated operator.
+%
+% This token list is employed when and only when \classoption{groupedaddress} is in effect.
+%
+% It must be empty at the beginning of the document.
+% \begin{macrocode}
+\appdef\class@inithook{%
+ \let\@AFG@list\@empty
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\@AFF@list}
+% The macro \cmd\@AFF@list\ accretes the list of affiliations
+% with \cmd\AFF@opr\ as the intercollated operator.
+% It must be empty at the beginning of the document.
+%
+% Each \cmd\affiliation\ command in the document contributes
+% to this list, as long as the argument of that instance of the command
+% is ``new''.
+%
+% Each entry in this list consists of the \cmd\AFF@opr\ intercollated operator
+% followed by three brace-delimited tokens, representing, in order:
+% \begin{enumerate}
+% \item
+% the affiliation's unique ID, a number increasing monotonically for
+% each new entry,
+% \item
+% the optional argument of the \cmd\affiliation\ command,
+% the footnote to the affiliation, and
+% \item
+% the text of the affiliation.
+% \end{enumerate}
+%
+% At various junctures in the code, the control sequence name
+% \cmd\AFF@opr\ assumes the meaning of
+% \cmd\@affil@footnote\ (FIXME: never used),
+% \cmd\@affil@group, \cmd\affil@script,
+% or is defined to expand to \cmd\@affil@match\cmd\@tempa.
+% and the list \cmd\@AFF@list\ is expanded.
+% \begin{macrocode}
+\appdef\class@inithook{%
+ \let\@AFF@list\@empty
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\AF@grp}
+% The macro \cmd\AF@grp\ accretes affiliation IDs into an affiliation group.
+% It must be empty at the beginning of the document.
+% \begin{macrocode}
+\let\AF@grp\@empty
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\@affil@match}
+% The definition for the \cmd\AFF@opr, the intercollated operator for \cmd\@AFF@list,
+% when adding a new affiliation: seeks a match with any earlier affiliation.
+% When the affiliation list is executed, this procedure tests each entry against
+% the `new' affiliation (which has been stored in \cmd\@tempa).
+%
+% Uses \cmd\temp@sw, \cmd\@tempc, and \cmd\@tempd\ to communicate back:
+% if it returns \cmd\temp@sw\ true, a match was found,
+% then \cmd\@tempc\ will be the footnote number of the matching entry
+% and \cmd\@tempd\ will be the auxiliary information of the matching entry.
+% \begin{macrocode}
+\def\@affil@match#1#2#3#4{%
+ \temp@sw{}{%
+ \def\@tempifx{#3}%
+ \@ifx{\@tempifx#1}{%
+ \@booleantrue\temp@sw
+ \def\@tempc{#2}%
+ \def\@tempd{#4}%
+ }{%
+ }%
+ }%
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\@tempaffil@script}
+% Give a superscript marker to the address,
+% and junk the text argument (|#2|).
+% \begin{macrocode}
+%\def\@tempaffil@script#1#2{%
+% ,\,#1%
+%}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\@tempaffil@group}
+%
+% Execute if we are grouping authors above `main addresses' (\classoption{groupedaddress}).
+% Put temp address in a footnote, with explanatory text.
+% Note: this procedure uses \cmd\@tempb\ and \cmd\@tempc\
+% to communicate to \cmd\AFF@opr.
+% \begin{macrocode}
+%\def\@tempaffil@group#1#2{%
+% \def\@tempc{#1}%
+% \def\@tempb{#2}%
+% \begingroup
+% \let\AFF@opr \@affil@footnote
+% \@AFF@list
+% \endgroup
+%}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\@affil@footnote}
+% The alias of \cmd\AFF@opr\ when \cmd\@AFF@list\
+% is executed by \cmd\@tempaffil@group.
+% If we find a match of the affiliation number to \cmd\@tempc,
+% then footnotes are formatted with the relevant content and \cmd\@tempb.
+% \begin{macrocode}
+%\def\@affil@footnote#1#2#3{%
+% \def\@tempifx{#1}%
+% \@ifx{\@tempifx\@tempc}{%
+% \comma@space
+% \expandafter\frontmatter@footnote\expandafter{\@tempb#2}%
+% \frontmatter@footnote{#3}%
+% }{%
+% }%
+%}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\@AF@join}
+% The \cmd\csname\ \cmd\@AF@join\ is a procedure used within
+% \cmd\email, \cmd\homepage, \cmd\thanks, and \cmd\altaffiliation.
+% When such a command appears after an \cmd\author, \cmd\collaboration, or \cmd\affiliation
+% command, the \cmd\@AF@join\ procedure appends the argument given to the appropriate
+% macro.
+% Its default is to give an error message, since these commands are legal only within
+% the particular context mentioned.
+% \begin{macrocode}
+\appdef\class@inithook{%
+ \let\@AF@join\@AF@join@error
+}%
+\def\@AF@join@error#1{%
+ \class@warn{%
+ \string\email, \string\homepage, \string\thanks, or \string\altaffiliation\space
+ appears in wrong context.
+ }%
+}%
+% \end{macrocode}
+% \end{macro}
+%
+%
+% \subsubsection{Commands for auxiliary information}%
+%
+% \begin{macro}{\email}
+% Just tacks the email address on to the current author or affiliation.
+% \begin{macrocode}
+\def\sanitize@url{%
+ \@makeother\%%
+ \@makeother\~%
+ \@makeother\_%
+}%
+\newcommand*\email[1][Electronic address: ]{\begingroup\sanitize@url\@email{#1}}%
+\def\@email#1#2{%
+ \endgroup
+ \@AF@join{#1\href{mailto:#2}{#2}}%
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\homepage}
+% Just tacks the URL on to the current author or affiliation.
+% Note: group opened in \cmd\homepage\ is closed in \cmd\@homepage.
+% \begin{macrocode}
+\newcommand*\homepage[1][URL:~]{\begingroup\sanitize@url\@homepage{#1}}%
+\def\@homepage#1#2{%
+ \endgroup
+ \@AF@join{#1\url{#2}}%
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\url}
+% \begin{macro}{\href}
+% \begin{macrocode}
+\appdef\class@documenthook{%
+ \providecommand\url{\texttt}%
+ \providecommand\href[2]{\url{#2}}%
+}%
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\thanks}
+%
+% The operative version of \cmd\thanks\ appends an item to \cmd\@author, or \cmd\@affil's auxiliary data.
+% \begin{macrocode}
+\def\thanks{% implicit #1
+ \@AF@join
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\altaffiliation}
+% \changes{v4.d}{1998/01/31}
+% {change name from \cs{tempaffiliation}}
+% \changes{v4.l}{1998/09/01}
+% {add optional argument handling back}
+% Implemented more or less like \cmd\thanks\ but shares the affiliation
+% counter.
+% Optional argument may be used to give explanatory text eg
+% `currently staying at'
+% This will be placed before the address, if used in a footnote.
+% \begin{macrocode}
+\newcommand*\altaffiliation[2][]{%
+ \@AF@join{#1#2}%
+}%
+% \end{macrocode}
+% \end{macro}
+%
+%
+% \subsubsection{Procedures for author, collaboration, and affiliation}%
+%
+% Macros that list off, say, authors, and which require punctuation, like ``A, B, and C'', will use
+% \cmd\set@listcomma@list, \cmd\set@listcomma@count, \cmd\@listand, and \cmd\@listcomma.
+% These macros use the \cmd\@tempcnta\ register to keep track of how many items remain to be listed off.
+%
+% At present, \cmd\@author@present\ will decrement that counter.
+%
+% \begin{macro}{\set@listcomma@list}
+% \begin{macro}{\set@listcomma@count}
+% \changes{v4.l}{1998/09/01}
+% {macro added}
+% Pop the author count for this collaboration.
+%
+% Note: by side effect, it assigns \cmd\@listcomma:
+% for a list of length two, suppress comma addition.
+% \begin{macrocode}
+\def\set@listcomma@list#1{%
+ \expandafter\@reset@ac\expandafter#1#1{0}\@reset@ac{%
+ \let\@listcomma\relax
+ }{%
+ \let\@listcomma\@listcomma@comma
+ }%
+}%
+\def\set@listcomma@count#1{%
+ \@ifnum{#1=\tw@}{%
+ \let\@listcomma\relax
+ }{%
+ \let\@listcomma\@listcomma@comma
+ }%
+}%
+% \end{macrocode}
+% Does the actual pop operation, then generates a Boolean which selects one of the two assignments for \cmd\@listcomma.
+%
+% Note: this procedure sets \cmd\@tempcnta\ to reflect the number of items to list off.
+% \begin{macrocode}
+\def\@reset@ac#1#2#3\@reset@ac{%
+ \def#1{#3}%
+ \@tempcnta#2\relax
+ \@ifnum{#2=\tw@}%
+}%
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\listand}
+% Might need extending with penalties etc.
+%
+% Note: this procedure expects \cmd\@tempcnta\ to reflect the current number of items remaining to list off.
+% \begin{macrocode}
+\def\@listand{\@ifnum{\@tempcnta=\tw@}{\andname\space}{}}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\@listcomma}
+% This definition, with \cmd\@ne\ puts a comma before and.
+%
+% David Carlisle says: In UK English (at least) would have \cmd\tw@\ here,
+% which would then implement the convention of indicating conjunction with comma or ``and'', but not both.
+%
+% Note: this procedure expects \cmd\@tempcnta\ to reflect the current number of items remaining to list off.
+% \begin{macrocode}
+\def\@listcomma@comma{\@ifnum{\@tempcnta>\@ne}{,}{}}%
+\def\@listcomma@comma@UK{\@ifnum{\@tempcnta>\tw@}{,}{}}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\@collaboration@err}
+% This command will be executed if are in a \classoption{groupedaddress} situation and
+% the document contained a \cmd\collaboration\ command.
+%
+% There already has been issued a warning message explaining that \cmd\collaboration\ is wrong in this case,
+% so here we simply gobble.
+% \begin{macrocode}
+\def\@collaboration@err#1#2{%
+%\class@warn{An illegal \string\collaboration\ command has been entered!}%
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\doauthor}
+% Main control over how authors are typeset.
+%
+% |#1| is loaded by \cmd\author
+%
+% |#2| is loaded by \cmd\email, \cmd\homepage, \cmd\altaffiliation, or \cmd\thanks
+%
+% |#3| is the superscript affiliation, if at all used.
+%
+% First, the author name is formatted, followed by a comma,
+% then come any marks relating to affiliation (if present),
+% then come the homepage URL and email address, if any, with
+% appropriate punctuation.
+% \begin{macrocode}
+\def\doauthor#1#2#3{%
+ \ignorespaces#1\unskip\@listcomma
+ \def\@tempifx{#3}%
+ \@ifx{\@tempifx\@empty}{%
+ \def\@tempifx{#2}%
+ \@ifx{\@tempifx\@empty}{}{\frontmatter@footnote{#2}}%
+ }{%
+ #3%
+ \def\@tempifx{#2}%
+ \@ifx{\@tempifx\@empty}{}{\comma@space\frontmatter@footnote{#2}}%
+ }%
+ \space \@listand
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\AF@opr}
+% DPC: The number(s) associated to the affiliation address(es).
+% \end{macro}
+%
+% \begin{macro}{\@thanks}
+% \changes{v4.d}{1998/01/31}
+% {\cs{comma@space} added}
+% Society-specific options may change this. \cmd\comma@space\ is a separator
+% between adjacent footnotes.
+% \begin{macrocode}
+\def\@thanks{\comma@space\footnote}
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\x@match}
+% Procedure \cmd\x@match\ is used as an alias \cmd\AFS@opr,
+% probing the list of affiliation groups
+% and setting a Boolean if a match is found to \#1 (usually the current affiliation group).
+%
+% Procedure \cmd\y@match\ is used as an alias to \cmd\@TBN@opr,
+% probing the list of unique title block footnotes for a match to the footnote text:
+% when the first match is found it sets a Boolean,
+% defining \cmd\@tempb\ to the number of that footnote.
+% \begin{macrocode}
+\def\x@match#1#2{%
+ \temp@sw{}{%
+ \def\@tempifx{#2}%
+ \@ifx{\@tempifx#1}{%
+ \@booleantrue\temp@sw
+ }{%
+ }%
+ }%
+}%
+\def\y@match#1#2#3{%
+ \temp@sw{}{%
+ \def\@tempifx{#3}%
+ \@ifx{\@tempifx#1}{%
+ \@booleantrue\temp@sw
+ \def\@tempb{#2}%
+ }{%
+ }%
+ }%
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\frontmatter@footnote}
+% Used when typesetting a title block footnote.
+% (Item 290: merge duplicates.)
+%
+% Maintains a list of frontmatter footnotes along with associated device.
+% Uniquifies each footnote by traversing the list and looking for a match to the footnote text.
+% If found, define \cmd\@tempb\ to the corresponding footnote device, and typeset it with \cmd\@footnotemark.
+% If never found, manually increment footnote counter,
+% determine the corresponding footnote device, and
+% add footnote text and device to the list of frontmatter footnotes.
+%
+% The list \cmd\@FMN@list\ is processed later in the title block production.
+%
+% Note that this method of making footnotes runs afoul of \classname{hyperref}'s patches of the \LaTeX\ kernel.
+% Therefore, we avoid trouble by refraining from using \cmd\@footnotemark\ and by creating our own hypertext links.
+% \begin{macrocode}
+\def\frontmatter@footnote#1{%
+ \begingroup
+ \@booleanfalse\temp@sw
+ \def\@tempa{#1}%
+ \let\@tempb\@empty
+ \def\@TBN@opr{\y@match\@tempa}%
+ \@FMN@list
+ \temp@sw{%
+ \expandafter\frontmatter@footnotemark
+ \expandafter{\@tempb}%
+ }{%
+ \stepcounter\@mpfn
+ \protected@xdef\@tempb{\the\csname c@\@mpfn\endcsname}%
+ \expandafter\frontmatter@footnotemark
+ \expandafter{\@tempb}%
+ \expandafter\g@addto@macro
+ \expandafter\@FMN@list
+ \expandafter{%
+ \expandafter \@TBN@opr
+ \expandafter {\@tempb}{#1}%
+ }%
+ }%
+ \endgroup
+}%
+\appdef\class@inithook{%
+ \global\let\@FMN@list\@empty
+}%
+\def\frontmatter@footnotemark#1{%
+ \leavevmode
+ \ifhmode\edef\@x@sf{\the\spacefactor}\nobreak\fi
+ \begingroup
+ \hyper@linkstart {link}{frontmatter@#1}%
+ \csname c@\@mpfn\endcsname#1\relax
+ \def\@thefnmark{\thempfn}%
+ \@makefnmark
+ \hyper@linkend
+ \endgroup
+ \ifhmode\spacefactor\@x@sf\fi
+ \relax
+}%
+% \end{macrocode}
+% \end{macro}
+%
+%
+%
+% \subsection{The keywords command}%
+% \begin{macro}{\keywords}
+% \begin{macro}{\@keywords}
+% Usual game, save text in a macro for processing by \cmd\maketitle.
+% \begin{macrocode}
+\def\keywords#1{%
+ \aftermaketitle@chk{\keywords}%
+ \gdef\@keywords{#1}%
+}%
+\def\@keywords{}%
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \subsection{The \cs{date} command and related commands}%
+% \begin{macro}{\date}
+% \begin{macro}{\received}
+% \begin{macro}{\revised}
+% \begin{macro}{\accepted}
+% \begin{macro}{\published}
+% \begin{macro}{\@date}
+% \begin{macro}{\@received}
+% \begin{macro}{\@revised}
+% \begin{macro}{\@accepted}
+% \begin{macro}{\@published}
+% Use the \cmd\date\ command to specify the document date,
+% the \cmd\received\ command to specify the date received,
+% \cmd\revised\ for date revised,
+% \cmd\accepted\ for date accepted, and
+% \cmd\published\ for date published.
+% Normally only \cmd\date\ to be used by author,
+% the remainder used only by editors.
+%
+% DPC: As for \cmd\keywords, but this time don't flag a warning if there is no
+% revision date specified.
+%
+% In each case the user-level command defines the value of a
+% macro which buffers the data entered by the user.
+% For instance, \cmd\received\ causes \cmd\@received\ to
+% acquire a value.
+% The optional argument allows the user to override the
+% text that will be typeset along with the date;
+% the default value of that text is itself
+% a localized macro.
+% \begin{macrocode}
+\renewcommand*\date[2][\Dated@name]{\def\@date{#1#2}}%
+\def\@date{}%
+\newcommand*\received[2][\Received@name]{\def\@received{#1#2}}%
+\def\@received{}%
+\newcommand*\revised[2][\Revised@name]{\def\@revised{#1#2}}%
+\def\@revised{}%
+\newcommand*\accepted[2][\Accepted@name]{\def\@accepted{#1#2}}%
+\def\@accepted{}%
+\newcommand*\published[2][\Published@name]{\def\@published{#1#2}}%
+\def\@published{}%
+% \end{macrocode}
+%
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \subsection{The pacs command}%
+% PACS, the Physics and Astronomy Classification Scheme.
+% \begin{macro}{\pacs}
+% \begin{macro}{\@pacs}
+% As for \cmd\keywords\
+% \begin{macrocode}
+\def\pacs#1{%
+ \aftermaketitle@chk{\pacs}%
+ \gdef\@pacs{#1}%
+}%
+\def\@pacs{}%{\class@warn{No PACS numbers}}%
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \subsection{The \cs{preprint} command}
+%
+% \changes{v4.g}{1998/06/10}
+% {multiple preprint commands}
+% \begin{macrocode}
+\def\preprint#1{\gappdef\@preprint{\preprint{#1}}}%
+\let\@preprint\@empty
+% \end{macrocode}
+%
+% \subsection{draft}
+% DPC: Name clash with draft option. \LaTeX\ doesn't care, but it confuses me.
+%
+% This command is a 3.1 relic, now removed.
+% \begin{macrocode}
+%\newcommand\draft{\@booleantrue\draft@sw}%
+% \end{macrocode}
+%
+% \subsection{The \env{abstract} environment}%
+%
+% \changes{v4.0beta 2}{1999/06/20}
+% {Gathered all code for the abstract environment together
+% and abstracted out the formatting details for journals to override.}
+%
+% \begin{macro}{\absbox}
+%
+% \begin{macrocode}
+\newbox\absbox
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{environment}{abstract}
+% Abstract, as in AMS classes, must be specified \emph{before}
+% \cmd\maketitle. It just saves everything up in \cmd\absbox.
+% \changes{v4.d}{1998/01/31}
+% {If empty \cs{abstractname} omit spacing}
+% \begin{macrocode}
+% \changes{v4.e}{1998/01/31}
+% {hookify}
+% Note that the specifics of how the abstract is to be
+% formatted are set by \cmd\frontmatter@abstractwidth,
+% \cmd\frontmatter@abstractheading, and \cmd\frontmatter@abstractfont.
+%
+% Here we wish to set the abstract into type but save it away in a box, much like
+% the \cmd\minipage\ command does.
+%
+% Note that the \cmd\endabstract\ portion of the environment
+% begins with code that mostly duplicates \cmd\endminipage, but without the \cmd\@iiiparbox.
+% At the end, we simply transfer the contents of the \cmd\minipage\
+% into a box of our own.
+%
+% However, we also have to end the gratuitous
+% paragraph begun by \cmd\minipage.
+% We ensure that no trace of that paragraph is left,
+% by absorbing the \cmd\parindent\ box.
+% Doing so destroys the paragraph entirely, except
+% for the \cmd\parskip\ glue, and that morsel is pruned, because we are at the top
+% of a page.
+%
+% This mechanism is vulnerable, however, because anything placed into the horizontal list
+% after the \cmd\parindent\ box will confound it.
+% And that is exactly what happens under Mik\TeX:
+% There seems to be an extension to the \TeX\ standard operating here which inserts a \cmd\special \ at the beginning of every paragraph.
+% (The mechnism is \emph{not} based on \cmd\everypar.)
+%
+% To accomodate this state of affairs,
+% we have a new mechanism that offers an even more complete rewrite of \env{minipage}.
+%
+% The user can elect to use this new mechanism via the \classoption{newabstract};
+% likewise the old mechanism via the \classoption{oldabstract}.
+%
+% \begin{macrocode}
+\newenvironment{abstract}{%
+ \aftermaketitle@chk{\begin{abstract}}%
+ \global\setbox\absbox\vbox\bgroup
+ \color@begingroup
+ \columnwidth\textwidth
+ \hsize\columnwidth
+ \@parboxrestore
+ \def\@mpfn{mpfootnote}\def\thempfn{\thempfootnote}\c@mpfootnote\z@
+ \let\@footnotetext\frontmatter@footnotetext
+ \minipagefootnote@init
+ \let\@listdepth\@mplistdepth \@mplistdepth\z@
+ \@minipagerestore
+ \@setminipage
+ \frontmatter@abstractheading
+ \frontmatter@abstractfont
+ \let\footnote\mini@note
+}{%
+ \par
+ \unskip
+ \minipagefootnote@here
+ \@minipagefalse %% added 24 May 89
+ \color@endgroup
+ \egroup
+}%
+% \end{macrocode}
+% \end{environment}
+%
+% \begin{macro}{\frontmatter@footnotetext}
+% We reimplement \cmd\@mpfootnotetext\ under a new name (so as not to be overridden by the \classname{hyperref} package)
+% and extend it to accomodate hyperrefs.
+%
+% Note that this procedure is very like \classname{ltxutil}'s \cmd\mp@footnotetext, except that it
+% uses \texttt{p@mpfootnote} instead of \texttt{p@mpfootnote}, and
+% \cmd\frontmatter@makefntext\ instead of \cmd\@makefntext.
+%
+% \begin{macrocode}
+\long\def\frontmatter@footnotetext#1{%
+ \minipagefootnote@pick
+ \reset@font\footnotesize
+ \interlinepenalty \interfootnotelinepenalty
+ \splittopskip \footnotesep
+ \splitmaxdepth \dp \strutbox
+ \set@footnotewidth
+ \@parboxrestore
+ \protected@edef\@currentlabel{\csname p@mpfootnote\endcsname\@thefnmark}%
+ \color@begingroup
+ \frontmatter@makefntext{%
+ \rule\z@\footnotesep\ignorespaces#1\@finalstrut\strutbox
+ }%
+ \color@endgroup
+ \minipagefootnote@drop
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\frontmatter@abstractheading}
+% The default abstract head; journals will override this procedure.
+% \begin{macrocode}
+\def\frontmatter@abstractheading{%
+ \begingroup
+ \centering\large
+ \abstractname
+ \par
+ \endgroup
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\frontmatter@abstractfont}
+% The default type specification for the body of the abstract.
+% Journals will override this setting.
+% \begin{macrocode}
+\def\frontmatter@abstractfont{}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\frontmatter@abstractwidth}
+% The default setting is the full text width;
+% journals can override this setting.
+% \begin{macrocode}
+\def\frontmatter@abstractwidth{\textwidth}
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\frontmatter@abstract@produce}
+% This procedure determines how the abstract is incorporated
+% into the title block. We split this out in anticipation
+% of the need to lift the limitation that the abstract not break over pages.
+%
+% If we are in preprint style, we provide an easy pagebreak point immediately above the abstract.
+% This means that the abstract will either fit entirely on the same page as the title block,
+% or it will start a page of its own (which may itself break onto a subsequent page if necessary).
+% \begin{macrocode}
+\def\frontmatter@abstract@produce{%
+ \par
+% \tracingall
+ \preprintsty@sw{%
+ \do@output@MVL{%
+ \vskip\frontmatter@preabstractspace
+ \vskip200\p@\@plus1fil
+ \penalty-200\relax
+ \vskip-200\p@\@plus-1fil
+ }%
+ }{%
+ \addvspace{\frontmatter@preabstractspace}%
+ }%
+% \leavevmode
+ \begingroup
+ \dimen@\baselineskip
+ \setbox\z@\vtop{\unvcopy\absbox}%
+ \advance\dimen@-\ht\z@\advance\dimen@-\prevdepth
+ \@ifdim{\dimen@>\z@}{\vskip\dimen@}{}%
+ \endgroup
+ \unvbox\absbox
+ \@ifx{\@empty\mini@notes}{}{\mini@notes\par}%
+% \par
+ \addvspace\frontmatter@postabstractspace
+% \showlists
+% \tracingplain
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\frontmatter@preabstractspace}
+% \begin{macro}{\frontmatter@postabstractspace}
+% Space above and space below abstract in title block
+% \begin{macrocode}
+\def\frontmatter@preabstractspace{.5\baselineskip}
+\def\frontmatter@postabstractspace{.5\baselineskip}
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+%
+%
+% \subsection{Formatting the title block}%
+%
+% \begin{environment}{titlepage}
+% This is \LaTeXe's \classname{article} class version,
+% with modifications.
+% \changes{v4.0beta 3}{1999/11/13}
+% {grid changes with ltxgrid}
+% \begin{macrocode}
+\newenvironment{titlepage}{%
+ \twocolumn@sw{\onecolumngrid}{\newpage}%
+ \thispagestyle{titlepage}%
+ \setcounter{page}\@ne
+}{%
+ \twocolumn@sw{\twocolumngrid}{\newpage}%
+ \twoside@sw{}{%
+ \setcounter{page}\@ne
+ }%
+}%
+% \end{macrocode}
+% \end{environment}
+%
+% \begin{macro}{\maketitle}
+% \changes{v4.d}{1998/01/31}
+% {hookify}
+% Put it all together to format the title block.
+% \begin{macrocode}
+\def\maketitle{%
+ \@author@finish
+ \title@column\titleblock@produce
+ \suppressfloats[t]%
+% \end{macrocode}
+% Now save some memory.
+% \begin{macrocode}
+ \let\and\relax
+ \let\affiliation\@gobble
+% \let\address\affiliation
+ \let\author\@gobble
+ \let\@AAC@list\@empty
+ \let\@AFF@list\@empty
+ \let\@AFG@list\@empty
+% \let\keywords\@gobble
+% \let\@keywords\@empty
+ \let\@AF@join\@AF@join@error
+ \let\email\@gobble
+ \let\@address\@empty
+ \let\maketitle\relax
+ \let\thanks\@gobble
+ \titlepage@sw{%
+ \vfil
+ \clearpage
+ }{}%
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\titleblock@produce}
+% This procedure produces the title block.
+%
+% It effectively executes inside a group
+% and always returns us to vertical mode.
+%
+% Note: we assume that it is OK to set the footnote counter to zero at this point.
+% Is this a safe assumption?
+% \begin{macrocode}
+\def\titleblock@produce{%
+ \begingroup
+% \end{macrocode}
+% Special footnote version [of footnotes] for \cmd\thanks,
+% copied from \classname{article} class \cmd\maketitle.
+% \changes{v4.0beta 2}{1999/06/20}
+% {Set \cs{footnote} to \LaTeX\ standard version for this scope.}
+% \begin{macrocode}
+ \let\footnote\footnote@latex
+ \let\@makefnmark\@makefnmark@latex
+ \let\@footnotemark\@footnotemark@latex
+ \let\thefootnote\frontmatter@thefootnote
+% \let\@makefntext\frontmatter@makefntext
+ \global\c@footnote\z@
+ \let\@makefnmark\frontmatter@makefnmark
+ \frontmatter@setup
+ \thispagestyle{titlepage}\label{FirstPage}%
+% \end{macrocode}
+% Produce the title:
+% \begin{macrocode}
+ \frontmatter@title@produce
+% \end{macrocode}
+% Produce the author list:
+% \begin{macrocode}
+ \groupauthors@sw{%
+ \frontmatter@author@produce@group
+ }{%
+ \frontmatter@author@produce@script
+ }%
+% \end{macrocode}
+% Produce the dates:
+% \begin{macrocode}
+ \frontmatter@RRAPformat{%
+ \expandafter\produce@RRAP\expandafter{\@date}%
+ \expandafter\produce@RRAP\expandafter{\@received}%
+ \expandafter\produce@RRAP\expandafter{\@revised}%
+ \expandafter\produce@RRAP\expandafter{\@accepted}%
+ \expandafter\produce@RRAP\expandafter{\@published}%
+ }%
+% \end{macrocode}
+% Produce the abstract, PACS, and keywords, and end any paragraph.
+% \begin{macrocode}
+ \frontmatter@abstract@produce
+ \@pacs@produce\@pacs
+ \@keywords@produce\@keywords
+ \par
+ \frontmatter@finalspace
+ \endgroup
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\frontmatter@title@produce}
+% The default formatting procedure for the article title.
+% This procedure should take care of the vertical space below the title.
+%
+% Journals may override this procedure, but it is more likely that they will
+% use the hooks \cmd\frontmatter@title@above, \cmd\frontmatter@title@format, and \cmd\frontmatter@title@below.
+%
+% At this point, we patch into \cmd\let@mark\ so that the common markup idiom \cmd\\ is taken care of.
+% \begin{macrocode}
+\def\frontmatter@title@produce{%
+ \begingroup
+ \frontmatter@title@above
+ \frontmatter@title@format
+ \@title
+ \@ifx{\@title@aux\@title@aux@cleared}{}{%
+ \expandafter\frontmatter@footnote\expandafter{\@title@aux}%
+ }%
+ \par
+% \@ifx{\@shorttitle\@empty}{%
+% \markboth{\@title}{\@title}%
+% }{%
+% \markboth{\@shorttitle}{\@shorttitle}%
+%% \class@warn{Unused short title ignored}%
+% }%
+ \frontmatter@title@below
+ \endgroup
+}%
+\appdef\let@mark{\let\\\relax}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\frontmatter@title@above}
+% \begin{macro}{\frontmatter@title@format}
+% \begin{macro}{\frontmatter@title@below}
+% The default values for formatting specifications for the article title.
+%
+% The procedure \cmd\frontmatter@title@above\ should take care of the vertical space above the title;
+% \cmd\frontmatter@title@below\ below.
+% The procedure \cmd\frontmatter@title@format\ should invoke any font switches, etc.
+% that may apply to the title.
+%
+% Journals will override this procedure.
+% \begin{macrocode}
+\def\frontmatter@title@above{}%
+\def\frontmatter@title@format{}%
+\def\frontmatter@title@below{\addvspace{\baselineskip}}%
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \subsubsection{Authors and affiliations in superscriptaddress style}
+%
+% \begin{macro}{\frontmatter@author@produce@script}
+% \cmd\frontmatter@author@produce@script\ is an alias for \cmd\frontmatter@author@produce,
+% the procedure that formats the author/affiliation list.
+%
+% In this case, the authors affiliations are being superscripted
+% (class option \classoption{superscriptaddress}).
+%
+% This procedure must ensure that any paragraph that it starts gets ended finally.
+% \begin{macrocode}
+\def\frontmatter@author@produce@script{%
+ \begingroup
+ \let\@author@present\@author@present@script
+% \end{macrocode}
+% \changes{v4.l}{1998/09/01}
+% {Changes to add collaboration processing, which now
+% means doing comma/and processing on each sublist}
+% When the author/affiliation list \cmd\@AAC@list\ is executed,
+% \cmd\@tempcnt\ counts each author,
+% and
+% \cmd\@tempa\ stores a list of author indices for the current collaboration.
+%
+% Note: this procedure uses \cmd\AU@temp\ to hold a list of author counts for each collaboration.
+% Note: also uses \cmd\@tempcnta\ to communicate between procedures.
+% This usage ot \cmd\@tempcnta\ is OK, because the expansion of \cmd\@AAC@list\ will trigger only the execution of
+% our own procedures.
+% \begin{macrocode}
+ \let\AU@temp\@empty
+ \@tempcnta\z@
+% \end{macrocode}
+%
+% We wish to know how many authors are in each collaboration, for the purposes of \cmd\listcomma\ and \cmd\listand.
+%
+% Here we assign values for the intercollated operators appearing within \cmd\@AAC@list,
+% then execute that macro, registering the authors in each collaboration.
+% Afterwards, clean up by emulating an empty collaboration.
+%
+% The result, stored in \cmd\AU@temp, is a list of brace-delmited tokens, each a number
+% representing the number of authors in that collaboration.
+% \begin{macrocode}
+ \let\AF@opr \@gobble
+ \def\AU@opr{\@author@count\@tempcnta}%
+ \def\CO@opr{\@collaboration@count\AU@temp\@tempcnta}%
+ \@AAC@list
+ \expandafter\CO@opr\@author@cleared
+% \end{macrocode}
+% Next, present the authors and their affiliations.
+%
+% We assign values for the intercollated operators appearing within \cmd\@AAC@list.
+% Next, assign \cmd\@listcomma\ based on the number of authors in the first collaboration.
+% Then expand \cmd\@AAC@list.
+%
+% Upon completion, there will be two space tokens following the last author because of \cmd\listand\ processing;
+% remove them, then end the last paragraph.
+%
+% FIXME: this juncture would be a good time to effectively eliminate the unused affiliations in \cmd\@AFF@list.
+% \begin{macrocode}
+ \begingroup
+ \frontmatter@authorformat
+ \let\AF@opr \@affilID@def
+ \let\AU@opr \@author@present
+ \def\CO@opr{\@collaboration@present\AU@temp}%
+ \set@listcomma@list\AU@temp
+ \@AAC@list
+ \unskip\unskip
+ \par
+ \endgroup
+% \end{macrocode}
+% We now list out the affiliations in the order they appeared.
+% \begin{macrocode}
+ \begingroup
+ \frontmatter@above@affiliation@script
+% \let\@AFU@list\@empty
+% \let\AFS@opr \affils@present@script
+% \@AFG@list
+% \let\AF@opr \affil@present@script
+% \@AFU@list
+ \let\AFF@opr \@affil@script
+% \say\@AFF@list
+ \@AFF@list
+ \frontmatter@footnote@produce
+ \par
+ \endgroup
+ \endgroup
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\@author@count}
+% This version of \cmd\AU@opr\ counts the number of authors it processes.
+% \begin{macrocode}
+\def\@author@count#1{\advance#1\@ne\@author@gobble}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\@collaboration@present}
+% Format a collaboration.
+% Note that we immediately end the paragraph thus begun, because we only support
+% \classoption{superscriptaddress}-style processing.
+%
+% Note also that the execution of the production procedures takes place inside of a group;
+% for reasons of protective programming.
+%
+% Bug note: originally, the number of authors in each collaboration was stored in \cmd\@tempa,
+% but this exposed us to procedures in our production that modified the meaning of \cmd\@tempa.
+%
+% The last action done in this procedure is to assign \cmd\@listcomma\ based on the number of authors in the next collaboration.
+% \begin{macrocode}
+\def\@collaboration@present#1#2#3{%
+ \par
+ \begingroup
+ \frontmatter@collaboration@above
+ \@author@present{(\ignorespaces#2\unskip)}{#3}%
+ \par
+ \endgroup
+ \set@listcomma@list#1%
+}%
+\def\frontmatter@collaboration@above{}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\@collaboration@count}
+% Simply register each author in this collaboration.
+% Note: \#1 is a \cmd\csname\ to hold the value, \#2 is the \cmd\count\ register holding the count of interest.
+% \begin{macrocode}
+\def\@collaboration@count#1#2{%
+ \appdef@eval#1{\the#2}#2\z@
+ \@author@gobble
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\@affilID@def}
+%
+% The \cmd\@affilID@def\ procedure, an alias of \cmd\AF@opr\ used during \cmd\frontmatter@author@produce@script\ processing,
+% memorizes the affiliation in \cmd\@affilID@temp.
+% \begin{macrocode}
+\def\@affilID@def{\def\@affilID@temp}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\@affilID@temp}
+%
+% The macro \cmd\@affilID@temp\ is used to communicate between
+% \cmd\doauthor\ and instances of \cmd\AF@opr\ within the author list.
+% \begin{macrocode}
+\let\@affilID@temp\@empty
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\affils@present@script}
+% An alias of \cmd\AFS@opr, this procedure is applied to \cmd\@AFG@list\ and builds
+% \cmd\@AFU@list, a list of unique affiliations found within affiliation groups.
+%
+% \begin{macrocode}
+%\def\affils@present@script#1{%
+% \get@affil#1{}%
+%}%
+%\def\get@affil#1{%
+% \def\@tempa{#1}%
+% \@ifx{\@empty\@tempa}{}{%
+% \@booleanfalse\temp@sw
+% \def\AF@opr{\x@match\@tempa}%
+% \@AFU@list
+% \temp@sw{}{%
+% \appdef\@AFU@list{\AF@opr{#1}}%
+% }%
+% \get@affil
+% }%
+%}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\affil@present@script}
+% An alias of \cmd\AF@opr, this procedure is applied to \cmd\@AFU@list\ when formatting the affiliations
+% under \classoption{superscriptaddress}.
+% It in turn applies \cmd\affil@script\ to \cmd\@AFF@list, thereby formatting the affiliation
+% associated with this affiliation index.
+%
+% Note: it traverses the \cmd\@AFF@list\ for each entry in the \cmd\@AFU@list, thereby making this
+% portion of code execute in $N^{2}$ time.
+% \begin{macrocode}
+%\def\affil@present@script#1{%
+% \def\@tempa{#1}%
+% \begingroup
+% \frontmatter@affiliationfont
+% \let\AFF@opr \affil@script
+% \@AFF@list
+% \endgroup
+%}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\affil@script}
+% Alias of \cmd\AFF@opr, this procedure is applied to \cmd\@AFF@list\ when formatting the affiliations
+% under \classoption{superscriptaddress}.
+% It rejects all entries not matching the affiliation index held in \cmd\@tempa.
+% \begin{macrocode}
+\def\affil@script#1#2#3{%
+ \def\@tempifx{#1}\@ifx{\@tempifx\@tempa}{%
+ \@if@empty{#2}{}{%
+ \par
+ \begingroup
+ \textsuperscript{\normalfont#1}%
+ #2%
+ \@if@empty{#3}{}{\frontmatter@footnote{#3}}%
+ \par
+ \endgroup
+ }%
+ }{}%
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\@affil@script}
+% Alias of \cmd\AFF@opr:
+% The affiliations are being formatted in the case
+% where affiliations are being superscripted:
+% make a list out of the affiliations with the numbers printed.
+% Society-specific code can change the formatting
+% by overriding the definition of \cmd\frontmatter@affiliationfont.
+%
+% FIXME: dead code.
+% \begin{macrocode}
+\def\@affil@script#1#2#3{%
+ \@ifnum{#1=\z@}{}{%
+ \par
+ \begingroup
+ \frontmatter@affiliationfont
+% \textsuperscript{\normalfont#1}%
+ \textsuperscript{#1}%
+ #2%
+ \@if@empty{#3}{}{\frontmatter@footnote{#3}}%
+ \par
+ \endgroup
+ }%
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\@author@present@script}
+% \changes{v4.h}{1998/06/25}
+% {Changes to support groups of affiliations}
+% This version of \cmd\AU@opr\ applies to the \classoption{superscriptaddress} class option.
+% Need to add commas between groups
+% of address numbers, which are passed in the macro \cmd\@affilID@temp\ in the form
+% |{3}{4}{7}| if this set of authors is related to addresses, 3, 4 and 7.
+% \begin{macrocode}
+\def\@author@present@script#1#2{%
+ \begingroup
+% \parskip\@author@parskip\relax
+ \gdef\comma@space{\textsuperscript{,\,}}%
+% \@ifx{\@affilID@temp\@empty}{%
+% \doauthor{#1}{#2}{}%
+% }{%
+% \expandafter\set@count@\@affilID@temp{}{}\@@nil
+% \@ifnum{\count@=\z@}{%
+% \doauthor{#1}{#2}{}%
+% }{%
+% \doauthor{#1}{#2}{%
+% \textsuperscript{\expandafter\@affilcomma\@affilID@temp\relax\relax}%
+% }%
+% }%
+% }%
+ \doauthor{#1}{#2}{%
+ \let\@tempa\@empty
+ \expandafter\@affil@present@script\@affilID@temp\relax
+ }%
+ \endgroup
+ \advance\@tempcnta\m@ne
+}%
+%\def\set@count@#1#2\@@nil{%
+% \count@#1\relax
+%}%
+\def\@affilcomma#1#2{%
+ \@ifx{\z@#1}{%
+ \@ifx{\relax#2}{}{%
+ \@affilcomma{#2}%
+ }%
+ }{%
+ #1%
+ \@ifx{\relax#2}{}{%
+ \@ifx{\z@#2}{%
+ \@affilcomma
+ }{%
+ ,\,\@affilcomma{#2}%
+ }%
+ }%
+ }%
+}%
+\def\@affil@present@script#1{%
+ \@ifx{\relax#1}{%
+ \@ifx{\@tempa\@empty}{}{%
+ \textsuperscript{\expandafter\@affilcomma\@tempa\relax\relax}%
+ }%
+ }{%
+ \@ifx{\z@#1}{}{\appdef\@tempa{{#1}}}%
+ \@affil@present@script
+ }%
+}%
+\@provide\@author@parskip{\z@skip}%
+% \end{macrocode}
+% \end{macro}
+%
+%
+% \subsubsection{Authors and affiliations in groupedaddress style}
+%
+% \begin{macro}{\frontmatter@author@produce@group}
+% \cmd\frontmatter@author@produce@group\ is an alias for \cmd\frontmatter@author@produce,
+% the procedure that formats the author/affiliation list when \classoption{groupedaddress} is in effect.
+%
+% In this case, the authors are being grouped above their shared addresses.
+% How it works: the \cmd\@AFG@list, a list of all unique affiliation groups, is expanded.
+% For each such group, all authors belonging to that group are formatted above the affiliation group.
+% Collaborations are silently passed over.
+%
+% This procedure must ensure that any paragraph that it starts gets ended finally.
+% \begin{macrocode}
+\def\frontmatter@author@produce@group{%
+ \begingroup
+ \let\@author@present\@author@present@group
+ \frontmatter@authorformat
+ \let\AFS@opr \affils@present@group
+ \let\@listcomma\relax
+ \@AFG@list
+ \frontmatter@footnote@produce
+ \par
+ \endgroup
+ \frontmatter@authorbelow
+}%
+\@provide\frontmatter@authorbelow{}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\affils@present@group}
+% Alias of \cmd\AFS@opr, this procedure is used on every element of
+% \cmd\@AFG@list, the list of unique affiliation groups,
+% when formatting the author/affiliation block:
+% It presents all authors sharing this affiliation group.
+%
+% This procedure is called via the procedure
+% \cmd\frontmatter@author@produce@group, an alias for \cmd\frontmatter@author@produce,
+%
+% Its use pertains to the class options
+% \classoption{groupedaddress}, \classoption{unsortedaddress}, and \classoption{runinaddress}
+% (that is, \cmd\groupauthors@sw\ is true).
+%
+% For each invocation, the entire \cmd\@AAC@list\ is run through (twice),
+% making this process execute in $N^{2}$ time. (Is this really true?)
+% \begin{macrocode}
+\def\affils@present@group#1{%
+% \end{macrocode}
+% Count (using \cmd\@tempcnta) the authors in the author list whose affiliation is |#1|.
+%
+% Note that \cmd\AF@temp\ is used as a storage register for the argument
+% of the procedure, for purposes of comparison using \cmd\ifx.
+%
+% Note: here we use \cmd\@tempcnta\ to communicate accumulate the number of authors in the current affiliation set.
+%
+% QUERY: does the value of \cmd\AF@temp\ and \cmd\AF@opr\ need to persist?
+%
+% QUERY: what value does \cmd\AU@opr\ have at this point?
+% ANSWER: the \cmd\@affilID@count\ and \cmd\@affilID@match\ procedures
+% assign meanings to \cmd\AU@opr\ and \cmd\CO@opr\ dynamically.
+% We initialize the two operators to \cmd\undefined\ so as to catch the
+% pathological cases.
+% \begin{macrocode}
+ \begingroup
+ \def\AF@temp{#1}% % 0B13
+ \@tempcnta\z@
+ \let\AU@opr \undefined % 0B13
+ \let\CO@opr \undefined % 0B13
+ \def\AF@opr{\@affilID@count\AF@temp\@tempcnta}% 0B13
+ \@AAC@list
+% \end{macrocode}
+%
+% If no authors using this affiliation group, skip it.
+% \begin{macrocode}
+ \@ifnum{\@tempcnta=\z@}{}{%
+% \end{macrocode}
+%
+% DPC: If we have a list of length two, need to locally zap a comma.
+% \begin{macrocode}
+ \begingroup
+ \frontmatter@above@affilgroup
+ \set@listcomma@count\@tempcnta
+% \end{macrocode}
+% Run through the author list, presenting (with \cmd\@author@present)
+% those authors whose affiliation matched the given one.
+% \begin{macrocode}
+ \let\AU@opr \undefined % 0B13
+ \let\CO@opr \undefined % 0B13
+ \def\AF@opr{\@affilID@match\AF@temp}%
+ \@AAC@list
+ \endgroup
+% \end{macrocode}
+% Now that the authors have all been presented, present the affiliations, grouped.
+%
+% Note: \cmd\@tempcnta\ is used to count the number of affiliations for this set of authors.
+% \begin{macrocode}
+ \begingroup
+ \par
+ \frontmatter@above@affiliation
+ \frontmatter@affiliationfont
+ \let\\\frontmatter@addressnewline
+ \@tempcnta\z@
+ \@tfor\AF@temp:=#1\do{%
+ \expandafter\@ifx\expandafter{\z@\AF@temp}{}{%
+ \advance\@tempcnta\@ne
+ }%
+ }%
+ \@ifnum{\@tempcnta=\tw@}{%
+ \let\@listcomma\relax
+ }{}%
+ \def\after@address{\@listcomma\ \@listand}%
+ \runinaddress@sw{%
+ }{%
+ \tightenlines@sw{%
+ }{%
+ \parskip\z@ %space between affilations in grouped mode
+ }%
+ \appdef\after@address\par
+ }%
+ \let\AFF@opr \@affil@group
+ \do@affil@fromgroup\@AFF@list#1\relax
+ \endgroup
+% \end{macrocode}
+% End of branch handling authors.
+% \begin{macrocode}
+ }%
+% \end{macrocode}
+%
+% \begin{macrocode}
+ \par
+ \endgroup
+%\smallskip
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\@affilID@count}
+% \begin{macro}{\@affilID@match}
+%
+% The control sequence name \cmd\AF@opr\ is placed in the
+% author/affiliation list (\cmd\@AAC@list) for each (new) affiliation, with that affiliation as its argument.
+%
+% The control sequence name \cmd\AF@opr\ is an alias of a procedure
+% to count the number of authors at a given affiliation (\cmd\@affilID@count)
+% or to present the given authors (\cmd\@affilID@match).
+% These two procedures are executed when formatting under the \classoption{groupedaddress} class option.
+% \begin{macrocode}
+\def\@affilID@count#1#2#3{%
+ \def\@tempifx{#3}%
+ \@ifx{\@tempifx#1}{%
+ \def\AU@opr{\@author@count#2}%
+ }{%
+ \let\AU@opr \@author@gobble
+ }%
+ \let\CO@opr \@collaboration@err
+}%
+\def\@affilID@match#1#2{%
+ \def\@tempifx{#2}%
+ \@ifx{\@tempifx#1}{%
+ \let\AU@opr \@author@present
+ }{%
+ \let\AU@opr \@author@gobble
+ }%
+ \let\CO@opr \@collaboration@err
+}%
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\do@affil@fromgroup}
+% Executes iteratively:
+% selects the next address ID to print,
+% and then re-execute the list of addresses to print the text of that address.
+%
+% Note: an argument of \cmd\relax\ breaks out of this iteration,
+% while an argument of \cmd\z@\ (\cmd\noaffiliation) is a no-op.
+%
+% Note: we have createed an $N^{2}$ computing process.
+% FIXME: can use hashing instead, you know.
+% \begin{macrocode}
+\def\do@affil@fromgroup#1#2{%
+ \@ifx{\relax#2}{}{%
+ \count@#2\relax
+ \@ifnum{\count@=\z@}{}{#1}%
+ \do@affil@fromgroup#1%
+ }%
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\@affil@group}
+% The alias of \cmd\AFF@opr\ when the list \cmd\@AFF@list\
+% is executed by \cmd\do@affil@fromgroup (\classoption{groupedaddress}):
+% Print the address if its number matches \cmd\count@\ (value set by \cmd\do@affil@fromgroup).
+%
+% Note that we detect a \cmd\noaffiliation\ here by the second parameter's being the token \cmd\blankaffiliation.
+% \begin{macrocode}
+\def\@affil@group#1#2#3{%
+ \@ifnum{\count@=#1\relax}{%
+ \@ifx{#2\blankaffiliation}{}{%
+ #2%
+ \@if@empty{#3}{}{%
+ \frontmatter@footnote{#3}%
+ }%
+ \after@address
+ }%
+ \advance\@tempcnta\m@ne
+ }{}%
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\@author@present@group}
+% \changes{v4.d}{1998/01/31}
+% {\cs{comma@space}}
+% This version of \cmd\AU@opr\ applies to the \classoption{groupedaddress} class option.
+% Traverse the list calling the \cmd\doauthor\ hook on each author.
+% \begin{macrocode}
+\def\@author@present@group#1#2{%
+ \gdef\comma@space{\gdef\comma@space{\textsuperscript{,\,}}}%
+ \doauthor{#1}{#2}{}%
+ \advance\@tempcnta\m@ne
+}%
+% \end{macrocode}
+% \end{macro}
+%
+%
+% \begin{macro}{\@pacs@produce}
+% PACS, keywords and dates.
+% \changes{v4.d}{1998/01/31}
+% {PACS boxed to width of abstract}
+% \changes{v4.0beta 2}{1999/06/20}
+% {Separate option now controls production}
+% \changes{v4.0beta 5}{2000/09/20}
+% {Warn if no production}
+% \begin{macrocode}
+\def\@pacs@produce#1{%
+ \showPACS@sw{%
+ \begingroup
+ \frontmatter@PACS@format
+ \@pacs@name#1\par
+ \endgroup
+ }{%
+ \@if@empty{#1}{}{%
+ \class@warn{\PACS@warn}%
+ }%
+ }%
+}%
+\def\PACS@warn{If you want your PACS to appear in your output, use document class option showpacs}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\@keywords@produce}
+% \changes{v4.0beta 5}{2000/09/20}
+% {Keywords on same footing as PACS}
+% \begin{macrocode}
+\def\@keywords@produce#1{%
+ \showKEYS@sw{%
+ \begingroup
+ \frontmatter@keys@format
+ \@keys@name#1\par
+ \endgroup
+ }{%
+ \@if@empty{#1}{}{%
+ \class@warn{If you want your keywords to appear in your output, use document class option showkeys}%
+ }%
+ }%
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\frontmatter@footnote@produce@footnote}
+% \begin{macro}{\frontmatter@footnote@produce@endnote}
+% Produce any footnotes to appear at the bottom of the title page.
+%
+% If frontmatter footnotes have been redirected to the bibliography, this will be a no-op.
+% ref.: \cmd\present@bibnote.
+% \begin{macrocode}
+\def\frontmatter@footnote@produce@footnote{%
+ \let\@TBN@opr\present@FM@footnote
+ \@FMN@list
+ \global\let\@FMN@list\@empty
+}%
+\def\present@FM@footnote#1#2{%
+ \begingroup
+ \csname c@\@mpfn\endcsname#1\relax
+ \def\@thefnmark{\thempfn}%
+ \frontmatter@footnotetext{#2}%
+ \endgroup
+}%
+\def\frontmatter@footnote@produce@endnote{%
+}%
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\frontmatter@thefootnote}
+% The in the title block, the footnote symbol (used for alternate affiliation and sundry)
+% is a \cmd\@fnsymbol\ or a lowercase letter, as selected by a boolean.
+% \begin{macrocode}
+\def\frontmatter@thefootnote{%
+ \altaffilletter@sw{\@alph}{\@fnsymbol}\c@footnote
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\frontmatter@makefnmark}
+% The default procedure for setting the footnote mark within the frontmatter.
+% Journals will override this procedure.
+% \begin{macrocode}
+\def\frontmatter@makefnmark{%
+ \@textsuperscript{%
+ \normalfont\@thefnmark
+ }%
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\frontmatter@makefntext}
+% This procedure is used when setting footnotes within the title block.
+% \begin{macrocode}
+\long\def\frontmatter@makefntext#1{%
+ \parindent 1em
+ \noindent
+ \Hy@raisedlink{\hyper@anchorstart{frontmatter@\the\c@footnote}\hyper@anchorend}%
+ \@makefnmark
+ #1%
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\frontmatter@setup}
+% The default type specification for the title page.
+% Journals will override this definition.
+% \begin{macrocode}
+\def\frontmatter@setup{}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\frontmatter@RRAPformat}
+% \begin{macro}{\punct@RRAP}
+% \begin{macro}{\produce@RRAP}
+% The procedure \cmd\frontmatter@RRAPformat\ is the formatting procedure
+% for the ``Received, Revised, etc.'' block on the title page.
+% The macro \cmd\punct@RRAP\ produces the punctuation between the entries.
+% Journals will override these two definitions.
+%
+% The procedure \cmd\produce@RRAP\ is fairly general, and should be used by most
+% journal substyles.
+%
+% The procedure \cmd\frontmatter@RRAP@format\ will be overridden by the journal.
+% \begin{macrocode}
+\def\frontmatter@RRAPformat#1{%
+ \removelastskip
+ \begingroup
+ \frontmatter@RRAP@format
+ #1\par
+ \endgroup
+}%
+\def\punct@RRAP{; }%
+\def\produce@RRAP#1{%
+ \@if@empty{#1}{}{%
+ \@ifvmode{\leavevmode}{\unskip\punct@RRAP\ignorespaces}%
+ #1%
+ }%
+}%
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\frontmatter@authorformat}
+% The default type specification for the author list.
+% Journals will override this definition.
+% \begin{macrocode}
+\def\frontmatter@authorformat{}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\frontmatter@above@affilgroup}
+% The default amount of space above an affiliation group
+% (in a \cmd\affils@present@group\ production).
+% Journals will override this definition.
+% \begin{macrocode}
+\def\frontmatter@above@affilgroup{}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\frontmatter@above@affiliation}
+% \begin{macro}{\frontmatter@above@affiliation@script}
+% The \cmd\frontmatter@above@affiliation\ is the amount of space above affiliations in the \classoption{groupedaddress}
+% style, and \cmd\frontmatter@above@affiliation@script\ is that for \classoption{superscriptaddress}.
+% Journals will override this definition.
+% \begin{macrocode}
+\def\frontmatter@above@affiliation{}%
+\def\frontmatter@above@affiliation@script{}%
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\frontmatter@affiliationfont}
+% The default type specification for the affiliation.
+% Journals will override this definition.
+% \begin{macrocode}
+\def\frontmatter@affiliationfont{}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\frontmatter@RRAP@format}
+% The default type specification for the dates.
+% Journals will override this definition.
+% \begin{macrocode}
+\def\frontmatter@RRAP@format{}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\frontmatter@PACS@format}
+% Specify the formatting of the title page PACS statement.
+%
+% Journals will override.
+% \begin{macrocode}
+\def\frontmatter@PACS@format{}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\frontmatter@keys@format}
+% Specify the formatting of the title page keywords statement.
+%
+% Journals will override.
+% \begin{macrocode}
+\def\frontmatter@keys@format{}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\frontmatter@finalspace}
+%
+% \begin{macrocode}
+\def\frontmatter@finalspace{\addvspace{18\p@}}
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\frontmatter@addressnewline}
+% \changes{v4.h}{1998/06/25}
+% {new hook}
+% The definition of \cmd\\\ for address handling. Default puts all the `lines'
+% on a run-in line, separated by comma and space.
+% DPC: was \newline space between lines of addresss.
+% \begin{macrocode}
+\def\frontmatter@addressnewline{%
+ \@ifhmode{\skip@\lastskip\unskip\unpenalty\break\hskip\skip@}{}%
+ % was: \vskip-.5ex
+}%
+% \end{macrocode}
+% \end{macro}
+%
+%
+% \begin{macro}{\frontmatter@preabstractspace}
+% \begin{macro}{\frontmatter@postabstractspace}
+% Space above and space below abstract in title block
+% \begin{macrocode}
+\def\frontmatter@preabstractspace{5.5\p@}
+\def\frontmatter@postabstractspace{6.5\p@}
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\aftermaketitle@chk}
+% Error-checking code.
+% \begin{macrocode}
+\def\aftermaketitle@chk#1{%
+ \@ifx{\maketitle\relax}{%
+ \class@err{\protect#1 must be used before \protect\maketitle}%
+ }{}%
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\ps@titlepage}
+% Default page style for title page.
+% Journals will override this procedure.
+% \begin{macrocode}
+\def\ps@titlepage{\ps@empty}%
+% \end{macrocode}
+% \end{macro}
+%
+%FIXME: there is a limitation to the default meaning of
+% \cmd\@startpage, the label ``FirstPage'' is only
+% defined if the document has a \cmd\maketitle\ command.
+%
+% \begin{macrocode}
+\def\volumeyear#1{\gdef\@volumeyear{#1}}%
+\def\@volumeyear{}%
+\def\volumenumber#1{\gdef\@volumenumber{#1}}%
+\def\@volumenumber{}%
+\def\issuenumber#1{\gdef\@issuenumber{#1}}%
+\def\@issuenumber{}%
+\def\eid#1{\gdef\@eid{#1}}%
+\def\@eid{}%
+%
+\def\startpage#1{\gdef\@startpage{#1}\c@page#1\relax}%
+\def\@startpage{\pageref{FirstPage}}%
+\def\endpage#1{\gdef\@endpage{#1}}%
+\def\@endpage{\pageref{LastPage}}%
+% \end{macrocode}
+%
+%
+% \subsection{Printing out the ``list-of'' elements}%
+%
+% FIXME: The \cmd\appendix@toc\ procedure should change the meaning
+% of \cmd\l@section\ so that the \cmd\section s can be appropriately formatted,
+% reflecting their status as appendices.
+%
+% \begin{macrocode}
+\def\print@toc#1{%
+ \begingroup
+% \c@secnumdepth-\maxdimen
+ \expandafter\section
+ \expandafter*%
+ \expandafter{%
+ \csname#1name\endcsname
+ }%
+ \let\appendix\appendix@toc
+ \@starttoc{#1}%
+ \endgroup
+}%
+\def\appendix@toc{}%
+% \end{macrocode}
+%
+%
+% \section{Body}
+%
+% \subsection{counters}
+% The following definitions are probably identical to those in
+% \file{classes.dtx}%
+% \begin{macrocode}
+\def\labelenumi{\theenumi.}
+\def\theenumi{\arabic{enumi}}
+% \end{macrocode}
+%
+% \begin{macrocode}
+\def\labelenumii{(\theenumii)}
+\def\theenumii{\alph{enumii}}
+\def\p@enumii{\theenumi}
+% \end{macrocode}
+%
+% \begin{macrocode}
+\def\labelenumiii{\theenumiii.}
+\def\theenumiii{\roman{enumiii}}
+\def\p@enumiii{\theenumi(\theenumii)}
+% \end{macrocode}
+%
+% \begin{macrocode}
+\def\labelenumiv{\theenumiv.}
+\def\theenumiv{\Alph{enumiv}}
+\def\p@enumiv{\p@enumiii\theenumiii}
+% \end{macrocode}
+%
+% \begin{macrocode}
+\def\labelitemi{\textbullet}
+\def\labelitemii{\normalfont\bfseries\textendash}
+\def\labelitemiii{\textasteriskcentered}
+\def\labelitemiv{\textperiodcentered}
+% \end{macrocode}
+%
+% \begin{macrocode}
+\pagenumbering{arabic}
+% \end{macrocode}
+%
+% \subsection{float parameters}
+%
+% from the old aps.sty. (DPC: same as article I think)
+% AO: here, \LaTeX's standard classes fail very poorly
+% (the price of backward compatability): the values for
+% \cmd\floatpagefraction\ and \cmd\dblfloatpagefraction\
+% need to be raised to avoid creating extremely short
+% float pages.
+% \begin{macrocode}
+\setcounter{topnumber}{2}
+\def\topfraction{.9}
+\setcounter{bottomnumber}{1}
+\def\bottomfraction{.9}
+\setcounter{totalnumber}{3}
+\def\textfraction{.1}
+\def\floatpagefraction{.9}
+\setcounter{dbltopnumber}{2}
+\def\dbltopfraction{.9}
+\def\dblfloatpagefraction{.9}
+% \end{macrocode}
+%
+% \subsection{List Environments}
+%
+% \begin{macrocode}
+\newenvironment{verse}{%
+ \let\\=\@centercr
+ \list{}{%
+ \itemsep\z@ \itemindent -1.5em\listparindent \itemindent
+ \rightmargin\leftmargin\advance\leftmargin 1.5em}\item[]%
+}{%
+ \endlist
+}%
+% \end{macrocode}
+%
+% \begin{macrocode}
+\newenvironment{quotation}{%
+ \list{}{%
+ \listparindent 1.5em
+ \itemindent\listparindent
+ \rightmargin\leftmargin \parsep \z@ \@plus\p@}\item[]%
+}{%
+ \endlist
+}%
+% \end{macrocode}
+%
+% \begin{macrocode}
+\newenvironment{quote}{%
+ \list{}{%
+ \rightmargin\leftmargin}\item[]%
+}{%
+ \endlist
+}%
+% \end{macrocode}
+%
+% \begin{macrocode}
+\def\descriptionlabel#1{%
+ \hspace\labelsep \normalfont\bfseries #1%
+}%
+% \end{macrocode}
+%
+% \begin{macrocode}
+\newenvironment{description}{%
+ \list{}{%
+ \labelwidth\z@ \itemindent-\leftmargin
+ \let\makelabel\descriptionlabel
+ }%
+}{%
+ \endlist
+}%
+% \end{macrocode}
+%
+% \subsection{Sectioning Commands}
+%
+% \subsubsection{Counters and Their Productions}%
+% \begin{macrocode}
+\newcounter{part}
+\newcounter{section}
+\newcounter{subsection}[section]
+\newcounter{subsubsection}[subsection]
+\newcounter{paragraph}[subsubsection]
+\newcounter{subparagraph}[paragraph]
+% \end{macrocode}
+%
+% \begin{macrocode}
+\def\thepart {\Roman{part}} %
+\def\thesection {\Roman{section}}
+\def\p@section {}
+\def\thesubsection {\Alph{subsection}}
+\def\p@subsection {\thesection\,}
+\def\thesubsubsection {\arabic{subsubsection}}
+\def\p@subsubsection {\thesection\,\thesubsection\,}
+\def\theparagraph {\alph{paragraph}}
+\def\p@paragraph {\thesection\,\thesubsection\,\thesubsubsection\,}
+\def\thesubparagraph {\arabic{subparagraph}}
+\def\p@subparagraph {\thesection\,\thesubsection\,\thesubsubsection\,\theparagraph\,}
+% \end{macrocode}
+%
+% \subsubsection{The Acknowledgments Environment}%
+% This user-level markup produces a head introducing
+% the acknowledgments, and acts as a wrapper for the text.
+% In this implementation, it is an unnumbered section,
+% but appears within the toc.
+%
+% For compatiability's sake, we implement it under the alternative
+% spelling \env{acknowledgements}.
+% \changes{v4.0beta 3}{1999/11/13}
+% {also spelled ``acknowledgements''.}
+% \changes{v4.0beta 4}{2000/05/18}
+% {But alternative spelling is deprecated.}
+% \begin{macrocode}
+\newenvironment{acknowledgments}{%
+ \acknowledgments@sw{%
+ \section*{\acknowledgmentsname}%
+ }{%
+ \par
+ \phantomsection
+ \addcontentsline{toc}{section}{\protect\numberline{}\acknowledgmentsname}%
+ }%
+}{%
+ \par
+}%
+\@booleantrue\acknowledgments@sw
+\newenvironment{acknowledgements}{%
+ \replace@environment{acknowledgements}{acknowledgments}%
+}{%
+ \endacknowledgments
+}%
+% \end{macrocode}
+%
+% \subsubsection{Part Opener}%
+% section setup copied verbatim from revtex3 aps/osa.
+% Does not explicitly depend on pointsize options.
+% \begin{macrocode}
+\def\part{\par
+ \addvspace{4ex}%
+ \@afterindentfalse
+ \secdef\@part\@spart}%
+% \end{macrocode}
+%
+% \begin{macrocode}
+\def\@part[#1]#2{%
+ \@ifnum{\c@secnumdepth >\m@ne}{%
+ \refstepcounter{part}%
+ \addcontentsline{toc}{part}{\thepart\hspace{1em}#1}%
+ }{%
+ \addcontentsline{toc}{part}{#1}%
+ }%
+ \begingroup
+ \parindent \z@ \raggedright
+ \interlinepenalty\@M
+ \@ifnum{\c@secnumdepth >\m@ne}{%
+ \Large \bf \partname~\thepart%
+ \par\nobreak
+ }{}%
+ \huge \bf
+ #2%
+ \markboth{}{}\par
+ \endgroup
+ \nobreak
+ \vskip 3ex
+ \@afterheading
+}%
+% \end{macrocode}
+%
+% \begin{macrocode}
+\def\@spart#1{{\parindent \z@ \raggedright
+ \interlinepenalty\@M
+ \huge \bf
+ #1\par}
+ \nobreak
+ \vskip 3ex
+ \@afterheading}
+% \end{macrocode}
+%
+% \subsubsection{Stacked Heads}%
+% Here are the class default definitions for sectioning commands.
+% A society or a journal substyle will likely override these definitions.
+%
+% In doing so, you can customize the formatting for a particular level by defining,
+% e.g., \cmd\@hangfrom@section\ or \cmd\@subsectioncntformat.
+% \begin{macrocode}
+\def\section{%
+ \@startsection
+ {section}%
+ {1}%
+ {\z@}%
+ {0.8cm \@plus1ex \@minus .2ex}%
+ {0.5cm}%
+ {\normalfont\small\bfseries}%
+}%
+% \end{macrocode}
+%
+% \begin{macrocode}
+\def\subsection{%
+ \@startsection
+ {subsection}%
+ {2}%
+ {\z@}%
+ {.8cm \@plus1ex \@minus .2ex}%
+ {.5cm}%
+ {\normalfont\small\bfseries}%
+}%
+% \end{macrocode}
+%
+% \begin{macrocode}
+\def\subsubsection{%
+ \@startsection
+ {subsubsection}%
+ {3}%
+ {\z@}%
+ {.8cm \@plus1ex \@minus .2ex}%
+ {.5cm}%
+ {\normalfont\small\itshape}%
+}%
+% \end{macrocode}
+%
+% \subsubsection{Runin Heads}%
+% \begin{macrocode}
+\def\paragraph{%
+ \@startsection
+ {paragraph}%
+ {4}%
+ {\parindent}%
+ {\z@}%
+ {-1em}%
+ {\normalfont\normalsize\itshape}%
+}%
+% \end{macrocode}
+%
+% \begin{macrocode}
+\def\subparagraph{%
+ \@startsection
+ {subparagraph}%
+ {5}%
+ {\parindent}%
+ {3.25ex \@plus1ex \@minus .2ex}%
+ {-1em}%
+ {\normalfont\normalsize\bfseries}%
+}%
+% \end{macrocode}
+%
+%
+% \subsection{Math}
+%
+% \begin{macro}{\theequation}
+% We change the production of the equation counter so that
+% we can accomodate the \classoption{eqsecnum} option.
+% \begin{macrocode}
+\def\theequation{%
+ \theequation@prefix\arabic{equation}%
+}%
+\def\theequation@prefix{}%
+% \end{macrocode}
+% \end{macro}
+%
+%
+% \subsection{Type Size-Dependent Settings}
+%
+% Note that many parameters, as well as the meaning of
+% \cmd\normalsize, are deferred until
+% \cmd\AtEndOfClass\ time.
+% Therefore, code elsewhere in this class \emph{should not}
+% assume these things' values have been set!
+% \begin{macrocode}
+\AtEndOfClass{%
+ \expandafter\input\expandafter{\@pointsize pt\substyle@ext}%
+}%
+% \end{macrocode}
+%
+%
+% \subsection{All Point Sizes}
+% \begin{macrocode}
+\setcounter{secnumdepth}{4}
+% \end{macrocode}
+%
+% \begin{macrocode}
+\lineskip 1pt
+\normallineskip 1pt
+\def\baselinestretch{1}%
+% \end{macrocode}
+%
+% \begin{macrocode}
+\@lowpenalty 51
+\@medpenalty 151
+\@highpenalty 301
+% \end{macrocode}
+%
+% \begin{macrocode}
+\@beginparpenalty -\@lowpenalty
+\@endparpenalty -\@lowpenalty
+\@itempenalty -\@lowpenalty
+% \end{macrocode}
+%
+% \begin{macrocode}
+\arraycolsep 3pt
+\tabcolsep 2pt
+\arrayrulewidth .4pt
+\doublerulesep 2pt
+% \end{macrocode}
+%
+% \begin{macrocode}
+\skip\@mpfootins = 0pt
+% \end{macrocode}
+%
+% \begin{macrocode}
+\fboxsep = 3.0pt
+\fboxrule = 0.4pt
+% \end{macrocode}
+%
+% \subsection{Figures}
+% \changes{v4.0beta 2}{1999/06/20}
+% {Processing delayed to \cs{AtBeginDocument} time}
+%
+% \begin{environment}{figure}
+% We define the \env{figure} environment.
+% Later, we will horse around with its meaning
+% in order to accomodate \cmd\floats@sw.
+% \begin{macrocode}
+\newenvironment{figure}
+ {\@float{figure}}
+ {\end@float}
+\newenvironment{figure*}
+ {\@dblfloat{figure}}
+ {\end@dblfloat}
+% \end{macrocode}
+% \end{environment}
+%
+% \begin{macro}{\@makecaption}
+% If caption is one line long, to be centered; if lines turn, then set justified.
+% \begin{macrocode}
+\newlength\abovecaptionskip
+\newlength\belowcaptionskip
+\setlength\abovecaptionskip{10\p@}
+\setlength\belowcaptionskip{2\p@}
+% \end{macrocode}
+%
+% \changes{v4.0beta 2}{1999/06/20}
+% {One-line caption sets flush left.}
+% \changes{v4.0beta 3}{1999/11/13}
+% {Prevent an inner footnote from performing twice}
+% \begin{macrocode}
+\long\def\@makecaption#1#2{%
+ \par
+% \nobreak
+ \vskip\abovecaptionskip
+ \begingroup
+ \small\rmfamily
+ \sbox\@tempboxa{%
+ \let\\\heading@cr
+ #1: #2%
+ }%
+ \@ifdim{\wd\@tempboxa >\hsize}{%
+ \begingroup
+ \samepage
+ \flushing
+ \let\footnote\@footnotemark@gobble
+ #1: #2\par
+ \endgroup
+ }{%
+ \global \@minipagefalse
+ \hb@xt@\hsize{\hfil\unhbox\@tempboxa\hfil}%
+ }%
+ \endgroup
+ \vskip\belowcaptionskip
+}%
+\def\@footnotemark@gobble{%
+ \@footnotemark
+ \@ifnextchar[{\@gobble@opt@i}{\@gobble}%
+}%
+\def\@gobble@opt@i[#1]#2{}%
+\def\@mpmakefntext#1{%
+ \flushing
+ \parindent=1em
+ \noindent
+ \hb@xt@1em{\hss\@makefnmark}%
+ #1%
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\thefigure}
+% The figure counter and float placement defaults.
+% \begin{macrocode}
+\newcounter{figure}
+\renewcommand \thefigure {\@arabic\c@figure}
+% \end{macrocode}
+% Note that we give the `!' modifier by default.
+% This is an effort to avoid the syndrome wherein
+% a deferred float finds itself unqualified for
+% placement, thereby getting carried until \cmd\clearpage.
+% \begin{macrocode}
+\def\fps@figure{tbp}
+\def\ftype@figure{1}
+\def\ext@figure{lof}
+\def\fnum@figure{\figurename~\thefigure}
+% \end{macrocode}
+% \end{macro}
+%
+% \subsubsection{Deferring \env{figure} Floats}%
+%
+% We install a hook at \cmd\AtBeginDocument\ time which determines if
+% figures will float or will be deferred until
+% \cmd\printfigures\ time.
+% \begin{macrocode}
+\appdef\class@documenthook{%
+ \do@if@floats{figure}{.fgx}%
+}%
+\appdef\class@enddocumenthook{%
+ \printfigures\relax
+}%
+% \end{macrocode}
+%
+% \begin{macro}{\printfigures}
+% The user-level command \cmd\printfigures\ determines
+% where the figures are to appear in a document in which
+% \cmd\floats@sw\ is false.
+% If the user invokes the \classoption{endfloats} class option and fails to
+% insert a \cmd\printfigures\ command,
+% the figures will be printed at the end of the document.
+% If the command is given, but floats are not being deferred,
+% it amounts to a no-op.
+% \changes{v4.0beta 2}{1999/06/20}
+% {call \cs{print@floats}}
+% \changes{v4.0beta 3}{1999/11/13}
+% {*-form mandates pagebreak}
+% \begin{macrocode}
+\newcommand\printfigures{%
+ \@ifstar{\true@sw}{\floatp@sw{\true@sw}{\false@sw}}%
+ {%
+ \print@float{figure}{\oneapage}%
+ }{%
+ \print@float{figure}{}%
+ }%
+}%
+% \end{macrocode}
+% \end{macro}
+%
+%
+% \begin{macro}{\@xfloat@prep}
+% We patch into the procedure \cmd\@xfloat@prep.
+% This patch applies to all floats (not \env{figure} alone)
+% and makes the type center.
+% \begin{macrocode}
+\appdef\@xfloat@prep{%
+ \appdef\@parboxrestore{\centering}%
+ \let\@makefnmark\@makefnmark@latex
+ \let\@footnotemark\@footnotemark@latex
+}%
+% \end{macrocode}
+% \end{macro}
+%
+%
+% \subsection{Tables}
+% \changes{v4.0beta 2}{1999/06/20}
+% {Processing delayed to \cs{AtBeginDocument} time}
+%
+% DPC: More or less taken from \file{revtex2} \file{aps.sty},
+% but using dcolumn for decimal alignment.
+% \begin{environment}{table}
+% We define the \env{table} environment.
+% Later, we will horse around with its meaning
+% in order to accomodate \cmd\floats@sw.
+% \begin{macrocode}
+\newenvironment{table}
+ {\@float{table}}
+ {\end@float}
+\newenvironment{table*}
+ {\@dblfloat{table}}
+ {\end@dblfloat}
+% \end{macrocode}
+% \end{environment}
+%
+% \begin{macro}{\thetable}
+% Table counter and default float placement declarations.
+% \begin{macrocode}
+\newcounter{table}
+\renewcommand\thetable{\@Roman\c@table}
+% \end{macrocode}
+% Note that we give the `!' modifier by default.
+% This is an effort to avoid the syndrome wherein
+% a deferred float finds itself unqualified for
+% placement, thereby getting carried until \cmd\clearpage.
+% \begin{macrocode}
+\def\fps@table{tbp}
+\def\ftype@table{2}
+\def\ext@table{lot}
+\def\fnum@table{\tablename~\thetable}
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\table@hook}
+% \begin{macro}{\squeezetable}
+% Assign a meaning to the hook installed into float processing.
+%
+% By default floats are \cmd\small. The \cmd\squeezetable\ declaration
+% makes them smaller (\cmd\scriptsize). In general you can locally
+% redefine \cmd\table@hook\ to be whatever you like.
+% (DPC: |\Huge\color{magenta}|\ldots?)
+% \begin{macrocode}
+\def\table@hook{\small}%
+\def\squeezetable{\def\table@hook{\scriptsize}}%
+\appdef\@floatboxreset{\table@hook}%
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \subsubsection{Deferring \env{table} Floats}%
+%
+% We install a hook at
+% \cmd\AtBeginDocument\ time which determines if
+% tables will float or will be deferred until
+% \cmd\printtables\ time.
+%
+% We also deal with the possibility of \env{longtable} environments.
+%
+% \begin{macrocode}
+\appdef\class@documenthook{%
+ \floats@sw{}{%
+ \let@environment{longtable@float}{longtable}%
+ \let@environment{longtable}{longtable@write}%
+ \let@environment{longtable*@float}{longtable*}%
+ \let@environment{longtable*}{longtable*@write}%
+ \let@environment{turnpage@float}{turnpage}%
+ \let@environment{turnpage}{turnpage@write}%
+ }%
+ \do@if@floats{table}{.tbx}%
+}%
+\appdef\class@enddocumenthook{%
+ \printtables\relax
+}%
+\newenvironment{longtable@write}{%
+ \write@@float{longtable}{table}%
+}{%
+ \endwrite@float
+}%
+\newenvironment{longtable*@write}{%
+ \write@@float{longtable*}{table}%
+}{%
+ \endwrite@float
+}%
+\newenvironment{turnpage@write}{%
+ \immediate\write\tablewrite{\string\begin{turnpage}}%
+}{%
+ \immediate\write\tablewrite{\string\end{turnpage}}%
+}%
+% \end{macrocode}
+%
+% \begin{macro}{\printtables}
+% The user-level command \cmd\printtables\ determines
+% where the tables are to appear in a document in which
+% \cmd\floats@sw\ is false.
+% If the user invokes the \classoption{nofloats} and fails to
+% insert a \cmd\printtables\ command,
+% the tables will be printed at the end of the document.
+% If the command is given, but floats are not being deferred,
+% it amounts to a no-op.
+% \changes{v4.0beta 2}{1999/06/20}
+% {only execute if appropriate}
+% \changes{v4.0beta 3}{1999/11/13}
+% {*-form mandates pagebreak}
+% \changes{v4.0beta 4}{2000/05/17}
+% {make longtable trigger the head, too}
+% \begin{macrocode}
+\newcommand\printtables{%
+ \begingroup
+ \let@environment{longtable}{longtable@float}%
+ \let@environment{longtable*}{longtable*@float}%
+ \let@environment{turnpage}{turnpage@anchored}%
+ \prepdef\longtable{\trigger@float@par}%
+ \expandafter\prepdef\csname longtable*\endcsname{\trigger@float@par}%
+ \expandafter\prepdef\csname table@floats\endcsname{%
+ \onecolumngrid@push
+ }%
+ \expandafter\appdef\csname endtable@floats\endcsname{%
+ \onecolumngrid@pop
+ }%
+ \@ifstar{\true@sw}{\floatp@sw{\true@sw}{\false@sw}}%
+ {%
+ \print@float{table}{\oneapage}%
+ }{%
+ \print@float{table}{}%
+ }%
+ \endgroup
+}%
+\newenvironment{turnpage@anchored}{%
+ \onecolumngrid@push
+ \setbox\z@\vbox to\textwidth\bgroup
+ \columnwidth\textheight
+}{%
+ \vfil
+ \egroup
+ \rotatebox{90}{\box\z@}%
+ \onecolumngrid@pop
+}%
+% \end{macrocode}
+% \end{macro}
+%
+%
+%
+% \section{Tabular}
+% Every tabular has a double (Scotch) rule above and below.
+% The column specifier ``d'' is implemented using the \classname{dcolumn}
+% package, if available.
+% FIXME: always load dcolumn!
+% \begin{macro}{\tabular@hook}
+% \begin{macro}{\endtabular@hook}
+% \begin{environment}{ruledtabular}
+% \begin{macrocode}
+\def\endtabular@hook{}%
+%\RequirePackage{dcolumn}%
+\appdef\class@documenthook{%
+ \@ifpackageloaded{dcolumn}{%
+ \newcolumntype{d}{D{.}{.}{-1}}%
+ }{}%
+}%
+\def\toprule{\hline\hline}%
+\def\colrule{\hline}%
+\def\botrule{\hline\hline}%
+\newenvironment{ruledtabular}{%
+ \def\array@default{v}%
+ \appdef\tabular@hook{\def\@halignto{to\hsize}}%
+ \let\tableft@skip@default\tableft@skip
+ \let\tableft@skip\tableft@skip@float
+ \let\tabmid@skip@default\tabmid@skip
+ \let\tabmid@skip\tabmid@skip@float
+ \let\tabright@skip@default\tabright@skip
+ \let\tabright@skip\tabright@skip@float
+ \let\array@row@pre@default\array@row@pre
+ \let\array@row@pre\array@row@pre@float
+ \let\array@row@pst@default\array@row@pst
+ \let\array@row@pst\array@row@pst@float
+ \appdef\array@row@rst{%
+ \let\array@row@pre\array@row@pre@default
+ \let\array@row@pst\array@row@pst@default
+ \let\tableft@skip\tableft@skip@default
+ \let\tabmid@skip\tabmid@skip@default
+ \let\tabright@skip\tabright@skip@default
+ \appdef\tabular@hook{\let\@halignto\@empty}%
+ }%
+}{%
+}%
+% \end{macrocode}
+% \end{environment}
+% \end{macro}
+% \end{macro}
+%
+% \section{Footnote Text}
+%
+% \changes{v4.g}{1998/06/10}
+% {single space footnotes}
+% \changes{v4.0beta 2}{1999/06/20}
+% {%
+% Frank Mittelbach, has stated in \protect\classname{multicol}:
+% ``The kernel command \cs{@footnotetext} should not be modified.''
+% Thus, I have removed David Carlisle's redefinition of that command.
+% Note, however, that later versions of \protect\classname{multicol} do not require
+% this workaround. Belt and suspenders.
+% }%
+%
+% \begin{macro}{\@makefntext}
+% \begin{macrocode}
+\def\@makefntext#1{%
+ \def\baselinestretch{1}%
+ \reset@font\footnotesize
+ \parindent 1em%
+ \noindent
+ \hb@xt@1.8em{%
+ \Hy@raisedlink{\hyper@anchorstart{footnote@\the\c@footnote}\hyper@anchorend}%
+ \hss\@makefnmark
+ }%
+ #1\par
+}%
+% \end{macrocode}
+% \end{macro}
+%
+%
+% \subsection{Citations, Bibliography, Endnotes}
+%
+% \subsubsection{Bibliography}
+%
+% Load Patrick Daly's \classname{natbib} package,
+% \url{ftp://ctan.tug.org/macros/latex/contrib/supported/natbib}
+%
+% Note that \classname{natbib} assumes that it loads
+% over a document class, such as the \classname{article} class,
+% that has already defined \env{thebibliography} and \cmd\@listi.
+%
+% Note also that \classname{natbib} also installs a command
+% \cmd\NAT@set@cites\ into \cmd\AtBeginDocument\ which presumes
+% that the proper \cmd\bibpunct\ command has been issued.
+%
+% We have a problem with \classname{natbib} in that it does not
+% accomodate late binding of its package options. This means
+% that those options must be known at the time the package
+% loads. Therefore, we select \classoption{sort\&compress}
+% now, and thereby foreclose on alternative uses of \classname{natbib}.
+%
+% Certain societies may prefer an alternative; they are out of luck.
+% If you find yourself in this position, please let the author
+% of \classname{natbib} know that his package should be modified to
+% accomodate late binding of package options.
+%
+% To remedy the situation, note that the macro \cmd\NAT@sort\ controls
+% whether citations are left alone (\cmd\NAT@sort=0), sorted (\cmd\NAT@sort=1), or sorted and compressed (\cmd\NAT@sort=2).
+% Since we give \classname{natbib} the \classoption{sort\&compress} option,
+% if you prefer \classoption{sort}, you need only define \cmd\NAT@sort\ to be 1.
+% However, if you prefer the effect of having neither \classoption{sort} nor \classoption{sort\&compress},
+% you must define \cmd\NAT@sort\ to be 0 and you must also define
+% \cmd\NAT@sort@cites:
+%\begin{verbatim}
+%\renewcommand\NAT@sort@cites[1]{\edef\NAT@cite@list{#1}}%
+%\end{verbatim}
+%
+% For other \classname{natbib} customizations, you may proceed as if you were going to use
+% the \file{natbib.cfg} file: anything that you can modify by this means is fair game.
+% Once \revtex\ is finished loading, you can assert any definitions for \classname{natbib} that you wish.
+%
+% \begin{macrocode}
+\newenvironment{thebibliography}{}{}%
+\let\@listi\@empty
+\RequirePackage[sort&compress]{natbib}[1999/05/28 7.0]%
+\let\NAT@thebibliography\thebibliography
+\let\NAT@endthebibliography\endthebibliography
+\let\bibliographystyle@latex\bibliographystyle
+% \end{macrocode}
+%
+% \begin{macro}{\NAT@citesuper}
+% We redefine \classname{natbib}'s procedure \cmd\NAT@citesuper,
+% which is executed when setting a superscript citation.
+% The \cmd\hspace\ is removed: in any case, it
+% should really be \cmd\hspace*, to prevent an unwanted pagebreak.
+% \begin{macrocode}
+\renewcommand\NAT@citesuper[3]{%
+ \ifNAT@swa
+ \leavevmode
+ \unskip
+% \hspace{1\p@}%
+ \textsuperscript{#1}%
+ \if*#3*\else\ (#3)\fi
+ \else
+ #1%
+ \fi
+ \endgroup
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\@makefnmark@cite}
+% We define a procedure that will set a footnote mark the same way that
+% a citation is set. If footnotes are put in the bibliography with
+% \cmd\footinbib@sw, then the corresponding mark should look the same
+% as the result of a \cmd\cite. This is how we do it.
+% \begin{macrocode}
+\def\@makefnmark@cite{\begingroup\NAT@swatrue\@cite{{\@thefnmark}}{}{}}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\bibliographystyle}
+% We arrange for the selection of bibliography style
+% to occur either due to the document's explicit
+% \cmd\bibliographystyle\ statement or
+% via the journal substyle.
+%
+% Note that \revtex\ is now incompatible with
+% any package that patches \cmd\bibliographystyle,
+% except for those (like \classname{natbib}) that load before here.
+% \begin{macrocode}
+\let\bibliographystyle@latex\bibliographystyle
+\def\bibliographystyle{\def\@bibstyle}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\bibsection}
+% \begin{macro}{\NAT@bibsetnum}
+% \begin{macro}{\NAT@bibsetup}
+% \begin{macro}{\bibpreamble}
+% \begin{macro}{\newblock}
+% \begin{macro}{\bibnumfmt}
+% We define the sectioning command to use when starting the bibliography
+% and gently coax \classname{natbib} into using
+% the formatting procedures that \emph{we} want it to use.
+%
+% This way of setting up \env{thebibliography}
+% automatically sets the label width based on
+% the largest number used within the bibliography.
+% This scheme will not work properly using the
+% author/year style of bib entry, though.
+%
+% We define \cmd\bibnumfmt\ to be \cmd\place@bibnumber, which is a macro
+% managed by \revtex. If the document defines \cmd\bibnumfmt, then that
+% definition will be used instead, which is what the \classname{natbib}
+% package gives as its programming interface.
+%
+% And if the substyle does not give a meaning to \cmd\place@bibnumber,
+% we give it that of the original \cmd\bibnumfmt.
+% \begin{macrocode}
+\def\bibsection{%
+ \@ifx@empty\refname{%
+ \par
+ }{%
+ \let\@hangfroms@section\@hang@froms
+ \section*{\refname}%
+ \@nobreaktrue
+ }%
+}%
+\let\NATx@bibsetnum\NAT@bibsetnum
+\def\NAT@bibsetnum#1{%
+ \setlength{\topsep}{\z@}%
+ \NATx@bibsetnum{\ref{LastBibItem}}%
+}%
+\let\NATx@bibsetup\NAT@bibsetup
+\def\NAT@bibsetup{%
+ \setlength{\labelwidth}{\z@}%
+ \setlength{\labelsep}{\z@}%
+ \setlength{\itemindent}{\z@}%
+ \setlength{\listparindent}{\z@}%
+ \setlength{\topsep}{\z@}%
+ \setlength{\parsep}{\z@}%
+ \NATx@bibsetup
+}%
+\let\bibpreamble\@empty
+\def\newblock{\ }%
+\let\NATx@bibnumfmt\bibnumfmt
+\def\bibnumfmt{\place@bibnumber}%
+\AtEndOfClass{%
+ \appdef\class@documenthook{%
+ \@ifxundefined\place@bibnumber{%
+ \let\place@bibnumber\NATx@bibnumfmt
+ }{}%
+ }%
+}%
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\NAT@onlinecite}
+% We anticipate using numerical citations.
+% If superscript citations are used, we nonetheless need a way to
+% use numerical citations as required by the author.
+% These should be accessible via the \cmd\citet\ command.
+%
+% Therefore we remember how to do a numerical citation
+% even when the superscript citation has been selected.
+%
+% Can you say ``inelegant hack''? I knew you could!
+% \begin{macrocode}
+\let\NAT@citeyear\citeyear
+\def\NAT@onlinecite#1{%
+ \begingroup
+ \let\@cite\NAT@citenum
+ \let\NAT@mbox\mbox
+ \let\citeyear\NAT@citeyear
+ \def\NAT@space{ }%
+ \cite{#1}%
+ \endgroup
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{+}
+% \begin{macro}{\textcite}
+% We extend \classname{natbib}'s syntax with two commands to set a citation
+% on the baseline (as opposed to superscripted) and as text (rather than parenthetical), respectively.
+% A journal substyle that makes citations be superscripted or parenthetical as the case may be, should ensure
+% that the author has continued access to these two styles.
+%
+% \begin{macrocode}
+\appdef\class@inithook{%
+ \@ifxundefined\@onlinecite{\def\@onlinecite{\citealp}}{}%
+ \@ifxundefined\@textcite{\def\@textcite{\citet}}{}%
+}%
+\DeclareRobustCommand\onlinecite{\@onlinecite}%
+\DeclareRobustCommand\textcite{\@textcite}%
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{environment}{thebibliography}%
+% \begin{macro}{\present@bibnote}%
+% We put a tail patch into \cmd\thebibliogrphy\ and
+% a headpatch into \cmd\endthebibliography.
+%
+% We also apply a fix to the \cmd\list\ procedure
+% so that, at least within the bibliography,
+% items are not formatted via \cmd\parshape,
+% which caused problems during column balancing.
+% (Other list-like environments may also require fixing,
+% but we will save that for another day.)
+%
+% Here we provide a default treatment for frontmatter notes
+% deferred to the bibliography; a journal substyle
+% might want to override the definition of \cmd\present@bibnote.
+% \begin{macrocode}
+\renewenvironment{thebibliography}[1]{%
+ \NAT@thebibliography{#1}%
+ \@clubpenalty\clubpenalty
+ \let\@TBN@opr\present@bibnote
+ \@FMN@list
+}{%
+ \@endnotesinbib
+ \edef\@currentlabel{\arabic{NAT@ctr}}%
+ \label{LastBibItem}%
+ \NAT@endthebibliography
+ \global\let\auto@bib\@empty
+}%
+\def\present@bibnote#1#2{%
+ \item[%
+ \textsuperscript{%
+ \normalfont
+ \Hy@raisedlink{\hyper@anchorstart{frontmatter@#1}\hyper@anchorend}%
+ \begingroup
+ \c@footnote#1\relax
+ \frontmatter@thefootnote
+ \endgroup
+ }%
+ ]#2\par
+}%
+% \end{macrocode}
+% \end{macro}
+% \end{environment}
+%
+% \begin{macro}{\place@bibnumber@sup}
+% \begin{macro}{\place@bibnumber@inl}
+% For producing the \cmd\bibitem\ device,
+% we define two procedures to select from.
+%
+% Note that we could have used \classname{natbib}'s \cmd\ifNAT@super\
+% switch, but it does not allow for altering the meaning of \cmd\bibnumfmt.
+% \begin{macrocode}
+\def\place@bibnumber@sup#1{\textsuperscript{#1}}%
+\def\place@bibnumber@inl#1{[#1]}%
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+%% \begin{macro}{\@cite}
+%% \begin{macro}{\nocite}
+% \begin{macro}{write@bibliographystyle}
+% We wish to delay committing the \cmd\bibliographystyle\ until as late as possible.
+% The journal substyle will define a default bibliography style, and
+% the document's explicit \cmd\bibliographystyle\ command, if any, will override that default.
+%% For the sake of Bib\TeX, we must execute the \cmd\bibliographystyle\ command
+%% (putting a \cmd\bibstyle\ command into the auxiliary file)
+%% somewhere before the first \cmd\cite\ command (or any other like command),
+%% which puts a \cmd\citation\ command into the auxiliary file.
+%%
+%% Therefore, we headpatch the two procedures in \classname{natbib} that can result in
+%% the appearance in the auxiliary file of a \cmd\citation\ command.
+%% These two procedures are \cmd\@cite\ and \cmd\nocite.
+%%
+%% Not!
+%
+% The \cmd\bibstyle\ command may appear quite late in the \filename{.aux} file.
+% We now delay the automatic writing of the \cmd\bibstyle\ command to the end of the job.
+%
+% The procedure \cmd\write@bibliographystyle\ tests whether a \cmd\bibliographystyle\ command has
+% already been given.
+% If not, it effectively executes the needed \cmd\bibliographystyle\ command, then neutralizes itself
+% (we only need to do this once per job).
+%
+% \begin{macrocode}
+%\AtBeginDocument{%
+% \let\natbib@@cite\@cite
+% \def\@cite{\write@bibliographystyle\natbib@@cite}%
+% \let\natbib@nocite\nocite
+% \def\nocite{\write@bibliographystyle\natbib@nocite}%
+%}%
+\def\write@bibliographystyle{%
+ \@ifxundefined\@bibstyle{}{%
+ \expandafter\bibliographystyle@latex\expandafter{\@bibstyle}%
+ }%
+ \global\let\write@bibliographystyle\relax
+}%
+\AtEndDocument{\write@bibliographystyle}%
+% \end{macrocode}
+% \end{macro}
+%% \end{macro}
+%% \end{macro}
+%
+% \begin{macro}{\super@cite}
+% \begin{macro}{\super@cite@}
+% The \cmd\super@cite\ procedure takes the place of \cmd\@citex\ and implements the feature of the
+% \classoption{citeautoscript} class option.
+% This means that the document should be marked up as if citations were \emph{not} superscripted,
+% and then if you select a journal substyle that has superscripted citations, \revtex\ will do its best
+% to alter the formatting of the \cmd\cite s to accomodate superscript style.
+%
+% Note that the commands \cmd\onlinecite\ and \cmd\textcite\ are not affected by this procedure.
+%
+% Here's a subtle one: when is the argument of \cmd\super@cite@@\ not the same as the token \cmd\@let@token?
+% Answer: when the latter is \cmd\@sptoken! This case has to be handled separately.
+%
+% \begin{macrocode}
+\def\super@cite[#1][#2]#3{%
+ \leavevmode
+ \skip@\lastskip
+ \unskip
+ \begingroup
+ \def\@tempa{[#1][#2]{#3}}%
+ \super@cite@
+}%
+\def\super@cite@{%
+ \futurelet\@let@token\super@cite@@
+}%
+\long\def\super@cite@@#1{%
+ \@ifx{\@let@token\@sptoken}{%
+ \aftergroup\@sptoken\false@sw
+ }{%
+ \@ifx{\@let@token.}{\true@sw}{%
+ \@ifx{\@let@token,}{\true@sw}{%
+ \@ifx{\@let@token'}{\true@sw}{%
+ \@ifx{\@let@token)}{\true@sw}{%
+ \false@sw
+ }%
+ }%
+ }%
+ }%
+ }%
+ {%
+ #1%
+ \super@cite@
+ }{%
+ \expandafter\@citex@NAT\@tempa\endgroup
+ #1%
+ }%
+}%
+\AtBeginDocument{%
+ \citeautoscript@sw{%
+ \@ifx{\place@bibnumber\place@bibnumber@sup}{%
+ \let\@citex@NAT\@citex\let\@citex\super@cite
+ }{}%
+ }{}%
+}%
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+%
+% Resolve an incompatability between \classname{natbib} and \classname{listings}.
+% The latter package tests \cmd\chapter (which has now been \cmd\let\ to \cmd\relax\ as a side effect \classname{natbib}'s use of \LaTeX's \cmd\@ifundefined).
+%
+% We couch our fix in such terms that will not be disruptive if \cmd\chapter\ is actually defined at this point.
+% \begin{macrocode}
+\@ifx{\chapter\relax}{\let\chapter\@undefined}{}%
+% \end{macrocode}
+%
+%
+% \subsubsection{\cs{endnote}s and \cs{rtx@bibnote}s}
+% FIXME: check hyperref compatibility!
+%
+% \begin{macro}{\mini@note}
+% \begin{macro}{\save@note}
+% \begin{macrocode}
+\def\mini@note{\save@note\mini@notes}%Implicit #1
+\def\save@note#1#2{%
+ \stepcounter{footnote}%
+ \protected@xdef\@thefnmark{\the\c@footnote}%
+ \@footnotemark
+ \expandafter\g@addto@macro
+ \expandafter#1%
+ \expandafter{%
+ \expandafter \@@footnotetext
+ \expandafter {\@thefnmark}{#2}%
+ }%
+}%
+\long\def\@@footnotetext#1{\def\@thefnmark{#1}\@footnotetext}%
+\let\mini@notes\@empty
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\endnote}
+% A version of footnote that appears at the end of the document.
+% Actually it just appears where \cmd\printendnotes\ appears.
+% \begin{macrocode}
+\def\endnote{%
+ \begingroup
+ \aftergroup\@footnotemark
+ \aftergroup\@endnotetext
+ \@ifnextchar[{%
+ \@xendnote
+ }{%
+ \stepcounter{footnote}%
+ \protected@xdef\@tempa{\thefootnote}%
+ \expandafter\@xendnote\expandafter[\@tempa]%
+ }%
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\@xendnote}
+% \begin{macrocode}
+\def\@xendnote[#1]{%
+ \begingroup
+ \c@footnote#1\relax
+ \unrestored@protected@xdef\@thefnmark{endnote\thefootnote}%
+ \endgroup
+ \endgroup
+}%
+\def\@endnotemark{%
+ \expandafter\cite\expandafter{\@thefnmark}%
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\endnote@ext}
+% \begin{macrocode}
+\def\endnote@ext{.end}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\@endnotetext}
+% The commands \cmd\label, \cmd\index, and \cmd\glossary,
+% which are robustified for \cmd\markright\ and \cmd\addcontentsline,
+% are likewise robustified here.
+%
+% Note the similarity to \cmd\protected@write.
+% \begin{macrocode}
+\long\def\@endnotetext#1{%
+ \@ifxundefined\@endnoteout{%
+ \newwrite\@endnoteout
+ \gdef\endnote@stream{\jobname\endnote@ext}%
+ \immediate\openout\@endnoteout\endnote@stream\relax
+ }{}%
+ \begingroup
+ \let\label\relax \let\index\relax \let\glossary\relax
+ \let\cite \relax \let\ref \relax \let\pageref \relax
+ \let\( \relax \let\) \relax \let\\ \relax
+% \let\protect\noexpand
+ \let \protect \@unexpandable@protect
+ \newlinechar`\^^M%
+% \newlinechar`\ %
+ \immediate\write\@endnoteout{\string\@doendnote{\@thefnmark}{#1}}%
+ \endgroup
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\endnotetext}
+% \begin{macrocode}
+\def\endnotetext{%
+ \@ifnextchar[{%
+ \@xendnotenext
+ }{%
+ \protected@xdef\@thefnmark{\thefootnote}%
+ \@endnotetext
+ }%
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\@xendnotenext}
+% \begin{macrocode}
+\def\@xendnotenext[#1]{%
+ \begingroup
+ \c@footnote#1\relax
+ \unrestored@protected@xdef\@thefnmark{\thefootnote}%
+ \endgroup
+ \@endnotetext
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\printendnotes}
+% \begin{macrocode}
+\def\printendnotes{%
+ \@ifx{\@endnoteout\@undefined}{}{%
+ \begingroup
+ \c@secnumdepth-\maxdimen
+ \section{%
+ \notesname
+ }%
+ \immediate\closeout\@endnoteout
+ \global\let\@endnoteout\@undefined
+ \begin{enumerate}%
+ \makeatletter
+ \input{\endnote@stream}%
+ \end{enumerate}%
+ \endgroup
+ }%
+}%
+\appdef\class@enddocumenthook{%
+ \printendnotes
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\firstnote@num}
+% \begin{macro}{\set@firstnote}
+%
+% In \revtex, we have permanently altered the way
+% that the footnote counter gets reset at the beginning
+% of the document.
+%
+% If footnotes appear in the bibliography, we initialise
+% the footnote counter to number of cites (found last time)
+% via \cmd\set@firstnote.
+%
+% \begin{macrocode}
+\mathchardef\firstnote@num\z@
+\appdef\class@documenthook{%
+ \global\c@footnote\firstnote@num
+ \appdef\maketitle{\global\c@footnote\firstnote@num}%
+}%
+\def\set@firstnote#1{%
+ \@ifnum{\firstnote@num=#1\relax}{}{%
+ \class@warn@end{Endnote numbers changed: rerun LaTeX}%
+ }%
+ \immediate\write\@mainaux{%
+ \global\chardef\string\firstnote@num#1\relax
+ }%
+}%
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\@endnotesinbib}
+% \begin{macro}{\@endnotesinbibliography}
+% We define a function \cmd\@endnotesinbib,
+% and a variant \cmd\@endnotesinbibliography.
+% The former is invoked at the start of
+% the end processing for \enve{thebibliography};
+% the latter is a synonym.
+%
+% The procedure typesets the footnotes that are to
+% appear in the bibliography; the default is to
+% simply arrange for the footnote counter to be
+% reset at the start of the document.
+%
+% The switchover to setting footnotes in the bibliography
+% is done via \cmd\AtBeginDocument\ code that
+% changes the meaning of \cmd\footnote\ and
+% that substitutes the synonym for
+% \cmd\@endnotesinbib.
+%
+% Note that this code make the assumption that
+% the counter used in \env{thebibliography} is \cmd\c@NAT@ctr.
+%
+% \begin{macrocode}
+\def\@endnotesinbib{%
+ \set@firstnote\z@
+}%
+\def\@endnotesinbibliography{%
+ \expandafter\set@firstnote\expandafter{\the\c@NAT@ctr}%
+ \@ifx{\@endnoteout\@undefined}{}{%
+ \immediate\closeout\@endnoteout
+ \global\let\@endnoteout\@undefined
+ \makeatletter
+ \input{\endnote@stream}%
+ }%
+}%
+\appdef\class@documenthook{%
+ \let\@makefnmark@latex\@makefnmark
+ \let\@footnotemark@latex\@footnotemark
+ \footinbib@sw{%
+ \let\footnote\endnote
+ \let\@footnotemark\@endnotemark
+ \let\@makefnmark\@makefnmark@cite
+ \let\@endnotesinbib\@endnotesinbibliography
+ \appdef\class@enddocumenthook{\auto@bib}%
+ \let\printendnotes\relax
+ }{}%
+}%
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\auto@bib}
+% Under some circumstances, we must typeset the bibliography automatically.
+% For instance, the document specifies footnotes in the bibliography, but contains
+% no bibliography.
+%
+% Note that this facility is not able to work more than once per document.
+% If multiple bibliographys are required (e.g., per article), it will be the responsibility of the journal style
+% to restore \cmd\auto@bib\ to its original meaning so it can be re-invoked.
+% \begin{macrocode}
+\def\auto@bib{%
+ \@ifx@empty\@FMN@list{%
+ \@ifxundefined\@endnoteout{%
+ \false@sw
+ }{%
+ \true@sw
+ }%
+ }{%
+ \true@sw
+ }{%
+ \begin{thebibliography}{}\end{thebibliography}%
+ }%
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\@doendnote}
+% \begin{macrocode}
+\def\@doendnote#1#2{\bibitem{#1}#2}
+% \end{macrocode}
+% \end{macro}
+%
+%
+%
+% \section{Initial setup}
+% \changes{v4.0beta 2}{1999/06/20}
+% {Defer assignment until \cs{AtBeginDocument} time.}
+% Note that this code is executed at \cmd\AtBeginDocument\ time
+% in order to handle a particular timing problem.
+% Please do not move this code any earlier!
+% \begin{macrocode}
+\appdef\class@documenthook{%
+% \end{macrocode}
+% We assert our own page style only if nobody else has done so.
+% Users wishing to customize their documents will be able to invoke a \cmd\pagestyle\
+% command anywhere in the preamble.
+% \begin{macrocode}
+ \@ifxundefined\@evenhead{%
+ \@ifxundefined\@oddhead{%
+ \@ifxundefined\@oddfoot{%
+ \@ifxundefined\@evenfoot{%
+ \preprintsty@sw{%
+ \ps@preprint
+ }{%
+ \ps@article
+ }%
+ }{}%
+ }{}%
+ }{}%
+ }{}%
+ \pagenumbering{arabic}%
+ \normalsize
+ \tabbingsep \labelsep
+ \leftmargin\leftmargini
+ \labelwidth\leftmargin\advance\labelwidth-\labelsep
+ \let\@listi\@listI
+ \@listi
+% \end{macrocode}
+%
+% \begin{macrocode}
+%\twoside@sw{}{%
+% \raggedbottom
+%}%
+% \end{macrocode}
+% We make sure that the ``environment'' component mark is set.
+% \begin{macrocode}
+ \def\curr@envir{document}%
+ \mark@envir{\curr@envir}%
+% \end{macrocode}
+%
+% \changes{v4.0beta 3}{1999/11/13}
+% {Grid changes with ltxgrid}
+% \changes{v4.0beta 4}{2000/04/10}
+% {Also alter how lists get indented.}
+% When setting the column grid, we have to alter the procedure
+% for formatting lists. Because \cmd\twocolumngrid\ requires
+% rebalancing columns at some points, lists must \emph{not}
+% involve the use of \cmd\moveleft, \cmd\moveright, or \cmd\parshape.
+% The alternative procedure only involves the manipulation of
+% \cmd\leftskip\ and \cmd\rightskip.
+%
+% FIXME: Always sloppy??
+% \begin{macrocode}
+ \twocolumn@sw{%
+ \open@column@mlt\tw@
+ \set@colht
+ \@floatplacement
+ \@dblfloatplacement
+ \sloppy
+ \@booleanfalse\raggedcolumn@sw
+ \let\set@listindent\set@listindent@
+ }{%
+ \open@column@one\@ne
+ \set@colht
+ \@floatplacement
+ \@dblfloatplacement
+ \sloppy
+% \raggedbottom
+ \let\set@listindent\set@listindent@parshape
+ }%
+}%
+% \end{macrocode}
+%
+%
+% \section{\cs{appendix}}
+%
+% \begin{macrocode}
+%\newif\ifappendixon
+% \end{macrocode}
+% Note that, within appendices,
+% equations are numbered within sections (appendices).
+% \begin{macrocode}
+\def\appendix{%
+ \par
+%\appendixontrue
+ \setcounter{section}\z@
+ \setcounter{subsection}\z@
+ \setcounter{subsubsection}\z@
+ \def\thesubsection{\arabic{subsection}}%
+ \def\thesubsubsection{\alph{subsubsection}}%
+ \@addtoreset{equation}{section}%
+ \def\theequation@prefix{\thesection}%
+ \addtocontents{toc}{\protect\appendix}%
+ \@ifstar{%
+ \def\thesection{\unskip}%
+ \def\theequation@prefix{A.}%
+ }{%
+ \def\thesection{\Alph{section}}%
+ }%
+}%
+% \end{macrocode}
+%
+% \section{Changing the page grid}
+% \changes{v4.0beta 2}{1999/06/20}
+% {Define three separate environments,
+% defer assignment to \cs{AtBeginDocument} time.
+% }
+%
+% \subsection{Avoiding Grid Changes}%
+%
+% In preprint styles, ``wide text'' is a no-op, and the title page
+% processing involves no grid change.
+%
+% \begin{macro}{\title@column@default}%
+% \begin{macro}{\close@column@default}%
+%
+% \begin{macrocode}
+\def\title@column@default#1{%
+ \minipagefootnote@init
+ #1%
+ \minipagefootnote@foot
+}%
+\def\close@column@default{%
+ \newpage
+}%
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+%
+% \subsection{Galley Style: Margin Changes}%
+% A variant of preprint processing. Emulate journal appearance somewhat.
+%
+% \begin{environment}{widetext@galley}
+% DPC: We're in galley style so do a lob sided display environment.
+%
+% QUERY: How can we be sure that we are in galley style?
+% ANSWER: as noted elsewhere, require both \cmd\twocolumn@sw\ and \cmd\preprintsty@sw\ to be false.
+% \begin{macrocode}
+\def\galley@outdent{\rightmargin-\columnwidth\advance\rightmargin-\columnsep}%
+\let\widetext@outdent\@empty
+\newenvironment{widetext@galley}{%
+ \list{}{%
+ \topsep \z@skip
+ \listparindent \parindent
+ \itemindent \parindent
+ \leftmargin \z@
+ \parsep \z@\@plus\p@
+ \widetext@outdent
+ \relax
+ }%
+ \item\relax
+}{
+ \endlist
+}%
+% \end{macrocode}
+% \end{environment}
+%
+%
+% \subsection{Grid Changing Via \classname{ltxgrid}}%
+%
+% In case \classoption{twocolumngrid} has been invoked,
+% switch column grid using the column grid-changing commands.
+% Supply stub definitions of those commands here.
+% \changes{v4.0beta 3}{1999/11/13}
+% {grid changes with ltxgrid}
+%
+% \begin{macro}{\title@column@grid}%
+% \begin{macro}{\close@column@grid}%
+% The title block always starts at the top of a new page.
+% \changes{v4.0beta 3}{1999/11/13}
+% {grid changes with ltxgrid}
+%
+% Note that, for the procedure \cmd\close@column@grid,
+% we balance columns by switching to the one-column page grid,
+% then we return to a two-column page grid in anticipation of the
+% upcoming \cmd\clearpage.
+% \begin{macrocode}
+\def\title@column@grid#1{%
+ \minipagefootnote@init
+ \onecolumngrid
+ \begingroup
+ \let\@footnotetext\frontmatter@footnotetext
+ \let\set@footnotewidth\set@footnotewidth@ii
+ #1%
+ \endgroup
+ \twocolumngrid
+ \minipagefootnote@foot
+}%
+\def\close@column@grid{%
+ \balancelastpage@sw{%
+ \onecolumngrid
+% \twocolumngrid
+ }{}%
+}%
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{environment}{widetext@grid}
+% \changes{v4.0beta 3}{1999/11/13}
+% {grid changes via ltxgrid procedures}
+% \changes{v4.0beta 5}{2000/11/21}
+% {adornments above and below.}
+% We slip into the one-column page grid
+% within the scope of this environment.
+%
+% Note that we set adornments above and below the \env{widettext}.
+% These are set as leaders, so they will disappear at a page break.
+% \begin{macrocode}
+\newenvironment{widetext@grid}{%
+ \par\ignorespaces
+ \setbox\widetext@top\vbox{%
+ \vskip15\p@
+ \hb@xt@\hsize{%
+ \leaders\hrule\hfil
+ \vrule\@height6\p@
+ }%
+ \vskip6\p@
+ }%
+ \setbox\widetext@bot\hb@xt@\hsize{%
+ \vrule\@depth6\p@
+ \leaders\hrule\hfil
+ }%
+ \onecolumngrid
+ \dimen@\ht\widetext@top\advance\dimen@\dp\widetext@top
+ \cleaders\box\widetext@top\vskip\dimen@
+ \let\set@footnotewidth\set@footnotewidth@ii
+}{%
+ \par
+ \setbox\widetext@bot\vbox{%
+ \hb@xt@\hsize{\hfil\box\widetext@bot}%
+ \vskip14\p@
+ }%
+ \dimen@\ht\widetext@bot\advance\dimen@\dp\widetext@bot
+ \cleaders\box\widetext@bot\vskip\dimen@
+ \twocolumngrid\global\@ignoretrue
+ \@endpetrue
+}%
+\newbox\widetext@top
+\newbox\widetext@bot
+% \end{macrocode}
+% \end{environment}
+%
+%
+% Decide, finally, how the page grid is to be manipulated.
+% \begin{macrocode}
+\appdef\class@documenthook{%
+ \twocolumn@sw{%
+ \let@environment{widetext}{widetext@grid}%
+ \let\title@column\title@column@grid
+ \let\close@column\close@column@grid
+ }{%
+ \let@environment{widetext}{widetext@galley}%
+ \preprintsty@sw{%
+% \end{macrocode}
+% Change the page grid not at all.
+% \begin{macrocode}
+ }{%
+% \end{macrocode}
+% If we are galley style, change the page margin only.
+% \begin{macrocode}
+ \galley@sw{%
+ \let\widetext@outdent\galley@outdent
+ }{%
+ }%
+ }%
+ }%
+}%
+% \end{macrocode}
+%
+% \begin{macro}{\title@column}
+% Provide a default meaning for \cmd\title@column\
+% in case it was never defined.
+% \begin{macrocode}
+\appdef\class@inithook{%
+ \@ifxundefined\title@column{%
+ \let\title@column\title@column@default
+ }{}%
+}%
+% \end{macrocode}
+% \end{macro}
+%
+%
+% \section{Old font commands}
+%
+% \begin{macrocode}
+\DeclareOldFontCommand{\rm}{\normalfont\rmfamily}{\mathrm}
+\DeclareOldFontCommand{\sf}{\normalfont\sffamily}{\mathsf}
+\DeclareOldFontCommand{\tt}{\normalfont\ttfamily}{\mathtt}
+\DeclareOldFontCommand{\bf}{\normalfont\bfseries}{\mathbf}
+\DeclareOldFontCommand{\it}{\normalfont\itshape}{\mathit}
+\DeclareOldFontCommand{\sl}{\normalfont\slshape}{\@nomath\sl}
+\DeclareOldFontCommand{\sc}{\normalfont\scshape}{\@nomath\sc}
+\DeclareRobustCommand*\cal{\@fontswitch\relax\mathcal}
+\DeclareRobustCommand*\mit{\@fontswitch\relax\mathnormal}
+% \end{macrocode}
+%
+% \section{English-Language Texts}
+% As this class is just for English language journals, we
+% could hardwire these texts,
+% but to make it easier to use this as a
+% basis for the code for similar journal styles, separate out all the
+% fixed text strings into babel-style macros of the form
+% |\|\ldots|name|
+%
+% Note: for babel compatability, use version 1999/05/05 v3.6x or later.
+%
+% Some of these might need changing in the society-specific code.
+%
+% \begin{macro}{\today}
+% Procedure \cmd\today\ is used in the article class, but not in
+% this document class.
+% \begin{macrocode}
+\def\today{\ifcase\month\or
+ January\or February\or March\or April\or May\or June\or
+ July\or August\or September\or October\or November\or December\fi
+ \space\number\day, \number\year}
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\notesname}
+% Text entity \cmd\notesname\ is used in \cmd\printendnotes.
+% \begin{macrocode}
+\def\notesname{Notes}
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\partname}
+% Text entity \cmd\partname\ is used in \cmd\@part.
+% \begin{macrocode}
+\def\partname{Part}
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\tocname}
+% Text entity \cmd\tocname\ is used in \cmd\tableofcontents,
+% as defined in the standard \LaTeX\ book class.
+% \begin{macrocode}
+\def\tocname{Contents}
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\lofname}
+% Text entity \cmd\lofname\ is used in \cmd\listoffigures,
+% as defined in the standard \LaTeX\ book class.
+% \begin{macrocode}
+\def\lofname{List of Figures}
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\lotname}
+% Text entity \cmd\lotname\ is used in \cmd\listoftables,
+% as defined in the standard \LaTeX\ book class.
+% \begin{macrocode}
+\def\lotname{List of Tables}
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\refname}
+% Text entity \cmd\refname\ is used in \env{thebibliography}.
+% \begin{macrocode}
+\def\refname{References}
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\indexname}
+% Text entity \cmd\indexname\ is used in \env{theindex},
+% as defined in the standard \LaTeX\ book class.
+% \begin{macrocode}
+\def\indexname{Index}
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\figurename}
+% \changes{v4.0beta 2}{1999/06/20}
+% {Added localization of \cs{figuresname}}
+% Text entity \cmd\figurename\ is used in \env{figure},
+% \cmd\figuresname\ in \cmd\printfigures.
+% \begin{macrocode}
+\def\figurename{FIG.}
+\def\figuresname{Figures}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\tablename}
+% \changes{v4.0beta 2}{1999/06/20}
+% {Added localization of \cs{tablesname}}
+% Text entity \cmd\tablename\ is used in \env{table},
+% \cmd\tablesname\ in \cmd\printtables.
+% \begin{macrocode}
+\def\tablename{TABLE}
+\def\tablesname{Tables}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\abstractname}
+% Text entity \cmd\abstractname\ is used in \env{abstract}.
+% \begin{macrocode}
+\def\abstractname{Abstract}
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\appendixesname}
+% \begin{macro}{\appendixname}
+% Text entity \cmd\appendixesname\ is used in TOC.
+% \begin{macrocode}
+\def\appendixesname{Appendixes}%
+\def\appendixname{Appendix}%
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\acknowledgmentsname}
+% Text entity \cmd\acknowledgmentsname\ is used in \env{acknowledgments}.
+% \begin{macrocode}
+\def\acknowledgmentsname{Acknowledgments}
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\journalname}
+% This should be set by the society journal options, eg `pra'.
+% \begin{macrocode}
+\def\journalname{??}
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\copyrightname}
+% Default layout does not assign copyright, but a journal that wants
+% to might use this.
+% \begin{macrocode}
+\def\copyrightname{??}
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\andname}
+% The text string ``and'' for use in author lists.
+% \begin{macrocode}
+\def\andname{and}
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\@pacs@name}
+% \begin{macro}{\@keys@name}
+% The text string prepended to PACS numbers, resp. to keywords.
+% \begin{macrocode}
+\def\@pacs@name{PACS numbers: }%
+\def\@keys@name{Keywords: }%
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\ppname}
+% The text string ``pp'' for use in page ranges.
+% \begin{macrocode}
+\def\ppname{pp}
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\numbername}
+% The text string ``number'' for use in article reference.
+% \begin{macrocode}
+\def\numbername{number}
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\volumename}
+% The text string ``volume'' for use in article reference.
+% \begin{macrocode}
+\def\volumename{volume}
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\Dated@name}
+% \begin{macro}{\Received@name}
+% \begin{macro}{\Revised@name}
+% \begin{macro}{\Accepted@name}
+% \begin{macro}{\Published@name}
+% These texts are used in the \cmd\date, et al. commands.
+% \begin{macrocode}
+\def\Dated@name{Dated: }%
+\def\Received@name{Received }%
+\def\Revised@name{Revised }%
+\def\Accepted@name{Accepted }%
+\def\Published@name{Published }%
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \section{Legacy Commands}
+% We deffine some commands left over from version 3.1, or give default meanings.
+% Some definitions can be overridden
+% in the document preamble or in included packages.
+%
+% Note on the namespace: command names like \cmd\REV@\emph{name} are used here,
+% because it is not clear that any of this code is generally useful.
+% \begin{macrocode}
+\def\address{\replace@command\address\affiliation}%
+\def\altaddress{\replace@command\altaddress\altaffiliation}%
+\newenvironment{references}{%
+ \class@warn@end{The references environment is not supported; use thebibliography instead.}
+ \gdef\references{\thebibliography{}}\references
+}{%
+ \endthebibliography
+}%
+\def\draft{%
+ \class@warn@end{Command \string\draft\space is obsolete;^^JInvoke option draft instead.}%
+ \@booleantrue\draft@sw
+}%
+\def\tighten{%
+ \class@warn@end{Command \string\tighten\space is obsolete;^^JInvoke option tightenlines instead.}%
+ \@booleantrue\tightenlines@sw
+}%
+\def\tableline{%
+ \noalign{%
+ \class@warn@end{Command \string\tableline\space is obsolete;^^JUse \string\colrule\space instead.}%
+ \global\let\tableline\colrule
+ }%
+ \tableline
+}%
+\def\case{\replace@command\case\frac}%
+\def\slantfrac{\replace@command\slantfrac\frac}%
+\def\tablenote{\replace@command\tablenote\footnote}%
+\def\tablenotemark{\replace@command\tablenotemark\footnotemark}%
+\def\tablenotetext{\replace@command\tablenotetext\footnotetext}%
+% Lose the following definition:
+\DeclareRobustCommand\REV@text[1]{%
+ \relax
+ \ifmmode
+ \mathchoice
+ {\hbox{{\everymath{\displaystyle }#1}}}%
+ {\hbox{{\everymath{\textstyle }#1}}}%
+ {\hbox{{\everymath{\scriptstyle }\let\f@size\sf@size\selectfont#1}}}%
+ {\hbox{{\everymath{\scriptscriptstyle}\let\f@size\ssf@size\selectfont#1}}}%
+ \glb@settings
+ \else
+ \mbox{#1}%
+ \fi
+}%
+% Lose the following definition:
+\DeclareRobustCommand\REV@bbox[1]{%
+ \relax
+ \ifmmode
+ \mathchoice
+ {\hbox{{\everymath{\displaystyle }\boldmath$#1$}}}%
+ {\hbox{{\everymath{\textstyle }\boldmath$#1$}}}%
+ {\hbox{{\everymath{\scriptstyle }\boldmath$#1$}}}%
+ {\hbox{{\everymath{\scriptscriptstyle}\boldmath$#1$}}}%
+ \glb@settings
+ \else
+ \mbox{#1}%
+ \fi
+}%
+\DeclareRobustCommand\REV@bm[1]{%
+ \class@warn@end{To use \string\bm, please load the bm package!}%
+ \global\let\bm\relax
+}%
+\def\FL{\obsolete@command\FL}%
+\def\FR{\obsolete@command\FR}%
+\def\narrowtext{\obsolete@command\narrowtext}%
+\def\mediumtext{\obsolete@command\mediumtext}%
+\newenvironment{quasitable}{%
+ \let@environment{tabular}{longtable}%
+}{%
+}%
+% \end{macrocode}
+%
+% \begin{macro}{\text}
+% \begin{macro}{\bm}
+% \begin{macro}{\bibinfo}
+% \begin{macro}{\eprint}
+% \begin{macro}{\url}
+%
+% If not otherwise defined, give default meanings to certain commands.
+% \changes{v4.0beta 4}{2000/04/10}
+% {eprint takes an aoptional argument, syntactical only in this case.}
+% \begin{macrocode}
+\appdef\class@inithook{%
+ \@ifxundefined\text{\let\text\REV@text}{}%
+ \@ifxundefined\bm{\let\bm\REV@bm}{}%
+ \providecommand\bibinfo[2]{#2}%
+ \providecommand\eprint[2][]{#2}%
+ \providecommand\url[1]{#1}%
+}%
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\bbox}
+%
+% \begin{macrocode}
+\def\bbox#1{%
+ \class@warn@end{\string\bbox\space is obsolete,^^Jload the bm package and use \string\bm\space instead.}%
+ \global\let\bbox\relax
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\mathletters}
+%
+% \begin{macrocode}
+\newenvironment{mathletters}{%
+ \class@warn@end{Environment {mathletters} is obsolete;^^Jload the amsmath package and use {subequations}!}%
+ \global\let\mathletters\@empty
+}{%
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\eqnum}
+%
+% \begin{macrocode}
+\def\eqnum#1{%
+ \class@warn@end{\string\eqnum\space is obsolete, load the amsmath package and use \string\tag!}%
+ \global\let\eqnum\@gobble
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% We read in the symbol definitions.
+% \begin{macrocode}
+\RequirePackage{revsymb}%
+\appdef\class@inithook{\revsymb@inithook}%
+% \end{macrocode}
+%
+% \section{Endgame for the Document Class}
+%
+% This section contains definitions related to the
+% end-of-class processing:
+% we want to control exactly what happens next.
+%
+% We provide for a ``job macro package'' that can override
+% definitions made by the class or any other packages it loads.
+% We act on the value of \cmd\secnumarabic@sw;
+% this code is a model for dealing with the job macro package.
+% We also install a the very last procedures into
+% \cmd\AtEndOfClass, ones that need to tail on to
+% any processing performed either by the
+% class, by any of its included packages, or
+% by the job macro package and any packages it may have loaded.
+%
+% \subsection{Job Macro Package}
+% You can create a ``job macro package'' for your document
+% that will be read in automatically every time
+% your document is processed.
+% Thus, if your job is a file called \file{myarticle.tex},
+% then the file \file{myarticle.rty} will be read in
+% just the same as if you had placed a
+% \cmd\usepackage|{myarticle.rty}| statement
+% immediately following your \cmd\documentclass\ statement.
+%
+% Within your \file{.rty} file,
+% you can define and use control sequence names that use the |@| character
+% and you can override any of the definitions or assignments made
+% by the \revtex\ document class or the selected journal substyle.
+% That is, you have the power to really mess things up badly.
+%
+% If you choose to have a job macro package, you are well
+% advised to read the \LaTeX\ guide to document classes,
+% \filename{clsguide.tex} or read up on the subject
+% in a book line the \LaTeX\ Companion.
+%
+% The file \filename{template.rty} contains a template for
+% creating your own job macro package.
+%
+% \begin{macrocode}
+\AtEndOfClass{%
+ \InputIfFileExists{\jobname.rty}{}{}%
+}%
+% \end{macrocode}
+%
+% \subsection{\cs{secnumarabic@sw}}\label{sec:secnumarabic}
+% \begin{macro}{\secnumarabic@sw}
+% The flag \cmd\secnumarabic@sw\ signifies that sectioning commands are
+% to be numbered arabic.
+% The effect of this command should override any settings
+% made by the journal substyle or by the job macro package.
+% However, it should be lower in precedence to definitions
+% appearing in the document preamble.
+% Therefore, we install its code into \cmd\AtEndOfClass,
+% but do it \emph{after} the corresponding code
+% for the journal substyle and the job macro package.
+%
+% Note: \cmd\thesection\ must \emph{not} be redefined,
+% say, at \cmd\AtBeginDocument\ time by some delayed code.
+% \begin{macrocode}
+\AtEndOfClass{%
+ \@ifxundefined\secnumarabic@sw{\@booleanfalse\secnumarabic@sw}{}%
+ \secnumarabic@sw{%
+ \def\p@section {}%
+ \def\thesection {\arabic{section}}%
+ \def\p@subsection {}%
+ \def\thesubsection {\thesection.\arabic{subsection}}%
+ \def\p@subsubsection {}%
+ \def\thesubsubsection {\thesubsection.\arabic{subsubsection}}%
+ }{}%
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \subsection{Hook for default values of Booleans}
+% This portion of the code for this class file
+% \emph{must} appear at the very end.
+% \begin{macro}{\defaults@hook}
+% The procedure \cmd\defaults@hook\ should hereby
+% be the very last to be installed into \cmd\AtBeginDocument.
+% We provide it with a default in case it has not been
+% used up til now.
+%
+% CHECKME: is any other hook installed later than \cmd\defaults@hook?
+% \begin{macrocode}
+\AtEndOfClass{%
+ \appdef\class@documenthook{%
+ \defaults@hook
+ }%
+}%
+\@ifxundefined\defaults@hook{\let\defaults@hook\@empty}{}%
+% \end{macrocode}
+% \end{macro}
+%
+% End of the class file.
+% \begin{macrocode}
+%</class-pst>
+% \end{macrocode}
+%
+%
+% \section{Symbols: the \texttt{revsymb} module}
+% We immediately define a utility command: this module's warning.
+% \begin{macrocode}
+%<*revsymb>
+\def\REVSYMB@warn#1{\PackageWarningNoLine{revsymb}{#1}}%
+% \end{macrocode}
+%
+% \begin{macro}{\lambdabar}
+% \begin{macrocode}
+\DeclareRobustCommand\lambdabar{%
+ \bgroup
+ \def\@tempa{%
+ \hbox{%
+ \raise.73\ht\z@
+ \hb@xt@\z@{%
+ \kern.25\wd\z@
+ \vrule \@width.5\wd\z@\@height.1\p@\@depth.1\p@
+ \hss
+ }%
+ \box\z@
+ }%
+ }%
+ \mathchoice
+ {\setbox\z@\hbox{$\displaystyle \lambda$}\@tempa}%
+ {\setbox\z@\hbox{$\textstyle \lambda$}\@tempa}%
+ {\setbox\z@\hbox{$\scriptstyle \lambda$}\@tempa}%
+ {\setbox\z@\hbox{$\scriptscriptstyle\lambda$}\@tempa}%
+ \egroup
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\openone}
+% DPC: Really should use a font that includes this glyph.
+% Unfortunately not in AMS ones, but is in bbold, cmbb.
+% (I think, must check),
+% \changes{v4.d}{1998/01/31}
+% {use font-dependent spacing}
+% FIXME: check for bbold.
+% \begin{macrocode}
+\DeclareRobustCommand\openone{\leavevmode\hbox{\small1\normalsize\kern-.33em1}}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\corresponds}
+% \begin{macro}{\overdots}
+% \begin{macro}{\overcirc}
+% J\"org Knappen suggests the replacements:
+% replace \cmd\corresponds\ with \cmd\triangleq, source \classname{amssymb};
+% replace \cmd\overcirc\ with \cmd\mathring, source \classname{latex2e};
+% replace \cmd\overdots\ with \cmd\dddot, source \classname{amsmath}.
+%
+% Any use of any of these commands will result in a warning message at the end
+% of the log file. If the corresponding package is not loaded, a
+% definition will quietly be provided.
+% \begin{macrocode}
+\DeclareRobustCommand\corresponds{\replace@command\corresponds\triangleq}%
+\DeclareRobustCommand\overcirc{\replace@command\overcirc\mathring}%
+\DeclareRobustCommand\overdots{\replace@command\overdots\dddot}%
+\DeclareRobustCommand\REV@triangleq{%
+ {\lower.2ex\hbox{=}}{\kern-.75em^\triangle}%
+}%
+\DeclareRobustCommand\REV@dddot[1]{%
+ \@ontopof{#1}{\cdots}{1.0}\mathord{\box2}%
+}%
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\succsim}
+% \begin{macro}{\precsim}
+% \begin{macro}{\lesssim}
+% \begin{macro}{\gtrsim}
+% \begin{macro}{\alt}
+% \begin{macro}{\agt}
+% These version 3.1 commands are always supplied,
+% but the definitions in \classname{amssymb} are preferred.
+% \begin{macrocode}
+\DeclareRobustCommand\altsuccsim{\succ\kern-.9em_\sim\kern.3em}%
+\DeclareRobustCommand\altprecsim{\prec\kern-1em_\sim\kern.3em}%
+\let\REV@succsim\altsuccsim
+\let\REV@precsim\altprecsim
+\DeclareRobustCommand\REV@lesssim{\mathrel{\mathpalette\vereq{<}}}%
+\DeclareRobustCommand\REV@gtrsim{\mathrel{\mathpalette\vereq{>}}}%
+\DeclareRobustCommand\alt{\lesssim}
+\DeclareRobustCommand\agt{\gtrsim}
+\def\vereq#1#2{%
+ \lower3\p@\vbox{%
+ \baselineskip1.5\p@
+ \lineskip1.5\p@
+ \ialign{$\m@th#1\hfill##\hfil$\crcr#2\crcr\sim\crcr}%
+ }%
+}%
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\tensor}
+% \begin{macro}{\overstar}
+% \begin{macro}{\loarrow}
+% \begin{macro}{\roarrow}
+% \changes{v4.0beta 2}{1999/06/20}
+% {AO: remove duplicates}
+% \begin{macrocode}
+\DeclareRobustCommand\tensor[1]{\@ontopof{#1}{\leftrightarrow}{1.15}\mathord{\box2}}
+\DeclareRobustCommand\overstar[1]{\@ontopof{#1}{\ast}{1.15}\mathord{\box2}}
+\DeclareRobustCommand\loarrow[1]{\@ontopof{#1}{\leftarrow}{1.15}\mathord{\box2}}
+\DeclareRobustCommand\roarrow[1]{\@ontopof{#1}{\rightarrow}{1.15}\mathord{\box2}}
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\@ontopof}
+% \begin{macrocode}
+\def\@ontopof#1#2#3{%
+ {%
+ \mathchoice
+ {\@@ontopof{#1}{#2}{#3}\displaystyle \scriptstyle }%
+ {\@@ontopof{#1}{#2}{#3}\textstyle \scriptstyle }%
+ {\@@ontopof{#1}{#2}{#3}\scriptstyle \scriptscriptstyle}%
+ {\@@ontopof{#1}{#2}{#3}\scriptscriptstyle\scriptscriptstyle}%
+ }%
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\@@ontopof}
+% Same as \revtex3, more or less.
+% \begin{macrocode}
+\def\@@ontopof#1#2#3#4#5{%
+ \setbox\z@\hbox{$#4#1$}%
+ \setbox\f@ur\hbox{$#5#2$}%
+ \setbox\tw@\null\ht\tw@\ht\z@ \dp\tw@\dp\z@
+ \@ifdim{\wd\z@>\wd\f@ur}{%
+ \setbox\f@ur\hb@xt@\wd\z@{\hss\box\f@ur\hss}%
+ \mathord{\rlap{\raise#3\ht\z@\box\f@ur}\box\z@}%
+ }{%
+ \setbox\f@ur\hb@xt@.9\wd\f@ur{\hss\box\f@ur\hss}%
+ \setbox\z@\hb@xt@\wd\f@ur{\hss$#4\relax#1$\hss}%
+ \mathord{\rlap{\copy\z@}\raise#3\ht\z@\box\f@ur}%
+ }%
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\frak}
+% Deal with legacy \cmd\frak:
+% if \classname{amsfonts} not loaded, defined in such a way as to ask for that package.
+% Also, says to use \cmd\mathfrak\ instead.
+% \begin{macrocode}
+\DeclareRobustCommand\frak{%
+ \REVSYMB@warn{%
+ Command \string\frak\space unsupported:^^J%
+ please use \string\mathfrak\space instead.%
+ }%
+ \global\let\frak\mathfrak
+ \frak
+}%
+\DeclareRobustCommand\REV@mathfrak{%
+ \REVSYMB@warn{%
+ Command \string\mathfrak\space undefined:^^J%
+ please specify the amsfonts or amssymb option!%
+ }%
+ \global\let\mathfrak\@firstofone
+ \mathfrak
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\Bbb}
+% Deal with legacy \cmd\Bbb:
+% if \classname{amsfonts} not loaded, defined in such a way as to ask for that package.
+% Also, says to use \cmd\mathbb\ instead.
+% \begin{macrocode}
+\DeclareRobustCommand\Bbb{%
+ \REVSYMB@warn{%
+ Command \string\Bbb\space unsupported:^^J%
+ please use \string\mathbb\space instead.%
+ }%
+ \global\let\Bbb\mathbb
+ \Bbb
+}%
+\DeclareRobustCommand\REV@mathfrak{%
+ \REVSYMB@warn{%
+ Command \string\mathbb\space undefined:^^J%
+ please specify the amsfonts or amssymb option!%
+ }%
+ \global\let\mathbb\@firstofone
+ \mathbb
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\Bigglb}
+% Deal with legacy bold delimiters.
+% Each of the following takes an implicit argument consisting of
+% the delimiter to be made big and bold.
+% FIXME: \cmd\DeclareBoldMathCommand\ is not the right tool!
+% \begin{macrocode}
+\def\Bigglb{\REV@boldopen \Bigg}%
+\def\Biglb {\REV@boldopen \Big }%
+\def\bigglb{\REV@boldopen \bigg}%
+\def\biglb {\REV@boldopen \big }%
+\def\Biggrb{\REV@boldclose\Bigg}%
+\def\Bigrb {\REV@boldclose\Big }%
+\def\biggrb{\REV@boldclose\bigg}%
+\def\bigrb {\REV@boldclose\big }%
+\def\REV@pmb#1{%
+ \hbox{%
+ \setbox\z@=\hbox{#1}%
+ \kern-.02em\copy\z@\kern-\wd\z@
+ \kern .04em\copy\z@\kern-\wd\z@
+ \kern-.02em
+ \raise.04em\copy\z@
+ }%
+}%
+\def\REV@boldopen #1#2{\mathopen {\REV@pmb{$#1#2$}}}%
+\def\REV@boldclose#1#2{\mathclose{\REV@pmb{$#1#2$}}}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\AtBeginDocument}
+% Package dependencies are taken care of
+% \cmd\AtBeginDocument\ time.
+% \begin{macrocode}
+\def\revsymb@inithook{%
+ \@ifxundefined\dddot{\let\dddot\REV@dddot}{}%
+ \@ifxundefined\triangleq{\let\triangleq\REV@triangleq}{}%
+ \@ifxundefined\succsim{\let\succsim\altsuccsim}{}%
+ \@ifxundefined\precsim{\let\precsim\altprecsim}{}%
+ \@ifxundefined\lesssim{\let\lesssim\REV@lesssim}{}%
+ \@ifxundefined\gtrsim {\let\gtrsim \REV@gtrsim }{}%
+ \@ifxundefined\mathfrak{\let\mathfrak\REV@mathfrak}{}%
+ \@ifxundefined\mathbb{\let\mathbb\REV@mathbb}{}%
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macrocode}
+%</revsymb>
+% \end{macrocode}
+%
+% \section{The \classoption{10pt} class option: the \texttt{10pt} module}
+% \changes{v4.0beta 2}{1999/06/20}
+% {AO: all code for \protect\classoption{10pt} is in this module.}
+% The file \file{10pt.rtx} is read in by the \classname{revtex4} document class
+% if \cmd\@pointsize\ has the value 10.
+%
+% \begin{macrocode}
+%<*10pt>
+% \end{macrocode}
+%
+% \begin{macrocode}
+\def\normalsize{%
+ \@setfontsize\normalsize\@xpt{11.5}%
+ \abovedisplayskip 10\p@ \@plus2\p@ \@minus5\p@
+ \belowdisplayskip \abovedisplayskip
+ \abovedisplayshortskip \abovedisplayskip
+ \belowdisplayshortskip \abovedisplayskip
+ \let\@listi\@listI
+}%
+% \end{macrocode}
+%
+% \begin{macrocode}
+\def\small{%
+ \@setfontsize\small\@ixpt{10.5}%
+ \abovedisplayskip 8.5\p@ \@plus3\p@ \@minus4\p@
+ \belowdisplayskip \abovedisplayskip
+ \abovedisplayshortskip \z@ \@plus2\p@
+ \belowdisplayshortskip 4\p@ \@plus2\p@ \@minus2\p@
+ \def\@listi{%
+ \leftmargin\leftmargini
+ \topsep 4\p@ \@plus2\p@ \@minus2\p@
+ \parsep 2\p@ \@plus\p@ \@minus\p@
+ \itemsep \parsep
+ }%
+}%
+% \end{macrocode}
+%
+% \begin{macrocode}
+\def\footnotesize{%
+ \@setfontsize\footnotesize\@viiipt{9.5pt}%
+ \abovedisplayskip 6\p@ \@plus2\p@ \@minus4\p@
+ \belowdisplayskip \abovedisplayskip
+ \abovedisplayshortskip \z@ \@plus\p@
+ \belowdisplayshortskip 3\p@ \@plus\p@ \@minus2\p@
+ \def\@listi{%
+ \leftmargin\leftmargini
+ \topsep 3\p@ \@plus\p@ \@minus\p@
+ \parsep 2\p@ \@plus\p@ \@minus\p@
+ \itemsep \parsep
+ }%
+}%
+% \end{macrocode}
+%
+% \begin{macrocode}
+\def\scriptsize{%
+ \@setfontsize\scriptsize\@viipt\@viiipt
+}%
+\def\tiny{%
+ \@setfontsize\tiny\@vpt\@vipt
+}%
+\def\large{%
+ \@setfontsize\large\@xiipt{14pt}%
+}%
+\def\Large{%
+ \@setfontsize\Large\@xivpt{18pt}%
+}%
+\def\LARGE{%
+ \@setfontsize\LARGE\@xviipt{22pt}%
+}%
+\def\huge{%
+ \@setfontsize\huge\@xxpt{25pt}%
+}%
+\def\Huge{%
+ \@setfontsize\Huge\@xxvpt{30pt}%
+}%
+% \end{macrocode}
+%
+% \begin{macrocode}
+%\appdef\class@documenthook{%
+ \@ifxundefined\twoside@sw{\@booleanfalse\twoside@sw}{}%
+ \twoside@sw{%
+% \oddsidemargin -.1in
+% \evensidemargin -.4in
+ \oddsidemargin -20pt
+ \evensidemargin -20pt
+ \marginparwidth 107pt
+ }{%
+ \oddsidemargin -.25in
+ \evensidemargin -.25in
+ \marginparwidth 30pt
+ }%
+%}%
+% \end{macrocode}
+%
+% \begin{macrocode}
+\marginparsep 6pt
+% \end{macrocode}
+%
+% \begin{macrocode}
+\topmargin -61pt
+% \end{macrocode}
+%
+% \begin{macrocode}
+\headheight 25pt
+\headsep 16pt
+% \end{macrocode}
+%
+% \begin{macrocode}
+\topskip 10pt
+% \end{macrocode}
+%
+% \begin{macrocode}
+\footskip 30pt
+% \end{macrocode}
+%
+% Was: 57\cmd\baselineskip + \cmd\topskip
+% \begin{macrocode}
+%\appdef\set@pica@hook{%
+ \textheight = 56pc
+%}%
+% \end{macrocode}
+%
+% \begin{macrocode}
+\textwidth42.5pc
+% \end{macrocode}
+%
+% \begin{macrocode}
+\columnsep 1.5pc
+\columnseprule 0pt
+% \end{macrocode}
+%
+% \begin{macrocode}
+\footnotesep 1pt
+\skip\footins 39pt plus 4pt minus 2pt
+\def\footnoterule{\kern-19pt\hrule width.5in\kern18.6pt\relax}%
+% \end{macrocode}
+%
+% \begin{macrocode}
+\floatsep 12pt plus 2pt minus 2pt
+\textfloatsep 20pt plus 2pt minus 4pt
+\intextsep 12pt plus 2pt minus 2pt
+% \end{macrocode}
+%
+% \begin{macrocode}
+\dblfloatsep 12pt plus 2pt minus 2pt
+\dbltextfloatsep 20pt plus 2pt minus 4pt
+% \end{macrocode}
+%
+% \begin{macrocode}
+\@fptop 0pt plus 1fil
+\@fpsep 8pt plus 2fil
+\@fpbot 0pt plus 1fil
+\@dblfptop 0pt plus 1fil
+\@dblfpsep 8pt plus 2fil
+\@dblfpbot 0pt plus 1fil
+% \end{macrocode}
+%
+% \begin{macrocode}
+\marginparpush 5pt
+% \end{macrocode}
+%
+% \begin{macrocode}
+\parskip 0pt plus 1pt
+\parindent 10pt
+\emergencystretch8\p@
+% \end{macrocode}
+%
+% \begin{macrocode}
+\partopsep 2pt plus 1pt minus 1pt
+% \end{macrocode}
+%
+% \begin{macrocode}
+\leftmargini 25pt
+\leftmarginii 22pt
+\leftmarginiii 18.7pt
+\leftmarginiv 17pt
+\leftmarginv 10pt
+\leftmarginvi 10pt
+% \end{macrocode}
+%
+% \begin{macrocode}
+\def\@listI{%
+ \leftmargin\leftmargini
+ \parsep 4\p@ plus2\p@ minus\p@
+ \topsep 8\p@ plus2\p@ minus4\p@
+ \itemsep 4\p@ plus2\p@ minus\p@
+}%
+% \end{macrocode}
+%
+% \begin{macrocode}
+\labelsep 4pt
+% \end{macrocode}
+%
+% \begin{macrocode}
+\def\@listii{%
+ \leftmargin\leftmarginii
+ \labelwidth\leftmarginii
+ \advance\labelwidth-\labelsep
+ \topsep 4\p@ plus2\p@ minus\p@
+ \parsep 2\p@ plus\p@ minus\p@
+ \itemsep \parsep
+}%
+% \end{macrocode}
+%
+% \begin{macrocode}
+\def\@listiii{%
+ \leftmargin\leftmarginiii
+ \labelwidth\leftmarginiii
+ \advance\labelwidth-\labelsep
+ \topsep 2\p@ plus\p@ minus\p@
+ \parsep \z@
+ \partopsep \p@ plus\z@ minus\p@
+ \itemsep \topsep
+}%
+% \end{macrocode}
+%
+% \begin{macrocode}
+\def\@listiv{%
+ \leftmargin\leftmarginiv
+ \labelwidth\leftmarginiv
+ \advance\labelwidth-\labelsep
+}%
+% \end{macrocode}
+%
+% \begin{macrocode}
+\def\@listv{%
+ \leftmargin\leftmarginv
+ \labelwidth\leftmarginv
+ \advance\labelwidth-\labelsep
+}%
+% \end{macrocode}
+%
+% \begin{macrocode}
+\def\@listvi{%
+ \leftmargin\leftmarginvi
+ \labelwidth\leftmarginvi
+ \advance\labelwidth-\labelsep
+}%
+% \end{macrocode}
+%
+% \begin{macrocode}
+%</10pt>
+% \end{macrocode}
+%
+% \section{The \classoption{11pt} class option: the \texttt{11pt} module}
+% \changes{v4.0beta 2}{1999/06/20}
+% {AO: all code for \protect\classoption{11pt} is in this module.}
+% The file \file{11pt.rtx} is read in by the \classname{revtex4} document class
+% if \cmd\@pointsize\ has the value 11.
+%
+% \begin{macrocode}
+%<*11pt>
+% \end{macrocode}
+%
+% \begin{macrocode}
+\def\normalsize{%
+ \@setfontsize\normalsize\@xipt{13.6}%
+ \abovedisplayskip 11\p@ \@plus3\p@ \@minus6\p@
+ \belowdisplayskip \abovedisplayskip
+ \abovedisplayshortskip \abovedisplayskip
+ \belowdisplayshortskip \abovedisplayskip
+ \let\@listi\@listI
+}%
+% \end{macrocode}
+%
+% \begin{macrocode}
+\def\small{%
+ \@setfontsize\small\@xpt\@xiipt
+ \abovedisplayskip 10\p@ \@plus2\p@ \@minus5\p@
+ \abovedisplayshortskip \z@ \@plus3\p@
+ \belowdisplayshortskip 6\p@ \@plus3\p@ \@minus3\p@
+ \def\@listi{\leftmargin\leftmargini
+ \topsep 6\p@ \@plus2\p@ \@minus2\p@
+ \parsep 3\p@ \@plus2\p@ \@minus\p@
+ \itemsep \parsep
+ }%
+ \belowdisplayskip \abovedisplayskip
+}%
+% \end{macrocode}
+%
+% \begin{macrocode}
+\def\footnotesize{%
+ \@setfontsize\footnotesize\@ixpt{11}%
+ \abovedisplayskip 8\p@ \@plus2\p@ \@minus4\p@
+ \abovedisplayshortskip \z@ \@plus\p@
+ \belowdisplayshortskip 4\p@ \@plus2\p@ \@minus2\p@
+ \def\@listi{\leftmargin\leftmargini
+ \topsep 4\p@ \@plus2\p@ \@minus2\p@
+ \parsep 2\p@ \@plus\p@ \@minus\p@
+ \itemsep \parsep
+ }%
+ \belowdisplayskip \abovedisplayskip
+}%
+% \end{macrocode}
+%
+% \begin{macrocode}
+\def\scriptsize{%
+ \@setfontsize\scriptsize\@viiipt{9.5}%
+}%
+\def\tiny{%
+ \@setfontsize\tiny\@vipt\@viipt
+}%
+\def\large{%
+ \@setfontsize\large\@xiipt{14}%
+}%
+\def\Large{%
+ \@setfontsize\Large\@xivpt{18}%
+}%
+\def\LARGE{%
+ \@setfontsize\LARGE\@xviipt{22}%
+}%
+\def\huge{%
+ \@setfontsize\huge\@xxpt{25pt}%
+}%
+\def\Huge{%
+ \@setfontsize\Huge\@xxvpt{30pt}%
+}%
+% \end{macrocode}
+%
+% \begin{macrocode}
+%</11pt>
+% \end{macrocode}
+%
+% \section{The \classoption{12pt} class option: the \texttt{12pt} module}
+% \changes{v4.0beta 2}{1999/06/20}
+% {AO: all code for \protect\classoption{12pt} is in this module.}
+% The file \file{12pt.rtx} is read in by the \classname{revtex4} document class
+% if \cmd\@pointsize\ has the value 12.
+%
+% \begin{macrocode}
+%<*12pt>
+% \end{macrocode}
+%
+% \begin{macrocode}
+\def\normalsize{%
+ \@setfontsize\normalsize\@xiipt{14pt}%
+ \abovedisplayskip 12\p@ \@plus3\p@ \@minus7\p@
+ \belowdisplayskip \abovedisplayskip
+ \abovedisplayshortskip \z@ plus3\p@
+ \belowdisplayshortskip 6.5\p@ \@plus3.5\p@ \@minus3\p@
+ \let\@listi\@listI
+}%
+% \end{macrocode}
+%
+% \begin{macrocode}
+ \def\small{%
+ \@setfontsize\small\@xipt{14.5pt}%
+ \abovedisplayskip 8\p@ \@plus3\p@ \@minus6\p@
+ \belowdisplayskip \abovedisplayskip
+ \abovedisplayshortskip \z@ \@plus3\p@
+ \belowdisplayshortskip 6.5\p@ \@plus3.5\p@ \@minus3\p@
+ \def\@listi{%
+ \leftmargin\leftmargini
+ \topsep 9\p@ \@plus3\p@ \@minus5\p@
+ \parsep 4.5\p@ \@plus2\p@ \@minus\p@
+ \itemsep \parsep
+ }%
+}%
+% \end{macrocode}
+%
+% Same baselineskip as \cmd\small\ ?
+% \begin{macrocode}
+ \def\footnotesize{%
+ \@setfontsize\footnotesize\@xpt{14.5pt}%
+ \abovedisplayskip 10\p@ \@plus2\p@ \@minus5\p@
+ \belowdisplayskip \abovedisplayskip
+ \abovedisplayshortskip \z@ \@plus3\p@
+ \belowdisplayshortskip 6\p@ \@plus3\p@ \@minus3\p@
+ \def\@listi{%
+ \leftmargin\leftmargini
+ \topsep 6\p@ \@plus2\p@ \@minus2\p@
+ \parsep 3\p@ \@plus2\p@ \@minus\p@
+ \itemsep \parsep
+ }%
+}%
+% \end{macrocode}
+%
+% \begin{macrocode}
+\def\scriptsize{%
+ \@setfontsize\scriptsize\@viiipt{9.5pt}%
+}%
+\def\tiny{%
+ \@setfontsize\tiny\@vipt{7pt}%
+}%
+\def\large{%
+ \@setfontsize\large\@xivpt{18pt}%
+}%
+\def\Large{%
+ \@setfontsize\Large\@xviipt{22pt}%
+}%
+\def\LARGE{%
+ \@setfontsize\LARGE\@xxpt{25pt}%
+}%
+\def\huge{%
+ \@setfontsize\huge\@xxvpt{30pt}%
+}%
+\let\Huge=\huge
+% \end{macrocode}
+%
+% \begin{macrocode}
+%</12pt>
+% \end{macrocode}
+%
+% \section{Page parameters}%
+% This code is common to both \classoption{11pt} and \classoption{12pt}.
+%
+% \begin{macrocode}
+%<*11pt|12pt>
+% \end{macrocode}
+%
+% \begin{macrocode}
+%\appdef\class@documenthook{%
+ \@ifxundefined\twoside@sw{\@booleanfalse\twoside@sw}{}%
+ \twoside@sw{%
+ \oddsidemargin 0pt
+ \evensidemargin 0pt
+ \marginparwidth 60pt
+ }{%
+ \oddsidemargin 0pt
+ \evensidemargin 0pt
+ \marginparwidth 44pt
+ }%
+%}%
+% \end{macrocode}
+%
+% \begin{macrocode}
+\marginparsep 10pt
+% \end{macrocode}
+%
+% \begin{macrocode}
+\topmargin -37pt
+% \end{macrocode}
+%
+% \begin{macrocode}
+\headheight 12pt
+\headsep 25pt
+% \end{macrocode}
+%
+% \begin{macrocode}
+\topskip 10pt
+% \end{macrocode}
+%
+% \begin{macrocode}
+\footskip 30pt
+% \end{macrocode}
+%
+% \begin{macrocode}
+%\appdef\set@pica@hook{%
+ \@ifxundefined\tightenlines@sw{\@booleanfalse\tightenlines@sw}{}%
+ \textheight=665.5\p@
+ \tightenlines@sw{%
+ \def\baselinestretch{1}%
+ }{%
+ \def\baselinestretch{1.5}%
+ }%
+%}%
+% \end{macrocode}
+%
+% \begin{macrocode}
+\textwidth 468pt
+% \end{macrocode}
+%
+% \begin{macrocode}
+\columnsep 10pt
+\columnseprule 0pt
+% \end{macrocode}
+%
+% \begin{macrocode}
+\footnotesep 1pt
+\skip\footins 25.25pt plus 4pt minus 2pt
+\def\footnoterule{\kern-5.25pt\hrule width.5in\kern4.85pt\relax}
+% \end{macrocode}
+%
+% \begin{macrocode}
+\floatsep 14pt plus 2pt minus 4pt
+\textfloatsep 20pt plus 2pt minus 4pt
+\intextsep 14pt plus 4pt minus 4pt
+% \end{macrocode}
+%
+% \begin{macrocode}
+\dblfloatsep 14pt plus 2pt minus 4pt
+\dbltextfloatsep 20pt plus 2pt minus 4pt
+% \end{macrocode}
+%
+% \begin{macrocode}
+\@fptop 0pt plus 1fil
+\@fpsep 10pt plus 2fil
+\@fpbot 0pt plus 1fil
+\@dblfptop 0pt plus 1fil
+\@dblfpsep 10pt plus 2fil%
+\@dblfpbot 0pt plus 1fil
+% \end{macrocode}
+%
+% \begin{macrocode}
+\marginparpush 7pt
+% \end{macrocode}
+%
+% \begin{macrocode}
+\parskip 0pt plus 1pt
+\parindent 15pt
+\emergencystretch8\p@
+\partopsep 3pt plus 2pt minus 2pt
+% \end{macrocode}
+%
+% \begin{macrocode}
+\leftmargini 30pt
+\leftmarginii 26pt
+\leftmarginiii 22pt
+\leftmarginiv 20pt
+\leftmarginv 12pt
+\leftmarginvi 12pt
+% \end{macrocode}
+%
+% \begin{macrocode}
+\def\@listI{\leftmargin\leftmargini \parsep 5\p@ plus2.5\p@ minus\p@
+ \topsep 10\p@ plus4\p@ minus6\p@
+ \itemsep 5\p@ plus2.5\p@ minus\p@
+}%
+% \end{macrocode}
+%
+% \begin{macrocode}
+\labelsep 6pt
+% \end{macrocode}
+%
+% \begin{macrocode}
+\def\@listii{\leftmargin\leftmarginii
+ \labelwidth\leftmarginii\advance\labelwidth-\labelsep
+ \topsep 5\p@ plus2.5\p@ minus\p@
+ \parsep 2.5\p@ plus\p@ minus\p@
+ \itemsep \parsep
+}%
+% \end{macrocode}
+%
+% \begin{macrocode}
+\def\@listiii{\leftmargin\leftmarginiii
+ \labelwidth\leftmarginiii\advance\labelwidth-\labelsep
+ \topsep 2.5\p@ plus\p@ minus\p@
+ \parsep \z@ \partopsep \p@ plus\z@ minus\p@
+ \itemsep \topsep
+}%
+% \end{macrocode}
+%
+% \begin{macrocode}
+\def\@listiv{\leftmargin\leftmarginiv
+ \labelwidth\leftmarginiv\advance\labelwidth-\labelsep
+}%
+% \end{macrocode}
+%
+% \begin{macrocode}
+\def\@listv{\leftmargin\leftmarginv
+ \labelwidth\leftmarginv\advance\labelwidth-\labelsep
+}%
+% \end{macrocode}
+%
+% \begin{macrocode}
+\def\@listvi{\leftmargin\leftmarginvi
+ \labelwidth\leftmarginvi\advance\labelwidth-\labelsep
+}%
+% \end{macrocode}
+%
+% \begin{macrocode}
+%</11pt|12pt>
+% \end{macrocode}
+%
+% \section{The \classoption{aps} class option: the \texttt{aps} module}
+% \changes{v4.0beta 2}{1999/06/20}
+% {AO: made aps.rtx part of revtex4.dtx}
+% The file \file{aps.rtx} is read in by the \classname{revtex4} document class
+% if \cmd\@society\ has the value \classoption{aps}.
+%
+% Here, code specific to APS journals is separated out from the \revtex\ document class.
+% (Other societies can customize \revtex\ by supplying their own \file{.rtx} file.)
+%
+% We first give some text entities (amounting to journal abbreviations),
+% then some APS-specific initialisations,
+% then code for particular APS journals.
+% In the latter case, the choice is
+% keyed off the macro \cmd\@journal.
+%
+% Note on \cmd\AtEndOfClass:
+% this file, like all journal substyles, is read in at \cmd\AtEndOfClass\ time,
+% so you should not use this command in this file.
+% Mote for \LaTeX\ developers: It would be an improvement in \LaTeX\ to
+% \cmd\let\cmd\AtEndOfClass\ to something like \cmd\@firstofone.
+% This change would be effected in \cmd\@onefilewithoptions.
+%
+% \begin{macrocode}
+%<*aps>
+% \end{macrocode}
+%
+% \subsection{Defend Against Forseeable Errors}%
+% Protect this file from being read in by anything but \revtex.
+% \begin{macrocode}
+\ifx\undefined\substyle@ext
+ \def\@tempa{%
+ \endinput
+ \GenericWarning{I must be read in by REVTeX! (Bailing out)}%
+ }%
+ \expandafter\else
+ \def\@tempa{}%
+ \expandafter\fi\@tempa
+% \end{macrocode}
+%
+% \subsection{Abbreviations}%
+% The following macros constitute typing shortcuts for
+% certain journal names.
+% \begin{macrocode}
+\def\ao{Appl.\ Opt.\ }
+\def\ap{Appl.\ Phys.\ }
+\def\apl{Appl.\ Phys.\ Lett.\ }
+\def\apj{Astrophys.\ J.\ }
+\def\bell{Bell Syst.\ Tech.\ J.\ }
+\def\jqe{IEEE J.\ Quantum Electron.\ }
+\def\assp{IEEE Trans.\ Acoust.\ Speech Signal Process.\ }
+\def\aprop{IEEE Trans.\ Antennas Propag.\ }
+\def\mtt{IEEE Trans.\ Microwave Theory Tech.\ }
+\def\iovs{Invest.\ Ophthalmol.\ Vis.\ Sci.\ }
+\def\jcp{J.\ Chem.\ Phys.\ }
+\def\jmo{J.\ Mod.\ Opt.\ }
+\def\josa{J.\ Opt.\ Soc.\ Am.\ }
+\def\josaa{J.\ Opt.\ Soc.\ Am.\ A }
+\def\josab{J.\ Opt.\ Soc.\ Am.\ B }
+\def\jpp{J.\ Phys.\ (Paris) }
+\def\nat{Nature (London) }
+\def\oc{Opt.\ Commun.\ }
+\def\ol{Opt.\ Lett.\ }
+\def\pl{Phys.\ Lett.\ }
+\def\pra{Phys.\ Rev.\ A }
+\def\prb{Phys.\ Rev.\ B }
+\def\prc{Phys.\ Rev.\ C }
+\def\prd{Phys.\ Rev.\ D }
+\def\pre{Phys.\ Rev.\ E }
+\def\prl{Phys.\ Rev.\ Lett.\ }
+\def\rmp{Rev.\ Mod.\ Phys.\ }
+\def\pspie{Proc.\ Soc.\ Photo-Opt.\ Instrum.\ Eng.\ }
+\def\sjqe{Sov.\ J.\ Quantum Electron.\ }
+\def\vr{Vision Res.\ }
+% \end{macrocode}
+%
+% \subsection{APS Setup}%
+% Here we define the default procedures for APS journals.
+% Individual APS journals may override these definitions.
+%
+% \subsubsection{Title block}%
+%
+% The specifics of the title block.
+% Apply to all APS journals; individual journals may override these settings.
+%
+% \begin{macro}{\@fnsymbol}
+% The \LaTeX\ kernel definition of \cmd\@fnsymbol\ is overridden.
+% \begin{macrocode}
+\def\@fnsymbol#1{%
+ \ensuremath{%
+ \ifcase#1\or
+ *\or \dagger \or
+ \ddagger \or \mathsection \or
+ \mathparagraph\or
+ **\or \dagger\dagger \or
+ \ddagger\ddagger \or \mathsection \mathsection \or
+ \mathparagraph\mathparagraph\or
+ *{*}*\ignorespaces \or \dagger\dagger\dagger \or
+ \ddagger\ddagger\ddagger\or \mathsection \mathsection \mathsection \or
+ \mathparagraph\mathparagraph\mathparagraph\or
+ \else
+ \@ctrerr
+ \fi
+ }%
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\titlepage}
+% \begin{macrocode}
+\renewenvironment{titlepage}{%
+ \let\wastwocol@sw\twocolumn@sw
+ \onecolumngrid
+ \newpage
+ \thispagestyle{titlepage}%
+ \c@page\z@% article sets this to one not zero???
+}{%
+ \wastwocol@sw{\twocolumngrid}{\newpage}%
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\frontmatter@abstractheading}
+% APS Journals all set the abstract head the same way, with no head.
+% However, if the user has specified the \classoption{preprint} class option,
+% then the abstract will have a head.
+% \begin{macrocode}
+\def\frontmatter@abstractheading{%
+ \preprintsty@sw{%
+ \begingroup
+ \centering\large
+ \abstractname
+ \par
+ \endgroup
+ }{}%
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\frontmatter@abstractwidth}
+% All APS journals set the abstract to the same width.
+% \begin{macrocode}
+\def\frontmatter@abstractwidth{400\p@}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\frontmatter@abstractfont}
+% All APS journals set the abstract body the same way.
+% \changes{v4.0beta 3}{1999/11/13}
+% {Bug 123: add parskip to the abstract.}
+% \begin{macrocode}
+\def\frontmatter@abstractfont{%
+ \small
+ \parindent1em\relax
+ \adjust@abstractwidth
+}%
+\def\adjust@abstractwidth{%
+ \dimen@\textwidth\advance\dimen@-\frontmatter@abstractwidth
+ \divide\dimen@\tw@
+ \galley@sw{%
+ \advance\rightskip\tw@\dimen@
+ }{%
+ \advance\leftskip\dimen@
+ \advance\rightskip\dimen@
+ }%
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% All APS journal preprints use separate titlepage and full-width abstract.
+%
+% Note that we defer polling the \cmd\preprintsty@sw\ Boolean until
+% \cmd\AtBeginDocument\ time, because the document preamble
+% may legitimately change the state of that variable.
+% \begin{macrocode}
+\appdef\class@documenthook{%
+ \preprintsty@sw{%
+ \@ifxundefined\titlepage@sw{\@booleantrue\titlepage@sw}{}%
+% \tightenlines@sw{%
+% \def\baselinestretch{1}%
+% }{%
+% \def\baselinestretch{1.5}%
+% }%
+ \def\frontmatter@abstractwidth{\textwidth}%
+ \def\frontmatter@affiliationfont{\it}%
+ \let\section\section@preprintsty
+% \let\@hangfrom@section\@hangfrom@section@preprintsty
+ \let\subsection\subsection@preprintsty
+ \let\subsubsection\subsubsection@preprintsty
+ }{}%
+ }%
+% \end{macrocode}
+%
+% \begin{macro}{\frontmatter@authorformat}
+% All APS journals set the author list the same.
+% The leading is 11.5 points, and
+% there is 11.5 points of extra space above the first author line
+% (which amounts to the same thing as 11.5 points extra below the title)
+% for a total of 23 points base-to-base.
+%
+% \begin{macrocode}
+\def\frontmatter@authorformat{%
+ \skip@\@flushglue
+ \@flushglue\z@ plus.3\hsize\relax
+ \centering
+ \advance\baselineskip\p@
+ \parskip11.5\p@\relax
+ \@flushglue\skip@
+%\preprintsty@sw{}{%
+% \addvspace{0\p@}%
+%}%
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\frontmatter@above@affilgroup}
+% The default amount of space above affiliation.
+% APS Journals have 24 points b-b above an affiliation group.
+% \begin{macrocode}
+\def\frontmatter@above@affilgroup{%
+%\preprintsty@sw{}{%
+% \addvspace{11\p@}%
+%}%
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\frontmatter@above@affiliation}
+% \begin{macro}{\frontmatter@above@affiliation@script}
+% The default amount of space above affiliation.
+% APS Journals have no extra space between author group down to common affiliation.
+% \begin{macrocode}
+\def\frontmatter@above@affiliation@script{%
+ \skip@\@flushglue
+ \@flushglue\z@ plus.3\hsize\relax
+ \centering
+ \@flushglue\skip@
+ \addvspace{3.5\p@}%
+}%
+\def\frontmatter@above@affiliation{%
+ \preprintsty@sw{}{%
+% \addvspace{12\p@}%
+ }%
+}%
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\frontmatter@affiliationfont}
+% All APS journals set the affiliation the same.
+% \begin{macrocode}
+\def\frontmatter@affiliationfont{%
+ \small\it
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\frontmatter@collaboration@above}
+% PRL: 1.5 points extra: 13 points base-to-base above.
+% \begin{macrocode}
+\def\frontmatter@collaboration@above{%
+ \preprintsty@sw{%
+ }{%
+ \parskip1.5\p@\relax
+ }%
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% Set up the default APS style for title block authors and affiliations.
+% \begin{macrocode}
+\@ifxundefined\groupauthors@sw{%
+ \clo@groupedaddress
+}{}%
+% \end{macrocode}
+%
+% \begin{macro}{\frontmatter@setup}
+% All APS journals set the title page using the same font and size.
+% However, justification varies for the title block elements, so
+% we assert none here.
+% \begin{macrocode}
+\def\frontmatter@setup{%
+ \normalfont
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\frontmatter@title@above}
+% \begin{macro}{\frontmatter@title@format}
+% \begin{macro}{\frontmatter@title@below}
+% All APS journals set the article title the same.
+%
+% Note: Spacing from title to author is 23 points base-to-base.
+% \begin{macrocode}
+\def\frontmatter@title@above{\addvspace{6\p@}}%
+\def\frontmatter@title@format{\large\bfseries\centering\parskip\z@skip}%
+\def\frontmatter@title@below{}%
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\frontmatter@makefnmark}
+% All APS journals share this procedure for setting the titlepage footnote text.
+% \begin{macrocode}
+\def\@author@parskip{3\p@}%
+\def\frontmatter@makefnmark{%
+ \@textsuperscript{%
+ \normalfont\@thefnmark
+ }%
+}%
+\def\frontmatter@authorbelow{%
+ \addvspace{3\p@}%
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\frontmatter@RRAP@format}
+% All APS journals use the same format for the ``Received, Revised, etc.'' block on the title page.
+%
+% Change note: 11.5 points b-b from author/affiliation down to date.
+% \begin{macrocode}
+\def\frontmatter@RRAP@format{%
+ \small
+ \centering
+% \preprintsty@sw{}{\parskip.5ex\relax}%
+ \everypar{\hbox\bgroup(\@gobble@leavemode@uppercase}%
+ \def\par{\@ifvmode{}{\unskip)\egroup\@@par}}%
+}%
+\def\punct@RRAP{;\egroup\ \hbox\bgroup}%
+\def\@gobble@leavemode@uppercase#1#2{\expandafter\MakeTextUppercase}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\frontmatter@PACS@format}
+%
+% \begin{macrocode}
+\def\frontmatter@PACS@format{%
+ \addvspace{11\p@}%
+ \footnotesize
+ \adjust@abstractwidth
+ \parindent\z@
+ \parskip\z@skip
+ \samepage
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\frontmatter@keys@format}
+%
+% \begin{macrocode}
+\def\frontmatter@keys@format{%
+ \footnotesize
+ \adjust@abstractwidth
+ \parindent\z@
+ \samepage
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\ps@titlepage}
+% \changes{v4.g}{1998/06/10}
+% {multiple preprint commands}
+% \changes{v4.0beta 2}{1999/06/20}
+% {Stack \cs{preprint} args flush right at right margin.}
+% Title page style. Currently empty except for preprint header,
+% which consists of all the \cmd\preprint\ arguments,
+% stacked flush right at the right margin.
+% \changes{v4.0beta 3}{1999/11/13}
+% {Insert procedure \cs{checkindate}}
+% \begin{macrocode}
+\def\ps@titlepage{%
+ \def\@oddhead{%
+ \hfill
+ \produce@preprints\@preprint
+ }%
+ \let\@evenhead\@oddhead
+ \def\@oddfoot{%
+ \hb@xt@\z@{\byrevtex\hss}%
+ \hfil
+ \preprintsty@sw{\thepage}{}%
+ \quad\checkindate
+ \hfil
+ }%
+ \let\@evenfoot\@oddfoot
+}%
+\def\byrevtex{\byrevtex@sw{Typeset by REV\TeX}{}}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\produce@preprints}
+% \changes{v4.0beta 3}{1999/11/13}
+% {Bug 115. If three or more preprints specified, set on single line, with commas.}
+% \begin{macrocode}
+\def\produce@preprints#1{%
+ \preprint@sw{%
+ \vtop to \z@{%
+ \def\baselinestretch{1}%
+ \small
+ \let\preprint\preprint@count
+ \count@\z@#1\@ifnum{\count@>\tw@}{%
+ \hbox{%
+ \let\preprint\preprint@hlist
+ #1\setbox\z@\lastbox
+ }%
+ }{%
+ \let\preprint\preprint@cr
+ \halign{\hfil##\cr#1\crcr}%
+ \par
+ \vss
+ }%
+ }%
+ }{}%
+}%
+\def\preprint@cr#1{#1\cr}%
+\def\preprint@count#1{\advance\count@\@ne}%
+\def\preprint@hlist#1{#1\hbox{, }}%
+% \end{macrocode}
+% \end{macro}
+%
+% \subsubsection{Stacked Heads}%
+% All APS journals put a period (.), followed by quad space, after the section number.
+% Also, no hanging section number.
+% \begin{macrocode}
+\def\@seccntformat#1{\csname the#1\endcsname.\quad}%
+\def\@hang@from#1#2#3{#1#2#3}%
+% \end{macrocode}
+%
+% Note that in the following, we wish to set the section head uppercase, so we use
+% David Carlisle's \cmd\MakeTextUppercase. However, because this procedure effectively
+% parses its argument (looking for things to \emph{not} translate), it has to be invoked
+% in such a way that the argument of the \cmd\section\ command is passed to it as its
+% own argument.
+%
+% To accomplish this, we use the \cmd\@hangfrom@\ hook, which was developed for this
+% purpose.
+%
+% \begin{macrocode}
+\def\section{%
+ \@startsection
+ {section}%
+ {1}%
+ {\z@}%
+ {0.8cm \@plus1ex \@minus .2ex}%
+ {0.5cm}%
+ {%
+ \normalfont\small\bfseries
+ \centering
+ }%
+}%
+\def\@hangfrom@section#1#2#3{\@hangfrom{#1#2}\MakeTextUppercase{#3}}%
+% \end{macrocode}
+%
+% \begin{macrocode}
+\def\subsection{%
+ \@startsection
+ {subsection}%
+ {2}%
+ {\z@}%
+ {.8cm \@plus1ex \@minus .2ex}%
+ {.5cm}%
+ {%
+ \normalfont\small\bfseries
+ \centering
+ }%
+}%
+% \end{macrocode}
+%
+% \begin{macrocode}
+\def\subsubsection{%
+ \@startsection
+ {subsubsection}%
+ {3}%
+ {\z@}%
+ {.8cm \@plus1ex \@minus .2ex}%
+ {.5cm}%
+ {%
+ \normalfont\small\itshape
+ \centering
+ }%
+}%
+% \end{macrocode}
+%
+% \subsubsection{Runin Heads}%
+% \begin{macrocode}
+\def\paragraph{%
+ \@startsection
+ {paragraph}%
+ {4}%
+ {\parindent}%
+ {\z@}%
+ {-1em}%
+ {\normalfont\normalsize\itshape}%
+}%
+% \end{macrocode}
+%
+% \begin{macrocode}
+\def\subparagraph{%
+ \@startsection
+ {subparagraph}%
+ {5}%
+ {\parindent}%
+ {3.25ex \@plus1ex \@minus .2ex}%
+ {-1em}%
+ {\normalfont\normalsize\bfseries}%
+}%
+% \end{macrocode}
+%
+%
+%
+% \begin{macro}{\section@preprintsty}
+% \begin{macro}{\subsection@preprintsty}
+% \begin{macro}{\subsubsection@preprintsty}
+% Here are the formatting procedures specific to the preprint style;
+% the only difference is that the heads are flush left instead of centered.
+%
+% \begin{macrocode}
+\def\section@preprintsty{%
+ \@startsection
+ {section}%
+ {1}%
+ {\z@}%
+ {0.8cm \@plus1ex \@minus .2ex}%
+ {0.5cm}%
+ {%
+ \normalfont\small\bfseries
+% \centering
+ }%
+}%
+%\def\@hangfrom@section@preprintsty#1#2#3{\@hangfrom{#1#2}\MakeTextUppercase{#3}}%
+% \end{macrocode}
+% \begin{macrocode}
+\def\subsection@preprintsty{%
+ \@startsection
+ {subsection}%
+ {2}%
+ {\z@}%
+ {.8cm \@plus1ex \@minus .2ex}%
+ {.5cm}%
+ {%
+ \normalfont\small\bfseries
+% \centering
+ }%
+}%
+% \end{macrocode}
+% \begin{macrocode}
+\def\subsubsection@preprintsty{%
+ \@startsection
+ {subsubsection}%
+ {3}%
+ {\z@}%
+ {.8cm \@plus1ex \@minus .2ex}%
+ {.5cm}%
+ {%
+ \normalfont\small\itshape
+% \centering
+ }%
+}%
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% Effectively invoke the \classoption{bibnote} class option,
+% if \cmd\frontmatter@footnote@produce\ has not yet been set.
+% \begin{macrocode}
+ \@ifxundefined\frontmatter@footnote@produce{%
+ \let\frontmatter@footnote@produce\frontmatter@footnote@produce@endnote
+ }{}%
+% \end{macrocode}
+%
+%
+% \subsubsection{Table of Contents}%
+% The toc will itself make an entry in the toc,
+% but we temporarily turn off toc formatting for the duration.
+% \begin{macrocode}
+\def\@pnumwidth{1.55em}
+\def\@tocrmarg {2.55em}
+\def\@dotsep{4.5pt}
+\setcounter{tocdepth}{3}
+% \end{macrocode}
+%
+% \begin{macrocode}
+\def\tableofcontents{%
+ \addtocontents{toc}{\string\tocdepth@munge}%
+ \print@toc{toc}%
+ \addtocontents{toc}{\string\tocdepth@restore}%
+}%
+% \end{macrocode}
+%
+% \begin{macrocode}
+\def\tocdepth@munge{%
+ \let\l@section@saved\l@section
+ \let\l@section\@gobble@tw@
+}%
+\def\@gobble@tw@#1#2{}%
+% \end{macrocode}
+%
+% \begin{macrocode}
+\def\tocdepth@restore{%
+ \let\l@section\l@section@saved
+}%
+% \end{macrocode}
+%
+% \begin{macrocode}
+\def\l@part#1#2{\addpenalty{\@secpenalty}%
+ \begingroup
+ \set@tocdim@pagenum{#2}%
+% \@tempdima 3em %
+ \parindent \z@
+ \rightskip\tocleft@pagenum plus 1fil\relax
+ \skip@\parfillskip\parfillskip\z@
+ \addvspace{2.25em plus\p@}%
+ \large \bf %
+ \leavevmode\ignorespaces#1\unskip\nobreak\hskip\skip@
+ \hb@xt@\rightskip{\hfil\unhbox\z@}\hskip-\rightskip\hskip\z@skip
+ \par
+ \nobreak %
+ \endgroup
+}%
+% \end{macrocode}
+%
+% \begin{macro}{\l@section}
+% Determine which TOC elements are automatically indented.
+% \begin{macrocode}
+\def\tocleft@{\z@}%
+\def\tocdim@min{5\p@}%
+\def\l@section{%
+ \l@@sections{}{section}% Implicit #3#4
+}%
+\def\l@f@section{%
+ \addpenalty{\@secpenalty}%
+ \addvspace{1.0em plus\p@}%
+ \bf
+}%
+\def\l@subsection{%
+ \l@@sections{section}{subsection}% Implicit #3#4
+}%
+\def\l@subsubsection{%
+ \l@@sections{subsection}{subsubsection}% Implicit #3#4
+}%
+\def\l@paragraph#1#2{}%
+\def\l@subparagraph#1#2{}%
+% \end{macrocode}
+% \end{macro}
+%
+% Activate the auto TOC processing.
+% \begin{macrocode}
+\let\toc@pre\toc@pre@auto
+\let\toc@post\toc@post@auto
+% \end{macrocode}
+%
+% \begin{macrocode}
+\def\listoffigures{\print@toc{lof}}%
+\def\l@figure{\@dottedtocline{1}{1.5em}{2.3em}}
+% \end{macrocode}
+%
+% \begin{macrocode}
+\def\listoftables{\print@toc{lot}}%
+\let\l@table\l@figure
+% \end{macrocode}
+%
+% \subsubsection{Default column bottom}%
+% All APS journal styles have flush bottoms.
+%
+% Note that we defer polling the \cmd\raggedcolumn@sw\ Boolean until
+% \cmd\AtBeginDocument\ time, because the document preamble
+% may legitimately change the state of that variable.
+% \begin{macrocode}
+\appdef\class@documenthook{%
+ \@ifxundefined\raggedcolumn@sw{\@booleantrue\raggedcolumn@sw}{}%
+ \raggedcolumn@sw{\raggedbottom}{\flushbottom}%
+}%
+% \end{macrocode}
+%
+%
+% \subsubsection{Table alignment style}%
+%
+% \begin{macro}{\tableft@skip@float}
+% \begin{macro}{\tabmid@skip@float}
+% \begin{macro}{\tabright@skip@float}
+% \begin{macro}{\array@row@pre@float}
+% \begin{macro}{\array@row@pst@float}
+% All APS publications have the same table specification:
+% Scotch rules above and below, centered in column.
+% \begin{macrocode}
+\def\tableft@skip@float{\z@ plus\hsize}%
+\def\tabmid@skip@float{\@flushglue}%
+\def\tabright@skip@float{\z@ plus\hsize}%
+\def\array@row@pre@float{\hline\hline\noalign{\vskip\doublerulesep}}%
+\def\array@row@pst@float{\noalign{\vskip\doublerulesep}\hline\hline}%
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+%
+% \subsubsection{Footnote formatting}%
+% We customize the formatting of footnotes for all APS journals.
+% \begin{macro}{\@makefntext}
+% \begin{macrocode}
+\def\@makefntext#1{%
+ \def\baselinestretch{1}%
+ \reset@font
+ \footnotesize
+ \leftskip1em
+ \parindent1em
+ \noindent\nobreak\hskip-\leftskip
+ \hb@xt@\leftskip{%
+ \Hy@raisedlink{\hyper@anchorstart{footnote@\the\c@footnote}\hyper@anchorend}%
+ \hss\@makefnmark\
+ }%
+ #1%
+ \par
+}%
+% \end{macrocode}
+% \end{macro}
+%
+%
+% \subsubsection{Appendix}%
+% \begin{macro}{\appendix}
+% \begin{macro}{\@hangfrom@appendix}
+% \begin{macro}{\@hangfroms@appendix}
+% \begin{macro}{\@appendixcntformat}
+% \changes{v4.0beta 3}{1999/11/13}
+% {Bug 129: section* within appendix was producing appendixname}
+% \begin{macrocode}
+\prepdef\appendix{%
+ \par
+ \let\@hangfrom@section\@hangfrom@appendix
+ \let\@hangfroms@section\@hangfroms@appendix
+ \let\@sectioncntformat\@appendixcntformat
+}%
+\def\@hangfrom@appendix#1#2#3{%
+ #1%
+ \@if@empty{#2}{%
+ #3%
+ }{%
+ #2\@if@empty{#3}{}{:\ #3}%
+ }%
+}%
+\def\@hangfroms@appendix#1#2{%
+ #1#2%
+}%
+\def\@appendixcntformat#1{\appendixname\ \csname the#1\endcsname}%
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \subsubsection{Bibliography}%
+%
+% Customize \revtex\ for the journal substyle;
+% this task requires three components:
+% the \BibTeX\ \file{apsrev.bst} style file,
+% customizing code for \file{natbib}, and
+% customizations of the \env{thebibliography} environment.
+%
+% \begin{macro}{\@bibstyle}
+% Define the argument of the \cmd\bibliographystyle\ command
+% (if the document does not do so).
+% The user must have installed a \file{.bst} file of the corresponding name.
+% This file will then be used by \BibTeX\ when compiling the document's \file{.bbl} file.
+%
+% The default bibliography style for the APS journal substyles, \file{apsrev.bst},
+% was originally called \file{revtex.bst}, but
+% that name is much too generic and should be given to a \file{.bst} file that
+% that applies generally to any \revtex\ jounal substyle, an impossibility.
+%
+% To generate \file{apsrev.bst}, use \classname{custom-bib} version 4.0b2 or later.
+% Run the \file{.bst} generator, \file{makebst.tex}, and
+% accept all defaults, with the following exceptions:
+%
+% \begin{enumerate}
+% \item
+% ORDERING OF REFERENCES: c: seq-no (references in order of Citation);
+% \item
+% AUTHOR NAMES: i: nm-init,ed-au (Initials + surname);
+% \item
+% NUMBER OF AUTHORS: l: max 12, min 12 (there will be three prompts total);
+% \item
+% TYPEFACE FOR AUTHORS IN LIST OF REFERENCES: u: nmft,nmft-def (User defined author font);
+% \item
+% FONT FOR FIRST NAMES: u: fnm-def (First names in user defined font);
+% \item
+% EDITOR NAMES IN INCOLLECTION ETC: a: nmfted (Editors incollection like authors);
+% \item
+% DATE FORMAT: p: yr-par (Date in parentheses);
+% \item
+% SUPPRESS MONTH: x: xmth (Date is year only);
+% \item
+% TITLE OF ARTICLE: i: tit-it (Title italic)
+% \item
+% ARTICLE TITLE PRESENT: x: jtit-x (No article title);
+% \item
+% JOURNAL NAME FONT: r: jttl-rm (Journal title normal);
+% \item
+% TECHNICAL REPORT TITLE: b: trtit-b (Tech. report title like books);
+% \item
+% JOURNAL VOLUME: b: vol-bf (Volume bold);
+% \item
+% VOLUME PUNCTUATION: c: volp-com (Volume with comma);
+% \item
+% PAGE NUMBERS: f: jpg-1 (Only start page number);
+% \item
+% JOURNAL NAME PUNCTUATION: x: jnm-x (Space after journal);
+% \item
+% PUBLISHER IN PARENTHESES: d: pub-date (Publisher and date in parentheses);
+% \item
+% PUBLISHER POSITION: p: pre-pub (Publisher before chapter, pages);
+% \item
+% ISBN NUMBER: isbn: (Include ISBN for books, booklets)
+% \item
+% ISSN NUMBER: issn: (Include ISSN for periodicals)
+% \item
+% EDITOR IN COLLECTIONS: b: edby (Booktitle, edited by \dots);
+% \item
+% PUNCTUATION BETWEEN SECTIONS (BLOCKS): c: blk-com (Comma between blocks);
+% \item
+% ABBREVIATE WORD `PAGES': a: pp (`Page' abbreviated);
+% \item
+% ABBREVIATE WORD `EDITORS': a: ed (`Editor' abbreviated);
+% \item
+% OTHER ABBREVIATIONS: a: abr (Abbreviations);
+% \item
+% ABBREVIATION FOR `EDITION': a: ednx (`Edition' abbreviated as `ed');
+% \item
+% EDITION NUMBERS: n: ord (Numerical editions);
+% \item
+% STORED JOURNAL NAMES: a: jabr (Abbreviated journal names);
+% \item
+% FONT OF `ET AL': i: etal-it (Italic et al);
+% \item
+% ADDITIONAL REVTeX DATA FIELDS: r: revdata, eprint, url, url-blk (Include REVTeX data fields collaboration, eid, eprint, numpages, url)
+% \item
+% NEW FONT SELECTION SCHEME: n: nfss (NFSS);
+% \item
+% ADDITIONAL REVTeX DATA FIELDS: y: revdata (additional data fields);
+% \item
+% REFERENCE COMPONENT TAGS: y: reference component tags;
+% \item
+% URL ADDRESS: n: URL as note;
+% \end{enumerate}
+%
+% A file \file{apsrev.dbj} file equivalent to the following should result:
+% \begin{verbatim}
+% \input docstrip
+% \preamble
+% ----------------------------------------
+% *** REVTeX-compatible apsrev.bst ***
+% \endpreamble
+% \postamble
+% End of customized bst file
+% \endpostamble
+% \keepsilent
+% \askforoverwritefalse
+% \def\MBopts{\from{merlin.mbs}{%
+% seq-no%: Citation order (unsorted, like unsrt.bst)
+% ,nm-init,ed-au%: Initials + surname (J. F. Smith)
+% ,nmlm%: Limited authors (et al replaces missing names)
+% ,x10%: Maximum of 10 authors
+% ,m10%: Minimum of 10 authors
+% ,nmft,nmft-def%: User defined author font (\bibnamefont)
+% ,fnm-def%: First names in user defined font (\bibfnamefont)
+% ,nmfted%: Editors incollection like authors font
+% ,yr-par%: Date in parentheses as (May 1993)
+% ,xmth%: Date is year only
+% ,tit-it%: Title italic (\em)
+% ,jtit-x%: No article title
+% ,jttl-rm%: Journal name normal font
+% ,trtit-b%: Tech. report title like books
+% ,vol-bf%: Volume bold as {\bf vol}(num)
+% ,volp-com%: Volume with comma as vol(num), ppp
+% ,jpg-1%: Only start page number
+% ,jnm-x%: Space after journal name
+% ,pub-date%: Publisher and date in parentheses (Oxford, 1994)
+% ,pre-pub%: Publisher before chapter, pages
+% ,isbn%: Include ISBN for books, booklets, etc.
+% ,issn%: Include ISSN for periodicals
+% ,edby%: In booktitle, edited by .. (where .. is names)
+% ,blk-com%: Comma between blocks
+% ,pp%: `Page' abbreviated as p. or pp.
+% ,ed%: `Editor' abbreviated as ed. or eds.
+% ,abr%: Abbreviations of such words
+% ,ednx%: `Edition' abbreviated as `ed'
+% ,ord%: Numerical editions as 1st, 2nd, 3rd, etc
+% ,jabr%: Abbreviated journal names
+% ,etal-it%: Italic et al
+% ,revdata,eprint,url,url-blk%: Include REVTeX data fields collaboration, eid, eprint, numpages, url
+% ,url,url-nt%: URL as note
+% ,bibinfo%: Reference component tags like \bibinfo in the content of \bibitem
+% ,nfss%: NFSS use \textbf, \emph, not \bf, \em
+% }}
+% \generate{\file{\jobname.bst}{\MBopts}}
+% \endbatchfile
+% \end{verbatim}
+%
+% To get a banner in the \file{.blg} file, patch by hand into \file{apsrev.bst}:
+% \begin{verbatim}
+% FUNCTION {banner.log} {"REVTeX-compatible apsrev.bst 9812" warning$}
+% \end{verbatim}
+% and invoke \texttt{banner.log} from within \texttt{begin.bib}
+%
+% We ensure that the journal substyle has the first word
+% in the matter by installing the (default) APS code
+% later on (see Section~\ref{sec:aps-defaults}).
+% \end{macro}
+%
+% \begin{macro}{\bibpunct}
+% The following commands effectively establish the style in which \cmd\cite\ commands are
+% formatted.
+% You can think of them as the second needed component for the bibliography.
+%
+% Set up for numerical citations.
+%
+% The journal substyles can override these choices.
+%
+% Note that setting \cmd\NAT@sort\ at this late hour is not totally effective.
+% You have to give \classname{natbib} the relevant options at load time.
+% \begin{macrocode}
+ \bibpunct{[}{]}{,}{n}{}{,}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\bibsection}
+% We define the sectioning command to use when starting the bibliography.
+%
+% \changes{v4.0rc1}{2001/06/18}
+% {grid changes with push and pop}
+% \begin{macrocode}
+ \def\bibsection{%
+ \par
+% \twocolumn@sw{\onecolumngrid}{}%
+ \onecolumngrid@push
+ \begingroup
+ \baselineskip26\p@
+% \def\@currentHref{section.bib}%
+ \bib@device{\textwidth}{245.5\p@}%
+ \endgroup
+ \nobreak\@nobreaktrue
+ \addvspace{19\p@}%
+ \par
+% \twocolumn@sw{\twocolumngrid}{}%
+ \onecolumngrid@pop
+ }%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\bib@device}
+% \begin{macro}{\bibpreamble}
+% \begin{macro}{\bibsep}
+% \begin{macro}{\newblock}
+% We define the sectioning command to use when starting the bibliography.
+% \begin{macrocode}
+\def\bib@device#1#2{%
+ \hb@xt@#1{%
+ \hfil
+ \phantomsection
+ \addcontentsline {toc}{section}{\protect\numberline{}\refname}%
+% \hyper@anchorstart {\@currentHref }%
+ \hb@xt@#2{%
+ \skip@\z@\@plus-1fil\relax
+ \leaders\hrule height.25 \p@ depth.25 \p@ \hskip\z@\@plus1fil
+ \hskip\skip@
+ \hskip\z@\@plus0.125fil\leaders\hrule height.375\p@ depth.375\p@ \hskip\z@\@plus0.75fil \hskip\z@\@plus0.125fil
+ \hskip\skip@
+ \hskip\z@\@plus0.25 fil\leaders\hrule height.5 \p@ depth.5 \p@ \hskip\z@\@plus0.5 fil \hskip\z@\@plus0.25 fil
+ \hskip\skip@
+ \hskip\z@\@plus0.375fil\leaders\hrule height.625\p@ depth.625\p@ \hskip\z@\@plus0.25fil \hskip\z@\@plus0.375fil
+ % \hskip\skip@
+ % \hfil
+ }%
+%\hyper@anchorend
+ \hfil
+ }%
+}%
+\let\bibpreamble\@empty
+\bibsep\z@\relax
+\def\newblock{\ }%
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\bibfont}
+% We define the font switch that applies to the body of the bibliography.
+%
+% \begin{macrocode}
+ \def\bibfont{%
+ \small
+ }%
+% \end{macrocode}
+% \end{macro}
+%
+%
+% \subsubsection{Index}%
+% FIXME: the following call to \cmd\twocolumn\ appears wrong if we were in two-column grid.
+% \changes{v4.0beta 3}{1999/11/13}
+% {grid changes with ltxgrid}
+% \changes{v4.0rc1}{2001/06/18}
+% {grid changes with push and pop}
+% \begin{macrocode}
+\newenvironment{theindex}{%
+%\let\wastwocol@sw\twocolumn@sw
+ \columnseprule \z@
+ \columnsep 35\p@
+ \c@secnumdepth-\maxdimen
+ \onecolumngrid@push
+ \section{\indexname}%
+ \thispagestyle{plain}%
+ \parindent\z@
+ \parskip\z@ plus.3\p@\relax
+ \let\item\@idxitem
+%\wastwocol@sw{\twocolumngrid}{}%
+ \onecolumngrid@pop
+}{%
+%\onecolumngrid@pop
+}%
+%
+\def\@idxitem{\par\hangindent 40\p@}
+%
+\def\subitem{\par\hangindent 40\p@ \hspace*{20\p@}}
+%
+\def\subsubitem{\par\hangindent 40\p@ \hspace*{30\p@}}
+%
+\def\indexspace{\par \vskip 10\p@ plus5\p@ minus3\p@\relax}
+% \end{macrocode}
+%
+%
+% \subsection{Journal-Specific Code}%
+%
+% For APS journals, we supply code specific to PRA, PRB, PRC, PRD, PRE, PRL, and RMP.
+% At present, they are identical,
+% with the exception of PRB's code for superscript citations.
+%
+% \begin{macrocode}
+\def\journal@pra{pra}%
+\def\journal@prb{prb}%
+\def\journal@prc{prc}%
+\def\journal@prd{prd}%
+\def\journal@pre{pre}%
+\def\journal@prl{prl}%
+\def\journal@prstab{prstab}%
+\def\journal@rmp{rmp}%
+% \end{macrocode}
+%
+% Note: the convention in this document class is that
+% the substyle must not override any explicit class options
+% declared by the document.
+% This means that the various Booleans of Section~\ref{sec:options}
+% may be assigned here only if they are still undefined at this point.
+%
+% For most all of the APS journals, the journal-dependent
+% code is relatively meager and is therefore embedded in this file.
+% However, the RMP code is sufficiently extensive that
+% splitting it out into a separate file is more convenient.
+%
+% \subsubsection{\classoption{pra}}%
+% \begin{macrocode}
+\@ifx{\@journal\journal@pra}{%
+% \end{macrocode}
+%
+% End of \classoption{pra} code.
+% \begin{macrocode}
+}{}%
+% \end{macrocode}
+%
+% \subsubsection{\classoption{prb}}%
+% \begin{macrocode}
+\@ifx{\@journal\journal@prb}{%
+% \end{macrocode}
+% PRB requires superscript citations.
+% We use Patrick Daly's \classname{natbib} package,
+% and hyperref and other packages are already set up to cope with this.
+%
+% \begin{macro}{\bibpunct}
+% By default, PRB selects the \classoption{bibnotes} option.
+% We also set up for numerical citations.
+% \begin{macrocode}
+ \@ifxundefined\bibnotes@sw{\@booleantrue\bibnotes@sw}{}%
+ \bibpunct{}{}{,}{s}{}{\textsuperscript{,}}%
+ \def\@onlinecite#1{\begingroup\let\@cite\NAT@citenum\citealp{#1}\endgroup}%
+% \end{macrocode}
+% \end{macro}
+%
+% Invoke \classoption{superbib} option if the document
+% has made no selection of its own.
+% \begin{macrocode}
+ \appdef\class@documenthook{%
+ \@ifxundefined\place@bibnumber{%
+ \let\place@bibnumber\place@bibnumber@sup
+ }{}%
+ }%
+% \end{macrocode}
+%
+% End of \classoption{prb} code.
+% \begin{macrocode}
+}{}%
+% \end{macrocode}
+%
+% \subsubsection{\classoption{prc}}%
+% \begin{macrocode}
+\@ifx{\@journal\journal@prc}{%
+% \end{macrocode}
+%
+% End of \classoption{prc} code.
+% \begin{macrocode}
+}{}%
+% \end{macrocode}
+%
+% \subsubsection{\classoption{prd}}%
+% \begin{macrocode}
+\@ifx{\@journal\journal@prd}{%
+% \end{macrocode}
+%
+% End of \classoption{prd} code.
+% \begin{macrocode}
+}{}%
+% \end{macrocode}
+%
+% \subsubsection{\classoption{pre}}%
+% \begin{macrocode}
+\@ifx{\@journal\journal@pre}{%
+% \end{macrocode}
+%
+% End of \classoption{prd} code.
+% \begin{macrocode}
+}{}%
+% \end{macrocode}
+%
+% \subsubsection{\classoption{prl}}%
+% \begin{macrocode}
+\@ifx{\@journal\journal@prl}{%
+% \end{macrocode}
+% In PRL, the default is the \classoption{bibnotes} option,
+% and the Acknowledgments section has no head.
+%
+% However, the References head is a device that may be described as a lozenge centered on the baseline,
+% 71 points wide by 2 points thick, with the ends tapering to a half point in thickness.
+% Space above 26 points base to base, below 31 base to base.
+% \begin{macrocode}
+ \@ifxundefined\bibnotes@sw{\@booleantrue\bibnotes@sw}{}%
+ \@booleanfalse\acknowledgments@sw
+%\let\acknowledgmentsname\@empty
+ \def\bibsection{%
+ \par
+ \begingroup
+ \baselineskip26\p@
+ \bib@device{\hsize}{72\p@}%
+ \endgroup
+ \nobreak\@nobreaktrue
+ \addvspace{19\p@}%
+ }%
+% \end{macrocode}
+% Implement length checking.
+% Use the \classname{times} and \classname{mathtime} packages,
+% plus whatever other processing is required
+% to make the formatted output be true
+% to the metrics of the journal.
+%
+% Wait. Do not use mathtime after all. APS has their own way of doing math pi, involving
+% Adobe Mathematical Pi and other fonts.
+% \begin{macrocode}
+%\appdef\class@documenthook{%
+ \@ifxundefined\lengthcheck@sw{\@booleanfalse\lengthcheck@sw}{}%
+ \lengthcheck@sw{%
+ \RequirePackage{times}%
+% \RequirePackage{mathtime}%
+ }{%
+ }%
+%}%
+% \end{macrocode}
+% A PRL does not have numbered sections.
+% \begin{macrocode}
+ \c@secnumdepth=-\maxdimen
+% \end{macrocode}
+% Note: we defer this code until after type size file is read in.
+% \begin{macrocode}
+ \appdef\set@typesize@hook{%
+ \@ifnum{\@pointsize=10\relax}{%
+ \lengthcheck@sw{%
+ \def\large{%
+ \@setfontsize\large{12.5}{14\p@}%
+ }%
+ \def\normalsize{%
+ \@setfontsize\normalsize{10.5}\@xiipt
+ \abovedisplayskip 6\p@ \@plus6\p@ \@minus5\p@
+ \belowdisplayskip \abovedisplayskip
+ \abovedisplayshortskip \abovedisplayskip
+ \belowdisplayshortskip \abovedisplayskip
+ \let\@listi\@listI
+ }%
+ \def\small{%
+ \@setfontsize\small{9.5}\@xipt
+ \abovedisplayskip 5\p@ \@plus5\p@ \@minus4\p@
+ \belowdisplayskip \abovedisplayskip
+ \abovedisplayshortskip \abovedisplayskip
+ \belowdisplayshortskip \abovedisplayskip
+ \let\@listi\@listI
+ }%
+ \DeclareMathSizes{12.5}{12.5}{9}{6}%
+ \DeclareMathSizes{10.5}{10.5}{7.5}{5}%
+ \DeclareMathSizes{9.5}{9.5}{7.0}{5}%
+ }{%
+ \def\normalsize{%
+ \@setfontsize\normalsize\@xpt\@xiipt
+ \abovedisplayskip 10\p@ \@plus2\p@ \@minus5\p@
+ \belowdisplayskip \abovedisplayskip
+ \abovedisplayshortskip \abovedisplayskip
+ \belowdisplayshortskip \abovedisplayskip
+ \let\@listi\@listI
+ }%
+ }%
+ }{}%
+ }%
+%\appdef\set@pica@hook{%
+ \@ifnum{\@pointsize=10\relax}{%
+ \textheight = 694.0\p@
+% \textheight = 55\baselineskip
+% \advance\textheight by \topskip
+ }{}%
+%}%
+% \end{macrocode}
+%
+% End of \classoption{prl} code.
+% \begin{macrocode}
+}{}%
+% \end{macrocode}
+%
+% \subsubsection{\classoption{prstab}}%
+% \begin{macrocode}
+\@ifx{\@journal\journal@prstab}{%
+% \end{macrocode}
+%
+% End of \classoption{prstab} code.
+% \begin{macrocode}
+}{}%
+% \end{macrocode}
+%
+% \subsubsection{\classoption{rmp}}%
+% Read in the code from a file.
+% \begin{macrocode}
+\@ifx{\@journal\journal@rmp}{\input{\@journal\substyle@ext}}{}%
+% \end{macrocode}
+%
+%
+% \subsection{Establish APS Defaults}\label{sec:aps-defaults}
+%
+% \begin{macro}{\footinbib@sw}
+% All APS journals except RMP invoke the \classoption{footinbib} option.
+% We rely on the RMP-specific code to give a value to this switch.
+% \begin{macrocode}
+\@ifxundefined\footinbib@sw{\@booleantrue\footinbib@sw}{}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\place@bibnumber}
+% \begin{macro}{\@bibstyle}
+% We install code that will select the presentation for \cmd\bibitem s
+% and govern the \BibTeX\ processing.
+% If the user has not selected any related option and if
+% the journal substyle also has not, then the \classoption{aps}
+% option choses these defaults.
+% \begin{macrocode}
+\appdef\class@documenthook{%
+ \@ifxundefined\place@bibnumber{%
+ \let\place@bibnumber\place@bibnumber@inl
+ }{}%
+ \@ifxundefined\@bibstyle{%
+ \def\@bibstyle{apsrev}%
+ }{}%
+}%
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+%
+% \begin{macrocode}
+%</aps>
+% \end{macrocode}
+%
+%
+% \section{The \classoption{rmp} journal substyle: the \texttt{rmp} module}
+%
+% The file \file{rmp.rtx} is read in by the \classname{revtex4} document class
+% if \cmd\@society\ has the value \classoption{aps}
+% and \cmd\@journal\ has the value \classoption{rmp}.
+%
+% It is read at the end of the \filename{aps.rtx}, so all definitions
+% and assignments in that file are operative unless overridden here.
+%
+%
+% \begin{macrocode}
+%<*rmp>
+% \end{macrocode}
+%
+% Protect this file from being read in by anything but \revtex.
+% \begin{macrocode}
+\ifx\undefined\substyle@ext
+ \def\@tempa{%
+ \endinput
+ \GenericWarning{I must be read in by REVTeX! (Bailing out)}%
+ }%
+ \expandafter\else
+ \def\@tempa{}%
+ \expandafter\fi\@tempa
+% \end{macrocode}
+%
+% Protect this file from being read in as a society instead of a journal.
+% In such a case, \cmd\@journal\ will be \cmd\@empty.
+% \begin{macrocode}
+\@ifx@empty\@journal{%
+ \class@warn{I have to read in the aps substyle first!}%
+ \endinput
+ \def\@society{aps}\def\@journal{rmp}%
+ \expandafter\input\expandafter{\@society\substyle@ext}%
+}{}%
+% \end{macrocode}
+%
+% \subsection{Frontmatter}%
+%
+% \begin{macro}{\frontmatter@setup}
+%
+% \begin{macrocode}
+\def\frontmatter@setup{%
+ \normalfont\sffamily\raggedright
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\PACS@warn}
+% Per Mark Doyle, RMP never displays the PACS, so they don't want the `use showpacs' warning spit out.
+% \begin{macrocode}
+\def\PACS@warn{RMP documents do not display PACS. Your \string\pacs\space will be ignored}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\frontmatter@title@above}
+% \begin{macro}{\frontmatter@title@format}
+% \begin{macro}{\frontmatter@title@below}
+%
+% \begin{macrocode}
+\def\frontmatter@title@above{}%
+\def\frontmatter@title@format{\Large\bfseries\raggedright}% HelveticaNeue-Medium(Italic) 14pt.
+\def\frontmatter@title@below{\addvspace{12\p@}}% 24pt b-b down to first author
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\frontmatter@authorformat}
+% Set the rag to a milder value, because we want to do true ragged right
+% typesetting, as opposed to the \LaTeX\ default, which gives very poor results.
+%
+% Note: author font is 9.8bp. 19.2bp/14.3bp above/below.
+% \begin{macrocode}
+\def\frontmatter@authorformat{%
+ \preprintsty@sw{\vskip0.5pc\relax}{}%
+ \@tempskipa\@flushglue
+ \@flushglue\z@ plus50\p@\relax
+ \raggedright\advance\leftskip.5in\relax
+ \@flushglue\@tempskipa
+ \parskip\z@skip
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\frontmatter@affiliationfont}
+% The hook \cmd\frontmatter@affiliationfont\ controls the formatting of affiliations and affiliation groups.
+% The hook \cmd\frontmatter@above@affilgroup\ is invoked just before proceeding with author/affiliation processing.
+% The \cmd\frontmatter@above@affiliation\ is the amount of space above affiliations in the \classoption{groupedaddress} style,
+% and \cmd\frontmatter@above@affiliation@script\ is that for \classoption{superscriptaddress}.
+%
+% Note: affiliation font is 9.03/10.4bp, 14.3bp/19.2bp b-b above/below
+% \begin{macrocode}
+\def\frontmatter@affiliationfont{% Helvetica 9/10.2
+ \small\slshape\selectfont\baselineskip10.5\p@\relax
+ \@tempskipa\@flushglue
+ \@flushglue\z@ plus50\p@\relax
+ \raggedright\advance\leftskip.5in\relax
+ \@flushglue\@tempskipa
+}%
+\def\frontmatter@above@affilgroup{\addvspace{7.2\p@}}% additional leading above an author
+\def\frontmatter@above@affiliation{\addvspace{5.3\p@}}%
+\def\frontmatter@above@affiliation@script{}%
+% \end{macrocode}
+% \end{macro}
+%
+% Set up the default RMP style for title block authors and affiliations.
+% This command should effectively override the effect of the
+% corresponding command in the parent substyle.
+% \begin{macrocode}
+\@ifxundefined\groupauthors@sw{%
+ \clo@groupedaddress
+}{}%
+% \end{macrocode}
+%
+% \begin{macro}{\frontmatter@RRAP@format}
+% Note: in RMP, if we are not in preprint mode, the date will not be produced.
+%
+% Note: Helvetica C/lc, 8.98bp, space above: 16.3bp b-b.
+% \begin{macrocode}
+\def\frontmatter@RRAP@format{%
+ \addvspace{7.3\p@}%
+ \small
+ \raggedright\advance\leftskip.5in\relax
+% \parskip.5ex\relax
+% \everypar{\hbox\bgroup(\@gobble@leavemode@uppercase}%
+% \def\par{\@ifvmode{}{\unskip)\egroup\@@par}}%
+% \preprintsty@sw{}{\let\@date\@empty}%
+}%
+%\def\@gobble@leavemode@uppercase#1#2{\expandafter\MakeTextUppercase}%
+\def\produce@RRAP#1{%
+ \@if@empty{#1}{}{%
+ \@ifvmode{\leavevmode}{}%
+ \unskip(\ignorespaces#1\unskip)\quad
+ }%
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\frontmatter@abstractheading}
+% Space above 21.8bp b-b.
+% \begin{macrocode}
+\def\frontmatter@abstractheading{%
+ \preprintsty@sw{%
+ \begingroup
+ \centering\large
+ \abstractname
+ \par
+ \endgroup
+ \vspace{.5pc}%
+ }{}%
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\frontmatter@abstractfont}
+% TimesTen 8.93bp/9.6bp X 360bp, indented 36bp, with 21.9/37.6bp b-b above/below
+% \begin{macrocode}
+\def\frontmatter@abstractfont{%
+ \footnotesize
+ \hsize360\p@
+ \leftskip=0.5in
+ \advance\hsize\leftskip
+%\rightskip=\leftskip
+ \parindent\z@
+%\hsize5.5in
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\frontmatter@preabstractspace}
+% \begin{macro}{\frontmatter@postabstractspace}
+% Space above and space below abstract in title block.
+% Should be 22/36 points base-to-base.
+% \begin{macrocode}
+\def\frontmatter@preabstractspace{7.7\p@}%
+\def\frontmatter@postabstractspace{24.6\p@}%
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% FIXME: Not done: PACS.
+% FIXME: TOC: Head is same as \section: HelveticaNeue 8.98. 32/22bp b-b above/below,
+% Body: TimesTen 8/10.5.
+%
+% \subsection{General Text}%
+%
+% If not in preprint mode, set the type size to 10/12 point. Note: s/b 11.6bp leading
+% \begin{macrocode}
+ \appdef\set@typesize@hook{%
+ \@ifxundefined\preprintsty@sw{}{%
+ \preprintsty@sw{}{%
+ \def\normalsize{%
+ \@setsize\normalsize{12pt}\xpt\@xpt
+ \abovedisplayskip 10\p@ plus2\p@ minus5\p@
+ \belowdisplayskip \abovedisplayskip
+ \abovedisplayshortskip \abovedisplayskip
+ \belowdisplayshortskip \abovedisplayskip
+ \let\@listi\@listI
+ }%
+ }%
+ }%
+ }%
+% \end{macrocode}
+%
+% Footnote mods:
+% \begin{macrocode}
+ \footnotesep 9.25pt
+ \skip\footins 36pt plus 4pt minus 2pt
+ \def\footnoterule{\kern-13pt\hrule width.5in\kern15.6pt\relax}%
+% \end{macrocode}
+%
+%
+% \subsection{Sectioning}
+%
+% \begin{macrocode}
+\def\thepart{\Roman{part}} %
+\def\thesection{\Roman{section}}
+\def\p@section{}
+\def\thesubsection{\Alph{subsection}}
+\def\p@subsection{\thesection.}
+\def\thesubsubsection{\arabic{subsubsection}}
+\def\p@subsubsection{\thesection.\thesubsection.}
+\def\theparagraph{\alph{paragraph}}
+\def\p@paragraph{\thesection.\thesubsection.\thesubsubsection.}
+\def\thesubparagraph{\arabic{subparagraph}}
+\def\p@subparagraph{\thesection.\thesubsection.\thesubsubsection.\theparagraph.}
+% \end{macrocode}
+%
+% In RMP, put a period (.), followed by nut space, after the section number.
+% Also, hang the section number (the \LaTeX\ default).
+% \begin{macrocode}
+\def\@seccntformat#1{\csname the#1\endcsname.\hskip0.5em\relax}%
+% \end{macrocode}
+%
+% Note that we wish to set the section head uppercase, so we use
+% David Carlisle's \cmd\MakeTextUppercase. However, because this procedure effectively
+% parses its argument (looking for things to \emph{not} translate), it has to be invoked
+% in such a way that the argument of the \cmd\section\ command is passed to it as its
+% own argument.
+%
+% To accomplish this, we use the \cmd\@hangfrom@\ hook, which was developed for this
+% purpose.
+% \begin{macrocode}
+ \def\section{%
+ \@startsection{section}{1}{\z@}{0.8cm plus1ex minus.2ex}{0.4cm}%
+ {%
+ \small\sffamily\bfseries\selectfont
+ \raggedright
+ \parindent\z@
+ }%
+ }%
+ \def\@hangfrom@section#1#2#3{\@hangfrom{#1#2}\MakeTextUppercase{#3}}%
+ \def\subsection{%
+ \@startsection{subsection}{2}{\z@}{0.8cm plus1ex minus.2ex}{0.4cm}%
+ {%
+ \small\sffamily\bfseries
+ \raggedright
+ \parindent\z@
+ }%
+ }%
+ \def\subsubsection{%
+ \@startsection{subsubsection}{3}{\z@}{.8cm plus1ex minus.2ex}{0.4cm}%
+ {%
+ \small\sffamily\selectfont
+ \raggedright
+ \parindent\z@
+ }%
+ }%
+ \def\paragraph{%
+ \@startsection{paragraph}{4}{\z@}{.8cm plus1ex minus.2ex}{-1em}%
+ {%
+ \small\slshape\selectfont
+ \raggedright
+ \parindent\z@
+ }%
+ }%
+ \def\subparagraph{%
+ \@startsection{subparagraph}{4}{\parindent}{3.25ex plus1ex minus.2ex}{-1em}%
+ {\normalsize\bfseries\selectfont}%
+ }%
+%
+ \setcounter{tocdepth}{4}% FIXME: has no effect
+% \end{macrocode}
+%
+% \begin{macro}{\appendix}
+% \begin{macro}{\@hangfrom@appendix}
+% \begin{macro}{\@hangfroms@appendix}
+% \begin{macro}{\@appendixcntformat}
+% \begin{macrocode}
+\appdef\appendix{%
+ \let\@hangfrom@section\@hangfrom@appendix
+ \let\@hangfroms@section\@hangfroms@appendix
+ \let\@sectioncntformat\@appendixcntformat
+}%
+\def\@hangfrom@appendix#1#2#3{%
+ #1%
+ \@if@empty{#2}{%
+ #3%
+ }{%
+ #2\@if@empty{#3}{}{:\ #3}%
+ }%
+}%
+\def\@hangfroms@appendix#1#2{%
+ #1\appendixname\@if@empty{#2}{}{:\ #2}%
+}%
+\def\@appendixcntformat#1{\appendixname\ \csname the#1\endcsname}%
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+%
+% \subsection{Figure and Table Caption Formatting}
+% \begin{macro}{\@makecaption}
+%
+% \begin{macrocode}
+\setlength\belowcaptionskip{2\p@}
+\long\def\@makecaption#1#2{%
+ \vskip\abovecaptionskip
+ \vbox{%
+ \flushing
+ \small\rmfamily
+% \hsize\@capwidth
+ \noindent
+ #1\nobreak\hskip.5em plus.2em\ignorespaces#2\par
+ }%
+ \vskip\belowcaptionskip
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \subsection{Citations and Bibliography}%
+%
+% Customize \revtex\ for the journal substyle;
+% this task requires three components:
+% a \BibTeX\ \file{.bst} style file,
+% customizing code for \file{natbib}, and
+% customizations of the \env{thebibliography} environment.
+%
+% \begin{macro}{\@bibstyle}
+% Define the argument of the \cmd\bibliographystyle\ command
+% (if the document does not do so).
+%
+% The user must have installed a \file{.bst} file of the corresponding name.
+% This file will then be used by \BibTeX\ when compiling the document's \file{.bbl} file.
+%
+% To generate \file{apsrmp.bst}, use \classname{custom-bib} version 3.89d1 or later.
+% Run the \file{.bst} generator, \file{makebst.tex}, with the following options:
+%
+% \begin{enumerate}
+% \item
+% STYLE OF CITATIONS: a: ay( Author-year with some non-standard interface)
+% \item
+% AUTHOR-YEAR SUPPORT SYSTEM: nat: (Natbib for use with natbib v5.3 or later)
+% \item
+% ORDERING OF REFERENCES: *: (Alphabetical);
+% \item
+% ORDER ON VON PART: x: vonx (Sort without von part---de la Maire after Mahone)
+% \item
+% AUTHOR NAMES: a: nm-rev1 (Only first name reversed, initials---AGU style: Smith, J. F., H. K. Jones)
+% \item
+% POSITION OF JUNIOR: jnrlst (Junior comes last as Smith, John, Jr.)
+% \item
+% NUMBER OF AUTHORS: l: max 12, min 12 (three prompts in all);
+% \item
+% TYPEFACE FOR AUTHORS IN LIST OF REFERENCES: u: nmft, nmft-def (User defined author font---\cmd\bibnamefont)
+% \item
+% FONT FOR FIRST NAMES: u: fnm-def (First names in user defined font---\cmd\bibfnamefont)
+% \item
+% EDITOR NAMES IN INCOLLECTION ETC: a: nmfted (Editors incollection like authors font)
+% \item
+% FONT FOR `AND' IN LIST: r: nmand-rm (`And' in normal font---JONES and JAMES)
+% \item
+% FONT OF CITATION LABELS IN TEXT: u: lab, lab-def (User-defined citation font---\cmd\citenamefont)
+% \item
+% LABEL WHEN AUTHORS MISSING: keyxyr (Year blank when KEY replaces missing author---for natbib 7.0)
+% \item
+% DATE POSITION: b: dt-beg (Date after authors)
+% \item
+% SUPPRESS MONTH: xmth (Date is year only)
+% \item
+% DATE FORMAT: yr-com (Date preceded by comma as `, 1993')
+% \item
+% TRUNCATE YEAR: note-yr (Year text full as 1990--1993 or `in press')
+% \item
+% ARTICLE TITLE PRESENT: x: jtit-x (No article title);
+% \item
+% JOURNAL TITLE FONT: r: jttl-rm (Journal title normal);
+% \item
+% TECHNICAL REPORT TITLE: b: trtit-b (Tech. report title like books);
+% \item
+% JOURNAL VOLUME: b: vol-bf (Volume bold);
+% \item
+% VOLUME PUNCTUATION: c: volp-com (Volume with comma);
+% \item
+% PAGE NUMBERS: f: jpg-1 (Only start page number);
+% \item
+% POSITION OF PAGES: e: pp-last (Pages at end, but before any notes)
+% \item
+% NUMBER AND SERIES FOR COLLECTIONS: num-xser (Allows number without series and suppresses word ``number'')
+% \item
+% JOURNAL NAME PUNCTUATION: x: jnm-x (Space after journal);
+% \item
+% PUBLISHER IN PARENTHESES: p: pub-par (Publisher in parentheses)
+% \item
+% PUBLISHER POSITION: p: pre-pub (Publisher before chapter, pages);
+% \item
+% ISBN NUMBER: *: isbn (Include ISBN);
+% \item
+% ISSN NUMBER: *: isbn (Include ISSN for periodicals);
+% \item
+% `EDITOR' AFTER NAMES: a: edpar (`Name (editor),' in parentheses, after name, comma after)
+% \item
+% EDITOR IN COLLECTIONS: b: edby (In booktitle, edited by\dots---where \dots is names)
+% \item
+% PUNCTUATION BETWEEN SECTIONS (BLOCKS): c: blk-com (Comma between blocks);
+% \item
+% `IN' WITH JOURNAL NAMES: i: injnl (Add `in' before journal name in style for incollection)
+% \item
+% ABBREVIATE WORD `PAGES': a: pp (`Page' abbreviated);
+% \item
+% ABBREVIATE WORD `EDITORS': a: ed (`Editor' abbreviated);
+% \item
+% ABBREVIATION FOR `EDITION': a: ednx (`Edition' abbreviated as `ed');
+% \item
+% EDITION NUMBERS: xedn (Editions as in database saving much processing memory)
+% \item
+% STORED JOURNAL NAMES: a: jabr (Abbreviated journal names);
+% \item
+% FONT OF `ET AL': i: etal-it (Italic et al);
+% \item
+% ADDITIONAL REVTeX DATA FIELDS: r: revdata, eprint, url, url-blk (Include REVTeX data fields)
+% \item
+% NEW FONT SELECTION SCHEME: n: nfss (NFSS);
+% \item
+% ADDITIONAL REVTeX DATA FIELDS: y: revdata (additional data fields);
+% \item
+% REFERENCE COMPONENT TAGS: r: bibinfo (reference component tags);
+% \end{enumerate}
+%
+% A file \file{apsrmp.dbj} file equivalent to the following should result:
+% \begin{verbatim}
+% \input docstrip
+% \preamble
+% ----------------------------------------
+% *** REVTeX-compatible apsrmp.bst ***
+% \endpreamble
+% \postamble
+% End of customized bst file
+% \endpostamble
+% \keepsilent
+% \askforoverwritefalse
+% \def\MBopts{\from{merlin.mbs}{%
+% ay%: Author-year with some non-standard interface
+% ,nat%: Natbib for use with natbib v5.3 or later
+% ,vonx%: Sort without von part (de la Maire after Mahone)
+% ,nm-rev1%: Only first name reversed, initials (AGU style: Smith, J. F., H. K. Jones)
+% ,jnrlst%: Junior comes last as Smith, John, Jr.
+% ,nmlm%: Limited authors (et al replaces missing names)
+% ,x10%: Maximum of 10 authors
+% ,m10%: Minimum of 10 authors
+% ,nmft,nmft-def%: User defined author font (\bibnamefont)
+% ,fnm-def%: First names in user defined font (\bibfnamefont)
+% ,nmfted%: Editors incollection like authors font
+% ,nmand-rm%: `And' in normal font (JONES and JAMES)
+% ,lab,lab-def%: User defined citation font (\citenamefont)
+% ,keyxyr%: Year blank when KEY replaces missing author (for natbib 7.0)
+% ,dt-beg%: Date after authors
+% ,xmth%: Date is year only
+% ,yr-com%: Date preceded by comma as `, 1993'
+% ,note-yr%: Year text full as 1990--1993 or `in press'
+% ,jtit-x%: No article title
+% ,jttl-rm%: Journal name normal font
+% ,trtit-b%: Tech. report title like books
+% ,vol-bf%: Volume bold as {\bf vol}(num)
+% ,volp-com%: Volume with comma as vol(num), ppp
+% ,jpg-1%: Only start page number
+% ,pp-last%: Pages at end but before any notes
+% ,num-xser%: Allows number without series and suppresses word "number"
+% ,jnm-x%: Space after journal name
+% ,pub-par%: Publisher in parentheses
+% ,pre-pub%: Publisher before chapter, pages
+% ,isbn%: Include ISBN for books, booklets, etc.
+% ,issn%: Include ISSN for periodicals
+% ,edpar%: `Name (editor),' in parentheses, after name, comma after
+% ,edby%: In booktitle, edited by .. (where .. is names)
+% ,blk-com%: Comma between blocks
+% ,injnl%: Add `in' before journal name in style for incollection
+% ,pp%: `Page' abbreviated as p. or pp.
+% ,ed%: `Editor' abbreviated as ed. or eds.
+% ,xedn%: Editions as in database saving much processing memory
+% ,jabr%: Abbreviated journal names
+% ,etal-it%: Italic et al
+% ,revdata,eprint,url,url-blk%: Include REVTeX data fields collaboration, eid, eprint, numpages, url
+% ,bibinfo%: Reference component tags like \bibinfo in the content of \bibitem
+% ,nfss%: NFSS use \textbf, \emph, not \bf, \em
+% }}
+% \generate{\file{\jobname.bst}{\MBopts}}
+% \endbatchfile
+% \end{verbatim}
+%
+% \begin{macrocode}
+\appdef\class@documenthook{%
+ \@ifxundefined\@bibstyle{%
+ \def\@bibstyle{apsrmp}%
+ }{}%
+}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\bibpunct}
+% The following commands effectively etablish the style in which \cmd\cite\ commands are
+% formatted.
+% You can think of them as the second needed component for the bibliography.
+%
+% Set up for author-year citations:
+% when \cmd\NAT@set@cites\ executes (at \envb{document} time), the \cmd\@biblabel\ will
+% be set to \cmd\NAT@biblabel.
+%
+% Per Karie Friedman (friedman@phys.washington.edu), multiple citations are separated by semicolons, e.g.,
+% (Jones, 1999; Abbott and Smith, 2000; Wortley, 2001a),
+% and
+% multiple citations by the same author by commas, e.g.,
+% Abela et al. (1995, 1997a, 1997b).
+% The third argument of \cmd\bibpunct\ handles the former.
+%
+% The fifth argument puts a comma after the author when the year is not in parens:
+% (Lee et al., 1996).
+%
+% \begin{macrocode}
+ \bibpunct{(}{)}{;}{a}{,}{,}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\cite}
+% Per Mark Doyle, \cmd\cite\ is mapped to \cmd\citep\ in RMP.
+%
+% \begin{macrocode}
+ \let\cite\citep
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\bibsection}
+% \begin{macro}{\bibpreamble}
+% \begin{macro}{\newblock}
+% \begin{macro}{\bibhang}
+% \begin{macro}{\bibsep}
+% We define the sectioning command to use when starting the bibliography.
+%
+% \begin{macrocode}
+\def\bibsection{%
+ \let\@hangfroms@section\@hang@froms
+ \section*{\refname}%
+ \@nobreaktrue
+}%
+\let\bibpreamble\@empty
+\def\newblock{\ }%
+\bibhang10\p@
+\bibsep\z@
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}{\footinbib@sw}
+% Footnotes in bibliography are inconsistent with author-year
+% references, and are particularly nast under \classname{natbib}:
+% the packcage will automatically change to numbered references if
+% any \cmd\bibitem\ commands lack the optional argument.
+% Therefore, we must uninvoke it now, even if invoked by the document.
+% \begin{macrocode}
+\@ifxundefined\footinbib@sw{}{%
+ \footinbib@sw{%
+ \class@warn{%
+ Footnotes in bibliography are incompatible with RMP.^^J%
+ Undoing the footinbib option.
+ }%
+ }{}%
+}%
+\@booleanfalse\footinbib@sw
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\eprint}
+% RMP requires the \cmd\eprint\ field in the bib entry to be set off with the word ``eprint''.
+% \begin{macrocode}
+\def\eprint#1{eprint #1}%
+% \end{macrocode}
+% \end{macro}
+%
+% \subsection{Table of Contents}
+%
+% We set up for auto-sizing of certain TOC elements.
+%
+% To do this, we override certain definitions for
+% the default TOC font (\cmd\toc@@font) and spacing (\@dotsep),
+% and define
+% formatting for the needed elements (\cmd\l@\dots).
+% Finally, we activate the autosizing by
+% assigning \cmd\toc@pre\ and \cmd\toc@post.
+%
+% \begin{macro}{\toc@@font}
+% Set the formatting characteristics of the auto-indenting part
+% of the TOC.
+% \begin{macrocode}
+\def\toc@@font{%
+ \footnotesize\rmfamily
+ \def\\{\space\ignorespaces}%
+}%
+\def\@dotsep{5.5pt}%
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\l@section}
+% Determine which TOC elements are automatically indented.
+% \begin{macrocode}
+\def\tocleft@{\z@}%
+\def\tocdim@min{5\p@}%
+\def\l@section{%
+ \l@@sections{}{section}% Implicit #3#4
+}%
+\def\l@subsection{%
+ \l@@sections{section}{subsection}% Implicit #3#4
+}%
+\def\l@subsubsection{%
+ \l@@sections{subsection}{subsubsection}% Implicit #3#4
+}%
+%\def\l@subsubsection#1#2{}%
+\def\l@paragraph#1#2{}%
+\def\l@subparagraph#1#2{}%
+% \end{macrocode}
+% \end{macro}
+%
+% Activate the TOC processing.
+% \begin{macrocode}
+\let\toc@pre\toc@pre@auto
+\let\toc@post\toc@post@auto
+% \end{macrocode}
+%
+%
+% \begin{macrocode}
+%</rmp>
+% \end{macrocode}
+%
+%
+% \Finale
+% %Here ends the programmer's documentation.
+% \endinput
+%
+\endinput
diff --git a/Master/texmf-dist/source/latex/revtex4/revtex4.ins b/Master/texmf-dist/source/latex/revtex4/revtex4.ins
new file mode 100644
index 00000000000..96cf5b09722
--- /dev/null
+++ b/Master/texmf-dist/source/latex/revtex4/revtex4.ins
@@ -0,0 +1,36 @@
+% This file requires docstrip version 2.4 or higher available from
+% ftp://ctan.tug.org/tex-archive/macros/latex/unpacked/docstrip.tex
+\input docstrip
+\preamble
+
+This file is part of the APS files in the REVTeX 4 distribution.
+For the version number, search on the string %FileInfo
+
+Original version by David Carlisle
+Modified by Arthur Ogawa (mailto:ogawa@teleport.com)
+
+Copyright (c) 1999 The American Physical Society.
+http://publish.aps.org/revtex4/
+mailto:revtex4@aps.org
+
+See the REVTeX 4 README file for restrictions and more information.
+
+\endpreamble
+\keepsilent
+\askforoverwritefalse
+\generate{
+ \file{revtex4.cls}{%
+ \from{revtex4.dtx}{class-pre}%
+ \from{ltxutil.dtx}{ltxutil-krn}%
+ \from{ltxgrid.dtx}{ltxgrid-krn}%
+ \from{textcase.dtx}{package}%
+ \from{revtex4.dtx}{class-pst}%
+ }%
+ \file{aps.rtx}{\from{revtex4.dtx}{aps}}
+ \file{rmp.rtx}{\from{revtex4.dtx}{rmp}}
+ \file{10pt.rtx}{\from{revtex4.dtx}{10pt}}
+ \file{11pt.rtx}{\from{revtex4.dtx}{11pt}}
+ \file{12pt.rtx}{\from{revtex4.dtx}{12pt}}
+ \file{revsymb.sty}{\from{revtex4.dtx}{revsymb}}
+}
+\endbatchfile
diff --git a/Master/texmf-dist/source/latex/revtex4/textcase.dtx b/Master/texmf-dist/source/latex/revtex4/textcase.dtx
new file mode 100644
index 00000000000..60010e5ee67
--- /dev/null
+++ b/Master/texmf-dist/source/latex/revtex4/textcase.dtx
@@ -0,0 +1,404 @@
+% \iffalse
+%% Source File: textcase.dtx
+%% Copyright 1997 1998 David Carlisle
+%%
+%% This file may be distributed under the terms of the LPPL.
+%% See 00readme.txt for details.
+%
+%<*dtx>
+ \ProvidesFile{textcase.dtx}
+%</dtx>
+%<package>\NeedsTeXFormat{LaTeX2e}
+%<package>\ProvidesPackage{textcase}
+%<driver> \ProvidesFile{textcase.drv}
+% \fi
+% \ProvidesFile{textcase.dtx}
+ [1998/11/12 v0.06 Text only upper/lower case changing (DPC)]
+%
+% \iffalse
+%<*driver>
+\documentclass{ltxdoc}
+\usepackage{textcase}
+\begin{document}
+\DocInput{textcase.dtx}
+\end{document}
+%</driver>
+% \fi
+%
+% \GetFileInfo{textcase.dtx}
+% \CheckSum{110}
+%
+% \changes{v0.01}{1997/09/28}
+% {Posted to c.t.t}
+% \changes{v0.02}{1997/09/28}
+% {Support \cs{ensuremath}}
+% \changes{v0.03}{1997/12/12}
+% {Donald Arseneau: support \cs{ref} \cs{cite} \cs{label}}
+% \changes{v0.04}{1997/12/14}
+% {Convert to dtx form, add \cs{NoCaseChange}}
+% \changes{v0.04}{1997/12/14}
+% {Support (most uses of) optional argument of \cs{cite}}
+% \changes{v0.05}{1997/12/16}
+% {More comments from Donald,
+% perhaps I should just give him this package for Christmas}
+% \changes{v0.06}{1998/11/12}
+% {overload option}
+%
+%
+% \title{The \textsf{textcase} package\thanks{This file
+% has version number \fileversion, last
+% revised \filedate.}}
+% \author{David Carlisle\thanks{%^^A
+% Reorganisation to suppport skipping a wider class of
+% things than just math done by Donald Arseneau}}
+%
+% \date{\filedate}
+%
+% \maketitle
+%
+% \section{Introduction}\label{intro}
+% |\MakeTextUppercase| and |\MakeTextLowercase| are versions of the
+% standard |\MakeUppercase| and |\MakeLowercase| that do not change
+% the case of any math sections in their arguments.
+%\begin{verbatim}
+% \MakeTextUppercase{abc\ae\ \( a = b \) and $\alpha \neq a$
+% or even \ensuremath{x=y} and $\ensuremath{x=y}$}
+%\end{verbatim}
+% Should produce:
+% \begin{quotation}
+% ABC\AE\ \( a = b \) AND $\alpha \neq a$
+% OR EVEN \ensuremath{x=y} AND $\ensuremath{x=y}$
+% \end{quotation}
+%
+% Version 0.03 of this package incorporates some changes suggested by
+% Donald Arseneau so that as well as math mode, the arguments of
+% |\cite|, |\label| and |\ref| are also prevented from being uppercased.
+% So you can now go
+%\begin{verbatim}
+% \MakeTextUppercase{%
+% Text in section~\ref{intro}, about \cite[pp 2--4]{bbb}}
+%\end{verbatim}
+% which produces
+% \begin{quotation}
+% \MakeTextUppercase{%
+% Text in section~\ref{intro}, about \cite[pp 2--4]{bbb}}
+% \end{quotation}
+% If, instead, the standard |\MakeUppercase| were used here, the keys
+% `into' and `bbb' would be uppercased and generate errors about
+% undefined references to INTRO and BBB.
+%
+% Sometimes there may be a special section of text that should not be
+% uppercased. This can be marked with |\NoCaseChange|, as follows.
+%\begin{verbatim}
+% \MakeTextUppercase{%
+% Text \NoCaseChange{More Text} yet more text}
+%\end{verbatim}
+% which produces
+% \begin{quotation}
+% \MakeTextUppercase{%
+% Text \NoCaseChange{More Text} yet more text}
+% \end{quotation}
+%
+% |\NoCaseChange| has other uses. If for some reason you need a
+% tabular environment within an uppercased section, then you need
+% to ensure that the name `tabular' and the preamble (eg `ll')
+% does not get uppercased:
+%\begin{verbatim}
+% \MakeTextUppercase{%
+% Text \NoCaseChange{\begin{tabular}{ll}}%
+% table&stuff\\goes&here
+% \NoCaseChange{\end{tabular}}
+% More text}
+%\end{verbatim}
+% which produces
+% \begin{quotation}
+% \MakeTextUppercase{%
+% Text \NoCaseChange{\begin{tabular}{ll}}%^^A
+% table&stuff\\goes&here
+% \NoCaseChange{\end{tabular}}
+% More text}
+% \end{quotation}
+%
+% \section{Features and Foibles}
+%
+% \subsection{Nested text}
+% The commands defined here only skip math sections and |\ref| arguments
+% if they are not `hidden' inside a |{ }| brace group. All text inside
+% such a group will be made uppercase just as with the standard
+% |\MakeUppercase|.
+%\begin{verbatim}
+% \MakeTextUppercase{a b {c $d$} $e$}
+%\end{verbatim}
+% produces
+% \begin{quotation}
+% \MakeTextUppercase{a b {c $d$} $e$}
+% \end{quotation}
+% Of course, this restriction does not apply to the arguments of the
+% supported commands |\ensuremath|, |\label|, |\ref|, and |\cite|.
+%
+% If you cannot arrange for your
+% mathematics to be at the outer level of brace grouping, you should
+% use the following basic technique (which works even with the standard
+% |\MakeUppercase| command). Define a new command that expands to your
+% math expression, and then use that command, with |\protect|, in the
+% text to be uppercased. Note that if the text being uppercased is in a
+% section title or other moving argument you may need to make the
+% definition in the document preamble, rather than just before the
+% section command, so that the command is defined when the table of
+% contents file is read.
+%\begin{verbatim}
+% \MakeTextUppercase{%
+% Text \fbox{$a=b$ and $x=y$}}%
+%
+% \newcommand{\mathexprone}{$a=b$}
+% \newcommand{\mathexprtwo}{$x=y$}
+% \MakeTextUppercase{%
+% Text \fbox{\protect\mathexprone\ and \protect\mathexprtwo}}%
+%\end{verbatim}
+% which produces
+% \begin{quotation}
+% \MakeTextUppercase{%
+% Text \fbox{$a=b$ and $x=y$}}%
+%
+% \newcommand{\mathexprone}{$a=b$}
+% \newcommand{\mathexprtwo}{$x=y$}
+% \MakeTextUppercase{%
+% Text \fbox{\protect\mathexprone\ and \protect\mathexprtwo}}%
+% \end{quotation}
+%
+% \subsection{Citations}
+% As documented above, |\cite| and |\ref| commands are not uppercased by
+% |\MakeTextUppercase|. If you are using a non-numeric citation scheme
+% you may want the replacement text for |\cite| to be uppercased.
+%
+% It is difficult to arrange that |\MakeTextUppercase| uppercases such
+% text, not least because this would lead to interaction with the many
+% bibliography packages which redefine |\cite| one way or another. One
+% possibility to achieve this is to use Donald Arseneau's cite package
+% and to locally redefine |\citeform| to add |\MakeUppercase| around
+% the final text string produced by |\cite|.
+%\begin{verbatim}
+% \MakeTextUppercase{%
+% Text \cite{bbb} and \cite{ccc}}
+%
+% {\renewcommand\citeform{\MakeUppercase}\MakeTextUppercase{%
+% Text \cite{bbb} and \cite{ccc}}}
+%\end{verbatim}
+% which produces\footnote{This is faked, so this document does not
+% rely on \texttt{cite.sty} being installed}
+% \begin{quotation}
+% TEXT [1] AND [David Carlisle 1997]
+%
+% TEXT [1] AND [DAVID CARLISLE 1997]
+% \end{quotation}
+%
+% \subsection{overload Option}
+% By default the package only defines new commands, |\MakeTextUppercase|
+% and |\MakeTextlowercase| as described above. You may wish to redefine
+% the standard |\MakeUppercase| and |\MakeLowercase| commands to be
+% the same as these new commands. So that for example headings in the
+% book class have this new feature without any further redefinition.
+% You may use the package option |[overload]| in which case these
+% new definitions will be overloaded on to the existing command names.
+%
+% \begin{thebibliography}{9}
+%
+% \bibitem{bbb} Something numeric
+% \bibitem[David Carlisle 1997]{ccc} Something textual
+% \end{thebibliography}
+%
+% \StopEventually{}
+%
+% \section{Implementation}
+%
+% \begin{macrocode}
+%<*package>
+% \end{macrocode}
+%
+% \begin{macro}{\@uclcnotmath}
+% This is the main macro of this package. It is basically
+% a copy of |\MakeTextUppercase| and |\MakeTextLowercase|
+% from the \LaTeX\ kernel, modified slightly so that they
+% can share code (that modification could be done to the
+% standard versions as well) and then further changed to
+% skip certain features like math mode and |\label| arguments.
+%
+% The arguments are:\\
+% |#1|: Extra commands to apply for case changing.
+% Used to locally redefine |\i| and |\j| for uppercasing.\\
+% |#2|: Either |##1##2| or |##2##1| to control the order
+% in which |\let| is applied to the pairs of control
+% sequences in |\@uclclist|.\\
+% |#3|: |\uppercase| or |\lowercase|.\\
+% |#4|: The text to be upper (or lower) cased.
+% \begin{macrocode}
+\def\@uclcnotmath#1#2#3#4{\begingroup
+% \end{macrocode}
+% Run extra commands (currently just to redefine |\i| and |\j|).
+% \begin{macrocode}
+ #1%
+% \end{macrocode}
+%
+% Locally set |\( \)| to be just |$ $|, so that the math skipping
+% code can be simplified, just to look for |$|.
+% \begin{macrocode}
+ \def\({$}\let\)\(%
+% \end{macrocode}
+%
+% Set up the `non-math' things that also have to be skipped.
+% \begin{macrocode}
+ \def\NoCaseChange##1{\noexpand\NoCaseChange{\noexpand##1}}%
+ \@nonchangecase\label
+ \@nonchangecase\ref
+ \@nonchangecase\ensuremath
+% \end{macrocode}
+%
+% |\cite| a bit trickier, as we want to uppercase any optional argument.
+% This will fail if the optional argument contains a brace group, but
+% should catch most cases.
+%
+% |text \cite[page 1]{foo} more text| ends up as
+%\begin{verbatim}
+% \uppercase{text \toks@{\cite[page1]}%
+% \the\toks@{foo}%
+% \uppercase{ more text}
+%\end{verbatim}
+% \begin{macrocode}
+ \def\cite##1##{\toks@{\noexpand\cite##1}\@citex}%
+ \def\@citex##1{\NoCaseChange{\the\toks@{##1}}}%
+% \end{macrocode}
+% (|\@citex| is a scratch macro here, not a redefinition of the existing
+% |\@citex|.)
+%
+% The following is essentially taken from |\MakeUppercase|.
+% Recursively execute |\reserved@a| to |\let| the pairs in |\@uclclist|.
+% The strange construction with |\@gobble| at the end just gobbles the
+% final recursive call.
+%
+% Incidentally, packages should not use the |\reserved@|\ldots\ scratch
+% macros, which are `reserved' for use within the \LaTeX\ kernel, but
+% (a) this code is essentially a copy from the kernel, and
+% (b) I'm allowed to break the rules, so there.
+% \begin{macrocode}
+ \def\reserved@a##1##2{\let#2\reserved@a}%
+ \expandafter\reserved@a\@uclclist\reserved@b{\reserved@b\@gobble}%
+% \end{macrocode}
+%
+% Expand everything first so that the `skipping' code can see what to
+% skip and so that tokens are revealed to |\uppercase|.
+% This makes the argument `moving'.
+% The |$\valign$| is just a fake math expression used to terminate
+% the parsing done by |\@skipmath|.
+% \begin{macrocode}
+ \protected@edef\reserved@a{\endgroup
+ \noexpand\@skipmath#3#4$\valign$}%
+% \end{macrocode}
+%
+% \begin{macrocode}
+ \reserved@a}
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\@nonchangecase}
+% \begin{macrocode}
+\def\@nonchangecase#1{\def#1##1{\NoCaseChange{#1{##1}}}}
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\NoCaseChange}
+% For hiding arbitrary text from |\uppercase|. This innocuous
+% definition is used for any occurrence of |\NoCaseChange|
+% in text that is not passed to |\MakeTextUppercase|. For example a
+% section heading may be uppercased, but the toc entry may not.
+% It is also used for nested definitions of |\cite| etc, where the
+% |\NoCaseChange| is inserted by expansion, but not removed as it is
+% hidden by the brace group.
+% \begin{macrocode}
+\let\NoCaseChange\@firstofone
+% \end{macrocode}
+% \end{macro}
+%
+%\begin{macro}{\@skipmath}
+% |#1|: operation |\uppercase| / |\lowercase|\\
+% |#2|: text up to first (next) |$|\\
+% |#3|: first math mode material (or |\valign| sentinel)
+% \begin{macrocode}
+\def\@skipmath#1#2$#3${%
+ \@skip@nonchangecase#1#2\NoCaseChange\valign
+ \ifx\valign#3%
+ \else
+ $#3$%
+ \expandafter\@skipmath\expandafter#1%
+ \fi}
+% \end{macrocode}
+% \end{macro}
+%
+%\begin{macro}{\@skip@nonchangecase}
+% |#1|: |\uppercase| or |\lowercase|\\
+% |#2|: Text up to the first command (such as |\cite|) made `safe'.
+% or the first use of |\NoCaseChange|.
+% \begin{macrocode}
+\def\@skip@nonchangecase#1#2\NoCaseChange#3{%
+ #1{#2}%
+ \ifx\valign#3%
+ \else
+ #3%
+ \expandafter\@skip@nonchangecase\expandafter#1%
+ \fi}
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\MakeTextUppercase}
+% \changes{v0.06}{1998/11/12}
+% {Add edef to make behaviour more like \cs{MakeUppercase}}
+% Put it all together. Arrange for \i\ and \j\ to uppercase, and
+% to read the pairs in the ucase list `forwards'.
+% \begin{macrocode}
+\DeclareRobustCommand\MakeTextUppercase{%
+ \@uclcnotmath{\def\i{I}\def\j{J}}{##1##2}\uppercase}
+% \end{macrocode}
+%
+% \begin{macrocode}
+\protected@edef\MakeTextUppercase#1{\MakeTextUppercase{#1}}
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\MakeTextLowercase}
+% \changes{v0.06}{1998/11/12}
+% {Add edef to make behaviour more like \cs{MakeLowercase}}
+% probably will never be used, but it is easy to implement.
+% Use |\lowercase| and read the pairs in the uppercase list `backwards'.
+% \begin{macrocode}
+\DeclareRobustCommand\MakeTextLowercase{%
+ \@uclcnotmath{}{##2##1}\lowercase}
+% \end{macrocode}
+%
+% \begin{macrocode}
+\protected@edef\MakeTextLowercase#1{\MakeTextLowercase{#1}}
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\MakeUppercase}
+% \begin{macro}{\MakeLowercase}
+% \changes{v0.06}{1998/11/12}
+% {Option to overload added. latex/2906}
+% \begin{macrocode}
+\DeclareOption{overload}{%
+ \expandafter\let\csname MakeUppercase \expandafter\endcsname
+ \csname MakeTextUppercase \endcsname
+ \expandafter\let\csname MakeLowercase \expandafter\endcsname
+ \csname MakeTextLowercase \endcsname}
+% \end{macrocode}
+%
+% \begin{macrocode}
+\ProcessOptions
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macrocode}
+%</package>
+% \end{macrocode}
+%
+% \Finale
+%