summaryrefslogtreecommitdiff
path: root/Master/texmf-dist/source/latex/l3kernel/l3str.dtx
diff options
context:
space:
mode:
authorKarl Berry <karl@freefriends.org>2015-09-28 21:17:44 +0000
committerKarl Berry <karl@freefriends.org>2015-09-28 21:17:44 +0000
commit472e2dc119ba3eca6a8672e7606df6a7dc847a19 (patch)
tree7c10e4ee61b72663387bc5e19e97ddb6c8e57c62 /Master/texmf-dist/source/latex/l3kernel/l3str.dtx
parent3b599d9910b4fe668c87ddd32a48d48f8a377560 (diff)
l3 (28sep15)
git-svn-id: svn://tug.org/texlive/trunk@38490 c570f23f-e606-0410-a88d-b1316a301751
Diffstat (limited to 'Master/texmf-dist/source/latex/l3kernel/l3str.dtx')
-rw-r--r--Master/texmf-dist/source/latex/l3kernel/l3str.dtx1070
1 files changed, 977 insertions, 93 deletions
diff --git a/Master/texmf-dist/source/latex/l3kernel/l3str.dtx b/Master/texmf-dist/source/latex/l3kernel/l3str.dtx
index 63129bb4fcb..34d57cf2ff7 100644
--- a/Master/texmf-dist/source/latex/l3kernel/l3str.dtx
+++ b/Master/texmf-dist/source/latex/l3kernel/l3str.dtx
@@ -1,6 +1,6 @@
% \iffalse meta-comment
%
-%% File: l3str.dtx Copyright (C) 2014-2015 The LaTeX3 Project
+%% File: l3str.dtx Copyright (C) 2011-2015 The LaTeX3 Project
%%
%% It may be distributed and/or modified under the conditions of the
%% LaTeX Project Public License (LPPL), either version 1.3c of this
@@ -37,7 +37,7 @@
\documentclass[full]{l3doc}
%</driver>
%<*driver|package>
-\GetIdInfo$Id: l3str.dtx 5983 2015-09-10 18:57:56Z joseph $
+\GetIdInfo$Id: l3str.dtx 6021 2015-09-20 15:10:27Z joseph $
{L3 Strings}
%</driver|package>
%<*driver>
@@ -79,33 +79,163 @@
% space characters which have category code $10$ (\enquote{space}). Thus
% at a technical level, a \TeX{} string is a token list with the appropriate
% category codes. In this documentation, these will simply be referred to as
-% strings: note that they can be stored in token lists as normal.
+% strings.
%
-% The functions documented here take literal token lists,
-% convert to strings and then carry out manipulations. Thus they may
-% informally be described as \enquote{ignoring} category code. Note that
-% the functions \cs{cs_to_str:N}, \cs{tl_to_str:n}, \cs{tl_to_str:N} and
+% String variables are simply specialised token lists, but by convention
+% should be named with the suffix \texttt{\ldots{}str}. Such variables
+% should contain characters with category code $12$ (other), except
+% spaces, which have category code $10$ (blank space). All the
+% functions in this module which accept a token list argument first
+% convert it to a string using \cs{tl_to_str:n} for internal processing,
+% and will not treat a token list or the corresponding string
+% representation differently.
+%
+% Note that as string variables are a special case of token list variables
+% the coverage of \cs{str_\ldots{}:N} functions is somewhat smaller than
+% \cs{tl_\ldots{}:N}.
+%
+% The functions \cs{cs_to_str:N}, \cs{tl_to_str:n}, \cs{tl_to_str:N} and
% \cs{token_to_str:N} (and variants) will generate strings from the appropriate
% input: these are documented in \pkg{l3basics}, \pkg{l3tl} and \pkg{l3token},
% respectively.
%
-% \section{The first character from a string}
+% Most expandable functions in this module come in three flavours:
+% \begin{itemize}
+% \item \cs{str_...:N}, which expect a token list or string
+% variable as their argument;
+% \item \cs{str_...:n}, taking any token list (or string) as an
+% argument;
+% \item \cs{str_..._ignore_spaces:n}, which ignores any space
+% encountered during the operation: these functions are typically
+% faster than those which take care of escaping spaces
+% appropriately.
+% \end{itemize}
+%
+% \section{Building strings}
%
-% \begin{function}[added = 2011-08-10, EXP]{\str_head:n, \str_tail:n}
+% \begin{function}[added = 2015-09-18]{\str_new:N, \str_new:c}
% \begin{syntax}
-% \cs{str_head:n} \Arg{token list}
-% \cs{str_tail:n} \Arg{token list}
+% \cs{str_new:N} \meta{str~var}
+% \end{syntax}
+% Creates a new \meta{str~var} or raises an error if the name is
+% already taken. The declaration is global. The \meta{str~var} will
+% initially be empty.
+% \end{function}
+%
+% \begin{function}[added = 2015-09-18]
+% {\str_const:Nn, \str_const:Nx, \str_const:cn, \str_const:cx}
+% \begin{syntax}
+% \cs{str_const:Nn} \meta{str~var} \Arg{token list}
+% \end{syntax}
+% Creates a new constant \meta{str~var} or raises an error if the name
+% is already taken. The value of the \meta{str~var} will be set
+% globally to the \meta{token list}, converted to a string.
+% \end{function}
+%
+% \begin{function}[added = 2015-09-18]
+% {\str_clear:N, \str_clear:c, \str_gclear:N, \str_gclear:c}
+% \begin{syntax}
+% \cs{str_clear:N} \meta{str~var}
+% \end{syntax}
+% Clears the content of the \meta{str~var}.
+% \end{function}
+%
+% \begin{function}[added = 2015-09-18]{\str_clear_new:N, \str_clear_new:c}
+% \begin{syntax}
+% \cs{str_clear_new:N} \meta{str~var}
+% \end{syntax}
+% Ensures that the \meta{str~var} exists globally by applying
+% \cs{str_new:N} if necessary, then applies \cs{str_(g)clear:N} to leave
+% the \meta{str~var} empty.
+% \end{function}
+%
+% \begin{function}[added = 2015-09-18]
+% {
+% \str_set_eq:NN, \str_set_eq:cN, \str_set_eq:Nc, \str_set_eq:cc,
+% \str_gset_eq:NN, \str_gset_eq:cN, \str_gset_eq:Nc, \str_gset_eq:cc
+% }
+% \begin{syntax}
+% \cs{str_set_eq:NN} \meta{str~var_1} \meta{str~var_2}
% \end{syntax}
-% Converts the \meta{token list} into a string, as described for
-% \cs{tl_to_str:n}. The \cs{str_head:n} function then leaves
-% the first character of this string in the input stream.
-% The \cs{str_tail:n} function leaves all characters except
-% the first in the input stream. The first character may be
-% a space. If the \meta{token list} argument is entirely empty,
-% nothing is left in the input stream.
+% Sets the content of \meta{str~var_1} equal to that of
+% \meta{str~var_2}.
% \end{function}
%
-% \subsection{Tests on strings}
+% \section{Adding data to string variables}
+%
+% \begin{function}[added = 2015-09-18]
+% {
+% \str_set:Nn, \str_set:Nx, \str_set:cn, \str_set:cx,
+% \str_gset:Nn, \str_gset:Nx, \str_gset:cn, \str_gset:cx
+% }
+% \begin{syntax}
+% \cs{str_set:Nn} \meta{str var} \Arg{token list}
+% \end{syntax}
+% Converts the \meta{token list} to a \meta{string}, and stores the
+% result in \meta{str var}.
+% \end{function}
+%
+% \begin{function}[added = 2015-09-18]
+% {
+% \str_put_left:Nn, \str_put_left:Nx,
+% \str_put_left:cn, \str_put_left:cx,
+% \str_gput_left:Nn, \str_gput_left:Nx,
+% \str_gput_left:cn, \str_gput_left:cx
+% }
+% \begin{syntax}
+% \cs{str_put_left:Nn} \meta{str var} \Arg{token list}
+% \end{syntax}
+% Converts the \meta{token list} to a \meta{string}, and prepends the
+% result to \meta{str var}. The current contents of the \meta{str
+% var} are not automatically converted to a string.
+% \end{function}
+%
+% \begin{function}[added = 2015-09-18]
+% {
+% \str_put_right:Nn, \str_put_right:Nx,
+% \str_put_right:cn, \str_put_right:cx,
+% \str_gput_right:Nn, \str_gput_right:Nx,
+% \str_gput_right:cn, \str_gput_right:cx
+% }
+% \begin{syntax}
+% \cs{str_put_right:Nn} \meta{str var} \Arg{token list}
+% \end{syntax}
+% Converts the \meta{token list} to a \meta{string}, and appends the
+% result to \meta{str var}. The current contents of the \meta{str
+% var} are not automatically converted to a string.
+% \end{function}
+%
+% \subsection{String conditionals}
+%
+% \begin{function}[EXP, pTF, added = 2015-09-18]
+% {\str_if_exist:N, \str_if_exist:c}
+% \begin{syntax}
+% \cs{str_if_exist_p:N} \meta{str~var}
+% \cs{str_if_exist:NTF} \meta{str~var} \Arg{true code} \Arg{false code}
+% \end{syntax}
+% Tests whether the \meta{str~var} is currently defined. This does not
+% check that the \meta{str~var} really is a string.
+% \end{function}
+%
+% \begin{function}[EXP,pTF, added = 2015-09-18]
+% {\str_if_empty:N, \str_if_empty:c}
+% \begin{syntax}
+% \cs{sr_if_empty_p:N} \meta{str~var}
+% \cs{str_if_empty:NTF} \meta{str~var} \Arg{true code} \Arg{false code}
+% \end{syntax}
+% Tests if the \meta{string variable} is entirely empty
+% (\emph{i.e.}~contains no characters at all).
+% \end{function}
+%
+% \begin{function}[EXP,pTF, added = 2015-09-18]
+% {\str_if_eq:NN, \str_if_eq:Nc, \str_if_eq:cN, \str_if_eq:cc}
+% \begin{syntax}
+% \cs{str_if_eq_p:NN} \meta{str~var_1} \meta{str~var_2}
+% \cs{str_if_eq:NNTF} \meta{str~var_1} \meta{str~var_2} \Arg{true code} \Arg{false code}
+% \end{syntax}
+% Compares the content of two \meta{str variables} and
+% is logically \texttt{true} if the two contain the same characters.
+% \end{function}
%
% \begin{function}[EXP,pTF]
% {
@@ -189,6 +319,116 @@
% not be used within this string.
% \end{function}
%
+% \section{Working with the content of strings}
+%
+% \begin{function}[EXP, added = 2015-09-18]{\str_use:N, \str_use:c}
+% \begin{syntax}
+% \cs{str_use:N} \meta{str~var}
+% \end{syntax}
+% Recovers the content of a \meta{str~var} and places it
+% directly in the input stream. An error will be raised if the variable
+% does not exist or if it is invalid. Note that it is possible to use
+% a \meta{str} directly without an accessor function.
+% \end{function}
+%
+% \begin{function}[EXP, added = 2015-09-18]
+% {\str_count:N, \str_count:c, \str_count:n, \str_count_ignore_spaces:n}
+% \begin{syntax}
+% \cs{str_count:n} \Arg{token list}
+% \end{syntax}
+% Leaves in the input stream the number of characters in the string
+% representation of \meta{token list}, as an integer denotation. The
+% functions differ in their treatment of spaces. In the case of
+% \cs{str_count:N} and \cs{str_count:n}, all characters including
+% spaces are counted. The \cs{str_count_ignore_spaces:n} function
+% leaves the number of non-space characters in the input stream.
+% \end{function}
+%
+% \begin{function}[EXP, added = 2015-09-18]
+% {\str_count_spaces:N, \str_count_spaces:c, \str_count_spaces:n}
+% \begin{syntax}
+% \cs{str_count_spaces:n} \Arg{token list}
+% \end{syntax}
+% Leaves in the input stream the number of space characters in the
+% string representation of \meta{token list}, as an integer
+% denotation. Of course, this function has no \texttt{_ignore_spaces}
+% variant.
+% \end{function}
+%
+% \begin{function}[EXP, added = 2015-09-18]
+% {\str_head:N, \str_head:c, \str_head:n, \str_head_ignore_spaces:n}
+% \begin{syntax}
+% \cs{str_head:n} \Arg{token list}
+% \end{syntax}
+% Converts the \meta{token list} into a \meta{string}. The first
+% character in the \meta{string} is then left in the input stream,
+% with category code \enquote{other}. The functions differ if the
+% first character is a space: \cs{str_head:N} and \cs{str_head:n}
+% return a space token with category code~$10$ (blank space), while
+% the \cs{str_head_ignore_spaces:n} function ignores this space
+% character and leaves the first non-space character in the input
+% stream. If the \meta{string} is empty (or only contains spaces in
+% the case of the \texttt{_ignore_spaces} function), then nothing is
+% left on the input stream.
+% \end{function}
+%
+% \begin{function}[EXP, added = 2015-09-18]
+% {\str_tail:N, \str_tail:c, \str_tail:n, \str_tail_ignore_spaces:n}
+% \begin{syntax}
+% \cs{str_tail:n} \Arg{token list}
+% \end{syntax}
+% Converts the \meta{token list} to a \meta{string}, removes the first
+% character, and leaves the remaining characters (if any) in the input
+% stream, with category codes $12$ and $10$ (for spaces). The
+% functions differ in the case where the first character is a space:
+% \cs{str_tail:N} and \cs{str_tail:n} will trim only that space, while
+% \cs{str_tail_ignore_spaces:n} removes the first non-space character
+% and any space before it. If the \meta{token list} is empty (or
+% blank in the case of the \texttt{_ignore_spaces} variant), then
+% nothing is left on the input stream.
+% \end{function}
+%
+% \begin{function}[EXP, added = 2015-09-18]
+% {\str_item:Nn, \str_item:nn, \str_item_ignore_spaces:nn}
+% \begin{syntax}
+% \cs{str_item:nn} \Arg{token list} \Arg{integer expression}
+% \end{syntax}
+% Converts the \meta{token list} to a \meta{string}, and leaves in the
+% input stream the character in position \meta{integer expression} of
+% the \meta{string}, starting at $1$ for the first (left-most)
+% character. In the case of \cs{str_item:Nn} and \cs{str_item:nn},
+% all characters including spaces are taken into account. The
+% \cs{str_item_ignore_spaces:nn} function skips spaces when counting
+% characters. If the \meta{integer expression} is negative,
+% characters are counted from the end of the \meta{string}. Hence,
+% $-1$ is the right-most character, \emph{etc.}
+% \end{function}
+%
+% \begin{function}[EXP, added = 2015-09-18]
+% {
+% \str_range:Nnn, \str_range:cnn, \str_range:nnn,
+% \str_range_ignore_spaces:nnn
+% }
+% \begin{syntax}
+% \cs{str_range:nnn} \Arg{token list} \Arg{start index} \Arg{end index}
+% \end{syntax}
+% Converts the \meta{token list} to a \meta{string}, and leaves in the
+% input stream the characters from the \meta{start index} to the
+% \meta{end index} inclusive. Positive \meta{indices} are counted
+% from the start of the string, $1$~being the first character, and
+% negative \meta{indices} are counted from the end of the string,
+% $-1$~being the last character. If either of \meta{start index} or
+% \meta{end index} is~$0$, the result is empty. For instance,
+% \begin{verbatim}
+% \iow_term:x { \str_range:nnn { abcdef } { 2 } { 5 } }
+% \iow_term:x { \str_range:nnn { abcdef } { -4 } { -1 } }
+% \iow_term:x { \str_range:nnn { abcdef } { -2 } { -1 } }
+% \iow_term:x { \str_range:nnn { abcdef } { 0 } { -1 } }
+% \end{verbatim}
+% will print \texttt{bcd}, \texttt{cdef}, \texttt{ef}, and an empty
+% line to the terminal.
+% \end{function}
+%
% \section{String manipulation}
%
% \begin{function}[rEXP, added = 2015-03-01]
@@ -294,6 +534,53 @@
% \end{texnote}
% \end{function}
%
+% \section{Viewing strings}
+%
+% \begin{function}[added = 2015-09-18]
+% {\str_show:N, \str_show:c, \str_show:n}
+% \begin{syntax}
+% \cs{str_show:N} \meta{str~var}
+% \end{syntax}
+% Displays the content of the \meta{str~var} on the terminal.
+% \end{function}
+%
+% \section{Constant token lists}
+%
+% \begin{variable}[added = 2016-09-19]
+% {
+% \c_ampersand_str,
+% \c_atsign_str,
+% \c_backslash_str,
+% \c_left_brace_str,
+% \c_right_brace_str,
+% \c_circumflex_str,
+% \c_colon_str,
+% \c_dollar_str,
+% \c_hash_str,
+% \c_percent_str,
+% \c_tilde_str,
+% \c_underscore_str
+% }
+% Constant strings, containing a single character token, with category
+% code $12$.
+% \end{variable}
+%
+% \section{Scratch strings}
+%
+% \begin{variable}{\l_tmpa_str, \l_tmpb_str}
+% Scratch strings for local assignment. These are never used by
+% the kernel code, and so are safe for use with any \LaTeX3-defined
+% function. However, they may be overwritten by other non-kernel
+% code and so should only be used for short-term storage.
+% \end{variable}
+%
+% \begin{variable}{\g_tmpa_str, \g_tmpb_str}
+% Scratch strings for global assignment. These are never used by
+% the kernel code, and so are safe for use with any \LaTeX3-defined
+% function. However, they may be overwritten by other non-kernel
+% code and so should only be used for short-term storage.
+% \end{variable}
+%
% \subsection{Internal string functions}
%
% \begin{function}[EXP]{\__str_if_eq_x:nn}
@@ -317,6 +604,38 @@
% of \cs{str_if_eq_x:nn(TF)} coded for speed.
% \end{function}
%
+% \begin{function}[EXP]{\__str_to_other:n}
+% \begin{syntax}
+% \cs{__str_to_other:n} \Arg{token list}
+% \end{syntax}
+% Converts the \meta{token list} to a \meta{other string}, where
+% spaces have category code \enquote{other}. This function can be
+% \texttt{f}-expanded without fear of losing a leading space, since
+% spaces do not have category code $10$ in its result. It takes a
+% time quadratic in the character count of the string.
+% \end{function}
+%
+% \begin{function}[EXP]{\__str_count:n}
+% \begin{syntax}
+% \cs{__str_count:n} \Arg{other string}
+% \end{syntax}
+% This function expects an argument that is entirely made of
+% characters with category \enquote{other}, as produced by
+% \cs{__str_to_other:n}. It leaves in the input stream the number of
+% character tokens in the \meta{other string}, faster than the
+% analogous \cs{str_count:n} function.
+% \end{function}
+%
+% \begin{function}[EXP]{\__str_range:nnn}
+% \begin{syntax}
+% \cs{__str_range:nnn} \Arg{other string} \Arg{start index} \Arg{end index}
+% \end{syntax}
+% Identical to \cs{str_range:nnn} except that the first argument is
+% expected to be entirely made of characters with category
+% \enquote{other}, as produced by \cs{__str_to_other:n}, and the
+% result is also an \meta{other string}.
+% \end{function}
+%
% \end{documentation}
%
% \begin{implementation}
@@ -331,54 +650,113 @@
%<@@=str>
% \end{macrocode}
%
-% \begin{macro}{\str_head:n, \str_tail:n}
-% \begin{macro}[aux]{\__str_head:w}
-% \begin{macro}[aux]{\__str_tail:w}
-% After \cs{tl_to_str:n}, we have a list of character tokens,
-% all with category code 12, except the space, which has category
-% code 10. Directly using \cs{tl_head:w} would thus lose leading spaces.
-% Instead, we take an argument delimited by an explicit space, and
-% then only use \cs{tl_head:w}. If the string started with a
-% space, then the argument of \cs{__str_head:w} is empty, and
-% the function correctly returns a space character. Otherwise,
-% it returns the first token of |#1|, which is the first token
-% of the string. If the string is empty, we return an empty result.
-%
-% To remove the first character of \cs{tl_to_str:n} |{#1}|,
-% we test it using \cs{if_charcode:w} \cs{scan_stop:},
-% always \texttt{false} for characters. If the argument was non-empty,
-% then \cs{__str_tail:w} returns everything until the first
-% \texttt{X} (with category code letter, no risk of confusing
-% with the user input). If the argument was empty, the first
-% \texttt{X} is taken by \cs{if_charcode:w}, and nothing
-% is returned. We use \texttt{X} as a \meta{marker}, rather than
-% a quark because the test \cs{if_charcode:w} \cs{scan_stop:}
-% \meta{marker} has to be \texttt{false}.
+% \subsection{Creating and setting string variables}
+%
+% \begin{macro}
+% {
+% \str_new:N, \str_new:c,
+% \str_use:N, \str_use:c,
+% \str_clear:N, \str_clear:c,
+% \str_gclear:N,\str_gclear:c,
+% \str_clear_new:N, \str_clear_new:c,
+% \str_gclear_new:N, \str_gclear_new:c
+% }
+% \begin{macro}
+% {
+% \str_set_eq:NN, \str_set_eq:cN, \str_set_eq:Nc, \str_set_eq:cc,
+% \str_gset_eq:NN, \str_gset_eq:cN, \str_gset_eq:Nc, \str_gset_eq:cc
+% }
+% A string is simply a token list. The full mapping system isn't set up
+% yet so do things by hand.
% \begin{macrocode}
-\cs_new:Npn \str_head:n #1
- {
- \exp_after:wN \__str_head:w
- \tl_to_str:n {#1}
- { { } } ~ \q_stop
- }
-\cs_new:Npn \__str_head:w #1 ~ %
- { \tl_head:w #1 { ~ } }
-\cs_new:Npn \str_tail:n #1
- {
- \exp_after:wN \__str_tail:w
- \reverse_if:N \if_charcode:w
- \scan_stop: \tl_to_str:n {#1} X X \q_stop
- }
-\cs_new:Npn \__str_tail:w #1 X #2 \q_stop { \fi: #1 }
+\group_begin:
+ \cs_set_protected:Npn \@@_tmp:n #1
+ {
+ \tl_if_blank:nF {#1}
+ {
+ \cs_new_eq:cc { str_ #1 :N } { tl_ #1 :N }
+ \exp_args:Nc \cs_generate_variant:Nn { str_ #1 :N } { c }
+ \@@_tmp:n
+ }
+ }
+ \@@_tmp:n
+ { new }
+ { use }
+ { clear }
+ { gclear }
+ { clear_new }
+ { gclear_new }
+ { }
+\group_end:
+\cs_new_eq:NN \str_set_eq:NN \tl_set_eq:NN
+\cs_new_eq:NN \str_gset_eq:NN \tl_gset_eq:NN
+\cs_generate_variant:Nn \str_set_eq:NN { c , Nc , cc }
+\cs_generate_variant:Nn \str_gset_eq:NN { c , Nc , cc }
% \end{macrocode}
% \end{macro}
% \end{macro}
+%
+% \begin{macro}
+% {
+% \str_set:Nn, \str_set:Nx,
+% \str_set:cn, \str_set:cx,
+% \str_gset:Nn, \str_gset:Nx,
+% \str_gset:cn, \str_gset:cx,
+% \str_const:Nn, \str_const:Nx,
+% \str_const:cn, \str_const:cx,
+% \str_put_left:Nn, \str_put_left:Nx,
+% \str_put_left:cn, \str_put_left:cx,
+% \str_gput_left:Nn, \str_gput_left:Nx,
+% \str_gput_left:cn, \str_gput_left:cx,
+% \str_put_right:Nn, \str_put_right:Nx,
+% \str_put_right:cn, \str_put_right:cx,
+% \str_gput_right:Nn, \str_gput_right:Nx,
+% \str_gput_right:cn, \str_gput_right:cx,
+% }
+% Simply convert the token list inputs to \meta{strings}.
+% \begin{macrocode}
+\group_begin:
+ \cs_set_protected:Npn \@@_tmp:n #1
+ {
+ \tl_if_blank:nF {#1}
+ {
+ \cs_new_protected:cpx { str_ #1 :Nn } ##1##2
+ { \exp_not:c { tl_ #1 :Nx } ##1 { \exp_not:N \tl_to_str:n {##2} } }
+ \exp_args:Nc \cs_generate_variant:Nn { str_ #1 :Nn } { Nx , cn , cx }
+ \@@_tmp:n
+ }
+ }
+ \@@_tmp:n
+ { set }
+ { gset }
+ { const }
+ { put_left }
+ { gput_left }
+ { put_right }
+ { gput_right }
+ { }
+\group_end:
+% \end{macrocode}
% \end{macro}
%
% \subsection{String comparisons}
%
-% \begin{macro}[int, EXP]{\__str_if_eq_x:nn}
-% \begin{macro}[aux, EXP]{\__str_escape_x:n}
+% \begin{macro}[pTF, EXP]
+% {
+% \str_if_empty:N, \str_if_empty:c,
+% \str_if_exist:N, \str_if_exist:c
+% }
+% More copy-paste!
+% \begin{macrocode}
+\prg_new_eq_conditional:NNn \str_if_exist:N \tl_if_exist:N { p , T , F , TF }
+\prg_new_eq_conditional:NNn \str_if_exist:c \tl_if_exist:c { p , T , F , TF }
+\prg_new_eq_conditional:NNn \str_if_empty:N \tl_if_empty:N { p , T , F , TF }
+\prg_new_eq_conditional:NNn \str_if_empty:c \tl_if_empty:c { p , T , F , TF }
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}[int, EXP]{\@@_if_eq_x:nn}
+% \begin{macro}[aux, EXP]{\@@_escape_x:n}
% String comparisons rely on the primitive \cs{(pdf)strcmp} if available:
% \LuaTeX{} does not have it, so emulation is required. As the net result
% is that we do not \emph{always} use the primitive, the correct approach
@@ -392,21 +770,21 @@
% otherwise will fail as \cs{luatex_luaescapestring:D} does not double
% such tokens.
% \begin{macrocode}
-\cs_new:Npn \__str_if_eq_x:nn #1#2 { \pdftex_strcmp:D {#1} {#2} }
+\cs_new:Npn \@@_if_eq_x:nn #1#2 { \pdftex_strcmp:D {#1} {#2} }
\cs_if_exist:NT \luatex_luatexversion:D
{
- \cs_set:Npn \__str_if_eq_x:nn #1#2
+ \cs_set:Npn \@@_if_eq_x:nn #1#2
{
\luatex_directlua:D
{
l3kernel.strcmp
(
- " \__str_escape_x:n {#1} " ,
- " \__str_escape_x:n {#2} "
+ " \@@_escape_x:n {#1} " ,
+ " \@@_escape_x:n {#2} "
)
}
}
- \cs_new:Npn \__str_escape_x:n #1
+ \cs_new:Npn \@@_escape_x:n #1
{
\luatex_luaescapestring:D
{
@@ -418,16 +796,16 @@
% \end{macro}
% \end{macro}
%
-% \begin{macro}[int, EXP]{\__str_if_eq_x_return:nn}
+% \begin{macro}[int, EXP]{\@@_if_eq_x_return:nn}
% It turns out that we often need to compare a token list
% with the result of applying some function to it, and
% return with \cs{prg_return_true/false:}. This test is
% similar to \cs{str_if_eq:nnTF} (see \pkg{l3str}),
% but is hard-coded for speed.
% \begin{macrocode}
-\cs_new:Npn \__str_if_eq_x_return:nn #1 #2
+\cs_new:Npn \@@_if_eq_x_return:nn #1 #2
{
- \if_int_compare:w \__str_if_eq_x:nn {#1} {#2} = \c_zero
+ \if_int_compare:w \@@_if_eq_x:nn {#1} {#2} = \c_zero
\prg_return_true:
\else:
\prg_return_false:
@@ -450,7 +828,7 @@
\prg_new_conditional:Npnn \str_if_eq:nn #1#2 { p , T , F , TF }
{
\if_int_compare:w
- \__str_if_eq_x:nn { \exp_not:n {#1} } { \exp_not:n {#2} }
+ \@@_if_eq_x:nn { \exp_not:n {#1} } { \exp_not:n {#2} }
= \c_zero
\prg_return_true: \else: \prg_return_false: \fi:
}
@@ -464,13 +842,27 @@
\cs_generate_variant:Nn \str_if_eq:nnTF { nV , no , VV }
\prg_new_conditional:Npnn \str_if_eq_x:nn #1#2 { p , T , F , TF }
{
- \if_int_compare:w \__str_if_eq_x:nn {#1} {#2} = \c_zero
+ \if_int_compare:w \@@_if_eq_x:nn {#1} {#2} = \c_zero
\prg_return_true: \else: \prg_return_false: \fi:
}
% \end{macrocode}
% \end{macro}
%
-% \begin{macro}[int, EXP]{\@@_if_eq_x_return:nn}
+% \begin{macro}[EXP, pTF]
+% {\str_if_eq:NN, \str_if_eq:Nc, \str_if_eq:cN, \str_if_eq:cc}
+% Note that \cs{str_if_eq:NN} is different from
+% \cs{tl_if_eq:NN} because it needs to ignore category codes.
+% \begin{macrocode}
+\prg_new_conditional:Npnn \str_if_eq:NN #1#2 { p , TF , T , F }
+ {
+ \if_int_compare:w \@@_if_eq_x:nn { \tl_to_str:N #1 } { \tl_to_str:N #2 }
+ = \c_zero \prg_return_true: \else: \prg_return_false: \fi:
+ }
+\cs_generate_variant:Nn \str_if_eq:NNT { c , Nc , cc }
+\cs_generate_variant:Nn \str_if_eq:NNF { c , Nc , cc }
+\cs_generate_variant:Nn \str_if_eq:NNTF { c , Nc , cc }
+\cs_generate_variant:Nn \str_if_eq_p:NN { c , Nc , cc }
+% \end{macrocode}
% \end{macro}
%
% \begin{macro}[EXP]
@@ -550,6 +942,444 @@
% \end{macro}
% \end{macro}
%
+% \subsection{Accessing specific characters in a string}
+%
+% \begin{macro}[EXP, int]{\@@_to_other:n}
+% \begin{macro}[EXP, aux]{\@@_to_other_loop:w, \@@_to_other_end:w}
+% First apply \cs{tl_to_str:n}, then replace all spaces by
+% \enquote{other} spaces, $8$ at a time, storing the converted part of
+% the string between the \cs{q_mark} and \cs{q_stop} markers. The end
+% is detected when \cs{@@_to_other_loop:w} finds one of the trailing
+% |A|, distinguished from any contents of the initial token list by
+% their category. Then \cs{@@_to_other_end:w} is called, and finds
+% the result between \cs{q_mark} and the first |A| (well, there is
+% also the need to remove a space).
+% \begin{macrocode}
+\cs_new:Npn \@@_to_other:n #1
+ {
+ \exp_after:wN \@@_to_other_loop:w
+ \tl_to_str:n {#1} ~ A ~ A ~ A ~ A ~ A ~ A ~ A ~ A ~ \q_mark \q_stop
+ }
+\group_begin:
+\tex_lccode:D `\* = `\ %
+\tex_lccode:D `\A = `\A
+\tex_lowercase:D
+ {
+ \group_end:
+ \cs_new:Npn \@@_to_other_loop:w
+ #1 ~ #2 ~ #3 ~ #4 ~ #5 ~ #6 ~ #7 ~ #8 ~ #9 \q_stop
+ {
+ \if_meaning:w A #8
+ \@@_to_other_end:w
+ \fi:
+ \@@_to_other_loop:w
+ #9 #1 * #2 * #3 * #4 * #5 * #6 * #7 * #8 * \q_stop
+ }
+ \cs_new:Npn \@@_to_other_end:w \fi: #1 \q_mark #2 * A #3 \q_stop
+ { \fi: #2 }
+ }
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}[EXP]
+% {\str_item:Nn, \str_item:cn, \str_item:nn, \str_item_ignore_spaces:nn}
+% \begin{macro}[EXP, aux]{\@@_item:nn, \@@_item:w}
+% The \cs{str_item:nn} hands its argument with spaces escaped to
+% \cs{@@_item:nn}, and makes sure to turn the result back into
+% a proper string (with category code~$10$ spaces) eventually. The
+% \cs{str_item_ignore_spaces:nn} function does not escape spaces,
+% which are thus ignored by \cs{@@_item:nn} since
+% everything else is done with undelimited arguments.
+% Evaluate the \meta{index} argument~|#2| and count characters in
+% the string, passing those two numbers to \cs{@@_item:w} for
+% further analysis. If the \meta{index} is negative, shift it by
+% the \meta{count} to know the how many character to discard, and if
+% that is still negative give an empty result. If the \meta{index}
+% is larger than the \meta{count}, give an empty result, and
+% otherwise discard $\meta{index}-1$ characters before returning the
+% following one. The shift by $-1$ is obtained by inserting an empty
+% brace group before the string in that case: that brace group also
+% covers the case where the \meta{index} is zero.
+% \begin{macrocode}
+\cs_new_nopar:Npn \str_item:Nn { \exp_args:No \str_item:nn }
+\cs_generate_variant:Nn \str_item:Nn { c }
+\cs_new:Npn \str_item:nn #1#2
+ {
+ \exp_args:Nf \tl_to_str:n
+ {
+ \exp_args:Nf \@@_item:nn
+ { \@@_to_other:n {#1} } {#2}
+ }
+ }
+\cs_new:Npn \str_item_ignore_spaces:nn #1
+ { \exp_args:No \@@_item:nn { \tl_to_str:n {#1} } }
+\cs_new:Npn \@@_item:nn #1#2
+ {
+ \exp_after:wN \@@_item:w
+ \int_use:N \__int_eval:w #2 \exp_after:wN ;
+ \__int_value:w \@@_count:n {#1} ;
+ #1 \q_stop
+ }
+\cs_new:Npn \@@_item:w #1; #2;
+ {
+ \int_compare:nNnTF {#1} < \c_zero
+ {
+ \int_compare:nNnTF {#1} < {-#2}
+ { \use_none_delimit_by_q_stop:w }
+ {
+ \exp_after:wN \use_i_delimit_by_q_stop:nw
+ \exp:w \exp_after:wN \@@_skip_exp_end:w
+ \int_use:N \__int_eval:w #1 + #2 ;
+ }
+ }
+ {
+ \int_compare:nNnTF {#1} > {#2}
+ { \use_none_delimit_by_q_stop:w }
+ {
+ \exp_after:wN \use_i_delimit_by_q_stop:nw
+ \exp:w \@@_skip_exp_end:w #1 ; { }
+ }
+ }
+ }
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}[EXP, aux]{\@@_skip_exp_end:w}
+% \begin{macro}[EXP, aux]
+% {\@@_skip_loop:wNNNNNNNN, \@@_skip_end:w, \@@_skip_end:NNNNNNNN}
+% Removes |max(#1,0)| characters from the input stream, and then
+% leaves \cs{exp_end:}. This should be expanded using
+% \cs{exp:w}. We remove characters $8$ at a time until
+% there are at most $8$ to remove. Then we do a dirty trick: the
+% \cs{if_case:w} construction leaves between $0$ and $8$ times the
+% \cs{or:} control sequence, and those \cs{or:} become arguments of
+% \cs{@@_skip_end:NNNNNNNN}. If the number of characters to remove
+% is $6$, say, then there are two \cs{or:} left, and the $8$ arguments
+% of \cs{@@_skip_end:NNNNNNNN} are the two \cs{or:}, and $6$
+% characters from the input stream, exactly what we wanted to
+% remove. Then close the \cs{if_case:w} conditional with \cs{fi:}, and
+% stop the initial expansion with \cs{exp_end:} (see places where
+% \cs{@@_skip_exp_end:w} is called).
+% \begin{macrocode}
+\cs_new:Npn \@@_skip_exp_end:w #1;
+ {
+ \if_int_compare:w #1 > \c_eight
+ \exp_after:wN \@@_skip_loop:wNNNNNNNN
+ \else:
+ \exp_after:wN \@@_skip_end:w
+ \int_use:N \__int_eval:w
+ \fi:
+ #1 ;
+ }
+\cs_new:Npn \@@_skip_loop:wNNNNNNNN #1; #2#3#4#5#6#7#8#9
+ { \exp_after:wN \@@_skip_exp_end:w \int_use:N \__int_eval:w #1 - \c_eight ; }
+\cs_new:Npn \@@_skip_end:w #1 ;
+ {
+ \exp_after:wN \@@_skip_end:NNNNNNNN
+ \if_case:w #1 \exp_stop_f: \or: \or: \or: \or: \or: \or: \or: \or:
+ }
+\cs_new:Npn \@@_skip_end:NNNNNNNN #1#2#3#4#5#6#7#8 { \fi: \exp_end: }
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}[EXP]
+% {\str_range:Nnn, \str_range:nnn, \str_range_ignore_spaces:nnn}
+% \begin{macro}[EXP, int]{\@@_range:nnn}
+% \begin{macro}[EXP, aux]{\@@_range:w, \@@_range:nnw}
+% Sanitize the string. Then evaluate the arguments. At this stage we
+% also decrement the \meta{start index}, since our goal is to know how
+% many characters should be removed. Then limit the range to be
+% non-negative and at most the length of the string (this avoids
+% needing to check for the end of the string when grabbing
+% characters), shifting negative numbers by the appropriate amount.
+% Afterwards, skip characters, then keep some more, and finally drop
+% the end of the string.
+% \begin{macrocode}
+\cs_new_nopar:Npn \str_range:Nnn { \exp_args:No \str_range:nnn }
+\cs_generate_variant:Nn \str_range:Nnn { c }
+\cs_new:Npn \str_range:nnn #1#2#3
+ {
+ \exp_args:Nf \tl_to_str:n
+ {
+ \exp_args:Nf \@@_range:nnn
+ { \@@_to_other:n {#1} } {#2} {#3}
+ }
+ }
+\cs_new:Npn \str_range_ignore_spaces:nnn #1
+ { \exp_args:No \@@_range:nnn { \tl_to_str:n {#1} } }
+\cs_new:Npn \@@_range:nnn #1#2#3
+ {
+ \exp_after:wN \@@_range:w
+ \__int_value:w \@@_count:n {#1} \exp_after:wN ;
+ \int_use:N \__int_eval:w #2 - \c_one \exp_after:wN ;
+ \int_use:N \__int_eval:w #3 ;
+ #1 \q_stop
+ }
+\cs_new:Npn \@@_range:w #1; #2; #3;
+ {
+ \exp_args:Nf \@@_range:nnw
+ { \@@_range_normalize:nn {#2} {#1} }
+ { \@@_range_normalize:nn {#3} {#1} }
+ }
+\cs_new:Npn \@@_range:nnw #1#2
+ {
+ \exp_after:wN \@@_collect_delimit_by_q_stop:w
+ \int_use:N \__int_eval:w #2 - #1 \exp_after:wN ;
+ \exp:w \@@_skip_exp_end:w #1 ;
+ }
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+% \begin{macro}[EXP, aux]{\@@_range_normalize:nn}
+% This function converts an \meta{index} argument into an explicit
+% position in the string (a result of $0$ denoting \enquote{out of
+% bounds}). Expects two explicit integer arguments: the
+% \meta{index} |#1| and the string count~|#2|. If |#1| is negative,
+% replace it by $|#1| + |#2| + 1$, then limit to the range $[0,
+% |#2|]$.
+% \begin{macrocode}
+\cs_new:Npn \@@_range_normalize:nn #1#2
+ {
+ \int_eval:n
+ {
+ \if_int_compare:w #1 < \c_zero
+ \if_int_compare:w #1 < -#2 \exp_stop_f:
+ \c_zero
+ \else:
+ #1 + #2 + \c_one
+ \fi:
+ \else:
+ \if_int_compare:w #1 < #2 \exp_stop_f:
+ #1
+ \else:
+ #2
+ \fi:
+ \fi:
+ }
+ }
+% \end{macrocode}
+% \end{macro}
+% \begin{macro}[EXP, aux]{\@@_collect_delimit_by_q_stop:w}
+% \begin{macro}[EXP, aux]
+% {
+% \@@_collect_loop:wn, \@@_collect_loop:wnNNNNNNN,
+% \@@_collect_end:wn, \@@_collect_end:nnnnnnnnw
+% }
+% Collects |max(#1,0)| characters, and removes everything else until
+% \cs{q_stop}. This is somewhat similar to \cs{@@_skip_exp_end:w}, but
+% accepts integer expression arguments. This time we can only grab
+% $7$ characters at a time. At the end, we use an \cs{if_case:w}
+% trick again, so that the $8$ first arguments of
+% \cs{@@_collect_end:nnnnnnnnw} are some \cs{or:}, followed by an
+% \cs{fi:}, followed by |#1| characters from the input stream. Simply
+% leaving this in the input stream will close the conditional properly
+% and the \cs{or:} disappear.
+% \begin{macrocode}
+\cs_new:Npn \@@_collect_delimit_by_q_stop:w #1;
+ { \@@_collect_loop:wn #1 ; { } }
+\cs_new:Npn \@@_collect_loop:wn #1 ;
+ {
+ \if_int_compare:w #1 > \c_seven
+ \exp_after:wN \@@_collect_loop:wnNNNNNNN
+ \else:
+ \exp_after:wN \@@_collect_end:wn
+ \fi:
+ #1 ;
+ }
+\cs_new:Npn \@@_collect_loop:wnNNNNNNN #1; #2 #3#4#5#6#7#8#9
+ {
+ \exp_after:wN \@@_collect_loop:wn
+ \int_use:N \__int_eval:w #1 - \c_seven ;
+ { #2 #3#4#5#6#7#8#9 }
+ }
+\cs_new:Npn \@@_collect_end:wn #1 ;
+ {
+ \exp_after:wN \@@_collect_end:nnnnnnnnw
+ \if_case:w \if_int_compare:w #1 > \c_zero #1 \else: 0 \fi: \exp_stop_f:
+ \or: \or: \or: \or: \or: \or: \fi:
+ }
+\cs_new:Npn \@@_collect_end:nnnnnnnnw #1#2#3#4#5#6#7#8 #9 \q_stop
+ { #1#2#3#4#5#6#7#8 }
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \subsection{Counting characters}
+%
+% \begin{macro}[EXP]
+% {\str_count_spaces:N, \str_count_spaces:c, \str_count_spaces:n}
+% \begin{macro}[EXP, aux]{\@@_count_spaces_loop:w}
+% To speed up this function, we grab and discard $9$ space-delimited
+% arguments in each iteration of the loop. The loop stops when the
+% last argument is one of the trailing |X|\meta{number}, and that
+% \meta{number} is added to the sum of $9$ that precedes, to adjust
+% the result.
+% \begin{macrocode}
+\cs_new_nopar:Npn \str_count_spaces:N
+ { \exp_args:No \str_count_spaces:n }
+\cs_generate_variant:Nn \str_count_spaces:N { c }
+\cs_new:Npn \str_count_spaces:n #1
+ {
+ \int_eval:n
+ {
+ \exp_after:wN \@@_count_spaces_loop:w
+ \tl_to_str:n {#1} ~
+ X 7 ~ X 6 ~ X 5 ~ X 4 ~ X 3 ~ X 2 ~ X 1 ~ X 0 ~ X -1 ~
+ \q_stop
+ }
+ }
+\cs_new:Npn \@@_count_spaces_loop:w #1~#2~#3~#4~#5~#6~#7~#8~#9~
+ {
+ \if_meaning:w X #9
+ \use_i_delimit_by_q_stop:nw
+ \fi:
+ \c_nine + \@@_count_spaces_loop:w
+ }
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}[EXP]
+% {\str_count:N, \str_count:c, \str_count:n, \str_count_ignore_spaces:n}
+% \begin{macro}[EXP, int]{\@@_count:n}
+% \begin{macro}[EXP, aux]{\@@_count_aux:n, \@@_count_loop:NNNNNNNNN}
+% To count characters in a string we could first escape all spaces
+% using \cs{@@_to_other:n}, then pass the result to \cs{tl_count:n}.
+% However, the escaping step would be quadratic in the number of
+% characters in the string, and we can do better. Namely, sum the
+% number of spaces (\cs{str_count_spaces:n}) and the result of
+% \cs{tl_count:n}, which ignores spaces. Since strings tend to be
+% longer than token lists, we use specialized functions to count
+% characters ignoring spaces. Namely, loop, grabbing $9$ non-space
+% characters at each step, and end as soon as we reach one of the $9$
+% trailing items. The internal function \cs{@@_count:n}, used in
+% \cs{str_item:nn} and \cs{str_range:nnn}, is similar to
+% \cs{str_count_ignore_spaces:n} but expects its argument to already
+% be a string or a string with spaces escaped.
+% \begin{macrocode}
+\cs_new_nopar:Npn \str_count:N { \exp_args:No \str_count:n }
+\cs_generate_variant:Nn \str_count:N { c }
+\cs_new:Npn \str_count:n #1
+ {
+ \@@_count_aux:n
+ {
+ \str_count_spaces:n {#1}
+ + \exp_after:wN \@@_count_loop:NNNNNNNNN \tl_to_str:n {#1}
+ }
+ }
+\cs_new:Npn \@@_count:n #1
+ {
+ \@@_count_aux:n
+ { \@@_count_loop:NNNNNNNNN #1 }
+ }
+\cs_new:Npn \str_count_ignore_spaces:n #1
+ {
+ \@@_count_aux:n
+ { \exp_after:wN \@@_count_loop:NNNNNNNNN \tl_to_str:n {#1} }
+ }
+\cs_new:Npn \@@_count_aux:n #1
+ {
+ \int_eval:n
+ {
+ #1
+ { X \c_eight } { X \c_seven } { X \c_six }
+ { X \c_five } { X \c_four } { X \c_three }
+ { X \c_two } { X \c_one } { X \c_zero }
+ \q_stop
+ }
+ }
+\cs_new:Npn \@@_count_loop:NNNNNNNNN #1#2#3#4#5#6#7#8#9
+ {
+ \if_meaning:w X #9
+ \exp_after:wN \use_none_delimit_by_q_stop:w
+ \fi:
+ \c_nine + \@@_count_loop:NNNNNNNNN
+ }
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \end{macro}
+%
+% \subsection{The first character in a string}
+%
+% \begin{macro}[EXP]
+% {\str_head:N, \str_head:c, \str_head:n, \str_head_ignore_spaces:n}
+% \begin{macro}[EXP, aux]{\@@_head:w}
+% The \texttt{_ignore_spaces} variant applies \cs{tl_to_str:n} then
+% grabs the first item, thus skipping spaces.
+% As usual, \cs{str_head:N} expands its argument and
+% hands it to \cs{str_head:n}. To circumvent the fact that \TeX{}
+% skips spaces when grabbing undelimited macro parameters,
+% \cs{@@_head:w} takes an argument delimited by a space. If |#1|
+% starts with a non-space character, \cs{use_i_delimit_by_q_stop:nw}
+% leaves that in the input stream. On the other hand, if |#1| starts
+% with a space, the \cs{@@_head:w} takes an empty argument, and the
+% single (initially braced) space in the definition of \cs{@@_head:w}
+% makes its way to the output. Finally, for an empty argument, the
+% (braced) empty brace group in the definition of \cs{str_head:n}
+% gives an empty result after passing through
+% \cs{use_i_delimit_by_q_stop:nw}.
+% \begin{macrocode}
+\cs_new_nopar:Npn \str_head:N { \exp_args:No \str_head:n }
+\cs_generate_variant:Nn \str_head:N { c }
+\cs_set:Npn \str_head:n #1
+ {
+ \exp_after:wN \@@_head:w
+ \tl_to_str:n {#1}
+ { { } } ~ \q_stop
+ }
+\cs_set:Npn \@@_head:w #1 ~ %
+ { \use_i_delimit_by_q_stop:nw #1 { ~ } }
+\cs_new:Npn \str_head_ignore_spaces:n #1
+ {
+ \exp_after:wN \use_i_delimit_by_q_stop:nw
+ \tl_to_str:n {#1} { } \q_stop
+ }
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
+% \begin{macro}[EXP]
+% {\str_tail:N, \str_tail:c, \str_tail:n, \str_tail_ignore_spaces:n}
+% \begin{macro}[EXP, aux]{\@@_tail_auxi:w, \@@_tail_auxii:w}
+% Getting the tail is a little bit more convoluted than the head of a
+% string. We hit the front of the string with \cs{reverse_if:N}
+% \cs{if_charcode:w} \cs{scan_stop:}. This removes the first
+% character, and necessarily makes the test true, since the character
+% cannot match \cs{scan_stop:}. The auxiliary function then inserts
+% the required \cs{fi:} to close the conditional, and leaves the tail
+% of the string in the input stream. The details are such that an
+% empty string has an empty tail (this requires in particular that the
+% end-marker |X| be unexpandable and not a control sequence). The
+% \texttt{_ignore_spaces} is rather simpler: after converting the
+% input to a string, \cs{@@_tail_auxii:w} removes one undelimited
+% argument and leaves everything else until an end-marker \cs{q_mark}.
+% One can check that an empty (or blank) string yields an empty
+% tail.
+% \begin{macrocode}
+\cs_new_nopar:Npn \str_tail:N { \exp_args:No \str_tail:n }
+\cs_generate_variant:Nn \str_tail:N { c }
+\cs_set:Npn \str_tail:n #1
+ {
+ \exp_after:wN \@@_tail_auxi:w
+ \reverse_if:N \if_charcode:w
+ \scan_stop: \tl_to_str:n {#1} X X \q_stop
+ }
+\cs_set:Npn \@@_tail_auxi:w #1 X #2 \q_stop { \fi: #1 }
+\cs_new:Npn \str_tail_ignore_spaces:n #1
+ {
+ \exp_after:wN \@@_tail_auxii:w
+ \tl_to_str:n {#1} \q_mark \q_mark \q_stop
+ }
+\cs_new:Npn \@@_tail_auxii:w #1 #2 \q_mark #3 \q_stop { #2 }
+% \end{macrocode}
+% \end{macro}
+% \end{macro}
+%
% \subsection{String manipulation}
%
% \begin{macro}[EXP]
@@ -558,12 +1388,12 @@
% \str_lower_case:n, \str_lower_case:f,
% \str_upper_case:n, \str_upper_case:f
% }
-% \begin{macro}[aux, EXP]{\__str_change_case:nn}
-% \begin{macro}[aux, EXP]{\__str_change_case_aux:nn}
-% \begin{macro}[aux, EXP]{\__str_change_case_loop:nw}
-% \begin{macro}[aux, EXP]{\__str_change_case_space:n}
-% \begin{macro}[aux, EXP]{\__str_change_case_char:nN}
-% \begin{macro}[aux, EXP]{\__str_change_case_char:NNNNNNNNn}
+% \begin{macro}[aux, EXP]{\@@_change_case:nn}
+% \begin{macro}[aux, EXP]{\@@_change_case_aux:nn}
+% \begin{macro}[aux, EXP]{\@@_change_case_loop:nw}
+% \begin{macro}[aux, EXP]{\@@_change_case_space:n}
+% \begin{macro}[aux, EXP]{\@@_change_case_char:nN}
+% \begin{macro}[aux, EXP]{\@@_change_case_char:NNNNNNNNn}
% Case changing for programmatic reasons is done by first detokenizing
% input then doing a simple loop that only has to worry about spaces
% and everything else. The output is detokenized to allow data sharing
@@ -571,45 +1401,45 @@
% data files based on the character code of the current character is
% shared with the text case changer too.
% \begin{macrocode}
-\cs_new:Npn \str_fold_case:n #1 { \__str_change_case:nn {#1} { fold } }
-\cs_new:Npn \str_lower_case:n #1 { \__str_change_case:nn {#1} { lower } }
-\cs_new:Npn \str_upper_case:n #1 { \__str_change_case:nn {#1} { upper } }
+\cs_new:Npn \str_fold_case:n #1 { \@@_change_case:nn {#1} { fold } }
+\cs_new:Npn \str_lower_case:n #1 { \@@_change_case:nn {#1} { lower } }
+\cs_new:Npn \str_upper_case:n #1 { \@@_change_case:nn {#1} { upper } }
\cs_generate_variant:Nn \str_fold_case:n { V }
\cs_generate_variant:Nn \str_lower_case:n { f }
\cs_generate_variant:Nn \str_upper_case:n { f }
-\cs_new:Npn \__str_change_case:nn #1
+\cs_new:Npn \@@_change_case:nn #1
{
- \exp_after:wN \__str_change_case_aux:nn \exp_after:wN
+ \exp_after:wN \@@_change_case_aux:nn \exp_after:wN
{ \tl_to_str:n {#1} }
}
-\cs_new:Npn \__str_change_case_aux:nn #1#2
+\cs_new:Npn \@@_change_case_aux:nn #1#2
{
- \__str_change_case_loop:nw {#2} #1 \q_recursion_tail \q_recursion_stop
+ \@@_change_case_loop:nw {#2} #1 \q_recursion_tail \q_recursion_stop
}
-\cs_new:Npn \__str_change_case_loop:nw #1#2 \q_recursion_stop
+\cs_new:Npn \@@_change_case_loop:nw #1#2 \q_recursion_stop
{
\tl_if_head_is_space:nTF {#2}
- { \__str_change_case_space:n }
- { \__str_change_case_char:nN }
+ { \@@_change_case_space:n }
+ { \@@_change_case_char:nN }
{#1} #2 \q_recursion_stop
}
\use:x
- { \cs_new:Npn \exp_not:N \__str_change_case_space:n ##1 \c_space_tl }
+ { \cs_new:Npn \exp_not:N \@@_change_case_space:n ##1 \c_space_tl }
{
\c_space_tl
- \__str_change_case_loop:nw {#1}
+ \@@_change_case_loop:nw {#1}
}
-\cs_new:Npn \__str_change_case_char:nN #1#2
+\cs_new:Npn \@@_change_case_char:nN #1#2
{
\quark_if_recursion_tail_stop:N #2
\exp_args:Nf \tl_to_str:n
{
- \exp_after:wN \__str_change_case_char:NNNNNNNNn
+ \exp_after:wN \@@_change_case_char:NNNNNNNNn
\int_use:N \__int_eval:w 1000000 + `#2 \__int_eval_end: #2 {#1}
}
- \__str_change_case_loop:nw {#1}
+ \@@_change_case_loop:nw {#1}
}
-\cs_new:Npn \__str_change_case_char:NNNNNNNNn #1#2#3#4#5#6#7#8#9
+\cs_new:Npn \@@_change_case_char:NNNNNNNNn #1#2#3#4#5#6#7#8#9
{
\str_case:nvF #8
{ c__unicode_ #9 _ #6 _X_ #7 _tl }
@@ -624,10 +1454,64 @@
% \end{macro}
% \end{macro}
%
+% \begin{variable}
+% {
+% \c_ampersand_str,
+% \c_atsign_str,
+% \c_backslash_str,
+% \c_left_brace_str,
+% \c_right_brace_str,
+% \c_circumflex_str,
+% \c_colon_str,
+% \c_dollar_str,
+% \c_hash_str,
+% \c_percent_str,
+% \c_tilde_str,
+% \c_underscore_str
+% }
+% For all of those strings, use \cs{cs_to_str:N} to get characters with
+% the correct category code without worries
+% \begin{macrocode}
+\str_const:Nx \c_ampersand_str { \cs_to_str:N \& }
+\str_const:Nx \c_atsign_str { \cs_to_str:N \@ }
+\str_const:Nx \c_backslash_str { \cs_to_str:N \\ }
+\str_const:Nx \c_left_brace_str { \cs_to_str:N \{ }
+\str_const:Nx \c_right_brace_str { \cs_to_str:N \} }
+\str_const:Nx \c_circumflex_str { \cs_to_str:N \^ }
+\str_const:Nx \c_colon_str { \cs_to_str:N \: }
+\str_const:Nx \c_dollar_str { \cs_to_str:N \$ }
+\str_const:Nx \c_hash_str { \cs_to_str:N \# }
+\str_const:Nx \c_percent_str { \cs_to_str:N \% }
+\str_const:Nx \c_tilde_str { \cs_to_str:N \~ }
+\str_const:Nx \c_underscore_str { \cs_to_str:N \_ }
+% \end{macrocode}
+% \end{variable}
+%
+% \begin{variable}{\l_tmpa_str, \l_tmpb_str, \g_tmpa_str, \g_tmpb_str}
+% Scratch strings.
+% \begin{macrocode}
+\str_new:N \l_tmpa_str
+\str_new:N \l_tmpb_str
+\str_new:N \g_tmpa_str
+\str_new:N \g_tmpb_str
+% \end{macrocode}
+% \end{variable}
+%
+% \subsection{Viewing strings}
+%
+% \begin{macro}{\str_show:n, \str_show:N, \str_show:c}
+% Displays a string on the terminal.
+% \begin{macrocode}
+\cs_new_eq:NN \str_show:n \tl_show:n
+\cs_new_eq:NN \str_show:N \tl_show:N
+\cs_generate_variant:Nn \str_show:N { c }
+% \end{macrocode}
+% \end{macro}
+%
% \begin{macrocode}
%</initex|package>
% \end{macrocode}
%
% \end{implementation}
%
-% \PrintIndex
+% \PrintIndex \ No newline at end of file