diff options
Diffstat (limited to 'macros/latex/required/latex-lab/latex-lab-math.dtx')
-rw-r--r-- | macros/latex/required/latex-lab/latex-lab-math.dtx | 1561 |
1 files changed, 1263 insertions, 298 deletions
diff --git a/macros/latex/required/latex-lab/latex-lab-math.dtx b/macros/latex/required/latex-lab/latex-lab-math.dtx index 285471addc..0dbc591185 100644 --- a/macros/latex/required/latex-lab/latex-lab-math.dtx +++ b/macros/latex/required/latex-lab/latex-lab-math.dtx @@ -2,7 +2,7 @@ % %% File: latex-lab-math.dtx % -% Copyright (C) 2022-2023 The LaTeX Project +% Copyright (C) 2022-2024 The LaTeX Project % % It may be distributed and/or modified under the conditions of the % LaTeX Project Public License (LPPL), either version 1.3c of this @@ -19,8 +19,8 @@ % for those people who are interested or want to report an issue. % % -\def\ltlabmathdate{2023-09-11} -\def\ltlabmathversion{0.5c} +\def\ltlabmathdate{2024-04-16} +\def\ltlabmathversion{0.5j} % %<*driver> \documentclass{l3doc} @@ -46,7 +46,14 @@ % \newcommand\NEW[1]{\marginpar{\mbox{}\hfill\fbox{New: #1}}} % \providecommand\class[1]{\texttt{#1.cls}} % \providecommand\pkg[1]{\texttt{#1}} -% \providecommand\hook[1]{\texttt{#1}} +% +% \providecommand\hook[1]{\texttt{#1\DescribeHook[noprint]{#1}}} +% \providecommand\socket[1]{\texttt{#1\DescribeSocket[noprint]{#1}}} +% \providecommand\plug[1]{\texttt{#1\DescribePlug[noprint]{#1}}} +% +% \NewDocElement[printtype=\textit{socket},idxtype=socket,idxgroup=Sockets]{Socket}{socketdecl} +% \NewDocElement[printtype=\textit{hook},idxtype=hook,idxgroup=Hooks]{Hook}{hookdecl} +% \NewDocElement[printtype=\textit{plug},idxtype=plug,idxgroup=Plugs]{Plug}{plugdecl} % % ^^A \car {...} for marginal comments % ^^A \car*{...} for longer inline comments @@ -59,8 +66,6 @@ % {\IfBooleanTF{#1}{\todo[inline,#2]{#3}}^^A % {\todo[#2]{#3}}} % -% -% % \begin{abstract} % This is an experimental prototype. It captures math material % (basically okay, but the interfaces for packages aren't yet @@ -75,6 +80,18 @@ % \car*{Todo: update all the documentation! Both here and % (what little there is!) in the implementation section.} % +% Tagging math involves a variety of tasks that require that math is captured before the +% typesetting +% \begin{itemize} +% \item When typesetting the math MC-tags and structure commands must +% be inserted at the begin and the end, and perhaps also around lines +% or other subparts of the equation. +% \item The source and/or a mathml-representation of the source must be available +% so that it can be (perhaps after some preprocessing) be used in an associated file +% or in an alternate text +% \item It must be possible to measure the math for e.g. a bbox setting. +% \end{itemize} +% % This file implements capture of all math mode material at the outer % level, i.e., a formula is captured in its entirety with inner text % blocks (possibly containing further math) absorbed as part of the @@ -84,7 +101,24 @@ %\end{verbatim} % would only result in a single capture of the tokens % ``\verb*/a \in A \text{ for all $a<5$}/''. +% % +% \section{Math capture} \label{sec:mathcapture} +% In the current setup +% \begin{itemize} +% \item |$|, |\(...\)| and |$$| grab (through a command in \cs{everymath}/cs{everydisplay}) +% if the boolean \cs{l_@@_collected_bool} is false. +% If the boolean is true they behave normally and can for example contain verbatim. +% +% \item All (registered) environments grab their body +% regardless of the state of the boolean. For +% |equation|, |equation*| and |math| this is a change as they no longer can +% contain verbatim. +% +% \item BUG: |\[...\]| grabs if \cs{l_@@_collected_bool} is false. If it is +% true it falls back to |equation*| and then errors because this can't find the end. +% \end{itemize} +% % \subsection{Code level interfaces} % % \begin{function}{\math_register_env:n, \math_register_env:nn} @@ -109,7 +143,8 @@ % \end{syntax} % Declares that the captured math content should be passed to the % \meta{tokens}, which will receive the environment type as |#1| and -% the content as |#2|. +% the content as |#2|. The processing is done before the typesetting. It is not +% applied if \cs{ifmeasuring@} is true. % \end{function} % % \subsection{Document level interfaces} @@ -129,38 +164,275 @@ % \end{itemize} % \end{function} % +% \section{Math tagging} +% +% The tagging code has to handle +% \begin{itemize} +% \item the embedding into the surrounding. This means +% \begin{itemize} +% \item closing and reopening MC-chunks +% \item closing and reopening text/P-structures +% \item handling interferences of the tagging code with penalties and spacing. +% \end{itemize} +% \item the actual tagging which means to do some or all of the following tasks: +% \begin{itemize} +% \item setup content for an associated source file +% \item setup content for an associated mathml file +% \item setup content for the /Alt key +% \item setup content for the /ActualText key +% \item setup attributes +% \item add associated files +% \item add a Formula structure +% \item surround subparts (e.g. lines) with Formula sub structures +% (perhaps with their own set of additional content) +% \item surround elements of the equation with mathml structure elements +% (currently only luatex with luamml) +% \end{itemize} +% \end{itemize} +% +% \subsection{Inline math} +% +% The embedding code is added through +% the sockets +% \begin{itemize} +% \item |tagsupport/math/inline/begin| +% \item |tagsupport/math/inline/end| +% \end{itemize} +% The sockets simply push and pop the MC currently. Without +% tagging they use the noop-plug. +% +% The actual tagging is in done through the sockets +% \begin{itemize} +% \item |tagsupport/math/inline/formula/begin| +% This socket takes the math as argument and its code +% should output it for typesetting. Without tagging the +% socket uses the identity plug. +% The |default| plug of the socket calls these three internal sockets +% for the tagging support: +% \begin{itemize} +% \item |tagsupport/math/content| This should set up the various +% content variables (empty variables are ignored by the structure code +% and so can be used to suppress a setting). +% \item |tagsupport/math/struct/begin| This calls \cs{tag_struct_begin:n}. +% It should also write the associated files if needed. +% \item |tagsupport/math/substruct/begin| this handles subparts. +% TODO: does it really make sense in inline math to have that?? +% \end{itemize} +% \item |tagsupport/math/inline/formula/end| +% This socket ends the formula structure(s). The |default| +% plug calls these internal sockets: +% \begin{itemize} +% \item |tagsupport/math/substruct/end| +% \item |tagsupport/math/struct/end| +% \end{itemize} +% \end{itemize} +% +% \subsection{Display math} +% +% \textit{to be written} +% +% \subsection{Associated Files} +% +% The current code allows the attachment of two types of associated file to the +% Formula structure: +% the \LaTeX\ source and a MathML representation. +% Technically both can be attached---AF is an array +% of file references---practically there can be problems with PDF consumers: +% e.g. ngpdf used both and so showed the equation twice +% (this has been corrected in the newest version) and +% Foxit seems to see only the first AF in the array (so we attach the +% mathml as first file). +% +% The \LaTeX\ source can be (and is) attached automatically. +% It can be suppressed by an option with +% \texttt{math/tex/AF=false}, see below. +% +% The MathML is attached if a file |\jobname-mathml.html| is found and +% if it contains a suitable MathML snippet for the current formula. +% The MathML processing can be suppressed globally by emptying the list of +% mathml files with |math/mathml/sources=|. Locally for a formula |math/mathml/AF=false| +% can be used. +% +% For a MathML representation a file with such representations must be provided. +% If the equation is numbered the numbering should be part of the MathML as +% the |Lbl| substructure is ignored if an MathML is used (see https://github.com/foxitsoftware/PDF_UA-2). +% +% The MathML representation is given in a special format. +% It is meant to be a valid html file +% that can be viewed in a browser. +% For this it can start with |<!DOCTYPE html><html>| and end with |</html>| +% It should have the extension \texttt{.html}. The \meta{mathml} content +% is read with special catcodes, so can contain ambersands, hashes, comment chars +% and unmatched braces such as |<mo>{</mo>| +% +% The file should contain a number of representations in this format: +% \begin{quote} +% |<div>| \\ +% | <h2>\mml| \meta{key}|</h2>|\\ +% | <p>|\meta{source}|</p>| \\ +% | <p>|\meta{hash}|</p>| \\ +% | <math | \meta{attributes} |>|\\ +% \meta{mathml}\\ +% | </math>|\\ +% |</div>| +% \end{quote} +% The keywords |<div>|, |<h2>\mml|, |<p>|, |<math|, |</math>| |</div>| are required as +% they are used to delimit the arguments by the \LaTeX{} code. +% +% \meta{key} and \meta{source} are only used for debugging, they help to identify +% the equation referred by this representation. The source should be used correctly escaped +% |&| and |<| so that if gives valid html! +% +% \meta{attributes} is not required either, but can e.g. contain attributes +% to improve the display in a browser: +% \begin{verbatim} +% <math alttext="\mathbf{G}" class="ltx_Math" display="inline"> +% \end{verbatim} +% It can also contain the name space declaration: |xmlns="http://www.w3.org/1998/Math/MathML"|% +% \footnote{But it is probably not needed and only blows up the PDF.} +% +% +% By default the code tries at the begin of the document +% to read a file |\jobname-mathml.html| in the |html|-format. +% The file name can be changed with |mathml/setfiles={filename1,filename2}| +% (without extension, |html| is added automatically). +% If there is a list, all files are loaded. +% If a file doesn't exist it is ignored, only an info is written to the log. +% +% Currently every MathML-snippet from a file is embedded into the PDF, +% it is not checked first if it is actually used (simply writing everything to the PDF +% is a bit easier than keeping everything in memory and also means that +% the snippets are one after the other in the PDF). +% +% As mentioned above the MathML-AF can be suppressed for the equations in a group with +% |math/mathml/AF=false|, or +% completly by setting |math/mathml/sources=| in the preamble. +% +% Files embedded in a PDF can be listed in the attachments panel of a PDF viewer. +% This is probably not so useful for lots of small files (but one could create +% collections), but as long as PDF editors or viewers don't offer +% proper support to access the AF it can help so have them there. The MathML are +% added by default, but the \LaTeX{} source not. This can be changed with +% |viewer/pane/mathsource=true| (anywhere in the document) and |viewer/pane/mathml=false| (in the +% preamble, before the external file is read). +% +% \subsection{Options} +% +% % \section{Known current bugs, etc.} % \car*{New Section, now with subsections.\\ % As indicated, these lists are probably incomplete.\\ % Some of these have been addressed in a more recent branch.} % % \subsection{Capture/grabbing problems} -% +% % \begin{enumerate} % \item Incorrect grabbing of |$|-math when there is also % explicit |$|-math within a \textit{text environment} % that is itself within the math that should all be grabbed. +% For example, +% \begin{verbatim} +% $a\begin{minipage}{1cm}$b$\end{minipage}$ +% \end{verbatim} +% would only result in the capture of the tokens +% ``\verb*/a\begin {minipage}{1cm}/''. +% This can be avoided by an additional brace group: +% \begin{verbatim} +% $a{\begin{minipage}{1cm}$b$\end{minipage}}$ +% \end{verbatim} +% % \item Similar incorrect grabbing with |$$| also. +% % \item The grabbing, for all the display environments (and |\) \]|), needs % to deal with nesting: \pkg{amsmath} contains code for this. -% \item +% +% \item The math can't contain verbatim and verbatim-like commands. This is +% nothing new for the \pkg{amsmath} environments but changes |$| and |\[\]| +% and |equation*| (see e.g. tagging-project issue \#30). +% +% \item Begin and end of the math or math environment can not be hidden in commands. +% For example \verb+>{$}l<{$}+ in a tabular would lead to errors. +% Defining |\[| to fall back to |equation*| doesn't work if |equation*| is +% a grabbing environment. +% +% \item The behaviour of |\[...\]| is faulty. See above. % \end{enumerate} % +% \subsection{Fake math} +% In a number of places in \LaTeX{} math commands (mainly |$|) is used +% only for technical reason, e.g. to access a math font, to setup a symbol +% or to use \cs{vcenter}. +% +% The code identifies such fake math mostly by making use of the \cs{m@th} command +% where two methods are used for the automatic detection: +% +% \begin{itemize} +% \item After grabbing math content the code checks if the content contains the token +% \cs{m@th} and if yes it doesn't call the processor before reinserting +% the content and perhaps adding tagging code. +% This method requires that the math can be grabbed (e.g. that the end dollar is visible) +% and that the \cs{m@th} is visible. It applies for example in \cs{@iiiparbox} where the +% code from |$\vcenter| to |\m@th$| is grabbed an put back. It does not work for +% example for |tabular| where the dollars and the \cs{m@th} token are spread around +% over three commands. |tabular| needs therefore manual intervention. +% +% A look in the list of usages (in \texttt{usage-of-m@th.md}) justifies this approach. +% All usages are either not math at all, or related to small elements that probably +% shouldn't be grabbed and processed on their own. +% +% \item \cs{m@th} is redefined so that it sets the boolean \cs{l_@@_collected_bool} +% to true. If \cs{m@th} is used inside math that has been grabbed +% this doesn't change much as the boolean is set by the grabbing anyway. For usages +% outside math the benefit is not so clear: The setting avoids that in \cs{LaTeXe} +% the epsilon is processed as math, but it also prevents that the content of the amsmath +% command \cs{boxed} is processed as math. +% It means that if one wants to reenable math processing inside some (fake) math +% one has to do it after \cs{m@th} calls. +% \end{itemize} +% +% Open problems +% \begin{enumerate} +% \item The grabbing code doesn't pass the info that it detected a \cs{m@th} token. +% This means that the tagging code has to do the same check (and doesn't do this +% in all cases yet). +% +% \item Commands are missing to locally disable the grabbing and processing, e.g. +% to handle |tabular|. +% +% \item It must be checked if setting the boolean in \cs{m@th} really makes sense +% or if commands like \cs{LaTeXe} should be handled manually. +% +% \end{enumerate} +% +% \subsection{Processor} +% +% The grabbed math is at first passed to the processor. The processor is not called +% in a measuring phase (from the amsmath \cs{ifmeasuring@}) and if the \cs{m@th} +% token is detected. +% It is not quite clear what purpose the processor has. As it is a public interface +% it can't be used for internal code. And typesetting happens later and the processor +% can't really change this. Currently it is mostly used for debugging and messages. +% If the \cs{m@th} is found the \cs{l_@@_fakemath_bool} is set, so if the code +% is changed this must be preserved. +% % \subsection{Other problems} % % \begin{enumerate} % \item -% The presence of \cs{m@th} in association with \cs{ensuremath} +% The presence of \cs{m@th} in association with \cs{ensuremath} % does not necessarily indicate fakemath. This is because % wanting mathsurround to be zero is very reasonable and common, % \emph{even when the math is genuine} (and hence needs to be collected). +% +% TODO: this claim needs some examples. +% % \item User-defined environments can create problems; but this area, of % new, copied and changed environments, has not yet been developed. % % \car*{Joseph wrote, inter alia:\\ % My thinking [regarding] \cs{RegisterMathEnvironment}\\ % - (New) Math environments should not be created-then-patched, but only -% generated by a [(future)] dedicated command (\cs{DeclareMathEnviornment}, +% generated by a [(future)] dedicated command (\cs{DeclareMathEnvironment}, % presumably)\\ % - Math environments created with \pkg{ltcmd} [commands] should not be copied, . . .\\ % - Package authors should be able to manually set up math environments with a public boolean.} @@ -172,7 +444,10 @@ % \begin{enumerate} % \item Add (some of) the math display commands that were \enquote{lifted from % plain}, e.g., \cs{displaylines} \cs{eqalign}(??). -% \item +% \item The breqn packages changes catcodes and that isn't yet covered +% by our mechanism. +% \item \cs{intertext} is not correctly taken into account by the +% code splitting multiline math into subformulas. % \end{enumerate} % % @@ -218,36 +493,823 @@ % have conditional definitions and deals with how to define \cs{[}/\cs{]} % neatly. % \begin{macrocode} -\tl_gput_right:Nn \@kernel@before@begindocument - { \RequirePackage { amsmath } } +\AddToHook{begindocument/before}{ \RequirePackage { amsmath } } % \end{macrocode} % % % \subsection{Data structures} % -% -% -% \begin{variable}{\l_@@_collected_bool} +% \begin{variable}{\l_@@_collected_bool} % Tracks whether math mode material has been collected, which happens inside % \pkg{amsmath} environments as well as those handled directly here. +% If true following math will not grab and/or process. +% See \ref{sec:mathcapture} for details. +% \begin{macrocode} +\bool_new:N \l_@@_collected_bool +% \end{macrocode} +% \end{variable} % +% \begin{variable}{\l_@@_fakemath_bool} +% Tracks whether math mode material has been identified as fake math during +% the grabbing phase, which happens currently if the +% grabbed contents contains the \cs{m@th} token. % % \begin{macrocode} -\bool_new:N \l_@@_collected_bool +\bool_new:N \l_@@_fakemath_bool % \end{macrocode} % \end{variable} -% +% +% % \car{Change first tl name below: `env' $=>$ `info'?\\ % Or do we need an extra storage tl?} % % \begin{variable}{\g_@@_grabbed_env_tl, \g_@@_grabbed_math_tl} -% \cs{g_@@_grabbed_env_tl} contains the name of the math environment (\texttt{math} in the case -% of inline math, \cs{g_@@_grabbed_math_tl} the math content. +% \cs{g_@@_grabbed_env_tl} contains the name of the math environment +% (\texttt{math} in the case of inline math, +% \cs{g_@@_grabbed_math_tl} the math content. % \begin{macrocode} \tl_new:N \g_@@_grabbed_env_tl \tl_new:N \g_@@_grabbed_math_tl % \end{macrocode} % \end{variable} +% +% \begin{variable}{\l_@@_tmpa_tl,\l_@@_tmpa_skip,\l_@@_tmpa_str} +% Temporary variables +% \begin{macrocode} +\tl_new:N \l_@@_tmpa_tl +\skip_new:N \l_@@_tmpa_skip +% \end{macrocode} +% \end{variable} +% +% \begin{variable}{\l_@@_content_alt_tl, +% \l_@@_content_actual_tl, +% \l_@@_content_AF_tl} +% Temporary variables to hold math content that should +% be used in actual or alt text and stored as AF. +% \begin{macrocode} +\tl_new:N \l_@@_content_alt_tl +\tl_new:N \l_@@_content_actual_tl +\tl_new:N \l_@@_content_AF_source_tl +\tl_new:N \l_@@_content_AF_source_tmpa_tl +\tl_new:N \l_@@_content_AF_mathml_tl +% \end{macrocode} +% \end{variable} +% +% \subsection{Tagging tools} +% The following commands implement small tagging code chunks. +% This should probably be collected and moved into tagpdf later. +% \begin{macro}{\__tag_tool_close_P:} +% This closes a P/text-chunk, both the MC and the structure and +% increases the counter manually. +% \begin{macrocode} +\cs_new_protected:Npn \__tag_tool_close_P: + { + \tag_if_active:T + { + \tag_mc_end: %end P-chunk, should perhaps be \tag_mc_end_push: ... + \__tag_gincr_para_end_int: + \__tag_check_para_end_show:nn{red}{} %debug: show para + \tag_struct_end: + } + } +% \end{macrocode} +% \end{macro} +% +% We add also an attribute. +% \begin{macrocode} +\tl_new:N\l_@@_attribute_class_tl +\tagpdfsetup + {role/new-attribute = {inline} {/O /Layout /Placement/Inline}, + role/new-attribute = {display} {/O /Layout /Placement/Block}, + } +% \end{macrocode} +% +% \subsection{Code related to AF} +% Booleans to handle the options. +% \begin{variable}{ +% \l__tag_math_texsource_AF_bool, +% \l__tag_math_texsource_pane_bool, +% \l__tag_math_mathml_AF_bool, +% \g__tag_math_mathml_AF_bool, +% \l__tag_math_mathml_pane_bool, +% \l__tag_math_alt_bool +% } +% \begin{macrocode} +\bool_new:N\l__tag_math_texsource_AF_bool +\bool_new:N\l__tag_math_texsource_pane_bool +\bool_new:N\l__tag_math_mathml_AF_bool +\bool_new:N\g__tag_math_mathml_AF_bool +\bool_new:N\l__tag_math_mathml_pane_bool +\bool_new:N\l__tag_math_alt_bool +% \end{macrocode} +% \end{variable} +% +% \begin{variable}{ +% \g_@@_mathml_total_int, +% \g_@@_mathml_int, +% \g_@@_math_total_int, +% \g_@@_mathml_AF_found_int, +% \g_@@_mathml_AF_attached_int, +% } +% \cs{g_@@_mml_total_int} records the mathml fragments read in. +% \cs{g_@@_mml_int} records the mathml fragments read in with a different hash. +% \cs{g_@@_AF_total_int} records the number of math structures that try to +% attach a mathml AF. +% \cs{g_@@_AF_found_int} records the number of math structures for which a fitting +% mathml is found. +% \cs{g_@@_AF_attached_int} records the number of math structures which got a mathml fragment +% (if mathml-AF are not disabled locally this should be the equal to the previous number. +% +% \begin{macrocode} +\int_new:N\g_@@_mathml_total_int +\int_new:N\g_@@_mathml_int +\int_new:N\g_@@_math_total_int +\int_new:N\g_@@_mathml_AF_found_int +\int_new:N\g_@@_mathml_AF_attached_int +% \end{macrocode} +% \end{variable} +% +% \begin{variable}{\l__tag_math_mathml_files_clist} +% A sequence to store the file list for the mathml. +% \begin{macrocode} +\clist_new:N\l__tag_math_mathml_files_clist +\clist_put_right:Ne\l__tag_math_mathml_files_clist {\c_sys_jobname_str-mathml} +% \end{macrocode} +% \end{variable} + +% This is the internal variant of the \cs{mml} command. +% \begin{macro}{\@@_AF_mml:nnnn} +% \begin{macrocode} +\cs_new_protected:Npn \@@_AF_mml:nnnn #1 #2 #3 #4 +%#1 number, #2 tex source for debugging, #3 hash, #4 mathml + { + \int_gincr:N \g_@@_mathml_total_int +% \end{macrocode} +% mathml with the same hash should be included only once: +% \begin{macrocode} + \tl_if_exist:cF { g_@@_mathml_#3_tl } + { + \int_gincr:N \g_@@_mathml_int +% \end{macrocode} +% a simple Desc key, take care that it is a valid string! +% \begin{macrocode} + \pdfdict_put:nne {l_pdffile/Filespec} {Desc}{(mathml-#1)} + \pdffile_embed_stream:nnN {#4}{mathml-#1.xml}\l_@@_tmpa_tl +% \end{macrocode} +% not strictly necessary but makes the files visible in the file attachment +% page +% \begin{macrocode} + \bool_if:NT \l__tag_math_mathml_pane_bool + {\pdfmanagement_add:nne {Catalog/Names}{EmbeddedFiles}{\l_@@_tmpa_tl}} + \tl_new:c{g_@@_mathml_#3_tl} + \tl_gset_eq:cN{g_@@_mathml_#3_tl}\l_@@_tmpa_tl + } + } +% \end{macrocode} +% \end{macro} +% +% The html reader. +% \begin{macrocode} +\cs_new_protected:Npn \@@_AF_html_reader:w#1</h2>#2<p>#3</p>#4<p>#5</p>#6<math{ + \begingroup + \char_set_catcode_other:N\{ + \char_set_catcode_other:N\} + \char_set_catcode_other:N\# + \char_set_catcode_other:N\% + \@@_AF_html_reader_verb:w{#1}{#3}{#5}<math +} +% \end{macrocode} +% \begin{macrocode} +\cs_new_protected:Npn\@@_AF_html_reader_verb:w#1#2#3#4~</div>{ + \endgroup + \@@_AF_mml:nnnn{#1}{#2}{#3}{#4} + } +% \end{macrocode} + +% +% \begin{socketdecl}{tagsupport/math/mathml/write} +% This writes a html-dummy with the hash and the math content. +% This should be optional, so it uses a socket that can be disabled +% +% \begin{macrocode} +\socket_new:nn {tagsupport/math/mathml/write}{0} +% \end{macrocode} +% \end{socketdecl} +% +% \begin{plugdecl}{On} +% \begin{macrocode} +\socket_new_plug:nnn{tagsupport/math/mathml/write}{On} + { + \str_set:NV\l_@@_tmpa_str\l_@@_content_AF_source_tl + \str_replace_all:Nnn\l_@@_tmpa_str{&}{&} + \str_replace_all:Nnn\l_@@_tmpa_str{<}{<} + \iow_now:Ne \g_@@_writedummy_iow + { + \iow_newline: + <div> + \iow_newline: + <h2>\c_backslash_str mml\c_space_tl \int_use:N \g_@@_math_total_int </h2> + \iow_newline: + <p>\l_@@_tmpa_str</p> + \iow_newline: + <p>\l_@@_content_hash_tl </p> + \iow_newline: + <math></math> + \iow_newline: + </div> + \iow_newline: + } + } +% \end{macrocode} +% \end{plugdecl} +% And now a key to activate the socket. +% \begin{macrocode} + +\keys_define:nn { __tag / setup } + { + math/mathml/write-dummy .code:n = + { + \bool_gset_true:N \g__tag_math_mathml_AF_bool + \tl_if_exist:NF\g_@@_writedummy_iow + { + \iow_new:N \g_@@_writedummy_iow + \iow_open:Nn \g_@@_writedummy_iow + { + \c_sys_jobname_str-mathml-dummy.html + } + \iow_now:Ne \g_@@_writedummy_iow + { + <!DOCTYPE~html> + \iow_newline: + <html> + } + \AssignSocketPlug {tagsupport/math/mathml/write}{On} + \AddToHook{enddocument/afterlastpage} + { + \iow_now:Nn \g_@@_writedummy_iow {</html>} + \iow_close:N \g_@@_writedummy_iow + } + } + }, + math/mathml/write-dummy .usage:n=preamble + } +% \end{macrocode} +% +% \begin{macro}{\@@_AF_process_mathml_files:} +% \begin{macrocode} +\box_new:N\l_@@_tmpa_box +\cs_new_protected:Npn \@@_AF_process_mathml_files: + { + \hbox_set:Nn \l_@@_tmpa_box + { + \pdfdict_put:nnn { l_pdffile/Filespec }{AFRelationship} { /Supplement } + \pdfdict_put:nne + { l_pdffile }{Subtype} + { \pdf_name_from_unicode_e:n{application/mathml+xml} } + \char_set_catcode_other:N \# + \cs_set_eq:NN\mml \@@_AF_html_reader:w + \clist_map_inline:Nn \l__tag_math_mathml_files_clist + { + \file_if_exist:nTF {##1.html} + { + \typeout{Info:~reading~mathml~file~##1} + \file_input:n {##1.html} + \bool_gset_true:N\g__tag_math_mathml_AF_bool + } + { + \typeout{Info:~mathml~file~##1~does~not~exist}%info message + } + } + } + \bool_if:NT\g__tag_math_mathml_AF_bool + { + \typeout{Info:~Activating~mathml~support} + \AssignSocketPlug{tagsupport/math/struct/begin}{test-mathml} + \AssignSocketPlug{tagsupport/math/struct/end}{test-mathml} +% \end{macrocode} +% mathml handling doesn't like subparts, so we disable them for now: +% \begin{macrocode} + \AssignSocketPlug{tagsupport/math/substruct/begin}{single} + \AssignSocketPlug{tagsupport/math/substruct/end}{single} + \AddToHook{enddocument/info} + { + \iow_term:n{MathML~statistic} + \iow_term:n{================} + \iow_term:e{==>~\int_use:N\g_@@_mathml_total_int\c_space_tl + MathML~fragments~read} + \iow_term:e{==>~\int_use:N\g_@@_mathml_int\c_space_tl + different~MathML~fragments} + \iow_term:e{==>~\int_use:N\g_@@_math_total_int\c_space_tl + math~fragments~found} + \iow_term:e{==>~\int_use:N\g_@@_mathml_AF_found_int\c_space_tl + fitting~MathML~AF~found} + \iow_term:e{==>~\int_use:N\g_@@_mathml_AF_attached_int\c_space_tl + MathML~AF~attached} + } + } + } +\AddToHook{begindocument}{\@@_AF_process_mathml_files:} +% \end{macrocode} +% \end{macro} +% +% \subsection{Mathstyle detection} +% In some cases we need to detect the mathstyle used in a \cs{mathchoice} +% command and to disable/enable tagging in the unused branches. +% This is currently only used in the amstext command \cs{text} +% but is perhaps also needed in other cases, so we create a general command. +% +%\begin{macro}{\l_@@_mathstyle_int,\g_@@_mathchoice_int,mathstyle} +% \begin{macrocode} +\int_new:N \l_@@_mathstyle_int +\int_new:N \g_@@_mathchoice_int +\property_new:nnnn{mathstyle}{now}{-1}{\int_use:N \l_@@_mathstyle_int } +% \end{macrocode} +%\end{macro} +% For now internal, but perhaps will need a public version. +% The command should be used in every branch of a \cs{mathchoice} +% (with the correct mathstyle number) and with an unique label (which should +% be the same in every branch). +% \cs{g_@@_mathchoice_int} can be e.g. increased before the mathchoice and +% then used. +% \begin{macro}{\@@_tag_if_mathstyle:nn} +% \begin{macrocode} +\cs_new_protected:Npn \@@_tag_if_mathstyle:nn #1 #2 + %#1 refers to label + %#2 is a number for the mathstyle (typically 0,2,4,6) + { + \int_set:Nn \l_@@_mathstyle_int {#2} + \property_record:nn {#1} { mathstyle } + \int_compare:nNnTF { \property_ref:nn {#1}{ mathstyle} } = { #2 } + { \tag_start: }{ \tag_stop: } + } +\cs_generate_variant:Nn \@@_tag_if_mathstyle:nn {en} +% \end{macrocode} +% \end{macro} +% +% \subsection{Tagging options} +% \begin{macrocode} +\keys_define:nn { __tag / setup } + { + math/mathml/sources .clist_set:N = \l__tag_math_mathml_files_clist, + math/alt/use .bool_set:N = \l__tag_math_alt_bool, + viewer/pane/mathml .bool_set:N = \l__tag_math_mathml_pane_bool, + viewer/pane/mathml .initial:n = true, + viewer/pane/mathsource .bool_set:N = \l__tag_math_texsource_pane_bool, + math/mathml/AF .bool_set:N = \l__tag_math_mathml_AF_bool, + math/mathml/AF .initial:n = true, + math/tex/AF .bool_set:N = \l__tag_math_texsource_AF_bool, + math/tex/AF .initial:n = true + } +% \end{macrocode} +% +% \subsection{Sockets} +% \subsubsection{Main inline math sockets} +% +% \begin{socketdecl} +% { +% tagsupport/math/inline/begin, +% tagsupport/math/inline/end, +% tagsupport/math/inline/formula/begin, +% tagsupport/math/inline/formula/end, +% } +% The first two sockets are meant to embed inline +% math into the surrounding (so to close/reopen e.g. MC-chunks). +% The other two implement the actual formula structure. +% The formula sockets are despite their naming not symmetric: +% the begin socket is issued after the math has started, while +% the end socket is after the math! +% \begin{macrocode} +\socket_new:nn {tagsupport/math/inline/begin}{0} +\socket_new:nn {tagsupport/math/inline/end}{0} +\socket_new:nn {tagsupport/math/inline/formula/begin}{1} % +\socket_new:nn {tagsupport/math/inline/formula/end}{0} +% \end{macrocode} +%\end{socketdecl} +% +% +% \begin{plugdecl}{MC} +% \begin{macrocode} +\socket_new_plug:nnn + {tagsupport/math/inline/begin} + {MC} + {\tag_mc_end_push:} +\socket_new_plug:nnn + {tagsupport/math/inline/end} + {MC} + {\tag_mc_begin_pop:n{}} +% \end{macrocode} +% \end{plugdecl} +% +% We probably will want to test different tagging receipes. +% \begin{plugdecl}{default} +% \begin{macrocode} +\socket_new_plug:nnn + {tagsupport/math/inline/formula/begin} + {default} + { + \socket_use:n{tagsupport/math/content} + \socket_use:n{tagsupport/math/struct/begin} +% \end{macrocode} +% TODO: does inline math need subformula handling? +% \begin{macrocode} + % inner formula if multiple parts (not really implemented yet) + \socket_use:n{tagsupport/math/substruct/begin} + #1 + \socket_use:n{tagsupport/math/end} + } +\socket_new_plug:nnn + {tagsupport/math/inline/formula/end} + {default} + { + \socket_use:n{tagsupport/math/substruct/end} + \socket_use:n{tagsupport/math/struct/end} + } +% \end{macrocode} +% \end{plugdecl} +% +% \subsubsection{Main display math sockets} +% +% \begin{socketdecl} +% { +% tagsupport/math/display/begin, +% tagsupport/math/display/end, +% tagsupport/math/display/formula/begin, +% tagsupport/math/display/formula/end, +% } +% The first two sockets are meant to embed display +% math into the surrounding (so to close/reopen e.g. MC-chunks and +% P-structure). +% The other two implement the actual formula structure. +% The formula sockets are despite their naming not symmetric: +% the begin socket is issued after the math has started, while +% the end socket is after the math! +% \begin{macrocode} +\socket_new:nn {tagsupport/math/display/begin}{0} +\socket_new:nn {tagsupport/math/display/end}{0} +\socket_new:nn {tagsupport/math/display/formula/begin}{1} % +\socket_new:nn {tagsupport/math/display/formula/end}{0} +% \end{macrocode} +%\end{socketdecl} + +% \begin{plugdecl}{default} +% \begin{macrocode} +\socket_new_plug:nnn + {tagsupport/math/display/begin} + {default} + { \__tag_tool_close_P: } +\socket_new_plug:nnn + {tagsupport/math/display/end} + {default} + { + } +% \end{macrocode} +% \end{plugdecl} + + +% \begin{plugdecl}{default} +% \begin{macrocode} +\socket_new_plug:nnn + {tagsupport/math/display/formula/begin} + {default} + { + \socket_use:n{tagsupport/math/content} + \socket_use:n{tagsupport/math/struct/begin} + \socket_use:n{tagsupport/math/substruct/begin} + #1 + \socket_use:n{tagsupport/math/end} + } +\socket_new_plug:nnn + {tagsupport/math/display/formula/end} + {default} + { + \socket_use:n{tagsupport/math/substruct/end} + \socket_use:n{tagsupport/math/struct/end} + } +% \end{macrocode} +% \end{plugdecl} +% +% \subsubsection{Internal sockets} +% +% \begin{variable}{\l_@@_content_template_tl} +% The default text used as alt or actual text. +% \begin{macrocode} +\tl_new:N\l_@@_content_template_tl +\tl_set:Nn \l_@@_content_template_tl + { + LaTeX~ formula~ starts~ + \exp_not:N\begin{\g_@@_grabbed_env_tl} + \c_space_tl + \exp_not:V\g_@@_grabbed_math_tl + \c_space_tl + \exp_not:N\end{\g_@@_grabbed_env_tl} + \c_space_tl LaTeX~ formula~ ends~ + } +% \end{macrocode} +% \end{variable} + +% \begin{variable}{\l_@@_texsource_template_tl} +% The default text used as texsource +% \begin{macrocode} +\tl_new:N\l_@@_texsource_template_tl +\tl_const:Nn\c_@@_inline_env_tl {math} +\tl_set:Nn \l_@@_texsource_template_tl + { + \tl_if_eq:NNTF\g_@@_grabbed_env_tl\c_@@_inline_env_tl + { + $ + \exp_not:V\g_@@_grabbed_math_tl + $ + } + { + \exp_not:N\begin{\g_@@_grabbed_env_tl} + \exp_not:V\g_@@_grabbed_math_tl + \exp_not:N\end{\g_@@_grabbed_env_tl} + } + } +% \end{macrocode} +% \end{variable} + +% +% \begin{socketdecl}{tagsupport/math/content} +% The math content is stored in associated files and used for +% actual and alternative text. As the exact text is still +% unclear we use a socket to be able to test variants. +% The socket should set all four tl vars above, if needed +% to identical values. It can use the two variables +% \cs{g_@@_grabbed_env_tl} and \cs{g_@@_grabbed_math_tl} +% \begin{macrocode} +\socket_new:nn {tagsupport/math/content}{0} +% \end{macrocode} +% \end{socketdecl} +% +% Some default sockets to set the contents. +% TODO: think about naming convention. +% TODO: think how this should organized so that one +% has options to change from the outside and so that +% there are less repetitions. +% \begin{plugdecl}{actual+source} +% \begin{macrocode} +\socket_new_plug:nnn + {tagsupport/math/content} + {actual+source} + { + \tl_set:Ne\l_@@_content_actual_tl + { + \l_@@_content_template_tl + } + \tl_set:Ne \l_@@_content_AF_source_tl + { + \l_@@_texsource_template_tl + } + \tl_set:Nn \l_@@_content_AF_mathml_tl {} + \tl_set:Nn \l_@@_content_alt_tl {} + } +% \end{macrocode} +% \end{plugdecl} +% +% \begin{plugdecl}{alt+source} +% \begin{macrocode} +\socket_new_plug:nnn + {tagsupport/math/content} + {alt+source} + { + \tl_set:Ne\l_@@_content_alt_tl + { + \l_@@_content_template_tl + } + \tl_set:Ne \l_@@_content_AF_source_tl + { + \l_@@_texsource_template_tl + } + \tl_set:Nn \l_@@_content_AF_mathml_tl {} + \tl_set:Nn \l_@@_content_actual_tl {} + } +% \end{macrocode} +% \end{plugdecl} +% \begin{macrocode} +\socket_assign_plug:nn {tagsupport/math/content}{alt+source} +% \end{macrocode} +% +% \begin{socketdecl}{tagsupport/math/struct/begin, +% tagsupport/math/struct/end} +% For the main structure we use a socket too. +% This allow e.g. to create a special one for luamml +% which setups additional objects. +% The begin socket can use the two variables +% \cs{g_@@_grabbed_env_tl} and \cs{g_@@_grabbed_math_tl} +% \begin{macrocode} +\socket_new:nn {tagsupport/math/struct/begin}{0} +\socket_new:nn {tagsupport/math/struct/end}{0} +% \end{macrocode} +% \end{socketdecl} +% +% \begin{plugdecl}{default} +% TODO: think about some naming convention ... +% \begin{macrocode} +\socket_new_plug:nnn + {tagsupport/math/struct/begin} + {default} + { + \bool_if:NTF\l__tag_math_texsource_AF_bool + { \tl_set_eq:NN \l_@@_content_AF_source_tmpa_tl \l_@@_content_AF_source_tl } + { \tl_clear:N \l_@@_content_AF_source_tmpa_tl } + \tag_struct_begin:n + { + tag=Formula, + texsource = \l_@@_content_AF_source_tmpa_tl, + title-o = \g_@@_grabbed_env_tl, + actualtext = \l_@@_content_actual_tl, + alt = \l_@@_content_alt_tl + } + } +\socket_new_plug:nnn + {tagsupport/math/struct/end} + {default} + { \tag_struct_end: } + +\socket_assign_plug:nn {tagsupport/math/struct/begin}{default} +\socket_assign_plug:nn {tagsupport/math/struct/end}{default} +% \end{macrocode} +% \end{plugdecl} +% +% \begin{plugdecl}{test-mathml} +% This (test-)socket tries to add a mathml-AF to formula. +% It is activated if a mathml.html has been found and loaded. +% Additionally it also sets an attribute +% (this can perhaps be done by default anyway.) +% As it disturbs the reading of the AF it currently deactivates the /Alt key, +% unless it has been reenabled with |math/alt/use=true| +% \begin{macrocode} +\cs_generate_variant:Nn \str_mdfive_hash:n {o} +\tl_new:N\l_@@_content_hash_tl +% \end{macrocode} +% we need to save the grabbed math: +% \begin{macrocode} +\tl_new:N\l_@@_grabbed_math_tl +% \end{macrocode} +% the socket definition +% \begin{macrocode} +\socket_new_plug:nnn + {tagsupport/math/struct/begin} + {test-mathml} + { + \int_gincr:N\g_@@_math_total_int + \tl_set:Ne\l_@@_content_hash_tl + {\str_mdfive_hash:o { \l_@@_content_AF_source_tl }} + \tl_set_eq:NN\l_@@_grabbed_math_tl\g_@@_grabbed_math_tl + \tl_if_eq:NnTF\g_@@_grabbed_env_tl {math} + { + \tl_set:Nn\l_@@_attribute_class_tl{inline} + } + { + \tl_set:Nn\l_@@_attribute_class_tl{display} + } + \bool_if:NF\l__tag_math_alt_bool + { \tl_set:Nn \l_@@_content_alt_tl{} } +% \end{macrocode} +% debugging option. TODO: hide in debug key. +% \begin{macrocode} + \tl_if_exist:cTF { g_@@_mathml_ \l_@@_content_hash_tl _tl } + { + \int_gincr:N\g_@@_mathml_AF_found_int + \bool_if:NTF \l__tag_math_mathml_AF_bool + { + \int_gincr:N\g_@@_mathml_AF_attached_int + \typeout {Inserting~mathml~with~Hash~\l_@@_content_hash_tl} + } + { + \typeout {Ignoring~mathml~with~Hash~\l_@@_content_hash_tl} + } + } + { + \typeout{WARNING:~mathml~missing~for~hash~\l_@@_content_hash_tl} + } + \socket_use:n {tagsupport/math/mathml/write} % write hash if request + \bool_if:NTF\l__tag_math_texsource_AF_bool + { \tl_set_eq:NN \l_@@_content_AF_source_tmpa_tl \l_@@_content_AF_source_tl } + { \tl_clear:N \l_@@_content_AF_source_tmpa_tl } + \tag_struct_begin:n + { + tag=Formula, + attribute-class=\l_@@_attribute_class_tl, % + AFref = + \bool_if:NT\l__tag_math_mathml_AF_bool + { + \cs_if_exist_use:c {g_@@_mathml_ \l_@@_content_hash_tl _tl} + }, + texsource = \l_@@_content_AF_source_tmpa_tl, % should be after mathml AF! + title-o = \g_@@_grabbed_env_tl, % + alt = \l_@@_content_alt_tl + } + } +% \end{macrocode} +% not really needed but looks more symmetric: +% \begin{macrocode} +\socket_new_plug:nnn + {tagsupport/math/struct/end} + {test-mathml} + { + \tag_struct_end: + } +% \end{macrocode} +% \end{plugdecl} +% +% \begin{socketdecl}{tagsupport/math/substruct/begin, +% tagsupport/math/substruct/end} +% This holds the code to handle subparts of the formula. +% \begin{macrocode} +\socket_new:nn {tagsupport/math/substruct/begin}{0} +\socket_new:nn {tagsupport/math/substruct/end}{0} +% \end{macrocode} +% \end{socketdecl} +% +% \begin{plugdecl}{default} +% \begin{macrocode} +\socket_new_plug:nnn + {tagsupport/math/substruct/begin} + {default} + { \grabaformulapartandstart } +\socket_new_plug:nnn + {tagsupport/math/substruct/end} + {default} + { + \tagmcend + \if@subformulas + \tagstructend + \fi + } +\socket_assign_plug:nn {tagsupport/math/substruct/begin}{default} +\socket_assign_plug:nn {tagsupport/math/substruct/end}{default} +% \end{macrocode} +% \end{plugdecl} + +% \begin{plugdecl}{single} +% We need an option to disable subparts as it is unclear +% if consumers can handle them: +% \begin{macrocode} +\socket_new_plug:nnn + {tagsupport/math/substruct/begin} + {single} + { + \typeout{====>subpart~splitting~deactivated} + \typeout{====>grabbed~math=\meaning\g_@@_grabbed_math_tl} + \tag_mc_begin:n{} + } +\socket_new_plug:nnn + {tagsupport/math/substruct/end} + {single} + { \tag_mc_end:} +% \end{macrocode} +% \end{plugdecl} +% +% \begin{socketdecl}{tagsupport/math/end} +% A socket used at the end of the math (before the closing dollar(s)) +% which can e.g. set a flag for luamml. +% \begin{macrocode} +\socket_new:nn {tagsupport/math/end}{0} +% \end{macrocode} +% \end{socketdecl} + +% +% \begin{macro}{\__tag_math_disable:} +% Similar to the table code we collect the plugs that should be +% assigned to do nothing if we don't want tagging +% \begin{macrocode} +\cs_new_protected:Npn \__tag_math_disable: + { + \socket_assign_plug:nn {tagsupport/math/inline/begin}{noop} + \socket_assign_plug:nn {tagsupport/math/inline/end}{noop} + \socket_assign_plug:nn {tagsupport/math/inline/formula/begin}{identity} + \socket_assign_plug:nn {tagsupport/math/inline/formula/end}{noop} + \socket_assign_plug:nn {tagsupport/math/display/begin}{noop} + \socket_assign_plug:nn {tagsupport/math/display/end}{noop} + \socket_assign_plug:nn {tagsupport/math/display/formula/begin}{identity} + \socket_assign_plug:nn {tagsupport/math/display/formula/end}{noop} + } +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\__tag_math_enable:} +% Similar to the table code we collect the default plugs that should be +% assigned if we want tagging +% \begin{macrocode} +\cs_new_protected:Npn \__tag_math_enable: + { + \socket_assign_plug:nn {tagsupport/math/inline/begin}{MC} + \socket_assign_plug:nn {tagsupport/math/inline/end}{MC} + \socket_assign_plug:nn {tagsupport/math/inline/formula/begin}{default} + \socket_assign_plug:nn {tagsupport/math/inline/formula/end}{default} + \socket_assign_plug:nn {tagsupport/math/display/begin}{default} + \socket_assign_plug:nn {tagsupport/math/display/end}{default} + \socket_assign_plug:nn {tagsupport/math/display/formula/begin}{default} + \socket_assign_plug:nn {tagsupport/math/display/formula/end}{default} + } +% \end{macrocode} +% \end{macro} +% At begin document we can activate: +% \begin{macrocode} +\AtBeginDocument{\tag_if_active:T{\__tag_math_enable: }} +% \end{macrocode} % % \subsection{Interface commands} % @@ -260,7 +1322,8 @@ { \legacy_if:nF { measuring@ } { - \tl_if_in:nnF {#2} { \m@th } + \tl_if_in:nnTF {#2} { \m@th } + { \bool_set_true:N\l_@@_fakemath_bool } { \tl_trim_spaces_apply:nN {#2} \@@_process_auxi:nn {#1} } } } @@ -287,15 +1350,15 @@ % % \begin{macro}{\@@_grab_dollar:w} % Grab up to a single |$|, for inline math mode, suppressing -% any processing if the first token is \tn{m@th}. +% any processing if the token is \tn{m@th} found in the content. +% +% \fmi{what's that test doing?} % -% \fmi{what's that test doing?} +% \car{It is some kind of fix, to avoid the remote +% possibility that the math is empty, making the code +% produce an unwanted \texttt{\$\$}.} % -% \car{It is some kind of fix, to avoid the remote -% possibility that the math is empty, making the code -% produce an unwanted \texttt{\$\$}.} -% -%\car{cf.~the code for this in \cs{@ensuredmath}} +%\car{cf.~the code for this in \cs{@ensuredmath}} % %\car{It is harmless but unnecessary in the dollardollar grabbing below.} % @@ -309,91 +1372,62 @@ \tl_if_blank:nF {#1} { \@@_process:nn { math } {#1} % $ -% fairly simple this one % \end{macrocode} -% We do not want math tagging in fakemath or when measuring, -% so we imitate the test inside \cs{@@_process:nn} for now, -% see https://github.com/latex3/tagging-project/issues/5 -% TODO: use socket to get more control about typesetting variants (tagged, drop etc)? +% We do not want math tagging in fakemath or when measuring, +% We also do not want math tagging if tagging has been suspended. % \begin{macrocode} - \legacy_if:nTF { measuring@ } - { #1 $ } - { - \tl_if_in:nnTF {#1} { \m@th } - { #1 $ } - { - \tagmcend %end P-chunk, in code: \tag_mc_end_push: - \@kernel@math@begin - #1 $ - \@kernel@math@end - \tagmcbegin{} % restart P-chunk (whatsits in pdftex) - } - } + \bool_lazy_or:nnTF + {\legacy_if_p:n { measuring@ }} + { \l_@@_fakemath_bool } + { #1 $ } + { + \tag_socket_use:n {math/inline/begin} %end P-MC + \tag_socket_use:nn {math/inline/formula/begin}{#1} + $ + \tag_socket_use:n {math/inline/formula/end} + \tag_socket_use:n {math/inline/end} % restart P-MC + } } } % \end{macrocode} % \end{macro} +% % \begin{macro}{\@@_grab_dollardollar:w} % And for the classical \TeX{} display structure. % \begin{macrocode} - -\skip_new:N \l_@@_tmpa_skip - \cs_new_protected:Npn \@@_grab_dollardollar:w % $$ #1 $$ { \tl_if_blank:nF {#1} { - \@@_tag_dollardollar_display:nn { equation* }{#1} - #1 + \@@_process:nn { equation* } {#1} + \socket_use:n {tagsupport/math/display/begin} + \socket_use:nn{tagsupport/math/display/formula/begin}{#1} $$ } } % \end{macrocode} -% To allow to use the code without tagging we guard. But probably tagpdf should -% provide some tools for such manual para-ends. +% +% The end code is added through a \cs{aftergroup} so we +% store it inside a command. % \begin{macrocode} -\cs_new_protected:Npn \@kernel@close@P { - \tag_if_active:T - { - \tagmcend %end P-chunk, should perhaps be \tag_mc_end_push: ... - \int_gincr:N \g__tag_para_end_int - \bool_if:NT \l__tag_para_show_bool - { \tag_mc_begin:n{artifact} - \rlap{\color_select:n{red}\tiny\ \int_use:N\g__tag_para_end_int} - \tag_mc_end: - } - \tag_struct_end: - } -} - - - -\cs_new_protected:Npn \@@_tag_dollardollar_display:nn #1#2 { - \@@_process:nn {#1} {#2} - \@kernel@close@P - \@kernel@math@begin -% \skip_set:Nn \belowdisplayskip {-\belowdisplayskip} -% \skip_set:Nn \belowdisplayshortskip {-\belowdisplayshortskip} -% \int_set:Nn \postdisplaypenalty {10000} -%% -% \group_insert_after:N \@@_tag_dollardollar_display_end: -} - -\cs_new_protected:Npn \@@_tag_dollardollar_display_end: { -% \typeout{== tag dollarldollar display end} -% \ShowTagging{struct-stack} - \tagpdfparaOff - \para_raw_end: - \tagpdfparaOn - \l_@@_tmpa_skip \lastskip - \@kernel@math@end - \penalty \postdisplaypenalty +\cs_new_protected:Npn \@@_tag_dollardollar_display_end: + { + % \typeout{== tag dollarldollar display end} + % \ShowTagging{struct-stack} + \para_raw_end: +% \end{macrocode} +% TODO why is that needed? where is para-tagging disabled? +% \begin{macrocode} + \tagpdfparaOn + \l_@@_tmpa_skip \lastskip + \socket_use:n{tagsupport/math/display/formula/end} + \penalty \postdisplaypenalty % \end{macrocode} % This reinserts the below display skips. It must be doubled to % get the right amount: % \begin{macrocode} - \skip_vertical:n { -\l_@@_tmpa_skip * 2 } + \skip_vertical:n { -\l_@@_tmpa_skip * 2 } % \@doendpe % this has no \end{...} to take care of it } @@ -410,13 +1444,13 @@ { \tl_if_blank:nF {#1} { - \@@_process:nn { math } {#1} $ #1 $ } \bool_set_false:N \l_@@_collected_bool } % \end{macrocode} % \end{macro} +% % \begin{macro}{\@@_grab_eqn:w} % For the most common use of \cs{[}/\cs{]}: turn into an environment. % \begin{macrocode} @@ -463,7 +1497,7 @@ % \begin{macro}{\math_register_env:n} % \begin{macro}{\RegisterMathEnvironment} % Set up to capture environment content and make available. -% +% % \begin{macrocode} \cs_new_protected:Npn \math_register_env:nn #1#2 { @@ -472,15 +1506,8 @@ \cs_gset_eq:cc { @@_env_ #1 _begin: } {#1} \cs_gset_eq:cc { @@_env_ #1 _end: } { end #1 } % - \ExpandArgs { nnx } \RenewDocumentEnvironment {#1} { b } + \ExpandArgs { nne } \RenewDocumentEnvironment {#1} { b } { -% \bool_set_true:N \exp_not:N \l_@@_collected_bool -% \cs_if_exist:cTF { @@_env #1 _aux: } -% { -% \exp_not:c { @@_env #1 _aux: } -% ####1 \exp_not:N \@@_env_end: {#1} -% } -% { \exp_not:N \@@_process:nn {#1} {####1} } \exp_not:N \bool_if:NTF \exp_not:N \l_@@_collected_bool { % \typeout{===>B1} @@ -490,34 +1517,22 @@ \cs_if_exist:cTF { @@_env #1 _aux: } { \exp_not:c { @@_env #1 _aux: } - ####1 \exp_not:N \@@_env_end: {#1} + ##1 \exp_not:N \@@_env_end: {#1} } - { \exp_not:N \@@_process:nn {#1} {####1} } + { \exp_not:N \@@_process:nn {#1} {##1} } \exp_not:n { \@kernel@math@registered@begin } \bool_set_true:N \exp_not:N \l_@@_collected_bool } -% \exp_not:N \tracingall +% \exp_not:N \tracingall \exp_not:c { @@_env_ #1 _begin: } - ####1 + ##1 \exp_not:c { @@_env_ #1 _end: } -% \exp_not:c { @@_env_ #1 _end: } % \exp_not:N \tracingnone -% \exp_not:n { \@kernel@math@registered@end } } - { - } + { + } } -\cs_set_protected:Npn \__cs_tmp:w #1 - { - \group_begin: - \exp_args:No \__cs_generate_internal_variant:n - { \tl_to_str:n {#1} } - \group_end: - } -\__cs_tmp:w { nnxx } - - \cs_new_protected:Npn \math_register_halign_env:nn #1#2 { \tl_set:Nn \l_@@_env_name_tl {#1} @@ -525,15 +1540,8 @@ \cs_gset_eq:cc { @@_env_ #1 _begin: } {#1} \cs_gset_eq:cc { @@_env_ #1 _end: } { end #1 } % - \ExpandArgs { nnxx } \RenewDocumentEnvironment {#1} { b } + \ExpandArgs { nnee } \RenewDocumentEnvironment {#1} { b } { -% \bool_set_true:N \exp_not:N \l_@@_collected_bool -% \cs_if_exist:cTF { @@_env #1 _aux: } -% { -% \exp_not:c { @@_env #1 _aux: } -% ####1 \exp_not:N \@@_env_end: {#1} -% } -% { \exp_not:N \@@_process:nn {#1} {####1} } \exp_not:N \bool_if:NTF \exp_not:N \l_@@_collected_bool { % \typeout{===>B1} @@ -543,23 +1551,24 @@ \cs_if_exist:cTF { @@_env #1 _aux: } { \exp_not:c { @@_env #1 _aux: } - ####1 \exp_not:N \@@_env_end: {#1} + ##1 \exp_not:N \@@_env_end: {#1} } - { \exp_not:N \@@_process:nn {#1} {####1} } + { \exp_not:N \@@_process:nn {#1} {##1} } \exp_not:n { \@kernel@math@registered@begin } \bool_set_true:N \exp_not:N \l_@@_collected_bool } -% \exp_not:N \tracingall +% \exp_not:N \tracingall \exp_not:c { @@_env_ #1 _begin: } - ####1 -% \exp_not:c { @@_env_ #1 _end: } + ##1 % \exp_not:N \tracingnone } - { - \exp_not:c { @@_env_ #1 _end: } - } + { + \exp_not:c { @@_env_ #1 _end: } + } } - +% \end{macrocode} +% TODO: the following command is neither documented nor used. Is is needed? +% \begin{macrocode} \cs_new_protected:Npn \math_register_odd_env:nn #1#2 { \tl_set:Nn \l_@@_env_name_tl {#1} @@ -567,7 +1576,7 @@ \cs_gset_eq:cc { @@_env_ #1 _begin: } {#1} \cs_gset_eq:cc { @@_env_ #1 _end: } { end #1 } % - \ExpandArgs { nnxx } \RenewDocumentEnvironment {#1} { b } + \ExpandArgs { nnee } \RenewDocumentEnvironment {#1} { b } { \exp_not:N \bool_if:NTF \exp_not:N \l_@@_collected_bool { @@ -578,19 +1587,19 @@ \cs_if_exist:cTF { @@_env #1 _aux: } { \exp_not:c { @@_env #1 _aux: } - ####1 \exp_not:N \@@_env_end: {#1} + ##1 \exp_not:N \@@_env_end: {#1} } - { \exp_not:N \@@_process:nn {#1} {####1} } + { \exp_not:N \@@_process:nn {#1} {##1} } \exp_not:n { \@kernel@math@registered@begin } \bool_set_true:N \exp_not:N \l_@@_collected_bool } -% \exp_not:N \tracingall +% \exp_not:N \tracingall \exp_not:c { @@_env_ #1 _begin: } - ####1 + ##1 } { \exp_not:c { @@_env_ #1 _end: } -% needed if we don't have $$...$$ +% needed if we don't have $$...$$ % \exp_not:n { \typeout{---> @kernel@math@registered@end }} \exp_not:n { \@kernel@math@registered@end } } @@ -614,7 +1623,7 @@ \cs_new:Npn \@kernel@math@registered@begin { -% \ShowTagging{struct-stack} +% \ShowTagging{struct-stack} %\typeout{==>A1}\ShowTagging{struct-stack,mc-current} \mode_if_vertical:TF { @@ -623,19 +1632,21 @@ % { \__block_list_beginpar_vmode: } % % \typeout{==>~ at:~ \g__tag_struct_tag_tl} -% - \exp_args:Noo\str_if_eq:nnF \g__tag_struct_tag_tl { \l__tag_para_main_tag_tl } % needs correction! +% + \tag_if_active:T + { + \exp_args:Noo\str_if_eq:nnF \g__tag_struct_tag_tl { \l__tag_para_main_tag_tl } % needs correction! { % \typeout{==>A2} \__block_beginpar_vmode: } % needs correction! + } } { % \typeout{==>A3} - \@kernel@close@P -% \tagmcend % needs correction! + \__tag_tool_close_P: } - \@kernel@math@begin + \socket_use:nn{tagsupport/math/display/formula/begin}{} \tagpdfparaOff % \typeout{==>MC1}\ShowTagging{mc-current} } @@ -644,13 +1655,14 @@ % \typeout{==>MC2}\ShowTagging{mc-current} \para_raw_end: \tagpdfparaOn - \@kernel@math@end + \socket_use:n{tagsupport/math/display/formula/end} % \typeout{==>MC3}\ShowTagging{mc-current} \@endpetrue } - + \cs_new_protected:Npn \math_register_env:n #1 { \math_register_env:nn {#1} { } } + \NewDocumentCommand \RegisterMathEnvironment { O{} m } { \math_register_env:nn {#2} {#1} } % \end{macrocode} @@ -660,12 +1672,12 @@ % % \begin{macro}{\@@_env_forward:w} % \begin{macrocode} -\cs_new_protected:Npn \@@_env_forward:w #1 \@@_env_end: #2 +\cs_new_protected:Npn \@@_env_forward:w #1 \@@_env_end: #2 { \@@_process:nn {#2} {#1} } % \end{macrocode} % \end{macro} -% -% \subsection{Document commands} +% +% \subsection{Document commands} % % \car*{Add one more here: \texttt{displaymath}, which % is equivalent to \cs{[} , \cs{]}\\ @@ -692,9 +1704,9 @@ % at the moment register_env can only do display math % \math_register_env:n { math } \RenewDocumentEnvironment{math} {b}{$#1$}{} -% and this one doesn't work either +% and this one doesn't work either % \math_register_env:n { displaymath } - \RenewDocumentEnvironment{displaymath} {b}{\[#1\]}{} + \RenewDocumentEnvironment{displaymath} {b}{\[#1\]}{} } % \end{macrocode} % \end{macro} @@ -737,29 +1749,31 @@ { \cs_gset_protected:Npn \[ % \] { - \bool_if:NTF \l_@@_collected_bool - { \begin { equation* } } - { \@@_grab_eqn:w } + \@@_grab_eqn:w + % \bool_if:NTF \l_@@_collected_bool +% { \begin { equation* } } +% { \@@_grab_eqn:w } } % \[ \cs_gset_protected:Npn \] { - \bool_if:NTF \l_@@_collected_bool - { \end{ equation* } } - { \@badmath } + \@badmath +% \bool_if:NTF \l_@@_collected_bool +% { \end{ equation* } } +% { \@badmath } } } % \end{macrocode} % \end{macro} -% +% % % \begin{macro}{\ensuremath} % A bit of nesting fun to make sure we collect only if required. % \fmi{why does ensuremath need handling at all?} -% -% \car{Indeed! Currently, this is setup to process the math that -% it has anyways already captured as its argument; thus it is more +% +% \car{Indeed! Currently, this is setup to process the math that +% it has anyways already captured as its argument; thus it is more % efficient than leaving the capture to be repeated by the \cs{everymath}} -% +% % \begin{macrocode} %\cs_gset_protected:Npn \ensuremath #1 % { @@ -786,47 +1800,6 @@ % have macro control. % \begin{macrocode} -\tl_new:N\tmpmathcontent - - -\def\@kernel@math@begin { -% \typeout{==>~math~begin} -% needs different handling if we support nesting - \tl_gset:Nx\tmpmathcontent - { - LaTeX~ formula~ starts~ - \exp_not:N\begin{\g_@@_grabbed_env_tl} - \space - \exp_not:V\g_@@_grabbed_math_tl - \space - \exp_not:N\end{\g_@@_grabbed_env_tl} - \space LaTeX~ formula~ ends~ - } - \tagstructbegin{tag=Formula, - AFinline-o=\tmpmathcontent, - title-o=\g_@@_grabbed_env_tl, - actualtext=\tmpmathcontent -% alt=\tmpmathcontent - } -% inner formula if multiple parts (not really implemented yet) - \grabaformulapartandstart -% the above does: -% \tagstructbegin{tag=Formula}\tagmcbegin{} -% or just -% \tagmcbegin{} -} -\def\@kernel@math@end { -% \typeout{==>~math~end} -% \ShowTagging{struct-stack} - \tagmcend - \if@subformulas - \tagstructend - \else - \fi - \tagstructend -% \ShowTagging{struct-stack} -} - \exp_args:No \tex_everymath:D { \tex_the:D \tex_everymath:D @@ -841,24 +1814,21 @@ { \tex_the:D \tex_everydisplay:D \iftrue % this may have to be a settable flag! -% { % \typeout{==>~ in~ everydisplay} % \end{macrocode} -% flipping the \cs{belowdisplay} values is done so that we get (assumption) -% a negative skip and not make the page bigger then we take that out, -% then we add the tagging code (in \cs{@@_tag_dollardollar_display_end} ) and -% then we put a real \cs{postdisplaypenalty} in and -% the right skip (of which we don't know if it is short or a -% normal \cs{belowdisplayskip}). This might need some refinement if that skip -% is actually negative from the start +% flipping the \cs{belowdisplay} values is done so that we get (assumption) +% a negative skip and not make the page bigger then we take that out, +% then we add the tagging code (in \cs{@@_tag_dollardollar_display_end} ) and +% then we put a real \cs{postdisplaypenalty} in and +% the right skip (of which we don't know if it is short or a +% normal \cs{belowdisplayskip}). This might need some refinement if that skip +% is actually negative from the start % (not sure it ever is and is worth bothering about) % \begin{macrocode} \skip_set:Nn \belowdisplayskip {-\belowdisplayskip} \skip_set:Nn \belowdisplayshortskip {-\belowdisplayshortskip} \int_set:Nn \postdisplaypenalty {10000} -% \group_insert_after:N \@@_tag_dollardollar_display_end: -% } \fi \bool_if:NF \l_@@_collected_bool { @@ -876,14 +1846,23 @@ \math_register_env:n { eqnarray* } % \end{macrocode} % -% Places where math mode is (ab)used. +% Tabulars currently contain a \$ that shouldn't trigger math +% tagging. % \begin{macrocode} -\clist_map_inline:nn - { tabular } - { - \AddToHook{ env / #1 / begin } - { \bool_set_true:N \l_@@_collected_bool } - } +\RequirePackage{array} +\tl_if_in:NnT\@tabular{$} + { + \def\@tabular{% + \leavevmode + \UseTaggingSocket{tbl/hmode/begin}% + \hbox \bgroup + \bool_set_true:N \l_@@_collected_bool + $ + \bool_set_false:N \l_@@_collected_bool + \col@sep\tabcolsep \let\d@llarbegin\begingroup + \let\d@llarend\endgroup + \@tabarray} + } % \end{macrocode} % % \begin{macro}{\@@_m@th:, \m@th} @@ -898,6 +1877,20 @@ } % \end{macrocode} % \end{macro} +% +% \subsection{Disable math grabbing in the begindocument hook} +% For example amsart uses math to measure text there. +% +% \begin{macrocode} +\tl_gput_right:Nn\@kernel@before@begindocument + { + \bool_set_true:N\l_@@_collected_bool + } +\tl_gput_right:Nn\@kernel@after@begindocument + { + \bool_set_false:N\l_@@_collected_bool + } +% \end{macrocode} % % \subsection{Modifying \pkg{amsmath}} % @@ -934,7 +1927,7 @@ $$\ignorespacesafterend } % \end{macrocode} -% +% % \begin{macrocode} \def\common@align@ending { \math@cr \black@\totwidth@ @@ -1007,9 +2000,11 @@ % \end{macrocode} % % \begin{macrocode} -% +% \math_register_halign_env:nn {align}{} \math_register_halign_env:nn {align*}{} + \math_register_halign_env:nn {alignat}{} + \math_register_halign_env:nn {alignat*}{} \math_register_halign_env:nn {flalign}{} \math_register_halign_env:nn {flalign*}{} \math_register_halign_env:nn {gather}{} @@ -1032,44 +2027,8 @@ \fi } % \end{macrocode} -% -% \begin{macrocode} - \def\intertext@{% - \def\intertext##1{% - \ifvmode\else\\\@empty\fi - \noalign{% -% we have to flip the sign on the skip because we flipped it on the outside - \penalty\postdisplaypenalty\vskip-\belowdisplayskip - \vbox{ -% \end{macrocode} -% Stop tagging when measuring: -% \begin{macrocode} - \ifmeasuring@\tag_stop:\fi - \normalbaselines - \ifdim\linewidth=\columnwidth - \else \parshape\@ne \@totalleftmargin \linewidth - \fi -% \end{macrocode} -% End the previous mc:\fmi{if we use 2 levels of formulas this would -% need changing} -% \begin{macrocode} - \tag_mc_end_push: -% \end{macrocode} -% We are already in a par so we change now to Span:\fmi{not true any longer} -% \begin{macrocode} - \tagpdfsetup{paratag=P}% - \tagpdfparaOn - \noindent\ignorespaces##1\par -% \end{macrocode} -% Restart the MC -% \begin{macrocode} - \tag_mc_begin_pop:n{}}% - \penalty\predisplaypenalty\vskip\abovedisplayskip% - }% - } - } -% \end{macrocode} -% +% +% % \begin{macrocode} \@namedef{math@cr @ @ @ gather}{% \ifst@rred\nonumber\fi @@ -1083,7 +2042,7 @@ \cr } % \end{macrocode} -% +% % \begin{macrocode} \@namedef{math@cr @ @ @ align}{% \ifst@rred\nonumber\fi @@ -1104,7 +2063,7 @@ \cr } % \end{macrocode} -% +% % \begin{macrocode} \def\restore@math@cr{\@namedef{math@cr @ @ @}{ % @@ -1113,7 +2072,7 @@ \cr}} \restore@math@cr % \end{macrocode} -% +% % \begin{macrocode} } % \end{macrocode} @@ -1125,9 +2084,18 @@ % % -% \begin{macro}{} +% \begin{macro}{\@@_split_at_nl:NN} +% This splits grabbed math at newlines. % % \begin{macrocode} +\cs_new:Npn \@@_split_at_nl:NN #1#2 { + \tl_set:Nf \l_@@_tmpa_tl { + \exp_after:wN \@@_split_at_nl_first:w #1 \\ \q_nil \\ \s_stop } + \exp_after:wN \@@_split_at_nl_aux:nnNN \l_@@_tmpa_tl #1 #2 +} +% \end{macrocode} +% and the auxiliary commands +% \begin{macrocode} \cs_new:Npn \@@_split_at_nl_first:w #1 \\ #2 \\ #3 \s_stop { \quark_if_nil:nTF {#2} @@ -1137,8 +2105,15 @@ #2 \\ #3 \s_stop } } -\cs_new:Npn \@@_split_chk_if_begin:ww #1 \begin #2 #3 \s_mark - #4 \\ \q_nil \\ \s_stop + +\cs_new_protected:Npn \@@_split_at_nl_aux:nnNN #1 #2 #3 #4 + { + \tl_gset:Nn #4 {#1} + \tl_gset:Nn #3 {#2} + } + +\cs_new:Npn \@@_split_chk_if_begin:ww + #1 \begin #2 #3 \s_mark #4 \\ \q_nil \\ \s_stop { \quark_if_nil:nTF {#2} { {#1} {#4} } @@ -1148,7 +2123,9 @@ { } { 1 } } } + \cs_new:Npn \@@_split_cleanup_begin_q_nil:w #1 \begin \q_nil {#1} + \cs_new:Npn \@@_split_collect_one_end:w #1 \end #2 #3 \s_stop #4 #5 { \exp_args:Nf \@@_split_check_count_begins:nnnn @@ -1157,14 +2134,16 @@ } \cs_new:Npn \@@_split_count_begins:n #1 { \int_eval:n { 0 \@@_split_count_begins:w #1 \begin \q_nil } } + \cs_new:Npn \@@_split_count_begins:w #1 \begin #2 { \quark_if_nil:nF {#2} { +1 \@@_split_count_begins:w } } + \cs_new:Npn \@@_split_check_count_begins:nnnn #1 #2 #3 #4 { \int_compare:nNnTF {#1} = {#2} { \exp_last_unbraced:Nf \@@_split_final_cleanup:nn - { \split:n { \@@_split_guard:n {#3} #4 } } + { \@@_split:n { \@@_split_guard:n {#3} #4 } } } { \exp_args:No \use_ii_i:nn @@ -1184,29 +2163,17 @@ { \exp_end: { #4 #1 } } { \@@_split_final_cleanup:w #3 \s_mark { #4 #1 #2 } } } -\NewDocumentCommand \splitnl { mm +m } - { - \tl_set:Nf \l_tmpa_tl { \split:n {#3} } - \show \l_tmpa_tl - \exp_after:wN \__splitnl_aux:nnNN \l_tmpa_tl #1 #2 - } - - -\cs_new:Npn \split:n #1 { + +\cs_new:Npn \@@_split:n #1 { \@@_split_at_nl_first:w #1 \\ \q_nil \\ \s_stop } - -\cs_new:Npn \@@_split_at_nl:NN #1#2 { - \tl_set:Nf \l_tmpa_tl { - \exp_after:wN \@@_split_at_nl_first:w #1 \\ \q_nil \\ \s_stop } - \exp_after:wN \@@_split_at_nl_aux:nnNN \l_tmpa_tl #1 #2 -} - -\cs_new_protected:Npn \@@_split_at_nl_aux:nnNN #1 #2 #3 #4 - { - \tl_gset:Nn #4 {#1} - \tl_gset:Nn #3 {#2} -} - + +% this looks unused. +%\NewDocumentCommand \splitnl { mm +m } +% { +% \tl_set:Nf \l_@@_tmpa_tl { \split:n {#3} } +% \show \l_@@_tmpa_tl +% \exp_after:wN \__splitnl_aux:nnNN \l_@@_tmpa_tl #1 #2 +% } % \end{macrocode} % \end{macro} % @@ -1233,9 +2200,11 @@ \edef\resulttitle{\g_@@_grabbed_env_tl\space (part)} \tagstructbegin{tag=Formula, % \end{macrocode} -% For now we don't put anything in /alt or /ActualText on subformulas +% For now we don't put real content in /alt or /ActualText on subformulas +% but we add a short text to satisfy the pdf/ua-2 validator % \begin{macrocode} % alt=\result, + alt = subformula, title-o=\resulttitle } } @@ -1301,10 +2270,6 @@ % \end{macrocode} % % \begin{macrocode} -% -% \end{macrocode} -% -% \begin{macrocode} %</kernel> % \end{macrocode} % |