%%
%% \iffalse ^^A we skip the debug setting
\IfFileExists{\string~/bin/gen_and_xela}{\errorcontextlines=1000 }{}
%% \fi
%% ^^A $Id: TheUltimateMaster.gmd,v 0.298 2011/02/02 18:35:58 natror Exp natror $
%% This is file ���gmdoc.sty��� generated with the DocStrip utility.
%%
%%
%% The original source files were:
%%
%% gmdoc/gmdoc.gmd (with options: `doc')
%%
%%
%% Copyright �� 2006, 2007, 2008, 2009, 2010
%% by Grzegorz ���Natror��� Murzynowski
%% natror (at) gmail (dot) com
%%
%% This program is subject to the LaTeX Project Public License.
%% See http://www.ctan.org/tex-archive/help/Catalogue/licenses.lppl.html
%% for the details of that license.
%%
%% LPPL status: "author-maintained".
%%
%%
%%
%% For the documentation please refer to the file(s)
%% gmdoc.{gmd,pdf}.
%%
%%
\NeedsTeXFormat{LaTeX2e}
\ProvidesClass{gmdoc}
[2010/09/25 v0.993 a documenting package (GM)]
%%
%% ^^A(
%% \chschange{v0.98j}{2006/10/19}{4372}
%% \chschange{v0.98j}{06/10/19}{4372}
%% \begin{macrocode}
%%\end{macrocode}
%% \skiplines we skip the driver
%%\endskiplines
\ifnum\catcode`\@=12
\errorcontextlines=100
\documentclass[countalllines, codespacesgrey, outeroff, debug, mwrep,
pagella, trebuchet, cursor, fontspec=quiet]{gmdocc}
\verbLongDashes
\DoNotIndex{\gmu@tempa \gmu@tempb \gmu@tempc \gmu@tempd \gmu@tempe \gmu@tempf}
\twocoltoc
\title{The \pk{gmdoc} Package\\ i.e., \pk{gmdoc.sty} and
\pk{gmdocc.cls}}
\author{Grzegorz `Natror' Murzynowski}
\date{\ifcase\month\relax\or January\or February\or March\or April\or May\or
June\or July\or August\or September\or October\or November\or
December\fi\ \the\year}
\begin{document}
\emptify\udigits % because orig. def. raises an error `\cs{@secondoftwo} has
% an extra \}'
\maketitle
\setcounter{page}{2}% \pk{hyperref} cries if it sees two pages
\tableofcontents
\DoIndex\maketitle
\SelfInclude
\DocInclude{gmdocc}
\skipgmlonely[\stanza The remarks about installation and compiling
of the documentation are analogous to those in the chapter
\pk{gmdoc.sty} and therefore omitted.\stanza]
\DocInclude{gmutils}
\DocInclude{gmiflink}
\DocInclude{gmverb}
\DocInclude{gmoldcomm}
\typeout{%
Produce change log with^^J%
makeindex -r -s gmglo.ist -o \jobname.gls \jobname.glo^^J
(gmglo.ist should be put into some texmf/makeindex directory.)^^J}
\PrintChanges
\typeout{%
Produce index with^^J%
makeindex -r \jobname^^J}
\PrintIndex
\afterfi{%
\end{document}
makeindex -r gmdoc
makeindex -r -s gmglo.ist -o gmdocDoc.gls gmdocDoc.glo
}\fi% of |\ifnum\catcode`\@=12|, of the driver that is.
\catcode`\^^C=9\relax
\RequirePackage{gmutils}[2008/08/30]% includes redefinition of
\RequirePackage{xkeyval}% we need key-vals later, but maybe we'll make
\newif\if@linesnotnum
\DeclareOption{linesnotnum}{\@linesnotnumtrue}
\newif\if@uresetlinecount
\DeclareOption{uresetlinecount}{\@uresetlinecounttrue}
\newif\if@countalllines
\newif\if@printalllinenos
\DeclareOption{countalllines}{% to use the \incs{inputlineno}
% primitive and print real line numbers % in a~file.
\@countalllinestrue
\@printalllinenosfalse}
\DeclareOption{countalllines*}{%
\@countalllinestrue
\@printalllinenostrue}
\newif\if@noindex
\DeclareOption{noindex}{\@noindextrue}
\newif\if@pageindex
\DeclareOption{pageindex}{\@pageindextrue}
\newif\if@indexallmacros
\DeclareOption{indexallmacros}{\@indexallmacrostrue}
\@ifundefined{if@marginparsused}{\newif\if@marginparsused}{}
\@ifclassloaded{article}{\@marginparsusedtrue}{}
\@ifclassloaded{report}{\@marginparsusedtrue}{}
\@ifclassloaded{book}{\@marginparsusedtrue}{}
\DeclareOption{withmarginpar}{\@marginparsusedtrue}
\DeclareOption{nomarginpar}{\@marginparsusedfalse}
\DeclareOption{codespacesblank}{%
\AtEndOfPackage{% to allow \inverb|codespacesgrey, codespacesblank|
\AtBeginDocument{\CodeSpacesBlank}}}
\DeclareOption{codespacesgrey}{%
% \changes{v0.99l}{2008/08/06}{added due to Will Robertson's
% suggestion}
\AtEndOfPackage{% to put the declaration into the begin-document
% hook after definition of \incs{visiblespace}.
\AtBeginDocument{\CodeSpacesGrey}}}
\ProcessOptions
\RequirePackage{gmutils}[2008/08/08]
\RequirePackage{xcolor}
\definecolor{deepblue}{rgb}{0,0,.85}
\@ifpackageloaded{hyperref}{\hypersetup{colorlinks=true,
linkcolor=deepblue, urlcolor=blue, filecolor=blue}}{%
\RequirePackage[colorlinks=true, linkcolor=deepblue, urlcolor=blue,
filecolor=blue, pdfstartview=FitH, pdfview=FitBH,
pdfpagemode=UseNone]{hyperref}}
\RequirePackage{gmiflink}
\RequirePackage{gmverb}[2010/08/12]
\Store@Macros{\@verbatim\verb}
\if@noindex
\AtBeginDocument{\gag@index}% for the latter macro see line
% \ref{gag@index}.
\else
\RequirePackage{makeidx}\makeindex
\fi
\def\CodeDelim{\@bsphack\gmu@ifstar\Code@Delim@St\Code@Delim}
\def\Code@Delim@St#1{%
{\escapechar\m@ne
\@xa\gdef\@xa\code@delim\@xa{\string#1}}%
\@esphack}
\def\Code@Delim#1{\VerbHyphen{#1}\Code@Delim@St{#1}}
\CodeDelim\%
\pdef\narrationmark{{\codett\verbhyphen}{\normalfont\enspace}\ignorespaces}
\newtoks\gmd@preverypar
\newcommand*\settexcodehangi{%
\hangindent=\verbatimhangindent \hangafter=\@ne}% we'll use
\@ifdefinable\@@settexcodehangi{\let\@@settexcodehangi=\settexcodehangi}
\newlength\TextIndent
\newlength\CodeIndent
\CodeIndent=1,5em\relax
\@ifundefined{stanzaskip}{\newlength\stanzaskip}{}
\stanzaskip=\medskipamount
%^^A \advance\stanzaskip by-.25\medskipamount% to preserve the stretch- and
%^^A % shrinkability.\par
\newskip\CodeTopsep
\newskip\MacroTopsep
\def\UniformSkips{%\label{UniformSkips}
% \Define\CodeTopsep \Define\MacroTopsep
\CodeTopsep=\stanzaskip
\MacroTopsep=\stanzaskip
\abovedisplayskip=\stanzaskip
%\nostanza \leftskip0sp \gmdnoindent
% |%% \abovedisplayshortskip|
% remains untouched as it is 0.0\,pt plus 3.0\,pt by default.
% \nostanza
\belowdisplayskip=\stanzaskip
\belowdisplayshortskip=.5\stanzaskip% due to DEK's idea of making the
% short below display skip half of the normal.
\advance\belowdisplayshortskip by\smallskipamount
\advance\belowdisplayshortskip by-1\smallskipamount% We advance
% \incs{be\+low\+dis\+play\+short\+skip} forth and back to
% give it the \incs{small\+skip\+am\+ount}'s shrink- and
% stretchability components.
\topsep=\stanzaskip
\partopsep=\z@
}
\UniformSkips
\AtBeginDocument{\UniformSkips}
\newcommand*\NonUniformSkips{\@relaxen\UniformSkips}
\newcommand*\chunkskip{%
\par\addvspace{%
\glueexpr\MacroTopsep
\if@codeskipput-\CodeTopsep\fi
\relax
}\@codeskipputgtrue}
\pdef\stanza{%
\par\addvspace{%
\glueexpr\stanzaskip
\if@codeskipput-\CodeTopsep\fi
\relax}\@codeskipputgtrue}
\newcommand*\nostanza{% \changes{v0.99n}{2008/08/21}{added adding
% negative skip if in vmode and \cs{par}}
\par
\if@codeskipput\unless\if@nostanza\vskip-\CodeTopsep\relax\fi\fi
\@codeskipputgtrue\@nostanzagtrue
\@afternarrgfalse\@aftercodegtrue}% In the `code
\newgif\if@newline
\newgif\if@dsdir
\begingroup\catcode`\^^M=\active%
\firstofone{\endgroup%
\newcommand*{\DocInput}[1]{\begingroup%
% \changes{v0.98}{06/09/05}{\cs{@makeother\protect\bslash_} added}
% \DefIndex\gmd@inputname
\edef\gmd@inputname{#1}% we'll use it in some notifications.
%
\NamedInput@prepare{#1}% to make this input ���named���, as with |\NamedInput|.
%
% \DefIndex\gmd@currentlabel@before
\let\gmd@currentlabel@before=\@currentlabel% we store it because
% we'll do |\xdef|s of |\@currentlabel| to make proper references
% to the line numbers so we want to restore current
% |\@currentlabel| after our group.\
% \CodeUsgIndex\clubpenalty \CodeUsgIndex\widowpenalty
\gmd@setclubpenalty% we wrapped the assignment of |\clubpenalty|
% in a~macro because we'll repeat it twice more.
\@clubpenalty\clubpenalty \widowpenalty=3333 % Most paragraphs of the code will be
% one-line most probably and many of the narration, too.\par\
% \changes[\DocInput]{v0.95}{06/08/15}{\cs{club-}
% and \cs{widowpenalty} set both to 3333}
%\CodeUsgIndex\tolerance
\tolerance=1000 % as in \docfm.
% ^^A \catcode`\^^M=\active% redundant with line 2304
% \CodeUsgIndex\code@delim
\@xa\@makeother\csname\code@delim\endcsname%
% \CodeUsgIndex\gmd@resetlinecount
\gmd@resetlinecount% due to the option |uresetlinecount|
% we reset the line number counter or do nothing.
% \Define*{^^M}
\QueerEOL% \changes{v0.99m}{2008/08/09}{there was
% \cs{let}\hathat{M} but \cs{QueerEOL} is better: it also
% redefines \cs{}\hathat M} It has to be before the
% begin-input-hook to allow change by that hook.
% \CodeUsgIndex\@beginputhook
\@beginputhook% my first use of it is to redefine |\maketitle|
% just at this point not globally. \CodeUsgIndex\everypar
\everypar=\@xa{\@xa\@codetonarrskip\the\everypar}%
% \Define\gmd@guardedinput
\edef\gmd@guardedinput{%
\@nx\@@input #1\relax% |\@nx| is
% |\noexpand|, see \pk{gmutils}.\DoNotIndex\@nx
% \incs{@@input} is the
% true \TeX's \incs{input}. \CodeUsgIndex\EOFMark
\gmd@iihook% cf.\ line \ref{iihook}
\@nx\EOFMark% to pretty finish the input, see
% line~\ref{eofMark}.\CodeUsgIndex\code@delim
\@nx\CodeDelim\@xanxcs{\code@delim}% to
% ensure the code delimiter is the same as at the beginning of
% input.
% \changes[\DocInput]{v0.99c}{2007/03/02}{added ensuring the
% code delimiter to be the same at the end as at the beginning}
\@nx^^M\code@delim%
% \label{guardians}
}% we add guardians after
% |\input|ing a~file; somehow an error occurred without them.
\catcode`\%=9 % for \docfm -compatibility.\label{ignorePercent}
\setcounter{CheckSum}{0}% we initialise the counter for the number
% of the escape chars (the assignment is |\global|).
\everyeof{\relax}% |\@nx| moved not to spoil input of toc e.g.
\@xa\@xa\@xa^^M\gmd@guardedinput%\label{eeeEOL}
\par%
% \CodeUsgIndex\@endinputhook
\@endinputhook% It's a~hook to let postpone
% some stuff till the end of input. We use it e.g.\ for the
% \docfm-(not)likeliness notifications.
\glet\@currentlabel=\gmd@currentlabel@before% we restore
% value from before this group. In a~very special case this could
% cause unexpected behaviour of cross-refs, but anyway we acted
% globally and so acts \pk{hyperref}.
%
\NamedInput@finish% to clean up after a ���named��� input, as with |\NamedInput|.
%
\endgroup%
}% end of |\Doc@Input|'s definition.
}% end of |\firstofone|'s argument.
\pdef\gmd@textEOL{ % a~space just like in normal
% \TeX. We put it first to % cooperate with \inverb|\^^M|'s
% |\expandafter\ignorespaces|. It's % no % problem since a~space
% | |${}_{10}$ doesn't drive \TeX\ out of % the vmode.
\ifhmode\@afternarrgtrue\@codeskipputgfalse\fi% being in
% the horizontal mode means we've just typeset some narration so we
% turn the respective switches: the one bringing the message `we are
% after narration' to True (|@afternarr|) and the `we have put the
% code-narration glue' to False (|@codeskipput|). Since we are in
% a~verbatim group and the information should be brought outside it,
% we switch the switches globally (the letter |g| in both).
\@newlinegtrue% to |\refstep| the lines' counter at the proper
% point.
\@dsdirgtrue% to handle the \ds\ directives.
\@xa\@trimandstore\the\everypar\@trimandstore%we store the
% previous value of |\everypar| register to restore it at
% a~proper point. See line \ref{@trimandstore} for the details.
\begingroup%^^A ##### maybe we prefer to restore the genuine |\par|
% ^^A~before the group?
\gmd@setclubpenalty% Most paragraphs will be
% one-line most probably. Since some sectioning commands may change
% \incs{clubpenalty}, we set it again here and also after this
% group.
\aftergroup\gmd@setclubpenalty%
\let\par\@@par% inside the verbatim group we wish |\par| to be genuine.
% \CodeUsgIndex\ttverbatim
\let\verbatimfont\codett %
\ttverbatim% it applies the code-layer font (|\tt| by default) and makes specials
% other or |\active|-and-breakable.
% to turn verbatim specials off in
% \incmd\scanverb s.
\gmd@DoTeXCodeSpace%
\@makeother\|% because |\ttverbatim| doesn't do that.
\MakePrivateLetters% see line \ref{MPL}.\par
\@xa\@makeother\code@delim% we are
% almost sure the code comment char is among the chars having
% been \catother ed already. For `almost' see the
% \gmiflink[IndexInput]{\cs{IndexInput}} macro's definition.\par
% So, we've opened a~verbatim group and want to peek at the next
% character. If it's |%|, then we just continue narration, else we
% process the leading spaces supposed there are any and, if after
% them is a~|%|, we just continue the commentary as in the
% previous case or else we typeset the \TeX\ code.
\texcode@hook% we add some special stuff, e.g.\ in \pk{gmdocc.cls} we
% make star low.
\@xa\@ifnextcharRS\@xa{\code@delim}{%\label{ifContNarr}
% \CodeUsgIndex\gmd@continuenarration
\gmd@continuenarration}{% \CodeUsgIndex\gmd@dolspaces
\gmd@dolspaces% it will launch |\gmd@typesettexcode|.
}% end of |\@ifnextcharRS|'s else.
}% end of |\gmd@textEOL|'s definition.
\emptify\texcode@hook
\def\gmd@setclubpenalty{\clubpenalty=3333 }
\def\AtEndInput{\g@addto@macro\@endinputhook}
\def\@endinputhook{}
\def\AtBegInput{\g@addto@macro\@beginputhook}
\def\@beginputhook{}
\emptify\gmd@iihook
\AtBegInput{\let\gmd@@toc\tableofcontents
\def\tableofcontents{% \label{straighttoc}
\@ifQueerEOL
{\StraightEOL\gmd@@toc\QueerEOL}%
{\gmd@@toc}%
}%
}
\def\@ifEOLactive{%\
% \begin{enumargs}
% \item what if end of line is active,
% \item what if not.
% \end{enumargs}
\ifnum\catcode`\^^M=\active \@xa\@firstoftwo\else\@xa\@secondoftwo\fi}
\foone\obeylines{%
\def\@ifQueerEOL{%\
% \begin{enumargs}
% \item what if line end is ���queer���,
% \item what if not ���queer���.
% \end{enumargs}
% \changes{v0.98a}{06/09/06}{added}
\@ifEOLactive{%
\ifx^^M\gmd@textEOL\@xa\@firstoftwo\else\@xa\@secondoftwo\fi}%
{\@secondoftwo}}% of \cs{@ifQueerEOL}
}% of \cs{foone}
\pdef\qfootnote{%
\@ifQueerEOL
{\begingroup\StraightEOL\qfootnote@}%
{\footnote}}
\DeclareCommand\qfootnote@{o>Lm}{%
\endgroup % yes, we close the group: the arguments are already
% parsed and passed to this macro.
\edef\gmu@tempa{%
\@nx\footnote \IfValueT{#1}{[#1]}}%
\gmu@tempa{#2}%
}
\pdef\qemph{%
\@ifQueerEOL
{\begingroup\StraightEOL\qemph@}%
{\emph}}
\pdef\qemph@#1{\endgroup\emph{#1}}
\@emptify\gmd@ABIOnce
\AtEndOfPackage{\AtBegInput\gmd@ABIOnce}
\long\def\AtBegInputOnce#1{%
\gaddtomacro\gmd@ABIOnce{\g@emptify\gmd@ABIOnce#1}}
\def\EOFMark{\<eof>}
\begin{obeyspaces}%
\gdef\CodeSpacesVisible{%
\def\gmd@DoTeXCodeSpace{%
\obeyspaces\let =\breakablevisspace}}%
\gdef\CodeSpacesBlank{%
\let\gmd@DoTeXCodeSpace\gmobeyspaces%
\let\gmd@texcodespace=\ }% the latter |\let| is for the |\if������|s.
\gdef\CodeSpacesSmall{%
\def\gmd@DoTeXCodeSpace{%
\obeyspaces\def {\,\hskip\z@}}%
\def\gmd@texcodespace{\,\hskip\z@}}%
\end{obeyspaces}
\def\CodeSpacesGrey{%
% \changes{v0.99l}{2008/08/06}{added due to Will Robertson's
% suggestion}
\CodeSpacesVisible
\VisSpacesGrey% defined in \pk{gmverb}
}%
\CodeSpacesVisible
\def\gmd@continuenarration{%
\endgroup
\gmd@cpnarrline% see \gmiflink[countnarr]{below}.
\@xa\@trimandstore\the\everypar\@trimandstore
\everypar=\@xa{\@xa\@codetonarrskip\the\everypar}%
\@xa\gmd@checkifEOL\@gobble}
\if@countalllines
\def\gmd@countnarrline@{%
\gmd@grefstep{codelinenum}\@newlinegfalse
\everypar=\@xa{%
\@xa\@codetonarrskip\the\gmd@preverypar}%^^A
% the \cs{hy\+p\+er\+lab\+el\+@\+li\+ne} macro puts
% a~hypertarget in a~|\raise| i.e., drives \TeX\ into the
% horizontal mode so |\everypar| shall be issued. Therefore we
% should restore it.
}% of \cs{gmd@countnarrline@}
\def\gmd@grefstep#1{% instead of diligent redefining all possible
% commands and environments we just assign the current value of
% the respective \TeX's primitive to the \env{codelinenum}
% counter. Note we decrease it by $-1$ to get the proper value
% for the next line. (Well, I~don't quite know why, but it
% works.) %^^A~we'll step its value by 1 because it will be put
% ^^Aat the beginning of the \emph{next} line.
\ifnum\value{#1}<\inputlineno
\csname c@#1\endcsname\numexpr\inputlineno-1\relax
\ifvmode\leavevmode\fi% this line is added 2008/08/10 after an
% ^^A(
% all-night debuggery ;-) that showed that at one point
% \incs{gmd@grefstep} was called in vmode which caused adding
% \incs{penalty 10000} to the main vertical list and thus
% forbidding page break during entire \inenv{oldmc}.
\grefstepcounter{#1}%
\fi}% We wrap stepping the counter in an \incs{ifnum} to avoid
% repetition of the same ref-value (what would result in the
% ``multiply defined labels'' warning).\par\
% The
% \cs{grefstepcounter} macro, defined in \pk{gmverb}, is
% a~global
% version of \cs{ref\-step\-count\-er}, observing the redefinition
% made
% to \incs{refstepcounter} by \pk{hyperref}.
\if@printalllinenos% Note that checking this switch makes only
% sense when |countalllines| is true.
\def\gmd@cpnarrline{% count and print narration line
\if@newline
\gmd@countnarrline@
\hyperlabel@line
{\LineNumFont\thecodelinenum}\,\ignorespaces}%
\fi}
\else% not |printalllinenos|
\emptify\gmd@cpnarrline
\fi
\def\gmd@ctallsetup{% In the \inenv{oldmc} environments and with the
% \incs{FileInfo} declaration (when % \inenv{countalllines} option
% is in force) the % code is gobbled as an argument of a~macro and
% % then processed at one place (at the end of % \inenv{oldmc}
% % e.g.) so if we used % \incs{inputlineno}, we would have got all
% % the % lines with the same number. But we only set the counter
% % not \incs{refstep} it to avoid putting a~hypertarget.
\setcounter{codelinenum}{\inputlineno}% it's global.
\let\gmd@grefstep\hgrefstepcounter}
\else% not |countalllines| (and therefore we won't print the narration
% lines' numbers either)
\@emptify\gmd@cpnarrline
\let\gmd@grefstep\hgrefstepcounter% \label{let grefstep} if we don't
% want to count all the lines, we only \incs{ref}-increase the
% counter in the code layer.
\emptify\gmd@ctallsetup
\fi% of |\if@countalllines|
\def\skiplines{\bgroup
\let\do\@makeother \dospecials % not \incs{@sanitize} because the
% latter doesn't recatcode braces and we want all to be quieten.\ilrr
\gmd@skiplines}
\edef\gmu@tempa{%
\long\def\@nx\gmd@skiplines##1\bslash endskiplines{\egroup}}
\gmu@tempa
\foone\obeylines{% \DefIndex\gmd@typesettexcode
\def\gmd@typesettexcode{%
\gmd@parfixclosingspace% it's to eat a~space closing the
% paragraph, see \gmiflink[closingspace]{below}. It contains
% |\par|.
%
% ^^A\special{color push gray 0.2645}%
% A~verbatim group has already been opened by \cs{ttverb\+at\+im} and
% additional \cs{cat\-code}.
\everypar={\@@settexcodehangi}% At first attempt we thought
% of giving the user a~|\toks| list to insert at the beginning of
% every code line, but what for?
% \Define*{^^M}^^A
\def^^M{% \TeX\ code \acro{EOL}
\@newlinegtrue% to |\refstep| the counter in proper place.
\@dsdirgtrue% to handle the \ds\ directives.
\global\gmd@closingspacewd=\z@% \label{noclosingspace}we
% don't wish to eat a~closing space after a~codeline, because
% there isn't any and a~negative rigid |\hskip| added to
% |\parfillskip| would produce a~blank line.
\ifhmode\par\@codeskipputgfalse\else%
\if@codeskipput%
\else\addvspace{\stanzaskip}\@codeskipputgtrue%
\fi% if we've just met a~blank (code) line, we insert
% a~|\stanzaskip| glue.
% \label{codeskip}
\fi%
\prevhmodegfalse% we want to know later that now we are in the
% vmode.
% ^^A\special{color push gray 0.2666}%
\@ifnextcharRS{\gmd@texcodespace}{%
\@dsdirgfalse\gmd@dolspaces}{\gmd@charbychar}%
}% end of |^^M|'s definition.\label{debug!2}
% \DefIndex\gmd@texcodeEOL
\let\gmd@texcodeEOL=^^M% for further checks inside |\gmd@charbychar|.
\raggedright\leftskip=\CodeIndent%
\if@aftercode%
\gmd@nocodeskip1{iaC}%
\else%
\if@afternarr%
% ^^A~\def^^C{\showthe\hyphenpenalty\showthe\rightskip}
\if@codeskipput\else%
\gmd@codeskip1\@aftercodegfalse%
\fi% \label{codeskip3}
\else\gmd@nocodeskip1{naN}%
\fi%
\fi% if now we are
% switching from the narration into the code, we insert a~proper
% vertical space.
\@aftercodegtrue\@afternarrgfalse%
%^^A\special{color push gray 0.2682}% before that penalty
\ifdim\gmd@ldspaceswd>\z@% and here the leading spaces.
\leavevmode\@dsdirgfalse%
\if@newline\gmd@grefstep{codelinenum}\@newlinegfalse%
\fi%
\printlinenumber% if we don't want the lines to be numbered,
% the respective option \incs{let}s this \CS to \incs{relax}.
\hyperlabel@line%
%^^A\special{color push gray 0.2689}% seems O.K.
\mark@envir% index and/or marginize an environment if there is
% some to be done so, see line \ref{mark@envir}.
\hskip\gmd@ldspaceswd%
\advance\hangindent by\gmd@ldspaceswd%
\xdef\settexcodehangi{%
\@nx\hangindent=\the\hangindent% and also set the
% hanging indent setting for the same line comment case. \acro{BTW}.,
% this |%| or rather lack of it costed me five hours of
% debugging and rewriting. Active line ends require extreme
% caution.
\@nx\hangafter=1\space}%^^A~\@nx\relax
\else%
\glet\settexcodehangi=\@@settexcodehangi%\\
%|%| |\printlinenumber| here produced line numbers for blank lines
% which is what we don't want.
\fi% of |\ifdim|
\gmd@ldspaceswd=\z@%
\prevhmodegfalse% we have done |\par| so we are not in the
% hmode.
\@aftercodegtrue% we want to know later that now we are
% typesetting a~codeline.
\if@ilgroup\aftergroup\egroup\@ilgroupfalse\fi% when we are in the in-line
% comment group (for ragged right or justified), we want
% to \label{inline.egroup.2}
% close it. But if we did it
% here, we would close the verbatim group for the code. But we set
% the switch false not to repeat \inverb|\aftergroup\egroup|.
% ^^A\special{color push gray 0.2712}% before that penalty
\gmd@charbychar% we'll eat the code char by char to scan all the
% macros and thus to deal properly with the case |\%| in which the
% |%| will be scanned and won't
% launch closing of the verbatim group.
}% of |\gmd@typesettexcode|.
}% of |\foone\obeylines|.
\newlength\gmd@spacewd% to store the width of a~(leading)
\newlength\gmd@ldspaceswd% to store total length of gobbled leading
%spaces.
\let\gmd@texcodespace=\breakablevisspace
\def\gmd@dolspaces{%
\ifx\gmd@texcodespace\@let@token
\@dsdirgfalse
\afterfi{\settowidth{\gmd@spacewd}{\visiblespace}%
\gmd@ldspaceswd=\z@
\gmd@eatlspace}%
\else\afterfi{% about this smart macro and other of its family see
% \pk{gmutils} sec.\,3.
% ^^A \special{color push gray 0.2748}% debug of \penalty10000
% ^^A~2008/08/10 was far
% ^^A \ifnum\inputlineno>1016 \ifnum\inputlineno<1050 \show\par\fi\fi
% ^^A~O.K.
\if@afternarr\if@aftercode
\ifilrr\bgroup \gmd@setilrr\fi
\fi\fi
\par% possibly after narration
\if@afternarr\if@aftercode
\ifilrr\egroup\fi
\fi\fi
\gmd@typesettexcode}%
\fi}
\def\gmd@eatlspace#1{%
\ifx\gmd@texcodespace#1%
\advance\gmd@ldspaceswd by\gmd@spacewd% we don't
% \incs{advance} it \incs{global}ly because the current group may be closed
% iff we meet \inverb|%| and then we'll won't indent the line anyway.\ilrr
\afteriffifi\gmd@eatlspace
\else
\if\code@delim\@nx#1%
\gmd@ldspaceswd=\z@
\afterfifi{\gmd@continuenarration\narrationmark}%\label{ContNarr2}
% \changes{v0.99n}{2008/08/30}{\cs{afterfifi} added---a~bug fix}
\else \afterfifi{\gmd@typesettexcode#1}%
\fi
\fi}%
\newgif\ifprevhmode
\def\gmd@charbychar#1{%
\ifhmode\prevhmodegtrue
\else\prevhmodegfalse
% ^^A~\special{color push gray 0.2789}%
\fi
\if\code@delim\@nx#1%
\def\next{% occurs when next a~\cs{hskip4.875pt} is to be put
% ^^A\special{color push gray 0.2791}% O.K.
\gmd@percenthack% to typeset |%| if a~comment
% continues the~codeline.
\endgroup%
\gmd@checkifEOLmixd}% to see if next is |^^M| and then do |\par|.
\else% i.e., we've not met the code delimiter
\ifx\relax#1\def\next{%
% ^^A\special{color push gray 0.2798}%
\endgroup}% special case of end of file thanks to |\everyeof|.
\else
\if\code@escape@char\@nx#1%
\@dsdirgfalse% yes, just here not before the whole |\if| because
% then we would discard checking for \ds\ directives doable by
% the active |%| at the `old macrocode' setting.
\def\next{%
\gmd@counttheline#1\scan@macro}%
\else
\def\next{%
\gmd@EOLorcharbychar#1}%
\fi
\fi
\fi\next}
\def\debug@special#1{%
\ifhmode\special{color push gray 0.#1}%
\else\special{color push gray 0.#1000}\fi}
\def\gmd@EOLorcharbychar#1{%
%^^A \debug@special{2829}%
\ifx\gmd@texcodeEOL#1%
\if@newline
% ^^A\special{color push gray 0.281600}% no occurrence
% ^^A \if@countalllines\global\advance\c@codelinenum by\@ne obsolete
% ^^A\fi
\@newlinegfalse
\fi
\afterfi{#1}%\label{printhashone1}here we print |#1|.
\else% i.e., |#1| is \emph{not} a~(very active) line end,
\afterfi
{%^^A \debug@special{2839}% this occurs frequently
\gmd@counttheline#1\gmd@charbychar}% \label{printhashone2}or here
% we print |#1|. Here we would also possibly mark an environment
% but there's no need of it because declaring an environment to
% be marked requires a~bit of commentary and here we are after
% a~code |^^M| with no commentary.
\fi}
\def\gmd@counttheline{%
\ifvmode
\if@newline
\leavevmode
%^^A \debug@special{2851}%
\gmd@grefstep{codelinenum}\@newlinegfalse
\hyperlabel@line
\fi
%^^A\debug@special{2853}%
\printlinenumber
% ^^A\debug@special{2855}%
\mark@envir
\else% not vmode
\if@newline
% ^^A\special{color push gray 0.2842}% didn't occur
\gmd@grefstep{codelinenum}\@newlinegfalse
\hyperlabel@line
\fi
\fi}
\def\gmd@percenthack{%%
\ifprevhmode\aftergroup\narrationmark% We add a~space after |%|,
% because I~think it looks better. It's done |\aftergroup| to make
% the spaces possible after the |%| not to be
% typeset.
% \changes{v0.99n}{2008/08/21}{\cs{space} replaced with a~tilde to
% forbid a~line break before an in-line comment}
\else\aftergroup\gmd@dsNarrChecker% remember that
% \cs{gmd\-@\-pre\-cent\-hack} is only called when we've the code
% delimiter and soon we'll close the verbatim group and right after
% |\endgroup| there waits |\gmd@checkifEOLmixd|.
\fi}
\newif\ifgmd@dsVerb
\foone{\obeylines}%
{%
\def\gmd@dsVerbChecker%
#1% stuff for checking normal directive
#2% line contents
^^M{%%
\typeout{verb checker l.\the\inputlineno}%
\ifnum\strcmp{\detokenize{#2}}{\gmd@dsVerbDelim}=\z@%
\global\gmd@dsVerbfalse%
\def\gmd@modulehashone{%
\ModuleVerbClose{\gmd@dsVerbDelim}%
\global\emptify\gmd@dsVerbDelim%
\@afternarrgfalse\@aftercodegtrue%
\@codeskipputgfalse %
}%
\@xa\@firstoftwo%
\else \@xa\@secondoftwo %
\fi%
{\gmd@textEOL\gmd@modulehashone^^M}%
{\begingroup%
\endlinechar=\m@ne %
\@XA{%
\endgroup#1}\scantokens{#2}^^M% note that |\scantokens| adds
% char |\endlinechar| which we assure to be |^^M|
}%
}% of |\gmd@dsVerbChecker|
}% of |\obeylines|
\def\gmd@dsChecker#1{%
\@dsdirgfalse
\ifgmd@dsVerb
\@xa\@firstofone
\else
\@xa\@secondoftwo
\fi
{\gmd@dsVerbChecker}%
{#1}%
}% of |\gmd@dsChecker|
\def\gmd@dsNarrChecker#1{%%
\gmd@dsChecker
{\@ifnextcharRS<{%
\@xa\gmd@docstripdirective\@gobble}{#1}}%
}% of |\gmd@dsNarrChecker|
\def\gmd@checkifEOL{%
\gmd@cpnarrline
\everypar=\@xa{\@xa\@codetonarrskip% we add the
% macro that'll insert a~vertical space if we leave the code and
% enter the narration.
\the\gmd@preverypar}%
\@ifnextcharRS{\gmd@textEOL}{%
%^^A\@ifnextMac{%^^A}
\@dsdirgfalse
\par\ignorespaces}{%
\gmd@narrcheckifds}}%
\def\gmd@narrcheckifds{%^^A\typeout{narr if ds \on@line}%
\gmd@dsNarrChecker{\ignorespaces}}
\def\gmd@checkifEOLmixd{%
\gmd@cpnarrline
\everypar=\@xa{\@xa\@codetonarrskip\the\gmd@preverypar}%
%
% \label{longlinethatshouldbebroken}
\@afternarrgfalse\@aftercodegtrue
\ifhmode\@codeskipputgfalse\fi
\@ifnextcharRS{\gmd@textEOL}{%
%^^A\@ifnextMac{%^^A}
{\raggedright\gmd@endpe\par}% without \incs{raggedright} this
% \incs{par}
% would be justified which is not appropriate for a~long codeline
% that should be broken, e.g., \ref{longlinethatshouldbebroken}.
\prevhmodegfalse
\gmd@endpe\ignorespaces}{%
% If a~codeline ends with |%|
% (|prevhmode|${}=={}$True) first |\gmd@endpe| sets the parameters
% at the \TeX\ code values and |\par| closes a~paragraph and the
% latter |\gmd@endpe| sets the parameters at the narration values.
% In the other case both |\gmd@endpe|s do the same
% and |\par| between them does nothing.
% \DefIndex\par
\def\par{% the narration \cs{par}.
\ifhmode% (I~added this |\ifhmode| as a~result of a~heavy
% debug.)
\if@afternarr\if@aftercode
\unless\if@ilgroup\bgroup\@ilgrouptrue\fi
\ifilrr\gmd@setilrr\fi
\fi\fi
\@@par
\if@afternarr
\if@aftercode
\if@ilgroup\egroup\fi% \label{inline.egroup.1}if we are both after code
% and after narration it means we are after an in-line
% comment. Then we probably end a~group opened in line
% \ref{inline.bgroup}
\if@codeskipput\else\gmd@codeskip2\@aftercodegfalse\fi
%\label{codeskip4}
\else\gmd@nocodeskip2{naC}%
\fi
\else\gmd@nocodeskip2{naN}%
\fi
\prevhmodegfalse\gmd@endpe% when taken out of |\ifhmode|, this
% line caused some codeline numbers were typeset with
% |\leftskip|${}=0$.
\everypar=\@xa{%
\@xa\@codetonarrskip\the\gmd@preverypar}%
\let\par\@@par%
\fi}% of \cs{par}.
\gmd@endpe\ignorespaces}}
\def\gmd@endpe{%
\ifprevhmode
\settexcodehangi%\unskip ndent
\leftskip=\CodeIndent
\else
\leftskip=\TextIndent
\hangindent=\z@
\everypar=\@xa{%
\@xa\@codetonarrskip\the\gmd@preverypar}%
\fi}
\newif\ifilrr
\def\ilrr{%\changes{v0.99n}{2008/08/21}{added}
\if@aftercode
\unless\if@ilgroup\bgroup\@ilgrouptrue\fi% \label{inline.bgroup} If we are
% `aftercode', then we are in an in-line comment. Then we open
% a~group to be able to declare e.g.\ \cs{raggedright} for that
% comment only. This group is closed in line \ref{inline.egroup.1}
% or \ref{inline.egroup.2}.
\ilrrtrue
\fi}
\newif\if@ilgroup
\def\gmd@setilrr{\rightskip0ptplus\textwidth}
\def\ilju{% when in-line comments are ragged right in general but we
% want just this one to be justified.
\if@aftercode
\unless\if@ilgroup\bgroup\@ilgrouptrue\fi
\ilrrfalse
\fi}
\def\verbcodecorr{%\changes{v0.99n}{2008/08/21}{added}
% a~correction of vertical spaces between a~\env{verbatim} and
% code. We put also a~\cs{par} to allow parindent in the next
% commentary.
\vskip-\lastskip\vskip-4\CodeTopsep\vskip3\CodeTopsep\par}
\if@uresetlinecount% with |uresetlinecount| option\dots
\@relaxen\gmd@resetlinecount% \dots\ we turn
% resetting the counter by \cs{Doc\+In\+put} off\dots
\newcommand*\resetlinecountwith[1]{%
\newcounter{codelinenum}[#1]}% \dots\ and provide a~new
% declaration of the counter.
\else% With the option turned off\dots
\newcounter{DocInputsCount}%
\newcounter{codelinenum}[DocInputsCount]% \dots\ we declare the
% |\DocInput|s' number counter and the codeline counter
% to be reset with stepping of it.
% \changes[\c@DocInputsCount]{v0.98c}{06/9/8}{added for fixing
% duplication of \pk{hyperref} labels in the case of a~multiple
% \cs{DocInput}}
\newcommand*\gmd@resetlinecount{\stepcounter{DocInputsCount}}% \dots
% and let the |\DocInput| increment the |\DocInput|s number count
% and thus reset the codeline count. It's for unique naming of the
% \pk{hyperref} labels.
\fi
\newcommand*\printlinenumber{%
\leavevmode\llap{\rlap{\LineNumFont$\phantom{999}$\llap{\thecodelinenum}}%
\hskip\leftskip}}
\def\LineNumFont{\normalfont\tiny}
\if@linesnotnum\@relaxen\printlinenumber\fi
\newcommand*\hyperlabel@line{%
\if@pageindex% It's good to be able to switch it any time not just
% define it once according to the value of the switch set by the
% option.
\else
\raisebox{2ex}[1ex][\z@]{\gmhypertarget[clnum.%
\HLPrefix\arabic{codelinenum}]{}}%
\fi}
\newcommand*\gmd@codeskip[1]{%
\@@par\addvspace\CodeTopsep
\@codeskipputgtrue\@nostanzagfalse}
\newgif\if@codeskipput
\newgif\if@nostanza
\newcommand*\gmd@nocodeskip[2]{}
\if1 1
\renewcommand*\gmd@codeskip[1]{%
\hbox{\rule{1cm}{3pt} #1!!!}}
\renewcommand*\gmd@nocodeskip[2]{%
\hbox{\rule{1cm}{0.5pt} #1: #2 }}
\fi
\newgif\if@aftercode
\newgif\if@afternarr
\def\@codetonarrskip{%
\if@codeskipput\else
\if@afternarr\gmd@nocodeskip4{iaN}\else
\if@aftercode
% We are at the beginning of |\everypar|, i.e., \TeX\ has just entered
% the hmode and put the |\parindent| box. Let's remove it then.
{\setbox0=\lastbox}%
% Now we can put the vertical space and state we are not `aftercode'.
\gmd@codeskip4%
% \label{codeskip5}
\else\gmd@nocodeskip4{naC}%
\fi
\fi
\fi
\leftskip\TextIndent% this line is a~patch against
% a~bug-or-feature that in certain cases the narration |\leftskip|
% is left equal the code leftskip. (It happens when there are
% subsequent code lines after an in-line comment not ended with
% an explicit |\par|.) Before v0.99n it was just after line \ref{codeskip5}.
\@aftercodegfalse\@nostanzagtrue
% \changes{v0.99o}{2008/09/04}{a~bug fix: added \cs{@nostanzagtrue}}
}
\long\def\@trimandstore#1\@trimandstore{%
\def\@trimandstore@hash{#1}%
\ifx\@trimandstore@hash\@empty% we check if |#1| is
% nonempty. The \incs{if} \inverb*|\relax#1\relax| trick is not
% recommended here because using it we couldn't avoid expanding |#1|
% if it'd be expandable.
\gmd@preverypar={}%
\else
\afterfi{\@xa\@trimandstore@ne\the\everypar\@trimandstore}%
\fi}
\long\def\@trimandstore@ne#1#2\@trimandstore{%\label{@trimandstore}
\def\trimmed@everypar{#2}%
\ifx\@codetonarrskip#1%
\gmd@preverypar=\@xa{\trimmed@everypar}%
\else
\gmd@preverypar=\@xa{\the\everypar}%
\fi}
\catcode`\^^B=\active% note we re|\catcode| \<char2> globally, for the
\catcode`\^^V=\active % the same for |^^V|.
\catcode`\^^U=\active % and for |^^U|.
\foone{\obeylines}%% \Define*{^^B}
{\pdef\QueerCharTwo{%
\protected\def^^B##1^^M{%
%^^A\@newlinegtrue\gmd@countnarrline
\ifhmode\unskip\space\ignorespaces\fi}}% It shouldn't be \incs{ } not to
% drive \TeX\ into hmode.
%
% \Define*{^^V}
% \Define*{^^U}
% The |^^V| char is intended to mark parts of code commented out
% which are to be typeset verbatim. Lines are begun with narration
% marks (\%'s by default) and the narration-verbatim-typewriter font
% is used.
%
% The |^^U| char is intended for the lines commented out which are
% to be typeset as almost-invisible (second-class). They are marked
% with the narration mark as with |^^V| and a special font setting
% is used, which is a gray colour by default (in addition to the narration-verbatim-typewriter).
%
\pdef\gmd@UVdefs {%
\def\gmd@UV@percent{\global\let\verb@balance@group\@empty %
% ^^V \hyphenchar\font=\gmv@storedhyphenchar % it works back
% for the current paragraph so destroys our special hyphenchar.
\egroup\endgroup %
}% of |\gmd@UV@percent|
%
\@xa\def\@xa\verb@egroup@UV\@xa{%
\gmd@UV@percent ^^M%
}% of |\verb@egroup@UV|.
%
\addtomacro\gmd@UV@percent{\narrationmark}%
%
\pdef\gmd@QueerUV ##1{%
\scantokens\@xa{\code@delim %
\fooatletter{\@ifQueerEOL\@gobble}{}%
}% of |\scantokens|
%
\par%
{\codett\verbhyphen}\narrationmark %
\begingroup %
\catcode`\^^M=\active %
%
\let\verb@egroup=\verb@egroup@UV %
\verb^^M%
%
##1% nothing in |^^V| case and a gray setting in the |^^U| case.
%
\begingroup %
\@xa\lccode\@xa`\@xa~\@xa`\code@delim%
\lowercase{\endgroup\let~\gmd@UV@percent }%
\@xa\catcode\@xa`\code@delim\active %
}% of |\gmd@QueerUV|
}% of |\gmd@UVdefs|
%
%
\pdef\QueerU {%
\unless\ifdefined\gmd@QueerU%
%
\gmd@UVdefs %
\pdef\gmd@QueerU {\gmd@QueerUV {\QueerUFont }}%
\let^^U\gmd@QueerU%
\catcode`\^^U=\active%
\fi %
}% of |\QueerU|
%
\pdef\QueerV {%
\unless\ifdefined\gmd@QueerV %
%
\gmd@UVdefs %
\pdef\gmd@QueerV {\gmd@QueerUV {}}%
\let^^V\gmd@QueerV%
\catcode`\^^V=\active%
\fi %
}% of |\QueerV|
%
}% of |\foone|
\QueerCharTwo
\QueerV
\QueerU
\def\QueerUFont{\color{black!35}}
\AtBegInput{\@ifEOLactive{\catcode`\^^B\active}{}\QueerCharTwo}% \label{QCh2}
\catcode`\^^A=\active% note we re|\catcode| \<char1> globally, for the
\foone\obeylines{%%\DefIndex\QueerCharOne \Define*{^^A}
\def\QueerCharOne{%
\def^^A{%^^A no need to write \incs{gmd@countnarrline} because
% ^^A \inverb|^^M| will contain it if \inenv{countalllines} is in force.
\bgroup\let\do\@makeother\dospecials\gmd@gobbleuntilM}}%
\def\gmd@gobbleuntilM#1^^M{\egroup\ignorespaces^^M}%
}
\QueerCharOne
\AtBegInput{\@ifEOLactive{\catcode`\^^A\active}\QueerCharOne}% see note
\def\StraightEOL{%\label{StraightEOL}
\catcode`\^^M=5
\catcode`\^^A=14
\catcode`\^^B=14
\def\^^M{\ }}
\foone\obeylines{%
\def\QueerEOL{%\label{QueerEOL}
\catcode`\^^M=\active%
\let^^M\gmd@textEOL%
\catcode`\^^A=\active%
\catcode`\^^B=\active% I~only re|\catcode| \<char1> and \<char2>
% hoping no one but me is \emph{that} perverse to make them
% |\active| and (re)define. (Let me know if I'm wrong at this point.)
\let\^^M=\gmd@bslashEOL}%
}
\def\gmd@parfixclosingspace{{%
\advance\parfillskip by-\gmd@closingspacewd
\if@aftercode\ifilrr \gmd@setilrr \fi\fi
\par}%
\if@ilgroup\aftergroup\egroup\@ilgroupfalse\fi% we are in the
% verbatim group so we
% close the in-line comment group after it if the closing is not yet set.
}
\newskip\gmd@closingspacewd
\newcommand*\gmd@setclosingspacewd{%
\global\gmd@closingspacewd=\fontdimen2\font%
plus\fontdimen3\font minus\fontdimen4\font\relax}
\foone\obeylines{%\DefIndex*{\^^M}
\if 1 1%
\protected\def\gmd@bslashEOL{\ \@xa\ignorespaces^^M}%
}% of \cs{foone}. Note we interlace here \incs{if} with a~group.
\else%
\protected\def\gmd@bslashEOL{%
\ifhmode\unskip\fi\ \ignorespaces}
%^^A~\if@countalllines\@newlinegtrue\gmd@cpnarrline\fi%
\fi
\gaddtomacro\@verbatim{\leftskip=\CodeIndent}
\AtBegInput{\long\def\check@percent#1{%
\gmd@cpnarrline% to count the verbatim lines and possibly print
% their numbers. This macro is used only by the verbatim end of line.
\@xa\ifx\code@delim#1\else\afterfi{#1}\fi}}
\def\AddtoPrivateOthers#1{%
\@xa\def\@xa\doprivateothers\@xa{%
\doprivateothers\do#1}}%
\begingroup
\obeylines% \DefIndex\gm@verb@eol
\AtBegInput{\def\gm@verb@eol{\obeylines%
\def^^M{\verb@egroup\@latex@error{%
\@nx\verb ended by end of line}%
\@ifEOLactive{^^M}{\@ehc}}}}%
\endgroup
\def\verbatimfont{\narrativett}
\def\codett{\tt}
\pdef\texttt#1{{\narrativett#1}}
\VerbatimPitch
\def\ResultsIn{results in:}
\DeclareEnvironment{verbatim@p}{}
{\begingroup
\verbatim
}
{\endverbatim
\endgroup
\ResultsIn
\[\parbox{0,85\textwidth}{%
\newlinechar=\endlinechar
\StraightEOL
\scantokens\@xa{\VerbatimContents}%
}% of parbox
\]%
}
\edef\actualchar{\string @}
\edef\quotechar{\string "}
\edef\encapchar{\xiiclub}
\edef\levelchar{\string !}
\def\CodeEscapeChar#1{%
\begingroup
\escapechar\m@ne
\xdef\code@escape@char{\string#1}%
\endgroup}
\CodeEscapeChar\\
\@ifundefined{MakePrivateLetters}{%\label{MPL}
\def\MakePrivateLetters{\makeatletter\catcode`\*=11 }}{}
\def\MakePrivateOthers{\let\do=\@makeother \doprivateothers}
\def\doprivateothers{\do\ \do\^}
\foone\obeylines{%
\def\scan@macro#1{%
\ifx#1^^M\@xa#1\else\afterfi{\scan@macro@#1}\fi%
}% of |\scan@macro|,
}% of |\foone|.
\def\scan@macro@#1{% we are sure to scan at least one token which is
% not the line end and
% therefore we define this macro as one-parameter.\par
\step@checksum% \label{checksumUse}(see line \ref{checksum} for
% details),\par
% Then, unlike in \docfm , we do \emph{not} check if the scanning is
% allowed, because here it's always allowed and required.\par
% Of course, I~can imagine horrible perversities, but I~don't think
% they should really be taken into account. Giving the letter |a|
% |\catcode| other than \catletter\ surely would be one of those
% perversities. Therefore I~feel safe to take the character |a| as
% a~benchmark letter.
\ifcat a\@nx#1%
\quote@char#1%
\xdef\macro@iname{\gmd@maybequote#1}% global for symmetry with
% line \ref{x474}.
\xdef\macro@pname{\string#1}%\label{stringing0} we'll print entire
% name of the macro later.\par
% We |\string| it here and in the lines \ref{stringing1} and
% \ref{stringing2} to be sure it is whole \catother\ for easy
% testing for special index entry formats, see line
% \ref{pnametestDef} etc. Here we are sure the result of |\string|
% is \catother\ since its argument is \catletter.
\afterfi{\@ifnextcat{a}{\gmd@finishifstar#1}{\finish@macroscan}}%
\else% |#1| is not a~letter, so we have just scanned a~one-char
% \CS.\par
% Another reasonable |\catcode|s assumption seems to be that the
% digits are \catother. Then we don't have to
% type (|%|)|\expandafter\@gobble\string\a|. We do the |\uccode|
% trick to be sure that the char we write as the macro's name is
% \catother.
{\uccode`9=`#1%
\uppercase{\xdef\macro@iname{9}}%\label{x474}
}%
\quote@char#1%
\xdef\macro@iname{\gmd@maybequote\macro@iname}%
\xdef\macro@pname{\xiistring#1}%\label{stringing1}
\afterfi \finish@macroscan
\fi}% of |\scan@macro@|.
\def\continue@macroscan#1{%
\quote@char#1%
\xdef\macro@iname{\macro@iname \gmd@maybequote#1}%
\xdef\macro@pname{\macro@pname \string#1}%\label{stringing2} we know
% \inverb*|#1| to be \catletter, so
% we don't need \incs{xiistring}.
\@ifnextcat{a}{\gmd@finishifstar#1}{\finish@macroscan}%
}
\def\gmd@finishifstar#1{%
\if*\@nx#1\afterfi\finish@macroscan% note we protect |#1| against
% expansion. In \pk{gmdoc} verbatim scopes some chars are active
% (e.g.\ \inverb|\|\,).
\else\afterfi\continue@macroscan
\fi}
\def\quote@char#1{{\uccode`9=`#1% at first I~took digit 1 for this
% |\uccode|ing but then |#1| meant |#|\<\#1> in |\uppercase|'s
% argument, of course.
\uppercase{% \DefIndex\gmd@maybequote
\@ifinmeaning 9\of \indexcontrols
{\glet\gmd@maybequote\quotechar}%
{\g@emptify\gmd@maybequote}%
}%
}}
\AtBeginDocument{\xdef\indexcontrols{%
\bslash\levelchar\encapchar\actualchar\quotechar}}
\newif\ifgmd@glosscs% we use this switch to keep the information
\newcommand*\finish@macroscan{%\label{506}\par
% First we check if the current \CS is not just being defined. The
% switch may be set true in line \ref{519}
\ifgmd@adef@cshook% if so, we throw it into marginpar and index as
% a~def entry\dots
\gmu@ifundefined{gmd/iexcl/\macro@pname\space}{% \dots\ if it's not excluded
% from indexing.
\@xa\Code@MarginizeMacro\@xa{\macro@pname}%
\@xa\@defentryze\@xa{\macro@pname}{1}}{}%% here we declare the kind of
% index entry and define |\last@defmark| used by \cs{changes}
\global\gmd@adef@cshookfalse% we falsify the hook that was set
% true just for this \CS.
\fi
% We have the \CS's name for indexing in |\macro@iname| and
% for print in |\macro@pname|. So we index it. We do it a~bit
% counter-crank way because we wish to use more general indexing
% macro.
\if\verbatimchar\macro@pname% \label{3039}it's important that |\verbatimchar|
% comes before the macro's name: when it was reverse, the |\tt| \CS
% turned this test true and left the |\verbatimchar| what resulted
% with `|\+tt|' typeset. Note that this test should turn true iff
% the scanned macro name shows to be the default
% |\verb|'s delimiter. In such a~case we give
% |\verb| another delimiter, namely |$|: ^^A$
\def\im@firstpar{[$%^^A$
]}%
\else\def\im@firstpar{}%
\fi
\@xa \index@macro\im@firstpar\macro@iname\macro@pname
% \label{3049}
\maybe@marginpar\macro@pname
\if\xiispace\macro@pname\relax\gmd@texcodespace
\else
{\noverbatimspecials\Restore@Macro\verb
\@xa\scanverb\@xa{\macro@pname}}% we typeset scanned \CS.
\fi
% \changes[\finish@macroscan]{v0.99n}{2008/09/30}{the case of
% \cs{\vs} taken care of}
\let\next\gmd@charbychar
\gmd@detectors% \label{519} for automatic detection of
% definitions. Defined and
% explained in the next section. It redefines
% |\next| if detects a~definition command and thus
% sets the switch of line \ref{506} true.
\next
% \label{next 3690}
}
\def\maybe@marginpar#1{%
\gmu@ifundefined{gmd/2marpar/\@xa\detokenize\@xa{#1}}{}{%
\edef\gmu@tempa{%
\unexpanded{\Text@Marginize*}%
{\bslash\@xa\unexpanded\@xa{#1}}% ^^A
}\gmu@tempa
% ^^A % |\expandafter|s
% ^^A % \possfil because the |\Text@Marginize| command applies |\string| to its
% ^^A % argument.
% \incs{macro@pname}, which will be the only possible
% argument to
% \incs{may\+be\+@mar\+g\+in\+par},
% contains the macro's name
% without the escape char so we added it here.
\@xa\g@relaxen
\csname gmd/2marpar/\@xa\detokenize\@xa{#1}\endcsname% we reset the switch.
}}
\outer\def\DeclareDefining{\begingroup
\MakePrivateLetters
\gmu@ifstar
{\gdef\gmd@adef@defaulttype{text}\Declare@Dfng}%
{\gdef\gmd@adef@defaulttype{cs}\Declare@Dfng}%
}
\newcommand*\Declare@Dfng[2][]{%
\endgroup
\Declare@Dfng@inner{#1}{#2}%
\ifgmd@adef@star% this switch may be set false in first
% \incs{Declare@Dfng@inner} (it's the |star| key).
\Declare@Dfng@inner{#1}{#2*}% The catcode of |*| doesn't matter since
% it's in
% \incs{csname\+���\+\bslash end\+cs\+na\+me}
% everywhere.
\fi}
\def\Declare@Dfng@inner#1#2{%
\edef\gmd@resa{%
\@nx\setkeys[gmd]{adef}{type=\gmd@adef@defaulttype}}%
\gmd@resa
{\escapechar\m@ne
\xdef\gmd@adef@currdef{\string#2}%
% ^^A~\typeout{@@@ gmd@adef@currdef:::\gmd@adef@currdef::::}%
}%
\gmd@adef@setkeysdefault
\setkeys[gmd]{adef}{#1}%
\@xa\@ifinmeaning
\csname gmd@detect@\gmd@adef@currdef\endcsname
% \label{550}
\of\gmd@detectors{}{%
\@xa\gaddtomacro\@xa\gmd@detectors\@xa{%
\csname gmd@detect@\gmd@adef@currdef\endcsname}}% we add a~\CS\\
% |%| |\gmd@detect@|\<def name> (a~\textbf{detector}) to the
% meaning of the \textbf{detectors' carrier}. And we define it to
% detect the \inverb|#2| command.
\@xa\xdef\csname gmd@detectname@\gmd@adef@currdef\endcsname{%
\gmd@adef@currdef}%
\edef\gmu@tempa{% this |\edef| is to expand |\gmd@adef@TYPE|.
\global\@nx\@namedef{gmd@detect@\gmd@adef@currdef}{%
\@nx\ifx
\@xanxcs{gmd@detectname@\gmd@adef@currdef}%
\@nx\macro@pname
\@nx\n@melet{next}{gmd@adef@\gmd@adef@TYPE}%
\@nx\n@melet{gmd@adef@currdef}{gmd@detectname@\gmd@adef@currdef}%
\@nx\fi}}%
\gmu@tempa
\SMglobal\Store@MacroSt {gmd@detect@\gmd@adef@currdef}% we store the \CS to
% allow its temporary discarding later.
}
\def\gmd@adef@setkeysdefault{%
\setkeys[gmd]{adef}{star,prefix,KVpref}}
\define@boolkey[gmd]{adef}{star}[true]{}
\define@key[gmd]{adef}{prefix}[]{%
\edef\gmd@resa{%
\def\@xanxcs{gmd@adef@prefix@\gmd@adef@currdef }{%
#1}}%
\gmd@resa}
\def\gmd@KVprefdefault{KV}% in a~separate macro because we'll need
% it in \cs{ifx}.
\define@key[gmd]{adef}{KVpref}[\gmd@KVprefdefault]{%
\edef\gmd@resa{#1}%
\ifx\gmd@resa\gmd@KVprefdefault
\else
\@namedef{gmd@adef@KVprefixset@\gmd@adef@currdef}{1}%
\gmd@adef@setKV% whenever the |KVpref|fix is set (not default), the
% declared command is assumed to be \pk{keyval}ish.
\fi
\edef\gmd@resa{#1}% because |\gmd@adef@setKV| redefined it.
\edef\gmd@resa{%
\def\@xanxcs{gmd@adef@KVpref@\gmd@adef@currdef}{%
\ifx\gmd@resa\empty
\else#1@\fi}}% as in \pk{xkeyval}, if the \acro{KV} prefix is not
% empty, we add \inverb|@| to it.
\gmd@resa}
\define@key[gmd]{adef}{KVfam}[]{%
\edef\gmd@resa{#1}%
\@namedef{gmd@adef@KVfamset@\gmd@adef@currdef}{1}%
\edef\gmd@resa{%
\def\@xanxcs{gmd@adef@KVfam@\gmd@adef@currdef}{%
\ifx\gmd@resa\empty
\else#1@\fi}}%
\gmd@resa
\gmd@adef@setKV}% whenever the |KVfam|ily is set, the declared command is
% assumed to be \pk{keyval}ish.
\define@choicekey[gmd]{adef}{type}
[\gmd@adef@typevals\gmd@adef@typenr]
{% the list of possible types of defining commands
def,
newcommand,
cs,% equivalent to the two above, covers all the cases of defining
% a~\CS, including the \PlainTeX\ \inverb|\new������| and
% \LaTeX\ |\newlength|.
newenvironment,
text,% equivalent to the one above, covers all the commands defining
% its first mandatory argument that should be text,
% \inverb|\DeclareOption| e.g.
define@key,% special case of more arguments important; covers the
% \pk{xkeyval} defining commands.
dk,% a~shorthand for the one above.
DeclareOptionX,% another case of special arguments configuration,
% covers the \pk{xkeyval} homonym.
dox,% a~shorthand for the one above.
kvo% one of option defining commands of the \pk{kvoptions} package
% by Heiko Oberdiek (a~package available o~\acro{CTAN} in the
% \pk{oberdiek} bundle).
}
{% In fact we collapse all the types just to four so far:
\ifcase\gmd@adef@typenr% if |def|
\gmd@adef@settype{cs}{0}%
\or% when |newcommand|
\gmd@adef@settype{cs}{0}%
\or% when |cs|
\gmd@adef@settype{cs}{0}%
\or% when |newenvironment|
\gmd@adef@settype{text}{0}%
\or% when |text|
\gmd@adef@settype{text}{0}%
\or% when |define@key|
\gmd@adef@settype{dk}{1}%
\or% when |dk|
\gmd@adef@settype{dk}{1}%
\or% when |DeclareOptionX|
\gmd@adef@settype{dox}{1}%
\or% when |dox|
\gmd@adef@settype{dox}{1}%
\or% when |kvo|
\gmd@adef@settype{text}{1}%% The \pk{kvoptions} option
%% definitions take first mandatory
% argument as the option name and they define a~\pk{keyval} key
% whose macro's name begins with the prefix/family, either default or
% explicitly declared. The \pk{kvoptions} prefix/family is
% supported in \pk{gmdoc} with \inverb|[KVpref=, KVfam=|\<family>|]|.
\fi}
\def\gmd@adef@settype#1#2{%
\def\gmd@adef@TYPE{#1}%
\ifnum1=#2 % now we define (or not) a~quasi-switch that fires for
% the \pk{keyval}ish definition commands.
\gmd@adef@setKV
\fi}
\def\gmd@adef@setKV{%
\edef\gmd@resa{%
\def\@xanxcs{gmd@adef@KV@\gmd@adef@currdef}{1}%
}%
\gmd@resa}
\emptify\gmd@detectors
\newif\ifgmd@adef@cshook
\def\gmd@adef@cs{\global\gmd@adef@cshooktrue\gmd@charbychar}
\def\gmd@adef@text{\gdef\gmd@lbracecase{1}\gmd@charbychar}
\foone{%
\catcode`\[\active
% ^^A>\]
\catcode`\<\active}
{%\par
% The detector of \pk{xkeyval} |\define@�����[���]��key|:
\def\gmd@adef@dk{%
\let[\gmd@adef@scanKVpref
\catcode`\[\active
% ^^A\]]
\gdef\gmd@lbracecase{2}%
\gmd@adef@dfKVpref\gmd@KVprefdefault% We set the default value of
% the \pk{xkeyval} prefix. Each time again because an assignment
% in \inverb|\gmd@adef@dfKVpref| is global.\ilrr
\gmd@adef@checklbracket}
% The detector of \pk{xkeyval} |\DeclareOptionX|:
\def\gmd@adef@dox{%
\let[\gmd@adef@scanKVpref
\let<\gmd@adef@scanDOXfam
\catcode`[\active
% ^^A]]
\catcode`<\active
\gdef\gmd@lbracecase{1}%
\gmd@adef@dfKVpref\gmd@KVprefdefault% We set the default values of
% the \pk{xkeyval} prefix\dots
\edef\gmd@adef@fam{\gmd@inputname}% \dots\ and family.
\gmd@adef@dofam
% \label{defDOXfam}
\gmd@adef@checkDOXopts}%
}
\def\gmd@adef@checklbracket{%
\@ifnextchar[%^^A]
\gmd@adef@scanKVpref\gmd@charbychar}% note that
\def\gmd@adef@checkDOXopts{%
\@ifnextchar[\gmd@adef@scanKVpref%^^A]
{\@ifnextchar<\gmd@adef@scanDOXfam\gmd@charbychar}}
\def\gmd@adef@scanKVpref#1#2]{%
\gmd@adef@dfKVpref{#2}%
[#2]\gmd@charbychar}
\def\gmd@adef@dfKVpref#1{%
\ifnum1=0\csname gmd@adef@KVprefixset@\gmd@adef@currdef\endcsname
\relax
\else
\edef\gmu@resa{%
\gdef\@xa\@nx
\csname gmd@adef@KVpref@\gmd@adef@currdef\endcsname{%
\ifx\relax#1\relax
\else#1@%
\fi}}%
\gmu@resa
\fi}
\def\gmd@adef@scanDOXfam{%
\ifnum12=\catcode`\>\relax
\let\next\gmd@adef@scanfamoth
\else
\ifnum13=\catcode`\>\relax
\let\next\gmd@adef@scanfamact
\else
\PackageError{gmdoc}{> neither `other' nor `active'! Make it
`other' with \bslash AddtoPrivateOthers\bslash\>.}{}%
\fi
\fi
\next}
\def\gmd@adef@scanfamoth#1>{%
\edef\gmd@adef@fam{\@gobble#1}% there is always
% \cs{gmd@charbychar} first.
\gmd@adef@dofam
<\gmd@adef@fam>%
\gmd@charbychar}
\foone{\catcode`\>\active}
{\def\gmd@adef@scanfamact#1>{%
\edef\gmd@adef@fam{\@gobble#1}% there is always
% \cs{gmd@charbychar} first.
\gmd@adef@dofam
<\gmd@adef@fam>%
\gmd@charbychar}%
}
\def\gm@lbracehook{%
\ifcase\gmd@lbracecase\relax
\or% when 1
\afterfi{%
\gdef\gmd@lbracecase{0}%
\gmd@adef@scanname}%
\or% when 2---the first mandatory argument of two (|\define@�����[���]��key|)
\afterfi{%
\gdef\gmd@lbracecase{3}%
\gmd@adef@scanDKfam}%
\or% when 3---the second mandatory argument of two (the key name).
\afterfi{%
\gdef\gmd@lbracecase{0}%
\gmd@adef@scanname}%
\fi}
\def\gmd@lbracecase{0}% we initialise the hook caser.
\foone{\catcode`\[1 \catcode`\]2 \catcode`\}12 }
[% Note that till line \ref{738} the square brackets are grouping
% and the right brace is `other'. ^^A{
\def\gmd@adef@scanDKfam#1}[%^^A{
\edef\gmd@adef@fam[\@gobble#1]% there is always
% \cs{gmd@charbychar} first.
\gmd@adef@dofam
\gmd@adef@fam}%
\gmd@charbychar]
% ^^A{
\def\gmd@adef@scanname#1}[%^^A{
\@makeother\[%^^A\]
\@makeother\<%
% The scanned name begins with |\gmd@charbychar|, we have to be
% careful.
\gmd@adef@deftext[#1]%
\@gobble#1}%
\gmd@charbychar]
]
\def\gmd@adef@dofam{%
\ifnum1=0\csname gmd@adef@KVfamset@\gmd@adef@currdef\endcsname
\relax% a~family declared with |\DeclareDefining| overrides the
% one currently scanned.
\else
\edef\gmu@resa{%
\gdef\@xa\@nx
\csname gmd@adef@KVfam@\gmd@adef@currdef\endcsname
{\ifx\gmd@adef@fam\empty
\else\gmd@adef@fam @%
\fi}}%
\gmu@resa
\fi}
\def\gmd@adef@deftext#1{%
\@xa\def\@xa\macro@pname\@xa{\@gobble#1}%% we gobble |\gmd@charbychar|, cf. above.
\edef\macro@pname{\@xa\detokenize\@xa{\macro@pname} }% note the
% space at the end.
\edef\macro@pname{\@xa\@xiispaces\macro@pname\@nil}%
\@xa\Text@Marginize\@xa{\macro@pname}%
\gmd@adef@indextext
\edef\gmd@adef@altindex{%
\csname gmd@adef@prefix@\gmd@adef@currdef \endcsname}%
%\hskip-\parindent and we add the \pk{xkeyval} header if we are in \pk{xkeyval}
% definition.
\ifnum1=0\csname gmd@adef@KV@\gmd@adef@currdef \endcsname\relax% The\\ \CS
% \inverb|\gmd@adef@KV@|\<def. command> is defined |{1}| (so \cs{ifnum}
% gets |1=01\relax|---\hskip0sptrue) iff \<def. command> is a~\pk{keyval}
% definition. In that case we check for the \inverb|KVpref|ix and
% \inverb|KVfam|ily. (Otherwise |\gmd@adef@KV@|\<def. command> is undefined
% so \cs{ifnum} gets |1=0\relax|---false.)\ilrr
\edef\gmd@adef@altindex{%
\gmd@adef@altindex
\csname gmd@adef@KVpref@\gmd@adef@currdef \endcsname}%
\edef\gmd@adef@altindex{%
\gmd@adef@altindex
\csname gmd@adef@KVfam@\gmd@adef@currdef \endcsname}%
\fi
\ifx\gmd@adef@altindex\empty
\else% we make another index entry of the definiendum with prefix/KVheader.
\edef\macro@pname{\gmd@adef@altindex\macro@pname}%
\gmd@adef@indextext
\fi}
\def\gmd@adef@indextext{%
\@xa\@defentryze\@xa{\macro@pname}{0}% declare the definiendum has to
% have a~definition entry and should appear
% without backslash in the changes history.
\gmd@doindexingtext% redefine |\do| to an indexing macro.
\@xa\do\@xa{\macro@pname}}
\DeclareDefining[star=false]\def
\DeclareDefining[star=false]\pdef% it's a~\pk{gmutils}' shorthand for \inverb|\protected\def|.
\DeclareDefining[star=false]\provide% a~\pk{gmutils}' conditional \incs{def}.
\DeclareDefining[star=false]\pprovide% a~\pk{gmutils}' conditional \incs{pdef}.
\def\UnDef{{% \changes{v0.99n}{2008/08/30}{a~bug fixed:
% \cs{gmd@charbychar} appended to \cs{next}---without it
% a~subsequent in-line comment was typeset verbatim}
% \UnDef
\gmd@adef@selfrestore\def
}}
\def\UnPdef{{\gmd@adef@selfrestore\pdef}}
\Store@Macro\UnDef% because the `hiding' commands relax it.
\def\HideDef{% \changes{v0.99n}{2008/08/30}{added the starred version
% that calls \cs{UnDef}}
\gmu@ifstar\UnDef{\HideDefining\def\relaxen\UnDef}}
\def\ResumeDef{%
\ResumeDefining\def
\Restore@Macro\UnDef}
\DeclareDefining[star=false]\newcount
\DeclareDefining[star=false]\newdimen
\DeclareDefining[star=false]\newskip
\DeclareDefining[star=false]\newif
\DeclareDefining[star=false]\newtoks
\DeclareDefining[star=false]\newbox
\DeclareDefining[star=false]\newread
\DeclareDefining[star=false]\newwrite
\DeclareDefining[star=false]\newlength
\DeclareDefining[star=false]\DeclareDocumentCommand
\DeclareDefining[star=false]\DeclareCommand
\DeclareDefining\newcommand
\DeclareDefining\renewcommand
\DeclareDefining\providecommand
\DeclareDefining\DeclareRobustCommand
\DeclareDefining\DeclareTextCommand
\DeclareDefining\DeclareTextCommandDefault
\DeclareDefining*\newenvironment
\DeclareDefining*\renewenvironment
\DeclareDefining*[star=false]\DeclareOption
\DeclareDefining*[prefix=\bslash c@]\newcounter% \label{newcounter}
\DeclareDefining[type=dk, prefix=\bslash]\define@key
\DeclareDefining[type=dk, prefix=\bslash if]\define@boolkey% the
% alternate index entry will be
% \cs{if}\<KVpref>|@|\<KVfam>|@|\<key name>
\DeclareDefining[type=dk, prefix=\bslash]\define@choicekey
\DeclareDefining[type=dox, prefix=\bslash]\DeclareOptionX% the
\newcommand*\DeclareDOXHead[2][\gmd@KVprefdefault]{%
\csname DeclareDefining\endcsname
[type=dox, prefix=\bslash, KVpref=#1, KVfam=#2]% \HideDefining\DeclareOptionX
\DeclareOptionX
}
\iffalse
\DeclareOptionX[Berg]<Lulu>{EvelynLear}{}
\DeclareOptionX<AntonW>{ChneOelze}
\fi
\DeclareDefining[type=kvo, prefix=\bslash, KVpref=]\DeclareStringOption
\DeclareDefining[type=kvo, prefix=\bslash, KVpref=]\DeclareBoolOption
\DeclareDefining[type=kvo, prefix=\bslash, KVpref=]\DeclareComplementaryOption
\DeclareDefining[type=kvo, prefix=\bslash, KVpref=]\DeclareVoidOption
\def\DeclareKVOFam#1{%
\def\do##1{%
\csname DeclareDefining\endcsname
[type=kvo, prefix=\bslash, KVpref=, KVfam=#1]##1}%
\do\DeclareStringOption
\do\DeclareBoolOption
\do\DeclareComplementaryOption
\do\DeclareVoidOption
}
\def\HideAllDefining{%
\ifnum0=0\csname gmd@adef@allstored\endcsname
\SMglobal\Store@Macro\gmd@detectors
\global\@namedef{gmd@adef@allstored}{1}%
\fi
\global\emptify\gmd@detectors}% we make the carrier |\empty| not |\relax| to
% be able to declare new defining command in
% the scope of |\HideAll������|
\def\ResumeAllDefining{%
\ifnum1=0\csname gmd@adef@allstored\endcsname\relax
\SMglobal\Restore@Macro\gmd@detectors
\SMglobal\Restore@Macro\UnDef
\global\@namedef{gmd@adef@allstored}{0}%
\fi}
\def\HideDefining{\begingroup
% \changes{v0.99n}{2008/08/30}{Added the starred version that hides
% the defining command only once}
\MakePrivateLetters
\gmu@ifstar\Hide@DfngOnce\Hide@Dfng}
\def\Hide@Dfng#1{%%\UnDef
\escapechar\m@ne
\gn@melet{gmd@detect@\string#1}{relax}%
\gn@melet{gmd@detect@\string#1*}{relax}%
\ifx\def#1\global\relaxen\UnDef\fi
\endgroup}
\def\Hide@DfngOnce#1{%%\UnDef
\gmd@adef@selfrestore#1%
\endgroup}
\def\gmd@adef@selfrestore#1{%
\@ifundefined{gmd@detect@\strip@bslash{#1}}{%
\SMglobal\@xa\Store@Macro
\csname gmd@detect@\strip@bslash{#1}\endcsname}{}%
% ^^A\typeout{:::::::::::::::gmd@detect@\string#1:::::::::::::}%
\global\@nameedef{gmd@detect@\strip@bslash{#1}}{%
\@nx\gmu@if x%
{\@xanxcs{gmd@detectname@\strip@bslash{#1}}%
\@nx\macro@pname}% we compare the detect(ed) name with
% |\macro@pname|.
%
{\def\@nx\next{% this \incs{next} will be executed in line
% % \ref{next 3690}.
\SMglobal\Restore@Macro % they both are \incs{protected}.
\@xanxcs{gmd@detect@\string#1}%
\@nx\gmd@charbychar}%^^A we define
% ^^A \inverb|\next| not restore the macro here just in case of
% ^^A multiple declaration of |\def| (in that case there would
% ^^A~be
% ^^A multiple occurrences of the macro to be restored in the
% ^^A~carrier
% ^^A \inverb|\gmd@detectors| and we wish
% ^^A all of them not to fire this one time) That's probably too
% ^^A much care since there's the test in line \ref{550}.
\@nx}%
{}% or do nothing if the \CS' names are unequal.
}% of \cs{@nameedef}.
}% of \cs{gmd@adef@selfrestore}.
\def\ResumeDefining{\begingroup
\MakePrivateLetters
\gmd@ResumeDfng}
\def\gmd@ResumeDfng#1{%
\escapechar\m@ne
\SMglobal\Restore@MacroSt{gmd@detect@\string#1}%
\SMglobal\Restore@MacroSt{gmd@detect@\string#1*}%
\endgroup}
\newcommand*\index@macro[3][\verbatimchar]{{%
\gmu@ifundefined{gmd/iexcl/\@xa\detokenize\@xa{#3 }}%\label{iexcltest}
{% |#3| is not excluded from index
% ^^A\typeout{@@@ :::\@xa\string\csname gmd/iexcl/\@xa\detokenize\@xa{#3 }\endcsname::::}%
\gmu@ifundefined{gmd/defentry/\@xa\detokenize\@xa{#3 }}% \label{pnametestDef}
{% |#3| is not def entry
\gmu@ifundefined{gmd/usgentry/\@xa\detokenize\@xa{#3 }}%\label{pnametestUsg}
{% |#3| is not usg.\ entry
\edef\kind@fentry{\CommonEntryCmd}}%\label{CECmd}
{% |#3| is usg.\ entry
\def\kind@fentry{UsgEntry}%
\un@usgentryze{#3}}%\label{usgentryrs}
}%
{% |#3| is def entry
\def\kind@fentry{DefEntry}%
\un@defentryze{#3}%\label{defentryrs}
}% of |gmd/defentry/| test's `else'
\if@pageindex\@pageinclindexfalse\fi% should it be here or
% there? Definitely here because we'll wish to switch the switch
% with a~declaration.
\if@pageinclindex
\edef\gmu@tempa{gmdindexpagecs{\HLPrefix}{\kind@fentry}{\EntryPrefix}}%
\else
\edef\gmu@tempa{gmdindexrefcs{\HLPrefix}{\kind@fentry}{\EntryPrefix}}%
\fi
\edef\gmu@tempa{\IndexPrefix#2\actualchar%
\quotechar\bslash verb*#1\quoted@eschar#2#1% The last macro in
% this line usually means the first two, but in some cases
% it's redefined to be empty (when we use |\index@macro| to
% index not a~\CS).
\encapchar\gmu@tempa}%
\@xa\special@index\@xa{\gmu@tempa}% We give the
% indexing macro the argument expanded so that \pk{hyperref} may
% see the explicit encap-char in order not to add its own
% encapsulation of \verb+|hyperpage+ when the (default)
% |hyperindex=true|\TextCommonIndex*{hyperindex} option is in
% force. (After this setting the |\edef|s in the above may be
% changed to |\def|s.)
% \changes[\index@macro]{v0.98f}{06/9/30}{explicit MakeIndex
% controls changed to corresponding macros. Therefore
% \cs[]{hyperindex} option of \pk{hyperref} didn't see the
% encapsulation and added its own. So I~expanded the
% argument of the very indexing macro}
}{}% closing of |gmd/iexcl/| test.
}}
\def\un@defentryze#1{%
\ifcsname gmd/defentry/\@xa\detokenize\@xa{#1 }\endcsname
\@xa\g@relaxen\csname gmd/defentry/\@xa\detokenize\@xa{#1 }\endcsname
\fi
\ifx\gmd@detectors\empty
\g@relaxen\last@defmark
\fi}% the last macro (assuming \cs{fi} is not a~macro :-)
% is only used by \cs{changes}. If we are in the scope of automatic
% detection of definitions, we want to be able not to use \inverb|\Define|
% but write |\changes| after a~definition and get proper entry. Note
% that in case of automatic detection of definitions
% |\last@defmark|'s value keeps until the next definition.
\def\un@usgentryze#1{%
\ifcsname gmd/usgentry/\@xa\detokenize\@xa{#1 }\endcsname
\@xa\g@relaxen\csname gmd/usgentry/\@xa\detokenize\@xa{#1 }\endcsname
\fi}
\@emptify\EntryPrefix% this macro seems to be obsolete now
\newif\if@pageinclindex
\newcommand*\quoted@eschar{\quotechar\bslash}% we'll redefine it when
\def\IndexPrefix{}
\if@linesnotnum\@pageindextrue\fi
\AtBeginDocument{%
\if@pageindex
\def\gmdindexrefcs#1#2#3#4{\csname#2\endcsname{\hyperpage{#4}}}%^^A
% in the page case we gobble the third argument that is supposed
% to be the entry prefix.\ilrr
\let\gmdindexpagecs=\gmdindexrefcs
\else
\def \gmdindexrefcs#1#2#3#4{\gmiflink[clnum.#4]{%
\csname#2\endcsname{#4}}}%
\def \gmdindexpagecs#1#2#3#4{\hyperlink{page.#4}{%
\csname#2\endcsname{\gmd@revprefix{#3}#4}}}%
% \stanza \DefIndex\gmd@revprefix
\def\gmd@revprefix#1{%
\def\gmu@tempa{#1}%
\ifx\gmu@tempa\@empty p.\,\fi}
\providecommand*\HLPrefix{}% it'll be the hypertargets names' prefix
% in mul\-ti-docs. Moreover, it showed that if it was empty,
% \pk{hyperref} saw duplicates of the hyper destinations, which
% was perfectly understandable (|codelinenum.123| made by
% |\refstepcounter| and |codelinenum.123| made by
% |\gmhypertarget|). But since v0.98 it is not a~problem anymore
% because during the automatic \inverb|\hypertarget|ing the
% lines are labelled |clnum.|\<number>. When |\HLPrefix| was
% defined as dot, MakeIndex rejected the entries as `illegal page
% number'. \changes[\HLPrefix]{v0.98a}{06/09/05}{again
% \cs{@empty}fied since \cs{hypertarget}ing the codelines names
% them \cs[]{clnum} (since v0.98).}
\fi}
\def\DefEntry#1{\underline{#1}}
\def\UsgEntry#1{\textit{#1}}
\def\CommonEntryCmd{relax}
\outer\def\DefIndex{\begingroup
\MakePrivateLetters
\gmu@ifstar
{\@sanitize\MakePrivateOthers%^^A\verbatim@specials
\Code@DefIndexStar}%
{\Code@DefIndex}}
\long\def\Code@DefIndex#1{\endgroup{%
\escapechar\m@ne% because we will compare the macro's name with
% a~string without the backslash.
\@defentryze{#1}{1}}}
\long\def\Code@DefIndexStar#1{%
\endgroup{%
\addto@estoindex{#1}%
\@defentryze{#1}{0}}%
}
\def\gmd@justadot{.}
\long\def\@defentryze#1#2{%
\@xa\glet\csname gmd/defentry/\detokenize{#1 }\endcsname\gmd@justadot% The\\
% \LaTeX\ \inverb|\@namedef| macro could not be used since it's not
% `long'. The space to sound with the checker. \ilrr
% \Define\last@defmark
\ifcat\relax\@xa\@nx\@firstofmany#1\@nil
% \nostanza if we meet a~\CS, then maybe it's a~\CS to be
% ���defentryzed��� or maybe it's a~���verbatim special��� \CS. The only way
% to distinguish those cases is to assume there shouldn't be
% a~verbatim containing only a~���verbatim special��� \CS.
\@xa\def\@xa\gmu@tempa\@xa{\@allbutfirstof#1\@nil}%
\ifx\gmu@tempa\@empty
\afterfifi\@firstoftwo% if |#1| is a~single \CS, we
% \incmd\xiistring\ it. Otherwise we \incmd\detokenize\ it.
\else\afterfifi\@secondoftwo
\fi
\else\@xa\@secondoftwo
\fi
{\xdef\last@defmark{\xiistring#1}}% we |\string| the argument just in case it's
% a~control sequence. But when it can be a~\CS, we |\@defentryze| in
% a~scope of |\escapechar=-1|, so there will never be a~backslash at
% the beginning of |\last@defmark|'s meaning (unless we
% |\@defentryze| |\\|).
{\xdef\last@defmark{\detokenize{#1}}}%
\@xa\gdef\csname gmd/isaCS/\last@defmark\endcsname{#2}% |#2| is
}% of |\@defentryze|.
\long\def\@usgentryze#1{%
\@xa\let\csname gmd/usgentry/\detokenize{#1}\endcsname\gmd@justadot}
\@emptify\envirs@toindex
\outer\def\CodeUsgIndex{\begingroup
\MakePrivateLetters
\gmu@ifstar
{\@sanitize\MakePrivateOthers%^^A\verbatim@specials
\Code@UsgIndexStar}%
{\Code@UsgIndex}}
\long\def\Code@UsgIndex#1{%
\endgroup{%
\escapechar\m@ne
\global\@usgentryze{#1}}}
\long\def\Code@UsgIndexStar#1{%
\endgroup
{%^^A\verbatim@specials
\addto@estoindex{#1}%
\@usgentryze{#1}}%
}
\outer\def\CodeCommonIndex{\begingroup
\MakePrivateLetters
\gmu@ifstar
{\MakePrivateOthers\@sanitize\Code@CommonIndexStar}%
{\Code@CommonIndex}}
\long\def\Code@CommonIndex#1{\endgroup}
\long\def\Code@CommonIndexStar#1{%
\endgroup\addto@estoindex{#1}}
\long\def\text@indexmacro#1{%
{\escapechar\m@ne \xdef\macro@pname{\xiistring#1}}%
% ^^A \typeout{@@@@ 4803}%
\@xa\quote@mname\macro@pname\relax% we process the \CS's
% name char by char and quote MakeIndex controls. |\relax| is the
% iterating macro's stopper. The scanned \CS's quoted name shall be the
% expansion of |\macro@iname|.
\if\verbatimchar\macro@pname
\def\im@firstpar{[$]}%^^A$
\else\def\im@firstpar{}%
\fi
{\do@properindex% see line \ref{do@properindex}.
\@xa \index@macro\im@firstpar\macro@iname\macro@pname}}
\def\quote@mname{%
\def\macro@iname{}%
\quote@charbychar}
\def\quote@charbychar#1{%
\ifx\relax#1% finish quoting when you meet |\relax| or:
\else
\ifnum0\ifcat\@nx#1\@nx~1\fi\ifcat\@nx#1\relax1\fi>0 % we can meet
% active char and/or control sequences (made by) verbatim
% specials, therefore we check whether \inhash1 is an active char
% and if it is a~\CS.
\afterfifi{% we can meet an active char or a~\CS iff we use
% verbatim specials.
\ifdefined\verbatim@specials@list
\afterfi{%
\begingroup
\escapechar\@xa\@xa\@xa`\@xa\@firstofmany\verbatim@specials@list\@nil
\@xa\endgroup
\@xa\quote@charbychar\detokenize{#1}%% for a~\CS\
% \incs{detokenize} adds a~space but if so, it will be ignored
% by the argument scanner.
}% of |\afterfi|.
\else\PackageError{gmdoc}{Please report a\space bug in
\bslash quote@charbychar in line 4934}{}%
\fi% of |\ifdefined\verbatim@specials@list|.
}% of |\afterfifi|.
\else
\quote@char#1%
\xdef\macro@iname{\macro@iname \gmd@maybequote#1}%
\afterfifi\quote@charbychar
\fi
\fi}
\def\TextUsgIndex{\begingroup
\MakePrivateLetters
\gmu@ifstar{\MakePrivateOthers\Text@UsgIndexStar}{\Text@UsgIndex}}
\long\def\Text@UsgIndex#1{%
\endgroup\@usgentryze#1%
\text@indexmacro#1}
\long\def\Text@UsgIndexStar#1{\endgroup\@usgentryze{#1}%
\text@indexenvir{#1}}
\long\def\text@indexenvir#1{%
{\verbatim@specials
\edef\macro@pname{\xiistring#1}%
\if\bslash\@xa\@firstofmany\macro@pname\@nil% if
% |\string|ed |#1| begins with a~backslash, we will gobble it
% to make MakeIndex not see it.
\edef\gmu@tempa{\@xa\@gobble\macro@pname}%
\@tempswatrue
\else
\let\gmu@tempa\macro@pname
\@tempswafalse
\fi
% ^^A \typeout{@@@@@ 4858}%
\@xa\quote@mname\gmu@tempa\relax% \label{quote@mname comm}we
% process |\string|ed |#1| char by char and quote MakeIndex
% controls. |\relax| is the iterating macro's stopper. The quoted
% |\string|ed |#1| shall be the meaning of |\macro@iname|.
\if@tempswa
\def\quoted@eschar{\quotechar\bslash}%
\else\@emptify\quoted@eschar\fi% we won't print any backslash before
% an environment's name, but we will before a~\CS's name.
\do@properindex% see line \ref{do@properindex}.
\index@macro\macro@iname\macro@pname}}
\def\TextCommonIndex{\begingroup
\MakePrivateLetters
\gmu@ifstar{\MakePrivateOthers\Text@CommonIndexStar}{\Text@CommonIndex}}
\long\def\Text@CommonIndex#1{\endgroup
\text@indexmacro#1}
\long\def\Text@CommonIndexStar#1{\endgroup
\text@indexenvir{#1}}
\outer\def\CodeMarginize{\begingroup
\MakePrivateLetters
\gmu@ifstar
{\MakePrivateOthers\egCode@MarginizeEnvir}
{\egCode@MarginizeMacro}}
\long\def\egCode@MarginizeMacro#1{\endgroup
\Code@MarginizeMacro#1}
\long\def\Code@MarginizeMacro#1{{% |#1| is always a~\CS.
% ^^A \typeout{@@@@ :::\xiistring#1:::}%
% ^^A \show\detokenize
\escapechar\m@ne
\@xa\glet\csname gmd/2marpar/\xiistring#1\endcsname\gmd@justadot
}}
\long\def\egCode@MarginizeEnvir#1{\endgroup
\Code@MarginizeEnvir{#1}}
\long\def\Code@MarginizeEnvir#1{\addto@estomarginpar{#1}}
\def\mark@envir{%\label{mark@envir}
\ifx\envirs@tomarginpar\@empty
\else
\def\do{\Text@Marginize*}%
\envirs@tomarginpar%
\g@emptify\envirs@tomarginpar%
\fi
\ifx\envirs@toindex\@empty
\else
{\verbatim@specials
\gmd@doindexingtext
\envirs@toindex
\g@emptify\envirs@toindex}%
\fi}
\def\gmd@doindexingtext{%
\def\do##1{% the |\envirs@toindex| list contains |\string|ed
% macros or environments' names in braces and each preceded
% with |\do|. We extract the definition because we use it also in
% line \ref{gmd@doindexingtext 2nd use}.
\if\bslash\@firstofmany##1\@nil% if
% |##1| begins with a~backslash, we will gobble it for
% MakeIndex not see it.
\edef\gmd@resa{\@gobble##1}%
\@tempswatrue
\else
\edef\gmd@resa{##1}\@tempswafalse
\fi
\@xa\quote@mname\gmd@resa\relax% see line
% \ref{quote@mname comm} \& subs. for commentary.
{\if@tempswa
\def\quoted@eschar{\quotechar\bslash}%
\else\@emptify\quoted@eschar
\fi
\index@macro\macro@iname{##1}}}%
}
\@emptify\envirs@tomarginpar
\@emptify\envirs@toindex
\outer\def\Define{% note that since it's |\outer|, it doesn't have to
% be \incmd\protected.
\begingroup
\MakePrivateLetters
\gmu@ifstar{\@sanitize%^^A\verbatim@specials
\Code@DefEnvir}{\Code@DefMacro}}
\outer\def\CodeUsage{\begingroup
\MakePrivateLetters
\gmu@ifstar{%
\@sanitize%^^A\verbatim@specials
\MakePrivateOthers
\Code@UsgEnvir}{\Code@UsgMacro}}
\DeclareCommand\Code@DefMacro\long{om}{%
\Code@DefIndex#2% we use the internal macro; it'll close the group.
\IfValueTF{#1}%
{\Code@MarginizeMacro#1}%
{\Code@MarginizeMacro#2}%
}
\DeclareCommand\Code@UsgMacro\long{om}{%
\Code@UsgIndex#2% here also the internal macro; it'll close the
% group
\IfValueTF{#1}%
{\Code@MarginizeMacro#1}%
{\Code@MarginizeMacro#2}%
}
\def\codeline@wrindex#1{\if@filesw
\immediate\write\@indexfile
{\string\indexentry{#1}%
{\HLPrefix\number\c@codelinenum}}\fi}
\def\codeline@glossary#1{% It doesn't need to establish a~group since
% it is always called in a~group.
\if@pageinclindex
\edef\gmu@tempa{gmdindexpagecs{\HLPrefix}{relax}{\EntryPrefix}}%
\else
\edef\gmu@tempa{gmdindexrefcs{\HLPrefix}{relax}{\EntryPrefix}}% \inverb|relax| stands for the formatting command. But we don't want to do anything special with the change history entries. \ilrr
\fi
\protected@edef\gmu@tempa{%
\@nx\protected@write\@nx\@glossaryfile{}%
{\string\glossaryentry{#1\encapchar\gmu@tempa}%
{\HLPrefix\number\c@codelinenum}}}%
\gmu@tempa
}
\AtBeginDocument{%
\if@pageindex
\let\special@index=\index
\let\gmd@glossary\glossary
\else
% \DefIndex\special@index
\let\special@index=\codeline@wrindex
\let\gmd@glossary\codeline@glossary
% \label{codeline indexnumber declaration}
\fi}% postponed till |\begin{document}| with respect of \docfm-like
\def\gag@index{\let\index=\@gobble
%\label{gag@index}
\let\codeline@wrindex=\@gobble}
\Store@Macros{\index\codeline@wrindex}
\def\ungag@index{\Restore@Macros {\index\@@codeline@wrindex}}
%\label{ungag@index}
\DeclareCommand\Code@DefEnvir\long{om}{%
\endgroup
% ^^A \typeout{@@@@@ 5069: \unexpanded{#1}}%
{%^^A\verbatim@specials
\IfValueTF{#1}%
{\addto@estomarginpar{#1}}%
{\addto@estomarginpar{#2}}%
\addto@estoindex{#2}%
\@defentryze{#2}{0}}}
\DeclareCommand\Code@UsgEnvir\long{om}{%
\endgroup
{%^^A~\verbatim@specials
\IfValueTF{#1}%
{\addto@estomarginpar{#1}}%
{\addto@estomarginpar{#2}}%
\addto@estoindex{#2}%
\@usgentryze{#2}}}
\long\def\addto@estomarginpar#1{%
% ^^A \def\gmu@tempa{\do{#1}}% we |\string|ed the argument to
% ^^A % allow it to be a~control sequence but now (2008/12/14) it's not
% ^^A % necessary because |#1| is sanitised.
% ^^A \@xa\gaddtomacro\@xa\envirs@tomarginpar\@xa{\gmu@tempa}
\gaddtomacro\envirs@tomarginpar{\do{#1}}}
\long\def\addto@estoindex#1{%
% ^^A \typeout{@@@@ \unexpanded{#1}}%
% ^^A \edef\gmu@tempa{\@nx\do{\xiistring#1}}
% ^^A \@xa\gaddtomacro\@xa\envirs@toindex\@xa{\gmu@tempa}
\gaddtomacro\envirs@toindex{\do{#1}}}
\def\TextUsage{\begingroup
% \chgs{2008/12/8 v0.99s added \cmd\@sanitize\ in the starred
% version}
% \label{TextUsage}
\MakePrivateLetters
\gmu@ifstar{\@sanitize\MakePrivateOthers
\Text@UsgEnvir}{\Text@UsgMacro}}
\DeclareCommand\Text@UsgMacro\long{om}{%
\endgroup
% \chgs[\narrativett]{2008/12/12 v0.99s introduced in \pk{gmutils}
% and employed in the narrative verbatims, including \cs{ampulexdef}
% of the \pk{gmverb} macros}
\IfValueTF{#1}%
{\Text@Marginize*{#1}{\scanverb*{#1}}}%
{\Text@Marginize*{#2}{\scanverb*{#2}}}%
\begingroup\Code@UsgIndex#2% we declare the kind of formatting of the entry.
\text@indexmacro#2}
\DeclareCommand\Text@UsgEnvir\long{om}{%
\endgroup
% ^^A \typeout{@@@@ :::\detokenize{#1}::::}%
% ^^A \typeout{@@@@ :::\detokenize{#2}::::}%
\IfValueTF{#1}%
{\Text@Marginize*{#1}{\scanverb*{#1}}}%
{\Text@Marginize*{#2}{\scanverb*{#2}}}%
\@usgentryze{#2}% we declare the `usage' kind of formatting of the
% entry and index the sequence |#1|.
\text@indexenvir{#2}}
\pdef\TextMarginize{\@bsphack\begingroup
% \chgs{2008/12/8 v0.99s added \cmd\@sanitize\ in the starred
% version}
\MakePrivateLetters
\gmu@ifstar{%
\MakePrivateOthers\egText@MarginizeEnv}{\egText@MarginizeCS}}
\long\def\egText@MarginizeEnv#1{\endgroup
\Text@Marginize*{#1}% \chgs{2008/12/07 v0.99s a~bug fixed: braces
% added around \hash1}
\@esphack}
\long\def\egText@MarginizeCS#1{%
\endgroup
\Text@Marginize*{#1}%
}
\if@marginparsused
\reversemarginpar
\marginparpush\z@
\marginparwidth8pc\relax
% ^^A \settowidth{\marginparsep}{\ \ }%
%
% You may wish to put not only macros and environments to
% a~marginpar.
\long\def\gmdmarginpar#1{%
\marginpar{\raggedleft\strut
\hskip0ptplus100ptminus100pt%
#1}}%\stanza
%
\else
\long\def\gmdmarginpar#1{}%
\fi
\let\gmu@tempa\all@stars
\@xa\addtomacro\@xa\gmu@tempa\@xa{\all@unders}
\@xa\DeclareCommand\@xa\Text@Marginize\@xa!%
\@xa{\@xa Q\@xa{\gmu@tempa}m}{%
\gmdmarginpar{%
\addtomacro\verb@lasthook{\marginpartt}%
\IfValueTF{#1}{\scanverb#1}{\scanverb}{#2}}%
}% of |\Text@Marginize|.
\let\marginpartt\narrativett
\def\do@properindex{%\label{do@properindex}
\if@printalllinenos\else
\@pageinclindextrue
\let\special@index=\index
\fi}
\newenvironment*{macrocode*}{%\label{macrocode*}
\if@codeskipput\else\par\addvspace\CodeTopsep\@codeskipputgtrue\fi
\QueerEOL}%^^A\docstrips
{\par\addvspace\CodeTopsep\CodeEscapeChar\\}
% Let's remind that the starred version makes | |
\newenvironment*{macrocode}{%\label{macrocode}
\if@codeskipput\else\par\addvspace\CodeTopsep\@codeskipputgtrue\fi
\QueerEOL}% \changes{v0.99l}{2008/08/06}{removed \cs{CodeSpacesBlank}}
{\par\addvspace\CodeTopsep\CodeEscapeChar\\}
\let\oldmc\macrocode
\let\endoldmc\endmacrocode
\n@melet{oldmc*}{macrocode*}
\n@melet{endoldmc*}{endmacrocode*}
\addtomacro\oldmc{\@oldmacrocode@launch}%
\@xa\addtomacro\csname oldmc*\endcsname{%
\@oldmacrocode@launch}
\def\@oldmacrocode@launch{%
\emptify\gmd@textEOL% to disable it in |\gmd@docstripdirective|
% launched within the code.
\gmd@ctallsetup
\glet\stored@code@delim\code@delim
\@makeother\^^B\CodeDelim*\^^B%
\ttverbatim \gmd@DoTeXCodeSpace
\@makeother\|% because |\ttverbatim| doesn't do that.
\MakePrivateLetters% see line \ref{MPL}.\par
% ^^A \@xa\@makeother\code@delim
\docstrips@percent \@makeother\>%
% sine qua non of the automatic delimiting is replacing possible
% |*|\catother in the environment's name with |*|\catletter.
% Not to complicate assume |*| may occur at most once and only at
% the end. We also assume the environment's name consists only of
% character tokens whose catcodes (except of |*|) will be the same
% in the verbatim text.
\@xa\gmd@currenvxistar\@currenvir*\relax
\@oldmacrocode}
\foone{\catcode`*11 }
{\def\gmu@xistar{*}}
\def\gmd@currenvxistar#1*#2\relax{%
\edef\@currenvir{#1\if*#2\gmu@xistar\fi}}
\foone{%
\catcode`[=1 \catcode`]=2
\catcode`\{=\active \@makeother\}
\@makeother\^^B
\catcode`/=0 \catcode`\\=\active
\catcode`&=14 \catcode`*=11
\catcode`\%=\active \obeyspaces}& % \CodeEscapeChar\/ \CDAnd
[& here the \cs{foone}'s second pseudo-argument begins
&%\stanza
/def/@oldmacrocode[&
/bgroup/let =/relax&% to avoid writing |/@nx | four times.
/xdef/oldmc@def[&
/def/@nx/oldmc@end####1/@nx% /@nx\end&
/@nx{/@currenvir}[&
####1^^B/@nx/gmd@oldmcfinis]]&
/egroup&% now |\oldmc@edef| is defined to have one parameter delimited
&% with \incmd\end\arg{current env.'s name}
/oldmc@def&
/oldmc@end]&% \CDPerc
]
\def\gmd@oldmcfinis{%
\def\gmu@tempa{\end{\@currenvir}}%
\@xa\gmu@tempa\@xa\def\@xa\gmd@lastenvir\@xa{\@currenvir}%
\@xa\CodeDelim\@xa*\stored@code@delim
\gmd@mchook}% see line \ref{gmd@mchook}
\def\OldMacrocodes{%% \changes{v0.99m}{2008/08/10}{renamed from
%% \cs{VerbMacrocodes}}
\let\macrocode\oldmc
\n@melet{macrocode*}{oldmc*}}
\foone{\catcode`\%\active}
{\def\docstrips@percent{\catcode`\%\active
\let%\gmd@codecheckifds}}
\def\gmd@codecheckifds#1#2{% note that |#1| is just to gobble
% \inverb|\gmd@charbychar| token.
\typeout{@@@@@ codecheckifds hash 1: ��\unexpanded{#1}��, 2: ��\unexpanded{#2}��}%
\ifnum \if@dsdir 1\else 0\fi\ifgmd@dsVerb 1\fi>\z@
\afterfi{%
\gmd@dsChecker{%
\if\@nx<\@nx#2\afterfi\gmd@docstripdirective
\else\afterfi{\xiipercent#1#2}%
\fi}% of the checker's arg
}% of |\afterfi|
\else\afterfi{\xiipercent#1#2}%
\fi}
\newenvironment{macro}{%\label{macro}
\@tempskipa=\MacroTopsep
\if@codeskipput\advance\@tempskipa by-\CodeTopsep\fi
\par\addvspace{\@tempskipa}\@codeskipputgtrue
\begingroup\MakePrivateLetters\MakePrivateOthers% we make also the
% `private others' to cover the case of other sequence in the
% argument. (We'll use the |\macro| macro also in the environment
% for describing and defining environments.)
\gmd@ifonetoken\Hybrid@DefMacro\Hybrid@DefEnvir}%
{\par\addvspace\MacroTopsep\@codeskipputgtrue}
% It came out that the \docfm's author(s) give the \env{macro}
% environment also starred versions of commands as argument. It's \acro{OK}
% since (the default version of) |\MakePrivateLetters| makes |*|
% a~letter and therefore such a~starred version is just one \CS.
% However, in \pk{doc.dtx} occur \env{macro}s that mark
% \emph{implicit} definitions i.e., such that the defined \CS is not
% scanned in the subsequent code.
\@namedef{macro*}{\let\gmd@ifonetoken\@secondoftwo\macro}
\@xa\let\csname endmacro*\endcsname\endmacro
\def\gmd@ifsingle#1#2\@nil{%
\def\gmu@tempa{#2}%
\ifx\gmu@tempa\@empty}
\def\gmd@ifonetoken#1#2#3{%
\def\gmu@tempb{#3}% We hide |#3| from \TeX\ in case it's |\if������| or
% so. \inverb|\gmu@tempa| is used in \inverb|\gmd@ifsingle|. \ilrr
\gmd@ifsingle#3\@nil
\afterfi{\@xa#1\gmu@tempb}%
\else
\edef\gmu@tempa{\@xa\string\gmu@tempb}%
\afterfi{\@xa#2\@xa{\gmu@tempa}}%
\fi}
\long\def\Hybrid@DefMacro#1{%
\Code@DefIndex{#1}% this macro closes the group opened by |\macro|.
\Text@MarginizeNext{*{#1}}}
\long\def\Hybrid@DefEnvir#1{%
\Code@DefIndexStar{#1}% this macro also closes the group begun by
% |\macro|.
\Text@MarginizeNext{*{#1}}}
\long\def\Text@MarginizeNext#1{%
\gmd@evpaddonce{\Text@Marginize#1\ignorespaces}}
\long\def\gmd@evpaddonce#1{%
\global\advance\gmd@oncenum\@ne
\@xa\long\@xa\edef% \CodeUsgIndex*{NeuroOncer}
\csname gmd/evp/NeuroOncer\the\gmd@oncenum\endcsname{%
\@nx\g@relaxen
\csname gmd/evp/NeuroOncer\the\gmd@oncenum\endcsname}% Why does it
% work despite it shouldn't? Because when the \CS got with
% \inverb|\csname������\endcsname| is undefined, it's equivalent to
% |\relax| and therefore unexpandable. That's why it passes
% \inverb|\edef| and is able to be assigned.\ilrr
\@xa\addtomacro\csname gmd/evp/NeuroOncer\the\gmd@oncenum\endcsname{#1}%
\@xa\addto@hook\@xa\everypar\@xa{%
\csname gmd/evp/NeuroOncer\the\gmd@oncenum\endcsname}%
}
\newcount\gmd@oncenum
\@xa\let\@xa\environment\csname macro*\endcsname
\@xa\let\@xa\endenvironment\csname endmacro*\endcsname
\def\DoNotIndex{\bgroup\MakePrivateLetters\DoNot@Index}
\long\def\DoNot@Index#1{\egroup% we close the group,
\let\gmd@iedir\gmd@justadot% we declare the direction of the \<?>cluding
% to be \emph{ex}cluding. We act this way to be able to reverse the
% exclusions easily later.\ilrr
\dont@index#1.}
\long\def\dont@index#1{%
\def\gmu@tempa{\@nx#1}% My \TeX\ Guru's trick to deal with |\fi|
% and such, i.e., to hide from \TeX\ when it is processing a~test's
% branch without expanding.
\if\gmu@tempa.% a~dot finishes expelling
\else
\if\gmu@tempa,% The list this macro is put before may
% contain commas and that's O.K., we just continue the work.
\afterfifi\dont@index
\else% what is else shall off the Index be expelled.
{\escapechar\m@ne
\xdef\gmu@tempa{\string#1 }}% its to sound with
% \incs{detokenize}s in tests.
\@xa\let%
\csname gmd/iexcl/\gmu@tempa\endcsname=\gmd@iedir% In the default
% case explained e.g.\ by the macro's name, the last macro's
% meaning is such that the test in line \ref{iexcltest} will
% turn false and the subject \CS shall not be indexed.
% We |\let| not |\def| to spare \TeX's memory.
\afterfifi\dont@index
\fi
\fi}
\DoNotIndex\{ \DoNotIndex\}% \label{DNIbraces}the index entries of
\begin{MakePrivateLetters}% Yes, |\DoNotIndex| does
% |\MakePrivateLetters| on its own but No, it won't have any effect
% if it's given in another macro's |\def|.
% \HideAllDefining
\gdef\DefaultIndexExclusions{%
\DoNotIndex{\@ \@@par \@beginparpenalty \@empty}%\label{DIE1}
\DoNotIndex{\@flushglue \@gobble \@input}%
\DoNotIndex{\@makefnmark \@makeother \@maketitle}%
\DoNotIndex{\@namedef \@ne \@spaces \@tempa}%
\DoNotIndex{\@tempb \@tempswafalse \@tempswatrue}%
\DoNotIndex{\@thanks \@thefnmark \@topnum}%
\DoNotIndex{\@@ \@elt \@forloop \@fortmp \@gtempa \@totalleftmargin}%
\DoNotIndex{\" \/ \@ifundefined \@nil \@verbatim \@vobeyspaces}%
\DoNotIndex{\| \~ \ \active \advance \aftergroup \begingroup \bgroup}%
\DoNotIndex{\mathcal \csname \def \documentstyle \dospecials \edef}%
\DoNotIndex{\egroup}%
\DoNotIndex{\else \endcsname \endgroup ���endinput \endtrivlist}%
\DoNotIndex{\expandafter \fi \fnsymbol \futurelet \gdef \global}%
\DoNotIndex{\hbox \hss \if \if@inlabel \if@tempswa \if@twocolumn}%
\DoNotIndex{\ifcase}%
\DoNotIndex{\ifcat \iffalse \ifx \ignorespaces \index \input \item}%
\DoNotIndex{\jobname \kern \leavevmode \leftskip \let \llap \lower}%
\DoNotIndex{\m@ne \next \newpage \nobreak \noexpand \nonfrenchspacing}%
\DoNotIndex{\obeylines \or \protect \raggedleft \rightskip \rm \sc}%
\DoNotIndex{\setbox \setcounter \small \space \string \strut}%
\DoNotIndex{\strutbox}%
\DoNotIndex{\thefootnote \thispagestyle \topmargin \trivlist \tt}%
\DoNotIndex{\twocolumn \typeout \vss \vtop \xdef \z@}%
\DoNotIndex{\, \@bsphack \@esphack \@noligs \@vobeyspaces \@xverbatim}%
\DoNotIndex{\` \catcode \end \escapechar \frenchspacing \glossary}%
\DoNotIndex{\hangindent \hfil \hfill \hskip \hspace \ht \it \langle}%
\DoNotIndex{\leaders \long \makelabel \marginpar \markboth \mathcode}%
\DoNotIndex{\mathsurround \mbox}%\verb+% \newcount \newdimen \newskip+
\DoNotIndex{\nopagebreak}%
\DoNotIndex{\parfillskip \parindent \parskip \penalty \raise \rangle}%
\DoNotIndex{\section \setlength \TeX \topsep \underline \unskip}%
\DoNotIndex{\vskip \vspace \widetilde \\ \% \@date \@defpar}%
\DoNotIndex{\[ \]}% see line \ref{DNIbraces}.
\DoNotIndex{\count@ \ifnum \loop \today \uppercase \uccode}%
\DoNotIndex{\baselineskip \begin \tw@}%
\DoNotIndex{\a \b \c \d \e \f \g \h \i \j \k \l \m \n \o \p \q}%
\DoNotIndex{\r \s \t \u \v \w \x \y \z \A \B \C \D \E \F \G \H}%
\DoNotIndex{\I \J \K \L \M \N \O \P \Q \R \S \T \U \V \W \X \Y \Z}%
\DoNotIndex{\1 \2 \3 \4 \5 \6 \7 \8 \9 \0}%
\DoNotIndex{\! \# \$ \& \' \( \) \. \: \; \< \= \> \? \_}% |\+|
% seems to be so rarely used that it may be advisable to index it.
\DoNotIndex{\discretionary \immediate \makeatletter \makeatother}%
\DoNotIndex{\meaning \newenvironment \par \relax \renewenvironment}%
\DoNotIndex{\repeat \scriptsize \selectfont \the \undefined}%
\DoNotIndex{\arabic \do \makeindex \null \number \show \write \@ehc}%
\DoNotIndex{\@author \@ehc \@ifstar \@sanitize \@title}%
\DoNotIndex{\if@minipage \if@restonecol \ifeof \ifmmode}%
\DoNotIndex{\lccode %|%\newtoks|
\onecolumn \openin \p@ \SelfDocumenting}%
\DoNotIndex{\settowidth \@resetonecoltrue \@resetonecolfalse \bf}%
\DoNotIndex{\clearpage \closein \lowercase \@inlabelfalse}%
\DoNotIndex{\selectfont \mathcode \newmathalphabet \rmdefault}%
\DoNotIndex{\bfdefault}%
% From the above list I~removed some |\new������| declarations because
% I~think it may be useful to see gathered the special |\new������|s of
% each kind. For the same reason I~would not recommend excluding
% from the index such declarations as |\AtBeginDocument|,
% |\AtEndDocument|, |\AtEndOfPackage|, |\DeclareOption|,
% |\DeclareRobustCommand| etc. But the
% common definitions, such as |\���(new���|provide���)command| and
% |\���(e���|g���|x���)def|s, as the most common, in my opinion excluded should
% be.^^A\)
%
% \stanza
% And some my exclusions:
\DoNotIndex{\@@input \@auxout \@currentlabel \@dblarg}%
\DoNotIndex{\@ifdefinable \@ifnextchar \@ifpackageloaded}%
\DoNotIndex{\@indexfile \@let@token \@sptoken \^}% the latter comes
% from \CSs like |\^^M|, see sec. \ref{Tasks}.
\DoNotIndex{\addto@hook \addvspace}%
\DoNotIndex{\CurrentOption}%
\DoNotIndex{\emph \empty \firstofone}%
\DoNotIndex{\font \fontdimen \hangindent \hangafter}%
\DoNotIndex{\hyperpage \hyperlink \hypertarget}%
\DoNotIndex{\ifdim \ifhmode \iftrue \ifvmode \medskipamount}%
\DoNotIndex{\message}%
\DoNotIndex{\NeedsTeXFormat \newcommand \newif}%
\DoNotIndex{\newlabel}%
\DoNotIndex{\of}%
% ^^A\PassOptionsToClass\PassOptionsToPackage
\DoNotIndex{\phantom \ProcessOptions \protected@edef}%
\DoNotIndex{\protected@xdef \protected@write}%
\DoNotIndex{\ProvidesPackage \providecommand}%
\DoNotIndex{\raggedright}%
\DoNotIndex{\raisebox \refstepcounter \ref \rlap}% ^^A\RequirePackage
\DoNotIndex{\reserved@a \reserved@b \reserved@c \reserved@d}%
\DoNotIndex{\stepcounter \subsection \textit \textsf \thepage \tiny}%
\DoNotIndex{\copyright \footnote \label \LaTeX}%
% \changes[\DefaultIndexExclusions]{v0.98a}{06/09/06}{more
% macros added to the `exclude list'}
\DoNotIndex{\@eha \@endparenv \if@endpe \@endpefalse \@endpetrue}%
\DoNotIndex{\@evenfoot \@oddfoot \@firstoftwo \@secondoftwo}%
\DoNotIndex{\@for \@gobbletwo \@idxitem \@ifclassloaded}%
\DoNotIndex{\@ignorefalse \@ignoretrue \if@ignore}%
\DoNotIndex{\@input@ \@input}%
\DoNotIndex{\@latex@error \@mainaux \@nameuse}%
\DoNotIndex{\@nomath \@oddfoot}%|%\@onlypreamble| should be indexed
% \acro{IMHO}.
\DoNotIndex{\@outerparskip \@partaux \@partlist \@plus}%
\DoNotIndex{\@sverb \@sxverbatim}%
\DoNotIndex{\@tempcnta \@tempcntb \@tempskipa \@tempskipb}%\\
% I~think the layout parameters even the kernel, should not be
% excluded: \inverb|\@topsep| \inverb|\@topsepadd|^^B
% \ \inverb|\abovedisplayskip| \inverb|\clubpenalty| etc.
\DoNotIndex{\@writeckpt}%
\DoNotIndex{\bfseries \chapter \part \section \subsection}%
\DoNotIndex{\subsubsection}%
\DoNotIndex{\char \check@mathfonts \closeout}%
\DoNotIndex{\fontsize \footnotemark \footnotetext \footnotesize}%
\DoNotIndex{\g@addto@macro \hfilneg \Huge \huge}%
\DoNotIndex{\hyphenchar \if@partsw \IfFileExists }%
\DoNotIndex{\include \includeonly \indexspace}%
\DoNotIndex{\itshape \language \LARGE \Large \large}%
\DoNotIndex{\lastbox \lastskip \m@th \makeglossary}%
\DoNotIndex{\maketitle \math@fontsfalse \math@fontstrue \mathsf}%
\DoNotIndex{\MessageBreak \noindent \normalfont \normalsize}%
\DoNotIndex{\on@line \openout \outer}%
\DoNotIndex{\parbox \part \rmfamily \rule \sbox}%
\DoNotIndex{\sf@size \sffamily \skip}%
\DoNotIndex{\textsc \textup \toks@ \ttfamily \vbox}%
% \changes[\DoNotIndex]{v0.97}{06/09/04}{Excluding some star-versions of
% commands}
% \nostanza\noindent
% |%% \DoNotIndex{\begin*}| maybe in the future, if the idea gets
% popular\dots \nostanza
\DoNotIndex{\hspace* \newcommand* \newenvironment* \providecommand*}%
\DoNotIndex{\renewenvironment* \section* \chapter*}%\label{DIE2}
}% of |\DefaultIndexExclusions|.\par
% I~put all the expellings into a~macro because I~want them
% to be optional.
\end{MakePrivateLetters}
\if@indexallmacros\else
\DefaultIndexExclusions
\fi
\def\DoIndex{\bgroup\MakePrivateLetters\Do@Index}
\long\def\Do@Index#1{\egroup\@relaxen\gmd@iedir\dont@index#1.}% note
\def\UndoDefaultIndexExclusions{%
\Store@Macro\DoNotIndex
% ^^A\dont@index
\let\DoNotIndex\DoIndex
% ^^A \let\dont@index=\do@index
\DefaultIndexExclusions
% ^^A \RestoreMacro\dont@index
\Restore@Macro\DoNotIndex}
\long\def\IndexPrologue#1{\@bsphack\def\index@prologue{#1}\@esphack}
\def\indexdiv{\@ifundefined{chapter}{\section*}{\chapter*}}
\@ifundefined{index@prologue} {\def\index@prologue{\indexdiv{Index}%
\markboth{Index}{Index}%
Numbers written in italic refer to the \if@pageindex pages \else
code lines \fi where the
corresponding entry is described; numbers underlined refer to the
\if@pageindex\else code line of the \fi definition; numbers in
roman refer to the \if@pageindex pages\else code lines \fi where
the entry is used.
\if@pageindex\else
\ifx\HLPrefix\@empty
The numbers preceded with `p.' are page numbers.
\else The numbers with no prefix are page numbers.
\fi\fi
\ifx\IndexLinksBlack\relax\else
All the numbers are hyperlinks.
% ^^A, they are made black just to let Adobe Reader work
% ^^A~faster.
\fi
\gmd@dip@hook% this hook is intended to let a~user add
% something without redefining the entire prologue, see below.
}}{}
\@emptify\gmd@dip@hook
\long\def\AtDIPrologue#1{\g@addto@macro\gmd@dip@hook{#1}}
\AtDIPrologue{%\label{verb:ampulex:index}
\ampulexdef\verb\ttverbatim\verbatim@specials
{\ttverbatim\verbatim@specials}}
\RequirePackage{multicol}
\newdimen\IndexMin \IndexMin = 133pt\relax% originally it was set
\newcount\c@IndexColumns \c@IndexColumns = 3
\renewenvironment{theindex}
{\begin{multicols}\c@IndexColumns[\index@prologue][\IndexMin]%
\IndexLinksBlack
\IndexParms \let\item\@idxitem \ignorespaces}%
{\end{multicols}}
\def\IndexLinksBlack{\hypersetup{linkcolor=black}}% To make Adobe
\@ifundefined{IndexParms}
{\def\IndexParms{%
% \label{IndexParms}
\parindent \z@
\columnsep 15pt
\parskip 0pt plus 1pt
\rightskip 15pt
\mathsurround \z@
\parfillskip=-15pt plus 1 fil % \docfm\ defines this parameter
% rigid but that's because of the stretchable space (more
% precisely, a~|\dotfill|) between the item and the entries. But
% in \pk{gmdoc} we define no such special delimiters, so we add
% an infinite stretch.
\small
\def\@idxitem{\par\hangindent 30pt}%
\def\subitem{\@idxitem\hspace*{15pt}}%
\def\subsubitem{\@idxitem\hspace*{25pt}}%
\def\indexspace{\par\vspace{10pt plus 2pt minus 3pt}}%
\ifx\EntryPrefix\@empty\else\raggedright\fi% long (actually,
% a~quite \emph{short but nonempty} entry prefix) made space
% stretches so terribly large in the justified paragraphs that
% we should make |\raggedright| rather.
\ifnum\c@IndexColumns>\tw@\raggedright\fi% the numbers in narrow
% col\-umns look better when they are |\raggedright| in my opinion.
}}{}
\def\PrintIndex{% we ensure the standard meaning of the line end
% character not to cause a~disaster.
\@ifQueerEOL{\StraightEOL\printindex\QueerEOL}%
{\printindex}}
\foone{\@makeother\<\@makeother\>
\glet\sgtleftxii=<}
{
\def\gmd@docstripdirective{%
\begingroup\let\do=\@makeother
\do\*\do\/\do\+\do\-\do\,\do\&\do\|\do\!\do\(\do\)\do\>\do\<% ^^Ayes, we
% ^^A make \inverb|<|\catother\ since a~directive |<<|\<any text till ^^B>>
% ^^A~the end of line> theoretically may occur.
\@ifnextchar{<}{%
\let\do=\@makeother \dospecials
\gmd@docstripverb}
{\gmd@docstripinner}}%
\def\gmd@docstripinner#1>{%
\endgroup
\def\gmd@modulehashone{%
\Module{#1}\space
\@afternarrgfalse\@aftercodegtrue\@codeskipputgfalse}%
% ^^A \gmd@docstripshook
\gmd@textEOL\gmd@modulehashone}
% A~word of explanation: first of all, we
% close the group for changed |\catcode|s; the directive's text has its
% |\catcode|s fixed. Then we put the directive's text wrapped with the
% formatting macro into one macro in order to give just one token
% the \gmdoc's \TeX\ code scanner.
% ^^A But before we launch the \TeX\ code
% ^^A scanning with all the b\&w, we virtually restore the meaning of
% ^^A further |<|s to let them start further \ds\ directives,
% ^^A and after this possible restore (if it \emph{actually} takes place
% ^^A depends on the declaration used: |\docstrip| or |\docstrips|) we
% Then launch this big \TeX\ code scanning machinery by calling
% |\gmd@textEOL| which is an alias for the `narrative' meaning of the
% line end. This macro opens the verbatim group and launches the
% char-by-char scanner. That is this scanner because of what we
% encapsulated the directive's text with the formatting into one
% macro: to let it pass the scanner.
%
% That's why in the `old' macrocodes case the active |%| closes the
% group before launching |\gmd@docstripdirective|.
%
% \stanza
% The `verbatim' directive macro works very similarly.
}
\foone{\@makeother\<\@makeother\>
\glet\sgtleftxii=<
\catcode`\^^M=\active}%
{%
\def\gmd@docstripverb<#1^^M{%
\endgroup%
\def\gmd@modulehashone{%
\ModuleVerb{#1}\@afternarrgfalse\@aftercodegtrue%
\@codeskipputgfalse}%
\xdef\gmd@dsVerbDelim{\detokenize{#1}}%
\gmd@textEOL\gmd@modulehashone^^M}%
}
\edef\gmd@dsVerbDelim{\detokenize{#@$$$$#_()(_)*981-71092519-40^^A^^B}}
\pdef\dsVerbClose{%
\global\gmd@dsVerbtrue}
\providecommand*\Module[1]{{%
\mod@math@codes$\langle\mathsf{#1}\rangle$}}
\providecommand*\ModuleVerb[1]{{%
\mod@math@codes$\langle\langle\mathsf{#1}$}}
\def\ModuleVerbClose#1{{%
\xiipercent
\mod@math@codes$\mathsf{#1}$
{\normalfont[\ds\ verbatim closing dir.]}}}
\def\mod@math@codes{\mathcode`\|="226A \mathcode`\&="2026 }
\def\gmd@DefineChanges{%
\outer\long\def\changes{%
\gmd@changes@init
\changes@}}
\def\gmd@changes@init{%
\@bsphack\begingroup\@sanitize
\catcode`\\\z@ \catcode`\ 10 \MakePercentIgnore
\catcode`\^=7
\MakePrivateLetters \StraightEOL
\MakeGlossaryControls}
\newcommand\changes[4][]{\PackageWarningNoLine{gmdoc}{%
^^JThe \bslash changes command used \on@line
^^Jwith no \string\RecordChanges\space declared.
^^JI shall not warn you again about it}%
%^^A\gmd@countlines{#4}%
\renewcommand\changes[4][]{%^^A\gmd@countlines{#4}
}}
\def\MakeGlossaryControls{%
\edef\actualchar{\string=}\edef\quotechar{\string!}%
\edef\levelchar{\string>}\edef\encapchar{\xiiclub}}%for the glossary the
% `actual', the `quote' and the `level' chars are respectively |=|,
% |!| and |>|, the `encap' char remains untouched. I~decided to
% preserve the \docfm's settings for the compatibility.
\newcommand\changes@[4][\generalname]{%
\if@RecentChange{#3}% if the date is later than the one stored in
% \cs{c@Chang\+es\+Start\+Date},
\@tempswafalse
\ifx\generalname#1% then we check whether a~\CS-entry is given
% in the optional first argument or is it unchanged.
\ifx\last@defmark\relax\else% if no particular \CS is
% specified in |#1|, we check whether |\last@defmark| contains
% something and if so, we put it into |\gmu@tempb| scratch macro.
\@tempswatrue
\edef\gmu@tempb{% it's a~bug fix: while typesetting traditional
% \file{.dtx}es, \inverb|\last@defmark| came out with \inverb|\| at the
% beginning (which resulted with \inverb|\\|\<name> in the change
% log) but while typesetting the `new' way, it occurred
% without the bslash. So we gobble the bslash if it's
% present and two lines below we handle the exception of
% \inverb|\last@defmark|\equals|{\}| (what would happen if
% a~definition of \inverb|\\| was marked in new way \gmdoc ing).%^^A]
\if\bslash\last@defmark\else\last@defmark\fi}%
\ifx\last@defmark\bslash\let\gmu@tempb\last@defmark\fi%
\n@melet{gmd@glossCStest}{gmd/isaCS/\last@defmark}%
\fi
\else%the first argument isx not |\generalname| i.e.,
%a~particular \CS is specified by it (if some day one
% wishes to |\changes| |\generalname|, \heshe\ should type
% |\changes[generalname]|\dots)
\@tempswatrue
{\escapechar\m@ne
\xdef\gmu@tempb{\string#1}}%
\if\bslash\@xa\@firstofmany\string#1\relax\@nil% we check
% whether \inverb|#1| is a~\CS\dots\ilrr
\def\gmd@glossCStest{1}%\dots\ and tell the glossary if so.
\fi
\fi
\@ifundefined{gmd@glossCStest}{\def\gmd@glossCStest{0}}{}%
\protected@edef\gmu@tempa{\@nx\gmd@glossary{%
\if\relax\GeneralName\relax\else
\GeneralName% it's for the |\DocInclude| case to precede
% every |\changes| of the same file with the file name, cf.\ line
% \ref{GeneralName}.
\fi
#2\levelchar%
\if@tempswa% If the macro |\last@defmark|
% doesn't contain any \CS name (i.e., is empty) nor |#1|
% specifies a~\CS, the current
% changes entry was done at top-level. In this case we precede
% it by |\generalname|.
\gmu@tempb
\actualchar\bslash verb*%
\if\verbatimchar\gmu@tempb$\else\verbatimchar\fi
\if1\gmd@glossCStest\quotechar\bslash\fi \gmu@tempb
\if\verbatimchar\gmu@tempb$\else\verbatimchar\fi
\else
\space\actualchar\generalname
\fi
:\levelchar%^^A\gmd@deeolize{
#4%^^A}
}}%
\gmu@tempa
\grelaxen\gmd@glossCStest
\fi% of |\if@recentchange|
%^^A\gmd@countlines{#4}%
\endgroup\@esphack}
\@relaxen\last@defmark
\@emptify\GeneralName
\def\ChangesGeneral{\grelaxen\last@defmark}% If automatic detection of
\AtBegInput{\ChangesGeneral}
\newcount\c@ChangesStartDate
\def\if@RecentChange#1{%
\gmd@setChDate#1\@nil\@tempcnta
\ifnum\@tempcnta>\c@ChangesStartDate}
\def\gmd@setChDate#1/#2/#3\@nil#4{% the last parameter will be a~|\count|
% register.
#4=\numexpr#1*\@M+#2*100+#3\relax
%\CH{2010/06/23 changed from \TeX's arithmetic to \cmd\numexpr}
}
\def\ChangesStart#1#2{%
\gmd@setChDate#2\@nil\c@ChangesStartDate
\typeout{^^JPackage gmdoc info: ^^JChanges' start date #1 memorised
as \string<\the\c@ChangesStartDate\string> \on@line.^^J}
\advance\c@ChangesStartDate\m@ne% we shall show the changes \emph{at ^^B
% the specified day} and later.
\ifnum\c@ChangesStartDate>19820900 %\unskip\StraightEOL\footnote{^^A
% DEK writes in \textit{\TeX, The Program} of September 1982 as the
% date of \TeX\ Version 0.}\QueerEOL\label{TeXVer0}
% see \gmiflink[personalchanges]{below}.
\edef\gmu@tempa{%
\@nx\g@addto@macro\@nx\glossary@prologue{%
The changes
\if\relax\GeneralName\relax\else of \GeneralName\space\fi
earlier than
#1 \if\relax#1\relax #2\else(#2)\fi\space are not shown.}}%
\gmu@tempa
\fi}
\def\generalname{General}
\def\RecordChanges{\makeglossary\gmd@DefineChanges
\@relaxen\RecordChanges}
\newdimen\GlossaryMin \GlossaryMin = 80pt
\newcount\c@GlossaryColumns \c@GlossaryColumns = 2
\newenvironment{theglossary}{%
\begin{multicols}\c@GlossaryColumns
[\glossary@prologue][\GlossaryMin]%
\GlossaryParms \IndexLinksBlack
\let\item\@idxitem \ignorespaces}%
{\end{multicols}}
\long\def\GlossaryPrologue#1{\@bsphack
\def\glossary@prologue{#1}%
\@esphack}
\@ifundefined{glossary@prologue}
{\def\glossary@prologue{\indexdiv{{Change History}}%
\markboth{{Change History}}{{Change History}}%
}}{}
\AtBeginDocument{%
\@ifundefined{GlossaryParms}{\let\GlossaryParms\IndexParms}{}}
\def\PrintChanges{% to avoid a~disaster among queer EOLs:
\@ifQueerEOL
{\StraightEOL\@input@{\jobname.gls}\QueerEOL}%
{\@input@{\jobname.gls}}%
\g@emptify\PrintChanges}
\pdef\toCTAN{%\
% \begin{enumargs}
% \item \<year/month/day>\textvisiblespace\<version number>
% \end{enumargs}
% \changes{v0.99r}{2008/11/22}{added}
% \chgs{2008/11/30 v0.99s made a~shorthand for \cs{chgs} not
% \cs{changes}}
\gmd@changes@init
\gmd@toCTAN@}
\def\gmd@toCTAN@#1{%
\edef\gmu@tempa{\gmd@chgs@parse#1 \@nil}%
\edef\gmu@tempa{%
\unexpanded{\changes@[\generalname]}%
{\@xa\@firstofthree\gmu@tempa}%
{\@xa\@secondofthree\gmu@tempa}%
{put to \acro{CTAN} on \@xa\@secondofthree\gmu@tempa}}%
\gmu@tempa}
\outer\pdef\chgs{\gmd@changes@init\gmd@chgs}
\DeclareCommand\gmd@chgs{%
o % the optional \CS the change refers to
>!m % change's date, version and text
}{%
\IfValueTF{#1}{%%
\edef\gmu@tempa{\@nx\changes@[\unexpanded{#1}]%
\@xa\unexpanded\@xa{\gmd@chgs@parse#2\@nil}}}%
{\edef\gmu@tempa{\@nx\changes@
\@xa\unexpanded\@xa{\gmd@chgs@parse#2\@nil}}}%
\gmu@tempa}% of |\gmd@chgs|
\long\def\gmd@chgs@parse#1 #2 #3\@nil{{#2}{#1}{#3}}%
\outer\pdef\CH{%
\gmd@changes@init\gmd@chgsplus}
\DeclareCommand\gmd@chgsplus{\SameAs\gmd@chgs}{%
\DCUse\gmd@chgs{#1}{#2}%
\gmd@threeway{#1}#2\@nil
}
\long\def\gmd@threeway
#1% opt. \CS that |\CH| refers to
#2 % (delimd.\ with a blank) date
#3 % (delimd.\ with a blank) version
#4\@nil % text
{%
\par (#2, #3\IfValueT{#1}{, \texttt{\detokenize\@xa{\string#1}}}:)
#4\scantokens{}% to provide proper line end which'll take care of
% |\par| \&c.
}
\newcount\check@sum
\def\CheckSum#1{\@bsphack\global\check@sum#1\relax\@esphack}
\newcounter{CheckSum}
\newcommand*\step@checksum{\stepcounter{CheckSum}}
\AtEndInput{\check@checksum}
\def\check@checksum{\relax
\ifnum\check@sum=\z@
\edef\gmu@tempa{% why \cs{edef}---see line \ref{wypis sumy}
\@nx\typeout{**********************************^^J%
* The input file \gmd@inputname\space has no Checksum
stated.^^J%
* The current checksum is \the\c@CheckSum.^^J%
\gmd@chschangeline% a~check sum changes history entry, see below.
* (package gmdoc info.)^^J%
**********************************^^J}}
\else
\ifnum\check@sum=\c@CheckSum
\edef\gmu@tempa{%
\@nx\typeout{*****+*+*+*+*+*+*+*+*+*+^^J%
* The input file \gmd@inputname: Checksum passed.^^J%
\gmd@chschangeline
* (package gmdoc info.)^^J%
*****+*+*+*+*+*+*+*+*+*+^^J}}
\else
\edef\gmu@tempa{%
\@nx\typeout{********\gmd@wykrzykniki^^J%
*! The input file \gmd@inputname:^^J%
*! The CheckSum stated: \the\check@sum\space<> my
count: \the\c@CheckSum.^^J%
\gmd@chschangeline
*! (package gmdoc info.)^^J%
********\gmd@wykrzykniki^^J}}%
% \chgs{2008/11/30 v0.99s
% \env{\quotechar!*\quotechar!*\quotechar!���} sequence changed to
% \env{\quotechar! \quotechar! \quotechar!���} for better
% distinction}
\fi
\fi
\gmu@tempa
\@xa\AtEndDocument\@xa{\gmu@tempa}%\label{wypis sumy} we print
% the checksum notification on the terminal immediately and at end
% of \TeX ing not to have to scroll the output far nor search the log.
\global\check@sum\z@}
\def\gmd@wykrzykniki{! ! ! ! ! ! ! ! ! ! ! !}
\def\gmd@chschangeline{%
\xiipercent\space\string\chschange
{\@ifundefined{fileversion}{v???}{\fileversion}}%
{\the\year/\the\month/\the\day}%
{\the\c@CheckSum}^^J%
\xiipercent\space\string\chschange
{\@ifundefined{fileversion}{v???}{\fileversion}}%
{\@xa\@gobbletwo\the\year/\the\month/\the\day}%
{% with two di\-g\-it year in case you use |\ChangesStart|.
\the\c@CheckSum}^^J}
\outer\pdef\chschange{% \
% \begin{enumargs}
% \dc m file version,
% \dc m date,
% \dc checksum,
% \dc o the reason of check sum change, possibly short.
% \end{enumargs}
\@ifQueerEOL
{\def\EOLwasQueer{11}}{\def\EOLwasQueer{10}}%
\gmd@changes@init
\chschange@}
\DeclareCommand\chschange@{mmmo}{%\label{chschange}
\changes@{#1}{#2}{CheckSum #3
\IfValueT{#4}{because of #4}%
}% |\csname������| because
% \incs{cha\+n\+g\+es} is \inverb|\outer|.
\CheckSum{#3}%
\IfValueF{#4}{%
\if\EOLwasQueer
\afterfi{%
\@ifnextchar\par{%
\@xa\gmd@textEOL\gobble}%
{}%
}% of |\afterfi|,
\fi}% of no value of |#4|,
}% of |\chschange@|.\nostanza
\newcommand*\ltxPageLayout{%
\setlength{\textwidth}{355pt}%
\setlength\marginparwidth{95pt}%
\setlength\oddsidemargin{82pt}%
\setlength\evensidemargin{82pt}}
\DeclareCommand\DocInclude{O{}mO{}}{%\
% \begin{enumargs}
% \dc o path (with closing slash), will not be printed
% \dc m file name without extension, will be printed
% \dc o file extension (with dot) if not \file{.sty}, \file{.cls},
% \file{.tex}, \file{.dtx} nor \file{.fdd}
% \end{enumargs}\
% originally it took just one
% argument. Here we make it take two, first of which is intended to
% be the path (with the closing \inverb |/|). This is intended not to print
% the path in the page footers only the filename.\par
% \CodeUsage\HLPrefix
\gdef\HLPrefix{\filesep}%
\gdef\EntryPrefix{\filesep}% we define two rather kernel parameters
% to expand to the file marker. The first will bring the information
% to one of the default \inverb|\IndexPrologue|'s
% |\if|s. Therefore the
% definition is global. The latter is such for symmetry.
\def\GeneralName{#2\actualchar\pk{#2} }% \label{GeneralName} for the
% changes' history main level entry.
%
% Now we check whether we try to include ourselves and if
% so---we'll (create and) read an \file{.auxx} file instead of
% (the main) \file{.aux} to avoid an infinite recursion of |\input|s.
%
\edef\gmd@jobname{\jobname}%
\edef\gmd@difilename{% we want the filename all `other', just as
% in \inverb|\jobname|.
\@xa\@xa\@xa\@gobble\@xa\string\csname#2\endcsname}%
\ifx\gmd@jobname\gmd@difilename
\def\gmd@auxext{auxx}%
\else
\def\gmd@auxext{aux}%
\fi
\relax
%^^A\if@ltxDocInclude
\clearpage
% ^^A\fi
\gmd@docincludeaux \def\currentfile{gmdoc-IncludeFileNotFound.000}%
\let\fullcurrentfile\currentfile
\@ifnonempty{#3}%
{%
\unless\if.\@firstofmany#3\relax\@nil
\PackageError{gmdoc}{Optional \xiihash3 of
\string\DocInclude\space
if present has to begin with a dot (.)}{}%
\fi
\edef\currentfile{#2#3}%
\IfFileExists{#1\currentfile}{}%
{\PackageError{gmdoc}{\string\DocInclude\space file
\currentfile\space not found}{}}%
}% of if extension given.
{% if extension not given:
\IfFileExists{#1#2.fdd}{\edef\currentfile{#2.fdd}}{% it's not \pk{.fdd},
\IfFileExists{#1#2.dtx}{\edef\currentfile{#2.dtx}}{% it's not \pk{.dtx}
% either,
\IfFileExists{#1#2.sty}{\edef\currentfile{#2.sty}}{% it's not
% \pk{.sty},
\IfFileExists{#1#2.cls}{\edef\currentfile{#2.cls}}{% it's
% not \pk{.cls},
\IfFileExists{#1#2.tex}{\edef\currentfile{#2.tex}}{% it's
% not \pk{.tex},
\IfFileExists{#1#2.fd}{\edef\currentfile{#2.fd}}{% so it
% must be \pk{.fd} or error.
\PackageError{gmdoc}{\string\DocInclude\space file
#1#2.fdd/dtx/sty/cls/tex/fd not found.}{}%
}}}}}}%
}% of if no extension given
% \changes{v0.98j}{06/10/16}{\pk{.fdd} added in \cs{DocInclude}'s
% search for the extension.}
\edef\currentfile{\@xa\detokenize\@xa{\currentfile}}%
\edef\fullcurrentfile{#1\currentfile}%
\ifnum\@auxout=\@partaux
\@latexerr{\string\DocInclude\space cannot be nested}\@eha
\else \@docinclude{#1}#2#3 \fi}% Why is |#2| delimited with | | not
\def\@docinclude#1#2 {% To match the macro's parameter string, is an
% answer. But why is |\@docinclude| defined so? Originally, in
% \pk{ltxdoc} it takes one argument and it's delimited with a~space
% probably in resemblance to the true |\input| \nlpercent(|\@@input| in
% \LaTeX). ^^A\if@ltxDocInclude
\clearpage
%^^A\fi
\if@filesw \gmd@writemauxinpaux{#2.\gmd@auxext}\fi% this strange macro with
% a~long name is another spurious thing to allow |_| in the filenames (see
% line \ref{gmd@writemauxinpaux}). which are allowed anyway unless
% active or \cat14.
\@tempswatrue
\if@partsw \@tempswafalse\edef\gmu@tempb{#2}%
\@for \gmu@tempa:=\@partlist\do{\ifx\gmu@tempa\gmu@tempb\@tempswatrue\fi}%
\fi
\if@tempswa% the file is on |\@partlist|
\let\@auxout\@partaux
\if@filesw
\immediate\openout\@partaux #2.\gmd@auxext\relax% Yes, only |#2|. It's to
% create and process the partial \pk{.aux(x)} files always in the main
% document's (driver's) directory.
% \changes[\DocInclude]{v0.97}{06/09/04}{\protect\TeX forced to
% process the partial \pk{.aux}es in the main document directory}
\immediate\write\@partaux{\relax}%
\fi
% \begin{quotation}We need to save (and later restore) various
% index-related commands which might be changed by the included
% file.\end{quotation}
% \changes{v0.98a}{06/09/06}{\cs{MacroStoringDo} renamed (in
% \pk{gmutils}) to \cs{StoringAndRelaxingDo}, and
% \cs{MacroRestoringDo} to \cs{RestoringDo}.}
\StoringAndRelaxingDo\gmd@doIndexRelated
\if@ltxDocInclude\part{\currentfile}% In the \pk{ltxdoc}-like setup
% we make a~part title page with only the filename and the file's
% |\maketitle| will typeset an \pk{article}-like title.
\else\let\maketitle=\InclMaketitle
\fi% In the default setup we
% redefine |\maketitle| to typeset a~common chapter or part heading.
\if@ltxDocInclude\xdef@filekey\fi
\GetFileInfo{\currentfile}% it's my (GM) addition with the account
% of using file info in the included files' title/\:heading etc.
\incl@DocInput{\fullcurrentfile}% originally just |\currentfile|.
\if@ltxDocInclude\else\xdef@filekey\fi% in the default case we add
% new file to the file key \emph{after} the input because in this
% case it's the files own |\maketitle| what launches the
% sectioning command that increases the counter.
% \par And here is the moment to restore the index-related
% commands.
\RestoringDo\gmd@doIndexRelated
% ^^A\if@ltxDocInclude
\clearpage
% ^^A\fi
\gmd@writeckpt{#1#2}%
\if@filesw \immediate\closeout\@partaux \fi
\else% the file isn't on |\@partlist|
\@nameuse{cp@#1#2}%
\g@emptify\gmd@ABIOnce
\fi
\let\@auxout\@mainaux}% end of |\@docinclude|.
\def\xdef@filekey{{\@relaxen\narrativett% \label{LetTTFRelax}This
% assignment is very trickily crafted: it makes \emph{all}
% |\narrativett|s present in the |\filekey|'s expansion unexpandable
% not only the one added in this step.
\xdef\filekey{\filekey, \thefilediv={\narrativett\currentfile}}}}
\def\gmd@writemauxinpaux#1{% \label{gmd@writemauxinpaux}
% this name comes from `\emph{write} out to \emph main
% \emph{\pk{.aux}} to \emph{in}put \emph partial
% \emph{\pk{.aux}}'.\par
% We wrap |\@input{|\<partial \pk{.aux}>|}| in a~|_|\catother\
% hacked scope. This hack is especially recommended here since the
% \pk{.aux} file may contain a~non-|\global| stuff that should not
% be localised by a~group that we would have to establish if we
% didn't use the hack. (Hope you understand it. If not, notify me
% and for now I'll only give a~hint: ``Look at it with the \TeX's
% eyes''. More uses of this hack are to be seen in \pk{gmutils}
% where they are a~bit more explained.)
\immediate\write\@mainaux{%
\unexpanded{%
\bgroup
\@makeother\_% to allow underscore
\@makeother\~% to allow paths beginning with |~/|
\firstofone}{\egroup
\string\@input{#1}}}}
\def\gmd@writeckpt#1{%
\immediate\write\@partaux{%
\unexpanded{%
\bgroup
\@makeother\_%
\@makeother\~%
\firstofone}\@charlb\egroup}%
\@writeckpt{#1}%
\immediate\write\@partaux{\@charrb}}
\def\gmd@doIndexRelated{%
\do\tableofcontents \do\makeindex \do\EnableCrossrefs
\do\PrintIndex \do\printindex \do\RecordChanges \do\PrintChanges
\do\theglossary \do\endtheglossary}
\@emptify\filesep
\def\aalph#1{\@aalph{\csname c@#1\endcsname}}
\def\@aalph#1{%
\ifcase#1\or a\or b\or c\or d\or e\or f\or g\or h\or i\or
j\or k\or l\or m\or n\or o\or p\or q\or r\or s\or
t\or u\or v\or w\or x\or y\or z\or A\or B\or C\or
D\or E\or F\or G\or H\or I\or J\or K\or L\or M\or
N\or O\or P\or Q\or R\or S\or T\or U\or V\or W\or
X\or Y\or Z\else\@ctrerr\fi}
\def\gmd@docincludeaux{%
% We set the things for including the files only once.
\global\@relaxen\gmd@docincludeaux
% By default, we will include multiple files into one document
% as chapters in the classes that provide |\chapter| and as parts
% elsewhere.
\ifx\filediv\relax
\ifx\filedivname\relax% (nor |\filediv| neither
% |\filedivname| is defined by the user)
% \changes[\filediv]{v0.98a}{06/09/06}{definition moved into
% \cs{DocInclude} and let \cs{relax} by default ($\quotechar=$a~bug fix).}
\@ifundefined{chapter}{%
\SetFileDiv{part}}%\changes[\DocInclude]{v0.98f}{06/9/29}{\cs{filediv(name)}
% defined as \cs[]{(\protect\bslash)part} for the classes that
% don't know \cs{chapter}}
{\SetFileDiv{chapter}}%
\else% (|\filedivname| is defined by the user, |\filediv| is not)
\SetFileDiv{\filedivname}% why not? Inside is |\edef| so it'll work.
\fi
\else% (|\filediv| is defined by the user
\ifx\filedivname\relax% and |\filedivname| is not)
%^^A~I~deleted a~six level test which one sectioning command
%^^A~|\filedivname| isx.
\PackageError{gmdoc}{You've redefined \string\filediv\space
without redefining \string\filedivname.}{Please redefine the
two macros accordingly. You may use \string\SetFileDiv{name
without bslash}.}%
\fi
\fi
% \changes[\DocInclude]{v0.99m}{2008/08/09}{
% resetting of codeline number with every
% \cs{filedivname} commented out because with the
% \env{countalllines} option it caused that reset at \cs{maketitle}
% after some lines of file}
% ^^A \@addtoreset{codelinenum}{\filedivname}% remember it has
% ^^A~a~|\global| effect in fact. For each file we'll reset
% ^^A~\env{codelinenum}.
\def\thefilediv{\aalph{\filedivname}}% The files will be numbered
% with letters, lowercase first.
\@xa\let\csname the\filedivname\endcsname=\thefilediv% This
% li\-ne lets |\the|\<chapter> etc.\ equal |\thefilediv|.
\def\filesep{\thefilediv-}% File separator (identifier) for the index.
\let\filekey=\@gobble
\g@addto@macro\index@prologue{%
\gdef\@oddfoot{\parbox{\textwidth}{\strut\footnotesize
\raggedright{\bfseries File Key:} \filekey}}% The footer for
% the pages of index.
\glet\@evenfoot\@oddfoot}% \label{oneside}anyway, it's intended to
% be oneside.
\g@addto@macro\glossary@prologue{%
\gdef\@oddfoot{\strut Change History\hfill\thepage}% The footer
% for the changes history.
\glet\@evenfoot\@oddfoot}%
% \changes[\DocInclude]{v0.98c}{06/9/10}{Change History footer
% defined}
\gdef\@oddfoot{% The footer of the file pages will be its name and,
% if there is a~file info, also the date and version.
\@xa\ifx\csname ver@\currentfile\endcsname\relax
File \thefilediv: {\narrativett\currentfile} %
\else
\GetFileInfo{\currentfile}%
File \thefilediv: {\narrativett\filename} %
Date: \filedate\ %
Version \fileversion
\fi
\hfill\thepage}%
\glet\@evenfoot\@oddfoot% see line \ref{oneside}.
% ^^A \show\filedivname% for debug
\@xa\def\csname\filedivname name\endcsname{File}% we
% redefine the name of the proper division to `File'.
\ifx\filediv\section
\let\division=\subsection
\let\subdivision=\subsubsection
\let\subsubdivision=\paragraph
% If |\filediv| is higher than
% |\section| we don't change the three divisions (they are
% |\section|, |\subsection| and |\subsubsection| by default).
% |\section| seems to me the lowest reasonable sectioning command
% for the file. If |\filediv| is lower you should rather rethink
% the level of a~file in your documentation not redefine the two
% divisions.
\fi}% end of |\gmd@docincludeaux|.
\def\SetFileDiv#1{%
\edef\filedivname{#1}%
\@xa\let\@xa\filediv\csname#1\endcsname}
\def\SelfInclude{\DocInclude{\jobname}}
\newif\if@ltxDocInclude
\newcommand*\ltxLookSetup{%
\SetFileDiv{part}%
\ltxPageLayout
\@ltxDocIncludetrue
}
\@onlypreamble\ltxLookSetup
\let\incl@DocInput=\DocInput
\@emptify\currentfile% for the pages outside the
\newcommand*\olddocIncludes{%
\let\incl@DocInput=\OldDocInput}
\newcommand*\gmdocIncludes{%
\let\incl@DocInput=\DocInput
\AtBegInput{\QueerEOL}}% to move back the |\StraightEOL| declaration put
\AtBegInput{%^^A~ why provide not just \cs{def}?
\providecommand*\maketitle{\par
\begingroup \def \thefootnote {\fnsymbol {footnote}}%
\setcounter {footnote}\z@
\def\@makefnmark{\rlap{\@textsuperscript{\normalfont\@thefnmark}}}%
\long\def\@makefntext##1{\parindent 1em\noindent
\hb@xt@1.8em{%
\hss\@textsuperscript{\normalfont\@thefnmark}}##1}%
\if@twocolumn \twocolumn [\@maketitle ]%
\else \newpage \global \@topnum \z@ \@maketitle \fi
\thispagestyle{titlepage}\@thanks \endgroup
\setcounter {footnote}\z@
\gdef\@date{\today}\g@emptify\@thanks%
\g@relaxen\@author\g@relaxen\@title%
}%\par
% \begin{quotation}When a number of articles are concatenated into
% a~journal, for example, it is not usual for the title pages of
% such documents to be formatted differently. Therefore, a class
% such as \pk{ltugboat} can define this macro in advance. However,
% if no such definition exists, we use page style plain for title
% pages.\end{quotation}
\@ifundefined{ps@titlepage}{\let\ps@titlepage=\ps@plain}{}%
% \par And let's provide |\@maketitle| just in case: an error occurred
% without it at \TeX ing with \pk{mwbk.cls} because this class with the
% default options does not define |\@maketitle|. The below definitions
% are taken from \pk{report.cls} and \pk{mwrep.cls}.
\providecommand*\@maketitle{%
\newpage\null \vskip 2em\relax%
\begin{center}%
\titlesetup
\let \footnote \thanks
{\LARGE \@title \par}%
\vskip 1.5em%
{\large \lineskip .5em%
\begin{tabular}[t]{c}%
\strut \@author
\end{tabular}\par}%
\vskip 1em%
{\large \@date}%
\end{center}%
\par \vskip 1.5em\relax}%\par
% We'd better restore the primary meanings of the macros making
% a~title. (\LaTeXe\ source, File F: ltsect.dtx Date: 1996/12/20 Version
% v1.0z, lines 3.5.7.9--12.14--17.)
\providecommand*\title[1]{\gdef\@title{#1}}
\providecommand*\author[1]{\gdef\@author{#1}}
\providecommand*\date[1]{\gdef\@date{#1}}
\providecommand*\thanks[1]{\footnotemark
\protected@xdef\@thanks{\@thanks
\protect\footnotetext[\the\c@footnote]{#1}}%
}%
\providecommand*\and{%| % \begin{tabular}|
\end{tabular}%
\hskip 1em \@plus.17fil%
\begin{tabular}[t]{c}}%| % \end{tabular}|
% And finally, let's initialise \cs{tit\+le\+set\+up} if it is not yet.
\providecommand*\titlesetup{}%
}% end of |\AtBegInput|.
\@relaxen\filediv
\@relaxen\filedivname
\@relaxen\thefilediv
\def\InclMaketitle{%
% \changes{v0.98a}{06/09/05}{renamed from \cs{incl@maketitle} because
% I~needed it in self-input hacks.}
% \chgs{2008/12/10 v0.99s a~bug fixed: \cmd\if\cmd\relax\cmd\@date\
% changed to \cmd\ifx}
%
{\def\and{, }% we make |\and| just a~comma.
{\let\thanks=\@gobble% for the toc version of the heading we
% discard |\thanks|.
\protected@xdef\incl@titletotoc{%
\@title\@ifauthor{%
\protect\space(\@author)}{}}% we add the author iff the `files have
% different authors' \emph{and} author exists \nlpercent(|@variousauthors|)
}%
\def\thanks##1{\footnotemark
\protected@xdef\@thanks{\@thanks% to keep the previous |\thanks|
% if there were any.
\protect\footnotetext[\the\c@footnote]{##1}}}% for some
% mysterious reasons so defined |\thanks| do typeset the footnote
% mark and text but they don't hyperlink it
% properly. A~\pk{hyperref} bug?
\@emptify\@thanks
\protected@xdef\incl@filedivtitle{%
[{\incl@titletotoc}]% braces to allow |[| and
% |]| in the title to toc.
{\protect\@title
{\protect\smallerr% this macro is provided by the \pk{gmutils}
% package after the \pk{relsize} package.
\@ifauthor
{\protect\\[0.15em]\@nx\@author
\ifx\relax\@date\else, \fi}% after use, |\@date| is let to
% \incmd\relax.
{\ifx\relax\@date\else\protect\\[0.15em]\fi}
% The default is that all the included files have the same
% author(s). In this case we won't print the author(s) in
% the headings. Otherwise we wish to print them. The
% information which case are we in is brought by the
% |\if@variousauthors| switch defined in line \ref{@variousauthors}.
\protect\@date}}% end of |\incl@filedivtitle|'s brace (2nd
% or 3rd argument).
}% end of |\incl@filedivtitle|'s |\protected@xdef|.\par
% We |\protect| all the title components to avoid expanding
% |\footnotemark| hidden in |\thanks| during |\protected@xdef|
% (and to let it be executed during the typesetting, of course).
}% end of the comma-|\and|'s group.
\@xa\filediv\incl@filedivtitle
\@thanks
\g@relaxen\@author \g@relaxen\@title \g@relaxen\@date
\g@emptify\@thanks
}% end of |\InclMaketitle|.
\newif\if@variousauthors
\newcommand*\PrintFilesAuthors{\@variousauthorstrue}
\newcommand*\SkipFilesAuthors{\@variousauthorsfalse}
\def\@ifauthor{%\
% \begin{enumargs}
% \item what if true
% \item what if false
% \end{enumargs}
\ifnum\numexpr\if@variousauthors1\else0\fi*
\ifx\@author\relax0\else\ifx\@author\@empty0\else1\fi\fi>0
\@xa\@firstoftwo
\else
\@xa\@secondoftwo
\fi
}
\def\GetFileInfo#1{%
\def\filename{#1}%
\def\gmu@tempb##1 ##2 ##3\relax##4\relax{%
\def\filedate{##1}%
\def\fileversion{##2}%
\def\fileinfo{##3}}%
\edef\gmu@tempa{\csname ver@#1\endcsname}%
\@xa\gmu@tempb\gmu@tempa\relax? ? \relax\relax}
\newcommand*\FileInfo{%
\bgroup
\gmd@ctallsetup
\bgroup% yes, we open two groups because we want to rescan tokens in
% `usual' catcodes. We cannot put \incs{gmd@ctallsetup} into the
% inner macro because when that will be executed, the
% \incs{inputlineno} will be too large (the last not the first line).
\let\do\@makeother
\do\ \do\{\do\}\do\^^M\do\\%
\gmd@fileinfo}
\foone{%
\catcode`!\z@
\catcode`(\@ne
\catcode`)\tw@
\let\do\@makeother
\do\ % we make space `other' to keep it for scanning the code where
% it may be leading.
\do\{\do\}\do\^^M\do\\}%\CodeEscapeChar\!
(%
!def!gmd@fileinfo#1Provides#2{#3}#4[#5]#6^^M%
(!egroup% we close the group of changed catcodes, the catcodes
!gmd@writeFI(#2)(#3)(#5)%
!gmd@FIrescan(#1Provides#2{#3}#4[#5]#6)% this macro will close the group.
)% \CodeEscapeChar\\
)
\def\gmd@writeFI#1#2#3{%
% ^^A \typeout{@@@ write FI}\show\relax
{\newlinechar=\endlinechar% \chgs{2008/11/30 v0.99s added assignment
% of \cmd\newlinechar}
\immediate\write\@auxout{%
\global\@nx\@namedef{%
ver@#2.\if P\@firstofmany#1\@nil sty\else cls\fi}{#3}}}}
\foone\obeylines{%
\def\gmd@FIrescan#1{%
% \changes{v0.99l}{2008/08/06}{\cs{scantokens} used instead of
% \cs{write} and \cs{@@input} which simplified the macro}
% ^^A \newwrite\gmd@docrescanfile
% ^^A \immediate\openout\gmd@docrescanfile=\jobname.docrescan\relax
{\newlinechar=\endlinechar\scantokens{#1}}\egroup^^M}}
\def\ProvideFileInfo#1{%
\begingroup
\catcode`\ 10 \catcode\endlinechar 10 %
\@makeother\/\@makeother\&%
\kernel@ifnextchar[{\gmd@providefii{#1}}{\gmd@providefii{#1}[]}%^^A]
}
% \DefIndex\gmd@providefii
\def\gmd@providefii#1[#2]{%\\
% (we \emph{don't} write the file info to \pk{.log})
\@xa\xdef\csname ver@#1\endcsname{#2}%
\endgroup}
\def\ProvideSelfInfo{\ProvideFileInfo{\jobname.tex}}
\def\gmd@upperDIV#1{%
\if d#1D\fi
\if i#1I\fi
\if v#1V\fi
}
\def\edefInfo
#1% name
#2% datum
{%
\edef\gmd@edefInfo@resa{\gmd@upperDIV #2}%
\@nameedef{file#2}{%
\ifcsname #1Leaf\gmd@edefInfo@resa\endcsname % e.g.~|gmbaseLeafVersion|
\xA\xA\xA\detokenize\xA\xA\xA{%
\csname #1Leaf\gmd@edefInfo@resa\endcsname
}%
\else
\ifcsname #1\gmd@edefInfo@resa\endcsname % e.g.~|gmbaseVersion|
\xA\xA\xA\detokenize\xA\xA\xA{%
\csname #1\gmd@edefInfo@resa\endcsname
}%
\else
\ifcsname \gmBundleFile \gmd@edefInfo@resa\endcsname % e.g.~|gmutilsVersion|
\xA\xA\xA\detokenize\xA\xA\xA{%
\csname \gmBundleFile \gmd@edefInfo@resa\endcsname
}%
\fi
\fi
\fi
}% of edefined macro
}% of |\edefInfo|
\def\FileInfoFromName#1{%
\edefInfo{#1}{date}%
\edefInfo{#1}{version}%
\edefInfo{#1}{info}%
\def\GeneralName{#1\actualchar\pk{#1} }% for the changes' history.
}
\def\BundleInfoFromName#1{%
\def\gmBundleFile{#1}%
\Store@MacroSt {#1LeafDate}%
\Store@MacroSt {#1LeafVersion}%
\Store@MacroSt {#1LeafInfo}%
\n@melet{#1LeafDate}{@undefined}%
\n@melet{#1LeafVersion}{@undefined}%
\n@melet{#1LeafInfo}{@undefined}%
\FileInfoFromName{#1}%
\Restore@MacroSt {#1LeafDate}%
\Restore@MacroSt {#1LeafVersion}%
\Restore@MacroSt {#1LeafInfo}%
}
\pdef\filenote{This file has version number \fileversion{} dated \filedate{}.}
\pdef\thfileinfo{\thanks\filenote}
\pdef\fnfileinfo{%
\gmu@ifedetokens{\@currext}{toc}%
{}%
{\footnote\filenote}%
}
\foone{\obeylines}%
{%
\pdef\IndexInput#1{% \changes{v0.98b}{06/09/07}{\cs{StoreMacro}ing
%% and \cs{RestoreMacro}ing
% \cs{code@delim} added}
\Store@Macro\code@delim%
\CodeDelim*\^^Z%
\def\gmd@iihook{% \label{iihook}this hook is |\edef|ed!
\@nx^^M%
\code@delim\relax\@nx\let\@nx\EOFMark\relax}%^^A\@nx^^M
\DocInput{#1}\Restore@Macro\code@delim}%
}
\newenvironment{gmlonely}{\StraightEOL}{}
\newcommand\skipgmlonely[1][]{%
\def\gmu@tempa{%\DefIndex\gmd@skipgmltext
\def\gmd@skipgmltext{%
\g@emptify\gmd@skipgmltext
% ^^A \gmd@deeolize{
#1%^^A}
}}% not to count the lines of the substituting
% text but only of the text omitted
\gmu@tempa
\@xa\AtBegInput\@xa{\gmu@tempa}%
\renewenvironment{gmlonely}{%
\StraightEOL
\@fileswfalse% to forbid writing to \pk{.toc}, \pk{.idx} etc.
\setbox0=\vbox\bgroup}{\egroup\gmd@skipgmltext}}
\foone{\catcode`\_=8 }% we ensure the standard |\catcode| of |_|.
{%
\newcommand*\catother{${}_{12}$}%\par
% Similarly, if we need to say some char is of category 13 (`active'),
% we'll write \catactive, got by |\catactive|
\newcommand*\catactive{${}_{13}$}%\par
% and a~letter, \catletter
\newcommand*\catletter{${}_{11}$}%.
}
\newenvironment*{copyrnote}{%
\StraightEOL\everypar{\hangindent3em\relax\hangafter1 }%
\par\addvspace\medskipamount\parindent\z@\obeylines}{%
\@codeskipputgfalse\stanza}
\StoreEnvironment{quotation}
\def\gmd@quotationname{quotation}
\renewenvironment{quotation}{%
% \changes{v0.99j}{2008/08/03}{Improved behaviour of redefined
% \env{quotation} to be the original if used by another environment.}
% The first non-me user complained that \env{abstract} comes out in
% quotation marks. That is because \env{abstract} uses \env{quotation}
% internally. So we first check whether the current environment is
% \env{quotation} or something else.
\ifx\@currenvir\gmd@quotationname
\afterfi{\par``\ignorespaces}%
\else\afterfi{\storedcsname{quotation}}%
\fi}
{\ifx\@currenvir\gmd@quotationname
\afterfi{\ifhmode\unskip\fi''\par}%
\else\afterfi{\storedcsname{endquotation}}%
\fi}
\def\gmdnoindent{%
\ifvmode\leavevmode\hskip-\parindent\ignorespaces
\fi}% \incs{ignorespaces} is added to eat a~space inserted% by
\pdef\inverb{%
% \changes{v0.99g}{2007/11/12}{added}
\gmu@ifstar{%
\def\gmu@tempa{\verbhyphen}% the pre-break.
\@emptify\gmu@tempb% the no-break.
\gmd@inverb}%
{\@emptify\gmu@tempa% the pre-break empty
\def\gmu@tempb{\gmboxedspace}% the no-break boxed space.
\gmd@inverb}}
\newcommand*\gmboxedspace{\hbox{\normalfont{ }}}
\pdef\gmd@nlperc{% \changes{v0.99n}{2008/08/22}{% added
% \cs{hbox}es in \cs{discretionary} to score \cs{hyphenpenalty} not
% \cs{exhyphenpenalty}}
% \chgs{2008/12/06 v0.99s \cs{newcommand*} replaced with \cs{pdef}
% and optional argument's declaration removed since nothing is done
% to \texttt{\#1} in the body of now-macro. Wrapped in a~group for
% setting \cmd\hyphenpenalty}
\ifhmode\unskip\fi
\begingroup\hyphenpenalty\inverbpenalty\relax
\discretionary{\hbox{\gmu@tempa}}% (pre-break). I~always put
% a~\incs{hbox} here to make this discretionary score the
% \incs{hyphenpenalty} not \incs{exhyphenpenalty} (\TB\ p.~96) since
% the latter may be 10,000 in Polish typesetting.
{\hbox{\narrationmark}}% (post-break)
{\gmu@tempb}% (no-break).
\endgroup
\penalty10000\hskip0sp\relax}
\def\inverbpenalty{-1000}
\pdef\gmd@inverb{%
\gmd@nlperc
\ifmmode\hbox\else\leavevmode\null\fi
\bgroup
\ttverbatim
\narrativett
\def\breakablevisspace{%
\discretionary{\visiblespace}{\narrationmark}{\visiblespace}}%
\def\breakbslash{%
\discretionary{}{\narrationmark\type@bslash}{\type@bslash}}%
\def\breaklbrace{%
\discretionary
{\xiilbrace\verbhyphen}%
{\narrationmark}%
{\xiilbrace}}%
\gm@verb@eol
%^^A \gmu@ifstar {\@sverb@chbsl}
%^^A~ {\gmobeyspaces\frenchspacing\@sverb@chbsl}
\@sverb@chbsl% It's always with visible spaces.
}
\pdef\nlperc{\newline\narrationmark\ignorespaces}
\pdef\nlpercent{%
% ^^A \gmu@ifstar{\def\gmu@tempa{{\tt\xiipercent}}%
% ^^A \@emptify\gmu@tempb
% ^^A \gmd@nlperc}%
% ^^A {
% \chgs{200/12/7 v0.99s starred version discarded as never used and
% undocumented}
% \UnDef
\@emptify\gmu@tempa
\def\gmu@tempb{\gmboxedspace}%
\gmd@nlperc
% ^^A}
}
\pdef\incs{% an in-line \cs{cs}
% \changes{v0.99m}{008/08/07}{added}
% \chgs{200/12/7 v0.99s starred version discarded as never used and
% undocumented}
% ^^A \gmu@ifstar{\def\gmu@tempa{{\tt\xiipercent}}%
% ^^A \@emptify\gmu@tempb
% ^^A \gmd@nlperc\cs}%
% ^^A {
% \UnDef
\@emptify\gmu@tempa
\def\gmu@tempb{\gmboxedspace}%
\gmd@nlperc\cs
% ^^A}
}
\def\inenv{\incs[]}% an in-line \cs{env}
\def\incmd{% it has to be |\def| to let it expand to let \incmd\cmd\
% convert its argument to a~safe string.
\nlpercent\cmd}
\def\inhash{\nlpercent\hash}
\@xa\ampulexdef\csname\@dc@InnerName\cs\endcsname
[#1] [#1] {\begingroup} {\ifdefined}
{\begingroup \def\+{\discre{\gmv@hyphen}{\narrationmark}{}}%
\ifdefined}
\providecommand*\ds{DocStrip}
\newcommand*\CDAnd{\CodeDelim\&}
\newcommand*\CDPerc{\CodeDelim\%}
\let\division=\section
\let\subdivision=\subsection
\let\subsubdivision=\subsubsection
\newcounter{gmd@mc}
\def\gmd@mchook{\stepcounter{gmd@mc}% \label{gmd@mchook}
\gmd@mcdiag
\ifcsname gmd@mchook\the\c@gmd@mc\endcsname
\afterfi{\csname gmd@mchook\the\c@gmd@mc\endcsname}%
\fi}
\long\def\AfterMacrocode#1#2{\@namedef{gmd@mchook#1}{#2}}
\@emptify\gmd@mcdiag
\def\mcdiagOn{\def\gmd@mcdiag{%
\typeout{^^J\bslash end{\gmd@lastenvir} No.\the\c@gmd@mc
\space\on@line, cln.\the\c@codelinenum.}}}
\def\mcdiagOff{\@emptify\gmd@mcdiag}
\DeclareEnvironment{enumargs}{o}%\
{%
\StraightEOL
\if@aftercode
\edef\gmu@tempa{\the\leftskip}%
\edef\gmu@tempb{\the\hangindent}%
\fi
\enumerate
\if@aftercode
\leftskip=\glueexpr\gmu@tempa+\gmu@tempb\relax
\fi
\edef\gmd@ea@hashes{%
\#\ifcase\IfValueTF{#1}{#1}{1}\relax
\or\or\#\or\or\#\#\#\fi}%
%\stanza
\@namedef{label\@enumctr}{%
\env{\if@aftercode\narrationmark\fi
\relax% to stop |\ignorespaces|
\gmd@ea@bwrap
\gmd@ea@hashes
\csname the\@enumctr\endcsname
\gmd@ea@ewrap}}% of |\label���<@enumctr>|.
\let\mand\item
\provide\gmd@ea@wraps{%
\emptify\gmd@ea@ewrap
\emptify\gmd@ea@bwrap}%
\gmd@ea@wraps
\def\opt{%\HideDef
\def\gmd@ea@bwrap{[}\def\gmd@ea@ewrap{]}%
\item
\gmd@ea@wraps}%
%\stanza
\settowidth{\@tempdima}{\narrativett x\gmd@ea@hashes7x}%
\edef\gmd@ea@xxxwd{\the\@tempdima}%
%\stanza
\DeclareCommand\dc !{%
Q{*>} % (1) we check whether there's a~sergeant right of the prefix or
% a~star to suppress parentheses,
Q{P!lL\long iI} % (2) an optional ���bare��� prefix for a~���long��� argument
% or for ignored
b % (3) prefix(es) in curly braces
% ( % This way we allow the prefix(es) to be braced or not
% at the author's option),
>\@xa T{\@dc@argtypes} % (4) (optional) argument type
% specifier,
b % (5) (optional) default value of the specified argument or
% (for |K| and |G|) mandatory.
b % (6) default of |K| and |G|.
}{%
\gmu@ifxany *{##1}%
{% a~|*| suppresses bracket/brace/parentheses decoration.
\def\gmd@ea@bwrap{\hbox to \gmd@ea@xxxwd\bgroup\hss}%
\def\gmd@ea@ewrap{\hss\egroup}%
}%
{% if there's no |*| in \hash1, be wrap the item label in
% braces/brackets/parentheses.
\gmu@ifxany ##4{bB}{% I~decide not to print |m| type
% arguments in braces because the braces are not mandatory for
% this type.
\def\gmd@ea@bwrap{\{}%
\def\gmd@ea@ewrap{\}}%
}{}%
\gmu@ifxany ##4{cC}{%
\def\gmd@ea@bwrap{(}%
\def\gmd@ea@ewrap{)}%
}{}%
\gmu@ifxany ##4{oO}{%
\def\gmd@ea@bwrap{[}%
\def\gmd@ea@ewrap{]}%
}{}%
\gmu@ifxany ##4{G}{%
\def\gmd@ea@bwrap{\detokenize\@xa{\@firstoftwo##5}}%
\def\gmd@ea@ewrap{\detokenize\@xa{\@secondoftwo##5}}%%
}{}%
\gmu@ifxany ##4{A}{%
\def\gmd@ea@bwrap{<}%
\def\gmd@ea@ewrap{>}%
}{}%
}% of if no |*| in \hash1.
\gmu@ifxany ##4{mQsSTK\afterassignment}{%
\def\gmd@ea@bwrap{\hbox to \gmd@ea@xxxwd\bgroup\hss}%
\def\gmd@ea@ewrap{\hss\egroup}%
}{}%
% we add a~normal space
\addtomacro\gmd@ea@ewrap{{\normalfont\ }}%
\IfValueT{##2}{%
\addtomacro\gmd@ea@ewrap{>\{\string##2\}}}%
\IfValueT{##3}{%
\addtomacro\gmd@ea@ewrap{>\{##3\}}}%
\IfValueT{##4}{%
\ifx s##4%
\addtomacro\gmd@ea@ewrap{%
\llap{\metachar[}\scanverb{*}\metachar]}%
\else\addtomacro\gmd@ea@ewrap{##4}%
\fi}%
\IfValueT{##5}{%
\addtomacro\gmd@ea@ewrap{\{%
%\nostanza|%\ttverbatim| breakable chars won't work because we are in
% the item's label's \incmd\hbox. \nostanza
\scanverb*{##5}%
\}}}%
\IfValueT{##6}{%
\addtomacro\gmd@ea@ewrap{\{%
%\nostanza|%\ttverbatim| breakable chars won't work because we are in
% the item's label's \incmd\hbox. \nostanza
\scanverb*{##6}%
\}}}%
\def\gmd@blubra{%
\addtomacro\gmd@ea@bwrap{%
\begingroup
\relaxen\gmd@ea@hashes
\@namedef{the\@enumctr}{\<ign.>}%
}%
\prependtomacro\gmd@ea@ewrap{%
\endgroup}%
\addtomacro\gmd@ea@ewrap{%
\global \advance \csname c@\@enumctr\endcsname \m@ne
}%
\emptify\gmd@blubra
}%
\gmu@ifsbintersect {##2}{Ii}{\gmd@blubra}{}%
\gmu@ifsbintersect {##3}{Ii}{\gmd@blubra}{}%
\gmu@ifxany ##4{\afterassignment}{\gmd@blubra}{}%
\item\relax}%
% \stanza
\IfNoValueT{#1}{\@ifnextac\@gobble{}}% to gobble a~possible active line end or
% active \hathat{A} or \hathat{B} that might
% have occurred because of \incmd\futurelet\ of the optional
% argument checker.
}% of begin definition \ResumeDef
{\endenumerate}
\newenvironment*{enumargs*}{%
\def\gmd@ea@wraps{% \HideDef
\def\gmd@ea@bwrap{ }\def\gmd@ea@ewrap{ }}% \ResumeDef
\enumargs}{\endenumargs}
\def\CharacterTable{\begingroup
\@makeother\{\@makeother\}%
\Character@Table}
\foone{%
\catcode`\[=1 \catcode`\]=2 %
\@makeother\{\@makeother\}}%
[
\def\Character@Table#1{#2}[\endgroup
\message[^^J^^J gmdoc.sty package:^^J
==== The input file contains the \bslash CharacterTable.^^J
==== If you really need to check the correctness of the chars,^^J
==== please notify the author of gmdoc.sty at the email address^^J
==== given in the legal notice in gmdoc.sty.^^J^^J]%
%^^A~\gmd@countlines[#1]\gmd@countlines[#2]
]]
\outer\def\DescribeMacro{%
\@bsphack
\begingroup\MakePrivateLetters
\gmd@ifonetoken\Describe@Macro\Describe@Env}
\outer\def\DescribeEnv{%
\@bsphack
\begingroup\MakePrivateOthers\Describe@Env}
\outer\def\Describe{% It doesn't typeset its argument in the point of
% occurrence.
\leavevmode
\@bsphack
\begingroup\MakePrivateLetters
\gmu@ifstar{\MakePrivateOthers\Describe@Env}{\Describe@Macro}}
\long\def\Describe@Macro#1{%
\endgroup
\strut\Text@Marginize*{#1}%
\@usgentryze#1% we declare kind of formatting the entry
\text@indexmacro#1%
\@esphack}
\def\Describe@Env#1{%
\endgroup
\strut\Text@Marginize*{#1}%
\@usgentryze{#1}% we declare the `usage' kind of
% formatting the entry and index the sequence |#1|.
\text@indexenvir{#1}%
\@esphack}
\let\MacroFont\tt
\let\MacroIndent\CodeIndent
\let\MacrocodeTopsep\CodeTopsep
\let\SpecialEscapechar\CodeEscapeChar
\AtEndInput{\@ifundefined{theCodelineNo}{}{\PackageInfo{gmdoc}{The
\string\theCodelineNo\space macro has no effect here, please use
\string\LineNumFont\space for setting the font and/or
\string\thecodelinenum\space to set the number format.}}}
\def\noeffect@info#1{\@ifundefined{#1}{}{\PackageInfo{gmdoc}{^^J%
The \bslash#1 macro is not supported by this package^^J
and therefore has no effect but this notification.^^J
If you think it should have, please contact the maintainer^^J
indicated in the package's legal note.^^J}}}
\AtEndInput{%
\noeffect@info{PrintDescribeMacro}%
\noeffect@info{PrintMacroName}%
\noeffect@info{PrintDescribeEnv}%
\noeffect@info{PrintEnvName}}
\def\CodelineNumbered{\AtBeginDocument{\gag@index}}
\@onlypreamble\CodelineNumbered
\let\CodelineIndex\@pageindexfalse
\@onlypreamble\CodelineIndex
\let\PageIndex\@pageindextrue
\@onlypreamble\PageIndex
\def\DisableCrossrefs{\@bsphack\gag@index\@esphack}
\def\EnableCrossrefs{\@bsphack\ungag@index
\def\DisableCrossrefs{\@bsphack\@esphack}\@esphack}
\newcommand*\AlsoImplementation{\@bsphack
\long\def\StopEventually##1{\gdef\Finale{##1}}% we define
% \incs{Fi\+n\+a\+le} just to expand to the argument of \incs{StopEventually} not
% to to add anything to the end input hook because \incs{Finale} should
% only be executed if entire document is typeset.\nostanza\par
%
% \hangindent\verbatimhangindent \hangafter1\relax
% \gmdnoindent \leftskip\CodeIndent
% |%\init@checksum| is obsolete in
% \pk{gmdoc} at this point: the \env{CheckSum} counter is reset just at
% the beginning of (each of probably numerous) input(s).
% \nostanza\par
\@esphack}
\AlsoImplementation
\def\OnlyDescription{\@bsphack\long\def\StopEventually##1{%
% \begin{quotation} In this case the argument of |\StopEventually|
% should be set and afterwards \TeX\ should stop reading from
% this file. Therefore we finish this macro with\end{quotation}
##1���endinput}\@esphack}
\@relaxen\Finale
\let\SpecialMainIndex=\DefIndex
\def\SpecialMainEnvIndex{\csname CodeDefIndex\endcsname*}% we don't
\let\SpecialIndex=\CodeCommonIndex
\let\SpecialUsageIndex=\TextUsgIndex
\def\SpecialEnvIndex{\csname TextUsgIndex\endcsname*}
\def\SortIndex#1#2{\index{#1\actualchar#2}}
\def\verbatimchar{&}
\@ifundefined{main}{}{\let\DefEntry=\main}
\@ifundefined{usage}{}{\let\UsgEntry=\usage}
\newcounter{StandardModuleDepth}
\noeffect@info{DocstyleParms}
\@emptify\DontCheckModules \noeffect@info{DontCheckModules}
\@emptify\CheckModules \noeffect@info{CheckModules}
\@emptify\AltMacroFont \noeffect@info{AltMacroFont}
\def\MakePercentIgnore{\catcode`\%9\relax}
\def\MakePercentComment{\catcode`\%14\relax}
\def\OldDocInput{% \changes{v0.99g}{2007/11/11}{obsolete redefinition
% of the \env{macro} environment removed}
\AtBegInputOnce{\StraightEOL
\let\@MakeShortVerb=\old@MakeShortVerb
% \label{oldmkshvrrb}
\OldMacrocodes}%
\bgroup\@makeother\_% it's to allow |_| in the filenames. The next
% macro will close the group.
\Doc@Input}
%% \AVerySpecialMacro % delete the first % when.\unskip|..|\CDPerc
\DeclareCommand\OCRInclude{O{}mO{}}{%
\Store@Macro\incl@DocInput
\def\incl@DocInput##1{%
\begingroup
\CodeSpacesBlank
\@beginputhook
\title{\currentfile}\maketitle
\noverbatimspecials
\relaxen\@xverbatim
\relaxen\check@percent
\Restore@Macro\@verbatim
\verbatimleftskip\z@skip
\verbatim
\@makeother\|% because |\ttverbatim| doesn't do that.
\texcode@hook% we add some special stuff, e.g.\ in \pk{gmdocc.cls} we
\@input{##1}%
\endgroup}%
\csname\@dc@InnerName\DocInclude\endcsname{#1}{#2}{#3}%
\Restore@Macro\incl@DocInput
}
\foone{\catcode`\^^M\active }{%
\def\@NoEOF#1^^M{%
\@relaxen\EOFMark���endinput}%
\def\@EOF#1^^M{���endinput}}
\def\NoEOF{\QueerEOL\@NoEOF}
\def\EOF{\QueerEOL\@EOF}
\endinput
%%
%% End of file `gmdoc.sty'.