new package gmdoc (18dec06)

git-svn-id: svn://tug.org/texlive/trunk@2803 c570f23f-e606-0410-a88d-b1316a301751

2 files changed, 6572 insertions, 0 deletions

diff --git a/Master/texmf-dist/tex/latex/gmdoc/gmdoc.sty b/Master/texmf-dist/tex/latex/gmdoc/gmdoc.sty
new file mode 100644
index 00000000000..cee0aa6b911
--- /dev/null
+++ b/Master/texmf-dist/tex/latex/gmdoc/gmdoc.sty
@@ -0,0 +1,6265 @@
+%
+% \GetFileInfo{gmdoc.sty}
+% \title{The \pk{gmdoc.sty} Package\thfileinfo}
+% \author{Grzegorz `Natror' Murzynowski}
+% \date{\today}
+% \maketitle
+% \label{gmd}
+%% 
+%% This is (a~documentation of) file \pk{gmdoc.sty},
+%% intended to be used with \LaTeXe\ as a~package for
+%% documenting \LaTeXpar\ files and to be documented with itself. 
+%% \stanza
+%% 
+% \begin{copyrnote}
+%
+%% Written by Grzegorz `Natror' Murzynowski,
+%% natror at o2 dot pl
+%%
+%% \copyright\,2006 by Grzegorz `Natror' Murzynowski.
+%%
+%% This program is subject to the \LaTeX\ Project Public License.
+%% See \url{http://www.ctan.org/tex-archive/help/Catalogue/licenses.lppl.html} ^^A
+%% for the details of that license.
+%%
+%% LPPL status: "author-maintained".\par
+%%
+%% Many thanks to my \TeX\ Guru Marcin Woli\'nski for his \TeX nical ^^A
+%% support.
+%
+%\end{copyrnote}
+%
+% \ChangesStart{v0.98c}{1000/0/0}
+%
+% \chschange{v0.96}{2006/08/22}{2395}
+% \chschange{v0.97}{06/09/04}{3636}
+% \chschange{v0.98}{06/09/04}{3636}
+% \chschange{v0.98a}{06/09/06}{4092}
+% \chschange{v0.98b}{06/9/7}{4119}
+% \chschange{v0.98c}{06/9/10}{4299}
+% \chschange{v0.98d}{06/9/15}{4409}
+% \chschange{v0.98e}{06/9/23}{4420}
+% \chschange{v0.98f}{06/9/30}{4430}
+% \chschange{v0.98g}{06/10/5}{4294}
+% \chschange{v0.98i}{06/10/14}{4352}
+% \chschange{v0.98j}{06/10/17}{4372}
+% \chschange{v0.98k}{06/10/21}{4416}
+% \chschange{v0.98l}{06/10/27}{4401}
+% \chschange{v0.98m}{06/11/14}{4349}
+% \chschange{v0.99}{06/11/29}{4462}
+% \chschange{v0.99a}{2006/12/1}{4479}
+% 
+%^^A \docstrip
+%<package>
+
+\NeedsTeXFormat{LaTeX2e}
+\ProvidesPackage{gmdoc}
+         [2006/12/30 v0.99a a documenting package (GM)]
+
+%
+%
+% \CharacterTable {In my \TeX\ Guru's opinion the idea of
+%   checking of the correctness of chars is a~bit obsolete
+%   nowadays. Therefore I~define the \CharacterTable<#1>
+%   macro to gobble its argument and to typeout a~request
+%   for request.}
+%
+%
+% \newcommand*\docfm{\pk{doc}}
+% \newcommand*\gmdoc{\pk{gmdoc}}
+%
+% \tableofcontents
+%
+%
+% \division{Readme}
+%
+%   This package is a~tool for documenting of \LaTeXpar\
+%   packages, classes etc., i.e., the \file{.sty}, \file{.cls} files
+%   etc. The author just writes the code and adds the commentary 
+%   preceded with |%| sign (or another properly declared). No special
+%   environments are necessary.
+%
+%   The package tends to be (optionally) compatible with the standard
+%   \pk{doc.sty} package, 
+%   i.e., the \file{.dtx} files are also compilable with \pk{gmdoc}
+%   (they may need very little adjustment, in some rather special
+%   cases).
+%
+%   The tools are integrated with \pk{hyperref}'s advantages such as
+%   hyperlinking of index entries, contents entries and
+%   cross-references.
+%
+% \begin{gmlonely}
+%   \subdivision{Installation}
+%
+%   Just put the \pk{gmdoc.sty} and \pk{gmdocc.cls} somewhere in the
+%   \file{texmf/\:tex/\:latex} branch. Creating
+%   a~\file{texmf/\:tex/\:latex/\:gm} directory may be advisable if
+%   you consider using other packages written by me.
+%
+%   You should also install \pk{gmverb.sty}, \pk{gmutils.sty}
+%   and \pk{gmiflink.sty} (e.g., put them into the same \file{gm}
+%   directory). These packages are available on CTAN as separate
+%   \file{.zip} archives.
+%
+%   Moreover, you should put the \file{gmglo.ist} file, a MakeIndex
+%   style for the changes' history, into some \file{texmf/\:makeindex}
+%   (sub)directory.
+%
+%   Then you should refresh your \TeX\ distribution's files' database
+%   most probably.
+% \end{gmlonely}
+%
+%
+% \subdivision{Contents of the \pk{gmdoc.zip} Archive}
+%
+% The distribution of the \pk{gmdoc} package consists of the
+% following six files.
+% \begin{verse}
+%   \pk{gmdoc.sty}\\
+%   \pk{gmdocc.cls}\\
+%   \pk{gmglo.ist}\\
+%   \pk{README}\\
+%   \pk{gmdocDoc.tex}\\
+%   \pk{gmdocDoc.pdf}
+% \end{verse}
+%
+% 
+% \begin{gmlonely}
+%   \subdivision{Compiling the Documentation}
+%
+%   The last of the above files (the \pk{.pdf}, i.e., \emph{this
+%     file}) is a~documentation compiled from the \pk{.sty} and
+%   \pk{.cls} files by running \LaTeX\ on the \file{gmdocDoc.tex}
+%   twice, then MakeIndex on the \file{gmdocDoc.idx} and
+%   \file{gmdocDoc.glo} files, and then \LaTeX\ on \file{gmdocDoc.tex}
+%   once more.
+%
+%   MakeIndex shell commands:
+%\begin{verbatim}
+%makeindex -r gmdocDoc
+%makeindex -r -s gmglo.ist -o gmdocDoc.gls gmdocDoc.glo
+%\end{verbatim}
+%   The |-r| switch is to forbid MakeIndex to make implicit ranges since
+%   the (code line) numbers will be hyperlinks.
+%
+%   Compiling the documentation requires the packages:
+%   \pk{gmdoc} (\pk{gmdoc.sty} and \pk{gmdocc.cls}), \pk{gmutils.sty},
+%   \pk{gmverb.sty}, \pk{gmiflink.sty} and also some standard packages:
+%   \pk{hyperref.sty}, \pk{color.sty}, \pk{geometry.sty},
+%   \pk{multicol.sty}, \pk{lmodern.sty}, \pk{fontenc.sty} that should
+%   be installed on your computer by default.
+%
+%   If you had not installed the \pk{mwcls} classes (available on
+%   CTAN and present in \TeX\ Live e.g.), the result of your
+%   compilation might differ a~bit from the \pk{.pdf} provided in this
+%   \pk{.zip} archive in formatting: If you had not installed
+%   \pk{mwcls}, the standard \pk{article.cls} class would be used.
+% \end{gmlonely}
+%
+% 
+% % \subdivision{Bonus: \file{base} Drivers}
+%
+% As a~bonus and example of \docfm-compatibility 
+% there are driver files included
+% (cf.\ Palestrina, \textit{Missa papae Marceli} ;-):
+%
+% \begin{verse}
+%   \pk{source2e_gmdoc.tex}\\
+%   \pk{docstrip_gmdoc.tex}\\
+%   \pk{doc_gmdoc.tex}
+%
+%   \pk{oldcomm.sty}\\
+%   \pk{gmeometric.sty}
+% \end{verse}
+%
+% These drivers typeset the respective files from the
+%\[\hbox{\file{ .../texmf-dist/source/latex/base}}\]
+% directory of the \TeX Live2005 distribution.
+%
+% Probably you should redefine the |\BasePath| macro in them
+% so that it points that directory on your computer.
+%
+%
+%\division*{Introduction}
+%
+% There are very sophisticated and effective tools for documenting
+% \LaTeX\ macro packages, namely the \pk{doc} package and the
+% \pk{ltxdoc} class.
+% Why did I~write another documenting package then? 
+%
+% I~like comfort and \pk{doc} is not comfortable enough for me. It
+% requires special marking of the macro code to be properly typeset
+% when documented. I~want \TeX\ to know `itself' where the code begins
+% and ends, without additional marks. 
+%
+% That's the difference. One more difference, more important for the
+% people for whom the \docfm's conventions are acceptable, is that
+% \gmdoc\ makes use of \pk{hyperref} advantages and makes
+% a~hyperlinking index and toc entries and the
+% cross-references, too. (The CSs in the code maybe in the future.) 
+%
+% The rest is striving to level the very high \pk{doc/ltxdoc}'s
+% standard, such as (optional) numbering of the codelines and
+% authomatic indexing the control sequences e.g.
+%
+% The \docfm\ package was and still is a~great inspiration for me and
+% I~would like this humble package to be considered as a~sort of
+% hommage to it\footnote{As Grieg's Piano Concerto is a~hommage to ^^B
+%   the Schumann's.}. If I~mention copying some code or narrative but do
+% not state the source explicitly, I~mean the \docfm\ package's
+% documentation (I~have v2.1b dated 2004/02/09).
+%
+%
+%
+% \division{The User Interface}
+%
+%
+% \subdivision{Used Terms}
+%
+% When I~write of a~\textbf{macro}, I~mean a~macro in \TeXbook's
+% meaning, i.e., a~control sequence whose meaning is
+% |\(e/g/x)def|ined. By a~\textbf{macro's parameter} I~mean each of
+% |#|\<digit>s in its definition. When I~write about ^^A\)
+% a~\textbf{macro's argument}, I~mean the value (list of tokens)
+% subsituting the corresponding parameter of this macro. 
+% (These understandings are according ^^A(
+% to \TeXbook, I~hope: \TeX\ is a~religion of Book ;-)\,.)
+%
+% I'll use a~shorthand for `control sequence', \textbf{CS}. 
+%
+% When I~talk of a~\textbf{declaration}, I~mean a~macro that expands
+% to a~certain assignment, such as |\itshape| or
+% |\@onlypreamble{|\<CS>|}|.
+%
+% Talking of declarations, I'll use the \textbf{OCSR} acronym as
+% a~shorthand for 'observes/ing common \TeX\ scoping rules'.
+%
+% By a~\textbf{command} I~mean a~certain abstract visible to the end
+% user as a~CS but consisting possibly of more than one macro. I'll
+% talk of a~\textbf{command's argument} also in
+% the `sense\:-for\:-the\:-end\:-user', e.g., I'll talk of the
+% |\verb| \emph{command's} argument although \emph{the macro}
+% |\verb| has no |#|\<digit> in its definition.
+%
+% The \textbf{code} to be typeset verbatim (and with all the bells
+% and whistles) is everything that's not commented out in the source
+% file and what is not a~leading space(s).
+% 
+% The \textbf{commentary} or \textbf{narrative} is everything after
+% the comment char till the end of a~line. The \textbf{comment char}
+% is a~character the |\catcode| of which is 14 usually i.e., when the
+% file works; if you don't play with
+% the |\catcode|s, it's just the |%|. When the file is documented with
+% \gmdoc, such a~char is re|\catcode|d and its r\^ole is else: it
+% becomes the \textbf{code delimiter}.
+%
+% A~line containing any \TeX\ code (not commented out) will be called
+% a~\textbf{codeline}. A~line that begins with (some leading spaces
+% and) a~code delimiter will be called a~\textbf{comment line} or
+% \textbf{narration line}. 
+%
+% The \textbf{user} of this package will also be addressed
+% as \textbf{you}.\stanza
+%
+% Not to favour any particular gender (of the amazingly rich variety,
+% I~mean, not of the vulgarly simplified two-element set), in this
+% documentation I~use alternating pronouns of third person ^^A(
+% (\TextUsage\heshe\ etc. commands provided by \pk{gmutils}), so let
+% one be not surprised if `\heshe' sees `\himher self' altered in the
+% same sentence :-)\,.
+% 
+%
+% \subdivision{Preparing the Source File}
+%
+%  When \LaTeXpar\ with \pk{gmdoc.sty} package loaded typesets the
+% comment lines, the code delimiter is ommitted. If the comment
+% continues a~codeline, the code delimiter is printed. It's done so
+% because ending a~\TeX\ code line with a~|%| is often very important
+% and significant. Comments longer than one line are typeset
+% continuously with the code delimiters ommitted.
+%
+% The user should just write \hisher\ splendid code and brilliant
+% commentary. In the latter \heshe\ may use usual \LaTeXpar\ commands.
+% The only requirement is, if an argument is divided in two
+% lines, to end such a~dividing line with \TextUsage*{^^B} sequence
+% that'll enter the (active) \<char2> which shall gobble the line end.
+% ^^A~and the leading code delimiter of the next line---obsoleted by
+% ^^A~making % ignored in the commentary.
+%
+% Moreover, if \heshe\ wants to add a~meta-comment i.e., a~text that
+% doesn't appear in the code layer nor in the narrative, \heshe\ may
+% use the \TextUsage*{^^A} sequence that'll be read by \TeX\ as
+% \<char1>, which 
+% is in \gmdoc\ active and defined to gobble the stuff between itself
+% and the next line end. 
+%
+% Note, however, that both |^^A| and |^^B| are usually macros in
+% \gmdoc\ and the text being gobbled is their argument so it has to be
+% balanced of braces and, if it occures inside an |\if...|, it has to
+% have all |\if...|s coupled with |\else|s and |\fi|s.
+% 
+% However, it may be a~bit confusing
+% for someone acquainted with the \docfm\ conventions. If you don't
+% fancy the |^^B| special sequence, instead you may restore the
+% standard meaning of the line end with the \TextUsage\StraightEOL\
+% declaration which
+% OCSR. As almost all the control
+% sequences, it may be used also as an environment, i.e.,
+% |\begin{StraightEOL}|\ \dots\ |\end{StraightEOL}|. However, if for any
+% reason you don't want to make an environment (a~group), there's
+% a~|\StraightEOL|'s counterpart, the \TextUsage\QueerEOL\ declaration that
+% restores again the 
+% \StraightEOL
+% queer\footnote{In my understanding
+%   `queer' and `straight' are not the opposites excluding each
+%   other but the counterparts that may cooperate in harmony for
+%   people's good. And, as I~try to show with the \cs{QueerEOL} and
+%   \cs{StraightEOL} declarations, `queer' may be very useful and
+%   recommended while `straight' is the standard but not necessarily
+%   normative. (Remember also Alice's in the Wonderland exclamations
+%   ``What a~queer day is today''.)}^^A
+% \QueerEOL
+% \gmdoc's meaning of the line end. It OCSR, too.  One more point to use
+% |\StraightEOL| is where you wish some code lines to be executed
+% both while loading the file and during the documentation pass
+% (it's analogous to \docfm's not embracing some code lines in
+% a~\env{macrocode} environment).
+%
+% \stanza
+%
+% As in standard \TeX ing, one gets a~paragraph by a~blank line.
+% Such a~line should be |%|ed of course. A~fully blank line is
+% considered a~blank \emph{code line} and hence results in
+% a~vertical space in the documentation. As in the environments for
+% poetry known to me, subsequent blank lines do not increase such
+% a~space.
+%
+% Then \heshe\ should prepare a~main document file,
+% a~\textbf{driver} henceforth, to set all the
+% required formattings such as |\documentclass|, paper size
+% etc., and load this package with a~standard command
+% i.e., |\usepackage{gmdoc}|, just as \docfm's documentation says:
+%
+% \begin{quotation}
+%   If one is going to document a set of macros with the \pk{[gm]doc}
+%   package one has to prepare a special driver file which produces
+%   the formatted document. This driver file has the following
+%   characteristics: \par\leftskip\CodeIndent
+%   \stanza
+%
+%   \noindent|\documentclass[|\<options>|]{|\<document-class>|}|\par
+%   \noindent|\usepackage[|\<options, probably none>|]{gmdoc}|\par
+%   \<preamble>\par
+%   \noindent|\begin{document}|\par
+%     \<special input commands>\par
+%     \noindent|\end{document}|\par
+% \end{quotation}
+%
+%
+% \subdivision{The Main Input Commands}
+%
+% To typeset a~source file you may use the \TextUsage\DocInput\
+% macro that takes the (path and) name of the file \emph{with ^^B
+%   the extension} as the only argument, e.g.,
+% |\DocInput{mybrilliantpackage.sty}|.
+%
+% (Note that an \emph{installed} package or class file is findable to
+% \TeX\ even if you don't specify the path.)
+%
+% If a~source file is written with rather \docfm\ than \gmdoc\ in
+% mind, then the \TextUsage\OldDocInput\ command may be more
+% appropriate (e.g., if you break the arguments of commands in the
+% commentary in lines). It also takes the file (path and) name as the
+% argument. 
+%
+% When using |\OldDocInput|, you have to wrap all the code in
+% \TextUsage*{macrocode} environments, which is not necessary when you
+% use |\DocInput|. Moreover, with |\OldDocInput| the
+% \env{macrocode(*)} environments require to be ended with
+% |%    \end{macrocode(*)}|.
+%
+% \stanza
+% If you wish to document many files in one document, you are
+% provided \TextUsage\DocInclude\ command, analogous to \LaTeX's
+% |\include| and very likely to \pk{ltxdoc}'s command of the same
+% name. In \gmdoc\ it has one mandatory argument that should be the
+% file name \emph{without extension}, just like for |\include|.
+%
+% The file extensions supported by |\DocInclude| are \pk{.fdd},
+% \pk{.dtx}, \pk{.cls}, \pk{.sty}, \pk{.tex} and \pk{.fd}. The macro
+% looks for 
+% one of those 
+% extensions in the order just given. If you need to document files
+% of other extensions, please let me know and most probably we'll make
+% it possible.
+%
+% |\DocInclude| has also an optional first argument that is intended
+% to be the path of the included file with the levels separated by |/|
+% (slash) and also ended with a~slash. The path given to |\DocInclude| as
+% the first and optional argument will not appear in the headings nor
+% in the footers. 
+%
+% |\DocInclude| redefines \TextUsage\maketitle\ so
+% that it makes a~chapter heading or, in the classes that don't
+% support |\chapter|, a~part heading, in both cases with
+% respective toc entries. The default assumption is that all the
+% files have the same author(s) so there's no need to print them in
+% the file heading. If you wish the authors names to be printed, you
+% should write \TextUsage\PrintFilesAuthors\ in the preamble or before
+% the relevant |\DocInclude|s. If you wish to undeclare printing the
+% authors names, there is \TextUsage\SkipFilesAuthors\ declaration.
+%
+% Like in \pk{ltxdoc}, the name of an included file appears in the
+% footer of each page with date and version info (if they are
+% provided).
+%
+% The |\DocInclude|d files are numbered with the letters, the
+% lowercase first, as in \pk{ltxdoc}. Such a~filemarker also precedes
+% the index entries, if the (default) codeline index option is in force.
+%
+% As with |\include|, you may declare
+% \TextUsage\includeonly|{|\<filenames separated by commas>|}| for
+% the draft versions.
+%
+% \stanza
+% If you wish to include the driver file into your documentation, you
+% may write |\DocInput{\jobname.tex}|, but a~try of
+% |\DocInclude{\jobname}| would result with input stack overflow
+% caused by infinite |\input{|\<jobname>|.aux}| recursion.
+% But there's \TextUsage\SelfInclude\ at your service that creates and
+% uses \pk{\<jobname>.auxx} file instead of the usual
+% \pk{\<jobname>.aux}. Its effect is analogous to the |\DocInclude|'s, but
+% \emph{the arguments it takes are totally different}: Since the
+% filename is known, there's no need to state it. The extension is
+% assumed to be \file{.tex}, but if it's different, you may state it in
+% the first and optional argument. The second argument is mandatory and
+% it's the stuff to be put at begin of file input, this one and no
+% else (with \TextCommonIndex\AtBegInputOnce|\AtBegInputOnce| hook).
+% For the example of usage see line \ref{SelfIncludeUsg} of
+% chapter~\ref*{Driver}.
+% 
+% \stanza
+% 
+% At the default settings, the |\Doc/SelfInclude|d files constitute
+% chapters if |\chapter| is known and parts otherwise. The
+% |\maketitle|s of those files result in the respective headings.
+%
+% If you prefer more \pk{ltxdoc}ish look, in which the files always
+% constitute the parts and those parts have a~part's title
+% pages with the file name and the files' |\maketitle|s result in
+% (article-like) titles not division headings, then you are provided
+% the \TextUsage\ltxLookSetup\ declaration (allowed only in the
+% preamble). However, even after this declaration the files will be
+% included according to \gmdoc's rules not necessarily to the \docfm's
+% ones (i.e., with minimal marking necessary at the price of active
+% line ends (therefore not allowed between a~command and its argument
+% nor inside an argument)).
+%
+% On the other hand, if you like the look offered by me but you have
+% the files prepared for \docfm\ not for \gmdoc, then you
+% should declare \TextUsage\olddocIncludes. Unlike the previous one,
+% this may be used anywhere, because I~have the account of including both
+% \docfm-like and \gmdoc-like files into one document. This
+% declaration just changes the internal input command and doesn't
+% change the sectioning settings.
+%
+% It seems possible that you wish to document the `old-doc' files
+% first and the `new-doc' ones after, so the above declaration has its
+% counterpart, \TextUsage\gmdocIncludes, that may be used anywhere,
+% too. Before the respective |\DocInclude|(s), of course.
+%
+% Both these declarations OCSR.
+%
+% If you wish to document your files as with \pk{ltxdoc} \emph{and}
+% as with \docfm, you should declare
+% |\ltxLookSetup| in the preamble \emph{and} |\olddocIncludes|.
+%
+% \stanza
+% Talking of analogies with \pk{ltxdoc}, if you like only the page
+% layout provided by that class, there is the
+% \TextUsage\ltxPageLayout\ declaration (allowed only in preamble)
+% that only changes the margins and the text width (it's intended to
+% be used with the default paper size). This declaration is
+% contained in the |\ltxLookSetup| declaration.
+%
+% \stanza
+% If you need to add something at the beginning of the input of file,
+% there's the \TextUsage\AtBegInput\ declaration that takes one and
+% mandatory argument which is the stuff to be added. This declaration
+% is global. It may be used more than one time and the arguments of
+% each occurrence of it add up and are put at the beginning of input
+% of every subsequent files. 
+%
+% Simili modo, for the end of input, there's the
+% \TextUsage\AtEndInput\  declaration, also one-argument, global and
+% cumulative.
+%
+% If you need to add something at the beginning of input of only one
+% file, put before the respective input command an
+% \TextUsage\AtBegInputOnce|{|\<the stuff to be added>|}|
+% declaration. It's also global which means that the groups do not
+% limit its scope but it adds its argument only at the first input
+% succeding it (the argument gets wrapped in a~macro that's |\relax|ed 
+% at the first use). |\AtBegInputOnce|s add up, too.
+%
+% \stanza
+% One more input command is \TextUsage\IndexInput\ (the name and idea
+% of effect comes from \docfm). It takes the same argument as
+% |\DocInput|, the file's (path and) name with extension. (It
+% \emph{has} |\DocInput| inside). It works properly if the input file
+% doesn't contain explicit \<char1> (|^^A| is OK). 
+%
+% The effect of this command is typesetting of all the input file
+% verbatim, with the code lines numbered and the CSs automatically
+% indexed (\pk{gmdoc.sty} options are in force).
+%
+%
+% \subdivision{Package Options}
+%
+% As many good packages, this also provides some options:
+% \stanza
+%
+% Due to best \TeX\ documenting traditions the codelines will be
+% numbered. But if the user doesn't wish that, \heshe\ may turn it off
+% with the \TextUsage*{linesnotnum} option.
+%
+% However, if \heshe\ agrees to have the lines numbered, \heshe\ may wish
+% to reset the counter of lines \himher self, e.g., when \heshe\
+% documents many source files in one document. Then \heshe\ may wish
+% the line numbers to be reset with every |{section}|'s turn for
+% instance. This is the r\^ole of the \TextUsage*{uresetlinecount}
+% option, which seems to be a~bit obsolete however, since the
+% \TextCommonIndex\DocInclude|\DocInclude| command takes care of
+% a~proper reset.
+%
+% Talking of line numbering further, a~tradition seems to exist to
+% number only the codelines and not to number the lines of
+% commentary. That's the default behaviour of \pk{gmdoc} but, if
+% someone wants the comment lines to be numbered too, \heshe\ is
+% provided the \TextUsage*{countalllines} option. ^^A(
+% {\LineNumFont 441}\,Then the narration acquires a~bit biblical look ;-),
+% {\LineNumFont 442}\,as shown in this short example. This option is intended
+% {\LineNumFont 443}\,for the draft versions and it is not perfect (as if anything
+% {\LineNumFont 444}\,in this package was). As you see, the lines 
+% {\LineNumFont 445}\,are typeset continuously with the numbers printed.
+%
+% \stanza
+% By default the \pk{makeidx} package is loaded and initialized and
+% the CSs occurring in the code are automatically
+% (hyper)indexed thanks to the \pk{hyperref} package. If the user
+% doesn't wish to index anything, she should use the
+% \TextUsage*{noindex} option.
+%
+% The index comes two possible ways: with the line numbers (if the
+% lines are numbered) and that's the default, or with the page numbers, if
+% the \TextUsage*{pageindex}\label{indexoptions} option is set.
+%
+% \stanza
+% By default, \gmdoc\ excludes some 300 CSs from being indexed.
+% They are the most common CSs, \LaTeX\ internal
+% macros and \TeX\ primitives. To learn what CSs are
+% excluded actually, see lines \ref{DIE1}--\ref{DIE2}.
+%
+% If you don't want all those exclusions, you may turn them off with
+% the \TextUsage*{indexallmacros} option. 
+%
+% If you have ambiguous feelings about whether to let the default
+% exclusions or forbid them, see p.\,\pageref{IEIDeclarations} to
+% feed this ambiguity with a~couple of declarations.
+%
+% \stanza
+% In \docfm\ package there's a~default behaviour of putting marked
+% macro's or environment's name to a~marginpar. In the standard
+% classes it's allright but not all the classes support marginpars.
+% That is the reason why this package enables marginparing when in
+% standard classes, enables or disables it due to the respective option
+% when with Marcin Woli\'nski's classes and in any case provides the
+% options \TextUsage*{withmarginpar} and \TextUsage*{nomarginpar}. So, in
+% non-standard classes the default behaviour is to disable marginpars.
+% If the marginpars are enabled in \pk{gmdoc}, it will put marked
+% control sequences and environments into marginpars (see
+% \gmiflink[textUsage]{\cs{TextUsage} etc.}). These options do
+% not affect common using marginpars, which depends on the
+% documentclass.
+%
+% \stanza
+% My suggestion is to make the spaces in the code visible except the
+% leading ones and that's the default. But if you wish all the code
+% spaces to be blank, I~give the option \TextUsage*{codespacesblank}
+% reluctantly. Moreover, if you wish the code spaces to be blank only
+% in some areas, then there's \TextUsage\CodeSpacesBlank\ declaration
+% (OCSR).
+%
+%
+% \subdivision{The Packages Required}
+%
+% \pk{gmdoc} requires (loads if they're not loaded yet) some other
+% packages of mine, namely \pk{gmutils}, \pk{gmverb}, analogous to
+% Frank Mittelbach's \pk{shortvrb}, and \pk{gmiflink} for conditional
+% making of hyperlinks. It also requires \pk{hyperref},
+% \pk{multicol}, \pk{color} and \pk{makeidx}. 
+%
+% The \pk{gmverb}\gmdmarginpar{\pk{gmverb}} package redefines the
+% |\verb| command and the \env{verbatim} environment in such a~way that 
+% | |, |{| and |\| are breakable, the first with no `hyphen' and the
+%   other two with the comment char as a~hyphen, i.e.,
+%   |{|\<subsequent text>|}| breaks into 
+%     |{%|\\ \<subsequent text>|}| ^^A] bal. braces for Emacs
+%       and \<text>|\mylittlemacro| breaks into \<text>|%|\\
+%       |\mylittlemacro|.
+%
+% As the standard \LaTeX\ one, my |\verb| issues an error when a~line
+% end occures in its scope. But, if you'd like to allow line ends in
+% short verbatims, there's the \TextUsage\verbeolOK\ declaration.  The
+% plain |\verb| typesets spaces blank and |\verb*| makes them visible,
+% as in the standard version(s).
+%
+% Moreover, \pk{gmverb} provides the \TextUsage\MakeShortVerb\
+% declaration 
+% that takes a~one-char control sequence as the only argument and
+% turns the char used into a~short verbatim delimiter, e.g., after
+% \[\hbox{\verb+\MakeShortVerb*\|+}\]
+% (as you see, the declaration has the starred version, which is for
+% visible spaces, and non-starred for blank spaces) to get
+% |\mylittlemacro| you may type \verb+|\mylittlemacro|+ instead of
+% |\verb+\mylittlemacro+|. Because the char used in the last example
+% is my favourite and is used this way by DEK in \TeXbook's
+% format, \pk{gmverb} provides a~macro \TextUsage\dekclubs\ that
+% expands to the example displayed above.
+%
+% Be careful because such active chars may interfere with other
+% things, e.g., the \verb+|+ with the vertical line marker in \env{tabular}s and
+% with the \pk{tikz} package. If this happens, you can declare e.g.,
+% \TextUsage\DeleteShortVerb\verb+\|+ and the previous meaning of the
+% char used shall be restored.
+%
+% One more difference between \pk{gmverb} and \pk{shortvrb} is that
+% the chars |\active|ated  by |\MakeShortVerb|, behave as if they were
+% `other' in math mode, so you may type e.g., \verb+$k|n$+ to get
+% $k|n$ etc.
+% \stanza
+%
+% The \pk{gmutils}\gmdmarginpar{\pk{gmutils}} package provides a~couple of
+% macros similar to some basic \LaTeXpar\ ones, rather strictly
+% technical and (I~hope) tricky, such as |\afterfi|, |\ifnextcat|,
+% |\addtomacro| etc. It's this package that provides the macros for
+% formatting of names of macros and files, such as |\cs|, |\marg|, |\pk|
+% etc.  \stanza
+% 
+% The \gmdoc\ package uses a~lot of hyperlinking possibilities
+% provided by \pk{hyperref}\gmdmarginpar{\pk{hyperref}} which is therefore
+% probably the most important package required. The recommended
+% situation is that the user loads \pk{hyperref} package with \hisher\
+% favourite options \emph{before} loading \pk{gmdoc}.
+%
+% If \heshe\ does not, \pk{gmdoc} shall load it with \emph{my}
+% favourite options. 
+%
+% To avoid an error if a~(hyper)referenced label does not exist,
+% \pk{gmdoc} uses the \pk{gmiflink}\gmdmarginpar{\pk{gmiflink}}
+% package. It works e.g., in the index when the codeline numbers have
+% been changed: then they are still typeset, only not as hyperlinks
+% but as a~common text.
+%
+% \stanza
+% To typeset the index and the change history in balanced columns
+% \gmdoc\ uses the \gmdmarginpar{\pk{multicol}}\pk{multicol} package
+% that seems to be standard these days.
+%
+% Also the \gmdmarginpar{\pk{color}}\pk{multicol} package, required to
+% define the default colour of the hyperlinks, seems to be standard
+% already, and \pk{makeidx}.
+%
+%
+% \subdivision[Macros for Marking the Macros]{^^B
+%   \gmhypertarget{Macros for Marking the Macros}}
+%
+% The concept (taken from \docfm ) is to index virtually all the
+% control sequences occurring in the code. \gmdoc\ does that by
+% default and needs no special command. (See below about exluding some
+% macros from being indexed.)
+%
+% The next concept (also taken from \docfm) is to ditinguish some
+% occurrences of some control sequences by putting such a~sequence
+% into a~marginpar and by special formatting of its index entry. That
+% is what I~call marking the macros. \gmdoc\ provides also
+% a~possibility of analogous marking for the environments' names and
+% other sequences such as |^^A|.
+%
+% This package provides two kinds of special formatting of the
+% index entries: `usage', with the reference number italic by default,
+% and `def' (in \docfm\ called `main'), with the reference number
+% roman (upright) and underlined by default. All the reference
+% numbers, also those with no special formatting, are made hyperlinks
+% to the page or the codeline according to the respective indexing
+% option (see p.\,\pageref{indexoptions}).
+%
+% The macros and environments to be marked appear either in the code
+% or in the commentary. But all the definitions appear in the code,
+% I~suppose. Therefore the `def' marking macro is provided only for
+% the code case. So we have\gmhypertarget[textUsage]{} the
+% \TextUsage\CodeDefine, \TextUsage\CodeUsage\ and
+% \TextUsage\TextUsage\ commands.
+%
+% All three take one argument and all three may be starred.  The
+% non-starred versions are intended to take a~control sequence as the
+% argument and the starred to take whatever (an
+% environment name or a~|^^A|-like and also a~CS).
+%
+% You don't have to bother whether |@| is a~letter while documenting
+% because even if not, these commands do make it a~letter, or more
+% precisely, they execute \TextUsage\MakePrivateLetters\ whatever 
+% it does: At the default settings this command makes |*|
+% a~letter, too, so a~starred version of a~command is a~proper argument
+% to any of the three `|\...Define/Usage| commands unstarred.
+%
+% The two |\Code...| commands, if unstarred, mark the next scanned
+% occurrence of their argument in the code. (By `scanned occurrence'
+% I~mean a~situation of the CS having been scanned in the
+% code which happens iff its name was preceded by the char declared as
+% |\CodeEscapeChar|). The starred versions of those commands
+% mark just the next codeline and don't make \TeX\ looks for the
+% scanned occurrence of their argument (which would never happen if
+% the argument is not a~CS). Therefore, if you want to mark
+% a~definition of an environment \env{foo}, you should put
+% \[ \hbox{|%\CodeDefine*{foo}|} \]
+%  right before the code line
+% \[\hbox{|\newenvironment{foo}{%|}\]^^A]
+%     i.e., not separated by any code line. The starred versions of
+%     the |\Code...| commands are also intended to mark implicit
+%     definitions of macros, e.g.,  |\CodeDefine*\@foofalse| before
+%     the line
+% \[\hbox{|\newif\if@foo|.}\]
+%
+% They both are |\outer|.
+%
+% The |\TextUsage| (one-argument) command is intended to mark usage of
+% a~verbatim occurrence of a~\TeX\ object in the commentary. Unlike
+% the two |\Code...|s, it typesets its argument which means among others\ that
+% the marginpar appears usually at the same line as the text you
+% wanted to mark. This command also has the starred version primarily
+% intended for the environments names, and secondarily for
+% |^^A|-likes and CSs, too. Currently, the most important difference
+% is that the unstarred version executes |\MakePrivateLetters| while
+% the starred does both |\MakePrivateLetters| and |\MakePrivateOthers|
+% before reading the argument.
+%
+% If you consider the marginpars a~sort of sub(sub\dots)section
+% marks, then you may wish to have a~command that makes a~marginpar of
+% the desired CS (or whatever) at the beginning of its description,
+% which may be fairly far from the first occurrence of its
+% object. Then you have the \TextUsage\Describe\ command which
+% puts its argument in a~marginpar and indexes it as a~`usage' entry
+% but doesn't print it in the text. It's |\outer|.
+%
+% All four commands  just described put their (|\string|ed) argument
+% into a~marginpar (if the marginpars are enabled) and create an index
+% entry (if indexing is enabled).
+%
+% But what if you want just to make a~marginpar with macro's or
+% environment's name? Then you have \TextUsage\CodeMarginize\
+% to declare what to put into a~marginpar in the \TeX\ code (it's
+% |\outer|) and
+% \TextUsage\TextMarginize\ to do so in the commentary. According to
+% the spirit of this part of the interface, these commands also take one
+% argument and have their starred versions for strings other than
+% control sequences.
+%
+% The marginpars (if enabled) are `reverse' i.e., at the left margin, and
+% their contents is flush right and typeset in a~font declared with
+% \TextUsage\marginpartt. By default, this declaration is |\let| to
+% |\tt| but it may be advisable to choose a~condensed font if there is
+% any. Such a~choice is made by \pk{gmdocc.cls} if the Latin Modern
+% fonts are available: in this case \pk{gmdocc.cls} uses Latin Modern
+% Typewriter Light Condensed.
+% 
+% If you need to put something in a~marginpar without making it typewriter
+% font, there's the \TextUsage\gmdmarginpar\ macro (that takes one
+% and mandatory argument) that only flushes its contents right.
+%
+% \stanza
+% On the other hand, if you don't want to put a~CS (or another
+% verbatim text) in a~marginpar but only to index it, then there are
+% \TextUsage\CodeDefIndex\ and \TextUsage\CodeUsgIndex\ to declare
+% special formatting of an entry. The unstarred versions of these
+% commands look for their argument's scanned occurrence in the code
+% (the argument should be a~CS), and the starred ones just take the
+% next code line as the reference point. Both these commands are
+% |\outer|.
+%
+% In the code all the control sequences (except the excluded ones, see
+% below) are indexed by default so no explicit command is needed for
+% that. But the environments and other special sequences are not and
+% the two commands described above in their |*|ed versions contain the
+% command for indexing their argument. But what if you wish to index
+% a~not scanned stuff as a~usual entry? The \TextUsage\CodeCommonIndex*\
+% comes in rescue, starred for the symmetry with the two previous commands
+%  (without |*| it just gobbles
+% it's argument). It's |\outer|.
+%
+% Similarly, to index a~\TeX\ object occurring verbatim in the narrative, you
+% have \TextUsage\TextUsgIndex\ and \TextUsage\TextCommonIndex\
+% commands with 
+% their starless versions for a~CS argument and the starred for all
+% kinds of the argument.
+%
+% \stanza
+% Moreover, as in \docfm, the \TextUsage*{macro} and 
+% \TextUsage*{environment} environments are provided. Both take one
+% argument that should be a~CS for \env{macro} and `whatever' for
+% \env{environment}. Both add the |\MacroTopsep| glue before and after
+% their contents, and put their argument in a~marginpar at the first
+% line of their contents (since it's done with |\strut|, you should
+% not put any blank line (|%|ed or not) between
+% |\begin{macro/environment}| and the first line of the
+% contents). Then \env{macro} commands the first scanned occurrence of
+% its argument to be indexed as `def' entry and \env{environment}
+% commands \TeX\ to index the argument as if it occurred in the next
+% code line (also as `def' entry).
+%
+% Since it's possible that you define a~CS implicitly i.e., in such a~way that it
+% cannot be scanned in the definition (with |\csname...\endcsname|
+% e.g.) and wrapping such a~definition (and description) in an
+% \env{environment} environment would look misguidedly ugly, there's
+% the \env{macro*} environment which \TeX nically is just an alias for
+% \env{environment}.
+%
+% (To be honest, if you give a~\env{macro} environment a~non-CS
+% argument, it will accept it and then it'll work as
+% \env{evironment}.)
+%
+%
+% \subdivision{Index Ex/Inclusions}
+%
+% It's\label{IEIDeclarations} understandable\footnote{After reading ^^B(
+%   \docfm's documentation ;-)\,.} that you don't
+% want some control sequences to be indexed in your documentation. The
+% \docfm\ package gives a~brilliant solution: the
+% \TextUsage\DoNotIndex\ declaration. So do I
+% (although here, \TeX nically it's done another
+% way). It OCSR. This declaration takes one
+% argument consisting of a~list of control sequences not to be
+% indexed. The items of this list may be separated with commas, as in
+% \docfm , but it's not obligatory. The whole list should come in
+% curly braces (except when it's one-element), e.g.,
+% \[\hbox{\verb+\DoNotIndex{\some@macros,\are* \too\auxiliary\?}+}\]
+% (The spaces after the control sequences are ignored.)
+% You may use as many |\DoNotIndex|es as you wish (about half as many as
+% many CSs may be declared, because for each CS excluded from indexing
+% a~special CS is declared that stores the ban sentence).
+%  Excluding the same CS more than once makes no problem.
+%
+% I~assume you wish most of \LaTeX\ macros, \TeX\ primitives etc.\ to
+% be excluded from your index (as I~do). Therefore \pk{gmdoc} excludes
+% some 300 CSs by default. If you don't like it, just set the
+% |indexallmacros| package option.
+%
+% On the third hand, if you like the default exclusions in general but
+% wish to undo just a~couple of them, you are given \TextUsage\DoIndex\
+% declaration (OCSR) that removes a~ban on all the CSs given in the
+% argument, e.g.,
+% \[\hbox{\verb+\DoIndex{\par \@@par \endgraf}+}\]
+%
+% Moreover, you are provided the \TextUsage\DefaultIndexExclusions\
+% and \Describe\UndoDefaultIndexExclusions^^A
+% \cs{Un\-do\-Def\-ault\-Ind\-ex\-Ex\-cl\-us\-ions} declarations that
+% act according to their names. You may use them in any configuration
+% with the |indexallmacros| option. Both of these declarations OCSR.
+%
+%
+% \subdivision{The \ds\ Directives}
+%
+% \gmdoc\ typesets the \ds\ directives and it does it quite likely as
+% \docfm, i.e., with math sans serif font. It does it automatically
+% whether you use the traditional settings or the new.
+%
+% Advised by my \TeX\ Guru, I~didn't implement the module nesting
+% recognition (MW told it's not that important.)
+%
+% So far \gmhypertarget{verbatim mode directive} is only
+% half-handled. That is, a~line beginning with |%<<|\<END-TAG> will be
+% typeset as a~\ds\ directive, but 
+% the closing line |%|\<END-TAG> will be not. It doesn't hard to
+% implement, I~only receive some message it's really useful for
+% someone.
+%
+%^^A If you use the \TextUsage*{macrocode} environment, you don't have to
+%^^A add any CS to get \ds\ directive typeset properly: within
+%^^A \env{macrocode} a~|<| sign in a~line commented out is read as
+%^^A a~beginning of a~\ds\ directive, which means that \TeX\ looks
+%^^A for the closing |>| or the very next |<| as the beginning of the
+%^^A special directive |<<|\<any text till the end of line>.  In the
+%^^A standard case of |<|\<directive>|>| \TeX\ typesets any further text
+%^^A in its line as the code.
+%^^A
+%^^A If you want the \ds\ directives to be typeset without
+%^^A \env{macrocode}, you shoud declare \TextUsage\docstrips\ to make
+%^^A \TeX\ read the |<| in the comment layer as the beginning of
+%^^A a~\ds\ directive. In |\docstrips|'s scope the |<| signs
+%^^A occurring in math mode remain just inequality signs.
+%^^A
+%^^A If you wish only the first |<| met to start a~\ds\ directive
+%^^A typesetting, then there's \TextUsage\docstrip\ declaration that
+%^^A restores the previous meaning of |<| after the first using it as
+%^^A a~\ds\ directive opener.
+%^^A
+%^^A Both |\docstrips| and |\docstrip| OCSR and, of course you may use
+%^^A them as environments (which doesn't make much sense with the
+%^^A latter, though).
+%^^A
+%
+% \subdivision{The Changes History}
+%
+% The \docfm's documentation reads:
+% \begin{quotation}
+% To maintain a change history within the file, the |\changes| command may
+% be placed amongst the description part of the changed code. It takes three
+% arguments, thus:
+% \[\hbox{|\changes{|\<version>|}{|\<YYYY/MM/DD date>|}{|\<text>|}|}\]
+%
+% The changes may be used to produce an auxiliary file (\LaTeX's
+% |\glossary| mechanism is used for this) which may be printed after
+% suitable formatting. The |\changes| [command] encloses the \<date> in
+% parentheses and appends the \<text> to form the printed entry in
+% such a change history [\dots\ obsolete remark ommitted].
+% 
+% To cause the change information to be written out, include
+% \TextUsage\RecordChanges\ in the driver['s preamble or just in the
+% source file (\pk{gmdocc.cls} does it for you)]. To read in and print
+% the sorted change history (in two 
+% columns), just put the \TextUsage\PrintChanges\ command as the last
+% (commented-out, and thus executed during the documentation pass
+% through the file) command in your package file [or in the driver].
+% Alternatively, this command may form one of the arguments of the
+% |\StopEventually| command, although a change history is probably not
+% required if only the description is being printed. The command
+% assumes that MakeIndex or some other program has processed the
+% \pk{.glo} file to generate a~sorted \pk{.gls} file. You need
+% a~special MakeIndex style file; a suitable one is supplied with
+% \docfm\ [and \gmdoc], called \iffalse\pk{gglo.ist}\fi^^A
+% [\dots\ \textbf{\pk{gmglo.ist}} for \gmdoc]. The \TextUsage\GlossaryMin,
+% \TextUsage\GlossaryPrologue\ and \TextUsage\GlossaryParms\ macros are
+% analagous to the |\Index...| versions [see
+% sec.~\gmiflink[IndexParams]{The Parameters}
+% p.\,\pageref*{IndexParams}]. (The \LaTeX\ `glossary' 
+% mechanism is used for the change entries.)
+% \end{quotation}
+%
+% In \gmdoc\ |\changes| is |\outer|.
+%
+% \stanza
+% As mentioned in the introduction, the glossary, the changes history
+% that is, uses a~special MakeIndex style, \pk{gmglo.ist}. This style
+% declares another set of the control chars but you don't have to
+% worry: |\changes| takes care of setting them properly. To be
+% precise, |\changes| executes \TextUsage\MakeGlossaryControls\ that
+% is defined as
+% \begin{verbatim}
+%  \def\actualchar{=} \def\quotechar{!}%
+%  \def\levelchar{>} \edef\encapchar{\twelveclub}
+%\end{verbatim}
+%
+% Only if you
+% want to add a~control character yourself in a~changes entry, to quote
+% some char, that is (using level or encapsulation chars is not
+% recommended since |\changes| uses them itself), use rather
+% |\quotechar|.
+%
+% \stanza
+% Before writing an entry to the \pk{.glo} file, |\changes|
+% checks if the date (the second mandatory\equals the third argument)
+% is later than the date stored in the counter
+% \TextUsage*{ChangesStartDate}. You may set this counter with a
+% \Describe\ChangesStart^^B
+% \[\hbox{|\ChangesStart{|\<version>|}{|\<year>|/|\<month>|/|\<day>|}|}\]
+% declaration.
+%
+% If the \env{ChangesStartDate} is set to a~date contemporary to \TeX\
+% i.e., not earlier than September 1982\footnote{DEK in \textit{\TeX\ ^^B
+% The Program} mentions that month as of \TeX\ Version 0 release.},
+% then a~note shall appear at the beginning of the changes history
+% that informs the reader of ommitting the earlier changes entries.
+%
+% If the date stored in \env{ChangesStartDate} is earlier than \TeX,
+% no notification of ommitting shall be printed. This is intended for
+% a~rather tricky usage of the changes start date feature: you may
+% establish two threads of the changes history: the one for the users,
+% dated with four digit year, and the other for yourself only, dated
+% with two or three digit year. If you declare 
+% \[\hbox{|\ChangesStart{|\<version?>|}{1000/00/00}|}\]
+% or so, the changes entries dated with less-than-four digit year
+% shall be ommitted and no notification shall be issued of that.
+%
+% \stanza
+%
+% While scanning the CSs in the code, \gmdoc\ counts them and prints
+% the information about their number on the terminal and in
+% \pk{.log}. Moreover, you may declare
+% \TextUsage\CheckSum|{|\<number>|}| before the code and \TeX\ will
+% inform you whether the number stated by you is
+% correct or not, and what it is. As you guess, it's not my original idea
+% but I~took it from \docfm. 
+%
+% There it is provided as a~tool for
+% testing whether the file is corrupted. 
+% My \TeX\ Guru says it's a~bit old-fashioned nowadays but I~like the
+% idea and use it to document the file's growth. For this purpose
+% \gmdoc\ types out lines like
+% \begin{verbatim}
+%% \chschange{v0.98j}{2006/10/19}{4372}
+%% \chschange{v0.98j}{06/10/19}{4372}
+%\end{verbatim}
+% and you may place them at the beginning of the source file. Such
+% a~line results in setting the check sum to the number contained in
+% the last pair of braces and in making a~`general' changes entry that
+% states the check sum for version \<first brace> dated \<second ^^B
+% brace> was \<third brace>.
+%
+% \subdivision{The Parameters}
+%
+% The \pk{gmdoc} package provides some parameters specific to
+% typesetting the \TeX\ code:
+%
+% \stanza
+% \TextUsage\stanzaskip\ is a~vertical space inserted when a~blank
+% (code) line is met. It's equal |0.75\medskipamount| by default
+% (with the \emph{entire} |\medskipamount|'s stretch- and
+% shrinkability). Subsequent blank code lines do not increase this
+% space.
+%
+% At the points where narration begins a~new line after the code or an
+% inline comment and where a~new code line begins after the narration
+% (that is not an inline comment), a~\TextUsage\CodeTopsep\ glue is
+% added. At the beginning and the end of a~\env{macro} or
+% \env{environment} environment a~|\MacroTopsep| glue is added. By
+% default, these two skips are set equal |\stanzaskip|. 
+%
+% The |\stanzaskip|'s value is assigned also to the display skips and
+% to |\topsep|. This is done with the \TextUsage\UniformSkips\
+% declaration executed by default. If you want to change some of those
+% values, you should declare \TextUsage\NonUniformSkips\ in the
+% preamble to discard the default declaration. (To be more precise, by
+% default |\UniformSkips| is executed twice: when loading \gmdoc\ and
+% again |\AtBeginDocument| to allow you to change |\stanzaskip| and
+% have the other glues set due to it. |\NonUniformSkips| relaxes the
+% |\UniformSkips|'s occurrence at |\begin{document}|.)
+%
+% If you want to add a~vertical space of |\CodeTopsep| (equal by
+% default |\stanzaskip|), you are provided the \TextUsage\stanza\
+% command. Similarly, if you want to add a~vertical space of the
+% |\MacroTopsep| amount (by default also equal |\stanzaskip|), you are
+% given the \TextUsage\chunkskip\ command. They both act analogously
+% to |\addvspace| i.e., don't add two consecutive glues but put the
+% bigger of them.
+%
+% Since |\CodeTopsep| glue is inserted automatically at each
+% transition from the code (or code with an inline comment) to the
+% narration and reverse, it may happen that you want not to add such
+% a~glue exceptionally. Then there's the \TextUsage\nostanza\
+% command.
+%
+% \stanza
+% The \TeX\ code is indented with the \TextUsage\CodeIndent\ glue
+% and a~leading space increases indentation of the line by its
+% (space's) width. The default value of |\CodeIndent| is 1.5\,em.
+%
+% There's also a~parameter for the indent of the narration,
+% \TextUsage\TextIndent, but you should use it only in emergency
+% (otherwise what would be the margins for?). It's 0\,sp by default.
+%
+% \stanza
+% By default, typesetting a~|\DocInput/Include|d file is ended with
+% a~codeline containing the text `\EOFMark' given by the
+% \TextUsage\EOFMark\ macro. If you don't like such an ending, you
+% should end the source file with the \TextUsage\NoEOF\ macro in
+% a~comment, i.e.,
+% \[\hbox{|%|\<some text, why not>|\NoEOF|}\]
+%   This macro redefines |\EOFMark| and suppresses the End
+% Of File token to close the input properly. It also has the
+% |\endinput| effect so you may put some text you don't want to 
+% document after it.
+%
+% \stanza
+%
+% The crucial concept of \gmdoc\ is to use the line end character
+% as a~verbatim group opener and the comment char, usually the |%|, as
+% its delimiter. Therefore the `knowledge' what char starts
+% a~commentary is for this package crucial and utterly
+% important. The default assumption is that you use |%| as we
+% all do. So, if you use another character, then you should declare it
+% with \TextUsage\CodeDelim\ typing the desired char
+% preceded by a~backslash, e.g., |\CodeDelim\&|\,. (As just
+% mentioned implicitly, \possfil|\CodeDelim\%| is declared by
+% deafult.) 
+%
+% This declaration is always global so when- and wherever you change
+% your mind you should express it with a~new |\CodeDelim|
+% declaration.
+%
+% The plain (unstarred) version of |\CodeDelim| changes also the verb
+% `hyphen', the char appearing at the verbatim line breaks that is. If
+% you don't want to change the verb `hyphen', use |\CodeDelim*|.
+%
+% \stanza
+% Talking of special chars, the escape char, |\| by default, is also
+% very important for this package as it marks control sequences and
+% allows automatic indexing them for instance. Therefore, if you for
+% any reason choose another than |\| character to be the escape char,
+% you should tell \pk{gmdoc} about it with the
+% \TextUsage\CodeEscapeChar\ declaration. As the previous one, this too
+% takes its argument preceded by a~backslash,
+% e.g., |\CodeEscapeChar\!|. (As you may deduct from the above,
+% |\CodeEscapeChar\\| is declared by default.)
+%
+% The tradition is that in the packages |@| char is a~letter
+% i.e., of catcode \catletter. Frank Mittelbach in \docfm\ takes into account
+% a~possibility that a~user wishes some other chars to be letters,
+% too, and therefore he (F.M.) provides the
+% \TextUsage\MakePrivateLetters\ macro.  
+% So do I~and like in \docfm, this macro makes |@| sign a~letter. It
+%  also makes  |*| a~letter in order to cover the starred versions
+% of commands.
+%
+% Analogously but for a~slightly different purpose, the
+% \TextUsage\AddtoPrivateOthers\ macro is provided here. It adds its
+% argument, which is supposed to be a~one-char CS, to the
+% |\doprivateothers| list, whose r\^ole is to allow some special chars
+% to appear in the marking commands' arguments (the commands described
+% in section \gmiflink{Macros for Marking the Macros}). The default
+% contents of this list is | | (the space) and |^| so you may mark the
+% environments names and special sequences like |^^A| safely. This
+% list is also extended with every char that is |\MakeShortVerb|ed.
+% (I~don't see a~need of removing chars from this list, but if
+% you do, please let me know.)
+%
+% \stanza
+% The line numbers (if enabled) are typeset in the
+% \TextUsage\LineNumFont\ declaration's scope, which
+% is defined as |{\normalfont\tiny}| by default. Let us also remember, that for
+% each counter there is a~|\the|\<counter> macro available. The
+% counter for the line numbers is called \TextUsage*{codelinenum} so
+% the macro printing it is |\thecodelinenum|. By default we don't
+% change its \LaTeX's definition which is equivalent
+% |\arabic{codelinenum}|.
+%
+%
+% Three more parameter macros, are \TextUsage\IndexPrefix,
+% \TextUsage\EntryPrefix\ and \TextUsage\HLPrefix. All three are
+% provided with the account of  including multiple files in one
+% document. 
+% They are equal (almost) |\@empty| by default. The first may store
+% main level index entry of which all indexed macros and environments
+% would be subentries, e.g., the name of the package. The third may or
+% even should store a~text to distinguish equal codeline numbers of
+% distinct source files. It may be the file name too, of course. The
+% second macro is intended for another concept, namely the one from
+% \pk{ltxdoc} class, to distinguish the codeline numbers from
+% different files \emph{in the index} by the file marker. Anyway, if
+% you document just one file per document, there's no need of
+% redefining those macros, nor when you input multiple files with
+% |\DocInclude|.
+%
+% \pk{gmdoc} automatically indexes the control sequences
+% occurring in the code. Their index entries may be `common' or
+% distinguished in two (more) ways. The concept is to distinguish the
+% entries indicating the \emph{usage} of the CS and the entries
+% indicating the \emph{definition} of the CS.
+%
+% The special formattings of `usage' and `def' index entries are
+% determined by \TextUsage\UsgEntry\ and \TextUsage\DefEntry\
+% one-parameter macros (the parameter shall be substituted with the
+% reference number) and by default are defined as |\textit| and
+% |\underline| respectively (as in \docfm).
+%
+% There's one more parameter macro, \TextUsage\CommonEntryCmd\ that
+% stores the name of the encapsulation for the `common' index
+% entries (not special) i.e., a~word that'll become a~CS that will 
+% be put before an entry in the \file{.ind} file. By default it's defined
+% as |{relax}| and a~nontrivial use of it you may see in line
+% \ref{UsgHack} of the driver file of this documentation, where it
+% makes all the index entries of the driver's code are formatted as `usage'.
+%
+% \stanza
+% ^^A~Index Parameters
+% \gmhypertarget[IndexParams]{The index} comes in a~\env{multicols}
+% environment whose columns number is determined by the
+% \TextUsage*{IndexColumns} counter set by default to~3. To save space,
+% the index begins at the same page as the previous text provided
+% there is at least \TextUsage\IndexMin\ of the page height free. By
+% default, |\IndexMin|\equals\the\IndexMin.
+%
+% The text put at the beginning of the index is declared with
+% a~one-argument \TextUsage\IndexPrologue. Its default text at current
+% index option you may \gmiflink[DIPrologue]{admire} on page
+% \pageref*{DIPrologue}.^^A
+% \AtDIPrologue{\raisebox{4cm}[0sp][0sp]{\gmhypertarget[DIPrologue]}}
+% Of course, 
+% you may write your own |\IndexPrologue{|\<brand new index ^^B
+% prologue>|}|, but if you like the default and want only to add
+% something to it, you are provided \TextUsage\AtDIPrologue\
+% one-argument declaration that adds the stuff after the default
+% text. For instance, I~used it to add a~label and hypertarget that is
+% referred to two sentences earlier.
+%
+% By default the colour of the index entry hyperlinks is set black to
+% let Adobe Reader work faster. If you don't want this,
+% \Describe\IndexLinksBlack|\let\IndexLinksBlack\relax|. That leaves
+% the index links colour alone and hides
+% the text about black links from the default index prologue.
+%
+% Other index parameters are set with the \TextUsage\IndexParms\ macro
+% defined in line \gmifref[IndexParms]{sixty-odd} of the code. If you
+% want to change some of them, you don't have to use
+% |\renewcommand*\IndexParms| and set all of the parameters: you may
+% \Describe\gaddtomacro^^A
+% |\gaddtomacro\IndexParms{|\<only the desired changes>|}|.
+% (|\gaddtomacro| is an alias for \LaTeX's |\g@addto@macro| provided
+% by \pk{gmutils}.)
+%
+% At the default \gmdoc\ settings the \file{.idx} file is prepared for
+% the default settings of MakeIndex (no special style). Therefore the
+% index control chars are as usual. But if you need to use other
+% chars as MakeIndex controls, know that they are stored in the four
+% macros: \TextUsage\actualchar, \TextUsage\quotechar,
+% \TextUsage\levelchar\ and \TextUsage\encapchar\ whose meaning you
+% infer from their names.  Any redefinition of them \emph{should be ^^B 
+% done in the preamble} because the first usage of them takes place at
+% |\begin{document}| and on it depends further tests telling \TeX\
+%   what characters of a~scanned CS name it should quote before
+%   writing it to the \pk{.idx} file.
+% 
+% \stanza
+% Frank Mittelbach in \docfm\ provides the |\verbatimchar| macro to
+% (re)define the |\verb|'s delimiter for the index entries of the
+% scanned CS names etc. \gmdoc\ also uses \TextUsage\verbatimchar\ but
+% defines it as |{&}|. Moreover, a~macro that wraps a~CS name in |\verb|
+% checks whether the wrapped CS isn't |\&| and if it is, |$| is taken
+% as the delimiter. So there's hardly chance that you'll need to
+% redefine |\verbatimchar|.
+%
+% So strange delimiters are chosen deliberately to allow any `other'
+% chars in the environments names.
+%
+% \stanza
+% There's a~quadratus of commands taken from \docfm:
+% \TextUsage\StopEventually, \TextUsage\Finale,
+% \TextUsage\AlsoImplementation\ and \TextUsage\OnlyDescription\ that
+% should be 
+% explained simultaneously (in a~polyphonic song
+% e.g.).
+%
+% The |\OnlyDescription| and |\AlsoImplementation|
+% declarations are intended to  exclude or include the code part from
+% the documentation. The point between the description and the
+% implementation part should be marked with |\StopEventually{|\<the ^^B
+% stuff to be executed anyway>|}| and |\Finale| should be typed
+% at the end of file. Then |\OnlyDescription| defines
+% |\StopEventually| to expand to its argument followed by |\endinput|
+% and |\AlsoImplementation| defines |\StopEventually| to do nothing
+% but pass its argument to |\Finale|.
+%
+%
+% \subdivision{The Narration Macros}
+%
+% To print the control sequences' names you have the \TextUsage\verb\ macro and
+% its `shortverb' version whatever you define. But they won't work if
+% you put them in an argument of another macro. For such a~situation,
+% or if you just prefer, \gmdoc\ (\pk{gmutils}) provides a~robust command
+% \TextUsage\cs, which takes one obligatory argument, the macro's name
+% without the backslash, e.g., |\cs{mymacro}| produces
+% \cs{mymacro}. I~take account of a~need of printing some other text
+% verbatim, too, and therefore |\cs| has the first argument
+% optional, which is the text to be typeset before the mandatory
+% argument. It's the backslash by default, but if you wish to typeset
+% something without the |\|, you may write {\verbeolOK |\cs[]{not
+% a~macro}|}. Moreover, for typesetting the environments' names,
+% \gmdoc\ (\pk{gmutils})
+% provides the \TextUsage\env\ macro, that prints its argument
+% verbatim and without a~backslash, e.g., |\env{an environment}|
+% produces \env{an environment}.
+%
+% To print packages' names sans serif there is a~\TextUsage\pk\
+% one-argument command, and the \TextUsage\file\ command intended for
+% the filenames. 
+%
+% Because we play a~lot with the |\catcode|s here and want to talk
+% about it, there are \TextUsage\catletter, \TextUsage\catother\ and
+% \TextUsage\catactive\ macros that print \catletter, \catother\ and
+% \catactive\ respectively to concisely mark the most used char
+% categories.
+%
+% I~wish my self-documenting code to be able to be typeset each
+% package separately or several in one document. Therefore I~need some
+% `flexible' sectioning commands and here they are:
+% \TextUsage\division\ and \TextUsage\subdivision\ so far, that by
+% default are |\let| to be |\section| and |\subsection| if such
+% commands are defined in the documentclass. (If not, |\division| and
+% |\subdivision| are |\let| to be |\relax|.)
+%
+% \stanza
+% One more kind of flexibility is to allow using \pk{mwcls} or the
+% standard classes for the same file. There was a~trouble with the
+% number and order of the optional arguments of the original
+% \pk{mwcls}'s sectioning commands.
+%
+% It's resolved in \pk{gmutils} so you are free at this point, and
+% even more free than in the standard classes: if you give
+% a~sectioning command just one optional argument, it will be the
+% title to toc and to the running head (that's standard in
+% scls\footnote{See \pk{gmutils} for some subtle details.}). If you give
+% \emph{two} optionals, the first will go to the running head and the
+% other to toc. (In both cases the mandatory argument goes only to the
+% page).
+%
+% \stanza
+% If you wish the |\DocInclude|d files make other sectionings than the
+% default, you may declare \TextUsage\SetFileDiv|{|\<sec name without ^^B
+% backslash>|}|.
+%
+% \stanza
+%
+% \pk{gmdoc.sty} provides also an environment \TextUsage*{gmlonely}
+% to wrap some text you think you may want to skip some day. When
+% that day comes, you write \TextUsage\skipgmlonely\ before the
+% instances of \env{gmlonely} you want to skip. This declaration has
+% an optional argument which is for a~text that'll appear in(stead of) the
+% first \env{gmlonely}'s instance in every |\DocInput| or
+% |\DocInclude|d file within |\skipgmlonely|'s scope.
+%
+% An example of use you may see in this documentation: the repeated
+% passages about the installation and compiling the documentation are
+% skipped in further chapters thanks to it.
+%
+% \stanza
+% \gmdoc\ provides some \TeX-related logos:\\
+% \Describe\AmSTeX typesets \AmSTeX,\\
+% \Describe\BibTeX\BibTeX,\\
+% \Describe\SliTeX\SliTeX,\\
+% \Describe\PlainTeX\PlainTeX,\\
+% \Describe\Web\Web,\\
+% \Describe\TeXbook\TeXbook,\par^^A~without par there raised error
+% ^^A `too many unprocessed floats'.
+% \noindent\Describe\eTeX\eTeX,\\
+% \Describe\pdfeTeX\pdfeTeX\\
+% \Describe\pdfTeX\pdfTeX\\
+% \Describe\XeTeX\XeTeX\ (the first |E| will be reversed if the
+% \pk{graphics} package is loaded) and\\
+% \Describe\LaTeXpar\LaTeXpar.\par
+% (The last logo is defined in \pk{gmutils}.)
+%
+% \noindent\Describe\ds\ds\ not quite a~logo, but still convenient.
+%
+% \stanza
+% The \TextUsage*{copyrnote} environment is provided to format the
+% copyright note flush left in |\obeylines|' scope.
+%
+% \stanza
+%
+% To put an arbitrary text into a~marginpar and have it flushed right
+% just like the macros' names, you are provided the
+% \TextUsage\gmdmarginpar\ macro that takes one and mandatory argument
+% which is the contents of the marginpar.
+%
+% \stanza
+% To make a~vertical space to separate some piece of text you are
+% given two macros: \TextUsage\stanza\ and \TextUsage\chunkskip. The
+% first adds |\stanzaskip| while the latter |\MacroTopsep|. Both of
+% them take care of not cumulating the vspaces.
+%
+%
+%\stanza
+% The \TextUsage*{quotation} environment is redefined just to enclose
+% its contents in double quotes.
+% \stanza
+%
+% The \TextUsage\GetFileInfo|{|\<file name with extension>|}| command
+% defines \TextUsage\filedate, \Describe\fileversion\cs{fil\-e\-vers\-ion} and
+% \TextUsage\fileinfo\ as the respective pieces of the info (the
+% optional argument) provided
+% by \cs{Pro\-vid\-es\-Class/Pack\-age/Fi\-le} declarations. The
+% information of the file you process with \gmdoc\
+% is provided (and therefore getable) if the file is also loaded (or
+% the |\Provide...| line occures in a~|\StraightEOL| scope).
+%
+% Since you may wish to process also files that you don't load,
+% there are commands \TextUsage\ProvideFileInfo|{|\<file name with ^^B
+% extension>|}[|\<info>|]| and
+% \TextUsage\ProvideSelfInfo|[|\<info>|]|. (\<info> should consist of:
+% \<year>|/|\<month>|/|\<day>| |\<version number>| |\<a~short note>.)
+%
+% A~macro for the standard note is provided, \TextUsage\filenote, that
+% expands to ``This file has version number \<version number> dated
+% \<date>.'' To place such a~note in the document's title (or heading,
+% with |\DocInclude| at the default settings), there's
+% \TextUsage\thfileinfo\ macro that puts |\fileinfo| in |\thanks|.
+%
+% \stanza
+% Since |\noindent| didn't want to cooperate with my code and
+% narration layers sometimes, I~provide \TextUsage\gmdnoindent\ that
+% forces a~not indented paragraph if |\noindent| could not.
+%
+% \stanza
+% If you declare the code delimiter other than |%| and then want |%|
+% back, you may write \TextUsage\CDPerc\ instead of |\CodeDelim\%|.
+%
+% If you like |&| as the code delimiter (as I~did twice), you may write
+% \TextUsage\CDAnd\ instead of |\CodeDelim\&|.
+%
+% 
+%^^A \stanza
+%^^A For an example driver file see chapter \gmiflink{The Driver}.
+%
+%
+% \subdivision{A~Queerness of \cs{label}}
+%
+% You should be loyally informed that |\label| in \gmdoc\
+% behaves slightly
+% non-standard in the \cs{Doc\-In\-put/\:Inc\-lud\-e}d files:
+% the automatic redefinitions of |\ref| at each code line
+% are \emph{global} (since the code is typeset in groups and the
+% |\ref|s will be out of those groups), so
+% a~|\ref|erence in the narrative will point at the last code line not
+% the last section, \emph{unlike} in the standard \LaTeX.
+%
+% 
+% \subdivision{\docfm-Compatibility}
+% 
+% One of my goals while writing \gmdoc\ was to make compilation of
+% \docfm-like files with \gmdoc\ possible. I~cannot guarantee the goal
+% has been reached but I~\emph{did} compile \pk{doc.dtx} with not
+% a~smallest change of that file (actually, there was a~tiny little buggie in
+% line 3299 which I~fixed remotely with |\AfterMacrocode| tool written
+% specially for that). So,
+% if you wish to compile a~\docfm-like file with my humble package,
+% just try.
+%
+% The \docfm\ commands most important in my opinion are supported by
+% \gmdoc. Some commands, mostly the obsolete in my opinion, are not
+% supported but give an info on the terminal and in \pk{.log}.
+%
+% I~assume that if one wishes to use \docfm's interface then \heshe\
+% won't use \gmdoc's options but just the default. (Some \gmdoc\
+% options may interfere with some \docfm\ commands, they may cancel
+% them e.g.)
+%
+% \stanza
+% The main input commands compatible with \docfm\ are
+% \TextUsage\OldDocInput\ and \TextUsage\DocInclude, the latter
+% however only in the \TextUsage\olddocIncludes\ declaration's scope.
+%
+% Within their scope/argument the \TextUsage*{macrocode} environments
+% behave as in \docfm, i.e.\ they are a~kind of verbatim and require to be
+% ended with |%    \end{macrocode(*)}|.
+%
+% The default behaviour of \env{macrocode(*)} with the `new' input
+% commands is different however. Remember at te `new' fashion the code
+% and narration layers philosophy is in force and that is sustained
+% within \env{macrocode(*)}. Which means basically that with `new'
+% settings when you write
+%\begin{verbatim}
+%% \begin{macrocode}
+%  \alittlemacro % change it to \blaargh
+%%\end{macrocode}
+%\end{verbatim}
+% and |\blaargh|'s definition is |{foo}|, you'll get
+%\[\hbox{|\alittlemacro %| change it to foo}\]
+% (Note that `my' \env{macrocode} doesn't require 
+% the magical |%    \end|.)
+%
+% If you are used to the traditional (\docfm's) \env{macrocode} and still wish to
+% use \gmdoc\ new way, you have at least two options: there is the
+% \TextUsage*{oldmc} environment analogous to the traditional (\docfm's)
+% \env{macrocode} (it also has the starred version), that's the first
+% option (I~needed the traditional behaviour once in this
+% documentation, find out where \& why). The other is to write
+% \TextUsage\VerbMacrocodes. That declaration (OCSR) redefines 
+% \env{macrocode} and \env{macrocode*} to behave the traditional way.
+% (It's always executed by |\OldDocInput| and |\olddocIncludes|.)
+%
+% \stanza
+% For a~more detailed discussion of what is \docfm-compatible and how,
+% see the code section \gmiflink[doccompa]{\docfm-Compatibiliy}.
+%
+% \StopEventually{\NoEOF}
+%
+%
+% \division{The Code}
+%
+%\iffalse For debug
+\catcode`\^^C9\relax
+% \fi
+% \catcode`\^^C\active
+%
+% \changes[^^C]{v0.98g}{06/10/10}{recatcoded for debug: in the working
+% pass it's made ignored and in the documenting pass it's made active}
+%
+%
+% \DoIndex{\par \@@par}
+% 
+% The basic idea of this package is to re|\catcode| |^^M| (the line
+% end char) and |%| (or any other comment char) so that they start and
+% finish typesetting of what's between them as the \TeX\ code i.e.,
+% verbatim and with the bells and whistles.
+%
+% The bells and whistles are (optional) numbering of the codelines,
+% and automatic indexing the CSs, possibly with
+% special format for the `def' and `usage' entries.
+%
+% As mentioned in the preface, this package aims at a~minimal markup
+% of the working code. A~package author writes \hisher\ splendid code
+% and adds a~brilliant comment in |%|ed lines and that's all. Of
+% course, if \heshe\ wants tomake a~|\section| or |\emph|asise,
+% \heshe\ has to type respective CSs.
+%
+% I~see the feature described above to be quite a~convenience,
+% however it has some price. See section 
+% \gmiflink[afc]{Life Among Queer EOLs}  for details,
+% here I~state only that in my opinion the price is not very high.
+% \stanza
+%
+% More detailedly, the idea is to make |^^M| (end of line char) active
+% and to define it to check if the next char i.e., the beginnig of the
+% next line is a~|%| and if so
+% to gobble it and just continue usual typesetting or else to start
+% a~verbatim scope. In fact, every such a~line end starts a~verbatim
+% scope which is immediately closed, if the next line begins with
+% (leading spaces and) the code delimiter.
+%
+% Further details are typographical parameters of verbatim scope and
+% how to restore normal settings after such a~scope so that a~code
+% line could be commented and still displayed, how to deal with
+% leading spaces, how to allow breaking a~moving argument in two lines
+% in the comment layer, how to index and marginpar macros etc.
+%
+%
+%\subdivision{The Package Options}
+%
+% Maybe someone wants the code lines not to be numbered.
+% \CodeDefIndex\if@linesnotnum
+% \CodeDefIndex*\@linesnotnumtrue
+% \CodeDefIndex*\@linesnotnumalse
+\newif\if@linesnotnum
+% \CodeDefine*{linesnotnum}
+\DeclareOption{linesnotnum}{\@linesnotnumtrue}
+
+% And maybe he or she wishes to declare resetting the line counter
+% along with some sectioning counter  him/herself.
+% \CodeDefIndex\if@uresetlinecount
+% \CodeDefIndex*\@uresetlinecounttrue
+% \CodeDefIndex*\@uresetlinecountfalse
+\newif\if@uresetlinecount
+% \CodeDefine*{uresetlinecount}
+\DeclareOption{uresetlinecount}{\@uresetlinecounttrue}
+
+% And let the user be given a~possibility to count the comment
+% lines.
+% \CodeDefIndex\if@countalllines
+\newif\if@countalllines
+% \CodeDefine*{countalllines}
+\DeclareOption{countalllines}{\@countalllinestrue}
+
+% Unlike in \docfm , indexing the macros is the default and the
+% default reference is the code line number.
+% \CodeDefIndex\if@noindex
+% \CodeDefIndex*\@noindextrue
+% \CodeDefIndex*\@noindexfalse
+\newif\if@noindex
+% \CodeDefine*{noindex}
+\DeclareOption{noindex}{\@noindextrue}
+
+% \CodeDefIndex\if@pageindex
+% \CodeDefIndex*\@pageindextrue
+% \CodeDefIndex*\@pageindexfalse
+\newif\if@pageindex
+% \CodeDefine*{pageindex}
+\DeclareOption{pageindex}{\@pageindextrue}
+
+% It would be a~great honour to me if someone would like to document
+% \LaTeX\ source with this humble package but I~don't think it's
+% really probable so let's make an option that'll switch index exclude
+% list properly (see sec.\ \gmiflink{Index Exclude List}).
+% \CodeDefIndex\if@indexallmacros
+% \CodeDefIndex*\@indexallmacrostrue
+% \CodeDefIndex*\@indexallmacrosfalse
+\newif\if@indexallmacros
+% \CodeDefine*{indexallmacros}
+\DeclareOption{indexallmacros}{\@indexallmacrostrue}
+
+% Some document classes don't support marginpars or disable them by
+% default (as my favourite Marcin Woli\'nski's classes).
+%
+% \CodeDefIndex\if@marginparsused
+% \changes{v0.98e}{06/09/23}{wrapped in \cs{@ifundefined} (a~bug fix:
+%   before the bare \cs{newif} falsified the \cs{if@marginparsused}
+%   switch when it had been defined and set True by the \pk{mwart}
+%   class.)}
+% \CodeDefIndex*\@marginparsusedtrue
+% \CodeDefIndex*\@marginparsusedfalse
+\@ifundefined{if@marginparsused}{\newif\if@marginparsused}{}
+% This switch is copied from \pk{mwbk.cls} for compatibility with
+% it. Thanks to it loading an \pk{mwcls} with |[withmarginpar]| option
+% shall switch marginpars on in this package, too. 
+%
+% To be compatible with the standard classes, let's |\let|:
+\@ifclassloaded{article}{\@marginparsusedtrue}{}
+%\label{mparclause1}
+\@ifclassloaded{report}{\@marginparsusedtrue}{}
+\@ifclassloaded{book}{\@marginparsusedtrue}{}
+% \label{mparclause2} And if you don't use \pk{mwcls} nor standard
+% classes, then you have the options:\CodeDefine*{withmarginpar}
+\DeclareOption{withmarginpar}{\@marginparsusedtrue}
+%\CodeDefine*{nomarginpar}
+\DeclareOption{nomarginpar}{\@marginparsusedfalse}
+% The order of the above conditional switches and options is significant.
+% Thanks to it the options are available also in the
+% standard classes and in \pk{mwcls}.
+%
+% \stanza
+% To make the code spaces blank (they are visible by default except
+% the leading ones).
+% \CodeDefIndex\if@codespacesblank
+\newif\if@codespacesblank
+% \CodeDefine*{codespacesblank}
+\DeclareOption{codespacesblank}{\@codespacesblanktrue}
+
+
+\ProcessOptions
+
+
+% \subdivision{The Dependencies and Preliminaries}
+%
+% We require another package of mine that provides some tricky macros
+% analogous to the \LaTeX\ standard ones, such as |\newgif| and
+% |\@ifnextcat|.
+\RequirePackage{gmutils}
+
+% A~standard package for defining colours,
+\RequirePackage{color}
+% and a~colour definition for the hyperlinks not to be too bright
+\definecolor{deepblue}{rgb}{0,0,.85}
+
+% And the standard package probably most important for \gmdoc:
+% If the user doesn't load \pk{hyperref} with \hisher\ favourite
+% options, we do, with \emph{ours}. If \heshe\ has
+% done it, we change only the links' colour.
+% 
+% \changes{v0.97}{06/09/04}{linkcolor made deep blue}
+\@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=None]{hyperref}}
+
+% Now a~little addition to \pk{hyperref}, a~conditional hyperlinking
+% possibility with the |\gmhypertarget| and |\gmiflink| macros. It
+% \emph{has} to be loaded \emph{after} \pk{hyperref}.
+\RequirePackage{gmiflink}
+% And a~slight redefinition of \env{verbatim}, |\verb(*)| and
+% providing of |\MakeShortVerb(*)|.
+\RequirePackage{gmverb}
+
+\if@noindex
+  \AtBeginDocument{\gag@index}% for the latter macro see line
+  % \ref{gag@index}.
+\else
+   \RequirePackage{makeidx}\makeindex
+\fi
+
+% Now, a~crucial statement about the code delimiter in the input file.
+% Providing a~special declaration for the assignment is intended for
+% documenting the packages that play with |%|'s
+% |\catcode|. Some macros for such plays are defined
+% \gmiflink[CDAnd]{further}.
+%
+% The declaration comes in the starred and unstarred version. The
+% unstarred version besides declaring the code delimiter declares
+% the same char as the verb(atim) `hyphen'. The starred version
+% doesn't change the verb 'hyphen'. That is intended for the special
+% tricks e.g.\ for the \env{oldmc} environment.
+%
+% If you want to change the verb `hyphen', there is the
+% |\VerbHyphen\|\<char> 
+% declaration provided by \pk{gmverb}.
+% 
+% \changes{v0.98c}{06/9/8}{\cs{CodeDelim} renamed from a~rather confusing
+% \cs{NewCommentChar}; also the internal \cs{code@delim} is renamed
+% from as much confusing \cs{commentchar} and a~\cs{glet} for
+% \cs{verb}s \cs{verbhyphen} is added; similarly
+% \cs{[Un]CodeDelimAnd} renamed from \cs{[un]commentand}}
+%
+% \changes{v0.98m}{06/11/10}{\cs{CDAnd} and \cs{CDPerc} renamed from
+% \cs{CodeDelimAnd} and \cs{UnCodeDelimAnd} and both those commands
+% simplified to just declaring the code delimiter}
+%
+% \changes{v0.99a}{06/12/1}{split into the starred and unstarred
+% versions}
+%
+% \CodeDefine\CodeDelim \CodeDefine\code@delim
+\def\CodeDelim{\@ifstar{\Code@Delim}{\Code@Delim@St}}
+
+\def\Code@Delim#1{%
+  {\escapechar\m@ne
+    \expandafter\gdef\expandafter\code@delim\expandafter{\string#1}}}
+
+\def\Code@Delim@St#1{\Code@Delim{#1}\VerbHyphen{#1}}
+
+% It is an invariant of \gmdoc ing that |\code@delim| stores the
+% current code delimiter (of catcode 12).
+%
+% The |\code@delim| should be \catother\ so a~space is not allowed as
+% a~code delimiter. I~don't think it \emph{really} to be a~limitation.
+%
+% And let's assume you do as we all do:
+\CodeDelim\%
+
+% We'll play with |\everypar|, a~bit, and if you use such things as
+% the |{itemize}| environment, an error would occure if we didn't
+% store the previous value of |\everypar| and didn't restore it at
+% return to the narration. So let's assign a~|\toks| list to store the
+% original |\everypar|
+% \CodeDefIndex\gmd@preverypar
+\newtoks\gmd@preverypar 
+
+% \CodeDefIndex\settexcodehangi
+\newcommand*\settexcodehangi{%
+  \hangindent=\verbatimhangindent \hangafter=\@ne}% we'll use
+% it in the inline comment case. |\verbatimhangindent| is provided by the
+% \pk{gmverb} package and\equals3\,em by default.
+% \CodeDefIndex\@@settexcodehangi
+\@ifdefinable\@@settexcodehangi{\let\@@settexcodehangi=\settexcodehangi}
+%^^A~\gmdeverycodeline{\@@settexcodehangi}
+
+% We'll play a~bit with |\leftskip|, so let the user have a~parameter
+% instead. For normal text (i.e.\ the comment):
+% \CodeDefine\TextIndent
+\newlength\TextIndent
+% I~assume it's originally equal |\leftskip|, i.e.\ |\z@|. And for
+% the \TeX\ code:
+\newlength\CodeIndent
+%\label{CodeIndent}
+% \CodeDefine\CodeIndent
+\CodeIndent=1,5em\relax
+% And the vertical space to be inserted where there are blank lines in the
+% source code:
+\@ifundefined{stanzaskip}{\newlength\stanzaskip}{}
+% I~use |\stanzaskip| in \pk{gmverse} package and
+% derivates for typesetting poetry. A~computer program code \emph{is}
+% poetry.
+% \CodeDefine\stanzaskip
+\stanzaskip=\medskipamount
+\advance\stanzaskip by-.25\medskipamount% to preserve the stretch- and
+% shrinkability.\par
+%
+%\stanza
+% A~vertical space between the commentary and the code seems
+% to enhance readability so declare
+\newskip\CodeTopsep
+\newskip\MacroTopsep
+
+% And let's set them.  For the \ae sthetical
+% minimality\StraightEOL\footnote{The terms `minimal' and `minimalist'
+%   used in \gmdoc\ are among others\ inspired by the \emph{South Park}
+%   cartoon's episode \emph{Mr.\ Hankey The~Christmas (\dots)}\/ in
+%   which `Philip Glass, a~Minimalist New York composer' appears in
+%   a~`non-denominational non-offensive Christmas play' ^^A(
+%   ;-)\,. (Philip Glass composed the music to the \emph{Qatsi}
+%   trilogy among others)}\QueerEOL\ let's unify them and other most important
+% vertical spaces used in \gmdoc. I~think a~macro that gathers all
+% these assignments may be handy.  
+% \CodeDefine\MacroTopsep 
+% \CodeDefine\CodeTopsep
+% \CodeDefine\UniformSkips
+\def\UniformSkips{%\label{UniformSkips}
+  % \CodeDefine\CodeTopsep \CodeDefine\MacroTopsep
+  \CodeTopsep=\stanzaskip
+  \MacroTopsep=\stanzaskip
+  \abovedisplayskip=\stanzaskip
+  %\nostanza
+  % \leftskip0sp \noindent |%\abovedisplayshortskip|
+  % remains untouched as it is 0.0\,pt plus 3.0\,pt by default.\par
+  % 
+  % \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
+  % \nlpercent\cs{be\+low\+dis\+play\+short\+skip} forth and back to
+  % give it the \nlpercent\cs{small\+skip\+am\+ount}'s shrink- and
+  % stretchability components.
+  \topsep=\stanzaskip
+  \partopsep=\z@
+}
+% We make it the default,
+\UniformSkips
+% \gmdnoindent but we allow you to change the benchmark glue
+% i.e., |\stanzaskip| in the preamble and still have the other glues
+% set due to it: we launch |\UniformSkips| again after the
+% preamble. 
+\AtBeginDocument{\UniformSkips}
+
+% So, if you don't want them at all i.e., you don't want to set other
+% glues due to |\stanzaskip|, you should use the following
+% declaration. That shall discard the unwanted setting already placed
+% in the |\begin{document}| hook.
+%
+% \CodeDefine\NonUniformSkips
+\newcommand*\NonUniformSkips{\@relaxen\UniformSkips}
+% Why do we launch |\UniformSkips| twice then? The first time is
+% to set all the \gmdoc-specific glues \emph{somehow}, which allows
+% you to set not all of them, and the second
+% time to set them due to a~possible change of |\stanzaskip|.
+%
+% \stanza
+% And let's define a~macro to insert a~space
+% for a~chunk of documentation, e.g., to mark the beginning of new
+% macro's explanation and code.
+% \CodeDefine\chunkskip
+\newcommand*\chunkskip{%
+  \skip0=\MacroTopsep
+  \if@codeskipput\advance\skip0 by-\CodeTopsep\fi
+  \par\addvspace{\skip0}\@codeskipputgtrue}
+
+% And, for a~smaller part of text,
+% \CodeDefine\stanza
+\newcommand*\stanza{%
+  \skip0=\stanzaskip
+  \if@codeskipput\advance\skip0 by-\CodeTopsep\fi
+  \par\addvspace{\skip0}\@codeskipputgtrue}
+
+% Since the stanza skips are inserted automatically most often (cf.\ lines
+% \ref{codeskip}, \ref{codeskip2}, \ref{codeskip3}, \ref{codeskip4},
+% \ref{codeskip5}), sometimes you may need to forbid them.
+% \CodeDefine\nostanza
+\newcommand*\nostanza{%
+  \@codeskipputgtrue\@afternarrgfalse\@aftercodegtrue}% In the `code
+% to narration' case the first switch is enough but in the countercase
+% `narration to code' both the second and third are necessary while
+% the first is not.
+%
+%
+% To count the lines where they have begun not before them
+% \CodeDefine\if@newline \CodeDefine*\@newlinegtrue
+% \CodeDefine*\@newlinegfalse
+\newgif\if@newline
+% |\newgif| is |\newif| with global effect i.e., it defines |\...gtrue|
+% and |\...gfalse| switchers that switch respective Boolean switch
+% \emph{globally}. See \pk{gmutils} package for details.
+% \stanza
+%
+% To handle the \ds\ directives not \emph{any} |%<...|.
+% \CodeDefine*\if@dsdir
+\newgif\if@dsdir
+% This switch will be falsified at the first char of a~code line. (We
+% need a~switch independent of the one indicationg whether the line
+% has or has not been counted because of two reasons: 1.\ line
+% numbering is optional, 2.\ counting the line falsifies that switch
+% \emph{before} the first char.)
+%
+% \subdivision{The Core}
+%
+% Now we define main |\input|ing command that'll change catcodes.
+% The macros used by it are defined later.
+% 
+% \CodeDefine\DocInput \CodeDefine\Doc@Input
+% \changes{v0.98}{06/09/05}{\cs{@makeother\protect\bslash_} added}
+
+\newcommand*\DocInput{\bgroup\@makeother\_\Doc@Input}
+
+\begingroup\catcode`\^^M=\active%
+\firstofone{\endgroup%
+  \newcommand*{\Doc@Input}[1]{\egroup\begingroup%
+  % \CodeDefIndex\gmd@inputname
+    \edef\gmd@inputname{#1}% we'll use it in some notifications.
+    % \CodeDefIndex\gmd@currentlabel@before
+    \let\gmd@currentlabel@before=\@currentlabel% we store it 'cause
+    % 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
+    %\vskip-\stanzaskip
+    % \changes[\DocInput]{v0.95}{06/08/15}{\cs{club-}
+    % and \cs{widowpenalty} set both to 3333}
+    %\CodeUsgIndex\tolerance
+    \tolerance=1000 % as in \docfm.
+    % \CodeUsgIndex\CodeSpacesBlank
+    \if@codespacesblank\CodeSpacesBlank\fi% \label{cspb}The default is
+    % that the code spaces are visible but here this may be cancelled
+    % due to the \cs{co\+de\+spa\+ces\+blank} option.
+    \catcode`\^^M=\active%
+    % \CodeUsgIndex\code@delim
+    \expandafter\@makeother\code@delim%
+    % \CodeUsgIndex\gmd@resetlinecount
+    \gmd@resetlinecount% due to the option |uresetlinecount|
+    % we reset the linenumber counter or do nothing.
+    % \CodeUsgIndex\@beginputhook
+    \@beginputhook% my first use of it is to redefine |\maketitle|
+    % just at this point not globally. \CodeUsgIndex\everypar
+    \everypar=\expandafter{\expandafter\@codetonarrskip\the\everypar}%
+    % \CodeDefine*{^^M}
+    \let^^M=\gmd@textEOL%
+    % \CodeDefIndex\gmd@guardedinput
+    \edef\gmd@guardedinput{%
+      \noexpand\@@input #1\relax% it's true \TeX's
+      % |\input|. \CodeUsgIndex\EOFMark
+      \noexpand\EOFMark% to pretty finish the input, see
+      % line~\ref{eofMark}.\CodeUsgIndex\code@delim
+      \noexpand^^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 initialize the counter for the number
+    % of the escape chars (the assignment is |\global|).
+    \expandafter\expandafter\expandafter^^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 crossrefs, but anyway we acted
+    % globally and so acts \pk{hyperref}.
+    \endgroup%
+  }% end of |\Doc@Input|'s definition.
+}% end of |\firstofone|'s argument.
+
+% So, having the main macro outlined, let's fill in the details.
+%
+% First, define the queer EOL. We define a~macro that |^^M| will be
+% let to. |\gmd@textEOL|
+% will be used also for checking the |%^^M| case (|\@ifnextchar|
+% does |\ifx|).
+%
+% \CodeDefine\gmd@textEOL
+
+\def\gmd@textEOL{ % a~space just like in normal \TeX. We put it first to
+  % cooperate with \nlpercent|\^^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.
+  \expandafter\@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?
+% ^^A      \let\gmd@narrpar\par% we store the possibly redefined |\par| to
+% ^^A~typeset the previous narration appropriately.
+% ^^A      \def\par{\gmd@narrpar\let\par\@@par}% the first |\par| will be
+% ^^A~the one brought from the narration but then, 
+  \gmd@setclubpenalty% Most paragraphs will be
+  % one-line most probably. Since some sectioning commands may change
+  % \nlpercent|\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
+  \ttverbatim% it does |\tt| and makes specials
+  % other or |\active|-and-breakable.
+  \gmd@DoTeXCodeSpace%
+  \@makeother\|% 'cause |\ttverbatim| doesn't do that.
+  \MakePrivateLetters% see line \ref{MPL}.\par
+  \expandafter\@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.
+  \expandafter\@ifnextchar\expandafter{\code@delim}{%\label{ifContNarr}
+    % \CodeUsgIndex\gmd@continuenarration
+    \gmd@continuenarration}{% \CodeUsgIndex\gmd@dolspaces
+    \gmd@dolspaces% it will launch |\gmd@typesettexcode|.
+  }% end of |\@ifnextchar|'s else.
+}% end of |\gmd@textEOL|'s definition.
+
+\def\gmd@setclubpenalty{\clubpenalty=3333 }
+
+% \stanza
+% For convenient adding things to the begin- and endinput hooks:
+% \CodeDefine\AtEndInput \CodeDefine\@endinputhook
+\def\AtEndInput{\g@addto@macro\@endinputhook}
+\def\@endinputhook{}
+%\label{@endinputhook}
+% Simili modo
+% \CodeDefine\AtBegInput \CodeDefine\@beginputhook
+\def\AtBegInput{\g@addto@macro\@beginputhook}
+\def\@beginputhook{}
+
+% And let's use it instantly to avoid a~disaster while reading in the
+% table of contents.
+% \CodeDefine\tableofcontents
+% \CodeDefIndex\gmd@@toc
+\AtBegInput{\let\gmd@@toc\tableofcontents
+  \def\tableofcontents{% \label{straighttoc}
+    \@ifQueerEOL{\StraightEOL\gmd@@toc\QueerEOL}{\gmd@@toc}}}
+
+% As you'll learn from lines \ref{QueerEOL} and \ref{StraightEOL}, we
+% use those two strange declarations to change and restore the very
+% special meaning of the line end. Without such changes
+% |\tableofcontents| would cause a~disaster (it did indeed). And to
+% check the catcode of |^^M| is the r\^ole of |\@ifQueerEOL|:
+%
+% \changes{v0.98a}{06/09/06}{\cs{tableofcontents}, \cs{PrintIndex}
+% and \cs{PrintChanges} modified to work properly in the `queer line
+% ends' environment.}
+% \CodeDefine\@ifQueerEOL
+% \changes{v0.98a}{06/09/06}{added}
+\long\def\@ifQueerEOL#1#2{%
+  \ifnum\the\catcode`\^^M=\active \afterelsefi#1\else\afterfi#2\fi}
+
+% \CodeDefine\AtBegInputOnce The declaration below is useful if you
+% wish to put sth.\ just in the nearest input/included file and no
+% else: at the moment of putting the stuff it will erase it from the
+% hook. You may declare several |\AtBegInputOnce|s, they add up.
+% \changes{v0.98a}{06/09/05}{The counter \cs[]{hook@oncecnt} added
+%   to number the self-destructing cs's (thence the hooks add up) and
+%   the prefix changed to \cs[]{gmd/AtBI/NeuroOncer}}
+%
+% \changes{v0.98l}{06/10/26}{After the Tiger's suggestion, defining
+% a~unique macro for each use of \* substituted with just one macro in
+% the begin input hook and adding to this macro.}
+%
+%^^A CodeDefine*{hook@oncecnt}
+%^^A CodeDefIndex*\c@hook@oncecnt
+%^^A\newcounter{hook@oncecnt}
+% \CodeDefine\gmd@ABIOnce
+\@emptify\gmd@ABIOnce
+\AtBegInput\gmd@ABIOnce
+
+\long\def\AtBegInputOnce#1{%
+%^^A  \stepcounter{hook@oncecnt}%
+%^^A  \expandafter\long\expandafter\edef% \CodeUsgIndex*{NeuroOncer}
+%^^A    \csname gmd/AtBI/NeuroOncer\the\c@hook@oncecnt\endcsname{%
+%^^A      \noexpand\glet\csname gmd/AtBI/NeuroOncer\the\c@hook@oncecnt\endcsname\relax}%
+%^^A  \expandafter\addtomacro\csname gmd/AtBI/NeuroOncer\the\c@hook@oncecnt\endcsname{#1}%
+%^^A  \expandafter\AtBegInput\csname gmd/AtBI/NeuroOncer\the\c@hook@oncecnt\endcsname
+  \gaddtomacro\gmd@ABIOnce{\g@emptify\gmd@ABIOnce#1}}
+
+% Many tries of finishing the input cleanly led me to setting the
+% guardians as in line \ref{guardians} and to
+% \CodeDefine\EOFMark
+\def\EOFMark{\<eof>}
+% \label{eofMark}Other solutions did print the last code delimiter or
+% would require managing a~special case for the macros typesetting
+% \TeX\ code to suppress the last line's numbering etc.
+%
+% If you don't like it, see line \ref{NoEOF}.
+% \stanza
+%
+% Due to the |codespacesblank| option in the line \ref{cspb} we launch
+% the macro defined below to change the meaning of a~\pk{gmdoc}-kernel
+% macro.
+% \CodeDefIndex\gmd@DoTeXCodeSpace
+\begin{obeyspaces}%
+\gdef\gmd@DoTeXCodeSpace{%
+\obeyspaces\let =\breakabletwelvespace}%
+% \stanza
+% \CodeDefine\CodeSpacesBlank
+% \changes{v0.98a}{06/09/06}{a~kernel \cs{gmd@verbatimspace} renamed
+%   to \cs{gmd@texcodespace} and a~bug fixed.}
+% \CodeUsgIndex\gmobeyspaces
+% \CodeDefIndex\gmd@texcodespace
+\gdef\CodeSpacesBlank{%
+\let\gmd@DoTeXCodeSpace\gmobeyspaces%
+\let\gmd@texcodespace=\ }% the latter |\let| is for the |\if...|s.
+%\stanza
+% \CodeDefine\CodeSpacesSmall
+\gdef\CodeSpacesSmall{%
+\def\gmd@DoTeXCodeSpace{%
+\obeyspaces\def {\,\hskip\z@}}%
+\def\gmd@texcodespace{\,\hskip\z@}}%
+%
+\end{obeyspaces}
+
+% How the continuing of the narration should look like?
+% \CodeDefIndex\gmd@continuenarration
+\def\gmd@continuenarration{%
+  \endgroup
+  \gmd@countnarrationline% see \gmiflink[countnarr]{below}.
+  \expandafter\@trimandstore\the\everypar\@trimandstore
+  \everypar=\expandafter{\expandafter\@codetonarrskip\the\everypar}%
+  \expandafter\gmd@checkifEOL\@gobble}
+% Simple, isn't it? (We gobble the `other' code delimiter. Despite of
+% |\egroup| it's \catother\ because it was touched by |\futurelet|
+% contained in |\@ifnextchar| in line \ref{ifContNarr}. And in line
+% \ref{ContNarr2} it's been read as \catother. That's why it
+% works in spite of that |%| is of category `ignored'.)
+
+%\gmhypertarget[countnarr]{Whether} we count the narration lines
+% depends on the option |countalllines| which is off by default.
+% \CodeDefIndex\gmd@countnarrationline
+\if@countalllines
+  \def\gmd@countnarrationline{%
+    \if@newline
+      \grefstepcounter{codelinenum}\@newlinegfalse% the
+      % |\grefstepcounter| macro, defined in \pk{gmverb}, is a~global
+      % version of |\refstepcounter|, observing the redefinition made
+      % to |\refstepcounter| by \pk{hyperref}.
+      \everypar=\expandafter{%
+        \expandafter\@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.
+      \hyperlabel@line
+      {\LineNumFont\thecodelinenum}\,\ignorespaces
+    \fi}% \CodeDefIndex\gmd@countnarrationline
+\else
+  \@emptify\gmd@countnarrationline%
+\fi
+
+% And typesetting the \TeX\ code?
+\begingroup\catcode`\^^M=\active%
+\firstofone{\endgroup% \CodeDefIndex\gmd@typesettexcode
+  \def\gmd@typesettexcode{%
+    \gmd@parfixclosingspace% it's to eat a~space closing the
+    % paragraph, see \gmiflink[closingspace]{below}. It contains
+    % |\par|.
+    %
+    % A~verbatim group has already been opened by \cs{ttverb\+at\+im} and
+    % additional |\catcode|.
+    \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?
+    % \CodeDefine*{^^M}^^A
+    \def^^M{%
+      \@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, 'cause
+      % 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. 
+      \@ifnextchar{\gmd@texcodespace}{%
+        \@dsdirgfalse\gmd@dolspaces}{\gmd@charbychar}%
+    }% end of |^^M|'s definition.\label{debug!2}
+    % \CodeDefIndex\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\@codeskipputgtrue\@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%
+    \ifdim\gmd@ldspaceswd>\z@% and here the leading spaces.
+      \leavevmode\@dsdirgfalse%
+      \if@newline\grefstepcounter{codelinenum}\@newlinegfalse%
+      \fi%
+      \printlinenumber% if we don't want the lines to be numbered,
+      % the respective option \nlpercent|\let|s this CS to |\relax|.
+      \hyperlabel@line%
+      \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{%
+        \noexpand\hangindent=\the\hangindent% and also set the
+        % hanging indent setting for the same line comment case. BTW.,
+        % this |%| or rather lack of it costed me five hours of
+        % debugging and rewriting. Active lineends require extreme
+        % caution. 
+        \noexpand\hangafter=1\space}%^^A~\noexpand\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.
+    \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.
+  }%
+}% end of |\gmd@typesettexcode|'s definitions's group's |\firstofone|.
+%
+% \stanza
+%
+% Now let's deal with the leading spaces once forever. We wish not to
+% typeset | |s but to add the width of every leading space to the
+% paragraph's indent and to the hanging indent, but only if there'll
+% be any code character not being |%| in this line (e.g., the end of
+% line). If there'll be only |%|, we want just to continue the comment
+% or start a~new one. (We don't have to worry about whether we should
+% |\par| or not.) 
+% \CodeDefIndex\gmd@spacewd
+\newlength\gmd@spacewd% to store the width of a~(leading) 
+% | |\catother.
+% \CodeDefIndex\gmd@ldspaceswd
+\newlength\gmd@ldspaceswd% to store total length of gobbled leading
+                          %spaces.
+
+% It costed me some time to reach that in my verbatim scope a~space
+% isn't \catother\ but \catactive, namely |\let| to
+% |\breakabletwelvespace|.
+% So let us |\let| for future:
+% \CodeDefIndex\gmd@texcodespace
+\let\gmd@texcodespace=\breakabletwelvespace
+% And now let's try to deal with those spaces.
+% \CodeDefIndex\gmd@dolspaces
+\def\gmd@dolspaces{%
+  \ifx\gmd@texcodespace\@let@token
+    \@dsdirgfalse
+    \afterelsefi\settowidth{\gmd@spacewd}{\twelvespace}%
+    \gmd@ldspaceswd=\z@
+    \gmd@eatlspace%
+  \else\afterfi% about this smart macro and other of its family see
+  % \pk{gmutils} sec.\,3.
+    \par\gmd@typesettexcode
+  \fi}
+
+% And now, the iterating inner macro that'll eat the leading spaces.
+% \CodeDefIndex\gmd@eatlspace
+\def\gmd@eatlspace#1{%
+  \ifx\gmd@texcodespace#1%
+    \advance\gmd@ldspaceswd by\gmd@spacewd% we don't
+    % |\advance| it |\global|ly 'cause the current group may be closed
+    % iff we meet |%| and then we'll won't indent the line anyway.
+    \afterelseiffifi\gmd@eatlspace
+  \else
+    \if\code@delim\noexpand#1%
+      \gmd@ldspaceswd=\z@
+      \gmd@continuenarration#1%\label{ContNarr2}
+    \else \afterfifi\gmd@typesettexcode#1%
+    \fi
+  \fi}%
+
+% We want to know whether we were in hmode before reading current
+% |\code@delim|. We'll need to switch the switch globally.
+% 
+% \CodeDefine\ifprevhmode \CodeDefine*\prevhmodegtrue
+% \CodeDefine*\prevhmodegfalse
+\newgif\ifprevhmode
+
+% \addvspace\medskipamount
+%
+% And the main iterating inner macro which eats every single char of
+% verbatim text to check the end. The case |\%| should be excluded
+% and it is indeed.
+% \CodeDefine*\gmd@charbychar
+\newcommand*\gmd@charbychar[1]{%
+  \ifhmode\prevhmodegtrue
+  \else\prevhmodegfalse\fi 
+  \if\code@delim\noexpand#1%
+    \afterelseiffifi\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
+    \if\code@escape@char\noexpand#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.
+      \afterelsefifi\gmd@counttheline#1\scan@macro
+    \else
+      \afterfifi\gmd@EOLorcharbychar#1%
+    \fi
+  \fi}
+
+% One more inner macro because |^^M| in \TeX\ code wants to peek at the
+% next char and possibly launch |\gmd@charbychar|. We deal with
+% counting the lines thorougly. Increasing the counter is divided into
+% cases and it's very low level in one case because |\refstepcounter| and
+% |\stepcounter| added some stuff that caused blank lines, at
+% least with \pk {hyperref} package loaded.
+% \CodeDefIndex\gmd@EOLorcharbychar
+\def\gmd@EOLorcharbychar#1{%
+  \ifx\gmd@texcodeEOL#1%
+    \if@newline
+      \if@countalllines\global\advance\c@codelinenum by\@ne
+      \@newlinegfalse\fi
+    \fi
+    \afterelsefi#1%\label{printhashone1}here we print |#1|.
+  \else% i.e., |#1| is \emph{not} a~(very active) line end,
+    \afterfi
+    \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}
+
+% \CodeDefIndex\gmd@counttheline
+\def\gmd@counttheline{%
+  \ifvmode
+    \if@newline
+      \grefstepcounter{codelinenum}\@newlinegfalse
+      \hyperlabel@line
+    \fi
+    \printlinenumber
+    \mark@envir
+  \else
+    \if@newline
+      \grefstepcounter{codelinenum}\@newlinegfalse
+      \hyperlabel@line
+    \fi
+  \fi}
+
+% If before reading current |%| char we were in horizontal mode, then
+% we wish to print |%| (or another code delimiter).
+% \CodeDefIndex\gmd@percenthack
+\def\gmd@percenthack{%
+  \ifprevhmode\code@delim\aftergroup\space% We add a~space after |%|,
+  % 'cause I~think it looks better. It's done |\aftergroup| to make
+  % the spaces possible after the |%| not to be
+  % typeset.
+  \else\aftergroup\gmd@narrcheckifds@ne% 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}
+
+\def\gmd@narrcheckifds@ne#1{%^^A\typeout{narr@ne if ds \on@line}%
+  \@dsdirgfalse\@ifnextchar<{%
+    \expandafter\gmd@docstripdirective\@gobble}{#1}}
+
+% The macro below is used to look for the |%^^M| case to make
+% a~commented blank line make a~new paragraph.
+% Long searched and very simple at last.
+% \CodeDefIndex\gmd@checkifEOL
+\def\gmd@checkifEOL{%
+  \gmd@countnarrationline
+  \everypar=\expandafter{\expandafter\@codetonarrskip% we add the
+    % macro that'll insert a~vertical space if we leave the code and
+    % enter the narration.
+    \the\gmd@preverypar}%
+  \@ifnextchar{\gmd@textEOL}{%
+    \@dsdirgfalse\par\ignorespaces}{\gmd@narrcheckifds}}%
+
+% We check if it's |%<|, a~\ds\ directive that is.
+\def\gmd@narrcheckifds{%^^A\typeout{narr if ds \on@line}%
+  \@dsdirgfalse\@ifnextchar<{%
+    \expandafter\gmd@docstripdirective\@gobble}{\ignorespaces}}
+    
+
+% In the `mixed' line case it should be a~bit more complex, though. On
+% the other hand, there's no need to checking for \ds\ directives.
+% \CodeDefIndex\gmd@checkifEOLmixd
+\def\gmd@checkifEOLmixd{%
+  \gmd@countnarrationline
+  \everypar=\expandafter{\expandafter\@codetonarrskip\the\gmd@preverypar}%
+  % 
+  % \label{longlinethatshouldbebroken}
+  \@afternarrgfalse\@aftercodegtrue
+  \ifhmode\@codeskipputgfalse\fi
+  \@ifnextchar{\gmd@textEOL}{%
+    {\raggedright\gmd@endpe\par}% without |\raggedright| this |\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.
+    % \CodeDefIndex\par
+    \def\par{%
+      \ifhmode% (I~added this |\ifhmode| as a~result of a~heavy
+      % debug.) 
+%^^A      \typeout{par of 1766 in hmode \on@line}%
+        \@@par
+        \if@afternarr
+          \if@aftercode
+            \if@codeskipput\else\gmd@codeskip2\@aftercodegfalse\@codeskipputgtrue\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=\expandafter{%
+          \expandafter\@codetonarrskip\the\gmd@preverypar}%
+        \let\par\@@par%
+%^^A     else
+%^^A        \typeout{par of 1766 in NOT hmode \on@line}%
+      \fi}%
+    \gmd@endpe\ignorespaces}}
+
+% As we announced, we play with |\leftskip| inside the verbatim
+% group and therefore we wish to restore normal |\leftskip| when back to
+% normal text i.e.\ the commentary. But, if normal text starts in the
+% same line as the code, then we still wish to indent such a~line.
+% \CodeDefIndex\gmd@endpe
+\def\gmd@endpe{%
+  \ifprevhmode
+    \settexcodehangi%\unskip ndent
+    \leftskip=\CodeIndent
+%^^A    \typeout{gmd@endpe in prevhmode \on@line}%
+  \else
+    \leftskip=\TextIndent
+    \hangindent=\z@
+    \everypar=\expandafter{%
+      \expandafter\@codetonarrskip\the\gmd@preverypar}%
+%^^A    \typeout{gmd@endpe in not prevhmode \on@line}%
+  \fi}
+
+%
+%
+% \subdivision{Numbering (or Not) of the Lines}
+%
+% Maybe you want codelines to be numbered and maybe you want to
+% reset the counter within sections.
+\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]{%
+    % \CodeDefIndex*\c@codelinenum
+    \newcounter{codelinenum}[#1]}% \dots\ and provide a~new
+  % declaration of the counter.
+\else% With the option turned off\dots \CodeDefine*{DocInputsCount}
+  \newcounter{DocInputsCount}% \CodeDefine*{codelinenum}
+  \newcounter{codelinenum}[DocInputsCount]% \dots\ we declare the
+  % |\DocInput|s' number counter andthe 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}}
+  % \CodeDefine\gmd@resetlinecount
+  \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
+
+% Let's define printing the line number as we did in \pk{gmvb}
+% package.
+% \CodeDefIndex\printlinenumber
+\newcommand*\printlinenumber{%
+  \leavevmode\llap{\rlap{\LineNumFont$\phantom{999}$\llap{\thecodelinenum}}%
+    \hskip\leftskip}}
+% \CodeDefine\LineNumFont
+\def\LineNumFont{\normalfont\tiny}
+
+\if@linesnotnum\@relaxen\printlinenumber\fi
+% \CodeDefIndex\hyperlabel@line
+\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}
+
+%
+% \subdivision{Spacing with \cs{everypar}}
+%
+% Last but not least, let's define the macro inserting a~vertical space
+% between the code and the narration.  Its parameter is
+% a~\gmhypertarget[relic parameter]{relic} of a~very heavy debug of
+% the automatic vspacing mechanism. Let it remain at least until this
+% package is 2.0 version.  \CodeDefIndex\gmd@codeskip
+\newcommand*\gmd@codeskip[1]{\@@par\addvspace\CodeTopsep\@codeskipputgtrue}
+%\label{codeskip2}
+
+% Sometimes we add the |\CodeTopsep| vertical space in
+% |\everypar|. When this happens, first we remove the |\parindent|
+% empty box, but this doesn't reverse putting |\parskip| to the main
+% vertical list. And if |\parskip| is put, |\addvspace| shall see it
+% not the `true' last skip. \Describe*{@codeskipput}Therefore we need
+% a~Boolean switch to keep the knowledge of putting similar vskip
+% before |\parskip|.
+% \CodeDefine\if@codeskipput \CodeDefIndex*\@codeskipputgtrue
+% \CodeDefIndex*\@codeskipputgfalse \CodeDefIndex*{@codeskipput}
+\newgif\if@codeskipput
+
+% The below is another relic of the heavy debug of the automatic
+% vspacing. Let's give it the same removal clause as 
+% \gmiflink[relic parameter]{above}. 
+% \CodeDefIndex\gmd@nocodeskip
+\newcommand*\gmd@nocodeskip[2]{}
+
+% And here is how the two relic macros looked like during the
+% debug. As you see, they are disabled by a~false |\if| (look at it
+% closely ;-).
+\if1 1
+  \renewcommand*\gmd@codeskip[1]{%
+    \hbox{\rule{1cm}{3pt} #1!!!}}
+  \renewcommand*\gmd@nocodeskip[2]{%
+    \hbox{\rule{1cm}{0.5pt} #1: #2 }}
+\fi
+
+% We'll wish to execute |\gmd@codeskip| wherever a~codeline (possibly with
+% an inline comment) is followed by a~homogenic comment line or
+% reverse. Let us dedicate a~Boolean switch to this then.
+% \CodeDefine\if@aftercode \CodeDefine*\@aftercodegtrue
+% \CodeDefine*\@aftercodegfalse
+\newgif\if@aftercode
+% This switch will be set true in the moments when we are able to
+% switch from the \TeX\ code into the narration and the below one when
+% we are able to switch reversely.
+% \CodeDefine\if@afternarr \CodeDefine*\@afternarrgtrue
+% \CodeDefine*\@afternarrgfalse
+\newgif\if@afternarr
+
+% To insert vertical glue between the \TeX\ code and the narration
+% we'll be playing with |\everypar|. More precisely, we'll add a~macro
+% that the |\parindent| box shall move and the glue shall put. 
+% \CodeDefine\@codetonarrskip
+\long\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\@codeskipputgtrue
+        %\label{codeskip5}
+        \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're
+        % subsequent code lines after an inline comment not ended with
+        % an explicit |\par|.)
+        \else\gmd@nocodeskip4{naC}%
+      \fi%
+    \fi
+  \fi\@aftercodegfalse}
+
+% But we play with |\everypar| for other reasons too, and while restoring it,
+% we don't want to add the |\@codetonarrskip| macro infinitely many
+% times. So let us define a~macro that'll check if |\everypar| begins
+% with |\@codetonarrskip| and trim it if so. We'll use this macro with
+% proper |\expandafter|ing in order to give it the contents of
+% |\everypar|. The work should be done in two steps first of which
+% will be checking whether |\everypar| is nonempty (we can't have two
+% delimited parameters for a~macro: if we define a~two-parameter
+% macro, the first is undelimited so it has to be nonempty; it costed
+% me some one hour to understand it).
+% \CodeDefine\@trimandstore
+% \CodeDefIndex\@trimandstore@hash
+\long\def\@trimandstore#1\@trimandstore{%
+  \def\@trimandstore@hash{#1}%
+  \ifx\@trimandstore@hash\@empty% we check if |#1| is
+  % nonempty. The |\if|\nlpercent*|\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\expandafter\@trimandstore@ne\the\everypar\@trimandstore
+  \fi}
+
+% \CodeDefIndex\@trimandstore@ne
+\long\def\@trimandstore@ne#1#2\@trimandstore{%\label{@trimandstore}
+  \def\trimmed@everypar{#2}%
+  \ifx\@codetonarrskip#1%
+    \gmd@preverypar=\expandafter{\trimmed@everypar}%
+  \else
+    \gmd@preverypar=\expandafter{\the\everypar}%
+  \fi}
+% We prefer not to repeat |#1| and |#2| within the |\if|s and we even
+% define an auxiliary macro because |\everypar| may contain some
+% |\if|s or |\fi|s.
+%
+%
+% \subdivision[Life Among Queer EOLs]{Life Among Queer \gmhypertarget[afc]{EOLs}}
+%
+% When I~showed this package to my \TeX\ Guru he commended it and
+% immediately pointed some disadvantages in the comparison with the
+% \docfm\ package.
+%
+% One of them was an expected difficulty of breaking a~moving argument
+% (e.g., of a~sectioning macro) in two lines. To work it around let's
+% define a~line-end eater:
+\catcode`\^^B=\active% note we re|\catcode| \<char2> globally, for the
+% entire document.
+\bgroup\catcode`\^^M=\active%
+\firstofone{\egroup% \CodeDefIndex\QueerCharTwo \CodeDefine*{^^B}
+  \def\QueerCharTwo{\long\def^^B##1^^M{\ignorespaces}}}
+
+\QueerCharTwo
+
+\AtBegInput{\@ifQueerEOL{\catcode`\^^B\active}{}\QueerCharTwo}% \label{QCh2}
+% We repeat redefinition of \<char2> at begin of the documenting
+% input, because \pk{doc.dtx} suggests that some packages (namely
+% \pk{inputenc}) may re|\catcode| such unusual characters.
+%
+% \changes{v0.98d}{06/9/11}{re\cs{catcode}ing and defining
+% \cs{char1} and \cs{char2} added atto begin doc input}
+%
+% \stanza
+% As you see the |^^B| active char is defined to gobble everything
+% since itself till the end of line and the very end of line. This is
+% intended for harmless continuing a~line. The price is affecting the
+% line numbering when |countalllines| option is enabled.
+%
+% I~also liked the \docfm's idea of comment${}^2$ i.e., the
+% possibility of marking some text so that it doesn't appear nor in the
+% working version neither in the documentation, got by making
+% |^^A| (i.e., \<char1>) a~comment char. 
+%
+% However, in this package such a~trick would work another way: here
+% the line ends are active, a~comment char would disable them and that
+% would cause disasters. So let's do it an |\active| way.
+\catcode`\^^A=\active% note we re|\catcode| \<char1> globally, for the
+% entire document.
+\bgroup\catcode`\^^M=\active%
+\firstofone{\egroup% \CodeDefIndex\QueerCharOne \CodeDefine*{^^A}
+  \def\QueerCharOne{\long\def^^A##1^^M{\ignorespaces^^M}}}
+
+\QueerCharOne
+
+\AtBegInput{\@ifQueerEOL{\catcode`\^^A\active}\QueerCharOne}% see note
+% after line \ref{QCh2}.
+
+
+% As I~suggested in the users' guide, |\StraightEOL| and |\QueerEOL| are
+% intended to cooperate in harmony for the user's good. They take care
+% not only of redefining the line end but also these little things
+% related to it.
+%
+% One usefulness of |\StraightEOL| is allowing linebreaking of the
+% command arguments. Another making possible executing some code lines
+% during the documentation pass.
+%
+% \changes{v0.98b}{06/09/07}{\cs{QueerM} and \cs{StraightM} renamed
+% to \cs{QueerEOL} and \cs{StraightEOL} and other (internal) macros
+% ending with \cs[]{M}}
+% \CodeDefine\StraightEOL
+\def\StraightEOL{%\label{StraightEOL}
+  \catcode`\^^M=5\relax
+  \catcode`\^^A=14\relax
+  \catcode`\^^B=14\relax
+  \def\^^M{\ }}
+
+% \CodeDefine\QueerEOL
+\def\QueerEOL{%\label{QueerEOL}
+  \catcode`\^^M=\active
+  \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}
+
+% \iffalse
+% They won't work in \verb+|verbatim|+ scopes but the clubs redefine
+% the active lineend to be a~breakable (and visible) space, so
+% breaking a~`clubbed' text in two lines will not cause a~disaster but
+% you get e.g., |...^^B %...|.
+% \fi
+
+% \gmhypertarget[closingspace]{To make} |^^M| behave more like a~`normal'
+% lineend I~command it to add a~| |${}\subs{10}$ at first. It works
+% but has one uwelcome feature: if the line has nearly |\textwidth|,
+% this closing space may cause line breaking and setting a~blank line.
+% To fix this I~|\advance| the |\parfillskip|:
+% \CodeDefIndex\gmd@parfixclosingspace
+\def\gmd@parfixclosingspace{{%
+    \advance\parfillskip by-\gmd@closingspacewd\par}}
+% We'll put it in a~group surrounding |\par| but we need to check if
+% this |\par| is executed after narration or after the code,
+% i.e., whether the closing space was added or not.
+% \CodeDefIndex\gmd@closingspacewd
+% \CodeDefIndex\gmd@setclosingspacewd
+\newskip\gmd@closingspacewd
+\newcommand*\gmd@setclosingspacewd{%
+  \global\gmd@closingspacewd=\fontdimen2\font%
+    plus\fontdimen3\font minus\fontdimen4\font\relax}
+%
+% See also line \ref{noclosingspace} to see what we do in the codeline
+% case when no closing space is added.
+
+% And one more detail:
+\bgroup\catcode`\^^M=\active%
+\firstofone{\egroup% \CodeDefIndex*{\^^M}
+  \def\^^M{\ \expandafter\ignorespaces^^M}}
+% To make |\^^M| behave properly. If this definition was ommitted, it
+% would just expand to |\ | and thus not gobble the leading |%| of the
+% next line leave alone typesetting the \TeX\ code.  I~type |\ | etc.
+% instead of just |^^M| which adds a~space itself because I~take
+% account of a~possibility of redefining the |\ | CS by the user, just
+% like in normal \TeX.
+% \CodeDefIndex\gmd@bslashEOL
+\let\gmd@bslashEOL=\^^M
+% We'll need it for restoring queer definitions for \docfm-compatibility.
+%
+%
+% \subdivision{Adjustment of \env{verbatim} and \cs{verb}}
+%
+% To make \env{verbatim(*)} typeset its contents with the \TeX\ code's
+% indentation: 
+% \CodeDefIndex\@verbatim
+\gaddtomacro\@verbatim{\leftskip=\CodeIndent}
+
+% And a~one more little definition to accomodate |\verb| and pals for the
+% lines commented out.
+% \CodeDefIndex\check@percent
+\AtBegInput{\long\def\check@percent#1{%
+    \expandafter\ifx\code@delim#1\else\afterfi#1\fi}}
+% We also redefine \pk{gmverb}'s |\AddtoPrivateOthers| that has been
+% provided just with \pk{gmdoc}'s need in mind.
+% \CodeDefine\AddtoPrivateOthers
+\def\AddtoPrivateOthers#1{%
+  \expandafter\def\expandafter\doprivateothers\expandafter{%
+    \doprivateothers\do#1}}%
+
+% We also redefine an internal |\verb|'s macro |\gm@verb@eol| to put a~proper
+% line end if a~line end char is met in a~short verbatim: we have to
+% check if we are in `queer' or `straight' EOLs area.
+\begingroup
+\obeylines% \CodeDefIndex\gm@verb@eol
+\AtBegInput{\def\gm@verb@eol{\obeylines%
+    \def^^M{\verb@egroup\@latex@error{%
+        \noexpand\verb ended by end of line}%
+      \@ifQueerEOL{\gmd@textEOL}{\@ehc}}}}%
+\endgroup
+
+% \subdivision{Macros for Marking The Macros}
+%
+% A~great inspiration for this part was the \docfm\ package again.
+% I~take some macros from it, and some tasks I~solve a~different way,
+% e.g., the |\| (or another escapechar) is not active, because anyway
+% all the chars of code are scanned one by one. And exclusions from
+% indexing are supported not with a~list stored as |\toks| register
+% but with separate control sequences for each excluded CS.
+%
+% \stanza
+%
+% The \docfm\ package shows a~very general approach to the indexing
+% issue. It assumes using a~special MakeIndex style and doesn't use
+% explicit MakeIndex controls but provides specific macros to hide
+% them. But here in \pk{gmdoc} we prefer no special style for the
+% index. 
+
+% \CodeDefine\actualchar \CodeDefine\quotechar \CodeDefine\encapchar 
+% \CodeDefine\levelchar 
+% \changes[\actualchar]{v0.98e}{06/09/23}{a~bug fix: making the
+% \cs[]{@} \protect\catother\space again}
+\edef\actualchar{\string @} 
+\def\quotechar{"}
+\edef\encapchar{\twelveclub} 
+\def\levelchar{!}
+% However, for the glossary, i.e., the change history, a~special style
+% is required, e.g., \pk{gmglo.ist}, and the above macros are
+% redefined by the |\changes| command due to \pk{gmglo.ist} and
+% \pk{gglo.ist} settings.
+%
+% Moreover, if you insist on using a~special MakeIndex style, you may
+% redefine the above four macros in the preamble. The |\edef|s that
+% process them further are postponed till |\begin{document}|.
+%
+%
+% \CodeDefine\CodeEscapeChar \CodeDefine\code@escape@char
+\def\CodeEscapeChar#1{%
+  \begingroup
+  \escapechar\m@ne
+  \xdef\code@escape@char{\string#1}%
+  \endgroup}
+% As you see, to make a~proper use of this macro you should give it
+% the~|\|\<char> CS as an argument. It's an invariant assertion that
+% |\code@escape@char| stores  `other' version of the code layer
+% escape char.
+% \CodeUsgIndex\CodeEscapeChar
+\CodeEscapeChar\\
+
+% As mentioned in \docfm, someone may have some chars \catletter ed.
+\@ifundefined{MakePrivateLetters}{%\label{MPL}
+  % \CodeDefine\MakePrivateLetters 
+  \def\MakePrivateLetters{\makeatletter\catcode`\*=11 }}{}
+% A~tradition seems to exist to write about e.g., `command |\section| and
+% command |\section*|' and such an uderstanding also of `macro' is
+% noticeable in \docfm. Making the |*| a~letter solves the problem of
+% scanning starred commands.
+%
+% \stanza
+% And you may wish some special chars to be \catother.
+% \CodeDefine\MakePrivateOthers
+\def\MakePrivateOthers{\let\do=\@makeother \doprivateothers}
+% This macro we use to re|\catcode| the space for marking the
+% environments' names and the caret for marking chars such as |^^M|, see
+% line \ref{TextUsage}. So let's define the list:
+% \CodeDefine\doprivateothers
+\def\doprivateothers{\do\ \do\^}
+% Two chars for the beginning, and also the |\MakeShortVerb| command
+% shall this list enlarge with the char(s) declared.
+% (There's no need to add the backslash to this list since all the
+% relevant commands |\string| their argument whatever it is.)
+%
+% \changes{v0.98c}{06/9/9}{\cs{} removed from \cs{doprivateothers}
+% list and \cs{string} added instead to commands that process the
+% `environments' names;}
+
+% Now the main macro indexing a~macro's name. It would be a~verbatim
+% :-) copy of the \docfm 's one if I~didn't ommit some lines
+% irrelevant with my approach.
+% \CodeDefine\scan@macro
+\def\scan@macro#1{% we are sure to scan at least one token and
+                  % therefore we define this macro as one-parameter.\par
+% Unlike in \docfm , here we have the escape char \catother\ so we
+% may just have it printed during main scan char by char, i.e., in the
+% lines \ref{printhashone1} and \ref{printhashone2}.\par
+% So, we step the checksum counter first,
+  \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\noexpand#1%
+    \quote@char#1%
+    \xdef\macro@iname{\maybe@quote#1}% global for symmetry with
+    % line \ref{x123}.
+    \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 indexentry formats, see line
+    % \ref{pnametestDef} etc.
+    \afterelsefi\@ifnextcat{a}{\continue@macroscan}{\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{x123}
+    \quote@char#1%
+    \xdef\macro@iname{\maybe@quote\macro@iname}%
+    \xdef\macro@pname{\xiistring#1}%\label{stringing1}
+    \afterfi \finish@macroscan
+  \fi}
+% This is the end\dots\ beautiful friend, the end\dots\ of
+% |\scan@macro|'s |\def| the end\dots\
+% The |\xiistring| macro, provided by \pk{gmutils}, is used instead of
+% original |\string| because we wish to get the space \catother.
+%
+% \stanza
+% Now, let's explain some details, i.e., let's define them. We call
+% the following macro having known |#1| to be \catletter.
+% \CodeDefIndex\continue@macroscan
+\def\continue@macroscan#1{%
+  \quote@char#1%
+  \xdef\macro@iname{\macro@iname \maybe@quote#1}%
+  \xdef\macro@pname{\macro@pname \string#1}%\label{stringing2}
+  \@ifnextcat{a}{\continue@macroscan}{\finish@macroscan}%
+}
+% The |\@ifnextcat| macro is defined analogously to |\@ifnextchar| but
+% the test done by it is not |\ifx| but |\ifcat|.
+% \CodeDefIndex\quote@char
+\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{%
+      \@ifismember 9\of \indexcontrols\glet\maybe@quote\quotechar%
+      \else\g@emptify\maybe@quote%
+      \fi}%
+  }}
+
+% And now let's take care of the MakeIndex control characters. We'll
+% define a~list of them to check whether we should quote a~char or
+% not. But we'll do it at |\begin{document}| to allow the user to use
+% some special MakeIndex style and in such a~case to redefine the four
+% MakeIndex controls' macros. We enrich this list with the backslash
+% because sometimes MakeIndex didn't like it unquoted.
+% \CodeDefine\indexcontrols
+% \changes{v0.98d}{06/9/15}{relativized to the index control macros:
+% previously the controls were given explicitly as the standard ones.}
+\AtBeginDocument{\xdef\indexcontrols{%
+    \bslash\levelchar\encapchar\actualchar\quotechar}}
+  % \CodeDefine\@ifismember
+  \long\def \@ifismember#1\of#2{%
+    \long\def\in@@##1#1##2\in@@{%
+      \if\noexpand^^A##2\noexpand^^A\afterelsefi\gmd@@iffalse
+      \else\afterfi\gmd@@iftrue
+      \fi}%
+    \expandafter\in@@#2#1\in@@}%
+
+% To hide the Booleans from \TeX\ in some contexts define
+% \CodeDefine\gmd@@iftrue \CodeDefine\gmd@@iffalse
+\def\gmd@@iftrue{\iftrue}
+\def\gmd@@iffalse{\iffalse}
+% A~try to |\let| instead of |\def| resulted with the `incomplete if'
+% error: \TeX\ looks at the meanings not the names while counting
+% |\if|s. 
+
+% A~word of commentary. \docfm\ in another situation solves the
+% problem of checking similar way, but a~little more complexely.  On
+% the other hand, \docfm 's solution is more general: here a~trouble
+% may arise if the list begins with |^^A|. It seems little possible to
+% me, though.  Btw., this macro is used only for catching chars that are
+% MakeIndex's controls so far. How does it work?
+%
+% |\quote@char| sort of re|\catcode|s its argument through the
+% |\uccode| trick: assigns the argument as the uppercase code of the
+% digit 9 and does further work in the |\uppercase|'s scope so the
+% digit 9 (a~benchmark `other') is substituted by |#1| but the
+% |\catcode| remains so thus the |\@ifismember| macro gets
+% |\quote@char|'s |#1|  `other'ed as the first argument. 
+%
+% |\@ifismember| is defined to take two parameters separated by |\of|.
+% In |\quote@char| the second argument for it is |\indexcontrols|
+% defined as the (expanded) sequence of the MakeIndex controls.
+% |\@ifismember| defines its inner macro |\in@@| to take two
+% parameters separated by the first and the second |\@ifismember|'s
+% parameter, which are here the char investigated by |\quote@char| and
+% the |\indexcontrols| list. The inner macro's parameter string is
+% delimited by the macro itself, why not. |\in@@| is put before
+% a~string consisting of |\@ifismember|'s second and first parameters
+% (in such a~reversed order) and |\in@@| itself. In such a~sequence it
+% looks for something fitting its parameter pattern. |\in@@| is sure
+% to find the parameters delimiter (|\in@@| itself) and the separator,
+% |\ifismember|'s |#1| i.e., the investigated char, because they are
+% just there. But the investigated char may be found not near the end,
+% where we put it, but among the MakeIndex controls' list. Then the
+% rest of this list and |\ifismember|'s |#1| put by us become the
+% secong argument of |\in@@|. What |\in@@| does with its arguments, is
+% just a~check whether the second one is empty. This may happen \emph{iff}
+% the investigated char hasn't been found among the MakeIndex controls'
+% list and then |\in@@| shall expand to |\iffalse|, otherwise it'll
+% expand to |\iftrue|. (The |\after...| macros are employed not to
+% (mis)match just got |\if...| with the test's |\fi|.)  ``(Deep
+% breath.) You got that?'' If not, try \docfm 's explanation of
+% |\ifnot@excluded|, pp.\,36--37 of the v2.1b dated 2004/02/09
+% documentation, where a~similar construction is attributed to Michael
+% Spivak.
+
+% \CodeDefIndex\finish@macroscan
+\newcommand*\finish@macroscan{%\par
+  % We have the macro's name for indexing in |\macro@iname| and
+  % for print in |\macro@pname|. So we index it. We do it a~bit
+  % countercrank way because we wish to use more general indexing
+  % macro.
+  \if\verbatimchar\macro@pname% 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 |$|:
+    \def\im@firstpar{[$]}%^^A$
+  \else\def\im@firstpar{}\fi
+  \expandafter \index@macro\im@firstpar\macro@iname\macro@pname
+  \maybe@marginpar\macro@pname
+  \macro@pname
+  \gmd@charbychar
+}
+
+% Now, the~macro that checks whether the just scanned macro should be
+% put into a~marginpar: it checks the meaning of a~very special CS:
+% whose name consists of |gmd/2marpar/| and of the examined macro's
+% name. 
+% \CodeDefIndex\maybe@marginpar
+\def\maybe@marginpar#1{%
+  %^^A\ifx#1\mname@tomarginpar
+  \@ifundefined{gmd/2marpar/#1}{}{%
+    \expandafter\Text@Marginize\expandafter{\bslash#1}% |\expandafter|s 
+    % \possfil'cause the |\Text@Marginize| command applies |\string| to its
+    % argument.  \nlpercent |\macro@pname|, which will be the only possible
+    % argument to
+    % \cs{may\+be\+@mar\+g\+in\+par},  
+    % contains the macro's name 
+    % without the escapechar so we added it here.
+    % ^^A\gdef\mname@tomarginpar{}
+    \expandafter\g@relaxen\csname gmd/2marpar/#1\endcsname% we reset the switch.
+  }}
+
+% The inner macro indexing macro. |#1| is the |\verb|'s delimiter;
+% |#2| is assumed to be the macro's name with MakeIndex-control chars
+% quoted. |#3| is a~macro storing the \catother\ macro's name, usually
+% |\macro@pname|, built with |\string|ing every char in lines ^^A
+% \ref{stringing0}, \ref{stringing1} and \ref{stringing2}. |#3| is
+% used only to test if the entry should be specially formatted.
+% \changes{v0.98}{06/09/05}{Indexing the codelines improved to sort
+%   with account of the \cs{filesep} (\cs{HLPrefix})}
+% \CodeDefine\index@macro
+\newcommand*\index@macro[3][\verbatimchar]{{%
+    \@ifundefined{gmd/iexcl/#3}%\label{iexcltest}
+    {% |#3| is not excluded from index
+      \@ifundefined{gmd/defentry/#3}% \label{pnametestDef} 
+      {% |#3| is not def entry
+        \@ifundefined{gmd/usgentry/#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\IndexRefCs{gmdindexpagecs{\HLPrefix}{\kind@fentry}{\EntryPrefix}}%        
+      \else
+        \edef\IndexRefCs{gmdindexrefcs{\HLPrefix}{\kind@fentry}{\EntryPrefix}}%
+      \fi
+      \edef\@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\IndexRefCs}% 
+      \expandafter\special@index\expandafter{\@tempa}% We give the
+      % indexing macro the argument expanded so that \pk{hyperref} may
+      % see the explicit encapchar 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.
+    }}
+
+%\CodeDefine\un@defentryze \CodeDefine\un@usgentryze
+\def\un@defentryze#1{%
+  \expandafter\g@relaxen\csname gmd/defentry/#1\endcsname
+  \g@relaxen\last@defmark}% we care to clear the last definition
+% marker checked by |\changes|.
+\def\un@usgentryze#1{%
+  \expandafter\g@relaxen\csname gmd/usgentry/#1\endcsname}
+
+\@emptify\EntryPrefix% this macro seems to be obsolete now
+% (v0.98d).
+
+% For the case of page-indexing a~macro in the commentary when
+% codeline index option is on:
+% \CodeDefine\if@pageinclindex \CodeDefine*\@pageinclindextrue
+% \CodeDefine*\@pageinclindexfalse
+\newif\if@pageinclindex
+
+% \CodeDefine\quoted@eschar
+\newcommand*\quoted@eschar{\quotechar\bslash}% we'll redefine it when
+% indexing an environment.
+
+% Let's initialize |\IndexPrefix|
+\def\IndexPrefix{}
+% The |\IndexPrefix| and |\HLPrefix| (`HyperLabel Prefix') macros are
+% given with account of a~possibility of documenting several files
+% in(to) one document.  In such case the user may for each file
+% |\def\IndexPrefix{|\<package name>|!}| for instance and it will work
+% as main level index entry and |\def\HLPrefix{|\<package name>|}|
+% as a~prefix in hypertargets in the codelines. They are redefined by
+% |\DocInclude| e.g.
+
+\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.
+    \let\gmdindexpagecs=\gmdindexrefcs
+  \else
+% \CodeDefIndex\gmdindexrefcs
+% \CodeDefIndex\gmdindexpagecs
+    \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 \CodeDefIndex\gmd@revprefix
+    \def\gmd@revprefix#1{%
+      \def\@tempa{#1}%
+      \ifx\@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
+    % 'cause during the authomatic |\hypertarget|ing the lines are labeled
+    % |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}
+% The definition is postponed till |\begin{document}| because of the
+%   |\PageIndex| declaration (added for \docfm-compatibility), see
+%   line \ref{PageIndex}.
+%
+% I~design the index to contain hyperlinking numbers whether they are
+% the line numbers or page numbers. In both cases the last parameter
+% is the number, the one before the last is the name of a~formatting macro
+% and in linenumber case the first parameter is a~prefix for proper
+% reference in multi-doc.
+
+
+% I~take account of three kinds of formatting the numbers: 1.~the
+% `def' entry, 2.~a~`usage' entry, 3.~a~common entry. As in \docfm , let
+% them be underlined, italic and upright respectively.
+% \CodeDefine\DefEntry \CodeDefine\UsgEntry
+\def\DefEntry#1{\underline{#1}}
+\def\UsgEntry#1{\textit{#1}}
+% The third option will be just |\relax| by default:
+% \CodeDefine\CommonEntryCmd
+\def\CommonEntryCmd{relax}
+% In line \ref{CECmd} it's |\edef|ed to allow an `unm\"oglich'
+% situation that the user wants to have the common index entries
+% specially formatted. I~use this to make \emph{all} the index entries
+% of the (|\SelfInclude|d) driver file to be `usage', see codeline
+% \ref{UsgHack} of \file{gmdocDoc.tex}.
+
+% Now let's |\def| the macros declaring a~CS to be indexed
+% special way. Each declaration puts the \catother ed name of the
+% macro given it as the argument into proper macro to be |\ifx|ed in
+% lines \ref{pnametestDef} and \ref{pnametestUsg} respectively.
+%
+% But first let's make a~version of |\@ifstar| that would work with
+% |*|\catletter. It's analogous to |\@ifstar|.
+% \CodeDefine\@ifstarl
+\bgroup\catcode`\*=11 %
+\firstofone{\egroup
+  \def\@ifstarl#1{\@ifnextchar *{\@firstoftwo{#1}}}}
+
+% Now we are ready to define a~couple of commands. % The |*| versions of
+% them are for marking environments and \emph{implicit} CSs.
+
+% \CodeDefine\CodeDefIndex
+% \CodeDefine*\CodeDefIndex*
+\outer\def\CodeDefIndex{\begingroup
+  \MakePrivateLetters
+  \@ifstarl{\MakePrivateOthers\Code@DefIndexStar}{\Code@DefIndex}}
+
+% \changes{v0.98d}{06/9/11}{The macros indexing and marginizing
+% macros (and other sequences) made \cs{long}}
+%\CodeDefIndex\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.
+    \global\@defentryze{#1}}}
+% \label{stringamacro}
+
+% \CodeDefIndex\Code@DefIndexStar
+\long\def\Code@DefIndexStar#1{%
+  \endgroup
+  \addto@estoindex{#1}%
+  \@defentryze{#1}}
+
+% \CodeDefine\gmd@justadot
+\def\gmd@justadot{.}
+
+%\CodeDefine\@defentryze \CodeDefine\@usgentryze
+\long\def\@defentryze#1{%
+  \expandafter\let\csname gmd/defentry/\string#1\endcsname\gmd@justadot% The
+  % \LaTeX\ |\@namedef| macro could not be used since it's not `long'.
+  % \CodeDefine\last@defmark
+  \xdef\last@defmark{\string#1}}% we |\string| the argument 'cause it's
+% a~control sequence most probably.
+\long\def\@usgentryze#1{%
+  \expandafter\let\csname gmd/usgentry/\string#1\endcsname\gmd@justadot}
+
+% Initialize |\envirs@toindex|
+\@emptify\envirs@toindex
+
+% Now we'll do the same for the `usage' entries:
+%\CodeDefine\CodeUsgIndex
+%\CodeDefine*\CodeUsgIndex*
+\outer\def\CodeUsgIndex{\begingroup
+  \MakePrivateLetters
+  \@ifstarl{\MakePrivateOthers\Code@UsgIndexStar}{\Code@UsgIndex}}
+% The |*| possibility is for marking environments etc.
+% \CodeDefIndex\Code@UsgIndex
+\long\def\Code@UsgIndex#1{\endgroup{%
+    \escapechar\m@ne
+    \global\@usgentryze{#1}}}
+
+% \CodeDefIndex\Code@UsgIndexStar
+\long\def\Code@UsgIndexStar#1{%
+  \endgroup
+  \addto@estoindex{#1}%
+  \@usgentryze{#1}}
+
+% For the symmetry, if we want to mark a~control sequence or an
+% environment's name to be indexed as a~`normal' entry, let's have:
+% \CodeDefine\CodeCommonIndex
+% \CodeDefine\CodeCommonIndex*
+\outer\def\CodeCommonIndex{\begingroup
+  \MakePrivateLetters
+  \@ifstarl{\MakePrivateOthers\Code@CommonIndexStar}{\Code@CommonIndex}}
+
+% \CodeDefIndex\Code@CommonIndex
+\long\def\Code@CommonIndex#1{\endgroup}
+
+% \CodeDefIndex\Code@CommonIndexStar
+\long\def\Code@CommonIndexStar#1{%
+  \endgroup\addto@estoindex{#1}}
+
+
+% And now let's define commands to index the control sequences and
+% environments occurring in the narrative.
+% \CodeDefine\text@indexmacro
+\long\def\text@indexmacro#1{%
+  {\escapechar\m@ne \xdef\macro@pname{\xiistring#1}}%
+  \expandafter\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}.
+    \expandafter \index@macro\im@firstpar\macro@iname\macro@pname}}
+
+% The macro defined below (and the next one) are executed only before
+% a~\catother\ macro's name i.e.\ a~nonempty sequence of \catother\
+% character(s). This sequence is delimited (guarded) by |\relax|.
+% \CodeDefine\quote@mname
+\def\quote@mname{%
+  \def\macro@iname{}%
+  \quote@charbychar}
+
+% \CodeDefIndex\quote@charbychar
+\def\quote@charbychar#1{%
+  \if\relax#1% finish quoting when you meet |\relax| or:
+  \else
+    \quote@char#1%
+    \xdef\macro@iname{\macro@iname \maybe@quote#1}%
+    \afterfi\quote@charbychar
+  \fi}
+
+% The next command will take one argument,
+% which in plain version should be a~control sequence and in the
+% starred version also a~sequence of chars allowed in environment names or
+% made other by |\MakePrivateOthers| macro, taken in the curly braces.
+% \CodeDefine\TextUsgIndex
+% \CodeDefine*\TextUsgIndex*
+\def\TextUsgIndex{\begingroup
+  \MakePrivateLetters
+  \@ifstarl{\MakePrivateOthers\Text@UsgIndexStar}{\Text@UsgIndex}}
+
+% \CodeDefIndex\Text@UsgIndex
+\long\def\Text@UsgIndex#1{%
+  \endgroup\@usgentryze#1%
+  \text@indexmacro#1}
+
+% \CodeDefIndex\Text@UsgIndexStar
+\long\def\Text@UsgIndexStar#1{\endgroup\@usgentryze{#1}%
+  \text@indexenvir{#1}}
+
+% \CodeDefine\text@indexenvir
+\long\def \text@indexenvir#1{%
+  \edef\macro@pname{\xiistring#1}%
+  \if\bslash\expandafter\@firstofmany\macro@pname\@@nil% if
+  % |\string|ed |#1| begins with a~backslash, we will gobble it
+  % to make MakeIndex not see it.
+    \edef\@tempa{\expandafter\@gobble\macro@pname}%
+    \@tempswatrue
+  \else
+    \let\@tempa\macro@pname
+    \@tempswafalse
+  \fi
+  \expandafter\quote@mname\@tempa\relax% \label{quote@mname comm}we 
+  % process |\sting|ed |#1| char by char and quote MakeIndex
+  % controls. |\relax| is the iterating macro's stopper. The quoted
+  % |\string|ed |#1| shall be the expansion 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}}
+
+% \CodeDefine\TextCommonIndex
+% \CodeDefine*\TextCommonIndex*
+\def\TextCommonIndex{\begingroup
+  \MakePrivateLetters
+  \@ifstarl{\MakePrivateOthers\Text@CommonIndexStar}{\Text@CommonIndex}}
+
+%\CodeDefIndex\Text@CommonIndex
+\long\def\Text@CommonIndex#1{\endgroup
+  \text@indexmacro#1}
+
+% \CodeDefIndex\Text@CommonIndexStar
+\long\def\Text@CommonIndexStar#1{\endgroup
+  \text@indexenvir{#1}}
+
+% As you see in the lines \ref{defentryrs} and \ref{usgentryrs}, the
+% markers of special formatting are reset after first use.
+
+% But we wish the CSs not only to be indexed special way but also
+% to be put in marginpars. So:
+% \CodeDefine\CodeMarginize
+% \CodeDefine*\CodeMarginize*
+\outer\def\CodeMarginize{\begingroup
+  \MakePrivateLetters
+  \@ifstarl{\MakePrivateOthers\egCode@MarginizeEnvir}{\egCode@MarginizeMacro}}
+
+% One more expansion level because we wish |\Code@MarginizeMacro| not
+% to begin with |\endgroup| because in the subsequent macros it's used
+% \emph{after} ending the re|\catcode|ing group.
+% \CodeDefIndex\egCode@MarginizeMacro
+\long\def\egCode@MarginizeMacro#1{\endgroup
+  \Code@MarginizeMacro#1}
+
+% \CodeDefIndex\Code@MarginizeMacro
+\long\def\Code@MarginizeMacro#1{{\escapechar\m@ne
+    \expandafter\glet\csname gmd/2marpar/\string#1\endcsname\gmd@justadot
+%^^A    \xdef\mname@tomarginpar{\xiistring#1}
+  }}
+
+% \CodeDefIndex\egCode@MarginizeEnvir
+\long\def\egCode@MarginizeEnvir#1{\endgroup
+  \Code@MarginizeEnvir{#1}}
+
+% \CodeDefIndex\Code@MarginizeEnvir
+\long\def\Code@MarginizeEnvir#1{\addto@estomarginpar{#1}}
+
+% And a~macro really putting the environment's name in a~marginpar
+% shall be trigged at the beginning of the nearest codeline.
+%
+% Here it is:
+% \CodeDefine\mark@envir
+\def\mark@envir{%\label{mark@envir}
+  \ifx\envirs@tomarginpar\@empty
+  \else
+    \let\do\Text@Marginize
+    \envirs@tomarginpar%
+    \g@emptify\envirs@tomarginpar%
+  \fi
+  \ifx\envirs@toindex\@empty
+  \else
+    \def\do##1{% the |\envirs@toindex| list contains |\string|ed
+      % macros or environments' names in braces and each  preceded
+      % with |\do|.
+      \if\bslash\@firstofmany##1\@@nil% if
+      % |##1| begins with a~backslash, we will gobble it for
+      % MakeIndex not see it.
+        \edef\@tempa{\@gobble##1}%
+        \@tempswatrue
+      \else
+        \edef\@tempa{##1}\@tempswafalse
+      \fi
+      \expandafter\quote@mname\@tempa\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}}}%
+    \envirs@toindex
+    \g@emptify\envirs@toindex%
+  \fi}
+
+% One very important thing: initialisation of the list macros:
+\@emptify\envirs@tomarginpar
+\@emptify\envirs@toindex
+
+
+% For convenience we'll make the `private letters' first not to bother
+% ourselves with |\makeatletter| for instance when we want mark some
+% CS. And |\MakePrivateOthers| for the environment and other string
+% case.
+% \CodeDefine\CodeDefine
+% \CodeDefine*\CodeDefine*
+\outer\def\CodeDefine{\begingroup
+  \MakePrivateLetters
+% We do |\MakePrivateLetters| before |\@ifstarl| in order to avoid
+% a~situation that \TeX\ sees a~control sequence with improper name
+% (another CS than we wished)
+% (because |\@ifstarl| establishes the |\catcode|s for the next token):
+  \@ifstarl{\MakePrivateOthers\Code@DefEnvir}{\Code@DefMacro}}
+
+% \CodeDefine\CodeUsage
+% \CodeDefine*\CodeUsage*
+\outer\def\CodeUsage{\begingroup
+  \MakePrivateLetters
+  \@ifstarl{\MakePrivateOthers\Code@UsgEnvir}{\Code@UsgMacro}}
+
+% And then we launch the macros that close the group and do the work. 
+% \CodeDefIndex\Code@DefMacro
+\long\def\Code@DefMacro#1{%
+  \Code@DefIndex#1% we use the internal macro; it'll close the group.
+  \Code@MarginizeMacro#1}
+
+% \CodeDefIndex\Code@UsgMacro
+\long\def\Code@UsgMacro#1{%
+  \Code@UsgIndex#1% here also the internal macro; it'll close the group
+  \Code@MarginizeMacro#1}
+
+% The next macro is taken verbatim ;-) from \docfm\ and the subsequent
+% |\let|s, too.
+% \CodeDefIndex\codeline@wrindex
+\def\codeline@wrindex#1{\if@filesw
+  \immediate\write\@indexfile
+  {\string\indexentry{#1}%
+    {\HLPrefix\number\c@codelinenum}}\fi}
+
+% We initialize it due to the option (or lack of the option):
+\AtBeginDocument{%
+  \if@pageindex
+    \let\special@index=\index
+  \else
+  % \CodeDefIndex\special@index
+    \let\special@index=\codeline@wrindex
+    % \label{codeline indexnumber declaration}
+  \fi}% postponed till |\begin{document}| with respect of \docfm-like
+% declarations.
+
+% And in case we don't want to index:
+% \CodeDefine\gag@index
+\def\gag@index{\let\index=\@gobble
+  %\label{gag@index}
+  \let\codeline@wrindex=\@gobble}
+% We'll use it in one more place or two. And we'll wish to be able to
+% undo it so
+% let's copy the original meanings:
+% \CodeDefIndex\@@index \CodeDefIndex\@@codeline@wrindex
+\StoreMacros{\index\codeline@wrindex}
+% \CodeDefine\ungag@index
+\def\ungag@index{\RestoreMacros{\index\@@codeline@wrindex}}
+  %\label{ungag@index}
+  
+
+
+% Our next task is to define macros that'll mark and index an
+% environment or other string in the code. Because of lack of
+% a~backslash, no environment's name is scanned so we have to proceed
+% different way. But we wish the user to have symmetric tools, i.e.,
+% the `def' or `usage' use of an environment should be declared before
+% the line where the environment occurs. Note the slight difference
+% between these and the commands to declare a~CS marking: the latter
+% do not require to be used \emph{immediately} before the line containig the
+% CS to be marked.  We separate indexing from marginizing to leave
+% a~possibility of doing only one of those things.
+% \CodeDefIndex\Code@DefEnvir
+\long\def\Code@DefEnvir#1{%
+  \endgroup
+  \addto@estomarginpar{#1}%
+  \addto@estoindex{#1}%
+  \@defentryze{#1}}
+
+% \CodeDefIndex\Code@UsgEnvir
+\long\def\Code@UsgEnvir#1{%
+  \endgroup
+  \addto@estomarginpar{#1}%
+  \addto@estoindex{#1}%
+  \@usgentryze{#1}}
+
+% \CodeDefIndex\addto@estomarginpar
+\long\def\addto@estomarginpar#1{%
+  \edef\@tempa{\noexpand\do{\xiistring#1}}% we |\string| the argument to
+  % allow it to be a~control sequence.
+  \expandafter\addtomacro\expandafter\envirs@tomarginpar\expandafter{\@tempa}}
+
+% \CodeDefIndex\addto@estoindex
+\long\def\addto@estoindex#1{%
+  \edef\@tempa{\noexpand\do{\xiistring#1}}
+  \expandafter\addtomacro\expandafter\envirs@toindex\expandafter{\@tempa}}
+
+
+% And now a~command to mark a~`usage' occurrence of a~CS, environment
+% or another string in the commentary. As the `code' commands this also
+% has plain and starred version, first for CSs appearing explicitly and the
+% latter for the strings and CSs appearing implicitly.
+% \CodeDefine\TextUsage
+% \CodeDefine*\TextUsage*
+\def\TextUsage{\begingroup
+  %\label{TextUsage}
+  \MakePrivateLetters
+  \@ifstarl{\MakePrivateOthers\Text@UsgEnvir}{\Text@UsgMacro}}
+
+% \CodeDefIndex\Text@UsgMacro
+\long\def\Text@UsgMacro#1{%
+  \endgroup{\tt\xiistring#1}%
+  \Text@Marginize#1%
+  \begingroup\Code@UsgIndex#1% we declare the kind of formatting of the entry.
+  \text@indexmacro#1}
+
+% \CodeDefIndex\Text@UsgEnvir
+\long\def\Text@UsgEnvir#1{%
+  \endgroup{\tt\xiistring#1}%
+  \Text@Marginize{#1}%
+  \@usgentryze{#1}% we declare the `usage' kind of formatting of the
+  % entry and index the sequence |#1|.
+  \text@indexenvir{#1}}
+
+% We don't provide commands to mark a~macro's or environment's
+% definition present within the narrative because we think there won't
+% be any: one defines macros and environments in the code not in
+% the commentary. 
+% \CodeDefine\TextMarginize
+% \CodeDefine*\TextMarginize*
+\def\TextMarginize{\begingroup
+  \MakePrivateLetters
+  \@ifstarl{\MakePrivateOthers\egText@Marginize}{\egText@Marginize}}
+
+% \CodeDefIndex\egText@Marginize
+\long\def\egText@Marginize#1{\endgroup
+  \Text@Marginize#1}
+
+% We check whether the margin pars are enabled and proceed
+% respectively in either case.
+\if@marginparsused
+  \reversemarginpar
+  \marginparpush\z@
+  \marginparwidth8pc\relax
+  % ^^A  \settowidth{\marginparsep}{\ \ }%
+  %
+  % You may wish to put not only macros and environments to
+  % a~marginpar. 
+  % \CodeDefine\gmdmarginpar
+  \long\def\gmdmarginpar#1{%
+    \marginpar{\raggedleft\strut
+      \hskip0ptplus100ptminus100pt%
+      #1}}%\stanza
+  % 
+\else
+  \long\def\gmdmarginpar#1{}%
+\fi
+
+% \CodeDefine\Text@Marginize
+\long\def\Text@Marginize#1{%
+  \gmdmarginpar{\marginpartt\xiistring#1}}
+% Note that the above macro will just gobble its argument if the
+% marginpars are disabled.^^A\NoEOF so far O.K.
+%
+% It may be advisable to choose a~condensed typewriter font for the
+% marginpars, if there is any. (The Latin Modern font family provides
+% a~light condensed typewriter font, it's set in \pk{gmdocc} class.)
+\let\marginpartt\tt
+
+% If we count all lines then the index entries for CSs and
+% environments marked in the commentary should have codeline numbers
+% not page numbers and that is |\let| in line 
+% \ref{codeline indexnumber declaration}.  On the other hand, if we
+% count only the codelines, then a~macro or an environment marked in
+% the commentary should have page number not codeline number. This we
+% declare here, among others\ we add the letter |p| before the page number.
+% \CodeDefIndex\do@properindex
+\def\do@properindex{%\label{do@properindex}
+  \if@countalllines\else
+    \@pageinclindextrue
+    \let\special@index=\index
+  \fi}
+
+% \stanza
+% In \docfm\ all the `working' \TeX\ code should be braced in(to) the
+% \env{macrocode} environments. Here another solutions are taken so to
+% be \docfm-compatible we only should
+%  nearly-ignore \env{macrocode(*)}s
+% with their Percent and The Four Spaces Preceding ;-)\,. I.e., to ensure
+% the line ends are `queer'. And that the \ds\ directives will be
+% typeset as the \ds\ directives. And that the usual code escape
+% char will be restored at |\end{macrocode}|. And to add the vertical
+% spaces.
+%
+% ^^A~\NoEOF so far OK
+% If you know \docfm\ conventions, note that \gmdoc\ \emph{does not}
+% require |\end{macrocode}| to be preceded ^^A(  
+% with any particular number of any char :-)\,.
+% \changes{v0.98b}{06/09/07}{To \env{macrocode(*)} default
+% definitions \cs{QueerEOL} added}
+%\CodeDefine*{macrocode*} 
+\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 | |
+% visible, which is the default in \pk{gmdoc} outside
+% \env{macrocode}.
+%
+% So we should make the spaces \emph{invisible} for the unstarred
+% version.
+%\CodeDefine*{macrocode}
+\newenvironment*{macrocode}{%\label{macrocode}
+  \if@codeskipput\else\par\addvspace\CodeTopsep\@codeskipputgtrue\fi
+  \CodeSpacesBlank\QueerEOL}%^^A\docstrips
+  {\par\addvspace\CodeTopsep\CodeEscapeChar\\}
+% Note that at the end of both the above environments the |\|'s r\^ole
+% as the code escape char is restored. This is crafted for the
+% |\SpecialEscapechar| macro's compatibility: this macro influences
+% only the first \env{macrocode} environment. The situation that the
+% user wants some queer escape char in general and in a~particular
+% \env{macrocode} yet another seems to me ``unm\"oglich,
+% Prinzessin''\footnote{Richard Strauss after Oscar Wilde, ^^B
+% \textit{Salome}.}. 
+%
+% \stanza
+% Since the first \file{.dtx} I~tried to compile after the first
+% published version of \gmdoc\ uses a~lot of commented out code in
+% \env{macrocode}s, it seems to me necessary to add a~possibility to
+% typeset \env{macrocode}s as if they were a~kind of \env{verbatim},
+% that is to leave the code layer and narration layer philosophy.
+%
+%\CodeDefine*{oldmc}
+%\CodeDefine*{oldmc*}
+\let\oldmc\macrocode
+\let\endoldmc\endmacrocode
+\@namelet{oldmc*}{macrocode*}
+\@namelet{endoldmc*}{endmacrocode*}
+
+% Now we arm \env{oldmc} and \env{olmc*} with the 
+% the macro looking for |%   \end|\arg{envir name}.
+\addtomacro\oldmc{\@oldmacrocode@launch}%
+\expandafter\addtomacro\csname oldmc*\endcsname{%
+  \@oldmacrocode@launch}
+
+% \CodeDefIndex\@oldmacrocode@launch
+\def\@oldmacrocode@launch{%
+  \emptify\gmd@textEOL% to disable it in |\gmd@docstripdirective|
+  % launched within the code.
+  \glet\stored@code@delim\code@delim
+  \@makeother\^^B\CodeDelim*\^^B%
+  \ttverbatim \gmd@DoTeXCodeSpace%
+  \@makeother\|% 'cause |\ttverbatim| doesn't do that.
+  \MakePrivateLetters% see line \ref{MPL}.\par
+  % ^^A  \expandafter\@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 occure 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.
+  \expandafter\gmd@currenvxistar\@currenvir*\relax
+  \@oldmacrocode}
+
+\bgroup\catcode`*11
+\firstofone{\egroup
+  \def\gm@xistar{*}}
+
+\def\gmd@currenvxistar#1*#2\relax{%
+  \edef\@currenvir{#1\if*#2\gm@xistar\fi}}
+% The trick is that |#2| may be either |*|\catother\ or empty. If it's
+% |*|, the test is satisfied and |\if...\fi| expands to
+% |\gm@xistar|. If |#2| is empty, the test is also satisfied since
+% |\gm@xistar| expands to |*| but there's nothing to expand to. So, if
+% the environment's name ends with |*|\catother, it'll be substituted
+% with |*|\catletter or else nothing will be added. (Note that a~|*|
+% not at the end of env.\ name would cause a~disaster.)
+
+% \CodeDefine\@oldmacrocode
+\bgroup
+\catcode`[=1 \catcode`]=2
+\catcode`\{=\active \@makeother\}
+\@makeother\^^B
+\catcode`!=0 \catcode`\\=\active
+% \CodeEscapeChar\! \CDAnd
+!catcode`&=14 !catcode`*=11
+!catcode`!%=!active !obeyspaces&
+!firstofone[!egroup&
+&%\stanza
+!def!@oldmacrocode[&
+!bgroup!let =!relax&% to avoid writing |!noexpand | four times.
+!xdef!oldmc@def[&
+!def!noexpand!oldmc@end####1!noexpand%    !noexpand\end&
+!noexpand{!@currenvir}[&
+####1^^B!noexpand!end[!@currenvir]!noexpand!gmd@oldmcfinis]]&
+!egroup&% now |\oldmc@edef| is defined to have one parameter delimited
+&% with |\end|\arg{current env.'s name}
+!oldmc@def&
+!oldmc@end]&% \CDPerc
+]
+
+\def\gmd@oldmcfinis{%
+  \expandafter\CodeDelim\stored@code@delim
+  \gmd@mchook}% see line \ref{gmd@mchook}
+
+
+% \CodeDefine\VerbMacrocodes
+\def\VerbMacrocodes{%
+  \let\macrocode\oldmc
+  \@namelet{macrocode*}{oldmc*}}
+
+
+% To handle \ds\ directives in the code (in the old macrocodes
+% case that is).
+
+%    \begin{oldmc}
+\bgroup\catcode`\%\active
+\firstofone{\egroup
+\def\docstrips@percent{\catcode`\%\active
+  \let%\gmd@codecheckifds}}
+%    \end{oldmc}
+
+% The point is, the active |%| will be expanded when just after it is
+% the |\gmd@charbychar| cs token and next is some char, the |^^B| code
+% delimiter at least. So, if that char is |<|, we wish to launch
+% \ds\ directive typesetting. (Thanks to |\ttverbatim| all the |<|
+% are `other'.)
+\def\gmd@codecheckifds#1#2{% note that |#1| is just to gobble
+  % \nlpercent|\gmd@charbychar| token.
+% ^^A  \typeout{code if ds \on@line}%
+  \if@dsdir\@dsdirgfalse
+    \if\noexpand<\noexpand#2\afterelsefifi\gmd@docstripdirective
+    \else\afterfifi\twelvepercent#1#2%
+    \fi
+  \else\afterfi\twelvepercent#1#2%
+  \fi}
+%
+% \begin{environment}{macro}
+% Almost  the same we do with the \env{macro(*)} environments, stating
+% only their argument to be processed as the `def' entry. Of course,
+% we should re|\catcode| it first.
+% \CodeDefIndex*\macro
+\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}%
+% \CodeDefIndex*\endmacro
+  {\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 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} occure \env{macro}s that mark
+  % \emph{implicit} definitions i.e., such that the defined CS is not
+  % scanned in the subsequent code.
+%
+% \begin{macro*}{macro*}
+% And for those who want to to use this environment
+% for marking implicit definitions, define the star
+% version:
+\@namedef{macro*}{\let\gmd@ifonetoken\@secondoftwo\macro}
+% \CodeDefIndex*\endmacro*
+\expandafter\let\csname endmacro*\endcsname\endmacro
+% Note that \env{macro} and \env{macro*} have the same effect for
+% more-than-one-token arguments thanks to |\gmd@ifonetoken|'s meaning
+% inside unstarred \env{macro} (it checks whether the argument is
+% one-token and if it isn't, |\gmd@ifonetoken| switches execution to
+% `other sequence' path).
+%
+% The two environments behave different only with a~one-token
+% argument: \env{macro} postpones indexing it till the first scanned
+% occurrence while \env{macro*} till the first code line met.
+% \end{macro*}
+% \end{environment}
+%
+% \stanza
+% Now, let's complete the details.
+% First define an |\if|-like macro  that turns true when the string
+% given to it consists of just one token (or one |{|\<text>|}|, to
+% tell the whole truth).
+% \CodeDefine\gmd@ifsingle
+\def\gmd@ifsingle#1#2\@@nil{%
+  \def\@tempa{#2}%
+  \ifx\@tempa\@empty}
+% Note it expands to an open |\if...| test (unbalanced with |\fi|) so it
+% has to be used as all the |\if|s, with optional |\else| and
+% obligatory |\fi|. And cannot be used in the possibly skipped
+% branches of other |\if...|s (then it would result with `extra
+% |\fi|/extra |\else|' errors). But the below usage is safe since both
+% |\gmd@ifsingle| and its |\else| and |\fi| are hidden in a~macro
+% (that will not be |\expandafter|ed).
+%
+% Note also that giving |\gmd@ifsingle| an |\if...| or so as the first
+% token of the argument will not confuse \TeX\ since the first token 
+% is just gobbled. The possibility of occurrence of  |\if...| or so as
+% a~not-first token seems to be negligible.
+% \CodeDefine\gmd@ifonetoken
+\def\gmd@ifonetoken#1#2#3{%
+  \def\@tempb{#3}% We hide |#3| from \TeX\ in case it's |\if...| or
+  % so. \nlpercent|\@tempa| is used in \nlpercent|\gmd@ifsingle|. 
+  \gmd@ifsingle#3\@@nil
+    \afterelsefi\expandafter#1\@tempb%
+  \else
+    \edef\@tempa{\expandafter\string\@tempb}%
+    \afterfi\expandafter#2\expandafter{\@tempa}%
+  \fi}
+% Now, define the mysterious |\Hybrid@DefMacro| and |\Hybrid@DefEnvir|
+% macros. They mark their argument with a~certain subtlety: they put
+% it in a~marginpar at the point where they are and postpone indexing
+% it till the first scanned occurrence or just the first code line met.
+% \CodeDefine\Hybrid@DefMacro \CodeDefine\Hybrid@DefEnvir
+\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}}
+
+%\CodeDefine\Text@MarginizeNext
+\long\def\Text@MarginizeNext#1{%
+  \gmd@evpaddonce{\Text@Marginize{#1}\ignorespaces}}
+%^^A~\strut from before Text@Marginize deleted as unnecessary in everypar.
+
+% The following macro adds its argument to |\everypar| using an auxiliary
+% macro to wrap the stuff in. The auxiliary macro has
+% a~self-destructor built in so it |\relax|es itself after first use.
+%\CodeDefine\gmd@evpaddonce
+\long\def\gmd@evpaddonce#1{%
+  \stepnummacro\gmd@oncenum
+  \expandafter\long\expandafter\edef% \CodeUsgIndex*{NeuroOncer}
+    \csname gmd/evp/NeuroOncer\gmd@oncenum\endcsname{%
+      \noexpand\g@relaxen
+      \csname gmd/evp/NeuroOncer\gmd@oncenum\endcsname}% Why does it
+    % work despite it shouldn't? Because when the CS got with
+    % \nlpercent|\csname...\endcsname| is undefined, it's equivalent
+    % |\relax| and therefore unexpandable. That's why it passes
+    % \nlpercent|\edef| and is able to be assigned.
+  \expandafter\addtomacro\csname gmd/evp/NeuroOncer\gmd@oncenum\endcsname{#1}%
+  \expandafter\addto@hook\expandafter\everypar\expandafter{%
+    \csname gmd/evp/NeuroOncer\gmd@oncenum\endcsname}%
+}
+
+\nummacro\gmd@oncenum% We store the number uniquifying the auxiliary
+% macro in a~macro to save count registers (cf.\ \pk{gmutils}
+% sec.\ \gmiflink{To Save Precious Count Registers}).
+
+% \begin{environment}{environment}
+% Wrapping a~description and definition of an environment in
+% a~\env{macro} environment would look inappropriate (`zgrzyta\l o by'
+% in Polish) although there's no \TeX nical obstacle to do
+% so. Therefore we define the \env{environment}, because of \ae^^B
+% sthetic and psychological reasons.
+\expandafter\let\expandafter\environment\csname macro*\endcsname
+\expandafter\let\expandafter\endenvironment\csname endmacro*\endcsname
+% \end{environment}
+%
+
+
+%\subdivision[Index Exclude List]{\gmhypertarget{Index Exclude List}}
+% We want some CSs not to be indexed, e.g., the \LaTeX\
+% internals and \TeX\ primitives.
+%
+% \docfm\ takes |\index@excludelist| to be a~|\toks| register to store
+% the list of expelled CSs. Here we'll deal another way. For each CS
+% to be excluded we'll make (|\let|, to be precise) a~control sequence
+% and then we'll be checking if it's undefined (|\ifx|-equivalent
+% |\relax|).\footnote{This idea comes from Marcin Woli\'nski.}
+%
+% \CodeDefine\DoNotIndex
+% \changes{v0.97}{06/09/04}{\cs{MakePrivateLetters} added}
+% \changes{v0.98}{06/09/05}{Removed since it had not worked. The
+% macros in the argument should be separated with commas. I~understood
+% why it didn't work: 'cause 't was iside a~command. So I~put it back}
+% \changes{v0.98d}{06/9/11}{Unmade \cs{global}}
+\def\DoNotIndex{\bgroup\MakePrivateLetters\DoNot@Index}
+
+% \CodeDefIndex\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.
+  \dont@index#1.}
+
+% \CodeDefIndex\dont@index
+\long\def\dont@index#1{%
+  \def\@tempa{\noexpand#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\@tempa.% a~dot finishes expelling
+  \else
+    \if\@tempa,% The list this macro is put before may
+    % contain commas and that's O.K., we just continue the work.
+      \afterelsefifi\dont@index
+    \else% what is else shall off the Index be expelled.
+      {\escapechar\m@ne
+        \xdef\@tempa{\string#1}}%
+      \expandafter\let%
+      \csname gmd/iexcl/\@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}
+
+%^^A(
+% Let's now give the exclude list copied \*verbatim ;-) from
+% \pk{doc.dtx}. I~give it in the code layer 'cause I~suppose one will
+% document not \LaTeX\ source but normal packages.
+% 
+% \begin{CodeSpacesSmall}
+% \DoNotIndex\DoNotIndex \CodeDefine\DefaultIndexExclusions
+\DoNotIndex\{ \DoNotIndex\}% \label{DNIbraces}the index entries of
+% these two CSs would be rejected by MakeIndex anyway.
+
+\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|.
+  \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 'cause
+    % 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/providecommand| 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
+    % IMO.
+    \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: \nlpercent|\@topsep| \nlpercent|\@topsepadd|^^B 
+    % \ \nlpercent|\abovedisplayskip| \nlpercent|\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 \in@@}%
+    \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
+    % |%%\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}
+% \end{CodeSpacesSmall}
+
+% And we execute it due to the (lack of) counter-corresponding option:
+\if@indexallmacros\else
+  \DefaultIndexExclusions
+\fi
+
+
+% If we expelled so many CSs, someone may like it in general but
+% he/she may need one or two expelled to be indexed back. So
+%
+% \CodeDefine\DoIndex
+% \changes{v0.97}{06/09/04}{\cs{MakePrivateLetters} added.}
+% \changes{v0.98}{06/09/05}{\cs{MakePrivateLetters} removed since it
+% hadn't worked. I~understood why it didn't work: 'cause
+% 't was iside a~command and I~put it back}
+% \changes{v0.98d}{06/9/11}{Unmade \cs{global}}
+\def\DoIndex{\bgroup\MakePrivateLetters\Do@Index}
+
+% \CodeDefIndex\Do@Index
+\long\def\Do@Index#1{\egroup\@relaxen\gmd@iedir\dont@index#1.}% note
+% we only redefine an auxiliary CS and launch also |\dont@index| inner
+% macro.
+
+% And if a~user wants here make default exclusions and there do not
+% make them, \heshe\ may use the |\DefaultIndexExclusions| declaration
+% \himher self. This declaration OCSR, but anyway let's provide the
+% counterpart. It OCSR, too.
+%
+% \CodeDefine\UndoDefaultIndexExclusions
+% \changes{v0.98d}{06/9/11}{Unmade \cs{global}}
+\def\UndoDefaultIndexExclusions{%
+  \StoreMacro\DoNotIndex
+  % ^^A\dont@index
+  \let\DoNotIndex\DoIndex
+  % ^^A  \let\dont@index=\do@index
+  \DefaultIndexExclusions
+  % ^^A  \RestoreMacro\dont@index
+  \RestoreMacro\DoNotIndex}
+
+%
+%
+% \subdivision{Index Parameters}
+%
+% \begin{quotation}The |\IndexPrologue| macro is used to place a short
+% message into the document above the index. It is implemented by
+% redefining |\index@prologue|, a macro which holds the default
+% text. We'd better make it a |\long| macro to allow |\par| commands 
+% in its argument.\end{quotation} \CodeDefine\IndexPrologue
+\long\def\IndexPrologue#1{\@bsphack\def\index@prologue{#1}\@esphack}
+% \label{IndexPrologue}
+
+% \CodeDefine\indexdiv
+% \changes{v0.98a}{06/09/06}{modified to contain the star so that
+% it may be redefined for tests.}
+\def\indexdiv{\@ifundefined{chapter}{\section*}{\chapter*}}
+% \CodeDefine\index@prologue
+\@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.
+  }}{}
+
+% During the preparation of this package for publishing I~needed
+% only to add something at the end of the default index prologue. So
+% \CodeDefine\AtDIPrologue
+% \changes{v0.98c}{06/9/8}{added 'cause of need}
+% \CodeDefIndex\gmd@dip@hook \CodeDefine\AtDIPrologue
+\@emptify\gmd@dip@hook
+\long\def\AtDIPrologue#1{\g@addto@macro\gmd@dip@hook{#1}}
+
+% The Author(s) of \docfm\ assume \pk{multicol} is known not to
+% everybody. My assumption is the other so
+\RequirePackage{multicol}
+
+% \begin{quotation}If \pk{multicol} is in use, when the index is
+%   started we compute 
+%   the remaining space on the current page; if it is greater than
+%   |\IndexMin|, the first part of the index will then be placed in the
+%   available space. The number of columns set is controlled by the
+%   counter |\c@IndexColumns| which can be changed with a
+%   |\setcounter| declaration.\end{quotation}
+% \CodeDefine\IndexMin
+\newdimen\IndexMin \IndexMin = 133pt\relax% originally it was set
+% 80\,pt, but with my default prologue there's at least 4.7\,cm needed
+% to place the prologue and some index entries on the same page.
+% \CodeDefine*{IndexColumns}
+% \CodeDefIndex*{\c@IndexColumns}
+\newcount\c@IndexColumns \c@IndexColumns = 3
+% \CodeDefine*{theindex}
+\renewenvironment{theindex}
+  {\begin{multicols}\c@IndexColumns[\index@prologue][\IndexMin]%
+      \IndexLinksBlack
+      \IndexParms \let\item\@idxitem \ignorespaces}%
+    {\end{multicols}}
+
+%\CodeDefine\IndexLinksBlack
+\def\IndexLinksBlack{\hypersetup{linkcolor=black}}% To make Adobe
+% Reader work faster.
+
+% \CodeDefine\IndexParms
+\@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 ifinite 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.
+    }}{}
+
+  % \CodeDefine\PrintIndex
+\def\PrintIndex{% we ensure the standard meaning of the line end
+  % character not to cause a~disaster.
+  \@ifQueerEOL{\StraightEOL\printindex\QueerEOL}{\printindex}}
+
+% Remember that if you want to change not all the parameters, you
+% don't have to redefine the entire |\IndexParms| macro but you may
+% use a~very nice \LaTeX\ command |\g@addto@macro| (it has |\global|
+% effect, also with an apeless name (|\gaddtomacro|) provided by
+% \pk{gmutils}. (It adds its second argument at the end of definition
+% of its first argument provided the first argument is a~no-argument
+% macro.) Moreover, \pk{gmutils} provides also 
+% |\addtomacro| that has the same effect except it's not |\global|.
+%
+%
+%
+% \subdivision{The \ds\ Directives}
+%
+% \StraightEOL
+% \iffalse
+%
+% This passage is obsoleted on 2006/11/30.
+%
+% In the \gmdoc\ thinking, the \ds\ directives belong to the
+% narration layer since they all begin with |%|. For now I~don't
+% have a~more minimal idea than to define a~pair of macros the first of
+% which would |\active|ate the |<| and the other, assigned to the
+% active |<|, would do the work and re`other' the |<| back.
+%
+% We provide two versions of the seargant-activating macro: the first
+% makes only the first subsequent seargant typeset a~\ds\
+% directive (and the previous meaning of the |<| will be restored, the
+% latter doesn't restore the previous meaning of |<| after (by) the
+% first occurrence and the latter obeys usual scoping rules while the
+% scoping rules of the first (|\docstrip|) are more strict: the scope
+% is delimited by the first occurrence of the subject~|<|.
+%
+% Let us remeber that, just as all the control sequences except
+% |\relax|, you can use |\docstrips| as an environment, i.e., in the
+% |\begin{docstrips}|\dots |\end{docstrips}| syntax. And, that
+% \env{macrocode} environment declares it.
+% \bgroup\catcode`\<=\active
+% \firstofone{\egroup
+%   % \stanza
+%   % \CodeDefine\docstrip
+%   \newcommand*\docstrip{%
+%     \gmd@storesearg
+%     \catcode`\<=\active
+%     \gmd@setdocstrips
+%     \let\gmd@docstripshook=\gmd@restoresearg}%
+%   % \stanza
+%   % \CodeDefine\docstrips
+%   \newcommand*\docstrips{%
+%     \catcode`\<=\active
+%     \gmd@setdocstrips
+%     \@emptify\gmd@docstripshook}%
+%   % \stanza
+%   % \CodeDefIndex\gmd@setdocstrips
+%   \def\gmd@setdocstrips{%
+%     \def<{\ifmmode\sgtleftxii\else\afterfi\gmd@docstripdirective\fi}}
+%   %\stanza
+%   % \CodeDefIndex\gmd@storesearg
+%   \def\gmd@storesearg{%
+%     \edef\gmd@SeargantCat{\the\catcode`\<}%
+%     \ifnum\gmd@SeargantCat=\active
+%       \let\gmd@SeargantMng=<%
+%     \fi}
+
+%   % \CodeDefIndex\gmd@restoresearg
+%   \def\gmd@restoresearg{%
+%     \catcode`\<=\gmd@SeargantCat\relax
+%     \ifnum\gmd@SeargantCat=\active
+%       \let<=\gmd@SeargantMng
+%     \fi}%
+%   % \stanza
+% }% of |\active| |<|'s |\firstofone|
+%
+% \fi ^^A~of the obsoletion iffalse
+% \QueerEOL
+
+% \CodeDefine\gmd@docstripdirective
+{\@makeother\<\@makeother\>
+  \glet\sgtleftxii=<
+
+\gdef\gmd@docstripdirective{%
+  \begingroup\let\do=\@makeother
+  \do\*\do\/\do\+\do\-\do\,\do\&\do\|\do\!\do\(\do\)\do\>\do\<% ^^Ayes, we
+  % ^^A make \nlpercent|<|\catother\ since a~directive |<<|\<any text till ^^B
+  % ^^A~the end of line> theoretically may occure.
+  \@ifnextchar{<}{%
+    \let\do=\@makeother \dospecials
+    \gmd@docstripverb}
+  {\gmd@docstripinner}}%
+
+\if1 1%^^A~as the other debugs, let them rest in peace alt least till
+%^^A the v2.0.
+\gdef\Debug@dstron{\ifnum\c@codelinenum>1178 \ifdtraceon\fi
+  \ifnum\c@codelinenum>1184 \ifdtraceoff\@emptify\Debug@dstron\fi}%
+\else
+  \global\@emptify\Debug@dstron
+\fi
+
+  % \CodeDefIndex\gmd@docstripinner
+  \gdef\gmd@docstripinner#1>{%
+    \endgroup
+    \def\gmd@modulehashone{%
+      \Module{#1}\space
+      \@afternarrgfalse\@aftercodegtrue\@codeskipputgfalse}%
+    % ^^A    \gmd@docstripshook
+    \gmd@textEOL\gmd@modulehashone\Debug@dstron}
+  % 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.
+  % \CodeDefIndex\gmd@docstripverb
+  \catcode`\^^M=\active%
+  \gdef\gmd@docstripverb<#1^^M{%
+    \endgroup%
+    \def\gmd@modulehashone{%
+      \ModuleVerb{#1}\@afternarrgfalse\@aftercodegtrue%
+      \@codeskipputgfalse}%
+    \gmd@docstripshook%
+    \gmd@textEOL\gmd@modulehashone^^M}%
+}
+
+% (\*Verbatim ;-) from \docfm:)
+% \CodeDefine\Module \CodeDefine\ModuleVerb
+\providecommand*\Module[1]{{\mod@math@codes$\langle\mathsf{#1}\rangle$}}
+
+\providecommand*\ModuleVerb[1]{{\mod@math@codes$\langle\langle\mathsf{#1}$}}
+
+% \CodeDefIndex\mod@math@codes
+\def\mod@math@codes{\mathcode`\|="226A \mathcode`\&="2026 }
+
+% \subdivision{The Changes History}
+% 
+% The contents of this section was copied \*verbatim from the
+% \docfm's documentation, with only smallest necessary changes. Then ^^A(( 
+% my additions were added :-))\,.
+%
+% \begin{quotation}To provide a change history log, the |\changes|
+% command has been
+%   introduced.  This takes [one optional and] three [mandatory]
+%   arguments, respectively, [the macro that'll become the entry's
+%   second level,] the version 
+%   number of the file, the date of the change, and some detail
+%   regarding what change has been made [i.e., the description of the
+%   change]. The [second] of these arguments 
+%   is otherwise ignored, but the others are written out and may be
+%   used to generate a history of changes, to be printed at the end of
+%   the document. [\dots\ I~ommit an obsolete remark about then-older
+%   MakeIndex's versions.]
+% 
+%   The output of the |\changes| command goes into the
+%   \<Glossary_File> and therefore uses the normal |\glossaryentry|
+%   commands. Thus MakeIndex or a~similar program can be used to
+%   process the output into a sorted ``glossary''. The |\changes|
+%   command commences by taking the usual measures to hide its
+%   spacing, and then redefines |\protect| for use within the argument
+%   of the generated |\indexentry| command.  We re-code nearly all
+%   chars found in |\@sanitize| to letter since the use of special
+%   package which make some characters active might upset the
+%   |\changes| command when writing its entries to the file. However
+%   we have to leave |%| as
+%   comment and | | as \<space> otherwise chaos will happen. And, of
+%   course the |\| should be available as escape character.\end{quotation}
+%
+% We put the definition inside a~macro that will be 
+% executed by (the first use of) |\RecordChanges|. And we provide
+% the default definition of |\changes| as a~macro just gobbling its
+% arguments. We do this to provide no changes' writing out if
+% |\RecordChanges| is not used.
+%
+% \CodeDefine\gmd@DefineChanges
+% \CodeDefine\changes
+% \changes{v0.98e}{06/09/23}{definition put into a~macro that shall
+% launch it when \cs{RecordChanges} is executed and the default
+% defining it as a~gobbling macro}
+%
+\def\gmd@DefineChanges{%
+  \outer\long\def\changes{\@bsphack\begingroup\@sanitize
+    \catcode`\\\z@ \catcode`\ 10 \MakePercentIgnore
+    \MakePrivateLetters \StraightEOL
+    \MakeGlossaryControls
+    \changes@}}
+
+\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}%
+  \renewcommand\changes[4][]{}}
+
+\def\MakeGlossaryControls{%
+  \def\actualchar{=}\def\quotechar{!}%
+  \def\levelchar{>}\edef\encapchar{\twelveclub}}%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.
+
+% \CodeDefine\changes@
+% \changes[\changes]{v0.97}{06/09/04}{The optional pseudo-argument
+%   added intended to be the macro-entry.}
+\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 |\@tempb| scratch macro.
+        \@tempswatrue
+        \edef\@tempb{% it's a~bug fix: while typesetting traditional
+          % \file{.dtx}es, |\last@defmark| came out with |\| at the
+          % beginning (which resulted with |\\|\<name> in the change
+          % log) but while typesetting the `new' way, it occured 
+          % without the bslash. So we gobble the bslash if it's
+          % present and two lines below we handle the exception of
+          % |\last@defmark|\equals|{\}| (what would happen if 
+          % a~definition of |\\| was marked in new way \gmdoc ing).%^^A]
+          \if\bslash\last@defmark\else\last@defmark\fi}%
+        \ifx\last@defmark\bslash\let\@tempb\last@defmark\fi%
+%^^A        \typeout{last dm @tempb: \@tempb}% (debugging)
+        % ^^A~\expandafter\quote@mname\macro@pname\relax
+      \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\@tempb{\string#1}}
+      % ^^A~\expandafter\quote@mname\@tempb\relax
+    \fi
+    \protected@edef\@tempa{\noexpand\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 preceed
+        % it by |\generalname|.
+          \@tempb
+          \actualchar\bslash verb*%
+          \if\verbatimchar\@tempb$\else\verbatimchar\fi
+          \quotechar\bslash \@tempb
+          \if\verbatimchar\@tempb$\else\verbatimchar\fi       
+        \else
+          \space\actualchar\generalname
+        \fi
+        :\levelchar#4\encapchar hyperpage}}%
+    \@tempa
+  \fi\endgroup\@esphack}
+
+% Let's initialize |\last@defmark| and |\GeneralName|.
+\@relaxen\last@defmark
+\@emptify\GeneralName
+% Let's explain |\if@RecentChange|. We wish to check whether the
+% change's date is later than date declared (if any limit date
+% \emph{was} declared). First of all, let's establish a~counter to store
+% the declared date. The untouched counters are equal 0 so if no date
+% is declared there'll be no problem. The date will have the
+% \<YYYYMMDD> shape both to be easily compared and readable.
+% \CodeDefine\c@ChangesStartDate
+\newcount\c@ChangesStartDate
+
+% \CodeDefine\if@RecentChange
+\def\if@RecentChange#1{%
+  \gmd@setChDate#1\@@nil\@tempcnta
+  \ifnum\@tempcnta>\c@ChangesStartDate}
+
+% \CodeDefIndex\gmd@setChDate
+\def\gmd@setChDate#1/#2/#3\@@nil#4{% the last parameter will be a~|\count|
+  % register.
+  #4=#1\relax
+  \multiply#4 by\@M
+  \count8=#2\relax% I~know it's a~bit messy not to check whether the
+  % |#4| |\count| is |\count8| but I~know this macro will only be used
+  % with |\count0 | \nlpercent(\cs{@\+te\+m\+p\+cn\+ta}) and some
+  % higher (not a~scratch) one.
+  \multiply\count8 by100 %
+  \advance#4 by\count8 \count8=\z@
+  \advance#4 by#3\relax}
+
+% Having the test defined, let's define the command setting the date
+% counter. |#1| is to be the version and |#2| the date
+% \cs[]{\{\<year>/\<month>/\<day>\}}. 
+% \CodeDefine\ChangesStart
+% \changes{v0.98e}{06/09/23}{\cs{string}s added before the seargants
+% in the message with account of \cs{docstrip(s)}}
+\def\ChangesStart#1#2{%
+  \gmd@setChDate#2\@@nil\c@ChangesStartDate
+  \typeout{^^JPackage gmdoc info: ^^JChanges' start date #1 memorized
+    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\@tempa{%
+      \noexpand\g@addto@macro\noexpand\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.}}%
+    \@tempa
+  \fi}
+% (Explanation to line \ref{TeXVer0}.)
+% \gmhypertarget[personalchanges]{My} \TeX\ Guru has remarked that the
+% change history tool should be used for documenting the changes that
+% may be significant for the users not only for the author and talking
+% of what may be significant to the user, no changes should be hidden
+% since the first published version. However, the changes' start date
+% may be used to provide hiding the author's `personal' notes: \heshe\
+% should only date the `public' changes with the four digit year and
+% the `personal' ones with two digit year and set
+% |\ChangesStart{}{1000/0/0}| or so.
+%
+% In line \ref{TeXVer0} I~establish a~test value that corresponds to
+% a~date earlier than any \TeX\ stuff and is not too small (early) to
+% ensure that hiding the two digit year changes shall not be mentioned
+% in the changes prologue.
+% 
+% \changes{v0.98d}{2006/9/15}{An entry to show the change history
+% works: watch and admire. Some sixty \cs{changes} entries irrelevant
+% for the users-other-than-myself are hidden due to the trick
+% described on
+% p.\protect\,\protect\pageref{personalchanges}.%
+% \protect\StoreMacro\protect\hyperpage 
+% \protect\def\protect\hyperpage#1{\protect\RestoreMacro\protect\hyperpage}%
+% \protect\gobble}
+% \stanza
+% \begin{quotation}The entries [of a~given version number] are sorted
+%   for convenience by the name of [the macro explicitly specified as
+%   the first argument or] the most recently introduced macroname
+%   (i.e., that in the most recent |\begin{macro}| command [or
+%     |\CodeDefine|]). We therefore provide [|\last@defmark|] to
+%     record that argument, and provide a default definition in case
+%     |\changes| is used outside a \env{macro} environment. (This is a
+%     wicked hack to get such entries at the beginning of the sorted
+%     list! It works providing no macro names start with |!| or |"|.)
+% 
+% This macro holds the string placed before changes entries on
+% top-level.\end{quotation} 
+% \CodeDefine\generalname
+\def\generalname{General}
+
+% \CodeDefine\chschange
+% \changes{v0.98a}{06/09/06}{added.}
+%
+% \stanza
+% \begin{quotation}To cause the changes to be written (to a \pk{.glo})
+% file, we define |\RecordChanges| 
+%   to invoke \LaTeX's usual |\makeglossary| command.\end{quotation}
+% 
+% I~add to it also the |\write|ing definition of the |\changes| macro
+% to ensure no changes are written out without |\RecordChanges|.
+%
+% \CodeDefine\RecordChanges
+% \changes{v0.98c}{06/09/08}{made self-\cs{relax}ing}
+% \changes{v0.98e}{06/09/23}{made it defining \cs{changes}}
+\def\RecordChanges{\makeglossary\gmd@DefineChanges
+  \@relaxen\RecordChanges}
+% \begin{quotation}The remaining macros are all analogues of those used
+%   for the \env{theindex} environment. 
+% When the glossary is started we compute the space which remains at
+% the bottom of the current page; if this is greater than |\GlossaryMin| then the
+% first part of the glossary will be placed in the available space. The number of
+% columns set [is] controlled by the counter |\c@GlossaryColumns| which can be
+% changed with a |\setcounter| declaration.\end{quotation}
+% \begin{CodeSpacesBlank}
+
+\newdimen\GlossaryMin         \GlossaryMin       = 80pt
+% \CodeDefIndex*\c@GlossaryColumns
+\newcount\c@GlossaryColumns   \c@GlossaryColumns = 2
+% \end{CodeSpacesBlank}
+%\begin{quotation}The environment \env{theglossary} is defined in the
+%same manner as the \env{theindex} 
+% environment.\end{quotation}
+%
+% \CodeDefine*{theglossary}
+% \CodeDefIndex*\theglossary
+\newenvironment{theglossary}{%
+  \begin{multicols}\c@GlossaryColumns
+    [\glossary@prologue][\GlossaryMin]%
+    \GlossaryParms \let\item\@idxitem \ignorespaces}%
+  {\end{multicols}}
+
+% Here is the MakeIndex style definition:
+% \index{gmglo.ist@\file{gmglo.ist}}
+%^^A \begin{docstrips}
+%</package>
+%<+gmglo>preamble
+%<+gmglo>"\n \\begin{theglossary} \n
+%<+gmglo>\\makeatletter\n"
+%<+gmglo>postamble
+%<+gmglo>"\n\n \\end{theglossary}\n"
+%<+gmglo>keyword "\\glossaryentry"
+%<+gmglo>actual '='
+%<+gmglo>quote '!'
+%<+gmglo>level '>'
+%<*package>
+%^^A \end{docstrips}
+% The MakeIndex shell command for the glossary should look as
+% follows:
+% \[\hbox{|makeindex -r -s gmglo.ist -o |\<myfile>|.gls |\<myfile>|.glo|}\]
+% ^^B
+% ^^B makeindex -r -s gmglo.ist -o <myfile>.gls <myfile>.glo 
+% ^^B
+% where |-r| commands MakeIndex not to make implicit page ranges,
+% |-s| commands MakeIndex to use the style stated next not the
+% default settings and the |-o| option with the subsequent filename
+% defines the name of the output.
+%
+% \begin{quotation}The |\GlossaryPrologue| macro is used to place a
+% short message above the 
+% glossary into the document. It is implemented by redefining |\glossary@prologue|,
+% a macro which holds the default text. We better make it a long macro to allow
+% |\par| commands in its argument.\end{quotation}
+%
+% \CodeDefine\GlossaryPrologue
+\long\def\GlossaryPrologue#1{\@bsphack
+  \def\glossary@prologue{#1}%
+  \@esphack}
+% \begin{quotation}Now we test whether the default is already defined
+% by another package file. If not we define it.\end{quotation}
+% \changes{v0.98a}{06/09/06}{a~bug fixed: \cs{filediv} replaced with
+% \cs{indexdiv} in the default prologue.}
+% \CodeDefine\glossary@prologue
+\@ifundefined{glossary@prologue}
+  {\def\glossary@prologue{\indexdiv{{Change History}}%
+      \markboth{{Change History}}{{Change History}}%
+    }}{}
+%\begin{quotation}Unless the user specifies otherwise, we set the
+%change history using the same parameters as for the index.\end{quotation}
+% \CodeDefine\GlossaryParms
+\AtBeginDocument{%
+  \@ifundefined{GlossaryParms}{\let\GlossaryParms\IndexParms}{}}
+
+%\begin{quotation}
+%  To read in and print the sorted change history, just put the
+%  |\PrintChanges| command as the last (commented-out, and thus
+%  executed during the documentation pass through the file) command in
+%  your package file. Alternatively, this command may form one of the
+%  arguments of the |\StopEventually| command, although a change history
+%  is probably not required if only the description is being printed.
+%  The command assumes that MakeIndex or some other program has
+%  processed the \pk{.glo} file to generate a sorted \pk{.gls}
+%  file.\end{quotation} \CodeDefine\PrintChanges
+\def\PrintChanges{% to avoid a~disaster among queer EOLs:
+  \@ifQueerEOL
+    {\StraightEOL\@input@{\jobname.gls}\QueerEOL}%
+    {\@input@{\jobname.gls}}%
+  \g@emptify\PrintChanges}
+
+% % \iffalse
+% % \begin{quotation}
+% %   The following definitions provide the |\pfill| command; if this is
+% %   specified in the index style file to MakeIndex as the delimiter to
+% %   appear after index items, then the intervening space before the
+% %   referenced page numbers will be filled with dots, with a little
+% %   white space interpolated at each end of the dots. If the line is
+% %   broken the dots will show up on both lines.\end{quotation}
+% \def\dotfill{\leaders\hbox to.6em{\hss .\hss}\hskip\z@ plus 1fill}%
+% \def\dotfil{\leaders\hbox to.6em{\hss .\hss}\hfil}%
+% \def\pfill{\unskip~\dotfill\penalty500\strut\nobreak
+%   \dotfil~\ignorespaces}%
+%%
+%%\begin{verbatim}
+%%</package>
+%%<+ist>delim_0 "\\pfill "
+%%<+ist>delim_1 "\\pfill "2
+%%<+ist>delim_2 "\\pfill "
+%%<*package>
+%\end{verbatim}
+% % skipped definitions of dotted fills\fi
+%
+%
+% \subdivision{The Checksum}
+%
+%
+% \docfm\ provides a~checksum mechanism that counts the backslashes 
+% in the scanned code. Let's do almost the same.
+%
+% At the beginning of the source file you may put the |\CheckSum|
+% macro with a~number (in one of \TeX's formats) as its argument and
+% \TeX\ with \pk{gmdoc} shall count the number of the \emph{escape ^^B
+%   chars} in the source file and tell you in the \file{.log} file (and
+%   on the terminal)
+% whether you have typed the right number. If you don't type |\CheckSum|,
+% \TeX\ anyway will tell you how much it is.
+% \CodeDefine\check@sum \CodeDefine\CheckSum 
+\newcount\check@sum
+\def\CheckSum#1{\@bsphack\global\check@sum#1\relax\@esphack}
+% \CodeDefine*{CheckSum} \CodeDefIndex*\c@CheckSum
+\newcounter{CheckSum}
+% \label{checksum}
+% \CodeDefIndex\step@checksum
+\newcommand*\step@checksum{\stepcounter{CheckSum}}
+% And we'll use it in the line \ref{checksumUse} (|\stepcounter| is
+% |\global|). See also the |\chschange| declaration,
+% l.\,\ref{chschange}.  
+%
+% However, the check sum mechanism in \gmdoc\ behaves slightly
+% different than in \docfm\ which
+% is nicely visible while \pk{gmdoc}ing \docfm: \docfm\ states its
+% check sum to be 2171 and our count counts 2126. The mystery lies in
+% the fact that \docfm's CheckSum mechanism counts the code's
+% backslashes no matter what they mean and the \gmdoc's the escape
+% chars so, among others, |\\| at the default settings increases \docfm's
+% CheckSum by 2 while the \gmdoc's by 1. (There are 38 occurrences of |\\|
+% in \pk{doc.dtx} \env{macrocode}s, I~counted
+% myself.)\begin{StraightEOL}\footnote{My opinion is that nowadays
+%     a~check sum 
+%     is not necessary for checking the completness of a~file but
+%     I~like it as a~marker of file development and this more than
+%     that is its r\^ole in \gmdoc.}
+% \end{StraightEOL}
+%
+%
+% \begin{quotation}But |\Finale| will be called at the very end of a
+%   file. This is exactly the point were we want to know if the file
+%   is uncorrupted. Therefore we also call |\check@checksum| at this
+%   point.\end{quotation}
+% In \pk{gmdoc} we have the |\AtEndInput| hook.
+\AtEndInput{\check@checksum}
+
+% Based on the lines 723--741 of \pk{doc.dtx}.
+% \CodeDefine\check@checksum
+\def\check@checksum{\relax
+  \ifnum\check@sum=\z@
+    \typeout{**********************************}%
+    \typeout{* The input file \gmd@inputname\space has no Checksum
+      stated.}%
+    \typeout{* The current checksum is \the\c@CheckSum.}%
+    \gmd@chschangeline% a~check sum changes history entry, see below.
+    \typeout{* (package gmdoc info.)}
+    \typeout{**********************************}%
+  \else
+    \ifnum\check@sum=\c@CheckSum
+      \typeout{*****+*+*+*+*+*+*+*+*+*+}%
+      \typeout{* The input file \gmd@inputname: Checksum passed.}%
+      \gmd@chschangeline
+      \typeout{* (package gmdoc info.)}
+      \typeout{*****+*+*+*+*+*+*+*+*+*+}%
+    \else
+      \typeout{********!*!*!*!*!*!*!*!*!*!*!*!}%
+      \typeout{*! The input file \gmd@inputname:}
+      \typeout{*! The CheckSum stated: \the\check@sum\space<> my
+        count: \the\c@CheckSum.} 
+      \gmd@chschangeline
+      \typeout{*! (package gmdoc info.)}
+      \typeout{********!*!*!*!*!*!*!*!*!*!*!*!}%
+    \fi
+  \fi
+  \global\check@sum\z@}
+
+% As I~mentioned above, I~use the check sum mechanism to mark the file
+% growth. Therefore I~provide a~macro that produces a~line on the
+% terminal to be put somewhere at the
+% beginning of the source file's commentary for instance.
+% \CodeDefIndex\gmd@chschangeline
+\def\gmd@chschangeline{%
+  \typeout{\twelvepercent\space\string\chschange{%
+      \fileversion}{\the\year/\the\month/\the\day}{\the\c@CheckSum}}
+  \typeout{\twelvepercent\space\string\chschange{\fileversion}{%
+      \expandafter\@gobbletwo\the\year/\the\month/\the\day}{%\relax
+      %  with two di\-g\-it year in case you use |\ChangesStart|.
+      \the\c@CheckSum}}}
+
+% And here the meaning of such a~line is defined:
+\newcommand*\chschange[3]{%\label{chschange}
+  \csname changes\endcsname{#1}{#2}{CheckSum #3}% |\csname...| 'cause
+  % \nlpercent\cs{cha\+n\+g\+es} is \nlpercent|\outer|.
+  \CheckSum{#3}}
+% It will make a~`General' entry in the change history unless
+% used in some |\CodeDefine|'s scope or inside a~\env{macro}
+% environment. It's intended to be put 
+% somewhere at the beginning of the documented file.
+%
+%
+% \changes[\CheckSum]{v0.98a}{06/09/06}{`package gmdoc info'
+% statement moved to the end of the checksum messages.}
+% \changes[\CheckSum]{v0.98b}{06/09/07}{typing out the
+% \cs{chschange} line added}
+
+%
+%
+% \subdivision{Macros from \pk{ltxdoc}}
+%
+% I'm not sure whether this package still remains `minimal' but
+% I~liked the macros provided by \pk{ltxdoc.cls} so much\dots
+% \stanza
+%
+% The next page setup declaration is intended to be used
+% with the \pk{article}'s default Letter paper size. But since 
+% \CodeDefine\ltxPageLayout
+\newcommand*\ltxPageLayout{%
+%\begin{quotation}Increase the text width slightly so that width the standard fonts
+% 72 columns of code may appear in a |macrocode| environment.\end{quotation}
+% 
+  \setlength{\textwidth}{355pt}%
+% \begin{quotation}Increase the marginpar width slightly, for long
+% command names. And increase the left margin by a similar
+% amount.\end{quotation}
+% To make these settings independent from the defaults (changed e.g.\
+% in \pk{gmdocc.cls}) we replace the original |\addtolength|s with
+% |\setlength|s.
+%^^A  \addtolength\marginparwidth{30pt}%
+%^^A  \addtolength\oddsidemargin{20pt}% 
+%^^A  \addtolength\evensidemargin{20pt}
+ \setlength\marginparwidth{95pt}%
+ \setlength\oddsidemargin{82pt}%
+ \setlength\evensidemargin{82pt}}
+
+% %\iffalse
+% \def\task#1#2{}
+% % What is this for? Probably to turn an obsolete command off. Or maybe
+% % a~TODO?
+% % \fi
+
+%\subdivision{\cs{DocInclude} and the \pk{ltxdoc}-Like Setup}
+%
+% Let's provide a~command for including
+% multiple files into one document. In the \pk{ltxdoc} class such
+% a~command is defined to include files as parts. But we prefer to
+% include them as chapters in the classes that provide
+% |\chapter|. We'll redefine |\maketitle| so that it make a~chapter or
+% a~part heading \emph{unlike} in \pk{ltxdoc} where the file parts
+% have their titlepages with only the filename and \pk{article}-like titles
+% made by |\maketitle|.
+%
+% But we will also provide a~possibility of typesetting multiple files
+% exactly like with the \pk{ltxdoc} class.
+%
+% \begin{macro}{\DocInclude}
+%
+%   So, define the |\DocInclude| command, that acts
+%   \begin{quotation}more or less exactly the same as |\include|, but
+%     uses |\DocInput| on a \pk{dtx} [or \pk{.fdd}] file, not |\input| on
+%     a \pk{tex} file.\end{quotation}
+%   Our version will accept also \pk{.sty}, \pk{.cls}, and \pk{.tex}
+%   files.
+%
+% \changes{v0.98}{06/09/05}{\cs{@makeother\protect\bslash_} added}
+\newcommand*\DocInclude{\bgroup\@makeother\_\Doc@Include}% First, we
+% make \nlpercent|_| `other' in order to allow it in the filenames.
+
+% \CodeDefine\Doc@Include
+\newcommand*{\Doc@Include}[2][]{% originally it took just one
+  % argument. Here we make it take two, first of which is intended to
+  % be the path (with the closing \nlpercent |/|). This is intended not to print
+  % the path in the page footers only the filename.\par
+  % \CodeUsage\HLPrefix \CodeDefine\EntryPrefix
+  \egroup% having the arguments read, we close the group opened by the
+  % previous macro for |_|\catother.
+  \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 \nlpercent|\IndexPrologue|'s
+  % |\if|s. Therefore the 
+  % definition is global. The latter is such for symmetry.
+  % \CodeDefine\GeneralName
+  \def\GeneralName{#2\actualchar\pk{#2} }% \label{GeneralName} for the
+    % changes'history main level entry.
+  \relax
+  %^^A\if@ltxDocInclude
+  \clearpage
+  % ^^A\fi
+  \docincludeaux
+  \def\currentfile{gmdoc-IncludeFileNotFound.000}%
+  \let\fullcurrentfile\currentfile
+  \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.}}}}}}}%
+  % \changes{v0.98j}{06/10/16}{\pk{.fdd} added in \cs{DocInclude}'s
+  % search for the extension.}
+  \edef\fullcurrentfile{#1\currentfile}%
+  \ifnum\@auxout=\@partaux
+    \@latexerr{\string\DocInclude\space cannot be nested}\@eha
+  \else \@docinclude{#1}#2 \fi}% Why is |#2| delimited with | | not
+% braced as we are used to, one may ask. 
+
+% \CodeDefine\@docinclude
+% \changes[\DocInclude]{v0.97}{06/09/03}{\cs{@docinclude} gets 2 params
+%   in order to \cs{includeonly} work as expected, with only the names
+%   (w.o. the paths)}
+\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.aux}\fi% this strange macro with
+  % a~long name is another thing to allow |_| in the filenames (see
+  % line \ref{gmd@writemauxinpaux}).
+  \@tempswatrue
+  \if@partsw \@tempswafalse\edef\@tempb{#2}%
+    \@for \@tempa:=\@partlist\do{\ifx\@tempa\@tempb\@tempswatrue\fi}%
+  \fi
+  \if@tempswa \let\@auxout\@partaux 
+    \if@filesw
+      \immediate\openout\@partaux #2.aux\relax% Yes, only |#2|. It's to
+      % create and process the partial \pk{.aux} 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\@nameuse{cp@#1#2}%
+  \fi
+  \let\@auxout\@mainaux}% end of |\@docinclude|.
+% \end{macro}
+%
+% (Two is a~sufficient number of iterations to define a~macro for.)
+% \CodeDefine\xdef@filekey
+\def\xdef@filekey{{\@relaxen\ttfamily% \label{LetTTFRelax}This
+    % assignment is very trickly crafted: it makes \emph{all}
+    % |\ttfamily|s present in the |\filekey|'s expansion unexpandable
+    % not only the one added in this step.
+    \xdef\filekey{\filekey, \thefilediv={\ttfamily\currentfile}}}}
+
+
+% To allow |_| in the filenames we must assure |_| will be \catother\
+% while reading the filename. Therefore define
+% \CodeDefIndex\gmd@writemauxinpaux
+\def\gmd@writemauxinpaux#1{% \label{gmd@writemauxinpaux}
+  % this name comes from `\emph{write} outto \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 localized 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{%
+    \bgroup\string\@makeother\string\_%
+    \string\firstofone{\egroup
+    \string\@input{#1}}}}
+
+% We also slightly modify a~\LaTeX\ kernel macro |\@writeckpt| to
+% allow |_| in the file name.
+% \changes[\DocInclude]{v0.98a}{06/09/06}{\cs{@writeckpt} wrapped
+% with \cs{firstofone} hack to allow \cs[]{_} in the file names.}
+% \CodeDefIndex\gmd@writeckpt
+\def\gmd@writeckpt#1{%
+  \immediate\write\@partaux{%
+    \string\bgroup\string\@makeother\string\_%
+    \string\firstofone\@charlb\string\egroup}
+  \@writeckpt{#1}%
+  \immediate\write\@partaux{\@charrb}}
+
+%\CodeDefine\gmd@doIndexRelated
+\def\gmd@doIndexRelated{%
+  \do\tableofcontents \do\makeindex \do\EnableCrossrefs
+  \do\PrintIndex \do\printindex \do\RecordChanges \do\PrintChanges
+  \do\theglossary \do\endtheglossary}
+
+
+\@emptify\filesep
+% The \pk{ltxdoc} class establishes a~special number format for
+% multiple file documentation numbering needed to document the \LaTeX\
+% sources. I~like it too, so\CodeDefine\aalph \CodeDefine\@aalph
+\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}
+
+% A~macro that initialises things for |\DocInclude|.
+% \CodeDefine\docincludeaux
+\def\docincludeaux{%
+  % We set the things for including the files only once.
+  \global\@relaxen\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
+  \@addtoreset{codelinenum}{\filedivname}% remember it has
+  % a~|\global| effect in fact. For each file we'll reset
+  % \env{codelinenum}.
+  \def\thefilediv{\aalph{\filedivname}}% The files will be numbered
+  % with letters, lowercase first.
+  \expandafter\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.
+    \expandafter\ifx\csname ver@\currentfile\endcsname\relax
+      File \thefilediv: {\ttfamily\currentfile} %
+    \else
+      \GetFileInfo{\currentfile}%
+      File \thefilediv: {\ttfamily\filename} %
+      Date: \filedate\ %
+      Version \fileversion
+    \fi
+    \hfill\thepage}%
+  \glet\@evenfoot\@oddfoot% see line \ref{oneside}.
+  % ^^A  \show\filedivname% for debug
+  \expandafter\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% If |\filediv| is higher than
+    % |\section| we don't change the two divisions (they are
+    % |\section| and |\subsection| 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 |\docincludeaux|.
+
+% The |\filediv| and |\filedivname| macros should always be set
+% together. Therefore provide a~macro that takes care of both at
+% once. Its |#1| should be a~sectioning name without the backslash.
+% \CodeDefine\SetFileDiv
+% \changes{v0.98c}{06/9/10}{added and a~bug fixed in \cs{docincludeaux}}
+\def\SetFileDiv#1{%
+  \edef\filedivname{#1}%
+  \expandafter\let\expandafter\filediv\csname#1\endcsname}
+
+% \subdivision{\cs{SelfInclude}}
+%
+% I~needed to include the driver file into
+% a~documentation so I~wrote
+% a~macro in case I'll need it again or you'll need it.
+% We define it immediately i.e.,
+% without the |\catcode|s trick because it uses |\jobname| inside so
+% the filename will be all \catother\ anyway.
+% \CodeDefine\SelfInclude
+% \changes{v0.98a}{06/09/06}{added in response to the needs}
+\newcommand*{\SelfInclude}[2][.tex]{% As you guess, the optional |#1|
+  % is the jobname's extension. The second parameter is \emph{not} for
+  % the filename (note \emph{it's known:} as |\jobname|!), but
+  % \emph{for the stuff to be put at begin input.}
+  \AtBegInputOnce{#2}%
+  \gdef\HLPrefix{\filesep}%
+  \gdef\EntryPrefix{\filesep}% we define two rather kernel parameters
+  % etc.\ as in \nlpercent|\DocInclude|.
+  \relax
+  %^^A\if@ltxDocInclude
+  \clearpage
+  % ^^A\fi
+  \docincludeaux
+  \edef\currentfile{\jobname#1}%
+  \let\fullcurrentfile\currentfile
+  \def\GeneralName{\jobname\actualchar\pk{\jobname} }% for the
+  % changes history main level entry.
+  \ifnum\@auxout=\@partaux
+    \@latexerr{\string\DocInclude\space cannot be nested}\@eha
+  \else
+    \if@filesw
+      \gmd@writemauxinpaux{\jobname.auxx}% ^^A
+      % this queer macro allows |_| in the file names. In this
+      % particular case |\string\jobname| would do, but anyway we
+      % provide a~more general solution. Note the \pk{.auxx} extension
+      % used instead of \pk{.aux}. This is done to avoid an infinite
+      % recurrence of \nlpercent|\input|s.
+    \fi
+    \@tempswatrue
+    \if@partsw \@tempswafalse\edef\@tempb{\jobname}%
+      \@for
+      \@tempa:=\@partlist\do{\ifx\@tempa\@tempb\@tempswatrue\fi}%
+    \fi
+    \if@tempswa \let\@auxout\@partaux
+      \if@filesw
+        \immediate\openout\@partaux \jobname.auxx\relax
+        \immediate\write\@partaux{\relax}
+      \fi
+      % \begin{quotation}We need to save (and later
+      %   restore)\dots\end{quotation} 
+      \StoringAndRelaxingDo% provided by \pk{gmutils}
+      \gmd@doIndexRelated
+      \if@ltxDocInclude\part{\currentfile}%
+      \else\let\maketitle=\InclMaketitle
+      \fi
+      \if@ltxDocInclude\xdef@filekey\fi
+      \GetFileInfo{\currentfile}% it's my (GM) addition with the account
+      % of using file info in the included files' title etc.
+      \incl@DocInput{\fullcurrentfile}% originally just |\currentfile|, no
+      % difference in |\SelfInclude|.
+      \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 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% among others, causes the |\writes| to be executed which is
+      % crucial for proper toc-ing e.g.
+      %^^A\fi
+      \gmd@writeckpt{\jobname.x}% note the |.x| in the checkpoint
+      % used to distinguish this instance (input) of the driver file
+      % from its main instance.
+      \if@filesw \immediate\closeout\@partaux \fi
+    \else\@nameuse{cp@\jobname.x}% note |.x|: it's used for the same
+    % reason as above. 
+    \fi
+    \let\@auxout\@mainaux
+  \fi}% end of |\SelfInclude|.
+
+% The \pk{ltxdoc} class makes some preparations for inputting multiple
+% files. We are not sure if the user wishes to use \pk{ltxdoc}-like
+% way of documenting (maybe \heshe\ will prefer what I~offer,
+% \pk{gmdocc.cls} e.g.), so we put
+% those preparations into a~declaration.
+%
+% \CodeDefine\ltxLookSetup
+% \CodeDefIndex\if@ltxDocInclude \CodeDefIndex*\@ltxDocIncludetrue
+% \CodeDefIndex*\@ltxDocIncludefalse
+\newif\if@ltxDocInclude
+
+\newcommand*\ltxLookSetup{%
+  \SetFileDiv{part}%
+  \ltxPageLayout
+  \@ltxDocIncludetrue
+}
+
+\@onlypreamble\ltxLookSetup
+
+% The default is that we |\DocInclude| the files due to the original
+% \gmdoc\ input settings.
+\let\incl@DocInput=\DocInput
+
+\@emptify\currentfile% for the pages outside the
+% |\DocInclude|'s scope. In force for all includes.
+%
+% \iffalse
+% % \stanza
+% % But it may be not a~usual situation to include the source files
+% % as with \pk{ltxdoc} but with the \emph{new} |\DocInput| macro. So
+% % let's |\def|
+% % \CodeDefine\ltxMultidocOldSetup
+% \newcommand*\ltxMultidocOldSetup{%
+%   \ltxMultidocSetup
+%   \let\incl@DocInput=\OldDocInput}
+%
+% \@onlypreamble\ltxMultidocOldSetup
+%
+% And, for the symmetry, if you prefer the look offered by
+% me, but 
+% \fi
+
+% If you want to |\Doc/SelfInclude| \docfm-likes:
+% \CodeDefine\olddocIncludes
+\newcommand*\olddocIncludes{%
+  \let\incl@DocInput=\OldDocInput}
+
+% And, if you have set the previous and want to set it back:
+% \CodeDefine\gmdocIncludes
+\newcommand*\gmdocIncludes{%
+  \let\incl@DocInput=\DocInput
+  \AtBegInput{\QueerEOL}}% to move back the |\StraightEOL| declaration put
+% at begin input by |\olddocIncludes|.
+%
+%
+% \subdivision{Redefinition of \cs{maketitle}}
+%
+% A~not-so-slight alteration of the \TextUsage\maketitle\ command in
+% order it allow multiple titles in one document seems to me very
+% clever. So let's copy again (\pk{ltxdoc.dtx} the lines 643--656):
+%
+% \begin{quotation}The macro to generate titles is easily altered in
+% order that it can be used more than once (an article with many
+% titles). In the original, diverse macros were concealed 
+% after use with |\relax|. We must cancel anything that may have been
+% put into |\@thanks|, etc., otherwise all titles will carry forward
+% any earlier such setting!\end{quotation} 
+% But here in \pk{gmdoc} we'll do it locally for (each) input 
+% not to change the main title settings if there are any.
+% \CodeDefine\maketitle
+\AtBegInput{%
+  \providecommand*\maketitle{\par
+    \begingroup \def \thefootnote {\fnsymbol {footnote}}%
+    \setcounter {footnote}\z@
+    \def\@makefnmark{\hbox to\z@{$\m@th^{\@thefnmark}$\hss}}%
+    \long\def\@makefntext##1{\parindent 1em\noindent
+      \hbox to1.8em{\hss$\m@th^{\@thefnmark}$}##1}%
+    \if@twocolumn \twocolumn [\@maketitle ]%
+    \else \newpage \global \@topnum \z@ \@maketitle \fi
+% \begin{quotation}For special formatting requirements (such as in
+% TUGboat), we use pagestyle |titlepage| for this; this is later defined
+% to be |plain|, unless already defined, as, for example, by
+% \pk{ltugboat.sty}.\end{quotation}
+    \thispagestyle{titlepage}\@thanks \endgroup 
+%\begin{quotation}If the driver file
+% documents many files, we don't want parts of a title of one to
+% propagate to the next, so we have to cancel these:\end{quotation}
+    \setcounter {footnote}\z@
+    \gdef\@date{\today}\g@emptify\@thanks%
+    \g@emptify\@author\g@emptify\@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 pagestyle 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}.
+  % \CodeDefine\@maketitle
+  \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.)
+  % \CodeDefine\title \CodeDefine\author \CodeDefine\date
+  % \CodeDefine\thanks 
+  \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}}%
+  }%\CodeDefIndex\and
+  \providecommand*\and{%| % \begin{tabular}|
+    \end{tabular}%
+    \hskip 1em \@plus.17fil%
+    \begin{tabular}[t]{c}}%| % \end{tabular}|
+    % And finally, let's initialize \cs{tit\+le\+set\+up} if it is not yet.
+    % \CodeDefine\titlesetup
+    \providecommand*\titlesetup{}%
+}% end of |\AtBegInput|.
+%
+% \changes[\titlesetup]{v0.98c}{06/09/08}{\cs{provide}d \cs[]{\{\}}}
+%
+%
+% The \pk{ltxdoc} class redefines the |\maketitle| command to allow
+% multiple titles in one document. We'll do the same and something
+% more: our |\Doc/SelfInclude| will turn the file's |\maketitle| into
+% a~part or chapter heading. But, if hte |\ltxLookSetup| declaration
+% is in force, |\Doc/SelfInclude| will make for an included file
+% a~part's title page and an \pk{article}-like title.
+%
+% \stanza
+% Let's initialize the file division macros.
+\@relaxen\filediv
+\@relaxen\filedivname
+
+% If we don't include files the \pk{ltxdoc}-like way, we wish to
+% redefine |\maketitle| so that it typesets a~division's heading.
+%
+
+% Now, we redefine |\maketitle| and its relatives.
+%
+% \CodeDefine\InclMaketitle
+% \changes{v0.98a}{06/09/05}{renamed from \cs{incl@maketitle} 'cause
+% I~needed it in self-input hacks.}
+\def\InclMaketitle{%\par\CodeDefIndex\and
+  {\def\and{, }% we make |\and| just a~comma.
+    {\let\thanks=\@gobble% for the toc version of the heading we
+      % discard |\thanks|.\tOnLine
+      \protected@xdef\incl@titletotoc{\@title\if@fshda\protect\space
+        (\@author)\fi}% we add the author iff the `files have
+      % different authors' \nlpercent(|@fshda|)
+    }% \CodeDefine\thanks
+    \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
+        {\smallerr% this macro is provided by the \pk{gmutils}
+          % package after the \pk{relsize} package.
+          \if@fshda\\[0.15em]\protect\@author
+            \if\relax\@date\relax\else, \fi
+          \else
+            \if\relax\@date\relax\else\\[0.15em]\fi
+          \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@fshda| switch defined in line \ref{@fshda}.
+% 
+%  If we wish to print the author's name
+% (|\if@fshda|), then we'll print the date after the author, separated
+% with a~comma. If we don't print the author, there still may be
+% a~date to be printed. In such a~case we break the line, too, and
+% print the date with no comma.
+          \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.
+    \expandafter\filediv\incl@filedivtitle
+    \@thanks
+    \g@relaxen\@author \g@relaxen\@title \g@relaxen\@date
+    \g@emptify\@thanks
+}% end of |\InclMaketitle|.
+
+% What I~make the default, is an assumption that all the
+% multi-documented files have the same author(s). 
+% And with the account of the other possibility I~provide
+% the below switch and declaration.
+%
+% \CodeDefIndex\if@fshda
+% \CodeDefIndex*\@fshdatrue
+% \CodeDefIndex*\@fshdafalse
+% \CodeDefine\PrintFilesAuthors
+\newif\if@fshda
+% \label{@fshda} (its name comes from \textit file\textit s
+% \textit have \textit different \textit authors).
+
+\newcommand*\PrintFilesAuthors{\@fshdatrue}
+% And the counterpart, if you change your mind:
+\newcommand*\SkipFilesAuthors{\@fshdafalse}
+
+%
+%
+% \subdivision{Miscellanea}
+%
+% The main inputting macro, |\DocInput| has been provided. But there's
+% another one in \docfm\ and it looks very reasonably:
+% \gmhypertarget[IndexInput]{\cs{IndexInput}}. Let's make analogous one
+% here:
+% \CodeDefine\IndexInput
+% \changes{v0.98b}{06/09/07}{\cs{StoreMacro}ing and \cs{RestoreMacro}ing
+%   \cs{code@delim} added}
+\def\IndexInput#1{%
+  \StoreMacro\code@delim
+  \CodeDelim*\^^A%
+  \DocInput{#1}\RestoreMacro\code@delim}
+% How does it work? We assume in the input file is no explicit 
+% \<char1>. This char is chosen as the code delimiter and will be put
+% at the end of input. So, entire file contents will be scanned char
+% by char as the code.
+% \stanza
+%
+% \iffalse First I~tried to use \specialcomment{gmlonely} provided by
+% the \pk{comment} package, but it caused errors.\fi
+%
+% The below environment I~designed to be able to skip some repeating
+% texts while documenting several packages of mine into one
+% document. At the default settings it's just a~|\StraightEOL| group
+% and in the |\skipgmlonely| declaration's scope it gobbles its
+% contents.
+%
+% \CodeDefine*{gmlonely}
+% \CodeDefine\skipgmlonely
+\newenvironment{gmlonely}{\StraightEOL}{}
+
+\newcommand\skipgmlonely[1][]{%
+  \def\@tempa{%\CodeDefIndex\gmd@skipgmltext
+    \def\gmd@skipgmltext{\g@emptify\gmd@skipgmltext#1}}%
+  \@tempa
+  \expandafter\AtBegInput\expandafter{\@tempa}%
+  \renewenvironment{gmlonely}{%
+    \StraightEOL
+    \@fileswfalse% to forbid writing to \pk{.toc}, \pk{.idx} etc.
+    \setbox0=\vbox\bgroup}{\egroup\gmd@skipgmltext}}
+
+% Sometimes in the commentary of this package, so maybe also others,
+% I~need to say some char is of category 12 (`other sign'). This I'll
+% mark just as \catother\ got by |\catother|.
+%
+% \CodeDefine\catother
+\bgroup\catcode`\_=8 % we ensure the standard |\catcode| of |_|.
+\firstofone{\egroup
+  \newcommand*\catother{${}_{12}$}%\par
+% Similarly, if we need to say some char is of category 13 (`active'),
+% we'll write \catactive, got by |\catactive|
+% \CodeDefine\catactive
+  \newcommand*\catactive{${}_{13}$}%\par
+% and a~letter, \catletter
+% \CodeDefine\catletter
+  \newcommand*\catletter{${}_{11}$}%.
+}
+%
+% For the copyright note first I~used just \env{verse} but it requires
+% marking the line ends with |\\| and indents its contents while
+% I~prefer the copyright note to be flushed left. So
+% \CodeDefine*{copyrnote}
+\newenvironment*{copyrnote}{%
+  \StraightEOL\everypar{\hangindent3em\relax\hangafter1 }%
+  \par\addvspace\medskipamount\parindent\z@\obeylines}{%
+  \@codeskipputgfalse\stanza}
+%
+% I~renew the \env{quotation} environment to make the fact of quoting
+% visible
+% \CodeDefine*{quotation}
+\renewenvironment{quotation}{\par``\ignorespaces}{\unskip''\par}
+% \chunkskip
+% For some mysterious reasons |\noindent| doesn't work with the first
+% (narrative) paragraph after the code so let's work it around:
+% \CodeDefine\gmdnoindent
+\newcommand*\gmdnoindent{\leavevmode\hskip-\parindent}
+%
+% When a~verbatim text occurres in an inline comment, it's advisable to
+% precede it with |%| if it begins a~not first line of such a~comment
+% not to mistake it for a~part of code. For this purpose provide 
+% \CodeDefine\nlpercent
+\newcommand*\nlpercent{%
+  \@ifstar{\def\@tempa{{\tt\twelvepercent}}%
+    \@emptify\@tempb\nl@percent}%
+  {\@emptify\@tempa
+    \def\@tempb{\gmboxedspace}\nl@percent}}
+
+\newcommand*\gmboxedspace{\hbox{\normalfont{ }}}
+
+\newcommand*\nl@percent[1][]{%
+  \unskip
+  \discretionary{\@tempa}{{\tt\twelvepercent\gmboxedspace}}{\@tempb}%
+  \penalty10000\hskip0sp\relax}
+% As you see, |\nlpercent| inserts a~discretionary that breaks to |%|
+% at the beginning of the lower line. Without the break it's a~space
+% (alas at its natural width i.e., not flexible) or, with the starred
+% version, nothing. The starred version puts |%| also at the end of
+% the upper line.
+%
+% TODO: make the space flexible (most probably it requires using sth.\
+% else than |\discretionary|).
+%\stanza
+%
+% An optional hyphen for CSs in the inline comment:
+% \CodeDefine\+
+\@ifundefined{+}{}{\typeout{^^Jgmdoc.sty: redefining \bslash+.}}
+\def\+{\discre{{\normalfont-}}{{\tt\twelvepercent\gmboxedspace}}{}}
+%
+%
+% And finally, some logos:
+% \begin{quotation} Here are a few definitions which can usefully be
+% employed when documenting 
+% package files: now we can readily refer to \AmSTeX, \BibTeX\ and \SliTeX, as well
+% as the usual \TeX\ and \LaTeX. There's even a \PlainTeX\ and
+% a~\Web.\end{quotation}
+% \CodeDefine\AmSTeX \CodeDefine\BibTeX \CodeDefine\SliTeX
+% \CodeDefine\PlainTeX \CodeDefine\Web \CodeDefine\TeXbook
+% \CodeDefine\eTeX \CodeDefine\pdfeTeX \CodeDefine\pdfTeX
+% \CodeDefine\XeTeX \CodeDefine\ds
+\@ifundefined{AmSTeX}
+  {\def\AmSTeX{\leavevmode\hbox{$\mathcal A\kern-.2em\lower.376ex%
+        \hbox{$\mathcal M$}\kern-.2em\mathcal S$-\TeX}}}{}
+
+\@ifundefined{BibTeX}
+  {\def\BibTeX{{\rmfamily B\kern-.05em%
+        \textsc{i\kern-.025em b}\kern-.08em%
+        \TeX}}}{}
+
+\@ifundefined{SliTeX}
+  {\def\SliTeX{{\rmfamily S\kern-.06emL\kern-.18em\raise.32ex\hbox
+        {\scshape i}\kern -.03em\TeX}}}{}
+
+\@ifundefined{PlainTeX}{\def\PlainTeX{\textsc{Plain}\kern2pt\TeX}}{}
+
+\@ifundefined{Web}{\def\Web{\textsc{Web}}}{}
+% There's also the \LaTeXpar\ logo got with the |\LaTeXpar| macro
+% provided by \pk{gmutils}. And here \TeXbook's logo:
+\def\TeXbook{The \TeX book}
+
+\@ifundefined{eTeX}{\def\eTeX{$\varepsilon$\TeX}}{}% The \eTeX\ manual
+% uses `$\varepsilon$-\TeX' but that would look strange in \pdfeTeX.
+
+\@ifundefined{pdfeTeX}{\def\pdfeTeX{pdf\eTeX}}{}
+
+\@ifundefined{pdfTeX}{\def\pdfTeX{pdf\TeX}}{}
+
+\@ifundefined{XeTeX}{\def\XeTeX{X\kern-.125em\relax
+    \@ifundefined{reflectbox}{%
+      \lower.5ex\hbox{E}\kern-.1667em\relax}{%
+      \lower.5ex\hbox{\reflectbox{E}}\kern-.1667em\relax}%
+    \TeX}}{}% As you see, if
+% \TeX\ doesn't recognize \nlpercent\cs{re\+flect\+box} (\pk{graphics}
+% isn't loaded), the first |E| will not be reversed. This version of
+% the command is intended for non-\XeTeX\ usage. With \XeTeX, the
+% reversed |E| you get as the Unicode Latin Letter Reversed E.
+
+\@ifundefined{ds}{\def\ds{DocStrip}}{}
+
+% Define |\filedate| and friends from info in the
+% |\ProvidesPackage| etc.\ commands.
+% \CodeDefine\GetFileInfo
+\def\GetFileInfo#1{%
+  \def\filename{#1}%
+  \def\@tempb##1 ##2 ##3\relax##4\relax{%
+    \def\filedate{##1}%
+    \def\fileversion{##2}%
+    \def\fileinfo{##3}}%
+  \edef\@tempa{\csname ver@#1\endcsname}%
+  \expandafter\@tempb\@tempa\relax? ? \relax\relax}
+
+% Since we may documentally input files that we don't load, as \docfm\
+% e.g., let's define a~macro for explicit providing the file
+% info. It's written in analogy to |\ProvidesFile|, \pk{source ^^B
+% 2${}_\epsilon$}, file L v1.1g, l.\,102.
+% \CodeDefine\ProvideFileInfo
+% \changes{v0.98a}{06/09/06}{added}
+\def\ProvideFileInfo#1{%
+  \begingroup
+    \catcode`\ 10 \catcode\endlinechar 10 %
+    \@makeother\/\@makeother\&%
+    \kernel@ifnextchar[{\gmd@providefii{#1}}{\gmd@providefii{#1}[]}%^^A]
+  }
+
+  % \CodeDefIndex\gmd@providefii
+\def\gmd@providefii#1[#2]{%\\
+  % (we \emph{don't} write the file info to \pk{.log})
+  \expandafter\xdef\csname ver@#1\endcsname{#2}%
+  \endgroup}
+
+% And a~self-reference abbreviation (intended for providing file info
+% for the driver):
+% \CodeDefine\ProvideSelfInfo
+\def\ProvideSelfInfo{\ProvideFileInfo{\jobname.tex}}
+
+% A~neat conventional statement used in \docfm's documentation e.g.,
+% to be put in |\thanks| to the title or in a~footnote:
+% \CodeDefine\filenote
+\newcommand*\filenote{This file has version number \fileversion{} dated \filedate{}.}
+% And exactly as  |\thanks|:
+% \CodeDefine\thfileinfo
+\newcommand*\thfileinfo{\thanks\filenote}
+
+
+% \gmhypertarget[CDAnd]{Finally,} a~couple of macros for documenting
+% files playing with |%|'s
+% catcode(s). Instead of |%| I~used |&|. They may be at the end
+% because they're used in the commented thread i.e.\ after package's
+% |\usepackage|.  \CodeDefine\CDAnd
+\newcommand*\CDAnd{\CodeDelim\&}
+% \CodeDefine\CDPerc
+\newcommand*\CDPerc{\CodeDelim\%}
+
+% And for documenting in general:
+%
+% A~general sectioning command because I~foresee a~possibility of
+% typesetting the same file once as independent document and another
+% time as a~part of bigger whole.
+\@ifdefinable\division{}% just to test (this \LaTeX\ check issues an
+% error if the first argument is already defined). \CodeDefine\division
+\@ifundefined{section}{%
+  \@relaxen\division}{% \CodeDefine\division
+  \let\division=\section}
+
+\@ifdefinable\subdivision{}% just to test (see above). \CodeDefine\subdivision
+\@ifundefined{subsection}{%
+  \@relaxen\subdivision}{% \CodeDefine\subdivision
+  \let\subdivision=\subsection}
+% The |\let|s are inside |\@ifundefined|s because I'm not sure
+% whether you will typeset a~documentation with usual classes. Maybe
+% too much care.
+%
+%
+% \stanza
+% To kill a~tiny little bug in \file{doc.dtx} (in line 3299 |\@tempb| and
+% |\@tempc| are written plain not verbatim):
+% \CodeDefIndex*{gmd@mc} \CodeDefine\AfterMacrocode
+\newcounter{gmd@mc}
+
+\def\gmd@mchook{\stepcounter{gmd@mc}% \label{gmd@mchook}
+  \gmd@mcdiag
+  \csname gmd@mchook\the\c@gmd@mc\endcsname}
+
+\long\def\AfterMacrocode#1#2{\@namedef{gmd@mchook#1}{#2}}
+% What have I~done? I~declare a~new counter and employ it to count the
+% \env{macrocode(*)}s (and \env{oldmc(*)}s too, in fact) and attach
+% a~hook to (after) the end of every such environment. That lets us to
+% put some stuff pretty far inside the compiled file (for the buggie
+% in \file{doc.dtx}, to redefine |\@tempb/c|).
+
+% One more detail to expalin and define: the |\gmd@mcdiag| macro may
+% be defined to type out a~diagnostic message (the
+% \env{macrocode(*)}'s number, code line number and input line number).
+% \CodeDefine\gmd@mcdiag \CodeDefine\mcdiagOn \CodeDefine\mcdiagOff
+\@emptify\gmd@mcdiag
+
+\def\mcdiagOn{\def\gmd@mcdiag{%
+    \typeout{^^J\bslash end{macrocode(*)} No.\the\c@gmd@mc
+    \space\on@line, cln.\the\c@codelinenum.}}}
+
+\def\mcdiagOff{\@emptify\gmd@mcdiag}
+%
+%
+% \subdivision[\docfm-Compatibility]{\gmhypertarget[doccompa]{\docfm-Compatibility}}
+%
+% \gmhypertarget[doc-likeness]{My} \TeX\ Guru recommended me to write
+% hyperlinking for \docfm. The suggestion came out when writing of
+% \gmdoc\ was at such a~stage that I~thought it to be much easier to write
+% a~couple of |\let|s to make \gmdoc\ able to typeset
+% sources written for \docfm\ than to write a~new package that adds
+% hyperlinking to \docfm. So\dots
+%
+% \stanza
+% The \docfm\ package makes |%| an ignored char. Here the |%| delimits
+% the code and therefore has to be `other'. But only the first one
+% after the code. The others we may re|\catcode| to be ignored and we
+% do it indeed in line \ref{ignorePercent}.
+% \stanza
+%
+% At the very beginning of a~\docfm-prepared file we meet a~nice
+% command \cs{Char\-act\-er\-Tab\-le}. My \TeX\ Guru says it's a~bit old
+% fashioned these days so let's just make it notify the user:
+% \CodeDefine\CharacterTable
+% \CodeDefIndex\Character@Table
+\def\CharacterTable{\begingroup
+  \@makeother\{\@makeother\}%
+  \Character@Table}
+
+\begingroup
+\catcode`\<=1 \catcode`\>=2 %
+\@makeother\{\@makeother\}%
+\firstofone<\endgroup
+  \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>>>
+% Similarly as \docfm, \gmdoc\ provides \env{macrocode}, \env{macro}
+% and \env{environment} environments. Unlike in \docfm,
+% |\end{macrocode}| \emph{does not} require to be preceded with any
+% particular number of spaces. Unlike in \docfm, it \emph{is not}
+% a~kind of \env{verbatim}, however, which means the code and
+% narration layers remains in force inside it which means that any
+% text after the first |%| in a~line will be processed as narration
+% (and its control sequences will be executed). For a~discussion of
+% a~possible workaround see line \ref{AVerySpecialMacro}.
+%
+% \stanza
+% Let us now look over other original \docfm's control sequences and
+% let's `domesticate' them if they are not yet.
+% 
+%  \gmdmarginpar{kici kici}
+% The \TextUsage\DescribeMacro\ and \TextUsage\DescribeEnv\ commands
+% seem to correspond with my |\TextUsage| macro in its plain and starred
+% version respectively except they don't typeset their arguments in the text
+% i.e., they do two things of the three. So let's |\def| them to do these
+% two things in this package, too:
+% \CodeDefine\DescribeMacro
+\outer\def\DescribeMacro{%
+  \begingroup\MakePrivateLetters
+  \gmd@ifonetoken\Describe@Macro\Describe@Env}
+% Note that if the argument to |\DescribeMacro| is not
+% a~(possibly starred) control sequence, then as an environment's name
+% shall it be processed \emph{except} the \cs{Mak\-ePriv\-at\-e\-Oth\-ers}
+% re|\catcode|ing shall not be done to it.
+% \CodeDefine\DescribeEnv
+\outer\def\DescribeEnv{%
+  \begingroup\MakePrivateOthers\Describe@Env}
+
+% Actually, I've used the |\Describe...| commands myself a~few times, so
+% let's |\def| a~common command with a~starred
+% version:\CodeDefine\Describe
+\outer\def\Describe{%
+  \begingroup\MakePrivateLetters
+  \@ifstarl{\MakePrivateOthers\Describe@Env}{\Describe@Macro}}
+
+% The below two definitions are adjusted \*s of |\Text@UsgMacro| and
+% |\Text@UsgEnvir|.
+% \CodeDefIndex\Describe@Macro
+\long\def\Describe@Macro#1{%
+  \endgroup
+  \strut\Text@Marginize#1%
+  \@usgentryze#1% we declare kind of formatting the entry
+  \text@indexmacro#1\ignorespaces}
+
+% \CodeDefIndex\Describe@Env
+\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}\ignorespaces}
+
+% Note that here the environments' names are typeset in |\tt| font
+% just like the macros', \emph{unlike} in \docfm.
+%
+% \stanza
+% My understanding of `minimality' includes avoiding too much freedom
+% as causing chaos not beauty. That's the philosophical and \ae^^B
+% sthetic reason why I~don't provide \TextUsage\MacroFont. In
+% my opinion there's a~noble tradition of typesetting the \TeX\ code
+% in |\tt| font nad this tradition sustained should be. If one wants
+% to change the tradition, let \himher\ redefine |\tt|, in \TeX\ it's
+% no problem. I~suppose |\MacroFont| is not used explicitly, and that
+% it's (re)defined at most, but just in case let's |\let|:
+
+\let\MacroFont\tt
+
+%
+% \stanza
+% We have provided \TextUsage\CodeIndent\ in line
+% \ref{CodeIndent}. And it corresponds with \docfm's
+% \Describe\MacroIndent\cs{Mac\-ro\-In\-dent} so
+% \CodeDefine\MacroIndent
+\let\MacroIndent\CodeIndent
+% And similarly the other skips: \CodeDefine\MacrocodeTopsep
+\let\MacrocodeTopsep\CodeTopsep
+% Note that \TextUsage\MacroTopsep\ is defined in \gmdoc\ and has
+% the same r\^ole as in \docfm.
+
+\let\SpecialEscapechar\CodeEscapeChar
+
+% \TextUsage\theCodelineNo\ is not used in
+% \pk{gmdoc}. Instead of it there is \TextUsage\LineNumFont\
+% declaration and a~possibility to redefine |\thecodelinenum| as for
+% all the counters. Here the |\LineNumFont| is used two different
+% ways, to set the benchmark width for a~linenumber among others, so it's not
+% appropriate to put two things into one macro. Thus let's give the
+% user a~notice if \heshe\ defined this macro: 
+%
+% Because of possible localness of the definitions it seems to be
+% better to add a~check at the end of each |\DocInput| or
+% |\IndexInput|. 
+
+\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.}}}
+% I~hope this lack will not cause big trouble.
+%
+% \stanza
+% For further notifications let's define a~shorthand:
+% \CodeDefine\noeffect@info
+\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}}}
+
+% The four macros formatting the macro and environment names, namely
+% \possfil\TextUsage\PrintDescribeMacro,
+% \possfil\TextUsage\PrintMacroName,
+% \possfil\TextUsage\PrintDescribeEnv\ and
+% \possfil\TextUsage\PrintEnvName\  are not
+% supported by \pk{gmdoc}. They seem to me to be too internal to take
+% care of them. Note that in the name of (\ae sthetical) minimality
+% and (my) convenience I~deprive you of easy knobs to set strange
+% formats for verbatim bits: I~think they are not advisable.
+%
+% Let us just notify the user. 
+\AtEndInput{%
+  \noeffect@info{PrintDescribeMacro}%
+  \noeffect@info{PrintMacroName}%
+  \noeffect@info{PrintDescribeEnv}%
+  \noeffect@info{PrintEnvName}}
+  
+% The \TextUsage\CodelineNumbered\ declaration of \docfm\ seems to be
+% equivalent to our |noindex| option with the |linesnotnum| option set
+% off so let's define it such a~way.
+\def\CodelineNumbered{\AtBeginDocument{\gag@index}}
+\@onlypreamble\CodelineNumbered
+% Note that if the |linesnotnum| option is in force, this declaration
+% shall not revert its effect. 
+%
+% I~assume that if one wishes to use \docfm's interface then \heshe'll
+% not use \gmdoc's options but just the default.
+%
+% \stanza
+% The |\CodelineIndex| and |\PageIndex| declarations
+% correspond with  the \pk{gmdoc}'s default and the |pageindex| option
+% respectively.
+% Therefore let's |\let|
+% \CodeDefine\CodelineIndex
+\let\CodelineIndex\@pageindexfalse
+\@onlypreamble\CodelineIndex
+
+% \CodeDefine\PageIndex 
+\let\PageIndex\@pageindextrue
+% \label{PageIndex}
+\@onlypreamble\PageIndex
+
+% The next two declarations I~find useful and
+% smart:\CodeDefine\DisableCrossrefs
+\def\DisableCrossrefs{\@bsphack\gag@index\@esphack}
+
+% \CodeDefine\EnableCrossrefs
+\def\EnableCrossrefs{\@bsphack\ungag@index
+  \def\DisableCrossrefs{\@bsphack\@esphack}\@esphack}
+% The latter definition is made due to the footnote 6 on p.\,8 of the
+% Frank Mittelbach's \docfm's documentation and both of them are
+% copies of lines 302--304 of it modulo |\|(|un|)|gag@index|.
+%
+% \stanza
+% The subsequent few lines I~copy almost verbatim ;-) from the lines
+% 611--620.\CodeDefine\AlsoImplementation
+\newcommand*\AlsoImplementation{\@bsphack%
+  \long\def\StopEventually##1{\gdef\Finale{##1}}% we define
+  % \cs{Fi\+n\+a\+le} just to expand to the argument of |\StopEventually| not
+  % to to add anything to the end input hook because |\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 virtually numerous) input(s).
+  % \nostanza\par
+    \@esphack}
+
+\AlsoImplementation
+
+% \begin{quotation} When the user places an |\OnlyDescription|
+%   declaration in the driver file the document should only be typeset
+%   up to |\StopEventually|. We therefore have to redefine this
+%   macro.\end{quotation} \CodeDefine\OnlyDescription 
+
+\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}
+
+% \begin{quotation} If no |\StopEventually| command is given we
+%   silently ignore a~|\Finale| issued.\end{quotation}
+
+\@relaxen\Finale
+
+% The \TextUsage\meta\ macro is so beautifully crafted in \docfm\ that
+% I~couldn't resist copying it into \pk{gmutils}. It's also available
+% in Knuthian (\TeXbook\ format's) disguise
+% \Describe*{\<...>}|\<|\<the argument>|>|.
+% \stanza
+% The checksum mechanism is provided and developed for a~slightly
+% different purpose.
+% \stanza
+% Most of \docfm's indexing commands have already been `almost
+% defined' in \pk{gmdoc}:
+% \CodeDefine\SpecialMainIndex
+\let\SpecialMainIndex=\CodeDefIndex
+
+% \DoNotIndex\endcsname*
+% \CodeDefine\SpecialMainEnvIndex
+\def\SpecialMainEnvIndex{\csname CodeDefIndex\endcsname*}% we don't
+% type |\CodeDefIndex| explicitly here 'cause it's |\outer|, remember?
+
+% \CodeDefine\SpecialIndex \CodeDefine\SpecialUsageIndex
+% \CodeDefine\SpecialEnvIndex \CodeDefine\SortIndex
+\let\SpecialIndex=\CodeCommonIndex
+
+\let\SpecialUsageIndex=\TextUsgIndex
+
+\def\SpecialEnvIndex{\csname TextUsgIndex\endcsname*}
+
+\def\SortIndex#1#2{\index{#1\actualchar#2}}
+% \begin{quotation}All these macros are usually used by other macros;
+% you will need them only in an emergency.\end{quotation}
+% 
+% Therefore I~made the assumption(s) that `Main' indexing macros are
+% used in my `Code' context and the 'Usage' ones in my `Text' context.
+%
+% \stanza
+% Frank Mittelbach in \docfm\ provides the \TextUsage\verbatimchar\
+% macro to (re)define the |\verb(*)|'s delimiter for the index
+% entries. The \gmdoc\ package uses the same macro and its default definition
+% is |{&}|. When you use \docfm\ you
+% may have to redefine |\verbatimchar| if you use (and index) the |\+|
+% control sequence. \pk{gmdoc} does a~check for the analogous situation
+% (i.e., for processing |\&|) and
+% if it occures it takes |$| as the |\verb*|'s delimiter. So strange
+% delimiters are chosen deliberately to allow any `other' chars in the
+% environments' names. If this would cause problems, please
+% notify me and we'll think of adjustments.
+%
+% \CodeDefine\verbatimchar
+% \changes{v0.98c}{06/9/10}{put into all indexing macros for the
+% accordance of the `macro' and the `environment' index entries; the 
+% \cs[]{\$} sign set as its alternative}
+\def\verbatimchar{&}
+
+% One more a~very neat macro provided by \docfm. I~copy it
+% verbatim and put into \pk{gmutils}, too. (|\DeclareRobustCommand|
+% doesn't issue an error if its argument has been defined, it only
+% informs about redefining.)
+% \CodeDefine\*\changes{v0.98}{06/09/05}{made robust.}
+\DeclareRobustCommand*\*{\leavevmode\lower.8ex\hbox{$\,\widetilde{\ }\,$}}
+
+
+% \TextUsage\IndexPrologue\ is defined in line
+% \ref{IndexPrologue}. And other \docfm\ index commands too.
+
+% \CodeDefine\main \CodeDefine\usage
+\@ifundefined{main}{}{\let\DefEntry=\main}
+
+\@ifundefined{usage}{}{\let\UsgEntry=\usage}
+
+% About how the \ds\ directives are supported by \pk{gmdoc},
+% see section \gmiflink[docstrip]{The \ds\dots}.
+% This support is not \emph{that} sophisticated as in \docfm, among others, it
+% doesn't count the modules' nesting. Therefore if we dont want an
+% error while \pk{gmdoc}umenting \docfm-prepared files, better let's
+% define \docfm's counter for the modules' depths.
+% \CodeDefine*{StandardModuleDepth}
+\newcounter{StandardModuleDepth}
+
+% \stanza
+% For now let's just mark the macro for further development
+% \CodeDefine*\DocstyleParms
+\noeffect@info{DocstyleParms}
+
+% For possible further development or to notify the user once and
+% forever:
+% \CodeDefine\DontCheckModules \CodeDefine\CheckModules
+\@emptify\DontCheckModules \noeffect@info{DontCheckModules}
+\@emptify\CheckModules \noeffect@info{CheckModules}
+% The \TextMarginize\Module|\Module| macro \emph{is} provided exactly
+% as in \docfm.
+% \CodeDefine\AltMacroFont
+\@emptify\AltMacroFont \noeffect@info{AltMacroFont}
+
+% \begin{quotation} And finally the most important bit: we change the
+% |\catcode| of |%| so that it is ignored (which is how we are able to
+% produce this \pk{doc}ument!). We provide two commands to do the
+% actual switching.\end{quotation}
+% \CodeDefine\MakePercentIgnore \CodeDefine\MakePercentComment
+\def\MakePercentIgnore{\catcode`\%9\relax}
+\def\MakePercentComment{\catcode`\%14\relax}
+%
+%
+% \subdivision{\gmdoc ing \file{doc.dtx}}
+%
+% The author(s) of \docfm\ suggest(s): \begin{quotation}For examples
+% of the use of most---if not all---of the features described above
+% consult the \file{doc.dtx} source itself.\end{quotation}
+% Therefore I~hope that after \file{doc.dtx} has been \gmdoc-ed, one
+% can say \gmdoc\ is \docfm-compatible ``at most---if not at all''.
+%
+% \TeX ing the original \docfm\ with my humble\footnote{What ^^B(
+% a~\emph{false} modesty! ;-)}  package was a~challenge and
+% a~milestone experience in my \TeX\ life.
+%
+% \stanza
+% One of minor errors was caused by my understanding of a~`shortverb'
+% char: due to \pk{gmverb}, in the math mode an active `shortverb' char
+% expands to itself's `other' version thanks to |\string| (It's done
+% with \verb+|+ in mind). \docfm's
+% concept is different, there a~`shortverb' char should in the math
+% mode work as shortverb. So let it be as they wish: \pk{gmverb}
+% provides |\OldMakeShortVerb| and the oldstyle input commands change
+% the inner macros so that also |\MakeShortVerb| works as in \docfm
+% (cf.\ line \ref{oldmkshvrrb}). 
+%
+% We also redefine the \env{macro} environment to make it mark the
+% first code line as the point of defining of its argument, because
+% \pk{doc.dtx} uses this environment also for implicit definitions.
+%
+% \changes[\OldMakeShortVerb]{v0.98a}{06/09/05}{and pals moved to
+% \pk{gmverb}.} 
+%\CodeDefine\OldDocInput
+% \changes{v0.98}{06/09/05}{\cs{@makeother\protect\bslash_} added}
+% \changes{v0.98a}{06/09/05}{\cs{AtBegInput} changed into
+% \cs{AtBegInputOnce}.}
+% \changes{v0.98b}{06/09/07}{enrichments of the \env{macrocode(*)}
+% definitions moved to the default definitions of these envs.}
+\def\OldDocInput{%
+  \AtBegInputOnce{\StraightEOL
+    \let\@MakeShortVerb=\old@MakeShortVerb
+    %\label{oldmkshvrrb}\CodeDefIndex\gmd@@macro
+    \let\gmd@@macro\macro
+    \def\macro{\let\gmd@ifonetoken\@secondoftwo\gmd@@macro}% (Of course,
+  % na\-\"\i{}\-ve
+  % \nlpercent|\exp...\let\exp...\macro\cs...\macro*\endcs...| caused
+  % an infinite loop since in the definition of \env{macro*} the
+  % |\macro| macro occures.)
+    \VerbMacrocodes}%
+  \bgroup\@makeother\_% it's to allow |_| in the filenames. The next
+  % macro will close the group.
+  \Doc@Input}
+% We don't swith the |@codeskipput| switch neither we check it because
+% in `old' world there's nothing to switch this switch in the
+% narration layer.
+%
+% \stanza
+% I~had a~hot and wild \TeX\ all the night nad what a~bliss when
+% the `Succesfully formated 67 page(s)' message appeared. 
+% 
+% My package needed fixing some bugs and adding some compatibility
+% adjustments (listed in the previous section) and the original
+% \pk{doc.dtx} source file needed a~few adjustments too because some
+% crucial differences came out. I'd like to write a~word about them now.
+%
+% \stanza
+% The first but not least is that the author(s) of \docfm\ give the
+% CS marking commands non-macro arguments sometimes, e.g.,
+% |\DescribeMacro{StandardModuleDepth}|. Therefore we should launch
+% the \emph{starred} versions of corresponding \gmdoc\ commands. This
+% means the \docfm-like commands will not look for the CS's occurrence
+% in the code but will mark the first codeline met.
+% \stanza
+%
+% Another crucial difference is that in \gmdoc\ the narrative and the
+% code layers are separated with only the code delimiter and
+% therefore may be much more mixed than in \docfm. among others, the \env{macro}
+% environment is \emph{not} a~typical \env{verbatim} like: the texts
+% commented out within \env{macrocode} are considered a~normal
+% commentary i.e., not verbatim. Therefore some macros `commented out'
+% to be shown verbatim as an example source must have been `additionally'
+% verbatimized for \gmdoc\ with the shortverb chars e.g. You may also
+% change the code delimiter for a~while, e.g., the line
+% \CodeDelim*\.
+% \AVerySpecialMacro % delete the first % when.\unskip|..|\par\CDPerc
+%
+% \gmdnoindent\label{AVerySpecialMacro}
+%  was got with
+% \begin{verbatim}
+%\CodeDelim*\.
+%% \AVerySpecialMacro % delete the first % when.\unskip|..|\CDPerc
+%\end{verbatim}
+%
+% \leftskip0sp\relax
+% One more difference is that my shortverb chars expand to their
+% \catother\ versions in the math mode while in \docfm\ remain
+% shortverb, so I~added a~declaration |\OldMakeShortVerb| etc.
+%
+% \stanza
+% Moreover, it's \TeX ing \docfm\ what inspired adding the
+% |\StraightEOL| and |\QueerEOL| declarations.
+%
+%
+%
+% \division{Polishing, Development and Bugs}
+% 
+% \begin*{bulletpars}
+%   \everypar{$\bullet$ }
+%   \label{Tasks}|\MakePrivateLetters| theoretically may interfere
+%   with |\active|ating some chars to allow linebreaks. But making
+%   a~space or an opening brace a~letter seems so perverse that we may
+%   feel safe not to take account of such a~possibility.
+%
+%   When |countalllines| option is enabled, the comment lines
+%   that don't produce any printed output result with a~(blank) line too
+%   because there's put a~hypertarget at the beginning of them. But for
+%   now let's assume this option is for draft versions so hasn't be
+%   perfect.
+%   
+%   Marcin Woli\'nski suggests to add the marginpar clauses for the
+%   AMS classes as we did for the standard ones in the lines
+%   \ref{mparclause1}--\ref{mparclause2}. Most probably I~can do it on
+%   request when I~only know the classes' names and their `marginpar
+%   status'.
+%
+%   When the |countalllines| option is in force, some |\list|
+%   environments shall raise the `missing |\item|' error if you don't
+%   put the first |\item| in the same line as
+%   |\begin{|\<environment>|}| because the (comment-) line number is
+%   printed.^^A~\end{} 
+%
+% I'm prone to make the control sequences hyperlinks to the(ir)
+% `definition' occurrences. It doesn't seem to be a~big work compared
+% with what has been done so far.
+%     
+% Is |\RecordChanges| really necessary these days? Shouldn't be the
+% |\makeglossary| command rather executed by default?\footnote{It's ^^B
+% understandable that ten years earlier writing things out to the files ^^B
+% remarkably decelerated \TeX, but nowadays it does not in most ^^B
+% cases. That's why \cs{makeindex} is launched by default in \gmdoc.}
+%
+% Do you use |\listoftables| and/or |\listoffigures| in your
+% documentations? If so, I~should `EOL-straighten' them like
+% |\tableofcontents|, I~suppose (cf.\ line \ref{straighttoc}).
+%
+% Some lines of non-printing stuff such as |\CodeDefine...| and
+% |\changes| connecting the narration with the code resulted with
+% unexpected large vertical space. Adding a~fully blank line between
+% the printed narration text and not printed stuff helped.
+%
+% My \TeX\ Guru remarked that |\jobname| expands to the main file name
+% without extension iff that extension is \file{.tex}. Should
+% I~replace |\jobname| with |\jobnamewoe| then? (The latter always
+% expands to the file name without extension.)
+%
+% About the \ds\ \gmiflink{verbatim mode directive} see above.
+%
+% \end{bulletpars}
+% 
+% \def\EOFMark{\<eof>}
+% \division{(No) \EOFMark} If the user doesn't wish the
+% documentation to be ended by \EOFMark, \heshe\
+% should end the file with a~comment ending with |\NoEOF| macro
+% defined below\footnote{Thanks to Bernd Raichle at Bacho\TeX\ 2006 ^^B
+%   Pearl Session where he presented \cs{input}ing a~file inside ^^B
+%   \cs{edef}.}:
+%
+% \CodeDefine\NoEOF
+% \changes{v0.98}{06/09/05}{extended to add it the \cs{endinput} effect}
+% \changes{v0.98l}{06/10/27}{divided in two macros first of which
+% makes queer EOL and the latter gobbles the stuff till the EOL to
+% suppress possible \cs{endinput} when used in \cs{StopEventually}}
+
+\bgroup\catcode`\^^M\active \firstofone{\egroup%
+  \def\@NoEOF#1^^M{%
+    \@relaxen\EOFMark\expandafter\noexpand\endinput^^M}}
+
+\def\NoEOF{\QueerEOL\@NoEOF}
+
+% \label{NoEOF} As you probably see, |\NoEOF| has also the
+% |\endinput| effect. 
+
+\endinput
+%^^A~\docstrip
+%</package>
+%\PrintChanges \PrintIndex 
+%
+%
+%^^A~place for general changes:
+%
+% \changes{v0.98a}{06/09/06}{\cs{AtBeginput}, \cs{AtEndinput},
+% \cs{AtBeginputOnce} renamed to \cs{AtBegInput}, \cs{AtEndInput}
+% \cs{AtBegInputOnce} respect.}
+%
+% \changes{v0.98c}{06/9/9}{making \cs{CodeDefine} and \cs{CodeUsage}
+% markers to add up; bug fixes in indexing macros and change of
+% concept of \cs{Code(Define$\quotechar|$Usage)*}: hence they serve not only for
+% environments but also for implicit def/use of macros}
+%
+% \changes{v0.98d}{06/9/11}{mostly editing the narrative, marking
+% special index entries etc.; queering \cs{char1} and \cs{char2}
+% repeated \cs{AtBegInput}}
+%
+% \changes{v0.98e}{06/09/24}{Some macros moved to \pk{gmutils}:
+% \cs{cs} and pals and one-chars with queer \cs{catcode}s}
+%
+% \changes{v0.98f}{06/09/27}{A~bug fixed: \cs{Code@MarginizeMacro} set
+% to define a~CS and the corresponding test set to check if it's
+% undefined. In all three such definitions and resets after the use,
+% \cs{def} is changed to \cs{(g)let}}
+%
+% \changes{v0.98g}{06/10/10}{among others, \cs{discretionary}s for breaking
+% a~CS to percent at the beginning of the lower line. Moreover, fixing
+% a~bug/feature that leaves the code leftskip in the narration when an
+% inline comment is followed by another codeline (w.o.\ explicit
+% \cs{par}). And lots of finishing touches to the text. A~special font
+% for the marginpar CSs among others}
+%
+% \changes{v0.98l}{06/10/26}{in \cs{AtBegInputOnce} an auxiliary macro
+% for each use substituted with one macro added at begin input. In
+% \cs{gmd@evpaddonce} a~counter substituted with a~numeric
+% macro. \cs{@ifQueerEOL} made polite i.e., a~two-argument not
+% expanding to an open \cs{if...}}
+%
+% \changes{v0.99}{06/11/24}{\env{oldmc(*)} implemented that's
+% (hope)fully  compatible with \docfm's \env{macrocode(*)}. Moreover,
+% a~declaration letting \gmdoc's \env{macrocode} to \env{oldmc(*)}}
+%
+% \changes{v0.99a}{06/11/30}{The \ds\ directives implemented fully
+% automatic (no more need of \cs{doctrip(s)} declarations). Moreover,
+% some minor changes due to \TeX ing The Source.}
+%
+% \Finale
+
+% (For my GNU Emacs:)
+%%% Local Variables: 
+%%% mode: latex
+%%% TeX-master: "e:/LaTeX/TeXGuru/gmdoc/gmdocDoc"
+%%% End: 
diff --git a/Master/texmf-dist/tex/latex/gmdoc/gmdocc.cls b/Master/texmf-dist/tex/latex/gmdoc/gmdocc.cls
new file mode 100644
index 00000000000..a03d4c4aa43
--- /dev/null
+++ b/Master/texmf-dist/tex/latex/gmdoc/gmdocc.cls
@@ -0,0 +1,307 @@
+%
+% \GetFileInfo{gmdocc.cls}
+% \title{The \pk{gmdocc} Class For \pk{gmdoc} Driver Files\thfileinfo}
+% \author{Grzegorz `Natror' Murzynowski}
+% \maketitle
+%
+% \begin{copyrnote}
+%
+%%Written by Grzegorz `Natror' Murzynowski,
+%% natror at o2 dot pl
+%%
+%% \copyright\,2006 by Grzegorz `Natror' Murzynowski.
+%%
+%% This program is subject to the \LaTeX\ Project Public License.
+%% See ^^A
+%% \url{http://www.ctan.org/tex-archive/help/Catalogue/licenses.lppl.html} ^^A
+%% for the details of that license.
+%%
+%% LPPL status: "author-maintained".\par
+%%
+%
+%\end{copyrnote}
+%
+% \RecordChanges
+% \ChangesStart{v0.677}{1000/00/00}
+%
+% \chschange{v0.69}{06/10/1}{81}
+% \chschange{v0.70}{06/10/10}{98}
+% \chschange{v0.71}{06/10/21}{114}
+% \chschange{v0.72}{06/11/14}{122}
+% \chschange{v0.73}{06/12/1}{128}
+%
+
+\NeedsTeXFormat{LaTeX2e}
+\ProvidesClass{gmdocc}
+            [2006/12/01 v0.73 a class for gmdoc driver files (GM)]
+
+%
+% \division{Intro}
+%%
+%% This file is a~part of \pk{gmdoc} bundle and provides a~document
+%% class for the driver files documenting \LaTeXpar\ packages \&a.\
+%% with my \pk{gmdoc.sty} package. It's not necessary, of course: most
+%% probably you may use another document class you like.
+%%
+%% By default this class loads \pk{mwart} class with a4paper (default)
+%% option and \pk{lmodern} package with QX fontencoding. It loads also
+%% my \pk{gmdoc} documenting package which loads some auxiliary
+%% packages of mine and the standard ones.
+%% 
+%% If the \pk{mwart} class is not found, the standard \pk{article}
+%% class is loaded instead. Similarly, if the \pk{lmodern} is not
+%% found, the standard Computer Modern font family is used in the
+%% default font encoding.
+%% 
+%
+% \division{Usage}
+%
+% For the ideas and details of \pk{gmdoc}ing of the \LaTeXpar\ files
+% see the \pk{gmdoc.sty} file's documentation (chapter \ref{gmd}). The
+% r\^ole of the \pk{gmdocc} document class is rather auxiliary and
+% exemplary.  Most probably, you may use your favourite document class
+% with the settings you wish. This class I~wrote to meet my needs of
+% fine formatting, such as not numbered sections and sans serif demi
+% bold headings.
+%
+% However, with the users other than myself in mind, I~added some
+% conditional clauses that make this class works also if an \pk{mwcls}
+% class or the \pk{lmodern} package are unknown.  \dekmedskip
+%
+% Of rather many options supported by \pk{gmdoc.sty}, this class
+% chooses my favourite, i.e., the default. An exception is made for
+% the \TextUsage*{noindex} option, which is provided by this class and
+% passed to \pk{gmdoc.sty}. This is intended for the case you don't
+% want to make an index.
+%
+% Simili modo, the \TextUsage*{nochanges} option is provided to turn
+% creating the change history off. 
+%
+% Both of the above options turn the \emph{writing out to the files}
+% off. They don't turn off |\PrintIndex| nor |\PrintChanges|. (Those
+% two commands are no-ops by themselves if there's no \file{.ind}
+% (n)or \file{.gls} file respectively.)
+% \dekmedskip
+%
+% One more option is \TextUsage*{outeroff}. It's intended for
+% compiling the documentation of macros defined with the |\outer|
+% prefix. It |\relax|es this prefix so the `|\outer|' macros' names can
+% appear in the arguments of other macros, which is necessary to
+% pretty mark and index them. 
+%
+% I~decided not to make discarding |\outer| the default because
+% it seems that \LaTeX\ writers don't use it in general and
+% \pk{gmdoc.sty} \emph{does} make some use of it.
+% \dekmedskip
+% 
+% This class provides also the \TextUsage*{debug} option. It turns the
+% |\if@debug| Boolean switch True and loads the \pk{trace} package that
+% was a~great help to me while debugging \pk{gmdoc.sty}.
+% \dekmedskip
+%
+% The default base document class loaded by \pk{gmdocc.cls} is Marcin
+% Woli\'nski's \pk{mwart}. If you have not installed it on your
+% computer, the standard \pk{article} will be used. 
+%
+% Moreover, if you like MW's classes (as I~do) and need |\chapter|
+% (for multiple files' input e.g.), you may declare another \pk{mwcls}
+% with the option homonimic with the class'es name: \TextUsage*{mwrep}
+% for \pk{mwrep} and \TextUsage*{mwbk} for \pk{mwbk}. For the symmetry
+% there's also \TextUsage*{mwart} option (equivalent to the default
+% setting).
+%
+% The existence test is done for any MW class option as it is in the
+% default case.
+%
+% \stanza
+% The \TextUsage\EOFMark\ is in this class typesets like this (of
+% course, you can redefine it as you wish):\par\EOFMark
+%
+% \StopEventually\NoEOF
+%
+% \division{The Code}
+
+% \CodeDefIndex\gmcc@baseclass
+\def\gmcc@baseclass{mwart}% the default is Marcin Woli\'nski's class (\pk{mwcls})
+% analogous to \pk{article}.
+
+% Since you can choose the standard \pk{article} class, we'd better
+% provide a~Boolean switch to keep the score of what was chosen. It's
+% to avoid unused options if \pk{article} is chosen.
+%\CodeDefIndex*{gmcc@mwcls}
+\newif\ifgmcc@mwcls
+\gmcc@mwclstrue
+% \CodeDefine*{mwart}
+\DeclareOption{mwart}{\def\gmcc@baseclass{mwart}}% The \pk{mwart} class may also
+% be declared explicitly.
+% \CodeDefine*{mwrep}
+\DeclareOption{mwrep}{\def\gmcc@baseclass{mwrep}}% If you need
+% chapters, this option chooses an MW's class that corresponds to
+% \pk{report},
+% \CodeDefine*{mwbk}
+\DeclareOption{mwbk}{\def\gmcc@baseclass{mwbk}}% and this MW's class
+% corresponds to \pk{book}.
+% \CodeDefine*{article}
+\DeclareOption{article}{\gmcc@mwclsfalse}% you can also
+% \emph{choose} \pk{article}.
+
+
+% \CodeDefine*{outeroff}
+\DeclareOption{outeroff}{\let\outer\relax}% This option allows
+% |\outer|-prefixed macros to be \pk{gmdoc}-processed with all the
+% bells and whistles.
+
+% \CodeDefine\if@debug
+\newif\if@debug
+% \CodeDefine*{debug}
+\DeclareOption{debug}{\@debugtrue}% This option causes \pk{trace} to
+% be loaded and the Boolean switch of this option may be used to hide
+% some things needed only while debugging.
+
+% \CodeDefine*{noindex}
+\DeclareOption{noindex}{%
+  \PassOptionsToPackage{\CurrentOption}{gmdoc}}% This
+% option turns the writing outto \file{.idx} file off.
+
+% \CodeDefIndex\if@gmccnochanges
+% \CodeDefIndex*\@gmccnochangestrue
+% \CodeDefIndex*\@gmccnochangesfalse
+\newif\if@gmccnochanges
+% \CodeDefine*{nochanges}
+\DeclareOption{nochanges}{\@gmccnochangestrue}% This option turns the
+% writing outto \file{.glo} file off.
+
+% \CodeDefine*{gmeometric}
+\def\gmTheGeometry{geometry}
+\DeclareOption{gmeometric}{%
+  \IfFileExists{gmeometric.sty}{%
+    \def\gmTheGeometry{gmeometric}}{}}% It's an experimental option
+% that causes the |\geometry| macro provided by \pk{geometry} package
+% is not restricted to the preamble. This option causes the
+% \pk{gmeometric} package is loaded (if available) that works the
+% limitation around.
+
+
+\ProcessOptions
+
+% \newif\ifSecondClass^^A~to assure \TeX knows ifSecondClass
+% ^^A switch and therefore it counts the fis properly.
+% \iffalse It's for cooperation with my another class. (Feature under
+% construction.)
+\@ifundefined{ifSecondClass}{%
+  \newif\ifSecondClass}{}
+%\fi\fi
+
+%\iffalse
+\ifSecondClass\else
+%\fi\fi
+\ifgmcc@mwcls
+  \IfFileExists{\gmcc@baseclass.cls}{}{\gmcc@mwclsfalse}% As
+  % announced, we do the ontological test to any \pk{mwcls}.
+\fi 
+\ifgmcc@mwcls
+  \LoadClass[fleqn, oneside, noindentfirst, 11pt, withmarginpar,
+  sfheadings]{\gmcc@baseclass}% 
+\else
+  \LoadClass[fleqn, 11pt]{article}% Otherwise the standard \pk{article} is
+  % loaded. 
+\fi
+
+\AtBeginDocument{\mathindent=\CodeIndent}
+% The |fleqn| option makes displayed formuals be flushed left and
+% |\mathindent| is their indentation. Therefore we ensure it is always
+% equal |\CodeIndent| just like |\leftskip| in \env{verbatim}. Thanks
+% to that you may display single verbatim lines with |\[...\]|:
+% \[\hbox{\verb+\[\hbox{|\verbatim\stuff|}\]+\,.}\]
+
+\IfFileExists{lmodern.sty}{% We also examine the ontological status of
+  % this package
+  \RequirePackage{lmodern}% and if it shows to be satisfactory (the
+  % package shows \emph{to be}), we load it and set the proper
+  % font encoding.
+  \RequirePackage[QX]{fontenc}%
+}{}
+%\iffalse\iffalse
+\fi
+% of ifSecondClass (a~feature under construction).\fi
+
+\RequirePackage[margin=2.7cm, left=4cm,
+   right=2.2cm]{\gmTheGeometry}% Now we set the
+% page layout.
+
+\def\gmdoccMargins{%
+  \geometry{margin=2.7cm, left=4cm, right=2.2cm}}
+
+\if@debug%  For debugging we load also the \pk{trace} package that was
+% very helpful to me.
+  \RequirePackage{trace}%
+  \errorcontextlines=100 % And we set an error info parameter.
+\fi
+
+%\CodeDefine\ifdtraceon
+%\CodeDefine\ifdtraceoff
+\newcommand*\ifdtraceon{\if@debug\afterfi\traceon\fi}
+\newcommand*\ifdtraceoff{\if@debug\traceoff\fi}
+
+%  We load the core package:
+\RequirePackage{gmdoc}
+
+\@ifpackageloaded{lmodern}{% The Latin Modern font family provides
+  % a~light condensed typewriter font that seems to be the most
+  % suitable for the marginpar CS marking.
+  \def\marginpartt{\normalfont\fontseries{lc}\ttfamily}}%
+
+
+% ^^A\setcounter{IndexColumns}{2} narrow columns do spare space and
+% ^^A~look better
+\raggedbottom
+
+\setcounter{secnumdepth}{0}% We wish only the parts and chapters to be
+% numbered. 
+
+\renewcommand*\thesection{\arabic{section}}% isn't it redundant at the
+% above setting?
+
+\@ifnotmw{}{%
+  \@ifclassloaded{mwart}{% We set the indentation of Contents:
+    \SetTOCIndents{{}{\quad}{\quad}{\quad}{\quad}{\quad}{\quad}}}{% for \pk{mwart}
+    \SetTOCIndents{{}{\bf9.\enspace}{\quad}{\quad}{\quad}{\quad}{\quad}}}% and for the two other \pk{mwcls}s.
+  \pagestyle{outer}}% We set the page numbers to be printed in the
+    % outer and bottom corner of the page.
+
+\def\titlesetup{\bfseries\sffamily}% We set the title(s) to be
+% boldface and sans serif.
+
+\if@gmccnochanges\let\RecordChanges\relax\fi% If the |nochanges|
+% option is on, we discard writing outto the \file{.glo} file.
+
+\RecordChanges% We turn the writing the |\changes| outto the
+% \file{.glo} file if not the above.
+
+% \CodeUsgIndex\dekclubs \CodeUsgIndex\smartunder
+\dekclubs% We declare the club sign \verb+|+ to be a~shorthand for |\verb*|
+\smartunder% and we declare the |_| char to behave as usual in the
+% math mode and outside math to be just an uderscore.
+
+\exhyphenpenalty\hyphenpenalty% 'cause \pk{mwcls} set it |=10000| due
+% to Polish customs.
+
+\RequirePackage{amssymb}
+\def\EOFMark{\rightline{\ensuremath{\square}}}
+
+
+\endinput
+% \changes{v0.677}{06/9/23}{Added checkings if \pk{mwart.cls} and
+% \pk{lmodern.sty} are available and more standard behaviour if they
+% are not}
+%
+% \changes{v0.70}{06/10/10}{Added some options and Latin Modern Light
+% Condensed to the marginpar CSs.}
+%
+% \Finale
+
+% (For my GNU Emacs:)
+%%% Local Variables: 
+%%% mode: latex
+%%% TeX-master: "e:/LaTeX/TeXGuru/gmdoc/gmdocDoc"
+%%% End: