home *** CD-ROM | disk | FTP | other *** search
Text File | 1992-09-08 | 187.0 KB | 4,555 lines |
- % \iffalse meta-comment
- %
- % Copyright (C) 1989-1992 by Frank Mittelbach. All rights reserved.
- %
- % This file is part of the doc package.
- %
- % IMPORTANT NOTICE:
- %
- % You are not allowed to change this file. You may however copy
- % this file to a file with a different name and then change the
- % copy if you obey the restrictions on file changes described in
- % readme.mz.
- %
- % You are NOT ALLOWED to distribute this file alone. You are NOT
- % ALLOWED to take money for the distribution or use of this file
- % (or a changed version) except for a nominal charge for copying
- % etc.
- %
- % You are allowed to distribute this file under the condition that
- % it is distributed together with all files mentioned in
- % readme.mz0.
- %
- % If you receive only some of these files from someone, complain!
- %
- % However, if these files are distributed by established suppliers
- % as part of a complete TeX distribution, and the structure of the
- % distribution would make it difficult to distribute the whole set
- % of files, *those parties* are allowed to distribute only some of
- % the files provided that it is made clear that the user will get
- % a complete distribution-set upon request to that supplier (not
- % me). Notice that this permission is not granted to the end
- % user.
- %
- %
- % For error reports in case of UNCHANGED versions see readme.mz
- %
- %
- % \fi
- % ^^A -*-LaTeX-*-
-
- % ^^A These shouldn't come out in .ist files, hence the module
- % ^^A comments, or in the printed version, hence temporary comment
- % ^^A category for `<'
- %\catcode`\<=14
- %<*style>
- \def\fileversion{v1.7k}
- \def\filedate{92/08/24}
- \def\docdate {92/08/28}
- %</style>
- %\catcode`\<=12
-
- % \CheckSum{1985}
- %% \CharacterTable
- %% {Upper-case \A\B\C\D\E\F\G\H\I\J\K\L\M\N\O\P\Q\R\S\T\U\V\W\X\Y\Z
- %% Lower-case \a\b\c\d\e\f\g\h\i\j\k\l\m\n\o\p\q\r\s\t\u\v\w\x\y\z
- %% Digits \0\1\2\3\4\5\6\7\8\9
- %% Exclamation \! Double quote \" Hash (number) \#
- %% Dollar \$ Percent \% Ampersand \&
- %% Acute accent \' Left paren \( Right paren \)
- %% Asterisk \* Plus \+ Comma \,
- %% Minus \- Point \. Solidus \/
- %% Colon \: Semicolon \; Less than \<
- %% Equals \= Greater than \> Question mark \?
- %% Commercial at \@ Left bracket \[ Backslash \\
- %% Right bracket \] Circumflex \^ Underscore \_
- %% Grave accent \` Left brace \{ Vertical bar \|
- %% Right brace \} Tilde \~}
- %%
- %\iffalse This is a METACOMMENT
- % Everything up to the next `\ fi' (without a blank) will
- % be ignored. This is necessary because `%' may no longer
- % be a comment mark when this file is read in.
- %
-
- % Style-option `doc' to use with LaTeX v2.09
- %% Copyright (C) 1989-1992 Frank Mittelbach, all rights reserved.
- %
- %
- % Version: Date: Changes:
- %
- % 1.0a 5.5.88 This is nothing but a collection of tests and
- % hacks. It is certainly going to be greatly
- % changed.
- % Better not to use it!
- % 1.1a 13.5.88 \theindex and \endtheindex redefined.
- % 1.1b 15.5.88 \bslash redefined.
- % \verbatim changed so that `%'s will be ignored.
- % 1.1c 19.5.88 Partly documented and \@temp... in
- % \theindex replaced by private commands.
- % Moreover, \pageshrink incorporated in the
- % computation of the amount of free space on the
- % page.
- % 1.1d 3.6.88 Something goes wrong during the computation of
- % \pageshrink, therefore a bit of empty space is
- % inserted by
- % hand, and some tracing information introduced.
- % It can't stay like that!!
- % 1.2a 23.9.88 Documentation of the macros in DOC-format,
- % i.e. so that this file can document itself.
- % 1.2b 7.10.88 Change of the index environment.
- % 1.2c 10.10.88 Further Documentation.
- % 1.2d 19.10.88 \the before \catcode eliminated. (thanks P.
- % Breitenlohner)
- % 1.3a 24.10.88 Scanning of macro names via \scanmacro
- % introduced to produce index entries.
- % 1.3b 16.1.89 Scanning of macro names via \scanmacro
- % improved.
- % At present only suitable for the makeindex
- % program
- % in conjunction with a special style file.
- % Macro environment in `macro' environment
- % renamed.
- % 1.3c 30.1.89 Index environment removed. Will be developed in
- % other style file.
- % 1.4a 7.2.89 \SpecialEscapechar added (may change its name).
- % 1.4b 7.3.89 \SpecialEscapechar reimplemented.
- % 1.4c 9.3.89 \parts of \short@macro reimplemented because
- % of some bugs in the macros and the makeindex
- % program.
- % Old definition of \b@slash used for \bslash.
- % \b@slash removed.
- % 1.4d 10.3.89 English documentation added.
- % 1.4e 10.3.89 private names for private macros.
- % \makespecialletters renamed to
- % \MakePrivateLetters
- % 1.4f 12.3.89 \short@macro changed (again)
- % documentation update.
- % 1.4g 13.3.89 \DoNotIndex macros changed.
- % 1.4h 13.3.89 \DoNotIndex macros changed to version with
- % \@elt
- % 1.4i 15.3.89 documentation update
- % 1.4j 15.3.89 \printindex inserted
- % 1.4k 21.3.89 documentation update
- % \PrintDescribe... and \Describe... added
- % macro env. + \SpecialIndex changed to accept \
- % 1.4l 29.3.89 macro env. changed once more (\macro@ added) to
- % correct bug introduced in v1.4k
- % \SpecialUsageIndex and \SpecialEnvIndex added.
- % 1.4m 3.4.89 \OnlyDescription added, \MakePrivateLetters
- % changed
- % 1.4n 10.4.89 \MakePrivateLetters changed back to
- % \makeatletter.
- % \ifnot@excluded changed to allow other things
- % than
- % macro names in the exclude list (e.g. \abc*)
- % 1.4o 14.4.89 Documentation update (Brian's correction)
- % \EnableCrossrefs changed to disable
- % \DisableCrossrefs
- % 1.4p 15.4.89 Documentation update \Finale added,
- % some changes to index.tex
- % 1.4q 19.4.89 Documentation update \verbatim changed (RmS)
- % twocolumns env. added index.tex changed.
- % 1.4r 22.4.89 Changed \macro@ to support \par and
- % conditionals (might be wrong.)
- % 1.4s 23.4.89 Getting nearer to the `final' version.
- % \@sphack ...
- % placed in several macros. Documentation update.
- % Tried a \changes macro (somewhere in the middle
- % of this file by using \glossary (might work!)
- % 1.4t 24.4.89 Should change to 1.5? Index now uses multicols
- % added \c@IndexColumns. Other changes forgotten.
- % Some macro names renamed.
- % 1.5a 26.4.89 Finally 1.5: changes to \makelabel in macro
- % env.
- % 1.5b and higher... are documented with the (undocumented) \changes
- % feature.
- % \changes{v1.5f}{89/4/29}{Thanks to Brian who documented the
- % {\tt\bslash changes} macro feature.}
- % \changes{v1.5g}{89/5/07}{MacroTopsep now called MacrocodeTopsep and
- % new MacroTopsep added}
- % \changes{v1.5h}{89/05/17}{All lines shortened to <72 characters}
- % \changes{v1.5j}{89/06/09}{Corrections by Ron Whitney added}
- % \changes{v1.5q}{89/11/03}{`\ldots{}Listing macros renamed to
- % `\ldots{}Input. Suggested by R. Wonneberger}
- % \changes{v1.5W}{90/02/05}{Counter codelineno renamed to CodelineNo}
- %
- %
- %\fi
- %
- % \hyphenation{make-index}
- %
- % \DoNotIndex{\@,\@@par,\@beginparpenalty,\@empty}
- % \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{\cal,\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,\newcount,\newdimen,\newskip}
- % \DoNotIndex{\nopagebreak}
- % \DoNotIndex{\parfillskip,\parindent,\parskip,\penalty,\raise,\rangle}
- % \DoNotIndex{\section,\setlength,\TeX,\topsep,\underline,\unskip,\verb}
- % \DoNotIndex{\vskip,\vspace,\widetilde,\\,\%,\@date,\@defpar}
- % \DoNotIndex{\[,\{,\},\]}
- % \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{\!,\#,\$,\&,\',\(,\),\+,\.,\:,\;,\<,\=,\>,\?,\_}
- % \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,\everypar}
- % \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}
- %
- % \MakeShortVerb{\"}
- % \setcounter{StandardModuleDepth}{1}
- %
- % {\catcode`\p=12 \catcode`\t=12 ^^A hack used later on to print
- % \gdef\dimenvalue#1pt{$#1$pt}} ^^A a register value with a - sign
- %
- % \makeatletter ^^A hack so that this can be printed without multicols
- % \newif \ifmulticols
- % \ifhave@multicol \multicolstrue \fi
- % \makeatother
- % \title{The {\tt doc}--Option\thanks{%
- % This file has version number \fileversion{} dated \filedate{}.
- % The documentation was last revised on \docdate.
- % }}
- % \author{Frank Mittelbach\thanks{Further commentary added at Royal
- % Military College of Science by B. Hamilton Kelly; English
- % translation of parts of the original German commentary
- % provided by Andrew Mills; fairly substantial additions,
- % particularly from {\tt newdoc}, and
- % documentation of post-v1.5q features added at v1.7a by Dave
- % Love (SERC Daresbury Lab).}\\
- % Gutenberg Universit\"at Mainz}
- %
- % \maketitle
- %
- % \begin{abstract}
- % This style option contains the definitions that are necessary to
- % format the documentation of style files. The style file was
- % developed in Mainz in cooperation with the Royal Military College
- % of Science. This is an update which documents various changes
- % and new features in {\sf doc} and integrates the features of {\sf
- % newdoc}.
- % \end{abstract}
- %
- % \ifmulticols
- % \addtocontents{toc}{\protect\begin{multicols}{2}}
- % \fi
- %
- % {\parskip 0pt ^^A We have to reset \parskip
- % ^^A (bug in \LaTeX)
- % \tableofcontents
- % }
- %
- % \changes{v1.7a}{92/02/25}{Miscellaneous small changes to the text}
- %
- % \ifmulticols
- % \begin{multicols}{2}[\section*{Preface to version 1.7}]
- % \else \section*{Preface to version 1.7} \fi
- %
- % This version of {\tt doc.doc} documents changes which have occurred
- % since the last published version \cite{art:doc} but which have been
- % present in distributed versions of {\tt doc.sty} for some time. It
- % also integrates the (undocumented) features of the distributed {\tt
- % newdoc.sty}.
- %
- % The following changes and additions have been made to the user
- % interface since the published version~\cite{art:doc}. See
- % \S\ref{sec:interface} for more details.
- % \begin{description}
- % \item[Driver mechanism] "\DocInput" is now used in the driver file
- % to input possibly multiple independent {\tt doc} files and {\tt doc}
- % no longer has to be the last style option. "\IndexListing" is
- % replaced by "\IndexInput";
- % \item[Indexing] is controlled by "\PageIndex" and
- % "\CodelineIndex", one of which must be specified to produce an
- % index---there is no longer a "\makeindex" in the default
- % "\DocstyleParms";
- % \item[The {\tt macro} environment] now takes as argument the macro
- % name {\em with\/} the backslash;
- % \item[Verbatim text] Newlines are now forbidden inside "\verb" and
- % commands "\MakeShortVerb" and "\DeleteShortVerb" are provided for
- % verbatim shorthand;
- % \item[{\tt \bslash par}] can now be used in "\DoNotIndex";
- % \item[Checksum/character table support] for ensuring the
- % integrity of distributions is added;
- % \item[{\tt \bslash printindex}] becomes "\PrintIndex";
- % \item[{\tt multicol.sty}] is no longer necessary to use {\tt doc} or
- % print the documentation (although it is recommended);
- % \item[`Docstrip' modules] are recognised and formatted specially.
- % \end{description}
- %
- % As well as adding some completely new stuff,
- % the opportunity has been taken to add some commentary to the code
- % formerly in {\tt newdoc.sty} and that added after version 1.5k of
- % {\tt doc.sty}. Since (as noted in the sections concerned) this
- % commentary wasn't written by Frank Mittelbach but the code was, it is
- % probably {\em not\/} true in this case that ``if the code and
- % comments disagree both are probably wrong''!
- %
- % \subsection*{Bugs}
- %
- % There are some known bugs in this version:
- % \begin{itemize}
- % \item The `General changes' glossary entry would come out after
- % macro names with a leading "!" and possibly a leading \verb+"+;
- % \item If you have an old version of {\sf makeindex} long "\changes"
- % entries will come out strangely and you may find the section
- % header amalgamated with the first changes entry. Try to get an
- % up-to-date one (see p.~\pageref{makeindex:version});
- % \item Because the accompanying {\sf makeindex} style files support
- % the inconsistent attribute specifications of older and newer
- % versions {\sf makeindex} always complains about three `unknown
- % specifier's when sorting the index and changes entries.
- % \item If "\MakeShortVerb" and "\DeleteShortVerb" are used with
- % single character arguments, e.g., "{|}" instead of "{\|}" chaos
- % may happen.
- % \end{itemize}
- % (Some `features' are documented below.)
- %
- % \subsection*{Wish list}
- %
- % \begin{itemize}
- % \item Hooks to allow "\DescribeMacro" and "\DescribeEnv" to write
- % out to a special file information about the style's `exported'
- % definitions which they describe. This could subsequently be
- % included in the {\tt docstrip}ped {\tt .sty} file in a suitable form
- % for use by smart editors in command completion, spelling checking
- % etc., based on the style options of a document. This would need
- % agreement on a `suitable form'.
- % \item Indexing of the modules used in {\tt docstrip}'s "%<"
- % directives. I'm not sure how to index directives containing
- % module combinations;
- % \item Writing out bibliographic information about the style;
- % \item Allow turning off use of the special font for, say, the next
- % guarded block.
- % \end{itemize}
- %
- % \ifmulticols
- % \end{multicols}
- %
- % \begin{multicols}{2}[\medskip \rule{\textwidth}{.3pt}
- % \section{Introduction}]
- % \else
- % \section{Introduction}
- % \fi
- %
- % The \TeX{} macros which are described here allow definitions and
- % documentation to be held in one and the same file. This has the
- % advantage that normally very complicated instructions are made
- % simpler to understand by comments inside the definition. In addition
- % to this, updates are easier and only one source file needs to be
- % changed. On the other hand, because of this, the style files are
- % considerably longer: thus \TeX{} takes longer to load them. If this
- % is a problem, there is an easy remedy: one needs only to run the
- % {\tt docstrip.tex} program that removes nearly all lines that begin
- % with a
- % percent sign.
- %
- % The idea of integrated documentation was born with the development
- % of the \TeX{} program; it was crystallized in Pascal with the \Web{}
- % system. The advantages of this method are plain to see (it's easy
- % to make comparisons \cite{art:Knuthliterat}). Since this
- % development, systems similar to \Web{} have been developed for other
- % programming languages. But for one of the most complicated
- % programming languages (\TeX) the documentation has however been
- % neglected. The \TeX{} world seems to be divided between:---
- % \begin{itemize} \item a couple of ``wizards'', who produce many
- % lines of completely unreadable code ``off the cuff'', and \item many
- % users who are amazed that it works just how they want it to do. Or
- % rather, who despair that certain macros refuse to do what is
- % expected of them.\end{itemize}
- %
- % I do not think that the \Web{} system is {\em the\/} reference work;
- % on the contrary, it is a prototype which suffices for the
- % development of programs within the \TeX{} world. It is sufficient,
- % but not totally sufficient.\footnote{I know that this will be seen
- % differently by a few people, but this product should not be seen as
- % the finished product, at least as far as applications concerning
- % \TeX{} are concerned. The long-standing debate over `multiple
- % change files' shows this well.} As a result of \Web, new programming
- % perspectives have been demonstrated; unfortunately, though, they
- % haven't been developed further for other programming languages.
- %
- % The method of documentation of \TeX{} macros which I have introduced
- % here should also only be taken as a first sketch. It is designed
- % explicitly to run under \LaTeX{} alone. Not because I was of the
- % opinion that this was the best starting point, but because from this
- % starting point it was the quickest to develop.\footnote{This
- % argument is a bad one, however, it is all too often trotted out.} As
- % a result of this design decision, I had to move away from the
- % concept of modularization; this was certainly a step backward.
- %
- % I would be happy if this article could spark off discussion over
- % \TeX\ documentation. I can only advise anyone who thinks that they
- % can cope without documentation to ``Stop Time'' until he or she
- % completely understands the \AmSTeX{} source code.
- %
- %
- %
- %
- %
- % \subsection{Using the {\sf doc} style option}
- %
- % Just like any other option, invoke it by including it amongst the
- % style options in the optional parameter list for the
- % \verb+\documentstyle+ command. {\sf Doc}'s use of
- % \verb+\reversemarginpars+ may make it incompatible with some style
- % options.
- % \changes{v1.7a}{92/02/25}{Altered usage info}
- %
- % \ifmulticols\end{multicols}\fi
- %
- %
- % \section{The User Interface}\label{sec:interface}
- % \subsection{The driver file}
- %
- % If one is going to document a set of macros with the {\tt doc}
- % option one has to prepare a special driver file which produces the
- % formatted document. This driver file has the following
- % characteristics:
- %
- % \noindent \verb+\documentstyle[+\meta{options}]^^A
- % \verb+{+\meta{document-style}\verb+}+\\[3pt]
- % \hspace*{10pt}\meta{preamble}\\[3pt]
- % \verb+\begin{document}+\\[3pt]
- % \hspace*{10pt}\meta{special input commands}\\[3pt]
- % \verb+\end{document}+
- %
- % The list of \meta{options} must contain the {\tt doc} option but it
- % is not necessary any longer to put this option on the end of the
- % list.
- %
- % The \meta{document-style} might be any document style, I normally
- % use {\tt article}.
- %
- % In the \meta{preamble} one should place declarations which
- % manipulate the behavior of the {\tt doc} option like
- % \verb+\DisableCrossrefs+ or \verb+\OnlyDescription+.
- %
- % \DescribeMacro\DocInput \DescribeMacro\IndexInput
- % Finally the \meta{special input commands} part should contain one or
- % more \verb+\DocInput+\meta{file name} and/or
- % \verb+\IndexInput+\meta{file name} commands. The
- % \verb+\DocInput+ command is used for files prepared for the {\tt
- % doc} option whereas \verb+\IndexInput+ can be used for all kinds of
- % macro files. See page \pageref{..Input} for more details of
- % "\IndexInput". Multiple "\DocInput"s can be used with a
- % number of included files which are each self-contained
- % self-documenting styles---for instance, each containing
- % "\maketitle".
- %
- % As an example, the driver file for the {\tt doc} option itself is
- % derived from the following text. This is meant to be extracted by
- % the {\tt docstrip} program which will remove the leading "<+driver>"
- % indicating the `module'; the line numbers are added by {\tt
- % doc}'s formatting.
- % \changes{v1.7a}{92/03/06}{Added docstrip-derivable driver file as
- % example.}
- % \changes{v1.7c}{92/04/01}{Expurgated ltugboat.sty from driver.}
- % \begin{macrocode}
- %<+driver>\documentstyle[doc]{article}
- %<+driver>% dimensions from ltugboat.sty:
- %<+driver>\setlength\textwidth{31pc} \setlength\textheight{54pc}
- %<+driver>\setlength{\parindent}{0pt}
- %<+driver>\setlength{\parskip}{2pt plus 1pt minus 1pt}
- %<+driver>\setlength{\oddsidemargin}{8pc}
- %<+driver>\setlength{\marginparwidth}{8pc}
- %<+driver>\setlength{\topmargin}{-2.5pc}
- %<+driver>\setlength{\headsep}{20pt}
- %<+driver>\setlength{\columnsep}{1.5pc}
- %<+driver>\setlength{\columnwidth}{18.75pc}
- %<+driver>\EnableCrossrefs
- %<+driver>%\DisableCrossrefs % Say \DisableCrossrefs if index is ready
- %<+driver>\RecordChanges % Gather update information
- %<+driver>\CodelineIndex % Index code by line number
- %<+driver>%\OnlyDescription % comment out for implementation details
- %<+driver>%\OldMakeindex % use if your MakeIndex is pre-v2.9
- %<+driver>\begin{document}
- %<+driver> \DocInput{doc.doc}
- %<+driver>\end{document}
- % \end{macrocode}
- %^^A As an example this is the driver file for the {\tt doc} option
- %^^A itself:
- %^^A\begin{verbatim}
- %^^A\documentstyle[ltugboat,doc]{article}
- %^^A
- %^^A % \DisableCrossrefs % Say \DisableCrossrefs if the index
- %^^A % is ready.
- %^^A
- %^^A \RecordChanges % Gather update-information
- %^^A
- %^^A % \OnlyDescription % Remove the comment char if you only want
- %^^A % the description
- %^^A
- %^^A\SelfDocumenting % Tugbot style macro (for doc)
- %^^A
- %^^A\begin{document}
- %^^A
- %^^A \DocInput{doc.doc}
- %^^A
- %^^A\end{document}
- %^^A\end{verbatim}
- %
- % \changes{v1.7a}{92/02/25}{Note on avoiding driver file}
- % \subsection{Avoiding using a driver file}
- %
- % It is possible to dispense with a driver file at the expense of some
- % flexibility using the following trick.\footnote{Due to David
- % Carlisle, Manchester University.} Before the documentation part of
- % the file starts, include (without leading "%"s!)\ something like
- % \begin{verbatim}
- %\ifcat a\noexpand @\let\next\relax\else\def\next{%
- % \documentstyle[doc]{article}\MakePercentIgnore}\fi\next
- %\end{verbatim}
- % with any necessary extra style options included, of course. Then
- % the file can either be used directly as a style file as normal or
- % have the documentation printed simply by using
- % "latex"~\meta{filename}. You can also arrange to have this process
- % ask questions about whether to format only the usage section, for
- % instance.
- %
- % \subsection{General conventions}
- %
- % A \TeX{} file prepared to be used with the `doc' style option
- % consists of `documentation parts' intermixed with `definition
- % parts'.
- %
- % Every line of a `documentation part' starts with a percent sign
- % (\verb+%+) in column one. It may contain arbitrary \TeX{} or
- % \LaTeX{} commands except that the character `\verb+%+' cannot be
- % used as a comment character.
- % \SortIndex{\string^\string^A}{\string\verb\verbatimchar
- % \string^\string^A\verbatimchar \encapchar usage} To allow user
- % comments, the \verb+^^A+ character is defined as a comment character
- % later on. Such `metacomments' may be also be included simply by
- % surrounding them with "\iffalse" \ldots~"\fi".
- %
- % All other parts of the file are called `definition parts'. They
- % contain fractions of the macros described in the `documentation
- % parts'.
- %
- % If the file is used to define new macros (e.g.\ as a style file in
- % the \verb+\documentstyle+ macro), the `documentation parts' are
- % bypassed at high speed and the macro definitions are pasted
- % together, even if they are split into several `definition parts'.
- %
- % \DescribeEnv{macrocode}
- % On the other hand, if the documentation of these macros is to be
- % produced, the `definition parts' should be typeset verbatim. To
- % achieve this, these parts are surrounded by the {\sf macrocode}
- % environment.
- % More exactly: before a `definition part' there should be a line
- % containing\\
- % \hspace*{\MacroIndent}\verb*+% \begin{macrocode}+\\
- % and after this part a line\\
- % \hspace*{\MacroIndent}\verb*+% \end{macrocode}+\\
- % There must be {\em exactly\/} four spaces between the \verb+%+
- % and \verb+\end{macrocode}+ --- \TeX{} is looking for this string
- % and not for the macro while processing a `definition part'.
- %
- % Inside a `definition part' all \TeX{} commands are allowed; even the
- % percent sign could be used to suppress unwanted spaces etc.
- %
- % \DescribeEnv{macrocode*}
- % Instead of the {\sf macrocode} environment one can also use the {\sf
- % macrocode$*$} environment which produces the same results except
- % that spaces are printed as \nopagebreak\verb*+ + characters.
- %
- %
- %
- % \subsection{Describing the usage of new macros}
- %
- % \DescribeMacro\DescribeMacro
- % When you describe a new macro you may use \verb+\DescribeMacro+ to
- % indicate that at this point the usage of a specific macro is
- % explained. It takes one argument which will be printed in the margin
- % and also produces a special index entry. For example, I used
- % \verb+\DescribeMacro{\DescribeMacro}+ to make clear that this is the
- % point where the usage of \verb+\DescribeMacro+ is explained.
- %
- % \DescribeMacro\DescribeEnv
- % An analogous macro \verb+\DescribeEnv+ should be used to indicate
- % that a \LaTeX{} environment is explained. It will produce a somewhat
- % different index entry. Below I used \verb+\DescribeEnv{verbatim}+.
- %
- % \DescribeEnv{verbatim}
- % It is often a good idea to include examples of the usage of new macros
- % in the text. Because of the \verb+%+ sign in the first column of every
- % row, the {\sf verbatim} environment is slightly altered to suppress
- % those
- % characters.\footnote{These macros were written by Rainer
- % Sch\"opf~\cite{art:verbatim}. He also
- % provided a new {\sf verbatim} environment which
- % can be used inside of other macros.}
- % \DescribeEnv{verbatim*}
- % The {\sf verbatim$*$} environment is changed in the same way.
- % \changes{v1.7a}{92/02/26}{Documented `verb change.}
- % \DescribeMacro\verb
- % The "\verb" command is re-implemented to give an error report if a
- % newline appears in its argument.
- % The {\sf verbatim} and {\sf verbatim$*$} environments set text in
- % the style defined by "\MacroFont"~(\S\ref{sec:macrofont}).
- %
- %
- %
- % \subsection{Describing the definition of new macros}
- %
- % \DescribeEnv{macro}
- % To describe the definition of a new macro we use the {\sf macro}
- % environment. It has one argument: the name of the new
- % macro.\footnote{This is a change to the style design I described in
- % ^^A \TUB ^^A removed in case ltugboat.sty not used
- % {\sl TUGboat\/}\ 10\#1 (Jan.~89). We finally decided
- % that it would
- % be better to use the macro name {\em with\/} the
- % backslash as an argument.}
- % This argument is also used to print the name in the margin and to
- % produce an index entry.
- % Actually the index entries for usage and definition are different to
- % allow an easy reference.
- % This environment might be nested. In this case the
- % labels in the margin are placed under each other.
- % \changes{v1.7a}{92/02/26}{Note on need for some text in macro env.}
- % There should be some text---even if it's just an empty
- % "\mbox{}"---in this environment before "\begin{macrocode}" or the
- % marginal label won't print in the right place.
- %
- % \DescribeMacro\MacrocodeTopsep
- % \DescribeMacro\MacroTopsep
- % There also exist four style parameters: \verb+\MacrocodeTopsep+ and
- % \verb+\MacroTopsep+ are used to control the vertical spacing above
- % and below the {\sf macrocode} and the {\sf macro}
- % \DescribeMacro\MacroIndent
- % environment, \verb+\MacroIndent+ is used to indent the lines of code
- % and
- % \DescribeMacro\MacroFont \label{sec:macrofont}
- % \verb+\MacroFont+ holds the font and a possible size change command
- % for the code lines, the "verbatim"["*"] environment and the macro
- % names printed in the margin. If you want
- % to change their default values in a
- % style file (like {\tt ltugboat.sty}) use the \verb+\DocstyleParms+
- % command described below.
- %
- %
- %
- %
- % \subsection{Formatting the margins}
- %
- % \DescribeMacro\PrintDescribeMacro
- % \DescribeMacro\PrintDescribeEnv
- % \DescribeMacro\PrintMacroName
- % As mentioned earlier, some macros and the {\sf macro} environment
- % print their arguments in the margin. This is actually done by three
- % macros which are user
- % definable.\footnote{You may place the changed definitions in a
- % separate style
- % file or at the beginning of the documentation
- % file.
- % For example, if you don't like any names in the
- % margin
- % but want a fine index you can simply
- % {\tt \bslash let}
- % these macros equal {\tt \bslash @gobble}.
- % The doc style option won't redefine any existing
- % definitions of these macros.}
- % They are named \verb+\PrintDescribeMacro+, \verb+\PrintDescribeEnv+
- % and \verb+\PrintMacroName+ (called by the {\sf macro} environment).
- %
- %
- % \subsection{Using a special escape character}
- %
- % \DescribeMacro\SpecialEscapechar
- % If one defines complicated macros it is sometimes necessary to
- % introduce a new escape character because the `\verb+\+' has got a
- % special \verb+\catcode+. In this case one can use
- % \verb+\SpecialEscapechar+ to indicate which character is actually
- % used to play the r\^ole of the `\verb+\+'. A scheme like this is
- % needed because the {\sf macrocode} environment and its counterpart
- % {\sf macrocode$*$} produce an index entry for every occurrence of a
- % macro name. They would be very confused if you didn't tell them that
- % you'd changed \verb+\catcode+$\,$s. The argument to
- % \verb+\SpecialEscapechar+ is a single-letter control sequence, that
- % is, one has to use \verb+\|+ for example to denote that `\verb+|+'
- % is used as an escape character. \verb+\SpecialEscapechar+ only
- % changes the behavior of the next {\sf macrocode} or {\sf
- % macrocode$*$} environment.
- %
- % The actual index entries created will all be printed with \verb+\+
- % rather than \verb+|+, but this probably reflects their usage, if not
- % their definition, and anyway must be preferable to not having any
- % entry at all. The entries {\em could\/} be formatted appropriately,
- % but the effort is hardly worth it, and the resulting index might be
- % more confusing (it would certainly be longer!).
- %
- %
- % \subsection{Cross-referencing all macros used}
- %
- % \DescribeMacro\DisableCrossrefs
- % \DescribeMacro\EnableCrossrefs
- % As already mentioned, every new macro name used within a {\sf
- % macrocode} or {\sf macrocode$*$} environment will produce an index
- % entry. In this way one can easily find out where a specific macro is
- % used. Since \TeX{} is considerably slower when it has to produce
- % such a bulk of index entries one can turn off this feature by using
- % \verb+\DisableCrossrefs+ in the driver file. To turn it on again
- % just use
- % \verb+\EnableCrossrefs+.\footnote{Actually,
- % {\tt\bslash EnableCrossrefs}
- % changes things more drastically; any following
- % {\tt\bslash DisableCrossrefs}
- % which might be present in the source will be ignored.}
- %
- %
- % \DescribeMacro\DoNotIndex
- % But also finer control is provided. The \verb+\DoNotIndex+ macro
- % takes a list of macro names separated by commas. Those names won't
- % show up in the index. You might use several \verb+\DoNotIndex+
- % commands: their lists will be concatenated. In this article I used
- % \verb+\DoNotIndex+ for
- % all macros which are already defined in \LaTeX.
- %
- % All three above declarations are local to the current group.
- %
- % Production (or not) of the index (via the "\makeindex" commend) is
- % controlled by using or omitting the following declarations in the
- % driver file preamble; if neither is used, no index is produced.
- % \DescribeMacro\PageIndex Using "\PageIndex" makes all index
- % entries refer to their page number; with
- % \DescribeMacro\CodelineIndex "\CodelineIndex", index entries
- % produced by "\DescribeMacro" and "\DescribeEnv" refer to page number
- % but those produced by the {\sf macro} environment refer to the
- % code lines, which will be numbered automatically.\footnote{The line
- % number is actually that of the first line of the first {\sf
- % macrocode} environment in the {\sf macro} environment.}
- % \DescribeMacro\theCodelineNo
- % The style of this numbering can be controlled by defining the macro
- % "\theCodelineNo". Its default definition is to use scriptsize
- % arabic numerals; a user-supplied definition won't be overwritten.
- %
- %
- % \subsection{Producing the actual index entries}
- %
- % Several of the aforementioned macros will produce some sort of index
- % entries. These entries have to be sorted by an external
- % program---the current implementation assumes that the {\sf
- % makeindex} program by Chen~\cite{art:Chen} is used.
- %
- % But this isn't built in: one has only to redefine some of the
- % following macros to be able to use any other index program. All
- % macros which are installation
- % dependent are defined in such a way that they won't overwrite a
- % previous definition. Therefore it is safe to put the changed
- % versions in a style file which might be read in before the doc style
- % option.
- %
- % To allow the user to change the specific characters recognized by
- % his or her index program all characters which have special meaning
- % in the {\sf makeindex} program are given symbolic
- % names.\footnote{I don't know if there exists a program which needs
- % more command characters, but I hope not.}
- % However, all characters used should be of \verb+\catcode+ other than
- % `letter' (11).
- %
- % \DescribeMacro{\actualchar}
- % The \verb+\actualchar+ is used to separate the `key' and the actual
- % index entry.
- % \DescribeMacro{\quotechar}
- % The \verb+\quotechar+ is used before a special index program
- % character to suppress its special meaning.
- % \DescribeMacro{\encapchar}
- % The \verb+\encapchar+ separates the indexing information from a
- % letter string which {\sf makeindex} uses as a \TeX{} command to
- % format the page number associated with a special entry. It is used
- % in this style to apply the \verb+\main+ and the \verb+\usage+
- % commands.
- % \DescribeMacro{\levelchar}
- % Additionally \verb+\levelchar+ is used to separate `item',
- % `subitem' and `subsubitem' entries.
- %
- % It is a good idea to stick to these symbolic names even if you know
- % which index program is used. In this way your files will be
- % portable.
- %
- % \DescribeMacro\SpecialMainIndex
- % To produce a main index entry for a macro the
- % \verb+\SpecialMainIndex+ macro\footnote{This macro is called by the
- % {\sf macro} environment.} may be used. It is called `special'
- % because it has to print its argument verbatim.
- % \DescribeMacro\SpecialIndex
- % If you want a normal index entry for a macro name
- % \verb+\SpecialIndex+ might be used.\footnote{This macro is called
- % within the {\sf macrocode} environment when encountering a macro
- % name.}
- % \DescribeMacro\SpecialUsageIndex
- % \DescribeMacro\SpecialEnvIndex
- % To index the usage of a macro or an environment
- % \verb+\SpecialUsageIndex+ and \verb+\SpecialEnvIndex+ may be used.
- % \DescribeMacro\SortIndex
- % Additionally a \verb+\SortIndex+ command is provided. It takes two
- % arguments---the sort key and the actual index entry.
- %
- % All these macros are normally used by other macros; you will need
- % them only in an emergency.
- %
- % \DescribeMacro\verbatimchar
- % But there is one characteristic worth mentioning: all macro names in
- % the index are typeset with the \verb+\verb*+ command. Therefore one
- % special character is needed to act as a delimiter for this command.
- % To allow a change in this respect, again this character is
- % referenced indirectly, by the macro \verb+\verbatimchar+. It expands
- % by default to \verb?+? but if your code lines contain macros with
- % `{\tt +}' characters in their names (e.g.\ when you use \verb?\+?)
- % you will end up with an index entry containing \verb?\verb+\++?
- % which will be typeset as `\verb+\++' and not as `\verb?\+?'. In this
- % case you should redefine \verb+\verbatimchar+ globally or locally to
- % overcome this problem.
- %
- % \DescribeMacro\*
- % We also provide a \verb+\*+ macro. This is intended to be used for
- % index entries like
- % \begin{quote}
- % index entries \\
- % \hspace*{30pt} Special macros for \*
- % \end{quote}
- % Such an entry might be produced with the line:
- %\begin{verbatim}
- % \index{index entries\levelchar Special macros for \*}
- %\end{verbatim}
- %
- % \DescribeMacro\OldMakeindex
- % Versions of {\sf makeindex} prior to 2.9 had some bugs affecting
- % {\sf doc}. One of these,
- % pertaining to the "%" character doesn't have a work-around
- % appropriate for versions with and without the
- % bug.\label{makeindex:version} If
- % you have an old version, invoke "\OldMakeindex" in a
- % style file or the driver file to prevent problems with index entries
- % such as "\%", although you'll probably normally want to turn off
- % indexing of "\%" anyway. Try to get an up-to-date {\sf makeindex}
- % from one of the \TeX{} repositories.
- %
- %
- % \subsection{Setting the index entries}
- %
- % \changes{v1.7a}{92/03/11}{Usage note on gind.ist.}
- % After the first formatting pass through the {\tt .doc} file you need
- % to sort the index entries written to the {\tt .idx} file using {\sf
- % makeindex} or your favourite alternative. You need a suitable style
- % file for {\sf makeindex} (specified by the {\tt -s} switch). A
- % suitable one is supplied with {\sf doc}, called {\tt gind.ist}.
- %
- % \DescribeMacro\PrintIndex
- % To read in and print the sorted index, just put the
- % \verb+\PrintIndex+ command as the last (commented-out, and thus
- % executed during the documentation pass through the file) command
- % in your style file. Precede it by any bibliography commands
- % necessary for your citations.
- % Alternatively, it may be more convenient to put all such calls
- % amongst the arguments of the \verb+\StopEventually+ macro, in
- % which case a \verb+\Finale+ command should appear at the end of
- % your file.
- %
- % \DescribeEnv{theindex}
- % Contrary to standard \LaTeX, the index is typeset in three columns
- % by default. This is controlled by the \LaTeX{} counter `{\sf
- % IndexColumns}' and can therefore be changed with a
- % \verb+\setcounter+ declaration. Additionally one doesn't want to
- % start a new page unnecessarily. Therefore the {\sf theindex}
- % environment is redefined.
- % \DescribeMacro\IndexMin
- % When the {\sf theindex} environment starts it will measure how much
- % space is left on the current page. If this is more than
- % \verb+\IndexMin+ then the index will start on this page. Otherwise
- % \verb+\newpage+ is called.
- %
- % Then a short introduction about the meaning of several index entries
- % is typeset (still in onecolumn mode). Afterwards the actual index
- % entries follow in multi-column mode.
- % \DescribeMacro\IndexPrologue
- % You can change this prologue with the help of the
- % \verb+\IndexPrologue+ macro. Actually the section heading is also
- % produced in this way, so you'd better write something like:
- % \begin{verbatim}
- % \IndexPrologue{\section*{Index} The index entries underlined ...}
- %\end{verbatim}
- % When the {\sf theindex} environment is finished the last page will
- % be reformatted to produce balanced columns. This improves the layout
- % and allows the next article to start on the same page.
- % \DescribeMacro\IndexParms
- % Formatting of the index columns (values for \verb+\columnssep+
- % etc.)\ is controlled by the \verb+\IndexParms+ macro. It assigns the
- % following values:
- % \SpecialUsageIndex{\parindent}\SpecialUsageIndex{\columnsep}^^A
- % \SpecialUsageIndex{\parskip}\SpecialUsageIndex{\rightskip}^^A
- % \SpecialUsageIndex{\mathsurround}\SpecialUsageIndex{\parfillskip}
- % \begin{center}
- % \begin{tabular}{l@{\,=\,}ll@{\,=\,}l}
- % \verb+\parindent+ & \IndexParms \the\parindent &
- % \verb+\columnsep+ & \IndexParms \the\columnsep \\
- % \verb+\parskip+ & \IndexParms \the\parskip &
- % \verb+\rightskip+ & \IndexParms
- % \expandafter\dimenvalue\the\rightskip \\
- % \verb+\mathsurround+ & \IndexParms \the\mathsurround &
- % \verb+\parfillskip+ & \IndexParms
- % \expandafter\dimenvalue\the\parfillskip
- % \end{tabular}
- % \end{center}
- % \DescribeMacro{\@idxitem}
- % Additionally it defines \verb+\@idxitem+ (which will be used when an
- % \verb+\item+ command is encountered) and selects \verb+\small+ size.
- % If you want to change any of these values you have to define them
- % all.
- %
- % \DescribeMacro\main
- % \DescribeMacro\usage
- % The page numbers for main index entries are encapsulated by the
- % \verb+\main+ macro (underlining its argument) and the numbers
- % denoting the description are encapsulated by the \verb+\usage+ macro
- % (which produces {\em italics}). As usual these commands are user
- % definable.
- %
- %
- % \subsection{Changing the default values of style parameters}
- %
- % \DescribeMacro\DocstyleParms
- % If you want to overwrite some default settings made by the {\tt doc}
- % style, you can either put your declarations in the driver file (that
- % is after {\tt doc.sty} is read in) or use a separate style file for
- % doing this work. In the latter case you can define the macro
- % \verb+\DocstyleParms+ to contain all assignments. This
- % indirect approach is necessary if your style file might be read
- % before the {\tt doc.sty}, when some of the registers are not
- % allocated. Its default definition is null.
- %
- % The doc style option currently assigns values to the following
- % registers:
- % \SpecialUsageIndex{\IndexMin}\SpecialUsageIndex{\MacrocodeTopsep}^^A
- % \SpecialUsageIndex{\MacroTopsep}^^A
- % \SpecialUsageIndex{\MacroIndent}\SpecialUsageIndex{\marginparpush}^^A
- % \SpecialUsageIndex{\marginparwidth}\SpecialUsageIndex{\tolerance}
- % \begin{center}
- % \begin{tabular}{l@{\,=\,}ll@{\,=\,}l}
- % \verb+\IndexMin+ & \the\IndexMin &
- % \verb+\MacroTopsep+ & \the\MacroTopsep \\
- % \verb+\marginparwidth+& \the\marginparwidth &
- % \verb+\MacroIndent+ & \the\MacroIndent \\
- % \verb+\marginparpush+ & \the\marginparpush &
- % \verb+\MacrocodeTopsep+ & \the\MacrocodeTopsep \\
- % \verb+\tolerance+ & \the\tolerance
- % \end{tabular}
- % \end{center}
- %
- %
- % \subsection{Additional bells and whistles}
- %
- % We provide macros for logos such as \Web, \AmSTeX, \BibTeX,
- % \SliTeX{} and \PlainTeX. Just type \verb+\Web+, \verb+\AmSTeX+,
- % \verb+\BibTeX+, \verb+\SliTeX+ or \verb+\PlainTeX+, respectively.
- % \LaTeX{} and \TeX{} are already defined in {\tt latex.tex}.
- %
- % \DescribeMacro\meta
- % Another useful macro is \verb+\meta+ which has one argument and
- % produces something like \meta{dimen parameter}.
- %
- % \DescribeMacro\OnlyDescription
- % \DescribeMacro\StopEventually
- % You can use the \verb+\OnlyDescription+ declaration in the driver
- % file to suppress the last part of your document (which presumably
- % exhibits the code). To make this work
- % you have to place the command \verb+\StopEventually+ at a suitable
- % point in your file. This macro has one argument in which you put
- % all information you want to see printed if your document ends at
- % this point (for example a bibliography which is normally printed at
- % the very end). When the \verb+\OnlyDescription+ declaration is
- % missing the \verb+\StopEventually+
- % \DescribeMacro\Finale
- % macro saves its argument in a macro called \verb+\Finale+ which can
- % afterwards be used to get things back (usually at the very end).
- % Such a scheme makes changes in two places unnecessary.
- %
- % Thus you can use this feature to produce a local guide for the
- % \TeX{} users which describes only the usage of macros (most of them
- % won't be interested in your definitions anyway). For the same
- % reason the \verb+\maketitle+
- % \DescribeMacro\maketitle
- % command is slightly changed to allow multiple titles in one
- % document. So you can make one driver file reading in several
- % articles at once.
- % \DescribeMacro{\ps@titlepage}
- % To avoid an unwanted {\sf pagestyle} on the title page the
- % \verb+\maketitle+ command issues a \verb+\thispagestyle{titlepage}+
- % declaration which produces a {\sf plain} page if the {\sf titlepage}
- % page style is undefined. This allows style files like {\sf
- % ltugboat.sty} to define their own page styles for title pages.
- %
- % \DescribeMacro\IndexInput \label{..Input}
- % Last but not least I defined an \verb+\IndexInput+ macro which
- % takes a file name as an argument and produces a verbatim listing of
- % the file, indexing every command as it goes along. This might be
- % handy, if you want to learn something about macros without enough
- % documentation. I used this feature to cross-reference {\tt
- % latex.tex} getting a verbatim copy with about 15 pages
- % index.\footnote{It took quit a long time and the resulting {\tt .idx}
- % file was longer than the {\tt .dvi} file.
- % Actually too long to be handled by the {\sf makeindex}
- % program directly (on our MicroVAX) but the final result
- % was worth the trouble.}
- %
- % \DescribeMacro\changes
- % To maintain a change history within the file, the \verb+\changes+
- % command may be placed amongst the description part of the changed
- % code. It takes three arguments, thus:
- % \begin{quote}
- % \verb+\changes{+\meta{version}\verb+}{+\meta{date}\verb+}{+^^A
- % \meta{text}\verb+}+
- % \end{quote}
- % The changes may be used to produce an auxiliary file (\LaTeX's
- % \verb+\glossary+ mechanism is used for this) which may be printed
- % after suitable formatting. The \verb+\changes+ macro encloses the
- % \meta{date} in parentheses and appends the \meta{text} to form the
- % printed entry in such a change history; because old
- % versions\footnote{Before 2.6.} of the {\sf makeindex}
- % program limit such fields to 64 characters, care should be taken
- % not to exceed this limit when describing the change. When
- % referring to macros in change descriptions it is conventional to use
- % "`"\meta{macroname} rather than attempting to format it properly and
- % using up valuable characters in the entry with old {\sf makeindex}
- % versions.
- %
- % \changes{v1.7a}{92/02/26}{Description of `RecordChanges etc. added
- % to interface section.}
- % \DescribeMacro\RecordChanges
- % To cause the change information to be written out, include
- % "\RecordChanges" in the driver file.
- % \DescribeMacro\PrintChanges
- % To read in and print the sorted change history (in two columns),
- % just put the \verb+\PrintChanges+ command as the last
- % (commented-out, and thus executed during the documentation pass
- % through the file) command in your style file. Alternatively, this
- % command may form one of the arguments of the \verb+\StopEventually+
- % command, although a change history is probably {\em not\/} required
- % if only the description is being printed.
- % The command assumes that {\sf makeindex} or some other program
- % has processed the {\tt.glo} file to generate a sorted {\tt.gls} file.
- % You need a special {\sf makeindex} style file; a suitable one is
- % supplied with {\sf doc}, called {\tt gglo.ist}.
- % \DescribeMacro\GlossaryMin \DescribeMacro\GlossaryPrologue
- % \DescribeMacro\GlossaryParms
- % The "\GlossaryMin", "\GlossaryPrologue" and "\GlossaryParms" macros
- % are analagous to the "\Index"\ldots\ versions. (The \LaTeX{}
- % `glossary' mechanism is used for the change entries.)
- %
- % \label{sec:checksum}
- % \DescribeMacro\CharacterTable
- % \DescribeMacro\CheckSum
- % To overcome some of the problems of sending files over the networks
- % we developed two macros which should detect corrupted files. If one
- % places the lines
- % \begin{verbatim}
- %%% \CharacterTable
- %%% {Upper-case \A\B\C\D\E\F\G\H\I\J\K\L\M\N\O\P\Q\R\S\T\U\V\W\X\Y\Z
- %%% Lower-case \a\b\c\d\e\f\g\h\i\j\k\l\m\n\o\p\q\r\s\t\u\v\w\x\y\z
- %%% Digits \0\1\2\3\4\5\6\7\8\9
- %%% Exclamation \! Double quote \" Hash (number) \#
- %%% Dollar \$ Percent \% Ampersand \&
- %%% Acute accent \' Left paren \( Right paren \)
- %%% Asterisk \* Plus \+ Comma \,
- %%% Minus \- Point \. Solidus \/
- %%% Colon \: Semicolon \; Less than \<
- %%% Equals \= Greater than \> Question mark \?
- %%% Commercial at \@ Left bracket \[ Backslash \\
- %%% Right bracket \] Circumflex \^ Underscore \_
- %%% Grave accent \` Left brace \{ Vertical bar \|
- %%% Right brace \} Tilde \~}
- %%%
- %\end{verbatim}
- % at the beginning of the file then character translation failures
- % will be detected, provided of course, that the used {\tt doc} option
- % has a correct default table. The percent signs\footnote{There are
- % two percent signs in each line. This has the effect that these lines
- % are not removed by the {\tt docstrip.tex} program.} at the beginning
- % of the lines should be typed in, since only the {\tt doc} option
- % should look at this command.
- %
- % Another problem of mailing files is possible truncation. To
- % detect these sort of errors we provide a \verb+\CheckSum+ macro.
- % The check-sum of a file is simply the number of backslashes in the
- % code, i.e.\ all lines between the {\sf macrocode} environments. But
- % don't be afraid: you don't have count the code-lines yourself; this
- % is done by the {\tt doc} style option for you. You simply have to
- % use the \verb+\StopEventually+ (which starts looking for backslashes)
- % and the \verb+\Finale+ command. The latter will inform you either
- % that your file has no check-sum (telling you the right number) or
- % that your number is incorrect (this time telling you both the
- % correct and the incorrect one). Then you go to the top of your file
- % inserting the line
- % \begin{quote}
- % \verb+%% \CheckSum{+\meta{number}\verb+}+
- % \end{quote}
- % and that's all. If you precede it only with one percent then the
- % line will not show up in {\tt docstrip} versions of the file.
- % You should do so whenever you are using conditional code (see
- % {\tt docstrip} documentation) since then the check-sum will not
- % reflect the number of backslashes in the stripped of versions.
- %
- % \DescribeMacro\bslash
- % From time to time, it is necessary to print a \verb+\+ without
- % being able to use the \verb+\verb+ command because the
- % \verb+\catcode+$\,$s of the symbols are already firmly
- % established. In this instance we can use the command
- % \verb+\bslash+ presupposing, of course, that the actual font in
- % use at this point contains a `backslash' as a symbol. Note that
- % this definition of \verb+\bslash+ is expandable; it inserts a
- % $"\"_{12}$. This means that you have to \verb+\protect+
- % it if it is used in `moving arguments'.
- %
- % \DescribeMacro\MakePrivateLetters
- % \changes{v1.7a}{92/02/26}{Documented `MakePrivateLetters in
- % interface section}^^A
- % If your macros "\catcode" anything other than "@" to `letter', you
- % should redefine "\MakePrivateLetters" so that it also makes the
- % relevant characters `letters' for the benefit of the indexing. The
- % default definition is just "\makeatletter".
- %
- % \DescribeMacro\MakeShortVerb \DescribeMacro\DeleteShortVerb
- % It is awkward to have to type, say, "\verb+"\ldots"+" continually
- % when quoting verbatim bits (like macro names) in the text, so an
- % abbreviation mechanism is provided. Pick a character
- % \meta{c}---one which normally has catcode `other' unless you have
- % very good reason not to---which
- % you don't envisage using in the text, or not using often. (I like
- % \verb+"+, but you may prefer "|" if you have \verb+"+ active to do
- % umlauts, for instance.) Then if you say
- % "\MakeShortVerb{\"\meta{c}"}" you can subsequently use
- % \meta{c}\meta{text}\meta{c} as the equivalent of
- % "\verb"\meta{c}\meta{text}\meta{c}. Use
- % "\DeleteShortVerb{\"\meta{c}"}" if you subsequently
- % want \meta{c} to revert to its previous meaning---you can
- % always turn it on again after the unusual section. The `short verb'
- % commands make global changes. The abbreviated "\verb" may not
- % appear in the argument of another command just like "\verb".
- % However the `short verb' character may be used freely in the {\sf
- % verbatim} and {\sf macrocode} environments without ill effect.
- % "\DeleteShortVerb" is silently ignored if its argument does not
- % currently represent a short verb character. Both commands type a
- % message to tell you the meaning of the character is being changed.
- %
- % \DescribeMacro\DontCheckModules \DescribeMacro\CheckModules
- % \DescribeMacro\Module \DescribeMacro\AltMacroFont
- % The `module' directives of the {\sf docstrip} system
- % \cite{art:docstrip} are normally
- % recognised and invoke special formatting. This can be turned on and
- % off in the {\tt .doc} file or the driver file using "\CheckModules"
- % and "\DontCheckModules". If checking for module directives is on
- % (the default) then code in the scope of the directives is set as
- % determined by the hook "\AltMacroFont", which gives {\small\tt\it
- % small italic type\-writer\/} by default in the New Font
- % Selection Scheme but just ordinary {\small\tt small type\-writer} in
- % the old one, where a font such as italic typewriter can't be used
- % portably (plug for NFSS); you will need to override
- % this if you don't have the italic typewriter font available.
- % Code is in such a scope if it's on a line beginning with "%<" or is
- % between lines starting with "%<*"\meta{name list}">" and
- % "%</"\meta{name list}">". The directive is formatted by the macro
- % "\Module" whose single argument is the text of the
- % directive between, but not including, the angle brackets; this macro
- % may be re-defined in the driver or style file and by default
- % produces results like \Module{+foo|bar} with no following space.
- %
- % \DescribeMacro{StandardModuleDepth} Sometimes (as in this file) the
- % whole code is surrounded by modules to produce several files from a
- % single source. In this case it is clearly not appropriate to format
- % all code lines in a special "\AltMacroFont". For this reason a
- % counter "StandardModuleDepth" is provided which defines the level of
- % module nesting which is still supposed to be formatted in
- % "\MacroFont" rather then "\AltMacroFont". The default setting is
- % "0", for this documentation it was set to
- %\begin{verbatim}
- % \setcounter{StandardModuleDepth}{1}
- %\end{verbatim}
- % at the beginning of the file.
- %
- %
- % \subsection{Basic usage summary}
- % \changes{v1.7a}{92/03/11}{Added basic usage summary to spell it out.}
- %
- % To sum up, the basic structure of a {\tt .doc} file without any
- % refinements is like this:
- % \begin{verse}\small
- % "% "\meta{waffle}\ldots\\
- % \quad\ldots \\
- % "% \DescribeMacro{\fred}"\\
- % "% "\meta{description of fred's use}\\
- % \quad\ldots\\
- % "% \StopEventually{"\meta{finale code}"}"\\
- % \quad\ldots\\
- % "% \begin{macro}{\fred}"\\
- % "% "\meta{commentary on macro fred}\\
- % \verb*+% \begin{macrocode}+\\
- % \meta{code for macro fred}\\
- % \verb*+% \end{macrocode}+\\
- % "% \end{macro}"\\
- % \quad\ldots\\
- % "% \Finale \PrintIndex \PrintChanges"
- % \end{verse}
- % For examples of the use of most---if not all---of the features
- % described above consult the {\tt doc.doc} source itself.
- %
- % \subsection{Acknowledgements}
- %
- % I would like to thank all folks at Mainz and at the Royal Military
- % College of Science for their help in this project. Especially Brian
- % and Rainer who pushed everything with their suggestions, bug fixes,
- % etc.
- %
- % A big thank you to David Love who brought the documentation
- % up-to-date again, after I neglected this file for more than two
- % years. This was most certainly a tough job as many features added to
- % {\sf doc.doc} after its publication in {\sl TUGboat\/} have been never
- % properly described. Beside this splendid work he kindly provided
- % additional code (like ``docstrip'' module formatting) which I think
- % every {\sf doc.doc} user will be grateful for.
- %
- %
- % \StopEventually{
- % \begin{thebibliography}{1}
- % \bibitem{book:Buerger} {\sc G. A. B\"urger}.
- % \newblock Wunderbare Reisen zu Wasser und zu Lande, Feldz\"uge
- % und lustige Abenteuer des Freyherrn v.\ M\"unchhausen.
- % \newblock London, 1786 \& 1788.
- % \bibitem{art:Knuthliterat} {\sc D. E. Knuth}.
- % \newblock Literate Programming.
- % \newblock Computer Journal, Vol.~27, {\it pp}.~97--111, May 1984.
- % \bibitem{book:KnuthA} {\sc D. E. Knuth}.
- % \newblock Computers \& Typesetting (The \TeX book).
- % \newblock Addison-Wesley, Vol. A, 1986.
- % \bibitem{art:Chen} {\sc L. Lamport}.
- % \newblock MakeIndex: An Index Processor for \LaTeX.
- % \newblock 17 February 1987.
- % \newblock (Taken from the file {\tt makeindex.tex} provided with
- % the program source code.)
- % \bibitem{art:doc} {\sc Frank Mittelbach}.
- % \newblock The {\tt doc}-option.
- % \newblock {\sl TUGboat}, Vol.~10(2), {\it pp}.~245--273, July
- % 1989.
- % \bibitem{art:docstrip} {\sc Frank Mittelbach, Denys Duchier and
- % Johannes Braams}.
- % \newblock {\tt docstrip.doc} (to appear).
- % \newblock The file is part of the DOC package.
- % \bibitem{book:Raspe} {\sc R. E. Raspe} (*1737, \dag 1797).
- % \newblock Baron M\"unchhausens narrative of his marvellous
- % travels and campaigns in Russia.
- % \newblock Oxford, 1785.
- % \bibitem{art:verbatim} {\sc Rainer Sch\"opf}.
- % \newblock A New Implementation of \LaTeX's {\tt verbatim} and
- % {\tt verbatim*} Environments.
- % \newblock File {\tt verbatim.doc}, version 1.4i.
- % \end{thebibliography}
- % ^^A\PrintIndex
- % ^^A\PrintChanges
- %
- % \ifmulticols
- % \addtocontents{toc}{\protect\end{multicols}}
- % \fi
- %
- % } ^^A end \StopEventually
- %
- %
- % \section{The Description of Macros}
- %
- % Most of the following code is destined for {\tt doc.sty} after
- % processing with {\tt docstrip} to include the module {\bf style}
- % indicated here. (All code in this file not appropriate to {\tt
- % doc.sty} has to be included explicitly by docstrip so that this {\tt
- % .doc} file can be used as directly as a style file rather than the
- % stripped version.) The usual font change for the
- % conditionally-included lines between the \Module{*style} and
- % \Module{/style} directives is suppressed since only the lines with
- % an explicit directive are special in this file.
- % \begin{macrocode}
- %<*style>
- % \end{macrocode}
- % As always, we begin by identifying the latest version of this file
- % on the VDU and in the {\sf log} file. But only if the macros are
- % unknown to the system.
- % \changes{v1.5i}{89/06/07}{Avoid reading the file twice.}
- % \begin{macrocode}
- \@ifundefined{macro@cnt}{}{\endinput} \typeout{Style-Option: `doc'
- \fileversion \@spaces\space\space <\filedate> (FMi)} \typeout{English
- Documentation \@spaces\@spaces\space <\docdate> (DLo, FMi, RMCS)}
- % \end{macrocode}
- % \DescribeMacro\fileversion
- % \DescribeMacro\filedate
- % \DescribeMacro\docdate
- % As you can see I used macros like \verb+\fileversion+ to denote the
- % version number and the date. They are defined at the very beginning
- % of the style file (without a surrounding {\sf macrocode}
- % environment), so I don't have to search for this place here when I
- % change the version number. You can see their actual outcome in a
- % footnote to the title.
- %
- %
- % The first thing that we do next is to get ourselves a new comment
- % sign. Because all sensible signs are already occupied, we will
- % choose one that can only be entered indirectly:
- % {\DoNotIndex{\^}^^A avoid misinterpretion !!!!! VERIFY
- % \begin{macrocode}
- \catcode`\^^A=14
- % \end{macrocode}
- % \SortIndex{\string^\string^A}{\string\verb\verbatimchar
- % \string^\string^A\verbatimchar
- % \encapchar main}
- % }
- %
- %
- % \subsection{Macros surrounding the `definition parts'}
- %
- % \begin{macro}{\macrocode}
- % Parts of the macro definition will be surrounded by the
- % environment {\sf macrocode}. Put more precisely, they will be
- % enclosed by a macro whose argument (the text to be set
- % `verbatim') is terminated by the string
- % \verb*+% \end{macrocode}+. Carefully note the number of spaces.
- % \verb+\macrocode+ is defined completely analogously to
- % \verb+\verbatim+, but because a few small changes were carried
- % out, almost all internal macros have got new names. We start by
- % calling the macro \verb+\macro@code+, the macro which bears the
- % brunt of most of the work, such as \verb+\catcode+ reassignments,
- % etc.
- % \changes{v1.5r}{89/11/04}{Support for code line no. (Undoc)}
- % \begin{macrocode}
- \def\macrocode{\macro@code
- % \end{macrocode}
- % Then we take care that all spaces have the same width, and that
- % they are not discarded.
- % \begin{macrocode}
- \frenchspacing \@vobeyspaces
- % \end{macrocode}
- % Before closing, we need to call \verb+\xmacro@code+. It is this
- % macro that expects an argument which is terminated by the above
- % string. This way it is possible to keep the \verb+\catcode+
- % changes local.
- % \changes{v1.5r}{89/11/04}{Support for code line no. (Undoc)}
- % \changes{v1.5t}{89/11/07}{Common code moved to `macro@code.}
- % \begin{macrocode}
- \xmacro@code}
- % \end{macrocode}
- % \end{macro}
- %
- %
- % \begin{macro}{\macro@code}
- % We will now begin with the macro that does the actual work:
- % \begin{macrocode}
- \def\macro@code{%
- % \end{macrocode}
- % In theory it should consist of a {\sf trivlist} environment, but
- % the empty space before and after the environment should not be
- % too large.
- % \begin{macrocode}
- \topsep \MacrocodeTopsep
- % \end{macrocode}
- % The next parameter we set is \verb+\@beginparpenalty+, in order
- % to prevent a page break before such an environment.
- % \begin{macrocode}
- \@beginparpenalty \predisplaypenalty
- % \end{macrocode}
- % We then start a \verb+\trivlist+, set \verb+\parskip+ back to
- % zero and start an empty \verb+\item+.
- % \begin{macrocode}
- \trivlist \parskip \z@ \item[]%
- % \end{macrocode}
- % Additionally, everything should be set in {\tt typewriter} font.
- % Some people might prefer it somewhat differently; because of this
- % the font choice is
- % macro-driven.\footnote{The font change has to be placed
- % {\em after\/}
- % the {\tt\bslash item}. Otherwise a change to
- % {\tt\bslash baselineskip} will affect the
- % paragraph above.}
- % \begin{macrocode}
- \macro@font
- % \end{macrocode}
- % Because \verb+\item+ sets various parameters, we have found it
- % necessary to alter some of these retrospectively.
- % \begin{macrocode}
- \leftskip\@totalleftmargin \advance\leftskip\MacroIndent
- \rightskip\z@ \parindent\z@ \parfillskip\@flushglue
- % \end{macrocode}
- % The next line consists of the \LaTeX{} definition of \verb+\par+
- % used in \verb+\verbatim+ and should result in blank lines being
- % shown as blank lines.
- % \changes{v1.5l}{89/09/10}{Code line numbers supported.}
- % \changes{v1.5t}{89/11/07}{Call `leavevmode to get `everypar on
- % blank lines.}
- % \changes{v1.7c}{92/3/24}{Added `interlinepenalty to `par from
- % verbatim.sty}
- % \begin{macrocode}
- \blank@linefalse \def\par{\ifblank@line
- \leavevmode\fi
- \blank@linetrue\@@par
- \penalty\interlinepenalty}
- % \end{macrocode}
- % What use is this definition of \verb+\par+\,? We use the macro
- % \verb+\obeylines+ of \cite{book:KnuthA} which changes all \verb+^^M+
- % to \verb+\par+ so that each can control its own indentation.
- % Next we must also ensure that all special signs are normalized;
- % that is, they must be given \verb+\catcode+ $12$.
- % \begin{macrocode}
- \obeylines \let\do\@makeother \catcode`\`\active \@noligs \dospecials
- % \end{macrocode}
- % \changes{v1.5t}{89/11/07}{Common code added.}
- % \changes{v1.5w}{90/02/05}{Skip of `@totalleftmargin added.}
- % If indexing by code lines is switched on the line number is
- % incremented and set appropriately. We also check whether the start of
- % the next line indicates a {\tt docstrip} module directive and process
- % it appropriately if so using "\check@module".
- % \begin{macrocode}
- \global\@newlistfalse
- \global\@minipagefalse
- \ifcodeline@index
- \everypar{\global\advance\c@CodelineNo\@ne
- \llap{\theCodelineNo\ \hskip\@totalleftmargin}%
- \check@module}%
- \else \everypar{\check@module}%
- \fi
- % \end{macrocode}
- % We also initialize the cross-referencing feature by calling
- % \verb+\init@crossref+. This will start the scanning mechanism
- % when encountering an escape character.
- % \begin{macrocode}
- \init@crossref}
- % \end{macrocode}
- % \end{macro}
- %
- %
- % \begin{macro}{\ifblank@line}
- % \begin{macro}{\blank@linetrue}
- % \begin{macro}{\blank@linefalse}
- % \verb+\ifblank@line+ is the switch used in the definition above.
- % In the original {\sf verbatim} environment the \verb+\if@tempswa+
- % switch is used. This is dangerous because its value may change
- % while processing lines in the {\sf macrocode} environment.
- % \begin{macrocode}
- \newif\ifblank@line
- % \end{macrocode}
- % \end{macro}
- % \end{macro}
- % \end{macro}
- %
- % \begin{macro}{\endmacrocode}
- % Because we have begun a {\sf trivlist} environment in the {\sf
- % macrocode} environment, we must also end it. We must also act on
- % the value of the "pm@module" flag (see below) and empty
- % "\everypar".
- % \changes{v1.5r}{89/11/04}{Support for code line no. (Undoc)}
- % \begin{macrocode}
- \def\endmacrocode{%
- \ifpm@module \endgroup \pm@modulefalse \fi
- \everypar{}%
- \global\@inlabelfalse
- \endtrivlist
- % \end{macrocode}
- % Additionally \verb+\close@crossref+ is used to do anything needed
- % to end the cross-referencing mechanism.
- % \begin{macrocode}
- \close@crossref}
- % \end{macrocode}
- % \end{macro}
- %
- %
- % \begin{macro}{\MacroFont}
- % Here is the default definition for the \verb+\MacroFont+ macro.
- % If the new font selection scheme is in use we suppress changes
- % of math fonts thereby making doc much faster.
- % \changes{v1.5x}{90/02/17}{`math@fontsfalse added for new font sel.}
- % \changes{v1.7a}{92/03/13}{Added `reset@font for NFSS.}
- % \begin{macrocode}
- \@ifundefined{MacroFont}{%
- \ifx\undefined\selectfont
- \def\MacroFont{\small\tt}\else
- \def\MacroFont{\math@fontsfalse\reset@font\small\tt}\fi
- }{}
- % \end{macrocode}
- % \end{macro}
- %
- % \begin{macro}{\AltMacroFont}
- % \begin{macro}{\macro@font}
- % \changes{v1.7a}{92/03/12}{Added to support distinction of modules.}
- % \changes{v1.7c}{92/03/26}{Altered font change for OFSS.}
- % Although most of the macro code is set in "\MacroFont" we want to be
- % able to switch to indicate module code set in "\AltMacroFont".
- % "\macro@font" keeps track of which one we're using. We can't do the
- % same thing sensibly in OFSS as in NFSS.
- % \begin{macrocode}
- \@ifundefined{AltMacroFont}{%
- \ifx\undefined\selectfont
- \def\AltMacroFont{\small\tt}\else
- \def\AltMacroFont{\math@fontsfalse\small\reset@font\it\tt}\fi
- }{}
- \let\macro@font=\MacroFont
- % \end{macrocode}
- % \end{macro}
- % \end{macro}
- %
- % \begin{macro}{\check@module}
- % \begin{macro}{\ifpm@module}
- % \changes{v1.7a}{92/03/12}{Added.}
- % This is inserted by "\everypar" at the start of each macrocode line to
- % check whether it starts with module information. (Such information is
- % of the form "%<"\meta{switch}">", where the "%" must be at the
- % start of the line and \meta{switch} comprises names with various
- % possible separators and a possible leading "+", "-", "*" or "/"
- % \cite{art:docstrip}. All that concerns us here is what the first
- % character of \meta{switch} is.) First it checks the "pm@module"
- % flag in case the previous line had a non-block module
- % directive i.e., not "%<*" or "%</"; if it did we need to close the
- % group it started and unset the flag. "\check@module" looks ahead at
- % the next token and then calls "\ch@percent" to take action depending
- % on whether or not it's a "%"; we don't want to expand the token at
- % this stage. This is all done conditionally so it can be turned off
- % if it causes problems with code that wasn't designed to be {\tt
- % docstrip}ped.
- % \begin{macrocode}
- \def\check@module{%
- \ifcheck@modules
- \ifpm@module \endgroup \pm@modulefalse \fi
- \expandafter\futurelet\expandafter\next\expandafter\ch@percent
- \fi}
- \newif\ifpm@module
- % \end{macrocode}
- % \end{macro}
- % \end{macro}
- % \begin{macro}{\DontCheckModules}
- % \begin{macro}{\CheckModules}
- % \changes{v1.7a}{92/03/12}{Added.}
- % \begin{macro}{\ifcheck@modules}
- % Here are two driver-file interface macros for turning the module
- % checking on and off using the "check@modules" switch.
- % \begin{macrocode}
- \def\DontCheckModules{\check@modulesfalse}
- \def\CheckModules{\check@modulestrue}
- \newif\ifcheck@modules \check@modulestrue
- % \end{macrocode}
- % \end{macro}
- % \end{macro}
- % \end{macro}
- % \begin{macro}{\ch@percent}
- % \changes{v1.7a}{92/03/12}{Added.}
- % If the lookahead token in "\next" is $"%"_{12}$ we go on to check
- % whether the following one is "<" and otherwise do nothing. Note the
- % "\expandafter" to get past the "\fi".
- % \begin{macrocode}
- \def\ch@percent{%
- \if \percentchar\next
- \expandafter\check@angle
- \fi}
- % \end{macrocode}
- % \end{macro}
- % \begin{macro}{\check@angle}
- % \changes{v1.7a}{92/03/12}{Added.}
- % Before looking ahead for the "<" the "%" is gobbled by the
- % argument here.
- % \begin{macrocode}
- \def\check@angle#1{\futurelet\next\ch@angle}
- % \end{macrocode}
- % \end{macro}
- % \begin{macro}{\ch@angle}
- % \changes{v1.7a}{92/03/12}{Added.}
- % If the current lookahead token is "<" we are defined to be
- % processing a module directive can go on to look for "+"
- % etc.; otherwise we must put back the gobbled "%".
- % \begin{macrocode}
- \def\ch@angle{\if<\next
- \expandafter\ch@plus@etc
- \else \percentchar \fi}
- % \end{macrocode}
- % \end{macro}
- % \begin{macro}{\ch@plus@etc}
- % \begin{macro}{\check@plus@etc}
- % \changes{v1.7a}{92/03/12}{Added.}
- % We now have to decide what sort of a directive we're dealing with
- % and do the right thing with it.
- % \begin{macrocode}
- \def\ch@plus@etc<{\futurelet\next\check@plus@etc}
- \def\check@plus@etc{%
- \if +\next
- \let\next\pm@module
- \else\if -\next
- \let\next\pm@module
- \else\if *\next
- \let\next\star@module
- \else\if /\next
- \let\next\slash@module
- \else
- \let\next\pm@module
- \fi\fi\fi\fi
- \next}
- % \end{macrocode}
- % \end{macro}
- % \end{macro}
- % \begin{macro}{\pm@module}
- % If we're not dealing with a block
- % directive ("*" or "/") i.e., it's a single special line, we set
- % everything up to the next ">" appropriately and then change to the
- % special macro font inside a group which will be ended at the start
- % of the next line. If the apparent module directive is missing the
- % terminating ">" this will lose, but then so will the {\tt docstrip}
- % implementation. An alternative strategy would be to have
- % "\pm@module" make ">" active and clear a flag set here to indicate
- % processing the directive. Appropriate action could then be taken if
- % the flag was found still to be set when processing the next line.
- % \changes{v1.7a}{92/03/12}{Added.}
- % \changes{v1.7i}{92/07/11}{Support for fonts depending on nesting.}
- % \begin{macrocode}
- \def\pm@module#1>{\pm@moduletrue
- \Module{#1}\begingroup
- % \end{macrocode}
- % We switch to a special font as soon the nesting is higher than
- % the current value of "\c@StandardModuleDepth". We do a local
- % update to the "\guard@level" here which will be restored after
- % the current input line.
- % \begin{macrocode}
- \advance\guard@level\@ne
- \ifnum\guard@level>\c@StandardModuleDepth\AltMacroFont\fi
- }
- % \end{macrocode}
- % \end{macro}
- % \begin{macro}{\star@module}
- % \begin{macro}{\slash@module}
- % \changes{v1.7a}{92/03/12}{Added.}
- % \changes{v1.7f}{92/05/16}{Take account of nested guards.}
- % \changes{v1.7i}{92/07/11}{Add counter to determine when to switch to
- % special font.}
- % If the start or end of a module {\em block\/} is indicated, after
- % setting the guard we have to check whether a change in the macrocode
- % font should be done. This will be the case if we are already inside
- % a block or are ending the outermost block. If so, we globally
- % toggle the font for subsequent macrocode sections between the normal
- % and special form, switching to the new one immediately.
- % \changes{v1.7i}{92/07/17}{Support for fonts depending on module
- % nesting}
- % \begin{macrocode}
- \def\star@module#1>{%
- \Module{#1}%
- \global \advance \guard@level\@ne
- \ifnum \guard@level>\c@StandardModuleDepth
- \global\let\macro@font=\AltMacroFont \macro@font
- \fi}
- \def\slash@module#1>{%
- \Module{#1}%
- \global \advance \guard@level\m@ne
- \ifnum \guard@level=\c@StandardModuleDepth
- \global\let\macro@font\MacroFont \macro@font
- \fi
- }
- % \end{macrocode}
- % \end{macro}
- % \end{macro}
- %
- %
- % \begin{macro}{\c@StandardModuleDepth}
- % \changes{v1.7i}{92/07/11}{Counter added.}
- % Counter defining up to which level modules are considered part of
- % the main code. If, for example, the whole code is surrounded by
- % a |%<*style>| module we better set this counter to |1| to avoid
- % getting the whole code be displayed in typewriter italic.
- % \begin{macrocode}
- \newcounter{StandardModuleDepth}
- % \end{macrocode}
- % \end{macro}
- %
- %
- % \begin{macro}{\guard@level}
- % \changes{v1.7f}{92/05/16}{Added.}
- % We need a counter to keep track of the guard nesting.
- % \begin{macrocode}
- \newcount \guard@level
- % \end{macrocode}
- % \end{macro}
- % \begin{macro}{\Module}
- % \changes{v1.7a}{92/03/12}{Added.}
- % \changes{v1.7d}{92/04/25}{Use sans font for modules.}
- % This provides a hook to determine the way the module directive is
- % set. It gets as argument everything between the angle brackets.
- % The default is to set the contents in sans serif text between
- % $\langle\,\rangle$ with the special characters suitably "\mathcode"d
- % by "\mod@math@codes". (You can't just set it in a sans text font
- % because normally "|" will print as an em-dash.) This is done
- % differently depending on whether we have the NFSS or the old one. In
- % the latter case we can easily change "\fam" appropriately.
- % \begin{macrocode}
- \@ifundefined{Module}{%
- \ifx\undefined\selectfont
- \def\Module#1{{\mod@math@codes$\fam\sffam\langle #1\rangle$}}
- % \end{macrocode}
- % With NFSS what we probably {\em should\/} do is change to a new
- % "\mathversion" but I (Dave Love) haven't spotted an easy way to do so
- % correctly if the document uses a version other than "normal". (We
- % need to know in what font to set the other groups.) This uses a new
- % math alphabet rather than version and consequently has to worry about
- % whether we're using {\sf oldlfont} or not. I expect there's a better
- % way\ldots
- % \begin{macrocode}
- \else
- \expandafter\ifx\csname ds@oldlfont\endcsname\relax
- \def\Module#1{{\mod@math@codes$\langle\sfmath{#1}\rangle$}}
- \else
- \def\Module#1{{\mod@math@codes$\langle{\sfmath #1}\rangle$}}
- \fi
- \fi}{}
- % \end{macrocode}
- % \end{macro}
- %
- % \begin{macro}{\mod@math@codes}
- % \changes{v1.7c}{92/03/26}{Added.}
- % As well as `words', the module directive text might contain any of the
- % characters "*/+-,&|!()" for the current version of {\sf docstrip}. We
- % only need special action for two of them in the math code changing
- % required above: "|" is changed to a "\mathop" (it's normally
- % \verb+"026A+) and "&" is also made a "\mathop", but in family 0.
- % Remember that "&" will not have a special catcode when it's
- % encountered.
- % \begin{macrocode}
- \def\mod@math@codes{\mathcode`\|="226A \mathcode`\&="2026}
- % \end{macrocode}
- % \end{macro}
- %
- % \begin{macro}{\sfmath}
- % \changes{v1.7c}{92/03/26}{Added.}
- % \changes{v1.7d}{92/04/25}{Use sans font for modules.}
- % If NFSS is in use we need a new math alphabet which uses a sans serif
- % font.
- % \begin{macrocode}
- \ifx\selectfont\undefined
- \else
- \ifx\sfmath\undefined
- \newmathalphabet*{\sfmath}{\sfdefault}{m}{n}\fi
- \fi
- % \end{macrocode}
- % \end{macro}
- %
- %
- %
- % \begin{macro}{\MacrocodeTopsep}
- % \begin{macro}{\MacroIndent}
- % In the code above, we have used two registers. Therefore we have
- % to allocate them. The default values might be overwritten with
- % the help of the \verb+\DocstyleParms+ macro.
- % \changes{v1.5s}{89/11/05}{Support for code line no. (Undoc)}
- % \changes{v1.5y}{90/02/24}{Default changed.}
- % \changes{v1.6b}{90/06/15}{`rm moved before `scriptsize to
- % avoid unnecessary fontwarning.}
- % \begin{macrocode}
- \newskip\MacrocodeTopsep \MacrocodeTopsep = 3pt plus 1.2pt minus 1pt
- \newdimen\MacroIndent
- \settowidth\MacroIndent{\rm\scriptsize 00\ }
- % \end{macrocode}
- % \end{macro}
- % \end{macro}
- %
- %
- %
- %
- % \begin{macro}{\macrocode*}
- % \begin{macro}{\endmacrocode*}
- % Just as in the {\sf verbatim} environment, there is also a `star'
- % variant of the {\sf macrocode} environment in which a space is
- % shown by the symbol \verb*+ +. Until this moment, I have not yet
- % used it (it will be used in the description of the definition of
- % \verb+\xmacro@code+ below) but it's exactly on this one occasion
- % {\em here\/} that you can't use it
- % (cf.\ M\"unchhausens Marsh problem)\footnote{Karl Friedrich
- % Hieronymus Frhr.\ v.\ M\"unchhausen (*1720, \dag1797).
- % Several books were written about fantastic adventures
- % supposedly told by him (see \cite{book:Raspe} or
- % \cite{book:Buerger}). In one story he escaped from the
- % marsh by pulling himself out by his hair.}
- % directly. Because of this, on this one occasion we'll cheat
- % around the problem with an additional comment character. But now
- % back to \verb+\macrocode*+. We start with the macro
- % \verb+\macro@code+ which prepares everything and then call the
- % macro \verb+\sxmacro@code+ whose argument is terminated by the
- % string \verb*+% \end{macrocode*}+.
- % \begin{macrocode}
- \@namedef{macrocode*}{\macro@code\sxmacro@code}
- % \end{macrocode}
- % As we know, \verb+\sxmacro@code+ and then \verb+\end{macrocode*}+
- % (the macro, not the string), will be executed, so that for a
- % happy ending we still need to define the macro
- % \verb+\endmacrocode*+.
- % \begin{macrocode}
- \expandafter\let\csname endmacrocode*\endcsname = \endmacrocode
- % \end{macrocode}
- % \end{macro}
- % \end{macro}
- %
- %
- %
- %
- %
- %
- %
- % \begin{macro}{\xmacro@code}
- \catcode`\!=\catcode`\% ^^A In this section there must not be
- ^^A any exclamation marks.
- ^^A
- % As already mentioned, the macro \verb+\xmacro@code+ expects an
- % argument delimited by the string \verb*+% \end{macrocode}+. At
- % the moment that this macro is called, the \verb+\catcode+ of
- % \TeX's special characters are 12 (`other') or 13 (`active').
- % Because of this we need to utilize a different escape character
- % during the definition. This happens locally.
- % \begin{macrocode*}
- \begingroup
- \catcode`\|=\z@ \catcode`\[=\@ne \catcode`\]=\tw@
- % \end{macrocode*}
- % Additionally, we need to ensure that the symbols in the above
- % string contain the \verb+\catcode+$\,$s which are available
- % within the {\sf macrocode} environment.
- % \begin{macrocode*}
- \catcode`\{=12 \catcode`\}=12
- \catcode`\%=12 \catcode`\ =\active \catcode`\\=\active
- !% \end{macrocode*}
- ! Next follows the actual definition of \verb+\macro@code+;
- ! notice the
- ! use of the new escape character. We manage to get the argument
- ! surrounded by the string \verb+\end{macrocode}+, but at the end
- ! however, in spite of the actual characters used during the
- ! definition of
- ! this macro, \verb+\end+ with the argument \verb+{macrocode}+
- ! will be executed, to ensure a balanced environment.
- ! \begin{macrocode*}
- |gdef|xmacro@code#1% \end{macrocode}[#1|end[macrocode]]
- !% \end{macrocode*}
- ! \begin{macro}{\sxmacro@code}
- ! The definition of \verb+\sxmacro@code+ is completely analogous,
- ! only
- ! here a slightly different terminating string will be used.
- ! Note that the space is not active in this environment.
- ! \begin{macrocode}
- |catcode`| =12
- |gdef|sxmacro@code#1% \end{macrocode*}[#1|end[macrocode*]]
- !% \end{macrocode}
- ! because the \verb+\catcode+ changes have been made local by
- ! commencing a
- ! new group, there now follows the matching \verb+\endgroup+
- ! in a rather
- ! unusual style of writing.
- ! \begin{macrocode}
- |endgroup
- !% \end{macrocode}
- \catcode`\!=12
- % \end{macro}
- % \end{macro}
- %
- %
- %
- %
- % \subsection{Macros for the `documentation parts'}
- %
- %
- % \begin{macro}{\DescribeMacro}
- % \begin{macro}{\Describe@Macro}
- % \changes{v1.5v}{90/01/28}{Macro added.}
- % \changes{v1.5j}{89/06/09}{ignorespaces added as a temporary fix.}
- % \begin{macro}{\DescribeEnv}
- % \begin{macro}{\Describe@Env}
- % \changes{v1.5v}{90/01/28}{Macro added.}
- % \changes{v1.5j}{89/06/09}{ignorespaces added as a temporary fix.}
- % The \verb+\DescribeMacro+ and \verb+\DescribeEnv+ macros should
- % print their arguments in the margin and produce an index entry.
- % We simply use \verb+\marginpar+ to get the desired result. This
- % is however not the best solution because the labels might be
- % slightly misplaced. One also might get a lot of `marginpar moved'
- % messages which are hard-wired into the \LaTeX{} output
- % routine.\footnote{It might be better to change these macros into
- % environments like the {\sf macro} environment.} First we change
- % to horizontal mode if necessary. The \LaTeX{} macros
- % \verb+\@bsphack+ and \verb+\@esphack+ are used to make those
- % commands invisible (i.e.\ to normalize the surrounding space and
- % to make the \verb+\spacefactor+ transparent).
- % \changes{v1.5v}{90/01/28}{`MakePrivateLetters added.}
- % \begin{macrocode}
- \def\DescribeMacro{\leavevmode\@bsphack
- % \end{macrocode}
- % When documenting the code for the {\tt amstex.sty} option we
- % encountered a bug: the \verb+\catcode+ of \verb+@+ was active and
- % therefore couldn't be used in command names. So we first have to
- % make sure that we get all \verb+\catcode+s right by calling
- % \verb+\MakePrivateLetters+ inside a group. Then we call
- % \verb+\Describe@Macro+ to do the work.
- % \begin{macrocode}
- \begingroup\MakePrivateLetters\Describe@Macro}
- \def\Describe@Macro#1{\endgroup
- \marginpar{\raggedleft\PrintDescribeMacro{#1}}%
- % \end{macrocode}
- % Note the use of \verb+\raggedleft+ to place the output flushed
- % right. Finally we call a macro which produces the actual index
- % entry and finish with \verb+\@esphack+ to leave no
- % trace.\footnote{The whole mechanism won't work because
- % of the {\tt\bslash leavevmode} in front.
- % As a temporary change {\tt\bslash ignorespaces}
- % is added.}
- % \begin{macrocode}
- \SpecialUsageIndex{#1}\@esphack\ignorespaces}
- % \end{macrocode}
- % The \verb+\DescribeEnv+ macro is completely analogous.
- % \changes{v1.5v}{90/01/28}{`MakePrivateLetters added.}
- % \begin{macrocode}
- \def\DescribeEnv{\leavevmode\@bsphack\begingroup\MakePrivateLetters
- \Describe@Env}
- \def\Describe@Env#1{\endgroup
- \marginpar{\raggedleft\PrintDescribeEnv{#1}}%
- \SpecialEnvIndex{#1}\@esphack\ignorespaces}
- % \end{macrocode}
- % To put the labels in the left margin we have to use the
- % \verb+\reversemarginpar+ declaration. (This means that the {\tt
- % doc.sty} can't be used with all style options.) We also make the
- % \verb+\marginparpush+ zero and \verb+\marginparwidth+ suitably
- % wide.
- % \changes{v1.5d}{89/4/28}{marginparwith setting added.}
- % \begin{macrocode}
- \reversemarginpar
- \setlength\marginparpush{0pt} \setlength\marginparwidth{8pc}
- % \end{macrocode}
- % \end{macro}
- % \end{macro}
- % \end{macro}
- % \end{macro}
- %
- % \begin{macro}{\bslash}
- % \changes{v1.7a}{92/02/26}{Moved `bslash documentation to `user
- % interface' part}
- % We start a new group in which to hide the alteration of
- % \verb+\catcode+$\,$s, and make \verb+|+ introduce commands,
- % whilst \verb+\+ becomes an `other' character.
- %
- % \begin{macrocode}
- {\catcode`\|=\z@ \catcode`\\=12
- % \end{macrocode}
- % Now we are able to define \verb+\bslash+ (globally) to generate a
- % backslash of \verb+\catcode+ `other'. We then close this group,
- % restoring original \verb+\catcode+$\,$s.
- % \SpecialEscapechar{\|}
- % \begin{macrocode}
- |gdef|bslash{\}}
- % \end{macrocode}
- % \end{macro}
- %
- %
- %
- % \begin{macro}{\verbatim}
- % \begin{macro}{\verbatim*}
- % \changes{v1.7i}{92/07/12}{Added changed definition for verbatim!*.}
- % The {\sf verbatim} environment holds no secrets; it consists of
- % the normal \LaTeX{} environment. We also set the
- % \verb+\@beginparpenalty+ and change to the font given by
- % \verb+\MacroFont+.
- % \begin{macrocode}
- \def\verbatim{\@beginparpenalty \predisplaypenalty \@verbatim
- \MacroFont \frenchspacing \@vobeyspaces \@xverbatim}
- % \end{macrocode}
- % We deal in a similar way with the star form of this environment.
- % \begin{macrocode}
- \@namedef{verbatim*}{\@beginparpenalty \predisplaypenalty \@verbatim
- \MacroFont \@sxverbatim}
- % \end{macrocode}
- % \end{macro}
- % \end{macro}
- %
- % \begin{macro}{\@verbatim}
- % Additionally we redefine the \verb+\@verbatim+ macro so that it
- % suppresses \verb+%+ characters at the beginning of the line. The
- % first lines are copied literally from {\tt latex.tex}.
- % \changes{v1.7i}{92/07/12}{Added `@@par to clear possible `parshape.}
- % \begin{macrocode}
- \def\@verbatim{\trivlist \item[]\if@minipage\else\vskip\parskip\fi
- \leftskip\@totalleftmargin\rightskip\z@
- \parindent\z@\parfillskip\@flushglue\parskip\z@
- \@@par
- \@tempswafalse
- % \end{macrocode}
- % \verb+\@verbatim+ sets \verb+^^M+, the end of line character, to
- % be equal to \verb+\par+. This control sequence is redefined
- % here; \verb+\@@par+ is the paragraph primitive of \TeX.
- % \changes{v1.7c}{92/3/24}{Added `interlinepenalty to `par from
- % verbatim.sty}
- % \begin{macrocode}
- \def\par{\if@tempswa\hbox{}\fi\@tempswatrue\@@par
- \penalty\interlinepenalty
- % \end{macrocode}
- % We add a control sequence \verb+\check@percent+ to the definition
- % of \verb+\par+ whose task it is to check for a percent character.
- % \begin{macrocode}
- \check@percent}%
- % \end{macrocode}
- % The rest is again copied literally from {\tt latex.tex} (less
- % "\tt").
- % \changes{v1.7a}{92/02/26}{Removed redundant `tt.}
- % \begin{macrocode}
- \obeylines \catcode`\`\active \@noligs \let\do\@makeother
- \dospecials}
- % \end{macrocode}
- % \end{macro}
- %
- % \begin{macro}{\check@percent}
- % Finally we define \verb+\check@percent+. Since this must compare
- % a character with a percent sign we must first (locally) change
- % percent's \verb+\catcode+ so that it is seen by \TeX. The
- % definition itself is nearly trivial: grab the following
- % character, check if it is a \verb+%+, and insert it again if not.
- % At the end of the {\sf verbatim} environment this macro will peek
- % at the next input line. In that case the argument to
- % \verb+\check@percent+ might be a \verb+\par+ or a macro with
- % arguments. Therefore we make the definition \verb+\long+
- % (\verb+\par+ allowed) and use the normal \verb+\next+ mechanism
- % to reinsert the argument after the \verb+\fi+ if necessary.
- % \changes{v1.5i}{89/06/07}{Definition changed to `long'}
- % \changes{v1.5i}{89/06/07}{Macro `next' used to guard against
- % macro with arguments}
- % There is a subtle problem here, the equal sign between
- % \verb+\next+ and \verb+#1+ is actually necessary. Do you see why?
- % The omission of this token once caused a funny error.
- % \changes{v1.5u}{89/11/14}{equal sign added.}
- % \begin{macrocode}
- {\catcode`\%=12
- \long\gdef\check@percent#1{\ifx #1%\let\next\@empty \else
- \let\next=#1\fi \next}}
- % \end{macrocode}
- % \end{macro}
- %
- % \begin{macro}{\verb}
- % \changes{v1.7a}{92/02/27}{Now warns about newlines (from
- % newdoc with `@noligs added).}
- % We re-define "\verb" to check for newlines in its argument since a
- % missing delimiter is difficult to detect in {\sf doc} source.
- % Although the code is somewhat different, the
- % method follows \cite{art:verbatim}, which should be
- % consulted for commentary. Perhaps there should be a font-changing
- % hook rather than just using "\tt", but if so it probably should be
- % different from "\MacroFont" since that normally includes "\small"
- % and would look wrong inline. (There's no particular reason for
- % using lower-casing here to
- % splice in the relevant character, rather than the upper-casing used in
- % the definition of "\SpecialEscapechar"
- % (\S\ref{sect:specialescapechar}); remember that the case
- % shift doesn't touch the control sequence tokens.)
- % \changes{v1.7a}{92/02/28}{Added math math check (from verbatim.sty).}
- % \begin{macrocode}
- \begingroup
- \lccode`\~=`\^^M
- \lowercase{%
- \gdef\verb{\relax
- \ifmmode \hbox \else \leavevmode\null \fi
- \bgroup
- \tt \catcode`\`\active \@noligs
- \let~\verb@err
- \catcode`\^^M\active
- \let\do\@makeother \dospecials
- \@ifstar\@sverb{\@vobeyspaces \frenchspacing \@sverb}}}
- \endgroup
- \def\verb@err{\egroup\@latexerr{\string\verb\space
- command ended by end of line.}\@ehc}
- % \end{macrocode}
- % \end{macro}
- % \begin{macro}{\@sverb}
- % \changes{v1.7a}{92/02/27}{Added for `verb change.}
- % \changes{v1.7a}{92/02/28}{Now same as in verbatim.sty.}
- % See \cite{art:verbatim} for commentary.
- % \begin{macrocode}
- \def\@sverb#1{%
- \catcode`#1\active \lccode`\~`#1%
- \lowercase{\let~\egroup}}
- % \end{macrocode}
- % \end{macro}
- %
- %
- % \begin{macro}{\macro}
- % \begin{macro}{\m@cro@}
- % \changes{v1.5v}{90/01/28}{`macro@ renamed to `m@cro@ since AmSTeX
- % defines another macro of the same name.}
- % \begin{macro}{\macro@cnt}
- % \begin{macro}{\macro@level}
- % \label{page:macro} The {\sf macro} environment is implemented as
- % a {\sf trivlist} environment, whereby in order that the macro
- % names can be placed under one another in the margin
- % (corresponding to the macro's nesting depth), the macro
- % \verb+\makelabel+ must be altered. In order to store the nesting
- % depth, we use a counter. We also need a counter to count the
- % number of nested {\sf macro} environments.
- % \changes{v1.5k}{89/08/17}{Fix for save stack problem.}
- % \begin{macrocode}
- \newcount\macro@cnt \macro@cnt=0
- \newcount\macro@level \macro@level=0
- % \end{macrocode}
- % The environment takes an argument---the macro name to be
- % described. Since this name may contain special `letters' we have
- % to re-\verb+\catcode+ them before scanning the argument. This is
- % done by the \verb+\MakePrivateLetters+ macro. On toplevel we
- % start with \verb+\begingroup+ otherwise we simply re-catcode the
- % special letters and call \verb+\m@cro@+. Therefore the
- % \verb+\endgroup+ at the start of this macro will cancel the group
- % opened by the \verb+\begin+ macro if we are already inside a {\sf
- % macro} environment. This avoids problems with the save-stack on
- % small systems since then the \verb+\trivlist+ used later on will
- % not put anything onto this stack.
- % \changes{v1.5k}{89/08/17}{Fix for save stack problem.}
- % \changes{v1.7a}{92/02/26}{Catcode backslash to other (from newdoc).}
- % \begin{macrocode}
- \def\macro{%
- \ifnum\macro@level=\z@ \begingroup \fi
- \catcode`\\12
- \MakePrivateLetters \m@cro@}
- % \end{macrocode}
- % After scanning the argument we close the group to get the normal
- % \verb+\catcode+$\,$s back. Then we assign a special value to
- % \verb+\topsep+ and start a {\sf trivlist} environment.
- % Additionally we advance the \verb+\macro@level+ counter to keep
- % track of the number of nested {\sf macro} environments.
- % \changes{v1.5f}{89/5/07}{MacroTopsep parameter added.}
- % \changes{v1.5k}{89/08/17}{Fix for save stack problem.}
- % \begin{macrocode}
- \long\def\m@cro@#1{\endgroup \topsep\MacroTopsep \trivlist
- \advance\macro@level\@ne
- % \end{macrocode}
- % We also save the name being described in \verb+\saved@macroname+ for
- % use in conjunction with the \verb+\changes+ macro.
- % \begin{macrocode}
- \edef\saved@macroname{\string#1}%
- % \end{macrocode}
- % Now there follows a variation of \verb+\makelabel+ which is used
- % should the environment not be nested, or should it lie between
- % two successive \verb+\begin{macro}+ instructions or explanatory
- % text. One can recognize this with the switch \verb+\if@inlabel+
- % which will be \verb+true+ in the case of successive \verb+\item+
- % commands.
- % \begin{macrocode}
- \def\makelabel##1{\llap{##1}}%
- % \end{macrocode}
- % If \verb+@inlabel+ is \verb+true+ and if $\verb=\macro@cnt= > 0$
- % then the above definition needs to be changed, because in this
- % case \LaTeX{} would otherwise put the labels all on the same line
- % and this would lead to them being overprinted on top of each
- % other. Because of this \verb+\makelabel+ needs to be redefined
- % in this case.
- % \begin{macrocode}
- \if@inlabel
- % \end{macrocode}
- % If \verb+\macro@cnt+ has the value $1$, then we redefine
- % \verb+\makelabel+ so that the label will be positioned in the
- % second line of the margin. As a result of this, two macro names
- % appear correctly, one under the other. It's important whilst
- % doing this that the generated label box is not allowed to have
- % more depth than a normal line since otherwise the distance
- % between the first two text lines of \TeX{} will be incorrectly
- % calculated. The definition should then look like:
- %\begin{verbatim}
- % \def\makelabel##1{\llap{\vtop to \baselineskip
- % {\hbox{\strut}\hbox{##1}\vss}}}
- %\end{verbatim}
- % Completely analogous to this is the case where labels need to be
- % placed one under the other. The lines above are only an example
- % typeset with the {\sf verbatim} environment. To produce the real
- % definition we save the value of \verb+\macro@cnt+ in
- % \verb+\count@+ and empty the temp macro \verb+\@tempa+ for later
- % use.
- % \begin{macrocode}
- \let\@tempa\@empty \count@\macro@cnt
- % \end{macrocode}
- % In the following loop we append for every already typeset label
- % an \verb+\hbox{\strut}+ to the definition of \verb+\@tempa+.
- % \begin{macrocode}
- \loop \ifnum\count@>\z@
- \edef\@tempa{\@tempa\hbox{\strut}}\advance\count@\m@ne \repeat
- % \end{macrocode}
- % Now be put the definition of \verb+\makelabel+ together.
- % \changes{v1.5b}{89/04/27}{vbox to vtop changed in makelabel (test)}
- % \changes{v1.5e}{89/04/28}{ht strutbox changed to baselineskip (test)}
- % \begin{macrocode}
- \edef\makelabel##1{\llap{\vtop to\baselineskip
- {\@tempa\hbox{##1}\vss}}}%
- % \end{macrocode}
- % Next we increment the value of the nesting depth counter. This
- % value inside the {\sf macro} environment is always at least one
- % after this point, but its toplevel definition is zero. Provided
- % this environment has been used correctly, $\verb+\macro@cnt+=0$
- % should not occur when $\verb+@inlabel+=\sf true$. It is however
- % possible if this environment is used within other list
- % environments (but this would have little point).
- % \begin{macrocode}
- \advance \macro@cnt \@ne
- % \end{macrocode}
- % If \verb+@inlabel+ is false we reset \verb+\macro@cnt+ assuming
- % that there is enough room to print the macro name without
- % shifting.
- % \begin{macrocode}
- \else \macro@cnt\@ne \fi
- % \end{macrocode}
- % Now the label will be produced using \verb+\item+. The following
- % line is only a hack saving the day until a better solution is
- % implemented. We have to face two problems: the argument might be
- % a \verb+\par+ which is forbidden in the argument of other macros
- % if they are not defined as \verb+\long+, or it is something like
- % \verb+\iffalse+ or \verb+\else+, i.e.\ something which will be
- % misinterpreted when \TeX{} is skipping conditional text. In both
- % cases \verb+\item+ will bomb, so we protect the argument by using
- % \verb+\string+.
- % \begin{macrocode}
- \edef\@tempa{\noexpand\item[\noexpand\PrintMacroName{\string#1}]}%
- \@tempa
- % \end{macrocode}
- % At this point we also produce an index entry. Because it is not
- % known which index sorting program will be used, we do not use the
- % command \verb+\index+, but rather a command
- % \verb+\SpecialMainIndex+ after advancing the counter for indexing
- % by line number. This may be redefined by the user in
- % order to generate an index entry which will be understood by the
- % index program in use (note the definition of
- % \verb+\SpecialMainIndex+ for our installation).
- % \changes{v1.5s}{89/11/05}{Support for code line no. (Undoc)}
- % \begin{macrocode}
- {\advance\c@CodelineNo\@ne\SpecialMainIndex{#1}\nobreak}%
- % \end{macrocode}
- % The \verb+\nobreak+ is needed to prevent a page break after the
- % \verb+\write+ produced by the \verb+\SpecialMainIndex+ macro. We
- % exclude the new macro in the cross-referencing feature, to
- % prevent spurious non-main entry references. Regarding possibly
- % problematic arguments, the implementation takes
- % care of \verb+\par+ and the conditionals are uncritical.
- % \changes{v1.7a}{92/03/02}{Removed redundant code checking for
- % `par.}^^A
- % \begin{macrocode}
- \DoNotIndex{#1}%
- % \end{macrocode}
- % Because the space symbol should be ignored between the
- % \verb+\begin{macro}{...}+ and the following text we must take
- % care of this with \verb+\ignorespaces+.
- % \begin{macrocode}
- \ignorespaces}
- % \end{macrocode}
- % \end{macro}
- % \end{macro}
- % \end{macro}
- % \end{macro}
- %
- %
- %
- % \begin{macro}{\endmacro}
- % When ending a {\sf macro} environment we have to cancel the
- % \verb+\endgroup+ generated by the \verb+\end+ macro as long as we
- % are still inside another {\sf macro} environment. Therefore
- % we check the value of the \verb+\macro@level+ counter. Of course
- % this counter also has to be decremented.
- % \SpecialIndex{\macro@cnt}
- % \changes{v1.5k}{89/08/17}{Fix for save stack problem.}
- % \begin{macrocode}
- \def\endmacro{%
- \endtrivlist
- \ifnum\macro@level>\@ne \advance\macro@level\m@ne \begingroup \fi}
- % \end{macrocode}
- % \end{macro}
- %
- % \begin{macro}{\MacroTopsep}
- % Here is the default value for the \verb+\MacroTopsep+ parameter
- % used above.
- % \begin{macrocode}
- \newskip\MacroTopsep \MacroTopsep = 7pt plus 2pt minus 2pt
- % \end{macrocode}
- % \end{macro}
- %
- %
- %
- %
- %
- % \subsection{Formatting the margin}
- %
- % The following three macros should be user definable.
- % Therefore we define those macros only if they have not already
- % been defined.
- %
- % \begin{macro}{\PrintMacroName}
- % \begin{macro}{\PrintDescribeMacro}
- % \begin{macro}{\PrintDescribeEnv}
- % The formatting of the macro name in the left margin is done by
- % these macros. We first set a \verb+\strut+ to get the height and
- % depth of the normal lines. Then we change to the
- % \verb+\MacroFont+ using \verb+\string+ to \verb+\catcode+ the
- % argument to other (assuming that it is a macro name). Finally we
- % print a space. The font change remains local since this macro
- % will be called inside an \verb+\hbox+.
- % \begin{macrocode}
- \@ifundefined{PrintMacroName}
- {\def\PrintMacroName#1{\strut \MacroFont \string #1\ }}{}
- % \end{macrocode}
- % We use the same formatting conventions when describing a macro.
- % \begin{macrocode}
- \@ifundefined{PrintDescribeMacro}
- {\def\PrintDescribeMacro#1{\strut \MacroFont \string #1\ }}{}
- % \end{macrocode}
- % To format the name of a new environment there is no need to use
- % \verb+\string+.
- % \begin{macrocode}
- \@ifundefined{PrintDescribeEnv}
- {\def\PrintDescribeEnv#1{\strut \MacroFont #1\ }}{}
- % \end{macrocode}
- % \end{macro}
- % \end{macro}
- % \end{macro}
- %
- %
- %
- % \subsection{Creating index entries by scanning `macrocode'}
- %
- % The following macros ensure that index entries are created for each
- % occurrence of a \TeX-like command (something starting with
- % `\verb+\+') providing indexing has been turned on with "\PageIndex"
- % or "\CodelineIndex". With the default definitions of
- % \verb+\SpecialMainIndex+, etc., the index file generated is
- % intended to be processed by Chen's {\sf makeindex} program
- % \cite{art:Chen}.
- %
- %
- % Of course, in {\em this\/} style file itself we've sometimes had to
- % make \verb+|+ take the r\^ole of \TeX's escape character to
- % introduce command names at places where \verb+\+ has to belong to
- % some other category. Therefore, we may also need to recognize
- % \verb+|+ as the introducer for a command when setting the text
- % inside the {\sf macrocode} environment. Other users may have the
- % need to make similar reassignments for their macros.
- %
- %
- % \begin{macro}{\SpecialEscapechar}\label{sect:specialescapechar}
- % \begin{macro}{\active@escape@char}
- % \begin{macro}{\special@escape@char}
- % The macro \verb+\SpecialEscapechar+ is used to denote a special
- % escape character for the next {\sf macrocode} environment. It has
- % one argument---the new escape character given as a
- % `single-letter' control sequence. Its main purpose is defining
- % \verb+\special@escape@char+ to produce the chosen escape
- % character \verb+\catcode+$\,$d to 12 and
- % \verb+\active@escape@char+ to produce the same character but with
- % \verb+\catcode+ 13.
- %
- % The macro \verb+\special@escape@char+ is used to {\em print\/}
- % the escape character while \verb+\active@escape@char+ is needed
- % in the definition of \verb+\init@crossref+ to start the scanning
- % mechanism.
- %
- % In the definition of \verb+\SpecialEscapechar+ we need an
- % arbitrary character with \verb+\catcode+ 13. We use `\~{}' and
- % ensure that it is active. The \verb+\begingroup+ is used to make
- % a possible change local to the expansion of
- % \verb+\SpecialEscapechar+.
- % \changes{v1.7g}{92/6/19}{Making tilde active moved outside
- % definition}
- % \begin{macrocode}
- \begingroup
- \catcode`\~\active
- \gdef\SpecialEscapechar#1{%
- \begingroup
- % \end{macrocode}
- % Now we are ready for the definition of
- % \verb+\active@escape@char+. It's a little tricky: we first
- % define locally the uppercase code of `\~{}' to be the new escape
- % character.
- % \begin{macrocode}
- \uccode`\~`#1%
- % \end{macrocode}
- % Around the definition of \verb+\active@escape@char+ we place an
- % \verb+\uppercase+ command. Recall that the expansion of
- % \verb+\uppercase+ changes characters according to their
- % \verb+\uccode+, but leaves their \verb+\catcode+$\,$s untouched
- % (cf.\ \TeX{}book page 41).
- % \begin{macrocode}
- \uppercase{\gdef\active@escape@char{~}}%
- % \end{macrocode}
- % The definition of \verb+\special@escape@char+ is easier, we use
- % \verb+\string+ to \verb+\catcode+ the argument of
- % \verb+\SpecialEscapechar+ to 12 and suppress the preceding
- % \verb+\escapechar+.
- % \begin{macrocode}
- \escapechar\m@ne \xdef\special@escape@char{\string#1}%
- % \end{macrocode}
- % Now we close the group and end the definition: the value of
- % \verb+\escapechar+ as well as the \verb+\uccode+ and
- % \verb+\catcode+ of `\~{}' will be restored.
- % \begin{macrocode}
- \endgroup}
- \endgroup
- % \end{macrocode}
- % \end{macro}
- % \end{macro}
- % \end{macro}
- %
- %
- %
- %
- % \begin{macro}{\init@crossref}
- % The replacement text of \verb+\init@crossref+ should fulfill the
- % following tasks:
- % \begin{itemize}
- % \parindent4em
- % \item[1)]
- % \verb+\catcode+ all characters used in macro names to
- % 11 (i.e.\ `letter').
- % \item[2)]
- % \verb+\catcode+ the `\verb+\+' character to 13
- % (i.e.\ `active').
- % \item[3a)]
- % \verb+\let+ the `\verb+\+' equal \verb+\scan@macro+
- % (i.e.\ start the macro scanning mechanism) if there is
- % no special escape character (i.e.\ the
- % \verb+\special@escape@char+ is `\verb+\+').
- % \item[3b)]
- % Otherwise \verb+\let+ it equal \verb+\bslash+, i.e.\
- % produce a printable \verb+\+.
- % \item[4)]
- % Make the \meta{special escape character} active.
- % \item[5)]
- % \verb+\let+ the active version of the special escape
- % character
- % (i.e.\ the expansion of \verb+\active@escape@char+) equal
- % \verb+\scan@macro+.
- % \end{itemize}
- % The reader might ask why we bother to \verb+\catcode+ the
- % `\verb+\+' first to 12 (at the end of \verb+\macro@code+) then
- % re-\verb+\catcode+ it to 13 in order to produce a $\verb+\+_{12}$
- % in case 3b) above. This is done because we have to ensure that
- % `\verb+\+' has \verb+\catcode+ 13 within the {\sf macrocode}
- % environment. Otherwise the delimiter for the argument of
- % \verb+\xmacro@code+ would not be found (parameter matching
- % depends on \verb+\catcode+$\,$s).
- %
- % Therefore we first re-\verb+\catcode+ some characters.
- % \begin{macrocode}
- \begingroup \catcode`\|=\z@ \catcode`\\=\active
- % \end{macrocode}
- % We carry out tasks 2) and 3b) first.
- % \SpecialEscapechar{\|}
- % \begin{macrocode}
- |gdef|init@crossref{|catcode`|\|active |let\|bslash
- % \end{macrocode}
- % Because of the popularity of the `\verb+@+' character as a
- % `letter' in macros, we normally have to change its
- % \verb+\catcode+ here, and thus fulfill task 1). But the macro
- % designer might use other characters as private letters as well,
- % so we use a macro to do the \verb+\catcode+ switching.
- % \SpecialEscapechar\|
- % \begin{macrocode}
- |MakePrivateLetters
- % \end{macrocode}
- % Now we \verb+\catcode+ the special escape character to 13 and
- % \verb+\let+ it equal \verb+\scan@macro+, i.e.\ fulfill tasks 4)
- % and 5). Note the use of \verb+\expandafter+ to insert the chosen
- % escape character saved in \verb+\special@escape@char+ and
- % \verb+\active@escape@char+.
- % \SpecialEscapechar\|
- % \begin{macrocode}
- |catcode|expandafter`|special@escape@char|active
- |expandafter|let|active@escape@char|scan@macro}
- |endgroup
- % \end{macrocode}
- % If there is no special escape character, i.e.\ if
- % \verb+\SpecialEscapechar+ is \verb+\\+, the second last line will
- % overwrite the previous definition of $\verb+\+_{13}$. In this
- % way all tasks are fulfilled.
- %
- % For happy documenting we give default values to
- % \verb+\special@escape@char+ and \verb+\active@escape@char+ with
- % the following line:
- % \begin{macrocode}
- \SpecialEscapechar{\\}
- % \end{macrocode}
- % \end{macro}
- %
- %
- %
- % \begin{macro}{\MakePrivateLetters}
- % Here is the default definition of this command, which makes just
- % the \verb+@+ into a letter. The user may change it if he/she
- % needs more or other characters masquerading as letters.
- % \begin{macrocode}
- \@ifundefined{MakePrivateLetters}
- {\let\MakePrivateLetters\makeatletter}{}
- % \end{macrocode}
- % \end{macro}
- %
- % \begin{macro}{\close@crossref}
- % At the end of a cross-referencing part we prepare ourselves for
- % the next one by setting the escape character to `\verb+\+'.
- % \begin{macrocode}
- \def\close@crossref{\SpecialEscapechar\\}
- % \end{macrocode}
- % \end{macro}
- %
- %
- %
- %
- % \subsection{Macros for scanning macro names}
- %
- % \begin{macro}{\scan@macro}
- % \changes{v1.5k}{89/09/04}{Support for checksum added.}
- % \begin{macro}{\macro@namepart}
- % The \verb+\init@crossref+ will have made \verb+\active+ our
- % \verb+\special@escape@char+, so that each
- % \verb+\active@escape@char+ will invoke \verb+\scan@macro+ when
- % within the {\sf macrocode} environment. By this means, we can
- % automatically add index entries for every \TeX-like command which
- % is met whilst setting (in verbatim) the contents of {\sf
- % macrocode} environments.
- % \begin{macrocode}
- \def\scan@macro{%
- % \end{macrocode}
- % First we output the character which triggered this macro. Its
- % version \verb+\catcode+$\,$d to 12 is saved in
- % \verb+\special@escape@char+. We also call \verb+\step@checksum+
- % to generate later on a proper check-sum (see section
- % \ref{sec:checksum} for details).
- % \begin{macrocode}
- \special@escape@char
- \step@checksum
- % \end{macrocode}
- % If the {\sf macrocode} environment contains, for example, the
- % command \verb+\\+, the second \verb+\+ should not start the
- % scanning mechanism. Therefore we use a switch to decide if
- % scanning of macro names is allowed.
- % \begin{macrocode}
- \ifscan@allowed
- % \end{macrocode}
- % The macro assembles the letters forming a \TeX\ command in
- % \verb+\macro@namepart+ so this is initially cleared; we then set
- % \verb+\next+ to the {\it first\/} character following the
- % \verb+\+ and call \verb+\macro@switch+ to determine whether that
- % character is a letter or not.
- % \begin{macrocode}
- \let\macro@namepart\@empty
- \def\next{\futurelet\next\macro@switch}%
- % \end{macrocode}
- % As you recognize, we actually did something else, because we have
- % to defer the \verb+\futurelet+ call until after the final
- % \verb+\fi+. If, on the other hand, the scanning is disabled we
- % simply \verb+\let+ \verb+\next+ equal `empty'.
- % \begin{macrocode}
- \else \let\next\@empty \fi
- % \end{macrocode}
- % Now we invoke \verb+\next+ to carry out what's needed.
- % \begin{macrocode}
- \next}
- % \end{macrocode}
- % \end{macro}
- % \end{macro}
- %
- %
- % \begin{macro}{\ifscan@allowed}
- % \begin{macro}{\scan@allowedtrue}
- % \begin{macro}{\scan@allowedfalse}
- % \verb+\ifscan@allowed+ is the switch used above to determine if
- % the \verb+\active@escape@char+\SpecialIndex{\active@escape@char}
- % should start the macro scanning mechanism.
- % \begin{macrocode}
- \newif\ifscan@allowed \scan@allowedtrue
- % \end{macrocode}
- % \end{macro}
- % \end{macro}
- % \end{macro}
- %
- %
- % \begin{macro}{\EnableCrossrefs}
- % \begin{macro}{\DisableCrossrefs}
- % At this point we might define two macros which allow the user to
- % disable or enable the cross-referencing mechanism. Processing of
- % files will be faster if only main index entries are generated
- % (i.e., if \verb+\DisableCrossrefs+ is in force).
- % \begin{macrocode}
- \def\DisableCrossrefs{\@bsphack\scan@allowedfalse\@esphack}
- % \end{macrocode}
- % The macro \verb+\EnableCrossrefs+ will also disable any
- % \verb+\DisableCrossrefs+ command encountered afterwards.
- % \begin{macrocode}
- \def\EnableCrossrefs{\@bsphack\scan@allowedtrue
- \def\DisableCrossrefs{\@bsphack\@esphack}\@esphack}
- % \end{macrocode}
- % \end{macro}
- % \end{macro}
- %
- %
- %
- %
- % \begin{macro}{\macro@switch}
- % Now that we have the character which follows the escape character
- % (in \verb+\next+), we can determine whether it's a `letter'
- % (which probably includes \verb+@+).
- %
- % If it is, we let \verb+\next+ invoke a macro which assembles the
- % full command name.
- % \begin{macrocode}
- \def\macro@switch{\ifcat\noexpand\next a%
- \let\next\macro@name
- % \end{macrocode}
- % Otherwise, we have a `single-character' command name. For all
- % those single-character names, we use \verb+\short@macro+ to
- % process them into suitable index entries.
- % \begin{macrocode}
- \else \let\next\short@macro \fi
- % \end{macrocode}
- % Now that we know what macro to use to process the macro name, we
- % invoke it~\ldots
- % \begin{macrocode}
- \next}
- % \end{macrocode}
- % \end{macro}
- %
- %
- % \begin{macro}{\short@macro}
- % \changes{v1.5c}{89/4/27}{Corrected bad bug by putting the
- % scan@allowedfalse macro before printing
- % the argument.}
- % \changes{v1.7a}{92/03/10}{Ensure character stored in `macro@namepart
- % as `letter' so index exclusion works.}
- % This macro will be invoked (with a single character as parameter)
- % when a single-character macro name has been spotted whilst
- % scanning within the {\sf macrocode} environment.
- %
- % First we take a look at the \verb+\index@excludelist+ to see
- % whether this macro name should produce an index entry. This is
- % done by the \verb+\ifnot@excluded+ macro which assumes that the
- % macro name is saved in \verb+\macro@namepart+. The character
- % mustn't be stored with a special category code or exclusion from
- % the index won't work, so we employ the case-changing trick used
- % elsewhere. Since the argument might be an active character,
- % \verb+\string+ is used to normalize it.
- % \begin{macrocode}
- \def\short@macro#1{\begingroup
- \catcode`\&=11 \uccode`\&=\expandafter`\string#1%
- \uppercase{\def\macro@namepart{&}}%
- \endgroup
- \ifnot@excluded
- % \end{macrocode}
- % If necessary the index entry is produced by the macro
- % \verb+\produce@index+. Depending on the actual character seen,
- % this macro has to do different things, so we pass the character
- % as an argument.
- % \begin{macrocode}
- \produce@index{#1}\fi
- % \end{macrocode}
- % Then we disable the cross-referencing mechanism with
- % \verb+\scan@allowedfalse+ and print the actual character. The
- % index entry was generated first to ensure that no page break
- % intervenes (recall that a \verb+^^M+ will start a new line).
- % \begin{macrocode}
- \scan@allowedfalse#1%
- % \end{macrocode}
- % After typesetting the character we can safely enable the
- % cross-referencing feature again. Note that this macro won't be
- % called (since \verb+\macro@switch+ won't be called) if
- % cross-referencing is globally disabled.
- % \begin{macrocode}
- \scan@allowedtrue }
- % \end{macrocode}
- % \end{macro}
- %
- %
- %
- % \begin{macro}{\produce@index}
- % \changes{v1.4s}{89/04/23}{Added noexpand to all {\tt\protect\bslash
- % if} tests
- % to avoid garbage produced by new active chars}
- % \changes{v1.4s}{89/04/23}{Used {\tt\protect\bslash string} for the
- % same reason.}
- % \changes{v1.5c}{89/4/27}{Corrected bad bug by placing the
- % scan@allowedfalse macro into short@macro}
- % This macro is supposed to generate a suitable \verb+\SortIndex+
- % command for a given single-letter control sequence. We test
- % first for the cases which involve active characters (i.e.\ the
- % backslash, the special escape character (if any), the space and
- % the \verb+^^M+). Using the \verb+\if+ test (testing for
- % character codes), we have to ensure that the argument isn't
- % expanded.
- % \begin{macrocode}
- \def\produce@index#1{%
- \if\noexpand#1\special@escape@char
- % \end{macrocode}
- % If the character is the special escape character (or the
- % `\verb+\+' in case there was none) the \verb+\it@is@a+ macro is
- % used to produce the actual \verb+\SortIndex+ call.
- % \begin{macrocode}
- \scan@allowedfalse \it@is@a\special@escape@char \else
- % \end{macrocode}
- % Next comes the test for a `\verb+\+' which must be the
- % $\verb+\+_{13}$ expanding to \verb+\bslash+.
- % \begin{macrocode}
- \if\noexpand#1\bslash \it@is@a\bslash \else
- % \end{macrocode}
- % Another possibility is \verb*+ +$_{13}$. Recall that \verb+\space+
- % produces a \verb*+ +$_{10}$.
- % \begin{macrocode}
- \if\noexpand#1\space \it@is@a\space \else
- % \end{macrocode}
- % The last\footnote{Well, it isn't the last active character after
- % all. I added {\tt \bslash @noligs} some days ago
- % and now {\tt `} too is active. So we have to make
- % sure that such characters don't get expanded in
- % the index.}
- % possibility of an active character is \verb+^^M+\@. In this case
- % we don't test for character codes, since it is easier to look if
- % the character is equal to \verb+\par+. (We are inside the {\sf
- % macrocode} environment.)
- % \begin{macrocode}
- \ifx#1\par
- % \end{macrocode}
- % If we end up here we have just scanned a \verb+\^^M+ or something
- % similar. Since this will be treated like \verb*+\ + by \TeX{} we
- % produce a corresponding index entry.
- % \begin{macrocode}
- \it@is@a\space \else
- % \end{macrocode}
- % If it is the token \verb+\relax+ we do nothing. This can't happen
- % when the `doc' option is used in the way described here, but was
- % added to allow extensions like the {\tt idxverb} option.
- % \changes{v1.5t}{89/11/14}{Added `relax as a possible token to allow
- % extensions.}
- % \begin{macrocode}
- \ifx#1\relax \else
- % \end{macrocode}
- % The next three branches are needed because of bugs in
- % our {\sf makeindex} program. You can't produce unbalanced index
- % entries\footnote{This is possible for \TeX{} if you use
- % {\tt\string{$_{12}$ \rm or \tt\string}$_{12}$},
- % but {\sf makeindex} will complain.}
- % and you have to double a percent character. To get around these
- % restrictions we use special macros to produce the \verb+\index+
- % calls.\footnote{Brian {\sc Hamilton Kelly} has written fixes for
- % all three
- % bugs. When they've found their way through all
- % installations,
- % the lines above will be removed. See
- % page~\pageref{bug:fixes} if you already have them.
- % (I'm not sure which versions incorporate these, but
- % 2.11 is OK. See also
- % \pageref{makeindex:version}.)}
- % \begin{macrocode}
- \if\noexpand#1\bgroup \LeftBraceIndex \else
- \if\noexpand#1\egroup \RightBraceIndex \else
- \if\noexpand#1\percentchar \PercentIndex \else
- % \end{macrocode}
- % All remaining characters are used directly to produce their index
- % entries. This is possible even for the characters which have
- % special meanings in the index program, provided we quote the
- % characters. (This is correctly done in \verb+\it@is@a+.)
- % \begin{macrocode}
- \it@is@a{\string#1}%
- % \end{macrocode}
- % We now need a whole pile of \verb+\fi+$\,$s to match up with
- % the \verb+\if+$\,$s.
- % \begin{macrocode}
- \fi \fi \fi \fi \fi \fi \fi \fi}
- % \end{macrocode}
- % \end{macro}
- %
- %
- %
- % \begin{macro}{\macro@name}
- % We now come to the macro which assembles command names which
- % consist of one or more `letters' (which might well include
- % \verb+@+ symbols, or anything else which has a \verb+\catcode+ of
- % 11).
- %
- % To do this we add the `letter' to the existing definition of
- % \verb+\macro@namepart+ (which you will recall was originally set
- % to \verb+\@empty+).
- % \begin{macrocode}
- \def\macro@name#1{\edef\macro@namepart{\macro@namepart#1}%
- % \end{macrocode}
- % Then we grab hold of the {\em next\/} single character and let
- % \verb+\more@macroname+ determine whether it belongs to the letter
- % string forming the command name or is a `non-letter'.
- % \begin{macrocode}
- \futurelet\next\more@macroname}
- % \end{macrocode}
- % \end{macro}
- %
- %
- %
- %
- %
- % \begin{macro}{\more@macroname}
- %
- % This causes another call of \verb+\macro@name+ to add in the next
- % character, if it is indeed a `letter'.
- % \begin{macrocode}
- \def\more@macroname{\ifcat\noexpand\next a%
- \let\next\macro@name
- % \end{macrocode}
- % Otherwise, it finishes off the index entry by invoking
- % \verb+\macro@finish+.
- % \begin{macrocode}
- \else \let\next\macro@finish \fi
- % \end{macrocode}
- % Here's where we invoke whatever macro was \verb+\let+ equal to
- % \verb+\next+.
- % \begin{macrocode}
- \next}
- % \end{macrocode}
- % \end{macro}
- %
- %
- %
- %
- %
- % \begin{macro}{\macro@finish}
- % When we've assembled the full `letter'-string which forms the
- % command name, we set the characters forming the entire command
- % name, and generate an appropriate \verb+\index+ command (provided
- % the command name is not on the list of exclusions). The
- % `\verb+\+' is already typeset; therefore we only have to output
- % all `letters' saved in \verb+\macro@namepart+.
- % \begin{macrocode}
- \def\macro@finish{%
- \macro@namepart
- % \end{macrocode}
- % Then we call \verb+\ifnot@excluded+ to decide whether we have to
- % produce an index entry. The construction with \verb+\@tempa+ is
- % needed because we want the expansion of \verb+\macro@namepart+ in
- % the \verb+\index+ command.\footnote{The {\tt \bslash index}
- % command will expand its argument in the {\tt\bslash output}
- % routine. At this time {\tt\bslash macro@namepart} might have a
- % new value.}
- % \begin{macrocode}
- \ifnot@excluded
- \edef\@tempa{\noexpand\SpecialIndex{\bslash\macro@namepart}}%
- \@tempa \fi}
- % \end{macrocode}
- % \end{macro}
- %
- %
- %
- %
- %
- % \subsection[The index exclude list]{The index exclude
- % list\footnotemark}
- % \footnotetext{Warning: the incomplete commentary on {\tt
- % \bslash DoNotIndex} and the macros it calls
- % was written by Dave Love.}
- %
- % The internal form of the index exclude list is
- % \begin{quote}
- % \meta{macro name}\verb+,+\meta{macro name}\verb+,+
- % \ldots\verb+,+
- % \end{quote}
- % where \meta{macro name} is a macro name like
- % $"\"_{12}"p"{_{11}}"@"_{11}$ or $"\"_{12}"$"_{11}$. Note that the "\"
- % has category `other' and the other characters in the name are all
- % `letter', regardless of their normal category.
- %
- % \begin{macro}{\DoNotIndex}
- % This macro is used to suppress macro names in the index. It
- % starts off with a new group because we have to change the
- % \verb+\catcode+$\,$s of all characters which belong to `letters'
- % while macros are defined.
- % \begin{macrocode}
- \def\DoNotIndex{\begingroup \MakePrivateLetters
- \catcode`\\12
- % \end{macrocode}
- % Then we call the macro which actually reads the argument given by
- % the user.
- % \begin{macrocode}
- \do@not@index}
- % \end{macrocode}
- % \end{macro}
- %
- %
- %
- % \begin{macro}{\do@not@index}
- % We make the \verb+\do@not@index+ macro \verb+\long+
- % since the user might want to exclude the \verb+\par+
- % macro.
- % \begin{macrocode}
- \long\def\do@not@index#1{%
- % \end{macrocode}
- % It just adds to a token list after finishing the group in which
- % the catcodes were changed.
- % \changes{v1.7a}{92/02/26}{Replaced with newdoc version.}
- % \begin{macrocode}
- \endgroup
- \addto@hook\index@excludelist{#1,}}
- % \end{macrocode}
- % \end{macro}
- %
- % \begin{macro}{\addto@hook}
- % The code for adding tokens (the second argument) to a token list
- % (the first argument) is taken from~\cite{art:verbatim}, but it needs
- % to be "\long" in case "\par" is amongst the tokens.
- % \begin{macrocode}
- \long\def\addto@hook#1#2{#1\expandafter{\the#1#2}}
- % \end{macrocode}
- % \end{macro}
- %
- % \begin{macro}{\index@excludelist}
- % We need an initially-empty register for the excluded list.
- % \begin{macrocode}
- \newtoks\index@excludelist
- \index@excludelist{}
- % \end{macrocode}
- % \end{macro}
- %
- % \begin{macro}{\ifnot@excluded}
- % \changes{v1.7a}{92/02/26}{Replaced with newdoc version.}
- % \begin{macro}{\expanded@notin}
- % Now we take a look at the \verb+\index@excludelist+ to see
- % whether a macro name saved in \verb+\macro@namepart+ should
- % produce an index entry. This macro is a pseudo \verb+\if+; it
- % should expand to \verb+\iftrue+ or \verb+\iffalse+ depending on
- % the contents of \verb+\index@excludelist+.
- % \begin{macrocode}
- \begingroup
- % \end{macrocode}
- % First we change "\catcode"s so that "\" is `other' and "|" a
- % temporary for the escape character. This is necessary since our
- % macro names are stored that way in the "\index@excludelist".
- % \begin{macrocode}
- \catcode`\|=0%
- \catcode`\\=12
- % \end{macrocode}
- % Then we define "\ifnot@excluded" to call "\expanded@notin" with
- % two arguments: the first is the string "\" followed by the
- % contents of "\macro@namepart" followed by a "," and the second is
- % "\the" followed by "\index@excludelist". To achieve the expansion
- % of "\macro@namepart", i.e.\ to pass its contents, we need a
- % suitable number of "\expandafter"s.
- % \SpecialEscapechar{\|}
- % \begin{macrocode}
- |gdef|ifnot@excluded{|expandafter
- |expanded@notin|expandafter{|expandafter
- \|macro@namepart,}{|the|index@excludelist}}
- |endgroup
- % \end{macrocode}
- % The macro "\expanded@notin" now does the dirty work. It first
- % defines a macro "\in@@" with a very special parameter text. If
- % you look closely "\in@@" has three arguments, the first one is
- % delimited by the first argument of "\expanded@notin" (i.e.\ by
- % the string starting with a "\" and ending with a "," above), the
- % second is undelimited (which means that it will get the next
- % token after our string, and the third is delimited again and will
- % get the rest up to the token "\in@@". In other words the token
- % "\in@@" is also used as a delimiter.
- % \begin{macrocode}
- \def\expanded@notin#1#2{%
- \def\in@@##1#1##2##3\in@@{%
- % \end{macrocode}
- % Now the replacement text simply compares the second argument
- % (i.e.\ the undelimited one after our string) to the token
- % "\expanded@notin". This is an unclosed "\ifx" statement which
- % means that this macro behaves similar to a normal \TeX{}
- % conditional.
- % \begin{macrocode}
- \ifx\expanded@notin##2}%
- % \end{macrocode}
- % After all these preparations we call "\in@@". First we expand the
- % token after "\in@@" (which is "\the" from the second argument to
- % "\expanded@notin"). As a result we get the contents of the
- % "\index@excludelist" inserted after "\in@@". After this contents
- % we add once more the string we are looking for, then the token
- % "\expanded@notin" and finally the token "\in@@".
- % \begin{macrocode}
- \expandafter\in@@#2#1\expanded@notin\in@@}
- % \end{macrocode}
- % Now what happens when the macro "\in@@" above gets called? The
- % first argument to "\in@@" is delimited by our string. In other
- % words it will get everything from the contents of
- % "\index@excludelist" before this string. If the string is not in
- % "\index@excludelist" then it gets the whole contents, since after
- % it we had inserted the string one more. In this case the next
- % token is "\expanded@notin" which gets assigned to the second
- % argument and the third argument will be empty. If, on the other
- % hand, the string was inside "\index@excludelist" then the second
- % argument will not be the token "\expanded@notin" and the third
- % argument will be all the garbage up to "\in@@". Therefore testing
- % the seconded argument, as done in the definition of "\in@@" will
- % tell us whether or not the string is in "\index@includelist" and
- % this was exactly what we wanted. (Deep breath.) You got
- % that?\footnote{\TeX{}book page 125. The code described above is
- % originally due to Michael Spivak who used a similar method within
- % the \AmSTeX{} macros.}
- % \end{macro}
- % \end{macro}
- %
- %
- %
- %
- %
- %
- %
- % \subsection{Macros for generating index entries}
- %
- % Here we provide default definitions for the macros invoked to create
- % index entries; these are either invoked explicitly, or automatically
- % by \verb+\scan@macro+. As already mentioned, the definitions given
- % here presuppose that the \verb+.idx+ file will be processed by
- % Chen's {\sf makeindex} program --- they may be redefined for use
- % with the user's favourite such program.
- %
- % To assist the reader in locating items in the index, all such
- % entries are sorted alphabetically {\em ignoring\/} the initial
- % `\verb+\+'; this is achieved by issuing an \verb+\index+ command
- % which contains the `actual' operator for {\sf makeindex}. The
- % default value for the latter operator is `\verb+@+', but the latter
- % character is so popular in \LaTeX\ style files that it is necessary
- % to substitute another character. This is indicated to {\sf
- % makeindex} by means of an `index style file'; the character selected
- % for this function is \verb+=+, and therefore this character too must
- % be specially treated when it is met in a \TeX\ command. A suitable
- % index style file is provided amongst the supporting files for this
- % style file in {\tt gind.ist} and is generated from this source by
- % processing with {\tt docstrip} to extract the module {\bf gind}. A
- % similar style file {\tt gglo.ist} is supplied for sorting the change
- % information in the glossary file and is extracted as module {\bf
- % gglo}. First of all we add some information to the front of the
- % {\tt .ist} files. \changes{v1.7a}{92/03/11}{Gglo.ist and gind.ist
- % now derivable from doc.doc with docstrip.}
- % \begin{macrocode}
- %</style>
- %<+gind|gglo>%% This is a MAKEINDEX style file which should be used to
- %<+gind>%% generate the formatted index for use with the doc
- %<+gglo>%% generate the formatted change history for use with the doc
- %<+gind|gglo>%% style option. The TeX commands used below are defined in
- %<+gind|gglo>%% doc.sty. The commands for MAKEINDEX like `level'
- %<+gind|gglo>%% `item_x1' are described in `` Makeindex, A General
- %<+gind|gglo>%% Purpose, Formatter-Independent Index Processor'' by
- %<+gind|gglo>%% Pehong Chen.
- %<+gind|gglo>
- % \end{macrocode}
- %
- % \begin{macro}{\actualchar}
- % \begin{macro}{\quotechar}
- % \begin{macro}{\levelchar}
- % First come the definitions of \verb+\actualchar+,
- % \verb+\quotechar+ and \verb+\levelchar+. Note, that our defaults
- % are not the ones used by the {\sf makeindex} program without a
- % style file.
- % \begin{macrocode}
- %<*style>
- \@ifundefined{actualchar}{\def\actualchar{=}}{}
- \@ifundefined{quotechar}{\def\quotechar{!}}{}
- \@ifundefined{levelchar}{\def\levelchar{>}}{}
- %</style>
- %<+gind|gglo>actual '='
- %<+gind|gglo>quote '!'
- %<+gind|gglo>level '>'
- %<*style>
- % \end{macrocode}
- % \end{macro}
- % \end{macro}
- % \end{macro}
- %
- %
- % \begin{macro}{\encapchar}
- % The {\sf makeindex} default for the \verb+\encapchar+ isn't
- % changed.
- % \begin{macrocode}
- \@ifundefined{encapchar}{\def\encapchar{|}}{}
- % \end{macrocode}
- % \end{macro}
- %
- %
- % \begin{macro}{\verbatimchar}
- % We also need a special character to be used as a delimiter for
- % the \verb+\verb*+ command used in the definitions below.
- % \begin{macrocode}
- \@ifundefined{verbatimchar}{\def\verbatimchar{+}}{}
- % \end{macrocode}
- % \end{macro}
- %
- %
- % \begin{macro}{\SpecialIndex}
- % The \verb+\SpecialIndex+ command creates index entries for
- % macros. If the argument is \verb+\+$xyz$, the command produces
- % \verb|\indexentry{|$xyz$\verb|=\verb!*+\|$xyz$\verb|+}{|$n$\verb|}|
- % given the above defined defaults for \verb+\actualchar+,
- % \verb+\quotechar+ and \verb+\verbatimchar+. We first remove the
- % initial `\verb+\+' to get a better index.
- % \changes{v1.5s}{89/11/05}{Support for code line no. (Undoc)}
- % \begin{macrocode}
- \def\SpecialIndex#1{\@bsphack\special@index{\expandafter\@gobble
- \string#1\actualchar
- % \end{macrocode}
- % Then follows the actual entry. A \verb+\quotechar+ is placed
- % before the \verb+*+ to allow its use as a special {\sf makeindex}
- % character. Again \verb+\@bsphack+ and \verb+\@esphack+ are used
- % to make the macros invisible.
- % \begin{macrocode}
- \string\verb\quotechar*\verbatimchar\string#1\verbatimchar}%
- \@esphack}
- % \end{macrocode}
- % \end{macro}
-
- % \begin{macro}{\SpecialMainIndex}
- % \begin{macro}{\SpecialUsageIndex}
- % The \verb+\SpecialMainIndex+ macro is used to cross-reference the
- % names introduced by the {\sf macro} environment. The action is
- % as for \verb+\SpecialIndex+, except that {\sf makeindex} is
- % instructed to `encap'sulate the entry with the string
- % \verb+|main+ to cause it to generate a call of the \verb+\main+
- % macro.
- % \changes{v1.5s}{89/11/05}{Support for code line no. (Undoc)}
- % \begin{macrocode}
- \def\SpecialMainIndex#1{\@bsphack\special@index{\expandafter\@gobble
- \string#1\actualchar
- \string\verb
- \quotechar*\verbatimchar
- \string#1\verbatimchar
- \encapchar main}%
- \@esphack}
- % \end{macrocode}
- % The \verb+\SpecialUsageIndex+ is literally the same---only we use
- % {\tt usage} instead of {\tt main}.
- % \begin{macrocode}
- \def\SpecialUsageIndex#1{\@bsphack\index{\expandafter\@gobble\string#1%
- \actualchar\string\verb\quotechar*\verbatimchar
- \string#1\verbatimchar
- \encapchar usage}\@esphack}
- % \end{macrocode}
- % \end{macro}
- % \end{macro}
- %
- %
- %
- %
- % \begin{macro}{\SpecialEnvIndex}
- % Indexing environments is done a little bit differently; we produce
- % two index entries with the \verb+\SpecialEnvIndex+ macro:
- % \begin{macrocode}
- \def\SpecialEnvIndex#1{\@bsphack
- % \end{macrocode}
- % First we sort the environment under its own name stating in the
- % actual entry that this is an environment.
- % \begin{macrocode}
- \index{#1\actualchar{\tt #1} (environment)\encapchar usage}%
- % \end{macrocode}
- % The second entry is sorted as a subitem under the key
- % `environments:'.
- % \begin{macrocode}
- \index{environments:\levelchar{\tt #1}\encapchar usage}\@esphack}
- % \end{macrocode}
- % Because both entries correspond to `descriptions' of the
- % environment, we encapsulate the page numbers with the
- % \verb+\usage+ macro.
- % \end{macro}
- %
- %
- %
- % \begin{macro}{\SortIndex}
- % This macro is used to generate the index entries for any
- % single-character command that \verb+\scan@macro+ encounters. The
- % first parameter specifies the lexical order for the character,
- % whilst the second gives the actual characters to be printed in
- % the entry. It can also be used directly to generate index entries
- % which differ in sort key and actual entry.
- % \begin{macrocode}
- \def\SortIndex#1#2{\index{#1\actualchar#2}}
- % \end{macrocode}
- % \end{macro}
- %
- %
- %
- % \begin{macro}{\it@is@a}
- % This macro is supposed to produce a correct \verb+\SortIndex+
- % entry for a given character. Since this character might be
- % recognised as a `command' character by the index program used,
- % all characters are quoted with the \verb+\quotechar+.
- % \changes{v1.5s}{89/11/05}{Support for code line no. (Undoc)}
- % \begin{macrocode}
- \def\it@is@a#1{\special@index{\quotechar #1\actualchar
- \string\verb\quotechar*\verbatimchar
- \quotechar\bslash\quotechar#1\verbatimchar}}
- % \end{macrocode}
- % \end{macro}
- %
- %
- %
- % \begin{macro}{\LeftBraceIndex}
- % \changes{v1.5s}{89/11/05}{Support for code line no. (Undoc)}
- % \begin{macro}{\RightBraceIndex}
- % \changes{v1.5s}{89/11/05}{Support for code line no. (Undoc)}
- % These two macros fix the problems with {\sf makeindex}. Note the
- % `hack' with \verb+\iffalse}\fi+ to satisfy both \TeX{} and the
- % {\sf makeindex} program. When this is written to the {\tt .idx}
- % file \TeX{} will see both braces (so we get a balanced text).
- % {\sf makeindex} will also see balanced braces but when the actual
- % index entry is again processed by \TeX{} the brace in between
- % \verb+\iffalse+ \verb+\fi+ will vanish.
- % \begin{macrocode}
- \@ifundefined{LeftBraceIndex}{\def\LeftBraceIndex{%
- \special@index{\bgroup\actualchar\string\verb\quotechar*\verbatimchar
- \quotechar\bslash{\verbatimchar\string\iffalse}\string\fi}}}{}
-
- \@ifundefined{RightBraceIndex}{\def\RightBraceIndex{%
- \special@index{\egroup\actualchar\string\iffalse{\string\fi\string\verb
- \quotechar*\verbatimchar\quotechar\bslash}\verbatimchar}}}{}
- % \end{macrocode}
- % \end{macro}
- % \end{macro}
- %
- %
- % \begin{macro}{\PercentIndex}
- % \changes{v1.5s}{89/11/05}{Support for code line no. (Undoc)}
- % \changes{v1.7c}{92/03/25}{Default now for bug-fixed makeindex}
- % By default we assume a version of {\sf makeindex} without the
- % percent bug is being used.
- % \begin{macrocode}
- \@ifundefined{PercentIndex}
- {\def\PercentIndex{\it@is@a\percentchar}}{}
- % \end{macrocode}
- % \end{macro}
- % \begin{macro}{\OldMakeindex}
- % \changes{v1.7c}{92/03/26}{Replaced `NewMakeIndex.}
- % \begin{macro}{\percentchar}
- % Here is one solution for the percent bug in {\sf makeindex}. The
- % macro \verb+\percentchar+ denotes a \verb+%+$_{12}$.
- % Calling this from a style option or the driver file sets things
- % up appropriately.\label{bug:fixes}
- % \begin{macrocode}
- \def\OldMakeindex{\def\PercentIndex{%
- \special@index{\quotechar\percentchar\actualchar\string\verb
- \quotechar*\verbatimchar\quotechar\bslash
- \percentchar\percentchar\verbatimchar}}}
- {\catcode`\%=12 \gdef\percentchar{%}}
- % \end{macrocode}
- % \end{macro}
- % \end{macro}
- %
- %
- %
- %
- %
- %
- % \subsection{Redefining the {\sf index} environment}
- %
- %\changes{v1.4r}{89/04/22}{twocols env. placed into separate file}
- %\changes{v1.4?}{89/04/19}{use DEK's algorithm and implement
- % a twocols env.}
- %\changes{v1.4?}{89/04/16}{changes to the index env.}
- %\changes{v1.5a}{89/04/26}{Now input multicol.sty instead of
- % multcols.sty}
- % \begin{macro}{\ifhave@multicol}
- % \changes{v1.7a}{92/03/04}{Added to support avoiding multicol.sty}
- % By default
- % the index is set in three columns, and will start on the same
- % page as, and underneath, the last part of the text of the
- % documented style file, if possible. The last page will be
- % reformatted with balanced columns. This requires the {\sf
- % multicols} environment which is described elsewhere.
- % ^^A described elsewhere (see page~\xrefto{mittelbachmulti}).
- % So that {\sf doc} can be run independently of {\tt multicol.sty}
- % we first check for its existence and set the "have@multicol" flag
- % appropriately for use below.
- % \begin{macrocode}
- \newif\ifhave@multicol
- \openin\@ne multicol.sty
- \ifeof\@ne \else \have@multicoltrue \fi
- \closein\@ne \relax
- % \end{macrocode}
- % On systems where it is impossible to determine the existence of a
- % file using the above method the docstrip program can be directed
- % to force using multicol by including the next line.
- % \changes{v1.7i}{92/07/14}{Force use of multicol.sty if necessary.}
- % \begin{macrocode}
- %<+multicol>\have@multicoltrue
- % \end{macrocode}
- % \end{macro}
- % If we found {\tt multicol.sty} we use it. It would be nice to
- % delay this (and the re-definition of "theindex") until we knew
- % whether an index was actually required \ldots
- % \begin{macrocode}
- \ifhave@multicol \input{multicol.sty} \fi
- % \end{macrocode}
- %
- % \begin{macro}{\IndexMin}
- % \begin{macro}{\c@IndexColumns}
- % \changes{v1.4t}{89/04/24}{Counter added.}
- % If {\tt multicol} is in use,
- % when the index is started we compute the remaining space on the
- % current page; if it is greater than \verb+\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
- % \verb+\c@IndexColumns+ which can be changed with a
- % \verb+\setcounter+ declaration.
- % \begin{macrocode}
- \newdimen\IndexMin \IndexMin = 80pt
- \newcount\c@IndexColumns \c@IndexColumns = 3
- % \end{macrocode}
- % \end{macro}
- % \end{macro}
- %
- %
- % \begin{macro}{\theindex}
- % Now we start the multi-column mechanism, if appropriate. We use the
- % \verb+\c@IndexColumns+ \LaTeX{} counter declared above to denote
- % the number of columns and insert the `index prologue' text (which
- % might contain a \verb+\section+ call, etc.). See the default
- % definition for an example.
- % \changes{v1.4t}{89/04/24}{Incorporated new multicols env.}
- % \changes{v1.5a}{89/04/26}{Call multicols first}
- % \changes{v1.6e}{91/04/03}{Turned into env definition.}
- % \changes{v1.7a}{92/03/04}{Include test for multicols.}
- % \begin{macrocode}
- \ifhave@multicol
- \renewenvironment{theindex}
- {\begin{multicols}\c@IndexColumns[\index@prologue][\IndexMin]%
- % \end{macrocode}
- % Then we make a few last minute assignments to read the individual
- % index \verb+\item+$\,$s and finish off by ignoring any initial
- % space.
- % \begin{macrocode}
- \IndexParms \let\item\@idxitem \ignorespaces}%
- % \end{macrocode}
- % \end{macro}
- % \begin{macro}{\endtheindex}
- % \changes{v1.4t}{89/04/24}{Incorporated new multicols env.}
- % At the end of the index, we have only to end the {\sf multicols}
- % environment.
- % \begin{macrocode}
- {\end{multicols}}
- % \end{macrocode}
- % If we can't use {\sf multicols} we warn the user and use an
- % environment that's basically the one from {\tt article.sty}.
- % \begin{macrocode}
- \else
- \typeout{Can't find multicols.sty -- will use normal index layout if
- necessary.}
- \def\theindex{\@restonecoltrue\if@twocolumn\@restonecolfalse\fi
- \columnseprule \z@ \columnsep 35\p@
- \twocolumn[\index@prologue]%
- \IndexParms \let\item\@idxitem \ignorespaces}
- \def\endtheindex{\if@restonecol\onecolumn\else\clearpage\fi}
- \fi
- % \end{macrocode}
- % \end{macro}
- % Here are the necessary {\sf makeindex} declarations. We disable
- % scanning of macro names inside the index with "\scan@allowedfalse\n"
- % to avoid recursion.
- % \begin{macrocode}
- %</style>
- %<+gind>preamble
- %<+gind>"\n \\begin{theindex} \n \\makeatletter\\scan@allowedfalse\n"
- %<+gind>postamble
- %<+gind>"\n\n \\end{theindex}\n"
- %<*style>
- % \end{macrocode}
- %
- %
- % \begin{macro}{\IndexPrologue}
- % \begin{macro}{\index@prologue}
- % The \verb+\IndexPrologue+ macro is used to place a short message
- % into the document above the index. It is implemented by
- % redefining \verb+\index@prologue+, a macro which holds the
- % default text. We'd better make it a \verb+\long+ macro to allow
- % \verb+\par+ commands in its argument.
- % \begin{macrocode}
- \long\def\IndexPrologue#1{\@bsphack\def\index@prologue{#1}\@esphack}
- % \end{macrocode}
- % Now we test whether the default is already defined by another
- % style file. If not we define it.
- % \begin{macrocode}
- \@ifundefined{index@prologue}
- {\def\index@prologue{\section*{Index}%
- \markboth{Index}{Index}%
- The italic numbers denote the pages where the
- corresponding entry is described,
- numbers underlined point to the definition,
- all others indicate the places where it is used.
- }}{}
- % \end{macrocode}
- % \end{macro}
- % \end{macro}
- %
- %
- %
- % \begin{macro}{\IndexParms}
- % These are some last-minute assignments for formatting the index
- % entries. They are defined in a separate macro so that a user can
- % substitute different definitions. We start by defining the
- % various parameters controlling leading and the separation between
- % the two columns. The entire index is set in \verb+\small+ size.
- % \begin{macrocode}
- \@ifundefined{IndexParms}
- {\def\IndexParms{%
- \parindent \z@
- \columnsep 15pt
- \parskip 0pt plus 1pt
- \rightskip 15pt
- \mathsurround \z@
- \parfillskip=-15pt
- \small
- % \end{macrocode}
- % \begin{macro}{\@idxitem}
- % \begin{macro}{\subitem}
- % \begin{macro}{\subsubitem}
- % Index items are formatted with hanging indentation for any items
- % which may require more than one line.
- % \begin{macrocode}
- \def\@idxitem{\par\hangindent 30pt}%
- % \end{macrocode}
- % Any sub-item in the index is formatted with a 15pt indentation
- % under its main heading.
- % \begin{macrocode}
- \def\subitem{\@idxitem\hspace*{15pt}}%
- % \end{macrocode}
- % Whilst sub-sub-items go in a further 10pt.
- % \begin{macrocode}
- \def\subsubitem{\@idxitem\hspace*{25pt}}%
- % \end{macrocode}
- % \begin{macro}{\indexspace}
- % The {\sf makeindex} program generates an \verb+\indexspace+
- % before each new alphabetic section commences. After this final
- % definition we end the \verb+\@ifundefined+ and the definition of
- % \verb+\IndexParms+.
- % \begin{macrocode}
- \def\indexspace{\par\vspace{10pt plus 2pt minus 3pt}}%
- }}{}
- % \end{macrocode}
- % \end{macro}
- % \end{macro}
- % \end{macro}
- % \end{macro}
- % \end{macro}
- %
- %
- % \begin{macro}{\efill}
- % This definition of \verb+\efill+ is intended to be used after
- % index items which have no following text (for example, ``{\it
- % see\/}'' entries). It just ensures that the current line is
- % filled, preventing ``\verb+Underfull \hbox+'' messages.
- % \begin{macrocode}
- \def\efill{\hfill\nopagebreak}%
- %</style>
- %<+gind|gglo>item_x1 "\\efill \n \\subitem "
- %<+gglo>item_x2 "\\ "
- %<+gind>item_x2 "\\efill \n \\subsubitem "
- %<*style>
- % \end{macrocode}
- % \end{macro}
- %
- %
- %
- % \begin{macro}{\pfill}
- % \begin{macro}{\dotfil}
- % \begin{macro}{\dotfill}
- % The following definitions provide the \verb+\pfill+ command; if
- % this is specified in the index style file to {\sf 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.
- % \begin{macrocode}
- \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}%
- %</style>
- %<+gind|gglo>delim_0 "\\pfill "
- %<+gind|gglo>delim_1 "\\pfill "
- %<+gind|gglo>delim_2 "\\pfill "
- %<*style>
- % \end{macrocode}
- % \end{macro}
- % \end{macro}
- % \end{macro}
- %
- %
- %
- % \begin{macro}{\*}
- % Here is the definition for the \verb+\*+ macro. It isn't used in
- % this set of macros.
- % \begin{macrocode}
- \def\*{\leavevmode\lower.8ex\hbox{$\,\widetilde{\ }\,$}}
- % \end{macrocode}
- % \end{macro}
- %
- %
- % \begin{macro}{\main}
- % The {\it defining\/} entry for a macro name is flagged with the
- % string {\tt\encapchar main}\footnote{With the current definition
- % of {\tt\bslash encapchar} substituted for {\tt\encapchar}} in the
- % \verb+\index+ command; {\sf makeindex} processes this so that the
- % \verb+\main+ macro will be invoked to underline the page
- % number(s) on which the {\em definition\/} of the macro will be
- % found.
- % \begin{macrocode}
- \@ifundefined{main}{\def\main#1{\underline{#1}}}{}
- % \end{macrocode}
- % \end{macro}
- %
- % \begin{macro}{\usage}
- % The \verb+\usage+ macro is used to indicate entries describing
- % the usage of a macro. The corresponding page number(s) will be
- % set in {\it italics}.
- % \begin{macrocode}
- \@ifundefined{usage}{\def\usage#1{{\it #1}}}{}
- % \end{macrocode}
- % \end{macro}
- %
- %
- % \begin{macro}{\PrintIndex}
- % \changes{v1.5k}{89/09/04}{`printindex changed to `PrintIndex.}
- % \changes{v1.7a}{92/02/26}{Documentation moved to interface section.}
- % \begin{macro}{\printindex}
- % This is the same as "\printindex" in the {\sf makeidx} style.
- % \begin{macrocode}
- \def\PrintIndex{\@input{\jobname.ind}}
- % \end{macrocode}
- % Since this macro was called \verb+\printindex+ in older versions
- % of {\tt doc.sty} we also provide the following definition.
- % \begin{macrocode}
- \def\printindex{\typeout{\string\printindex\space is obsolete!}%
- \typeout{Please use \string\PrintIndex\space
- if you are a macro implementor^^J
- or get a newer version of the documented
- software if you are a user}%
- \PrintIndex}
- % \end{macrocode}
- % \end{macro}
- % \end{macro}
- %
- % We want headings in the index (and changes list) according to the
- % initial character of the next block of entries and have to instruct
- % {\sf makeindex} appropriately. Unfortunately the specification for
- % this changed sometime between versions 2.4 and 2.11 of {\sf
- % makeindex}. We provide both ways of doing it but unfortunately this
- % will always produce a warning message from {\sf makeindex}. This is
- % for older versions:
- % \changes{v1.7h}{92/07/01}{Turn off headings in gls file}
- % \begin{macrocode}
- %</style>
- %<+gind,gglo>% The next lines will produce some warnings when
- %<+gind,gglo>% running Makeindex as they try to cover two different
- %<+gind,gglo>% versions of the program:
- %<+gind,gglo>lethead_prefix "{\\bf\\hfil "
- %<+gind,gglo>lethead_suffix "\\hfil}\\nopagebreak\n"
- %<+gind>lethead_flag 1
- %<+gglo>lethead_flag 0
- % \end{macrocode}
- % This works for newer ones:
- % \begin{macrocode}
- %<+gind,gglo>heading_prefix "{\\bf\\hfil "
- %<+gind,gglo>heading_suffix "\\hfil}\\nopagebreak\n"
- %<+gind>headings_flag 1
- %<+gglo>headings_flag 0
- %<*style>
- % \end{macrocode}
- %
- %
- %
- % \subsection[Dealing with the change history]
- % {Dealing with the change history\footnotemark}
- % \footnotetext{The whole section was proposed by Brian {\sc Hamilton
- % Kelly}. He also documented and debugged the macros as
- % well as many other parts of this style option.}
- %
- % To provide a change history log, the \verb+\changes+ command has
- % been introduced. This takes three arguments, respectively, the
- % version number of the file, the date of the change, and some detail
- % regarding what change has been made. The first 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. However, note that older versions of Chen's standard {\sf
- % makeindex}
- % program limit any textual field to just 64 characters; therefore,
- % is important that the number of characters in the second and third
- % parameters should not exceed 61 altogether (to allow for the
- % parentheses placed around the date).
- %
- % \begin{macro}{\changes}
- % \changes{BHK}{89/04/26}{Documented {\tt\protect\bslash changes}
- % command.}
- % \changes{BHK}{89/04/26}{Changed definition of
- % {\tt\protect\bslash protect}.}
- % The output of
- % the \verb+\changes+ command goes into the \meta{Glossary\_File}
- % and therefore uses the normal \verb+\glossaryentry+
- % commands.\footnote{Note that a recent change in \LaTeX{} 2.09
- % changed the command name in the {\tt.glo} file from {\tt\bslash
- % indexentry} to {\tt\bslash glossaryentry}. It is therefore
- % necessary to have a special {\sf makeindex} style file called {\tt
- % gglo.ist} to process this file correctly.} Thus {\sf makeindex}
- % or a similar program can be used to process the output into a
- % sorted ``glossary''. The \verb+\changes+ command commences by
- % taking the usual measures to hide its spacing, and then redefines
- % \verb+\protect+ for use within the argument of the generated
- % \verb+\indexentry+ command.
- %
- % We re-code nearly all chars found in \verb+\sanitize+ to letter
- % since the use of special style options which make some characters
- % active might upset the \verb+\changes+ command when writing its
- % entries to the file. However we have to leave \verb+%+ as comment
- % and \verb*+ + as \meta{space} otherwise chaos will happen.
- % And, of course the \verb+\+ should be available as escape
- % character.
- % \changes{v1.5v}{90/01/28}{`Re-code a lot of chars.}
- % \changes{v1.5m}{89/09/20}{`actualchar in second level removed.}
- % \changes{v1.5o}{89/09/24}{New sorting.}
- % \changes{v1.6c}{90/06/29}{Again new sorting.}
- % \begin{macrocode}
- \def\changes{\@bsphack\begingroup\@sanitize
- \catcode`\\\z@ \catcode`\ 10 \MakePercentIgnore
- \changes@}
- \def\changes@#1#2#3{%
- \def\protect##1{\string##1\space}%
- \edef\@tempa{\noexpand\glossary{#1\levelchar
- \expandafter\@gobble
- \saved@macroname\actualchar
- \string\verb\quotechar*%
- \verbatimchar\saved@macroname
- \verbatimchar:\levelchar #3}}%
- \@tempa\endgroup\@esphack}
- % \end{macrocode}
- % \begin{macro}{\saved@macroname}
- % \changes{BHK}{89/04/26}{Provided for sorting outside
- % {\sf macro} environment}
- % The entries are sorted for convenience by the name of the most
- % recently introduced macroname (i.e., that in the most recent
- % \verb+\begin{macro}+ command). We therefore provide
- % \verb+\saved@macroname+ to record that argument, and provide a
- % default definition in case \verb+\changes+ is used outside a {\sf
- % macro} environment. (This is a {\em wicked\/} hack to get such
- % entries at the beginning of the sorted list! It works providing no
- % macro names start with "!" or \verb+"+.)
- % \changes{v1.7a}{92/03/02}{Changed string used for better sorting.}
- % \begin{macrocode}
- \def\saved@macroname{"General"}
- % \end{macrocode}
- % \end{macro}
- %
- % \begin{macro}{\RecordChanges}
- % \changes{BHK}{89/04/26}{Renames former {\tt\protect\bslash
- % PrintChanges} command.}
- % To cause the changes to be written (to a {\tt.glo}) file, we
- % define \verb+\RecordChanges+ to invoke \LaTeX's usual
- % \verb+\makeglossary+ command.
- % \begin{macrocode}
- \let\RecordChanges\makeglossary
- % \end{macrocode}
- % \end{macro}
- % \end{macro}
- %
- %
- % \begin{macro}{\GlossaryMin}
- % \changes{BHK}{89/04/26}{Added to support
- % {\tt\protect\bslash changes}.}
- % \begin{macro}{\c@GlossaryColumns}
- % \changes{BHK}{89/04/26}{Added to support
- % {\tt\protect\bslash changes}.}
- % The remaining macros are all analogues of those used for the {\sf
- % 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 \verb+\GlossaryMin+ then the first part of
- % the glossary will be placed in the available space. The number
- % of columns set are controlled by the counter
- % \verb+\c@GlossaryColumns+ which can be changed with a
- % \verb+\setcounter+ declaration.
- % \begin{macrocode}
- \newdimen\GlossaryMin \GlossaryMin = 80pt
- \newcount\c@GlossaryColumns \c@GlossaryColumns = 2
- % \end{macrocode}
- % \end{macro}
- % \end{macro}
- %
- %
- % \begin{macro}{\theglossary}
- % \changes{BHK}{89/04/26}{Added to support
- % {\tt\protect\bslash changes}.}
- % \changes{v1.5p}{89/09/28}{Now call `multicols first.}
- % \changes{v1.6e}{91/04/03}{Turned into env definition.}
- % \changes{v1.7a}{92/03/10}{Changed to work without multicols if
- % necessary.}
- % \begin{macro}{\endglossary}
- % \changes{BHK}{89/04/26}{Added to support
- % {\tt\protect\bslash changes}.}
- % The environment {\sf theglossary} is defined in the same manner
- % as the {\sf theindex} environment.
- % \begin{macrocode}
- \ifhave@multicol
- \newenvironment{theglossary}{%
- \begin{multicols}\c@GlossaryColumns
- [\glossary@prologue][\GlossaryMin]%
- \GlossaryParms \let\item\@idxitem \ignorespaces}%
- {\end{multicols}}
- \else
- \newenvironment{theglossary}{%
- \@restonecoltrue\if@twocolumn\@restonecolfalse\fi
- \columnseprule \z@ \columnsep 35\p@
- \twocolumn[\glossary@prologue]%
- \GlossaryParms \let\item\@idxitem \ignorespaces}
- {\if@restonecol\onecolumn\else\clearpage\fi}
- \fi
- % \end{macrocode}
- % \end{macro}
- % \end{macro}
- % Here are the necessary {\sf makeindex} declarations with scanning
- % disabled as for the index.
- % \begin{macrocode}
- %</style>
- %<+gglo>preamble
- %<+gglo>"\n \\begin{theglossary} \n
- %<+gglo> \\makeatletter\\scan@allowedfalse\n"
- %<+gglo>postamble
- %<+gglo>"\n\n \\end{theglossary}\n"
- % \end{macrocode}
- % This difference from {\tt gind.ist} is necessary if you have an
- % up-to-date \LaTeX.
- % \begin{macrocode}
- %<+gglo>keyword "\\glossaryentry"
- %<*style>
- % \end{macrocode}
- %
- %
- % \begin{macro}{\GlossaryPrologue}
- % \changes{BHK}{89/04/26}{Added to support
- % {\tt\protect\bslash changes}.}
- % \begin{macro}{\glossary@prologue}
- % \changes{BHK}{89/04/26}{Added to support
- % {\tt\protect\bslash changes}.}
- % The \verb+\GlossaryPrologue+ macro is used to place a short
- % message above the glossary into the document. It is implemented
- % by redefining \verb+\glossary@prologue+, a macro which holds the
- % default text. We better make it a long macro to allow
- % \verb+\par+ commands in its argument.
- % \begin{macrocode}
- \long\def\GlossaryPrologue#1{\@bsphack
- \def\glossary@prologue{#1}%
- \@esphack}
- % \end{macrocode}
- % Now we test whether the default is already defined by another
- % style file. If not we define it.
- % \begin{macrocode}
- \@ifundefined{glossary@prologue}
- {\def\glossary@prologue{\section*{{Change History}}%
- \markboth{{Change History}}{{Change History}}%
- }}{}
- % \end{macrocode}
- % \end{macro}
- % \end{macro}
- %
- % \begin{macro}{\GlossaryParms}
- % \changes{BHK}{89/04/26}{Added to support
- % {\tt\protect\bslash changes}.}
- % Unless the user specifies otherwise, we set the change history
- % using the same parameters as for the index.
- % \begin{macrocode}
- \@ifundefined{GlossaryParms}{\let\GlossaryParms\IndexParms}{}
- % \end{macrocode}
- % \end{macro}
- % \begin{macro}{\PrintChanges}
- % \changes{BHK}{89/04/26}{Added to support
- % {\tt\protect\bslash changes}.}
- % To read in and print the sorted change history, just put the
- % \verb+\PrintChanges+ command as the last (commented-out, and thus
- % executed during the documentation pass through the file) command
- % in your style file. Alternatively, this command may form one of
- % the arguments of the \verb+\StopEventually+ command, although a
- % change history is probably {\em not\/} required if only the
- % description is being printed.
- %
- % The command assumes that {\sf makeindex} or some other program
- % has processed the {\tt.glo} file to generate a sorted {\tt.gls}
- % file.
- % \begin{macrocode}
- \def\PrintChanges{\@input{\jobname.gls}}
- % \end{macrocode}
- % \end{macro}
- %
- %
- %
- %
- %
- % \subsection{Bells and whistles}
- %
- % \begin{macro}{\StopEventually}
- % \changes{v1.5k}{89/09/04}{Support for checksum.}
- % \begin{macro}{\Finale}
- % \changes{v1.5k}{89/09/04}{Support for checksum.}
- % \changes{v1.5z}{90/04/22}{Define `Finale globally.}
- % \begin{macro}{\OnlyDescription}
- % Here is the default definition for \verb+\StopEventually+, we
- % simply save its argument in the macro \verb+\Finale+.
- % \begin{macrocode}
- \long\def\StopEventually#1{\@bsphack\gdef\Finale{#1%
- % \end{macrocode}
- % But \verb+\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 call \verb+\check@checksum+ at this
- % point.
- % \begin{macrocode}
- \check@checksum}%
- % \end{macrocode}
- % On the other hand: \verb+\StopEventually+ is more or less a
- % dividing point between description and code. So we start to look
- % for the check-sum of the documented file by calling
- % \verb+\init@checksum+.
- % \begin{macrocode}
- \init@checksum
- \@esphack}
- % \end{macrocode}
- % When the user places an \verb+\OnlyDescription+ declaration in
- % the driver file the document should only be typeset up to
- % \verb+\StopEventually+. We therefore have to redefine this macro.
- % \begin{macrocode}
- \def\OnlyDescription{\@bsphack\long\def\StopEventually##1{%
- % \end{macrocode}
- % In this case the argument of \verb+\StopEventually+ should be set
- % and afterwards \TeX{} should stop reading from this file.
- % Therefore we finish this macro with
- % \begin{macrocode}
- ##1\endinput}\@esphack}
- % \end{macrocode}
- % \end{macro}
- % \end{macro}
- % \end{macro}
- %
- % \begin{macro}{\meta}
- % \changes{v1.4t}{89/04/24}{Macro added.}
- % \begin{macro}{\m@ta}
- % \changes{v1.5w}{90/02/03}{Breaks at space allowed.}
- % \changes{v1.6a}{90/05/24}{Extra space bug corrected.}
- % The \verb+\meta+ macro is a bit tricky. We want to allow line
- % breaks at blanks in the argument but we don't want a break
- % in between. We therefore define \verb+\meta+ in a way that a
- % \verb*+ + is active when the argument is scanned. Words are then
- % scanned into \verb+\hbox+es. The active \verb*+ + will end the
- % preceding \verb+\hbox+ add an ordinary space and open a new
- % \verb+\hbox+. In this way breaks are only possible at spaces. It
- % would be even better to forbid page breaks but this is not
- % possible in an all cases.
- % \begin{macrocode}
- \begingroup
- \obeyspaces%
- \catcode`\^^M\active%
- % \end{macrocode}
- % We have to be careful to end all lines with a \verb+%+ sign in
- % this definition.
- % \begin{macrocode}
- \gdef\meta{\begingroup\obeyspaces\catcode`\^^M\active%
- \let^^M\do@space\let \do@space%
- % \end{macrocode}
- % To allow to break up words inside the \verb+\meta+ command we
- % redefine the \verb+\-+ command. It now has to end the last
- % open box, add a discretionary and start the next one for the
- % rest of the current word. See below for more details.
- %
- % Finally we call \verb+\m@ta+ which
- % will scan the argument of \verb+\meta+.
- % \begin{macrocode}
- \def\-{\egroup\discretionary{-}{}{}\hbox\bgroup\it}%
- \m@ta}%
- \endgroup
- % \end{macrocode}
- % We start \verb+\m@ta+ by opening an \verb+\hbox+.
- % Inside this box there will be angle brackets and the argument
- % typeset in italic typeface. If there are no spaces or \verb+\-+
- % commands in this argument the result will be a single box. But
- % when a space is encountered (which has \verb+\catcode+ 13) then
- % it will expand into \verb+\do@space+ which will close the current
- % box, output a space (so that we have a legitimate break point, and
- % then opens an new box to catch the rest of the argument.
- % \changes{v1.6d}{90/11/16}{`leavevmode added.}
- % \begin{macrocode}
- \def\m@ta#1{\leavevmode\hbox\bgroup$\langle$\it#1\/$\rangle$\egroup
- % \end{macrocode}
- % Finally, we have to close the group which was started in
- % \verb+\meta+
- % to restore all our changes.
- % \begin{macrocode}
- \endgroup}
- % \end{macrocode}
- % The \verb+\do@space+ macro will produce the possible breakpoint
- % by ending the current box (\verb+\egroup+) and adding the
- % \verb+\space+
- % into the surrounding paragraph.
- % \begin{macrocode}
- \def\do@space{\egroup\space
- % \end{macrocode}
- % Then we start a new box, switching again to italic to catch the
- % rest of the argument of \verb+\meta+. But we also have to make sure
- % that any space following directly will be ignored. Therefore we
- % check the following token and discard it as long as it is a token
- % with the meaning \verb+\do@space+, i.e.\ a $\verb*+ +_{13}$ or a
- % $\verb+^^M+_{13}$.
- % \begin{macrocode}
- \hbox\bgroup\it\futurelet\next\sp@ce}
- \def\sp@ce{\ifx\next\do@space\expandafter\sp@@ce\fi}
- \def\sp@@ce#1{\futurelet\next\sp@ce}
- % \end{macrocode}
- % \end{macro}
- % \end{macro}
- %
- %
- % \begin{macro}{\IndexInput}
- % This next macro may be used to read in a separate file (possibly
- % a style file that is {\em not\/} documented by this means) and
- % set it verbatim, whilst scanning for macro names and indexing the
- % latter. This could be a useful first pass in preparing to
- % generate documentation for the file read.
- % \begin{macrocode}
- \def\IndexInput#1{%
- % \end{macrocode}
- % We commence by setting up a group, and initializing a
- % \verb+\trivlist+ as is normally done by a
- % \verb+\begin{macrocode}+ command.
- % \begin{macrocode}
- \begingroup \macro@code
- % \end{macrocode}
- % We also make spacing behave as in the {\sf macrocode}
- % environment, because otherwise all the spaces will be shown
- % explicitly.
- % \begin{macrocode}
- \frenchspacing \@vobeyspaces
- % \end{macrocode}
- % Then it only remains to read in the specified file, and finish
- % off the \verb+\trivlist+.
- % \changes{v1.5t}{89/11/07}{Call `endmacrocode instead of `endtrivlist.}
- % \begin{macrocode}
- \input{#1}\endmacrocode
- % \end{macrocode}
- % Of course, we need to finish off the group as well.
- % \begin{macrocode}
- \endgroup}
- % \end{macrocode}
- % \end{macro}
- %
- %
- % \begin{macro}{\maketitle}
- % 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
- % \verb+\relax+. We must cancel anything that may have been put
- % into \verb+\@thanks+, etc., otherwise {\em all\/} titles will
- % carry forward any earlier such setting!
- % \changes{v1.5j}{89/06/09}{thispagestyle plain removed}
- % \begin{macrocode}
- \def\maketitle{\par
- \begingroup \def \thefootnote {\fnsymbol {footnote}}%
- \setcounter {footnote}\z@
- \def \@makefnmark {\hbox to \z@{$^{\@thefnmark }$\hss }}%
- \if@twocolumn \twocolumn [\@maketitle ]%
- \else \newpage \global \@topnum \z@ \@maketitle \fi
- % \end{macrocode}
- % \changes{v1.5k}{89/09/04}{Added {\tt\protect\bslash ps@titlepage}}
- % For special formatting requirements (such as in TUGboat), we use
- % pagestyle \verb+titlepage+ for this; this is later defined to be
- % \verb+plain+, unless already defined, as, for example, by
- % \verb+ltugboat.sty+.
- % \begin{macrocode}
- \thispagestyle{titlepage}\@thanks \endgroup
- % \end{macrocode}
- % 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:
- % \begin{macrocode}
- \setcounter {footnote}\z@
- \gdef\@date{\today}\gdef\@thanks{}%
- \gdef\@author{}\gdef\@title{}}
- % \end{macrocode}
- % \end{macro}
- %
- %
- % \begin{macro}{\ps@titlepage}
- % \changes{v1.5k}{89/09/04}{Added {\tt\protect\bslash ps@titlepage}}
- % 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 style option such as {\sf
- % ltugboat} can define this macro in advance. However, if no such
- % definition exists, we use pagestyle {\tt plain} for title pages.
- % \begin{macrocode}
- \@ifundefined{ps@titlepage}
- {\let\ps@titlepage=\ps@plain}{}
- % \end{macrocode}
- % \end{macro}
- %
- % \begin{macro}{\MakeShortVerb}
- % \changes{v1.7a}{92/02/27}{Added (from newdoc but now alters
- % `dospecials, `@sanitize).}
- % This arranges an abbreviation for "\verb" such that if you say
- % "\MakeShortVerb{\"\meta{c}"}" subsequently using
- % \meta{c}\meta{text}\meta{c} is equivalent to
- % "\verb"\meta{c}\meta{text}\meta{c}.\footnote{Warning:
- % the commentary in the rest of this section was written by Dave
- % Love.} In addition, the fact
- % that \meta{c} is made active is recorded for the benefit of the
- % {\sf verbatim} and {\sf macrocode} environments.
- % Note particularly that the definitions below are global.
- % The first thing we do (it needn't be first) is to record
- % the---presumably new---special character in "\dospecials" and
- % "\@sanitize" using "\add@special".
- % \begin{macrocode}
- \def\MakeShortVerb#1{%
- \typeout{*** Made \expandafter\@gobble\string#1\space a short
- reference for \string\verb \on@line\space ***}%
- \add@special{#1}%
- % \end{macrocode}
- % Then the character's current catcode is stored in "\cc\"\meta{c}.
- % \begin{macrocode}
- \expandafter
- \xdef\csname cc\string#1\endcsname{\the\catcode`#1}%
- % \end{macrocode}
- % The character is spliced into the definition using the same trick as
- % used in "\verb" (for instance), having activated "~" in a group.
- % \begin{macrocode}
- \begingroup
- \catcode`\~\active \lccode`\~`#1%
- \lowercase{%
- % \end{macrocode}
- % The character's old meaning is recorded in "\ac\"\meta{c} prior to
- % assigning it a new one.
- % \begin{macrocode}
- \global\expandafter\let
- \csname ac\string#1\endcsname~%
- \gdef~{\verb~}}%
- \endgroup
- % \end{macrocode}
- % Finally the character is made active.
- % \begin{macrocode}
- \global\catcode`#1\active}
- % \end{macrocode}
- % \end{macro}
- % \begin{macro}{\DeleteShortVerb}
- % \changes{v1.7a}{92/02/27}{Added (from newdoc but now alters
- % `dospecials, `@sanitize).}
- % Here's the means of undoing a "\MakeShortVerb", for instance in a
- % region where you need to use the character outside a verbatim
- % environment. It arranges for "\dospecials" and "\@sanitize" to be
- % altered appropriately, restores the saved catcode and, if necessary,
- % the character's meaning (as stored by
- % "\MakeShortVerb"). If the catcode wasn't stored in
- % "\cc\"\meta{c} (by "\MakeShortVerb") the command is silently
- % ignored.
- % \changes{v1.7a}{92/02/28}{Check for previous matched `MakeShortVerb
- % to avoid error.}
- % \begin{macrocode}
- \def\DeleteShortVerb#1{%
- \expandafter\ifx\csname cc\string#1\endcsname\relax
- \else
- \typeout{*** Deleted \expandafter\@gobble\string#1\space as short
- reference for \string\verb \on@line\space ***}%
- \rem@special{#1}%
- \global\catcode`#1\csname cc\string#1\endcsname
- \ifnum\catcode`#1=\active
- \begingroup
- \catcode`\~\active \lccode`\~`#1%
- \lowercase{%
- \global\expandafter\let\expandafter~%
- \csname ac\string#1\endcsname}%
- \endgroup \fi \fi}
- % \end{macrocode}
- % \end{macro}
- % \begin{macro}{\add@special}
- % \changes{v1.7a}{92/02/27}{Added for short verb facility.}
- % This helper macro adds its argument to the
- % "\dospecials" macro which is conventionally used by verbatim macros
- % to alter the catcodes of the currently active characters. We need
- % to add "\do\"\meta{c} to the expansion of "\dospecials" after
- % removing the character if it was already there to avoid multiple
- % copies building up should "\MakeShortVerb" not be balanced by
- % "\DeleteShortVerb" (in case anything that uses "\dospecials" cares
- % about repetitions).
- % \begin{macrocode}
- \def\add@special#1{%
- \rem@special{#1}%
- \expandafter\gdef\expandafter\dospecials\expandafter
- {\dospecials \do #1}%
- % \end{macrocode}
- % Similarly we have to add "\@makeother\"\meta{c} to "\@sanitize"
- % (which is used in things like "\index" to re-catcode all special
- % characters except braces).
- % \begin{macrocode}
- \expandafter\gdef\expandafter\@sanitize\expandafter
- {\@sanitize \@makeother #1}}
- % \end{macrocode}
- % \end{macro}
- % \begin{macro}{\rem@special}
- % \changes{v1.7a}{92/02/27}{Added for short verb facility.}
- % The inverse of "\add@special" is slightly trickier. "\do" is
- % re-defined to expand to nothing if its argument is the character of
- % interest, otherwise to expand simply to the argument. We can then
- % re-define "\dospecials" to be the expansion of itself. The space
- % after "=`##1" prevents an expansion to "\relax"!
- % \begin{macrocode}
- \def\rem@special#1{%
- \def\do##1{%
- \ifnum`#1=`##1 \else \noexpand\do\noexpand##1\fi}%
- \xdef\dospecials{\dospecials}%
- % \end{macrocode}
- % Fixing "\@sanitize" is the same except that we need to re-define
- % "\@makeother" which obviously needs to be done in a group.
- % \begin{macrocode}
- \begingroup
- \def\@makeother##1{%
- \ifnum`#1=`##1 \else \noexpand\@makeother\noexpand##1\fi}%
- \xdef\@sanitize{\@sanitize}%
- \endgroup}
- % \end{macrocode}
- % \end{macro}
- % \begin{macro}{\MakeShortverb}
- % \begin{macro}{\DeleteShortverb}
- % \changes{v1.7a}{92/02/27}{Added (from newdoc).}
- % These commands from {\sf newdoc} are now obsolete.
- % \begin{macrocode}
- \def\MakeShortverb{\typeout{*** Switch to \noexpand\MakeShortVerb
- syntax, this is obsolete ***}\MakeShortVerb}
- \def\DeleteShortverb{\typeout{*** Switch to \noexpand\DeleteShortVerb
- syntax, this is obsolete ***}\DeleteShortVerb}
- % \end{macrocode}
- % \end{macro}
- % \end{macro}
- %
- %
- %
- % \subsection[Providing a checksum and character table]
- % {Providing a checksum and character table\footnotemark}
- % \footnotetext{Warning: the commentary in this section was
- % written by Dave Love. }
- %
- %
- % \begin{macro}{\init@checksum}
- % The checksum mechanism works by counting backslashes in the
- % macrocode. This initialises the count (when called from
- % "\StopEventually").
- % \changes{v1.5k}{89/09/04}{Macro added to support checksum.}
- % \begin{macrocode}
- \def\init@checksum{\relax
- \global\bslash@cnt\z@}
- % \end{macrocode}
- % \end{macro}
- %
- %
- % \begin{macro}{\check@checksum}
- % \changes{v1.5k}{89/09/04}{Macro added to support checksum.}
- % This reports the sum compared with the value ("\bslash@cnt") the
- % file advertises. It's called from "\Finale" (if that hasn't been
- % re-defined).
- % \begin{macrocode}
- \def\check@checksum{\relax
- \ifnum\check@sum=\z@
- \typeout{**********************************}%
- \typeout{* This macro file has no checksum!}%
- \typeout{* The checksum should be \the\bslash@cnt!}%
- \typeout{**********************************}%
- \else
- \ifnum\check@sum=\bslash@cnt
- \typeout{*******************}%
- \typeout{* Checksum passed *}%
- \typeout{*******************}%
- \else
- \errhelp\wrong@checksum
- \errmessage{Checksum not passed
- (\the\check@sum<>\the\bslash@cnt)}%
- \fi
- \fi
- \global\check@sum\z@}
- % \end{macrocode}
- % \end{macro}
- %
- %
- % \begin{macro}{\check@sum}
- % \changes{v1.5k}{89/09/04}{Macro added to support checksum.}
- % \begin{macro}{\bslash@cnt}
- % \changes{v1.5k}{89/09/04}{Macro added to support checksum.}
- % \begin{macro}{\wrong@checksum}
- % \changes{v1.5k}{89/09/04}{Macro added to support checksum.}
- % We need to define the counter for the number of backslashes counted
- % ("\bslash@cnt") and the value advertised by the file ("\check@sum")
- % as well as a help message for when things go wrong.
- % \begin{macrocode}
- \newcount\check@sum \check@sum = \z@
- \newcount\bslash@cnt \bslash@cnt = \z@
- \newhelp\wrong@checksum
- {The currently documented file seems to be wrong.^^J%
- Try to get a correct version.}%
- % \end{macrocode}
- % \end{macro}
- % \end{macro}
- % \end{macro}
- %
- %
- % \begin{macro}{\CheckSum}
- % \changes{v1.5k}{89/09/04}{Macro added to support checksum.}
- % This is the interface to setting "\check@sum".
- % \begin{macrocode}
- \def\CheckSum#1{\@bsphack\global\check@sum#1\relax\@esphack}
- % \end{macrocode}
- % \end{macro}
- %
- %
- %
- % \begin{macro}{\step@checksum}
- % \changes{v1.5k}{89/09/04}{Macro added to support checksum.}
- % This advances the count when a backslash is encountered in the
- % macrocode.
- % \begin{macrocode}
- \def\step@checksum{\global\advance\bslash@cnt\@ne}
- % \end{macrocode}
- % \end{macro}
- %
- % \begin{macro}{\CharacterTable}
- % The user interface to the character table-checking does some
- % "\catcode"ing and then compares the following table with the
- % stored version. We need to have "@" of type ``other'' within the
- % table since this is the way it is usually returned when reading
- % in a normal document. To nevertheless have a private letter we
- % use "~" for this purpose. "~" does no harm as a ``letter'' as it
- % comes last in the table and therefore will not gobble following
- % space.
- % \changes{v1.5m}{89/09/20}{Macro added to check character translation
- % problems.}
- % \changes{v1.5q}{89/11/01}{Made character table more readable.}
- % \changes{v1.5t}{89/11/07}{Make \string\~{} letter in chartable
- % macros.}
- % \changes{v1.5u}{89/11/14}{Made @ other in default table.}
- % \begin{macrocode}
- \def\CharacterTable{\begingroup \CharTableChanges \character@table}
- % \end{macrocode}
- % \end{macro}
- % \def\MakePrivateLetters{\catcode`\~=11\makeatletter}
- % \begin{macro}{\character@table}
- % This does the work of comparing the tables and reporting the result.
- % Note that the following code is enclosed in a group
- % with "~" catcoded to letter.
- % \begin{macrocode}
- \begingroup
- \catcode`\~=11
- \gdef\character@table#1{\def\used~table{#1}%
- \ifx\used~table\default~table
- \typeout{***************************}%
- \typeout{* Character table correct *}%
- \typeout{***************************}%
- \else
- \errhelp\wrong@table
- \errmessage{Character table corrupted}%
- \show\default~table
- \show\used~table
- \fi
- \endgroup}
- % \end{macrocode}
- % \end{macro}
- %
- % \begin{macro}{\CharTableChanges}
- % When the character table is read in we need to scan it with a
- % fixed set of "\catcode"s. The reference table below was defined by
- % assuming the normal "\catcode"s of \TeX{}, i.e.\ "@" is of type
- % other and the only token of type ``letter'' are the usual letters
- % of the alphabet. If, for some reason, other characters are made
- % ``letters'' then their "\catcode"s need to be restored before
- % checking the table. Otherwise spaces in the table are gobbled and
- % we get the information that the tables are different, even if
- % they are actually equal. For this reason "\CharTableChanges" can
- % be set up to locally restore the "\catcode"s of such ``letters''
- % to ``other''.
- % \begin{macrocode}
- \global\let\CharTableChanges\@empty
- % \end{macrocode}
- % \end{macro}
- %
- % \begin{macro}{\default~table}
- % Here's what the table {\em should\/} look like (modulo spaces).
- % \begin{macrocode}
- \makeatother
- \gdef\default~table
- {Upper-case \A\B\C\D\E\F\G\H\I\J\K\L\M\N\O\P\Q\R\S\T\U\V\W\X\Y\Z
- Lower-case \a\b\c\d\e\f\g\h\i\j\k\l\m\n\o\p\q\r\s\t\u\v\w\x\y\z
- Digits \0\1\2\3\4\5\6\7\8\9
- Exclamation \! Double quote \" Hash (number) \#
- Dollar \$ Percent \% Ampersand \&
- Acute accent \' Left paren \( Right paren \)
- Asterisk \* Plus \+ Comma \,
- Minus \- Point \. Solidus \/
- Colon \: Semicolon \; Less than \<
- Equals \= Greater than \> Question mark \?
- Commercial at \@ Left bracket \[ Backslash \\
- Right bracket \] Circumflex \^ Underscore \_
- Grave accent \` Left brace \{ Vertical bar \|
- Right brace \} Tilde \~}
- \endgroup
- % \end{macrocode}
- % \end{macro}
- % \let\MakePrivateLetters=\makeatletter
- %
- % \begin{macro}{\wrong@table}
- % \changes{v1.7a}{92/02/28}{Moved to where the catcodes are right so it
- % works.}
- % We need a help message in case of problems.
- % \begin{macrocode}
- \newhelp\wrong@table{Some of the ASCII characters are corrupted.^^J
- I now \string\show\space you both tables for comparison.}
- % \end{macrocode}
- % \end{macro}
- %
- %
- % \subsection[Attaching line numbers to code lines]
- % {Attaching line numbers to code lines\footnotemark}
- % \footnotetext{Warning: the commentary was written by Dave
- % Love.}
- %
- %
- % The code in this section allows index entries to refer to code line
- % numbers---the number of the first line of macrocode in the {\sf macro}
- % environment.
- %
- %
- % \begin{macro}{\codeline@index}
- % Indexing by code line is controlled by the "codeline@index" switch.
- % \changes{v1.5s}{89/11/05}{Support for code line no. (Undoc)}
- % \changes{v1.7a}{92/02/24}{Documented code line no. support.}
- % \begin{macrocode}
- \newif\ifcodeline@index \codeline@indexfalse
- % \end{macrocode}
- % \end{macro}
- % \begin{macro}{\codeline@wrindex}
- % The code index entries are written out by "\special@index". If
- % indexing is by code line this is "\let" to "\codeline@wrindex";
- % if indexing is by page it is just "\index". However, if
- % "\nofiles" is given, we omit writing such an index entry at all.
- % \changes{v1.7j}{92/08/14}{Added `if@filesw.}
- % \begin{macrocode}
- \def\codeline@wrindex#1{\if@filesw
- \immediate\write\@indexfile
- {\string\indexentry{#1}%
- {\number\c@CodelineNo}}\fi}
- % \end{macrocode}
- % \end{macro}
- % \begin{macro}{\special@index}
- % By default no index entries are written out.
- % \begin{macrocode}
- \let\special@index = \@gobble
- % \end{macrocode}
- % \end{macro}
- % \begin{macro}{\CodelineIndex}
- % \changes{v1.5u}{89/11/14}{Added `PageIndex and `CodelineIndex (Undoc)}
- % This switches on use of the index file with "\makeindex", sets the
- % switch to indicate code line numbering and defines "\special@index"
- % appropriately.
- % \begin{macrocode}
- \def\CodelineIndex{\makeindex
- \codeline@indextrue
- \let\special@index\codeline@wrindex}
- % \end{macrocode}
- % \end{macro}
- % \begin{macro}{\PageIndex}
- % "\PageIndex" is similar.
- % \begin{macrocode}
- \def\PageIndex{\makeindex
- \codeline@indexfalse
- \let\special@index\index}
- % \end{macrocode}
- % \end{macro}
- %
- %
- % \begin{macro}{\c@CodelineNo}
- % \changes{v1.5l}{89/09/10}{Counter added to support code line numbers.}
- % \changes{v1.5y}{90/02/24}{Default changed.}
- % \changes{v1.6b}{90/06/15}{`rm moved before `scriptsize to
- % avoid unnecessary fontwarning.}
- % We need a counter to keep track of the line number.
- % \begin{macrocode}
- \newcount\c@CodelineNo \c@CodelineNo\z@
- % \end{macrocode}
- % \end{macro}
- % \begin{macro}{\theCodelineNo}
- % \changes{v1.7a}{92/02/25}{Existing definition not overwritten.}
- % \changes{v1.7a}{92/03/12}{Use `reset@font for NFSS.}
- % This provides a hook to control the format of line numbers which may
- % be defined in a style file.
- % \begin{macrocode}
- \@ifundefined{theCodelineNo}
- {\ifx\selectfont\undefined
- \def\theCodelineNo{\rm\scriptsize\arabic{CodelineNo}}%
- \else
- \def\theCodelineNo{\reset@font\scriptsize\arabic{CodelineNo}}%
- \fi}
- {}
- % \end{macrocode}
- % \end{macro}
- %
- %
- %
- %
- % \subsection{Layout Parameters for documenting style files}
- %
- % \begin{macro}{\tolerance}
- % People documenting style files would probably rather have things
- % ``sticking out'' in overfull \verb+\hbox+es and poorish spacing,
- % because they probably don't want to spend a lot of time on making
- % all the line breaks perfect!
- % \begin{macrocode}
- \tolerance=1000\relax
- % \end{macrocode}
- % \end{macro}
- %
- % \DeleteShortVerb{\"}
- %
- % The following \verb+\mathcode+ definitions allow the characters
- % `\verb+\+'
- % and `{\tt @}' to appear in \verb+\tt+ font when invoked in math
- % mode;\footnote{You may wonder why the definitions state that both
- % characters belong to the {\em variable family\/}
- % (i.e.\ the number 7 in front). The reason is this:
- % Originally the {\tt\bslash mathcode} of
- % {\tt\bslash} was defined to be {\tt "075C},
- % i.e.\ ordinary character number 92 (hex 5C) in
- % math family number 7 which is the typewriter family in
- % standard \LaTeX.
- % But this file should not depend on this specific
- % setting, so I changed these {\tt\bslash mathcode}$\,$s
- % to work with any family assignments. For an example
- % see the article about the new font selection scheme.}
- % particularly for something like $\verb+\@abc+=1$.
- %
- % If an {\em old\/} version of the {\sf german} style option is in
- % force, then the `\verb+"+' character is active and would upset the
- % definition of the \meta{16-bit number} quantities below, therefore
- % we change the \verb+\catcode+ of \verb+"+ inside a group, and use
- % \verb+\global+.
- % \begin{macrocode}
- { \catcode`\"=12
- \global\mathcode`\\="705C \global\mathcode`\@="7040 }
- % \end{macrocode}
- % \MakeShortVerb{\"}
- %
- % \begin{macro}{\DocstyleParms}
- % This macro can be used, for example, to assign new values to
- % \verb+\MacrocodeTopsep+ and \verb+\MacroIndent+ and some other
- % internal registers. If it is already defined, the default
- % definition won't be carried out. Note that it is necessary to
- % assign new values via this macro if it should be done in a style
- % file (like {\tt ltugboat.sty} for example) since the registers are
- % undefined before {\tt doc.sty} is read in. The default values
- % for the internal registers are scattered over this file.
- % \changes{v1.5u}{89/11/14}{`DocStyleParms now empty}
- % \begin{macrocode}
- \@ifundefined{DocstyleParms}{}{}
- % \end{macrocode}
- % Now we allow overwriting the values by calling
- % \verb+\DocstyleParms+.
- % \begin{macrocode}
- \DocstyleParms \let\DocstyleParms\relax
- % \end{macrocode}
- % \end{macro}
- %
- %
- %
- % \begin{macro}{\AmSTeX}
- % \changes{v1.5j}{89/06/09}{Macro AmsTeX renamed to AmSTeX}
- % \begin{macro}{\BibTeX}
- % \begin{macro}{\SliTeX}
- % Here are a few definitions which can usefully be employed when
- % documenting style files: now we can readily refer to \AmSTeX,
- % \BibTeX\ and \SliTeX, as well as the usual \TeX\ and \LaTeX.
- % \begin{macrocode}
- \@ifundefined{AmSTeX}
- {\def\AmSTeX{\leavevmode\hbox{$\cal A\kern-.2em\lower.376ex%
- \hbox{$\cal M$}\kern-.2em\cal S$-\TeX}}}{}
- \@ifundefined{BibTeX}
- {\def\BibTeX{{\rm B\kern-.05em{\sc i\kern-.025em b}\kern-.08em%
- T\kern-.1667em\lower.7ex\hbox{E}\kern-.125emX}}}{}
- \@ifundefined{SliTeX}
- {\def\SliTeX{{\rm S\kern-.06emL\kern-.18em\raise.32ex\hbox
- {\sc i}\kern -.03em\TeX}}}{}
- % \end{macrocode}
- % \end{macro}
- % \end{macro}
- % \end{macro}
- % \begin{macro}{\PlainTeX}
- % \changes{v1.5g}{89/05/07}{space between plain and TeX changed.}
- % \begin{macro}{\Web}
- % There's even a \PlainTeX{} and a \Web.
- % \begin{macrocode}
- \@ifundefined{PlainTeX}{\def\PlainTeX{{\sc Plain}\kern2pt\TeX}}{}
- \@ifundefined{Web}{\def\Web{{\sc Web}}}{}
- % \end{macrocode}
- % \end{macro}
- % \end{macro}
- %
- %
- % \subsection{Changing the {\tt\protect\bslash catcode} of \%}
- %
- % \begin{macro}{\MakePercentIgnore}
- % \begin{macro}{\MakePercentComment}
- % And finally the most important bit: we change the \verb+\catcode+
- % of `\verb+%+' so that it is ignored (which is how we are able to
- % produce this document!). We provide two commands to do the actual
- % switching.
- %^^A The \verb+\MakePercentIgnore+ is then called as the
- %^^A last command in this file.
- % \begin{macrocode}
- \def\MakePercentIgnore{\catcode`\%9\relax}
- \def\MakePercentComment{\catcode`\%14\relax}
- % \end{macrocode}
- % \end{macro}
- % \end{macro}
- %
- % \begin{macro}{\DocInput}
- % The two macros above are now used to define the
- % \verb+\DocInput+ macro which was introduced in version v1.5l
- % (or so) of
- % the {\tt doc} style option. In older versions
- % \verb+\MakePercentIgnore+ was placed at the very end of {\tt
- % doc.sty}.
- % \begin{macrocode}
- \def\DocInput#1{\MakePercentIgnore\input{#1}\MakePercentComment}
- % \end{macrocode}
- % \end{macro}
- % We can now finish the {\tt docstrip} main module.
- % \begin{macrocode}
- %</style>
- % \end{macrocode}
- %
- %
- % \begin{macro}{\on@line}
- % Finally something for people with an old \LaTeX:
- % \changes{v1.7k}{92/08/24}{Macro and test added.}
- % \begin{macrocode}
- \ifx\on@line\undefined
- \def\on@line{ on input line \the\inputlineno}
- \errhelp{Support for input line numbers has been added
- to latex.tex <dec91>.^^J^^J%
- Please update to a newer LaTeX release.}
- \errmessage{Obsolete LaTeX release (older than Dec.91)}
- \fi
- % \end{macrocode}
- % \end{macro}
- %
- % \Finale
- % \PrintIndex \PrintChanges
- \endinput
-