diff options
author | Karl Berry <karl@freefriends.org> | 2007-10-24 23:29:14 +0000 |
---|---|---|
committer | Karl Berry <karl@freefriends.org> | 2007-10-24 23:29:14 +0000 |
commit | c5199a08bb8edc16006e6a5fc72cc704d7eae776 (patch) | |
tree | 729971967a99529b74598d27dd7cc585e0218e41 /Master/texmf-dist/doc/latex/cweb-latex/src | |
parent | bfaf341636c39818969bd384bb1e5fda9195f3a5 (diff) |
new (to TL) cweb-latex package (4feb07)
git-svn-id: svn://tug.org/texlive/trunk@5271 c570f23f-e606-0410-a88d-b1316a301751
Diffstat (limited to 'Master/texmf-dist/doc/latex/cweb-latex/src')
44 files changed, 7489 insertions, 0 deletions
diff --git a/Master/texmf-dist/doc/latex/cweb-latex/src/Imakefile b/Master/texmf-dist/doc/latex/cweb-latex/src/Imakefile new file mode 100644 index 00000000000..64bd313b8f5 --- /dev/null +++ b/Master/texmf-dist/doc/latex/cweb-latex/src/Imakefile @@ -0,0 +1,106 @@ +XCOMM $Id: Imakefile,v 1.7 1995/11/21 00:41:31 schrod Exp $ +/* ---------------------------------------------------------------------- */ + +/* + * Imakefile for cweb bundle + * + * (history at end) + */ + + +/* needed modules */ +#include <TeX.imk> +#include <makeprog.imk> + +#include <install/TeX.imk> + +XCOMM ------------------------------------------------------------ +XCOMM continue in Imakefile + + +/* exported files */ + +MpDoc2TeXTarget(cwebbase,tex) +MpDoc2TeXTarget(cweb,cls) +MpDoc2TeXTarget(cwebparts,sty) + +TeXTarget(latex,cweb-user) +TeXTarget(latex,cweb-conf) + + +/* for some time, until it's moved to an own bundle */ + +AllTarget(keyvald.sty) + +keyvald.sty: keyvald.dtx + latex keyvald.ins + +clean:: + $(RM) keyvald.sty + + +/* installation */ + +#ifdef ITI_Site + DESTDIR = /archives/tex/Install +#endif + +InstallTeXLaTeX(cweb,cweb.cls cwebbase.tex cwebarray.sty keyvald.sty \ + cwebparts.sty cwbl-german.sty) +InstallTeXDoc(latex/styles/cweb,cweb-user.dvi) +InstallTeXDoc(latex/styles/cweb,cweb-conf.dvi) + + +/* internal documentation */ + +TeXFileTarget(latex,cweb,doc) +TeXFileTarget(latex,cwebbase,doc) +TeXFileTarget(latex,cwebparts,doc) +TeXTarget(latex,cweave-spec) + + +/* distribution */ + +distrib:: all cweb.dvi cwebbase.dvi cwebparts.dvi + mv cweb-*.dvi *.cls cwebbase.tex keyvald.sty .. + cp cwebarray.sty cwbl-german.sty .. + cp $(HOME)/Notes/README.latex.doc-src README + $(MAKE) realclean + +realclean:: clean + $(RM) Makefile + + + +/* ============================================================ + * + * $Log: Imakefile,v $ + * Revision 1.7 1995/11/21 00:41:31 schrod + * German language package. + * + * Revision 1.6 1995/09/17 14:21:25 schrod + * Add support for German CWEB documents. + * + * Revision 1.5 1995/09/16 17:00:18 schrod + * Extract cwebbase.tex from cweb.cls, add cwebparts.sty. + * + * Revision 1.4 1995/08/29 17:34:39 schrod + * LaTeX CWEB, version 1.0. + * + * Revision 1.3 1995/08/29 15:19:47 schrod + * Added (hint to non-existence of) description of the internal interface. + * + * Revision 1.2 1995/08/25 19:00:32 schrod + * Add enhanced version of keyval package, with ability to specify + * default action for unknown keyword options. + * + * Revision 1.1 1995/08/08 00:14:24 schrod + * Updated to \LaTeXe{}, the |cweb| style is now a document class. Used + * my standard templates for that, no changes in functionality. + * + + * ------------------------------------------------------------ + * Local Variables: + * mode: Indented-Text + * End: + */ diff --git a/Master/texmf-dist/doc/latex/cweb-latex/src/README b/Master/texmf-dist/doc/latex/cweb-latex/src/README new file mode 100644 index 00000000000..0e7a4836715 --- /dev/null +++ b/Master/texmf-dist/doc/latex/cweb-latex/src/README @@ -0,0 +1,31 @@ +latex-style:src/README 29 Aug 95 +---------------------------------------------------------------------- + +Please note: This directory is part of the distribution primarily for your +information. It's not prepared to be usable by you, sorry. + +In particular, probably you won't have tools and style files I take +for granted during development. Please don't complain, either use the +supplied installable files (as listed in ../MANIFEST) or search & get +the missing files. I will neither stop adding this directory to the +distribution, nor will I add all missing files. + +On the one hand I won't restrict my own development environment by not +using available tools, on the other hand I'll want to make my source +available for those who have a similar environment to mine. + + +For your orientation: + + *.doc -- master file for TeX macro files + *.dvi -- typeset version of *.doc + *-user.tex -- user manual + *-conf.tex -- (optional) internal interface description + *.tex -- (optional) other documentation parts + *.{fig,latex} -- (optional) figures prepared with xfig + *.sty -- needed to process the above documents + + TODO -- (optional) personal notes for future revisions + Imakefile -- does not use MIT's Imake templates + style/ -- (optional) files for AUC-TeX + test/ -- test cases diff --git a/Master/texmf-dist/doc/latex/cweb-latex/src/TODO b/Master/texmf-dist/doc/latex/cweb-latex/src/TODO new file mode 100644 index 00000000000..82fddd2ee5c --- /dev/null +++ b/Master/texmf-dist/doc/latex/cweb-latex/src/TODO @@ -0,0 +1,168 @@ +% $Id: TODO,v 1.10 1995/11/30 15:44:51 schrod Exp $ +%---------------------------------------------------------------------- + + +---------------------------------------------------------------------- + +functionality [problem] + +In plain CWEB chunks are not split over pages, LaTeX CWEB does so. Add +an option that re-establishs the behaviour of plain CWEB. + +Problem reported by Denis B. R\"ogel <denis.roegel@loria.fr> + +---------------------------------------------------------------------- + +functionality [problem] + +User index would use with standard LaTeX markup (\index) would use +*.idx as raw index file. This file is used by CWEAVE as tagged index +file. + +Shall I redefine \makeindex? + +---------------------------------------------------------------------- + +functionality [proposal] + +Add support to create the illusion of more than one refinement in a +chunk. + +\cwebNoChunkIntro; see cwebx.sty. +Proposed by Christopher Higgins <C.Higgins@cit.gu.edu.au>. +Ack him for that. + +---------------------------------------------------------------------- + +functionality [proposal] + +Add support for CWEB code examples. + +For that code from cwebparts.doc must be moved to cwebbase.doc; see +cwebx.sty. That code has also several improvements over cwebparts.doc! +cwebparts.doc is needed later for typesetting sections, etc. Well, if +that can really be realized in a document/project independent way. +Don't know if I really need it. (Except as a user front end. But I +can also use cwebbase for that.) + +---------------------------------------------------------------------- + +functionality [proposal] + +Add support for HyperTeX. + +Proposed by Christopher Higgins <C.Higgins@cit.gu.edu.au>. + +---------------------------------------------------------------------- + +list headers [error] + +Use \chaptermark & \sectionmark, in \CwebListSection. + +Maybe discard \@mkboth from \CwebIdIndex, \CwebRefList, etc. It's +supposed to be changed by a pagestyle, but this doesn't happen with +fancyheadings anyhow -- and face it, fancyheadings is the method of +choice for head-/footline creation. + +The actual code was caused by stupid copying from LaTeX standard +classes; with the change it's possible to have other list heading +layout (add to toc, other marks, etc.) + +But check if I can use that in \CwebIdIndex: There \CwebListSection is +used in an \twocolumn optional arg, that will eventually trigger the +output routine and will be set as a float. Does marks correctly +migrate out of floats? + +---------------------------------------------------------------------- + +layout [error] + +In chunks without documentation, a refinement is too near to the chunk +marker. In addition, the distance is not independent of +\CwebIndentUnit, making it impossible to change the latter. + +NOTE: Might be a problem of \cwebNoChunk that should start a +paragraph. *When* will CWEAVE add `\4' after `\B'? Have to make tests +for that. How can I detect an empty documentation section? + +\B kind of repaired in cwebx.sty. + +---------------------------------------------------------------------- + +layout [error] + +Chunk numbers are not exactly aligned in all circumstances, see fixme +remark in cwebbase.doc. + +---------------------------------------------------------------------- + +interface [idea] + +Name all \cwbb@... macros used in other modules \Cwbb... (or even +\cwbb...). + +---------------------------------------------------------------------- + +markup [proposal] + +Support delimiting section titles by braces. + +Maybe one needs an option to turn that on. + +Supported by Peter Wainwright <prw-b@ugrad.cs.york.ac.uk>. + +---------------------------------------------------------------------- + +documentation [ack] + +Who sent the mail concerning \hb@xt@? Add to acks. Change log message +for it, too (cweb.doc: 3.4; cwebbase.doc: 1.4). + +---------------------------------------------------------------------- + +documentation [problem] + +cweb-conf.tex: + +Cseq name of chunk marker (paragraph symbol). +Ack Christopher Higgins \path|<c.higgins@cit.gu.edu.au>|. + +\CwebRankNoEject. (Implies explication of rank numbers. Should be +explicated in cwebbase.doc as well, btw.) +Ack Nicholas Urbanik \path|<nicku@iohk.com>| + +Support for another language must be stored in cwbl-*.sty. Check +cwbl-german.sty for the cseqs that must be redefined, list their +original English text for clarification. + +---------------------------------------------------------------------- + +distribution [problem] + +cweb.cls & cwebbase.tex is > 100KB now. Export it in uncommented form. + +---------------------------------------------------------------------- + +distribution [problem] + +It's better to collect stuff from contrib, and have each file that +shall be installed in top dir. + +This means a Makefile in contrib/; and changes in README, INSTALL, and +maybe even MANIFEST. + +---------------------------------------------------------------------- + +distribution [proposal] + +Add hint to TODO in README. Maybe point out that src/ sometimes holds +prereleases of future styles. + +---------------------------------------------------------------------- + +distribution [proposal] + +Install TODO in doc dir. The end user should see open problems and +plans, too. + +---------------------------------------------------------------------- diff --git a/Master/texmf-dist/doc/latex/cweb-latex/src/cwbl-german.sty b/Master/texmf-dist/doc/latex/cweb-latex/src/cwbl-german.sty new file mode 100644 index 00000000000..68a7ce6c054 --- /dev/null +++ b/Master/texmf-dist/doc/latex/cweb-latex/src/cwbl-german.sty @@ -0,0 +1,109 @@ +% $Id: cwbl-german.sty,v 1.1 1995/09/17 14:21:25 schrod Exp $ +%---------------------------------------------------------------------- + +% +% LaTeX2e package cwbl-german +% adapt CWEB documents to language `german' +% +% [LaTeX2e] +% (history at end) + + +% Identification. The code below is explained in the implementation +% documentation of the rcs package. + +\begingroup + \def\RCSPackage#1#2 $#3: #4 #5\endRCS $#6: #7 #8\endRCS{% + \def\date{#4}\def\id{v#7}% + \ProvidesPackage{#1}[\date\space\id\space #2]% + } + + \RCSPackage{cwbl-german}{German CWEB documents} + $Date: 1995/09/17 14:21:25 $: 9999/00/00 \endRCS + $Revision: 1.1 $: 0.0 \endRCS +\endgroup + + +% +% What's this package for? +% +% XXX +% + + +% This class is supported. Send bug reports, comments and repairs. +% +% The reference version may be retrieved via anonymous ftp from +% |ftp.th-darmstadt.de| [130.83.47.112], directory |pub/tex/latex|. It's +% placed there as a gzipped tar file. (The information on the +% IP~number is dated 17 Sep 95. It might have changed, also +% this is very unlikely. Use your friendly nameserver.) + + +%%% ------------------------------------------------------------ + + +% at the end of chunks + +\def\CwebCRAlso{\CwebCrossRef{Siehe auch Baustein}} +\def\CwebCRsAlso{\CwebCrossRef{Siehe auch Bausteine}} + +\def\CwebCRCite{\CwebCrossRef{Dies wird zitiert im Baustein}} +\def\CwebCRsCite{\CwebCrossRef{Dies wird zitiert in den Bausteinen}} + +\def\CwebCRUse{\CwebCrossRef{Dies wird benutzt im Baustein}} +\def\CwebCRsUse{\CwebCrossRef{Dies wird benutzt in den Bausteinen}} + +\def\CwebCREt{ und~} +\def\CwebCRsEt{ und~} + + +% back matter + +\def\CwebCRChanged{% + \CwebCrossRef{% + Die folgenden Bausteine wurden durch ein Changefile ge\"andert:% + }% + } +\def\CwebChangedListName{Ge\"anderte Bausteine} + + +\def\CwebRefListName{Verzeichnis der Verfeinerungen} + +\def\CwebXRCite{\CwebCrossRef{Zitiert im Baustein}} +\def\CwebXRCites{\CwebCrossRef{Zitiert in den Bausteinen}} +\def\CwebXRUse{\CwebCrossRef{Benutzt im Baustein}} +\def\CwebXRUses{\CwebCrossRef{Benutzt in den Bausteinen}} + + +\def\CwebMacrosHere{% + \begingroup + \def\CwebRefNumber##1{}% + \CwebRefName :Pr\"aprozessor Definitionen\X + \endgroup + } + + + +\endinput + + +%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% +% +% $Log: cwbl-german.sty,v $ +% Revision 1.1 1995/09/17 14:21:25 schrod +% Add support for German CWEB documents. +% + + + +%%%------------------------------------------------------------ + +% Local Variables: +% mode: LaTeX +% TeX-brace-indent-level: 4 +% TeX-parse-self: nil +% TeX-auto-save: nil +% TeX-master: t +% page-delimiter: "^%%%" +% End: diff --git a/Master/texmf-dist/doc/latex/cweb-latex/src/cweave-spec.tex b/Master/texmf-dist/doc/latex/cweb-latex/src/cweave-spec.tex new file mode 100644 index 00000000000..c1f5b4f07ff --- /dev/null +++ b/Master/texmf-dist/doc/latex/cweb-latex/src/cweave-spec.tex @@ -0,0 +1,403 @@ +% $DocId: cweave-spec.tex,v 1.8 1995/09/12 22:56:02 schrod Exp $ +%------------------------------------------------------------ + +% +% specification of CWEAVE/TeX interface +% +% [LaTeX] +% (history at end) + + +% This file is either a subdocument of the style doc or a document on +% its own. In the former case it's a chapter, in the latter it's a +% ``normal'' LaTeX progltx document. +% If it's a subdocument, this file will be included after +% \begin{document}. We can detect this: \document redefines +% \documentclass to be \@twoclasseserror. Then we also have to define +% how this document is ended: Either by \endinput or by an additional +% revision log. +% Of course, this test works only if LaTeX 2e is used for processing. + +\expandafter\ifx \csname @twoclasseserror\endcsname \documentclass + + \let\endSubDocument=\endinput + + \chap The interface between \cweave{} and \TeX{}. + \label{sec:cweave-spec} + +\else + + \let\endSubDocument=\relax + + \documentclass{progltx} + + \usepackage{cweb-doc} % additional document-specific markup + \usepackage{a4-9} % Tschichold's A4 layout + + + \begin{document} + + \title{The Interface between \cweave{} and \TeX{}} + \author{Joachim Schrod% + \thanks{Email: \texttt{schrod@iti.informatik.th-darmstadt.de}}% + } + \RCSdate $DocDate: 1995/09/12 22:56:02 $ + + \maketitle + + \sect + +\fi + + + + +Here we present all tags output by \cweave{}% + \footnote{This information refers to \cweb{}~3.4. Differences to + other 3.x versions of \cweave{} are minor and can be ignored in + our context. See the revision log of the complete interface specs + document for more information.} +in an ordered fashion. First we look at those +tags which are part of the `protected interface,' i.e., they are +visible to a \cweb{} user, but he must not use them. Then we consider +the private tags, some are used in the documentation part, others +are needed to typeset program code, and there are a few tags for +typesetting special characters in strings. + + +\sect Some tags output by \cweave{} are part of the protected +interface even though they are not prefixed by |cweb|. We'll present +them in the order they'll arrive in the document instance. + +The following table specifies in the second column if this tag takes +arguments. If the entry is non-empty, it's either a number listing +just how many arguments are expected; then usual argument passing is +used. Or it displays the context required. + % +\begin{cseqtab} +|\ATL| & \verb*|#1 #2 | & \cweb{} operator: |@l| + && (how to output non-ASCII chars in \ctangle{}) + && \arg1 hex code of mapped character + && \arg2 string output by \ctangle{} +|\M| & 1 & \cweb{} structure tag: start of a chunk + && \arg1 chunk number +|\N| & |#1#2#3.| & \cweb{} structure tag: start of a section + && \arg1 section rank, $0 \le |#1| \le 10$ + && \arg2 chunk number + && \arg3 section title +|\PB| & 1 & restricted program mode material + && \arg1 program code +|\Y| & & between major pieces of a program part +|\B| & & start program material +|\fi| & & \cweb{} structure tag: end of a chunk +|\9| & 1 & index entry, user defined layout + && \arg1 text of index entry +\end{cseqtab} + +|\ATL| does only appear in front of the very first chunk. The +chunk number is an explicit \TeX{} number which might be followed +by a `changed flag' (see chunk~\ref{sec:text-tag.priv}). Note the +usage of |\fi|, ie, each chunk must open an according |\if|. |\PB| +might appear within its own argument (created by restricted program +mode material in a refinement name within restricted program mode). + +|\9| deserves a further explanation: It is expected, though not +defined, that it expands to an empty token list. The parameter will +be a sort key, actually; the real key to be typeset will appear +afterwards. So a \cweb{} user might index \TeX{} as +`|@:TeX}{\TeX@>|', the index will feature it in the ``T''~chunk, +not in the ``|\|''~chunk. + +With the exception of |\PB| the tags above will not appear in math mode. + + +\sect The following tags might appear to be public ones, but they +are, in fact, never used. That's because they will be placed after +the |\fi| which terminates the last chunk. Our problem is that we +have to produce the information which were produced by these tags in +the Plain macros. Namely the optional lists poses us the problems. We +must read behind the document end to gather the information which +lists to produce. + +%% This label is used in cweb.doc +\label{spec:sec:docend} + +We can depend on the following rules. + % +\begin{enumerate} + +\item If |\end| is found, there will be no other of these tags. + +\item |\ch| comes first, but is optional. + +\item |\inx\fin| come next, they are mandatory. + +\item |\con| is optional again; if it is found it's the very last +token to look at in this document. + +\end{enumerate} + % +\begin{fixme} + +Perhaps |\inx| should be defined as |\end{document}|, and both |\ch| +and |\inx| should be added to the protected interface. Then the user +may simply omit |\end{document}|. But we should take care of a +possible occurence anyhow. + +\end{fixme} + +Here are the tags mentioned in this chunk, and their respective +meaning in the Plain version: + % +\begin{cseqtab} +|\ch| & |#1.| & Note which chunks are changed + && \arg1 list of chunk numbers, like in |\U| and such. +|\inx| & & Create index +|\fin| & & Create the table of refinement names +|\con| & & Create the table of contents +|\end| & & output if index and all lists were suppressed +\end{cseqtab} + + + +\sect \label{sec:text-tag.priv} + Some tags appear only in special circumstances and may therefore be +considered as private tags. The largest part of them concern the tagging of +program code, we'll have a look at them later. First we present the +tags used in other areas. + +Lists of chunk numbers occurs on several places: At the chunk +start (where the list has actually only one element), within +refinement names, in the identifier index, and for cross reference +purposes. Cross references can be made at the end of a chunk, and +in the refinement name list at the very end. Everywhere where a +chunk number can occur it can be followed by a tag which shows that +this chunk was changed by the changefile. + % +\begin{cseqtab} +|\*| & & tag after chunk number: this chunk is changed. +\end{cseqtab} + +Within the identifier index we have also special tags. The identifiers +are tagged like in the program mode, ie, with |\\| and |\&|. Remember +that |\9|, listed above, appears also in the index. + % +\begin{cseqtab} +|\I| & \verb*|#1, | & start of an index entry + && \arg1 index entry +|\[| & |#1]| & underlined chunk number in index + && \arg1 chunk number +|\.| & 1 & |@.| index entry + && \arg1 index entry +\end{cseqtab} + +In the list of refinement names the entries are marked similar to +the index entries. But note that |\I| has no arg here. + % +\begin{cseqtab} +|\I| & & start of a new refinement name +\end{cseqtab} + + +\sect \label{sec:prog-tag} + OK, now we can have a look at the large amount of tags used +for tagging program code. First, we have those which represent +directly C or \C++ tokens. + % +\begin{cseqtab} +|\?| & & C operator: conditional expression +|\AND| & & C operator: logical and +|\CM| & & C operator: binary complement +|\DC| & & \C++ operator: scope resolution +|\E| & & C operator: equivalence + && and equivalence sign after refinement name on it's definition +|\G| & & C operator: greater or equal +|\GG| & & C operator: shift right +|\I| & & C operator: not equal +|\K| & & C operator: assignment +|\LL| & & C operator: shift left +|\MG| & & C operator: pointer to struct component +|\MGA| & & \C++ operator: pointer to pointer to member +|\MM| & & C operator: decrement +|\MOD| & & C operator: modulo (actually, remainder) +|\NULL| & & `quoted' identifier +|\OR| & & C operator: binary or +|\PA| & & \C++ operator: pointer to member +|\PP| & & C operator: increment +|\R| & & C operator: logical negation +|\this| & & `quoted' identifier +|\TeX| & & `customized' identifier +|\V| & & C operator: logical or +|\W| & & C operator: logical and +|\XOR| & & C operator: binary exclusive or +|\Z| & & C operator: less or equal +\end{cseqtab} + +Other tokens have variable parts, passed as arguments. + % +\begin{cseqtab} +|\.| & 1 & C string + && \arg1 string +|\)| & & discretionary break between string parts +|\&| & 1 & reserved identifier + && \arg1 identifier +|\\| & 1 & ``normal'' identifier with more than one chars + && \arg1 identifier +\verb"\|" & 1 & ``normal'' identifier with one char + && \arg1 character +|\C| & 1 & C comment + && \arg1 comment text +|\MRL| & 1 & C operator: combined binary operators + && \arg1 operators, |\K| must print as `$=$' +|\SHC| & 1 & \C++ comment + && \arg1 comment text +|\T| & 1 & numeric constants + && \arg1 constant +|\X| & |#1: #2\X| & refinement name + && \arg1 chunk number + && \arg2 refinement name +\end{cseqtab} + % +The refinement name (second argument of |\X|) may be a file name +tagged by `|\.|'. Then the name must be set like a C~string. + +\cweb{} itself introduces its own operators: + % +\begin{cseqtab} +|\ATH| & & |@h| (place the preprocessor definitions here) +|\D| & & |@d| (define macro) +|\F| & & |@f| (format identifier like another one) +|\J| & & |@&| (join) +|\vb| & 1 & |@=| (pass arg verbatim by \ctangle{}) + && \arg1 string +\end{cseqtab} + +And we have some tags used for cross referencing. Chunk +cross-referencing (actually, for program parts) is started with some +tag, then follows a list of chunk numbers. + % +\begingroup + \hfuzz=22pt % 21.76602pt too wide (with a4-9.sty) +\begin{cseqtab} +|\A| & & one add-on definition (``See also chunk'') +|\As| & & more than one add-on definition (``See also chunks'') +|\ET| & & separator between the last two numbers if there are only two +|\ETs| & & separator between the last two numbers if there are more than two +|\Q| & & the chunk where it is cited (``This code is cited in chunk'') +|\Qs| & & more than one chunk where it is cited (``This code is cited in chunk'') +|\U| & & used in one chunk (``This code is used in chunk'') +|\Us| & & used in more than one chunk (``This code is used in chunks'') +\end{cseqtab} +\endgroup + +Program code must be indented according to its structure. Each +``larger'' statement is typeset as one paragraph, usually only one +line long. The basic indentation of the next statement may be +incremented and decremented in given units. If a statement has to be +broken in more than one line nevertheless, a hanging indentation is +added to the basic indentation for the subsequent lines. One can mark +an `optional' statement start, ie, some kind of optional paragraph +start. If a line break must be inserted it should be inserted here and +no hanging indentation should be added to the basic indentation. + % +\begin{cseqtab} +|\1| & & future stmts indented one more unit +|\2| & & future stmts indented one less unit +|\3| & 1 & optional line break within a statement + && \arg1 digit, penalty for line break +|\4| & & backspace one indentation unit +|\5| & & optional line break or small space (between run-in statements) +|\6| & & line break +|\7| & & line break and additional vertical space +|\8| & & line is a preprocessor directive (must be issued at start of line) +\end{cseqtab} + + +\sect Some args are designated as strings above. Within these args the +following tags are used to represent special characters: + % +\begin{cseqtab} +\verb*|\ | & & space +|\&| & & ampersand +|\\| & & backslash +|\^| & & hat +|\_| & & underscore +|\{| & & left brace +|\}| & & right brace +|\~| & & tilde +\end{cseqtab} + + +\sect Last, we have tags which are used in their ``standard'' +\LaTeX{} meaning. We don't have much work with them: + % +\begin{cseqtab} +|\#| & & hash mark in program or in string +|\$| & & dollar +|\%| & & percent +|\{| & & block start (is set in math mode) +|\}| & & block end (is set in math mode) +|\,| & & thin space (like the \LaTeX{} default) +|\langle| & & left delimiter in include directives, or in templates +|\le| & & C operator: less or equal +|\ldots| & & Function protoypes: Variable number of arguments +|\rangle| & & right delimiter in include directives, or in templates +\end{cseqtab} + + + + +\endSubDocument + + +%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% + +\vskip \PltxPreSectSkip + +\begin{rcslog} +$DocLog: cweave-spec.tex,v $ +\Revision 1.8 1995/09/12 22:56:02 schrod +|cweave -x| doesn't output |\vfill| in front of |\end| any more. +Change noticed in \cweb{} 3.4. + +\Revision 1.7 1995/08/29 01:40:24 schrod +Checked the interface description for \cweb{}~3.4. + +\Revision 1.6 1995/08/25 19:09:57 schrod +Cleaned up terminology: Chunks are not named sections any more. +Explicate possible range of section ranks. + +\Revision 1.5 1995/08/08 00:14:26 schrod +Updated to \LaTeXe{}, the |cweb| style is now a document class. Used +my standard templates for that, no changes in functionality. + +\Revision 1.4 1993/06/17 15:09:28 schrod +\cweb{} 3.0 was released officially on June 16, 1993. Mentioned in +the documentation that this version is needed for the |cweb| style. + +\Revision 1.3 1993/06/14 17:50:14 schrod +Clarify the occurence pattern of the lists after the last section: +which are optional, the expected order, etc. + +Changed `cseq' to `tag'. On this level of abstraction (that's the +specification!), we're still not considering \TeX{}nical details. + +\Revision 1.2 1993/06/14 15:40:02 schrod +Adapted to April 93 changes to \cweave{}: |\N| has three parameters +(added group level), |\M| param is no data-tag any more, |\Z| and +|\MRL| are new. + +Added further explanation of |\9| and |\X|. + +Replaced `program part name' by `refinement name'. + +\Revision 1.1 1993/04/09 15:00:37 schrod +Initial revision + +\end{rcslog} + + + +\end{document} + +% LocalWords: diff --git a/Master/texmf-dist/doc/latex/cweb-latex/src/cweb-conf.tex b/Master/texmf-dist/doc/latex/cweb-latex/src/cweb-conf.tex new file mode 100644 index 00000000000..86dd2ae04e4 --- /dev/null +++ b/Master/texmf-dist/doc/latex/cweb-latex/src/cweb-conf.tex @@ -0,0 +1,103 @@ +% $DocId: cweb-conf.tex,v 2.1 1995/08/29 15:19:45 schrod Exp $ +%------------------------------------------------------------ + +% +% internal interface of cweb class +% +% [LaTeX] +% (history at end) + + +% This file is either a subdocument of the style doc or a document on +% its own. In the former case it's a chapter, in the latter it's a +% ``normal'' LaTeX progltx document. +% If it's a subdocument, this file will be included after +% \begin{document}. We can detect this: \document redefines +% \documentclass to be \@twoclasseserror. Then we also have to define +% how this document is ended: Either by \endinput or by an additional +% revision log. +% Of course, this test works only if LaTeX 2e is used for processing. + +\expandafter\ifx \csname @twoclasseserror\endcsname \documentclass + + \let\endSubDocument=\endinput + + \chap The internal interface. + +\else + + \let\endSubDocument=\relax + + \documentclass{progltx} + + \usepackage{cweb-doc} % additional document-specific markup + \usepackage{a4-9} % Tschichold's A4 layout + + \nofiles % no crossreferences used + + \begin{document} + + \title{The \texttt{cweb} Class\\ + {\large (Internal Interface)}% + } + \author{Joachim Schrod% + \thanks{Email: \texttt{schrod@iti.informatik.th-darmstadt.de}}% + } + + \RCS $DocDate: 1995/08/29 15:19:45 $ + \date{% + \RCSDocDate\\[3pt] % LaTeX Error: Paragraph terminated too early + (Revision \RCSStyleRevision{} of \texttt{cweb.cls})% + } + + \maketitle + + \sect + +\fi + + + + +This +% +\ifx \endSubDocument\relax + document +\else + section +\fi +% +shall describe the internal interface for \TeX{} programmers who want +to write packages or subclasses to enhance the \cls{cweb} class. In +particular, it shall describe all control sequences a programmer may +redefine to adapt this class to special needs. (There are an awful lot +of them, all cseqs that start with `|Cweb|'.) + +Due to time constraints this description has not been written yet. It +depends on the demand when it will be written. I.e., if \emph{you} are +a package/class writer and if you want to enhance the \cls{cweb} +class, send me email. Please mention the kind of enhancement you want +to do. I will then add information to this description as needed. + + + + +\endSubDocument + + +%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% + +\vskip \PltxPreSectSkip + +\begin{rcslog} +$DocLog: cweb-conf.tex,v $ +\Revision 2.1 1995/08/29 15:19:45 schrod +Added (hint to non-existence of) description of the internal interface. + +\end{rcslog} + + + +\end{document} + +% LocalWords: diff --git a/Master/texmf-dist/doc/latex/cweb-latex/src/cweb-doc.sty b/Master/texmf-dist/doc/latex/cweb-latex/src/cweb-doc.sty new file mode 100644 index 00000000000..2859584f6cb --- /dev/null +++ b/Master/texmf-dist/doc/latex/cweb-latex/src/cweb-doc.sty @@ -0,0 +1,133 @@ +% $Id: cweb-doc.sty,v 1.3 1995/08/25 19:07:54 schrod Exp $ +%---------------------------------------------------------------------- + +% +% tags for the documentation of cweb class +% + + +\ProvidesPackage{cweb-doc} + + +% +% Used packages +% + +\begingroup + \let\begingroup=\endinput + \input{cweb.cls} % get style revision + \global\let\cweb@loaded=\cweb@loaded +\endgroup +\providecommand*\cweb@loaded{$StyleRevision$} % bootstrap + + + +% +% Configuration of RCS fields +% + +% keywords in style are prefixed by `Style', in documentation by `Doc'. +\let\RcsHandleStyleDate=\RcsHandleDate % transform StyleDate value +\let\RcsHandleDocDate=\RcsHandleDate % transform DocDate field + +% the revision of cweb.cls is needed by the documentation +\def\RcsEmptyValue{$\langle$unknown$\rangle$} +\expandafter\RCS \cweb@loaded % creates \RCSStyleRevision +\def\RcsEmptyValue{} + +% author name +\rcsAuthor{schrod}{Joachim Schrod} + + + +% +% new environments +% + +\font\manfnt=manfnt % with the dangerous bend sign +\chardef\bend="7F % I like the non-S curve ;-) +\newenvironment{fixme}{% + \trivlist + \hangindent 2em + \hangafter -2 + \item[\hbox to 0pt{\hss \manfnt \bend}]% + \ignorespaces + }{% + \endtrivlist + } + + +\newenvironment{cseqtab}{% + \quote + \obeycr + \tabular{lrl}% + \@gobblecr % eat following newline + }{% + \endtabular + \endquote + } + + +\newenvironment{options}{% + \list{}{% + \labelwidth\z@ + \itemindent -.5\leftmargin + \labelsep 1em + \def\makelabel##1{% + \hspace\labelsep + \normalfont\ttfamily ##1% + }% + }% + }{% + \endlist + } + + + +% +% new tags +% + +% logos +\def\cweb{\texttt{CWEB}} +\def\ctangle{\texttt{CTANGLE}} +\def\cweave{\texttt{CWEAVE}} + +% other +\renewcommand*\arg[1]{\textsl{Arg.~#1}:} % I don't use the arg operator + +\def\<#1>{\textit{#1}} % syntactic category + +\newcommand\cls[1]{\textsf{#1}} % LaTeX classes +\newcommand\pkg[1]{\textsf{#1}} % LaTeX packages + + +% +% change of LaTeX parameters, for layout etc. +% + +% underfull hboxes are not such a problem in program documentation + +\hbadness=4000 + + +% figure + +\def\textfraction{0.1} % must be at least 10% text on page +\def\topfraction{0.9} % rest may be top insertion +\def\floatpagefraction{0.8} % float pages only if the float is larger than 80% + + + + +\endinput + + + +%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% +Local Variables: +mode: LaTeX +TeX-brace-indent-level: 4 +indent-tabs-mode: t +TeX-auto-untabify: nil +End: diff --git a/Master/texmf-dist/doc/latex/cweb-latex/src/cweb-fsa.fig b/Master/texmf-dist/doc/latex/cweb-latex/src/cweb-fsa.fig new file mode 100644 index 00000000000..729c992d272 --- /dev/null +++ b/Master/texmf-dist/doc/latex/cweb-latex/src/cweb-fsa.fig @@ -0,0 +1,78 @@ +#FIG 2.1 +80 2 +6 35 82 134 122 +1 1 0 1 -1 0 0 0 0.00000 1 0.000 83 100 47 18 83 100 130 119 +4 1 0 10 0 -1 0 0.00000 0 7 63 83 104 Documentation +-6 +6 35 152 134 192 +6 62 158 107 180 +4 1 0 10 0 -1 0 0.00000 0 7 43 83 167 Restricted +4 1 0 10 0 -1 0 0.00000 0 9 36 83 180 Program +-6 +1 1 0 1 -1 0 0 0 0.00000 1 0.000 83 170 47 18 83 170 130 189 +-6 +6 116 17 215 57 +6 140 23 188 48 +4 1 0 10 0 -1 0 0.00000 0 9 48 164 32 (CR) Cross +4 1 0 10 0 -1 0 0.00000 0 7 40 164 48 Reference +-6 +1 1 0 1 -1 0 0 0 0.00000 1 0.000 164 35 47 18 164 35 211 54 +-6 +6 203 152 302 192 +1 1 0 1 -1 0 0 0 0.00000 1 0.000 251 170 47 18 251 170 298 189 +4 1 0 10 0 -1 0 0.00000 2 7 21 251 174 \TeX +-6 +6 203 82 302 122 +1 1 0 1 -1 0 0 0 0.00000 1 0.000 251 100 47 18 251 100 298 119 +4 1 0 10 0 -1 0 0.00000 0 9 36 251 104 Program +-6 +6 257 122 299 143 +6 257 131 287 143 +4 0 5 10 0 -1 0 0.00000 2 10 30 257 143 \\SHC +-6 +4 0 5 10 0 -1 0 0.00000 2 10 42 257 132 \\C \\X +-6 +6 221 47 266 71 +4 0 5 10 0 -1 0 0.00000 2 10 42 223 57 \\A \\Q +4 0 5 10 0 -1 0 0.00000 2 8 18 235 69 \\U +-6 +2 1 0 1 -1 0 0 0 0.000 -1 1 0 + 0 0 1.000 4.000 8.000 + 254 119 254 152 9999 9999 +2 1 0 1 -1 0 0 0 0.000 -1 1 0 + 0 0 1.000 4.000 8.000 + 245 83 208 46 9999 9999 +2 1 0 1 -1 0 0 0 0.000 -1 1 0 + 0 0 1.000 4.000 8.000 + 119 41 89 81 9999 9999 +2 1 0 1 -1 0 0 0 0.000 -1 1 0 + 0 0 1.000 4.000 8.000 + 101 83 125 51 9999 9999 +2 1 0 1 -1 0 0 0 0.000 -1 1 0 + 0 0 1.000 4.000 8.000 + 131 174 203 174 9999 9999 +2 1 0 1 -1 0 0 0 0.000 -1 0 1 + 0 0 1.000 4.000 8.000 + 131 164 203 164 9999 9999 +2 1 0 1 -1 0 0 0 0.000 -1 0 1 + 0 0 1.000 4.000 8.000 + 89 119 89 152 9999 9999 +2 1 0 1 -1 0 0 0 0.000 -1 1 0 + 0 0 1.000 4.000 8.000 + 131 95 203 95 9999 9999 +2 1 0 1 -1 0 0 0 0.000 -1 0 1 + 0 0 1.000 4.000 8.000 + 131 105 203 105 9999 9999 +2 1 0 1 -1 0 0 0 0.000 -1 1 0 + 0 0 1.000 4.000 8.000 + 74 119 74 152 9999 9999 +2 1 0 1 -1 0 0 0 0.000 -1 0 1 + 0 0 1.000 4.000 8.000 + 239 119 239 152 9999 9999 +4 2 5 10 0 -1 0 0.00000 2 10 24 72 136 \\PB +4 1 5 10 0 -1 0 0.00000 2 8 24 167 160 \\PB +4 1 5 10 0 -1 0 0.00000 2 8 18 166 92 \\B +4 1 5 10 0 -1 0 0.00000 2 8 42 166 117 \\M \\N +4 2 5 10 0 -1 0 0.00000 2 8 42 101 60 \\M \\N +4 0 5 10 0 -1 0 0.00000 2 8 24 113 76 \\ch +4 1 5 10 0 -1 0 0.00000 2 8 18 167 188 \\X diff --git a/Master/texmf-dist/doc/latex/cweb-latex/src/cweb-fsa.latex b/Master/texmf-dist/doc/latex/cweb-latex/src/cweb-fsa.latex new file mode 100644 index 00000000000..8f624ed30af --- /dev/null +++ b/Master/texmf-dist/doc/latex/cweb-latex/src/cweb-fsa.latex @@ -0,0 +1,62 @@ +\setlength{\unitlength}{0.012500in}% +% +\ifx \SetFigFont\undefined + \begingroup\makeatletter + % extract first six characters in \fmtname + \def\x#1#2#3#4#5#6#7\relax{\def\x{#1#2#3#4#5#6}}% + \expandafter\x\fmtname xxxxxx\relax \def\y{splain}% + \ifx\x\y % LaTeX or SliTeX? + \gdef\SetFigFont#1#2#3{% + \ifnum #1<17\tiny\else \ifnum #1<20\small\else + \ifnum #1<24\normalsize\else \ifnum #1<29\large\else + \ifnum #1<34\Large\else \ifnum #1<41\LARGE\else + \huge\fi\fi\fi\fi\fi\fi + \csname #3\endcsname}% + \else + \gdef\SetFigFont#1#2#3{\begingroup + \count@#1\relax \ifnum 25<\count@\count@25\fi + \def\x{\endgroup\@setsize\SetFigFont{#2pt}}% + \expandafter\x + \csname \romannumeral\the\count@ pt\expandafter\endcsname + \csname @\romannumeral\the\count@ pt\endcsname + \csname #3\endcsname}% + \fi + \endgroup +\fi +\begin{picture}(262,171)(37,651) +\put( 84,735){\makebox(0,0)[b]{\smash{\SetFigFont{10}{12.0}{}Documentation}}} +\put( 84,672){\makebox(0,0)[b]{\smash{\SetFigFont{10}{12.0}{}Restricted}}} +\put( 84,659){\makebox(0,0)[b]{\smash{\SetFigFont{10}{12.0}{}Program}}} +\thinlines +\put( 84,669){\oval(94,36)} +\put(165,807){\makebox(0,0)[b]{\smash{\SetFigFont{10}{12.0}{}(CR) Cross}}} +\put(165,791){\makebox(0,0)[b]{\smash{\SetFigFont{10}{12.0}{}Reference}}} +\put(165,804){\oval(94,36)} +\put(252,669){\oval(94,36)} +\put(252,665){\makebox(0,0)[b]{\smash{\SetFigFont{10}{12.0}{}\TeX}}} +\put(252,739){\oval(94,36)} +\put(252,735){\makebox(0,0)[b]{\smash{\SetFigFont{10}{12.0}{}Program}}} +\put(258,696){\makebox(0,0)[lb]{\smash{\SetFigFont{10}{12.0}{tt}\\SHC}}} +\put(258,707){\makebox(0,0)[lb]{\smash{\SetFigFont{10}{12.0}{tt}\\C \\X}}} +\put(224,782){\makebox(0,0)[lb]{\smash{\SetFigFont{10}{12.0}{tt}\\A \\Q}}} +\put(236,770){\makebox(0,0)[lb]{\smash{\SetFigFont{10}{12.0}{tt}\\U}}} +\put(255,720){\vector( 0,-1){ 33}} +\put( 84,739){\oval(94,36)} +\put(246,756){\vector(-1, 1){ 37}} +\put(168,651){\makebox(0,0)[b]{\smash{\SetFigFont{10}{12.0}{tt}\\X}}} +\put(120,798){\vector(-3,-4){ 30}} +\put(102,756){\vector( 3, 4){ 24}} +\put(132,665){\vector( 1, 0){ 72}} +\put(204,675){\vector(-1, 0){ 72}} +\put( 90,687){\vector( 0, 1){ 33}} +\put(132,744){\vector( 1, 0){ 72}} +\put(204,734){\vector(-1, 0){ 72}} +\put( 75,720){\vector( 0,-1){ 33}} +\put(240,687){\vector( 0, 1){ 33}} +\put( 73,703){\makebox(0,0)[rb]{\smash{\SetFigFont{10}{12.0}{tt}\\PB}}} +\put(168,679){\makebox(0,0)[b]{\smash{\SetFigFont{10}{12.0}{tt}\\PB}}} +\put(167,747){\makebox(0,0)[b]{\smash{\SetFigFont{10}{12.0}{tt}\\B}}} +\put(167,722){\makebox(0,0)[b]{\smash{\SetFigFont{10}{12.0}{tt}\\M \\N}}} +\put(102,779){\makebox(0,0)[rb]{\smash{\SetFigFont{10}{12.0}{tt}\\M \\N}}} +\put(114,763){\makebox(0,0)[lb]{\smash{\SetFigFont{10}{12.0}{tt}\\ch}}} +\end{picture} diff --git a/Master/texmf-dist/doc/latex/cweb-latex/src/cweb-fsa.ltx b/Master/texmf-dist/doc/latex/cweb-latex/src/cweb-fsa.ltx new file mode 100644 index 00000000000..aa5cdce2f26 --- /dev/null +++ b/Master/texmf-dist/doc/latex/cweb-latex/src/cweb-fsa.ltx @@ -0,0 +1,29 @@ +% $Id: cweb-fsa.ltx,v 1.2 1995/08/08 00:14:27 schrod Exp $ +%------------------------------------------------------------ + +% +% test of xfig created pictures +% +% [LaTeX] + + + + +\documentstyle{article} + + +\begin{document} + +\noindent Test of figure \texttt{cweb-fsa.latex}: Finite State +Automaton for different CWEB modi. + +\bigskip + +\begin{center} + \chardef\\=`\\ + \input{cweb-fsa.latex} +\end{center} + + + +\end{document} diff --git a/Master/texmf-dist/doc/latex/cweb-latex/src/cweb-user.tex b/Master/texmf-dist/doc/latex/cweb-latex/src/cweb-user.tex new file mode 100644 index 00000000000..2f21f53150c --- /dev/null +++ b/Master/texmf-dist/doc/latex/cweb-latex/src/cweb-user.tex @@ -0,0 +1,779 @@ +% $DocId: cweb-user.tex,v 2.10 1995/11/20 22:34:16 schrod Exp $ +%------------------------------------------------------------ + +% +% user manual/requirement definition for cweb class +% +% [LaTeX] +% (history at end) + + +% This file is either a subdocument of the style doc or a document on +% its own. In the former case it's a chapter, in the latter it's a +% ``normal'' LaTeX progltx document. +% If it's a subdocument, this file will be included after +% \begin{document}. We can detect this: \document redefines +% \documentclass to be \@twoclasseserror. Then we also have to define +% how this document is ended: Either by \endinput or by an additional +% revision log. +% Of course, this test works only if LaTeX 2e is used for processing. + +\expandafter\ifx \csname @twoclasseserror\endcsname \documentclass + + \let\endSubDocument=\endinput + + \chap What's this class for?. + +\else + + \let\endSubDocument=\relax + + \documentclass{progltx} + + \usepackage{cweb-doc} % additional document-specific markup + \usepackage{a4-9} % Tschichold's A4 layout + + + \begin{document} + + \title{The \texttt{cweb} Class} + \author{Joachim Schrod% + \thanks{Email: \texttt{jschrod@acm.org}}% + } + + \RCS $DocDate: 1995/11/20 22:34:16 $ + \date{% + \RCSDocDate\\[3pt] % LaTeX Error: Paragraph terminated too early + (Revision \RCSStyleRevision{} of \texttt{cweb.cls})% + } + + \maketitle + + \sect + +\fi + +\cweb{} is a Literate Programming tool that enables you to combine +documentation and actual C or \C++ code in one document. The \cls{cweb} +class allows you to use \LaTeX{} for the documentation of your \cweb{} +programs. Now a \cweb{} document is like a usual \LaTeX{} +document---as you have probably written dozens before. With all the +features of \LaTeX{}, and with all the quirks of \LaTeX{}. + +Originally, \cweb{} was only usable with Plain \TeX{}, as \cweave{} +outputs \TeX{} code that looks as if it were Plain \TeX{}. Well, it +isn't---it's just not the normal \LaTeX{} markup. No change in +\cweave{} is necessary to use \LaTeX{}. + +For the sake of those \cweb{} programmers who have used the original +markup interface (i.e., the Plain \TeX{} macros) before, I'll +sometimes contrast the approach presented here with the original one. +Then I'll call the original interface ``Plain \cweb{}''. The new +interface is called ``\LaTeX{} \cweb{}''. + +If you never used \cweb{} before, you won't find enough information in +this manual. Please read the user manual of \cweb{} first, and then +come back again. I assume a working knowledge, and will only present +the new stuff introduced by this class. + + +\sect This bundle does \emph{not} support inclusion of \cweb{} code in +other (general) \LaTeX{} documents. This is not the task of the +\cls{cweb} class -- its task is the usage of \LaTeX{} for creation of +\cweb{} documents. + +Note also that you will not be able to \LaTeX{} your \cweb{} documents +(i.e., the program files) directly. You must still process them first +by \cweave{}, then you can use \LaTeX{} to process the resulting +\TeX{} file. + + +\sect You have to use \cweb{} version~3 and you need at least \LaTeX{} +version \mbox{$\langle$1994/12/01$\rangle$} (or any later one), this +class does not work with older versions. + + +\sect A \cweb{} document has two concurrent structures, a +\textsl{document structure} and a \textsl{program structure}. + +The basic building block of the document structure, the bricks that +build up our document, are called \textsl{chunks}. We use chunks to +present one point of our work to our reader, the presentation has both +informal (explanation) and formal parts (code). This is the essence of +Literate Programming, that we tell each point twice, once on an +abstract and once on a technical level. + +A chunk consists of three parts: (1)~the \textsl{documentation part}, +(2)~the \textsl{definition part}, and (3)~the \textsl{program part}. +Each of these parts can be empty. The documentation part is mostly +text with \LaTeX{} tags. In this text material from \textsl{restricted +program mode} can appear. The definition part consists of a series of +either \textsl{macro} or \textsl{format definitions}. The program part +is one piece of a refinement, identified by a name (see below). + +A chunk starts with `\verb*|@ |', and gets assigned a unique +identification, a number; we'll use that number to refer to that +chunk. In particular, the number is of interest when we want to refer +to the code part of the chunk, otherwise it's often regarded as +superfluous. + +\sect And we have higher-level structure elements as well. They combine a +sequence of chunks and a sequence of structure elements on a lower +level, both sequences may be empty. As usual in \LaTeX{}, we call +these elements \textsl{section divisions} or short \textsl{sections}. +We do not give them names +like `part,' `chapter,' `subsection,' etc., to denote their level; +most often we'll use level-numbers (sometimes called \textit{rank}) +instead. + +A section division starts with `|@*|' and is followed by a character, the rank +indicator. The highest rank is~`|@**|', you can think of it as an +equivalence to \LaTeX{}'s |\part|. (Often we don't use that rank in +smaller programs, as we don't use |\part| in smaller articles.) A +section on the next rank is started with~`\verb*|@* |' (synonym to +`|@*0|'), roughly equivalent to chapters. Then we'll use numbers to +mark the next ranks, from |@*1| to~|@*9| (sections to +sub$^{9}$sections). That's outrageous---I'll hope you'll never use +11~hierarchy levels in your documents, they would be unreadable. + +A section has a title, that title comes after the section start tag +and ends with a full stop. If you need periods in your title, enclose +them in braces. (That behavior can be changed if there's enough +interest. I can also make it possible to use braces to delimit the +section title, as in |\section|, etc. Send mail with your opinion.) + +As a matter of convenience, a section implicitly starts a chunk; +i.e., one does not need to use `\verb*|@ |' after the section title. +Of course, it doesn't hurt to use it either. + +\medskip + +The explanation might have been a bit too abstract, so let me make one +thing plain clear: You should \emph{never ever} use |\part|, +|\chapter|, |\section|, or any other \LaTeX{} sectioning command in +your \cweb{} documents. Use the sectioning commands of \cweb{}, i.e., +`|@*|\textit{rank}'. + + + +\sect The description above does not mean that the output is fixed to +the style you'll know from the \cls{book} or \cls{report} document +classes. Quite in contrary, the default layout is oriented along the +\cls{article} class. + +There is a very easy possibility to change that default layout: With +the option |baseclass| (see chunk~\ref{chunk:baseclass}) you can +specify the document class \cls{cweb} inherits from. If that document +class has chapters (like \cls{report} or \cls{book}), you'll get the +layout as you're used from these classes. + + +\sect In Plain \cweb{}, the term \textit{section} was used for +`chunk.' We abstain from that usage, as sections are used in \LaTeX{} +(and in almost all other documentation systems, for that matter) to +designate higher-level structure elements. We prefer to follow the +traditional usage of terms here, even if that means that we'll have to +confront Plain \cweb{} users with a slight shift in semantics. + +And you'll find the term \textit{starred section} in Plain \cweb{} and +might think these are our sections. Well, almost. There's a subtle, +but important difference between these starred sections and our +terminology: A starred section is just a chunk with a special +attribute (a title). That title will appear in a table of contents, +but that does not mean that the starred section is an element to +denote a structural \emph{hierarchy}. + +You won't find a section number like `1.5.2' in Plain \cweb{}, do you? +--- In \LaTeX{} \cweb{}, you've got the choice. You have the new +support of the hierarchical structure, or you may stay with Plain +\cweb{}'s (flat) structure. + + +\sect We still have one open point in our explanation of the +document's structure. As mentioned at the start, we've got two +concurrent structures, let's have a look at the program structure +finally. + +A \cweb{} program consists of a tree of \textsl{refinements}. A +refinement is a sequence of program parts with the same name, ordered +in appearance in the document. The root of the tree is the refinement +with the special name~|@c|. + +The complete program text is defined by the depth-first traversal +(DFS, i.e.\ infix-order) of the tree, the tool \ctangle{} extracts the +program text from a \cweb{} document. + + + + +\chap Markup for your document structure. + +Like every \LaTeX{} document, a \cweb{} document starts with a +document class specification; then comes the preamble, terminated by +the document start. This way we brought some structure in the \cweb{} +limbo, but this shouldn't be a problem for you. You must tag the end +of the document, as you do in other \LaTeX{} files. But here you must +take a bit more care: +% +\begin{quote} + \itshape + + Assert that the\/ |\end{document}| is in the documentation part of + a\/ chunk, neither in the definition nor in the program part. + +\end{quote} +% +If it would be in the program part, \cweave{} would readily process it +and \LaTeX{} would never see this tag. Nevertheless, if you make this +mistake, you will detect it early: \ctangle{} will copy this tag to +the C file as well, and your program will be erroneous\,\dots + +And there's another difference you'll have to take care of: Usually +you can put arbitrary text after the end of your document, \LaTeX{} +will not see it. Please note that both \ctangle{} and \cweave{} will +still see it---don't put a program part or even complete chunks there. +For technical reasons, there must not be any |\fi| token after the +document end, too. + + +\sect The resulting structure of your \cweb{} document is exemplified +in figure~\ref{fig:structure}. Note that |\end{document}| is placed +in a new, empty, chunk. This isn't necessary, you could write more +documentation there---but it's considered good style to use a sole +chunk for document finish. + +%% my little joke +\begin{figure} + +\setbox0=\vbox{\parindent=0pt \hsize=.85\textwidth +\begin{minipage}{\hsize} +\begin{verbatim} +\documentclass{cweb} + +\begin{document} + +\title{My Program, Doomed for the ACM Software Systems Award} +\author{Joe L. User} + +\maketitle + +\tableofcontents % if you want + + +@* A PSPACE solution for the Traveling Salesman. + +< insert your program here > + + + +@ +\end{document} +\end{verbatim} +\end{minipage} +} + +\begin{center} +\fboxsep=1em +\fbox{\box0} +\end{center} + +\caption{Exemplified \cweb{} document structure} +\label{fig:structure} +\end{figure} + + +\sect If you don't call \cweave{} with the |-x| option, an identifier +index and a list of all refinements is created at the end of the +document. You can specify an introductionary text for the index with +the tag |\cwebIndexIntro|, the introduction is the argument of this +tag. + + + + +\chap Configuration. + +You may configure the \cls{cweb} class in several ways, with class +options, redefinitions in the preamble, supplying additional +packages, and by subclassing it. + +At start of the processing, the file |cweb.cfg| is read if it exists. +That file may contain configuration code that shall apply to all your +\cweb{} documents. You can set the paper size there, etc. Please don't +set the default structure there (see +chunk~\ref{chunk:structure-option}), others would get different results +when they process your document. + + +\sect Options are specified with the new ``keyword$=$value''-scheme, +first introduced with the \pkg{graphicx} package. I.e., the options have a +set of possible values, you may choose one. For instance, the option +|structure| has the possible values |flat| and |hierarchic|. You tell +that you want a flat document structure by +% +\begin{quote} + |\documentclass[structure=flat]{cweb}| +\end{quote} +% +As usual, options are separated by commas. From the set of possible +option values, two ones are special: The \textsl{predefined value} and +the \textsl{default value}. The predefined value is the one that's +used when you don't specify the option at all. The default value is +used of you don't specify the value in the option; e.g., +% +\begin{quote} + |\documentclass[suppress]{cweb}| +\end{quote} +% +selects the default value of the option |suppress|. (Not that there is +one currently. :--) + +There is always a predefined value, a default value may not be there. +If a default value exists, it is always different from the predefined +one. The default value gives you an easy specification of an option +value that is expected to be common. If that value would be the +predefined one, you won't need to specify the option in the first +hand. + +The following options are available and explained below: +% +\begin{options} +\item[structure] Select the structure model used by the \cls{cweb} + class. +\item[suppress] Suppress output of different document parts. +\item[baseclass] Select the base document class. +\item[language] Select a language for inserted texts. +\end{options} + + +\sect \label{chunk:structure-option} + +We have two possibilities to output chunks and sections, either +as a flat or as a hierarchic structure. This configuration is selected +by the option |structure|, either through the value |flat| or +|hierarchic|. +% +\begin{options} + +\item[flat] The flat structure is the ``classic'' approach, + the way Plain \cweb{} renders its chunks. Then each chunk starts + with its number. The chunk numbers are output in boldface, followed + by a dot and a quad. + + Sections show chunk numbers as well (remember, they implicitly + start a chunk). ``Important'' sections are added to the table of + contents. + +\item[hierarchic] The hierarchic structure is the + ``modernistic'' (aka \LaTeX{}) approach, then we mark the start of a + chunk by a start sign. (By default, that's a paragraph sign, `\P'). + Chunk numbers are rendered in the margin of the program part. If + there's no program part, no chunk number will be rendered---after + all, they are only used for identification of program parts. + + Section numbers are hierarchic, like in other \LaTeX{} document + classes. + +\end{options} +% +The predefined value of |structure| is |hierarchic|. There doesn't +exist a default value. + + +\sect The option |suppress| allows to select suppression of different +document parts. You have different parts you can suppress. The value +of the |suppress| option is a comma-separated list of identifiers, +enclosed in braces. If you have only one thing you want to suppress, +you don't need to surround that identifier by braces. Possible +identifiers are: + +\begin{options} + +\item[changehints] suppress output of hints that a changefile was + used: The change flag (by default a star at the start of a chunk) + and the list of changed chunks at the end of the document. + +\item[unchanged] suppress output of chunks that are not changed by + a changefile. This value implies |changehints|---by definition all + printed chunks are changed, it doesn't make any sense to hint to + that fact in addition. + +\item[index] suppress the identifier index. + +\item[reflist] suppress the refinement list. + +\item[format] suppress output of `|@f|' directives. + +\end{options} + +The predefined value is the empty list, i.e., nothing is suppressed. +There is no default value since I don't know which parts are +suppressed most often. Send email with your demands, I'll add a +default if I see that there is one major application of this option. + + +\sect \label{chunk:baseclass} The option |baseclass| allows to select +the base document class used by the \cls{cweb} class. In particular, +if the base class has |\chapter| defined the layout of major section +divisions (`|@**|' and `\verb*|@* |') will be changed to be in a +chapter layout. + +The predefined value is |article|, the default value is |report|. + + +\sect The option |language| allows to adapt inserted texts to +different languages. E.g., with the option `|language=german|' German +texts are inserted for cross references, headings, etc. That option +will neither switch on language-specific hyphenation, nor will it +trigger any other processing -- that's the task of Babel or any other +\LaTeX{} language style. + +There is neither a predefined nor a default value. (Without the +|language| option, inserted texts are in English.) + +That option actually triggers the usage of a package +\pkg{cwbl-\textit{language}}. To see if a language is supported, you +have to look for respective files in the \LaTeX{} \cweb{} macro +directory.% + \footnote{That's |texmf/tex/latex/cweb|, on a TDS compliant + installation. Please beware that pseudo operating systems like DOS + cripple file names and shorten them to eight characters; if you + see a file |cwbl-ger.sty| you'll know that the option value + |german| is supported.} +Of course, the easiest way to check the availability is to try it +out -- you'll get a comprehensible error message if there is no +respective \cweb{} language package. In November 1995, |german|, +|french|, and |italian| were supported, more might have been added +later. + +\smallskip + +For compatibility with Babel, some other options (e.g., |german| or +|french|) are also supported as nicknames for +|language=|\textit{option}. If you use packages that make also use of +these options (e.g., \pkg{babel} or \pkg{varioref}), you'll probably +prefer them over |language| since they are picked up by those packages +as well. + + + + + +\chap Restrictions. + +The following restrictions will not be withdrawn as far as I can +see---except if somebody will send me a patch with changes. I tried to +sort the restrictions in order of severity. + + +\sect \emph{(Restrictions due to the design of \cweave{}.)} + +Please be aware that the vertical bar (`\verb"|"') is used by \cweb{} +to delimit small C code pieces in the documentation parts, and is +therefore processed (and mangled) by \cweave{}. \emph{You cannot use +it for \TeX{} any more.} + +In particular, you cannot specify rules for the |tabular| or the +|array| environment. Since you most certainly want to do so: You have +two choices left: +% +\begin{enumerate} + +\item Make sure you have the \pkg{array} package (by Frank Mittelbach and + David Carlisle, from the Tools bundle) installed. Then you may use + the package \pkg{cwebarray}, it defines `|I|' (that's an uppercase~`i') + as a specifier for rules. I.e., instead of + \verb"\begin{tabular}{l|l}" you have to write + |\begin{tabular}{lIl}|. Not the most elegant solution, but it + works\,\dots + +\item + Use `|^^7c|' instead of `\verb"|"'. I.e., instead of + \verb"\begin{tabular}{l|l}" you may write |\begin{tabular}{l^^7cl}|. + +\end{enumerate} +% +These two choices are compatible, you may use both in one document. +Needless to say, I consider the first alternative the better one. + + +\sect Neither a refinement name nor an index entry made by~|@^| may +consist of a \emph{single} dot-accented term. I.e., you must not +write `|@<\.O@>|', `|@^\.o@>|', or even `|@^\.{foolish}@>|'. Of +course you may write `|@^\.o, accent@>|' or `|@< Handle accent \.o @>|'. + + +\sect \emph{(Restriction due to the implementation of \LaTeX{}.)} + +One cannot use restricted C mode in moving arguments. Most notably, +this is disturbing in the section titles and in |\caption| +tags. Sorry, folks. + +(Basically, that's because the definition of |\addcontentsline|, etc., +in the \LaTeX{} kernel is brain damaged; it wants to expand its +argument. And I don't want to maintain redefinitions of such basic +macros.) + + +\sect \emph{(Restrictions due to \TeX{}.)} + +\C++ comments (i.e., from |//| to the end of the line) are typeset as C +comments. This is especially bad if they are used for a whole block of +comment lines, as it is quite common. Please put such comment blocks +in the documentation part. (An |\everyline| implementation would be +needed to lift that restriction, and that's impossible in general.) + + + + +\chap Details. + +The following tags are reserved and must neither be used nor redefined: + |\ATL|, + |\B|, + |\M|, + |\N|, + |\PB|, + and~|\Y|. + +|\9| is already explained in the \cweb{} user manual: It's a special +control sequence used for the index entries tagged with~`|@:|'. Its +default definition is setup in such a way that you can cheat +\cweave{} concerning the sort order of this entry: If you enter +`|@:sort}{print@>|' you will get an index entry ``print'' next to the +place where the index entry ``sort'' would be. But you're allowed to +change this default definition.% + \footnote{Remember: A package is a very good place to place such + redefinitions; your document should be concerned with contents, not + with appearance.} + +The names of all other control sequences defined by this +class---besides the common \LaTeX{} control sequences---start with +|cweb|, |Cweb|, or |cwbb|. Please don't define new control sequences starting +with this prefix. (The control sequences starting with |Cweb| may be +redefined in a package to change the appearance of the \cweb{} +document or the behavior in a controlled manner, check the +description of the internal interface for this possibility.) + + +\sect Plain \cweb{} users +should note that the macros from |cwebmac.tex| are not available any +more. E.g., you cannot use |\.| to typeset typewriter material; use +either |\texttt| or~|\verb|, as it fits the situation. On the other hand, +now you're able to use |\.| for the dot accent, you can define |\3| +for the `sharp~s', |\C++| for the \C++ logo,% + \footnote{I will not define any such logos in this class. A + package like \pkg{logos} with lots of name definitions is the + appropriate place for this.} +etc. + +Another detail for ex-plainies: The table of contents is produced by +the |\tableofcontents| tag (during the second \LaTeX{} run), not +automatically. But this is the standard \LaTeX{} behavior. + + + + +\chap Known Problems. +\label{sec:problems} + +There is a bunch of known bugs, problems, and ommisions. + +\bigbreak + +\noindent \textit{Bugs:} + % +\begin{itemize} + +\item The presentation of |@l| redefinitions is not proper. But it +wasn't in Plain \cweb{}, either. + +\end{itemize} + +\noindent \textit{Problems:} + % +\begin{itemize} + +\item You can't use |\index| and |\makeindex| as-is to create an + index. Add your entries instead to \cweave{}'s index using `|@^|', + `|@.|', or `|@:|'. That problem is caused by a clash in the usage of + the filename extension |idx| by \cweave{} and \LaTeX{}. The + \pkg{index} package by \textsc{David Jones} will help you to define + additional index categories (that use an other extension) if you + don't want to mix identifiers and your other index entries. + +\end{itemize} + +\noindent \textit{Ommisions} (``wouldn't it be nice, if''s): + % +\begin{itemize} + +\item The current chunk number and most section titles (below rank 1 + or~2, depending on the base class) are not available as a mark for + inclusion in a running head. + +\item Better integration of \pkg{rcs} package. That package is used to + support including information from the revision management system + RCS into the document. E.g., the revision log at the end of that + document was produced by that package. + +\end{itemize} + + +\sect I would like to thank those who helped me to improve this bundle. +% In particular, XXX provided XXXsubstantial parts of the code. + +\textsc{Michael M\"uller} and \textsc{Zden\v{e}k Wagner} did thorough +checks that helped me to improve the alpha test version. + +\textsc{Christian Kumpf} triggered the addition of language support +and pointed out where configuration could be made easier by supplying +more macros in the protected interface. \textsc{Andreas Scherer} +provided packages for French and Italian \cweb{} documents. + +\textsc{Laurent Desnogues}, +\textsc{Felix G\"artner}, +and +\textsc{John S. Robinson} +provided error and problem reports. +\textsc{Bronne Louis} pointed out the embarrassing error that section +numbers were not reset when a higher-level section began. + + + +\endSubDocument + + +%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% + +\newpage + +\begin{rcslog} +$DocLog: cweb-user.tex,v $ +\Revision 2.10 1995/11/20 22:34:16 schrod +Add language support.\\ +Triggered by Christian Kumpf \path|<smurf@igd.fhg.de>| and Andreas +Scherer \path|<scherer@physik.rwth-aachen.de>|. + +Use my `official' (ACM) email address. + +\Revision 2.9 1995/11/20 10:39:34 schrod + Explicate that one must not use |\section| etc., that one cannot +\LaTeX{} the \cweb{} program directly but must run \cweave{} first, +and that the \cls{cweb} class does not provide inclusion of \cweb{} +programs in other \LaTeX{} documents.\\ +Clarification triggered by Laurent Desnogues +\path|<laurent.desnogues@aiguemarine.unice.fr>|. + +\Revision 2.8 1995/09/15 10:22:58 schrod +The \cls{cweb} class needs at least \LaTeX{} version +\mbox{$\langle$1994/12/01$\rangle$}.\\ +Problem reported by John S. Robinson \path|jsrobin@umiacs.umd.edu|. + +\Revision 2.7 1995/09/12 22:58:38 schrod +|cwbb| is also a reserved namespace. + +Added acknowledgements. + +\Revision 2.6 1995/08/29 17:34:14 schrod + \LaTeX{} \cweb{}, version 1.0. + +\Revision 2.5 1995/08/29 02:07:28 schrod +Discard dependencies on 10\,pt fonts. + +Support suppression of format directives. + +\Revision 2.4 1995/08/27 19:31:11 schrod +Suppression of index and reference list is supported. + +Discard documentation on |\cwebSecNoEject|, that's part of the +protected interface. + +\Revision 2.3 1995/08/27 17:24:45 schrod +Make usage of baseclass with chapters work. + +\Revision 2.2 1995/08/27 13:25:59 schrod +Add possibility to suppress change hints. Suppression of unchanged +chunks suppress change hints as well, they are meaningless as all +printed chunks are changed by definition. + +Update problems section, they were partly resolved by the current +changes. Added hint to \pkg{index} package that resolves the problem +of extensions clashes with \MakeIndex{}. + +\Revision 2.1 1995/08/25 19:11:17 schrod +Add keyword-value option style, with new \pkg{keyvald} package. + +Hierarchic strucutures are supported now, in addition to the flat +structure of the beta-test version. One can choose with an option. For +that step, the terminology was cleaned up, too: Chunks are not named +sections any more. (That change involved reimplementation of almost +all the structure and toc stuff.) + +The chunk number supplied by \cweave{} is used now, not some computed +number. Change flags are printed, too. + +One can suppress output of unchanged sections. + +One can select the baseclass with an option. That may be used to use +\cls{report} or \cls{book} to get chapter-style layout. Of course, +using an arbitrary baseclass is dangerous, it must conform to the +conventions of \LaTeX{} standard classes. + +\Revision 1.10 1995/08/08 00:14:28 schrod +Updated to \LaTeXe{}, the |cweb| style is now a document class. Used +my standard templates for that, no changes in functionality. + +\Revision 1.9 1993/08/10 14:15:40 schrod +New page on main section only if group level $<$ |\cwebSecNoEject|. +Default for the latter is 3. + +Document that logos will not be defined in this style file. + +Copy of plain macros for |\CwebNumber| does not work. Repaired the +most important one (subscript must be accessed via |\sb|). +Incompatibility to NFSS will be addressed later.\\ +(Problems reported by Zden\v{e}k Wagner \path|<wagner@csearn.bitnet>|.) + +\Revision 1.8 1993/08/10 11:21:03 schrod +Reference to section number does not render a period after the +number any more. + +\Revision 1.7 1993/08/09 18:05:54 schrod +Mentioned that `\verb"|"' cannot be used for LaTeX purposes, in +particular, not for ruled tables. Described workarounds, one of them +is the new style option |cwebarray|.\\ +(Problem reported by Felix G\"artner +\path|<theedge@rbg.informatik.th-darmstadt.de>|.) + +\Revision 1.6 1993/06/17 16:10:06 schrod +`|cweb|' was still an option in the example. + +\Revision 1.5 1993/06/17 15:09:30 schrod +\cweb{} 3.0 was released officially on June 16, 1993. Mentioned in +the documentation that this version is needed for the |cweb| style. + +\Revision 1.4 1993/06/15 15:19:11 schrod +It's a style now, was an option formerly. + +\Revision 1.3 1993/06/14 15:40:42 schrod +Added section about restrictions (no restricted C mode material +in moving arguments, no single dot-accented term in an index entry). + +\Revision 1.2 1993/05/12 20:23:32 schrod +Adapted to recent changes of \cweave{} (of April 93): Main sections +have a group level, represented in the table of contents. + +Boxed the example, it was not legible formerly. + +\Revision 1.1 1993/04/09 15:00:37 schrod +Initial revision + +\end{rcslog} + + + +\end{document} + +% LocalWords: baseclass DFS PSPACE graphicx changehints changefile reflist idx +% LocalWords: cwebarray cwebmac plainies ommisions if''s rcs cwbb uller Zden diff --git a/Master/texmf-dist/doc/latex/cweb-latex/src/cweb.doc b/Master/texmf-dist/doc/latex/cweb-latex/src/cweb.doc new file mode 100644 index 00000000000..8fb7a0a5582 --- /dev/null +++ b/Master/texmf-dist/doc/latex/cweb-latex/src/cweb.doc @@ -0,0 +1,1396 @@ +% $StyleId: cweb.doc,v 3.6 1995/11/30 15:44:53 schrod Exp $ +%---------------------------------------------------------------------- +% Written by Joachim Schrod <schrod@iti.informatik.th-darmstadt.de>. +% Copyright conditions see below (GPL). + +% +% LaTeX class cweb +% support for LaTeX markup in CWEB sources +% +% [LaTeX in MAKEPROG] +% (history at end) + +% If you have received this style file without the user manual (in the +% file cweb-user.dvi or the respective LaTeX file), it's incomplete and +% near to useless. If it was given to you as something that you shall +% use as an author -- complain bitterly to your provider. You need the +% documentation and you have a right on it! (Below you can find info +% where to get the reference version.) + + + + +\documentclass{progltx} + +\usepackage{cweb-doc} % document-specific markup +\usepackage{fullpage} + + +\RCS $StyleRevision: 3.6 $ +\RCS $StyleDate: 1995/11/30 15:44:53 $ + + +\begin{document} + + +\title{The \texttt{cweb} Class\\ + {\large (Implementation)}% + } +\author{% % LaTeX does not discard unnecessary glue... + Joachim Schrod% + \thanks{% + \protect\raggedright + TU~Darmstadt, Computer Science Department, WG Systems Programming, + Alexanderstr.~10, D-64283~Darmstadt, Germany. + Email: \texttt{jschrod@acm.org}% + }% + } +\date{% + Revision \RCSStyleRevision\\ + (as of \RCSStyleDate)% + } + +\maketitle + + +% doesn't work with progltx yet +%\tableofcontents + + + +% ------------------------------------------------------------ +% +% subdocument: The user interface of cweb.cls +% + +\input{cweb-user} + +% +% ------------------------------------------------------------ +% +% subdocument: The internal interface +% + +\input{cweb-conf} + +% +% ------------------------------------------------------------ + + + + +\chap Implementation. + +This implementation must typeset complete files output by \cweave{}. +Large parts of this task are available from the module \pkg{cwebbase}. +It remains mainly to implement document-level markup and front and +back matter (table of contents, index, etc.) In particular, the back +matter material is not trivial; its tags are behind |\end{document}|, +we have to read carefully behind the end of our \cweb{} document to +detect if we have to typeset them at all. + + +\sect This module reserves the namespace |cweb|. + +\beginprog +\ifx \cweb@loaded\undefined + \def\cweb@loaded{$StyleRevision: 3.6 $} +\else + \PackageWarningNoLine{cweb}% + {Some other package already uses namespace `cweb'} +\fi +\endprog + + +\sect Let's identify this class. + +\noindent The code below is explained in the implementation +documentation of the \pkg{rcs} package. + +\beginprog +\begingroup + \def\RCSClass#1#2 $#3: #4 #5\endRCS $#6: #7 #8\endRCS{% + \def\date{#4}\def\id{v#7}% + \ProvidesClass{#1}[\date\space\id\space #2]% + } + + \RCSClass{cweb}{LaTeX markup for CWEB sources} + $StyleDate: 1995/11/30 15:44:53 $: 9999/00/00 \endRCS + $StyleRevision: 3.6 $: 0.0 \endRCS +\endgroup +\endprog + + +\sect This module is from a supported bundle. Send bug reports, +comments and repairs. + +The reference version may be retrieved via anonymous ftp from +|ftp.th-darmstadt.de| [130.83.47.112], directory +|pub/programming/literate-programming/c.c++/|. It's +placed there as a gzipped tar file. (The information on the IP~number +is dated August~29,~1995. +It might have changed, although this is very unlikely. Use +your friendly nameserver.) + + +\sect This is freely distributable software; you can redistribute it +and/or modify it under the terms of the GNU General Public License as +published by the Free Software Foundation; either version~2 of the +License, or (at your option) any later version. + +This software is distributed in the hope that it will be useful, but +\textbf{without any warranty}; without even the implied warranty of +\textbf{merchantability} or \textbf{fitness for a particular purpose}. +See the GNU General Public License for more details. + +You should have received a copy of the GNU General Public License in +the file |License| along with this package; if not, write to the Free +Software Foundation, Inc., 675~Mass Ave, Cambridge, MA~02139,~USA. + + +\sect Before we start we declare some shorthands for category codes. +By declaring the underscore~`(|_|)' as letter we can use it in our +macros. (I agree with \textsc{D.~Knuth} that +|\identifier_several_words_long| is more readable than +|\IdentifierSeveralWordsLong| and in every case better than |\p@@@s|.) + +As this is a \LaTeX{} style file the at sign is a letter anyhow; so +we can use the ``private'' \LaTeX{} macros; and with the underscore we +can make our own macros more readable. But as we have to restore this +category code at the end of this macro file we store its former value +in the control sequence |\CatUsCode|. This method is better than to +use a group because not all macros have to be defined global this way. + +Since somebody might use more styles from me, this cseqs might be +defined already. + +\beginprog +\ifx \CatEscape\undefined + \chardef\CatEscape=0 + \chardef\CatOpen=1 + \chardef\CatClose=2 + \chardef\CatIgnore=9 + \chardef\CatLetter=11 + \chardef\CatOther=12 + \chardef\CatActive=13 % is defined in Plain already + + \chardef\CatUsCode=\catcode`\_ +\fi + +\catcode`\_=\CatLetter % top level macro file +\endprog + + +\sect Problems over problems with different \LaTeX{} versions. I have +decided that I will take the same stand as the \LaTeX{} team: I will +only support the last two official \LaTeX{} versions, folks who run +older software might need to upgrade. + +The implementation will use `star'-forms of \LaTeX{}'s command +definition macros. They were introduced with \LaTeX{} +\mbox{$\langle$1994/12/01$\rangle$}. + +Currently (as of 07 Nov 95), 1995/06/01 is the most recent version. +I want to use a features introduced with that version: |\newif| is not +outer any more. So I check here for the version and supply the new +definition if necessary. + +\beginprog +\NeedsTeXFormat{LaTeX2e}[1994/12/01] +\@ifl@t@r \fmtversion{1995/06/01}{}% else + { + \@namedef{newif}#1{% + \count@\escapechar \escapechar\m@ne + \let#1\iffalse + \@if#1\iftrue + \@if#1\iffalse + \escapechar\count@ + } + \def\@if#1#2{% + \expandafter\def + \csname \expandafter\@gobbletwo\string#1% + \expandafter\@gobbletwo\string#2\endcsname {% + \let#1#2% + }% + } + } +\endprog + + + + +\chap Options. + +Our options are of the ``keyword-value'' category. We realize them by +the \pkg{keyvald} package, an enhancement of the \pkg{keyval} package +from the Graphics bundle that supports defaulting for unknown keys. + +\begin{fixme} + Most of the code below should be extracted and moved to a + \pkg{keyvalx} package (or a new version of \pkg{keyval}). In + particular, support for package/class option processing, enumeration + values, and set values for keyword options. +\end{fixme} + +|\ProcessKeyValOptions| may be used instead of |\ProcessOptions|, it +takes the keyval set identifier as argument. At the end it evaluates +and resets a hook. That hook may be used at option definition to +define further dependencies. (E.g., enabling that option implies +enabling an other option, too.) +% +\begin{fixme} + Like |\ProcessOptions|, I define |\@curroptions|, since I don't know + if that's needed later. The only place where it could been needed is + in the check for unprocessed options. +\end{fixme} + +\beginprog +\RequirePackage{keyvald} + +\let\ProcessKeyValOptionsHook\empty +\providecommand*\ProcessKeyValOptions[1]{% + \edef\@curroptions{\@ptionlist{\@currname.\@currext}}% + \edef\next{\noexpand\setkeys{#1}{\@curroptions}}% + \next + \ProcessKeyValOptionsHook + \let\ProcessKeyValOptionsHook\empty + } +\endprog + + +\sect The enumeration option |structure| selects hierarchic~(0) or +flat~(1) structure. The predefined value is `hierarchic', there is no +default. The option value is bound to |\cweb@structure|. + +\beginprog +%% \define@key@enum{cweb}{structure}{hierarchic,flat}[hierarchic] +\@namedef{KV@cweb@structure@enum:hierarchic}{0 } % <-- space! +\@namedef{KV@cweb@structure@enum:flat}{1 } % <-- space! +\define@key{cweb}{structure}{% + \expandafter\let \expandafter\@tempa + \csname KV@cweb@structure@enum:#1\endcsname % \relax if undefined + \ifx \@tempa\relax + \PackageError{cweb}% + {% + Invalid value `#1' for option structure% + }{% +Possible valid values are `hierarchic' and `flat'.\MessageBreak +The predefined value is `hierarchic'.\MessageBreak +There's no default value, you have to specify one.% + }% + \else + \let\cweb@structure\@tempa + \fi + } +\setkeys{cweb}{structure=hierarchic} +\endprog + + +\sect With the set option |suppress| one can select suppression of +different document parts: |changehints| suppresses output of hints +that a changefile was involved, |unchanged| output of unchanged +chunks, |index| and |reflist| output of index and reference list, and +|format| output of `|@f|' directives. + +The predefined value is no suppression. There is no default value. +% +\begin{fixme} + Shouldn't I supply a default value? `|{index,reflist}|'? +\end{fixme} + +Set options are represented by flags named +|@cweb@|\<option>|@|\<value>|@|. + +\beginprog +%% \define@key@set{cweb}{suppress}{changehints,unchanged,index,reflist,format} +\@for \@tempa :=changehints,unchanged,index,reflist,format\do {% + \expandafter\newif \csname if@cweb@suppress@\@tempa @\endcsname + \csname @cweb@suppress@\@tempa @false\endcsname + }% +\define@key{cweb}{suppress}{% + \@for \@tempa :=#1\do {% + \expandafter\let \expandafter\@tempb + \csname @cweb@suppress@\@tempa @true\endcsname + %% \relax if undefined + \ifx \@tempb\relax + \PackageError{cweb}% + {% + Invalid value `#1' for set option suppress% + }{% +Possible valid values are `changehints', `unchanged',\MessageBreak +or a comma-separated list of these words enclosed in braces.\MessageBreak +Specified values are turned on, you can't turn them off.\MessageBreak +No value is on initially.\MessageBreak +There's no default value, you have to specify one.% + }% + \else + \@tempb + \fi + }% + } + +\g@addto@macro\ProcessKeyValOptionsHook{% + \if@cweb@suppress@unchanged@ + \@cweb@suppress@changehints@true + \fi + } +\endprog + + +\sect The base document class may be specified by the option +|baseclass|. + +\beginprog +\def\CwebBaseClass{article} +\define@key{cweb}{baseclass}[report]{\def\CwebBaseClass{#1}} +\endprog + + +\sect Language-specific adaptions are stored in files +\texttt{cwbl-{\it language}.sty}. They may be accessed as packages or +by the |language| option that takes the \textit{language} as value. We +have to require the respective package at the end of the class, then +all cseqs with default names are defined and may be overwritten. + +For some languages, we provide Babel-compatible option names, to ease +the usage of global class options taken up by other packages (like +\pkg{varioref}), too. + +\beginprog +\def\cweb_set_language#1{% + \IfFileExists{cwbl-#1.sty}{% + %% true + \RequirePackage{cwbl-#1}% + }{% + %% false + \ClassError{cweb}{Language `#1' is unknown.}{% +Language definitions are stored in files `cwbl-<language>.sty'.\MessageBreak +I could not locate the file `cwbl-#1.sty'. If you didn't make a% +\MessageBreak +typo, please check if there is such a file in a directory searched\MessageBreak +by LaTeX.% + }% + }% + } + +\define@key{cweb}{language}{\AtEndOfClass{\cweb_set_language{#1}}} + +\define@key{cweb}{german}[]{\setkeys{cweb}{language=german}} +\define@key{cweb}{french}[]{\setkeys{cweb}{language=french}} +\define@key{cweb}{italian}[]{\setkeys{cweb}{language=italian}} +\endprog + + + + +\chap Configuration, inheritance, and aggregation. + +This chapter sets up access to other modules, therefore we restore the +original catcodes for the time being. + +\beginprog +\catcode`\_=\CatUsCode +\endprog + + +\sect As usual, we provide the possibility to configure this class. + +\beginprog +\InputIfFileExists{cweb.cfg}{% + \typeout{******************************************^^J% + *^^J% + * Using local configuration file cweb.cfg^^J% + *^^J% + ******************************************% + }% + }{} +\endprog + + +\sect Inherit article class. + +\beginprog +\ifx \CwebBaseClass\undefined + \def\CwebBaseClass{article} +\fi +\define@key{cweb}*[]{\PassOptionsToClass{\KV@key}{\CwebBaseClass}} +\ProcessKeyValOptions{cweb} +\LoadClass{\CwebBaseClass} +\endprog + + +\sect OK, now comes our own code; switch on again our lexical +conventions. + +\beginprog +\catcode`\_=\CatLetter +\endprog + + + + +\chap Typesetting \cweave{} output. + +This task has been out-sourced to an other module, \pkg{cwebbase}. As +mentioned at the start of the implementation, there remain mainly some +bigger tasks, they will be implemented in further sections below. For +the most part, the necessary preconditions are actually our class +option definitions, we don't have to do anything. (Gosh, what a +coincidence. :--) We need to tell \pkg{cwebbase}, that it should use +class warnings on problems. + +\beginprog +\let\cweb_warning=\ClassWarningNoLine +\input{cwebbase} +\endprog + + +\sect But there remain also some small things to be done right now: +For flat structure, the counter |secnumdepth| must be set to~11, then +we want to number chunks in the whole document. + +\beginprog +\ifcase \cweb@structure + %% hierarchic + \or + %% flat + \setcounter{secnumdepth}{11} +\fi +\endprog + + +\sect \pkg{cwebbase} uses |\part|, |\chapter| (if bound), and +|\section|. We have no use for any other section division that came +from the base class, let's discard the standard ones. + +\beginprog +\let\subsection\undefined +\let\subsubsection\undefined +\let\paragraph\undefined +\let\subparagraph\undefined +\endprog + + + + +\chap Table of contents. + +The table of contents features only sections, chunks don't have +titles, after all. Titles are indented proportionally to the rank, +with an basic indentation of |\CwebTocIndent|. An entry on major +ranks may be differentiated by layout instead of indentation, as it's +done in standard \LaTeX{} classes, too. + +The counter |\CwebTocIndentMaxLevel| constitutes an upper limit for a +recognized indent (concerning indentation, that is). + +\beginprog +\newdimen\CwebTocIndent + \CwebTocIndent=1.5em +\newcount\CwebTocIndentMaxLevel + \CwebTocIndentMaxLevel=6 % increase indentation if <= max level +\endprog + + +\sect For hierarchic structure, we can ignore chunk entries +completely, but have to supply a bunch of definitions for entries of +type `|section|$l$'. But for flat structure, sections are passed as +chunks, too. Then the current rank is stored in the title, as +|\cwbbRank{|$r$|}|. These differences are so large that we'll have +to make a very large |\ifcase|, over many chunks in this macro source +file. Let's start with the hierarchic stuff. + +\beginprog +\ifcase \cweb@structure + %% hierarchic + \def\l@chunk#1#2{} +\endprog + + +\sect Let's assume that we're subclassed from one of \LaTeX{}'s +standard classes; otherwise this stuff must be redefined, too. They +indent |\section| entries always by the same amount (1.5\,em), our +layout must fit to that. + +If we have chapters, they are also indented by 1.5\,em and +distinguished by layout. The first minor section starts on rank~3 and +is placed on level~2, it should be indented larger. Let's simply add +1.5\,em for each rank, that makes 3\,em for sections of type +|section2|. I.e., we use $f=l$ as a factor for our basic indentation. + +If we don't have chapters, the first minor section starts on rank~2 +and is placed also on level~2, it should be indented like sections +(1.5\,em), as it will be in another layout. I.e., we use $f=l-1$ as +the indentation factor. + +The actual definitions are generated in a loop. `|\l@section|$l$' +expands to `|\cweb@toc_entry{|$l$|}|'. The indentation will be +computed within that macro. + +\beginprog + \count@=2 + \@whilenum \count@<11 \do { + \expandafter\edef \csname l@section\number\count@\endcsname{% + \noexpand\cweb@toc_entry{\number\count@}% + } + \advance \count@ by 1 + } +\endprog + + +\sect A toc entry may be produced by |\@dottedtocline|. That cseq +needs the level (to decide if that entry is produced after all), the +entry indentation, the inner-entry indentation used for the number, +entry text, and page number. + +As explained above, we use an indentation of $f \cdot +|\CwebTocIndent|$, if the level is not larger than +|\CwebTocIndentMaxLevel|. Otherwise we use the indentation of that +level. + +Each entry is kind of an item with the number as the label. We need an +inner-entry indentation to be used for subsequent lines within this +entry. If a number is specified, it's tagged by |\numberline| in the +entry text. With a proper local definition and a pro-forma evaluation +of the text, we get at the width of the number (with a fitting space +for separation to text) and can use that as the inner-entry +indentation. + +\beginprog + \def\cweb@toc_entry#1#2{% % #3 (page) will follow + %% \count@ == factor for basic indentation + %% min(#1,TocIndentMaxRank) - correct_level + \ifnum #1>\CwebTocIndentMaxLevel + \count@\CwebTocIndentMaxLevel + \else + \count@ #1 % <-- space! + \fi + \ifx \chapter\undefined + \advance \count@\m@ne + \fi + %% \dimen@ == inner-entry indentation + %% width of number & space, if existing + \begingroup + \global\dimen@ \z@ % assert value + \def\numberline##1{\settowidth{\global\dimen@}{##1\enskip}}% + \setbox\z@ \hbox{#2}% % evaluates \numberline if specified + \endgroup + \@dottedtocline + {#1}% % level + {\count@ \CwebTocIndent}% % basic indent + \dimen@ % num width + {#2}% % entry; page will follow + } +\endprog + + +\sect We're done with toc configuration for hierarchic structure, +since entry definitions for |chapter| and |section| types exist +already. Let's turn over to the flat structure. + +\beginprog + \or + %% flat +\endprog + + +\sect The section rank determines how they are featured in the table. +Titles on rank~0 are typeset boldface, other titles are dotted +contents lines in roman. + +The rank is specified within the title, i.e., in the +first argument to |\l@chunk|, as an argument for the |\cwbbRank| tag. +With an appropriate definition we might set |\count@| to the +respective value. If it is not set, an explicite invocation of +|\section| or |\addcontentsline| is responsible for this entry. Then +we assume that the rank is~0. + +\beginprog + \def\l@chunk#1{% % page will be processed later + \count@\@ne % default value of group level + \begingroup + \let\numberline\@gobble % width not known yet + \def\cwbbRank##1{\global\count@ ##1\relax}% + \setbox\z@ \hbox{#1}% + \endgroup + \ifnum \count@=\z@ + \let\next\cweb@bold_toc_line + \else + \let\next\cweb@normal_toc_line + \fi + \next{#1}% + } +\endprog + + +\sect A boldface line is typeset similar to the definition of +|\@dottedtocline|, only with fewer parameters. It takes two +parameters, the entry and the page number. + +\beginprog + \def\cweb@bold_toc_line#1#2{% + \addpenalty{\@secpenalty}% + \addvspace{1em plus\p@}% + \begingroup + \noindent \bf + \hangindent\CwebTocIndent + \rightskip\@tocrmarg \parfillskip -\rightskip + \interlinepenalty\@M + \@tempdima\CwebTocIndent % for \numberline + #1\nobreak\hfill \hbox to\@pnumwidth{\hss #2}% + \par + \endgroup + } +\endprog + + +\sect A normal line uses |\@dottedtocline| like the hierarchic stuff. +Rank~1 shall still be on the same level as the bold toc line, so we +have to use $r-1$ as the proportional factor for |\CwebTocIndent|. + +\beginprog + \def\cweb@normal_toc_line#1{% % page will be processed later + \edef\next{\noexpand\@dottedtocline{\the\count@}}% % toc entry level + \ifnum \count@>\CwebTocIndentMaxLevel + \count@ \CwebTocIndentMaxLevel + \fi + \advance\count@\m@ne % group level -= 1 + \next{\count@\CwebTocIndent}% % basic indent + \CwebTocIndent % numwidth + {#1}% % entry + } +\endprog + + +\sect The first argument of |\@dottedtocline| is used to specify a +depth of the issued entry. All entries with a depth larger than +|tocdepth| are discarded. We typeset all entries, our |tocdepth| +value is the largest possible section level. + +\beginprog + \setcounter{tocdepth}{10} +\endprog + + +\sect That's it. Close our distinction between hierarchic and flat +structure. + +\beginprog +\fi +\endprog + + + + +\chap The end of the document. + +The end of a \cweave{} document is a rather complicated thing. It is +described in detail in the \cweave{} interface specification, we repeat the +most important points shortly. +% +\begin{itemize} + +\item The user has issued an |\end{document}| tag in the documentation +part of the last chunk. + +This tag is therefore part of a chunk, the chunk end (i.e., the +|\fi| cseq) does appear later in the document. Particularly, this +means that at the accurence of this tag there is still an open |\if|. + +\item Between the |\end{document}| and the |\fi| a lot of text may +come which must be ignored. Within this text no |\fi| is allowed. + +\item After the |\fi| the document may be finished by |\end|. +(Then the |-x| option of \cweave{} was used.) + +Or additional information about the whole document is added: An +optional tag |\ch|, two mandatory tags |\inx\fin|, and another +optional tag |\con|. + +\end{itemize} + +Btw, the last chunk is marked as changed if any changefile entry has +been applied in the document. (In Plain \cweb{} that chunk is +colloquially used as the introduction to the index, and that shall be +printed always.) If we suppress unchanged chunks, the last +chunk will therefore always be processed, independent if it was really +changed or not. Yet another reason to use a whole chunk just for +|\end{document}|. + +We'll use a the following approach for the implementation of the +iidentifier index and the refinement list (remember that we do not +produce a table of contents): We will check if there is any of these +two lists available. The available ones are typeset. Then we'll +finish the document. This functionality will be bound to |\fin|. +|\inx| will be simply a no-op. |\con| will never be reached, so we +don't have to cope with it. + + +\sect With the introduction of an end-document hook in \LaTeXe{}, one +might have hoped that we can utilize that hook for our task. But this +is not possible: We don't want to add some additional actions +\emph{into} the end-document handling, we want to look ahead and issue +lots of commands \emph{before} the end-document handling has happened. +(If you know \TeX{} by heart, you may skip the rest of the explanation +and go to the next chunk.) + +As a macro language, \TeX{}'s cseqs take those tokens as arguments +that are behind them in the token stream. The token stream is +initially the document source, evaluation of a macro pushes its +expansion (a token list) on front of the stream. I.e., the evaluation +of |\end| will push the expansion on the token stream before +|\enddocument| is even looked at. And |\enddocument| will also push +its expansion before the hook is evaluated. + +If we would allow |\enddocument| to push its expansion, all these +tokens have to be evaluated after we have typeset the backmatter -- +after all, we need a proper document end, don't we? One would need to +skip over them, look at and eventually evaluate tokens that come +behind (from the initial document stream, the document source), and +then evaluate the skipped tokens. That's too hairy, we don't know +exactly how many tokens to skip, this is probable to change in future +\LaTeX{} revisions, etc. + +The pushed tokens from |\end| don't pose so much problems. As +|\enddocument| will terminate the document processing, they are not +evaluated anyhow, we can just throw them away. And |\end| has changed +only once in the last few years, so it seems to be a stable macro -- +necessary as we need to know its expansion for propper skipping. + +A very first consequence of the demand to process tokens after +|\end{document}| is that we have to save the ``standard'' \LaTeX{} +document termination, since we will execute that later. + +\beginprog +\let\cweb@save@enddocument=\enddocument +\endprog + + +\sect When we encounter |\end{document}| we have to skip tokens +until we reach a |\fi|. Then we establish the correct +environment for the material behind. + +The skip itself is actually a bit difficil, so we'll look at it more +closely. |\enddocument| is executed within |\end|. As mentioned above, +the original +definition of |\enddocument| cheats, it terminates \LaTeX{} before +the rest of |\end| is executed. |\end| first checks if we're in +the correct current environment, then it establishs the end +processing for paragraph building environments, closes the +group---each environment is in its own group, executes the end +processing established before, and it may ignore following white +space. + +The |document| pseudo-environment is different in that it isn't in a +group. I.e., we have to open a group which can be closed by |\end| +afterwards. The environment check will be done in this group. With +|\aftergroup| we establish a macro which will skip the rest of the +|\end| processing. This processing is not needed anyhow: |document| +is no paragraph building environment and the following white space is +ignored by ourselves. Since |document| is no paragraph building +environment, the end processing described above is empty, i.e., not +existant. Then come a few tokens which ignore the white space. One of +these tokens is a |\fi|, so we have to skip them first separately. +They are discarded by |\cweb@skip_end|, we can start to look +for our |\fi| afterwards. + +If |\end{document}| is the first (i.e., the only) text in this chunk +we discard the chunk start marker. This is possible since it isn't set +until now, it's in the |\everypar| token register. Of course, then we +have to do the house keeping work stored in |\everypar| ourselves. + +\beginprog +\def\enddocument{% + \if@noskipsec + \global\everypar{}% + \global\@noskipsecfalse + \global\@nobreakfalse + \global\clubpenalty\@clubpenalty + \fi + \begingroup + \aftergroup\cweb@skip_end + } +\endprog + + +\sect Let's make sure that the expansion of |\end| is the one that we +expect. Then we can define |\cweb@skip_end| appropriately, we have to +skip 5~tokens. + +\beginprog +\CheckCommand*\end[1]{% + \csname end#1\endcsname\@checkend{#1}% + \expandafter\endgroup\if@endpe\@doendpe\fi +% 1 2 3 4 5 + \if@ignore\global\@ignorefalse\ignorespaces\fi} + +\def\cweb@skip_end#1#2#3#4#5{\cweb@check_fi} +\endprog + + +\sect |\cweb@check_fi| gets the next token, and invokes +|\cweb@do_check_fi| who really checks it. If this token is |\fi| we +can prepare to finish our document. Otherwise we have to check the +next token. + +Of course, first the token we have checked already must be discarded. +Therefore we insert the followup action by |\afterassignment|. It's +important that the assignment gets \emph{only} the next token, we +have to supply both the equals sign and the optional blank which are +allowed between the both arguments to |\let|. + +\beginprog +\def\cweb@check_fi{% + \futurelet\next \cweb@do_check_fi + } +\def\cweb@do_check_fi{% + \ifx \next\fi + \afterassignment\cweb@finish + \else + \afterassignment\cweb@check_fi + \fi + \let\next= % <-- blank!! + } +\endprog + + +\sect When we found the |\fi|, we also discard it. But we're still +within a chunk, i.e., there is an open |\if|. So we must supply the +|\fi|. + +\TeX{}nical note: |\next| is now equivalent to |\fi|. It must not be +used within a loop where it denotes the tail recursion's action. I.e., +code like + % +\begin{verbatim} + \ifx \foo\bar + \let\next\relax + \else + \let\next\continue + \fi + \next +\end{verbatim} + % +will fail. This trap is closed by a redefinition of |\next|. + +The output of all following lists may be suppressed by some option. +That is tested at the place where the respective list shall be +produced. + +All following lists are typeset in CR state. + +\beginprog +\def\cweb@finish{% + \fi + \let\next\relax + \let\ch\cweb@changed_list + \let\inx\relax + \let\fin\cweb@end_document + \let\end\cweb@save@enddocument + \cwbb@CR + } +\endprog + + +\sect The following lists are started with unnumbered section +divisions. They are either chapters or sections. + +\beginprog +\ifx \chapter\undefined + \let\CwebListSectionDiv=\section +\else + \let\CwebListSectionDiv=\chapter +\fi +\def\CwebListSection{\CwebListSectionDiv*} +\endprog + + +\sect The list of changed chunks is a cross reference list, (nearly) +like all others at the end of a chunk. The only difference is that +we do not show changeflags any more---each chunk number in this list +carries a change flag by definition. + +\TeX{}nical note: The redefinition of |\*| is part of the second +argument of |\CwebCrossRef|. It is \emph{not} a global redefinition. + +\beginprog +\def\CwebCRChanged{% + \CwebCrossRef{The following chunks were changed by the change file:}% + } +\def\CwebChangedListName{Changed Chunks} + +\if@cweb@suppress@changehints@ + \def\cweb@changed_list#1.{} +\else + \def\cweb@changed_list{% + \CwebListSection{\CwebChangedListName}% + \@mkboth{\uppercase{\CwebChangedListName}}% + {\uppercase{\CwebChangedListName}}% + \message{\CwebChangedListName}% + \let\CwebCRSize\relax + \CwebCRChanged \let\*\relax + } +\fi +\endprog + + +\sect As outlined above, we want to check if there are any entries in +the identifier index or the refinement list. The former is stored in +the file |\jobname.idx|, the latter in the file |\jobname.scn|. + +We write a macro |\cweb@if_file_not_empty| that's like |\IfFileExists|. + +\beginprog +\def\cweb@end_document{% + \cweb@if_file_not_empty{\jobname.idx}{% + \if@cweb@suppress@index@ \else + \CwebIdIndex + \fi + }{}% + \cweb@if_file_not_empty{\jobname.scn}{% + \if@cweb@suppress@reflist@ \else + \CwebRefList + \fi + }{}% + \cweb@save@enddocument + } +\endprog + + +\sect The paragraph layout for both the index and the refinement list +is much the same, it is taken from Plain \cweb{}: Each entry is a +paragraph, nearly no skip between the paragraphs (just a bit to +prevent underfull vboxes), no paragraph indentation, ragged right. +Overfull hboxes in the lists don't make sense, so we prevent them. And +we don't allow hyphenation. + +\beginprog +\def\cweb@list_par_layout{% + \parskip \z@ plus .5\p@ + \parindent\z@ + \rightskip \z@ plus 2.5em + \tolerance\@M \hyphenpenalty\@M + } +\endprog + + +\sect We consider a file as ``not empty'' if (1)~the file exists, +(2)~is not empty, and (3)~has a non-empty line at the very front. (In +fact, an empty line at the front means that there should not be +anything behind it---but we can't test this portably.) + +|\cweb@if_file_not_empty| tests the property, it's parametrized by the +file name and gets passed actions for the true and the false case. + +We open the file first. Then we check if it doesn't exist or if it's +empty, both conditions deliver true on |\ifeof|; in this case we +pretend that there was an empty line. (The \LaTeX{} kernel already +once provided a macro for an empty line, |\@defpar|, we repeat that +definition here since it was discarded there.) Otherwise we read the +first line. At this state, the emptiness of the first line is +equivalent to the non-availability of entries, we can easily construct +an appropriate macro call to set the flag. + +\beginprog +\def\@defpar{\par} +\def\cweb@if_file_not_empty#1#2#3{% + \openin\@inputcheck #1\relax + \ifeof \@inputcheck + \let\next\@defpar + \else + \read\@inputcheck to \next + \fi + \closein\@inputcheck + \ifx \next\@defpar + \def\reserved@a{#3}% + \else + \def\reserved@a{#2}% + \fi + \reserved@a + } +\endprog + + +\sect The identifier index is available in the file |\jobname.idx|. +The setup for the index is a mixture of the |theindex| environment of +the \cls{article} class and DEK's index macros. It's typeset in two +columns; the user may specify an introductionary text for the index +by |\cwebIndexIntro|. If there is any introductionary text we add a +medium skip below. + +For the index, |\parfillskip| set in such a way that almost empty +lines are avoided. + +Before we read in this file, we have to bind the special cseqs used +therein. + +\begin{fixme} + \LaTeX{} feature or bug? The optional argument of |\twocolumn| is + typeset as a vbox, therefore the space at top of the section heading + is not discarded at the top of a page. But we know that |\twocolumn| + will start a page in any case, we must prevent the addition of this + space. Looking in the source, we see that setting the flag + |@nobreak| will prevent this addition, let's do that as a + workaround. File a \LaTeX{} bug report? +\end{fixme} + +\begin{fixme} + Surprise: If |@nobreak| is set \emph{and} if there are no index + entries, |\twocolumn| will drop its optional argument. (That + happened to be noticed when there was an error in + |\cweb@if_file_not_empty|.) File a \LaTeX{} bug report? +\end{fixme} + +\begin{fixme} + If |\CwebListSection| is a chapter tag of the standard classes, that + macro will check if it's in twocolumn mode and will try to set + itself at the top of the current page in full width, as it would + have been given in an optional argument of |\twocolumn|. As the + method that optional argument uses (a float) must not be nested, we + must prevent this to happen. We set the |@twocolumn| flag to false, + chapter macros will then think they're evaluated in a one-column + document. It's debatable if that is a kludge or even \emph{the} + solution -- after all, within the optional argument we aren't in + twocolumn mode any more. File a \LaTeX{} bug report? +\end{fixme} + +\beginprog +\newtoks\cwebIndexIntro + \cwebIndexIntro={} +\def\CwebIdIndex{% + %% two column format + \columnseprule\z@ + \columnsep 35\p@ % value is from article.sty + %% introduction of index + \twocolumn[% + \cwbb@tex + \@nobreaktrue % don't add space in front of section + \@twocolumnfalse % \CwebListSection might be \chapter + \CwebListSection{\indexname}% + \edef\intro{\the\cwebIndexIntro}% % is a local def + \ifx \intro\empty + \else + \the\cwebIndexIntro\unskip + \par\medskip + \fi + ]% + \@mkboth{\uppercase{\indexname}}{\uppercase{\indexname}}% + \message{\indexname:}% % tell the user what we're doing + %% paragraph layout + \cweb@list_par_layout + \parfillskip \z@ plus .6\hsize + %% bindings + \let\I\CwebIndexEntry + \let\[\CwebIndexDeclared + \let\*\cwbb@change_flag + %% read the index + \CwebReadIndex + } +\def\CwebReadIndex{\@input{\jobname.idx}} +\endprog + + +\sect An index entry is typeset with the same hanging indentation like +a cross reference list. + +The entry is tagged with |\.|, if it was entered by the \cweb{} +operator~`|@.|'. Then it shall be typeset as a string. But an indexed +name may also want to use |\.| as an accent. This is the same +situation as we had at the refinement names (in module +\pkg{cwebbase}), where we introduced |\cwbb@check_dot| to handle this +case. The same minor restriction as there holds here, a refinement +name may not consist of a single dot-accented expression. (|@:|~helps +in this singular case.) + +\beginprog +\def\CwebIndexEntry#1, {% + \par + \hangindent\CwebNumberListHangindent + \leavevmode + \cwbb@check_dot{#1}:\quad + } +\endprog + + +\sect The chunks where identifiers are declared are noted with +underlined numbers. We also must not forget the default declaration of +|\9|, the tag for the user definable index layout. + +\beginprog +\def\CwebIndexDeclared#1]{$\underline{#1}$} +\def\9#1{} +\endprog + + +\sect The list of the refinement names is available in the file +|\jobname.scn|. The layout is taken from the plain version: ragged +right, each entry is a paragraph, the different cross reference +categories are separated by a quad. + +We must initialize our whole paragraph layout, as the index might have +been suppressed and no initialization has happened there then. If +there was an index, it changed |\parfillskip| and |\*|, let's restore it. + +\beginprog +\def\CwebRefListName{List of Refinements} + +\def\CwebRefList{% + %% paragraph layout: like in index, but + \onecolumn + \cweb@list_par_layout + \parfillskip\@flushglue % must be reset before section heading + \CwebListSection{\CwebRefListName}% + \@mkboth{\uppercase{\CwebRefListName}}% + {\uppercase{\CwebRefListName}}% + \message{\CwebRefListName:}% + %% different bindings + \def\I{\par \hangindent\CwebNumberListHangindent}% + \def\CwebCrossRef##1##2.{\quad {\reset@font\footnotesize ##1~##2.}}% + \let\Q\CwebXRCite + \let\Qs\CwebXRCites + \let\U\CwebXRUse + \let\Us\CwebXRUses + \let\*\cwbb@change_flag + %% Read the list, allow configuration first. + \csname CwebRefListHook\endcsname + \CwebReadPPList + } +\def\CwebReadPPList{\@input{\jobname.scn}} +\endprog + + +\sect Extra definitions for the sake of internationalization. + +\beginprog +\def\CwebXRCite{\CwebCrossRef{Cited in chunk}} +\def\CwebXRCites{\CwebCrossRef{Cited in chunks}} +\def\CwebXRUse{\CwebCrossRef{Used in chunk}} +\def\CwebXRUses{\CwebCrossRef{Used in chunks}} +\endprog + + + +\chap The end. + +Well, after all we're finished with this class. We must not forget to +restore the underscore catcode. + +\beginprog +\catcode`\_=\CatUsCode + +\endinput +\endprog + + +\sect I would like to thank those who helped me to improve this module. +% In particular, XXX provided XXXsubstantial parts of the code. + +\textsc{Michael M\"uller} and \textsc{Zden\v{e}k Wagner} did thorough +checks that helped me to improve the alpha test version. + + + +%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% + +\vskip \PltxPreSectSkip + +\begin{rcslog} +$StyleLog: cweb.doc,v $ +\Revision 3.6 1995/11/30 15:44:53 schrod +Still didn't work with version \mbox{$\langle$1994/12/01$\rangle$}, +the copy of |\newif| was not complete.\\ +Problem reported by Michael Seel \path|<seel@mpi-sb.mpg.de>|. + +\Revision 3.5 1995/11/20 22:34:18 schrod +Add language support.\\ +Triggered by Christian Kumpf \path|<smurf@igd.fhg.de>| and Andreas +Scherer \path|<scherer@physik.rwth-aachen.de>|. + +Use my `official' (ACM) email address. + +\Revision 3.4 1995/11/07 17:55:18 schrod + \LaTeX{} \cweb{} should work at least with the last two \LaTeX{} +versions; make it work with the previous-to-last one, version +\mbox{$\langle$1994/12/01$\rangle$}. For that, one has to install the +\mbox{$\langle$1995/06/01$\rangle$} (non-outer) definition of +|\newif|, and |\hb@xt@| must not be used.\\ +Problem reported by Laurent Desnogues +\path|<laurent.desnogues@aiguemarine.unice.fr>| and somebody else (XXX +-- add name). + +\Revision 3.3 1995/09/17 14:15:36 schrod +Initialize list paragraph layout in refinement list, too. With the +introduction of index suppression, the initialization -- located in +the index code -- might have been skipped. + +\Revision 3.2 1995/09/15 10:22:59 schrod +The \cls{cweb} class needs at least \LaTeX{} version +\mbox{$\langle$1994/12/01$\rangle$}.\\ +Problem reported by John S. Robinson \path|jsrobin@umiacs.umd.edu|. + +\Revision 3.1 1995/09/12 23:00:48 schrod +Moved all code that does the actual typesetting of \cweave{} tags and +is therefore also needed for a |cweb| environment. It's now an own +module named \pkg{cwebbase}. + +\Revision 2.8 1995/09/12 11:44:33 schrod +Standard base classes with chapters (i.e., report \& book) produced an +error at the start of the index: |\chapter| may not be called in the +optional argument of |\twocolumn|. Add a workaround to |\twocolumn|.\\ +Problem reported by Christopher Higgins \path|<C.Higgins@cit.gu.edu.au>|. + +\Revision 2.7 1995/08/29 17:32:36 schrod +Add hook for refinement list. + +Allow internationalization of refinement list.\\ +(Problem reported by Christian Kumpf \path|<kumpf@igd.fhg.de>|.) + +Cweb bundle, version 1.0. + +\Revision 2.6 1995/08/29 15:19:50 schrod +Added (hint to non-existence of) description of the internal interface. + +\Revision 2.5 1995/08/29 02:07:26 schrod +Discard dependencies on 10\,pt fonts. + +Support suppression of format directives. + +\Revision 2.4 1995/08/27 19:31:43 schrod +Discard superfluous space at the top of index. Spacing in refinement +heading was wrong. Repair test on empty auxilliary files. Make +section heading for index and refinement list divisions (more) +configurable. Suppression of index and reference list is supported. + +Make configuration of change flag easier. + +Put section title of changed chunks list in marks, and tell the user +that it got typeset. + +\Revision 2.3 1995/08/27 17:24:46 schrod +Make usage of baseclass with chapters work. + +\Revision 2.2 1995/08/27 13:26:22 schrod +Add possibility to suppress change hints. Suppression of unchanged +chunks suppress change hints as well, they are meaningless as all +printed chunks are changed by definition. + +Realize dependencies in options by |\ProcessKeyValOptionsHook|. +Illegal values for enumeration options don't alter the current value +any more. + +List of changed chunks is an unnumbered section now. + +\Revision 2.1 1995/08/25 19:11:18 schrod +Add keyword-value option style, with new \pkg{keyvald} package. + +Hierarchic strucutures are supported now, in addition to the flat +structure of the beta-test version. One can choose with an option. For +that step, the terminology was cleaned up, too: Chunks are not named +sections any more. (That change involved reimplementation of almost +all the structure and toc stuff.) + +The chunk number supplied by \cweave{} is used now, not some computed +number. Change flags are printed, too. + +One can suppress output of unchanged sections. + +One can select the baseclass with an option. That may be used to use +\cls{report} or \cls{book} to get chapter-style layout. Of course, +using an arbitrary baseclass is dangerous, it must conform to the +conventions of \LaTeX{} standard classes. + +\Revision 1.13 1995/08/08 00:14:29 schrod +Updated to \LaTeXe{}, the |cweb| style is now a document class. Used +my standard templates for that, no changes in functionality. + +\Revision 1.12 1993/08/10 14:15:43 schrod +New page on main section only if group level $<$ |\cwebSecNoEject|. +Default for the latter is 3. + +Document that logos will not be defined in this style file. + +Copy of plain macros for |\CwebNumber| does not work. Repaired the +most important one (subscript must be accessed via |\sb|). +Incompatibility to NFSS will be addressed later.\\ +(Problems reported by Zden\v{e}k Wagner \path|<wagner@csearn.bitnet>|.) + +\Revision 1.11 1993/08/10 11:21:07 schrod +Reference to section number does not render a period after the +number any more. + +\Revision 1.10 1993/08/09 20:08:20 schrod +|\cweb@cweave_bindings| is now a no-op if \cweave{} bindings are in +effect already.\\ +(Problem reported by Michael M\"uller \path|<mimu@mpi-sb.mpg.de>|.) + +\Revision 1.9 1993/08/09 18:05:28 schrod +Left shift operator wasn't defined correctly.\\ +(Problem reported by Michael M\"uller \path|<mimu@mpi-sb.mpg.de>|.) + +\Revision 1.8 1993/06/15 15:22:30 schrod +\textbf{Version 0.2:} Moved from |pub/incoming| to official place, in +Literate Programming Archive. + +\Revision 1.7 1993/06/15 13:25:48 schrod +First attempt to make it a style: |article| is imported. Warns the +user if he tries to use |cweb| as a style option. + +\Revision 1.6 1993/06/15 08:49:23 schrod +|\cweb@check_dot| must not evaluate its argument in an |\edef|, this +causes problems if a |\PB| is within. Now I try hard not to evaluate any +tokens outside of my control. + +Can use |\@defpar| for an empty line in |\cweb@has_entries|, don't +need an own macro. + +\Revision 1.5 1993/06/14 17:50:30 schrod +Handle a missing |\con|: |\inx| and |\con| are now never looked at. +Instead |\fin| checks for the existence of the lists and typeset them +if there are entries available. + +\Revision 1.4 1993/06/14 15:54:18 schrod +Made it work with older versions of \LaTeX{}, too. The |\enddocument| +implementation of cweb.sty depends on the |\end| implementation; check +if the current implementation is one we know of (currently: 91-01-14 +and 92-03-25).\\ +(Problem reported by Michael M\"uller \path|<mimu@mpi-sb.mpg.de>|.) + +Add FSA diagram about processing states. CR state is also switched +to from \TeX{} state (that happens with |\ch| at the document end). + +\Revision 1.3 1993/05/13 17:51:21 schrod +Refinements may also be filenames (`|@(|'). Then the complete name +consists of a |\.| macro call, which is handled now.\\ +(Problem reported by Michael M\"uller \path|<mimu@mpi-sb.mpg.de>|.) + +Made the detection of `|@.|' index entries more robust. + +\Revision 1.2 1993/05/12 18:28:59 schrod +Adapted to recent changes of \cweave{} (of April 93): + +Main sections have a group level, represented in the table of +contents. This changed the complete implementation of section tags. + +New C token cseqs: |\Z| and |\MRL|, implemented as |\CwebLe| and +|\CwebCombinedOp|. + +\Revision 1.1 1993/04/09 15:00:37 schrod +Initial revision + +\end{rcslog} + + + +\end{document} + + + +%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% +Local Variables: +mode: LaTeX +TeX-brace-indent-level: 4 +indent-tabs-mode: t +TeX-auto-untabify: nil +TeX-auto-regexp-list: LaTeX-auto-regexp-list +compile-command: "make cweb.cls" +End: diff --git a/Master/texmf-dist/doc/latex/cweb-latex/src/cwebarray.sty b/Master/texmf-dist/doc/latex/cweb-latex/src/cwebarray.sty new file mode 100644 index 00000000000..1eb73a0d48b --- /dev/null +++ b/Master/texmf-dist/doc/latex/cweb-latex/src/cwebarray.sty @@ -0,0 +1,50 @@ +% $Id: cwebarray.sty,v 1.2 1995/08/08 00:14:31 schrod Exp $ +%---------------------------------------------------------------------- + +% +% LaTeX style option cwebarray +% depends on array style option +% +% [LaTeX] +% (history at end) + +\ProvidesPackage{cwebarray} + + +% +% What's this style option for? +% +% If you use CWEB, you cannot use vertical bars (`|') as rule specifiers in +% tables any more; CWEAVE processes them already. This style option +% defines `I' (that's an uppercase i) as a replacement for the +% vertical bar. +% +% I.e., write +% +% \begin{tabular}{lIl} +% +% instead of \begin{tabular}{l|l}. + +% This option won't work without the array style. + +\ifx \d@llarbegin\undefined + \RequirePackage{array} +\fi + +\newcolumntype{I}{|} + + + +%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% +% +% $Log: cwebarray.sty,v $ +% Revision 1.2 1995/08/08 00:14:31 schrod +% Updated to \LaTeXe{}, the |cweb| style is now a document class. Used +% my standard templates for that, no changes in functionality. +% +% Revision 1.1 1993/08/09 18:05:55 schrod +% Mentioned that `|' cannot be used for LaTeX purposes, in +% particular, not for ruled tables. Described workarounds, one of them +% is the new style option cwebarray. +% [Reported by Felix G\"artner] +% diff --git a/Master/texmf-dist/doc/latex/cweb-latex/src/cwebbase.doc b/Master/texmf-dist/doc/latex/cweb-latex/src/cwebbase.doc new file mode 100644 index 00000000000..280e53f82fc --- /dev/null +++ b/Master/texmf-dist/doc/latex/cweb-latex/src/cwebbase.doc @@ -0,0 +1,2059 @@ +% $StyleId: cwebbase.doc,v 1.4 1995/11/07 17:55:19 schrod Exp $ +%---------------------------------------------------------------------- +% Written by Joachim Schrod <schrod@iti.informatik.th-darmstadt.de>. +% Copyright conditions see below (GPL). + +% +% LaTeX macros cwebbase +% support for incorporation of CWEAVE markup in LaTeX documents +% +% [LaTeX in MAKEPROG] +% (history at end) + + + + +\documentclass{progltx} + +\usepackage{cweb-doc} % document-specific markup +\usepackage{fullpage} + + +\RCS $StyleRevision: 1.4 $ +\RCS $StyleDate: 1995/11/07 17:55:19 $ + + +\begin{document} + + +\title{Typesetting \cweave{} Output} +\author{% % LaTeX does not discard unnecessary glue... + Joachim Schrod% + \thanks{% + \protect\raggedright + TU~Darmstadt, Computer Science Department, WG Systems Programming, + Alexanderstr.~10, D-64283~Darmstadt, Germany. + Email: \texttt{schrod@iti.informatik.th-darmstadt.de}.% + }% + } +\date{% + Revision \RCSStyleRevision\\ + (as of \RCSStyleDate)% + } + +\maketitle + + +% doesn't work with progltx yet +%\tableofcontents + + + + +\chap Introduction. + +This code shall interpret and render the tag set output by \cweave{}. +That's realized in this module as it's needed in multiple places -- +when we want to use \LaTeX{} markup in \cweb{}, when we want to +incorporate \cweb{} pieces in \LaTeX{} documents (as examples, or in +textbooks), and when we want to incorporate whole \cweb{} documents in +a \LaTeX{} document. (That's by far the most difficult task.) + +As \cweb{} operates in modes, for obvious reasons our implementation +is based on a finite state automaton (FSA). The whole tag set is +separated into subsets that are valid in certain states. + +The complete set of tags does not obey any name conventions, some tags +are even in conflict with standard \LaTeX{} tags (e.g.\ |\\|). Most +tags are used to render code pieces. With the introduction of states +we can restrict the visibilities of these tags to those states where +we typeset program parts. The needed functionality will be available +in cseqs of our namespace, we'll map that to \cweave{}'s names in a +group. + +Then we have to implement the rendering of chunks and sections, that's +(more or less) a straight-forward task. Actually, we only set up the +basic code pieces, the integration of section divisions in the actual +\LaTeX{} document is a task of higher-level code; that integration +depends heavily on the type of document. That also implies that we +won't implement front or back matter material (document markup, table +of contents, index, list of refinements, etc.)\ in this module. + + +\sect Let's identify this module. + +\noindent The code below is explained in the implementation +documentation of the \pkg{rcs} package. + +\beginprog +\begingroup + \def\RCSFile#1#2 $#3: #4 #5\endRCS $#6: #7 #8\endRCS{% + \def\date{#4}\def\id{v#7}% + \ProvidesFile{#1}[\date\space\id\space #2]% + } + + \RCSFile{cwebbase.tex}{CWEAVE tags for LaTeX markup} + $StyleDate: 1995/11/07 17:55:19 $: 9999/00/00 \endRCS + $StyleRevision: 1.4 $: 0.0 \endRCS +\endgroup +\endprog + + +\sect This module is part of the supported \textsl{cweb-sty} bundle. +Send bug reports, comments and repairs. + +It assumes that the canonical setup of my modules has happened and can +be used herein. (In particular, the underscore is a letter.) + + +\sect This module reserves the namespace |cwbb|. It assumes that +|\cweb_warning| is bound either to |\PackageWarningNoLine| or +|\ClassWarningNoLine|, depending on the environment. + +\beginprog +\ifx \cwbb@loaded\undefined + \def\cwbb@loaded{$StyleRevision: 1.4 $} +\else + \cweb_warning{cwebbase}% + {Some other package already uses namespace `cwbb'} +\fi +\endprog + + +\sect This is freely distributable software; you can redistribute it +and/or modify it under the terms of the GNU General Public License as +published by the Free Software Foundation; either version~2 of the +License, or (at your option) any later version. + +This software is distributed in the hope that it will be useful, but +\textbf{without any warranty}; without even the implied warranty of +\textbf{merchantability} or \textbf{fitness for a particular purpose}. +See the GNU General Public License for more details. + +You should have received a copy of the GNU General Public License in +the file |License| along with this package; if not, write to the Free +Software Foundation, Inc., 675~Mass Ave, Cambridge, MA~02139,~USA. + + + + +% ------------------------------------------------------------ +% +% subdocument: The interface between CWEAVE and TeX +% + +\input{cweave-spec} + +% +% ------------------------------------------------------------ + + + + +\chap Interface of \pkg{cwebbase} module. + +\pkg{cwebbase} implements most of \cweave{} tags. It sets up tons of +redefinable cseqs, all of them start with |Cweb|. +% +\begin{fixme} + How do I document them? We need a configuration document anyhow. Do + I present here just a list of those cseqs realized in this module + and refer to the config docs? Need to wait until I have it, then + we'll see how the stuff here will be added. +\end{fixme} + + +\sect This module supports both hierarchic and flat section +structuring. You have to bind |\cweb@structure| to 0 (hierarchic) or 1 +(flat) before importing it, to show which structure is wanted. + + +\sect This module supports suppression of different document parts. +You have to define flags |@cweb@suppress@|\<foo>|@| with +$\mbox{\<foo>} \in \{ |changehints|, |unchanged|, |format| \}$ to +denote if suppression shall happen. + + +\sect The counter |secnumdepth| must be~11 if chunks shall be +numbered. That counter doesn't need to be set when this module is +imported, it can be set locally during typesetting \cweb{} code. + +The definitions for flat structure assume that |secnumdepth| is~11 +during typesetting. + + +\sect |\part|, |\chapter| (if available), and |\section| are used for +\cweb{} section divisions (those with the lowest ranks). These +commands will change the respective counters globally and will +add toc lines and marks. You might want to prevent that by redefining +|\addcontentsline| and |\mark|. + +Table of contents macros are not defined for the section divisions +introduced by this module. You must either prevent that they are +written to the |AUX| file at all; or you must define them. +% +\begin{fixme} + Add explanation of introduced section divisions. +\end{fixme} + +For flat structure, all section divisions are of type `|chunk|', like +chunks. The section names start with `|\cwbbRank{|\<rank>|}|'. By +default, |\cwbbRank| takes one argument and expands to the empty list +-- this can be redefined for the table of contents. + + +\sect The document end, as explained in the \cweave{} interface specs +(chunk~\ref{spec:sec:docend}), is not implemented. That must be done +by the user of this module -- if the back matter is to be processed at +all. + + + + +\chap Processing states. + +We have to typeset five different categories of material: +Documentation, program pieces---embedded within the documentation and +as large chunks, \TeX{} material within program pieces (i.e., comments +and refinement names), and cross reference information. Since we need +a complete other environment for the program pieces than for the rest +we design ``states'' where we switch to appropriately. +% +\begin{enumerate} + +\item A chunk starts in the \textsl{documentation state}. + +\item |\B| switches to \textsl{program state}. This can happen in +documentation and program state. + +\item While we process the argument of |\PB| we're in \textsl{restricted +program state}; |\PB| may appear in documentation and in \TeX{} +state. Since \TeX{} state can be switched on within restricted +program state, |\PB| can appear within the argument of itself. + +\item In the arguments of |\C|, |\SHC|, and |\X| we switch to +\textsl{\TeX{} state}. All these cseqs appear only in (restricted) +program state, their official names are actually different. I.e., only +in (restricted) program state these cseq are bound to the meaning +described here. + +\item Cross reference information are attached to most chunks with +refinements. This information is processed in \textsl{CR state}. +After CR state material +comes always the next chunk or the document end, i.e., material in +documentation state. + +\end{enumerate} +% +This FSA is illustrated by the diagram in figure~\ref{fig:fsa}. + +\begin{figure} + +\begin{center} +\chardef\\=`\\ +\input{cweb-fsa.latex} +\end{center} + +\caption{The processing state's FSA. The automaton starts and ends in +the ``documentation'' state.} +\label{fig:fsa} + +\end{figure} + +|\cwbb@documentation| will switch to documentation state, +|\cwbb@program| to program state, |\cwbb@Rprogram| to restricted +program state, |\cwbb@tex| to \TeX{} state, and |\cwbb@CR| to CR state. +If we're already in a state, the switch to this state shall be a +permissible null operation. + + +\sect The basic difference between these states can be named with +two parameters: (1)~The cseq bindings in effect and (2)~the layout +parameters used for paragraph makeup. + +In (restricted) program state and in CR state the text is output +under the control of \cweave{}, and tagged by \cweave{}. The used +tags are from a global namespace and should only be in effect during +these states. We call this tag set the \textsl{\cweave{} bindings}. In the +other two states the tags are largely defined by the user, the tag +set is called the \textsl{user bindings}. The switch to another binding +is always done locally, i.e., if we switch from documentation to +restricted program state within a group we don't have to bother about +the restauration of the user binding; it will be done automatically +by \TeX{} at the end of the group. Nevertheless we must be able to +switch from the \cweave{} bindings back to the user bindings which +were in effect when we activated the \cweave{} bindings. This is +needed for the \TeX{} state which is always activated within +(restricted) program state. + +The parameters for \textsl{program layout} are really special ones +since they need to support the indentation which shows the program +structure. These parameters are used in program and in \TeX{} state. +The \textsl{document layout} parameters established by the user are +used in the other three states. + +The following table shall summarize this. $C$~denotes \cweave{} +bindings, $U$~user bindings, $P$~program layout, and $D$~document +layout. If an entry is empty, its value is not changed on entry in +this state. +% +\begin{center} + \DeleteShortVerb\| +\begin{tabular}{r|c|c} +\multicolumn1{c|}{\textsc{State}} & \textsc{Binding} & \textsc{Layout} \\[1pt] +\hline +documentation & $U$ & $D$ \\ +program & $C$ & $P$ \\ +restricted program & $C$ & \\ +\TeX{} & $U$ & \\ +CR & $C$ & $D$ \\ +\end{tabular} + \MakeShortVerb\| +\end{center} + + +\sect Since the user bindings and the document layout is defined +initially, we don't have to do anything if it's requested. Only if we +change it, i.e., within |\cwbb@cweave_bindings| and +|\cwbb@program_layout|, we redefine |\cwbb@user_bindings| and +|\cwbb@doc_layout|. If they are eventually executed, they shall +rebind themselves back to |\relax|. This way we can switch to +documentation state as often as we want. + +\beginprog +\let\cwbb@user_bindings=\relax +\let\cwbb@doc_layout=\relax + +\def\cwbb@documentation{% + \cwbb@user_bindings + \cwbb@doc_layout + } +\def\cwbb@Rprogram{\cwbb@cweave_bindings} + +\def\cwbb@program{% + \cwbb@cweave_bindings + \cwbb@program_layout + } +\def\cwbb@tex{\cwbb@user_bindings} + +\def\cwbb@CR{% + \cwbb@cweave_bindings + \cwbb@doc_layout + } +\endprog + + + + +\chap Saving and restoring control sequences. + +We have a lot of cseqs which are defined within the namespace of this +module and which will be used with other names. This usage is in a +controlled environment, namely neither in documentation nor in \TeX{} +state. (I.e., the text processed is tagged by \cweave{}, not by humans; +therefore we have a precise specification of the cseqs we have to +accept.) We cannot work with groups where a cseq is just redefined +and \TeX{} takes care for establishing the old binding again; when we +switch from program state to \TeX{} state all the bindings which were +in effect before the program state get active, i.e., in the +documentation state, must be in effect again. We cannot simply +consider the \TeX{} state as something parallel to program state, it +must be a hierarchical relationship: In the program state values are +set up which must be available after switching back from \TeX{} to +program state. + +We save the current binding of a cseq in another cseq, but only if +there exists a binding currently. This is done to save valuable +\TeX{} main memory. (Why don't all those people who talk about +extending \TeX{} not just implement dynamic memory management for the +existing \TeX{}? It would still be \TeX{} and would \emph{really} be +a worthy activity. But not enough fun, I assume\,\dots) Actually, one +can assume that nearly no cseq bindings must be saved at all---the +used names are strange enough. The bindings of |\foo|, i.e., of the +cseq with the name \fbox{\ttfamily foo}, is saved as the cseq with the +name \fbox{\ttfamily cweb\char `\\ foo}, i.e., as +|\csname cweb\string\foo \endcsname|. + + +\sect The save process is not done statically, but by the macro +|\cwbb@rebind| which interprets a list of tuples $({\it old\_name}, +{\it new\_name})$, terminated by the tuple $(|\stop|, |\stop|)$. +Eventually it constructs two new lists, |\cwbb@to_restore| with the +cseqs which had a binding, and |\cwbb@undefined| with the names which +didn't have one. + +The saving is actually done by |\cwbb@save_binding|, |\cwbb@rebind| +is responsible for the initialization, |\cwbb@do_rebind| for the +effective rebinding and the tail recursion on the list. + +\begin{fixme} + We could pull the |\next| assignment in the |\else| branch out of + the loop to get a better performance. Should measure if this is of + interest. +\end{fixme} + +\beginprog +\newtoks\cwbb@undefined +\newtoks\cwbb@to_restore + +\def\cwbb@rebind{% + \cwbb@undefined{}% + \cwbb@to_restore{}% + \cwbb@do_rebind + } +\def\cwbb@do_rebind#1#2{% + \ifx #1\stop + \let\next\relax + \else + \cwbb@save_binding #2% + \let #2=#1% + \let\next\cwbb@do_rebind + \fi + \next + } +\endprog + + +\sect If the cseq to be saved is undefined, it may just be added to +the ``undefined list.'' Otherwise its binding is saved and it's added +to the ``to-be-restored list.'' + +\TeX{}nical note: The cseq-name for the saved binding must be created +before the |\let| is executed. + +\beginprog +\def\cwbb@save_binding#1{% + \ifx #1\undefined + \cwbb@undefined \expandafter{\the\cwbb@undefined #1}% + \else + \expandafter\let \csname cweb\string#1\endcsname =#1% + \cwbb@to_restore \expandafter{\the\cwbb@to_restore #1}% + \fi + } +\endprog + + +\sect The restoration of rebound cseqs is a two-tied activity: All +previously undefined cseqs must be made undefined again, and all saved +cseqs must be restored. Actually, we don't need to reset the two token +lists, but we do it to save space. + +Both |\cwbb@undefine| and |\cwbb@restore_binding| iterate over a list +of cseqs terminated by |\stop|. + +\begin{fixme} + And here the |\next| assignment could be prepended to the loop as + well. +\end{fixme} + +\beginprog +\def\cwbb@restore_bindings{% + \expandafter\cwbb@undefine \the\cwbb@undefined \stop + \cwbb@undefined{}% + \expandafter\cwbb@restore_binding \the\cwbb@to_restore \stop + \cwbb@to_restore{}% + } + +\def\cwbb@undefine#1{% + \ifx #1\stop + \let\next\relax + \else + \let#1\undefined + \let\next\cwbb@undefine + \fi + \next + } +\endprog + + +\sect \TeX{}nical note: As in |\cwbb@save_binding|, the cseq-name for +the saved binding must be created before the |\let|s are executed. + +\begin{fixme} + Another |\next| assignment. +\end{fixme} + +\beginprog +\def\cwbb@restore_binding#1{% + \ifx #1\stop + \let\next\relax + \else + \expandafter\let \expandafter#1\csname cweb\string#1\endcsname + \expandafter\let \csname cweb\string#1\endcsname \undefined + \let\next\cwbb@restore_binding + \fi + \next + } +\endprog + + + + +\chap Chunks. + +All chunks are numbered subsequently. The chunk number is supplied as +part of the markup. In addition, the `changed' tag |\*| may be +appended to the chunk number. + +By default, chunks have approximately two picas vertical space in +front. + +If we use an hierarchic structure, we output the chunk number at the +refinement. (It's only needed to identify the refinement for cross +references.) But chunks are important structural units and thus should +be marked by typographic means that's more visible than the space +mentioned above, we add a `chunk start marker'. + +For a flat structure, we output the chunk number as the chunk start +marker, that is the traditional \WEB{} layout. + +In any case, the documentation part shall be run in after the chunk +start marker. + +\beginprog +\newskip\CwebChunkPreSkip + \CwebChunkPreSkip=2pc plus 1pc minus 6pt +\endprog + + +\sect In \LaTeX{}, structural divisions are best realized by +|\@startsection|, we'll use it also for chunk starts. The chunk start +marker is the section heading. For hierarchic structure, we use a +traditional paragraph sign, used since medieval times to tag new text +units. Our first test showed that the paragraph sign has typically an +underlength, that doesn't look good at the start of a chunk. We raise +the sign by its depth. + +For flat structure, we don't have text as the section heading, +but we have to cope for a deficiency of \LaTeX{}: It always leaves a +quad after the section number (by the way, this is not mentioned in +the documentation---you have to look in the macros to figure this +out). So we use a backskip of $\rm -1\,em$ as the chunk start marker +then. + +Changed chunks are marked by a changeflag, by default a star lapping +to the right. We need the information if that chunk is changed to +decide if we shall output the changeflag. We demand from the chunk +macros that they record that info in the flag |@cwbb_changed_chunk@|. + +Each structural division must be categorized, it belongs to a +\textsl{type}. This type determines also the name of the counter, the +mark handling for this category, and how this section tag is +represented in the table of contents. The type of chunks is `|chunk|' +(trivial, isn't it?). + +Furthermore, a structural division in \LaTeX{} has an associated +\textsl{level}. That level is used to determine if that division get +numbered in text or table of contents. For sections, we'll compute the +level from the rank, it will not be larger than the rank.% + \footnote{Check the next section for a more precise definition of + `rank' and `level'.} +Since the highest section rank is~10, we may use 11 as chunk level. + +For flat structure, we output the number for all section headings, +including chunks. For hierarchic structure we use the default maximum +level as supplied by our document class (e.g., 3~for \cls{article}, +and 2~for \cls{report} or \cls{book}). + +As a last point, we have to care already for sections in flat +structure. They will use the |chunk| category as well, we have to +distinguish them in the table of contents. For that, we'll record the +level in the text, it will get written to the table of contents. + +\beginprog +\newif\if@cwbb_changed_chunk@ +\let\CwebChangeFlag=* +\if@cweb@suppress@changehints@ + \let\cwbb@change_flag=\relax +\else + \let\cwbb@change_flag=\CwebChangeFlag +\fi +\def\cwbb@lap_change_flag{\rlap{\cwbb@change_flag}} + +\ifcase \cweb@structure + %% hierarchic + \let\CwebChunkHeadingStyle\relax + \newbox\cweb@chunk_marker + \setbox\cweb@chunk_marker=\hbox{\P} % tradition from medivial times + \setbox\cweb@chunk_marker=\hbox{% + \raise\dp\cweb@chunk_marker \box\cweb@chunk_marker + } + \def\CwebChunkHeadingMarker{\copy\cweb@chunk_marker} + \def\CwebChunkHeading#1{% + \CwebChunkHeadingMarker + \if@cwbb_changed_chunk@ \cwbb@lap_change_flag \fi + } + \or + %% flat + \def\cwbbRank#1{} + \def\CwebChunkHeadingStyle{\bfseries} + \def\CwebChunkHeading#1{% + \protect\cwbbRank{#1}% + \kern -1em + \if@cwbb_changed_chunk@ \cwbb@lap_change_flag \fi + } +\fi +\endprog + + +\sect Chunks are tagged with |\M| and get passed the chunk number as +argument. Before we evaluate |\@startsection|, we'll start the chunk. +That's common code to be used by sections as well, as they also +implicitely start a chunk. + +\beginprog +\def\M#1{% + \cwbb@start_chunk{#1}% + \@startsection + {chunk}% % category, == counter etc. + {11}% % level + {\z@}% % no indentation for heading + {\CwebChunkPreSkip}% % skip above heading + {\m@ne em}% % run-in heading, 1em distance to text + {\normalfont\CwebChunkHeadingStyle}% % layout + {\CwebChunkHeading{11}}% % text, level might go to toc file + } +\endprog + + +\sect But let's first declare how a chunk number is represented. The +principle rendering is controlled by |\thechunk|, namely, we get +arabic numbers. This rendering is used both for creation of the +section header (if in flat mode) and for cross references. I.e., +|\ref| expands into it. + +But for flat structure we need different expansions: In the section +header a period shall be appended, a cross reference shall appear +without such a period. We use two different macros: |\thechunk| +creates the number with the period, |\cwbb@refchunk| only the number. +The latter is used for a redefinition of |\@currentlabel|. Of course, +for hierarchic structure, both are the same. + +\beginprog +\newcounter{chunk} + +\def\cwbb@refchunk{\arabic{chunk}} + +\ifcase \cweb@structure + %% hierarchic + \let\thechunk=\cwbb@refchunk + \or + %% flat + \def\thechunk{\arabic{chunk}.} +\fi +\endprog + + +\sect Each chunk is started by |\cwbb@start_chunk|, which does the +common actions. First of all, we switch to documentation state. + +Then we store the chunk number in the |chunk| counter. That implies +that the chunk start must not step this counter! We'll take care for +that later by redefining |\refstepcounter|. The actual number is +passed as the argument, including an optional |\*| used to denote +changed chunks. We'll redefine that cseq to update the flag +|@cwbb_changed_chunk@|. + +Each chunk must be typeset with an open |\if| since it is finished by +|\fi|. If we shall suppress the output of unchanged chunks, we have +to use |\iffalse| if not |@cwbb_changed_chunk@|, otherwise |\iftrue|. +The flag |@cwbb_print_chunk@| is set up to provide an appropriate +test. + +When we typeset a chunk, we evaluate |\cwbb@init_print_chunk|. For +hierarchic structure, that cseq will be used to initialize the +typesetting of the chunk number at program parts. + +\beginprog +\if@cweb@suppress@unchanged@ + \def\if@cwbb_print_chunk@{\if@cwbb_changed_chunk@} + \@gobble\fi % close \if in case it's skipped +\else + \let\if@cwbb_print_chunk@=\iftrue + \@gobble\fi % same here +\fi + +\def\cwbb@start_chunk#1{% + \cwbb@documentation + \begingroup + \def\*{\global\@cwbb_changed_chunk@true}% + \global\@cwbb_changed_chunk@false + \global \c@chunk #1\relax % might expand \* + \endgroup + \if@cwbb_print_chunk@ + \cwbb@init_print_chunk + } +\endprog + + +\sect |\@startsection| will step the section counter (i.e., |chunk|). +Since we just assigned it in |\cwbb@start_chunk|, that would leave us +with a counter that's one too high. And our output in cross references +may be wrong, as a period is appended in flat structure. + +That damage all comes from one macro, |\refstepcounter|; let's +advice it to do the Right Thing for |chunk| counters. + +\beginprog +\let\cwbb@refstepcounter=\refstepcounter + +\def\cwbb@string@chunk{chunk} +\def\refstepcounter#1{% + \def\@tempa{#1}% + \ifx \@tempa\cwbb@string@chunk + \protected@edef\@currentlabel{\cwbb@refchunk}% + \else + \cwbb@refstepcounter{#1}% + \fi + } +\endprog + + + + +\chap Sections. + +Important sections are started on a new page, normal sections have the +same space as chunks in front. The very first section does not +automatically start on a new page since a title may be in front of it. + +Each section division has a \textsl{rank}, that denotes the section +depth as seen by \cweave{}. + +For this depth we compute a \textsl{level}, used to denote the section +depth for \LaTeX{}. This section depth is used to decide where +numbering shall stop, or if that entry shall be presented in the table +of contents, etc. + +These two numbers are not the same, due to the support of document +classes with chapters. + +We consider a section as important if it's rank is smaller than~3, +i.e., then we'll start a new page. This limit is controlled by the +number |\CwebRankNoEject|. The rank is used as the section level +and is therefore available to configure depth of typeset numbering and +appearance in table of contents. + +\beginprog +\def\CwebRankNoEject{3} +\endprog + + +\sect Sections are started with the macro |\N|. It's first parameter +is the section's rank. The second parameter is the chunk number, the +third the title of the section. This last parameter must be terminated +by a dot. + +Each section implicitly starts a chunk. As explained above, major +sections may eject the page, and we add our actual section heading +with |\CwebSection|. + +We also tell the user that we have reached the next section by telling +him about the current chunk number. The message is output after the +section header is set, it shall be on the correct page. Since the +token list to be output is expanded at shipout time, we must take care +for the immediate expansion ourselves. + +\beginprog +\def\N#1#2#3.{% + \cwbb@start_chunk{#2}% + \CwebEjectSection{#1}% + \CwebSection{#1}{#3}% + \begingroup + \def\*{}% + \edef\tmp{*#2}% + \message \expandafter{\tmp}% + \endgroup + } +\endprog + + +\sect At the start of a section we usually eject the page if our rank +is not too low. But if the user had a title above the first section, +he won't want this page break. So we try to guess it. If we guess +wrong, the user may either add |\newpage| himself or he may redefine +|\CwebEjectSection| in a package. + +\beginprog +\def\CwebEjectSection#1{% % #1 == rank + \gdef\CwebEjectSection##1{% + \relax + \ifnum ##1<\CwebRankNoEject\relax + \newpage + \fi + }% + } + +\if@titlepage + \CwebEjectSection{} +\fi +\endprog + + +\sect Now comes a heary piece. We have 11~section ranks, and need 11 +different section categories, section counters, etc. For flat +structure, they shall all do the same; for hierarchic structure, they +shall be really different. + +\noindent Let's have a look at the hierarchic structure first, it's +more complex. +% +\begin{itemize} + +\item We associate section rank~0 (that's `|@**|') with document + parts, i.e., map it to |\part|. Many smaller programs won't use it. + +\item If we are used in a class that has a bound cseq |\chapter|, we + associate rank~1 (`|@*0|') with it and rank~2 (`|@*1|') with + |\section|. Otherwise we associate rank~1 with |\section|. These + tags (|\chapter| and/or |\section|) must be defined because many + macros that want to create section divisions will use them. + +\item All other section ranks are placed below |\section|. + |\subsection| and other \LaTeX{} style markup is not supported. + + For hierarchic structure, we identify these sections with category + names `|section|$l$', where $l$ is the respective level, computed + from the rank. Those section names do not exist yet, we have to + create them. + +\end{itemize} + +Now we come to the point why we can't use the rank as the section +level passed to \LaTeX{}: When we assume the level usage of \LaTeX{}'s +standard classes, |\section| is \emph{always} on level~1, independent +of the existence of |\chapter|. The latter is placed on level~0, its +existence only influences the section level of |\part|, which is +either $-1$ or~0. + +This implies that |\subsection| in a standard class is always on +level~2, but our equivalent section division is either `|@*0|' +(rank~1) or `|@*1|' (rank~2). I.e., when we have chapters, we use a +level that's one lower than the rank, otherwise we use the rank. + +For flat structure, we use type |chunk|. This way we'll use the chunk +numbers, they are already set up. + +For flat structure, we will map all ranks to the same section markup. +This markup will also be accessible by |\section| or |\chapter|. They +will be declared to be on a level that is the same as the rank. + + +\sect The printed representation of a section divison number is simply +the presentation from the next upper level, a full stop, and an arabic +section number. When we compute the name of the previous level we must +pay attention to our first section division created this way. This +division must be placed below |section|, without any following number. + +Since registers are a valuable resource in \TeX{}, we have to take +care not to allocate a superfluous one. + +\beginprog +\ifcase \cweb@structure + %% hierarchic + \newcounter{section2}[section] + \@namedef{thesection2}{\@nameuse{thesection}.\arabic{section2}} + \newcounter{section3}[section2] + \@namedef{thesection3}{\@nameuse{thesection2}.\arabic{section3}} + \newcounter{section4}[section3] + \@namedef{thesection4}{\@nameuse{thesection3}.\arabic{section4}} + \newcounter{section5}[section4] + \@namedef{thesection5}{\@nameuse{thesection4}.\arabic{section5}} + \newcounter{section6}[section5] + \@namedef{thesection6}{\@nameuse{thesection5}.\arabic{section6}} + \newcounter{section7}[section6] + \@namedef{thesection7}{\@nameuse{thesection6}.\arabic{section7}} + \newcounter{section8}[section7] + \@namedef{thesection8}{\@nameuse{thesection7}.\arabic{section8}} + \newcounter{section9}[section8] + \@namedef{thesection9}{\@nameuse{thesection8}.\arabic{section9}} + \ifx \chapter\undefined + \newcounter{section10}[section9] + \@namedef{thesection10}{\@nameuse{thesection9}.\arabic{section10}} + \fi +\fi +\endprog + + +\sect Sections are typeset by |\CwebSection|; rank and title are the +arguments. + +For hierarchic structure, we select the proper cseq according to rank: +|\part|, |\chapter| if this is a structure with chapters, |\section|, +and |\cwbb@section{|\<rank>|}| for other ranks. For the last cseq we +also have to supply the section category, it's `|section|$l$' where +$l$ is the rank~$r$ if we don't have chapters, $r-1$ otherwise. + +That category cannot be computed by |\cwbb@section| itself since we +want to use that macro also for flat structure. But there all sections +are basically chunks, we have to use the |chunk| category there. Since +we want to differentiate them in the table of contents nevertheless, +we record the current rank in the title. + +\beginprog +\ifcase \cweb@structure + %% hierarchic + \ifx \chapter\undefined + \def\CwebSection#1#2{% + \ifcase #1 + %% 0 + \let\next\part + \or %% 1 + \let\next\section + \else %% level = rank + \def\cwbb@sect_name{section#1}% + \def\next{\cwbb@section{#1}}% + \fi + \next{#2}% + } + \else + \def\CwebSection#1#2{% + \ifcase #1 + %% 0 + \let\next\part + \or %% 1 + \let\next\chapter + \or %% 2 + \let\next\section + \else %% level = rank - 1 + \count@=#1 \advance \count@\m@ne + \edef\cwbb@sect_name{section\number\count@}% + \edef\next{\noexpand\cwbb@section{\number\count@}}% + \fi + \next{#2}% + } + \fi + \or + %% flat + \def\CwebSection#1#2{% + \def\cwbb@sect_name{chunk}% + \cwbb@section{#1}{\protect\cwbbRank{#1}#2}% + } + \def\section{% + \def\cwbb@sect_name{chunk}% + \cwbb@section{1}% + } +\fi +\endprog + + +\sect We use the same layout for all minor sections. That layout can +be changed by ammended by defining |\CwebSect|$r$|Hook| to redefine +|\CwebSectPreSkip| and/or |\CwebSectLayout| locally. (Of couse, it may +be redefined once globally, too.) These are the values to be used for +arguments 4 and~5 of |\@startsection|, pay attention to the skip sign, +telling about indentation of the first paragraph. And don't forget +that the skip is only used for sections on ranks below +|\CwebRankNoEject|, the others start on a new page anyhow. + +The default definition sets the section title in bold and leaves the +same space as before chunks. The first paragraph of the following +documentation has no indentation. + +Other layout changes (indentation for heading, etc.)\ must be made by +redefining |\cwbb@section|. Or---email me, if you want to do it. Then +I might be enclined to introduce further possibilities. + +\beginprog +\newskip\CwebSectPreSkip + \CwebSectPreSkip = -\CwebChunkPreSkip +\def\CwebSectLayout{\normalfont\bfseries} + +\def\cwbb@section#1{% % #2 (next token) is title + \csname CwebSect#1Hook\endcsname % \relax if undefined + \expandafter\@startsection + \expandafter{\cwbb@sect_name}% % category; for counter, etc. + {#1}% % level is rank + \z@ % no indentation for heading + \CwebSectPreSkip % skip above, maybe no next par indent + \medskipamount % skip below heading + \CwebSectLayout + } +\endprog + + + + +\chap Marks. + +\begin{fixme} + Since I don't know yet how to handle marks I ignore most of them for + the time being. + + Have to change this later. +\end{fixme} + +\beginprog +\let\chunkmark\@gobble + +\count@=2 +\@whilenum \count@<11 \do { + \expandafter\let \csname section\number\count@ mark\endcsname \@gobble + \advance \count@ by 1 + } +\endprog + + + + +\chap Typesetting programs. + +Program pieces come in two flavours: as argument of |\PB|, or as +material after |\B|. In the former case we can use a group for the +switch to the restricted program state, the group end will restore the +previous state again. In the latter case we use |\B| for the switch to +program state, the cross reference list or the next chunk will go to +another state. Since we have to do more for the material after |\B|, +we define this cseq later. + +\beginprog +\def\PB#1{% + \begingroup + \cwbb@Rprogram + \leavevmode + #1% + \endgroup + } +\endprog + + +\sect Note, that |\Y| cannot just be |\smallskip|, as in Plain +\cweb{}. We must assert that the current paragraph is ended before the +vertical glue is inserted, and that's not done by the \LaTeX{} +definition of |\smallskip|. In addition we add some negative penalty, +here is a good place for a page break. As the penalty value we use +half the chunk break penalty---of course a chunk start is an even +better place for a page break\,\dots + +It might be that |\Y| is the very first token in a chunk. Then the +chunk start marker isn't set already since it is regarded as a run-in +section heading. In that case we don't set the vertical skip but +simply start the chunk with the program part. + +\TeX{}nical note: The flag |@noskipsec| may be used to test if we're +immediately after a run-in section heading. + +\beginprog +\newcount\CwebProgPenalty + \CwebProgPenalty=\@secpenalty + \divide\CwebProgPenalty by 2 +\def\Y{% + \if@noskipsec + \else + \par + \penalty\CwebProgPenalty + \smallskip + \fi + } +\endprog + + +\sect Yep, let's unfold the ``official'' names of the cseqs used in +program state. + +If we turn on the \cweave{} bindings, they might be in effect already, +we don't need to establish them again. We can test +|\cwbb@user_bindings| for this case, it will be redefined then. + +\beginprog +\def\cwbb@cweave_bindings{% + \ifx \cwbb@user_bindings\relax + \cwbb@rebind + % indentation and paragraph layout + \cwbb@incr_indent \1% + \cwbb@decr_indent \2% + \cwbb@expr_break \3% + \cwbb@backup \4% + \cwbb@opt_break \5% + \cwbb@break \6% + \cwbb@big_break \7% + \cwbb@noindent \8% + % C/C++ tokens + \CwebRel \?% + \CwebAddress \AND + \CwebComplement \CM + \CwebScope \DC + \CwebEquiv \E + \CwebGe \G + \CwebRightShift \GG + \CwebNe \I + \CwebAssign \K + \CwebLeftShift \LL + \CwebMod \MOD + \CwebNull \NULL + \CwebNot \R + \CwebBinOr \OR + \CwebMemberRef \PA + \CwebThis \this + \CwebOr \V + \CwebAnd \W + \CwebXor \XOR + \CwebLe \Z + \CwebPointer \MG + \CwebPointerMemberRef \MGA + \CwebDecr \MM + \CwebIncr \PP + % more tokens + \CwebId \\% + \CwebIdLetter \|% + \CwebRes \&% + \CwebString \.% %% ( ...Emacs... + \CwebStringBreak \)% + \CwebNumber \T + \CwebCombinedOp \MRL + % goes to TeX state + \CwebComment \C + \CwebCxxComment \SHC + \CwebRefName \X + % CWEB tokens + \CwebMacrosHere \ATH + \CwebDefine \D + \cwbb@format \F + \CwebIdCat \J + \CwebVerbString \vb + % cross reference tags + \cwbb@change_flag \*% + \CwebCRAlso \A + \CwebCRsAlso \As + \CwebCRCite \Q + \CwebCRsCite \Qs + \CwebCRUse \U + \CwebCRsUse \Us + \CwebCREt \ET + \CwebCRsEt \ETs + % finish the list + \stop\stop + \def\cwbb@user_bindings{% + \cwbb@restore_bindings + \let\cwbb@user_bindings\relax + }% + \fi + } +\endprog + + +\sect Since most of the \cweave{} bindings are simple and tedious +coding, we'll have a look at the program layout next. Between two +paragraphs there must not be any skip, the skip used in document +layout is saved in |\cwbb@save@parskip|. A few other layout +parameters from the document layout must be saved as well. + +\beginprog +\newskip\cwbb@save@parskip +\newskip\cwbb@save@rightskip +\newcount\cwbb@save@sem_sfcode +\newcount\cwbb@save@pretolerance +\newcount\cwbb@save@hyphenpenalty +\newcount\cwbb@save@exhyphenpenalty + +\def\cwbb@save_doc_layout{% + \cwbb@save@parskip\parskip + \cwbb@save@rightskip\rightskip + \cwbb@save@sem_sfcode\sfcode`\; + \cwbb@save@pretolerance\pretolerance + \cwbb@save@hyphenpenalty\hyphenpenalty + \cwbb@save@exhyphenpenalty\exhyphenpenalty + } +\def\cwbb@restore_doc_layout{% + \parskip\cwbb@save@parskip + \rightskip\cwbb@save@rightskip + \sfcode`\;=\cwbb@save@sem_sfcode + \pretolerance\cwbb@save@pretolerance + \hyphenpenalty\cwbb@save@hyphenpenalty + \exhyphenpenalty\cwbb@save@exhyphenpenalty + } +\endprog + + +\sect When |\cwbb@program_layout| is called for the very first time +we have to set up some variables. This cannot be done in advance +since they depend on values which might be changed by the user in the +preamble. This setup is done in |\cwbb@layout_init|. + +If the program layout is already active, we must not switch it on +another time. This can be tested by the current binding of +|\cwbb@doc_layout|. + +\beginprog +\def\cwbb@program_layout{% + \ifx \cwbb@doc_layout\relax + \cwbb@layout_init + \cwbb@save_doc_layout + \def\cwbb@doc_layout{% + \cwbb@restore_doc_layout + \let\cwbb@doc_layout\relax + }% + % set new values + \parskip\z@skip + \rightskip\z@ plus 100\p@ minus 10\p@ + \sfcode`\;=3000 + \pretolerance\@M + \hyphenpenalty 1000 % strings can be broken this way + \exhyphenpenalty\@M + \fi + } +\endprog + + +\sect The unit of the basic indentation is stored in +|\CwebIndentUnit|. Continuation lines are indented two units further, +i.e., the initial hanging indentation is 3~units. The current hanging +indentation is kept in |\cwbb@indent|. + +We need undiscardable items which can be used as backspaces, of one +and two units, respectively. This is done best by boxes. The +initialization of the boxes must be postponed until the user had the +chance to change |\CwebIndentUnit|, it's done in the program state +initialization. + +\beginprog +\newdimen\CwebIndentUnit + \CwebIndentUnit=1em +\newdimen\cwbb@indent + +\newbox\cwbb@bak % backspace one unit +\newbox\cwbb@bakk % backspace two units + +\def\cwbb@layout_init{% + \global\setbox\cwbb@bak \hbox to -1\CwebIndentUnit{}% + \global\setbox\cwbb@bakk \hbox to -2\CwebIndentUnit{}% + \global\let\cwbb@layout_init\relax + } +\endprog + + +\sect Now we can formulate how to start typesetting program pieces: We +switch to program state, set the initial indentation, and add the +basic indentation of the first code line. Then we might add a marker +to this program part; for hierarchic structure we want to typeset the +chunk number in the margin of the chunk's program part. That's not +really possible, 'though: definitions and format requests are also +introduced with |\B|, so the chunk number will happen to be at the +start of the definition part if one exists. + +If the code line is the very first text to be typeset in this chunk, +we don't add the basic indentation---we're already indented from the +run-in chunk start marker. But we must not start our chunk with +|\noindent|, the marker would be typeset in the left margin then. (I +consider this as a \LaTeX{} error.) So we just emulate an empty +documentation part, the marker is now set correctly. |\B| can test for +this case: |@noskipsec| is still true then. + +\beginprog +\def\B{% + \cwbb@program + \cwbb@indent 3\CwebIndentUnit \hangindent\cwbb@indent + \ifvmode + \if@noskipsec + \indent % add an empty documentation part + \else + \noindent \kern\CwebIndentUnit + \fi + \fi + \CwebNumberProgramPart + } +\endprog + + +\sect The chunk number shall be added at the first program part. We +have a hook in the chunk start where we can set up +|\CwebNumberProgramPart|. It will redefine itself at first call, to +disable further output on more |\B| occurences in a chunk. + +The chunk number should always be in the left margin. If it's in the +right margin, there would be too much horizontal space between the +code start and the actual number. (The first code line is typically +not very long.) \LaTeX{} does not allow to select the margin for +marginpars, therefore we have to typeset the marginal note `by hand'. +As we know that we're in horizontal mode when we're called, |\vadjust| +is used to create the note. The standard \LaTeX{} dimen register for +the distance between marginal notes and text matter is used, of +course. + +Note that marginal notes may overwrite the chunk number. As it's very +unlikely that somebody issues marginal notes in program mode (that's +quite difficult and involves ingenious usage of the implementation of +this macros), I won't care for that case. + +\begin{fixme} + The baseline of the marginal note shall be aligned to the baseline + of the preceding line. That does not happen by default. Currently, I + assume that the preceding line has a depth of appropriately the + depth of a~`g'. Experiments showed that this gives good results for + most \cweb{} documents. But nevertheless one should determine the + exact depth of the preceding line, to know how much |box0| should be + raised. If anybody knows how this can be done, please tell me. (I + know about the route over the output routine, implementing an own + marginal note category, but this seems too much effort.) +\end{fixme} + +For flat structure, we don't add any program part marker and we +don't need to initialize it. + +\beginprog +\ifcase \cweb@structure + %% hierarchic + \def\cwbb@marginal_chunkno{% + \vadjust{% + \setbox\z@ \hbox to \z@{% + \hss \vphantom{g}% + \small \thechunk + \hskip \marginparsep + }% + \vbox to \z@{% + \hsize\z@ + \vss + \noindent \raise \dp\z@ \box\z@ + }% + }% + } + \def\cwbb@init_print_chunk{% + \def\CwebNumberProgramPart{% + \cwbb@marginal_chunkno + \let\CwebNumberProgramPart\relax + }% + } + \or + %% flat + \let\cwbb@init_print_chunk\relax + \let\CwebNumberProgramPart\relax +\fi +\endprog + + +\sect If a statement is finished, a new paragraph with the basic +indentation has to be started. + +An optional statement break is implemented by a low penalty which +will be selected if the line has to be broken. We assume that the +hanging indentation of the new line is already set correctly, and +have to backup two units to get the basic indentation. It might be +that the line break is not chosen by \TeX{}, we compensate the +backspace for this case. This compensation is discarded at the start +of a new line. + +\beginprog +\def\cwbb@break{% % forced break, between statements + \ifmmode\else + \endgraf % in LaTeX it isn't sure what \par is really... + \noindent + \hangindent\cwbb@indent \kern\cwbb@indent + \copy\cwbb@bakk % go back to basic indentation + \ignorespaces + \fi + } +\def\cwbb@opt_break{% % optional break between statements + \hfil \penalty\m@ne + \hfilneg \kern .5em \kern 2\CwebIndentUnit % discarded on line break + \copy\cwbb@bakk + \ignorespaces + } +\def\cwbb@big_break{% % forced break and a little extra space + \Y + \cwbb@break + } +\def\cwbb@expr_break#1{% % break with penalty #1 * 10 + \hfil \penalty#10 + \hfilneg % discarded on line break + } +\endprog + + +\sect When we increment the indentation, we must not forget to set +the hanging indentation immediately since |\cwbb@opt_break| relies on +the new value. If a continuation line is needed it will also be +indented one unit more, which is ok since it should be distinguishable +from the next line. +% +\begin{fixme} + Shouldn't the hanging indentation be set anew at the end of + |\cwbb@decr_indent| as well? +\end{fixme} + +\beginprog +\def\cwbb@incr_indent{% + \global\advance\cwbb@indent\CwebIndentUnit + \hangindent\cwbb@indent + } +\def\cwbb@decr_indent{\global\advance\cwbb@indent -\CwebIndentUnit} +\def\cwbb@backup{\copy\cwbb@bak} +\def\cwbb@noindent{% % no indentation + \hskip -\cwbb@indent \hskip 2\CwebIndentUnit + } +\endprog + + + + +\chap Program (C or \C++) tokens. + +Since the user might want to change the (somewhat unusual) way some +of the operators are typeset, we supply names for them. + +\beginprog +\let\CwebAnd=\land % logical and, && +\let\CwebEquiv=\equiv % equiv sign, for == +\let\CwebGe=\ge % greater or equal +\let\CwebLe=\le % less or equal +\let\CwebNe=\ne % unequal, != +\let\CwebNull=\Lambda % NULL pointer +\let\CwebNot=\lnot % logical not, ! +\let\CwebOr=\lor % logical or, || +\let\CwebXor=\oplus % bitwise exclusive or, ^ +\endprog + + +\sect Some symbols have to be shifted around, to save computation time +we put them in boxes. + +\begin{fixme} + It's not sure if the following symbols still look good in other font + families than Computer Modern. The movements might need to get + adapted when more experience is available. They were chosen from + Plain \cweb{}, where these were explicit movements measured in + points. I changed the dimension to quads, to achieve at least the + support for different sizes in the Computer Modern family. +\end{fixme} + +\TeX{}nical note: The pointer symbol must not be declared as a math +symbol (e.g., with |\DeclareMathSymbol|). Then \TeX{} would insert an +italic correction behind it, according to rule~17 from appendix~G. +(Please send email if you know of an easy way to prevent that +addition.) So I use the symbol as an accent, but then I have to supply +a nucleus for this math list. That nucleus also determines the total +width of this construct. As the vector accent has a width of +approximately 5\,pt in |cmmi10|, I'm going to use 0.5\,em. (That +heuristic is as good as any other I can think of.) + +\beginprog +%% pointer to struct component (`->'), use symbol of \vec accent +\newbox\cwbb@pointer +\setbox\cwbb@pointer=\hbox{% + \kern -.2em + \lower .3em \hbox{$\vec{\kern .5em}$}% + \kern .1em + } + +\newbox\cwbb@decr % decrement, -- +\setbox\cwbb@decr=\hbox{% + \kern .05em + \raise .1em \hbox{$\scriptstyle {-}\kern -.1em{-}$}% + \kern .05em + } +\newbox\cwbb@incr % increment, ++ +\setbox\cwbb@incr=\hbox{% + \kern .05em + \raise .1em \hbox{$\scriptstyle {+}\kern -.1em{+}$}% + \kern .05em + } +\endprog + + +\sect Although the following symbols are typeset like an ``ordinary +\C++ programmer'' would expect them, we provide own module names +nevertheless. They can now be changed as well, i.e., orthogonality is +enhanced. (And we can use |\cwbb@rebind| for assigning them to their +names while we switch to program state\,\dots). + +Note: The comment at |\CwebMod| says ``modulo/remainder'', as this is +not defined by the C standard, i.e., implementation-dependent. + +\beginprog +\mathchardef\CwebAddress="2026 % `&', as binary op +\let\CwebAssign== % assignment +\let\CwebBinOr=\mid % binary or +\def\CwebComplement{{\sim}} % `~', as ordinary symbol +\def\CwebDecr{\copy\cwbb@decr} % decrement +\def\CwebIncr{\copy\cwbb@incr} % increment +\let\CwebLeftShift=\ll % left shift, << +\def\CwebMod{\mathbin{\hbox{\footnotesize\rm\%}}} % modulo/remainder, % +\def\CwebMemberRef{\mathbin{.*}} % ptr to member (on object) +\def\CwebPointer{\copy\cwbb@pointer} +\def\CwebPointerMemberRef{\mathbin{\CwebPointer*}} % ptr to member (on ptr) +\def\CwebRel{\mathrel?} % relation operator +\let\CwebRightShift=\gg % right shift, >> +\def\CwebScope{\kern.1em{::}\kern.1em} % scope resolution + +\def\CwebThis{\CwebRes{this}} % reserved identifier `this' +\endprog + + +\sect The special identifier |\TeX| remains. In math mode it shall be +the identifier, in other modes the usual logo. We use the definition +supplied by DEK (or \textsc{Silvio Levy}?)\ from Plain \cweb{}. + +\beginprog +\def\TeX{% + {% + \ifmmode\it\fi + \mbox{T\kern-.1667em\lower.424ex\hbox{E}\hskip-.125emX}% + }} +\endprog + + +\sect Some tokens don't have constant names, the name is supplied as +the argument. + +Identifiers are typeset in italics, reserved words and type names in +boldface, and strings in typewriter. For an underscore in the bold +words we use a line that's 50\,\% thicker as the usual line thickness. +-- But what is the usual line thickness. As Knuth uses 0.4\,pt for a +10\,pt Computer Modern font, we'll use 0.06\,em for a CM font in an +arbitrary size. If that line thickness is still OK for other fonts is +doubtful, as it should fit to the stem width. (CM is a rather light +font, after all.) But since there is no font dimension that gives a +hint to an appropriate line thickness, it's our best try. + +\beginprog +\def\CwebCombinedOp#1{\mathrel{\let\K==#1}} % e.g., += operator +\def\CwebId#1{\hbox{\it#1\/\kern.05em}} % identifier, more than one char +\def\CwebIdLetter#1{\hbox{$#1$}} % identifier, one letter +\def\CwebRes#1{% % reserverd words and type names + \hbox{\bf + \def\_{\kern.04em\vbox{\hrule width.3em height.06em}\kern.08em}% + #1\/\kern .05em + }% + } +\endprog + + +\sect In Plain \cweb{} strings are typeset in the font |cmtex10|, +which is a typewriter font with extended ASCII characters. This font +does not have the usual accents and can therefore not used for +typesetting national characters. (Very often they are input in some +8-bit encoding, the respective character code is made active and is +substituted by a cseq which expands to the correct glyph. Of course, +there are better ways, but that's the reality we have to cope with.) +Instead of this special font we use the standard typewriter font. +This has the further advantage that |cmtex10| may not be introduced +to NFSS\@. + +\beginprog +\def\CwebString{% + \hbox \bgroup + \tt + \cwbb@string_setup + \cwbb@string + } +\def\cwbb@string#1{% + #1\kern .05em + \egroup + } +\endprog + + +\sect Within strings certain cseqs have a special meaning; this is +introduced by |\cwbb@string_setup|. The cseqs within strings are +mostly special \TeX{} characters that are to be typeset by cseqs with +the same name. + +But a problem remains: active characters. They will still expand, and +will mess up the string visualization. This is a particular problem +with |german.sty| where the double quote is active. (Of course, that's +a problem -- I'm German. :--) When one typesets strings, a double +quote shall be typeset as double quotes, and must not expand to the +special meaning. Remember that double quotes happen to appear quite +often in strings. As the string argument is already tokenized, one has +to redefine the binding of $(\textit{\textup{symbol}}\ \textbf{.}\ +|#\"|)$. + +\begin{fixme} + Perhaps one might use a loop over all characters except the braces, + and set their catcode to other. That would give us enough security. + + For now, we handle only the double quote. +\end{fixme} + +Every once in a while we have discretionary breaks in a string, +denoted by |\CwebStringBreak|. This break is shown with the C +convention of escaped newlines. + +\beginprog +\def\cwbb@string_setup{% + \chardef\ =`\ % % <-- two spaces ! + \chardef\&=`\& + \chardef\\=`\\ + \chardef\^=`\^ + \chardef\_=`\_ + \chardef\{=`\{ + \chardef\}=`\} + \chardef\~=`\~ + \catcode`\"=\CatOther + } + +\def\CwebStringBreak{\discretionary{\hbox{\tt\char`\\}}{}{}} +\endprog + + +\sect Numbers are typeset in different ways. We use the definition of +Plain \cweb{}. +% +\begin{fixme} + Should add a specification of the possible input and an explanation + of the macros below. In particular, that the closing brace after + |\aftergroup| is used much later is probably not grokked by most + \TeX{} programmers. +\end{fixme} + +\beginprog +\def\cwbb@oct{\hbox{$^\circ$\kern-.1em\it\aftergroup\?\aftergroup}} +\def\cwbb@hex{\hbox{$^{\scriptscriptstyle\#}$\tt\aftergroup}} + +\def\CwebNumber#1{% % octal, hex, or decimal constant + \hbox{% + $% + \def\?{\kern.2em}% + \def\$##1{\egroup\sb{\,\rm##1}\bgroup}% suffix to constant + \def\_{\cdot 10^{\aftergroup}}% power of ten (via dirty trick) + \let\~\cwbb@oct \let\^\cwbb@hex + {#1}% + $}% + } +\endprog + + +\sect Comments are typeset in \TeX{} state. We add a hook, the user +shall be able to change the layout (e.g., he might want another font). +% +\begin{fixme} + Currently \C++ comments are typeset like C comments. This is + horrible in usual circumstances, i.e., when complete blocks of text + are prefixed with |//|. We should simply catenate all these text and + typeset it as one paragraph, each line prefixed by |//|. But then we + have to implement an |\everyline| first, and since that's not so + easy we postpone it\,\dots +\end{fixme} +% +In front of a C comment there is an optional stmt break, with 2\,quad +in front if the line is not broken and 1.5\,quad if the line is broken. + +\beginprog +\def\CwebComment#1{% + \5% % 0.5em will be discarded on line break + \hskip 1.5em + $/\ast\,$% + {\cwbb@tex + \CwebCommentHook + #1% + }% + $\,\ast/$% + } +\let\CwebCommentHook\relax +\let\CwebCxxComment\CwebComment +\endprog + + + + + +\chap \cweb{} tokens. + +We distinguish three categories of \cweb{} tokens: (1) Those which +output constant text, (2) those which have attributes to be displayed +in a special way, and (3) those which start a new structure element, +namely |@d| and~|@f|. Let's consider them in this order. + + +\sect \cweb{} tokens which expand in a constant string are the +identifier catenation operator~(`|@&|') and the macro placement +directive~(`|@h|'). + +\beginprog +\def\CwebIdCat{\CwebString{@\&}} +\def\CwebMacrosHere{% + \begingroup + \def\CwebRefNumber##1{}% + \CwebRefName :Preprocessor definitions\X + \endgroup + } +\endprog + + +\sect Verbatim program strings, i.e., strings passed verbatim by +\ctangle{}~(`|@=|') are typeset like normal strings, but within a box. +We use 2\,pt as the separating distance, this is set locally. + +\beginprog +\def\CwebVerbString#1{{\fboxsep\tw@\p@ \fbox{\CwebString{#1}}}} +\endprog + + +\sect The refinement names are typeset in angles, this has a long +tradition. We burry the typesetting of the chunk number (which is +the first parameter) in a macro call; the user may change this to +achieve special effects. +% +\begin{fixme} + One could think about handling special values of the chunk numbers + differently. E.g., an empty argument is not typeset at all, a~0 + triggers a marginal note about a missing definition, etc. +\end{fixme} + +A refinement name may be typeset both in math and horizontal mode. +The name itself is typeset in horizontal mode, of course; for the +angles we need math mode. Therefore we assert at the start of the +macro that we're not in math mode any more. At the end we switch back +to math mode if we've started in it. This conditional switch from and +back to math mode is done by |\cwbb@toggle_text|. + +\TeX{}nical note: If we're in math mode |\cwbb@toggle_text| must be +defined globally, as it will turn off math mode and the definition +would be un-made then. The second invocation of |\cwbb@toggle_text| +would be undefined then. + +A refinement may also consist of the file name the expansion of this +refinement shall be written to. +% +\label{sec:refname-dot} +% +This file name is tagged with |\.|, it shall be typeset like a string. +But the user shall be able to use the dot accent in the refinement +name as well. We check if the text consists solely of the tag and its +argument; in this case we substitute |\.| with |\CwebString|. +Otherwise we leave it as it is. Then a refinement name may not +consist of a single dot-accented expression---well, that's highly +unlikely. (Nevertheless I'll document it in the user's manual\,\dots) + +\begin{fixme} + One should be able to use |\EnsureMath| here. But the refinement + name must not be set in math mode and must not be set in an hbox; + the latter would prevent line breaks. +\end{fixme} + +\beginprog +\def\CwebRefName#1:#2\X{% + \ifmmode \gdef\cwbb@toggle_text{\null$\null}% + \else \let\cwbb@toggle_text\relax + \fi + \cwbb@toggle_text + $\langle\,${\cwbb@tex \cwbb@check_dot{#2}\CwebRefNumber{#1}}$\,\rangle$% + \cwbb@toggle_text + } +\def\CwebRefNumber#1{% + {\reset@font \footnotesize \kern .5em#1}% + } +\endprog + + +\sect The next macro, |\cwbb@check_dot| is a hassle to implement. We +must check if it's argument consists solely of a `|\.|'~tag, together +with the argument to this tag. Then the `|\.|'-argument shall be +typeset like a string, |\.| itself must be ignored. + +The check for |\.| must not evaluate the parameter. Especially |\PB| +must not be evaluated within, it would lead to havoc with the +redefinition of all bindings, etc. (At least that's the empirical +result---it happened; although I don't know where the problems are. +Anyhow, it wasn't in the specs of the rebinding process.) All other +cseqs which shouldn't be used in moving arguments may cause problems +here, too. Note that |\protect| is no solution here, the user often is +not aware of the presence of \cweave{}-generated tags. As an example, +consider the refinement \verb"@< local variables of |foo| @>|" which +will lead to a call with the argument `|local variables of \PB{\\{foo}}|'. +|\\| must not be evaluated here. + +I use an approach where I hope that it will work---but to be honest, +I'm not sure. I want to define a macro |\next| which shall have an +empty expansion iff |#1| was a `|\.|'~tag with its argument. Empty +refinement names are not allowed so we can assume that |#1| consists +of at least one token. I evaluate the first token of |#1| before the +|\def| is done. With an appropriate binding a |\.| will disappear, +perhaps leaving an empty token list. I also take care that |\PB| is +not evaluated then, by temporary rebinding it to |\relax|. Thereby I +have taken care for all cseqs \cweave{} might have introduced at the +very front of |#1|. Other cseqs are user-tags. If they cause +problems, the user might circumstance them by adding |\protect|. + +\beginprog +\def\cwbb@check_dot#1{% + \begingroup + \let\.\@gobble + \let\PB\relax + \expandafter\def \expandafter\next \expandafter{#1}% + \ifx \next\empty + \gdef\next##1{\CwebString}% + \else + \global\let\next\relax + \fi + \endgroup + \next #1% + } +\endprog + + +\sect \cweb{} macro definitions (done by~|@d|) and format directives are +basically program text. When they are started we are already in +program state, i.e., |\B| has appeared in front of it. But since we have +an introductionary identifier (either |#define| or |format|) to set at +the very front we increase the indentation by three units. + +The output of format directives may be suppressed. Then we have to +gather both identifiers that will appear after |\F|. Both of them +consist of one group-token (a cseq and the identifier name), after +each identifier one other token separates them. I.e., we have to +gobble six group-tokens. As the very last token is |\par| if a format +directive follows directly on another one, the definition must be +long. + +\TeX{}nical note: |\cwbb@format| must expand to |\CwebFormat|, not to +its expansion. Otherwise a redefinition of |\CwebFormat| in a package +would not be used. + +\beginprog +\def\CwebDefine{\cwbb@macro{\#define}} +\def\CwebFormat{\cwbb@macro{format}} +\if@cweb@suppress@format@ + \long\def\cwbb@format#1#2#3#4#5#6{} +\else + \def\cwbb@format{\CwebFormat} +\fi + +\def\cwbb@macro#1{% + \global\advance\cwbb@indent \tw@\CwebIndentUnit + \cwbb@incr_indent + \CwebRes{#1 }% % <-- blank! + } +\endprog + + + + +\chap Refinement cross references and changes. + +At the end of chunks with refinements, \cweave{} may output cross +reference information: In which chunks additional definitions for +this refinement are found, where this refinement is used, and where it +is cited. This cross reference information is always introduced by a +tag followed by a non-empty list of chunk numbers. Different tags +are used for lists with only one element and for lists with more than +one element, this way the introductionary text may be adapted. + +Within a list of $n$~chunk numbers the first $n-1$~numbers are +separated by commas with one following blank. The last two numbers +are separated by |\ET| if~$n=2$, and by |\ETs| if~$n>2$. If a chunk +is changed by the changefile, a changeflag~(`|\*|') is appended to +its number. The list is eventually terminated by a full stop. + + +\sect We separate the cross reference information by a smallskip from +the refinement or from a previous cross reference information. The +information itself is typeset in a smaller font, as it is auxilliary, +inserted stuff. The number list has a hanging indentation of +|\CwebNumberListHangindent|. But beware: This isn't a dimension +register, it's a macro. This way one can use ems as dimensions. + +The actual size change is represented by a macro. We will redefine it +when we need an other size, e.g., in the lists at the end of the +document. + +We must assert that we're in CR state while we set a cross reference. + +The first parameter of a cross reference information unit is the +introductionary text, the second is the number list. The parameters +must be evaluated in a group---local parameter changes therein must +not influence the environment. + +\beginprog +\def\CwebNumberListHangindent{2em} +\let\CwebCRSize\footnotesize +\def\CwebCrossRef#1#2.{% + \par\smallskip + \cwbb@CR + \begingroup + \reset@font \CwebCRSize + \noindent \hangindent\CwebNumberListHangindent\relax + #1~#2.\par + \endgroup + } +\endprog + + +\sect Well, let's define all introducing tags, which in fact start the +cross reference information unit. The number list is gathered by +|\CwebCrossRef|. + +\beginprog +\def\CwebCRAlso{\CwebCrossRef{See also chunk}} +\def\CwebCRsAlso{\CwebCrossRef{See also chunks}} + +\def\CwebCRCite{\CwebCrossRef{This code is cited in chunk}} +\def\CwebCRsCite{\CwebCrossRef{This code is cited in chunks}} + +\def\CwebCRUse{\CwebCrossRef{This code is used in chunk}} +\def\CwebCRsUse{\CwebCrossRef{This code is used in chunks}} + +\def\CwebCREt{ and~} +\def\CwebCRsEt{, and~} +\endprog + + + + +\chap Bells and whistles. + +\begin{fixme} + Just copied from Plain \cweb{}. It doesn't work + anyhow---|\char"|\<xy> usually doesn't typeset the correct + character. That's because this character is most probably an active + character; at least that's typical for the way \TeX{} systems are + used in Europe. I have to think about the implementation. +\end{fixme} + +\beginprog +\def\ATL{\par\noindent\bgroup\catcode`\_=12 \postATL} % print @l in limbo +\def\postATL#1 #2 {\bf letter \\{\uppercase{\char"#1}} + tangles as \tentex "#2"\egroup\par} +\def\noATL#1 #2 {} +\def\noatl{\let\ATL=\noATL} % suppress output from @l +\endprog + + + +\chap The end. + +Well, after all we're finished with this module + +\beginprog +\endinput +\endprog + + + +%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% + +\vskip \PltxPreSectSkip + +\rcsLogIntro{After the actual revision log of this module, we give an + extract of log messages of the \cls{cweb} class. This module was + extracted from revision~2.8 of that class, we present to you the + log entries that concern the code of the new module.} + +\begin{rcslog} +$StyleLog: cwebbase.doc,v $ +\Revision 1.4 1995/11/07 17:55:19 schrod + \LaTeX{} \cweb{} should work at least with the last two \LaTeX{} +versions; make it work with the previous-to-last one, version +\mbox{$\langle$1994/12/01$\rangle$}. For that, one has to install the +\mbox{$\langle$1995/06/01$\rangle$} (non-outer) definition of +|\newif|, and |\hb@xt@| must not be used.\\ +Problem reported by Laurent Desnogues +\path|<laurent.desnogues@aiguemarine.unice.fr>| and somebody else (XXX +-- add name). + +\Revision 1.3 1995/11/07 14:13:12 schrod +Reset section numbers when higher-level section begins.\\ +Error reported by Bronne Louis \path|<bronne@montefiore.ulg.ac.be>|. + +\Revision 1.2 1995/11/06 10:58:47 schrod + Support active double quotes in strings. That happens when +|german.sty| is used. Problems with other active characters may still +appear. + +\Revision 1.1 1995/09/12 23:02:22 schrod +Moved all code that does the actual typesetting of \cweave{} tags and +is therefore also needed for a |cweb| environment. It's now an own +module named \pkg{cwebbase}. + +Chunk markers and chunk marginal numbers don't hang below the +baseline any more. + + +\item[] \noindent\hrulefill\break + \textbf{Extract of the revision log for |cweb.doc|:} + + \vskip 2ex + +\Revision 2.5 1995/08/29 02:07:26 schrod +Discard dependencies on 10\,pt fonts. + +Support suppression of format directives. + +\Revision 2.4 1995/08/27 19:31:43 schrod +Make configuration of change flag easier. + +Put section title of changed chunks list in marks, and tell the user +that it got typeset. + +\Revision 2.3 1995/08/27 17:24:46 schrod +Make usage of baseclass with chapters work. + +\Revision 2.2 1995/08/27 13:26:22 schrod +Add possibility to suppress change hints. + +\Revision 2.1 1995/08/25 19:11:18 schrod +Hierarchic structures are supported now, in addition to the flat +structure of the beta-test version. One can choose with an option. For +that step, the terminology was cleaned up, too: Chunks are not named +sections any more. (That change involved reimplementation of almost +all the structure and toc stuff.) + +The chunk number supplied by \cweave{} is used now, not some computed +number. Change flags are printed, too. + +One can suppress output of unchanged sections. + +\Revision 1.12 1993/08/10 14:15:43 schrod +New page on main section only if group level $<$ |\cwebSecNoEject|. +Default for the latter is 3. + +\Revision 1.11 1993/08/10 11:21:07 schrod +Reference to section number does not render a period after the +number any more. + +\Revision 1.10 1993/08/09 20:08:20 schrod +|\cweb@cweave_bindings| is now a no-op if \cweave{} bindings are in +effect already.\\ +(Problem reported by Michael M\"uller \path|<mimu@mpi-sb.mpg.de>|.) + +\Revision 1.9 1993/08/09 18:05:28 schrod +Left shift operator wasn't defined correctly.\\ +(Problem reported by Michael M\"uller \path|<mimu@mpi-sb.mpg.de>|.) + +\Revision 1.6 1993/06/15 08:49:23 schrod +|\cweb@check_dot| must not evaluate its argument in an |\edef|, this +causes problems if a |\PB| is within. Now I try hard not to evaluate any +tokens outside of my control. + +Can use |\@defpar| for an empty line in |\cweb@has_entries|, don't +need an own macro. + +\Revision 1.4 1993/06/14 15:54:18 schrod +Add FSA diagram about processing states. CR state is also switched +to from \TeX{} state (that happens with |\ch| at the document end). + +\Revision 1.3 1993/05/13 17:51:21 schrod +Refinements may also be filenames (`|@(|'). Then the complete name +consists of a |\.| macro call, which is handled now.\\ +(Problem reported by Michael M\"uller \path|<mimu@mpi-sb.mpg.de>|.) + +Made the detection of `|@.|' index entries more robust. + +\Revision 1.2 1993/05/12 18:28:59 schrod +Adapted to recent changes of \cweave{} (of April 93): + +Main sections have a group level, represented in the table of +contents. This changed the complete implementation of section tags. + +New C token cseqs: |\Z| and |\MRL|, implemented as |\CwebLe| and +|\CwebCombinedOp|. + +\Revision 1.1 1993/04/09 15:00:37 schrod +Initial revision + +\end{rcslog} + + + +\end{document} + + + +%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% +Local Variables: +mode: LaTeX +TeX-brace-indent-level: 4 +indent-tabs-mode: t +TeX-auto-untabify: nil +TeX-auto-regexp-list: LaTeX-auto-regexp-list +compile-command: "make cwebbase.tex" +End: diff --git a/Master/texmf-dist/doc/latex/cweb-latex/src/cwebparts.doc b/Master/texmf-dist/doc/latex/cweb-latex/src/cwebparts.doc new file mode 100644 index 00000000000..3eec51882b1 --- /dev/null +++ b/Master/texmf-dist/doc/latex/cweb-latex/src/cwebparts.doc @@ -0,0 +1,452 @@ +% $StyleId: cwebparts.doc,v 1.1 1995/09/16 17:01:59 schrod Exp $ +%---------------------------------------------------------------------- +% Written by Joachim Schrod <schrod@iti.informatik.th-darmstadt.de>. +% Copyright conditions see below. + +% +% LaTeX package cwebparts +% include parts of CWEB programs in LaTeX documents +% +% [LaTeX in MAKEPROG] +% (history at end) + +%% FIXME: The user manual is not in cwebparts-user.tex. Most probably +%% I'll make one user manual for the whole bundle; due to the usage of +%% cwebbase class and package have lots in common. +%% +%% % If you have received this style file without the user manual (in the +%% % file cwebparts-user.tex or the respective DVI file), it's incomplete and +%% % near to useless. If it was given to you as something that you shall +%% % use as an author -- complain bitterly to your provider. You need the +%% % documentation and you have a right on it! (Below you can find info +%% % where to get the reference version.) + + + + +\documentclass{progltx} + +\usepackage{cweb-doc} % document-specific markup +\usepackage{fullpage} + + +\RCS $StyleRevision: 1.1 $ +\RCS $StyleDate: 1995/09/16 17:01:59 $ + + +\begin{document} + + +\title{The \texttt{cwebparts} Package\\ + {\large (Implementation)}% + } +\author{% % LaTeX does not discard unnecessary glue... + Joachim Schrod% + \thanks{% + \protect\raggedright + TU~Darmstadt, Computer Science Department, WG Systems Programming, + Alexanderstr.~10, D-64283~Darmstadt, Germany. + Email: \texttt{schrod@iti.informatik.th-darmstadt.de}.% + }% + } +\date{% + Revision \RCSStyleRevision\\ + (as of \RCSStyleDate)% + } + +\maketitle + + +% doesn't work with progltx yet +%\tableofcontents + + + +\chap What's this style for?. + +This style supports inclusion of \cweb{} program parts in \LaTeX{} +documents. For that it provides the environment |cweb| where \cweave{} +output may be stuffed in. + +A more versatile explanation is available in the user manual of the +\cweb{} bundle; this package is part of that bundle. +% +\begin{fixme} + Read: ``should be available'' -- there is no bundle user manual + (yet). |cweb-user.tex| will be enhanced after first testing of this + code. +\end{fixme} + + +\sect The internal interface is also not described in this source, +refer to the configuration documentation for this bundle. +% +\begin{fixme} + Even worse: I don't even have written the internal specs for the + \cls{cweb} class, not to speak of this package. +\end{fixme} + + + + +\chap Implementation. + +Most of the implementation is made available by the module +\pkg{cwebbase}. We have to supply macros it demands from a user, and +have to deliver the |cweb| environment. + + +\sect This module reserves the namespace |cweb|. (This implies that it +cannot be used with the \cls{cweb} class. But you don't want to do +that anyhow. :--) + +\beginprog +\ifx \cweb@loaded\undefined + \def\cweb@loaded{$StyleRevision: 1.1 $} +\else + \PackageWarningNoLine{cwebparts}% + {Some other package already uses namespace `cweb'} +\fi +\endprog + + +\sect Let's identify this package. + +\noindent The code below is explained in the implementation +documentation of the |rcs| package. + +\beginprog +\begingroup + \def\RCSPackage#1#2 $#3: #4 #5\endRCS $#6: #7 #8\endRCS{% + \def\date{#4}\def\id{v#7}% + \ProvidesPackage{#1}[\date\space\id\space #2]% + } + + \RCSPackage{cwebparts}{include CWEB program parts in LaTeX documents} + $StyleDate: 1995/09/16 17:01:59 $: 9999/00/00 \endRCS + $StyleRevision: 1.1 $: 0.0 \endRCS +\endgroup +\endprog + + +\sect This module is supported. Send bug reports, comments and +repairs. + +The reference version may be retrieved via anonymous ftp from +|ftp.th-darmstadt.de| [130.83.47.112], directory |pub/tex/latex|. It's +placed there as a gzipped tar file. (The information on the IP~number +is dated September~13,~1995. +It might have changed, although this is very unlikely. Use your +friendly nameserver.) + + +\sect This is freely distributable software; you can redistribute it +and/or modify it under the terms of the GNU General Public License as +published by the Free Software Foundation; either version~2 of the +License, or (at your option) any later version. + +This software is distributed in the hope that it will be useful, but +\textbf{without any warranty}; without even the implied warranty of +\textbf{merchantability} or \textbf{fitness for a particular purpose}. +See the GNU General Public License for more details. + +You should have received a copy of the GNU General Public License in +the file |License| along with this package; if not, write to the Free +Software Foundation, Inc., 675~Mass Ave, Cambridge, MA~02139,~USA. + + +\sect Before we start we declare some shorthands for category codes. +By declaring the underscore~`(|_|)' as letter we can use it in our +macros. (I agree with \textsc{D.~Knuth} that +|\identifier_several_words_long| is more readable than +|\IdentifierSeveralWordsLong| and in every case better than |\p@@@s|.) + +As this is a \LaTeX{} style file the at sign is a letter anyhow; so +we can use the ``private'' \LaTeX{} macros; and with the underscore we +can make our own macros more readable. But as we have to restore this +category code at the end of this macro file we store its former value +in the control sequence |\CatUsCode|. This method is better than to +use a group because not all macros have to be defined global this way. + +Since somebody might use more styles from me, this cseqs might be +defined already. + +\beginprog +\ifx \CatEscape\undefined + \chardef\CatEscape=0 + \chardef\CatOpen=1 + \chardef\CatClose=2 + \chardef\CatIgnore=9 + \chardef\CatLetter=11 + \chardef\CatOther=12 + \chardef\CatActive=13 % is defined in Plain already + + \chardef\CatUsCode=\catcode`\_ +\fi + +\catcode`\_=\CatLetter % top level macro file +\endprog + + + + +\chap Typesetting \cweave{} output. + +The module \pkg{cwebbase} handles that. It demands some cseqs to be +defined before it may be imported. +% +\begin{fixme} + These cseqs should actually be implemented as package options. Have + to extract the \pkg{keyvalx} code from the \cls{cweb} class first. + Now I fix the usage I need for our Literate Programming book. +\end{fixme} + +\begin{fixme} + The options should not only be available as package options, but + also on a per environment base. We might to typeset one part with + flat and another one with hierarchic structure. Or we might to + suppress format directives only in some parts, while other parts + feature them. But this needs a change in \pkg{cwebbase}, there code + is \emph{defined} depending on the cseqs below. It must be evaluated + differently to support the stuff below. OTOH, this will cost + performance -- is it worth it? +\end{fixme} + +\beginprog +\def\cweb@structure{0 } % hierarchic structure +\let\if@cweb@suppress@changehints@=\iffalse % no suppression whatsoever +\let\if@cweb@suppress@unchanged@=\iffalse +\let\if@cweb@suppress@format@=\iffalse + +\input{cwebbase} +\endprog + + +\sect Now we can typeset pieces of \cweave{} output -- in principle. +If these pieces have sections in them, they'll write entries to the +|TOC| file. These entries will either mess up the structure of the +document's table of contents or lead to \TeX{} error messages. That's +because either toc entries for section types |part|, |chapter|, or +|section| are written (causing the conflict); or toc entries for +section types |chunk| and `|section|$i$' are written (being +undefined). + +The best way seems to stop the included source to add anything to the +table of contents or to do any mark processing. For that, we provide a +macro |\CwebNoLists| that will locally redefine the \LaTeX{} cseqs +that cause such actions to happen. That macro will be used at the +start of the environment. + +\beginprog +\def\CwebNoLists{% + \let\addtocontents\@gobbletwo + \let\mark\@gobble +%% \let\markboth\@gobbletwo +%% \let\markright\@gobble + } +\endprog + + +\sect Furthermore, section divisions should not start a new page. +% +\begin{fixme} + Currently, sections in the included parts are typeset exactly the + same way as they're typeset in the main document. That should be + different -- but how do I change it in a general way? +\end{fixme} + +\beginprog +\let\CwebEjectSection\@gobble +\endprog + + +\sect OK, what needs to be done for the environment? We need to save +and restore counters that are changed globally by section commands. + +\beginprog +\ifx \chapter\undefined + \let\cweb@save_c@chapter\@gobble + \let\c@chapter\@gobble +\else + \newcount\cweb@save_c@chapter +\fi +\newcount\cweb@save_c@part +\newcount\cweb@save_c@section + +\def\cweb@save_for_env{% + %% save values changed globally by cwebbase macros + \cweb@save_c@chapter\c@chapter + \cweb@save_c@part\c@part + \cweb@save_c@section\c@section + \let\cweb@save_thechapter\thechapter + \let\cweb@save_thesection\thesection + %% set up typesetting of CWEB source + \@nobreaktrue % don't add space from first chunk + } +\def\cweb@restore_env{% + \global\c@chapter\cweb@save_c@chapter + \global\c@part\cweb@save_c@part + \global\c@section\cweb@save_c@section + \global\let\thechapter\cweb@save_thechapter + \global\let\thesection\cweb@save_thesection + } +\endprog + + +\sect For the moment, we use list spacing around a |cweb| environment. +\LaTeX{} supplies the |trivlist| environment for such purposes. + +But the environment content itself must not be typeset in a list +context. Two problems exist: There must be at least one |\item|, it +will actually add the vertical space and must therefore not be +discarded. And list environments redefine |\par|, but the +\pkg{cwebbase} macros need a more `normal' definition. + +An item label is stored by |\item| in the |\everypar| register. The +label is only typeset if |@inlabel| is true. The flag |@newlist| tells +most other \LaTeX{} macros, that there was no item yet; typically +they'll add an empty paragraph if necessary. As the content will +start with a \LaTeX{} section division, or even with document markup, +we don't want that empty paragraph and should skip the empty item +label. (Often enough, due to the initialisation of |\everypar| in +|\@startsection|, it will be thrown away anyhow). So we establish its +postconditions, the next macro won't notice that the item hasn't been +set actually. + +\begin{fixme} + Experimentally, we reduce the skip in front of a chunk to 1\,pc, + otherwise the distance between two chunks is far too wide compared + to other vertical distances in the text around the included stuff. +\end{fixme} + +\beginprog +\def\CwebBeginEnvLayout{% + %% use spacing of list, but switch off list handling afterwards + \trivlist + \@inlabelfalse % don't go in horizontal mode + \item\relax + \global\@newlistfalse + \global\@inlabelfalse\the\everypar + \@setpar{\@@par}% + \CwebChunkPreSkip 1pc plus 6pt minus 3pt + } +\let\CwebEndEnvLayout=\endtrivlist +\endprog + + +\sect The environment gets trivial now. We supply a hook for a whole +document and also environment-specific ones (as an optional +environment argument). + +\beginprog +\newenvironment{cweb}[1][]% + {% + \CwebBeginEnvLayout + \cweb@save_for_env + \CwebNoLists + \csname CwebEnvHook\endcsname + #1 + }{% + \CwebEndEnvLayout + } +\endprog + + +\sect An a convenience tag to include a complete \cweave{} file. Since +we want to test the typesetting of our included \cweb{} program alone, +we allow but ignore most kinds of document markup. Many of them have +two arguments, one is optional. + +In the preambel, only |\documentclass| and |\usepackage| may appear. +Titlepage material is handled, table of contents is discarded. The +document environment does nothing, i.e., unlike the \cls{cweb} class, +material behind |\end{document}| is not ignored. + +An |\end| tag is inserted by \cweave{} and must be redefined, we don't +end an environment there. We'll open a group and will use that tag to +close it again, as we know that we'll have reached the end of our +\cweb{} file. Of course, |\end| must not be redefined during +processing the actual program source -- environments are valid there. +The redefinition is best be triggered by material inserted from +\cweave{} at the very end, e.g., from |\inx|. But |\inx| may not be +there (when \cweave{} is called with the `|-x|' option), we'll use +|\enddocument| to add a |\inx| tag for safety reason. If the user +didn't add an |\end{document}| to his file and used `|-x|', we're +going to loose. + +\TeX{}nical note: |\document| must not have the same binding as +|\relax|, as this would be interpreted as a non-existing environment +by \LaTeX{}. (That's due to the |\csname| lossage.) + +\TeX{}nical note: Evaluating |\inx| in |\enddocument| will not have +any effect, we're inside a group. That evaluation must be postponed +after the group has been closed. + +\beginprog +\newcommand\cweb@doc_markup_gobble[2][]{} +\newcommand*\cwebInput[1]{% + \begin{cweb} + [% + \begingroup + \let\documentclass\cweb@doc_markup_gobble + \let\usepackage\cweb@doc_markup_gobble + \let\title\@gobble + \let\author\@gobble + \let\date\@gobble + \let\maketitle\relax + \let\tableofcontents\relax + \let\document\empty + \def\inx{\let\end\endgroup}% + \let\fin\relax + \let\con\relax + \def\enddocument{\aftergroup\inx}% + ]% + \input{#1}% + \end{cweb}% + } +\endprog + + + + +\chap The end. + +Well, after all we're finished with this style. We must not forget to +restore the underscore catcode. + +\beginprog +\catcode`\_=\CatUsCode + +\endinput +\endprog + + +%\sect I would like to thank those who helped me to improve this style. +% In particular, XXX provided XXXsubstantial parts of the code. + + + +%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% + +\begin{rcslog} +$StyleLog: cwebparts.doc,v $ +\Revision 1.1 1995/09/16 17:01:59 schrod +Add support for inclusion of \cweb{} files in \LaTeX{} documents. + +\end{rcslog} + + + +\end{document} + + + +%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% +Local Variables: +mode: LaTeX +TeX-brace-indent-level: 4 +indent-tabs-mode: t +TeX-auto-untabify: nil +TeX-auto-regexp-list: LaTeX-auto-regexp-list +compile-command: "make cwebparts.sty" +End: diff --git a/Master/texmf-dist/doc/latex/cweb-latex/src/cwebx.sty b/Master/texmf-dist/doc/latex/cweb-latex/src/cwebx.sty new file mode 100644 index 00000000000..d20db59b171 --- /dev/null +++ b/Master/texmf-dist/doc/latex/cweb-latex/src/cwebx.sty @@ -0,0 +1,406 @@ +% $Id: cwebx.sty,v 3.1 1995/11/21 00:42:49 schrod Exp $ +%---------------------------------------------------------------------- + +% +% LaTeX2e package cwebx +% extra stuff for cweb bundle, to be added there sometimes +% +% [LaTeX2e] +% (history at end) + + +% Identification. The code below is explained in the implementation +% documentation of the rcs package. + +\begingroup + \def\RCSPackage#1#2 $#3: #4 #5\endRCS $#6: #7 #8\endRCS{% + \def\date{#4}\def\id{v#7}% + \ProvidesPackage{#1}[\date\space\id\space #2]% + } + + \RCSPackage{cwebx}{extra stuff for cweb bundle} + $Date: 1995/11/21 00:42:49 $: 9999/00/00 \endRCS + $Revision: 3.1 $: 0.0 \endRCS +\endgroup + + +% +% What's this package for? +% +% Additional macros for cweb.cls or cwebpart.sty, to be added there +% sometimes. +% + + +% requires that cweb base is loaded +\ifx \cwbb@loaded\undefined + \PackageError{cwebx}{cweb base is not loaded} + {This package requires the usage of either cweb.cls or + cwebparts.sty.} +\fi + + +% This style is supported. Send bug reports, comments and repairs. +% +% The reference version may be retrieved via anonymous ftp from +% |ftp.th-darmstadt.de| [130.83.47.112], directory |pub/tex/latex|. It's +% placed there as a gzipped tar file. (The information on the +% IP~number is dated 27 Sep 95. It might have changed, also +% this is very unlikely. Use your friendly nameserver.) + + + +\catcode`\_=\CatLetter + + + +%%% ====================================================================== + + +%% +%% Errors in cwebbase.doc +%% + + +% As \4 follows \B always, the \kern\CwebIndentUnit to neutralize that +% backspace must be made after run-in section, too. Between the chunk +% marker and the program code start is 1em then (inserted by LaTeX +% section start). + +% FIXME: \4 does _not_ follow \B all the time. Check when it is +% issued. The code below was introduced to repair a problem that gets +% visible if one uses \cwebNoChunk with an indentation unit of 2em, +% obviously I have to look for a better repair possibility. + +\CheckCommand*\B{% + \cwbb@program + \cwbb@indent 3\CwebIndentUnit \hangindent\cwbb@indent + \ifvmode + \if@noskipsec + \indent % add an empty documentation part + \else + \noindent \kern\CwebIndentUnit + \fi + \fi + \CwebNumberProgramPart + } + +\renewcommand\B{% + \cwbb@program + \cwbb@indent 3\CwebIndentUnit \hangindent\cwbb@indent + \ifvmode + \if@noskipsec + \indent % add an empty documentation part + \else + \noindent + \fi + \fi + \kern\CwebIndentUnit + \CwebNumberProgramPart + } + + +%%% ------------------------------------------------------------ + + +%% +%% Macros from cwebparts.doc, to be moved to cwebbase.doc. +%% + +% They are partly repaired and improved, where noted with `CHANGE'. + + +% \sect Now we can typeset pieces of \cweave{} output -- in principle. +% If these pieces have sections in them, they'll write entries to the +% |TOC| file. These entries will either mess up the structure of the +% document's table of contents or lead to \TeX{} error messages. That's +% because either toc entries for section types |part|, |chapter|, or +% |section| are written (causing the conflict); or toc entries for +% section types |chunk| and `|section|$i$' are written (being +% undefined). + +% The best way seems to stop the included source to add anything to the +% table of contents or to do any mark processing. For that, we provide a +% macro |\CwebNoLists| that will locally redefine the \LaTeX{} cseqs +% that cause such actions to happen. That macro will be used at the +% start of the environment. + +% CHANGE: Discard try code that redefines |\markboth| and |\markright|. + +% \beginprog +\def\CwebNoLists{% + \let\addtocontents\@gobbletwo + \let\mark\@gobble + } +% \endprog + + +% \sect Furthermore, section divisions should not start a new page. +% % +% \begin{fixme} +% Currently, sections in the included parts are typeset exactly the +% same way as they're typeset in the main document. That should be +% different -- but how do I change it in a general way? +% \end{fixme} +% +% CHANGE: Moved to Cweb environment. + + +% CHANGE: Improved question. +% +% \sect OK, what gets destroyed by the environment? We need to save +% and restore counters that are changed globally by section commands. + +% \beginprog +\ifx \chapter\undefined + \let\cweb@save_c@chapter\@gobble + \let\c@chapter\@gobble +\else + \newcount\cweb@save_c@chapter +\fi +\newcount\cweb@save_c@part +\newcount\cweb@save_c@section +\newcount\cweb@save_c@chunk + +\def\cweb@save_for_env{% + %% save values changed globally by cwebbase macros + \cweb@save_c@chapter\c@chapter + \cweb@save_c@part\c@part + \cweb@save_c@section\c@section + \cweb@save_c@chunk\c@chunk + \let\cweb@save_thechapter\thechapter + \let\cweb@save_thesection\thesection + %% set up typesetting of CWEB source + \@nobreaktrue % don't add space from first chunk + } +\def\cweb@restore_env{% + \global\c@chapter\cweb@save_c@chapter + \global\c@part\cweb@save_c@part + \global\c@section\cweb@save_c@section + \global\c@chunk\cweb@save_c@chunk + \global\let\thechapter\cweb@save_thechapter + \global\let\thesection\cweb@save_thesection + } +% \endprog + + +% \sect For the moment, we use list spacing around a |cweb| environment. +% \LaTeX{} supplies the |trivlist| environment for such purposes. + +% But the environment content itself must not be typeset in a list +% context. Two problems exist: There must be at least one |\item|, it +% will actually add the vertical space and must therefore not be +% discarded. And list environments redefine |\par|, but the +% \pkg{cwebbase} macros need a more `normal' definition. + +% An item label is stored by |\item| in the |\everypar| register. The +% label is only typeset if |@inlabel| is true. The flag |@newlist| tells +% most other \LaTeX{} macros, that there was no item yet; typically +% they'll add an empty paragraph if necessary. As the content will +% start with a \LaTeX{} section division, or even with document markup, +% we don't want that empty paragraph and should skip the empty item +% label. (Often enough, due to the initialisation of |\everypar| in +% |\@startsection|, it will be thrown away anyhow). So we establish its +% postconditions, the next macro won't notice that the item hasn't been +% set actually. + +% CHANGE: Added \CwebNoLists, as I consider it part of the env layout. + +% CHANGE: Added paragraph below. +% +% Furthermore, section divisions should not start a new page. +% % +% \begin{fixme} +% Currently, sections in the included parts are typeset exactly the +% same way as they're typeset in the main document. That should be +% different -- but how do I change it in a general way? +% \end{fixme} + +% \begin{fixme} +% Experimentally, we reduce the skip in front of a chunk to 1\,pc, +% otherwise the distance between two chunks is far too wide compared +% to other vertical distances in the text around the included stuff. +% \end{fixme} + +% \beginprog +\def\CwebBeginEnvLayout{% + %% use spacing of list, but switch off list handling afterwards + \trivlist + \@inlabelfalse % don't go in horizontal mode + \item\relax + \global\@newlistfalse + \global\@inlabelfalse\the\everypar + \@setpar{\@@par}% + %% now change cweb base layout, meant for classes + \CwebNoLists + \let\CwebEjectSection\@gobble + \CwebChunkPreSkip 1pc plus 6pt minus 3pt + } +\let\CwebEndEnvLayout=\endtrivlist +% \endprog + + +% \sect The environment gets trivial now. We supply a hook for a whole +% document and also environment-specific ones (as an optional +% environment argument). + +% CHANGE: Renamed to `Cweb'. Lusers should actually not call that env +% directly, it's too dangerous. + +% CHANGE: Added forgotten evaluation of \cweb@restore_env. + +% \beginprog +\newenvironment{Cweb}[1][]% + {% + \CwebBeginEnvLayout + \cweb@save_for_env + \csname CwebEnvHook\endcsname + #1 + }{% + \cweb@restore_env + \CwebEndEnvLayout + } +% \endprog + + +% CHANGE: Have to improve the introduction of |cwebInput|. Now it has +% an additional optional parameter, where instance specific cseqs may +% be passed (like the cweb env). + +% \sect An a convenience tag to include a complete \cweave{} file. Since +% we want to test the typesetting of our included \cweb{} program alone, +% we allow but ignore most kinds of document markup. Many of them have +% two arguments, one is optional. + +% In the preambel, only |\documentclass| and |\usepackage| may appear. +% Titlepage material is handled, table of contents is discarded. The +% document environment does nothing, i.e., unlike the \cls{cweb} class, +% material behind |\end{document}| is not ignored. + +% An |\end| tag is inserted by \cweave{} and must be redefined, we don't +% end an environment there. We'll open a group and will use that tag to +% close it again, as we know that we'll have reached the end of our +% \cweb{} file. Of course, |\end| must not be redefined during +% processing the actual program source -- environments are valid there. +% The redefinition is best be triggered by material inserted from +% \cweave{} at the very end, e.g., from |\inx|. But |\inx| may not be +% there (when \cweave{} is called with the `|-x|' option), we'll use +% |\enddocument| to add a |\inx| tag for safety reason. If the user +% didn't add an |\end{document}| to his file and used `|-x|', we're +% going to loose. + +% CHANGE: Added handling of empty chunk. (@noskipsec was set after the +% environment, what messed up later invocations of program code if +% used with \cwebCodeExample.) + +% \TeX{}nical note: |\document| must not have the same binding as +% |\relax|, as this would be interpreted as a non-existing environment +% by \LaTeX{}. (That's due to the |\csname| lossage.) + +% \TeX{}nical note: Evaluating |\inx| in |\enddocument| will not have +% any effect, we're inside a group. That evaluation must be postponed +% after the group has been closed. + +% \beginprog +\providecommand\cweb@doc_markup_gobble[2][]{} +\let\cwebInput\undefined +\newcommand*\cwebInput[2][]{% + \begin{Cweb} + [% + \begingroup + \let\documentclass\cweb@doc_markup_gobble + \let\usepackage\cweb@doc_markup_gobble + \let\title\@gobble + \let\author\@gobble + \let\date\@gobble + \let\maketitle\relax + \let\tableofcontents\relax + \let\document\empty + \def\inx{\let\end\endgroup}% + \let\fin\relax + \let\con\relax + \def\enddocument{% + \if@noskipsec % empty chunk + \global\@noskipsecfalse + \global\everypar{}% + \fi + \aftergroup\inx + }% + #1% + ]% + \input{#2}% + \end{Cweb}% + } +% \endprog + + +%%% ------------------------------------------------------------ + + +%% +%% New macros for cwebbase.doc. +%% + + +% For production of code examples in CWEB documents. It's expected +% that the document consists of only \emph{one} chunk that has no +% documentation. Just the code is of interest. Neither chunk markers, +% nor chunk numbers, nor refinement numbers are typeset. +% +% But one may change that with the optional parameter. + +\providecommand*\cwebCodeExample[2][]{% + \cwebInput[% + \def\CwebChunkHeading##1{\kern -1em}% + \let\cwbb@marginal_chunkno\relax + \let\CwebRefNumber\@gobble + \CwebChunkPreSkip\z@skip + #1% + ]{#2}% + } + + +% Often one has an explanation for more than one refinement. (E.g., +% when there's needed a definition and a declaration.) Then the +% documentation of the follow-up refinements shall consist solely of +% |\cwebNoChunk|. Then no chunk marker is set, and the refinement +% is typeset nearer to the previous refinement. Optically, we create +% one chunk with many refinements this way. + +% \TeX{}nical note: No chunk marker is created by redefining macro +% |\@svsechd| that creates header for run-in \LaTeX{} section. Every +% header is followed by 1\,em space, we have to annulate that skip. +% (The refinement shall start at the beginning of the line.) One must +% not use |\hskip| to backup, an |\unskip| comes after the header and +% would discard that skip. + +\providecommand\cwebNoChunk{% + \vskip -\lastskip \bigskip % discard LaTeX's inter-section skip + \def\@svsechd{\kern -1em}% + } + + + + +\catcode`\_=\CatUsCode +\endinput + + +%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% +% +% $Log: cwebx.sty,v $ +% Revision 3.1 1995/11/21 00:42:49 schrod +% version 1.1 +% + + + +%%%------------------------------------------------------------ + +% Local Variables: +% mode: LaTeX +% TeX-brace-indent-level: 4 +% indent-tabs-mode: t +% TeX-auto-untabify: nil +% page-delimiter: "^%%%" +% End: diff --git a/Master/texmf-dist/doc/latex/cweb-latex/src/keyvald.dtx b/Master/texmf-dist/doc/latex/cweb-latex/src/keyvald.dtx new file mode 100644 index 00000000000..6873fde7110 --- /dev/null +++ b/Master/texmf-dist/doc/latex/cweb-latex/src/keyvald.dtx @@ -0,0 +1,347 @@ +%% $ITI: keyvald.dtx,v 1.2 1995/08/25 19:00:31 schrod Exp $ +%%------------------------------------------------------------ +%% (history at end) +% +% \iffalse +%% File: keyvald.dtx; changed version of +%% File: keyval.dtx Copyright (C) 1993-1994 David Carlisle +% +%<*dtx> + \ProvidesFile{keyvald.dtx} +%</dtx> +%<*!plain> +%<package&!plain>\NeedsTeXFormat{LaTeX2e} +%<package&!plain>\ProvidesPackage{keyvald} +%<driver> \ProvidesFile{keyvald.drv} +% \fi +% \ProvidesFile{keyvald.dtx} + [1994/08/23 v1.08.1 key=value parser (DPC) w/ defaults (js)] +% +% \iffalse +%</!plain> +%<*driver> +\documentclass{ltxdoc} +\usepackage{keyvald} +\begin{document} + \DocInput{keyvald.dtx} +\end{document} +%</driver> +% \fi +% +% \GetFileInfo{keyvald.dtx} +% \title{The \textsf{keyvald} package\thanks{This file +% has version number \fileversion, last +% revised \filedate.}} +% \author{David Carlisle\\carlisle@cs.man.ac.uk +% \and +% changed by\\Joachim Schrod\\ +% \texttt{<schrod@iti.informatik.th-darmstadt.de>} +% } +% \date{\filedate} +% \maketitle +% +% \begin{abstract} +% A \LaTeX\ package implementing a system allowing the setting +% of parameters (or `named arguments' with a +% \meta{key}${}={}$\meta{value} syntax. +% +% Eg: |\foo[height=3in, shadow = true ]{bar}| +% \end{abstract} +% +% \CheckSum{110} +% +% \changes{v1.03} +% {1993/10/13}{Initial version} +% \changes{v1.04}{1993/11/15} +% {Upgrade after comments from Timothy van Zandt} +% \changes{v1.05}{1993/11/17} +% {Further small improvements} +% \changes{v1.06}{1994/02/01} +% {Update to LaTeX2e} +% \changes{v1.07}{1994/03/15} +% {New style ltxdoc} +% \changes{v1.08}{1994/09/12} +% {Improve docstrip handling} +% \changes{v1.08.1}{1995/08/23} +% {Default keys} +% +% +% This package implements a system of defining and using sets of +% parameters, which are set using the syntax \meta{key}=\meta{value}. +% +% For each keyword in such a set, there exists a function which is +% called whenever the parameter appears in a parameter list. For +% instance if the set |dpc| is to have the keyword |scale| then I +% would define.\\ +% | \define@key{dpc}{scale}{scale ({\tt\string#1})\\}|\\ +% The first argument of |\define@key| is the set of keywords being +% used, the second is the keyword, and the third is the function to +% call. This function will be given as |#1| the \meta{value} specified +% by the user. +% +% Normally it is an error to omit the `=\meta{value}' however if an +% optional \meta{value} is supplied when the keyword is defined, then +% just the keyword need be supplied.\\ +% |\define@key{dpc}{clip}[true]{...}|\\ +% For `|clip|' you can go `|clip = true|' or `|clip = false|' or +% just `|clip|', which is the same as `|clip = true|' +% +% To use these keywords, just call `|\setkeys|' with a comma +% separated list of settings, each of the form +% \meta{key}=\meta{value}, or just \meta{key}. Any white space around +% the `|=|' and `|,|' is ignored. +% +% As the \meta{key} is passed as a macro argument, if it consists +% entirely of a |{ }| group, the outer braces are stripped off. Thus +% |,key=foo,| and |key=,{foo},| are equivalent. This fact enables one to +% `hide' any commas or equals signs that must appear in the value. i.e.\ +% in |foo={1,2,3},bar=4|, |foo| gets the value |1,2,3|, the comma after +% |1| does not terminate the keyval pair, as it is `hidden' by the +% braces. +% +% Empty entries, with nothing between the commas, are silently ignored. +% This means that it is not an error to have a comma after the last +% term, or before the first. +% +% \section{Example} +% +% We may extend the examples above to give a `fake' graphics +% inclusion macro, with a syntax similar to that used in the psfig +% macros. +% +% \makeatletter +% \def\dpcgraphics{\@ifnextchar[\@dpcgraphics{\@dpcgraphics[]}} +% +% |\dpcgraphics| has one optional argument which is passed through +% |\setkeys|, and one mandatory argument, the filename. It actually +% just typesets its arguments, for demonstration. +% +% \def\@dpcgraphics[#1]#2{{\setkeys{dpc}{#1}INPUT: #2}}% +% +% \define@key{dpc}{scale}{scale ({\tt\string#1\relax})\\} +% \define@key{dpc}{height}{height ({\tt#1})\\} +% \define@key{dpc}{width}{width ({\tt#1})\\} +% \define@key{dpc}{bb}{bounding box ({\tt#1})\\} +% \define@key{dpc}{clip}[true]{clip ({\tt\string#1\relax})\\} +% \makeatother +% +% The declared keys are: |scale|, |height|, |width|, |bb|, +% and |clip|. Except for the last, they must all be given a value if +% used. +% +% Note how in the following, any white space arround |=| or |,| is +% ignored, as are the `empty' arguments caused by extra commas. Note +% also that each macro receives {\em exactly\/} the tokens that you +% specify as arguments, no premature expansion is done. +% +% \begin{verbatim} +% \def\dpcgraphics{\@ifnextchar[\@dpcgraphics{\@dpcgraphics[]}} +% \def\@dpcgraphics[#1]#2{{\setkeys{dpc}{#1}INPUT: #2}} +% +% \define@key{dpc}{scale}{scale ({\tt\string#1\relax})\\} +% \define@key{dpc}{clip}[true]{clip ({\tt\string#1\relax})\\} +% \end{verbatim} +% +% \begin{minipage}{.4\textwidth} +% \begin{verbatim} +% \def\scalemacro{9} +% \dpcgraphics +% [ height =4in, , +% width = 3in, +% scale = \scalemacro, +% bb = 20 20 300 400 , +% clip, +% ]{aaa} +% \end{verbatim} +% \end{minipage} +% \hfill +% \begin{minipage}{.4\textwidth} +% \def\scalemacro{9} +%\dpcgraphics +% [ height =4in, , +% width = 3in, +% scale = \scalemacro , +% bb = 20 20 300 400 , +% clip , +% ]{aaa} +% \end{minipage} +% +% +% \section{The Internal Interface} +% A declaration of the form:\\ +% |\define@key{family}{key}{...}|\\ +% Defines a macro |\KV@prefix@key| with one argument. When used in a +% keyval list, the macro receives the value as its argument. +% +% A declaration of the form:\\ +% |\define@key{family}{key}[default]{...}|\\ +% Defines a macro |\KV@family@key| as above, however it also defines the +% macro |\KV@family@key@default| as a macro with no arguments, and +% definition\\ +% |\KV@family@key{default}|. +% +% Thus if macros are defined using |\define@key|, the use of a key with +% no value \ldots|,foo,|\ldots\ is always equivalent to the use of the +% key with some value, \ldots|,foo=default,|\ldots. However a package +% writer may wish that the `default' behaviour for some key is not +% directly equivalent to using that key with a value. (In particular, as +% pointed out to me by Timothy Van Zandt, you may wish to omit error +% checking on the default value as you know it is correct.) In these +% cases one simply needs to define the two macros +% |\KV@|\meta{family}|@key| and |\KV@|\meta{family}|@key@default| +% directly using |\def| (or |\newcommand|). I do not supply a user +% interface for this type of definition, but it is supported in the +% sense that I will try to ensure that any future upgrades of this +% package do not break styles making use of these `low level' +% definitions. +% +% \StopEventually{} +% +% \section{The Macros} +% +% From version~1.05, all `internal' macros associated to keys have names +% of the form:\\ +% |\KV@|\meta{family}|@|\meta{key} or +% |\KV@|\meta{family}|@|\meta{key}|@|\meta{default} +% +% \begin{macrocode} +%<*package> +% \end{macrocode} +% +% \begin{macro}{\setkeys} +% The top level macro. |#2| should be a comma separated values of the +% form \meta{key} |=| \meta{value} or just simply \meta{key}. +% The macro associated with this key in the `family' |#1| is called with +% argument \meta{value}. The second form is only allowed if the key was +% declared with a default value. +% \begin{macrocode} +\def\setkeys#1#2{% +% \end{macrocode} +% Save the `family' for later. Then begin acting on the comma +% separated list. +% \begin{macrocode} + \def\KV@prefix{KV@#1@}% + \KV@do#2,\relax,} +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\KV@do} +% Iterate down the list of comma separated argument pairs. +% \begin{macrocode} +\def\KV@do#1,{% + \ifx\relax#1\empty\else + \KV@split#1==\relax + \expandafter\KV@do\fi} +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\KV@split} +% Split up the keyword and value, and call the appropriate command. +% This macro was slightly reorganised for version 1.04, after some +% suggestions from Timothy Van Zandt. +% \begin{macrocode} +\def\KV@split#1=#2=#3\relax{% + \KV@@sp@def\KV@key{#1}% + \ifx\KV@key\@empty\else + \expandafter\let\expandafter\@tempc + \csname\KV@prefix\KV@key\endcsname + \ifx \@tempc\relax + \expandafter\let \expandafter\@tempc + \csname\KV@prefix*\endcsname + \def\@tempa{*}% + \else + \let\@tempa\KV@key + \fi + \ifx\@tempc\relax + \KV@err{\KV@key\space undefined}% + \else + \ifx\@empty#3\@empty + \KV@default + \else + \KV@@sp@def\@tempb{#2}% + \expandafter\@tempc\expandafter{\@tempb}\relax + \fi + \fi + \fi} +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\KV@default} +% Run the default code, or raise an error. +% \begin{macrocode} +\def\KV@default{% + \expandafter\let\expandafter\@tempb + \csname\KV@prefix\@tempa @default\endcsname + \ifx\@tempb\relax + \KV@err{No value specified for \KV@key}% + \else + \@tempb\relax + \fi} +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\KV@err} +% Error messages. +% \begin{macrocode} +\def\KV@err#1{\PackageError{keyvald}{#1}{}} +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\KV@@sp@def} +% \begin{macro}{\KV@@sp@b} +% \begin{macro}{\KV@@sp@c} +% |\KV@@sp@def|\meta{cmd}\meta{token list} is like |\def|, except that +% a space token at the beginning or end of \meta{token list} is +% removed before making the assignment. \meta{token list} may not +% contain the token |\@nil|, unless it is within a brace group. +% The names of these commands were changed at version~1.05 to ensure +% that they do not clash with `internal' macros in a key family `sp'. +% +% The following three lines are equivalent to\\ +% |\def\KV@@sp@def#1#2{\KV@@sp@b#2\@nil\@nil~\@nil\relax#1}|\\ +% Where |~| is an explicit space token. +% \begin{macrocode} +\def\@tempa#1{% +\def\KV@@sp@def##1##2{\KV@@sp@b##2\@nil\@nil#1\@nil\relax##1}} +\@tempa{ } +\def\KV@@sp@b#1#2 \@nil{\KV@@sp@c#1#2} +\def\KV@@sp@c#1\@nil#2\relax#3{\def#3{#1}} +% \end{macrocode} +% \end{macro} +% \end{macro} +% \end{macro} +% +% \begin{macro}{\define@key} +% Define the command associated to the key |#2| in the family |#1|. +% First looks for a default argument (the default value for the +% key) +% \begin{macrocode} +\def\define@key#1#2{% + \@ifnextchar[{\KV@def{#1}{#2}}{\@namedef{KV@#1@#2}####1}} +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\KV@def} +% Make the definitions of the command, and the default value. +% \begin{macrocode} +\def\KV@def#1#2[#3]{% + \@namedef{KV@#1@#2@default\expandafter}\expandafter + {\csname KV@#1@#2\endcsname{#3}}% + \@namedef{KV@#1@#2}##1} +% \end{macrocode} +% \end{macro} +% +% \begin{macrocode} +%</package> +% \end{macrocode} +% \Finale +% +\endinput + +%------------------------------------------------------------ +% +% $ITIlog: keyvald.dtx,v $ +% Revision 1.2 1995/08/25 19:00:31 schrod +% Add enhanced version of keyval package, with ability to specify +% default action for unknown keyword options. +% diff --git a/Master/texmf-dist/doc/latex/cweb-latex/src/keyvald.ins b/Master/texmf-dist/doc/latex/cweb-latex/src/keyvald.ins new file mode 100644 index 00000000000..0fe873fd74f --- /dev/null +++ b/Master/texmf-dist/doc/latex/cweb-latex/src/keyvald.ins @@ -0,0 +1,6 @@ +\def\batchfile{keyvald.ins} +\input docstrip + +\keepsilent + +\generateFile{keyvald.sty}{f}{\from{keyvald.dtx}{package}} diff --git a/Master/texmf-dist/doc/latex/cweb-latex/src/style/cweb-doc.el b/Master/texmf-dist/doc/latex/cweb-latex/src/style/cweb-doc.el new file mode 100644 index 00000000000..eb801786499 --- /dev/null +++ b/Master/texmf-dist/doc/latex/cweb-latex/src/style/cweb-doc.el @@ -0,0 +1,22 @@ +;; $Id: cweb-doc.el,v 1.1 1995/08/29 15:15:30 schrod Exp $ +;;---------------------------------------------------------------------- + +;;; +;;; AUC-TeX configuration for cweb-doc.sty +;;; + +(TeX-add-style-hook "cweb-doc" + (function + (lambda () + (TeX-add-symbols "cweb" "ctangle" "cweave" + '("arg" 1) + '("<" TeX-arg-free ">") + '("cls" 1) + '("pkg" 1) + ) + (LaTeX-add-environments "fixme" "cseqtab" + '("options" LaTeX-env-item) + ) + (setq LaTeX-item-list (cons '("options" . LaTeX-item-argument) + LaTeX-item-list)) + ))) diff --git a/Master/texmf-dist/doc/latex/cweb-latex/src/test/Imakefile b/Master/texmf-dist/doc/latex/cweb-latex/src/test/Imakefile new file mode 100644 index 00000000000..0d32d8301c3 --- /dev/null +++ b/Master/texmf-dist/doc/latex/cweb-latex/src/test/Imakefile @@ -0,0 +1,142 @@ +XCOMM $Id: Imakefile,v 1.9 1995/11/20 22:34:44 schrod Exp $ +/* ---------------------------------------------------------------------- */ + +/* + * Imakefile for test of cweb bundle + * + * (history at end) + */ + + +/* needed modules */ +#define idx NO_IDX_CLEAN /* don't delete *.idx on make clean */ +#include <TeX.imk> +#undef idx + +XCOMM ------------------------------------------------------------ +XCOMM continue in Imakefile + +.SUFFIXES: .tex .w + +.w.tex: @@\ + if [ -r $*.ch ] ; \ @@\ + then cweave $* $* ; \ @@\ + else cweave $* ; \ @@\ + fi + + +regression:: clean + $(MAKE) all + + +/* ============================================================ + * + * tests for cweb class & cwebarray package + */ + +TeXTarget(latex,minimal) /* minimal document */ +TeXTarget(latex,token) /* rendering of tokens */ +TeXTarget(latex,refname) /* refinement names, filename ref. */ +TeXTarget(latex,modes) /* cweb FSA */ +TeXTarget(latex,section) /* hierarchic document structure */ +TeXTarget(latex,flat) /* flat document structure */ +TeXTarget(latex,report) /* report as base class */ +TeXTarget(latex,nolists) /* nolists via CWEAVE option -x */ +TeXTarget(latex,newif) /* idiosyncrasy of old LateX version */ + +TeXTarget(latex,enddocbegin) /* skip stuff after \end{document} */ +TeXTarget(latex,vbar) /* ruled table [cwebarray] */ +TeXTarget(latex,titlepage) /* titlepage option */ + +TeXTarget(latex,language-german) /* language option german */ +TeXTarget(latex,german) /* babel-compatible option german */ + +TeXTarget(latex,change) /* changefile-related output */ +TeXTarget(latex,sup-changes) /* changefile-related suppression */ + +TeXTarget(latex,sup-lists) /* suppression of index & reflist */ +TeXTarget(latex,sup-format) /* suppression of format directives */ + +TeXTarget(latex '\nonstopmode \input',badend) /* unknown \end expansion */ +TeXTarget(latex '\nonstopmode \input',badopts) /* unknown & illegal options */ + + +/* + * Special (additional) rules for targets above + */ + +nolists.tex: nolists.w + cweave -x nolists + +change.tex: change.ch + +sup-changes.tex: sup-changes.w change.ch + cweave sup-changes change + + + +/* ============================================================ + * + * tests for packages + */ + +/* cwebparts */ + +TeXTarget(latex,parts) /* CWEB parts in LaTeX document */ + +parts.dvi: parts-code.tex + +parts-code.tex: parts-code.w + cweave -x parts-code + + + +/* ============================================================ + * + * Tests for stuff that is not strictly part of the cweb bundle. + */ + +TeXTarget(latex,keyvald) /* keyword-value options w/ defaults */ + + +realclean:: clean + rm -f *.idx *.scn *.tex + + + +/* ============================================================ + * + * $Log: Imakefile,v $ + * Revision 1.9 1995/11/20 22:34:44 schrod + * Test language support, actually for German. + * + * Revision 1.8 1995/11/07 18:28:52 schrod + * Add test newif, for old LaTeX version. + * + * Revision 1.7 1995/09/17 14:21:36 schrod + * Test package cwbl-german. + * + * Revision 1.6 1995/09/16 17:02:09 schrod + * Test cwebparts.sty. + * + * Revision 1.5 1995/09/12 22:32:21 schrod + * Add test case for output of CWEAVE with `-x' option. + * + * Revision 1.4 1995/08/29 15:22:14 schrod + * Add test of format directive suppression. + * + * Revision 1.3 1995/08/27 13:29:38 schrod + * Tests of options & changefile-related suppression. + * + * Revision 1.2 1995/08/25 19:12:39 schrod + * Test implementation of hierarchic and flat structure. + * + * Revision 1.1 1995/08/08 00:14:46 schrod + * Updated to \LaTeXe{}, the |cweb| style is now a document class. + * + + * ------------------------------------------------------------ + * Local Variables: + * mode: Indented-Text + * End: + */ diff --git a/Master/texmf-dist/doc/latex/cweb-latex/src/test/badend.w b/Master/texmf-dist/doc/latex/cweb-latex/src/test/badend.w new file mode 100644 index 00000000000..51060a65cfd --- /dev/null +++ b/Master/texmf-dist/doc/latex/cweb-latex/src/test/badend.w @@ -0,0 +1,21 @@ +% $Id: badend.w,v 1.3 1995/08/08 00:14:46 schrod Exp $ +%---------------------------------------------------------------------- + +% tests unknown \end macro +% assumes "h" and "x" user reaction +% note: produces no output! + +\makeatletter +\def\end#1{\csname end#1\endcsname\@@checkend{#1}% + \expandafter\endgroup \if@@endpe \@@doendpe \fi +\relax % <-- added + \if@@ignore \global\@@ignorefalse \ignorespaces\fi} +\makeatother + +\documentclass{cweb} + +\begin{document} + +@ Test + +\end{document} diff --git a/Master/texmf-dist/doc/latex/cweb-latex/src/test/badopts.w b/Master/texmf-dist/doc/latex/cweb-latex/src/test/badopts.w new file mode 100644 index 00000000000..5fff5f7add7 --- /dev/null +++ b/Master/texmf-dist/doc/latex/cweb-latex/src/test/badopts.w @@ -0,0 +1,13 @@ +% $Id: badopts.w,v 1.1 1995/08/27 13:29:39 schrod Exp $ +%------------------------------------------------------------ + +% test bad options + +\documentclass[undef,structure,structure=chaos]{cweb} + +\begin{document} + +Text. + +@ +\end{document} diff --git a/Master/texmf-dist/doc/latex/cweb-latex/src/test/change.ch b/Master/texmf-dist/doc/latex/cweb-latex/src/test/change.ch new file mode 100644 index 00000000000..2796a280e68 --- /dev/null +++ b/Master/texmf-dist/doc/latex/cweb-latex/src/test/change.ch @@ -0,0 +1,5 @@ +@x +change here +@y +did the change +@z diff --git a/Master/texmf-dist/doc/latex/cweb-latex/src/test/change.w b/Master/texmf-dist/doc/latex/cweb-latex/src/test/change.w new file mode 100644 index 00000000000..d787065b933 --- /dev/null +++ b/Master/texmf-dist/doc/latex/cweb-latex/src/test/change.w @@ -0,0 +1,20 @@ +% $Id: change.w,v 1.4 1995/08/27 13:29:40 schrod Exp $ +%---------------------------------------------------------------------- + +% tests changefile output + +\documentclass{cweb} + +\begin{document} + +\title{Test for changefiles} +\author{Coogar} +\maketitle + +@ test + +@ +change here + +@ +\end{document} diff --git a/Master/texmf-dist/doc/latex/cweb-latex/src/test/enddocbegin.w b/Master/texmf-dist/doc/latex/cweb-latex/src/test/enddocbegin.w new file mode 100644 index 00000000000..64d77368e1f --- /dev/null +++ b/Master/texmf-dist/doc/latex/cweb-latex/src/test/enddocbegin.w @@ -0,0 +1,16 @@ +% $Id: enddocbegin.w,v 1.3 1995/08/08 00:14:47 schrod Exp $ +%---------------------------------------------------------------------- + +% tests skip of stuff after \end{document} + +\documentclass{cweb} + +\begin{document} + +@ Test. + +\end{document} + +\begin{itemize} +\item an item +\end{itemize} diff --git a/Master/texmf-dist/doc/latex/cweb-latex/src/test/flat.w b/Master/texmf-dist/doc/latex/cweb-latex/src/test/flat.w new file mode 100644 index 00000000000..993193971f3 --- /dev/null +++ b/Master/texmf-dist/doc/latex/cweb-latex/src/test/flat.w @@ -0,0 +1,47 @@ +% $Id: flat.w,v 1.1 1995/08/25 19:12:40 schrod Exp $ +%---------------------------------------------------------------------- + +% tests section levels and references for flat structure documents. + +\documentclass[structure=flat]{cweb} + +\begin{document} + +\tableofcontents + +\newpage + +@** s title. +\label{main-sec} + +Some text. + +@* e title. + +Some text. +And some references: Part~\ref{main-sec} and section~\ref{normal-sec}. + +@*0 0 title. + +Some text. + +@*1 1 title. + +Some text. + +@*2 2 title. + +Some text. + +@ \label{normal-sec} +Doc. + +@d def foo +@f def int + +@c +for ( int i=1; i<10; i++ ) + @< do something @> + +@ +\end{document} diff --git a/Master/texmf-dist/doc/latex/cweb-latex/src/test/german.w b/Master/texmf-dist/doc/latex/cweb-latex/src/test/german.w new file mode 100644 index 00000000000..dd6f5d64bbe --- /dev/null +++ b/Master/texmf-dist/doc/latex/cweb-latex/src/test/german.w @@ -0,0 +1,33 @@ +% $Id: german.w,v 1.3 1995/11/20 22:34:44 schrod Exp $ +%---------------------------------------------------------------------- + +% tests german language support + + +% check if option after german works as well +\documentclass[german,suppress=index]{cweb} + +\usepackage{german} + +\begin{document} + +\title{Test der \texttt{german} Option} +\author{Puma} +\maketitle + +@ Aktives Anf\"uhrungszeichen im String. +@c +#include "foo.h" + +@ @( foo.h@>= +@< etwas |code| @>@; +@< mehr Code @> + +@ @< etwas ...@>= +for ( int i=1; i<10; i++ ) + @< mehr... @> + +@ @< mehr... @>= + +@ +\end{document} diff --git a/Master/texmf-dist/doc/latex/cweb-latex/src/test/language-german.ch b/Master/texmf-dist/doc/latex/cweb-latex/src/test/language-german.ch new file mode 100644 index 00000000000..1e41e64e1e3 --- /dev/null +++ b/Master/texmf-dist/doc/latex/cweb-latex/src/test/language-german.ch @@ -0,0 +1,5 @@ +@x +@< mehr Code @> +@y +@< mehr -- ge\"anderter -- Code @> +@z diff --git a/Master/texmf-dist/doc/latex/cweb-latex/src/test/language-german.w b/Master/texmf-dist/doc/latex/cweb-latex/src/test/language-german.w new file mode 100644 index 00000000000..bd145c4f921 --- /dev/null +++ b/Master/texmf-dist/doc/latex/cweb-latex/src/test/language-german.w @@ -0,0 +1,31 @@ +% $Id: language-german.w,v 1.1 1995/11/20 22:34:45 schrod Exp $ +%---------------------------------------------------------------------- + +% tests german language support + +\documentclass[language=german]{cweb} + +\usepackage{german} + +\begin{document} + +\title{Test der \texttt{language} Option} +\author{Puma} +\maketitle + +@ Aktives Anf\"uhrungszeichen im String. +@c +#include "foo.h" + +@ @( foo.h@>= +@< etwas |code| @>@; +@< mehr Code @> + +@ @< etwas ...@>= +for ( int i=1; i<10; i++ ) + @< mehr... @> + +@ @< mehr... @>= + +@ +\end{document} diff --git a/Master/texmf-dist/doc/latex/cweb-latex/src/test/minimal.w b/Master/texmf-dist/doc/latex/cweb-latex/src/test/minimal.w new file mode 100644 index 00000000000..fc0cdaa492a --- /dev/null +++ b/Master/texmf-dist/doc/latex/cweb-latex/src/test/minimal.w @@ -0,0 +1,12 @@ +% $Id: minimal.w,v 1.4 1995/08/25 19:12:41 schrod Exp $ +%---------------------------------------------------------------------- + +% tests minimal CWEB w/ LaTeX input file + +\documentclass{cweb} +\begin{document} + +Test. + +@ +\end{document} diff --git a/Master/texmf-dist/doc/latex/cweb-latex/src/test/modes.w b/Master/texmf-dist/doc/latex/cweb-latex/src/test/modes.w new file mode 100644 index 00000000000..48525ae38d5 --- /dev/null +++ b/Master/texmf-dist/doc/latex/cweb-latex/src/test/modes.w @@ -0,0 +1,16 @@ +% $Id: modes.w,v 1.3 1995/08/27 19:28:49 schrod Exp $ +%---------------------------------------------------------------------- + +\documentclass{cweb} + +\begin{document} + +@ program $\to$ program with documentation of |bar(i+1)|. + +@d foo bar +@< refinement @>= + stmt; + + +@ +\end{document} diff --git a/Master/texmf-dist/doc/latex/cweb-latex/src/test/newif.w b/Master/texmf-dist/doc/latex/cweb-latex/src/test/newif.w new file mode 100644 index 00000000000..53ae2775145 --- /dev/null +++ b/Master/texmf-dist/doc/latex/cweb-latex/src/test/newif.w @@ -0,0 +1,39 @@ +% $Id: newif.w,v 1.1 1995/11/07 18:28:52 schrod Exp $ +%------------------------------------------------------------ + +% test that proper definition of \newif is established. + + +% First, restore problematic definition and pretend we're an old LaTeX. + +\makeatletter + +\outer\def\newif#1{\count@@\escapechar \escapechar\m@@ne + \expandafter\expandafter\expandafter + \edef\@@if#1{true}{\let\noexpand#1=\noexpand\iftrue}% + \expandafter\expandafter\expandafter + \edef\@@if#1{false}{\let\noexpand#1=\noexpand\iffalse}% + \@@if#1{false}\escapechar\count@@} % the condition starts out false + +\let\OldNewif=\newif + +\edef\fmtversion{1994/12/01} + +\makeatother + + +% Now, start the document and check if \newif got redefined. + +\documentclass{cweb} + +\begin{document} + +\ifx \OldNewif\newif + \errmessage{\string\newif\space was not redefined.} +\else + \message{\string\newif\space got redefined.} +\fi + +@ Test. + +\end{document} diff --git a/Master/texmf-dist/doc/latex/cweb-latex/src/test/nolists.w b/Master/texmf-dist/doc/latex/cweb-latex/src/test/nolists.w new file mode 100644 index 00000000000..190c22c3c21 --- /dev/null +++ b/Master/texmf-dist/doc/latex/cweb-latex/src/test/nolists.w @@ -0,0 +1,16 @@ +% $Id: nolists.w,v 1.1 1995/09/12 22:32:22 schrod Exp $ +%------------------------------------------------------------ + +% test suppression of index & ref list + +\documentclass{cweb} + +\begin{document} + +@ We need something to add to the lists. + +@< refinement @>= +foo(bar); + +@ +\end{document} diff --git a/Master/texmf-dist/doc/latex/cweb-latex/src/test/parts-code.w b/Master/texmf-dist/doc/latex/cweb-latex/src/test/parts-code.w new file mode 100644 index 00000000000..ad24b1e2b29 --- /dev/null +++ b/Master/texmf-dist/doc/latex/cweb-latex/src/test/parts-code.w @@ -0,0 +1,32 @@ +% $Id: parts-code.w,v 1.1 1995/09/16 17:02:10 schrod Exp $ +%------------------------------------------------------------ + +% test of `cwebparts' package: +% code parts to be included in a LaTeX document +% no sections + +\documentclass{cweb} + +\usepackage{rcs} + +\begin{document} + +\title{Believe it} +\author{Coogar} + +\maketitle + +@ Some explanation of the code below. It's very hairy, isn't it? + +@( foo.h@>= +@< some |code| @>@; +@< more code @> + +@ @< some ...@>= +for ( int i=1; i<10; i++ ) + @< more... @> + +@ @< more... @>= + +@ +\end{document} diff --git a/Master/texmf-dist/doc/latex/cweb-latex/src/test/parts.tex b/Master/texmf-dist/doc/latex/cweb-latex/src/test/parts.tex new file mode 100644 index 00000000000..252b2129918 --- /dev/null +++ b/Master/texmf-dist/doc/latex/cweb-latex/src/test/parts.tex @@ -0,0 +1,47 @@ +% $Id: parts.tex,v 1.1 1995/09/16 17:02:10 schrod Exp $ +%---------------------------------------------------------------------- + +% +% test of inclusion of CWEB program parts +% +% [LaTeX] +% (history at end) + + + +\documentclass{article} + +\usepackage{cwebparts} + + +\begin{document} + + + +\title{Foobar Algorithm} +\author{Cougar} + +\maketitle + +\makeatletter + + +\section{C Implementation} + +Let's have a look at the implementation in C: + +\cwebInput{parts-code} + +\noindent Got it? + + + +\end{document} + + +%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% + +$Log: parts.tex,v $ +Revision 1.1 1995/09/16 17:02:10 schrod + Test cwebparts.sty. + diff --git a/Master/texmf-dist/doc/latex/cweb-latex/src/test/refname.w b/Master/texmf-dist/doc/latex/cweb-latex/src/test/refname.w new file mode 100644 index 00000000000..e15fa990ddd --- /dev/null +++ b/Master/texmf-dist/doc/latex/cweb-latex/src/test/refname.w @@ -0,0 +1,21 @@ +% $Id: refname.w,v 1.3 1995/08/08 00:14:49 schrod Exp $ +%---------------------------------------------------------------------- + +% tests refinement names, incl. pseudo-refinements w/ file names + +\documentclass{cweb} + +\begin{document} + +@ @( foo.h@>= +@< some |code| @>@; +@< more code @> + +@ @< some ...@>= +for ( int i=1; i<10; i++ ) + @< more... @> + +@ @< more... @>= + +@ +\end{document} diff --git a/Master/texmf-dist/doc/latex/cweb-latex/src/test/report.w b/Master/texmf-dist/doc/latex/cweb-latex/src/test/report.w new file mode 100644 index 00000000000..15242e7908d --- /dev/null +++ b/Master/texmf-dist/doc/latex/cweb-latex/src/test/report.w @@ -0,0 +1,48 @@ +% $Id: report.w,v 1.1 1995/08/27 17:25:40 schrod Exp $ +%------------------------------------------------------------ + +% test baseclass option +% default is `report' + +\documentclass[baseclass]{cweb} + +\begin{document} + +\tableofcontents + +\newpage + +@** s title. +\label{main-sec} + +Some text. + +@* e title. + +Some text. +And some references: Part~\ref{main-sec} and section~\ref{normal-sec}. + +@*0 0 title. + +Some text. + +@*1 1 title. + +Some text. + +@*2 2 title. + +Some text. + +@ \label{normal-sec} +Doc. + +@d def foo +@f def int + +@c +for ( int i=1; i<10; i++ ) + @< do something @> + +@ +\end{document} diff --git a/Master/texmf-dist/doc/latex/cweb-latex/src/test/section.w b/Master/texmf-dist/doc/latex/cweb-latex/src/test/section.w new file mode 100644 index 00000000000..13159e2fc6a --- /dev/null +++ b/Master/texmf-dist/doc/latex/cweb-latex/src/test/section.w @@ -0,0 +1,55 @@ +% $Id: section.w,v 1.6 1995/11/07 14:14:25 schrod Exp $ +%---------------------------------------------------------------------- + +% tests section levels and references. + +\documentclass{cweb} + +\begin{document} + +\tableofcontents + +\newpage + +@** s title. +\label{main-sec} + +Some text. + +@* e title. + +Some text. +And some references: Part~\ref{main-sec} and section~\ref{normal-sec}. + +@*0 0 title. + +Some text. + +@*1 1 title. + +Some text. + +@*2 2 title. + +Some text. + +@*1 1 title. + +Reset counter. + +@*2 2 title. + +Some text. + +@ \label{normal-sec} +Doc. + +@d def foo +@f def int + +@c +for ( int i=1; i<10; i++ ) + @< do something @> + +@ +\end{document} diff --git a/Master/texmf-dist/doc/latex/cweb-latex/src/test/sup-changes.w b/Master/texmf-dist/doc/latex/cweb-latex/src/test/sup-changes.w new file mode 100644 index 00000000000..019dc8388e0 --- /dev/null +++ b/Master/texmf-dist/doc/latex/cweb-latex/src/test/sup-changes.w @@ -0,0 +1,20 @@ +% $Id: sup-changes.w,v 1.1 1995/08/27 13:29:40 schrod Exp $ +%---------------------------------------------------------------------- + +% tests suppress=unchanged + +\documentclass[suppress=unchanged]{cweb} + +\begin{document} + +\title{Test for suppression of unchanged chunks} +\author{Coogar} +\maketitle + +@ test + +@ +change here + +@ +\end{document} diff --git a/Master/texmf-dist/doc/latex/cweb-latex/src/test/sup-format.w b/Master/texmf-dist/doc/latex/cweb-latex/src/test/sup-format.w new file mode 100644 index 00000000000..eec493ff4fd --- /dev/null +++ b/Master/texmf-dist/doc/latex/cweb-latex/src/test/sup-format.w @@ -0,0 +1,17 @@ +% $Id: sup-format.w,v 1.1 1995/08/29 15:22:17 schrod Exp $ +%------------------------------------------------------------ + +% test suppression of format directives + +\documentclass[suppress=format]{cweb} + +\begin{document} + +@ test +@f XmMainWindow int +@f foo int +@f bar i +XmMainWindow dir = XtCreateWidget("foobar"); + +@ +\end{document} diff --git a/Master/texmf-dist/doc/latex/cweb-latex/src/test/sup-lists.w b/Master/texmf-dist/doc/latex/cweb-latex/src/test/sup-lists.w new file mode 100644 index 00000000000..2049f361512 --- /dev/null +++ b/Master/texmf-dist/doc/latex/cweb-latex/src/test/sup-lists.w @@ -0,0 +1,16 @@ +% $Id: sup-lists.w,v 1.1 1995/08/27 19:29:07 schrod Exp $ +%------------------------------------------------------------ + +% test suppression of index & ref list + +\documentclass[suppress={index,reflist}]{cweb} + +\begin{document} + +@ We need something to add to the lists. + +@< refinement @>= +foo(bar); + +@ +\end{document} diff --git a/Master/texmf-dist/doc/latex/cweb-latex/src/test/titlepage.w b/Master/texmf-dist/doc/latex/cweb-latex/src/test/titlepage.w new file mode 100644 index 00000000000..b8230cebeb2 --- /dev/null +++ b/Master/texmf-dist/doc/latex/cweb-latex/src/test/titlepage.w @@ -0,0 +1,19 @@ +% $Id: titlepage.w,v 1.1 1995/08/27 13:29:41 schrod Exp $ +%------------------------------------------------------------ + +% test titlepage option +% test also forwarding of option to baseclass + +\documentclass[titlepage]{cweb} + +\begin{document} + +\title{Test title} +\author{Coogar} + +\maketitle + +Test text. + +@ +\end{document} diff --git a/Master/texmf-dist/doc/latex/cweb-latex/src/test/token.w b/Master/texmf-dist/doc/latex/cweb-latex/src/test/token.w new file mode 100644 index 00000000000..7765f1b9b37 --- /dev/null +++ b/Master/texmf-dist/doc/latex/cweb-latex/src/test/token.w @@ -0,0 +1,29 @@ +% $Id: token.w,v 1.3 1995/08/29 15:21:08 schrod Exp $ +%---------------------------------------------------------------------- + +% tests: rendering of tokens + +\documentclass{cweb} + +\begin{document} + +@ +@c +a = 100L; +a = 100UL; + +@ shift operators: +@c +cout << "hello, world" << endl; +cin >> var; + +@ formatting that depend on font sizes in Plain \texttt{CWEB}: +@f foo_bar int +@c +foo++; bar--; +foo_bar obj; +obj->method(); + + +@ +\end{document} diff --git a/Master/texmf-dist/doc/latex/cweb-latex/src/test/vbar.w b/Master/texmf-dist/doc/latex/cweb-latex/src/test/vbar.w new file mode 100644 index 00000000000..144b2810b36 --- /dev/null +++ b/Master/texmf-dist/doc/latex/cweb-latex/src/test/vbar.w @@ -0,0 +1,30 @@ +% $Id: vbar.w,v 1.2 1995/08/08 00:14:51 schrod Exp $ +%---------------------------------------------------------------------- + +% tests ways to get ruled tables + +\documentclass{cweb} + +\usepackage{cwebarray} + +\begin{document} + +@ Test 1: + +\begin{tabular}{^^7cl^^7cl^^7c} +\hline +1st column & 2nd column\\ +\hline +\end{tabular} + + +@ Test 2: + +\begin{tabular}{IlIlI} +\hline +1st column & 2nd column\\ +\hline +\end{tabular} + + +\end{document} |