summaryrefslogtreecommitdiff
path: root/macros/latex/contrib/mdwtools/sverb.dtx
diff options
context:
space:
mode:
authorNorbert Preining <norbert@preining.info>2019-09-02 13:46:59 +0900
committerNorbert Preining <norbert@preining.info>2019-09-02 13:46:59 +0900
commite0c6872cf40896c7be36b11dcc744620f10adf1d (patch)
tree60335e10d2f4354b0674ec22d7b53f0f8abee672 /macros/latex/contrib/mdwtools/sverb.dtx
Initial commit
Diffstat (limited to 'macros/latex/contrib/mdwtools/sverb.dtx')
-rw-r--r--macros/latex/contrib/mdwtools/sverb.dtx1186
1 files changed, 1186 insertions, 0 deletions
diff --git a/macros/latex/contrib/mdwtools/sverb.dtx b/macros/latex/contrib/mdwtools/sverb.dtx
new file mode 100644
index 0000000000..df12c8c1df
--- /dev/null
+++ b/macros/latex/contrib/mdwtools/sverb.dtx
@@ -0,0 +1,1186 @@
+% \begin{meta-comment}
+%
+% $Id: sverb.dtx,v 1.3 1996/11/19 21:01:18 mdw Exp $
+%
+% Verbatim typesetting done properly (ahem)
+%
+% (c) 1996 Mark Wooding
+%
+%----- Revision history -----------------------------------------------------
+%
+% $Log: sverb.dtx,v $
+% Revision 1.3 1996/11/19 21:01:18 mdw
+% Entered into RCS
+%
+%
+% \end{meta-comment}
+%
+% \begin{meta-comment} <general public licence>
+%%
+%% sverb package -- handling of verbatim text
+%% Copyright (c) 1996 Mark Wooding
+%%
+%% This program is free 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 program is distributed in the hope that it will be useful,
+%% but WITHOUT ANY WARRANTY; without even the implied warranty of
+%% MERCHANTABILITY or 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
+%% along with this program; if not, write to the Free Software
+%% Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
+%%
+% \end{meta-comment}
+%
+% \begin{meta-comment} <Package preamble>
+%<+package>\NeedsTeXFormat{LaTeX2e}
+%<+package>\ProvidesPackage{sverb}
+%<+package> [1996/05/08 1.3 Verbatim typesetting]
+% \end{meta-comment}
+%
+% \CheckSum{651}
+%% \CharacterTable
+%% {Upper-case \A\B\C\D\E\F\G\H\I\J\K\L\M\N\O\P\Q\R\S\T\U\V\W\X\Y\Z
+%% Lower-case \a\b\c\d\e\f\g\h\i\j\k\l\m\n\o\p\q\r\s\t\u\v\w\x\y\z
+%% Digits \0\1\2\3\4\5\6\7\8\9
+%% Exclamation \! Double quote \" Hash (number) \#
+%% Dollar \$ Percent \% Ampersand \&
+%% Acute accent \' Left paren \( Right paren \)
+%% Asterisk \* Plus \+ Comma \,
+%% Minus \- Point \. Solidus \/
+%% Colon \: Semicolon \; Less than \<
+%% Equals \= Greater than \> Question mark \?
+%% Commercial at \@ Left bracket \[ Backslash \\
+%% Right bracket \] Circumflex \^ Underscore \_
+%% Grave accent \` Left brace \{ Vertical bar \|
+%% Right brace \} Tilde \~}
+%%
+%
+% \begin{meta-comment}
+%
+%<*driver>
+\input{mdwtools}
+\describespackage{sverb}
+\mdwdoc
+%</driver>
+%
+% \end{meta-comment}
+%
+% \section{User guide}
+%
+% The \package{sverb} package provides some useful commands and environments
+% for doing things with verbatim text. I prefer this code to the standard
+% \package{verbatim} package (by Rainer Sch\"opf et al.)\ although I'm
+% biased.
+%
+% The package was written to fulfil a particular purpose: I wanted to be able
+% to typeset ARM assembler code, 77~columns wide, on A5~paper, with the
+% fields separated by \textit{tab} characters. It's grown up fairly
+% organically from that, and I've tidied it when I've seen the code get too
+% ugly.
+%
+% The current features are:
+%
+% \begin{itemize}
+%
+% \item A `listing' environment which typesets verbatim text nicely.
+%
+% \item A command to read verbatim text from an external file.
+%
+% \item Support for arbitrary-sized chunks of text without overflowing \TeX's
+% memory.
+%
+% \item Support for \textit{tab} characters in the verbatim text.
+%
+% \item An environment for typesetting demonstrations of \LaTeX\ markup.
+%
+% \item It all works correctly with the \package{doc} system for documenting
+% \LaTeX\ packages.
+%
+% \item A fairly hairy but quite powerful programmer interface to the yukky
+% bits of the package.
+%
+% \end{itemize}
+%
+% The interface is described in its own section, so that more timid readers
+% can avoid it. That said, some of the stuff in this section gets rather
+% technical.
+%
+% Note that this package doesn't even try to do anything with short bits of
+% verbatim text (as handled by the |\verb:...:| command). I have a separate
+% package (\package{syntax}) which does all sorts of horrible things along
+% those lines.
+%
+% \subsection{The \env{listing} environment}
+%
+% \DescribeEnv{listing}
+% The main method for typesetting verbatim text is the \env{listing}
+% environment. This works pretty much the same as the standard
+% \env{verbatim} environment, with some exceptions, which are described
+% below.
+%
+% So that you know exactly what you're getting, here are the rules by which
+% \package{sverb} decides what the verbatim text actually is:
+%
+% \begin{itemize}
+%
+% \item If there's any text, other than spaces, on the same line as the
+% `|\begin{listing}|', then the contents of the environment begins
+% immediately after the closing brace (with all leading spaces
+% preserved). Otherwise, the text begins on the following line.
+%
+% \item If there is any text, other than spaces, before the
+% `|\end{listing}|', but on the same line, this is considered to be the
+% last line of the text; otherwise the text is presumed to have ended
+% at the end of the previous line.
+%
+% \item Any text following the |\end{listing}| on the same line is thrown
+% away. There are good reasons for this, but they're technical.
+% Essentially there's nothing I can do about it.
+%
+% \end{itemize}
+%
+% \begin{figure}
+% \begin{demo}[w]{The \env{listing} environment}
+%\dots in the following code:
+%
+%\begin{listing}
+%init MOV R0,#200 ;Version 2.00 please
+% LDR R1,=&4B534154 ;Magic number (`TASK')
+% ADR R2,appName ;Find application name
+% SWI Wimp_Initialise ;Register as a WIMP task
+%\end{listing}
+%
+%The next step is to \dots
+% \end{demo}
+% \end{figure}
+%
+% Tab characters are supported within the environment: tab stops are set
+% every eighth column, although this can be modified.
+%
+% \subsubsection{Configuring the \env{listing} environment}
+%
+% The text size used in the \env{listing} environment is set by the
+% |\listingsize| command. By default, this is set to |\small|, although you
+% can redefine it in the document preamble, or it can be set in the document
+% class.
+%
+% The amount by which the listing text is indented is controlled by the
+% |\listingindent| length parameter. This is a fixed length, whose default
+% value is 1\,em.
+%
+% \subsubsection{Choosing a different end-text}
+%
+% \DescribeEnv{listing*}
+% The \env{listing} environment is terminated by the exact character sequence
+% `|\end{listing}|'. This isn't too much of a problem, unless you want to
+% include this string in the text. This is achieved by the \env{listing$*$}
+% environment, which allows you to specify the end-text to find as an
+% argument.
+%
+% For example:
+%
+% \begin{demo}{The \env{listing$*$} environment}
+%Type a listing as follows:
+%
+%\begin{listing*}{<end-listing*>}
+%\begin{listing}
+%This is a listing. Yes.
+%\end{listing}
+%<end-listing*>
+%\end{demo}
+%
+% Don't include `special' characters in your chosen end-text unless you know
+% what you're doing.
+%
+% \subsection{Writing text to a file}
+%
+% \DescribeEnv{verbwrite}
+% You can write verbatim text to a file using the \env{verbwrite}
+% environment. The syntax is fairly straightforward:
+%
+% \begin{quote}
+% \syntax{"\\begin{verbwrite}{"<file-name>"}" \dots "\\end{verbwrite}"}
+% \end{quote}
+%
+% The text of the environment is written to the named file. The rules about
+% where the text actually starts and ends are the same as for the
+% \env{listing} environment.
+%
+% There is also a $*$-variant, like \env{listing$*$}, which allows you to
+% choose the end-text. The end-text is the first argument, the filename
+% comes second.
+%
+% There is a restriction on the characters you can write to the file: they
+% must all be considered `printable' by \TeX; otherwise they will be read
+% back in as `\syntax{"^^"<chars>}' which isn't too good. Unfortunately,
+% this includes tab characters, so you can't write them.\footnote{^^A
+% Well, not without doing serious surgery on \TeX\ itself, anyway. }
+%
+% \iffalse [Example time... Ho hum. There is evilness here.] \fi
+%\begin{verbwrite*}{<end-write>}{wrdemo1.tmp}
+%\begin{verbwrite}{wrdemo.tmp}
+%This is some text written to
+%a file near the beginning of
+%the file.
+%\end{verbwrite}
+%<end-write>
+%
+% For example: \verbinput{wrdemo1.tmp}
+%
+% \input{wrdemo1.tmp} \iffalse [Now build the file ;-) ] \fi
+%
+% \subsection{The \cmd\verbinput\ command}
+%
+% \DescribeMacro{\verbinput}
+% You can input a pre-prepared text file exactly as it is in the input using
+% the |\verbinput| command. The filename is given as an argument. For
+% example:
+%
+% \begin{demo}{The \cmd\verbinput\ command}
+%\verbinput{wrdemo.tmp}
+% \end{demo}
+%
+% \subsection{The \env{demo} environment}
+%
+% Package authors need to document their packages, and it's common to want
+% to display examples showing the original text and the output side-by-side
+% (or, when space doesn't permit this, one above the other). Both the
+% \LaTeX\ book and \textit{The \LaTeX\ Companion} contain such examples.
+%
+% The \env{demo} environment allows such displays to be created easily. The
+% syntax of the environment is as follows:
+%
+% \begin{quote}
+% \syntax{"\\begin{demo}["<shape>"]{"<title>"}" \dots "\\end{demo}"}
+% \end{quote}
+%
+% The optional \synt{shape} argument can be either `|w|' (wide), or `|n|'
+% (narrow). A `wide' shape places the input and output one above the other,
+% while the `narrow' shape puts them side-by-side. The default shape is
+% `narrow'. An attractive border is drawn around the display to finish it
+% off nicely.
+%
+% An example:
+%
+%\begin{demo*}{<end-demo>}[w]{The \env{demo} environment}
+%\begin{demo}{From the \textit{\TeX book}}
+%\[ \sum_{p\;\rm prime}
+% f(p) = \int_{t>1}
+% f(t)\,{\rm d}\pi(t) \]
+%\end{demo}
+%<end-demo>
+%
+% \DescribeEnv{demo*}
+% As with the other environments created by this package, there's a
+% $*$-variant which takes the end-text as an argument.
+%
+%
+% \section{Programmer interface}
+%
+% This section describes the publicly available routines provided by the
+% \package{sverb} package. Routines not described here are libable to be
+% changed or even removed without warning, so don't use them.
+%
+% \subsection{Environment hooks}
+%
+% Each of the environments created here works in the same way. For each
+% environment \env{foo}, there's a main command responsible for doing the
+% work, called |\sv@foo|. This is given all the arguments of the normal
+% environment, and two more:
+%
+% \begin{itemize}
+%
+% \item The `end-text' to search for, which marks the end of the environment.
+%
+% \item Some actions to perform after the text has been read and processed.
+% This allows the calling macro to do some extra actions, like closing
+% boxes, etc.
+%
+% \end{itemize}
+%
+% All the environments do is call the main command with appropriate
+% arguments.
+%
+% \subsection{Reading the verbatim text}
+%
+% \DescribeMacro{\sv@read}
+% The main scanning routine is |\sv@read|. It is called with three
+% arguments:
+%
+% \begin{itemize}
+%
+% \item The end-text marking the end of the environment.
+%
+% \item The name of a macro (which must be a single token) which is called
+% with a line of text as its single argument. This is given each
+% line of text which is read from the environment in turn.
+%
+% \item A macro, or other sort of action, which is to be done when the text
+% has been read and processed.
+%
+% \end{itemize}
+%
+% The macro |\sv@read| assumes that the caller has already made some
+% provision for removing the category codes of the following text, by either
+% calling |\@verbatim| or using the construction
+% \begin{listing}
+%\let\do=\@makeother
+%\dospecials
+% \end{listing}
+%
+% \DescribeMacro{\sv@safespc}
+% Note that any space characters you read using |\sv@read| will be catcoded
+% as |\active|. Normally this is OK because |\obeyspaces| (or
+% |\@vobeyspaces|) will be in effect. If you're doing something more exotic,
+% like writing text to a file or building a command string, you can call
+% |\sv@safespc| which defines the active-space character to be a normal
+% whitespace-space when expanded.
+%
+% \implementation
+%
+% \section{Implementation}
+%
+% This section defines several macros and environments which allow verbatim
+% typing, with a high degree of configurability. OK, so this sort of
+% thing's been done so often before that it isn't true, but I don't really
+% care.
+%
+% \begin{macrocode}
+%<*package>
+% \end{macrocode}
+%
+% \subsection{Simple things}
+%
+% To help us build funny macros which involve strange and different category
+% codes, I'll write some simple macros which I can use while building my
+% complicated and clever ones.
+%
+% \begin{macro}{\@cspecials}
+%
+% This macro is used to assist the definition of some of the environments.
+% It makes `|\|', `|{|' and `|}|' into `other' characters, and replaces them
+% with `\verb"|"', `|<|' and `|>|' respectively. Note that `|[|' and `|]|'
+% aren't used, because they make defining commands which take optional
+% arguments awkward. Note that we open a group here. This should be closed
+% using \verb"|endgroup" at the end of the special section.
+%
+% \begin{macrocode}
+\def\@cspecials{%
+ \begingroup%
+ \catcode`|0%
+ \catcode`<1%
+ \catcode`>2%
+ \catcode`\{12%
+ \catcode`\}12%
+ \catcode`\\12%
+}
+% \end{macrocode}
+% \end{macro}
+%
+% \begin{macro}{\sv@startlisting}
+%
+% This macro sets everything up nicely for a \env{listing}-type verbatim
+% environment.
+%
+% \begin{macrocode}
+\def\sv@startlisting{%
+ \def\par{\@@par\penalty\interlinepenalty}%
+ \@@par%
+ \leftskip\@totalleftmargin%
+ \obeylines%
+ \@noligs%
+ \let\do\@makeother\dospecials%
+ \verbatim@font%
+ \frenchspacing%
+ \@vobeyspaces%
+ \settabwidth%
+ \catcode9\active%
+ \lccode`\~9\lowercase{\let~\sv@vtab}%
+ \lccode`\~13\lowercase{\let~\vinput@cr}%
+ \interlinepenalty500%
+}
+% \end{macrocode}
+%
+% \end{macro}
+%
+% \subsection{Tab character handling}
+%
+% One of the things we want to do here is handle tab characters properly.
+% (Here, `properly' means `moving to the next column which is a multiple of
+% eight', the way these things were always meant to.)
+%
+% \begin{macro}{\settabwidth}
+%
+% The tabs used by our tabbed verbatim environments are set up by this
+% routine. It sets the tab width parameter |\svtab| to 8 times the width
+% of a |\tt| space. If you really want, you can redefine this macro.
+%
+% \begin{macrocode}
+\newdimen\svtab
+\def\settabwidth{\setbox\z@\hbox{\texttt{\space}}\svtab8\wd\z@}
+% \end{macrocode}
+%
+% \end{macro}
+%
+% \begin{macro}{\sv@vtab}
+%
+% Here we handle tabs inside verbatim environments. We expect each line to
+% be typeset as a box, using something like
+%
+% \begin{listing}
+%\setbox0\hbox{#1}
+%\leavevmode
+%\box0
+%\par
+% \end{listing}
+%
+% The idea is that you make tab active, and set it to this macro. We stop
+% the current box, stretch it to the right width, and start another one
+% straight after, so nobody know the difference. The code here is straight
+% from Appendix~D of \textit{The \TeX book}.
+%
+% \begin{macrocode}
+\def\sv@vtab{%
+ \hfill\egroup%
+ \@tempdima\wd\z@%
+ \divide\@tempdima\svtab%
+ \multiply\@tempdima\svtab%
+ \advance\@tempdima\svtab%
+ \wd\z@\@tempdima%
+ \leavevmode\box\z@%
+ \setbox\z@\hbox\bgroup%
+}
+% \end{macrocode}
+%
+% \end{macro}
+%
+% \begin{macro}{\verbinput}
+%
+% We allow input from a file, by the |\verbinput| command. We display the
+% text pretty much the same as the \env{listing} environment below.
+%
+% We set tab and return active, and get them to do appropriate things. This
+% isn't actually all that hard.
+%
+% \begin{macrocode}
+\def\verbinput#1{%
+ \begin{listinglist}%
+ \listingsize%
+ \sv@startlisting%
+ \setbox\z@\hbox\bgroup%
+ \input{#1}%
+ \sv@stripspc%
+ \egroup%
+ \ifdim\wd\z@=\z@%
+ \ifhmode\par\fi%
+ \else%
+ \leavevmode\box\z@\par%
+ \fi%
+ \end{listinglist}%
+}
+% \end{macrocode}
+%
+% \end{macro}
+%
+% \begin{macro}{\vinput@cr}
+%
+% This macro handles return characters while inputting text in |\verbinput|.
+% We just output our current box, and start another.
+%
+% \begin{macrocode}
+\def\vinput@cr{%
+ \egroup%
+ \leavevmode\box\z@%
+ \par%
+ \setbox\z@\hbox\bgroup%
+}
+% \end{macrocode}
+%
+% \end{macro}
+%
+% \subsection{Reading verbatim text}
+%
+% The traditional way of reading verbatim text is to use a delimited
+% argument, as described in the \textit{\TeX book}. This works well-ish if
+% the text isn't very long. A better solution would be to pick out the text
+% line-by-line and process it like that. So this is what we do.
+%
+% \begin{macro}{\matcher}
+%
+% For long verbatim environments, we need to be able to find the end text.
+% This is rather tricky. The solution here is rather horrible. The
+% environment picks out each line of the text at a time, as an argument, and
+% tests to see if it contains the text we're after. We do the test in a
+% particularly yukky way: we add the actual target text to the end of the
+% line, and inspect the text following the match to see if the match is at
+% the end.
+%
+% The |\matcher| macro creates a `matcher' which will test strings to see if
+% they contain something interesting.
+%
+% To create a matcher, say
+% \syntax{"\\matcher{"<cmd-name>"}{"<target>"}{"<process-cmd>"}"}. The
+% command \synt{cmd-name} accepts a line of text as an argument and calls
+% the \synt{process-cmd} with the text of the line before the match, or the
+% whole lot. It also sets |\@ifmatched| appropriately.
+%
+% (Having spent ages coming up with this cruft myself, I found some very
+% similar, but slightly better, code in Appendix~D. So I've changed mine to
+% match Donald's. Anyway, credit where it's due: cheers Don.)
+%
+% \begin{macrocode}
+\newif\if@matched
+\def\matcher#1#2#3{%
+ \expandafter\def\csname\string#1$match\endcsname##1#2##2##3\end{%
+ \ifx##2\relax%
+ \@matchedfalse%
+ \else%
+ \@matchedtrue%
+ \fi%
+ #3{##1}%
+ }%
+ \expandafter\def\expandafter#1\expandafter##\expandafter1\expandafter{%
+ \csname\string#1$match\endcsname##1#2\relax\end%
+ }%
+}
+% \end{macrocode}
+%
+% \end{macro}
+%
+% \begin{macro}{\sv@stripspc}
+%
+% This macro strips any trailing glue in the current horizontal list. This
+% is fairly simple, actually: we just loop while glue is the last item. It's
+% slightly complicated by penalties which \TeX\ puts into the list between
+% the glue items, but we just remove them too.
+%
+% \begin{macrocode}
+\def\sv@stripspc{%
+ \unpenalty%
+ \ifdim\lastskip=\z@\else%
+ \unskip\expandafter\sv@stripspc%
+ \fi%
+}
+% \end{macrocode}
+%
+% \end{macro}
+%
+% \begin{macro}{\sv@percent}
+%
+% This macro strips a single leading percent character if there is one, and
+% if the \env{doc} package is loaded. We store the possibly stripped text in
+% |\@tempa|.
+%
+% \begin{macrocode}
+\begingroup
+\catcode`\%=12
+\gdef\sv@percent#1#2\relax
+ {\ifx\check@percent\@@undefined
+ \ifx#1\relax\def\@tempa{}\else
+ \def\@tempa{#1#2}\fi\else
+ \ifx#1\relax\def\@tempa{}\else
+ \ifx#1%\def\@tempa{#2}\else
+ \def\@tempa{#1#2}\fi\fi\fi}
+\endgroup
+% \end{macrocode}
+%
+% \end{macro}
+%
+% \begin{macro}{\@isspaces}
+%
+% We want to avoid writing the first and last lines of the environment to the
+% file if there's nothing in them. To do this, we need to know whether a
+% piece of text contains only space characters. This macro does this, in a
+% rather nasty way. See the other macros below for details of how this
+% works.
+%
+% We define |\sv@safespc| at the same time: this makes space active and
+% expand to a space character which is not active. Neat, huh?
+%
+% \begin{macrocode}
+\lccode`\~32
+\lccode`\!32
+\lowercase{%
+ \def\@isspaces#1{%
+ \ifx#1\relax%
+ \def\@tempb{\@tempswafalse}%
+ \else\ifx#1~%
+ \let\@tempb\@isspaces%
+ \else%
+ \def\@tempb##1\relax{}%
+ \fi\fi%
+ \@tempb%
+ }
+ \def\sv@safespc{%
+ \catcode32\active%
+ \def~{ }%
+ }
+}
+% \end{macrocode}
+%
+% \end{macro}
+%
+% \begin{macro}{\sv@read}
+%
+% This macro does the main job of reading a chunk of verbatim text. You call
+% it like this:
+%
+% \begin{quote}
+% \syntax{"\\sv@read{"<end-text>"}{"<process-line-proc>"}{"<end-proc>"}"}
+% \end{quote}
+%
+% The \synt{end-text} is the text to find at the end of the `environment': we
+% stop when we find it.
+%
+% The \synt{process-line-proc} is a macro which is passed as an argument each
+% line which we read from the text.
+%
+% The \synt{end-proc} is a macro to call once we've finished reading all of
+% the text. This can tidy up an environment or close a file or whatever.
+%
+% We read the text by picking out newlines using a delimited macro. We have
+% to be a little clever, because newlines are active in verbatim text.
+%
+% We will also strip `|%|' signs off the beginning if the \package{doc}
+% package is here (\package{doc} tries to play with \LaTeX's verbatim stuff,
+% and doesn't understand the way we do things).
+%
+% \begin{macrocode}
+\def\sv@read#1#2#3{%
+% \end{macrocode}
+%
+% This code does all sorts of evil things, so I'll start by opening a group.
+%
+% \begin{macrocode}
+ \begingroup%
+% \end{macrocode}
+%
+% So that I can spot the end-text, I'll create a matcher macro.
+%
+% \begin{macrocode}
+ \matcher\@match{#1}\sv@read@ii%
+% \end{macrocode}
+%
+% So that I can identify line ends, I'll make them active. I'll also make
+% spaces active so that they can expand to whatever they ought to expand
+% to (spaces in files, or funny \verb*" " characters or whatever.
+%
+% \begin{macrocode}
+ \catcode13\active%
+ \catcode32\active%
+% \end{macrocode}
+%
+% I'll use the |\if@tempswa| flag to tell me whether I ought to output the
+% current line. This is a little messy, so I'll describe it later. I'll
+% initialise it to false because this is the correct thing to do.
+%
+% \begin{macrocode}
+ \@tempswafalse%
+% \end{macrocode}
+%
+% Most of the job is done by two submacros. I'll define them in terms of
+% my current arguments (to save lots of token munging). The first just
+% extracts the next line (which ends at the next newline character) and
+% tries to match it.
+%
+% \begin{macrocode}
+ \lccode`\~13\lowercase{%
+ \def\sv@read@i##1~{\@match{##1}}%
+ }%
+% \end{macrocode}
+%
+% The results of the match get passed here, along with the text of the
+% line up to the matched text.
+%
+% \begin{macrocode}
+ \def\sv@read@ii##1{%
+% \end{macrocode}
+%
+% The first job to do is to maybe strip off percent signs from the beginning,
+% to keep \package{doc} happy.
+%
+% \begin{macrocode}
+ \sv@percent##1\relax\relax%
+% \end{macrocode}
+%
+% Now I need to decide whether I ought to output this line. The method goes
+% like this: if this is the first line (|\if@tempswa| is false) or the last
+% (|\if@matched| is true), \emph{and} the text consists only of spaces, then
+% I'll ignore it.
+%
+% The first thing to do is to notice the last line -- if |\if@matched| is
+% true, then I'll make |\if@tempswa| false to make the first-line and
+% last-line cases work the same way.
+%
+% \begin{macrocode}
+ \if@matched\@tempswafalse\fi%
+% \end{macrocode}
+%
+% Now if this is the first or last line, I'll examine it for spaces. This
+% is done in a separate macro. It will set |\if@tempswa| false if the
+% text contains only spaces.
+%
+% \begin{macrocode}
+ \if@tempswa\else\@tempswatrue\expandafter\@isspaces\@tempa\relax\fi%
+% \end{macrocode}
+%
+% Now, if |\if@tempswa| is still true, perform the \<process-line-proc> on
+% the line of text. I'll provide a group, so that it doesn't upset me
+% too much.
+%
+% \begin{macrocode}
+ \if@tempswa%
+ \begingroup%
+ \expandafter#2\expandafter{\@tempa}%
+ \endgroup%
+ \fi%
+% \end{macrocode}
+%
+% The next line won't be the first one, so I'll set the flag true in
+% readiness.
+%
+% \begin{macrocode}
+ \@tempswatrue%
+% \end{macrocode}
+%
+% Now, if that wasn't the last line, go round again; otherwise end the group
+% I started ages ago, and do the user's \<end-proc>.
+%
+% \begin{macrocode}
+ \if@matched\def\@tempa{\endgroup#3}\else\let\@tempa\sv@read@i\fi%
+ \@tempa%
+ }%
+% \end{macrocode}
+%
+% Now to start the thing up. I'll read the first line.
+%
+% \begin{macrocode}
+ \sv@read@i%
+}
+% \end{macrocode}
+%
+% \end{macro}
+%
+% \begin{macro}{\sv@readenv}
+%
+% This macro works out an appropriate end-text for the current environment.
+% If you say \syntax{"\\sv@readenv{"<macro-name>"}"}, it will expand do
+% \begin{listinglist} \listingsize \synshorts
+% <macro-name>"{\\"$_{12}$"end{"$_{12}$<current-env-name>"}"$_{12}$"}"^^A
+% "{\\end{"<current-env-name>"}}"
+% \end{listinglist}
+% Easy, no?
+%
+% This is all done with mirrors. No, err\dots\ it's done with
+% |\expandafter|.
+%
+% \begin{macrocode}
+\begingroup
+\lccode`\<=`\{
+\lccode`\>=`\}
+\lccode`\|=`\\
+\lowercase{\endgroup
+\def\sv@readenv#1{%
+ \expandafter\expandafter\expandafter%
+ #1\expandafter\sv@readenv@i\@currenvir\@@%
+}
+\def\sv@readenv@i#1\@@{{|end<#1>}{\end{#1}}}
+}
+% \end{macrocode}
+%
+% \end{macro}
+%
+% \begin{macro}{\sv@verbline}
+%
+% This macro typesets a line in a verbatim way, so you can construct a real
+% verbatim environment from it. It's a bit tricky in the way that it catches
+% the last line. Don't worry about this: it's easy really. Note the
+% |\relax| after the |\par| -- this is because \package{doc} tries to do
+% clever things with |\par| to strip `|%|' signs out.
+%
+% \begin{macrocode}
+\def\sv@verbline#1{%
+ \setbox\z@\hbox{#1\sv@stripspc}%
+ \ifdim\wd\z@=\z@%
+ \if@matched\ifhmode\par\relax\fi\else\leavevmode\par\relax\fi%
+ \else%
+ \leavevmode\box\z@\par\relax%
+ \fi%
+}
+% \end{macrocode}
+%
+% \end{macro}
+%
+% \subsection{Listing environments}
+%
+% The \env{listing} environment is our equivalent of the standard
+% \env{verbatim} environment. We do some slightly cleverer things, though,
+% to make sure (for example) that even text which contains |\end{listing}|
+% can be typeset.
+%
+% \begin{macro}{\listinglist}
+% \begin{environment}{listinglist}
+%
+% This defines the layout for the \env{listing} environment. It starts a
+% list with the appropriate shape. It's also made into an environment, so
+% that the end-paragraph-environment bits work correctly.
+%
+% The |\listingindent| length parameter sets up the indentation of the
+% listings. If there's a |\parindent| setting, I'll line listings up with
+% that; otherwise I'll just choose something which looks right.
+%
+% \begin{macrocode}
+\newdimen\listingindent
+\AtBeginDocument{%
+ \ifdim\parindent=\z@\listingindent1em\else\listingindent\parindent\fi%
+}
+% \end{macrocode}
+%
+% Now to define a size hook for the environment. This is fairly simple
+% stuff.
+%
+% \begin{macrocode}
+\ifx\listingsize\@@undefined
+ \let\listingsize\small
+\fi
+% \end{macrocode}
+%
+% Now to define the environment itself. Suppress the indentation if we're
+% first thing on a new list item, so that the listing lines up with
+% everything else.
+%
+% \begin{macrocode}
+\def\listinglist{%
+ \list{}{%
+ \if@inlabel%
+ \leftmargin\z@%
+ \else%
+ \leftmargin\listingindent%
+ \fi%
+ \rightmargin\z@%
+ \labelwidth\z@%
+ \labelsep\z@%
+ \itemindent\z@%
+ \listparindent\z@%
+ \let\makelabel\relax%
+ \parsep\z@skip%
+ }%
+ \parfillskip\@flushglue%
+ \item\relax%
+}
+\let\endlistinglist\endlist
+% \end{macrocode}
+%
+% \end{environment}
+% \end{macro}
+%
+% \begin{environment}{listing}
+%
+% The \env{listing} environment is the only real verbatim-like environment we
+% create will all this kit, although it does the job very nicely.
+%
+% The environment indents its contents slightly, unlike \env{verbatim}, and
+% uses a smaller typeface in an attempt to fit 77-column text on an A5~page.
+% There is also a $*$-variant, which allows you to specify the terminating
+% text. This enables you to include absolutely any text in the environment,
+% including |\end{listing}|.
+%
+% First, we must define the |\listing| command.
+%
+% \begin{macrocode}
+\def\listing{%
+ \listinglist%
+ \listingsize%
+ \sv@readenv\sv@listing%
+}
+% \end{macrocode}
+%
+% Now we define the |\@listing| command, which does most of the work. We
+% base the \env{listing} environment on a \env{list}.
+%
+% \begin{macrocode}
+\def\sv@listing#1#2{%
+ \sv@startlisting%
+ \sv@read{#1}\sv@verbline{\endlistinglist#2}%
+}
+% \end{macrocode}
+%
+% Now we define the starred version. The command name needs to include the
+% `|*|' character, so we must use |\csname|. There's some hacking here to
+% allow us to read the name using the appropriate catcodes for otherwise
+% normal characters: \LaTeX\ activates some characters and makes them typeset
+% themselves to suppress some ligaturing.
+%
+% \begin{macrocode}
+\expandafter\def\csname listing*\endcsname{%
+ \listinglist%
+ \listingsize%
+ \begingroup%
+ \@noligs%
+ \def\@tempa##1{\endgroup\sv@listing{##1}{\end{listing*}}}%
+ \@tempa%
+}
+% \end{macrocode}
+%
+% \end{environment}
+%
+% \begin{environment}{ignore}
+%
+% The \env{ignore} environment entirely ignores its contents. Anything at
+% all may be put into the environment: it is discarded utterly.
+%
+% We define some macros for defining ignoring environments, because this can
+% be useful for version control, possibly.
+%
+% \begin{macrocode}
+\def\sv@ignore#1#2{%
+ \@bsphack%
+ \let\do\@makeother\dospecials%
+ \sv@read{#1}\@gobble{\@esphack#2}%
+}
+\def\ignore{\sv@readenv\sv@ignore}
+\def\ignoreenv#1{%
+ \expandafter\let\csname #1\endcsname\ignore%
+}
+\def\unignoreenv#1{%
+ \expandafter\def\csname #1\endcsname{\endgroup}%
+ \expandafter\def\csname end#1\endcsname%
+ {\begingroup\def\@currenvir{#1}}%
+}
+% \end{macrocode}
+%
+% \end{environment}
+%
+% \subsection{The \env{verbwrite} environment}
+%
+% The \env{verbwrite} environment allows text to be written to a file in a
+% verbatim way. Note that tab characters don't work, because \TeX\ refuses
+% to be nice.
+%
+% \begin{macro}{\sv@write}
+%
+% As seems to be traditional now, we first define a general hookable macro
+% which allows a caller to specify the end-text and what to do afterwards.
+%
+% \begin{macrocode}
+\newwrite\sv@writefile
+\def\sv@write#1#2{%
+ \begingroup%
+ \@bsphack%
+ \let\do\@makeother\dospecials%
+ \sv@safespc%
+ \sv@read{#1}\sv@writeline{\sv@endwrite#2}%
+}
+\def\sv@writeline#1{%
+ \immediate\write\sv@writefile{#1}%
+}
+\def\sv@endwrite{%
+ \@esphack%
+ \endgroup%
+}
+% \end{macrocode}
+%
+% \end{macro}
+%
+% \begin{environment}{verbwrite}
+%
+% Now we can define the actual environment. We define a $*$-variant which
+% allows the user to specify the end-text, just to make sure.
+%
+% \begin{macrocode}
+\def\verbwrite#1{%
+ \immediate\openout\sv@writefile#1\relax%
+ \sv@readenv\sv@write%
+}
+\def\endverbwrite{\immediate\closeout\sv@writefile}
+\expandafter\def\csname verbwrite*\endcsname#1#2{%
+ \immediate\openout\sv@writefile#2\relax%
+ \sv@write{#1}{\immediate\closeout\sv@writefile\end{verbwrite*}}%
+}
+% \end{macrocode}
+%
+% \end{environment}
+%
+% \subsection{The \env{demo} environment}
+%
+% By way of tying all of this together, I present an environment for
+% displaying demonstrations of \LaTeX\ markup. We read the contents of the
+% environment, write it to a temporary file, and read it back twice,
+% typesetting it the first time and displaying it verbatim the second time.
+%
+% \begin{macro}{\sv@demoname}
+%
+% This macro expands to the filename to use for the temporary data. To
+% allow the package documentation to demonstrate the \env{demo} environment
+% itself, we need to keep a nesting count. This avoids too much hackery,
+% which unfortunately appears to plague all of my \TeX\ code.
+%
+% \begin{macrocode}
+\newcount\sv@nestcount
+\def\sv@demoname{demo\number\sv@nestcount.tmp}
+% \end{macrocode}
+%
+% \end{macro}
+%
+% \begin{macro}{\sv@demo}
+%
+% As for listing, we do all the business through a private macro. This is
+% good because it means we can leave the main macro readable. The argument
+% is the end-text to spot.
+%
+% \begin{macrocode}
+\def\sv@demo#1#2{%
+ \@ifnextchar[{\sv@demo@i{#1}{#2}}{\sv@demo@i{#1}{#2}[n]}%
+}
+\def\sv@demo@i#1#2[#3]#4{%
+ \advance\sv@nestcount by\@ne%
+ \immediate\openout\sv@writefile\sv@demoname\relax%
+ \sv@write{#1}{%
+ \immediate\closeout\sv@writefile%
+ \sv@dodemo{#2}{#3}{#4}%
+ }%
+}
+% \end{macrocode}
+%
+% \end{macro}
+%
+% \begin{environment}{demo}
+%
+% This is the real environment. We provide \env{demo$*$} too, to allow the
+% user to choose the end-text.
+%
+% \begin{macrocode}
+\def\demo{\sv@readenv\sv@demo}
+\expandafter\def\csname demo*\endcsname#1{\sv@demo{#1}{\end{demo*}}}
+% \end{macrocode}
+%
+% \end{environment}
+%
+% \begin{macro}{\sv@dodemo}
+%
+% First, let's define some common bits of code in the stuff below. The
+% minipages used to typeset the material has some clever stuff to avoid
+% strange spacing in the output.
+%
+% \begin{macrocode}
+\def\sv@demosmp{%
+ \begin{minipage}[t]{\@tempdima}%
+ \vskip8\p@%
+ \hrule\@height\z@%
+ \raggedright%
+ \vbox\bgroup%
+}
+\def\sv@demoemp{%
+ \par\unpenalty\unskip%
+ \egroup%
+ \vskip8\p@%
+ \hrule\@height\z@%
+ \end{minipage}%
+}
+% \end{macrocode}
+%
+% This is the macro which actually typesets the demonstration.
+%
+% \begin{macrocode}
+\def\sv@dodemo#1#2#3{%
+% \end{macrocode}
+%
+% Now work out some values. We set |\hsize| to the line width leaving 2\,em
+% of space on either side. The size of the minipages is calculated depending
+% on the shape of the demonstration. This is all fairly simple.
+%
+% \begin{macrocode}
+ \begingroup%
+ \@tempdima\linewidth%
+ \advance\@tempdima-2em%
+ \hsize\@tempdima%
+ \if#2w%
+ \advance\@tempdima-2em%
+ \else%
+ \advance\@tempdima-3em%
+ \divide\@tempdima2%
+ \fi%
+% \end{macrocode}
+%
+% Now we open a big vertical box, and put in a header to mark off the
+% demonstration.
+%
+% \begin{macrocode}
+ \par%
+ \setbox\z@\hbox{\strut\enspace#3\enspace\strut}%
+ \@tempdimb.5\dp\z@%
+ \advance\@tempdimb-.5\ht\z@%
+ \ht\z@\@tempdimb\dp\z@\@tempdimb%
+ \noindent\hskip1em\vtop{%
+ \hb@xt@\hsize{%
+ \hrulefill%
+ \raise\@tempdimb\box\z@%
+ \hrulefill%
+ }%
+ \nointerlineskip%
+ \hb@xt@\hsize{\vrule\@height5\p@\hfil\vrule\@height5\p@}%
+ \nointerlineskip%
+% \end{macrocode}
+%
+% Now we insert the output text in the first minipage. I'll force `|%|'
+% to be a comment character, in case something like \package{doc} has had its
+% wicked way.
+%
+% \begin{macrocode}
+ \vskip-\parskip%
+ \noindent\hbox{}\hskip1em%
+ \sv@demosmp%
+ \catcode`\%14\relax%
+ \input{\sv@demoname}%
+ \sv@demoemp%
+% \end{macrocode}
+%
+% Insert some kind of separation between the two. In `wide' format, we start
+% a new line, and put a ruleoff between the two. In `narrow' format, we just
+% leave some space.
+%
+% \begin{macrocode}
+ \if#2w%
+ \vskip8\p@\hrule\vskip8\p@%
+ \noindent\hbox{}%
+ \fi%
+ \hskip1em%
+% \end{macrocode}
+%
+% Now we put the verbatim copy of the text in the other minipage.
+%
+% \begin{macrocode}
+ \sv@demosmp%
+ \listingindent\z@%
+ \verbinput\sv@demoname%
+ \sv@demoemp%
+ \par%
+ \nointerlineskip%
+ \hb@xt@\hsize{\vrule\@height5\p@\hfil\vrule\@height5\p@}%
+ \hrule%
+ }%
+ \endgroup%
+ \par%
+ \vskip\baselineskip%
+ #1%
+}
+% \end{macrocode}
+%
+% \end{macro}
+%
+% That's all there is. Have fun.
+%
+% \begin{macrocode}
+%</package>
+% \end{macrocode}
+%
+% \hfill Mark Wooding, \today
+%
+% \Finale
+%
+\endinput