summaryrefslogtreecommitdiff
path: root/Master
diff options
context:
space:
mode:
Diffstat (limited to 'Master')
-rw-r--r--Master/texmf-dist/doc/latex/tcldoc/examples/README.txt (renamed from Master/texmf-dist/source/latex/tcldoc/examples/README.txt)0
-rw-r--r--Master/texmf-dist/doc/latex/tcldoc/examples/parsetcl.dtx (renamed from Master/texmf-dist/source/latex/tcldoc/examples/parsetcl.dtx)0
-rw-r--r--Master/texmf-dist/doc/latex/tcldoc/examples/parsetcl.ins (renamed from Master/texmf-dist/source/latex/tcldoc/examples/parsetcl.ins)0
-rw-r--r--Master/texmf-dist/doc/latex/tcldoc/examples/pdf.dtx (renamed from Master/texmf-dist/source/latex/tcldoc/examples/pdf.dtx)0
-rw-r--r--Master/texmf-dist/doc/latex/tcldoc/examples/pdf.ins (renamed from Master/texmf-dist/source/latex/tcldoc/examples/pdf.ins)0
-rw-r--r--Master/texmf-dist/doc/latex/tcldoc/tclldoc.pdfbin0 -> 436262 bytes
-rw-r--r--Master/texmf-dist/doc/latex/tcldoc/tools/README.txt (renamed from Master/texmf-dist/source/latex/tcldoc/tools/README.txt)0
-rw-r--r--Master/texmf-dist/doc/latex/tcldoc/tools/eemenu.dtx (renamed from Master/texmf-dist/source/latex/tcldoc/tools/eemenu.dtx)0
-rw-r--r--Master/texmf-dist/doc/latex/tcldoc/tools/eemenu.ins (renamed from Master/texmf-dist/source/latex/tcldoc/tools/eemenu.ins)0
-rw-r--r--Master/texmf-dist/doc/latex/tcldoc/tools/sourcedtx.dtx (renamed from Master/texmf-dist/source/latex/tcldoc/tools/sourcedtx.dtx)0
-rw-r--r--Master/texmf-dist/doc/latex/tcldoc/tools/sourcedtx.ins (renamed from Master/texmf-dist/source/latex/tcldoc/tools/sourcedtx.ins)0
-rw-r--r--Master/texmf-dist/source/latex/tcldoc/dtxload.dtx1256
-rw-r--r--Master/texmf-dist/source/latex/tcldoc/dtxload.ins54
-rw-r--r--Master/texmf-dist/source/latex/tcldoc/dtxload.tcl361
-rw-r--r--Master/texmf-dist/source/latex/tcldoc/pstokens.dtx979
-rw-r--r--Master/texmf-dist/source/latex/tcldoc/tcldoc.dtx3402
-rw-r--r--Master/texmf-dist/source/latex/tcldoc/tcldoc.ins45
-rwxr-xr-xMaster/tlpkg/bin/tlpkg-ctan-check2
18 files changed, 1 insertions, 6098 deletions
diff --git a/Master/texmf-dist/source/latex/tcldoc/examples/README.txt b/Master/texmf-dist/doc/latex/tcldoc/examples/README.txt
index 2d3d8abe5da..2d3d8abe5da 100644
--- a/Master/texmf-dist/source/latex/tcldoc/examples/README.txt
+++ b/Master/texmf-dist/doc/latex/tcldoc/examples/README.txt
diff --git a/Master/texmf-dist/source/latex/tcldoc/examples/parsetcl.dtx b/Master/texmf-dist/doc/latex/tcldoc/examples/parsetcl.dtx
index 22609d6fa4e..22609d6fa4e 100644
--- a/Master/texmf-dist/source/latex/tcldoc/examples/parsetcl.dtx
+++ b/Master/texmf-dist/doc/latex/tcldoc/examples/parsetcl.dtx
diff --git a/Master/texmf-dist/source/latex/tcldoc/examples/parsetcl.ins b/Master/texmf-dist/doc/latex/tcldoc/examples/parsetcl.ins
index 645dc77f8a8..645dc77f8a8 100644
--- a/Master/texmf-dist/source/latex/tcldoc/examples/parsetcl.ins
+++ b/Master/texmf-dist/doc/latex/tcldoc/examples/parsetcl.ins
diff --git a/Master/texmf-dist/source/latex/tcldoc/examples/pdf.dtx b/Master/texmf-dist/doc/latex/tcldoc/examples/pdf.dtx
index 249a28ee6c3..249a28ee6c3 100644
--- a/Master/texmf-dist/source/latex/tcldoc/examples/pdf.dtx
+++ b/Master/texmf-dist/doc/latex/tcldoc/examples/pdf.dtx
diff --git a/Master/texmf-dist/source/latex/tcldoc/examples/pdf.ins b/Master/texmf-dist/doc/latex/tcldoc/examples/pdf.ins
index 00bea9f775e..00bea9f775e 100644
--- a/Master/texmf-dist/source/latex/tcldoc/examples/pdf.ins
+++ b/Master/texmf-dist/doc/latex/tcldoc/examples/pdf.ins
diff --git a/Master/texmf-dist/doc/latex/tcldoc/tclldoc.pdf b/Master/texmf-dist/doc/latex/tcldoc/tclldoc.pdf
new file mode 100644
index 00000000000..c2646fff73c
--- /dev/null
+++ b/Master/texmf-dist/doc/latex/tcldoc/tclldoc.pdf
Binary files differ
diff --git a/Master/texmf-dist/source/latex/tcldoc/tools/README.txt b/Master/texmf-dist/doc/latex/tcldoc/tools/README.txt
index c8fd744afd8..c8fd744afd8 100644
--- a/Master/texmf-dist/source/latex/tcldoc/tools/README.txt
+++ b/Master/texmf-dist/doc/latex/tcldoc/tools/README.txt
diff --git a/Master/texmf-dist/source/latex/tcldoc/tools/eemenu.dtx b/Master/texmf-dist/doc/latex/tcldoc/tools/eemenu.dtx
index d2383b7c1e0..d2383b7c1e0 100644
--- a/Master/texmf-dist/source/latex/tcldoc/tools/eemenu.dtx
+++ b/Master/texmf-dist/doc/latex/tcldoc/tools/eemenu.dtx
diff --git a/Master/texmf-dist/source/latex/tcldoc/tools/eemenu.ins b/Master/texmf-dist/doc/latex/tcldoc/tools/eemenu.ins
index 6b28dcfde27..6b28dcfde27 100644
--- a/Master/texmf-dist/source/latex/tcldoc/tools/eemenu.ins
+++ b/Master/texmf-dist/doc/latex/tcldoc/tools/eemenu.ins
diff --git a/Master/texmf-dist/source/latex/tcldoc/tools/sourcedtx.dtx b/Master/texmf-dist/doc/latex/tcldoc/tools/sourcedtx.dtx
index 3b024427ab6..3b024427ab6 100644
--- a/Master/texmf-dist/source/latex/tcldoc/tools/sourcedtx.dtx
+++ b/Master/texmf-dist/doc/latex/tcldoc/tools/sourcedtx.dtx
diff --git a/Master/texmf-dist/source/latex/tcldoc/tools/sourcedtx.ins b/Master/texmf-dist/doc/latex/tcldoc/tools/sourcedtx.ins
index f26acf7ee73..f26acf7ee73 100644
--- a/Master/texmf-dist/source/latex/tcldoc/tools/sourcedtx.ins
+++ b/Master/texmf-dist/doc/latex/tcldoc/tools/sourcedtx.ins
diff --git a/Master/texmf-dist/source/latex/tcldoc/dtxload.dtx b/Master/texmf-dist/source/latex/tcldoc/dtxload.dtx
deleted file mode 100644
index f68a0ca3a8b..00000000000
--- a/Master/texmf-dist/source/latex/tcldoc/dtxload.dtx
+++ /dev/null
@@ -1,1256 +0,0 @@
-%
-%
-% \iffalse Meta-comment
-%
-%<*driver>
-\documentclass{tcldoc}
-
-\DeclareRobustCommand{\package}[1]{\textsf{#1}}
-\newcommand{\Alpha}{\textit{Alpha}}
-\newcommand{\AlphaTk}{\textit{Alphatk}}
-
-\PageIndex
-\setcounter{IndexColumns}{2}
-\CodelineNumbered
-\RecordChanges
-\IndexPrologue{\section*{Index}\markboth{Index}{Index}%
- Numbers written in italic refer to the page where
- the corresponding entry is described; numbers underlined refer
- to the definition; numbers in roman refer to the pages where
- the entry is used.}
-\expandafter\def \expandafter\GlossaryParms \expandafter{\GlossaryParms
- \hbadness=10000\emergencystretch=2cm}
-
-\begin{document}
-\DocInput{dtxload.dtx}
-\PrintChanges
-\PrintIndex
-\end{document}
-%</driver>
-% \fi
-%
-% \title{\package{dtxload} --- loading \Tcllogo\ code directly from
-% \texttt{.dtx} source files}
-% \author{Lars Hellstr\"om^^A
-% \thanks{E-mail: \texttt{Lars.Hellstrom@math.umu.se}}}
-% \date{December 20, 2000}
-% \maketitle
-%
-% \begin{abstract}
-% \package{dtxload} is a feature for the text editor \Alpha. It
-% implements a command for loading a piece of \Tcllogo\ code directly
-% from a \texttt{.dtx} file.
-% \end{abstract}
-%
-% \tableofcontents
-%
-% \section{Usage}
-%
-% \subsection{About \Alpha}
-%
-% About \Alpha\ itself, the \Alpha~v\,7.3 manual says the following:
-% \begin{quotation}
-% \Alpha\ is a very powerful, multi-purpose text editor.
-% \Alpha\ is multi-modal, which means it switches between different
-% modes depending on which type of document you are editing. Everybody
-% who has used Emacs knows about this concept. The idea with different
-% modes is to change \Alpha's behaviour depending on which kind of text
-% you're editing. For example if you're programming in C$^{++}$ or
-% editing a \LaTeX\ document you want different features.
-%
-% \Alpha\ is very customizable. Much of \Alpha's functionality is
-% implemented in `packages' or `extensions'. [\dots]
-% \Alpha\ uses Dr.~Ousterhout's \emph{Tool Command Language} (\Tcllogo) as
-% an extension language. Many of the functions bound to keystrokes,
-% as well as many of the functions in the menus, are written in \Tcllogo.
-% All the `packages' or `extensions' just mentioned are all written
-% in \Tcllogo. Anybody who wants can write new packages for \Alpha!
-%
-% Of course, you don't have to learn \Tcllogo\ to use \Alpha. You can
-% have lots of use of \Alpha\ without knowing anything about \Tcllogo.
-% But if you want to become a real power user, you should consider
-% trying it out. Then the possibilities of customizing \Alpha\ are
-% unlimited.
-% \end{quotation}
-% Originally \Alpha\ only existed on the \mbox{MacOS} platform, but
-% recently the program was rewritten in Tk.\footnote{Tk is an extension of
-% \Tcllogo\ which includes GUI support.} In that form (which is called
-% \AlphaTk), \Alpha\ can be run on any platform for which there exists
-% a port of Tk!
-%
-% \medskip
-%
-% There are also some \Alpha-related URLs which should be mentioned.
-% \Alpha's official home page is~\cite{AlphaHome}, but it is currently
-% rather out-of-date. More recent information can be found on
-% e.g.~\cite{DocScriptingAlpha}, which also has links to many other
-% \Alpha-related sites. The home page for \AlphaTk\ is~\cite{Alphatk}.
-%
-%
-%
-% \subsection{Installation}
-%
-% To install \package{dtxload} as an \Alpha\ feature, you first \LaTeX\
-% the file \texttt{dtxload.ins}---this will generate a file
-% \texttt{dtxload.tcl}, which contains the \Tcllogo\ code for the
-% package in a form that \Alpha\ understands. Then follow the normal
-% procedure for installing an \Alpha\ package, i.e., open
-% \texttt{dtxload.tcl}\footnote{Which should be somewhere outside
-% \Alpha's directory subtree.} in \Alpha---this will trigger the
-% automatic installation dialog for the package and you proceed as
-% usual.
-%
-%
-% \subsection{Loading \Tcllogo\ code}
-% \label{Ssec:Loading}
-%
-% To load (evaluate) a piece of \Tcllogo\ code in a \texttt{.dtx}
-% document, simply select that code and press \marginpar{Loading code:
-% \textsf{Cmd--L}}Command--L. The commands are evaluated one by one and
-% the result of each command will be shown at the current log destination
-% (by default the status line, but that can be changed). In case one of
-% the commands return with an error, that error will be shown as well,
-% evaluation of further commands is cancelled, and the insertion point is
-% moved to the beginning of the erroneous command. Furthermore the mark
-% will be left at the end of the selection, so that the part of the
-% original selection which has not yet been successfully evaluated can
-% easily be reselected using the \textsf{Mark Hilite} command.
-%
-% Loading \Tcllogo\ code does not necessary mean all text in the
-% selection will be evaluated; only extractable code will be passed on
-% to the \Tcllogo\ interpreter. There are several factors which affect
-% whether a piece of text is extractable code or not, but the foremost
-% is that text on \package{docstrip} comment lines (i.e., lines which
-% begin by a `|%|' that is not immediately followed by a `|<|') is
-% \emph{not} extractable code. Secondly, if the selection contains a
-% \package{docstrip} guard line which marks the start of a module then
-% \package{dtxload} will ask you whether you want the code in that
-% module to be loaded. If you answer no, it naturally won't be.
-%
-% Thus far the factors which may restrict which text is extractable
-% code correspond exactly to what the \package{docstrip}
-% program~\cite{docstrip} might do when generating a file of extracted
-% code, but \package{dtxload} has two additional mechanisms which may
-% categorize text as ``not extractable code''. Both are however
-% controlled by preferences in the \Alpha\ \TeX\ mode, so they can be
-% switched off if you don't like them.
-% The first is the \describestring[var.][]{onlyLoadDtxFiles}
-% \textsf{Only Load Dtx Files} preference, which when on restricts
-% loading to files with suffix \texttt{.dtx}. For \TeX\ files with other
-% suffixes I suspect that Command--L is pressed more often by accident
-% than because there actually is some \Tcllogo\ code that should be
-% loaded \textellipsis\space The second mechanism is controlled by the
-% \describestring[var.][]{onlyLoadTclEnvs} \textsf{Only Load Tcl Envs}
-% preference, which when on adds the restriction that extractable code
-% must appear inside a \texttt{tcl} or \texttt{tcl*} environment
-% (defined by the \package{tcldoc} \LaTeX\ package~\cite{tcldoc}). This
-% is primarily useful when a \texttt{.dtx} file contains both \Tcllogo\
-% code in \texttt{tcl} environments and code in some other language in
-% e.g.\ \texttt{macrocode} environments, since if the preference had been
-% off then \package{dtxload} would stop at the \package{docstrip} guard
-% that must appear at the beginning of each non-\Tcllogo\ code section
-% and ask whether the code should be included, whereas if it is on then
-% \package{dtxload} doesn't even have to ask whether the code in the
-% \texttt{macrocode} environments should be loaded.
-%
-% \medskip
-%
-% As an example of how the load \Tcllogo\ code command works, consider
-% the following code:
-% \DontCheckModules\iffalse
-%<*example>
-%<<EXAMPLE
-% \fi\begin{macrocode}
-% Text text text text text
-% \begin{tcl}
-set a A
-%<*andB>
-append a B
-%</andB>
-append a C
-%<*andD>
-append a D
-%</andB>
-expr $a
-string length $a
-% \end{tcl}
-% text text text.
-% \end{macrocode}\iffalse
-%EXAMPLE
-%</example>
-% \fi \CheckModules
-% If lines 1--7 above are selected and Command--L is pressed,
-% \package{dtxload} will display the message ``\texttt{Tcl eval OK}'' on
-% the status line and the following information will have been added to
-% the log.\footnote{This shows the window log format, but there are
-% other formats. Furthermore the `\texttt{<}' and `\texttt{>}' will
-% actually be guillemets in \Alpha\ (but not in \AlphaTk).}
-% \begin{trivlist}\item\small
-% |<dtxload> set a A|\\
-% |A|\\[\medskipamount]
-% |<dtxload> append a B|\\
-% |AB|\\[\medskipamount]
-% |<dtxload> append a C|\\
-% |ABC|
-% \end{trivlist}
-% The |<dtxload>| lines show the commands that were evaluated and the
-% lines after these contain the respective results of these commands.
-% Before the |append a B| command is evaluated, \package{dtxload} will
-% stop and ask
-% \begin{quote}
-% \textsf{Should \textless andB\textgreater\ modules be included?}
-% \end{quote}
-% In the case above the answer given was `Yes'; if it had been `No' then
-% the |append a B| command would neither be logged nor evaluated, and
-% the result of the |append a C| command would have been |AC|.
-%
-% If one continues by selecting lines 9--12 and pressing Command--L
-% then \package{dtxload} will display the message ``\texttt{Tcl eval
-% error}'' on the status line, the insertion point will be moved to the
-% beginning of line 11, and the following information will be found in
-% the log:
-% \begin{trivlist}\item\small
-% |<dtxload> append a D|\\
-% |ABCD|\\[\medskipamount]
-% |<dtxload> expr $a|\\
-% |Error: syntax error in expression "ABCD"|
-% \end{trivlist}
-% If then line 11 is changed to |expr {$a}| and lines 11--12 are again
-% selected (e.g. using \textsf{Mark Hilite}) and loaded then
-% \package{dtxload} will log:
-% \begin{trivlist}\item\small
-% |<dtxload> expr {$a}|\\
-% |ABCD|\\[\medskipamount]
-% |<dtxload> string length $a|\\
-% |4|
-% \end{trivlist}
-%
-% Finally, if lines 3--12 are selected and loaded then these are the
-% questions \package{dtxload} will ask:
-% \begin{itemize}
-% \item \textsf{Should \textless andB\textgreater\ modules be included?}
-% (Possible answers are `Yes' and `No'.)
-% \item \textsf{Should \textless andD\textgreater\ modules be included?}
-% (Possible answers are `Yes' and `No'.)
-% \item \textsf{Module nesting error: \textless*andD\textgreater\
-% module ended by \textless/andB\textgreater. For which
-% guards should the positions be pushed?}
-% (Possible answers are `None', `Start', `End', and `Both'.)
-% \end{itemize}
-% In the last case, \package{dtxload} has noticed that the guard
-% expressions that began (\textsf{andD}) and ended (\textsf{andB}) a
-% module are not equal (which is an error). Answering `Both' here will
-% push the positions of (the beginning of) both guard lines (8 and 10) onto
-% the bookmark stack so that they can easily be jumped to and corrected
-% later, but \package{dtxload} will continue evaluating the rest of the
-% selection after you answer.
-%
-%
-% \subsection{Selecting log destination}
-%
-% At startup, \package{dtxload} is logging results on the status line,
-% i.e., after a command has been evaluated it either displays the message
-% \begin{quote}
-% \texttt{Tcl eval OK:\ }\meta{result}
-% \end{quote}
-% or
-% \begin{quote}
-% \texttt{Tcl eval error:\ }\meta{error message}
-% \end{quote}
-% depending on whether the command returned an error or not. The
-% evaluation log can however be sent to a window instead.
-%
-% To set the destination for the log, one uses the procedure
-% \describestring[proc][dtxload]{logdest}|dtxload::logdest|. To set the
-% log destination to the window \word{win}, one gives the command
-% \begin{quote}
-% |dtxload::logdest window| \word{win}
-% \end{quote}
-% When the log destination is a window, the log shows the commands
-% that were evaluated as well as their result. If there is no window
-% named \word{win} then it is created. A slight
-% variation on this is the command
-% \begin{quote}
-% |dtxload::logdest shell| \word{win}
-% \end{quote}
-% which makes the \word{win} window a \Alpha\ \Tcllogo\ shell window if
-% it has to create it. When the log destination is a shell window, the
-% commands that are executed will be saved on the history list as well
-% as inserted in the window.
-%
-% A short form which may be convenient is
-% \begin{quote}
-% |dtxload::logdest here|
-% \end{quote}
-% which simply makes the current window the log destination. Finally,
-% the log destination is reset to the status line after the command
-% \begin{quote}
-% |dtxload::logdest status|
-% \end{quote}
-% Evaluation will also be logged on the status line if the log
-% destination window is closed.
-%
-%
-% \subsection{Other features and surprises}
-%
-% If nothing is selected when Command--L is pressed, \package{dtxload}
-% will continue as if all text was selected.
-%
-% Like \Alpha's \Tcllogo\ shell, \package{dtxload} can send the commands
-% to evaluate to a remote \Tcllogo\ interpreter rather than \Alpha's
-% built-in interpreter. In both cases the choice of target interpreter is
-% controlled by the global \Alpha\ variable
-% \describestring[var.][]{evaluateRemotely}|evaluateRemotely|. The
-% easiest way of changing this variable is via the \Tcllogo-tk submenu
-% of \Alpha's \Tcllogo\ menu.
-%
-% If the \textsf{Only Load Tcl Envs} preference is on and lines 4--6 in
-% the code example of Subsection~\ref{Ssec:Loading} are loaded then
-% \package{dtxload} will simply display the message
-% \begin{quote}
-% \texttt{No extractable code found.}
-% \end{quote}
-% The reason for this is that the first character of the selection is
-% a |%|; when the first character in the interval to load is a percent
-% character then \package{dtxload} will guess that this position is not
-% in a \texttt{tcl} or \texttt{tcl*} environment, and therefore it won't
-% try to extract any code until it has seen a comment line that contains
-% |\begin{tcl}| or |\begin{tcl*}|. You can always avoid making a
-% \package{docstrip} guard the first thing in the selection to load.
-%
-%
-%
-% \StopEventually{
-% \begin{thebibliography}{9}
-% \bibitem{Alphatk}
-% Vince Darley: \textit{Alphatk} ({\small WWW} page),
-% \texttt{http}://\texttt{www.santafe.edu}\slash
-% \texttt{\~{}vince}\slash \texttt{alphatk}\slash
-% \texttt{about.html}
-% \bibitem{DocScriptingAlpha}
-% Donavan Hall: \textit{Alpha} ({\small WWW} page),
-% \texttt{http}://\texttt{magnet.fsu.edu}\slash
-% \texttt{\~{}hall}\slash\texttt{docscripting}\slash
-% \texttt{alpha}/.
-% \bibitem{tcldoc}
-% Lars Hellstr\"om:
-% \textit{The \package{tcldoc} package and class},
-% \texttt{ftp}:/\slash\texttt{ftp.tex.ac.uk}\slash
-% \texttt{tex-archive}\slash \texttt{macros}\slash
-% \texttt{latex}\slash \texttt{contrib}\slash
-% \texttt{supported}\slash \texttt{tcldoc.dtx}.
-% \bibitem{AlphaHome}
-% Pete Keleher: \textit{Alpha Home Page} ({\small WWW} page),
-% \texttt{http}://\texttt{alpha.olm.net}/.
-% \bibitem{docstrip}
-% Frank Mittelbach, Denys Duchier, Johannes Braams, Marcin
-% Woli\'nski, and Mark Wooding: \textit{The \textsf{DocStrip}
-% program}, The \LaTeX3 Project;
-% \texttt{ftp}:/\slash\texttt{ftp.tex.ac.uk}\slash
-% \texttt{tex-archive}\slash \texttt{macros}\slash
-% \texttt{latex}\slash \texttt{base}\slash \texttt{docstrip.dtx}.
-% \end{thebibliography}
-% }
-%
-% \section{Implementation}
-%
-%
-% \subsection{Interface to \Alpha}
-%
-% The package begins with an |alpha::feature| command to declare itself
-% to \Alpha. As it is hardly useful outside the \texttt{TeX} mode, it
-% declares itself as being specific to that mode.
-% \begin{tcl}
-%<*tcl>
-alpha::feature dtxload 1.1 {TeX} {
-% \end{tcl}
-%
-% \begin{variable}{onlyLoadDtxFiles}
-% \begin{variable}{onlyLoadTclEnvs}
-% Next comes the initialization script for the feature, which
-% declares the two preference variables used by \package{dtxload}.
-% Both variables are flags. |onlyLoadDtxFiles| is |1| if loading of
-% code is restricted to files with suffix \texttt{.dtx} and |0| if
-% there is no restriction. |onlyLoadTclEnvs| is |1| if only code
-% surrounded by a \texttt{tcl} or \texttt{tcl*} environment is loaded
-% and |0| if there is no such restriction. Both preferences are by
-% default |1|.
-% \begin{tcl}
- # When set, the Load Tcl Code command will be disabled in
- # TeX files whose suffix is not dtx.
- newPref flag onlyLoadDtxFiles 1 TeX
- # When set, the Load Tcl Code command will only extract code
- # from code lines in tcl and tcl* environments.
- newPref flag onlyLoadTclEnvs 1 TeX
-% \end{tcl}
-% \end{variable}\end{variable}
-% \begin{variable}{loadTclCode}
-% The initialization script also declares a keybinding preference (to
-% the |dtxload::load| procedure), which is used for
-% \package{dtxload}'s only command. The default value is Command--L
-% and the binding is specific to the \texttt{TeX} mode.
-% \begin{tcl}
- # This is the key binding for dtxload's Load Tcl Code command.
- newPref binding loadTclCode <O/L TeX "" dtxload::load TeX
-% \end{tcl}
-% \end{variable}
-%
-% After the initialization script comes the activation and deactivation
-% scripts (which, if I understood things correctly, are called when
-% switching between modes), but these need not contain anything for
-% \package{dtxload}. Instead the arguments of |alpha::feature| continue
-% with the uninstallation procedure
-% \begin{tcl}
-} "" "" uninstall {this-file} \
-% \end{tcl}
-% the maintainer specification (|\x9a| is \texttt{\"{o}})
-% \begin{tcl}
- maintainer {"Lars Hellstr\x9am" Lars.Hellstrom@math.umu.se} \
-% \end{tcl}
-% and a short help text
-% \begin{tcl}
- help {The dtxload package implements a command for loading
- Tcl code that is embedded in a .dtx file without docstripping
- the file first. For more help, LaTeX the file dtxload.dtx.}
-% \end{tcl}
-% It is possible that the help should rather be |file dtxload.dtx|, but
-% then this file would have to be installed somewhere that \Alpha\
-% would find it. I haven't studied the matter.
-%
-% Next the |dtxload| namespace is declared to \Tcllogo.
-% \begin{tcl}
-namespace eval dtxload {}
-% \end{tcl}
-% \setnamespace{dtxload}
-%
-%
-% \subsection{Code extraction}
-%
-%
-% The code extraction algorithm of \package{dtxload} lack a few of the
-% features of proper \package{docstrip}, but I don't think any of them
-% are essential for the kind of work \package{dtxload} is designed for.
-% The differences to \package{docstrip} are:
-% \begin{itemize}
-% \item The values of guard expressions are not determined from a
-% base set of options, instead the user is inquired for the value of
-% each guard expression.
-% \item Extraction of \package{docstrip} verbatim mode lines is not
-% supported, but it isn't needed for \Tcllogo\ code either.
-% \item Metacomments aren't extracted, but they hardly contain any
-% code of interest to the \Tcllogo\ interpreter anyway.
-% \item \package{dtxload} may scan the comment lines for certain
-% known markup elements (beginning and end of \texttt{tcl} and
-% \texttt{tcl*} environments) and choose to enable or disable
-% extraction based on that.
-% \end{itemize}
-% \package{dtxload} does however act on the guard modifiers and it also
-% looks for errors in the syntax of guards.
-%
-% \begin{variable}{module_stack}
-% \begin{variable}{next_module_idx}
-% The |dtxload::module_stack| variable is keeps track of how modules
-% are nested around the currently considered line. The variable is a
-% stack implemented as a list in which the last element corresponds to
-% the innermost module. Each element in this list is itself a list
-% with the structure
-% \begin{quote}
-% \word{expression} \word{start pos.} \word{previous}\regopt
-% \end{quote}
-% The \word{expression} is the guard expression, as a string. The
-% \word{start pos.} is the position in the file where the module
-% started. Finally, the \word{previous} is a boolean value which
-% records the inclusion status (value of |dtxload::module_included|)
-% that was in force before the module was entered.
-%
-% The |dtxload::next_module_idx| variable holds the index in the
-% |dtxload:|\penalty\hyphenpenalty|:module_stack| list of the next
-% module whose expression has not yet been evaluated; it is equal to
-% |llength ${dtxload::module_stack}| if all module expressions have
-% been evaluated. Those elements in the |dtxload::module_stack| list
-% which have index $\geq$ |dtxload::next_module_idx| do not have a
-% \word{previous} element.
-% \end{variable}\end{variable}
-%
-% \begin{arrayvar}{known_expressions}[expression]
-% The |dtxload::known_expressions| array records the values of all
-% known expressions, i.e., those expressions about whose values the
-% user has been inquired. The entries are either |0| (for ``don't
-% include'') or |1| (for ``include'').
-% \end{arrayvar}
-%
-% \begin{proc}{eval_guard}
-% The |dtxload::eval_guard| procedure takes the string that is a guard
-% expression as argument. It returns |1| if the expression evaluates
-% to true and |0| if it evaluates to false.
-%
-% If the expression has an entry in the |dtxload::known_expressions|
-% array then the |dtxload::eval_guard| procedure returns that value,
-% and if it does not the user is inquired for the value.
-% \begin{tcl}
-proc dtxload::eval_guard {e} {
- global dtxload::known_expressions
- if {![info exists dtxload::known_expressions($e)]} then {
- set dtxload::known_expressions($e)\
- [expr {"yes" ==\
- "[askyesno "Should <$e> modules be included?"]"}]
- }
- set dtxload::known_expressions($e)
-}
-% \end{tcl}
-% \end{proc}
-%
-% \begin{proc}{push_module}
-% The |dtxload::push_module| procedure pushes a block module onto the
-% module stack, but it does not modify the value of
-% |dtxload::module_included|. The syntax of |dtxload::push_module| is
-% \begin{quote}
-% |dtxload::push_module| \word{expression} \word{position}
-% \end{quote}
-% where \word{expression} is the expression string and
-% \word{position} is the position where the guard begins.
-% \begin{tcl}
-proc dtxload::push_module {e p} {
- global dtxload::module_stack
- lappend dtxload::module_stack [list $e $p]
-}
-% \end{tcl}
-% \end{proc}
-%
-% \begin{proc}{pop_module}
-% The |dtxload::pop_module| procedure pops a block module off the
-% module stack and checks that the modules are properly nested. Its
-% syntax is
-% \begin{quote}
-% |dtxload::pop_module| \word{expression} \word{position}
-% \end{quote}
-% where \word{expression} is the expression string of the module to
-% pop and \word{position} is the position where the (end of module)
-% guard begins.
-%
-% If the module nesting is incorrect then the user is presented with
-% a dialog with buttons for pushing the positions of the incorrect
-% modules. It is not an error if the module stack is empty, since it
-% is perfectly reasonable that the beginning of the code to evaluate
-% is in the middle of a module.
-% \begin{tcl}
-proc dtxload::pop_module {e p} {
- global dtxload::module_stack dtxload::module_included\
- dtxload::next_module_idx
- set len [llength ${dtxload::module_stack}]
- if {$len==0} then {return}
- set L [lindex ${dtxload::module_stack} [expr {$len-1}]]
- if {[lindex $L 0] != $e} then {
- switch [buttonAlert "Module nesting error: <*[lindex $L 0]>\
- module ended by </$e>. For which guards should the positions\
- be pushed?" None Start End Both]\
- {
- Start {dtxload::push_bookmarks [lindex $L 1]}
- End {dtxload::push_bookmarks $p}
- Both {dtxload::push_bookmarks [lindex $L 1] $p}
- }
- }
- set dtxload::module_stack\
- [lreplace ${dtxload::module_stack} end end]
- if {$len>=${dtxload::next_module_idx}} then {
- incr dtxload::next_module_idx -1
- }
- if {[llength $L]>2} then {
- set dtxload::module_included [lindex $L 2]
- }
-}
-% \end{tcl}
-% \end{proc}
-%
-% \begin{proc}{push_bookmarks}
-% The |dtxload::push_bookmarks| procedure pushes one or several
-% positions in the current window onto the bookmarks stack. The
-% arguments are the positions to push.
-% \begin{tcl}
-proc dtxload::push_bookmarks {args} {
- global markStack markName
- foreach pos $args {
- set name mark$markName
- incr markName
- createTMark $name $pos
- set fileName [win::Current]
- set markStack [linsert $markStack 0 [list $fileName $name]]
- }
-}
-% \end{tcl}
-% \end{proc}
-%
-% \begin{variable}{module_included}
-% \begin{variable}{in_tcl_env}
-% The |dtxload::module_included| variable keeps track of whether the
-% current position is in an included module. It is |1| if the current
-% position should be included and |0| if it shouldn't.
-% The |dtxload::in_tcl_env| variable is used for keeping track of
-% what markup context surrounds the currently considered line in the
-% file. It is set to |1| when a |\begin{tcl}| or |\begin{tcl*}| is
-% scanned, and set to |0| when a |\end{tcl}| or |\end{tcl*}| is scanned.
-% \end{variable}\end{variable}
-%
-% \begin{proc}{update_included}
-% Simply pushing a module onto the module stack does not update the
-% |dtxload:|\penalty\hyphenpenalty|:module_included| variable; that is
-% instead done by the |dtxload::update_included| procedure, which
-% should be called right before the value of |dtxload::module_included|
-% is used. The reason for this arrangement is that it avoids evaluating
-% guard expressions whose values do not matter.
-%
-% |dtxload::update_included| has no arguments.
-% \begin{tcl}
-proc dtxload::update_included {} {
- global dtxload::module_included dtxload::module_stack\
- dtxload::next_module_idx
- while {${dtxload::module_included} &&\
- ${dtxload::next_module_idx}<[llength ${dtxload::module_stack}]}\
- {
- set L [lindex ${dtxload::module_stack} ${dtxload::next_module_idx}]
- lappend L ${dtxload::module_included}
- set dtxload::module_stack [lreplace ${dtxload::module_stack}\
- ${dtxload::next_module_idx} ${dtxload::next_module_idx} $L]
- incr dtxload::next_module_idx
- set dtxload::module_included [dtxload::eval_guard [lindex $L 0]]
- }
-}
-% \end{tcl}
-% \end{proc}
-%
-%
-% \begin{proc}{extract_line}
-% The |dtxload::extract_line| procedure steps forward in the file one
-% line at a time until it has found a codeline that can be extracted or
-% it reaches the end of the part of the file from which code is to be
-% extracted. Its syntax is
-% \begin{quote}
-% |dtxload::extract_line| \word{startpos} \word{endpos}
-% \end{quote}
-% Here \word{startpos} is the position at which it should start to
-% look for an extractable code line, and \word{endpos} is the
-% position at which the search should stop if no such line is found.
-% \word{startpos} is assumed to be at the beginning of a line (for
-% identification of code, comment, and guard line purposes) but no
-% test for this is made.
-%
-% The return value of |dtxload::extract_line| is a list with the
-% structure
-% \begin{quote}
-% \word{nextpos} \word{line}\regopt\ \word{foundpos}\regopt
-% \end{quote}
-% \word{nextpos} is the first position in the file after the part that
-% was searched. \word{line} is the extracted codeline and
-% \word{foundpos} is the position at which it began, but these two items
-% are missing if there was no more codeline.
-%
-% The overall structure of |dtxload::extract_line| is a loop
-% which steps through the lines until it reaches the |endpos|. When a
-% line to extract is actually found the procedure returns from within
-% the loop using the |return| command. The local variable |pos|
-% stores the current position, and |pos2| stores the position of the
-% beginning of the next line.
-% \begin{tcl}
-proc dtxload::extract_line {startpos endpos} {
- global dtxload::in_tcl_env onlyLoadTclEnvs dtxload::module_included
- for {set pos $startpos} {[pos::compare {$pos<$endpos}]}\
- {set pos $pos2} {
-% \end{tcl}
-% Inside the loop, the next line is extracted and stored in the local
-% variable |line|. Then |switch| is used for classifying the line and
-% taking appropriate actions.
-% \begin{tcl}
- set pos2 [nextLineStart $pos]
- if {$pos2>$endpos} then {set pos2 $endpos}
- set line [getText $pos $pos2]
- switch -regexp -- $line {
-% \end{tcl}
-% \changes{v\,1.0}{2000/10/15}{Corrected regular expressions; only
-% percents on the beginning of a line are significant. (LH)}
-% \changes{v\,1.1}{2000/12/03}{Guard lines will affect code extraction.
-% (LH)}
-% \changes{v\,1.1}{2000/12/18}{The code extraction procedures can
-% now handle \package{docstrip} verbatim mode guards, provided
-% they do not appear in a block of extractable code. (LH)}
-% Lines that begin with |%<<| are \package{docstrip} guard lines that
-% start verbatim mode. Extracting code from such lines is currently
-% not supported by \package{dtxload}, so if such a guard is ever
-% encountered the extraction is stopped, but \package{dtxload} can
-% skip verbatim sections from which it shouldn't extract code.
-% \begin{tcl}
- {^%<<} {
- if {!$onlyLoadTclEnvs || ${dtxload::in_tcl_env}} then {
- dtxload::update_included
- if {${dtxload::module_included}} then {
- alertnote "docstrip verbatim mode guard line\
- encountered. Verbatim mode extraction is\
- currently not supported, so I stop here."
- goto $pos
- return [list $endpos]
- }
- }
-% \end{tcl}
-% The part of the guard line which comes after the |%<<| is called an
-% end-tag. Verbatim mode ends with the next line which consists of a
-% percent character followed by this end-tag, and therefore all lines
-% up to and including that one are skipped. A technical matter here
-% is that spaces at the end of the line does not count (\TeX\ cannot
-% see them), but spaces inside the end-tag do count!
-% \begin{tcl}
- set endtag %[string trimright [string trimright\
- [string range $line 3 end] \n\r] \ ]
- while {[string trimright [string trimright $line \n\r] \ ]\
- != $endtag} {
- set pos $pos2
- if {[pos::compare {$pos>=$endpos}]}\
- then {return [list $pos]}
- set pos2 [nextLineStart $pos]
- if {$pos2>$endpos} then {set pos2 $endpos}
- set line [getText $pos $pos2]
- }
- }
-% \end{tcl}
-% Lines that begin with |%<*| or |%</| are \package{docstrip} guard
-% lines that delimit block modules, but they do not contain any code
-% themselves. The module is pushed onto or popped off the module stack
-% when the guard line is encountered, but evaluating it is deferred
-% until the first extractable code line.
-% \begin{tcl}
- {^%<[*/]} {
- if {![regexp {^%<(\*|/)([^>]+)>} $line foo modifier\
- expression]}\
- then {
- if {"[askyesno "Malformed guard line \"$line\"\
- encountered. Push position?"]"=="yes"}\
-%
- then {dtxload::push_bookmarks $pos}
- } elseif {$modifier=="*"} then {
- dtxload::push_module $expression $pos
- } else {
- dtxload::pop_module $expression $pos
- }
- }
-% \end{tcl}
-%
-% Other lines that begin with |%<| are also \package{docstrip}
-% guard lines, but in these cases the lines (i) contain some code
-% and (ii) the guard only affect the code on that line. There must be
-% a |>| somewhere on the line, and the part of the line after the |>|
-% is code that \package{docstrip} can extract.
-% \begin{tcl}
- {^%<} {
- if {![regexp {^%<(-|\+|)([^>]+)>(.*)$} $line\
- foo modifier expression code]}\
- then {
- if {"[askyesno "Malformed guard line \"$line\"\
- encountered. Push position?"]"=="yes"}\
-%
- then {dtxload::push_bookmarks $pos}
- } elseif {!$onlyLoadTclEnvs || ${dtxload::in_tcl_env}} then {
- dtxload::update_included
- if {${dtxload::module_included}} then {
- if {[dtxload::eval_guard $expression] !=\
- ($modifier=="-")}\
- then {return [list $pos2 $code $pos]}
- }
- }
- }
-% \end{tcl}
-% Remaining lines that begin with |%| are considered comment lines,
-% but they could technically be metacomment lines (begin with |%%|) or
-% a line that \package{docstrip} would process in verbatim mode. In
-% both these cases \package{docstrip} would copy the line, but as there
-% is little point in using these \package{docstrip} features for lines
-% that the \Tcllogo\ interpreter would miss, \package{dtxload} ignores
-% those subtleties.
-% \begin{tcl}
- {^%} {
- while {[regexp -indices {\\(begin|end) *{tcl\*?}} $line\
- match type]} {
- set dtxload::in_tcl_env\
- [expr {"[eval string range \$line $type]"=="begin"}]
- set line [string range $line\
- [expr {[lindex $match 1]+1}] end]
- }
- }
-% \end{tcl}
-% Finally, all lines that do not begin with a |%| are code lines.
-% \begin{tcl}
- default {
- if {!$onlyLoadTclEnvs || ${dtxload::in_tcl_env}} then {
- dtxload::update_included
- if {${dtxload::module_included}}\
-%
- then {return [list $pos2 $line $pos]}
- }
- }
- }
- }
- list $endpos
-}
-% \end{tcl}
-% \end{proc}
-%
-%
-%
-% \subsection{Evaluating the code}
-%
-% The code in this subsection is based on the \Alpha\ procedures
-% |Alpha::evaluate| and |tcltk::evaluate|, but |dtxload::evaluate|
-% tries to return a bit more information.
-%
-% \begin{proc}{evaluate}
-% The |dtxload::evaluate| procedure takes a script as argument and
-% evaluates it in the global variable context. Any errors that occur
-% are caught. The returned value is a list with the structure
-% \begin{quote}
-% \word{code} \word{message}
-% \begin{regexp}[\regopt]\word{errorInfo}
-% \word{errorCode}\end{regexp}
-% \end{quote}
-% \word{code} is the \Tcllogo\ return code, as returned by |catch|
-% when evaluating the script. \word{message} is the value returned by
-% the script; it is an error message if an error occurred.
-% \word{errorInfo} is the value that the global variable |errorInfo|
-% got after the script was evaluated and \word{errorCode} is the
-% corresponding value of |errorCode|.
-%
-% When the \Alpha\ global variable |evaluateRemotely| is |0| the
-% evaluation is carried out by \Alpha\ itself, but when
-% |evaluateRemotely| is |1| the script is instead sent to a separate
-% \Tcllogo\ shell for evaluation. The current implementation does not
-% return the values of the |errorInfo| and |errorCode| variables in
-% that case, which is why they may be absent from the return value of
-% |dtxload::evaluate|.
-% \changes{v\,1.0}{2000/10/17}{Moved \texttt{history} interaction to
-% \texttt{dtxload::shell\_log}. (LH)}
-% \changes{v\,1.1}{2000/12/05}{Added code for remote evaluation. (LH)}
-% \begin{tcl}
-proc dtxload::evaluate {t} {
- global errorInfo errorCode evaluateRemotely tcl_platform
- if {!$evaluateRemotely} then {
- if {[set code [catch {uplevel \#0 $t} msg]] == 1} then {
-% \end{tcl}
-% If there was an error, the last part of the value of |errorInfo|
-% will be due to the |uplevel| command. As that command was not
-% requested by the user, it would be confusing and it is therefore
-% best to remove it.
-% \begin{tcl}
- set L [split $errorInfo \n]
- set errorInfo [join [lrange $L 0 [expr {[llength $L] - 4}]] \n]
- }
- list $code $msg $errorInfo $errorCode
- } elseif {$tcl_platform(platform) == "macintosh"} then {
-% \end{tcl}
-% On the \mbox{MacOS} platform, the script is sent to a remote
-% \Tcllogo\ shell via AppleEvents.
-% \begin{tcl}
- global tclshSig
- app::ensureRunning $tclshSig
- list [catch {
- aebuild::result -t 30000 '${tclshSig}' misc dosc ----\
- [aebuild::TEXT $t]
- } res]\
- $res
- } else {
-% \end{tcl}
-% On the other platforms I don't really know what should be done. The
-% following code is based on that in |tcltk::evaluate|, but I haven't
-% been able to test it (help, please?).
-% \begin{tcl}
- global tclshInterp
- if {![info exists tclshInterp]} {
- if {[catch {tcltk::findTclshInterp}]} {
- return [list 1 "No shell selected"]
- }
- }
- if {$tcl_platform(platform) == "windows"} {
- if {[dde services Tk $tclshInterp] == ""} {
- alertnote "The remote shell has died, please select a new one."
- unset tclshInterp
- return [dtxload::evaluate $t]
- }
- list [catch {dde execute Tk $tclshInterp $t} res] $res
- } else {
- list [catch {send $tclshInterp $t} res] $res
- }
- }
-}
-% \end{tcl}
-% \end{proc}
-%
-%
-% \subsection{Logging}
-%
-% The code in this subsection is based on the \Alpha\ procedures
-% |Shel::carriageReturn| and |Alpha::evaluate|, but many details had to
-% be changed because, amongst other things, the window to which text is
-% written is not the frontmost one.
-%
-% \begin{proc}{message_log}
-% The |dtxload::message_log| procedure handles logging a script
-% evaluation on the status line. It takes one argument, which is the
-% script to evaluate, and returns a boolean value that specifies whether
-% an error occurred (|0| is no error, |1| is that there was an error).
-%
-% Since the status line is rather limited in size there isn't very
-% much that can be logged on it. |dtxload::message_log| simply displays
-% the result or error that occurred.
-% \begin{tcl}
-proc dtxload::message_log {script} {
- set res [dtxload::evaluate $script]
- if {[lindex $res 0]==1} then {
- message "Tcl eval error: [lindex $res 1]"
- expr 1
- } else {
- if {[string length [lindex $res 1]] == 0} then {
- message "Tcl eval OK."
- } else {
- message "Tcl eval OK: [lindex $res 1]"
- }
- expr 0
- }
-}
-% \end{tcl}
-% \end{proc}
-%
-%
-% \begin{proc}{window_log}
-% \changes{v\,1.1}{2000/12/05}{Renamed procedure (old name was
-% \texttt{dtxload:\discretionary{}{}{}:shell\_log}). (LH)}
-% \begin{variable}{log_window_name}
-% \changes{v\,1.1}{2000/12/05}{Renamed variable (old name was
-% \texttt{dtxload:\discretionary{}{}{}:shell\_window\_name}). (LH)}
-% The |dtxload::window_log| procedure handles logging a script
-% evaluation in a window, which must be open when this procedure is
-% called. It takes one argument, which is the script to evaluate, and
-% returns a boolean value that specifies whether an error occurred
-% (|0| is no error, |1| is that there was an error).
-%
-% When it is set the |dtxload::log_window_name| variable holds the
-% name of the window which is used in logging, and when it isn't set
-% this means that logging should be in the status line. Initially it
-% is unset.
-% \begin{tcl}
-if {[info exists dtxload::log_window_name]} then {
- unset dtxload::log_window_name
-}
-% \end{tcl}
-% \end{variable}
-%
-% If the log window has mode |Shel| then the |script| is recorded in
-% the history list, but since the shells have trouble handling history
-% items that contain more than one line, all newlines in the recorded
-% script will be replaced by semicolons. This \emph{can} change the
-% meaning of the script, but hopefully that happens so rarely that one
-% can live with it.
-% \changes{v\,1.0}{2000/11/01}{Scripts added to the history list are
-% converted to one-liners. (LH)}
-% \begin{tcl}
-proc dtxload::window_log {script} {
- upvar #0 dtxload::log_window_name win
- global win::Modes Shel::startPrompt Shel::endPrompt
- if {[set win::Modes($win)] == "Shel"} then {
- global Shel::histnum
- regsub -all \n [string trimright $script] " ; " text
- history add $text
- set Shel::histnum [history nextid]
- }
-% \end{tcl}
-%
-% Next the script is inserted into (at the end of) the log window.
-% The lines on which logged scripts are written are marked out using
-% a special \package{dtxload} prompt, which is constructed as
-% \begin{quote}
-% |"${Shel::startPrompt}dtxload${Shel::endPrompt} "|
-% \end{quote}
-% If the log window mode is |Shel| and the last thing in the window
-% is the current prompt for the window, then that prompt is replaced
-% by the \package{dtxload} prompt, otherwise a newline followed by a
-% \package{dtxload} prompt will be appended to the window. The result
-% of a logged script is put on the line after the last line of the
-% script.
-% \begin{tcl}
- set pos [maxPos -w $win]
- select -w $win $pos $pos
- set dtxprompt "${Shel::startPrompt}dtxload${Shel::endPrompt} "
- if {[set win::Modes($win)] != "Shel"} then {
- insertText -w $win \r$dtxprompt
- } else {
- set stdprompt [Alpha::Prompt]
- set pos2 [pos::math $pos - [string length $stdprompt]]
- if {[getText -w $win $pos2 $pos] != $stdprompt} then {
- insertText -w $win \r$dtxprompt
- } else {
- replaceText -w $win $pos2 $pos $dtxprompt
- refresh
- }
- }
-% \end{tcl}
-% There currently is a bug (it is fixed in \Alpha~8) in
-% |replaceText| which causes it to always draw in the current window.
-% The |refresh| restores the proper window contents.
-% \changes{v\,1.0}{2000/10/16}{\texttt{refresh} inserted to overcome
-% \texttt{replaceText} bug. (LH)}
-%
-% Inserting the actual script text is basically an |insertText|, but
-% before the script is inserted all newlines in it (except the last)
-% are replaced by newlines followed by a \package{dtxload} prompt.
-% Since |dtxload::load| converts all newlines to linefeeds, it is
-% sufficient to replace linefeeds. It seems that newlines in \Alpha\
-% are always carriage returns (regardless of the window's ``platform''
-% attribute), and I assume that the \package{dtxload} prompt doesn't
-% contain any regexp metacharacter.
-% \changes{v\,1.0}{2000/10/16}{Assuming newlines in the script to log
-% are linefeeds. (LH)}
-% \begin{tcl}
- regsub -all \n [string trimright $script] \r$dtxprompt text
- insertText -w $win "$text\r"
-% \end{tcl}
-% After the script has been logged, it is time to evaluate it and log
-% the result.
-% \begin{tcl}
- set res [dtxload::evaluate $script]
- if {[lindex $res 0]==1} then {
- insertText -w $win "Error: [lindex $res 1]\r"
- if {[llength $res]>2} then {
- global errorInfo
- set errorInfo [lindex $res 2]
- }
- message "Tcl eval error"
- } else {
- if {[string length [lindex $res 1]]>0}\
- then {insertText -w $win "[lindex $res 1]\r"}
- message "Tcl eval OK"
- }
-% \end{tcl}
-% If the window has mode |Shel| then the window's current prompt is
-% inserted. It is necessary to call |Alpha::Prompt| again because
-% evaluating the script might have changed the prompt. Then the return
-% value is finally computed.
-% \begin{tcl}
- if {[set win::Modes($win)] == "Shel"}\
- then {insertText -w $win [Alpha::Prompt]}
- expr {[lindex $res 0]==1}
-}
-% \end{tcl}
-% \end{proc}
-%
-%
-%
-%
-% \subsection{Putting it all together}
-%
-% \begin{proc}{load}
-% The |dtxload::load| procedure implements the \package{dtxload} user
-% command---it extracts and evaluates the code in the current
-% selection (if something is selected) or the entire front window (if
-% nothing is selected). The evaluation is logged a window or on the
-% status line depending on the current value of
-% |dtxload::log_window_name|.
-%
-% Code is extracted and evaluated one command (or as close as one gets)
-% at a time. In case an error occurs the loading is stopped and the
-% insertion point is moved to the beginning of the problematic script,
-% whereas the mark is put at the end of the code to load. This way a
-% simple \textsf{Mark Hilite} will select the part of the code that
-% remains to be loaded.
-%
-% The very first thing |dtxload::load| does is to check whether it is
-% allowed to load any code in this window.
-% \begin{tcl}
-proc dtxload::load {} {
- global onlyLoadDtxFiles
- if {$onlyLoadDtxFiles && ![string match *.dtx [win::CurrentTail]]}\
- then {
- message "Not a .dtx file."
- return
- }
-% \end{tcl}
-% Then the interval from which code should be extracted is
-% determined.
-% \begin{tcl}
- set startpos [getPos]
- set endpos [selEnd]
- if {[pos::compare {$startpos == $endpos}]} then {
- set startpos [minPos]
- set endpos [maxPos]
- }
-% \end{tcl}
-% Next it is determined what logging method should be used. The name
-% of the procedure gets stored in the local variable |log|.
-% \begin{tcl}
- global dtxload::log_window_name
- if {
- [info exists dtxload::log_window_name] &&
- [lsearch -exact [winNames] ${dtxload::log_window_name}]!=-1
- } then {set log dtxload::window_log}\
- else {set log dtxload::message_log}
-% \end{tcl}
-% After that a couple of extraction-related variables are initialized.
-% The idea behind how the |dtxload::in_tcl_env| variable is set is that
-% if one is loading a short piece of code inside a \texttt{tcl}
-% environment then the first line probably doesn't start with a |%|,
-% whereas if one is loading an entire file (and thus starts in
-% non-\texttt{tcl}-environment) the first character quite likely is a
-% |%|.
-% \begin{tcl}
- global dtxload::in_tcl_env dtxload::module_stack\
- dtxload::next_module_idx dtxload::module_included
- set dtxload::in_tcl_env\
- [expr {[getText $startpos [pos::math {$startpos+1}]] != "%"}]
- set dtxload::module_stack {}
- set dtxload::next_module_idx 0
- set dtxload::module_included 1
-% \end{tcl}
-% The last piece of initialization concerns the |script| local
-% variable, in which the extracted lines are collected until they
-% make up an entire command. The |escaped| variable is a boolean
-% which is true when the newline at the end of the last line of
-% |script| was escaped. The |spos| variable stores the position at
-% which the command currently being collected began, but it is
-% initialized to |none| here to help distinguishing the case that no
-% extractable code was found.
-% \begin{tcl}
- set script ""
- set escaped 0
- set spos none
- while {[pos::compare $startpos < $endpos]} {
- set extract [dtxload::extract_line $startpos $endpos]
- if {[llength $extract]>1} then {
-% \end{tcl}
-% When a line of code has been extracted, it must be appended to
-% |script|. This isn't as easy as one might imagine, because \Tcllogo\
-% code being |source|d (which is what loading must function as) is fed
-% through a few more conversion steps than code being |eval|ed. In
-% particular all escaped newlines are converted to spaces and all
-% other newlines are converted to linefeeds, so that must be done
-% explicitly. A newline is considered to be escaped if is is preceded
-% by an odd number of backslashes, and if it is then the last backslash
-% and the newline are replaced by a single space. If the last line of
-% |script| ended by an escaped newline then the script is considered
-% incomplete and more code will be collected before evaluation is
-% attempted. Evaluation of |script| is also conditioned by that |info\
-% complete $script| returns |1|.
-% \changes{v\,1.0}{2000/10/16}{Converting nonescaped newlines to
-% linefeeds, since \Tcllogo\ doesn't consider carriage returns to
-% be command separators. (LH)}
-% \begin{tcl}
- if {[string length $script]==0} then {
- set spos [lindex $extract 2]
- }
- set line [lindex $extract 1]
- if {$escaped} then {set line [string trimleft $line]}
- set escaped [regsub -all\
- "((^|\[^\\\\\])(\\\\\\\\)*)\\\\(\r|\n|\r\n)$" $line {\1 } line]
- if {!$escaped} then {regsub "\r\n?\$" $line \n line}
- append script $line
- if {!$escaped && [info complete $script]} then {
- if {[$log $script]} then {
- select $spos $endpos
- markHilite
- return
- } else {
- set script ""
- }
- }
- } elseif {$spos=="none"} then {
- message "No extractable code found."
- }
- set startpos [lindex $extract 0]
- }
-% \end{tcl}
-% If the |script| variable isn't empty, then try to evaluate whatever
-% is in it before returning. If not even that causes an error the
-% |dtxload:|\penalty\hyphenpenalty|:known_expressions| array is unset
-% so that one can use other options for the next piece of code loaded.
-% \begin{tcl}
- if {[string length $script]>0 && [$log $script]} then {
- select $spos $endpos
- markHilite
- } else {
- global dtxload::known_expressions
- if {[info exists dtxload::known_expressions]}\
- then {unset dtxload::known_expressions}
- }
-}
-% \end{tcl}
-% \end{proc}
-%
-%
-% \subsection{Setting the log destination}
-%
-% \begin{proc}{logdest}
-% \changes{v\,1.0}{2000/10/16}{Procedure added. (LH)}
-% \changes{v\,1.1}{2000/12/07}{Procedure renamed from
-% \texttt{dtxload::redirect} and the syntax was changed. (LH)}
-% The log destination is set using the |dtxload::logdest| procedure.
-% The general syntax for this procedure is
-% \begin{quote}
-% |dtxload::logdest| \word{keyword} \word{window name}\regopt
-% \end{quote}
-% where the \word{keyword} (mainly) specifies the type of log
-% destination and \word{window name} is the name of the target
-% window (ignored for some \word{keyword}s).
-% \begin{tcl}
-proc dtxload::logdest {type {win "dtxload log"}} {
- global dtxload::log_window_name
- switch -exact -- $type {
-% \end{tcl}
-% The first form of |dtxload::logdest| has the syntax
-% \begin{quote}
-% |dtxload::logdest status|
-% \end{quote}
-% It makes the status line the log destination.
-% \begin{tcl}
- status {
- if {[info exists dtxload::log_window_name]}\
- then {unset dtxload::log_window_name}
- }
-% \end{tcl}
-% The second form has the syntax
-% \begin{quote}
-% |dtxload::logdest here|
-% \end{quote}
-% It makes the current window the log destination, regardless of what
-% may already be in it or which mode it has.
-% \begin{tcl}
- here {set dtxload::log_window_name [win::CurrentTail]}
-% \end{tcl}
-% The third form has the syntax
-% \begin{quote}
-% |dtxload::logdest window| \word{window name}
-% \end{quote}
-% It sets the log destination to the window named \word{window name},
-% regardless of what may already be in it or which mode it has. If
-% the window does not already exist it is created.
-% \begin{tcl}
- window {
- if {[lsearch -exact [winNames] $win] == -1} then {
- new -n $win -shell 1 -text "This is a dtxload log window.\r"
- }
- set dtxload::log_window_name $win
- }
-% \end{tcl}
-% The fourth and final form has the syntax
-% \begin{quote}
-% |dtxload::logdest shell| \word{window name}
-% \end{quote}
-% It calls |Shel::start| to open a \Tcllogo\ shell window and sets the
-% log destination to that window.
-% \begin{tcl}
- shell {
- Shel::start "Alpha" $win\
- "This is a dtxload log window (and Alpha Tcl shell).\r"
- set dtxload::log_window_name $win
- }
- }
-}
-%</tcl>
-% \end{tcl}
-% \end{proc}
-%
-%
-% \Finale
-%
-\endinput
diff --git a/Master/texmf-dist/source/latex/tcldoc/dtxload.ins b/Master/texmf-dist/source/latex/tcldoc/dtxload.ins
deleted file mode 100644
index 4e33fb87512..00000000000
--- a/Master/texmf-dist/source/latex/tcldoc/dtxload.ins
+++ /dev/null
@@ -1,54 +0,0 @@
-% dtxload.ins --- DOCSTRIP installation script for dtxload.
-\input docstrip
-
-% Redefine the \MetaPrefix; it should be something which starts a
-% until-end-of-line comment:
-\edef\MetaPrefix{\string#\string#}
-
-
-% Redefine the file preamble and postamble; this is necessary because
-% otherwise the old \metaPrefix is inserted at the beginning of these
-% lines.
-\preamble
-
-This file may be distributed and/or modified under the conditions
-of the LaTeX Project Public License, either version 1.2 of this
-license or (at your option) any later version. The latest version
-of this license is in
- http://www.latex-project.org/lppl.txt
-and version 1.2 or later is part of all distributions of LaTeX
-version 1999/12/01 or later.
-
-This file may only be distributed together with a copy of the source
-file from which it was generated. You may distribute that source file
-without this generated file.
-
-\endpreamble
-
-\postamble
-\endpostamble
-
-% Add a (auto-install) line at the top of the preamble, so that
-% Alpha will auto-install it.
-\edef\defaultpreamble{%
- \MetaPrefix\space (auto-install)^^J%
- \defaultpreamble
-}
-
-% Actually make dtxload.tcl:
-\generate{\file{dtxload.tcl}{\from{dtxload.dtx}{tcl}}}
-
-
-\Msg{}
-\Msg{**********************************************}
-\Msg{*}
-\Msg{* To complete installation,}
-\Msg{* have Alpha open the file dtxload.tcl}
-\Msg{* somewhere outside Alpha's folder hierarchy.}
-\Msg{*}
-\Msg{**********************************************}
-\Msg{}
-
-
-\end
-
diff --git a/Master/texmf-dist/source/latex/tcldoc/dtxload.tcl b/Master/texmf-dist/source/latex/tcldoc/dtxload.tcl
deleted file mode 100644
index af26546a774..00000000000
--- a/Master/texmf-dist/source/latex/tcldoc/dtxload.tcl
+++ /dev/null
@@ -1,361 +0,0 @@
-## (auto-install)
-##
-## This is file `dtxload.tcl',
-## generated with the docstrip utility.
-##
-## The original source files were:
-##
-## dtxload.dtx (with options: `tcl')
-##
-## This file may be distributed and/or modified under the conditions
-## of the LaTeX Project Public License, either version 1.2 of this
-## license or (at your option) any later version. The latest version
-## of this license is in
-## http://www.latex-project.org/lppl.txt
-## and version 1.2 or later is part of all distributions of LaTeX
-## version 1999/12/01 or later.
-##
-## This file may only be distributed together with a copy of the source
-## file from which it was generated. You may distribute that source file
-## without this generated file.
-##
-alpha::feature dtxload 1.1 {TeX} {
- # When set, the Load Tcl Code command will be disabled in
- # TeX files whose suffix is not dtx.
- newPref flag onlyLoadDtxFiles 1 TeX
- # When set, the Load Tcl Code command will only extract code
- # from code lines in tcl and tcl* environments.
- newPref flag onlyLoadTclEnvs 1 TeX
- # This is the key binding for dtxload's Load Tcl Code command.
- newPref binding loadTclCode <O/L TeX "" dtxload::load TeX
-} "" "" uninstall {this-file} \
- maintainer {"Lars Hellstr\x9am" Lars.Hellstrom@math.umu.se} \
- help {The dtxload package implements a command for loading
- Tcl code that is embedded in a .dtx file without docstripping
- the file first. For more help, LaTeX the file dtxload.dtx.}
-namespace eval dtxload {}
-proc dtxload::eval_guard {e} {
- global dtxload::known_expressions
- if {![info exists dtxload::known_expressions($e)]} then {
- set dtxload::known_expressions($e)\
- [expr {"yes" ==\
- "[askyesno "Should <$e> modules be included?"]"}]
- }
- set dtxload::known_expressions($e)
-}
-proc dtxload::push_module {e p} {
- global dtxload::module_stack
- lappend dtxload::module_stack [list $e $p]
-}
-proc dtxload::pop_module {e p} {
- global dtxload::module_stack dtxload::module_included\
- dtxload::next_module_idx
- set len [llength ${dtxload::module_stack}]
- if {$len==0} then {return}
- set L [lindex ${dtxload::module_stack} [expr {$len-1}]]
- if {[lindex $L 0] != $e} then {
- switch [buttonAlert "Module nesting error: <*[lindex $L 0]>\
- module ended by </$e>. For which guards should the positions\
- be pushed?" None Start End Both]\
- {
- Start {dtxload::push_bookmarks [lindex $L 1]}
- End {dtxload::push_bookmarks $p}
- Both {dtxload::push_bookmarks [lindex $L 1] $p}
- }
- }
- set dtxload::module_stack\
- [lreplace ${dtxload::module_stack} end end]
- if {$len>=${dtxload::next_module_idx}} then {
- incr dtxload::next_module_idx -1
- }
- if {[llength $L]>2} then {
- set dtxload::module_included [lindex $L 2]
- }
-}
-proc dtxload::push_bookmarks {args} {
- global markStack markName
- foreach pos $args {
- set name mark$markName
- incr markName
- createTMark $name $pos
- set fileName [win::Current]
- set markStack [linsert $markStack 0 [list $fileName $name]]
- }
-}
-proc dtxload::update_included {} {
- global dtxload::module_included dtxload::module_stack\
- dtxload::next_module_idx
- while {${dtxload::module_included} &&\
- ${dtxload::next_module_idx}<[llength ${dtxload::module_stack}]}\
- {
- set L [lindex ${dtxload::module_stack} ${dtxload::next_module_idx}]
- lappend L ${dtxload::module_included}
- set dtxload::module_stack [lreplace ${dtxload::module_stack}\
- ${dtxload::next_module_idx} ${dtxload::next_module_idx} $L]
- incr dtxload::next_module_idx
- set dtxload::module_included [dtxload::eval_guard [lindex $L 0]]
- }
-}
-proc dtxload::extract_line {startpos endpos} {
- global dtxload::in_tcl_env onlyLoadTclEnvs dtxload::module_included
- for {set pos $startpos} {[pos::compare {$pos<$endpos}]}\
- {set pos $pos2} {
- set pos2 [nextLineStart $pos]
- if {$pos2>$endpos} then {set pos2 $endpos}
- set line [getText $pos $pos2]
- switch -regexp -- $line {
- {^%<<} {
- if {!$onlyLoadTclEnvs || ${dtxload::in_tcl_env}} then {
- dtxload::update_included
- if {${dtxload::module_included}} then {
- alertnote "docstrip verbatim mode guard line\
- encountered. Verbatim mode extraction is\
- currently not supported, so I stop here."
- goto $pos
- return [list $endpos]
- }
- }
- set endtag %[string trimright [string trimright\
- [string range $line 3 end] \n\r] \ ]
- while {[string trimright [string trimright $line \n\r] \ ]\
- != $endtag} {
- set pos $pos2
- if {[pos::compare {$pos>=$endpos}]}\
- then {return [list $pos]}
- set pos2 [nextLineStart $pos]
- if {$pos2>$endpos} then {set pos2 $endpos}
- set line [getText $pos $pos2]
- }
- }
- {^%<[*/]} {
- if {![regexp {^%<(\*|/)([^>]+)>} $line foo modifier\
- expression]}\
- then {
- if {"[askyesno "Malformed guard line \"$line\"\
- encountered. Push position?"]"=="yes"}\
- then {dtxload::push_bookmarks $pos}
- } elseif {$modifier=="*"} then {
- dtxload::push_module $expression $pos
- } else {
- dtxload::pop_module $expression $pos
- }
- }
- {^%<} {
- if {![regexp {^%<(-|\+|)([^>]+)>(.*)$} $line\
- foo modifier expression code]}\
- then {
- if {"[askyesno "Malformed guard line \"$line\"\
- encountered. Push position?"]"=="yes"}\
- then {dtxload::push_bookmarks $pos}
- } elseif {!$onlyLoadTclEnvs || ${dtxload::in_tcl_env}} then {
- dtxload::update_included
- if {${dtxload::module_included}} then {
- if {[dtxload::eval_guard $expression] !=\
- ($modifier=="-")}\
- then {return [list $pos2 $code $pos]}
- }
- }
- }
- {^%} {
- while {[regexp -indices {\\(begin|end) *{tcl\*?}} $line\
- match type]} {
- set dtxload::in_tcl_env\
- [expr {"[eval string range \$line $type]"=="begin"}]
- set line [string range $line\
- [expr {[lindex $match 1]+1}] end]
- }
- }
- default {
- if {!$onlyLoadTclEnvs || ${dtxload::in_tcl_env}} then {
- dtxload::update_included
- if {${dtxload::module_included}}\
- then {return [list $pos2 $line $pos]}
- }
- }
- }
- }
- list $endpos
-}
-proc dtxload::evaluate {t} {
- global errorInfo errorCode evaluateRemotely tcl_platform
- if {!$evaluateRemotely} then {
- if {[set code [catch {uplevel \#0 $t} msg]] == 1} then {
- set L [split $errorInfo \n]
- set errorInfo [join [lrange $L 0 [expr {[llength $L] - 4}]] \n]
- }
- list $code $msg $errorInfo $errorCode
- } elseif {$tcl_platform(platform) == "macintosh"} then {
- global tclshSig
- app::ensureRunning $tclshSig
- list [catch {
- aebuild::result -t 30000 '${tclshSig}' misc dosc ----\
- [aebuild::TEXT $t]
- } res]\
- $res
- } else {
- global tclshInterp
- if {![info exists tclshInterp]} {
- if {[catch {tcltk::findTclshInterp}]} {
- return [list 1 "No shell selected"]
- }
- }
- if {$tcl_platform(platform) == "windows"} {
- if {[dde services Tk $tclshInterp] == ""} {
- alertnote "The remote shell has died, please select a new one."
- unset tclshInterp
- return [dtxload::evaluate $t]
- }
- list [catch {dde execute Tk $tclshInterp $t} res] $res
- } else {
- list [catch {send $tclshInterp $t} res] $res
- }
- }
-}
-proc dtxload::message_log {script} {
- set res [dtxload::evaluate $script]
- if {[lindex $res 0]==1} then {
- message "Tcl eval error: [lindex $res 1]"
- expr 1
- } else {
- if {[string length [lindex $res 1]] == 0} then {
- message "Tcl eval OK."
- } else {
- message "Tcl eval OK: [lindex $res 1]"
- }
- expr 0
- }
-}
-if {[info exists dtxload::log_window_name]} then {
- unset dtxload::log_window_name
-}
-proc dtxload::window_log {script} {
- upvar #0 dtxload::log_window_name win
- global win::Modes Shel::startPrompt Shel::endPrompt
- if {[set win::Modes($win)] == "Shel"} then {
- global Shel::histnum
- regsub -all \n [string trimright $script] " ; " text
- history add $text
- set Shel::histnum [history nextid]
- }
- set pos [maxPos -w $win]
- select -w $win $pos $pos
- set dtxprompt "${Shel::startPrompt}dtxload${Shel::endPrompt} "
- if {[set win::Modes($win)] != "Shel"} then {
- insertText -w $win \r$dtxprompt
- } else {
- set stdprompt [Alpha::Prompt]
- set pos2 [pos::math $pos - [string length $stdprompt]]
- if {[getText -w $win $pos2 $pos] != $stdprompt} then {
- insertText -w $win \r$dtxprompt
- } else {
- replaceText -w $win $pos2 $pos $dtxprompt
- refresh
- }
- }
- regsub -all \n [string trimright $script] \r$dtxprompt text
- insertText -w $win "$text\r"
- set res [dtxload::evaluate $script]
- if {[lindex $res 0]==1} then {
- insertText -w $win "Error: [lindex $res 1]\r"
- if {[llength $res]>2} then {
- global errorInfo
- set errorInfo [lindex $res 2]
- }
- message "Tcl eval error"
- } else {
- if {[string length [lindex $res 1]]>0}\
- then {insertText -w $win "[lindex $res 1]\r"}
- message "Tcl eval OK"
- }
- if {[set win::Modes($win)] == "Shel"}\
- then {insertText -w $win [Alpha::Prompt]}
- expr {[lindex $res 0]==1}
-}
-proc dtxload::load {} {
- global onlyLoadDtxFiles
- if {$onlyLoadDtxFiles && ![string match *.dtx [win::CurrentTail]]}\
- then {
- message "Not a .dtx file."
- return
- }
- set startpos [getPos]
- set endpos [selEnd]
- if {[pos::compare {$startpos == $endpos}]} then {
- set startpos [minPos]
- set endpos [maxPos]
- }
- global dtxload::log_window_name
- if {
- [info exists dtxload::log_window_name] &&
- [lsearch -exact [winNames] ${dtxload::log_window_name}]!=-1
- } then {set log dtxload::window_log}\
- else {set log dtxload::message_log}
- global dtxload::in_tcl_env dtxload::module_stack\
- dtxload::next_module_idx dtxload::module_included
- set dtxload::in_tcl_env\
- [expr {[getText $startpos [pos::math {$startpos+1}]] != "%"}]
- set dtxload::module_stack {}
- set dtxload::next_module_idx 0
- set dtxload::module_included 1
- set script ""
- set escaped 0
- set spos none
- while {[pos::compare $startpos < $endpos]} {
- set extract [dtxload::extract_line $startpos $endpos]
- if {[llength $extract]>1} then {
- if {[string length $script]==0} then {
- set spos [lindex $extract 2]
- }
- set line [lindex $extract 1]
- if {$escaped} then {set line [string trimleft $line]}
- set escaped [regsub -all\
- "((^|\[^\\\\\])(\\\\\\\\)*)\\\\(\r|\n|\r\n)$" $line {\1 } line]
- if {!$escaped} then {regsub "\r\n?\$" $line \n line}
- append script $line
- if {!$escaped && [info complete $script]} then {
- if {[$log $script]} then {
- select $spos $endpos
- markHilite
- return
- } else {
- set script ""
- }
- }
- } elseif {$spos=="none"} then {
- message "No extractable code found."
- }
- set startpos [lindex $extract 0]
- }
- if {[string length $script]>0 && [$log $script]} then {
- select $spos $endpos
- markHilite
- } else {
- global dtxload::known_expressions
- if {[info exists dtxload::known_expressions]}\
- then {unset dtxload::known_expressions}
- }
-}
-proc dtxload::logdest {type {win "dtxload log"}} {
- global dtxload::log_window_name
- switch -exact -- $type {
- status {
- if {[info exists dtxload::log_window_name]}\
- then {unset dtxload::log_window_name}
- }
- here {set dtxload::log_window_name [win::CurrentTail]}
- window {
- if {[lsearch -exact [winNames] $win] == -1} then {
- new -n $win -shell 1 -text "This is a dtxload log window.\r"
- }
- set dtxload::log_window_name $win
- }
- shell {
- Shel::start "Alpha" $win\
- "This is a dtxload log window (and Alpha Tcl shell).\r"
- set dtxload::log_window_name $win
- }
- }
-}
-##
-##
-## End of file `dtxload.tcl'.
diff --git a/Master/texmf-dist/source/latex/tcldoc/pstokens.dtx b/Master/texmf-dist/source/latex/tcldoc/pstokens.dtx
deleted file mode 100644
index 26f2a71e28b..00000000000
--- a/Master/texmf-dist/source/latex/tcldoc/pstokens.dtx
+++ /dev/null
@@ -1,979 +0,0 @@
-%
-% \iffalse
-%<*driver>
-\documentclass{tcldoc}
-
-\usepackage{array}
-
-
-% Logos:
-%
-\DeclareFontFamily{U}{logo}{}
-\DeclareFontShape{U}{logo}{m}{n}{
- <-9> logo8
- <9-10> logo9
- <10-> logo10
-}{}
-\DeclareFontShape{U}{logo}{bx}{n}{<-> logobf10}{}
-\DeclareFontShape{U}{logo}{b}{n}{<-> ssub * logo/bx/n}{}
-\DeclareFontShape{U}{logo}{m}{sl}{<-> logosl10}{}
-\DeclareFontShape{U}{logo}{m}{it}{<-> ssub * logo/m/sl}{}
-
-% \DeclareRobustCommand\META{{%
-% \fontencoding{U}\fontfamily{logo}\selectfont META%
-% }}
-\DeclareRobustCommand\MF{%
- {\fontencoding{U}\fontfamily{logo}\selectfont META}\-%
- {\fontencoding{U}\fontfamily{logo}\selectfont FONT}%
-}
-\DeclareRobustCommand\MP{%
- {\fontencoding{U}\fontfamily{logo}\selectfont META}\-%
- {\fontencoding{U}\fontfamily{logo}\selectfont POST}%
-}
-\DeclareRobustCommand\PS{PostScript}
-\DeclareRobustCommand\TOne{Type~1}
-\DeclareRobustCommand\package[1]{\textsf{#1}}
-
-
-% Useful environments:
-%
-% \begin{environment}{isyntax}
-% |isyntax| is for ``informal'' syntax specifications, such as
-% \begin{quote}
-% A \meta{varchar command} can be one of\\
-% \begin{quote}
-% |\vartop|\marg{glyph}\\
-% |\varmid|\marg{glyph}\\
-% |\varbot|\marg{glyph}\\
-% |\varrep|\marg{glyph}
-% \end{quote}
-% \end{quote}
-% The implementation is based on that of the \texttt{verse}
-% environment.It usually works just as fine with \texttt{quote}, but
-% \texttt{isyntax} handles breaking long lines better (and more
-% elegantly).
-% \changes{1.37}{2000/05/14}{Increased \cs{rightskip} stretchability
-% and added setting of \cs{linepenalty}. (LH)}
-% \begin{macrocode}
-\makeatletter
-\newenvironment{isyntax}{%
- \let\\\@centercr
- \list{}{%
- \itemsep \z@
- \itemindent -1.5em%
- \listparindent \itemindent
- \advance \leftmargin 1.5em%
- }%
- \advance \rightskip \z@\@plus0.7\linewidth \relax
- \linepenalty=100\relax
- \item\relax
-}{\endlist}
-\makeatother
-% \end{macrocode}
-% \end{environment}
-
-
-% Tabular notes:
-\makeatletter
-\newcommand\tabnotemark[1]{%
- \leavevmode
- \ifhmode\edef\@x@sf{\the\spacefactor}\nobreak\fi
- \hbox{\@textsuperscript{%
- \normalfont\itshape
- \count@=#1\relax\@alph\count@
- }}%
- \ifhmode\spacefactor\@x@sf\fi
- \relax
-}
-\newenvironment{tabnotes}{%
- \list{\the\c@mpfootnote}{%
- \usecounter{mpfootnote}%
- \def\makelabel##1{\hfil\tabnotemark{##1}}%
- \normalfont\footnotesize
- \setlength\leftmargin{0pt}%
- \setlength\labelsep{0pt}%
- \setlength\labelwidth{1.8em}%
- \setlength\itemindent{\labelwidth}%
- \setlength\listparindent{1em}%
- \setlength\parsep{0pt plus 1pt}%
- \setlength\itemsep{\parsep}%
- \setlength\topsep{\footnotesep}%
- }%
-}{\endlist}
-\makeatother
-
-
-\CodelineIndex
-
-\title{\texttt{pstokens.dtx}}
-\author{Lars Hellstr\"om}
-\date{Work in progress}
-\begin{document}
-\maketitle
-
-\DocInput{pstokens.dtx}
-
-\end{document}
-%</driver>
-% \fi
-%
-% \begin{abstract}
-% \package{pstokens} is a collection of utility routines for
-% \Tcllogo\ programs that need to view a file as a sequence of
-% \PS\ tokens rather than as a sequence of characters or lines.
-% \end{abstract}
-%
-%
-% \setnamespace{pstokens}
-%
-% First of all, the \texttt{pstokens} namespace must be defined.
-% \begin{tcl}
-namespace eval pstokens {}
-% \end{tcl}
-%
-% \section{Reading and writing tokens}
-%
-% This section contains the routines for reading a text file as a
-% sequence of tokens (tokenizing) and writing a sequence of tokens
-% to a text file (detokenizing).
-%
-%
-% \subsection{Data structures for tokens}
-%
-% A \emph{token} is a \Tcllogo\ list with the format
-% \begin{quote}
-% \word{type} \word{data}\regstar
-% \end{quote}
-% The interpretation of the \word{data} depends on what the \word{type}
-% is, as specified below. Some of the token types only occur internally
-% in the tokenizer and some others have no meaning to a \PS\ interpreter,
-% but this organisation seems to be the most convenient here.
-%
-% A primary goal for the format of tokens is that if a sequence of
-% tokens are read from one file and written to another then the only
-% things that may have changed should be the whitespace between the
-% tokens, not the tokens themselves. Therefore the tokenizer generally
-% avoids interpreting the tokens any more than necessary;
-% interpretation of tokens is instead handled in the next section.
-%
-% The token types that occur in \package{pstokens} are
-% \begin{description}
-% \item[\texttt{number}, \texttt{radixnumber}]
-% \describestring[token type]{number}
-% \describestring[token type]{radixnumber}
-% A number (integer or real), and radix number, respectively. There
-% is one \word{data} item, which contains the string that
-% represented the number in the tokenized text.
-% \item[\texttt{string}, \texttt{hexstring}, \texttt{base-85}]
-% \describestring[token type]{string}
-% \describestring[token type]{hexstring}
-% \describestring[token type]{base-85}
-% A parenthesis-delimited string, hexadecimal string, and ASCII
-% base-85 string, respectively. There is one \word{data} item,
-% which contains the string (minus delimiters) that represented
-% the string in the tokenized text.
-% \item[\texttt{executable}, \texttt{literal}, \texttt{immediate}]
-% \describestring[token type]{executable}
-% \describestring[token type]{literal}
-% \describestring[token type]{immediate}
-% An executable (normal), literal, and immediately evaluated name
-% respectively. There is one \word{data} item, which simply contains
-% the name as a \Tcllogo\ string, not including any initial slashes.
-% \item[\texttt{procedure}]
-% \describestring[token type]{procedure}
-% An executionable array. The data is simply the sequence of tokens
-% in the array.
-% \item[\texttt{comment}]
-% \describestring[token type]{comment}
-% This is not a \PS\ token at all, but a representation of a
-% comment. The reason comments are tokenized is that they often
-% contain interesting information (especially structured comments).
-% The structure of this kind of token is
-% \begin{quote}
-% |comment| \word{head} \word{tail}
-% \end{quote}
-% The \word{head} is the part of the line which is after the |%|
-% that started the comment and before the first whitespace character
-% after it. The \word{tail} is the rest of the line (including the
-% whitespace character, if there is one). Any of \word{head} and
-% \word{tail} may be empty. The structured comment
-% \begin{quote}
-% |%%Pages: 2|
-% \end{quote}
-% would have \word{head} equal to \verb*"%Pages:" and \word{tail}
-% equal to \verb*" 2".
-% \item[\texttt{eof}]
-% \describestring[token type]{eof}
-% This is a pseudotoken which denotes that the end of the file was
-% reached while looking for the next token. There structure is
-% \begin{quote}
-% |eof| \word{token}\regopt
-% \end{quote}
-% If the \word{token} is present then it is a composite token which
-% was being read when the file unexpectedly ended. It is an error
-% if the \word{token} is present.
-% \item[\texttt{special}]
-% \describestring[token type]{special}
-% This token type can be used to extend the token type syntax
-% without having to change the actual routines for reading and
-% writing tokens. The structure of a |special| token is
-% \begin{quote}
-% |special| \word{token} \word{extra data}\regstar
-% \end{quote}
-% If a |special| token gets sent to the token writer then it will be
-% treated as the token \word{token} (which usually has another
-% type), but other routines may treat it differently based on the
-% information in the \word{extra data}.
-% \item[\texttt{error}]
-% \describestring[token type]{error}
-% A token of this type is returned from the tokenizer whenever it
-% encounters anything which is syntactically incorrect. The
-% general structure of this kind of token is
-% \begin{quote}
-% |error| \word{error type} \word{data}\regstar
-% \end{quote}
-% where the \word{error type} more precisely specifies the error.
-% So far, the following error types are used
-% \begin{quote}
-% |error| |extra| \word{string} \word{line no.}\\
-% |error| |missing| \word{string} \word{line no.}\\
-% |error| |missing/extra| \word{m-string} \word{e-string}
-% \word{line no.}
-% \end{quote}
-% |extra| means that there is an occurence of \word{string} which
-% shouldn't have been there, whereas |missing| means that there
-% should have been a \word{string} there, but isn't. |missing/extra|
-% is the combination that there either should have been an
-% \word{m-string} there, but isn't, or there is an \word{e-string}
-% there, but there shouldn't be.
-% \word{line no.} is the line on which the error was encountered.
-% \end{description}
-%
-%
-% \subsection{Reading tokens}
-%
-% \begin{arrayvar}{input_line}[file-id]
-% \begin{arrayvar}{input_line_no}[file-id]
-% Since the |gets| procedure always reads an entire line (which
-% usually contain several tokens) each time it is used, it is
-% necessary to store the most recently read line somewhere. This is
-% the purpose of the |input_line| array, which is indexed by the file
-% id. In general, text is removed from the entires in |input_line| as
-% it is tokenized.
-%
-% The |input_line_no| array stores the number of the line in the
-% corresponding |input_line| entry. The first line in a file has
-% number one.
-% \end{arrayvar}\end{arrayvar}
-%
-% \begin{proc}{open_input}
-% \begin{proc}{close_input}
-% These procedures are wrappers around \Tcllogo's |open| and |close|
-% procedures respectively which additionally initialize and unset the
-% relevant entry of the |input_line| and |input_line_no| arrays. Thus
-% their syntaxes are
-% \begin{quote}
-% |pstokens::open_input| \word{file-name} \word{access}\regopt\
-% \word{permissions}\regopt\\
-% |pstokens::close_input| \word{file-id}
-% \end{quote}
-% and they return the same things as |open| and |close| respectively.
-%
-% A possible development of tokenizer is to implement filtered input,
-% in which case more support data structures would have to be
-% initialized, but then the |open_input| and |close_input| procedures
-% would be redefined to cope with this.
-% \begin{tcl}
-%<*tokenio>
-proc pstokens::open_input {args} {
- set res [eval open $args]
- global pstokens::input_line pstokens::input_line_no
- set pstokens::input_line($res) ""
- set pstokens::input_line_no($res) 0
- return $res
-}
-% \medskip
-proc pstokens::close_input {file_id} {
- global pstokens::input_line pstokens::input_line_no
- unset pstokens::input_line($file_id)
- unset pstokens::input_line_no($file_id)
- close $file_id
-}
-% \end{tcl}
-% \end{proc}\end{proc}
-%
-% The tokenizer has a couple of extra internal token types for things
-% which later become parts of other tokens.
-% \begin{description}
-% \item[\texttt{beginproc}, \texttt{endproc}]
-% \describestring[token type]{beginproc}
-% \describestring[token type]{endproc}
-% These are the |{| and |}| at the beginning and end respectively of
-% a \PS\ procedure. There is no \word{data}.
-% \item[\texttt{beginstring}, \texttt{beginhex}, \texttt{beginbase85}]
-% \describestring[token type]{beginstring}
-% \describestring[token type]{beginhex}
-% \describestring[token type]{beginbase85}
-% These token types are the |(|, |<|, and |<~| at the beginning of
-% a \PS\ literal, hex, and base-85 string respectively. There is no
-% \word{data}.
-% \end{description}
-%
-%
-% \begin{proc}{get_primitive_token}
-% The |get_primitive_token| procedure returns the next primitive
-% token in the input file. It takes the file id of the source to read
-% from as its only argument.
-%
-% A \emph{primitive token} is a smallest syntactic unit in the \PS\
-% file. Smallest syntactic units are characterized by that the
-% insertion of a space anywhere in them changes its meaning. One
-% exception is however the left parenthesis characters; it is
-% convenient to include |(| among the primitive tokens even though
-% inserting a space after it will change the meaning of
-% the corresponding token.
-%
-% The first step in |get_primitive_token| is to read ahead until some
-% non-whitespace (or the end of the file) is encountered.
-% \begin{tcl}
-proc pstokens::get_primitive_token {file} {
- global pstokens::input_line pstokens::input_line_no
- set pstokens::input_line($file)\
- [string trimleft [set pstokens::input_line($file)] \t\n\f\r\ ]
- while {[string length [set pstokens::input_line($file)]]==0} {
- if {[gets $file pstokens::input_line($file)]==-1} then {
- return {eof}
- }
- incr pstokens::input_line_no($file)
- set pstokens::input_line($file)\
- [string trimleft [set pstokens::input_line($file)] \t\n\f\r\ ]
- }
-% \end{tcl}
-% The second step is to isolate the part of the text that constitutes
-% the next token. This is done using regular expressions, but as these
-% get rather complicated it seems best to explain how they are
-% constructed. The ASCII characters null (|\x00|), tab (|\x09|), line
-% feed (|\x0a|), form feed (|\x0c|), carriage return (|\x0d|), and
-% space (|\x20|) count as whitespace and delimit tokens, but null is
-% currently ignored because \Tcllogo\ before v\,8 cannot keep them in
-% strings anyway. The ASCII characters |(|, |)|, |<|, |>|, |[|, |]|,
-% |{|, |}|, |/|, and |%| also delimit tokens, and therefore the
-% regular expression for a \PS\ name is
-% \begin{quote}
-% |(/?/?)([^][()<>{}/%|\meta{tab}\meta{LF}\meta{FF}\meta{CR}^^A
-% \verb*" ]+)|//?"
-% \end{quote}
-% (Note that the first right bracket above does not end a character
-% range, but is part of the list of characters which are exculded
-% from it. Also note that |/| and |//| are valid \PS\ tokens.)
-% The number of initial slashes determine whether the name is
-% \texttt{executable}, \texttt{literal}, or \texttt{immediate}.
-%
-% The executable tokens are however somewhat tricky, because the above
-% regular expression will also match all numbers. An \emph{integer} is
-% any token which matches the regular expression \verb"[+-]?[0-9]+",
-% whereas a \emph{real number} is anything else which matches the
-% regular expression
-% \begin{quote}
-% \verb"[+-]?([0-9]+(\.[0-9]*)?|\.[0-9]+)([eE][+-]?[0-9]+)?"
-% \end{quote}
-% and a \emph{radix number} is anything which matches the regular
-% expression\footnote{Actually there is also the restriction that
-% none of the characters after the \texttt{\#} may represent a digit
-% greater than or equal to the decimal number before the \texttt{\#},
-% but I'm ignoring that, for now.}
-% \begin{quote}
-% |[0-9][0-9]?#[0-9A-Za-z]+|
-% \end{quote}
-% \begin{tcl}
- if {[regexp "^(/?/?)(\[^\]\[()<>\{\}/%\t\n\f\r\ \]+)|^//?"\
- [set pstokens::input_line($file)] tokenstr slashes name]}\
- then {
- if {[regexp\
- {^[+-]?([0-9]+(\.[0-9]*)?|\.[0-9]+)([eE][+-]?[0-9]+)?$}\
- $tokenstr]}\
- then {
- set token [list number $tokenstr]
- } elseif {[regexp {^[0-9][0-9]?#[0-9A-Za-z]+$} $tokenstr]} then {
- set token [list radixnumber $tokenstr]
- } else {
- if {"$tokenstr"=="/" || "$tokenstr"=="//"}\
- then {set slashes $tokenstr}
- set token [list [lindex {executable literal immediate}\
- [string length $slashes]] $name]
- }
- }\
-% \end{tcl}
-% There are also a couple of primitive tokens which consist of
-% non-whitespace delimiting characters, namely
-% \begin{quote}
-% |{|, |}|, |<|, |<~|, |<<|, |>>|, |[|, |]|, |(|,
-% \end{quote}
-% The right parenthesis and greater than characters does not appear in
-% the list because they may not appear outside a string. Finally
-% there are the comments, which have the regular expression
-% \begin{quote}
-% |^%([^|\meta{tab}\meta{LF}\meta{FF}\meta{CR}\verb*" ]*)(.*)$"
-% \end{quote}
-% Here the first parenthesis is the head of the comment and the second
-% is the tail.
-% \begin{tcl}
- else {
- switch -regexp -- [set pstokens::input_line($file)] {
- ^\{ {set token beginproc ; set tokenstr \{}
- ^\} {set token endproc ; set tokenstr \}}
- ^\\\[ {set token {executable [} ; set tokenstr \[}
- ^\\\] {set token {executable ]} ; set tokenstr \]}
- ^< {set token beginhex ; set tokenstr <}
- ^<~ {set token beginbase85 ; set tokenstr <~}
- ^<< {set token {executable <<} ; set tokenstr <<}
- ^>> {set token {executable >>} ; set tokenstr >>}
- ^\\( {set token beginstring ; set tokenstr (}
- ^% {
- regexp "^%(\[^\t\n\f\r\ \]*)(.*)\$"\
- [set pstokens::input_line($file)] tokenstr head tail
- set token [list comment $head $tail]
- }
- default {
-% \end{tcl}
-% But if none of the above matched the next character must be
-% erroneous.
-% \begin{tcl}
- set tokenstr\
- [string index [set pstokens::input_line($file)] 0]
- set token [list error extra $tokenstr\
- [set pstokens::input_line_no($file)] ]
- }
- }
- }
- set pstokens::input_line($file)\
- [string range [set pstokens::input_line($file)]\
- [string length $tokenstr] end]
- return $token
-}
-% \end{tcl}
-%
-% \end{proc}
-%
-% \begin{proc}{get_string}
-% The |get_string| procedure reads a \PS\ literal string and converts
-% it to a \texttt{string} token, which it returns. The only argument
-% is the file id of the input file. It is assumed that the next thing
-% on the input file line is the character after the left parenthesis
-% that started the string.
-%
-% The only conversion of the data that |get_string| performs is that
-% escaped newlines are skipped and non-escaped newlines are replaced
-% by |\n| escapes. If the end-of-file is encountered while reading a
-% string then the necessary number of right parentheses are added to
-% the string and the resulting string token is built into an |eof|
-% token, which is returned.
-%
-% The local variable |level| keeps track of the parenthesis nesting
-% level, the local variable |data| is the string read so far, and the
-% local variable is a flag which is |1| if the next character should
-% be escaped and |0| otherwise.
-% \begin{tcl}
-proc pstokens::get_string {file} {
- upvar #0 pstokens::input_line line
- global pstokens::input_line_no
- set level 1
- set data ""
- set escaped 0
- while {$level>0} {
- while {$level>0 && [string length $line($file)]>0} {
- set c [string index $line($file) 0]
- if {$escaped} then {
- append data $c
- set escaped 0
- } else {
- switch -exact -- $c {
- ( {incr level ; append data $c}
- ) {
- incr level -1
- if {$level>0} then {append data $c}
- }
- \\ {set escaped 1 ; append data $c}
- default {append data $c}
- }
- }
- set line($file) [string range $line($file) 1 end]
- }
- if {$level>0} then {
- if {!$escaped} then {append data \\n}
- set escaped 0
- if {[gets $file line($file)]==-1} then {
- while {$level>1} {append data ) ; incr level -1}
- return [list eof [list string $data]]
- }
- incr pstokens::input_line_no($file)
- }
- }
- list string $data
-}
-% \end{tcl}
-% \end{proc}
-%
-%
-% \begin{proc}{get_hexstring}
-% The |get_hexstring| procedure reads a \PS\ hexadecimal string and
-% converts it to a \texttt{hexstring} token, which it returns. The
-% only argument is the file id of the input file. It is assumed that
-% the next thing on the input file line is the character after the
-% less than character that started the string.
-%
-% The only conversion of the data that |get_hexstring| performs is that
-% embedded whitespace is removed. If a character which is neither
-% whitespace, a hexadecimal digit, or the closing |>| is encountered
-% then an \texttt{error} token is returned and the text read is put
-% back into the |input_line| buffert (followed by an |>| which is
-% inserted to terminate the string. If the end-of-file is
-% encountered while reading a string then what has been read so far
-% is returned as a hexstring token built into an |eof| token.
-% \begin{tcl}
-proc pstokens::get_hexstring {file} {
- upvar #0 pstokens::input_line line
- global pstokens::input_line_no
- set data ""
- while 1 {
- while {[string length $line($file)]==0} {
- if {[gets $file line($file)]==-1} then {
- return [list eof [list hexstring $data]]
- }
- incr pstokens::input_line_no($file)
- }
- set c [string index $line($file) 0]
- set line($file) [string range $line($file) 1 end]
- switch -regexp -- $c {
- {[0-9a-fA-F]} {append data $c}
- \[\t\n\f\r\ \] {}
- > {break}
- default {
- set line($file) <$data>$c$line($file)
- return [list error missing/extra > $c\
- [set pstokens::input_line_no($file)]]
- }
- }
- }
- list hexstring $data
-}
-% \end{tcl}
-% \end{proc}
-%
-%
-% \begin{proc}{get_base85string}
-% The |get_base85string| procedure reads a \PS\ base-85 string and
-% converts it to a \texttt{base-85} token, which it returns. The
-% only argument is the file id of the input file. It is assumed that
-% the next thing on the input file line is the character after the
-% |<~| characters that started the string.
-%
-% The only conversion of the data that |get_base85string| performs is
-% that embedded whitespace is removed. If a character which is neither
-% whitespace, a base-85 digit, the letter |z|, or the closing |~| is
-% encountered then an \texttt{error} token is returned and the text
-% read is put back into the |input_line| buffert (followed by a |~>|
-% that terminates the string). If the end-of-file is encountered while
-% reading a string then what has been read so far is returned as a
-% base-85 string token built into an |eof| token.
-% \begin{tcl}
-proc pstokens::get_base85string {file} {
- upvar #0 pstokens::input_line line pstokens::input_line_no line_no
- set data ""
- while 1 {
- while {[string length $line($file)]==0} {
- if {[gets $file line($file)]==-1} then {
- return [list eof [list base-85 $data]]
- }
- incr line_no($file)
- }
- set c [string index $line($file) 0]
- set line($file) [string range $line($file) 1 end]
- switch -regexp -- $c {
- {[!-uz]} {append data $c}
- \[\t\n\f\r\ \] {}
- ~ {break}
- default {
- set line($file) <~$data~>$c$line($file)
- return [list error missing/extra ~> $c $line_no($file)]
- }
- }
- }
-% \end{tcl}
-% If the character after the closing |~| is not a |>| then the entire
-% string is put back in the |input_line| buffert and a |missing| error
-% is issued.
-% \begin{tcl}
- if {[string length $line($file)]>0 &&\
- "[string index $line($file) 0]"==">"} then {
- set line($file) [string range $line($file) 1 end]
- list base-85 $data
- } else {
- set line($file) <~$data~>$line($file)
- list error missing > $line_no($file)
- }
-}
-% \end{tcl}
-% \end{proc}
-%
-%
-% \begin{proc}{get_token}
-% \begin{proc}{get_token_rek}
-% The |get_token| procedure has the syntax
-% \begin{quote}
-% |pstokens::get_token| \word{file-id}
-% \end{quote}
-% and returns the next \PS\ token in the \word{file-id} file.
-%
-% Most of the work is however done by the |get_token_rek| procedure,
-% which is a recursive form of |get_token|. |get_token_rek| handles
-% converting the various |begin|\textellipsis\ token types to their
-% complete forms, and it calls itself recursively to get the tokens
-% that form a procedure. |get_token| only checks that there are no
-% lone |endproc| tokens.
-% \begin{tcl}
-proc pstokens::get_token {file} {
- set token [pstokens::get_token_rek $file]
- if {"[lindex $token 0]" == "endproc"} then {
- global pstokens::input_line_no
- set token [list error extra \} [set pstokens::input_line_no($file)]]
- }
- return $token
-}
-% \medskip
-proc pstokens::get_token_rek {file} {
- set token [pstokens::get_primitive_token $file]
- switch -exact [lindex $token 0] {
- beginproc {
- set token {procedure}
- while {"[lindex [set token2 [pstokens::get_token_rek $file]] 0]"\
- != "endproc" && "[lindex $token2 0]" != "eof"} {
- lappend token $token2
- }
- if {"[lindex $token2 0]" == "eof"} then {
- if {[llength $token2]>1} then {
- lappend token [lindex $token2 1]
- }
- set token [list eof $token]
- }
- }
- beginstring {set token [pstokens::get_string $file]}
- beginhex {set token [pstokens::get_hexstring $file]}
- beginbase85 {set token [pstokens::get_base85string $file]}
- }
- return $token
-}
-% \end{tcl}
-%
-% These procedures, and in particular |get_token_rek|, could be made
-% more intelligent when it comes to handling errors that occur inside
-% procedures. Currently they just embed the |error| token in the
-% procedure.
-% \end{proc}\end{proc}
-%
-%
-% \subsection{Writing tokens}
-%
-% Writing tokens is simpler than reading them; the only real
-% complication is that the length of lines must be bounded. In order to
-% achieve this, data is written linewise to the file (much like the way
-% it is read).
-%
-% \begin{arrayvar}{output_line}[file-id]
-% The |output_line| array stores each output line until it is
-% actually written to file. Like with the |input_line| array, each
-% active output file has its own entry in this array.
-% \end{arrayvar}
-%
-% \begin{proc}{flush_line}
-% The |flush_line| procedure has the syntax
-% \begin{quote}
-% |pstokens::flush_line| \word{file-id} \word{physical}\regopt
-% \end{quote}
-% It flushes buffered data that is to be written to the file with id
-% \word{file-id}. The \word{physical} argument controls whether the
-% data actually should be physically written to file (when |1|) or
-% merely flushed from the |output_line| array (when |0|). The default
-% is |0|. Calling |flush_line| starts a new line in the output file
-% unless the |output_line| entry is empty.
-% \begin{tcl}
-proc pstokens::flush_line {fileid {physical 0}} {
- global pstokens::output_line
- if {[string length [set pstokens::output_line($fileid)]]>0} then {
- puts $fileid [set pstokens::output_line($fileid)]
- set pstokens::output_line($fileid) ""
- }
- if {$physical} then {flush $fileid}
-}
-% \end{tcl}
-% \end{proc}
-%
-% \begin{proc}{open_output}
-% \begin{proc}{close_output}
-% These procedures are wrappers around \Tcllogo's |open| and |close|
-% procedures respectively which additionally initialize and unset the
-% relevant entry of the |output_line|. Thus their syntaxes are
-% \begin{quote}
-% |pstokens::open_output| \word{file-name} \word{access}\regopt\
-% \word{permissions}\regopt\\
-% |pstokens::close_output| \word{file-id}
-% \end{quote}
-% and they return the same things as |open| and |close| respectively.
-% Note however that unlike the case with |open|, the \word{access}
-% argument of |open_output| defaults to |w|.
-%
-% A possible development of the routines for writing tokens would be
-% to support output channels other than files. In that case more
-% support data structures would have to be initialized, but then the
-% |open_output| and |close_output| procedures would be redefined to
-% cope with this.
-% \begin{tcl}
-proc pstokens::open_output {name {access w} args} {
- set res [eval open \$name \$access $args]
- global pstokens::output_line
- set pstokens::output_line($res) ""
- return $res
-}
-% \medskip
-proc pstokens::close_output {file_id} {
- pstokens::flush_line $file_id 1
- close $file_id
- global pstokens::output_line
- unset pstokens::output_line($file_id)
-}
-% \end{tcl}
-% \end{proc}\end{proc}
-%
-% \begin{variable}{wrap_length}
-% The |wrap_length| variable stores the maximal number of characters
-% that may appear on a single output line. The same value applies for
-% all output files.
-% \begin{tcl}
-set pstokens::wrap_length 72
-% \end{tcl}
-% \end{variable}
-%
-% \begin{proc}{put_word}
-% The |put_word| procedure has the syntax
-% \begin{quote}
-% |pstokens::put_word| \word{file-id} \word{string}
-% \end{quote}
-% It writes \meta{string} to the file with id \word{file-id},
-% preceeded by a whitespace character if the current line is
-% nonempty. The whitespace character is usually a space, but it
-% will be a newline if the length of what has already been written
-% to the current line is too large to also fit the \word{string}
-% within the length specified by the |wrap_length| variable.
-% \begin{tcl}
-proc pstokens::put_word {file str} {
- global pstokens::output_line pstokens::wrap_length
- set l [string length [set pstokens::output_line($file)]]
- if {$l==0} then {
- set pstokens::output_line($file) $str
- } elseif {$l + [string length $str] < ${pstokens::wrap_length}}\
- then {
- append pstokens::output_line($file) " $str"
- } else {
- puts $file [set pstokens::output_line($file)]
- set pstokens::output_line($file) $str
- }
-}
-% \end{tcl}
-% \end{proc}
-%
-% \begin{proc}{put_breakable}
-% The |put_breakable| procedure has the syntax
-% \begin{quote}
-% |pstokens::put_breakable| \word{file-id}
-% \word{prefix} \word{text} \word{suffix}
-% \end{quote}
-% It writes \meta{prefix}\meta{text}\meta{suffix} to the file with
-% id \word{file-id} and inserts whitspace (newlines) into the
-% \meta{text} in such a way that the |wrap_length| linewidth isn't
-% exceeded. This is primarily useful for hexadecimal and base-85
-% strings.
-% \begin{tcl}
-proc pstokens::put_breakable {file prefix text suffix} {
- upvar #0 pstokens::output_line line pstokens::wrap_length wrap
- if {[string length $line($file)] + [string length $prefix] +\
- [string length $text] + [string length $suffix] < $wrap} then {
- append line($file) " $prefix$text$suffix"
- } else {
- pstokens::flush_line $file
- set line($file) $prefix
- while {[string length $line($file)] + [string length $text] +\
- [string length $suffix] < $wrap} {
- set t [expr {$wrap -[string length $line($file)]}]
- append line($file) [string range $text 0 [expr {$t-1}]]
- set text [string range $text $t end]
- pstokens::flush_line $file
- }
- append line($file) $text$suffix
- }
-}
-% \end{tcl}
-% \end{proc}
-%
-% \begin{proc}{put_string}
-% The |put_string| procedure has the syntax
-% \begin{quote}
-% |pstokens::put_string| \word{file-id} \word{text}
-% \end{quote}
-% It writes |(|\meta{text}|)| to the file with id \word{file-id} and
-% inserts backslash+newline items into the \meta{text} in such a way
-% that the |wrap_length| linewidth isn't exceeded and the
-% tokenization of the string is unaffected.
-%
-% The only places where it is unsafe to insert a backslash+newline is
-% inside an escape squence. A position can only be inside such a
-% sequence if the part of the \meta{text} which is before it gets a
-% match against the regular expression
-% \begin{quote}
-% \verb"(^|[^\])(\\\\)*(\\[0-7]?[0-7]?)$"
-% \end{quote}
-% and in this case the position between the second and third
-% parenthesis is safe.
-% \begin{tcl}
-proc pstokens::put_string {file text} {
- upvar #0 pstokens::output_line line pstokens::wrap_length wrap
- if {[string length $line($file)] + [string length $text] + 2 <\
- $wrap} then {
- append line($file) " ($text)"
- } else {
- pstokens::flush_line $file
- set line($file) (
- while {[string length $line($file)] + [string length $text] +\
- 1 < $wrap} {
- set t [expr {$wrap -[string length $line($file)]}]
- set s [string range $text 0 [expr {$t-1}]]
- if {[regexp {(^|[^\])(\\\\)*(\\[0-7]?[0-7]?)$} $s\
- foo bar baz escape]}\
- then {
- incr t -[string length $escape]
- set s [string range $text [expr {$t-1}]]
- }
- append line($file) $s
- set text [string range $text $t end]
- pstokens::flush_line $file
- }
- append line($file) $text)
- }
-}
-% \end{tcl}
-% \end{proc}
-%
-% \begin{proc}{put_token}
-% The |put_token| procedure has the syntax
-% \begin{quote}
-% |pstokens::put_token| \word{file-id} \word{token}
-% \end{quote}
-% It writes the \PS\ token \word{token} to the file with id
-% \word{file-id}.
-% \begin{tcl}
-proc pstokens::put_token {file token} {
- switch [lindex $token 0] {
- number -
- radixnumber -
- executable {pstokens::put_word $file [lindex $token 1]}
- literal {pstokens::put_word $file /[lindex $token 1]}
- immediate {pstokens::put_word $file //[lindex $token 1]}
- string {pstokens::put_string $file [lindex $token 1]}
- comment {
- pstokens::flush_line $file 0
- puts $file %[lindex $token 1][lindex $token 2]
- }
- special {pstoken::put_token $file [lindex $token 1]}
- procedure {
- pstokens::put_word $file \{
- foreach subtok [lrange $token 1 end] {
- pstokens::put_token $file $subtok
- }
- pstokens::put_word $file \}
- }
- hexstring {pstokens::put_breakable $file < [lindex $token 1] >}
- base-85 {pstokens::put_breakable $file <~ [lindex $token 1] ~>}
- eof {
- if {[llength $token]>1} then {
- pstokens::put_token $file [lindex $token 1]
- }
- pstokens::terminal_print "Token unexpectedly terminated by\
- end of file."
- }
- error {
- pstokens::terminal_print "Error token intercepted at put:\
- [switch [lindex $token 1] {
- extra {expr {"extra [lindex $token 2] at input line\
- [lindex $token 3]"}}
- missing {expr {"missing [lindex $token 2] at input\
- line [lindex $token 3]"}}
- missing/extra {expr {"missing [lindex $token 2], or\
- extra [lindex $token 3], at input line\
- [lindex $token 4]"}}
- default {expr {"!!!-unimplemented error-!!!"}}
- }]."
- }
- default {
- pstokens::terminal_print\
- "Token of unknown type encountered: $token"
- }
- }
-}
-% \end{tcl}
-% \end{proc}
-
-
-% \begin{tcl}
-%</tokenio>
-% \end{tcl}
-%
-% \section{Token stream editing}
-%
-% This section I still have ToDo.
-%
-%
-% \begin{thebibliography}{99}
-%
-% \bibitem{PSman}
-% Adobe Systems Incorporated:
-% \textit{\PS\ language reference manual, 3rd ed.},
-% Addison--Wesley, 1999; ISBN 0-201-37922-8;
-% http:/\slash\texttt{partners.adobe.com}\slash
-% \texttt{asn}\slash\texttt{developer}\slash\texttt{PDFS}\slash
-% \texttt{TN}\slash\texttt{PLRM.pdf}.
-%
-% \bibitem{MPinTUG}
-% John D.\ Hobby: \textit{A \MF-like System with \PS\ output},
-% TUGboat \textbf{10} (4) (1989), 505--512.
-% %^^A 2 or 4 ? Hobby's bibliography says 2, but Bebee's
-% %^^A tugboat.bib says 4.
-% %^^A @article{Hobby89a,
-% %^^A author = {John D. Hobby},
-% %^^A title = {A {METAFONT}-like System with PostScript Output},
-% %^^A journal = {{TUG}boat},
-% %^^A volume = {10},
-% %^^A number = {2},
-% %^^A pages = {505--512},
-% %^^A year = {1989}
-% %^^A }
-% \bibitem{MPman}
-% John D.\ Hobby: \textit{A User's Manual for \MP},
-% AT\&T Bell Laboratories Computing Science Technical Report no.~162
-% (1992);
-% \textit{see} http:/\slash \texttt{cm.bell-labs.com}\slash
-% \texttt{who}\slash\texttt{hobby}\slash\texttt{MetaPost.html}.
-% %^^A @techreport{Hobby92,
-% %^^A title = {A User's manual for {MetaPost}},
-% %^^A author = {John D. Hobby},
-% %^^A institution = {AT\&T Bell Laboratories},
-% %^^A address = {Murray Hill, New Jersey},
-% %^^A type = {Computing Science Technical Report},
-% %^^A number = {no.~162},
-% %^^A year = {1992}
-% %^^A }
-%
-% \end{thebibliography}
-%
-%
-% \Finale
-%
-\endinput
diff --git a/Master/texmf-dist/source/latex/tcldoc/tcldoc.dtx b/Master/texmf-dist/source/latex/tcldoc/tcldoc.dtx
deleted file mode 100644
index ce2114822b4..00000000000
--- a/Master/texmf-dist/source/latex/tcldoc/tcldoc.dtx
+++ /dev/null
@@ -1,3402 +0,0 @@
-% \CheckSum{1549}
-%
-% \title{The \package{tcldoc} package and class}
-% \author{Lars Hellstr\"om^^A
-% \thanks{E-mail: \texttt{Lars.Hellstrom@math.umu.se}}
-% \emph{et al.}}
-% \date{2001/01/04}
-% \maketitle
-%
-% \iffalse
-%
-%<class|pkg>\NeedsTeXFormat{LaTeX2e}
-%<class>\ProvidesClass{tcldoc}
-%<pkg>\ProvidesPackage{tcldoc}
-%<class|pkg> [2000/12/20 v2.13 Tcl documentation
-%<class> class]
-%<pkg> package]
-%
-%<*driver>
-\documentclass{ltxdoc}
-\usepackage{amsmath}
-\IfFileExists{xdoc2.sty}{\usepackage[dolayout]{xdoc2}[2000/11/18]}{}
-
-\makeatletter
-\@ifpackageloaded{xdoc2}{}{%
- \@ifpackagelater{doc}{2000/05/20}{}{%
- \let\XD@fragile@meta=\meta
- \def\meta{%
- \ifx \protect\@typeset@protect
- \expandafter\futurelet \expandafter\@let@token
- \expandafter\XD@fragile@meta
- \else
- \noexpand\meta
- \fi
- }%
- }%
-}
-\@ifundefined{option}{%
- \newenvironment{option}[1]{%
- \trivlist
- \if@inlabel\else \macro@cnt=\z@ \fi
- \item[]\ignorespaces
- }{\endtrivlist}%
-}{}
-\makeatother
-
-\DeclareRobustCommand\package[1]{\textsf{#1}}
-\DeclareRobustCommand\Tcl{T\kern-0.1em cl}
-\DeclareRobustCommand\word[1]{\mbox{$\{$\itshape#1\/$\}$}}
-
-\providecommand*\describeoption[1]{\marginpar{\texttt{#1} option}}
-\providecommand*\PrintChar[1]{\char#1\relax}
-\providecommand*{\DoNotIndexHarmless}[1]{}
-\providecommand*{\DoNotIndexBy}[1]{}
-\DoNotIndexBy{@}
-\DoNotIndexBy{@@}
-\DoNotIndexBy{TD@}
-\newenvironment{isyntax}{%
- \list{}{%
- \setlength\parsep{0pt plus 1pt}%
- \setlength\listparindent{-1em}%
- \setlength\itemindent{\listparindent}%
- \addtolength\leftmargin{1em}%
- }%
- \item\relax \rightskip=1.0\rightskip plus 5cm
-}{\endlist}
-
-\setcounter{IndexColumns}{2}
-\hfuzz=10pt
-\expandafter\def \expandafter\GlossaryParms \expandafter{%
- \GlossaryParms \hbadness=10000
-}
-
-\CodelineIndex
-% \EnableCrossrefs
-% \RecordChanges
-
-\begin{document}
-\DocInput{tcldoc.dtx}
-\PrintChanges
-\PrintIndex
-\end{document}
-%</driver>
-%
-% \fi
-%
-% \begin{abstract}
-% The \package{tcldoc} package defines a couple of environments and
-% macros for documenting \Tcl\ source code in \texttt{.dtx}-style
-% documented source files, much like what the \package{doc}
-% package~\cite{doc} does for \TeX\ source code. The \package{tcldoc}
-% class is analogous to the \package{ltxdoc} document
-% class~\cite{ltxdoc}---it loads the package to gain the basic
-% functionality and changes some layout parameters to values that
-% are better suited for documented source than those set by the
-% standard \package{article} document class.
-%
-% The \package{tcldoc} package builds on the \package{doc} and
-% \package{xdoc}~\cite{xdoc} packages.
-% \end{abstract}
-%
-% \tableofcontents
-%
-% \DoNotIndex{\advance,\begingroup,\bgroup,\big,\Big,\bigg,\Bigg,\box}
-% \DoNotIndex{\catcode,\char,\chardef,\cleaders,\color@begingroup}
-% \DoNotIndex{\color@endgroup,\copy,\csname,\DeclareOption}
-% \DoNotIndex{\DeclareRobustCommand,\DeclareTextCommand}
-% \DoNotIndex{\DeclareTextSymbol,\def,\discretionary,\divide,\do}
-% \DoNotIndex{\edef,\egroup,\else,\@empty,\endcsname,\endgroup}
-% \DoNotIndex{\expandafter,\fi,\@firstoftwo,\font,\fontdimen}
-% \DoNotIndex{\futurelet,\gdef,\global,\@gobble,\hb@xt@,\hbox}
-% \DoNotIndex{\hfill,\if,\ifcase,\ifcat,\ifmmode,\ifnum,\ifvbox}
-% \DoNotIndex{\ifvmode,\ifvoid,\ifx,\ignorespaces,\immediate,\indent}
-% \DoNotIndex{\input,\@@input,\InputIfFileExists,\IfFileExists}
-% \DoNotIndex{\item,\kern,\language,\lastbox,\lastpenalty,\lastskip}
-% \DoNotIndex{\lccode,\leavevmode,\let,\llap,\LoadClass,\lowercase}
-% \DoNotIndex{\m@ne,\m@th,\@makeother,\mathord,\mkern,\multiply}
-% \DoNotIndex{\@ne,\newbox,\newcommand,\newcount,\NewDescribeCommand}
-% \DoNotIndex{\newenvironment,\newif,\NewMacroEnvironment,\nfss@text}
-% \DoNotIndex{\noexpand,\normalfont,\null,\number,\or,\p@,\@@par}
-% \DoNotIndex{\parshape,\PassOptionsToClass,\PassOptionsToPackage}
-% \DoNotIndex{\penalty,\@plus,\predisplaypenalty,\prevdepth,\prevgraf}
-% \DoNotIndex{\ProcessOptions,\protect,\protected@edef}
-% \DoNotIndex{\providecommand,\ProvideTextCommandDefault,\relax}
-% \DoNotIndex{\RequirePackage,\RequirePackageWithOptions,\rightarrow}
-% \DoNotIndex{\sbox,\selectfont,\setbox,\skewchar,\space,\string}
-% \DoNotIndex{\TextSymbolUnavailable,\texttt,\the,\tracingparagraphs}
-% \DoNotIndex{\ttfamily,\tw@,\typeout,\unhbox,\unpenalty,\unskip}
-% \DoNotIndex{\unrestored@protected@xdef,\unskip,\unvbox,\vadjust}
-% \DoNotIndex{\vbox,\vert,\vrule,\vskip,\wd,\write,\xdef,\z@,\z@skip}
-%
-% \section{Introduction}
-% \label{Sec:Introduction}
-%
-% This introduction is meant to be comprehensible even for readers
-% unfamiliar with writing \texttt{.dtx} files and using the \package{doc}
-% package. Readers who are experienced with this will probably want to
-% skip right to the next section.
-%
-% A \texttt{.dtx} file has a dual nature. On one hand it is a container
-% for some lines of code---it could be a program, a macro package, a
-% configuration file for some program, merely a part of any of the
-% aforementioned, or even arbitrary combinations of the above---and on
-% the other hand it is a \LaTeX\ document which documents this code. One
-% important advantage with this arrangement is that one can keep all the
-% pieces of a project that has to do with a specific task at one place;
-% experience has shown that this greatly furthers really keeping all
-% parts of a project up to date with each other.
-%
-% Slightly simplified, one can say that a \texttt{.dtx} file contains
-% three kinds of lines. A \emph{code line} is a line that doesn't begin
-% with a `|%|' character; such lines can be extracted (copied) using the
-% \package{docstrip} program~\cite{docstrip}. A \emph{guard line} is a
-% line that begins with the two characters `|%<|'; guards are used to
-% structure the set of code lines so that \package{docstrip} can extract
-% different code lines to different generated files. A \emph{comment
-% line}, finally, is a line that begins with a `|%|' character that is
-% not immediately followed by a `|<|' character. The comment lines are
-% ignored by \package{docstrip}, but are part of (and usually make up
-% most of) the \LaTeX\ document in the \texttt{.dtx} file.
-%
-%
-% \subsection{Special conventions and basic features in a \texttt{.dtx}
-% \LaTeX\ document}
-%
-% An important difference between normal \LaTeX\ documents and
-% \texttt{.dtx} \LaTeX\ documents is that the percent character `|%|'
-% doesn't start a comment in the latter; in fact it is usually ignored.
-% This allows \LaTeX\ to see and typeset the text in the comment lines
-% of a \texttt{.dtx} file. Hence if one wants to include the sentence
-% ``Your hovercraft is full of eels!'', which in a normal \LaTeX\
-% document could have been written as the line\iffalse
-%<*example>
-% \fi
-%\begin{verbatim}
-%Your hovercraft is full of eels!
-%\end{verbatim}
-% one would instead write the line as
-%\begin{verbatim}
-%% Your hovercraft is full of eels!
-%\end{verbatim}
-% in a \texttt{.dtx} document. The space after the |%| is not
-% necessary, but most \texttt{.dtx} documents you see include
-% it---probably because the ``comment out \TeX\ code'' action of most
-% text editors consists of inserting a percent \emph{and} a space at
-% the beginning of each line.
-%
-% The code lines present the opposite problem, as they usually
-% shouldn't be treated as normal \LaTeX\ code although the normal \LaTeX\
-% reading conventions would make them the entire document. The usual way
-% around this is to surround each group of code lines with two comment
-% lines that begin and end an environment in which the code lines get
-% typeset verbatim. The \package{tcldoc} package provides the \texttt{tcl}
-% environment for this purpose, so the code lines
-%\begin{verbatim}
-%proc factorial {n} {
-% set result 1
-% for {set i 1} {$i<=$n} {incr i} {
-% set result [expr {$result * $i}]
-% }
-% return $result
-%}
-%\end{verbatim}
-% could be included in a \texttt{.dtx} document as the lines
-%\begin{verbatim}
-%% \begin{tcl}
-%proc factorial {n} {
-% set result 1
-% for {set i 1} {$i<=$n} {incr i} {
-% set result [expr {$result * $i}]
-% }
-% return $result
-%}
-%% \end{tcl}
-%\end{verbatim}
-% When typeset, this will look as
-% \begin{macrocode}
-proc factorial {n} {
- set result 1
- for {set i 1} {$i<=$n} {incr i} {
- set result [expr {$result * $i}]
- }
- return $result
-}
-% \end{macrocode}
-% The tiny numbers at the beginning of each line enumerate the code lines.
-% Index references to code usually specify such code line numbers, but
-% the enumeration can be switched off.
-%
-% In mathematical papers, the statements of e.g.\ theorems are usually
-% made inside a \texttt{theorem} (or whatever) environment which
-% provides certain text formatting, a heading, and a position in the
-% document that can be referenced from other parts of it. In \texttt{.dtx}
-% documents one usually does something similar for each named piece of
-% code: macros, environments, templates, etc. In particular, the
-% \package{tcldoc} package provides two environments \texttt{proc} (for
-% procedures) and \texttt{variable} (for variables).
-% Figure~\ref{Fig:procex} contains an example of how \texttt{proc}
-% might be used in describing a procedure for computing the greatest
-% common divisor of two integers.
-%
-% \begin{figure}
-%
-% \begin{trivlist}
-% \item[]\leavevmode\llap{\texttt{gcd} (proc)\kern\marginparsep}^^A
-% The |gcd| procedure takes two arguments $a$ and $b$ which must be
-% integers and returns their greatest common divisor $\gcd(a,b)$,
-% which is computed using Euclid's algorithm. As a special case,
-% $\gcd(0,0)$ is considered to be $0$, so formally |gcd $a $b|
-% computes $\lvert a \rvert \wedge \lvert b \rvert$, where $\wedge$
-% denotes the meet operation in the divisor lattice of non-negative
-% integers.
-% \begin{macrocode}
-proc gcd {a b} {
- set a [expr {abs($a)}]
- set b [expr {abs($b)}]
- while {$b>0} {
- set r [expr {$a%$b}]
- set a $b
- set b $r
- }
- return $a
-}
-% \end{macrocode}
-% \end{trivlist}
-%
-% \begin{center}
-% (a) A typeset procedure with description
-% \end{center}
-%
-%\begin{verbatim}
-%% \begin{proc}{gcd}
-%% The |gcd| procedure takes two arguments $a$ and $b$ which must be
-%% integers and returns their greatest common divisor $\gcd(a,b)$,
-%% which is computed using Euclid's algorithm. As a special case,
-%% $\gcd(0,0)$ is considered to be $0$, so formally |gcd $a $b|
-%% computes $\lvert a \rvert \wedge \lvert b \rvert$, where $\wedge$
-%% denotes the meet operation in the divisor lattice of non-negative
-%% integers.
-%% \begin{tcl}
-%proc gcd {a b} {
-% set a [expr {abs($a)}]
-% set b [expr {abs($b)}]
-% while {$b>0} {
-% set r [expr {$a%$b}]
-% set a $b
-% set b $r
-% }
-% return $a
-%}
-%% \end{tcl}
-%% \end{proc}
-%\end{verbatim}
-%
-% \begin{center}
-% (b) The code for the example in (a)
-% \end{center}
-%
-% \caption{An example of the \texttt{proc} environment}
-% \label{Fig:procex}
-% \end{figure}
-%
-% What does the \texttt{proc} environment do more precisely? First
-% there's the marginal heading which can be seen in Figure~^^A
-% \ref{Fig:procex}. Such headings make it easier to find the procedure in
-% the typeset form of the document. Then the \texttt{proc} environment
-% makes an index entry which tells you where the procedure is defined,
-% and finally it stores the procedure name in a variable so that
-% subsequent |\changes|\footnote{The \cs{changes} command is defined by
-% the \package{doc} package~\cite{doc}. It is used for adding entries
-% to a global list of changes for code in the \texttt{.dtx} document.}
-% commands know to what the change that they are recording was made.
-%
-% The \texttt{variable} environment does the same things except that it
-% writes ``(var.)'' rather than ``(proc)''. This environment wasn't used
-% for describing the three local variables |a|, |b|, and |r| in the
-% example; this is since there is no point in referring to these
-% variables from elsewhere in the program. Instead the \texttt{variable}
-% environment is primarily meant for global variables (although it could
-% also be useful for local variables that are meant to be accessed using
-% |upvar| or |uplevel|), and as such it can often be of great help, since
-% the description of a global variable can otherwise be hard to find,
-% especially with languages like \Tcl\ where variables don't have to be
-% declared and thus have no natural ``home'' in the code.
-%
-% Another noteworthy feature in the example is the use of vertical
-% bar `\verb+|+' characters to delimit short pieces of verbatim \Tcl\
-% code in the comment lines. It is often necessary for the explanation
-% to include short examples of code in the documentation, and the
-% standard \LaTeX\ |\verb| command is exactly what one would need for
-% this. As such code sections are rather frequent however, it has become
-% the custom to use a single character for both starting and ending such
-% a piece of code. The \package{tcldoc} document class defines \verb+|+
-% as a shorthand for \verb+\tclverb|+, where |\tclverb| is a variant of
-% |\verb| which has been designed specifically for \Tcl\ code.
-%
-% The above description was meant to give a basic understanding of how
-% \Tcl\ code and documentation thereof can be mixed in a \texttt{.dtx}
-% file, it neither explains all the environments and commands that
-% the \package{tcldoc} package provides, nor mentions all the
-% features of the environments that were described. That information can
-% instead be found in Section~\ref{Sec:Manual} of this paper. It should
-% also be mentioned that the \package{doc} package~\cite{doc} defines
-% several commands and environments that may be of use for describing
-% code, and it is well worth getting acquainted with the features of
-% that package as well (although parts of its documentation has become
-% rather archaic).
-%
-%
-% \subsection{Guards and \package{docstrip} installation scripts}
-%
-% The central command in a \package{docstrip} installation script is
-% |\generate|, since this is the command which actually causes code to
-% be extracted. |\generate|'s syntax is
-% \begin{quote}
-% |\generate|\marg{files}
-% \end{quote}
-% where \meta{files} consists of one or several |\file| commands, each
-% of which has the syntax
-% \begin{quote}
-% |\file|\marg{output}\marg{sources}
-% \end{quote}
-% where \meta{output} is a filename and \meta{sources} consists of one
-% or several |\from| commands, each of which has the syntax
-% \begin{quote}
-% |\from|\marg{input}\marg{options}
-% \end{quote}
-% where, finally, \meta{input} is a filename and \meta{options} is a
-% comma-separated list of alphanumerical strings. Thus a |\generate|
-% command might look like
-%\begin{verbatim}
-% \generate{\file{p1.sty}{\from{s1.dtx}{foo,bar}}
-% \file{p2.sty}{\from{s2.dtx}{baz}
-% \from{s3.dtx}{baz}}
-% \file{p3.sty}{\from{s1.dtx}{zip}
-% \from{s2.dtx}{zap}}
-% }
-%\end{verbatim}
-% The meaning of this command is
-% \begin{quote}
-% Generate the three files \texttt{p1.sty}, \texttt{p2.sty}, and
-% \texttt{p3.sty}. Extract the code for \texttt{p1.sty} from
-% \texttt{s1.dtx} with options \texttt{foo} and \texttt{bar}, extract
-% the code for \texttt{p2.sty} from \texttt{s2.dtx} with option
-% \texttt{baz} and \texttt{s3.dtx} (the code from \texttt{s2.dtx}
-% will be put before the code from \texttt{s3.dtx}) with option
-% \texttt{baz}, and finally extract the code for \texttt{p3.sty} from
-% \texttt{s1.dtx} with option \texttt{zip} and \texttt{s2.dtx} with
-% option \texttt{zap}.
-% \end{quote}
-% The \emph{options} are used to control which parts of the source files
-% should be extracted to which generated file. A source file can
-% contain a number of \emph{modules}, and at the beginning of each
-% module \package{docstrip} decides, for each output file separately,
-% whether the code lines in that module should be extracted to the
-% output file. The beginning of a module is marked by a guard line which
-% has the syntax
-% \begin{quote}
-% |%<*|\meta{expression}|>|
-% \end{quote}
-% and the end by a corresponding
-% \begin{quote}
-% |%</|\meta{expression}|>|
-% \end{quote}
-% guard line. In their simplest form, the \meta{expression}s are names
-% of options, and in that case the code lines in the module are only
-% extracted if that option appears in the \meta{options} for that
-% combination of input file and output file. The \meta{expression}s can
-% however be arbitrarily complicated boolean expressions;
-% see~\cite{docstrip} for more information.
-% Modules may nest inside each other, and in that case the code lines in
-% an inner module can only be included if all surrounding modules are
-% being included. It is checked that matching |*| and |/| guard lines
-% contain the same (as strings) \meta{expression}, and case is
-% significant in the names of options.
-%
-% One application of modules which has already been mentioned is to
-% bundle code for several different generated files in the same
-% \texttt{.dtx} file---one example of this is the file
-% \texttt{doc.dtx} (part of the \LaTeX\ base distribution) which
-% contains both the \package{doc} package (\texttt{doc.sty}) and two
-% \package{makeindex} style files (\texttt{gglo.ist} and
-% \texttt{gind.ist}). Another application is to keep variant sections
-% of code---such as special code for debugging or gathering
-% statistics---in the \texttt{.dtx} source file for a program without
-% thereby making it a part of the normal form of that program. It is
-% quite possible to use \package{docstrip} as a simple pre-processor
-% for languages whose compiler\slash interpreter haven't got one built
-% in.
-%
-% There are many other commands available in a \package{docstrip}
-% installation script beside those listed above, but those are well
-% described in the \package{docstrip} manual~\cite{docstrip} and need
-% little attention here. Instead I'm going to finish this subsection
-% with a quick guide to the particular difficulties one faces when using
-% \package{docstrip} to extract \Tcl\ code, and how to overcome them.
-%
-% The main problem is that \package{docstrip} insert a few comment lines
-% at the beginning and end of each file it generates. This is a good
-% thing, because a file consisting entirely of extracted code lines
-% would normally be completely void of commentary and quite
-% unintelligible for the casual user. These few comment lines explain
-% that the file was generated by \package{docstrip} from other files
-% (which contain the documentation of the code), lists those files,
-% and normally also contains a copyright (or more commonly some kind of
-% copyleft) notice. The problem lies in that comments look different in
-% different languages, and as the default is to write \TeX\ style
-% comments, one must tell \package{docstrip} to write \Tcl\ style
-% comments. This can be done through the command
-%\begin{verbatim}
-%\edef\MetaPrefix{\string#}
-%\end{verbatim}
-% which tells \package{docstrip} to begin each inserted comment line
-% with the character `|#|'.
-%
-% The comment lines inserted at the beginning of a generated file is
-% called the \emph{preamble} and those of the end the \emph{postamble}.
-% To set the preamble, one writes
-% \begin{trivlist}
-% \item |\preamble|\\
-% \meta{preamble lines}\\
-% |\endpreamble|
-% \end{trivlist}
-% and correspondingly to set the postamble
-% \begin{trivlist}
-% \item |\postamble|\\
-% \meta{postamble lines}\\
-% |\endpostamble|
-% \end{trivlist}
-% The \meta{preamble lines} and \meta{postamble lines} can be any number
-% of lines (including zero). Unlike the text in source files, the text in
-% these preamble and postamble lines is not read verbatim, so things in
-% these lines which have special meaning to \TeX\ (such as control
-% sequences) will be treated as such; the only exception is that spaces
-% and newlines are preserved (instead of concatenated to single spaces
-% as they normally would). It is important that the preamble and
-% postamble \emph{are} set after |\MetaPrefix| is changed, because each
-% line specified between |\preamble| and |\endpreamble| or |\postamble|
-% and |\endpostamble| respectively will be prefixed by the current value
-% of |\MetaPrefix|.
-%
-% Finally, some programs assign special meaning to the first line of a
-% file, so one might want to control what gets put there. Merely using
-% |\preamble| doesn't achieve this, because the \meta{preamble lines}
-% specified that way are put after the lines saying ``this is a
-% generated file \textellipsis''. You can however add things to the
-% preamble by explicitly setting the macro |\defaultpreamble|, which is
-% where \package{docstrip} stores the preamble. To make the first line
-% a comment which simply contains the text `|-*-Tcl-*-|', you could give
-% the command
-%\begin{verbatim}
-%\edef\defaultpreamble{\MetaPrefix\space -*-Tcl-*-^^J\defaultpreamble}
-%\end{verbatim}
-% but an explanation of this is a bit beyond this introduction.
-%
-% In summary, a \package{docstrip} installation script for extracting a
-% file \texttt{foo.tcl} from \texttt{foo.dtx}, using \Tcl\ style
-% comments, inserting a GPL copyleft notice in the preamble, and
-% beginning with the line |# -*-Tcl-*-| could look as follows:
-%\begin{verbatim}
-%\input docstrip.tex
-%
-%\edef\MetaPrefix{\string#}
-%
-%\preamble
-%
-%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.
-%See the file "license.terms" for information on usage and
-%redistribution of this file, and for a DISCLAIMER OF ALL
-%WARRANTIES.
-%
-%\endpreamble
-%
-%\postamble
-%\endpostamble
-%
-%\edef\defaultpreamble{\MetaPrefix\space -*-Tcl-*-^^J\defaultpreamble}
-%
-%\generate{\file{foo.tcl}{\from{foo.dtx}{bar}}}
-%
-%\end
-%\end{verbatim}
-%
-% The generated file \texttt{foo.tcl} will contain
-%\begin{verbatim}
-%# -*-Tcl-*-
-%#
-%# This is file `foo.tcl',
-%# generated with the docstrip utility.
-%#
-%# The original source files were:
-%#
-%# foo.dtx (with options: `bar')
-%#
-%# 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.
-%# See the file "license.terms" for information on usage and
-%# redistribution of this file, and for a DISCLAIMER OF ALL
-%# WARRANTIES.
-%#
-%\end{verbatim}
-% \meta{lines extracted from \texttt{foo.dtx}}
-%\begin{verbatim}
-%#
-%#
-%# End of file `foo.tcl'.
-%\end{verbatim}
-%
-%
-%
-% \subsection{The structure of the \LaTeX\ document}
-%
-% All this has been about the local appearance of a \texttt{.dtx} file,
-% but what about the overall structure? There are several points to
-% raise about that as well.
-%
-% The first is that \LaTeX\ to begin with treat \texttt{.dtx} documents
-% just like any other document---a `|%|' starts a comment and only lines
-% \emph{not} beginning with a `|%|' contain anything that \LaTeX\ can
-% see. Somehow \LaTeX\ must be instructed to start applying the special
-% reading rules that were described above. This is the job of the
-% so-called \emph{driver}, which (for a file \texttt{myfile.dtx}) in its
-% simplest form can look like
-%\begin{verbatim}
-%\documentclass{tcldoc}
-%\begin{document}
-%\DocInput{myfile.dtx}
-%\end{document}
-%\end{verbatim}
-% The important command here is |\DocInput|, because that is what tells
-% \LaTeX\ to apply the special \texttt{.dtx} reading rules. More
-% precisely it means ``Start ignoring `|%|' characters in the text you
-% read, input the file \texttt{myfile.dtx}, and when you're done return
-% to treating `|%|' characters as before.''
-%
-% The driver is usually put in the very first stretch of code lines in
-% the \texttt{.dtx} file. This means that \LaTeX, when ordered to typeset
-% the \texttt{.dtx} file, will start to read along, possibly ignoring
-% hundreds of lines beginning with `|%|' because they are comments. Then
-% it encounters the driver, and after the |\documentclass| and
-% |\begin{document}| commands it executes the |\DocInput|. This will
-% cause it to not ignore lines beginning with `|%|', so when it starts
-% reading the file again it will see all the lines it skipped the first
-% time through. The file will be read to end, after which \LaTeX\
-% returns to the command after the |\DocInput|. As that command happens
-% to be |\end{document}|, it finishes the typeset document and stops.
-% This stop prevents it from seeing and interpreting as \LaTeX\ commands
-% the remaining code lines in the file.
-%
-% The second time through the driver shouldn't be interpreted as
-% \LaTeX\ commands, since for example the |\documentclass| command may
-% only be used once in a \LaTeX\ document. One way of achieving this is
-% to put an |\iffalse| command right before the driver and a |\fi|
-% command right after it. This says to \LaTeX\ that the driver code is
-% conditional material, and since the condition evaluates to false
-% (|\iffalse| always evaluates to false), this conditional material
-% should be skipped. Thus the first few lines of \texttt{myfile.dtx}
-% typically might be
-%\begin{verbatim}
-%% \iffalse
-%%<*driver>
-%\documentclass{tcldoc}
-%\begin{document}
-%\DocInput{myfile.dtx}
-%\end{document}
-%%</driver>
-%% \fi
-%\end{verbatim}
-% The \texttt{driver} guard lines are there to stop \package{docstrip}
-% from including the driver code in the generated files.
-%
-% After the driver comes the actual \LaTeX\ document. The document
-% usually consists of two parts, where the first part is a manual for
-% the \emph{usage} of the code defined in the file, and the second part
-% contains the actual \emph{implementation} (documented code). The idea
-% is that most people are (at least the first time) quite content with
-% learning how to use something, so one should make it simple for them
-% to find that information.\footnote{One needn't take this as an absolute
-% rule---I for one haven't written all my packages that way---but
-% structuring the document like this generally makes it more accessible.}
-% To further this approach one puts the command |\StopEventually| at the
-% start of the implementation part and the command |\Finale| at the end
-% of it. Normally |\StopEventually| doesn't make itself felt, but if one
-% previously has given the command |\OnlyDescription| then rest of the
-% file will not be read; this can be used to produce a ``manual only''
-% version of the documentation. |\StopEventually| takes one argument and
-% the code in this argument is executed at the |\Finale| (if the
-% implementation part is being included) or immediately (if the
-% implementation part isn't being included). Thus this argument is the
-% place to put things that should appear at the very end of the
-% document.
-%
-% The \texttt{tcl}, \texttt{proc}, and \texttt{variable} environments
-% described above all typically appear in the implementation part of the
-% document.
-%
-%
-% \section{Usage of commands and environments}
-% \label{Sec:Manual}
-%
-%
-% \subsection{The actual source code}
-%
-% The \DescribeEnv{tcl}\texttt{tcl} environment is used for wrapping up
-% a group of code lines which contain \Tcl\ code. Lines inside this
-% environment which begin with a percent character are called
-% \emph{command} lines and can contain \LaTeX\ commands which get
-% executed, whereas lines that do not begin with a percent character are
-% called \emph{normal} lines and get typeset verbatim (or nearly
-% verbatim). Lines that begin with |%<| (\package{docstrip} guard lines)
-% do however constitute a special case, as the guard expression will get
-% typeset as in \package{doc}'s \texttt{macrocode} environment and the
-% remainder of the line will get processed in command mode if it is a
-% |*| or |/| guard, but in normal mode if the guard line was of any other
-% type.
-%
-% The \texttt{tcl} environment uses the same general formatting
-% parameters as \package{doc}'s \texttt{macrocode} environment. In
-% particular this means that the text on a normal line is typeset in
-% |\MacroFont| (by default the same thing as \cs{normalfont}\penalty0
-% \cs{ttfamily}\penalty0\cs{small}) or |\AltMacroFont| (by default the same
-% thing as \cs{normalfont}\penalty0\cs{ttfamily}\penalty0\cs{itshape}^^A
-% \penalty0\cs{small}) depending on the current \package{docstrip} module
-% nesting level. The \package{tcldoc} class sets the
-% \texttt{StandardModuleDepth} counter to 1, which means that the
-% |\AltMacroFont| is used when the modules two levels deep or more.
-%
-% If a normal line is too long to fit on one line then the \texttt{tcl}
-% environment will try to break it. Legal breakpoints are spaces which
-% could be replaced by a backslash and a newline without changing the
-% meaning of the command; thus most spaces are legal breakpoints. When a
-% line is broken at a space like this, the space is replaced by a
-% backslash so that the line is still ``syntactically correct''. The
-% opposite happens to lines which actually end with an escaped newline;
-% such lines are concatenated with the following line and are treated as
-% one long line. This is so that a sequence of breakpoints can be chosen
-% which is optimal for the actual line width of the document (as opposed
-% to the line width used in the text file, which can be something quite
-% different). For example
-%\begin{verbatim}
-%% \begin{tcl}
-%lsearch -exact\
-% [concat $a [lrange $b $first end] c d e f]\
-% [lindex $g $h]
-%% \end{tcl}
-%\end{verbatim}\iffalse
-%</example>
-% \fi
-% could be typeset as any of the code examples in
-% Figure~\ref{Fig:Radbrytning}, depending on how wide a line is.
-% \begin{figure}
-% \begin{center}
-% \small
-% \(\left\lvert
-% \begin{minipage}{350pt}
-% \begin{flushleft}
-% |lsearch -exact [concat $a [lrange $b $first end] c d e f]|^^A
-% | [lindex $g $h]|
-% \end{flushleft}
-% \end{minipage}
-% \right\rvert\)
-%
-% \smallskip
-%
-% \(\left\lvert
-% \begin{minipage}{298pt}
-% \begin{flushright}
-% \noindent |lsearch -exact [concat $a [lrange $b $first|^^A
-% | end] c d e f]|\textbackslash\hfill\vadjust{}\\
-% |[lindex $g $h]|
-% \end{flushright}
-% \end{minipage}
-% \right\rvert\)
-%
-% \smallskip
-%
-% \(\left\lvert
-% \begin{minipage}{241pt}
-% \begin{flushright}
-% \noindent |lsearch -exact|\textbackslash\hfill\vadjust{}\\
-% |[concat $a [lrange $b $first end] c d e f]|\textbackslash\\
-% |[lindex $g $h]|
-% \end{flushright}
-% \end{minipage}
-% \right\rvert\)
-%
-% \smallskip
-%
-% \(\left\lvert
-% \begin{minipage}{184pt}
-% \begin{flushright}
-% \noindent |lsearch -exact [concat $a|\textbackslash\hfill
-% \vadjust{}\\
-% |[lrange $b $first end] c d e f]|\textbackslash\\
-% |[lindex $g $h]|
-% \end{flushright}
-% \end{minipage}
-% \right\rvert\)
-% \end{center}
-%
-% \caption{The same \Tcl\ code, set in different linewidths}
-% \label{Fig:Radbrytning}
-% \end{figure}
-% The environment tries to put the linebreaks at the lowest possible
-% nesting (of braces and brackets) level; I believe this gives the best
-% readability.\footnote{When I first started programming in \Tcl\ I
-% used a completely different method for breaking long lines in the
-% code---I mainly implemented the current method because it was
-% simple to program---but I was quite surprised by how readable it made
-% the code.} There is however a way to override this automatic choice
-% of breakpoints: if a normal line which ends with an escaped newline is
-% followed by a command line (even a command line that doesn't contain
-% any commands) then it will not be concatenated with the next normal
-% line.
-%
-% There are a couple of restrictions on the code in command lines. First
-% of all it is not allowed to start a new paragraph (there will be an
-% error message). Secondly a command may not be broken across several
-% lines---all the arguments must appear on the same line as the
-% control sequence. Thirdly some characters have other catcodes than
-% in normal \LaTeX, so it is not certain that all commands work. Some
-% commands that do work and may be useful are:
-% \begin{itemize}
-% \item Vertical space commands (|\smallskip|, |\medskip|, etc.)
-% The command line `|% \medskip|' is more to type than a blank
-% normal line, but it looks slightly better.
-% \item Indexing commands (|\index|, |\IndexEntry|,\footnote{This
-% command is defined by the \package{xdoc} package~\cite{xdoc}.}
-% etc.)
-% \item The \cs{TclInput} and |\settabsize| commands (see below).
-% \end{itemize}
-% And of course the |\end{tcl}| command works in a command line, since
-% that is how one ends a \texttt{tcl} environment.
-%
-% Besides the \texttt{tcl} environment there is also a
-% \DescribeEnv{tcl*}\texttt{tcl*} environment which is different from
-% \texttt{tcl} only in that spaces and tabs are typeset as special
-% visible space `\textvisiblespace' and visible tab
-% `\makebox[6\fontdimen2\font]{\( - \mkern-7mu
-% \cleaders\hbox{$\mkern-2mu - \mkern-2mu$}\hfill \mkern-7mu
-% \mathord\rightarrow \mkern-1mu \vrule \mkern1mu\)}' characters. This
-% can be useful for pieces of code where the exact number of spaces in
-% a sequence is significant, such as code for writing tables that align.
-%
-% For shorter pieces of \Tcl\ code, e.g.\ examples, there's the
-% \DescribeMacro{\tclverb}|\tclverb| command. |\tclverb| is very similar
-% to the standard \LaTeX\ command |\verb|, but there are two differences.
-% The first is that text typeset by |\tclverb| can contain breakpoints
-% at whitespace; these behave just as in the \texttt{tcl} environment.
-% The second is that the verbatim text that follows |\tclverb| may contain
-% newlines, provided that these newlines are escaped by a backslash. Like
-% a \Tcl\ interpreter, |\tclverb| ignores whitespace following an escaped
-% newline. Unlike a \Tcl\ interpreter, |\tclverb| also ignores one percent
-% character before the ignored whitespace, if it is the first character
-% on the following line. Thus
-%\begin{verbatim}
-%% \tclverb|append a $b| is much more efficient than \tclverb|set a\
-%% $a$b| if \tclverb|$a| is long.
-%\end{verbatim}
-% is perfectly legal, and the escaped newline between |a| and |$a$b| is
-% treated just like the space between |set| and |a|. Like |\verb|, the
-% |\tclverb| command has a starred form \DescribeMacro{\tclverb*}^^A
-% |\tclverb*| which also typesets spaces and tabs as visible characters.
-%
-%
-% The \DescribeMacro{\MakeShortTclverb}|\MakeShortTclverb| command works
-% just like the \package{doc}\slash\package{shortvrb} command
-% |\MakeShortVerb|, except that it makes the active character a shorthand
-% reference for |\tclverb|\textellipsis\ instead of |\verb|\textellipsis.
-% Use |\DeleteShortVerb| to undo the effect of a |\MakeShortTclverb|. The
-% \package{tcldoc} class executes the command
-% \begin{quote}
-% \verb"\MakeShortTclverb{\|}"
-% \end{quote}
-% which makes \verb"|" a shorthand for \verb"\tclverb|".
-%
-% Since there is no universally accepted standard for the size
-% (equivalent number of spaces) of a tab, there is a command
-% \DescribeMacro{\settabsize}|\settabsize| for changing this.
-% |\settabsize| takes as its only argument the new tab size, which must
-% be an integer in the range 2--255. The default value is 8.
-% |\settabsize| makes a local assignment to the tab size. The tab size
-% can be changed inside a \texttt{tcl} (or \texttt{tcl*}) environment.
-%
-% There is also a command \DescribeMacro{\TclInput}|\TclInput| which
-% is used for typesetting ``raw'' (not in \texttt{.dtx} format) \Tcl\ code
-% files. |\TclInput| is meant to be used on a \emph{command} line of a
-% \texttt{tcl} or \texttt{tcl*} environment, and it efficiently makes
-% things look as if the |\TclInput| command had been replaced by the
-% inputted file in its entirety (preceded by a newline, and followed
-% by a percent and a space). |\TclInput| takes as its only argument the
-% name of the file to input.
-%
-% To typeset the file \texttt{myscript.tcl} one would write
-%\begin{verbatim}
-%% \begin{tcl}
-%% \TclInput{myscript.tcl}
-%% \end{tcl}
-%\end{verbatim}
-% or even
-%\begin{verbatim}
-% \begin{tcl}\TclInput{myscript.tcl}\end{tcl}
-%\end{verbatim}
-% anywhere in a \package{tcldoc} document. This works since the
-% \texttt{tcl} environment is in command mode right after the initial
-% |\begin{tcl}|, and the |\end{tcl}| need not be the first command on a
-% command mode line.
-%
-% Finally, if the \package{rtkinenc} package~\cite{rtkinenc} is being
-% used then the verbatim code environments and commands described above
-% are to some extent able to catch errors that would otherwise occur
-% because the fonts in use do not contain all characters in the input
-% encoding. Rather than printing nothing, missing characters will be
-% written as the corresponding |\x|\meta{hh} backslash sequence in a
-% slightly different font than the rest of the text. This provides an
-% unambiguous representation regardless of input encoding.
-%
-%
-% \subsection{Markup of named things}
-%
-% The two environments \DescribeEnv{proc}\texttt{proc} and
-% \DescribeEnv{variable}\texttt{variable}, which are analogues of
-% \package{doc}'s \texttt{macro} environment, for procedures and
-% variables respectively have already been mentioned in
-% Section~\ref{Sec:Introduction}. In addition to those there are two
-% environments \DescribeEnv{arrayentry}\texttt{arrayentry} and
-% \DescribeEnv{arrayvar}\texttt{arrayvar} which are meant for entries
-% in array variables and array variables as a whole. The complete
-% syntaxes of these environments are
-% \begin{quote}
-% |\begin{proc}|\oarg{namespace}\marg{proc name}\\
-% \vadjust{}\quad$\vdots$\\
-% |\end{proc}|
-%
-% |\begin{variable}|\oarg{namespace}\marg{variable name}\\
-% \vadjust{}\quad$\vdots$\\
-% \quad|\end{variable}|
-%
-% |\begin{arrayentry}|\oarg{namespace}\marg{array name}^^A
-% \marg{entry name}\\
-% \vadjust{}\quad$\vdots$\\
-% |\end{arrayentry}|
-%
-% |\begin{arrayvar}|\oarg{namespace}\marg{array name}^^A
-% \oarg{index-des}\\
-% \vadjust{}\quad$\vdots$\\
-% |\end{arrayvar}|
-% \end{quote}
-% The \meta{proc name}, \meta{variable name}, and \meta{array name}
-% arguments are quite evidently the names of the procedure, variable,
-% and array respectively. The \meta{namespace} argument can be to
-% specify the namespace part of a qualified name; having the name split
-% like this makes it easier to treat the namespace differently from the
-% rest of the qualified name.\footnote{By redefining \cs{TD@main@index}
-% one can for example make the index sorted first by name, and then by
-% namespace.} The command \DescribeMacro{\buildname}|\buildname|
-% constructs a qualified name from a namespace and a name. If there is
-% no \meta{namespace} argument then the namespace used will be the
-% default namespace. The default namespace is set using the
-% \DescribeMacro{\setnamespace}|\setnamespace| command, which takes the
-% namespace name as its only argument. The default namespace at the
-% beginning of the document is the global namespace, whose name is the
-% empty string.
-%
-% The \texttt{arrayentry} environment is intended for certain
-% distinguished entries in an array, such as entries inserted to
-% make the boundary cases of an algorithm work correctly and entries
-% which have a special meaning to the program. Not all arrays contain
-% such special entries, but when they do it is a good practice
-% to explain them explicitly. The \meta{index-des} argument of the
-% \texttt{arrayvar} environment can be used to specify what is used as
-% index into the array; the text in this argument will appear both in
-% the margin and in the index, but note that \meta{index-des} is a
-% moving argument. There is little difference between the
-% \texttt{variable} and \texttt{arrayvar} environments when the
-% \meta{index-des} argument of the latter isn't used, but the index
-% entries they make behave differently with respect to
-% \texttt{arrayentry} index entries. An \texttt{arrayentry} index entry
-% will be put as a subentry of the corresponding \texttt{arrayvar}
-% entry, whereas a \texttt{variable} entry would appear separately.
-%
-% The above environments usually only appear in the implementation part
-% of a \texttt{.dtx} file. For the usage part there is a command
-% \DescribeMacro{\describestring}|\describestring| which produces
-% marginal headings and index entries. The syntax of |\describestring|
-% is
-% \begin{quote}
-% |\describestring|\oarg{type}\oarg{namespace}\marg{text}
-% \end{quote}
-% The \meta{text} is the string for which a heading and index entry will
-% be made, whereas the \meta{type} (if given) is put after the text.
-% If the \meta{namespace} is given then the thing described is supposed
-% to be the name of something namespace-relative (like a procedure or
-% global variable) and in this case the complete name is formed by
-% passing \meta{namespace} and \meta{text} to |\buildname|. If
-% \meta{type} is \texttt{proc}, \texttt{var.}, or \texttt{array} and a
-% namespace is given then the index entry made will fit that made by
-% a corresponding \texttt{proc}, \texttt{variable}, or
-% \texttt{arrayvar} respectively environment. The \meta{type} argument
-% is, in \LaTeX\ terminology, moving.
-%
-% The \meta{text} and \meta{namespace} arguments can contain arbitrary
-% characters and most characters can be entered verbatim. Amongst the
-% exceptions are `|%|', `|\|', `|{|', and `|}|', which instead can be
-% entered as
-% |\PrintChar|\discretionary{}{}{}|{`\%}|,
-% |\PrintChar|\discretionary{}{}{}|{`\\}|,
-% |\PrintChar|\discretionary{}{}{}|{`\{}|, and
-% |\PrintChar|\discretionary{}{}{}|{`\}}| respectively. See the
-% \package{xdoc} package~\cite{xdoc} documentation for an explanation of
-% the |\PrintChar| command. The \meta{text} argument can also contain
-% ``variant'' parts made using the \DescribeMacro{\meta}|\meta| command.
-% As an example,
-% \begin{quote}
-% |\describestring[array]{\meta{mode}modeVars}|
-% \end{quote}
-% puts the text
-% \begin{quote}
-% \meta{mode}|modeVars| (array)
-% \end{quote}
-% in the margin and index. The arguments of such |\meta| commands are
-% moving.
-%
-%
-% \subsection{Describing command syntaxes}
-%
-% One important part of documentation is to describe the syntaxes of
-% commands. The previous subsection contains examples of the conventions
-% for this that has been developed for \LaTeX\
-% commands---mandatory arguments are denoted as `\marg{argument}' and
-% optional arguments are denoted as `\oarg{argument}'. These two classes
-% suffice rather well for \LaTeX\ commands, but the syntaxes of \Tcl\
-% commands are not seldom much more complex. Therefore a more powerful
-% form of syntax specification is called for, and one which is close at
-% hand is that used in regular expressions since it is already part of
-% the \Tcl\ language anyway.
-%
-% \DescribeMacro\regopt
-% \DescribeMacro\regstar
-% \DescribeMacro\regplus
-% The simplest commands available are the modifiers |\regopt|,
-% |\regstar|, and |\regplus|, which correspond to the |?|,
-% |*|, and |+| metacharacters in a regular expression; using |\regopt|
-% after a term says that it is optional, |\regstar| says that the term
-% can be repeated an arbitrary number of times (including zero), and
-% |\regplus| says that the term occurs at least once. The typeset
-% results of these commands are $^?$, $^*$, and $^+$ respectively
-% (recall that exponents are sometimes used to denote repetition).
-%
-% The terminals in the expression are best made using |\tclverb| (for
-% ``fixed'' material, e.g.\ procedure names) and
-% \DescribeMacro{\word}|\word| (for variable material, e.g.\
-% arguments). The syntax of |\word| is
-% \begin{quote}
-% |\word|\marg{text}
-% \end{quote}
-% and e.g.\ |\word{script}| gets typeset as
-% \begin{quote}
-% \word{script}
-% \end{quote}
-% Using these, one can for example specify the syntaxes of the \Tcl\
-% commands |append| and |catch| through
-% \begin{quote}
-% \verb"|append| \word{var-name} \word{value}\regplus"\\
-% \verb"|catch| \word{script} \word{var-name}\regopt"
-% \end{quote}
-% (recall that `\verb"|"' is a shorthand for `\verb"\tclverb|"'). These
-% get typeset as
-% \begin{quote}
-% |append| \word{var-name} \word{value}\(^+\)\\
-% |catch| \word{script} \word{var-name}\(^?\)
-% \end{quote}
-%
-% Terms in regular expressions can also consist of parenthesised
-% subexpressions, which are made using the \DescribeEnv{regexp}^^A
-% \texttt{regexp} environment. The syntax of this environment is
-% \begin{quote}
-% |\begin{regexp}|\oarg{modifier}\quad\dots\quad |\end{regexp}|
-% \end{quote}
-% If \texttt{regexp} environments are nested then the parentheses of
-% the outer environment will grow to be larger than those of the inner
-% environment. A side-effect of this is that the \texttt{regexp}
-% environment wants to know if a modifier will be applied to it, since
-% the amount by which the modifier should be raised in this case depends
-% on the size of the parenthesis before it, and this is what the
-% \meta{modifier} optional argument is for. \LaTeX\ does not provide
-% for arguments at the |\end| of an environment, so it has to be placed
-% at the |\begin|. Using these elements, the syntax of |if| can be
-% specified through
-% \begin{quote}
-% \verb"|if| \word{expression} |then|\regopt\ \word{script}"\\
-% \verb"\begin{regexp}[\regstar]|elseif| \word{expression}"\\
-% \verb"|then|\regopt\ \word{script}\end{regexp}"\\
-% \verb"\begin{regexp}[\regopt]|else| \word{script}\end{regexp}"
-% \end{quote}
-% which typesets as
-% \begin{isyntax}
-% |if| \word{expression} |then|\(^?\) \word{script} \(\bigl(\)^^A
-% |elseif| \word{expression} |then|\(^?\) \word{script}\(\bigr)^*\)
-% \(\bigl(\)|else|~\word{script}\(\bigr)^?\)
-% \end{isyntax}
-%
-% The final regular expression construction that is supported is that
-% of branches of a regular expression. A \texttt{regexp} environment
-% consist of one or several branches that are separated by
-% \DescribeMacro{\regalt}|\regalt| commands. Visually the |\regalt|
-% command gets typeset as a vertical bar that has the same size as the
-% parentheses of the surrounding \texttt{regexp} environment. The
-% |\regalt| command may only be used inside a \texttt{regexp} environment.
-% An example of the use of |\regalt| is the following specification of
-% the syntax of \Tcl's |regexp| command:
-% \begin{quote}
-% \verb"|regexp| \begin{regexp}[\regstar]|-nocase|\regalt"\\
-% \verb"|-indices|\end{regexp} |--|\regopt \word{regular expression}"\\
-% \verb"\word{string} \word{var-name}\regstar"
-% \end{quote}
-% which typesets as
-% \begin{isyntax}
-% |regexp| \((\)|-nocase|~\(\mid\) |-indices|\()^*\) |--|\(^?\)
-% \word{regular expression} \word{string} \word{var-name}\(^*\)
-% \end{isyntax}
-%
-% Finally a note about the relationship between the |\word| command and
-% \package{doc}'s |\meta| command. Whereas the argument of |\word| is
-% encapsulated in braces (and thus ought to be a separate word for a
-% \Tcl\ interpreter), the argument of |\meta| is encapsulated in angle
-% brackets. The idea is that |\word| should be used for things which are
-% separate words to \Tcl, whereas |\meta| should be used for things which
-% are merely part of words. Thus in the command |set b Z${a}Y|, the second
-% word |b| would be a `\word{var-name}' and the third word |Z${a}Y| would
-% be a `|Z|\meta{string}|Y|'.
-%
-%
-%
-% \subsection{Options and customisation}
-%
-% The \package{tcldoc} package does not have any options, but all
-% options passed to it are passed on to the \package{xdoc} package.
-% The \package{tcldoc} class accepts all options that the standard
-% \LaTeX\ document class \package{article} accepts, with the exception
-% of \texttt{a5paper}.
-%
-% In addition, the \package{tcldoc} class also accepts the option
-% \describeoption{macinputenc}\texttt{macinputenc}, which causes a lot
-% of things to be set up so that input characters from outside 7-bit
-% ASCII can be interpreted according to the Macintosh character set.
-%
-% Like the \package{ltxdoc} class, the \package{tcldoc} class will look
-% for a special configuration file \texttt{tcldoc.cfg} and input that
-% file if it exists. This file can be used to declare extra options for
-% the class, have certain options always given, etc. Section~2 of
-% \texttt{ltxdoc.dtx}~\cite{ltxdoc} is a good introduction to how such
-% configuration files can be used with \texttt{.dtx} sources in general.
-%
-% When customising the \package{tcldoc} document class, you affect how
-% all documents using that class will be typeset in your particular \TeX\
-% installation. It is \emph{not} something you have to do, but it can
-% make \package{tcldoc} documents work better with the printers, paper
-% formats, fonts, etc.\ that are available in your installation. It
-% will usually cause line and page breaks to occur at other places than
-% they would do if typeset using an uncustomised \package{tcldoc}
-% class, so the typographical quality of the document can be decreased,
-% but it is uncommon to find an \texttt{.dtx} document whose author
-% have given these matters much attention anyway. Hence the typographic
-% arguments against customisation are weak.
-%
-% A common form of customisation is to use additional packages, since
-% various kinds of document-wide font selection is often done by
-% packages. Due to that the code in \texttt{tcldoc.cfg} is executed when
-% the \package{tcldoc} class does its option processing, at which time
-% \LaTeX\ does not allow loading packages, such customisation is not
-% straightforward. There is a way around that however; to load e.g.\ the
-% \package{textcomp} package, use the command
-%\begin{verbatim}
-%\AtEndOfClass{\usepackage{textcomp}}
-%\end{verbatim}
-% Using |\AtEndOfClass| like this delays the command until it may be
-% executed.
-%
-% Finally, and in particular if you are typesetting documents that use
-% the \texttt{macinputenc} option, you can put code that define
-% ``fakes'' for various text commands in the \texttt{tcldoc.cfg} file.
-% For example, the `diamond' character in the Macintosh encoding is not
-% available in any of the standard \LaTeX\ text fonts, and therefore the
-% \texttt{tcl} environment will typeset this character as the escape
-% code `\textbackslash xd7'. There is a diamond character in the math
-% fonts however, and you can make the \texttt{tcl} environment use that
-% character instead by adding the commands
-%\begin{verbatim}
-%\DeclareTextSymbol{\textdiamond}{OMS}{5}
-%\DeclareTextSymbolDefault{\textdiamond}{OMS}
-%\end{verbatim}
-% to you \texttt{tcldoc.cfg} file. Note however, that this kind of
-% configuration is \emph{advanced} \LaTeX, and you should probably
-% familiarise yourself with the contents of the file
-% \texttt{ltoutenc.dtx}~\cite{ltoutenc} before attempting to design
-% anything along these lines yourself.
-%
-%
-% \subsection{Miscellanea}
-%
-% For writing ``\Tcl'', the \package{tcldoc} package defines the command
-% \DescribeMacro{\Tcllogo}|\Tcllogo|, which for most fonts look slightly
-% better than simply typing |Tcl|. (|\Tcllogo| becomes \Tcl, whereas
-% |Tcl| becomes Tcl.)
-%
-%
-% \section{Acknowledgements}
-%
-% The \package{tcldoc} document class and \LaTeX\ package were
-% constructed starting from three other sources: (i)~the \package{ltxdoc}
-% document class~\cite{ltxdoc} by David Carlisle,
-% (ii)~the \package{doc} package~\cite{doc} by Frank Mittelbach,
-% B.~Hamilton Kelly, Andrew Mills, Dave Love, and Joachim Schrod, and
-% (iii)~my own \package{pasdoc} document class. Hence the `et al.'\ in
-% the author field above. This complicated heritage in the code is
-% mirrored by the documented source---there are paragraphs below that
-% are rather about one of (i)--(iii), than about \package{tcldoc}.
-%
-%
-%
-% \StopEventually{
-% \begin{thebibliography}{9}
-% \bibitem{ltoutenc}
-% Johannes Braams, David Carlisle, Alan Jeffrey, Frank
-% Mittelbach, Chris Rowley, and Rainer Sch\"opf:
-% \textit{ltoutenc.dtx}, The \LaTeX3 Project;
-% CTAN:\discretionary{}{}{\thinspace}\texttt{macros}\slash
-% \texttt{latex}\slash \texttt{base}\slash \texttt{ltoutenc.dtx}.
-% \bibitem{ltxdoc}
-% David Carlisle:
-% \textit{The file \texttt{ltxdoc.dtx} for use with \LaTeXe},
-% The \LaTeX3 Project;
-% CTAN:\discretionary{}{}{\thinspace}\texttt{macros}\slash
-% \texttt{latex}\slash \texttt{base}\slash \texttt{ltxdoc.dtx}.
-% \bibitem{rtkinenc}
-% Lars Hellstr\"om:
-% \textit{The \package{rtkinenc} package};
-% CTAN:\discretionary{}{}{\thinspace}\texttt{macros}\slash
-% \texttt{latex}\slash \texttt{contrib}\slash
-% \texttt{supported}\slash \texttt{rtkinenc}\slash
-% \texttt{rtkinenc.dtx}.
-% \bibitem{xdoc}
-% Lars Hellstr\"om:
-% \textit{The \package{xdoc} package --- experimental
-% reimplementations of features from \package{doc},
-% second~prototype}, 2000;
-% CTAN:\discretionary{}{}{\thinspace}\texttt{macros}\slash
-% \texttt{latex}\slash \texttt{exptl}\slash \texttt{xdoc}\slash
-% \texttt{xdoc2.dtx}.
-% \bibitem{docstrip}
-% Frank Mittelbach, Denys Duchier, Johannes Braams, Marcin
-% Woli\'nski, and Mark Wooding: \textit{The \textsf{DocStrip}
-% program}, The \LaTeX3 Project;
-% CTAN:\discretionary{}{}{\thinspace}\texttt{macros}\slash
-% \texttt{latex}\slash \texttt{base}\slash \texttt{docstrip.dtx}.
-% \bibitem{doc}
-% Frank Mittelbach, B.~Hamilton Kelly, Andrew Mills, Dave Love, and
-% Joachim \mbox{Schrod}: \textit{The \package{doc} and
-% \package{shortvrb} Packages}, The \LaTeX3 Project;
-% CTAN:\discretionary{}{}{\thinspace}\texttt{macros}\slash
-% \texttt{latex}\slash \texttt{base}\slash \texttt{doc.dtx}.
-% \end{thebibliography}
-% \begin{flushleft}\footnotesize
-% The ``CTAN:'' above is short for ``any of the servers in the
-% Comprehensive \TeX\ Archive Network (or any of their mirrors)''.
-% You get a working URL if you replace this by e.g.\
-% ``\texttt{ftp://ftp.tex.ac.uk/tex-archive/}''.%
-% \end{flushleft}
-% }
-%
-%
-% \changes{v\,1.90}{2000/01/02}{Took a copy of \texttt{pasdoc.dtx} and
-% started modifying it into \texttt{tcldoc.dtx}. (LH)}
-%
-% \section{Initial stuff in the \package{tcldoc} package}
-%
-% The |\NeedsTeXFormat| and |\ProvidesPackage| commands for the
-% \package{tcldoc} package appear at the top of \texttt{tcldoc.dtx} (not
-% shown in the typeset document). Apart from that the only initial
-% action required is to load the \package{xdoc} package (and thereby
-% the \package{doc} package).
-%
-% \begin{macrocode}
-%<*pkg>
-\RequirePackageWithOptions{xdoc2}[2000/11/26]
-% \end{macrocode}
-%
-%
-% \section{Verbatim typesetting of \Tcl\ code}
-%
-% \changes{v\,2.10}{2000/10/03}{Major overhaul of the \texttt{tcl}
-% environment, to make it possible to assign different penalties to
-% different breakpoints. (LH)}
-%
-% The main feature in the \package{tcldoc} package is the \texttt{tcl}
-% environment, which typesets embedded code in a way more adapted to
-% \Tcl\ code than what the \texttt{macrocode} environment does.
-%
-%
-% \subsection{Beginning of line processing}
-%
-% The organization of the text in \emph{lines} is crucial for its
-% interpretation by the \texttt{tcl} environment. In particular the
-% following features rely on being at the beginning of a line:
-% \begin{itemize}
-% \item indentation of \Tcl\ code,
-% \item invoking command mode,
-% \item recognizing \package{docstrip} guards.
-% \end{itemize}
-%
-% \begin{macro}{\TD@percent@token}
-% A control sequence |\let| to a |%|$_{12}$, mainly for use in |\ifx|
-% comparisons.
-% \DoNotIndex{\%}
-% \begin{macrocode}
-\begingroup
- \catcode`\%=12
- \global\let\TD@percent@token=%
-\endgroup
-% \end{macrocode}
-% \end{macro}
-%
-% \begin{macro}{\TD@start@n@line}
-% \begin{macro}{\TD@start@n@line@}
-% \begin{macro}{\TD@start@n@line@i}
-% The macro |\TD@start@n@line| is used at the beginning of every line.
-% The macro first clears the |\TD@line@indent| and |\TD@nesting@level|
-% registers and then it checks if the next character is a percent. If
-% it is then this is a command line. If it isn't then it is a normal
-% line and it is time to start a codeline paragraph and count the
-% character indentation.
-%
-% |\TD@|\-|start@|\-|n@|\-|line@| is the macro which actually does
-% these things; |\TD@|\-|start@|\-|n@|\-|line| is usually |\let| to
-% |\TD@|\-|start@|\-|n@|\-|line@|. When a file is being |\TclInput|ed
-% however, |\TD@|\-|start@|\-|n@|\-|line| is |\let| to
-% |\TD@|\-|start@|\-|n@|\-|line@i| which peeks at the next token
-% before calling |\TD@|\-|start@|\-|n@|\-|line@|. If it didn't,
-% |\TD@|\-|start@|\-|n@|\-|line@| would have a problem at the end of
-% the file, since the end of a file counts as being |\outer|.
-% \begin{macrocode}
-\def\TD@start@n@line@#1{%
- \global\TD@line@indent=\z@
- \TD@nesting@level=\z@
- \ifx #1\TD@percent@token
- \expandafter\TD@module
- \else
- \global\advance \c@codelineno \@ne
- \TD@begin@tclpar
- \expandafter\TD@count@indent \expandafter#1%
- \fi
-}
-% \end{macrocode}
-% \begin{macrocode}
-\let\TD@start@n@line=\TD@start@n@line@
-\def\TD@start@n@line@i{\futurelet\next\TD@start@n@line@}
-% \end{macrocode}
-% \end{macro}\end{macro}\end{macro}
-%
-%
-% \begin{macro}{\TD@count@indent}
-% The purpose of the |\TD@count@indent| macro is to compute the
-% indentation of the current line. It grabs the next character, and
-% as long as that is a space or tab it will just increment
-% |\TD@line@indent| accordingly and continue with the next character.
-% |!| and |"| are used to make space and tab respectively with
-% suitable catcode.
-% \DoNotIndex{\!,\",\ ,\^}
-% \begin{macrocode}
-\begingroup
- \catcode`\!=\active
- \catcode`\"=\active
- \lccode`\!=`\ %
- \lccode`\"=`\^^I%
- \lowercase{%
-\endgroup
- \def\TD@count@indent#1{%
- \ifx !#1%
- \global\advance \TD@line@indent \@ne
- \expandafter\TD@count@indent
- \else \ifx "#1%
- \global\divide \TD@line@indent \TD@tab@size
- \global\advance \TD@line@indent \@ne
- \global\multiply \TD@line@indent \TD@tab@size
- \expandafter\expandafter \expandafter\TD@count@indent
- \else
- \TD@setup@parshape
- \expandafter\expandafter \expandafter#1%
- \fi\fi
- }%
-}
-% \end{macrocode}
-% \end{macro}
-%
-%
-% \begin{macro}{\TD@module}
-% When command mode has been properly established it is time to
-% check if the next character is |<|, in which case the following
-% \package{docstrip} guard must be processed.
-% \changes{v\,2.13}{2000/12/20}{Always check whether the current line
-% is a guard line, since the \texttt{\PrintChar{\number`\<}} in
-% the guard would otherwise start a new paragraph, which is an
-% error. (LH)}
-% \begin{macrocode}
-\def\TD@module{%
- \TD@command
- \futurelet\next
- \TD@ch@angle
-}
-% \end{macrocode}
-% \end{macro}
-%
-%
-% \begin{macro}{\TD@ch@angle}
-% The module checking in \package{tcldoc} takes advantage of the
-% macros for this in \package{doc}---|\ch@plus@etc| is a \package{doc}
-% macro. |<| is active because it is in the nolig list
-% (|\verbatim@|\-|nolig@|\-|list|).
-% \begin{macrocode}
-\begingroup
- \catcode`<=\active
- \gdef\TD@ch@angle{%
- \ifx <\next
- \global\advance \c@codelineno \@ne
- \TD@begin@tclpar
- \expandafter\futurelet \expandafter\next
- \expandafter\ch@plus@etc
- \fi
- }
-\endgroup
-% \end{macrocode}
-% \end{macro}
-%
-% \begin{macro}{\TD@pm@module}
-% The |\TD@pm@module| macro is used instead of \package{doc}'s
-% |\pm@module| when in the \texttt{tcl} environment. This is necessary
-% since the paragraph man\oe uvres of the \texttt{tcl} environment would
-% not work right with \package{doc}'s |\pm@module|.
-%
-% If we're not dealing with a block directive (|*| or |/|), i.e.,
-% it's a single special line, we set everything up to the next |>|
-% appropriately, return to normal mode, possibly change font, and
-% start counting indentation spaces.
-% \DoNotIndex{\>}
-% \begin{macrocode}
-\begingroup
- \catcode`\>=\active
- \gdef\TD@pm@module#1>{%
- \Module{#1}%
- \TD@normal
- \ifnum \guard@level<\c@StandardModuleDepth \else
- \AltMacroFont
- \fi
- \TD@count@indent
- }
-\endgroup
-% \end{macrocode}
-% \end{macro}
-%
-%
-% \begin{macro}{\TD@star@module}
-% \begin{macro}{\TD@slash@module}
-% The |\TD@star@module| and |\TD@slash@module| macros are, like
-% |\TD@pm@module|, used instead of \package{doc}'s |\star@module| and
-% |\slash@module| respectively when in the \texttt{tcl} environment.
-% The only difference is that the |\TD@|\textellipsis\ macros also end
-% the paragraph started in |\TD@ch@angle|, but without that
-% difference, the module guard would incorrectly be reported as
-% ``horizontal material on command line''.
-% \DoNotIndex{\>}
-% \begin{macrocode}
-\begingroup
- \catcode`\>=\active
- \gdef\TD@star@module#1>{%
- \Module{#1}\TD@end@tclpar
- \global\advance \guard@level \@ne
- \ifnum \c@StandardModuleDepth<\guard@level
- \global\let\macro@font=\AltMacroFont
- \macro@font
- \fi
- }
- \gdef\TD@slash@module#1>{%
- \Module{#1}\TD@end@tclpar
- \global\advance \guard@level \m@ne
- \ifnum \guard@level=\c@StandardModuleDepth
- \global\let\macro@font\MacroFont
- \macro@font
- \fi
- }
-\endgroup
-% \end{macrocode}
-% \end{macro}\end{macro}
-%
-%
-% \begin{macro}{\TD@gobble@whitespace}
-% This macro gobbles all normal mode spaces and tabs following it.
-% \begin{macrocode}
-\def\TD@gobble@whitespace#1{%
- \if \ifx #1\TD@active@space
- 0%
- \else
- \ifx #1\TD@active@tab 0\else 1\fi
- \fi 0%
- \expandafter\TD@gobble@whitespace
- \else
- \expandafter#1%
- \fi
-}
-% \end{macrocode}
-% \end{macro}
-%
-%
-%
-% \subsection{Typographical elements}
-%
-% This subsection contains definitions of the typographical
-% elements---special symbols and the like--that are used in the
-% \texttt{tcl} and \texttt{tcl*} environments.
-%
-%
-% \begin{macro}{\TD@typography}
-% \begin{macro}{\TD@nontcl@font}
-% The |\TD@typography| macro does all the typographic set-up that is
-% common to the \texttt{tcl} environment, the \texttt{tcl*}
-% environment, and the |\tclverb| command.
-%
-% The |\TD@nontcl@font| macro selects an ``escape'' font for \Tcl\
-% material. This font is used for various kinds of escape codes that
-% are typeset instead of some character---either because the
-% character is not available, or because the line needs to be broken.
-% The escape font will have the same font family (and same encoding)
-% as the text surrounding the \texttt{tcl} environment or |\tclverb|
-% command. Therefore |\TD@nontcl@font| must be defined before the
-% \Tcl\ material font is selected. I don't think it is necessary to
-% use |\protected@edef| below, but it can't hurt either.
-% \begin{macrocode}
-\def\TD@typography{%
- \protected@edef\TD@nontcl@font{%
- \noexpand\fontencoding{\cf@encoding}%
- \noexpand\fontfamily{\f@family}\noexpand\selectfont
- }%
-% \end{macrocode}
-% The most frequent escape symbol is the single backslash used for
-% escaping the end of lines, and this is stored in the
-% |\TD@backslash@box| box.
-%
-% The following code have to do with how characters outside
-% visible ASCII are typeset when they appear in the \texttt{tcl} or
-% \texttt{tcl*} environment. The commands are defined by the
-% \package{rtkinenc} package.
-% \begin{macrocode}
- \InputModeCode
- \SetUnavailableAction{\leavevmode{%
- \TD@nontcl@font\textbackslash x\TypesetHexNumber{##1}%
- }}%
- \DeclareInputMath{0}{\RIE@undefined{0}}%
- \DeclareInputMath{12}{\RIE@undefined{12}}%
-}
-% \end{macrocode}
-% \end{macro}\end{macro}
-%
-% \begin{macro}{\TD@typeset@space}
-% This macro typesets a space, as it should be in the \texttt{tcl}
-% environment (and other places too). It generates a kern, not glue, so
-% that it can be used in a |\discretionary|.
-% \begin{macrocode}
-\def\TD@typeset@space{\kern\fontdimen\tw@\font}
-% \end{macrocode}
-% \end{macro}
-%
-% \begin{macro}{\TD@discretionary@space}
-% This is the macro which offers all the breakpoints in |\tclverb|
-% code. The name comes from the fact that all the breakpoints are
-% discretionaries which appear to be spaces when not broken. When
-% they are broken, they are backslashes. To give some visual hint
-% that these backslashes need not be backslashes in the source, they
-% are typeset in the same font family (and same encoding) as the
-% text surrounding the \texttt{tcl} environment.
-% \begin{macrocode}
-\def\TD@discretionary@space{%
- \discretionary{\copy\TD@backslash@box}{}{\TD@typeset@space}%
-}%
-% \end{macrocode}
-% \end{macro}
-%
-% \begin{macro}{\TD@breakable@space}
-% The valid breakpoints inside \texttt{tcl} and \texttt{tcl*}
-% environments are made in the |\TD@|\-|breakable@|\-|space| macro.
-% This macro contributes an |\hbox| (which contains a
-% |\TD@typeset@space|), a penalty, and an empty |\vadjust| to the
-% current horizontal list. If a line break is made at the penalty,
-% the |\hbox| will later be removed and replaced by a non-macro font
-% backslash (from |\TD@backslash@box|). The |\vadjust| is there to
-% prevent that any discardable items disappears after the breakpoint.
-% \begin{macrocode}
-\def\TD@breakable@space{%
- \hbox{\TD@typeset@space}%
- \TD@nesting@penalty
- \vadjust{}%
-}
-% \end{macrocode}
-% \end{macro}
-%
-% \begin{macro}{\TD@typeset@tab}
-% This is an unbreakable space whose width equals that of
-% |\TD@tab@size| spaces.
-% \begin{macrocode}
-\def\TD@typeset@tab{\kern\TD@tab@size\fontdimen\tw@\font}
-% \end{macrocode}
-% \end{macro}
-%
-% \begin{macro}{\TD@visible@whitespace}
-% The |\TD@visible@whitespace| macro locally redefines the macros
-% which typesets spaces and tabs to make visible symbols. The visible
-% tab is roughly a |\rightarrowfill| and a |\vrule|, but a difference
-% is that the minuses that form the extensible ``tail'' to the
-% |\rightarrow| are not smashed. Not doing that gives the |\vrule| a
-% suitable height, but I had rather expected the nominal height of
-% the |\rightarrow| to be larger \textellipsis\ It's yet another
-% mysterious math font feature, I suppose.
-%
-% \begin{macrocode}
-\def\TD@visible@whitespace{%
- \def\TD@typeset@space{\char32 }%
- \def\TD@typeset@tab{%
- \hb@xt@\TD@tab@size\fontdimen\tw@\font{%
- $\m@th-\mkern-7mu%
- \cleaders\hbox{$\mkern-2mu-\mkern-2mu$}\hfill
- \mkern-7mu\mathord\rightarrow\mkern-1mu\vrule\mkern1mu$%
- }%
- }%
-}
-% \end{macrocode}
-% \end{macro}
-%
-%
-% \subsection{Catcodes and active characters}
-%
-% Like any \texttt{verbatim}-like environment, the \texttt{tcl}
-% environment does some rather extensive alterations of catcodes. This
-% is handled by the three macros |\TD@general|, |\TD@normal|, and
-% |\TD@command|. Set-up that is common for both modes is made by
-% |\TD@general| when the \texttt{tcl} environment is entered.
-% Mode-dependent set-up is made by |\TD@normal| and |\TD@command|
-% whenever a mode switch occurs.
-%
-% \begin{macro}{\TD@let@active}
-% One technicality here is that all of these routines need to set the
-% definition of some active character, and this character is often not
-% active with normal \LaTeX\ catcodes. To overcome this difficulty,
-% there is a macro |\TD@let@active| which takes two arguments: a
-% control sequence with a single character name, and an arbitrary
-% control sequence. The active character with the same name as the
-% first control sequence is |\let| locally to the second control
-% sequence.
-% \begin{macrocode}
-\begingroup
- \catcode\z@=\active
- \gdef\TD@let@active#1{%
- \begingroup
- \lccode\z@=`#1%
- \lowercase{%
- \endgroup
- \let^^@%
- }%
- }%
-\endgroup
-% \end{macrocode}
-% \end{macro}
-%
-%
-% \begin{macro}{\TD@general}
-% The |\TD@general| macro does the initial set-ups that are common for
-% both modes of the \texttt{tcl} environment and for the |\tclverb|
-% command. These include definitions of various active characters and
-% |\catcode| assignments.
-%
-% \DoNotIndex{\^,\ ,\\,\[,\]}
-% \DoNotIndexHarmless{\PrintChar{`\{}}
-% \DoNotIndexHarmless{\PrintChar{`\}}}
-% \begin{macrocode}
-\def\TD@general{%
- \let\do\do@noligs
- \verbatim@nolig@list
- \let\do\@makeother
- \dospecials
- \catcode`\^^M=\active
- \TD@let@active\^^I\TD@active@tab
- \TD@let@active\^^M\TD@active@CR
- \TD@let@active\ \TD@active@space
- \TD@let@active\\\TD@active@backslash
- \TD@let@active\{\TD@active@braceleft
- \TD@let@active\}\TD@active@braceright
- \TD@let@active\[\TD@active@bracketleft
- \TD@let@active\]\TD@active@bracketright
-}
-% \end{macrocode}
-% \end{macro}
-%
-%
-% \begin{macro}{\TD@normal}
-% This macro sets up the normal mode, changing everything that is
-% different in any of the other modes. Note that the catcodes in
-% normal mode are such that no character is a space. This makes it
-% possible for macros used in normal mode to look ahead by simply
-% grabbing the next character in an undelimited argument.
-% \DoNotIndex{\^,\ ,\\,\[,\]}
-% \DoNotIndexHarmless{\PrintChar{`\{}}
-% \DoNotIndexHarmless{\PrintChar{`\}}}
-% \begin{macrocode}
-\def\TD@normal{%
- \catcode`\^^I=\active
- \catcode`\ =\active
- \catcode`\\=\active
- \catcode`\{=\active
- \catcode`\}=\active
- \catcode`\[=\active
- \catcode`\]=\active
-}
-% \end{macrocode}
-% \end{macro}
-%
-% \begin{macro}{\TD@command}
-% This macro switches from normal mode to command mode.
-% \DoNotIndex{\^,\ ,\\,\[,\]}
-% \DoNotIndexHarmless{\PrintChar{`\{}}
-% \DoNotIndexHarmless{\PrintChar{`\}}}
-% \begin{macrocode}
-\def\TD@command{%
- \catcode`\^^I=10%
- \catcode`\ =10%
- \catcode`\[=12%
- \catcode`\]=12%
- \catcode`\\=\z@
- \catcode`\{=\@ne
- \catcode`\}=\tw@
-}
-% \end{macrocode}
-% \end{macro}
-%
-%
-% \begin{macro}{\TD@active@space}
-% \begin{macro}{\TD@active@space@}
-% This macro is used for spaces in normal mode. Its main problem is
-% to determine whether it is followed by more whitespace or not. If it
-% is, then it is an unbreakable space. If it isn't, then it is a
-% breakable space.
-% \begin{macrocode}
-\def\TD@active@space#1{%
- \ifx #1\TD@active@backslash
- \expandafter\TD@active@space@
- \else
- \ifx #1\TD@active@space
- \TD@typeset@space
- \else\ifx #1\TD@active@tab
- \TD@typeset@space
- \else
- \TD@breakable@space
- \fi\fi
- \expandafter#1%
- \fi
-}
-% \end{macrocode}
-% The special case to look out for is that the next character is a
-% backslash which escapes a newline.
-% \begin{macrocode}
-\def\TD@active@space@#1{%
- \ifx #1\TD@active@CR
- \TD@typeset@space
- \else
- \TD@breakable@space
- \fi
- \TD@active@backslash #1%
-}
-% \end{macrocode}
-% \end{macro}\end{macro}
-%
-% \begin{macro}{\TD@active@tab}
-% The |\TD@active@tab| macro is what the active character tab is
-% |\let| to inside the \texttt{tcl} and \texttt{tcl*} environments.
-% \begin{macrocode}
-\def\TD@active@tab{\TD@typeset@tab}
-% \end{macrocode}
-% \end{macro}
-%
-% \begin{macro}{\TD@active@CR}
-% This macro is called at the end of a line. The |\catcode| tests the
-% current mode.
-% \DoNotIndex{\\}
-% \begin{macrocode}
-\def\TD@active@CR{%
- \ifnum \catcode`\\=\z@
- \ifvmode\else
- \PackageError{tcldoc}{Horizontal material on command line}\@ehc
- \@@par
- \fi
- \TD@normal
- \else
- \TD@end@tclpar
- \fi
- \TD@start@n@line
-}
-% \end{macrocode}
-% \end{macro}
-%
-% \begin{macro}{\TD@active@CRv}
-% This is what |\tclverb| defines the |^^M| active character as
-% being. Cf.\ the standard \LaTeX\ macro |\verb@eol@error|.
-% \begin{macrocode}
-\def\TD@active@CRv{%
- \verb@egroup
- \PackageError{tcldoc}{\protect\tclverb\space ended by end of line}%
- \@ehc
-}
-% \end{macrocode}
-% \end{macro}
-%
-%
-% \begin{macro}{\TD@active@backslash}
-% \changes{v\,2.10}{2000/10/08}{First checking if next character is
-% active, and only expands the macro containing the detailed tests
-% if it is. (LH)}
-% \begin{macro}{\TD@active@backslash@}
-% \begin{macro}{\TD@active@backslash@v}
-% |\TD@active@backslash| is used for backslashes in normal mode. Its
-% main problem is to determine whether the backslash is escaping
-% something (which would normally be treated specially by the
-% environment) or not. The primary test for this is whether the
-% argument grabbed is an active character, and if it is then
-% processing continues (in |\TD@active@backslash@|) to determine what
-% should be done.
-% \begin{macrocode}
-\def\TD@active@backslash#1{%
- \ifcat \noexpand#1\noexpand~%
- \expandafter\TD@active@backslash@
- \else
- \@backslashchar
- \fi
- #1%
-}
-% \end{macrocode}
-%
-% \changes{v\,2.11}{2000/11/06}{Escaped spaces are no longer breakable.
-% (LH)}
-% \begin{macrocode}
-\def\TD@active@backslash@#1{%
- \ifcase
- \ifx #1\TD@active@CR 0%
- \else\ifx #1\TD@active@backslash 1%
- \else\ifx #1\TD@active@braceleft 2%
- \else\ifx #1\TD@active@braceright 2%
- \else\ifx #1\TD@active@bracketleft 2%
- \else\ifx #1\TD@active@bracketright 2%
- \else\ifx #1\TD@active@space 3%
- \else 4\fi\fi\fi\fi\fi\fi\fi
- \space
- \expandafter\TD@active@backslash@i
- \or
- \@backslashchar\@backslashchar
- \or
- \@backslashchar \string#1%
- \or
- \@backslashchar \TD@typeset@space
- \else
- \@backslashchar
- \expandafter#1%
- \fi
-}
-% \end{macrocode}
-% Escaped newlines that are followed by command mode lines are not
-% converted to discretionary spaces.
-% \begin{macrocode}
-\def\TD@active@backslash@i#1{%
- \ifx #1\TD@percent@token
- \copy\TD@backslash@box
- \expandafter\TD@active@CR
- \else
- \TD@breakable@space
- \global\advance \c@codelineno \@ne
- \expandafter\TD@gobble@whitespace
- \fi
- #1%
-}
-% \end{macrocode}
-% The |\tclverb| command uses |\TD@active|\-|@backslash@v| instead of
-% |\TD|\-|@active|\-|@backslash@i|, since it is intended for use on
-% lines that may begin with |%|.
-% \begin{macrocode}
-\def\TD@active@backslash@v#1{%
- \TD@discretionary@space
- \ifx #1\TD@percent@token
- \expandafter\@firstoftwo
- \fi
- \TD@gobble@whitespace #1%
-}
-% \end{macrocode}
-% \end{macro}\end{macro}\end{macro}
-%
-%
-% \begin{macro}{\TD@active@braceleft}
-% \begin{macro}{\TD@active@braceright}
-% \begin{macro}{\TD@active@bracketleft}
-% \begin{macro}{\TD@active@bracketright}
-% These macros mainly typeset the corresponding characters, but they
-% also increase or decrease the |\TD@nesting@level| count by one.
-% \DoNotIndex{\[,\]}
-% \DoNotIndexHarmless{\PrintChar{`\{}}
-% \DoNotIndexHarmless{\PrintChar{`\}}}
-% \begin{macrocode}
-\begingroup
- \catcode`\{=12 \catcode`\}=12
- \catcode`\[=1 \catcode`\]=2
- \gdef\TD@active@braceleft[{\advance\TD@nesting@level\@ne]
- \gdef\TD@active@braceright[}\advance\TD@nesting@level\m@ne]
-\endgroup
-% \end{macrocode}
-% \begin{macrocode}
-\def\TD@active@bracketleft{[\advance\TD@nesting@level\@ne}
-\def\TD@active@bracketright{]\advance\TD@nesting@level\m@ne}
-% \end{macrocode}
-% \end{macro}\end{macro}\end{macro}\end{macro}
-%
-%
-%
-% \subsection{Paragraph formatting}
-%
-% This subsection contains definitions of control sequences that are
-% mainly related to the formatting of paragraphs.
-%
-% The new implementation of the \texttt{tcl} and \texttt{tcl*}
-% environments is based on inserting the ``discretionary'' backslashes
-% at the linebreaks \emph{after} the paragraph has been broken.
-% Doing this involves reboxing every line in the paragraph, and in this
-% process the line adjustments are changed as well.
-%
-% \begin{macro}{\TD@backslash@box}
-% The |\TD@backslash@box| box register is set to a backslash in the
-% special |\TD@nontcl@font| at the beginning of each \texttt{tcl} and
-% \texttt{tcl*} environment. These backslashes are used to denote
-% ``backslash escaping newline''. The primary reason for keeping this
-% glyph in a box is not that this might be slightly faster to
-% typeset, but to have the \emph{width} of it easily accessible.
-% \begin{macrocode}
-\newbox\TD@backslash@box
-% \end{macrocode}
-% \end{macro}
-%
-% \begin{macro}{\TD@nesting@level}
-% \begin{macro}{\TD@nesting@penalty}
-% |\TD@nesting@level| is a count register which keeps track of how
-% many brace\slash bracket groups surrounds the current position in
-% the line. It is cleared to zero at the beginning of each paragraph.
-% \begin{macrocode}
-\newcount\TD@nesting@level
-% \end{macrocode}
-% |\TD@nesting@penalty| makes a |\penalty| whose value should depend
-% on the nesting level (the deeper the nest, the larger the penalty).
-% It may also adjust the value of |\linepenalty| to counter that
-% the nesting level has become negative. This is typically needed for
-% lines which say things like ``|} else {|''. The default definition is
-% to make the nesting penalty $100$ times the nesting level and to
-% make the |\linepenalty| equal to
-% \(10 - \min\{0,\mbox{least penalty}\}\).
-% \begin{macrocode}
-\def\TD@nesting@penalty{%
- \penalty \the\TD@nesting@level 00\relax
- \ifnum \lastpenalty<-\linepenalty
- \linepenalty=10%
- \advance \linepenalty -\lastpenalty
- \fi
-}
-% \end{macrocode}
-% I doubt that this is necessarily the best algorithm for computing
-% these penalties, so if you are really thorough about your
-% typography you may well want to experiment with other definitions of
-% |\TD@nesting@penalty|. In case you should then find something which
-% works even better then I'm interested to learn about it.
-% \end{macro}\end{macro}
-%
-% \begin{macro}{\TD@reformat@lines}
-% The |\TD@reformat@lines| macro calls itself recursively to reformat
-% all lines on the current vertical list. The first line will remain
-% flush left, but all other lines will be reset flush right. The
-% visible material on the last line will be left as it is, but the
-% last box in all other lines will be replaced by a non-macro font
-% backslash.
-%
-% It is very important that the current vertical list is not the main
-% vertical list.
-% It is assumed that the current vertical list consists of a sequence
-% of \meta{box}, \meta{penalty}, \meta{glue}, with an extra glue item
-% at the top of the list. It is OK if some penalty or glue item is
-% missing. In case the list contains other material as well the line
-% reformatting may be stopped prematurely, but there is a trick that
-% allows one to put arbitrary material between the lines of the
-% reformatted paragraph: rather than doing e.g.
-% \begin{quote}
-% |\mark|\marg{text}
-% \end{quote}
-% in the paragraph, do
-% \begin{quote}
-% |\vadjust{\vbox{\mark{|\meta{text}|}}}|
-% \end{quote}
-% The |\vbox| will be recognised by the paragraph reformatting
-% mechanism as a container for vertical mode material that appears
-% between the lines of the paragraph, so it will simply be unboxed.
-% \changes{v\,2.12}{2000/11/11}{\cs{vbox} containers for vertical
-% material are allowed between the lines of a reformatted
-% paragraph. (LH)}
-%
-% Each line's horizontal list ends with
-% \begin{itemize}
-% \item a box (which contains the space that is to be replaced by a
-% backslash),
-% \item a penalty (at which the paragraph was broken), and
-% \item a glue item (the |\rightskip|).
-% \end{itemize}
-%
-% A tricky feature in the implementation is that the
-% |\bgroup|--|\egroup| nesting will be off by one. The |\bgroup| at
-% the beginning of a |\TD@reformat@line| will be matched by the
-% |\egroup| at the end of the |\TD@reformat@line| that the first one
-% calls!
-% \begin{macrocode}
-\def\TD@reformat@lines{%
- \bgroup
- \unskip
- \count@=\lastpenalty \unpenalty
- \setbox\z@=\lastbox
- \ifvoid\z@
-% \end{macrocode}
-% The recursion had already descended down to the last line of the
-% paragraph, and it is now time to reformat it.
-% \begin{macrocode}
- \egroup
- \prevdepth=\TD@prevdepth
- \hbox{%
- \unhbox\z@
- \unskip \unpenalty
- \setbox\z@=\lastbox
- \copy\TD@backslash@box
- }%
- \else
-% \end{macrocode}
-% Else there may be another line, and the |\TD@reformat@lines| recursion
-% must continue to descend. Upon return the box currently in box
-% register zero must be reformatted as a non-first line (flush
-% right) and it cannot be the last line in the paragraph, so it is
-% always correct to replace the last box by a backslash.
-% \begin{macrocode}
- \TD@reformat@lines
- \ifvbox\z@ \unvbox\z@ \else
- \hb@xt@\dimen@{%
- \hfill
- \unhbox\z@
- \unskip \unpenalty
- \setbox\z@=\lastbox
- \copy\TD@backslash@box
- }%
- \fi
- \fi
- \ifnum \count@=\z@ \else \penalty\count@ \fi
- \egroup
-}
-% \end{macrocode}
-% \end{macro}
-%
-% \begin{macro}{\TD@reformat@par}
-% The |\TD@reformat@par| macro reformats all lines (they're supposed
-% to constitute a paragraph, but that isn't so important) in the
-% current vertical list. The restrictions of |\TD@reformat@lines| on
-% what may appear in the list apply and there must be at least two
-% lines in the list. |\dimen@| is used to hold the desired width of
-% reformatted paragraphs.
-%
-% More precisely |\TD@reformat@par| takes care of the last line of
-% the paragraph and the possible |\vbox| containers for vertical
-% material that may follow it. Everything in the paragraph that comes
-% before the last line is handled by |\TD@reformat@lines|.
-% \changes{v\,2.12}{2000/11/11}{\cs{vbox} containers for vertical
-% material are allowed after the last line of a reformatted
-% paragraph. (LH)}
-% \begin{macrocode}
-\def\TD@reformat@par{%
- \unskip
- \count@=\lastpenalty \unpenalty
- \setbox\z@=\lastbox
- \ifvbox\z@
- \bgroup
- \TD@reformat@par
- \egroup
- \unvbox\z@
- \else\ifnum \prevgraf>\@ne
- \dimen@=\@totalleftmargin
- \advance \dimen@ \linewidth
- \bgroup
- \unskip
- \count@=\lastpenalty \unpenalty
- \setbox\z@=\lastbox
- \TD@reformat@lines
- \hb@xt@\dimen@{\hfill \unhbox\z@ \unskip}%
- \else
- \unskip
- \prevdepth=\TD@prevdepth
- \box\z@
- \fi\fi
- \ifnum \count@=\z@ \else \penalty\count@ \fi
-}
-% \end{macrocode}
-% \end{macro}
-%
-% \begin{macro}{\TD@prevdepth}
-% |\TD@prevdepth| is a macro which is used for storing the value of
-% |\prevdepth| at times where \TeX\ modifies the primitive in unwanted
-% ways. It should always be set globally.
-% \end{macro}
-%
-%
-% \begin{macro}{\TD@begin@tclpar}
-% The |\TD@begin@tclpar| macro is called when a paragraph in a
-% \texttt{tcl} or \texttt{tcl*} environment is about to start. It
-% takes care of setting up things so that the paragraph can later be
-% reformatted using |\TD@reformat@par|, but it also has to make sure
-% that this reformatting doesn't affect the way the paragraph blends
-% in with vertical material before and after it.
-%
-% Reformatting requires that the paragraph is first built in
-% restricted vertical mode, i.e., it has to be built in an explicit
-% |\vbox|. A problem with this is however that it changes the value of
-% |\prevdepth|, which must therefore be explicitly restored.
-% \begin{macrocode}
-\def\TD@begin@tclpar{%
- \xdef\TD@prevdepth{\the\prevdepth}%
- \setbox\z@=\vbox\bgroup
- \color@begingroup
- \prevdepth=\TD@prevdepth
- \indent
-}
-% \end{macrocode}
-% \end{macro}
-%
-% \begin{macro}{\TD@end@tclpar}
-% The |\TD@end@tclpar| macro ends a paragraph begun by
-% |\TD@begin@tclpar|, reformats it (|\TD@reformat@par|), and
-% contributes it to the surrounding vertical list. The |\begingroup|
-% and |\endgroup| are there to sort things out in case the recursion
-% in |\TD@reformat@par| fails to match as intended. The second
-% |\@@par| sees to that the page builder is exercised (without it,
-% several pages may go onto the main vertical list without anything
-% being shipped out).
-% \begin{macrocode}
-\def\TD@end@tclpar{%
- \@@par
- \begingroup
- \skip@=\lastskip
- \TD@reformat@par
- \vskip\skip@
- \endgroup
- \xdef\TD@prevdepth{\the\prevdepth}%
- \color@endgroup
- \egroup
- \unvbox\z@
- \prevdepth=\TD@prevdepth
- \@@par
-}
-% \end{macrocode}
-% \end{macro}
-%
-%
-% \begin{macro}{\TD@line@indent}
-% |\TD@line@indent| is a |\count| register holding the indentation of
-% the current line, in number of spaces. |\TD@line@indent| should
-% always be assigned globally.
-% \begin{macrocode}
-\newcount\TD@line@indent
-% \end{macrocode}
-% \end{macro}
-%
-% \begin{macro}{\TD@par@setup}
-% This macro sets up various paragraph formatting parameters for the
-% normal mode. It mainly consists of code from |\macro@code| of
-% \package{doc}.
-% \begin{macrocode}
-\def\TD@par@setup{%
-% \end{macrocode}
-%
-% In theory a \texttt{tcl} environment should consist of a
-% \texttt{trivlist} environment, but the empty space before and after
-% the environment should not be too large.
-% \begin{macrocode}
- \topsep\MacrocodeTopsep
-% \end{macrocode}
-%
-% The next parameter we set is |\@beginparpenalty|, in order
-% to prevent a page break before such an environment.
-% \begin{macrocode}
- \@beginparpenalty\predisplaypenalty
-% \end{macrocode}
-%
-% We then start a |\trivlist|, set |\parskip| back to
-% zero and start an empty |\item|.
-% \changes{ \package{doc} v\,1.9b}{1993/12/03}{Forcing any label
-% from macro env.}
-% \begin{macrocode}
- \if@inlabel\leavevmode\fi
- \trivlist
- \parskip\z@skip
- \item[]%
-% \end{macrocode}
-% However, the \texttt{trivlist} environment is only started to make
-% standard \LaTeX\ happy (and to have the various list-related
-% measurements properly calculated). Everything below will instead
-% by done with \TeX\ primitives.
-% \changes{ \package{pasdoc} v\,1.31}{1999/09/02}{\cs{parshape}
-% assignment added. (LH)}
-% \changes{v\,2.10}{2000/10/08}{\cs{parshape} assignment removed (now
-% done in \cs{TD@begin@tclpar}). (LH)}
-%
-% It is important that the change to a special font does not take
-% place before the above |\item|, otherwise a change to
-% |\baselineskip| will affect the paragraph above.
-% \begin{macrocode}
- \macro@font
- \frenchspacing
-% \end{macrocode}
-%
-% \leavevmode
-% \changes{v\,2.12}{2000/11/11}{Assignment to \cs{TD@backslash@box}
-% moved from \cs{TD@typography} to here, so that the size will be
-% right. (LH)}^^A
-% One escape symbol that is particularly common is a single
-% backslash, and this is stored in the |\TD@backslash@box| box. As
-% some encodings (e.g.\ \texttt{OT1}) does not contain a backslash,
-% these escape backslashes are made with |\textbackslash|, rather
-% than an explicit character.
-% \begin{macrocode}
- \sbox\TD@backslash@box{\TD@nontcl@font\textbackslash}%
-% \end{macrocode}
-%
-% The |\rightskip| is given a generous stretchability of |\linewidth|
-% so that it doesn't matter too much if the lengths of lines in the
-% paragraph varies. Decreasing this value would make the nesting in
-% the code less important.
-% \begin{macrocode}
- \parindent=\@totalleftmargin
- \advance \parindent \MacroIndent
- \leftskip=\z@skip
- \rightskip=\z@ \@plus \linewidth\relax
-% \end{macrocode}
-% The next two lines are from the definition of the
-% \texttt{macrocode} environment, and I (LH) have no idea what they are
-% good for, but I suppose I can keep them, at least for now.
-% \begin{macrocode}
- \global\@newlistfalse
- \global\@minipagefalse
-% \end{macrocode}
-% \changes{ \package{pasdoc} v\,1.31}{1999/09/02}{Made list
-% environments indent line numbers. (LH)}
-% Line numbers are inserted using |\everypar|.
-% \begin{macrocode}
- \ifcodeline@index
- \everypar={\llap{\PrintCodelineNo\ }}%
- \else
- \everypar={}%
- \fi
-% \end{macrocode}
-%
-% These commands replace \package{doc}'s final module processing
-% macros with macros that work in the \texttt{tcl} environment.
-% \begin{macrocode}
- \let\pm@module=\TD@pm@module
- \let\star@module=\TD@star@module
- \let\slash@module=\TD@slash@module
-}
-% \end{macrocode}
-% \end{macro}
-%
-% \begin{macro}{\TD@setup@parshape}
-% This macro sets the shape of the current paragraph, primarily based
-% on the value in |\TD@line@indent|, and makes the final indentation
-% of the first line. By the time it gets called, there usually already
-% is some material (|\indent|, codeline number, and possibly a
-% \package{docstrip} guard) in the paragraph horizontal list.
-%
-% The basic area in which the paragraph material is to be put has a
-% left margin at |\@totalleftmargin|${}+{}$|\MacroIndent| and a right
-% margin at |\@totalleftmargin|${}+{}$|\linewidth| (in both cases the
-% position is that much to the right of the left edge of the galley).
-% The printed codeline number (if there is one) will be put to the
-% left of the left margin, but all other material is indented at least
-% the width of another |\TD@line@indent| spaces. In the first line
-% that will be done through an explicit kern, but in subsequent lines
-% it will instead be done through a |\parshape| assignment. On these
-% subsequent lines, there will be an additional indentation of\/
-% $1\,\mathrm{em}$ to mark the continuation.
-%
-% When the paragraph is broken, the right margin will be moved to the
-% right by the width of a space minus the width of the
-% |\TD@backslash@box|. The reason for this is that the paragraph will
-% be broken with spaces (from |\TD@breakable@space|) at the end of
-% lines, but that these spaces will later be replaced by the
-% backslash in |\TD@backslash@box|. As that doesn't happen in the
-% very last line of the paragraph however, something special must be
-% done there. The option chosen is to give the |\parfillskip| a
-% nonzero natural width.
-%
-% The length |\@totalleftmargin|${}+{}$|\MacroIndent| is put in
-% |\parindent| at the beginning of the environment.
-% \begin{macrocode}
-\def\TD@setup@parshape{%
- \parfillskip=\fontdimen\tw@\font \@plus 1fil%
- \advance \parfillskip -\wd\TD@backslash@box
- \dimen@=\MacroIndent
- \advance \dimen@ \TD@line@indent\fontdimen\tw@\font
- \advance \dimen@ 1em%
- \dimen@ii=\linewidth
- \advance \dimen@ii -\dimen@
- \dimen4=\linewidth
- \advance \dimen@ \@totalleftmargin
- \advance \dimen4 \@totalleftmargin
- \advance \dimen@ii -\parfillskip
- \advance \dimen4 -\parfillskip
- \parshape \tw@ \z@ \dimen4 \dimen@ \dimen@ii
-% \end{macrocode}
-% Then one just needs one more kern to indent the first line
-% properly.
-% \begin{macrocode}
- \kern \TD@line@indent\fontdimen\tw@\font
-}
-% \end{macrocode}
-% \end{macro}
-%
-% \begin{macro}{\TD@trace@pardimen}
-% This macro is for tracing dimens that are of interest for paragraph
-% breaking. Its call syntax is
-% \begin{quote}
-% |\TD@trace@pardimen|\marg{text}\marg{dimen}
-% \end{quote}
-% where \meta{text} will be written to the log file to identify the
-% dimension, and \meta{dimen} is the dimen register in question (it
-% can be anything that can appear after |\the|).
-%
-% To reduce the amount of text written, this macro only does anything
-% if the \TeX\ parameter |\tracingparagraphs| is positive.
-% \begin{macrocode}
-%<*partrace>
-\def\TD@trace@pardimen#1#2{%
- \ifnum \tracingparagraphs>\z@
- \immediate\write\m@ne{#1 \the#2}%
- \fi
-}
-%</partrace>
-% \end{macrocode}
-% \changes{ \package{pasdoc} v\,1.21}{1999/08/15}{Added macro and
-% some calls of it. All of those are however surrounded by
-% \texttt{partrace} guards. (LH)}
-% \end{macro}
-%
-%
-%
-% \subsection{User interface}
-%
-% This subsection defines the actual environment and the
-% commands and parameters intended to be user-definable.
-%
-%
-% \begin{macro}{\settabsize}
-% \changes{v\,2.12}{2000/11/09}{New command for setting the tab size.
-% (LH)}
-% \begin{macro}{\TD@tab@size}
-% \changes{v\,2.12}{2000/11/09}{New name for \cs{c@tabSize}; no longer a
-% count register. (LH)}
-% The |\TD@tab@size| control sequence stores (as a |\chardef| token)
-% the number of spaces that are equivalent to one tab; the default
-% value is 8. To set this number, one uses the command |\settabsize|,
-% which has the syntax
-% \begin{quote}
-% |\settabsize|\marg{new size}
-% \end{quote}
-% The effect of |\settabsize| is local.
-% \begin{macrocode}
-\newcommand*\settabsize[1]{\chardef\TD@tab@size=#1\relax}
-\settabsize{8}
-% \end{macrocode}
-% \end{macro}\end{macro}
-%
-%
-% \begin{environment}{tcl}
-% \begin{environment}{tcl*}
-% This is some kind of a definition. There are aspects of the
-% interaction with \texttt{trivlist} that I (LH) don't understand,
-% but it appears to work.
-% \begin{macrocode}
-\newenvironment{tcl}{%
- \TD@general
- \TD@typography
- \TD@par@setup
- \TD@command
-}{%
- \global\@inlabelfalse
- \endtrivlist
-}
-% \end{macrocode}
-% \begin{macrocode}
-\newenvironment{tcl*}{%
- \TD@visible@whitespace
- \TD@general
- \TD@typography
- \TD@par@setup
- \TD@command
-}{%
- \global\@inlabelfalse
- \endtrivlist
-}
-% \end{macrocode}
-% \end{environment}\end{environment}
-%
-% \begin{macro}{\tclverb}
-% \begin{macro}{\TD@verb}
-% The |\tclverb| macro is for typesetting short passages of \Tcl\ code
-% which one does not want to have a special paragraph for. It is
-% similar to the standard \LaTeX\ command |\verb|, but it has a couple
-% of extra features. One is that spaces that are not followed by more
-% whitespace are discretionary breakpoints, just as in the \texttt{tcl}
-% environment. Another is that a newline can be escaped by putting a
-% backslash at the end of an input line. Such an escaped newline will
-% count as a space, and whitespace at the beginning of the line
-% following it will be ignored. A percent character immediately
-% following an escaped newline will also be ignored.
-%
-% |\tclverb| has a |*| form |\tclverb*|, which is analogous to the
-% standard \LaTeX\ |\verb*|, i.e., spaces are typeset using ``visible
-% space'' symbols.
-% \begin{macrocode}
-\newcommand\tclverb{%
- \relax\ifmmode\hbox\else\leavevmode\null\fi
- \bgroup
- \@ifstar{\TD@visible@whitespace\TD@verb}\TD@verb
-}
-% \end{macrocode}
-% \DoNotIndex{\ ,\^,\\}
-% \begin{macrocode}
-\def\TD@verb{%
- \let\TD@active@backslash@i=\TD@active@backslash@v
- \let\TD@active@CR=\TD@active@CRv
- \let\TD@breakable@space=\TD@discretionary@space
- \TD@general
- \catcode`\^^I=\active
- \catcode`\ =\active
- \catcode`\\=\active
- \TD@typography
- \verbatim@font
- \sbox\TD@backslash@box{\TD@nontcl@font\textbackslash}%
- \@sverb
-}
-% \end{macrocode}
-% \end{macro}\end{macro}
-%
-%
-% \begin{macro}{\TclInput}
-% The |\TclInput| macro is very much like the standard \LaTeX\
-% |\input|, but it is intended to be used in command mode, for
-% inputting \Tcl\ files that have not been marked up. |\input| would
-% work in this case, but there would be errors at the first and last
-% lines of the |\input|ted file.
-% \begin{macrocode}
-\newcommand\TclInput[1]{%
- \IfFileExists{#1}{%
- \@addtofilelist{#1}%
- \begingroup
- \TD@normal
- \let\TD@start@n@line=\TD@start@n@line@i
- \expandafter\TD@start@n@line
- \@@input\@filef@und\TD@percent@token
- \endgroup
- }{\PackageError{tcldoc}{No file #1}\@eha}%
-}
-% \end{macrocode}
-% \end{macro}
-%
-% \begin{macro}{\MakeShortTclverb}
-% This definition follows \package{doc}'s |\MakeShortVerb| pretty
-% close.
-% \DoNotIndex{\~}
-% \begin{macrocode}
-\newcommand\MakeShortTclverb[1]{%
- \expandafter\ifx \csname cc\string#1\endcsname\relax
- \PackageInfo{tcldoc}{%
- Made \expandafter\@gobble\string#1 a short %
- reference for \string\tclverb}%
- \add@special{#1}%
- \expandafter\xdef \csname cc\string#1\endcsname{\the\catcode`#1}%
- \begingroup
- \lccode`\~=`#1%
- \lowercase{%
- \global\expandafter\let
- \csname ac\string#1\endcsname ~%
- \gdef~{\tclverb~}}%
- \endgroup
- \global\catcode`#1\active
- \else
- \PackageInfo{tcldoc}{%
- \expandafter\@gobble\string#1 already a short verb %
- reference}%
- \fi
-}
-% \end{macrocode}
-% \end{macro}
-%
-%
-% \section{Miscellaneous markup features}
-%
-% \subsection{Namespaces}
-%
-% \changes{v\,2.10}{2000/10/13}{Added namespace support. (LH)}
-%
-% \begin{macro}{\setnamespace}
-% \begin{macro}{\TD@namespace}
-% Since it is common that many identifiers in the same namespace are
-% defined in sequence, one can specify a default namespace to use for
-% all commands where no explicit namespace is given. This is done
-% using the |\setnamespace| command, whose syntax is
-% \begin{quote}
-% |\setnamespace|\marg{namespace}
-% \end{quote}
-% This converts the \meta{namespace} to a harmless character string
-% and locally assigns it to the |\TD@namespace| macro, which stores
-% the current default namespace.
-% \begin{macrocode}
-\newcommand\setnamespace[1]{\MakeHarmless\TD@namespace{#1}}
-% \end{macrocode}
-% At the beginning of a document, the default namespace is the global
-% namespace, which is represented by the empty string.
-% \begin{macrocode}
-\let\TD@namespace\@empty
-% \end{macrocode}
-% \end{macro}\end{macro}
-%
-% \begin{macro}{\buildname}
-% \changes{v\,2.13}{2000/12/16}{Added \cs{@empty} after test for
-% \cs{NoValue}, so that it works for empty arguments. (LH)}
-% \begin{macro}{\namespacesepator}
-% \changes{v\,2.12}{2000/11/25}{Command added. (LH)}
-% The |\buildname| macro builds a qualified name from a namespace and
-% a name tail. It has the syntax
-% \begin{quote}
-% |\buildname|\marg{namespace}\marg{name}
-% \end{quote}
-% where \meta{name} is a harmless character sequence and
-% \meta{namespace} is a harmless character sequence or the token
-% |\NoValue|. If \meta{namespace} is |\NoValue| then the default
-% namespace is used. |\buildname| does all processing at expand-time.
-%
-% If the namespace is the empty string (the global namespace) then
-% \meta{name} is returned without a `|::|' prefix; this is for the
-% sake of compatibility with pre-v\,8 \Tcl s.
-% \begin{macrocode}
-\newcommand\buildname[2]{%
- \ifx \NoValue#1\@empty
- \ifx \@empty\TD@namespace \else
- \TD@namespace\namespaceseparator
- \fi
- \else
- \ifx $#1$\else #1\namespaceseparator \fi
- \fi
- #2%
-}
-% \end{macrocode}
-% The |\namespaceseparator| command typesets a namespace separator. By
-% default it is two colons separated by a penalty, but the user may
-% redefine it.
-% \begin{macrocode}
-\DeclareRobustCommand\namespaceseparator{:\penalty\hyphenpenalty:}
-% \end{macrocode}
-% \end{macro}\end{macro}
-%
-%
-% \subsection{\texttt{macro}-like environments}
-%
-% This subsection contains the definitions of a couple of environments
-% which, like \package{doc}'s \texttt{macro}, are for marking up the
-% document (what is defined and where). These environments are defined
-% using the |\New|\-|Macro|\-|Environment| command of the \package{xdoc2}
-% package.
-% \changes{ \package{pasdoc} v\,1.40}{1999/12/05}{\texttt{macro}-like
-% environments redefined. (LH)}
-% \changes{v\,2.00}{2000/07/17}{\texttt{macro}-like environments are
-% defined using the \cs{NewMacroEnvironment} command. The
-% \cs{TD@identifier}, \cs{SpecialMainTclIndex}, and
-% \cs{PrintTclName} macros were removed. (LH)}
-% \changes{v\,2.12}{2000/11/26}{Using starred form of
-% \cs{NewMacroEnvironment} for the \texttt{variable}, \texttt{proc},
-% \texttt{arrayvar}, and \texttt{arrayentry} environments, so that
-% the marginal headings can be broken. (LH)}
-%
-%
-% \begin{environment}{variable}
-% The \texttt{variable} environment has the syntax
-% \begin{quote}
-% |\begin{variable}|\oarg{namespace}\marg{variable}
-% \end{quote}
-% It makes a marginal heading, index entry, etc.\@ for the
-% \meta{variable} variable in the \meta{namespace} namespace
-% (if specified, otherwise in the default namespace).
-% \begin{macrocode}
-\NewMacroEnvironment*{variable}%
- {\XD@grab@harmless@oarg\XD@grab@harmless\relax}{2}%
- {\XDParToMargin{\MacroFont\buildname{#1}{#2} \normalfont (var.)}}%
- {\TD@main@index{#1}{#2}{var.}{}}%
- {{\buildname{#1}{#2}}{\texttt{\buildname{#1}{#2}} variable}}
- {}%
-% \end{macrocode}
-% \end{environment}
-%
-% \begin{environment}{proc}
-% The \texttt{proc} environment has the syntax
-% \begin{quote}
-% |\begin{proc}|\oarg{namespace}\marg{proc}
-% \end{quote}
-% It makes a marginal heading, index entry, etc.\@ for the \meta{proc}
-% procedure in the \meta{namespace} namespace (if specified, otherwise
-% in the default namespace).
-% \begin{macrocode}
-\NewMacroEnvironment*{proc}%
- {\XD@grab@harmless@oarg\XD@grab@harmless\relax}{2}%
- {\XDParToMargin{\MacroFont\buildname{#1}{#2} \normalfont (proc)}}%
- {\TD@main@index{#1}{#2}{proc}{}}%
- {{\buildname{#1}{#2}}{\texttt{\buildname{#1}{#2}} proc}}
- {}%
-% \end{macrocode}
-% \end{environment}
-%
-% \begin{environment}{arrayvar}
-% \changes{v\,2.00}{2000/07/17}{Environment added. (LH)}
-% \changes{v\,2.12}{2000/11/09}{Made last argument optional. (LH)}
-% \changes{v\,2.13}{2000/12/15}{Not using \cs{meta} for the
-% \meta{index-des} argument. (LH)}
-% \begin{environment}{arrayentry}
-% \changes{v\,2.02}{2000/08/21}{Environment added. (LH)}
-% The \texttt{arrayvar} and \texttt{arrayentry} environments are for
-% array variables; \texttt{arrayvar} is for the array as a whole,
-% whereas \texttt{arrayentry} is for individual entries in the array
-% that (for some reason) are important enough to warrant special
-% attention. The respective syntaxes are
-% \begin{quote}
-% |\begin{arrayvar}|\oarg{namespace}\marg{array}\oarg{index-des}\\
-% |\begin{arrayentry}|\oarg{namespace}\marg{array}\marg{entry}
-% \end{quote}
-% Here \meta{array} is the name of an array variable, \meta{entry} is
-% the name of an entry, and \meta{index-des} is a short piece of text
-% (seldom more than one word) which describes what is used as indices
-% into that array.
-% \begin{macrocode}
-\NewMacroEnvironment*{arrayvar}{%
- \XD@grab@harmless@oarg\XD@grab@harmless\relax\XD@grab@harmless@oarg
-}{3}%
- {\XDParToMargin{\MetaNormalfont\MacroFont \buildname{#1}{#2}%
- \ifx \NoValue#3%
- \space\normalfont (array)%
- \else
- \penalty\hyphenpenalty(\mbox{\meta@font@select#3})%
- \fi
- }}%
- {\TD@main@index{#1}{#2}{array}{%
- \ifx \NoValue#3\else
- \LevelSorted{#3}{\protect\mbox{\textit{#3}} entries}%
- \fi
- }}%
- {{\buildname{#1}{#2}}{\texttt{\buildname{#1}{#2}} array}}
- {}%
-% \end{macrocode}
-% \begin{macrocode}
-\NewMacroEnvironment*{arrayentry}{%
- \XD@grab@harmless@oarg\XD@grab@harmless\relax\XD@grab@harmless\relax
-}{3}%
- {\XDParToMargin{%
- \MetaNormalfont\MacroFont \buildname{#1}{#2}%
- \penalty\hyphenpenalty(#3)%
- }}%
- {\TD@main@index{#1}{#2}{array}{\LevelSorted{#3}{\texttt{#3}}}}%
- {{\buildname{#1}{#2}(#3)}{\texttt{\buildname{#1}{#2}(#3)}}}
- {}%
-% \end{macrocode}
-% \end{environment}\end{environment}
-%
-% \begin{macro}{\describestring}
-% \changes{v\,2.01}{2000/08/12}{Command added. (LH)}
-% \changes{v\,2.12}{2000/11/21}{Optional argument added. (LH)}
-% \changes{v\,2.13}{2000/12/16}{Second optional argument added. (LH)}
-% The |\describestring| command has the syntax
-% \begin{quote}
-% |\describestring|\oarg{type}\oarg{namespace}\marg{string}
-% \end{quote}
-% It is similar to \package{xdoc}'s own |\describe|\-|cs|\-|family|
-% command, but it has a couple of extra features which are connected
-% to the optional arguments and it doesn't prepend a backslash to
-% the name of the thing being described. When the \meta{type} argument
-% is present, it is taken as a ``type declaration'' of the
-% \meta{string} being described and `(\meta{type})' will be put as a
-% suffix to the \meta{string}. When the \oarg{namespace} argument is
-% present it means that the string being described is really the name
-% of something (e.g.\ a procedure or variable) which exists in a
-% namespace, and the complete name is formed by passing both
-% \meta{namespace} and \meta{string} as arguments to |\buildname|.
-% Note that an empty \meta{namespace} argument is not quite the
-% same thing as no \meta{namespace} argument.
-% \begin{macrocode}
-\NewDescribeCommand{\describestring}{%
- \XD@grab@oarg\XD@grab@harmless@oarg\XD@grab@harmless{}%
-}{3}{%
- \GenericDescribePrint{%
- \MetaNormalfont\MacroFont
- \ifx \NoValue#2\@empty
- #3%
- \else
- \buildname{#2}{#3}%
- \fi
- \ifx \NoValue#1\@empty \else\ \normalfont(#1)\fi
- }%
- \begingroup
- \def\meta##1{(##1)}%
- \unrestored@protected@xdef\@gtempa{%
- #3\ifx \NoValue#2\@empty \else\ifx $#2$\else, #2\fi\fi
- }%
- \endgroup
- \IndexEntry{%
- \LevelSorted{\@gtempa}{%
- \texttt{%
- \ifx \NoValue#2\@empty
- \protect\MetaNormalfont#3%
- \else
- \buildname{#2}{#3}%
- \fi
- }%
- \ifx \NoValue#1\@empty \else\space(#1)\fi
- }%
- }{usage}{\thepage}%
-}
-% \end{macrocode}
-% \end{macro}
-%
-% \begin{macro}{\TD@main@index}
-% \changes{v\,2.10}{2000/10/14}{Generalized macro. (LH)}
-% \changes{v\,2.12}{2000/11/25}{Sorting by base name first, not by
-% namespace first. (LH)}
-% The |\TD@main@index| macro has the syntax
-% \begin{quote}
-% |\TD@main@index|\marg{namespace}\marg{name}\marg{type}^^A
-% \marg{extra}
-% \end{quote}
-% It makes a \texttt{main} index entry. The top level of this entry
-% is like an entry for the \meta{name} object of type \meta{type} in
-% the \meta{namespace} namespace. The entry may have one or more
-% sublevels\footnote{But no more than one extra level is recommended,
-% as this macro is intended to be redefinable to make a two-level
-% entry by default and \package{makeindex} can handle at most three
-% levels.} specified in the \meta{extra}; these are then made using
-% explicit |\LevelSame| or |\LevelSorted| commands. \meta{name} and
-% \meta{namespace} are as for the |\buildname| macro, whereas
-% \meta{type} is a description, e.g. `\texttt{proc}', that gets
-% appended to the printed text of the entry's top level.
-% \begin{macrocode}
-\def\TD@main@index#1#2#3#4{%
- \XDMainIndex{%
- \LevelSorted{%
- #2%
- \ifx \NoValue#1%
- \ifx \@empty\TD@namespace \else , \TD@namespace\fi
- \else
- \ifx $#1$\else , #1\fi
- \fi
- }{\texttt{\buildname{#1}{#2}} (#3)}%
- #4%
- }%
-}
-% \end{macrocode}
-% \end{macro}
-%
-%
-% \subsection{Regular expressions and friends}
-%
-% I have noticed that when one is describing the syntax of some \Tcl\
-% command\slash procedure, one often ends up writing argument lists with
-% lots of optional, repeated, or variant parts. In general one can of
-% course specify these using BNF syntax diagrams, but whereas occurrences
-% of the three syntactic constructions mentioned above are plenty,
-% there are much fewer occurrences of for example nesting. Since
-% furthermore even the simplest optional or repeated element
-% requires an extra syntactic term to be defined, it becomes interesting
-% to search for an alternative way of specifying these constructions.
-% One that is close at hand is that of regular expressions, as that is
-% anyway a part of the \Tcl\ language.
-%
-% \begin{macro}{\regstar}
-% \begin{macro}{\regplus}
-% \begin{macro}{\regopt}
-% These commands are used to modify the meaning of another syntactic
-% element. |\regstar| puts a star $^*$ on something, meaning it is
-% repeated zero or more times. |\regplus| puts a plus $^+$ on
-% something, meaning it is repeated one or more times. |\regopt| puts
-% a question mark $^?$ on something, meaning it is optional. Normally
-% these commands are simply put right after what they should modify,
-% but if the ``something'' is a regular expression block, i.e., a
-% \texttt{regexp} environment, then these things should be put as the
-% \meta{modifier} in
-% \begin{quote}
-% |\begin{regexp}|\oarg{modifier}\\
-% | |\textellipsis\\
-% |\end{regexp}|
-% \end{quote}
-% In particular, they should \emph{not} be put after the
-% |\end{regexp}|, as that may impair the placement of the exponent.
-% \begin{macrocode}
-\newcommand\regstar{\ensuremath{^*}}
-\newcommand\regplus{\ensuremath{^+}}
-\newcommand\regopt{\ensuremath{^?}}
-% \end{macrocode}
-% \end{macro}\end{macro}\end{macro}
-%
-% \begin{macro}{\TD@delim@size}
-% \begin{macro}{\TD@delim@size@G}
-% These control sequences are used in determining what size the
-% delimiters (parentheses) around a regular expression block should
-% have. The values specify what should be the least size of a
-% delimiter that bracket a piece of text, and are interpreted as
-% follows: \(0={}\)normal size, \(1={}\)|\big| size, \(2={}\)|\Big|
-% size, \(3={}\)|\bigg| size, and $4$ and above is |\Bigg| size.
-%
-% |\TD@delim@size@G| is a macro which is always set globally and
-% which keeps track of the delimiter size at the innermost level.
-% |\TD@delim@size| is a |\count| register that is assigned locally
-% and is used for keeping track of delimiter sizes at all other
-% levels. Commands like |\word|, which contribute a delimiter to some
-% piece of text, should set |\TD@delim@size@G|.
-% \begin{macrocode}
-\newcount\TD@delim@size
-\gdef\TD@delim@size@G{0}
-% \end{macrocode}
-% \end{macro}\end{macro}
-%
-% \begin{environment}{regexp}
-% The \texttt{regexp} environment embarks upon a rather tricky scheme
-% of boxing and unboxing to determine the proper delimiter size. The
-% basic idea is that the contents of the environment are first typeset
-% in an \texttt{lrbox} environment, and |\TD@delim@size@G| is set to
-% its proper value. Then the left delimiter is set at this size, the
-% saved box is unboxed, and finally the right delimiter is set. All
-% through the process, the value of |\TD@delim@size@G| at the
-% beginning of the \texttt{regexp} environment has been saved in
-% |\TD@delim@size| and at the end |\TD@delim@size@G| is set to the
-% maximum of its old value and one plus its value inside the
-% \texttt{regexp} environment. This makes it possible to nest
-% \texttt{regexp} environments.
-%
-% It is not quite that easy, though. A regular expression can consist
-% of several branches, which must be separated by vertical bars.
-% These bars should be the same size as the delimiters around the
-% regular expression, and hence the bars cannot be typeset before
-% every branch has been typeset. Therefore the \texttt{regexp}
-% environment starts an extra horizontal list (the |\hbox|) for the
-% sole purpose of storing the branches that have already been typeset
-% (and \emph{only} these). Each branch is put on the list as a box.
-% When all branches have been typeset, they are first removed from the
-% list (by |\TD@join@branches|), and then they are inserted again, but
-% this time they are unboxed and separated by suitable delimiters.
-% Finally the |\hbox| is ended and almost immediately unboxed.
-%
-% \begin{macrocode}
-\newenvironment{regexp}[1][]{%
- \leavevmode
- \def\TD@modifier{#1}%
- \TD@delim@size=\TD@delim@size@G
- \let\regalt=\TD@regalt
- \setbox\z@=\hbox\bgroup
- \TD@delim@size=\z@
- \begin{lrbox}{\z@}%
- \gdef\TD@delim@size@G{0}%
- \ignorespaces
-}{%
- \end{lrbox}%
- \ifnum \TD@delim@size@G>\TD@delim@size
- \TD@delim@size=\TD@delim@size@G\relax
- \fi
- {\TD@join@branches}%
- \unhbox\z@
- $\TD@size@delimiter)\TD@modifier$%
- \advance \TD@delim@size \@ne
- \xdef\TD@delim@size@G{\the\TD@delim@size}%
- \egroup
- \ifnum \TD@delim@size@G<\TD@delim@size
- \xdef\TD@delim@size@G{\the\TD@delim@size}%
- \fi
- \unhbox\z@
-}
-% \end{macrocode}
-% \end{environment}
-%
-% \begin{macro}{\TD@join@branches}
-% \changes{v\,2.12}{2000/11/13}{Made penalty for linebreak after
-% \cs{regalt} depend on the delimiter size. (LH)}
-% The |\TD@join@branches| macro uses |\lastbox| to retrieve all boxes
-% in the current horizontal list, and unboxes them. Between the
-% unboxed boxes it puts a |\vert| delimiter, and before the first it
-% puts a left parenthesis. Both will be the current delimiter size.
-% \begin{macrocode}
-\def\TD@join@branches{%
- \setbox\z@=\lastbox
- \ifvoid\z@
- $\TD@size@delimiter($%
- \else
- {\TD@join@branches}%
- \unhbox\z@
- ~$\TD@size@delimiter\vert$%
- \count@=\TD@delim@size
- \advance \count@ \@ne
- \multiply \count@ -\@lowpenalty
- \penalty\count@\ %
- \fi
-}
-% \end{macrocode}
-% \end{macro}
-%
-% \begin{macro}{\regalt}
-% \begin{macro}{\TD@regalt}
-% The |\regalt| command is used in a \texttt{regexp} environment
-% to separate two branches. It ends an \texttt{lrbox} environment,
-% puts the box collected in the surrounding horizontal list (which
-% was begun by the |\hbox| in |\lrbox|), and begins a new
-% \texttt{lrbox} environment. As this would not make sense outside a
-% \texttt{regexp} environment however, the |\regalt| command does by
-% default only produce an error message. The ``real'' definition is
-% kept in |\TD@regalt| and the \texttt{regexp} environment locally
-% redefines |\regalt|.
-% \begin{macrocode}
-\newcommand\regalt{%
- \PackageError{tcldoc}{%
- Lonely \protect\regalt--perhaps a missing regexp environment%
- }\@eha
-}
-\def\TD@regalt{%
- \end{lrbox}%
- \box\z@
- \ifnum \TD@delim@size@G>\TD@delim@size
- \TD@delim@size=\TD@delim@size@G\relax
- \fi
- \begin{lrbox}{\z@}%
- \gdef\TD@delim@size@G{0}%
-}
-% \end{macrocode}
-% \end{macro}\end{macro}
-%
-% \begin{macro}{\TD@size@delimiter}
-% This macro expands to nothing, |\big|, |\Big|, |\bigg|, or |\Bigg|
-% depending on the value of |\TD@delim@size|.
-% \begin{macrocode}
-\def\TD@size@delimiter{%
- \ifcase\TD@delim@size
- \or \expandafter\big \or \expandafter\Big \or
- \expandafter\bigg \else \expandafter\Bigg
- \fi
-}
-% \end{macrocode}
-% \end{macro}
-%
-% \begin{macro}{\word}
-% \changes{v\,2.10}{2000/10/13}{Reimplemented like the new \cs{meta}.
-% (LH)}
-% \changes{v\,2.12}{2000/11/21}{Using braces instead of angle brackets.
-% (LH)}
-% This command works like |\meta|, but the argument is enclosed in
-% braces rather than angle brackets and it communicates with the
-% \texttt{regexp} environment (the braces count as a smallest-size
-% delimiter). Something marked out using |\word| rather than |\meta|
-% is meant to always denote a separate \Tcl\ word.
-% \begin{macrocode}
-\DeclareRobustCommand\word[1]{%
- \leavevmode
- \ifmmode \expandafter \nfss@text \fi
- {%
- $\{$%
- \meta@font@select
- \edef\meta@hyphen@restore
- {\hyphenchar\the\font\the\hyphenchar\font}%
- \hyphenchar\font\m@ne
- \language\l@nohyphenation
- #1\/%
- \meta@hyphen@restore
- $\}$%
- }%
- \ifnum \TD@delim@size@G<\@ne \gdef\TD@delim@size@G{1}\fi
-}
-% \end{macrocode}
-% \end{macro}
-%
-%
-% \subsection{Logos}
-%
-% \begin{macro}{\Tcllogo}
-% \changes{v\,2.03}{2000/09/15}{Macro added. (LH)}
-% The |\Tcllogo| command makes a \Tcl\ ``logo'', which simply consists
-% of the three letters `T', `c', and `l' in the current font. The only
-% special thing about it is that there is a kern between the `T' and
-% the `c', since most fonts doesn't have that kern. No work has been
-% put into tuning the kern, but the size is about right. As a
-% comparison the kern between `T' and `o' in \texttt{cmr10} is
-% $-0.083334\,\mathrm{em}$.
-% \begin{macrocode}
-\DeclareRobustCommand\Tcllogo{T\kern-0.1em cl}
-%</pkg>
-% \end{macrocode}
-% \end{macro}
-%
-% \changes{v\,2.03}{2000/09/15}{\cs{DescribeResource} command removed.
-% Just in case anyone is interested in using it, I have included it as
-% a code example in the source for the \package{xdoc} package. (LH)}
-%
-%
-% \section{The \package{tcldoc} class}
-%
-% \subsection{Pre-options definitions}
-%
-% \begin{macro}{\if@rtkinenc@}
-% Switch. True iff the \package{rtkinenc} package should be loaded.
-% It starts out false.
-% \begin{macrocode}
-%<*class>
-\newif\if@rtkinenc@
-% \end{macrocode}
-% \end{macro}
-%
-%
-% \subsection{Option declarations and processing}
-%
-% \begin{option}{a5paper}
-% \begin{option}{olddocinclude}
-% The \texttt{a5paper} (to the \package{article} document class) and
-% the \texttt{olddocinclude} (to the \package{xdoc} package) are not
-% supported for the \package{tcldoc} class.
-% \begin{macrocode}
-\DeclareOption{a5paper}{%
- \ClassError{tcldoc}{The a5paper option is not supported}\@eha
-}
-\DeclareOption{olddocinclude}{%
- \ClassError{tcldoc}{The olddocinclude option is not supported}\@eha
-}
-% \end{macrocode}
-% \end{option}\end{option}
-%
-% \begin{option}{macinputenc}
-% The \texttt{macinputenc} option enables some special code for
-% supporting the typesetting of \Tcl\ code written for the Macintosh
-% input encoding (cf.\ the \texttt{applemac} option to the
-% \package{inputenc} package).
-% \begin{macrocode}
-\DeclareOption{macinputenc}{%
- \PassOptionsToPackage{applemac}{rtkinenc}%
- \@rtkinenc@true
-}
-% \end{macrocode}
-% \end{option}
-%
-%
-% \begin{macrocode}
-\DeclareOption*{%
- \PassOptionsToClass{\CurrentOption}{article}}
-% \end{macrocode}
-%
-%
-%
-% Input a local configuration file, if it exists.
-% \begin{macrocode}
-\InputIfFileExists{tcldoc.cfg}
- {\typeout{*************************************^^J%
- * Local config file tcldoc.cfg used^^J%
- *************************************}}
- {}
-% \end{macrocode}
-%
-%
-%
-% \begin{macrocode}
-\ProcessOptions\relax
-% \end{macrocode}
-%
-% \subsection{Loading the base class}
-%
-% \begin{macrocode}
-\LoadClass{article}
-% \end{macrocode}
-%
-%
-% \subsection{Input encoding and font representations}
-%
-% There are some rather tricky issues concerning interpretation of the
-% input in \Tcl\ code. If the |\MacroFont| font contains a character that
-% correctly represents a character in the input, then that character is
-% obviously what one would like to appear in the typeset output. If no
-% font in the entire \TeX\ system contains a character that looks like
-% the one in the input, then just as obviously must one examine whether
-% there exists some other method of specifying that character. In the
-% case of \Tcl\ code, a suitable method seems to be to typeset the
-% equivalent |\x|\meta{dd} escape sequence instead. It is for the
-% purpose of being able to detect these cases that the \package{tcldoc}
-% class (if the user so requests) loads the \package{rtkinenc} package,
-% rather than the standard \LaTeX\ \package{inputenc} package.
-%
-% The problem of determining for each input character which method of
-% representation that should be used---single character or escape
-% sequence---does not have a unique solution. Therefore the default
-% set-up is to not define any special fakes at all, which means the
-% escape sequence method will be chosen in the doubtful cases. There are
-% also a couple of options for the \package{tcldoc} class that define
-% all (or nearly all) possible fakes---the difference between them lie
-% in which fonts are assumed to be present in the system. A user who
-% does not like any of these predefined options also has the option of
-% putting the corresponding |\Declare|\-|Text|\-|Default|\textellipsis
-% \ commands directly into the \texttt{tcldoc.cfg} file.
-%
-% \begin{macrocode}
-\if@rtkinenc@
- \RequirePackage{rtkinenc}
-%</class>
-% \end{macrocode}
-%
-% \begin{macro}{\InputModeCode}
-% \begin{macro}{\SetUnavailableAction}
-% \begin{macro}{\DeclareInputMath}
-% \begin{macro}{\RIE@undefined}
-% The \package{tcldoc} package uses some commands defined by the
-% \package{rtkinenc} package, but since \package{rtkinenc} is not
-% required by the \package{tcldoc} package, the \package{tcldoc}
-% package must provide its own definitions, even though these can be
-% no-ops. Definition of these commands is delayed until
-% |\begin|\nolinebreak[1]|{document}| to avoid problems if the user
-% loads the \package{rtkinenc} package after the \package{tcldoc}
-% package.
-% \begin{macrocode}
-%<*pkg>
-\AtBeginDocument{%
- \providecommand*\InputModeCode{}%
- \providecommand*\SetUnavailableAction[1]{}%
- \providecommand*\DeclareInputMath[2]{}%
- \providecommand*\RIE@undefined[1]{\@inpenc@undefined}%
-}
-%</pkg>
-% \end{macrocode}
-% \end{macro}\end{macro}\end{macro}\end{macro}
-%
-% \begin{macro}{\.}
-% \begin{macro}{\H}
-% \begin{macro}{\textemdash}
-% \begin{macro}{\textendash}
-% \begin{macro}{\textexclamdown}
-% \begin{macro}{\textquestiondown}
-% \begin{macro}{\textquotedblleft}
-% The following text commands are defined by \texttt{OT1enc.def}, but
-% their definitions will not work for \texttt{OT1}-encoded typewriter
-% fonts. Therefore they are redefined, either so that they work
-% anyway (|\textexclamdown| and |\textquestiondown|) or so that they
-% call |\TextSymbolUnavailable| when \package{rtkinenc} is in code
-% mode. This should make things work for the best in most
-% reasonable situations.
-%
-% It could be noted here that the definitions below for |\.| and |\H|
-% does not quite behave as if the commands had not been defined at
-% all (I'm assuming that \package{rtkinenc} is in code mode here); an
-% undefined command would leave its argument to be typeset, but the
-% definitions below will swallow it in those cases.
-% \begin{macrocode}
-%<*class>
- \DeclareTextCommand{\.}{OT1}[1]{%
- \IfInputModeCode{\TextSymbolUnavailable\.}%
- {\add@accent{95}{#1}}%
- }
- \DeclareTextCommand{\H}{OT1}[1]{%
- \IfInputModeCode{\TextSymbolUnavailable\H}%
- {\add@accent{125}{#1}}%
- }
- \DeclareTextCommand{\textemdash}{OT1}{%
- \IfInputModeCode{\TextSymbolUnavailable\textemdash}{\char 124 }%
- }
- \DeclareTextCommand{\textendash}{OT1}{%
- \IfInputModeCode{\TextSymbolUnavailable\textendash}{\char 123 }%
- }
- \DeclareTextCommand{\textexclamdown}{OT1}{!`}
- \DeclareTextCommand{\textquestiondown}{OT1}{?`}
- \DeclareTextCommand{\textquotedblleft}{OT1}{%
- \IfInputModeCode{\TextSymbolUnavailable\textquotedblleft}%
- {\char 92 }%
- }
-% \end{macrocode}
-% \end{macro}\end{macro}\end{macro}\end{macro}\end{macro}\end{macro}
-% \end{macro}
-%
-% Several of the definitions in \texttt{applemac.def} can be improved.
-% Most of it is about seeing to that all characters have a proper text
-% mode definition, so that their unavailability can be trapped.
-%
-% \begin{macrocode}
- \def\@tempa{applemac}
- \ifx \@tempa\RIE@encoding
- \DeclareInputBoth{173}{\textneq}{\neq}
- \DeclareInputBoth{176}{\textinfty}{\infty}
- \DeclareInputBoth{177}{\textpm}{\pm}
- \DeclareInputBoth{178}{\textleq}{\leq}
- \DeclareInputBoth{179}{\textgeq}{\geq}
- \DeclareInputBoth{181}{\textmu}{\mu}
- \DeclareInputBoth{182}{\textpartial}{\partial}
- \DeclareInputBoth{183}{\textSigma}{\Sigma}
- \DeclareInputBoth{184}{\textPi}{\Pi}
- \DeclareInputBoth{185}{\textpi}{\pi}
- \DeclareInputBoth{186}{\textint}{\int}
- \DeclareInputBoth{189}{\textohm}{\Omega}
- \DeclareInputBoth{194}{\textlnot}{\lnot}
- \DeclareInputBoth{195}{\textsurd}{\surd}
- \DeclareInputBoth{197}{\textapprox}{\approx}
- \DeclareInputBoth{198}{\textDelta}{\Delta}
- \DeclareInputBoth{214}{\textdiv}{\div}
- \DeclareInputBoth{215}{\textdiamond}{\diamond}
- \DeclareInputBoth{218}{\textfractionsolidus}{/}
-% \end{macrocode}
-%
-% Several of the above text commands are not defined in the \LaTeX\
-% kernel. Therefore one must either define them in some encoding, or
-% define a default for them, to make them proper text commands. If they
-% are not then their unavailability cannot be trapped. |\textapplelogo|
-% appears in \texttt{applemac.def}.
-% \begin{macrocode}
- \ProvideTextCommandDefault{\textneq}{%
- \TextSymbolUnavailable\textneq
- }
- \ProvideTextCommandDefault{\textapplelogo}{%
- \TextSymbolUnavailable\textapplelogo
- }
- \DeclareTextSymbol{\textinfty}{OMS}{49}
- \DeclareTextSymbol{\textpm}{OMS}{6}
- \DeclareTextSymbol{\textleq}{OMS}{20}
- \DeclareTextSymbol{\textgeq}{OMS}{21}
-% \DeclareTextSymbol{\textmu}{TS1}{181}
- \DeclareTextSymbol{\textpartial}{OML}{64}
- \DeclareTextSymbol{\textSigma}{OT1}{6}
- \DeclareTextSymbol{\textPi}{OT1}{5}
- \DeclareTextSymbol{\textpi}{OML}{25}
- \DeclareTextSymbol{\textint}{OMS}{115}
- \DeclareTextSymbol{\textohm}{OT1}{10} % Kind of \textOmega
-% \DeclareTextSymbol{\textlnot}{TS1}{172}
-% \DeclareTextSymbol{\textsurd}{TS1}{187}
- \DeclareTextSymbol{\textapprox}{OMS}{25}
- \DeclareTextSymbol{\textDelta}{OT1}{1}
-% \DeclareTextSymbol{\textdiv}{TS1}{246}
- \DeclareTextSymbol{\textdiamond}{OMS}{5}
-% \DeclareTextSymbol{\textfractionsolidus}{TS1}{47}
- \input{TS1enc.def}
- \fi
-\fi
-% \end{macrocode}
-% \begin{macrocode}
-\let\if@rtkinenc@=\@undefined
-\let\@rtkinenc@true=\@undefined
-\let\@rtkinenc@false=\@undefined
-% \end{macrocode}
-% The |@rtkinenc@| switch isn't needed any more, so it might just as
-% well be undefined.
-%
-%
-% \subsection{The layout}
-%
-% Increase the text width slightly so that with the standard fonts
-% 72 columns of code may appear in a |macrocode| environment.
-% \changes{ \package{ltxdoc} v\,2.0c}{1994/03/15}{Set \cs{textwidth}.}
-% \begin{macrocode}
-\setlength{\textwidth}{355pt}
-% \end{macrocode}
-%
-% Increase the marginpar width slightly, for long command names.
-% And increase the left margin by a similar amount
-% \changes{ \package{ltxdoc} v\,2.0l}
-% {1994/05/25}{Increase \cs{marginparwidth}}
-% \changes{ \package{ltxdoc} v\,2.0q}{1995/11/28}
-% {Increase \cs{marginparwidth} and page margin.}
-% \begin{macrocode}
-\addtolength\oddsidemargin{20pt}
-\addtolength\evensidemargin{20pt}
-% \end{macrocode}
-%
-%
-% Change some defaults for list formatting. In particular, continue to
-% indent paragraphs in |\list| environments and don't put extra space
-% between them.
-% \changes{ \package{pasdoc} v\,1.30}{1999/08/31}{Modifying the list
-% formatting. (LH)}
-%
-% \begin{macro}{\TD@list}
-% \begin{macro}{\@listi}
-% \begin{macro}{\@listI}
-% \begin{macro}{\@listii}
-% \begin{macro}{\@listiii}
-% \begin{macro}{\@listiv}
-% \begin{macro}{\@listv}
-% \begin{macro}{\@listvi}
-% \begin{macrocode}
-\def\@tempa#1{%
- \expandafter\def \expandafter#1\expandafter{#1\TD@list}%
-}
-\def\TD@list{%
- \advance \itemsep \parsep
- \parsep=\z@\@plus\p@\relax
- \advance \itemsep -\parsep
- \listparindent=1em\relax
-}
-\@tempa\@listi
-\let\@listI\@listi
-\@tempa\@listii
-\@tempa\@listiii
-\@tempa\@listiv
-\@tempa\@listv
-\@tempa\@listvi
-% \end{macrocode}
-% \end{macro}\end{macro}\end{macro}\end{macro}\end{macro}\end{macro}
-% \end{macro}\end{macro}
-%
-%
-% \subsection{Loading the \package{tcldoc} package}
-%
-% \begin{macrocode}
-\RequirePackage[dolayout]{tcldoc}
-% \end{macrocode}
-%
-% \begingroup
-% Make \verb+|+ be a `short verb' character, but not in the document
-% preamble, where an active character may interfere with packages that
-% are loaded.
-% \DoNotIndex{\|}
-% \begin{macrocode}
-\AtBeginDocument{\MakeShortTclverb{\|}}
-% \end{macrocode}
-% \endgroup
-%
-% As \texttt{.dtx} documents tend to have a lot of monospaced material,
-% set up some \texttt{cmtt} substitutions to occur silently.
-% \changes{ \package{ltxdoc} v\,2.0p}{1995/11/02}{Add font substitutions}
-% \begin{macrocode}
-\DeclareFontShape{OT1}{cmtt}{bx}{n}{<-> ssub * cmtt/m/n}{}
-\DeclareFontFamily{OMS}{cmtt}{\skewchar\font'60}
-\DeclareFontShape{OMS}{cmtt}{m}{n}{<-> ssub * cmsy/m/n}{}
-\DeclareFontShape{OMS}{cmtt}{bx}{n}{<-> ssub * cmsy/b/n}{}
-% \end{macrocode}
-% This substitution is in the standard fd file, but not silent.
-% \begin{macrocode}
-\DeclareFontShape{OT1}{cmss}{m}{it}{<->ssub*cmss/m/sl}{}
-% \end{macrocode}
-%
-%
-% \subsection{More layout}
-%
-% \changes{v\,2.12}{2000/11/26}{Setting \cs{marginparwidth} so that the
-% outer edge of the marginpar is $1\,\mathrm{cm}$ from the paper
-% edge. (LH)}
-% \changes{v\,2.12}{2000/11/27}{Rearranged code so that
-% the \cs{marginparwidth} is set after \package{tcldoc} has been
-% loaded; otherwise \package{doc} might overwrite it! (LH)}
-% \begin{macrocode}
-\setlength\marginparwidth{\evensidemargin}
-\addtolength\marginparwidth{1in}
-\addtolength\marginparwidth{-\marginparsep}
-\addtolength\marginparwidth{-1cm}
-% \end{macrocode}
-%
-%
-% \begin{macrocode}
-\CodelineNumbered
-\DisableCrossrefs
-% \end{macrocode}
-%
-% \begin{macrocode}
-\setcounter{StandardModuleDepth}{1}
-% \end{macrocode}
-%
-%
-% \subsection{Useful abbreviations}
-% \changes{v\,2.10}{2000/10/11}{\LaTeX\ abbreviation commands from
-% \package{ltxdoc} commented out. (LH)}
-%
-% The commands currently found in this section are inherited from the
-% \package{ltxdoc} package, but as they aren't of much use with \Tcl,
-% they have been commented out.
-%
-% |\cmd{\foo}| Prints |\foo| verbatim. It may be used inside moving
-% arguments. |\cs{foo}| also prints |\foo|, for those who prefer that
-% syntax. (This second form may even be used when |\foo| is |\outer|).
-% \begin{macro}{\cmd}
-% \begin{macro}{\cs}
-% \DoNotIndex{\\}
-% \begin{macrocode}
-% \def\cmd#1{\cs{\expandafter\cmd@to@cs\string#1}}
-% \def\cmd@to@cs#1#2{\char\number`#2\relax}
-% \DeclareRobustCommand\cs[1]{\texttt{\char`\\#1}}
-% \end{macrocode}
-% \end{macro}
-% \end{macro}
-%
-% \begin{macro}{\marg}
-% |\marg{text}| prints \marg{text}, `mandatory argument'.
-% \DoNotIndexHarmless{\PrintChar{`\{}}
-% \DoNotIndexHarmless{\PrintChar{`\}}}
-% \begin{macrocode}
-% \providecommand\marg[1]{%
-% {\ttfamily\char`\{}\meta{#1}{\ttfamily\char`\}}}
-% \end{macrocode}
-% \end{macro}
-%
-% \begin{macro}{\oarg}
-% |\oarg{text}| prints \oarg{text}, `optional argument'.
-% \begin{macrocode}
-% \providecommand\oarg[1]{%
-% {\ttfamily[}\meta{#1}{\ttfamily]}}
-% \end{macrocode}
-% \end{macro}
-%
-% \begin{macro}{\parg}
-% |\parg{te,xt}| prints \parg{te,xt}, `picture mode argument'.
-% \begin{macrocode}
-% \providecommand\parg[1]{%
-% {\ttfamily(}\meta{#1}{\ttfamily)}}
-%</class>
-% \end{macrocode}
-% \end{macro}
-%
-%
-%
-%
-% \changes{v\,2.00}{2000/07/17}{Section with \cs{DocInclude} removed,
-% since that is defined by \package{xdoc}. (LH)}
-%
-% \Finale
-%
-\endinput
-
diff --git a/Master/texmf-dist/source/latex/tcldoc/tcldoc.ins b/Master/texmf-dist/source/latex/tcldoc/tcldoc.ins
deleted file mode 100644
index f4ae5b3c944..00000000000
--- a/Master/texmf-dist/source/latex/tcldoc/tcldoc.ins
+++ /dev/null
@@ -1,45 +0,0 @@
-% This file is the docstrip installation script for the tcldoc package.
-%
-% It may be distributed under the conditions of the LaTeX
-% Project Public License, either version 1.2 of this license or
-% (at your option) any later version. The latest version of this
-% license is in
-% http://www.latex-project.org/lppl.txt
-% and version 1.2 or later is part of all distributions of LaTeX
-% version 1999/12/01 or later.
-%
-% Copyright 2000 Lars Hellstr\"om
-
-\input docstrip.tex
-
-\preamble
-
-This file may be distributed and/or modified under the conditions
-of the LaTeX Project Public License, either version 1.2 of this
-license or (at your option) any later version. The latest version
-of this license is in
- http://www.latex-project.org/lppl.txt
-and version 1.2 or later is part of all distributions of LaTeX
-version 1999/12/01 or later.
-
-This file may only be distributed together with a copy of the source
-file from which it was generated. You may distribute that source file
-without this generated file.
-
-\endpreamble
-
-\usedir{tex/latex/misc}
-
-\keepsilent
-
-\generate{%
- \file{tcldoc.cls}{%
- \from{tcldoc.dtx}{class}
- }%
- \file{tcldoc.sty}{%
- \from{tcldoc.dtx}{pkg}
- }%
-}
-
-\endbatchfile
-\endinput
diff --git a/Master/tlpkg/bin/tlpkg-ctan-check b/Master/tlpkg/bin/tlpkg-ctan-check
index e879ea9c44a..865bb2accd1 100755
--- a/Master/tlpkg/bin/tlpkg-ctan-check
+++ b/Master/tlpkg/bin/tlpkg-ctan-check
@@ -337,7 +337,7 @@ my @TLP_working = qw(
tableaux tablenotes tablists tablor tabls tabto-generic tabto-ltx
tabu tabularborder tabularcalc tabularew
tabulars-e tabulary tabvar talk
- tamethebeast tapir tdclock tdsfrmath
+ tamethebeast tapir tcldoc tdclock tdsfrmath
technics ted templates-fenn templates-sommer tengwarscript
tensor termcal termlist teubner
tex-ewd tex-font-errors-cheatsheet tex-gyre tex-label tex-overview