diff options
author | Norbert Preining <preining@logic.at> | 2009-12-23 18:57:23 +0000 |
---|---|---|
committer | Norbert Preining <preining@logic.at> | 2009-12-23 18:57:23 +0000 |
commit | 954a951fc174a8b1b5e1ce335a3d0fa8ae77a25b (patch) | |
tree | 8d87babe00944a8357cb3309bb2e00fd5e00e0ba /Master/texmf-dist/source/latex | |
parent | 8f0d30f2d461ba69f47843ce84fd1a0732cc7066 (diff) |
add revtex4 package that ships the old revtex version
taken from TL 2007 (the only place where I could find it for now)
git-svn-id: svn://tug.org/texlive/trunk@16488 c570f23f-e606-0410-a88d-b1316a301751
Diffstat (limited to 'Master/texmf-dist/source/latex')
-rw-r--r-- | Master/texmf-dist/source/latex/revtex4/ltxdocext.dtx | 1208 | ||||
-rw-r--r-- | Master/texmf-dist/source/latex/revtex4/ltxgrid.dtx | 4815 | ||||
-rw-r--r-- | Master/texmf-dist/source/latex/revtex4/ltxutil.dtx | 3534 | ||||
-rw-r--r-- | Master/texmf-dist/source/latex/revtex4/revtex4.dtx | 8728 | ||||
-rw-r--r-- | Master/texmf-dist/source/latex/revtex4/revtex4.ins | 36 | ||||
-rw-r--r-- | Master/texmf-dist/source/latex/revtex4/textcase.dtx | 404 |
6 files changed, 18725 insertions, 0 deletions
diff --git a/Master/texmf-dist/source/latex/revtex4/ltxdocext.dtx b/Master/texmf-dist/source/latex/revtex4/ltxdocext.dtx new file mode 100644 index 00000000000..ee9d71c9c04 --- /dev/null +++ b/Master/texmf-dist/source/latex/revtex4/ltxdocext.dtx @@ -0,0 +1,1208 @@ +% \iffalse &pdflatex ltxdoc klootch +% ltxdocext.dtx: package to extend the ltxdoc class of standard LaTeX +% Copyright (c) 1999 Arthur Ogawa +% +% Disclaimer +% This file is distributed WITHOUT ANY WARRANTY; +% without even the implied warranty of +% MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. +% ReadMe +% For the documentation and more detailed instructions for +% installation, typeset this document with \LaTeX. +% \fi +% \CheckSum{938} +%% \CharacterTable +%% {Upper-case \A\B\C\D\E\F\G\H\I\J\K\L\M\N\O\P\Q\R\S\T\U\V\W\X\Y\Z +%% Lower-case \a\b\c\d\e\f\g\h\i\j\k\l\m\n\o\p\q\r\s\t\u\v\w\x\y\z +%% Digits \0\1\2\3\4\5\6\7\8\9 +%% Exclamation \! Double quote \" Hash (number) \# +%% Dollar \$ Percent \% Ampersand \& +%% Acute accent \' Left paren \( Right paren \) +%% Asterisk \* Plus \+ Comma \, +%% Minus \- Point \. Solidus \/ +%% Colon \: Semicolon \; Less than \< +%% Equals \= Greater than \> Question mark \? +%% Commercial at \@ Left bracket \[ Backslash \\ +%% Right bracket \] Circumflex \^ Underscore \_ +%% Grave accent \` Left brace \{ Vertical bar \| +%% Right brace \} Tilde \~}% +% +% \iffalse ltxdoc klootch +%%% @LaTeX-file{ +%%% filename = "ltxdocext.dtx", +%%% version = "0.0a0", +%%% date = "1999/06/04", +%%% time = "11:23:00 GMT", +%%% checksum = "929", +%%% author = "Arthur Ogawa (mailto:ogawa@teleport.com)", +%%% copyright = "Copyright (C) 1999 Arthur Ogawa, +%%% all rights reserved. Copying of this file is +%%% authorized only if either: +%%% (1) you make absolutely no changes to your copy, +%%% including name; OR +%%% (2) if you do make changes, you first rename it +%%% to some other name.", +%%% address = "Arthur Ogawa, +%%% USA", +%%% telephone = "", +%%% FAX = "", +%%% email = "ogawa@teleport.com", +%%% codetable = "ISO/ASCII", +%%% keywords = "latex, ltxdoc", +%%% supported = "yes", +%%% abstract = "extensions to the ltxdoc class", +%%% docstring = "The checksum field above generated by ltxdoc", +%%% } +% \fi +% +% \iffalse ltxdoc klootch +% The following references the \file{00readme.txt} file, +% which contains basic information about this package. +% The contents of this file are generated when +% you typeset the programmer's documentation. +% Search on "{filecontents*}{00readme.txt}" to locate it. +% \fi\input{00readme.txt}% +% +% \subsection{Bill of Materials} +% +% Following is a list of the files in this distribution arranged +% according to provenance. +% +% \subsubsection{Primary Source}% +% One single file generates all. +%\begin{verbatim} +%ltxdocext.dtx +%\end{verbatim} +% +% \subsubsection{Generated by \texttt{latex ltxdocext.dtx}}% +% Typesetting the source file under \LaTeX\ +% generates the readme and the installer. +%\begin{verbatim} +%00readme.txt ltxdocext.ins +%\end{verbatim} +% +% \subsubsection{Generated by \texttt{tex ltxdocext.ins}}% +% Typesetting the installer generates +% the package files. +%\begin{verbatim} +%ltxdocext.sty acrofont.sty +%\end{verbatim} +% +% \subsubsection{Documentation}% +% The following are the online documentation: +% \begin{verbatim} +%ltxdocext.pdf +% \end{verbatim} +% +% \subsubsection{Auxiliary}% +% The following are auxiliary files generated +% in the course of running \LaTeX: +% \begin{verbatim} +%ltxdocext.aux ltxdocext.idx ltxdocext.ind ltxdocext.log ltxdocext.toc +% \end{verbatim} +% +% \section{Code common to all modules}% +% +% The following may look a bit klootchy, but we +% want to require only one place in this file +% where the version number is stated, +% and we also want to ensure that the version +% number is embedded into every generated file. +% +% Now we declare that +% these files can only be used with \LaTeXe. +% An appropriate message is displayed if +% a different \TeX{} format is used. +% \begin{macrocode} +%<*doc|extensions|fonts> +\NeedsTeXFormat{LaTeX2e}[1995/12/01]% +%</doc|extensions|fonts> +% \end{macrocode} +% As desired, the following modules all +% take common version information: +% \begin{macrocode} +%<extensions>\ProvidesFile{ltxdocext.sty}% +%<fonts>\ProvidesFile{acrofont.sty}% +%<*doc> +\expandafter\ProvidesFile\expandafter{\jobname.dtx}% +%</doc> +% \end{macrocode} +% +% The following line contains, for once and for all, +% the version and date information. +% By various means, this information is reproduced +% consistently in all generated files and in the +% typeset documentation. +% \begin{macrocode} + [1999/06/04 0.0a0 ltxdoc extensions package]% \fileversion +% \end{macrocode} +% +% +% \section{The driver module \texttt{doc}} +% +% This module, consisting of the present section, +% typesets the programmer's documentation, +% generating the \file{.ins} installer and \file{00readme.txt} as required. +% +% Because the only uncommented-out lines of code at the beginning of +% this file constitute the \file{doc} module itself, +% we can simply typeset the \file{.dtx} file directly, +% and there is thus rarely any need to +% generate the ``doc'' {\sc docstrip} module. +% Module delimiters are nonetheless required so that +% this code does not find its way into the other modules. +% +% The \enve{document} command concludes the typesetting run. +% +% \begin{macrocode} +%<*doc> +% \end{macrocode} +% +% \subsection{Stubs for \file{ltxdocext.sty} and \file{acrofont.sty}} +% +% This \env{filecontents} environment generates stub files +% \file{ltxdocext.sty} and \file{acrofont.sty}. +% Stubs are needed because the documentation for this package +% requires them and they might not yet be present. +% +% A real version of these files will be generated +% when you typeset \file{ltxdocext.ins}, at which point +% the stub file will be overwritten (you will have to give {\sc docstrip} +% permission to do this). +% \begin{macrocode} +\begin{filecontents}{ltxdocext.sty} +% Stub version of ltxdocext.sty +\end{filecontents} +\begin{filecontents}{acrofont.sty} +% Stub version of acrofont.sty +\end{filecontents} +% \end{macrocode} +% +% \subsection{The Preamble} +% The programmers documentation is formatted +% with the \classname{ltxdoc} class with local customizations, +% and with the usual code line indexing. +% \begin{macrocode} +\documentclass[draft]{ltxdoc} +\RequirePackage{ltxdocext}% +\RequirePackage{acrofont}% +\CodelineIndex\EnableCrossrefs +% \end{macrocode} +% +% \subsubsection{Docstrip and info directives} +% We use so many {\sc docstrip} modules that we set the +% \texttt{StandardModuleDepth} counter to 1. +% \begin{macrocode} +\setcounter{StandardModuleDepth}{1} +% \end{macrocode} +% The following command retrieves the date and version information +% from this file. +% \begin{macrocode} +\expandafter\GetFileInfo\expandafter{\jobname.dtx}% +% \end{macrocode} +% +% +% \subsection{The installer file} +% +% The installer \file{ltxdocext.ins} appears here. +% If you have retrieved the standard distribution of this package, +% the installer file is already on your filesystem. +% If you are bootstrapping, +% the first typesetting of the \file{.dtx} file +% will cause the installer to be generated. +% +% The following modules are used to direct +% {\sc docstrip} in generating the external files: +% \begin{center} +% \begin{tabular}{lll} +% \textbf{Module}&\textbf{File}&\textbf{Description}\\ +% doc &\file{ltxdocext.drv}&driver for programmer's documantation\\ +% extensions&\file{ltxdocext.sty}<xdoc extensions package\\ +% fonts &\file{acrofont.sty} &package to use only Acrobat fonts +% \end{tabular} +% \end{center} +% +% \begin{macrocode} +\begin{filecontents}{ltxdocext.ins} +%% This file will generate documentation and runtime files +%% from ltxdocext.dtx when run through LaTeX or TeX. +\input docstrip +\preamble + +This is a generated file; +altering it directly is inadvisable; +instead, modify the original source file. +See the URL in the file 00readme.txt. + +Copyright notice. + + These files are distributed + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. + +\endpreamble +\keepsilent + \generate{% + \file{ltxdocext.drv}{\from{ltxdocext.dtx}{doc}}% + \file{ltxdocext.sty}{\from{ltxdocext.dtx}{extensions}}% + \file{acrofont.sty}{\from{ltxdocext.dtx}{fonts}}% + }% +\ifToplevel{ +\Msg{***********************************************************} +\Msg{*} +\Msg{* To finish the installation, please move} +\Msg{* ltxdocext.sty} +\Msg{* into a directory searched by TeX.} +\Msg{*} +\Msg{* To produce the user documentation, + run ltxdocext.tex through LaTeX.} +\Msg{*} +\Msg{* To produce the programmer documentation, + run ltxdocext.dtx through LaTeX.} +\Msg{*} +\Msg{* Happy TeXing} +\Msg{***********************************************************} +} +\endbatchfile +\end{filecontents} +% \end{macrocode} +% Note that, because all of the files generated by the installer +% are part of the standard distribution, it will +% be necessary to run the installer only when bootstrapping +% (or, of course, during development). +% Note, too, that it is rare to generate the \file{doc} +% module because it suffices to simply typeset the \file{.dtx} file itself. +% +% \subsection{The ``Read Me'' File} +% As promised above, here is the contents of the +% ``Read Me'' file. That file serves a double purpose, +% since it also constitutes the beginining of the +% programmer's documentation. What better thing, after +% all, to have appear at the beginning of the +% typeset documentation? +% +% A good discussion of how to write a ReadMe file can be found in +% Engst, Tonya, ``Writing a ReadMe File? Read This'' +% \emph{MacTech} October 1998, p. 58. +% +% Note the appearance of the +% \cmd\StopEventually\ command, which marks the +% dividing line between the user documentation +% and the programmer documentation. +% +% The usual user will not be asked to +% do a full build, not to speak +% of the bootstrap. +% Instructions for carrying these processes +% begin the programmer's manual. +% +% \begin{macrocode} +\begin{filecontents*}{00readme.txt} +\title{% + Extensions to the \classname{ltxdoc} class% + \thanks{% + This file has version number \fileversion, + last revised \filedate.% + % For version number and date, search on "\fileversion" in the .dtx file, + % or see the end of the 00readme.txt file. + }% +}% + +\author{% +Arthur Ogawa (\texttt{mailto:ogawa@teleport.com}), +\fileversion\\Copyright (C) 1999 Arthur Ogawa +}% +\maketitle + +This file embodies the \classname{ltxdocext} package, +the implementation and its user documentation. + +The distribution point for this work is +\url{ftp://ftp.teleport.com/users/ogawa/macros/latex/contrib/supported/ltxdocext...}, +which contains fully unpacked, prebuilt runtime files and documentation. + +To use this document class, you must have a working +\TeX\ installation equipped with \LaTeXe\ +and possibly pdftex and Adobe Acrobat Reader or equivalent. + +To install, retrieve the distribution, +unpack it into a directory on the target computer, +and move the files \file{ltxdocext.sty} and \file{acrofont.sty} +into a location in your filesystem where they will be found by \LaTeX. + +If you will be using the \classname{acrofont} package, you must +also install the virtual fonts +\file{zpsynocmrv}, \file{zptmnocmr}, +\file{zptmnocmrm}, and \file{zpzcnocmry}. +The corresponding \file{.tfm}, \file{.vf}, and \file{.vpl} +files are part of this distribution. + +To use, read the user documentation \file{ltxdocext.pdf}. +The \file{.dtx} file, \file{ltxdocext.dtx}, constitutes +in itself an instance of use of the \classname{ltxdocext} +package and the \classname{acrofont} package. + +\tableofcontents + +\section{Processing Instructions} + +The package files \file{ltxdocext.sty} and \file{acrofont.sty} +are generated from this file, \file{ltxdocext.dtx}, +via the {\sc docstrip} facility of \LaTeX +via |tex ltxdocext.ins|. +The typeset documentation that you are now reading is generated from +the same file by typesetting it with \LaTeX\ or pdftex +via |latex ltxdocext.dtx| or |pdflatex ltxdocext.dtx|. + +\subsection{Build Instructions} + +You may bootstrap this suite of files solely from \file{ltxdocext.dtx}. +Prepare by installing \LaTeXe\ (and either tex or pdftex) on your computer, +then carry out the following steps: +\begin{enumerate} +\item +Within an otherwise empty directory, +typeset \file{ltxdocext.dtx} with \LaTeX\ or pdflatex; +you will obtain the typeset documentation you are now reading, +along with +the installer \file{ltxdocext.ins}, +and the file \file{00readme.txt}. +\item +Now typeset \file{ltxdocext.ins}, +thereby generating the package file \file{ltxdocext.sty}, +and the package file \file{acrofont.sty}. +Make sure that {\sc docstrip} receives permission +to overwrite existing versions of these packages. +\item +Install \classname{ltxdocext.sty} and \classname{acrofont.sty} +by moving them to a location +in your filesystem where they will be found by \LaTeX. +\item +Now complete the typesetting of the documentation by +retypesetting \file{ltxdocext.dtx}. +Note: you will have to run \LaTeX\ twice, then \file{makeindex}, then +\LaTeX\ again in order to obtain a valid index and table of contents. +\end{enumerate} +\end{filecontents*} +% \end{macrocode} +% +% \subsection{The Document Body} +% +% Here is the document body, containing only a +% \cmd\DocInput\ directive---referring to this very file. +% This very cute self-reference is a common \classname{ltxdoc} idiom. +% \begin{macrocode} +\begin{document}% +\expandafter\DocInput\expandafter{\jobname.dtx}% +% ^^A\PrintChanges +\end{document} +% \end{macrocode} +% +% \begin{macrocode} +%</doc> +% \end{macrocode} +% +% \section{Using the \classname{ltxdoc} and \classname{acrofont} packages}% +% These packages are an adjunct to the +% standard \LaTeX\ \classname{ltxdoc} class and may be +% simply invoked as follows: +%\begin{verbatim} +%\documentclass[draft]{ltxdoc} +%\RequirePackage{ltxdocext}% +%\RequirePackage{acrofont}% +%\CodelineIndex\EnableCrossrefs +%\end{verbatim} +% +% Your document should simply cleave to the standards of the +% \classname{ltxdoc} class, with extensions and alterations as noted. +% +% \subsection{Extensions to the \classname{ltxdoc} class}% +% +% \subsubsection{Extensions to the \env{verbatim} environment and \cs{verb} command}% +% +% The delimiters \verb$<<$ and \verb$>>$ within the scope of the verbatim environment +% or within the argument of the \cmd\verb\ command produce +% italics surrounded by angle brackets. +% This typographic convention usually indicates +% \emph{metalanguage}, i.e., a placeholder. +% +% To obtain the angle bracket character per se, +% double the character, viz., ``\verb$the delimiter \verb+<<<<+$''. +% +% \subsubsection{The -\texttt{matter} Commands Work} +% The sectioning commands \cmd\frontmatter, \cmd\mainmatter, and \cmd\backmatter +% of the standard \LaTeX\ \classname{book} class are operative in the +% \classname{ltxdoc} class. +% +% \subsubsection{The \cs{GetFileInfo} command}\label{sec:GetFileInfo} +% You can use the \cmd\GetFileInfo\ command to extract +% the date, version, and file info of a file which has registered itself +% via the \cmd\ProvidesFile\ or \cmd\ProvidesClass\ command +% (employing the optional argument thereto). +% +% For instance, if your document contains the following: +%\begin{verbatim} +%\RequirePackage{ltxdocext}% +%\GetFileInfo{ltxdocext.sty}% +%\end{verbatim} +% then the following control sequence names will have +% a value corresponding to that package's \cmd\ProvidesFile\ +% command: +% \cmd\filedate: the file's date, +% \cmd\fileversion: the file's version, and +% \cmd\fileinfo: the file's info. +% +% \subsubsection{Self-Indexing Commands}% +% Certain commands automatically produce an index entry +% (or several related entries) according to the meaning. +% +% \begin{unnumtable} +% \begin{tabular}{ll} +% meta-text &\cmd\m\arg{text}\\ +% command &\cmd\cmd\cmd\csname\\ +% environment name &\cmd\env\arg{name}\\ +% \cmd\begin\verb`{foo}`&\cmd\envb\arg{foo}\\ +% \cmd\end\verb`{foo}` &\cmd\enve\arg{foo}\\ +% argument &\cmd\arg\arg{name}\\ +% optional &\cmd\oarg\arg{name}\\ +% filename &\cmd\file\arg{name}\\ +% url &\cmd\url\arg{name}\\ +% document class &\cmd\classname\arg{name}\\ +% document substyle &\cmd\substyle\arg{name}\\ +% class option &\cmd\classoption\arg{name} +% \end{tabular} +% \end{unnumtable} +% +% \subsubsection{Unnumbered Tables}% +% +% When your documentation requires the use of an unnumbered table, +% use the \env{unnumtable} environment: +% \begin{verbatim} +%\begin{unnumtable} +%\begin{tabular}{ll} +%<table rows> +%\end{tabular} +%\end{unnumtable} +% \end{verbatim} +% +% \subsubsection{Structuring Tables}% +% The commands \cmd\toprule, \cmd\colrule, and \cmd\botrule +% allow you to mark the beginning of the column heads +% the beginning of the table body, and the end of +% the table body, respectively. +% In context, +% \begin{verbatim} +%\begin{tabular}{ll} +%\toprule +%<table head rows> +%\colrule +%<table rows> +%\botrule +%\end{tabular} +% \end{verbatim} +% +% \subsubsection{A Sectioning Command Below \cs{subsection}}% +% The \cmd\subsubsection\ command is defined. +% +% +% \subsection{Alterations to the \classname{ltxdoc} class}% +% The following involve no new markup, but they +% do change the appearance of your formatted documentation: +% \begin{enumerate} +% \item +% Using the \classname{acrofont} package causes your +% document to be formatted using the standard +% Acrobat fonts to the greatest extent possible. +% This means that for most documents, Computer Modern +% is not used at all. +% Math that unavoidable must use CM still exists, however. +% \item +% An index will be produced at the end of the document +% without your needing to explicitly mark it up, and +% it will have an entry in the TOC. +% \item +% The \env{quote} environment has a slightly smaller left margin. +% \item +% Array columns are set tight by default. +% \item +% A host of \cmd\DoNotIndex\ directives are invoked. +% I intend this list to grow to encompass +% even more commands. Send me your suggestions. +% \end{enumerate} +% +% +% +%\StopEventually{} +% +% \section{Extensions to the ltxdoc class} +% The \file{extensions} {\sc docstrip} module comprises the +% package \classname{ltxdocext.sty}, which provides extensions +% to the standard \LaTeX\ \classname{ltxdoc} class. +% +% \subsection{Beginning of the \file{extensions} {\sc docstrip} module} +% \begin{macrocode} +%<*extensions> +\def\class@name{ltxdocext}% +\expandafter\PackageInfo\expandafter{\class@name}{% + An extension to the \protect\LaTeXe\space ltxdoc class + by A. Ogawa (ogawa@teleport.com)% +}% +% \end{macrocode} +% +% \subsection{Incorporate \classname{ltxguide.cls} extensions} +% +% Code extracted from \classname{ltxguide.cls}, by Alan Jeffrey. +% ``This code stolen from \classname{ltxguide.cls}: +% Some hacks with verbatim... NB: this would be better done with the +% verbatim package, but this document has to run on any \LaTeX +% installation.'' +% \begin{macrocode} +\RequirePackage{verbatim}% +\let\o@verbatim\verbatim +\def\verbatim{% + \ifhmode\unskip\par\fi +% \nopagebreak % Overridden by list penalty + \ifx\@currsize\normalsize + \small + \fi + \o@verbatim +}% +% \end{macrocode} +% +% Here we extend the font-setting command to include making \texttt{<>} active +% (i.e., adjusting the input encoding). +% \begin{macrocode} +\renewcommand \verbatim@font {% + \normalfont \ttfamily + \catcode`\<=\active + \catcode`\>=\active +}% +% \end{macrocode} +% +% Make \verb$|...|$ a synonym for \cmd\verb\verb$|...|$. +% \begin{macrocode} +\RequirePackage{shortvrb} +\AtBeginDocument{% + \MakeShortVerb{\|}% +}% +% \end{macrocode} +% +% Make active bracket characters produce italics surrounded by angle brackets +% (used in \env{verbatim} and \cmd\verb). +% \verb|<<| produces a less-than, and \verb|>>| produces a greater-than. +% \begin{macrocode} +\begingroup + \catcode`\<=\active + \catcode`\>=\active + \gdef<{\@ifnextchar<\@lt\@meta} + \gdef>{\@ifnextchar>\@gt\@gtr@err} + \gdef\@meta#1>{\m{#1}} + \gdef\@lt<{\char`\<} + \gdef\@gt>{\char`\>} +\endgroup +\def\@gtr@err{% + \ClassError{ltxguide}{% + Isolated \protect>% + }{% + In this document class, \protect<...\protect> + is used to indicate a parameter.\MessageBreak + I've just found a \protect> on its own. + Perhaps you meant to type \protect>\protect>? + }% +} +\def\verbatim@nolig@list{\do\`\do\,\do\'\do\-} +% \end{macrocode} +% End of code stolen from \file{ltxguide.cls}. Thanks, Alan. +% +% Add functionality from doc.dtx: +% (code stolen from doc.dtx): +% \begin{macrocode} +\def\GetFileInfo#1{% + \def\filename{#1}% + \def\@tempb##1 ##2 ##3\relax##4\relax{% + \def\filedate{##1}% + \def\fileversion{##2}% + \def\fileinfo{##3}}% + \edef\@tempa{\csname ver@#1\endcsname}% + \expandafter\@tempb\@tempa\relax? ? \relax\relax} +% \end{macrocode} +% (end of code stolen from doc.dtx. Thanks FMi.) +% +% Various forms of self-indexing commands: +% \begin{macrocode} +\DeclareRobustCommand{\m}[1]{% + \meta{#1}% + \index{#1=\string\meta{#1} placeholder}\index{placeholder>#1=\string\meta{#1}}% +}% +\DeclareRobustCommand\meta[1]{% + \mbox{\LANGLE\itshape#1\/\RANGLE}% +}% +\def\LANGLE{$\langle$}% +\def\RANGLE{$\rangle$}% +\DeclareRobustCommand{\arg}[1]{% + {\ttfamily\string{}\meta{#1}{\ttfamily\string}}% + \index{#1=\string\ttt{#1}, argument}\index{argument>#1=\string\ttt{#1}}% +}% +\let\oarg\undefined +\DeclareRobustCommand{\oarg}[1]{% + {\ttfamily[%] + }\meta{#1}{\ttfamily%[ + ]}% + \index{#1=\string\ttt{#1}, optional argument}% + \index{argument, optional>#1=\string\ttt{#1}}% +}% +\DeclareRobustCommand\cmd{\begingroup\makeatletter\@cmd}% +\def\@cmd#1{% + \endgroup + \cs{\expandafter\cmd@to@cs\string#1}% + \expandafter\cmd@to@index\string#1\@nil +}% +\def\cmd@to@cs#1#2{\char\number`#2\relax}% +\def\cmd@to@index#1#2\@nil{% + \index{#2=\string\cmd#1#2}%\index{command>#2=\string\cmd#1#2}% +}% +\DeclareRobustCommand\cs[1]{{\ttfamily\char`\\#1}}% +\def\scmd#1{% + \cs{\expandafter\cmd@to@cs\string#1}% + \expandafter\scmd@to@index\string#1\@nil +}% +\def\scmd@to@index#1#2\@nil#3{% + \index{\string$#3=\string\cmd#1#2---#3}% +%\index{command>\string$#3=\string\cmd#1#2---#3}% +}% +\DeclareRobustCommand\env{\name@idx{environment}}% +\DeclareRobustCommand\envb[1]{% + {\ttfamily\string\begin\string{}\env{#1}{\ttfamily\string}}% +}% +\DeclareRobustCommand\enve[1]{{\ttfamily\string\end\string{}\env{#1}{\ttfamily\string}}}% +\DeclareRobustCommand*{\file}[1]{% + {\ttfamily#1}% + \index{#1=\string\ttt{#1}}\index{file>#1=\string\ttt{#1}}% +}% +\DeclareRobustCommand\substyle{\name@idx{document substyle}}% +\DeclareRobustCommand\classoption{\name@idx{document class option}}% +\DeclareRobustCommand\classname{\name@idx{document class}}% +\def\name@idx#1#2{% + {\ttfamily#2}% + \index{#2\space#1=\string\ttt{#2}\space#1}\index{#1>#2=\string\ttt{#2}}% +}% +\DeclareRobustCommand\url@ltxdocext{\begingroup\catcode`\/\active\catcode`\.\active\catcode`\:\active\@url}% +\AtBeginDocument{% + \ifx\url\undefined\let\url\url@ltxdocext\fi +}% +\def\@url#1{% + \url@break{\ttfamily#1}% + \url@char\edef\@tempa{#1=\string\url{#1}}% + \expandafter\index\expandafter{\@tempa}% + \expandafter\index\expandafter{\expandafter u\expandafter r\expandafter l\expandafter >\@tempa}% + \endgroup +}% +{\catcode`\:\active\aftergroup\def\aftergroup:}{\active@colon}% +\def\colon@break{\colon@char\allowbreak}% +\def\colon@char{:}% +{\catcode`\/\active\aftergroup\def\aftergroup/}{\active@slash}% +\def\slash@break{\slash@char\allowbreak}% +\def\slash@char{/}% +{\catcode`\.\active\aftergroup\def\aftergroup.}{\active@dot}% +\def\dot@break{\dot@char\allowbreak}% +\def\dot@char{.}% +\def\url@break{\let\active@slash\slash@break\let\active@dot\dot@break\let\active@colon\colon@break}% +\def\url@char{\let\active@slash\slash@char\let\active@dot\dot@char\let\active@colon\colon@char}% +% \end{macrocode} +% +% \subsection{Changes to the base class of the ltxdoc class} +% Modify \env{theindex} environment so that it produces a TOC entry +% \begin{macrocode} +\renewenvironment{theindex} + {\if@twocolumn + \@restonecolfalse + \else + \@restonecoltrue + \fi + \columnseprule \z@ + \columnsep 35\p@ +\def\see##1##2{\textit{See} ##1}% +\def\seealso##1##2{\textit{See also} ##1}% +\def\cmd##1{\cs{\expandafter\cmd@to@cs\string##1}}% +\def\@url##1{\url@break\ttt{##1}\endgroup}% +\def\ttt##1{{\ttfamily##1}}% +\mathchardef\save@secnumdepth\c@secnumdepth +\c@secnumdepth\m@ne + \twocolumn[\section{\indexname}]% +% \@mkboth{\MakeUppercase\indexname}% +% {\MakeUppercase\indexname}% +\c@secnumdepth\save@secnumdepth + \thispagestyle{plain}\parindent\z@ + \parskip\z@ \@plus .3\p@\relax + \let\item\@idxitem} + {\if@restonecol\onecolumn\else\clearpage\fi} +\renewenvironment{quote} + {\list{}{% + \leftmargin1em\relax + \rightmargin\leftmargin + }% + \item\relax} + {\endlist} +% \end{macrocode} +% +% \subsection{Extensions to the base class of \classname{ltxdoc.cls}} +% +% Matter commands from \classname{book.cls} +% \begin{macrocode} +\newif\if@mainmatter +\newif\if@openright +\@openrighttrue +\DeclareRobustCommand\frontmatter{% + \cleartorecto + \@mainmatterfalse + \pagenumbering{roman}% +}% +\DeclareRobustCommand\mainmatter{% + \cleartorecto + \@mainmattertrue + \pagenumbering{arabic}% +}% +\DeclareRobustCommand\backmatter{% + \if@openright + \cleartorecto + \else + \clearpage + \fi + \@mainmatterfalse +}% +\ifx\undefined\cleartorecto + \def\cleartorecto{\cleardoublepage}% +\fi +% \end{macrocode} +% +% Unnumbered tables +% +% \begin{environment}{unnumtable} +% An unnumbered table does not float. +% \begin{macrocode} +\def\@to{to}% +\newenvironment{unnumtable}{% + \par + \addpenalty\predisplaypenalty + \addvspace\abovedisplayskip + \hbox\@to\hsize\bgroup\hfil\ignorespaces + \let\@Hline\@empty +}{% + \unskip\hfil\egroup + \penalty\postdisplaypenalty + \vskip\belowdisplayskip + \aftergroup\ignorespaces + \@endpetrue +}% +% \end{macrocode} +% \end{environment} +% +% Emulate \cmd\toprule\ and friends +% \begin{macrocode} +\newcommand\toprule{\hline\hline}% +\newcommand\colrule{\\\hline}% +\newcommand\botrule{\\\hline\hline}% +% \end{macrocode} +% +% Define sectioning command below \cmd\subsubsection. +% \begin{macrocode} +\DeclareRobustCommand\subsubsubsection{% + \@startsection{subsubsection}{4}% + {\z@}{-15\p@\@plus-5\p@\@minus-2\p@}% + {5\p@}{\normalfont\normalsize\itshape}% +}% +% \end{macrocode} +% +% \subsection{In lieu of \file{ltxdoc.cfg}} +% We don't want everything to appear in the index +% \begin{macrocode} +\DoNotIndex{\',\.,\@M,\@@input,\@Alph,\@alph,\@addtoreset,\@arabic} +\DoNotIndex{\@badmath,\@centercr,\@cite} +\DoNotIndex{\@dotsep,\@empty,\@float,\@gobble,\@gobbletwo,\@ignoretrue} +\DoNotIndex{\@input,\@ixpt,\@m,\@minus,\@mkboth} +\DoNotIndex{\@ne,\@nil,\@nomath,\@plus,\roman,\@set@topoint} +\DoNotIndex{\@tempboxa,\@tempcnta,\@tempdima,\@tempdimb} +\DoNotIndex{\@tempswafalse,\@tempswatrue,\@viipt,\@viiipt,\@vipt} +\DoNotIndex{\@vpt,\@warning,\@xiipt,\@xipt,\@xivpt,\@xpt,\@xviipt} +\DoNotIndex{\@xxpt,\@xxvpt,\\,\ ,\addpenalty,\addtolength,\addvspace} +\DoNotIndex{\advance,\ast,\begin,\begingroup,\bfseries,\bgroup,\box} +\DoNotIndex{\bullet} +\DoNotIndex{\cdot,\cite,\CodelineIndex,\cr,\day,\DeclareOption} +\DoNotIndex{\def,\DisableCrossrefs,\divide,\DocInput,\documentclass} +\DoNotIndex{\DoNotIndex,\egroup,\ifdim,\else,\fi,\em,\endtrivlist} +\DoNotIndex{\EnableCrossrefs,\end,\end@dblfloat,\end@float,\endgroup} +\DoNotIndex{\endlist,\everycr,\everypar,\ExecuteOptions,\expandafter} +\DoNotIndex{\fbox} +\DoNotIndex{\filedate,\filename,\fileversion,\fontsize,\framebox,\gdef} +\DoNotIndex{\global,\halign,\hangindent,\hbox,\hfil,\hfill,\hrule} +\DoNotIndex{\hsize,\hskip,\hspace,\hss,\if@tempswa,\ifcase,\or,\fi,\fi} +\DoNotIndex{\ifhmode,\ifvmode,\ifnum,\iftrue,\ifx,\fi,\fi,\fi,\fi,\fi} +\DoNotIndex{\input} +\DoNotIndex{\jobname,\kern,\leavevmode,\let,\leftmark} +\DoNotIndex{\list,\llap,\long,\m@ne,\m@th,\mark,\markboth,\markright} +\DoNotIndex{\month,\newcommand,\newcounter,\newenvironment} +\DoNotIndex{\NeedsTeXFormat,\newdimen} +\DoNotIndex{\newlength,\newpage,\nobreak,\noindent,\null,\number} +\DoNotIndex{\numberline,\OldMakeindex,\OnlyDescription,\p@} +\DoNotIndex{\pagestyle,\par,\paragraph,\paragraphmark,\parfillskip} +\DoNotIndex{\penalty,\PrintChanges,\PrintIndex,\ProcessOptions} +\DoNotIndex{\protect,\ProvidesClass,\raggedbottom,\raggedright} +\DoNotIndex{\refstepcounter,\relax,\renewcommand} +\DoNotIndex{\rightmargin,\rightmark,\rightskip,\rlap,\rmfamily} +\DoNotIndex{\secdef,\selectfont,\setbox,\setcounter,\setlength} +\DoNotIndex{\settowidth,\sfcode,\skip,\sloppy,\slshape,\space} +\DoNotIndex{\symbol,\the,\trivlist,\typeout,\tw@,\undefined,\uppercase} +\DoNotIndex{\usecounter,\usefont,\usepackage,\vfil,\vfill,\viiipt} +\DoNotIndex{\viipt,\vipt,\vskip,\vspace} +\DoNotIndex{\wd,\xiipt,\year,\z@} +\DoNotIndex{\next} +% \end{macrocode} +% +% Direct \classname{ltxdoc} to produce an index. +% \begin{macrocode} +\AtEndDocument{\PrintIndex}% +% \end{macrocode} +% +% \subsection{Extension to \LaTeX's \env{filecontents} Environment} +% We want to +% coax the version number into \env{filecontents}-generated files. +% Note that we expect \cmd\fileversion\ and \cmd\filedate\ to +% hold the needed information. For this to be the case, +% your document should execute the \cmd\GetFileInfo\ command +% (as documented in section~\ref{sec:GetFileInfo}) before +% any instances of \env{filecontents}. +% \begin{macrocode} +\makeatletter +\def\endfilecontents{% + \immediate\write\reserved@c{% + \string\iffalse\space ltxdoc klootch^^J% + \ifx\undefined\fileversion\else + \ifx\undefined\filedate\else + This file has version number \fileversion, last revised \filedate.% + \fi\fi + \string\fi + }% + \immediate\closeout\reserved@c + \def\T##1##2##3{% + \ifx##1\@undefined\else + \@latex@warning@no@line{##2 has been converted to Blank ##3e}% + \fi + }% + \T\L{Form Feed}{Lin}% + \T\I{Tab}{Spac}% + \immediate\write\@unused{}% +}% +\expandafter\let\csname endfilecontents*\endcsname\endfilecontents +\makeatother +% \end{macrocode} +% +% Alter formatting in arrays; set them tight. +% \begin{macrocode} +\setlength\arraycolsep{0pt}% +% \end{macrocode} +% +% \subsection{End of the \file{extensions} {\sc docstrip} module} +% Here ends the module. +% \begin{macrocode} +%</extensions> +% \end{macrocode} +% +% +% \section{Font Package for Acrobat Compatability}% +% The package \classname{acrofont} +% substitutes Acrobat-standard fonts for Computer Modern where possible, +% even in math mode. +% Documents typeset with this package in effect will require as little +% downloaded font data as possible, but will not be exemplars of +% fine math typesetting. +% +% \subsection{Beginning of the \file{fonts} {\sc docstrip} module} +% The document class module comprises this and the next +% four sections. +% \begin{macro}{\class@base} +% We define in exactly one spot the base class. +% Typically that class will be one of \classname{book}, +% \classname{article}, or \classname{report}. +% The base class effectively defines the use and the markup scheme +% of the class of documents to be handled by this class. +% +% This class is a variant of the standard \LaTeX\ book class: +% \url{ftp://ctan.tug.org/tex-archive/macros/latex/unpacked}. +% \begin{macrocode} +%<*fonts> +\def\class@name{ltxdocext}% +\expandafter\ClassInfo\expandafter{\class@name}{% + Written for \protect\LaTeXe\space + by A. Ogawa (ogawa@teleport.com)% +}% +% \end{macrocode} +% \end{macro} +% +% \subsection{Variants on psfonts packages}% +% The following uses \file{times.sty} from \url{/packages/psnfss/psfonts.dtx} +% \begin{macrocode} +\RequirePackage{times}% +% \end{macrocode} +% +% The following uses \file{mathptm.sty} from \url{/packages/psnfss/psfonts.dtx} +% \begin{macrocode} +\RequirePackage{mathptm}% +% \end{macrocode} +% +% The following is a customization of \file{ot1ptmcm.fd}. +% The virtual font referred to here \file{zptmnocmr} is +% a variant of Sebastian Rahtz's \file{zptmcmr}, but with +% even more glyphs moved from CM to Acrobat-standard fonts. +% \begin{macrocode} +\DeclareFontFamily{OT1}{ptmcm}{} +\DeclareFontShape{OT1}{ptmcm}{m}{n}{ + <-> zptmnocmr +}{} +\DeclareFontShape{OT1}{ptmcm}{l}{n}{<->ssub * ptmnocm/m/n}{} +% \end{macrocode} +% +% The following is a customization of \file{omlptmcm.fd} +% The virtual font referred to here \file{zptmnocmrm} is +% a variant of Sebastian Rahtz's \file{zptmcmrm}, but with +% even more glyphs moved from CM to Acrobat-standard fonts. +% \begin{macrocode} +\DeclareFontFamily{OML}{ptmcm}{\skewchar \font =127} +\DeclareFontShape{OML}{ptmcm}{m}{it}{ + <-> zptmnocmrm +}{} +\DeclareFontShape{OML}{ptmcm}{l}{it}{<->ssub * ptmcm/m/it}{} +\DeclareFontShape{OML}{ptmcm}{m}{sl}{<->ssub * ptmcm/m/it}{} +\DeclareFontShape{OML}{ptmcm}{l}{sl}{<->ssub * ptmcm/m/sl}{} +% \end{macrocode} +% +% The following is a customization of \file{omspzccm.fd} +% The virtual font referred to here \file{zpzcnocmry} is +% a variant of Sebastian Rahtz's \file{zpzccmry}, but with +% even more glyphs moved from CM to Acrobat-standard fonts. +% \begin{macrocode} +\DeclareFontFamily{OMS}{pzccm}{} +\DeclareFontShape{OMS}{pzccm}{m}{n}{ + <-> zpzcnocmry +}{}% cmsy10 Symbol Zapf Chancery Medium-Italic Times-Roman +\DeclareFontShape{OMS}{pzccm}{l}{n}{<->ssub * pzccm/m/n}{} +% \end{macrocode} +% +% The following is a customization of \file{omxpsycm.fd} +% The virtual font referred to here \file{zpsynocmrv} is +% a variant of Sebastian Rahtz's \file{zpsycmrv}, but with +% even more glyphs moved from CM to Acrobat-standard fonts. +% \begin{macrocode} +\DeclareFontFamily{OMX}{psycm}{} +\DeclareFontShape{OMX}{psycm}{m}{n}{ + <-> zpsynocmrv +}{} +\DeclareFontShape{OMX}{psycm}{l}{n}{<->ssub * psycm/m/n}{} +% +\DeclareFontEncoding{8r}{}{}% from file: 8renc.def +\DeclareFontFamily{8r}{cmr}{\hyphenchar\font45 }% from file: 8rcmr.fd +\DeclareFontShape{8r}{cmr}{m}{n}{ + <-> cmr10 +}{} +% \end{macrocode} +% +% \subsection{Font definition files}% +% +% The following forces \LaTeX\ to do now what it would do anyway: +% load the `font definition' information for the fonts that we +% use. In this way, we prepare for faster processing through +% the \cmd\dump\ of a preformatted macro package that will not +% need to read in any packages or font definitions from disk. +% \begin{macrocode} +\input{8rphv.fd}% +\input{8rptm.fd}% +\input{ot1phv.fd}% +\input{ot1ptm.fd}% +\input{t1ptm.fd}% +% \end{macrocode} +% +% \subsection{More math substitutions}% +% +% The following definitions arrange to get certain glyphs from the +% text font instead of out of math pi fonts. +% In particular, the copyright and registered symbols +% are single glyphs instead of composites involving +% the big circle from the \file{cmsy} font. +% \begin{macrocode} + \def\eightRChar#1{{\def\encodingdefault{8r}\fontencoding\encodingdefault\selectfont\char"#1}}% + \def\LANGLE{$<$}%{\eightRChar{8B}}% + \def\RANGLE{$>$}%{\eightRChar{9B}}% +%\def\ASTER{\eightRChar{2A}}% +%\def\DAGGER{\eightRChar{86}}% +%\def\DDAGGER{\eightRChar{87}}% + \def\BULLET{\eightRChar{95}}% +%\def\SECTION{\eightRChar{A7}}% +%\def\PARAGRAPH{\eightRChar{B6}}% + \def\VERTBAR{\eightRChar{7C}}% + \def\COPYRIGHT{\eightRChar{A9}}% + \def\REGISTERED{\eightRChar{AE}}% +% \end{macrocode} +% +% \begin{macrocode} + \def\textbar{\VERTBAR}% + \def\textbullet{\BULLET}% + \def\textcopyright{\COPYRIGHT}% + \def\textregistered{\REGISTERED}% +% \end{macrocode} +% +% I have removed \cmd\ensuremath\ from the following definition, and +% all commands like \cmd\mathsection have been converted to +% e.g., \cmd\textsection. +% \begin{macrocode} +\def\@makefnmark{\@thefnmark}% +\def\@fnsymbol#1{{\ifcase#1\or *\or \dagger\or \ddagger\or + \textsection\or \textparagraph\or \|\or **\or \dagger\dagger + \or \ddagger\ddagger \else\@ctrerr\fi}} +% \end{macrocode} +% +% \subsection{End of the \file{fonts} {\sc docstrip} module} +% Here ends the module. +% \begin{macrocode} +%</fonts> +% \end{macrocode} +% +% +% \section{Programming Conventions}% +% In writing the above code, I cleave to certain conventions, noted +% here. +% My goal in explaining them is to encourage others maintaining this +% body of code to consider following them as well. +% The benefits are twofold: +% Some of the coding conventions aim to avoid programming pitfalls; +% following them reduces maintenance costs. +% Others make the code easier to read; following these eases the +% process of understanding how the code works. +% +% And, for my part, I prefer to read code of this type. +% +% \subsection{Whitespace Conventions}% +% Exactly where code lines break and indent, and where additional +% whitespace is inserted is explained here. +% \begin{itemize} +% \item +% Each new macro definition or register assignment begins a new line. +% Therefore, \cmd\def, \cmd\newcommand, and their ilk will start in column +% 1. +% \item +% Code is indented one space for each level of nesting within braces +% \verb|{}|. +% \item +% Likewise, if possible, for \cmd\if\dots and matching \cmd\fi. +% \item +% However, the closing brace or \cmd\fi\ is outdented by one so that it +% falls +% at the same level of indentation as its matching brace or \cmd\if, +% and it appears alone on its line. +% \item +% Use of the |tab| character is deprecated +% (tabs are not standardized across all applications and operating +% systems). +% \item +% Lines of code are limited to 72 characters. +% I follow this convention mostly to ease the transmission of files +% via email (a deprecated practice) and to accomodate people with +% small monitors. +% But \classname{ltxdoc} output looks better with the shorter lines, too. +% \item +% Extraneous whitespace is avoided by using the comment character |%|. +% In most cases, if falling at the end of a line of code, +% a brace will be immediately followed by a comment character, +% as will the macro parameter |#1| and any one-letter control sequence, +% like |\\|. +% \end{itemize} +% These conventions taken together are illustrated by the following: +% \begin{verbatim} +%\def\prepdef#1#2{% +% \@ifxundefined#1{\toks@{}}{\toks@\expandafter{#1}}% +% \toks@ii{#2}% +% \edef#1{\the\toks@ii\the\toks@}% +%}% +% \end{verbatim} +% In the above, the definition of \cmd\prepdef\ would not fit on a single +% line, +% and required breaking. The first and last lines have matching braces, +% and are a the same level of indentation, with the last line containing +% a single brace. +% +% Each of the three intervening lines has balanced braces and is +% indented by one space. Each line that would otherwise end in a single +% brace character is terminated by a comment character. +% +% Some coders rely on the fact that a space character seen by \TeX's +% scanner while in vertical mode is a no-op. +% Be that as it may, I eliminate them unless actually intentional. +% +% \subsection{Conventions For Procedures}% +% Here are some of my preferences when writing procedures: +% \begin{itemize} +% +% \item +% I dislike defining a macro within another macro, especially when the +% pattern part is non-nil. +% You know, you are not saving much space in |mem| when you do this, +% right? +% You do save space in the hash table and the string pool, though. +% On the other hand, we are not dealing with small \TeX\ engines +% anymore; Team \LaTeX\ has made sure of this. +% +% \item +% If two or more macros have very similar replacement parts, consider +% layering. +% +% \item +% Macros may perform parsing, may maintain tokens or registers, or may +% set type (produce marks). I try to avoid combining the three functions +% in a single macro. +% +% \item +% When a procedure both does assignments and sets type, I try to have a +% clean separations between the two activities. Try to avoid: +% \begin{verbatim} +% \vskip10pt +% \parindent=0pt +% \leavevmode +% \end{verbatim} +% +% \item +% The Boolean calculus (cf. \cmd\@ifx) +% is very useful and produces code that executes nicely. +% Using it also helps avoid your having to debug problems where +% \cmd\if\dots\ and \cmd\fi\ are not properly balanced +% (a nightmare in case you didn't already experience it). +% +% \end{itemize} +% +% \subsection{Conventions For \LaTeX}% +% Team \LaTeX\ make certain recommendations in \file{clsguide.tex}. +% Ones that I particularly pay attention to are: +% \begin{itemize} +% +% \item +% For the sake of ``color safety'', +% use \cmd\sbox\ rather than \cmd\setbox, \cmd\mbox\ rather than \cmd\hbox, and +% \cmd\parbox\ or \env{minipage} rather than \cmd\vbox. +% +% \item +% Use \cmd\newcommand\ and \cmd\newenvironment\ to declare user-level commands +% and environments. Avoid the idiom \cmd\def\cmd\foo, \cmd\def\cmd\endfoo\ to define +% an environment. +% Ideally, all user-level markup could be extracted from the +% document class by grepping on \cmd\newcommand\ and \cmd\newenvironment. +% +% \item +% Prefer to use \cmd\setlength\ to assign registers. +% +% \end{itemize} +% I cannot help but notice that much of the code of \LaTeX\ itself fails +% to comply with many of the coding recommendations of Team \LaTeX. +% +% +% \Finale +% %Here ends the programmer's documentation. +% \endinput +% +\endinput diff --git a/Master/texmf-dist/source/latex/revtex4/ltxgrid.dtx b/Master/texmf-dist/source/latex/revtex4/ltxgrid.dtx new file mode 100644 index 00000000000..ab0209ed098 --- /dev/null +++ b/Master/texmf-dist/source/latex/revtex4/ltxgrid.dtx @@ -0,0 +1,4815 @@ +% \iffalse ltxdoc klootch +% ltxgrid.dtx: package to change page grid, MVL. +% Copyright (c) 1999 Arthur Ogawa +% +% Disclaimer +% This file is distributed WITHOUT ANY WARRANTY; +% without even the implied warranty of +% MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. +% ReadMe +% For the documentation and more detailed instructions for +% installation, typeset this document with \LaTeX. +% \fi +% \GetFileInfo{ltxgrid.dtx}\CheckSum{4234} +% +% \iffalse ltxdoc klootch +%<*ltxgrid> +%%% @LaTeX-file{ +%%% filename = "ltxgrid.dtx", +%%% version = "1.0rc5", +%%% date = "2001/07/26", +%%% time = "12:23:00 GMT+8", +%%% checksum = "4234", +%%% author = "Arthur Ogawa (mailto:ogawa@teleport.com), +%%% commissioned by the American Physical Society. +%%% ", +%%% copyright = "Copyright (C) 1999, 2000 Arthur Ogawa, +%%% distributed under the terms of the +%%% LaTeX Project Public License, see +%%% ftp://ctan.tug.org/macros/latex/base/lppl.txt +%%% ", +%%% address = "Arthur Ogawa, +%%% USA", +%%% telephone = "", +%%% FAX = "", +%%% email = "ogawa@teleport.com", +%%% codetable = "ISO/ASCII", +%%% keywords = "latex, page grid, main vertical list", +%%% supported = "yes", +%%% abstract = "package to change page grid, MVL", +%%% docstring = "The checksum field above generated by ltxdoc", +%%% } +%</ltxgrid> +% \fi +% +% \iffalse ltxdoc klootch +% The following references the \file{00readme.txt} file, +% which contains basic information about this package. +% The contents of this file are generated when +% you typeset the programmer's documentation. +% Search on "{filecontents*}{00readme.txt}" to locate it. +% \fi\input{00readme.txt}% +% +% \subsection{Bill of Materials} +% +% Following is a list of the files in this distribution arranged +% according to provenance. +% +% \subsubsection{Primary Source}% +% One single file generates all. +%\begin{verbatim} +%ltxgrid.dtx +%\end{verbatim} +% +% \subsubsection{Generated by \texttt{latex ltxgrid.dtx}}% +% Typesetting the source file under \LaTeX\ +% generates the readme and the installer. +%\begin{verbatim} +%00readme.txt ltxgrid.ins +%\end{verbatim} +% +% \subsubsection{Generated by \texttt{tex ltxgrid.ins}}% +% Typesetting the installer generates +% the package files. +%\begin{verbatim} +%ltxgrid.sty +%\end{verbatim} +% +% \subsubsection{Documentation}% +% The following are the online documentation: +% \begin{verbatim} +%ltxgrid.pdf +% \end{verbatim} +% +% \subsubsection{Auxiliary}% +% The following are auxiliary files generated +% in the course of running \LaTeX: +% \begin{verbatim} +%ltxgrid.aux ltxgrid.idx ltxgrid.ind ltxgrid.log ltxgrid.toc +% \end{verbatim} +% +% \section{Code common to all modules}% +% +% The following may look a bit klootchy, but we +% want to require only one place in this file +% where the version number is stated, +% and we also want to ensure that the version +% number is embedded into every generated file. +% +% Now we declare that +% these files can only be used with \LaTeXe. +% An appropriate message is displayed if +% a different \TeX{} format is used. +% \begin{macrocode} +%<*doc|ltxgrid> +\NeedsTeXFormat{LaTeX2e}[1995/12/01]% +%</doc|ltxgrid> +% \end{macrocode} +% As desired, the following modules all +% take common version information: +% \begin{macrocode} +%<ltxgrid>\ProvidesFile{ltxgrid.sty}% +%<*doc> +\expandafter\ProvidesFile\expandafter{\jobname.dtx}% +%</doc> +% \end{macrocode} +% +% The following line contains, for once and for all, +% the version and date information. +% By various means, this information is reproduced +% consistently in all generated files and in the +% typeset documentation. +% \begin{macrocode} +%<*doc|ltxgrid> + [2001/07/26 1.0rc5 page grid package]% \fileversion +%</doc|ltxgrid> +% \end{macrocode} +% +% +% \section{The driver module \texttt{doc}} +% +% This module, consisting of the present section, +% typesets the programmer's documentation, +% generating the \file{.ins} installer and \file{00readme.txt} as required. +% +% Because the only uncommented-out lines of code at the beginning of +% this file constitute the \file{doc} module itself, +% we can simply typeset the \file{.dtx} file directly, +% and there is thus rarely any need to +% generate the ``doc'' {\sc docstrip} module. +% Module delimiters are nonetheless required so that +% this code does not find its way into the other modules. +% +% The \enve{document} command concludes the typesetting run. +% +% \begin{macrocode} +%<*doc> +% \end{macrocode} +% +% \subsection{The Preamble} +% The programmers documentation is formatted +% with the \classname{ltxdoc} class with local customizations, +% and with the usual code line indexing. +% \begin{macrocode} +\documentclass{ltxdoc} +\RequirePackage{ltxdocext}% +\RequirePackage[colorlinks=true,linkcolor=blue]{hyperref}% +\ifx\package@font\@undefined\else + \expandafter\expandafter + \expandafter\RequirePackage + \expandafter\expandafter + \expandafter{% + \csname package@font\endcsname + }% +\fi +\CodelineIndex\EnableCrossrefs +% \end{macrocode} +% +% \subsubsection{Docstrip and info directives} +% We use so many {\sc docstrip} modules that we set the +% \texttt{StandardModuleDepth} counter to 1. +% \begin{macrocode} +\setcounter{StandardModuleDepth}{1} +% \end{macrocode} +% The following command retrieves the date and version information +% from this file. +% \begin{macrocode} +\expandafter\GetFileInfo\expandafter{\jobname.dtx}% +% \end{macrocode} +% +% +% \subsection{The installer file} +% +% The installer \file{ltxgrid.ins} appears here. +% If you have retrieved the standard distribution of this package, +% the installer file is already on your filesystem. +% If you are bootstrapping, +% the first typesetting of the \file{.dtx} file +% will cause the installer to be generated. +% +% The following modules are used to direct +% {\sc docstrip} in generating the external files: +% \begin{center} +% \begin{tabular}{lll} +% \textbf{Module}&\textbf{File}&\textbf{Description}\\ +% doc &\file{ltxgrid.drv}&driver for programmer's documentation\\ +% ltxgrid,ltxgrid-krn &\file{ltxgrid.sty}&this package\\ +% ltxgrid-krn& &the portion of this package suitable for inclusion within another package +% \end{tabular} +% \end{center} +% +% \begin{macrocode} +\begin{filecontents}{ltxgrid.ins} +%% This file will generate documentation and runtime files +%% from ltxgrid.dtx when run through LaTeX or TeX. +\input docstrip +\preamble + +This is a generated file; +altering it directly is inadvisable; +instead, modify the original source file. +See the URL in the file 00readme.txt. + +Copyright notice. + + These files are distributed + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. + +\endpreamble +\keepsilent + \generate{% + \file{ltxgrid.drv}{\from{ltxgrid.dtx}{doc}}% + \file{ltxgrid.sty}{% + \from{ltxgrid.dtx}{ltxgrid,ltxgrid-krn}% + }% + }% +\ifToplevel{ +\Msg{***********************************************************} +\Msg{*} +\Msg{* To finish the installation, please move} +\Msg{* ltxgrid.sty} +\Msg{* into a directory searched by TeX;} +\Msg{* in a TDS-compliant installation:} +\Msg{* texmf/tex/macros/latex/ao/.} +\Msg{*} +\Msg{* To produce the documentation, + run ltxgrid.dtx through LaTeX.} +\Msg{*} +\Msg{* Happy TeXing} +\Msg{***********************************************************} +} +\endbatchfile +\end{filecontents} +% \end{macrocode} +% Note that, because all of the files generated by the installer +% are part of the standard distribution, it will +% be necessary to run the installer only when bootstrapping +% (or, of course, during development). +% Note, too, that it is rare to generate the \file{doc} +% module because it suffices to simply typeset the \file{.dtx} file itself. +% +% \subsection{The ``Read Me'' File} +% As promised above, here is the contents of the +% ``Read Me'' file. That file serves a double purpose, +% since it also constitutes the beginining of the +% programmer's documentation. What better thing, after +% all, to have appear at the beginning of the +% typeset documentation? +% +% A good discussion of how to write a ReadMe file can be found in +% Engst, Tonya, ``Writing a ReadMe File? Read This'' +% \emph{MacTech} October 1998, p. 58. +% +% Note the appearance of the +% \cmd\StopEventually\ command, which marks the +% dividing line between the user documentation +% and the programmer documentation. +% +% The usual user will not be asked to +% do a full build, not to speak +% of the bootstrap. +% Instructions for carrying these processes +% begin the programmer's manual. +% +% \begin{macrocode} +\begin{filecontents*}{00readme.txt} +\title{% + A \LaTeX\ Package for changing the page grid and MVL% + \thanks{% + This file has version number \fileversion, + last revised \filedate.% + % For version number and date, + % search on "\fileversion" in the .dtx file, + % or see the end of the 00readme.txt file. + }% +}% + +\author{% +Arthur Ogawa (\texttt{mailto:ogawa@teleport.com}), +\fileversion\\Copyright (C) 1999, 2000 Arthur Ogawa +}% +\maketitle + +This file embodies the \classname{ltxgrid} package, +the implementation and its user documentation. + +The distribution point for this work is +\url{ftp://ftp.teleport.com/users/ogawa/macros/latex/contrib/supported/ltxgrid...}, +which contains fully unpacked, prebuilt runtime files and documentation. + +The \classname{ltxgrid} package was commissioned by the American Physical Society +and is distributed under the terms of the \LaTeX\ Project Public License, +the same license under which all the portions of \LaTeX\ itself is distributed. +Please see \url{http://ctan.tug.org/macros/latex/base/lppl.txt} for details. + +To use this document class, you must have a working +\TeX\ installation equipped with \LaTeXe\ +and possibly pdftex and Adobe Acrobat Reader or equivalent. + +To install, retrieve the distribution, +unpack it into a directory on the target computer, +into a location in your filesystem where it will be found by \LaTeX; +in a TDS-compliant installation this would be: +\file{texmf/tex/macros/latex/ao/.} + +To use, read the user documentation \file{ltxgrid.pdf}. + +\tableofcontents + +\section{Processing Instructions} + +The package file \file{ltxgrid.sty} +is generated from this file, \file{ltxgrid.dtx}, +using the {\sc docstrip} facility of \LaTeX +via |tex ltxgrid.ins|. +The typeset documentation that you are now reading is generated from +the same file by typesetting it with \LaTeX\ or pdftex +via |latex ltxgrid.dtx| or |pdflatex ltxgrid.dtx|. + +\subsection{Build Instructions} + +You may bootstrap this suite of files solely from \file{ltxgrid.dtx}. +Prepare by installing \LaTeXe\ (and either tex or pdftex) on your computer, +then carry out the following steps: +\begin{enumerate} +\item +Within an otherwise empty directory, +typeset \file{ltxgrid.dtx} with \LaTeX\ or pdflatex; +you will obtain the typeset documentation you are now reading, +along with +the installer \file{ltxgrid.ins}, +and the file \file{00readme.txt}. + +Note: you will have to run \LaTeX\ twice, then \file{makeindex}, then +\LaTeX\ again in order to obtain a valid index and table of contents. +\item +Now typeset \file{ltxgrid.ins}, +thereby generating the package file \file{ltxgrid.sty}. +\item +Install \classname{ltxgrid.sty} +by moving it to a location +in your filesystem where they will be found by \LaTeX. +\end{enumerate} +\end{filecontents*} +% \end{macrocode} +% +% \subsection{The Document Body} +% +% Here is the document body, containing only a +% \cmd\DocInput\ directive---referring to this very file. +% This very cute self-reference is a common \classname{ltxdoc} idiom. +% \begin{macrocode} +\begin{document}% +\expandafter\DocInput\expandafter{\jobname.dtx}% +% ^^A\PrintChanges +\end{document} +% \end{macrocode} +% +% \begin{macrocode} +%</doc> +% \end{macrocode} +% +% \section{Using this package} +% Once this package is installed on your filesystem, you can employ it in +% adding functionality to \LaTeX\ by invoking it in your document or document class. +% +% \subsection{Invoking the package} +% In your document, you can simply call it up in your preamble: +% \begin{verbatim} +%\documentclass{book}% +%\usepackage{ltxgrid}% +%\begin{document} +%<your document here> +%\end{document} +% \end{verbatim} +% However, the preferred way is to invoke this package from within your +% customized document class: +% \begin{verbatim} +%\NeedsTeXFormat{LaTeX2e}[1995/12/01]% +%\ProvidesClass{myclass}% +%\LoadClass{book}% +%\RequirePackage{ltxgrid}% +%<class customization commands> +%\endinput +% \end{verbatim} +% +% Note that this package requires the features of the \classname{ltxutil} package, +% available at +% \url{ftp://ftp.teleport.com/users/ogawa/macros/latex/contrib/supported/ltxutil/}. +% +% Once loaded, the package gives you acccess to certain procedures, +% usually to be invoked by a \LaTeX\ command or environment, but not at the document level. +% +% \subsection{Changing the page grid}% +% This package provides two procedures, \cmd\onecolumngrid, \cmd\twocolumngrid, +% that change the page grid (it can be extended to more columns and to other page grids). +% +% They differ from standard \LaTeX's \cmd\onecolumn\ and \cmd\twocolumn\ commands in that +% they do not force a page break. Also, upon leaving a multiple-column grid, the columns are balanced. +% In other respects they work same. +% +% They differ from the grid-changing commands of Frank Mittelbach's \classname{multicol} package +% in that they allow floats of all types (single- and double column floats, that is) and +% preserve compatability with the \classname{longtable} package. +% +% These commands must be issued in vertical mode (conceivably via a \cmd\vadjust) such that +% they are ultimately present in the MVL, where they can do their work. +% Because they do not work in \LaTeX's left-right mode, they are unsuitable at the +% document level. +% Furthermore, packaging a grid command in a \cmd\vadjust, although possible, will probably +% not acheive satisfactory page layout. +% +% Page grid commands are not intended to be issued unnecessarily: only the first of +% two successive \cmd\onecolumngrid\ commands is effective; the second will be silently ignored. +% +% \DescribeMacro\onecolumngrid +% You command \LaTeX\ to return to the one-column grid with the +% \cmd\onecolumngrid\ command. If you are already in the one-column grid, this +% is a no-op. The one-column grid is considered special of all page grids, in that +% no portion of the page is held back (in \cmd\pagesofar); all items that might go +% on the current page (with the exception of floats and footnotes) are on the MVL. +% +% \DescribeMacro\twocolumngrid +% You command \LaTeX\ to return to the two-column grid with the +% \cmd\twocolumngrid\ command. If you are already in the two-column grid, this +% is a no-op. +% +% These two commands should be issued by a macro procedure that can ensure that +% \TeX\ is in outer vertical mode. +% +% \subsection{Changing the MVL}% +% +% This package also provides commands to modify the main vertical list (MVL) in a safe way. +% The scheme here is to structure, insofar possible, \TeX's MVL as follows: +%\begin{quotation} +%box or boxes\\ +%penalty\\ +%glue +%\end{quotation} +% This should be a familiar sequence. It is the prototype sequence for a vertical list, +% and is followed when \TeX\ breaks paragraphs into lines, and when \TeX\ generates +% a display math equation. +% +% If you (as a macro programmer) wish to modify the value of the penalty or glue item, +% you can use one of the MVL-altering commands to do so. Certain operations are implemented +% here; you can make up your own. +% +% Note that these commands must be issued in vertical mode, perhaps via a \cmd\vadjust\ or a \cmd\noalign. +% They can work directly if you are in inner mode (say within a parbox or a minipage). +% +% \DescribeMacro\removestuff +% You instruct \LaTeX\ to remove both the penalty and the glue item with this command. +% +% \DescribeMacro\addstuff +% You issue the \cmd\addstuff\arg{penalty}\arg{glue} command to add a penalty, glue, or both. +% If you do not wish to add one or the other, the corresponding argument should be nil. +% Note that the effect of \cmd\addstuff\ is to stack the penalties and glue items. +% Therefore, the lesser of the two penalties takes effect, +% and the two glue items add together. +% +% \cmd\addstuff\ is limited because once applied, it cannot be applied again with correct results. +% +% \DescribeMacro\replacestuff +% The \cmd\replacestuff\ command is syntactically the same as \cmd\addstuff, but works +% differently: the existing penalty and glue are replaced or modified. +% +% The specified penalty is not inserted if the existing penalty is greater than 10000 +% (that is, in case of a \cmd\nobreak), otherwise, the lower (non-zero) of the two penalties is inserted. +% +% If the specified glue has a larger natural component than the existing glue, we replace the glue. +% However, if the specified glue's natural component is negative, then the existing glue's natural component is +% changed by that amount. +% +% \cmd\replacestuff\ can be applied mutiple times bceause it retains the list structure in the canonical form. +% +% Note that we treat two penalties specially (as does \TeX): a penalty of 10000 is considered +% a garbage value, to be replaced if found. This is the signal value that \TeX\ inserts on +% the MVL replacing the penalty that caused the page break (if the page break occurred at a penalty). +% Also, a penalty of zero is indistinguishable from no penalty at all, so it will always +% be replaced by the given value. +% +% Therefore, it is highly recommended to never set any of \TeX's penalty parameters to +% zero (a value of, say, 1, is practically the same), nor should a skip parameter be set to zero +% (instead, use, say, 1sp). Also, to prevent a pagebreak, do not use a penalty of 10000, use, say +% 10001 instead. +% +% You can define your own construct that modifies the MVL: +% Define a command, say, \cmd\myadjust, as follows: +%\begin{verbatim} +%\def\myadjust#1{\noexpand\do@main@vlist{\noexpand\@myadjust{#1}}\@tempa}% +%\end{verbatim} +% that is, \cmd\myadjust\ invokes \cmd\do@main@vlist, passing it the procedure name +% \cmd\@myadjust\ along with the arguments thereof pre-expanded. +% Next, define the procedure \cmd\@myadjust: +%\begin{verbatim} +%\def\@myadjust#1{<meddle with the MVL>}% +%\end{verbatim} +% when \cmd\@myadjust\ executes, you will be in the output routine (in inner vertical mode) +% and the MVL will be that very vertical list. +% +% +% +% +% \section{Compatability with \LaTeX's Required Packages} +% Certain packages, usually ones written by members of the +% \LaTeX\ Project itself, have been designated ``required'' and +% are distributed as part of standard \LaTeX. +% These packages have been placed in a priviledged position +% vis \'a vis the \LaTeX\ kernel in that they override the definitions of certain kernel macros. +% +% Compatability between \classname{ltxgrid} and these packages is complicated +% by a number of factors. First is that \classname{ltxgrid} alters the meaning of some of the same +% kernel macros as certain of the ``required'' packages. +% Second is that fact that certain of the ``required'' packages of \LaTeX\ are incompatible with +% each other. +% +% Examples of the first kind are the \classname{ftnright}, \classname{multicol}, and \classname{longtable} +% packages. +% The \classname{ltxgrid} package is not compatible with \classname{multicol}, +% but if you are using \classname{ltxgrid}, you do not need to use \classname{ftnright} or \classname{multicol} +% anyway. The \classname{ltxgrid} package does however attempt to be compatible with \classname{longtable}. +% +% Among the ``required'' packages that are mutually incompatible are \classname{multicol} and \classname{longtable}, +% the incompatibility arising because both packages replace \LaTeX's output routine: +% if one package is active, the other must not be so. +% This state of affairs has remained essentially unchanged since the introduction of the two as \LaTeX2.09 packages in the late 1980s. +% +% The reason that \classname{ltxgrid} can remain compatible with \classname{longtable} is due to the +% introduction of a more modern architecture, the ``output routine dispatcher'', which allows all macro packages access to the +% safe processing environment of the output routine, on an equal footing. +% The relevant portions of the \classname{longtable} package are reimplemented in \classname{ltxgrid} +% to take advantage of this mechanism. +% +% Timing is critical: +% the \classname{ltxgrid} package will be incompatible with any package that +% redefines any of the kernel macros that \classname{ltxgrid} patches---if that +% package is loaded \emph{after} \classname{ltxgrid}. +% +% Hereinafter follows some notes on specific \LaTeX\ packages. +% +% \subsection{ftnright} +% Frank Mittelbach's \classname{ftnright} package effects a change to \LaTeX's +% \cmd\twocolumn\ mode such that footnotes are set at the bottom of the right-hand +% column instead of at the foot of each of the two columns. +% +% Note that it overwrites three \LaTeX\ kernel macros: \cmd\@outputdblcol, \cmd\@startcolumn, and \cmd\@makecol. +% Fortunately none of the three are patched by \classname{ltxgrid}, so that compatability +% is not excluded on this basis. +% +% At the same time, it changes the meaning of \cmd\footnotesize, the macro that is automatically +% invoked when setting a document's footnote into type. +% One might well argue that it is an error for the meaning of \cmd\footnotesize\ to be determined by +% a package such as \classname{ftnright}, that indeed such a choice should be made in the +% document class, or in a file such as \file{bk10.clo}. +% +% To avoid being tripped up by this misfeature in \classname{ftnright}, it is only necessary to +% reassert our meaning for \cmd\footnotesize\ later on, after \classname{ftnright} has been loaded. +% +% Note that \classname{ftnright} inserts code that demands that \LaTeX's flag \cmd\if@twocolumn\ +% is true, that is, it will complain if deployed in a \cmd\onecolumn\ document. +% It is therefore necessary for any other multicolumn package to assert that flag in order to +% avoid this package's complaint. It is an interesting question exactly why this package has +% this limitation. After all, a one-column page grid is just a degenerate case of the +% two column. +% +% \subsection{longtable} +% David Carlisle's \classname{longtable} package sets tables that can be so long as to break over pages. +% According to its author, it uses the same override of \LaTeX's output routine as +% Frank Mittelbach's \classname{multicol} package. By implication, then, it has a hard +% incompatability with the latter. +% +% The \classname{longtable} package also performs a check of whether the document is in +% \cmd\twocolumn\ mode, and declines to work if this is the case. It is not clear, however, +% that there is any true incompatability present if so. It's just that David did not see any reason +% anyone would want to set such long tables in a multicolumn document, hence the check. +% +% There does not appear to be any indication that \classname{longtable} would work less +% well under \classname{ltxgrid} than under standard \LaTeX's \cmd\twocolumn\ mode. +% Therefore, this \classname{ltxgrid} patches \classname{longtable} (if loaded) so as to provide +% compatability. In the course of which, \classname{longtable} becomes more robust +% (\classname{longtable} has mumerous bugs and incompatabilities of long standing, +% some of which are repaired by \classname{ltxgrid}). +% +% One problem remains, namely that, if a \env{longtable} environment breaks over columns +% and thereby inserts its special headers and footers at that break, and those columms are then +% balanced (due to a return to the one-column page grid), then those inserted rows +% will remain, and may no longer fall at the column break. This will, of course look +% wrong. +% +% The only way to fix this problem is to avoid doing column balancing in the way +% I have implemented here; such an enhancement to this package is possible. +% +% \subsection{multicol} +% Frank Mittelbach's \classname{multicol} package provides a page grid with many columns, +% albeit denies the placement of floats in individual columns. +% +% It esablishes its own \cmd\output\ routine, which is the reason it runs afoul of +% the \classname{longtable} package. On the other hand, \classname{ltxgrid} specifically +% allows for the case where a package installs its own \cmd\output\ routine, so +% there is no incompatability on that basis. +% +% Still, it is pointless to use \classname{multicol} if you are using \classname{ltxgrid}, +% since both packages provide multicolumn page layouts. +% Therefore, \classname{multicol} is not supported by \classname{ltxgrid}. +% +% \subsection{ltxgrid} +% It has been pointed out that one of the disadvantages of adopting the \classname{ltxgrid} package is that +% it does alter the \LaTeX\ kernel. +% Any package that itself alters the \LaTeX\ kernel may be incompatible with \classname{ltxgrid}, and new packages +% (destined perhaps to become part of the successor to \LaTeXe) may break \classname{ltxgrid}. +% +% The consequence is that packages introduced in future, and future changes to \LaTeX\ may be incompatible +% with \classname{ltxgrid}. +% This is, of course, true. +% The development plan for \classname{ltxgrid} is that when such packages and \LaTeX\ kernel changes come about, +% the burden will be on \classname{ltxgrid} to change in a way that provides for continued compatability with +% those packages and \LaTeX\ kernel changes. +% +% +% \section{How \classname{ltxgrid} places footnotes} +% +% In conventional multicolumn layouts, a footnote will appear at the bottom of the column in which it is called out. +% The \classname{ltxgrid} package implements this conventional layout choice by default. +% However, other choices are possible (a la \classname{ftnright}, whose compatability with \classname{ltxgrid} has not been tested). +% +% One unusual feature of \classname{ltxgrid}'s default implementation must be mentioned, though, +% namely the case in a two-column page grid, where a footnote is followed by a temporary change to the one-column page grid +% (e.g., for a wide equation). +% In such a case, the material above the wide material is split into two columns, and a footnote whose callout +% appears in the right-hand column will nonetheless be set at the base of the left column. +% +% This arrangement was chosen because it ensures that the footnotes at the bottom of any page will appear in +% numerical order. It can be argued that this choice is ``incorrect'', but be that as it may, +% the \classname{ltxgrid} package does not foreclose on other arrangements for the footnotes. +% The package can be adapted to accomodate any page design desired. +% +% \section{Limitations in \classname{ltxgrid}'s default column balancing method}% +% +% In a multicolumn page grid, when encountering a page that is not completely full, +% it is customary to set the material in balanced columns (typically with the last column no longer than any of the others). +% Such a case also crops up when temporarily interrupting the multicolumn grid to set material on the full width of the page: +% the material on the page above the break is customarily set in balanced columns. +% +% An awkward case arises when we have already set one or more complete columns of type before encountering the need to +% balance columns. In this subset of cases, the default in \classname{ltxgrid} is to +% do an operation I call ``re-balancing'': +% the material on the page so far is pasted back together into a single column, and new, balanced column breaks are +% calculated. +% +% This scheme typically works fine, but it has a significant vulnerability: +% any discardable items trimmed at the original column break is lost, never to be retrieved. +% Consequently, after re-balancing, an element like, say, a section head can fail to have the correct amount of whitespace above. +% +% This problem is due to an unfortunate optimization in \TeX, wherein a certain class of nodes is trimmed from the +% top of main vertical list upon returning from the output routine: +% any penalty, glue, or leader node falls in to this class of discardable nodes, +% and trimming proceeds until a non-discardable node (such as a box, or rule) is encountered. +% It gets better: a third class of nodes is transparent to this trimming process; +% they are neither discarded nor do they halt the process of trimming: +% mark nodes and all whatsits fall into this class of transparent nodes; +% they are quietly passed over during trimming. +% +% An alternative approach for \TeX\ to take would have been, +% rather than discarding the node entirely, to simply mark it as discarded. +% (Implementors of NTS, please note!) +% Then, upon shipping out, such nodes would not make it into the DVI. +% \TeX's optimization, driven by the small computer architectures current when it was developed, +% does save mem, but at the cost of revisiting page breaks in a reliable way. +% +% FIXME: how to fix a column break in the above case? Widetext? +% +%\StopEventually{} +% +% \section{Implementation of package} +% +% Special acknowledgment: this package uses concepts pioneered +% and first realized by William Baxter (mailto:web@superscript.com) +% in his SuperScript line of commercial typesetting tools, and +% which are used here with his permission. His thorough understanding +% of \TeX's output routine underpins the entire \classname{ltxgrid} +% package. +% +% \subsection{Beginning of the \file{ltxgrid} {\sc docstrip} module} +% Requires the underpinnings of the \classname{ltxkrnext} package. +% \begin{macrocode} +%<*ltxgrid> +\def\package@name{ltxgrid}% +\expandafter\PackageInfo\expandafter{\package@name}{% + Page grid for \protect\LaTeXe, + by A. Ogawa (ogawa@teleport.com)% +}% +\RequirePackage{ltxutil}% +%</ltxgrid> +% \end{macrocode} +% +% \subsection{Banner}% +% Credit where due. +% \begin{macrocode} +%<*ltxgrid-krn> +\typeout{% + ltxgrid: portions licensed from W. E. Baxter (web@superscript.com)% +}% +% \end{macrocode} +% +% \subsection{Sundry}% +% Here are assorted macro definitions. +% \begin{macro}{\lineloop} +% The document-level command \cmd\lineloop\ sets numbered lines until the +% specified count is reached. +% This command is mainly used to construct test documents. +% \begin{macrocode} +\newcounter{linecount} +\def\lineloop#1{% + \loop + \ifnum\c@linecount<#1\relax + \global\advance\c@linecount\@ne + \par + \hb@xt@\hsize{% + \ifnum\c@linecount<100 0\fi\ifnum\c@linecount<10 0\fi\number\c@linecount + \vrule depth2.5\p@ + \leaders\hrule\hfil + }% + \penalty\interlinepenalty + \repeat +}% +% \end{macrocode} +% \end{macro} +% +% +% \subsection{Mark Components}% +% +% Override LaTeX's mark macros to allow more components. +% +% We remain bound by the weakness of LaTeX's scheme in that +% one cannot emulate the action of \TeX\ whereby +% material with marks can be inserted in the middle of +% a vertical list such that the marks are reliably calculated. +% If we did that, \cmd\@themark\ would no longer be utilized. +% +% A more robust scheme involves placing all marks (component and value) +% into a list (using global scoping, i.e., \cmd\gdef), +% and using \cmd\@@mark to place an index on that list into the MVL. +% Then, e.g., \cmd\@@botmark signifies the place where that list is to be cut, +% and the \cmd\botmark\ of any component is +% the value of the last element of the cut +% list having the given component. The \cmd\firstmark\ and \cmd\topmark\ +% can likewise be defined relative to \cmd\@@firstmark\ and \cmd\@@topmark, +% except in the latter case, we want the first following the cut instead of the last +% preceding the cut. +% +% The limitation of this scheme is its demands upon \TeX's mem. +% The list of marks would need to be trimmed back to, effectively, +% \cmd\topmark\ at the beginning of every page. +% +% This approach is not yet part of the extended LaTeX kernel. +% +% \begin{macro}{\@@mark} +% \begin{macro}{\@@topmark} +% \begin{macro}{\@@firstmark} +% \begin{macro}{\@@botmark} +% \begin{macro}{\@@splitfirstmark} +% \begin{macro}{\@@splitbotmark} +% Remember primitives under a new set of names. +% \begin{macrocode} +\let\@@mark\mark +\let\@@topmark\topmark +\let\@@firstmark\firstmark +\let\@@botmark\botmark +\let\@@splitfirstmark\splitfirstmark +\let\@@splitbotmark\splitbotmark +% \end{macrocode} +% \end{macro} +% \end{macro} +% \end{macro} +% \end{macro} +% \end{macro} +% \end{macro} +% +% \subsubsection{Procedures that expose the component data structure}% +% This portion of the code exposes the internal representation of +% the mark components. If we wish to add more components, we will have to revise +% these macro definitions: +% \cmd\@themark, +% \cmd\nul@mark, +% \cmd\set@mark@netw@, +% \cmd\set@marktw@, +% \cmd\set@markthr@@, +% \cmd\get@mark@@ne, +% \cmd\get@mark@tw@, +% \cmd\get@mark@thr@@, +% \cmd\get@mark@f@ur. +% +% \begin{macro}{\@themark} +%FIXME: is it safer to eliminate \cmd\@themark\ in favor of a message that evaluates \cmd\@@botmark? +% +% Note: these definitions expose the data structure of mark components. +% \begin{macrocode} +\def\@themark{{}{}{}{}}% +\def\nul@mark{{}{}{}{}\@@nul}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\set@mark@netw@} +% \begin{macro}{\set@marktw@} +% \begin{macro}{\set@markthr@@} +% These procedures insert the new value of a particular mark component into the given argument. +% They expose the data structure of mark components. +% +% \begin{macrocode} +\def\set@mark@netw@#1#2#3#4#5#6#7{\gdef#1{{#6}{#7}{#4}{#5}}\do@mark}% +\def\set@marktw@#1#2#3#4#5#6{\gdef#1{{#2}{#6}{#4}{#5}}\do@mark}% +\def\set@markthr@@#1#2#3#4#5#6{\gdef#1{{#2}{#3}{#6}{#5}}\do@mark}% +% \end{macrocode} +% \end{macro} +% \end{macro} +% \end{macro} +% +% \begin{macro}{\get@mark@@ne} +% \begin{macro}{\get@mark@tw@} +% \begin{macro}{\get@mark@thr@@} +% \begin{macro}{\get@mark@f@ur} +% These procedures retreive the value of a particular mark component. +% They expose the data structure of mark components. +% \begin{macrocode} +\def\get@mark@@ne#1#2#3#4#5\@@nul{#1}% +\def\get@mark@tw@#1#2#3#4#5\@@nul{#2}% +\def\get@mark@thr@@#1#2#3#4#5\@@nul{#3}% +\def\get@mark@f@ur#1#2#3#4#5\@@nul{#4}% +% \end{macrocode} +% \end{macro} +% \end{macro} +% \end{macro} +% \end{macro} +% +% +% \subsubsection{Procedures that do not expose the component data structure}% +% +% \begin{macro}{\mark@netw@} +% \begin{macro}{\marktw@} +% \begin{macro}{\markthr@@} +% These procedures insert the new value of a particular mark component into \cmd\@themark, +% then execute \cmd\do@mark. +% They constitute the implementation layer for mark components one, two, and three. +% An analogous procedure for component four could be defined; call it \cmd\markf@ur. +% +% \begin{macrocode} +\def\mark@netw@{\expandafter\set@mark@netw@\expandafter\@themark\@themark}% +\def\marktw@{\expandafter\set@marktw@\expandafter\@themark\@themark}% +\def\markthr@@{\expandafter\set@markthr@@\expandafter\@themark\@themark}% +% \end{macrocode} +% \end{macro} +% \end{macro} +% \end{macro} +% +% \begin{macro}{\do@mark} +% \begin{macro}{\do@@mark} +% Access procedures \cmd\mark (AKA \cmd\@@mark). +% The \cmd\do@mark\ procedure is used when a mark is being put down into the MVL; +% \cmd\do@@mark\ when this happens in the output routine. +% \begin{macrocode} +\def\do@mark{\do@@mark\@themark\nobreak@mark}% +\def\do@@mark#1{% + \begingroup + \let@mark + \@@mark{#1}% + \endgroup +}% +% \end{macrocode} +% \end{macro} +% \end{macro} +% +% \begin{macro}{\let@mark} +% \begin{macro}{\nobreak@mark}% +% The procedure that makes \cmd\csname s robust within a mark. +% Use \cmd\appdef\ and \cmd\robust@\ to extend the list. +% \begin{macrocode} +\def\let@mark{% + \let\protect\@unexpandable@protect + \let\label\relax + \let\index\relax + \let\glossary\relax +}% +\def\nobreak@mark{% + \@if@sw\if@nobreak\fi{\@ifvmode{\nobreak}{}}{}% +}% +% \end{macrocode} +% \end{macro} +% \end{macro} +% +% +% \subsubsection{Using mark components}% +% +% These procedures use the component mark mechanism to implement +% a mark component that remembers the current environment (used in page makeup) +% and the the two mark components left over from the original \LaTeX. +% The fourth component is presently unused. +% +% \begin{macro}{\mark@envir} +% The third mark component's access procedures. +% The \cmd\mark@envir\ and \cmd\bot@envir\ commands are a good model of how to write +% access procedures for a new mark component. +% \begin{macrocode} +\def\mark@envir{\markthr@@}% +\def\bot@envir{% + \expandafter\expandafter + \expandafter\get@mark@thr@@ + \expandafter\@@botmark + \nul@mark +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\markboth} +% \begin{macro}{\markright} +% \begin{macro}{\leftmark} +% \begin{macro}{\rightmark} +% Set procedures for legacy components. +% \begin{macrocode} +\def\markboth{\mark@netw@}% +\def\markright{\marktw@}% +% \end{macrocode} +% +% Retrieval procedures for legacy mark components. +% The procedure for retrieving the first component from \cmd\botmark\ +% and the second component from \cmd\firstmark have names in \LaTeX; +% they are called, respectively, \cmd\leftmark\ and \cmd\rightmark. +% +% It is possible to retrieve the components of \cmd\topmark\ +% as well: use \cmd\saved@@topmark. +% \begin{macrocode} +\def\leftmark{% + \expandafter\expandafter + \expandafter\get@mark@@ne + \expandafter\saved@@botmark + \nul@mark +}% +\def\rightmark{% + \expandafter\expandafter + \expandafter\get@mark@tw@ + \expandafter\saved@@firstmark + \nul@mark +}% +% \end{macrocode} +% \end{macro} +% \end{macro} +% \end{macro} +% \end{macro} +% +% +% +% \subsection{Output Super-routine}% +% +% We want to change \LaTeX's output routine, but do not wish to remain vulnerable +% to interference from such ``required'' packages as +% \classname{multicol} (authored by Frank Mittelbach) +% and \classname{longtable} (authored by David P. Carlisle), which +% swap in their own output routines when the respective package is active. +% +% The better mechanism, used here, is due to William Baxter (web@superscript.com), +% who has allowed his several ideas to be used in this package. +% +% In what follows, we effectively wrap up the old \LaTeX\ output routine inside +% a new, more flexible ``super routine''. When the output routine is called, +% the ``super routine'' acts as a dispatcher. If the old routine is needed, it is called. +% +% If a package attempts to substitute in their own output routine, they will effectively +% be modifying a token register by the name of \cmd\output. +% The primitive \cmd\output\ is now known by a different name, which should no longer be +% necessary to use. +% +% Usage note: to make a visit to the output routine employing the dispatcher, enter +% with a value of \cmd\outputpenalty\ that corresponds to a macro. Defining as follows: +%\begin{verbatim} +%\@namedef{output@10000}{<your code here>}% +%\end{verbatim} +% by convention, your output routine should void out \cmd\box\cmd\@cclv. +% +% In rewriting \LaTeX's output dispatcher +% in a much simpler form, we also avoid the sin of multiple \cmd\shipout s +% within a single visit to the output routine. +% +% Conceptually, we divide visits to the output routine into two classes. +% The first involves natural page breaks +% (at a \cmd\newpage\ or when \cmd\pagetotal $>$ \cmd\pagegoal) +% and usually resulting in \cmd\box\cmd\@cclv\ either being shipped out or +% salted away (e.g., each column in a multicolumn layout). +% We might call this class the ``natural output routines''; the \cmd\outputpenalty\ +% will never be less than $-10000$. +% Furthermore, we ensure that \cmd\holdinginserts\ is cleared when +% calling such routines. +% +% The other class involves a forced visit to the output routine +% via a large negative penalty ($< -10000$). They do not generally +% result in a \cmd\shipout\ of \cmd\box\cmd\@cclv: they may be dead cycles. +% We provide a mechanism (call it a ``one-off'' output routine) that allows +% us to specify certain processing to be done when \TeX\ reaches +% the current position on the page. +% +% One-off output routines themselves fall into two divisions, ones +% that process \cmd\box\cmd\@cclv, and ones that work on the main vertical list (MVL). +% The former are typified by changes to the page grid, perhaps +% even column balancing. +% The latter involve the insertion of penalties or glue and the processing of floats. +% +% The natural output routine is a single procedure. We have not introduced multiple +% natural output routines based on the \cmd\outputpenalty\ because \TeX\ does not +% support such a thing: \TeX\ sometimes lays down a penalty whose value is the sum +% of other penalties. Because of this, we cannot depend on the value of \cmd\outputpenalty\ +% in such areas. +% +% We do introduce flexibility in the form of a mechanism for patching into the +% natural output routine. Three hooks are offered, allowing a procedure +% to prepare for the upcoming visit to the output routine, +% access to \cmd\box\cmd\@cclv, and after shipping out (or otherwise +% committing the material to the page). +% +% Environments, commands, and even packages can install their +% own procedures into these hooks. +% For instance, if the longtable package is loaded, it will install +% its procedures, but those procedures will punt if the page break +% being processed does not actually fall within a longtable environment. +% +% \begin{macro}{\primitive@output} +% Here we remember the \TeX\ primitive \cmd\output\ and its value, +% and then proceed to take over the \cmd\csname\ of \cmd\output, +% making it a \cmd\toks\ register and installing the old value of +% the output routine. +% \begin{macrocode} +\let\primitive@output\output +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\output} +% Grab the tokens in \cmd\the\cmd\output\ (but without the extra set of braces). +% The value of \cmd\toks@\ must remain untouched until loaded into the appropriate token +% register; this is done a few lines below. +% \begin{macrocode} +\long\def\@tempa#1\@@nil{#1}% +\toks@ +\expandafter\expandafter +\expandafter{% +\expandafter \@tempa + \the\output + \@@nil + }% +\newtoks\output +\output\expandafter{\the\toks@}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\dispatch@output} +% We now install our own output routine in place of the +% old one, which is still available as \cmd\the\cmd\output. +% +% The output routine is simply the procedure \cmd\dispatch@output. +% It either dispatches to a procedure based on a particular value of +% \cmd\outputpenalty\ or it executes \cmd\the\cmd\output\ tokens. +% \begin{macrocode} +\primitive@output{\dispatch@output}% +\def\dispatch@output{% + \let\par\@@par + \expandafter\let\expandafter\@tempa\csname output@\the\outputpenalty\endcsname + \outputdebug@sw{% + \saythe\badness + \saythe\outputpenalty + \saythe\holdinginserts + \say\thepagegrid + \saythe\pagegrid@col + \saythe\pagegrid@cur + %\say\bot@envir + \saythe\insertpenalties + %\say\@@topmark + %\say\saved@@topmark + %\say\@@firstmark + %\say\saved@@firstmark + \say\@@botmark + %\say\saved@@botmark + \saythe\pagegoal + \saythe\pagetotal + \saythe{\badness\@cclv}% + \expandafter\@ifx\expandafter{\csname output@-\the\execute@message@pen\endcsname\@tempa}{% + \say\@message@saved + }{% + \expandafter\say\csname output@\the\outputpenalty\endcsname + }% + \say\@toplist + \say\@botlist + \say\@dbltoplist + \say\@deferlist + {\tracingall\scrollmode + \showbox\@cclv + \showbox\@cclv@saved + \showbox\pagesofar + \showbox\footbox + \showbox\footins@saved + \showbox\footins + \showlists + }% + }{}% + \@ifnotrelax\@tempa{\@tempa}{\the\output}% +}% +\@ifxundefined{\outputdebug@sw}{% + \@booleanfalse\outputdebug@sw +}{}% +% \end{macrocode} +% \end{macro} +% +% +% \subsection{Further thoughts about inserts} +% +% The only safe way to deal with inserts is to either set \cmd\holdininserts\ or +% to commit to using whatever insert comes your way: you cannot change your mind +% once you see a non-void \cmd\box\cmd\footins, say. +% +% Therefore all output routine processing must proceed with \cmd\holdinginserts\ set +% until you are sure of the material to be committed to the page. At that point, you +% can clear \cmd\holdinginserts, spew \cmd\box\cmd\@cclv, put down the appropriate penalty, +% and exit, with the knowledge that \TeX\ will re-find the same pagebreak, this time +% visiting the output routine with everything, including inserts, in their proper +% place. +% This technique applies to split elements (screens, longtable, index) as well as to +% manufactured pages (float pages and clearpage pages). +% +% Therefore, the output routine must not make assumptions about whether \cmd\holdinginserts\ +% should be cleared; instead this must be left to the one-off output routines or the natural output routine. +% +% If we are manufacturing pages (``float page processing''), and if \cmd\pagegoal\ is not equal to +% \cmd\vsize, then inserts are at hand, and our criterion should take into account the insert +% material, even though we cannot measure its height based on the size of \cmd\box\cmd\footins\ +% (because \cmd\holdinginserts\ is set, you see). +% +% It would be better to take the complement of \cmd\floatpagefraction\ and use that +% as a standard for the looseness of the page. Since \cmd\pagegoal\ reflects the inserted material, +% the criterion becomes the difference of the aggregate height of the floats and the \cmd\pagegoal\ +% versus this "page looseness" standard. +% +% As a check, consider what happens if we bail out: \cmd\@deferlist\ has never been touched, so it +% requires no attention. Also, \cmd\holdinginserts\ has never been cleared, so inserts require +% no attention. So we only have to ensure that marks are preserved, which is already taken +% care of by the message handler mechanism. +% +% If we are doing ordinary page cutting, then the scheme would be to detect whether we are within +% a screen (or longtable as may be), do the adjustment to the page height, and return, but this time +% with \cmd\holdinginserts\ cleared. Upon reentering the output routine, we may or may not be within +% the screen environment, but we are now sure to have a final page break, and we can commit this +% material (by shipping out or by saving it out as a full column). +% +% In the above, the first of the two visits to the output routine is a dead cycle and requires +% propagation of marks, but nothing else. +% +% The natural output routine +% +% Here is the portion of the output routine that fields cases not handled by +% the dispatcher. +% +% The default is to ship out a page and then look around for more material +% that might constitute a ``float page''. However, because \cmd\holdinginserts\ +% is normally set, this output routine must first have a dead cycle and +% come back again with \cmd\holdinginserts\ cleared. +% Then, after shipping out, it puts down a message that +% will manufacture zero or more float pages, finally terminating +% with a procedure that commits floats to a new unfinished page. +% +% To accomodate special processing, we execute hooks whose name is based +% on the value of the "envir" mark component. The default is "document", +% ensured by an initial mark of that value; the associated procedures +% are all nil. Any unknown envir value will "\cmd\relax\ out". +% +% The code \cmd\move@insert@sw\ tells whether we are on our first visit to +% the output routine (with \cmd\holdinginserts\ still set), or our second +% (with \cmd\holdinginserts\ cleared). The output routine will toggle the +% setting. +% +% The commands \cmd\hold@insertions\ and \cmd\move@insertions\ respectively +% clear and set the state of \cmd\move@insert@sw, so this procedure effectively +% clears \cmd\holdinginserts\ just long enough to pick up the insertions. +% Important: any output routine that clears \cmd\holdinginserts\ +% must guarentee that it is restored on the subsequent visit to the output routine. +% Or, to put it another way, if an output routine detects that \cmd\holdinginserts\ +% is cleared, it should take it upon itself to restore it before exiting. +% +% The branch with \cmd\holdinginserts\ set is executed first; the other +% branch follows on practically immediately thereafter. In the first branch, +% we simply execute the appropriate hook and then execute a dead cycle. +% +% In the branch with \cmd\holdinginserts\ cleared, the procedure +% builds up the current column, which is now complete, with \cmd\@makecol, then +% dispatches to the shipout routine associated with the current page grid, \cmd\output@column@. +% At the end, it triggers the execution of an output routine to prepare the next column (or page). +% +% \subsection{Natural output routine}% +% +% \begin{macro}{\output} +% Here is what has become of the output routine of \LaTeX. +% It is of necessity divided into phases, \cmd\output@holding\ is executed upon first encountering the natural page-breaking point, while inserts are being held. +% The second phase, \cmd\output@moving, is set in motion by the first: here the same material (in most cases) will be processed with \cmd\holdinginserts\ cleared. +% +% \begin{macrocode} +\output={\toggle@insert\output@holding\output@moving}% +% \end{macrocode} +% +% The procedure \cmd\output@holding\ +% is our first cycle through the output routine; \cmd\holdinginserts\ is still set. +% We give the current environment a heads up +% (it is through this means that \classname{longtable} sets its running header and footer), +% then we execute a dead cycle, which should propagate marks. +% +% One corner case that can crop up is the presence of a single unbreakable chunk whose size is larger +% than \cmd\vsize. +% Doing a dead cycle under such circumstances will not find the same breakpoint as this time +% (remember we threw in a \cmd\mark\ node). +% Instead, we attempt to remove the excess height of the material, so we can continue to propagate marks. +% +% The corner case is at hand if the natural size of \cmd\box\cmd\@cclv\ exceeds \cmd\pagegoal\ and +% the contents cannot be shrunk to fit. +% +% \begin{macrocode} +\def\output@holding{% + \csname output@init@\bot@envir\endcsname +%\vbadness\@M +%\vfuzz\maxdimen + \@if@exceed@pagegoal{\unvcopy\@cclv}{% + \setbox\z@\vbox{\unvcopy\@cclv}% + \outputdebug@sw{{\tracingall\scrollmode\showbox\z@}}{}% + \dimen@\ht\@cclv\advance\dimen@-\ht\z@ + \dead@cycle@repair\dimen@ + }{% + \dead@cycle + }% +}% +\def\@if@exceed@pagegoal#1{% + \begingroup + \setbox\z@\vbox{#1}% + \dimen@\ht\z@\advance\dimen@\dp\z@ + \outputdebug@sw{\saythe\dimen@}{}% + \@ifdim{\dimen@>\pagegoal}{% + \setbox\z@\vbox{\@@mark{}\unvbox\z@}% + \splittopskip\topskip + \splitmaxdepth\maxdepth + \vbadness\@M + \vfuzz\maxdimen + \setbox\tw@\vsplit\z@ to\pagegoal + \outputdebug@sw{{\tracingall\scrollmode\showbox\tw@\showbox\z@}}{}% + \setbox\tw@\vbox{\unvbox\tw@}% + \@ifdim{\ht\tw@=\z@}{% + \ltxgrid@info{Found overly large chunk while preparing to move insertions. Attempting repairs}% + \aftergroup\true@sw + }{% + \aftergroup\false@sw + }% + }{% + \aftergroup\false@sw + }% + \endgroup +}% +% \end{macrocode} +% +% The procedure \cmd\output@moving\ +% is our second cycle through the output routine; \cmd\holdinginserts\ is now cleared, +% and \cmd\insert s will have been split off into their respective box registers, like \cmd\footins. +% +% \begin{enumerate} +% \item +% Set the values of \cmd\topmark\ and \cmd\firstmark. +% \item +% If we got here because of a \cmd\clearpage\ command, remove the protection box that this mechanism has left on the MVL. +% \item +% If the contents of \cmd\box\cmd\@cclv\ are non-trivial, commit it to the current page or ship it out as the case may call for. +% \item +% If not, discard it (we are at the end of \cmd\clearpage\ processing). +% \item +% Set various values, including the available space for setting type on the next column (\cmd\@colroom). +% \end{enumerate} +% +% The processing for a non-trivial \cmd\box\cmd\@cclv\ are: +% \begin{enumerate} +% \item +% Execute the head procedure for the current environment. +% \item +% Make up a column and ship it out (or commit it to the current page) via a procedure keyed to the current page grid. +% \item +% Put down an interrupt for \cmd\do@startcolumn@pen: this will force a visit to the output routine for the +% purpose of committing floats to the next column. +% \item +% Possibly put down an interrupt to continue \cmd\clearpage\ proccessing. +% \item +% Execute the tail procedure for the current environment. +% \end{enumerate} +% +% +% The processing for a trivial \cmd\box\cmd\@cclv\ are: +% \begin{enumerate} +% \item +% Void out \cmd\box\cmd\@cclv\ and give appropriate warning messages and diagnostics. +%% \item +%% Put down the same interrupts as for the non-trivial case above. +% \end{enumerate} +% +% \begin{macrocode} +\def\output@moving{% + \set@top@firstmark + \@ifnum{\outputpenalty=\do@newpage@pen}{% + \setbox\@cclv\vbox{% + \unvbox\@cclv + \setbox\z@\lastbox + \@ifdim{\ht\z@=\ht\@protection@box}{\box\lastbox}{\unskip}% + }% + }{}% + \@cclv@nontrivial@sw{% + \csname output@prep@\bot@envir \endcsname + \@makecol\csname output@column@\thepagegrid\endcsname + \protect@penalty\do@startcolumn@pen + \clearpage@sw{% + \protect@penalty\do@endpage@pen + }{}% + \csname output@post@\bot@envir \endcsname + }{% + {\setbox\z@\box\@cclv}% + }% + \set@colroom + \global\@mparbottom\z@ + \global\@textfloatsheight\z@ %FIXME: this legacy LaTeX variable is set, but never queried! +}% +% \end{macrocode} +% +% The procedure \cmd\@cclv@nontrivial@sw\ determines if this visit to \cmd\output@moving\ +% is a trivial one, which happens at the end of \cmd\clearpage\ processing and under some pathological circumstances. +% It emits a Boolean, so it is syntactically like \cmd\true@sw, albeit does not execute solely via expansion. +% +% Note: the case where \cmd\box\cmd\@cclv\ is void comes up at the very beginning of the job, when +% typesetting a (full-page-width) title block in a two-column layout. +% +% Note: the code that removes the last box and skip from the output is intended to detect the case +% where the output has whatit nodes followed by topskip and a protection box. +% This is what happens under normal circumstances at the end of \cmd\clearpage\ processing. +% \begin{macrocode} +\def\@cclv@nontrivial@sw{% + \@ifx@empty\@toplist{% + \@ifx@empty\@botlist{% + \@ifvoid\footins{% + \@ifvoid\@cclv{% + \false@sw + }{% + \setbox\z@\vbox{\unvcopy\@cclv}% + \@ifdim{\ht\z@=\topskip}{% + \setbox\z@\vbox{% + \unvbox\z@ + \setbox\z@\lastbox\dimen@\lastskip\unskip + \@ifdim{\ht\z@=\ht\@protection@box}{% + \advance\dimen@\ht\z@ + \@ifdim{\dimen@=\topskip}{% + \aftergroup\true@sw + }{% + \aftergroup\false@sw + }% + }{% + \aftergroup\false@sw + }% + }% + {% + \false@sw % Normal for \clearpage + }{% + \true@sw + }% + }{% + \@ifdim{\ht\z@=\z@}{% + \ltxgrid@info{Found trivial column. Discarding it}% + \outputdebug@sw{{\tracingall\scrollmode\showbox\@cclv}}{}% + \false@sw + }{% + \true@sw + }% + }% + }% + }{% + \true@sw + }% + }{% + \true@sw + }% + }{% + \true@sw + }% +}% +% \end{macrocode} +% +% +% \end{macro} +% +% \begin{macro}{\protect@penalty} +% The procedure \cmd\protect@penalty\ is the utility procedure for invoking a +% one-off output routine. Such a routine can expect to find the protection box +% above it in \cmd\box\cmd\@cclv: it should remove that box. +% +% Note that \cmd\execute@message\ does the same thing as \cmd\protect@penalty, but +% in a slightly different way. +% +% We create a specially formulated box that will be universally used when a protection box is needed. +% In this way, we can always recognize when \cmd\box\cmd\@cclv\ is trivial: +% it will consist of whatsits followed by \cmd\topskip\ glue and the \cmd\@protection@box. +% \begin{macrocode} +\def\protect@penalty#1{\protection@box\penalty-#1\relax}% +\newbox\@protection@box +\setbox\@protection@box\vbox to1986sp{\vfil}% +\def\protection@box{\nointerlineskip\copy\@protection@box}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\dead@cycle} +% \begin{macro}{\dead@cycle@repair} +% The procedure \cmd\dead@cycle\ is defined separately as a utility which can be used by +% any output processing routine to emulate what takes place in the standard output routine. +% +% Here, we have entered the output routine with \cmd\holdinginserts\ enabled, which means that we +% are not yet ready to ship out material, because the \cmd\insert\ registers are being held. +% We want to clear \cmd\holdinginserts\ and come back here with the same page break as before, whereupon +% we may properly proceed with page makeup. +% +% To do this, we +% propagate marks, then spew the contents of \cmd\box\cmd\@cclv\ followed by the +% original output penalty that landed us here (but only if it is not 10000, +% the flag value for a pagebreak not at a penalty). +% +% However, the natural output routine should do this only if \cmd\box\cmd\@cclv\ is nontrivial. +% A pathological case exists wherein a box of height greater than \cmd\textheight\ would cause an infinite loop involving the output routine. +% The procedure \cmd\dead@cycle@repair, attempts to catch this case and avoid the loop. +% +% The test of the height of \cmd\box\cmd\@cclv\ is not the correct one, because this test will run afoul in +% the case where \cmd\box\cmd\@cclv\ contains nothing but an \cmd\insert\ node. What to do? +% +% It is possible that the pathological case can be detected by looking at \cmd\pagetotal. If that quantity is +% zero, then \cmd\box\cmd\@cclv\ really is trivial. +% +% In the procedure \cmd\dead@cycle@repair, if \cmd\box\cmd\@cclv\ is nontrivial, we execute \cmd\dead@cycle, +% otherwise it contains nothing but a mark, so we dispense with propagating marks +% and we simply spew out \cmd\box\cmd\@cclv\ without an accompanying mark. +% This has the effect of failing to propagate marks, but this problem is preferrable to the infinite loop, +% which in principle could crash even a robust operating system by filling up the file system. +% +% If a document has such a large chunk, it should be fixed, so we give a message in the log. +% +% You ask, ``In what way does this infinite loop come about?'' Good question! +% +% The setup is a chunk in the MVL that is taller than \cmd\textheight. +% (Yes, it's that simple.) +% As soon as the previous page ships out, the MVL will contain a mark (propagated from the previous page) followed +% by that large chunk (call it the `big bad box', albeit does not need to be a single box). +% The next visit to the output routine will be a natural page break, but +% \TeX\ will select the juncture between the mark and the big bad box as the least-cost page break. +% Unless the test in \cmd\dead@cycle\ is done, the cycle is perpetuated when the macro +% reinserts the mark. +% +% The crux matter is achieving, in a robust way, the goal of going from a \cmd\holdinginserts\ state to one +% where the insertions are moving. +% +% \begin{macrocode} +\def\dead@cycle@repair#1{% + \expandafter\do@@mark + \expandafter{% + \@@botmark + }% + \unvbox\@cclv + \nointerlineskip + \vbox to#1{\vss}% + \@ifnum{\outputpenalty<\@M}{\penalty\outputpenalty}{}% +}% +\def\dead@cycle@repair@protected#1{% + \expandafter\do@@mark + \expandafter{% + \@@botmark + }% + \begingroup + \unvbox\@cclv + \setbox\z@\lastbox % Remove protection box + \nointerlineskip + \advance#1-\ht\@protection@box + \vbox to#1{\vss}% + \protection@box % Reinsert protection box + \@ifnum{\outputpenalty<\@M}{\penalty\outputpenalty}{}% + \endgroup +}% +\def\dead@cycle{% + \expandafter\do@@mark + \expandafter{% + \@@botmark + }% + \unvbox\@cclv + \@ifnum{\outputpenalty<\@M}{\penalty\outputpenalty}{}% +}% +% \end{macrocode} +% \end{macro} +% \end{macro} +% +% \begin{macro}{\output@init@document} +% \begin{macro}{\output@prep@document} +% \begin{macro}{\output@post@document} +% The default processing simply provides for insertion of held-over footnotes. +% At a natural page break, we are either at the bottom of a column or at the bottom +% of a page. In either case, the \cmd\output@init@\ processing adjusts for the height +% of the held-over footnotes and bails out. +% Upon our return, at \cmd\output@prep@\ time, the page break will accomodate the material; +% it is now actually inserted by concatenating it with the contents of \cmd\footins. +% The default processing for \cmd\output@post@\ is nil. +% \begin{macrocode} +\def\output@init@document{% + \@ifvoid\footbox{}{% + \global\advance\vsize-\ht\footbox + \global\advance\vsize-\dp\footbox + }% +}% +\def\output@prep@document{% + \@ifvoid\footbox{}{% +% {\tracingall\scrollmode\showbox\footbox\showbox\footins}% + \setbox\footins\vbox{\unvbox\footbox\unvbox\footins}% + }% +}% +\def\output@post@document{}% +% \end{macrocode} +% \end{macro} +% \end{macro} +% \end{macro} +% +% \begin{macro}{\@opcol} +% The standard \LaTeX\ procedure \cmd\@opcol\ is now completely obsoleted. +% \begin{macrocode} +\let\@opcol\@undefined +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\@makecol} +% The procedure \cmd\@makecol\ packages up a page along with all its insertions and floats. +% Therefore it is essential that it be executed with \cmd\holdininserts\ cleared. +% +% Note that there is a corner case when in a multi-column grid, where the change back to +% one-column grid occurs just after a complete page ships out. We want to detect when +% \cmd\@cclv\ contains nothing but a \cmd\mark, but this is a \TeX\ impossibility. +% +% Note on \cmd\@kludgeins: we have removed this mechanism from \LaTeX, because the implementation +% of \cmd\enlargethispage\ no longer requires it. +% Here, for consistency sake, we remove \cmd\@makespecialcolbox. +% \begin{macrocode} +\def\@makecol{% + \setbox\@outputbox\vbox{% + \boxmaxdepth\@maxdepth + \@tempdima\dp\@cclv + \unvbox\@cclv + \vskip-\@tempdima + }% + \xdef\@freelist{\@freelist\@midlist}\global\let\@midlist\@empty + \@combinefloats + \@combineinserts\@outputbox\footins +%\@ifvbox\@kludgeins{% +% \@makespecialcolbox +%}{% + \set@adj@colht\dimen@ + \count@\vbadness + \vbadness\@M + \setbox\@outputbox\vbox to\dimen@{% + \@texttop + \dimen@\dp\@outputbox + \unvbox\@outputbox + \vskip-\dimen@ + \@textbottom + }% + \vbadness\count@ +%}% + \global\maxdepth\@maxdepth +}% +\let\@makespecialcolbox\@undefined +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\@combineinserts} +% We split out the procedure to add the \cmd\footins\ insertions to the packaged-up page. +% Any other non-trivial insertions should also be dealt with at this time. +% \begin{macrocode} +\def\@combineinserts#1#2{% + \setbox#1\vbox{% + \unvbox#1% +% {\tracingall\scrollmode\showbox#2}% + \vbox{% + \@ifvoid#2{}{% + \vskip\skip\footins + \color@begingroup + \normalcolor + \footnoterule + \nointerlineskip + \box#2% + \color@endgroup + }{}% + }% + }% +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\@floatplacement} +% In standard \LaTeX, someone (DPC?) makes the assumption that \cmd\@fpmin\ can be assigned +% locally. This is no longer true now that we ship no more than one page per visit to the output routine. +% We apply a bandaid. +% \begin{macrocode} +\appdef\@floatplacement{% + \global\@fpmin\@fpmin +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\pagebreak@pen} +% While we are in the way of registering certain penalty values, +% let us register the smallest one that will force a visit to the output routine. +% However, this penalty will not have an assciated macro: we wish to execute the +% natural output routine instead. +% +% Note that this penalty is invoked by \cmd\clearpage\ and \cmd\newpage. +% \begin{macrocode} +\mathchardef\pagebreak@pen=\@M +\expandafter\let\csname output@-\the\pagebreak@pen\endcsname\relax +% \end{macrocode} +% \end{macro} +% +% +% \subsection{Float placement}% +% +% \begin{macro}{\do@startcolumn@pen} +% The procedure \cmd\do@startcolumn@pen\ is executed as a one-off output routine +% just after a page is shipped out (or, in a multicolumn page grid, a column is salted away). +% +% Its job is to either generate a ``float page'' (in reality a column) for shipping out, +% or to commit deferred floats to the fresh column, concluding with a dead cycle. +% In the former case, we accomodate split footnotes and other insertions (by comparing \cmd\vsize\ and \cmd\pagegoal): +% the floats are spewed onto the page, whereupon \LaTeX's output routine will place the footnotes and ship out, +% iterating the process once again. +% +% Note that when this procedure is invoked, \cmd\box\cmd\@cclv\ still has within it the protection box, so we +% start by removing it. Note also that if there was a split insertion held over from the previous page, the +% insert node will be present in \cmd\box\cmd\@cclv, \emph{prior to} the protection box. For this reason, we cannot +% just throw away that box, as we might be tempted to do. +% +% FIXME: where else do we possibly inappropriately discard \cmd\box\cmd\@cclv? +% +% Note that, because a column or page page had previously just been completed, +% we can assume that there is nothing of importance on the page, +% and because no message is being passed, we can preserve marks in a simple way. +% +% A Note on terminology: +% In a single-column page grid, you might expect that we would execute the procedure \cmd\do@startpage. +% But this is not so. +% \LaTeX\ has a confustion of long standing, +% in which the procedures that handle full-page width floats in a two-column page grid all have in their names +% the string `dbl', which erroneously suggests having something to do with ``double''. It does not: +% when you see `dbl', think ``full page width''. +% \begin{macrocode} +\mathchardef\do@startcolumn@pen=10005 +\@namedef{output@-\the\do@startcolumn@pen}{\do@startcolumn}% +\def\do@startcolumn{% + \setbox\@cclv\vbox{\unvbox\@cclv\setbox\z@\lastbox\unskip}% + \clearpage@sw{\@clearfloatplacement}{\@floatplacement}% + \set@colroom + \@booleanfalse\pfloat@avail@sw + \begingroup + \@colht\@colroom + \@booleanfalse\float@avail@sw + \@tryfcolumn\test@colfloat + \float@avail@sw{\aftergroup\@booleantrue\aftergroup\pfloat@avail@sw}{}% + \endgroup + \fcolmade@sw{% + \setbox\@cclv\vbox{\unvbox\@outputbox\unvbox\@cclv}% +% \csname float@column@\thepagegrid\endcsname +% \csname output@column@\thepagegrid\endcsname + \outputpenalty-\pagebreak@pen % ask for a return visit, this time with insertions and all. + \dead@cycle + }{% + \begingroup + \let\@elt\@scolelt + \let\reserved@b\@deferlist\global\let\@deferlist\@empty\reserved@b + \endgroup + \clearpage@sw{% + \outputpenalty\@M + }{% + \outputpenalty\do@newpage@pen + }% + \dead@cycle + }% + \check@deferlist@stuck\do@startcolumn + \set@vsize +}% +\def\@scolelt#1{\def\@currbox{#1}\@addtonextcol}% +\def\test@colfloat#1{% + \csname @floatselect@sw@\thepagegrid\endcsname#1{}{\@testtrue}% + \@if@sw\if@test\fi{}{\aftergroup\@booleantrue\aftergroup\float@avail@sw}% +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\@addtonextcol} +% We must adjust \cmd\@addtonextcol\ to take held-over inserts into account. +% Now that all deferred floats are queued up together (in order), we must have a way of +% differentiating them; this is done by the page grid-dependent procedure \cmd\@floatselect@sw@. +% \begin{macrocode} +\def\@addtonextcol{% + \begingroup + \@insertfalse + \@setfloattypecounts + \csname @floatselect@sw@\thepagegrid\endcsname\@currbox{% + \@ifnum{\@fpstype=8 }{}{% + \@ifnum{\@fpstype=24 }{}{% + \@flsettextmin + \@reqcolroom \ht\@currbox + \advance \@reqcolroom \@textmin + \advance \@reqcolroom \vsize % take into account split insertions + \advance \@reqcolroom -\pagegoal + \@ifdim{\@colroom>\@reqcolroom}{% + \@flsetnum \@colnum + \@ifnum{\@colnum>\z@}{% + \@bitor\@currtype\@deferlist + \@if@sw\if@test\fi{}{% + \@addtotoporbot + }% + }{}% + }{}% + }% + }% + }{}% + \@if@sw\if@insert\fi{}{% + \@cons\@deferlist\@currbox + }% + \endgroup +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\do@startpage@pen} +% \begin{macro}{\forcefloats@sw} +% \begin{macro}{\@sdblcolelt} +% \begin{macro}{\test@dblfloat} +% \begin{macro}{\@if@notdblfloat} +% Similar to \cmd\do@startcolumn, +% the procedure \cmd\do@startpage\ starts up a new page (not column) in a multi-column page grid. +% It is invoked after a page is shipped out in a multi-column page grid, and +% it commits full-page-width floats to the fresh page, possibly resulting in a float page. +% In implementation, it is similar to \cmd\do@startcolumn, except that +% it commits effectively via \cmd\@addtodblcol\ instead of \cmd\@addtonextcol. +% Note that this procedure will inevitably be followed by \cmd\do@startcolumn. +% +% Some details of the procedure: +% +% We begin by removing the protection box from \cmd\box\cmd\@cclv, then setting the values of the +% float placement parameters appropriately, and resetting \cmd\@colht, \cmd\@colroom, and \cmd\vsize\ to base values. +% +% Next we attempt to compose a float page, a page consisting entirely of floats. If successful, +% we ship out the float page and lay down an interrupt that will send us back here for another try. +% +% If no float page is formed, we attempt to commit full-page-width floats to the text page, and return with a dead cycle. +% We are now ready to compose columns of text. +% +% Note that all floats (both column floats and full-page-width floats) move through a single queue. +% To differentiate between the two, the width of the float is compared to \cmd\textwidth. +% This comparison is encapsulated in the macro \cmd\@if@notdblfloat, which should be used whenever +% such a determination must be made. This procedure returns a Boolean. +% \begin{macrocode} +\mathchardef\do@startpage@pen=10006 +\@namedef{output@-\the\do@startpage@pen}{\do@startpage}% +\def\do@startpage{% + \setbox\@cclv\vbox{\unvbox\@cclv\setbox\z@\lastbox\unskip}% + \clearpage@sw{\@clearfloatplacement}{\@dblfloatplacement}% + \set@colht + \@booleanfalse\pfloat@avail@sw + \begingroup + \@booleanfalse\float@avail@sw + \@tryfcolumn\test@dblfloat + \float@avail@sw{\aftergroup\@booleantrue\aftergroup\pfloat@avail@sw}{}% + \endgroup + \fcolmade@sw{% + \global\setbox\pagesofar\vbox{\unvbox\pagesofar\unvbox\@outputbox}% + \@combinepage + \@combinedblfloats + \@outputpage + \global\pagegrid@cur\@ne + \protect@penalty\do@startpage@pen + }{% + \begingroup + \@booleanfalse\float@avail@sw + \let\@elt\@sdblcolelt + \let\reserved@b\@deferlist\global\let\@deferlist\@empty\reserved@b + \endgroup + \@ifdim{\@colht=\textheight}{% No luck... + \pfloat@avail@sw{% ...but a float *was* available! + \forcefloats@sw{% + \ltxgrid@warn{Forced dequeueing of floats stalled}% + }{% + \ltxgrid@warn{Dequeueing of floats stalled}% + }% + }{}% + }{}% + \outputpenalty\@M + \dead@cycle + }% + \check@deferlist@stuck\do@startpage + \set@colht +%\set@colroom +}% +\def\@sdblcolelt#1{\def\@currbox{#1}\@addtodblcol}% +\def\test@dblfloat#1{% + \@if@notdblfloat{#1}{\@testtrue}{}% + \@if@sw\if@test\fi{}{\aftergroup\@booleantrue\aftergroup\float@avail@sw}% +}% +\def\@if@notdblfloat#1{\@ifdim{\wd#1<\textwidth}}% +\@booleanfalse\forcefloats@sw +% \end{macrocode} +% \end{macro} +% \end{macro} +% \end{macro} +% \end{macro} +% \end{macro} +% +% \begin{macro}{\@addtodblcol} +% +% The procedure \cmd\@addtodblcol\ is called into play at the beginning of each fresh page +% and operates on each deferred float, in the hopes of placing one or more such floats +% at the top of the current page. +% +% We alter the procedure of standard \LaTeX\ by putting failed floats into +% \cmd\@deferlist\ instead of \cmd\@dbldeferlist. Having done so, we must have a means +% of differentiating full-page-width floats from column-width floats. +% We assume that the latter will always be narrower than \cmd\textwidth. +% +% In aid of detecting a stalled float flushing process, we set a Boolean if we encounter +% a qualified full-page-width float here. Any that qualify but fail the rest of the tests +% might still pass when reconsidered on an otherwise blank page. +% \begin{macrocode} +\def\@addtodblcol{% + \begingroup + \@if@notdblfloat{\@currbox}{% + \false@sw + }{% + \@setfloattypecounts + \@getfpsbit \tw@ + \@bitor \@currtype \@deferlist + \@if@sw\if@test\fi{% + \false@sw + }{% + \@ifodd\@tempcnta{% + \aftergroup\@booleantrue\aftergroup\float@avail@sw + \@flsetnum \@dbltopnum + \@ifnum{\@dbltopnum>\z@}{% + \@ifdim{\@dbltoproom>\ht\@currbox}{% + \true@sw + }{% + \@ifnum{\@fpstype<\sixt@@n}{% + \begingroup + \advance \@dbltoproom \@textmin + \@ifdim{\@dbltoproom>\ht\@currbox}{% + \endgroup\true@sw + }{% + \endgroup\false@sw + }% + }{% + \false@sw + }% + }% + }{% + \false@sw + }% + }{% + \false@sw + }% + }% + }% + {% + \@tempdima -\ht\@currbox + \advance\@tempdima + -\@ifx{\@dbltoplist\@empty}{\dbltextfloatsep}{\dblfloatsep}% + \global \advance \@dbltoproom \@tempdima + \global \advance \@colht \@tempdima + \global \advance \@dbltopnum \m@ne + \@cons \@dbltoplist \@currbox + }{% + \@cons \@deferlist \@currbox + }% + \endgroup +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\@tryfcolumn} +% \begin{macro}{\@wtryfc} +% \begin{macro}{\@xtryfc} +% \begin{macro}{\@ztryfc} +% Whenever a page is shipped out, \LaTeX\ automatically tries out a float column: +% a page containing nothing but floats (and, as we have added here, split footnotes). +% +% The following four procedures employ certain macros to communicate between each other: +% +% \cmd\fcolmade@sw, a boolean, says whether we were successful in making a float column. +% +% \cmd\if@test, a \cmd\newif\ switch, says a float has failed some test. +% +% \cmd\@deferlist, is the input to the process, a list, of deferred floats. +% +% \cmd\@trylist, a list, stores the deferred floats to be tried out on the float column. +% +% \cmd\@failedlist, a list of floats that have failed the selection for the float column. +% +% \cmd\@flfail, a list of floats that have failed the second selection for the float column. +% +% \cmd\@flsucceed, a list, the floats that have been successfully placed on the float column. +% +% \cmd\@freelist, a list, receives any freed floats. +% +% \cmd\@colht, a dimen, the available space for the column, including column floats and insertions (footnotes). +% +% \cmd\@fpmin, a dimen, the required minimum height for the float column. +% +% \cmd\@outputbox, a box, the output of the process. +% +% \cmd\@fptop, \cmd\@fpsep, \cmd\@fpbot, glue, placed above, between, and below floats on the float column. +% +% \cmd\@currtype, a count, used temporarily for the float's bits. +% +% \cmd\@tempcnta, a count, used temporarily for the float's bits. +% +% In \cmd\@tryfcolumn, we alter the criterion for a float page, because if footnotes are present at this point +% (presumably due to a split insertion) then \cmd\@fpmin is no longer the right threshold to apply. +% +% Note that we have changed \cmd\@tryfcolumn, \cmd\@xtryfc, and \cmd\@ztryfc\ syntactically so that the procedure +% to test for the float's being a column float versus a full-page-width float is passed in as an +% argument. +% +% \begin{macrocode} +\def\@tryfcolumn#1{% + \global\@booleanfalse\fcolmade@sw + \@ifx@empty\@deferlist{}{% + \global\let\@trylist\@deferlist + \global\let\@failedlist\@empty + \begingroup + \dimen@\vsize\advance\dimen@-\pagegoal\@ifdim{\dimen@>\z@}{% + \advance\@fpmin-\dimen@ + }{}% + \def\@elt{\@xtryfc#1}\@trylist + \endgroup + \fcolmade@sw{% + \global\setbox\@outputbox\vbox{\vskip \@fptop}% + \let \@elt \@wtryfc \@flsucceed + \global\setbox\@outputbox\vbox{\unvbox\@outputbox + \unskip \vskip \@fpbot + }% + \let \@elt \relax + \xdef\@deferlist{\@failedlist\@flfail}% + \xdef\@freelist{\@freelist\@flsucceed}% + }{}% + }% +}% +\def\@wtryfc #1{% + \global\setbox\@outputbox\vbox{\unvbox\@outputbox + \box #1\vskip\@fpsep + }% +}% +\def\@xtryfc#1#2{% + \@next\reserved@a\@trylist{}{}% trim \@trylist. Ugly! + \@currtype \count #2% + \divide\@currtype\@xxxii\multiply\@currtype\@xxxii + \@bitor \@currtype \@failedlist + \@testfp #2% + #1#2% + \@ifdim{\ht #2>\@colht }{\@testtrue}{}% + \@if@sw\if@test\fi{% + \@cons\@failedlist #2% + }{% + \begingroup + \gdef\@flsucceed{\@elt #2}% + \global\let\@flfail\@empty + \@tempdima\ht #2% + \def \@elt {\@ztryfc#1}\@trylist + \@ifdim{\@tempdima >\@fpmin}{% + \global\@booleantrue\fcolmade@sw + }{% + \@cons\@failedlist #2% + }% + \endgroup + \fcolmade@sw{% + \let \@elt \@gobble + }{}% + }% +}% +\def\@ztryfc #1#2{% + \@tempcnta \count#2% + \divide\@tempcnta\@xxxii\multiply\@tempcnta\@xxxii + \@bitor \@tempcnta {\@failedlist \@flfail}% + \@testfp #2% + #1#2% + \@tempdimb\@tempdima + \advance\@tempdimb \ht#2\advance\@tempdimb\@fpsep + \@ifdim{\@tempdimb >\@colht}{% + \@testtrue + }{}% + \@if@sw\if@test\fi{% + \@cons\@flfail #2% + }{% + \@cons\@flsucceed #2% + \@tempdima\@tempdimb + }% +}% +% \end{macrocode} +% \end{macro} +% \end{macro} +% \end{macro} +% \end{macro} +% +% +% \subsection{Clearing pages}% +% +% Clearing the page is an elaboration of ending the page: it entails flushing all floats. +% +% This package might make number of float flushing algorithms available, +% a very simple one that does not try to produce excellent pages, +% another that tries to make the best use of space, +% and a more complex one that tries to balance columns. +% +% At the beginning of the page-clearing process, by definition all of the paragraph text involved is on the MVL and all floats have been encountered. +% There may be material in \cmd\pagesofar, and (in a multi-column page grid) any number of columns of the page have been composed. +% Also, there might be footnote material saved up in \cmd\footbox. +% +% Because we did not want to perform multiple \cmd\shipout s per visit to the output routine, +% our multi-column page makeup will not compose multiple columns per visit. +% This implementation detail may not require alteration, but it is not a limitation that is truly necessary: +% it is only multiple \cmd\shipout s per visit that must be avoided. +% +% The crux matter is how to continue with flushing floats even after the material in the MVL is exhausted. +% At that point, we must, upon completion of the output routine, +% insert into the MVL an interrupt that triggers the next step in the processing. +% +% Therefore, after processing a \cmd\do@startcolumn\ interrupt, we must somehow force the completion of that column. +% This could be done by inserting a \cmd\do@newpage@pen\ interrupt. +% +% And after processing a \cmd\do@startpage@pen\ interrupt, that results in \cmd\@dbltopinsert s, +% we must ensure that the multiple columns on the page get completed, so that the page itself finally gets shipped out. +% This part will proceed automatically given that \cmd\do@startcolumn\ processing completes successfully. +% +% The process will not be complete until all deferred floats have been placed and shipped out, and all saved-up footnotes have been inserted. +% +% Full-page-width floats can get out of order of column floats. This problem can be remedied by holding them all in the same list. +% We therefore stop using \cmd\@dbldeferlist\ entirely, and all of the procedures that formerly used it have been rewritten to +% use \cmd\@deferlist\ instead. When traversing the list, we apply a selector on the given box that determines whether it is a column-width or page-width float. +% This selector is different depending on the page grid. +% +% When the \cmd\@deferlist\ is processed (by any means), we have to take care of the case where a float of one category is passed over but we are looking for a float of the other category. +% Here, we must terminate processing, to avoid disordering the floats. This we do by the usual means. +% +% The system has a Boolean that says we are clearing pages: \cmd\clearpage@sw; if it is true, +% then at the tail of \cmd\do@startcolumn\ processing, we should put down a (\cmd\vfil?) \cmd\do@newpage@pen\ interrupt. +% This is because the MVL is now empty, so we have to force the columns to complete. +% +% One potential very pathological case would be where there is one or more deferred floats that never successfully get placed: +% placing floats has stalled, and we will ship out blank pages indefinitely. How to detect this case? +% +% First, \cmd\do@startpage\ will evidently be stalled if the following are all true: +% a) \cmd\@tryfcolumn\ and \cmd\@sdblcolelt\ both fail, +% b) there are deferred floats available for page placement, and +% c) the \cmd\@colht=\cmd\textheight, that is, the full page height is available for placement of column floats. +% +% Second, \cmd\do@startcolumn\ will evidently be stalled if the following are all true: +% a) tryfcolumn fails, +% b) there are deferred floats available for column placement, and +% a) the \cmd\@colroom=\cmd\textheight, that is, the full page height is available for placement of column floats. +% +% +% \begin{macro}{\cleardoublepage} +% \begin{macro}{\clearpage} +% \begin{macro}{\newpage} +% \begin{macro}{\newpage@prep} +% The function of \cmd\clearpage\ is to end the current page with \cmd\newpage\ and then +% ship out additional pages until (\footins) inserts and (deferred) floats are exhausted. +% +% The method involves setting the float placement parameters to completely permissive values +% and kicking out the current page (using a non-discardable penalty). +% A possibly short page will be shipped out, followed +% by any number of float pages. However these float pages, because using permissive float placement, +% will exhaust all inserts and deferred floats. +% +% Bug Note: in the code for \cmd\clearpage, the first penalty we output is an unprotected \cmd\pagebreak@pen. +% I tried using a protected \cmd\do@newpage@pen, but that gave rise to a corner case where a blank page +% was output. +% +% At present, the \cmd\clearpage\ procedure does the same as \cmd\newpage, +% except that \cmd\clearpage@sw\ is turned on, +% and the (discardable) \cmd\newpage\ is inevitably followed by the same procedures +% that are executed if a page is shipped out. +% +% FIXME: it seems that better than \cmd\pagebreak@pen\ would be an unprotected penalty of a special value that would +% entail output routine processing consisting of the following steps: +% 3) \cmd\unvbox\cmd\@cclv, +% 1) set \cmd\clearpage@sw\ to \cmd\true@sw, +% 2) put down a protected \cmd\do@startcolumn@pen, +% 4) take a dead cycle. +% +% The effect would be to liberalize float placement options for the current column as well as further columns that may be output as part of \cmd\clearpage\ processing. +% Of course, it would still be necessary to set \cmd\clearpage@sw\ again via an interrupt. +% +% An optimization might be to clear \cmd\clearpage@sw\ as part of the same interrupt, +% but that would actually not work properly, because it is necessary for \cmd\do@endpage\ to +% possibly invoke furhter visits to the output routine before clearpage processing ceases. +% \begin{macrocode} +\def\newpage@prep{% + \if@noskipsec + \ifx \@nodocument\relax + \leavevmode + \global \@noskipsecfalse + \fi + \fi + \if@inlabel + \leavevmode + \global \@inlabelfalse + \fi + \if@nobreak \@nobreakfalse \everypar{}\fi + \par +}% +\def \newpage {% + \newpage@prep + \do@output@MVL{% + \vfil + \penalty-\pagebreak@pen + }% +}% +\def\clearpage{% + \newpage@prep + \do@output@MVL{% + \vfil + \penalty-\pagebreak@pen + \global\@booleantrue\clearpage@sw + \protect@penalty\do@startcolumn@pen + \protect@penalty\do@endpage@pen + }% + \do@output@MVL{% + \global\@booleanfalse\clearpage@sw + }% +}% +\def\cleardoublepage{% + \clearpage + \@if@sw\if@twoside\fi{% + \@ifodd\c@page{}{% + \null\clearpage + }% + }{}% +}% +\@booleanfalse\clearpage@sw +% \end{macrocode} +% \end{macro} +% \end{macro} +% \end{macro} +% \end{macro} +% +% \begin{macro}{\do@endpage@pen} +% The penalty \cmd\do@endpage@pen\ simply dispatches to the page grid procedure that forces an end page. +% That procedure should test whether there is anything to ship out (say committed floats), then act accordingly. +% Note that as part of this work, it should \cmd\unvbox\cmd\@cclv, which has been left boxed up so it can be +% measured. +% \begin{macrocode} +\mathchardef\do@endpage@pen=10007 +\@namedef{output@-\the\do@endpage@pen}{% + \csname end@column@\thepagegrid\endcsname +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\do@newpage@pen} +% The penalty \cmd\do@newpage@pen\ allows a ``non-discardable \cmd\newpage'' command: +% a \cmd\newpage\ command that will not disappear at a pagebreak. +% This visit to the output routine will not be dispatched to an interrupt, +% rather the natural output routine will be executed, where it +% will remove the protection box. +% +% Call this routine by executing \cmd\protect@penalty\cmd\do@newpage@pen. +% \begin{macrocode} +\mathchardef\do@newpage@pen=10001 +\expandafter\let\csname output@-\the\do@newpage@pen\endcsname\relax +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\@clearfloatplacement} +% The procedure \cmd\@clearfloatplacement\ sets all of the float placement parameters +% to completely permissive values. The standard values appear as comments. +% \begin{macrocode} +\def\@clearfloatplacement{% + \global\@topnum \maxdimen % \c@topnumber + \global\@toproom \maxdimen % \topfraction\@colht + \global\@botnum \maxdimen % \c@bottomnumber + \global\@botroom \maxdimen % \bottomfraction\@colht + \global\@colnum \maxdimen % \c@totalnumber +%\global\@fpmin \z@ % \floatpagefraction\@colht + \global\@dbltopnum \maxdimen % \c@dbltopnumber + \global\@dbltoproom \maxdimen % \dbltopfraction\@colht + \global\@textmin \z@ % \@colht\advance \@textmin -\@dbltoproom + \global\@fpmin \z@ % \dblfloatpagefraction\textheight + \let\@testfp\@gobble + \appdef\@setfloattypecounts{\@fpstype16\advance\@fpstype\m@ne}% +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\@doclearpage} +% The \cmd\@doclearpage\ procedure is now obsoleted, as is \cmd\@makefcolumn, which it invoked. +% \begin{macrocode} +\let\@doclearpage\@undefined +\let\@makefcolumn\@undefined +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\clr@top@firstmark} +% \begin{macro}{\set@top@firstmark} +% \begin{macro}{\@outputpage} +% We want accurate values of \cmd\topmark\ and \cmd\firstmark, but we must deal with +% the fact that there are many different ways of contributing material +% to the page. Only upon the first contribution to the page is the value of \cmd\topmark\ +% accurate. However, with \cmd\firstmark\ we must potentially examine each contribution +% because the first mark on the page may happen to fall in the last piece of material contributed. +% +% To begin, we define the procedure that initializes the macros to appropriate flag values. +% \begin{macrocode} +\def\clr@top@firstmark{% + \global\let\saved@@topmark\@undefined + \global\let\saved@@firstmark\@empty + \global\let\saved@@botmark\@empty +}% +\clr@top@firstmark +% \end{macrocode} +% +% Note that the flag value for \cmd\saved@@topmark\ is \cmd\@undefined, just as one would +% expect. But that for \cmd\saved@@firstmark\ and \cmd\saved@@botmark\ is \cmd\@empty. +% +% Next, we define procedure \cmd\set@top@firstmark; it will be exercised everywhere material is contributed, +% capturing the mark values if appropriate. +% \begin{macrocode} +\def\set@top@firstmark{% + \@ifxundefined\saved@@topmark{\expandafter\gdef\expandafter\saved@@topmark\expandafter{\@@topmark}}{}% + \@if@empty\saved@@firstmark{\expandafter\gdef\expandafter\saved@@firstmark\expandafter{\@@firstmark}}{}% + \@if@empty\@@botmark{}{\expandafter\gdef\expandafter\saved@@botmark\expandafter{\@@botmark}}% +}% +% \end{macrocode} +% When should \cmd\set@top@firstmark\ be called? +% A good candidate for a universal procedure for handling contributed material is +% the natural output routine; are any other calls needed? +% +% Yes, in \cmd\save@column\ we must execute \cmd\set@top@firstmark\ because we are about to +% save away \cmd\box\cmd\@cclv, and we will never see its marks again (unless it is unboxed into the MVL), +% because \TeX\ lets one access a box's marks only within an output routine that has put that box into \cmd\box\cmd\@cclv. +% +% As soon as a page is shipped out, we initialize the two macros that +% hold the values of \cmd\topmark\ and \cmd\firstmark, respectively. +% \LaTeX\ has exactly one procedure \cmd\@outputpage\ that does \cmd\shipout, +% which is as it should be: we tailpatch it, and the job is done. +% \begin{macrocode} +\appdef\@outputpage{% + \clr@top@firstmark +}% +% \end{macrocode} +% \end{macro} +% \end{macro} +% \end{macro} +% +% +% \subsection{Other interfaces to \LaTeX}% +% +% \begin{macro}{\@float} +% \begin{macro}{\@dblfloat} +% \begin{macro}{\@yfloat} +% \begin{macro}{\fps@} +% \begin{macro}{\fpsd@} +% The \LaTeX\ kernel procedures \cmd\@float\ and \cmd\@dblfloat\ are treated on an equal footing. +% Each now takes environment-specific float placement defaults. +% If none are defined for the calling environment, we apply a default. +% +% A parameter is passed that will set the width of text within the float, normally \cmd\columnwidth, +% and in the "dbl" version, \cmd\textwidth. However, an environment such as \env{turnpage} +% may change the meanings of these macros to allow turnpage floats. +% +% \begin{macrocode} +\def\@float#1{% + \@ifnextchar[{%}]{%Brace-matching klootch + \@yfloat\width@float{#1}% + }{% + \@ifxundefined@cs{fps@#1}{% + \edef\reserved@a{\noexpand\@yfloat\noexpand\width@float{#1}[\csname fps@\endcsname]}\reserved@a + }{% + \edef\reserved@a{\noexpand\@yfloat\noexpand\width@float{#1}[\csname fps@#1\endcsname]}\reserved@a + }% + }% +}% +\def\@dblfloat#1{% + \@ifnum{\pagegrid@col=\@ne}{% + \@float{#1}% + }{% + \@ifnextchar[{%}]{%Brace-matching klootch + \@yfloat\widthd@float{#1}% + }{% + \@ifxundefined@cs{fpsd@#1}{% + \edef\reserved@a{\noexpand\@yfloat\noexpand\widthd@float{#1}[\csname fpsd@\endcsname]}\reserved@a + }{% + \edef\reserved@a{\noexpand\@yfloat\noexpand\widthd@float{#1}[\csname fpsd@#1\endcsname]}\reserved@a + }% + }% + }% +}% +\def\@yfloat#1#2[#3]{% + \@xfloat{#2}[#3]% + \hsize#1\linewidth\hsize + \minipagefootnote@init +}% +\def\fps@{tbp}% +\def\fpsd@{tp}% +\def\width@float{\columnwidth}% +\def\widthd@float{\textwidth}% +% \end{macrocode} +% \end{macro} +% \end{macro} +% \end{macro} +% \end{macro} +% \end{macro} +% +% \begin{macro}{\end@float} +% \begin{macro}{\end@dblfloat} +% \begin{macro}{\end@@float} +% \begin{macro}{\check@currbox@count} +% \begin{macro}{\minipagefootnote@init} +% \begin{macro}{\minipagefootnote@here} +% \LaTeX\ kernel procedures \cmd\end@float\ and \cmd\end@dblfloat\ +% have been changed to work alike; in particular, floats of both classes +% are deferred into the same queue. +% This measure ensures that they will be placed in their original order, +% an aspect in which \LaTeX\ is broken. +% +% Note: when retrieving floats from the queues, we can differentiate those of the two categories +% by the width of the box. +% +% Floats are processed via an output routine message, and are checked for +% sanity in re the float placement options. In the case of full-page-width floats, +% we ensure that the h and b float placement options are never asserted, because they +% make no sense. +% +% Note that if we get to the end of the float box and still have pending +% footnotes, we put then out. +% +% LaTeX Bug note: if a user types \cmd\begin{table*}[h], the float will never succeed in being placed! +% we try to catch such cases. +% +% Note that the macro \cmd\check@currbox@count\ trys to catch cases where the float placement options +% are such that the float can never be placed. +% \begin{macrocode} +\def\end@float{% + \end@@float{% + \check@currbox@count + }% +}% +\def\end@dblfloat{% + \@ifnum{\pagegrid@col=\@ne}{% + \end@float + }{% + \end@@float{% + \@boxfpsbit\@currbox{1}\@ifodd\@tempcnta{\global\advance\count\@currbox\m@ne}{}% + \@boxfpsbit\@currbox{4}\@ifodd\@tempcnta{\global\advance\count\@currbox-4\relax}{}% + \global\wd\@currbox\textwidth % Klootch + \check@currbox@count + }% + }% +}% +\def\end@@float#1{% + \minipagefootnote@here +%\minipagefootnotes + \@endfloatbox + #1% + \@ifnum{\@floatpenalty <\z@}{% + \@largefloatcheck + \@cons\@currlist\@currbox + \@ifnum{\@floatpenalty <-\@Mii}{% + \do@output@cclv{\@add@float}% + }{% + \vadjust{\do@output@cclv{\@add@float}}% + \@Esphack + }% + }{}% +}% +\def\check@currbox@count{% + \@ifnum{\count\@currbox>\z@}{% + \count@\count\@currbox\divide\count@\sixt@@n\multiply\count@\sixt@@n + \@tempcnta\count\@currbox\advance\@tempcnta-\count@ + \@ifnum{\@tempcnta=\z@}{% + \ltxgrid@warn{Float cannot be placed}% + }{}% + }{% + % Is a \marginpar + }% +}% +\providecommand\minipagefootnote@init{}% +\providecommand\minipagefootnote@here{}% +% \end{macrocode} +% \end{macro} +% \end{macro} +% \end{macro} +% \end{macro} +% \end{macro} +% \end{macro} +% +% \begin{macro}{\@specialoutput} +% The \cmd\@add@float\ procedure used to reside in standard \LaTeX's \cmd\@specialoutput, +% which is no more. +% +% Historical Note: \cmd\@specialoutput\ and Lamport's method of an output routine dispatcher +% is the genesis of our more powerful and refined way of using \TeX's output routine to +% safely accomplish page makeup tasks. To it and to him we owe acknowledgement and thanks. +% \begin{macrocode} +\let\@specialoutput\@undefined +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\@add@float} +% In the following, we do not need to execute \cmd\@reinserts, which was wrong anyway, as you cannot +% reliably recover insertions when they split (unless you have a way of reinserting the captured insertion +% ahead of the split-off part). +% +% Now that full-page-width floats are being processed the same as column floats, we +% have to nip in here and cause them always to be deferred. +% +% At the very end, the \cmd\vsize\ is adjusted for any newly committed float. +% \begin{macrocode} +\def\@add@float{% + \@pageht\ht\@cclv\@pagedp\dp\@cclv + \unvbox\@cclv + \@next\@currbox\@currlist{% + \csname @floatselect@sw@\thepagegrid\endcsname\@currbox{% + \@ifnum{\count\@currbox>\z@}{% + \advance \@pageht \@pagedp + \advance \@pageht \vsize \advance \@pageht -\pagegoal % do not assume \holdinginserts is cleared! +% \@ifvbox\@kludgeins{% +% \@ifdim{\wd\@kludgeins=\z@}{% +% \advance \@pageht \ht\@kludgeins +% }{}% +% }{}% +% \@reinserts + \@addtocurcol % Commit an h float + }{% +% \@reinserts + \@addmarginpar + }% + }{% + \@resethfps + \@cons\@deferlist\@currbox + }% + }{\@latexbug}% + \@ifnum{\outputpenalty<\z@}{% + \@if@sw\if@nobreak\fi{% + \nobreak + }{% + \addpenalty \interlinepenalty + }% + }{}% + \set@vsize +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\@reinserts} +% The \cmd\@reinserts\ procedure of standard \LaTeX\ is now obsoleted. +% \begin{macrocode} +\let\@reinserts\@undefined +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\@addtocurcol} +% We modify the \cmd\@addtocurcol\ procedure of standard \LaTeX\ +% so that a float placed ``here'' may break over pages. +% \begin{macrocode} +\def \@addtocurcol {% + \@insertfalse + \@setfloattypecounts + \ifnum \@fpstype=8 + \else + \ifnum \@fpstype=24 + \else + \@flsettextmin + \advance \@textmin \@textfloatsheight + \@reqcolroom \@pageht + \ifdim \@textmin>\@reqcolroom + \@reqcolroom \@textmin + \fi + \advance \@reqcolroom \ht\@currbox + \ifdim \@colroom>\@reqcolroom + \@flsetnum \@colnum + \ifnum \@colnum>\z@ + \@bitor\@currtype\@deferlist + \if@test + \else + \@bitor\@currtype\@botlist + \if@test + \@addtobot + \else + \ifodd \count\@currbox + \advance \@reqcolroom \intextsep + \ifdim \@colroom>\@reqcolroom + \global \advance \@colnum \m@ne + \global \advance \@textfloatsheight \ht\@currbox + \global \advance \@textfloatsheight 2\intextsep + \@cons \@midlist \@currbox + \if@nobreak + \nobreak + \@nobreakfalse + \everypar{}% + \else + \addpenalty \interlinepenalty + \fi + \vskip \intextsep + \unvbox\@currbox %AO + \penalty\interlinepenalty + \vskip\intextsep + \ifnum\outputpenalty <-\@Mii \vskip -\parskip\fi + \outputpenalty \z@ + \@inserttrue + \fi + \fi + \if@insert + \else + \@addtotoporbot + \fi + \fi + \fi + \fi + \fi + \fi + \fi + \if@insert + \else + \@resethfps + \@cons\@deferlist\@currbox + \fi +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\if@twocolumn} +% The \cmd\newif\ switch \cmd\if@twocolumn\ is entirely unused. However its access words are invoked by +% \LaTeX's \cmd\document\ procedure, so we de-fang it. +% \begin{macrocode} +\@twocolumnfalse +\let\@twocolumntrue\@twocolumnfalse +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\@addmarginpar} +% The procedure \cmd\@addmarginpar\ used to access \cmd\if@twocolumn, but that switch is not reliable; +% the better way is to use \cmd\thepagegrid. We establish a convention for a page-grid-oriented +% procedure, e.g., \cmd\@addmarginpar@one, that emits a boolean, telling this procedure +% whether to set the marginpar on the left or right. +% \begin{macrocode} +\def\@addmarginpar{% + \@next\@marbox\@currlist{% + \@cons\@freelist\@marbox\@cons\@freelist\@currbox + }\@latexbug + \setbox\@marbox\hb@xt@\columnwidth{% + \csname @addmarginpar@\thepagegrid\endcsname{% + \hskip-\marginparsep\hskip-\marginparwidth + \box\@currbox + }{% + \hskip\columnwidth\hskip\marginparsep + \box\@marbox + }% + \hss + }% + \setbox\z@\box\@currbox + \@tempdima\@mparbottom + \advance\@tempdima -\@pageht + \advance\@tempdima\ht\@marbox + \@ifdim{\@tempdima >\z@}{% + \@latex@warning@no@line {Marginpar on page \thepage\space moved}% + }{% + \@tempdima\z@ + }% + \global\@mparbottom\@pageht + \global\advance\@mparbottom\@tempdima + \global\advance\@mparbottom\dp\@marbox + \global\advance\@mparbottom\marginparpush + \advance\@tempdima -\ht\@marbox + \global\setbox \@marbox + \vbox {\vskip \@tempdima + \box \@marbox}% + \global \ht\@marbox \z@ + \global \dp\@marbox \z@ + \kern -\@pagedp + \nointerlineskip + \box\@marbox + \nointerlineskip + \hbox{\vrule \@height\z@ \@width\z@ \@depth\@pagedp}% +}% +% \end{macrocode} +% \end{macro} +% +% \begin{environment}{turnpage} +% Any float (viz., \env{figure} or \env{table}) within the scope of this environment +% will be a turnpage float: It will be assumed to occupy an entire page (constitute a float page), +% the width will be \cmd\textheight, the height \cmd\textwidth, and the entire float will be presented +% rotated 90 degrees. +% +% The implementation requires the services of the \cmd\rotatebox\ command, so we supply a dummy definition +% that explains things to the user. +% \begin{macrocode} +\newenvironment{turnpage}{% + \def\width@float{\textheight}% + \def\widthd@float{\textheight}% + \appdef\@endfloatbox{% + \@ifxundefined\@currbox{% + \ltxgrid@warn{Cannot rotate! Not a float}% + }{% + \setbox\@currbox\vbox to\textwidth{\vfil\unvbox\@currbox\vfil}% + \global\setbox\@currbox\vbox{\rotatebox{90}{\box\@currbox}}% + }% + }% +}{% +}% +\def\rotatebox@dummy#1#2{% + \ltxgrid@warn{You must load the graphics or graphicx package in order to use the turnpage environment}% + #2% +}% +\AtBeginDocument{% + \@ifxundefined\rotatebox{\let\rotatebox\rotatebox@dummy}{}% +}% +% \end{macrocode} +% \end{environment} +% +% +% \subsection{One-off output routines} +% +% These procedures are executed in lieu of \cmd\the\cmd\output\ when the output penalty has the associated flag value. +% +% \begin{macro}{output@-1073741824} +% The first one-off output routine handles the end of the job, wherein +% \LaTeX\ executes \cmd\@@end, and breaks to the output with a penalty of +% $"40000000 = 2^{32}/4$. We simply discard \cmd\box\cmd\@cclv\ and leave. +% This means that \LaTeX\ is obligated to do \cmd\clearpage\ as part of +% its \enve{document} processing, otherwise material will be lost. +% \begin{macrocode} +\@namedef{output@-1073741824}{%"40000000 + \deadcycles\z@ +%\showbox\@cclv + \setbox\z@\box\@cclv +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\save@column@pen} +% The one-off output routine associated with \cmd\penalty\cmd\save@column@pen\ +% will be called within a sequence of three such routines by \cmd\execute@message +% or its companion routine \cmd\execute@message@insert. +% This procedure must save away any the current page and preserve marks. +% \begin{macrocode} +\mathchardef\save@column@pen=10016 +\@namedef{output@-\the\save@column@pen}{\save@column}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\@cclv@saved} +% We take over the \cmd\@holdpg\ box register. Hereafter, +% we no longer use the \cmd\@holdpg\ box register, so let the world know. +% This should decisively break packages that assume standard \LaTeX. +% Breaking decisively is preferred to quietly proceeding erroneously. +% \begin{macrocode} +\let \@cclv@saved \@holdpg +\let \@holdpg \@undefined +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\save@column} +% The procedure \cmd\save@column\ does the actual work of saving away the material +% on the page. It is invoked both by \cmd\save@column@pen\ and by \cmd\save@column@insert@pen. +% We save \cmd\box\cmd\@cclv\ and the primitive \cmd\@@topmark. +% \begin{macrocode} +\def\save@column{% + \@ifvoid\@cclv@saved{% + \set@top@firstmark + \global\@topmark@saved\expandafter{\@@topmark}% + }{}% + \global\setbox\@cclv@saved\vbox{% + \@ifvoid\@cclv@saved{}{% + \unvbox\@cclv@saved + \marry@baselines + }% + \unvbox\@cclv + \lose@breaks + \setbox\z@\lastbox + }% +}% +\newtoks\@topmark@saved +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\prep@cclv} +% The procedure \cmd\prep@cclv\ is used by message handlers to set up their environment +% to ape that of the usual output routine, with the boxed-up page in \cmd\box\cmd\@cclv. +% Here, we retrieve the material from \cmd\@cclv@saved, where it was saved away by +% the one-off output routine associated with \cmd\save@column@pen. +% \begin{macrocode} +\def\prep@cclv{% + \setbox\z@\box\@cclv + \setbox\@cclv\box\@cclv@saved + \vbadness\@M +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\save@column@insert@pen} +% The one-off output routine associated with \cmd\penalty\cmd\save@column@insert@pen\ +% is similar to that of \cmd\save@column@pen\ augmented with the processing of insertions. +% It is called by \cmd\execute@message@insert\ (i.e., at a grid change) +% and saves away the current page and preserves marks. +% In addition, it saves away any insertions that fall on the current page. +% As with the regular output routine, it executes in two phases, +% first with \cmd\holdinginserts\ set, then with it cleared. +% \begin{macrocode} +\mathchardef\save@column@insert@pen=10017 +\@namedef{output@-\the\save@column@insert@pen}{\toggle@insert\savecolumn@holding\savecolumn@moving}% +% \end{macrocode} +% The procedure \cmd\savecolumn@holding\ is the first phase of saving a column with its inserts. +% This phase must detect and remedy the one circumstance that will confound our efforts to propagate marks. +% It is similar to \cmd\output@holding, except that we have to deal with the protection box, which must +% remain, because the messaging mechanism is being used. +% +% If it appears that we have the pathological ``Big Bad Box'' case at hand, we use the \cmd\dead@cycle@repair@protected\ +% procedure instead of \cmd\dead@cycle\ to do our dead cycle. +% \begin{macrocode} +\def\savecolumn@holding{% + \@if@exceed@pagegoal{\unvcopy\@cclv\setbox\z@\lastbox}{% + \setbox\z@\vbox{\unvcopy\@cclv\setbox\z@\lastbox}% + \outputdebug@sw{{\tracingall\scrollmode\showbox\z@}}{}% + \dimen@\ht\@cclv\advance\dimen@-\ht\z@ + \dead@cycle@repair@protected\dimen@ + }{% + \dead@cycle + }% +}% +% \end{macrocode} +% The procedure \cmd\save@column@moving\ is the second phase of saving a column with its inserts. +% Now that \cmd\holdinginserts\ is cleared, we can look in the various \cmd\insert\ registers for +% our inserts (at present there is only one, \cmd\footins). +% if anything is there, we save it away and ask for another cycle (because it may have split). +% +% Note that the message that is about to be executed had better deal properly with the contents of +% the \cmd\footins@saved\ box. +% \begin{macrocode} +\def\savecolumn@moving{% + \@cclv@nontrivial@sw{% + \save@column + }{% + {\setbox\z@\box\@cclv}% + }% + \@ifvoid\footins{}{% + \outputdebug@sw{{\tracingall\scrollmode\showbox\footins}}{}% + \global\setbox\footins@saved\vbox{\unvbox\footins@saved\marry@baselines\unvbox\footins}% + \protect@penalty\save@column@insert@pen + }% +}% +\newbox\footins@saved +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\save@message@pen} +% The one-off output routine associated with \cmd\penalty\cmd\save@message@pen\ +% saves away the message that has been passed. +% This procedure is penultimate in a sequence of one-off output routine calls; +% earlier ones have saved away the MVL and preserved marks, the last executes the message. +% +% Note that we are passing tokens to \TeX's primitive \cmd\mark\ mechanism, so we must ensure +% that they are not inappropriately expanded. We use the same mechanism for all such cases, +% namely \cmd\let@mark. +% +% Note: we expect that \cmd\box\cmd\@cclv's contents are well known: +% \cmd\topskip, protection box, and a \cmd\mark, the latter havin the +% message. +% But if we came here via \cmd\penalty 10017, there might be an \cmd\insert\ node here as well, because +% a footnote split. +% Because this procedure simply voids out \cmd\box\cmd\@cclv, such material would be lost. +% Perhaps we can repair things by manipulating the \cmd\insert \ mechanism temporarily. +% \begin{macrocode} +\mathchardef\save@message@pen=10018 +\@namedef{output@-\the\save@message@pen}{\save@message}% +\def\save@message{% + \setbox\z@\box\@cclv %FIXME: what if \box\@cclv is not empty? + \toks@\expandafter{\@@firstmark}% + \expandafter\gdef\expandafter\@message@saved\expandafter{\the\toks@}% + \expandafter\do@@mark\expandafter{\the\@topmark@saved}% +}% +\gdef\@message@saved{}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\execute@message@pen} +% The one-off output routine associated with \cmd\execute@message@pen\ +% simply executes the given message. It is last in a sequence of one-off output routine calls; +% earlier ones have saved all that require saving. +% \begin{macrocode} +\mathchardef\execute@message@pen=10019 +\@namedef{output@-\the\execute@message@pen}{\@message@saved}% +% \end{macrocode} +% \end{macro} +% +% +% \subsection{Output messages} +% +% Message handlers are procedures that execute output messages, tokens that are passed to +% the output routine for execution in an environment appropriate to page makeup. +% +% How it works. We put down three large negative penalties, each of which will be handled by +% the output dispatcher (not the output routine), each penalty being protected by +% a removable, non-discardable item (i.e., a box). +% Either three or four invocations of one-off output routines are involved per message. +% +% We make the last of the three protection boxes have a depth equal to the value of \cmd\prevdepth\ +% that was current when the procedure is called. This effectively restores \cmd\prevdepth. +% +% In each case, the one-off output routine will remove the extraneous box we have inserted. +% And the second and third one-off routines will simply void \cmd\box\cmd\@cclv, because its contents +% are entirely artificial. +% +% FIXME: not so! If \cmd\holdinginserts\ is cleared, that box may have an insert node; it must be preserved, too. +% +% The first routine saves away the current column contents and remembers the \cmd\topmark\ for later +% use. There is a variant routine that first clears \cmd\holdinginserts, so that the message can +% handle any inserts present in the boxed-up page; this of course entails yet another +% visit to the output routine. +% +% The penultimate routine saves away the tokens transmitted in via the \cmd\@@mark: the argument of +% the macro. These tokens are of course the very thing we wish to execute within the safety +% of the output routine. It also puts down a mark containing the \cmd\topmark\ tokens saved +% by the first routine. By this means, the mark, which we have clobbered, is restored. +% +% The last routine simply executes the given tokens. +% In the course of doing this, it must take care of \cmd\box\cmd\@cclv, either by shipping it out, +% or by \cmd\unvbox ing it onto the MVL. +% +% \begin{macro}{\execute@message} +% The procedure \cmd\execute@message\ simply calls the utility procedure \cmd\@execute@message\ +% with a penalty value for the standard treatment. +% \begin{macrocode} +\def\execute@message{% + \@execute@message\save@column@pen %Implicit #2 +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\execute@message@insert} +% The procedure \cmd\execute@message@insert\ is like \cmd\execute@message\ in all respects +% except that the penalty value is \cmd\save@column@insert@pen, which arranges for the +% message handler involved to deal with the page's insertions. +% At the same time, we prepare the \cmd\footins\ box so that these insertions can be dealt with. +% +% Note: +% If more insertions are added to \LaTeX\ (presumably via \cmd\newinsert), then they +% must be dealt with in a way entirely analogous to \cmd\footins. +% \begin{macrocode} +\def\execute@message@insert#1{% + \@execute@message\save@column@insert@pen{\setbox\footins\box\footins@saved#1}% +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\@execute@message} +% The utility procedure \cmd\@execute@message\ is called by \cmd\execute@message\ and \cmd\execute@message@insert. +% We prepare by creating a \cmd\vbox\ containing all the needed nodes and proceed by simply +% \cmd\unvbox ing that box onto the MVL. +% We ensure that \cmd\box\cmd\@cclv\ is properly set up for the output message handler +% by always inserting \cmd\prep@cclv\ in advance of the argument. +% +% Note that each one-off output routine is invoked effectively the same as +% \cmd\protect@penalty, except that the second invocation involves an additional +% \cmd\mark\ node, and the third a specially prepared protection box. +% +% Note also that \TeX's primitive \cmd\mark\ is called here without any expansion protection. +% This is the only place where it is called that way, but it's OK because those tokens +% have have been pre-expanded by procedures that call \cmd\execute@message. +% FIXME: all procedures calling \cmd\execute@message\ must pre-expand their tokens! +% \begin{macrocode} +\long\def\@execute@message#1#2{% + \begingroup + \dimen@\prevdepth\@ifdim{\dimen@<\z@}{\dimen@\z@}{}% + \setbox\z@\vbox{% + \protect@penalty#1% + \protection@box + \toks@{\prep@cclv#2}% + \@@mark{\the\toks@}% + \penalty-\save@message@pen +% \hbox{\vrule\@height\z@\@width\z@\@depth\dimen@}% + \setbox\z@\null\dp\z@\dimen@\ht\z@-\dimen@ + \nointerlineskip\box\z@ + \penalty-\execute@message@pen + }\unvbox\z@ + \endgroup +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\do@output@cclv} +% The procedure \cmd\do@output@cclv\ provides access to message handlers at their simplest. +% The message will execute in the usual environment of the output routine, with +% the boxed-up page in \cmd\box\cmd\@cclv, and we assume that \cmd\holdinginserts\ remains set. +% This procedure must be invoked within main vertical mode; +% it is the obligation of the macro writer to ensure that this is the case. +% \begin{macrocode} +\def\do@output@cclv{\execute@message}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\do@output@MVL} +% The procedure \cmd\do@output@MVL, like \cmd\do@output@cclv, is an interface for messages, +% but provides two additional services: the command may also be invoked in horizontal mode, and +% the message handler will execute with the MVL unboxed. +% \begin{macrocode} +\def\do@output@MVL#1{% + \@ifvmode{% + \begingroup\execute@message{\unvbox\@cclv#1}\endgroup + }{% + \@ifhmode{% + \vadjust{\execute@message{\unvbox\@cclv#1}}% + }{% + \@latexerr{\string\do@output@MVL\space cannot be executed in this mode!}\@eha + }% + }% +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\lose@breaks} +% The purpose of this procedure is to get rid of all the extraneous +% \cmd\penalty\cmd\@M\ nodes that tend to build up in the MVL. +% \begin{macrocode} +\def\lose@breaks{% + \loopwhile{% + \count@\lastpenalty + \@ifnum{\count@=\@M}{% 10000 is a TeX magic number! + \unpenalty\true@sw + }{% + \false@sw + }% + }% +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\removestuff} +% \cmd\removestuff\ is a document-level command that removes the bottom skip glue item +% from the MVL. +% \begin{macrocode} +\def\removestuff{\do@output@MVL{\unskip\unpenalty}}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\removephantombox} +% The procedure \cmd\removephantombox\ is a special-purpose message handler exclusively for +% preventing incorrect spacing above display math. It must be issued in +% horizontal mode within the phantom paragraph generated when display math starts up in +% vertical mode. +% \begin{macrocode} +\def\removephantombox{% + \vadjust{% + \execute@message{% + \unvbox\@cclv + \setbox\z@\lastbox + \unskip + \unskip + \unpenalty + \penalty\predisplaypenalty + \vskip\abovedisplayskip + }% + }% +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\addstuff} +% \cmd\addstuff\ is a document-level command that adds penalty, glue, or both to the +% MVL. The penalty and glue items are rearranged so that all penalties nodes precede all the glue nodes, +% which is the canonical arrangement. +% \begin{macrocode} +\def\addstuff#1#2{\edef\@tempa{\noexpand\do@output@MVL{\noexpand\@addstuff{#1}{#2}}}\@tempa}% +\def\@addstuff#1#2{% + \skip@\lastskip\unskip + \count@\lastpenalty\unpenalty + \@if@empty{#1}{}{\penalty#1\relax}% + \@ifnum{\count@=\z@}{}{\penalty\count@}% + \vskip\skip@ + \@if@empty{#2}{}{\vskip#2\relax}% +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\replacestuff} +% \cmd\replacestuff\ is a document-level command similar to \cmd\addstuff; +% but it replaces penalty, glue, or both in the MVL. +% The penalty and glue items are rearranged so that all penalties nodes precede all the glue nodes, +% which is the canonical arrangement. +% \begin{macrocode} +\def\replacestuff#1#2{\edef\@tempa{\noexpand\do@output@MVL{\noexpand\@replacestuff{#1}{#2}}}\@tempa}% +\def\@replacestuff#1#2{% + \skip@\lastskip\unskip + \count@\lastpenalty\unpenalty + \@if@empty{#1}{}{% + \@ifnum{\count@>\@M}{}{% + \@ifnum{\count@=\z@}{\count@=#1\relax}{% + \@ifnum{\count@<#1\relax}{}{% + \count@=#1\relax + }% + }% + }% + }% + \@ifnum{\count@=\z@}{}{\penalty\count@}% + \@if@empty{#2}{}{% + \@tempskipa#2\relax + \@ifdim{\z@>\@tempskipa}{% + \advance\skip@-\@tempskipa + }{% + \@ifdim{\skip@>\@tempskipa}{}{% + \skip@\@tempskipa + }% + }% + }% + \vskip\skip@ +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\move@insertions} +% \begin{macro}{\hold@insertions} +% \begin{macro}{\move@insert@sw} +% In order to avoid bolluxing up \cmd\insert\ registers by our one-off +% output routines, we set \cmd\holdinginserts\ to zero by default and only +% clear it (briefly) while we handle cases where we want inserts to show up. +% \begin{macrocode} +\def\move@insertions{\global\holdinginserts\z@}% +\def\hold@insertions{\global\holdinginserts\@ne}% +\hold@insertions +\def\move@insert@sw{\@ifnum{\holdinginserts=\z@}}% +\def\toggle@insert#1#2{% + \@ifnum{\holdinginserts=\z@}{\hold@insertions#2}{\move@insertions#1}% +}% +% \end{macrocode} +% \end{macro} +% \end{macro} +% \end{macro} +% +% +% \subsection{Messages to alter the page grid}% +% +% Here is the implementation of the grid-switching procedures. +% We perform two checks when changing the page grid; first to ensure that +% the target page grid is known (defensive programming), second to ensure that +% the switch is a non-trivial one. The latter check must be performed within +% the safety of the output routine, so requires using an output message. +% Thus, a grid change requires two messages, for a total of six visits to the output routine. +% +% \begin{macro}{\do@columngrid} +% Utility procedure \cmd\do@columngrid\ changes the page grid. +% Note that this command forces an end to the current paragraph. This is necessary, because +% a page grid change makes no sense unless we can alter the \cmd\hsize\ before commencing to +% typeset the following paragraph. So the command should never be executed in horizontal mode +% anyway. +% \begin{macrocode} +\def\do@columngrid#1#2{% + \par + \expandafter\let\expandafter\@tempa\csname open@column@#1\endcsname + \@ifx{\relax\@tempa}{% + \ltxgrid@warn{Unknown page grid #1. No action taken}% + }{% + \do@output@MVL{\start@column{#1}{#2}}% + }% +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\start@column} +% Procedure \cmd\start@column\ lays down the interrupts to switch the page grid. +% If the change to the page grid would have been trivial, it bails out. +% It seems a reasonable tradeoff of processing versus security: once we commit +% to changing the page grid, we clear \cmd\holdinginserts, so there is no turning back. +% +% Note that the second argument to the macro allows us to pass an argument to the +% page grid that is starting up. This can be handy, because a single procedure can +% handle multiple page grids, differing only by the value of a parameter. +% +% FIXME: this means that you cannot switch between mlt page grids in a single step. +% But do we want to do this, at all, at all? +% \begin{macrocode} +\def\start@column#1#2{% + \def\@tempa{#1}\@ifx{\@tempa\thepagegrid}{% + \ltxgrid@info{Already in page grid \thepagegrid. No action taken}% + }{% + \expandafter\execute@message@insert + \expandafter{% + \csname shut@column@\thepagegrid\expandafter\endcsname + \csname open@column@#1\endcsname{#2}% + \set@vsize + }% + }% +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\thepagegrid} +% The macro \cmd\thepagegrid\ tracks what kind of page grid we are in. +% +% Note: Access \cmd\thepagegrid\ only within the safety of the output routine. +% +% Warning: The page grid should be changed only within the safety of the output routine. +% People who write multicol page grid mechanisms appear not to understand the matter, so they +% should particularly heed this warning. Think about it: obviously Lamport did so, which is why +% his \cmd\twocolumn\ command forced a pagebreak, which is limiting, but safe. +% \begin{macrocode} +\def\thepagegrid{one}% +% \end{macrocode} +% \end{macro} +% +% +% \subsection{Application Note: implementing a page grid} +% +% If you want to create a new page grid for \LaTeX, +% you must define five procedures with specific names: +% \cmd\open@column@ name, \cmd\shut@column@ name, \cmd\end@column@ name, +% \cmd\output@column@ name, and \cmd\@addmarginpar@ name, where ``name'' is the +% name of your page grid. +% +% The procedure \cmd\open@column@ name starts the new page grid. It should define \cmd\thepagegrid, +% deal with \cmd\box\cmd\pagesofar\ and \cmd\box\cmd\footbox\ (perhaps by leaving them alone), +% and it should set the values of \LaTeX's page layout parameters for the column size and height. +% +% The procedure \cmd\shut@column@ name should expect to be called with \cmd\holdinginserts\ +% cleared (it can assume that \cmd\holdinginserts\ will automatically be restored). +% It should properly deal with insertions (like footnotes); calling \cmd\@makecol\ +% is a good way to do this. It should know that the page grid is being terminated +% in the middle of a page, so it should make arrangements to carry the footnotes down to the +% bottom of the column or page, and it should possibly salt away the material for later +% incorporation into the page. The box registers \cmd\footbox\ and \cmd\pagesofar\ are customarily used +% for this purpose. +% +% The procedure \cmd\end@column@ name should kick out a possibly short page containing all the +% floats committed to the page. It will be invoked during \cmd\clearpage\ processing. +% After that, it should \cmd\unvbox\cmd\@cclv. +% +% The procedure \cmd\output@column@ name should ship out or commit the current \cmd\@outputbox. +% In a one-column layout, you ship out; in a multicolumn layout, you commit the box as the contents +% of a particular column, and if that column is the last, you ship out. +% +% The procedure \cmd\@addmarginpar@ name should return a boolean (either \cmd\true@sw\ or \cmd\false@sw\ or an equivalent) +% to tell the marginpar mechanism to place the marginal material to the right or left, respectively. +% +% You can use the existing page grids ``one'' and ``mlt'' as a point of departure for creating others. +% The former can be the basis for, say, a single-column page grid with a side column. +% +% \begin{macro}{\pagesofar} +% \begin{macro}{\footbox} +% \begin{macrocode} +\newbox\pagesofar +\newbox\footbox +% \end{macrocode} +% \end{macro} +% \end{macro} +% +% +% \subsubsection{One-column page grid} +% +% \begin{macro}{\onecolumngrid} +% \begin{macro}{\open@column@one} +% \begin{macro}{\shut@column@one} +% \begin{macro}{\float@column@one} +% \begin{macro}{\end@column@one} +% \begin{macro}{\output@column@one} +% \begin{macro}{\@addmarginpar@one} +% Here are all the procedures necessary for the standard page grid named ``one'': +% a single column layout. It is, of course, \LaTeX's familiar \cmd\onecolumn\ layout. +% We begin with the procedure exposed to the style writer. +% This is, however, not a \LaTeX\ command; users should not change the page grid. +% \begin{macrocode} +\newcommand\onecolumngrid{\do@columngrid{one}{\@ne}}% +% \end{macrocode} +% +% Note that a document class that issues the command \cmd\onecolumn\ will break. This includes +% \LaTeX's standard classes.dtx-based classes: if your class descends from one of these, you +% must expunge it of all such commands. +% \begin{macrocode} +\let\onecolumn\@undefined +% \end{macrocode} +% +% The procedure \cmd\open@column@one\ takes advantage of the special nature of the one-column +% page grid to deal with \cmd\box\cmd\pagesofar, therefore it must also reset \cmd\@colroom. +% \begin{macrocode} +\def\open@column@one#1{% + \unvbox\pagesofar + \gdef\thepagegrid{one}% + \global\pagegrid@col#1% + \global\pagegrid@cur\@ne + \set@colht +%\set@colroom + \set@column@hsize\pagegrid@col +}% +% \end{macrocode} +% +% The procedure \cmd\shut@column@one\ saves away the one-column material into the box register \cmd\pagesofar. +% Because it is called from a message handler, we are assured that marks are properly taken care of. +% \begin{macrocode} +\def\shut@column@one{% + \@makecol + \global\setbox\pagesofar\vbox{\unvbox\@outputbox\recover@footins}% + \set@colht +%\set@colroom +}% +% \end{macrocode} +% +% The procedure \cmd\float@column@one\ takes care of a float column that has been built by \cmd\@tryfcolumn, +% in the single-column page grid. +% \begin{macrocode} +\def\float@column@one{% + \@makecol + \@outputpage +}% +% \end{macrocode} +% +% The procedure \cmd\end@column@one\ is executed at the end of \cmd\clearpage\ processing, +% if we were in a one-column page grid, once all permissive float pages have been shipped out. +% At this point, one could perhaps +% assume that nothing more need be done, but let us anyway test for committed floats and force a shipout. +% +% FIXME: this procedure does the same as \cmd\end@column@mlt +% (except for the test of \cmd\@ifx@empty\cmd\@dbltoplist): +% the two could almost be the same procedure. +% +% I have changed this procedure to avoid the testing it once did: it simply puts down interrupts, +% upon which it relies to correctly do what \cmd\clearpage\ requires. +% \begin{macrocode} +\def\end@column@one{% + \unvbox\@cclv\setbox\z@\lastbox + \protect@penalty\do@newpage@pen +}% +% \end{macrocode} +% +% The procedure \cmd\output@column@one\ is dispatched from the output routine when +% we have completed a page (that is, a column in a one-column page grid). +% It ships out the page using the \cmd\@outputpage\ of standard \LaTeX, +% which has been retained (it is needed also in \cmd\output@column@mlt, +% and in any case should remain as the sole procedure in \LaTeX\ where \cmd\shipout\ is performed). +% It will be followed up with an output routine message to prepare a new column. +% \begin{macrocode} +\def\output@column@one{% + \@outputpage +}% +% \end{macrocode} +% +% The following procedure determines which side of the page a marginpar will appear. +% It reproduces the behavior of standard \LaTeX. +% \begin{macrocode} +\def\@addmarginpar@one{% + \@if@sw\if@mparswitch\fi{% + \@ifodd\c@page{\false@sw}{\true@sw}% + }{\false@sw}{% + \@if@sw\if@reversemargin\fi{\false@sw}{\true@sw}% + }{% + \@if@sw\if@reversemargin\fi{\true@sw}{\false@sw}% + }% +}% +% \end{macrocode} +% +% The following procedure yields a Boolean value; it determines whether a float in the deferred queue +% is appropriate for placing. In the one-column grid, all floats are so. +% \begin{macrocode} +\def\@floatselect@sw@one#1{\true@sw}% +% \end{macrocode} +% +% \begin{macrocode} +\def\onecolumngrid@push{% + \do@output@MVL{% + \@ifnum{\pagegrid@col=\@ne}{% + \global\let\restorecolumngrid\@empty + }{% + \xdef\restorecolumngrid{% + \noexpand\start@column{\thepagegrid}{\the\pagegrid@col}% + }% + \start@column{one}{\@ne}% + }% + }% +}% +\def\onecolumngrid@pop{% + \do@output@MVL{\restorecolumngrid}% +}% +% \end{macrocode} +% \end{macro} +% \end{macro} +% \end{macro} +% \end{macro} +% \end{macro} +% \end{macro} +% \end{macro} +% +% +% \subsubsection{Two-column page grid} +% +% \begin{macro}{\twocolumngrid} +% \begin{macro}{\open@column@mlt} +% \begin{macro}{\shut@column@mlt} +% \begin{macro}{\end@column@mlt} +% \begin{macro}{\output@column@mlt} +% \begin{macro}{\@addmarginpar@mlt} +% Here are all the procedures necessary for the standard page grid named ``mlt'': +% the multi-column page grid. With an argument of "2", it is, +% of course, \LaTeX's familiar \cmd\twocolumn\ layout. +% +% We start with the procedure to switch to the two-column page grid. +% \begin{macrocode} +\newcommand\twocolumngrid{\do@columngrid{mlt}{\tw@}}% +% \end{macrocode} +% +% The corresponding command of \LaTeX\ is obsolete. +% \begin{macrocode} +\let\twocolumn\@undefined +% \end{macrocode} +% +% Of course, \cmd\@topnewpage\ is also obsolete. Just do +%\begin{quote} +% \cmd\clearpage\cmd\onecolumngrid<vertical mode material>\cmd\twocolumngrid. +%\end{quote} +% \begin{macrocode} +\let\@topnewpage\@undefined +% \end{macrocode} +% +% If your document class descends from one of \LaTeX's standard classes.dtx-derived +% classes, it will break. You must expunge from it all such commands. +% +% \begin{macrocode} +\def\open@column@mlt#1{% + \gdef\thepagegrid{mlt}% + \global\pagegrid@col#1% + \global\pagegrid@cur\@ne + \set@column@hsize\pagegrid@col + \set@colht +%\set@colroom +}% +% \end{macrocode} +% +% The procedure \cmd\shut@column@mlt\ ends the current column, balances the columns, and +% salts away all in \cmd\pagesofar. Because it is called in a message handler, +% we are assured that marks are handled properly. +% Attention: because this procedure balances columns, all footnotes are +% held aside in \cmd\footbox\ for placement at the bottom of the page. +% +% Bug note: the last macro executed by this procedure is \cmd\set@colht, but had been erroneously \cmd\set@colroom. +% I now believe that the latter should be changed pretty much everywhere to the former. +% \begin{macrocode} +\def\shut@column@mlt{% + \@cclv@nontrivial@sw{% + \@makecol + \@ifnum{\pagegrid@cur<\pagegrid@col}{% + \expandafter\global\expandafter\setbox\csname col@\the\pagegrid@cur\endcsname\box\@outputbox + \global\advance\pagegrid@cur\@ne + }{}% + }{% + {\setbox\z@\box\@cclv}% + }% + \@ifnum{\pagegrid@cur>\@ne}{% + \csname balance@\the\pagegrid@col\endcsname + \grid@column{}% + \@combinepage + \@combinedblfloats + \global\setbox\pagesofar\box\@outputbox + }{}% + \set@colht +}% +% \end{macrocode} +% +% The procedure \cmd\float@column@mlt\ takes care of a float page that has been built by \cmd\@tryfcolumn, +% in the multi-column page grid. +% \begin{macrocode} +\def\float@column@mlt{% + \@combinepage + \@combinedblfloats + \@outputpage + \global\pagegrid@cur\@ne + \protect@penalty\do@startpage@pen +}% +% \end{macrocode} +% +% The procedure \cmd\end@column@mlt\ is executed at the end of \cmd\clearpage\ processing, +% if we were in a multi-column page grid, once all permissive float pages have been shipped out. +% If no floats are committed and if no columns are yet filled, we have nothing to do. +% Otherwise, we kick out a column and try again. +% +% Note that in our code to kick out a column, we must deal properly with the case where the column +% is trivial: it will have nothing but \cmd\topskip\ glue plus a protection box. We substitute an ordinary +% \cmd\null\ for the protection box. +% +% \begin{macrocode} +\def\end@column@mlt{% + \@ifx@empty\@toplist{% + \@ifx@empty\@botlist{% + \@ifx@empty\@dbltoplist{% + \@ifx@empty\@deferlist{% + \@ifnum{\pagegrid@cur=\@ne}{% + \false@sw + }{% + \true@sw + }% + }{% + \true@sw + }% + }{% + \true@sw + }% + }{% + \true@sw + }% + }{% + \true@sw + }% + % true = kick out a column and try again + {% + \@cclv@nontrivial@sw{% + \unvbox\@cclv\setbox\z@\lastbox + }{% + \unvbox\@cclv\setbox\z@\lastbox\unskip\null + }% + \protect@penalty\do@newpage@pen + \protect@penalty\do@endpage@pen + }{% + \unvbox\@cclv\setbox\z@\lastbox + }% +}% +% \end{macrocode} +% The procedure \cmd\output@column@mlt (cf. \cmd\output@column@one) +% is dispatched from the output routine when +% we have completed a column in a multi-column page grid). +% (It replaces the \cmd\@outputdblcol\ of standard \LaTeX.) +% If a complete set of columns is at hand, it ships out the page and +% lays down an interrupt for \cmd\do@startpage@pen, which will commit the +% full-page-width floats to the next page. +% Like \cmd\output@column@mlt, this is followed by +% an output routine message to prepare a new column. +% \begin{macrocode} +\def\output@column@mlt{% + \@ifnum{\pagegrid@cur<\pagegrid@col}{% + \expandafter\global\expandafter\setbox\csname col@\the\pagegrid@cur\endcsname\box\@outputbox + \global\advance\pagegrid@cur\@ne + }{% + \set@adj@colht\dimen@ +% \advance\dimen@-\topskip + \grid@column{}%{\dimen@}% + \@combinepage + \@combinedblfloats + \@outputpage + \global\pagegrid@cur\@ne + \protect@penalty\do@startpage@pen + }% +}% +% \end{macrocode} +% The procedure \cmd\output@column@mlt\ obsoletes \LaTeX's \cmd\@outputdblcol +% \begin{macrocode} +\let\@outputdblcol\@undefined +% \end{macrocode} +% +% The following procedure yields a Boolean value; it determines whether a float in the deferred queue +% is appropriate for placement in the column. In the multi-column grid, only those narrower than \cmd\textwidth\ are so. +% \begin{macrocode} +\def\@floatselect@sw@mlt#1{\@if@notdblfloat{#1}}% +% \end{macrocode} +% +% The following procedure determines which side of the page a marginpar will appear. +% It reproduces the behavior of standard \LaTeX. +% \begin{macrocode} +\def\@addmarginpar@mlt{% emits a boolean + \@ifnum{\pagegrid@cur=\@ne}% +}% +% \end{macrocode} +% \end{macro} +% \end{macro} +% \end{macro} +% \end{macro} +% \end{macro} +% \end{macro} +% +% +% \subsubsection{Page grid utility procedures} +% +% \begin{macro}{\pagegrid@cur} +% \begin{macro}{\pagegrid@col} +% \begin{macro}{\col@} +% \begin{macro}{\pagegrid@init} +% We take over \LaTeX's \cmd\col@number\ and \cmd\@leftcolumn, which are obsolete. +% We create two counters to hold the columns in the page grid and the current column within. +% We also create the first of a set of box registers to hold the committted columns. +% \begin{macrocode} +\let\pagegrid@cur\col@number +\let\col@number\@undefined +\newcount\pagegrid@col +\pagegrid@cur\@ne +\expandafter\let\csname col@\the\pagegrid@cur\endcsname\@leftcolumn +\let\@leftcolumn\@undefined +% \end{macrocode} +% +% The default is for maximum two columns. +% If your class will require more columns, assign that number to \cmd\pagegrid@col\ +% before \envb{document} time. +% \begin{macrocode} +\pagegrid@col\tw@ +% \end{macrocode} +% +% The procedure \cmd\pagegrid@init\ exercises \cmd\newbox\ sufficiently to create the +% boxes for holding the columns in the page grid. +% \begin{macrocode} +\def\pagegrid@init{% + \advance\pagegrid@cur\@ne + \@ifnum{\pagegrid@cur<\pagegrid@col}{% + \csname newbox\expandafter\endcsname\csname col@\the\pagegrid@cur\endcsname + \pagegrid@init + }{% + }% +}% +\appdef\class@documenthook{% + \pagegrid@init +}% +% \end{macrocode} +% \end{macro} +% \end{macro} +% \end{macro} +% \end{macro} +% +% \begin{macro}{\grid@column} +% The procedure \cmd\grid@column\ knows how to lay up the columns in a multi-column page grid. +% It uses utility procedures \cmd\append@column\ and \cmd\box@column. +% +% \begin{macrocode} +\def\grid@column#1{% + \global\setbox\@outputbox\vbox{% + \hb@xt@\textwidth{% + \vrule\@height\z@\@width\z@\@if@empty{#1}{}{\@depth#1}% + \pagegrid@cur\@ne + \append@column + \box@column\@outputbox + }% + \vskip\z@skip % FIXME: page depth! + }% +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\append@column} +% \begin{macro}{\box@column} +% \begin{macro}{\marry@baselines} +% The procedure \cmd\append@column\ appends columns for \cmd\grid@column, +% \cmd\box@column\ builds the columns for \cmd\append@column, +% and \cmd\marry@baselines\ pastes vertical things back together. +% \changes{v4.0rc1}{2001/06/18} +% {Introoduce \cs{marry@height} } +% +% Note that \cmd\box@column\ makes an attempt to prevent excessive \cmd\topskip\ +% or \cmd\baselineskip\ glue +% from being applied by \TeX\ when \cmd\@outputbox\ is contributed to the MVL. +% If this is not done, it is possible to get into an infinite loop in the corner case, +% wherein the page grid is changed to one column and the balanced-up columns are +% already sufficient to fill the page. +% +% Note (AO 0920): I have changed the dimension involved with \cmd\box@column\ from +% \cmd\vsize\ to \cmd\textheight, because the former is certainly not the correct value +% to use: it will change if floats have been placed in the last column of the page. +% I believe \cmd\textheight\ is the correct parameter to use here. +% +% A REVTeX4 beta user, Sergey Strelkov (strelkov@maik.rssi.ru), wants the option +% of ragged-bottom columns. Implementing this feature properly means reboxing the +% columns to their natural height only if \cmd\raggedcolumn@sw\ is true. +% Otherwise, they get reboxed to their common height (\cmd\@colht?). +% +% Note that the default has hereby changed from ragged to flush. +% It's not clear that anyone but Sergey will notice. +% +% The macro \cmd\marry@skip\ addresses (in a limited way) +% the fact that neither the value of \cmd\baselineskip\ nor that of \cmd\topskip\ +% can be relied upon for the purpose of marrying the baselines of two split columns. +% (Because there might have been a local change to their values at the point where +% the output routine got triggered.) +% +% For best results, your document class should call for grid changes only when in basal text +% settings. The \cmd\marry@baselines\ procedure will use the values appropriate to that point +% when attempting to put the columns back together. +% +% In any case, we are not attempting to solve the more general problem of how to marry baselines +% where the leading can change arbitrarily within the galley or where glue could have been trimmed +% at a page top. +% \begin{macrocode} +\def\append@column{% + \@ifnum{\pagegrid@cur<\pagegrid@col}{% + \expandafter\box@column\csname col@\the\pagegrid@cur\endcsname + \hfil + \vrule \@width\columnseprule + \hfil + \advance\pagegrid@cur\@ne + \append@column + }{% + }% +}% +\def\box@column#1{% + \raise\topskip + \hb@xt@\columnwidth{% + \dimen@\ht#1\@ifdim{\dimen@>\@colht}{\dimen@\@colht}{}% +% \advance\dimen@-\topskip + \count@\vbadness\vbadness\@M + \dimen@ii\vfuzz\vfuzz\maxdimen + \outputdebug@sw{\saythe\@colht\saythe\dimen@}{}% + \vtop to\dimen@ +% \@ifdim{\ht#1>\textheight}{to\textheight}{}% + {\hrule\@height\z@ + \unvbox#1% + \raggedcolumn@skip + }% + \vfuzz\dimen@ii + \vbadness\count@ + \hss + }% +}% +\def\marry@baselines{% +%{\tracingall\scrollmode\showlists}% +%\skip@\baselineskip\advance\skip@-\topskip %FIXME: cannot assume \baselineskip nor \topskip + \vskip\marry@skip\relax +}% +\gdef\marry@skip{\z@skip}% +\def\set@marry@skip{% + \begingroup + \skip@\baselineskip\advance\skip@-\topskip + \@ifdim{\skip@>\z@}{% + \xdef\marry@skip{\the\skip@}% + }{}% + \endgroup +}% +\AtBeginDocument{% + \@ifxundefined\raggedcolumn@sw{\@booleanfalse\raggedcolumn@sw}{}% +}% +\def\raggedcolumn@skip{% + \vskip\z@\raggedcolumn@sw{\@plus.0001fil\@minus.0001fil}{}\relax +}% +% \end{macrocode} +% \end{macro} +% \end{macro} +% \end{macro} +% +% \begin{macro}{\@combinepage} +% The procedure \cmd\@combinepage\ prepends the stored page to \cmd\@outputbox. +% \begin{macrocode} +\def\@combinepage{% + \@ifvoid\pagesofar{}{% + \setbox\@outputbox\vbox{% + \unvbox\pagesofar + \marry@baselines + \unvbox\@outputbox + }% + }% + \@ifvoid\footbox{}{% + \setbox\@outputbox\vbox{% + \unvbox\@outputbox + \marry@baselines + \unvbox\footbox + }% + }% +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\@combinedblfloats} +% We modify \LaTeX's \cmd\@combinedblfloats\ to be more appropriate for incremental page building: +% we \cmd\unvbox\ the \cmd\@outputbox. +% \begin{macrocode} +\def\@combinedblfloats{% + \@ifx@empty\@dbltoplist{}{% + \setbox\@tempboxa\vbox{}% + \let\@elt\@comdblflelt\@dbltoplist + \let\@elt\relax\xdef\@freelist{\@freelist\@dbltoplist}% + \global\let\@dbltoplist\@empty + \setbox\@outputbox\vbox{% + %\boxmaxdepth\maxdepth %% probably not needed, CAR + \unvbox\@tempboxa\unskip + \@ifnum{\@dbltopnum>\m@ne}{\dblfigrule}{}%FIXME: how is \@dbltopnum maintained? + \vskip\dbltextfloatsep + \unvbox\@outputbox + }% + }% +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\set@column@hsize}% +% The procedure \cmd\set@column@hsize\ takes care of setting up the horizontal dimensions +% for the current page grid. The present routine will certainly not be adequate for more +% complex page layouts (e.g., with a side column), but works for the common ones. +% \changes{v4.0rc1}{2001/06/18} +% {Introoduce \cs{set@marry@height} } +% \begin{macrocode} +\def\set@column@hsize#1{% + \pagegrid@col#1% + \global\columnwidth\textwidth + \global\advance\columnwidth\columnsep + \global\divide\columnwidth\pagegrid@col + \global\advance\columnwidth-\columnsep + \global\hsize\columnwidth + \global\linewidth\columnwidth + \skip@\baselineskip\advance\skip@-\topskip + \@ifnum{\pagegrid@col>\@ne}{\set@marry@skip}{}% +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\set@colht}% +% \begin{macro}{\set@colroom}% +% \begin{macro}{\set@vsize}% +% \begin{macro}{\set@adj@colht}% +% The story of \cmd\textheight, \cmd\@colht, \cmd\@colroom, and \cmd\vsize. +% +% \cmd\textheight---height of the text column. Not a running parameter, however, each time a page is +% shipped out, the \cmd\textheight\ could in principle be altered. This must be done before +% +% \cmd\@colht---\cmd\textheight\ minus the height of any full-page-width floats. The latter are committed +% only just after shipping out, and only if we are in a multicolumn page grid. +% Therefore, \cmd\@colht\ should be set after a \cmd\shipout\ (by \cmd\@outputpage) and +% will be adjusted when full-page-width floats are committed to the fresh page by \cmd\do@startpage. +% +% \cmd\@colroom---\cmd\@colht\ (adjusted by \cmd\pagesofar) minus the height of any column-width floats. +% The latter are committed anywhere on the page, at which point \cmd\@colroom\ must be adjusted. +% Therefore, \cmd\@colroom\ should be set (by \cmd\set@colroom) whenever a column is prepared (by ). +%FIXME: committed (by \cmd\output@column@) and +% will be adjusted (by \cmd\@add@float\ or \cmd\do@startcolumn) whenever a float is committted to the column. +% +% \cmd\vsize---\cmd\@colroom. +% Therefore, \cmd\vsize\ should be set (by \cmd\set@vsize) whenever +% the \cmd\@colroom\ is set (by \cmd\set@colroom) or adjusted (by \cmd\@add@float\ or \cmd\do@startcolumn) +%FIXME: or when the \cmd\pagesofar\ box is changed (after invoking \cmd\open@column@). +% +% Question: what if there are committed floats? Footnotes? +% Answer: full-page-width floats are only committed at top, and they are already reckoned with in \cmd\@colht. +% Column-width committed floats are incorporated by \cmd\@makecol; footnotes need help. +% +% Note: FIXME: adjusting for \cmd\pagesofar\ is done at not quite the right time. I need to reexamine \cmd\set@colht, +% because \cmd\@dbltoplist\ and \cmd\pagesofar\ really should be on the same footing. +% Perhaps \cmd\@colht\ and \cmd\@colroom\ should both deal with their respective ``lists'' in the same way? +% +% These concerns will be particularly germane if we ever extend this package to deal with full-page-width floats +% placed at the bottom of the page, or committed on the same page as called out. +% +% It occurs to me that we should ditch \cmd\set@colroom\ and only ever execute \cmd\set@colht, which sets \cmd\@colroom\ as a side effect. +% If so, we can make \cmd\@colht\ take \cmd\pagesofar\ into account, as it should. Then \cmd\@colht\ will return to its +% original significance as the value that \cmd\@colroom\ is set to after a column is committed. +% +% On the other hand, why not simply forget all this caching and (re-)calculate \cmd\vsize\ as late as possible? +% Paticularly, \cmd\@colht\ is an artifact of the old way of doing things, where once it was set, it would never change. +% +% \begin{macrocode} +\def\set@colht{% + \set@adj@textheight\@colht + \global\let\enlarge@colroom\@empty + \set@colroom +}% +\def\set@adj@textheight#1{% + #1\textheight + \def\@elt{\adj@page#1}% + \@booleantrue\firsttime@sw\@dbltoplist + \let\@elt\relax +%\@ifvoid\pagesofar{}{% +% \advance#1-\ht\pagesofar\advance#1-\dp\pagesofar +%}% + \global#1#1\relax + \outputdebug@sw{\saythe#1}{}% +}% +\def\set@colroom{% + \set@adj@colht\@colroom + \@if@empty\enlarge@colroom{}{% + \global\advance\@colroom\enlarge@colroom\relax + }% + \outputdebug@sw{\saythe\@colroom}{}% + \@ifdim{\@colroom>\topskip}{}{% + \ltxgrid@info{Not enough room: \string\@colroom=\the\@colroom; increasing to \the\topskip}% + \@colroom\topskip + }% + \global\@colroom\@colroom + \set@vsize +}% +% +\def\set@vsize{% + \global\vsize\@colroom + \outputdebug@sw{\saythe\vsize}{}% +}% +% +\def\set@adj@colht#1{% + #1\@colht + \@ifvoid\pagesofar{}{% + \advance#1-\ht\pagesofar\advance#1-\dp\pagesofar + }% + \@ifvoid\footbox{}{% + \advance#1-\ht\footbox\advance#1-\dp\footbox + }% + \def\@elt{\adj@column#1}% + \@booleantrue\firsttime@sw\@toplist + \@booleantrue\firsttime@sw\@botlist + \let\@elt\relax + \outputdebug@sw{\saythe#1}{}% +}% +\def\adj@column#1#2{% + \advance#1-\ht#2% + \advance#1-\firsttime@sw{\textfloatsep\@booleanfalse\firsttime@sw}{\floatsep}% +}% +\def\adj@page#1#2{% + \advance#1-\ht#2% + \advance#1-\firsttime@sw{\dbltextfloatsep\@booleanfalse\firsttime@sw}{\dblfloatsep}% +}% +% \end{macrocode} +% \end{macro} +% \end{macro} +% \end{macro} +% \end{macro} +% +% \begin{macro}{\@outputpage}% +% At the tail of \cmd\@outputpage, we set \cmd\@colht\ and the float placement parameters +% (this is the one point where it is appropriate to set \cmd\@colht). +% At \cmd\do@startpage\ time, we adjust \cmd\@colht's value to reflect committed +% full-page-width floats. +% +% Note: with a correctly written output routine, a call to \cmd\@outputpage\ will inevitably be +% followed by a call to \cmd\do@startpage, so these procedure calls would be unneeded. +% \begin{macrocode} +\appdef\@outputpage{% + \set@colht % FIXME: needed? + \@floatplacement % FIXME: needed? + \@dblfloatplacement % FIXME: needed? +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{balance@2} +% We define procedures for balancing columns in a multicolumn layout. +% For now, we define only one: a procedure for the two-column grid. +% All others will simply \cmd\relax\ out. +% \begin{macrocode} +\@namedef{balance@2}{% + \expandafter\balance@two\csname col@1\endcsname\@outputbox + % Avoid a bug by preventing a restore when leaving this group + \global\setbox\csname col@1\endcsname\box\csname col@1\endcsname + \@ifvoid\footbox{}{% + \global\setbox\footbox\vbox{% + \setbox\z@\box\@tempboxa + \let\recover@footins\relax + \balance@two\footbox\@tempboxa + \hb@xt@\textwidth{\box\footbox\hfil\box\@tempboxa}% + }% + }% +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\balance@two} +% The procedure \cmd\balance@two\ takes two columns and balances them; in the process it removes +% any footnotes that may be present to a place of safety, for later placement at the foot of the +% shipped-out page. +% The box register \cmd\box\cmd\@ne\ is the aggregate of all columns. +% The box register \cmd\box\cmd \z@\ is the last column. +% The box register \cmd\box\cmd\tw@\ is the first column. +% The \cmd\dimen\ register \cmd\dimen@\ is the trial value to balance to, +% initially half the height of \cmd\box\cmd\@ne. +% The \cmd\dimen\ register \cmd\dimen@i\ is the increment for the next trial; +% its initial value is equal to the initial value of \cmd\dimen@. +% The \cmd\dimen\ register \cmd\dimen@ii\ is the difference of the heights of the two columns. +% +% The procedure uses a binary search for that value of \cmd\dimen@\ which is stable to within .5\cmd\p@\ and which +% makes the last column be shorter than the others. +% +% This procedure can be extended to multiple columns simply by changing it to execute \cmd\vsplit\ multiple +% times (one less than the total number of columns in the page layout) and to calculating \cmd\dimen@ii\ +% to be the difference of the heights of last column and the \cmd\dimen@. +% Upon termination of the search, one would excute the \cmd\vsplit s once again, this time +% using the actual \cmd\col@\ box registers to store the +% balanced columns, thereby clobbering their former contents. +% +% Bug Note: +% as originally written, this macro had a bug, which is well worth avoiding under similar circumstances anywhere. +% So, learn from the mistakes of others, as they say. +% In trying to remove the depth of the boxes created via \cmd\vsplit\ within the \cmd\loopwhile\ control, +% I originally coded +% \cmd\unvbox +% \cmd\z@\ +% \cmd\setbox +% \cmd\z@ +% \cmd\lastbox\ +% \cmd\dimen@ +% \cmd\dp +% \cmd\z@\ +% \cmd\box +% \cmd\z@\ +% \cmd\vskip-% +% \cmd\dimen@. +% The error here is that the shift of the last box in the vertical list will be lost in the process. +% Simply put, \cmd\setbox\cmd\z@\cmd\lastbox\ fails to retain the shift of the box node in the vertical list, +% and when it is put down again via \cmd\box\cmd\z@, it will no longer have the correct shift. +% +% This bug affected things placed in the MVL with \cmd\moveleft, \cmd\moveright, \cmd\parshape, and +% \cmd\hangindent, as well as things shifted by \TeX's primitive mechanisms. +% +% A superior strategy for removing the depth of the last line of the list is more expensive, but safer: +% make a separate copy of the list, measure the depth of the last box as above, but then discard +% the list, retaining only the value of the dimension. +% +% Note that this procedure will not work if the material within is excessively chunky. +% A particular failure mode exists where none of the material is allocated to the last (right) column. +% We detect this case and revert to unbalanced columns. +% +% Another failure mode is where a large chunk occurs at the beginning of the composite box. +% In this case, the left column may fill up even when \cmd\dimen@\ is very small. +% If this configuration leaves the left column longer than the right, then we are done, +% but \cmd\dimen@\ by no means represents the height of either finished box. +% +% Therefore the last step in the process is to rebox the two columns to a common height determined +% independently of the balancing process. +% +% The dimension involved is checked against the current \cmd\@colroom\ to guard against the case where +% excessive material happens to fall in either column. +% \begin{macrocode} +\def\balance@two#1#2{% + \outputdebug@sw{{\tracingall\scrollmode\showbox#1\showbox#2}}{}% + \setbox\@ne\vbox{% + \@ifvoid#1{}{% + \unvcopy#1\recover@footins + \@ifvoid#2{}{\marry@baselines}% + }% + \@ifvoid#2{}{% + \unvcopy#2\recover@footins + }% + }% + \dimen@\ht\@ne\divide\dimen@\tw@ + \dimen@i\dimen@ + \vbadness\@M + \vfuzz\maxdimen + \loopwhile{% + \dimen@i=.5\dimen@i + \outputdebug@sw{\saythe\dimen@\saythe\dimen@i\saythe\dimen@ii}{}% + \setbox\z@\copy\@ne\setbox\tw@\vsplit\z@ to\dimen@ + \setbox\z@ \vbox{% + \unvcopy\z@ + \setbox\z@\vbox{\unvbox\z@ \setbox\z@\lastbox\aftergroup\vskip\aftergroup-\expandafter}\the\dp\z@\relax + }% + \setbox\tw@\vbox{% + \unvcopy\tw@ + \setbox\z@\vbox{\unvbox\tw@\setbox\z@\lastbox\aftergroup\vskip\aftergroup-\expandafter}\the\dp\z@\relax + }% + \dimen@ii\ht\tw@\advance\dimen@ii-\ht\z@ + \@ifdim{\dimen@i>.5\p@}{% + \advance\dimen@\@ifdim{\dimen@ii<\z@}{}{-}\dimen@i + \true@sw + }{% + \@ifdim{\dimen@ii<\z@}{% + \advance\dimen@\tw@\dimen@i + \true@sw + }{% + \false@sw + }% + }% + }% + \outputdebug@sw{\saythe\dimen@\saythe\dimen@i\saythe\dimen@ii}{}% + \@ifdim{\ht\z@=\z@}{% + \@ifdim{\ht\tw@=\z@}{% + \true@sw + }{% + \false@sw + }% + }{% + \true@sw + }% + {% + }{% + \ltxgrid@info{Unsatifactorily balanced columns: giving up}% + \setbox\tw@\box#1% + \setbox\z@ \box#2% + }% + \setbox\tw@\vbox{\unvbox\tw@\vskip\z@skip}% + \setbox\z@ \vbox{\unvbox\z@ \vskip\z@skip}% + \set@colroom + \dimen@\ht\z@\@ifdim{\dimen@<\ht\tw@}{\dimen@\ht\tw@}{}% + \@ifdim{\dimen@>\@colroom}{\dimen@\@colroom}{}% + \outputdebug@sw{\saythe{\ht\z@}\saythe{\ht\tw@}\saythe\@colroom\saythe\dimen@}{}% + \setbox#1\vbox to\dimen@{\unvbox\tw@\unskip\raggedcolumn@skip}% + \setbox#2\vbox to\dimen@{\unvbox\z@ \unskip\raggedcolumn@skip}% + \outputdebug@sw{{\tracingall\scrollmode\showbox#1\showbox#2}}{}% +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\recover@footins} +% The procedure \cmd\recover@footins\ is the utility procedure for recovering the footnotes +% from the bottom of a column. It is used when the page grid is changed, so that footnotes can be +% set at the bottom of the shipped out page. +% \begin{macrocode} +\def\recover@footins{% + \skip\z@ \lastskip\unskip + \skip\@ne\lastskip\unskip + \setbox\z@\lastbox + \@ifvbox\z@{% + \setbox\z@\vbox{% + \unvbox\z@ + \setbox\z@\lastbox +% \outputdebug@sw{{\tracingall\showbox\lastbox}}{}% + \@ifvoid\z@{}{% + \global\setbox\footbox\vbox{% + \unvbox\footbox + \@ifvbox\z@{% + \unvbox\z@ + }{% + \box\z@ + }% + }% + }% + }% + }{}% + \outputdebug@sw{{\tracingall\scrollmode\showbox\footbox}}{}% +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\@begindocumenthook} +% Initialization: +% we initialize to the page grid named ``one''. +% If the class decides to initially set type in a different grid, it +% should execute these same commands, but changing the first to the appropriate procedure. +% +% Note that the point where this sequence is executed would be an excellent place to arrange for +% floats to be committed to the first page of a document. +% That is, we execute \cmd\do@startpage, which triggers \cmd\do@startcolumn. +% +% FIXME: it should be the job of the page grid to determine the procedure to execute at +% the start of the job. Make this a hook. +% \begin{macrocode} +\prepdef\@begindocumenthook{% + \open@column@one\@ne + \set@colht + \@floatplacement + \@dblfloatplacement +}% +% \end{macrocode} +% \end{macro} +% +% Comment: our technique of balancing columns is severely limited, because it cannot properly work +% with \env{longtable}, which places material at the bottom and top of the column break. +% +% The proper way to handle a grid change in the middle of the page is to accumulate all the material for +% an entire article (or chapter) and then assemble finished pages therefrom. This approach is fundamentally +% superior for complex layouts: it corresponds to real-world workflows. +% Such a scheme is an excellent subject for another \LaTeX\ package. +% +% +% \subsection{Patches for the longtable package}% +% +% \LaTeX's ``required'' package \classname{longtable} (written by David P. Carlilsle), +% which is part of /latex/required/tools, is incmpatible with both +% \LaTeX's ``required'' package \classname{multicol} and with +% \LaTeX's native \cmd\twocolumn\ capability. There is no essential reason +% for this incompatability, aside from implementation details, and the +% \classname{ltxgrid} package gives us the ability to lift them. +% +% Only four of \classname{longtable}'s procedures require rewriting: +% \cmd\longtable, +% \cmd\endlongtable, +% \cmd\LT@start, and +% \cmd\LT@end@hd@ft. +% The procedure \cmd\switch@longtable\ checks against their expected meanings +% and, if all is as expected, applies the patches. +% In the process, we simplify things considerably and also make them more +% secure. +% +% Why does \classname{longtable} need to access the output routine, anyway? +% What it comes down to, is what happens when a pagebreak falls within a +% long table. If this happens, we would like to append a row at the bottom of +% the broken table and add a row at the top of the next page. +% +% These things can be accomodated easily by the \classname{ltxgrid} output +% routine hooks. +% +% +% \begin{macro}{\longtable} +% \begin{macrocode} +\def\longtable@longtable{% + \par + \ifx\multicols\@undefined\else\ifnum\col@number>\@ne\@twocolumntrue\fi\fi + \if@twocolumn\LT@err{longtable not in 1-column mode}\@ehc\fi + \begingroup + \@ifnextchar[\LT@array{\LT@array[x]}% +}% +\def\longtable@new{% + \par + \@ifnextchar[\LT@array{\LT@array[x]}% +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\endlongtable} +% \begin{macrocode} +\def\endlongtable@longtable{% + \crcr + \noalign{% + \let\LT@entry\LT@entry@chop + \xdef\LT@save@row{\LT@save@row}}% + \LT@echunk + \LT@start + \unvbox\z@ + \LT@get@widths + \if@filesw + {\let\LT@entry\LT@entry@write\immediate\write\@auxout{% + \gdef\expandafter\noexpand + \csname LT@\romannumeral\c@LT@tables\endcsname + {\LT@save@row}}}% + \fi + \ifx\LT@save@row\LT@@save@row + \else + \LT@warn{Column \@width s have changed\MessageBreak + in table \thetable}% + \LT@final@warn + \fi + \endgraf\penalty -\LT@end@pen + \endgroup + \global\@mparbottom\z@ + \pagegoal\vsize + \endgraf\penalty\z@\addvspace\LTpost + \ifvoid\footins\else\insert\footins{}\fi +}% +\def\endlongtable@new{% + \crcr + \noalign{% + \let\LT@entry\LT@entry@chop + \xdef\LT@save@row{\LT@save@row}% + }% + \LT@echunk + \LT@start + \unvbox\z@ + \LT@get@widths + \@if@sw\if@filesw\fi{% + {% + \let\LT@entry\LT@entry@write + \immediate\write\@auxout{% + \gdef\expandafter\noexpand\csname LT@\romannumeral\c@LT@tables\endcsname + {\LT@save@row}% + }% + }% + }{}% + \@ifx\LT@save@row\LT@@save@row{}{% + \LT@warn{% + Column \@width s have changed\MessageBreak in table \thetable + }\LT@final@warn + }% + \endgraf + \nobreak + \box\@ifvoid\LT@lastfoot{\LT@foot}{\LT@lastfoot}% + \global\@mparbottom\z@ + \endgraf + \LT@post +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\LT@start} +% \begin{macrocode} +\def\LT@start@longtable{% + \let\LT@start\endgraf + \endgraf + \penalty\z@ + \vskip\LTpre + \dimen@\pagetotal + \advance\dimen@ \ht\ifvoid\LT@firsthead\LT@head\else\LT@firsthead\fi + \advance\dimen@ \dp\ifvoid\LT@firsthead\LT@head\else\LT@firsthead\fi + \advance\dimen@ \ht\LT@foot + \dimen@ii\vfuzz\vfuzz\maxdimen + \setbox\tw@\copy\z@ + \setbox\tw@\vsplit\tw@ to \ht\@arstrutbox + \setbox\tw@\vbox{\unvbox\tw@}% + \vfuzz\dimen@ii + \advance\dimen@ \ht + \ifdim\ht\@arstrutbox>\ht\tw@\@arstrutbox\else\tw@\fi + \advance\dimen@\dp + \ifdim\dp\@arstrutbox>\dp\tw@\@arstrutbox\else\tw@\fi + \advance\dimen@ -\pagegoal + \ifdim \dimen@>\z@\vfil\break\fi + \global\@colroom\@colht + \ifvoid\LT@foot\else + \advance\vsize-\ht\LT@foot + \global\advance\@colroom-\ht\LT@foot + \dimen@\pagegoal\advance\dimen@-\ht\LT@foot\pagegoal\dimen@ + \maxdepth\z@ + \fi + \ifvoid\LT@firsthead\copy\LT@head\else\box\LT@firsthead\fi + \output{\LT@output}% +}% +\def\LT@start@new{% + \let\LT@start\endgraf + \endgraf + \markthr@@{}% + \LT@pre + \@ifvoid\LT@firsthead{\LT@top}{\box\LT@firsthead\nobreak}% + \mark@envir{longtable}% +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\LT@end} +% \begin{macrocode} +\def\LT@end@hd@ft@longtable#1{% + \LT@echunk + \ifx\LT@start\endgraf + \LT@err{Longtable head or foot not at start of table}{Increase LTchunksize}% + \fi + \setbox#1\box\z@ + \LT@get@widths\LT@bchunk +}% +\def\LT@end@hd@ft@new#1{% + \LT@echunk + \@ifx{\LT@start\endgraf}{% + \LT@err{Longtable head or foot not at start of table}{Increase LTchunksize}% + }% + \global\setbox#1\box\z@ + \LT@get@widths + \LT@bchunk +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\LT@array} +% \begin{macrocode} +\def\LT@array@longtable[#1]#2{% + \refstepcounter{table}\stepcounter{LT@tables}% + \if l#1% + \LTleft\z@ \LTright\fill + \else\if r#1% + \LTleft\fill \LTright\z@ + \else\if c#1% + \LTleft\fill \LTright\fill + \fi\fi\fi + \let\LT@mcol\multicolumn + \let\LT@@tabarray\@tabarray + \let\LT@@hl\hline + \def\@tabarray{% + \let\hline\LT@@hl + \LT@@tabarray}% + \let\\\LT@tabularcr\let\tabularnewline\\% + \def\newpage{\noalign{\break}}% + \def\pagebreak{\noalign{\ifnum`}=0\fi\@testopt{\LT@no@pgbk-}4}% + \def\nopagebreak{\noalign{\ifnum`}=0\fi\@testopt\LT@no@pgbk4}% + \let\hline\LT@hline \let\kill\LT@kill\let\caption\LT@caption + \@tempdima\ht\strutbox + \let\@endpbox\LT@endpbox + \ifx\extrarowheight\@undefined + \let\@acol\@tabacol + \let\@classz\@tabclassz \let\@classiv\@tabclassiv + \def\@startpbox{\vtop\LT@startpbox}% + \let\@@startpbox\@startpbox + \let\@@endpbox\@endpbox + \let\LT@LL@FM@cr\@tabularcr + \else + \advance\@tempdima\extrarowheight + \col@sep\tabcolsep + \let\@startpbox\LT@startpbox\let\LT@LL@FM@cr\@arraycr + \fi + \setbox\@arstrutbox\hbox{\vrule + \@height \arraystretch \@tempdima + \@depth \arraystretch \dp \strutbox + \@width \z@}% + \let\@sharp##\let\protect\relax + \begingroup + \@mkpream{#2}% + \xdef\LT@bchunk{% + \global\advance\c@LT@chunks\@ne + \global\LT@rows\z@\setbox\z@\vbox\bgroup + \LT@setprevdepth + \tabskip\LTleft\halign to\hsize\bgroup + \tabskip\z@ \@arstrut \@preamble \tabskip\LTright \cr}% + \endgroup + \expandafter\LT@nofcols\LT@bchunk&\LT@nofcols + \LT@make@row + \m@th\let\par\@empty + \everycr{}\lineskip\z@\baselineskip\z@ + \LT@bchunk +}% +\def\LT@LR@l{\LTleft\z@ \LTright\fill}% +\def\LT@LR@r{\LTleft\fill \LTright\z@ }% +\def\LT@LR@c{\LTleft\fill \LTright\fill}% +\def\LT@array@new[#1]#2{% + \refstepcounter{table}\stepcounter{LT@tables}% + \table@hook + \LTleft\fill \LTright\fill + \csname LT@LR@#1\endcsname + \let\LT@mcol\multicolumn + \let\LT@@hl\hline + \prepdef\@tabarray{\let\hline\LT@@hl}% + \let\\\LT@tabularcr + \let\tabularnewline\\% + \def\newpage{\noalign{\break}}% + \def\pagebreak{\noalign{\ifnum`}=0\fi\@testopt{\LT@no@pgbk-}4}% + \def\nopagebreak{\noalign{\ifnum`}=0\fi\@testopt\LT@no@pgbk4}% + \let\hline\LT@hline + \let\kill\LT@kill + \let\caption\LT@caption + \@tempdima\ht\strutbox + \let\@endpbox\LT@endpbox + \@ifxundefined\extrarowheight{% + \let\@acol\@tabacol + \let\@classz\@tabclassz + \let\@classiv\@tabclassiv + \def\@startpbox{\vtop\LT@startpbox}% + \let\@@startpbox\@startpbox + \let\@@endpbox\@endpbox + \let\LT@LL@FM@cr\@tabularcr + }{% + \advance\@tempdima\extrarowheight + \col@sep\tabcolsep + \let\@startpbox\LT@startpbox + \let\LT@LL@FM@cr\@arraycr + }% +% + \let\@acoll\@tabacoll + \let\@acolr\@tabacolr + \let\@acol\@tabacol +% + \setbox\@arstrutbox\hbox{% + \vrule + \@height \arraystretch \@tempdima + \@depth \arraystretch \dp \strutbox + \@width \z@ + }% + \let\@sharp##% + \let\protect\relax + \begingroup + \@mkpream{#2}% + \@mkpream@relax + \edef\@preamble{\@preamble}% + \prepdef\@preamble{% + \global\advance\c@LT@chunks\@ne + \global\LT@rows\z@ + \setbox\z@\vbox\bgroup + \LT@setprevdepth + \tabskip\LTleft + \halign to\hsize\bgroup + \tabskip\z@ + \@arstrut + }% + \appdef\@preamble{% + \tabskip\LTright + \cr + }% + \global\let\LT@bchunk\@preamble + \endgroup + \expandafter\LT@nofcols\LT@bchunk&\LT@nofcols + \LT@make@row + \m@th + \let\par\@empty + \everycr{}% + \lineskip\z@ + \baselineskip\z@ + \LT@bchunk +}% +\appdef\table@hook{}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\switch@longtable} +% +% Here is the switch from standard \classname{longtable} to the new, \classname{ltxgrid}-compatible values. +% +% At this point, we extend \env{longtable} with a \env{longtable*} form, which signifies that we want to +% use the full page width for setting the table. +% You can think this way: \env{longtable*} is to \env{longtable} as \env{table*} is to \env{table}. +% +%FIXME: the following is no longer true: +%% Note that it is not enough to define the environment itself; we also have to create the corresponding +%% \cmd\output\ routine procedures, which provide for continued footers and headers +%% (the very feature of \env{longtable} requiring support in the output routine). +% +%% This same consideration would arise in defining any syntactic extension to \env{longtable}, because +%% the environment name itself is exposed in the output routine. +% +% \begin{macrocode} +\def\switch@longtable{% + \@ifpackageloaded{longtable}{% + \@ifx{\longtable\longtable@longtable}{% + \@ifx{\endlongtable\endlongtable@longtable}{% + \@ifx{\LT@start\LT@start@longtable}{% + \@ifx{\LT@end@hd@ft\LT@end@hd@ft@longtable}{% + \@ifx{\LT@array\LT@array@longtable}{% + \true@sw + }{\false@sw}% + }{\false@sw}% + }{\false@sw}% + }{\false@sw}% + }{\false@sw}% + {% + \class@info{Patching longtable package}% + }{% + \class@info{Patching unrecognized longtable package. (Proceeding with fingers crossed)}% + }% + \let\longtable\longtable@new + \let\endlongtable\endlongtable@new + \let\LT@start\LT@start@new + \let\LT@end@hd@ft\LT@end@hd@ft@new + \let\LT@array\LT@array@new + \newenvironment{longtable*}{% + \onecolumngrid@push + \longtable + }{% + \endlongtable + \onecolumngrid@pop + }% +% \expandafter\let\csname output@init@longtable*\endcsname\output@init@longtable +% \expandafter\let\csname output@prep@longtable*\endcsname\output@prep@longtable +% \expandafter\let\csname output@post@longtable*\endcsname\output@post@longtable + }{}% +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\LT@pre} +% \begin{macro}{\LT@bot} +% \begin{macro}{\LT@top} +% \begin{macro}{\LT@post} +% \begin{macro}{\LT@adj} +% Note that at the end of the longtable environment, we reestablish the \cmd\mark@envir\ of the +% containing environment. We have left \cmd\curr@envir\ alone, so this will work. +% \begin{macrocode} +\def\LT@pre{\penalty\z@\vskip\LTpre}% +\def\LT@bot{\nobreak\copy\LT@foot\vfil}% +\def\LT@top{\copy\LT@head\nobreak}% +\def\LT@post{\penalty\z@\addvspace\LTpost\mark@envir{\curr@envir}}% +\def\LT@adj{% + \setbox\z@\vbox{\null}\dimen@-\ht\z@ + \setbox\z@\vbox{\unvbox\z@\LT@bot}\advance\dimen@\ht\z@ + \global\advance\vsize-\dimen@ +}% +% \end{macrocode} +% \end{macro} +% \end{macro} +% \end{macro} +% \end{macro} +% \end{macro} +% +% \begin{macro}{output@init} +% \begin{macro}{output@prep} +% \begin{macro}{output@post} +% \begin{macrocode} +\def\output@init@longtable{\LT@adj}% +\def\output@prep@longtable{\setbox\@cclv\vbox{\unvbox\@cclv\LT@bot}}% +\def\output@post@longtable{\LT@top}% +% \end{macrocode} +% \end{macro} +% \end{macro} +% \end{macro} +% +% +% \subsection{Patches for index processing}% +% +% Another feature that uses the output routine hooks occurs within +% an index, where one wishes to apply a ``continue head'' when a +% column breaks within a primary index entry. +% Some book designs call for the continue head to only be applied +% at a turnpage break. +% +% In any case, it is easy enough for \cmd\output@post@theindex\ +% to do this in conjunction with component marks. +% Only the bare outlines are shown here. +% +% \begin{macro}{\output@init} +% \begin{macro}{\output@prep} +% \begin{macro}{\output@post} +% \begin{macrocode} +\let\output@init@theindex\@empty +\let\output@prep@theindex\@empty +\def\output@post@theindex{% + \@ifodd\c@page{}{% + \@ifnum{\pagegrid@cur=\@ne}{% we have the leftmost column of a verso page + % insert the current top-level continued head + }% + }% +}% +% \end{macrocode} +% \end{macro} +% \end{macro} +% \end{macro} +% +% +% \subsection{Checking the auxiliary file}% +% +% We relegate the checking of the auxiliary file to the output routine. +% This task must wait until the last page is shipped out, because otherwise +% the stream might get closed before the last page is shipped out. +% Obviously, we must use \cmd\do@output@MVL\ for the job. +% +% \begin{macro}{\check@aux} +% \begin{macrocode} +\def\check@aux{\do@output@MVL{\do@check@aux}}% +% \end{macrocode} +% \end{macro} +% +% +% +% \subsection{Dealing with stuck floats and stalled float dequeueing}% +% +% \LaTeX's float placement mechanism is fundamentally flawed, as evidenced by +% its warning message ``too many unprocessed floats'', which users understandably find frustrating. +% The \classname{ltxgrid} package provides tools for ameliorating the situation somewhat. +% +% Two cases require detection and rectification: +% \begin{enumerate} +% \item +% A float is ``stuck'' in the \cmd\@deferlist: for whatever reason, the float fails to be committed, +% even at the start of a fresh page. +% Once this condition prevails, following floats can never be committed, subsequently all of \LaTeX's +% float registers are used up. +% +% If this condition is detected, we reconsider float dequeueing under permissive (\cmd\clearpage-style) processing. +% +% \item +% The \cmd\@freelist\ is exhausted: +% a large concentration of floats, say, uses up all of \LaTeX's float registers all at once. +% This condition commonly occurs when the user collects floats at the end of the document, for some reason. +% +% When a float is encountered, \LaTeX\ uses a float register (allocated from a pool of free registers) to contain it until it can be placed. +% However, no further action is taken until the pagebuilder is visited, so floats can accumulate. +% Also, even after the pagebuilder is visited, deferred floats can accumulate, and these are not committed +% until a column (or page) of text is completed. +% +% Once the last free float register is used, action should be taken that will commit some of the deferred floats, +% even if this might require ending the page right where we are (resulting in a short page). +% +% Perhaps, committed floats should be stored using some mechanism other than a list, as is currently done. +% A feasible alternative storage method would be to use a \cmd\box\ register in place of +% \cmd\@toplist, +% \cmd\@botlist, and +% \cmd\@dbltoplist. +% This is probably just fine, since such committed floats are not reconsidered (I think). +% +% \end{enumerate} +% +% The emergency processing implemented here immediately ends the current page and begins to output float pages under (\cmd\clearpage-style) rules. +% It proceeds until all deferred floats have been flushed. +% +% Users should expect non-optimal page makeup under these circumstances. +% +% Note that there is a weakness in our approach that we have not attempted to repair: if floats are being +% added as part of a paragraph, we will not be able to take these remedial steps until the paragraph ends. +% This means that the approach implemented here cannot fix all \LaTeX\ documents. Users can still construct +% documents that exhaust \LaTeX's pool of float registers! +% +% \begin{macro}{\check@deferlist@stuck} +% \begin{macro}{\@outputpage} +% We detect the case where, at the start of a fresh page, there are deferred floats, but none are +% committed. We memorize the \cmd\@deferlist\ at \cmd\shipout\ time, then examine it at the point where +% our efforts to commit floats to the new page are complete. +% If it has not changed, the first float must be stuck, and we +% attempt to fix things via \cmd\force@deferlist@stuck. +% +% This simple approach is comp[letely effective in for typical documents. +% +% Note that we try to avoid an infinite loop by examining the value of \cmd\clearpage@sw: +% if we come here with that boolean true, we are in a loop. +% \begin{macrocode} +\def\check@deferlist@stuck#1{% + \@ifx{\@deferlist@postshipout\@empty}{}{% + \@ifx{\@deferlist@postshipout\@deferlist}{% + \@fltstk + \clearpage@sw{% + \ltxgrid@warn{Deferred float stuck during \string\clearpage\space processing}% + }{% + \force@deferlist@stuck#1% + }% + }{% + %Successfully committed float(s) + }% + \global\let\@deferlist@postshipout\@empty + }% +}% +\def\@fltstk{% + \@latex@warning{A float is stuck (cannot be placed without \string\clearpage)}% +}% +\appdef\@outputpage{% + \global\let\@deferlist@postshipout\@deferlist +}% +% \end{macrocode} +% \end{macro} +% \end{macro} +% +% \begin{macro}{\@next} +% \begin{macro}{\@xnext} +% We rewrite the \LaTeX\ kernel macros that dequeue float registers from, e.g., \cmd\@deferlist, +% providing a test for the condition where the pool of free registers is about to underflow. +% +% In this case, we attempt to fix things via \cmd\force@deferlist@empty. +% \begin{macrocode} +\def\@next#1#2{% + \@ifx{#2\@empty}{\false@sw}{% + \expandafter\@xnext#2\@@#1#2% + \true@sw + }% +}% +\def\@xnext\@elt#1#2\@@#3#4{% + \def#3{#1}% + \gdef#4{#2}% + \def\@tempa{#4}\def\@tempb{\@freelist}% + \@ifx{\@tempa\@tempb}{% + \@ifx{#4\@empty}{% + \force@deferlist@empty%{Float register pool exhausted}% + }{}% + }{}% +}% +% \end{macrocode} +% \end{macro} +% \end{macro} +% +% \begin{macro}{\force@deferlist@stuck} +% \begin{macro}{\force@deferlist@empty} +% \begin{macro}{\force@deferlist@sw} +% \begin{macro}{\do@forcecolumn@pen} +% \begin{macro}{\do@forcecolumn} +% The procedure \cmd\force@deferlist@empty\ is an attempt to rectify a situation where \LaTeX's float placement mechanism +% may fail (``too many unprocessed floats''). +% +% We put down interrupts that call for the float placement to be redone, but under permissive conditions, +% just the same as if \cmd\clearpage\ had been invoked. +% +% Note that the attempt to rectify the error is contingent on the setting of \cmd\force@deferlist@sw, +% default false. A document class using this package that wishes to enable this error recovery mechanism should +% set this boolean to true. +% +% The interrupt \cmd\do@forcecolumn@pen, which invokes the procedure \cmd\do@forcecolumn, +% does the same as \cmd\do@startcolumn, except under permissive conditions: +% we are trying to empty out the float registers completely. +% +% In order to properly with the case where there is material in \cmd\box\cmd\@cclv, +% \cmd\@toplist, \cmd\@botlist, \cmd\@dbltoplist, etc, +% we do what amounts to \cmd\newpage\ to get things rolling. +% +% In \cmd\force@deferlist@stuck, we take advantage of already being in the output routine: +% simply reinvoke \cmd\do@startcolumn\ under permissive conditions. +% +% \begin{macrocode} +\def\force@deferlist@stuck#1{% + \force@deferlist@sw{% + \@booleantrue\clearpage@sw + \@booleantrue\forcefloats@sw + #1% + }{% + }% +}% +\def\force@deferlist@empty{% + \force@deferlist@sw{% +% \ltxgrid@info{#1, attempting rectification}% + \penalty-\pagebreak@pen + \protect@penalty\do@forcecolumn@pen + }{% +% \ltxgrid@info{#1}% + }% +}% +\@booleanfalse\force@deferlist@sw +\mathchardef\do@forcecolumn@pen=10009 +\@namedef{output@-\the\do@forcecolumn@pen}{\do@forcecolumn}% +\def\do@forcecolumn{% + \@booleantrue\clearpage@sw + \@booleantrue\forcefloats@sw +%\unvbox\@cclv +%\vfil +%\penalty-\pagebreak@pen + \do@startcolumn +}% +% \end{macrocode} +% \end{macro} +% \end{macro} +% \end{macro} +% \end{macro} +% \end{macro} +% +% A more thorough revision of \LaTeX's float placement mechanism would involve substituting a single \cmd\box\ +% register for the \cmd\@deferlist. This way, \LaTeX's ability to have latent floats would be limited by +% box memory alone. +% +% Because only the \cmd\box\ and \cmd\count\ components of the float box register are actually used by \LaTeX, +% our scheme can be accomplished if we can find a way to encode the information held in the \cmd\count\ component. +% +% A first-in, first-out mechanism exists, wherein a box-penalty pair is dequeued by \cmd\lastbox\cmd\lastpenalty\cmd\unpenalty\ and enqueued +% by \cmd\setbox\cmd\foo=\cmd\hbox\cmd\bgroup\cmd\penalty\cmd\floatpenalty\cmd\box\cmd\floatbox\cmd\unhbox\cmd\foo\cmd\egroup. +% +% Note that this scheme is made possible by our change to \LaTeX's float placement mechanism, +% wherein we consolidated the two \cmd\@deferlist s into one. +% +% \section{Support for legacy \LaTeX\ commands} +% +% We provide support for the \cmd\enlargethispage\ command. +% +% Note: using a command of this sort is questionable. +% Instead, people should enlarge the entire spread. +% +% Timing Note: In a multicolumn page grid, the user should issue the \cmd\enlargethispage\ command +% while the first column of the page is being typeset. +% We provide a helpful message if the timing is wrong. +% +% This code can serve as a model for introducing commands that need to execute within the safety of the output routine. +% We ensure that the arguments are fully expanded, then execute \cmd\do@output@MVL\ to cause an output procedure, +% \cmd\@@enlargethispage, to execute. When it does execute, the MVL will be exposed. +% +% The \cmd\@@enlargethispage\ procedure simply adjusts the vertical dimensions of the page. +% The adjustment will persist until the column is committed, at which point the page dimension +% will revert to its standard value. +% \begin{macrocode} +\def\enlargethispage{% + \@ifstar{% + \@enlargethispage{}% + }{% + \@enlargethispage{}% + }% +}% +\def\@enlargethispage#1#2{% + \begingroup + \dimen@#2\relax + \edef\@tempa{#1}% + \edef\@tempa{\noexpand\@@enlargethispage{\@tempa}{\the\dimen@}}% + \expandafter\do@output@MVL\expandafter{\@tempa}% + \endgroup +}% +\def\@@enlargethispage#1#2{% + \def\@tempa{one}% + \@ifx{\thepagegrid\@tempa}{% + \true@sw + }{% + \def\@tempa{mlt}% + \@ifx{\thepagegrid\@tempa}{% + \@ifnum{\pagegrid@cur=\@ne}{% OK to adjust this page + \gdef\enlarge@colroom{#2}% + \true@sw + }{% Can only adjust this column; give up + \ltxgrid@warn{Too late to enlarge this page; move the command to the first column.}% + \false@sw + }% + }{% Unknown page grid + \ltxgrid@warn{Unable to enlarge a page of this kind.}% + \false@sw + }% + }% + {% + \class@info{Enlarging page \thepage\space by #2}% + \global\advance\@colroom#2\relax + \set@vsize + }{% + % Could not adjust this page + }% +}% +\let\enlarge@colroom\@empty +% \end{macrocode} +% The \cmd\@kludgeins\ insert register is now unneeded. +% Ensure that packages using this mechanism break (preferrable to subtle bugs). +% \begin{macrocode} +\let\@kludgeins\@undefined +% \end{macrocode} +% +% \subsubsection{Building the page for shipout} +% +% \begin{macro}{\@outputpage} +% The procedures that build \cmd\@outputbox\ just before a page is shipped out by \cmd\@outputpage\ are: +% \cmd\@makecol, +% \cmd\@combinepage, and +% \cmd\@combinedblfloats. +% We headpatch \cmd\@outputpage\ to make the \cmd\@outputbox\ be of fixed height. +% \begin{macrocode} +\@booleantrue\textheight@sw +\prepdef\@outputpage{% + \textheight@sw{% + \count@\vbadness\vbadness\@M + \dimen@\vfuzz\vfuzz\maxdimen + \setbox\@outputbox\vbox to\textheight{\unvbox\@outputbox}% + \vfuzz\dimen@ + \vbadness\count@ + }{}% +}% +% \end{macrocode} +% \end{macro} +% +% \subsubsection{Warning message} +% +% \begin{macro}{\ltxgrid@info} +% \begin{macro}{\ltxgrid@warn} +% Something has happened that the user might be interested in. +% Print a message to the log, but only if the user selected the verbose option. +% \begin{macrocode} +\def\ltxgrid@info{% + \ltxgrid@info@sw{\class@info}{\@gobble}% +}% +\@booleanfalse\ltxgrid@info@sw +\def\ltxgrid@warn{% + \ltxgrid@warn@sw{\class@warn}{\@gobble}% +}% +\@booleantrue\ltxgrid@warn@sw +% \end{macrocode} +% \end{macro} +% \end{macro} +% +% \section{End of the \file{ltxgrid} {\sc docstrip} module} +% Here ends the module. +% \begin{macrocode} +%</ltxgrid-krn> +% \end{macrocode} +% +% \Finale +% %Here ends the programmer's documentation. +% \endinput +% +\endinput +%%EOF diff --git a/Master/texmf-dist/source/latex/revtex4/ltxutil.dtx b/Master/texmf-dist/source/latex/revtex4/ltxutil.dtx new file mode 100644 index 00000000000..5553e3a863e --- /dev/null +++ b/Master/texmf-dist/source/latex/revtex4/ltxutil.dtx @@ -0,0 +1,3534 @@ +% \iffalse ltxdoc klootch +% ltxutil.dtx: package to add utilties to LaTeX +% Copyright (c) 2000 Arthur Ogawa +% +% Disclaimer +% This file is distributed WITHOUT ANY WARRANTY; +% without even the implied warranty of +% MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. +% ReadMe +% For the documentation and more detailed instructions for +% installation, typeset this document with \LaTeX. +% \fi +% \GetFileInfo{ltxutil.dtx}\CheckSum{3641} +% +% \iffalse ltxdoc klootch +%<*ltxutil> +%%% @LaTeX-file{ +%%% filename = "ltxutil.dtx", +%%% version = "1.0rc5b", +%%% date = "2001/07/31", +%%% time = "12:23:00 GMT+8", +%%% checksum = "3641", +%%% author = "Arthur Ogawa (mailto:ogawa@teleport.com), +%%% commissioned by the American Physical Society. +%%% ", +%%% copyright = "Copyright (C) 1999 Arthur Ogawa, +%%% distributed under the terms of the +%%% LaTeX Project Public License, see +%%% ftp://ctan.tug.org/macros/latex/base/lppl.txt +%%% ", +%%% address = "Arthur Ogawa, +%%% USA", +%%% telephone = "", +%%% FAX = "", +%%% email = "ogawa@teleport.com", +%%% codetable = "ISO/ASCII", +%%% keywords = "latex, utility, kernel", +%%% supported = "yes", +%%% abstract = "package to add utilties to LaTeX", +%%% docstring = "The checksum field above generated by ltxdoc", +%%% } +%</ltxutil> +% \fi +% +% \iffalse ltxdoc klootch +% The following references the \file{00readme.txt} file, +% which contains basic information about this package. +% The contents of this file are generated when +% you typeset the programmer's documentation. +% Search on "{filecontents*}{00readme.txt}" to locate it. +% \fi\input{00readme.txt}% +% +% \subsection{Bill of Materials} +% +% Following is a list of the files in this distribution arranged +% according to provenance. +% +% \subsubsection{Primary Source}% +% One single file generates all. +%\begin{verbatim} +%ltxutil.dtx +%\end{verbatim} +% +% \subsubsection{Generated by \texttt{latex ltxutil.dtx}}% +% Typesetting the source file under \LaTeX\ +% generates the readme and the installer. +%\begin{verbatim} +%00readme.txt ltxutil.ins +%\end{verbatim} +% +% \subsubsection{Generated by \texttt{tex ltxutil.ins}}% +% Typesetting the installer generates +% the package files. +%\begin{verbatim} +%ltxutil.sty +%\end{verbatim} +% +% \subsubsection{Documentation}% +% The following are the online documentation: +% \begin{verbatim} +%ltxutil.pdf +% \end{verbatim} +% +% \subsubsection{Auxiliary}% +% The following are auxiliary files generated +% in the course of running \LaTeX: +% \begin{verbatim} +%ltxutil.aux ltxutil.idx ltxutil.ind ltxutil.log ltxutil.toc +% \end{verbatim} +% +% \section{Code common to all modules}% +% +% The following may look a bit klootchy, but we +% want to require only one place in this file +% where the version number is stated, +% and we also want to ensure that the version +% number is embedded into every generated file. +% +% Now we declare that +% these files can only be used with \LaTeXe. +% An appropriate message is displayed if +% a different \TeX{} format is used. +% \begin{macrocode} +%<*doc|ltxutil> +\NeedsTeXFormat{LaTeX2e}[1995/12/01]% +%</doc|ltxutil> +% \end{macrocode} +% As desired, the following modules all +% take common version information: +% \begin{macrocode} +%<ltxutil>\ProvidesFile{ltxutil.sty}% +%<*doc> +\expandafter\ProvidesFile\expandafter{\jobname.dtx}% +%</doc> +% \end{macrocode} +% +% The following line contains, for once and for all, +% the version and date information. +% By various means, this information is reproduced +% consistently in all generated files and in the +% typeset documentation. +% \begin{macrocode} +%<*doc|ltxutil> + [2001/07/31 1.0rc5b utilities package]% \fileversion +%</doc|ltxutil> +% \end{macrocode} +% +% +% \section{The driver module \texttt{doc}} +% +% This module, consisting of the present section, +% typesets the programmer's documentation, +% generating the \file{.ins} installer and \file{00readme.txt} as required. +% +% Because the only uncommented-out lines of code at the beginning of +% this file constitute the \file{doc} module itself, +% we can simply typeset the \file{.dtx} file directly, +% and there is thus rarely any need to +% generate the ``doc'' {\sc docstrip} module. +% Module delimiters are nonetheless required so that +% this code does not find its way into the other modules. +% +% The \enve{document} command concludes the typesetting run. +% +% \begin{macrocode} +%<*doc> +% \end{macrocode} +% +% \subsection{The Preamble} +% The programmers documentation is formatted +% with the \classname{ltxdoc} class with local customizations, +% and with the usual code line indexing. +% \begin{macrocode} +\documentclass{ltxdoc} +\RequirePackage{ltxdocext}% +\let\url\undefined +\RequirePackage[colorlinks=true,linkcolor=blue]{hyperref}% +\expandafter\ifx\csname package@font\endcsname\@undefined\else + \expandafter\RequirePackage\expandafter{\csname package@font\endcsname}% +\fi +\CodelineIndex\EnableCrossrefs +% \end{macrocode} +% +% \subsubsection{Docstrip and info directives} +% We use so many {\sc docstrip} modules that we set the +% \texttt{StandardModuleDepth} counter to 1. +% \begin{macrocode} +\setcounter{StandardModuleDepth}{1} +% \end{macrocode} +% The following command retrieves the date and version information +% from this file. +% \begin{macrocode} +\expandafter\GetFileInfo\expandafter{\jobname.dtx}% +% \end{macrocode} +% +% +% \subsection{The installer file} +% +% The installer \file{ltxutil.ins} appears here. +% If you have retrieved the standard distribution of this package, +% the installer file is already on your filesystem. +% If you are bootstrapping, +% the first typesetting of the \file{.dtx} file +% will cause the installer to be generated. +% +% The following modules are used to direct +% {\sc docstrip} in generating the external files: +% \begin{center} +% \begin{tabular}{lll} +% \textbf{Module}&\textbf{File}&\textbf{Description}\\ +% doc &\file{ltxutil.drv}&driver for programmer's documentation\\ +% ltxutil,ltxutil-krn &\file{ltxutil.sty}&this package\\ +% ltxutil-krn& &the portion of this package suitable for inclusion within another package +% \end{tabular} +% \end{center} +% +% \begin{macrocode} +\begin{filecontents}{ltxutil.ins} +%% This file will generate documentation and runtime files +%% from ltxutil.dtx when run through LaTeX or TeX. +\input docstrip +\preamble + +This is a generated file; +altering it directly is inadvisable; +instead, modify the original source file. +See the URL in the file 00readme.txt. + +Copyright notice. + + These files are distributed + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. + +\endpreamble +\keepsilent + \generate{% + \file{ltxutil.drv}{\from{ltxutil.dtx}{doc}}% + \file{ltxutil.sty}{% + \from{ltxutil.dtx}{ltxutil,ltxutil-krn}% + }% + }% +\ifToplevel{ +\Msg{***********************************************************} +\Msg{*} +\Msg{* To finish the installation, please move} +\Msg{* ltxutil.sty} +\Msg{* into a directory searched by TeX.} +\Msg{*} +\Msg{* To produce the documentation, + run ltxutil.dtx through LaTeX.} +\Msg{*} +\Msg{* Happy TeXing} +\Msg{***********************************************************} +} +\endbatchfile +\end{filecontents} +% \end{macrocode} +% Note that, because all of the files generated by the installer +% are part of the standard distribution, it will +% be necessary to run the installer only when bootstrapping +% (or, of course, during development). +% Note, too, that it is rare to generate the \file{doc} +% module because it suffices to simply typeset the \file{.dtx} file itself. +% +% \subsection{The ``Read Me'' File} +% As promised above, here is the contents of the +% ``Read Me'' file. That file serves a double purpose, +% since it also constitutes the beginining of the +% programmer's documentation. What better thing, after +% all, to have appear at the beginning of the +% typeset documentation? +% +% A good discussion of how to write a ReadMe file can be found in +% Engst, Tonya, ``Writing a ReadMe File? Read This'' +% \emph{MacTech} October 1998, p. 58. +% +% Note the appearance of the +% \cmd\StopEventually\ command, which marks the +% dividing line between the user documentation +% and the programmer documentation. +% +% The usual user will not be asked to +% do a full build, not to speak +% of the bootstrap. +% Instructions for carrying these processes +% begin the programmer's manual. +% +% \begin{macrocode} +\begin{filecontents*}{00readme.txt} +\title{% + A \LaTeX\ Package of utility macros% + \thanks{% + This file has version number \fileversion, + last revised \filedate.% + % For version number and date, + % search on "\fileversion" in the .dtx file, + % or see the end of the 00readme.txt file. + }% +}% + +\author{% +Arthur Ogawa (\texttt{mailto:ogawa@teleport.com}), +\fileversion\\Copyright (C) 1999 Arthur Ogawa +}% +\maketitle + +This file embodies the \classname{ltxutil} package, +the implementation and its user documentation. + +The distribution point for this work is +\url{ftp://ftp.teleport.com/users/ogawa/macros/latex/contrib/supported/ltxutil...}, +which contains fully unpacked, prebuilt runtime files and documentation. + +The \classname{ltxutil} package was commissioned by the American Physical Society +and is distributed under the terms of the \LaTeX\ Project Public License, +the same license under which all the portions of \LaTeX\ itself is distributed. +Please see \url{http://ctan.tug.org/macros/latex/base/lppl.txt} for details. + +To use this document class, you must have a working +\TeX\ installation equipped with \LaTeXe\ +and possibly pdftex and Adobe Acrobat Reader or equivalent. + +To install, retrieve the distribution, +unpack it into a directory on the target computer, +and move the file \file{ltxutil.sty} +into a location in your filesystem where it will be found by \LaTeX. + +To use, read the user documentation \file{ltxutil.pdf}. + +\tableofcontents + +\section{Processing Instructions} + +The package file \file{ltxutil.sty} +is generated from this file, \file{ltxutil.dtx}, +using the {\sc docstrip} facility of \LaTeX +via |tex ltxutil.ins|. +The typeset documentation that you are now reading is generated from +the same file by typesetting it with \LaTeX\ or pdftex +via |latex ltxutil.dtx| or |pdflatex ltxutil.dtx|. + +\subsection{Build Instructions} + +You may bootstrap this suite of files solely from \file{ltxutil.dtx}. +Prepare by installing \LaTeXe\ (and either tex or pdftex) on your computer, +then carry out the following steps: +\begin{enumerate} +\item +Within an otherwise empty directory, +typeset \file{ltxutil.dtx} with \LaTeX\ or pdflatex; +you will obtain the typeset documentation you are now reading, +along with +the installer \file{ltxutil.ins}, +and the file \file{00readme.txt}. + +Note: you will have to run \LaTeX\ twice, then \file{makeindex}, then +\LaTeX\ again in order to obtain a valid index and table of contents. +\item +Now typeset \file{ltxutil.ins}, +thereby generating the package file \file{ltxutil.sty}. +\item +Install \classname{ltxutil.sty} +by moving it to a location +in your filesystem where they will be found by \LaTeX. +\end{enumerate} +\end{filecontents*} +% \end{macrocode} +% +% \subsection{The Document Body} +% +% Here is the document body, containing only a +% \cmd\DocInput\ directive---referring to this very file. +% This very cute self-reference is a common \classname{ltxdoc} idiom. +% \begin{macrocode} +\begin{document}% +\expandafter\DocInput\expandafter{\jobname.dtx}% +% ^^A\PrintChanges +\end{document} +% \end{macrocode} +% +% \begin{macrocode} +%</doc> +% \end{macrocode} +% +% \section{Using this package} +% Once this package is installed on your filesystem, you can employ it in +% adding functionality to \LaTeX\ by invoking it in your document or document class. +% +% \subsection{Invoking the package} +% In your document, you can simply call it up in your preamble: +% \begin{verbatim} +%\documentclass{book}% +%\usepackage{ltxutil}% +%\begin{document} +%<your document here> +%\end{document} +% \end{verbatim} +% However, the preferred way is to invoke this package from within your +% customized document class: +% \begin{verbatim} +%\NeedsTeXFormat{LaTeX2e}[1995/12/01]% +%\ProvidesClass{myclass}% +%\RequirePackage{ltxutil}% +%\LoadClass{book}% +%<class customization commands> +%\endinput +% \end{verbatim} +% +% Once loaded, the package gives you acccess to certain procedures, +% usually to be invoked by a \LaTeX\ command or environment, but not at the document level. +% +% +% \section{Compatability with \LaTeX's Required Packages} +% Certain packages, usually ones written by members of the +% \LaTeX\ Project itself, have been designated ``required'' and +% are distributed as part of standard \LaTeX. +% These packages have been placed in a priviledged position +% vis \'a vis the \LaTeX\ kernel in that they override the definitions of certain kernel macros. +% +% The \classname{ltxutil} package will be incompatible with any package that +% redefines any of the kernel macros that \classname{ltxutil} patches---if that +% package is loaded \emph{after} \classname{ltxutil}. This means that for +% greatest compatability, \classname{ltxutil} should be loaded \emph{after}, +% say, \classname{ftnright}, which overwrites \LaTeX's kernel +% procedures \cmd\@outputdblcol, \cmd\@startcolumn, and \cmd\@makecol. +% +% Hereinafter follows some notes on specific \LaTeX\ packages. +% +% \subsection{array} +% This package alters the way tabular environments are done, +% therefore it could run afoul of the \LaTeX\ ``required'' package \classname{array} or any +% package that calls for it to be loaded. +% However, this package has provisions for remaining compatible with \classname{array}. +% So long as the version of \classname{array} that is used with this package has the appropriate +% meanings for the procedures it overwrites, all should be well. +% +% \subsection{longtable} +% David Carlisle's \classname{longtable} package modifies both the \LaTeX\ kernel and the +% \classname{array} package. This package must therefore alter \cmd\LT@array. +% For now, that job is handled by \classname{ltxgrid}. +% +% +%\StopEventually{} +% +% \section{Implementation of package} +% +% Special acknowledgment: this package uses concepts pioneered +% and first realized by William Baxter (mailto:web@superscript.com) +% in his SuperScript line of commercial typesetting tools, and +% which are used here with his permission. +% +% \subsection{Beginning of the \file{ltxutil} {\sc docstrip} module} +% \begin{macrocode} +%<*ltxutil> +\def\package@name{ltxutil}% +\expandafter\PackageInfo\expandafter{\package@name}{% + Utility macros for \protect\LaTeXe, + by A. Ogawa (ogawa@teleport.com)% +}% +%</ltxutil> +% \end{macrocode} +% +% \subsection{Banner}% +% Credit where due. +% \begin{macrocode} +%<*ltxutil-krn> +\typeout{% + ltxutil: portions licensed from W. E. Baxter (web@superscript.com)% +}% +% \end{macrocode} +% +% \subsection{Errors and warnings} +% +% \begin{macro}{\class@err} +% \begin{macro}{\class@warn} +% \begin{macro}{\class@info} +% A few shorthands for Class messages. +% Your document class should define \cmd\class@name. +% \begin{macrocode} +\def\class@err#1{\ClassError{\class@name}{#1}\@eha}% +\def\class@warn#1{\ClassWarningNoLine{\class@name}{#1}}% +\def\class@info#1{\ClassInfo{\class@name}{#1}}% +\def\obsolete@command#1{% + \class@warn@end{Command \string#1\space is obsolete.^^JPlease remove from your document}% + \global\let#1\@empty + #1% +}% +\def\replace@command#1#2{% + \class@warn@end{Command \string#1\space is obsolete;^^JUse \string#2\space instead}% + \global\let#1#2% + #1% +}% +\def\replace@environment#1#2{% + \class@warn@end{Environment #1 is obsolete;^^JUse #2 instead}% + \glet@environment{#1}{#2}% + \@nameuse{#1}% +}% +\def\incompatible@package#1{% + \@ifpackageloaded{#1}{% + \def\@tempa{I cannot continue. You must remove the \string\usepackage\ statement that caused that package to be loaded.}% + \ClassError{\class@name}{The #1 package cannot be used with \class@name}% + \@tempa\stop + }{% + \class@info{#1 was not loaded (OK!)}% + }% +}% +\def\class@warn@end#1{% + \gappdef\class@enddocumenthook{\class@warn{#1}}% +}% +\AtEndOfClass{% + \@ifxundefined\class@name{\def\class@name{Generic Class}}{}% +}% +% \end{macrocode} +% \end{macro} +% \end{macro} +% \end{macro} +% +% \subsection{New Tools}% +% +% \begin{macro}{\t@} +% \begin{macrocode} +\def\t@{to}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\dimen@iii} +% \begin{macrocode} +\dimendef\dimen@iii\thr@@ +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\halignt@} +% \begin{macrocode} +\def\halignt@{\halign\t@}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\f@ur} +% Analogous to \cmd\@ne, \cmd\tw@, and \cmd\thr@@. +% \begin{macrocode} +\chardef\f@ur=4\relax +\chardef\cat@letter=11\relax +\chardef\other=12\relax +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\let@environment} +% \begin{macro}{\glet@environment} +% The directive \cmd\let@environment\ takes care of a common programming +% idiom whereby one environment is made a synonym for another. +% \begin{macrocode} +\def\let@environment#1#2{% + \expandafter\let + \csname#1\expandafter\endcsname\csname#2\endcsname + \expandafter\let + \csname end#1\expandafter\endcsname\csname end#2\endcsname +}% +\def\glet@environment#1#2{% + \global\expandafter\let + \csname#1\expandafter\endcsname\csname#2\endcsname + \global\expandafter\let + \csname end#1\expandafter\endcsname\csname end#2\endcsname +}% +% \end{macrocode} +% \end{macro} +% \end{macro} +% +% \begin{macro}{\tracingplain} +% The command \cmd\tracingplain\ causes \TeX's tracing parameters to +% return to the values set by default. This command is sometimes +% useful when you have said \cmd\tracingall\ somewhere and want to +% restore. +% The \cmd\traceoutput\ command causes \cmd\tracingoutput\ diagnostics +% upon \cmd\shipout. +% \begin{macrocode} +\newcommand\tracingplain{% + \tracingonline\z@\tracingcommands\z@\tracingstats\z@ + \tracingpages\z@\tracingoutput\z@\tracinglostchars\@ne + \tracingmacros\z@\tracingparagraphs\z@\tracingrestores\z@ + \showboxbreadth5\showboxdepth3\relax %\errorstopmode + }% +\newcommand\traceoutput{% + \appdef\@resetactivechars{\showoutput}% +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\say} +% \begin{macro}{\saythe} +% The commands \cmd\say\ and \cmd\saythe\ cause diagnostic messages in the +% \TeX\ log that give the value of a control sequence name or a register +% respectively. +% \begin{macrocode} +\newcommand\say[1]{\typeout{<\noexpand#1=\meaning#1>}}% +\newcommand\saythe[1]{\typeout{<\noexpand#1=\the#1>}}% +% \end{macrocode} +% \end{macro} +% \end{macro} +% +% \begin{macro}{\fullinterlineskip} +% Resets the \cmd\prevdepth\ so that the full amount of \cmd\baselineskip\ glue will be inserted by +% the \cmd\baselinesklip\ mechanism. +% Can be invoked just after a \cmd\hrule\ to undo its default suppression of base line skip. +% \begin{macrocode} +\def\fullinterlineskip{\prevdepth\z@}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\count@i} +% \begin{macro}{\count@ii} +% +% \begin{macrocode} +\countdef\count@i\@ne +\countdef\count@ii\tw@ +% \end{macrocode} +% \end{macro} +% \end{macro} +% +% +% \subsection{Boolean Control}% +% We introduce just enough of the Boolean calculus for \TeX. +% Alan Jeffrey was the pioneer here, with an article in TUGboat +% (Vol. 11, No. 2, page 237). +% This implementation owes a debt to +% William Baxter (web@superscript.com). +% See articles by Baxter and Ogawa in the proceedings of the +% 1994 TUG meeting, TUGboat Vol.~15, No.~3. +% +% \begin{macro}{\prepdef} +% \begin{macro}{\appdef} +% \begin{macro}{\gappdef} +% +% Provide the capability of performing head- and tail patches. +% The procedure \cmd\prepdef\ prepends to the given macro +% the tokens specified in its second argument. +% Likewise for \cmd\appdef, except that it appends. +% Note that the first 10 toks registers are utility registers, +% and we simply make a control sequence name, \cmd\toks@ii, for one of +% them. +% \begin{macrocode} +\long\def\prepdef#1#2{% + \@ifxundefined#1{\toks@{}}{\toks@\expandafter{#1}}% + \toks@ii{#2}% + \edef#1{\the\toks@ii\the\toks@}% +}% +\long\def\appdef#1#2{% + \@ifxundefined#1{\toks@{}}{\toks@\expandafter{#1}}% + \toks@ii{#2}% + \edef#1{\the\toks@\the\toks@ii}% +}% +\long\def\gappdef#1#2{% + \@ifxundefined#1{\toks@{}}{\toks@\expandafter{#1}}% + \toks@ii{#2}% + \global\edef#1{\the\toks@\the\toks@ii}% +}% +\long\def\appdef@val#1#2{% + \appdef#1{{#2}}% +}% +\long\def\appdef@e#1#2{% + \expandafter\appdef + \expandafter#1% + \expandafter{#2}% +}% +\long\def\appdef@eval#1#2{% + \expandafter\appdef@val + \expandafter#1% + \expandafter{#2}% +}% +\toksdef\toks@ii=\tw@ +% \end{macrocode} +% \end{macro} +% \end{macro} +% \end{macro} +% +% \begin{macro}{\@ifxundefined} +% \begin{macro}{\@ifnotrelax} +% \begin{macro}{\@argswap} +% \begin{macro}{\@argswap@val} +% +% Certain utility procedures use \cmd\@ifxundefined, +% which is defined here in terms of \cmd\@ifx. +% Others use \cmd\@ifnotrelax, namely when +% the control sequence name is manufactured by +% the use of \cmd\csname. +% +% The procedures \cmd\@argswap and \cmd\@argswap@val +% are used to facilitate control of expansion. +% +% \begin{macrocode} +\long\def\@ifxundefined#1{\@ifx{\undefined#1}}% +\long\def\@ifnotrelax#1#2#3{\@ifx{\relax#1}{#3}{#2}}% +\long\def\@argswap#1#2{#2#1}% +\long\def\@argswap@val#1#2{#2{#1}}% +\def\@ifxundefined@cs#1{\expandafter\@ifx\expandafter{\csname#1\endcsname\relax}}% +% \end{macrocode} +% \end{macro} +% \end{macro} +% \end{macro} +% \end{macro} +% +% \begin{macro}{\@boolean} +% \begin{macro}{\@boole@def} +% In order to define \cmd\@ifx, we first must create the +% ``defining word'' (term taken form our Forth vocabulary) +% \cmd\@boole@def, which employs \cmd\@boolean\ to do its job. +% \begin{macrocode} +\def\@boolean#1#2{% + \long\def#1{% + #2% \if<something> + \expandafter\true@sw + \else + \expandafter\false@sw + \fi + }% +}% +\def\@boole@def#1#{\@boolean{#1}}% Implicit #2 +% \end{macrocode} +% \end{macro} +% \end{macro} +% +% \begin{macro}{\@booleantrue} +% \begin{macro}{\@booleanfalse} +% The procedures \cmd\@booleantrue\ and +% \cmd\@booleanfalse\ are assignment operators +% for Boolean flags. +% \begin{macrocode} +\def\@booleantrue#1{\let#1\true@sw}% +\def\@booleanfalse#1{\let#1\false@sw}% +% \end{macrocode} +% \end{macro} +% \end{macro} +% +% \begin{macro}{\@ifx} +% \begin{macro}{\@ifx@empty} +% \begin{macro}{\@if@empty} +% \begin{macro}{\@ifcat}% +% \begin{macro}{\@ifdim}% +% \begin{macro}{\@ifeof}% +% \begin{macro}{\@ifhbox}% +% \begin{macro}{\@ifhmode}% +% \begin{macro}{\@ifinner}% +% \begin{macro}{\@ifmmode}% +% \begin{macro}{\@ifnum}% +% \begin{macro}{\@ifodd}% +% \begin{macro}{\@ifvbox}% +% \begin{macro}{\@ifvmode}% +% \begin{macro}{\@ifvoid}% +% We can now invoke the defining word to create +% the procedures \cmd\@ifx\ and friends. +% +% Compatability Note: earlier versions of this package +% defined a procedure \cmd\@ifempty. However, for compatability with AMS\LaTeX, +% we must avoid the following three names: +% \cmd\@ifempty, \cmd\@xifempty, and \cmd\@ifnotempty. +% +% \begin{macrocode} +\@boole@def\@ifx#1{\ifx#1}% +\@boole@def\@ifx@empty#1{\ifx\@empty#1}% +\@boole@def\@if@empty#1{\if!#1!}% +%\@boole@def\@if@sw#1{\csname if#1\endcsname}% +\def\@if@sw#1#2{#1\expandafter\true@sw\else\expandafter\false@sw#2}% +\@boole@def\@ifdim#1{\ifdim#1}% +\@boole@def\@ifeof#1{\ifeof#1}% +\@boole@def\@ifhbox#1{\ifhbox#1}% +\@boole@def\@ifhmode{\ifhmode}% +\@boole@def\@ifinner{\ifinner}% +\@boole@def\@ifmmode{\ifmmode}% +\@boole@def\@ifnum#1{\ifnum#1}% +\@boole@def\@ifodd#1{\ifodd#1}% +\@boole@def\@ifvbox#1{\ifvbox#1}% +\@boole@def\@ifvmode{\ifvmode}% +\@boole@def\@ifvoid#1{\ifvoid#1}% +% \end{macrocode} +% \end{macro} +% \end{macro} +% \end{macro} +% \end{macro} +% \end{macro} +% \end{macro} +% \end{macro} +% \end{macro} +% \end{macro} +% \end{macro} +% \end{macro} +% \end{macro} +% \end{macro} +% \end{macro} +% \end{macro} +% +% \begin{macro}{\true@sw} +% \begin{macro}{\false@sw} +% +% Note that when a Boolean operator expands, it +% employs two macros that act as selectors, defined here. +% +% \begin{macrocode} +\long\def\true@sw#1#2{#1}% +\long\def\false@sw#1#2{#2}% +% \end{macrocode} +% \end{macro} +% \end{macro} +% +% \begin{macro}{\loopuntil} +% \begin{macro}{\loopwhile} +% +% Loop control using the Boolean idiom. +% Superior to \cmd\loop\dots\cmd\repeat\ because these can be nested. +% The tail of the argument must have a Boolean predicate. +% +% \begin{macrocode} +\long\def\loopuntil#1{#1{}{\loopuntil{#1}}}% +\long\def\loopwhile#1{#1{\loopwhile{#1}}{}}% +% \end{macrocode} +% \end{macro} +% \end{macro} +% +% \begin{macro}{\@provide} +% +% A defining word that refuses to clobber a prior meaning. +% +% \begin{macrocode} +\def\@provide#1{% + \@ifx{\undefined#1}{\true@sw}{\@ifx{\relax#1}{\true@sw}{\false@sw}}% + {\def#1}{\def\j@nk}% +}% +% \end{macrocode} +% \end{macro} +% +% +% \subsection{Begin Document Structure} +% The standard \LaTeX\ mechanism \cmd\AtBeginDocument\ +% is inadequate because the \cmd\vsize\ is bound much too early. +% We supply here a mechanism whereby decisions about the +% page layout can be deferred until \cmd\AtBeginDocument\ time. +% +% The problem we are working around is that the \cmd\AtBeginDocument\ +% hook in \cmd\document\ appears long after the calculation of +% \cmd\vsize\ and \cmd\hsize, that is, \LaTeX\ provides no mechanism +% for deferring the decision about the page grid until \cmd\AtBeginDocument\ time. +% We fix things by prepending a hook at the very beginning of \cmd\document. +% +% The price we pay for this facility is to depend on the stability of this part +% of \LaTeX's kernel code (the first token of \cmd\document), +% which could change, you see. But considering that \LaTeX\ is +% at this point essentially stagnant once more, we risk it. +% +% \begin{macro}{\document} +% We begin by installing hooks into \cmd\document\ that +% we will manage ourselves. +% First, we do as \cmd\document\ does: end the group +% begun by \cmd\begin. Last, we conclude our shenanigans by absorbing +% the first token of the expansion of \cmd\document, which +% we assume to be \cmd\endgroup. +% \begin{macrocode} +\prepdef\document{% + \endgroup + \init@documenthook + \set@typesize@hook + \normalsize + \set@pica@hook + \true@sw{}% +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\class@documenthook} +% \begin{macro}{\class@enddocumenthook} +% We install the first \cmd\AtBeginDocument\ hook, namely the +% procedure \cmd\class@documenthook. Within the document class, +% we will use this hook exclusively, so as to avoid interference from other packages. +% Similarly with \cmd\class@enddocumenthook, installed via \cmd\AtEndDocument. +% +% A document class using this package should do as this package does and +% just say, \cmd\appdef\cmd\class@documenthook\ and \cmd\appdef\cmd\class@enddocumenthook\ +% instead of \cmd\AtBeginDocument\ and \cmd\AtEndDocument. +% \begin{macrocode} +\def\init@documenthook{}% +\AtBeginDocument{% + \class@documenthook +}% +\AtEndDocument{% + \class@enddocumenthook +}% +\def\class@documenthook{}% +\def\class@enddocumenthook{}% +% \end{macrocode} +% \end{macro} +% \end{macro} +% +% \begin{macro}{\set@typesize@hook} +% \begin{macro}{\set@pica@hook} +% The macros \cmd\set@typesize@hook\ and \cmd\set@pica@hook\ provide everything we need. +% To use, simply \cmd\appdef\ your tokens to the appropriate hook. +% \begin{macrocode} +\def\set@typesize@hook{}% +\def\set@pica@hook{}% +% \end{macrocode} +% \end{macro} +% \end{macro} +% +% \begin{macro}{\enddocument} +% \begin{macro}{\check@aux} +% \begin{macro}{\do@check@aux} +% The standard \LaTeX\ \enve{document} processing is a potential problem, +% particularly when the output routine has been changed by \classname{ltxgrid}. +% We separate out the procedure that checks the auxiliary file at the end of +% the job so that later it can be called from the safety of the output +% routine. +% We will do this to ensure that the \cmd\@mainaux\ stream is not closed until +% the last page of the job is shipped out, and that can only be done by coordinating +% with the output routine. +% \begin{macrocode} +\def\enddocument{% + \@enddocumenthook + \@checkend{document}% + \clear@document + \check@aux + \deadcycles\z@ + \@@end +}% +\def\clear@document{\clearpage}% +\def\check@aux{\do@check@aux}% +\def\do@check@aux{% + \@if@sw\if@filesw\fi{% + \immediate\closeout\@mainaux + \let\@setckpt\@gobbletwo + \let\@newl@bel\@testdef + \@tempswafalse + \makeatletter + \input\jobname.aux\relax + }{}% + \@dofilelist + \@ifdim{\font@submax >\fontsubfuzz\relax}{% + \@font@warning{% + Size substitutions with differences\MessageBreak + up to \font@submax\space have occured.\@gobbletwo + }% + }{}% + \@defaultsubs + \@refundefined + \@if@sw\if@filesw\fi{% + \@ifx{\@multiplelabels\relax}{% + \@if@sw\if@tempswa\fi{% + \@latex@warning@no@line{% + Label(s) may have changed. + Rerun to get cross-references right + }% + }{}% + }{% + \@multiplelabels + }% + }{}% +}% +% \end{macrocode} +% \end{macro} +% \end{macro} +% \end{macro} +% +% +% \subsection{Type Tools}% +% +% \begin{macro}{\flushing} +% Undoes \cmd\centering. Should also undo \cmd\raggedleft\ and \cmd\raggedright. +% \begin{macrocode} +\def\flushing{% + \let\\\@normalcr + \leftskip\z@skip + \rightskip\z@skip + \@rightskip\z@skip + \parfillskip\@flushglue +}% +% \end{macrocode} +% \end{macro} +% +% +% \subsection{Display Math}% +% +% \begin{macro}{\eqnarray@LaTeX} +% \begin{macro}{\eqnarray@fleqn@fixed} +% Team \LaTeX\ has stated they will never repair Leslie's broken definition of \env{eqnarray}. +% Let us be bold\dots. +% +% Note on \classname{hyperref} package compatability: that package overrides +% \cmd\eqnarray\ by wrapping it up in a larger procedure, so its changes +% are compatible with this package's changes. +% +% \begin{macrocode} +\def\eqnarray@LaTeX{% + \stepcounter{equation}% + \def\@currentlabel{\p@equation\theequation}% + \global\@eqnswtrue + \m@th + \global\@eqcnt\z@ + \tabskip\@centering + \let\\\@eqncr + $$\everycr{}\halign\t@\displaywidth\bgroup + \hskip\@centering$\displaystyle\tabskip\z@skip{##}$\@eqnsel + &\global\@eqcnt\@ne\hskip \tw@\arraycolsep \hfil${##}$\hfil + &\global\@eqcnt\tw@ \hskip \tw@\arraycolsep + $\displaystyle{##}$\hfil\tabskip\@centering + &\global\@eqcnt\thr@@ \hb@xt@\z@\bgroup\hss##\egroup + \tabskip\z@skip + \cr +} +\long\def\eqnarray@fleqn@fixed{% + \stepcounter{equation}\def\@currentlabel{\p@equation\theequation}% + \global\@eqnswtrue\m@th\global\@eqcnt\z@ + \tabskip\mathindent + \let\\=\@eqncr + \setlength\abovedisplayskip{\topsep}% + \ifvmode\addtolength\abovedisplayskip{\partopsep}\fi + \addtolength\abovedisplayskip{\parskip}% + \setlength\belowdisplayskip{\abovedisplayskip}% + \setlength\belowdisplayshortskip{\abovedisplayskip}% + \setlength\abovedisplayshortskip{\abovedisplayskip}% + $$% + \everycr{}% + \halignt@\linewidth\bgroup + \hskip\@centering$\displaystyle\tabskip\z@skip{##}$\@eqnsel + &\global\@eqcnt\@ne + \hskip\tw@\eqncolsep + \hfil${{}##{}}$\hfil + &\global\@eqcnt\tw@ + \hskip\tw@\eqncolsep + $\displaystyle{##}$\hfil\tabskip\@centering + &\global\@eqcnt\thr@@\hb@xt@\z@\bgroup\hss##\egroup + \tabskip\z@skip + \cr +}% +\@ifx{\eqnarray\eqnarray@LaTeX}{% + \class@info{Repairing broken LaTeX eqnarray}% + \let\eqnarray\eqnarray@fleqn@fixed + \newlength\eqncolsep + \setlength\eqncolsep\z@ + \let\eqnarray@LaTeX\relax + \let\eqnarray@fleqn@fixed\relax +}{}% +\def\mathindent{\@centering}% +\def\set@eqnarray@skips{}% +% \end{macrocode} +% \end{macro} +% \end{macro} +% +% +% \subsection{Footnotes} +% +% \changes{v4.0beta 4}{2000/04/10} +% {New kernel fix. For bug 174} +% +% \begin{macro}{\footnote} +% \begin{macro}{\footnotemark} +% \begin{macro}{\@xfootnote} +% \begin{macro}{\@xfootnotemark} +% \begin{macro}{\@yfootnote} +% We repair an error in the \LaTeX\ kernel (see \file{ltfloat.dtx}) involving footnotes. +% The symptom is that the \cmd\footnotemark\ command does not work properly within a \env{minipage} environment. +% The source of the problem is in the way the \cmd\footnotemark\ and \cmd\@xfootnotemark\ procedures are defined: +% they do not share the method used by the \cmd\footnote\ and other procedures that allows a context switch to +% change the way footnotes behave within a minipage environment. +% This is a \LaTeX\ bug of long standing; this fix dates to 1987. +% +% While we are at it, we rewrite both the \cmd\footnote\ and \cmd\footnotemark\ procedures, +% achieving a slightly cleaner separation of syntax and semantics. +% Note that the \cmd\@footnotemark\ and \cmd\@footnotetext\ procedures are not altered here; +% they continue as the methods of formatting the footnote mark and footnote text, respectively. +% +% A note about the context switch mentioned above: +% the \env{minipage} environment executes the following in order to alter the way footnotes +% behave: +%\begin{verbatim} +%\def\@mpfn{mpfootnote}% +%\def\thempfn{\thempfootnote}% +%\let\@footnotetext\@mpfootnotetext +%\c@mpfootnote\z@ +%\end{verbatim} +% This code changes the counter used in autonumbered footnotes, the choice of footnote marker, +% and the procedure used on the footnote text. Changing the counter is needed because minipage +% footnotes are in their own sequence, and the footnote marker is customarily different within +% a minipage. The procedure that works on the footnote text must be different because the footnotes +% are placed at the bottom of the minipage, not the bottom of the text column. +% +% Any procedure that establishes a minipage-like context (e.g., floats) can do the same. +% \begin{macrocode} +\def\footnote{% + \@ifnextchar[\@xfootnote{\@yfootnote\@footnotetext}% +}% +\def\footnotemark{% + \@ifnextchar[\@xfootnotemark{\@yfootnote}% +}% +\def\@xfootnote[#1]{% + \@xfootnotemark[#1]% + \@footnotetext +}% +\def\@xfootnotemark@ltx[#1]{% + \begingroup + \csname c@\@mpfn\endcsname #1\relax + \unrestored@protected@xdef\@thefnmark{\thempfn}% + \endgroup + \H@@footnotemark +}% +\def\@yfootnote{% + \stepcounter\@mpfn + \protected@xdef\@thefnmark{\thempfn}% + \H@@footnotemark +}% +% \end{macrocode} +% +% Note on \classname{hyperref} compatability: +% In its ``Automated \LaTeX\ hypertext cross-references'', +% the \classname{hyperref} package alters footnote processing, +% thereby imperiling these fixes and necessiating defensive measures. +% +% The main thing \classname{hyperref} does is to take over the \cmd\@mpfootnotetext\ and +% \cmd\@footnotetext\ procedures, feeding its own arguments to these macros. +% It also rewrites \cmd\@footnotemark, making it a hyperlink. +% +% But at the same time, it attempts to turn off these changes during +% \cmd\maketitle\ processing, necessitating rewriting \cmd\@xfootnotemark. +% At this point it is on the slippery slope. +% +% We make ourself \classname{hyperref} friendly: +% we give \classname{hyperref} what it needs, but avoid its change to +% \cmd\@xfootnotemark. +% +% Any other package that rewrites \LaTeX's footnote macros will be incompatible +% with this package. +% \begin{macrocode} +\appdef\class@documenthook{% + \@ifxundefined\H@@footnotemark{% + \let\H@@footnotemark\@footnotemark + }{}% + \let\@xfootnotemark\@xfootnotemark@ltx +}% +% \end{macrocode} +% +% Two thoughts about \classname{hyperref}: what for does it define \cmd\realfootnote? +% Also: a document class that desires high hypertext capabilities might +% well wish to reimplement \cmd\maketitle\ so that footnotes called out from there +% are hypertext links: the \classname{hyperref} package's +% ``Automated \LaTeX\ hypertext cross-references'' +% does not do any of this: +%\begin{quotation} +% But the special footnotes +% in |\maketitle| are much too hard to deal with +% properly. Let them revert to plain behaviour. +%\end{quotation} +% Note that the document class, in reimplementing \cmd\maketitle, must ensure +% that the \classname{hyperref} package does not clobber its own definition! +% +% \end{macro} +% \end{macro} +% \end{macro} +% \end{macro} +% \end{macro} +% +% \begin{macro}{\@footnotetext} +% \begin{macro}{\@mpfootnotetext} +% \begin{macro}{\@tpfootnotetext} +% \begin{macro}{\make@footnotetext} +% \begin{macro}{\set@footnotewidth} +% The two procedures \cmd\@footnotetext\ and \cmd\@mpfootnotetext\ share code. +% We make that explicit here. +% +% Note that the procedure calling \cmd\make@footnotetext\ will open a group +% with \cmd\bgroup\ which is then closed by \cmd\minipagefootnote@drop. +% +% Difference from \LaTeX: here we do not set \cmd\floatingpenalty\ to infinity. +% Doing this must date back to a time when \LaTeX\ could not accomodate split insertions (footnotes). +% I cannot think of any other reason to do have done this. +% At any rate, with the \classname{ltxgrid} package, split insertions are specifically properly +% taken care of, so we allow it. +% +% We provide the hook \cmd\set@footnotewidth\ +% that sets the footnote on a particular measure. +% Some page grids are such as to set a footnote in a context where \cmd\columnwidth +% is not the right parameter to use for the set width of a footnote. +% In such a case, for the applicable scope, you should define +% \cmd\set@footnotewidth\ to perform this job correctly. +% +% A procedure, \cmd\set@footnotewidth@ii, illustrates how to do this when in a two-column page grid. +% In general, remember that footnotes, like all insertions (including floats), +% are a step outside of the galley context, and all aspects of insertions need to be +% properly handled, including the set width. +% \begin{macrocode} +\long\def\@footnotetext{% + \insert\footins\bgroup + \make@footnotetext +}% +\long\def\@mpfootnotetext{% + \minipagefootnote@pick + \make@footnotetext +}% +\def\make@footnotetext#1{% + \reset@font\footnotesize + \interlinepenalty\interfootnotelinepenalty + \splittopskip\footnotesep + \splitmaxdepth\dp\strutbox +% \floatingpenalty\@MM + \set@footnotewidth + \@parboxrestore + \protected@edef\@currentlabel{% + \csname p@footnote\endcsname\@thefnmark + }% + \color@begingroup + \@makefntext{% + \rule\z@\footnotesep\ignorespaces#1\@finalstrut\strutbox + }% + \color@endgroup + \minipagefootnote@drop +}% +\def\set@footnotewidth{% + \hsize\columnwidth + \linewidth\hsize +}% +\def\set@footnotewidth@ii{% + \hsize\textwidth + \advance\hsize\columnsep + \divide\hsize\tw@ + \advance\hsize-\columnsep + \linewidth\hsize +}% +% \end{macrocode} +% \end{macro} +% \end{macro} +% \end{macro} +% \end{macro} +% \end{macro} +% +% +% \subsection{Floats} +% +% \subsubsection{Usage notes}% +% We extend the \LaTeX\ kernel for three purposes: +% \begin{enumerate} +% +% \item +% When the \cmd\footnote\ command is used within the +% scope of a float, we do as \env{minipage} does. +% +% \item +% We provide a mechanism to write floats out to an external +% stream for temporary storage (deferred floats). +% +% \item +% We provide mechanism for placing a float \texttt{here} +% invariably, that is, floats are unfloated. +% This mechanism is used to read the external stream mentioned above. +% +% \end{enumerate} +% +% To use these mechanisms, the document class should +% define a float, say, \env{figure} as per usual, and in addition: +% \begin{enumerate} +% +% \item +% Optionally define an alternative, say \env{figure@write} as follows: +% \begin{verbatim} +%\newenvironment{figure@write}{% +% \write@float{figure}% +%}{% +% \endwrite@float +%}% +% \end{verbatim} +% That is, the alternative environment executes \cmd\write@float\ +% instead of \cmd\@float. +% Note that this step is not needed if the float environment +% is defined in the simple way of \file{classes.dtx}. +% However, an environment like \env{longtable} will require it. +% +% \item +% Install into \cmd\AtBeginDocument\ a call to \cmd\do@if@floats, +% with the float name and an appropriate file extension as its arguments. +% \begin{verbatim} +%\AtBeginDocument{\do@if@floats{figure}{.fgx}}% +% \end{verbatim} +% +% \item +% Optionally define a text entity \cmd\figuresname\ that will +% be the text of the head that is set over the +% deferred floats. +% If not defined, there will be no head. +% +% \item +% Optionally define a user-level command to allow +% the document to determine where the figures are printed out +% (default is to print at end of document). E.g., +% \begin{verbatim} +%\newcommand\printfigures{\print@float{figure}}% +% \end{verbatim} +% \item +% Install into \cmd\appdef\cmd\class@enddocumenthook\ a call to \cmd\printfigures, +% or, if the latter is not defined, as follows: +% \begin{verbatim} +%\appdef\class@enddocumenthook{\print@float{figure}}% +% \end{verbatim} +% Note that installing this command into \cmd\AtBeginDocument +% is best done earlier than calls that assume the last page of +% the document is at hand. +% +% \end{enumerate} +% +% \subsubsection{Robustifying fragile commands}% +% Certain of \LaTeX's commands cannot be written out to a file or appear within a \cmd\mark\ command argument +% because they do calculations during expansion. +% We provide for a little help, but without changing the meanings of these commands. +% +% \begin{macro}{\addtocontents} +% +% \begin{macrocode} +\def\addtocontents#1#2{% + \protected@write\@auxout{% + \let \label \@gobble \let \index \@gobble \let \glossary \@gobble + \def\({\string\(}% + \def\){\string\)}% + \def\\{\string\\}% + }{\string \@writefile {#1}{#2}}% +}% +% \end{macrocode} +% +% \end{macro} +% +% \subsubsection{Preparing for the \classname{hyperref package}}% +% +% \begin{macro}{\addcontentsline} +% \begin{macro}{\contentsline} +% The \classname{hyperref} package assumes that the \cmd\contentsline\ command will be given four arguments. +% Therefore it cannot successfully process a \filename{.toc} file that had been written by standard \LaTeX. +% We fix things up by always writing that fourth argument and by supplying a \cmd\contentsline\ command that +% can read them. +% +% We also give the \cmd\newlabel\ command's second argument five tokens. +% +% This means that a document class that uses this package will itself have trouble taking over a \filename{.toc} file that was written by standard \LaTeX. Sigh. +% \begin{macrocode} +\def\addcontentsline#1#2#3{% + \addtocontents{#1}{% + \protect\contentsline{#2}{#3}{\thepage}{}% + }% +}% +\def\contentsline#1#2#3#4{% + \csname l@#1\endcsname{#2}{#3}% +}% +\def\label#1{% + \@bsphack + \protected@write\@auxout{}{% + \string\newlabel{#1}{{\@currentlabel}{\thepage}{}{}{}}% + }% + \@esphack +}% +% \end{macrocode} +% +% \end{macro} +% \end{macro} +% +% \subsubsection{Footnotes within floats, unfloating floats, float font}% +% +% \begin{macro}{\caption} +% DPC: Er a bit of a hack, but seems best way of supporting normal +% \LaTeX\ syntax at this point: If a caption is used below a table, +% then put out the footnotes before the caption. +% \changes{v4.0beta 2}{1999/06/20} +% {Support the hack with \cs{prepdef}, and delay until \cs{AtBeginDocument} time, +% since \classname{hyperref} clobbers \cs{caption}.} +% \begin{macrocode} +\appdef\class@documenthook{% + \prepdef\caption{\minipagefootnote@here}% +}% +% \end{macrocode} +% +% Note on \classname{hyperref} compatability: +% this change to the \cmd\caption\ command is compatible with the +% ``Automated \LaTeX\ hypertext cross-references'' patches of +% that package. +% +% All the same, I think Sebastian's changes to \cmd\caption\ and \cmd\@caption\ +% could bear with some improvement. +% The following implementation requires knowing only the pattern part of the +% \cmd\@caption\ macro: +%\begin{verbatim} +%\def\caption{% +% \H@refstepcounter\@captype +% \hyper@makecurrent{\@captype}% +% \@dblarg{\H@caption\@captype}% +%}% +%\def\H@caption#1[#2]#3{% +% \@caption{#1}[#2]{% +% \ifHy@nesting +% \hyper@@anchor{\@currentHref}{#3}% +% \else +% \hyper@@anchor{\@currentHref}{\relax}#3% +% \fi +% }% +%}% +% +%\end{verbatim} +% +% \end{macro} +% +% \begin{macro}{\minipagefootnote@init} +% \begin{macro}{\minipagefootnote@here} +% \begin{macro}{\minipagefootnote@foot} +% \begin{macro}{\minipagefootnote@pick} +% \begin{macro}{\minipagefootnote@drop} +% Procedure to deal with footnotes accumulated within a minipage environment. +% These procedures encapsulate all uses of the \cmd\@mpfootins\ box. +% +% +% Note: \cmd\minipagefootnote@here\ must \emph{not} be executed within the MVL! +% \begin{macrocode} +\def\minipagefootnote@init{% + \setbox\@mpfootins\box\voidb@x +}% +\def\minipagefootnote@pick{% + \global\setbox\@mpfootins\vbox\bgroup + \unvbox\@mpfootins +}% +\def\minipagefootnote@drop{% + \egroup +}% +\def\minipagefootnote@here{% + \par + \@ifvoid\@mpfootins{}{% + \vskip\skip\@mpfootins + \fullinterlineskip + \@ifinner{% + \vtop{\unvcopy\@mpfootins}% + {\setbox\z@\lastbox}% + }{}% + \unvbox\@mpfootins + }% +}% +\def\minipagefootnote@foot{% + \@ifvoid\@mpfootins{}{% + \insert\footins\bgroup\unvbox\@mpfootins\egroup + }% +}% +\def\endminipage{% + \par + \unskip + \minipagefootnote@here + \@minipagefalse %% added 24 May 89 + \color@endgroup + \egroup + \expandafter\@iiiparbox\@mpargs{\unvbox\@tempboxa}% +}% +% \end{macrocode} +% \end{macro} +% \end{macro} +% \end{macro} +% \end{macro} +% \end{macro} +% +% \begin{macro}{\floats@sw} +% The Boolean \cmd\floats@sw\ signifies that floats are to be floated; +% if false, that floats are to be deferred to the end of the document. +% Note that the state of this Boolean is to be changed by +% the document class in response to user-selected options. +% Here we display model code that assigns a default value +% at \cmd\AtBeginDocument\ time. +% \begin{verbatim} +%\AtBeginDocument{% +% \@ifxundefined\floats@sw{\@booleantrue\floats@sw}{}% +%}% +% \end{verbatim} +% \end{macro} +% +% \begin{macro}{\@xfloat} +% \begin{macro}{\@mpmakefntext} +% The float start-code is redefined to set up footnotes in the style of minipage. +% Also, the \cmd\floats@sw\ Boolean informs us that floats are to be +% all placed \texttt{here}. +% Note that, to protect against the Boolean being undefined +% at this late hour, we default it globally to true. +% +% \changes{v4.0beta 2}{1999/06/20} +% {AO: Removed superfluous \cs{def}s, changed to using \cs{floats@sw} as the flag. +% Also stopped using DPC's \cs{if@twocolumn} flag: using \cs{floats@sw} instead. +% Also added \cs{par}\cs{vskip}\cs{z@skip} after the \cs{minipagefootnotes} +% so that the float box would have zero depth like the kernel one. +% } +% \changes{v4.0beta 3}{1999/11/13} +% {bug fix 110. Install hooks for endfloats processing} +% \changes{v4.0beta 4}{2000/04/10} +% {bug fix 127. Floats placed [h] to allow page breaks} +% \changes{v4.0beta 4}{2000/05/19} +% {bug fix 224. Hyperref compatability.} +% \changes{v4.0beta 5}{2000/11/16} +% {bug fix 221. Remove samepage command from @xfloat@prep: If the float can break over pages, we want better control.} +%FIXME: why does hyperref override \cmd\@xfootnotenext? +% +% \begin{macrocode} +\let\@xfloat@LaTeX\@xfloat +\def\@xfloat#1[#2]{% + \@xfloat@prep + \@nameuse{fp@proc@#2}% + \@ifxundefined\floats@sw{\global\@booleantrue\floats@sw}{}% + \floats@sw{\@xfloat@LaTeX{#1}[#2]}{\@xfloat@anchored{#1}[]}% +}% +\def\@xfloat@prep{% + \let\footnote\footnote@latex + \def\@mpfn{mpfootnote}% + \def\thempfn{\thempfootnote}% +% \def\thefootnote{\thempfootnote}% + \c@mpfootnote\z@ + \let\@footnotetext\@mpfootnotetext + \let\H@@footnotetext\@mpfootnotetext + \let\@makefntext\@mpmakefntext +% \samepage +}% +\appdef\class@documenthook{% + \let\footnote@latex\footnote +}% +%\def\fp@proc@h{\@booleanfalse\floats@sw}% +%\def\fp@proc@H{\@booleanfalse\floats@sw}% +\def\@xfloat@anchored#1[#2]{% + \def\@captype{#1}% + \begin@float@pagebreak +%\vbox\bgroup + \let\end@float\end@float@anchored + \let\end@dblfloat\end@float@anchored +% do as \@xfloat does: + \hsize\columnwidth + \@parboxrestore + \@floatboxreset + \minipagefootnote@init +% \pagegrid@col\@ne % Klootch to avoid processing as a float +}% +\def\end@float@anchored{% + \minipagefootnote@here + \par\vskip\z@skip %% \par\vskip\z@ added 15 Dec 87 +%\egroup + \par + \end@float@pagebreak +}% +\def\begin@float@pagebreak{\par\addvspace\intextsep}% +\def\end@float@pagebreak{\par\addvspace\intextsep}% +\def\@mpmakefntext#1{% + \parindent=1em + \noindent + \hb@xt@1em{\hss\@makefnmark}% + #1% +}% +% \end{macrocode} +% \end{macro} +% \end{macro} +% +% +% \subsubsection{Writing floats out to a file}% +% +% \begin{macro}{\do@if@floats} +% The procedure \cmd\do@if@floats\ should be executed at +% \cmd\AtBeginDocument\ time, and arranges to write out +% the floats of the given class to a temporary file, to be +% read back later (deferred floats), +% given that \cmd\floats@sw\ is false. +% Note that, to protect against the Boolean being undefined +% at this late hour, we default it globally to true. +% \begin{macrocode} +\def\do@if@floats#1#2{% + \@ifxundefined\floats@sw{\global\@booleantrue\floats@sw}{}% + \floats@sw{}{% +% \end{macrocode} +% Open the stream to save out the document's floats of this class. +% \begin{macrocode} + \expandafter\newwrite + \csname#1write\endcsname + \expandafter\def + \csname#1@stream\endcsname{\jobname#2}% + \expandafter\immediate + \expandafter\openout + \csname#1write\endcsname + \csname#1@stream\endcsname\relax +% \end{macrocode} +% Swap environments. +% If the class writer has defined, e.g., \env{figure@write}, +% then we use this as the procedure to execute for writing +% the float out to the external stream. +% Otherwise, the replacement of \cmd\@float\ by \cmd\write@float\ +% should do the right thing for float environments defined +% in the simple way of \classname{classes.dtx}. +% \begin{macrocode} + \@ifxundefined\@float@LaTeX{% + \let\@float@LaTeX\@float + \let\@dblfloat@LaTeX\@dblfloat + \let\@float\write@float + \let\@dblfloat\write@floats + }{}% + \let@environment{#1@float}{#1}% + \let@environment{#1@floats}{#1*}% + \@ifxundefined@cs{#1@write}{}{% + \let@environment{#1}{#1@write}% + }% + }% +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\print@float} +% The procedure \cmd\print@float\ prints out the +% deferred floats. +% \changes{v4.0beta 2}{1999/06/20} +% {only execute if there really were floats of the given type} +% \changes{v4.0beta 3}{1999/11/13} +% {*-form mandates pagebreak at each float; +% only print section head if there is something there.} +% \changes{v4.0beta 4}{2000/05/23} +% {Allow things to break over pages by setting array@default.} +% +% \begin{macrocode} +\def\triggerpar{\leavevmode\@@par}% +\def\oneapage{\def\begin@float@pagebreak{\newpage}\def\end@float@pagebreak{\newpage}}% +\def\print@float#1#2{% + \@ifxundefined@cs{#1write}{}{% + \begingroup + \@booleanfalse\floats@sw + #2% + \raggedbottom + \def\array@default{v}% floats must + \let\@float\@float@LaTeX + \let\@dblfloat\@dblfloat@LaTeX + \let\trigger@float@par\triggerpar + \let@environment{#1}{#1@float}% + \let@environment{#1*}{#1@floats}% + \expandafter\prepdef\csname#1\endcsname{\trigger@float@par}% + \expandafter\prepdef\csname#1*\endcsname{\trigger@float@par}% + \@namedef{fps@#1}{h!}% + \expandafter\immediate + \expandafter\closeout + \csname#1write\endcsname + \everypar{% + \global\let\trigger@float@par\relax + \global\everypar{}\setbox\z@\lastbox + \@ifxundefined@cs{#1sname}{}{% + \begin@float@pagebreak + \expandafter\section + \expandafter*% + \expandafter{% + \csname#1sname\endcsname + }% + }% + }% + \input{\csname#1@stream\endcsname}% + \endgroup + \global\expandafter\let\csname#1write\endcsname\relax + }% +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\write@float} +% \begin{macro}{\write@floats} +% \begin{macro}{\write@@float} +% Handles the case where the name of the float is the same as +% that of the stream. Note that +% \env{longtable} does \emph{not} fit this case. +% Note also: \cmd\write@float\ is \emph{not} a user-level environment +% therefore it is properly not defined with \cmd\newenvironment. +% \begin{macrocode} +\def\write@float#1{\write@@float{#1}{#1}}% +\def\endwrite@float{\@Esphack}% +\def\write@floats#1{\write@@float{#1*}{#1}}% +\def\endwrite@floats{\@Esphack}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\write@@float} +% \changes{v4.0beta 2}{1999/06/20} +% {AO: Fixed spurious \texttt{CR} and (return) characters in output file. +% Also, if the document did not have the \cs{end}\texttt{figure} on a line of its own, +% the macro wouldn't work. Fixed.} +% \begin{macrocode} +\def\write@@float#1#2{% + \ifhmode + \@bsphack + \fi + \chardef\@tempc\csname#2write\endcsname + \toks@{\begin{#1}}% + \def\@tempb{#1}% + \expandafter\let\csname end#1\endcsname\endwrite@float + \catcode`\^^M\active + \@makeother\{\@makeother\}\@makeother\% + \write@floatline +}% +% \end{macrocode} +% \end{macro} +% \end{macro} +% \end{macro} +% +% \begin{macro}{\write@floatline} +% \begin{macro}{\@write@floatline} +% \begin{macro}{\float@end@tag} +% The procedure \cmd\write@floatline\ only parses, and passes +% its result to \cmd\@write@floatline, which +% writes the line to output, then tests the line +% for the \cmd\end\arg{float} tokens with +% aid of the \cmd\float@end@tag\ procedure. +% \begin{macrocode} +\begingroup + \catcode`\[\the\catcode`\{\catcode`\]\the\catcode`\}\@makeother\{\@makeother\}% + \gdef\float@end@tag#1\end{#2}#3\@nul[% + \def\@tempa[#2]% + \@ifx[\@tempa\@tempb][\end[#2]][\write@floatline]% + ]% + \obeylines% + \gdef\write@floatline#1^^M[% + \begingroup% + \newlinechar`\^^M% + \toks@\expandafter[\the\toks@#1]\immediate\write\@tempc[\the\toks@]% + \endgroup% + \toks@[]% + \float@end@tag#1\end{}\@nul% + ]% +\endgroup +% \end{macrocode} +% \end{macro} +% \end{macro} +% \end{macro} +% +% +% +% \subsection{Counters} +% The following definitions override those of the \LaTeX\ kernel, +% providing for a greater range of inputs. +% \begin{macrocode} +\def\@alph#1{\ifcase#1\or a\or b\or c\or d\else\@ialph{#1}\fi} +% \end{macrocode} +% +% \begin{macrocode} +\def\@ialph#1{\ifcase#1\or \or \or \or \or e\or f\or g\or h\or i\or j\or + k\or l\or m\or n\or o\or p\or q\or r\or s\or t\or u\or v\or w\or x\or + y\or z\or aa\or bb\or cc\or dd\or ee\or ff\or gg\or hh\or ii\or jj\or + kk\or ll\or mm\or nn\or oo\or pp\or qq\or rr\or ss\or tt\or uu\or + vv\or ww\or xx\or yy\or zz\else\@ctrerr\fi} +% \end{macrocode} +% +% +% \subsection{Customization of Sections}% +% +% Patch the standard \LaTeX\ sectioning procedure to: +%\begin{itemize} +%\item +% Allow a sectioning command to trigger the title page, or more generally +% to recognize that it is the first object in the document, +% so we headpatch \cmd\@startsection. +% +%\item +% Allow a tail command in |#6| to uppercase the title, so we retain +% DPC's braces. +% +%\item +% Allow each type of sectioning command to format its number differently, +% so we generalize \cmd\@seccntformat. +% +%\item +% Allow each type of sectioning command to format its argument differently, +% so we generalize \cmd\@hangfrom. +% +%\item +% Allow the starred form of the command to +% mark (the running head) and +% make an entry in the TOC, +% so we put \cmd\@ssect\ on the same footing as \cmd\@sect. +% +% Note that the tokens passed to the TOC now are \emph{not} +% the optional argument of the command, but the required. +% This means that the user can no longer use the former +% to put variant content in to the TOC as the Manual says. +% +% Instead, the optional argument is used to put an alternative +% title into the running headers, a better choice. +% +%\end{itemize} +% +% \begin{macro}{\@startsection} +% Patch a head hook into the basic sectioning command. +% Treat \cmd\@sect\ and \cmd\@ssect\ on an equal footing: +% now their pattern parts are identical. +% \begin{macrocode} +\def\@startsection#1#2#3#4#5#6{% + \@startsection@hook + \if@noskipsec \leavevmode \fi + \par + \@tempskipa #4\relax + \@afterindenttrue + \ifdim \@tempskipa <\z@ + \@tempskipa -\@tempskipa \@afterindentfalse + \fi + \if@nobreak + \everypar{}% + \else + \addpenalty\@secpenalty\addvspace\@tempskipa + \fi + \@ifstar + {\@dblarg{\@ssect@ltx{#1}{#2}{#3}{#4}{#5}{#6}}}% + {\@dblarg{\@sect@ltx {#1}{#2}{#3}{#4}{#5}{#6}}}% +}% +\def\@startsection@hook{}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\@sect} +% When defining \cmd\@svsec, do not expand \cmd\@seccntformat. +% Put brace characters back where they were before David Carlisle got at them +% (i.e., as if \cmd\@hangfrom\ had two arguments). +% Protect the mark mechanism from an undefined meaning. +% Pass |#8| to the TOC instead of |#7|. +% Remove \cmd\relax\ from the replacement part of \cmd\@svsec. +% +% The procedure \cmd\@hangfrom\ and \cmd\@runin@to\ can +% be used to process the argument of the head. +% The head can define, e.g., \cmd\@hangfrom@section, to +% do its own processing. +% +% In using \cmd\H@refstepcounter\ in place of \cmd\refstepcounter\ we rely on +% either loading before any package that patches the latter, or +% the convention that the former is the original \LaTeX\ procedure. +% +% \begin{macrocode} +\class@info + {Repairing broken LateX \string\@sect}% +\def\@sect@ltx#1#2#3#4#5#6[#7]#8{% + \@ifnum{#2>\c@secnumdepth}{% + \def\H@svsec{\phantomsection}% + \let\@svsec\@empty + }{% + \H@refstepcounter{#1}% + \def\H@svsec{% + \phantomsection + }% + \protected@edef\@svsec{{#1}}% + \@ifundefined{@#1cntformat}{% + \prepdef\@svsec\@seccntformat + }{% + \expandafter\prepdef + \expandafter\@svsec + \csname @#1cntformat\endcsname + }% + }% + \@tempskipa #5\relax + \@ifdim{\@tempskipa>\z@}{% + \begingroup + \interlinepenalty \@M + #6{% + \@ifundefined{@hangfrom@#1}{\@hang@from}{\csname @hangfrom@#1\endcsname}% + {\hskip#3\relax\H@svsec}{\@svsec}{#8}% + }% + \@@par + \endgroup + \@ifundefined{#1mark}{\@gobble}{\csname #1mark\endcsname}{#7}% + \addcontentsline{toc}{#1}{% + \@ifnum{#2>\c@secnumdepth}{% + \protect\numberline{}% + }{% + \protect\numberline{\csname the#1\endcsname}% + }% + #8}% + }{% + \def\@svsechd{% + #6{% + \@ifundefined{@runin@to@#1}{\@runin@to}{\csname @runin@to@#1\endcsname}% + {\hskip#3\relax\H@svsec}{\@svsec}{#8}% + }% + \@ifundefined{#1mark}{\@gobble}{\csname #1mark\endcsname}{#7}% + \addcontentsline{toc}{#1}{% + \@ifnum{#2>\c@secnumdepth}{% + \protect\numberline{}% + }{% + \protect\numberline{\csname the#1\endcsname}% + }% + #8}% + }% + }% + \@xsect{#5}% +}% +\def\@hang@from#1#2#3{\@hangfrom{#1#2}#3}% +\def\@runin@to #1#2#3{#1#2#3}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\@ssect} +% Put brace characters back where they were before David Carlisle got at them +% (as if \cmd\@hangfrom\ has two arguments). +% Possibly set a mark. +% Make a TOC entry. +% +% Note that, for compatability with the \classname{hyperref} package, we +% need to provide the interface required by that package +% (actually required by \file{pdfmark.def} and \file{nameref.sty}), +% namely +% the definition of \cmd\@currentlabelname\ (but now removed), +% the insertion of the procedure \cmd\Sectionformat\ (but why is this needed?), and +% the call to \cmd\phantomsection\ (which must precede the call to \cmd\addcontentsline). +% We also have to sidestep the patch to \cmd\@ssect\ in that same file, therefore +% we use a different control sequence name in the call from \cmd\@startsection. +% \changes{v4.0beta 3}{1999/11/13} +% {Bug 116: Hyperref compatability} +% \changes{v4.0rc3b}{2001/07/13} +% {Bug 404: Hyperref compatability} +% \begin{macrocode} +\def\@ssect@ltx#1#2#3#4#5#6[#7]#8{% +% \def\@currentlabelname{#8}% + \def\H@svsec{\phantomsection}% + \@tempskipa #5\relax + \@ifdim{\@tempskipa>\z@}{% + \begingroup + \interlinepenalty \@M + #6{% + \@ifundefined{@hangfroms@#1}{\@hang@froms}{\csname @hangfroms@#1\endcsname}% +% {\hskip#3\relax\H@svsec}{\Sectionformat{#8}{#1}}% + {\hskip#3\relax\H@svsec}{#8}% + }% + \@@par + \endgroup + \@ifundefined{#1smark}{\@gobble}{\csname #1smark\endcsname}{#7}% + \addcontentsline{toc}{#1}{\protect\numberline{}#8}% + }{% + \def\@svsechd{% + #6{% + \@ifundefined{@runin@tos@#1}{\@runin@tos}{\csname @runin@tos@#1\endcsname}% +% {\hskip#3\relax\H@svsec}{\Sectionformat{#8}{#1}}% + {\hskip#3\relax\H@svsec}{#8}% + }% + \@ifundefined{#1smark}{\@gobble}{\csname #1smark\endcsname}{#7}% + \addcontentsline{toc}{#1}{\protect\numberline{}#8}% + }% + }% + \@xsect{#5}% +}% +\def\@hang@froms#1#2{#1#2}% +\def\@runin@tos #1#2{#1#2}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\init@documenthook} +% Document classes that incorporate this package will be \classname{hyperref}-savvy. +% (To accomplish this, we ensure that \cmd\hyperanchor\ and \cmd\hyper@last\ are both defined.) +% Being \classname{hyperref}-savvy levels some requirements on us, but the benefits are many. +% +% One is that the TOC will not get amnesia and require a full set of three typesetting runs before its formatting is stable. +% Instead, only two runs are required: the first updates the auxiliary file, the second the TOC. +% However, the formatting of the document does not change. +% +% Another aspect of being \classname{hyperref}-savvy is that the syntax of commands in the \filename{.aux} file will now change +% if \classname{hyperref} is turned on or off. +% +% Note that \cmd\hyper@anchorstart\ and \cmd\hyper@anchorend\ constitute the programming interface +% for a hypertext anchor (the target of a hypertext link); \cmd\hyper@linkstart\ and \cmd\hyper@linkend\ +% are the interface for a hypertext link. +% \begin{macrocode} +\appdef\init@documenthook{% + \providecommand\phantomsection{}% +%\@ifx{\Sectionformat\@undefined}{\let\Sectionformat\@firstoftwo}{}% + \providecommand\hyper@anchor[1]{}% + \providecommand\hyper@last{}% + \providecommand\Hy@raisedlink[1]{#1}% + \providecommand\hyper@anchorstart[1]{}% + \providecommand\hyper@anchorend{}% + \providecommand\hyper@linkstart[2]{}% + \providecommand\hyper@linkend{}% +}% +\let\H@refstepcounter\refstepcounter +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\sec@upcase} +% Upper case for sections (optional upper case items). These are +% created so that some headings can be toggled between mixed case and +% upper case readily. +% Headings that might be changed can be wrapped in the style file in +% \cmd\sec@upcase\arg{text} constructs; +% the expansion of \cmd\sec@upcase\ is +% controlled here. It is \cmd\relax\ by default (mixed case heads), and +% can easily be changed to \cmd\uppercase\ if desired. +% If mixed-case headings are wanted by the editor, authors {\em must} +% supply mixed case text, although this is what authors should be doing +% anyway. +% (Mixed can be converted to upper, +% but the reverse transformation cannot be automated.) +% +% The following setting gives the \LaTeX\ default. +% \begin{macrocode} +\def\sec@upcase#1{\relax{#1}}% +% \end{macrocode} +% \end{macro} +% +% +% \subsection{Patch the \env{tabular} and \env{array} Environments} +% +% \begin{macro}{\endtabular} +% \begin{macro}{\endarray} +% We headpatch the begin processing and tailpatch the end processing +% of the \env{tabular} and \env{array} environments. +% A document class can define these hooks as needed. +% +% We proceed with care to make further patches to +% support tabulars that break over pages. +% Our patches will not necessarily be effective for +% other packages that replace the \LaTeX\ \env{array} and \env{tabular} +% environments. I know of none that do so. +% \begin{macrocode} +\appdef\class@documenthook{% + \@ifpackageloaded{array}{\switch@array}{\switch@tabular}% + \prepdef\endtabular{\endtabular@hook}% + \@provide\endtabular@hook{}% + \prepdef\endarray{\endarray@hook}% + \@provide\endarray@hook{}% + \providecommand\array@hook{}% +% \end{macrocode} +% Install, effectively, a head patch to \cmd\tabular. +% In order to avoid interference from, e.g., the \classname{array} package, +% we must perform this patch only \emph{after} packages load. +% \changes{v4.0beta 3}{1999/11/13} +% {Bug 130. Interference from array package} +% \begin{macrocode} + \prepdef\@tabular{\tabular@hook}% + \@provide\tabular@hook{}% +}% +% \end{macrocode} +% \end{macro} +% \end{macro} +% +% \begin{macro}{\switch@tabular} +% \begin{macro}{\switch@array} +% The two procedures \cmd\switch@tabular\ and \cmd\switch@array\ +% apply needed patches to the various tabular procedures, +% the former applying to the \LaTeX\ kernel, the latter to the +% required \classname{array} package (and to the number of other +% required packages that load it). +% +% \begin{macrocode} +\def\switch@tabular{% + \let\@array@sw\@array@sw@array + \@ifx{\@array\@array@LaTeX}{% + \@ifx{\multicolumn\multicolumn@LaTeX}{% + \@ifx{\@tabular\@tabular@LaTeX}{% + \@ifx{\@tabarray\@tabarray@LaTeX}{% + \@ifx{\array\array@LaTeX}{% + \@ifx{\endarray\endarray@LaTeX}{% + \@ifx{\endtabular\endtabular@LaTeX}{% + \@ifx{\@mkpream\@mkpream@LaTeX}{% + \@ifx{\@addamp\@addamp@LaTeX}{% + \@ifx{\@arrayacol\@arrayacol@LaTeX}{% + \@ifx{\@tabacol\@tabacol@LaTeX}{% + \@ifx{\@arrayclassz\@arrayclassz@LaTeX}{% + \@ifx{\@tabclassiv\@tabclassiv@LaTeX}{% + \@ifx{\@arrayclassiv\@arrayclassiv@LaTeX}{% + \@ifx{\@tabclassz\@tabclassz@LaTeX}{% + \@ifx{\@classv\@classv@LaTeX}{% + \@ifx{\hline\hline@LaTeX}{% + \@ifx{\@tabularcr\@tabularcr@LaTeX}{% + \@ifx{\@xtabularcr\@xtabularcr@LaTeX}{% + \@ifx{\@xargarraycr\@xargarraycr@LaTeX}{% + \@ifx{\@yargarraycr\@yargarraycr@LaTeX}{% + \true@sw + }{% + \false@sw + }% + }{% + \false@sw + }% + }{% + \false@sw + }% + }{% + \false@sw + }% + }{% + \false@sw + }% + }{% + \false@sw + }% + }{% + \false@sw + }% + }{% + \false@sw + }% + }{% + \false@sw + }% + }{% + \false@sw + }% + }{% + \false@sw + }% + }{% + \false@sw + }% + }{% + \false@sw + }% + }{% + \false@sw + }% + }{% + \false@sw + }% + }{% + \false@sw + }% + }{% + \false@sw + }% + }{% + \false@sw + }% + }{% + \false@sw + }% + }{% + \false@sw + }% + }{% + \false@sw + }% + {% + \class@info{Patching LaTeX tabular.}% + }{% + \class@info{Unrecognized LaTeX tabular. Please update this document class! (Proceeding with fingers crossed.)}% + }% + \let\@array\@array@ltx + \let\multicolumn\multicolumn@ltx + \let\@tabular\@tabular@ltx + \let\@tabarray\@tabarray@ltx + \let\array\array@ltx + \let\endarray\endarray@ltx + \let\endtabular\endtabular@ltx + \let\@mkpream\@mkpream@ltx + \let\@addamp\@addamp@ltx + \let\@arrayacol\@arrayacol@ltx + \let\@tabacol\@tabacol@ltx + \let\@arrayclassz\@arrayclassz@ltx + \let\@tabclassiv\@tabclassiv@ltx + \let\@arrayclassiv\@arrayclassiv@ltx + \let\@tabclassz\@tabclassz@ltx + \let\@classv\@classv@ltx + \let\hline\hline@ltx + \let\@tabularcr\@tabularcr@ltx + \let\@xtabularcr\@xtabularcr@ltx + \let\@xargarraycr\@xargarraycr@ltx + \let\@yargarraycr\@yargarraycr@ltx +}% +\def\switch@array{% + \let\@array@sw\@array@sw@LaTeX + \@ifx{\@array\@array@array}{% + \@ifx{\@tabular\@tabular@array}{% + \@ifx{\@tabarray\@tabarray@array}{% + \@ifx{\array\array@array}{% + \@ifx{\endarray\endarray@array}{% + \@ifx{\endtabular\endtabular@array}{% + \@ifx{\@mkpream\@mkpream@array}{% + \@ifx{\@classx\@classx@array}{% + \@ifx{\insert@column\insert@column@array}{% + \@ifx{\@arraycr\@arraycr@array}{% + \@ifx{\@xarraycr\@xarraycr@array}{% + \@ifx{\@xargarraycr\@xargarraycr@array}{% + \@ifx{\@yargarraycr\@yargarraycr@array}{% + \true@sw + }{% + \false@sw + }% + }{% + \false@sw + }% + }{% + \false@sw + }% + }{% + \false@sw + }% + }{% + \false@sw + }% + }{% + \false@sw + }% + }{% + \false@sw + }% + }{% + \false@sw + }% + }{% + \false@sw + }% + }{% + \false@sw + }% + }{% + \false@sw + }% + }{% + \false@sw + }% + }{% + \false@sw + }{% + \class@info{Patching array package.}% + }{% + \class@info{Unrecognized array package. Please update this document class! (Proceeding with fingers crossed.)}% + }% + \let\@array \@array@array@new + \let\@@array \@array % Cosi fan tutti + \let\@tabular \@tabular@array@new + \let\@tabarray \@tabarray@array@new + \let\array \array@array@new + \let\endarray \endarray@array@new + \let\endtabular\endtabular@array@new + \let\@mkpream \@mkpream@array@new + \let\@classx \@classx@array@new + \let\@arrayacol\@arrayacol@ltx + \let\@tabacol \@tabacol@ltx + \let\insert@column\insert@column@array@new + \expandafter\let\csname endtabular*\endcsname\endtabular % Cosi fan tutti + \let\@arraycr \@arraycr@new + \let\@xarraycr \@xarraycr@new + \let\@xargarraycr\@xargarraycr@new + \let\@yargarraycr\@yargarraycr@new +}% +% \end{macrocode} +% \end{macro} +% \end{macro} +% +% \begin{macro}{\@array@sw} +% The Boolean \cmd\@array@sw\ must be different depending on +% whether the \classname{array} package is loaded. +% \begin{macrocode} +\def\@array@sw@LaTeX{\@ifx{\\\@tabularcr}}% +\def\@array@sw@array{\@ifx{\d@llarbegin\begingroup}}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\@tabular} +% We provide the old versions of \cmd\@tabular\ along with the respective new versions. +% The change here is to avoid committing to LR mode. That will be done later (as late as possible, naturally). +% \begin{macrocode} +\def\@tabular@LaTeX{% + \leavevmode + \hbox\bgroup$% + \let\@acol\@tabacol + \let\@classz\@tabclassz + \let\@classiv\@tabclassiv + \let\\\@tabularcr + \@tabarray +}% +\def\@tabular@ltx{% + \let\@acoll\@tabacoll + \let\@acolr\@tabacolr + \let\@acol\@tabacol + \let\@classz\@tabclassz + \let\@classiv\@tabclassiv + \let\\\@tabularcr + \@tabarray +}% +\def\@tabular@array{% + \leavevmode + \hbox\bgroup$% + \col@sep\tabcolsep + \let\d@llarbegin\begingroup + \let\d@llarend\endgroup + \@tabarray +}% +\def\@tabular@array@new{% + \let\@acoll\@tabacoll + \let\@acolr\@tabacolr + \let\@acol\@tabacol + \let\col@sep\@undefined + \let\d@llarbegin\begingroup + \let\d@llarend\endgroup + \@tabarray +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\@tabarray} +% Here we provide old and new versions of the \cmd\@tabarray\ procedure. +% The change here is to parametrize the default vertical alignment, +% which is 'c' in standard \LaTeX. +% Under some circumstances, we want to change this to, say, 'v'. +% +% FIXME: must decouple \env{array} and \env{tabular}. +% \begin{macrocode} +\def\@tabarray@LaTeX{% + \m@th\@ifnextchar[\@array{\@array[c]}% +}% +\def\@tabarray@ltx{% + \m@th\@ifnextchar[\@array{\expandafter\@array\expandafter[\array@default]}% +}% +\def\@tabarray@array{% + \@ifnextchar[{\@@array}{\@@array[c]}% +}% +\def\@tabarray@array@new{% + \@ifnextchar[{\@@array}{\expandafter\@@array\expandafter[\array@default]}% +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\@tabularcr} +% \begin{macro}{\@tbpen} +% \begin{macro}{\@tabularcr} +% \begin{macro}{\@xtabularcr} +% \begin{macro}{\@xargarraycr} +% \begin{macro}{\@yargarraycr} +% \begin{macro}{\@arraycr} +% \begin{macro}{\@xarraycr} +% We provide for the \cmd\\ command within \env{tabular} to provide control over page breaking, just the same as +% that of \env{eqnarray}. +% +% The count register \cmd\intertabularlinepenalty\ is similar to \cmd\interdisplaylinepenalty: it is the penalty +% associated with each row of a tabular. When it is set to \cmd\@M, the tabular will cleave together. +% +% The count register \cmd\@tbpen\ is similar to \cmd\@eqpen: it memorizes the penalty to use after the current tabular row. +% If the \cmd\\ command is in its star form, then \cmd\@eqpen\ is set to \cmd\@M. +% +% We append code to \cmd\samepage\ so that a tabular within its scope will cleave together. +% +% We keep the standard definition of \cmd\@tabularcr\ in \cmd\@tabularcr@LaTeX\ for reference, +% and provide a new definition that works like \cmd\@eqncr: it sets \cmd\@tbpen\ to \cmd\@M\ if the star was given. +% +% We also provide new versions of \cmd\@xtabularcr, \cmd\@xargarraycr, and \cmd\@yargarraycr, all of which invoke \cmd\@tbpen. +% +% The \cmd\switch@tabular\ procedure switches in the new definitions. +% \begin{macrocode} +\newcount\intertabularlinepenalty +\intertabularlinepenalty=100 +\newcount\@tbpen +\appdef\samepage{\intertabularlinepenalty\@M}% +\def\@tabularcr@LaTeX{{\ifnum 0=`}\fi \@ifstar \@xtabularcr \@xtabularcr}% +\def\@tabularcr@ltx{{\ifnum 0=`}\fi \@ifstar {\global \@tbpen \@M \@xtabularcr }{\global \@tbpen \intertabularlinepenalty \@xtabularcr }}% +\def\@xtabularcr@LaTeX{\@ifnextchar [\@argtabularcr {\ifnum 0=`{\fi }\cr }}% +\def\@xtabularcr@ltx{\@ifnextchar [\@argtabularcr {\ifnum 0=`{\fi }\cr \noalign {\penalty \@tbpen }}}% +\def\@xargarraycr@LaTeX#1{\@tempdima #1\advance \@tempdima \dp \@arstrutbox \vrule \@height \z@ \@depth \@tempdima \@width \z@ \cr}% +\def\@xargarraycr@ltx#1{\@tempdima #1\advance \@tempdima \dp \@arstrutbox \vrule \@height \z@ \@depth \@tempdima \@width \z@ \cr \noalign {\penalty \@tbpen }}% +\def\@yargarraycr@LaTeX#1{\cr \noalign {\vskip #1}}% +\def\@yargarraycr@ltx#1{\cr \noalign {\penalty \@tbpen \vskip #1}}% +% \end{macrocode} +% +% If the \classname{array} package has been loaded, we must alter the meanings of +% \cmd\@arraycr, \cmd\@xarraycr, \cmd\@xargarraycr, and \cmd\@yargarraycr. +% In this case, it is \cmd\switch@array\ that switches in the new definitions. +% \begin{macrocode} +\def\@arraycr@array{% + \relax + \iffalse{\fi\ifnum 0=`}\fi + \@ifstar \@xarraycr \@xarraycr +}% +\def\@arraycr@new{% + \relax + \iffalse{\fi\ifnum 0=`}\fi + \@ifstar {\global \@tbpen \@M \@xarraycr }{\global \@tbpen \intertabularlinepenalty \@xarraycr }% +}% +\def\@xarraycr@array{% + \@ifnextchar [%] + \@argarraycr {\ifnum 0=`{}\fi\cr}% +}% +\def\@xarraycr@new{% + \@ifnextchar [%] + \@argarraycr {\ifnum 0=`{}\fi\cr \noalign {\penalty \@tbpen }}% +}% +\def\@xargarraycr@array#1{% + \unskip + \@tempdima #1\advance\@tempdima \dp\@arstrutbox + \vrule \@depth\@tempdima \@width\z@ + \cr +}% +\def\@xargarraycr@new#1{% + \unskip + \@tempdima #1\advance\@tempdima \dp\@arstrutbox + \vrule \@depth\@tempdima \@width\z@ + \cr + \noalign {\penalty \@tbpen }% +}% +\def\@yargarraycr@array#1{% + \cr + \noalign{\vskip #1}% +}% +\def\@yargarraycr@new#1{% + \cr + \noalign{\penalty \@tbpen \vskip #1}% +}% +% \end{macrocode} +% \end{macro} +% \end{macro} +% \end{macro} +% \end{macro} +% \end{macro} +% \end{macro} +% \end{macro} +% \end{macro} +% +% \begin{macro}{\array} +% We provide old and new versions of the \cmd\array\ procedure for both \LaTeX\ and the \classname{array} package. +% The change here is to accomodate the new procedures that will be called for the array boundaries, even +% though at present they are not special. +% A thought: here is where matrices can be readily accomodated. +% \begin{macrocode} +\def\array@LaTeX{% + \let\@acol\@arrayacol + \let\@classz\@arrayclassz + \let\@classiv\@arrayclassiv + \let\\\@arraycr + \let\@halignto\@empty + \@tabarray +}% +\def\array@ltx{% + \@ifmmode{}{\@badmath$}% + \let\@acoll\@arrayacol + \let\@acolr\@arrayacol + \let\@acol\@arrayacol + \let\@classz\@arrayclassz + \let\@classiv\@arrayclassiv + \let\\\@arraycr + \let\@halignto\@empty + \@tabarray +}% +\def\array@array{% + \col@sep\arraycolsep + \def\d@llarbegin{$}\let\d@llarend\d@llarbegin\gdef\@halignto{}% + \@tabarray +} +\def\array@array@new{% + \@ifmmode{}{\@badmath$}% + \let\@acoll\@arrayacol + \let\@acolr\@arrayacol + \let\@acol\@arrayacol +\let\col@sep\@undefined + \def\d@llarbegin{$}% + \let\d@llarend\d@llarbegin + \gdef\@halignto{}% + \@tabarray +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\@array} +% Here we provide old and new versions of \cmd\@array. +% The change here is to provide a convenient, flexible, and extensible +% mechanism for new vertical alignment options. +% +% Instead of testing the optional argument with \cmd\if, we +% use a dispatcher based on \cmd\csname. +% +% We also refrain from using \cmd\ialign, which would set +% the \cmd\tabskip\ to the wrong value. +% +% Finally, the procedure to set the \cmd\@arstrutbox\ +% is broken out so that it can be patched. +% \begin{macrocode} +\def\@array@LaTeX[#1]#2{% + \if #1t\vtop \else \if#1b\vbox \else \vcenter \fi\fi + \bgroup + \setbox\@arstrutbox\hbox{% + \vrule \@height\arraystretch\ht\strutbox + \@depth\arraystretch \dp\strutbox + \@width\z@}% + \@mkpream{#2}% + \edef\@preamble{% + \ialign \noexpand\@halignto + \bgroup \@arstrut \@preamble \tabskip\z@skip \cr}% + \let\@startpbox\@@startpbox \let\@endpbox\@@endpbox + \let\tabularnewline\\% + \let\par\@empty + \let\@sharp##% + \set@typeset@protect + \lineskip\z@skip\baselineskip\z@skip + \ifhmode \@preamerr\z@ \@@par\fi + \@preamble +}% +\def\@array@ltx[#1]#2{% + \@nameuse{@array@align@#1}% + \set@arstrutbox + \@mkpream{#2}% + \prepdef\@preamble{% + \tabskip\tabmid@skip + \@arstrut + }% + \appdef\@preamble{% + \tabskip\tabright@skip + \cr + \array@row@pre + }% +% \let\@startpbox\@@startpbox +% \let\@endpbox\@@endpbox + \let\tabularnewline\\% + \let\par\@empty + \let\@sharp##% + \set@typeset@protect + \lineskip\z@skip\baselineskip\z@skip + \tabskip\tableft@skip\relax + \ifhmode \@preamerr\z@ \@@par\fi + \everycr{}% + \expandafter\halign\expandafter\@halignto\expandafter\bgroup\@preamble +}% +% +\def\set@arstrutbox{% + \setbox\@arstrutbox\hbox{% + \vrule \@height\arraystretch\ht\strutbox + \@depth\arraystretch \dp\strutbox + \@width\z@ + }% +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\@array@array} +% +% \begin{macrocode} +\def\@array@array[#1]#2{% + \@tempdima \ht \strutbox + \advance \@tempdima by\extrarowheight + \setbox \@arstrutbox \hbox{\vrule + \@height \arraystretch \@tempdima + \@depth \arraystretch \dp \strutbox + \@width \z@}% + \begingroup + \@mkpream{#2}% + \xdef\@preamble{\noexpand \ialign \@halignto + \bgroup \@arstrut \@preamble + \tabskip \z@ \cr}% + \endgroup + \@arrayleft + \if #1t\vtop \else \if#1b\vbox \else \vcenter \fi \fi + \bgroup + \let \@sharp ##\let \protect \relax + \lineskip \z@ + \baselineskip \z@ + \m@th + \let\\\@arraycr \let\tabularnewline\\\let\par\@empty \@preamble +}% +\def\@array@array@new[#1]#2{% + \@tempdima\ht\strutbox + \advance\@tempdima by\extrarowheight + \setbox\@arstrutbox\hbox{% + \vrule \@height\arraystretch\@tempdima + \@depth \arraystretch\dp\strutbox + \@width \z@ + }% + \begingroup + \@mkpream{#2}% + \xdef\@preamble{\@preamble}% + \endgroup + \prepdef\@preamble{% + \tabskip\tabmid@skip + \@arstrut + }% + \appdef\@preamble{% + \tabskip\tabright@skip + \cr + \array@row@pre + }% + \@arrayleft + \@nameuse{@array@align@#1}% + \m@th + \let\\\@arraycr + \let\tabularnewline\\% + \let\par\@empty + \let\@sharp##% + \set@typeset@protect + \lineskip\z@\baselineskip\z@ + \tabskip\tableft@skip + \everycr{}% + \expandafter\halign\expandafter\@halignto\expandafter\bgroup\@preamble +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\endarray} +% Here we provide old and new versions of \cmd\endarray. +% The change here is to use a single procedure to close +% out any array-like structure, namely \cmd\endarray@ltx. +% It merely closes out the \cmd\halign. +% \begin{macrocode} +\def\endarray@LaTeX{% + \crcr\egroup\egroup +}% +\def\endarray@ltx{% + \crcr\array@row@pst\egroup\egroup +}% +\def\endarray@array{% + \crcr \egroup \egroup \@arrayright \gdef\@preamble{}% +}% +\def\endarray@array@new{% + \crcr\array@row@pst\egroup\egroup % Same as \endarray@ltx + \@arrayright + \global\let\@preamble\@empty +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\endtabular} +% +% \begin{macrocode} +\def\endtabular@LaTeX{% + \crcr\egroup\egroup $\egroup +}% +\def\endtabular@ltx{% + \endarray +}% +\def\endtabular@array{% + \endarray $\egroup +}% +\def\endtabular@array@new{% + \endarray +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{endtabular*} +% Here we provide a proper definition for the star-form of \enve{endtabular}. +% It is one of the enduring curiosities that the \LaTeX\ kernel continues to use +% dangerously and inappropriately ``optimized'' definitions for such commands. +% \begin{macrocode} +\@namedef{endtabular*}{\endtabular}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\multicolumn} +% +% \begin{macrocode} +\long\def\multicolumn@LaTeX#1#2#3{% + \multispan{#1}\begingroup + \@mkpream{#2}% + \def\@sharp{#3}\set@typeset@protect + \let\@startpbox\@@startpbox\let\@endpbox\@@endpbox + \@arstrut \@preamble\hbox{}\endgroup\ignorespaces +}% +\long\def\multicolumn@ltx#1#2#3{% + \multispan{#1}% + \begingroup + \@mkpream{#2}% + \def\@sharp{#3}% + \set@typeset@protect + %\let\@startpbox\@@startpbox\let\@endpbox\@@endpbox + \@arstrut + \@preamble + \hbox{}% + \endgroup + \ignorespaces +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\@array@align@} +% \begin{macro}{\array@default} +% Here are the various procedures for the vertical alignment options. +% The change from standard \LaTeX\ is that we do not go into math mode +% in every case: only when required by \cmd\vcenter. +% Also, we use \cmd\aftergroup\ to close out the boxes and modes we have started. +% It requires only that each procedure issue exactly one unmatched \cmd\bgroup. +% +% We establish here the default vertical alignment. +% \begin{macrocode} +\def\@array@align@t{\leavevmode\vtop\bgroup}% +\def\@array@align@b{\leavevmode\vbox\bgroup}% +\def\@array@align@c{\leavevmode\@ifmmode{\vcenter\bgroup}{$\vcenter\bgroup\aftergroup$\aftergroup\relax}}% +\def\@array@align@v{% + \@ifmmode{% + \@badmath + \vcenter\bgroup + }{% + \@ifinner{% + $\vcenter\bgroup\aftergroup$ + }{% + \@@par\bgroup + }% + }% +}% +\def\array@default{c}% +% \end{macrocode} +% \end{macro} +% \end{macro} +% +% \begin{macro}{\array@row@pre} +% \begin{macro}{\array@row@pst} +% \begin{macro}{\array@row@rst} +% The procedure \cmd\array@row@rst\ reestablishes a default context for +% an alignment, so that they can be nested. +% Any environment or procedure that alters the way alignments are formatted +% must patch this procedure to restore from that alteration. +% To start things off, we equate \cmd\@array@align@v\ to \cmd\@array@align@c, +% because it does not make sense to do the former in any context other +% than the MVL or in a list that will be unboxed onto the MVL. +% \begin{macrocode} +\def\array@row@rst{% + \let\@array@align@v\@array@align@c +}% +\def\array@row@pre{}% +\def\array@row@pst{}% +% \end{macrocode} +% \end{macro} +% \end{macro} +% \end{macro} +% +% \begin{macro}{\toprule} +% \begin{macro}{\colrule} +% \begin{macro}{\botrule} +% Default definitions for \cmd\toprule, \cmd\colrule, \cmd\botrule +% \begin{macrocode} +\newcommand\toprule{\tab@rule{\column@font}{\column@fil}{\frstrut}}% +\newcommand\colrule{\unskip\lrstrut\\\tab@rule{\body@font}{}{\frstrut}}% +\newcommand\botrule{\unskip\lrstrut\\\noalign{\hline@rule}{}}% +% \end{macrocode} +% \end{macro} +% \end{macro} +% \end{macro} +% +% \begin{macro}{\hline} +% \begin{macrocode} +\def\hline@LaTeX{% + \noalign{\ifnum0=`}\fi\hrule \@height \arrayrulewidth \futurelet + \reserved@a\@xhline +}% +\def\hline@ltx{% + \noalign{% + \ifnum0=`}\fi + \hline@rule + \futurelet\reserved@a\@xhline + % \noalign ended in \@xhline +}% +\def\@xhline@unneeded{% + \say\reserved@a + \ifx\reserved@a\hline + \vskip\doublerulesep + \vskip-\arrayrulewidth + \fi + \ifnum0=`{\fi}% +}% +\def\tab@rule#1#2#3{% + \crcr + \noalign{% + \hline@rule + \gdef\@arstrut@hook{% + \global\let\@arstrut@hook\@empty + #3% + }% + \gdef\cell@font{#1}% + \gdef\cell@fil{#2}% + }% +}% +\def\column@font{}% +\def\column@fil{}% +\def\body@font{}% +\def\cell@font{}% +\def\frstrut{}% +\def\lrstrut{}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\@arstrut@hline} +% \begin{macro}{\@arstrut@org} +% \begin{macro}{\@arstrut@hook} +% \begin{macro}{\@arstrutbox@hline} +% \begin{macro}{\set@arstrutbox} +% \begin{macro}{\hline@rule} +% The procedure \cmd\@arstrut@hline\ is substantially the same as +% \cmd\@arstrut, except the strut copied in is \cmd\@arstrutbox@hline +% instead of \cmd\@arstrutbox. +% +% The procedure \cmd\@arstrut@hook\ is redefined in \cmd\tab@rule! +% +% The register \cmd\@arstrutbox@hline. +% +% We append to \cmd\set@arstrutbox\ the code necessary to set a strut following an \cmd\hline. +% +% The procedure \cmd\hline@rule\ lays down a rule, and changes the meaning of \cmd\@arstrut\ +% so that the next line will be correctly strutted. +% +% The \cmd\@arstrut@hline@clnc\ is a klootch, a magic number. +% \begin{macrocode} +\def\@arstrut@hline{% + \relax + \@ifmmode{\copy}{\unhcopy}\@arstrutbox@hline + \@arstrut@hook +}% +% +\let\@arstrut@org\@arstrut +\def\@arstrut@hook{% + \global\let\@arstrut\@arstrut@org +}% +% +\newbox\@arstrutbox@hline +\appdef\set@arstrutbox{% + \setbox\@arstrutbox@hline\hbox{% + \setbox\z@\hbox{$0^{0}_{}$}% + \dimen@\ht\z@\advance\dimen@\@arstrut@hline@clnc + \@ifdim{\dimen@<\arraystretch\ht\strutbox}{\dimen@=\arraystretch\ht\strutbox}{}% + \vrule \@height\dimen@ + \@depth\arraystretch \dp\strutbox + \@width\z@ + }% +}% +% +\def\hline@rule{% + \hrule \@height \arrayrulewidth + \global\let\@arstrut\@arstrut@hline +}% +\def\@arstrut@hline@clnc{2\p@}% % Klootch: magic number +% \end{macrocode} +% \end{macro} +% \end{macro} +% \end{macro} +% \end{macro} +% \end{macro} +% \end{macro} +% +% \begin{macro}{\tableft@skip} +% \begin{macrocode} +\def\tableft@skip{\z@skip}% +\def\tabmid@skip{\z@skip}%\@flushglue +\def\tabright@skip{\z@skip}% +\def\tableftsep{\tabcolsep}% +\def\tabmidsep{\tabcolsep}% +\def\tabrightsep{\tabcolsep}% +\def\cell@fil{}% +\def\pbox@hook{}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\@arstrut} +% \begin{macrocode} +\appdef\@arstrut{\@arstrut@hook}% +\let\@arstrut@hook\@empty +\def\@addtopreamble{\appdef\@preamble}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\@mkpream} +% \begin{macrocode} +\def\@mkpream@LaTeX#1{% + \@firstamptrue\@lastchclass6 + \let\@preamble\@empty + \let\protect\@unexpandable@protect + \let\@sharp\relax + \let\@startpbox\relax\let\@endpbox\relax + \@expast{#1}% + \expandafter\@tfor \expandafter + \@nextchar \expandafter:\expandafter=\reserved@a\do + {\@testpach\@nextchar + \ifcase \@chclass \@classz \or \@classi \or \@classii \or \@classiii + \or \@classiv \or\@classv \fi\@lastchclass\@chclass}% + \ifcase \@lastchclass \@acol + \or \or \@preamerr \@ne\or \@preamerr \tw@\or \or \@acol \fi +}% +\def\@mkpream@ltx#1{% + \@firstamptrue + \@lastchclass6 + \let\@preamble\@empty + \let\protect\@unexpandable@protect + \let\@sharp\relax +%\let\@startpbox\relax\let\@endpbox\relax + \@expast{#1}% + \expandafter\@tfor\expandafter\@nextchar\expandafter:\expandafter=\reserved@a + \do{% + \expandafter\@testpach\expandafter{\@nextchar}% + \ifcase\@chclass + \@classz + \or + \@classi + \or + \@classii + \or + \@classiii + \or + \@classiv + \or + \@classv + \fi + \@lastchclass\@chclass + }% + \ifcase\@lastchclass + \@acolr % right-hand column + \or + \or + \@preamerr\@ne + \or + \@preamerr\tw@ + \or + \or + \@acolr % right-hand column + \fi +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\insert@column} +% \begin{macrocode} +\def\insert@column@array{% + \the@toks \the \@tempcnta + \ignorespaces \@sharp \unskip + \the@toks \the \count@ \relax +}% +\def\insert@column@array@new{% + \the@toks\the\@tempcnta + \array@row@rst\cell@font + \ignorespaces\@sharp\unskip + \the@toks\the\count@ + \relax +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\@mkpream@relax} +% The procedure \cmd\@mkpream@relax\ participates in a strange and wonderful +% method of binding the alignment procedure---but only certain parts thereof. +% +% Here is how it works: in \LaTeX, the \classname{array} package, and in the +% \classname{longtable} package alike, there is a need to create an alignment +% preamble (using \cmd\@mkpream) for use by the upcoming \cmd\halign. +% Then, in both \classname{array} and \classname{longtable}, \TeX's \cmd\edef\ +% is used to `compile in place' that alignment preamble. +% +% In the case of \classname{array}, the operation is done in order to +% pre-expand the use of \texttt{*}, in \classname{longtable}, it is to +% set the widths of the columns. +% +% Now, during this \cmd\edef, certain control sequence names must \emph{not} +% be expanded, and those are robustified by \cmd\@mkpream@relax. +% +% \begin{macrocode} +\def\@mkpream@relax{% + \let\tableftsep\relax + \let\tabmidsep\relax + \let\tabrightsep\relax + \let\array@row@rst\relax + \let\cell@font\relax + \let\@startpbox\relax +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\@mkpream} +% \begin{macrocode} +\def\@mkpream@array#1{% + \gdef\@preamble{}\@lastchclass 4 \@firstamptrue + \let\@sharp\relax \let\@startpbox\relax \let\@endpbox\relax + \@temptokena{#1}\@tempswatrue + \@whilesw\if@tempswa\fi{\@tempswafalse\the\NC@list}% + \count@\m@ne + \let\the@toks\relax + \prepnext@tok + \expandafter \@tfor \expandafter \@nextchar + \expandafter :\expandafter =\the\@temptokena \do + {\@testpach + \ifcase \@chclass \@classz \or \@classi \or \@classii + \or \save@decl \or \or \@classv \or \@classvi + \or \@classvii \or \@classviii + \or \@classx + \or \@classx \fi + \@lastchclass\@chclass}% + \ifcase\@lastchclass + \@acol \or + \or + \@acol \or + \@preamerr \thr@@ \or + \@preamerr \tw@ \@addtopreamble\@sharp \or + \or + \else \@preamerr \@ne \fi + \def\the@toks{\the\toks}% +}% +\def\@mkpream@array@new#1{% + \gdef\@preamble{}% + \@lastchclass\f@ur + \@firstamptrue + \let\@sharp\relax + \@mkpream@relax +%\let\@startpbox\relax\let\@endpbox\relax + \@temptokena{#1}\@tempswatrue + \@whilesw\if@tempswa\fi{\@tempswafalse\the\NC@list}% + \count@\m@ne + \let\the@toks\relax + \prepnext@tok + \expandafter\@tfor\expandafter\@nextchar\expandafter:\expandafter=\the\@temptokena + \do{% + \@testpach + \ifcase\@chclass + \@classz + \or + \@classi + \or + \@classii + \or + \save@decl + \or + \or + \@classv + \or + \@classvi + \or + \@classvii + \or + \@classviii + \or + \@classx + \or + \@classx + \fi + \@lastchclass\@chclass + }% + \ifcase\@lastchclass + \@acolr % right-hand column + \or + \or + \@acolr % right-hand column + \or + \@preamerr\thr@@ + \or + \@preamerr\tw@\@addtopreamble\@sharp + \or + \or + \else + \@preamerr\@ne + \fi + \def\the@toks{\the\toks}% +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\@addamp} +% \begin{macrocode} +\def\@addamp@LaTeX{% + \if@firstamp\@firstampfalse\else\edef\@preamble{\@preamble &}\fi +}% +\def\@addamp@ltx{% + \if@firstamp\@firstampfalse\else\@addtopreamble{&}\fi +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\@arrayacol} +% \begin{macrocode} +\def\@arrayacol@LaTeX{% + \edef\@preamble{\@preamble \hskip \arraycolsep}% +}% +\def\@arrayacol@ltx{% + \@addtopreamble{\hskip\arraycolsep}% +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\@tabacol} +% \begin{macrocode} +\def\@tabacoll{% + \@addtopreamble{\hskip\tableftsep\relax}% +}% +\def\@tabacol@LaTeX{% + \edef\@preamble{\@preamble \hskip \tabcolsep}% +}% +\def\@tabacol@ltx{% + \@addtopreamble{\hskip\tabmidsep\relax}% +}% +\def\@tabacolr{% + \@addtopreamble{\hskip\tabrightsep\relax}% +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\@arrayclassz} +% \begin{macrocode} +\def\@arrayclassz@LaTeX{% + \ifcase \@lastchclass \@acolampacol \or \@ampacol \or + \or \or \@addamp \or + \@acolampacol \or \@firstampfalse \@acol \fi + \edef\@preamble{\@preamble + \ifcase \@chnum + \hfil$\relax\@sharp$\hfil \or $\relax\@sharp$\hfil + \or \hfil$\relax\@sharp$\fi}% +}% +\def\@arrayclassz@ltx{% + \ifcase\@lastchclass + \@acolampacol + \or + \@ampacol + \or + \or + \or + \@addamp + \or + \@acolampacol + \or + \@firstampfalse\@acoll + \fi + \ifcase\@chnum + \@addtopreamble{% + \hfil\array@row@rst$\relax\@sharp$\hfil + }% + \or + \@addtopreamble{% + \array@row@rst$\relax\@sharp$\hfil + }% + \or + \@addtopreamble{% + \hfil\array@row@rst$\relax\@sharp$% + }% + \fi +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\@tabclassz} +% \begin{macrocode} +\def\@tabclassz@LaTeX{% + \ifcase\@lastchclass + \@acolampacol + \or + \@ampacol + \or + \or + \or + \@addamp + \or + \@acolampacol + \or + \@firstampfalse\@acol + \fi + \edef\@preamble{% + \@preamble{% + \ifcase\@chnum + \hfil\ignorespaces\@sharp\unskip\hfil + \or + \hskip1sp\ignorespaces\@sharp\unskip\hfil + \or + \hfil\hskip1sp\ignorespaces\@sharp\unskip + \fi}}% +}% +\def\@tabclassz@ltx{% + \ifcase\@lastchclass + \@acolampacol + \or + \@ampacol + \or + \or + \or + \@addamp + \or + \@acolampacol + \or + \@firstampfalse\@acoll + \fi + \ifcase\@chnum + \@addtopreamble{% + {\hfil\array@row@rst\cell@font\ignorespaces\@sharp\unskip\hfil}% + }% + \or + \@addtopreamble{% + {\cell@fil\hskip1sp\array@row@rst\cell@font\ignorespaces\@sharp\unskip\hfil}% + }% + \or + \@addtopreamble{% + {\hfil\hskip1sp\array@row@rst\cell@font\ignorespaces\@sharp\unskip\cell@fil}% + }% + \fi +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\@tabclassiv} +% \begin{macrocode} +\def\@tabclassiv@LaTeX{% + \@addtopreamble\@nextchar +}% +\def\@tabclassiv@ltx{% + \expandafter\@addtopreamble\expandafter{\@nextchar}% +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\@arrayclassiv} +% \begin{macrocode} +\def\@arrayclassiv@LaTeX{% + \@addtopreamble{$\@nextchar$}% +}% +\def\@arrayclassiv@ltx{% + \expandafter\@addtopreamble\expandafter{\expandafter$\@nextchar$}% +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\@classv} +% \begin{macrocode} +\def\@classv@LaTeX{% + \@addtopreamble{\@startpbox{\@nextchar}\ignorespaces + \@sharp\@endpbox}% +}% +\def\@classv@ltx{% + \expandafter\@addtopreamble + \expandafter{% + \expandafter \@startpbox + \expandafter {\@nextchar}% + \pbox@hook\array@row@rst\cell@font\ignorespaces\@sharp\@endpbox + }% +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\@classx} +% \begin{macrocode} +\def\@classx@array{% + \ifcase \@lastchclass + \@acolampacol \or + \@addamp \@acol \or + \@acolampacol \or + \or + \@acol \@firstampfalse \or + \@addamp + \fi +}% +\def\@classx@array@new{% + \ifcase \@lastchclass + \@acolampacol + \or + \@addamp \@acol + \or + \@acolampacol + \or + \or + \@firstampfalse\@acoll + \or + \@addamp + \fi +}% +% \end{macrocode} +% \end{macro} +% +% +% \subsection{Repair other broken parts of \LaTeX} +% +% \begin{macro}{\@xbitor} +% Expansion part has extraneous space token. Removed. +% \begin{macrocode} +\def\@xbitor@LaTeX #1{\@tempcntb \count#1 + \ifnum \@tempcnta =\z@ + \else + \divide\@tempcntb\@tempcnta + \ifodd\@tempcntb \@testtrue\fi + \fi}% +\def\@xbitor@ltx#1{% + \@tempcntb\count#1% + \@ifnum{\@tempcnta=\z@}{}{% + \divide\@tempcntb\@tempcnta + \@ifodd\@tempcntb{\@testtrue}{}% + }% +}% +\@ifx{\@xbitor\@xbitor@LaTeX}{% + \class@info{Repairing broken LaTeX \string\@xbitor}% +}{% + \class@info{Unrecognized LaTeX \string\@xbitor. Please update this document class! (Proceeding with fingers crossed.)}% +}% +\let\@xbitor\@xbitor@ltx +% \end{macrocode} +% \end{macro} +% +% +% \subsection{Syntax} +% \begin{macro}{\@gobble@opt@one} +% The \cmd\@gobble@opt@one\ command eats up an optional argument +% and one required argument. +% \begin{macrocode} +\newcommand*\@gobble@opt@one[2][]{}% +% \end{macrocode} +% \end{macro} +% +% \subsection{Auto-indented Contents} +% Facility to automatically determine the proper indentation of +% the TOC entries. +% +% Note on \classname{hyperref} compatibility: +% We must respect that +% \cmd\contentsline now has a 4th argument. +% So, instead of trying to override the meaning of \cmd\contentsline, +% we use the aux file to remember max values from one run to the next. +% +% In this respect, this package retains compatability with +% \classname{hyperref}. +% +% \begin{macro}{\@starttoc} +% Install hooks at beginning and end of the TOC processing. +% \begin{macrocode} +\def\@starttoc#1{% + \begingroup + \toc@pre + \makeatletter + \@input{\jobname.#1}% + \if@filesw + \expandafter\newwrite\csname tf@#1\endcsname + \immediate\openout \csname tf@#1\endcsname \jobname.#1\relax + \fi + \@nobreakfalse + \toc@post + \endgroup +}% +\def\toc@pre{}% +\def\toc@post{}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\toc@@font} +% Interface for setting the formatting characteristics of this part +% of the TOC. +% +% Note: \cmd\toc@@font\ is the common font for all auto-sizing toc commands, +% although this, too, could become a dispatcher. +% \begin{macrocode} +\def\toc@@font{}%{\footnotesize\rmfamily}% +\def\@dotsep{\z@}%{5.5pt}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\l@section} +% Interface for determining which TOC elements are automatically indented. +% +% All of the \cmd\l@\dots\ commands simply go through the +% bottleproc \cmd\l@@sections. The calling convention is +% to pass the name of self and the name of parent. +% If you want to exclude any of these from the indentation +% scheme, simply leave the \cmd\l@\dots\ command undefined. +% +% Note that the parent of ``section'' is nil, so we have to +% define a stub. +% \begin{verbatim} +%\def\l@section{% +% \l@@sections{}{section}% Implicit #3#4 +%}% +%\def\tocleft@{\z@}% +%\def\l@subsection{% +% \l@@sections{section}{subsection}% Implicit #3#4 +%}% +%\def\l@subsubsection{% +% \l@@sections{subsection}{subsubsection}% Implicit #3#4 +%}% +%\def\l@paragraph{% +% \l@@sections{subsubsection}{paragraph}% Implicit #3#4 +%}% +%\def\l@subparagraph#1#2{% +% \l@@sections{paragraph}{subparagraph}% Implicit #3#4 +%}% +% \end{verbatim} +% \end{macro} +% +% Glom some \cmd\dimen\ registers. +% \begin{macrocode} +\let\tocdim@section \leftmargini +\let\tocdim@subsection \leftmarginii +\let\tocdim@subsubsection \leftmarginiii +\let\tocdim@paragraph \leftmarginiv +\let\tocdim@appendix \leftmarginv +\let\tocdim@pagenum \leftmarginvi +% \end{macrocode} +% +% \begin{macro}{\toc@pre@auto} +% \begin{macro}{\toc@post@auto} +% We patch \cmd\@starttoc\ to: +% 1) before TOC processing, +% initialize the max registers and +% set the needed dimensions from +% the values stored in the auxiliary file, and +% 2) after TOC processing, +% store out those max register values into the auxiliary file. +% +% Note that the font is set here: all other TOC entries must +% override these font settings. +% +% To activate this override of the standard \LaTeX\ processing, +% the substyle does: \cmd\let\cmd\toc@pre\cmd\toc@pre@auto\ +% and \cmd\let\cmd\toc@post\cmd\toc@post@auto. +% \begin{macrocode} +\def\toc@pre@auto{% + \toc@@font + \@tempdima\z@ + \toc@setindent\@tempdima{section}% + \toc@setindent\@tempdima{subsection}% + \toc@setindent\@tempdima{subsubsection}% + \toc@setindent\@tempdima{paragraph}% + \toc@letdimen{appendix}% + \toc@letdimen{pagenum}% +}% +\def\toc@post@auto{% + \if@filesw + \begingroup + \toc@writedimen{section}% + \toc@writedimen{subsection}% + \toc@writedimen{subsubsection}% + \toc@writedimen{paragraph}% + \toc@writedimen{appendix}% + \toc@writedimen{pagenum}% + \endgroup + \fi +}% +% \end{macrocode} +% \end{macro} +% \end{macro} +% +% \begin{macro}{\toc@setindent} +% \begin{macrocode} +\def\toc@setindent#1#2{% + \csname tocdim@#2\endcsname\tocdim@min\relax + \@ifundefined{tocmax@#2}{\@namedef{tocmax@#2}{\z@}}{}% + \advance#1\@nameuse{tocmax@#2}\relax + \expandafter\edef\csname tocleft@#2\endcsname{\the#1}% +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\toc@letdimen} +% \begin{macrocode} +\def\toc@letdimen#1{% + \csname tocdim@#1\endcsname\tocdim@min\relax + \@ifundefined{tocmax@#1}{\@namedef{tocmax@#1}{\z@}}{}% + \expandafter\let\csname tocleft@#1\expandafter\endcsname\csname tocmax@#1\endcsname +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\toc@writedimen} +% \begin{macrocode} +\def\toc@writedimen#1{% + \immediate\write\@auxout{% + \gdef\expandafter\string\csname tocmax@#1\endcsname{% + \expandafter\the\csname tocdim@#1\endcsname + }% + }% +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\l@@sections} +% The procedure for formatting the indented TOC entries. +% We use control sequence names such as \cmd\tocmax@section\ and +% \cmd\tocleft@section, the former being written to the auxiliary file +% and the latter only defined for the duration of the TOC processing. +% +% Note that the assignment of \cmd\box\cmd\z@\ must endure +% over the invocation of |#3|. +% \begin{macrocode} +\def\l@@sections#1#2#3#4{% + % #1 - superior section + % #2 - this section + % #3 - content, including possible \numberline + % #4 - page number + \begingroup + \everypar{}% + \set@tocdim@pagenum{#4}% + \global\@tempdima\csname tocdim@#2\endcsname + \leftskip\csname tocleft@#2\endcsname\relax + \dimen@\csname tocleft@#1\endcsname\relax + \parindent-\leftskip\advance\parindent\dimen@ + \rightskip\tocleft@pagenum plus 1fil\relax + \skip@\parfillskip\parfillskip\z@ + \let\numberline\numberline@@sections + \@nameuse{l@f@#2}% + \ignorespaces#3\unskip\nobreak\hskip\skip@ + \hb@xt@\rightskip{\hfil\unhbox\@tempboxa}\hskip-\rightskip\hskip\z@skip + \par + \expandafter\aftergroup\csname tocdim@#2\endcsname\expandafter + \endgroup\the\@tempdima\relax +}% +\def\set@tocdim@pagenum#1{% + \setbox\@tempboxa\hbox{\ignorespaces#1}% + \@ifdim{\tocdim@pagenum<\wd\z@}{\global\tocdim@pagenum\wd\z@}{}% +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\numberline@@sections} +% The bottleproc for all \cmd\numberline\ processing in indented TOC entries. +% The first argument is self. +% +% We use \cmd\@tempdima\ to pass a value around (via global assignment) because +% \cmd\numberline\ executes inside a group if the +% \classname{hyperref} package is loaded. +% Would that it were not so! +% \begin{macrocode} +\def\numberline@@sections#1{% + \leavevmode\hb@xt@-\parindent{% + \hfil + \@if@empty{#1}{}{% + \setbox\z@\hbox{#1.\kern\@dotsep}% + \@ifdim{\@tempdima<\wd\z@}{\global\@tempdima\wd\z@}{}% + \unhbox\z@ + }% + }% + \ignorespaces +}% +\def\tocdim@min{\z@}% +% \end{macrocode} +% \end{macro} +% +% +% \subsection{Lists} +% \begin{macro}{\list} +% Using \cmd\parshape\ to implement lists was always suspect +% (can you get behind \cmd\parshape\cmd\@ne?) and we now see that +% it was a mistake all along. Why? Because \cmd\parshape, like +% \cmd\hangindent, achieves its effect via ``shifting'' the \cmd\hbox es +% in a paragraph +% instead of using \cmd\leftskip\ and \cmd\parindent, which is +% robust during column balancing. +% +% We introduce the alternative method with a hook into +% the \LaTeX\ kernel procedure \cmd\list, which is +% the implementation of all lists. +% +% \begin{macrocode} +\def\list#1#2{% + \ifnum \@listdepth >5\relax + \@toodeep + \else + \global\advance\@listdepth\@ne + \fi + \rightmargin\z@ + \listparindent\z@ + \itemindent\z@ + \csname @list\romannumeral\the\@listdepth\endcsname + \def\@itemlabel{#1}% + \let\makelabel\@mklab + \@nmbrlistfalse + #2\relax + \@trivlist + \parskip\parsep + \set@listindent + \ignorespaces +}% +\def\set@listindent@parshape{% + \parindent\listparindent + \advance\@totalleftmargin\leftmargin + \advance\linewidth-\rightmargin + \advance\linewidth-\leftmargin + \parshape\@ne\@totalleftmargin\linewidth +}% +\def\set@listindent@{% + \parindent\listparindent + \advance\@totalleftmargin\leftmargin + \advance\rightskip\rightmargin + \advance\leftskip\@totalleftmargin +}% +\let\set@listindent\set@listindent@parshape +% \end{macrocode} +% \end{macro} +% +% \subsection{End of the \file{ltxutil} {\sc docstrip} module} +% Here ends the module. +% \begin{macrocode} +%</ltxutil-krn> +% \end{macrocode} +% +% +% \Finale +% %Here ends the programmer's documentation. +% \endinput +% +\endinput +%%EOF diff --git a/Master/texmf-dist/source/latex/revtex4/revtex4.dtx b/Master/texmf-dist/source/latex/revtex4/revtex4.dtx new file mode 100644 index 00000000000..693747ef47f --- /dev/null +++ b/Master/texmf-dist/source/latex/revtex4/revtex4.dtx @@ -0,0 +1,8728 @@ +% \iffalse +% ****** Start of file revtex4.dtx ****** % +% +% This file is part of the APS files in the REVTeX 4 distribution. +% For version number, search on the string "%FileInfo" +% +% First revision by David Carlisle, +% later versions by Arthur Ogawa. +% +% Copyright (c) 2000 The American Physical Society. +% http://publish.aps.org/revtex4/ +% mailto:revtex@aps.org +% +% +% See the REVTeX 4 README file for restrictions and more information. +% \fi +% \iffalse +%<*driver> + \ProvidesFile{revtex4.dtx} +%</driver> +%<class-pre|aps|rmp>\NeedsTeXFormat{LaTeX2e}[1996/12/01] +%<class-pre>\ProvidesClass{revtex4} +%<aps>\ProvidesFile{aps.rtx} +%<rmp>\ProvidesFile{rmp.rtx} +%<10pt>\ProvidesFile{10pt.rtx} +%<11pt>\ProvidesFile{11pt.rtx} +%<12pt>\ProvidesFile{12pt.rtx} +%<revsymb&!class-pre>\ProvidesPackage{revsymb} +%<driver>\ProvidesFile{revtex.drv} +%<*!class-pst> + [2001/08/02 v4.0 (http://publish.aps.org/revtex4/ for documentation)]%FileInfo +%</!class-pst> +%<*driver> +\listfiles +\documentclass{ltxdoc} +\RequirePackage{ltxdocext}% +\let\url\undefined +\RequirePackage[colorlinks=true,linkcolor=blue]{hyperref}% +\expandafter\ifx\csname package@font\endcsname\relax\else + \expandafter\expandafter + \expandafter\RequirePackage + \expandafter\expandafter + \expandafter{\csname package@font\endcsname}% +\fi +\CodelineIndex\EnableCrossrefs +\begin{document} +\gdef\lquote{`}\gdef\rquote{'}% +%{\tracingall\showlists} +\DocInput{revtex4.dtx} +\end{document} +%</driver> +% \fi +% +% \GetFileInfo{revtex4.dtx}\CheckSum{7177} +% +% \title{The \classname{revtex4} document class of the American Physical Society} +% \author{Arthur Ogawa% +% \thanks{First revision by David Carlisle}% +% }% +% \date{Version \fileversion, dated \filedate}% +% \newcommand\revtex{REV\TeX} +% +% \maketitle +% +% This file embodies the implementation of the APS \revtex\ 4 document class +% for electronic submissions to journals. +% +% The distribution point for this work is +% \url{http://publish.aps.org/revtex4/}, +% which contains fully unpacked, prebuilt runtime files and documentation. +% +% \tableofcontents +% +% \section{Using \protect\revtex} +% +% The file \file{README} has retrieval and installation information. +% +% User documentation is presented separately in \file{revguide.tex}. +% +% The file \file{template.aps} is a boilerplate file. +% +% \StopEventually{} +% +% \changes{v4.a}{1998/01/16} +% {Initial version} +% \changes{v4.0beta 2}{1999/06/20} +% {First modifications by Arthur Ogawa (mailto:ogawa@teleport.com)} +% \changes{v4.0beta 3}{1999/11/13} +% {New ltxgrid-based code, other bug fixes} +% \changes{v4.0beta 4}{2000/05/10} +% {More features and bug fixes: compatability with longtable and array packages. +% Now certainly incompatible with multicol.} +% \changes{v4.0beta 5}{2000/11/14} +% {Bug fixes and minor new features: title block affiliations can have ancillary data, just like authors; +% clearpage processing revamped, with floats staying in order; widetext ornaments.} +% \changes{v4.0beta 6}{2001/02/13} +% {Last bug fixes before release.} +% +% \section{Overview}% +% \revtex\ is a \LaTeXe\ document class, aomewhat like a hybrid of +% the standard \LaTeX\ \classname{book} and \classname{article} classes. +% +% Certain packages are (should be) loaded by this class in any case: +% \classname{amssymb}, \classname{amsmath}, \classname{bm}, \classname{natbib}. +% +% Certain packages are automatically loaded by this class +% when a corresponding class option has been invoked: +% +% \begin{tabular}{ll} +% \textbf{\revtex\ option}&\textbf{package}\\ +% \classoption{amsfonts}&\classname{amsfonts}\\ +% \classoption{amssymb}&\classname{amssymb}\\ +% \classoption{aps}&\classname{overcite}\\ +% \end{tabular} +% +% Certain other packages are to be loaded by the +% document through explicit use of \cmd\usepackage. +% Some mentioned in the user documentation are +% \classname{graphicx}, \classname{longtable}, +% and \classname{bm}. +% +% Certain commonly used packages are known to be incompatible with +% \revtex, among them \classname{multicol} and \classname{cite}. +% If such a package is found to be loaded, \revtex\ issues an error message +% and halts the job. +% Halting might be considered severe punishment for loading an incompatible +% package, but if we were to proceed, an even weirder error might be encountered +% further down the road. +% +% This document class implements the substyle: a set of mutually +% exclusive class options that, in this case, allow the +% document class to address multiple societies. +% It also implements a sub-substyle, giving the +% journal of the given society. +% The only society currently implemented is \classoption{aps}, +% however, code is in place for another society, the \classoption{osa}. +% +% FIXME: should always load the \classname{graphicx} package. +% +% QUERY: since \classname{amsfonts} and \classname{amssymb} extend +% syntax, why not load them in any case? +% +% QUERY: if you say \cmd\documentclass|[prb,osa]{revtex4}|, then +% the value of \cmd\@society\ will be wrong, or? Fixed. +% +% Certain events occur at \cmd\AtEndOfClass\ time: +% \begin{enumerate} +% \item +% Optionally read in packages +% \classname{amsfonts}, \classname{amssymb}. +% \item +% Read in society substyle \file{.rtx} file. +% \item +% Read in type size \file{.rtx} file. +% \item +% Read in the job's style file, the \file{.rty} file. +% \item +% Define things based on \cmd\secnumarabic@sw. +% \item +% Install procedures to execute at the very end of the class's +% \cmd\AtBeginDocument\ processing, such as +% \begin{enumerate} +% \item +% closing out the page grid +% \item +% putting out the LastPage label. +% \item +% issuing a \cmd\bibliographystyle\ command, based on the value set +% by the society substyle. +% \item +% setting default values for parameters used in the document. +%FIXME: differentiate between class's parameters that can wait until +% they are used in the document, and parameters that are used at \cmd\AtBeginDocument\ time. +% \end{enumerate} +% \item +% Install procedures to execute the very last at \cmd\AtEndDocument\ time, +% such as the \cmd\clearpage\ processing. +% \end{enumerate} +% +% Certain events are optionally scheduled for \cmd\AtBeginDocument\ time: +% \begin{enumerate} +% \item +% Setting default values for the Booleans and for other procedures used in formatting. +% \item +% In response to class options options, adjusting parameters and procedures used in formatting. +% \item +% Implementing the \classoption{eqsecnum} option, if required. +% \item +% Setting the state engine for data commands. +% \item +% Memorizing procedures for later use. +% \item +% Setting type size and area, for use by later calculations. +% \end{enumerate} +% +% Certain events are scheduled for \cmd\class@enddocumenthook\ time: +% \begin{enumerate} +% \item +% Print out the migrated floats or the end notes, if needed. +% \item +% Close out the page grid. +% \item +% Label the last page of the document +% \item +% (\classname{natbib}) prepare to read in the \file{.aux} file. +% \end{enumerate} +% +% +% \section{The installer file} +% We include here a reference copy of the installer file, \file{revtex4.ins}. +% It is not intended that one actually generate that file using \file{docstrip} +% via the revtex4 installer file, because that would involve overwriting the +% file that \TeX\ is typesetting. Instead, the installer file is provided +% pre-generated. +% +% For the most part, the installer is simple: it generates \file{revtex4.cls} and +% a collection of \file{.rtx} files from \file{revtex4.dtx}. +% +% However, the process of generating \file{revtex4.cls} involves the incorporation +% of portions of two of the \file{ltx} series of files, \file{ltxutil.dtx}, +% and \file{ltxgrid.dtx}. The former is a collection of macro definitions that +% serve as the lowest level of extensions to \LaTeX, the latter implements +% an alternative output routine and page gridding mechanism that provides for +% multicolumn page layouts, but without the sorry limitations of the +% \LaTeX\ required package, \classname{multicol}, and is also compatible with +% the \classname{longtable} package (which \classname{multicol} is not). +% \begin{macrocode} +%<*ins> +% This file requires docstrip version 2.4 or higher available from +% ftp://ctan.tug.org/tex-archive/macros/latex/unpacked/docstrip.tex +\input docstrip +\preamble + +This file is part of the APS files in the REVTeX 4 distribution. +For the version number, search on the string %FileInfo + +Copyright (c) 2000 The American Physical Society. +http://publish.aps.org/revtex4/ +mailto:revtex@aps.org + +Maintained by Arthur Ogawa (mailto:ogawa@teleport.com) +under contract to the American Physical Society. + +See the REVTeX 4 README file for restrictions and more information. + +\endpreamble +\keepsilent +\askforoverwritefalse +\generate{% + \file{revtex4.cls}{% + \from{revtex4.dtx}{class-pre}% + \from{ltxutil.dtx}{ltxutil-krn}% + \from{ltxgrid.dtx}{ltxgrid-krn}% + \from{textcase.dtx}{package}% + \from{revtex4.dtx}{class-pst}% + }% + \file{aps.rtx}{\from{revtex4.dtx}{aps}}% + \file{rmp.rtx}{\from{revtex4.dtx}{rmp}}% + \file{10pt.rtx}{\from{revtex4.dtx}{10pt}}% + \file{11pt.rtx}{\from{revtex4.dtx}{11pt}}% + \file{12pt.rtx}{\from{revtex4.dtx}{12pt}}% + \file{revsymb.sty}{\from{revtex4.dtx}{revsymb}}% +}% +\endbatchfile +%</ins> +% \end{macrocode} +% +% +% \section{Writing journal-specifc extensions to \revtex} +% +% With this version of \revtex, we introduce a somewhat different +% scheme for adapting \revtex\ to the needs of a specific journal. +% +% To create a journal substyle, you create new class options in +% \revtex\ for the society, say \classoption{osa}, and any of that society's journals, +% one of which is, say, \classoption{josaa}, +% using the code for the APS as a guide. +% In particular, each of your new options should separately define \cmd\@society\ +% and \cmd\@journal. That for the former will be the same for +% all options relating to a particular society. +% +% Then, for the society, you create a corresponding \file{.rtx} +% file, in our case \file{osa.rtx}. +% Within that file, you override procedures and parameter assignments as you see fit. +% Ideally they will be generally applicable to all of that society's +% journals (see the file \file{aps.rtx} for a realization of this scheme). +% Also within that file, you include a +% section of code for each journal, that for josaa looks like: +% \begin{verbatim} +% \@ifx{\@journal\journal@josaa}{% +% <code specific to the josaa> +% }{}% +% \end{verbatim} +% +% Thus far, the scheme is similar to that used in \revtex\ 3.1. +% However, the new scheme does differ from the old in that +% the \file{.rtx} file should define no syntactical extensions +% to \revtex. +% +% \section{The \classname{revtex4} Document Class} +% +% \begin{macrocode} +%<*class-pre> +\def\class@name{revtex4}% +% \end{macrocode} +% +% Print a banner in the log: +% \begin{macrocode} +\GenericInfo{}{\space + Copyright (c) 2001 The American Physical Society.^^J + mailto:revtex@aps.org^^J + Licensed under the LPPL:^^Jhttp://www.ctan.org/tex-archive/macros/latex/base/lppl.txt^^J + Arthur Ogawa <ogawa@teleport.com>^^J + Based on work by David Carlisle <david@dcarlisle.demon.co.uk>% + \@gobble +}% +% \end{macrocode} +% +% \subsection{Compatability Processing} +% +% If the document has \cmd\documentstyle|{revtex4}|, +% then, instead of attempting to run in compatability mode, +% just complain and exit. +% \changes{v4.0beta 3}{1999/11/13} +% {Lose compatability mode.} +% \begin{macrocode} +\if@compatibility + \edef\reserved@a{\errhelp{% + Change your \string\documentstyle\space statement to + \string\documentclass\space and rerun. + }}\reserved@a + \errmessage{You cannot run \class@name\space in compatability mode}% + \expandafter\@@end +\fi +% \end{macrocode} +% +% +% \section{Extensions to the \LaTeX\ Kernel} +% I am introducing +% some portions of the \classname{ltxkrnext} package +% (\url{ftp://ftp.teleport.com/users/ogawa/macros/latex/contrib/supported/ltxkrnext/}). +% +% First, de-fang unwanted commands and turn \cmd\ProvidesPackage\ into an informative message. +% \begin{macrocode} +\let\DeclareOption@latex\DeclareOption +\let\DeclareOption\@gobbletwo +\let\ProcessOptions@latex\ProcessOptions +\let\ProcessOptions\@empty +\let\ProvidesPackage@latex\ProvidesPackage +\def\ProvidesPackage#1[#2]{% + \class@info{Incorporating package #1 [#2].}% +}% +%</class-pre> +% \end{macrocode} +% Now load the utility and page grid packages. +% The \classname{ltxutil} and \classname{ltxgrid} packages are distributed with +% \revtex. +% \begin{macrocode} +%\RequirePackage{ltxutil}% +%\RequirePackage{ltxgrid}% +% \end{macrocode} +% +% The \classname{textcase} package is available from +% \url{CTAN:/macros/latex/contrib/supported/carlisle/textcase.dtx} and is +% used at the suggestion, and by permission, of David~P. Carlisle. +% \begin{quotation} +% \cmd\ProvidesFile{textcase.dtx} +% [1998/11/12 v0.06 Text only upper/lower case changing (DPC)] +% \end{quotation} +% +% David, in turn, suggests that Donald Arseneau be credited for some of the ideas in +% the \classname{textcase} package, so let it hereby be so understood. +% +% Note that David's package has a bug, which is fixed as follows: +% Change: +%\begin{verbatim} +% \def\NoCaseChange##1{\noexpand\NoCaseChange{\noexpand##1}} +%\end{verbatim} +% To: +%\begin{verbatim} +% \def\NoCaseChange##1{\noexpand\NoCaseChange{\noexpand##1}}% +%\end{verbatim} +% +% \begin{macrocode} +%\RequirePackage{textcase}% +% \end{macrocode} +% +% Next, restore those cammands we de-fanged: +% \begin{macrocode} +%<*class-pst> +\let\ProvidesPackage\ProvidesPackage@latex +\let\ProcessOptions\ProcessOptions@latex +\let\DeclareOption\DeclareOption@latex +% \end{macrocode} +% And let the \classname{textcase} commands override those of \LaTeX: +% \begin{macrocode} +\expandafter +\let\csname MakeUppercase \expandafter\endcsname + \csname MakeTextUppercase \endcsname +\expandafter +\let\csname MakeLowercase \expandafter\endcsname + \csname MakeTextLowercase \endcsname +% \end{macrocode} +% +% Try to patch \classname{longtable}. +% \begin{macrocode} +\appdef\class@documenthook{% + \switch@longtable +}% +% \end{macrocode} +% +% Try to provide a more robust version of \cmd\boldmath. +% If Team \LaTeX\ ever gets around to fixing this problem, we will bow out. +% (This relates to bug \#394.) +% \begin{macrocode} +\appdef\class@documenthook{% + \expandafter\@ifnotrelax\csname boldmath \endcsname{}{% + \expandafter\let\csname boldmath \endcsname\boldmath + \expandafter\def\expandafter\boldmath\expandafter{\expandafter\protect\csname boldmath \endcsname}% + }% +}% +% \end{macrocode} +% +% +% +% \section{Options}\label{sec:options}% +% +% \subsection{Define Booleans Used in Options}% +% The following Booleans are used within the document class to +% allow the document or the substyle to make selections +% of formatting. +% Because an explicit document class option always overrides a substyle, +% any substyle wishing to assign a value to a flag must first test the flag to +% confirm that it is still undefined. +% This further means that the default value of the flag must be +% assigned late, at the end of \cmd\AtBeginDocument\ time, and that +% the flag may be queried no earlier than that time. +% +% \begin{macro}{\class@inithook} +% The hook \cmd\class@inithook\ is installed into \cmd\class@documenthook, +% and thereafter serves as the focal point for all code that +% gives default values to the class's parameters. +% +% A typical code might be: +% \begin{verbatim} +%\appdef\class@inithook{% +% \@ifxundefined\twocolumn@sw{\@booleanfalse\twocolumn@sw}{}% +%}% +% \end{verbatim} +% which ensures that the switch \cmd\twocolumn@sw\ is never left undefined. +% +% \begin{macrocode} +\appdef\class@documenthook{% + \class@inithook +}% +\def\class@inithook{}% +% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\twocolumn@sw} +% The boolean \cmd\twocolumn@sw\ signifies that we are to use a two-column grid. +% +% \begin{macrocode} +\appdef\class@inithook{% + \@ifxundefined\twocolumn@sw{\@booleanfalse\twocolumn@sw}{}% +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\footinbib@sw} +% The boolean \cmd\footinbib@sw\ signifies that +% footnotes are to be set in the bibliography, as endnotes. +% \begin{macrocode} +\appdef\class@inithook{% + \@ifxundefined\footinbib@sw{\@booleanfalse\footinbib@sw}{}% +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\preprintsty@sw} +% The boolean \cmd\preprintsty@sw\ signifies that +% the document is to be formatted in preprint style. +% \begin{macrocode} +\appdef\class@inithook{% + \@ifxundefined\preprintsty@sw{\@booleanfalse\preprintsty@sw}{}% +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\preprint@sw} +% The boolean \cmd\preprint@sw\ signifies that +% the preprints (cf. \cmd\preprint) are to be formatted (usually on the title page). +% The default is to do so; the option \classoption{nopreprintnumbers} +% declares not to. +% \begin{macrocode} +\appdef\class@inithook{% + \@ifxundefined\preprint@sw{\@booleanfalse\preprint@sw}{}% + \preprintsty@sw{\@booleantrue\preprint@sw}{}% +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\galley@sw} +% The boolean \cmd\galley@sw\ signifies that +% the document is to be formatted in galley style. +% +% Asserting both \cmd\galley@sw\ and \cmd\preprintsty@sw\ +% may produce strange formatting results, but it is not illegal. +% However, it \emph{is} illegal to assert galley and any +% twocolumn option. +% \begin{macrocode} +\appdef\class@inithook{% + \@ifxundefined\galley@sw{\@booleanfalse\galley@sw}{}% +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\groupauthors@sw} +% The flag \cmd\groupauthors@sw\ signifies that authors are to be grouped. +% This affects the meaning of \cmd\@author@present\ and \cmd\@tempaffil. +% +% FIXME: apparently \cmd\@tempaffil\ and its aliases are never used. +% \end{macro} +% +% \begin{macro}{\runinaddress@sw} +% The flag \cmd\runinaddress@sw\ signifies that author addresses are to be +% run in. +% \end{macro} +% +% \begin{macro}{\@affils@sw} +% If \cmd\@affils@sw\ is false, an address is never recognised as ``new'' +% and is therefore always entered into the affiliation list, +% stopping groups of authors at the same address being amalgamated +% into the same list: +% the address will be printed the same number of times it is entered. +% \end{macro} +% +% \begin{macro}{\showPACS@sw} +% \begin{macro}{\showKEYS@sw} +% If \cmd\showPACS@sw\ is true, print the PACS information in the title block, +% otherwise not. +% Similarly for \cmd\showKEYS@sw\ and the keywords. +% \begin{macrocode} +\appdef\class@inithook{% + \@ifxundefined\showPACS@sw{\@booleanfalse\showPACS@sw}{}% + \@ifxundefined\showKEYS@sw{\@booleanfalse\showKEYS@sw}{}% +}% +% \end{macrocode} +% \end{macro} +% \end{macro} +% +% \begin{macro}{\floats@sw} +% The Boolean \cmd\floats@sw\ signifies that floats are to be floated; +% if false, that floats are to be deferred to the end of the document. +% By default, the former. +% Note that the state of this Boolean is to be changed by +% the document class in response to user-selected options. +% +% This boolean and the assignment of its default value is done by +% the \classname{ltxutil} package. +% \end{macro} +% +% \begin{macro}{\floatp@sw} +% The Boolean \cmd\floatp@sw\ signifies that endfloats are to be set one per page; +% if false, that endfloats are to be set with multiple floats per page permitted. +% By default, the latter. +% Note that the state of this Boolean is to be changed by +% the document class in response to user-selected options. +% The default is established here. +% \end{macro} +% +% \begin{macro}{\twoside@sw} +% The flag \cmd\twoside@sw\ signifies that the document is to be +% formatted for duplex printing. +% At \cmd\AtBeginDocument\ time, we must align the value +% of the kernel \cmd\newif\ switch \cmd\if@twoside\ +% to that of \cmd\twoside@sw. +% \begin{macrocode} +\appdef\class@inithook{% + \@ifxundefined\twoside@sw{\@booleanfalse\twoside@sw}{}% + \twoside@sw{\@twosidetrue}{\@twosidefalse}% +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\draft@sw} +% The flag \cmd\draft@sw\ signifies that the document is to be +% formatted in draft mode. +% Certain packages may pay attention to the +% class option \classoption{draft} that sets this Boolean. +% \begin{macrocode} +\appdef\class@inithook{% + \@ifxundefined\draft@sw{\@booleanfalse\draft@sw}{}% + \draft@sw{\overfullrule 5\p@}{\overfullrule\z@}% +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\tightenlines@sw} +% The boolean \cmd\tightenlines@sw\ signifies that +% the leading is to be made standard amount. +% If false, it means that the leading is to be set extra open. +% Has no effect on 10pt size option. +% \begin{macrocode} +\appdef\class@inithook{% + \@ifxundefined\tightenlines@sw{\@booleanfalse\tightenlines@sw}{}% +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\lengthcheck@sw} +% The flag \cmd\lengthcheck@sw\ signifies that +% the length checking is in effect. +% It is up to the individual journal substyle +% to alter its formatting accordingly. +% \begin{macrocode} +\appdef\class@inithook{% + \@ifxundefined\lengthcheck@sw{\@booleanfalse\lengthcheck@sw}{}% +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\eqsecnum@sw} +% The flag \cmd\eqsecnum@sw\ signifies that equations are to be +% numbered with the section, e.g., ``Eq. (2.13)''. +% This flag is neither set nor queried. +% \begin{macrocode} +\appdef\class@inithook{% + \@ifxundefined\eqsecnum@sw{\@booleanfalse\eqsecnum@sw}{}% + \eqsecnum@sw{% + \@addtoreset{equation}{section}% + \def\theequation@prefix{\arabic{section}.}% + }{}% +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\byrevtex@sw} +% The flag \cmd\byrevtex@sw\ signifies that the document should +% bear an imprint to the effect that it was formatted by this document class. +% \begin{macrocode} +\appdef\class@inithook{% + \@ifxundefined\byrevtex@sw{\@booleanfalse\byrevtex@sw}{}% +}% +% \end{macrocode} +% \end{macro} +% +% +% \subsection{Declare Options}% +% +% \subsubsection{Checkin: for Editorial Use}% +% A document class option declaring that the document is being processed by the +% editorial staff. +% +% This option should: +% \begin{itemize} +% \item +% put date in footer along with folio +% \item +% Have the effect of selecting the \classoption{preprint} class option. +% \item +% Have the effect of selecting the \classoption{showpacs} class option. +% \item +% specify that when a float is placed h or H, it will be allowed to break over pages. +% (Note: be sure that if the enclosed \env{tabular} has an optional argument, you change it to [v], or remove it entirely.) +% \end{itemize} +% \changes{v4.0beta 3}{1999/11/13} +% {New option ``checkin''} +% \begin{macrocode} +\DeclareOption{checkin}{% + \@booleantrue\dateinRH@sw + \@booleantrue\preprintsty@sw + \def\@pointsize{12}% + \@booleantrue\showPACS@sw + \@booleantrue\showKEYS@sw + \def\fp@proc@h{\allow@breaking@tables}% + \def\fp@proc@H{\allow@breaking@tables}% +}% +\@booleanfalse\dateinRH@sw +\def\checkindate{\dateinRH@sw{{\tiny(\today)}}{}}% +\def\allow@breaking@tables{% + \def\array@default{v}% tabular can break over pages + \@booleanfalse\floats@sw % table can break over pages +}% +% \end{macrocode} +% +% \subsubsection{Preprint Style}% +% \changes{v4.d}{1998/01/31} +% {Rearrange the ordering so numerical ones come first. AO: David, what does this mean?} +% +% \changes{v4.g}{1998/06/10} +% {comma not space between email and homepage} +% +% \changes{v4.0beta 2}{1999/06/20} +% {Move this ``complex'' option to the front, where it can be overridden by ``simple'' options.} +% \begin{macrocode} +\DeclareOption{preprint}{% + \@booleantrue\preprintsty@sw + \def\@pointsize{12}% +}% +\DeclareOption{manuscript}{% + \class@warn{Document class option manuscript is obsolete; use preprint instead}% + \ExecuteOptions{preprint}% +}% +% \end{macrocode} +% +% \subsubsection{Showing PACS and keywords}% +% \changes{v4.0beta 2}{1999/06/20} +% {New option} +% \changes{v4.0beta 5}{2000/09/20} +% {New option showkeys} +% \begin{macrocode} +\DeclareOption{showpacs}{% + \@booleantrue\showPACS@sw +}% +\DeclareOption{noshowpacs}{% + \@booleanfalse\showPACS@sw +}% +\DeclareOption{showkeys}{% + \@booleantrue\showKEYS@sw +}% +\DeclareOption{noshowkeys}{% + \@booleanfalse\showKEYS@sw +}% +% \end{macrocode} +% +% \subsubsection{Balance the last page when in two-column page grid}% +% If we are in a two-column page grid, we may wish to balance the columns +% of the last page. This will be done automatically if the +% \classoption{twocolumn} document class option is chosen. +% This action will be turned off by the \classoption{nobalancelastpage}. +% A complementary class option, \classoption{balancelastpage} is also provided. +% \begin{macrocode} +\DeclareOption{balancelastpage}{% + \@booleantrue\balancelastpage@sw +}% +\DeclareOption{nobalancelastpage}{% + \@booleanfalse\balancelastpage@sw +}% +\appdef\class@inithook{% + \@ifxundefined\balancelastpage@sw{% + \@booleantrue\balancelastpage@sw + }{}% +}% +% \end{macrocode} +% +% \subsubsection{Showing preprint numbers}% +% \changes{v4.0beta 4}{2000/04/10} +% {New option} +% \begin{macrocode} +\DeclareOption{nopreprintnumbers}{\@booleanfalse\preprint@sw}% +\DeclareOption{preprintnumbers}{\@booleantrue\preprint@sw}% +% \end{macrocode} +% +% \subsubsection{Hypertext Option}% +% +% \changes{v4.e}{1998/01/31} +% {protect against hyperref revtex kludges which are not needed now} +% \changes{v4.0beta 2}{1999/06/20} +% {Move this ``complex'' option to the front, where it can be overridden by ``simple'' options.} +% \changes{v4.0rc4}{2001/07/23} +% {hyperref is no longer loaded via class option: use a usepackage statement instead} +% \begin{macrocode} +\DeclareOption{hyperref}{% +%\AtEndOfClass{% +% \begingroup +% \edef\@tempa{% +% \let +% \noexpand\@clsextension +% \noexpand\@empty +% \noexpand\RequirePackage{hyperref}% +% \def\noexpand +% \@clsextension{\@clsextension}% +% }% +% \expandafter +% \endgroup +% \@tempa +%}% +% \end{macrocode} +% +% If you have a hyper-foo enabled browser you may prefer this format +% which does not print the URL for the home page, but just makes the +% name a link, but by default print it so it works on paper. +% \changes{v4.e}{1998/01/31} +% {print homepage} +% +% \begin{macrocode} +% \def\@pointsize{10}% + \class@warn{Class option "hyperref" is no longer supported.^^JEmploy \string\usepackage{hyperref} instead}% +}% +% \end{macrocode} +% +% \subsubsection{Type Size}% +% Use \cmd\@pointsize|=10| rather than \cmd\@ptsize|=0| to allow easy extensions +% to 9pt or whatever. +% Note: the three alternatives are mutually exclusive. Furthermore +% if \cmd\@pointsize\ is still undefined at \cmd\AtEndOfClass\ time, +% then the substyle can conclude that the user accepts the default. +% \begin{macrocode} +\DeclareOption{10pt}{\def\@pointsize{10}} +\DeclareOption{11pt}{\def\@pointsize{11}} +\DeclareOption{12pt}{\def\@pointsize{12}} +% \end{macrocode} +% +% \subsubsection{Media Size}% +% \changes{v4.0beta 2}{1999/06/20} +% {Restore all media size class option of \protect\file{classes.dtx}} +% \begin{macrocode} +\DeclareOption{a4paper} + {\setlength\paperheight {297mm}% + \setlength\paperwidth {210mm}} +\DeclareOption{a5paper} + {\setlength\paperheight {210mm}% + \setlength\paperwidth {148mm}} +\DeclareOption{b5paper} + {\setlength\paperheight {250mm}% + \setlength\paperwidth {176mm}} +\DeclareOption{letterpaper} + {\setlength\paperheight {11in}% + \setlength\paperwidth {8.5in}} +\DeclareOption{legalpaper} + {\setlength\paperheight {14in}% + \setlength\paperwidth {8.5in}} +\DeclareOption{executivepaper} + {\setlength\paperheight {10.5in}% + \setlength\paperwidth {7.25in}} +\DeclareOption{landscape} + {\setlength\@tempdima {\paperheight}% + \setlength\paperheight {\paperwidth}% + \setlength\paperwidth {\@tempdima}} +% \end{macrocode} +% +% \subsubsection{Bibnotes}% +% Frontmatter footnotes result from frontmatter commands like +% \cmd\email, \cmd\homepage, \cmd\altaffiliation, and \cmd\thanks. +% The default for \cmd\frontmatter@footnote@produce\ is +% \cmd\frontmatter@footnote@produce@footnote, which formats the frontmatter footnotes +% at the foot of the title page. The \classoption{bibnotes} class option +% defers them to the bibliography. +% \begin{macrocode} +\DeclareOption{bibnotes}{\@booleantrue\bibnotes@sw}% +\DeclareOption{nobibnotes}{\@booleanfalse\bibnotes@sw}% +\appdef\class@enddocumenthook{\auto@bib}% +\appdef\class@inithook{% + \@ifxundefined\bibnotes@sw{\@booleanfalse\bibnotes@sw}{}% + \bibnotes@sw{% + \let\frontmatter@footnote@produce\frontmatter@footnote@produce@endnote + }{% + \let\frontmatter@footnote@produce\frontmatter@footnote@produce@footnote + }% +}% +% \end{macrocode} +% +% \subsubsection{Footinbib}% +% \begin{macrocode} +\DeclareOption{footinbib}{\@booleantrue\footinbib@sw} +\DeclareOption{nofootinbib}{\@booleanfalse\footinbib@sw} +% \end{macrocode} +% +% \subsubsection{altaffilletter}% +% +% \begin{macrocode} +\DeclareOption{altaffilletter}{\@booleantrue\altaffilletter@sw}% +\DeclareOption{altaffilsymbol}{\@booleanfalse\altaffilletter@sw}% +\appdef\class@inithook{% + \@ifxundefined\altaffilletter@sw{\@booleanfalse\altaffilletter@sw}{}% +}% +% \end{macrocode} +% +% \subsubsection{superbib}% +% The document may invoke superscript bibliography style. +% If not, the journal substyle may make a selection. +% If neither, the class itself selects a default. +% This is really a quite deep nesting of defaults. +% \begin{macrocode} +\DeclareOption{superbib}{% + \def\place@bibnumber{\place@bibnumber@sup}% +}% +% \end{macrocode} +% +% \subsubsection{citeautoscript}% +% This class option allows you to automatically accomodate a change from non-superscripted, numbered references +% to superscripted, bunmbered references. +% +% Note: you should always mark up your document with the assumption that references are \emph{not} going to be superscripted. +% Otherwise this option has no hope of working properly. +% +% \begin{macrocode} +\DeclareOption{citeautoscript}{% + \@booleantrue\citeautoscript@sw +}% +\@booleanfalse\citeautoscript@sw +% \end{macrocode} +% +% \subsubsection{Simplex/Duplex Pages}% +% \changes{v4.e}{1998/01/31} +% {4.d had twoside option setting twoside switch to false} +% The complementary options \classoption{twoside} and \classoption{oneside} +% assert formatting for duplex or simplex printing, respectively. +% At the same time, we arrange for the selection of the page grid +% with respect to the marginal column: +% Because \cmd\if@reversemargin\ remains default (false), +% if duplex printing, this column will always be on the (right), +% if simplex printing, it will always be on the (outside). +% QUERY: correct choice? +% \begin{macrocode} +\DeclareOption{twoside}{\@booleantrue\twoside@sw \@mparswitchfalse} +\DeclareOption{oneside}{\@booleanfalse\twoside@sw\@mparswitchtrue} +% \end{macrocode} +% +% \subsubsection{Two-Column Page Grid}% +% +% \begin{macro}{\twocolumn@sw} +% The flag \cmd\twocolumn@sw\ signifies that the document is to be +% formatted in the two-column page grid. +% +% If no options relating to page grid are invoked by +% \cmd\AtBeginDcoument\ time, we set default values. +% Up to that point, the class can check if \cmd\twocolumn@sw\ +% is \cmd\undefined\ to see if any related options have been invoked. +% +% At \cmd\AtBeginDocument\ time, we must align the value +% of the kernel \cmd\newif\ switch \cmd\if@twocolumn\ +% to that of \cmd\twocolumn@sw. +% +% We install into \cmd\class@enddocumenthook\ +% all tasks that we know must be completed at \cmd\AtEndDocument\ time. +% The first thing it does is end the current paragraph. +% The last is +% to return us to one-column page grid. +% (This task is installed as late as possible, +% i.e., after items like \cmd\printtables.) +% \begin{macrocode} +\AtEndOfClass{% + \def\clear@document{% + \close@column + }% + \appdef\class@inithook{% + \@ifxundefined\close@column{% + \let\close@column\close@column@default + }{}% + }% +}% +% \end{macrocode} +% \end{macro} +% +% The \classoption{twocolumn} class option employs the standard mechanism for +% changing the column grid: the \classname{ltxgrid} package. +% \begin{macrocode} +\DeclareOption{twocolumn}{% + \@booleantrue\twocolumn@sw +}% +% \end{macrocode} +% +% \classoption{onecolumn} specifies one-column page grid. +% \begin{macrocode} +\DeclareOption{onecolumn}{% + \@booleanfalse\twocolumn@sw +}% +% \end{macrocode} +% +% \classoption{galley} emulates setting the galleys of +% a two-column journal. +% FIXME: this option should effectively set \cmd\preprintsty@sw\ false. +% NOTE: it makes no sense to assert both galley and twocolumn. +% +% \begin{macrocode} +\DeclareOption{galley}{% + \ExecuteOptions{onecolumn}% + \@booleantrue\galley@sw + \appdef\set@pica@hook{% + \advance\textwidth-\columnsep + \textwidth.5\textwidth + }% +}% +% \end{macrocode} +% +% +% \subsubsection{raggedbottom or flushbottom}% +% The class options \classoption{raggedbottom} and \classoption{flushbottom} determine whether the columns (page) +% are ragged bottom or flush bottom. +% Note that we do not select a default here; that is done by the journal substyle. +% \begin{macrocode} +\DeclareOption{raggedbottom}{\@booleantrue\raggedcolumn@sw} +\DeclareOption{flushbottom}{\@booleanfalse\raggedcolumn@sw} +% \end{macrocode} +% +% \subsubsection{tightenlines}% +% This class option specifies that standard leading +% is to be used to set the type. +% If lacking, the leading will be loose. +% +% \begin{macrocode} +\DeclareOption{tightenlines}{\@booleantrue\tightenlines@sw} +% \end{macrocode} +% +% \subsubsection{lengthcheck}% +% This class option specifies that the formatted document +% should approach as closely as possible the +% formatting of an actual journal article +% to facilitate the author's performance of a length check. +% +% \begin{macrocode} +\DeclareOption{lengthcheck}{% + \@booleantrue\lengthcheck@sw + \@booleantrue\twoside@sw + \@booleantrue\twocolumn@sw +}% +% \end{macrocode} +% +% \subsubsection{Draft and Final}% +% \begin{macrocode} +\DeclareOption{draft}{% + \@booleantrue\draft@sw +}% +\DeclareOption{final}{% + \@booleanfalse\draft@sw +}% +% \end{macrocode} +% +% \subsubsection{eqsecnum}% +% The \classoption{eqsecnum} class option signifies that equations are +% to be numbered within sections. +% \begin{macrocode} +\DeclareOption{eqsecnum}{\@booleantrue\eqsecnum@sw} +% \end{macrocode} +% +% \subsubsection{secnumarabic}% +% The \classoption{secnumarabic} class option signifies that sectioning commands are +% to be numbered arabic. +% \begin{macrocode} +\DeclareOption{secnumarabic}{\@booleantrue\secnumarabic@sw} +% \end{macrocode} +% The code that polls the Boolean \cmd\secnumarabic@sw\ appears +% in Section~\ref{sec:secnumarabic}. +% +% FIXME: model fleqn after amsfonts. I no longer understand why I said this. +% \begin{macrocode} +\DeclareOption{fleqn}{\input{fleqn.clo}} +% \end{macrocode} +% +% \subsubsection{floats/endfloats}% +% These options control, via the Boolean \cmd\floats@sw, +% whether floats are to be migrated to the end of the document. +% \begin{macrocode} +\DeclareOption{floats}{\@booleantrue\floats@sw\@booleanfalse\floatp@sw} +\DeclareOption{endfloats}{\@booleanfalse\floats@sw\@booleanfalse\floatp@sw} +\DeclareOption{endfloats*}{\@booleanfalse\floats@sw\@booleantrue\floatp@sw} +\appdef\class@inithook{% + \@ifxundefined\floats@sw{\@booleantrue\floats@sw}{}% + \@ifxundefined\floatp@sw{\@booleantrue\floatp@sw}{}% +}% +% \end{macrocode} +% +% \subsubsection{titlepage/notitlepage}% +% These options control, via \cmd\titlepage@sw, +% whether the title block is to be set on a separate page. +% \begin{macro}{\titlepage@sw} +% The flag \cmd\titlepage@sw\ signifies that a forced +% page break is to follow the title page: the article +% title appears on a page by itself. +% +% The Boolean \cmd\titlepage@sw\ is in an odd situation: +% its value is set by one or more substyles, contingent +% upon the value of another Boolean. +% Therefore that code must be invoked at \cmd\AtBeginDocument\ time. +% Therefore we must assign the default value at the very end +% of \cmd\AtBeginDocument\ processing: we use \cmd\defaults@hook. +% Luckily, this Boolean does not need to be queried at +% \cmd\AtBeginDocument\ time, or we'd be in dire straights. +% \begin{macrocode} +\DeclareOption{titlepage}{\@booleantrue\titlepage@sw} +\DeclareOption{notitlepage}{\@booleanfalse\titlepage@sw} +\appdef\defaults@hook{% + \@ifxundefined\titlepage@sw{\@booleanfalse\titlepage@sw}{}% +}% +% \end{macrocode} +% \end{macro} +% +% \subsubsection{Substyle and Sub-substyle}% +% +% If the society has already been assigned, +% notify user that \cmd\@society\ is being overridden. +% +% \begin{macrocode} +\def\change@society#1{% + \def\@tempa{#1}% + \@ifxundefined\@society{}{% + \@ifx{\@tempa\@society}{}{% + \class@warn{Conflicting society \@tempa<>\@society.}% + }% + }% + \let\@society\@tempa +}% +% \end{macrocode} +% Here are the class options relating to the APS +% \begin{macrocode} +\DeclareOption{aps}{\change@society{aps}\def\@journal{}}% +\DeclareOption{pra}{\change@society{aps}\def\@journal{pra}}% +\DeclareOption{prb}{\change@society{aps}\def\@journal{prb}}% +\DeclareOption{prc}{\change@society{aps}\def\@journal{prc}}% +\DeclareOption{prd}{\change@society{aps}\def\@journal{prd}}% +\DeclareOption{pre}{\change@society{aps}\def\@journal{pre}}% +\DeclareOption{prl}{\change@society{aps}\def\@journal{prl}}% +\DeclareOption{prstab}{\change@society{aps}\def\@journal{prstab}}% +\DeclareOption{rmp}{\change@society{aps}\def\@journal{rmp}}% +% \end{macrocode} +% Here are the class options relating to the Optical Society of America +% \begin{macrocode} +\DeclareOption{osa}{\change@society{osa}\def\@journal{}}% +\DeclareOption{osameet}{\change@society{osa}\def\@journal{osameet}}% +\DeclareOption{opex}{\change@society{osa}\def\@journal{opex}}% +\DeclareOption{tops}{\change@society{osa}\def\@journal{tops}}% +\DeclareOption{josa}{\change@society{osa}\def\@journal{josa}}% +% \end{macrocode} +% +% NOTE: \file{revguide.tex} does not mention any \cmd\@journal\ options +% besides \classoption{prb}. +% +% \begin{macro}{amsfonts} +% The class option \classoption{amsfonts} has the +% same effect as if the document preamble contained +% a \cmd\usepackage\verb${amsfonts}$ statement. +% \begin{macrocode} +\DeclareOption{amsfonts}{% + \def\class@amsfonts{\RequirePackage{amsfonts}}% +}% +\DeclareOption{noamsfonts}{% + \let\class@amsfonts\@empty +}% +\AtEndOfClass{% + \@ifxundefined\class@amsfonts{}{\class@amsfonts}% +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{amssymb} +% The class option \classoption{amssymb} has the +% same effect as if the document preamble contained +% a \cmd\usepackage\verb${amssymb}$ statement. +% \begin{macrocode} +\DeclareOption{amssymb}{% + \def\class@amssymb{\RequirePackage{amssymb}}% +}% +\DeclareOption{noamssymb}{% + \let\class@amssymb\@empty +}% +\AtEndOfClass{% + \@ifxundefined\class@amssymb{}{\class@amssymb}% +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{amsmath} +% The class option \classoption{amsmath} has the +% same effect as if the document preamble contained +% a \cmd\usepackage\verb${amsmath}$ statement. +% +% We require version 1.2 (datestamped 1997/03/20) or later. +% The \expandafter\cmd\csname ver@amsmath.sty\endcsname, will +% +% \LaTeX\ note: +% Certain \LaTeX\ procedures have an arbitrary and pointless restriction that they may be used only within the preamble. +% We get around this by preserving the procedures in private \cmd\csname s. +% +% FIXME note: it is difficult to ensure that an error summary will be printed on the console at the very end, +% but \classname{ltxgrid} allows accomplishing this via an interrupt, put down at \cmd\AtEndDocument\ time. +% +% \begin{macrocode} +\DeclareOption{amsmath}{% + \def\class@amsmath{\RequirePackage{amsmath}[\ver@amsmath@prefer]}% +}% +\DeclareOption{noamsmath}{% + \let\class@amsmath\@empty +}% +\AtEndOfClass{% + \preserve@LaTeX + \@ifxundefined\class@amsmath{}{\class@amsmath}% + \appdef\class@enddocumenthook{\test@amsmath@ver}% +}% +\appdef\preserve@LaTeX{% + \let\@ifl@aded@LaTeX\@ifl@aded + \let\@ifpackageloaded@LaTeX\@ifpackageloaded + \let\@pkgextension@LaTeX\@pkgextension + \let\@ifpackagelater@LaTeX\@ifpackagelater + \let\@ifl@ter@LaTeX\@ifl@ter + \let\@ifl@t@r@LaTeX\@ifl@t@r + \let\@parse@version@LaTeX\@parse@version +}% +\appdef\restore@LaTeX{% + \let\@ifl@aded\@ifl@aded@LaTeX + \let\@ifpackageloaded\@ifpackageloaded@LaTeX + \let\@pkgextension\@pkgextension@LaTeX + \let\@ifpackagelater\@ifpackagelater@LaTeX + \let\@ifl@ter\@ifl@ter@LaTeX + \let\@ifl@t@r\@ifl@t@r@LaTeX + \let\@parse@version\@parse@version@LaTeX +}% +\def\test@amsmath@ver{% + \begingroup + \restore@LaTeX + \@ifpackageloaded{amsmath}{% + \@ifpackagelater{amsmath}{\ver@amsmath@prefer}{}{% + \class@warn{% + You have loaded amsmath, version "\csname ver@amsmath.sty\endcsname",\MessageBreak + but this class requires version "\ver@amsmath@prefer", or later.\MessageBreak + Please update your LaTeX installation. + }% + }% + }{% + }% + \endgroup +}% +\def\ver@amsmath@prefer{2000/01/15 v2.05 AMS math features}% +% \end{macrocode} +% \end{macro} +% +% \subsubsection{Presenting Authors and Their Affiliations}% +% \changes{v4.0beta 2}{1999/06/20} +% {These options are now mutually exclusive. But is this correct?} +% +% The \classoption{runinaddress} option is the only one that sets \cmd\runinaddress@sw\ to true. +% +% The \classoption{unsortedaddress} option is the only one that sets \cmd\@affils@sw\ to false. +% +% If we take the \classoption{groupedaddress} option as the default, then we can look upon +% the \classoption{superscriptaddress} option as simply turning \cmd\groupauthors@sw\ to false. +% +% \classoption{groupedaddress}, the default, groups authors above a common address. +% In order to assert the default and still provide for proper precedence +% for document-supplied options, we carry out our action at \cmd\AtBeginDocument\ time, +% and only if none of the related class options were invoked. +% \begin{macrocode} +\DeclareOption{groupedaddress}{\clo@groupedaddress}% +\def\clo@groupedaddress{% + \@booleantrue\groupauthors@sw + \@booleantrue\@affils@sw + \@booleanfalse\runinaddress@sw +}% +\appdef\class@inithook{% + \@ifxundefined\groupauthors@sw{\clo@groupedaddress}{}% +}% +% \end{macrocode} +% +% \classoption{unsortedaddress} is similar to \classoption{groupedaddress}, +% but turns off \cmd\@affils@sw, +% with the result that each address that is entered will be printed. +% \begin{macrocode} +\DeclareOption{unsortedaddress}{\clo@unsortedaddress}% +\def\clo@unsortedaddress{% + \@booleantrue\groupauthors@sw + \@booleanfalse\@affils@sw + \@booleanfalse\runinaddress@sw +}% +% \end{macrocode} +% +% \classoption{runinaddress} is similar to \classoption{groupedaddress}, +% with the addition of the Boolean \cmd\runinaddress@sw, +% which causes the authors associated with a particular address group +% to be formatted in a paragraph instead of on separate lines. +% \begin{macrocode} +\DeclareOption{runinaddress}{\clo@runinaddress}% +\def\clo@runinaddress{% + \@booleantrue\groupauthors@sw + \@booleantrue\@affils@sw + \@booleantrue\runinaddress@sw +}% +% \end{macrocode} +% +% \classoption{superscriptaddress} presents author affiliations +% as superscripts. Authors with like affiliations share the same +% superscript. +% +% FIXME: if \cmd\groupauthors@sw\ is false, would not \cmd\@affils@sw\ be a don't care? +% \begin{macrocode} +\DeclareOption{superscriptaddress}{\clo@superscriptaddress}% +\def\clo@superscriptaddress{% + \@booleanfalse\groupauthors@sw + \@booleantrue\@affils@sw + \@booleanfalse\runinaddress@sw +}% +% \end{macrocode} +% +% +% \subsubsection{Typeset by \revtex}% +% +% The class option \classoption{byrevtex} signifies that you want the +% ``Typeset by \revtex'' byline to appear on your formatted output. +% By default, no such byline appears. +% \begin{macrocode} +\DeclareOption{byrevtex}{\@booleantrue\byrevtex@sw}% +% \end{macrocode} +% +% +% \subsection{Attempt to fix float placement failure}% +% +% \revtex\ uses the \classname{ltxgrid} package, which provides the ability to attempt repairs when +% \LaTeX's float placement mechanism is about to fail, but that facility is turned off by default. +% Users should invoke the \classoption{floatfix} document class option to enable this \LaTeX\ extension. +% If not, a helpful message is printed in the log, indicating how to work around the difficulty. +% \begin{macrocode} +\DeclareOption{floatfix}{\@booleantrue\force@deferlist@sw}% +\DeclareOption{nofloatfix}{\@booleanfalse\force@deferlist@sw}% +\appdef\class@inithook{% + \@ifxundefined\force@deferlist@sw{\@booleanfalse\force@deferlist@sw}{}% +}% +% \end{macrocode} +% +% \begin{macro}{\@fltovf} +% \begin{macro}{\@fltstk} +% The \LaTeX\ kernel error message \cmd\@fltovf\ may now be a bit more helpful to the user; +% likewise for the \cmd\@fltstk\ message of \classname{ltxgrid}. +% \begin{macrocode} +\gdef\@fltovf{% + \@latex@error{% + Too many unprocessed floats% + \force@deferlist@sw{}{; try class option [floatfix]}% + }\@ehb +}% +\def\@fltstk{% + \@latex@warning{% + A float is stuck (cannot be placed)% + \force@deferlist@sw{}{; try class option [floatfix]}% + }% +}% +% \end{macrocode} +% \end{macro} +% \end{macro} +% +% \begin{macro}{\ltxgrid@info@sw} +% \begin{macro}{\outputdebug@sw} +% The two options \classoption{ltxgridinfo} and \classoption{outputdebug} turn on informative diagnostics within the +% package \classname{ltxgrid}. Only people who really want to see this output will select these class options. +% Consult documentation for the \classname{ltxgrid} package to see what output the related switches enable. +% \begin{macrocode} +\DeclareOption{ltxgridinfo}{\@booleantrue\ltxgrid@info@sw}% +\DeclareOption{outputdebug}{\@booleantrue\outputdebug@sw}% +% \end{macrocode} +% \end{macro} +% \end{macro} +% +% \subsection{Option to relax page height}% +% +% The \classname{ltxgrid} package can set text pages to their natural height or force them to the full +% text height; the latter is the default. +% If setting the pages with a variable length, +% the running foot will move up or down with the natural length of the text column. +% While I recommend against doing so, this option will turn that switch to the latter setting. +% \begin{macrocode} +\DeclareOption{raggedfooter}{\@booleanfalse\textheight@sw}% +\DeclareOption{noraggedfooter}{\@booleantrue\textheight@sw}% +% \end{macrocode} +% +% \subsection{Selecting procedure for processing abstract}% +% +% There are two different procedures for processing an abstract; the options +% \classoption{newabstract} and \classoption{oldabstract} determine which one the user +% prefers. At present, the default is to use the old, but this may change. +% \begin{macrocode} +\DeclareOption{newabstract}{\@booleanfalse\oldabstract@sw}% +\DeclareOption{oldabstract}{\@booleantrue \oldabstract@sw}% +\appdef\class@inithook{% + \@ifxundefined\oldabstract@sw{\@booleanfalse\oldabstract@sw}{}% + \oldabstract@sw{% + \let@environment{abstract}{abstract-old}% + }{}% +}% +% \end{macrocode} +% +% +% +% +% \subsection{Default Option}% +% +% The default handling for a document class option is to +% treat it as the name of a society and +% hunt for a corresponding \file{.rtx} file to read in. +% This behavior is much like the \LaTeX2.09 handling, where +% one looked for a \file{.sty} file, except that +% in this case, we must catch the case where multiple +% unknown options are declared. +% +% \begin{macrocode} +\DeclareOption*{% + \IfFileExists{\CurrentOption\substyle@ext}{% + \expandafter\change@society\expandafter{\CurrentOption}% + }{% + \OptionNotUsed + }% +}% +\def\substyle@ext{.rtx}% +\AtEndOfClass{% + \@ifxundefined\@society{% + \class@warn{No Society specified, using generic APS style}% + \def\@society{aps}\def\@journal{}% + }{}% + \expandafter\input\expandafter{\@society\substyle@ext}% +}% +% \end{macrocode} +% +% \subsection{Class-Asserted Options}% +% \changes{v4.0beta 2}{1999/06/20} +% {% +% Removed invocation of nonexistent class option \protect\classoption{groupauthors} +% and all other class options that should only be invoked by the document. +% (Otherwise precedence of class options does not work.) +% } +% Here we establish the default document class options. +% Those of the document itself will override these. +% \begin{macrocode} +\ExecuteOptions{10pt} +% \end{macrocode} +% +% \subsection{Execute Options}% +% The document class options are processed in the order +% they are declared in this document class file. +% +% Precedence is established in the following way: +% +% First, the single \file{.rtx} file is read in. +% It may set values to any of the Booleans that represent class options, +% and it may elect to do so only if that Boolean is undefined, signifying +% that its corresponding document class options were absent from +% the document's option list. +% In this way correct precedence of class options is maintained. +% +% Second, the type size option is acted upon. Type size code +% should not require altering the values of any of the Booleans carrying +% class options. +% +% Next, at \cmd\AtEndOfClass\ time, certain packages are read in. +% +% Finally, at \cmd\AtBeginDocument\ time, the state of the Booleans determines +% how the formatting will go. +% +% \begin{macrocode} +\ProcessOptions\relax +% \end{macrocode} +% +% +% \section{Procedures Dependent Upon Options} +% +% Here we introduce \classname{classes.dtx} definitions for +% the page styles that people will expect to be able to use. +% \begin{macro}{\ps@headings} +% \begin{macro}{\ps@myheadings} +% \begin{macrocode} + \def\ps@headings{% + \let\@oddfoot\@empty\let\@evenfoot\@empty + \def\@evenhead{\thepage\hfil\slshape\leftmark}% + \def\@oddhead{{\slshape\rightmark}\hfil\thepage}% + \let\@mkboth\markboth + \def\sectionmark##1{% + \markboth {\MakeUppercase{% + \ifnum \c@secnumdepth >\z@ + \thesection\quad + \fi + ##1}}{}}% + \def\subsectionmark##1{% + \markright {% + \ifnum \c@secnumdepth >\@ne + \thesubsection\quad + \fi + ##1}}}% +\def\ps@myheadings{% + \let\@oddfoot\@empty\let\@evenfoot\@empty + \def\@evenhead{\thepage\hfil\slshape\leftmark}% + \def\@oddhead{{\slshape\rightmark}\hfil\thepage}% + \let\@mkboth\@gobbletwo + \let\sectionmark\@gobble + \let\subsectionmark\@gobble + }% +% \end{macrocode} +% \end{macro} +% \end{macro} +% +% \begin{macro}{\ps@article@twoside} +% \begin{macro}{\ps@article@oneside} +% \begin{macro}{\ps@preprint} +% \changes{v4.e}{1998/01/31} +% {Move after process options, so the following test works} +% \changes{v4.0beta 2}{1999/06/20} +% {Defer decision until \cs{AtBeginDocument} time} +% \changes{v4.0beta 3}{1999/11/13} +% {Insert procedure \cs{checkindate}} +% \changes{v4.0rc1}{2001/06/17} +% {Running headers always as if two-sided} +% \begin{macrocode} +\def\ps@article{% + \@provide\@evenhead{\let\\\heading@cr\thepage\quad\checkindate\hfil{\leftmark}}% + \@provide\@oddhead{\let\\\heading@cr{\rightmark}\hfil\checkindate\quad\thepage}% + \@provide\@oddfoot{}% + \@provide\@evenfoot{}% + \let\@mkboth\markboth + \let\sectionmark\@gobble + \let\subsectionmark\@gobble +}% +\def\ps@article@final{% + \@provide\@evenhead{\let\\\heading@cr\thepage\quad\checkindate\hfil{\leftmark}}% + \@provide\@oddhead{\let\\\heading@cr{\rightmark}\hfil\checkindate\quad\thepage}% + \@provide\@oddfoot{}% + \@provide\@evenfoot{}% + \let\@mkboth\markboth + \def\sectionmark##1{% + \markboth{% + \MakeTextUppercase{% + \@ifnum{\c@secnumdepth >\z@}{\thesection\hskip 1em\relax}{}% + ##1% + }% + }{}% + }% + \def\subsectionmark##1{% + \markright {% + \@ifnum{\c@secnumdepth >\@ne}{\thesubsection\hskip 1em\relax}{}% + ##1% + }% + }% +}% +\appdef\class@inithook{% +%\twoside@sw{% +%}{% +%}% +}% +% \end{macrocode} +% +% \begin{macrocode} +\def\heading@cr{\unskip\space\ignorespaces}% +% \end{macrocode} +% +% \changes{v4.0beta 3}{1999/11/13} +% {Do not put by REVTeX in every page foot} +% \changes{v4.0beta 3}{1999/11/13} +% {Insert procedure \cs{checkindate}} +% \begin{macrocode} +\def\ps@preprint{% + \@provide\@oddfoot{\hfil\thepage\quad\checkindate\hfil}% + \@provide\@evenfoot{\hfil\thepage\quad\checkindate\hfil}% + \@provide\@oddhead{}% + \@provide\@evenhead{}% + \let\@mkboth\@gobbletwo + \let\sectionmark\@gobble + \let\subsectionmark\@gobble +}% +% \end{macrocode} +% +% \begin{macrocode} +\let\@oddhead\@undefined +\let\@evenhead\@undefined +\let\@oddfoot\@undefined +\let\@evenfoot\@undefined +\AtEndOfClass{% + \appdef\class@documenthook{% + \@provide\@oddhead{}% + \@provide\@evenhead{}% + \@provide\@oddfoot{}% + \@provide\@evenfoot{}% + }% +}% +% \end{macrocode} +% \end{macro} +% \end{macro} +% \end{macro} +% +% \begin{macro}{\lastpage@putlabel} +% Support the default meaning of \cmd\@endpage. +% \changes{v4.d}{1998/01/31} +% {Move after process options, so \cs{clearpage} not in scope of +% twocolumn} +% Name of this macro (and the \cmd\label\ key) taken from +% \url{CTAN:/macros/latex/contrib/other/lastpage}% +% with code optimised slightly. +% \begin{macrocode} +\def\lastpage@putlabel{% + \if@filesw + \begingroup + \advance\c@page\m@ne + \immediate\write\@auxout{\string\newlabel{LastPage}{{}{\thepage}}}% + \endgroup + \fi +}% +% \end{macrocode} +% Install a late hook into \cmd\class@enddocumenthook\ that +% labels the last page of the document. +% This is done just before the \file{.aux} file +% is closed, and does not require a \cmd\shipout, because it writes +% directly to the \file{.aux} file. +% Note that we invoke \cmd\clearpage: this means that no more typesetting +% should be done past this point. +% CHECKME: are there any more things installed via +% \cmd\AtEndOfClass\allowbreak\cmd\class@enddocumenthook? +% \begin{macrocode} +\AtEndOfClass{% + \appdef\clear@document{% + \clearpage + \lastpage@putlabel + }% +}% +% \end{macrocode} +% \end{macro} +% +% +% \section{Required Packages} +% \url{CTAN:macros/latex/contrib/other/misc/url.sty}% +% \begin{macrocode} +\RequirePackage{url}% +% \end{macrocode} +% +% \section{Incompatible Packages} +% Known to be incompatible are \classname{multicol} and \classname{cite}. +% The former is incompatible with \classname{ltxgrid}, which replaces it. +% The latter is incompatible with \classname{natbib}, which replaces its functionality. +% \begin{macrocode} +\appdef\class@inithook{% + \incompatible@package{cite}% + \incompatible@package{multicol}% +}% +% \end{macrocode} +% +% \section{Society- and Journal-Specific Code} +% +% \begin{macro}{\@journal} +% Journal test helper, used as +% \begin{verbatim} +%\@ifx{\@journal\journal@pra}{% +% <journal-specific setup> +%\fi +% \end{verbatim} +% Journal code might like to further specify (if as yet undefined) or distinguish on +% the following Booleans. +% +% Note: the journal substyle code should only alter the value of one of these Booleans if +% the Boolean is \cmd\undefined. This convention is what makes the document's +% options take precedence over the values set by the journal. +% +% FIXME: make this table an exhaustive listing of all the parameters set by +% the class options. +% +% \begin{tabular}{ll} +% \cmd\@pointsize &(10|11|12), depending on the type size\\ +% \cmd\footinbib@sw &true if footnotes are to be formatted in the bibliography\\ +% \cmd\preprintsty@sw &true for preprint and hyperpreprint\\ +% \cmd\eqsecnum@sw &true means that equations are numbered within sections\\ +% \cmd\secnumarabic@sw &true means the sections are to be numbered arabic\\ +% \cmd\groupauthors@sw &true means authors listed separately for each address\\ +% \cmd\preprint@sw &true means to produce the preprint numbers as part of the title block\\ +% \cmd\showPACS@sw &true means to produce the PACS as part of the title block\\ +% \cmd\showKEYS@sw &true means to produce the keywords as part of the title block\\ +% \cmd\@affils@sw &true means each affiliation is printed, for each author\\ +% \cmd\runinaddress@sw &true means author addresses are printed run-in\\ +% \cmd\draft@sw &true implies that PACS will be printed\\ +% \cmd\tightenlines@sw &true if preprint single spaced\\ +% \cmd\lengthcheck@sw &true if length checking is in effect\\ +% \cmd\byrevtex@sw &true means to announce ``typeset by \revtex''\\ +% \cmd\titlepage@sw &true for title is to be set on a separate page\\ +% \cmd\twocolumn@sw &true if two-column page grid\\ +% \cmd\twocolumn@sw &true if we are to automatically balance the columns of the last page\\ +% \cmd\twoside@sw &true means to format pages for duplex printing\\ +% \cmd\floats@sw &false means floats are migrated to end of document\\ +% \cmd\floatp@sw &true means endfloats are set one to a page\\ +% \cmd\class@amsfonts &if \cmd\@empty, means that \classname{amsfonts} will \emph{not} be loaded\\ +% \cmd\class@amssymb &if \cmd\@empty, means that \classname{amssymb} will \emph{not} be loaded\\ +% \cmd\frontmatter@footnote &if \cmd\undefined, means that the default (\cmd\footnote) will be used\\ +% \cmd\place@bibnumber &if \cmd\undefined, means that the default (inline) will be used\\ +% \end{tabular} +% +% Note: if \cmd\twocolumn@sw\ and \cmd\preprintsty@sw\ are both false, +% then `galley' style is in effect. The \classoption{galley} option invokes +% \classoption{onecolumn}, but does not affect the \cmd\preprintsty@sw. +% +% Note: \cmd\paperwidth\ and \cmd\paperheight\ are not integrated into this scheme, +% and should be selected by the document alone. +% \end{macro} +% +% +% \section{Front Matter} +% +% \subsection{The title command}% +% +% \begin{macro}{\title} +% \begin{macro}{\do@title} +% The author uses \cmd\title\ to specify the title. As in the AMS +% classes, \cmd\title\ has an optional argument specifying a short form +% for use in running heads. +% +% Once the \cmd\title\ is given, you can specify any combination of +% \cmd\thanks, \cmd\email, \cmd\homepage, and \cmd\altaffiliation\ commands, +% applying to the title of the document. +% \begin{macrocode} +\renewcommand\title[2][]{% + \def\@title{#2}% + \def\@shorttitle{#1}% + \let\@AF@join\@title@join +}% +% \end{macrocode} +% \end{macro} +% \end{macro} +% +% \begin{macro}{\@title} +% \begin{macro}{\@shorttitle} +% The arguments to \cmd\title\ are saved in these internal +% macros, which are set up to produce a warning if \cmd\title\ has not +% been used before \cmd\maketitle. +% \begin{macrocode} +\appdef\class@inithook{% + \def\@title{\class@warn{No title}}% + \let\@shorttitle\@empty + \let\@title@aux\@title@aux@cleared +}% +% \end{macrocode} +% \end{macro} +% \end{macro} +% +% \begin{macro}{\@title@join} +% \begin{macro}{\@title@join@} +% \begin{macro}{\@title@aux@cleared} +% +% \begin{macrocode} +\def\@title@join{\expandafter\@title@join@\@title@aux}% +\def\@title@join@#1#2{% + \def\@title@aux{{\@join{\@separator}{#1}{#2}}}% +}% +\def\@title@aux@cleared{{}}% +% \end{macrocode} +% \end{macro} +% \end{macro} +% \end{macro} +% +% +% \subsection{The author, affiliation, and related commands}% +% +% \begin{macro}{\c@affil} +% A counter to store the affiliation numbers used for the +% superscript marks. This command defines \cmd\theaffil. +% \begin{macrocode} +\newcounter{affil} +% \end{macrocode} +% \end{macro} +% +% \subsubsection{Commands for author or collaboration}% +% +% \begin{macro}{\author} +% Each author is given in a \emph{separate} \cmd\author\ command. +% This is similar to the AMS classes, but the AMS also try to support +% the \classname{article} class's \cmd\and\ command for putting multiple +% authors in one \cmd\author\ command. +% \revtex, however, does not support the \cmd\and\ command. +% +% When the \cmd\author\ command is encountered, the previous author, author group, affiliation, and affiliation group are at an end. +% This procedure moves the previous author to the author group (\cmd\move@AU), +% moves the previous affiliation to the affiliation group (\cmd\move@AF), +% and terminates the previous author/affiliation group (\cmd\move@AUAF) as the case may require. +% \begin{macrocode} +\renewcommand\author{% implicit #1 + \@author@def{}% implicit #2 +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\collaboration} +% \begin{macro}{\@collaboration@} +% Specify the collaboration (given after a group of authors). +% +% This procedure does exactly what \cmd\author\ does, and sets +% a flag signifying that the \cmd\collaboration\ command was given. +% +% Note: earlier versions attempted to ignore the collaboration command if it was considered inappropriate. +% We must not do this, however, because it throws the state machine out of kilter. +% \begin{macrocode} +\def\collaboration{% implicit #1 + \groupauthors@sw{% + \class@warn{% + \string\collaboration\space is only allowed with superscriptaddress option. + Use superscriptaddress in place of groupedaddress, unsortedaddress, or runinaddress% + }% + }{}% + \@author@def{\@booleantrue\collaboration@sw}% implicit #2 +}% +% \end{macrocode} +% \end{macro} +% \end{macro} +% +% \begin{macro}{\@author@init} +% \begin{macro}{\@author@cleared} +% \begin{macro}{\@authorclear@sw} +% \begin{macro}{\@author@def} +% \begin{macro}{\@author@join@} +% \begin{macro}{\@author@join} +% \changes{v4.0beta 2}{1999/06/20} +% {Change handling of \cs{@author} to prevent parsing problems. +% (Helps in error recovery when syntax errors exist)} +% +% Here are all the access procedures for the author data structure. +% +% We define a flag value for \cmd\@author (private), +% a procedure for setting \cmd\@author\ to the flag value (\cmd\@author@init), +% a Boolean to test against the flag value (\cmd\@authorclear@sw), +% and a procedure to add new ancillary data to the author. +% \begin{macrocode} +\def\@author@cleared{{}{}}% +\def\@author@init{% + \let\@author\@author@cleared +}% +\def\@authorclear@sw{\@ifx{\@author\@author@cleared}}% +\appdef\class@inithook{% + \@author@init +}% +\def\@author@def#1#2{% + \move@AU\move@AF\move@AUAF + \let\@AF@join\@author@join + #1% + \def\@author{{#2}{}}% +}% +\def\@author@join@#1#2#3{% + \def\@author{{#1}{\@join{\@separator}{#2}{#3}}}% +}% +\def\@author@join{\expandafter\@author@join@\@author}% +\def\@author@gobble#1#2{}% +% \end{macrocode} +% \end{macro} +% \end{macro} +% \end{macro} +% \end{macro} +% \end{macro} +% \end{macro} +% +% \begin{macro}{\move@AU} +% Move the most recent author to the list of current authors. +% If we recently did a \cmd\@author@init, this is a no-op. +% If we recently encountered a \cmd\collaboration, +% we preface the author with the \cmd\CO@opr \ intercollated operator, +% otherwise we apply \cmd\AU@opr. +% \begin{macrocode} +\def\move@AU{% + \@authorclear@sw{}{% + \collaboration@sw{% + \expandafter\appdef + \expandafter\AU@grp + \expandafter{% + \expandafter \CO@opr + \@author + }% + }{% + \expandafter\appdef + \expandafter\AU@grp + \expandafter{% + \expandafter \AU@opr + \@author + }% + }% + }% + \@author@init + \@booleanfalse\collaboration@sw +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\@author@finish} +% We define a procedure to terminate author/affiliation parsing just before the title block is formatted. +% We detect the case where an author group is not followed by an affiliation group. +% \changes{v4.0beta 6}{2001/02/13} +% {Actually append AU@grp to @AAClist.} +% +% We first finish off any author- or affiliation groups in progress. +% Next, we detect the case where there is an author group in progress, but no affiliation group. +% If so, we emit a diagnositic message and act as if a \cmd\noaffiliation\ command had been given. +% In any case, we finish off any author/affiliation pair that may remain. +% Formatting of the title block may now proceed. +% \begin{macrocode} +\def\@author@finish{% + \move@AU\move@AF + \@ifx{\AU@grp\@empty}{}{% + \@ifx{\AF@grp\@empty}{% + \begingroup + \let\href\@secondoftwo + \let\AU@opr\@firstoftwo + \let\CO@opr\@firstoftwo + \class@warn{Assuming \string\noaffiliation\space for \AU@grp}% + \endgroup + \begingroup\@affiliation{\blankaffiliation}%\endgroup in \@affiliation + \move@AF +% +% \appdef \@AAC@list{\AF@opr}% +% \appdef@eval\@AAC@list{\AF@grp}% +% \appdef@e \@AAC@list{\AU@grp}% +% \let\AU@grp\@empty +% \let\AF@grp\@empty + }{}% + }% + \move@AUAF +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\@join} +% The procedure \cmd\@join\ inserts +% a separator between two tokens, or, if the first token is +% nil, elides both that token and the separator. +% \begin{macrocode} +\def\@join#1#2#3{% + \@if@empty{#2}{#3}{#2#1#3}% +}% +\def\@separator{;\space}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\surname} +% \begin{macro}{\firstname} +% No-op to allow better post-processing of the file. +% \begin{macrocode} +\let\surname\@firstofone +\let\firstname\@firstofone +% \end{macrocode} +% \end{macro} +% \end{macro} +% +% \begin{macro}{\AU@grp} +% The control sequence name \cmd\AU@grp\ accretes +% consecutive \cmd\author\ entries (with \cmd\AU@opr\ as the intercollated operator) +% and \cmd\collaboration\ entries (with \cmd\CO@opr\ as the intercollated operator). +% Ultimately, its contents are appended to the author/affiliation list \cmd\@AAC@list. +% It must be initialized to \cmd\@empty. +% +% Note on \cmd\AU@opr\ and \cmd\CO@opr: it is essential to treat these two operators to the greatest extent possible on an equal footing. +% Therefore we invariably assign values to the pair of them within the same procedure. +% +% \begin{macrocode} +\let\AU@grp\@empty +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\and} +% The original \LaTeX\ idea of using one \cmd\author\ command to capture +% all authors and their address just doesn't work with multiple authors +% possibly sharing addresses, so in this class disable \cmd\and. +% \begin{macrocode} +\renewcommand\and{\class@err{\protect\and\space is not supported}} +% \end{macrocode} +% \end{macro} +% +% +% \subsubsection{Commands for affiliation}% +% +% \begin{macro}{\affiliation} +% After each group of authors with the same address, +% give that address in \cmd\affiliation. +% If later in the list you have an author +% with the address of an earlier author, repeat the \cmd\affiliation\ +% command: the system will detect the similarity and (if using the \classoption{superscriptaddress} option) only print the +% affiliation once, reusing the superscript marker. +% +% When the \cmd\affiliation\ command is encountered, the current author and author group (if any) are at an end. +% Also, the current affiliation is at an end. +% +% FIXME: changes to catcode required elsewhere now? +% \begin{macrocode} +\def\cat@comma@active{\catcode`\,\active}% +{\cat@comma@active\gdef,{\active@comma}}% +\def\active@comma{,\penalty-300\relax}% +\newcommand\affiliation{% + \move@AU\move@AF + \begingroup + \cat@comma@active + \@affiliation +}% +\def\@affiliation#1{% + \endgroup + \let\@AF@join\@affil@join + \@affil@def{#1}% +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\noaffiliation} +% User-level command signifying that the current group of authors has \emph{no} affiliation. +% +% This is implemented by acting as if the \cmd\affiliation\ command has been given, and using a +% flag value of \cmd\blankaffiliation, which \cmd\@affil@group\ can recognize as a no-op. +% \begin{macrocode} +\newcommand\noaffiliation{% + \move@AU\move@AF +% \edef\AF@grp{\AF@grp}% + \begingroup\@affiliation{\blankaffiliation}% +% \move@AF +}% +\def\blankaffiliation{}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\@affil@cleared} +% \begin{macro}{\@affil@init} +% \begin{macro}{\@affilclear@sw} +% \begin{macro}{\@affil@def} +% \begin{macro}{\@affil@join@} +% \begin{macro}{\@affil@join} +% +% Here are all the access procedures for the affiliation data structure. +% Note the similarity to those of the \cmd\@author\ data structure. +% +% We define a flag value for \cmd\@affil\ (private), +% a procedure for setting \cmd\@affil\ to the flag value (\cmd\@affil@init), +% a Boolean to test against the flag value (\cmd\@affilclear@sw), +% and a procedure to add new ancillary data to the affiliation. +% \begin{macrocode} +\def\@affil@cleared{{}{}}% +\def\@affil@init{% + \let\@affil\@affil@cleared +}% +\def\@affilclear@sw{\@ifx{\@affil\@affil@cleared}}% +\appdef\class@inithook{% + \@affil@init +}% +\def\@affil@def#1{\def\@affil{{#1}{}}}% +\def\@affil@join@#1#2#3{% + \def\@affil{{#1}{\@join{\@separator}{#2}{#3}}}% +}% +\def\@affil@join{\expandafter\@affil@join@\@affil}% +% \end{macrocode} +% \end{macro} +% \end{macro} +% \end{macro} +% \end{macro} +% \end{macro} +% \end{macro} +% +% \begin{macro}{\move@AF} +% Move the most recent affiliation to the affiliation group. +% If we recently did a \cmd\@affil@init, this is a no-op. +% \begin{macrocode} +\def\move@AF{% + \@affilclear@sw{}{% +% \end{macrocode} +% Set the \cmd\temp@sw\ to false, then execute the affiliation list (\cmd\@AFF@list) +% with the intercollated operator set to \cmd\@affil@match\ +% and with \cmd\@tempa\ holding the first component of \cmd\@affil, the current affiliation, which is to be matched. +% If the current affiliation has been seen before, \cmd\temp@sw\ will be set to true, +% \cmd\@tempc\ will be set to the existing affiliation number, +% and \cmd\@tempd\ will be set to the affiliation's ancillary data. +% The Boolean \cmd\@affils@sw\ being false prevents this test from ever returning a positive result. +% +% This procedure uses \cmd\@tempa\ and sets \cmd\@tempc, \cmd\@tempd, and \cmd\temp@sw. +% \begin{macrocode} + \@booleanfalse\temp@sw + \let\@tempd\@empty + \@affils@sw{% + \expandafter\@firstoftwodef\expandafter\@tempa\@affil + \def\AFF@opr{\@affil@match\@tempa}% + \@AFF@list + }{}\temp@sw +% \end{macrocode} +% True clause: This affiliation has been seen before. +% If ancillary data for the affiliation have been given but are not identical to those seen before, give a warning. +% \begin{macrocode} + {% + \expandafter\@secondoftwodef\expandafter\@tempb\@affil + \@ifx{\@tempb\@empty}{}{% + \@ifx{\@tempb\@tempd}{}{% + \class@warn{% + Ancillary information for \@tempa\space must not be different! + Please put it on the first instance% + }% + }% + }% + \appdef@eval\AF@grp\@tempc + }% +% \end{macrocode} +% False clause: This is a new address: +% increment the counter to get a unique ID for the new affiliation +% and append the new number, address, and footnote to the list \cmd\@AFF@list. +% \begin{macrocode} + {% + \@if@empty{\expandafter\@firstoftwo\@affil}{% + \edef\@tempc{\z@}% + }{% + \stepcounter{affil}% + \edef\@tempc{\theaffil}% + }% + \appdef \@AFF@list{\AFF@opr}% + \appdef@eval\@AFF@list\@tempc + \appdef@e \@AFF@list{\@affil}% + \appdef@eval\AF@grp\@tempc + }% +% \end{macrocode} +% Append the ID of the affiliation to the affiliation group of the current author group. +% \begin{macrocode} + \@affil@init + }% +}% +\def\@firstoftwodef#1#2#3{% + \def#1{#2}% +}% +\def\@secondoftwodef#1#2#3{% + \def#1{#3}% +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\move@AUAF} +% \changes{v4.h}{1998/06/25} +% {Changes to support groups of affiliations} +% Append \cmd\AF@grp\ and \cmd\AU@grp\ to \cmd\@AAC@list\ % +% and merge any new \cmd\AF@grp\ to \cmd\@AFG@list. +% \begin{macrocode} +\def\move@AUAF{% +% \end{macrocode} +% Entire procedure is predicated on something non-trivial to move. +% +% If both author and affiliation are nil, bail out. +% +% No, try that again: +% +% If we have seen any affiliations, then it is time to move the current +% affiliation group and author group to \cmd\@AAC@list. +% If not, we are picking up authors into an author group, and we +% should bail out. +% \begin{macrocode} +%\@ifx{\AU@grp\@empty}{\@ifx{\AF@grp\@empty}{\true@sw}{\false@sw}}{\false@sw}% + \@ifx{\AF@grp\@empty}% + {}{% +% \end{macrocode} +% First, append current affiliation group (expansion of \cmd\AF@grp) +% to the author/affiliation list (\cmd\@AAC@list), +% using \cmd\AF@opr\ as the intercollated operator; +% also append the current authors (\cmd\AU@grp). +% \begin{macrocode} + \appdef \@AAC@list{\AF@opr}% + \appdef@eval\@AAC@list{\AF@grp}% + \appdef@e \@AAC@list{\AU@grp}% +% \end{macrocode} +% Next, append the current affiliation group to \cmd\@AFG@list, +% (if it is not already there) +% using \cmd\AFS@opr\ as the intercollated operator. +% +% Note that \cmd\@AFG@list\ is a list of \emph{unique} affiliation groups. +% \begin{macrocode} + \@ifx{\@empty\AU@grp}{}{% + \@ifx{\@empty\AF@grp}{}{% + \@booleanfalse\temp@sw + \def\AFS@opr{\x@match\AF@grp}% + \@AFG@list + \temp@sw{}{% + \appdef \@AFG@list{\AFS@opr}% + \appdef@eval\@AFG@list{\AF@grp}% + }% + }% + }% +% \end{macrocode} +% Finally, null out the macros holding the author group and affiliation group, respectively. +% \begin{macrocode} + \let\AU@grp\@empty + \let\AF@grp\@empty + }% +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\@AAC@list} +% The \cmd\@AAC@list\ macro accretes +% authors (using \cmd\AU@opr\ as the intercollated operator), +% collaborations (using \cmd\CO@opr\ as the intercollated operator), +% and +% affiliations (using \cmd\AF@opr\ as the intercollated operator). +% \begin{macrocode} +\appdef\class@inithook{% + \let\@AAC@list\@empty +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\@AFG@list} +% The macro \cmd\@AFG@list\ accretes \cmd\AF@grp, +% with \cmd\AFS@opr\ as the intercollated operator. +% +% This token list is employed when and only when \classoption{groupedaddress} is in effect. +% +% It must be empty at the beginning of the document. +% \begin{macrocode} +\appdef\class@inithook{% + \let\@AFG@list\@empty +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\@AFF@list} +% The macro \cmd\@AFF@list\ accretes the list of affiliations +% with \cmd\AFF@opr\ as the intercollated operator. +% It must be empty at the beginning of the document. +% +% Each \cmd\affiliation\ command in the document contributes +% to this list, as long as the argument of that instance of the command +% is ``new''. +% +% Each entry in this list consists of the \cmd\AFF@opr\ intercollated operator +% followed by three brace-delimited tokens, representing, in order: +% \begin{enumerate} +% \item +% the affiliation's unique ID, a number increasing monotonically for +% each new entry, +% \item +% the optional argument of the \cmd\affiliation\ command, +% the footnote to the affiliation, and +% \item +% the text of the affiliation. +% \end{enumerate} +% +% At various junctures in the code, the control sequence name +% \cmd\AFF@opr\ assumes the meaning of +% \cmd\@affil@footnote\ (FIXME: never used), +% \cmd\@affil@group, \cmd\affil@script, +% or is defined to expand to \cmd\@affil@match\cmd\@tempa. +% and the list \cmd\@AFF@list\ is expanded. +% \begin{macrocode} +\appdef\class@inithook{% + \let\@AFF@list\@empty +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\AF@grp} +% The macro \cmd\AF@grp\ accretes affiliation IDs into an affiliation group. +% It must be empty at the beginning of the document. +% \begin{macrocode} +\let\AF@grp\@empty +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\@affil@match} +% The definition for the \cmd\AFF@opr, the intercollated operator for \cmd\@AFF@list, +% when adding a new affiliation: seeks a match with any earlier affiliation. +% When the affiliation list is executed, this procedure tests each entry against +% the `new' affiliation (which has been stored in \cmd\@tempa). +% +% Uses \cmd\temp@sw, \cmd\@tempc, and \cmd\@tempd\ to communicate back: +% if it returns \cmd\temp@sw\ true, a match was found, +% then \cmd\@tempc\ will be the footnote number of the matching entry +% and \cmd\@tempd\ will be the auxiliary information of the matching entry. +% \begin{macrocode} +\def\@affil@match#1#2#3#4{% + \temp@sw{}{% + \def\@tempifx{#3}% + \@ifx{\@tempifx#1}{% + \@booleantrue\temp@sw + \def\@tempc{#2}% + \def\@tempd{#4}% + }{% + }% + }% +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\@tempaffil@script} +% Give a superscript marker to the address, +% and junk the text argument (|#2|). +% \begin{macrocode} +%\def\@tempaffil@script#1#2{% +% ,\,#1% +%}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\@tempaffil@group} +% +% Execute if we are grouping authors above `main addresses' (\classoption{groupedaddress}). +% Put temp address in a footnote, with explanatory text. +% Note: this procedure uses \cmd\@tempb\ and \cmd\@tempc\ +% to communicate to \cmd\AFF@opr. +% \begin{macrocode} +%\def\@tempaffil@group#1#2{% +% \def\@tempc{#1}% +% \def\@tempb{#2}% +% \begingroup +% \let\AFF@opr \@affil@footnote +% \@AFF@list +% \endgroup +%}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\@affil@footnote} +% The alias of \cmd\AFF@opr\ when \cmd\@AFF@list\ +% is executed by \cmd\@tempaffil@group. +% If we find a match of the affiliation number to \cmd\@tempc, +% then footnotes are formatted with the relevant content and \cmd\@tempb. +% \begin{macrocode} +%\def\@affil@footnote#1#2#3{% +% \def\@tempifx{#1}% +% \@ifx{\@tempifx\@tempc}{% +% \comma@space +% \expandafter\frontmatter@footnote\expandafter{\@tempb#2}% +% \frontmatter@footnote{#3}% +% }{% +% }% +%}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\@AF@join} +% The \cmd\csname\ \cmd\@AF@join\ is a procedure used within +% \cmd\email, \cmd\homepage, \cmd\thanks, and \cmd\altaffiliation. +% When such a command appears after an \cmd\author, \cmd\collaboration, or \cmd\affiliation +% command, the \cmd\@AF@join\ procedure appends the argument given to the appropriate +% macro. +% Its default is to give an error message, since these commands are legal only within +% the particular context mentioned. +% \begin{macrocode} +\appdef\class@inithook{% + \let\@AF@join\@AF@join@error +}% +\def\@AF@join@error#1{% + \class@warn{% + \string\email, \string\homepage, \string\thanks, or \string\altaffiliation\space + appears in wrong context. + }% +}% +% \end{macrocode} +% \end{macro} +% +% +% \subsubsection{Commands for auxiliary information}% +% +% \begin{macro}{\email} +% Just tacks the email address on to the current author or affiliation. +% \begin{macrocode} +\def\sanitize@url{% + \@makeother\%% + \@makeother\~% + \@makeother\_% +}% +\newcommand*\email[1][Electronic address: ]{\begingroup\sanitize@url\@email{#1}}% +\def\@email#1#2{% + \endgroup + \@AF@join{#1\href{mailto:#2}{#2}}% +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\homepage} +% Just tacks the URL on to the current author or affiliation. +% Note: group opened in \cmd\homepage\ is closed in \cmd\@homepage. +% \begin{macrocode} +\newcommand*\homepage[1][URL:~]{\begingroup\sanitize@url\@homepage{#1}}% +\def\@homepage#1#2{% + \endgroup + \@AF@join{#1\url{#2}}% +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\url} +% \begin{macro}{\href} +% \begin{macrocode} +\appdef\class@documenthook{% + \providecommand\url{\texttt}% + \providecommand\href[2]{\url{#2}}% +}% +% \end{macrocode} +% \end{macro} +% \end{macro} +% +% \begin{macro}{\thanks} +% +% The operative version of \cmd\thanks\ appends an item to \cmd\@author, or \cmd\@affil's auxiliary data. +% \begin{macrocode} +\def\thanks{% implicit #1 + \@AF@join +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\altaffiliation} +% \changes{v4.d}{1998/01/31} +% {change name from \cs{tempaffiliation}} +% \changes{v4.l}{1998/09/01} +% {add optional argument handling back} +% Implemented more or less like \cmd\thanks\ but shares the affiliation +% counter. +% Optional argument may be used to give explanatory text eg +% `currently staying at' +% This will be placed before the address, if used in a footnote. +% \begin{macrocode} +\newcommand*\altaffiliation[2][]{% + \@AF@join{#1#2}% +}% +% \end{macrocode} +% \end{macro} +% +% +% \subsubsection{Procedures for author, collaboration, and affiliation}% +% +% Macros that list off, say, authors, and which require punctuation, like ``A, B, and C'', will use +% \cmd\set@listcomma@list, \cmd\set@listcomma@count, \cmd\@listand, and \cmd\@listcomma. +% These macros use the \cmd\@tempcnta\ register to keep track of how many items remain to be listed off. +% +% At present, \cmd\@author@present\ will decrement that counter. +% +% \begin{macro}{\set@listcomma@list} +% \begin{macro}{\set@listcomma@count} +% \changes{v4.l}{1998/09/01} +% {macro added} +% Pop the author count for this collaboration. +% +% Note: by side effect, it assigns \cmd\@listcomma: +% for a list of length two, suppress comma addition. +% \begin{macrocode} +\def\set@listcomma@list#1{% + \expandafter\@reset@ac\expandafter#1#1{0}\@reset@ac{% + \let\@listcomma\relax + }{% + \let\@listcomma\@listcomma@comma + }% +}% +\def\set@listcomma@count#1{% + \@ifnum{#1=\tw@}{% + \let\@listcomma\relax + }{% + \let\@listcomma\@listcomma@comma + }% +}% +% \end{macrocode} +% Does the actual pop operation, then generates a Boolean which selects one of the two assignments for \cmd\@listcomma. +% +% Note: this procedure sets \cmd\@tempcnta\ to reflect the number of items to list off. +% \begin{macrocode} +\def\@reset@ac#1#2#3\@reset@ac{% + \def#1{#3}% + \@tempcnta#2\relax + \@ifnum{#2=\tw@}% +}% +% \end{macrocode} +% \end{macro} +% \end{macro} +% +% \begin{macro}{\listand} +% Might need extending with penalties etc. +% +% Note: this procedure expects \cmd\@tempcnta\ to reflect the current number of items remaining to list off. +% \begin{macrocode} +\def\@listand{\@ifnum{\@tempcnta=\tw@}{\andname\space}{}}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\@listcomma} +% This definition, with \cmd\@ne\ puts a comma before and. +% +% David Carlisle says: In UK English (at least) would have \cmd\tw@\ here, +% which would then implement the convention of indicating conjunction with comma or ``and'', but not both. +% +% Note: this procedure expects \cmd\@tempcnta\ to reflect the current number of items remaining to list off. +% \begin{macrocode} +\def\@listcomma@comma{\@ifnum{\@tempcnta>\@ne}{,}{}}% +\def\@listcomma@comma@UK{\@ifnum{\@tempcnta>\tw@}{,}{}}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\@collaboration@err} +% This command will be executed if are in a \classoption{groupedaddress} situation and +% the document contained a \cmd\collaboration\ command. +% +% There already has been issued a warning message explaining that \cmd\collaboration\ is wrong in this case, +% so here we simply gobble. +% \begin{macrocode} +\def\@collaboration@err#1#2{% +%\class@warn{An illegal \string\collaboration\ command has been entered!}% +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\doauthor} +% Main control over how authors are typeset. +% +% |#1| is loaded by \cmd\author +% +% |#2| is loaded by \cmd\email, \cmd\homepage, \cmd\altaffiliation, or \cmd\thanks +% +% |#3| is the superscript affiliation, if at all used. +% +% First, the author name is formatted, followed by a comma, +% then come any marks relating to affiliation (if present), +% then come the homepage URL and email address, if any, with +% appropriate punctuation. +% \begin{macrocode} +\def\doauthor#1#2#3{% + \ignorespaces#1\unskip\@listcomma + \def\@tempifx{#3}% + \@ifx{\@tempifx\@empty}{% + \def\@tempifx{#2}% + \@ifx{\@tempifx\@empty}{}{\frontmatter@footnote{#2}}% + }{% + #3% + \def\@tempifx{#2}% + \@ifx{\@tempifx\@empty}{}{\comma@space\frontmatter@footnote{#2}}% + }% + \space \@listand +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\AF@opr} +% DPC: The number(s) associated to the affiliation address(es). +% \end{macro} +% +% \begin{macro}{\@thanks} +% \changes{v4.d}{1998/01/31} +% {\cs{comma@space} added} +% Society-specific options may change this. \cmd\comma@space\ is a separator +% between adjacent footnotes. +% \begin{macrocode} +\def\@thanks{\comma@space\footnote} +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\x@match} +% Procedure \cmd\x@match\ is used as an alias \cmd\AFS@opr, +% probing the list of affiliation groups +% and setting a Boolean if a match is found to \#1 (usually the current affiliation group). +% +% Procedure \cmd\y@match\ is used as an alias to \cmd\@TBN@opr, +% probing the list of unique title block footnotes for a match to the footnote text: +% when the first match is found it sets a Boolean, +% defining \cmd\@tempb\ to the number of that footnote. +% \begin{macrocode} +\def\x@match#1#2{% + \temp@sw{}{% + \def\@tempifx{#2}% + \@ifx{\@tempifx#1}{% + \@booleantrue\temp@sw + }{% + }% + }% +}% +\def\y@match#1#2#3{% + \temp@sw{}{% + \def\@tempifx{#3}% + \@ifx{\@tempifx#1}{% + \@booleantrue\temp@sw + \def\@tempb{#2}% + }{% + }% + }% +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\frontmatter@footnote} +% Used when typesetting a title block footnote. +% (Item 290: merge duplicates.) +% +% Maintains a list of frontmatter footnotes along with associated device. +% Uniquifies each footnote by traversing the list and looking for a match to the footnote text. +% If found, define \cmd\@tempb\ to the corresponding footnote device, and typeset it with \cmd\@footnotemark. +% If never found, manually increment footnote counter, +% determine the corresponding footnote device, and +% add footnote text and device to the list of frontmatter footnotes. +% +% The list \cmd\@FMN@list\ is processed later in the title block production. +% +% Note that this method of making footnotes runs afoul of \classname{hyperref}'s patches of the \LaTeX\ kernel. +% Therefore, we avoid trouble by refraining from using \cmd\@footnotemark\ and by creating our own hypertext links. +% \begin{macrocode} +\def\frontmatter@footnote#1{% + \begingroup + \@booleanfalse\temp@sw + \def\@tempa{#1}% + \let\@tempb\@empty + \def\@TBN@opr{\y@match\@tempa}% + \@FMN@list + \temp@sw{% + \expandafter\frontmatter@footnotemark + \expandafter{\@tempb}% + }{% + \stepcounter\@mpfn + \protected@xdef\@tempb{\the\csname c@\@mpfn\endcsname}% + \expandafter\frontmatter@footnotemark + \expandafter{\@tempb}% + \expandafter\g@addto@macro + \expandafter\@FMN@list + \expandafter{% + \expandafter \@TBN@opr + \expandafter {\@tempb}{#1}% + }% + }% + \endgroup +}% +\appdef\class@inithook{% + \global\let\@FMN@list\@empty +}% +\def\frontmatter@footnotemark#1{% + \leavevmode + \ifhmode\edef\@x@sf{\the\spacefactor}\nobreak\fi + \begingroup + \hyper@linkstart {link}{frontmatter@#1}% + \csname c@\@mpfn\endcsname#1\relax + \def\@thefnmark{\thempfn}% + \@makefnmark + \hyper@linkend + \endgroup + \ifhmode\spacefactor\@x@sf\fi + \relax +}% +% \end{macrocode} +% \end{macro} +% +% +% +% \subsection{The keywords command}% +% \begin{macro}{\keywords} +% \begin{macro}{\@keywords} +% Usual game, save text in a macro for processing by \cmd\maketitle. +% \begin{macrocode} +\def\keywords#1{% + \aftermaketitle@chk{\keywords}% + \gdef\@keywords{#1}% +}% +\def\@keywords{}% +% \end{macrocode} +% \end{macro} +% \end{macro} +% +% \subsection{The \cs{date} command and related commands}% +% \begin{macro}{\date} +% \begin{macro}{\received} +% \begin{macro}{\revised} +% \begin{macro}{\accepted} +% \begin{macro}{\published} +% \begin{macro}{\@date} +% \begin{macro}{\@received} +% \begin{macro}{\@revised} +% \begin{macro}{\@accepted} +% \begin{macro}{\@published} +% Use the \cmd\date\ command to specify the document date, +% the \cmd\received\ command to specify the date received, +% \cmd\revised\ for date revised, +% \cmd\accepted\ for date accepted, and +% \cmd\published\ for date published. +% Normally only \cmd\date\ to be used by author, +% the remainder used only by editors. +% +% DPC: As for \cmd\keywords, but this time don't flag a warning if there is no +% revision date specified. +% +% In each case the user-level command defines the value of a +% macro which buffers the data entered by the user. +% For instance, \cmd\received\ causes \cmd\@received\ to +% acquire a value. +% The optional argument allows the user to override the +% text that will be typeset along with the date; +% the default value of that text is itself +% a localized macro. +% \begin{macrocode} +\renewcommand*\date[2][\Dated@name]{\def\@date{#1#2}}% +\def\@date{}% +\newcommand*\received[2][\Received@name]{\def\@received{#1#2}}% +\def\@received{}% +\newcommand*\revised[2][\Revised@name]{\def\@revised{#1#2}}% +\def\@revised{}% +\newcommand*\accepted[2][\Accepted@name]{\def\@accepted{#1#2}}% +\def\@accepted{}% +\newcommand*\published[2][\Published@name]{\def\@published{#1#2}}% +\def\@published{}% +% \end{macrocode} +% +% \end{macro} +% \end{macro} +% \end{macro} +% \end{macro} +% \end{macro} +% \end{macro} +% \end{macro} +% \end{macro} +% \end{macro} +% \end{macro} +% +% \subsection{The pacs command}% +% PACS, the Physics and Astronomy Classification Scheme. +% \begin{macro}{\pacs} +% \begin{macro}{\@pacs} +% As for \cmd\keywords\ +% \begin{macrocode} +\def\pacs#1{% + \aftermaketitle@chk{\pacs}% + \gdef\@pacs{#1}% +}% +\def\@pacs{}%{\class@warn{No PACS numbers}}% +% \end{macrocode} +% \end{macro} +% \end{macro} +% +% \subsection{The \cs{preprint} command} +% +% \changes{v4.g}{1998/06/10} +% {multiple preprint commands} +% \begin{macrocode} +\def\preprint#1{\gappdef\@preprint{\preprint{#1}}}% +\let\@preprint\@empty +% \end{macrocode} +% +% \subsection{draft} +% DPC: Name clash with draft option. \LaTeX\ doesn't care, but it confuses me. +% +% This command is a 3.1 relic, now removed. +% \begin{macrocode} +%\newcommand\draft{\@booleantrue\draft@sw}% +% \end{macrocode} +% +% \subsection{The \env{abstract} environment}% +% +% \changes{v4.0beta 2}{1999/06/20} +% {Gathered all code for the abstract environment together +% and abstracted out the formatting details for journals to override.} +% +% \begin{macro}{\absbox} +% +% \begin{macrocode} +\newbox\absbox +% \end{macrocode} +% \end{macro} +% +% \begin{environment}{abstract} +% Abstract, as in AMS classes, must be specified \emph{before} +% \cmd\maketitle. It just saves everything up in \cmd\absbox. +% \changes{v4.d}{1998/01/31} +% {If empty \cs{abstractname} omit spacing} +% \begin{macrocode} +% \changes{v4.e}{1998/01/31} +% {hookify} +% Note that the specifics of how the abstract is to be +% formatted are set by \cmd\frontmatter@abstractwidth, +% \cmd\frontmatter@abstractheading, and \cmd\frontmatter@abstractfont. +% +% Here we wish to set the abstract into type but save it away in a box, much like +% the \cmd\minipage\ command does. +% +% Note that the \cmd\endabstract\ portion of the environment +% begins with code that mostly duplicates \cmd\endminipage, but without the \cmd\@iiiparbox. +% At the end, we simply transfer the contents of the \cmd\minipage\ +% into a box of our own. +% +% However, we also have to end the gratuitous +% paragraph begun by \cmd\minipage. +% We ensure that no trace of that paragraph is left, +% by absorbing the \cmd\parindent\ box. +% Doing so destroys the paragraph entirely, except +% for the \cmd\parskip\ glue, and that morsel is pruned, because we are at the top +% of a page. +% +% This mechanism is vulnerable, however, because anything placed into the horizontal list +% after the \cmd\parindent\ box will confound it. +% And that is exactly what happens under Mik\TeX: +% There seems to be an extension to the \TeX\ standard operating here which inserts a \cmd\special \ at the beginning of every paragraph. +% (The mechnism is \emph{not} based on \cmd\everypar.) +% +% To accomodate this state of affairs, +% we have a new mechanism that offers an even more complete rewrite of \env{minipage}. +% +% The user can elect to use this new mechanism via the \classoption{newabstract}; +% likewise the old mechanism via the \classoption{oldabstract}. +% +% \begin{macrocode} +\newenvironment{abstract}{% + \aftermaketitle@chk{\begin{abstract}}% + \global\setbox\absbox\vbox\bgroup + \color@begingroup + \columnwidth\textwidth + \hsize\columnwidth + \@parboxrestore + \def\@mpfn{mpfootnote}\def\thempfn{\thempfootnote}\c@mpfootnote\z@ + \let\@footnotetext\frontmatter@footnotetext + \minipagefootnote@init + \let\@listdepth\@mplistdepth \@mplistdepth\z@ + \@minipagerestore + \@setminipage + \frontmatter@abstractheading + \frontmatter@abstractfont + \let\footnote\mini@note +}{% + \par + \unskip + \minipagefootnote@here + \@minipagefalse %% added 24 May 89 + \color@endgroup + \egroup +}% +% \end{macrocode} +% \end{environment} +% +% \begin{macro}{\frontmatter@footnotetext} +% We reimplement \cmd\@mpfootnotetext\ under a new name (so as not to be overridden by the \classname{hyperref} package) +% and extend it to accomodate hyperrefs. +% +% Note that this procedure is very like \classname{ltxutil}'s \cmd\mp@footnotetext, except that it +% uses \texttt{p@mpfootnote} instead of \texttt{p@mpfootnote}, and +% \cmd\frontmatter@makefntext\ instead of \cmd\@makefntext. +% +% \begin{macrocode} +\long\def\frontmatter@footnotetext#1{% + \minipagefootnote@pick + \reset@font\footnotesize + \interlinepenalty \interfootnotelinepenalty + \splittopskip \footnotesep + \splitmaxdepth \dp \strutbox + \set@footnotewidth + \@parboxrestore + \protected@edef\@currentlabel{\csname p@mpfootnote\endcsname\@thefnmark}% + \color@begingroup + \frontmatter@makefntext{% + \rule\z@\footnotesep\ignorespaces#1\@finalstrut\strutbox + }% + \color@endgroup + \minipagefootnote@drop +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\frontmatter@abstractheading} +% The default abstract head; journals will override this procedure. +% \begin{macrocode} +\def\frontmatter@abstractheading{% + \begingroup + \centering\large + \abstractname + \par + \endgroup +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\frontmatter@abstractfont} +% The default type specification for the body of the abstract. +% Journals will override this setting. +% \begin{macrocode} +\def\frontmatter@abstractfont{}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\frontmatter@abstractwidth} +% The default setting is the full text width; +% journals can override this setting. +% \begin{macrocode} +\def\frontmatter@abstractwidth{\textwidth} +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\frontmatter@abstract@produce} +% This procedure determines how the abstract is incorporated +% into the title block. We split this out in anticipation +% of the need to lift the limitation that the abstract not break over pages. +% +% If we are in preprint style, we provide an easy pagebreak point immediately above the abstract. +% This means that the abstract will either fit entirely on the same page as the title block, +% or it will start a page of its own (which may itself break onto a subsequent page if necessary). +% \begin{macrocode} +\def\frontmatter@abstract@produce{% + \par +% \tracingall + \preprintsty@sw{% + \do@output@MVL{% + \vskip\frontmatter@preabstractspace + \vskip200\p@\@plus1fil + \penalty-200\relax + \vskip-200\p@\@plus-1fil + }% + }{% + \addvspace{\frontmatter@preabstractspace}% + }% +% \leavevmode + \begingroup + \dimen@\baselineskip + \setbox\z@\vtop{\unvcopy\absbox}% + \advance\dimen@-\ht\z@\advance\dimen@-\prevdepth + \@ifdim{\dimen@>\z@}{\vskip\dimen@}{}% + \endgroup + \unvbox\absbox + \@ifx{\@empty\mini@notes}{}{\mini@notes\par}% +% \par + \addvspace\frontmatter@postabstractspace +% \showlists +% \tracingplain +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\frontmatter@preabstractspace} +% \begin{macro}{\frontmatter@postabstractspace} +% Space above and space below abstract in title block +% \begin{macrocode} +\def\frontmatter@preabstractspace{.5\baselineskip} +\def\frontmatter@postabstractspace{.5\baselineskip} +% \end{macrocode} +% \end{macro} +% \end{macro} +% +% +% +% \subsection{Formatting the title block}% +% +% \begin{environment}{titlepage} +% This is \LaTeXe's \classname{article} class version, +% with modifications. +% \changes{v4.0beta 3}{1999/11/13} +% {grid changes with ltxgrid} +% \begin{macrocode} +\newenvironment{titlepage}{% + \twocolumn@sw{\onecolumngrid}{\newpage}% + \thispagestyle{titlepage}% + \setcounter{page}\@ne +}{% + \twocolumn@sw{\twocolumngrid}{\newpage}% + \twoside@sw{}{% + \setcounter{page}\@ne + }% +}% +% \end{macrocode} +% \end{environment} +% +% \begin{macro}{\maketitle} +% \changes{v4.d}{1998/01/31} +% {hookify} +% Put it all together to format the title block. +% \begin{macrocode} +\def\maketitle{% + \@author@finish + \title@column\titleblock@produce + \suppressfloats[t]% +% \end{macrocode} +% Now save some memory. +% \begin{macrocode} + \let\and\relax + \let\affiliation\@gobble +% \let\address\affiliation + \let\author\@gobble + \let\@AAC@list\@empty + \let\@AFF@list\@empty + \let\@AFG@list\@empty +% \let\keywords\@gobble +% \let\@keywords\@empty + \let\@AF@join\@AF@join@error + \let\email\@gobble + \let\@address\@empty + \let\maketitle\relax + \let\thanks\@gobble + \titlepage@sw{% + \vfil + \clearpage + }{}% +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\titleblock@produce} +% This procedure produces the title block. +% +% It effectively executes inside a group +% and always returns us to vertical mode. +% +% Note: we assume that it is OK to set the footnote counter to zero at this point. +% Is this a safe assumption? +% \begin{macrocode} +\def\titleblock@produce{% + \begingroup +% \end{macrocode} +% Special footnote version [of footnotes] for \cmd\thanks, +% copied from \classname{article} class \cmd\maketitle. +% \changes{v4.0beta 2}{1999/06/20} +% {Set \cs{footnote} to \LaTeX\ standard version for this scope.} +% \begin{macrocode} + \let\footnote\footnote@latex + \let\@makefnmark\@makefnmark@latex + \let\@footnotemark\@footnotemark@latex + \let\thefootnote\frontmatter@thefootnote +% \let\@makefntext\frontmatter@makefntext + \global\c@footnote\z@ + \let\@makefnmark\frontmatter@makefnmark + \frontmatter@setup + \thispagestyle{titlepage}\label{FirstPage}% +% \end{macrocode} +% Produce the title: +% \begin{macrocode} + \frontmatter@title@produce +% \end{macrocode} +% Produce the author list: +% \begin{macrocode} + \groupauthors@sw{% + \frontmatter@author@produce@group + }{% + \frontmatter@author@produce@script + }% +% \end{macrocode} +% Produce the dates: +% \begin{macrocode} + \frontmatter@RRAPformat{% + \expandafter\produce@RRAP\expandafter{\@date}% + \expandafter\produce@RRAP\expandafter{\@received}% + \expandafter\produce@RRAP\expandafter{\@revised}% + \expandafter\produce@RRAP\expandafter{\@accepted}% + \expandafter\produce@RRAP\expandafter{\@published}% + }% +% \end{macrocode} +% Produce the abstract, PACS, and keywords, and end any paragraph. +% \begin{macrocode} + \frontmatter@abstract@produce + \@pacs@produce\@pacs + \@keywords@produce\@keywords + \par + \frontmatter@finalspace + \endgroup +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\frontmatter@title@produce} +% The default formatting procedure for the article title. +% This procedure should take care of the vertical space below the title. +% +% Journals may override this procedure, but it is more likely that they will +% use the hooks \cmd\frontmatter@title@above, \cmd\frontmatter@title@format, and \cmd\frontmatter@title@below. +% +% At this point, we patch into \cmd\let@mark\ so that the common markup idiom \cmd\\ is taken care of. +% \begin{macrocode} +\def\frontmatter@title@produce{% + \begingroup + \frontmatter@title@above + \frontmatter@title@format + \@title + \@ifx{\@title@aux\@title@aux@cleared}{}{% + \expandafter\frontmatter@footnote\expandafter{\@title@aux}% + }% + \par +% \@ifx{\@shorttitle\@empty}{% +% \markboth{\@title}{\@title}% +% }{% +% \markboth{\@shorttitle}{\@shorttitle}% +%% \class@warn{Unused short title ignored}% +% }% + \frontmatter@title@below + \endgroup +}% +\appdef\let@mark{\let\\\relax}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\frontmatter@title@above} +% \begin{macro}{\frontmatter@title@format} +% \begin{macro}{\frontmatter@title@below} +% The default values for formatting specifications for the article title. +% +% The procedure \cmd\frontmatter@title@above\ should take care of the vertical space above the title; +% \cmd\frontmatter@title@below\ below. +% The procedure \cmd\frontmatter@title@format\ should invoke any font switches, etc. +% that may apply to the title. +% +% Journals will override this procedure. +% \begin{macrocode} +\def\frontmatter@title@above{}% +\def\frontmatter@title@format{}% +\def\frontmatter@title@below{\addvspace{\baselineskip}}% +% \end{macrocode} +% \end{macro} +% \end{macro} +% \end{macro} +% +% \subsubsection{Authors and affiliations in superscriptaddress style} +% +% \begin{macro}{\frontmatter@author@produce@script} +% \cmd\frontmatter@author@produce@script\ is an alias for \cmd\frontmatter@author@produce, +% the procedure that formats the author/affiliation list. +% +% In this case, the authors affiliations are being superscripted +% (class option \classoption{superscriptaddress}). +% +% This procedure must ensure that any paragraph that it starts gets ended finally. +% \begin{macrocode} +\def\frontmatter@author@produce@script{% + \begingroup + \let\@author@present\@author@present@script +% \end{macrocode} +% \changes{v4.l}{1998/09/01} +% {Changes to add collaboration processing, which now +% means doing comma/and processing on each sublist} +% When the author/affiliation list \cmd\@AAC@list\ is executed, +% \cmd\@tempcnt\ counts each author, +% and +% \cmd\@tempa\ stores a list of author indices for the current collaboration. +% +% Note: this procedure uses \cmd\AU@temp\ to hold a list of author counts for each collaboration. +% Note: also uses \cmd\@tempcnta\ to communicate between procedures. +% This usage ot \cmd\@tempcnta\ is OK, because the expansion of \cmd\@AAC@list\ will trigger only the execution of +% our own procedures. +% \begin{macrocode} + \let\AU@temp\@empty + \@tempcnta\z@ +% \end{macrocode} +% +% We wish to know how many authors are in each collaboration, for the purposes of \cmd\listcomma\ and \cmd\listand. +% +% Here we assign values for the intercollated operators appearing within \cmd\@AAC@list, +% then execute that macro, registering the authors in each collaboration. +% Afterwards, clean up by emulating an empty collaboration. +% +% The result, stored in \cmd\AU@temp, is a list of brace-delmited tokens, each a number +% representing the number of authors in that collaboration. +% \begin{macrocode} + \let\AF@opr \@gobble + \def\AU@opr{\@author@count\@tempcnta}% + \def\CO@opr{\@collaboration@count\AU@temp\@tempcnta}% + \@AAC@list + \expandafter\CO@opr\@author@cleared +% \end{macrocode} +% Next, present the authors and their affiliations. +% +% We assign values for the intercollated operators appearing within \cmd\@AAC@list. +% Next, assign \cmd\@listcomma\ based on the number of authors in the first collaboration. +% Then expand \cmd\@AAC@list. +% +% Upon completion, there will be two space tokens following the last author because of \cmd\listand\ processing; +% remove them, then end the last paragraph. +% +% FIXME: this juncture would be a good time to effectively eliminate the unused affiliations in \cmd\@AFF@list. +% \begin{macrocode} + \begingroup + \frontmatter@authorformat + \let\AF@opr \@affilID@def + \let\AU@opr \@author@present + \def\CO@opr{\@collaboration@present\AU@temp}% + \set@listcomma@list\AU@temp + \@AAC@list + \unskip\unskip + \par + \endgroup +% \end{macrocode} +% We now list out the affiliations in the order they appeared. +% \begin{macrocode} + \begingroup + \frontmatter@above@affiliation@script +% \let\@AFU@list\@empty +% \let\AFS@opr \affils@present@script +% \@AFG@list +% \let\AF@opr \affil@present@script +% \@AFU@list + \let\AFF@opr \@affil@script +% \say\@AFF@list + \@AFF@list + \frontmatter@footnote@produce + \par + \endgroup + \endgroup +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\@author@count} +% This version of \cmd\AU@opr\ counts the number of authors it processes. +% \begin{macrocode} +\def\@author@count#1{\advance#1\@ne\@author@gobble}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\@collaboration@present} +% Format a collaboration. +% Note that we immediately end the paragraph thus begun, because we only support +% \classoption{superscriptaddress}-style processing. +% +% Note also that the execution of the production procedures takes place inside of a group; +% for reasons of protective programming. +% +% Bug note: originally, the number of authors in each collaboration was stored in \cmd\@tempa, +% but this exposed us to procedures in our production that modified the meaning of \cmd\@tempa. +% +% The last action done in this procedure is to assign \cmd\@listcomma\ based on the number of authors in the next collaboration. +% \begin{macrocode} +\def\@collaboration@present#1#2#3{% + \par + \begingroup + \frontmatter@collaboration@above + \@author@present{(\ignorespaces#2\unskip)}{#3}% + \par + \endgroup + \set@listcomma@list#1% +}% +\def\frontmatter@collaboration@above{}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\@collaboration@count} +% Simply register each author in this collaboration. +% Note: \#1 is a \cmd\csname\ to hold the value, \#2 is the \cmd\count\ register holding the count of interest. +% \begin{macrocode} +\def\@collaboration@count#1#2{% + \appdef@eval#1{\the#2}#2\z@ + \@author@gobble +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\@affilID@def} +% +% The \cmd\@affilID@def\ procedure, an alias of \cmd\AF@opr\ used during \cmd\frontmatter@author@produce@script\ processing, +% memorizes the affiliation in \cmd\@affilID@temp. +% \begin{macrocode} +\def\@affilID@def{\def\@affilID@temp}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\@affilID@temp} +% +% The macro \cmd\@affilID@temp\ is used to communicate between +% \cmd\doauthor\ and instances of \cmd\AF@opr\ within the author list. +% \begin{macrocode} +\let\@affilID@temp\@empty +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\affils@present@script} +% An alias of \cmd\AFS@opr, this procedure is applied to \cmd\@AFG@list\ and builds +% \cmd\@AFU@list, a list of unique affiliations found within affiliation groups. +% +% \begin{macrocode} +%\def\affils@present@script#1{% +% \get@affil#1{}% +%}% +%\def\get@affil#1{% +% \def\@tempa{#1}% +% \@ifx{\@empty\@tempa}{}{% +% \@booleanfalse\temp@sw +% \def\AF@opr{\x@match\@tempa}% +% \@AFU@list +% \temp@sw{}{% +% \appdef\@AFU@list{\AF@opr{#1}}% +% }% +% \get@affil +% }% +%}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\affil@present@script} +% An alias of \cmd\AF@opr, this procedure is applied to \cmd\@AFU@list\ when formatting the affiliations +% under \classoption{superscriptaddress}. +% It in turn applies \cmd\affil@script\ to \cmd\@AFF@list, thereby formatting the affiliation +% associated with this affiliation index. +% +% Note: it traverses the \cmd\@AFF@list\ for each entry in the \cmd\@AFU@list, thereby making this +% portion of code execute in $N^{2}$ time. +% \begin{macrocode} +%\def\affil@present@script#1{% +% \def\@tempa{#1}% +% \begingroup +% \frontmatter@affiliationfont +% \let\AFF@opr \affil@script +% \@AFF@list +% \endgroup +%}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\affil@script} +% Alias of \cmd\AFF@opr, this procedure is applied to \cmd\@AFF@list\ when formatting the affiliations +% under \classoption{superscriptaddress}. +% It rejects all entries not matching the affiliation index held in \cmd\@tempa. +% \begin{macrocode} +\def\affil@script#1#2#3{% + \def\@tempifx{#1}\@ifx{\@tempifx\@tempa}{% + \@if@empty{#2}{}{% + \par + \begingroup + \textsuperscript{\normalfont#1}% + #2% + \@if@empty{#3}{}{\frontmatter@footnote{#3}}% + \par + \endgroup + }% + }{}% +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\@affil@script} +% Alias of \cmd\AFF@opr: +% The affiliations are being formatted in the case +% where affiliations are being superscripted: +% make a list out of the affiliations with the numbers printed. +% Society-specific code can change the formatting +% by overriding the definition of \cmd\frontmatter@affiliationfont. +% +% FIXME: dead code. +% \begin{macrocode} +\def\@affil@script#1#2#3{% + \@ifnum{#1=\z@}{}{% + \par + \begingroup + \frontmatter@affiliationfont +% \textsuperscript{\normalfont#1}% + \textsuperscript{#1}% + #2% + \@if@empty{#3}{}{\frontmatter@footnote{#3}}% + \par + \endgroup + }% +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\@author@present@script} +% \changes{v4.h}{1998/06/25} +% {Changes to support groups of affiliations} +% This version of \cmd\AU@opr\ applies to the \classoption{superscriptaddress} class option. +% Need to add commas between groups +% of address numbers, which are passed in the macro \cmd\@affilID@temp\ in the form +% |{3}{4}{7}| if this set of authors is related to addresses, 3, 4 and 7. +% \begin{macrocode} +\def\@author@present@script#1#2{% + \begingroup +% \parskip\@author@parskip\relax + \gdef\comma@space{\textsuperscript{,\,}}% +% \@ifx{\@affilID@temp\@empty}{% +% \doauthor{#1}{#2}{}% +% }{% +% \expandafter\set@count@\@affilID@temp{}{}\@@nil +% \@ifnum{\count@=\z@}{% +% \doauthor{#1}{#2}{}% +% }{% +% \doauthor{#1}{#2}{% +% \textsuperscript{\expandafter\@affilcomma\@affilID@temp\relax\relax}% +% }% +% }% +% }% + \doauthor{#1}{#2}{% + \let\@tempa\@empty + \expandafter\@affil@present@script\@affilID@temp\relax + }% + \endgroup + \advance\@tempcnta\m@ne +}% +%\def\set@count@#1#2\@@nil{% +% \count@#1\relax +%}% +\def\@affilcomma#1#2{% + \@ifx{\z@#1}{% + \@ifx{\relax#2}{}{% + \@affilcomma{#2}% + }% + }{% + #1% + \@ifx{\relax#2}{}{% + \@ifx{\z@#2}{% + \@affilcomma + }{% + ,\,\@affilcomma{#2}% + }% + }% + }% +}% +\def\@affil@present@script#1{% + \@ifx{\relax#1}{% + \@ifx{\@tempa\@empty}{}{% + \textsuperscript{\expandafter\@affilcomma\@tempa\relax\relax}% + }% + }{% + \@ifx{\z@#1}{}{\appdef\@tempa{{#1}}}% + \@affil@present@script + }% +}% +\@provide\@author@parskip{\z@skip}% +% \end{macrocode} +% \end{macro} +% +% +% \subsubsection{Authors and affiliations in groupedaddress style} +% +% \begin{macro}{\frontmatter@author@produce@group} +% \cmd\frontmatter@author@produce@group\ is an alias for \cmd\frontmatter@author@produce, +% the procedure that formats the author/affiliation list when \classoption{groupedaddress} is in effect. +% +% In this case, the authors are being grouped above their shared addresses. +% How it works: the \cmd\@AFG@list, a list of all unique affiliation groups, is expanded. +% For each such group, all authors belonging to that group are formatted above the affiliation group. +% Collaborations are silently passed over. +% +% This procedure must ensure that any paragraph that it starts gets ended finally. +% \begin{macrocode} +\def\frontmatter@author@produce@group{% + \begingroup + \let\@author@present\@author@present@group + \frontmatter@authorformat + \let\AFS@opr \affils@present@group + \let\@listcomma\relax + \@AFG@list + \frontmatter@footnote@produce + \par + \endgroup + \frontmatter@authorbelow +}% +\@provide\frontmatter@authorbelow{}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\affils@present@group} +% Alias of \cmd\AFS@opr, this procedure is used on every element of +% \cmd\@AFG@list, the list of unique affiliation groups, +% when formatting the author/affiliation block: +% It presents all authors sharing this affiliation group. +% +% This procedure is called via the procedure +% \cmd\frontmatter@author@produce@group, an alias for \cmd\frontmatter@author@produce, +% +% Its use pertains to the class options +% \classoption{groupedaddress}, \classoption{unsortedaddress}, and \classoption{runinaddress} +% (that is, \cmd\groupauthors@sw\ is true). +% +% For each invocation, the entire \cmd\@AAC@list\ is run through (twice), +% making this process execute in $N^{2}$ time. (Is this really true?) +% \begin{macrocode} +\def\affils@present@group#1{% +% \end{macrocode} +% Count (using \cmd\@tempcnta) the authors in the author list whose affiliation is |#1|. +% +% Note that \cmd\AF@temp\ is used as a storage register for the argument +% of the procedure, for purposes of comparison using \cmd\ifx. +% +% Note: here we use \cmd\@tempcnta\ to communicate accumulate the number of authors in the current affiliation set. +% +% QUERY: does the value of \cmd\AF@temp\ and \cmd\AF@opr\ need to persist? +% +% QUERY: what value does \cmd\AU@opr\ have at this point? +% ANSWER: the \cmd\@affilID@count\ and \cmd\@affilID@match\ procedures +% assign meanings to \cmd\AU@opr\ and \cmd\CO@opr\ dynamically. +% We initialize the two operators to \cmd\undefined\ so as to catch the +% pathological cases. +% \begin{macrocode} + \begingroup + \def\AF@temp{#1}% % 0B13 + \@tempcnta\z@ + \let\AU@opr \undefined % 0B13 + \let\CO@opr \undefined % 0B13 + \def\AF@opr{\@affilID@count\AF@temp\@tempcnta}% 0B13 + \@AAC@list +% \end{macrocode} +% +% If no authors using this affiliation group, skip it. +% \begin{macrocode} + \@ifnum{\@tempcnta=\z@}{}{% +% \end{macrocode} +% +% DPC: If we have a list of length two, need to locally zap a comma. +% \begin{macrocode} + \begingroup + \frontmatter@above@affilgroup + \set@listcomma@count\@tempcnta +% \end{macrocode} +% Run through the author list, presenting (with \cmd\@author@present) +% those authors whose affiliation matched the given one. +% \begin{macrocode} + \let\AU@opr \undefined % 0B13 + \let\CO@opr \undefined % 0B13 + \def\AF@opr{\@affilID@match\AF@temp}% + \@AAC@list + \endgroup +% \end{macrocode} +% Now that the authors have all been presented, present the affiliations, grouped. +% +% Note: \cmd\@tempcnta\ is used to count the number of affiliations for this set of authors. +% \begin{macrocode} + \begingroup + \par + \frontmatter@above@affiliation + \frontmatter@affiliationfont + \let\\\frontmatter@addressnewline + \@tempcnta\z@ + \@tfor\AF@temp:=#1\do{% + \expandafter\@ifx\expandafter{\z@\AF@temp}{}{% + \advance\@tempcnta\@ne + }% + }% + \@ifnum{\@tempcnta=\tw@}{% + \let\@listcomma\relax + }{}% + \def\after@address{\@listcomma\ \@listand}% + \runinaddress@sw{% + }{% + \tightenlines@sw{% + }{% + \parskip\z@ %space between affilations in grouped mode + }% + \appdef\after@address\par + }% + \let\AFF@opr \@affil@group + \do@affil@fromgroup\@AFF@list#1\relax + \endgroup +% \end{macrocode} +% End of branch handling authors. +% \begin{macrocode} + }% +% \end{macrocode} +% +% \begin{macrocode} + \par + \endgroup +%\smallskip +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\@affilID@count} +% \begin{macro}{\@affilID@match} +% +% The control sequence name \cmd\AF@opr\ is placed in the +% author/affiliation list (\cmd\@AAC@list) for each (new) affiliation, with that affiliation as its argument. +% +% The control sequence name \cmd\AF@opr\ is an alias of a procedure +% to count the number of authors at a given affiliation (\cmd\@affilID@count) +% or to present the given authors (\cmd\@affilID@match). +% These two procedures are executed when formatting under the \classoption{groupedaddress} class option. +% \begin{macrocode} +\def\@affilID@count#1#2#3{% + \def\@tempifx{#3}% + \@ifx{\@tempifx#1}{% + \def\AU@opr{\@author@count#2}% + }{% + \let\AU@opr \@author@gobble + }% + \let\CO@opr \@collaboration@err +}% +\def\@affilID@match#1#2{% + \def\@tempifx{#2}% + \@ifx{\@tempifx#1}{% + \let\AU@opr \@author@present + }{% + \let\AU@opr \@author@gobble + }% + \let\CO@opr \@collaboration@err +}% +% \end{macrocode} +% \end{macro} +% \end{macro} +% +% \begin{macro}{\do@affil@fromgroup} +% Executes iteratively: +% selects the next address ID to print, +% and then re-execute the list of addresses to print the text of that address. +% +% Note: an argument of \cmd\relax\ breaks out of this iteration, +% while an argument of \cmd\z@\ (\cmd\noaffiliation) is a no-op. +% +% Note: we have createed an $N^{2}$ computing process. +% FIXME: can use hashing instead, you know. +% \begin{macrocode} +\def\do@affil@fromgroup#1#2{% + \@ifx{\relax#2}{}{% + \count@#2\relax + \@ifnum{\count@=\z@}{}{#1}% + \do@affil@fromgroup#1% + }% +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\@affil@group} +% The alias of \cmd\AFF@opr\ when the list \cmd\@AFF@list\ +% is executed by \cmd\do@affil@fromgroup (\classoption{groupedaddress}): +% Print the address if its number matches \cmd\count@\ (value set by \cmd\do@affil@fromgroup). +% +% Note that we detect a \cmd\noaffiliation\ here by the second parameter's being the token \cmd\blankaffiliation. +% \begin{macrocode} +\def\@affil@group#1#2#3{% + \@ifnum{\count@=#1\relax}{% + \@ifx{#2\blankaffiliation}{}{% + #2% + \@if@empty{#3}{}{% + \frontmatter@footnote{#3}% + }% + \after@address + }% + \advance\@tempcnta\m@ne + }{}% +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\@author@present@group} +% \changes{v4.d}{1998/01/31} +% {\cs{comma@space}} +% This version of \cmd\AU@opr\ applies to the \classoption{groupedaddress} class option. +% Traverse the list calling the \cmd\doauthor\ hook on each author. +% \begin{macrocode} +\def\@author@present@group#1#2{% + \gdef\comma@space{\gdef\comma@space{\textsuperscript{,\,}}}% + \doauthor{#1}{#2}{}% + \advance\@tempcnta\m@ne +}% +% \end{macrocode} +% \end{macro} +% +% +% \begin{macro}{\@pacs@produce} +% PACS, keywords and dates. +% \changes{v4.d}{1998/01/31} +% {PACS boxed to width of abstract} +% \changes{v4.0beta 2}{1999/06/20} +% {Separate option now controls production} +% \changes{v4.0beta 5}{2000/09/20} +% {Warn if no production} +% \begin{macrocode} +\def\@pacs@produce#1{% + \showPACS@sw{% + \begingroup + \frontmatter@PACS@format + \@pacs@name#1\par + \endgroup + }{% + \@if@empty{#1}{}{% + \class@warn{\PACS@warn}% + }% + }% +}% +\def\PACS@warn{If you want your PACS to appear in your output, use document class option showpacs}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\@keywords@produce} +% \changes{v4.0beta 5}{2000/09/20} +% {Keywords on same footing as PACS} +% \begin{macrocode} +\def\@keywords@produce#1{% + \showKEYS@sw{% + \begingroup + \frontmatter@keys@format + \@keys@name#1\par + \endgroup + }{% + \@if@empty{#1}{}{% + \class@warn{If you want your keywords to appear in your output, use document class option showkeys}% + }% + }% +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\frontmatter@footnote@produce@footnote} +% \begin{macro}{\frontmatter@footnote@produce@endnote} +% Produce any footnotes to appear at the bottom of the title page. +% +% If frontmatter footnotes have been redirected to the bibliography, this will be a no-op. +% ref.: \cmd\present@bibnote. +% \begin{macrocode} +\def\frontmatter@footnote@produce@footnote{% + \let\@TBN@opr\present@FM@footnote + \@FMN@list + \global\let\@FMN@list\@empty +}% +\def\present@FM@footnote#1#2{% + \begingroup + \csname c@\@mpfn\endcsname#1\relax + \def\@thefnmark{\thempfn}% + \frontmatter@footnotetext{#2}% + \endgroup +}% +\def\frontmatter@footnote@produce@endnote{% +}% +% \end{macrocode} +% \end{macro} +% \end{macro} +% +% \begin{macro}{\frontmatter@thefootnote} +% The in the title block, the footnote symbol (used for alternate affiliation and sundry) +% is a \cmd\@fnsymbol\ or a lowercase letter, as selected by a boolean. +% \begin{macrocode} +\def\frontmatter@thefootnote{% + \altaffilletter@sw{\@alph}{\@fnsymbol}\c@footnote +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\frontmatter@makefnmark} +% The default procedure for setting the footnote mark within the frontmatter. +% Journals will override this procedure. +% \begin{macrocode} +\def\frontmatter@makefnmark{% + \@textsuperscript{% + \normalfont\@thefnmark + }% +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\frontmatter@makefntext} +% This procedure is used when setting footnotes within the title block. +% \begin{macrocode} +\long\def\frontmatter@makefntext#1{% + \parindent 1em + \noindent + \Hy@raisedlink{\hyper@anchorstart{frontmatter@\the\c@footnote}\hyper@anchorend}% + \@makefnmark + #1% +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\frontmatter@setup} +% The default type specification for the title page. +% Journals will override this definition. +% \begin{macrocode} +\def\frontmatter@setup{}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\frontmatter@RRAPformat} +% \begin{macro}{\punct@RRAP} +% \begin{macro}{\produce@RRAP} +% The procedure \cmd\frontmatter@RRAPformat\ is the formatting procedure +% for the ``Received, Revised, etc.'' block on the title page. +% The macro \cmd\punct@RRAP\ produces the punctuation between the entries. +% Journals will override these two definitions. +% +% The procedure \cmd\produce@RRAP\ is fairly general, and should be used by most +% journal substyles. +% +% The procedure \cmd\frontmatter@RRAP@format\ will be overridden by the journal. +% \begin{macrocode} +\def\frontmatter@RRAPformat#1{% + \removelastskip + \begingroup + \frontmatter@RRAP@format + #1\par + \endgroup +}% +\def\punct@RRAP{; }% +\def\produce@RRAP#1{% + \@if@empty{#1}{}{% + \@ifvmode{\leavevmode}{\unskip\punct@RRAP\ignorespaces}% + #1% + }% +}% +% \end{macrocode} +% \end{macro} +% \end{macro} +% \end{macro} +% +% \begin{macro}{\frontmatter@authorformat} +% The default type specification for the author list. +% Journals will override this definition. +% \begin{macrocode} +\def\frontmatter@authorformat{}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\frontmatter@above@affilgroup} +% The default amount of space above an affiliation group +% (in a \cmd\affils@present@group\ production). +% Journals will override this definition. +% \begin{macrocode} +\def\frontmatter@above@affilgroup{}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\frontmatter@above@affiliation} +% \begin{macro}{\frontmatter@above@affiliation@script} +% The \cmd\frontmatter@above@affiliation\ is the amount of space above affiliations in the \classoption{groupedaddress} +% style, and \cmd\frontmatter@above@affiliation@script\ is that for \classoption{superscriptaddress}. +% Journals will override this definition. +% \begin{macrocode} +\def\frontmatter@above@affiliation{}% +\def\frontmatter@above@affiliation@script{}% +% \end{macrocode} +% \end{macro} +% \end{macro} +% +% \begin{macro}{\frontmatter@affiliationfont} +% The default type specification for the affiliation. +% Journals will override this definition. +% \begin{macrocode} +\def\frontmatter@affiliationfont{}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\frontmatter@RRAP@format} +% The default type specification for the dates. +% Journals will override this definition. +% \begin{macrocode} +\def\frontmatter@RRAP@format{}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\frontmatter@PACS@format} +% Specify the formatting of the title page PACS statement. +% +% Journals will override. +% \begin{macrocode} +\def\frontmatter@PACS@format{}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\frontmatter@keys@format} +% Specify the formatting of the title page keywords statement. +% +% Journals will override. +% \begin{macrocode} +\def\frontmatter@keys@format{}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\frontmatter@finalspace} +% +% \begin{macrocode} +\def\frontmatter@finalspace{\addvspace{18\p@}} +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\frontmatter@addressnewline} +% \changes{v4.h}{1998/06/25} +% {new hook} +% The definition of \cmd\\\ for address handling. Default puts all the `lines' +% on a run-in line, separated by comma and space. +% DPC: was \newline space between lines of addresss. +% \begin{macrocode} +\def\frontmatter@addressnewline{% + \@ifhmode{\skip@\lastskip\unskip\unpenalty\break\hskip\skip@}{}% + % was: \vskip-.5ex +}% +% \end{macrocode} +% \end{macro} +% +% +% \begin{macro}{\frontmatter@preabstractspace} +% \begin{macro}{\frontmatter@postabstractspace} +% Space above and space below abstract in title block +% \begin{macrocode} +\def\frontmatter@preabstractspace{5.5\p@} +\def\frontmatter@postabstractspace{6.5\p@} +% \end{macrocode} +% \end{macro} +% \end{macro} +% +% \begin{macro}{\aftermaketitle@chk} +% Error-checking code. +% \begin{macrocode} +\def\aftermaketitle@chk#1{% + \@ifx{\maketitle\relax}{% + \class@err{\protect#1 must be used before \protect\maketitle}% + }{}% +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\ps@titlepage} +% Default page style for title page. +% Journals will override this procedure. +% \begin{macrocode} +\def\ps@titlepage{\ps@empty}% +% \end{macrocode} +% \end{macro} +% +%FIXME: there is a limitation to the default meaning of +% \cmd\@startpage, the label ``FirstPage'' is only +% defined if the document has a \cmd\maketitle\ command. +% +% \begin{macrocode} +\def\volumeyear#1{\gdef\@volumeyear{#1}}% +\def\@volumeyear{}% +\def\volumenumber#1{\gdef\@volumenumber{#1}}% +\def\@volumenumber{}% +\def\issuenumber#1{\gdef\@issuenumber{#1}}% +\def\@issuenumber{}% +\def\eid#1{\gdef\@eid{#1}}% +\def\@eid{}% +% +\def\startpage#1{\gdef\@startpage{#1}\c@page#1\relax}% +\def\@startpage{\pageref{FirstPage}}% +\def\endpage#1{\gdef\@endpage{#1}}% +\def\@endpage{\pageref{LastPage}}% +% \end{macrocode} +% +% +% \subsection{Printing out the ``list-of'' elements}% +% +% FIXME: The \cmd\appendix@toc\ procedure should change the meaning +% of \cmd\l@section\ so that the \cmd\section s can be appropriately formatted, +% reflecting their status as appendices. +% +% \begin{macrocode} +\def\print@toc#1{% + \begingroup +% \c@secnumdepth-\maxdimen + \expandafter\section + \expandafter*% + \expandafter{% + \csname#1name\endcsname + }% + \let\appendix\appendix@toc + \@starttoc{#1}% + \endgroup +}% +\def\appendix@toc{}% +% \end{macrocode} +% +% +% \section{Body} +% +% \subsection{counters} +% The following definitions are probably identical to those in +% \file{classes.dtx}% +% \begin{macrocode} +\def\labelenumi{\theenumi.} +\def\theenumi{\arabic{enumi}} +% \end{macrocode} +% +% \begin{macrocode} +\def\labelenumii{(\theenumii)} +\def\theenumii{\alph{enumii}} +\def\p@enumii{\theenumi} +% \end{macrocode} +% +% \begin{macrocode} +\def\labelenumiii{\theenumiii.} +\def\theenumiii{\roman{enumiii}} +\def\p@enumiii{\theenumi(\theenumii)} +% \end{macrocode} +% +% \begin{macrocode} +\def\labelenumiv{\theenumiv.} +\def\theenumiv{\Alph{enumiv}} +\def\p@enumiv{\p@enumiii\theenumiii} +% \end{macrocode} +% +% \begin{macrocode} +\def\labelitemi{\textbullet} +\def\labelitemii{\normalfont\bfseries\textendash} +\def\labelitemiii{\textasteriskcentered} +\def\labelitemiv{\textperiodcentered} +% \end{macrocode} +% +% \begin{macrocode} +\pagenumbering{arabic} +% \end{macrocode} +% +% \subsection{float parameters} +% +% from the old aps.sty. (DPC: same as article I think) +% AO: here, \LaTeX's standard classes fail very poorly +% (the price of backward compatability): the values for +% \cmd\floatpagefraction\ and \cmd\dblfloatpagefraction\ +% need to be raised to avoid creating extremely short +% float pages. +% \begin{macrocode} +\setcounter{topnumber}{2} +\def\topfraction{.9} +\setcounter{bottomnumber}{1} +\def\bottomfraction{.9} +\setcounter{totalnumber}{3} +\def\textfraction{.1} +\def\floatpagefraction{.9} +\setcounter{dbltopnumber}{2} +\def\dbltopfraction{.9} +\def\dblfloatpagefraction{.9} +% \end{macrocode} +% +% \subsection{List Environments} +% +% \begin{macrocode} +\newenvironment{verse}{% + \let\\=\@centercr + \list{}{% + \itemsep\z@ \itemindent -1.5em\listparindent \itemindent + \rightmargin\leftmargin\advance\leftmargin 1.5em}\item[]% +}{% + \endlist +}% +% \end{macrocode} +% +% \begin{macrocode} +\newenvironment{quotation}{% + \list{}{% + \listparindent 1.5em + \itemindent\listparindent + \rightmargin\leftmargin \parsep \z@ \@plus\p@}\item[]% +}{% + \endlist +}% +% \end{macrocode} +% +% \begin{macrocode} +\newenvironment{quote}{% + \list{}{% + \rightmargin\leftmargin}\item[]% +}{% + \endlist +}% +% \end{macrocode} +% +% \begin{macrocode} +\def\descriptionlabel#1{% + \hspace\labelsep \normalfont\bfseries #1% +}% +% \end{macrocode} +% +% \begin{macrocode} +\newenvironment{description}{% + \list{}{% + \labelwidth\z@ \itemindent-\leftmargin + \let\makelabel\descriptionlabel + }% +}{% + \endlist +}% +% \end{macrocode} +% +% \subsection{Sectioning Commands} +% +% \subsubsection{Counters and Their Productions}% +% \begin{macrocode} +\newcounter{part} +\newcounter{section} +\newcounter{subsection}[section] +\newcounter{subsubsection}[subsection] +\newcounter{paragraph}[subsubsection] +\newcounter{subparagraph}[paragraph] +% \end{macrocode} +% +% \begin{macrocode} +\def\thepart {\Roman{part}} % +\def\thesection {\Roman{section}} +\def\p@section {} +\def\thesubsection {\Alph{subsection}} +\def\p@subsection {\thesection\,} +\def\thesubsubsection {\arabic{subsubsection}} +\def\p@subsubsection {\thesection\,\thesubsection\,} +\def\theparagraph {\alph{paragraph}} +\def\p@paragraph {\thesection\,\thesubsection\,\thesubsubsection\,} +\def\thesubparagraph {\arabic{subparagraph}} +\def\p@subparagraph {\thesection\,\thesubsection\,\thesubsubsection\,\theparagraph\,} +% \end{macrocode} +% +% \subsubsection{The Acknowledgments Environment}% +% This user-level markup produces a head introducing +% the acknowledgments, and acts as a wrapper for the text. +% In this implementation, it is an unnumbered section, +% but appears within the toc. +% +% For compatiability's sake, we implement it under the alternative +% spelling \env{acknowledgements}. +% \changes{v4.0beta 3}{1999/11/13} +% {also spelled ``acknowledgements''.} +% \changes{v4.0beta 4}{2000/05/18} +% {But alternative spelling is deprecated.} +% \begin{macrocode} +\newenvironment{acknowledgments}{% + \acknowledgments@sw{% + \section*{\acknowledgmentsname}% + }{% + \par + \phantomsection + \addcontentsline{toc}{section}{\protect\numberline{}\acknowledgmentsname}% + }% +}{% + \par +}% +\@booleantrue\acknowledgments@sw +\newenvironment{acknowledgements}{% + \replace@environment{acknowledgements}{acknowledgments}% +}{% + \endacknowledgments +}% +% \end{macrocode} +% +% \subsubsection{Part Opener}% +% section setup copied verbatim from revtex3 aps/osa. +% Does not explicitly depend on pointsize options. +% \begin{macrocode} +\def\part{\par + \addvspace{4ex}% + \@afterindentfalse + \secdef\@part\@spart}% +% \end{macrocode} +% +% \begin{macrocode} +\def\@part[#1]#2{% + \@ifnum{\c@secnumdepth >\m@ne}{% + \refstepcounter{part}% + \addcontentsline{toc}{part}{\thepart\hspace{1em}#1}% + }{% + \addcontentsline{toc}{part}{#1}% + }% + \begingroup + \parindent \z@ \raggedright + \interlinepenalty\@M + \@ifnum{\c@secnumdepth >\m@ne}{% + \Large \bf \partname~\thepart% + \par\nobreak + }{}% + \huge \bf + #2% + \markboth{}{}\par + \endgroup + \nobreak + \vskip 3ex + \@afterheading +}% +% \end{macrocode} +% +% \begin{macrocode} +\def\@spart#1{{\parindent \z@ \raggedright + \interlinepenalty\@M + \huge \bf + #1\par} + \nobreak + \vskip 3ex + \@afterheading} +% \end{macrocode} +% +% \subsubsection{Stacked Heads}% +% Here are the class default definitions for sectioning commands. +% A society or a journal substyle will likely override these definitions. +% +% In doing so, you can customize the formatting for a particular level by defining, +% e.g., \cmd\@hangfrom@section\ or \cmd\@subsectioncntformat. +% \begin{macrocode} +\def\section{% + \@startsection + {section}% + {1}% + {\z@}% + {0.8cm \@plus1ex \@minus .2ex}% + {0.5cm}% + {\normalfont\small\bfseries}% +}% +% \end{macrocode} +% +% \begin{macrocode} +\def\subsection{% + \@startsection + {subsection}% + {2}% + {\z@}% + {.8cm \@plus1ex \@minus .2ex}% + {.5cm}% + {\normalfont\small\bfseries}% +}% +% \end{macrocode} +% +% \begin{macrocode} +\def\subsubsection{% + \@startsection + {subsubsection}% + {3}% + {\z@}% + {.8cm \@plus1ex \@minus .2ex}% + {.5cm}% + {\normalfont\small\itshape}% +}% +% \end{macrocode} +% +% \subsubsection{Runin Heads}% +% \begin{macrocode} +\def\paragraph{% + \@startsection + {paragraph}% + {4}% + {\parindent}% + {\z@}% + {-1em}% + {\normalfont\normalsize\itshape}% +}% +% \end{macrocode} +% +% \begin{macrocode} +\def\subparagraph{% + \@startsection + {subparagraph}% + {5}% + {\parindent}% + {3.25ex \@plus1ex \@minus .2ex}% + {-1em}% + {\normalfont\normalsize\bfseries}% +}% +% \end{macrocode} +% +% +% \subsection{Math} +% +% \begin{macro}{\theequation} +% We change the production of the equation counter so that +% we can accomodate the \classoption{eqsecnum} option. +% \begin{macrocode} +\def\theequation{% + \theequation@prefix\arabic{equation}% +}% +\def\theequation@prefix{}% +% \end{macrocode} +% \end{macro} +% +% +% \subsection{Type Size-Dependent Settings} +% +% Note that many parameters, as well as the meaning of +% \cmd\normalsize, are deferred until +% \cmd\AtEndOfClass\ time. +% Therefore, code elsewhere in this class \emph{should not} +% assume these things' values have been set! +% \begin{macrocode} +\AtEndOfClass{% + \expandafter\input\expandafter{\@pointsize pt\substyle@ext}% +}% +% \end{macrocode} +% +% +% \subsection{All Point Sizes} +% \begin{macrocode} +\setcounter{secnumdepth}{4} +% \end{macrocode} +% +% \begin{macrocode} +\lineskip 1pt +\normallineskip 1pt +\def\baselinestretch{1}% +% \end{macrocode} +% +% \begin{macrocode} +\@lowpenalty 51 +\@medpenalty 151 +\@highpenalty 301 +% \end{macrocode} +% +% \begin{macrocode} +\@beginparpenalty -\@lowpenalty +\@endparpenalty -\@lowpenalty +\@itempenalty -\@lowpenalty +% \end{macrocode} +% +% \begin{macrocode} +\arraycolsep 3pt +\tabcolsep 2pt +\arrayrulewidth .4pt +\doublerulesep 2pt +% \end{macrocode} +% +% \begin{macrocode} +\skip\@mpfootins = 0pt +% \end{macrocode} +% +% \begin{macrocode} +\fboxsep = 3.0pt +\fboxrule = 0.4pt +% \end{macrocode} +% +% \subsection{Figures} +% \changes{v4.0beta 2}{1999/06/20} +% {Processing delayed to \cs{AtBeginDocument} time} +% +% \begin{environment}{figure} +% We define the \env{figure} environment. +% Later, we will horse around with its meaning +% in order to accomodate \cmd\floats@sw. +% \begin{macrocode} +\newenvironment{figure} + {\@float{figure}} + {\end@float} +\newenvironment{figure*} + {\@dblfloat{figure}} + {\end@dblfloat} +% \end{macrocode} +% \end{environment} +% +% \begin{macro}{\@makecaption} +% If caption is one line long, to be centered; if lines turn, then set justified. +% \begin{macrocode} +\newlength\abovecaptionskip +\newlength\belowcaptionskip +\setlength\abovecaptionskip{10\p@} +\setlength\belowcaptionskip{2\p@} +% \end{macrocode} +% +% \changes{v4.0beta 2}{1999/06/20} +% {One-line caption sets flush left.} +% \changes{v4.0beta 3}{1999/11/13} +% {Prevent an inner footnote from performing twice} +% \begin{macrocode} +\long\def\@makecaption#1#2{% + \par +% \nobreak + \vskip\abovecaptionskip + \begingroup + \small\rmfamily + \sbox\@tempboxa{% + \let\\\heading@cr + #1: #2% + }% + \@ifdim{\wd\@tempboxa >\hsize}{% + \begingroup + \samepage + \flushing + \let\footnote\@footnotemark@gobble + #1: #2\par + \endgroup + }{% + \global \@minipagefalse + \hb@xt@\hsize{\hfil\unhbox\@tempboxa\hfil}% + }% + \endgroup + \vskip\belowcaptionskip +}% +\def\@footnotemark@gobble{% + \@footnotemark + \@ifnextchar[{\@gobble@opt@i}{\@gobble}% +}% +\def\@gobble@opt@i[#1]#2{}% +\def\@mpmakefntext#1{% + \flushing + \parindent=1em + \noindent + \hb@xt@1em{\hss\@makefnmark}% + #1% +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\thefigure} +% The figure counter and float placement defaults. +% \begin{macrocode} +\newcounter{figure} +\renewcommand \thefigure {\@arabic\c@figure} +% \end{macrocode} +% Note that we give the `!' modifier by default. +% This is an effort to avoid the syndrome wherein +% a deferred float finds itself unqualified for +% placement, thereby getting carried until \cmd\clearpage. +% \begin{macrocode} +\def\fps@figure{tbp} +\def\ftype@figure{1} +\def\ext@figure{lof} +\def\fnum@figure{\figurename~\thefigure} +% \end{macrocode} +% \end{macro} +% +% \subsubsection{Deferring \env{figure} Floats}% +% +% We install a hook at \cmd\AtBeginDocument\ time which determines if +% figures will float or will be deferred until +% \cmd\printfigures\ time. +% \begin{macrocode} +\appdef\class@documenthook{% + \do@if@floats{figure}{.fgx}% +}% +\appdef\class@enddocumenthook{% + \printfigures\relax +}% +% \end{macrocode} +% +% \begin{macro}{\printfigures} +% The user-level command \cmd\printfigures\ determines +% where the figures are to appear in a document in which +% \cmd\floats@sw\ is false. +% If the user invokes the \classoption{endfloats} class option and fails to +% insert a \cmd\printfigures\ command, +% the figures will be printed at the end of the document. +% If the command is given, but floats are not being deferred, +% it amounts to a no-op. +% \changes{v4.0beta 2}{1999/06/20} +% {call \cs{print@floats}} +% \changes{v4.0beta 3}{1999/11/13} +% {*-form mandates pagebreak} +% \begin{macrocode} +\newcommand\printfigures{% + \@ifstar{\true@sw}{\floatp@sw{\true@sw}{\false@sw}}% + {% + \print@float{figure}{\oneapage}% + }{% + \print@float{figure}{}% + }% +}% +% \end{macrocode} +% \end{macro} +% +% +% \begin{macro}{\@xfloat@prep} +% We patch into the procedure \cmd\@xfloat@prep. +% This patch applies to all floats (not \env{figure} alone) +% and makes the type center. +% \begin{macrocode} +\appdef\@xfloat@prep{% + \appdef\@parboxrestore{\centering}% + \let\@makefnmark\@makefnmark@latex + \let\@footnotemark\@footnotemark@latex +}% +% \end{macrocode} +% \end{macro} +% +% +% \subsection{Tables} +% \changes{v4.0beta 2}{1999/06/20} +% {Processing delayed to \cs{AtBeginDocument} time} +% +% DPC: More or less taken from \file{revtex2} \file{aps.sty}, +% but using dcolumn for decimal alignment. +% \begin{environment}{table} +% We define the \env{table} environment. +% Later, we will horse around with its meaning +% in order to accomodate \cmd\floats@sw. +% \begin{macrocode} +\newenvironment{table} + {\@float{table}} + {\end@float} +\newenvironment{table*} + {\@dblfloat{table}} + {\end@dblfloat} +% \end{macrocode} +% \end{environment} +% +% \begin{macro}{\thetable} +% Table counter and default float placement declarations. +% \begin{macrocode} +\newcounter{table} +\renewcommand\thetable{\@Roman\c@table} +% \end{macrocode} +% Note that we give the `!' modifier by default. +% This is an effort to avoid the syndrome wherein +% a deferred float finds itself unqualified for +% placement, thereby getting carried until \cmd\clearpage. +% \begin{macrocode} +\def\fps@table{tbp} +\def\ftype@table{2} +\def\ext@table{lot} +\def\fnum@table{\tablename~\thetable} +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\table@hook} +% \begin{macro}{\squeezetable} +% Assign a meaning to the hook installed into float processing. +% +% By default floats are \cmd\small. The \cmd\squeezetable\ declaration +% makes them smaller (\cmd\scriptsize). In general you can locally +% redefine \cmd\table@hook\ to be whatever you like. +% (DPC: |\Huge\color{magenta}|\ldots?) +% \begin{macrocode} +\def\table@hook{\small}% +\def\squeezetable{\def\table@hook{\scriptsize}}% +\appdef\@floatboxreset{\table@hook}% +% \end{macrocode} +% \end{macro} +% \end{macro} +% +% \subsubsection{Deferring \env{table} Floats}% +% +% We install a hook at +% \cmd\AtBeginDocument\ time which determines if +% tables will float or will be deferred until +% \cmd\printtables\ time. +% +% We also deal with the possibility of \env{longtable} environments. +% +% \begin{macrocode} +\appdef\class@documenthook{% + \floats@sw{}{% + \let@environment{longtable@float}{longtable}% + \let@environment{longtable}{longtable@write}% + \let@environment{longtable*@float}{longtable*}% + \let@environment{longtable*}{longtable*@write}% + \let@environment{turnpage@float}{turnpage}% + \let@environment{turnpage}{turnpage@write}% + }% + \do@if@floats{table}{.tbx}% +}% +\appdef\class@enddocumenthook{% + \printtables\relax +}% +\newenvironment{longtable@write}{% + \write@@float{longtable}{table}% +}{% + \endwrite@float +}% +\newenvironment{longtable*@write}{% + \write@@float{longtable*}{table}% +}{% + \endwrite@float +}% +\newenvironment{turnpage@write}{% + \immediate\write\tablewrite{\string\begin{turnpage}}% +}{% + \immediate\write\tablewrite{\string\end{turnpage}}% +}% +% \end{macrocode} +% +% \begin{macro}{\printtables} +% The user-level command \cmd\printtables\ determines +% where the tables are to appear in a document in which +% \cmd\floats@sw\ is false. +% If the user invokes the \classoption{nofloats} and fails to +% insert a \cmd\printtables\ command, +% the tables will be printed at the end of the document. +% If the command is given, but floats are not being deferred, +% it amounts to a no-op. +% \changes{v4.0beta 2}{1999/06/20} +% {only execute if appropriate} +% \changes{v4.0beta 3}{1999/11/13} +% {*-form mandates pagebreak} +% \changes{v4.0beta 4}{2000/05/17} +% {make longtable trigger the head, too} +% \begin{macrocode} +\newcommand\printtables{% + \begingroup + \let@environment{longtable}{longtable@float}% + \let@environment{longtable*}{longtable*@float}% + \let@environment{turnpage}{turnpage@anchored}% + \prepdef\longtable{\trigger@float@par}% + \expandafter\prepdef\csname longtable*\endcsname{\trigger@float@par}% + \expandafter\prepdef\csname table@floats\endcsname{% + \onecolumngrid@push + }% + \expandafter\appdef\csname endtable@floats\endcsname{% + \onecolumngrid@pop + }% + \@ifstar{\true@sw}{\floatp@sw{\true@sw}{\false@sw}}% + {% + \print@float{table}{\oneapage}% + }{% + \print@float{table}{}% + }% + \endgroup +}% +\newenvironment{turnpage@anchored}{% + \onecolumngrid@push + \setbox\z@\vbox to\textwidth\bgroup + \columnwidth\textheight +}{% + \vfil + \egroup + \rotatebox{90}{\box\z@}% + \onecolumngrid@pop +}% +% \end{macrocode} +% \end{macro} +% +% +% +% \section{Tabular} +% Every tabular has a double (Scotch) rule above and below. +% The column specifier ``d'' is implemented using the \classname{dcolumn} +% package, if available. +% FIXME: always load dcolumn! +% \begin{macro}{\tabular@hook} +% \begin{macro}{\endtabular@hook} +% \begin{environment}{ruledtabular} +% \begin{macrocode} +\def\endtabular@hook{}% +%\RequirePackage{dcolumn}% +\appdef\class@documenthook{% + \@ifpackageloaded{dcolumn}{% + \newcolumntype{d}{D{.}{.}{-1}}% + }{}% +}% +\def\toprule{\hline\hline}% +\def\colrule{\hline}% +\def\botrule{\hline\hline}% +\newenvironment{ruledtabular}{% + \def\array@default{v}% + \appdef\tabular@hook{\def\@halignto{to\hsize}}% + \let\tableft@skip@default\tableft@skip + \let\tableft@skip\tableft@skip@float + \let\tabmid@skip@default\tabmid@skip + \let\tabmid@skip\tabmid@skip@float + \let\tabright@skip@default\tabright@skip + \let\tabright@skip\tabright@skip@float + \let\array@row@pre@default\array@row@pre + \let\array@row@pre\array@row@pre@float + \let\array@row@pst@default\array@row@pst + \let\array@row@pst\array@row@pst@float + \appdef\array@row@rst{% + \let\array@row@pre\array@row@pre@default + \let\array@row@pst\array@row@pst@default + \let\tableft@skip\tableft@skip@default + \let\tabmid@skip\tabmid@skip@default + \let\tabright@skip\tabright@skip@default + \appdef\tabular@hook{\let\@halignto\@empty}% + }% +}{% +}% +% \end{macrocode} +% \end{environment} +% \end{macro} +% \end{macro} +% +% \section{Footnote Text} +% +% \changes{v4.g}{1998/06/10} +% {single space footnotes} +% \changes{v4.0beta 2}{1999/06/20} +% {% +% Frank Mittelbach, has stated in \protect\classname{multicol}: +% ``The kernel command \cs{@footnotetext} should not be modified.'' +% Thus, I have removed David Carlisle's redefinition of that command. +% Note, however, that later versions of \protect\classname{multicol} do not require +% this workaround. Belt and suspenders. +% }% +% +% \begin{macro}{\@makefntext} +% \begin{macrocode} +\def\@makefntext#1{% + \def\baselinestretch{1}% + \reset@font\footnotesize + \parindent 1em% + \noindent + \hb@xt@1.8em{% + \Hy@raisedlink{\hyper@anchorstart{footnote@\the\c@footnote}\hyper@anchorend}% + \hss\@makefnmark + }% + #1\par +}% +% \end{macrocode} +% \end{macro} +% +% +% \subsection{Citations, Bibliography, Endnotes} +% +% \subsubsection{Bibliography} +% +% Load Patrick Daly's \classname{natbib} package, +% \url{ftp://ctan.tug.org/macros/latex/contrib/supported/natbib} +% +% Note that \classname{natbib} assumes that it loads +% over a document class, such as the \classname{article} class, +% that has already defined \env{thebibliography} and \cmd\@listi. +% +% Note also that \classname{natbib} also installs a command +% \cmd\NAT@set@cites\ into \cmd\AtBeginDocument\ which presumes +% that the proper \cmd\bibpunct\ command has been issued. +% +% We have a problem with \classname{natbib} in that it does not +% accomodate late binding of its package options. This means +% that those options must be known at the time the package +% loads. Therefore, we select \classoption{sort\&compress} +% now, and thereby foreclose on alternative uses of \classname{natbib}. +% +% Certain societies may prefer an alternative; they are out of luck. +% If you find yourself in this position, please let the author +% of \classname{natbib} know that his package should be modified to +% accomodate late binding of package options. +% +% To remedy the situation, note that the macro \cmd\NAT@sort\ controls +% whether citations are left alone (\cmd\NAT@sort=0), sorted (\cmd\NAT@sort=1), or sorted and compressed (\cmd\NAT@sort=2). +% Since we give \classname{natbib} the \classoption{sort\&compress} option, +% if you prefer \classoption{sort}, you need only define \cmd\NAT@sort\ to be 1. +% However, if you prefer the effect of having neither \classoption{sort} nor \classoption{sort\&compress}, +% you must define \cmd\NAT@sort\ to be 0 and you must also define +% \cmd\NAT@sort@cites: +%\begin{verbatim} +%\renewcommand\NAT@sort@cites[1]{\edef\NAT@cite@list{#1}}% +%\end{verbatim} +% +% For other \classname{natbib} customizations, you may proceed as if you were going to use +% the \file{natbib.cfg} file: anything that you can modify by this means is fair game. +% Once \revtex\ is finished loading, you can assert any definitions for \classname{natbib} that you wish. +% +% \begin{macrocode} +\newenvironment{thebibliography}{}{}% +\let\@listi\@empty +\RequirePackage[sort&compress]{natbib}[1999/05/28 7.0]% +\let\NAT@thebibliography\thebibliography +\let\NAT@endthebibliography\endthebibliography +\let\bibliographystyle@latex\bibliographystyle +% \end{macrocode} +% +% \begin{macro}{\NAT@citesuper} +% We redefine \classname{natbib}'s procedure \cmd\NAT@citesuper, +% which is executed when setting a superscript citation. +% The \cmd\hspace\ is removed: in any case, it +% should really be \cmd\hspace*, to prevent an unwanted pagebreak. +% \begin{macrocode} +\renewcommand\NAT@citesuper[3]{% + \ifNAT@swa + \leavevmode + \unskip +% \hspace{1\p@}% + \textsuperscript{#1}% + \if*#3*\else\ (#3)\fi + \else + #1% + \fi + \endgroup +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\@makefnmark@cite} +% We define a procedure that will set a footnote mark the same way that +% a citation is set. If footnotes are put in the bibliography with +% \cmd\footinbib@sw, then the corresponding mark should look the same +% as the result of a \cmd\cite. This is how we do it. +% \begin{macrocode} +\def\@makefnmark@cite{\begingroup\NAT@swatrue\@cite{{\@thefnmark}}{}{}}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\bibliographystyle} +% We arrange for the selection of bibliography style +% to occur either due to the document's explicit +% \cmd\bibliographystyle\ statement or +% via the journal substyle. +% +% Note that \revtex\ is now incompatible with +% any package that patches \cmd\bibliographystyle, +% except for those (like \classname{natbib}) that load before here. +% \begin{macrocode} +\let\bibliographystyle@latex\bibliographystyle +\def\bibliographystyle{\def\@bibstyle}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\bibsection} +% \begin{macro}{\NAT@bibsetnum} +% \begin{macro}{\NAT@bibsetup} +% \begin{macro}{\bibpreamble} +% \begin{macro}{\newblock} +% \begin{macro}{\bibnumfmt} +% We define the sectioning command to use when starting the bibliography +% and gently coax \classname{natbib} into using +% the formatting procedures that \emph{we} want it to use. +% +% This way of setting up \env{thebibliography} +% automatically sets the label width based on +% the largest number used within the bibliography. +% This scheme will not work properly using the +% author/year style of bib entry, though. +% +% We define \cmd\bibnumfmt\ to be \cmd\place@bibnumber, which is a macro +% managed by \revtex. If the document defines \cmd\bibnumfmt, then that +% definition will be used instead, which is what the \classname{natbib} +% package gives as its programming interface. +% +% And if the substyle does not give a meaning to \cmd\place@bibnumber, +% we give it that of the original \cmd\bibnumfmt. +% \begin{macrocode} +\def\bibsection{% + \@ifx@empty\refname{% + \par + }{% + \let\@hangfroms@section\@hang@froms + \section*{\refname}% + \@nobreaktrue + }% +}% +\let\NATx@bibsetnum\NAT@bibsetnum +\def\NAT@bibsetnum#1{% + \setlength{\topsep}{\z@}% + \NATx@bibsetnum{\ref{LastBibItem}}% +}% +\let\NATx@bibsetup\NAT@bibsetup +\def\NAT@bibsetup{% + \setlength{\labelwidth}{\z@}% + \setlength{\labelsep}{\z@}% + \setlength{\itemindent}{\z@}% + \setlength{\listparindent}{\z@}% + \setlength{\topsep}{\z@}% + \setlength{\parsep}{\z@}% + \NATx@bibsetup +}% +\let\bibpreamble\@empty +\def\newblock{\ }% +\let\NATx@bibnumfmt\bibnumfmt +\def\bibnumfmt{\place@bibnumber}% +\AtEndOfClass{% + \appdef\class@documenthook{% + \@ifxundefined\place@bibnumber{% + \let\place@bibnumber\NATx@bibnumfmt + }{}% + }% +}% +% \end{macrocode} +% \end{macro} +% \end{macro} +% \end{macro} +% \end{macro} +% \end{macro} +% \end{macro} +% +% \begin{macro}{\NAT@onlinecite} +% We anticipate using numerical citations. +% If superscript citations are used, we nonetheless need a way to +% use numerical citations as required by the author. +% These should be accessible via the \cmd\citet\ command. +% +% Therefore we remember how to do a numerical citation +% even when the superscript citation has been selected. +% +% Can you say ``inelegant hack''? I knew you could! +% \begin{macrocode} +\let\NAT@citeyear\citeyear +\def\NAT@onlinecite#1{% + \begingroup + \let\@cite\NAT@citenum + \let\NAT@mbox\mbox + \let\citeyear\NAT@citeyear + \def\NAT@space{ }% + \cite{#1}% + \endgroup +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{+} +% \begin{macro}{\textcite} +% We extend \classname{natbib}'s syntax with two commands to set a citation +% on the baseline (as opposed to superscripted) and as text (rather than parenthetical), respectively. +% A journal substyle that makes citations be superscripted or parenthetical as the case may be, should ensure +% that the author has continued access to these two styles. +% +% \begin{macrocode} +\appdef\class@inithook{% + \@ifxundefined\@onlinecite{\def\@onlinecite{\citealp}}{}% + \@ifxundefined\@textcite{\def\@textcite{\citet}}{}% +}% +\DeclareRobustCommand\onlinecite{\@onlinecite}% +\DeclareRobustCommand\textcite{\@textcite}% +% \end{macrocode} +% \end{macro} +% \end{macro} +% +% \begin{environment}{thebibliography}% +% \begin{macro}{\present@bibnote}% +% We put a tail patch into \cmd\thebibliogrphy\ and +% a headpatch into \cmd\endthebibliography. +% +% We also apply a fix to the \cmd\list\ procedure +% so that, at least within the bibliography, +% items are not formatted via \cmd\parshape, +% which caused problems during column balancing. +% (Other list-like environments may also require fixing, +% but we will save that for another day.) +% +% Here we provide a default treatment for frontmatter notes +% deferred to the bibliography; a journal substyle +% might want to override the definition of \cmd\present@bibnote. +% \begin{macrocode} +\renewenvironment{thebibliography}[1]{% + \NAT@thebibliography{#1}% + \@clubpenalty\clubpenalty + \let\@TBN@opr\present@bibnote + \@FMN@list +}{% + \@endnotesinbib + \edef\@currentlabel{\arabic{NAT@ctr}}% + \label{LastBibItem}% + \NAT@endthebibliography + \global\let\auto@bib\@empty +}% +\def\present@bibnote#1#2{% + \item[% + \textsuperscript{% + \normalfont + \Hy@raisedlink{\hyper@anchorstart{frontmatter@#1}\hyper@anchorend}% + \begingroup + \c@footnote#1\relax + \frontmatter@thefootnote + \endgroup + }% + ]#2\par +}% +% \end{macrocode} +% \end{macro} +% \end{environment} +% +% \begin{macro}{\place@bibnumber@sup} +% \begin{macro}{\place@bibnumber@inl} +% For producing the \cmd\bibitem\ device, +% we define two procedures to select from. +% +% Note that we could have used \classname{natbib}'s \cmd\ifNAT@super\ +% switch, but it does not allow for altering the meaning of \cmd\bibnumfmt. +% \begin{macrocode} +\def\place@bibnumber@sup#1{\textsuperscript{#1}}% +\def\place@bibnumber@inl#1{[#1]}% +% \end{macrocode} +% \end{macro} +% \end{macro} +% +%% \begin{macro}{\@cite} +%% \begin{macro}{\nocite} +% \begin{macro}{write@bibliographystyle} +% We wish to delay committing the \cmd\bibliographystyle\ until as late as possible. +% The journal substyle will define a default bibliography style, and +% the document's explicit \cmd\bibliographystyle\ command, if any, will override that default. +%% For the sake of Bib\TeX, we must execute the \cmd\bibliographystyle\ command +%% (putting a \cmd\bibstyle\ command into the auxiliary file) +%% somewhere before the first \cmd\cite\ command (or any other like command), +%% which puts a \cmd\citation\ command into the auxiliary file. +%% +%% Therefore, we headpatch the two procedures in \classname{natbib} that can result in +%% the appearance in the auxiliary file of a \cmd\citation\ command. +%% These two procedures are \cmd\@cite\ and \cmd\nocite. +%% +%% Not! +% +% The \cmd\bibstyle\ command may appear quite late in the \filename{.aux} file. +% We now delay the automatic writing of the \cmd\bibstyle\ command to the end of the job. +% +% The procedure \cmd\write@bibliographystyle\ tests whether a \cmd\bibliographystyle\ command has +% already been given. +% If not, it effectively executes the needed \cmd\bibliographystyle\ command, then neutralizes itself +% (we only need to do this once per job). +% +% \begin{macrocode} +%\AtBeginDocument{% +% \let\natbib@@cite\@cite +% \def\@cite{\write@bibliographystyle\natbib@@cite}% +% \let\natbib@nocite\nocite +% \def\nocite{\write@bibliographystyle\natbib@nocite}% +%}% +\def\write@bibliographystyle{% + \@ifxundefined\@bibstyle{}{% + \expandafter\bibliographystyle@latex\expandafter{\@bibstyle}% + }% + \global\let\write@bibliographystyle\relax +}% +\AtEndDocument{\write@bibliographystyle}% +% \end{macrocode} +% \end{macro} +%% \end{macro} +%% \end{macro} +% +% \begin{macro}{\super@cite} +% \begin{macro}{\super@cite@} +% The \cmd\super@cite\ procedure takes the place of \cmd\@citex\ and implements the feature of the +% \classoption{citeautoscript} class option. +% This means that the document should be marked up as if citations were \emph{not} superscripted, +% and then if you select a journal substyle that has superscripted citations, \revtex\ will do its best +% to alter the formatting of the \cmd\cite s to accomodate superscript style. +% +% Note that the commands \cmd\onlinecite\ and \cmd\textcite\ are not affected by this procedure. +% +% Here's a subtle one: when is the argument of \cmd\super@cite@@\ not the same as the token \cmd\@let@token? +% Answer: when the latter is \cmd\@sptoken! This case has to be handled separately. +% +% \begin{macrocode} +\def\super@cite[#1][#2]#3{% + \leavevmode + \skip@\lastskip + \unskip + \begingroup + \def\@tempa{[#1][#2]{#3}}% + \super@cite@ +}% +\def\super@cite@{% + \futurelet\@let@token\super@cite@@ +}% +\long\def\super@cite@@#1{% + \@ifx{\@let@token\@sptoken}{% + \aftergroup\@sptoken\false@sw + }{% + \@ifx{\@let@token.}{\true@sw}{% + \@ifx{\@let@token,}{\true@sw}{% + \@ifx{\@let@token'}{\true@sw}{% + \@ifx{\@let@token)}{\true@sw}{% + \false@sw + }% + }% + }% + }% + }% + {% + #1% + \super@cite@ + }{% + \expandafter\@citex@NAT\@tempa\endgroup + #1% + }% +}% +\AtBeginDocument{% + \citeautoscript@sw{% + \@ifx{\place@bibnumber\place@bibnumber@sup}{% + \let\@citex@NAT\@citex\let\@citex\super@cite + }{}% + }{}% +}% +% \end{macrocode} +% \end{macro} +% \end{macro} +% +% +% Resolve an incompatability between \classname{natbib} and \classname{listings}. +% The latter package tests \cmd\chapter (which has now been \cmd\let\ to \cmd\relax\ as a side effect \classname{natbib}'s use of \LaTeX's \cmd\@ifundefined). +% +% We couch our fix in such terms that will not be disruptive if \cmd\chapter\ is actually defined at this point. +% \begin{macrocode} +\@ifx{\chapter\relax}{\let\chapter\@undefined}{}% +% \end{macrocode} +% +% +% \subsubsection{\cs{endnote}s and \cs{rtx@bibnote}s} +% FIXME: check hyperref compatibility! +% +% \begin{macro}{\mini@note} +% \begin{macro}{\save@note} +% \begin{macrocode} +\def\mini@note{\save@note\mini@notes}%Implicit #1 +\def\save@note#1#2{% + \stepcounter{footnote}% + \protected@xdef\@thefnmark{\the\c@footnote}% + \@footnotemark + \expandafter\g@addto@macro + \expandafter#1% + \expandafter{% + \expandafter \@@footnotetext + \expandafter {\@thefnmark}{#2}% + }% +}% +\long\def\@@footnotetext#1{\def\@thefnmark{#1}\@footnotetext}% +\let\mini@notes\@empty +% \end{macrocode} +% \end{macro} +% \end{macro} +% +% \begin{macro}{\endnote} +% A version of footnote that appears at the end of the document. +% Actually it just appears where \cmd\printendnotes\ appears. +% \begin{macrocode} +\def\endnote{% + \begingroup + \aftergroup\@footnotemark + \aftergroup\@endnotetext + \@ifnextchar[{% + \@xendnote + }{% + \stepcounter{footnote}% + \protected@xdef\@tempa{\thefootnote}% + \expandafter\@xendnote\expandafter[\@tempa]% + }% +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\@xendnote} +% \begin{macrocode} +\def\@xendnote[#1]{% + \begingroup + \c@footnote#1\relax + \unrestored@protected@xdef\@thefnmark{endnote\thefootnote}% + \endgroup + \endgroup +}% +\def\@endnotemark{% + \expandafter\cite\expandafter{\@thefnmark}% +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\endnote@ext} +% \begin{macrocode} +\def\endnote@ext{.end}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\@endnotetext} +% The commands \cmd\label, \cmd\index, and \cmd\glossary, +% which are robustified for \cmd\markright\ and \cmd\addcontentsline, +% are likewise robustified here. +% +% Note the similarity to \cmd\protected@write. +% \begin{macrocode} +\long\def\@endnotetext#1{% + \@ifxundefined\@endnoteout{% + \newwrite\@endnoteout + \gdef\endnote@stream{\jobname\endnote@ext}% + \immediate\openout\@endnoteout\endnote@stream\relax + }{}% + \begingroup + \let\label\relax \let\index\relax \let\glossary\relax + \let\cite \relax \let\ref \relax \let\pageref \relax + \let\( \relax \let\) \relax \let\\ \relax +% \let\protect\noexpand + \let \protect \@unexpandable@protect + \newlinechar`\^^M% +% \newlinechar`\ % + \immediate\write\@endnoteout{\string\@doendnote{\@thefnmark}{#1}}% + \endgroup +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\endnotetext} +% \begin{macrocode} +\def\endnotetext{% + \@ifnextchar[{% + \@xendnotenext + }{% + \protected@xdef\@thefnmark{\thefootnote}% + \@endnotetext + }% +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\@xendnotenext} +% \begin{macrocode} +\def\@xendnotenext[#1]{% + \begingroup + \c@footnote#1\relax + \unrestored@protected@xdef\@thefnmark{\thefootnote}% + \endgroup + \@endnotetext +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\printendnotes} +% \begin{macrocode} +\def\printendnotes{% + \@ifx{\@endnoteout\@undefined}{}{% + \begingroup + \c@secnumdepth-\maxdimen + \section{% + \notesname + }% + \immediate\closeout\@endnoteout + \global\let\@endnoteout\@undefined + \begin{enumerate}% + \makeatletter + \input{\endnote@stream}% + \end{enumerate}% + \endgroup + }% +}% +\appdef\class@enddocumenthook{% + \printendnotes +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\firstnote@num} +% \begin{macro}{\set@firstnote} +% +% In \revtex, we have permanently altered the way +% that the footnote counter gets reset at the beginning +% of the document. +% +% If footnotes appear in the bibliography, we initialise +% the footnote counter to number of cites (found last time) +% via \cmd\set@firstnote. +% +% \begin{macrocode} +\mathchardef\firstnote@num\z@ +\appdef\class@documenthook{% + \global\c@footnote\firstnote@num + \appdef\maketitle{\global\c@footnote\firstnote@num}% +}% +\def\set@firstnote#1{% + \@ifnum{\firstnote@num=#1\relax}{}{% + \class@warn@end{Endnote numbers changed: rerun LaTeX}% + }% + \immediate\write\@mainaux{% + \global\chardef\string\firstnote@num#1\relax + }% +}% +% \end{macrocode} +% \end{macro} +% \end{macro} +% +% \begin{macro}{\@endnotesinbib} +% \begin{macro}{\@endnotesinbibliography} +% We define a function \cmd\@endnotesinbib, +% and a variant \cmd\@endnotesinbibliography. +% The former is invoked at the start of +% the end processing for \enve{thebibliography}; +% the latter is a synonym. +% +% The procedure typesets the footnotes that are to +% appear in the bibliography; the default is to +% simply arrange for the footnote counter to be +% reset at the start of the document. +% +% The switchover to setting footnotes in the bibliography +% is done via \cmd\AtBeginDocument\ code that +% changes the meaning of \cmd\footnote\ and +% that substitutes the synonym for +% \cmd\@endnotesinbib. +% +% Note that this code make the assumption that +% the counter used in \env{thebibliography} is \cmd\c@NAT@ctr. +% +% \begin{macrocode} +\def\@endnotesinbib{% + \set@firstnote\z@ +}% +\def\@endnotesinbibliography{% + \expandafter\set@firstnote\expandafter{\the\c@NAT@ctr}% + \@ifx{\@endnoteout\@undefined}{}{% + \immediate\closeout\@endnoteout + \global\let\@endnoteout\@undefined + \makeatletter + \input{\endnote@stream}% + }% +}% +\appdef\class@documenthook{% + \let\@makefnmark@latex\@makefnmark + \let\@footnotemark@latex\@footnotemark + \footinbib@sw{% + \let\footnote\endnote + \let\@footnotemark\@endnotemark + \let\@makefnmark\@makefnmark@cite + \let\@endnotesinbib\@endnotesinbibliography + \appdef\class@enddocumenthook{\auto@bib}% + \let\printendnotes\relax + }{}% +}% +% \end{macrocode} +% \end{macro} +% \end{macro} +% +% \begin{macro}{\auto@bib} +% Under some circumstances, we must typeset the bibliography automatically. +% For instance, the document specifies footnotes in the bibliography, but contains +% no bibliography. +% +% Note that this facility is not able to work more than once per document. +% If multiple bibliographys are required (e.g., per article), it will be the responsibility of the journal style +% to restore \cmd\auto@bib\ to its original meaning so it can be re-invoked. +% \begin{macrocode} +\def\auto@bib{% + \@ifx@empty\@FMN@list{% + \@ifxundefined\@endnoteout{% + \false@sw + }{% + \true@sw + }% + }{% + \true@sw + }{% + \begin{thebibliography}{}\end{thebibliography}% + }% +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\@doendnote} +% \begin{macrocode} +\def\@doendnote#1#2{\bibitem{#1}#2} +% \end{macrocode} +% \end{macro} +% +% +% +% \section{Initial setup} +% \changes{v4.0beta 2}{1999/06/20} +% {Defer assignment until \cs{AtBeginDocument} time.} +% Note that this code is executed at \cmd\AtBeginDocument\ time +% in order to handle a particular timing problem. +% Please do not move this code any earlier! +% \begin{macrocode} +\appdef\class@documenthook{% +% \end{macrocode} +% We assert our own page style only if nobody else has done so. +% Users wishing to customize their documents will be able to invoke a \cmd\pagestyle\ +% command anywhere in the preamble. +% \begin{macrocode} + \@ifxundefined\@evenhead{% + \@ifxundefined\@oddhead{% + \@ifxundefined\@oddfoot{% + \@ifxundefined\@evenfoot{% + \preprintsty@sw{% + \ps@preprint + }{% + \ps@article + }% + }{}% + }{}% + }{}% + }{}% + \pagenumbering{arabic}% + \normalsize + \tabbingsep \labelsep + \leftmargin\leftmargini + \labelwidth\leftmargin\advance\labelwidth-\labelsep + \let\@listi\@listI + \@listi +% \end{macrocode} +% +% \begin{macrocode} +%\twoside@sw{}{% +% \raggedbottom +%}% +% \end{macrocode} +% We make sure that the ``environment'' component mark is set. +% \begin{macrocode} + \def\curr@envir{document}% + \mark@envir{\curr@envir}% +% \end{macrocode} +% +% \changes{v4.0beta 3}{1999/11/13} +% {Grid changes with ltxgrid} +% \changes{v4.0beta 4}{2000/04/10} +% {Also alter how lists get indented.} +% When setting the column grid, we have to alter the procedure +% for formatting lists. Because \cmd\twocolumngrid\ requires +% rebalancing columns at some points, lists must \emph{not} +% involve the use of \cmd\moveleft, \cmd\moveright, or \cmd\parshape. +% The alternative procedure only involves the manipulation of +% \cmd\leftskip\ and \cmd\rightskip. +% +% FIXME: Always sloppy?? +% \begin{macrocode} + \twocolumn@sw{% + \open@column@mlt\tw@ + \set@colht + \@floatplacement + \@dblfloatplacement + \sloppy + \@booleanfalse\raggedcolumn@sw + \let\set@listindent\set@listindent@ + }{% + \open@column@one\@ne + \set@colht + \@floatplacement + \@dblfloatplacement + \sloppy +% \raggedbottom + \let\set@listindent\set@listindent@parshape + }% +}% +% \end{macrocode} +% +% +% \section{\cs{appendix}} +% +% \begin{macrocode} +%\newif\ifappendixon +% \end{macrocode} +% Note that, within appendices, +% equations are numbered within sections (appendices). +% \begin{macrocode} +\def\appendix{% + \par +%\appendixontrue + \setcounter{section}\z@ + \setcounter{subsection}\z@ + \setcounter{subsubsection}\z@ + \def\thesubsection{\arabic{subsection}}% + \def\thesubsubsection{\alph{subsubsection}}% + \@addtoreset{equation}{section}% + \def\theequation@prefix{\thesection}% + \addtocontents{toc}{\protect\appendix}% + \@ifstar{% + \def\thesection{\unskip}% + \def\theequation@prefix{A.}% + }{% + \def\thesection{\Alph{section}}% + }% +}% +% \end{macrocode} +% +% \section{Changing the page grid} +% \changes{v4.0beta 2}{1999/06/20} +% {Define three separate environments, +% defer assignment to \cs{AtBeginDocument} time. +% } +% +% \subsection{Avoiding Grid Changes}% +% +% In preprint styles, ``wide text'' is a no-op, and the title page +% processing involves no grid change. +% +% \begin{macro}{\title@column@default}% +% \begin{macro}{\close@column@default}% +% +% \begin{macrocode} +\def\title@column@default#1{% + \minipagefootnote@init + #1% + \minipagefootnote@foot +}% +\def\close@column@default{% + \newpage +}% +% \end{macrocode} +% \end{macro} +% \end{macro} +% +% +% \subsection{Galley Style: Margin Changes}% +% A variant of preprint processing. Emulate journal appearance somewhat. +% +% \begin{environment}{widetext@galley} +% DPC: We're in galley style so do a lob sided display environment. +% +% QUERY: How can we be sure that we are in galley style? +% ANSWER: as noted elsewhere, require both \cmd\twocolumn@sw\ and \cmd\preprintsty@sw\ to be false. +% \begin{macrocode} +\def\galley@outdent{\rightmargin-\columnwidth\advance\rightmargin-\columnsep}% +\let\widetext@outdent\@empty +\newenvironment{widetext@galley}{% + \list{}{% + \topsep \z@skip + \listparindent \parindent + \itemindent \parindent + \leftmargin \z@ + \parsep \z@\@plus\p@ + \widetext@outdent + \relax + }% + \item\relax +}{ + \endlist +}% +% \end{macrocode} +% \end{environment} +% +% +% \subsection{Grid Changing Via \classname{ltxgrid}}% +% +% In case \classoption{twocolumngrid} has been invoked, +% switch column grid using the column grid-changing commands. +% Supply stub definitions of those commands here. +% \changes{v4.0beta 3}{1999/11/13} +% {grid changes with ltxgrid} +% +% \begin{macro}{\title@column@grid}% +% \begin{macro}{\close@column@grid}% +% The title block always starts at the top of a new page. +% \changes{v4.0beta 3}{1999/11/13} +% {grid changes with ltxgrid} +% +% Note that, for the procedure \cmd\close@column@grid, +% we balance columns by switching to the one-column page grid, +% then we return to a two-column page grid in anticipation of the +% upcoming \cmd\clearpage. +% \begin{macrocode} +\def\title@column@grid#1{% + \minipagefootnote@init + \onecolumngrid + \begingroup + \let\@footnotetext\frontmatter@footnotetext + \let\set@footnotewidth\set@footnotewidth@ii + #1% + \endgroup + \twocolumngrid + \minipagefootnote@foot +}% +\def\close@column@grid{% + \balancelastpage@sw{% + \onecolumngrid +% \twocolumngrid + }{}% +}% +% \end{macrocode} +% \end{macro} +% \end{macro} +% +% \begin{environment}{widetext@grid} +% \changes{v4.0beta 3}{1999/11/13} +% {grid changes via ltxgrid procedures} +% \changes{v4.0beta 5}{2000/11/21} +% {adornments above and below.} +% We slip into the one-column page grid +% within the scope of this environment. +% +% Note that we set adornments above and below the \env{widettext}. +% These are set as leaders, so they will disappear at a page break. +% \begin{macrocode} +\newenvironment{widetext@grid}{% + \par\ignorespaces + \setbox\widetext@top\vbox{% + \vskip15\p@ + \hb@xt@\hsize{% + \leaders\hrule\hfil + \vrule\@height6\p@ + }% + \vskip6\p@ + }% + \setbox\widetext@bot\hb@xt@\hsize{% + \vrule\@depth6\p@ + \leaders\hrule\hfil + }% + \onecolumngrid + \dimen@\ht\widetext@top\advance\dimen@\dp\widetext@top + \cleaders\box\widetext@top\vskip\dimen@ + \let\set@footnotewidth\set@footnotewidth@ii +}{% + \par + \setbox\widetext@bot\vbox{% + \hb@xt@\hsize{\hfil\box\widetext@bot}% + \vskip14\p@ + }% + \dimen@\ht\widetext@bot\advance\dimen@\dp\widetext@bot + \cleaders\box\widetext@bot\vskip\dimen@ + \twocolumngrid\global\@ignoretrue + \@endpetrue +}% +\newbox\widetext@top +\newbox\widetext@bot +% \end{macrocode} +% \end{environment} +% +% +% Decide, finally, how the page grid is to be manipulated. +% \begin{macrocode} +\appdef\class@documenthook{% + \twocolumn@sw{% + \let@environment{widetext}{widetext@grid}% + \let\title@column\title@column@grid + \let\close@column\close@column@grid + }{% + \let@environment{widetext}{widetext@galley}% + \preprintsty@sw{% +% \end{macrocode} +% Change the page grid not at all. +% \begin{macrocode} + }{% +% \end{macrocode} +% If we are galley style, change the page margin only. +% \begin{macrocode} + \galley@sw{% + \let\widetext@outdent\galley@outdent + }{% + }% + }% + }% +}% +% \end{macrocode} +% +% \begin{macro}{\title@column} +% Provide a default meaning for \cmd\title@column\ +% in case it was never defined. +% \begin{macrocode} +\appdef\class@inithook{% + \@ifxundefined\title@column{% + \let\title@column\title@column@default + }{}% +}% +% \end{macrocode} +% \end{macro} +% +% +% \section{Old font commands} +% +% \begin{macrocode} +\DeclareOldFontCommand{\rm}{\normalfont\rmfamily}{\mathrm} +\DeclareOldFontCommand{\sf}{\normalfont\sffamily}{\mathsf} +\DeclareOldFontCommand{\tt}{\normalfont\ttfamily}{\mathtt} +\DeclareOldFontCommand{\bf}{\normalfont\bfseries}{\mathbf} +\DeclareOldFontCommand{\it}{\normalfont\itshape}{\mathit} +\DeclareOldFontCommand{\sl}{\normalfont\slshape}{\@nomath\sl} +\DeclareOldFontCommand{\sc}{\normalfont\scshape}{\@nomath\sc} +\DeclareRobustCommand*\cal{\@fontswitch\relax\mathcal} +\DeclareRobustCommand*\mit{\@fontswitch\relax\mathnormal} +% \end{macrocode} +% +% \section{English-Language Texts} +% As this class is just for English language journals, we +% could hardwire these texts, +% but to make it easier to use this as a +% basis for the code for similar journal styles, separate out all the +% fixed text strings into babel-style macros of the form +% |\|\ldots|name| +% +% Note: for babel compatability, use version 1999/05/05 v3.6x or later. +% +% Some of these might need changing in the society-specific code. +% +% \begin{macro}{\today} +% Procedure \cmd\today\ is used in the article class, but not in +% this document class. +% \begin{macrocode} +\def\today{\ifcase\month\or + January\or February\or March\or April\or May\or June\or + July\or August\or September\or October\or November\or December\fi + \space\number\day, \number\year} +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\notesname} +% Text entity \cmd\notesname\ is used in \cmd\printendnotes. +% \begin{macrocode} +\def\notesname{Notes} +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\partname} +% Text entity \cmd\partname\ is used in \cmd\@part. +% \begin{macrocode} +\def\partname{Part} +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\tocname} +% Text entity \cmd\tocname\ is used in \cmd\tableofcontents, +% as defined in the standard \LaTeX\ book class. +% \begin{macrocode} +\def\tocname{Contents} +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\lofname} +% Text entity \cmd\lofname\ is used in \cmd\listoffigures, +% as defined in the standard \LaTeX\ book class. +% \begin{macrocode} +\def\lofname{List of Figures} +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\lotname} +% Text entity \cmd\lotname\ is used in \cmd\listoftables, +% as defined in the standard \LaTeX\ book class. +% \begin{macrocode} +\def\lotname{List of Tables} +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\refname} +% Text entity \cmd\refname\ is used in \env{thebibliography}. +% \begin{macrocode} +\def\refname{References} +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\indexname} +% Text entity \cmd\indexname\ is used in \env{theindex}, +% as defined in the standard \LaTeX\ book class. +% \begin{macrocode} +\def\indexname{Index} +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\figurename} +% \changes{v4.0beta 2}{1999/06/20} +% {Added localization of \cs{figuresname}} +% Text entity \cmd\figurename\ is used in \env{figure}, +% \cmd\figuresname\ in \cmd\printfigures. +% \begin{macrocode} +\def\figurename{FIG.} +\def\figuresname{Figures}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\tablename} +% \changes{v4.0beta 2}{1999/06/20} +% {Added localization of \cs{tablesname}} +% Text entity \cmd\tablename\ is used in \env{table}, +% \cmd\tablesname\ in \cmd\printtables. +% \begin{macrocode} +\def\tablename{TABLE} +\def\tablesname{Tables}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\abstractname} +% Text entity \cmd\abstractname\ is used in \env{abstract}. +% \begin{macrocode} +\def\abstractname{Abstract} +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\appendixesname} +% \begin{macro}{\appendixname} +% Text entity \cmd\appendixesname\ is used in TOC. +% \begin{macrocode} +\def\appendixesname{Appendixes}% +\def\appendixname{Appendix}% +% \end{macrocode} +% \end{macro} +% \end{macro} +% +% \begin{macro}{\acknowledgmentsname} +% Text entity \cmd\acknowledgmentsname\ is used in \env{acknowledgments}. +% \begin{macrocode} +\def\acknowledgmentsname{Acknowledgments} +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\journalname} +% This should be set by the society journal options, eg `pra'. +% \begin{macrocode} +\def\journalname{??} +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\copyrightname} +% Default layout does not assign copyright, but a journal that wants +% to might use this. +% \begin{macrocode} +\def\copyrightname{??} +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\andname} +% The text string ``and'' for use in author lists. +% \begin{macrocode} +\def\andname{and} +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\@pacs@name} +% \begin{macro}{\@keys@name} +% The text string prepended to PACS numbers, resp. to keywords. +% \begin{macrocode} +\def\@pacs@name{PACS numbers: }% +\def\@keys@name{Keywords: }% +% \end{macrocode} +% \end{macro} +% \end{macro} +% +% \begin{macro}{\ppname} +% The text string ``pp'' for use in page ranges. +% \begin{macrocode} +\def\ppname{pp} +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\numbername} +% The text string ``number'' for use in article reference. +% \begin{macrocode} +\def\numbername{number} +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\volumename} +% The text string ``volume'' for use in article reference. +% \begin{macrocode} +\def\volumename{volume} +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\Dated@name} +% \begin{macro}{\Received@name} +% \begin{macro}{\Revised@name} +% \begin{macro}{\Accepted@name} +% \begin{macro}{\Published@name} +% These texts are used in the \cmd\date, et al. commands. +% \begin{macrocode} +\def\Dated@name{Dated: }% +\def\Received@name{Received }% +\def\Revised@name{Revised }% +\def\Accepted@name{Accepted }% +\def\Published@name{Published }% +% \end{macrocode} +% \end{macro} +% \end{macro} +% \end{macro} +% \end{macro} +% \end{macro} +% +% \section{Legacy Commands} +% We deffine some commands left over from version 3.1, or give default meanings. +% Some definitions can be overridden +% in the document preamble or in included packages. +% +% Note on the namespace: command names like \cmd\REV@\emph{name} are used here, +% because it is not clear that any of this code is generally useful. +% \begin{macrocode} +\def\address{\replace@command\address\affiliation}% +\def\altaddress{\replace@command\altaddress\altaffiliation}% +\newenvironment{references}{% + \class@warn@end{The references environment is not supported; use thebibliography instead.} + \gdef\references{\thebibliography{}}\references +}{% + \endthebibliography +}% +\def\draft{% + \class@warn@end{Command \string\draft\space is obsolete;^^JInvoke option draft instead.}% + \@booleantrue\draft@sw +}% +\def\tighten{% + \class@warn@end{Command \string\tighten\space is obsolete;^^JInvoke option tightenlines instead.}% + \@booleantrue\tightenlines@sw +}% +\def\tableline{% + \noalign{% + \class@warn@end{Command \string\tableline\space is obsolete;^^JUse \string\colrule\space instead.}% + \global\let\tableline\colrule + }% + \tableline +}% +\def\case{\replace@command\case\frac}% +\def\slantfrac{\replace@command\slantfrac\frac}% +\def\tablenote{\replace@command\tablenote\footnote}% +\def\tablenotemark{\replace@command\tablenotemark\footnotemark}% +\def\tablenotetext{\replace@command\tablenotetext\footnotetext}% +% Lose the following definition: +\DeclareRobustCommand\REV@text[1]{% + \relax + \ifmmode + \mathchoice + {\hbox{{\everymath{\displaystyle }#1}}}% + {\hbox{{\everymath{\textstyle }#1}}}% + {\hbox{{\everymath{\scriptstyle }\let\f@size\sf@size\selectfont#1}}}% + {\hbox{{\everymath{\scriptscriptstyle}\let\f@size\ssf@size\selectfont#1}}}% + \glb@settings + \else + \mbox{#1}% + \fi +}% +% Lose the following definition: +\DeclareRobustCommand\REV@bbox[1]{% + \relax + \ifmmode + \mathchoice + {\hbox{{\everymath{\displaystyle }\boldmath$#1$}}}% + {\hbox{{\everymath{\textstyle }\boldmath$#1$}}}% + {\hbox{{\everymath{\scriptstyle }\boldmath$#1$}}}% + {\hbox{{\everymath{\scriptscriptstyle}\boldmath$#1$}}}% + \glb@settings + \else + \mbox{#1}% + \fi +}% +\DeclareRobustCommand\REV@bm[1]{% + \class@warn@end{To use \string\bm, please load the bm package!}% + \global\let\bm\relax +}% +\def\FL{\obsolete@command\FL}% +\def\FR{\obsolete@command\FR}% +\def\narrowtext{\obsolete@command\narrowtext}% +\def\mediumtext{\obsolete@command\mediumtext}% +\newenvironment{quasitable}{% + \let@environment{tabular}{longtable}% +}{% +}% +% \end{macrocode} +% +% \begin{macro}{\text} +% \begin{macro}{\bm} +% \begin{macro}{\bibinfo} +% \begin{macro}{\eprint} +% \begin{macro}{\url} +% +% If not otherwise defined, give default meanings to certain commands. +% \changes{v4.0beta 4}{2000/04/10} +% {eprint takes an aoptional argument, syntactical only in this case.} +% \begin{macrocode} +\appdef\class@inithook{% + \@ifxundefined\text{\let\text\REV@text}{}% + \@ifxundefined\bm{\let\bm\REV@bm}{}% + \providecommand\bibinfo[2]{#2}% + \providecommand\eprint[2][]{#2}% + \providecommand\url[1]{#1}% +}% +% \end{macrocode} +% \end{macro} +% \end{macro} +% \end{macro} +% \end{macro} +% \end{macro} +% +% \begin{macro}{\bbox} +% +% \begin{macrocode} +\def\bbox#1{% + \class@warn@end{\string\bbox\space is obsolete,^^Jload the bm package and use \string\bm\space instead.}% + \global\let\bbox\relax +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\mathletters} +% +% \begin{macrocode} +\newenvironment{mathletters}{% + \class@warn@end{Environment {mathletters} is obsolete;^^Jload the amsmath package and use {subequations}!}% + \global\let\mathletters\@empty +}{% +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\eqnum} +% +% \begin{macrocode} +\def\eqnum#1{% + \class@warn@end{\string\eqnum\space is obsolete, load the amsmath package and use \string\tag!}% + \global\let\eqnum\@gobble +}% +% \end{macrocode} +% \end{macro} +% +% We read in the symbol definitions. +% \begin{macrocode} +\RequirePackage{revsymb}% +\appdef\class@inithook{\revsymb@inithook}% +% \end{macrocode} +% +% \section{Endgame for the Document Class} +% +% This section contains definitions related to the +% end-of-class processing: +% we want to control exactly what happens next. +% +% We provide for a ``job macro package'' that can override +% definitions made by the class or any other packages it loads. +% We act on the value of \cmd\secnumarabic@sw; +% this code is a model for dealing with the job macro package. +% We also install a the very last procedures into +% \cmd\AtEndOfClass, ones that need to tail on to +% any processing performed either by the +% class, by any of its included packages, or +% by the job macro package and any packages it may have loaded. +% +% \subsection{Job Macro Package} +% You can create a ``job macro package'' for your document +% that will be read in automatically every time +% your document is processed. +% Thus, if your job is a file called \file{myarticle.tex}, +% then the file \file{myarticle.rty} will be read in +% just the same as if you had placed a +% \cmd\usepackage|{myarticle.rty}| statement +% immediately following your \cmd\documentclass\ statement. +% +% Within your \file{.rty} file, +% you can define and use control sequence names that use the |@| character +% and you can override any of the definitions or assignments made +% by the \revtex\ document class or the selected journal substyle. +% That is, you have the power to really mess things up badly. +% +% If you choose to have a job macro package, you are well +% advised to read the \LaTeX\ guide to document classes, +% \filename{clsguide.tex} or read up on the subject +% in a book line the \LaTeX\ Companion. +% +% The file \filename{template.rty} contains a template for +% creating your own job macro package. +% +% \begin{macrocode} +\AtEndOfClass{% + \InputIfFileExists{\jobname.rty}{}{}% +}% +% \end{macrocode} +% +% \subsection{\cs{secnumarabic@sw}}\label{sec:secnumarabic} +% \begin{macro}{\secnumarabic@sw} +% The flag \cmd\secnumarabic@sw\ signifies that sectioning commands are +% to be numbered arabic. +% The effect of this command should override any settings +% made by the journal substyle or by the job macro package. +% However, it should be lower in precedence to definitions +% appearing in the document preamble. +% Therefore, we install its code into \cmd\AtEndOfClass, +% but do it \emph{after} the corresponding code +% for the journal substyle and the job macro package. +% +% Note: \cmd\thesection\ must \emph{not} be redefined, +% say, at \cmd\AtBeginDocument\ time by some delayed code. +% \begin{macrocode} +\AtEndOfClass{% + \@ifxundefined\secnumarabic@sw{\@booleanfalse\secnumarabic@sw}{}% + \secnumarabic@sw{% + \def\p@section {}% + \def\thesection {\arabic{section}}% + \def\p@subsection {}% + \def\thesubsection {\thesection.\arabic{subsection}}% + \def\p@subsubsection {}% + \def\thesubsubsection {\thesubsection.\arabic{subsubsection}}% + }{}% +}% +% \end{macrocode} +% \end{macro} +% +% \subsection{Hook for default values of Booleans} +% This portion of the code for this class file +% \emph{must} appear at the very end. +% \begin{macro}{\defaults@hook} +% The procedure \cmd\defaults@hook\ should hereby +% be the very last to be installed into \cmd\AtBeginDocument. +% We provide it with a default in case it has not been +% used up til now. +% +% CHECKME: is any other hook installed later than \cmd\defaults@hook? +% \begin{macrocode} +\AtEndOfClass{% + \appdef\class@documenthook{% + \defaults@hook + }% +}% +\@ifxundefined\defaults@hook{\let\defaults@hook\@empty}{}% +% \end{macrocode} +% \end{macro} +% +% End of the class file. +% \begin{macrocode} +%</class-pst> +% \end{macrocode} +% +% +% \section{Symbols: the \texttt{revsymb} module} +% We immediately define a utility command: this module's warning. +% \begin{macrocode} +%<*revsymb> +\def\REVSYMB@warn#1{\PackageWarningNoLine{revsymb}{#1}}% +% \end{macrocode} +% +% \begin{macro}{\lambdabar} +% \begin{macrocode} +\DeclareRobustCommand\lambdabar{% + \bgroup + \def\@tempa{% + \hbox{% + \raise.73\ht\z@ + \hb@xt@\z@{% + \kern.25\wd\z@ + \vrule \@width.5\wd\z@\@height.1\p@\@depth.1\p@ + \hss + }% + \box\z@ + }% + }% + \mathchoice + {\setbox\z@\hbox{$\displaystyle \lambda$}\@tempa}% + {\setbox\z@\hbox{$\textstyle \lambda$}\@tempa}% + {\setbox\z@\hbox{$\scriptstyle \lambda$}\@tempa}% + {\setbox\z@\hbox{$\scriptscriptstyle\lambda$}\@tempa}% + \egroup +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\openone} +% DPC: Really should use a font that includes this glyph. +% Unfortunately not in AMS ones, but is in bbold, cmbb. +% (I think, must check), +% \changes{v4.d}{1998/01/31} +% {use font-dependent spacing} +% FIXME: check for bbold. +% \begin{macrocode} +\DeclareRobustCommand\openone{\leavevmode\hbox{\small1\normalsize\kern-.33em1}}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\corresponds} +% \begin{macro}{\overdots} +% \begin{macro}{\overcirc} +% J\"org Knappen suggests the replacements: +% replace \cmd\corresponds\ with \cmd\triangleq, source \classname{amssymb}; +% replace \cmd\overcirc\ with \cmd\mathring, source \classname{latex2e}; +% replace \cmd\overdots\ with \cmd\dddot, source \classname{amsmath}. +% +% Any use of any of these commands will result in a warning message at the end +% of the log file. If the corresponding package is not loaded, a +% definition will quietly be provided. +% \begin{macrocode} +\DeclareRobustCommand\corresponds{\replace@command\corresponds\triangleq}% +\DeclareRobustCommand\overcirc{\replace@command\overcirc\mathring}% +\DeclareRobustCommand\overdots{\replace@command\overdots\dddot}% +\DeclareRobustCommand\REV@triangleq{% + {\lower.2ex\hbox{=}}{\kern-.75em^\triangle}% +}% +\DeclareRobustCommand\REV@dddot[1]{% + \@ontopof{#1}{\cdots}{1.0}\mathord{\box2}% +}% +% \end{macrocode} +% \end{macro} +% \end{macro} +% \end{macro} +% +% \begin{macro}{\succsim} +% \begin{macro}{\precsim} +% \begin{macro}{\lesssim} +% \begin{macro}{\gtrsim} +% \begin{macro}{\alt} +% \begin{macro}{\agt} +% These version 3.1 commands are always supplied, +% but the definitions in \classname{amssymb} are preferred. +% \begin{macrocode} +\DeclareRobustCommand\altsuccsim{\succ\kern-.9em_\sim\kern.3em}% +\DeclareRobustCommand\altprecsim{\prec\kern-1em_\sim\kern.3em}% +\let\REV@succsim\altsuccsim +\let\REV@precsim\altprecsim +\DeclareRobustCommand\REV@lesssim{\mathrel{\mathpalette\vereq{<}}}% +\DeclareRobustCommand\REV@gtrsim{\mathrel{\mathpalette\vereq{>}}}% +\DeclareRobustCommand\alt{\lesssim} +\DeclareRobustCommand\agt{\gtrsim} +\def\vereq#1#2{% + \lower3\p@\vbox{% + \baselineskip1.5\p@ + \lineskip1.5\p@ + \ialign{$\m@th#1\hfill##\hfil$\crcr#2\crcr\sim\crcr}% + }% +}% +% \end{macrocode} +% \end{macro} +% \end{macro} +% \end{macro} +% \end{macro} +% \end{macro} +% \end{macro} +% +% \begin{macro}{\tensor} +% \begin{macro}{\overstar} +% \begin{macro}{\loarrow} +% \begin{macro}{\roarrow} +% \changes{v4.0beta 2}{1999/06/20} +% {AO: remove duplicates} +% \begin{macrocode} +\DeclareRobustCommand\tensor[1]{\@ontopof{#1}{\leftrightarrow}{1.15}\mathord{\box2}} +\DeclareRobustCommand\overstar[1]{\@ontopof{#1}{\ast}{1.15}\mathord{\box2}} +\DeclareRobustCommand\loarrow[1]{\@ontopof{#1}{\leftarrow}{1.15}\mathord{\box2}} +\DeclareRobustCommand\roarrow[1]{\@ontopof{#1}{\rightarrow}{1.15}\mathord{\box2}} +% \end{macrocode} +% \end{macro} +% \end{macro} +% \end{macro} +% \end{macro} +% +% \begin{macro}{\@ontopof} +% \begin{macrocode} +\def\@ontopof#1#2#3{% + {% + \mathchoice + {\@@ontopof{#1}{#2}{#3}\displaystyle \scriptstyle }% + {\@@ontopof{#1}{#2}{#3}\textstyle \scriptstyle }% + {\@@ontopof{#1}{#2}{#3}\scriptstyle \scriptscriptstyle}% + {\@@ontopof{#1}{#2}{#3}\scriptscriptstyle\scriptscriptstyle}% + }% +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\@@ontopof} +% Same as \revtex3, more or less. +% \begin{macrocode} +\def\@@ontopof#1#2#3#4#5{% + \setbox\z@\hbox{$#4#1$}% + \setbox\f@ur\hbox{$#5#2$}% + \setbox\tw@\null\ht\tw@\ht\z@ \dp\tw@\dp\z@ + \@ifdim{\wd\z@>\wd\f@ur}{% + \setbox\f@ur\hb@xt@\wd\z@{\hss\box\f@ur\hss}% + \mathord{\rlap{\raise#3\ht\z@\box\f@ur}\box\z@}% + }{% + \setbox\f@ur\hb@xt@.9\wd\f@ur{\hss\box\f@ur\hss}% + \setbox\z@\hb@xt@\wd\f@ur{\hss$#4\relax#1$\hss}% + \mathord{\rlap{\copy\z@}\raise#3\ht\z@\box\f@ur}% + }% +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\frak} +% Deal with legacy \cmd\frak: +% if \classname{amsfonts} not loaded, defined in such a way as to ask for that package. +% Also, says to use \cmd\mathfrak\ instead. +% \begin{macrocode} +\DeclareRobustCommand\frak{% + \REVSYMB@warn{% + Command \string\frak\space unsupported:^^J% + please use \string\mathfrak\space instead.% + }% + \global\let\frak\mathfrak + \frak +}% +\DeclareRobustCommand\REV@mathfrak{% + \REVSYMB@warn{% + Command \string\mathfrak\space undefined:^^J% + please specify the amsfonts or amssymb option!% + }% + \global\let\mathfrak\@firstofone + \mathfrak +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\Bbb} +% Deal with legacy \cmd\Bbb: +% if \classname{amsfonts} not loaded, defined in such a way as to ask for that package. +% Also, says to use \cmd\mathbb\ instead. +% \begin{macrocode} +\DeclareRobustCommand\Bbb{% + \REVSYMB@warn{% + Command \string\Bbb\space unsupported:^^J% + please use \string\mathbb\space instead.% + }% + \global\let\Bbb\mathbb + \Bbb +}% +\DeclareRobustCommand\REV@mathfrak{% + \REVSYMB@warn{% + Command \string\mathbb\space undefined:^^J% + please specify the amsfonts or amssymb option!% + }% + \global\let\mathbb\@firstofone + \mathbb +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\Bigglb} +% Deal with legacy bold delimiters. +% Each of the following takes an implicit argument consisting of +% the delimiter to be made big and bold. +% FIXME: \cmd\DeclareBoldMathCommand\ is not the right tool! +% \begin{macrocode} +\def\Bigglb{\REV@boldopen \Bigg}% +\def\Biglb {\REV@boldopen \Big }% +\def\bigglb{\REV@boldopen \bigg}% +\def\biglb {\REV@boldopen \big }% +\def\Biggrb{\REV@boldclose\Bigg}% +\def\Bigrb {\REV@boldclose\Big }% +\def\biggrb{\REV@boldclose\bigg}% +\def\bigrb {\REV@boldclose\big }% +\def\REV@pmb#1{% + \hbox{% + \setbox\z@=\hbox{#1}% + \kern-.02em\copy\z@\kern-\wd\z@ + \kern .04em\copy\z@\kern-\wd\z@ + \kern-.02em + \raise.04em\copy\z@ + }% +}% +\def\REV@boldopen #1#2{\mathopen {\REV@pmb{$#1#2$}}}% +\def\REV@boldclose#1#2{\mathclose{\REV@pmb{$#1#2$}}}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\AtBeginDocument} +% Package dependencies are taken care of +% \cmd\AtBeginDocument\ time. +% \begin{macrocode} +\def\revsymb@inithook{% + \@ifxundefined\dddot{\let\dddot\REV@dddot}{}% + \@ifxundefined\triangleq{\let\triangleq\REV@triangleq}{}% + \@ifxundefined\succsim{\let\succsim\altsuccsim}{}% + \@ifxundefined\precsim{\let\precsim\altprecsim}{}% + \@ifxundefined\lesssim{\let\lesssim\REV@lesssim}{}% + \@ifxundefined\gtrsim {\let\gtrsim \REV@gtrsim }{}% + \@ifxundefined\mathfrak{\let\mathfrak\REV@mathfrak}{}% + \@ifxundefined\mathbb{\let\mathbb\REV@mathbb}{}% +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macrocode} +%</revsymb> +% \end{macrocode} +% +% \section{The \classoption{10pt} class option: the \texttt{10pt} module} +% \changes{v4.0beta 2}{1999/06/20} +% {AO: all code for \protect\classoption{10pt} is in this module.} +% The file \file{10pt.rtx} is read in by the \classname{revtex4} document class +% if \cmd\@pointsize\ has the value 10. +% +% \begin{macrocode} +%<*10pt> +% \end{macrocode} +% +% \begin{macrocode} +\def\normalsize{% + \@setfontsize\normalsize\@xpt{11.5}% + \abovedisplayskip 10\p@ \@plus2\p@ \@minus5\p@ + \belowdisplayskip \abovedisplayskip + \abovedisplayshortskip \abovedisplayskip + \belowdisplayshortskip \abovedisplayskip + \let\@listi\@listI +}% +% \end{macrocode} +% +% \begin{macrocode} +\def\small{% + \@setfontsize\small\@ixpt{10.5}% + \abovedisplayskip 8.5\p@ \@plus3\p@ \@minus4\p@ + \belowdisplayskip \abovedisplayskip + \abovedisplayshortskip \z@ \@plus2\p@ + \belowdisplayshortskip 4\p@ \@plus2\p@ \@minus2\p@ + \def\@listi{% + \leftmargin\leftmargini + \topsep 4\p@ \@plus2\p@ \@minus2\p@ + \parsep 2\p@ \@plus\p@ \@minus\p@ + \itemsep \parsep + }% +}% +% \end{macrocode} +% +% \begin{macrocode} +\def\footnotesize{% + \@setfontsize\footnotesize\@viiipt{9.5pt}% + \abovedisplayskip 6\p@ \@plus2\p@ \@minus4\p@ + \belowdisplayskip \abovedisplayskip + \abovedisplayshortskip \z@ \@plus\p@ + \belowdisplayshortskip 3\p@ \@plus\p@ \@minus2\p@ + \def\@listi{% + \leftmargin\leftmargini + \topsep 3\p@ \@plus\p@ \@minus\p@ + \parsep 2\p@ \@plus\p@ \@minus\p@ + \itemsep \parsep + }% +}% +% \end{macrocode} +% +% \begin{macrocode} +\def\scriptsize{% + \@setfontsize\scriptsize\@viipt\@viiipt +}% +\def\tiny{% + \@setfontsize\tiny\@vpt\@vipt +}% +\def\large{% + \@setfontsize\large\@xiipt{14pt}% +}% +\def\Large{% + \@setfontsize\Large\@xivpt{18pt}% +}% +\def\LARGE{% + \@setfontsize\LARGE\@xviipt{22pt}% +}% +\def\huge{% + \@setfontsize\huge\@xxpt{25pt}% +}% +\def\Huge{% + \@setfontsize\Huge\@xxvpt{30pt}% +}% +% \end{macrocode} +% +% \begin{macrocode} +%\appdef\class@documenthook{% + \@ifxundefined\twoside@sw{\@booleanfalse\twoside@sw}{}% + \twoside@sw{% +% \oddsidemargin -.1in +% \evensidemargin -.4in + \oddsidemargin -20pt + \evensidemargin -20pt + \marginparwidth 107pt + }{% + \oddsidemargin -.25in + \evensidemargin -.25in + \marginparwidth 30pt + }% +%}% +% \end{macrocode} +% +% \begin{macrocode} +\marginparsep 6pt +% \end{macrocode} +% +% \begin{macrocode} +\topmargin -61pt +% \end{macrocode} +% +% \begin{macrocode} +\headheight 25pt +\headsep 16pt +% \end{macrocode} +% +% \begin{macrocode} +\topskip 10pt +% \end{macrocode} +% +% \begin{macrocode} +\footskip 30pt +% \end{macrocode} +% +% Was: 57\cmd\baselineskip + \cmd\topskip +% \begin{macrocode} +%\appdef\set@pica@hook{% + \textheight = 56pc +%}% +% \end{macrocode} +% +% \begin{macrocode} +\textwidth42.5pc +% \end{macrocode} +% +% \begin{macrocode} +\columnsep 1.5pc +\columnseprule 0pt +% \end{macrocode} +% +% \begin{macrocode} +\footnotesep 1pt +\skip\footins 39pt plus 4pt minus 2pt +\def\footnoterule{\kern-19pt\hrule width.5in\kern18.6pt\relax}% +% \end{macrocode} +% +% \begin{macrocode} +\floatsep 12pt plus 2pt minus 2pt +\textfloatsep 20pt plus 2pt minus 4pt +\intextsep 12pt plus 2pt minus 2pt +% \end{macrocode} +% +% \begin{macrocode} +\dblfloatsep 12pt plus 2pt minus 2pt +\dbltextfloatsep 20pt plus 2pt minus 4pt +% \end{macrocode} +% +% \begin{macrocode} +\@fptop 0pt plus 1fil +\@fpsep 8pt plus 2fil +\@fpbot 0pt plus 1fil +\@dblfptop 0pt plus 1fil +\@dblfpsep 8pt plus 2fil +\@dblfpbot 0pt plus 1fil +% \end{macrocode} +% +% \begin{macrocode} +\marginparpush 5pt +% \end{macrocode} +% +% \begin{macrocode} +\parskip 0pt plus 1pt +\parindent 10pt +\emergencystretch8\p@ +% \end{macrocode} +% +% \begin{macrocode} +\partopsep 2pt plus 1pt minus 1pt +% \end{macrocode} +% +% \begin{macrocode} +\leftmargini 25pt +\leftmarginii 22pt +\leftmarginiii 18.7pt +\leftmarginiv 17pt +\leftmarginv 10pt +\leftmarginvi 10pt +% \end{macrocode} +% +% \begin{macrocode} +\def\@listI{% + \leftmargin\leftmargini + \parsep 4\p@ plus2\p@ minus\p@ + \topsep 8\p@ plus2\p@ minus4\p@ + \itemsep 4\p@ plus2\p@ minus\p@ +}% +% \end{macrocode} +% +% \begin{macrocode} +\labelsep 4pt +% \end{macrocode} +% +% \begin{macrocode} +\def\@listii{% + \leftmargin\leftmarginii + \labelwidth\leftmarginii + \advance\labelwidth-\labelsep + \topsep 4\p@ plus2\p@ minus\p@ + \parsep 2\p@ plus\p@ minus\p@ + \itemsep \parsep +}% +% \end{macrocode} +% +% \begin{macrocode} +\def\@listiii{% + \leftmargin\leftmarginiii + \labelwidth\leftmarginiii + \advance\labelwidth-\labelsep + \topsep 2\p@ plus\p@ minus\p@ + \parsep \z@ + \partopsep \p@ plus\z@ minus\p@ + \itemsep \topsep +}% +% \end{macrocode} +% +% \begin{macrocode} +\def\@listiv{% + \leftmargin\leftmarginiv + \labelwidth\leftmarginiv + \advance\labelwidth-\labelsep +}% +% \end{macrocode} +% +% \begin{macrocode} +\def\@listv{% + \leftmargin\leftmarginv + \labelwidth\leftmarginv + \advance\labelwidth-\labelsep +}% +% \end{macrocode} +% +% \begin{macrocode} +\def\@listvi{% + \leftmargin\leftmarginvi + \labelwidth\leftmarginvi + \advance\labelwidth-\labelsep +}% +% \end{macrocode} +% +% \begin{macrocode} +%</10pt> +% \end{macrocode} +% +% \section{The \classoption{11pt} class option: the \texttt{11pt} module} +% \changes{v4.0beta 2}{1999/06/20} +% {AO: all code for \protect\classoption{11pt} is in this module.} +% The file \file{11pt.rtx} is read in by the \classname{revtex4} document class +% if \cmd\@pointsize\ has the value 11. +% +% \begin{macrocode} +%<*11pt> +% \end{macrocode} +% +% \begin{macrocode} +\def\normalsize{% + \@setfontsize\normalsize\@xipt{13.6}% + \abovedisplayskip 11\p@ \@plus3\p@ \@minus6\p@ + \belowdisplayskip \abovedisplayskip + \abovedisplayshortskip \abovedisplayskip + \belowdisplayshortskip \abovedisplayskip + \let\@listi\@listI +}% +% \end{macrocode} +% +% \begin{macrocode} +\def\small{% + \@setfontsize\small\@xpt\@xiipt + \abovedisplayskip 10\p@ \@plus2\p@ \@minus5\p@ + \abovedisplayshortskip \z@ \@plus3\p@ + \belowdisplayshortskip 6\p@ \@plus3\p@ \@minus3\p@ + \def\@listi{\leftmargin\leftmargini + \topsep 6\p@ \@plus2\p@ \@minus2\p@ + \parsep 3\p@ \@plus2\p@ \@minus\p@ + \itemsep \parsep + }% + \belowdisplayskip \abovedisplayskip +}% +% \end{macrocode} +% +% \begin{macrocode} +\def\footnotesize{% + \@setfontsize\footnotesize\@ixpt{11}% + \abovedisplayskip 8\p@ \@plus2\p@ \@minus4\p@ + \abovedisplayshortskip \z@ \@plus\p@ + \belowdisplayshortskip 4\p@ \@plus2\p@ \@minus2\p@ + \def\@listi{\leftmargin\leftmargini + \topsep 4\p@ \@plus2\p@ \@minus2\p@ + \parsep 2\p@ \@plus\p@ \@minus\p@ + \itemsep \parsep + }% + \belowdisplayskip \abovedisplayskip +}% +% \end{macrocode} +% +% \begin{macrocode} +\def\scriptsize{% + \@setfontsize\scriptsize\@viiipt{9.5}% +}% +\def\tiny{% + \@setfontsize\tiny\@vipt\@viipt +}% +\def\large{% + \@setfontsize\large\@xiipt{14}% +}% +\def\Large{% + \@setfontsize\Large\@xivpt{18}% +}% +\def\LARGE{% + \@setfontsize\LARGE\@xviipt{22}% +}% +\def\huge{% + \@setfontsize\huge\@xxpt{25pt}% +}% +\def\Huge{% + \@setfontsize\Huge\@xxvpt{30pt}% +}% +% \end{macrocode} +% +% \begin{macrocode} +%</11pt> +% \end{macrocode} +% +% \section{The \classoption{12pt} class option: the \texttt{12pt} module} +% \changes{v4.0beta 2}{1999/06/20} +% {AO: all code for \protect\classoption{12pt} is in this module.} +% The file \file{12pt.rtx} is read in by the \classname{revtex4} document class +% if \cmd\@pointsize\ has the value 12. +% +% \begin{macrocode} +%<*12pt> +% \end{macrocode} +% +% \begin{macrocode} +\def\normalsize{% + \@setfontsize\normalsize\@xiipt{14pt}% + \abovedisplayskip 12\p@ \@plus3\p@ \@minus7\p@ + \belowdisplayskip \abovedisplayskip + \abovedisplayshortskip \z@ plus3\p@ + \belowdisplayshortskip 6.5\p@ \@plus3.5\p@ \@minus3\p@ + \let\@listi\@listI +}% +% \end{macrocode} +% +% \begin{macrocode} + \def\small{% + \@setfontsize\small\@xipt{14.5pt}% + \abovedisplayskip 8\p@ \@plus3\p@ \@minus6\p@ + \belowdisplayskip \abovedisplayskip + \abovedisplayshortskip \z@ \@plus3\p@ + \belowdisplayshortskip 6.5\p@ \@plus3.5\p@ \@minus3\p@ + \def\@listi{% + \leftmargin\leftmargini + \topsep 9\p@ \@plus3\p@ \@minus5\p@ + \parsep 4.5\p@ \@plus2\p@ \@minus\p@ + \itemsep \parsep + }% +}% +% \end{macrocode} +% +% Same baselineskip as \cmd\small\ ? +% \begin{macrocode} + \def\footnotesize{% + \@setfontsize\footnotesize\@xpt{14.5pt}% + \abovedisplayskip 10\p@ \@plus2\p@ \@minus5\p@ + \belowdisplayskip \abovedisplayskip + \abovedisplayshortskip \z@ \@plus3\p@ + \belowdisplayshortskip 6\p@ \@plus3\p@ \@minus3\p@ + \def\@listi{% + \leftmargin\leftmargini + \topsep 6\p@ \@plus2\p@ \@minus2\p@ + \parsep 3\p@ \@plus2\p@ \@minus\p@ + \itemsep \parsep + }% +}% +% \end{macrocode} +% +% \begin{macrocode} +\def\scriptsize{% + \@setfontsize\scriptsize\@viiipt{9.5pt}% +}% +\def\tiny{% + \@setfontsize\tiny\@vipt{7pt}% +}% +\def\large{% + \@setfontsize\large\@xivpt{18pt}% +}% +\def\Large{% + \@setfontsize\Large\@xviipt{22pt}% +}% +\def\LARGE{% + \@setfontsize\LARGE\@xxpt{25pt}% +}% +\def\huge{% + \@setfontsize\huge\@xxvpt{30pt}% +}% +\let\Huge=\huge +% \end{macrocode} +% +% \begin{macrocode} +%</12pt> +% \end{macrocode} +% +% \section{Page parameters}% +% This code is common to both \classoption{11pt} and \classoption{12pt}. +% +% \begin{macrocode} +%<*11pt|12pt> +% \end{macrocode} +% +% \begin{macrocode} +%\appdef\class@documenthook{% + \@ifxundefined\twoside@sw{\@booleanfalse\twoside@sw}{}% + \twoside@sw{% + \oddsidemargin 0pt + \evensidemargin 0pt + \marginparwidth 60pt + }{% + \oddsidemargin 0pt + \evensidemargin 0pt + \marginparwidth 44pt + }% +%}% +% \end{macrocode} +% +% \begin{macrocode} +\marginparsep 10pt +% \end{macrocode} +% +% \begin{macrocode} +\topmargin -37pt +% \end{macrocode} +% +% \begin{macrocode} +\headheight 12pt +\headsep 25pt +% \end{macrocode} +% +% \begin{macrocode} +\topskip 10pt +% \end{macrocode} +% +% \begin{macrocode} +\footskip 30pt +% \end{macrocode} +% +% \begin{macrocode} +%\appdef\set@pica@hook{% + \@ifxundefined\tightenlines@sw{\@booleanfalse\tightenlines@sw}{}% + \textheight=665.5\p@ + \tightenlines@sw{% + \def\baselinestretch{1}% + }{% + \def\baselinestretch{1.5}% + }% +%}% +% \end{macrocode} +% +% \begin{macrocode} +\textwidth 468pt +% \end{macrocode} +% +% \begin{macrocode} +\columnsep 10pt +\columnseprule 0pt +% \end{macrocode} +% +% \begin{macrocode} +\footnotesep 1pt +\skip\footins 25.25pt plus 4pt minus 2pt +\def\footnoterule{\kern-5.25pt\hrule width.5in\kern4.85pt\relax} +% \end{macrocode} +% +% \begin{macrocode} +\floatsep 14pt plus 2pt minus 4pt +\textfloatsep 20pt plus 2pt minus 4pt +\intextsep 14pt plus 4pt minus 4pt +% \end{macrocode} +% +% \begin{macrocode} +\dblfloatsep 14pt plus 2pt minus 4pt +\dbltextfloatsep 20pt plus 2pt minus 4pt +% \end{macrocode} +% +% \begin{macrocode} +\@fptop 0pt plus 1fil +\@fpsep 10pt plus 2fil +\@fpbot 0pt plus 1fil +\@dblfptop 0pt plus 1fil +\@dblfpsep 10pt plus 2fil% +\@dblfpbot 0pt plus 1fil +% \end{macrocode} +% +% \begin{macrocode} +\marginparpush 7pt +% \end{macrocode} +% +% \begin{macrocode} +\parskip 0pt plus 1pt +\parindent 15pt +\emergencystretch8\p@ +\partopsep 3pt plus 2pt minus 2pt +% \end{macrocode} +% +% \begin{macrocode} +\leftmargini 30pt +\leftmarginii 26pt +\leftmarginiii 22pt +\leftmarginiv 20pt +\leftmarginv 12pt +\leftmarginvi 12pt +% \end{macrocode} +% +% \begin{macrocode} +\def\@listI{\leftmargin\leftmargini \parsep 5\p@ plus2.5\p@ minus\p@ + \topsep 10\p@ plus4\p@ minus6\p@ + \itemsep 5\p@ plus2.5\p@ minus\p@ +}% +% \end{macrocode} +% +% \begin{macrocode} +\labelsep 6pt +% \end{macrocode} +% +% \begin{macrocode} +\def\@listii{\leftmargin\leftmarginii + \labelwidth\leftmarginii\advance\labelwidth-\labelsep + \topsep 5\p@ plus2.5\p@ minus\p@ + \parsep 2.5\p@ plus\p@ minus\p@ + \itemsep \parsep +}% +% \end{macrocode} +% +% \begin{macrocode} +\def\@listiii{\leftmargin\leftmarginiii + \labelwidth\leftmarginiii\advance\labelwidth-\labelsep + \topsep 2.5\p@ plus\p@ minus\p@ + \parsep \z@ \partopsep \p@ plus\z@ minus\p@ + \itemsep \topsep +}% +% \end{macrocode} +% +% \begin{macrocode} +\def\@listiv{\leftmargin\leftmarginiv + \labelwidth\leftmarginiv\advance\labelwidth-\labelsep +}% +% \end{macrocode} +% +% \begin{macrocode} +\def\@listv{\leftmargin\leftmarginv + \labelwidth\leftmarginv\advance\labelwidth-\labelsep +}% +% \end{macrocode} +% +% \begin{macrocode} +\def\@listvi{\leftmargin\leftmarginvi + \labelwidth\leftmarginvi\advance\labelwidth-\labelsep +}% +% \end{macrocode} +% +% \begin{macrocode} +%</11pt|12pt> +% \end{macrocode} +% +% \section{The \classoption{aps} class option: the \texttt{aps} module} +% \changes{v4.0beta 2}{1999/06/20} +% {AO: made aps.rtx part of revtex4.dtx} +% The file \file{aps.rtx} is read in by the \classname{revtex4} document class +% if \cmd\@society\ has the value \classoption{aps}. +% +% Here, code specific to APS journals is separated out from the \revtex\ document class. +% (Other societies can customize \revtex\ by supplying their own \file{.rtx} file.) +% +% We first give some text entities (amounting to journal abbreviations), +% then some APS-specific initialisations, +% then code for particular APS journals. +% In the latter case, the choice is +% keyed off the macro \cmd\@journal. +% +% Note on \cmd\AtEndOfClass: +% this file, like all journal substyles, is read in at \cmd\AtEndOfClass\ time, +% so you should not use this command in this file. +% Mote for \LaTeX\ developers: It would be an improvement in \LaTeX\ to +% \cmd\let\cmd\AtEndOfClass\ to something like \cmd\@firstofone. +% This change would be effected in \cmd\@onefilewithoptions. +% +% \begin{macrocode} +%<*aps> +% \end{macrocode} +% +% \subsection{Defend Against Forseeable Errors}% +% Protect this file from being read in by anything but \revtex. +% \begin{macrocode} +\ifx\undefined\substyle@ext + \def\@tempa{% + \endinput + \GenericWarning{I must be read in by REVTeX! (Bailing out)}% + }% + \expandafter\else + \def\@tempa{}% + \expandafter\fi\@tempa +% \end{macrocode} +% +% \subsection{Abbreviations}% +% The following macros constitute typing shortcuts for +% certain journal names. +% \begin{macrocode} +\def\ao{Appl.\ Opt.\ } +\def\ap{Appl.\ Phys.\ } +\def\apl{Appl.\ Phys.\ Lett.\ } +\def\apj{Astrophys.\ J.\ } +\def\bell{Bell Syst.\ Tech.\ J.\ } +\def\jqe{IEEE J.\ Quantum Electron.\ } +\def\assp{IEEE Trans.\ Acoust.\ Speech Signal Process.\ } +\def\aprop{IEEE Trans.\ Antennas Propag.\ } +\def\mtt{IEEE Trans.\ Microwave Theory Tech.\ } +\def\iovs{Invest.\ Ophthalmol.\ Vis.\ Sci.\ } +\def\jcp{J.\ Chem.\ Phys.\ } +\def\jmo{J.\ Mod.\ Opt.\ } +\def\josa{J.\ Opt.\ Soc.\ Am.\ } +\def\josaa{J.\ Opt.\ Soc.\ Am.\ A } +\def\josab{J.\ Opt.\ Soc.\ Am.\ B } +\def\jpp{J.\ Phys.\ (Paris) } +\def\nat{Nature (London) } +\def\oc{Opt.\ Commun.\ } +\def\ol{Opt.\ Lett.\ } +\def\pl{Phys.\ Lett.\ } +\def\pra{Phys.\ Rev.\ A } +\def\prb{Phys.\ Rev.\ B } +\def\prc{Phys.\ Rev.\ C } +\def\prd{Phys.\ Rev.\ D } +\def\pre{Phys.\ Rev.\ E } +\def\prl{Phys.\ Rev.\ Lett.\ } +\def\rmp{Rev.\ Mod.\ Phys.\ } +\def\pspie{Proc.\ Soc.\ Photo-Opt.\ Instrum.\ Eng.\ } +\def\sjqe{Sov.\ J.\ Quantum Electron.\ } +\def\vr{Vision Res.\ } +% \end{macrocode} +% +% \subsection{APS Setup}% +% Here we define the default procedures for APS journals. +% Individual APS journals may override these definitions. +% +% \subsubsection{Title block}% +% +% The specifics of the title block. +% Apply to all APS journals; individual journals may override these settings. +% +% \begin{macro}{\@fnsymbol} +% The \LaTeX\ kernel definition of \cmd\@fnsymbol\ is overridden. +% \begin{macrocode} +\def\@fnsymbol#1{% + \ensuremath{% + \ifcase#1\or + *\or \dagger \or + \ddagger \or \mathsection \or + \mathparagraph\or + **\or \dagger\dagger \or + \ddagger\ddagger \or \mathsection \mathsection \or + \mathparagraph\mathparagraph\or + *{*}*\ignorespaces \or \dagger\dagger\dagger \or + \ddagger\ddagger\ddagger\or \mathsection \mathsection \mathsection \or + \mathparagraph\mathparagraph\mathparagraph\or + \else + \@ctrerr + \fi + }% +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\titlepage} +% \begin{macrocode} +\renewenvironment{titlepage}{% + \let\wastwocol@sw\twocolumn@sw + \onecolumngrid + \newpage + \thispagestyle{titlepage}% + \c@page\z@% article sets this to one not zero??? +}{% + \wastwocol@sw{\twocolumngrid}{\newpage}% +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\frontmatter@abstractheading} +% APS Journals all set the abstract head the same way, with no head. +% However, if the user has specified the \classoption{preprint} class option, +% then the abstract will have a head. +% \begin{macrocode} +\def\frontmatter@abstractheading{% + \preprintsty@sw{% + \begingroup + \centering\large + \abstractname + \par + \endgroup + }{}% +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\frontmatter@abstractwidth} +% All APS journals set the abstract to the same width. +% \begin{macrocode} +\def\frontmatter@abstractwidth{400\p@}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\frontmatter@abstractfont} +% All APS journals set the abstract body the same way. +% \changes{v4.0beta 3}{1999/11/13} +% {Bug 123: add parskip to the abstract.} +% \begin{macrocode} +\def\frontmatter@abstractfont{% + \small + \parindent1em\relax + \adjust@abstractwidth +}% +\def\adjust@abstractwidth{% + \dimen@\textwidth\advance\dimen@-\frontmatter@abstractwidth + \divide\dimen@\tw@ + \galley@sw{% + \advance\rightskip\tw@\dimen@ + }{% + \advance\leftskip\dimen@ + \advance\rightskip\dimen@ + }% +}% +% \end{macrocode} +% \end{macro} +% +% All APS journal preprints use separate titlepage and full-width abstract. +% +% Note that we defer polling the \cmd\preprintsty@sw\ Boolean until +% \cmd\AtBeginDocument\ time, because the document preamble +% may legitimately change the state of that variable. +% \begin{macrocode} +\appdef\class@documenthook{% + \preprintsty@sw{% + \@ifxundefined\titlepage@sw{\@booleantrue\titlepage@sw}{}% +% \tightenlines@sw{% +% \def\baselinestretch{1}% +% }{% +% \def\baselinestretch{1.5}% +% }% + \def\frontmatter@abstractwidth{\textwidth}% + \def\frontmatter@affiliationfont{\it}% + \let\section\section@preprintsty +% \let\@hangfrom@section\@hangfrom@section@preprintsty + \let\subsection\subsection@preprintsty + \let\subsubsection\subsubsection@preprintsty + }{}% + }% +% \end{macrocode} +% +% \begin{macro}{\frontmatter@authorformat} +% All APS journals set the author list the same. +% The leading is 11.5 points, and +% there is 11.5 points of extra space above the first author line +% (which amounts to the same thing as 11.5 points extra below the title) +% for a total of 23 points base-to-base. +% +% \begin{macrocode} +\def\frontmatter@authorformat{% + \skip@\@flushglue + \@flushglue\z@ plus.3\hsize\relax + \centering + \advance\baselineskip\p@ + \parskip11.5\p@\relax + \@flushglue\skip@ +%\preprintsty@sw{}{% +% \addvspace{0\p@}% +%}% +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\frontmatter@above@affilgroup} +% The default amount of space above affiliation. +% APS Journals have 24 points b-b above an affiliation group. +% \begin{macrocode} +\def\frontmatter@above@affilgroup{% +%\preprintsty@sw{}{% +% \addvspace{11\p@}% +%}% +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\frontmatter@above@affiliation} +% \begin{macro}{\frontmatter@above@affiliation@script} +% The default amount of space above affiliation. +% APS Journals have no extra space between author group down to common affiliation. +% \begin{macrocode} +\def\frontmatter@above@affiliation@script{% + \skip@\@flushglue + \@flushglue\z@ plus.3\hsize\relax + \centering + \@flushglue\skip@ + \addvspace{3.5\p@}% +}% +\def\frontmatter@above@affiliation{% + \preprintsty@sw{}{% +% \addvspace{12\p@}% + }% +}% +% \end{macrocode} +% \end{macro} +% \end{macro} +% +% \begin{macro}{\frontmatter@affiliationfont} +% All APS journals set the affiliation the same. +% \begin{macrocode} +\def\frontmatter@affiliationfont{% + \small\it +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\frontmatter@collaboration@above} +% PRL: 1.5 points extra: 13 points base-to-base above. +% \begin{macrocode} +\def\frontmatter@collaboration@above{% + \preprintsty@sw{% + }{% + \parskip1.5\p@\relax + }% +}% +% \end{macrocode} +% \end{macro} +% +% Set up the default APS style for title block authors and affiliations. +% \begin{macrocode} +\@ifxundefined\groupauthors@sw{% + \clo@groupedaddress +}{}% +% \end{macrocode} +% +% \begin{macro}{\frontmatter@setup} +% All APS journals set the title page using the same font and size. +% However, justification varies for the title block elements, so +% we assert none here. +% \begin{macrocode} +\def\frontmatter@setup{% + \normalfont +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\frontmatter@title@above} +% \begin{macro}{\frontmatter@title@format} +% \begin{macro}{\frontmatter@title@below} +% All APS journals set the article title the same. +% +% Note: Spacing from title to author is 23 points base-to-base. +% \begin{macrocode} +\def\frontmatter@title@above{\addvspace{6\p@}}% +\def\frontmatter@title@format{\large\bfseries\centering\parskip\z@skip}% +\def\frontmatter@title@below{}% +% \end{macrocode} +% \end{macro} +% \end{macro} +% \end{macro} +% +% \begin{macro}{\frontmatter@makefnmark} +% All APS journals share this procedure for setting the titlepage footnote text. +% \begin{macrocode} +\def\@author@parskip{3\p@}% +\def\frontmatter@makefnmark{% + \@textsuperscript{% + \normalfont\@thefnmark + }% +}% +\def\frontmatter@authorbelow{% + \addvspace{3\p@}% +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\frontmatter@RRAP@format} +% All APS journals use the same format for the ``Received, Revised, etc.'' block on the title page. +% +% Change note: 11.5 points b-b from author/affiliation down to date. +% \begin{macrocode} +\def\frontmatter@RRAP@format{% + \small + \centering +% \preprintsty@sw{}{\parskip.5ex\relax}% + \everypar{\hbox\bgroup(\@gobble@leavemode@uppercase}% + \def\par{\@ifvmode{}{\unskip)\egroup\@@par}}% +}% +\def\punct@RRAP{;\egroup\ \hbox\bgroup}% +\def\@gobble@leavemode@uppercase#1#2{\expandafter\MakeTextUppercase}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\frontmatter@PACS@format} +% +% \begin{macrocode} +\def\frontmatter@PACS@format{% + \addvspace{11\p@}% + \footnotesize + \adjust@abstractwidth + \parindent\z@ + \parskip\z@skip + \samepage +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\frontmatter@keys@format} +% +% \begin{macrocode} +\def\frontmatter@keys@format{% + \footnotesize + \adjust@abstractwidth + \parindent\z@ + \samepage +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\ps@titlepage} +% \changes{v4.g}{1998/06/10} +% {multiple preprint commands} +% \changes{v4.0beta 2}{1999/06/20} +% {Stack \cs{preprint} args flush right at right margin.} +% Title page style. Currently empty except for preprint header, +% which consists of all the \cmd\preprint\ arguments, +% stacked flush right at the right margin. +% \changes{v4.0beta 3}{1999/11/13} +% {Insert procedure \cs{checkindate}} +% \begin{macrocode} +\def\ps@titlepage{% + \def\@oddhead{% + \hfill + \produce@preprints\@preprint + }% + \let\@evenhead\@oddhead + \def\@oddfoot{% + \hb@xt@\z@{\byrevtex\hss}% + \hfil + \preprintsty@sw{\thepage}{}% + \quad\checkindate + \hfil + }% + \let\@evenfoot\@oddfoot +}% +\def\byrevtex{\byrevtex@sw{Typeset by REV\TeX}{}}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\produce@preprints} +% \changes{v4.0beta 3}{1999/11/13} +% {Bug 115. If three or more preprints specified, set on single line, with commas.} +% \begin{macrocode} +\def\produce@preprints#1{% + \preprint@sw{% + \vtop to \z@{% + \def\baselinestretch{1}% + \small + \let\preprint\preprint@count + \count@\z@#1\@ifnum{\count@>\tw@}{% + \hbox{% + \let\preprint\preprint@hlist + #1\setbox\z@\lastbox + }% + }{% + \let\preprint\preprint@cr + \halign{\hfil##\cr#1\crcr}% + \par + \vss + }% + }% + }{}% +}% +\def\preprint@cr#1{#1\cr}% +\def\preprint@count#1{\advance\count@\@ne}% +\def\preprint@hlist#1{#1\hbox{, }}% +% \end{macrocode} +% \end{macro} +% +% \subsubsection{Stacked Heads}% +% All APS journals put a period (.), followed by quad space, after the section number. +% Also, no hanging section number. +% \begin{macrocode} +\def\@seccntformat#1{\csname the#1\endcsname.\quad}% +\def\@hang@from#1#2#3{#1#2#3}% +% \end{macrocode} +% +% Note that in the following, we wish to set the section head uppercase, so we use +% David Carlisle's \cmd\MakeTextUppercase. However, because this procedure effectively +% parses its argument (looking for things to \emph{not} translate), it has to be invoked +% in such a way that the argument of the \cmd\section\ command is passed to it as its +% own argument. +% +% To accomplish this, we use the \cmd\@hangfrom@\ hook, which was developed for this +% purpose. +% +% \begin{macrocode} +\def\section{% + \@startsection + {section}% + {1}% + {\z@}% + {0.8cm \@plus1ex \@minus .2ex}% + {0.5cm}% + {% + \normalfont\small\bfseries + \centering + }% +}% +\def\@hangfrom@section#1#2#3{\@hangfrom{#1#2}\MakeTextUppercase{#3}}% +% \end{macrocode} +% +% \begin{macrocode} +\def\subsection{% + \@startsection + {subsection}% + {2}% + {\z@}% + {.8cm \@plus1ex \@minus .2ex}% + {.5cm}% + {% + \normalfont\small\bfseries + \centering + }% +}% +% \end{macrocode} +% +% \begin{macrocode} +\def\subsubsection{% + \@startsection + {subsubsection}% + {3}% + {\z@}% + {.8cm \@plus1ex \@minus .2ex}% + {.5cm}% + {% + \normalfont\small\itshape + \centering + }% +}% +% \end{macrocode} +% +% \subsubsection{Runin Heads}% +% \begin{macrocode} +\def\paragraph{% + \@startsection + {paragraph}% + {4}% + {\parindent}% + {\z@}% + {-1em}% + {\normalfont\normalsize\itshape}% +}% +% \end{macrocode} +% +% \begin{macrocode} +\def\subparagraph{% + \@startsection + {subparagraph}% + {5}% + {\parindent}% + {3.25ex \@plus1ex \@minus .2ex}% + {-1em}% + {\normalfont\normalsize\bfseries}% +}% +% \end{macrocode} +% +% +% +% \begin{macro}{\section@preprintsty} +% \begin{macro}{\subsection@preprintsty} +% \begin{macro}{\subsubsection@preprintsty} +% Here are the formatting procedures specific to the preprint style; +% the only difference is that the heads are flush left instead of centered. +% +% \begin{macrocode} +\def\section@preprintsty{% + \@startsection + {section}% + {1}% + {\z@}% + {0.8cm \@plus1ex \@minus .2ex}% + {0.5cm}% + {% + \normalfont\small\bfseries +% \centering + }% +}% +%\def\@hangfrom@section@preprintsty#1#2#3{\@hangfrom{#1#2}\MakeTextUppercase{#3}}% +% \end{macrocode} +% \begin{macrocode} +\def\subsection@preprintsty{% + \@startsection + {subsection}% + {2}% + {\z@}% + {.8cm \@plus1ex \@minus .2ex}% + {.5cm}% + {% + \normalfont\small\bfseries +% \centering + }% +}% +% \end{macrocode} +% \begin{macrocode} +\def\subsubsection@preprintsty{% + \@startsection + {subsubsection}% + {3}% + {\z@}% + {.8cm \@plus1ex \@minus .2ex}% + {.5cm}% + {% + \normalfont\small\itshape +% \centering + }% +}% +% \end{macrocode} +% \end{macro} +% \end{macro} +% \end{macro} +% +% Effectively invoke the \classoption{bibnote} class option, +% if \cmd\frontmatter@footnote@produce\ has not yet been set. +% \begin{macrocode} + \@ifxundefined\frontmatter@footnote@produce{% + \let\frontmatter@footnote@produce\frontmatter@footnote@produce@endnote + }{}% +% \end{macrocode} +% +% +% \subsubsection{Table of Contents}% +% The toc will itself make an entry in the toc, +% but we temporarily turn off toc formatting for the duration. +% \begin{macrocode} +\def\@pnumwidth{1.55em} +\def\@tocrmarg {2.55em} +\def\@dotsep{4.5pt} +\setcounter{tocdepth}{3} +% \end{macrocode} +% +% \begin{macrocode} +\def\tableofcontents{% + \addtocontents{toc}{\string\tocdepth@munge}% + \print@toc{toc}% + \addtocontents{toc}{\string\tocdepth@restore}% +}% +% \end{macrocode} +% +% \begin{macrocode} +\def\tocdepth@munge{% + \let\l@section@saved\l@section + \let\l@section\@gobble@tw@ +}% +\def\@gobble@tw@#1#2{}% +% \end{macrocode} +% +% \begin{macrocode} +\def\tocdepth@restore{% + \let\l@section\l@section@saved +}% +% \end{macrocode} +% +% \begin{macrocode} +\def\l@part#1#2{\addpenalty{\@secpenalty}% + \begingroup + \set@tocdim@pagenum{#2}% +% \@tempdima 3em % + \parindent \z@ + \rightskip\tocleft@pagenum plus 1fil\relax + \skip@\parfillskip\parfillskip\z@ + \addvspace{2.25em plus\p@}% + \large \bf % + \leavevmode\ignorespaces#1\unskip\nobreak\hskip\skip@ + \hb@xt@\rightskip{\hfil\unhbox\z@}\hskip-\rightskip\hskip\z@skip + \par + \nobreak % + \endgroup +}% +% \end{macrocode} +% +% \begin{macro}{\l@section} +% Determine which TOC elements are automatically indented. +% \begin{macrocode} +\def\tocleft@{\z@}% +\def\tocdim@min{5\p@}% +\def\l@section{% + \l@@sections{}{section}% Implicit #3#4 +}% +\def\l@f@section{% + \addpenalty{\@secpenalty}% + \addvspace{1.0em plus\p@}% + \bf +}% +\def\l@subsection{% + \l@@sections{section}{subsection}% Implicit #3#4 +}% +\def\l@subsubsection{% + \l@@sections{subsection}{subsubsection}% Implicit #3#4 +}% +\def\l@paragraph#1#2{}% +\def\l@subparagraph#1#2{}% +% \end{macrocode} +% \end{macro} +% +% Activate the auto TOC processing. +% \begin{macrocode} +\let\toc@pre\toc@pre@auto +\let\toc@post\toc@post@auto +% \end{macrocode} +% +% \begin{macrocode} +\def\listoffigures{\print@toc{lof}}% +\def\l@figure{\@dottedtocline{1}{1.5em}{2.3em}} +% \end{macrocode} +% +% \begin{macrocode} +\def\listoftables{\print@toc{lot}}% +\let\l@table\l@figure +% \end{macrocode} +% +% \subsubsection{Default column bottom}% +% All APS journal styles have flush bottoms. +% +% Note that we defer polling the \cmd\raggedcolumn@sw\ Boolean until +% \cmd\AtBeginDocument\ time, because the document preamble +% may legitimately change the state of that variable. +% \begin{macrocode} +\appdef\class@documenthook{% + \@ifxundefined\raggedcolumn@sw{\@booleantrue\raggedcolumn@sw}{}% + \raggedcolumn@sw{\raggedbottom}{\flushbottom}% +}% +% \end{macrocode} +% +% +% \subsubsection{Table alignment style}% +% +% \begin{macro}{\tableft@skip@float} +% \begin{macro}{\tabmid@skip@float} +% \begin{macro}{\tabright@skip@float} +% \begin{macro}{\array@row@pre@float} +% \begin{macro}{\array@row@pst@float} +% All APS publications have the same table specification: +% Scotch rules above and below, centered in column. +% \begin{macrocode} +\def\tableft@skip@float{\z@ plus\hsize}% +\def\tabmid@skip@float{\@flushglue}% +\def\tabright@skip@float{\z@ plus\hsize}% +\def\array@row@pre@float{\hline\hline\noalign{\vskip\doublerulesep}}% +\def\array@row@pst@float{\noalign{\vskip\doublerulesep}\hline\hline}% +% \end{macrocode} +% \end{macro} +% \end{macro} +% \end{macro} +% \end{macro} +% \end{macro} +% +% +% \subsubsection{Footnote formatting}% +% We customize the formatting of footnotes for all APS journals. +% \begin{macro}{\@makefntext} +% \begin{macrocode} +\def\@makefntext#1{% + \def\baselinestretch{1}% + \reset@font + \footnotesize + \leftskip1em + \parindent1em + \noindent\nobreak\hskip-\leftskip + \hb@xt@\leftskip{% + \Hy@raisedlink{\hyper@anchorstart{footnote@\the\c@footnote}\hyper@anchorend}% + \hss\@makefnmark\ + }% + #1% + \par +}% +% \end{macrocode} +% \end{macro} +% +% +% \subsubsection{Appendix}% +% \begin{macro}{\appendix} +% \begin{macro}{\@hangfrom@appendix} +% \begin{macro}{\@hangfroms@appendix} +% \begin{macro}{\@appendixcntformat} +% \changes{v4.0beta 3}{1999/11/13} +% {Bug 129: section* within appendix was producing appendixname} +% \begin{macrocode} +\prepdef\appendix{% + \par + \let\@hangfrom@section\@hangfrom@appendix + \let\@hangfroms@section\@hangfroms@appendix + \let\@sectioncntformat\@appendixcntformat +}% +\def\@hangfrom@appendix#1#2#3{% + #1% + \@if@empty{#2}{% + #3% + }{% + #2\@if@empty{#3}{}{:\ #3}% + }% +}% +\def\@hangfroms@appendix#1#2{% + #1#2% +}% +\def\@appendixcntformat#1{\appendixname\ \csname the#1\endcsname}% +% \end{macrocode} +% \end{macro} +% \end{macro} +% \end{macro} +% \end{macro} +% +% \subsubsection{Bibliography}% +% +% Customize \revtex\ for the journal substyle; +% this task requires three components: +% the \BibTeX\ \file{apsrev.bst} style file, +% customizing code for \file{natbib}, and +% customizations of the \env{thebibliography} environment. +% +% \begin{macro}{\@bibstyle} +% Define the argument of the \cmd\bibliographystyle\ command +% (if the document does not do so). +% The user must have installed a \file{.bst} file of the corresponding name. +% This file will then be used by \BibTeX\ when compiling the document's \file{.bbl} file. +% +% The default bibliography style for the APS journal substyles, \file{apsrev.bst}, +% was originally called \file{revtex.bst}, but +% that name is much too generic and should be given to a \file{.bst} file that +% that applies generally to any \revtex\ jounal substyle, an impossibility. +% +% To generate \file{apsrev.bst}, use \classname{custom-bib} version 4.0b2 or later. +% Run the \file{.bst} generator, \file{makebst.tex}, and +% accept all defaults, with the following exceptions: +% +% \begin{enumerate} +% \item +% ORDERING OF REFERENCES: c: seq-no (references in order of Citation); +% \item +% AUTHOR NAMES: i: nm-init,ed-au (Initials + surname); +% \item +% NUMBER OF AUTHORS: l: max 12, min 12 (there will be three prompts total); +% \item +% TYPEFACE FOR AUTHORS IN LIST OF REFERENCES: u: nmft,nmft-def (User defined author font); +% \item +% FONT FOR FIRST NAMES: u: fnm-def (First names in user defined font); +% \item +% EDITOR NAMES IN INCOLLECTION ETC: a: nmfted (Editors incollection like authors); +% \item +% DATE FORMAT: p: yr-par (Date in parentheses); +% \item +% SUPPRESS MONTH: x: xmth (Date is year only); +% \item +% TITLE OF ARTICLE: i: tit-it (Title italic) +% \item +% ARTICLE TITLE PRESENT: x: jtit-x (No article title); +% \item +% JOURNAL NAME FONT: r: jttl-rm (Journal title normal); +% \item +% TECHNICAL REPORT TITLE: b: trtit-b (Tech. report title like books); +% \item +% JOURNAL VOLUME: b: vol-bf (Volume bold); +% \item +% VOLUME PUNCTUATION: c: volp-com (Volume with comma); +% \item +% PAGE NUMBERS: f: jpg-1 (Only start page number); +% \item +% JOURNAL NAME PUNCTUATION: x: jnm-x (Space after journal); +% \item +% PUBLISHER IN PARENTHESES: d: pub-date (Publisher and date in parentheses); +% \item +% PUBLISHER POSITION: p: pre-pub (Publisher before chapter, pages); +% \item +% ISBN NUMBER: isbn: (Include ISBN for books, booklets) +% \item +% ISSN NUMBER: issn: (Include ISSN for periodicals) +% \item +% EDITOR IN COLLECTIONS: b: edby (Booktitle, edited by \dots); +% \item +% PUNCTUATION BETWEEN SECTIONS (BLOCKS): c: blk-com (Comma between blocks); +% \item +% ABBREVIATE WORD `PAGES': a: pp (`Page' abbreviated); +% \item +% ABBREVIATE WORD `EDITORS': a: ed (`Editor' abbreviated); +% \item +% OTHER ABBREVIATIONS: a: abr (Abbreviations); +% \item +% ABBREVIATION FOR `EDITION': a: ednx (`Edition' abbreviated as `ed'); +% \item +% EDITION NUMBERS: n: ord (Numerical editions); +% \item +% STORED JOURNAL NAMES: a: jabr (Abbreviated journal names); +% \item +% FONT OF `ET AL': i: etal-it (Italic et al); +% \item +% ADDITIONAL REVTeX DATA FIELDS: r: revdata, eprint, url, url-blk (Include REVTeX data fields collaboration, eid, eprint, numpages, url) +% \item +% NEW FONT SELECTION SCHEME: n: nfss (NFSS); +% \item +% ADDITIONAL REVTeX DATA FIELDS: y: revdata (additional data fields); +% \item +% REFERENCE COMPONENT TAGS: y: reference component tags; +% \item +% URL ADDRESS: n: URL as note; +% \end{enumerate} +% +% A file \file{apsrev.dbj} file equivalent to the following should result: +% \begin{verbatim} +% \input docstrip +% \preamble +% ---------------------------------------- +% *** REVTeX-compatible apsrev.bst *** +% \endpreamble +% \postamble +% End of customized bst file +% \endpostamble +% \keepsilent +% \askforoverwritefalse +% \def\MBopts{\from{merlin.mbs}{% +% seq-no%: Citation order (unsorted, like unsrt.bst) +% ,nm-init,ed-au%: Initials + surname (J. F. Smith) +% ,nmlm%: Limited authors (et al replaces missing names) +% ,x10%: Maximum of 10 authors +% ,m10%: Minimum of 10 authors +% ,nmft,nmft-def%: User defined author font (\bibnamefont) +% ,fnm-def%: First names in user defined font (\bibfnamefont) +% ,nmfted%: Editors incollection like authors font +% ,yr-par%: Date in parentheses as (May 1993) +% ,xmth%: Date is year only +% ,tit-it%: Title italic (\em) +% ,jtit-x%: No article title +% ,jttl-rm%: Journal name normal font +% ,trtit-b%: Tech. report title like books +% ,vol-bf%: Volume bold as {\bf vol}(num) +% ,volp-com%: Volume with comma as vol(num), ppp +% ,jpg-1%: Only start page number +% ,jnm-x%: Space after journal name +% ,pub-date%: Publisher and date in parentheses (Oxford, 1994) +% ,pre-pub%: Publisher before chapter, pages +% ,isbn%: Include ISBN for books, booklets, etc. +% ,issn%: Include ISSN for periodicals +% ,edby%: In booktitle, edited by .. (where .. is names) +% ,blk-com%: Comma between blocks +% ,pp%: `Page' abbreviated as p. or pp. +% ,ed%: `Editor' abbreviated as ed. or eds. +% ,abr%: Abbreviations of such words +% ,ednx%: `Edition' abbreviated as `ed' +% ,ord%: Numerical editions as 1st, 2nd, 3rd, etc +% ,jabr%: Abbreviated journal names +% ,etal-it%: Italic et al +% ,revdata,eprint,url,url-blk%: Include REVTeX data fields collaboration, eid, eprint, numpages, url +% ,url,url-nt%: URL as note +% ,bibinfo%: Reference component tags like \bibinfo in the content of \bibitem +% ,nfss%: NFSS use \textbf, \emph, not \bf, \em +% }} +% \generate{\file{\jobname.bst}{\MBopts}} +% \endbatchfile +% \end{verbatim} +% +% To get a banner in the \file{.blg} file, patch by hand into \file{apsrev.bst}: +% \begin{verbatim} +% FUNCTION {banner.log} {"REVTeX-compatible apsrev.bst 9812" warning$} +% \end{verbatim} +% and invoke \texttt{banner.log} from within \texttt{begin.bib} +% +% We ensure that the journal substyle has the first word +% in the matter by installing the (default) APS code +% later on (see Section~\ref{sec:aps-defaults}). +% \end{macro} +% +% \begin{macro}{\bibpunct} +% The following commands effectively establish the style in which \cmd\cite\ commands are +% formatted. +% You can think of them as the second needed component for the bibliography. +% +% Set up for numerical citations. +% +% The journal substyles can override these choices. +% +% Note that setting \cmd\NAT@sort\ at this late hour is not totally effective. +% You have to give \classname{natbib} the relevant options at load time. +% \begin{macrocode} + \bibpunct{[}{]}{,}{n}{}{,}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\bibsection} +% We define the sectioning command to use when starting the bibliography. +% +% \changes{v4.0rc1}{2001/06/18} +% {grid changes with push and pop} +% \begin{macrocode} + \def\bibsection{% + \par +% \twocolumn@sw{\onecolumngrid}{}% + \onecolumngrid@push + \begingroup + \baselineskip26\p@ +% \def\@currentHref{section.bib}% + \bib@device{\textwidth}{245.5\p@}% + \endgroup + \nobreak\@nobreaktrue + \addvspace{19\p@}% + \par +% \twocolumn@sw{\twocolumngrid}{}% + \onecolumngrid@pop + }% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\bib@device} +% \begin{macro}{\bibpreamble} +% \begin{macro}{\bibsep} +% \begin{macro}{\newblock} +% We define the sectioning command to use when starting the bibliography. +% \begin{macrocode} +\def\bib@device#1#2{% + \hb@xt@#1{% + \hfil + \phantomsection + \addcontentsline {toc}{section}{\protect\numberline{}\refname}% +% \hyper@anchorstart {\@currentHref }% + \hb@xt@#2{% + \skip@\z@\@plus-1fil\relax + \leaders\hrule height.25 \p@ depth.25 \p@ \hskip\z@\@plus1fil + \hskip\skip@ + \hskip\z@\@plus0.125fil\leaders\hrule height.375\p@ depth.375\p@ \hskip\z@\@plus0.75fil \hskip\z@\@plus0.125fil + \hskip\skip@ + \hskip\z@\@plus0.25 fil\leaders\hrule height.5 \p@ depth.5 \p@ \hskip\z@\@plus0.5 fil \hskip\z@\@plus0.25 fil + \hskip\skip@ + \hskip\z@\@plus0.375fil\leaders\hrule height.625\p@ depth.625\p@ \hskip\z@\@plus0.25fil \hskip\z@\@plus0.375fil + % \hskip\skip@ + % \hfil + }% +%\hyper@anchorend + \hfil + }% +}% +\let\bibpreamble\@empty +\bibsep\z@\relax +\def\newblock{\ }% +% \end{macrocode} +% \end{macro} +% \end{macro} +% \end{macro} +% \end{macro} +% +% \begin{macro}{\bibfont} +% We define the font switch that applies to the body of the bibliography. +% +% \begin{macrocode} + \def\bibfont{% + \small + }% +% \end{macrocode} +% \end{macro} +% +% +% \subsubsection{Index}% +% FIXME: the following call to \cmd\twocolumn\ appears wrong if we were in two-column grid. +% \changes{v4.0beta 3}{1999/11/13} +% {grid changes with ltxgrid} +% \changes{v4.0rc1}{2001/06/18} +% {grid changes with push and pop} +% \begin{macrocode} +\newenvironment{theindex}{% +%\let\wastwocol@sw\twocolumn@sw + \columnseprule \z@ + \columnsep 35\p@ + \c@secnumdepth-\maxdimen + \onecolumngrid@push + \section{\indexname}% + \thispagestyle{plain}% + \parindent\z@ + \parskip\z@ plus.3\p@\relax + \let\item\@idxitem +%\wastwocol@sw{\twocolumngrid}{}% + \onecolumngrid@pop +}{% +%\onecolumngrid@pop +}% +% +\def\@idxitem{\par\hangindent 40\p@} +% +\def\subitem{\par\hangindent 40\p@ \hspace*{20\p@}} +% +\def\subsubitem{\par\hangindent 40\p@ \hspace*{30\p@}} +% +\def\indexspace{\par \vskip 10\p@ plus5\p@ minus3\p@\relax} +% \end{macrocode} +% +% +% \subsection{Journal-Specific Code}% +% +% For APS journals, we supply code specific to PRA, PRB, PRC, PRD, PRE, PRL, and RMP. +% At present, they are identical, +% with the exception of PRB's code for superscript citations. +% +% \begin{macrocode} +\def\journal@pra{pra}% +\def\journal@prb{prb}% +\def\journal@prc{prc}% +\def\journal@prd{prd}% +\def\journal@pre{pre}% +\def\journal@prl{prl}% +\def\journal@prstab{prstab}% +\def\journal@rmp{rmp}% +% \end{macrocode} +% +% Note: the convention in this document class is that +% the substyle must not override any explicit class options +% declared by the document. +% This means that the various Booleans of Section~\ref{sec:options} +% may be assigned here only if they are still undefined at this point. +% +% For most all of the APS journals, the journal-dependent +% code is relatively meager and is therefore embedded in this file. +% However, the RMP code is sufficiently extensive that +% splitting it out into a separate file is more convenient. +% +% \subsubsection{\classoption{pra}}% +% \begin{macrocode} +\@ifx{\@journal\journal@pra}{% +% \end{macrocode} +% +% End of \classoption{pra} code. +% \begin{macrocode} +}{}% +% \end{macrocode} +% +% \subsubsection{\classoption{prb}}% +% \begin{macrocode} +\@ifx{\@journal\journal@prb}{% +% \end{macrocode} +% PRB requires superscript citations. +% We use Patrick Daly's \classname{natbib} package, +% and hyperref and other packages are already set up to cope with this. +% +% \begin{macro}{\bibpunct} +% By default, PRB selects the \classoption{bibnotes} option. +% We also set up for numerical citations. +% \begin{macrocode} + \@ifxundefined\bibnotes@sw{\@booleantrue\bibnotes@sw}{}% + \bibpunct{}{}{,}{s}{}{\textsuperscript{,}}% + \def\@onlinecite#1{\begingroup\let\@cite\NAT@citenum\citealp{#1}\endgroup}% +% \end{macrocode} +% \end{macro} +% +% Invoke \classoption{superbib} option if the document +% has made no selection of its own. +% \begin{macrocode} + \appdef\class@documenthook{% + \@ifxundefined\place@bibnumber{% + \let\place@bibnumber\place@bibnumber@sup + }{}% + }% +% \end{macrocode} +% +% End of \classoption{prb} code. +% \begin{macrocode} +}{}% +% \end{macrocode} +% +% \subsubsection{\classoption{prc}}% +% \begin{macrocode} +\@ifx{\@journal\journal@prc}{% +% \end{macrocode} +% +% End of \classoption{prc} code. +% \begin{macrocode} +}{}% +% \end{macrocode} +% +% \subsubsection{\classoption{prd}}% +% \begin{macrocode} +\@ifx{\@journal\journal@prd}{% +% \end{macrocode} +% +% End of \classoption{prd} code. +% \begin{macrocode} +}{}% +% \end{macrocode} +% +% \subsubsection{\classoption{pre}}% +% \begin{macrocode} +\@ifx{\@journal\journal@pre}{% +% \end{macrocode} +% +% End of \classoption{prd} code. +% \begin{macrocode} +}{}% +% \end{macrocode} +% +% \subsubsection{\classoption{prl}}% +% \begin{macrocode} +\@ifx{\@journal\journal@prl}{% +% \end{macrocode} +% In PRL, the default is the \classoption{bibnotes} option, +% and the Acknowledgments section has no head. +% +% However, the References head is a device that may be described as a lozenge centered on the baseline, +% 71 points wide by 2 points thick, with the ends tapering to a half point in thickness. +% Space above 26 points base to base, below 31 base to base. +% \begin{macrocode} + \@ifxundefined\bibnotes@sw{\@booleantrue\bibnotes@sw}{}% + \@booleanfalse\acknowledgments@sw +%\let\acknowledgmentsname\@empty + \def\bibsection{% + \par + \begingroup + \baselineskip26\p@ + \bib@device{\hsize}{72\p@}% + \endgroup + \nobreak\@nobreaktrue + \addvspace{19\p@}% + }% +% \end{macrocode} +% Implement length checking. +% Use the \classname{times} and \classname{mathtime} packages, +% plus whatever other processing is required +% to make the formatted output be true +% to the metrics of the journal. +% +% Wait. Do not use mathtime after all. APS has their own way of doing math pi, involving +% Adobe Mathematical Pi and other fonts. +% \begin{macrocode} +%\appdef\class@documenthook{% + \@ifxundefined\lengthcheck@sw{\@booleanfalse\lengthcheck@sw}{}% + \lengthcheck@sw{% + \RequirePackage{times}% +% \RequirePackage{mathtime}% + }{% + }% +%}% +% \end{macrocode} +% A PRL does not have numbered sections. +% \begin{macrocode} + \c@secnumdepth=-\maxdimen +% \end{macrocode} +% Note: we defer this code until after type size file is read in. +% \begin{macrocode} + \appdef\set@typesize@hook{% + \@ifnum{\@pointsize=10\relax}{% + \lengthcheck@sw{% + \def\large{% + \@setfontsize\large{12.5}{14\p@}% + }% + \def\normalsize{% + \@setfontsize\normalsize{10.5}\@xiipt + \abovedisplayskip 6\p@ \@plus6\p@ \@minus5\p@ + \belowdisplayskip \abovedisplayskip + \abovedisplayshortskip \abovedisplayskip + \belowdisplayshortskip \abovedisplayskip + \let\@listi\@listI + }% + \def\small{% + \@setfontsize\small{9.5}\@xipt + \abovedisplayskip 5\p@ \@plus5\p@ \@minus4\p@ + \belowdisplayskip \abovedisplayskip + \abovedisplayshortskip \abovedisplayskip + \belowdisplayshortskip \abovedisplayskip + \let\@listi\@listI + }% + \DeclareMathSizes{12.5}{12.5}{9}{6}% + \DeclareMathSizes{10.5}{10.5}{7.5}{5}% + \DeclareMathSizes{9.5}{9.5}{7.0}{5}% + }{% + \def\normalsize{% + \@setfontsize\normalsize\@xpt\@xiipt + \abovedisplayskip 10\p@ \@plus2\p@ \@minus5\p@ + \belowdisplayskip \abovedisplayskip + \abovedisplayshortskip \abovedisplayskip + \belowdisplayshortskip \abovedisplayskip + \let\@listi\@listI + }% + }% + }{}% + }% +%\appdef\set@pica@hook{% + \@ifnum{\@pointsize=10\relax}{% + \textheight = 694.0\p@ +% \textheight = 55\baselineskip +% \advance\textheight by \topskip + }{}% +%}% +% \end{macrocode} +% +% End of \classoption{prl} code. +% \begin{macrocode} +}{}% +% \end{macrocode} +% +% \subsubsection{\classoption{prstab}}% +% \begin{macrocode} +\@ifx{\@journal\journal@prstab}{% +% \end{macrocode} +% +% End of \classoption{prstab} code. +% \begin{macrocode} +}{}% +% \end{macrocode} +% +% \subsubsection{\classoption{rmp}}% +% Read in the code from a file. +% \begin{macrocode} +\@ifx{\@journal\journal@rmp}{\input{\@journal\substyle@ext}}{}% +% \end{macrocode} +% +% +% \subsection{Establish APS Defaults}\label{sec:aps-defaults} +% +% \begin{macro}{\footinbib@sw} +% All APS journals except RMP invoke the \classoption{footinbib} option. +% We rely on the RMP-specific code to give a value to this switch. +% \begin{macrocode} +\@ifxundefined\footinbib@sw{\@booleantrue\footinbib@sw}{}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\place@bibnumber} +% \begin{macro}{\@bibstyle} +% We install code that will select the presentation for \cmd\bibitem s +% and govern the \BibTeX\ processing. +% If the user has not selected any related option and if +% the journal substyle also has not, then the \classoption{aps} +% option choses these defaults. +% \begin{macrocode} +\appdef\class@documenthook{% + \@ifxundefined\place@bibnumber{% + \let\place@bibnumber\place@bibnumber@inl + }{}% + \@ifxundefined\@bibstyle{% + \def\@bibstyle{apsrev}% + }{}% +}% +% \end{macrocode} +% \end{macro} +% \end{macro} +% +% +% \begin{macrocode} +%</aps> +% \end{macrocode} +% +% +% \section{The \classoption{rmp} journal substyle: the \texttt{rmp} module} +% +% The file \file{rmp.rtx} is read in by the \classname{revtex4} document class +% if \cmd\@society\ has the value \classoption{aps} +% and \cmd\@journal\ has the value \classoption{rmp}. +% +% It is read at the end of the \filename{aps.rtx}, so all definitions +% and assignments in that file are operative unless overridden here. +% +% +% \begin{macrocode} +%<*rmp> +% \end{macrocode} +% +% Protect this file from being read in by anything but \revtex. +% \begin{macrocode} +\ifx\undefined\substyle@ext + \def\@tempa{% + \endinput + \GenericWarning{I must be read in by REVTeX! (Bailing out)}% + }% + \expandafter\else + \def\@tempa{}% + \expandafter\fi\@tempa +% \end{macrocode} +% +% Protect this file from being read in as a society instead of a journal. +% In such a case, \cmd\@journal\ will be \cmd\@empty. +% \begin{macrocode} +\@ifx@empty\@journal{% + \class@warn{I have to read in the aps substyle first!}% + \endinput + \def\@society{aps}\def\@journal{rmp}% + \expandafter\input\expandafter{\@society\substyle@ext}% +}{}% +% \end{macrocode} +% +% \subsection{Frontmatter}% +% +% \begin{macro}{\frontmatter@setup} +% +% \begin{macrocode} +\def\frontmatter@setup{% + \normalfont\sffamily\raggedright +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\PACS@warn} +% Per Mark Doyle, RMP never displays the PACS, so they don't want the `use showpacs' warning spit out. +% \begin{macrocode} +\def\PACS@warn{RMP documents do not display PACS. Your \string\pacs\space will be ignored}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\frontmatter@title@above} +% \begin{macro}{\frontmatter@title@format} +% \begin{macro}{\frontmatter@title@below} +% +% \begin{macrocode} +\def\frontmatter@title@above{}% +\def\frontmatter@title@format{\Large\bfseries\raggedright}% HelveticaNeue-Medium(Italic) 14pt. +\def\frontmatter@title@below{\addvspace{12\p@}}% 24pt b-b down to first author +% \end{macrocode} +% \end{macro} +% \end{macro} +% \end{macro} +% +% \begin{macro}{\frontmatter@authorformat} +% Set the rag to a milder value, because we want to do true ragged right +% typesetting, as opposed to the \LaTeX\ default, which gives very poor results. +% +% Note: author font is 9.8bp. 19.2bp/14.3bp above/below. +% \begin{macrocode} +\def\frontmatter@authorformat{% + \preprintsty@sw{\vskip0.5pc\relax}{}% + \@tempskipa\@flushglue + \@flushglue\z@ plus50\p@\relax + \raggedright\advance\leftskip.5in\relax + \@flushglue\@tempskipa + \parskip\z@skip +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\frontmatter@affiliationfont} +% The hook \cmd\frontmatter@affiliationfont\ controls the formatting of affiliations and affiliation groups. +% The hook \cmd\frontmatter@above@affilgroup\ is invoked just before proceeding with author/affiliation processing. +% The \cmd\frontmatter@above@affiliation\ is the amount of space above affiliations in the \classoption{groupedaddress} style, +% and \cmd\frontmatter@above@affiliation@script\ is that for \classoption{superscriptaddress}. +% +% Note: affiliation font is 9.03/10.4bp, 14.3bp/19.2bp b-b above/below +% \begin{macrocode} +\def\frontmatter@affiliationfont{% Helvetica 9/10.2 + \small\slshape\selectfont\baselineskip10.5\p@\relax + \@tempskipa\@flushglue + \@flushglue\z@ plus50\p@\relax + \raggedright\advance\leftskip.5in\relax + \@flushglue\@tempskipa +}% +\def\frontmatter@above@affilgroup{\addvspace{7.2\p@}}% additional leading above an author +\def\frontmatter@above@affiliation{\addvspace{5.3\p@}}% +\def\frontmatter@above@affiliation@script{}% +% \end{macrocode} +% \end{macro} +% +% Set up the default RMP style for title block authors and affiliations. +% This command should effectively override the effect of the +% corresponding command in the parent substyle. +% \begin{macrocode} +\@ifxundefined\groupauthors@sw{% + \clo@groupedaddress +}{}% +% \end{macrocode} +% +% \begin{macro}{\frontmatter@RRAP@format} +% Note: in RMP, if we are not in preprint mode, the date will not be produced. +% +% Note: Helvetica C/lc, 8.98bp, space above: 16.3bp b-b. +% \begin{macrocode} +\def\frontmatter@RRAP@format{% + \addvspace{7.3\p@}% + \small + \raggedright\advance\leftskip.5in\relax +% \parskip.5ex\relax +% \everypar{\hbox\bgroup(\@gobble@leavemode@uppercase}% +% \def\par{\@ifvmode{}{\unskip)\egroup\@@par}}% +% \preprintsty@sw{}{\let\@date\@empty}% +}% +%\def\@gobble@leavemode@uppercase#1#2{\expandafter\MakeTextUppercase}% +\def\produce@RRAP#1{% + \@if@empty{#1}{}{% + \@ifvmode{\leavevmode}{}% + \unskip(\ignorespaces#1\unskip)\quad + }% +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\frontmatter@abstractheading} +% Space above 21.8bp b-b. +% \begin{macrocode} +\def\frontmatter@abstractheading{% + \preprintsty@sw{% + \begingroup + \centering\large + \abstractname + \par + \endgroup + \vspace{.5pc}% + }{}% +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\frontmatter@abstractfont} +% TimesTen 8.93bp/9.6bp X 360bp, indented 36bp, with 21.9/37.6bp b-b above/below +% \begin{macrocode} +\def\frontmatter@abstractfont{% + \footnotesize + \hsize360\p@ + \leftskip=0.5in + \advance\hsize\leftskip +%\rightskip=\leftskip + \parindent\z@ +%\hsize5.5in +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\frontmatter@preabstractspace} +% \begin{macro}{\frontmatter@postabstractspace} +% Space above and space below abstract in title block. +% Should be 22/36 points base-to-base. +% \begin{macrocode} +\def\frontmatter@preabstractspace{7.7\p@}% +\def\frontmatter@postabstractspace{24.6\p@}% +% \end{macrocode} +% \end{macro} +% \end{macro} +% +% FIXME: Not done: PACS. +% FIXME: TOC: Head is same as \section: HelveticaNeue 8.98. 32/22bp b-b above/below, +% Body: TimesTen 8/10.5. +% +% \subsection{General Text}% +% +% If not in preprint mode, set the type size to 10/12 point. Note: s/b 11.6bp leading +% \begin{macrocode} + \appdef\set@typesize@hook{% + \@ifxundefined\preprintsty@sw{}{% + \preprintsty@sw{}{% + \def\normalsize{% + \@setsize\normalsize{12pt}\xpt\@xpt + \abovedisplayskip 10\p@ plus2\p@ minus5\p@ + \belowdisplayskip \abovedisplayskip + \abovedisplayshortskip \abovedisplayskip + \belowdisplayshortskip \abovedisplayskip + \let\@listi\@listI + }% + }% + }% + }% +% \end{macrocode} +% +% Footnote mods: +% \begin{macrocode} + \footnotesep 9.25pt + \skip\footins 36pt plus 4pt minus 2pt + \def\footnoterule{\kern-13pt\hrule width.5in\kern15.6pt\relax}% +% \end{macrocode} +% +% +% \subsection{Sectioning} +% +% \begin{macrocode} +\def\thepart{\Roman{part}} % +\def\thesection{\Roman{section}} +\def\p@section{} +\def\thesubsection{\Alph{subsection}} +\def\p@subsection{\thesection.} +\def\thesubsubsection{\arabic{subsubsection}} +\def\p@subsubsection{\thesection.\thesubsection.} +\def\theparagraph{\alph{paragraph}} +\def\p@paragraph{\thesection.\thesubsection.\thesubsubsection.} +\def\thesubparagraph{\arabic{subparagraph}} +\def\p@subparagraph{\thesection.\thesubsection.\thesubsubsection.\theparagraph.} +% \end{macrocode} +% +% In RMP, put a period (.), followed by nut space, after the section number. +% Also, hang the section number (the \LaTeX\ default). +% \begin{macrocode} +\def\@seccntformat#1{\csname the#1\endcsname.\hskip0.5em\relax}% +% \end{macrocode} +% +% Note that we wish to set the section head uppercase, so we use +% David Carlisle's \cmd\MakeTextUppercase. However, because this procedure effectively +% parses its argument (looking for things to \emph{not} translate), it has to be invoked +% in such a way that the argument of the \cmd\section\ command is passed to it as its +% own argument. +% +% To accomplish this, we use the \cmd\@hangfrom@\ hook, which was developed for this +% purpose. +% \begin{macrocode} + \def\section{% + \@startsection{section}{1}{\z@}{0.8cm plus1ex minus.2ex}{0.4cm}% + {% + \small\sffamily\bfseries\selectfont + \raggedright + \parindent\z@ + }% + }% + \def\@hangfrom@section#1#2#3{\@hangfrom{#1#2}\MakeTextUppercase{#3}}% + \def\subsection{% + \@startsection{subsection}{2}{\z@}{0.8cm plus1ex minus.2ex}{0.4cm}% + {% + \small\sffamily\bfseries + \raggedright + \parindent\z@ + }% + }% + \def\subsubsection{% + \@startsection{subsubsection}{3}{\z@}{.8cm plus1ex minus.2ex}{0.4cm}% + {% + \small\sffamily\selectfont + \raggedright + \parindent\z@ + }% + }% + \def\paragraph{% + \@startsection{paragraph}{4}{\z@}{.8cm plus1ex minus.2ex}{-1em}% + {% + \small\slshape\selectfont + \raggedright + \parindent\z@ + }% + }% + \def\subparagraph{% + \@startsection{subparagraph}{4}{\parindent}{3.25ex plus1ex minus.2ex}{-1em}% + {\normalsize\bfseries\selectfont}% + }% +% + \setcounter{tocdepth}{4}% FIXME: has no effect +% \end{macrocode} +% +% \begin{macro}{\appendix} +% \begin{macro}{\@hangfrom@appendix} +% \begin{macro}{\@hangfroms@appendix} +% \begin{macro}{\@appendixcntformat} +% \begin{macrocode} +\appdef\appendix{% + \let\@hangfrom@section\@hangfrom@appendix + \let\@hangfroms@section\@hangfroms@appendix + \let\@sectioncntformat\@appendixcntformat +}% +\def\@hangfrom@appendix#1#2#3{% + #1% + \@if@empty{#2}{% + #3% + }{% + #2\@if@empty{#3}{}{:\ #3}% + }% +}% +\def\@hangfroms@appendix#1#2{% + #1\appendixname\@if@empty{#2}{}{:\ #2}% +}% +\def\@appendixcntformat#1{\appendixname\ \csname the#1\endcsname}% +% \end{macrocode} +% \end{macro} +% \end{macro} +% \end{macro} +% \end{macro} +% +% +% \subsection{Figure and Table Caption Formatting} +% \begin{macro}{\@makecaption} +% +% \begin{macrocode} +\setlength\belowcaptionskip{2\p@} +\long\def\@makecaption#1#2{% + \vskip\abovecaptionskip + \vbox{% + \flushing + \small\rmfamily +% \hsize\@capwidth + \noindent + #1\nobreak\hskip.5em plus.2em\ignorespaces#2\par + }% + \vskip\belowcaptionskip +}% +% \end{macrocode} +% \end{macro} +% +% \subsection{Citations and Bibliography}% +% +% Customize \revtex\ for the journal substyle; +% this task requires three components: +% a \BibTeX\ \file{.bst} style file, +% customizing code for \file{natbib}, and +% customizations of the \env{thebibliography} environment. +% +% \begin{macro}{\@bibstyle} +% Define the argument of the \cmd\bibliographystyle\ command +% (if the document does not do so). +% +% The user must have installed a \file{.bst} file of the corresponding name. +% This file will then be used by \BibTeX\ when compiling the document's \file{.bbl} file. +% +% To generate \file{apsrmp.bst}, use \classname{custom-bib} version 3.89d1 or later. +% Run the \file{.bst} generator, \file{makebst.tex}, with the following options: +% +% \begin{enumerate} +% \item +% STYLE OF CITATIONS: a: ay( Author-year with some non-standard interface) +% \item +% AUTHOR-YEAR SUPPORT SYSTEM: nat: (Natbib for use with natbib v5.3 or later) +% \item +% ORDERING OF REFERENCES: *: (Alphabetical); +% \item +% ORDER ON VON PART: x: vonx (Sort without von part---de la Maire after Mahone) +% \item +% AUTHOR NAMES: a: nm-rev1 (Only first name reversed, initials---AGU style: Smith, J. F., H. K. Jones) +% \item +% POSITION OF JUNIOR: jnrlst (Junior comes last as Smith, John, Jr.) +% \item +% NUMBER OF AUTHORS: l: max 12, min 12 (three prompts in all); +% \item +% TYPEFACE FOR AUTHORS IN LIST OF REFERENCES: u: nmft, nmft-def (User defined author font---\cmd\bibnamefont) +% \item +% FONT FOR FIRST NAMES: u: fnm-def (First names in user defined font---\cmd\bibfnamefont) +% \item +% EDITOR NAMES IN INCOLLECTION ETC: a: nmfted (Editors incollection like authors font) +% \item +% FONT FOR `AND' IN LIST: r: nmand-rm (`And' in normal font---JONES and JAMES) +% \item +% FONT OF CITATION LABELS IN TEXT: u: lab, lab-def (User-defined citation font---\cmd\citenamefont) +% \item +% LABEL WHEN AUTHORS MISSING: keyxyr (Year blank when KEY replaces missing author---for natbib 7.0) +% \item +% DATE POSITION: b: dt-beg (Date after authors) +% \item +% SUPPRESS MONTH: xmth (Date is year only) +% \item +% DATE FORMAT: yr-com (Date preceded by comma as `, 1993') +% \item +% TRUNCATE YEAR: note-yr (Year text full as 1990--1993 or `in press') +% \item +% ARTICLE TITLE PRESENT: x: jtit-x (No article title); +% \item +% JOURNAL TITLE FONT: r: jttl-rm (Journal title normal); +% \item +% TECHNICAL REPORT TITLE: b: trtit-b (Tech. report title like books); +% \item +% JOURNAL VOLUME: b: vol-bf (Volume bold); +% \item +% VOLUME PUNCTUATION: c: volp-com (Volume with comma); +% \item +% PAGE NUMBERS: f: jpg-1 (Only start page number); +% \item +% POSITION OF PAGES: e: pp-last (Pages at end, but before any notes) +% \item +% NUMBER AND SERIES FOR COLLECTIONS: num-xser (Allows number without series and suppresses word ``number'') +% \item +% JOURNAL NAME PUNCTUATION: x: jnm-x (Space after journal); +% \item +% PUBLISHER IN PARENTHESES: p: pub-par (Publisher in parentheses) +% \item +% PUBLISHER POSITION: p: pre-pub (Publisher before chapter, pages); +% \item +% ISBN NUMBER: *: isbn (Include ISBN); +% \item +% ISSN NUMBER: *: isbn (Include ISSN for periodicals); +% \item +% `EDITOR' AFTER NAMES: a: edpar (`Name (editor),' in parentheses, after name, comma after) +% \item +% EDITOR IN COLLECTIONS: b: edby (In booktitle, edited by\dots---where \dots is names) +% \item +% PUNCTUATION BETWEEN SECTIONS (BLOCKS): c: blk-com (Comma between blocks); +% \item +% `IN' WITH JOURNAL NAMES: i: injnl (Add `in' before journal name in style for incollection) +% \item +% ABBREVIATE WORD `PAGES': a: pp (`Page' abbreviated); +% \item +% ABBREVIATE WORD `EDITORS': a: ed (`Editor' abbreviated); +% \item +% ABBREVIATION FOR `EDITION': a: ednx (`Edition' abbreviated as `ed'); +% \item +% EDITION NUMBERS: xedn (Editions as in database saving much processing memory) +% \item +% STORED JOURNAL NAMES: a: jabr (Abbreviated journal names); +% \item +% FONT OF `ET AL': i: etal-it (Italic et al); +% \item +% ADDITIONAL REVTeX DATA FIELDS: r: revdata, eprint, url, url-blk (Include REVTeX data fields) +% \item +% NEW FONT SELECTION SCHEME: n: nfss (NFSS); +% \item +% ADDITIONAL REVTeX DATA FIELDS: y: revdata (additional data fields); +% \item +% REFERENCE COMPONENT TAGS: r: bibinfo (reference component tags); +% \end{enumerate} +% +% A file \file{apsrmp.dbj} file equivalent to the following should result: +% \begin{verbatim} +% \input docstrip +% \preamble +% ---------------------------------------- +% *** REVTeX-compatible apsrmp.bst *** +% \endpreamble +% \postamble +% End of customized bst file +% \endpostamble +% \keepsilent +% \askforoverwritefalse +% \def\MBopts{\from{merlin.mbs}{% +% ay%: Author-year with some non-standard interface +% ,nat%: Natbib for use with natbib v5.3 or later +% ,vonx%: Sort without von part (de la Maire after Mahone) +% ,nm-rev1%: Only first name reversed, initials (AGU style: Smith, J. F., H. K. Jones) +% ,jnrlst%: Junior comes last as Smith, John, Jr. +% ,nmlm%: Limited authors (et al replaces missing names) +% ,x10%: Maximum of 10 authors +% ,m10%: Minimum of 10 authors +% ,nmft,nmft-def%: User defined author font (\bibnamefont) +% ,fnm-def%: First names in user defined font (\bibfnamefont) +% ,nmfted%: Editors incollection like authors font +% ,nmand-rm%: `And' in normal font (JONES and JAMES) +% ,lab,lab-def%: User defined citation font (\citenamefont) +% ,keyxyr%: Year blank when KEY replaces missing author (for natbib 7.0) +% ,dt-beg%: Date after authors +% ,xmth%: Date is year only +% ,yr-com%: Date preceded by comma as `, 1993' +% ,note-yr%: Year text full as 1990--1993 or `in press' +% ,jtit-x%: No article title +% ,jttl-rm%: Journal name normal font +% ,trtit-b%: Tech. report title like books +% ,vol-bf%: Volume bold as {\bf vol}(num) +% ,volp-com%: Volume with comma as vol(num), ppp +% ,jpg-1%: Only start page number +% ,pp-last%: Pages at end but before any notes +% ,num-xser%: Allows number without series and suppresses word "number" +% ,jnm-x%: Space after journal name +% ,pub-par%: Publisher in parentheses +% ,pre-pub%: Publisher before chapter, pages +% ,isbn%: Include ISBN for books, booklets, etc. +% ,issn%: Include ISSN for periodicals +% ,edpar%: `Name (editor),' in parentheses, after name, comma after +% ,edby%: In booktitle, edited by .. (where .. is names) +% ,blk-com%: Comma between blocks +% ,injnl%: Add `in' before journal name in style for incollection +% ,pp%: `Page' abbreviated as p. or pp. +% ,ed%: `Editor' abbreviated as ed. or eds. +% ,xedn%: Editions as in database saving much processing memory +% ,jabr%: Abbreviated journal names +% ,etal-it%: Italic et al +% ,revdata,eprint,url,url-blk%: Include REVTeX data fields collaboration, eid, eprint, numpages, url +% ,bibinfo%: Reference component tags like \bibinfo in the content of \bibitem +% ,nfss%: NFSS use \textbf, \emph, not \bf, \em +% }} +% \generate{\file{\jobname.bst}{\MBopts}} +% \endbatchfile +% \end{verbatim} +% +% \begin{macrocode} +\appdef\class@documenthook{% + \@ifxundefined\@bibstyle{% + \def\@bibstyle{apsrmp}% + }{}% +}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\bibpunct} +% The following commands effectively etablish the style in which \cmd\cite\ commands are +% formatted. +% You can think of them as the second needed component for the bibliography. +% +% Set up for author-year citations: +% when \cmd\NAT@set@cites\ executes (at \envb{document} time), the \cmd\@biblabel\ will +% be set to \cmd\NAT@biblabel. +% +% Per Karie Friedman (friedman@phys.washington.edu), multiple citations are separated by semicolons, e.g., +% (Jones, 1999; Abbott and Smith, 2000; Wortley, 2001a), +% and +% multiple citations by the same author by commas, e.g., +% Abela et al. (1995, 1997a, 1997b). +% The third argument of \cmd\bibpunct\ handles the former. +% +% The fifth argument puts a comma after the author when the year is not in parens: +% (Lee et al., 1996). +% +% \begin{macrocode} + \bibpunct{(}{)}{;}{a}{,}{,}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\cite} +% Per Mark Doyle, \cmd\cite\ is mapped to \cmd\citep\ in RMP. +% +% \begin{macrocode} + \let\cite\citep +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\bibsection} +% \begin{macro}{\bibpreamble} +% \begin{macro}{\newblock} +% \begin{macro}{\bibhang} +% \begin{macro}{\bibsep} +% We define the sectioning command to use when starting the bibliography. +% +% \begin{macrocode} +\def\bibsection{% + \let\@hangfroms@section\@hang@froms + \section*{\refname}% + \@nobreaktrue +}% +\let\bibpreamble\@empty +\def\newblock{\ }% +\bibhang10\p@ +\bibsep\z@ +% \end{macrocode} +% \end{macro} +% \end{macro} +% \end{macro} +% \end{macro} +% \end{macro} +% +% \begin{macro}{\footinbib@sw} +% Footnotes in bibliography are inconsistent with author-year +% references, and are particularly nast under \classname{natbib}: +% the packcage will automatically change to numbered references if +% any \cmd\bibitem\ commands lack the optional argument. +% Therefore, we must uninvoke it now, even if invoked by the document. +% \begin{macrocode} +\@ifxundefined\footinbib@sw{}{% + \footinbib@sw{% + \class@warn{% + Footnotes in bibliography are incompatible with RMP.^^J% + Undoing the footinbib option. + }% + }{}% +}% +\@booleanfalse\footinbib@sw +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\eprint} +% RMP requires the \cmd\eprint\ field in the bib entry to be set off with the word ``eprint''. +% \begin{macrocode} +\def\eprint#1{eprint #1}% +% \end{macrocode} +% \end{macro} +% +% \subsection{Table of Contents} +% +% We set up for auto-sizing of certain TOC elements. +% +% To do this, we override certain definitions for +% the default TOC font (\cmd\toc@@font) and spacing (\@dotsep), +% and define +% formatting for the needed elements (\cmd\l@\dots). +% Finally, we activate the autosizing by +% assigning \cmd\toc@pre\ and \cmd\toc@post. +% +% \begin{macro}{\toc@@font} +% Set the formatting characteristics of the auto-indenting part +% of the TOC. +% \begin{macrocode} +\def\toc@@font{% + \footnotesize\rmfamily + \def\\{\space\ignorespaces}% +}% +\def\@dotsep{5.5pt}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\l@section} +% Determine which TOC elements are automatically indented. +% \begin{macrocode} +\def\tocleft@{\z@}% +\def\tocdim@min{5\p@}% +\def\l@section{% + \l@@sections{}{section}% Implicit #3#4 +}% +\def\l@subsection{% + \l@@sections{section}{subsection}% Implicit #3#4 +}% +\def\l@subsubsection{% + \l@@sections{subsection}{subsubsection}% Implicit #3#4 +}% +%\def\l@subsubsection#1#2{}% +\def\l@paragraph#1#2{}% +\def\l@subparagraph#1#2{}% +% \end{macrocode} +% \end{macro} +% +% Activate the TOC processing. +% \begin{macrocode} +\let\toc@pre\toc@pre@auto +\let\toc@post\toc@post@auto +% \end{macrocode} +% +% +% \begin{macrocode} +%</rmp> +% \end{macrocode} +% +% +% \Finale +% %Here ends the programmer's documentation. +% \endinput +% +\endinput diff --git a/Master/texmf-dist/source/latex/revtex4/revtex4.ins b/Master/texmf-dist/source/latex/revtex4/revtex4.ins new file mode 100644 index 00000000000..96cf5b09722 --- /dev/null +++ b/Master/texmf-dist/source/latex/revtex4/revtex4.ins @@ -0,0 +1,36 @@ +% This file requires docstrip version 2.4 or higher available from +% ftp://ctan.tug.org/tex-archive/macros/latex/unpacked/docstrip.tex +\input docstrip +\preamble + +This file is part of the APS files in the REVTeX 4 distribution. +For the version number, search on the string %FileInfo + +Original version by David Carlisle +Modified by Arthur Ogawa (mailto:ogawa@teleport.com) + +Copyright (c) 1999 The American Physical Society. +http://publish.aps.org/revtex4/ +mailto:revtex4@aps.org + +See the REVTeX 4 README file for restrictions and more information. + +\endpreamble +\keepsilent +\askforoverwritefalse +\generate{ + \file{revtex4.cls}{% + \from{revtex4.dtx}{class-pre}% + \from{ltxutil.dtx}{ltxutil-krn}% + \from{ltxgrid.dtx}{ltxgrid-krn}% + \from{textcase.dtx}{package}% + \from{revtex4.dtx}{class-pst}% + }% + \file{aps.rtx}{\from{revtex4.dtx}{aps}} + \file{rmp.rtx}{\from{revtex4.dtx}{rmp}} + \file{10pt.rtx}{\from{revtex4.dtx}{10pt}} + \file{11pt.rtx}{\from{revtex4.dtx}{11pt}} + \file{12pt.rtx}{\from{revtex4.dtx}{12pt}} + \file{revsymb.sty}{\from{revtex4.dtx}{revsymb}} +} +\endbatchfile diff --git a/Master/texmf-dist/source/latex/revtex4/textcase.dtx b/Master/texmf-dist/source/latex/revtex4/textcase.dtx new file mode 100644 index 00000000000..60010e5ee67 --- /dev/null +++ b/Master/texmf-dist/source/latex/revtex4/textcase.dtx @@ -0,0 +1,404 @@ +% \iffalse +%% Source File: textcase.dtx +%% Copyright 1997 1998 David Carlisle +%% +%% This file may be distributed under the terms of the LPPL. +%% See 00readme.txt for details. +% +%<*dtx> + \ProvidesFile{textcase.dtx} +%</dtx> +%<package>\NeedsTeXFormat{LaTeX2e} +%<package>\ProvidesPackage{textcase} +%<driver> \ProvidesFile{textcase.drv} +% \fi +% \ProvidesFile{textcase.dtx} + [1998/11/12 v0.06 Text only upper/lower case changing (DPC)] +% +% \iffalse +%<*driver> +\documentclass{ltxdoc} +\usepackage{textcase} +\begin{document} +\DocInput{textcase.dtx} +\end{document} +%</driver> +% \fi +% +% \GetFileInfo{textcase.dtx} +% \CheckSum{110} +% +% \changes{v0.01}{1997/09/28} +% {Posted to c.t.t} +% \changes{v0.02}{1997/09/28} +% {Support \cs{ensuremath}} +% \changes{v0.03}{1997/12/12} +% {Donald Arseneau: support \cs{ref} \cs{cite} \cs{label}} +% \changes{v0.04}{1997/12/14} +% {Convert to dtx form, add \cs{NoCaseChange}} +% \changes{v0.04}{1997/12/14} +% {Support (most uses of) optional argument of \cs{cite}} +% \changes{v0.05}{1997/12/16} +% {More comments from Donald, +% perhaps I should just give him this package for Christmas} +% \changes{v0.06}{1998/11/12} +% {overload option} +% +% +% \title{The \textsf{textcase} package\thanks{This file +% has version number \fileversion, last +% revised \filedate.}} +% \author{David Carlisle\thanks{%^^A +% Reorganisation to suppport skipping a wider class of +% things than just math done by Donald Arseneau}} +% +% \date{\filedate} +% +% \maketitle +% +% \section{Introduction}\label{intro} +% |\MakeTextUppercase| and |\MakeTextLowercase| are versions of the +% standard |\MakeUppercase| and |\MakeLowercase| that do not change +% the case of any math sections in their arguments. +%\begin{verbatim} +% \MakeTextUppercase{abc\ae\ \( a = b \) and $\alpha \neq a$ +% or even \ensuremath{x=y} and $\ensuremath{x=y}$} +%\end{verbatim} +% Should produce: +% \begin{quotation} +% ABC\AE\ \( a = b \) AND $\alpha \neq a$ +% OR EVEN \ensuremath{x=y} AND $\ensuremath{x=y}$ +% \end{quotation} +% +% Version 0.03 of this package incorporates some changes suggested by +% Donald Arseneau so that as well as math mode, the arguments of +% |\cite|, |\label| and |\ref| are also prevented from being uppercased. +% So you can now go +%\begin{verbatim} +% \MakeTextUppercase{% +% Text in section~\ref{intro}, about \cite[pp 2--4]{bbb}} +%\end{verbatim} +% which produces +% \begin{quotation} +% \MakeTextUppercase{% +% Text in section~\ref{intro}, about \cite[pp 2--4]{bbb}} +% \end{quotation} +% If, instead, the standard |\MakeUppercase| were used here, the keys +% `into' and `bbb' would be uppercased and generate errors about +% undefined references to INTRO and BBB. +% +% Sometimes there may be a special section of text that should not be +% uppercased. This can be marked with |\NoCaseChange|, as follows. +%\begin{verbatim} +% \MakeTextUppercase{% +% Text \NoCaseChange{More Text} yet more text} +%\end{verbatim} +% which produces +% \begin{quotation} +% \MakeTextUppercase{% +% Text \NoCaseChange{More Text} yet more text} +% \end{quotation} +% +% |\NoCaseChange| has other uses. If for some reason you need a +% tabular environment within an uppercased section, then you need +% to ensure that the name `tabular' and the preamble (eg `ll') +% does not get uppercased: +%\begin{verbatim} +% \MakeTextUppercase{% +% Text \NoCaseChange{\begin{tabular}{ll}}% +% table&stuff\\goes&here +% \NoCaseChange{\end{tabular}} +% More text} +%\end{verbatim} +% which produces +% \begin{quotation} +% \MakeTextUppercase{% +% Text \NoCaseChange{\begin{tabular}{ll}}%^^A +% table&stuff\\goes&here +% \NoCaseChange{\end{tabular}} +% More text} +% \end{quotation} +% +% \section{Features and Foibles} +% +% \subsection{Nested text} +% The commands defined here only skip math sections and |\ref| arguments +% if they are not `hidden' inside a |{ }| brace group. All text inside +% such a group will be made uppercase just as with the standard +% |\MakeUppercase|. +%\begin{verbatim} +% \MakeTextUppercase{a b {c $d$} $e$} +%\end{verbatim} +% produces +% \begin{quotation} +% \MakeTextUppercase{a b {c $d$} $e$} +% \end{quotation} +% Of course, this restriction does not apply to the arguments of the +% supported commands |\ensuremath|, |\label|, |\ref|, and |\cite|. +% +% If you cannot arrange for your +% mathematics to be at the outer level of brace grouping, you should +% use the following basic technique (which works even with the standard +% |\MakeUppercase| command). Define a new command that expands to your +% math expression, and then use that command, with |\protect|, in the +% text to be uppercased. Note that if the text being uppercased is in a +% section title or other moving argument you may need to make the +% definition in the document preamble, rather than just before the +% section command, so that the command is defined when the table of +% contents file is read. +%\begin{verbatim} +% \MakeTextUppercase{% +% Text \fbox{$a=b$ and $x=y$}}% +% +% \newcommand{\mathexprone}{$a=b$} +% \newcommand{\mathexprtwo}{$x=y$} +% \MakeTextUppercase{% +% Text \fbox{\protect\mathexprone\ and \protect\mathexprtwo}}% +%\end{verbatim} +% which produces +% \begin{quotation} +% \MakeTextUppercase{% +% Text \fbox{$a=b$ and $x=y$}}% +% +% \newcommand{\mathexprone}{$a=b$} +% \newcommand{\mathexprtwo}{$x=y$} +% \MakeTextUppercase{% +% Text \fbox{\protect\mathexprone\ and \protect\mathexprtwo}}% +% \end{quotation} +% +% \subsection{Citations} +% As documented above, |\cite| and |\ref| commands are not uppercased by +% |\MakeTextUppercase|. If you are using a non-numeric citation scheme +% you may want the replacement text for |\cite| to be uppercased. +% +% It is difficult to arrange that |\MakeTextUppercase| uppercases such +% text, not least because this would lead to interaction with the many +% bibliography packages which redefine |\cite| one way or another. One +% possibility to achieve this is to use Donald Arseneau's cite package +% and to locally redefine |\citeform| to add |\MakeUppercase| around +% the final text string produced by |\cite|. +%\begin{verbatim} +% \MakeTextUppercase{% +% Text \cite{bbb} and \cite{ccc}} +% +% {\renewcommand\citeform{\MakeUppercase}\MakeTextUppercase{% +% Text \cite{bbb} and \cite{ccc}}} +%\end{verbatim} +% which produces\footnote{This is faked, so this document does not +% rely on \texttt{cite.sty} being installed} +% \begin{quotation} +% TEXT [1] AND [David Carlisle 1997] +% +% TEXT [1] AND [DAVID CARLISLE 1997] +% \end{quotation} +% +% \subsection{overload Option} +% By default the package only defines new commands, |\MakeTextUppercase| +% and |\MakeTextlowercase| as described above. You may wish to redefine +% the standard |\MakeUppercase| and |\MakeLowercase| commands to be +% the same as these new commands. So that for example headings in the +% book class have this new feature without any further redefinition. +% You may use the package option |[overload]| in which case these +% new definitions will be overloaded on to the existing command names. +% +% \begin{thebibliography}{9} +% +% \bibitem{bbb} Something numeric +% \bibitem[David Carlisle 1997]{ccc} Something textual +% \end{thebibliography} +% +% \StopEventually{} +% +% \section{Implementation} +% +% \begin{macrocode} +%<*package> +% \end{macrocode} +% +% \begin{macro}{\@uclcnotmath} +% This is the main macro of this package. It is basically +% a copy of |\MakeTextUppercase| and |\MakeTextLowercase| +% from the \LaTeX\ kernel, modified slightly so that they +% can share code (that modification could be done to the +% standard versions as well) and then further changed to +% skip certain features like math mode and |\label| arguments. +% +% The arguments are:\\ +% |#1|: Extra commands to apply for case changing. +% Used to locally redefine |\i| and |\j| for uppercasing.\\ +% |#2|: Either |##1##2| or |##2##1| to control the order +% in which |\let| is applied to the pairs of control +% sequences in |\@uclclist|.\\ +% |#3|: |\uppercase| or |\lowercase|.\\ +% |#4|: The text to be upper (or lower) cased. +% \begin{macrocode} +\def\@uclcnotmath#1#2#3#4{\begingroup +% \end{macrocode} +% Run extra commands (currently just to redefine |\i| and |\j|). +% \begin{macrocode} + #1% +% \end{macrocode} +% +% Locally set |\( \)| to be just |$ $|, so that the math skipping +% code can be simplified, just to look for |$|. +% \begin{macrocode} + \def\({$}\let\)\(% +% \end{macrocode} +% +% Set up the `non-math' things that also have to be skipped. +% \begin{macrocode} + \def\NoCaseChange##1{\noexpand\NoCaseChange{\noexpand##1}}% + \@nonchangecase\label + \@nonchangecase\ref + \@nonchangecase\ensuremath +% \end{macrocode} +% +% |\cite| a bit trickier, as we want to uppercase any optional argument. +% This will fail if the optional argument contains a brace group, but +% should catch most cases. +% +% |text \cite[page 1]{foo} more text| ends up as +%\begin{verbatim} +% \uppercase{text \toks@{\cite[page1]}% +% \the\toks@{foo}% +% \uppercase{ more text} +%\end{verbatim} +% \begin{macrocode} + \def\cite##1##{\toks@{\noexpand\cite##1}\@citex}% + \def\@citex##1{\NoCaseChange{\the\toks@{##1}}}% +% \end{macrocode} +% (|\@citex| is a scratch macro here, not a redefinition of the existing +% |\@citex|.) +% +% The following is essentially taken from |\MakeUppercase|. +% Recursively execute |\reserved@a| to |\let| the pairs in |\@uclclist|. +% The strange construction with |\@gobble| at the end just gobbles the +% final recursive call. +% +% Incidentally, packages should not use the |\reserved@|\ldots\ scratch +% macros, which are `reserved' for use within the \LaTeX\ kernel, but +% (a) this code is essentially a copy from the kernel, and +% (b) I'm allowed to break the rules, so there. +% \begin{macrocode} + \def\reserved@a##1##2{\let#2\reserved@a}% + \expandafter\reserved@a\@uclclist\reserved@b{\reserved@b\@gobble}% +% \end{macrocode} +% +% Expand everything first so that the `skipping' code can see what to +% skip and so that tokens are revealed to |\uppercase|. +% This makes the argument `moving'. +% The |$\valign$| is just a fake math expression used to terminate +% the parsing done by |\@skipmath|. +% \begin{macrocode} + \protected@edef\reserved@a{\endgroup + \noexpand\@skipmath#3#4$\valign$}% +% \end{macrocode} +% +% \begin{macrocode} + \reserved@a} +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\@nonchangecase} +% \begin{macrocode} +\def\@nonchangecase#1{\def#1##1{\NoCaseChange{#1{##1}}}} +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\NoCaseChange} +% For hiding arbitrary text from |\uppercase|. This innocuous +% definition is used for any occurrence of |\NoCaseChange| +% in text that is not passed to |\MakeTextUppercase|. For example a +% section heading may be uppercased, but the toc entry may not. +% It is also used for nested definitions of |\cite| etc, where the +% |\NoCaseChange| is inserted by expansion, but not removed as it is +% hidden by the brace group. +% \begin{macrocode} +\let\NoCaseChange\@firstofone +% \end{macrocode} +% \end{macro} +% +%\begin{macro}{\@skipmath} +% |#1|: operation |\uppercase| / |\lowercase|\\ +% |#2|: text up to first (next) |$|\\ +% |#3|: first math mode material (or |\valign| sentinel) +% \begin{macrocode} +\def\@skipmath#1#2$#3${% + \@skip@nonchangecase#1#2\NoCaseChange\valign + \ifx\valign#3% + \else + $#3$% + \expandafter\@skipmath\expandafter#1% + \fi} +% \end{macrocode} +% \end{macro} +% +%\begin{macro}{\@skip@nonchangecase} +% |#1|: |\uppercase| or |\lowercase|\\ +% |#2|: Text up to the first command (such as |\cite|) made `safe'. +% or the first use of |\NoCaseChange|. +% \begin{macrocode} +\def\@skip@nonchangecase#1#2\NoCaseChange#3{% + #1{#2}% + \ifx\valign#3% + \else + #3% + \expandafter\@skip@nonchangecase\expandafter#1% + \fi} +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\MakeTextUppercase} +% \changes{v0.06}{1998/11/12} +% {Add edef to make behaviour more like \cs{MakeUppercase}} +% Put it all together. Arrange for \i\ and \j\ to uppercase, and +% to read the pairs in the ucase list `forwards'. +% \begin{macrocode} +\DeclareRobustCommand\MakeTextUppercase{% + \@uclcnotmath{\def\i{I}\def\j{J}}{##1##2}\uppercase} +% \end{macrocode} +% +% \begin{macrocode} +\protected@edef\MakeTextUppercase#1{\MakeTextUppercase{#1}} +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\MakeTextLowercase} +% \changes{v0.06}{1998/11/12} +% {Add edef to make behaviour more like \cs{MakeLowercase}} +% probably will never be used, but it is easy to implement. +% Use |\lowercase| and read the pairs in the uppercase list `backwards'. +% \begin{macrocode} +\DeclareRobustCommand\MakeTextLowercase{% + \@uclcnotmath{}{##2##1}\lowercase} +% \end{macrocode} +% +% \begin{macrocode} +\protected@edef\MakeTextLowercase#1{\MakeTextLowercase{#1}} +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\MakeUppercase} +% \begin{macro}{\MakeLowercase} +% \changes{v0.06}{1998/11/12} +% {Option to overload added. latex/2906} +% \begin{macrocode} +\DeclareOption{overload}{% + \expandafter\let\csname MakeUppercase \expandafter\endcsname + \csname MakeTextUppercase \endcsname + \expandafter\let\csname MakeLowercase \expandafter\endcsname + \csname MakeTextLowercase \endcsname} +% \end{macrocode} +% +% \begin{macrocode} +\ProcessOptions +% \end{macrocode} +% \end{macro} +% \end{macro} +% +% \begin{macrocode} +%</package> +% \end{macrocode} +% +% \Finale +% |