%^^A* legal notices
% \iffalse
%
% This program is part of the Frankenstein bundle for LaTeX.
%
% Copyright 1995-2001 Matt Swift <swift@alum.mit.edu>.
%
% This file contains both the code and documentation for the
% titles LaTeX package. It will work ONLY if it is placed in a
% proper directory. Files called README, INSTALL, titles.tex
% and titles.ins should have also been distributed to you
% with this file. See them for more information on how to typeset
% the documentation with LaTeX and how to generate a version of this
% file that will work faster than this one.
%
% This program is free software; you may redistribute it and/or
% modify it under the conditions of the LaTeX Project Public
% License, either version 1.2 or (at your option) any later version.
% The latest version of this license is in
% http://www.latex-project.org/lppl.txt, and version 1.2 or later is
% part of all distributions of LaTeX version 1999/12/01 or later.
%
% This program is distributed in the hope that it will be useful,
% but without any warranty; without even the implied warranty of
% merchantability or fitness for a particular purpose. See the
% LaTeX Project Public License for more details.
%
% \fi
%
%^^A* checks
%
%^^A NOTE: The character table, with two %'s, will get written to all files.
%% \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 \~}
%
% \CheckSum{783}
%
%^^A** abstract
% \begin{abstract}
% The \package{titles} package defines macros that typeset the titles of
% books, journals, etc. and handle following spacing and punctuation
% intelligently, based on context. These are useful for bibliographic
% databases, for example. Also defined is other markup like \cs\word,
% \cs\defn, \cs\phrase, etc.
% \end{abstract}
% \tableofcontents
%
% \part{Discussion}
%
% \section{Options}
%
% There are two package options, \option{british} and \option{american}, the default
% is \option{american}. They select the conventional way to use quotation marks:
% British style is use single quotes, and do not suck following period or comma
% inside; American style is to use double quotes and to suck following period
% or comma inside.
%
% \section{Words and phrases}
%
% \DescribeMacro\word
% Typeset a word or phrase referred to as a noun with \cs\word\marg{word}.
% The argument is not expected to contain punctuation.
%
% \begin{bothexample}
% \word{Elephant} is such a silly word.
% \end{bothexample}
%
% \DescribeMacro\phrase
% Typeset a phrase used as a noun rather than direct quotation with
% \cs\phrase\marg{phrase}. The argument might well have punctuation,
% including final punctuation, which should not be considered to be
% punctuation of the containing sentence.
%
% \begin{bothexample}
% The sentence \phrase{And stop calling me Shirley!} occurs
% twenty-seven times.
% \end{bothexample}
%
% \DescribeMacro{\foreign}
% Typeset a foreign word or phrase with \cs\foreign\marg{foreign text}.
%
% \begin{bothexample}
% I couldn't think of the \foreign{mot juste} at the time.
% \end{bothexample}
%
% \DescribeMacro{\foreignword}
% Typeset a foreign word or phrase referred to as a noun with
% \cs\foreignword\marg{foreign word}.
%
% \begin{bothexample}
% Only later did I realize that the right word was
% \foreignword{bouffon}.
% \end{bothexample}
%
% \caveat {Notice that writing \code{\foreign{\word{text}}} or vice versa is
% not necessarily going to do the right thing. Suppose \cs\foreign and
% \cs\word were both set to \cs\textitswitch (which are in fact the default
% settings below). Then \code{\foreign{\word{text}}} is going to cancel out
% and look just like the surrounding text. This is not the most intuitive
% fact, but it's not worth it to try to make \cs\foreign and \cs\word smart
% enough to see each other inside themselves.}
%
% \DescribeMacro{\term}
% \cs\term\marg{technical term} typesets a techincal term in a different
% font. You might want to use this where a techincal term is first used, or
% defined. One could enhance this macro and \cs\defn to help build an
% automatic glossary
%
% \begin{bothexample}
% This sort of thing is called a \term{blibnil}.
% \end{bothexample}
%
% \DescribeMacro{\defn}
% \cs\defn\marg{definition} typesets a definition, perhaps of a technical
% term. One could enhance this macro and \cs\defn to help build an automatic
% glossary
%
% \begin{bothexample}
% We may describe a \term{blibnil} as \defn{a slibnil with
% three arms}.
% \end{bothexample}
%
% \section{Titles}
%
% \DescribeMacro\book
% \cs\book\marg{book title} typesets a book title.
%
% \begin{bothexample}
% Some people find \book{Moby-Dick} dull, but I thought
% it was exciting.
% \end{bothexample}
%
% \DescribeMacro\journal
% \cs\journal\marg{journal title} typesets a journal title.
%
% \begin{bothexample}
% I liked it so much I started a scholarly journal called
% \journal{The Melville Times} with the inheritance from
% my grandmother.
% \end{bothexample}
%
% \DescribeMacro\music
% \cs\music\marg{music title} typesets a music title.
%
% \begin{bothexample}
% My journal didn't do very well; I moped around my office
% and listened to Schubert's \music{Winterreise}.
% \end{bothexample}
%
% \DescribeMacro\article
% \cs\article\marg{article title} typesets a article title.
%
% \begin{bothexample}
% Then one day I received an article, \article{Pip and
% the Milk of Human Kindness}, by express mail from Wales.
% \end{bothexample}
%
% \DescribeMacro\poemtitle
% \cs\poemtitle\marg{poem title} typesets a poem title.
%
% \begin{bothexample}
% I then wrote my famous poem \poemtitle{Jump for Joy like
% the Butterflies of Troy} in five minutes.
% \end{bothexample}
%
% Sometimes longer poems are distinguished from shorter ones in type when
% they have been published separately as a book [FIX give reference]. This
% package defines a macro \cs\longpoem in the configuration file in the
% following way:
% \begin{codeexample}
% \newlet\longpoem\textitswitch
% \end{codeexample}
%
% \DescribeMacro\play
% \cs\play\marg{play title} typesets a play title.
%
% \begin{bothexample}
% To celebrate the popularity of the article, I took the
% author to the theater to see the acclaimed play
% \play{Grave in Waterloo}, starring Vincent Price.
% \end{bothexample}
%
% \DescribeMacro\craft
% \cs\craft\marg{craft title} typesets a title of a craft or ship.
%
% \begin{bothexample}
% With tears in my heart, I put the author on the \craft{HMS Shangrila}
% bound for Wales.
% \end{bothexample}
%
% \DescribeMacro\species
% \cs\species\marg{[ genus ] species [ subspecies ]} typesets the Latin
% generic and/or specific names for an organism.
%
% \begin{lesson}
% \manual specifies italic type. Genus names should be capitalized, and
% may be abbreviated on subsequent appearances with the initial letter.
% Following designations should be in roman. E.g., ``var.''\ for
% ``variant'' following species name and ``sp.''\ for ``species'' following
% genus name, meaning ``any species in the genus.'' \attrib{7.102--4}
%
% Higher groupings should be in capitalized roman. English derivatives of
% scientific names, e.g., amoeba, are lowercased. \attrib{7.105--6}
%
% \todo{\package{abbrevs}
% category for genus/species and/or datemark for suffixes}
% \end{lesson}
%
% \begin{warning}
% Right now there is a small discrepancy between the behavior of
% \cs\textitswitch and \cs\Wrapquotes regarding what happens when followed by
% a command sequence such as \cs\footnote. I hope to make these things
% completely parallel one day, but for now, realize that after using a
% titling macro that uses \cs\Wrapquotes, you must use |{}| before any
% following command sequence that you want to immediately follow the title
% with no intervening space. The only case I can think of is \cs\footnote.
% If you forget the |{}|, you will have an extra space after the title and before the
% footnotemark. The following example illustrates this behavior and
% contrasts it with \cs\textitswitch:
% \begin{bothexample}
% \newabbrev\foo{Foo}
%
% \book{Foo}\foo
%
% \book{Foo} \foo
%
% \book{Foo}\footnote{footie}
%
% \book{Foo}{}\footnote{footie}
%
% \poemtitle{Foo}\foo
%
% \poemtitle{Foo} \foo
%
% \poemtitle{Foo}\footnote{footie}
%
% \poemtitle{Foo}{}\footnote{footie}
%
% \poemtitle{Foo}.\footnote{footie}
% \end{bothexample}
% \end{warning}
%
% \section{Programmer's interface}
%
% \DescribeMacro\Wrapquotes
% \DescribeMacro\WrapquotesNS
% \DescribeMacro\WrapquotesIS
% \DescribeMacro\WrapquotesNN
% \DescribeMacro\WrapquotesIN
% \DescribeMacro\WrapquotesSN
% \DescribeMacro\WrapquotesDN
% \DescribeMacro\WrapquotesSK
% \cs\Wrapquotes\marg{text} wraps \meta{text} in quotes. Single quotes are
% used initially if the \option{singlequotes} option is given to the
% package, and double quotes if no option or the \option{doublequotes}
% option is given to the package.
%
% When quotation marks inserted by \cs\Wrapquotes and friends are doubled up
% (this occurs sometimes when nesting them), a thinspace (|\,|) is inserted
% between the abutted quotes.
%
% \cs\Wrapquotes will be \cs\let to one of the six macros
% \cs\Wrapquotes\meta{XY}.
%
% In the two-letter suffix \meta{XY}, first letter |N| means ``normal'' and
% |I| means ``inverse.'' These are macros that switch between single and
% double quotes when they nest: an inverse wrapquotes wraps with single
% quotes when a normal wrapquotes would wrap with double quotes, and vice
% versa. First letter |S| for ``single'' and |D| for ``double'' are for
% macros that always wrap with single or double quotes. Spacing and
% punctuation following the closing quotes are handled intelligently by
% macros with second letter |S|, which means means suck a following period
% or comma into the closing quote, that is, if what follows is a comma or
% period, it is pulled inside the quotes (following American practice).
% Second letter |N| means ``nosuck,'' that is, don't suck. Second letter
% |K| means ``kill'': the same as |N| but suppress the effect of any
% punctuation in the quoted argument on spacing that follows the closing
% quotes (i.e., execute \cs\@, which sets the spacefactor to~1000). This is
% only useful in certain technical writing where punctuation in the quoted
% argument should not be considered puncutation of the containing sentence.
%
% A space is inserted after the closing quotes unless what follows is in the
% set |;?:!-)]'\textquoteright{|, in which case no space is inserted.^^A for Emacs: }
% FIX: that would be |\nospacelist|
%
% \DescribeMacro\IfQuestionOrExclamation
% \cs\IfQuestionOrExclamation\marg{text}\marg{true}\marg{false}
% executes \meta{true} clause iff \meta{text} ends with a question mark or
% an exclamation point; executes \meta{false} clause otherwise.
%
% \subsection{Limitations of Wrapquotes and friends}
% ^^A FIX: hmm, can't use \cs or \protect\cs
% \subsubsection{Nesting}
% \begin{warning}
% For proper nesting of \cs\Wrapquotes and friends, user commands must be
% \cs\let to \cs\Wrapquotes or one of the six \cs\Wrapquotes\meta{XY}
% commands instead of using a \cs\def-like defining command. It's OK to
% \cs\let % a user macro to something like \cs\Wrapquotes which itself
% has been % \cs\let to one of the six \cs\Wrapquotes\meta{XY} macros.
%
% The user command which is \cs\let to one of the \cs\Wrapquotes commands
% must furthermore appear in the source. That is, it must not appear as
% the result of an expansion. Among other things, this means that
% nesting won't work properly if you put \cs\Wrapquotes into an abbrev
% (see the \package{abbrevs} package in the \Frankenstein bundle).
%
% For applications where nesting will not occur, there should be nothing
% to worry about.
% \end{warning}
%
% \subsubsection{Italic corrections}
%
% \begin{warning}
% The question of when to insert an italic correction is not nearly as
% simple as it might seem. I cannot figure good rules which cover all
% cases, and I do not trust the behavior of the kernel's macros as a
% guide. So I can not tell you whether this package handles italic
% corrections properly. If you discover behavior which you think is
% wrong, please let me know with an example and an argument.
% \end{warning}
%
% \subsubsection{A slight bug}
% \begin{warning}
% Right now there is a small bug in cases where closing quotes fall at
% the end of italic text, such as
% \begin{bothexample}
% \normalfont
% \book{My love of \poemtitle{Daffodils}}, by H.~Moneysworth.
% \end{bothexample}
% These cases loose because the closing quotation marks and any
% sucked-in punctuation are going to be in roman, not italic, or italic,
% not roman. Only the more obsessive will notice this flaw. I'm sure I
% will come up with a way to handle this for a future version of this
% package.
% \end{warning}
%
% \StopEventually{}
%
% \part{Implementation}
%
% \section{Version control}
%
% \begin{macro}{\fileinfo}
% \begin{macro}{\DoXUsepackagE}
% \begin{macro}{\HaveECitationS}
% \begin{macro}{\fileversion}
% \begin{macro}{\filedate}
% \begin{macro}{\docdate}
% \begin{macro}{\PPOptArg}
% These definitions must be the first ones in the file.
% \begin{macrocode}
\def\fileinfo{title macros (Frankenstein's references)}
\def\DoXPackageS {}
\def\initelyHavECitationS {}
\def\fileversion{v1.2}
\def\filedate{2001/08/31}
\def\docdate{2001/08/31}
\edef\PPOptArg {%
\filedate\space \fileversion\space \fileinfo
}
% \end{macrocode}
% \end{macro}
% \end{macro}
% \end{macro}
% \end{macro}
% \end{macro}
% \end{macro}
% \end{macro}
%
% If we're loading this file from a \cs\ProcessDTXFile command (see the
% \package{compsci} package), then \cs\JusTLoaDInformatioN will be defined;
% othewise we assume it is not (that's why the FunkY NamE).
%
% If we're loading from \cs\ProcessDTXFile, we want to load the packages listed
% in \cs\DoXPackageS (needed to typeset the documentation for this file) and
% then bail out. Otherwise, we're using this file in a normal way as a
% package, so do nothing. \cs\DoXPackageS, if there are any, are declared in
% the \ext{dtx} file, and, if you're reading the typeset documentation of this
% package, would appear just above. (It's OK to call \cs\usepackage with an
% empty argument or \cs\relax, by the way.)
% \begin{macrocode}
\makeatletter% A special comment to help create bst files. Don't change!
\@ifundefined{JusTLoaDInformatioN} {%
}{% ELSE (we know the compsci package is already loaded, too)
\UndefineCS\JusTLoaDInformatioN
\SaveDoXVarS
\eExpand\csname DoXPackageS\endcsname\In {%use \csname in case it's undefined
\usepackage{#1}%
}%
\RestoreDoXVarS
\makeatother
\endinput
}% A special comment to help create bst files. Don't change!
% \end{macrocode}
%
% Now we check for \LaTeX2e and declare the LaTeX package.
% \begin{macrocode}
\NeedsTeXFormat{LaTeX2e}
\ProvidesPackage{titles}[\PPOptArg]
% \end{macrocode}^^A special comment to help create bst files. Don't change!
%
% ^^A NOTE: We have to compensate for the above backslashes, which are not
% ^^A actually in the .dtx file the author works on, by adding to the
% ^^A CheckSum.
%%
% \AddToCheckSum{17}^^A `dtx-update-checksum' automatically handles this.
% \AddToCheckSum{7}^^A The half a macrocode env. at the top is missed, however...
% \AddToCheckSum{10}^^A ... and so are the 5 \defs from the .dtx file
% ^^A that precede it.
% \IfCitations {%
% \AddToCheckSum{2}^^A When \initelyHavECitationS is defined in
% } ^^A the .dtx file, we need 2 more in the CheckSum.
%
%
% \section{Requirements}
%
% \begin{macrocode}
\RequirePackage{moredefs,slemph}
% \end{macrocode}
%
% \section{Options}
%
% \begin{macro}{\ti@domelater}
% \begin{macrocode}
\ReserveCS\ti@domelater
\DeclareOption{british} {%
\def\ti@domelater {%
\let\Wrapquotes\WrapquotesNN
\@doublequotes@false
}
}
\DeclareOption{american} {%
\def\ti@domelater {%
\let\Wrapquotes\WrapquotesNS
\@doublequotes@true
}
}
\ExecuteOptions{american}
\ProcessOptions
% \end{macrocode}
% \end{macro}
%
% \section{Wrapquotes}
%
% Here we go! This is not a picnic, so leave your jelly jar home.
%
% \subsection{Titles that are Questions or Exclamations}
%
% \begin{macro}{\IfQuestionOrExclamation}
% \begin{macro}{\ti@checkfor@q}
% \begin{macro}{\ti@checkfor@e}
% \begin{macro}{\ti@prev}
% \begin{macro}{\ti@prev@prev}
% \begin{macro}{\@ti@sw@true}
% \begin{macro}{\@ti@sw@false}
% \begin{macro}{\if@ti@sw@}
% \mbox{}
% \begin{macrocode}
\newcommand\IfQuestionOrExclamation [1] {%
\@tempswafalse
\ti@checkfor@q #1?\@nil
\ti@checkfor@e #1!\@nil
\if@tempswa
\expandafter\@firstoftwo
\else
\expandafter\@secondoftwo
\fi
}
% \end{macrocode}
% The large majority of titles will not contain a question mark or exclamation
% point. The large majority of those that do will have a single mark or point
% at the end. We could (I think) use a simpler check that processed all titles
% by looping through to examine the end, but a slightly more complicated check
% will handle the majority of cases very quickly (and at a constant speed,
% rather than proportional to title length) and not greatly slow down
% processing the remaining two unusual cases. We divide our argument (with an
% extra question mark tacked onto the end) into what's before the first
% question mark and what's after it. Then we examine what's after it and
% interpret the results thus:
% \begin{description}
% \item [empty] no question mark in title
% \item [question mark] title ends with question mark (and there are no other
% question marks)
% \item [text ending with one question mark] a question mark occurs in the
% title, but not at the end
% \item [text ending with two question marks] title ends with a question mark
% (and there is a previous question mark)
% \end{description}
% We set switch |a| to true if the title ends with a question mark.
% \begin{macrocode}
\newboolean{@ti@sw@}
\ReserveCS\ti@prev
\ReserveCS\ti@prev@prev
\NewName{ti@checkfor@q} {#1?#2\@nil} {%
\def\sc@t@a{#2}%
\def\sc@t@b{?}%
\ifx\sc@t@a\ShortEmpty
\else
\ifx\sc@t@a\sc@t@b
\@tempswatrue
\else
% \end{macrocode}
% We use a loop to whittle down |#2| until \cname{ti@prev} contains the last
% character and \cname{ti@prev@prev} contains the second-to-last. We know that
% \cname{ti@prev} is going to be a question mark. Iff \cname{ti@prev@prev} is a
% question mark, we are in the final case above.
% \begin{macrocode}
\let\ti@prev\sc@t@a
\let\ti@prev@prev\sc@t@a
\@ti@sw@true
\@whilesw \if@ti@sw@ \fi {%
\ifx\sc@t@a\ShortEmpty
\@ti@sw@false
\else
\let\ti@prev@prev\ti@prev
\let\ti@prev\sc@t@a
\edef\sc@t@a{\E@cdr\sc@t@a\@nil}%
\fi
}%
\edef\ti@prev@prev{\E@car\ti@prev@prev\@nil}%
\ifx\ti@prev@prev\sc@t@b
\@tempswatrue
\fi
\fi
\fi
}
% \end{macrocode}
% Exact same logic applies to exclamation points.
% \begin{macrocode}
\NewName{ti@checkfor@e} {#1!#2\@nil} {%
\def\sc@t@a{#2}%
\def\sc@t@b{!}%
\ifx\sc@t@a\ShortEmpty
\else
\ifx\sc@t@a\sc@t@b
\@tempswatrue
\else
\let\ti@prev\sc@t@a
\let\ti@prev@prev\sc@t@a
\@ti@sw@true
\@whilesw \if@ti@sw@ \fi {%
\ifx\sc@t@a\ShortEmpty
\@ti@sw@false
\else
\let\ti@prev@prev\ti@prev
\let\ti@prev\sc@t@a
\edef\sc@t@a{\E@cdr\sc@t@a\@nil}%
\fi
}%
\edef\ti@prev@prev{\E@car\ti@prev@prev\@nil}%
\ifx\ti@prev@prev\sc@t@b
\@tempswatrue
\fi
\fi
\fi
}
% \end{macrocode}
% \end{macro}
% \end{macro}
% \end{macro}
% \end{macro}
% \end{macro}
% \end{macro}
% \end{macro}
% \end{macro}
%
% \subsection{Highlevel macros}
%
% \begin{macro}{\ti@wrapquotes@suck}
% \begin{macro}{\ti@wrapquotes@nosuck}
% These two are the top-level internal macros, and they are pretty sane. One
% sucks in a following period or comma, the other does not.
% \cname{ti@wrapquotes@suck} does not suck, however, when the title ends in a
% question or exclamation point.
%
% The group here is necessary to scope the \cname{@doublequotes@} boolean.
% \begin{macrocode}
\newcommand*\ti@wrapquotes@suck [1] {%
% \DTypeout{top of wrapquotes@suck}%
\IfQuestionOrExclamation {#1} {%
\ti@wrapquotes@nosuck{#1}%
}{% ELSE
% \DTypeout{top of wrapquotes@suck ELSE}%
\begingroup
\if@doublequotes@
% \DTypeout{double true in suck}%
\@doublequotes@false
\def\sc@t@a {\ti@open@double #1\ti@close@double@suck}%
\else
% \DTypeout{double false in suck}%
\@doublequotes@true
\def\sc@t@a {\ti@open@single #1\ti@close@single@suck}%
\fi
\sc@t@a
\endgroup
}%
}
\newcommand*\ti@wrapquotes@nosuck [1] {%
\begingroup
\if@doublequotes@
% \DTypeout{double true in nosuck}%
\@doublequotes@false
\def\sc@t@a {\ti@open@double #1\ti@close@double@nosuck}%
\else
% \DTypeout{double false in nosuck}%
\@doublequotes@true
\def\sc@t@a {\ti@open@single #1\ti@close@single@nosuck}%
\fi
\sc@t@a
\endgroup
}
% \end{macrocode}
% \end{macro}
% \end{macro}
%
% \begin{macro}{\WrapquotesNS}
% \begin{macro}{\WrapquotesIS}
% \begin{macro}{\WrapquotesNN}
% \begin{macro}{\WrapquotesIN}
% \begin{macro}{\WrapquotesSN}
% \begin{macro}{\WrapquotesDN}
% \begin{macro}{\WrapquotesSK}
% \begin{macro}{\Wrapquotes}
% \begin{macro}{\if@doublequotes@}
% \begin{macro}{\@doublequotes@true}
% \begin{macro}{\@doublequotes@false}
% Now we can define the secondary programmers' macros.
%
% We simply reserve \cs\Wrapquotes here, and assign it in the user options
% section above.
% \begin{macrocode}
\newboolean{@doublequotes@}
\newcommand*\WrapquotesNS {%
% \DTypeout{starting wrapquotes NS}%
\ti@wrapquotes@suck
}
\newcommand*\WrapquotesIS {%
% \DTypeout{starting wrapquotes IS}%
\ToggleBoolean{@doublequotes@}%
\ti@wrapquotes@suck
}
\newcommand*\WrapquotesNN {%
% \DTypeout{starting wrapquotes NN}%
\ti@wrapquotes@nosuck
}
\newcommand*\WrapquotesIN {%
% \DTypeout{starting wrapquotes IN}%
\ToggleBoolean{@doublequotes@}%
\ti@wrapquotes@nosuck
}
\newcommand*\WrapquotesSN [1] {%
% \DTypeout{starting wrapquotes SN}%
\begingroup
\ti@open@single #1\ti@close@single@nosuck
\endgroup
}
\newcommand*\WrapquotesDN [1] {%
% \DTypeout{starting wrapquotes DN}%
\begingroup
\ti@open@double #1\ti@close@double@nosuck
\endgroup
}
\newcommand*\WrapquotesSK [1] {% FIX: test
% \DTypeout{starting wrapquotes SK}%
\begingroup
\ti@open@single #1\ti@close@single@nosuck\@%
\endgroup
}
\ReserveCS\Wrapquotes
\ti@domelater
% \end{macrocode}
% \end{macro}
% \end{macro}
% \end{macro}
% \end{macro}
% \end{macro}
% \end{macro}
% \end{macro}
% \end{macro}
% \end{macro}
% \end{macro}
% \end{macro}
%
% \subsection{Opening quotes}
%
% \begin{macro}{\ti@open@double}
% \begin{macro}{\ti@open@single}
% \begin{macro}{\ti@openquote}
% We start by putting an opening mark in scratch |f| with a global
% definition.
%
% I can't remember why it's global. In the macros that close quotes, we want
% to keep that information around past a group end because we're using
% \cs\aftergroup, but that doesn't seem to apply for opening them. Best not
% to change what's not broke, however.
% \begin{macrocode}
\newcommand\ti@open@double {%
\gdef\sc@t@f {\textquotedblleft}%
\ti@openquote
}
\newcommand\ti@open@single {%
\gdef\sc@t@f {\textquoteleft}%
\ti@openquote
}
% \end{macrocode}
% Then we look ahead with scratch |a|. We are looking ahead at the first
% character of the contents of the \cs\Wrapquotes.
% \begin{macrocode}
\newcommand\ti@openquote {%
\futurelet\sc@t@a\ti@@openquote
}
% \end{macrocode}
% Insert the opening mark. Then, if we are about to open another quote, insert
% the space appropriate to separate contiguous quotation marks.
% \begin{macrocode}
\newcommand\ti@@openquote {%
\sc@t@f
\ifx\sc@t@a\WrapquotesNS
% \DTypeout{Quotation marks are doubled up (next is NS); inserting padding.}%
\,%
\else \ifx\sc@t@a\WrapquotesNN
% \DTypeout{Quotation marks are doubled up (next is NN); inserting padding.}%
\,%
\else \ifx\sc@t@a\WrapquotesIN
% \DTypeout{Quotation marks are doubled up (next is IN); inserting padding.}%
\,%
\else \ifx\sc@t@a\WrapquotesIS
% \DTypeout{Quotation marks are doubled up (next is IS); inserting padding.}%
\,%
\else \ifx\sc@t@a\WrapquotesSN
% \DTypeout{Quotation marks are doubled up (next is SN); inserting padding.}%
\,%
\else \ifx\sc@t@a\WrapquotesDN
% \DTypeout{Quotation marks are doubled up (next is DN); inserting padding.}%
\,%
\else \ifx\sc@t@a\WrapquotesSK
% \DTypeout{Quotation marks are doubled up (next is SK); inserting padding.}%
\,%
\else
\fi \fi \fi \fi \fi \fi \fi
}
% \end{macrocode}
% \end{macro}
% \end{macro}
% \end{macro}
%
% \subsection{Closing macros that don't suck}
%
% This case that doesn't suck is easier, so we warm up with it.
%
% \begin{macro}{\ti@close@single@nosuck}
% \begin{macro}{\ti@close@double@nosuck}
% \begin{macro}{\ti@close@single@@nosuck}
% \begin{macro}{\ti@close@double@@nosuck}
% \mbox{}
% \begin{macrocode}
\newcommand*\ti@close@single@nosuck {%
\aftergroup\ti@close@single@@nosuck
}
\newcommand*\ti@close@double@nosuck {%
\aftergroup\ti@close@double@@nosuck
}
\newcommand*\ti@close@single@@nosuck {%
\gdef\sc@t@f {\textquoteright}%
\ti@close@quote@nosuck
}
\newcommand*\ti@close@double@@nosuck {%
\gdef\sc@t@f {\textquotedblright}%
\ti@close@quote@nosuck
}
% \end{macrocode}
% \end{macro}
% \end{macro}
% \end{macro}
% \end{macro}
%
% \begin{macro}{\ti@close@quote@nosuck}
% \begin{macro}{\if@look@nosuck@}
% \begin{macro}{\@look@nosuck@true}
% \begin{macro}{\@look@nosuck@false}
% \todo{Document this flag. It's a hack, we must set it before each call
% to \cname{ti@q@ifnextcharin} I think. What it stands for is something
% like the presence of the tokens \cname{ti@close@single@nosuck} and
% \cname{ti@close@double@nosuck} in the list of chars to look for, but since
% they aren't really chars they can't go in the list, so instead we set the
% flag. Somewhat cleaner would be putting a flag char in the list, but I
% can't think of what char I could safely use.}
% \begin{macrocode}
\newboolean{@look@nosuck@}
\@look@nosuck@false
\newcommand\ti@close@quote@nosuck {%
% \DTypeout{Starting ti@close@quote@nosuck}%
\@look@nosuck@true
% \end{macrocode}
%% FIX
% Aha, but here is a good reason to leave in |.,| in our substitute for \cs\nospacelist.
% \begin{macrocode}
\expandafter \ti@q@ifnextcharin \expandafter {\nospacelist} {%
% \DTypeout{Found a nosuck no-spacer. C=[\meaning\sc@t@c] F=[\meaning\sc@t@f]}%
\sc@t@f
}{% ELSE
% \DTypeout{Found a nosuck spacer. C=[\meaning\sc@t@c] F=[\meaning\sc@t@f]}%
\sc@t@f\space
}%
}
% \end{macrocode}
% \end{macro}
% \end{macro}
% \end{macro}
% \end{macro}
%
% \subsection{Closing macros that suck}
%
% \begin{macro}{\ti@close@double@suck}
% \begin{macro}{\ti@close@single@suck}
% We need to look ahead beyond the \cs\endgroup that ends
% \cname{Wrapquotes??}. The lookahead mechanism that gets invoked in scratch
% |a| below could handle looking past the \cs\endgroup, but I think it is
% more efficient to skip it by using \cs\aftergroup.
% \begin{macrocode}
\newcommand\ti@close@double@suck {%
\aftergroup\ti@close@double@@suck
}
\newcommand\ti@close@single@suck {%
\aftergroup\ti@close@single@@suck
}
% \end{macrocode}
% \end{macro}
% \end{macro}
%
% \begin{macro}{\ti@close@double@@suck}
% \begin{macro}{\ti@close@single@@suck}
% This part isn't so bad yet. To close the quotes, we again start with the
% closing mark in scratch |f|, with a global definition.
% \begin{macrocode}
\newcommand\ti@close@double@@suck {%
\gdef\sc@t@f {\textquotedblright}%
\ti@close@quote@suck
}
\newcommand\ti@close@single@@suck {%
\gdef\sc@t@f {\textquoteright}%
\ti@close@quote@suck
}
% \end{macrocode}
% \end{macro}
% \end{macro}
%
% \begin{macro}{\nospacelist}
% Put these in the order of their frequency. Anything in \cs\nocorrlist
% should also be in here, most likely. I'm putting in \cname{@xobeysp}
% because it's in the \package{xspace} package, but I can't tell you when it
% would come up.
% \begin{macrocode}
\requirecommand\nospacelist {%
,.':;?-/\slash~!)]\bgroup\egroup\@sptoken\ \space\/\@xobeysp
}
% \end{macrocode}
% \end{macro}
%
% \begin{macro}{\ti@close@quote@suck}
% Then we use \cname{ti@q@ifnextcharin} to look as far ahead as necessary for
% a significant character. The latest significant character found is
% available in scratch |c|. The work of handling all the cases of what we
% might find while looking ahead is divided up between
% \cname{ti@close@quote@suck} and \cname{ti@q@ifnextcharin}.
% \cname{ti@close@quote@suck} handles the last step in the process, and
% \cname{ti@q@ifnextcharin} handles all the steps up to the last.
%
% Here is what \cname{ti@close@quote@suck} does, in English. If we find a
% comma or period, we put it inside the closing quote, and gobble the one we
% found. That is, we print out scratch |c|, then scratch |f|, then gobble a
% character. If we find something in the set given in \cs\nospacelist, do
% not leave a space after the closing mark. That is, just print out scratch
% |f|. If we find something else, we leave a space after the closing mark.
% That is, print scratch |f| and a space.
% \begin{macrocode}
\newcommand\ti@close@quote@suck {%
% \DTypeout{Starting ti@close@quote@suck}%
\@look@nosuck@false
\ti@q@ifnextcharin {.,} {%
% \DTypeout{Found a comma or period. C=[\meaning\sc@t@c] F=[\meaning\sc@t@f]}%
\sc@t@c\sc@t@f\DGobbleM % This gobbles the original punctuation.
}{% ELSE
% \DTypeout {Before second ti@qifnextcharin. C=[\meaning\sc@t@c] F=[\meaning\sc@t@f]}%
\@look@nosuck@true
% \end{macrocode}
% \begin{todoenv}
% Using \cs\nospacelist is inefficient here, since some of the cases,
%^^A for Emacs {
% namely \verb*|,.\@sptoken}|, are never going to be there and shouldn't be
% checked for, since they are passed over by \cname{ti@q@ifnextcharin} before
% the list is compared. But it would be good to have this parallelism between
% \package{abbrevs} and \package{titles}.
% \end{todoenv}
% \begin{macrocode}
\expandafter \ti@q@ifnextcharin \expandafter {\nospacelist} {%
% \DTypeout{Found a suck no-spacer. C=[\meaning\sc@t@c] F=[\meaning\sc@t@f]}%
\sc@t@f
}{% ELSE
% \DTypeout{Found a suck spacer. C=[\meaning\sc@t@c] F=[\meaning\sc@t@f]}%
\sc@t@f\space
}%
}%
}
% \end{macrocode}
% \end{macro}
%
% \subsection{Looking ahead}
%
% Now things are getting fun.
%
% \begin{macro}{\ti@q@ifnextcharin}
% \begin{macro}{\ti@q@check}
% \begin{macro}{\ti@q@ifnch}
% \begin{macro}{\ti@q@@ifnch}
% These macros are modeled after the definition of \cname{@ifnextchar} which
% skips spaces. While looking ahead for the next significant character,
% these macros skip spaces, \cs\egroup, \cs\endgroup, \cname{check@icr},
% \cname{ti@close@double} and \cname{ti@close@single} while doing the right
% thing after each.
%
% The first argument should be a list of tokens. If the next significant
% char is in the list, then the |true| clause is executed, otherwise the
% |false| clause is executed. The next significant char is left in scratch
% |c| so it can be accessed by the clauses.
%
% The three arguments to \cname{ti@q@ifnextcharin} are saved in global
% variables because while looking ahead we must continue past the ends of
% groups.
%
%% FIX
% Not sure I need gdef for scratch |e|.
% \begin{macrocode}
\newcommand\ti@q@ifnextcharin [3] {% args: charlist true false
% \sc@toks@a{#1}%
% \DTypeout{charlist unexpanded is =[\the\sc@toks@a]}%
\gdef\sc@t@e {#1}%
\gdef\sc@t@a {#2}%
\gdef\sc@t@b {#3}%
\ti@q@check
}
% \end{macrocode}
% Having saved the arguments, we look ahead with scratch |c|. This step is not
% in the macro above so that we can jump back to \cname{ti@q@check} whenever we
% want to look ahead another character.
% \begin{macrocode}
\newcommand\ti@q@check {%
\futurelet\sc@t@c\ti@q@ifnch
}
% \end{macrocode}
% Scratch |c| contains the current char. Scratch |d| is the action to take at
% the end of this macro. We attempt to order these possibilities to make
% \cs\Wrapquotes most efficient, though it is a guess which items will be
% encountered most frequently.
%
% The actions taken for each of the possibilities are the following:
% \begin{description}
% \item[\cs\ifvmode] Assume that the \cs\Wrapquotes was the argument of a
% \cs\TextFontCommand from certain \LaTeX{} kernels. Gobble three more tokens
% expected to follow the \cs\ifvmode, execute them, and continue on to look
% ahead another character. See documentation of
% \cname{ti@q@handle@ifvmode} for more details.
% \item[\cname{check@icr}] This means the \cs\Wrapquotes was the
% argument of a \cs\TextFontCommand. Gobble the \cname{check@icr}
% and look ahead another character after we exit the
% group that the \cs\TextFontCommand has given us.
% \item[\cs\endgroup and {\normalfont\ttfamily \}}] Pass right by an \cs\endgroup or
% |}| and look ahead another char.
% \item[\cname{@sptoken} (a non-explicit space)] Handle a non-explicit space by calling
% \cname{ti@q@handle@space}, which gobbles the space and looks ahead
% another char. When the user or a macro has followed the titles with an
% \emph{explicit} space such as a tie, or the
%% \verb*|\ |
% or \cs\space macros, we do nothing and let this be caught by the
% comparison to the tokens in the argument of \cname{ti@q@ifnextcharin}.
% \item[\cname{ti@close@double@suck},
% \cname{ti@close@single@suck}]
% \item[\cname{ti@close@double@nosuck}, and
% \cname{ti@close@single@nosuck}] We are in a nested
% \cs\Wrapquotes. Call \cname{ti@q@handle@single/double@suck/nosuck}
% as appropriate, which
% gobbles the closequotes token, adds properly-padded closing quotes to
% scratch |f|, and then goes on to look ahead another character.
% \end{description}
% The lookahead process stops when it finds something not on this list. Then
% it compares what it found to the list of characters given to
% \cname{ti@q@ifnextcharin} and executes the |true| or |false| clause as
% appropriate.
%
% First we have to handle the case of finding an \cs\ifvmode. We can't bundle
% this test in with the tests for other tokens, so it gets its own macro,
% \cname{ti@q@handle@ifvmode}, which see for details. The remaining cases are
% handled in \cname{ti@q@@ifnch}.
% \begin{macrocode}
\newcommand\ti@q@ifnch {%
% \DTypeout{The lookahead in ti@q@ifnch: [\meaning\sc@t@c]}%
\ifx\sc@t@c\ifvmode
\let\sc@t@d\ti@q@handle@ifvmode
\else
\let\sc@t@d\ti@q@@ifnch
\fi
\sc@t@d
}
\newcommand\ti@q@@ifnch {%
% \DTypeout{entering ti@q@@ifnch}%
% \expandafter\sc@toks@a\expandafter{\sc@t@c}%
% \DTypeout{ti@q@@ifnch: C expanded once is =[\the\sc@toks@a]}%
\ifx\sc@t@c\check@icr
% \DTypeout{Handling check@icr}%
\defcommand\sc@t@d [1] {%
% \DTypeout{check@icr handler: gobbling [\meaning ##1]}%
##1\aftergroup\ti@q@check
}%
\else \ifx\sc@t@c\endgroup
% \DTypeout{Handling endgroup}%
\def\sc@t@d {\aftergroup\ti@q@check}%
\else \ifx\sc@t@c\@sptoken
% \DTypeout{Handling space}%
\let\sc@t@d\ti@q@handle@space
\else \ifx\sc@t@c\egroup
% \DTypeout{Handling egroup}%
\def\sc@t@d {\aftergroup\ti@q@check}%
\else \ifx\sc@t@c\ti@close@double@suck
% \DTypeout{Handling ti@close@double@suck}%
\let\sc@t@d\ti@q@handle@double@suck
\else \ifx\sc@t@c\ti@close@single@suck
% \DTypeout{Handling ti@close@single@suck}%
\let\sc@t@d\ti@q@handle@single@suck
\else \ifx\sc@t@c\ti@close@double@nosuck
% \DTypeout{Handling ti@close@double@nosuck}%
\let\sc@t@d\ti@q@handle@double@nosuck
\else \ifx\sc@t@c\ti@close@single@nosuck
% \DTypeout{Handling ti@close@single@nosuck}%
\let\sc@t@d\ti@q@handle@single@nosuck
\else
% \end{macrocode}
% We've handled all the lookahead cases, so now we are left with the simple
% comparison of the next char with the charlist.
% \begin{macrocode}
\@tempswafalse
\expandafter \@tfor
\expandafter \sc@t@g
\expandafter :%
\expandafter =%
\sc@t@e
\do {%
\expandafter\ifx\sc@t@g\sc@t@c
% \DTypeout{We have a match of [\meaning\sc@t@c]
% with [\expandafter\meaning\sc@t@g]}%
\@tempswatrue
\@break@tfor
\else
% \DTypeout{We have NO match between [\meaning\sc@t@c]
% with [\meaning\sc@t@g]}%
\fi
}%
\if@tempswa
% \DTypeout{Choosing true clause [\meaning\sc@t@a]}%
\let\sc@t@d\sc@t@a % the ``true'' clause
\else
% \DTypeout{Choosing false clause [\meaning\sc@t@b]}%
\let\sc@t@d\sc@t@b % the ``false'' clause
\fi
\fi \fi \fi \fi \fi \fi \fi \fi
% \DTypeout{About to fall out of ti@q@@ifnch and do this [\meaning\sc@t@d]}%
\sc@t@d
}
% \end{macrocode}
% \end{macro}
% \end{macro}
% \end{macro}
% \end{macro}
%
% \begin{macro}{\ti@q@handle@ifvmode}
% This is in its own macro for clarity and to avoid problems with skipping
% over clauses.
%
% \cname{ti@q@ifnch} has to take two different kinds of \LaTeX{} kernel into
% account. The 1996/12/01 and 1997/06/01 kernels used a different definition
% of \cs\DeclareTextFontCommand:
% \begin{macrocode}
%\def \DeclareTextFontCommand #1#2{%
% \DeclareRobustCommand#1[1] {%
% \ifmmode
% \nfss@text{#2##1}%
% \else
% \leavevmode
% {\text@command{##1}%
% #2\check@icl ##1\ifvmode\else\check@icr\fi
% \expandafter}%
% \fi
% }%
%}
% \end{macrocode}
% All other kernels leave out the check for vertical mode (kernels from
% 1997/12/01 include it when necessary inside \cname{check@icr}). The macro
% \cname{ti@q@ifnch}, which will be called immediately before this point of
% difference, handles both cases by looking for both \cs\ifvmode and
% \cname{check@icr}. For the history, see \LaTeX{} bug report 2646.
%
% The check for \cs\ifvmode must not be part of a nested conditional. \TeX{}
% can't match \cs\if{}s with \cs\fi{}s properly when you nest tests for
% \cs\if-type tokens. See p.~211 of the \TeX{}book.
%
% When we encounter an \cs\ifvmode, we must assume we are inside a
% TextFontCommand declared by one of the two kernel versions mentioned above.
% If not, we are in an unknown situation and we will bomb. Since the error
% message in this case won't be helpful, we warn the user in the log file.
% We use scratch |d| to gobble both the \cs\ifvmode and what we expect will
% follow the \cs\ifvmode, namely \code{\else\check@icr\fi}. After swallowing
% those, we reissue those same commands and then proceed with our lookahead.
% We want to issue those commands, which conditionally introduce an italic
% correction, before looking further ahead.
% \begin{macrocode}
\newcommand\ti@q@handle@ifvmode {%
% \DTypeout{Handling ifvmode}%
\FrankenInfo{titles}
{Handling an \string\ifvmode\space following a title.\MessageBreak
If you now get an error that \string\sc@t@d\space does not\MessageBreak
match its definition, this \string\ifvmode\space is\MessageBreak
unexpected}%
\DefName{sc@t@d} {\ifvmode\else\check@icr\fi} {%
\ifvmode
\else
\check@icr
\fi
\aftergroup\ti@q@check
}%
\sc@t@d
}
% \end{macrocode}
% \end{macro}
%
% \begin{macro}{\ti@q@handle@space}
% Handle the case of a following space: gobble the space and call
% \cname{ti@q@check}.
%
% This little bit of trickery sneaks a space in as the \cs\def template,
% thereby causing a space following \cname{ti@q@handle@space} to get gobbled.
% We use the control character \cname{:} and restore its value.
% \begin{macrocode}
\ReserveCS\ti@q@handle@space
\let\sc@t@a\:
\def\:{\ti@q@handle@space} \expandafter\def\: {\ti@q@check}
\let\:\sc@t@a
% \end{macrocode}
% \end{macro}
%
% \begin{macro}{\ti@q@handle@single@suck}
% \begin{macro}{\ti@q@handle@double@suck}
% Handle the single and double sucking cases: gobble the closequotes token
% with a \cs\def template, add some stuff to scratch |f| and call
% \cname{ti@q@check}. These are put in their own macros only to avoid
% clutter above.
% \begin{macrocode}
\newcommand*\ti@q@handle@double@suck [1] {%
% \DTypeout{handle double suck: gobbling [\meaning#1]}%
% \DTypeout{scratch f before: [\meaning\sc@t@f]}%
\g@addto@macro\sc@t@f {\,\textquotedblright}%
% \DTypeout{scratch f after: [\meaning\sc@t@f]}%
\ti@q@check
}
\newcommand*\ti@q@handle@single@suck [1] {%
% \DTypeout{handle single suck: gobbling [\meaning#1]}%
% \DTypeout{scratch f before: [\meaning\sc@t@f]}%
\g@addto@macro\sc@t@f {\,\textquoteright}%
% \DTypeout{scratch f after: [\meaning\sc@t@f]}%
\ti@q@check
}
% \end{macrocode}
% \end{macro}
% \end{macro}
%
% \begin{macro}{\ti@q@handle@single@nosuck}
% \begin{macro}{\ti@q@handle@double@nosuck}
% Handle the single and double nosucking cases. Add inter-quote space to
% scratch |f| and exit \cname{ti@q@@ifnch} with |true| or |false| depending
% on whether we were looking for it. We had to do it this way instead of the
% normal \cname{if} test above at the end of \cname{ti@q@@ifnch} because
% \cname{ti@close@double@nosuck} is more than one character long.
% \begin{macrocode}
\newcommand*\ti@q@handle@double@nosuck [1] {%
% \DTypeout{handle double nosuck: gobbling [\meaning#1]}%
\if@look@nosuck@
% \DTypeout{And we're looking for \string\ti@close@double@nosuck.}%
\g@addto@macro\sc@t@f {\,\textquotedblright}%
% \DTypeout{After adding padding, F=[\meaning\sc@t@f]}%
\let\sc@t@d\sc@t@a % the ``true'' clause of ti@q@ifnextcharin
\else
% \DTypeout{But we're not looking for \string\ti@close@double@nosuck.}%
% \DTypeout{F is unchanged, F=[\meaning\sc@t@f]}%
\let\sc@t@d\sc@t@b % the ``false'' clause of ti@q@ifnextcharin
\fi
\ti@q@check
}
\newcommand*\ti@q@handle@single@nosuck [1] {%
% \DTypeout{handle single nosuck: gobbling [\meaning#1]}%
\if@look@nosuck@
% \DTypeout{And we're looking for \string\ti@close@single@nosuck.}%
\g@addto@macro\sc@t@f {\,\textquoteright}%
% \DTypeout{After adding padding, F=[\meaning\sc@t@f]}%
\let\sc@t@d\sc@t@a % the ``true'' clause of ti@q@ifnextcharin
\else
% \DTypeout{But we're not looking for \string\ti@close@single@nosuck.}%
% \DTypeout{F is unchanged, F=[\meaning\sc@t@f]}%
\let\sc@t@d\sc@t@b % the ``false'' clause of ti@q@ifnextcharin
\fi
\ti@q@check
}
% \end{macrocode}
% \end{macro}
% \end{macro}
%
% \section{Words and phrases}
%
% \begin{macro}{\word}
% \begin{macro}{\foreign}
% \begin{macro}{\foreignword}
% \begin{macro}{\phrase}
% \begin{macro}{\term}
% \begin{macro}{\defn}
% \mbox{}
% \begin{macrocode}
\newlet\word\textitswitch
\newlet\foreign\textitswitch
% \end{macrocode}
% \todo{\cs\phrase is the result of expansion here: what effect will this have
% on its proper nesting, and is this something to worry about?}
% \begin{macrocode}
\newcommand\foreignword [1] {%
\phrase{\word{#1}}%
}
% \end{macrocode}
% The \cs\@ cancels the effect on spacing of any final punctuation in the
% argument.
% \begin{macrocode}
% \newlet\phrase\WrapquotesSK -- whoops, doesn't work as intended,
% \phrase{foo}s puts a space before the following `s'
% old definition:
\newcommand\phrase [1] {%
\textquoteleft #1\textquoteright\@%
}
\newlet\term\textitswitch
\newlet\defn\textslswitch
% \end{macrocode}
% \end{macro}
% \end{macro}
% \end{macro}
% \end{macro}
% \end{macro}
% \end{macro}
%
% \section{Titles}
%
% \begin{macro}{\book}
% \begin{macro}{\journal}
% \begin{macro}{\music}
% \begin{macro}{\article}
% \begin{macro}{\storytitle}
% \begin{macro}{\poemtitle}
% \begin{macro}{\play}
% \begin{macro}{\craft}
% \begin{macro}{\species}
% \mbox{}
% \begin{macrocode}
\newlet\book\textitswitch
\newlet\journal\textitswitch
\newlet\music\textitswitch
\newlet\article\Wrapquotes
\newlet\storytitle\Wrapquotes
\newlet\poemtitle\Wrapquotes
\newlet\play\textitswitch % \manualref{7.145}
\newlet\craft\textitswitch
\newlet\species\textitswitch
% \end{macrocode}
% \end{macro}
% \end{macro}
% \end{macro}
% \end{macro}
% \end{macro}
% \end{macro}
% \end{macro}
% \end{macro}
% \end{macro}
%
% \part{Configuration}
%
% User alterations and additions and package testing are in a configuration
% file.
% \begin{macrocode}
\InputIfFileExists{titles.cfg}{}{}
% \end{macrocode}
%
% The contents of the \emph{distributed} configuration file are below.
%
% \InputIfFileExists{titles.cfg}{}{}
%
% \Finale