[BACK]Return to texinfo.tex CVS log [TXT][DIR] Up to [local] / OpenXM / src / asir-doc

Diff for /OpenXM/src/asir-doc/Attic/texinfo.tex between version 1.1 and 1.3

version 1.1, 1999/12/21 02:47:29 version 1.3, 2017/08/31 01:24:32
Line 1 
Line 1 
 % texinfo.tex -- TeX macros to handle Texinfo files.  % texinfo.tex -- TeX macros to handle Texinfo files.
 %  %
 % Load plain if necessary, i.e., if running under initex.  % Load plain if necessary, i.e., if running under initex.
 \expandafter\ifx\csname fmtname\endcsname\relax\input plain\fi  \expandafter\ifx\csname fmtname\endcsname\relax\input plain\fi
 %  %
 \def\texinfoversion{1999-09-25.10}  \def\texinfoversion{2016-08-16.20}
 %  %
 % Copyright (C) 1985, 86, 88, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99  % Copyright 1985, 1986, 1988, 1990, 1991, 1992, 1993, 1994, 1995,
   % 1996, 1997, 1998, 1999, 2000, 2001, 2002, 2003, 2004, 2005, 2006,
   % 2007, 2008, 2009, 2010, 2011, 2012, 2013, 2014, 2015, 2016
 % Free Software Foundation, Inc.  % Free Software Foundation, Inc.
 %  %
 % This texinfo.tex file is free software; you can redistribute it and/or  % This texinfo.tex file is free software: you can redistribute it and/or
 % modify it under the terms of the GNU General Public License as  % modify it under the terms of the GNU General Public License as
 % published by the Free Software Foundation; either version 2, or (at  % published by the Free Software Foundation, either version 3 of the
 % your option) any later version.  % License, or (at your option) any later version.
 %  %
 % This texinfo.tex file is distributed in the hope that it will be  % This texinfo.tex file is distributed in the hope that it will be
 % useful, but WITHOUT ANY WARRANTY; without even the implied warranty  % useful, but WITHOUT ANY WARRANTY; without even the implied warranty
Line 19 
Line 21 
 % General Public License for more details.  % General Public License for more details.
 %  %
 % You should have received a copy of the GNU General Public License  % You should have received a copy of the GNU General Public License
 % along with this texinfo.tex file; see the file COPYING.  If not, write  % along with this program.  If not, see <http://www.gnu.org/licenses/>.
 % to the Free Software Foundation, Inc., 59 Temple Place - Suite 330,  
 % Boston, MA 02111-1307, USA.  
 %  %
 % In other words, you are welcome to use, share and improve this program.  % As a special exception, when this file is read by TeX when processing
 % You are forbidden to forbid anyone else to use, share and improve  % a Texinfo source document, you may use the result without
 % what you give them.   Help stamp out software-hoarding!  % restriction. This Exception is an additional permission under section 7
   % of the GNU General Public License, version 3 ("GPLv3").
 %  %
 % Please try the latest version of texinfo.tex before submitting bug  % Please try the latest version of texinfo.tex before submitting bug
 % reports; you can get the latest version from:  % reports; you can get the latest version from:
 %   ftp://ftp.gnu.org/gnu/texinfo.tex  %   http://ftp.gnu.org/gnu/texinfo/ (the Texinfo release area), or
 %   (and all GNU mirrors, see http://www.gnu.org/order/ftp.html)  %   http://ftpmirror.gnu.org/texinfo/ (same, via a mirror), or
 %   ftp://texinfo.org/tex/texinfo.tex  %   http://www.gnu.org/software/texinfo/ (the Texinfo home page)
 %   ftp://us.ctan.org/macros/texinfo/texinfo.tex  % The texinfo.tex in any given distribution could well be out
 %   (and all CTAN mirrors, finger ctan@us.ctan.org for a list).  
 %   /home/gd/gnu/doc/texinfo.tex on the GNU machines.  
 % The texinfo.tex in any given Texinfo distribution could well be out  
 % of date, so if that's what you're using, please check.  % of date, so if that's what you're using, please check.
 % Texinfo has a small home page at http://texinfo.org/.  
 %  %
 % Send bug reports to bug-texinfo@gnu.org.  Please include including a  % Send bug reports to bug-texinfo@gnu.org.  Please include including a
 % complete document in each bug report with which we can reproduce the  % complete document in each bug report with which we can reproduce the
Line 50 
Line 47 
 %   texindex foo.??  %   texindex foo.??
 %   tex foo.texi  %   tex foo.texi
 %   tex foo.texi  %   tex foo.texi
 %   dvips foo.dvi -o # or whatever, to process the dvi file; this makes foo.ps.  %   dvips foo.dvi -o  # or whatever; this makes foo.ps.
 % The extra runs of TeX get the cross-reference information correct.  % The extra TeX runs get the cross-reference information correct.
 % Sometimes one run after texindex suffices, and sometimes you need more  % Sometimes one run after texindex suffices, and sometimes you need more
 % than two; texi2dvi does it as many times as necessary.  % than two; texi2dvi does it as many times as necessary.
 %  %
 % It is possible to adapt texinfo.tex for other languages.  You can get  % It is possible to adapt texinfo.tex for other languages, to some
 % the existing language-specific files from ftp://ftp.gnu.org/gnu/texinfo/.  % extent.  You can get the existing language-specific files from the
   % full Texinfo distribution.
   %
   % The GNU Texinfo home page is http://www.gnu.org/software/texinfo.
   
   
 \message{Loading texinfo [version \texinfoversion]:}  \message{Loading texinfo [version \texinfoversion]:}
   
 % If in a .fmt file, print the version number  % If in a .fmt file, print the version number
Line 66 
Line 67 
 \everyjob{\message{[Texinfo version \texinfoversion]}%  \everyjob{\message{[Texinfo version \texinfoversion]}%
   \catcode`+=\active \catcode`\_=\active}    \catcode`+=\active \catcode`\_=\active}
   
 % Save some parts of plain tex whose names we will redefine.  % LaTeX's \typeout.  This ensures that the messages it is used for
   % are identical in format to the corresponding ones from latex/pdflatex.
   \def\typeout{\immediate\write17}%
   
   \chardef\other=12
   
   % We never want plain's \outer definition of \+ in Texinfo.
   % For @tex, we can use \tabalign.
   \let\+ = \relax
   
   % Save some plain tex macros whose names we will redefine.
 \let\ptexb=\b  \let\ptexb=\b
 \let\ptexbullet=\bullet  \let\ptexbullet=\bullet
 \let\ptexc=\c  \let\ptexc=\c
Line 76 
Line 87 
 \let\ptexend=\end  \let\ptexend=\end
 \let\ptexequiv=\equiv  \let\ptexequiv=\equiv
 \let\ptexexclam=\!  \let\ptexexclam=\!
   \let\ptexfootnote=\footnote
   \let\ptexgtr=>
   \let\ptexhat=^
 \let\ptexi=\i  \let\ptexi=\i
   \let\ptexindent=\indent
   \let\ptexinsert=\insert
 \let\ptexlbrace=\{  \let\ptexlbrace=\{
   \let\ptexless=<
   \let\ptexnewwrite\newwrite
   \let\ptexnoindent=\noindent
   \let\ptexplus=+
   \let\ptexraggedright=\raggedright
 \let\ptexrbrace=\}  \let\ptexrbrace=\}
   \let\ptexslash=\/
   \let\ptexsp=\sp
 \let\ptexstar=\*  \let\ptexstar=\*
   \let\ptexsup=\sup
 \let\ptext=\t  \let\ptext=\t
   \let\ptextop=\top
   {\catcode`\'=\active \global\let\ptexquoteright'}% active in plain's math mode
   
 % We never want plain's outer \+ definition in Texinfo.  
 % For @tex, we can use \tabalign.  
 \let\+ = \relax  
   
 \message{Basics,}  
 \chardef\other=12  
   
 % If this character appears in an error message or help string, it  % If this character appears in an error message or help string, it
 % starts a new line in the output.  % starts a new line in the output.
 \newlinechar = `^^J  \newlinechar = `^^J
   
   % Use TeX 3.0's \inputlineno to get the line number, for better error
   % messages, but if we're using an old version of TeX, don't do anything.
   %
   \ifx\inputlineno\thisisundefined
     \let\linenumber = \empty % Pre-3.0.
   \else
     \def\linenumber{l.\the\inputlineno:\space}
   \fi
   
 % Set up fixed words for English if not already set.  % Set up fixed words for English if not already set.
 \ifx\putwordAppendix\undefined  \gdef\putwordAppendix{Appendix}\fi  \ifx\putwordAppendix\undefined  \gdef\putwordAppendix{Appendix}\fi
 \ifx\putwordChapter\undefined   \gdef\putwordChapter{Chapter}\fi  \ifx\putwordChapter\undefined   \gdef\putwordChapter{Chapter}\fi
   \ifx\putworderror\undefined     \gdef\putworderror{error}\fi
 \ifx\putwordfile\undefined      \gdef\putwordfile{file}\fi  \ifx\putwordfile\undefined      \gdef\putwordfile{file}\fi
 \ifx\putwordin\undefined        \gdef\putwordin{in}\fi  \ifx\putwordin\undefined        \gdef\putwordin{in}\fi
 \ifx\putwordIndexIsEmpty\undefined     \gdef\putwordIndexIsEmpty{(Index is empty)}\fi  \ifx\putwordIndexIsEmpty\undefined       \gdef\putwordIndexIsEmpty{(Index is empty)}\fi
 \ifx\putwordIndexNonexistent\undefined \gdef\putwordIndexNonexistent{(Index is nonexistent)}\fi  \ifx\putwordIndexNonexistent\undefined   \gdef\putwordIndexNonexistent{(Index is nonexistent)}\fi
 \ifx\putwordInfo\undefined      \gdef\putwordInfo{Info}\fi  \ifx\putwordInfo\undefined      \gdef\putwordInfo{Info}\fi
 \ifx\putwordInstanceVariableof\undefined \gdef\putwordInstanceVariableof{Instance Variable of}\fi  \ifx\putwordInstanceVariableof\undefined \gdef\putwordInstanceVariableof{Instance Variable of}\fi
 \ifx\putwordMethodon\undefined  \gdef\putwordMethodon{Method on}\fi  \ifx\putwordMethodon\undefined  \gdef\putwordMethodon{Method on}\fi
Line 131 
Line 160 
 \ifx\putwordDefspec\undefined   \gdef\putwordDefspec{Special Form}\fi  \ifx\putwordDefspec\undefined   \gdef\putwordDefspec{Special Form}\fi
 \ifx\putwordDefvar\undefined    \gdef\putwordDefvar{Variable}\fi  \ifx\putwordDefvar\undefined    \gdef\putwordDefvar{Variable}\fi
 \ifx\putwordDefopt\undefined    \gdef\putwordDefopt{User Option}\fi  \ifx\putwordDefopt\undefined    \gdef\putwordDefopt{User Option}\fi
 \ifx\putwordDeftypevar\undefined\gdef\putwordDeftypevar{Variable}\fi  
 \ifx\putwordDeffunc\undefined   \gdef\putwordDeffunc{Function}\fi  \ifx\putwordDeffunc\undefined   \gdef\putwordDeffunc{Function}\fi
 \ifx\putwordDeftypefun\undefined\gdef\putwordDeftypefun{Function}\fi  
   
   % Give the space character the catcode for a space.
   \def\spaceisspace{\catcode`\ =10\relax}
   
   \chardef\dashChar  = `\-
   \chardef\slashChar = `\/
   \chardef\underChar = `\_
   
 % Ignore a token.  % Ignore a token.
 %  %
 \def\gobble#1{}  \def\gobble#1{}
   
 \hyphenation{ap-pen-dix}  % The following is used inside several \edef's.
 \hyphenation{mini-buf-fer mini-buf-fers}  \def\makecsname#1{\expandafter\noexpand\csname#1\endcsname}
 \hyphenation{eshell}  
 \hyphenation{white-space}  
   
 % Margin to add to right of even pages, to left of odd pages.  % Hyphenation fixes.
 \newdimen \bindingoffset  \hyphenation{
 \newdimen \normaloffset    Flor-i-da Ghost-script Ghost-view Mac-OS Post-Script
 \newdimen\pagewidth \newdimen\pageheight    ap-pen-dix bit-map bit-maps
     data-base data-bases eshell fall-ing half-way long-est man-u-script
     man-u-scripts mini-buf-fer mini-buf-fers over-view par-a-digm
     par-a-digms rath-er rec-tan-gu-lar ro-bot-ics se-vere-ly set-up spa-ces
     spell-ing spell-ings
     stand-alone strong-est time-stamp time-stamps which-ever white-space
     wide-spread wrap-around
   }
   
 % Sometimes it is convenient to have everything in the transcript file  % Sometimes it is convenient to have everything in the transcript file
 % and nothing on the terminal.  We don't just call \tracingall here,  % and nothing on the terminal.  We don't just call \tracingall here,
 % since that produces some useless output on the terminal.  % since that produces some useless output on the terminal.  We also make
   % some effort to order the tracing commands to reduce output in the log
   % file; cf. trace.sty in LaTeX.
 %  %
 \def\gloggingall{\begingroup \globaldefs = 1 \loggingall \endgroup}%  \def\gloggingall{\begingroup \globaldefs = 1 \loggingall \endgroup}%
 \ifx\eTeXversion\undefined  \def\loggingall{%
 \def\loggingall{\tracingcommands2 \tracingstats2    \tracingstats2
    \tracingpages1 \tracingoutput1 \tracinglostchars1    \tracingpages1
    \tracingmacros2 \tracingparagraphs1 \tracingrestores1    \tracinglostchars2  % 2 gives us more in etex
    \showboxbreadth\maxdimen\showboxdepth\maxdimen    \tracingparagraphs1
     \tracingoutput1
     \tracingmacros2
     \tracingrestores1
     \showboxbreadth\maxdimen \showboxdepth\maxdimen
     \ifx\eTeXversion\thisisundefined\else % etex gives us more logging
       \tracingscantokens1
       \tracingifs1
       \tracinggroups1
       \tracingnesting2
       \tracingassigns1
     \fi
     \tracingcommands3  % 3 gives us more in etex
     \errorcontextlines16
 }%  }%
 \else  
 \def\loggingall{\tracingcommands3 \tracingstats2  
    \tracingpages1 \tracingoutput1 \tracinglostchars1  
    \tracingmacros2 \tracingparagraphs1 \tracingrestores1  
    \tracingscantokens1 \tracingassigns1 \tracingifs1  
    \tracinggroups1 \tracingnesting2  
    \showboxbreadth\maxdimen\showboxdepth\maxdimen  
 }%  
 \fi  
   
 % For @cropmarks command.  % @errormsg{MSG}.  Do the index-like expansions on MSG, but if things
   % aren't perfect, it's not the end of the world, being an error message,
   % after all.
   %
   \def\errormsg{\begingroup \indexnofonts \doerrormsg}
   \def\doerrormsg#1{\errmessage{#1}}
   
   % add check for \lastpenalty to plain's definitions.  If the last thing
   % we did was a \nobreak, we don't want to insert more space.
   %
   \def\smallbreak{\ifnum\lastpenalty<10000\par\ifdim\lastskip<\smallskipamount
     \removelastskip\penalty-50\smallskip\fi\fi}
   \def\medbreak{\ifnum\lastpenalty<10000\par\ifdim\lastskip<\medskipamount
     \removelastskip\penalty-100\medskip\fi\fi}
   \def\bigbreak{\ifnum\lastpenalty<10000\par\ifdim\lastskip<\bigskipamount
     \removelastskip\penalty-200\bigskip\fi\fi}
   
   % Output routine
   %
   
   % For a final copy, take out the rectangles
   % that mark overfull boxes (in case you have decided
   % that the text looks ok even though it passes the margin).
   %
   \def\finalout{\overfullrule=0pt }
   
 % Do @cropmarks to get crop marks.  % Do @cropmarks to get crop marks.
 %  %
 \newif\ifcropmarks  \newif\ifcropmarks
Line 184 
Line 254 
 \newdimen\cornerthick \cornerthick=.3pt  \newdimen\cornerthick \cornerthick=.3pt
 \newdimen\topandbottommargin \topandbottommargin=.75in  \newdimen\topandbottommargin \topandbottommargin=.75in
   
   % Output a mark which sets \thischapter, \thissection and \thiscolor.
   % We dump everything together because we only have one kind of mark.
   % This works because we only use \botmark / \topmark, not \firstmark.
   %
   % A mark contains a subexpression of the \ifcase ... \fi construct.
   % \get*marks macros below extract the needed part using \ifcase.
   %
   % Another complication is to let the user choose whether \thischapter
   % (\thissection) refers to the chapter (section) in effect at the top
   % of a page, or that at the bottom of a page.
   
   % \domark is called twice inside \chapmacro, to add one
   % mark before the section break, and one after.
   %   In the second call \prevchapterdefs is the same as \lastchapterdefs,
   % and \prevsectiondefs is the same as \lastsectiondefs.
   %   Then if the page is not broken at the mark, some of the previous
   % section appears on the page, and we can get the name of this section
   % from \firstmark for @everyheadingmarks top.
   %   @everyheadingmarks bottom uses \botmark.
   %
   % See page 260 of The TeXbook.
   \def\domark{%
     \toks0=\expandafter{\lastchapterdefs}%
     \toks2=\expandafter{\lastsectiondefs}%
     \toks4=\expandafter{\prevchapterdefs}%
     \toks6=\expandafter{\prevsectiondefs}%
     \toks8=\expandafter{\lastcolordefs}%
     \mark{%
                      \the\toks0 \the\toks2  % 0: marks for @everyheadingmarks top
         \noexpand\or \the\toks4 \the\toks6  % 1: for @everyheadingmarks bottom
       \noexpand\else \the\toks8             % 2: color marks
     }%
   }
   
   % \gettopheadingmarks, \getbottomheadingmarks,
   % \getcolormarks - extract needed part of mark.
   %
   % \topmark doesn't work for the very first chapter (after the title
   % page or the contents), so we use \firstmark there -- this gets us
   % the mark with the chapter defs, unless the user sneaks in, e.g.,
   % @setcolor (or @url, or @link, etc.) between @contents and the very
   % first @chapter.
   \def\gettopheadingmarks{%
     \ifcase0\topmark\fi
     \ifx\thischapter\empty \ifcase0\firstmark\fi \fi
   }
   \def\getbottomheadingmarks{\ifcase1\botmark\fi}
   \def\getcolormarks{\ifcase2\topmark\fi}
   
   % Avoid "undefined control sequence" errors.
   \def\lastchapterdefs{}
   \def\lastsectiondefs{}
   \def\lastsection{}
   \def\prevchapterdefs{}
   \def\prevsectiondefs{}
   \def\lastcolordefs{}
   
   % Margin to add to right of even pages, to left of odd pages.
   \newdimen\bindingoffset
   \newdimen\normaloffset
   \newdimen\txipagewidth \newdimen\txipageheight
   
 % Main output routine.  % Main output routine.
   %
 \chardef\PAGE = 255  \chardef\PAGE = 255
 \output = {\onepageout{\pagecontents\PAGE}}  \output = {\onepageout{\pagecontents\PAGE}}
   
 \newbox\headlinebox  \newbox\headlinebox
 \newbox\footlinebox  \newbox\footlinebox
   
 % \onepageout takes a vbox as an argument.  Note that \pagecontents  % \onepageout takes a vbox as an argument.
 % does insertions, but you have to call it yourself.  % \shipout a vbox for a single page, adding an optional header, footer,
   % cropmarks, and footnote.  This also causes index entries for this page
   % to be written to the auxiliary files.
   %
 \def\onepageout#1{%  \def\onepageout#1{%
   \ifcropmarks \hoffset=0pt \else \hoffset=\normaloffset \fi    \ifcropmarks \hoffset=0pt \else \hoffset=\normaloffset \fi
   %    %
   \ifodd\pageno  \advance\hoffset by \bindingoffset    \ifodd\pageno  \advance\hoffset by \bindingoffset
   \else \advance\hoffset by -\bindingoffset\fi    \else \advance\hoffset by -\bindingoffset\fi
   %    %
     % Common context changes for both heading and footing.
   % Do this outside of the \shipout so @code etc. will be expanded in    % Do this outside of the \shipout so @code etc. will be expanded in
   % the headline as they should be, not taken literally (outputting ''code).    % the headline as they should be, not taken literally (outputting ''code).
   \setbox\headlinebox = \vbox{\let\hsize=\pagewidth \makeheadline}%    \def\commmonheadfootline{\let\hsize=\txipagewidth \texinfochars}
   \setbox\footlinebox = \vbox{\let\hsize=\pagewidth \makefootline}%  
   %    %
     % Retrieve the information for the headings from the marks in the page,
     % and call Plain TeX's \makeheadline and \makefootline, which use the
     % values in \headline and \footline.
     %
     % This is used to check if we are on the first page of a chapter.
     \ifcase1\topmark\fi
     \let\prevchaptername\thischaptername
     \ifcase0\firstmark\fi
     \let\curchaptername\thischaptername
     %
     \ifodd\pageno \getoddheadingmarks \else \getevenheadingmarks \fi
     \ifodd\pageno \getoddfootingmarks \else \getevenfootingmarks \fi
     %
     \ifx\curchaptername\prevchaptername
       \let\thischapterheading\thischapter
     \else
       % \thischapterheading is the same as \thischapter except it is blank
       % for the first page of a chapter.  This is to prevent the chapter name
       % being shown twice.
       \def\thischapterheading{}%
     \fi
     %
     \global\setbox\headlinebox = \vbox{\commmonheadfootline \makeheadline}%
     \global\setbox\footlinebox = \vbox{\commmonheadfootline \makefootline}%
     %
   {%    {%
       % Set context for writing to auxiliary files like index files.
     % Have to do this stuff outside the \shipout because we want it to      % Have to do this stuff outside the \shipout because we want it to
     % take effect in \write's, yet the group defined by the \vbox ends      % take effect in \write's, yet the group defined by the \vbox ends
     % before the \shipout runs.      % before the \shipout runs.
     %      %
     \escapechar = `\\     % use backslash in output files.  
     \indexdummies         % don't expand commands in the output.      \indexdummies         % don't expand commands in the output.
     \normalturnoffactive  % \ in index entries must not stay \, e.g., if      \normalturnoffactive  % \ in index entries must not stay \, e.g., if
                    % the page break happens to be in the middle of an example.                 % the page break happens to be in the middle of an example.
                  % We don't want .vr (or whatever) entries like this:
                  % \entry{{\indexbackslash }acronym}{32}{\code {\acronym}}
                  % "\acronym" won't work when it's read back in;
                  % it needs to be
                  % {\code {{\backslashcurfont }acronym}
     \shipout\vbox{%      \shipout\vbox{%
         % Do this early so pdf references go to the beginning of the page.
         \ifpdfmakepagedest \pdfdest name{\the\pageno} xyz\fi
         %
       \ifcropmarks \vbox to \outervsize\bgroup        \ifcropmarks \vbox to \outervsize\bgroup
         \hsize = \outerhsize          \hsize = \outerhsize
         \vskip-\topandbottommargin          \vskip-\topandbottommargin
Line 237 
Line 406 
       \pagebody{#1}%        \pagebody{#1}%
       \ifdim\ht\footlinebox > 0pt        \ifdim\ht\footlinebox > 0pt
         % Only leave this space if the footline is nonempty.          % Only leave this space if the footline is nonempty.
         % (We lessened \vsize for it in \oddfootingxxx.)          % (We lessened \vsize for it in \oddfootingyyy.)
         % The \baselineskip=24pt in plain's \makefootline has no effect.          % The \baselineskip=24pt in plain's \makefootline has no effect.
         \vskip 2\baselineskip          \vskip 24pt
         \unvbox\footlinebox          \unvbox\footlinebox
       \fi        \fi
       %        %
       \ifpdfmakepagedest \pdfmkdest{\the\pageno} \fi  
       %  
       \ifcropmarks        \ifcropmarks
           \egroup % end of \vbox\bgroup            \egroup % end of \vbox\bgroup
         \hfil\egroup % end of (centering) \line\bgroup          \hfil\egroup % end of (centering) \line\bgroup
Line 262 
Line 429 
       \egroup % \vbox from first cropmarks clause        \egroup % \vbox from first cropmarks clause
       \fi        \fi
     }% end of \shipout\vbox      }% end of \shipout\vbox
   }% end of group with \turnoffactive    }% end of group with \indexdummies
   \advancepageno    \advancepageno
   \ifnum\outputpenalty>-20000 \else\dosupereject\fi    \ifnum\outputpenalty>-20000 \else\dosupereject\fi
 }  }
   
 \newinsert\margin \dimen\margin=\maxdimen  \newinsert\margin \dimen\margin=\maxdimen
   
 \def\pagebody#1{\vbox to\pageheight{\boxmaxdepth=\maxdepth #1}}  % Main part of page, including any footnotes
   \def\pagebody#1{\vbox to\txipageheight{\boxmaxdepth=\maxdepth #1}}
 {\catcode`\@ =11  {\catcode`\@ =11
 \gdef\pagecontents#1{\ifvoid\topins\else\unvbox\topins\fi  \gdef\pagecontents#1{\ifvoid\topins\else\unvbox\topins\fi
 % marginal hacks, juha@viisa.uucp (Juha Takala)  % marginal hacks, juha@viisa.uucp (Juha Takala)
 \ifvoid\margin\else % marginal info is present  \ifvoid\margin\else % marginal info is present
   \rlap{\kern\hsize\vbox to\z@{\kern1pt\box\margin \vss}}\fi    \rlap{\kern\hsize\vbox to\z@{\kern1pt\box\margin \vss}}\fi
 \dimen@=\dp#1 \unvbox#1  \dimen@=\dp#1\relax \unvbox#1\relax
 \ifvoid\footins\else\vskip\skip\footins\footnoterule \unvbox\footins\fi  \ifvoid\footins\else\vskip\skip\footins\footnoterule \unvbox\footins\fi
 \ifr@ggedbottom \kern-\dimen@ \vfil \fi}  \ifr@ggedbottom \kern-\dimen@ \vfil \fi}
 }  }
Line 291 
Line 459 
 \def\nsbot{\vbox  \def\nsbot{\vbox
   {\hrule height\cornerlong depth\cornerthick width\cornerthick}}    {\hrule height\cornerlong depth\cornerthick width\cornerthick}}
   
   
   % Argument parsing
   
 % Parse an argument, then pass it to #1.  The argument is the rest of  % Parse an argument, then pass it to #1.  The argument is the rest of
 % the input line (except we remove a trailing comment).  #1 should be a  % the input line (except we remove a trailing comment).  #1 should be a
 % macro which expects an ordinary undelimited TeX argument.  % macro which expects an ordinary undelimited TeX argument.
   % For example, \def\foo{\parsearg\fooxxx}.
 %  %
 \def\parsearg#1{%  \def\parsearg{\parseargusing{}}
   \let\next = #1%  \def\parseargusing#1#2{%
     \def\argtorun{#2}%
   \begingroup    \begingroup
     \obeylines      \obeylines
     \futurelet\temp\parseargx      \spaceisspace
       #1%
       \parseargline\empty% Insert the \empty token, see \finishparsearg below.
 }  }
   
 % If the next token is an obeyed space (from an @example environment or  
 % the like), remove it and recurse.  Otherwise, we're done.  
 \def\parseargx{%  
   % \obeyedspace is defined far below, after the definition of \sepspaces.  
   \ifx\obeyedspace\temp  
     \expandafter\parseargdiscardspace  
   \else  
     \expandafter\parseargline  
   \fi  
 }  
   
 % Remove a single space (as the delimiter token to the macro call).  
 {\obeyspaces %  
  \gdef\parseargdiscardspace {\futurelet\temp\parseargx}}  
   
 {\obeylines %  {\obeylines %
   \gdef\parseargline#1^^M{%    \gdef\parseargline#1^^M{%
     \endgroup % End of the group started in \parsearg.      \endgroup % End of the group started in \parsearg.
     %      \argremovecomment #1\comment\ArgTerm%
     % First remove any @c comment, then any @comment.  
     % Result of each macro is put in \toks0.  
     \argremovec #1\c\relax %  
     \expandafter\argremovecomment \the\toks0 \comment\relax %  
     %  
     % Call the caller's macro, saved as \next in \parsearg.  
     \expandafter\next\expandafter{\the\toks0}%  
   }%    }%
 }  }
   
 % Since all \c{,omment} does is throw away the argument, we can let TeX  % First remove any @comment, then any @c comment.  Also remove a @texinfoc
 % do that for us.  The \relax here is matched by the \relax in the call  % comment (see \scanmacro for details).  Pass the result on to \argcheckspaces.
 % in \parseargline; it could be more or less anything, its purpose is  \def\argremovecomment#1\comment#2\ArgTerm{\argremovec #1\c\ArgTerm}
 % just to delimit the argument to the \c.  \def\argremovec#1\c#2\ArgTerm{\argremovetexinfoc #1\texinfoc\ArgTerm}
 \def\argremovec#1\c#2\relax{\toks0 = {#1}}  \def\argremovetexinfoc#1\texinfoc#2\ArgTerm{\argcheckspaces#1\^^M\ArgTerm}
 \def\argremovecomment#1\comment#2\relax{\toks0 = {#1}}  
   
 % \argremovec{,omment} might leave us with trailing spaces, though; e.g.,  % Each occurrence of `\^^M' or `<space>\^^M' is replaced by a single space.
   %
   % \argremovec might leave us with trailing space, e.g.,
 %    @end itemize  @c foo  %    @end itemize  @c foo
 % will have two active spaces as part of the argument with the  % This space token undergoes the same procedure and is eventually removed
 % `itemize'.  Here we remove all active spaces from #1, and assign the  % by \finishparsearg.
 % result to \toks0.  
 %  %
 % This loses if there are any *other* active characters besides spaces  \def\argcheckspaces#1\^^M{\argcheckspacesX#1\^^M \^^M}
 % in the argument -- _ ^ +, for example -- since they get expanded.  \def\argcheckspacesX#1 \^^M{\argcheckspacesY#1\^^M}
 % Fortunately, Texinfo does not define any such commands.  (If it ever  \def\argcheckspacesY#1\^^M#2\^^M#3\ArgTerm{%
 % does, the catcode of the characters in questionwill have to be changed    \def\temp{#3}%
 % here.)  But this means we cannot call \removeactivespaces as part of    \ifx\temp\empty
 % \argremovec{,omment}, since @c uses \parsearg, and thus the argument      % Do not use \next, perhaps the caller of \parsearg uses it; reuse \temp:
 % that \parsearg gets might well have any character at all in it.      \let\temp\finishparsearg
 %    \else
 \def\removeactivespaces#1{%      \let\temp\argcheckspaces
   \begingroup    \fi
     \ignoreactivespaces    % Put the space token in:
     \edef\temp{#1}%    \temp#1 #3\ArgTerm
     \global\toks0 = \expandafter{\temp}%  
   \endgroup  
 }  }
   
 % Change the active space to expand to nothing.  % If a _delimited_ argument is enclosed in braces, they get stripped; so
   % to get _exactly_ the rest of the line, we had to prevent such situation.
   % We prepended an \empty token at the very beginning and we expand it now,
   % just before passing the control to \argtorun.
   % (Similarly, we have to think about #3 of \argcheckspacesY above: it is
   % either the null string, or it ends with \^^M---thus there is no danger
   % that a pair of braces would be stripped.
 %  %
 \begingroup  % But first, we have to remove the trailing space token.
   %
   \def\finishparsearg#1 \ArgTerm{\expandafter\argtorun\expandafter{#1}}
   
   
   % \parseargdef - define a command taking an argument on the line
   %
   % \parseargdef\foo{...}
   %       is roughly equivalent to
   % \def\foo{\parsearg\Xfoo}
   % \def\Xfoo#1{...}
   \def\parseargdef#1{%
     \expandafter \doparseargdef \csname\string#1\endcsname #1%
   }
   \def\doparseargdef#1#2{%
     \def#2{\parsearg#1}%
     \def#1##1%
   }
   
   % Several utility definitions with active space:
   {
   \obeyspaces    \obeyspaces
   \gdef\ignoreactivespaces{\obeyspaces\let =\empty}    \gdef\obeyedspace{ }
 \endgroup  
   
     % Make each space character in the input produce a normal interword
     % space in the output.  Don't allow a line break at this space, as this
     % is used only in environments like @example, where each line of input
     % should produce a line of output anyway.
     %
     \gdef\sepspaces{\obeyspaces\let =\tie}
   
     % If an index command is used in an @example environment, any spaces
     % therein should become regular spaces in the raw index file, not the
     % expansion of \tie (\leavevmode \penalty \@M \ ).
     \gdef\unsepspaces{\let =\space}
   }
   
   
 \def\flushcr{\ifx\par\lisppar \def\next##1{}\else \let\next=\relax \fi \next}  \def\flushcr{\ifx\par\lisppar \def\next##1{}\else \let\next=\relax \fi \next}
   
 %% These are used to keep @begin/@end levels from running away  % Define the framework for environments in texinfo.tex.  It's used like this:
 %% Call \inENV within environments (after a \begingroup)  %
 \newif\ifENV \ENVfalse \def\inENV{\ifENV\relax\else\ENVtrue\fi}  %   \envdef\foo{...}
 \def\ENVcheck{%  %   \def\Efoo{...}
 \ifENV\errmessage{Still within an environment; press RETURN to continue}  %
 \endgroup\fi} % This is not perfect, but it should reduce lossage  % It's the responsibility of \envdef to insert \begingroup before the
   % actual body; @end closes the group after calling \Efoo.  \envdef also
   % defines \thisenv, so the current environment is known; @end checks
   % whether the environment name matches.  The \checkenv macro can also be
   % used to check whether the current environment is the one expected.
   %
   % Non-false conditionals (@iftex, @ifset) don't fit into this, so they
   % are not treated as environments; they don't open a group.  (The
   % implementation of @end takes care not to call \endgroup in this
   % special case.)
   
 % @begin foo  is the same as @foo, for now.  
 \newhelp\EMsimple{Press RETURN to continue.}  
   
 \outer\def\begin{\parsearg\beginxxx}  % At run-time, environments start with this:
   \def\startenvironment#1{\begingroup\def\thisenv{#1}}
   % initialize
   \let\thisenv\empty
   
 \def\beginxxx #1{%  % ... but they get defined via ``\envdef\foo{...}'':
 \expandafter\ifx\csname #1\endcsname\relax  \long\def\envdef#1#2{\def#1{\startenvironment#1#2}}
 {\errhelp=\EMsimple \errmessage{Undefined command @begin #1}}\else  \def\envparseargdef#1#2{\parseargdef#1{\startenvironment#1#2}}
 \csname #1\endcsname\fi}  
   
 % @end foo executes the definition of \Efoo.  % Check whether we're in the right environment:
 %  \def\checkenv#1{%
 \def\end{\parsearg\endxxx}    \def\temp{#1}%
 \def\endxxx #1{%    \ifx\thisenv\temp
   \removeactivespaces{#1}%  
   \edef\endthing{\the\toks0}%  
   %  
   \expandafter\ifx\csname E\endthing\endcsname\relax  
     \expandafter\ifx\csname \endthing\endcsname\relax  
       % There's no \foo, i.e., no ``environment'' foo.  
       \errhelp = \EMsimple  
       \errmessage{Undefined command `@end \endthing'}%  
     \else  
       \unmatchedenderror\endthing  
     \fi  
   \else    \else
     % Everything's ok; the right environment has been started.      \badenverr
     \csname E\endthing\endcsname  
   \fi    \fi
 }  }
   
 % There is an environment #1, but it hasn't been started.  Give an error.  % Environment mismatch, #1 expected:
 %  \def\badenverr{%
 \def\unmatchedenderror#1{%  
   \errhelp = \EMsimple    \errhelp = \EMsimple
   \errmessage{This `@end #1' doesn't have a matching `@#1'}%    \errmessage{This command can appear only \inenvironment\temp,
       not \inenvironment\thisenv}%
 }  }
   \def\inenvironment#1{%
     \ifx#1\empty
       outside of any environment%
     \else
       in environment \expandafter\string#1%
     \fi
   }
   
 % Define the control sequence \E#1 to give an unmatched @end error.  % @end foo executes the definition of \Efoo.
   % But first, it executes a specialized version of \checkenv
 %  %
 \def\defineunmatchedend#1{%  \parseargdef\end{%
   \expandafter\def\csname E#1\endcsname{\unmatchedenderror{#1}}%    \if 1\csname iscond.#1\endcsname
     \else
       % The general wording of \badenverr may not be ideal.
       \expandafter\checkenv\csname#1\endcsname
       \csname E#1\endcsname
       \endgroup
     \fi
 }  }
   
   \newhelp\EMsimple{Press RETURN to continue.}
   
 % Single-spacing is done by various environments (specifically, in  
 % \nonfillstart and \quotations).  
 \newskip\singlespaceskip \singlespaceskip = 12.5pt  
 \def\singlespace{%  
   % Why was this kern here?  It messes up equalizing space above and below  
   % environments.  --karl, 6may93  
   %{\advance \baselineskip by -\singlespaceskip  
   %\kern \baselineskip}%  
   \setleading \singlespaceskip  
 }  
   
 %% Simple single-character @ commands  
   
 % @@ prints an @  
 % Kludge this until the fonts are right (grr).  
 \def\@{{\tt\char64}}  
   
 % This is turned off because it was never documented  
 % and you can use @w{...} around a quote to suppress ligatures.  
 %% Define @` and @' to be the same as ` and '  
 %% but suppressing ligatures.  
 %\def\`{{`}}  
 %\def\'{{'}}  
   
 % Used to generate quoted braces.  
 \def\mylbrace {{\tt\char123}}  
 \def\myrbrace {{\tt\char125}}  
 \let\{=\mylbrace  
 \let\}=\myrbrace  
 \begingroup  
   % Definitions to produce actual \{ & \} command in an index.  
   \catcode`\{ = 12 \catcode`\} = 12  
   \catcode`\[ = 1 \catcode`\] = 2  
   \catcode`\@ = 0 \catcode`\\ = 12  
   @gdef@lbracecmd[\{]%  
   @gdef@rbracecmd[\}]%  
 @endgroup  
   
 % Accents: @, @dotaccent @ringaccent @ubaraccent @udotaccent  
 % Others are defined by plain TeX: @` @' @" @^ @~ @= @v @H.  
 \let\, = \c  
 \let\dotaccent = \.  
 \def\ringaccent#1{{\accent23 #1}}  
 \let\tieaccent = \t  
 \let\ubaraccent = \b  
 \let\udotaccent = \d  
   
 % Other special characters: @questiondown @exclamdown  
 % Plain TeX defines: @AA @AE @O @OE @L (and lowercase versions) @ss.  
 \def\questiondown{?`}  
 \def\exclamdown{!`}  
   
 % Dotless i and dotless j, used for accents.  
 \def\imacro{i}  
 \def\jmacro{j}  
 \def\dotless#1{%  
   \def\temp{#1}%  
   \ifx\temp\imacro \ptexi  
   \else\ifx\temp\jmacro \j  
   \else \errmessage{@dotless can be used only with i or j}%  
   \fi\fi  
 }  
   
 % Be sure we're in horizontal mode when doing a tie, since we make space  % Be sure we're in horizontal mode when doing a tie, since we make space
 % equivalent to this in @example-like environments. Otherwise, a space  % equivalent to this in @example-like environments. Otherwise, a space
 % at the beginning of a line will start with \penalty -- and  % at the beginning of a line will start with \penalty -- and
Line 501 
Line 640 
 \def\:{\spacefactor=1000 }  \def\:{\spacefactor=1000 }
   
 % @* forces a line break.  % @* forces a line break.
 \def\*{\hfil\break\hbox{}\ignorespaces}  \def\*{\unskip\hfil\break\hbox{}\ignorespaces}
   
   % @/ allows a line break.
   \let\/=\allowbreak
   
 % @. is an end-of-sentence period.  % @. is an end-of-sentence period.
 \def\.{.\spacefactor=3000 }  \def\.{.\spacefactor=\endofsentencespacefactor\space}
   
 % @! is an end-of-sentence bang.  % @! is an end-of-sentence bang.
 \def\!{!\spacefactor=3000 }  \def\!{!\spacefactor=\endofsentencespacefactor\space}
   
 % @? is an end-of-sentence query.  % @? is an end-of-sentence query.
 \def\?{?\spacefactor=3000 }  \def\?{?\spacefactor=\endofsentencespacefactor\space}
   
   % @frenchspacing on|off  says whether to put extra space after punctuation.
   %
   \def\onword{on}
   \def\offword{off}
   %
   \parseargdef\frenchspacing{%
     \def\temp{#1}%
     \ifx\temp\onword \plainfrenchspacing
     \else\ifx\temp\offword \plainnonfrenchspacing
     \else
       \errhelp = \EMsimple
       \errmessage{Unknown @frenchspacing option `\temp', must be on|off}%
     \fi\fi
   }
   
 % @w prevents a word break.  Without the \leavevmode, @w at the  % @w prevents a word break.  Without the \leavevmode, @w at the
 % beginning of a paragraph, when TeX is still in vertical mode, would  % beginning of a paragraph, when TeX is still in vertical mode, would
 % produce a whole line of output instead of starting the paragraph.  % produce a whole line of output instead of starting the paragraph.
Line 525 
Line 682 
 % therefore, no glue is inserted, and the space between the headline and  % therefore, no glue is inserted, and the space between the headline and
 % the text is small, which looks bad.  % the text is small, which looks bad.
 %  %
 \def\group{\begingroup  % Another complication is that the group might be very large.  This can
   \ifnum\catcode13=\active \else  % cause the glue on the previous page to be unduly stretched, because it
   % does not have much material.  In this case, it's better to add an
   % explicit \vfill so that the extra space is at the bottom.  The
   % threshold for doing this is if the group is more than \vfilllimit
   % percent of a page (\vfilllimit can be changed inside of @tex).
   %
   \newbox\groupbox
   \def\vfilllimit{0.7}
   %
   \envdef\group{%
     \ifnum\catcode`\^^M=\active \else
     \errhelp = \groupinvalidhelp      \errhelp = \groupinvalidhelp
     \errmessage{@group invalid in context where filling is enabled}%      \errmessage{@group invalid in context where filling is enabled}%
   \fi    \fi
     \startsavinginserts
   %    %
   % The \vtop we start below produces a box with normal height and large    \setbox\groupbox = \vtop\bgroup
   % depth; thus, TeX puts \baselineskip glue before it, and (when the  
   % next line of text is done) \lineskip glue after it.  (See p.82 of  
   % the TeXbook.)  Thus, space below is not quite equal to space  
   % above.  But it's pretty close.  
   \def\Egroup{%  
     \egroup           % End the \vtop.  
     \endgroup         % End the \group.  
   }%  
   %  
   \vtop\bgroup  
     % We have to put a strut on the last line in case the @group is in  
     % the midst of an example, rather than completely enclosing it.  
     % Otherwise, the interline space between the last line of the group  
     % and the first line afterwards is too small.  But we can't put the  
     % strut in \Egroup, since there it would be on a line by itself.  
     % Hence this just inserts a strut at the beginning of each line.  
     \everypar = {\strut}%  
     %  
     % Since we have a strut on every line, we don't need any of TeX's  
     % normal interline spacing.  
     \offinterlineskip  
     %  
     % OK, but now we have to do something about blank  
     % lines in the input in @example-like environments, which normally  
     % just turn into \lisppar, which will insert no space now that we've  
     % turned off the interline space.  Simplest is to make them be an  
     % empty paragraph.  
     \ifx\par\lisppar  
       \edef\par{\leavevmode \par}%  
       %  
       % Reset ^^M's definition to new definition of \par.  
       \obeylines  
     \fi  
     %  
     % Do @comment since we are called inside an environment such as      % Do @comment since we are called inside an environment such as
     % @example, where each end-of-line in the input causes an      % @example, where each end-of-line in the input causes an
     % end-of-line in the output.  We don't want the end-of-line after      % end-of-line in the output.  We don't want the end-of-line after
Line 575 
Line 709 
     \comment      \comment
 }  }
 %  %
   % The \vtop produces a box with normal height and large depth; thus, TeX puts
   % \baselineskip glue before it, and (when the next line of text is done)
   % \lineskip glue after it.  Thus, space below is not quite equal to space
   % above.  But it's pretty close.
   \def\Egroup{%
       % To get correct interline space between the last line of the group
       % and the first line afterwards, we have to propagate \prevdepth.
       \endgraf % Not \par, as it may have been set to \lisppar.
       \global\dimen1 = \prevdepth
     \egroup           % End the \vtop.
     \addgroupbox
     \prevdepth = \dimen1
     \checkinserts
   }
   
   \def\addgroupbox{
     % \dimen0 is the vertical size of the group's box.
     \dimen0 = \ht\groupbox  \advance\dimen0 by \dp\groupbox
     % \dimen2 is how much space is left on the page (more or less).
     \dimen2 = \txipageheight   \advance\dimen2 by -\pagetotal
     % if the group doesn't fit on the current page, and it's a big big
     % group, force a page break.
     \ifdim \dimen0 > \dimen2
       \ifdim \pagetotal < \vfilllimit\txipageheight
         \page
       \fi
     \fi
     \box\groupbox
   }
   
   %
 % TeX puts in an \escapechar (i.e., `@') at the beginning of the help  % TeX puts in an \escapechar (i.e., `@') at the beginning of the help
 % message, so this ends up printing `@group can only ...'.  % message, so this ends up printing `@group can only ...'.
 %  %
Line 587  where each line of input produces a line of output.}
Line 752  where each line of input produces a line of output.}
   
 \newdimen\mil  \mil=0.001in  \newdimen\mil  \mil=0.001in
   
 \def\need{\parsearg\needx}  \parseargdef\need{%
   
 % Old definition--didn't work.  
 %\def\needx #1{\par %  
 %% This method tries to make TeX break the page naturally  
 %% if the depth of the box does not fit.  
 %{\baselineskip=0pt%  
 %\vtop to #1\mil{\vfil}\kern -#1\mil\nobreak  
 %\prevdepth=-1000pt  
 %}}  
   
 \def\needx#1{%  
   % Ensure vertical mode, so we don't make a big box in the middle of a    % Ensure vertical mode, so we don't make a big box in the middle of a
   % paragraph.    % paragraph.
   \par    \par
Line 637  where each line of input produces a line of output.}
Line 791  where each line of input produces a line of output.}
   \fi    \fi
 }  }
   
 % @br   forces paragraph break  % @br   forces paragraph break (and is undocumented).
   
 \let\br = \par  \let\br = \par
   
 % @dots{} output an ellipsis using the current font.  % @page forces the start of a new page.
 % We do .5em per period so that it has the same spacing in a typewriter  
 % font as three actual period characters.  
 %  %
 \def\dots{%  
   \leavevmode  
   \hbox to 1.5em{%  
     \hskip 0pt plus 0.25fil minus 0.25fil  
     .\hss.\hss.%  
     \hskip 0pt plus 0.5fil minus 0.5fil  
   }%  
 }  
   
 % @enddots{} is an end-of-sentence ellipsis.  
 %  
 \def\enddots{%  
   \leavevmode  
   \hbox to 2em{%  
     \hskip 0pt plus 0.25fil minus 0.25fil  
     .\hss.\hss.\hss.%  
     \hskip 0pt plus 0.5fil minus 0.5fil  
   }%  
   \spacefactor=3000  
 }  
   
   
 % @page    forces the start of a new page  
 %  
 \def\page{\par\vfill\supereject}  \def\page{\par\vfill\supereject}
   
 % @exdent text....  % @exdent text....
Line 679  where each line of input produces a line of output.}
Line 807  where each line of input produces a line of output.}
 \newskip\exdentamount  \newskip\exdentamount
   
 % This defn is used inside fill environments such as @defun.  % This defn is used inside fill environments such as @defun.
 \def\exdent{\parsearg\exdentyyy}  \parseargdef\exdent{\hfil\break\hbox{\kern -\exdentamount{\rm#1}}\hfil\break}
 \def\exdentyyy #1{{\hfil\break\hbox{\kern -\exdentamount{\rm#1}}\hfil\break}}  
   
 % This defn is used inside nofill environments such as @example.  % This defn is used inside nofill environments such as @example.
 \def\nofillexdent{\parsearg\nofillexdentyyy}  \parseargdef\nofillexdent{{\advance \leftskip by -\exdentamount
 \def\nofillexdentyyy #1{{\advance \leftskip by -\exdentamount    \leftline{\hskip\leftskip{\rm#1}}}}
 \leftline{\hskip\leftskip{\rm#1}}}}  
   
 % @inmargin{TEXT} puts TEXT in the margin next to the current paragraph.  % @inmargin{WHICH}{TEXT} puts TEXT in the WHICH margin next to the current
   % paragraph.  For more general purposes, use the \margin insertion
 \def\inmargin#1{%  % class.  WHICH is `l' or `r'.  Not documented, written for gawk manual.
 \strut\vadjust{\nobreak\kern-\strutdepth  %
   \vtop to \strutdepth{\baselineskip\strutdepth\vss  
   \llap{\rightskip=\inmarginspacing \vbox{\noindent #1}}\null}}}  
 \newskip\inmarginspacing \inmarginspacing=1cm  \newskip\inmarginspacing \inmarginspacing=1cm
 \def\strutdepth{\dp\strutbox}  \def\strutdepth{\dp\strutbox}
   %
   \def\doinmargin#1#2{\strut\vadjust{%
     \nobreak
     \kern-\strutdepth
     \vtop to \strutdepth{%
       \baselineskip=\strutdepth
       \vss
       % if you have multiple lines of stuff to put here, you'll need to
       % make the vbox yourself of the appropriate size.
       \ifx#1l%
         \llap{\ignorespaces #2\hskip\inmarginspacing}%
       \else
         \rlap{\hskip\hsize \hskip\inmarginspacing \ignorespaces #2}%
       \fi
       \null
     }%
   }}
   \def\inleftmargin{\doinmargin l}
   \def\inrightmargin{\doinmargin r}
   %
   % @inmargin{TEXT [, RIGHT-TEXT]}
   % (if RIGHT-TEXT is given, use TEXT for left page, RIGHT-TEXT for right;
   % else use TEXT for both).
   %
   \def\inmargin#1{\parseinmargin #1,,\finish}
   \def\parseinmargin#1,#2,#3\finish{% not perfect, but better than nothing.
     \setbox0 = \hbox{\ignorespaces #2}%
     \ifdim\wd0 > 0pt
       \def\lefttext{#1}%  have both texts
       \def\righttext{#2}%
     \else
       \def\lefttext{#1}%  have only one text
       \def\righttext{#1}%
     \fi
     %
     \ifodd\pageno
       \def\temp{\inrightmargin\righttext}% odd page -> outside is right margin
     \else
       \def\temp{\inleftmargin\lefttext}%
     \fi
     \temp
   }
   
 %\hbox{{\rm#1}}\hfil\break}}  % @| inserts a changebar to the left of the current line.  It should
   % surround any changed text.  This approach does *not* work if the
   % change spans more than two lines of output.  To handle that, we would
   % have adopt a much more difficult approach (putting marks into the main
   % vertical list for the beginning and end of each change).  This command
   % is not documented, not supported, and doesn't work.
   %
   \def\|{%
     % \vadjust can only be used in horizontal mode.
     \leavevmode
     %
     % Append this vertical mode material after the current line in the output.
     \vadjust{%
       % We want to insert a rule with the height and depth of the current
       % leading; that is exactly what \strutbox is supposed to record.
       \vskip-\baselineskip
       %
       % \vadjust-items are inserted at the left edge of the type.  So
       % the \llap here moves out into the left-hand margin.
       \llap{%
         %
         % For a thicker or thinner bar, change the `1pt'.
         \vrule height\baselineskip width1pt
         %
         % This is the space between the bar and the text.
         \hskip 12pt
       }%
     }%
   }
   
 % @include file    insert text of that file as input.  % @include FILE -- \input text of FILE.
 % Allow normal characters that  we make active in the argument (a file name).  %
 \def\include{\begingroup  \def\include{\parseargusing\filenamecatcodes\includezzz}
   \catcode`\\=12  \def\includezzz#1{%
   \catcode`~=12    \pushthisfilestack
   \catcode`^=12  
   \catcode`_=12  
   \catcode`|=12  
   \catcode`<=12  
   \catcode`>=12  
   \catcode`+=12  
   \parsearg\includezzz}  
 % Restore active chars for included file.  
 \def\includezzz#1{\endgroup\begingroup  
   % Read the included file in a group so nested @include's work.  
   \def\thisfile{#1}%    \def\thisfile{#1}%
   \input\thisfile    {%
 \endgroup}      \makevalueexpandable  % we want to expand any @value in FILE.
       \turnoffactive        % and allow special characters in the expansion
       \indexnofonts         % Allow `@@' and other weird things in file names.
       \wlog{texinfo.tex: doing @include of #1^^J}%
       \edef\temp{\noexpand\input #1 }%
       %
       % This trickery is to read FILE outside of a group, in case it makes
       % definitions, etc.
       \expandafter
     }\temp
     \popthisfilestack
   }
   \def\filenamecatcodes{%
     \catcode`\\=\other
     \catcode`~=\other
     \catcode`^=\other
     \catcode`_=\other
     \catcode`|=\other
     \catcode`<=\other
     \catcode`>=\other
     \catcode`+=\other
     \catcode`-=\other
     \catcode`\`=\other
     \catcode`\'=\other
   }
   
   \def\pushthisfilestack{%
     \expandafter\pushthisfilestackX\popthisfilestack\StackTerm
   }
   \def\pushthisfilestackX{%
     \expandafter\pushthisfilestackY\thisfile\StackTerm
   }
   \def\pushthisfilestackY #1\StackTerm #2\StackTerm {%
     \gdef\popthisfilestack{\gdef\thisfile{#1}\gdef\popthisfilestack{#2}}%
   }
   
   \def\popthisfilestack{\errthisfilestackempty}
   \def\errthisfilestackempty{\errmessage{Internal error:
     the stack of filenames is empty.}}
   %
 \def\thisfile{}  \def\thisfile{}
   
 % @center line   outputs that line, centered  % @center line
   % outputs that line, centered.
   %
   \parseargdef\center{%
     \ifhmode
       \let\centersub\centerH
     \else
       \let\centersub\centerV
     \fi
     \centersub{\hfil \ignorespaces#1\unskip \hfil}%
     \let\centersub\relax % don't let the definition persist, just in case
   }
   \def\centerH#1{{%
     \hfil\break
     \advance\hsize by -\leftskip
     \advance\hsize by -\rightskip
     \line{#1}%
     \break
   }}
   %
   \newcount\centerpenalty
   \def\centerV#1{%
     % The idea here is the same as in \startdefun, \cartouche, etc.: if
     % @center is the first thing after a section heading, we need to wipe
     % out the negative parskip inserted by \sectionheading, but still
     % prevent a page break here.
     \centerpenalty = \lastpenalty
     \ifnum\centerpenalty>10000 \vskip\parskip \fi
     \ifnum\centerpenalty>9999 \penalty\centerpenalty \fi
     \line{\kern\leftskip #1\kern\rightskip}%
   }
   
 \def\center{\parsearg\centerzzz}  
 \def\centerzzz #1{{\advance\hsize by -\leftskip  
 \advance\hsize by -\rightskip  
 \centerline{#1}}}  
   
 % @sp n   outputs n lines of vertical space  % @sp n   outputs n lines of vertical space
   %
   \parseargdef\sp{\vskip #1\baselineskip}
   
 \def\sp{\parsearg\spxxx}  
 \def\spxxx #1{\vskip #1\baselineskip}  
   
 % @comment ...line which is ignored...  % @comment ...line which is ignored...
 % @c is the same as @comment  % @c is the same as @comment
 % @ignore ... @end ignore  is another way to write a comment  % @ignore ... @end ignore  is another way to write a comment
   %
   \def\comment{\begingroup \catcode`\^^M=\active%
   \catcode`\@=\other \catcode`\{=\other \catcode`\}=\other\commentxxx}%
   
 \def\comment{\begingroup \catcode`\^^M=\other%  {\catcode`\^^M=\active%
   \gdef\commentxxx#1^^M{\endgroup%
   \futurelet\nexttoken\commentxxxx}%
   \gdef\commentxxxx{\ifx\nexttoken\aftermacro\expandafter\comment\fi}%
   }
   
   \def\c{\begingroup \catcode`\^^M=\active%
 \catcode`\@=\other \catcode`\{=\other \catcode`\}=\other%  \catcode`\@=\other \catcode`\{=\other \catcode`\}=\other%
 \commentxxx}  \cxxx}
 {\catcode`\^^M=\other \gdef\commentxxx#1^^M{\endgroup}}  {\catcode`\^^M=\active \gdef\cxxx#1^^M{\endgroup}}
   % See comment in \scanmacro about why the definitions of @c and @comment differ
   
 \let\c=\comment  
   
 % @paragraphindent NCHARS  % @paragraphindent NCHARS
 % We'll use ems for NCHARS, close enough.  % We'll use ems for NCHARS, close enough.
 % We cannot implement @paragraphindent asis, though.  % NCHARS can also be the word `asis' or `none'.
 %  % We cannot feasibly implement @paragraphindent asis, though.
   %
 \def\asisword{asis} % no translation, these are keywords  \def\asisword{asis} % no translation, these are keywords
 \def\noneword{none}  \def\noneword{none}
 %  %
 \def\paragraphindent{\parsearg\doparagraphindent}  \parseargdef\paragraphindent{%
 \def\doparagraphindent#1{%  
   \def\temp{#1}%    \def\temp{#1}%
   \ifx\temp\asisword    \ifx\temp\asisword
   \else    \else
Line 767  where each line of input produces a line of output.}
Line 1021  where each line of input produces a line of output.}
 % We'll use ems for NCHARS like @paragraphindent.  % We'll use ems for NCHARS like @paragraphindent.
 % It seems @exampleindent asis isn't necessary, but  % It seems @exampleindent asis isn't necessary, but
 % I preserve it to make it similar to @paragraphindent.  % I preserve it to make it similar to @paragraphindent.
 \def\exampleindent{\parsearg\doexampleindent}  \parseargdef\exampleindent{%
 \def\doexampleindent#1{%  
   \def\temp{#1}%    \def\temp{#1}%
   \ifx\temp\asisword    \ifx\temp\asisword
   \else    \else
Line 780  where each line of input produces a line of output.}
Line 1033  where each line of input produces a line of output.}
   \fi    \fi
 }  }
   
 % @asis just yields its argument.  Used with @table, for example.  % @firstparagraphindent WORD
   % If WORD is `none', then suppress indentation of the first paragraph
   % after a section heading.  If WORD is `insert', then do indent at such
   % paragraphs.
 %  %
 \def\asis#1{#1}  % The paragraph indentation is suppressed or not by calling
   % \suppressfirstparagraphindent, which the sectioning commands do.
   % We switch the definition of this back and forth according to WORD.
   % By default, we suppress indentation.
   %
   \def\suppressfirstparagraphindent{\dosuppressfirstparagraphindent}
   \def\insertword{insert}
   %
   \parseargdef\firstparagraphindent{%
     \def\temp{#1}%
     \ifx\temp\noneword
       \let\suppressfirstparagraphindent = \dosuppressfirstparagraphindent
     \else\ifx\temp\insertword
       \let\suppressfirstparagraphindent = \relax
     \else
       \errhelp = \EMsimple
       \errmessage{Unknown @firstparagraphindent option `\temp'}%
     \fi\fi
   }
   
 % @math means output in math mode.  % Here is how we actually suppress indentation.  Redefine \everypar to
 % We don't use $'s directly in the definition of \math because control  % \kern backwards by \parindent, and then reset itself to empty.
 % sequences like \math are expanded when the toc file is written.  Then,  
 % we read the toc file back, the $'s will be normal characters (as they  
 % should be, according to the definition of Texinfo).  So we must use a  
 % control sequence to switch into and out of math mode.  
 %  %
 % This isn't quite enough for @math to work properly in indices, but it  % We also make \indent itself not actually do anything until the next
 % seems unlikely it will ever be needed there.  % paragraph.
 %  %
 \let\implicitmath = $  \gdef\dosuppressfirstparagraphindent{%
 \def\math#1{\implicitmath #1\implicitmath}    \gdef\indent  {\restorefirstparagraphindent \indent}%
     \gdef\noindent{\restorefirstparagraphindent \noindent}%
     \global\everypar = {\kern -\parindent \restorefirstparagraphindent}%
   }
   %
   \gdef\restorefirstparagraphindent{%
     \global\let\indent = \ptexindent
     \global\let\noindent = \ptexnoindent
     \global\everypar = {}%
   }
   
 % @bullet and @minus need the same treatment as @math, just above.  
 \def\bullet{\implicitmath\ptexbullet\implicitmath}  
 \def\minus{\implicitmath-\implicitmath}  
   
 % @refill is a no-op.  % @refill is a no-op.
 \let\refill=\relax  \let\refill=\relax
   
 % If working on a large document in chapters, it is convenient to  % @setfilename INFO-FILENAME - ignored
 % be able to disable indexing, cross-referencing, and contents, for test runs.  \let\setfilename=\comment
 % This is done with @novalidate (before @setfilename).  
 %  
 \newif\iflinks \linkstrue % by default we want the aux files.  
 \let\novalidate = \linksfalse  
   
 % @setfilename is done at the beginning of every texinfo file.  
 % So open here the files we need to have open while reading the input.  
 % This makes it possible to make a .fmt file for texinfo.  
 \def\setfilename{%  
    \iflinks  
      \readauxfile  
    \fi % \openindices needs to do some work in any case.  
    \openindices  
    \fixbackslash  % Turn off hack to swallow `\input texinfo'.  
    \global\let\setfilename=\comment % Ignore extra @setfilename cmds.  
    %  
    % If texinfo.cnf is present on the system, read it.  
    % Useful for site-wide @afourpaper, etc.  
    % Just to be on the safe side, close the input stream before the \input.  
    \openin 1 texinfo.cnf  
    \ifeof1 \let\temp=\relax \else \def\temp{\input texinfo.cnf }\fi  
    \closein1  
    \temp  
    %  
    \comment % Ignore the actual filename.  
 }  
   
 % Called from \setfilename.  
 %  
 \def\openindices{%  
   \newindex{cp}%  
   \newcodeindex{fn}%  
   \newcodeindex{vr}%  
   \newcodeindex{tp}%  
   \newcodeindex{ky}%  
   \newcodeindex{pg}%  
 }  
   
 % @bye.  % @bye.
 \outer\def\bye{\pagealignmacro\tracingstats=1\ptexend}  \outer\def\bye{\pagealignmacro\tracingstats=1\ptexend}
   
Line 860  where each line of input produces a line of output.}
Line 1099  where each line of input produces a line of output.}
 \newtoks\toksC  \newtoks\toksC
 \newtoks\toksD  \newtoks\toksD
 \newbox\boxA  \newbox\boxA
   \newbox\boxB
 \newcount\countA  \newcount\countA
 \newif\ifpdf  \newif\ifpdf
 \newif\ifpdfmakepagedest  \newif\ifpdfmakepagedest
   
 \ifx\pdfoutput\undefined  %
   \pdffalse  % For LuaTeX
   \let\pdfmkdest = \gobble  %
   \let\pdfurl = \gobble  
   \let\endlink = \relax  \newif\iftxiuseunicodedestname
   \let\linkcolor = \relax  \txiuseunicodedestnamefalse % For pdfTeX etc.
   \let\pdfmakeoutlines = \relax  
   \ifx\luatexversion\thisisundefined
 \else  \else
   \pdftrue    % Use Unicode destination names
   \pdfoutput = 1    \txiuseunicodedestnametrue
   \input pdfcolor    % Escape PDF strings with converting UTF-16 from UTF-8
     \begingroup
       \catcode`\%=12
       \directlua{
         function UTF16oct(str)
           tex.sprint(string.char(0x5c) .. '376' .. string.char(0x5c) .. '377')
           for c in string.utfvalues(str) do
             if c < 0x10000 then
               tex.sprint(
                 string.format(string.char(0x5c) .. string.char(0x25) .. '03o' ..
                               string.char(0x5c) .. string.char(0x25) .. '03o',
                               (c / 256), (c % 256)))
             else
               c = c - 0x10000
               local c_hi = c / 1024 + 0xd800
               local c_lo = c % 1024 + 0xdc00
               tex.sprint(
                 string.format(string.char(0x5c) .. string.char(0x25) .. '03o' ..
                               string.char(0x5c) .. string.char(0x25) .. '03o' ..
                               string.char(0x5c) .. string.char(0x25) .. '03o' ..
                               string.char(0x5c) .. string.char(0x25) .. '03o',
                               (c_hi / 256), (c_hi % 256),
                               (c_lo / 256), (c_lo % 256)))
             end
           end
         end
       }
     \endgroup
     \def\pdfescapestrutfsixteen#1{\directlua{UTF16oct('\luaescapestring{#1}')}}
     % Escape PDF strings without converting
     \begingroup
       \directlua{
         function PDFescstr(str)
           for c in string.bytes(str) do
             if c <= 0x20 or c >= 0x80 or c == 0x28 or c == 0x29 or c == 0x5c then
               tex.sprint(
                 string.format(string.char(0x5c) .. string.char(0x25) .. '03o',
                               c))
             else
               tex.sprint(string.char(c))
             end
           end
         end
       }
     \endgroup
     \def\pdfescapestring#1{\directlua{PDFescstr('\luaescapestring{#1}')}}
     \ifnum\luatexversion>84
       % For LuaTeX >= 0.85
       \def\pdfdest{\pdfextension dest}
       \let\pdfoutput\outputmode
       \def\pdfliteral{\pdfextension literal}
       \def\pdfcatalog{\pdfextension catalog}
       \def\pdftexversion{\numexpr\pdffeedback version\relax}
       \let\pdfximage\saveimageresource
       \let\pdfrefximage\useimageresource
       \let\pdflastximage\lastsavedimageresourceindex
       \def\pdfendlink{\pdfextension endlink\relax}
       \def\pdfoutline{\pdfextension outline}
       \def\pdfstartlink{\pdfextension startlink}
       \def\pdffontattr{\pdfextension fontattr}
       \def\pdfobj{\pdfextension obj}
       \def\pdflastobj{\numexpr\pdffeedback lastobj\relax}
       \let\pdfpagewidth\pagewidth
       \let\pdfpageheight\pageheight
       \edef\pdfhorigin{\pdfvariable horigin}
       \edef\pdfvorigin{\pdfvariable vorigin}
     \fi
   \fi
   
   % when pdftex is run in dvi mode, \pdfoutput is defined (so \pdfoutput=1
   % can be set).  So we test for \relax and 0 as well as being undefined.
   \ifx\pdfoutput\thisisundefined
   \else
     \ifx\pdfoutput\relax
     \else
       \ifcase\pdfoutput
       \else
         \pdftrue
       \fi
     \fi
   \fi
   
   % PDF uses PostScript string constants for the names of xref targets,
   % for display in the outlines, and in other places.  Thus, we have to
   % double any backslashes.  Otherwise, a name like "\node" will be
   % interpreted as a newline (\n), followed by o, d, e.  Not good.
   %
   % See http://www.ntg.nl/pipermail/ntg-pdftex/2004-July/000654.html and
   % related messages.  The final outcome is that it is up to the TeX user
   % to double the backslashes and otherwise make the string valid, so
   % that's what we do.  pdftex 1.30.0 (ca.2005) introduced a primitive to
   % do this reliably, so we use it.
   
   % #1 is a control sequence in which to do the replacements,
   % which we \xdef.
   \def\txiescapepdf#1{%
     \ifx\pdfescapestring\thisisundefined
       % No primitive available; should we give a warning or log?
       % Many times it won't matter.
       \xdef#1{#1}%
     \else
       % The expandable \pdfescapestring primitive escapes parentheses,
       % backslashes, and other special chars.
       \xdef#1{\pdfescapestring{#1}}%
     \fi
   }
   \def\txiescapepdfutfsixteen#1{%
     \ifx\pdfescapestrutfsixteen\thisisundefined
       % No UTF-16 converting macro available.
       \txiescapepdf{#1}%
     \else
       \xdef#1{\pdfescapestrutfsixteen{#1}}%
     \fi
   }
   
   \newhelp\nopdfimagehelp{Texinfo supports .png, .jpg, .jpeg, and .pdf images
   with PDF output, and none of those formats could be found.  (.eps cannot
   be supported due to the design of the PDF format; use regular TeX (DVI
   output) for that.)}
   
   \ifpdf
     %
     % Color manipulation macros using ideas from pdfcolor.tex,
     % except using rgb instead of cmyk; the latter is said to render as a
     % very dark gray on-screen and a very dark halftone in print, instead
     % of actual black. The dark red here is dark enough to print on paper as
     % nearly black, but still distinguishable for online viewing.  We use
     % black by default, though.
     \def\rgbDarkRed{0.50 0.09 0.12}
     \def\rgbBlack{0 0 0}
     %
     % rg sets the color for filling (usual text, etc.);
     % RG sets the color for stroking (thin rules, e.g., normal _'s).
     \def\pdfsetcolor#1{\pdfliteral{#1 rg  #1 RG}}
     %
     % Set color, and create a mark which defines \thiscolor accordingly,
     % so that \makeheadline knows which color to restore.
     \def\setcolor#1{%
       \xdef\lastcolordefs{\gdef\noexpand\thiscolor{#1}}%
       \domark
       \pdfsetcolor{#1}%
     }
     %
     \def\maincolor{\rgbBlack}
     \pdfsetcolor{\maincolor}
     \edef\thiscolor{\maincolor}
     \def\lastcolordefs{}
     %
     \def\makefootline{%
       \baselineskip24pt
       \line{\pdfsetcolor{\maincolor}\the\footline}%
     }
     %
     \def\makeheadline{%
       \vbox to 0pt{%
         \vskip-22.5pt
         \line{%
           \vbox to8.5pt{}%
           % Extract \thiscolor definition from the marks.
           \getcolormarks
           % Typeset the headline with \maincolor, then restore the color.
           \pdfsetcolor{\maincolor}\the\headline\pdfsetcolor{\thiscolor}%
         }%
         \vss
       }%
       \nointerlineskip
     }
     %
     %
     \pdfcatalog{/PageMode /UseOutlines}
     %
     % #1 is image name, #2 width (might be empty/whitespace), #3 height (ditto).
   \def\dopdfimage#1#2#3{%    \def\dopdfimage#1#2#3{%
     \def\imagewidth{#2}%      \def\pdfimagewidth{#2}\setbox0 = \hbox{\ignorespaces #2}%
     \def\imageheight{#3}%      \def\pdfimageheight{#3}\setbox2 = \hbox{\ignorespaces #3}%
       %
       % pdftex (and the PDF format) support .pdf, .png, .jpg (among
       % others).  Let's try in that order, PDF first since if
       % someone has a scalable image, presumably better to use that than a
       % bitmap.
       \let\pdfimgext=\empty
       \begingroup
         \openin 1 #1.pdf \ifeof 1
           \openin 1 #1.PDF \ifeof 1
             \openin 1 #1.png \ifeof 1
               \openin 1 #1.jpg \ifeof 1
                 \openin 1 #1.jpeg \ifeof 1
                   \openin 1 #1.JPG \ifeof 1
                     \errhelp = \nopdfimagehelp
                     \errmessage{Could not find image file #1 for pdf}%
                   \else \gdef\pdfimgext{JPG}%
                   \fi
                 \else \gdef\pdfimgext{jpeg}%
                 \fi
               \else \gdef\pdfimgext{jpg}%
               \fi
             \else \gdef\pdfimgext{png}%
             \fi
           \else \gdef\pdfimgext{PDF}%
           \fi
         \else \gdef\pdfimgext{pdf}%
         \fi
         \closein 1
       \endgroup
       %
       % without \immediate, ancient pdftex seg faults when the same image is
       % included twice.  (Version 3.14159-pre-1.0-unofficial-20010704.)
     \ifnum\pdftexversion < 14      \ifnum\pdftexversion < 14
       \pdfimage        \immediate\pdfimage
     \else      \else
       \pdfximage        \immediate\pdfximage
     \fi      \fi
       \ifx\empty\imagewidth\else width \imagewidth \fi        \ifdim \wd0 >0pt width \pdfimagewidth \fi
       \ifx\empty\imageheight\else height \imageheight \fi        \ifdim \wd2 >0pt height \pdfimageheight \fi
       {#1.pdf}%        \ifnum\pdftexversion<13
            #1.\pdfimgext
          \else
            {#1.\pdfimgext}%
          \fi
     \ifnum\pdftexversion < 14 \else      \ifnum\pdftexversion < 14 \else
       \pdfrefximage \pdflastximage        \pdfrefximage \pdflastximage
     \fi}      \fi}
   \def\pdfmkdest#1{\pdfdest name{#1@} xyz}    %
   \def\pdfmkpgn#1{#1@}    \def\setpdfdestname#1{{%
   \let\linkcolor = \Cyan      % We have to set dummies so commands such as @code, and characters
   \def\endlink{\Black\pdfendlink}      % such as \, aren't expanded when present in a section title.
       \indexnofonts
       \makevalueexpandable
       \turnoffactive
       \iftxiuseunicodedestname
         \ifx \declaredencoding \latone
           % Pass through Latin-1 characters.
           % LuaTeX with byte wise I/O converts Latin-1 characters to Unicode.
         \else
           \ifx \declaredencoding \utfeight
             % Pass through Unicode characters.
           \else
             % Use ASCII approximations in destination names.
             \passthroughcharsfalse
           \fi
         \fi
       \else
         % Use ASCII approximations in destination names.
         \passthroughcharsfalse
       \fi
       \def\pdfdestname{#1}%
       \txiescapepdf\pdfdestname
     }}
     %
     \def\setpdfoutlinetext#1{{%
       \indexnofonts
       \makevalueexpandable
       \turnoffactive
       \ifx \declaredencoding \latone
         % The PDF format can use an extended form of Latin-1 in bookmark
         % strings.  See Appendix D of the PDF Reference, Sixth Edition, for
         % the "PDFDocEncoding".
         \passthroughcharstrue
         % Pass through Latin-1 characters.
         %   LuaTeX: Convert to Unicode
         %   pdfTeX: Use Latin-1 as PDFDocEncoding
         \def\pdfoutlinetext{#1}%
       \else
         \ifx \declaredencoding \utfeight
           \ifx\luatexversion\thisisundefined
             % For pdfTeX  with UTF-8.
             % TODO: the PDF format can use UTF-16 in bookmark strings,
             % but the code for this isn't done yet.
             % Use ASCII approximations.
             \passthroughcharsfalse
             \def\pdfoutlinetext{#1}%
           \else
             % For LuaTeX with UTF-8.
             % Pass through Unicode characters for title texts.
             \passthroughcharstrue
             \def\pdfoutlinetext{#1}%
           \fi
         \else
           % For non-Latin-1 or non-UTF-8 encodings.
           % Use ASCII approximations.
           \passthroughcharsfalse
           \def\pdfoutlinetext{#1}%
         \fi
       \fi
       % LuaTeX: Convert to UTF-16
       % pdfTeX: Use Latin-1 as PDFDocEncoding
       \txiescapepdfutfsixteen\pdfoutlinetext
     }}
     %
     \def\pdfmkdest#1{%
       \setpdfdestname{#1}%
       \safewhatsit{\pdfdest name{\pdfdestname} xyz}%
     }
     %
     % used to mark target names; must be expandable.
     \def\pdfmkpgn#1{#1}
     %
     % by default, use black for everything.
     \def\urlcolor{\rgbBlack}
     \def\linkcolor{\rgbBlack}
     \def\endlink{\setcolor{\maincolor}\pdfendlink}
     %
   % Adding outlines to PDF; macros for calculating structure of outlines    % Adding outlines to PDF; macros for calculating structure of outlines
   % come from Petr Olsak    % come from Petr Olsak
   \def\expnumber#1{\expandafter\ifx\csname#1\endcsname\relax 0%    \def\expnumber#1{\expandafter\ifx\csname#1\endcsname\relax 0%
     \else \csname#1\endcsname \fi}      \else \csname#1\endcsname \fi}
   \def\advancenumber#1{\tempnum=\expnumber{#1}\relax    \def\advancenumber#1{\tempnum=\expnumber{#1}\relax
     \advance\tempnum by1      \advance\tempnum by 1
     \expandafter\xdef\csname#1\endcsname{\the\tempnum}}      \expandafter\xdef\csname#1\endcsname{\the\tempnum}}
   \def\pdfmakeoutlines{{%    %
     \openin 1 \jobname.toc    % #1 is the section text, which is what will be displayed in the
     \ifeof 1\else\bgroup    % outline by the pdf viewer.  #2 is the pdf expression for the number
       \closein 1    % of subentries (or empty, for subsubsections).  #3 is the node text,
       \indexnofonts    % which might be empty if this toc entry had no corresponding node.
       \def\tt{}    % #4 is the page number
       % thanh's hack / proper braces in bookmarks    %
       \edef\mylbrace{\iftrue \string{\else}\fi}\let\{=\mylbrace    \def\dopdfoutline#1#2#3#4{%
       \edef\myrbrace{\iffalse{\else\string}\fi}\let\}=\myrbrace      % Generate a link to the node text if that exists; else, use the
       %      % page number.  We could generate a destination for the section
       \def\chapentry ##1##2##3{}      % text in the case where a section has no node, but it doesn't
       \def\unnumbchapentry ##1##2{}      % seem worth the trouble, since most documents are normally structured.
       \def\secentry ##1##2##3##4{\advancenumber{chap##2}}      \setpdfoutlinetext{#1}
       \def\unnumbsecentry ##1##2{}      \setpdfdestname{#3}
       \def\subsecentry ##1##2##3##4##5{\advancenumber{sec##2.##3}}      \ifx\pdfdestname\empty
       \def\unnumbsubsecentry ##1##2{}        \def\pdfdestname{#4}%
       \def\subsubsecentry ##1##2##3##4##5##6{\advancenumber{subsec##2.##3.##4}}  
       \def\unnumbsubsubsecentry ##1##2{}  
       \input \jobname.toc  
       \def\chapentry ##1##2##3{%  
         \pdfoutline goto name{\pdfmkpgn{##3}}count-\expnumber{chap##2}{##1}}  
       \def\unnumbchapentry ##1##2{%  
         \pdfoutline goto name{\pdfmkpgn{##2}}{##1}}  
       \def\secentry ##1##2##3##4{%  
         \pdfoutline goto name{\pdfmkpgn{##4}}count-\expnumber{sec##2.##3}{##1}}  
       \def\unnumbsecentry ##1##2{%  
         \pdfoutline goto name{\pdfmkpgn{##2}}{##1}}  
       \def\subsecentry ##1##2##3##4##5{%  
         \pdfoutline goto name{\pdfmkpgn{##5}}count-\expnumber{subsec##2.##3.##4}{##1}}  
       \def\unnumbsubsecentry ##1##2{%  
         \pdfoutline goto name{\pdfmkpgn{##2}}{##1}}  
       \def\subsubsecentry ##1##2##3##4##5##6{%  
         \pdfoutline goto name{\pdfmkpgn{##6}}{##1}}  
       \def\unnumbsubsubsecentry ##1##2{%  
         \pdfoutline goto name{\pdfmkpgn{##2}}{##1}}  
       \input \jobname.toc  
     \egroup\fi  
   }}  
   \def\makelinks #1,{%  
     \def\params{#1}\def\E{END}%  
     \ifx\params\E  
       \let\nextmakelinks=\relax  
     \else  
       \let\nextmakelinks=\makelinks  
       \ifnum\lnkcount>0,\fi  
       \picknum{#1}%  
       \startlink attr{/Border [0 0 0]}  
         goto name{\pdfmkpgn{\the\pgn}}%  
       \linkcolor #1%  
       \advance\lnkcount by 1%  
       \endlink  
     \fi      \fi
     \nextmakelinks      %
       \pdfoutline goto name{\pdfmkpgn{\pdfdestname}}#2{\pdfoutlinetext}%
   }    }
   \def\picknum#1{\expandafter\pn#1}    %
   \def\pn#1{%    \def\pdfmakeoutlines{%
     \def\p{#1}%      \begingroup
     \ifx\p\lbrace        % Read toc silently, to get counts of subentries for \pdfoutline.
       \let\nextpn=\ppn        \def\partentry##1##2##3##4{}% ignore parts in the outlines
     \else        \def\numchapentry##1##2##3##4{%
       \let\nextpn=\ppnn          \def\thischapnum{##2}%
       \def\first{#1}          \def\thissecnum{0}%
     \fi          \def\thissubsecnum{0}%
     \nextpn        }%
         \def\numsecentry##1##2##3##4{%
           \advancenumber{chap\thischapnum}%
           \def\thissecnum{##2}%
           \def\thissubsecnum{0}%
         }%
         \def\numsubsecentry##1##2##3##4{%
           \advancenumber{sec\thissecnum}%
           \def\thissubsecnum{##2}%
         }%
         \def\numsubsubsecentry##1##2##3##4{%
           \advancenumber{subsec\thissubsecnum}%
         }%
         \def\thischapnum{0}%
         \def\thissecnum{0}%
         \def\thissubsecnum{0}%
         %
         % use \def rather than \let here because we redefine \chapentry et
         % al. a second time, below.
         \def\appentry{\numchapentry}%
         \def\appsecentry{\numsecentry}%
         \def\appsubsecentry{\numsubsecentry}%
         \def\appsubsubsecentry{\numsubsubsecentry}%
         \def\unnchapentry{\numchapentry}%
         \def\unnsecentry{\numsecentry}%
         \def\unnsubsecentry{\numsubsecentry}%
         \def\unnsubsubsecentry{\numsubsubsecentry}%
         \readdatafile{toc}%
         %
         % Read toc second time, this time actually producing the outlines.
         % The `-' means take the \expnumber as the absolute number of
         % subentries, which we calculated on our first read of the .toc above.
         %
         % We use the node names as the destinations.
         \def\numchapentry##1##2##3##4{%
           \dopdfoutline{##1}{count-\expnumber{chap##2}}{##3}{##4}}%
         \def\numsecentry##1##2##3##4{%
           \dopdfoutline{##1}{count-\expnumber{sec##2}}{##3}{##4}}%
         \def\numsubsecentry##1##2##3##4{%
           \dopdfoutline{##1}{count-\expnumber{subsec##2}}{##3}{##4}}%
         \def\numsubsubsecentry##1##2##3##4{% count is always zero
           \dopdfoutline{##1}{}{##3}{##4}}%
         %
         % PDF outlines are displayed using system fonts, instead of
         % document fonts.  Therefore we cannot use special characters,
         % since the encoding is unknown.  For example, the eogonek from
         % Latin 2 (0xea) gets translated to a | character.  Info from
         % Staszek Wawrykiewicz, 19 Jan 2004 04:09:24 +0100.
         %
         % TODO this right, we have to translate 8-bit characters to
         % their "best" equivalent, based on the @documentencoding.  Too
         % much work for too little return.  Just use the ASCII equivalents
         % we use for the index sort strings.
         %
         \indexnofonts
         \setupdatafile
         % We can have normal brace characters in the PDF outlines, unlike
         % Texinfo index files.  So set that up.
         \def\{{\lbracecharliteral}%
         \def\}{\rbracecharliteral}%
         \catcode`\\=\active \otherbackslash
         \input \tocreadfilename
       \endgroup
   }    }
   \def\ppn#1{\pgn=#1\gobble}    {\catcode`[=1 \catcode`]=2
   \def\ppnn{\pgn=\first}     \catcode`{=\other \catcode`}=\other
   \def\pdfmklnk#1{\lnkcount=0\makelinks #1,END,}     \gdef\lbracecharliteral[{]%
   \def\addtokens#1#2{\edef\addtoks{\noexpand#1={\the#1#2}}\addtoks}     \gdef\rbracecharliteral[}]%
     ]
     %
   \def\skipspaces#1{\def\PP{#1}\def\D{|}%    \def\skipspaces#1{\def\PP{#1}\def\D{|}%
     \ifx\PP\D\let\nextsp\relax      \ifx\PP\D\let\nextsp\relax
     \else\let\nextsp\skipspaces      \else\let\nextsp\skipspaces
       \ifx\p\space\else\addtokens{\filename}{\PP}%        \addtokens{\filename}{\PP}%
         \advance\filenamelength by 1        \advance\filenamelength by 1
       \fi  
     \fi      \fi
     \nextsp}      \nextsp}
   \def\getfilename#1{\filenamelength=0\expandafter\skipspaces#1|\relax}    \def\getfilename#1{%
       \filenamelength=0
       % If we don't expand the argument now, \skipspaces will get
       % snagged on things like "@value{foo}".
       \edef\temp{#1}%
       \expandafter\skipspaces\temp|\relax
     }
   \ifnum\pdftexversion < 14    \ifnum\pdftexversion < 14
     \let \startlink \pdfannotlink      \let \startlink \pdfannotlink
   \else    \else
     \let \startlink \pdfstartlink      \let \startlink \pdfstartlink
   \fi    \fi
     % make a live url in pdf output.
   \def\pdfurl#1{%    \def\pdfurl#1{%
     \begingroup      \begingroup
       \normalturnoffactive\def\@{@}%        % it seems we really need yet another set of dummies; have not
       \leavevmode\Red        % tried to figure out what each command should do in the context
         % of @url.  for now, just make @/ a no-op, that's the only one
         % people have actually reported a problem with.
         %
         \normalturnoffactive
         \def\@{@}%
         \let\/=\empty
         \makevalueexpandable
         % do we want to go so far as to use \indexnofonts instead of just
         % special-casing \var here?
         \def\var##1{##1}%
         %
         \leavevmode\setcolor{\urlcolor}%
       \startlink attr{/Border [0 0 0]}%        \startlink attr{/Border [0 0 0]}%
         user{/Subtype /Link /A << /S /URI /URI (#1) >>}%          user{/Subtype /Link /A << /S /URI /URI (#1) >>}%
         % #1  
     \endgroup}      \endgroup}
   \def\pdfgettoks#1.{\setbox\boxA=\hbox{\toksA={#1.}\toksB={}\maketoks}}    \def\pdfgettoks#1.{\setbox\boxA=\hbox{\toksA={#1.}\toksB={}\maketoks}}
   \def\addtokens#1#2{\edef\addtoks{\noexpand#1={\the#1#2}}\addtoks}    \def\addtokens#1#2{\edef\addtoks{\noexpand#1={\the#1#2}}\addtoks}
   \def\adn#1{\addtokens{\toksC}{#1}\global\countA=1\let\next=\maketoks}    \def\adn#1{\addtokens{\toksC}{#1}\global\countA=1\let\next=\maketoks}
   \def\poptoks#1#2|ENDTOKS|{\let\first=#1\toksD={#1}\toksA={#2}}    \def\poptoks#1#2|ENDTOKS|{\let\first=#1\toksD={#1}\toksA={#2}}
   \def\maketoks{%    \def\maketoks{%
     \expandafter\poptoks\the\toksA|ENDTOKS|      \expandafter\poptoks\the\toksA|ENDTOKS|\relax
     \ifx\first0\adn0      \ifx\first0\adn0
     \else\ifx\first1\adn1 \else\ifx\first2\adn2 \else\ifx\first3\adn3      \else\ifx\first1\adn1 \else\ifx\first2\adn2 \else\ifx\first3\adn3
     \else\ifx\first4\adn4 \else\ifx\first5\adn5 \else\ifx\first6\adn6      \else\ifx\first4\adn4 \else\ifx\first5\adn5 \else\ifx\first6\adn6
     \else\ifx\first7\adn7 \else\ifx\first8\adn8 \else\ifx\first9\adn9      \else\ifx\first7\adn7 \else\ifx\first8\adn8 \else\ifx\first9\adn9
     \else      \else
       \ifnum0=\countA\else\makelink\fi        \ifnum0=\countA\else\makelink\fi
       \ifx\first.\let\next=\done\else        \ifx\first.\let\next=\done\else
Line 1013  where each line of input produces a line of output.}
Line 1584  where each line of input produces a line of output.}
   \def\makelink{\addtokens{\toksB}%    \def\makelink{\addtokens{\toksB}%
     {\noexpand\pdflink{\the\toksC}}\toksC={}\global\countA=0}      {\noexpand\pdflink{\the\toksC}}\toksC={}\global\countA=0}
   \def\pdflink#1{%    \def\pdflink#1{%
     \startlink attr{/Border [0 0 0]} goto name{\mkpgn{#1}}      \startlink attr{/Border [0 0 0]} goto name{\pdfmkpgn{#1}}
     \linkcolor #1\endlink}      \setcolor{\linkcolor}#1\endlink}
   \def\mkpgn#1{#1@}  
   \def\done{\edef\st{\global\noexpand\toksA={\the\toksB}}\st}    \def\done{\edef\st{\global\noexpand\toksA={\the\toksB}}\st}
 \fi % \ifx\pdfoutput  \else
     % non-pdf mode
     \let\pdfmkdest = \gobble
     \let\pdfurl = \gobble
     \let\endlink = \relax
     \let\setcolor = \gobble
     \let\pdfsetcolor = \gobble
     \let\pdfmakeoutlines = \relax
   \fi  % \ifx\pdfoutput
   
   %
   % For XeTeX
   %
   \ifx\XeTeXrevision\thisisundefined
   \else
     %
     % XeTeX version check
     %
     \ifnum\strcmp{\the\XeTeXversion\XeTeXrevision}{0.99996}>-1
       % TeX Live 2016 contains XeTeX 0.99996 and xdvipdfmx 20160307.
       % It can use the `dvipdfmx:config' special (from TeX Live SVN r40941).
       % For avoiding PDF destination name replacement, we use this special
       % instead of xdvipdfmx's command line option `-C 0x0010'.
       \special{dvipdfmx:config C 0x0010}
       % XeTeX 0.99995+ comes with xdvipdfmx 20160307+.
       % It can handle Unicode destination names for PDF.
       \txiuseunicodedestnametrue
     \else
       % XeTeX < 0.99996 (TeX Live < 2016) cannot use the
       % `dvipdfmx:config' special.
       % So for avoiding PDF destination name replacement,
       % xdvipdfmx's command line option `-C 0x0010' is necessary.
       %
       % XeTeX < 0.99995 can not handle Unicode destination names for PDF
       % because xdvipdfmx 20150315 has a UTF-16 conversion issue.
       % It is fixed by xdvipdfmx 20160106 (TeX Live SVN r39753).
       \txiuseunicodedestnamefalse
     \fi
     %
     % Color support
     %
     \def\rgbDarkRed{0.50 0.09 0.12}
     \def\rgbBlack{0 0 0}
     %
     \def\pdfsetcolor#1{\special{pdf:scolor [#1]}}
     %
     % Set color, and create a mark which defines \thiscolor accordingly,
     % so that \makeheadline knows which color to restore.
     \def\setcolor#1{%
       \xdef\lastcolordefs{\gdef\noexpand\thiscolor{#1}}%
       \domark
       \pdfsetcolor{#1}%
     }
     %
     \def\maincolor{\rgbBlack}
     \pdfsetcolor{\maincolor}
     \edef\thiscolor{\maincolor}
     \def\lastcolordefs{}
     %
     \def\makefootline{%
       \baselineskip24pt
       \line{\pdfsetcolor{\maincolor}\the\footline}%
     }
     %
     \def\makeheadline{%
       \vbox to 0pt{%
         \vskip-22.5pt
         \line{%
           \vbox to8.5pt{}%
           % Extract \thiscolor definition from the marks.
           \getcolormarks
           % Typeset the headline with \maincolor, then restore the color.
           \pdfsetcolor{\maincolor}\the\headline\pdfsetcolor{\thiscolor}%
         }%
         \vss
       }%
       \nointerlineskip
     }
     %
     % PDF outline support
     %
     % Emulate pdfTeX primitive
     \def\pdfdest name#1 xyz{%
       \special{pdf:dest (#1) [@thispage /XYZ @xpos @ypos null]}%
     }
     %
     \def\setpdfdestname#1{{%
       % We have to set dummies so commands such as @code, and characters
       % such as \, aren't expanded when present in a section title.
       \indexnofonts
       \makevalueexpandable
       \turnoffactive
       \iftxiuseunicodedestname
         % Pass through Unicode characters.
       \else
         % Use ASCII approximations in destination names.
         \passthroughcharsfalse
       \fi
       \def\pdfdestname{#1}%
       \txiescapepdf\pdfdestname
     }}
     %
     \def\setpdfoutlinetext#1{{%
       \turnoffactive
       % Always use Unicode characters in title texts.
       \def\pdfoutlinetext{#1}%
       % For XeTeX, xdvipdfmx converts to UTF-16.
       % So we do not convert.
       \txiescapepdf\pdfoutlinetext
     }}
     %
     \def\pdfmkdest#1{%
       \setpdfdestname{#1}%
       \safewhatsit{\pdfdest name{\pdfdestname} xyz}%
     }
     %
     % by default, use black for everything.
     \def\urlcolor{\rgbBlack}
     \def\linkcolor{\rgbBlack}
     \def\endlink{\setcolor{\maincolor}\pdfendlink}
     %
     \def\dopdfoutline#1#2#3#4{%
       \setpdfoutlinetext{#1}
       \setpdfdestname{#3}
       \ifx\pdfdestname\empty
         \def\pdfdestname{#4}%
       \fi
       %
       \special{pdf:out [-] #2 << /Title (\pdfoutlinetext) /A
         << /S /GoTo /D (\pdfdestname) >> >> }%
     }
     %
     \def\pdfmakeoutlines{%
       \begingroup
         %
         % For XeTeX, counts of subentries are not necessary.
         % Therefore, we read toc only once.
         %
         % We use node names as destinations.
         \def\partentry##1##2##3##4{}% ignore parts in the outlines
         \def\numchapentry##1##2##3##4{%
           \dopdfoutline{##1}{1}{##3}{##4}}%
         \def\numsecentry##1##2##3##4{%
           \dopdfoutline{##1}{2}{##3}{##4}}%
         \def\numsubsecentry##1##2##3##4{%
           \dopdfoutline{##1}{3}{##3}{##4}}%
         \def\numsubsubsecentry##1##2##3##4{%
           \dopdfoutline{##1}{4}{##3}{##4}}%
         %
         \let\appentry\numchapentry%
         \let\appsecentry\numsecentry%
         \let\appsubsecentry\numsubsecentry%
         \let\appsubsubsecentry\numsubsubsecentry%
         \let\unnchapentry\numchapentry%
         \let\unnsecentry\numsecentry%
         \let\unnsubsecentry\numsubsecentry%
         \let\unnsubsubsecentry\numsubsubsecentry%
         %
         % For XeTeX, xdvipdfmx converts strings to UTF-16.
         % Therefore, the encoding and the language may not be considered.
         %
         \indexnofonts
         \setupdatafile
         % We can have normal brace characters in the PDF outlines, unlike
         % Texinfo index files.  So set that up.
         \def\{{\lbracecharliteral}%
         \def\}{\rbracecharliteral}%
         \catcode`\\=\active \otherbackslash
         \input \tocreadfilename
       \endgroup
     }
     {\catcode`[=1 \catcode`]=2
      \catcode`{=\other \catcode`}=\other
      \gdef\lbracecharliteral[{]%
      \gdef\rbracecharliteral[}]%
     ]
   
     \special{pdf:docview << /PageMode /UseOutlines >> }
     % ``\special{pdf:tounicode ...}'' is not necessary
     % because xdvipdfmx converts strings from UTF-8 to UTF-16 without it.
     % However, due to a UTF-16 conversion issue of xdvipdfmx 20150315,
     % ``\special{pdf:dest ...}'' cannot handle non-ASCII strings.
     % It is fixed by xdvipdfmx 20160106 (TeX Live SVN r39753).
   %
     \def\skipspaces#1{\def\PP{#1}\def\D{|}%
       \ifx\PP\D\let\nextsp\relax
       \else\let\nextsp\skipspaces
         \addtokens{\filename}{\PP}%
         \advance\filenamelength by 1
       \fi
       \nextsp}
     \def\getfilename#1{%
       \filenamelength=0
       % If we don't expand the argument now, \skipspaces will get
       % snagged on things like "@value{foo}".
       \edef\temp{#1}%
       \expandafter\skipspaces\temp|\relax
     }
     % make a live url in pdf output.
     \def\pdfurl#1{%
       \begingroup
         % it seems we really need yet another set of dummies; have not
         % tried to figure out what each command should do in the context
         % of @url.  for now, just make @/ a no-op, that's the only one
         % people have actually reported a problem with.
         %
         \normalturnoffactive
         \def\@{@}%
         \let\/=\empty
         \makevalueexpandable
         % do we want to go so far as to use \indexnofonts instead of just
         % special-casing \var here?
         \def\var##1{##1}%
         %
         \leavevmode\setcolor{\urlcolor}%
         \special{pdf:bann << /Border [0 0 0]
           /Subtype /Link /A << /S /URI /URI (#1) >> >>}%
       \endgroup}
     \def\endlink{\setcolor{\maincolor}\special{pdf:eann}}
     \def\pdfgettoks#1.{\setbox\boxA=\hbox{\toksA={#1.}\toksB={}\maketoks}}
     \def\addtokens#1#2{\edef\addtoks{\noexpand#1={\the#1#2}}\addtoks}
     \def\adn#1{\addtokens{\toksC}{#1}\global\countA=1\let\next=\maketoks}
     \def\poptoks#1#2|ENDTOKS|{\let\first=#1\toksD={#1}\toksA={#2}}
     \def\maketoks{%
       \expandafter\poptoks\the\toksA|ENDTOKS|\relax
       \ifx\first0\adn0
       \else\ifx\first1\adn1 \else\ifx\first2\adn2 \else\ifx\first3\adn3
       \else\ifx\first4\adn4 \else\ifx\first5\adn5 \else\ifx\first6\adn6
       \else\ifx\first7\adn7 \else\ifx\first8\adn8 \else\ifx\first9\adn9
       \else
         \ifnum0=\countA\else\makelink\fi
         \ifx\first.\let\next=\done\else
           \let\next=\maketoks
           \addtokens{\toksB}{\the\toksD}
           \ifx\first,\addtokens{\toksB}{\space}\fi
         \fi
       \fi\fi\fi\fi\fi\fi\fi\fi\fi\fi
       \next}
     \def\makelink{\addtokens{\toksB}%
       {\noexpand\pdflink{\the\toksC}}\toksC={}\global\countA=0}
     \def\pdflink#1{%
       \special{pdf:bann << /Border [0 0 0]
         /Type /Annot /Subtype /Link /A << /S /GoTo /D (#1) >> >>}%
       \setcolor{\linkcolor}#1\endlink}
     \def\done{\edef\st{\global\noexpand\toksA={\the\toksB}}\st}
   %
     %
     % @image support
     %
     % #1 is image name, #2 width (might be empty/whitespace), #3 height (ditto).
     \def\doxeteximage#1#2#3{%
       \def\xeteximagewidth{#2}\setbox0 = \hbox{\ignorespaces #2}%
       \def\xeteximageheight{#3}\setbox2 = \hbox{\ignorespaces #3}%
       %
       % XeTeX (and the PDF format) supports .pdf, .png, .jpg (among
       % others).  Let's try in that order, PDF first since if
       % someone has a scalable image, presumably better to use that than a
       % bitmap.
       \let\xeteximgext=\empty
       \begingroup
         \openin 1 #1.pdf \ifeof 1
           \openin 1 #1.PDF \ifeof 1
             \openin 1 #1.png \ifeof 1
               \openin 1 #1.jpg \ifeof 1
                 \openin 1 #1.jpeg \ifeof 1
                   \openin 1 #1.JPG \ifeof 1
                     \errmessage{Could not find image file #1 for XeTeX}%
                   \else \gdef\xeteximgext{JPG}%
                   \fi
                 \else \gdef\xeteximgext{jpeg}%
                 \fi
               \else \gdef\xeteximgext{jpg}%
               \fi
             \else \gdef\xeteximgext{png}%
             \fi
           \else \gdef\xeteximgext{PDF}%
           \fi
         \else \gdef\xeteximgext{pdf}%
         \fi
         \closein 1
       \endgroup
       %
       \def\xetexpdfext{pdf}%
       \ifx\xeteximgext\xetexpdfext
         \XeTeXpdffile "#1".\xeteximgext ""
       \else
         \def\xetexpdfext{PDF}%
         \ifx\xeteximgext\xetexpdfext
           \XeTeXpdffile "#1".\xeteximgext ""
         \else
           \XeTeXpicfile "#1".\xeteximgext ""
         \fi
       \fi
       \ifdim \wd0 >0pt width \xeteximagewidth \fi
       \ifdim \wd2 >0pt height \xeteximageheight \fi \relax
     }
   \fi
   
 \message{fonts,}  \message{fonts,}
 % Font-change commands.  
   
   % Change the current font style to #1, remembering it in \curfontstyle.
   % For now, we do not accumulate font styles: @b{@i{foo}} prints foo in
   % italics, not bold italics.
   %
   \def\setfontstyle#1{%
     \def\curfontstyle{#1}% not as a control sequence, because we are \edef'd.
     \csname ten#1\endcsname  % change the current font
   }
   
   % Select #1 fonts with the current style.
   %
   \def\selectfonts#1{\csname #1fonts\endcsname \csname\curfontstyle\endcsname}
   
   \def\rm{\fam=0 \setfontstyle{rm}}
   \def\it{\fam=\itfam \setfontstyle{it}}
   \def\sl{\fam=\slfam \setfontstyle{sl}}
   \def\bf{\fam=\bffam \setfontstyle{bf}}\def\bfstylename{bf}
   \def\tt{\fam=\ttfam \setfontstyle{tt}}
   
   % Unfortunately, we have to override this for titles and the like, since
   % in those cases "rm" is bold.  Sigh.
   \def\rmisbold{\rm\def\curfontstyle{bf}}
   
 % Texinfo sort of supports the sans serif font style, which plain TeX does not.  % Texinfo sort of supports the sans serif font style, which plain TeX does not.
 % So we set up a \sf analogous to plain's \rm, etc.  % So we set up a \sf.
 \newfam\sffam  \newfam\sffam
 \def\sf{\fam=\sffam \tensf}  \def\sf{\fam=\sffam \setfontstyle{sf}}
 \let\li = \sf % Sometimes we call it \li, not \sf.  \let\li = \sf % Sometimes we call it \li, not \sf.
   
 % We don't need math for this one.  % We don't need math for this font style.
 \def\ttsl{\tenttsl}  \def\ttsl{\setfontstyle{ttsl}}
   
 % Use Computer Modern fonts at \magstephalf (11pt).  
 \newcount\mainmagstep  
 \mainmagstep=\magstephalf  
   
 % Set the font macro #1 to the font named #2, adding on the  % Set the baselineskip to #1, and the lineskip and strut size
 % specified font prefix (normally `cm').  % correspondingly.  There is no deep meaning behind these magic numbers
 % #3 is the font's design size, #4 is a scale factor  % used as factors; they just match (closely enough) what Knuth defined.
 \def\setfont#1#2#3#4{\font#1=\fontprefix#2#3 scaled #4}  %
   \def\lineskipfactor{.08333}
   \def\strutheightpercent{.70833}
   \def\strutdepthpercent {.29167}
   %
   % can get a sort of poor man's double spacing by redefining this.
   \def\baselinefactor{1}
   %
   \newdimen\textleading
   \def\setleading#1{%
     \dimen0 = #1\relax
     \normalbaselineskip = \baselinefactor\dimen0
     \normallineskip = \lineskipfactor\normalbaselineskip
     \normalbaselines
     \setbox\strutbox =\hbox{%
       \vrule width0pt height\strutheightpercent\baselineskip
                       depth \strutdepthpercent \baselineskip
     }%
   }
   
   % PDF CMaps.  See also LaTeX's t1.cmap.
   %
   % do nothing with this by default.
   \expandafter\let\csname cmapOT1\endcsname\gobble
   \expandafter\let\csname cmapOT1IT\endcsname\gobble
   \expandafter\let\csname cmapOT1TT\endcsname\gobble
   
   % if we are producing pdf, and we have \pdffontattr, then define cmaps.
   % (\pdffontattr was introduced many years ago, but people still run
   % older pdftex's; it's easy to conditionalize, so we do.)
   \ifpdf \ifx\pdffontattr\thisisundefined \else
     \begingroup
       \catcode`\^^M=\active \def^^M{^^J}% Output line endings as the ^^J char.
       \catcode`\%=12 \immediate\pdfobj stream {%!PS-Adobe-3.0 Resource-CMap
   %%DocumentNeededResources: ProcSet (CIDInit)
   %%IncludeResource: ProcSet (CIDInit)
   %%BeginResource: CMap (TeX-OT1-0)
   %%Title: (TeX-OT1-0 TeX OT1 0)
   %%Version: 1.000
   %%EndComments
   /CIDInit /ProcSet findresource begin
   12 dict begin
   begincmap
   /CIDSystemInfo
   << /Registry (TeX)
   /Ordering (OT1)
   /Supplement 0
   >> def
   /CMapName /TeX-OT1-0 def
   /CMapType 2 def
   1 begincodespacerange
   <00> <7F>
   endcodespacerange
   8 beginbfrange
   <00> <01> <0393>
   <09> <0A> <03A8>
   <23> <26> <0023>
   <28> <3B> <0028>
   <3F> <5B> <003F>
   <5D> <5E> <005D>
   <61> <7A> <0061>
   <7B> <7C> <2013>
   endbfrange
   40 beginbfchar
   <02> <0398>
   <03> <039B>
   <04> <039E>
   <05> <03A0>
   <06> <03A3>
   <07> <03D2>
   <08> <03A6>
   <0B> <00660066>
   <0C> <00660069>
   <0D> <0066006C>
   <0E> <006600660069>
   <0F> <00660066006C>
   <10> <0131>
   <11> <0237>
   <12> <0060>
   <13> <00B4>
   <14> <02C7>
   <15> <02D8>
   <16> <00AF>
   <17> <02DA>
   <18> <00B8>
   <19> <00DF>
   <1A> <00E6>
   <1B> <0153>
   <1C> <00F8>
   <1D> <00C6>
   <1E> <0152>
   <1F> <00D8>
   <21> <0021>
   <22> <201D>
   <27> <2019>
   <3C> <00A1>
   <3D> <003D>
   <3E> <00BF>
   <5C> <201C>
   <5F> <02D9>
   <60> <2018>
   <7D> <02DD>
   <7E> <007E>
   <7F> <00A8>
   endbfchar
   endcmap
   CMapName currentdict /CMap defineresource pop
   end
   end
   %%EndResource
   %%EOF
       }\endgroup
     \expandafter\edef\csname cmapOT1\endcsname#1{%
       \pdffontattr#1{/ToUnicode \the\pdflastobj\space 0 R}%
     }%
   %
   % \cmapOT1IT
     \begingroup
       \catcode`\^^M=\active \def^^M{^^J}% Output line endings as the ^^J char.
       \catcode`\%=12 \immediate\pdfobj stream {%!PS-Adobe-3.0 Resource-CMap
   %%DocumentNeededResources: ProcSet (CIDInit)
   %%IncludeResource: ProcSet (CIDInit)
   %%BeginResource: CMap (TeX-OT1IT-0)
   %%Title: (TeX-OT1IT-0 TeX OT1IT 0)
   %%Version: 1.000
   %%EndComments
   /CIDInit /ProcSet findresource begin
   12 dict begin
   begincmap
   /CIDSystemInfo
   << /Registry (TeX)
   /Ordering (OT1IT)
   /Supplement 0
   >> def
   /CMapName /TeX-OT1IT-0 def
   /CMapType 2 def
   1 begincodespacerange
   <00> <7F>
   endcodespacerange
   8 beginbfrange
   <00> <01> <0393>
   <09> <0A> <03A8>
   <25> <26> <0025>
   <28> <3B> <0028>
   <3F> <5B> <003F>
   <5D> <5E> <005D>
   <61> <7A> <0061>
   <7B> <7C> <2013>
   endbfrange
   42 beginbfchar
   <02> <0398>
   <03> <039B>
   <04> <039E>
   <05> <03A0>
   <06> <03A3>
   <07> <03D2>
   <08> <03A6>
   <0B> <00660066>
   <0C> <00660069>
   <0D> <0066006C>
   <0E> <006600660069>
   <0F> <00660066006C>
   <10> <0131>
   <11> <0237>
   <12> <0060>
   <13> <00B4>
   <14> <02C7>
   <15> <02D8>
   <16> <00AF>
   <17> <02DA>
   <18> <00B8>
   <19> <00DF>
   <1A> <00E6>
   <1B> <0153>
   <1C> <00F8>
   <1D> <00C6>
   <1E> <0152>
   <1F> <00D8>
   <21> <0021>
   <22> <201D>
   <23> <0023>
   <24> <00A3>
   <27> <2019>
   <3C> <00A1>
   <3D> <003D>
   <3E> <00BF>
   <5C> <201C>
   <5F> <02D9>
   <60> <2018>
   <7D> <02DD>
   <7E> <007E>
   <7F> <00A8>
   endbfchar
   endcmap
   CMapName currentdict /CMap defineresource pop
   end
   end
   %%EndResource
   %%EOF
       }\endgroup
     \expandafter\edef\csname cmapOT1IT\endcsname#1{%
       \pdffontattr#1{/ToUnicode \the\pdflastobj\space 0 R}%
     }%
   %
   % \cmapOT1TT
     \begingroup
       \catcode`\^^M=\active \def^^M{^^J}% Output line endings as the ^^J char.
       \catcode`\%=12 \immediate\pdfobj stream {%!PS-Adobe-3.0 Resource-CMap
   %%DocumentNeededResources: ProcSet (CIDInit)
   %%IncludeResource: ProcSet (CIDInit)
   %%BeginResource: CMap (TeX-OT1TT-0)
   %%Title: (TeX-OT1TT-0 TeX OT1TT 0)
   %%Version: 1.000
   %%EndComments
   /CIDInit /ProcSet findresource begin
   12 dict begin
   begincmap
   /CIDSystemInfo
   << /Registry (TeX)
   /Ordering (OT1TT)
   /Supplement 0
   >> def
   /CMapName /TeX-OT1TT-0 def
   /CMapType 2 def
   1 begincodespacerange
   <00> <7F>
   endcodespacerange
   5 beginbfrange
   <00> <01> <0393>
   <09> <0A> <03A8>
   <21> <26> <0021>
   <28> <5F> <0028>
   <61> <7E> <0061>
   endbfrange
   32 beginbfchar
   <02> <0398>
   <03> <039B>
   <04> <039E>
   <05> <03A0>
   <06> <03A3>
   <07> <03D2>
   <08> <03A6>
   <0B> <2191>
   <0C> <2193>
   <0D> <0027>
   <0E> <00A1>
   <0F> <00BF>
   <10> <0131>
   <11> <0237>
   <12> <0060>
   <13> <00B4>
   <14> <02C7>
   <15> <02D8>
   <16> <00AF>
   <17> <02DA>
   <18> <00B8>
   <19> <00DF>
   <1A> <00E6>
   <1B> <0153>
   <1C> <00F8>
   <1D> <00C6>
   <1E> <0152>
   <1F> <00D8>
   <20> <2423>
   <27> <2019>
   <60> <2018>
   <7F> <00A8>
   endbfchar
   endcmap
   CMapName currentdict /CMap defineresource pop
   end
   end
   %%EndResource
   %%EOF
       }\endgroup
     \expandafter\edef\csname cmapOT1TT\endcsname#1{%
       \pdffontattr#1{/ToUnicode \the\pdflastobj\space 0 R}%
     }%
   \fi\fi
   
   
   % Set the font macro #1 to the font named \fontprefix#2.
   % #3 is the font's design size, #4 is a scale factor, #5 is the CMap
   % encoding (only OT1, OT1IT and OT1TT are allowed, or empty to omit).
   % Example:
   % #1 = \textrm
   % #2 = \rmshape
   % #3 = 10
   % #4 = \mainmagstep
   % #5 = OT1
   %
   \def\setfont#1#2#3#4#5{%
     \font#1=\fontprefix#2#3 scaled #4
     \csname cmap#5\endcsname#1%
   }
   % This is what gets called when #5 of \setfont is empty.
   \let\cmap\gobble
   %
   % (end of cmaps)
   
 % Use cm as the default font prefix.  % Use cm as the default font prefix.
 % To specify the font prefix, you must define \fontprefix  % To specify the font prefix, you must define \fontprefix
 % before you read in texinfo.tex.  % before you read in texinfo.tex.
 \ifx\fontprefix\undefined  \ifx\fontprefix\thisisundefined
 \def\fontprefix{cm}  \def\fontprefix{cm}
 \fi  \fi
 % Support font families that don't use the same naming scheme as CM.  % Support font families that don't use the same naming scheme as CM.
 \def\rmshape{r}  \def\rmshape{r}
 \def\rmbshape{bx}               %where the normal face is bold  \def\rmbshape{bx}               % where the normal face is bold
 \def\bfshape{b}  \def\bfshape{b}
 \def\bxshape{bx}  \def\bxshape{bx}
 \def\ttshape{tt}  \def\ttshape{tt}
Line 1064  where each line of input produces a line of output.}
Line 2246  where each line of input produces a line of output.}
 \def\scshape{csc}  \def\scshape{csc}
 \def\scbshape{csc}  \def\scbshape{csc}
   
 \ifx\bigger\relax  % Definitions for a main text size of 11pt.  (The default in Texinfo.)
 \let\mainmagstep=\magstep1  %
 \setfont\textrm\rmshape{12}{1000}  \def\definetextfontsizexi{%
 \setfont\texttt\ttshape{12}{1000}  % Text fonts (11.2pt, magstep1).
 \else  \def\textnominalsize{11pt}
 \setfont\textrm\rmshape{10}{\mainmagstep}  \edef\mainmagstep{\magstephalf}
 \setfont\texttt\ttshape{10}{\mainmagstep}  \setfont\textrm\rmshape{10}{\mainmagstep}{OT1}
 \fi  \setfont\texttt\ttshape{10}{\mainmagstep}{OT1TT}
 % Instead of cmb10, you many want to use cmbx10.  \setfont\textbf\bfshape{10}{\mainmagstep}{OT1}
 % cmbx10 is a prettier font on its own, but cmb10  \setfont\textit\itshape{10}{\mainmagstep}{OT1IT}
 % looks better when embedded in a line with cmr10.  \setfont\textsl\slshape{10}{\mainmagstep}{OT1}
 \setfont\textbf\bfshape{10}{\mainmagstep}  \setfont\textsf\sfshape{10}{\mainmagstep}{OT1}
 \setfont\textit\itshape{10}{\mainmagstep}  \setfont\textsc\scshape{10}{\mainmagstep}{OT1}
 \setfont\textsl\slshape{10}{\mainmagstep}  \setfont\textttsl\ttslshape{10}{\mainmagstep}{OT1TT}
 \setfont\textsf\sfshape{10}{\mainmagstep}  
 \setfont\textsc\scshape{10}{\mainmagstep}  
 \setfont\textttsl\ttslshape{10}{\mainmagstep}  
 \font\texti=cmmi10 scaled \mainmagstep  \font\texti=cmmi10 scaled \mainmagstep
 \font\textsy=cmsy10 scaled \mainmagstep  \font\textsy=cmsy10 scaled \mainmagstep
   \def\textecsize{1095}
   
 % A few fonts for @defun, etc.  % A few fonts for @defun names and args.
 \setfont\defbf\bxshape{10}{\magstep1} %was 1314  \setfont\defbf\bfshape{10}{\magstep1}{OT1}
 \setfont\deftt\ttshape{10}{\magstep1}  \setfont\deftt\ttshape{10}{\magstep1}{OT1TT}
 \def\df{\let\tentt=\deftt \let\tenbf = \defbf \bf}  \setfont\defsl\slshape{10}{\magstep1}{OT1TT}
   \setfont\defttsl\ttslshape{10}{\magstep1}{OT1TT}
   \def\df{\let\tentt=\deftt \let\tenbf = \defbf
   \let\tenttsl=\defttsl \let\tensl=\defsl \bf}
   
 % Fonts for indices, footnotes, small examples (9pt).  % Fonts for indices, footnotes, small examples (9pt).
 \setfont\smallrm\rmshape{9}{1000}  \def\smallnominalsize{9pt}
 \setfont\smalltt\ttshape{9}{1000}  \setfont\smallrm\rmshape{9}{1000}{OT1}
 \setfont\smallbf\bfshape{10}{900}  \setfont\smalltt\ttshape{9}{1000}{OT1TT}
 \setfont\smallit\itshape{9}{1000}  \setfont\smallbf\bfshape{10}{900}{OT1}
 \setfont\smallsl\slshape{9}{1000}  \setfont\smallit\itshape{9}{1000}{OT1IT}
 \setfont\smallsf\sfshape{9}{1000}  \setfont\smallsl\slshape{9}{1000}{OT1}
 \setfont\smallsc\scshape{10}{900}  \setfont\smallsf\sfshape{9}{1000}{OT1}
 \setfont\smallttsl\ttslshape{10}{900}  \setfont\smallsc\scshape{10}{900}{OT1}
   \setfont\smallttsl\ttslshape{10}{900}{OT1TT}
 \font\smalli=cmmi9  \font\smalli=cmmi9
 \font\smallsy=cmsy9  \font\smallsy=cmsy9
   \def\smallecsize{0900}
   
 % Fonts for title page:  % Fonts for small examples (8pt).
 \setfont\titlerm\rmbshape{12}{\magstep3}  \def\smallernominalsize{8pt}
 \setfont\titleit\itbshape{10}{\magstep4}  \setfont\smallerrm\rmshape{8}{1000}{OT1}
 \setfont\titlesl\slbshape{10}{\magstep4}  \setfont\smallertt\ttshape{8}{1000}{OT1TT}
 \setfont\titlett\ttbshape{12}{\magstep3}  \setfont\smallerbf\bfshape{10}{800}{OT1}
 \setfont\titlettsl\ttslshape{10}{\magstep4}  \setfont\smallerit\itshape{8}{1000}{OT1IT}
 \setfont\titlesf\sfbshape{17}{\magstep1}  \setfont\smallersl\slshape{8}{1000}{OT1}
   \setfont\smallersf\sfshape{8}{1000}{OT1}
   \setfont\smallersc\scshape{10}{800}{OT1}
   \setfont\smallerttsl\ttslshape{10}{800}{OT1TT}
   \font\smalleri=cmmi8
   \font\smallersy=cmsy8
   \def\smallerecsize{0800}
   
   % Fonts for title page (20.4pt):
   \def\titlenominalsize{20pt}
   \setfont\titlerm\rmbshape{12}{\magstep3}{OT1}
   \setfont\titleit\itbshape{10}{\magstep4}{OT1IT}
   \setfont\titlesl\slbshape{10}{\magstep4}{OT1}
   \setfont\titlett\ttbshape{12}{\magstep3}{OT1TT}
   \setfont\titlettsl\ttslshape{10}{\magstep4}{OT1TT}
   \setfont\titlesf\sfbshape{17}{\magstep1}{OT1}
 \let\titlebf=\titlerm  \let\titlebf=\titlerm
 \setfont\titlesc\scbshape{10}{\magstep4}  \setfont\titlesc\scbshape{10}{\magstep4}{OT1}
 \font\titlei=cmmi12 scaled \magstep3  \font\titlei=cmmi12 scaled \magstep3
 \font\titlesy=cmsy10 scaled \magstep4  \font\titlesy=cmsy10 scaled \magstep4
 \def\authorrm{\secrm}  \def\titleecsize{2074}
   
 % Chapter (and unnumbered) fonts (17.28pt).  % Chapter (and unnumbered) fonts (17.28pt).
 \setfont\chaprm\rmbshape{12}{\magstep2}  \def\chapnominalsize{17pt}
 \setfont\chapit\itbshape{10}{\magstep3}  \setfont\chaprm\rmbshape{12}{\magstep2}{OT1}
 \setfont\chapsl\slbshape{10}{\magstep3}  \setfont\chapit\itbshape{10}{\magstep3}{OT1IT}
 \setfont\chaptt\ttbshape{12}{\magstep2}  \setfont\chapsl\slbshape{10}{\magstep3}{OT1}
 \setfont\chapttsl\ttslshape{10}{\magstep3}  \setfont\chaptt\ttbshape{12}{\magstep2}{OT1TT}
 \setfont\chapsf\sfbshape{17}{1000}  \setfont\chapttsl\ttslshape{10}{\magstep3}{OT1TT}
   \setfont\chapsf\sfbshape{17}{1000}{OT1}
 \let\chapbf=\chaprm  \let\chapbf=\chaprm
 \setfont\chapsc\scbshape{10}{\magstep3}  \setfont\chapsc\scbshape{10}{\magstep3}{OT1}
 \font\chapi=cmmi12 scaled \magstep2  \font\chapi=cmmi12 scaled \magstep2
 \font\chapsy=cmsy10 scaled \magstep3  \font\chapsy=cmsy10 scaled \magstep3
   \def\chapecsize{1728}
   
 % Section fonts (14.4pt).  % Section fonts (14.4pt).
 \setfont\secrm\rmbshape{12}{\magstep1}  \def\secnominalsize{14pt}
 \setfont\secit\itbshape{10}{\magstep2}  \setfont\secrm\rmbshape{12}{\magstep1}{OT1}
 \setfont\secsl\slbshape{10}{\magstep2}  \setfont\secrmnotbold\rmshape{12}{\magstep1}{OT1}
 \setfont\sectt\ttbshape{12}{\magstep1}  \setfont\secit\itbshape{10}{\magstep2}{OT1IT}
 \setfont\secttsl\ttslshape{10}{\magstep2}  \setfont\secsl\slbshape{10}{\magstep2}{OT1}
 \setfont\secsf\sfbshape{12}{\magstep1}  \setfont\sectt\ttbshape{12}{\magstep1}{OT1TT}
   \setfont\secttsl\ttslshape{10}{\magstep2}{OT1TT}
   \setfont\secsf\sfbshape{12}{\magstep1}{OT1}
 \let\secbf\secrm  \let\secbf\secrm
 \setfont\secsc\scbshape{10}{\magstep2}  \setfont\secsc\scbshape{10}{\magstep2}{OT1}
 \font\seci=cmmi12 scaled \magstep1  \font\seci=cmmi12 scaled \magstep1
 \font\secsy=cmsy10 scaled \magstep2  \font\secsy=cmsy10 scaled \magstep2
   \def\sececsize{1440}
   
 % \setfont\ssecrm\bxshape{10}{\magstep1}    % This size an font looked bad.  
 % \setfont\ssecit\itshape{10}{\magstep1}    % The letters were too crowded.  
 % \setfont\ssecsl\slshape{10}{\magstep1}  
 % \setfont\ssectt\ttshape{10}{\magstep1}  
 % \setfont\ssecsf\sfshape{10}{\magstep1}  
   
 %\setfont\ssecrm\bfshape{10}{1315}      % Note the use of cmb rather than cmbx.  
 %\setfont\ssecit\itshape{10}{1315}      % Also, the size is a little larger than  
 %\setfont\ssecsl\slshape{10}{1315}      % being scaled magstep1.  
 %\setfont\ssectt\ttshape{10}{1315}  
 %\setfont\ssecsf\sfshape{10}{1315}  
   
 %\let\ssecbf=\ssecrm  
   
 % Subsection fonts (13.15pt).  % Subsection fonts (13.15pt).
 \setfont\ssecrm\rmbshape{12}{\magstephalf}  \def\ssecnominalsize{13pt}
 \setfont\ssecit\itbshape{10}{1315}  \setfont\ssecrm\rmbshape{12}{\magstephalf}{OT1}
 \setfont\ssecsl\slbshape{10}{1315}  \setfont\ssecit\itbshape{10}{1315}{OT1IT}
 \setfont\ssectt\ttbshape{12}{\magstephalf}  \setfont\ssecsl\slbshape{10}{1315}{OT1}
 \setfont\ssecttsl\ttslshape{10}{1315}  \setfont\ssectt\ttbshape{12}{\magstephalf}{OT1TT}
 \setfont\ssecsf\sfbshape{12}{\magstephalf}  \setfont\ssecttsl\ttslshape{10}{1315}{OT1TT}
   \setfont\ssecsf\sfbshape{12}{\magstephalf}{OT1}
 \let\ssecbf\ssecrm  \let\ssecbf\ssecrm
 \setfont\ssecsc\scbshape{10}{\magstep1}  \setfont\ssecsc\scbshape{10}{1315}{OT1}
 \font\sseci=cmmi12 scaled \magstephalf  \font\sseci=cmmi12 scaled \magstephalf
 \font\ssecsy=cmsy10 scaled 1315  \font\ssecsy=cmsy10 scaled 1315
 % The smallcaps and symbol fonts should actually be scaled \magstep1.5,  \def\ssececsize{1200}
 % but that is not a standard magnification.  
   
   % Reduced fonts for @acro in text (10pt).
   \def\reducednominalsize{10pt}
   \setfont\reducedrm\rmshape{10}{1000}{OT1}
   \setfont\reducedtt\ttshape{10}{1000}{OT1TT}
   \setfont\reducedbf\bfshape{10}{1000}{OT1}
   \setfont\reducedit\itshape{10}{1000}{OT1IT}
   \setfont\reducedsl\slshape{10}{1000}{OT1}
   \setfont\reducedsf\sfshape{10}{1000}{OT1}
   \setfont\reducedsc\scshape{10}{1000}{OT1}
   \setfont\reducedttsl\ttslshape{10}{1000}{OT1TT}
   \font\reducedi=cmmi10
   \font\reducedsy=cmsy10
   \def\reducedecsize{1000}
   
   \textleading = 13.2pt % line spacing for 11pt CM
   \textfonts            % reset the current fonts
   \rm
   } % end of 11pt text font size definitions, \definetextfontsizexi
   
   
   % Definitions to make the main text be 10pt Computer Modern, with
   % section, chapter, etc., sizes following suit.  This is for the GNU
   % Press printing of the Emacs 22 manual.  Maybe other manuals in the
   % future.  Used with @smallbook, which sets the leading to 12pt.
   %
   \def\definetextfontsizex{%
   % Text fonts (10pt).
   \def\textnominalsize{10pt}
   \edef\mainmagstep{1000}
   \setfont\textrm\rmshape{10}{\mainmagstep}{OT1}
   \setfont\texttt\ttshape{10}{\mainmagstep}{OT1TT}
   \setfont\textbf\bfshape{10}{\mainmagstep}{OT1}
   \setfont\textit\itshape{10}{\mainmagstep}{OT1IT}
   \setfont\textsl\slshape{10}{\mainmagstep}{OT1}
   \setfont\textsf\sfshape{10}{\mainmagstep}{OT1}
   \setfont\textsc\scshape{10}{\mainmagstep}{OT1}
   \setfont\textttsl\ttslshape{10}{\mainmagstep}{OT1TT}
   \font\texti=cmmi10 scaled \mainmagstep
   \font\textsy=cmsy10 scaled \mainmagstep
   \def\textecsize{1000}
   
   % A few fonts for @defun names and args.
   \setfont\defbf\bfshape{10}{\magstephalf}{OT1}
   \setfont\deftt\ttshape{10}{\magstephalf}{OT1TT}
   \setfont\defsl\slshape{10}{\magstephalf}{OT1TT}
   \setfont\defttsl\ttslshape{10}{\magstephalf}{OT1TT}
   \def\df{\let\tentt=\deftt \let\tenbf = \defbf
   \let\tensl=\defsl \let\tenttsl=\defttsl \bf}
   
   % Fonts for indices, footnotes, small examples (9pt).
   \def\smallnominalsize{9pt}
   \setfont\smallrm\rmshape{9}{1000}{OT1}
   \setfont\smalltt\ttshape{9}{1000}{OT1TT}
   \setfont\smallbf\bfshape{10}{900}{OT1}
   \setfont\smallit\itshape{9}{1000}{OT1IT}
   \setfont\smallsl\slshape{9}{1000}{OT1}
   \setfont\smallsf\sfshape{9}{1000}{OT1}
   \setfont\smallsc\scshape{10}{900}{OT1}
   \setfont\smallttsl\ttslshape{10}{900}{OT1TT}
   \font\smalli=cmmi9
   \font\smallsy=cmsy9
   \def\smallecsize{0900}
   
   % Fonts for small examples (8pt).
   \def\smallernominalsize{8pt}
   \setfont\smallerrm\rmshape{8}{1000}{OT1}
   \setfont\smallertt\ttshape{8}{1000}{OT1TT}
   \setfont\smallerbf\bfshape{10}{800}{OT1}
   \setfont\smallerit\itshape{8}{1000}{OT1IT}
   \setfont\smallersl\slshape{8}{1000}{OT1}
   \setfont\smallersf\sfshape{8}{1000}{OT1}
   \setfont\smallersc\scshape{10}{800}{OT1}
   \setfont\smallerttsl\ttslshape{10}{800}{OT1TT}
   \font\smalleri=cmmi8
   \font\smallersy=cmsy8
   \def\smallerecsize{0800}
   
   % Fonts for title page (20.4pt):
   \def\titlenominalsize{20pt}
   \setfont\titlerm\rmbshape{12}{\magstep3}{OT1}
   \setfont\titleit\itbshape{10}{\magstep4}{OT1IT}
   \setfont\titlesl\slbshape{10}{\magstep4}{OT1}
   \setfont\titlett\ttbshape{12}{\magstep3}{OT1TT}
   \setfont\titlettsl\ttslshape{10}{\magstep4}{OT1TT}
   \setfont\titlesf\sfbshape{17}{\magstep1}{OT1}
   \let\titlebf=\titlerm
   \setfont\titlesc\scbshape{10}{\magstep4}{OT1}
   \font\titlei=cmmi12 scaled \magstep3
   \font\titlesy=cmsy10 scaled \magstep4
   \def\titleecsize{2074}
   
   % Chapter fonts (14.4pt).
   \def\chapnominalsize{14pt}
   \setfont\chaprm\rmbshape{12}{\magstep1}{OT1}
   \setfont\chapit\itbshape{10}{\magstep2}{OT1IT}
   \setfont\chapsl\slbshape{10}{\magstep2}{OT1}
   \setfont\chaptt\ttbshape{12}{\magstep1}{OT1TT}
   \setfont\chapttsl\ttslshape{10}{\magstep2}{OT1TT}
   \setfont\chapsf\sfbshape{12}{\magstep1}{OT1}
   \let\chapbf\chaprm
   \setfont\chapsc\scbshape{10}{\magstep2}{OT1}
   \font\chapi=cmmi12 scaled \magstep1
   \font\chapsy=cmsy10 scaled \magstep2
   \def\chapecsize{1440}
   
   % Section fonts (12pt).
   \def\secnominalsize{12pt}
   \setfont\secrm\rmbshape{12}{1000}{OT1}
   \setfont\secit\itbshape{10}{\magstep1}{OT1IT}
   \setfont\secsl\slbshape{10}{\magstep1}{OT1}
   \setfont\sectt\ttbshape{12}{1000}{OT1TT}
   \setfont\secttsl\ttslshape{10}{\magstep1}{OT1TT}
   \setfont\secsf\sfbshape{12}{1000}{OT1}
   \let\secbf\secrm
   \setfont\secsc\scbshape{10}{\magstep1}{OT1}
   \font\seci=cmmi12
   \font\secsy=cmsy10 scaled \magstep1
   \def\sececsize{1200}
   
   % Subsection fonts (10pt).
   \def\ssecnominalsize{10pt}
   \setfont\ssecrm\rmbshape{10}{1000}{OT1}
   \setfont\ssecit\itbshape{10}{1000}{OT1IT}
   \setfont\ssecsl\slbshape{10}{1000}{OT1}
   \setfont\ssectt\ttbshape{10}{1000}{OT1TT}
   \setfont\ssecttsl\ttslshape{10}{1000}{OT1TT}
   \setfont\ssecsf\sfbshape{10}{1000}{OT1}
   \let\ssecbf\ssecrm
   \setfont\ssecsc\scbshape{10}{1000}{OT1}
   \font\sseci=cmmi10
   \font\ssecsy=cmsy10
   \def\ssececsize{1000}
   
   % Reduced fonts for @acro in text (9pt).
   \def\reducednominalsize{9pt}
   \setfont\reducedrm\rmshape{9}{1000}{OT1}
   \setfont\reducedtt\ttshape{9}{1000}{OT1TT}
   \setfont\reducedbf\bfshape{10}{900}{OT1}
   \setfont\reducedit\itshape{9}{1000}{OT1IT}
   \setfont\reducedsl\slshape{9}{1000}{OT1}
   \setfont\reducedsf\sfshape{9}{1000}{OT1}
   \setfont\reducedsc\scshape{10}{900}{OT1}
   \setfont\reducedttsl\ttslshape{10}{900}{OT1TT}
   \font\reducedi=cmmi9
   \font\reducedsy=cmsy9
   \def\reducedecsize{0900}
   
   \divide\parskip by 2  % reduce space between paragraphs
   \textleading = 12pt   % line spacing for 10pt CM
   \textfonts            % reset the current fonts
   \rm
   } % end of 10pt text font size definitions, \definetextfontsizex
   
   
   % We provide the user-level command
   %   @fonttextsize 10
   % (or 11) to redefine the text font size.  pt is assumed.
   %
   \def\xiword{11}
   \def\xword{10}
   \def\xwordpt{10pt}
   %
   \parseargdef\fonttextsize{%
     \def\textsizearg{#1}%
     %\wlog{doing @fonttextsize \textsizearg}%
     %
     % Set \globaldefs so that documents can use this inside @tex, since
     % makeinfo 4.8 does not support it, but we need it nonetheless.
     %
    \begingroup \globaldefs=1
     \ifx\textsizearg\xword \definetextfontsizex
     \else \ifx\textsizearg\xiword \definetextfontsizexi
     \else
       \errhelp=\EMsimple
       \errmessage{@fonttextsize only supports `10' or `11', not `\textsizearg'}
     \fi\fi
    \endgroup
   }
   
 % In order for the font changes to affect most math symbols and letters,  % In order for the font changes to affect most math symbols and letters,
 % we have to define the \textfont of the standard families.  Since  % we have to define the \textfont of the standard families.  We don't
 % texinfo doesn't allow for producing subscripts and superscripts, we  % bother to reset \scriptfont and \scriptscriptfont; awaiting user need.
 % don't bother to reset \scriptfont and \scriptscriptfont (which would  
 % also require loading a lot more fonts).  
 %  %
 \def\resetmathfonts{%  \def\resetmathfonts{%
   \textfont0 = \tenrm \textfont1 = \teni \textfont2 = \tensy    \textfont0=\tenrm \textfont1=\teni \textfont2=\tensy
   \textfont\itfam = \tenit \textfont\slfam = \tensl \textfont\bffam = \tenbf    \textfont\itfam=\tenit \textfont\slfam=\tensl \textfont\bffam=\tenbf
   \textfont\ttfam = \tentt \textfont\sffam = \tensf    \textfont\ttfam=\tentt \textfont\sffam=\tensf
 }  }
   
   
 % The font-changing commands redefine the meanings of \tenSTYLE, instead  % The font-changing commands redefine the meanings of \tenSTYLE, instead
 % of just \STYLE.  We do this so that font changes will continue to work  % of just \STYLE.  We do this because \STYLE needs to also set the
 % in math mode, where it is the current \fam that is relevant in most  % current \fam for math mode.  Our \STYLE (e.g., \rm) commands hardwire
 % cases, not the current font.  Plain TeX does \def\bf{\fam=\bffam  % \tenSTYLE to set the current font.
 % \tenbf}, for example.  By redefining \tenbf, we obviate the need to  %
 % redefine \bf itself.  % Each font-changing command also sets the names \lsize (one size lower)
   % and \lllsize (three sizes lower).  These relative commands are used
   % in, e.g., the LaTeX logo and acronyms.
   %
   % This all needs generalizing, badly.
   %
 \def\textfonts{%  \def\textfonts{%
   \let\tenrm=\textrm \let\tenit=\textit \let\tensl=\textsl    \let\tenrm=\textrm \let\tenit=\textit \let\tensl=\textsl
   \let\tenbf=\textbf \let\tentt=\texttt \let\smallcaps=\textsc    \let\tenbf=\textbf \let\tentt=\texttt \let\smallcaps=\textsc
   \let\tensf=\textsf \let\teni=\texti \let\tensy=\textsy \let\tenttsl=\textttsl    \let\tensf=\textsf \let\teni=\texti \let\tensy=\textsy
   \resetmathfonts}    \let\tenttsl=\textttsl
     \def\curfontsize{text}%
     \def\lsize{reduced}\def\lllsize{smaller}%
     \resetmathfonts \setleading{\textleading}}
 \def\titlefonts{%  \def\titlefonts{%
   \let\tenrm=\titlerm \let\tenit=\titleit \let\tensl=\titlesl    \let\tenrm=\titlerm \let\tenit=\titleit \let\tensl=\titlesl
   \let\tenbf=\titlebf \let\tentt=\titlett \let\smallcaps=\titlesc    \let\tenbf=\titlebf \let\tentt=\titlett \let\smallcaps=\titlesc
   \let\tensf=\titlesf \let\teni=\titlei \let\tensy=\titlesy    \let\tensf=\titlesf \let\teni=\titlei \let\tensy=\titlesy
   \let\tenttsl=\titlettsl    \let\tenttsl=\titlettsl
   \resetmathfonts \setleading{25pt}}    \def\curfontsize{title}%
 \def\titlefont#1{{\titlefonts\rm #1}}    \def\lsize{chap}\def\lllsize{subsec}%
     \resetmathfonts \setleading{27pt}}
   \def\titlefont#1{{\titlefonts\rmisbold #1}}
 \def\chapfonts{%  \def\chapfonts{%
   \let\tenrm=\chaprm \let\tenit=\chapit \let\tensl=\chapsl    \let\tenrm=\chaprm \let\tenit=\chapit \let\tensl=\chapsl
   \let\tenbf=\chapbf \let\tentt=\chaptt \let\smallcaps=\chapsc    \let\tenbf=\chapbf \let\tentt=\chaptt \let\smallcaps=\chapsc
   \let\tensf=\chapsf \let\teni=\chapi \let\tensy=\chapsy \let\tenttsl=\chapttsl    \let\tensf=\chapsf \let\teni=\chapi \let\tensy=\chapsy
     \let\tenttsl=\chapttsl
     \def\curfontsize{chap}%
     \def\lsize{sec}\def\lllsize{text}%
   \resetmathfonts \setleading{19pt}}    \resetmathfonts \setleading{19pt}}
 \def\secfonts{%  \def\secfonts{%
   \let\tenrm=\secrm \let\tenit=\secit \let\tensl=\secsl    \let\tenrm=\secrm \let\tenit=\secit \let\tensl=\secsl
   \let\tenbf=\secbf \let\tentt=\sectt \let\smallcaps=\secsc    \let\tenbf=\secbf \let\tentt=\sectt \let\smallcaps=\secsc
   \let\tensf=\secsf \let\teni=\seci \let\tensy=\secsy \let\tenttsl=\secttsl    \let\tensf=\secsf \let\teni=\seci \let\tensy=\secsy
   \resetmathfonts \setleading{16pt}}    \let\tenttsl=\secttsl
     \def\curfontsize{sec}%
     \def\lsize{subsec}\def\lllsize{reduced}%
     \resetmathfonts \setleading{17pt}}
 \def\subsecfonts{%  \def\subsecfonts{%
   \let\tenrm=\ssecrm \let\tenit=\ssecit \let\tensl=\ssecsl    \let\tenrm=\ssecrm \let\tenit=\ssecit \let\tensl=\ssecsl
   \let\tenbf=\ssecbf \let\tentt=\ssectt \let\smallcaps=\ssecsc    \let\tenbf=\ssecbf \let\tentt=\ssectt \let\smallcaps=\ssecsc
   \let\tensf=\ssecsf \let\teni=\sseci \let\tensy=\ssecsy \let\tenttsl=\ssecttsl    \let\tensf=\ssecsf \let\teni=\sseci \let\tensy=\ssecsy
     \let\tenttsl=\ssecttsl
     \def\curfontsize{ssec}%
     \def\lsize{text}\def\lllsize{small}%
   \resetmathfonts \setleading{15pt}}    \resetmathfonts \setleading{15pt}}
 \let\subsubsecfonts = \subsecfonts % Maybe make sssec fonts scaled magstephalf?  \let\subsubsecfonts = \subsecfonts
   \def\reducedfonts{%
     \let\tenrm=\reducedrm \let\tenit=\reducedit \let\tensl=\reducedsl
     \let\tenbf=\reducedbf \let\tentt=\reducedtt \let\reducedcaps=\reducedsc
     \let\tensf=\reducedsf \let\teni=\reducedi \let\tensy=\reducedsy
     \let\tenttsl=\reducedttsl
     \def\curfontsize{reduced}%
     \def\lsize{small}\def\lllsize{smaller}%
     \resetmathfonts \setleading{10.5pt}}
 \def\smallfonts{%  \def\smallfonts{%
   \let\tenrm=\smallrm \let\tenit=\smallit \let\tensl=\smallsl    \let\tenrm=\smallrm \let\tenit=\smallit \let\tensl=\smallsl
   \let\tenbf=\smallbf \let\tentt=\smalltt \let\smallcaps=\smallsc    \let\tenbf=\smallbf \let\tentt=\smalltt \let\smallcaps=\smallsc
   \let\tensf=\smallsf \let\teni=\smalli \let\tensy=\smallsy    \let\tensf=\smallsf \let\teni=\smalli \let\tensy=\smallsy
   \let\tenttsl=\smallttsl    \let\tenttsl=\smallttsl
   \resetmathfonts \setleading{11pt}}    \def\curfontsize{small}%
     \def\lsize{smaller}\def\lllsize{smaller}%
     \resetmathfonts \setleading{10.5pt}}
   \def\smallerfonts{%
     \let\tenrm=\smallerrm \let\tenit=\smallerit \let\tensl=\smallersl
     \let\tenbf=\smallerbf \let\tentt=\smallertt \let\smallcaps=\smallersc
     \let\tensf=\smallersf \let\teni=\smalleri \let\tensy=\smallersy
     \let\tenttsl=\smallerttsl
     \def\curfontsize{smaller}%
     \def\lsize{smaller}\def\lllsize{smaller}%
     \resetmathfonts \setleading{9.5pt}}
   
 % Set up the default fonts, so we can use them for creating boxes.  % Fonts for short table of contents.
 %  \setfont\shortcontrm\rmshape{12}{1000}{OT1}
 \textfonts  \setfont\shortcontbf\bfshape{10}{\magstep1}{OT1}  % no cmb12
   \setfont\shortcontsl\slshape{12}{1000}{OT1}
   \setfont\shortconttt\ttshape{12}{1000}{OT1TT}
   
 % Define these so they can be easily changed for other fonts.  % Define these just so they can be easily changed for other fonts.
 \def\angleleft{$\langle$}  \def\angleleft{$\langle$}
 \def\angleright{$\rangle$}  \def\angleright{$\rangle$}
   
   % Set the fonts to use with the @small... environments.
   \let\smallexamplefonts = \smallfonts
   
   % About \smallexamplefonts.  If we use \smallfonts (9pt), @smallexample
   % can fit this many characters:
   %   8.5x11=86   smallbook=72  a4=90  a5=69
   % If we use \scriptfonts (8pt), then we can fit this many characters:
   %   8.5x11=90+  smallbook=80  a4=90+  a5=77
   % For me, subjectively, the few extra characters that fit aren't worth
   % the additional smallness of 8pt.  So I'm making the default 9pt.
   %
   % By the way, for comparison, here's what fits with @example (10pt):
   %   8.5x11=71  smallbook=60  a4=75  a5=58
   % --karl, 24jan03.
   
   % Set up the default fonts, so we can use them for creating boxes.
   %
   \definetextfontsizexi
   
   
   \message{markup,}
   
   % Check if we are currently using a typewriter font.  Since all the
   % Computer Modern typewriter fonts have zero interword stretch (and
   % shrink), and it is reasonable to expect all typewriter fonts to have
   % this property, we can check that font parameter.
   %
   \def\ifmonospace{\ifdim\fontdimen3\font=0pt }
   
   % Markup style infrastructure.  \defmarkupstylesetup\INITMACRO will
   % define and register \INITMACRO to be called on markup style changes.
   % \INITMACRO can check \currentmarkupstyle for the innermost
   % style and the set of \ifmarkupSTYLE switches for all styles
   % currently in effect.
   \newif\ifmarkupvar
   \newif\ifmarkupsamp
   \newif\ifmarkupkey
   %\newif\ifmarkupfile % @file == @samp.
   %\newif\ifmarkupoption % @option == @samp.
   \newif\ifmarkupcode
   \newif\ifmarkupkbd
   %\newif\ifmarkupenv % @env == @code.
   %\newif\ifmarkupcommand % @command == @code.
   \newif\ifmarkuptex % @tex (and part of @math, for now).
   \newif\ifmarkupexample
   \newif\ifmarkupverb
   \newif\ifmarkupverbatim
   
   \let\currentmarkupstyle\empty
   
   \def\setupmarkupstyle#1{%
     \csname markup#1true\endcsname
     \def\currentmarkupstyle{#1}%
     \markupstylesetup
   }
   
   \let\markupstylesetup\empty
   
   \def\defmarkupstylesetup#1{%
     \expandafter\def\expandafter\markupstylesetup
       \expandafter{\markupstylesetup #1}%
     \def#1%
   }
   
   % Markup style setup for left and right quotes.
   \defmarkupstylesetup\markupsetuplq{%
     \expandafter\let\expandafter \temp
       \csname markupsetuplq\currentmarkupstyle\endcsname
     \ifx\temp\relax \markupsetuplqdefault \else \temp \fi
   }
   
   \defmarkupstylesetup\markupsetuprq{%
     \expandafter\let\expandafter \temp
       \csname markupsetuprq\currentmarkupstyle\endcsname
     \ifx\temp\relax \markupsetuprqdefault \else \temp \fi
   }
   
   {
   \catcode`\'=\active
   \catcode`\`=\active
   
   \gdef\markupsetuplqdefault{\let`\lq}
   \gdef\markupsetuprqdefault{\let'\rq}
   
   \gdef\markupsetcodequoteleft{\let`\codequoteleft}
   \gdef\markupsetcodequoteright{\let'\codequoteright}
   }
   
   \let\markupsetuplqcode \markupsetcodequoteleft
   \let\markupsetuprqcode \markupsetcodequoteright
   %
   \let\markupsetuplqexample \markupsetcodequoteleft
   \let\markupsetuprqexample \markupsetcodequoteright
   %
   \let\markupsetuplqkbd     \markupsetcodequoteleft
   \let\markupsetuprqkbd     \markupsetcodequoteright
   %
   \let\markupsetuplqsamp \markupsetcodequoteleft
   \let\markupsetuprqsamp \markupsetcodequoteright
   %
   \let\markupsetuplqverb \markupsetcodequoteleft
   \let\markupsetuprqverb \markupsetcodequoteright
   %
   \let\markupsetuplqverbatim \markupsetcodequoteleft
   \let\markupsetuprqverbatim \markupsetcodequoteright
   
   % Allow an option to not use regular directed right quote/apostrophe
   % (char 0x27), but instead the undirected quote from cmtt (char 0x0d).
   % The undirected quote is ugly, so don't make it the default, but it
   % works for pasting with more pdf viewers (at least evince), the
   % lilypond developers report.  xpdf does work with the regular 0x27.
   %
   \def\codequoteright{%
     \expandafter\ifx\csname SETtxicodequoteundirected\endcsname\relax
       \expandafter\ifx\csname SETcodequoteundirected\endcsname\relax
         '%
       \else \char'15 \fi
     \else \char'15 \fi
   }
   %
   % and a similar option for the left quote char vs. a grave accent.
   % Modern fonts display ASCII 0x60 as a grave accent, so some people like
   % the code environments to do likewise.
   %
   \def\codequoteleft{%
     \expandafter\ifx\csname SETtxicodequotebacktick\endcsname\relax
       \expandafter\ifx\csname SETcodequotebacktick\endcsname\relax
         % [Knuth] pp. 380,381,391
         % \relax disables Spanish ligatures ?` and !` of \tt font.
         \relax`%
       \else \char'22 \fi
     \else \char'22 \fi
   }
   
   % Commands to set the quote options.
   %
   \parseargdef\codequoteundirected{%
     \def\temp{#1}%
     \ifx\temp\onword
       \expandafter\let\csname SETtxicodequoteundirected\endcsname
         = t%
     \else\ifx\temp\offword
       \expandafter\let\csname SETtxicodequoteundirected\endcsname
         = \relax
     \else
       \errhelp = \EMsimple
       \errmessage{Unknown @codequoteundirected value `\temp', must be on|off}%
     \fi\fi
   }
   %
   \parseargdef\codequotebacktick{%
     \def\temp{#1}%
     \ifx\temp\onword
       \expandafter\let\csname SETtxicodequotebacktick\endcsname
         = t%
     \else\ifx\temp\offword
       \expandafter\let\csname SETtxicodequotebacktick\endcsname
         = \relax
     \else
       \errhelp = \EMsimple
       \errmessage{Unknown @codequotebacktick value `\temp', must be on|off}%
     \fi\fi
   }
   
   % [Knuth] pp. 380,381,391, disable Spanish ligatures ?` and !` of \tt font.
   \def\noligaturesquoteleft{\relax\lq}
   
 % Count depth in font-changes, for error checks  % Count depth in font-changes, for error checks
 \newcount\fontdepth \fontdepth=0  \newcount\fontdepth \fontdepth=0
   
 % Fonts for short table of contents.  % Font commands.
 \setfont\shortcontrm\rmshape{12}{1000}  
 \setfont\shortcontbf\bxshape{12}{1000}  
 \setfont\shortcontsl\slshape{12}{1000}  
   
 %% Add scribe-like font environments, plus @l for inline lisp (usually sans  % #1 is the font command (\sl or \it), #2 is the text to slant.
 %% serif) and @ii for TeX italic  % If we are in a monospaced environment, however, 1) always use \ttsl,
   % and 2) do not add an italic correction.
   \def\dosmartslant#1#2{%
     \ifusingtt
       {{\ttsl #2}\let\next=\relax}%
       {\def\next{{#1#2}\futurelet\next\smartitaliccorrection}}%
     \next
   }
   \def\smartslanted{\dosmartslant\sl}
   \def\smartitalic{\dosmartslant\it}
   
 % \smartitalic{ARG} outputs arg in italics, followed by an italic correction  % Output an italic correction unless \next (presumed to be the following
 % unless the following character is such as not to need one.  % character) is such as not to need one.
 \def\smartitalicx{\ifx\next,\else\ifx\next-\else\ifx\next.\else\/\fi\fi\fi}  \def\smartitaliccorrection{%
 \def\smartslanted#1{{\sl #1}\futurelet\next\smartitalicx}    \ifx\next,%
 \def\smartitalic#1{{\it #1}\futurelet\next\smartitalicx}    \else\ifx\next-%
     \else\ifx\next.%
     \else\ifx\next\.%
     \else\ifx\next\comma%
     \else\ptexslash
     \fi\fi\fi\fi\fi
     \aftersmartic
   }
   
   % Unconditional use \ttsl, and no ic.  @var is set to this for defuns.
   \def\ttslanted#1{{\ttsl #1}}
   
   % @cite is like \smartslanted except unconditionally use \sl.  We never want
   % ttsl for book titles, do we?
   \def\cite#1{{\sl #1}\futurelet\next\smartitaliccorrection}
   
   \def\aftersmartic{}
   \def\var#1{%
     \let\saveaftersmartic = \aftersmartic
     \def\aftersmartic{\null\let\aftersmartic=\saveaftersmartic}%
     \smartslanted{#1}%
   }
   
 \let\i=\smartitalic  \let\i=\smartitalic
 \let\var=\smartslanted  \let\slanted=\smartslanted
 \let\dfn=\smartslanted  \let\dfn=\smartslanted
 \let\emph=\smartitalic  \let\emph=\smartitalic
 \let\cite=\smartslanted  
   
   % Explicit font changes: @r, @sc, undocumented @ii.
   \def\r#1{{\rm #1}}              % roman font
   \def\sc#1{{\smallcaps#1}}       % smallcaps font
   \def\ii#1{{\it #1}}             % italic font
   
   % @b, explicit bold.  Also @strong.
 \def\b#1{{\bf #1}}  \def\b#1{{\bf #1}}
 \let\strong=\b  \let\strong=\b
   
   % @sansserif, explicit sans.
   \def\sansserif#1{{\sf #1}}
   
 % We can't just use \exhyphenpenalty, because that only has effect at  % We can't just use \exhyphenpenalty, because that only has effect at
 % the end of a paragraph.  Restore normal hyphenation at the end of the  % the end of a paragraph.  Restore normal hyphenation at the end of the
 % group within which \nohyphenation is presumably called.  % group within which \nohyphenation is presumably called.
Line 1261  where each line of input produces a line of output.}
Line 2869  where each line of input produces a line of output.}
 \def\nohyphenation{\hyphenchar\font = -1  \aftergroup\restorehyphenation}  \def\nohyphenation{\hyphenchar\font = -1  \aftergroup\restorehyphenation}
 \def\restorehyphenation{\hyphenchar\font = `- }  \def\restorehyphenation{\hyphenchar\font = `- }
   
   % Set sfcode to normal for the chars that usually have another value.
   % Can't use plain's \frenchspacing because it uses the `\x notation, and
   % sometimes \x has an active definition that messes things up.
   %
   \catcode`@=11
     \def\plainfrenchspacing{%
       \sfcode`\.=\@m \sfcode`\?=\@m \sfcode`\!=\@m
       \sfcode`\:=\@m \sfcode`\;=\@m \sfcode`\,=\@m
       \def\endofsentencespacefactor{1000}% for @. and friends
     }
     \def\plainnonfrenchspacing{%
       \sfcode`\.3000\sfcode`\?3000\sfcode`\!3000
       \sfcode`\:2000\sfcode`\;1500\sfcode`\,1250
       \def\endofsentencespacefactor{3000}% for @. and friends
     }
   \catcode`@=\other
   \def\endofsentencespacefactor{3000}% default
   
   % @t, explicit typewriter.
 \def\t#1{%  \def\t#1{%
   {\tt \rawbackslash \frenchspacing #1}%    {\tt \rawbackslash \plainfrenchspacing #1}%
   \null    \null
 }  }
 \let\ttfont=\t  
 \def\samp#1{`\tclose{#1}'\null}  
 \setfont\keyrm\rmshape{8}{1000}  
 \font\keysy=cmsy9  
 \def\key#1{{\keyrm\textfont2=\keysy \leavevmode\hbox{%  
   \raise0.4pt\hbox{\angleleft}\kern-.08em\vtop{%  
     \vbox{\hrule\kern-0.4pt  
      \hbox{\raise0.4pt\hbox{\vphantom{\angleleft}}#1}}%  
     \kern-0.4pt\hrule}%  
   \kern-.06em\raise0.4pt\hbox{\angleright}}}}  
 % The old definition, with no lozenge:  
 %\def\key #1{{\ttsl \nohyphenation \uppercase{#1}}\null}  
 \def\ctrl #1{{\tt \rawbackslash \hat}#1}  
   
 % @file, @option are the same as @samp.  % @samp.
 \let\file=\samp  \def\samp#1{{\setupmarkupstyle{samp}\lq\tclose{#1}\rq\null}}
 \let\option=\samp  
   
 % @code is a modification of @t,  % @indicateurl is \samp, that is, with quotes.
 % which makes spaces the same size as normal in the surrounding text.  \let\indicateurl=\samp
   
   % @code (and similar) prints in typewriter, but with spaces the same
   % size as normal in the surrounding text, without hyphenation, etc.
   % This is a subroutine for that.
 \def\tclose#1{%  \def\tclose#1{%
   {%    {%
     % Change normal interword space to be same as for the current font.      % Change normal interword space to be same as for the current font.
Line 1300  where each line of input produces a line of output.}
Line 2917  where each line of input produces a line of output.}
     \nohyphenation      \nohyphenation
     %      %
     \rawbackslash      \rawbackslash
     \frenchspacing      \plainfrenchspacing
     #1%      #1%
   }%    }%
   \null    \null % reset spacefactor to 1000
 }  }
   
 % We *must* turn on hyphenation at `-' and `_' in \code.  % We *must* turn on hyphenation at `-' and `_' in @code.
   % (But see \codedashfinish below.)
 % Otherwise, it is too hard to avoid overfull hboxes  % Otherwise, it is too hard to avoid overfull hboxes
 % in the Emacs manual, the Library manual, etc.  % in the Emacs manual, the Library manual, etc.
   %
 % Unfortunately, TeX uses one parameter (\hyphenchar) to control  % Unfortunately, TeX uses one parameter (\hyphenchar) to control
 % both hyphenation at - and hyphenation within words.  % both hyphenation at - and hyphenation within words.
 % We must therefore turn them both off (\tclose does that)  % We must therefore turn them both off (\tclose does that)
 % and arrange explicitly to hyphenate at a dash.  % and arrange explicitly to hyphenate at a dash. -- rms.
 %  -- rms.  
 {  {
   \catcode`\-=\active    \catcode`\-=\active \catcode`\_=\active
   \catcode`\_=\active    \catcode`\'=\active \catcode`\`=\active
     \global\let'=\rq \global\let`=\lq  % default definitions
   %    %
   \global\def\code{\begingroup    \global\def\code{\begingroup
     \catcode`\-=\active \let-\codedash      \setupmarkupstyle{code}%
     \catcode`\_=\active \let_\codeunder      % The following should really be moved into \setupmarkupstyle handlers.
       \catcode\dashChar=\active  \catcode\underChar=\active
       \ifallowcodebreaks
        \let-\codedash
        \let_\codeunder
       \else
        \let-\normaldash
        \let_\realunder
       \fi
       % Given -foo (with a single dash), we do not want to allow a break
       % after the hyphen.
       \global\let\codedashprev=\codedash
       %
     \codex      \codex
   }    }
   %    %
   % If we end up with any active - characters when handling the index,    \gdef\codedash{\futurelet\next\codedashfinish}
   % just treat them as a normal -.    \gdef\codedashfinish{%
   \global\def\indexbreaks{\catcode`\-=\active \let-\realdash}      \normaldash % always output the dash character itself.
       %
       % Now, output a discretionary to allow a line break, unless
       % (a) the next character is a -, or
       % (b) the preceding character is a -.
       % E.g., given --posix, we do not want to allow a break after either -.
       % Given --foo-bar, we do want to allow a break between the - and the b.
       \ifx\next\codedash \else
         \ifx\codedashprev\codedash
         \else \discretionary{}{}{}\fi
       \fi
       % we need the space after the = for the case when \next itself is a
       % space token; it would get swallowed otherwise.  As in @code{- a}.
       \global\let\codedashprev= \next
     }
 }  }
   \def\normaldash{-}
 \def\realdash{-}  %
 \def\codedash{-\discretionary{}{}{}}  
 \def\codeunder{\ifusingtt{\normalunderscore\discretionary{}{}{}}{\_}}  
 \def\codex #1{\tclose{#1}\endgroup}  \def\codex #1{\tclose{#1}\endgroup}
   
 %\let\exp=\tclose  %Was temporary  \def\codeunder{%
     % this is all so @math{@code{var_name}+1} can work.  In math mode, _
     % is "active" (mathcode"8000) and \normalunderscore (or \char95, etc.)
     % will therefore expand the active definition of _, which is us
     % (inside @code that is), therefore an endless loop.
     \ifusingtt{\ifmmode
                  \mathchar"075F % class 0=ordinary, family 7=ttfam, pos 0x5F=_.
                \else\normalunderscore \fi
                \discretionary{}{}{}}%
               {\_}%
   }
   
 % @kbd is like @code, except that if the argument is just one @key command,  % An additional complication: the above will allow breaks after, e.g.,
 % then @kbd has no effect.  % each of the four underscores in __typeof__.  This is bad.
   % @allowcodebreaks provides a document-level way to turn breaking at -
   % and _ on and off.
   %
   \newif\ifallowcodebreaks  \allowcodebreakstrue
   
 % @kbdinputstyle -- arg is `distinct' (@kbd uses slanted tty font always),  \def\keywordtrue{true}
 %   `example' (@kbd uses ttsl only inside of @example and friends),  \def\keywordfalse{false}
 %   or `code' (@kbd uses normal tty font always).  
 \def\kbdinputstyle{\parsearg\kbdinputstylexxx}  \parseargdef\allowcodebreaks{%
 \def\kbdinputstylexxx#1{%    \def\txiarg{#1}%
   \def\arg{#1}%    \ifx\txiarg\keywordtrue
   \ifx\arg\worddistinct      \allowcodebreakstrue
     \gdef\kbdexamplefont{\ttsl}\gdef\kbdfont{\ttsl}%    \else\ifx\txiarg\keywordfalse
   \else\ifx\arg\wordexample      \allowcodebreaksfalse
     \gdef\kbdexamplefont{\ttsl}\gdef\kbdfont{\tt}%    \else
   \else\ifx\arg\wordcode      \errhelp = \EMsimple
     \gdef\kbdexamplefont{\tt}\gdef\kbdfont{\tt}%      \errmessage{Unknown @allowcodebreaks option `\txiarg', must be true|false}%
   \fi\fi\fi    \fi\fi
 }  }
 \def\worddistinct{distinct}  
 \def\wordexample{example}  
 \def\wordcode{code}  
   
 % Default is kbdinputdistinct.  (Too much of a hassle to call the macro,  % For @command, @env, @file, @option quotes seem unnecessary,
 % the catcodes are wrong for parsearg to work.)  % so use \code rather than \samp.
 \gdef\kbdexamplefont{\ttsl}\gdef\kbdfont{\ttsl}  \let\command=\code
   \let\env=\code
   \let\file=\code
   \let\option=\code
   
 \def\xkey{\key}  % @uref (abbreviation for `urlref') aka @url takes an optional
 \def\kbdfoo#1#2#3\par{\def\one{#1}\def\three{#3}\def\threex{??}%  % (comma-separated) second argument specifying the text to display and
 \ifx\one\xkey\ifx\threex\three \key{#2}%  % an optional third arg as text to display instead of (rather than in
 \else{\tclose{\kbdfont\look}}\fi  % addition to) the url itself.  First (mandatory) arg is the url.
 \else{\tclose{\kbdfont\look}}\fi}  
   
 % For @url, @env, @command quotes seem unnecessary, so use \code.  % TeX-only option to allow changing PDF output to show only the second
 \let\url=\code  % arg (if given), and not the url (which is then just the link target).
 \let\env=\code  \newif\ifurefurlonlylink
 \let\command=\code  
   
 % @uref (abbreviation for `urlref') takes an optional (comma-separated)  % The main macro is \urefbreak, which allows breaking at expected
 % second argument specifying the text to display and an optional third  % places within the url.  (There used to be another version, which
 % arg as text to display instead of (rather than in addition to) the url  % didn't support automatic breaking.)
 % itself.  First (mandatory) arg is the url.  Perhaps eventually put in  \def\urefbreak{\begingroup \urefcatcodes \dourefbreak}
 % a hypertex \special here.  \let\uref=\urefbreak
 %  %
 \def\uref#1{\douref #1,,,\finish}  \def\dourefbreak#1{\urefbreakfinish #1,,,\finish}
 \def\douref#1,#2,#3,#4\finish{\begingroup  \def\urefbreakfinish#1,#2,#3,#4\finish{% doesn't work in @example
   \unsepspaces    \unsepspaces
   \pdfurl{#1}%    \pdfurl{#1}%
   \setbox0 = \hbox{\ignorespaces #3}%    \setbox0 = \hbox{\ignorespaces #3}%
   \ifdim\wd0 > 0pt    \ifdim\wd0 > 0pt
     \unhbox0 % third arg given, show only that      \unhbox0 % third arg given, show only that
   \else    \else
     \setbox0 = \hbox{\ignorespaces #2}%      \setbox0 = \hbox{\ignorespaces #2}% look for second arg
     \ifdim\wd0 > 0pt      \ifdim\wd0 > 0pt
       \ifpdf        \ifpdf
         \unhbox0             % PDF: 2nd arg given, show only it          % For pdfTeX and LuaTeX
           \ifurefurlonlylink
             % PDF plus option to not display url, show just arg
             \unhbox0
           \else
             % PDF, normally display both arg and url for consistency,
             % visibility, if the pdf is eventually used to print, etc.
             \unhbox0\ (\urefcode{#1})%
           \fi
       \else        \else
         \unhbox0\ (\code{#1})% DVI: 2nd arg given, show both it and url          \ifx\XeTeXrevision\thisisundefined
             \unhbox0\ (\urefcode{#1})% DVI, always show arg and url
           \else
             % For XeTeX
             \ifurefurlonlylink
               % PDF plus option to not display url, show just arg
               \unhbox0
             \else
               % PDF, normally display both arg and url for consistency,
               % visibility, if the pdf is eventually used to print, etc.
               \unhbox0\ (\urefcode{#1})%
             \fi
           \fi
       \fi        \fi
     \else      \else
       \code{#1}% only url given, so show it        \urefcode{#1}% only url given, so show it
     \fi      \fi
   \fi    \fi
   \endlink    \endlink
 \endgroup}  \endgroup}
   
   % Allow line breaks around only a few characters (only).
   \def\urefcatcodes{%
     \catcode`\&=\active \catcode`\.=\active
     \catcode`\#=\active \catcode`\?=\active
     \catcode`\/=\active
   }
   {
     \urefcatcodes
     %
     \global\def\urefcode{\begingroup
       \setupmarkupstyle{code}%
       \urefcatcodes
       \let&\urefcodeamp
       \let.\urefcodedot
       \let#\urefcodehash
       \let?\urefcodequest
       \let/\urefcodeslash
       \codex
     }
     %
     % By default, they are just regular characters.
     \global\def&{\normalamp}
     \global\def.{\normaldot}
     \global\def#{\normalhash}
     \global\def?{\normalquest}
     \global\def/{\normalslash}
   }
   
   % we put a little stretch before and after the breakable chars, to help
   % line breaking of long url's.  The unequal skips make look better in
   % cmtt at least, especially for dots.
   \def\urefprestretchamount{.13em}
   \def\urefpoststretchamount{.1em}
   \def\urefprestretch{\urefprebreak \hskip0pt plus\urefprestretchamount\relax}
   \def\urefpoststretch{\urefpostbreak \hskip0pt plus\urefprestretchamount\relax}
   %
   \def\urefcodeamp{\urefprestretch \&\urefpoststretch}
   \def\urefcodedot{\urefprestretch .\urefpoststretch}
   \def\urefcodehash{\urefprestretch \#\urefpoststretch}
   \def\urefcodequest{\urefprestretch ?\urefpoststretch}
   \def\urefcodeslash{\futurelet\next\urefcodeslashfinish}
   {
     \catcode`\/=\active
     \global\def\urefcodeslashfinish{%
       \urefprestretch \slashChar
       % Allow line break only after the final / in a sequence of
       % slashes, to avoid line break between the slashes in http://.
       \ifx\next/\else \urefpoststretch \fi
     }
   }
   
   % One more complication: by default we'll break after the special
   % characters, but some people like to break before the special chars, so
   % allow that.  Also allow no breaking at all, for manual control.
   %
   \parseargdef\urefbreakstyle{%
     \def\txiarg{#1}%
     \ifx\txiarg\wordnone
       \def\urefprebreak{\nobreak}\def\urefpostbreak{\nobreak}
     \else\ifx\txiarg\wordbefore
       \def\urefprebreak{\allowbreak}\def\urefpostbreak{\nobreak}
     \else\ifx\txiarg\wordafter
       \def\urefprebreak{\nobreak}\def\urefpostbreak{\allowbreak}
     \else
       \errhelp = \EMsimple
       \errmessage{Unknown @urefbreakstyle setting `\txiarg'}%
     \fi\fi\fi
   }
   \def\wordafter{after}
   \def\wordbefore{before}
   \def\wordnone{none}
   
   \urefbreakstyle after
   
   % @url synonym for @uref, since that's how everyone uses it.
   %
   \let\url=\uref
   
 % rms does not like angle brackets --karl, 17may97.  % rms does not like angle brackets --karl, 17may97.
 % So now @email is just like @uref, unless we are pdf.  % So now @email is just like @uref, unless we are pdf.
 %  %
 %\def\email#1{\angleleft{\tt #1}\angleright}  %\def\email#1{\angleleft{\tt #1}\angleright}
 \ifpdf  \ifpdf
   \def\email#1{\doemail#1,,\finish}    \def\email#1{\doemail#1,,\finish}
Line 1415  where each line of input produces a line of output.}
Line 3167  where each line of input produces a line of output.}
     \endlink      \endlink
   \endgroup}    \endgroup}
 \else  \else
   \let\email=\uref    \ifx\XeTeXrevision\thisisundefined
       \let\email=\uref
     \else
       \def\email#1{\doemail#1,,\finish}
       \def\doemail#1,#2,#3\finish{\begingroup
         \unsepspaces
         \pdfurl{mailto:#1}%
         \setbox0 = \hbox{\ignorespaces #2}%
         \ifdim\wd0>0pt\unhbox0\else\code{#1}\fi
         \endlink
       \endgroup}
     \fi
 \fi  \fi
   
 % Check if we are currently using a typewriter font.  Since all the  % @kbdinputstyle -- arg is `distinct' (@kbd uses slanted tty font always),
 % Computer Modern typewriter fonts have zero interword stretch (and  %   `example' (@kbd uses ttsl only inside of @example and friends),
 % shrink), and it is reasonable to expect all typewriter fonts to have  %   or `code' (@kbd uses normal tty font always).
 % this property, we can check that font parameter.  \parseargdef\kbdinputstyle{%
     \def\txiarg{#1}%
     \ifx\txiarg\worddistinct
       \gdef\kbdexamplefont{\ttsl}\gdef\kbdfont{\ttsl}%
     \else\ifx\txiarg\wordexample
       \gdef\kbdexamplefont{\ttsl}\gdef\kbdfont{\tt}%
     \else\ifx\txiarg\wordcode
       \gdef\kbdexamplefont{\tt}\gdef\kbdfont{\tt}%
     \else
       \errhelp = \EMsimple
       \errmessage{Unknown @kbdinputstyle setting `\txiarg'}%
     \fi\fi\fi
   }
   \def\worddistinct{distinct}
   \def\wordexample{example}
   \def\wordcode{code}
   
   % Default is `distinct'.
   \kbdinputstyle distinct
   
   % @kbd is like @code, except that if the argument is just one @key command,
   % then @kbd has no effect.
   \def\kbd#1{{\def\look{#1}\expandafter\kbdsub\look??\par}}
   
   \def\xkey{\key}
   \def\kbdsub#1#2#3\par{%
     \def\one{#1}\def\three{#3}\def\threex{??}%
     \ifx\one\xkey\ifx\threex\three \key{#2}%
     \else{\tclose{\kbdfont\setupmarkupstyle{kbd}\look}}\fi
     \else{\tclose{\kbdfont\setupmarkupstyle{kbd}\look}}\fi
   }
   
   % definition of @key that produces a lozenge.  Doesn't adjust to text size.
   %\setfont\keyrm\rmshape{8}{1000}{OT1}
   %\font\keysy=cmsy9
   %\def\key#1{{\keyrm\textfont2=\keysy \leavevmode\hbox{%
   %  \raise0.4pt\hbox{\angleleft}\kern-.08em\vtop{%
   %    \vbox{\hrule\kern-0.4pt
   %     \hbox{\raise0.4pt\hbox{\vphantom{\angleleft}}#1}}%
   %    \kern-0.4pt\hrule}%
   %  \kern-.06em\raise0.4pt\hbox{\angleright}}}}
   
   % definition of @key with no lozenge.  If the current font is already
   % monospace, don't change it; that way, we respect @kbdinputstyle.  But
   % if it isn't monospace, then use \tt.
 %  %
 \def\ifmonospace{\ifdim\fontdimen3\font=0pt }  \def\key#1{{\setupmarkupstyle{key}%
     \nohyphenation
     \ifmonospace\else\tt\fi
     #1}\null}
   
   % @clicksequence{File @click{} Open ...}
   \def\clicksequence#1{\begingroup #1\endgroup}
   
   % @clickstyle @arrow   (by default)
   \parseargdef\clickstyle{\def\click{#1}}
   \def\click{\arrow}
   
 % Typeset a dimension, e.g., `in' or `pt'.  The only reason for the  % Typeset a dimension, e.g., `in' or `pt'.  The only reason for the
 % argument is to make the input look right: @dmn{pt} instead of @dmn{}pt.  % argument is to make the input look right: @dmn{pt} instead of @dmn{}pt.
 %  %
 \def\dmn#1{\thinspace #1}  \def\dmn#1{\thinspace #1}
   
 \def\kbd#1{\def\look{#1}\expandafter\kbdfoo\look??\par}  % @acronym for "FBI", "NATO", and the like.
   % We print this one point size smaller, since it's intended for
   % all-uppercase.
   %
   \def\acronym#1{\doacronym #1,,\finish}
   \def\doacronym#1,#2,#3\finish{%
     {\selectfonts\lsize #1}%
     \def\temp{#2}%
     \ifx\temp\empty \else
       \space ({\unsepspaces \ignorespaces \temp \unskip})%
     \fi
     \null % reset \spacefactor=1000
   }
   
 % @l was never documented to mean ``switch to the Lisp font'',  % @abbr for "Comput. J." and the like.
 % and it is not used as such in any manual I can find.  We need it for  % No font change, but don't do end-of-sentence spacing.
 % Polish suppressed-l.  --karl, 22sep96.  %
 %\def\l#1{{\li #1}\null}  \def\abbr#1{\doabbr #1,,\finish}
   \def\doabbr#1,#2,#3\finish{%
     {\plainfrenchspacing #1}%
     \def\temp{#2}%
     \ifx\temp\empty \else
       \space ({\unsepspaces \ignorespaces \temp \unskip})%
     \fi
     \null % reset \spacefactor=1000
   }
   
 % Explicit font changes: @r, @sc, undocumented @ii.  % @asis just yields its argument.  Used with @table, for example.
 \def\r#1{{\rm #1}}              % roman font  %
 \def\sc#1{{\smallcaps#1}}       % smallcaps font  \def\asis#1{#1}
 \def\ii#1{{\it #1}}             % italic font  
   
 % @acronym downcases the argument and prints in smallcaps.  % @math outputs its argument in math mode.
 \def\acronym#1{{\smallcaps \lowercase{#1}}}  %
   % One complication: _ usually means subscripts, but it could also mean
   % an actual _ character, as in @math{@var{some_variable} + 1}.  So make
   % _ active, and distinguish by seeing if the current family is \slfam,
   % which is what @var uses.
   {
     \catcode`\_ = \active
     \gdef\mathunderscore{%
       \catcode`\_=\active
       \def_{\ifnum\fam=\slfam \_\else\sb\fi}%
     }
   }
   % Another complication: we want \\ (and @\) to output a math (or tt) \.
   % FYI, plain.tex uses \\ as a temporary control sequence (for no
   % particular reason), but this is not advertised and we don't care.
   %
   % The \mathchar is class=0=ordinary, family=7=ttfam, position=5C=\.
   \def\mathbackslash{\ifnum\fam=\ttfam \mathchar"075C \else\backslash \fi}
   %
   \def\math{%
     \ifmmode\else % only go into math if not in math mode already
       \tex
       \mathunderscore
       \let\\ = \mathbackslash
       \mathactive
       % make the texinfo accent commands work in math mode
       \let\"=\ddot
       \let\'=\acute
       \let\==\bar
       \let\^=\hat
       \let\`=\grave
       \let\u=\breve
       \let\v=\check
       \let\~=\tilde
       \let\dotaccent=\dot
       % have to provide another name for sup operator
       \let\mathopsup=\sup
     $\expandafter\finishmath\fi
   }
   \def\finishmath#1{#1$\endgroup}  % Close the group opened by \tex.
   
 % @pounds{} is a sterling sign.  % Some active characters (such as <) are spaced differently in math.
   % We have to reset their definitions in case the @math was an argument
   % to a command which sets the catcodes (such as @item or @section).
   %
   {
     \catcode`^ = \active
     \catcode`< = \active
     \catcode`> = \active
     \catcode`+ = \active
     \catcode`' = \active
     \gdef\mathactive{%
       \let^ = \ptexhat
       \let< = \ptexless
       \let> = \ptexgtr
       \let+ = \ptexplus
       \let' = \ptexquoteright
     }
   }
   
   % for @sub and @sup, if in math mode, just do a normal sub/superscript.
   % If in text, use math to place as sub/superscript, but switch
   % into text mode, with smaller fonts.  This is a different font than the
   % one used for real math sub/superscripts (8pt vs. 7pt), but let's not
   % fix it (significant additions to font machinery) until someone notices.
   %
   \def\sub{\ifmmode \expandafter\sb \else \expandafter\finishsub\fi}
   \def\finishsub#1{$\sb{\hbox{\selectfonts\lllsize #1}}$}%
   %
   \def\sup{\ifmmode \expandafter\ptexsp \else \expandafter\finishsup\fi}
   \def\finishsup#1{$\ptexsp{\hbox{\selectfonts\lllsize #1}}$}%
   
   % @inlinefmt{FMTNAME,PROCESSED-TEXT} and @inlineraw{FMTNAME,RAW-TEXT}.
   % Ignore unless FMTNAME == tex; then it is like @iftex and @tex,
   % except specified as a normal braced arg, so no newlines to worry about.
   %
   \def\outfmtnametex{tex}
   %
   \long\def\inlinefmt#1{\doinlinefmt #1,\finish}
   \long\def\doinlinefmt#1,#2,\finish{%
     \def\inlinefmtname{#1}%
     \ifx\inlinefmtname\outfmtnametex \ignorespaces #2\fi
   }
   %
   % @inlinefmtifelse{FMTNAME,THEN-TEXT,ELSE-TEXT} expands THEN-TEXT if
   % FMTNAME is tex, else ELSE-TEXT.
   \long\def\inlinefmtifelse#1{\doinlinefmtifelse #1,,,\finish}
   \long\def\doinlinefmtifelse#1,#2,#3,#4,\finish{%
     \def\inlinefmtname{#1}%
     \ifx\inlinefmtname\outfmtnametex \ignorespaces #2\else \ignorespaces #3\fi
   }
   %
   % For raw, must switch into @tex before parsing the argument, to avoid
   % setting catcodes prematurely.  Doing it this way means that, for
   % example, @inlineraw{html, foo{bar} gets a parse error instead of being
   % ignored.  But this isn't important because if people want a literal
   % *right* brace they would have to use a command anyway, so they may as
   % well use a command to get a left brace too.  We could re-use the
   % delimiter character idea from \verb, but it seems like overkill.
   %
   \long\def\inlineraw{\tex \doinlineraw}
   \long\def\doinlineraw#1{\doinlinerawtwo #1,\finish}
   \def\doinlinerawtwo#1,#2,\finish{%
     \def\inlinerawname{#1}%
     \ifx\inlinerawname\outfmtnametex \ignorespaces #2\fi
     \endgroup % close group opened by \tex.
   }
   
   % @inlineifset{VAR, TEXT} expands TEXT if VAR is @set.
   %
   \long\def\inlineifset#1{\doinlineifset #1,\finish}
   \long\def\doinlineifset#1,#2,\finish{%
     \def\inlinevarname{#1}%
     \expandafter\ifx\csname SET\inlinevarname\endcsname\relax
     \else\ignorespaces#2\fi
   }
   
   % @inlineifclear{VAR, TEXT} expands TEXT if VAR is not @set.
   %
   \long\def\inlineifclear#1{\doinlineifclear #1,\finish}
   \long\def\doinlineifclear#1,#2,\finish{%
     \def\inlinevarname{#1}%
     \expandafter\ifx\csname SET\inlinevarname\endcsname\relax \ignorespaces#2\fi
   }
   
   
   \message{glyphs,}
   % and logos.
   
   % @@ prints an @, as does @atchar{}.
   \def\@{\char64 }
   \let\atchar=\@
   
   % @{ @} @lbracechar{} @rbracechar{} all generate brace characters.
   \def\lbracechar{{\ifmonospace\char123\else\ensuremath\lbrace\fi}}
   \def\rbracechar{{\ifmonospace\char125\else\ensuremath\rbrace\fi}}
   \let\{=\lbracechar
   \let\}=\rbracechar
   
   % @comma{} to avoid , parsing problems.
   \let\comma = ,
   
   % Accents: @, @dotaccent @ringaccent @ubaraccent @udotaccent
   % Others are defined by plain TeX: @` @' @" @^ @~ @= @u @v @H.
   \let\, = \ptexc
   \let\dotaccent = \ptexdot
   \def\ringaccent#1{{\accent23 #1}}
   \let\tieaccent = \ptext
   \let\ubaraccent = \ptexb
   \let\udotaccent = \d
   
   % Other special characters: @questiondown @exclamdown @ordf @ordm
   % Plain TeX defines: @AA @AE @O @OE @L (plus lowercase versions) @ss.
   \def\questiondown{?`}
   \def\exclamdown{!`}
   \def\ordf{\leavevmode\raise1ex\hbox{\selectfonts\lllsize \underbar{a}}}
   \def\ordm{\leavevmode\raise1ex\hbox{\selectfonts\lllsize \underbar{o}}}
   
   % Dotless i and dotless j, used for accents.
   \def\imacro{i}
   \def\jmacro{j}
   \def\dotless#1{%
     \def\temp{#1}%
     \ifx\temp\imacro \ifmmode\imath \else\ptexi \fi
     \else\ifx\temp\jmacro \ifmmode\jmath \else\j \fi
     \else \errmessage{@dotless can be used only with i or j}%
     \fi\fi
   }
   
   % The \TeX{} logo, as in plain, but resetting the spacing so that a
   % period following counts as ending a sentence.  (Idea found in latex.)
   %
   \edef\TeX{\TeX \spacefactor=1000 }
   
   % @LaTeX{} logo.  Not quite the same results as the definition in
   % latex.ltx, since we use a different font for the raised A; it's most
   % convenient for us to use an explicitly smaller font, rather than using
   % the \scriptstyle font (since we don't reset \scriptstyle and
   % \scriptscriptstyle).
   %
   \def\LaTeX{%
     L\kern-.36em
     {\setbox0=\hbox{T}%
      \vbox to \ht0{\hbox{%
        \ifx\textnominalsize\xwordpt
          % for 10pt running text, \lllsize (8pt) is too small for the A in LaTeX.
          % Revert to plain's \scriptsize, which is 7pt.
          \count255=\the\fam $\fam\count255 \scriptstyle A$%
        \else
          % For 11pt, we can use our lllsize.
          \selectfonts\lllsize A%
        \fi
        }%
        \vss
     }}%
     \kern-.15em
     \TeX
   }
   
   % Some math mode symbols.  Define \ensuremath to switch into math mode
   % unless we are already there.  Expansion tricks may not be needed here,
   % but safer, and can't hurt.
   \def\ensuremath{\ifmmode \expandafter\asis \else\expandafter\ensuredmath \fi}
   \def\ensuredmath#1{$\relax#1$}
   %
   \def\bullet{\ensuremath\ptexbullet}
   \def\geq{\ensuremath\ge}
   \def\leq{\ensuremath\le}
   \def\minus{\ensuremath-}
   
   % @dots{} outputs an ellipsis using the current font.
   % We do .5em per period so that it has the same spacing in the cm
   % typewriter fonts as three actual period characters; on the other hand,
   % in other typewriter fonts three periods are wider than 1.5em.  So do
   % whichever is larger.
   %
   \def\dots{%
     \leavevmode
     \setbox0=\hbox{...}% get width of three periods
     \ifdim\wd0 > 1.5em
       \dimen0 = \wd0
     \else
       \dimen0 = 1.5em
     \fi
     \hbox to \dimen0{%
       \hskip 0pt plus.25fil
       .\hskip 0pt plus1fil
       .\hskip 0pt plus1fil
       .\hskip 0pt plus.5fil
     }%
   }
   
   % @enddots{} is an end-of-sentence ellipsis.
   %
   \def\enddots{%
     \dots
     \spacefactor=\endofsentencespacefactor
   }
   
   % @point{}, @result{}, @expansion{}, @print{}, @equiv{}.
   %
   % Since these characters are used in examples, they should be an even number of
   % \tt widths. Each \tt character is 1en, so two makes it 1em.
   %
   \def\point{$\star$}
   \def\arrow{\leavevmode\raise.05ex\hbox to 1em{\hfil$\rightarrow$\hfil}}
   \def\result{\leavevmode\raise.05ex\hbox to 1em{\hfil$\Rightarrow$\hfil}}
   \def\expansion{\leavevmode\hbox to 1em{\hfil$\mapsto$\hfil}}
   \def\print{\leavevmode\lower.1ex\hbox to 1em{\hfil$\dashv$\hfil}}
   \def\equiv{\leavevmode\hbox to 1em{\hfil$\ptexequiv$\hfil}}
   
   % The @error{} command.
   % Adapted from the TeXbook's \boxit.
   %
   \newbox\errorbox
   %
   {\tentt \global\dimen0 = 3em}% Width of the box.
   \dimen2 = .55pt % Thickness of rules
   % The text. (`r' is open on the right, `e' somewhat less so on the left.)
   \setbox0 = \hbox{\kern-.75pt \reducedsf \putworderror\kern-1.5pt}
   %
   \setbox\errorbox=\hbox to \dimen0{\hfil
      \hsize = \dimen0 \advance\hsize by -5.8pt % Space to left+right.
      \advance\hsize by -2\dimen2 % Rules.
      \vbox{%
         \hrule height\dimen2
         \hbox{\vrule width\dimen2 \kern3pt          % Space to left of text.
            \vtop{\kern2.4pt \box0 \kern2.4pt}% Space above/below.
            \kern3pt\vrule width\dimen2}% Space to right.
         \hrule height\dimen2}
       \hfil}
   %
   \def\error{\leavevmode\lower.7ex\copy\errorbox}
   
   % @pounds{} is a sterling sign, which Knuth put in the CM italic font.
   %
 \def\pounds{{\it\$}}  \def\pounds{{\it\$}}
   
   % @euro{} comes from a separate font, depending on the current style.
   % We use the free feym* fonts from the eurosym package by Henrik
   % Theiling, which support regular, slanted, bold and bold slanted (and
   % "outlined" (blackboard board, sort of) versions, which we don't need).
   % It is available from http://www.ctan.org/tex-archive/fonts/eurosym.
   %
   % Although only regular is the truly official Euro symbol, we ignore
   % that.  The Euro is designed to be slightly taller than the regular
   % font height.
   %
   % feymr - regular
   % feymo - slanted
   % feybr - bold
   % feybo - bold slanted
   %
   % There is no good (free) typewriter version, to my knowledge.
   % A feymr10 euro is ~7.3pt wide, while a normal cmtt10 char is ~5.25pt wide.
   % Hmm.
   %
   % Also doesn't work in math.  Do we need to do math with euro symbols?
   % Hope not.
   %
   %
   \def\euro{{\eurofont e}}
   \def\eurofont{%
     % We set the font at each command, rather than predefining it in
     % \textfonts and the other font-switching commands, so that
     % installations which never need the symbol don't have to have the
     % font installed.
     %
     % There is only one designed size (nominal 10pt), so we always scale
     % that to the current nominal size.
     %
     % By the way, simply using "at 1em" works for cmr10 and the like, but
     % does not work for cmbx10 and other extended/shrunken fonts.
     %
     \def\eurosize{\csname\curfontsize nominalsize\endcsname}%
     %
     \ifx\curfontstyle\bfstylename
       % bold:
       \font\thiseurofont = \ifusingit{feybo10}{feybr10} at \eurosize
     \else
       % regular:
       \font\thiseurofont = \ifusingit{feymo10}{feymr10} at \eurosize
     \fi
     \thiseurofont
   }
   
   % Glyphs from the EC fonts.  We don't use \let for the aliases, because
   % sometimes we redefine the original macro, and the alias should reflect
   % the redefinition.
   %
   % Use LaTeX names for the Icelandic letters.
   \def\DH{{\ecfont \char"D0}} % Eth
   \def\dh{{\ecfont \char"F0}} % eth
   \def\TH{{\ecfont \char"DE}} % Thorn
   \def\th{{\ecfont \char"FE}} % thorn
   %
   \def\guillemetleft{{\ecfont \char"13}}
   \def\guillemotleft{\guillemetleft}
   \def\guillemetright{{\ecfont \char"14}}
   \def\guillemotright{\guillemetright}
   \def\guilsinglleft{{\ecfont \char"0E}}
   \def\guilsinglright{{\ecfont \char"0F}}
   \def\quotedblbase{{\ecfont \char"12}}
   \def\quotesinglbase{{\ecfont \char"0D}}
   %
   % This positioning is not perfect (see the ogonek LaTeX package), but
   % we have the precomposed glyphs for the most common cases.  We put the
   % tests to use those glyphs in the single \ogonek macro so we have fewer
   % dummy definitions to worry about for index entries, etc.
   %
   % ogonek is also used with other letters in Lithuanian (IOU), but using
   % the precomposed glyphs for those is not so easy since they aren't in
   % the same EC font.
   \def\ogonek#1{{%
     \def\temp{#1}%
     \ifx\temp\macrocharA\Aogonek
     \else\ifx\temp\macrochara\aogonek
     \else\ifx\temp\macrocharE\Eogonek
     \else\ifx\temp\macrochare\eogonek
     \else
       \ecfont \setbox0=\hbox{#1}%
       \ifdim\ht0=1ex\accent"0C #1%
       \else\ooalign{\unhbox0\crcr\hidewidth\char"0C \hidewidth}%
       \fi
     \fi\fi\fi\fi
     }%
   }
   \def\Aogonek{{\ecfont \char"81}}\def\macrocharA{A}
   \def\aogonek{{\ecfont \char"A1}}\def\macrochara{a}
   \def\Eogonek{{\ecfont \char"86}}\def\macrocharE{E}
   \def\eogonek{{\ecfont \char"A6}}\def\macrochare{e}
   %
   % Use the European Computer Modern fonts (cm-super in outline format)
   % for non-CM glyphs.  That is ec* for regular text and tc* for the text
   % companion symbols (LaTeX TS1 encoding).  Both are part of the ec
   % package and follow the same conventions.
   %
   \def\ecfont{\etcfont{e}}
   \def\tcfont{\etcfont{t}}
   %
   \def\etcfont#1{%
     % We can't distinguish serif/sans and italic/slanted, but this
     % is used for crude hacks anyway (like adding French and German
     % quotes to documents typeset with CM, where we lose kerning), so
     % hopefully nobody will notice/care.
     \edef\ecsize{\csname\curfontsize ecsize\endcsname}%
     \edef\nominalsize{\csname\curfontsize nominalsize\endcsname}%
     \ifmonospace
       % typewriter:
       \font\thisecfont = #1ctt\ecsize \space at \nominalsize
     \else
       \ifx\curfontstyle\bfstylename
         % bold:
         \font\thisecfont = #1cb\ifusingit{i}{x}\ecsize \space at \nominalsize
       \else
         % regular:
         \font\thisecfont = #1c\ifusingit{ti}{rm}\ecsize \space at \nominalsize
       \fi
     \fi
     \thisecfont
   }
   
   % @registeredsymbol - R in a circle.  The font for the R should really
   % be smaller yet, but lllsize is the best we can do for now.
   % Adapted from the plain.tex definition of \copyright.
   %
   \def\registeredsymbol{%
     $^{{\ooalign{\hfil\raise.07ex\hbox{\selectfonts\lllsize R}%
                  \hfil\crcr\Orb}}%
       }$%
   }
   
   % @textdegree - the normal degrees sign.
   %
   \def\textdegree{$^\circ$}
   
   % Laurent Siebenmann reports \Orb undefined with:
   %  Textures 1.7.7 (preloaded format=plain 93.10.14)  (68K)  16 APR 2004 02:38
   % so we'll define it if necessary.
   %
   \ifx\Orb\thisisundefined
   \def\Orb{\mathhexbox20D}
   \fi
   
   % Quotes.
   \chardef\quotedblleft="5C
   \chardef\quotedblright=`\"
   \chardef\quoteleft=`\`
   \chardef\quoteright=`\'
   
   
 \message{page headings,}  \message{page headings,}
   
 \newskip\titlepagetopglue \titlepagetopglue = 1.5in  \newskip\titlepagetopglue \titlepagetopglue = 1.5in
Line 1458  where each line of input produces a line of output.}
Line 3719  where each line of input produces a line of output.}
 \newif\ifseenauthor  \newif\ifseenauthor
 \newif\iffinishedtitlepage  \newif\iffinishedtitlepage
   
 % Do an implicit @contents or @shortcontents after @end titlepage if the  % @setcontentsaftertitlepage used to do an implicit @contents or
 % user says @setcontentsaftertitlepage or @setshortcontentsaftertitlepage.  % @shortcontents after @end titlepage, but it is now obsolete.
 %  \def\setcontentsaftertitlepage{%
 \newif\ifsetcontentsaftertitlepage    \errmessage{@setcontentsaftertitlepage has been removed as a Texinfo
  \let\setcontentsaftertitlepage = \setcontentsaftertitlepagetrue                command; move your @contents command if you want the contents
 \newif\ifsetshortcontentsaftertitlepage                after the title page.}}%
  \let\setshortcontentsaftertitlepage = \setshortcontentsaftertitlepagetrue  \def\setshortcontentsaftertitlepage{%
     \errmessage{@setshortcontentsaftertitlepage has been removed as a Texinfo
                 command; move your @shortcontents and @contents commands if you
                 want the contents after the title page.}}%
   
 \def\shorttitlepage{\parsearg\shorttitlepagezzz}  \parseargdef\shorttitlepage{%
 \def\shorttitlepagezzz #1{\begingroup\hbox{}\vskip 1.5in \chaprm \centerline{#1}%    \begingroup \hbox{}\vskip 1.5in \chaprm \centerline{#1}%
         \endgroup\page\hbox{}\page}    \endgroup\page\hbox{}\page}
   
 \def\titlepage{\begingroup \parindent=0pt \textfonts  \envdef\titlepage{%
    \let\subtitlerm=\tenrm    % Open one extra group, as we want to close it in the middle of \Etitlepage.
    \def\subtitlefont{\subtitlerm \normalbaselineskip = 13pt \normalbaselines}%    \begingroup
    %      \parindent=0pt \textfonts
    \def\authorfont{\authorrm \normalbaselineskip = 16pt \normalbaselines}%      % Leave some space at the very top of the page.
    %      \vglue\titlepagetopglue
    % Leave some space at the very top of the page.      % No rule at page bottom unless we print one at the top with @title.
    \vglue\titlepagetopglue      \finishedtitlepagetrue
    %      %
    % Now you can print the title using @title.      % Most title ``pages'' are actually two pages long, with space
    \def\title{\parsearg\titlezzz}%      % at the top of the second.  We don't want the ragged left on the second.
    \def\titlezzz##1{\leftline{\titlefonts\rm ##1}      \let\oldpage = \page
                     % print a rule at the page bottom also.      \def\page{%
                     \finishedtitlepagefalse  
                     \vskip4pt \hrule height 4pt width \hsize \vskip4pt}%  
    % No rule at page bottom unless we print one at the top with @title.  
    \finishedtitlepagetrue  
    %  
    % Now you can put text using @subtitle.  
    \def\subtitle{\parsearg\subtitlezzz}%  
    \def\subtitlezzz##1{{\subtitlefont \rightline{##1}}}%  
    %  
    % @author should come last, but may come many times.  
    \def\author{\parsearg\authorzzz}%  
    \def\authorzzz##1{\ifseenauthor\else\vskip 0pt plus 1filll\seenauthortrue\fi  
       {\authorfont \leftline{##1}}}%  
    %  
    % Most title ``pages'' are actually two pages long, with space  
    % at the top of the second.  We don't want the ragged left on the second.  
    \let\oldpage = \page  
    \def\page{%  
       \iffinishedtitlepage\else        \iffinishedtitlepage\else
          \finishtitlepage           \finishtitlepage
       \fi        \fi
       \oldpage  
       \let\page = \oldpage        \let\page = \oldpage
       \hbox{}}%        \page
 %   \def\page{\oldpage \hbox{}}        \null
       }%
 }  }
   
 \def\Etitlepage{%  \def\Etitlepage{%
    \iffinishedtitlepage\else      \iffinishedtitlepage\else
       \finishtitlepage          \finishtitlepage
    \fi      \fi
    % It is important to do the page break before ending the group,      % It is important to do the page break before ending the group,
    % because the headline and footline are only empty inside the group.      % because the headline and footline are only empty inside the group.
    % If we use the new definition of \page, we always get a blank page      % If we use the new definition of \page, we always get a blank page
    % after the title page, which we certainly don't want.      % after the title page, which we certainly don't want.
    \oldpage      \oldpage
    \endgroup    \endgroup
    %    %
    % If they want short, they certainly want long too.    % Need this before the \...aftertitlepage checks so that if they are
    \ifsetshortcontentsaftertitlepage    % in effect the toc pages will come out with page numbers.
      \shortcontents    \HEADINGSon
      \contents  
      \global\let\shortcontents = \relax  
      \global\let\contents = \relax  
    \fi  
    %  
    \ifsetcontentsaftertitlepage  
      \contents  
      \global\let\contents = \relax  
      \global\let\shortcontents = \relax  
    \fi  
    %  
    \ifpdf \pdfmakepagedesttrue \fi  
    %  
    \HEADINGSon  
 }  }
   
 \def\finishtitlepage{%  \def\finishtitlepage{%
    \vskip4pt \hrule height 2pt width \hsize    \vskip4pt \hrule height 2pt width \hsize
    \vskip\titlepagebottomglue    \vskip\titlepagebottomglue
    \finishedtitlepagetrue    \finishedtitlepagetrue
 }  }
   
 %%% Set up page headings and footings.  % Settings used for typesetting titles: no hyphenation, no indentation,
   % don't worry much about spacing, ragged right.  This should be used
   % inside a \vbox, and fonts need to be set appropriately first.  Because
   % it is always used for titles, nothing else, we call \rmisbold.  \par
   % should be specified before the end of the \vbox, since a vbox is a group.
   %
   \def\raggedtitlesettings{%
     \rmisbold
     \hyphenpenalty=10000
     \parindent=0pt
     \tolerance=5000
     \ptexraggedright
   }
   
   % Macros to be used within @titlepage:
   
   \let\subtitlerm=\tenrm
   \def\subtitlefont{\subtitlerm \normalbaselineskip = 13pt \normalbaselines}
   
   \parseargdef\title{%
     \checkenv\titlepage
     \vbox{\titlefonts \raggedtitlesettings #1\par}%
     % print a rule at the page bottom also.
     \finishedtitlepagefalse
     \vskip4pt \hrule height 4pt width \hsize \vskip4pt
   }
   
   \parseargdef\subtitle{%
     \checkenv\titlepage
     {\subtitlefont \rightline{#1}}%
   }
   
   % @author should come last, but may come many times.
   % It can also be used inside @quotation.
   %
   \parseargdef\author{%
     \def\temp{\quotation}%
     \ifx\thisenv\temp
       \def\quotationauthor{#1}% printed in \Equotation.
     \else
       \checkenv\titlepage
       \ifseenauthor\else \vskip 0pt plus 1filll \seenauthortrue \fi
       {\secfonts\rmisbold \leftline{#1}}%
     \fi
   }
   
   
   % Set up page headings and footings.
   
 \let\thispage=\folio  \let\thispage=\folio
   
 \newtoks\evenheadline    % headline on even pages  \newtoks\evenheadline    % headline on even pages
Line 1555  where each line of input produces a line of output.}
Line 3834  where each line of input produces a line of output.}
 \newtoks\evenfootline    % footline on even pages  \newtoks\evenfootline    % footline on even pages
 \newtoks\oddfootline     % footline on odd pages  \newtoks\oddfootline     % footline on odd pages
   
 % Now make Tex use those variables  % Now make \makeheadline and \makefootline in Plain TeX use those variables
 \headline={{\textfonts\rm \ifodd\pageno \the\oddheadline  \headline={{\textfonts\rm \ifodd\pageno \the\oddheadline
                             \else \the\evenheadline \fi}}                              \else \the\evenheadline \fi}}
 \footline={{\textfonts\rm \ifodd\pageno \the\oddfootline  \footline={{\textfonts\rm \ifodd\pageno \the\oddfootline
Line 1569  where each line of input produces a line of output.}
Line 3848  where each line of input produces a line of output.}
 % @evenfooting @thisfile||  % @evenfooting @thisfile||
 % @oddfooting ||@thisfile  % @oddfooting ||@thisfile
   
 \def\evenheading{\parsearg\evenheadingxxx}  
 \def\oddheading{\parsearg\oddheadingxxx}  
 \def\everyheading{\parsearg\everyheadingxxx}  
   
 \def\evenfooting{\parsearg\evenfootingxxx}  \def\evenheading{\parsearg\evenheadingxxx}
 \def\oddfooting{\parsearg\oddfootingxxx}  \def\evenheadingxxx #1{\evenheadingyyy #1\|\|\|\|\finish}
 \def\everyfooting{\parsearg\everyfootingxxx}  \def\evenheadingyyy #1\|#2\|#3\|#4\finish{%
   
 {\catcode`\@=0 %  
   
 \gdef\evenheadingxxx #1{\evenheadingyyy #1@|@|@|@|\finish}  
 \gdef\evenheadingyyy #1@|#2@|#3@|#4\finish{%  
 \global\evenheadline={\rlap{\centerline{#2}}\line{#1\hfil#3}}}  \global\evenheadline={\rlap{\centerline{#2}}\line{#1\hfil#3}}}
   
 \gdef\oddheadingxxx #1{\oddheadingyyy #1@|@|@|@|\finish}  \def\oddheading{\parsearg\oddheadingxxx}
 \gdef\oddheadingyyy #1@|#2@|#3@|#4\finish{%  \def\oddheadingxxx #1{\oddheadingyyy #1\|\|\|\|\finish}
   \def\oddheadingyyy #1\|#2\|#3\|#4\finish{%
 \global\oddheadline={\rlap{\centerline{#2}}\line{#1\hfil#3}}}  \global\oddheadline={\rlap{\centerline{#2}}\line{#1\hfil#3}}}
   
 \gdef\everyheadingxxx#1{\oddheadingxxx{#1}\evenheadingxxx{#1}}%  \parseargdef\everyheading{\oddheadingxxx{#1}\evenheadingxxx{#1}}%
   
 \gdef\evenfootingxxx #1{\evenfootingyyy #1@|@|@|@|\finish}  \def\evenfooting{\parsearg\evenfootingxxx}
 \gdef\evenfootingyyy #1@|#2@|#3@|#4\finish{%  \def\evenfootingxxx #1{\evenfootingyyy #1\|\|\|\|\finish}
   \def\evenfootingyyy #1\|#2\|#3\|#4\finish{%
 \global\evenfootline={\rlap{\centerline{#2}}\line{#1\hfil#3}}}  \global\evenfootline={\rlap{\centerline{#2}}\line{#1\hfil#3}}}
   
 \gdef\oddfootingxxx #1{\oddfootingyyy #1@|@|@|@|\finish}  \def\oddfooting{\parsearg\oddfootingxxx}
 \gdef\oddfootingyyy #1@|#2@|#3@|#4\finish{%  \def\oddfootingxxx #1{\oddfootingyyy #1\|\|\|\|\finish}
   \def\oddfootingyyy #1\|#2\|#3\|#4\finish{%
   \global\oddfootline = {\rlap{\centerline{#2}}\line{#1\hfil#3}}%    \global\oddfootline = {\rlap{\centerline{#2}}\line{#1\hfil#3}}%
   %    %
   % Leave some space for the footline.  Hopefully ok to assume    % Leave some space for the footline.  Hopefully ok to assume
   % @evenfooting will not be used by itself.    % @evenfooting will not be used by itself.
   \global\advance\pageheight by -\baselineskip    \global\advance\txipageheight by -12pt
   \global\advance\vsize by -\baselineskip    \global\advance\vsize by -12pt
 }  }
   
 \gdef\everyfootingxxx#1{\oddfootingxxx{#1}\evenfootingxxx{#1}}  \parseargdef\everyfooting{\oddfootingxxx{#1}\evenfootingxxx{#1}}
   
   % @evenheadingmarks top     \thischapter <- chapter at the top of a page
   % @evenheadingmarks bottom  \thischapter <- chapter at the bottom of a page
 %  %
 }% unbind the catcode of @.  % The same set of arguments for:
   %
   % @oddheadingmarks
   % @evenfootingmarks
   % @oddfootingmarks
   % @everyheadingmarks
   % @everyfootingmarks
   
   % These define \getoddheadingmarks, \getevenheadingmarks,
   % \getoddfootingmarks, and \getevenfootingmarks, each to one of
   % \gettopheadingmarks, \getbottomheadingmarks.
   %
   \def\evenheadingmarks{\headingmarks{even}{heading}}
   \def\oddheadingmarks{\headingmarks{odd}{heading}}
   \def\evenfootingmarks{\headingmarks{even}{footing}}
   \def\oddfootingmarks{\headingmarks{odd}{footing}}
   \parseargdef\everyheadingmarks{\headingmarks{even}{heading}{#1}
                             \headingmarks{odd}{heading}{#1} }
   \parseargdef\everyfootingmarks{\headingmarks{even}{footing}{#1}
                             \headingmarks{odd}{footing}{#1} }
   % #1 = even/odd, #2 = heading/footing, #3 = top/bottom.
   \def\headingmarks#1#2#3 {%
     \expandafter\let\expandafter\temp \csname get#3headingmarks\endcsname
     \global\expandafter\let\csname get#1#2marks\endcsname \temp
   }
   
   \everyheadingmarks bottom
   \everyfootingmarks bottom
   
 % @headings double      turns headings on for double-sided printing.  % @headings double      turns headings on for double-sided printing.
 % @headings single      turns headings on for single-sided printing.  % @headings single      turns headings on for single-sided printing.
 % @headings off         turns them off.  % @headings off         turns them off.
Line 1617  where each line of input produces a line of output.}
Line 3921  where each line of input produces a line of output.}
 % By default, they are off at the start of a document,  % By default, they are off at the start of a document,
 % and turned `on' after @end titlepage.  % and turned `on' after @end titlepage.
   
 \def\headings #1 {\csname HEADINGS#1\endcsname}  \parseargdef\headings{\csname HEADINGS#1\endcsname}
   
 \def\HEADINGSoff{  \def\headingsoff{% non-global headings elimination
 \global\evenheadline={\hfil} \global\evenfootline={\hfil}    \evenheadline={\hfil}\evenfootline={\hfil}%
 \global\oddheadline={\hfil} \global\oddfootline={\hfil}}     \oddheadline={\hfil}\oddfootline={\hfil}%
 \HEADINGSoff  }
   
   \def\HEADINGSoff{{\globaldefs=1 \headingsoff}} % global setting
   \HEADINGSoff  % it's the default
   
 % When we turn headings on, set the page number to 1.  % When we turn headings on, set the page number to 1.
 % For double-sided printing, put current file name in lower left corner,  % For double-sided printing, put current file name in lower left corner,
 % chapter name on inside top of right hand pages, document  % chapter name on inside top of right hand pages, document
 % title on inside top of left hand pages, and page numbers on outside top  % title on inside top of left hand pages, and page numbers on outside top
 % edge of all pages.  % edge of all pages.
 \def\HEADINGSdouble{  \def\HEADINGSdouble{%
 \global\pageno=1  \global\pageno=1
 \global\evenfootline={\hfil}  \global\evenfootline={\hfil}
 \global\oddfootline={\hfil}  \global\oddfootline={\hfil}
 \global\evenheadline={\line{\folio\hfil\thistitle}}  \global\evenheadline={\line{\folio\hfil\thistitle}}
 \global\oddheadline={\line{\thischapter\hfil\folio}}  \global\oddheadline={\line{\thischapterheading\hfil\folio}}
 \global\let\contentsalignmacro = \chapoddpage  \global\let\contentsalignmacro = \chapoddpage
 }  }
 \let\contentsalignmacro = \chappager  \let\contentsalignmacro = \chappager
   
 % For single-sided printing, chapter title goes across top left of page,  % For single-sided printing, chapter title goes across top left of page,
 % page number on top right.  % page number on top right.
 \def\HEADINGSsingle{  \def\HEADINGSsingle{%
 \global\pageno=1  \global\pageno=1
 \global\evenfootline={\hfil}  \global\evenfootline={\hfil}
 \global\oddfootline={\hfil}  \global\oddfootline={\hfil}
 \global\evenheadline={\line{\thischapter\hfil\folio}}  \global\evenheadline={\line{\thischapterheading\hfil\folio}}
 \global\oddheadline={\line{\thischapter\hfil\folio}}  \global\oddheadline={\line{\thischapterheading\hfil\folio}}
 \global\let\contentsalignmacro = \chappager  \global\let\contentsalignmacro = \chappager
 }  }
 \def\HEADINGSon{\HEADINGSdouble}  \def\HEADINGSon{\HEADINGSdouble}
Line 1656  where each line of input produces a line of output.}
Line 3964  where each line of input produces a line of output.}
 \global\evenfootline={\hfil}  \global\evenfootline={\hfil}
 \global\oddfootline={\hfil}  \global\oddfootline={\hfil}
 \global\evenheadline={\line{\folio\hfil\thistitle}}  \global\evenheadline={\line{\folio\hfil\thistitle}}
 \global\oddheadline={\line{\thischapter\hfil\folio}}  \global\oddheadline={\line{\thischapterheading\hfil\folio}}
 \global\let\contentsalignmacro = \chapoddpage  \global\let\contentsalignmacro = \chapoddpage
 }  }
   
Line 1664  where each line of input produces a line of output.}
Line 3972  where each line of input produces a line of output.}
 \def\HEADINGSsinglex{%  \def\HEADINGSsinglex{%
 \global\evenfootline={\hfil}  \global\evenfootline={\hfil}
 \global\oddfootline={\hfil}  \global\oddfootline={\hfil}
 \global\evenheadline={\line{\thischapter\hfil\folio}}  \global\evenheadline={\line{\thischapterheading\hfil\folio}}
 \global\oddheadline={\line{\thischapter\hfil\folio}}  \global\oddheadline={\line{\thischapterheading\hfil\folio}}
 \global\let\contentsalignmacro = \chappager  \global\let\contentsalignmacro = \chappager
 }  }
   
 % Subroutines used in generating headings  % Subroutines used in generating headings
 % Produces Day Month Year style of output.  % This produces Day Month Year style of output.
   % Only define if not already defined, in case a txi-??.tex file has set
   % up a different format (e.g., txi-cs.tex does this).
   \ifx\today\thisisundefined
 \def\today{%  \def\today{%
   \number\day\space    \number\day\space
   \ifcase\month    \ifcase\month
Line 1679  where each line of input produces a line of output.}
Line 3990  where each line of input produces a line of output.}
   \or\putwordMSep\or\putwordMOct\or\putwordMNov\or\putwordMDec    \or\putwordMSep\or\putwordMOct\or\putwordMNov\or\putwordMDec
   \fi    \fi
   \space\number\year}    \space\number\year}
   \fi
   
 % @settitle line...  specifies the title of the document, for headings.  % @settitle line...  specifies the title of the document, for headings.
 % It generates no output of its own.  % It generates no output of its own.
 \def\thistitle{\putwordNoTitle}  \def\thistitle{\putwordNoTitle}
 \def\settitle{\parsearg\settitlezzz}  \def\settitle{\parsearg{\gdef\thistitle}}
 \def\settitlezzz #1{\gdef\thistitle{#1}}  
   
   
 \message{tables,}  \message{tables,}
 % Tables -- @table, @ftable, @vtable, @item(x), @kitem(x), @xitem(x).  % Tables -- @table, @ftable, @vtable, @item(x).
   
 % default indentation of table text  % default indentation of table text
 \newdimen\tableindent \tableindent=.8in  \newdimen\tableindent \tableindent=.8in
Line 1700  where each line of input produces a line of output.}
Line 4011  where each line of input produces a line of output.}
 % used internally for \itemindent minus \itemmargin  % used internally for \itemindent minus \itemmargin
 \newdimen\itemmax  \newdimen\itemmax
   
 % Note @table, @vtable, and @vtable define @item, @itemx, etc., with  % Note @table, @ftable, and @vtable define @item, @itemx, etc., with
 % these defs.  % these defs.
 % They also define \itemindex  % They also define \itemindex
 % to index the item name in whatever manner is desired (perhaps none).  % to index the item name in whatever manner is desired (perhaps none).
Line 1712  where each line of input produces a line of output.}
Line 4023  where each line of input produces a line of output.}
 \def\internalBitem{\smallbreak \parsearg\itemzzz}  \def\internalBitem{\smallbreak \parsearg\itemzzz}
 \def\internalBitemx{\itemxpar \parsearg\itemzzz}  \def\internalBitemx{\itemxpar \parsearg\itemzzz}
   
 \def\internalBxitem "#1"{\def\xitemsubtopix{#1} \smallbreak \parsearg\xitemzzz}  
 \def\internalBxitemx "#1"{\def\xitemsubtopix{#1} \itemxpar \parsearg\xitemzzz}  
   
 \def\internalBkitem{\smallbreak \parsearg\kitemzzz}  
 \def\internalBkitemx{\itemxpar \parsearg\kitemzzz}  
   
 \def\kitemzzz #1{\dosubind {kw}{\code{#1}}{for {\bf \lastfunction}}%  
                  \itemzzz {#1}}  
   
 \def\xitemzzz #1{\dosubind {kw}{\code{#1}}{for {\bf \xitemsubtopic}}%  
                  \itemzzz {#1}}  
   
 \def\itemzzz #1{\begingroup %  \def\itemzzz #1{\begingroup %
   \advance\hsize by -\rightskip    \advance\hsize by -\rightskip
   \advance\hsize by -\tableindent    \advance\hsize by -\tableindent
   \setbox0=\hbox{\itemfont{#1}}%    \setbox0=\hbox{\itemindicate{#1}}%
   \itemindex{#1}%    \itemindex{#1}%
   \nobreak % This prevents a break before @itemx.    \nobreak % This prevents a break before @itemx.
   %    %
Line 1743  where each line of input produces a line of output.}
Line 4042  where each line of input produces a line of output.}
     \begingroup      \begingroup
       \advance\leftskip by-\tableindent        \advance\leftskip by-\tableindent
       \advance\hsize by\tableindent        \advance\hsize by\tableindent
       \advance\rightskip by0pt plus1fil        \advance\rightskip by0pt plus1fil\relax
       \leavevmode\unhbox0\par        \leavevmode\unhbox0\par
     \endgroup      \endgroup
     %      %
Line 1751  where each line of input produces a line of output.}
Line 4050  where each line of input produces a line of output.}
     % \parskip glue -- logically it's part of the @item we just started.      % \parskip glue -- logically it's part of the @item we just started.
     \nobreak \vskip-\parskip      \nobreak \vskip-\parskip
     %      %
     % Stop a page break at the \parskip glue coming up.  Unfortunately      % Stop a page break at the \parskip glue coming up.  However, if
     % we can't prevent a possible page break at the following      % what follows is an environment such as @example, there will be no
     % \baselineskip glue.      % \parskip glue; then the negative vskip we just inserted would
     \nobreak      % cause the example and the item to crash together.  So we use this
       % bizarre value of 10001 as a signal to \aboveenvbreak to insert
       % \parskip glue after all.  Section titles are handled this way also.
       %
       \penalty 10001
     \endgroup      \endgroup
     \itemxneedsnegativevskipfalse      \itemxneedsnegativevskipfalse
   \else    \else
Line 1773  where each line of input produces a line of output.}
Line 4076  where each line of input produces a line of output.}
   \fi    \fi
 }  }
   
 \def\item{\errmessage{@item while not in a table}}  \def\item{\errmessage{@item while not in a list environment}}
 \def\itemx{\errmessage{@itemx while not in a table}}  \def\itemx{\errmessage{@itemx while not in a list environment}}
 \def\kitem{\errmessage{@kitem while not in a table}}  
 \def\kitemx{\errmessage{@kitemx while not in a table}}  
 \def\xitem{\errmessage{@xitem while not in a table}}  
 \def\xitemx{\errmessage{@xitemx while not in a table}}  
   
 % Contains a kludge to get @end[description] to work.  
 \def\description{\tablez{\dontindex}{1}{}{}{}{}}  
   
 % @table, @ftable, @vtable.  % @table, @ftable, @vtable.
 \def\table{\begingroup\inENV\obeylines\obeyspaces\tablex}  \envdef\table{%
 {\obeylines\obeyspaces%    \let\itemindex\gobble
 \gdef\tablex #1^^M{%    \tablecheck{table}%
 \tabley\dontindex#1        \endtabley}}  
   
 \def\ftable{\begingroup\inENV\obeylines\obeyspaces\ftablex}  
 {\obeylines\obeyspaces%  
 \gdef\ftablex #1^^M{%  
 \tabley\fnitemindex#1        \endtabley  
 \def\Eftable{\endgraf\afterenvbreak\endgroup}%  
 \let\Etable=\relax}}  
   
 \def\vtable{\begingroup\inENV\obeylines\obeyspaces\vtablex}  
 {\obeylines\obeyspaces%  
 \gdef\vtablex #1^^M{%  
 \tabley\vritemindex#1        \endtabley  
 \def\Evtable{\endgraf\afterenvbreak\endgroup}%  
 \let\Etable=\relax}}  
   
 \def\dontindex #1{}  
 \def\fnitemindex #1{\doind {fn}{\code{#1}}}%  
 \def\vritemindex #1{\doind {vr}{\code{#1}}}%  
   
 {\obeyspaces %  
 \gdef\tabley#1#2 #3 #4 #5 #6 #7\endtabley{\endgroup%  
 \tablez{#1}{#2}{#3}{#4}{#5}{#6}}}  
   
 \def\tablez #1#2#3#4#5#6{%  
 \aboveenvbreak %  
 \begingroup %  
 \def\Edescription{\Etable}% Necessary kludge.  
 \let\itemindex=#1%  
 \ifnum 0#3>0 \advance \leftskip by #3\mil \fi %  
 \ifnum 0#4>0 \tableindent=#4\mil \fi %  
 \ifnum 0#5>0 \advance \rightskip by #5\mil \fi %  
 \def\itemfont{#2}%  
 \itemmax=\tableindent %  
 \advance \itemmax by -\itemmargin %  
 \advance \leftskip by \tableindent %  
 \exdentamount=\tableindent  
 \parindent = 0pt  
 \parskip = \smallskipamount  
 \ifdim \parskip=0pt \parskip=2pt \fi%  
 \def\Etable{\endgraf\afterenvbreak\endgroup}%  
 \let\item = \internalBitem %  
 \let\itemx = \internalBitemx %  
 \let\kitem = \internalBkitem %  
 \let\kitemx = \internalBkitemx %  
 \let\xitem = \internalBxitem %  
 \let\xitemx = \internalBxitemx %  
 }  }
   \envdef\ftable{%
     \def\itemindex ##1{\doind {fn}{\code{##1}}}%
     \tablecheck{ftable}%
   }
   \envdef\vtable{%
     \def\itemindex ##1{\doind {vr}{\code{##1}}}%
     \tablecheck{vtable}%
   }
   \def\tablecheck#1{%
     \ifnum \the\catcode`\^^M=\active
       \endgroup
       \errmessage{This command won't work in this context; perhaps the problem is
         that we are \inenvironment\thisenv}%
       \def\next{\doignore{#1}}%
     \else
       \let\next\tablex
     \fi
     \next
   }
   \def\tablex#1{%
     \def\itemindicate{#1}%
     \parsearg\tabley
   }
   \def\tabley#1{%
     {%
       \makevalueexpandable
       \edef\temp{\noexpand\tablez #1\space\space\space}%
       \expandafter
     }\temp \endtablez
   }
   \def\tablez #1 #2 #3 #4\endtablez{%
     \aboveenvbreak
     \ifnum 0#1>0 \advance \leftskip by #1\mil \fi
     \ifnum 0#2>0 \tableindent=#2\mil \fi
     \ifnum 0#3>0 \advance \rightskip by #3\mil \fi
     \itemmax=\tableindent
     \advance \itemmax by -\itemmargin
     \advance \leftskip by \tableindent
     \exdentamount=\tableindent
     \parindent = 0pt
     \parskip = \smallskipamount
     \ifdim \parskip=0pt \parskip=2pt \fi
     \let\item = \internalBitem
     \let\itemx = \internalBitemx
   }
   \def\Etable{\endgraf\afterenvbreak}
   \let\Eftable\Etable
   \let\Evtable\Etable
   \let\Eitemize\Etable
   \let\Eenumerate\Etable
   
 % This is the counter used by @enumerate, which is really @itemize  % This is the counter used by @enumerate, which is really @itemize
   
 \newcount \itemno  \newcount \itemno
   
 \def\itemize{\parsearg\itemizezzz}  \envdef\itemize{\parsearg\doitemize}
   
 \def\itemizezzz #1{%  \def\doitemize#1{%
   \begingroup % ended by the @end itemize    \aboveenvbreak
   \itemizey {#1}{\Eitemize}    \itemmax=\itemindent
     \advance\itemmax by -\itemmargin
     \advance\leftskip by \itemindent
     \exdentamount=\itemindent
     \parindent=0pt
     \parskip=\smallskipamount
     \ifdim\parskip=0pt \parskip=2pt \fi
     %
     % Try typesetting the item mark so that if the document erroneously says
     % something like @itemize @samp (intending @table), there's an error
     % right away at the @itemize.  It's not the best error message in the
     % world, but it's better than leaving it to the @item.  This means if
     % the user wants an empty mark, they have to say @w{} not just @w.
     \def\itemcontents{#1}%
     \setbox0 = \hbox{\itemcontents}%
     %
     % @itemize with no arg is equivalent to @itemize @bullet.
     \ifx\itemcontents\empty\def\itemcontents{\bullet}\fi
     %
     \let\item=\itemizeitem
 }  }
   
 \def\itemizey #1#2{%  % Definition of @item while inside @itemize and @enumerate.
 \aboveenvbreak %  %
 \itemmax=\itemindent %  \def\itemizeitem{%
 \advance \itemmax by -\itemmargin %    \advance\itemno by 1  % for enumerations
 \advance \leftskip by \itemindent %    {\let\par=\endgraf \smallbreak}% reasonable place to break
 \exdentamount=\itemindent    {%
 \parindent = 0pt %     % If the document has an @itemize directly after a section title, a
 \parskip = \smallskipamount %     % \nobreak will be last on the list, and \sectionheading will have
 \ifdim \parskip=0pt \parskip=2pt \fi%     % done a \vskip-\parskip.  In that case, we don't want to zero
 \def#2{\endgraf\afterenvbreak\endgroup}%     % parskip, or the item text will crash with the heading.  On the
 \def\itemcontents{#1}%     % other hand, when there is normal text preceding the item (as there
 \let\item=\itemizeitem}     % usually is), we do want to zero parskip, or there would be too much
      % space.  In that case, we won't have a \nobreak before.  At least
      % that's the theory.
      \ifnum\lastpenalty<10000 \parskip=0in \fi
      \noindent
      \hbox to 0pt{\hss \itemcontents \kern\itemmargin}%
      %
      \ifinner\else
        \vadjust{\penalty 1200}% not good to break after first line of item.
      \fi
      % We can be in inner vertical mode in a footnote, although an
      % @itemize looks awful there.
     }%
     \flushcr
   }
   
 % Set sfcode to normal for the chars that usually have another value.  
 % These are `.?!:;,'  
 \def\frenchspacing{\sfcode46=1000 \sfcode63=1000 \sfcode33=1000  
   \sfcode58=1000 \sfcode59=1000 \sfcode44=1000 }  
   
 % \splitoff TOKENS\endmark defines \first to be the first token in  % \splitoff TOKENS\endmark defines \first to be the first token in
 % TOKENS, and \rest to be the remainder.  % TOKENS, and \rest to be the remainder.
 %  %
Line 1874  where each line of input produces a line of output.}
Line 4201  where each line of input produces a line of output.}
 % or number, to specify the first label in the enumerated list.  No  % or number, to specify the first label in the enumerated list.  No
 % argument is the same as `1'.  % argument is the same as `1'.
 %  %
 \def\enumerate{\parsearg\enumeratezzz}  \envparseargdef\enumerate{\enumeratey #1  \endenumeratey}
 \def\enumeratezzz #1{\enumeratey #1  \endenumeratey}  
 \def\enumeratey #1 #2\endenumeratey{%  \def\enumeratey #1 #2\endenumeratey{%
   \begingroup % ended by the @end enumerate  
   %  
   % If we were given no argument, pretend we were given `1'.    % If we were given no argument, pretend we were given `1'.
   \def\thearg{#1}%    \def\thearg{#1}%
   \ifx\thearg\empty \def\thearg{1}\fi    \ifx\thearg\empty \def\thearg{1}\fi
Line 1949  where each line of input produces a line of output.}
Line 4273  where each line of input produces a line of output.}
   }%    }%
 }  }
   
 % Call itemizey, adding a period to the first argument and supplying the  % Call \doitemize, adding a period to the first argument and supplying the
 % common last two arguments.  Also subtract one from the initial value in  % common last two arguments.  Also subtract one from the initial value in
 % \itemno, since @item increments \itemno.  % \itemno, since @item increments \itemno.
 %  %
 \def\startenumeration#1{%  \def\startenumeration#1{%
   \advance\itemno by -1    \advance\itemno by -1
   \itemizey{#1.}\Eenumerate\flushcr    \doitemize{#1.}\flushcr
 }  }
   
 % @alphaenumerate and @capsenumerate are abbreviations for giving an arg  % @alphaenumerate and @capsenumerate are abbreviations for giving an arg
Line 1966  where each line of input produces a line of output.}
Line 4290  where each line of input produces a line of output.}
 \def\Ealphaenumerate{\Eenumerate}  \def\Ealphaenumerate{\Eenumerate}
 \def\Ecapsenumerate{\Eenumerate}  \def\Ecapsenumerate{\Eenumerate}
   
 % Definition of @item while inside @itemize.  
   
 \def\itemizeitem{%  
 \advance\itemno by 1  
 {\let\par=\endgraf \smallbreak}%  
 \ifhmode \errmessage{In hmode at itemizeitem}\fi  
 {\parskip=0in \hskip 0pt  
 \hbox to 0pt{\hss \itemcontents\hskip \itemmargin}%  
 \vadjust{\penalty 1200}}%  
 \flushcr}  
   
 % @multitable macros  % @multitable macros
 % Amy Hendrickson, 8/18/94, 3/6/96  % Amy Hendrickson, 8/18/94, 3/6/96
 %  %
Line 2002  where each line of input produces a line of output.}
Line 4316  where each line of input produces a line of output.}
 %   @multitable {Column 1 template} {Column 2 template} {Column 3 template}  %   @multitable {Column 1 template} {Column 2 template} {Column 3 template}
 %   @item ...  %   @item ...
 %   using the widest term desired in each column.  %   using the widest term desired in each column.
 %  
 % For those who want to use more than one line's worth of words in  
 % the preamble, break the line within one argument and it  
 % will parse correctly, i.e.,  
 %  
 %     @multitable {Column 1 template} {Column 2 template} {Column 3  
 %      template}  
 % Not:  
 %     @multitable {Column 1 template} {Column 2 template}  
 %      {Column 3 template}  
   
 % Each new table line starts with @item, each subsequent new column  % Each new table line starts with @item, each subsequent new column
 % starts with @tab. Empty columns may be produced by supplying @tab's  % starts with @tab. Empty columns may be produced by supplying @tab's
 % with nothing between them for as many times as empty columns are needed,  % with nothing between them for as many times as empty columns are needed,
 % ie, @tab@tab@tab will produce two empty columns.  % ie, @tab@tab@tab will produce two empty columns.
   
 % @item, @tab, @multitable or @end multitable do not need to be on their  % @item, @tab do not need to be on their own lines, but it will not hurt
 % own lines, but it will not hurt if they are.  % if they are.
   
 % Sample multitable:  % Sample multitable:
   
Line 2063  where each line of input produces a line of output.}
Line 4367  where each line of input produces a line of output.}
 \def\xcolumnfractions{\columnfractions}  \def\xcolumnfractions{\columnfractions}
 \newif\ifsetpercent  \newif\ifsetpercent
   
 % #1 is the part of the @columnfraction before the decimal point, which  % #1 is the @columnfraction, usually a decimal number like .5, but might
 % is presumably either 0 or the empty string (but we don't check, we  % be just 1.  We just use it, whatever it is.
 % just throw it away).  #2 is the decimal part, which we use as the  %
 % percent of \hsize for this column.  \def\pickupwholefraction#1 {%
 \def\pickupwholefraction#1.#2 {%  
   \global\advance\colcount by 1    \global\advance\colcount by 1
   \expandafter\xdef\csname col\the\colcount\endcsname{.#2\hsize}%    \expandafter\xdef\csname col\the\colcount\endcsname{#1\hsize}%
   \setuptable    \setuptable
 }  }
   
Line 2086  where each line of input produces a line of output.}
Line 4389  where each line of input produces a line of output.}
          \let\go\pickupwholefraction           \let\go\pickupwholefraction
       \else        \else
          \global\advance\colcount by 1           \global\advance\colcount by 1
          \setbox0=\hbox{#1\unskip }% Add a normal word space as a separator;           \setbox0=\hbox{#1\unskip\space}% Add a normal word space as a
                             % typically that is always in the input, anyway.                     % separator; typically that is always in the input, anyway.
          \expandafter\xdef\csname col\the\colcount\endcsname{\the\wd0}%           \expandafter\xdef\csname col\the\colcount\endcsname{\the\wd0}%
       \fi        \fi
     \fi      \fi
Line 2102  where each line of input produces a line of output.}
Line 4405  where each line of input produces a line of output.}
   \go    \go
 }  }
   
 % This used to have \hskip1sp.  But then the space in a template line is  % multitable-only commands.
 % not enough.  That is bad.  So let's go back to just & until we  %
 % encounter the problem it was intended to solve again.  % @headitem starts a heading row, which we typeset in bold.  Assignments
 % --karl, nathan@acm.org, 20apr99.  % have to be global since we are inside the implicit group of an
 \def\tab{&}  % alignment entry.  \everycr below resets \everytab so we don't have to
   % undo it ourselves.
   \def\headitemfont{\b}% for people to use in the template row; not changeable
   \def\headitem{%
     \checkenv\multitable
     \crcr
     \gdef\headitemcrhook{\nobreak}% attempt to avoid page break after headings
     \global\everytab={\bf}% can't use \headitemfont since the parsing differs
     \the\everytab % for the first item
   }%
   %
   % default for tables with no headings.
   \let\headitemcrhook=\relax
   %
   % A \tab used to include \hskip1sp.  But then the space in a template
   % line is not enough.  That is bad.  So let's go back to just `&' until
   % we again encounter the problem the 1sp was intended to solve.
   %                                       --karl, nathan@acm.org, 20apr99.
   \def\tab{\checkenv\multitable &\the\everytab}%
   
 % @multitable ... @end multitable definitions:  % @multitable ... @end multitable definitions:
 %  %
 \def\multitable{\parsearg\dotable}  \newtoks\everytab  % insert after every tab.
 \def\dotable#1{\bgroup  %
   \envdef\multitable{%
   \vskip\parskip    \vskip\parskip
   \let\item\crcr    \startsavinginserts
     %
     % @item within a multitable starts a normal row.
     % We use \def instead of \let so that if one of the multitable entries
     % contains an @itemize, we don't choke on the \item (seen as \crcr aka
     % \endtemplate) expanding \doitemize.
     \def\item{\crcr}%
     %
   \tolerance=9500    \tolerance=9500
   \hbadness=9500    \hbadness=9500
   \setmultitablespacing    \setmultitablespacing
Line 2121  where each line of input produces a line of output.}
Line 4450  where each line of input produces a line of output.}
   \parindent=\multitableparindent    \parindent=\multitableparindent
   \overfullrule=0pt    \overfullrule=0pt
   \global\colcount=0    \global\colcount=0
   \def\Emultitable{\global\setpercentfalse\cr\egroup\egroup}%  
   %    %
     \everycr = {%
       \noalign{%
         \global\everytab={}% Reset from possible headitem.
         \global\colcount=0 % Reset the column counter.
         %
         % Check for saved footnotes, etc.:
         \checkinserts
         %
         % Perhaps a \nobreak, then reset:
         \headitemcrhook
         \global\let\headitemcrhook=\relax
       }%
     }%
     %
     \parsearg\domultitable
   }
   \def\domultitable#1{%
   % To parse everything between @multitable and @item:    % To parse everything between @multitable and @item:
   \setuptable#1 \endsetuptable    \setuptable#1 \endsetuptable
   %    %
   % \everycr will reset column counter, \colcount, at the end of  
   % each line. Every column entry will cause \colcount to advance by one.  
   % The table preamble  
   % looks at the current \colcount to find the correct column width.  
   \everycr{\noalign{%  
   %  
   % \filbreak%% keeps underfull box messages off when table breaks over pages.  
   % Maybe so, but it also creates really weird page breaks when the table  
   % breaks over pages. Wouldn't \vfil be better?  Wait until the problem  
   % manifests itself, so it can be fixed for real --karl.  
     \global\colcount=0\relax}}%  
   %  
   % This preamble sets up a generic column definition, which will    % This preamble sets up a generic column definition, which will
   % be used as many times as user calls for columns.    % be used as many times as user calls for columns.
   % \vtop will set a single line and will also let text wrap and    % \vtop will set a single line and will also let text wrap and
   % continue for many paragraphs if desired.    % continue for many paragraphs if desired.
   \halign\bgroup&\global\advance\colcount by 1\relax    \halign\bgroup &%
     \multistrut\vtop{\hsize=\expandafter\csname col\the\colcount\endcsname      \global\advance\colcount by 1
   %      \multistrut
   % In order to keep entries from bumping into each other      \vtop{%
   % we will add a \leftskip of \multitablecolspace to all columns after        % Use the current \colcount to find the correct column width:
   % the first one.        \hsize=\expandafter\csname col\the\colcount\endcsname
   %        %
   % If a template has been used, we will add \multitablecolspace        % In order to keep entries from bumping into each other
   % to the width of each template entry.        % we will add a \leftskip of \multitablecolspace to all columns after
   %        % the first one.
   % If the user has set preamble in terms of percent of \hsize we will        %
   % use that dimension as the width of the column, and the \leftskip        % If a template has been used, we will add \multitablecolspace
   % will keep entries from bumping into each other.  Table will start at        % to the width of each template entry.
   % left margin and final column will justify at right margin.        %
   %        % If the user has set preamble in terms of percent of \hsize we will
   % Make sure we don't inherit \rightskip from the outer environment.        % use that dimension as the width of the column, and the \leftskip
   \rightskip=0pt        % will keep entries from bumping into each other.  Table will start at
   \ifnum\colcount=1        % left margin and final column will justify at right margin.
     % The first column will be indented with the surrounding text.        %
     \advance\hsize by\leftskip        % Make sure we don't inherit \rightskip from the outer environment.
   \else        \rightskip=0pt
     \ifsetpercent \else        \ifnum\colcount=1
       % If user has not set preamble in terms of percent of \hsize          % The first column will be indented with the surrounding text.
       % we will advance \hsize by \multitablecolspace.          \advance\hsize by\leftskip
       \advance\hsize by \multitablecolspace        \else
     \fi          \ifsetpercent \else
    % In either case we will make \leftskip=\multitablecolspace:            % If user has not set preamble in terms of percent of \hsize
   \leftskip=\multitablecolspace            % we will advance \hsize by \multitablecolspace.
   \fi            \advance\hsize by \multitablecolspace
   % Ignoring space at the beginning and end avoids an occasional spurious          \fi
   % blank line, when TeX decides to break the line at the space before the         % In either case we will make \leftskip=\multitablecolspace:
   % box from the multistrut, so the strut ends up on a line by itself.        \leftskip=\multitablecolspace
   % For example:        \fi
   % @multitable @columnfractions .11 .89        % Ignoring space at the beginning and end avoids an occasional spurious
   % @item @code{#}        % blank line, when TeX decides to break the line at the space before the
   % @tab Legal holiday which is valid in major parts of the whole country.        % box from the multistrut, so the strut ends up on a line by itself.
   % Is automatically provided with highlighting sequences respectively marking        % For example:
   % characters.        % @multitable @columnfractions .11 .89
   \noindent\ignorespaces##\unskip\multistrut}\cr        % @item @code{#}
         % @tab Legal holiday which is valid in major parts of the whole country.
         % Is automatically provided with highlighting sequences respectively
         % marking characters.
         \noindent\ignorespaces##\unskip\multistrut
       }\cr
 }  }
   \def\Emultitable{%
     \crcr
     \egroup % end the \halign
     \global\setpercentfalse
   }
   
 \def\setmultitablespacing{% test to see if user has set \multitablelinespace.  \def\setmultitablespacing{%
 % If so, do nothing. If not, give it an appropriate dimension based on    \def\multistrut{\strut}% just use the standard line spacing
 % current baselineskip.    %
     % Compute \multitablelinespace (if not defined by user) for use in
     % \multitableparskip calculation.  We used define \multistrut based on
     % this, but (ironically) that caused the spacing to be off.
     % See bug-texinfo report from Werner Lemberg, 31 Oct 2004 12:52:20 +0100.
 \ifdim\multitablelinespace=0pt  \ifdim\multitablelinespace=0pt
 \setbox0=\vbox{X}\global\multitablelinespace=\the\baselineskip  \setbox0=\vbox{X}\global\multitablelinespace=\the\baselineskip
 \global\advance\multitablelinespace by-\ht0  \global\advance\multitablelinespace by-\ht0
 %% strut to put in table in case some entry doesn't have descenders,  \fi
 %% to keep lines equally spaced  % Test to see if parskip is larger than space between lines of
 \let\multistrut = \strut  % table. If not, do nothing.
 \else  %        If so, set to same dimension as multitablelinespace.
 %% FIXME: what is \box0 supposed to be?  
 \gdef\multistrut{\vrule height\multitablelinespace depth\dp0  
 width0pt\relax} \fi  
 %% Test to see if parskip is larger than space between lines of  
 %% table. If not, do nothing.  
 %%        If so, set to same dimension as multitablelinespace.  
 \ifdim\multitableparskip>\multitablelinespace  \ifdim\multitableparskip>\multitablelinespace
 \global\multitableparskip=\multitablelinespace  \global\multitableparskip=\multitablelinespace
 \global\advance\multitableparskip-7pt %% to keep parskip somewhat smaller  \global\advance\multitableparskip-7pt % to keep parskip somewhat smaller
                                       %% than skip between lines in the table.                                        % than skip between lines in the table.
 \fi%  \fi%
 \ifdim\multitableparskip=0pt  \ifdim\multitableparskip=0pt
 \global\multitableparskip=\multitablelinespace  \global\multitableparskip=\multitablelinespace
 \global\advance\multitableparskip-7pt %% to keep parskip somewhat smaller  \global\advance\multitableparskip-7pt % to keep parskip somewhat smaller
                                       %% than skip between lines in the table.                                        % than skip between lines in the table.
 \fi}  \fi}
   
   
 \message{conditionals,}  \message{conditionals,}
 % Prevent errors for section commands.  
 % Used in @ignore and in failing conditionals.  
 \def\ignoresections{%  
   \let\chapter=\relax  
   \let\unnumbered=\relax  
   \let\top=\relax  
   \let\unnumberedsec=\relax  
   \let\unnumberedsection=\relax  
   \let\unnumberedsubsec=\relax  
   \let\unnumberedsubsection=\relax  
   \let\unnumberedsubsubsec=\relax  
   \let\unnumberedsubsubsection=\relax  
   \let\section=\relax  
   \let\subsec=\relax  
   \let\subsubsec=\relax  
   \let\subsection=\relax  
   \let\subsubsection=\relax  
   \let\appendix=\relax  
   \let\appendixsec=\relax  
   \let\appendixsection=\relax  
   \let\appendixsubsec=\relax  
   \let\appendixsubsection=\relax  
   \let\appendixsubsubsec=\relax  
   \let\appendixsubsubsection=\relax  
   \let\contents=\relax  
   \let\smallbook=\relax  
   \let\titlepage=\relax  
 }  
   
 % Used in nested conditionals, where we have to parse the Texinfo source  % @iftex, @ifnotdocbook, @ifnothtml, @ifnotinfo, @ifnotplaintext,
 % and so want to turn off most commands, in case they are used  % @ifnotxml always succeed.  They currently do nothing; we don't
 % incorrectly.  % attempt to check whether the conditionals are properly nested.  But we
   % have to remember that they are conditionals, so that @end doesn't
   % attempt to close an environment group.
 %  %
 \def\ignoremorecommands{%  \def\makecond#1{%
   \let\defcodeindex = \relax    \expandafter\let\csname #1\endcsname = \relax
   \let\defcv = \relax    \expandafter\let\csname iscond.#1\endcsname = 1
   \let\deffn = \relax  
   \let\deffnx = \relax  
   \let\defindex = \relax  
   \let\defivar = \relax  
   \let\defmac = \relax  
   \let\defmethod = \relax  
   \let\defop = \relax  
   \let\defopt = \relax  
   \let\defspec = \relax  
   \let\deftp = \relax  
   \let\deftypefn = \relax  
   \let\deftypefun = \relax  
   \let\deftypeivar = \relax  
   \let\deftypeop = \relax  
   \let\deftypevar = \relax  
   \let\deftypevr = \relax  
   \let\defun = \relax  
   \let\defvar = \relax  
   \let\defvr = \relax  
   \let\ref = \relax  
   \let\xref = \relax  
   \let\printindex = \relax  
   \let\pxref = \relax  
   \let\settitle = \relax  
   \let\setchapternewpage = \relax  
   \let\setchapterstyle = \relax  
   \let\everyheading = \relax  
   \let\evenheading = \relax  
   \let\oddheading = \relax  
   \let\everyfooting = \relax  
   \let\evenfooting = \relax  
   \let\oddfooting = \relax  
   \let\headings = \relax  
   \let\include = \relax  
   \let\lowersections = \relax  
   \let\down = \relax  
   \let\raisesections = \relax  
   \let\up = \relax  
   \let\set = \relax  
   \let\clear = \relax  
   \let\item = \relax  
 }  }
   \makecond{iftex}
   \makecond{ifnotdocbook}
   \makecond{ifnothtml}
   \makecond{ifnotinfo}
   \makecond{ifnotplaintext}
   \makecond{ifnotxml}
   
 % Ignore @ignore ... @end ignore.  % Ignore @ignore, @ifhtml, @ifinfo, and the like.
 %  %
 \def\ignore{\doignore{ignore}}  \def\direntry{\doignore{direntry}}
   \def\documentdescription{\doignore{documentdescription}}
 % Ignore @ifinfo, @ifhtml, @ifnottex, @html, @menu, and @direntry text.  \def\docbook{\doignore{docbook}}
 %  \def\html{\doignore{html}}
 \def\ifinfo{\doignore{ifinfo}}  \def\ifdocbook{\doignore{ifdocbook}}
 \def\ifhtml{\doignore{ifhtml}}  \def\ifhtml{\doignore{ifhtml}}
   \def\ifinfo{\doignore{ifinfo}}
 \def\ifnottex{\doignore{ifnottex}}  \def\ifnottex{\doignore{ifnottex}}
 \def\html{\doignore{html}}  \def\ifplaintext{\doignore{ifplaintext}}
   \def\ifxml{\doignore{ifxml}}
   \def\ignore{\doignore{ignore}}
 \def\menu{\doignore{menu}}  \def\menu{\doignore{menu}}
 \def\direntry{\doignore{direntry}}  \def\xml{\doignore{xml}}
   
 % @dircategory CATEGORY  -- specify a category of the dir file  % Ignore text until a line `@end #1', keeping track of nested conditionals.
 % which this file should belong to.  Ignore this in TeX.  
 \let\dircategory = \comment  
   
 % Ignore text until a line `@end #1'.  
 %  %
   % A count to remember the depth of nesting.
   \newcount\doignorecount
   
 \def\doignore#1{\begingroup  \def\doignore#1{\begingroup
   % Don't complain about control sequences we have declared \outer.    % Scan in ``verbatim'' mode:
   \ignoresections    \obeylines
     \catcode`\@ = \other
     \catcode`\{ = \other
     \catcode`\} = \other
   %    %
   % Define a command to swallow text until we reach `@end #1'.  
   % This @ is a catcode 12 token (that is the normal catcode of @ in  
   % this texinfo.tex file).  We change the catcode of @ below to match.  
   \long\def\doignoretext##1@end #1{\enddoignore}%  
   %  
   % Make sure that spaces turn into tokens that match what \doignoretext wants.    % Make sure that spaces turn into tokens that match what \doignoretext wants.
   \catcode32 = 10    \spaceisspace
   %    %
   % Ignore braces, too, so mismatched braces don't cause trouble.    % Count number of #1's that we've seen.
   \catcode`\{ = 9    \doignorecount = 0
   \catcode`\} = 9  
   %    %
   % We must not have @c interpreted as a control sequence.    % Swallow text until we reach the matching `@end #1'.
   \catcode`\@ = 12    \dodoignore{#1}%
   %  
   % Make the letter c a comment character so that the rest of the line  
   % will be ignored. This way, the document can have (for example)  
   %   @c @end ifinfo  
   % and the @end ifinfo will be properly ignored.  
   % (We've just changed @ to catcode 12.)  
   \catcode`\c = 14  
   %  
   % And now expand that command.  
   \doignoretext  
 }  }
   
 % What we do to finish off ignored text.  { \catcode`_=11 % We want to use \_STOP_ which cannot appear in texinfo source.
 %    \obeylines %
 \def\enddoignore{\endgroup\ignorespaces}%  
   
 \newif\ifwarnedobs\warnedobsfalse  
 \def\obstexwarn{%  
   \ifwarnedobs\relax\else  
   % We need to warn folks that they may have trouble with TeX 3.0.  
   % This uses \immediate\write16 rather than \message to get newlines.  
     \immediate\write16{}  
     \immediate\write16{WARNING: for users of Unix TeX 3.0!}  
     \immediate\write16{This manual trips a bug in TeX version 3.0 (tex hangs).}  
     \immediate\write16{If you are running another version of TeX, relax.}  
     \immediate\write16{If you are running Unix TeX 3.0, kill this TeX process.}  
     \immediate\write16{  Then upgrade your TeX installation if you can.}  
     \immediate\write16{  (See ftp://ftp.gnu.org/pub/gnu/TeX.README.)}  
     \immediate\write16{If you are stuck with version 3.0, run the}  
     \immediate\write16{  script ``tex3patch'' from the Texinfo distribution}  
     \immediate\write16{  to use a workaround.}  
     \immediate\write16{}  
     \global\warnedobstrue  
     \fi  
 }  
   
 % **In TeX 3.0, setting text in \nullfont hangs tex.  For a  
 % workaround (which requires the file ``dummy.tfm'' to be installed),  
 % uncomment the following line:  
 %%%%%\font\nullfont=dummy\let\obstexwarn=\relax  
   
 % Ignore text, except that we keep track of conditional commands for  
 % purposes of nesting, up to an `@end #1' command.  
 %  
 \def\nestedignore#1{%  
   \obstexwarn  
   % We must actually expand the ignored text to look for the @end  
   % command, so that nested ignore constructs work.  Thus, we put the  
   % text into a \vbox and then do nothing with the result.  To minimize  
   % the change of memory overflow, we follow the approach outlined on  
   % page 401 of the TeXbook: make the current font be a dummy font.  
   %    %
   \setbox0 = \vbox\bgroup    \gdef\dodoignore#1{%
     % Don't complain about control sequences we have declared \outer.      % #1 contains the command name as a string, e.g., `ifinfo'.
     \ignoresections  
     %      %
     % Define `@end #1' to end the box, which will in turn undefine the      % Define a command to find the next `@end #1'.
     % @end command again.      \long\def\doignoretext##1^^M@end #1{%
     \expandafter\def\csname E#1\endcsname{\egroup\ignorespaces}%        \doignoretextyyy##1^^M@#1\_STOP_}%
     %      %
     % We are going to be parsing Texinfo commands.  Most cause no      % And this command to find another #1 command, at the beginning of a
     % trouble when they are used incorrectly, but some commands do      % line.  (Otherwise, we would consider a line `@c @ifset', for
     % complicated argument parsing or otherwise get confused, so we      % example, to count as an @ifset for nesting.)
     % undefine them.      \long\def\doignoretextyyy##1^^M@#1##2\_STOP_{\doignoreyyy{##2}\_STOP_}%
     %      %
     % We can't do anything about stray @-signs, unfortunately;      % And now expand that command.
     % they'll produce `undefined control sequence' errors.      \doignoretext ^^M%
     \ignoremorecommands    }%
     %  
     % Set the current font to be \nullfont, a TeX primitive, and define  
     % all the font commands to also use \nullfont.  We don't use  
     % dummy.tfm, as suggested in the TeXbook, because not all sites  
     % might have that installed.  Therefore, math mode will still  
     % produce output, but that should be an extremely small amount of  
     % stuff compared to the main input.  
     %  
     \nullfont  
     \let\tenrm=\nullfont \let\tenit=\nullfont \let\tensl=\nullfont  
     \let\tenbf=\nullfont \let\tentt=\nullfont \let\smallcaps=\nullfont  
     \let\tensf=\nullfont  
     % Similarly for index fonts (mostly for their use in smallexample).  
     \let\smallrm=\nullfont \let\smallit=\nullfont \let\smallsl=\nullfont  
     \let\smallbf=\nullfont \let\smalltt=\nullfont \let\smallsc=\nullfont  
     \let\smallsf=\nullfont  
     %  
     % Don't complain when characters are missing from the fonts.  
     \tracinglostchars = 0  
     %  
     % Don't bother to do space factor calculations.  
     \frenchspacing  
     %  
     % Don't report underfull hboxes.  
     \hbadness = 10000  
     %  
     % Do minimal line-breaking.  
     \pretolerance = 10000  
     %  
     % Do not execute instructions in @tex  
     \def\tex{\doignore{tex}}%  
     % Do not execute macro definitions.  
     % `c' is a comment character, so the word `macro' will get cut off.  
     \def\macro{\doignore{ma}}%  
 }  }
   
   \def\doignoreyyy#1{%
     \def\temp{#1}%
     \ifx\temp\empty                       % Nothing found.
       \let\next\doignoretextzzz
     \else                                 % Found a nested condition, ...
       \advance\doignorecount by 1
       \let\next\doignoretextyyy           % ..., look for another.
       % If we're here, #1 ends with ^^M\ifinfo (for example).
     \fi
     \next #1% the token \_STOP_ is present just after this macro.
   }
   
   % We have to swallow the remaining "\_STOP_".
   %
   \def\doignoretextzzz#1{%
     \ifnum\doignorecount = 0      % We have just found the outermost @end.
       \let\next\enddoignore
     \else                         % Still inside a nested condition.
       \advance\doignorecount by -1
       \let\next\doignoretext      % Look for the next @end.
     \fi
     \next
   }
   
   % Finish off ignored text.
   { \obeylines%
     % Ignore anything after the last `@end #1'; this matters in verbatim
     % environments, where otherwise the newline after an ignored conditional
     % would result in a blank line in the output.
     \gdef\enddoignore#1^^M{\endgroup\ignorespaces}%
   }
   
   
 % @set VAR sets the variable VAR to an empty value.  % @set VAR sets the variable VAR to an empty value.
 % @set VAR REST-OF-LINE sets VAR to the value REST-OF-LINE.  % @set VAR REST-OF-LINE sets VAR to the value REST-OF-LINE.
 %  %
 % Since we want to separate VAR from REST-OF-LINE (which might be  % Since we want to separate VAR from REST-OF-LINE (which might be
 % empty), we can't just use \parsearg; we have to insert a space of our  % empty), we can't just use \parsearg; we have to insert a space of our
 % own to delimit the rest of the line, and then take it out again if we  % own to delimit the rest of the line, and then take it out again if we
 % didn't need it.  Make sure the catcode of space is correct to avoid  % didn't need it.
 % losing inside @example, for instance.  % We rely on the fact that \parsearg sets \catcode`\ =10.
 %  %
 \def\set{\begingroup\catcode` =10  \parseargdef\set{\setyyy#1 \endsetyyy}
   \catcode`\-=12 \catcode`\_=12 % Allow - and _ in VAR.  
   \parsearg\setxxx}  
 \def\setxxx#1{\setyyy#1 \endsetyyy}  
 \def\setyyy#1 #2\endsetyyy{%  \def\setyyy#1 #2\endsetyyy{%
   \def\temp{#2}%    {%
   \ifx\temp\empty \global\expandafter\let\csname SET#1\endcsname = \empty      \makevalueexpandable
   \else \setzzz{#1}#2\endsetzzz % Remove the trailing space \setxxx inserted.      \def\temp{#2}%
   \fi      \edef\next{\gdef\makecsname{SET#1}}%
   \endgroup      \ifx\temp\empty
         \next{}%
       \else
         \setzzz#2\endsetzzz
       \fi
     }%
 }  }
 % Can't use \xdef to pre-expand #2 and save some time, since \temp or  % Remove the trailing space \setxxx inserted.
 % \next or other control sequences that we've defined might get us into  \def\setzzz#1 \endsetzzz{\next{#1}}
 % an infinite loop. Consider `@set foo @cite{bar}'.  
 \def\setzzz#1#2 \endsetzzz{\expandafter\gdef\csname SET#1\endcsname{#2}}  
   
 % @clear VAR clears (i.e., unsets) the variable VAR.  % @clear VAR clears (i.e., unsets) the variable VAR.
 %  %
 \def\clear{\parsearg\clearxxx}  \parseargdef\clear{%
 \def\clearxxx#1{\global\expandafter\let\csname SET#1\endcsname=\relax}    {%
       \makevalueexpandable
       \global\expandafter\let\csname SET#1\endcsname=\relax
     }%
   }
   
 % @value{foo} gets the text saved in variable foo.  % @value{foo} gets the text saved in variable foo.
   \def\value{\begingroup\makevalueexpandable\valuexxx}
   \def\valuexxx#1{\expandablevalue{#1}\endgroup}
 {  {
   \catcode`\_ = \active    \catcode`\-=\active \catcode`\_=\active
   %    %
   % We might end up with active _ or - characters in the argument if    \gdef\makevalueexpandable{%
   % we're called from @code, as @code{@value{foo-bar_}}.  So \let any      \let\value = \expandablevalue
   % such active characters to their normal equivalents.      % We don't want these characters active, ...
   \gdef\value{\begingroup      \catcode`\-=\other \catcode`\_=\other
     \catcode`\-=12 \catcode`\_=12      % ..., but we might end up with active ones in the argument if
     \indexbreaks \let_\normalunderscore      % we're called from @code, as @code{@value{foo-bar_}}, though.
     \valuexxx}      % So \let them to their normal equivalents.
       \let-\normaldash \let_\normalunderscore
     }
 }  }
 \def\valuexxx#1{\expandablevalue{#1}\endgroup}  
   
 % We have this subroutine so that we can handle at least some @value's  % We have this subroutine so that we can handle at least some @value's
 % properly in indexes (we \let\value to this in \indexdummies).  Ones  % properly in indexes (we call \makevalueexpandable in \indexdummies).
 % whose names contain - or _ still won't work, but we can't do anything  % The command has to be fully expandable (if the variable is set), since
 % about that.  The command has to be fully expandable, since the result  % the result winds up in the index file.  This means that if the
 % winds up in the index file.  This means that if the variable's value  % variable's value contains other Texinfo commands, it's almost certain
 % contains other Texinfo commands, it's almost certain it will fail  % it will fail (although perhaps we could fix that with sufficient work
 % (although perhaps we could fix that with sufficient work to do a  % to do a one-level expansion on the result, instead of complete).
 % one-level expansion on the result, instead of complete).  %
 %  % Unfortunately, this has the consequence that when _ is in the *value*
   % of an @set, it does not print properly in the roman fonts (get the cmr
   % dot accent at position 126 instead).  No fix comes to mind, and it's
   % been this way since 2003 or earlier, so just ignore it.
   %
 \def\expandablevalue#1{%  \def\expandablevalue#1{%
   \expandafter\ifx\csname SET#1\endcsname\relax    \expandafter\ifx\csname SET#1\endcsname\relax
     {[No value for ``#1'']}%      {[No value for ``#1'']}%
       \message{Variable `#1', used in @value, is not set.}%
   \else    \else
     \csname SET#1\endcsname      \csname SET#1\endcsname
   \fi    \fi
 }  }
   
 % @ifset VAR ... @end ifset reads the `...' iff VAR has been defined  % Like \expandablevalue, but completely expandable (the \message in the
 % with @set.  % definition above operates at the execution level of TeX).  Used when
   % writing to auxiliary files, due to the expansion that \write does.
   % If flag is undefined, pass through an unexpanded @value command: maybe it
   % will be set by the time it is read back in.
 %  %
 \def\ifset{\parsearg\ifsetxxx}  % NB flag names containing - or _ may not work here.
 \def\ifsetxxx #1{%  \def\dummyvalue#1{%
   \expandafter\ifx\csname SET#1\endcsname\relax    \expandafter\ifx\csname SET#1\endcsname\relax
     \expandafter\ifsetfail      \noexpand\value{#1}%
   \else    \else
     \expandafter\ifsetsucceed      \csname SET#1\endcsname
   \fi    \fi
 }  }
 \def\ifsetsucceed{\conditionalsucceed{ifset}}  
 \def\ifsetfail{\nestedignore{ifset}}  
 \defineunmatchedend{ifset}  
   
 % @ifclear VAR ... @end ifclear reads the `...' iff VAR has never been  % Used for @value's in index entries to form the sort key: expand the @value
 % defined with @set, or has been undefined with @clear.  % if possible, otherwise sort late.
 %  \def\indexnofontsvalue#1{%
 \def\ifclear{\parsearg\ifclearxxx}  
 \def\ifclearxxx #1{%  
   \expandafter\ifx\csname SET#1\endcsname\relax    \expandafter\ifx\csname SET#1\endcsname\relax
     \expandafter\ifclearsucceed      ZZZZZZZ
   \else    \else
     \expandafter\ifclearfail      \csname SET#1\endcsname
   \fi    \fi
 }  }
 \def\ifclearsucceed{\conditionalsucceed{ifclear}}  
 \def\ifclearfail{\nestedignore{ifclear}}  
 \defineunmatchedend{ifclear}  
   
 % @iftex, @ifnothtml, @ifnotinfo always succeed; we read the text  % @ifset VAR ... @end ifset reads the `...' iff VAR has been defined
 % following, through the first @end iftex (etc.).  Make `@end iftex'  % with @set.
 % (etc.) valid only after an @iftex.  %
   % To get the special treatment we need for `@end ifset,' we call
   % \makecond and then redefine.
 %  %
 \def\iftex{\conditionalsucceed{iftex}}  \makecond{ifset}
 \def\ifnothtml{\conditionalsucceed{ifnothtml}}  \def\ifset{\parsearg{\doifset{\let\next=\ifsetfail}}}
 \def\ifnotinfo{\conditionalsucceed{ifnotinfo}}  \def\doifset#1#2{%
 \defineunmatchedend{iftex}    {%
 \defineunmatchedend{ifnothtml}      \makevalueexpandable
 \defineunmatchedend{ifnotinfo}      \let\next=\empty
       \expandafter\ifx\csname SET#2\endcsname\relax
         #1% If not set, redefine \next.
       \fi
       \expandafter
     }\next
   }
   \def\ifsetfail{\doignore{ifset}}
   
 % We can't just want to start a group at @iftex (for example) and end it  % @ifclear VAR ... @end executes the `...' iff VAR has never been
 % at @end iftex, since then @set commands inside the conditional have no  % defined with @set, or has been undefined with @clear.
 % effect (they'd get reverted at the end of the group).  So we must  
 % define \Eiftex to redefine itself to be its previous value.  (We can't  
 % just define it to fail again with an ``unmatched end'' error, since  
 % the @ifset might be nested.)  
 %  %
 \def\conditionalsucceed#1{%  % The `\else' inside the `\doifset' parameter is a trick to reuse the
   \edef\temp{%  % above code: if the variable is not set, do nothing, if it is set,
     % Remember the current value of \E#1.  % then redefine \next to \ifclearfail.
     \let\nece{prevE#1} = \nece{E#1}%  %
     %  \makecond{ifclear}
     % At the `@end #1', redefine \E#1 to be its previous value.  \def\ifclear{\parsearg{\doifset{\else \let\next=\ifclearfail}}}
     \def\nece{E#1}{\let\nece{E#1} = \nece{prevE#1}}%  \def\ifclearfail{\doignore{ifclear}}
   }%  
   \temp  
 }  
   
 % We need to expand lots of \csname's, but we don't want to expand the  % @ifcommandisdefined CMD ... @end executes the `...' if CMD (written
 % control sequences after we've constructed them.  % without the @) is in fact defined.  We can only feasibly check at the
   % TeX level, so something like `mathcode' is going to considered
   % defined even though it is not a Texinfo command.
   %
   \makecond{ifcommanddefined}
   \def\ifcommanddefined{\parsearg{\doifcmddefined{\let\next=\ifcmddefinedfail}}}
 %  %
 \def\nece#1{\expandafter\noexpand\csname#1\endcsname}  \def\doifcmddefined#1#2{{%
       \makevalueexpandable
       \let\next=\empty
       \expandafter\ifx\csname #2\endcsname\relax
         #1% If not defined, \let\next as above.
       \fi
       \expandafter
     }\next
   }
   \def\ifcmddefinedfail{\doignore{ifcommanddefined}}
   
   % @ifcommandnotdefined CMD ... handled similar to @ifclear above.
   \makecond{ifcommandnotdefined}
   \def\ifcommandnotdefined{%
     \parsearg{\doifcmddefined{\else \let\next=\ifcmdnotdefinedfail}}}
   \def\ifcmdnotdefinedfail{\doignore{ifcommandnotdefined}}
   
   % Set the `txicommandconditionals' variable, so documents have a way to
   % test if the @ifcommand...defined conditionals are available.
   \set txicommandconditionals
   
   % @dircategory CATEGORY  -- specify a category of the dir file
   % which this file should belong to.  Ignore this in TeX.
   \let\dircategory=\comment
   
 % @defininfoenclose.  % @defininfoenclose.
 \let\definfoenclose=\comment  \let\definfoenclose=\comment
   
Line 2565  width0pt\relax} \fi
Line 4832  width0pt\relax} \fi
 % Index generation facilities  % Index generation facilities
   
 % Define \newwrite to be identical to plain tex's \newwrite  % Define \newwrite to be identical to plain tex's \newwrite
 % except not \outer, so it can be used within \newindex.  % except not \outer, so it can be used within macros and \if's.
 {\catcode`\@=11  \edef\newwrite{\makecsname{ptexnewwrite}}
 \gdef\newwrite{\alloc@7\write\chardef\sixt@@n}}  
   
 % \newindex {foo} defines an index named foo.  % \newindex {foo} defines an index named IX.
 % It automatically defines \fooindex such that  % It automatically defines \IXindex such that
 % \fooindex ...rest of line... puts an entry in the index foo.  % \IXindex ...rest of line... puts an entry in the index IX.
 % It also defines \fooindfile to be the number of the output channel for  % It also defines \IXindfile to be the number of the output channel for
 % the file that accumulates this index.  The file's extension is foo.  % the file that accumulates this index.  The file's extension is IX.
 % The name of an index should be no more than 2 characters long  % The name of an index should be no more than 2 characters long
 % for the sake of vms.  % for the sake of vms.
 %  %
 \def\newindex#1{%  \def\newindex#1{%
   \iflinks    \expandafter\chardef\csname#1indfile\endcsname=0
     \expandafter\newwrite \csname#1indfile\endcsname  
     \openout \csname#1indfile\endcsname \jobname.#1 % Open the file  
   \fi  
   \expandafter\xdef\csname#1index\endcsname{%     % Define @#1index    \expandafter\xdef\csname#1index\endcsname{%     % Define @#1index
     \noexpand\doindex{#1}}      \noexpand\doindex{#1}}
 }  }
   
 % @defindex foo  ==  \newindex{foo}  % @defindex foo  ==  \newindex{foo}
   %
 \def\defindex{\parsearg\newindex}  \def\defindex{\parsearg\newindex}
   
 % Define @defcodeindex, like @defindex except put all entries in @code.  % Define @defcodeindex, like @defindex except put all entries in @code.
   %
   \def\defcodeindex{\parsearg\newcodeindex}
   %
 \def\newcodeindex#1{%  \def\newcodeindex#1{%
   \iflinks    \expandafter\chardef\csname#1indfile\endcsname=0
     \expandafter\newwrite \csname#1indfile\endcsname  
     \openout \csname#1indfile\endcsname \jobname.#1  
   \fi  
   \expandafter\xdef\csname#1index\endcsname{%    \expandafter\xdef\csname#1index\endcsname{%
     \noexpand\docodeindex{#1}}      \noexpand\docodeindex{#1}}%
 }  }
   
 \def\defcodeindex{\parsearg\newcodeindex}  % The default indices:
   \newindex{cp}%      concepts,
   \newcodeindex{fn}%  functions,
   \newcodeindex{vr}%  variables,
   \newcodeindex{tp}%  types,
   \newcodeindex{ky}%  keys
   \newcodeindex{pg}%  and programs.
   
   
 % @synindex foo bar    makes index foo feed into index bar.  % @synindex foo bar    makes index foo feed into index bar.
 % Do this instead of @defindex foo if you don't want it as a separate index.  % Do this instead of @defindex foo if you don't want it as a separate index.
 % The \closeout helps reduce unnecessary open files; the limit on the  %
 % Acorn RISC OS is a mere 16 files.  
 \def\synindex#1 #2 {%  
   \expandafter\let\expandafter\synindexfoo\expandafter=\csname#2indfile\endcsname  
   \expandafter\closeout\csname#1indfile\endcsname  
   \expandafter\let\csname#1indfile\endcsname=\synindexfoo  
   \expandafter\xdef\csname#1index\endcsname{% define \xxxindex  
     \noexpand\doindex{#2}}%  
 }  
   
 % @syncodeindex foo bar   similar, but put all entries made for index foo  % @syncodeindex foo bar   similar, but put all entries made for index foo
 % inside @code.  % inside @code.
 \def\syncodeindex#1 #2 {%  %
   \expandafter\let\expandafter\synindexfoo\expandafter=\csname#2indfile\endcsname  \def\synindex#1 #2 {\dosynindex\doindex{#1}{#2}}
   \expandafter\closeout\csname#1indfile\endcsname  \def\syncodeindex#1 #2 {\dosynindex\docodeindex{#1}{#2}}
   \expandafter\let\csname#1indfile\endcsname=\synindexfoo  
   \expandafter\xdef\csname#1index\endcsname{% define \xxxindex  % #1 is \doindex or \docodeindex, #2 the index getting redefined (foo),
     \noexpand\docodeindex{#2}}%  % #3 the target index (bar).
   \def\dosynindex#1#2#3{%
     \requireopenindexfile{#3}%
     % redefine \fooindfile:
     \expandafter\let\expandafter\temp\expandafter=\csname#3indfile\endcsname
     \expandafter\let\csname#2indfile\endcsname=\temp
     % redefine \fooindex:
     \expandafter\xdef\csname#2index\endcsname{\noexpand#1{#3}}%
 }  }
   
 % Define \doindex, the driver for all \fooindex macros.  % Define \doindex, the driver for all index macros.
 % Argument #1 is generated by the calling \fooindex macro,  % Argument #1 is generated by the calling \fooindex macro,
 %  and it is "foo", the name of the index.  % and it is the two-letter name of the index.
   
 % \doindex just uses \parsearg; it calls \doind for the actual work.  \def\doindex#1{\edef\indexname{#1}\parsearg\doindexxxx}
 % This is because \doind is more useful to call from other macros.  \def\doindexxxx #1{\doind{\indexname}{#1}}
   
 % There is also \dosubind {index}{topic}{subtopic}  
 % which makes an entry in a two-level index such as the operation index.  
   
 \def\doindex#1{\edef\indexname{#1}\parsearg\singleindexer}  
 \def\singleindexer #1{\doind{\indexname}{#1}}  
   
 % like the previous two, but they put @code around the argument.  % like the previous two, but they put @code around the argument.
 \def\docodeindex#1{\edef\indexname{#1}\parsearg\singlecodeindexer}  \def\docodeindex#1{\edef\indexname{#1}\parsearg\docodeindexxxx}
 \def\singlecodeindexer #1{\doind{\indexname}{\code{#1}}}  \def\docodeindexxxx #1{\doind{\indexname}{\code{#1}}}
   
   
   % Used when writing an index entry out to an index file to prevent
   % expansion of Texinfo commands that can appear in an index entry.
   %
 \def\indexdummies{%  \def\indexdummies{%
 \def\ { }%    \escapechar = `\\     % use backslash in output files.
 % Take care of the plain tex accent commands.    \definedummyletter\@%
 \def\"{\realbackslash "}%    \definedummyletter\ %
 \def\`{\realbackslash `}%    %
 \def\'{\realbackslash '}%    % For texindex which always views { and } as separators.
 \def\^{\realbackslash ^}%    \def\{{\lbracechar{}}%
 \def\~{\realbackslash ~}%    \def\}{\rbracechar{}}%
 \def\={\realbackslash =}%    %
 \def\b{\realbackslash b}%    % Do the redefinitions.
 \def\c{\realbackslash c}%    \definedummies
 \def\d{\realbackslash d}%  }
 \def\u{\realbackslash u}%  
 \def\v{\realbackslash v}%  % Used for the aux and toc files, where @ is the escape character.
 \def\H{\realbackslash H}%  
 % Take care of the plain tex special European modified letters.  
 \def\oe{\realbackslash oe}%  
 \def\ae{\realbackslash ae}%  
 \def\aa{\realbackslash aa}%  
 \def\OE{\realbackslash OE}%  
 \def\AE{\realbackslash AE}%  
 \def\AA{\realbackslash AA}%  
 \def\o{\realbackslash o}%  
 \def\O{\realbackslash O}%  
 \def\l{\realbackslash l}%  
 \def\L{\realbackslash L}%  
 \def\ss{\realbackslash ss}%  
 % Take care of texinfo commands likely to appear in an index entry.  
 % (Must be a way to avoid doing expansion at all, and thus not have to  
 % laboriously list every single command here.)  
 \def\@{@}% will be @@ when we switch to @ as escape char.  
 % Need these in case \tex is in effect and \{ is a \delimiter again.  
 % But can't use \lbracecmd and \rbracecmd because texindex assumes  
 % braces and backslashes are used only as delimiters.  
 \let\{ = \mylbrace  
 \let\} = \myrbrace  
 \def\_{{\realbackslash _}}%  
 \def\w{\realbackslash w }%  
 \def\bf{\realbackslash bf }%  
 %\def\rm{\realbackslash rm }%  
 \def\sl{\realbackslash sl }%  
 \def\sf{\realbackslash sf}%  
 \def\tt{\realbackslash tt}%  
 \def\gtr{\realbackslash gtr}%  
 \def\less{\realbackslash less}%  
 \def\hat{\realbackslash hat}%  
 \def\TeX{\realbackslash TeX}%  
 \def\dots{\realbackslash dots }%  
 \def\result{\realbackslash result}%  
 \def\equiv{\realbackslash equiv}%  
 \def\expansion{\realbackslash expansion}%  
 \def\print{\realbackslash print}%  
 \def\error{\realbackslash error}%  
 \def\point{\realbackslash point}%  
 \def\copyright{\realbackslash copyright}%  
 \def\tclose##1{\realbackslash tclose {##1}}%  
 \def\code##1{\realbackslash code {##1}}%  
 \def\uref##1{\realbackslash uref {##1}}%  
 \def\url##1{\realbackslash url {##1}}%  
 \def\env##1{\realbackslash env {##1}}%  
 \def\command##1{\realbackslash command {##1}}%  
 \def\option##1{\realbackslash option {##1}}%  
 \def\dotless##1{\realbackslash dotless {##1}}%  
 \def\samp##1{\realbackslash samp {##1}}%  
 \def\,##1{\realbackslash ,{##1}}%  
 \def\t##1{\realbackslash t {##1}}%  
 \def\r##1{\realbackslash r {##1}}%  
 \def\i##1{\realbackslash i {##1}}%  
 \def\b##1{\realbackslash b {##1}}%  
 \def\sc##1{\realbackslash sc {##1}}%  
 \def\cite##1{\realbackslash cite {##1}}%  
 \def\key##1{\realbackslash key {##1}}%  
 \def\file##1{\realbackslash file {##1}}%  
 \def\var##1{\realbackslash var {##1}}%  
 \def\kbd##1{\realbackslash kbd {##1}}%  
 \def\dfn##1{\realbackslash dfn {##1}}%  
 \def\emph##1{\realbackslash emph {##1}}%  
 \def\acronym##1{\realbackslash acronym {##1}}%  
 %  %
 % Handle some cases of @value -- where the variable name does not  \def\atdummies{%
 % contain - or _, and the value does not contain any    \definedummyletter\@%
 % (non-fully-expandable) commands.    \definedummyletter\ %
 \let\value = \expandablevalue    \definedummyletter\{%
     \definedummyletter\}%
     %
     % Do the redefinitions.
     \definedummies
     \otherbackslash
   }
   
   % \definedummyword defines \#1 as \string\#1\space, thus effectively
   % preventing its expansion.  This is used only for control words,
   % not control letters, because the \space would be incorrect for
   % control characters, but is needed to separate the control word
   % from whatever follows.
 %  %
 \unsepspaces  % These can be used both for control words that take an argument and
 % Turn off macro expansion  % those that do not.  If it is followed by {arg} in the input, then
 \turnoffmacros  % that will dutifully get written to the index (or wherever).
   %
   % For control letters, we have \definedummyletter, which omits the
   % space.
   %
   \def\definedummyword  #1{\def#1{\string#1\space}}%
   \def\definedummyletter#1{\def#1{\string#1}}%
   \let\definedummyaccent\definedummyletter
   
   % Called from \indexdummies and \atdummies, to effectively prevent
   % the expansion of commands.
   %
   \def\definedummies{%
     %
     \let\commondummyword\definedummyword
     \let\commondummyletter\definedummyletter
     \let\commondummyaccent\definedummyaccent
     \commondummiesnofonts
     %
     \definedummyletter\_%
     \definedummyletter\-%
     %
     % Non-English letters.
     \definedummyword\AA
     \definedummyword\AE
     \definedummyword\DH
     \definedummyword\L
     \definedummyword\O
     \definedummyword\OE
     \definedummyword\TH
     \definedummyword\aa
     \definedummyword\ae
     \definedummyword\dh
     \definedummyword\exclamdown
     \definedummyword\l
     \definedummyword\o
     \definedummyword\oe
     \definedummyword\ordf
     \definedummyword\ordm
     \definedummyword\questiondown
     \definedummyword\ss
     \definedummyword\th
     %
     % Although these internal commands shouldn't show up, sometimes they do.
     \definedummyword\bf
     \definedummyword\gtr
     \definedummyword\hat
     \definedummyword\less
     \definedummyword\sf
     \definedummyword\sl
     \definedummyword\tclose
     \definedummyword\tt
     %
     \definedummyword\LaTeX
     \definedummyword\TeX
     %
     % Assorted special characters.
     \definedummyword\atchar
     \definedummyword\arrow
     \definedummyword\bullet
     \definedummyword\comma
     \definedummyword\copyright
     \definedummyword\registeredsymbol
     \definedummyword\dots
     \definedummyword\enddots
     \definedummyword\entrybreak
     \definedummyword\equiv
     \definedummyword\error
     \definedummyword\euro
     \definedummyword\expansion
     \definedummyword\geq
     \definedummyword\guillemetleft
     \definedummyword\guillemetright
     \definedummyword\guilsinglleft
     \definedummyword\guilsinglright
     \definedummyword\lbracechar
     \definedummyword\leq
     \definedummyword\mathopsup
     \definedummyword\minus
     \definedummyword\ogonek
     \definedummyword\pounds
     \definedummyword\point
     \definedummyword\print
     \definedummyword\quotedblbase
     \definedummyword\quotedblleft
     \definedummyword\quotedblright
     \definedummyword\quoteleft
     \definedummyword\quoteright
     \definedummyword\quotesinglbase
     \definedummyword\rbracechar
     \definedummyword\result
     \definedummyword\sub
     \definedummyword\sup
     \definedummyword\textdegree
     %
     % We want to disable all macros so that they are not expanded by \write.
     \macrolist
     \let\value\dummyvalue
     %
     \normalturnoffactive
 }  }
   
 % If an index command is used in an @example environment, any spaces  % \commondummiesnofonts: common to \definedummies and \indexnofonts.
 % therein should become regular spaces in the raw index file, not the  % Define \commondummyletter, \commondummyaccent and \commondummyword before
 % expansion of \tie (\\leavevmode \penalty \@M \ ).  % using.  Used for accents, font commands, and various control letters.
 {\obeyspaces  %
  \gdef\unsepspaces{\obeyspaces\let =\space}}  \def\commondummiesnofonts{%
     % Control letters and accents.
     \commondummyletter\!%
     \commondummyaccent\"%
     \commondummyaccent\'%
     \commondummyletter\*%
     \commondummyaccent\,%
     \commondummyletter\.%
     \commondummyletter\/%
     \commondummyletter\:%
     \commondummyaccent\=%
     \commondummyletter\?%
     \commondummyaccent\^%
     \commondummyaccent\`%
     \commondummyaccent\~%
     \commondummyword\u
     \commondummyword\v
     \commondummyword\H
     \commondummyword\dotaccent
     \commondummyword\ogonek
     \commondummyword\ringaccent
     \commondummyword\tieaccent
     \commondummyword\ubaraccent
     \commondummyword\udotaccent
     \commondummyword\dotless
     %
     % Texinfo font commands.
     \commondummyword\b
     \commondummyword\i
     \commondummyword\r
     \commondummyword\sansserif
     \commondummyword\sc
     \commondummyword\slanted
     \commondummyword\t
     %
     % Commands that take arguments.
     \commondummyword\abbr
     \commondummyword\acronym
     \commondummyword\anchor
     \commondummyword\cite
     \commondummyword\code
     \commondummyword\command
     \commondummyword\dfn
     \commondummyword\dmn
     \commondummyword\email
     \commondummyword\emph
     \commondummyword\env
     \commondummyword\file
     \commondummyword\image
     \commondummyword\indicateurl
     \commondummyword\inforef
     \commondummyword\kbd
     \commondummyword\key
     \commondummyword\math
     \commondummyword\option
     \commondummyword\pxref
     \commondummyword\ref
     \commondummyword\samp
     \commondummyword\strong
     \commondummyword\tie
     \commondummyword\U
     \commondummyword\uref
     \commondummyword\url
     \commondummyword\var
     \commondummyword\verb
     \commondummyword\w
     \commondummyword\xref
   }
   
 % \indexnofonts no-ops all font-change commands.  % For testing: output @{ and @} in index sort strings as \{ and \}.
 % This is used when outputting the strings to sort the index by.  \newif\ifusebracesinindexes
 \def\indexdummyfont#1{#1}  
 \def\indexdummytex{TeX}  
 \def\indexdummydots{...}  
   
   \let\indexlbrace\relax
   \let\indexrbrace\relax
   
   {\catcode`\@=0
   \catcode`\\=13
     @gdef@backslashdisappear{@def\{}}
   }
   
   {
   \catcode`\<=13
   \catcode`\-=13
   \catcode`\`=13
     \gdef\indexnonalnumdisappear{%
       \expandafter\ifx\csname SETtxiindexlquoteignore\endcsname\relax\else
         % @set txiindexlquoteignore makes us ignore left quotes in the sort term.
         % (Introduced for FSFS 2nd ed.)
         \let`=\empty
       \fi
       %
       \expandafter\ifx\csname SETtxiindexbackslashignore\endcsname\relax\else
         \backslashdisappear
       \fi
       %
       \expandafter\ifx\csname SETtxiindexhyphenignore\endcsname\relax\else
         \def-{}%
       \fi
       \expandafter\ifx\csname SETtxiindexlessthanignore\endcsname\relax\else
         \def<{}%
       \fi
       \expandafter\ifx\csname SETtxiindexatsignignore\endcsname\relax\else
         \def\@{}%
       \fi
     }
   
     \gdef\indexnonalnumreappear{%
       \useindexbackslash
       \let-\normaldash
       \let<\normalless
       \def\@{@}%
     }
   }
   
   
   % \indexnofonts is used when outputting the strings to sort the index
   % by, and when constructing control sequence names.  It eliminates all
   % control sequences and just writes whatever the best ASCII sort string
   % would be for a given command (usually its argument).
   %
 \def\indexnofonts{%  \def\indexnofonts{%
 % Just ignore accents.    % Accent commands should become @asis.
 \let\,=\indexdummyfont    \def\commondummyaccent##1{\let##1\asis}%
 \let\"=\indexdummyfont    % We can just ignore other control letters.
 \let\`=\indexdummyfont    \def\commondummyletter##1{\let##1\empty}%
 \let\'=\indexdummyfont    % All control words become @asis by default; overrides below.
 \let\^=\indexdummyfont    \let\commondummyword\commondummyaccent
 \let\~=\indexdummyfont    \commondummiesnofonts
 \let\==\indexdummyfont    %
 \let\b=\indexdummyfont    % Don't no-op \tt, since it isn't a user-level command
 \let\c=\indexdummyfont    % and is used in the definitions of the active chars like <, >, |, etc.
 \let\d=\indexdummyfont    % Likewise with the other plain tex font commands.
 \let\u=\indexdummyfont    %\let\tt=\asis
 \let\v=\indexdummyfont    %
 \let\H=\indexdummyfont    \def\ { }%
 \let\dotless=\indexdummyfont    \def\@{@}%
 % Take care of the plain tex special European modified letters.    \def\_{\normalunderscore}%
 \def\oe{oe}%    \def\-{}% @- shouldn't affect sorting
 \def\ae{ae}%    %
 \def\aa{aa}%    \uccode`\1=`\{ \uppercase{\def\{{1}}%
 \def\OE{OE}%    \uccode`\1=`\} \uppercase{\def\}{1}}%
 \def\AE{AE}%    \let\lbracechar\{%
 \def\AA{AA}%    \let\rbracechar\}%
 \def\o{o}%    %
 \def\O{O}%    % Non-English letters.
 \def\l{l}%    \def\AA{AA}%
 \def\L{L}%    \def\AE{AE}%
 \def\ss{ss}%    \def\DH{DZZ}%
 \let\w=\indexdummyfont    \def\L{L}%
 \let\t=\indexdummyfont    \def\OE{OE}%
 \let\r=\indexdummyfont    \def\O{O}%
 \let\i=\indexdummyfont    \def\TH{TH}%
 \let\b=\indexdummyfont    \def\aa{aa}%
 \let\emph=\indexdummyfont    \def\ae{ae}%
 \let\strong=\indexdummyfont    \def\dh{dzz}%
 \let\cite=\indexdummyfont    \def\exclamdown{!}%
 \let\sc=\indexdummyfont    \def\l{l}%
 %Don't no-op \tt, since it isn't a user-level command    \def\oe{oe}%
 % and is used in the definitions of the active chars like <, >, |...    \def\ordf{a}%
 %\let\tt=\indexdummyfont    \def\ordm{o}%
 \let\tclose=\indexdummyfont    \def\o{o}%
 \let\code=\indexdummyfont    \def\questiondown{?}%
 \let\url=\indexdummyfont    \def\ss{ss}%
 \let\uref=\indexdummyfont    \def\th{th}%
 \let\env=\indexdummyfont    %
 \let\acronym=\indexdummyfont    \def\LaTeX{LaTeX}%
 \let\command=\indexdummyfont    \def\TeX{TeX}%
 \let\option=\indexdummyfont    %
 \let\file=\indexdummyfont    % Assorted special characters.  \defglyph gives the control sequence a
 \let\samp=\indexdummyfont    % definition that removes the {} that follows its use.
 \let\kbd=\indexdummyfont    \defglyph\atchar{@}%
 \let\key=\indexdummyfont    \defglyph\arrow{->}%
 \let\var=\indexdummyfont    \defglyph\bullet{bullet}%
 \let\TeX=\indexdummytex    \defglyph\comma{,}%
 \let\dots=\indexdummydots    \defglyph\copyright{copyright}%
 \def\@{@}%    \defglyph\dots{...}%
     \defglyph\enddots{...}%
     \defglyph\equiv{==}%
     \defglyph\error{error}%
     \defglyph\euro{euro}%
     \defglyph\expansion{==>}%
     \defglyph\geq{>=}%
     \defglyph\guillemetleft{<<}%
     \defglyph\guillemetright{>>}%
     \defglyph\guilsinglleft{<}%
     \defglyph\guilsinglright{>}%
     \defglyph\leq{<=}%
     \defglyph\lbracechar{\{}%
     \defglyph\minus{-}%
     \defglyph\point{.}%
     \defglyph\pounds{pounds}%
     \defglyph\print{-|}%
     \defglyph\quotedblbase{"}%
     \defglyph\quotedblleft{"}%
     \defglyph\quotedblright{"}%
     \defglyph\quoteleft{`}%
     \defglyph\quoteright{'}%
     \defglyph\quotesinglbase{,}%
     \defglyph\rbracechar{\}}%
     \defglyph\registeredsymbol{R}%
     \defglyph\result{=>}%
     \defglyph\textdegree{o}%
     %
     % We need to get rid of all macros, leaving only the arguments (if present).
     % Of course this is not nearly correct, but it is the best we can do for now.
     % makeinfo does not expand macros in the argument to @deffn, which ends up
     % writing an index entry, and texindex isn't prepared for an index sort entry
     % that starts with \.
     %
     % Since macro invocations are followed by braces, we can just redefine them
     % to take a single TeX argument.  The case of a macro invocation that
     % goes to end-of-line is not handled.
     %
     \macrolist
     \let\value\indexnofontsvalue
 }  }
   \def\defglyph#1#2{\def#1##1{#2}} % see above
   
 % To define \realbackslash, we must make \ not be an escape.  
 % We must first make another character (@) an escape  
 % so we do not become unable to do a definition.  
   
 {\catcode`\@=0 \catcode`\\=\other  
  @gdef@realbackslash{\}}  
   
 \let\indexbackslash=0  %overridden during \printindex.  
 \let\SETmarginindex=\relax % put index entries in margin (undocumented)?  \let\SETmarginindex=\relax % put index entries in margin (undocumented)?
   
 % For \ifx comparisons.  
 \def\emptymacro{\empty}  
   
 % Most index entries go through here, but \dosubind is the general case.  % Most index entries go through here, but \dosubind is the general case.
 %  % #1 is the index name, #2 is the entry text.
 \def\doind#1#2{\dosubind{#1}{#2}\empty}  \def\doind#1#2{\dosubind{#1}{#2}{}}
   
 % Workhorse for all \fooindexes.  % There is also \dosubind {index}{topic}{subtopic}
   % which makes an entry in a two-level index such as the operation index.
   % TODO: Two-level index?  Operation index?
   
   % Workhorse for all indexes.
 % #1 is name of index, #2 is stuff to put there, #3 is subentry --  % #1 is name of index, #2 is stuff to put there, #3 is subentry --
 % \empty if called from \doind, as we usually are.  The main exception  % empty if called from \doind, as we usually are (the main exception
 % is with defuns, which call us directly.  % is with most defuns, which call us directly).
 %  %
 \def\dosubind#1#2#3{%  \def\dosubind#1#2#3{%
     \iflinks
     {%
       \requireopenindexfile{#1}%
       % Store the main index entry text (including the third arg).
       \toks0 = {#2}%
       % If third arg is present, precede it with a space.
       \def\thirdarg{#3}%
       \ifx\thirdarg\empty \else
         \toks0 = \expandafter{\the\toks0 \space #3}%
       \fi
       %
       \edef\writeto{\csname#1indfile\endcsname}%
       %
       \safewhatsit\dosubindwrite
     }%
     \fi
   }
   
   % Check if an index file has been opened, and if not, open it.
   \def\requireopenindexfile#1{%
   \ifnum\csname #1indfile\endcsname=0
     \expandafter\newwrite \csname#1indfile\endcsname
     \edef\suffix{#1}%
     % A .fls suffix would conflict with the file extension for the output
     % of -recorder, so use .f1s instead.
     \ifx\suffix\indexisfl\def\suffix{f1}\fi
     % Open the file
     \immediate\openout\csname#1indfile\endcsname \jobname.\suffix
     % Using \immediate above here prevents an object entering into the current
     % box, which could confound checks such as those in \safewhatsit for
     % preceding skips.
     \typeout{Writing index file \jobname.\suffix}%
   \fi}
   \def\indexisfl{fl}
   
   % Output \ as {\indexbackslash}, because \ is an escape character in
   % the index files.
   \let\indexbackslash=\relax
   {\catcode`\@=0 \catcode`\\=\active
     @gdef@useindexbackslash{@def\{{@indexbackslash}}}
   }
   
   % Definition for writing index entry text.
   \def\sortas#1{\ignorespaces}%
   
   % Definition for writing index entry sort key.  Should occur at the at
   % the beginning of the index entry, like
   %     @cindex @sortas{september} \september
   % The \ignorespaces takes care of following space, but there's no way
   % to remove space before it.
   {
   \catcode`\-=13
   \gdef\indexwritesortas{%
     \begingroup
     \indexnonalnumreappear
     \indexwritesortasxxx}
   \gdef\indexwritesortasxxx#1{%
     \xdef\indexsortkey{#1}\endgroup}
   }
   
   
   % Write the entry in \toks0 to the index file.
   %
   \def\dosubindwrite{%
   % Put the index entry in the margin if desired.    % Put the index entry in the margin if desired.
   \ifx\SETmarginindex\relax\else    \ifx\SETmarginindex\relax\else
     \insert\margin{\hbox{\vrule height8pt depth3pt width0pt #2}}%      \insert\margin{\hbox{\vrule height8pt depth3pt width0pt \the\toks0}}%
   \fi    \fi
   {%    %
     \count255=\lastpenalty    % Remember, we are within a group.
     {%    \indexdummies % Must do this here, since \bf, etc expand at this stage
       \indexdummies % Must do this here, since \bf, etc expand at this stage    \useindexbackslash % \indexbackslash isn't defined now so it will be output
       \escapechar=`\\                       % as is; and it will print as backslash.
       {%    % The braces around \indexbrace are recognized by texindex.
         \let\folio = 0% We will expand all macros now EXCEPT \folio.    %
         \def\rawbackslashxx{\indexbackslash}% \indexbackslash isn't defined now    % Get the string to sort by, by processing the index entry with all
         % so it will be output as is; and it will print as backslash.    % font commands turned off.
         %    {\indexnofonts
         \def\thirdarg{#3}%     \def\lbracechar{{\indexlbrace}}%
         %     \def\rbracechar{{\indexrbrace}}%
         % If third arg is present, precede it with space in sort key.     \let\{=\lbracechar
         \ifx\thirdarg\emptymacro     \let\}=\rbracechar
           \let\subentry = \empty     \indexnonalnumdisappear
         \else     \xdef\indexsortkey{}%
           \def\subentry{ #3}%     \let\sortas=\indexwritesortas
         \fi     \edef\temp{\the\toks0}%
         %     \setbox\dummybox = \hbox{\temp}% Make sure to execute any \sortas
         % First process the index entry with all font commands turned     \ifx\indexsortkey\empty
         % off to get the string to sort by.       \xdef\indexsortkey{\temp}%
         {\indexnofonts \xdef\indexsorttmp{#2\subentry}}%       \ifx\indexsortkey\empty\xdef\indexsortkey{ }\fi
         %     \fi
         % Now the real index entry with the fonts.  
         \toks0 = {#2}%  
         %  
         % If third (subentry) arg is present, add it to the index  
         % string.  And include a space.  
         \ifx\thirdarg\emptymacro \else  
           \toks0 = \expandafter{\the\toks0 \space #3}%  
         \fi  
         %  
         % Set up the complete index entry, with both the sort key  
         % and the original text, including any font commands.  We write  
         % three arguments to \entry to the .?? file, texindex reduces to  
         % two when writing the .??s sorted result.  
         \edef\temp{%  
           \write\csname#1indfile\endcsname{%  
             \realbackslash entry{\indexsorttmp}{\folio}{\the\toks0}}%  
         }%  
         %  
         % If a skip is the last thing on the list now, preserve it  
         % by backing up by \lastskip, doing the \write, then inserting  
         % the skip again.  Otherwise, the whatsit generated by the  
         % \write will make \lastskip zero.  The result is that sequences  
         % like this:  
         % @end defun  
         % @tindex whatever  
         % @defun ...  
         % will have extra space inserted, because the \medbreak in the  
         % start of the @defun won't see the skip inserted by the @end of  
         % the previous defun.  
         %  
         % But don't do any of this if we're not in vertical mode.  We  
         % don't want to do a \vskip and prematurely end a paragraph.  
         %  
         % Avoid page breaks due to these extra skips, too.  
         %  
         \iflinks  
           \ifvmode  
             \skip0 = \lastskip  
             \ifdim\lastskip = 0pt \else \nobreak\vskip-\lastskip \fi  
           \fi  
           %  
           \temp % do the write  
           %  
           %  
           \ifvmode \ifdim\skip0 = 0pt \else \nobreak\vskip\skip0 \fi \fi  
         \fi  
       }%  
     }%  
     \penalty\count255  
   }%    }%
     %
     % Set up the complete index entry, with both the sort key and
     % the original text, including any font commands.  We write
     % three arguments to \entry to the .?? file (four in the
     % subentry case), texindex reduces to two when writing the .??s
     % sorted result.
     \edef\temp{%
       \write\writeto{%
         \string\entry{\indexsortkey}{\noexpand\folio}{\the\toks0}}%
     }%
     \temp
 }  }
   \newbox\dummybox % used above
   
   % Take care of unwanted page breaks/skips around a whatsit:
   %
   % If a skip is the last thing on the list now, preserve it
   % by backing up by \lastskip, doing the \write, then inserting
   % the skip again.  Otherwise, the whatsit generated by the
   % \write or \pdfdest will make \lastskip zero.  The result is that
   % sequences like this:
   % @end defun
   % @tindex whatever
   % @defun ...
   % will have extra space inserted, because the \medbreak in the
   % start of the @defun won't see the skip inserted by the @end of
   % the previous defun.
   %
   % But don't do any of this if we're not in vertical mode.  We
   % don't want to do a \vskip and prematurely end a paragraph.
   %
   % Avoid page breaks due to these extra skips, too.
   %
   % But wait, there is a catch there:
   % We'll have to check whether \lastskip is zero skip.  \ifdim is not
   % sufficient for this purpose, as it ignores stretch and shrink parts
   % of the skip.  The only way seems to be to check the textual
   % representation of the skip.
   %
   % The following is almost like \def\zeroskipmacro{0.0pt} except that
   % the ``p'' and ``t'' characters have catcode \other, not 11 (letter).
   %
   \edef\zeroskipmacro{\expandafter\the\csname z@skip\endcsname}
   %
   \newskip\whatsitskip
   \newcount\whatsitpenalty
   %
   % ..., ready, GO:
   %
   \def\safewhatsit#1{\ifhmode
     #1%
    \else
     % \lastskip and \lastpenalty cannot both be nonzero simultaneously.
     \whatsitskip = \lastskip
     \edef\lastskipmacro{\the\lastskip}%
     \whatsitpenalty = \lastpenalty
     %
     % If \lastskip is nonzero, that means the last item was a
     % skip.  And since a skip is discardable, that means this
     % -\whatsitskip glue we're inserting is preceded by a
     % non-discardable item, therefore it is not a potential
     % breakpoint, therefore no \nobreak needed.
     \ifx\lastskipmacro\zeroskipmacro
     \else
       \vskip-\whatsitskip
     \fi
     %
     #1%
     %
     \ifx\lastskipmacro\zeroskipmacro
       % If \lastskip was zero, perhaps the last item was a penalty, and
       % perhaps it was >=10000, e.g., a \nobreak.  In that case, we want
       % to re-insert the same penalty (values >10000 are used for various
       % signals); since we just inserted a non-discardable item, any
       % following glue (such as a \parskip) would be a breakpoint.  For example:
       %   @deffn deffn-whatever
       %   @vindex index-whatever
       %   Description.
       % would allow a break between the index-whatever whatsit
       % and the "Description." paragraph.
       \ifnum\whatsitpenalty>9999 \penalty\whatsitpenalty \fi
     \else
       % On the other hand, if we had a nonzero \lastskip,
       % this make-up glue would be preceded by a non-discardable item
       % (the whatsit from the \write), so we must insert a \nobreak.
       \nobreak\vskip\whatsitskip
     \fi
   \fi}
   
 % The index entry written in the file actually looks like  % The index entry written in the file actually looks like
 %  \entry {sortstring}{page}{topic}  %  \entry {sortstring}{page}{topic}
 % or  % or
Line 2938  width0pt\relax} \fi
Line 5500  width0pt\relax} \fi
 % @printindex causes a particular index (the ??s file) to get printed.  % @printindex causes a particular index (the ??s file) to get printed.
 % It does not print any chapter heading (usually an @unnumbered).  % It does not print any chapter heading (usually an @unnumbered).
 %  %
 \def\printindex{\parsearg\doprintindex}  \parseargdef\printindex{\begingroup
 \def\doprintindex#1{\begingroup  
   \dobreak \chapheadingskip{10000}%    \dobreak \chapheadingskip{10000}%
   %    %
   \smallfonts \rm    \smallfonts \rm
   \tolerance = 9500    \tolerance = 9500
   \indexbreaks    \plainfrenchspacing
     \everypar = {}% don't want the \kern\-parindent from indentation suppression.
   %    %
   % See if the index file exists and is nonempty.    % See if the index file exists and is nonempty.
   % Change catcode of @ here so that if the index file contains    % Change catcode of @ here so that if the index file contains
   % \initial {@}    % \initial {@}
   % as its first line, TeX doesn't complain about mismatched braces    % as its first line, TeX doesn't complain about mismatched braces
   % (because it thinks @} is a control sequence).    % (because it thinks @} is a control sequence).
   \catcode`\@ = 11    \catcode`\@ = 12
   \openin 1 \jobname.#1s    % See comment in \requireopenindexfile.
     \def\indexname{#1}\ifx\indexname\indexisfl\def\indexname{f1}\fi
     \openin 1 \jobname.\indexname s
   \ifeof 1    \ifeof 1
     % \enddoublecolumns gets confused if there is no text in the index,      % \enddoublecolumns gets confused if there is no text in the index,
     % and it loses the chapter title and the aux file entries for the      % and it loses the chapter title and the aux file entries for the
     % index.  The easiest way to prevent this problem is to make sure      % index.  The easiest way to prevent this problem is to make sure
     % there is some text.      % there is some text.
     \putwordIndexNonexistent      \putwordIndexNonexistent
       \typeout{No file \jobname.\indexname s.}%
   \else    \else
       \catcode`\\ = 0
     %      %
     % If the index file exists but is empty, then \openin leaves \ifeof      % If the index file exists but is empty, then \openin leaves \ifeof
     % false.  We have to make TeX try to read something from the file, so      % false.  We have to make TeX try to read something from the file, so
     % it can discover if there is anything in it.      % it can discover if there is anything in it.
     \read 1 to \temp      \read 1 to \thisline
     \ifeof 1      \ifeof 1
       \putwordIndexIsEmpty        \putwordIndexIsEmpty
     \else      \else
       % Index files are almost Texinfo source, but we use \ as the escape        % Index files are almost Texinfo source, but we use \ as the escape
       % character.  It would be better to use @, but that's too big a change        % character.  It would be better to use @, but that's too big a change
       % to make right now.        % to make right now.
       \def\indexbackslash{\rawbackslashxx}%        \def\indexbackslash{\ttbackslash}%
       \catcode`\\ = 0        \let\indexlbrace\{   % Likewise, set these sequences for braces
       \escapechar = `\\        \let\indexrbrace\}   % used in the sort key.
       \begindoublecolumns        \begindoublecolumns
       \input \jobname.#1s        \let\entrywidowpenalty=\indexwidowpenalty
         %
         % Read input from the index file line by line.
         \loopdo
           \ifeof1
             \let\firsttoken\relax
           \else
             \read 1 to \nextline
             \edef\act{\gdef\noexpand\firsttoken{\getfirsttoken\nextline}}%
             \act
           \fi
           \thisline
           %
           \ifeof1\else
           \let\thisline\nextline
         \repeat
         %%
       \enddoublecolumns        \enddoublecolumns
     \fi      \fi
   \fi    \fi
   \closein 1    \closein 1
 \endgroup}  \endgroup}
   
   \def\getfirsttoken#1{\expandafter\getfirsttokenx#1\endfirsttoken}
   \long\def\getfirsttokenx#1#2\endfirsttoken{\noexpand#1}
   
   \def\loopdo#1\repeat{\def\body{#1}\loopdoxxx}
   \def\loopdoxxx{\let\next=\relax\body\let\next=\loopdoxxx\fi\next}
   
 % These macros are used by the sorted index file itself.  % These macros are used by the sorted index file itself.
 % Change them to control the appearance of the index.  % Change them to control the appearance of the index.
   
 \def\initial#1{{%  {\catcode`\/=13 \catcode`\-=13 \catcode`\^=13 \catcode`\~=13 \catcode`\_=13
   % Some minor font changes for the special characters.  \catcode`\|=13 \catcode`\<=13 \catcode`\>=13 \catcode`\+=13 \catcode`\"=13
   \let\tentt=\sectt \let\tt=\sectt \let\sf=\sectt  \catcode`\$=3
   \gdef\initialglyphs{%
     % Some changes for non-alphabetic characters.  Using the glyphs from the
     % math fonts looks more consistent than the typewriter font used elsewhere
     % for these characters.
     \def\indexbackslash{\math{\backslash}}%
     \let\\=\indexbackslash
   %    %
     % Can't get bold backslash so don't use bold forward slash
     \catcode`\/=13
     \def/{{\secrmnotbold \normalslash}}%
     \def-{{\normaldash\normaldash}}% en dash `--'
     \def^{{\chapbf \normalcaret}}%
     \def~{{\chapbf \normaltilde}}%
     \def\_{%
        \leavevmode \kern.07em \vbox{\hrule width.3em height.1ex}\kern .07em }%
     \def|{$\vert$}%
     \def<{$\less$}%
     \def>{$\gtr$}%
     \def+{$\normalplus$}%
   }}
   
   \def\initial{%
     \bgroup
     \initialglyphs
     \initialx
   }
   
   \def\initialx#1{%
   % Remove any glue we may have, we'll be inserting our own.    % Remove any glue we may have, we'll be inserting our own.
   \removelastskip    \removelastskip
   %    %
   % We like breaks before the index initials, so insert a bonus.    % We like breaks before the index initials, so insert a bonus.
   \penalty -300    % The glue before the bonus allows a little bit of space at the
     % bottom of a column to reduce an increase in inter-line spacing.
     \nobreak
     \vskip 0pt plus 5\baselineskip
     \penalty -300
     \vskip 0pt plus -5\baselineskip
   %    %
   % Typeset the initial.  Making this add up to a whole number of    % Typeset the initial.  Making this add up to a whole number of
   % baselineskips increases the chance of the dots lining up from column    % baselineskips increases the chance of the dots lining up from column
Line 3001  width0pt\relax} \fi
Line 5621  width0pt\relax} \fi
   % we need before each entry, but it's better.    % we need before each entry, but it's better.
   %    %
   % No shrink because it confuses \balancecolumns.    % No shrink because it confuses \balancecolumns.
   \vskip 1.67\baselineskip plus .5\baselineskip    \vskip 1.67\baselineskip plus 1\baselineskip
   \leftline{\secbf #1}%    \leftline{\secfonts \kern-0.05em \secbf #1}%
   \vskip .33\baselineskip plus .1\baselineskip    % \secfonts is inside the argument of \leftline so that the change of
   %    % \baselineskip will not affect any glue inserted before the vbox that
     % \leftline creates.
   % Do our best not to break after the initial.    % Do our best not to break after the initial.
   \nobreak    \nobreak
 }}    \vskip .33\baselineskip plus .1\baselineskip
     \egroup % \initialglyphs
   }
   
 % This typesets a paragraph consisting of #1, dot leaders, and then #2  \newdimen\entryrightmargin
 % flush to the right margin.  It is used for index and table of contents  \entryrightmargin=0pt
 % entries.  The paragraph is indented by \leftskip.  
   % \entry typesets a paragraph consisting of the text (#1), dot leaders, and
   % then page number (#2) flushed to the right margin.  It is used for index
   % and table of contents entries.  The paragraph is indented by \leftskip.
 %  %
 \def\entry#1#2{\begingroup  \def\entry{%
   %    \begingroup
   % Start a new paragraph if necessary, so our assignments below can't  
   % affect previous text.  
   \par  
   %  
   % Do not fill out the last line with white space.  
   \parfillskip = 0in  
   %  
   % No extra space above this paragraph.  
   \parskip = 0in  
   %  
   % Do not prefer a separate line ending with a hyphen to fewer lines.  
   \finalhyphendemerits = 0  
   %  
   % \hangindent is only relevant when the entry text and page number  
   % don't both fit on one line.  In that case, bob suggests starting the  
   % dots pretty far over on the line.  Unfortunately, a large  
   % indentation looks wrong when the entry text itself is broken across  
   % lines.  So we use a small indentation and put up with long leaders.  
   %  
   % \hangafter is reset to 1 (which is the value we want) at the start  
   % of each paragraph, so we need not do anything with that.  
   \hangindent = 2em  
   %  
   % When the entry text needs to be broken, just fill out the first line  
   % with blank space.  
   \rightskip = 0pt plus1fil  
   %  
   % A bit of stretch before each entry for the benefit of balancing columns.  
   \vskip 0pt plus1pt  
   %  
   % Start a ``paragraph'' for the index entry so the line breaking  
   % parameters we've set above will have an effect.  
   \noindent  
   %  
   % Insert the text of the index entry.  TeX will do line-breaking on it.  
   #1%  
   % The following is kludged to not output a line of dots in the index if  
   % there are no page numbers.  The next person who breaks this will be  
   % cursed by a Unix daemon.  
   \def\tempa{{\rm }}%  
   \def\tempb{#2}%  
   \edef\tempc{\tempa}%  
   \edef\tempd{\tempb}%  
   \ifx\tempc\tempd\ \else%  
     %      %
     % If we must, put the page number on a line of its own, and fill out      % For pdfTeX and XeTeX.
     % this line with blank space.  (The \hfil is overwhelmed with the      % The redefinition of \domark stops marks being added in \pdflink to
     % fill leaders glue in \indexdotfill if the page number does fit.)      % preserve coloured links across page boundaries.  Otherwise the marks
     \hfil\penalty50      % would get in the way of \lastbox in \insertindexentrybox.
     \null\nobreak\indexdotfill % Have leaders before the page number.      \let\domark\relax
     %      %
     % The `\ ' here is removed by the implicit \unskip that TeX does as      % Start a new paragraph if necessary, so our assignments below can't
     % part of (the primitive) \par.  Without it, a spurious underfull      % affect previous text.
     % \hbox ensues.      \par
     \ifpdf      %
       \pdfgettoks#2.\ \the\toksA % The page number ends the paragraph.      % No extra space above this paragraph.
       \parskip = 0in
       %
       % When reading the text of entry, convert explicit line breaks
       % from @* into spaces.  The user might give these in long section
       % titles, for instance.
       \def\*{\unskip\space\ignorespaces}%
       \def\entrybreak{\hfil\break}% An undocumented command
       %
       % Swallow the left brace of the text (first parameter):
       \afterassignment\doentry
       \let\temp =
   }
   \def\entrybreak{\unskip\space\ignorespaces}%
   \def\doentry{%
       % Save the text of the entry
       \global\setbox\boxA=\hbox\bgroup
       \bgroup % Instead of the swallowed brace.
         \noindent
         \aftergroup\finishentry
         % And now comes the text of the entry.
         % Not absorbing as a macro argument reduces the chance of problems
         % with catcodes occurring.
   }
   {\catcode`\@=11
   \gdef\finishentry#1{%
       \egroup % end box A
       \dimen@ = \wd\boxA % Length of text of entry
       \global\setbox\boxA=\hbox\bgroup\unhbox\boxA
       % #1 is the page number.
       %
       % Get the width of the page numbers, and only use
       % leaders if they are present.
       \global\setbox\boxB = \hbox{#1}%
       \ifdim\wd\boxB = 0pt
         \null\nobreak\hfill\ %
     \else      \else
       \ #2% The page number ends the paragraph.        %
         \null\nobreak\indexdotfill % Have leaders before the page number.
         %
         \ifpdf
           \pdfgettoks#1.%
           \hskip\skip\thinshrinkable\the\toksA
         \else
           \ifx\XeTeXrevision\thisisundefined
             \hskip\skip\thinshrinkable #1%
           \else
             \pdfgettoks#1.%
             \hskip\skip\thinshrinkable\the\toksA
           \fi
         \fi
     \fi      \fi
   \fi%      \egroup % end \boxA
   \par      \ifdim\wd\boxB = 0pt
 \endgroup}        \global\setbox\entryindexbox=\vbox{\unhbox\boxA}%
       \else
       \global\setbox\entryindexbox=\vbox\bgroup
         % We want the text of the entries to be aligned to the left, and the
         % page numbers to be aligned to the right.
         %
         \parindent = 0pt
         \advance\leftskip by 0pt plus 1fil
         \advance\leftskip by 0pt plus -1fill
         \rightskip = 0pt plus -1fil
         \advance\rightskip by 0pt plus 1fill
         % Cause last line, which could consist of page numbers on their own
         % if the list of page numbers is long, to be aligned to the right.
         \parfillskip=0pt plus -1fill
         %
         \advance\rightskip by \entryrightmargin
         % Determine how far we can stretch into the margin.
         % This allows, e.g., "Appendix H  GNU Free Documentation License" to
         % fit on one line in @letterpaper format.
         \ifdim\entryrightmargin>2.1em
           \dimen@i=2.1em
         \else
           \dimen@i=0em
         \fi
         \advance \parfillskip by 0pt minus 1\dimen@i
         %
         \dimen@ii = \hsize
         \advance\dimen@ii by -1\leftskip
         \advance\dimen@ii by -1\entryrightmargin
         \advance\dimen@ii by 1\dimen@i
         \ifdim\wd\boxA > \dimen@ii % If the entry doesn't fit in one line
         \ifdim\dimen@ > 0.8\dimen@ii   % due to long index text
           \dimen@ = 0.7\dimen@ % Try to split the text roughly evenly
           \dimen@ii = \hsize
           \ifnum\dimen@>\dimen@ii
             % If the entry is too long, use the whole line
             \dimen@ = \dimen@ii
           \fi
           \advance\leftskip by 0pt plus 1fill % ragged right
           \advance \dimen@ by 1\rightskip
           \parshape = 2 0pt \dimen@ 0em \dimen@ii
           % Ideally we'd add a finite glue at the end of the first line only,
           % instead of using \parshape with explicit line lengths, but TeX
           % doesn't seem to provide a way to do such a thing.
           %
           \leftskip = 1em
           \parindent = -1em
         \fi\fi
         \indent % start paragraph
         \unhbox\boxA
         %
         % Do not prefer a separate line ending with a hyphen to fewer lines.
         \finalhyphendemerits = 0
         %
         % Word spacing - no stretch
         \spaceskip=\fontdimen2\font minus \fontdimen4\font
         %
         \linepenalty=1000  % Discourage line breaks.
         \hyphenpenalty=5000  % Discourage hyphenation.
         %
         \par % format the paragraph
       \egroup % The \vbox
       \fi
     \endgroup
     % delay text of entry until after penalty
     \bgroup\aftergroup\insertindexentrybox
     \entrywidowpenalty
   }}
   
 % Like \dotfill except takes at least 1 em.  \newskip\thinshrinkable
   \skip\thinshrinkable=.15em minus .15em
   
   \newbox\entryindexbox
   \def\insertindexentrybox{%
     \ourunvbox\entryindexbox
   }
   
   % Use \lastbox to take apart vbox box by box, and add each sub-box
   % to the current vertical list.
   \def\ourunvbox#1{%
   \bgroup % for local binding of \delayedbox
     % Remove the last box from box #1
     \global\setbox#1=\vbox{%
       \unvbox#1%
       \unskip % remove any glue
       \unpenalty
       \global\setbox\interbox=\lastbox
     }%
     \setbox\delayedbox=\box\interbox
     \ifdim\ht#1=0pt\else
       \ourunvbox#1 % Repeat on what's left of the box
       \nobreak
     \fi
     \box\delayedbox
   \egroup
   }
   \newbox\delayedbox
   \newbox\interbox
   
   % Default is no penalty
   \let\entrywidowpenalty\egroup
   
   % Used from \printindex.  \firsttoken should be the first token
   % after the \entry.  If it's not another \entry, we are at the last
   % line of a group of index entries, so insert a penalty to discourage
   % widowed index entries.
   \long\def\indexwidowpenalty{%
     \def\isentry{\entry}%
     \ifx\firsttoken\isentry
     \else
       \penalty 9000
     \fi
     \egroup % now comes the box added with \aftergroup
   }
   
   % Like plain.tex's \dotfill, except uses up at least 1 em.
   % The filll stretch here overpowers both the fil and fill stretch to push
   % the page number to the right.
 \def\indexdotfill{\cleaders  \def\indexdotfill{\cleaders
   \hbox{$\mathsurround=0pt \mkern1.5mu ${\it .}$ \mkern1.5mu$}\hskip 1em plus 1fill}    \hbox{$\mathsurround=0pt \mkern1.5mu.\mkern1.5mu$}\hskip 1em plus 1filll}
   
   
 \def\primary #1{\line{#1\hfil}}  \def\primary #1{\line{#1\hfil}}
   
 \newskip\secondaryindent \secondaryindent=0.5cm  \newskip\secondaryindent \secondaryindent=0.5cm
   \def\secondary#1#2{{%
 \def\secondary #1#2{    \parfillskip=0in
 {\parfillskip=0in \parskip=0in    \parskip=0in
 \hangindent =1in \hangafter=1    \hangindent=1in
 \noindent\hskip\secondaryindent\hbox{#1}\indexdotfill #2\par    \hangafter=1
     \noindent\hskip\secondaryindent\hbox{#1}\indexdotfill
     \ifpdf
       \pdfgettoks#2.\ \the\toksA % The page number ends the paragraph.
     \else
       \ifx\XeTeXrevision\thisisundefined
         #2
       \else
         \pdfgettoks#2.\ \the\toksA % The page number ends the paragraph.
       \fi
     \fi
     \par
 }}  }}
   
 % Define two-column mode, which we use to typeset indexes.  % Define two-column mode, which we use to typeset indexes.
 % Adapted from the TeXbook, page 416, which is to say,  % Adapted from the TeXbook, page 416, which is to say,
 % the manmac.tex format used to print the TeXbook itself.  % the manmac.tex format used to print the TeXbook itself.
 \catcode`\@=11  \catcode`\@=11  % private names
   
 \newbox\partialpage  \newbox\partialpage
 \newdimen\doublecolumnhsize  \newdimen\doublecolumnhsize
   
   % Use inside an output routine to save \topmark and \firstmark
   \def\savemarks{%
     \global\savedtopmark=\expandafter{\topmark }%
     \global\savedfirstmark=\expandafter{\firstmark }%
   }
   \newtoks\savedtopmark
   \newtoks\savedfirstmark
   
   % Set \topmark and \firstmark for next time \output runs.
   % Can't be run from withinside \output (because any material
   % added while an output routine is active, including
   % penalties, is saved for after it finishes).  The page so far
   % should be empty, otherwise what's on it will be thrown away.
   \def\restoremarks{%
     \mark{\the\savedtopmark}%
     \bgroup\output = {%
       \setbox\dummybox=\box\PAGE
     }abc\eject\egroup
     % "abc" because output routine doesn't fire for a completely empty page.
     \mark{\the\savedfirstmark}%
   }
   
 \def\begindoublecolumns{\begingroup % ended by \enddoublecolumns  \def\begindoublecolumns{\begingroup % ended by \enddoublecolumns
     % If not much space left on page, start a new page.
     \ifdim\pagetotal>0.8\vsize\vfill\eject\fi
     %
   % Grab any single-column material above us.    % Grab any single-column material above us.
   \output = {%    \output = {%
     %      %
Line 3120  width0pt\relax} \fi
Line 5902  width0pt\relax} \fi
       \unvbox\PAGE        \unvbox\PAGE
       \kern-\topskip \kern\baselineskip        \kern-\topskip \kern\baselineskip
     }%      }%
       \savemarks
   }%    }%
   \eject % run that output routine to set \partialpage    \eject % run that output routine to set \partialpage
     \restoremarks
   %    %
     % We recover the two marks that the last output routine saved in order
     % to propagate the information in marks added around a chapter heading,
     % which could be otherwise be lost by the time the final page is output.
     %
     %
   % Use the double-column output routine for subsequent pages.    % Use the double-column output routine for subsequent pages.
   \output = {\doublecolumnout}%    \output = {\doublecolumnout}%
   %    %
Line 3149  width0pt\relax} \fi
Line 5938  width0pt\relax} \fi
   %    %
   % Double the \vsize as well.  (We don't need a separate register here,    % Double the \vsize as well.  (We don't need a separate register here,
   % since nobody clobbers \vsize.)    % since nobody clobbers \vsize.)
   \advance\vsize by -\ht\partialpage  
   \vsize = 2\vsize    \vsize = 2\vsize
     %
     % For the benefit of balancing columns
     \advance\baselineskip by 0pt plus 0.5pt
 }  }
   
 % The double-column output routine for all double-column pages except  % The double-column output routine for all double-column pages except
 % the last.  % the last, which is done by \balancecolumns.
 %  %
 \def\doublecolumnout{%  \def\doublecolumnout{%
     %
   \splittopskip=\topskip \splitmaxdepth=\maxdepth    \splittopskip=\topskip \splitmaxdepth=\maxdepth
   % Get the available space for the double columns -- the normal    % Get the available space for the double columns -- the normal
   % (undoubled) page height minus any material left over from the    % (undoubled) page height minus any material left over from the
   % previous page.    % previous page.
   \dimen@ = \vsize    \dimen@ = \vsize
   \divide\dimen@ by 2    \divide\dimen@ by 2
     \advance\dimen@ by -\ht\partialpage
   %    %
   % box0 will be the left-hand column, box2 the right.    % box0 will be the left-hand column, box2 the right.
   \setbox0=\vsplit255 to\dimen@ \setbox2=\vsplit255 to\dimen@    \setbox0=\vsplit255 to\dimen@ \setbox2=\vsplit255 to\dimen@
Line 3170  width0pt\relax} \fi
Line 5963  width0pt\relax} \fi
   \unvbox255    \unvbox255
   \penalty\outputpenalty    \penalty\outputpenalty
 }  }
   %
   % Re-output the contents of the output page -- any previous material,
   % followed by the two boxes we just split, in box0 and box2.
 \def\pagesofar{%  \def\pagesofar{%
   % Re-output the contents of the output page -- any previous material,  
   % followed by the two boxes we just split, in box0 and box2.  
   \unvbox\partialpage    \unvbox\partialpage
   %    %
   \hsize = \doublecolumnhsize    \hsize = \doublecolumnhsize
   \wd0=\hsize \wd2=\hsize    \wd0=\hsize \wd2=\hsize
   \hbox to\pagewidth{\box0\hfil\box2}%    \hbox to\txipagewidth{\box0\hfil\box2}%
 }  }
   
   
   % Finished with with double columns.
 \def\enddoublecolumns{%  \def\enddoublecolumns{%
     % The following penalty ensures that the page builder is exercised
     % _before_ we change the output routine.  This is necessary in the
     % following situation:
     %
     % The last section of the index consists only of a single entry.
     % Before this section, \pagetotal is less than \pagegoal, so no
     % break occurs before the last section starts.  However, the last
     % section, consisting of \initial and the single \entry, does not
     % fit on the page and has to be broken off.  Without the following
     % penalty the page builder will not be exercised until \eject
     % below, and by that time we'll already have changed the output
     % routine to the \balancecolumns version, so the next-to-last
     % double-column page will be processed with \balancecolumns, which
     % is wrong:  The two columns will go to the main vertical list, with
     % the broken-off section in the recent contributions.  As soon as
     % the output routine finishes, TeX starts reconsidering the page
     % break.  The two columns and the broken-off section both fit on the
     % page, because the two columns now take up only half of the page
     % goal.  When TeX sees \eject from below which follows the final
     % section, it invokes the new output routine that we've set after
     % \balancecolumns below; \onepageout will try to fit the two columns
     % and the final section into the vbox of \txipageheight (see
     % \pagebody), causing an overfull box.
     %
     % Note that glue won't work here, because glue does not exercise the
     % page builder, unlike penalties (see The TeXbook, pp. 280-281).
     \penalty0
     %
   \output = {%    \output = {%
     % Split the last of the double-column material.  Leave it on the      % Split the last of the double-column material.
     % current page, no automatic page break.      \savemarks
     \balancecolumns      \balancecolumns
     %      %
     % If we end up splitting too much material for the current page,      % Having called \balancecolumns once, we do not
     % though, there will be another page break right after this \output  
     % invocation ends.  Having called \balancecolumns once, we do not  
     % want to call it again.  Therefore, reset \output to its normal      % want to call it again.  Therefore, reset \output to its normal
     % definition right away.  (We hope \balancecolumns will never be      % definition right away.
     % called on to balance too much material, but if it is, this makes  
     % the output somewhat more palatable.)  
     \global\output = {\onepageout{\pagecontents\PAGE}}%      \global\output = {\onepageout{\pagecontents\PAGE}}%
   }%    }%
   \eject    \eject
   \endgroup % started in \begindoublecolumns    \endgroup % started in \begindoublecolumns
     \restoremarks
     % Leave the double-column material on the current page, no automatic
     % page break.
     \box\balancedcolumns
   %    %
   % \pagegoal was set to the doubled \vsize above, since we restarted    % \pagegoal was set to the doubled \vsize above, since we restarted
   % the current page.  We're now back to normal single-column    % the current page.  We're now back to normal single-column
Line 3203  width0pt\relax} \fi
Line 6028  width0pt\relax} \fi
   % \endgroup where \vsize got restored).    % \endgroup where \vsize got restored).
   \pagegoal = \vsize    \pagegoal = \vsize
 }  }
   \newbox\balancedcolumns
   \setbox\balancedcolumns=\vbox{shouldnt see this}%
   %
   % Only called for the last of the double column material.  \doublecolumnout
   % does the others.
 \def\balancecolumns{%  \def\balancecolumns{%
   % Called at the end of the double column material.  
   \setbox0 = \vbox{\unvbox255}% like \box255 but more efficient, see p.120.    \setbox0 = \vbox{\unvbox255}% like \box255 but more efficient, see p.120.
   \dimen@ = \ht0    \dimen@ = \ht0
   \advance\dimen@ by \topskip    \advance\dimen@ by \topskip
   \advance\dimen@ by-\baselineskip    \advance\dimen@ by-\baselineskip
   \divide\dimen@ by 2 % target to split to    \ifdim\dimen@<5\baselineskip
   %debug\message{final 2-column material height=\the\ht0, target=\the\dimen@.}%      % Don't split a short final column in two.
   \splittopskip = \topskip      \setbox2=\vbox{}%
   % Loop until we get a decent breakpoint.    \else
   {%      \divide\dimen@ by 2 % target to split to
     \vbadness = 10000      \dimen@ii = \dimen@
     \loop      \splittopskip = \topskip
       \global\setbox3 = \copy0      % Loop until left column is at least as high as the right column.
       \global\setbox1 = \vsplit3 to \dimen@      {%
     \ifdim\ht3>\dimen@        \vbadness = 10000
       \global\advance\dimen@ by 1pt        \loop
     \repeat          \global\setbox3 = \copy0
   }%          \global\setbox1 = \vsplit3 to \dimen@
   %debug\message{split to \the\dimen@, column heights: \the\ht1, \the\ht3.}%        \ifdim\ht1<\ht3
   \setbox0=\vbox to\dimen@{\unvbox1}%          \global\advance\dimen@ by 1pt
   \setbox2=\vbox to\dimen@{\unvbox3}%        \repeat
       }%
       % Now the left column is in box 1, and the right column in box 3.
       % Check whether the left column has come out higher than the page itself.
       % (Note that we have doubled \vsize for the double columns, so
       % the actual height of the page is 0.5\vsize).
       \ifdim2\ht1>\vsize
         % Just split the last of the double column material roughly in half.
         \setbox2=\box0
         \setbox0 = \vsplit2 to \dimen@ii
         \setbox0=\vbox to \dimen@ii {\unvbox0\vfill}%
         \setbox2=\vbox to \dimen@ii {\unvbox2\vfill}%
       \else
         % Compare the heights of the two columns.
         \ifdim4\ht1>5\ht3
           % Column heights are too different, so don't make their bottoms
           % flush with each other.
           \setbox2=\vbox to \ht1 {\unvbox3\vfill}%
           \setbox0=\vbox to \ht1 {\unvbox1\vfill}%
         \else
           % Make column bottoms flush with each other.
           \setbox2=\vbox to\ht1{\unvbox3\unskip}%
           \setbox0=\vbox to\ht1{\unvbox1\unskip}%
         \fi
       \fi
     \fi
   %    %
   \pagesofar    \global\setbox\balancedcolumns=\vbox{\pagesofar}%
 }  }
 \catcode`\@ = \other  \catcode`\@ = \other
   
Line 3234  width0pt\relax} \fi
Line 6088  width0pt\relax} \fi
 \message{sectioning,}  \message{sectioning,}
 % Chapters, sections, etc.  % Chapters, sections, etc.
   
   % Let's start with @part.
   \outer\parseargdef\part{\partzzz{#1}}
   \def\partzzz#1{%
     \chapoddpage
     \null
     \vskip.3\vsize  % move it down on the page a bit
     \begingroup
       \noindent \titlefonts\rmisbold #1\par % the text
       \let\lastnode=\empty      % no node to associate with
       \writetocentry{part}{#1}{}% but put it in the toc
       \headingsoff              % no headline or footline on the part page
       % This outputs a mark at the end of the page that clears \thischapter
       % and \thissection, as is done in \startcontents.
       \let\pchapsepmacro\relax
       \chapmacro{}{Yomitfromtoc}{}%
       \chapoddpage
     \endgroup
   }
   
   % \unnumberedno is an oxymoron.  But we count the unnumbered
   % sections so that we can refer to them unambiguously in the pdf
   % outlines by their "section number".  We avoid collisions with chapter
   % numbers by starting them at 10000.  (If a document ever has 10000
   % chapters, we're in trouble anyway, I'm sure.)
   \newcount\unnumberedno \unnumberedno = 10000
 \newcount\chapno  \newcount\chapno
 \newcount\secno        \secno=0  \newcount\secno        \secno=0
 \newcount\subsecno     \subsecno=0  \newcount\subsecno     \subsecno=0
Line 3241  width0pt\relax} \fi
Line 6120  width0pt\relax} \fi
   
 % This counter is funny since it counts through charcodes of letters A, B, ...  % This counter is funny since it counts through charcodes of letters A, B, ...
 \newcount\appendixno  \appendixno = `\@  \newcount\appendixno  \appendixno = `\@
   %
 % \def\appendixletter{\char\the\appendixno}  % \def\appendixletter{\char\the\appendixno}
 % We do the following for the sake of pdftex, which needs the actual  % We do the following ugly conditional instead of the above simple
   % construct for the sake of pdftex, which needs the actual
 % letter in the expansion, not just typeset.  % letter in the expansion, not just typeset.
   %
 \def\appendixletter{%  \def\appendixletter{%
   \ifnum\appendixno=`A A%    \ifnum\appendixno=`A A%
   \else\ifnum\appendixno=`B B%    \else\ifnum\appendixno=`B B%
Line 3279  width0pt\relax} \fi
Line 6161  width0pt\relax} \fi
   \fi\fi\fi\fi\fi\fi\fi\fi\fi\fi\fi\fi\fi    \fi\fi\fi\fi\fi\fi\fi\fi\fi\fi\fi\fi\fi
   \fi\fi\fi\fi\fi\fi\fi\fi\fi\fi\fi\fi\fi}    \fi\fi\fi\fi\fi\fi\fi\fi\fi\fi\fi\fi\fi}
   
 % Each @chapter defines this as the name of the chapter.  % Each @chapter defines these (using marks) as the number+name, number
 % page headings and footings can use it.  @section does likewise.  % and name of the chapter.  Page headings and footings can use
   % these.  @section does likewise.
 \def\thischapter{}  \def\thischapter{}
   \def\thischapternum{}
   \def\thischaptername{}
 \def\thissection{}  \def\thissection{}
   \def\thissectionnum{}
   \def\thissectionname{}
   
 \newcount\absseclevel % used to calculate proper heading level  \newcount\absseclevel % used to calculate proper heading level
 \newcount\secbase\secbase=0 % @raise/lowersections modify this count  \newcount\secbase\secbase=0 % @raisesections/@lowersections modify this count
   
 % @raisesections: treat @section as chapter, @subsection as section, etc.  % @raisesections: treat @section as chapter, @subsection as section, etc.
 \def\raisesections{\global\advance\secbase by -1}  \def\raisesections{\global\advance\secbase by -1}
Line 3295  width0pt\relax} \fi
Line 6182  width0pt\relax} \fi
 \def\lowersections{\global\advance\secbase by 1}  \def\lowersections{\global\advance\secbase by 1}
 \let\down=\lowersections % original BFox name  \let\down=\lowersections % original BFox name
   
 % Choose a numbered-heading macro  % we only have subsub.
 % #1 is heading level if unmodified by @raisesections or @lowersections  \chardef\maxseclevel = 3
 % #2 is text for heading  %
 \def\numhead#1#2{\absseclevel=\secbase\advance\absseclevel by #1  % A numbered section within an unnumbered changes to unnumbered too.
 \ifcase\absseclevel  % To achieve this, remember the "biggest" unnum. sec. we are currently in:
   \chapterzzz{#2}  \chardef\unnlevel = \maxseclevel
 \or  %
   \seczzz{#2}  % Trace whether the current chapter is an appendix or not:
 \or  % \chapheadtype is "N" or "A", unnumbered chapters are ignored.
   \numberedsubseczzz{#2}  \def\chapheadtype{N}
 \or  
   \numberedsubsubseczzz{#2}  % Choose a heading macro
 \else  % #1 is heading type
   \ifnum \absseclevel<0  % #2 is heading level
     \chapterzzz{#2}  % #3 is text for heading
   \def\genhead#1#2#3{%
     % Compute the abs. sec. level:
     \absseclevel=#2
     \advance\absseclevel by \secbase
     % Make sure \absseclevel doesn't fall outside the range:
     \ifnum \absseclevel < 0
       \absseclevel = 0
   \else    \else
     \numberedsubsubseczzz{#2}      \ifnum \absseclevel > 3
         \absseclevel = 3
       \fi
   \fi    \fi
 \fi    % The heading type:
 }    \def\headtype{#1}%
     \if \headtype U%
 % like \numhead, but chooses appendix heading levels      \ifnum \absseclevel < \unnlevel
 \def\apphead#1#2{\absseclevel=\secbase\advance\absseclevel by #1        \chardef\unnlevel = \absseclevel
 \ifcase\absseclevel      \fi
   \appendixzzz{#2}  
 \or  
   \appendixsectionzzz{#2}  
 \or  
   \appendixsubseczzz{#2}  
 \or  
   \appendixsubsubseczzz{#2}  
 \else  
   \ifnum \absseclevel<0  
     \appendixzzz{#2}  
   \else    \else
     \appendixsubsubseczzz{#2}      % Check for appendix sections:
       \ifnum \absseclevel = 0
         \edef\chapheadtype{\headtype}%
       \else
         \if \headtype A\if \chapheadtype N%
           \errmessage{@appendix... within a non-appendix chapter}%
         \fi\fi
       \fi
       % Check for numbered within unnumbered:
       \ifnum \absseclevel > \unnlevel
         \def\headtype{U}%
       \else
         \chardef\unnlevel = 3
       \fi
   \fi    \fi
 \fi    % Now print the heading:
 }    \if \headtype U%
       \ifcase\absseclevel
 % like \numhead, but chooses numberless heading levels          \unnumberedzzz{#3}%
 \def\unnmhead#1#2{\absseclevel=\secbase\advance\absseclevel by #1      \or \unnumberedseczzz{#3}%
 \ifcase\absseclevel      \or \unnumberedsubseczzz{#3}%
   \unnumberedzzz{#2}      \or \unnumberedsubsubseczzz{#3}%
 \or      \fi
   \unnumberedseczzz{#2}  
 \or  
   \unnumberedsubseczzz{#2}  
 \or  
   \unnumberedsubsubseczzz{#2}  
 \else  
   \ifnum \absseclevel<0  
     \unnumberedzzz{#2}  
   \else    \else
     \unnumberedsubsubseczzz{#2}      \if \headtype A%
         \ifcase\absseclevel
             \appendixzzz{#3}%
         \or \appendixsectionzzz{#3}%
         \or \appendixsubseczzz{#3}%
         \or \appendixsubsubseczzz{#3}%
         \fi
       \else
         \ifcase\absseclevel
             \chapterzzz{#3}%
         \or \seczzz{#3}%
         \or \numberedsubseczzz{#3}%
         \or \numberedsubsubseczzz{#3}%
         \fi
       \fi
   \fi    \fi
 \fi    \suppressfirstparagraphindent
 }  }
   
 % @chapter, @appendix, @unnumbered.  % an interface:
 \def\thischaptername{No Chapter Title}  \def\numhead{\genhead N}
 \outer\def\chapter{\parsearg\chapteryyy}  \def\apphead{\genhead A}
 \def\chapteryyy #1{\numhead0{#1}} % normally numhead0 calls chapterzzz  \def\unnmhead{\genhead U}
 \def\chapterzzz #1{%  
 \secno=0 \subsecno=0 \subsubsecno=0  % @chapter, @appendix, @unnumbered.  Increment top-level counter, reset
 \global\advance \chapno by 1 \message{\putwordChapter\space \the\chapno}%  % all lower-level sectioning counters to zero.
 \chapmacro {#1}{\the\chapno}%  %
 \gdef\thissection{#1}%  % Also set \chaplevelprefix, which we prepend to @float sequence numbers
 \gdef\thischaptername{#1}%  % (e.g., figures), q.v.  By default (before any chapter), that is empty.
 % We don't substitute the actual chapter name into \thischapter  \let\chaplevelprefix = \empty
 % because we don't want its macros evaluated now.  %
 \xdef\thischapter{\putwordChapter{} \the\chapno: \noexpand\thischaptername}%  \outer\parseargdef\chapter{\numhead0{#1}} % normally numhead0 calls chapterzzz
 \toks0 = {#1}%  \def\chapterzzz#1{%
 \edef\temp{\noexpand\writetocentry{\realbackslash chapentry{\the\toks0}%    % section resetting is \global in case the chapter is in a group, such
                                   {\the\chapno}}}%    % as an @include file.
 \temp    \global\secno=0 \global\subsecno=0 \global\subsubsecno=0
 \donoderef      \global\advance\chapno by 1
 \global\let\section = \numberedsec    %
 \global\let\subsection = \numberedsubsec    % Used for \float.
 \global\let\subsubsection = \numberedsubsubsec    \gdef\chaplevelprefix{\the\chapno.}%
     \resetallfloatnos
     %
     % \putwordChapter can contain complex things in translations.
     \toks0=\expandafter{\putwordChapter}%
     \message{\the\toks0 \space \the\chapno}%
     %
     % Write the actual heading.
     \chapmacro{#1}{Ynumbered}{\the\chapno}%
     %
     % So @section and the like are numbered underneath this chapter.
     \global\let\section = \numberedsec
     \global\let\subsection = \numberedsubsec
     \global\let\subsubsection = \numberedsubsubsec
 }  }
   
 \outer\def\appendix{\parsearg\appendixyyy}  \outer\parseargdef\appendix{\apphead0{#1}} % normally calls appendixzzz
 \def\appendixyyy #1{\apphead0{#1}} % normally apphead0 calls appendixzzz  %
 \def\appendixzzz #1{%  \def\appendixzzz#1{%
 \secno=0 \subsecno=0 \subsubsecno=0    \global\secno=0 \global\subsecno=0 \global\subsubsecno=0
 \global\advance \appendixno by 1      \global\advance\appendixno by 1
 \message{\putwordAppendix\space \appendixletter}%    \gdef\chaplevelprefix{\appendixletter.}%
 \chapmacro {#1}{\putwordAppendix{} \appendixletter}%    \resetallfloatnos
 \gdef\thissection{#1}%    %
 \gdef\thischaptername{#1}%    % \putwordAppendix can contain complex things in translations.
 \xdef\thischapter{\putwordAppendix{} \appendixletter: \noexpand\thischaptername}%    \toks0=\expandafter{\putwordAppendix}%
 \toks0 = {#1}%    \message{\the\toks0 \space \appendixletter}%
 \edef\temp{\noexpand\writetocentry{\realbackslash chapentry{\the\toks0}%    %
                        {\putwordAppendix{} \appendixletter}}}%    \chapmacro{#1}{Yappendix}{\appendixletter}%
 \temp    %
 \appendixnoderef    \global\let\section = \appendixsec
 \global\let\section = \appendixsec    \global\let\subsection = \appendixsubsec
 \global\let\subsection = \appendixsubsec    \global\let\subsubsection = \appendixsubsubsec
 \global\let\subsubsection = \appendixsubsubsec  
 }  }
   
   % normally unnmhead0 calls unnumberedzzz:
   \outer\parseargdef\unnumbered{\unnmhead0{#1}}
   \def\unnumberedzzz#1{%
     \global\secno=0 \global\subsecno=0 \global\subsubsecno=0
       \global\advance\unnumberedno by 1
     %
     % Since an unnumbered has no number, no prefix for figures.
     \global\let\chaplevelprefix = \empty
     \resetallfloatnos
     %
     % This used to be simply \message{#1}, but TeX fully expands the
     % argument to \message.  Therefore, if #1 contained @-commands, TeX
     % expanded them.  For example, in `@unnumbered The @cite{Book}', TeX
     % expanded @cite (which turns out to cause errors because \cite is meant
     % to be executed, not expanded).
     %
     % Anyway, we don't want the fully-expanded definition of @cite to appear
     % as a result of the \message, we just want `@cite' itself.  We use
     % \the<toks register> to achieve this: TeX expands \the<toks> only once,
     % simply yielding the contents of <toks register>.  (We also do this for
     % the toc entries.)
     \toks0 = {#1}%
     \message{(\the\toks0)}%
     %
     \chapmacro{#1}{Ynothing}{\the\unnumberedno}%
     %
     \global\let\section = \unnumberedsec
     \global\let\subsection = \unnumberedsubsec
     \global\let\subsubsection = \unnumberedsubsubsec
   }
   
 % @centerchap is like @unnumbered, but the heading is centered.  % @centerchap is like @unnumbered, but the heading is centered.
 \outer\def\centerchap{\parsearg\centerchapyyy}  \outer\parseargdef\centerchap{%
 \def\centerchapyyy #1{{\let\unnumbchapmacro=\centerchapmacro \unnumberedyyy{#1}}}    \let\centerparametersmaybe = \centerparameters
     \unnmhead0{#1}%
     \let\centerparametersmaybe = \relax
   }
   
 % @top is like @unnumbered.  % @top is like @unnumbered.
 \outer\def\top{\parsearg\unnumberedyyy}  \let\top\unnumbered
   
 \outer\def\unnumbered{\parsearg\unnumberedyyy}  
 \def\unnumberedyyy #1{\unnmhead0{#1}} % normally unnmhead0 calls unnumberedzzz  
 \def\unnumberedzzz #1{%  
 \secno=0 \subsecno=0 \subsubsecno=0  
 %  
 % This used to be simply \message{#1}, but TeX fully expands the  
 % argument to \message.  Therefore, if #1 contained @-commands, TeX  
 % expanded them.  For example, in `@unnumbered The @cite{Book}', TeX  
 % expanded @cite (which turns out to cause errors because \cite is meant  
 % to be executed, not expanded).  
 %  
 % Anyway, we don't want the fully-expanded definition of @cite to appear  
 % as a result of the \message, we just want `@cite' itself.  We use  
 % \the<toks register> to achieve this: TeX expands \the<toks> only once,  
 % simply yielding the contents of <toks register>.  (We also do this for  
 % the toc entries.)  
 \toks0 = {#1}\message{(\the\toks0)}%  
 %  
 \unnumbchapmacro {#1}%  
 \gdef\thischapter{#1}\gdef\thissection{#1}%  
 \toks0 = {#1}%  
 \edef\temp{\noexpand\writetocentry{\realbackslash unnumbchapentry{\the\toks0}}}%  
 \temp  
 \unnumbnoderef  
 \global\let\section = \unnumberedsec  
 \global\let\subsection = \unnumberedsubsec  
 \global\let\subsubsection = \unnumberedsubsubsec  
 }  
   
 % Sections.  % Sections.
 \outer\def\numberedsec{\parsearg\secyyy}  %
 \def\secyyy #1{\numhead1{#1}} % normally calls seczzz  \outer\parseargdef\numberedsec{\numhead1{#1}} % normally calls seczzz
 \def\seczzz #1{%  \def\seczzz#1{%
 \subsecno=0 \subsubsecno=0 \global\advance \secno by 1 %    \global\subsecno=0 \global\subsubsecno=0  \global\advance\secno by 1
 \gdef\thissection{#1}\secheading {#1}{\the\chapno}{\the\secno}%    \sectionheading{#1}{sec}{Ynumbered}{\the\chapno.\the\secno}%
 \toks0 = {#1}%  
 \edef\temp{\noexpand\writetocentry{\realbackslash secentry{\the\toks0}%  
                                   {\the\chapno}{\the\secno}}}%  
 \temp  
 \donoderef  
 \nobreak  
 }  }
   
 \outer\def\appendixsection{\parsearg\appendixsecyyy}  % normally calls appendixsectionzzz:
 \outer\def\appendixsec{\parsearg\appendixsecyyy}  \outer\parseargdef\appendixsection{\apphead1{#1}}
 \def\appendixsecyyy #1{\apphead1{#1}} % normally calls appendixsectionzzz  \def\appendixsectionzzz#1{%
 \def\appendixsectionzzz #1{%    \global\subsecno=0 \global\subsubsecno=0  \global\advance\secno by 1
 \subsecno=0 \subsubsecno=0 \global\advance \secno by 1 %    \sectionheading{#1}{sec}{Yappendix}{\appendixletter.\the\secno}%
 \gdef\thissection{#1}\secheading {#1}{\appendixletter}{\the\secno}%  
 \toks0 = {#1}%  
 \edef\temp{\noexpand\writetocentry{\realbackslash secentry{\the\toks0}%  
                                   {\appendixletter}{\the\secno}}}%  
 \temp  
 \appendixnoderef  
 \nobreak  
 }  }
   \let\appendixsec\appendixsection
   
 \outer\def\unnumberedsec{\parsearg\unnumberedsecyyy}  % normally calls unnumberedseczzz:
 \def\unnumberedsecyyy #1{\unnmhead1{#1}} % normally calls unnumberedseczzz  \outer\parseargdef\unnumberedsec{\unnmhead1{#1}}
 \def\unnumberedseczzz #1{%  \def\unnumberedseczzz#1{%
 \plainsecheading {#1}\gdef\thissection{#1}%    \global\subsecno=0 \global\subsubsecno=0  \global\advance\secno by 1
 \toks0 = {#1}%    \sectionheading{#1}{sec}{Ynothing}{\the\unnumberedno.\the\secno}%
 \edef\temp{\noexpand\writetocentry{\realbackslash unnumbsecentry{\the\toks0}}}%  
 \temp  
 \unnumbnoderef  
 \nobreak  
 }  }
   
 % Subsections.  % Subsections.
 \outer\def\numberedsubsec{\parsearg\numberedsubsecyyy}  %
 \def\numberedsubsecyyy #1{\numhead2{#1}} % normally calls numberedsubseczzz  % normally calls numberedsubseczzz:
 \def\numberedsubseczzz #1{%  \outer\parseargdef\numberedsubsec{\numhead2{#1}}
 \gdef\thissection{#1}\subsubsecno=0 \global\advance \subsecno by 1 %  \def\numberedsubseczzz#1{%
 \subsecheading {#1}{\the\chapno}{\the\secno}{\the\subsecno}%    \global\subsubsecno=0  \global\advance\subsecno by 1
 \toks0 = {#1}%    \sectionheading{#1}{subsec}{Ynumbered}{\the\chapno.\the\secno.\the\subsecno}%
 \edef\temp{\noexpand\writetocentry{\realbackslash subsecentry{\the\toks0}%  
                                     {\the\chapno}{\the\secno}{\the\subsecno}}}%  
 \temp  
 \donoderef  
 \nobreak  
 }  }
   
 \outer\def\appendixsubsec{\parsearg\appendixsubsecyyy}  % normally calls appendixsubseczzz:
 \def\appendixsubsecyyy #1{\apphead2{#1}} % normally calls appendixsubseczzz  \outer\parseargdef\appendixsubsec{\apphead2{#1}}
 \def\appendixsubseczzz #1{%  \def\appendixsubseczzz#1{%
 \gdef\thissection{#1}\subsubsecno=0 \global\advance \subsecno by 1 %    \global\subsubsecno=0  \global\advance\subsecno by 1
 \subsecheading {#1}{\appendixletter}{\the\secno}{\the\subsecno}%    \sectionheading{#1}{subsec}{Yappendix}%
 \toks0 = {#1}%                   {\appendixletter.\the\secno.\the\subsecno}%
 \edef\temp{\noexpand\writetocentry{\realbackslash subsecentry{\the\toks0}%  
                                 {\appendixletter}{\the\secno}{\the\subsecno}}}%  
 \temp  
 \appendixnoderef  
 \nobreak  
 }  }
   
 \outer\def\unnumberedsubsec{\parsearg\unnumberedsubsecyyy}  % normally calls unnumberedsubseczzz:
 \def\unnumberedsubsecyyy #1{\unnmhead2{#1}} %normally calls unnumberedsubseczzz  \outer\parseargdef\unnumberedsubsec{\unnmhead2{#1}}
 \def\unnumberedsubseczzz #1{%  \def\unnumberedsubseczzz#1{%
 \plainsubsecheading {#1}\gdef\thissection{#1}%    \global\subsubsecno=0  \global\advance\subsecno by 1
 \toks0 = {#1}%    \sectionheading{#1}{subsec}{Ynothing}%
 \edef\temp{\noexpand\writetocentry{\realbackslash unnumbsubsecentry%                   {\the\unnumberedno.\the\secno.\the\subsecno}%
                                     {\the\toks0}}}%  
 \temp  
 \unnumbnoderef  
 \nobreak  
 }  }
   
 % Subsubsections.  % Subsubsections.
 \outer\def\numberedsubsubsec{\parsearg\numberedsubsubsecyyy}  %
 \def\numberedsubsubsecyyy #1{\numhead3{#1}} % normally numberedsubsubseczzz  % normally numberedsubsubseczzz:
 \def\numberedsubsubseczzz #1{%  \outer\parseargdef\numberedsubsubsec{\numhead3{#1}}
 \gdef\thissection{#1}\global\advance \subsubsecno by 1 %  \def\numberedsubsubseczzz#1{%
 \subsubsecheading {#1}    \global\advance\subsubsecno by 1
   {\the\chapno}{\the\secno}{\the\subsecno}{\the\subsubsecno}%    \sectionheading{#1}{subsubsec}{Ynumbered}%
 \toks0 = {#1}%                   {\the\chapno.\the\secno.\the\subsecno.\the\subsubsecno}%
 \edef\temp{\noexpand\writetocentry{\realbackslash subsubsecentry{\the\toks0}%  
   {\the\chapno}{\the\secno}{\the\subsecno}{\the\subsubsecno}}}%  
 \temp  
 \donoderef  
 \nobreak  
 }  }
   
 \outer\def\appendixsubsubsec{\parsearg\appendixsubsubsecyyy}  % normally appendixsubsubseczzz:
 \def\appendixsubsubsecyyy #1{\apphead3{#1}} % normally appendixsubsubseczzz  \outer\parseargdef\appendixsubsubsec{\apphead3{#1}}
 \def\appendixsubsubseczzz #1{%  \def\appendixsubsubseczzz#1{%
 \gdef\thissection{#1}\global\advance \subsubsecno by 1 %    \global\advance\subsubsecno by 1
 \subsubsecheading {#1}    \sectionheading{#1}{subsubsec}{Yappendix}%
   {\appendixletter}{\the\secno}{\the\subsecno}{\the\subsubsecno}%                   {\appendixletter.\the\secno.\the\subsecno.\the\subsubsecno}%
 \toks0 = {#1}%  
 \edef\temp{\noexpand\writetocentry{\realbackslash subsubsecentry{\the\toks0}%  
   {\appendixletter}{\the\secno}{\the\subsecno}{\the\subsubsecno}}}%  
 \temp  
 \appendixnoderef  
 \nobreak  
 }  }
   
 \outer\def\unnumberedsubsubsec{\parsearg\unnumberedsubsubsecyyy}  % normally unnumberedsubsubseczzz:
 \def\unnumberedsubsubsecyyy #1{\unnmhead3{#1}} %normally unnumberedsubsubseczzz  \outer\parseargdef\unnumberedsubsubsec{\unnmhead3{#1}}
 \def\unnumberedsubsubseczzz #1{%  \def\unnumberedsubsubseczzz#1{%
 \plainsubsubsecheading {#1}\gdef\thissection{#1}%    \global\advance\subsubsecno by 1
 \toks0 = {#1}%    \sectionheading{#1}{subsubsec}{Ynothing}%
 \edef\temp{\noexpand\writetocentry{\realbackslash unnumbsubsubsecentry%                   {\the\unnumberedno.\the\secno.\the\subsecno.\the\subsubsecno}%
                                     {\the\toks0}}}%  
 \temp  
 \unnumbnoderef  
 \nobreak  
 }  }
   
 % These are variants which are not "outer", so they can appear in @ifinfo.  
 % Actually, they should now be obsolete; ordinary section commands should work.  
 \def\infotop{\parsearg\unnumberedzzz}  
 \def\infounnumbered{\parsearg\unnumberedzzz}  
 \def\infounnumberedsec{\parsearg\unnumberedseczzz}  
 \def\infounnumberedsubsec{\parsearg\unnumberedsubseczzz}  
 \def\infounnumberedsubsubsec{\parsearg\unnumberedsubsubseczzz}  
   
 \def\infoappendix{\parsearg\appendixzzz}  
 \def\infoappendixsec{\parsearg\appendixseczzz}  
 \def\infoappendixsubsec{\parsearg\appendixsubseczzz}  
 \def\infoappendixsubsubsec{\parsearg\appendixsubsubseczzz}  
   
 \def\infochapter{\parsearg\chapterzzz}  
 \def\infosection{\parsearg\sectionzzz}  
 \def\infosubsection{\parsearg\subsectionzzz}  
 \def\infosubsubsection{\parsearg\subsubsectionzzz}  
   
 % These macros control what the section commands do, according  % These macros control what the section commands do, according
 % to what kind of chapter we are in (ordinary, appendix, or unnumbered).  % to what kind of chapter we are in (ordinary, appendix, or unnumbered).
 % Define them by default for a numbered chapter.  % Define them by default for a numbered chapter.
 \global\let\section = \numberedsec  \let\section = \numberedsec
 \global\let\subsection = \numberedsubsec  \let\subsection = \numberedsubsec
 \global\let\subsubsection = \numberedsubsubsec  \let\subsubsection = \numberedsubsubsec
   
 % Define @majorheading, @heading and @subheading  % Define @majorheading, @heading and @subheading
   
 % NOTE on use of \vbox for chapter headings, section headings, and such:  \def\majorheading{%
 %       1) We use \vbox rather than the earlier \line to permit    {\advance\chapheadingskip by 10pt \chapbreak }%
 %          overlong headings to fold.    \parsearg\chapheadingzzz
 %       2) \hyphenpenalty is set to 10000 because hyphenation in a  }
 %          heading is obnoxious; this forbids it.  
 %       3) Likewise, headings look best if no \parindent is used, and  
 %          if justification is not attempted.  Hence \raggedright.  
   
   \def\chapheading{\chapbreak \parsearg\chapheadingzzz}
   \def\chapheadingzzz#1{%
     \vbox{\chapfonts \raggedtitlesettings #1\par}%
     \nobreak\bigskip \nobreak
     \suppressfirstparagraphindent
   }
   
 \def\majorheading{\parsearg\majorheadingzzz}  
 \def\majorheadingzzz #1{%  
 {\advance\chapheadingskip by 10pt \chapbreak }%  
 {\chapfonts \vbox{\hyphenpenalty=10000\tolerance=5000  
                   \parindent=0pt\raggedright  
                   \rm #1\hfill}}\bigskip \par\penalty 200}  
   
 \def\chapheading{\parsearg\chapheadingzzz}  
 \def\chapheadingzzz #1{\chapbreak %  
 {\chapfonts \vbox{\hyphenpenalty=10000\tolerance=5000  
                   \parindent=0pt\raggedright  
                   \rm #1\hfill}}\bigskip \par\penalty 200}  
   
 % @heading, @subheading, @subsubheading.  % @heading, @subheading, @subsubheading.
 \def\heading{\parsearg\plainsecheading}  \parseargdef\heading{\sectionheading{#1}{sec}{Yomitfromtoc}{}
 \def\subheading{\parsearg\plainsubsecheading}    \suppressfirstparagraphindent}
 \def\subsubheading{\parsearg\plainsubsubsecheading}  \parseargdef\subheading{\sectionheading{#1}{subsec}{Yomitfromtoc}{}
     \suppressfirstparagraphindent}
   \parseargdef\subsubheading{\sectionheading{#1}{subsubsec}{Yomitfromtoc}{}
     \suppressfirstparagraphindent}
   
 % These macros generate a chapter, section, etc. heading only  % These macros generate a chapter, section, etc. heading only
 % (including whitespace, linebreaking, etc. around it),  % (including whitespace, linebreaking, etc. around it),
 % given all the information in convenient, parsed form.  % given all the information in convenient, parsed form.
   
 %%% Args are the skip and penalty (usually negative)  % Args are the skip and penalty (usually negative)
 \def\dobreak#1#2{\par\ifdim\lastskip<#1\removelastskip\penalty#2\vskip#1\fi}  \def\dobreak#1#2{\par\ifdim\lastskip<#1\removelastskip\penalty#2\vskip#1\fi}
   
 \def\setchapterstyle #1 {\csname CHAPF#1\endcsname}  
   
 %%% Define plain chapter starts, and page on/off switching for it  
 % Parameter controlling skip before chapter headings (if needed)  % Parameter controlling skip before chapter headings (if needed)
   
 \newskip\chapheadingskip  \newskip\chapheadingskip
   
   % Define plain chapter starts, and page on/off switching for it.
 \def\chapbreak{\dobreak \chapheadingskip {-4000}}  \def\chapbreak{\dobreak \chapheadingskip {-4000}}
   
   % Start a new page
 \def\chappager{\par\vfill\supereject}  \def\chappager{\par\vfill\supereject}
 \def\chapoddpage{\chappager \ifodd\pageno \else \hbox to 0pt{} \chappager\fi}  
   
 \def\setchapternewpage #1 {\csname CHAPPAG#1\endcsname}  % \chapoddpage - start on an odd page for a new chapter
   % Because \domark is called before \chapoddpage, the filler page will
   % get the headings for the next chapter, which is wrong.  But we don't
   % care -- we just disable all headings on the filler page.
   \def\chapoddpage{%
     \chappager
     \ifodd\pageno \else
       \begingroup
         \headingsoff
         \null
         \chappager
       \endgroup
     \fi
   }
   
   \parseargdef\setchapternewpage{\csname CHAPPAG#1\endcsname}
   
 \def\CHAPPAGoff{%  \def\CHAPPAGoff{%
 \global\let\contentsalignmacro = \chappager  \global\let\contentsalignmacro = \chappager
 \global\let\pchapsepmacro=\chapbreak  \global\let\pchapsepmacro=\chapbreak
Line 3637  width0pt\relax} \fi
Line 6502  width0pt\relax} \fi
 \global\let\pagealignmacro=\chappager  \global\let\pagealignmacro=\chappager
 \global\def\HEADINGSon{\HEADINGSsingle}}  \global\def\HEADINGSon{\HEADINGSsingle}}
   
 \def\CHAPPAGodd{  \def\CHAPPAGodd{%
 \global\let\contentsalignmacro = \chapoddpage  \global\let\contentsalignmacro = \chapoddpage
 \global\let\pchapsepmacro=\chapoddpage  \global\let\pchapsepmacro=\chapoddpage
 \global\let\pagealignmacro=\chapoddpage  \global\let\pagealignmacro=\chapoddpage
Line 3645  width0pt\relax} \fi
Line 6510  width0pt\relax} \fi
   
 \CHAPPAGon  \CHAPPAGon
   
 \def\CHAPFplain{  % \chapmacro - Chapter opening.
 \global\let\chapmacro=\chfplain  %
 \global\let\unnumbchapmacro=\unnchfplain  % #1 is the text, #2 is the section type (Ynumbered, Ynothing,
 \global\let\centerchapmacro=\centerchfplain}  % Yappendix, Yomitfromtoc), #3 the chapter number.
   % Not used for @heading series.
 % Plain chapter opening.  %
 % #1 is the text, #2 the chapter number or empty if unnumbered.  % To test against our argument.
 \def\chfplain#1#2{%  \def\Ynothingkeyword{Ynothing}
   \def\Yappendixkeyword{Yappendix}
   \def\Yomitfromtockeyword{Yomitfromtoc}
   %
   \def\chapmacro#1#2#3{%
     \expandafter\ifx\thisenv\titlepage\else
       \checkenv{}% chapters, etc., should not start inside an environment.
     \fi
     % FIXME: \chapmacro is currently called from inside \titlepage when
     % \setcontentsaftertitlepage to print the "Table of Contents" heading, but
     % this should probably be done by \sectionheading with an option to print
     % in chapter size.
     %
     % Insert the first mark before the heading break (see notes for \domark).
     \let\prevchapterdefs=\lastchapterdefs
     \let\prevsectiondefs=\lastsectiondefs
     \gdef\lastsectiondefs{\gdef\thissectionname{}\gdef\thissectionnum{}%
                           \gdef\thissection{}}%
     %
     \def\temptype{#2}%
     \ifx\temptype\Ynothingkeyword
       \gdef\lastchapterdefs{\gdef\thischaptername{#1}\gdef\thischapternum{}%
                             \gdef\thischapter{\thischaptername}}%
     \else\ifx\temptype\Yomitfromtockeyword
       \gdef\lastchapterdefs{\gdef\thischaptername{#1}\gdef\thischapternum{}%
                             \gdef\thischapter{}}%
     \else\ifx\temptype\Yappendixkeyword
       \toks0={#1}%
       \xdef\lastchapterdefs{%
         \gdef\noexpand\thischaptername{\the\toks0}%
         \gdef\noexpand\thischapternum{\appendixletter}%
         % \noexpand\putwordAppendix avoids expanding indigestible
         % commands in some of the translations.
         \gdef\noexpand\thischapter{\noexpand\putwordAppendix{}
                                    \noexpand\thischapternum:
                                    \noexpand\thischaptername}%
       }%
     \else
       \toks0={#1}%
       \xdef\lastchapterdefs{%
         \gdef\noexpand\thischaptername{\the\toks0}%
         \gdef\noexpand\thischapternum{\the\chapno}%
         % \noexpand\putwordChapter avoids expanding indigestible
         % commands in some of the translations.
         \gdef\noexpand\thischapter{\noexpand\putwordChapter{}
                                    \noexpand\thischapternum:
                                    \noexpand\thischaptername}%
       }%
     \fi\fi\fi
     %
     % Output the mark.  Pass it through \safewhatsit, to take care of
     % the preceding space.
     \safewhatsit\domark
     %
     % Insert the chapter heading break.
   \pchapsepmacro    \pchapsepmacro
     %
     % Now the second mark, after the heading break.  No break points
     % between here and the heading.
     \let\prevchapterdefs=\lastchapterdefs
     \let\prevsectiondefs=\lastsectiondefs
     \domark
     %
   {%    {%
     \chapfonts \rm      \chapfonts \rmisbold
     \def\chapnum{#2}%      \let\footnote=\errfootnoteheading % give better error message
     \setbox0 = \hbox{#2\ifx\chapnum\empty\else\enspace\fi}%      %
     \vbox{\hyphenpenalty=10000 \tolerance=5000 \parindent=0pt \raggedright      % Have to define \lastsection before calling \donoderef, because the
           \hangindent = \wd0 \centerparametersmaybe      % xref code eventually uses it.  On the other hand, it has to be called
       % after \pchapsepmacro, or the headline will change too soon.
       \gdef\lastsection{#1}%
       %
       % Only insert the separating space if we have a chapter/appendix
       % number, and don't print the unnumbered ``number''.
       \ifx\temptype\Ynothingkeyword
         \setbox0 = \hbox{}%
         \def\toctype{unnchap}%
       \else\ifx\temptype\Yomitfromtockeyword
         \setbox0 = \hbox{}% contents like unnumbered, but no toc entry
         \def\toctype{omit}%
       \else\ifx\temptype\Yappendixkeyword
         \setbox0 = \hbox{\putwordAppendix{} #3\enspace}%
         \def\toctype{app}%
       \else
         \setbox0 = \hbox{#3\enspace}%
         \def\toctype{numchap}%
       \fi\fi\fi
       %
       % Write the toc entry for this chapter.  Must come before the
       % \donoderef, because we include the current node name in the toc
       % entry, and \donoderef resets it to empty.
       \writetocentry{\toctype}{#1}{#3}%
       %
       % For pdftex, we have to write out the node definition (aka, make
       % the pdfdest) after any page break, but before the actual text has
       % been typeset.  If the destination for the pdf outline is after the
       % text, then jumping from the outline may wind up with the text not
       % being visible, for instance under high magnification.
       \donoderef{#2}%
       %
       % Typeset the actual heading.
       \nobreak % Avoid page breaks at the interline glue.
       \vbox{\raggedtitlesettings \hangindent=\wd0 \centerparametersmaybe
           \unhbox0 #1\par}%            \unhbox0 #1\par}%
   }%    }%
   \nobreak\bigskip % no page break after a chapter title    \nobreak\bigskip % no page break after a chapter title
   \nobreak    \nobreak
 }  }
   
 % Plain opening for unnumbered.  
 \def\unnchfplain#1{\chfplain{#1}{}}  
   
 % @centerchap -- centered and unnumbered.  % @centerchap -- centered and unnumbered.
 \let\centerparametersmaybe = \relax  \let\centerparametersmaybe = \relax
 \def\centerchfplain#1{{%  \def\centerparameters{%
   \def\centerparametersmaybe{%    \advance\rightskip by 3\rightskip
     \advance\rightskip by 3\rightskip    \leftskip = \rightskip
     \leftskip = \rightskip    \parfillskip = 0pt
     \parfillskip = 0pt  
   }%  
   \chfplain{#1}{}%  
 }}  
   
 \CHAPFplain % The default  
   
 \def\unnchfopen #1{%  
 \chapoddpage {\chapfonts \vbox{\hyphenpenalty=10000\tolerance=5000  
                        \parindent=0pt\raggedright  
                        \rm #1\hfill}}\bigskip \par\nobreak  
 }  }
   
 \def\chfopen #1#2{\chapoddpage {\chapfonts  
 \vbox to 3in{\vfil \hbox to\hsize{\hfil #2} \hbox to\hsize{\hfil #1} \vfil}}%  
 \par\penalty 5000 %  
 }  
   
 \def\centerchfopen #1{%  % Section titles.  These macros combine the section number parts and
 \chapoddpage {\chapfonts \vbox{\hyphenpenalty=10000\tolerance=5000  % call the generic \sectionheading to do the printing.
                        \parindent=0pt  %
                        \hfill {\rm #1}\hfill}}\bigskip \par\nobreak  
 }  
   
 \def\CHAPFopen{  
 \global\let\chapmacro=\chfopen  
 \global\let\unnumbchapmacro=\unnchfopen  
 \global\let\centerchapmacro=\centerchfopen}  
   
   
 % Section titles.  
 \newskip\secheadingskip  \newskip\secheadingskip
 \def\secheadingbreak{\dobreak \secheadingskip {-1000}}  \def\secheadingbreak{\dobreak \secheadingskip{-1000}}
 \def\secheading#1#2#3{\sectionheading{sec}{#2.#3}{#1}}  
 \def\plainsecheading#1{\sectionheading{sec}{}{#1}}  
   
 % Subsection titles.  % Subsection titles.
 \newskip \subsecheadingskip  \newskip\subsecheadingskip
 \def\subsecheadingbreak{\dobreak \subsecheadingskip {-500}}  \def\subsecheadingbreak{\dobreak \subsecheadingskip{-500}}
 \def\subsecheading#1#2#3#4{\sectionheading{subsec}{#2.#3.#4}{#1}}  
 \def\plainsubsecheading#1{\sectionheading{subsec}{}{#1}}  
   
 % Subsubsection titles.  % Subsubsection titles.
 \let\subsubsecheadingskip = \subsecheadingskip  \def\subsubsecheadingskip{\subsecheadingskip}
 \let\subsubsecheadingbreak = \subsecheadingbreak  \def\subsubsecheadingbreak{\subsecheadingbreak}
 \def\subsubsecheading#1#2#3#4#5{\sectionheading{subsubsec}{#2.#3.#4.#5}{#1}}  
 \def\plainsubsubsecheading#1{\sectionheading{subsubsec}{}{#1}}  
   
   
 % Print any size section title.  % Print any size, any type, section title.
 %  %
 % #1 is the section type (sec/subsec/subsubsec), #2 is the section  % #1 is the text of the title,
 % number (maybe empty), #3 the text.  % #2 is the section level (sec/subsec/subsubsec),
 \def\sectionheading#1#2#3{%  % #3 is the section type (Ynumbered, Ynothing, Yappendix, Yomitfromtoc),
   % #4 is the section number.
   %
   \def\seckeyword{sec}
   %
   \def\sectionheading#1#2#3#4{%
   {%    {%
     \expandafter\advance\csname #1headingskip\endcsname by \parskip      \def\sectionlevel{#2}%
     \csname #1headingbreak\endcsname      \def\temptype{#3}%
   }%      %
   {%      % It is ok for the @heading series commands to appear inside an
       % environment (it's been historically allowed, though the logic is
       % dubious), but not the others.
       \ifx\temptype\Yomitfromtockeyword\else
         \checkenv{}% non-@*heading should not be in an environment.
       \fi
       \let\footnote=\errfootnoteheading
       %
     % Switch to the right set of fonts.      % Switch to the right set of fonts.
     \csname #1fonts\endcsname \rm      \csname #2fonts\endcsname \rmisbold
     %      %
     % Only insert the separating space if we have a section number.      % Insert first mark before the heading break (see notes for \domark).
     \def\secnum{#2}%      \let\prevsectiondefs=\lastsectiondefs
     \setbox0 = \hbox{#2\ifx\secnum\empty\else\enspace\fi}%      \ifx\temptype\Ynothingkeyword
         \ifx\sectionlevel\seckeyword
           \gdef\lastsectiondefs{\gdef\thissectionname{#1}\gdef\thissectionnum{}%
                                 \gdef\thissection{\thissectionname}}%
         \fi
       \else\ifx\temptype\Yomitfromtockeyword
         % Don't redefine \thissection.
       \else\ifx\temptype\Yappendixkeyword
         \ifx\sectionlevel\seckeyword
           \toks0={#1}%
           \xdef\lastsectiondefs{%
             \gdef\noexpand\thissectionname{\the\toks0}%
             \gdef\noexpand\thissectionnum{#4}%
             % \noexpand\putwordSection avoids expanding indigestible
             % commands in some of the translations.
             \gdef\noexpand\thissection{\noexpand\putwordSection{}
                                        \noexpand\thissectionnum:
                                        \noexpand\thissectionname}%
           }%
         \fi
       \else
         \ifx\sectionlevel\seckeyword
           \toks0={#1}%
           \xdef\lastsectiondefs{%
             \gdef\noexpand\thissectionname{\the\toks0}%
             \gdef\noexpand\thissectionnum{#4}%
             % \noexpand\putwordSection avoids expanding indigestible
             % commands in some of the translations.
             \gdef\noexpand\thissection{\noexpand\putwordSection{}
                                        \noexpand\thissectionnum:
                                        \noexpand\thissectionname}%
           }%
         \fi
       \fi\fi\fi
     %      %
     \vbox{\hyphenpenalty=10000 \tolerance=5000 \parindent=0pt \raggedright      % Go into vertical mode.  Usually we'll already be there, but we
           \hangindent = \wd0 % zero if no section number      % don't want the following whatsit to end up in a preceding paragraph
           \unhbox0 #3}%      % if the document didn't happen to have a blank line.
       \par
       %
       % Output the mark.  Pass it through \safewhatsit, to take care of
       % the preceding space.
       \safewhatsit\domark
       %
       % Insert space above the heading.
       \csname #2headingbreak\endcsname
       %
       % Now the second mark, after the heading break.  No break points
       % between here and the heading.
       \global\let\prevsectiondefs=\lastsectiondefs
       \domark
       %
       % Only insert the space after the number if we have a section number.
       \ifx\temptype\Ynothingkeyword
         \setbox0 = \hbox{}%
         \def\toctype{unn}%
         \gdef\lastsection{#1}%
       \else\ifx\temptype\Yomitfromtockeyword
         % for @headings -- no section number, don't include in toc,
         % and don't redefine \lastsection.
         \setbox0 = \hbox{}%
         \def\toctype{omit}%
         \let\sectionlevel=\empty
       \else\ifx\temptype\Yappendixkeyword
         \setbox0 = \hbox{#4\enspace}%
         \def\toctype{app}%
         \gdef\lastsection{#1}%
       \else
         \setbox0 = \hbox{#4\enspace}%
         \def\toctype{num}%
         \gdef\lastsection{#1}%
       \fi\fi\fi
       %
       % Write the toc entry (before \donoderef).  See comments in \chapmacro.
       \writetocentry{\toctype\sectionlevel}{#1}{#4}%
       %
       % Write the node reference (= pdf destination for pdftex).
       % Again, see comments in \chapmacro.
       \donoderef{#3}%
       %
       % Interline glue will be inserted when the vbox is completed.
       % That glue will be a valid breakpoint for the page, since it'll be
       % preceded by a whatsit (usually from the \donoderef, or from the
       % \writetocentry if there was no node).  We don't want to allow that
       % break, since then the whatsits could end up on page n while the
       % section is on page n+1, thus toc/etc. are wrong.  Debian bug 276000.
       \nobreak
       %
       % Output the actual section heading.
       \vbox{\hyphenpenalty=10000 \tolerance=5000 \parindent=0pt \ptexraggedright
             \hangindent=\wd0  % zero if no section number
             \unhbox0 #1}%
   }%    }%
   \ifdim\parskip<10pt \nobreak\kern10pt\nobreak\kern-\parskip\fi \nobreak    % Add extra space after the heading -- half of whatever came above it.
     % Don't allow stretch, though.
     \kern .5 \csname #2headingskip\endcsname
     %
     % Do not let the kern be a potential breakpoint, as it would be if it
     % was followed by glue.
     \nobreak
     %
     % We'll almost certainly start a paragraph next, so don't let that
     % glue accumulate.  (Not a breakpoint because it's preceded by a
     % discardable item.)  However, when a paragraph is not started next
     % (\startdefun, \cartouche, \center, etc.), this needs to be wiped out
     % or the negative glue will cause weirdly wrong output, typically
     % obscuring the section heading with something else.
     \vskip-\parskip
     %
     % This is so the last item on the main vertical list is a known
     % \penalty > 10000, so \startdefun, etc., can recognize the situation
     % and do the needful.
     \penalty 10001
 }  }
   
   
Line 3754  width0pt\relax} \fi
Line 6798  width0pt\relax} \fi
 \newwrite\tocfile  \newwrite\tocfile
   
 % Write an entry to the toc file, opening it if necessary.  % Write an entry to the toc file, opening it if necessary.
 % Called from @chapter, etc.  We supply {\folio} at the end of the  % Called from @chapter, etc.
 % argument, which will end up as the last argument to the \...entry macro.  
 %  %
 % We open the .toc file here instead of at @setfilename or any other  % Example usage: \writetocentry{sec}{Section Name}{\the\chapno.\the\secno}
 % given time so that @contents can be put in the document anywhere.  % We append the current node name (if any) and page number as additional
   % arguments for the \{chap,sec,...}entry macros which will eventually
   % read this.  The node name is used in the pdf outlines as the
   % destination to jump to.
 %  %
   % We open the .toc file for writing here instead of at @setfilename (or
   % any other fixed time) so that @contents can be anywhere in the document.
   % But if #1 is `omit', then we don't do anything.  This is used for the
   % table of contents chapter openings themselves.
   %
 \newif\iftocfileopened  \newif\iftocfileopened
 \def\writetocentry#1{%  \def\omitkeyword{omit}%
   \iftocfileopened\else  %
     \immediate\openout\tocfile = \jobname.toc  \def\writetocentry#1#2#3{%
     \global\tocfileopenedtrue    \edef\writetoctype{#1}%
     \ifx\writetoctype\omitkeyword \else
       \iftocfileopened\else
         \immediate\openout\tocfile = \jobname.toc
         \global\tocfileopenedtrue
       \fi
       %
       \iflinks
         {\atdummies
          \edef\temp{%
            \write\tocfile{@#1entry{#2}{#3}{\lastnode}{\noexpand\folio}}}%
          \temp
         }%
       \fi
   \fi    \fi
   \iflinks \write\tocfile{#1{\folio}}\fi    %
     % Tell \shipout to create a pdf destination on each page, if we're
     % writing pdf.  These are used in the table of contents.  We can't
     % just write one on every page because the title pages are numbered
     % 1 and 2 (the page numbers aren't printed), and so are the first
     % two pages of the document.  Thus, we'd have two destinations named
     % `1', and two named `2'.
     \ifpdf
       \global\pdfmakepagedesttrue
     \else
       \ifx\XeTeXrevision\thisisundefined
       \else
         \global\pdfmakepagedesttrue
       \fi
     \fi
 }  }
   
   
   % These characters do not print properly in the Computer Modern roman
   % fonts, so we must take special care.  This is more or less redundant
   % with the Texinfo input format setup at the end of this file.
   %
   \def\activecatcodes{%
     \catcode`\"=\active
     \catcode`\$=\active
     \catcode`\<=\active
     \catcode`\>=\active
     \catcode`\\=\active
     \catcode`\^=\active
     \catcode`\_=\active
     \catcode`\|=\active
     \catcode`\~=\active
   }
   
   
   % Read the toc file, which is essentially Texinfo input.
   \def\readtocfile{%
     \setupdatafile
     \activecatcodes
     \input \tocreadfilename
   }
   
 \newskip\contentsrightmargin \contentsrightmargin=1in  \newskip\contentsrightmargin \contentsrightmargin=1in
 \newcount\savepageno  \newcount\savepageno
 \newcount\lastnegativepageno \lastnegativepageno = -1  \newcount\lastnegativepageno \lastnegativepageno = -1
   
 % Finish up the main text and prepare to read what we've written  % Prepare to read what we've written to \tocfile.
 % to \tocfile.  
 %  %
 \def\startcontents#1{%  \def\startcontents#1{%
    % If @setchapternewpage on, and @headings double, the contents should    % If @setchapternewpage on, and @headings double, the contents should
    % start on an odd page, unlike chapters.  Thus, we maintain    % start on an odd page, unlike chapters.  Thus, we maintain
    % \contentsalignmacro in parallel with \pagealignmacro.    % \contentsalignmacro in parallel with \pagealignmacro.
    % From: Torbjorn Granlund <tege@matematik.su.se>    % From: Torbjorn Granlund <tege@matematik.su.se>
    \contentsalignmacro    \contentsalignmacro
    \immediate\closeout\tocfile    \immediate\closeout\tocfile
    %    %
    % Don't need to put `Contents' or `Short Contents' in the headline.    % Don't need to put `Contents' or `Short Contents' in the headline.
    % It is abundantly clear what they are.    % It is abundantly clear what they are.
    \unnumbchapmacro{#1}\def\thischapter{}%    \chapmacro{#1}{Yomitfromtoc}{}%
    \savepageno = \pageno    %
    \begingroup                  % Set up to handle contents files properly.    \savepageno = \pageno
       \catcode`\\=0  \catcode`\{=1  \catcode`\}=2  \catcode`\@=11    \begingroup                  % Set up to handle contents files properly.
       % We can't do this, because then an actual ^ in a section      \raggedbottom              % Worry more about breakpoints than the bottom.
       % title fails, e.g., @chapter ^ -- exponentiation.  --karl, 9jul97.      \entryrightmargin=\contentsrightmargin % Don't use the full line length.
       %\catcode`\^=7 % to see ^^e4 as \"a etc. juha@piuha.ydi.vtt.fi      %
       \raggedbottom             % Worry more about breakpoints than the bottom.      % Roman numerals for page numbers.
       \advance\hsize by -\contentsrightmargin % Don't use the full line length.      \ifnum \pageno>0 \global\pageno = \lastnegativepageno \fi
       %  
       % Roman numerals for page numbers.  
       \ifnum \pageno>0 \pageno = \lastnegativepageno \fi  
 }  }
   
   % redefined for the two-volume lispref.  We always output on
   % \jobname.toc even if this is redefined.
   %
   \def\tocreadfilename{\jobname.toc}
   
 % Normal (long) toc.  % Normal (long) toc.
   %
 \def\contents{%  \def\contents{%
    \startcontents{\putwordTOC}%    \startcontents{\putwordTOC}%
      \openin 1 \jobname.toc      \openin 1 \tocreadfilename\space
      \ifeof 1 \else      \ifeof 1 \else
        \closein 1        \readtocfile
        \input \jobname.toc      \fi
      \fi      \vfill \eject
      \vfill \eject      \contentsalignmacro % in case @setchapternewpage odd is in effect
      \contentsalignmacro % in case @setchapternewpage odd is in effect      \ifeof 1 \else
      \pdfmakeoutlines        \pdfmakeoutlines
    \endgroup      \fi
    \lastnegativepageno = \pageno      \closein 1
    \pageno = \savepageno    \endgroup
     \lastnegativepageno = \pageno
     \global\pageno = \savepageno
 }  }
   
 % And just the chapters.  % And just the chapters.
 \def\summarycontents{%  \def\summarycontents{%
    \startcontents{\putwordShortTOC}%    \startcontents{\putwordShortTOC}%
       %      %
       \let\chapentry = \shortchapentry      \let\partentry = \shortpartentry
       \let\unnumbchapentry = \shortunnumberedentry      \let\numchapentry = \shortchapentry
       % We want a true roman here for the page numbers.      \let\appentry = \shortchapentry
       \secfonts      \let\unnchapentry = \shortunnchapentry
       \let\rm=\shortcontrm \let\bf=\shortcontbf \let\sl=\shortcontsl      % We want a true roman here for the page numbers.
       \rm      \secfonts
       \hyphenpenalty = 10000      \let\rm=\shortcontrm \let\bf=\shortcontbf
       \advance\baselineskip by 1pt % Open it up a little.      \let\sl=\shortcontsl \let\tt=\shortconttt
       \def\secentry ##1##2##3##4{}      \rm
       \def\unnumbsecentry ##1##2{}      \hyphenpenalty = 10000
       \def\subsecentry ##1##2##3##4##5{}      \advance\baselineskip by 1pt % Open it up a little.
       \def\unnumbsubsecentry ##1##2{}      \def\numsecentry##1##2##3##4{}
       \def\subsubsecentry ##1##2##3##4##5##6{}      \let\appsecentry = \numsecentry
       \def\unnumbsubsubsecentry ##1##2{}      \let\unnsecentry = \numsecentry
       \openin 1 \jobname.toc      \let\numsubsecentry = \numsecentry
       \ifeof 1 \else      \let\appsubsecentry = \numsecentry
         \closein 1      \let\unnsubsecentry = \numsecentry
         \input \jobname.toc      \let\numsubsubsecentry = \numsecentry
       \fi      \let\appsubsubsecentry = \numsecentry
      \vfill \eject      \let\unnsubsubsecentry = \numsecentry
      \contentsalignmacro % in case @setchapternewpage odd is in effect      \openin 1 \tocreadfilename\space
    \endgroup      \ifeof 1 \else
    \lastnegativepageno = \pageno        \readtocfile
    \pageno = \savepageno      \fi
       \closein 1
       \vfill \eject
       \contentsalignmacro % in case @setchapternewpage odd is in effect
     \endgroup
     \lastnegativepageno = \pageno
     \global\pageno = \savepageno
 }  }
 \let\shortcontents = \summarycontents  \let\shortcontents = \summarycontents
   
 \ifpdf  % Typeset the label for a chapter or appendix for the short contents.
   \pdfcatalog{/PageMode /UseOutlines}%  % The arg is, e.g., `A' for an appendix, or `3' for a chapter.
 \fi  %
   \def\shortchaplabel#1{%
     % This space should be enough, since a single number is .5em, and the
     % widest letter (M) is 1em, at least in the Computer Modern fonts.
     % But use \hss just in case.
     % (This space doesn't include the extra space that gets added after
     % the label; that gets put in by \shortchapentry above.)
     %
     % We'd like to right-justify chapter numbers, but that looks strange
     % with appendix letters.  And right-justifying numbers and
     % left-justifying letters looks strange when there is less than 10
     % chapters.  Have to read the whole toc once to know how many chapters
     % there are before deciding ...
     \hbox to 1em{#1\hss}%
   }
   
 % These macros generate individual entries in the table of contents.  % These macros generate individual entries in the table of contents.
 % The first argument is the chapter or section name.  % The first argument is the chapter or section name.
 % The last argument is the page number.  % The last argument is the page number.
 % The arguments in between are the chapter number, section number, ...  % The arguments in between are the chapter number, section number, ...
   
 % Chapter-level things, for both the long and short contents.  % Parts, in the main contents.  Replace the part number, which doesn't
 \def\chapentry#1#2#3{\dochapentry{#2\labelspace#1}{#3}}  % exist, with an empty box.  Let's hope all the numbers have the same width.
   % Also ignore the page number, which is conventionally not printed.
   \def\numeralbox{\setbox0=\hbox{8}\hbox to \wd0{\hfil}}
   \def\partentry#1#2#3#4{\dochapentry{\numeralbox\labelspace#1}{}}
   %
   % Parts, in the short toc.
   \def\shortpartentry#1#2#3#4{%
     \penalty-300
     \vskip.5\baselineskip plus.15\baselineskip minus.1\baselineskip
     \shortchapentry{{\bf #1}}{\numeralbox}{}{}%
   }
   
 % See comments in \dochapentry re vbox and related settings  % Chapters, in the main contents.
 \def\shortchapentry#1#2#3{%  \def\numchapentry#1#2#3#4{\dochapentry{#2\labelspace#1}{#4}}
   \tocentry{\shortchaplabel{#2}\labelspace #1}{\doshortpageno\bgroup#3\egroup}%  
   % Chapters, in the short toc.
   % See comments in \dochapentry re vbox and related settings.
   \def\shortchapentry#1#2#3#4{%
     \tocentry{\shortchaplabel{#2}\labelspace #1}{\doshortpageno\bgroup#4\egroup}%
 }  }
   
 % Typeset the label for a chapter or appendix for the short contents.  % Appendices, in the main contents.
 % The arg is, e.g. `Appendix A' for an appendix, or `3' for a chapter.  % Need the word Appendix, and a fixed-size box.
 % We could simplify the code here by writing out an \appendixentry  
 % command in the toc file for appendices, instead of using \chapentry  
 % for both, but it doesn't seem worth it.  
 %  %
 \newdimen\shortappendixwidth  \def\appendixbox#1{%
     % We use M since it's probably the widest letter.
     \setbox0 = \hbox{\putwordAppendix{} M}%
     \hbox to \wd0{\putwordAppendix{} #1\hss}}
 %  %
 \def\shortchaplabel#1{%  \def\appentry#1#2#3#4{\dochapentry{\appendixbox{#2}\hskip.7em#1}{#4}}
   % Compute width of word "Appendix", may change with language.  
   \setbox0 = \hbox{\shortcontrm \putwordAppendix}%  
   \shortappendixwidth = \wd0  
   %  
   % We typeset #1 in a box of constant width, regardless of the text of  
   % #1, so the chapter titles will come out aligned.  
   \setbox0 = \hbox{#1}%  
   \dimen0 = \ifdim\wd0 > \shortappendixwidth \shortappendixwidth \else 0pt \fi  
   %  
   % This space should be plenty, since a single number is .5em, and the  
   % widest letter (M) is 1em, at least in the Computer Modern fonts.  
   % (This space doesn't include the extra space that gets added after  
   % the label; that gets put in by \shortchapentry above.)  
   \advance\dimen0 by 1.1em  
   \hbox to \dimen0{#1\hfil}%  
 }  
   
 \def\unnumbchapentry#1#2{\dochapentry{#1}{#2}}  % Unnumbered chapters.
 \def\shortunnumberedentry#1#2{\tocentry{#1}{\doshortpageno\bgroup#2\egroup}}  \def\unnchapentry#1#2#3#4{\dochapentry{#1}{#4}}
   \def\shortunnchapentry#1#2#3#4{\tocentry{#1}{\doshortpageno\bgroup#4\egroup}}
   
 % Sections.  % Sections.
 \def\secentry#1#2#3#4{\dosecentry{#2.#3\labelspace#1}{#4}}  \def\numsecentry#1#2#3#4{\dosecentry{#2\labelspace#1}{#4}}
 \def\unnumbsecentry#1#2{\dosecentry{#1}{#2}}  \let\appsecentry=\numsecentry
   \def\unnsecentry#1#2#3#4{\dosecentry{#1}{#4}}
   
 % Subsections.  % Subsections.
 \def\subsecentry#1#2#3#4#5{\dosubsecentry{#2.#3.#4\labelspace#1}{#5}}  \def\numsubsecentry#1#2#3#4{\dosubsecentry{#2\labelspace#1}{#4}}
 \def\unnumbsubsecentry#1#2{\dosubsecentry{#1}{#2}}  \let\appsubsecentry=\numsubsecentry
   \def\unnsubsecentry#1#2#3#4{\dosubsecentry{#1}{#4}}
   
 % And subsubsections.  % And subsubsections.
 \def\subsubsecentry#1#2#3#4#5#6{%  \def\numsubsubsecentry#1#2#3#4{\dosubsubsecentry{#2\labelspace#1}{#4}}
   \dosubsubsecentry{#2.#3.#4.#5\labelspace#1}{#6}}  \let\appsubsubsecentry=\numsubsubsecentry
 \def\unnumbsubsubsecentry#1#2{\dosubsubsecentry{#1}{#2}}  \def\unnsubsubsecentry#1#2#3#4{\dosubsubsecentry{#1}{#4}}
   
 % This parameter controls the indentation of the various levels.  % This parameter controls the indentation of the various levels.
 \newdimen\tocindent \tocindent = 3pc  % Same as \defaultparindent.
   \newdimen\tocindent \tocindent = 15pt
   
 % Now for the actual typesetting. In all these, #1 is the text and #2 is the  % Now for the actual typesetting. In all these, #1 is the text and #2 is the
 % page number.  % page number.
Line 3918  width0pt\relax} \fi
Line 7046  width0pt\relax} \fi
 \def\dochapentry#1#2{%  \def\dochapentry#1#2{%
    \penalty-300 \vskip1\baselineskip plus.33\baselineskip minus.25\baselineskip     \penalty-300 \vskip1\baselineskip plus.33\baselineskip minus.25\baselineskip
    \begingroup     \begingroup
        % Move the page numbers slightly to the right
        \advance\entryrightmargin by -0.05em
      \chapentryfonts       \chapentryfonts
      \tocentry{#1}{\dopageno\bgroup#2\egroup}%       \tocentry{#1}{\dopageno\bgroup#2\egroup}%
    \endgroup     \endgroup
Line 3939  width0pt\relax} \fi
Line 7069  width0pt\relax} \fi
   \tocentry{#1}{\dopageno\bgroup#2\egroup}%    \tocentry{#1}{\dopageno\bgroup#2\egroup}%
 \endgroup}  \endgroup}
   
 % Final typesetting of a toc entry; we use the same \entry macro as for  % We use the same \entry macro as for the index entries.
 % the index entries, but we want to suppress hyphenation here.  (We  \let\tocentry = \entry
 % can't do that in the \entry macro, since index entries might consist  
 % of hyphenated-identifiers-that-do-not-fit-on-a-line-and-nothing-else.)  
 \def\tocentry#1#2{\begingroup  
   \vskip 0pt plus1pt % allow a little stretch for the sake of nice page breaks  
   % Do not use \turnoffactive in these arguments.  Since the toc is  
   % typeset in cmr, so characters such as _ would come out wrong; we  
   % have to do the usual translation tricks.  
   \entry{#1}{#2}%  
 \endgroup}  
   
 % Space between chapter (or whatever) number and the title.  % Space between chapter (or whatever) number and the title.
 \def\labelspace{\hskip1em \relax}  \def\labelspace{\hskip1em \relax}
Line 3959  width0pt\relax} \fi
Line 7080  width0pt\relax} \fi
   
 \def\chapentryfonts{\secfonts \rm}  \def\chapentryfonts{\secfonts \rm}
 \def\secentryfonts{\textfonts}  \def\secentryfonts{\textfonts}
 \let\subsecentryfonts = \textfonts  \def\subsecentryfonts{\textfonts}
 \let\subsubsecentryfonts = \textfonts  \def\subsubsecentryfonts{\textfonts}
   
   
 \message{environments,}  \message{environments,}
 % @foo ... @end foo.  % @foo ... @end foo.
   
 % Since these characters are used in examples, it should be an even number of  % @tex ... @end tex    escapes into raw TeX temporarily.
 % \tt widths. Each \tt character is 1en, so two makes it 1em.  
 % Furthermore, these definitions must come after we define our fonts.  
 \newbox\dblarrowbox    \newbox\longdblarrowbox  
 \newbox\pushcharbox    \newbox\bullbox  
 \newbox\equivbox       \newbox\errorbox  
   
 %{\tentt  
 %\global\setbox\dblarrowbox = \hbox to 1em{\hfil$\Rightarrow$\hfil}  
 %\global\setbox\longdblarrowbox = \hbox to 1em{\hfil$\mapsto$\hfil}  
 %\global\setbox\pushcharbox = \hbox to 1em{\hfil$\dashv$\hfil}  
 %\global\setbox\equivbox = \hbox to 1em{\hfil$\ptexequiv$\hfil}  
 % Adapted from the manmac format (p.420 of TeXbook)  
 %\global\setbox\bullbox = \hbox to 1em{\kern.15em\vrule height .75ex width .85ex  
 %                                      depth .1ex\hfil}  
 %}  
   
 % @point{}, @result{}, @expansion{}, @print{}, @equiv{}.  
 \def\point{$\star$}  
 \def\result{\leavevmode\raise.15ex\hbox to 1em{\hfil$\Rightarrow$\hfil}}  
 \def\expansion{\leavevmode\raise.1ex\hbox to 1em{\hfil$\mapsto$\hfil}}  
 \def\print{\leavevmode\lower.1ex\hbox to 1em{\hfil$\dashv$\hfil}}  
 \def\equiv{\leavevmode\lower.1ex\hbox to 1em{\hfil$\ptexequiv$\hfil}}  
   
 % Adapted from the TeXbook's \boxit.  
 {\tentt \global\dimen0 = 3em}% Width of the box.  
 \dimen2 = .55pt % Thickness of rules  
 % The text. (`r' is open on the right, `e' somewhat less so on the left.)  
 \setbox0 = \hbox{\kern-.75pt \tensf error\kern-1.5pt}  
   
 \global\setbox\errorbox=\hbox to \dimen0{\hfil  
    \hsize = \dimen0 \advance\hsize by -5.8pt % Space to left+right.  
    \advance\hsize by -2\dimen2 % Rules.  
    \vbox{  
       \hrule height\dimen2  
       \hbox{\vrule width\dimen2 \kern3pt          % Space to left of text.  
          \vtop{\kern2.4pt \box0 \kern2.4pt}% Space above/below.  
          \kern3pt\vrule width\dimen2}% Space to right.  
       \hrule height\dimen2}  
     \hfil}  
   
 % The @error{} command.  
 \def\error{\leavevmode\lower.7ex\copy\errorbox}  
   
 % @tex ... @end tex    escapes into raw Tex temporarily.  
 % One exception: @ is still an escape character, so that @end tex works.  % One exception: @ is still an escape character, so that @end tex works.
 % But \@ or @@ will get a plain tex @ character.  % But \@ or @@ will get a plain @ character.
   
 \def\tex{\begingroup  \envdef\tex{%
     \setupmarkupstyle{tex}%
   \catcode `\\=0 \catcode `\{=1 \catcode `\}=2    \catcode `\\=0 \catcode `\{=1 \catcode `\}=2
   \catcode `\$=3 \catcode `\&=4 \catcode `\#=6    \catcode `\$=3 \catcode `\&=4 \catcode `\#=6
   \catcode `\^=7 \catcode `\_=8 \catcode `\~=13 \let~=\tie    \catcode `\^=7 \catcode `\_=8 \catcode `\~=\active \let~=\tie
   \catcode `\%=14    \catcode `\%=14
   \catcode 43=12 % plus    \catcode `\+=\other
   \catcode`\"=12    \catcode `\"=\other
   \catcode`\==12    \catcode `\|=\other
   \catcode`\|=12    \catcode `\<=\other
   \catcode`\<=12    \catcode `\>=\other
   \catcode`\>=12    \catcode `\`=\other
   \escapechar=`\\    \catcode `\'=\other
   %    %
     % ' is active in math mode (mathcode"8000).  So reset it, and all our
     % other math active characters (just in case), to plain's definitions.
     \mathactive
     %
     % Inverse of the list at the beginning of the file.
   \let\b=\ptexb    \let\b=\ptexb
   \let\bullet=\ptexbullet    \let\bullet=\ptexbullet
   \let\c=\ptexc    \let\c=\ptexc
Line 4036  width0pt\relax} \fi
Line 7119  width0pt\relax} \fi
   \let\equiv=\ptexequiv    \let\equiv=\ptexequiv
   \let\!=\ptexexclam    \let\!=\ptexexclam
   \let\i=\ptexi    \let\i=\ptexi
     \let\indent=\ptexindent
     \let\noindent=\ptexnoindent
   \let\{=\ptexlbrace    \let\{=\ptexlbrace
   \let\+=\tabalign    \let\+=\tabalign
   \let\}=\ptexrbrace    \let\}=\ptexrbrace
     \let\/=\ptexslash
     \let\sp=\ptexsp
   \let\*=\ptexstar    \let\*=\ptexstar
     %\let\sup=\ptexsup % do not redefine, we want @sup to work in math mode
   \let\t=\ptext    \let\t=\ptext
     \expandafter \let\csname top\endcsname=\ptextop  % we've made it outer
     \let\frenchspacing=\plainfrenchspacing
   %    %
   \def\endldots{\mathinner{\ldots\ldots\ldots\ldots}}%    \def\endldots{\mathinner{\ldots\ldots\ldots\ldots}}%
   \def\enddots{\relax\ifmmode\endldots\else$\mathsurround=0pt \endldots\,$\fi}%    \def\enddots{\relax\ifmmode\endldots\else$\mathsurround=0pt \endldots\,$\fi}%
   \def\@{@}%    \def\@{@}%
 \let\Etex=\endgroup}  }
   % There is no need to define \Etex.
   
 % Define @lisp ... @endlisp.  % Define @lisp ... @end lisp.
 % @lisp does a \begingroup so it can rebind things,  % @lisp environment forms a group so it can rebind things,
 % including the definition of @endlisp (which normally is erroneous).  % including the definition of @end lisp (which normally is erroneous).
   
 % Amount to narrow the margins by for @lisp.  % Amount to narrow the margins by for @lisp.
 \newskip\lispnarrowing \lispnarrowing=0.4in  \newskip\lispnarrowing \lispnarrowing=0.4in
Line 4059  width0pt\relax} \fi
Line 7150  width0pt\relax} \fi
 % have any width.  % have any width.
 \def\lisppar{\null\endgraf}  \def\lisppar{\null\endgraf}
   
 % Make each space character in the input produce a normal interword  
 % space in the output.  Don't allow a line break at this space, as this  
 % is used only in environments like @example, where each line of input  
 % should produce a line of output anyway.  
 %  
 {\obeyspaces %  
 \gdef\sepspaces{\obeyspaces\let =\tie}}  
   
 % Define \obeyedspace to be our active space, whatever it is.  This is  
 % for use in \parsearg.  
 {\sepspaces%  
 \global\let\obeyedspace= }  
   
 % This space is always present above and below environments.  % This space is always present above and below environments.
 \newskip\envskipamount \envskipamount = 0pt  \newskip\envskipamount \envskipamount = 0pt
   
 % Make spacing and below environment symmetrical.  We use \parskip here  % Make spacing and below environment symmetrical.  We use \parskip here
 % to help in doing that, since in @example-like environments \parskip  % to help in doing that, since in @example-like environments \parskip
 % is reset to zero; thus the \afterenvbreak inserts no space -- but the  % is reset to zero; thus the \afterenvbreak inserts no space -- but the
 % start of the next paragraph will insert \parskip  % start of the next paragraph will insert \parskip.
 %  %
 \def\aboveenvbreak{{\advance\envskipamount by \parskip  \def\aboveenvbreak{{%
 \endgraf \ifdim\lastskip<\envskipamount    % =10000 instead of <10000 because of a special case in \itemzzz and
 \removelastskip \penalty-50 \vskip\envskipamount \fi}}    % \sectionheading, q.v.
     \ifnum \lastpenalty=10000 \else
       \advance\envskipamount by \parskip
       \endgraf
       \ifdim\lastskip<\envskipamount
         \removelastskip
         \ifnum\lastpenalty<10000
           % Penalize breaking before the environment, because preceding text
           % often leads into it.
           \penalty100
         \fi
         \vskip\envskipamount
       \fi
     \fi
   }}
   
 \let\afterenvbreak = \aboveenvbreak  \def\afterenvbreak{{%
     % =10000 instead of <10000 because of a special case in \itemzzz and
     % \sectionheading, q.v.
     \ifnum \lastpenalty=10000 \else
       \advance\envskipamount by \parskip
       \endgraf
       \ifdim\lastskip<\envskipamount
         \removelastskip
         % it's not a good place to break if the last penalty was \nobreak
         % or better ...
         \ifnum\lastpenalty<10000 \penalty-50 \fi
         \vskip\envskipamount
       \fi
     \fi
   }}
   
 % \nonarrowing is a flag.  If "set", @lisp etc don't narrow margins.  % \nonarrowing is a flag.  If "set", @lisp etc don't narrow margins; it will
   % also clear it, so that its embedded environments do the narrowing again.
 \let\nonarrowing=\relax  \let\nonarrowing=\relax
   
 % @cartouche ... @end cartouche: draw rectangle w/rounded corners around  % @cartouche ... @end cartouche: draw rectangle w/rounded corners around
Line 4110  width0pt\relax} \fi
Line 7217  width0pt\relax} \fi
 %  %
 \newskip\lskip\newskip\rskip  \newskip\lskip\newskip\rskip
   
 \long\def\cartouche{%  \envdef\cartouche{%
 \begingroup    \ifhmode\par\fi  % can't be in the midst of a paragraph.
         \lskip=\leftskip \rskip=\rightskip    \startsavinginserts
         \leftskip=0pt\rightskip=0pt %we want these *outside*.    \lskip=\leftskip \rskip=\rightskip
         \cartinner=\hsize \advance\cartinner by-\lskip    \leftskip=0pt\rightskip=0pt % we want these *outside*.
                           \advance\cartinner by-\rskip    \cartinner=\hsize \advance\cartinner by-\lskip
         \cartouter=\hsize    \advance\cartinner by-\rskip
         \advance\cartouter by 18.4pt % allow for 3pt kerns on either    \cartouter=\hsize
 %                                    side, and for 6pt waste from    \advance\cartouter by 18.4pt  % allow for 3pt kerns on either
 %                                    each corner char, and rule thickness                                  % side, and for 6pt waste from
         \normbskip=\baselineskip \normpskip=\parskip \normlskip=\lineskip                                  % each corner char, and rule thickness
         % Flag to tell @lisp, etc., not to narrow margin.    \normbskip=\baselineskip \normpskip=\parskip \normlskip=\lineskip
         \let\nonarrowing=\comment    %
         \vbox\bgroup    % If this cartouche directly follows a sectioning command, we need the
                 \baselineskip=0pt\parskip=0pt\lineskip=0pt    % \parskip glue (backspaced over by default) or the cartouche can
                 \carttop    % collide with the section heading.
                 \hbox\bgroup    \ifnum\lastpenalty>10000 \vskip\parskip \penalty\lastpenalty \fi
                         \hskip\lskip    %
                         \vrule\kern3pt    \setbox\groupbox=\vbox\bgroup
                         \vbox\bgroup        \baselineskip=0pt\parskip=0pt\lineskip=0pt
                                 \hsize=\cartinner        \carttop
                                 \kern3pt        \hbox\bgroup
                                 \begingroup            \hskip\lskip
                                         \baselineskip=\normbskip            \vrule\kern3pt
                                         \lineskip=\normlskip            \vbox\bgroup
                                         \parskip=\normpskip                \kern3pt
                                         \vskip -\parskip                \hsize=\cartinner
                 \baselineskip=\normbskip
                 \lineskip=\normlskip
                 \parskip=\normpskip
                 \vskip -\parskip
                 \comment % For explanation, see the end of def\group.
   }
 \def\Ecartouche{%  \def\Ecartouche{%
                                 \endgroup                \ifhmode\par\fi
                                 \kern3pt                \kern3pt
                         \egroup            \egroup
                         \kern3pt\vrule            \kern3pt\vrule
                         \hskip\rskip            \hskip\rskip
                 \egroup        \egroup
                 \cartbot        \cartbot
         \egroup    \egroup
 \endgroup    \addgroupbox
 }}    \checkinserts
   }
   
   
 % This macro is called at the beginning of all the @example variants,  % This macro is called at the beginning of all the @example variants,
 % inside a group.  % inside a group.
   \newdimen\nonfillparindent
 \def\nonfillstart{%  \def\nonfillstart{%
   \aboveenvbreak    \aboveenvbreak
   \inENV % This group ends at the end of the body    \ifdim\hfuzz < 12pt \hfuzz = 12pt \fi % Don't be fussy
   \hfuzz = 12pt % Don't be fussy  
   \sepspaces % Make spaces be word-separators rather than space tokens.    \sepspaces % Make spaces be word-separators rather than space tokens.
   \singlespace  
   \let\par = \lisppar % don't ignore blank lines    \let\par = \lisppar % don't ignore blank lines
   \obeylines % each line of input is a line of output    \obeylines % each line of input is a line of output
   \parskip = 0pt    \parskip = 0pt
     % Turn off paragraph indentation but redefine \indent to emulate
     % the normal \indent.
     \nonfillparindent=\parindent
   \parindent = 0pt    \parindent = 0pt
     \let\indent\nonfillindent
     %
   \emergencystretch = 0pt % don't try to avoid overfull boxes    \emergencystretch = 0pt % don't try to avoid overfull boxes
   % @cartouche defines \nonarrowing to inhibit narrowing  
   % at next level down.  
   \ifx\nonarrowing\relax    \ifx\nonarrowing\relax
     \advance \leftskip by \lispnarrowing      \advance \leftskip by \lispnarrowing
     \exdentamount=\lispnarrowing      \exdentamount=\lispnarrowing
     \let\exdent=\nofillexdent    \else
     \let\nonarrowing=\relax      \let\nonarrowing = \relax
   \fi    \fi
     \let\exdent=\nofillexdent
 }  }
   
 % Define the \E... control sequence only if we are inside the particular  \begingroup
 % environment, so the error checking in \end will work.  \obeyspaces
   % We want to swallow spaces (but not other tokens) after the fake
   % @indent in our nonfill-environments, where spaces are normally
   % active and set to @tie, resulting in them not being ignored after
   % @indent.
   \gdef\nonfillindent{\futurelet\temp\nonfillindentcheck}%
   \gdef\nonfillindentcheck{%
   \ifx\temp %
   \expandafter\nonfillindentgobble%
   \else%
   \leavevmode\nonfillindentbox%
   \fi%
   }%
   \endgroup
   \def\nonfillindentgobble#1{\nonfillindent}
   \def\nonfillindentbox{\hbox to \nonfillparindent{\hss}}
   
   % If you want all examples etc. small: @set dispenvsize small.
   % If you want even small examples the full size: @set dispenvsize nosmall.
   % This affects the following displayed environments:
   %    @example, @display, @format, @lisp
 %  %
 % To end an @example-like environment, we first end the paragraph (via  \def\smallword{small}
 % \afterenvbreak's vertical glue), and then the group.  That way we keep  \def\nosmallword{nosmall}
 % the zero \parskip that the environments set -- \parskip glue will be  \let\SETdispenvsize\relax
 % inserted at the beginning of the next paragraph in the document, after  \def\setnormaldispenv{%
 % the environment.    \ifx\SETdispenvsize\smallword
 %      % end paragraph for sake of leading, in case document has no blank
 \def\nonfillfinish{\afterenvbreak\endgroup}      % line.  This is redundant with what happens in \aboveenvbreak, but
       % we need to do it before changing the fonts, and it's inconvenient
       % to change the fonts afterward.
       \ifnum \lastpenalty=10000 \else \endgraf \fi
       \smallexamplefonts \rm
     \fi
   }
   \def\setsmalldispenv{%
     \ifx\SETdispenvsize\nosmallword
     \else
       \ifnum \lastpenalty=10000 \else \endgraf \fi
       \smallexamplefonts \rm
     \fi
   }
   
 % @lisp: indented, narrowed, typewriter font.  % We often define two environments, @foo and @smallfoo.
 \def\lisp{\begingroup  % Let's do it in one command.  #1 is the env name, #2 the definition.
   \nonfillstart  \def\makedispenvdef#1#2{%
   \let\Elisp = \nonfillfinish    \expandafter\envdef\csname#1\endcsname {\setnormaldispenv #2}%
   \tt    \expandafter\envdef\csname small#1\endcsname {\setsmalldispenv #2}%
   \let\kbdfont = \kbdexamplefont % Allow @kbd to do something special.    \expandafter\let\csname E#1\endcsname \afterenvbreak
   \gobble       % eat return    \expandafter\let\csname Esmall#1\endcsname \afterenvbreak
 }  }
   
 % @example: Same as @lisp.  % Define two environment synonyms (#1 and #2) for an environment.
 \def\example{\begingroup \def\Eexample{\nonfillfinish\endgroup}\lisp}  \def\maketwodispenvdef#1#2#3{%
     \makedispenvdef{#1}{#3}%
 % @small... is usually equivalent to the non-small (@smallbook    \makedispenvdef{#2}{#3}%
 % redefines).  We must call \example (or whatever) last in the  }
 % definition, since it reads the return following the @example (or  
 % whatever) command.  
 %  %
 % This actually allows (for example) @end display inside an  % @lisp: indented, narrowed, typewriter font;
 % @smalldisplay.  Too bad, but makeinfo will catch the error anyway.  % @example: same as @lisp.
 %  %
 \def\smalldisplay{\begingroup\def\Esmalldisplay{\nonfillfinish\endgroup}\display}  % @smallexample and @smalllisp: use smaller fonts.
 \def\smallexample{\begingroup\def\Esmallexample{\nonfillfinish\endgroup}\lisp}  
 \def\smallformat{\begingroup\def\Esmallformat{\nonfillfinish\endgroup}\format}  
 \def\smalllisp{\begingroup\def\Esmalllisp{\nonfillfinish\endgroup}\lisp}  
   
 % Real @smallexample and @smalllisp (when @smallbook): use smaller fonts.  
 % Originally contributed by Pavel@xerox.  % Originally contributed by Pavel@xerox.
 \def\smalllispx{\begingroup  %
   \def\Esmalllisp{\nonfillfinish\endgroup}%  \maketwodispenvdef{lisp}{example}{%
   \def\Esmallexample{\nonfillfinish\endgroup}%    \nonfillstart
   \smallfonts    \tt\setupmarkupstyle{example}%
   \lisp    \let\kbdfont = \kbdexamplefont % Allow @kbd to do something special.
     \gobble % eat return
 }  }
   % @display/@smalldisplay: same as @lisp except keep current font.
 % @display: same as @lisp except keep current font.  
 %  %
 \def\display{\begingroup  \makedispenvdef{display}{%
   \nonfillstart    \nonfillstart
   \let\Edisplay = \nonfillfinish  
   \gobble    \gobble
 }  }
   
 % @smalldisplay (when @smallbook): @display plus smaller fonts.  % @format/@smallformat: same as @display except don't narrow margins.
 %  %
 \def\smalldisplayx{\begingroup  \makedispenvdef{format}{%
   \def\Esmalldisplay{\nonfillfinish\endgroup}%    \let\nonarrowing = t%
   \smallfonts \rm    \nonfillstart
   \display    \gobble
 }  }
   
 % @format: same as @display except don't narrow margins.  % @flushleft: same as @format, but doesn't obey \SETdispenvsize.
 %  \envdef\flushleft{%
 \def\format{\begingroup    \let\nonarrowing = t%
   \let\nonarrowing = t  
   \nonfillstart    \nonfillstart
   \let\Eformat = \nonfillfinish  
   \gobble    \gobble
 }  }
   \let\Eflushleft = \afterenvbreak
   
 % @smallformat (when @smallbook): @format plus smaller fonts.  % @flushright.
 %  %
 \def\smallformatx{\begingroup  \envdef\flushright{%
   \def\Esmallformat{\nonfillfinish\endgroup}%    \let\nonarrowing = t%
   \smallfonts \rm    \nonfillstart
   \format    \advance\leftskip by 0pt plus 1fill\relax
     \gobble
 }  }
   \let\Eflushright = \afterenvbreak
   
 % @flushleft (same as @format).  
   % @raggedright does more-or-less normal line breaking but no right
   % justification.  From plain.tex.  Don't stretch around special
   % characters in urls in this environment, since the stretch at the right
   % should be enough.
   \envdef\raggedright{%
     \rightskip0pt plus2.4em \spaceskip.3333em \xspaceskip.5em\relax
     \def\urefprestretchamount{0pt}%
     \def\urefpoststretchamount{0pt}%
   }
   \let\Eraggedright\par
   
   \envdef\raggedleft{%
     \parindent=0pt \leftskip0pt plus2em
     \spaceskip.3333em \xspaceskip.5em \parfillskip=0pt
     \hbadness=10000 % Last line will usually be underfull, so turn off
                     % badness reporting.
   }
   \let\Eraggedleft\par
   
   \envdef\raggedcenter{%
     \parindent=0pt \rightskip0pt plus1em \leftskip0pt plus1em
     \spaceskip.3333em \xspaceskip.5em \parfillskip=0pt
     \hbadness=10000 % Last line will usually be underfull, so turn off
                     % badness reporting.
   }
   \let\Eraggedcenter\par
   
   
   % @quotation does normal linebreaking (hence we can't use \nonfillstart)
   % and narrows the margins.  We keep \parskip nonzero in general, since
   % we're doing normal filling.  So, when using \aboveenvbreak and
   % \afterenvbreak, temporarily make \parskip 0.
 %  %
 \def\flushleft{\begingroup \def\Eflushleft{\nonfillfinish\endgroup}\format}  \makedispenvdef{quotation}{\quotationstart}
   %
   \def\quotationstart{%
     \indentedblockstart % same as \indentedblock, but increase right margin too.
     \ifx\nonarrowing\relax
       \advance\rightskip by \lispnarrowing
     \fi
     \parsearg\quotationlabel
   }
   
 % @flushright.  % We have retained a nonzero parskip for the environment, since we're
   % doing normal filling.
 %  %
 \def\flushright{\begingroup  \def\Equotation{%
   \let\nonarrowing = t    \par
   \nonfillstart    \ifx\quotationauthor\thisisundefined\else
   \let\Eflushright = \nonfillfinish      % indent a bit.
   \advance\leftskip by 0pt plus 1fill      \leftline{\kern 2\leftskip \sl ---\quotationauthor}%
   \gobble    \fi
     {\parskip=0pt \afterenvbreak}%
 }  }
   \def\Esmallquotation{\Equotation}
   
 % @quotation does normal linebreaking (hence we can't use \nonfillstart)  % If we're given an argument, typeset it in bold with a colon after.
 % and narrows the margins.  \def\quotationlabel#1{%
     \def\temp{#1}%
     \ifx\temp\empty \else
       {\bf #1: }%
     \fi
   }
   
   % @indentedblock is like @quotation, but indents only on the left and
   % has no optional argument.
   %
   \makedispenvdef{indentedblock}{\indentedblockstart}
 %  %
 \def\quotation{%  \def\indentedblockstart{%
   \begingroup\inENV %This group ends at the end of the @quotation body  
   {\parskip=0pt \aboveenvbreak}% because \aboveenvbreak inserts \parskip    {\parskip=0pt \aboveenvbreak}% because \aboveenvbreak inserts \parskip
   \singlespace  
   \parindent=0pt    \parindent=0pt
   % We have retained a nonzero parskip for the environment, since we're  
   % doing normal filling. So to avoid extra space below the environment...  
   \def\Equotation{\parskip = 0pt \nonfillfinish}%  
   %    %
   % @cartouche defines \nonarrowing to inhibit narrowing at next level down.    % @cartouche defines \nonarrowing to inhibit narrowing at next level down.
   \ifx\nonarrowing\relax    \ifx\nonarrowing\relax
     \advance\leftskip by \lispnarrowing      \advance\leftskip by \lispnarrowing
     \advance\rightskip by \lispnarrowing  
     \exdentamount = \lispnarrowing      \exdentamount = \lispnarrowing
     \else
     \let\nonarrowing = \relax      \let\nonarrowing = \relax
   \fi    \fi
 }  }
   
   % Keep a nonzero parskip for the environment, since we're doing normal filling.
   %
   \def\Eindentedblock{%
     \par
     {\parskip=0pt \afterenvbreak}%
   }
   \def\Esmallindentedblock{\Eindentedblock}
   
 \message{defuns,}  
 % @defun etc.  
   
 % Allow user to change definition object font (\df) internally  % LaTeX-like @verbatim...@end verbatim and @verb{<char>...<char>}
 \def\setdeffont #1 {\csname DEF#1\endcsname}  % If we want to allow any <char> as delimiter,
   % we need the curly braces so that makeinfo sees the @verb command, eg:
 \newskip\defbodyindent \defbodyindent=.4in  % `@verbx...x' would look like the '@verbx' command.  --janneke@gnu.org
 \newskip\defargsindent \defargsindent=50pt  %
 \newskip\deftypemargin \deftypemargin=12pt  % [Knuth]: Donald Ervin Knuth, 1996.  The TeXbook.
 \newskip\deflastargmargin \deflastargmargin=18pt  %
   % [Knuth] p.344; only we need to do the other characters Texinfo sets
 \newcount\parencount  % active too.  Otherwise, they get lost as the first character on a
 % define \functionparens, which makes ( and ) and & do special things.  % verbatim line.
 % \functionparens affects the group it is contained in.  \def\dospecials{%
 \def\activeparens{%    \do\ \do\\\do\{\do\}\do\$\do\&%
 \catcode`\(=\active \catcode`\)=\active \catcode`\&=\active    \do\#\do\^\do\^^K\do\_\do\^^A\do\%\do\~%
 \catcode`\[=\active \catcode`\]=\active}    \do\<\do\>\do\|\do\@\do+\do\"%
     % Don't do the quotes -- if we do, @set txicodequoteundirected and
 % Make control sequences which act like normal parenthesis chars.    % @set txicodequotebacktick will not have effect on @verb and
 \let\lparen = ( \let\rparen = )    % @verbatim, and ?` and !` ligatures won't get disabled.
     %\do\`\do\'%
 {\activeparens % Now, smart parens don't turn on until &foo (see \amprm)  
   
 % Be sure that we always have a definition for `(', etc.  For example,  
 % if the fn name has parens in it, \boldbrax will not be in effect yet,  
 % so TeX would otherwise complain about undefined control sequence.  
 \global\let(=\lparen \global\let)=\rparen  
 \global\let[=\lbrack \global\let]=\rbrack  
   
 \gdef\functionparens{\boldbrax\let&=\amprm\parencount=0 }  
 \gdef\boldbrax{\let(=\opnr\let)=\clnr\let[=\lbrb\let]=\rbrb}  
 % This is used to turn on special parens  
 % but make & act ordinary (given that it's active).  
 \gdef\boldbraxnoamp{\let(=\opnr\let)=\clnr\let[=\lbrb\let]=\rbrb\let&=\ampnr}  
   
 % Definitions of (, ) and & used in args for functions.  
 % This is the definition of ( outside of all parentheses.  
 \gdef\oprm#1 {{\rm\char`\(}#1 \bf \let(=\opnested  
   \global\advance\parencount by 1  
 }  }
 %  %
 % This is the definition of ( when already inside a level of parens.  % [Knuth] p. 380
 \gdef\opnested{\char`\(\global\advance\parencount by 1 }  \def\uncatcodespecials{%
     \def\do##1{\catcode`##1=\other}\dospecials}
 %  %
 \gdef\clrm{% Print a paren in roman if it is taking us back to depth of 0.  % Setup for the @verb command.
   % also in that case restore the outer-level definition of (.  
   \ifnum \parencount=1 {\rm \char `\)}\sl \let(=\oprm \else \char `\) \fi  
   \global\advance \parencount by -1 }  
 % If we encounter &foo, then turn on ()-hacking afterwards  
 \gdef\amprm#1 {{\rm\&#1}\let(=\oprm \let)=\clrm\ }  
 %  %
 \gdef\normalparens{\boldbrax\let&=\ampnr}  % Eight spaces for a tab
 } % End of definition inside \activeparens  \begingroup
 %% These parens (in \boldbrax) actually are a little bolder than the    \catcode`\^^I=\active
 %% contained text.  This is especially needed for [ and ]    \gdef\tabeightspaces{\catcode`\^^I=\active\def^^I{\ \ \ \ \ \ \ \ }}
 \def\opnr{{\sf\char`\(}\global\advance\parencount by 1 }  \endgroup
 \def\clnr{{\sf\char`\)}\global\advance\parencount by -1 }  %
 \let\ampnr = \&  \def\setupverb{%
 \def\lbrb{{\bf\char`\[}}    \tt  % easiest (and conventionally used) font for verbatim
 \def\rbrb{{\bf\char`\]}}    \def\par{\leavevmode\endgraf}%
     \setupmarkupstyle{verb}%
 % Active &'s sneak into the index arguments, so make sure it's defined.    \tabeightspaces
 {    % Respect line breaks,
   \catcode`& = 13    % print special symbols as themselves, and
   \global\let& = \ampnr    % make each space count
     % must do in this order:
     \obeylines \uncatcodespecials \sepspaces
 }  }
   
 % First, defname, which formats the header line itself.  % Setup for the @verbatim environment
 % #1 should be the function name.  %
 % #2 should be the type of definition, such as "Function".  % Real tab expansion.
   \newdimen\tabw \setbox0=\hbox{\tt\space} \tabw=8\wd0 % tab amount
   %
   % We typeset each line of the verbatim in an \hbox, so we can handle
   % tabs.  The \global is in case the verbatim line starts with an accent,
   % or some other command that starts with a begin-group.  Otherwise, the
   % entire \verbbox would disappear at the corresponding end-group, before
   % it is typeset.  Meanwhile, we can't have nested verbatim commands
   % (can we?), so the \global won't be overwriting itself.
   \newbox\verbbox
   \def\starttabbox{\global\setbox\verbbox=\hbox\bgroup}
   %
   \begingroup
     \catcode`\^^I=\active
     \gdef\tabexpand{%
       \catcode`\^^I=\active
       \def^^I{\leavevmode\egroup
         \dimen\verbbox=\wd\verbbox % the width so far, or since the previous tab
         \divide\dimen\verbbox by\tabw
         \multiply\dimen\verbbox by\tabw % compute previous multiple of \tabw
         \advance\dimen\verbbox by\tabw  % advance to next multiple of \tabw
         \wd\verbbox=\dimen\verbbox \box\verbbox \starttabbox
       }%
     }
   \endgroup
   
 \def\defname #1#2{%  % start the verbatim environment.
 % Get the values of \leftskip and \rightskip as they were  \def\setupverbatim{%
 % outside the @def...    \let\nonarrowing = t%
 \dimen2=\leftskip    \nonfillstart
 \advance\dimen2 by -\defbodyindent    \tt % easiest (and conventionally used) font for verbatim
 \noindent    % The \leavevmode here is for blank lines.  Otherwise, we would
 \setbox0=\hbox{\hskip \deflastargmargin{\rm #2}\hskip \deftypemargin}%    % never \starttabox and the \egroup would end verbatim mode.
 \dimen0=\hsize \advance \dimen0 by -\wd0 % compute size for first line    \def\par{\leavevmode\egroup\box\verbbox\endgraf}%
 \dimen1=\hsize \advance \dimen1 by -\defargsindent %size for continuations    \tabexpand
 \parshape 2 0in \dimen0 \defargsindent \dimen1    \setupmarkupstyle{verbatim}%
 % Now output arg 2 ("Function" or some such)    % Respect line breaks,
 % ending at \deftypemargin from the right margin,    % print special symbols as themselves, and
 % but stuck inside a box of width 0 so it does not interfere with linebreaking    % make each space count.
 {% Adjust \hsize to exclude the ambient margins,    % Must do in this order:
 % so that \rightline will obey them.    \obeylines \uncatcodespecials \sepspaces
 \advance \hsize by -\dimen2    \everypar{\starttabbox}%
 \rlap{\rightline{{\rm #2}\hskip -1.25pc }}}%  
 % Make all lines underfull and no complaints:  
 \tolerance=10000 \hbadness=10000  
 \advance\leftskip by -\defbodyindent  
 \exdentamount=\defbodyindent  
 {\df #1}\enskip        % Generate function name  
 }  }
   
 % Actually process the body of a definition  % Do the @verb magic: verbatim text is quoted by unique
 % #1 should be the terminating control sequence, such as \Edefun.  % delimiter characters.  Before first delimiter expect a
 % #2 should be the "another name" control sequence, such as \defunx.  % right brace, after last delimiter expect closing brace:
 % #3 should be the control sequence that actually processes the header,  %
 %    such as \defunheader.  %    \def\doverb'{'<char>#1<char>'}'{#1}
   %
   % [Knuth] p. 382; only eat outer {}
   \begingroup
     \catcode`[=1\catcode`]=2\catcode`\{=\other\catcode`\}=\other
     \gdef\doverb{#1[\def\next##1#1}[##1\endgroup]\next]
   \endgroup
   %
   \def\verb{\begingroup\setupverb\doverb}
   %
   %
   % Do the @verbatim magic: define the macro \doverbatim so that
   % the (first) argument ends when '@end verbatim' is reached, ie:
   %
   %     \def\doverbatim#1@end verbatim{#1}
   %
   % For Texinfo it's a lot easier than for LaTeX,
   % because texinfo's \verbatim doesn't stop at '\end{verbatim}':
   % we need not redefine '\', '{' and '}'.
   %
   % Inspired by LaTeX's verbatim command set [latex.ltx]
   %
   \begingroup
     \catcode`\ =\active
     \obeylines %
     % ignore everything up to the first ^^M, that's the newline at the end
     % of the @verbatim input line itself.  Otherwise we get an extra blank
     % line in the output.
     \xdef\doverbatim#1^^M#2@end verbatim{#2\noexpand\end\gobble verbatim}%
     % We really want {...\end verbatim} in the body of the macro, but
     % without the active space; thus we have to use \xdef and \gobble.
   \endgroup
   %
   \envdef\verbatim{%
       \setupverbatim\doverbatim
   }
   \let\Everbatim = \afterenvbreak
   
 \def\defparsebody #1#2#3{\begingroup\inENV% Environment for definitionbody  
 \medbreak %  
 % Define the end token that this defining construct specifies  
 % so that it will exit this group.  
 \def#1{\endgraf\endgroup\medbreak}%  
 \def#2{\begingroup\obeylines\activeparens\spacesplit#3}%  
 \parindent=0in  
 \advance\leftskip by \defbodyindent  
 \exdentamount=\defbodyindent  
 \begingroup %  
 \catcode 61=\active % 61 is `='  
 \obeylines\activeparens\spacesplit#3}  
   
 % #1 is the \E... control sequence to end the definition (which we define).  % @verbatiminclude FILE - insert text of file in verbatim environment.
 % #2 is the \...x control sequence for consecutive fns (which we define).  
 % #3 is the control sequence to call to resume processing.  
 % #4, delimited by the space, is the class name.  
 %  %
 \def\defmethparsebody#1#2#3#4 {\begingroup\inENV %  \def\verbatiminclude{\parseargusing\filenamecatcodes\doverbatiminclude}
 \medbreak %  %
 % Define the end token that this defining construct specifies  \def\doverbatiminclude#1{%
 % so that it will exit this group.    {%
 \def#1{\endgraf\endgroup\medbreak}%      \makevalueexpandable
 \def#2##1 {\begingroup\obeylines\activeparens\spacesplit{#3{##1}}}%      \setupverbatim
 \parindent=0in      \indexnofonts       % Allow `@@' and other weird things in file names.
 \advance\leftskip by \defbodyindent      \wlog{texinfo.tex: doing @verbatiminclude of #1^^J}%
 \exdentamount=\defbodyindent      \input #1
 \begingroup\obeylines\activeparens\spacesplit{#3{#4}}}      \afterenvbreak
     }%
   }
   
 % Used for @deftypemethod and @deftypeivar.  % @copying ... @end copying.
 % #1 is the \E... control sequence to end the definition (which we define).  % Save the text away for @insertcopying later.
 % #2 is the \...x control sequence for consecutive fns (which we define).  
 % #3 is the control sequence to call to resume processing.  
 % #4, delimited by a space, is the class name.  
 % #5 is the method's return type.  
 %  %
 \def\deftypemethparsebody#1#2#3#4 #5 {\begingroup\inENV  % We save the uninterpreted tokens, rather than creating a box.
   \medbreak  % Saving the text in a box would be much easier, but then all the
   \def#1{\endgraf\endgroup\medbreak}%  % typesetting commands (@smallbook, font changes, etc.) have to be done
   \def#2##1 ##2 {\begingroup\obeylines\activeparens\spacesplit{#3{##1}{##2}}}%  % beforehand -- and a) we want @copying to be done first in the source
   \parindent=0in  % file; b) letting users define the frontmatter in as flexible order as
   \advance\leftskip by \defbodyindent  % possible is desirable.
   \exdentamount=\defbodyindent  %
   \begingroup\obeylines\activeparens\spacesplit{#3{#4}{#5}}}  \def\copying{\checkenv{}\begingroup\scanargctxt\docopying}
   \def\docopying#1@end copying{\endgroup\def\copyingtext{#1}}
   %
   \def\insertcopying{%
     \begingroup
       \parindent = 0pt  % paragraph indentation looks wrong on title page
       \scanexp\copyingtext
     \endgroup
   }
   
 % Used for @deftypeop.  The change from \deftypemethparsebody is an  
 % extra argument at the beginning which is the `category', instead of it  
 % being the hardwired string `Method' or `Instance Variable'.  We have  
 % to account for this both in the \...x definition and in parsing the  
 % input at hand.  Thus also need a control sequence (passed as #5) for  
 % the \E... definition to assign the category name to.  
 %  
 \def\deftypeopparsebody#1#2#3#4#5 #6 {\begingroup\inENV  
   \medbreak  
   \def#1{\endgraf\endgroup\medbreak}%  
   \def#2##1 ##2 ##3 {%  
     \def#4{##1}%  
     \begingroup\obeylines\activeparens\spacesplit{#3{##2}{##3}}}%  
   \parindent=0in  
   \advance\leftskip by \defbodyindent  
   \exdentamount=\defbodyindent  
   \begingroup\obeylines\activeparens\spacesplit{#3{#5}{#6}}}  
   
 \def\defopparsebody #1#2#3#4#5 {\begingroup\inENV %  \message{defuns,}
 \medbreak %  % @defun etc.
 % Define the end token that this defining construct specifies  
 % so that it will exit this group.  
 \def#1{\endgraf\endgroup\medbreak}%  
 \def#2##1 ##2 {\def#4{##1}%  
 \begingroup\obeylines\activeparens\spacesplit{#3{##2}}}%  
 \parindent=0in  
 \advance\leftskip by \defbodyindent  
 \exdentamount=\defbodyindent  
 \begingroup\obeylines\activeparens\spacesplit{#3{#5}}}  
   
 % These parsing functions are similar to the preceding ones  \newskip\defbodyindent \defbodyindent=.4in
 % except that they do not make parens into active characters.  \newskip\defargsindent \defargsindent=50pt
 % These are used for "variables" since they have no arguments.  \newskip\deflastargmargin \deflastargmargin=18pt
   \newcount\defunpenalty
   
 \def\defvarparsebody #1#2#3{\begingroup\inENV% Environment for definitionbody  % Start the processing of @deffn:
 \medbreak %  \def\startdefun{%
 % Define the end token that this defining construct specifies    \ifnum\lastpenalty<10000
 % so that it will exit this group.      \medbreak
 \def#1{\endgraf\endgroup\medbreak}%      \defunpenalty=10003 % Will keep this @deffn together with the
 \def#2{\begingroup\obeylines\spacesplit#3}%                          % following @def command, see below.
 \parindent=0in    \else
 \advance\leftskip by \defbodyindent      % If there are two @def commands in a row, we'll have a \nobreak,
 \exdentamount=\defbodyindent      % which is there to keep the function description together with its
 \begingroup %      % header.  But if there's nothing but headers, we need to allow a
 \catcode 61=\active %      % break somewhere.  Check specifically for penalty 10002, inserted
 \obeylines\spacesplit#3}      % by \printdefunline, instead of 10000, since the sectioning
       % commands also insert a nobreak penalty, and we don't want to allow
 % This is used for \def{tp,vr}parsebody.  It could probably be used for      % a break between a section heading and a defun.
 % some of the others, too, with some judicious conditionals.      %
 %      % As a further refinement, we avoid "club" headers by signalling
 \def\parsebodycommon#1#2#3{%      % with penalty of 10003 after the very first @deffn in the
   \begingroup\inENV %      % sequence (see above), and penalty of 10002 after any following
   \medbreak %      % @def command.
   % Define the end token that this defining construct specifies      \ifnum\lastpenalty=10002 \penalty2000 \else \defunpenalty=10002 \fi
   % so that it will exit this group.      %
   \def#1{\endgraf\endgroup\medbreak}%      % Similarly, after a section heading, do not allow a break.
   \def#2##1 {\begingroup\obeylines\spacesplit{#3{##1}}}%      % But do insert the glue.
       \medskip  % preceded by discardable penalty, so not a breakpoint
     \fi
     %
   \parindent=0in    \parindent=0in
   \advance\leftskip by \defbodyindent    \advance\leftskip by \defbodyindent
   \exdentamount=\defbodyindent    \exdentamount=\defbodyindent
   \begingroup\obeylines  
 }  }
   
 \def\defvrparsebody#1#2#3#4 {%  \def\dodefunx#1{%
   \parsebodycommon{#1}{#2}{#3}%    % First, check whether we are in the right environment:
   \spacesplit{#3{#4}}%    \checkenv#1%
     %
     % As above, allow line break if we have multiple x headers in a row.
     % It's not a great place, though.
     \ifnum\lastpenalty=10002 \penalty3000 \else \defunpenalty=10002 \fi
     %
     % And now, it's time to reuse the body of the original defun:
     \expandafter\gobbledefun#1%
 }  }
   \def\gobbledefun#1\startdefun{}
   
 % This loses on `@deftp {Data Type} {struct termios}' -- it thinks the  % \printdefunline \deffnheader{text}
 % type is just `struct', because we lose the braces in `{struct  
 % termios}' when \spacesplit reads its undelimited argument.  Sigh.  
 % \let\deftpparsebody=\defvrparsebody  
 %  %
 % So, to get around this, we put \empty in with the type name.  That  \def\printdefunline#1#2{%
 % way, TeX won't find exactly `{...}' as an undelimited argument, and    \begingroup
 % won't strip off the braces.      % call \deffnheader:
 %      #1#2 \endheader
 \def\deftpparsebody #1#2#3#4 {%      % common ending:
   \parsebodycommon{#1}{#2}{#3}%      \interlinepenalty = 10000
   \spacesplit{\parsetpheaderline{#3{#4}}}\empty      \advance\rightskip by 0pt plus 1fil\relax
       \endgraf
       \nobreak\vskip -\parskip
       \penalty\defunpenalty  % signal to \startdefun and \dodefunx
       % Some of the @defun-type tags do not enable magic parentheses,
       % rendering the following check redundant.  But we don't optimize.
       \checkparencounts
     \endgroup
 }  }
   
 % Fine, but then we have to eventually remove the \empty *and* the  \def\Edefun{\endgraf\medbreak}
 % braces (if any).  That's what this does.  
   % \makedefun{deffn} creates \deffn, \deffnx and \Edeffn;
   % the only thing remaining is to define \deffnheader.
 %  %
 \def\removeemptybraces\empty#1\relax{#1}  \def\makedefun#1{%
     \expandafter\let\csname E#1\endcsname = \Edefun
     \edef\temp{\noexpand\domakedefun
       \makecsname{#1}\makecsname{#1x}\makecsname{#1header}}%
     \temp
   }
   
 % After \spacesplit has done its work, this is called -- #1 is the final  % \domakedefun \deffn \deffnx \deffnheader { (defn. of \deffnheader) }
 % thing to call, #2 the type name (which starts with \empty), and #3  
 % (which might be empty) the arguments.  
 %  %
 \def\parsetpheaderline#1#2#3{%  % Define \deffn and \deffnx, without parameters.
   #1{\removeemptybraces#2\relax}{#3}%  % \deffnheader has to be defined explicitly.
 }%  %
   \def\domakedefun#1#2#3{%
     \envdef#1{%
       \startdefun
       \doingtypefnfalse    % distinguish typed functions from all else
       \parseargusing\activeparens{\printdefunline#3}%
     }%
     \def#2{\dodefunx#1}%
     \def#3%
   }
   
 \def\defopvarparsebody #1#2#3#4#5 {\begingroup\inENV %  \newif\ifdoingtypefn       % doing typed function?
 \medbreak %  \newif\ifrettypeownline    % typeset return type on its own line?
 % Define the end token that this defining construct specifies  
 % so that it will exit this group.  
 \def#1{\endgraf\endgroup\medbreak}%  
 \def#2##1 ##2 {\def#4{##1}%  
 \begingroup\obeylines\spacesplit{#3{##2}}}%  
 \parindent=0in  
 \advance\leftskip by \defbodyindent  
 \exdentamount=\defbodyindent  
 \begingroup\obeylines\spacesplit{#3{#5}}}  
   
 % Split up #2 at the first space token.  % @deftypefnnewline on|off says whether the return type of typed functions
 % call #1 with two arguments:  % are printed on their own line.  This affects @deftypefn, @deftypefun,
 %  the first is all of #2 before the space token,  % @deftypeop, and @deftypemethod.
 %  the second is all of #2 after that space token.  %
 % If #2 contains no space token, all of it is passed as the first arg  \parseargdef\deftypefnnewline{%
 % and the second is passed as empty.    \def\temp{#1}%
     \ifx\temp\onword
       \expandafter\let\csname SETtxideftypefnnl\endcsname
         = \empty
     \else\ifx\temp\offword
       \expandafter\let\csname SETtxideftypefnnl\endcsname
         = \relax
     \else
       \errhelp = \EMsimple
       \errmessage{Unknown @txideftypefnnl value `\temp',
                   must be on|off}%
     \fi\fi
   }
   
 {\obeylines  % Untyped functions:
 \gdef\spacesplit#1#2^^M{\endgroup\spacesplitfoo{#1}#2 \relax\spacesplitfoo}%  
 \long\gdef\spacesplitfoo#1#2 #3#4\spacesplitfoo{%  
 \ifx\relax #3%  
 #1{#2}{}\else #1{#2}{#3#4}\fi}}  
   
 % So much for the things common to all kinds of definitions.  % @deffn category name args
   \makedefun{deffn}{\deffngeneral{}}
   
 % Define @defun.  % @deffn category class name args
   \makedefun{defop}#1 {\defopon{#1\ \putwordon}}
   
 % First, define the processing that is wanted for arguments of \defun  % \defopon {category on}class name args
 % Use this to expand the args and terminate the paragraph they make up  \def\defopon#1#2 {\deffngeneral{\putwordon\ \code{#2}}{#1\ \code{#2}} }
   
 \def\defunargs#1{\functionparens \sl  % \deffngeneral {subind}category name args
 % Expand, preventing hyphenation at `-' chars.  %
 % Note that groups don't affect changes in \hyphenchar.  \def\deffngeneral#1#2 #3 #4\endheader{%
 % Set the font temporarily and use \font in case \setfont made \tensl a macro.    % Remember that \dosubind{fn}{foo}{} is equivalent to \doind{fn}{foo}.
 {\tensl\hyphenchar\font=0}%    \dosubind{fn}{\code{#3}}{#1}%
 #1%    \defname{#2}{}{#3}\magicamp\defunargs{#4\unskip}%
 {\tensl\hyphenchar\font=45}%  
 \ifnum\parencount=0 \else \errmessage{Unbalanced parentheses in @def}\fi%  
 \interlinepenalty=10000  
 \advance\rightskip by 0pt plus 1fil  
 \endgraf\nobreak\vskip -\parskip\nobreak  
 }  }
   
 \def\deftypefunargs #1{%  % Typed functions:
 % Expand, preventing hyphenation at `-' chars.  
 % Note that groups don't affect changes in \hyphenchar.  
 % Use \boldbraxnoamp, not \functionparens, so that & is not special.  
 \boldbraxnoamp  
 \tclose{#1}% avoid \code because of side effects on active chars  
 \interlinepenalty=10000  
 \advance\rightskip by 0pt plus 1fil  
 \endgraf\nobreak\vskip -\parskip\nobreak  
 }  
   
 % Do complete processing of one @defun or @defunx line already parsed.  % @deftypefn category type name args
   \makedefun{deftypefn}{\deftypefngeneral{}}
   
 % @deffn Command forward-char nchars  % @deftypeop category class type name args
   \makedefun{deftypeop}#1 {\deftypeopon{#1\ \putwordon}}
   
 \def\deffn{\defmethparsebody\Edeffn\deffnx\deffnheader}  % \deftypeopon {category on}class type name args
   \def\deftypeopon#1#2 {\deftypefngeneral{\putwordon\ \code{#2}}{#1\ \code{#2}} }
   
 \def\deffnheader #1#2#3{\doind {fn}{\code{#2}}%  % \deftypefngeneral {subind}category type name args
 \begingroup\defname {#2}{#1}\defunargs{#3}\endgroup %  %
 \catcode 61=\other % Turn off change made in \defparsebody  \def\deftypefngeneral#1#2 #3 #4 #5\endheader{%
     \dosubind{fn}{\code{#4}}{#1}%
     \doingtypefntrue
     \defname{#2}{#3}{#4}\defunargs{#5\unskip}%
 }  }
   
 % @defun == @deffn Function  % Typed variables:
   
 \def\defun{\defparsebody\Edefun\defunx\defunheader}  % @deftypevr category type var args
   \makedefun{deftypevr}{\deftypecvgeneral{}}
   
 \def\defunheader #1#2{\doind {fn}{\code{#1}}% Make entry in function index  % @deftypecv category class type var args
 \begingroup\defname {#1}{\putwordDeffunc}%  \makedefun{deftypecv}#1 {\deftypecvof{#1\ \putwordof}}
 \defunargs {#2}\endgroup %  
 \catcode 61=\other % Turn off change made in \defparsebody  
 }  
   
 % @deftypefun int foobar (int @var{foo}, float @var{bar})  % \deftypecvof {category of}class type var args
   \def\deftypecvof#1#2 {\deftypecvgeneral{\putwordof\ \code{#2}}{#1\ \code{#2}} }
   
 \def\deftypefun{\defparsebody\Edeftypefun\deftypefunx\deftypefunheader}  % \deftypecvgeneral {subind}category type var args
   %
 % #1 is the data type.  #2 is the name and args.  \def\deftypecvgeneral#1#2 #3 #4 #5\endheader{%
 \def\deftypefunheader #1#2{\deftypefunheaderx{#1}#2 \relax}    \dosubind{vr}{\code{#4}}{#1}%
 % #1 is the data type, #2 the name, #3 the args.    \defname{#2}{#3}{#4}\defunargs{#5\unskip}%
 \def\deftypefunheaderx #1#2 #3\relax{%  
 \doind {fn}{\code{#2}}% Make entry in function index  
 \begingroup\defname {\defheaderxcond#1\relax$$$#2}{\putwordDeftypefun}%  
 \deftypefunargs {#3}\endgroup %  
 \catcode 61=\other % Turn off change made in \defparsebody  
 }  }
   
 % @deftypefn {Library Function} int foobar (int @var{foo}, float @var{bar})  % Untyped variables:
   
 \def\deftypefn{\defmethparsebody\Edeftypefn\deftypefnx\deftypefnheader}  % @defvr category var args
   \makedefun{defvr}#1 {\deftypevrheader{#1} {} }
   
 % \defheaderxcond#1\relax$$$  % @defcv category class var args
 % puts #1 in @code, followed by a space, but does nothing if #1 is null.  \makedefun{defcv}#1 {\defcvof{#1\ \putwordof}}
 \def\defheaderxcond#1#2$$${\ifx#1\relax\else\code{#1#2} \fi}  
   
 % #1 is the classification.  #2 is the data type.  #3 is the name and args.  % \defcvof {category of}class var args
 \def\deftypefnheader #1#2#3{\deftypefnheaderx{#1}{#2}#3 \relax}  \def\defcvof#1#2 {\deftypecvof{#1}#2 {} }
 % #1 is the classification, #2 the data type, #3 the name, #4 the args.  
 \def\deftypefnheaderx #1#2#3 #4\relax{%  
 \doind {fn}{\code{#3}}% Make entry in function index  
 \begingroup  
 \normalparens % notably, turn off `&' magic, which prevents  
 %               at least some C++ text from working  
 \defname {\defheaderxcond#2\relax$$$#3}{#1}%  
 \deftypefunargs {#4}\endgroup %  
 \catcode 61=\other % Turn off change made in \defparsebody  
 }  
   
 % @defmac == @deffn Macro  % Types:
   
 \def\defmac{\defparsebody\Edefmac\defmacx\defmacheader}  % @deftp category name args
   \makedefun{deftp}#1 #2 #3\endheader{%
 \def\defmacheader #1#2{\doind {fn}{\code{#1}}% Make entry in function index    \doind{tp}{\code{#2}}%
 \begingroup\defname {#1}{\putwordDefmac}%    \defname{#1}{}{#2}\defunargs{#3\unskip}%
 \defunargs {#2}\endgroup %  
 \catcode 61=\other % Turn off change made in \defparsebody  
 }  }
   
 % @defspec == @deffn Special Form  % Remaining @defun-like shortcuts:
   \makedefun{defun}{\deffnheader{\putwordDeffunc} }
   \makedefun{defmac}{\deffnheader{\putwordDefmac} }
   \makedefun{defspec}{\deffnheader{\putwordDefspec} }
   \makedefun{deftypefun}{\deftypefnheader{\putwordDeffunc} }
   \makedefun{defvar}{\defvrheader{\putwordDefvar} }
   \makedefun{defopt}{\defvrheader{\putwordDefopt} }
   \makedefun{deftypevar}{\deftypevrheader{\putwordDefvar} }
   \makedefun{defmethod}{\defopon\putwordMethodon}
   \makedefun{deftypemethod}{\deftypeopon\putwordMethodon}
   \makedefun{defivar}{\defcvof\putwordInstanceVariableof}
   \makedefun{deftypeivar}{\deftypecvof\putwordInstanceVariableof}
   
 \def\defspec{\defparsebody\Edefspec\defspecx\defspecheader}  % \defname, which formats the name of the @def (not the args).
   % #1 is the category, such as "Function".
 \def\defspecheader #1#2{\doind {fn}{\code{#1}}% Make entry in function index  % #2 is the return type, if any.
 \begingroup\defname {#1}{\putwordDefspec}%  % #3 is the function name.
 \defunargs {#2}\endgroup %  
 \catcode 61=\other % Turn off change made in \defparsebody  
 }  
   
 % @defop CATEGORY CLASS OPERATION ARG...  
 %  %
 \def\defop #1 {\def\defoptype{#1}%  % We are followed by (but not passed) the arguments, if any.
 \defopparsebody\Edefop\defopx\defopheader\defoptype}  
 %  %
 \def\defopheader#1#2#3{%  \def\defname#1#2#3{%
 \dosubind {fn}{\code{#2}}{\putwordon\ #1}% Make entry in function index    \par
 \begingroup\defname {#2}{\defoptype\ \putwordon\ #1}%    % Get the values of \leftskip and \rightskip as they were outside the @def...
 \defunargs {#3}\endgroup %    \advance\leftskip by -\defbodyindent
     %
     % Determine if we are typesetting the return type of a typed function
     % on a line by itself.
     \rettypeownlinefalse
     \ifdoingtypefn  % doing a typed function specifically?
       % then check user option for putting return type on its own line:
       \expandafter\ifx\csname SETtxideftypefnnl\endcsname\relax \else
         \rettypeownlinetrue
       \fi
     \fi
     %
     % How we'll format the category name.  Putting it in brackets helps
     % distinguish it from the body text that may end up on the next line
     % just below it.
     \def\temp{#1}%
     \setbox0=\hbox{\kern\deflastargmargin \ifx\temp\empty\else [\rm\temp]\fi}
     %
     % Figure out line sizes for the paragraph shape.  We'll always have at
     % least two.
     \tempnum = 2
     %
     % The first line needs space for \box0; but if \rightskip is nonzero,
     % we need only space for the part of \box0 which exceeds it:
     \dimen0=\hsize  \advance\dimen0 by -\wd0  \advance\dimen0 by \rightskip
     %
     % If doing a return type on its own line, we'll have another line.
     \ifrettypeownline
       \advance\tempnum by 1
       \def\maybeshapeline{0in \hsize}%
     \else
       \def\maybeshapeline{}%
     \fi
     %
     % The continuations:
     \dimen2=\hsize  \advance\dimen2 by -\defargsindent
     %
     % The final paragraph shape:
     \parshape \tempnum  0in \dimen0  \maybeshapeline  \defargsindent \dimen2
     %
     % Put the category name at the right margin.
     \noindent
     \hbox to 0pt{%
       \hfil\box0 \kern-\hsize
       % \hsize has to be shortened this way:
       \kern\leftskip
       % Intentionally do not respect \rightskip, since we need the space.
     }%
     %
     % Allow all lines to be underfull without complaint:
     \tolerance=10000 \hbadness=10000
     \exdentamount=\defbodyindent
     {%
       % defun fonts. We use typewriter by default (used to be bold) because:
       % . we're printing identifiers, they should be in tt in principle.
       % . in languages with many accents, such as Czech or French, it's
       %   common to leave accents off identifiers.  The result looks ok in
       %   tt, but exceedingly strange in rm.
       % . we don't want -- and --- to be treated as ligatures.
       % . this still does not fix the ?` and !` ligatures, but so far no
       %   one has made identifiers using them :).
       \df \tt
       \def\temp{#2}% text of the return type
       \ifx\temp\empty\else
         \tclose{\temp}% typeset the return type
         \ifrettypeownline
           % put return type on its own line; prohibit line break following:
           \hfil\vadjust{\nobreak}\break
         \else
           \space  % type on same line, so just followed by a space
         \fi
       \fi           % no return type
       #3% output function name
     }%
     {\rm\enskip}% hskip 0.5 em of \tenrm
     %
     \boldbrax
     % arguments will be output next, if any.
 }  }
   
 % @deftypeop CATEGORY CLASS TYPE OPERATION ARG...  % Print arguments in slanted roman (not ttsl), inconsistently with using
   % tt for the name.  This is because literal text is sometimes needed in
   % the argument list (groff manual), and ttsl and tt are not very
   % distinguishable.  Prevent hyphenation at `-' chars.
 %  %
 \def\deftypeop #1 {\def\deftypeopcategory{#1}%  \def\defunargs#1{%
   \deftypeopparsebody\Edeftypeop\deftypeopx\deftypeopheader    % use sl by default (not ttsl),
                        \deftypeopcategory}    % tt for the names.
 %    \df \sl \hyphenchar\font=0
 % #1 is the class name, #2 the data type, #3 the operation name, #4 the args.    %
 \def\deftypeopheader#1#2#3#4{%    % On the other hand, if an argument has two dashes (for instance), we
   \dosubind{fn}{\code{#3}}{\putwordon\ \code{#1}}% entry in function index    % want a way to get ttsl.  We used to recommend @var for that, so
   \begingroup    % leave the code in, but it's strange for @var to lead to typewriter.
     \defname{\defheaderxcond#2\relax$$$#3}    % Nowadays we recommend @code, since the difference between a ttsl hyphen
             {\deftypeopcategory\ \putwordon\ \code{#1}}%    % and a tt hyphen is pretty tiny.  @code also disables ?` !`.
     \deftypefunargs{#4}%    \def\var##1{{\setupmarkupstyle{var}\ttslanted{##1}}}%
   \endgroup    #1%
     \sl\hyphenchar\font=45
 }  }
   
 % @deftypemethod CLASS TYPE METHOD ARG...  % We want ()&[] to print specially on the defun line.
 %  %
 \def\deftypemethod{%  \def\activeparens{%
   \deftypemethparsebody\Edeftypemethod\deftypemethodx\deftypemethodheader}    \catcode`\(=\active \catcode`\)=\active
 %    \catcode`\[=\active \catcode`\]=\active
 % #1 is the class name, #2 the data type, #3 the method name, #4 the args.    \catcode`\&=\active
 \def\deftypemethodheader#1#2#3#4{%  
   \dosubind{fn}{\code{#3}}{\putwordon\ \code{#1}}% entry in function index  
   \begingroup  
     \defname{\defheaderxcond#2\relax$$$#3}{\putwordMethodon\ \code{#1}}%  
     \deftypefunargs{#4}%  
   \endgroup  
 }  }
   
 % @deftypeivar CLASS TYPE VARNAME  % Make control sequences which act like normal parenthesis chars.
 %  \let\lparen = ( \let\rparen = )
 \def\deftypeivar{%  
   \deftypemethparsebody\Edeftypeivar\deftypeivarx\deftypeivarheader}  
 %  
 % #1 is the class name, #2 the data type, #3 the variable name.  
 \def\deftypeivarheader#1#2#3{%  
   \dosubind{vr}{\code{#3}}{\putwordof\ \code{#1}}% entry in variable index  
   \begingroup  
     \defname{#3}{\putwordInstanceVariableof\ \code{#1}}%  
     \defvarargs{#3}%  
   \endgroup  
 }  
   
 % @defmethod == @defop Method  % Be sure that we always have a definition for `(', etc.  For example,
 %  % if the fn name has parens in it, \boldbrax will not be in effect yet,
 \def\defmethod{\defmethparsebody\Edefmethod\defmethodx\defmethodheader}  % so TeX would otherwise complain about undefined control sequence.
 %  {
 % #1 is the class name, #2 the method name, #3 the args.    \activeparens
 \def\defmethodheader#1#2#3{%    \global\let(=\lparen \global\let)=\rparen
   \dosubind{fn}{\code{#2}}{\putwordon\ \code{#1}}% entry in function index    \global\let[=\lbrack \global\let]=\rbrack
   \begingroup    \global\let& = \&
     \defname{#2}{\putwordMethodon\ \code{#1}}%  
     \defunargs{#3}%    \gdef\boldbrax{\let(=\opnr\let)=\clnr\let[=\lbrb\let]=\rbrb}
   \endgroup    \gdef\magicamp{\let&=\amprm}
 }  }
   
 % @defcv {Class Option} foo-class foo-flag  \newcount\parencount
   
 \def\defcv #1 {\def\defcvtype{#1}%  % If we encounter &foo, then turn on ()-hacking afterwards
 \defopvarparsebody\Edefcv\defcvx\defcvarheader\defcvtype}  \newif\ifampseen
   \def\amprm#1 {\ampseentrue{\bf\&#1 }}
   
 \def\defcvarheader #1#2#3{%  \def\parenfont{%
 \dosubind {vr}{\code{#2}}{\putwordof\ #1}% Make entry in var index    \ifampseen
 \begingroup\defname {#2}{\defcvtype\ \putwordof\ #1}%      % At the first level, print parens in roman,
 \defvarargs {#3}\endgroup %      % otherwise use the default font.
       \ifnum \parencount=1 \rm \fi
     \else
       % The \sf parens (in \boldbrax) actually are a little bolder than
       % the contained text.  This is especially needed for [ and ] .
       \sf
     \fi
 }  }
   \def\infirstlevel#1{%
     \ifampseen
       \ifnum\parencount=1
         #1%
       \fi
     \fi
   }
   \def\bfafterword#1 {#1 \bf}
   
 % @defivar CLASS VARNAME == @defcv {Instance Variable} CLASS VARNAME  \def\opnr{%
 %    \global\advance\parencount by 1
 \def\defivar{\defvrparsebody\Edefivar\defivarx\defivarheader}    {\parenfont(}%
 %    \infirstlevel \bfafterword
 \def\defivarheader#1#2#3{%  
   \dosubind {vr}{\code{#2}}{\putwordof\ #1}% entry in var index  
   \begingroup  
     \defname{#2}{\putwordInstanceVariableof\ #1}%  
     \defvarargs{#3}%  
   \endgroup  
 }  }
   \def\clnr{%
     {\parenfont)}%
     \infirstlevel \sl
     \global\advance\parencount by -1
   }
   
 % @defvar  \newcount\brackcount
 % First, define the processing that is wanted for arguments of @defvar.  \def\lbrb{%
 % This is actually simple: just print them in roman.    \global\advance\brackcount by 1
 % This must expand the args and terminate the paragraph they make up    {\bf[}%
 \def\defvarargs #1{\normalparens #1%  
 \interlinepenalty=10000  
 \endgraf\nobreak\vskip -\parskip\nobreak}  
   
 % @defvr Counter foo-count  
   
 \def\defvr{\defvrparsebody\Edefvr\defvrx\defvrheader}  
   
 \def\defvrheader #1#2#3{\doind {vr}{\code{#2}}%  
 \begingroup\defname {#2}{#1}\defvarargs{#3}\endgroup}  
   
 % @defvar == @defvr Variable  
   
 \def\defvar{\defvarparsebody\Edefvar\defvarx\defvarheader}  
   
 \def\defvarheader #1#2{\doind {vr}{\code{#1}}% Make entry in var index  
 \begingroup\defname {#1}{\putwordDefvar}%  
 \defvarargs {#2}\endgroup %  
 }  }
   \def\rbrb{%
     {\bf]}%
     \global\advance\brackcount by -1
   }
   
 % @defopt == @defvr {User Option}  \def\checkparencounts{%
     \ifnum\parencount=0 \else \badparencount \fi
 \def\defopt{\defvarparsebody\Edefopt\defoptx\defoptheader}    \ifnum\brackcount=0 \else \badbrackcount \fi
   
 \def\defoptheader #1#2{\doind {vr}{\code{#1}}% Make entry in var index  
 \begingroup\defname {#1}{\putwordDefopt}%  
 \defvarargs {#2}\endgroup %  
 }  }
   % these should not use \errmessage; the glibc manual, at least, actually
   % has such constructs (when documenting function pointers).
   \def\badparencount{%
     \message{Warning: unbalanced parentheses in @def...}%
     \global\parencount=0
   }
   \def\badbrackcount{%
     \message{Warning: unbalanced square brackets in @def...}%
     \global\brackcount=0
   }
   
 % @deftypevar int foobar  
   
 \def\deftypevar{\defvarparsebody\Edeftypevar\deftypevarx\deftypevarheader}  \message{macros,}
   % @macro.
   
 % #1 is the data type.  #2 is the name, perhaps followed by text that  % To do this right we need a feature of e-TeX, \scantokens,
 % is actually part of the data type, which should not be put into the index.  % which we arrange to emulate with a temporary file in ordinary TeX.
 \def\deftypevarheader #1#2{%  \ifx\eTeXversion\thisisundefined
 \dovarind#2 \relax% Make entry in variables index    \newwrite\macscribble
 \begingroup\defname {\defheaderxcond#1\relax$$$#2}{\putwordDeftypevar}%    \def\scantokens#1{%
 \interlinepenalty=10000      \toks0={#1}%
 \endgraf\nobreak\vskip -\parskip\nobreak      \immediate\openout\macscribble=\jobname.tmp
 \endgroup}      \immediate\write\macscribble{\the\toks0}%
 \def\dovarind#1 #2\relax{\doind{vr}{\code{#1}}}      \immediate\closeout\macscribble
       \input \jobname.tmp
     }
   \fi
   
 % @deftypevr {Global Flag} int enable  \let\aftermacroxxx\relax
   \def\aftermacro{\aftermacroxxx}
   
 \def\deftypevr{\defvrparsebody\Edeftypevr\deftypevrx\deftypevrheader}  % alias because \c means cedilla in @tex or @math
   \let\texinfoc=\c
   
 \def\deftypevrheader #1#2#3{\dovarind#3 \relax%  \newcount\savedcatcodeone
 \begingroup\defname {\defheaderxcond#2\relax$$$#3}{#1}  \newcount\savedcatcodetwo
 \interlinepenalty=10000  
 \endgraf\nobreak\vskip -\parskip\nobreak  
 \endgroup}  
   
 % Now define @deftp  % Used at the time of macro expansion.
 % Args are printed in bold, a slight difference from @defvar.  % Argument is macro body with arguments substituted
   \def\scanmacro#1{%
     \newlinechar`\^^M
     \def\xeatspaces{\eatspaces}%
     %
     % Temporarily undo catcode changes of \printindex.  Set catcode of @ to
     % 0 so that @-commands in macro expansions aren't printed literally when
     % formatting an index file, where \ is used as the escape character.
     \savedcatcodeone=\catcode`\@
     \savedcatcodetwo=\catcode`\\
     \catcode`\@=0
     \catcode`\\=\active
     %
     % Process the macro body under the current catcode regime.
     \scantokens{#1@texinfoc}\aftermacro%
     %
     \catcode`\@=\savedcatcodeone
     \catcode`\\=\savedcatcodetwo
     %
     % The \texinfoc is to remove the \newlinechar added by \scantokens, and
     % can be noticed by \parsearg.
     %   The \aftermacro allows a \comment at the end of the macro definition
     % to duplicate itself past the final \newlinechar added by \scantokens:
     % this is used in the definition of \group to comment out a newline.  We
     % don't do the same for \c to support Texinfo files with macros that ended
     % with a @c, which should no longer be necessary.
     %   We avoid surrounding the call to \scantokens with \bgroup and \egroup
     % to allow macros to open or close groups themselves.
   }
   
 \def\deftpargs #1{\bf \defvarargs{#1}}  % Used for copying and captions
   \def\scanexp#1{%
 % @deftp Class window height width ...    \expandafter\scanmacro\expandafter{#1}%
   
 \def\deftp{\deftpparsebody\Edeftp\deftpx\deftpheader}  
   
 \def\deftpheader #1#2#3{\doind {tp}{\code{#2}}%  
 \begingroup\defname {#2}{#1}\deftpargs{#3}\endgroup}  
   
 % These definitions are used if you use @defunx (etc.)  
 % anywhere other than immediately after a @defun or @defunx.  
 %  
 \def\defcvx#1 {\errmessage{@defcvx in invalid context}}  
 \def\deffnx#1 {\errmessage{@deffnx in invalid context}}  
 \def\defivarx#1 {\errmessage{@defivarx in invalid context}}  
 \def\defmacx#1 {\errmessage{@defmacx in invalid context}}  
 \def\defmethodx#1 {\errmessage{@defmethodx in invalid context}}  
 \def\defoptx #1 {\errmessage{@defoptx in invalid context}}  
 \def\defopx#1 {\errmessage{@defopx in invalid context}}  
 \def\defspecx#1 {\errmessage{@defspecx in invalid context}}  
 \def\deftpx#1 {\errmessage{@deftpx in invalid context}}  
 \def\deftypefnx#1 {\errmessage{@deftypefnx in invalid context}}  
 \def\deftypefunx#1 {\errmessage{@deftypefunx in invalid context}}  
 \def\deftypeivarx#1 {\errmessage{@deftypeivarx in invalid context}}  
 \def\deftypemethodx#1 {\errmessage{@deftypemethodx in invalid context}}  
 \def\deftypeopx#1 {\errmessage{@deftypeopx in invalid context}}  
 \def\deftypevarx#1 {\errmessage{@deftypevarx in invalid context}}  
 \def\deftypevrx#1 {\errmessage{@deftypevrx in invalid context}}  
 \def\defunx#1 {\errmessage{@defunx in invalid context}}  
 \def\defvarx#1 {\errmessage{@defvarx in invalid context}}  
 \def\defvrx#1 {\errmessage{@defvrx in invalid context}}  
   
   
 \message{macros,}  
 % @macro.  
   
 % To do this right we need a feature of e-TeX, \scantokens,  
 % which we arrange to emulate with a temporary file in ordinary TeX.  
 \ifx\eTeXversion\undefined  
  \newwrite\macscribble  
  \def\scanmacro#1{%  
    \begingroup \newlinechar`\^^M  
    % Undo catcode changes of \startcontents and \doprintindex  
    \catcode`\@=0 \catcode`\\=12 \escapechar=`\@  
    % Append \endinput to make sure that TeX does not see the ending newline.  
    \toks0={#1\endinput}%  
    \immediate\openout\macscribble=\jobname.tmp  
    \immediate\write\macscribble{\the\toks0}%  
    \immediate\closeout\macscribble  
    \let\xeatspaces\eatspaces  
    \input \jobname.tmp  
    \endgroup  
 }  }
 \else  
 \def\scanmacro#1{%  
 \begingroup \newlinechar`\^^M  
 % Undo catcode changes of \startcontents and \doprintindex  
 \catcode`\@=0 \catcode`\\=12 \escapechar=`\@  
 \let\xeatspaces\eatspaces\scantokens{#1\endinput}\endgroup}  
 \fi  
   
 \newcount\paramno   % Count of parameters  \newcount\paramno   % Count of parameters
 \newtoks\macname    % Macro name  \newtoks\macname    % Macro name
 \newif\ifrecursive  % Is it recursive?  \newif\ifrecursive  % Is it recursive?
 \def\macrolist{}    % List of all defined macros in the form  
                     % \do\macro1\do\macro2...  
   
   % List of all defined macros in the form
   %    \commondummyword\macro1\commondummyword\macro2...
   % Currently is also contains all @aliases; the list can be split
   % if there is a need.
   \def\macrolist{}
   
   % Add the macro to \macrolist
   \def\addtomacrolist#1{\expandafter \addtomacrolistxxx \csname#1\endcsname}
   \def\addtomacrolistxxx#1{%
        \toks0 = \expandafter{\macrolist\commondummyword#1}%
        \xdef\macrolist{\the\toks0}%
   }
   
 % Utility routines.  % Utility routines.
 % Thisdoes \let #1 = #2, except with \csnames.  % This does \let #1 = #2, with \csnames; that is,
   %   \let \csname#1\endcsname = \csname#2\endcsname
   % (except of course we have to play expansion games).
   %
 \def\cslet#1#2{%  \def\cslet#1#2{%
 \expandafter\expandafter    \expandafter\let
 \expandafter\let    \csname#1\expandafter\endcsname
 \expandafter\expandafter    \csname#2\endcsname
 \csname#1\endcsname  }
 \csname#2\endcsname}  
   
 % Trim leading and trailing spaces off a string.  % Trim leading and trailing spaces off a string.
 % Concepts from aro-bend problem 15 (see CTAN).  % Concepts from aro-bend problem 15 (see CTAN).
Line 4898  width0pt\relax} \fi
Line 8159  width0pt\relax} \fi
 }  }
   
 % Trim a single trailing ^^M off a string.  % Trim a single trailing ^^M off a string.
 {\catcode`\^^M=12\catcode`\Q=3%  {\catcode`\^^M=\other \catcode`\Q=3%
 \gdef\eatcr #1{\eatcra #1Q^^MQ}%  \gdef\eatcr #1{\eatcra #1Q^^MQ}%
 \gdef\eatcra#1^^MQ{\eatcrb#1Q}%  \gdef\eatcra#1^^MQ{\eatcrb#1Q}%
 \gdef\eatcrb#1Q#2Q{#1}%  \gdef\eatcrb#1Q#2Q{#1}%
Line 4906  width0pt\relax} \fi
Line 8167  width0pt\relax} \fi
   
 % Macro bodies are absorbed as an argument in a context where  % Macro bodies are absorbed as an argument in a context where
 % all characters are catcode 10, 11 or 12, except \ which is active  % all characters are catcode 10, 11 or 12, except \ which is active
 % (as in normal texinfo). It is necessary to change the definition of \.  % (as in normal texinfo). It is necessary to change the definition of \
   % to recognize macro arguments; this is the job of \mbodybackslash.
   %
   % Non-ASCII encodings make 8-bit characters active, so un-activate
   % them to avoid their expansion.  Must do this non-globally, to
   % confine the change to the current group.
   %
 % It's necessary to have hard CRs when the macro is executed. This is  % It's necessary to have hard CRs when the macro is executed. This is
 % done by  making ^^M (\endlinechar) catcode 12 when reading the macro  % done by making ^^M (\endlinechar) catcode 12 when reading the macro
 % body, and then making it the \newlinechar in \scanmacro.  % body, and then making it the \newlinechar in \scanmacro.
   %
   \def\scanctxt{% used as subroutine
     \catcode`\"=\other
     \catcode`\+=\other
     \catcode`\<=\other
     \catcode`\>=\other
     \catcode`\^=\other
     \catcode`\_=\other
     \catcode`\|=\other
     \catcode`\~=\other
     \passthroughcharstrue
   }
   
 \def\macrobodyctxt{%  \def\scanargctxt{% used for copying and captions, not macros.
   \catcode`\~=12    \scanctxt
   \catcode`\^=12    \catcode`\@=\other
   \catcode`\_=12    \catcode`\\=\other
   \catcode`\|=12    \catcode`\^^M=\other
   \catcode`\<=12  }
   \catcode`\>=12  
   \catcode`\+=12  
   \catcode`\{=12  
   \catcode`\}=12  
   \catcode`\@=12  
   \catcode`\^^M=12  
   \usembodybackslash}  
   
   \def\macrobodyctxt{% used for @macro definitions
     \scanctxt
     \catcode`\ =\other
     \catcode`\@=\other
     \catcode`\{=\other
     \catcode`\}=\other
     \catcode`\^^M=\other
     \usembodybackslash
   }
   
   % Used when scanning braced macro arguments.  Note, however, that catcode
   % changes here are ineffectual if the macro invocation was nested inside
   % an argument to another Texinfo command.
 \def\macroargctxt{%  \def\macroargctxt{%
   \catcode`\~=12    \scanctxt
   \catcode`\^=12    \catcode`\ =\active
   \catcode`\_=12    \catcode`\^^M=\other
   \catcode`\|=12    \catcode`\\=\active
   \catcode`\<=12  }
   \catcode`\>=12  
   \catcode`\+=12  
   \catcode`\@=12  
   \catcode`\\=12}  
   
   \def\macrolineargctxt{% used for whole-line arguments without braces
     \scanctxt
     \catcode`\{=\other
     \catcode`\}=\other
   }
   
 % \mbodybackslash is the definition of \ in @macro bodies.  % \mbodybackslash is the definition of \ in @macro bodies.
 % It maps \foo\ => \csname macarg.foo\endcsname => #N  % It maps \foo\ => \csname macarg.foo\endcsname => #N
 % where N is the macro parameter number.  % where N is the macro parameter number.
 % We define \csname macarg.\endcsname to be \realbackslash, so  % We define \csname macarg.\endcsname to be \realbackslash, so
 % \\ in macro replacement text gets you a backslash.  % \\ in macro replacement text gets you a backslash.
   %
 {\catcode`@=0 @catcode`@\=@active  {\catcode`@=0 @catcode`@\=@active
  @gdef@usembodybackslash{@let\=@mbodybackslash}   @gdef@usembodybackslash{@let\=@mbodybackslash}
  @gdef@mbodybackslash#1\{@csname macarg.#1@endcsname}   @gdef@mbodybackslash#1\{@csname macarg.#1@endcsname}
 }  }
 \expandafter\def\csname macarg.\endcsname{\realbackslash}  \expandafter\def\csname macarg.\endcsname{\realbackslash}
   
   \def\margbackslash#1{\char`\#1 }
   
 \def\macro{\recursivefalse\parsearg\macroxxx}  \def\macro{\recursivefalse\parsearg\macroxxx}
 \def\rmacro{\recursivetrue\parsearg\macroxxx}  \def\rmacro{\recursivetrue\parsearg\macroxxx}
   
 \def\macroxxx#1{%  \def\macroxxx#1{%
   \getargs{#1}%           now \macname is the macname and \argl the arglist    \getargs{#1}% now \macname is the macname and \argl the arglist
   \ifx\argl\empty       % no arguments    \ifx\argl\empty       % no arguments
      \paramno=0%       \paramno=0\relax
   \else    \else
      \expandafter\parsemargdef \argl;%       \expandafter\parsemargdef \argl;%
        \if\paramno>256\relax
          \ifx\eTeXversion\thisisundefined
            \errhelp = \EMsimple
            \errmessage{You need eTeX to compile a file with macros with more than 256 arguments}
          \fi
        \fi
   \fi    \fi
   \if1\csname ismacro.\the\macname\endcsname    \if1\csname ismacro.\the\macname\endcsname
      \message{Warning: redefining \the\macname}%       \message{Warning: redefining \the\macname}%
   \else    \else
      \expandafter\ifx\csname \the\macname\endcsname \relax       \expandafter\ifx\csname \the\macname\endcsname \relax
      \else \errmessage{The name \the\macname\space is reserved}\fi       \else \errmessage{Macro name \the\macname\space already defined}\fi
      \global\cslet{macsave.\the\macname}{\the\macname}%       \global\cslet{macsave.\the\macname}{\the\macname}%
      \global\expandafter\let\csname ismacro.\the\macname\endcsname=1%       \global\expandafter\let\csname ismacro.\the\macname\endcsname=1%
      % Add the macroname to \macrolist       \addtomacrolist{\the\macname}%
      \toks0 = \expandafter{\macrolist\do}%  
      \xdef\macrolist{\the\toks0  
        \expandafter\noexpand\csname\the\macname\endcsname}%  
   \fi    \fi
   \begingroup \macrobodyctxt    \begingroup \macrobodyctxt
   \ifrecursive \expandafter\parsermacbody    \ifrecursive \expandafter\parsermacbody
   \else \expandafter\parsemacbody    \else \expandafter\parsemacbody
   \fi}    \fi}
   
 \def\unmacro{\parsearg\unmacroxxx}  \parseargdef\unmacro{%
 \def\unmacroxxx#1{%  
   \if1\csname ismacro.#1\endcsname    \if1\csname ismacro.#1\endcsname
     \global\cslet{#1}{macsave.#1}%      \global\cslet{#1}{macsave.#1}%
     \global\expandafter\let \csname ismacro.#1\endcsname=0%      \global\expandafter\let \csname ismacro.#1\endcsname=0%
     % Remove the macro name from \macrolist      % Remove the macro name from \macrolist:
     \begingroup      \begingroup
       \edef\tempa{\expandafter\noexpand\csname#1\endcsname}%        \expandafter\let\csname#1\endcsname \relax
       \def\do##1{%        \let\commondummyword\unmacrodo
         \def\tempb{##1}%        \xdef\macrolist{\macrolist}%
         \ifx\tempa\tempb  
           % remove this  
         \else  
           \toks0 = \expandafter{\newmacrolist\do}%  
           \edef\newmacrolist{\the\toks0\expandafter\noexpand\tempa}%  
         \fi}%  
       \def\newmacrolist{}%  
       % Execute macro list to define \newmacrolist  
       \macrolist  
       \global\let\macrolist\newmacrolist  
     \endgroup      \endgroup
   \else    \else
     \errmessage{Macro #1 not defined}%      \errmessage{Macro #1 not defined}%
   \fi    \fi
 }  }
   
 % This makes use of the obscure feature that if the last token of a  % Called by \do from \dounmacro on each macro.  The idea is to omit any
 % <parameter list> is #, then the preceding argument is delimited by  % macro definitions that have been changed to \relax.
 % an opening brace, and that opening brace is not consumed.  %
   \def\unmacrodo#1{%
     \ifx #1\relax
       % remove this
     \else
       \noexpand\commondummyword \noexpand#1%
     \fi
   }
   
   % \getargs -- Parse the arguments to a @macro line.  Set \macname to
   % the name of the macro, and \argl to the braced argument list.
 \def\getargs#1{\getargsxxx#1{}}  \def\getargs#1{\getargsxxx#1{}}
 \def\getargsxxx#1#{\getmacname #1 \relax\getmacargs}  \def\getargsxxx#1#{\getmacname #1 \relax\getmacargs}
 \def\getmacname #1 #2\relax{\macname={#1}}  \def\getmacname#1 #2\relax{\macname={#1}}
 \def\getmacargs#1{\def\argl{#1}}  \def\getmacargs#1{\def\argl{#1}}
   % This made use of the feature that if the last token of a
   % <parameter list> is #, then the preceding argument is delimited by
   % an opening brace, and that opening brace is not consumed.
   
 % Parse the optional {params} list.  Set up \paramno and \paramlist  % Parse the optional {params} list to @macro or @rmacro.
 % so \defmacro knows what to do.  Define \macarg.blah for each blah  % Set \paramno to the number of arguments,
 % in the params list, to be ##N where N is the position in that list.  % and \paramlist to a parameter text for the macro (e.g. #1,#2,#3 for a
   % three-param macro.)  Define \macarg.BLAH for each BLAH in the params
   % list to some hook where the argument is to be expanded.  If there are
   % less than 10 arguments that hook is to be replaced by ##N where N
   % is the position in that list, that is to say the macro arguments are to be
   % defined `a la TeX in the macro body.
   %
 % That gets used by \mbodybackslash (above).  % That gets used by \mbodybackslash (above).
   
 % We need to get `macro parameter char #' into several definitions.  
 % The technique used is stolen from LaTeX:  let \hash be something  
 % unexpandable, insert that wherever you need a #, and then redefine  
 % it to # just before using the token list produced.  
 %  %
 % The same technique is used to protect \eatspaces till just before  % If there are 10 or more arguments, a different technique is used: see
 % the macro is used.  % \parsemmanyargdef.
   %
 \def\parsemargdef#1;{\paramno=0\def\paramlist{}%  \def\parsemargdef#1;{%
         \let\hash\relax\let\xeatspaces\relax\parsemargdefxxx#1,;,}    \paramno=0\def\paramlist{}%
     \let\hash\relax
     % \hash is redefined to `#' later to get it into definitions
     \let\xeatspaces\relax
     \parsemargdefxxx#1,;,%
     \ifnum\paramno<10\relax\else
       \paramno0\relax
       \parsemmanyargdef@@#1,;,% 10 or more arguments
     \fi
   }
 \def\parsemargdefxxx#1,{%  \def\parsemargdefxxx#1,{%
   \if#1;\let\next=\relax    \if#1;\let\next=\relax
   \else \let\next=\parsemargdefxxx    \else \let\next=\parsemargdefxxx
     \advance\paramno by 1%      \advance\paramno by 1
     \expandafter\edef\csname macarg.\eatspaces{#1}\endcsname      \expandafter\edef\csname macarg.\eatspaces{#1}\endcsname
         {\xeatspaces{\hash\the\paramno}}%          {\xeatspaces{\hash\the\paramno}}%
     \edef\paramlist{\paramlist\hash\the\paramno,}%      \edef\paramlist{\paramlist\hash\the\paramno,}%
   \fi\next}    \fi\next}
   
 % These two commands read recursive and nonrecursive macro bodies.  % \parsemacbody, \parsermacbody
 % (They're different since rec and nonrec macros end differently.)  %
   % Read recursive and nonrecursive macro bodies. (They're different since
   % rec and nonrec macros end differently.)
   %
   % We are in \macrobodyctxt, and the \xdef causes backslashshes in the macro
   % body to be transformed.
   % Set \macrobody to the body of the macro, and call \defmacro.
   %
   {\catcode`\ =\other\long\gdef\parsemacbody#1@end macro{%
   \xdef\macrobody{\eatcr{#1}}\endgroup\defmacro}}%
   {\catcode`\ =\other\long\gdef\parsermacbody#1@end rmacro{%
   \xdef\macrobody{\eatcr{#1}}\endgroup\defmacro}}%
   
 \long\def\parsemacbody#1@end macro%  % Make @ a letter, so that we can make private-to-Texinfo macro names.
 {\xdef\temp{\eatcr{#1}}\endgroup\defmacro}%  \edef\texiatcatcode{\the\catcode`\@}
 \long\def\parsermacbody#1@end rmacro%  \catcode `@=11\relax
 {\xdef\temp{\eatcr{#1}}\endgroup\defmacro}%  
   
 % This defines the macro itself. There are six cases: recursive and  %%%%%%%%%%%%%% Code for > 10 arguments only   %%%%%%%%%%%%%%%%%%
 % nonrecursive macros of zero, one, and many arguments.  
 % Much magic with \expandafter here.  % If there are 10 or more arguments, a different technique is used, where the
   % hook remains in the body, and when macro is to be expanded the body is
   % processed again to replace the arguments.
   %
   % In that case, the hook is \the\toks N-1, and we simply set \toks N-1 to the
   % argument N value and then \edef the body (nothing else will expand because of
   % the catcode regime under which the body was input).
   %
   % If you compile with TeX (not eTeX), and you have macros with 10 or more
   % arguments, no macro can have more than 256 arguments (else error).
   %
   % In case that there are 10 or more arguments we parse again the arguments
   % list to set new definitions for the \macarg.BLAH macros corresponding to
   % each BLAH argument. It was anyhow needed to parse already once this list
   % in order to count the arguments, and as macros with at most 9 arguments
   % are by far more frequent than macro with 10 or more arguments, defining
   % twice the \macarg.BLAH macros does not cost too much processing power.
   \def\parsemmanyargdef@@#1,{%
     \if#1;\let\next=\relax
     \else
       \let\next=\parsemmanyargdef@@
       \edef\tempb{\eatspaces{#1}}%
       \expandafter\def\expandafter\tempa
          \expandafter{\csname macarg.\tempb\endcsname}%
       % Note that we need some extra \noexpand\noexpand, this is because we
       % don't want \the  to be expanded in the \parsermacbody  as it uses an
       % \xdef .
       \expandafter\edef\tempa
         {\noexpand\noexpand\noexpand\the\toks\the\paramno}%
       \advance\paramno by 1\relax
     \fi\next}
   
   
   \let\endargs@\relax
   \let\nil@\relax
   \def\nilm@{\nil@}%
   \long\def\nillm@{\nil@}%
   
   % This macro is expanded during the Texinfo macro expansion, not during its
   % definition.  It gets all the arguments' values and assigns them to macros
   % macarg.ARGNAME
   %
   % #1 is the macro name
   % #2 is the list of argument names
   % #3 is the list of argument values
   \def\getargvals@#1#2#3{%
     \def\macargdeflist@{}%
     \def\saveparamlist@{#2}% Need to keep a copy for parameter expansion.
     \def\paramlist{#2,\nil@}%
     \def\macroname{#1}%
     \begingroup
     \macroargctxt
     \def\argvaluelist{#3,\nil@}%
     \def\@tempa{#3}%
     \ifx\@tempa\empty
       \setemptyargvalues@
     \else
       \getargvals@@
     \fi
   }
   \def\getargvals@@{%
     \ifx\paramlist\nilm@
         % Some sanity check needed here that \argvaluelist is also empty.
         \ifx\argvaluelist\nillm@
         \else
           \errhelp = \EMsimple
           \errmessage{Too many arguments in macro `\macroname'!}%
         \fi
         \let\next\macargexpandinbody@
     \else
       \ifx\argvaluelist\nillm@
          % No more arguments values passed to macro.  Set remaining named-arg
          % macros to empty.
          \let\next\setemptyargvalues@
       \else
         % pop current arg name into \@tempb
         \def\@tempa##1{\pop@{\@tempb}{\paramlist}##1\endargs@}%
         \expandafter\@tempa\expandafter{\paramlist}%
          % pop current argument value into \@tempc
         \def\@tempa##1{\longpop@{\@tempc}{\argvaluelist}##1\endargs@}%
         \expandafter\@tempa\expandafter{\argvaluelist}%
          % Here \@tempb is the current arg name and \@tempc is the current arg value.
          % First place the new argument macro definition into \@tempd
          \expandafter\macname\expandafter{\@tempc}%
          \expandafter\let\csname macarg.\@tempb\endcsname\relax
          \expandafter\def\expandafter\@tempe\expandafter{%
            \csname macarg.\@tempb\endcsname}%
          \edef\@tempd{\long\def\@tempe{\the\macname}}%
          \push@\@tempd\macargdeflist@
          \let\next\getargvals@@
       \fi
     \fi
     \next
   }
   
   \def\push@#1#2{%
     \expandafter\expandafter\expandafter\def
     \expandafter\expandafter\expandafter#2%
     \expandafter\expandafter\expandafter{%
     \expandafter#1#2}%
   }
   
   % Replace arguments by their values in the macro body, and place the result
   % in macro \@tempa.
   %
   \def\macvalstoargs@{%
     %  To do this we use the property that token registers that are \the'ed
     % within an \edef  expand only once. So we are going to place all argument
     % values into respective token registers.
     %
     % First we save the token context, and initialize argument numbering.
     \begingroup
       \paramno0\relax
       % Then, for each argument number #N, we place the corresponding argument
       % value into a new token list register \toks#N
       \expandafter\putargsintokens@\saveparamlist@,;,%
       % Then, we expand the body so that argument are replaced by their
       % values. The trick for values not to be expanded themselves is that they
       % are within tokens and that tokens expand only once in an \edef .
       \edef\@tempc{\csname mac.\macroname .body\endcsname}%
       % Now we restore the token stack pointer to free the token list registers
       % which we have used, but we make sure that expanded body is saved after
       % group.
       \expandafter
     \endgroup
     \expandafter\def\expandafter\@tempa\expandafter{\@tempc}%
     }
   
   % Define the named-macro outside of this group and then close this group.
   %
   \def\macargexpandinbody@{%
     \expandafter
     \endgroup
     \macargdeflist@
     % First the replace in body the macro arguments by their values, the result
     % is in \@tempa .
     \macvalstoargs@
     % Then we point at the \norecurse or \gobble (for recursive) macro value
     % with \@tempb .
     \expandafter\let\expandafter\@tempb\csname mac.\macroname .recurse\endcsname
     % Depending on whether it is recursive or not, we need some tailing
     % \egroup .
     \ifx\@tempb\gobble
        \let\@tempc\relax
     \else
        \let\@tempc\egroup
     \fi
     % And now we do the real job:
     \edef\@tempd{\noexpand\@tempb{\macroname}\noexpand\scanmacro{\@tempa}\@tempc}%
     \@tempd
   }
   
   \def\putargsintokens@#1,{%
     \if#1;\let\next\relax
     \else
       \let\next\putargsintokens@
       % First we allocate the new token list register, and give it a temporary
       % alias \@tempb .
       \toksdef\@tempb\the\paramno
       % Then we place the argument value into that token list register.
       \expandafter\let\expandafter\@tempa\csname macarg.#1\endcsname
       \expandafter\@tempb\expandafter{\@tempa}%
       \advance\paramno by 1\relax
     \fi
     \next
   }
   
   % Trailing missing arguments are set to empty.
   %
   \def\setemptyargvalues@{%
     \ifx\paramlist\nilm@
       \let\next\macargexpandinbody@
     \else
       \expandafter\setemptyargvaluesparser@\paramlist\endargs@
       \let\next\setemptyargvalues@
     \fi
     \next
   }
   
   \def\setemptyargvaluesparser@#1,#2\endargs@{%
     \expandafter\def\expandafter\@tempa\expandafter{%
       \expandafter\def\csname macarg.#1\endcsname{}}%
     \push@\@tempa\macargdeflist@
     \def\paramlist{#2}%
   }
   
   % #1 is the element target macro
   % #2 is the list macro
   % #3,#4\endargs@ is the list value
   \def\pop@#1#2#3,#4\endargs@{%
      \def#1{#3}%
      \def#2{#4}%
   }
   \long\def\longpop@#1#2#3,#4\endargs@{%
      \long\def#1{#3}%
      \long\def#2{#4}%
   }
   
   
   %%%%%%%%%%%%%% End of code for > 10 arguments %%%%%%%%%%%%%%%%%%
   
   
   % This defines a Texinfo @macro or @rmacro, called by \parsemacbody.
   %    \macrobody has the body of the macro in it, with placeholders for
   % its parameters, looking like "\xeatspaces{\hash 1}".
   %    \paramno is the number of parameters
   %    \paramlist is a TeX parameter text, e.g. "#1,#2,#3,"
   % There are four cases: macros of zero, one, up to nine, and many arguments.
 % \xdef is used so that macro definitions will survive the file  % \xdef is used so that macro definitions will survive the file
 % they're defined in; @include reads the file inside a group.  % they're defined in: @include reads the file inside a group.
   %
 \def\defmacro{%  \def\defmacro{%
   \let\hash=##% convert placeholders to macro parameter chars    \let\hash=##% convert placeholders to macro parameter chars
   \ifrecursive    \ifnum\paramno=1
     \ifcase\paramno      \def\xeatspaces##1{##1}%
     % 0      % This removes the pair of braces around the argument.  We don't
       \expandafter\xdef\csname\the\macname\endcsname{%      % use \eatspaces, because this can cause ends of lines to be lost
         \noexpand\scanmacro{\temp}}%      % when the argument to \eatspaces is read, leading to line-based
     \or % 1      % commands like "@itemize" not being read correctly.
       \expandafter\xdef\csname\the\macname\endcsname{%  
          \bgroup\noexpand\macroargctxt  
          \noexpand\braceorline  
          \expandafter\noexpand\csname\the\macname xxx\endcsname}%  
       \expandafter\xdef\csname\the\macname xxx\endcsname##1{%  
          \egroup\noexpand\scanmacro{\temp}}%  
     \else % many  
       \expandafter\xdef\csname\the\macname\endcsname{%  
          \bgroup\noexpand\macroargctxt  
          \noexpand\csname\the\macname xx\endcsname}%  
       \expandafter\xdef\csname\the\macname xx\endcsname##1{%  
           \expandafter\noexpand\csname\the\macname xxx\endcsname ##1,}%  
       \expandafter\expandafter  
       \expandafter\xdef  
       \expandafter\expandafter  
         \csname\the\macname xxx\endcsname  
           \paramlist{\egroup\noexpand\scanmacro{\temp}}%  
     \fi  
   \else    \else
     \ifcase\paramno      \let\xeatspaces\relax % suppress expansion
     % 0    \fi
     \ifcase\paramno
     % 0
       \expandafter\xdef\csname\the\macname\endcsname{%
         \noexpand\scanmacro{\macrobody}}%
     \or % 1
       \expandafter\xdef\csname\the\macname\endcsname{%
          \bgroup
          \noexpand\braceorline
          \expandafter\noexpand\csname\the\macname @@@\endcsname}%
       \expandafter\xdef\csname\the\macname @@@\endcsname##1{%
         \egroup
         \noexpand\scanmacro{\macrobody}%
         }%
     \else % at most 9
       \ifnum\paramno<10\relax
         % @MACNAME sets the context for reading the macro argument
         % @MACNAME@@ gets the argument, processes backslashes and appends a
         % comma.
         % @MACNAME@@@ removes braces surrounding the argument list.
         % @MACNAME@@@@ scans the macro body with arguments substituted.
       \expandafter\xdef\csname\the\macname\endcsname{%        \expandafter\xdef\csname\the\macname\endcsname{%
         \noexpand\norecurse{\the\macname}%          \bgroup
         \noexpand\scanmacro{\temp}\egroup}%          \noexpand\expandafter  % This \expandafter skip any spaces after the
     \or % 1          \noexpand\macroargctxt % macro before we change the catcode of space.
       \expandafter\xdef\csname\the\macname\endcsname{%          \noexpand\expandafter
          \bgroup\noexpand\macroargctxt          \expandafter\noexpand\csname\the\macname @@\endcsname}%
          \noexpand\braceorline        \expandafter\xdef\csname\the\macname @@\endcsname##1{%
          \expandafter\noexpand\csname\the\macname xxx\endcsname}%            \noexpand\passargtomacro
       \expandafter\xdef\csname\the\macname xxx\endcsname##1{%            \expandafter\noexpand\csname\the\macname @@@\endcsname{##1,}}%
         \egroup        \expandafter\xdef\csname\the\macname @@@\endcsname##1{%
         \noexpand\norecurse{\the\macname}%            \expandafter\noexpand\csname\the\macname @@@@\endcsname ##1}%
         \noexpand\scanmacro{\temp}\egroup}%  
     \else % many  
       \expandafter\xdef\csname\the\macname\endcsname{%  
          \bgroup\noexpand\macroargctxt  
          \expandafter\noexpand\csname\the\macname xx\endcsname}%  
       \expandafter\xdef\csname\the\macname xx\endcsname##1{%  
           \expandafter\noexpand\csname\the\macname xxx\endcsname ##1,}%  
       \expandafter\expandafter        \expandafter\expandafter
       \expandafter\xdef        \expandafter\xdef
       \expandafter\expandafter        \expandafter\expandafter
       \csname\the\macname xxx\endcsname          \csname\the\macname @@@@\endcsname\paramlist{%
       \paramlist{%            \egroup\noexpand\scanmacro{\macrobody}}%
           \egroup      \else % 10 or more:
           \noexpand\norecurse{\the\macname}%        \expandafter\xdef\csname\the\macname\endcsname{%
           \noexpand\scanmacro{\temp}\egroup}%          \noexpand\getargvals@{\the\macname}{\argl}%
         }%
         \global\expandafter\let\csname mac.\the\macname .body\endcsname\macrobody
         \global\expandafter\let\csname mac.\the\macname .recurse\endcsname\gobble
     \fi      \fi
   \fi}    \fi}
   
   \catcode `\@\texiatcatcode\relax % end private-to-Texinfo catcodes
   
 \def\norecurse#1{\bgroup\cslet{#1}{macsave.#1}}  \def\norecurse#1{\bgroup\cslet{#1}{macsave.#1}}
   
 % \braceorline decides whether the next nonwhitespace character is a  
 % {.  If so it reads up to the closing }, if not, it reads the whole  
 % line.  Whatever was read is then fed to the next control sequence  
 % as an argument (by \parsebrace or \parsearg)  
 \def\braceorline#1{\let\next=#1\futurelet\nchar\braceorlinexxx}  
 \def\braceorlinexxx{%  
   \ifx\nchar\bgroup\else  
     \expandafter\parsearg  
   \fi \next}  
   
 % We mant to disable all macros during \shipout so that they are not  %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
 % expanded by \write.  %
 \def\turnoffmacros{\begingroup \def\do##1{\let\noexpand##1=\relax}%  {\catcode`\@=0 \catcode`\\=13  % We need to manipulate \ so use @ as escape
   \edef\next{\macrolist}\expandafter\endgroup\next}  @catcode`@_=11  % private names
   @catcode`@!=11  % used as argument separator
   
   % \passargtomacro#1#2 -
   % Call #1 with a list of tokens #2, with any doubled backslashes in #2
   % compressed to one.
   %
   % This implementation works by expansion, and not execution (so we cannot use
   % \def or similar).  This reduces the risk of this failing in contexts where
   % complete expansion is done with no execution (for example, in writing out to
   % an auxiliary file for an index entry).
   %
   % State is kept in the input stream: the argument passed to
   % @look_ahead, @gobble_and_check_finish and @add_segment is
   %
   % THE_MACRO ARG_RESULT ! {PENDING_BS} NEXT_TOKEN  (... rest of input)
   %
   % where:
   % THE_MACRO - name of the macro we want to call
   % ARG_RESULT - argument list we build to pass to that macro
   % PENDING_BS - either a backslash or nothing
   % NEXT_TOKEN - used to look ahead in the input stream to see what's coming next
   
   @gdef@passargtomacro#1#2{%
     @add_segment #1!{}@relax#2\@_finish\%
   }
   @gdef@_finish{@_finishx} @global@let@_finishx@relax
   
   % #1 - THE_MACRO ARG_RESULT
   % #2 - PENDING_BS
   % #3 - NEXT_TOKEN
   % #4 used to look ahead
   %
   % If the next token is not a backslash, process the rest of the argument;
   % otherwise, remove the next token.
   @gdef@look_ahead#1!#2#3#4{%
     @ifx#4\%
      @expandafter@gobble_and_check_finish
     @else
      @expandafter@add_segment
     @fi#1!{#2}#4#4%
   }
   
   % #1 - THE_MACRO ARG_RESULT
   % #2 - PENDING_BS
   % #3 - NEXT_TOKEN
   % #4 should be a backslash, which is gobbled.
   % #5 looks ahead
   %
   % Double backslash found.  Add a single backslash, and look ahead.
   @gdef@gobble_and_check_finish#1!#2#3#4#5{%
     @add_segment#1\!{}#5#5%
   }
   
   @gdef@is_fi{@fi}
   
   % #1 - THE_MACRO ARG_RESULT
   % #2 - PENDING_BS
   % #3 - NEXT_TOKEN
   % #4 is input stream until next backslash
   %
   % Input stream is either at the start of the argument, or just after a
   % backslash sequence, either a lone backslash, or a doubled backslash.
   % NEXT_TOKEN contains the first token in the input stream: if it is \finish,
   % finish; otherwise, append to ARG_RESULT the segment of the argument up until
   % the next backslash.  PENDING_BACKSLASH contains a backslash to represent
   % a backslash just before the start of the input stream that has not been
   % added to ARG_RESULT.
   @gdef@add_segment#1!#2#3#4\{%
   @ifx#3@_finish
     @call_the_macro#1!%
   @else
     % append the pending backslash to the result, followed by the next segment
     @expandafter@is_fi@look_ahead#1#2#4!{\}@fi
     % this @fi is discarded by @look_ahead.
     % we can't get rid of it with \expandafter because we don't know how
     % long #4 is.
   }
   
   % #1 - THE_MACRO
   % #2 - ARG_RESULT
   % #3 discards the res of the conditional in @add_segment, and @is_fi ends the
   % conditional.
   @gdef@call_the_macro#1#2!#3@fi{@is_fi #1{#2}}
   
   }
   %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
   
   % \braceorline MAC is used for a one-argument macro MAC.  It checks
   % whether the next non-whitespace character is a {.  It sets the context
   % for reading the argument (slightly different in the two cases).  Then,
   % to read the argument, in the whole-line case, it then calls the regular
   % \parsearg MAC; in the lbrace case, it calls \passargtomacro MAC.
   %
   \def\braceorline#1{\let\macnamexxx=#1\futurelet\nchar\braceorlinexxx}
   \def\braceorlinexxx{%
     \ifx\nchar\bgroup
       \macroargctxt
       \expandafter\passargtomacro
     \else
       \macrolineargctxt\expandafter\parsearg
     \fi \macnamexxx}
   
   
 % @alias.  % @alias.
 % We need some trickery to remove the optional spaces around the equal  % We need some trickery to remove the optional spaces around the equal
 % sign.  Just make them active and then expand them all to nothing.  % sign.  Make them active and then expand them all to nothing.
 \def\alias{\begingroup\obeyspaces\parsearg\aliasxxx}  %
   \def\alias{\parseargusing\obeyspaces\aliasxxx}
 \def\aliasxxx #1{\aliasyyy#1\relax}  \def\aliasxxx #1{\aliasyyy#1\relax}
 \def\aliasyyy #1=#2\relax{\ignoreactivespaces  \def\aliasyyy #1=#2\relax{%
 \edef\next{\global\let\expandafter\noexpand\csname#1\endcsname=%    {%
            \expandafter\noexpand\csname#2\endcsname}%      \expandafter\let\obeyedspace=\empty
 \expandafter\endgroup\next}      \addtomacrolist{#1}%
       \xdef\next{\global\let\makecsname{#1}=\makecsname{#2}}%
     }%
     \next
   }
   
   
 \message{cross references,}  \message{cross references,}
 % @xref etc.  
   
 \newwrite\auxfile  \newwrite\auxfile
   
 \newif\ifhavexrefs    % True if xref values are known.  \newif\ifhavexrefs    % True if xref values are known.
 \newif\ifwarnedxrefs  % True if we warned once that they aren't known.  \newif\ifwarnedxrefs  % True if we warned once that they aren't known.
   
 % @inforef is relatively simple.  % @inforef is relatively simple.
 \def\inforef #1{\inforefzzz #1,,,,**}  \def\inforef #1{\inforefzzz #1,,,,**}
 \def\inforefzzz #1,#2,#3,#4**{\putwordSee{} \putwordInfo{} \putwordfile{} \file{\ignorespaces #3{}},  \def\inforefzzz #1,#2,#3,#4**{%
     \putwordSee{} \putwordInfo{} \putwordfile{} \file{\ignorespaces #3{}},
   node \samp{\ignorespaces#1{}}}    node \samp{\ignorespaces#1{}}}
   
 % @node's job is to define \lastnode.  % @node's only job in TeX is to define \lastnode, which is used in
 \def\node{\ENVcheck\parsearg\nodezzz}  % cross-references.  The @node line might or might not have commas, and
 \def\nodezzz#1{\nodexxx [#1,]}  % might or might not have spaces before the first comma, like:
 \def\nodexxx[#1,#2]{\gdef\lastnode{#1}}  % @node foo , bar , ...
   % We don't want such trailing spaces in the node name.
   %
   \parseargdef\node{\checkenv{}\donode #1 ,\finishnodeparse}
   %
   % also remove a trailing comma, in case of something like this:
   % @node Help-Cross,  ,  , Cross-refs
   \def\donode#1 ,#2\finishnodeparse{\dodonode #1,\finishnodeparse}
   \def\dodonode#1,#2\finishnodeparse{\gdef\lastnode{#1}}
   
 \let\nwnode=\node  \let\nwnode=\node
 \let\lastnode=\relax  \let\lastnode=\empty
   
 % The sectioning commands (@chapter, etc.) call these.  % Write a cross-reference definition for the current node.  #1 is the
 \def\donoderef{%  % type (Ynumbered, Yappendix, Ynothing).
   \ifx\lastnode\relax\else  %
     \expandafter\expandafter\expandafter\setref{\lastnode}%  \def\donoderef#1{%
       {Ysectionnumberandtype}%    \ifx\lastnode\empty\else
     \global\let\lastnode=\relax      \setref{\lastnode}{#1}%
       \global\let\lastnode=\empty
   \fi    \fi
 }  }
 \def\unnumbnoderef{%  
   \ifx\lastnode\relax\else  
     \expandafter\expandafter\expandafter\setref{\lastnode}{Ynothing}%  
     \global\let\lastnode=\relax  
   \fi  
 }  
 \def\appendixnoderef{%  
   \ifx\lastnode\relax\else  
     \expandafter\expandafter\expandafter\setref{\lastnode}%  
       {Yappendixletterandtype}%  
     \global\let\lastnode=\relax  
   \fi  
 }  
   
   
 % @anchor{NAME} -- define xref target at arbitrary point.  % @anchor{NAME} -- define xref target at arbitrary point.
 %  %
 \newcount\savesfregister  \newcount\savesfregister
 \gdef\savesf{\relax \ifhmode \savesfregister=\spacefactor \fi}  %
 \gdef\restoresf{\relax \ifhmode \spacefactor=\savesfregister \fi}  \def\savesf{\relax \ifhmode \savesfregister=\spacefactor \fi}
 \gdef\anchor#1{\savesf \setref{#1}{Ynothing}\restoresf \ignorespaces}  \def\restoresf{\relax \ifhmode \spacefactor=\savesfregister \fi}
   \def\anchor#1{\savesf \setref{#1}{Ynothing}\restoresf \ignorespaces}
   
 % \setref{NAME}{SNT} defines a cross-reference point NAME, namely  % \setref{NAME}{SNT} defines a cross-reference point NAME (a node or an
 % NAME-title, NAME-pg, and NAME-SNT.  Called from \foonoderef.  We have  % anchor), which consists of three parts:
 % to set \indexdummies so commands such as @code in a section title  % 1) NAME-title - the current sectioning name taken from \lastsection,
 % aren't expanded.  It would be nicer not to expand the titles in the  %                 or the anchor name.
 % first place, but there's so many layers that that is hard to do.  % 2) NAME-snt   - section number and type, passed as the SNT arg, or
   %                 empty for anchors.
   % 3) NAME-pg    - the page number.
 %  %
 \def\setref#1#2{{%  % This is called from \donoderef, \anchor, and \dofloat.  In the case of
   \indexdummies  % floats, there is an additional part, which is not written here:
   % 4) NAME-lof   - the text as it should appear in a @listoffloats.
   %
   \def\setref#1#2{%
   \pdfmkdest{#1}%    \pdfmkdest{#1}%
   \dosetq{#1-title}{Ytitle}%    \iflinks
   \dosetq{#1-pg}{Ypagenumber}%      {%
   \dosetq{#1-snt}{#2}%        \requireauxfile
 }}        \atdummies  % preserve commands, but don't expand them
         % match definition in \xrdef, \refx, \xrefX.
         \def\value##1{##1}%
         \edef\writexrdef##1##2{%
           \write\auxfile{@xrdef{#1-% #1 of \setref, expanded by the \edef
             ##1}{##2}}% these are parameters of \writexrdef
         }%
         \toks0 = \expandafter{\lastsection}%
         \immediate \writexrdef{title}{\the\toks0 }%
         \immediate \writexrdef{snt}{\csname #2\endcsname}% \Ynumbered etc.
         \safewhatsit{\writexrdef{pg}{\folio}}% will be written later, at \shipout
       }%
     \fi
   }
   
   % @xrefautosectiontitle on|off says whether @section(ing) names are used
   % automatically in xrefs, if the third arg is not explicitly specified.
   % This was provided as a "secret" @set xref-automatic-section-title
   % variable, now it's official.
   %
   \parseargdef\xrefautomaticsectiontitle{%
     \def\temp{#1}%
     \ifx\temp\onword
       \expandafter\let\csname SETxref-automatic-section-title\endcsname
         = \empty
     \else\ifx\temp\offword
       \expandafter\let\csname SETxref-automatic-section-title\endcsname
         = \relax
     \else
       \errhelp = \EMsimple
       \errmessage{Unknown @xrefautomaticsectiontitle value `\temp',
                   must be on|off}%
     \fi\fi
   }
   
   %
 % @xref, @pxref, and @ref generate cross-references.  For \xrefX, #1 is  % @xref, @pxref, and @ref generate cross-references.  For \xrefX, #1 is
 % the node name, #2 the name of the Info cross-reference, #3 the printed  % the node name, #2 the name of the Info cross-reference, #3 the printed
 % node name, #4 the name of the Info file, #5 the name of the printed  % node name, #4 the name of the Info file, #5 the name of the printed
 % manual.  All but the node name can be omitted.  % manual.  All but the node name can be omitted.
 %  %
 \def\pxref#1{\putwordsee{} \xrefX[#1,,,,,,,]}  \def\pxref{\putwordsee{} \xrefXX}
 \def\xref#1{\putwordSee{} \xrefX[#1,,,,,,,]}  \def\xref{\putwordSee{} \xrefXX}
 \def\ref#1{\xrefX[#1,,,,,,,]}  \def\ref{\xrefXX}
   
   \def\xrefXX#1{\def\xrefXXarg{#1}\futurelet\tokenafterxref\xrefXXX}
   \def\xrefXXX{\expandafter\xrefX\expandafter[\xrefXXarg,,,,,,,]}
   %
   \newbox\toprefbox
   \newbox\printedrefnamebox
   \newbox\infofilenamebox
   \newbox\printedmanualbox
   %
 \def\xrefX[#1,#2,#3,#4,#5,#6]{\begingroup  \def\xrefX[#1,#2,#3,#4,#5,#6]{\begingroup
   \unsepspaces    \unsepspaces
     %
     % Get args without leading/trailing spaces.
     \def\printedrefname{\ignorespaces #3}%
     \setbox\printedrefnamebox = \hbox{\printedrefname\unskip}%
     %
     \def\infofilename{\ignorespaces #4}%
     \setbox\infofilenamebox = \hbox{\infofilename\unskip}%
     %
   \def\printedmanual{\ignorespaces #5}%    \def\printedmanual{\ignorespaces #5}%
   \def\printednodename{\ignorespaces #3}%    \setbox\printedmanualbox  = \hbox{\printedmanual\unskip}%
   \setbox1=\hbox{\printedmanual}%    %
   \setbox0=\hbox{\printednodename}%    % If the printed reference name (arg #3) was not explicitly given in
   \ifdim \wd0 = 0pt    % the @xref, figure out what we want to use.
     \ifdim \wd\printedrefnamebox = 0pt
     % No printed node name was explicitly given.      % No printed node name was explicitly given.
     \expandafter\ifx\csname SETxref-automatic-section-title\endcsname\relax      \expandafter\ifx\csname SETxref-automatic-section-title\endcsname \relax
       % Use the node name inside the square brackets.        % Not auto section-title: use node name inside the square brackets.
       \def\printednodename{\ignorespaces #1}%        \def\printedrefname{\ignorespaces #1}%
     \else      \else
       % Use the actual chapter/section title appear inside        % Auto section-title: use chapter/section title inside
       % the square brackets.  Use the real section title if we have it.        % the square brackets if we have it.
       \ifdim \wd1 > 0pt        \ifdim \wd\printedmanualbox > 0pt
         % It is in another manual, so we don't have it.          % It is in another manual, so we don't have it; use node name.
         \def\printednodename{\ignorespaces #1}%          \def\printedrefname{\ignorespaces #1}%
       \else        \else
         \ifhavexrefs          \ifhavexrefs
           % We know the real title if we have the xref values.            % We (should) know the real title if we have the xref values.
           \def\printednodename{\refx{#1-title}{}}%            \def\printedrefname{\refx{#1-title}{}}%
         \else          \else
           % Otherwise just copy the Info node name.            % Otherwise just copy the Info node name.
           \def\printednodename{\ignorespaces #1}%            \def\printedrefname{\ignorespaces #1}%
         \fi%          \fi%
       \fi        \fi
     \fi      \fi
   \fi    \fi
   %    %
   % If we use \unhbox0 and \unhbox1 to print the node names, TeX does not    % Make link in pdf output.
   % insert empty discretionaries after hyphens, which means that it will  
   % not find a line break at a hyphen in a node names.  Since some manuals  
   % are best written with fairly long node names, containing hyphens, this  
   % is a loss.  Therefore, we give the text of the node name again, so it  
   % is as if TeX is seeing it for the first time.  
   \ifpdf    \ifpdf
     \leavevmode      % For pdfTeX and LuaTeX
     \getfilename{#4}%      {\indexnofonts
     \ifnum\filenamelength>0       \makevalueexpandable
       \startlink attr{/Border [0 0 0]}%       \turnoffactive
         goto file{\the\filename.pdf} name{#1@}%       % This expands tokens, so do it after making catcode changes, so _
        % etc. don't get their TeX definitions.  This ignores all spaces in
        % #4, including (wrongly) those in the middle of the filename.
        \getfilename{#4}%
        %
        % This (wrongly) does not take account of leading or trailing
        % spaces in #1, which should be ignored.
        \setpdfdestname{#1}%
        %
        \ifx\pdfdestname\empty
          \def\pdfdestname{Top}% no empty targets
        \fi
        %
        \leavevmode
        \startlink attr{/Border [0 0 0]}%
        \ifnum\filenamelength>0
          goto file{\the\filename.pdf} name{\pdfdestname}%
        \else
          goto name{\pdfmkpgn{\pdfdestname}}%
        \fi
       }%
       \setcolor{\linkcolor}%
     \else
       \ifx\XeTeXrevision\thisisundefined
     \else      \else
       \startlink attr{/Border [0 0 0]}%        % For XeTeX
         goto name{#1@}%        {\indexnofonts
          \makevalueexpandable
          \turnoffactive
          % This expands tokens, so do it after making catcode changes, so _
          % etc. don't get their TeX definitions.  This ignores all spaces in
          % #4, including (wrongly) those in the middle of the filename.
          \getfilename{#4}%
          %
          % This (wrongly) does not take account of leading or trailing
          % spaces in #1, which should be ignored.
          \setpdfdestname{#1}%
          %
          \ifx\pdfdestname\empty
            \def\pdfdestname{Top}% no empty targets
          \fi
          %
          \leavevmode
          \ifnum\filenamelength>0
            % With default settings,
            % XeTeX (xdvipdfmx) replaces link destination names with integers.
            % In this case, the replaced destination names of
            % remote PDFs are no longer known.  In order to avoid a replacement,
            % you can use xdvipdfmx's command line option `-C 0x0010'.
            % If you use XeTeX 0.99996+ (TeX Live 2016+),
            % this command line option is no longer necessary
            % because we can use the `dvipdfmx:config' special.
            \special{pdf:bann << /Border [0 0 0] /Type /Annot /Subtype /Link /A
              << /S /GoToR /F (\the\filename.pdf) /D (\pdfdestname) >> >>}%
          \else
            \special{pdf:bann << /Border [0 0 0] /Type /Annot /Subtype /Link /A
              << /S /GoTo /D (\pdfdestname) >> >>}%
          \fi
         }%
         \setcolor{\linkcolor}%
     \fi      \fi
     \linkcolor  
   \fi    \fi
     {%
       % Have to otherify everything special to allow the \csname to
       % include an _ in the xref name, etc.
       \indexnofonts
       \turnoffactive
       \def\value##1{##1}%
       \expandafter\global\expandafter\let\expandafter\Xthisreftitle
         \csname XR#1-title\endcsname
     }%
   %    %
   \ifdim \wd1 > 0pt    % Float references are printed completely differently: "Figure 1.2"
     \putwordsection{} ``\printednodename'' \putwordin{} \cite{\printedmanual}%    % instead of "[somenode], p.3".  \iffloat distinguishes them by
     % \Xthisreftitle being set to a magic string.
     \iffloat\Xthisreftitle
       % If the user specified the print name (third arg) to the ref,
       % print it instead of our usual "Figure 1.2".
       \ifdim\wd\printedrefnamebox = 0pt
         \refx{#1-snt}{}%
       \else
         \printedrefname
       \fi
       %
       % If the user also gave the printed manual name (fifth arg), append
       % "in MANUALNAME".
       \ifdim \wd\printedmanualbox > 0pt
         \space \putwordin{} \cite{\printedmanual}%
       \fi
   \else    \else
     % _ (for example) has to be the character _ for the purposes of the      % node/anchor (non-float) references.
     % control sequence corresponding to the node, but it has to expand      %
     % into the usual \leavevmode...\vrule stuff for purposes of      % If we use \unhbox to print the node names, TeX does not insert
     % printing. So we \turnoffactive for the \refx-snt, back on for the      % empty discretionaries after hyphens, which means that it will not
     % printing, back off for the \refx-pg.      % find a line break at a hyphen in a node names.  Since some manuals
     {\normalturnoffactive      % are best written with fairly long node names, containing hyphens,
      % Only output a following space if the -snt ref is nonempty; for      % this is a loss.  Therefore, we give the text of the node name
      % @unnumbered and @anchor, it won't be.      % again, so it is as if TeX is seeing it for the first time.
      \setbox2 = \hbox{\ignorespaces \refx{#1-snt}{}}%      %
      \ifdim \wd2 > 0pt \refx{#1-snt}\space\fi      \ifdim \wd\printedmanualbox > 0pt
     }%        % Cross-manual reference with a printed manual name.
     % [mynode],        %
     [\printednodename],\space        \crossmanualxref{\cite{\printedmanual\unskip}}%
     % page 3      %
     \turnoffactive \putwordpage\tie\refx{#1-pg}{}%      \else\ifdim \wd\infofilenamebox > 0pt
         % Cross-manual reference with only an info filename (arg 4), no
         % printed manual name (arg 5).  This is essentially the same as
         % the case above; we output the filename, since we have nothing else.
         %
         \crossmanualxref{\code{\infofilename\unskip}}%
       %
       \else
         % Reference within this manual.
         %
         % _ (for example) has to be the character _ for the purposes of the
         % control sequence corresponding to the node, but it has to expand
         % into the usual \leavevmode...\vrule stuff for purposes of
         % printing. So we \turnoffactive for the \refx-snt, back on for the
         % printing, back off for the \refx-pg.
         {\turnoffactive
          % Only output a following space if the -snt ref is nonempty; for
          % @unnumbered and @anchor, it won't be.
          \setbox2 = \hbox{\ignorespaces \refx{#1-snt}{}}%
          \ifdim \wd2 > 0pt \refx{#1-snt}\space\fi
         }%
         % output the `[mynode]' via the macro below so it can be overridden.
         \xrefprintnodename\printedrefname
         %
         % But we always want a comma and a space:
         ,\space
         %
         % output the `page 3'.
         \turnoffactive \putwordpage\tie\refx{#1-pg}{}%
         % Add a , if xref followed by a space
         \if\space\noexpand\tokenafterxref ,%
         \else\ifx\        \tokenafterxref ,% @TAB
         \else\ifx\*\tokenafterxref ,%   @*
         \else\ifx\ \tokenafterxref ,%   @SPACE
         \else\ifx\
                   \tokenafterxref ,%    @NL
         \else\ifx\tie\tokenafterxref ,% @tie
         \fi\fi\fi\fi\fi\fi
       \fi\fi
   \fi    \fi
   \endlink    \endlink
 \endgroup}  \endgroup}
   
 % \dosetq is the interface for calls from other macros  % Output a cross-manual xref to #1.  Used just above (twice).
   %
 % Use \normalturnoffactive so that punctuation chars such as underscore  % Only include the text "Section ``foo'' in" if the foo is neither
 % and backslash work in node names.  (\turnoffactive doesn't do \.)  % missing or Top.  Thus, @xref{,,,foo,The Foo Manual} outputs simply
 \def\dosetq#1#2{%  % "see The Foo Manual", the idea being to refer to the whole manual.
   {\let\folio=0%  %
    \normalturnoffactive  % But, this being TeX, we can't easily compare our node name against the
    \edef\next{\write\auxfile{\internalsetq{#1}{#2}}}%  % string "Top" while ignoring the possible spaces before and after in
    \iflinks  % the input.  By adding the arbitrary 7sp below, we make it much less
      \next  % likely that a real node name would have the same width as "Top" (e.g.,
    \fi  % in a monospaced font).  Hopefully it will never happen in practice.
   }%  %
   % For the same basic reason, we retypeset the "Top" at every
   % reference, since the current font is indeterminate.
   %
   \def\crossmanualxref#1{%
     \setbox\toprefbox = \hbox{Top\kern7sp}%
     \setbox2 = \hbox{\ignorespaces \printedrefname \unskip \kern7sp}%
     \ifdim \wd2 > 7sp  % nonempty?
       \ifdim \wd2 = \wd\toprefbox \else  % same as Top?
         \putwordSection{} ``\printedrefname'' \putwordin{}\space
       \fi
     \fi
     #1%
 }  }
   
 % \internalsetq {foo}{page} expands into  % This macro is called from \xrefX for the `[nodename]' part of xref
 % CHARACTERS 'xrdef {foo}{...expansion of \Ypage...}  % output.  It's a separate macro only so it can be changed more easily,
 % When the aux file is read, ' is the escape character  % since square brackets don't work well in some documents.  Particularly
   % one that Bob is working on :).
   %
   \def\xrefprintnodename#1{[#1]}
   
 \def\internalsetq #1#2{'xrdef {#1}{\csname #2\endcsname}}  % Things referred to by \setref.
   %
 % Things to be expanded by \internalsetq  
   
 \def\Ypagenumber{\folio}  
   
 \def\Ytitle{\thissection}  
   
 \def\Ynothing{}  \def\Ynothing{}
   \def\Yomitfromtoc{}
   \def\Ynumbered{%
     \ifnum\secno=0
       \putwordChapter@tie \the\chapno
     \else \ifnum\subsecno=0
       \putwordSection@tie \the\chapno.\the\secno
     \else \ifnum\subsubsecno=0
       \putwordSection@tie \the\chapno.\the\secno.\the\subsecno
     \else
       \putwordSection@tie \the\chapno.\the\secno.\the\subsecno.\the\subsubsecno
     \fi\fi\fi
   }
   \def\Yappendix{%
     \ifnum\secno=0
        \putwordAppendix@tie @char\the\appendixno{}%
     \else \ifnum\subsecno=0
        \putwordSection@tie @char\the\appendixno.\the\secno
     \else \ifnum\subsubsecno=0
       \putwordSection@tie @char\the\appendixno.\the\secno.\the\subsecno
     \else
       \putwordSection@tie
         @char\the\appendixno.\the\secno.\the\subsecno.\the\subsubsecno
     \fi\fi\fi
   }
   
 \def\Ysectionnumberandtype{%  % \refx{NAME}{SUFFIX} - reference a cross-reference string named NAME.  SUFFIX
 \ifnum\secno=0 \putwordChapter\xreftie\the\chapno %  % is output afterwards if non-empty.
 \else \ifnum \subsecno=0 \putwordSection\xreftie\the\chapno.\the\secno %  
 \else \ifnum \subsubsecno=0 %  
 \putwordSection\xreftie\the\chapno.\the\secno.\the\subsecno %  
 \else %  
 \putwordSection\xreftie\the\chapno.\the\secno.\the\subsecno.\the\subsubsecno %  
 \fi \fi \fi }  
   
 \def\Yappendixletterandtype{%  
 \ifnum\secno=0 \putwordAppendix\xreftie'char\the\appendixno{}%  
 \else \ifnum \subsecno=0 \putwordSection\xreftie'char\the\appendixno.\the\secno %  
 \else \ifnum \subsubsecno=0 %  
 \putwordSection\xreftie'char\the\appendixno.\the\secno.\the\subsecno %  
 \else %  
 \putwordSection\xreftie'char\the\appendixno.\the\secno.\the\subsecno.\the\subsubsecno %  
 \fi \fi \fi }  
   
 \gdef\xreftie{'tie}  
   
 % Use TeX 3.0's \inputlineno to get the line number, for better error  
 % messages, but if we're using an old version of TeX, don't do anything.  
 %  
 \ifx\inputlineno\thisisundefined  
   \let\linenumber = \empty % Non-3.0.  
 \else  
   \def\linenumber{\the\inputlineno:\space}  
 \fi  
   
 % Define \refx{NAME}{SUFFIX} to reference a cross-reference string named NAME.  
 % If its value is nonempty, SUFFIX is output afterward.  
   
 \def\refx#1#2{%  \def\refx#1#2{%
   \expandafter\ifx\csname X#1\endcsname\relax    \requireauxfile
     {%
       \indexnofonts
       \otherbackslash
       \def\value##1{##1}%
       \expandafter\global\expandafter\let\expandafter\thisrefX
         \csname XR#1\endcsname
     }%
     \ifx\thisrefX\relax
     % If not defined, say something at least.      % If not defined, say something at least.
     \angleleft un\-de\-fined\angleright      \angleleft un\-de\-fined\angleright
     \iflinks      \iflinks
       \ifhavexrefs        \ifhavexrefs
         \message{\linenumber Undefined cross reference `#1'.}%          {\toks0 = {#1}% avoid expansion of possibly-complex value
            \message{\linenumber Undefined cross reference `\the\toks0'.}}%
       \else        \else
         \ifwarnedxrefs\else          \ifwarnedxrefs\else
           \global\warnedxrefstrue            \global\warnedxrefstrue
Line 5352  width0pt\relax} \fi
Line 9147  width0pt\relax} \fi
     \fi      \fi
   \else    \else
     % It's defined, so just use it.      % It's defined, so just use it.
     \csname X#1\endcsname      \thisrefX
   \fi    \fi
   #2% Output the suffix in any case.    #2% Output the suffix in any case.
 }  }
   
 % This is the macro invoked by entries in the aux file.  % This is the macro invoked by entries in the aux file.  Define a control
   % sequence for a cross-reference target (we prepend XR to the control sequence
   % name to avoid collisions).  The value is the page number.  If this is a float
   % type, we have more work to do.
 %  %
 \def\xrdef#1{\begingroup  \def\xrdef#1#2{%
   % Reenable \ as an escape while reading the second argument.    {% Expand the node or anchor name to remove control sequences.
   \catcode`\\ = 0     % \turnoffactive stops 8-bit characters being changed to commands
   \afterassignment\endgroup     % like @'e.  \refx does the same to retrieve the value in the definition.
   \expandafter\gdef\csname X#1\endcsname      \indexnofonts
       \turnoffactive
       \def\value##1{##1}%
       \xdef\safexrefname{#1}%
     }%
     %
     \expandafter\gdef\csname XR\safexrefname\endcsname{#2}% remember this xref
     %
     % Was that xref control sequence that we just defined for a float?
     \expandafter\iffloat\csname XR\safexrefname\endcsname
       % it was a float, and we have the (safe) float type in \iffloattype.
       \expandafter\let\expandafter\floatlist
         \csname floatlist\iffloattype\endcsname
       %
       % Is this the first time we've seen this float type?
       \expandafter\ifx\floatlist\relax
         \toks0 = {\do}% yes, so just \do
       \else
         % had it before, so preserve previous elements in list.
         \toks0 = \expandafter{\floatlist\do}%
       \fi
       %
       % Remember this xref in the control sequence \floatlistFLOATTYPE,
       % for later use in \listoffloats.
       \expandafter\xdef\csname floatlist\iffloattype\endcsname{\the\toks0
         {\safexrefname}}%
     \fi
 }  }
   
   % If working on a large document in chapters, it is convenient to
   % be able to disable indexing, cross-referencing, and contents, for test runs.
   % This is done with @novalidate at the beginning of the file.
   %
   \newif\iflinks \linkstrue % by default we want the aux files.
   \let\novalidate = \linksfalse
   
   % Used when writing to the aux file, or when using data from it.
   \def\requireauxfile{%
     \iflinks
       \tryauxfile
       % Open the new aux file.  TeX will close it automatically at exit.
       \immediate\openout\auxfile=\jobname.aux
     \fi
     \global\let\requireauxfile=\relax   % Only do this once.
   }
   
 % Read the last existing aux file, if any.  No error if none exists.  % Read the last existing aux file, if any.  No error if none exists.
 \def\readauxfile{\begingroup  %
   \def\tryauxfile{%
     \openin 1 \jobname.aux
     \ifeof 1 \else
       \readdatafile{aux}%
       \global\havexrefstrue
     \fi
     \closein 1
   }
   
   \def\setupdatafile{%
   \catcode`\^^@=\other    \catcode`\^^@=\other
   \catcode`\^^A=\other    \catcode`\^^A=\other
   \catcode`\^^B=\other    \catcode`\^^B=\other
Line 5395  width0pt\relax} \fi
Line 9246  width0pt\relax} \fi
   \catcode`\^^]=\other    \catcode`\^^]=\other
   \catcode`\^^^=\other    \catcode`\^^^=\other
   \catcode`\^^_=\other    \catcode`\^^_=\other
   \catcode`\@=\other    % It was suggested to set the catcode of ^ to 7, which would allow ^^e4 etc.
   \catcode`\^=\other  
   % It was suggested to define this as 7, which would allow ^^e4 etc.  
   % in xref tags, i.e., node names.  But since ^^e4 notation isn't    % in xref tags, i.e., node names.  But since ^^e4 notation isn't
   % supported in the main text, it doesn't seem desirable.  Furthermore,    % supported in the main text, it doesn't seem desirable.  Furthermore,
   % that is not enough: for node names that actually contain a ^    % that is not enough: for node names that actually contain a ^
Line 5410  width0pt\relax} \fi
Line 9259  width0pt\relax} \fi
   % \def\auxhat{\def^{'hat }}% extra space so ok if followed by letter    % \def\auxhat{\def^{'hat }}% extra space so ok if followed by letter
   % and then to call \auxhat in \setq.    % and then to call \auxhat in \setq.
   %    %
     \catcode`\^=\other
     %
     % Special characters.  Should be turned off anyway, but...
   \catcode`\~=\other    \catcode`\~=\other
   \catcode`\[=\other    \catcode`\[=\other
   \catcode`\]=\other    \catcode`\]=\other
Line 5421  width0pt\relax} \fi
Line 9273  width0pt\relax} \fi
   \catcode`\$=\other    \catcode`\$=\other
   \catcode`\#=\other    \catcode`\#=\other
   \catcode`\&=\other    \catcode`\&=\other
   \catcode`+=\other % avoid \+ for paranoia even though we've turned it off  
   % Make the characters 128-255 be printing characters  
   {%  
     \count 1=128  
     \def\loop{%  
       \catcode\count 1=\other  
       \advance\count 1 by 1  
       \ifnum \count 1<256 \loop \fi  
     }%  
   }%  
   % The aux file uses ' as the escape (for now).  
   % Turn off \ as an escape so we do not lose on  
   % entries which were dumped with control sequences in their names.  
   % For example, 'xrdef {$\leq $-fun}{page ...} made by @defun ^^  
   % Reference to such entries still does not work the way one would wish,  
   % but at least they do not bomb out when the aux file is read in.  
   \catcode`\{=1  
   \catcode`\}=2  
   \catcode`\%=\other    \catcode`\%=\other
   \catcode`\'=0    \catcode`+=\other % avoid \+ for paranoia even though we've turned it off
     %
     % This is to support \ in node names and titles, since the \
     % characters end up in a \csname.  It's easier than
     % leaving it active and making its active definition an actual \
     % character.  What I don't understand is why it works in the *value*
     % of the xrdef.  Seems like it should be a catcode12 \, and that
     % should not typeset properly.  But it works, so I'm moving on for
     % now.  --karl, 15jan04.
   \catcode`\\=\other    \catcode`\\=\other
   %    %
   \openin 1 \jobname.aux    % @ is our escape character in .aux files, and we need braces.
   \ifeof 1 \else    \catcode`\{=1
     \closein 1    \catcode`\}=2
     \input \jobname.aux    \catcode`\@=0
     \global\havexrefstrue  }
     \global\warnedobstrue  
   \fi  \def\readdatafile#1{%
   % Open the new aux file.  TeX will close it automatically at exit.  \begingroup
   \openout\auxfile=\jobname.aux    \setupdatafile
     \input\jobname.#1
 \endgroup}  \endgroup}
   
   
 % Footnotes.  \message{insertions,}
   % including footnotes.
   
 \newcount \footnoteno  \newcount \footnoteno
   
Line 5466  width0pt\relax} \fi
Line 9310  width0pt\relax} \fi
 % space to prevent strange expansion errors.)  % space to prevent strange expansion errors.)
 \def\supereject{\par\penalty -20000\footnoteno =0 }  \def\supereject{\par\penalty -20000\footnoteno =0 }
   
 % @footnotestyle is meaningful for info output only.  % @footnotestyle is meaningful for Info output only.
 \let\footnotestyle=\comment  \let\footnotestyle=\comment
   
 \let\ptexfootnote=\footnote  
   
 {\catcode `\@=11  {\catcode `\@=11
 %  %
 % Auto-number footnotes.  Otherwise like plain.  % Auto-number footnotes.  Otherwise like plain.
Line 5481  width0pt\relax} \fi
Line 9323  width0pt\relax} \fi
   % In case the footnote comes at the end of a sentence, preserve the    % In case the footnote comes at the end of a sentence, preserve the
   % extra spacing after we do the footnote number.    % extra spacing after we do the footnote number.
   \let\@sf\empty    \let\@sf\empty
   \ifhmode\edef\@sf{\spacefactor\the\spacefactor}\/\fi    \ifhmode\edef\@sf{\spacefactor\the\spacefactor}\ptexslash\fi
   %    %
   % Remove inadvertent blank space before typesetting the footnote number.    % Remove inadvertent blank space before typesetting the footnote number.
   \unskip    \unskip
   \thisfootno\@sf    \thisfootno\@sf
   \footnotezzz    \dofootnote
 }%  }%
   
 % Don't bother with the trickery in plain.tex to not require the  % Don't bother with the trickery in plain.tex to not require the
 % footnote text as a parameter.  Our footnotes don't need to be so general.  % footnote text as a parameter.  Our footnotes don't need to be so general.
 %  %
 % Oh yes, they do; otherwise, @ifset and anything else that uses  % Oh yes, they do; otherwise, @ifset (and anything else that uses
 % \parseargline fail inside footnotes because the tokens are fixed when  % \parseargline) fails inside footnotes because the tokens are fixed when
 % the footnote is read.  --karl, 16nov96.  % the footnote is read.  --karl, 16nov96.
 %  %
 \long\gdef\footnotezzz{\insert\footins\bgroup  \gdef\dofootnote{%
     \insert\footins\bgroup
     %
     % Nested footnotes are not supported in TeX, that would take a lot
     % more work.  (\startsavinginserts does not suffice.)
     \let\footnote=\errfootnotenest
     %
   % We want to typeset this text as a normal paragraph, even if the    % We want to typeset this text as a normal paragraph, even if the
   % footnote reference occurs in (for example) a display environment.    % footnote reference occurs in (for example) a display environment.
   % So reset some parameters.    % So reset some parameters.
     \hsize=\txipagewidth
   \interlinepenalty\interfootnotelinepenalty    \interlinepenalty\interfootnotelinepenalty
   \splittopskip\ht\strutbox % top baseline for broken footnotes    \splittopskip\ht\strutbox % top baseline for broken footnotes
   \splitmaxdepth\dp\strutbox    \splitmaxdepth\dp\strutbox
Line 5512  width0pt\relax} \fi
Line 9361  width0pt\relax} \fi
   %    %
   \smallfonts \rm    \smallfonts \rm
   %    %
   % Hang the footnote text off the number.    % Because we use hanging indentation in footnotes, a @noindent appears
   \hang    % to exdent this text, so make it be a no-op.  makeinfo does not use
     % hanging indentation so @noindent can still be needed within footnote
     % text after an @example or the like (not that this is good style).
     \let\noindent = \relax
     %
     % Hang the footnote text off the number.  Use \everypar in case the
     % footnote extends for more than one paragraph.
     \everypar = {\hang}%
   \textindent{\thisfootno}%    \textindent{\thisfootno}%
   %    %
   % Don't crash into the line above the footnote text.  Since this    % Don't crash into the line above the footnote text.  Since this
   % expands into a box, it must come within the paragraph, lest it    % expands into a box, it must come within the paragraph, lest it
   % provide a place where TeX can split the footnote.    % provide a place where TeX can split the footnote.
   \footstrut    \footstrut
     %
     % Invoke rest of plain TeX footnote routine.
   \futurelet\next\fo@t    \futurelet\next\fo@t
 }  }
 \def\fo@t{\ifcat\bgroup\noexpand\next \let\next\f@@t  
   \else\let\next\f@t\fi \next}  
 \def\f@@t{\bgroup\aftergroup\@foot\let\next}  
 \def\f@t#1{#1\@foot}  
 \def\@foot{\strut\par\egroup}  
   
 }%end \catcode `\@=11  }%end \catcode `\@=11
   
 % Set the baselineskip to #1, and the lineskip and strut size  \def\errfootnotenest{%
 % correspondingly.  There is no deep meaning behind these magic numbers    \errhelp=\EMsimple
 % used as factors; they just match (closely enough) what Knuth defined.    \errmessage{Nested footnotes not supported in texinfo.tex,
       even though they work in makeinfo; sorry}
   }
   
   \def\errfootnoteheading{%
     \errhelp=\EMsimple
     \errmessage{Footnotes in chapters, sections, etc., are not supported}
   }
   
   % In case a @footnote appears in a vbox, save the footnote text and create
   % the real \insert just after the vbox finished.  Otherwise, the insertion
   % would be lost.
   % Similarly, if a @footnote appears inside an alignment, save the footnote
   % text to a box and make the \insert when a row of the table is finished.
   % And the same can be done for other insert classes.  --kasal, 16nov03.
 %  %
 \def\lineskipfactor{.08333}  % Replace the \insert primitive by a cheating macro.
 \def\strutheightpercent{.70833}  % Deeper inside, just make sure that the saved insertions are not spilled
 \def\strutdepthpercent {.29167}  % out prematurely.
 %  %
 \def\setleading#1{%  \def\startsavinginserts{%
   \normalbaselineskip = #1\relax    \ifx \insert\ptexinsert
   \normallineskip = \lineskipfactor\normalbaselineskip      \let\insert\saveinsert
   \normalbaselines    \else
   \setbox\strutbox =\hbox{%      \let\checkinserts\relax
     \vrule width0pt height\strutheightpercent\baselineskip    \fi
                     depth \strutdepthpercent \baselineskip  
   }%  
 }  }
   
 % @| inserts a changebar to the left of the current line.  It should  % This \insert replacement works for both \insert\footins{foo} and
 % surround any changed text.  This approach does *not* work if the  % \insert\footins\bgroup foo\egroup, but it doesn't work for \insert27{foo}.
 % change spans more than two lines of output.  To handle that, we would  
 % have adopt a much more difficult approach (putting marks into the main  
 % vertical list for the beginning and end of each change).  
 %  %
 \def\|{%  \def\saveinsert#1{%
   % \vadjust can only be used in horizontal mode.    \edef\next{\noexpand\savetobox \makeSAVEname#1}%
   \leavevmode    \afterassignment\next
   %    % swallow the left brace
   % Append this vertical mode material after the current line in the output.    \let\temp =
   \vadjust{%  
     % We want to insert a rule with the height and depth of the current  
     % leading; that is exactly what \strutbox is supposed to record.  
     \vskip-\baselineskip  
     %  
     % \vadjust-items are inserted at the left edge of the type.  So  
     % the \llap here moves out into the left-hand margin.  
     \llap{%  
       %  
       % For a thicker or thinner bar, change the `1pt'.  
       \vrule height\baselineskip width1pt  
       %  
       % This is the space between the bar and the text.  
       \hskip 12pt  
     }%  
   }%  
 }  }
   \def\makeSAVEname#1{\makecsname{SAVE\expandafter\gobble\string#1}}
   \def\savetobox#1{\global\setbox#1 = \vbox\bgroup \unvbox#1}
   
 % For a final copy, take out the rectangles  \def\checksaveins#1{\ifvoid#1\else \placesaveins#1\fi}
 % that mark overfull boxes (in case you have decided  
 % that the text looks ok even though it passes the margin).  
 %  
 \def\finalout{\overfullrule=0pt}  
   
   \def\placesaveins#1{%
     \ptexinsert \csname\expandafter\gobblesave\string#1\endcsname
       {\box#1}%
   }
   
   % eat @SAVE -- beware, all of them have catcode \other:
   {
     \def\dospecials{\do S\do A\do V\do E} \uncatcodespecials  %  ;-)
     \gdef\gobblesave @SAVE{}
   }
   
   % initialization:
   \def\newsaveins #1{%
     \edef\next{\noexpand\newsaveinsX \makeSAVEname#1}%
     \next
   }
   \def\newsaveinsX #1{%
     \csname newbox\endcsname #1%
     \expandafter\def\expandafter\checkinserts\expandafter{\checkinserts
       \checksaveins #1}%
   }
   
   % initialize:
   \let\checkinserts\empty
   \newsaveins\footins
   \newsaveins\margin
   
   
 % @image.  We use the macros from epsf.tex to support this.  % @image.  We use the macros from epsf.tex to support this.
 % If epsf.tex is not installed and @image is used, we complain.  % If epsf.tex is not installed and @image is used, we complain.
 %  %
Line 5591  width0pt\relax} \fi
Line 9462  width0pt\relax} \fi
 % undone and the next image would fail.  % undone and the next image would fail.
 \openin 1 = epsf.tex  \openin 1 = epsf.tex
 \ifeof 1 \else  \ifeof 1 \else
   \closein 1    % Do not bother showing banner with epsf.tex v2.7k (available in
   % Do not bother showing banner with post-v2.7 epsf.tex (available in    % doc/epsf.tex and on ctan).
   % doc/epsf.tex until it shows up on ctan).  
   \def\epsfannounce{\toks0 = }%    \def\epsfannounce{\toks0 = }%
   \input epsf.tex    \input epsf.tex
 \fi  \fi
   \closein 1
 %  %
 % We will only complain once about lack of epsf.tex.  % We will only complain once about lack of epsf.tex.
 \newif\ifwarnednoepsf  \newif\ifwarnednoepsf
Line 5605  width0pt\relax} \fi
Line 9476  width0pt\relax} \fi
   it from ftp://tug.org/tex/epsf.tex.}    it from ftp://tug.org/tex/epsf.tex.}
 %  %
 \def\image#1{%  \def\image#1{%
   \ifx\epsfbox\undefined    \ifx\epsfbox\thisisundefined
     \ifwarnednoepsf \else      \ifwarnednoepsf \else
       \errhelp = \noepsfhelp        \errhelp = \noepsfhelp
       \errmessage{epsf.tex not found, images will be ignored}%        \errmessage{epsf.tex not found, images will be ignored}%
       \global\warnednoepsftrue        \global\warnednoepsftrue
     \fi      \fi
   \else    \else
     \imagexxx #1,,,\finish      \imagexxx #1,,,,,\finish
   \fi    \fi
 }  }
 %  %
 % Arguments to @image:  % Arguments to @image:
 % #1 is (mandatory) image filename; we tack on .eps extension.  % #1 is (mandatory) image filename; we tack on .eps extension.
 % #2 is (optional) width, #3 is (optional) height.  % #2 is (optional) width, #3 is (optional) height.
 % #4 is just the usual extra ignored arg for parsing this stuff.  % #4 is (ignored optional) html alt text.
 \def\imagexxx#1,#2,#3,#4\finish{%  % #5 is (ignored optional) extension.
   % #6 is just the usual extra ignored arg for parsing stuff.
   \newif\ifimagevmode
   \def\imagexxx#1,#2,#3,#4,#5,#6\finish{\begingroup
     \catcode`\^^M = 5     % in case we're inside an example
     \normalturnoffactive  % allow _ et al. in names
     \def\xprocessmacroarg{\eatspaces}% in case we are being used via a macro
     % If the image is by itself, center it.
     \ifvmode
       \imagevmodetrue
     \else \ifx\centersub\centerV
       % for @center @image, we need a vbox so we can have our vertical space
       \imagevmodetrue
       \vbox\bgroup % vbox has better behavior than vtop herev
     \fi\fi
     %
     \ifimagevmode
       \nobreak\medskip
       % Usually we'll have text after the image which will insert
       % \parskip glue, so insert it here too to equalize the space
       % above and below.
       \nobreak\vskip\parskip
       \nobreak
     \fi
     %
     % Leave vertical mode so that indentation from an enclosing
     %  environment such as @quotation is respected.
     % However, if we're at the top level, we don't want the
     %  normal paragraph indentation.
     % On the other hand, if we are in the case of @center @image, we don't
     %  want to start a paragraph, which will create a hsize-width box and
     %  eradicate the centering.
     \ifx\centersub\centerV\else \noindent \fi
     %
     % Output the image.
   \ifpdf    \ifpdf
     \centerline{\dopdfimage{#1}{#2}{#3}}%      % For pdfTeX and LuaTeX <= 0.80
       \dopdfimage{#1}{#2}{#3}%
   \else    \else
     % \epsfbox itself resets \epsf?size at each figure.      \ifx\XeTeXrevision\thisisundefined
     \setbox0 = \hbox{\ignorespaces #2}\ifdim\wd0 > 0pt \epsfxsize=#2\relax \fi        % For epsf.tex
     \setbox0 = \hbox{\ignorespaces #3}\ifdim\wd0 > 0pt \epsfysize=#3\relax \fi        % \epsfbox itself resets \epsf?size at each figure.
     \begingroup        \setbox0 = \hbox{\ignorespaces #2}%
       \catcode`\^^M = 5 % in case we're inside an example          \ifdim\wd0 > 0pt \epsfxsize=#2\relax \fi
       % If the image is by itself, center it.        \setbox0 = \hbox{\ignorespaces #3}%
       \ifvmode          \ifdim\wd0 > 0pt \epsfysize=#3\relax \fi
         \nobreak\bigskip        \epsfbox{#1.eps}%
         % Usually we'll have text after the image which will insert      \else
         % \parskip glue, so insert it here too to equalize the space        % For XeTeX
         % above and below.        \doxeteximage{#1}{#2}{#3}%
         \nobreak\vskip\parskip      \fi
         \nobreak    \fi
         \centerline{\epsfbox{#1.eps}}%    %
         \bigbreak    \ifimagevmode
       \else      \medskip  % space after a standalone image
         % In the middle of a paragraph, no extra space.    \fi
         \epsfbox{#1.eps}%    \ifx\centersub\centerV \egroup \fi
   \endgroup}
   
   
   % @float FLOATTYPE,LABEL,LOC ... @end float for displayed figures, tables,
   % etc.  We don't actually implement floating yet, we always include the
   % float "here".  But it seemed the best name for the future.
   %
   \envparseargdef\float{\eatcommaspace\eatcommaspace\dofloat#1, , ,\finish}
   
   % There may be a space before second and/or third parameter; delete it.
   \def\eatcommaspace#1, {#1,}
   
   % #1 is the optional FLOATTYPE, the text label for this float, typically
   % "Figure", "Table", "Example", etc.  Can't contain commas.  If omitted,
   % this float will not be numbered and cannot be referred to.
   %
   % #2 is the optional xref label.  Also must be present for the float to
   % be referable.
   %
   % #3 is the optional positioning argument; for now, it is ignored.  It
   % will somehow specify the positions allowed to float to (here, top, bottom).
   %
   % We keep a separate counter for each FLOATTYPE, which we reset at each
   % chapter-level command.
   \let\resetallfloatnos=\empty
   %
   \def\dofloat#1,#2,#3,#4\finish{%
     \let\thiscaption=\empty
     \let\thisshortcaption=\empty
     %
     % don't lose footnotes inside @float.
     %
     % BEWARE: when the floats start float, we have to issue warning whenever an
     % insert appears inside a float which could possibly float. --kasal, 26may04
     %
     \startsavinginserts
     %
     % We can't be used inside a paragraph.
     \par
     %
     \vtop\bgroup
       \def\floattype{#1}%
       \def\floatlabel{#2}%
       \def\floatloc{#3}% we do nothing with this yet.
       %
       \ifx\floattype\empty
         \let\safefloattype=\empty
       \else
         {%
           % the floattype might have accents or other special characters,
           % but we need to use it in a control sequence name.
           \indexnofonts
           \turnoffactive
           \xdef\safefloattype{\floattype}%
         }%
       \fi
       %
       % If label is given but no type, we handle that as the empty type.
       \ifx\floatlabel\empty \else
         % We want each FLOATTYPE to be numbered separately (Figure 1,
         % Table 1, Figure 2, ...).  (And if no label, no number.)
         %
         \expandafter\getfloatno\csname\safefloattype floatno\endcsname
         \global\advance\floatno by 1
         %
         {%
           % This magic value for \lastsection is output by \setref as the
           % XREFLABEL-title value.  \xrefX uses it to distinguish float
           % labels (which have a completely different output format) from
           % node and anchor labels.  And \xrdef uses it to construct the
           % lists of floats.
           %
           \edef\lastsection{\floatmagic=\safefloattype}%
           \setref{\floatlabel}{Yfloat}%
         }%
       \fi
       %
       % start with \parskip glue, I guess.
       \vskip\parskip
       %
       % Don't suppress indentation if a float happens to start a section.
       \restorefirstparagraphindent
   }
   
   % we have these possibilities:
   % @float Foo,lbl & @caption{Cap}: Foo 1.1: Cap
   % @float Foo,lbl & no caption:    Foo 1.1
   % @float Foo & @caption{Cap}:     Foo: Cap
   % @float Foo & no caption:        Foo
   % @float ,lbl & Caption{Cap}:     1.1: Cap
   % @float ,lbl & no caption:       1.1
   % @float & @caption{Cap}:         Cap
   % @float & no caption:
   %
   \def\Efloat{%
       \let\floatident = \empty
       %
       % In all cases, if we have a float type, it comes first.
       \ifx\floattype\empty \else \def\floatident{\floattype}\fi
       %
       % If we have an xref label, the number comes next.
       \ifx\floatlabel\empty \else
         \ifx\floattype\empty \else % if also had float type, need tie first.
           \appendtomacro\floatident{\tie}%
       \fi        \fi
         % the number.
         \appendtomacro\floatident{\chaplevelprefix\the\floatno}%
       \fi
       %
       % Start the printed caption with what we've constructed in
       % \floatident, but keep it separate; we need \floatident again.
       \let\captionline = \floatident
       %
       \ifx\thiscaption\empty \else
         \ifx\floatident\empty \else
           \appendtomacro\captionline{: }% had ident, so need a colon between
         \fi
         %
         % caption text.
         \appendtomacro\captionline{\scanexp\thiscaption}%
       \fi
       %
       % If we have anything to print, print it, with space before.
       % Eventually this needs to become an \insert.
       \ifx\captionline\empty \else
         \vskip.5\parskip
         \captionline
         %
         % Space below caption.
         \vskip\parskip
       \fi
       %
       % If have an xref label, write the list of floats info.  Do this
       % after the caption, to avoid chance of it being a breakpoint.
       \ifx\floatlabel\empty \else
         % Write the text that goes in the lof to the aux file as
         % \floatlabel-lof.  Besides \floatident, we include the short
         % caption if specified, else the full caption if specified, else nothing.
         {%
           \requireauxfile
           \atdummies
           %
           \ifx\thisshortcaption\empty
             \def\gtemp{\thiscaption}%
           \else
             \def\gtemp{\thisshortcaption}%
           \fi
           \immediate\write\auxfile{@xrdef{\floatlabel-lof}{\floatident
             \ifx\gtemp\empty \else : \gtemp \fi}}%
         }%
       \fi
     \egroup  % end of \vtop
     %
     \checkinserts
   }
   
   % Append the tokens #2 to the definition of macro #1, not expanding either.
   %
   \def\appendtomacro#1#2{%
     \expandafter\def\expandafter#1\expandafter{#1#2}%
   }
   
   % @caption, @shortcaption
   %
   \def\caption{\docaption\thiscaption}
   \def\shortcaption{\docaption\thisshortcaption}
   \def\docaption{\checkenv\float \bgroup\scanargctxt\defcaption}
   \def\defcaption#1#2{\egroup \def#1{#2}}
   
   % The parameter is the control sequence identifying the counter we are
   % going to use.  Create it if it doesn't exist and assign it to \floatno.
   \def\getfloatno#1{%
     \ifx#1\relax
         % Haven't seen this figure type before.
         \csname newcount\endcsname #1%
         %
         % Remember to reset this floatno at the next chap.
         \expandafter\gdef\expandafter\resetallfloatnos
           \expandafter{\resetallfloatnos #1=0 }%
     \fi
     \let\floatno#1%
   }
   
   % \setref calls this to get the XREFLABEL-snt value.  We want an @xref
   % to the FLOATLABEL to expand to "Figure 3.1".  We call \setref when we
   % first read the @float command.
   %
   \def\Yfloat{\floattype@tie \chaplevelprefix\the\floatno}%
   
   % Magic string used for the XREFLABEL-title value, so \xrefX can
   % distinguish floats from other xref types.
   \def\floatmagic{!!float!!}
   
   % #1 is the control sequence we are passed; we expand into a conditional
   % which is true if #1 represents a float ref.  That is, the magic
   % \lastsection value which we \setref above.
   %
   \def\iffloat#1{\expandafter\doiffloat#1==\finish}
   %
   % #1 is (maybe) the \floatmagic string.  If so, #2 will be the
   % (safe) float type for this float.  We set \iffloattype to #2.
   %
   \def\doiffloat#1=#2=#3\finish{%
     \def\temp{#1}%
     \def\iffloattype{#2}%
     \ifx\temp\floatmagic
   }
   
   % @listoffloats FLOATTYPE - print a list of floats like a table of contents.
   %
   \parseargdef\listoffloats{%
     \def\floattype{#1}% floattype
     {%
       % the floattype might have accents or other special characters,
       % but we need to use it in a control sequence name.
       \indexnofonts
       \turnoffactive
       \xdef\safefloattype{\floattype}%
     }%
     %
     % \xrdef saves the floats as a \do-list in \floatlistSAFEFLOATTYPE.
     \expandafter\ifx\csname floatlist\safefloattype\endcsname \relax
       \ifhavexrefs
         % if the user said @listoffloats foo but never @float foo.
         \message{\linenumber No `\safefloattype' floats to list.}%
       \fi
     \else
       \begingroup
         \leftskip=\tocindent  % indent these entries like a toc
         \let\do=\listoffloatsdo
         \csname floatlist\safefloattype\endcsname
     \endgroup      \endgroup
   \fi    \fi
 }  }
   
   % This is called on each entry in a list of floats.  We're passed the
   % xref label, in the form LABEL-title, which is how we save it in the
   % aux file.  We strip off the -title and look up \XRLABEL-lof, which
   % has the text we're supposed to typeset here.
   %
   % Figures without xref labels will not be included in the list (since
   % they won't appear in the aux file).
   %
   \def\listoffloatsdo#1{\listoffloatsdoentry#1\finish}
   \def\listoffloatsdoentry#1-title\finish{{%
     % Can't fully expand XR#1-lof because it can contain anything.  Just
     % pass the control sequence.  On the other hand, XR#1-pg is just the
     % page number, and we want to fully expand that so we can get a link
     % in pdf output.
     \toksA = \expandafter{\csname XR#1-lof\endcsname}%
     %
     % use the same \entry macro we use to generate the TOC and index.
     \edef\writeentry{\noexpand\entry{\the\toksA}{\csname XR#1-pg\endcsname}}%
     \writeentry
   }}
   
   
 \message{localization,}  \message{localization,}
 % and i18n.  
   
 % @documentlanguage is usually given very early, just after  % For single-language documents, @documentlanguage is usually given very
 % @setfilename.  If done too late, it may not override everything  % early, just after @documentencoding.  Single argument is the language
 % properly.  Single argument is the language abbreviation.  % (de) or locale (de_DE) abbreviation.
 % It would be nice if we could set up a hyphenation file here.  
 %  %
 \def\documentlanguage{\parsearg\dodocumentlanguage}  {
 \def\dodocumentlanguage#1{%    \catcode`\_ = \active
     \globaldefs=1
   \parseargdef\documentlanguage{%
   \tex % read txi-??.tex file in plain TeX.    \tex % read txi-??.tex file in plain TeX.
   % Read the file if it exists.      % Read the file by the name they passed if it exists.
       \let_ = \normalunderscore  % normal _ character for filename test
       \openin 1 txi-#1.tex
       \ifeof 1
         \documentlanguagetrywithoutunderscore #1_\finish
       \else
         \globaldefs = 1  % everything in the txi-LL files needs to persist
         \input txi-#1.tex
       \fi
       \closein 1
     \endgroup % end raw TeX
   }
   %
   % If they passed de_DE, and txi-de_DE.tex doesn't exist,
   % try txi-de.tex.
   %
   \gdef\documentlanguagetrywithoutunderscore#1_#2\finish{%
   \openin 1 txi-#1.tex    \openin 1 txi-#1.tex
   \ifeof1    \ifeof 1
     \errhelp = \nolanghelp      \errhelp = \nolanghelp
     \errmessage{Cannot read language file txi-#1.tex}%      \errmessage{Cannot read language file txi-#1.tex}%
     \let\temp = \relax  
   \else    \else
     \def\temp{\input txi-#1.tex }%      \globaldefs = 1  % everything in the txi-LL files needs to persist
       \input txi-#1.tex
   \fi    \fi
   \temp    \closein 1
   \endgroup  
 }  }
   }% end of special _ catcode
   %
 \newhelp\nolanghelp{The given language definition file cannot be found or  \newhelp\nolanghelp{The given language definition file cannot be found or
 is empty.  Maybe you need to install it?  In the current directory  is empty.  Maybe you need to install it?  Putting it in the current
 should work if nowhere else does.}  directory should work if nowhere else does.}
   
   % This macro is called from txi-??.tex files; the first argument is the
   % \language name to set (without the "\lang@" prefix), the second and
   % third args are \{left,right}hyphenmin.
   %
   % The language names to pass are determined when the format is built.
   % See the etex.log file created at that time, e.g.,
   % /usr/local/texlive/2008/texmf-var/web2c/pdftex/etex.log.
   %
   % With TeX Live 2008, etex now includes hyphenation patterns for all
   % available languages.  This means we can support hyphenation in
   % Texinfo, at least to some extent.  (This still doesn't solve the
   % accented characters problem.)
   %
   \catcode`@=11
   \def\txisetlanguage#1#2#3{%
     % do not set the language if the name is undefined in the current TeX.
     \expandafter\ifx\csname lang@#1\endcsname \relax
       \message{no patterns for #1}%
     \else
       \global\language = \csname lang@#1\endcsname
     \fi
     % but there is no harm in adjusting the hyphenmin values regardless.
     \global\lefthyphenmin = #2\relax
     \global\righthyphenmin = #3\relax
   }
   
 % @documentencoding should change something in TeX eventually, most  % XeTeX and LuaTeX can handle Unicode natively.
 % likely, but for now just recognize it.  % Their default I/O uses UTF-8 sequences instead of a byte-wise operation.
 \let\documentencoding = \comment  % Other TeX engines' I/O (pdfTeX, etc.) is byte-wise.
   %
   \newif\iftxinativeunicodecapable
   \newif\iftxiusebytewiseio
   
   \ifx\XeTeXrevision\thisisundefined
     \ifx\luatexversion\thisisundefined
       \txinativeunicodecapablefalse
       \txiusebytewiseiotrue
     \else
       \txinativeunicodecapabletrue
       \txiusebytewiseiofalse
     \fi
   \else
     \txinativeunicodecapabletrue
     \txiusebytewiseiofalse
   \fi
   
 % Page size parameters.  % Set I/O by bytes instead of UTF-8 sequence for XeTeX and LuaTex
   % for non-UTF-8 (byte-wise) encodings.
 %  %
   \def\setbytewiseio{%
     \ifx\XeTeXrevision\thisisundefined
     \else
       \XeTeXdefaultencoding "bytes"  % For subsequent files to be read
       \XeTeXinputencoding "bytes"  % For document root file
       % Unfortunately, there seems to be no corresponding XeTeX command for
       % output encoding.  This is a problem for auxiliary index and TOC files.
       % The only solution would be perhaps to write out @U{...} sequences in
       % place of non-ASCII characters.
     \fi
   
     \ifx\luatexversion\thisisundefined
     \else
       \directlua{
       local utf8_char, byte, gsub = unicode.utf8.char, string.byte, string.gsub
       local function convert_char (char)
         return utf8_char(byte(char))
       end
   
       local function convert_line (line)
         return gsub(line, ".", convert_char)
       end
   
       callback.register("process_input_buffer", convert_line)
   
       local function convert_line_out (line)
         local line_out = ""
         for c in string.utfvalues(line) do
            line_out = line_out .. string.char(c)
         end
         return line_out
       end
   
       callback.register("process_output_buffer", convert_line_out)
       }
     \fi
   
     \txiusebytewiseiotrue
   }
   
   
   % Helpers for encodings.
   % Set the catcode of characters 128 through 255 to the specified number.
   %
   \def\setnonasciicharscatcode#1{%
      \count255=128
      \loop\ifnum\count255<256
         \global\catcode\count255=#1\relax
         \advance\count255 by 1
      \repeat
   }
   
   \def\setnonasciicharscatcodenonglobal#1{%
      \count255=128
      \loop\ifnum\count255<256
         \catcode\count255=#1\relax
         \advance\count255 by 1
      \repeat
   }
   
   % @documentencoding sets the definition of non-ASCII characters
   % according to the specified encoding.
   %
   \def\documentencoding{\parseargusing\filenamecatcodes\documentencodingzzz}
   \def\documentencodingzzz#1{%
     %
     % Encoding being declared for the document.
     \def\declaredencoding{\csname #1.enc\endcsname}%
     %
     % Supported encodings: names converted to tokens in order to be able
     % to compare them with \ifx.
     \def\ascii{\csname US-ASCII.enc\endcsname}%
     \def\latnine{\csname ISO-8859-15.enc\endcsname}%
     \def\latone{\csname ISO-8859-1.enc\endcsname}%
     \def\lattwo{\csname ISO-8859-2.enc\endcsname}%
     \def\utfeight{\csname UTF-8.enc\endcsname}%
     %
     \ifx \declaredencoding \ascii
        \asciichardefs
     %
     \else \ifx \declaredencoding \lattwo
        \iftxinativeunicodecapable
          \setbytewiseio
        \fi
        \setnonasciicharscatcode\active
        \lattwochardefs
     %
     \else \ifx \declaredencoding \latone
        \iftxinativeunicodecapable
          \setbytewiseio
        \fi
        \setnonasciicharscatcode\active
        \latonechardefs
     %
     \else \ifx \declaredencoding \latnine
        \iftxinativeunicodecapable
          \setbytewiseio
        \fi
        \setnonasciicharscatcode\active
        \latninechardefs
     %
     \else \ifx \declaredencoding \utfeight
        \iftxinativeunicodecapable
          % For native Unicode handling (XeTeX and LuaTeX)
          \nativeunicodechardefs
        \else
          % For treating UTF-8 as byte sequences (TeX, eTeX and pdfTeX)
          \setnonasciicharscatcode\active
          % since we already invoked \utfeightchardefs at the top level
          % (below), do not re-invoke it, otherwise our check for duplicated
          % definitions gets triggered.  Making non-ascii chars active is
          % sufficient.
        \fi
     %
     \else
       \message{Ignoring unknown document encoding: #1.}%
     %
     \fi % utfeight
     \fi % latnine
     \fi % latone
     \fi % lattwo
     \fi % ascii
     %
     \ifx\XeTeXrevision\thisisundefined
     \else
       \ifx \declaredencoding \utfeight
       \else
         \ifx \declaredencoding \ascii
         \else
           \message{Warning: XeTeX with non-UTF-8 encodings cannot handle %
           non-ASCII characters in auxiliary files.}%
         \fi
       \fi
     \fi
   }
   
   % emacs-page
   % A message to be logged when using a character that isn't available
   % the default font encoding (OT1).
   %
   \def\missingcharmsg#1{\message{Character missing, sorry: #1.}}
   
   % Take account of \c (plain) vs. \, (Texinfo) difference.
   \def\cedilla#1{\ifx\c\ptexc\c{#1}\else\,{#1}\fi}
   
   % First, make active non-ASCII characters in order for them to be
   % correctly categorized when TeX reads the replacement text of
   % macros containing the character definitions.
   \setnonasciicharscatcode\active
   %
   
   \def\gdefchar#1#2{%
   \gdef#1{%
      \ifpassthroughchars
        \string#1%
      \else
        #2%
      \fi
   }}
   
   % Latin1 (ISO-8859-1) character definitions.
   \def\latonechardefs{%
     \gdefchar^^a0{\tie}
     \gdefchar^^a1{\exclamdown}
     \gdefchar^^a2{{\tcfont \char162}} % cent
     \gdefchar^^a3{\pounds}
     \gdefchar^^a4{{\tcfont \char164}} % currency
     \gdefchar^^a5{{\tcfont \char165}} % yen
     \gdefchar^^a6{{\tcfont \char166}} % broken bar
     \gdefchar^^a7{\S}
     \gdefchar^^a8{\"{}}
     \gdefchar^^a9{\copyright}
     \gdefchar^^aa{\ordf}
     \gdefchar^^ab{\guillemetleft}
     \gdefchar^^ac{\ensuremath\lnot}
     \gdefchar^^ad{\-}
     \gdefchar^^ae{\registeredsymbol}
     \gdefchar^^af{\={}}
     %
     \gdefchar^^b0{\textdegree}
     \gdefchar^^b1{$\pm$}
     \gdefchar^^b2{$^2$}
     \gdefchar^^b3{$^3$}
     \gdefchar^^b4{\'{}}
     \gdefchar^^b5{$\mu$}
     \gdefchar^^b6{\P}
     \gdefchar^^b7{\ensuremath\cdot}
     \gdefchar^^b8{\cedilla\ }
     \gdefchar^^b9{$^1$}
     \gdefchar^^ba{\ordm}
     \gdefchar^^bb{\guillemetright}
     \gdefchar^^bc{$1\over4$}
     \gdefchar^^bd{$1\over2$}
     \gdefchar^^be{$3\over4$}
     \gdefchar^^bf{\questiondown}
     %
     \gdefchar^^c0{\`A}
     \gdefchar^^c1{\'A}
     \gdefchar^^c2{\^A}
     \gdefchar^^c3{\~A}
     \gdefchar^^c4{\"A}
     \gdefchar^^c5{\ringaccent A}
     \gdefchar^^c6{\AE}
     \gdefchar^^c7{\cedilla C}
     \gdefchar^^c8{\`E}
     \gdefchar^^c9{\'E}
     \gdefchar^^ca{\^E}
     \gdefchar^^cb{\"E}
     \gdefchar^^cc{\`I}
     \gdefchar^^cd{\'I}
     \gdefchar^^ce{\^I}
     \gdefchar^^cf{\"I}
     %
     \gdefchar^^d0{\DH}
     \gdefchar^^d1{\~N}
     \gdefchar^^d2{\`O}
     \gdefchar^^d3{\'O}
     \gdefchar^^d4{\^O}
     \gdefchar^^d5{\~O}
     \gdefchar^^d6{\"O}
     \gdefchar^^d7{$\times$}
     \gdefchar^^d8{\O}
     \gdefchar^^d9{\`U}
     \gdefchar^^da{\'U}
     \gdefchar^^db{\^U}
     \gdefchar^^dc{\"U}
     \gdefchar^^dd{\'Y}
     \gdefchar^^de{\TH}
     \gdefchar^^df{\ss}
     %
     \gdefchar^^e0{\`a}
     \gdefchar^^e1{\'a}
     \gdefchar^^e2{\^a}
     \gdefchar^^e3{\~a}
     \gdefchar^^e4{\"a}
     \gdefchar^^e5{\ringaccent a}
     \gdefchar^^e6{\ae}
     \gdefchar^^e7{\cedilla c}
     \gdefchar^^e8{\`e}
     \gdefchar^^e9{\'e}
     \gdefchar^^ea{\^e}
     \gdefchar^^eb{\"e}
     \gdefchar^^ec{\`{\dotless i}}
     \gdefchar^^ed{\'{\dotless i}}
     \gdefchar^^ee{\^{\dotless i}}
     \gdefchar^^ef{\"{\dotless i}}
     %
     \gdefchar^^f0{\dh}
     \gdefchar^^f1{\~n}
     \gdefchar^^f2{\`o}
     \gdefchar^^f3{\'o}
     \gdefchar^^f4{\^o}
     \gdefchar^^f5{\~o}
     \gdefchar^^f6{\"o}
     \gdefchar^^f7{$\div$}
     \gdefchar^^f8{\o}
     \gdefchar^^f9{\`u}
     \gdefchar^^fa{\'u}
     \gdefchar^^fb{\^u}
     \gdefchar^^fc{\"u}
     \gdefchar^^fd{\'y}
     \gdefchar^^fe{\th}
     \gdefchar^^ff{\"y}
   }
   
   % Latin9 (ISO-8859-15) encoding character definitions.
   \def\latninechardefs{%
     % Encoding is almost identical to Latin1.
     \latonechardefs
     %
     \gdefchar^^a4{\euro}
     \gdefchar^^a6{\v S}
     \gdefchar^^a8{\v s}
     \gdefchar^^b4{\v Z}
     \gdefchar^^b8{\v z}
     \gdefchar^^bc{\OE}
     \gdefchar^^bd{\oe}
     \gdefchar^^be{\"Y}
   }
   
   % Latin2 (ISO-8859-2) character definitions.
   \def\lattwochardefs{%
     \gdefchar^^a0{\tie}
     \gdefchar^^a1{\ogonek{A}}
     \gdefchar^^a2{\u{}}
     \gdefchar^^a3{\L}
     \gdefchar^^a4{\missingcharmsg{CURRENCY SIGN}}
     \gdefchar^^a5{\v L}
     \gdefchar^^a6{\'S}
     \gdefchar^^a7{\S}
     \gdefchar^^a8{\"{}}
     \gdefchar^^a9{\v S}
     \gdefchar^^aa{\cedilla S}
     \gdefchar^^ab{\v T}
     \gdefchar^^ac{\'Z}
     \gdefchar^^ad{\-}
     \gdefchar^^ae{\v Z}
     \gdefchar^^af{\dotaccent Z}
     %
     \gdefchar^^b0{\textdegree}
     \gdefchar^^b1{\ogonek{a}}
     \gdefchar^^b2{\ogonek{ }}
     \gdefchar^^b3{\l}
     \gdefchar^^b4{\'{}}
     \gdefchar^^b5{\v l}
     \gdefchar^^b6{\'s}
     \gdefchar^^b7{\v{}}
     \gdefchar^^b8{\cedilla\ }
     \gdefchar^^b9{\v s}
     \gdefchar^^ba{\cedilla s}
     \gdefchar^^bb{\v t}
     \gdefchar^^bc{\'z}
     \gdefchar^^bd{\H{}}
     \gdefchar^^be{\v z}
     \gdefchar^^bf{\dotaccent z}
     %
     \gdefchar^^c0{\'R}
     \gdefchar^^c1{\'A}
     \gdefchar^^c2{\^A}
     \gdefchar^^c3{\u A}
     \gdefchar^^c4{\"A}
     \gdefchar^^c5{\'L}
     \gdefchar^^c6{\'C}
     \gdefchar^^c7{\cedilla C}
     \gdefchar^^c8{\v C}
     \gdefchar^^c9{\'E}
     \gdefchar^^ca{\ogonek{E}}
     \gdefchar^^cb{\"E}
     \gdefchar^^cc{\v E}
     \gdefchar^^cd{\'I}
     \gdefchar^^ce{\^I}
     \gdefchar^^cf{\v D}
     %
     \gdefchar^^d0{\DH}
     \gdefchar^^d1{\'N}
     \gdefchar^^d2{\v N}
     \gdefchar^^d3{\'O}
     \gdefchar^^d4{\^O}
     \gdefchar^^d5{\H O}
     \gdefchar^^d6{\"O}
     \gdefchar^^d7{$\times$}
     \gdefchar^^d8{\v R}
     \gdefchar^^d9{\ringaccent U}
     \gdefchar^^da{\'U}
     \gdefchar^^db{\H U}
     \gdefchar^^dc{\"U}
     \gdefchar^^dd{\'Y}
     \gdefchar^^de{\cedilla T}
     \gdefchar^^df{\ss}
     %
     \gdefchar^^e0{\'r}
     \gdefchar^^e1{\'a}
     \gdefchar^^e2{\^a}
     \gdefchar^^e3{\u a}
     \gdefchar^^e4{\"a}
     \gdefchar^^e5{\'l}
     \gdefchar^^e6{\'c}
     \gdefchar^^e7{\cedilla c}
     \gdefchar^^e8{\v c}
     \gdefchar^^e9{\'e}
     \gdefchar^^ea{\ogonek{e}}
     \gdefchar^^eb{\"e}
     \gdefchar^^ec{\v e}
     \gdefchar^^ed{\'{\dotless{i}}}
     \gdefchar^^ee{\^{\dotless{i}}}
     \gdefchar^^ef{\v d}
     %
     \gdefchar^^f0{\dh}
     \gdefchar^^f1{\'n}
     \gdefchar^^f2{\v n}
     \gdefchar^^f3{\'o}
     \gdefchar^^f4{\^o}
     \gdefchar^^f5{\H o}
     \gdefchar^^f6{\"o}
     \gdefchar^^f7{$\div$}
     \gdefchar^^f8{\v r}
     \gdefchar^^f9{\ringaccent u}
     \gdefchar^^fa{\'u}
     \gdefchar^^fb{\H u}
     \gdefchar^^fc{\"u}
     \gdefchar^^fd{\'y}
     \gdefchar^^fe{\cedilla t}
     \gdefchar^^ff{\dotaccent{}}
   }
   
   % UTF-8 character definitions.
   %
   % This code to support UTF-8 is based on LaTeX's utf8.def, with some
   % changes for Texinfo conventions.  It is included here under the GPL by
   % permission from Frank Mittelbach and the LaTeX team.
   %
   \newcount\countUTFx
   \newcount\countUTFy
   \newcount\countUTFz
   
   \gdef\UTFviiiTwoOctets#1#2{\expandafter
      \UTFviiiDefined\csname u8:#1\string #2\endcsname}
   %
   \gdef\UTFviiiThreeOctets#1#2#3{\expandafter
      \UTFviiiDefined\csname u8:#1\string #2\string #3\endcsname}
   %
   \gdef\UTFviiiFourOctets#1#2#3#4{\expandafter
      \UTFviiiDefined\csname u8:#1\string #2\string #3\string #4\endcsname}
   
   \gdef\UTFviiiDefined#1{%
     \ifx #1\relax
       \message{\linenumber Unicode char \string #1 not defined for Texinfo}%
     \else
       \expandafter #1%
     \fi
   }
   
   % Give non-ASCII bytes the active definitions for processing UTF-8 sequences
   \begingroup
     \catcode`\~13
     \catcode`\$12
     \catcode`\"12
   
     % Loop from \countUTFx to \countUTFy, performing \UTFviiiTmp
     % substituting ~ and $ with a character token of that value.
     \def\UTFviiiLoop{%
       \global\catcode\countUTFx\active
       \uccode`\~\countUTFx
       \uccode`\$\countUTFx
       \uppercase\expandafter{\UTFviiiTmp}%
       \advance\countUTFx by 1
       \ifnum\countUTFx < \countUTFy
         \expandafter\UTFviiiLoop
       \fi}
   
     % For bytes other than the first in a UTF-8 sequence.  Not expected to
     % be expanded except when writing to auxiliary files.
     \countUTFx = "80
     \countUTFy = "C2
     \def\UTFviiiTmp{%
       \gdef~{%
           \ifpassthroughchars $\fi}}%
     \UTFviiiLoop
   
     \countUTFx = "C2
     \countUTFy = "E0
     \def\UTFviiiTmp{%
       \gdef~{%
           \ifpassthroughchars $%
           \else\expandafter\UTFviiiTwoOctets\expandafter$\fi}}%
     \UTFviiiLoop
   
     \countUTFx = "E0
     \countUTFy = "F0
     \def\UTFviiiTmp{%
       \gdef~{%
           \ifpassthroughchars $%
           \else\expandafter\UTFviiiThreeOctets\expandafter$\fi}}%
     \UTFviiiLoop
   
     \countUTFx = "F0
     \countUTFy = "F4
     \def\UTFviiiTmp{%
       \gdef~{%
           \ifpassthroughchars $%
           \else\expandafter\UTFviiiFourOctets\expandafter$\fi
           }}%
     \UTFviiiLoop
   \endgroup
   
   \def\globallet{\global\let} % save some \expandafter's below
   
   % @U{xxxx} to produce U+xxxx, if we support it.
   \def\U#1{%
     \expandafter\ifx\csname uni:#1\endcsname \relax
       \iftxinativeunicodecapable
         % All Unicode characters can be used if native Unicode handling is
         % active.  However, if the font does not have the glyph,
         % letters are missing.
         \begingroup
           \uccode`\.="#1\relax
           \uppercase{.}
         \endgroup
       \else
         \errhelp = \EMsimple
         \errmessage{Unicode character U+#1 not supported, sorry}%
       \fi
     \else
       \csname uni:#1\endcsname
     \fi
   }
   
   % These macros are used here to construct the name of a control
   % sequence to be defined.
   \def\UTFviiiTwoOctetsName#1#2{%
     \csname u8:#1\string #2\endcsname}%
   \def\UTFviiiThreeOctetsName#1#2#3{%
     \csname u8:#1\string #2\string #3\endcsname}%
   \def\UTFviiiFourOctetsName#1#2#3#4{%
     \csname u8:#1\string #2\string #3\string #4\endcsname}%
   
   % For UTF-8 byte sequences (TeX, e-TeX and pdfTeX),
   % provide a definition macro to replace a Unicode character;
   % this gets used by the @U command
   %
   \begingroup
     \catcode`\"=12
     \catcode`\<=12
     \catcode`\.=12
     \catcode`\,=12
     \catcode`\;=12
     \catcode`\!=12
     \catcode`\~=13
     \gdef\DeclareUnicodeCharacterUTFviii#1#2{%
       \countUTFz = "#1\relax
       \begingroup
         \parseXMLCharref
   
         % Give \u8:... its definition.  The sequence of seven \expandafter's
         % expands after the \gdef three times, e.g.
         %
         % 1.  \UTFviiTwoOctetsName B1 B2
         % 2.  \csname u8:B1 \string B2 \endcsname
         % 3.  \u8: B1 B2  (a single control sequence token)
         %
         \expandafter\expandafter
         \expandafter\expandafter
         \expandafter\expandafter
         \expandafter\gdef       \UTFviiiTmp{#2}%
         %
         \expandafter\ifx\csname uni:#1\endcsname \relax \else
          \message{Internal error, already defined: #1}%
         \fi
         %
         % define an additional control sequence for this code point.
         \expandafter\globallet\csname uni:#1\endcsname \UTFviiiTmp
       \endgroup}
     %
     % Given the value in \countUTFz as a Unicode code point, set \UTFviiiTmp
     % to the corresponding UTF-8 sequence.
     \gdef\parseXMLCharref{%
       \ifnum\countUTFz < "A0\relax
         \errhelp = \EMsimple
         \errmessage{Cannot define Unicode char value < 00A0}%
       \else\ifnum\countUTFz < "800\relax
         \parseUTFviiiA,%
         \parseUTFviiiB C\UTFviiiTwoOctetsName.,%
       \else\ifnum\countUTFz < "10000\relax
         \parseUTFviiiA;%
         \parseUTFviiiA,%
         \parseUTFviiiB E\UTFviiiThreeOctetsName.{,;}%
       \else
         \parseUTFviiiA;%
         \parseUTFviiiA,%
         \parseUTFviiiA!%
         \parseUTFviiiB F\UTFviiiFourOctetsName.{!,;}%
       \fi\fi\fi
     }
   
     % Extract a byte from the end of the UTF-8 representation of \countUTFx.
     % It must be a non-initial byte in the sequence.
     % Change \uccode of #1 for it to be used in \parseUTFviiiB as one
     % of the bytes.
     \gdef\parseUTFviiiA#1{%
       \countUTFx = \countUTFz
       \divide\countUTFz by 64
       \countUTFy = \countUTFz  % Save to be the future value of \countUTFz.
       \multiply\countUTFz by 64
   
       % \countUTFz is now \countUTFx with the last 5 bits cleared.  Subtract
       % in order to get the last five bits.
       \advance\countUTFx by -\countUTFz
   
       % Convert this to the byte in the UTF-8 sequence.
       \advance\countUTFx by 128
       \uccode `#1\countUTFx
       \countUTFz = \countUTFy}
   
     % Used to put a UTF-8 byte sequence into \UTFviiiTmp
     % #1 is the increment for \countUTFz to yield a the first byte of the UTF-8
     %    sequence.
     % #2 is one of the \UTFviii*OctetsName macros.
     % #3 is always a full stop (.)
     % #4 is a template for the other bytes in the sequence.  The values for these
     %    bytes is substituted in here with \uppercase using the \uccode's.
     \gdef\parseUTFviiiB#1#2#3#4{%
       \advance\countUTFz by "#10\relax
       \uccode `#3\countUTFz
       \uppercase{\gdef\UTFviiiTmp{#2#3#4}}}
   \endgroup
   
   % For native Unicode handling (XeTeX and LuaTeX),
   % provide a definition macro that sets a catcode to `other' non-globally
   %
   \def\DeclareUnicodeCharacterNativeOther#1#2{%
     \catcode"#1=\other
   }
   
   % https://en.wikipedia.org/wiki/Plane_(Unicode)#Basic_M
   % U+0000..U+007F = https://en.wikipedia.org/wiki/Basic_Latin_(Unicode_block)
   % U+0080..U+00FF = https://en.wikipedia.org/wiki/Latin-1_Supplement_(Unicode_block)
   % U+0100..U+017F = https://en.wikipedia.org/wiki/Latin_Extended-A
   % U+0180..U+024F = https://en.wikipedia.org/wiki/Latin_Extended-B
   %
   % Many of our renditions are less than wonderful, and all the missing
   % characters are available somewhere.  Loading the necessary fonts
   % awaits user request.  We can't truly support Unicode without
   % reimplementing everything that's been done in LaTeX for many years,
   % plus probably using luatex or xetex, and who knows what else.
   % We won't be doing that here in this simple file.  But we can try to at
   % least make most of the characters not bomb out.
   %
   \def\unicodechardefs{%
     \DeclareUnicodeCharacter{00A0}{\tie}%
     \DeclareUnicodeCharacter{00A1}{\exclamdown}%
     \DeclareUnicodeCharacter{00A2}{{\tcfont \char162}}% 0242=cent
     \DeclareUnicodeCharacter{00A3}{\pounds}%
     \DeclareUnicodeCharacter{00A4}{{\tcfont \char164}}% 0244=currency
     \DeclareUnicodeCharacter{00A5}{{\tcfont \char165}}% 0245=yen
     \DeclareUnicodeCharacter{00A6}{{\tcfont \char166}}% 0246=brokenbar
     \DeclareUnicodeCharacter{00A7}{\S}%
     \DeclareUnicodeCharacter{00A8}{\"{ }}%
     \DeclareUnicodeCharacter{00A9}{\copyright}%
     \DeclareUnicodeCharacter{00AA}{\ordf}%
     \DeclareUnicodeCharacter{00AB}{\guillemetleft}%
     \DeclareUnicodeCharacter{00AC}{\ensuremath\lnot}%
     \DeclareUnicodeCharacter{00AD}{\-}%
     \DeclareUnicodeCharacter{00AE}{\registeredsymbol}%
     \DeclareUnicodeCharacter{00AF}{\={ }}%
     %
     \DeclareUnicodeCharacter{00B0}{\ringaccent{ }}%
     \DeclareUnicodeCharacter{00B1}{\ensuremath\pm}%
     \DeclareUnicodeCharacter{00B2}{$^2$}%
     \DeclareUnicodeCharacter{00B3}{$^3$}%
     \DeclareUnicodeCharacter{00B4}{\'{ }}%
     \DeclareUnicodeCharacter{00B5}{$\mu$}%
     \DeclareUnicodeCharacter{00B6}{\P}%
     \DeclareUnicodeCharacter{00B7}{\ensuremath\cdot}%
     \DeclareUnicodeCharacter{00B8}{\cedilla{ }}%
     \DeclareUnicodeCharacter{00B9}{$^1$}%
     \DeclareUnicodeCharacter{00BA}{\ordm}%
     \DeclareUnicodeCharacter{00BB}{\guillemetright}%
     \DeclareUnicodeCharacter{00BC}{$1\over4$}%
     \DeclareUnicodeCharacter{00BD}{$1\over2$}%
     \DeclareUnicodeCharacter{00BE}{$3\over4$}%
     \DeclareUnicodeCharacter{00BF}{\questiondown}%
     %
     \DeclareUnicodeCharacter{00C0}{\`A}%
     \DeclareUnicodeCharacter{00C1}{\'A}%
     \DeclareUnicodeCharacter{00C2}{\^A}%
     \DeclareUnicodeCharacter{00C3}{\~A}%
     \DeclareUnicodeCharacter{00C4}{\"A}%
     \DeclareUnicodeCharacter{00C5}{\AA}%
     \DeclareUnicodeCharacter{00C6}{\AE}%
     \DeclareUnicodeCharacter{00C7}{\cedilla{C}}%
     \DeclareUnicodeCharacter{00C8}{\`E}%
     \DeclareUnicodeCharacter{00C9}{\'E}%
     \DeclareUnicodeCharacter{00CA}{\^E}%
     \DeclareUnicodeCharacter{00CB}{\"E}%
     \DeclareUnicodeCharacter{00CC}{\`I}%
     \DeclareUnicodeCharacter{00CD}{\'I}%
     \DeclareUnicodeCharacter{00CE}{\^I}%
     \DeclareUnicodeCharacter{00CF}{\"I}%
     %
     \DeclareUnicodeCharacter{00D0}{\DH}%
     \DeclareUnicodeCharacter{00D1}{\~N}%
     \DeclareUnicodeCharacter{00D2}{\`O}%
     \DeclareUnicodeCharacter{00D3}{\'O}%
     \DeclareUnicodeCharacter{00D4}{\^O}%
     \DeclareUnicodeCharacter{00D5}{\~O}%
     \DeclareUnicodeCharacter{00D6}{\"O}%
     \DeclareUnicodeCharacter{00D7}{\ensuremath\times}%
     \DeclareUnicodeCharacter{00D8}{\O}%
     \DeclareUnicodeCharacter{00D9}{\`U}%
     \DeclareUnicodeCharacter{00DA}{\'U}%
     \DeclareUnicodeCharacter{00DB}{\^U}%
     \DeclareUnicodeCharacter{00DC}{\"U}%
     \DeclareUnicodeCharacter{00DD}{\'Y}%
     \DeclareUnicodeCharacter{00DE}{\TH}%
     \DeclareUnicodeCharacter{00DF}{\ss}%
     %
     \DeclareUnicodeCharacter{00E0}{\`a}%
     \DeclareUnicodeCharacter{00E1}{\'a}%
     \DeclareUnicodeCharacter{00E2}{\^a}%
     \DeclareUnicodeCharacter{00E3}{\~a}%
     \DeclareUnicodeCharacter{00E4}{\"a}%
     \DeclareUnicodeCharacter{00E5}{\aa}%
     \DeclareUnicodeCharacter{00E6}{\ae}%
     \DeclareUnicodeCharacter{00E7}{\cedilla{c}}%
     \DeclareUnicodeCharacter{00E8}{\`e}%
     \DeclareUnicodeCharacter{00E9}{\'e}%
     \DeclareUnicodeCharacter{00EA}{\^e}%
     \DeclareUnicodeCharacter{00EB}{\"e}%
     \DeclareUnicodeCharacter{00EC}{\`{\dotless{i}}}%
     \DeclareUnicodeCharacter{00ED}{\'{\dotless{i}}}%
     \DeclareUnicodeCharacter{00EE}{\^{\dotless{i}}}%
     \DeclareUnicodeCharacter{00EF}{\"{\dotless{i}}}%
     %
     \DeclareUnicodeCharacter{00F0}{\dh}%
     \DeclareUnicodeCharacter{00F1}{\~n}%
     \DeclareUnicodeCharacter{00F2}{\`o}%
     \DeclareUnicodeCharacter{00F3}{\'o}%
     \DeclareUnicodeCharacter{00F4}{\^o}%
     \DeclareUnicodeCharacter{00F5}{\~o}%
     \DeclareUnicodeCharacter{00F6}{\"o}%
     \DeclareUnicodeCharacter{00F7}{\ensuremath\div}%
     \DeclareUnicodeCharacter{00F8}{\o}%
     \DeclareUnicodeCharacter{00F9}{\`u}%
     \DeclareUnicodeCharacter{00FA}{\'u}%
     \DeclareUnicodeCharacter{00FB}{\^u}%
     \DeclareUnicodeCharacter{00FC}{\"u}%
     \DeclareUnicodeCharacter{00FD}{\'y}%
     \DeclareUnicodeCharacter{00FE}{\th}%
     \DeclareUnicodeCharacter{00FF}{\"y}%
     %
     \DeclareUnicodeCharacter{0100}{\=A}%
     \DeclareUnicodeCharacter{0101}{\=a}%
     \DeclareUnicodeCharacter{0102}{\u{A}}%
     \DeclareUnicodeCharacter{0103}{\u{a}}%
     \DeclareUnicodeCharacter{0104}{\ogonek{A}}%
     \DeclareUnicodeCharacter{0105}{\ogonek{a}}%
     \DeclareUnicodeCharacter{0106}{\'C}%
     \DeclareUnicodeCharacter{0107}{\'c}%
     \DeclareUnicodeCharacter{0108}{\^C}%
     \DeclareUnicodeCharacter{0109}{\^c}%
     \DeclareUnicodeCharacter{010A}{\dotaccent{C}}%
     \DeclareUnicodeCharacter{010B}{\dotaccent{c}}%
     \DeclareUnicodeCharacter{010C}{\v{C}}%
     \DeclareUnicodeCharacter{010D}{\v{c}}%
     \DeclareUnicodeCharacter{010E}{\v{D}}%
     \DeclareUnicodeCharacter{010F}{d'}%
     %
     \DeclareUnicodeCharacter{0110}{\DH}%
     \DeclareUnicodeCharacter{0111}{\dh}%
     \DeclareUnicodeCharacter{0112}{\=E}%
     \DeclareUnicodeCharacter{0113}{\=e}%
     \DeclareUnicodeCharacter{0114}{\u{E}}%
     \DeclareUnicodeCharacter{0115}{\u{e}}%
     \DeclareUnicodeCharacter{0116}{\dotaccent{E}}%
     \DeclareUnicodeCharacter{0117}{\dotaccent{e}}%
     \DeclareUnicodeCharacter{0118}{\ogonek{E}}%
     \DeclareUnicodeCharacter{0119}{\ogonek{e}}%
     \DeclareUnicodeCharacter{011A}{\v{E}}%
     \DeclareUnicodeCharacter{011B}{\v{e}}%
     \DeclareUnicodeCharacter{011C}{\^G}%
     \DeclareUnicodeCharacter{011D}{\^g}%
     \DeclareUnicodeCharacter{011E}{\u{G}}%
     \DeclareUnicodeCharacter{011F}{\u{g}}%
     %
     \DeclareUnicodeCharacter{0120}{\dotaccent{G}}%
     \DeclareUnicodeCharacter{0121}{\dotaccent{g}}%
     \DeclareUnicodeCharacter{0122}{\cedilla{G}}%
     \DeclareUnicodeCharacter{0123}{\cedilla{g}}%
     \DeclareUnicodeCharacter{0124}{\^H}%
     \DeclareUnicodeCharacter{0125}{\^h}%
     \DeclareUnicodeCharacter{0126}{\missingcharmsg{H WITH STROKE}}%
     \DeclareUnicodeCharacter{0127}{\missingcharmsg{h WITH STROKE}}%
     \DeclareUnicodeCharacter{0128}{\~I}%
     \DeclareUnicodeCharacter{0129}{\~{\dotless{i}}}%
     \DeclareUnicodeCharacter{012A}{\=I}%
     \DeclareUnicodeCharacter{012B}{\={\dotless{i}}}%
     \DeclareUnicodeCharacter{012C}{\u{I}}%
     \DeclareUnicodeCharacter{012D}{\u{\dotless{i}}}%
     \DeclareUnicodeCharacter{012E}{\ogonek{I}}%
     \DeclareUnicodeCharacter{012F}{\ogonek{i}}%
     %
     \DeclareUnicodeCharacter{0130}{\dotaccent{I}}%
     \DeclareUnicodeCharacter{0131}{\dotless{i}}%
     \DeclareUnicodeCharacter{0132}{IJ}%
     \DeclareUnicodeCharacter{0133}{ij}%
     \DeclareUnicodeCharacter{0134}{\^J}%
     \DeclareUnicodeCharacter{0135}{\^{\dotless{j}}}%
     \DeclareUnicodeCharacter{0136}{\cedilla{K}}%
     \DeclareUnicodeCharacter{0137}{\cedilla{k}}%
     \DeclareUnicodeCharacter{0138}{\ensuremath\kappa}%
     \DeclareUnicodeCharacter{0139}{\'L}%
     \DeclareUnicodeCharacter{013A}{\'l}%
     \DeclareUnicodeCharacter{013B}{\cedilla{L}}%
     \DeclareUnicodeCharacter{013C}{\cedilla{l}}%
     \DeclareUnicodeCharacter{013D}{L'}% should kern
     \DeclareUnicodeCharacter{013E}{l'}% should kern
     \DeclareUnicodeCharacter{013F}{L\U{00B7}}%
     %
     \DeclareUnicodeCharacter{0140}{l\U{00B7}}%
     \DeclareUnicodeCharacter{0141}{\L}%
     \DeclareUnicodeCharacter{0142}{\l}%
     \DeclareUnicodeCharacter{0143}{\'N}%
     \DeclareUnicodeCharacter{0144}{\'n}%
     \DeclareUnicodeCharacter{0145}{\cedilla{N}}%
     \DeclareUnicodeCharacter{0146}{\cedilla{n}}%
     \DeclareUnicodeCharacter{0147}{\v{N}}%
     \DeclareUnicodeCharacter{0148}{\v{n}}%
     \DeclareUnicodeCharacter{0149}{'n}%
     \DeclareUnicodeCharacter{014A}{\missingcharmsg{ENG}}%
     \DeclareUnicodeCharacter{014B}{\missingcharmsg{eng}}%
     \DeclareUnicodeCharacter{014C}{\=O}%
     \DeclareUnicodeCharacter{014D}{\=o}%
     \DeclareUnicodeCharacter{014E}{\u{O}}%
     \DeclareUnicodeCharacter{014F}{\u{o}}%
     %
     \DeclareUnicodeCharacter{0150}{\H{O}}%
     \DeclareUnicodeCharacter{0151}{\H{o}}%
     \DeclareUnicodeCharacter{0152}{\OE}%
     \DeclareUnicodeCharacter{0153}{\oe}%
     \DeclareUnicodeCharacter{0154}{\'R}%
     \DeclareUnicodeCharacter{0155}{\'r}%
     \DeclareUnicodeCharacter{0156}{\cedilla{R}}%
     \DeclareUnicodeCharacter{0157}{\cedilla{r}}%
     \DeclareUnicodeCharacter{0158}{\v{R}}%
     \DeclareUnicodeCharacter{0159}{\v{r}}%
     \DeclareUnicodeCharacter{015A}{\'S}%
     \DeclareUnicodeCharacter{015B}{\'s}%
     \DeclareUnicodeCharacter{015C}{\^S}%
     \DeclareUnicodeCharacter{015D}{\^s}%
     \DeclareUnicodeCharacter{015E}{\cedilla{S}}%
     \DeclareUnicodeCharacter{015F}{\cedilla{s}}%
     %
     \DeclareUnicodeCharacter{0160}{\v{S}}%
     \DeclareUnicodeCharacter{0161}{\v{s}}%
     \DeclareUnicodeCharacter{0162}{\cedilla{T}}%
     \DeclareUnicodeCharacter{0163}{\cedilla{t}}%
     \DeclareUnicodeCharacter{0164}{\v{T}}%
     \DeclareUnicodeCharacter{0165}{\v{t}}%
     \DeclareUnicodeCharacter{0166}{\missingcharmsg{H WITH STROKE}}%
     \DeclareUnicodeCharacter{0167}{\missingcharmsg{h WITH STROKE}}%
     \DeclareUnicodeCharacter{0168}{\~U}%
     \DeclareUnicodeCharacter{0169}{\~u}%
     \DeclareUnicodeCharacter{016A}{\=U}%
     \DeclareUnicodeCharacter{016B}{\=u}%
     \DeclareUnicodeCharacter{016C}{\u{U}}%
     \DeclareUnicodeCharacter{016D}{\u{u}}%
     \DeclareUnicodeCharacter{016E}{\ringaccent{U}}%
     \DeclareUnicodeCharacter{016F}{\ringaccent{u}}%
     %
     \DeclareUnicodeCharacter{0170}{\H{U}}%
     \DeclareUnicodeCharacter{0171}{\H{u}}%
     \DeclareUnicodeCharacter{0172}{\ogonek{U}}%
     \DeclareUnicodeCharacter{0173}{\ogonek{u}}%
     \DeclareUnicodeCharacter{0174}{\^W}%
     \DeclareUnicodeCharacter{0175}{\^w}%
     \DeclareUnicodeCharacter{0176}{\^Y}%
     \DeclareUnicodeCharacter{0177}{\^y}%
     \DeclareUnicodeCharacter{0178}{\"Y}%
     \DeclareUnicodeCharacter{0179}{\'Z}%
     \DeclareUnicodeCharacter{017A}{\'z}%
     \DeclareUnicodeCharacter{017B}{\dotaccent{Z}}%
     \DeclareUnicodeCharacter{017C}{\dotaccent{z}}%
     \DeclareUnicodeCharacter{017D}{\v{Z}}%
     \DeclareUnicodeCharacter{017E}{\v{z}}%
     \DeclareUnicodeCharacter{017F}{\missingcharmsg{LONG S}}%
     %
     \DeclareUnicodeCharacter{01C4}{D\v{Z}}%
     \DeclareUnicodeCharacter{01C5}{D\v{z}}%
     \DeclareUnicodeCharacter{01C6}{d\v{z}}%
     \DeclareUnicodeCharacter{01C7}{LJ}%
     \DeclareUnicodeCharacter{01C8}{Lj}%
     \DeclareUnicodeCharacter{01C9}{lj}%
     \DeclareUnicodeCharacter{01CA}{NJ}%
     \DeclareUnicodeCharacter{01CB}{Nj}%
     \DeclareUnicodeCharacter{01CC}{nj}%
     \DeclareUnicodeCharacter{01CD}{\v{A}}%
     \DeclareUnicodeCharacter{01CE}{\v{a}}%
     \DeclareUnicodeCharacter{01CF}{\v{I}}%
     %
     \DeclareUnicodeCharacter{01D0}{\v{\dotless{i}}}%
     \DeclareUnicodeCharacter{01D1}{\v{O}}%
     \DeclareUnicodeCharacter{01D2}{\v{o}}%
     \DeclareUnicodeCharacter{01D3}{\v{U}}%
     \DeclareUnicodeCharacter{01D4}{\v{u}}%
     %
     \DeclareUnicodeCharacter{01E2}{\={\AE}}%
     \DeclareUnicodeCharacter{01E3}{\={\ae}}%
     \DeclareUnicodeCharacter{01E6}{\v{G}}%
     \DeclareUnicodeCharacter{01E7}{\v{g}}%
     \DeclareUnicodeCharacter{01E8}{\v{K}}%
     \DeclareUnicodeCharacter{01E9}{\v{k}}%
     %
     \DeclareUnicodeCharacter{01F0}{\v{\dotless{j}}}%
     \DeclareUnicodeCharacter{01F1}{DZ}%
     \DeclareUnicodeCharacter{01F2}{Dz}%
     \DeclareUnicodeCharacter{01F3}{dz}%
     \DeclareUnicodeCharacter{01F4}{\'G}%
     \DeclareUnicodeCharacter{01F5}{\'g}%
     \DeclareUnicodeCharacter{01F8}{\`N}%
     \DeclareUnicodeCharacter{01F9}{\`n}%
     \DeclareUnicodeCharacter{01FC}{\'{\AE}}%
     \DeclareUnicodeCharacter{01FD}{\'{\ae}}%
     \DeclareUnicodeCharacter{01FE}{\'{\O}}%
     \DeclareUnicodeCharacter{01FF}{\'{\o}}%
     %
     \DeclareUnicodeCharacter{021E}{\v{H}}%
     \DeclareUnicodeCharacter{021F}{\v{h}}%
     %
     \DeclareUnicodeCharacter{0226}{\dotaccent{A}}%
     \DeclareUnicodeCharacter{0227}{\dotaccent{a}}%
     \DeclareUnicodeCharacter{0228}{\cedilla{E}}%
     \DeclareUnicodeCharacter{0229}{\cedilla{e}}%
     \DeclareUnicodeCharacter{022E}{\dotaccent{O}}%
     \DeclareUnicodeCharacter{022F}{\dotaccent{o}}%
     %
     \DeclareUnicodeCharacter{0232}{\=Y}%
     \DeclareUnicodeCharacter{0233}{\=y}%
     \DeclareUnicodeCharacter{0237}{\dotless{j}}%
     %
     \DeclareUnicodeCharacter{02DB}{\ogonek{ }}%
     %
     % Greek letters upper case
     \DeclareUnicodeCharacter{0391}{{\it A}}%
     \DeclareUnicodeCharacter{0392}{{\it B}}%
     \DeclareUnicodeCharacter{0393}{\ensuremath{\mit\Gamma}}%
     \DeclareUnicodeCharacter{0394}{\ensuremath{\mit\Delta}}%
     \DeclareUnicodeCharacter{0395}{{\it E}}%
     \DeclareUnicodeCharacter{0396}{{\it Z}}%
     \DeclareUnicodeCharacter{0397}{{\it H}}%
     \DeclareUnicodeCharacter{0398}{\ensuremath{\mit\Theta}}%
     \DeclareUnicodeCharacter{0399}{{\it I}}%
     \DeclareUnicodeCharacter{039A}{{\it K}}%
     \DeclareUnicodeCharacter{039B}{\ensuremath{\mit\Lambda}}%
     \DeclareUnicodeCharacter{039C}{{\it M}}%
     \DeclareUnicodeCharacter{039D}{{\it N}}%
     \DeclareUnicodeCharacter{039E}{\ensuremath{\mit\Xi}}%
     \DeclareUnicodeCharacter{039F}{{\it O}}%
     \DeclareUnicodeCharacter{03A0}{\ensuremath{\mit\Pi}}%
     \DeclareUnicodeCharacter{03A1}{{\it P}}%
     %\DeclareUnicodeCharacter{03A2}{} % none - corresponds to final sigma
     \DeclareUnicodeCharacter{03A3}{\ensuremath{\mit\Sigma}}%
     \DeclareUnicodeCharacter{03A4}{{\it T}}%
     \DeclareUnicodeCharacter{03A5}{\ensuremath{\mit\Upsilon}}%
     \DeclareUnicodeCharacter{03A6}{\ensuremath{\mit\Phi}}%
     \DeclareUnicodeCharacter{03A7}{{\it X}}%
     \DeclareUnicodeCharacter{03A8}{\ensuremath{\mit\Psi}}%
     \DeclareUnicodeCharacter{03A9}{\ensuremath{\mit\Omega}}%
     %
     % Vowels with accents
     \DeclareUnicodeCharacter{0390}{\ensuremath{\ddot{\acute\iota}}}%
     \DeclareUnicodeCharacter{03AC}{\ensuremath{\acute\alpha}}%
     \DeclareUnicodeCharacter{03AD}{\ensuremath{\acute\epsilon}}%
     \DeclareUnicodeCharacter{03AE}{\ensuremath{\acute\eta}}%
     \DeclareUnicodeCharacter{03AF}{\ensuremath{\acute\iota}}%
     \DeclareUnicodeCharacter{03B0}{\ensuremath{\acute{\ddot\upsilon}}}%
     %
     % Standalone accent
     \DeclareUnicodeCharacter{0384}{\ensuremath{\acute{\ }}}%
     %
     % Greek letters lower case
     \DeclareUnicodeCharacter{03B1}{\ensuremath\alpha}%
     \DeclareUnicodeCharacter{03B2}{\ensuremath\beta}%
     \DeclareUnicodeCharacter{03B3}{\ensuremath\gamma}%
     \DeclareUnicodeCharacter{03B4}{\ensuremath\delta}%
     \DeclareUnicodeCharacter{03B5}{\ensuremath\epsilon}%
     \DeclareUnicodeCharacter{03B6}{\ensuremath\zeta}%
     \DeclareUnicodeCharacter{03B7}{\ensuremath\eta}%
     \DeclareUnicodeCharacter{03B8}{\ensuremath\theta}%
     \DeclareUnicodeCharacter{03B9}{\ensuremath\iota}%
     \DeclareUnicodeCharacter{03BA}{\ensuremath\kappa}%
     \DeclareUnicodeCharacter{03BB}{\ensuremath\lambda}%
     \DeclareUnicodeCharacter{03BC}{\ensuremath\mu}%
     \DeclareUnicodeCharacter{03BD}{\ensuremath\nu}%
     \DeclareUnicodeCharacter{03BE}{\ensuremath\xi}%
     \DeclareUnicodeCharacter{03BF}{{\it o}}% omicron
     \DeclareUnicodeCharacter{03C0}{\ensuremath\pi}%
     \DeclareUnicodeCharacter{03C1}{\ensuremath\rho}%
     \DeclareUnicodeCharacter{03C2}{\ensuremath\varsigma}%
     \DeclareUnicodeCharacter{03C3}{\ensuremath\sigma}%
     \DeclareUnicodeCharacter{03C4}{\ensuremath\tau}%
     \DeclareUnicodeCharacter{03C5}{\ensuremath\upsilon}%
     \DeclareUnicodeCharacter{03C6}{\ensuremath\phi}%
     \DeclareUnicodeCharacter{03C7}{\ensuremath\chi}%
     \DeclareUnicodeCharacter{03C8}{\ensuremath\psi}%
     \DeclareUnicodeCharacter{03C9}{\ensuremath\omega}%
     %
     % More Greek vowels with accents
     \DeclareUnicodeCharacter{03CA}{\ensuremath{\ddot\iota}}%
     \DeclareUnicodeCharacter{03CB}{\ensuremath{\ddot\upsilon}}%
     \DeclareUnicodeCharacter{03CC}{\ensuremath{\acute o}}%
     \DeclareUnicodeCharacter{03CD}{\ensuremath{\acute\upsilon}}%
     \DeclareUnicodeCharacter{03CE}{\ensuremath{\acute\omega}}%
     %
     % Variant Greek letters
     \DeclareUnicodeCharacter{03D1}{\ensuremath\vartheta}%
     \DeclareUnicodeCharacter{03D6}{\ensuremath\varpi}%
     \DeclareUnicodeCharacter{03F1}{\ensuremath\varrho}%
     %
     \DeclareUnicodeCharacter{1E02}{\dotaccent{B}}%
     \DeclareUnicodeCharacter{1E03}{\dotaccent{b}}%
     \DeclareUnicodeCharacter{1E04}{\udotaccent{B}}%
     \DeclareUnicodeCharacter{1E05}{\udotaccent{b}}%
     \DeclareUnicodeCharacter{1E06}{\ubaraccent{B}}%
     \DeclareUnicodeCharacter{1E07}{\ubaraccent{b}}%
     \DeclareUnicodeCharacter{1E0A}{\dotaccent{D}}%
     \DeclareUnicodeCharacter{1E0B}{\dotaccent{d}}%
     \DeclareUnicodeCharacter{1E0C}{\udotaccent{D}}%
     \DeclareUnicodeCharacter{1E0D}{\udotaccent{d}}%
     \DeclareUnicodeCharacter{1E0E}{\ubaraccent{D}}%
     \DeclareUnicodeCharacter{1E0F}{\ubaraccent{d}}%
     %
     \DeclareUnicodeCharacter{1E1E}{\dotaccent{F}}%
     \DeclareUnicodeCharacter{1E1F}{\dotaccent{f}}%
     %
     \DeclareUnicodeCharacter{1E20}{\=G}%
     \DeclareUnicodeCharacter{1E21}{\=g}%
     \DeclareUnicodeCharacter{1E22}{\dotaccent{H}}%
     \DeclareUnicodeCharacter{1E23}{\dotaccent{h}}%
     \DeclareUnicodeCharacter{1E24}{\udotaccent{H}}%
     \DeclareUnicodeCharacter{1E25}{\udotaccent{h}}%
     \DeclareUnicodeCharacter{1E26}{\"H}%
     \DeclareUnicodeCharacter{1E27}{\"h}%
     %
     \DeclareUnicodeCharacter{1E30}{\'K}%
     \DeclareUnicodeCharacter{1E31}{\'k}%
     \DeclareUnicodeCharacter{1E32}{\udotaccent{K}}%
     \DeclareUnicodeCharacter{1E33}{\udotaccent{k}}%
     \DeclareUnicodeCharacter{1E34}{\ubaraccent{K}}%
     \DeclareUnicodeCharacter{1E35}{\ubaraccent{k}}%
     \DeclareUnicodeCharacter{1E36}{\udotaccent{L}}%
     \DeclareUnicodeCharacter{1E37}{\udotaccent{l}}%
     \DeclareUnicodeCharacter{1E3A}{\ubaraccent{L}}%
     \DeclareUnicodeCharacter{1E3B}{\ubaraccent{l}}%
     \DeclareUnicodeCharacter{1E3E}{\'M}%
     \DeclareUnicodeCharacter{1E3F}{\'m}%
     %
     \DeclareUnicodeCharacter{1E40}{\dotaccent{M}}%
     \DeclareUnicodeCharacter{1E41}{\dotaccent{m}}%
     \DeclareUnicodeCharacter{1E42}{\udotaccent{M}}%
     \DeclareUnicodeCharacter{1E43}{\udotaccent{m}}%
     \DeclareUnicodeCharacter{1E44}{\dotaccent{N}}%
     \DeclareUnicodeCharacter{1E45}{\dotaccent{n}}%
     \DeclareUnicodeCharacter{1E46}{\udotaccent{N}}%
     \DeclareUnicodeCharacter{1E47}{\udotaccent{n}}%
     \DeclareUnicodeCharacter{1E48}{\ubaraccent{N}}%
     \DeclareUnicodeCharacter{1E49}{\ubaraccent{n}}%
     %
     \DeclareUnicodeCharacter{1E54}{\'P}%
     \DeclareUnicodeCharacter{1E55}{\'p}%
     \DeclareUnicodeCharacter{1E56}{\dotaccent{P}}%
     \DeclareUnicodeCharacter{1E57}{\dotaccent{p}}%
     \DeclareUnicodeCharacter{1E58}{\dotaccent{R}}%
     \DeclareUnicodeCharacter{1E59}{\dotaccent{r}}%
     \DeclareUnicodeCharacter{1E5A}{\udotaccent{R}}%
     \DeclareUnicodeCharacter{1E5B}{\udotaccent{r}}%
     \DeclareUnicodeCharacter{1E5E}{\ubaraccent{R}}%
     \DeclareUnicodeCharacter{1E5F}{\ubaraccent{r}}%
     %
     \DeclareUnicodeCharacter{1E60}{\dotaccent{S}}%
     \DeclareUnicodeCharacter{1E61}{\dotaccent{s}}%
     \DeclareUnicodeCharacter{1E62}{\udotaccent{S}}%
     \DeclareUnicodeCharacter{1E63}{\udotaccent{s}}%
     \DeclareUnicodeCharacter{1E6A}{\dotaccent{T}}%
     \DeclareUnicodeCharacter{1E6B}{\dotaccent{t}}%
     \DeclareUnicodeCharacter{1E6C}{\udotaccent{T}}%
     \DeclareUnicodeCharacter{1E6D}{\udotaccent{t}}%
     \DeclareUnicodeCharacter{1E6E}{\ubaraccent{T}}%
     \DeclareUnicodeCharacter{1E6F}{\ubaraccent{t}}%
     %
     \DeclareUnicodeCharacter{1E7C}{\~V}%
     \DeclareUnicodeCharacter{1E7D}{\~v}%
     \DeclareUnicodeCharacter{1E7E}{\udotaccent{V}}%
     \DeclareUnicodeCharacter{1E7F}{\udotaccent{v}}%
     %
     \DeclareUnicodeCharacter{1E80}{\`W}%
     \DeclareUnicodeCharacter{1E81}{\`w}%
     \DeclareUnicodeCharacter{1E82}{\'W}%
     \DeclareUnicodeCharacter{1E83}{\'w}%
     \DeclareUnicodeCharacter{1E84}{\"W}%
     \DeclareUnicodeCharacter{1E85}{\"w}%
     \DeclareUnicodeCharacter{1E86}{\dotaccent{W}}%
     \DeclareUnicodeCharacter{1E87}{\dotaccent{w}}%
     \DeclareUnicodeCharacter{1E88}{\udotaccent{W}}%
     \DeclareUnicodeCharacter{1E89}{\udotaccent{w}}%
     \DeclareUnicodeCharacter{1E8A}{\dotaccent{X}}%
     \DeclareUnicodeCharacter{1E8B}{\dotaccent{x}}%
     \DeclareUnicodeCharacter{1E8C}{\"X}%
     \DeclareUnicodeCharacter{1E8D}{\"x}%
     \DeclareUnicodeCharacter{1E8E}{\dotaccent{Y}}%
     \DeclareUnicodeCharacter{1E8F}{\dotaccent{y}}%
     %
     \DeclareUnicodeCharacter{1E90}{\^Z}%
     \DeclareUnicodeCharacter{1E91}{\^z}%
     \DeclareUnicodeCharacter{1E92}{\udotaccent{Z}}%
     \DeclareUnicodeCharacter{1E93}{\udotaccent{z}}%
     \DeclareUnicodeCharacter{1E94}{\ubaraccent{Z}}%
     \DeclareUnicodeCharacter{1E95}{\ubaraccent{z}}%
     \DeclareUnicodeCharacter{1E96}{\ubaraccent{h}}%
     \DeclareUnicodeCharacter{1E97}{\"t}%
     \DeclareUnicodeCharacter{1E98}{\ringaccent{w}}%
     \DeclareUnicodeCharacter{1E99}{\ringaccent{y}}%
     %
     \DeclareUnicodeCharacter{1EA0}{\udotaccent{A}}%
     \DeclareUnicodeCharacter{1EA1}{\udotaccent{a}}%
     %
     \DeclareUnicodeCharacter{1EB8}{\udotaccent{E}}%
     \DeclareUnicodeCharacter{1EB9}{\udotaccent{e}}%
     \DeclareUnicodeCharacter{1EBC}{\~E}%
     \DeclareUnicodeCharacter{1EBD}{\~e}%
     %
     \DeclareUnicodeCharacter{1ECA}{\udotaccent{I}}%
     \DeclareUnicodeCharacter{1ECB}{\udotaccent{i}}%
     \DeclareUnicodeCharacter{1ECC}{\udotaccent{O}}%
     \DeclareUnicodeCharacter{1ECD}{\udotaccent{o}}%
     %
     \DeclareUnicodeCharacter{1EE4}{\udotaccent{U}}%
     \DeclareUnicodeCharacter{1EE5}{\udotaccent{u}}%
     %
     \DeclareUnicodeCharacter{1EF2}{\`Y}%
     \DeclareUnicodeCharacter{1EF3}{\`y}%
     \DeclareUnicodeCharacter{1EF4}{\udotaccent{Y}}%
     %
     \DeclareUnicodeCharacter{1EF8}{\~Y}%
     \DeclareUnicodeCharacter{1EF9}{\~y}%
     %
     % Punctuation
     \DeclareUnicodeCharacter{2013}{--}%
     \DeclareUnicodeCharacter{2014}{---}%
     \DeclareUnicodeCharacter{2018}{\quoteleft}%
     \DeclareUnicodeCharacter{2019}{\quoteright}%
     \DeclareUnicodeCharacter{201A}{\quotesinglbase}%
     \DeclareUnicodeCharacter{201C}{\quotedblleft}%
     \DeclareUnicodeCharacter{201D}{\quotedblright}%
     \DeclareUnicodeCharacter{201E}{\quotedblbase}%
     \DeclareUnicodeCharacter{2020}{\ensuremath\dagger}%
     \DeclareUnicodeCharacter{2021}{\ensuremath\ddagger}%
     \DeclareUnicodeCharacter{2022}{\bullet}%
     \DeclareUnicodeCharacter{202F}{\thinspace}%
     \DeclareUnicodeCharacter{2026}{\dots}%
     \DeclareUnicodeCharacter{2039}{\guilsinglleft}%
     \DeclareUnicodeCharacter{203A}{\guilsinglright}%
     %
     \DeclareUnicodeCharacter{20AC}{\euro}%
     %
     \DeclareUnicodeCharacter{2192}{\expansion}%
     \DeclareUnicodeCharacter{21D2}{\result}%
     %
     % Mathematical symbols
     \DeclareUnicodeCharacter{2200}{\ensuremath\forall}%
     \DeclareUnicodeCharacter{2203}{\ensuremath\exists}%
     \DeclareUnicodeCharacter{2208}{\ensuremath\in}%
     \DeclareUnicodeCharacter{2212}{\minus}%
     \DeclareUnicodeCharacter{2217}{\ast}%
     \DeclareUnicodeCharacter{221E}{\ensuremath\infty}%
     \DeclareUnicodeCharacter{2225}{\ensuremath\parallel}%
     \DeclareUnicodeCharacter{2227}{\ensuremath\wedge}%
     \DeclareUnicodeCharacter{2229}{\ensuremath\cap}%
     \DeclareUnicodeCharacter{2261}{\equiv}%
     \DeclareUnicodeCharacter{2264}{\ensuremath\leq}%
     \DeclareUnicodeCharacter{2265}{\ensuremath\geq}%
     \DeclareUnicodeCharacter{2282}{\ensuremath\subset}%
     \DeclareUnicodeCharacter{2287}{\ensuremath\supseteq}%
     %
     \DeclareUnicodeCharacter{2016}{\ensuremath\Vert}%
     \DeclareUnicodeCharacter{2032}{\ensuremath\prime}%
     \DeclareUnicodeCharacter{210F}{\ensuremath\hbar}%
     \DeclareUnicodeCharacter{2111}{\ensuremath\Im}%
     \DeclareUnicodeCharacter{2113}{\ensuremath\ell}%
     \DeclareUnicodeCharacter{2118}{\ensuremath\wp}%
     \DeclareUnicodeCharacter{211C}{\ensuremath\Re}%
     \DeclareUnicodeCharacter{2127}{\ensuremath\mho}%
     \DeclareUnicodeCharacter{2135}{\ensuremath\aleph}%
     \DeclareUnicodeCharacter{2190}{\ensuremath\leftarrow}%
     \DeclareUnicodeCharacter{2191}{\ensuremath\uparrow}%
     \DeclareUnicodeCharacter{2193}{\ensuremath\downarrow}%
     \DeclareUnicodeCharacter{2194}{\ensuremath\leftrightarrow}%
     \DeclareUnicodeCharacter{2195}{\ensuremath\updownarrow}%
     \DeclareUnicodeCharacter{2196}{\ensuremath\nwarrow}%
     \DeclareUnicodeCharacter{2197}{\ensuremath\nearrow}%
     \DeclareUnicodeCharacter{2198}{\ensuremath\searrow}%
     \DeclareUnicodeCharacter{2199}{\ensuremath\swarrow}%
     \DeclareUnicodeCharacter{21A6}{\ensuremath\mapsto}%
     \DeclareUnicodeCharacter{21A9}{\ensuremath\hookleftarrow}%
     \DeclareUnicodeCharacter{21AA}{\ensuremath\hookrightarrow}%
     \DeclareUnicodeCharacter{21BC}{\ensuremath\leftharpoonup}%
     \DeclareUnicodeCharacter{21BD}{\ensuremath\leftharpoondown}%
     \DeclareUnicodeCharacter{21BE}{\ensuremath\upharpoonright}%
     \DeclareUnicodeCharacter{21C0}{\ensuremath\rightharpoonup}%
     \DeclareUnicodeCharacter{21C1}{\ensuremath\rightharpoondown}%
     \DeclareUnicodeCharacter{21CC}{\ensuremath\rightleftharpoons}%
     \DeclareUnicodeCharacter{21D0}{\ensuremath\Leftarrow}%
     \DeclareUnicodeCharacter{21D1}{\ensuremath\Uparrow}%
     \DeclareUnicodeCharacter{21D3}{\ensuremath\Downarrow}%
     \DeclareUnicodeCharacter{21D4}{\ensuremath\Leftrightarrow}%
     \DeclareUnicodeCharacter{21D5}{\ensuremath\Updownarrow}%
     \DeclareUnicodeCharacter{21DD}{\ensuremath\leadsto}%
     \DeclareUnicodeCharacter{2201}{\ensuremath\complement}%
     \DeclareUnicodeCharacter{2202}{\ensuremath\partial}%
     \DeclareUnicodeCharacter{2205}{\ensuremath\emptyset}%
     \DeclareUnicodeCharacter{2207}{\ensuremath\nabla}%
     \DeclareUnicodeCharacter{2209}{\ensuremath\notin}%
     \DeclareUnicodeCharacter{220B}{\ensuremath\owns}%
     \DeclareUnicodeCharacter{220F}{\ensuremath\prod}%
     \DeclareUnicodeCharacter{2210}{\ensuremath\coprod}%
     \DeclareUnicodeCharacter{2211}{\ensuremath\sum}%
     \DeclareUnicodeCharacter{2213}{\ensuremath\mp}%
     \DeclareUnicodeCharacter{2218}{\ensuremath\circ}%
     \DeclareUnicodeCharacter{221A}{\ensuremath\surd}%
     \DeclareUnicodeCharacter{221D}{\ensuremath\propto}%
     \DeclareUnicodeCharacter{2220}{\ensuremath\angle}%
     \DeclareUnicodeCharacter{2223}{\ensuremath\mid}%
     \DeclareUnicodeCharacter{2228}{\ensuremath\vee}%
     \DeclareUnicodeCharacter{222A}{\ensuremath\cup}%
     \DeclareUnicodeCharacter{222B}{\ensuremath\smallint}%
     \DeclareUnicodeCharacter{222E}{\ensuremath\oint}%
     \DeclareUnicodeCharacter{223C}{\ensuremath\sim}%
     \DeclareUnicodeCharacter{2240}{\ensuremath\wr}%
     \DeclareUnicodeCharacter{2243}{\ensuremath\simeq}%
     \DeclareUnicodeCharacter{2245}{\ensuremath\cong}%
     \DeclareUnicodeCharacter{2248}{\ensuremath\approx}%
     \DeclareUnicodeCharacter{224D}{\ensuremath\asymp}%
     \DeclareUnicodeCharacter{2250}{\ensuremath\doteq}%
     \DeclareUnicodeCharacter{2260}{\ensuremath\neq}%
     \DeclareUnicodeCharacter{226A}{\ensuremath\ll}%
     \DeclareUnicodeCharacter{226B}{\ensuremath\gg}%
     \DeclareUnicodeCharacter{227A}{\ensuremath\prec}%
     \DeclareUnicodeCharacter{227B}{\ensuremath\succ}%
     \DeclareUnicodeCharacter{2283}{\ensuremath\supset}%
     \DeclareUnicodeCharacter{2286}{\ensuremath\subseteq}%
     \DeclareUnicodeCharacter{228E}{\ensuremath\uplus}%
     \DeclareUnicodeCharacter{228F}{\ensuremath\sqsubset}%
     \DeclareUnicodeCharacter{2290}{\ensuremath\sqsupset}%
     \DeclareUnicodeCharacter{2291}{\ensuremath\sqsubseteq}%
     \DeclareUnicodeCharacter{2292}{\ensuremath\sqsupseteq}%
     \DeclareUnicodeCharacter{2293}{\ensuremath\sqcap}%
     \DeclareUnicodeCharacter{2294}{\ensuremath\sqcup}%
     \DeclareUnicodeCharacter{2295}{\ensuremath\oplus}%
     \DeclareUnicodeCharacter{2296}{\ensuremath\ominus}%
     \DeclareUnicodeCharacter{2297}{\ensuremath\otimes}%
     \DeclareUnicodeCharacter{2298}{\ensuremath\oslash}%
     \DeclareUnicodeCharacter{2299}{\ensuremath\odot}%
     \DeclareUnicodeCharacter{22A2}{\ensuremath\vdash}%
     \DeclareUnicodeCharacter{22A3}{\ensuremath\dashv}%
     \DeclareUnicodeCharacter{22A4}{\ensuremath\ptextop}%
     \DeclareUnicodeCharacter{22A5}{\ensuremath\bot}%
     \DeclareUnicodeCharacter{22A8}{\ensuremath\models}%
     \DeclareUnicodeCharacter{22B4}{\ensuremath\unlhd}%
     \DeclareUnicodeCharacter{22B5}{\ensuremath\unrhd}%
     \DeclareUnicodeCharacter{22C0}{\ensuremath\bigwedge}%
     \DeclareUnicodeCharacter{22C1}{\ensuremath\bigvee}%
     \DeclareUnicodeCharacter{22C2}{\ensuremath\bigcap}%
     \DeclareUnicodeCharacter{22C3}{\ensuremath\bigcup}%
     \DeclareUnicodeCharacter{22C4}{\ensuremath\diamond}%
     \DeclareUnicodeCharacter{22C5}{\ensuremath\cdot}%
     \DeclareUnicodeCharacter{22C6}{\ensuremath\star}%
     \DeclareUnicodeCharacter{22C8}{\ensuremath\bowtie}%
     \DeclareUnicodeCharacter{2308}{\ensuremath\lceil}%
     \DeclareUnicodeCharacter{2309}{\ensuremath\rceil}%
     \DeclareUnicodeCharacter{230A}{\ensuremath\lfloor}%
     \DeclareUnicodeCharacter{230B}{\ensuremath\rfloor}%
     \DeclareUnicodeCharacter{2322}{\ensuremath\frown}%
     \DeclareUnicodeCharacter{2323}{\ensuremath\smile}%
     %
     \DeclareUnicodeCharacter{25A1}{\ensuremath\Box}%
     \DeclareUnicodeCharacter{25B3}{\ensuremath\triangle}%
     \DeclareUnicodeCharacter{25B7}{\ensuremath\triangleright}%
     \DeclareUnicodeCharacter{25BD}{\ensuremath\bigtriangledown}%
     \DeclareUnicodeCharacter{25C1}{\ensuremath\triangleleft}%
     \DeclareUnicodeCharacter{25C7}{\ensuremath\Diamond}%
     \DeclareUnicodeCharacter{2660}{\ensuremath\spadesuit}%
     \DeclareUnicodeCharacter{2661}{\ensuremath\heartsuit}%
     \DeclareUnicodeCharacter{2662}{\ensuremath\diamondsuit}%
     \DeclareUnicodeCharacter{2663}{\ensuremath\clubsuit}%
     \DeclareUnicodeCharacter{266D}{\ensuremath\flat}%
     \DeclareUnicodeCharacter{266E}{\ensuremath\natural}%
     \DeclareUnicodeCharacter{266F}{\ensuremath\sharp}%
     \DeclareUnicodeCharacter{26AA}{\ensuremath\bigcirc}%
     \DeclareUnicodeCharacter{27B9}{\ensuremath\rangle}%
     \DeclareUnicodeCharacter{27C2}{\ensuremath\perp}%
     \DeclareUnicodeCharacter{27E8}{\ensuremath\langle}%
     \DeclareUnicodeCharacter{27F5}{\ensuremath\longleftarrow}%
     \DeclareUnicodeCharacter{27F6}{\ensuremath\longrightarrow}%
     \DeclareUnicodeCharacter{27F7}{\ensuremath\longleftrightarrow}%
     \DeclareUnicodeCharacter{27FC}{\ensuremath\longmapsto}%
     \DeclareUnicodeCharacter{29F5}{\ensuremath\setminus}%
     \DeclareUnicodeCharacter{2A00}{\ensuremath\bigodot}%
     \DeclareUnicodeCharacter{2A01}{\ensuremath\bigoplus}%
     \DeclareUnicodeCharacter{2A02}{\ensuremath\bigotimes}%
     \DeclareUnicodeCharacter{2A04}{\ensuremath\biguplus}%
     \DeclareUnicodeCharacter{2A06}{\ensuremath\bigsqcup}%
     \DeclareUnicodeCharacter{2A1D}{\ensuremath\Join}%
     \DeclareUnicodeCharacter{2A3F}{\ensuremath\amalg}%
     \DeclareUnicodeCharacter{2AAF}{\ensuremath\preceq}%
     \DeclareUnicodeCharacter{2AB0}{\ensuremath\succeq}%
     %
     \global\mathchardef\checkmark="1370% actually the square root sign
     \DeclareUnicodeCharacter{2713}{\ensuremath\checkmark}%
   }% end of \unicodechardefs
   
   % UTF-8 byte sequence (pdfTeX) definitions (replacing and @U command)
   % It makes the setting that replace UTF-8 byte sequence.
   \def\utfeightchardefs{%
     \let\DeclareUnicodeCharacter\DeclareUnicodeCharacterUTFviii
     \unicodechardefs
   }
   
   % Whether the active definitions of non-ASCII characters expand to
   % non-active tokens with the same character code.  This is used to
   % write characters literally, instead of using active definitions for
   % printing the correct glyphs.
   \newif\ifpassthroughchars
   \passthroughcharsfalse
   
   % For native Unicode handling (XeTeX and LuaTeX),
   % provide a definition macro to replace/pass-through a Unicode character
   %
   \def\DeclareUnicodeCharacterNative#1#2{%
     \catcode"#1=\active
     \def\dodeclareunicodecharacternative##1##2##3{%
       \begingroup
         \uccode`\~="##2\relax
         \uppercase{\gdef~}{%
           \ifpassthroughchars
             ##1%
           \else
             ##3%
           \fi
         }
       \endgroup
     }
     \begingroup
       \uccode`\.="#1\relax
       \uppercase{\def\UTFNativeTmp{.}}%
       \expandafter\dodeclareunicodecharacternative\UTFNativeTmp{#1}{#2}%
     \endgroup
   }
   
   % Native Unicode handling (XeTeX and LuaTeX) character replacing definition.
   % It activates the setting that replaces Unicode characters.
   \def\nativeunicodechardefs{%
     \let\DeclareUnicodeCharacter\DeclareUnicodeCharacterNative
     \unicodechardefs
   }
   
   % For native Unicode handling (XeTeX and LuaTeX),
   % make the character token expand
   % to the sequences given in \unicodechardefs for printing.
   \def\DeclareUnicodeCharacterNativeAtU#1#2{%
     \def\UTFAtUTmp{#2}
     \expandafter\globallet\csname uni:#1\endcsname \UTFAtUTmp
   }
   
   % @U command definitions for native Unicode handling (XeTeX and LuaTeX).
   \def\nativeunicodechardefsatu{%
     \let\DeclareUnicodeCharacter\DeclareUnicodeCharacterNativeAtU
     \unicodechardefs
   }
   
   % US-ASCII character definitions.
   \def\asciichardefs{% nothing need be done
      \relax
   }
   
   % define all Unicode characters we know about, for the sake of @U.
   \iftxinativeunicodecapable
     \nativeunicodechardefsatu
   \else
     \utfeightchardefs
   \fi
   
   
   % Make non-ASCII characters printable again for compatibility with
   % existing Texinfo documents that may use them, even without declaring a
   % document encoding.
   %
   \setnonasciicharscatcode \other
   
   
   \message{formatting,}
   
 \newdimen\defaultparindent \defaultparindent = 15pt  \newdimen\defaultparindent \defaultparindent = 15pt
   
 \chapheadingskip = 15pt plus 4pt minus 2pt  \chapheadingskip = 15pt plus 4pt minus 2pt
Line 5692  should work if nowhere else does.}
Line 11267  should work if nowhere else does.}
 % Prevent underfull vbox error messages.  % Prevent underfull vbox error messages.
 \vbadness = 10000  \vbadness = 10000
   
 % Don't be so finicky about underfull hboxes, either.  % Don't be very finicky about underfull hboxes, either.
 \hbadness = 2000  \hbadness = 6666
   
 % Following George Bush, just get rid of widows and orphans.  % Following George Bush, get rid of widows and orphans.
 \widowpenalty=10000  \widowpenalty=10000
 \clubpenalty=10000  \clubpenalty=10000
   
Line 5713  should work if nowhere else does.}
Line 11288  should work if nowhere else does.}
   \fi    \fi
 }  }
   
 % Parameters in order: 1) textheight; 2) textwidth; 3) voffset;  % Parameters in order: 1) textheight; 2) textwidth;
 % 4) hoffset; 5) binding offset; 6) topskip.  Then whoever calls us can  % 3) voffset; 4) hoffset; 5) binding offset; 6) topskip;
 % set \parskip and call \setleading for \baselineskip.  % 7) physical page height; 8) physical page width.
 %  %
 \def\internalpagesizes#1#2#3#4#5#6{%  % We also call \setleading{\textleading}, so the caller should define
   % \textleading.  The caller should also set \parskip.
   %
   \def\internalpagesizes#1#2#3#4#5#6#7#8{%
   \voffset = #3\relax    \voffset = #3\relax
   \topskip = #6\relax    \topskip = #6\relax
   \splittopskip = \topskip    \splittopskip = \topskip
Line 5726  should work if nowhere else does.}
Line 11304  should work if nowhere else does.}
   \advance\vsize by \topskip    \advance\vsize by \topskip
   \outervsize = \vsize    \outervsize = \vsize
   \advance\outervsize by 2\topandbottommargin    \advance\outervsize by 2\topandbottommargin
   \pageheight = \vsize    \txipageheight = \vsize
   %    %
   \hsize = #2\relax    \hsize = #2\relax
   \outerhsize = \hsize    \outerhsize = \hsize
   \advance\outerhsize by 0.5in    \advance\outerhsize by 0.5in
   \pagewidth = \hsize    \txipagewidth = \hsize
   %    %
   \normaloffset = #4\relax    \normaloffset = #4\relax
   \bindingoffset = #5\relax    \bindingoffset = #5\relax
   %    %
     \ifpdf
       \pdfpageheight #7\relax
       \pdfpagewidth #8\relax
       % if we don't reset these, they will remain at "1 true in" of
       % whatever layout pdftex was dumped with.
       \pdfhorigin = 1 true in
       \pdfvorigin = 1 true in
     \else
       \ifx\XeTeXrevision\thisisundefined
       \else
         \pdfpageheight #7\relax
         \pdfpagewidth #8\relax
         % XeTeX does not have \pdfhorigin and \pdfvorigin.
       \fi
     \fi
     %
     \setleading{\textleading}
     %
   \parindent = \defaultparindent    \parindent = \defaultparindent
   \setemergencystretch    \setemergencystretch
 }  }
Line 5743  should work if nowhere else does.}
Line 11339  should work if nowhere else does.}
 % @letterpaper (the default).  % @letterpaper (the default).
 \def\letterpaper{{\globaldefs = 1  \def\letterpaper{{\globaldefs = 1
   \parskip = 3pt plus 2pt minus 1pt    \parskip = 3pt plus 2pt minus 1pt
   \setleading{13.2pt}%    \textleading = 13.2pt
   %    %
   % If page is nothing but text, make it come out even.    % If page is nothing but text, make it come out even.
   \internalpagesizes{46\baselineskip}{6in}{\voffset}{.25in}{\bindingoffset}{36pt}%    \internalpagesizes{607.2pt}{6in}% that's 46 lines
                       {\voffset}{.25in}%
                       {\bindingoffset}{36pt}%
                       {11in}{8.5in}%
 }}  }}
   
 % Use @smallbook to reset parameters for 7x9.5 (or so) format.  % Use @smallbook to reset parameters for 7x9.25 trim size.
 \def\smallbook{{\globaldefs = 1  \def\smallbook{{\globaldefs = 1
   \parskip = 2pt plus 1pt    \parskip = 2pt plus 1pt
   \setleading{12pt}%    \textleading = 12pt
   %    %
   \internalpagesizes{7.5in}{5.in}{\voffset}{.25in}{\bindingoffset}{16pt}%    \internalpagesizes{7.5in}{5in}%
                       {-.2in}{0in}%
                       {\bindingoffset}{16pt}%
                       {9.25in}{7in}%
   %    %
   \lispnarrowing = 0.3in    \lispnarrowing = 0.3in
   \tolerance = 700    \tolerance = 700
   \hfuzz = 1pt    \hfuzz = 1pt
   \contentsrightmargin = 0pt    \contentsrightmargin = 0pt
   \deftypemargin = 0pt  
   \defbodyindent = .5cm    \defbodyindent = .5cm
   }}
   
   % Use @smallerbook to reset parameters for 6x9 trim size.
   % (Just testing, parameters still in flux.)
   \def\smallerbook{{\globaldefs = 1
     \parskip = 1.5pt plus 1pt
     \textleading = 12pt
   %    %
   \let\smalldisplay = \smalldisplayx    \internalpagesizes{7.4in}{4.8in}%
   \let\smallexample = \smalllispx                      {-.2in}{-.4in}%
   \let\smallformat = \smallformatx                      {0pt}{14pt}%
   \let\smalllisp = \smalllispx                      {9in}{6in}%
     %
     \lispnarrowing = 0.25in
     \tolerance = 700
     \hfuzz = 1pt
     \contentsrightmargin = 0pt
     \defbodyindent = .4cm
 }}  }}
   
 % Use @afourpaper to print on European A4 paper.  % Use @afourpaper to print on European A4 paper.
 \def\afourpaper{{\globaldefs = 1  \def\afourpaper{{\globaldefs = 1
   \setleading{12pt}%  
   \parskip = 3pt plus 2pt minus 1pt    \parskip = 3pt plus 2pt minus 1pt
     \textleading = 13.2pt
   %    %
   \internalpagesizes{53\baselineskip}{160mm}{\voffset}{4mm}{\bindingoffset}{44pt}%    % Double-side printing via postscript on Laserjet 4050
     % prints double-sided nicely when \bindingoffset=10mm and \hoffset=-6mm.
     % To change the settings for a different printer or situation, adjust
     % \normaloffset until the front-side and back-side texts align.  Then
     % do the same for \bindingoffset.  You can set these for testing in
     % your texinfo source file like this:
     % @tex
     % \global\normaloffset = -6mm
     % \global\bindingoffset = 10mm
     % @end tex
     \internalpagesizes{673.2pt}{160mm}% that's 51 lines
                       {\voffset}{\hoffset}%
                       {\bindingoffset}{44pt}%
                       {297mm}{210mm}%
   %    %
   \tolerance = 700    \tolerance = 700
   \hfuzz = 1pt    \hfuzz = 1pt
     \contentsrightmargin = 0pt
     \defbodyindent = 5mm
 }}  }}
   
 % A specific text layout, 24x15cm overall, intended for A4 paper.  Top margin  % Use @afivepaper to print on European A5 paper.
 % 29mm, hence bottom margin 28mm, nominal side margin 3cm.  % From romildo@urano.iceb.ufop.br, 2 July 2000.
 \def\afourlatex{{\globaldefs = 1  % He also recommends making @example and @lisp be small.
   \setleading{13.6pt}%  \def\afivepaper{{\globaldefs = 1
     \parskip = 2pt plus 1pt minus 0.1pt
     \textleading = 12.5pt
   %    %
     \internalpagesizes{160mm}{120mm}%
                       {\voffset}{\hoffset}%
                       {\bindingoffset}{8pt}%
                       {210mm}{148mm}%
     %
     \lispnarrowing = 0.2in
     \tolerance = 800
     \hfuzz = 1.2pt
     \contentsrightmargin = 0pt
     \defbodyindent = 2mm
     \tableindent = 12mm
   }}
   
   % A specific text layout, 24x15cm overall, intended for A4 paper.
   \def\afourlatex{{\globaldefs = 1
   \afourpaper    \afourpaper
   \internalpagesizes{237mm}{150mm}{3.6mm}{3.6mm}{3mm}{7mm}%    \internalpagesizes{237mm}{150mm}%
                       {\voffset}{4.6mm}%
                       {\bindingoffset}{7mm}%
                       {297mm}{210mm}%
   %    %
     % Must explicitly reset to 0 because we call \afourpaper.
   \globaldefs = 0    \globaldefs = 0
 }}  }}
   
 % Use @afourwide to print on European A4 paper in wide format.  % Use @afourwide to print on A4 paper in landscape format.
 \def\afourwide{%  \def\afourwide{{\globaldefs = 1
   \afourpaper    \afourpaper
   \internalpagesizes{9.5in}{6.5in}{\hoffset}{\normaloffset}{\bindingoffset}{7mm}%    \internalpagesizes{241mm}{165mm}%
   %                      {\voffset}{-2.95mm}%
                       {\bindingoffset}{7mm}%
                       {297mm}{210mm}%
   \globaldefs = 0    \globaldefs = 0
 }  }}
   
 % @pagesizes TEXTHEIGHT[,TEXTWIDTH]  % @pagesizes TEXTHEIGHT[,TEXTWIDTH]
 % Perhaps we should allow setting the margins, \topskip, \parskip,  % Perhaps we should allow setting the margins, \topskip, \parskip,
 % and/or leading, also. Or perhaps we should compute them somehow.  % and/or leading, also. Or perhaps we should compute them somehow.
 %  %
 \def\pagesizes{\parsearg\pagesizesxxx}  \parseargdef\pagesizes{\pagesizesyyy #1,,\finish}
 \def\pagesizesxxx#1{\pagesizesyyy #1,,\finish}  
 \def\pagesizesyyy#1,#2,#3\finish{{%  \def\pagesizesyyy#1,#2,#3\finish{{%
   \setbox0 = \hbox{\ignorespaces #2}\ifdim\wd0 > 0pt \hsize=#2\relax \fi    \setbox0 = \hbox{\ignorespaces #2}\ifdim\wd0 > 0pt \hsize=#2\relax \fi
   \globaldefs = 1    \globaldefs = 1
   %    %
   \parskip = 3pt plus 2pt minus 1pt    \parskip = 3pt plus 2pt minus 1pt
   \setleading{13.2pt}%    \setleading{\textleading}%
   %    %
   \internalpagesizes{#1}{\hsize}{\voffset}{\normaloffset}{\bindingoffset}{44pt}%    \dimen0 = #1\relax
     \advance\dimen0 by \voffset
     %
     \dimen2 = \hsize
     \advance\dimen2 by \normaloffset
     %
     \internalpagesizes{#1}{\hsize}%
                       {\voffset}{\normaloffset}%
                       {\bindingoffset}{44pt}%
                       {\dimen0}{\dimen2}%
 }}  }}
   
 % Set default to letter.  % Set default to letter.
Line 5822  should work if nowhere else does.}
Line 11482  should work if nowhere else does.}
   
 \message{and turning on texinfo input format.}  \message{and turning on texinfo input format.}
   
   \def^^L{\par} % remove \outer, so ^L can appear in an @comment
   
   % DEL is a comment character, in case @c does not suffice.
   \catcode`\^^? = 14
   
 % Define macros to output various characters with catcode for normal text.  % Define macros to output various characters with catcode for normal text.
 \catcode`\"=\other  \catcode`\"=\other \def\normaldoublequote{"}
 \catcode`\~=\other  \catcode`\$=\other \def\normaldollar{$}%$ font-lock fix
 \catcode`\^=\other  \catcode`\+=\other \def\normalplus{+}
 \catcode`\_=\other  \catcode`\<=\other \def\normalless{<}
 \catcode`\|=\other  \catcode`\>=\other \def\normalgreater{>}
 \catcode`\<=\other  \catcode`\^=\other \def\normalcaret{^}
 \catcode`\>=\other  \catcode`\_=\other \def\normalunderscore{_}
 \catcode`\+=\other  \catcode`\|=\other \def\normalverticalbar{|}
 \catcode`\$=\other  \catcode`\~=\other \def\normaltilde{~}
 \def\normaldoublequote{"}  
 \def\normaltilde{~}  
 \def\normalcaret{^}  
 \def\normalunderscore{_}  
 \def\normalverticalbar{|}  
 \def\normalless{<}  
 \def\normalgreater{>}  
 \def\normalplus{+}  
 \def\normaldollar{$}  
   
 % This macro is used to make a character print one way in ttfont  % This macro is used to make a character print one way in \tt
 % where it can probably just be output, and another way in other fonts,  % (where it can probably be output as-is), and another way in other fonts,
 % where something hairier probably needs to be done.  % where something hairier probably needs to be done.
 %  %
 % #1 is what to print if we are indeed using \tt; #2 is what to print  % #1 is what to print if we are indeed using \tt; #2 is what to print
Line 5859  should work if nowhere else does.}
Line 11515  should work if nowhere else does.}
 % this is not a problem.  % this is not a problem.
 \def\ifusingit#1#2{\ifdim \fontdimen1\font>0pt #1\else #2\fi}  \def\ifusingit#1#2{\ifdim \fontdimen1\font>0pt #1\else #2\fi}
   
 % Turn off all special characters except @  % Set catcodes for Texinfo file
 % (and those which the user can use as if they were ordinary).  
   % Active characters for printing the wanted glyph.
 % Most of these we simply print from the \tt font, but for some, we can  % Most of these we simply print from the \tt font, but for some, we can
 % use math or other variants that look better in normal text.  % use math or other variants that look better in normal text.
   %
 \catcode`\"=\active  \catcode`\"=\active
 \def\activedoublequote{{\tt\char34}}  \def\activedoublequote{{\tt\char34}}
 \let"=\activedoublequote  \let"=\activedoublequote
 \catcode`\~=\active  \catcode`\~=\active \def\activetilde{{\tt\char126}} \let~ = \activetilde
 \def~{{\tt\char126}}  \chardef\hatchar=`\^
 \chardef\hat=`\^  \catcode`\^=\active \def\activehat{{\tt \hatchar}} \let^ = \activehat
 \catcode`\^=\active  
 \def^{{\tt \hat}}  
   
 \catcode`\_=\active  \catcode`\_=\active
 \def_{\ifusingtt\normalunderscore\_}  \def_{\ifusingtt\normalunderscore\_}
 % Subroutine for the previous macro.  \def\_{\leavevmode \kern.07em \vbox{\hrule width.3em height.1ex}\kern .07em }
 \def\_{\leavevmode \kern.06em \vbox{\hrule width.3em height.1ex}}  \let\realunder=_
   
 \catcode`\|=\active  \catcode`\|=\active \def|{{\tt\char124}}
 \def|{{\tt\char124}}  
 \chardef \less=`\<  \chardef \less=`\<
 \catcode`\<=\active  \catcode`\<=\active \def\activeless{{\tt \less}}\let< = \activeless
 \def<{{\tt \less}}  
 \chardef \gtr=`\>  \chardef \gtr=`\>
 \catcode`\>=\active  \catcode`\>=\active \def\activegtr{{\tt \gtr}}\let> = \activegtr
 \def>{{\tt \gtr}}  \catcode`\+=\active \def+{{\tt \char 43}}
 \catcode`\+=\active  \catcode`\$=\active \def${\ifusingit{{\sl\$}}\normaldollar}%$ font-lock fix
 \def+{{\tt \char 43}}  \catcode`\-=\active \let-=\normaldash
 \catcode`\$=\active  
 \def${\ifusingit{{\sl\$}}\normaldollar}  
 %\catcode 27=\active  
 %\def^^[{$\diamondsuit$}  
   
 % Set up an active definition for =, but don't enable it most of the time.  
 {\catcode`\==\active  
 \global\def={{\tt \char 61}}}  
   
 \catcode`+=\active  % used for headline/footline in the output routine, in case the page
 \catcode`\_=\active  % breaks in the middle of an @tex block.
   \def\texinfochars{%
     \let< = \activeless
     \let> = \activegtr
     \let~ = \activetilde
     \let^ = \activehat
     \markupsetuplqdefault \markupsetuprqdefault
     \let\b = \strong
     \let\i = \smartitalic
     % in principle, all other definitions in \tex have to be undone too.
   }
   
 % If a .fmt file is being used, characters that might appear in a file  % Used sometimes to turn off (effectively) the active characters even after
 % name cannot be active until we have parsed the command line.  % parsing them.
 % So turn them off again, and have \everyjob (or @setfilename) turn them on.  \def\turnoffactive{%
 % \otherifyactive is called near the end of this file.    \normalturnoffactive
 \def\otherifyactive{\catcode`+=\other \catcode`\_=\other}    \otherbackslash
   }
   
 \catcode`\@=0  \catcode`\@=0
   
 % \rawbackslashxx output one backslash character in current font  % \backslashcurfont outputs one backslash character in current font,
 \global\chardef\rawbackslashxx=`\\  % as in \char`\\.
 %{\catcode`\\=\other  \global\chardef\backslashcurfont=`\\
 %@gdef@rawbackslashxx{\}}  \global\let\rawbackslashxx=\backslashcurfont  % let existing .??s files work
   
 % \rawbackslash redefines \ as input to do \rawbackslashxx.  % \realbackslash is an actual character `\' with catcode other, and
 {\catcode`\\=\active  % \doublebackslash is two of them (for the pdf outlines).
 @gdef@rawbackslash{@let\=@rawbackslashxx }}  {\catcode`\\=\other @gdef@realbackslash{\} @gdef@doublebackslash{\\}}
   
 % \normalbackslash outputs one backslash in fixed width font.  % In Texinfo, backslash is an active character; it prints the backslash
 \def\normalbackslash{{\tt\rawbackslashxx}}  % in fixed width font.
   \catcode`\\=\active  % @ for escape char from now on.
   
 % \catcode 17=0   % Define control-q  % Print a typewriter backslash.  For math mode, we can't simply use
 \catcode`\\=\active  % \backslashcurfont: the story here is that in math mode, the \char
   % of \backslashcurfont ends up printing the roman \ from the math symbol
   % font (because \char in math mode uses the \mathcode, and plain.tex
   % sets \mathcode`\\="026E).  Hence we use an explicit \mathchar,
   % which is the decimal equivalent of "715c (class 7, e.g., use \fam;
   % ignored family value; char position "5C).  We can't use " for the
   % usual hex value because it has already been made active.
   
 % Used sometimes to turn off (effectively) the active characters  @def@ttbackslash{{@tt @ifmmode @mathchar29020 @else @backslashcurfont @fi}}
 % even after parsing them.  @let@backslashchar = @ttbackslash % @backslashchar{} is for user documents.
 @def@turnoffactive{@let"=@normaldoublequote  
 @let\=@realbackslash  
 @let~=@normaltilde  
 @let^=@normalcaret  
 @let_=@normalunderscore  
 @let|=@normalverticalbar  
 @let<=@normalless  
 @let>=@normalgreater  
 @let+=@normalplus  
 @let$=@normaldollar}  
   
 @def@normalturnoffactive{@let"=@normaldoublequote  % \rawbackslash defines an active \ to do \backslashcurfont.
 @let\=@normalbackslash  % \otherbackslash defines an active \ to be a literal `\' character with
 @let~=@normaltilde  % catcode other.  We switch back and forth between these.
 @let^=@normalcaret  @gdef@rawbackslash{@let\=@backslashcurfont}
 @let_=@normalunderscore  @gdef@otherbackslash{@let\=@realbackslash}
 @let|=@normalverticalbar  
 @let<=@normalless  
 @let>=@normalgreater  
 @let+=@normalplus  
 @let$=@normaldollar}  
   
 % Make _ and + \other characters, temporarily.  % Same as @turnoffactive except outputs \ as {\tt\char`\\} instead of
 % This is canceled by @fixbackslash.  % the literal character `\'.
 @otherifyactive  %
   {@catcode`- = @active
    @gdef@normalturnoffactive{%
      @passthroughcharstrue
      @let-=@normaldash
      @let"=@normaldoublequote
      @let$=@normaldollar %$ font-lock fix
      @let+=@normalplus
      @let<=@normalless
      @let>=@normalgreater
      @let^=@normalcaret
      @let_=@normalunderscore
      @let|=@normalverticalbar
      @let~=@normaltilde
      @let\=@ttbackslash
      @markupsetuplqdefault
      @markupsetuprqdefault
      @unsepspaces
    }
   }
   
   % If a .fmt file is being used, characters that might appear in a file
   % name cannot be active until we have parsed the command line.
   % So turn them off again, and have @fixbackslash turn them back on.
   @catcode`+=@other @catcode`@_=@other
   
   % \enablebackslashhack - allow file to begin `\input texinfo'
   %
 % If a .fmt file is being used, we don't want the `\input texinfo' to show up.  % If a .fmt file is being used, we don't want the `\input texinfo' to show up.
 % That is what \eatinput is for; after that, the `\' should revert to printing  % That is what \eatinput is for; after that, the `\' should revert to printing
 % a backslash.  % a backslash.
 %  % If the file did not have a `\input texinfo', then it is turned off after
 @gdef@eatinput input texinfo{@fixbackslash}  % the first line; otherwise the first `\' in the file would cause an error.
 @global@let\ = @eatinput  % This is used on the very last line of this file, texinfo.tex.
   % We also use @c to call @fixbackslash, in case ends of lines are hidden.
   {
   @catcode`@^=7
   @catcode`@^^M=13@gdef@enablebackslashhack{%
     @global@let\ = @eatinput%
     @catcode`@^^M=13%
     @def@c{@fixbackslash@c}%
     % Definition for the newline at the end of this file.
     @def ^^M{@let^^M@secondlinenl}%
     % Definition for a newline in the main Texinfo file.
     @gdef @secondlinenl{@fixbackslash}%
   }}
   
 % On the other hand, perhaps the file did not have a `\input texinfo'. Then  {@catcode`@^=7 @catcode`@^^M=13%
 % the first `\{ in the file would cause an error. This macro tries to fix  @gdef@eatinput input texinfo#1^^M{@fixbackslash}}
 % that, assuming it is called before the first `\' could plausibly occur.  
 % Also back turn on active characters that might appear in the input  % Emergency active definition of newline, in case an active newline token
 % file name, in case not using a pre-dumped format.  % appears by mistake.
 %  {@catcode`@^=7 @catcode13=13%
   @gdef@enableemergencynewline{%
     @gdef^^M{%
       @par%
       %<warning: active newline>@par%
   }}}
   
   
 @gdef@fixbackslash{%  @gdef@fixbackslash{%
   @ifx\@eatinput @let\ = @normalbackslash @fi    @ifx\@eatinput @let\ = @ttbackslash @fi
     @catcode13=5 % regular end of line
     @enableemergencynewline
     @let@c=@texinfoc
     % Also turn back on active characters that might appear in the input
     % file name, in case not using a pre-dumped format.
   @catcode`+=@active    @catcode`+=@active
   @catcode`@_=@active    @catcode`@_=@active
     %
     % If texinfo.cnf is present on the system, read it.
     % Useful for site-wide @afourpaper, etc.  This macro, @fixbackslash, gets
     % called at the beginning of every Texinfo file.  Not opening texinfo.cnf
     % directly in this file, texinfo.tex, makes it possible to make a format
     % file for Texinfo.
     %
     @openin 1 texinfo.cnf
     @ifeof 1 @else @input texinfo.cnf @fi
     @closein 1
 }  }
   
   
 % Say @foo, not \foo, in error messages.  % Say @foo, not \foo, in error messages.
 @escapechar = `@@  @escapechar = `@@
   
 % These look ok in all fonts, so just make them not special.  % These (along with & and #) are made active for url-breaking, so need
 @catcode`@& = @other  % active definitions as the normal characters.
 @catcode`@# = @other  @def@normaldot{.}
 @catcode`@% = @other  @def@normalquest{?}
   @def@normalslash{/}
   
 @c Set initial fonts.  % These look ok in all fonts, so just make them not special.
 @textfonts  % @hashchar{} gets its own user-level command, because of #line.
 @rm  @catcode`@& = @other @def@normalamp{&}
   @catcode`@# = @other @def@normalhash{#}
   @catcode`@% = @other @def@normalpercent{%}
   
   @let @hashchar = @normalhash
   
   @c Finally, make ` and ' active, so that txicodequoteundirected and
   @c txicodequotebacktick work right in, e.g., @w{@code{`foo'}}.  If we
   @c don't make ` and ' active, @code will not get them as active chars.
   @c Do this last of all since we use ` in the previous @catcode assignments.
   @catcode`@'=@active
   @catcode`@`=@active
   @markupsetuplqdefault
   @markupsetuprqdefault
   
 @c Local variables:  @c Local variables:
 @c eval: (add-hook 'write-file-hooks 'time-stamp)  @c eval: (add-hook 'write-file-hooks 'time-stamp)
 @c page-delimiter: "^\\\\message"  @c page-delimiter: "^\\\\message\\|emacs-page"
 @c time-stamp-start: "def\\\\texinfoversion{"  @c time-stamp-start: "def\\\\texinfoversion{"
 @c time-stamp-format: "%:y-%02m-%02d.%02H"  @c time-stamp-format: "%:y-%02m-%02d.%02H"
 @c time-stamp-end: "}"  @c time-stamp-end: "}"
 @c End:  @c End:
   
   @c vim:sw=2:
   
   @ignore
      arch-tag: e1b36e32-c96e-4135-a41a-0b2efa2ea115
   @end ignore
   @enablebackslashhack
Legend:
Removed from v.1.1  
changed lines
  Added in v.1.3
FreeBSD-CVSweb <freebsd-cvsweb@FreeBSD.org>