% This is part of the book TeX for the Impatient.
% Copyright (C) 2003, 2014 Paul W. Abrahams, Kathryn A. Hargreaves, Karl Berry.
% See file fdl.tex for copying conditions.

\input macros 
\chapter{A compendium \linebreak of useful macros}

\chapterdef{eplain}

This section describes |eplain.tex|, a collection of macros and other
definitions that extend  plain \TeX.
\bix^^|eplain.tex|
The descriptions of the various macros explain their purposes,
but usually do not explain how they
work or provide explicit details on
how to use them.  That information is contained in the source files for
|eplain.tex| and in the documentation that comes with it.
See \headcit{Resources}{resources} for how to obtain |eplain.tex|.


\section{Preliminaries}

We start with some macros for changing category codes and convenient
definitions for two of the commonly used ones.
^^{category codes//useful definitions for}

\pix^^|\makeactive|
\pix^^|\letter|
\pix^^|\other|
\pix^^|\uncatcodespecials|

\csdisplay
\def\makeactive#1{\catcode`#1 = \active \ignorespaces}%
\chardef\letter = 11 \chardef\other = 12
\def\uncatcodespecials{%
   \def\do##1{\catcode`##1 = \other}%
   \dospecials}% Defined in plain.
|

In order to define `|^^M|' as an active character, you need to encase
the definition in a group and invoke some extra machinery.
\ttidxref{^^M}
The \pix^|\letreturn| macro lets you define `|^^M|' without that
extra machinery (which you can see in the definition below).

\csdisplay
{\makeactive\^^M \long\gdef\letreturn#1{\let^^M = #1}}%
|

These macros consume one, two, or three arguments.

\pix^^|\gobble|
\pix^^|\gobbletwo|
\pix^^|\gobblethree|
\csdisplay
\def\gobble#1{}\def\gobbletwo#1#2{}%
\def\gobblethree#1#2#3{}%
|

Now we establish some conventions for reading the rest of the file.
Within the file we allow ``private'' control sequences that contain
`|@|' in their names.
These control sequences aren't accessible outside of this file (unless
you change the category code of `|@|' again).
\xrdef{eplainconv}

\csdisplay
\catcode`@ = \letter    % Allow control sequences with @.
\let\@plainwlog = \wlog % Don't log register allocations.
\let\wlog = \gobble
\newlinechar = `^^J
|

The next two macros provide convenient forms of diagnostic output.
\pix^|\loggingall|  turns on all tracing, but causes the trace
output to appear only in the log file and not at your terminal.
\pix^|\tracingboxes| causes boxes to be displayed completely when
they're traced.  (\TeX\ normally shows only three levels of boxing
and five items within each box.)


\csdisplay
\def\loggingall{\tracingcommands\tw@\tracingstats\tw@
   \tracingpages\@ne\tracingoutput\@ne
   \tracinglostchars\@ne\tracingmacros\tw@
   \tracingparagraphs\@ne\tracingrestores\@ne
   \showboxbreadth\maxdimen\showboxdepth\maxdimen}%
\def\tracingboxes{\showboxbreadth = \maxdimen
   \showboxdepth = \maxdimen}%
|


The default thickness of rules is $0.4$\pt.
You can produce rules of any default thickness you choose by redefining
|\vruledefaultwidth|, |\hruledefaultheight|, and |\hruledefaultdepth|
and then using
\pix^^|\ehrule|
|\eh!-rule|
and \pix^|\evrule|
instead of |\hrule| and |\vrule|. (The `e'  stands for
``eplain''.)
^^{rules//thickness of}
If you give an explicit
dimension (e.g., |\ehrule height 16pt|), \TeX\ will use it.

\csdisplay
\newdimen\hruledefaultheight  \hruledefaultheight = 0.4pt
\newdimen\hruledefaultdepth   \hruledefaultdepth = 0.0pt
\newdimen\vruledefaultwidth   \vruledefaultwidth = 0.4pt
\def\ehrule{\hrule height\hruledefaultheight
   depth\hruledefaultdepth}%
\def\evrule{\vrule width\vruledefaultwidth}%
|

The |\%| convention for writing a `|%|' character doesn't work when you 
want to include that character in the token list of |\write|.
^^|\write//writing \b\tt\%\e\ with|
^^|%//writing out|
You can use ^|\percentchar| to achieve this.
We also redefine ^|^^L| to be nonouter so that you can use it in a
macro definition or argument.

\csdisplay
{\catcode`\% = \other \gdef\percentchar{%}}%
 \def^^L{\par
}%
|

\pix^|\tokstostring| converts its argument into a list of ^{character
tokens}.  
It uses only expansions that are handled in \TeX's gullet.
This property is necessary
for it to work with |\edef|.  It is used by the cross-referencing
macros (\xref{xrefs}).

In order to split the argument up at spaces, we have to use two
subsidiary macros. |\@ttsA| finds the spaces, and |\@ttsB| handles a
token sequence without any spaces.  Each space is replaced by the
expansion of \pix^|\spacesub|.

\csdisplay
\def\tokstostring#1{\@ttsA#1 \ttsmarkA}%
\def\@ttsA#1 #2\ttsmarkA{\ifempty{#1}\else
      \@ttsB #1\@ttsmarkB
      \ifempty{#2}\else
         \spacesub\@ttsA#2\ttsmarkA\fi\fi}%
\def\@ttsB#1{\ifx #1\@ttsmarkB\else
      \string #1%
      \expandafter\@ttsB\fi}%
\def\@ttsmarkB{\@ttsmarkB}% should never be expanded
\def\spacesub{+}%
|

\noindent
\pix^|\ifempty| tests if its argument is empty.

\csdisplay
\def\ifempty#1{\@ifempty #1\@emptymarkA\@emptymarkB}%
\def\@ifempty#1#2\@emptymarkB{\ifx #1\@emptymarkA}%
\def\@emptymarkA{\@emptymarkA}%
|

The ^|\for|  macro implements a \TeX\ version of the ^{for loop} in
traditional programming languages.  These macros come directly from
\LaTeX.

\csdisplay
\def\for#1:=#2\do#3{\edef\@fortmp{#2}%
   \ifx\@fortmp\empty \else
      \expandafter\@forloop#2,\@nil,\@nil\@@#1{#3}\fi}%
\def\@nnil{\@nil}%
\def\@fornoop#1\@@#2#3{}%
\def\@forloop#1,#2,#3\@@#4#5{\def#4{#1}\ifx #4\@nnil
   \else #5\def#4{#2} ifx #4\@nnil \else
      #5\@iforloop #3\@@#4{#5}\fi\fi}%
\def\@iforloop#1,#2\@@#3#4{\def#3{#1}\ifx #3\@nnil 
   \let\@nextwhile=\@fornoop \else #4\relax
      \let\@nextwhile=\@iforloop\fi
         \@nextwhile#2\@@#3{#4}}%
|

\pix^|\obeywhitespace| is useful for reproducing line breaks, blank lines,
and spaces in your input. It combines the effects of |\obey!-lines| and
|\obey!-spaces|,
^^|\obeylines| ^^|\obeyspaces|
and also causes spaces at the start of a line to
be printed. Tab characters are not affected by this; they still produce
normal glue.
\xrdef{ewhitesp}
^^{space characters}
^^|\fontdimen|
^^|\font|
^^|\letreturn|

\csdisplay
\def\alwaysspace{\hglue\fontdimen2\the\font \relax}%
{\makeactive\^^M \makeactive\ %
\gdef\obeywhitespace{%
\makeactive\^^M\def^^M{\par\indent}%
\aftergroup\@removebox% Kill extra paragraph at end.
\makeactive\ \let =\alwaysspace}}%
\def\@removebox{\setbox0=\lastbox}
|

\pix^|\frac| is a good way to print fractions in text when you don't
want to use |\over| and
``1/2'' just doesn't look right.  This macro is the answer to
\knuth{Exercise~11.6}.^^{fractions//slashed form}

\csdisplay
\def\frac#1/#2{\leavevmode
   \kern.1em \raise .5ex \hbox{\the\scriptfont0 #1}%
   \kern-.1em $/$%
   \kern-.15em \lower .25ex \hbox{\the\scriptfont0 #2}}%
|

The following macros produce logos that are useful in the \TeX\ world.
The \AMSTeX\ logo is from \knuth{page~420}.  The \LaTeX\ logo is
slightly modified from the one in |latex.tex| (we use a different font
for the `A'); similarly, the \BibTeX\ logo uses |\sevenrm| instead of a
true caps-and-small-caps font. The |.mf| source file for the \MF\ logo
is given in the \MF\ manual:
\smallskip
{\narrower\noindent
^{Knuth, Donald E.},{\sl The {\logosl METAFONT}book}.
  Reading, Mass.: Addison-Wesley, 1986.\par}
\smallskip
\pix^^|\LaTeX|
\pix^^|\AMSTeX|
\pix^^|\BibTeX|
\pix^^|\MF|

\csdisplay
\def\LaTeX{L\kern-.26em \raise.6ex\hbox{\fiverm A}%
   \kern-.15em TeX}%
\def\AMSTeX{$\cal A\kern-.1667em \lower.5ex\hbox{$\cal M$}%
   \kern-.125em S$-\TeX}%
\def\BibTeX{{\rm B\kern-.05em {\sevenrm I\kern-.025em B}%
   \kern-.08em T\kern-.1667em \lower.7ex\hbox{E}%
   \kern-.125emX}}%
\font\mflogo = logo10
\def\MF{{\mflogo META}{\tenrm \-}{\mflogo FONT}}%
|

The next two macros produce boxes.  \pix^|\blackbox| produces a ``square
bullet'', used in the list macros (\xref{listmacs}).
\pix^|\makeblankbox| (from \knuth{page~311}) produces an unfilled
rectangle, with the thickness of the border rules given by the
arguments.

\csdisplay
\def\blackbox{\vrule height .8ex width .6ex depth -.2ex}%
\def\makeblankbox#1#2{%
   \hbox{\lower\dp0\vbox{\hidehrule{#1}{#2}%
      \kern -#1% overlap rules
      \hbox to \wd0{\hidevrule{#1}{#2}%
         \raise\ht0\vbox to #1{}% vrule height
         \lower\dp0\vtop to #1{}% vrule depth
         \hfil\hidevrule{#2}{#1}}%
      \kern-#1\hidehrule{#2}{#1}}}}%
\def\hidehrule#1#2{\kern-#1\hrule height#1 depth#2
   \kern-#2}%
\def\hidevrule#1#2{\kern-#1{\dimen0 = #1
   \advance\dimen0 by #2 \vrule width\dimen0}\kern-#2}%
|

\pix^|\numbername| produces the written-out form of a number.  (If the
number is greater than ten, the macro just reproduces the numerals
of its argument.)

\csdisplay
\def\numbername#1{\ifcase#1%
   zero\or one\or two\or three\or four\or five%
   \or six\or seven\or eight\or nine\or ten\or #1\fi}%
|

\pix^|\testfileexistence| determines whether a file |\jobname.#1| is
non\-empty and sets |\if!-fileexists| appropriately.
^^{files//testing for existence of}
The file name in the argument need not end in a space token since
the macro provides the space token.

\csdisplay
\newif\iffileexists
\def\testfileexistence#1{\begingroup
      \immediate\openin0 = \jobname.#1\space
      \ifeof 0\global\fileexistsfalse
      \else \global\fileexiststrue\fi
      \immediate\closein0
   \endgroup}%
|


\section Displays

By default, \TeX\ centers displayed material (the material between
|$$|'s).  ^^{displays, formatting} \pix^|\leftdisplays| causes displays
to be left-justified by default.  You can return to centered displays
with \pix^|\centereddisplays|.

The macros here are more general than they need to be just for doing
left-justified displays.  For every display,
\pix^|\ifeqno| will be true if an |\eqno| occurred in the display.
\pix^|\ifleqno| will be true if an |\leqno| occurred.  If either
kind of equation number occurred, \pix^|\eqn| produces the text of the
equation number.  \pix^|\eq|  always produces the text of the
equation itself.

These macros are based on the code on \knuth{page~376}.

\csdisplay
\newif\ifeqno \newif\ifleqno
\newtoks\@eqtoks \newtoks\@eqnotoks
\def\eq{\the\@eqtoks}\def\eqn{\the\@eqnotoks}%
\def\displaysetup#1$${%
   \@displaytest#1\eqno\eqno\@displaytest}%
\def\@displaytest#1\eqno#2\eqno#3\@displaytest{%
   \if !#3!% No \eqno, check for \leqno:
      \@ldisplaytest#1\leqno\leqno\@ldisplaytest
   \else
      \eqnotrue \leqnofalse % Have \eqno, not \leqno.
      \@eqnotoks = {#2}\@eqtoks = {#1}%
   \fi
   \generaldisplay$$}%
\def\@ldisplaytest#1\leqno#2\leqno#3\@ldisplaytest{%
   \@eqtoks = {#1}%
   \if !#3!%
      \eqnofalse % No \leqno; we're done.
   \else
      \eqnotrue \leqnotrue % Have \leqno.
      \@eqnotoks = {#2}%
   \fi}%
|

You can format displays differently by defining your own macro,
analogous to \pix^|\leftdisplays|.
The macro definition must 
place a call on |\display!-setup| in |\every!-display| ^^|\everydisplay|
so as to ensure
that |\dis!-play!-setup| ^^|\displaysetup| is called at the start of
every display. The macro definition must also include a definition of
^|\generaldisplay|.

\csdisplay
\newtoks\previouseverydisplay
\def\leftdisplays{%
   \previouseverydisplay = \everydisplay
   \everydisplay = 
     {\the\previouseverydisplay \displaysetup}%
   \def\generaldisplay{%
      \leftline{%
         \strut \indent \hskip\leftskip
         \dimen0 = \parindent
         \advance\dimen0 by \leftskip
         \advance\displaywidth by -\dimen0
         \@redefinealignmentdisplays
         \ifeqno \ifleqno
            \kern-\dimen0
            \rlap{$\displaystyle\eqn$}%
            \kern\dimen0
         \fi\fi
         $\displaystyle{\eq}$%
         \ifeqno \ifleqno\else
            \hfill $\displaystyle{\eqn}$%
         \fi\fi}}}%
\def\centereddisplays{\let\displaysetup = \relax}%
|

\filbreak
\pix^|\leftdisplays| must go to some pains to make sure that
|\dis!-play!-lines|, ^^|\displaylines|
^|\eqalignno|,  and ^|\leqalignno| still work properly.
|\eq| is typeset in math mode,
and |\halign| is illegal in math mode.
^^|\halign//illegal in math mode|
We use ^|\vcenter| to change the context so that  |\halign| becomes
legal again. We also remove the |\hfil| commands at the left of the template
to obtain the flush left formatting. Other than those changes, the macros
are the same as in |plain.tex|.

\csdisplay
\def\@redefinealignmentdisplays{%
   \def\displaylines##1{\displ@y
      \vcenter{\halign{\hbox to\displaywidth{$\@lign
            \displaystyle####\hfil$}\crcr##1\crcr}}}%
   \def\eqalignno##1{\displ@y
      \vcenter{\halign to\displaywidth{%
             $\@lign\displaystyle{####}$\tabskip\z@skip
            &$\@lign\displaystyle{{}####}$
               \hfil\tabskip\centering
            &\llap{$\@lign####$}\tabskip\z@skip\crcr
            ##1\crcr}}}%
   \def\leqalignno##1{\displ@y
      \vcenter{\halign to\displaywidth{%
             $\@lign\displaystyle{####}$\tabskip\z@skip
            &$\@lign\displaystyle{{}####}
               $\hfil\tabskip\centering
            &\kern-\displaywidth
             \rlap{\kern-\parindent\kern-\leftskip$
                \@lign####$}%
             \tabskip\displaywidth\crcr
            ##1\crcr}}}}%
|


\section Time of day

When \TeX\ starts up, it sets the values of the
^|\time|, ^|\day|, ^|\month|, and ^|\year| parameters.
^^{time of day}^^{date}
\pix^|\monthname|
produces the name of the month, abbreviated to three letters.
\pix^|\timestring| produces the current time, as in ``1:14\thinspace
p.m.''. \pix^|\timestamp| produces the text of the complete date, as
in ``23 Apr 1964\quad 1:14\thinspace p.m.''.

\csdisplay
\def\monthname{%
   \ifcase\month
      \or Jan\or Feb\or Mar\or Apr\or May\or Jun%
      \or Jul\or Aug\or Sep\or Oct\or Nov\or Dec%
   \fi}%
\def\timestring{\begingroup
   \count0 = \time \divide\count0 by 60
   \count2 = \count0   % The hour.
   \count4 = \time \multiply\count0 by 60
   \advance\count4 by -\count0   % The minute.
   \ifnum\count4<10 \toks1 = {0}% Get a leading zero.
   \else            \toks1 = {}%
   \fi
   \ifnum\count2<12 \toks0 = {a.m.}%
   \else            \toks0 = {p.m.}%
      \advance\count2 by -12
   \fi
   \ifnum\count2=0 \count2 = 12 \fi % Make midnight `12'.
   \number\count2:\the\toks1 \number\count4
   \thinspace \the\toks0
\endgroup}%
\def\timestamp{\number\day\space\monthname\space
   \number\year\quad\timestring}%
|


\section Lists

\null
\xrdef{listmacs}
\bix^^{itemized lists}
\bix^^{enumerations}
\pix^|\numberedlist| produces numbered lists; |\endnumberedlist| ends
them.
\pix^|\unorderedlist| is analogous.
For either of these, items inside the lists begin
with \pix^|\li| (``list item'').  You can put \pix^|\listcompact|
at the beginning of a list if
you don't want any additional space between the items of that list. 
Lists can be nested arbitrarily.

You can control the spacing between the items more generally by
assigning values to the registers listed below.  If the items in
your lists tend to be long, you might want to make |\interitemskip|
nonzero.
The left indentation of each list item is given by |\par!-indent| plus
|\list!-left!-indent|; the right indentation of each list item
is given by |\list!-right!-indent|.

\csdisplay
\newskip\abovelistskip  \abovelistskip = .5\baselineskip 
\newskip\interitemskip  \interitemskip = 0pt
\newskip\belowlistskip  \belowlistskip = .5\baselineskip
\newdimen\listleftindent  \listleftindent = \parindent   
\newdimen\listrightindent \listrightindent = 0pt        
\def\listcompact{\interitemskip = 0pt \relax}%
|

Both numbered and unnumbered lists use the macros that follow.
We don't change |\parindent|, since many existing macros, e.g., 
|\footnote|, depend on |\parindent|.
We must account for the
possibility that items are more than one paragraph long. In this case, all
paragraphs after the first will be indented.
We use 
|\leftskip| and |\rightskip| to indent the list items.
Indentation of displays is accounted for by changes to |\every!-dis!-play|.
^^|\everydisplay|

\csdisplay
\newdimen\@listindent
\def\beginlist{%
   \@listindent = \parindent
   \advance\@listindent by \listleftindent
   \everydisplay = \expandafter{\the\everydisplay
      % Don't lose user's \everydisplay:
      \advance\displayindent by \@listindent
      \advance\displaywidth by -\@listindent
      \advance\displaywidth by -\listrightindent}%
   \nobreak\vskip\abovelistskip
   \parskip = 0pt
   % \leftskip shifts nested lists to the right on the page.
   \advance\leftskip by \@listindent
   \advance\rightskip by \listrightindent}%
\def\printitem{\par\noindent
   \llap{\hskip-\listleftindent \marker \enspace}}%
\def\endlist{\vskip\belowlistskip}%
|

\noindent
You can change the way the item labels are typeset by
redefining the |\numbered!-marker| macro.
\pix^^|\numberedmarker|

\csdisplay
\newcount\numberedlistdepth \newcount\itemnumber
\newcount\itemletter
\def\numberedmarker{%
   \ifcase\numberedlistdepth
       (impossible)%
   \or \itemnumberout)%
   \or \itemletterout)%
   \else *%
   \fi}%
|

\noindent Here are the definitions of |\numberedlist| and
|\unorderedlist|.
Both definitions have the same structure.

\csdisplay
\def\numberedlist{\environment{@numbered-list}%
   \advance\numberedlistdepth by 1
   \itemnumber = 1 \itemletter = `a
   \beginlist \let\marker = \numberedmarker
   \def\li{%
      \ifnum\itemnumber=1\else \vskip\interitemskip \fi
      \printitem
      \advance\itemnumber by 1 \advance\itemletter by 1
   }}%
\def\itemnumberout{\number\itemnumber}%
\def\itemletterout{\char\itemletter}%
\def\endnumberedlist{\par
   \endenvironment{@numbered-list}\endlist}%
!bigskip
\newcount\unorderedlistdepth
\def\unorderedmarker{%
   \ifcase\unorderedlistdepth
       (impossible)%
   \or \blackbox
   \or ---%
   \else *%
   \fi}%
\def\unorderedlist{\environment{@unordered-list}%
   \advance\unorderedlistdepth by 1
   \beginlist \itemnumber = 1
   \let\marker = \unorderedmarker
   \def\li{%
      \ifnum\itemnumber=1\else \vskip\interitemskip \fi
      \printitem \advance\itemnumber by 1
   }}%
\def\endunorderedlist{\par
   \endenvironment{@unordered-list}\endlist}%
|

\eix^^{itemized lists}
\eix^^{enumerations}


\section Verbatim listing

The \pix^|\listing| macro produces a verbatim listing of a specified
file in the |\tt| font.
It is based on the code on \knuth{page~380}.
Tabs produce a fixed amount of space, and form
feeds produce a page break.  Other control characters produce whatever
happens to be at that font position, which is generally not very useful.
By redefining |\setup!-listing!-hook|,
\pix^^|\setuplistinghook|
you can take additional actions that are appropriate for your particular
fonts and\slash or environment before the file is read in.

\csdisplay
\def\listing#1{%
   \par \begingroup \@setuplisting \setuplistinghook
   \input #1 \endgroup}%
\let\setuplistinghook = \empty
\def\@setuplisting{%
   \uncatcodespecials
   \obeywhitespace \makeactive\` \makeactive\^^I
   \def^^L{\vfill\eject}\tt}%
{\makeactive\` \gdef`{\relax\lq}}% Defeat ligatures.
{\makeactive\^^I\gdef^^I{\hskip8\fontdimen2\tt \relax}}%
|


\section Tables of contents

\null ^^{table of contents}
The macro \pix^|\writetocentry| writes a macro call to the file
|\jobname.toc|.  The first argument of |\writetocentry|, e.g.,
``chapter'', is used to compose the name of the called macro. The second
argument is the text to appear in the table of contents entry.
|\writetocentry| appends the page number to the macro call.  For
example: 
\csdisplay
\writetocentry{chapter}{Introduction}
|
\noindent
will produce the line:
\csdisplay
\tocchapterentry{Introduction}{2}
|
\noindent in the |.toc| file, indicating
that `Introduction' started on page 2.

You can use |\writenumberedtocentry| to provide a third parameter, such
as a chapter number. For example:
\csdisplay
\writenumberedtocentry{chapter}{The second chapter}{2}
|
\noindent will write a line:
\csdisplay
\tocchapterentry{The second chapter}{2}{14}
|
\noindent
You can also |\write| to |\tocfile| yourself.

\csdisplay
\newwrite\tocfile \newif\iftocfileopened
\def\opentocfile{\iftocfileopened\else
      \tocfileopenedtrue
      \immediate\openout\tocfile = \jobname.toc
\fi}%
\def\writetocentry#1#2{\ifrewritetocfile
   \opentocfile
   \write\tocfile{%
      \expandafter\noexpand \csname toc#1entry\endcsname
      {#2}{\folio}}%
\fi\ignorespaces}%
%
\def\writenumberedtocentry#1#2#3{\ifrewritetocfile
   \opentocfile
   \write\tocfile{%
      \expandafter\noexpand \csname toc#1entry\endcsname
      {#2}{#3}{\folio}}%
\fi\ignorespaces}%
|

To produce a table of contents, read the |.toc| file with
\pix^|\readtocfile|. You should call |\read!-tocfile| before the first
|\write!-toc!-entry|.  When you're processing the table of contents
without regenerating it, you should not rewrite
the |.toc| file---if you do, its contents will be lost.
The command
|\rewrite!-tocfile!-false| will prevent the rewrite.

\csdisplay
\newif\ifrewritetocfile \rewritetocfiletrue
\def\readtocfile{\testfileexistence{toc}%
   \iffileexists
      \input \jobname.toc
      \ifrewritetocfile \opentocfile \fi
   \fi}%
|

Here are some definitions of possible |\toc|\dots|entry| macros.  These
definitions are meant only as examples---running leaders across the line
is usually not the best way to typeset a table of contents.

\csdisplay
\def\tocchapterentry#1#2{\line{\bf #1 \dotfill\ #2}}%
\def\tocsectionentry#1#2{%
   \line{\quad\sl #1 \dotfill\ \rm #2}}%
\def\tocsubsectionentry#1#2{%
   \line{\qquad\rm #1 \dotfill\ #2}}%
|


\section Cross-references

\null ^^{cross-references}
\xrdef{xrefs}
The macros that follow provide symbolic
cross-referencing, so that you can refer to something in another part of
a document by name instead of by its actual page number.
\pix^|\xrdef||{foo}| defines a label |foo| to be the current page
number, and \pix^|\xrefn||{foo}| produces that page number, e.g., $77$.
More often you'll want to say something like ``see p.\thinspace77'', so
\pix^|\xref||{foo}| produces ``p.\thinspace 77''.  If |foo| is not
defined, a warning message will be given.  |\xrefwarningfalse|
suppresses the warning.

These macros provide no protection against duplicate definitions.  You can
check for duplicate definitions by sorting the cross-reference file and
checking, either mechanically or by eye, for adjacent definitions of
the same symbol.

\csdisplay
\newif\ifxrefwarning  \xrefwarningtrue
\def\xrdef#1{\begingroup
   \xrlabel{#1}%
   \edef\@wr{\@writexrdef{\the\@xrlabeltoks}}%
   \@wr
   \endgroup \ignorespaces}%
\def\@writexrdef#1{\write\reffile{%
     \string\gdef
        \expandafter\string\csname#1\endcsname
        {\noexpand\folio}\percentchar}}%
\def\xrefnumber#1{%
   \xrlabel{#1}%
   % \@xrlabeltoks now has the control sequence name.
   \toks0 = 
      \expandafter{\csname\the\@xrlabeltoks\endcsname}%
   \expandafter \ifx\the\toks0\relax
      \ifxrefwarning \message{Undefined label
         `\tokstostring{#1}'.}\fi
      {\let\spacesub = \space
       \expandafter\xdef\the\toks0
          {`{\tt \tokstostring{#1}}'}}\fi
   \the\toks0}%
\def\xref#1{p.\thinspace\xrefnumber{#1}}%
\def\xrefn#1{\xrefnumber{#1}}%
|

This macro turns a label into a list of character tokens in the token
register |\labeltoks|.  A label can include blanks and control sequences in
it as well as normal characters, but it can't include braces.

\csdisplay
\newtoks\@xrlabeltoks
\def\xrlabel#1{\begingroup
      \escapechar = `\_ \edef\tts{\tokstostring{#1_}}%
      \global\@xrlabeltoks = \expandafter{\tts}%
   \endgroup}%
|

It takes two passes to get the cross-references right, since the
definitions are written out to the auxiliary file |\jobname.aux|.
\pix^|\readreffile| reads them back in. If you don't issue this command
before the first definition, you'll lose all the definitions from the
previous run.

\csdisplay
\newwrite\reffile \newif\ifreffileopened
\def\openreffile{\ifreffileopened\else
      \reffileopenedtrue
      \immediate\openout\reffile = \jobname.aux
   \fi}%
\def\readreffile{%
   \testfileexistence{aux}%
   \iffileexists
      \begingroup
         \@setletters
         \input \jobname.aux
      \endgroup
   \else
      \message{No cross-reference file; I won't give you
         warnings about undefined labels.}%
      \xrefwarningfalse
   \fi
   \openreffile}%
\def\@setletters{%
   \catcode`_ = \letter \catcode`+ = \letter
   \catcode`- = \letter \catcode`@ = \letter
   \catcode`0 = \letter \catcode`1 = \letter
   \catcode`2 = \letter \catcode`3 = \letter
   \catcode`4 = \letter \catcode`5 = \letter
   \catcode`6 = \letter \catcode`7 = \letter
   \catcode`8 = \letter \catcode`9 = \letter
   \catcode`( = \letter \catcode`) = \letter}%
|

^^{equations, labelling}
You can give symbolic names to equations in a similar way, using
\pix^|\eqdef| and \pix^|\eqref|.  |\eqdef| inserts its own |\eqno|
command, so it must be invoked in a place where |\eqno| is legal.

\csdisplay
\newcount\eqnumber
\def\eqdef#1{\global\advance\eqnumber by 1
   \expandafter\xdef
      \csname#1eqref\endcsname{\the\eqnumber}%
   \immediate\write\reffile{\string\def
      \expandafter\string\csname#1eqref\endcsname
         {\the\eqnumber}}%
   \eqno
   \eqprint{\the\eqnumber}}%
|

|\eqref| produces ``(equation number)''.  You can handle fancier
formatting by redefining \pix^|\eqprint|. For example, you could redefine
it so that the equation numbers include
the chapter number.

\csdisplay
\def\eqref#1{%
   \expandafter \ifx \csname#1eqref\endcsname \relax
      \ifxrefwarning \message{Undefined equation label
         `#1'.}\fi
      \expandafter\def\csname#1eqref\endcsname{00}%
   \else \eqprint{\csname#1eqref\endcsname}%
   \fi}%
\def\eqprint#1{(#1)}%
|


\section Environments

\null ^^{environments} These macros let you define your own named groups
(environments) for parts of your manuscript.  Like \TeX\ groups, these
groups can be nested, and in fact their nesting can be intertwined with
the nesting of \TeX\ groups.  If the names at the beginning and end of
an environment don't match, you'll get an error message. The macros are
designed so that the message you get when such an error occurs will give
you a good chance of localizing the cause of the error easily.

You begin an environment with
\pix^^|\environment|
|\envi!-ron!-ment| |{foo}| and end it with |\endenvironment{foo}|, where
|foo| is the name of the environment.  Our macros slightly improve on
the answer to Exercise~5.7 of \texbook, by doing some checks on
|\begingroup| and |\endgroup| pairs, as well as making sure
|\environment| and |\endenvironment| pairs match.

\csdisplay
\def\environment#1{\ifx\@groupname\undefined\else
      \errhelp = \@unnamedendgrouphelp
      \errmessage{`\@groupname' was not closed by
         \string\endenvironment}\fi
   \def\@groupname{#1}%
   \begingroup
      \let\@groupname = \undefined \ignorespaces}%
\def\endenvironment#1{\endgroup
   \def\@thearg{#1}%
   \ifx\@groupname\@thearg
   \else
      \ifx\@groupname\undefined
         \errhelp = \@isolatedendenvironmenthelp
         \errmessage{Isolated
            \string\endenvironment\space for `#1'}%
      \else
         \errhelp = \@mismatchedenvironmenthelp
         \errmessage{Environment `#1' ended,
            but `\@groupname' started}%
         \endgroup % Probably a typo in the names.
      \fi
   \fi
   \let\@groupname = \undefined \ignorespaces}%
|

We also define help messages for each of the errors above.
^^|\newhelp|

\csdisplay
\newhelp\@unnamedendgrouphelp{%
   Most likely, you just forgot an^^J%
   \string\endenvironment. 
   Maybe you should try inserting another^^J%
   \string\endgroup to recover.}%
\newhelp\@isolatedendenvironmenthelp{%
   You ended an environment X, but^^J%
   no \string\environment\space to start it
   is anywhere in sight.^^J%
   You might also be at an
   \string\endenvironment\space that would match^^J%
   a \string\begingroup, i.e., you forgot an
   \string\endgroup.}%
\newhelp\@mismatchedenvironmenthelp{%
   You started an environment X, but^^J%
   you ended it with Y.  Maybe you made a typo
   in one or the other^^J%
   of the names.}%
|

Some environments should not be allowed to occur within
another environment.  Let's call these environments
``outer environments''. |\check!-env| checks that no outer environment
is currently in effect and complains if one is. To use |\check!-env|, you
must issue the command |\environment!-true| at the beginning of every
outer environment.

\csdisplay
\newif\ifenvironment
\def\checkenv{%
   \ifenvironment
      \errhelp = \@interwovenenvhelp
      \errmessage{Interwoven environments}%
      \endgroup
   \fi}%
\newhelp\@interwovenenvhelp{%
   Perhaps you forgot to end the previous^^J%
   environment? I'm finishing off the current group,^^J%
   hoping that will fix it.}%
|


\section Justification

\bix^^{flush left}
\bix^^{flush right}
\bix^^{centered text}
The three macros \pix^|\flushleft|, \pix^|\flushright|, and
\xrdef{eplaincenter}
\pix^|\center| justify  the text on the following lines in the indicated way.
The command should appear on a line by itself.
Both the command and the text should be enclosed in a group---%
the end of the group indicates the end of the text.
The entire group is set as a single paragraph, with lines filled out
on one side or another as appropriate.  Blank lines are reproduced.

\csdisplay
\begingroup
   \catcode `\^^M = \active
   \globaldefs = 1 %
   \def\flushleft{\beforejustify %
      \aftergroup\@endflushleft %
      \def^^M{\null\hfil\break}%
      \def\@eateol^^M{}\@eateol}%
   \def\flushright{\beforejustify %
      \aftergroup\@endflushright %
      \def^^M{\break\null\hfil}%
      \def\@eateol^^M{\hfil\null}\@eateol}%
   \def\center {\beforejustify %
      \aftergroup\@endcenter %
      \def^^M{\hfil\break\null\hfil}%
      \def\@eateol^^M{\hfil\null}\@eateol}%
\endgroup
|

The following commands are called as a result of the |\after!-group|
^^|\aftergroup|
in the definitions of |\flush!-left|, |\flush!-right|, and |\center|.
They perform the necessary cleanup operations.

\csdisplay
\def\@endflushleft{\unpenalty
   {\parfillskip = 0pt plus 1 fil\par}%
   \ignorespaces}%
\def\@endflushright{%
   % Remove the \hfil\null\break we just put on.
   \unskip \setbox0=\lastbox \unpenalty
   % We have fil glue at the left of the line;
   % \parfillskip shouldn't affect that.
   {\parfillskip = 0pt \par}\ignorespaces}%
\def\@endcenter{%
   % Remove the \hfil\null\break we just put on.
   \unskip \setbox0=\lastbox \unpenalty
   % We have fil glue at the left of the line;
   % \parfillskip must balance it.
   {\parfillskip = 0pt plus 1fil \par}\ignorespaces}%
\def\beforejustify{%
   \par\noindent
   \catcode`\^^M = \active
   \checkenv \environmenttrue}%
|

\eix^^{flush left}
\eix^^{flush right}
\eix^^{centered text}


\section Tables

The \pix^|\makecolumns| macro enables you to give all the entries in a
table without having to worry about where the columns break.  For
example, if you're typing a long alphabetical list that will be
formatted in several columns, you usually won't know in advance where
one column ends and the next begins.  Moreover, if another item gets
added, the column breaks will change.

|\makecolumns| takes two (delimited) arguments: the total number of
entries in the table and the number of columns in the table.  For
example, `|\makecolumns 37/3:|' specifies a three-column table whose
entries are the next $37$ lines.  You can adjust the positioning of the
table on the page by changing |\parindent|, which determines the space
to the left, and |\hsize|, which determines the space from the left
margin of the page to the right of the block.  You can allow a page
break above the |\valign|
^^|\valign//used in {\tt\\makecolumns}|
by changing
\pix^|\abovecolumnspenalty|.

\csdisplay
\newcount\abovecolumnspenalty
\abovecolumnspenalty = 10000
\newcount\@linestogo      % Lines remaining to process.
\newcount\@linestogoincolumn % Lines remaining in column.
\newcount\@columndepth    % Number of lines in a column.
\newdimen\@columnwidth    % Width of each column.
\newtoks\crtok  \crtok = {\cr}%
\def\makecolumns#1/#2: {\par \begingroup
   \@columndepth = #1 \advance\@columndepth by #2
   \advance\@columndepth by -1
   \divide \@columndepth by #2
   \@linestogoincolumn = \@columndepth \@linestogo = #1
   \def\@endcolumnactions{%
      \ifnum \@linestogo<2 
         \the\crtok \egroup \endgroup \par
            % End \valign and \makecolumns.
      \else
         \global\advance\@linestogo by -1
         \ifnum\@linestogoincolumn<2
            \global\@linestogoincolumn = \@columndepth
            \the\crtok
         \else &\global\advance\@linestogoincolumn by -1
         \fi
      \fi}%
   \makeactive\^^M\letreturn\@endcolumnactions
   \@columnwidth = \hsize
   \advance\@columnwidth by -\parindent
   \divide\@columnwidth by #2
   \penalty\abovecolumnspenalty
   \noindent % It's not a paragraph (usually).
   \valign\bgroup
      &\hbox to \@columnwidth{\strut ##\hfil}\cr
}% The next end-of-line starts everything going.
|


\section Footnotes

\null ^^{footnotes}
Footnotes are most commonly typeset by using a raised number as the
reference mark.  We define the \pix^|\numberedfootnote| macro to do
this.  It also redefines ^|\vfootnote| to allow slightly more general
formatting of footnotes than \plainTeX\ does. The dimension register
\pix^^|\footnotemarkseparation| |\foot!-note!-mark!-sepa!-ra!-tion|
controls the space between the footnote mark (e.g., the number) and the
beginning of the text. The \pix^|\everyfootnote| tokens are inserted
before producing the footnote.

The \plainTeX\ definitions of |\footnote| and |\vfootnote| are
preserved in |\@plain!-footnote| and |\@plain!-vfootnote| in case you should
need them.
\csdisplay
\newcount\footnotenumber \newtoks\everyfootnote
\newdimen\footnotemarkseparation
\footnotemarkseparation = .5em
\let\@plainfootnote = \footnote
\let\@plainvfootnote = \vfootnote
\def\vfootnote#1{\insert\footins\bgroup
  \interlinepenalty\interfootnotelinepenalty
  \splittopskip\ht\strutbox \splitmaxdepth\dp\strutbox
  \floatingpenalty\@MM
  \leftskip\z@skip \rightskip\z@skip \spaceskip\z@skip
  \xspaceskip\z@skip
  \everypar = {}%
  \the\everyfootnote
  \indent\llap{#1\kern\footnotemarkseparation}\footstrut
  \futurelet\next\fo@t}%
\def\numberedfootnote{\global\advance\footnotenumber by 1
  \@plainfootnote{$^{\number\footnotenumber}$}}%
|


\section Double columns

\null ^^{double columns} The |\doublecolumns| command begins
double-column output, while the |\single!-column|
\pix^^|\singlecolumn|
command restores
single-column output.  You can switch back and forth between them
on a single page.
The glue specified by |\above!-double!-column!-skip| and
|\below!-double!-column!-skip| is inserted before and after the
double-column material.

The approach is derived from \knuth{page~417}.

\csdisplay
\newskip\abovedoublecolumnskip
\newskip\belowdoublecolumnskip
\abovedoublecolumnskip = \bigskipamount
\belowdoublecolumnskip = \bigskipamount
\newdimen\gutter            \gutter = 2pc
\newdimen\doublecolumnhsize \doublecolumnhsize = \hsize
\newbox\@partialpage        \newdimen\singlecolumnhsize
\newdimen\singlecolumnvsize \newtoks\previousoutput
\def\doublecolumns{\par % Don't start in horizontal mode.
   \previousoutput = \expandafter{\the\output}
   \advance\doublecolumnhsize by -\gutter
   \divide\doublecolumnhsize by 2
   \output = {\global\setbox\@partialpage =
         \vbox{\unvbox255\vskip\abovedoublecolumnskip}}%
   \pagegoal = \pagetotal \break % Expands \output above.
   \output = {\doublecolumnoutput}%
   \singlecolumnhsize = \hsize
   \singlecolumnvsize = \vsize
   \hsize = \doublecolumnhsize \vsize = 2\vsize}%
|

The |\@double!-column!-split| macro does the actual splitting.
Insertions are assumed to be single-column material.  If you don't want
this to be the case, you'll have to modify the output routine.  After
|\@double!-column!-split| has done its work, |\box255| will have the
double-column material.  The double-column material will be preceded by
any single-column material that was typeset before |\doublecolumns| was
invoked. |\box4| will have the material that didn't fit on the page.

\csdisplay
\def\@doublecolumnsplit{%
   \splittopskip = \topskip \splitmaxdepth = \maxdepth
   \dimen0 = \singlecolumnvsize
      \advance\dimen0 by -\ht\@partialpage
      \advance\dimen0 by -\ht\footins
      \advance\dimen0 by -\skip\footins
      \advance\dimen0 by -\ht\topins
   \begingroup
      \vbadness = 10000
      \global\setbox1=\vsplit255 to \dimen0 \wd1=\hsize
      \global\setbox3=\vsplit255 to \dimen0 \wd3=\hsize
   \endgroup
   \global\setbox4=\vbox{\unvbox255
      \penalty\outputpenalty}%
   \global\setbox255=\vbox{\unvbox\@partialpage
      \hbox to \singlecolumnhsize{\box1\hfil\box3}%
      \vfill}}%
|
\needspace{.5in}
|\double!-column!-out!-put| is the real output routine.  We call the old
|\output| to do the work of actually shipping out the box.

\csdisplay
\def\doublecolumnoutput{\@doublecolumnsplit
   \hsize = \singlecolumnhsize \vsize = \singlecolumnvsize
   \previousoutput \unvbox4}%
|

|\singlecolumn| resumes typesetting in one column.  It assumes that
|\doublecolumns| has been called.

\csdisplay
\def\singlecolumn{\par % Don't start in horizontal mode.
   \output = {\global\setbox1 =
      \vbox{\unvbox255\vskip\abovedoublecolumnskip}}%
   \pagegoal = \pagetotal \break \setbox255 = \box1
   {\singlecolumnvsize = \ht255
      \divide\singlecolumnvsize by 2
      \advance\singlecolumnvsize by +\ht\@partialpage
      \advance\singlecolumnvsize by +\ht\footins
      \advance\singlecolumnvsize by +\skip\footins
      \advance\singlecolumnvsize by +\ht\topins
   \@doublecolumnsplit}%
   \hsize = \singlecolumnhsize
   \vsize = \singlecolumnvsize
   \output = \expandafter{\the\previousoutput}%
   \unvbox255}%
|

\margin{`Sensible paragraph skips' section deleted}


\section Finishing up

We now must undo the changes that we made when we started (see
\xref{eplainconv}).  We also give a version identification, which is
subsequently available in |\fmt!-name| and |\fmt!-version|.

\csdisplay
\let\wlog = \@plainwlog \catcode`@ = \other
\def\fmtname{eplain}%
{\edef\plainversion{\fmtversion}%
 \xdef\fmtversion{1.0: 15 May 1990
   (and plain \plainversion)}%
}%
|

\eix^^|eplain.tex|

\endchapter
\byebye