% pdcsty -- documentation and source code for several style files -*-tex-*-

%%%@TeX-document-file {
%%% title 	= "Damian's random style files",
%%% filename 	= "$texmf/doc/plain/pdcmac/pdcsty.dtx",
%%% version 	= "$Revision: 1.2 $",
%%% package 	= "pdcmac 1.0",
%%% date 	= "$Date: 1995/03/30 16:12:17 $",
%%% author	= "P. Damian Cugley",
%%% email	= "damian.cugley@comlab.ox.ac.uk",
%%% address	= "Oxford University Computing Laboratory,"
%%%                Parks Road, Oxford  OX1 3QD, UK",
%%% abstract	= "This document describes and is the source code for 
%%%                the TeX style files pdccmdoc.tex, ma55doc.tex,
%%%                etc. and definitioon files pdcdoc1.tex and pdclft1.tex.
%%%                Running plain TeX on this file produces both the
%%%                definitions files and the printed documentation.",
%%% copyright	= "Copyright (c) 1991-1995 P. Damian Cugley",
%%% copying	= "This program is free software; you can redistribute
%%%		   it and/or modify it under the terms of the GNU
%%%		   General Public License as published by the Free
%%%		   Software Foundation; either version 2 of the License,
%%%		   or (at your option) any later version.",
%%% notice	= "This program is distributed in the hope that it will
%%%		   be useful, but WITHOUT ANY WARRANTY; without even the
%%%		   implied warranty of MERCHANTABILITY or FITNESS FOR A
%%%		   PARTICULAR PURPOSE.  See the GNU General Public
%%%		   License for more details.",
%%% notice	= "You should have received a copy of the GNU General
%%%		   Public License along with this program; if not, write
%%%		   to the Free Software Foundation, Inc., 675 Mass Ave,
%%%		   Cambridge, MA 02139, USA.",
%%% codetable	= "USASCII",
%%% dependencies = "pdccode2.tex"
%%%}
	

%{{{ pdcsty.dtx
%{{{  preamble

\input pdccode2
\rcs$Id: pdcsty.dtx,v 1.2 1995/03/30 16:12:17 pdc Exp $\endrcs

\everycode{\tt \baselineskip=0.8\baselineskip \lineskiplimit=0pt }
\newcount\lineno
\everycodeline{\llap{$ \global\advance\lineno1 
	\scriptstyle\the\lineno $\enspace}}


\codefile{cmX}{pdccmlft.tex} \codefile{cmT}{pdccmdoc.tex} 
\codefile{maX}{ma55lft.tex} \codefile{maT}{ma55doc.tex}
\codefile{psX}{pdcpslft.tex} \codefile{psT}{pdcpsdoc.tex}
\codefile{X}{pdclft.def} \codefile{T}{pdcdoc.def}
	
\def\maonly{\global\maXtrue \global\maTtrue 
    \global\cmXfalse \global\cmTfalse
    \global\psXfalse \global\psTfalse
    \global\Xfalse \global\Tfalse 
    \banner{Malvern}}
\def\cmonly{\global\maXfalse \global\maTfalse
    \global\cmXtrue  \global\cmTtrue
    \global\psXfalse  \global\psTfalse
    \global\Xfalse \global\Tfalse
    \banner{Computer Modern}}
\def\psonly{\global\maXfalse \global\maTfalse
    \global\cmXfalse  \global\cmTfalse
    \global\psXtrue  \global\psTtrue
    \global\Xfalse \global\Tfalse
    \banner{PostScript fonts}}

\def\styonly {\global\maXtrue \global\maTtrue
    \global\cmXtrue \global\cmTtrue
    \global\psXtrue \global\psTtrue
    \global\Xfalse \global\Tfalse
    \banner{Style files}}
\def\Tstyonly{\global\maXfalse \global\maTtrue
    \global\cmXfalse \global\cmTtrue
    \global\psXfalse \global\psTtrue
    \global\Xfalse \global\Tfalse
    \banner{Doc style files}}
\def\Xstyonly{\global\maXtrue \global\maTfalse
    \global\cmXtrue \global\cmTfalse
    \global\psXtrue \global\psTfalse
    \global\Xfalse \global\Tfalse
    \banner{Lft style files}}

\def\Tonly {\global\maXfalse \global\maTfalse
    \global\cmXfalse \global\cmTfalse
    \global\psXfalse \global\psTfalse
    \global\Xfalse \global\Ttrue
    \banner{Sectioning macros for Docs}}
\def\Xonly {\global\maXfalse \global\maTfalse
    \global\cmXfalse \global\cmTfalse
    \global\psXfalse \global\psTfalse
    \global\Xtrue \global\Tfalse
    \banner{Sectioning macros for Lfts}}

\def\banner#1{
    \par
    \line{%
	\vrule height 1ex width 0.2ex
	\leaders\vrule height 1ex depth -0.8ex \hskip2em
	\enspace{\it \strut #1\/}\enspace
	\leaders\vrule height 1ex depth -0.8ex \hfil
    }
}

\def\onlyone#1{
    \allfilesfalse \global\csname#1true\endcsname
    \banner{{\tt\csname filename#1\endcsname}}
}
\def\onlytwo#1#2{
    \allfilesfalse \global\csname#1true\endcsname 
	\global\csname#2true\endcsname
    \banner{{\tt\csname filename#1\endcsname}, 
	{\tt\csname filename#2\endcsname}}
}

\def\common{
    \par
    \allfilesfalse \global\Xtrue \global\Ttrue
    \closebanner
}

\def\closebanner{
    \line{%
	\vrule height 1.8ex depth -0.5ex width 0.2ex
	\leaders\vrule height 0.7ex depth -0.5ex \hfil
    }
}



%}}}  preamble
%{{{  introduction

\null\vskip2\baselineskip\hrule\bigskip
\leftline{Style files for the pdc- macros \fileversion\space(\filedate)}
\bigskip\hrule\bigskip
\headline{\hfil\global\headline={\hfil
	\sevenbf Style files for the pdc- macros 
	\fileversion\space(\filedate)}} 

\section{Introduction}

\subsec{About the style files}

	This document describes a family of plain \TeX\ style files for
	typesetting documents which are short enough not to need any
	fancy indexing or cross-references, but long enough to want
	numbered section headings and a table of contents.

	A {\it style file} in this context means a definitions file that
	specifies the formatting of for a document: layout, choice of
	fonts, instructions for how to format lists, and so on.  The
	style files use the macros defined in other definition files to
	supply most of the features they need.  The document only needs
	one `|\input|' command to set up most of the commands it needs.

\subsec{No cross-references}

	Omitting cross-references is important because it means we can
	avoid having documents that have to be run through \TeX\ more
	than once (the first time to generate a raw index or
	cross-reference file, the second to incorporate the references
	into the document).  This greatly simplifies the macros.

	So that the table of contents may be generated automatically
	without requiring a second pass, it is printed {\it last\/} in
	the file.  It follows that the table of contents belongs at the
	end of the front matter of the document (on pages numbered in
	roman numerals), immediately before page~1.  It is easiest if
	all the front matter is printed last.  After the end of the
	document proper, put the front matter enclosed in |\frontmatter|
	\dots |\endfrontmatter|; this automatically switches to page
	numbering in roman numerals, and puts the table of contents at
	the end.

	The file |pdccmdoc.tex| does documents using the Computer Modern
	fonts.  There is a variant called |pdccmlft.tex| intended for
	`leaflets' that are even shorter than the `docs' produced with
	|pdccmdoc.tex|.  Leaflets have no front matter section, no table
	of contents, and section headings are much less prominent---they
	should only be a few pages long at most.  The Malvern versions
	are called |ma55doc.tex| and |ma55lft.tex|.

\subsec{About this document}

	The file that generates this printed documentation, |pdcsty.dtx|
	is a `documented \TeX' file.  As well as briefly describing what
	the macros are for and how they work, it includes a complete
	listing of their definitions.  The text of the definitions is
	also written to files in the current directory, generating fresh
	copies of the style files.  This way this documentation and the
	files it describes are guaranteed to match.

	The lines of all definitions are numbered continuously, even
	though only a subset of all the definition lines will go in each
	file.
	
%}}}  introduction
%{{{  starting up

\section{Starting up}

\subsec{File identification}
	
	Nowadays, macro files start with some comments identifying the
	file, for the benefit of people wondering what the file is for.
	The macro file otherwise has no other comments, in the hope that
	this will let \TeX\ read it in faster.  (It should be possible
	for the identification comments to go {\it after} an |\endinput|
	at the close of the file, but the convention is to have them at
	the start.)

	The following segment is common to the style files and the
	definiton files.

\code
	\|\% \filename~\fileversion~\filedate~-- macros for short documents
\smallbreak
	|%%%@TeX-definition-file {
	\|\%\%\% filename~~~~~~~= "\filename",
	\|\%\%\% version~~~~~~~~= "\fileversion",
	\|\%\%\% date~~~~~~~~~~~= "\filedate",
	|%%% package        = "pdcmac 1.0",
	|%%% author         = "P. Damian Cugley",
	|%%% email          = "damian.cugley@comlab.ox.ac.uk",
	|%%% address        = "Oxford University Computing Laboratory,
	|%%%                   Parks Road, Oxford  OX1 3QD, UK",
	|%%% codetable      = "USASCII",
	|%%% keywords       = "TeX, plain TeX, style file",
	|%%% supported      = "Maybe",
\cmonly	|%%% abstract       = "Top-level macro file for documents set in
	|%%%                   Computer Modern fonts.  
\maonly	|%%% abstract       = "Top-level macro file for documents set in
	|%%%                   Malvern 55, 56 and 75 fonts.  
\onlytwo{X}{T}
	|%%% abstract       = "Sectioning macros for style files.
\allfilestrue\closebanner
	|%%%                   This macro file was generated by running plain
	\|\%\%\%~~~~~~~~~~~~~~~~~~~TeX on \jobname.dtx",
	|%%% copyright      = "Copyright (c) 1991-1995 P. Damian Cugley",
	|%%% copying        = "DO NOT DISTRIBUTE THIS FILE.
	\|\%\%\%~~~~~~~~~~~~~~~~~~~Distribute \jobname.dtx only as part of the
	|%%%                   package it came in.",
\onlytwo{X}{T}
	|%%% dependencies   = ""
\cmonly	|%%% dependencies   = "pdcfmt2.tex, pdcfsel.tex, pdchyex.tex,
	|%%%                   pdcimth.tex, pdccmsub.tex, pdcmigr.tex,
\maonly	|%%% dependencies   = "maamac.tex, pdcfmt2.tex, pdcfsel.tex,
	|%%%                   pdchyex.tex, pdcimth.tex, pdcmigr.tex,
\psonly	|%%% dependencies   = "pdcadobe.tex, pdcfmt2.tex, pdcfsel.tex,
	|%%%                   pdchyex.tex, pdcimth.tex, pdcmigr.tex,
\Tstyonly
	|%%%                   pdcoput5.tex, pdcdoc.def",
\Xstyonly
	|%%%                   pdcoput5.tex, pdclft.def",
\allfilestrue\closebanner
	|%%%}
	|
	\|\\message\{\fileversion~<\fileauthor~\filedate>\}
	|
\endcode

	This is followed immediately by a list of the macro files
	included by this format:
\code
\styonly
	|\input pdcimth
	|\input pdcmigr
	|\input pdcfmt2
	|\input pdchyex
	|\input pdcoput5
	|\input pdcfsel
\cmonly
	|\input pdccmsub
\maonly
	|\input maamac
\psonly
	|\input pdcadobe
\common
\endcode


%}}}  starting up
%{{{  layout
	
\section{Layout}

\subsec{Paper sizes}

	The section defining the paper size is at the very top of the
	file to make it easy for someone adapting the macros to American
	Letter size paper to find them.

	Nevertheless, we set the top and bottom margins to be large
	enough that even if it is printed on American paper, nothing
	important will be cut off.  This requires decent-sized margins
	at the top and bottom of pages (since the end of the page that
	will be lost depends on whether the printer driver has its real
	origin at top-left or bottom-left).  The edge of the printing
	area is guessed at as being $7\,{\rm mm}$ from the edge of the
	paper.
\code
	|\topmg=297mm \advance\topmg-11in \advance\topmg7mm
	|\botmg=\topmg
	|\advance\topmg\ht\strutbox \advance\topmg\headlineskip
	|\ifdim \topmg<20mm \topmg=20mm \fi
	|\ifdim \botmg<27mm \botmg=27mm \fi
\smallbreak
	|\setpaper{210mm}{297mm}
\endcode

	The document has `tab-index' headings, that is, headings go in a
	wide left margin to the left of the body text.  The underlying
	structure is a grid is 4~cells wide, with the text spanning
	across the rightmost 3~cells.  Definitions in the fonts section
	below specify that the text will be set ragged-right.
\code
	|\setnkgrid{4}{3}
	|\colrule=0pt
 	|\leftmargin=\gridwd \advance\leftmargin\colsep
\endcode
	The output routine's |\leftbox| is set at the start of each page
	to make space for the left margin.
\code
	|\everypage={\ifdim\leftmargin>0pt 
	|    \setbox\leftbox=\hbox to \gridwd{}\fi}
\endcode

	While we're here, we customize some of the environments supplied
	by pdcFMT2:
\code
	|\everyfootnote={\notefonts}
	|\everylinenum={\smallfonts}
\maonly
	|\def\footnotetextmark#1{{\rmsy#1}}
	|\def\footnotenotemark#1{{\rmsy#1}}
\common
	|\everylinenum={\smallfonts}
	|\everytag={\bf}
\endcode

%}}}  layout
%{{{  fonts
\section{Fonts}

	This section describes the fonts available in the document.

	We introduce extra nicknames |\bi| for bold italic (used for the
	first mention of new techical terms), |\mf| for the METAFONT
	logo, |\mfi| for the METAFONT logo in italic (Something of an
	affectation) and |\ttb| for an `emphasized' typewriter (usually
	it is best if this is slanted, not bold).
\code
\styonly
	|\autoloadfonts
\endcode

	The style files differ because of encoding schemes more than
	because of the particular fonts.  For example, a style for Times
	Roman will work for Palatino too,if we change the names of the
	fonts used.  To do this, a macro |\basefontname| may be defined
	before the style file is read in; if it isn't defined it is
	given a default value.

%{{{   cmr

\subsec{Fonts for Computer Modern documents}
\cmonly
	We use Computer Modern Sans Serif for the headings in CM
	documents.  I have not put much effort into making formulas
	work in headings.  
\code
	|\def\STYheadingT{%
	|    \m{rm}{cmss}\m{mi}{cmmi}\m{sy}{cmsy}%
	|    \f{it}{cmssi}\f{tt}{cmtt}\f{mf}{logo}\f{mfi}{logosl}%
	|}
\smallbreak
	|\font\headingrm=cmss17 \font\headingit=cmssi17
	|\fontset{subheading}\STYheadingT{10 scaled 1200}{14pt}
	|    {small}{tiny}
	|\fontset{heading}\STYheadingT{10 scaled 1728}{20pt}
	|    {subheading}{small}
\endcode

	We use Computer Modern for body text.
\code
	|\def\STYbodyT{\m{rm}{cmr}\m{it}{cmti}%
	|    \m{bf}{cmbx}\@\f{bi}{cmbxti10}%
	|    \m{mi}{cmmi}\m{sy}{cmsy}%
	|    \@\f{mf}{logo10}\@\f{mfi}{logosl10}%
	|    \f{tt}{cmtt}\@\f{ttb}{cmsltt10}%
	|}
\smallbreak
	|\font\tinytt=cmtt8 at 7pt
	|\font\bodysy=cmsy10 at 12pt
\endcode
	
%}}}   cmr
%{{{   malvern
	
\subsec{Fonts for Malvern documents}
\maonly

	Headings are in Malvern~75.  The maths fonts are scaled for
	expediency; I~do not make much effort to make formulas work in
	headings.
\code
	|\def\STYheadingT{%
	|    \m{rm}{ma75a}\@\m{mi}{cmmi10}\@\m{sy}{cmsy10}%
	|    \f{it}{ma76a}\@\f{tt}{cmtt10}\@\f{mf}{logo10}\@\f{mfi}{logosl10}%
	|}
\smallbreak
	|\fontset{subheading}\STYheadingT{12}{14pt}{subheading}{subheading}
	|\fontset{heading}\STYheadingT{18}{20pt}{subheading}{subheading}
\endcode


	We use Malvern~55 for body text.  The supplementary font with
	encoding~B is added, with font nickname |\rmsy|.
\code
	|\newfam\rmsyfam
\smallbreak
	|\def\STYbodyT{%
	|    \m{rm}{ma55a}\f{rmsy}{ma55b}\m{it}{ma56a}%
	|    \m{bf}{ma75a}\f{bi}{ma76a}%
	|    \m{mi}{cmmi}\m{sy}{cmsy}%
	|    \@\f{mf}{logo10}\@\f{mfi}{logosl10}%
	|    \f{tt}{cmtt}\f{tb}{cmsltt}%
	|}
\endcode
%	The main problem with defining the body fonts is that the
%	$x$-height of Malvern is larger then Computer Modern. This makes
%	in necessary to load the maths fonts at a larger size to match.
%	(A better solution might be to make a parameter file for
%	Computer Modern so that maths fonts may be generated with
%	compatible proportions.)
%\code
%	|\loadfont\bodytt{cmtt10 scaled \magstep2}
%	|\loadfont\notett{cmtt10 scaled \magstep1}
%	|\loadfont\smalltt{cmtt10 scaled \magstephalf}
%	|\loadfont\bodytb{cmsltt10 scaled \magstep2}
%	|\loadfont\smalltb{cmsltt10 scaled \magstep1}
%	|\loadfont\notetb{cmsltt10 scaled \magstephalf}
%\endcode
	There is no point using |\loadfont| on the maths symbol fonts,
	since they have to be loaded anyway the first time |\bodyfonts|
	is called.
\code
	|\font\bodymi=cmmi10 scaled \magstep2
	|\font\bodysy=cmsy10 scaled \magstep2
	|\font\notemi=cmmi10 scaled \magstep1
	|\font\notesy=cmsy10 scaled \magstep1
	|\font\smallmi=cmmi10 scaled \magstephalf
	|\font\smallsy=cmsy10 scaled \magstephalf
\endcode
%}}}   malvern
%{{{   PostScript
	
\subsec{Fonts for Times documents}
\psonly
	The big problem with PostScript fonts is what they are called;
	I~shall be assuming that Fontname~$1.6$\footnote*{Karl Berry,
	{\it File Names for \TeX\ Fonts} $1.6$ ({\tt
	ftp:\penalty\exhyphenpenalty //ftp.umb.edu\slash pub\slash
	tex\slash fontname/}).} applies: fonts with |afm2tfm|'s
	variation on the \TeX\ Text encoding are called |ptmr| and fonts
	with Adobe's StandardRoman encoding are called |ptmr0| (rather
	than |rptmr|).

	This macro attempts to make it possible to guess the standard
	name for the italic variant of a given font.
	|\KBcanonical{|\<tentative font name>|}| takes a font name which
	is formed from an upright name optionally followed by `|i|'
	representing an italic variation.  It returns the corresponding
	standard name.  For example, |ptmri| is left unchanged, but
	|phvri| is changed to |phvro| and |phvrrni| is changed to
	|phvron|.
\code
	|\def\KBcanonical#1{%
	|    \expcs\ifx{KB#1}\relax #1\else\csname KB#1\endcsname \fi
	|}
\smallbreak
	|\ifx\KBadobe\UNDEFINED \def\KBadobe#1{\KBcanonical{#10}} \fi
	|\def\KBit#1{\KBcanonical{#1i}}
	|\def\KBbi#1{\KBbf{#1i}}
	|\def\KBbf#1{\expandafter\KBbfx#1.}
	|\def\KBbfx#1#2#3#4#5.{\KBcanonical{#1#2#3b#5}}
\smallbreak
	|\def\KBpagbr{pagdr}        \def\KBpagbri{pagdro}
	|\def\KBpbkbr{pbkdr}        \def\KBpbkbri{pbkdri}
	|\def\KBpagri{pagro}        \def\KBpagbi{pagbo}
	|\def\KBphvri{phvro}        \def\KBphvbi{phvbo}
	|\def\KBpcrri{pcrro}        \def\KBpcrbi{pcrbo}
	|\def\KBphvrrni{phvron}     \def\KBphvbrni{phvbon}
	|\def\KBphvrrn0{phvr0n}     \def\KBphvbrn0{phvb0n}
	|\def\KBphvrrni0{phvro0n}   \def\KBphvbrni0{phvbo0n}
\endcode

	Headings are in Helvetica.  The maths fonts are scaled for
	expediency; I~do not make much effort to make formulas work in
	headings.
\code

	|\ifx\headingfontname\UNDEFINED \def\headingfontname{phvr} \fi
	|\ifx\headingttfontname\UNDEFINED \def\headingttfontname{pcrb} \fi
\smallbreak
	|\def\STYheadingT{%
	|    \@\f{rm}{\headingfontname}\@\m{mi}{cmmi10}\@\m{sy}{cmsy10}%
	|    \@\f{asy}{psyr}%
	|    \@\f{it}{\KBit\headingfontname}%
	|    \@\f{tt}{\headingttfontname}\@\f{mf}{logo10}\@\f{mfi}{logosl10}%
	|}
\smallbreak
	|\loadfont\bodysy{cmsy10 at 12pt}
\smallbreak
	|\fontset{subheading}\STYheadingT{12}{14pt}{subheading}{subheading}
	|\fontset{heading}\STYheadingT{18}{20pt}{subheading}{subheading}
\endcode


	We use Times Roman by default for body text, and Courier for
	typewriter text.  Adobe's Symbol font is included, with font
	nickname |\asy|; a family called |\asyfam| has already been
	created by |pdcadobe.tex|.
\code
	|\ifx\bodyfontname\UNDEFINED \def\bodyfontname{ptmr} \fi
	|\ifx\bodyttfontname\UNDEFINED \def\bodyttfontname{pcrr} \fi
\smallbreak
	|\def\STYbodyT{%
	|    \@\m{rm}{\bodyfontname}\@\m{it}{\KBit\bodyfontname}%
	|    \@\m{bf}{\KBbf\bodyfontname}\@\f{bi}{\KBbi\bodyfontname}%
	|    \@\m{asy}{psyr}\m{mi}{cmmi}\m{sy}{cmsy}%
	|    \@\f{mf}{logo10}\@\f{mfi}{logosl10}%
	|    \@\f{tt}{\bodyttfontname}\@\f{tb}{\KBit\bodyttfontname}%
	|}
\endcode


	Finally the alphabet-switching commands defined for Malvern
	fonts are given dummy definitions.
\code
	|\let\sc=\relax  \let\mc=\relax  \let\csc=\relax
\endcode

%}}}   malvern
	
\common
\subsec{Definitions common to all formats}

\code
\styonly
	|\fontset{tiny}\STYbodyT{7}{9pt}{tiny}{tiny}
	|\fontset{small}\STYbodyT{9}{11pt}{tiny}{tiny}
	|\fontset{note}\STYbodyT{10}{12pt}{tiny}{tiny}
	|\fontset{body}\STYbodyT{12}{14pt}{small}{tiny}
\endcode

	The text is set left-justified (ragged-right).  To make this
	work, the fontdimens that control word spacing are set to force
	interword spaces to all be the same width.
\code
	|\def\everyloadfont#1#2{\fontdimen3#1=0pt \fontdimen4#1=0pt}
	|\rightskip=0pt plus 2em \tolerance=1000
\smallbreak
	|\bodyfonts
	|{\tt \hyphenchar\font=-1 }
\common
\endcode

%}}}  fonts
%{{{  minimal section headings
\section{Simple sectioning}

\onlyone{X}
	A `leaflet' is divided into short sections, which may contain
	subsections (but usually will not).  There may well be more than
	one section on a page.  

	The section number is held in two count registers, and printed
	(enclosed in |\n{|\dots|}|) with the macro |\thesecno|:

\code
	|\newcount\secno \newcount\subsecno
\smallbreak
	|\def\thesecno{%
	|    \n{%
	|        \number\secno
	|        \ifnum\subsecno>0
	|            .\number\subsecno
	|        \fi
	|    }%
	|}
\endcode

	The section and subsection commands are implemented
	straightforwardly with macros from |pdcfmt2.tex|, since there is
	no need for marks or fiddling with table-of-contents files.
\code
	|\def\section{
	|    \global\advance\secno1 \global\subsecno0
	|    \doheading\medskipamount{\subheadingfonts\thesecno\enspace}{}
	|}
\smallbreak
	|\def\subsec{
	|    \global\advance\subsecno1
	|    \doheading\smallskipamount{\bodyfonts\thesecno\enspace}{}
	|}
\endcode

	The page header includes only the title of the file, which the
	user sets with the command |\majorheadline{|\<text>|}|
\code
	|\newtoks\majorheadline 
	|\expandafter\majorheadline	
	|    \expandafter{\expandafter\uppercase\expandafter{\jobname}}
\endcode
	All those |\expandafter|s are to get the initial value in upper
	case.

	The flag |noheadline| signals that the headline is to be omitted
	on this page (e.g., the title page).  The |twosided| flag is set
	for two-sided binding.
\code
	|\newif\ifnoheadline \newif\iftwosided 
\smallbreak
	|\headline{%
	|    \ifnoheadline
	|        \hfil
	|    \else\iftwosided
	|        \ifodd\pageno
	|            \hfil \the\majorheadline \quad{\bf\folio}%
	|        \else
	|            {\bf\folio}\quad \the\majorheadline\hfil
	|        \fi
	|    \else
	|        \hfil \the\majorheadline \quad{\bf\folio}%
	|    \fi\fi
	|}
\smallbreak
	|\footline{%
	|    \ifnoheadline
	|        \global\noheadlinefalse
	|        \hfil{\bf\folio}\hfil
	|    \else
	|        \hfil
	|    \fi
	|}
\endcode

%}}}  minimal section headings
%{{{  sectioning
\section{Sectioning in PDCCMDOC documents}

\onlyone{T}
	Now a document is divided into {\it sections}, which are further
	divided into so-called {\it subsecs} and {\it subsubsecs}.
	(Division into subsubsecs should be done sparingly.)

	The sections may be grouped into {\it parts}---or, to look at it
	another way, parts are a larger division of the document than
	sections, but sections are not numbered within part numbers.
	Each |\part| command should be almost immediately followed by
	|\section|.

\subsec{Macros for section numbers}

	The part and section numbers are kept in \TeX\ number registers,
	with corresponding `|\the|-' macros to print the value
	formatted.
\code
	|\newcount\partno
\smallbreak
	|\def\thepartno{%
	|    \ifcase\partno O\or I\or II\or III\or IV\or V\else
	|        \expandafter\uppercase\expandafter{\romannumeral\partno}%
	|    \fi
	|}
\endcode
	Formatted section and subsec numbers are generated by one macro
	|\thesecno|.  It automatically omits (sub)subsec numbers if they
	are zero.  The whole number is enclosed in |\n{|\dots|}| so that
	the formatting of numbers may be customized.
\code
	|\newcount\secno \newcount\subsecno \newcount\subsubsecno
\smallbreak
	|\def\thesecno{%
	|    \n{%
	|        \number\secno
	|        \ifnum\subsecno>0 
	|            .\number\subsecno
	|            \ifnum\subsubsecno>0
	|        	.\number\subsubsecno
	|            \fi
	|        \fi
	|    }%
	|}
\endcode

\subsec{Parts}

	The |\part| macro takes one argument, the title.  It sets a
	token register |\majorheadline| which is printed on the left
	page of two-sided documents.
	
	The |\majorheadline{|\<text>|}| command may be used immediately
	after a |\part| command if the part title is not suitable to go
	in the headline.  This should also be used at the start of the
	document to the title of the document, if division into parts is
	not used.
\code
	|\newtoks\majorheadline
	|\expandafter\majorheadline	
	|    \expandafter{\expandafter\uppercase\expandafter{\jobname}}
\smallbreak
	|\def\part#1{%
	|    \global\advance\partno1
	|    \newpageheading{2\baselineskip}{\headingfonts}
	|        {}{\thepartno.\quad#1}
	|    \TOCwrite\TOCentrypart{\thepartno}{#1}
	|}
\endcode

\subsec{Section headings}

	Section headings will start on a new page if the current page is
	almost finished.  The section heading is set across the full
	width of the page, including the margin usually used for
	tab-indexed headings.

	To force the next section to start a new page, do
	|\sectionpenalty-10000| immediately before it.  
\code
	|\newcount\sectionpenalty 
\smallbreak
	|\def\sectionheading#1{
	|    \par
	|    \vskip 0pt plus 0.3\vsize 
	|    \penalty\sectionpenalty \sectionpenalty-200
	|    \vskip 1\bigskipamount plus -0.3\vsize
	|    \begingroup \advance\leftskip-\leftmargin
	|        \parskip=0pt \parindent=0pt
	|        \hyphenpenalty=10000 \exhyphenpenalty=500 
	|        \interlinepenalty=100
	|        \headingfonts \nointerlineskip
	|        \hbox{\hskip-\leftmargin 
	|        	\vrule height 2pt depth 0pt width \leftmargin
	|        	\vrule height 2pt depth 0pt width \hsize}
%	|        \nobreak\medskip
	|       #1\par
	|        \nobreak\smallskip
	|    \endgroup
	|    \global\noindenttrue
	|}
\endcode
	As well as generating a heading, the section macro generates
	marks used for the page headings and a TOC entry.  The marks are
	of the form in a way similar to that discussed in Chapter~23 of
	the {\it\TeX book}.  More precisely, a mark is placed on either
	side of the section heading of the form
$$\hbox{\tt\arg{\<rhs>}\arg{\<lhs>}}$$ where \<rhs> is the section
	number that belongs on the right-hand page, reflecting the
	situation at the bottom of the page, whereas \<lhs> is what
	belongs on following left-hand pages.
\code
	|\newtoks\sectiontoks
	|\newif\ifsecno \secnotrue
\smallbreak
	|\def\section#1{
	|    \ifsecno
	|        \global\advance\secno1 \global\subsecno0 \global\subsubsecno0
	|    \fi
	|    \toks0={#1}
	|    \mark{{\the\sectiontoks}{\ifsecno\S\thesecno\space\fi\the\toks0 }}
	|    \edef\tmp{\ifsecno\noexpand\S\the\secno\space\fi}
	|    \global\sectiontoks\expandafter{\tmp#1}
	|    \sectionheading{\ifsecno\thesecno\quad\fi #1}
	|    \mark{{\the\sectiontoks}{\the\sectiontoks}}
	|    \TOCwrite\TOCentrysection{\ifsecno\thesecno\fi}{#1}
	|}
\endcode
	Subsecs and subsubsecs use the |\doheading| macro defined in
	|pdcfmt.tex|.  They do not appear in page headings, but may
	appear in the TOC.
\code
	|\def\subsec#1{%
	|    \ifsecno \global\advance\subsecno1  \global\subsubsecno0 \fi
	|    \doheading\medskipamount{\subheadingfonts}{}
	|        {\ifsecno\thesecno\quad\fi#1}
	|    \TOCwrite\TOCentrysubsec{\ifsecno\thesecno\fi}{#1}
	|}
\smallbreak
	|\def\subsubsec#1{%
	|    \ifsecno \global\advance\subsubsecno+1 \fi
	|    \doheading\smallskipamount{\rm}{}{\ifsecno\thesecno\quad\fi#1}
	|    \TOCwrite\TOCentrysubsubsec{\ifsecno\thesecno\fi}{#1}
	|}
\endcode

	The headline and footline are controlled by two flags.  The
	command |\noheadlinetrue| suppresses the headline for this page
	(e.g., a title page).  The |twosided| flag says whether the
	document should be formatted for two-sided printing.
\code
	|\newif\ifnoheadline
	|\newif\iftwosided
\smallbreak
	|\footline={%
	|    \ifnoheadline 
	|        \global\noheadlinefalse
	|        \hfil\bodyfonts\n\folio\hfil 
	|    \else
	|        \hfil
	|    \fi
	|}
\endcode
	The headline varies according to whether it is a one-sided page,
	a left-hand page or a right-hand page.  On a one-sided page, it
	has the major headline (part or document title) at the left, and
	the section title of the {\it first} section on the page on the
	right with the folio.  On a left-hand page, there is the folio
	and the major headline, and on a right-hand page there is the
	section title for the {\it last} section on the page, and the
	folio.
\code
	|\headline={%
	|    \ifnoheadline
	|        \hfil
	|    \else
	|        \bodyfonts 
	|        \iftwosided
	|            \ifodd\pageno
	|        	\hfil 
	|        	\expandafter\firstoftwo\botmark\quad
	|        	{\bf\n\folio}%
	|            \else
	|        	{\bf\n\folio}\quad
	|        	\the\majorheadline
	|        	\hfil
	|            \fi
	|        \else
	|            \the\majorheadline
	|            \hfil 
	|           \expandafter\secondoftwo\topmark\quad
	|            {\bf\n\folio}%
	|        \fi
	|    \fi
	|}
\smallbreak
	|\def\firstoftwo#1#2{#1}
	|\def\secondoftwo#1{}
\endcode
%}}}  sectioning
%{{{  table of contents
\section{Table of contents intrinsics}

	This section defines the macros used to generate a table of
	contents.  The entries are written to a file \<jobname>|.toc| in
	the current directory.  At the end of the manuscript this files
	is closed and then read to generate the table of contents.

	Macros special to the TOC-generating section have names starting
	with `|\TOC|-'.
\code
	|\newwrite\TOCfile
	|\openout\TOCfile=\jobname.toc
\endcode

	Entries in the TOC file are of the form 
$$\hbox{\tt\cs{TOCentry}\<foo>\arg{\<secno>}\arg{\<title>}\arg{\<folio>}}$$
	where |\TOCentry|\<foo> is a csname, \<secno> is some
	automatically-generated section number like `|1.2|', \<title> is
	the section title as supplied by the user, and \<folio> is the
	folio for this page, in arabic or lower-case roman numeral form.

	They are generated with calls to |\TOCwrite| like this
$$\hbox{\tt\cs{TOCwrite}\cs{TOCentry}\<foo>\arg{\<secno>}\arg{\<title>}}$$
	The \<secno> is evaluated when |\TOCwrite| is called, whereas
	the \<title> is carefully written without being expanded so that
	its expansion happens when the TOC file is read in.  The
	\<folio> is automatically appended when it is written.
\code
	|\def\TOCwrite#1#2#3{%
	|    \begingroup\def\n{\string\n}
	|    \def\tmp{#3}%
	|    \edef\tmp{\string#1{#2}{\expandafter\TOCtrim\meaning\tmp}}
	|    \write\TOCfile\expandafter{\tmp{\folio}}\endgroup
	|}
\smallbreak
	|\def\TOCtrim#1:->{}
\endcode


	Now we define all the `|\TOCentry|-' macros.  
\code
	|\def\TOCentrypart#1#2#3{
	|    \smallskip
	|    \dimen0=\hsize \advance\dimen0\leftmargin
	|    \moveleft\leftmargin\hbox to\dimen0{\strut\bf #1. #2\hfil}
	|    \smallskip
	|}
\smallbreak
	|\def\TOCentrysection#1#2#3{
	|    \smallskip\noindent\llap{#1\kern1pc}%
	|        {\bf \ignorespaces #2\quad\rm \n{#3}}\par
	|}
\smallbreak
	|\def\TOCentrysubsec#1#2#3{
	|    \ifvmode \noindent \else \quad \fi
	|    {#1}~{\ignorespaces #2}
	|}
\smallbreak
	|\def\TOCentrysubsubsec#1#2#3{}
\endcode
%}}}  table of contents
%{{{  front matter
\section{Printing the front matter}

	The front matter of the document---the title page, preface,
	forword, etc.---must be printed {\it last}, with the table of
	contents at the end of the front matter; this is so that the
	table of contents may be accumulated during the \TeX ing of the
	file.\footnote*{It has a benefit for people reading the document
	with a browser: page~1 of the document is the first page of the
	DVI file, which makes selecting a given page easy, and the table
	of contents is at the very end, so the browsers `go to last
	page' command can be used to find the table of contents
	quickly.}
     	The front matter starts with the command |\frontmatter| and may
	contain |\section| commands.  Such sections will be unnumbered
	and will not appear in the table of contents.
	
	For a short document, a separate title page is probably
	excessive, and an abstract may be preferable to a preface.  In
	this case the first page after |\frontmatter| could have the
	title of the document (with author etc.)\ followed by an
	abstract, any copyright information (or other small print), and
	the contents (generated by |\endfrontmatter|).  In other words,
	something like this:
$$\halign{\indent#\hfil\cr
	|\frontmatter|\cr
	\quad\it commands to print the title, etc.\cr
	\quad|\abstract|\cr
	\qquad\it the text of the abstract\cr
	\quad|\endabstract|\cr
	\quad\it copyright information, etc.\cr
	|\endfrontmatter|\cr
}$$
 	For a longer document, there will be a separate title page and
	perhaps a preface.
$$\halign{\indent#\hfil\cr
	|\frontmatter|\cr
	\quad|\titlepage|\cr
	\qquad\it commands to print the title, etc.\cr
	\quad|\splittitlepage|\cr
	\qquad\it print copyright information, etc.\cr
	\quad|\endtitlepage|\cr
	\quad|\section{Preface}|\cr
	\qquad\it text of preface, etc.\cr
	|\endfrontmatter|\cr
}$$ 
 	The macro |\splittitlepage| marks the division between the title
	page (title recto, page~i) and the back of the title page (title
	verso, page~ii), which is where copyright information goes.  On
	a one-sided document, the copyright information belongs on the
	title recto, because the title verso will be blank, so
	|\splittitle| instead does |\vfill|.

	Setting |\pageno| nagative makes it print in roman numerals.
	Calling |\secnofalse| suppresses section numbering.  The first
	page has no headline.
\code
	|\def\frontmatter{
	|    \write\TOCfile{\string\endinput}
	|    \write\TOCfile{\folio}
	|    \closeout\TOCfile
	|    \vfill\supereject
	|    \ifvoid\partialpage\else \null\vfill\eject \fi
	|    \beginthe{frontmatter}
	|    \pageno=-1 \secnofalse \noheadlinetrue
	|    \def\TOCwrite##1##2##3{}
	|}
\endcode
	The front matter is finshed off with the contents list,
	generated by |\endfrontmatter|.  This reads the TOC file, which
	was closed at the end of the last page of the body of the
	document.
\code
	|\def\endfrontmatter{
	|    \section{Contents}
	|    \input\jobname.toc
	|    \vfill\supereject
	|    \endthe{frontmatter}
	|}
\endcode

	Here's macros for generating the titlepage---it shoul;d enclose
	commands to generate the title, author and publisher information
	for the document.  If it is used, it goes immediately after
	|\frontmatter|, so that the title page is page~i.
\code
	|\def\titlepage{
	|    \vfill\eject
	|    \beginthe{titlepage}
	|    \noheadlinetrue
	|    \def\splittitlepage{
	|        \iftwosided \vfill\eject\null\vfill \else \vfill \fi
	|    }
	|}
\smallbreak
	|\def\endtitlepage{
	|    \eject
	|    \endthe{titlepage}
	|}
\endcode

	This |abstract| environment doesn't do anything fancy.
\code
	|\def\abstract{
	|    \smallskip 
	|    \beginthe{abstract}
	|        \setbox\parbox\hbox to \gridwd{\subheadingfonts Abstract\hfil}
	|        \noindenttrue \bodyfonts
	|}
\smallbreak
	|\def\endabstract{
	|        \smallskip
	|    \endthe{abstract}
	|}
\endcode		
%}}}  front matter
	This concludes the code that is included only in documents with TOC.
\common

%{{{  abbreviations

\section{abbrevs}


	The definitions of |\pt| and |\mm| work only in maths mode.
\code
	|\def\pt{\,{\rm pt}}
	|\def\mm{\,{\rm mm}}
\smallbreak
	|\def\MF{{\ifdim\fontdimen1\font>0pt \mfi \else \mf \fi META}\-%
	|        {\ifdim\fontdimen1\font>0pt \mfi \else \mf \fi FONT}}
	|\def\MSDOS{\leavevmode\hbox{MS-DOS}}
\endcode

\subsec{Program names with small caps}

	Some program names use small capitals in their names.  For
	normal fonts there is no point trying to duplicate this, since
	small capitals aren't common to all fonts.  But small capitals
	come `for free' with Malvern, so we might as well use them.  The
	following definitions apply to Malvern~A fonts.
\code
\maonly
	|\def\PS{\csc{PostScript}}
	|\def\La{L\negthinspace\flushtop{\char"C1 }}
\cmonly
	|\def\PS{PostScript}
\common
\endcode

\font\sf=cmss10
	Here's a redefinition of |\TeX|.  This adds a spacefactor 1000
	to the end (so that \TeX\ doesn't apply its usual heuristic that
	a full stop following a capoital {\it X} isn't a sentence end).
	It also loosens up the logo, producing `\TeX' rather than the
	usual `T\negthinspace\lower0.5ex\hbox{E}\kern-0.125emX'.  This
	design is more forgiving of changes in font (for example, in
	sanserif it makes `{\sf\TeX}' instead of `{\sf
	T\negthinspace\lower0.5ex\hbox{E}\kern-0.125emX}').
\code
	|\def\TeX{T\kern-0.1em\lower0.5ex\hbox{E}X\spacefactor1000 }
\endcode
	
	The macro |\dfn{...}| formats a defining instance of a word or
	phrase (usually used the first time a new technical term is
	introduced).  When generating an index with a command |\index|,
	this also adds the phrase to the index.
\code
	|\begingroup\catcode`\|=12 \toks0={\endgroup
	|    \def\dfn#1{\ifx\index\UNDEFINED\else\index{#1|INDdfn}\fi{\bi #1}}
	|} \the\toks0
\endcode

	Here's a hack for using email-style emphasis using
	`asterisking'.  Text enclosed in matching asterisks is
	emphasized.
\code
	|\declareactivechar\* \def*#1*{{\it#1\/}}
\endcode


\begingroup\defverbatim\~
	When documenting computer progams it is useful to be able to
	present dialogues between the user and the computer, with the
	user's commands distinguished by a slightly different font.
	Here we define ~\|~\<text>~|~ so that it behaves like
	~|~\<text>~|~ except that |\ttb| is used instead of |\tt|.  In
	maths mode it is still equivalent to |\Vert|.
\code
	|\def\|{%
	|    \relax\ifmmode
	|        \Vert
	|    \else
	|        \begingroup
	|        \ttb \setupverbatim 
	|        \catcode`\|=13 \let|\endgroup
	|    \fi
	|}
\endcode
\endgroup
%}}}  abbreviations
%{{{  postamble
\section{Finally}

	The definitions common to all style files are read last, so that
	they can take account of changes to |\baselineskip|.
\code
\Tstyonly
	|\input pdcdoc.def
\Xstyonly
	|\input pdclft.def
\common
\endcode

	And, at last, a file {\tt\<filename>.cfg} is read, if it exists. 
\code	
	|\def\inputifexists#1{%
	|    \openin15=#1
	|    \ifeof15 \else \input #1 \fi
	|    \closein15
	|}
\styonly
	\|\\inputifexists\{\filenamesanssuffix.cfg\}
\common
\endcode

\endcodefile{cmX} \endcodefile{cmT} \endcodefile{maX} \endcodefile{maT}
\endcodefile{psX} \endcodefile{psT} \endcodefile{X} \endcodefile{T}
\bye
%}}}  postamble	
%}}} pdccmcode.dtx

%Local variables:
%fold-folded-p: t
%fill-prefix: "	"
%End: