path: root/Master/texmf-dist/source/startex/base/startex.dtx

\def \CodeVersion {1.04}
\def \CodeDate    {11th March 1999}
%
% \iffalse ^^A Comment out driver code, so LaTeX can read this file as it is.
%
%<*driver>
\documentclass{ltxdoc}
\usepackage{varioref,amssymbols}
\EnableCrossrefs         
\RecordChanges
\CodelineIndex
\renewcommand{\levelchar}{?}

\title{The \starTeX{} source code}
\author{Dag Langmyhr\\ Department of Informatics\\
  University of Oslo\\[3mm]
  \textsf{dag@ifi.uio.no}}%
\date{Version \CodeVersion, \CodeDate}

\newcommand{\hex}[1]{\texttt{``#1}}
\newcommand{\stcs}[1]{\texttt{<#1>}}
\newcommand{\stenv}[1]{\stcs{#1}\dots\hspace{0.1pt}\stcs{/#1}}
\newcommand{\ParX}{\texttt{\textit{x}}}
\newcommand{\ps}{\textsc{Post\-Script}}
\newcommand{\starTeX}{Star\TeX}
\newcommand{\xx}{\textit{x}}

\newfont{\StdECR}{ecrm1000}
\newfont{\CompECR}{tcrm1000}

\begin{document}
\maketitle

\DoNotIndex{\#,\$,\&,\\,\{,\},\^,\_,\~}
\DoNotIndex{\active,\advance,\alpha}
\DoNotIndex{\begingroup,\beta,\bgroup}
\DoNotIndex{\catcode,\Cdef,\chi,\csname}
\DoNotIndex{\day,\def,\Delta,\delta,\Df,\dots,\dump}
\DoNotIndex{\edef,\egroup,\endcsname,\endgraf,\endgroup,\epsilon,\eta}
\DoNotIndex{\Gamma,\gamma,\gdef,\global}
\DoNotIndex{\halign,\hangafter,\hangindent,\hbadness,\hbox,\hfil,\hfuzz}
\DoNotIndex{\hoffset,\hrule,\hsize}
\DoNotIndex{\iota,\input,\inputlineno}
\DoNotIndex{\kappa,\kern}
\DoNotIndex{\Lambda,\lambda,\language,\LaTeX,\leavevmode,\leftskip,\let,\llap}
\DoNotIndex{\magstep,\magstephalf,\mathcode,\message,\month,\mu}
\DoNotIndex{\newbox,\newcount,\newdimen,\newif,\newlinechar,\newread}
\DoNotIndex{\newskip,\newwrite,\next,\Next,\noalign,\nu}
\DoNotIndex{\obeylines,\obeyspaces,\offinterlineskip,\Omega,\omega}
\DoNotIndex{\openin,\openout,\over}
\DoNotIndex{\pageno,\par,\parindent,\parskip,\penalty,\Phi,\phi,\pm,\Pr}
\DoNotIndex{\Pi,\pi,\Psi,\psi}
\DoNotIndex{\raggedright,\Re,\relax,\rightskip,\rho,\rlap}
\DoNotIndex{\setbox,\Sigma,\sigma,\space,\string,\strut,\supereject}
\DoNotIndex{\tau,\TeX,\the,\Theta,\theta,\time,\times,\TmpA,\TmpB,\tolerance}
\DoNotIndex{\topinsert}
\DoNotIndex{\tracingcommands,\tracingmacros,\tracingoutput,\tracingrestores}
\DoNotIndex{\uccode,\unskip,\Uparrow,\Updownarrow,\uppercase,\Upsilon,\upsilon}
\DoNotIndex{\varepsilon,\varphi,\varpi,\varrho,\varsigma,\vartheta}
\DoNotIndex{\vbox,\vdots,\vfill,\voffset,\vrule,\vsize,\vskip}
\DoNotIndex{\wd}
\DoNotIndex{\Xi,\xi}
\DoNotIndex{\year}
\DoNotIndex{\zeta}

\DeleteShortVerb{\|}\MakeShortVerb{\"}
\DocInput{startex.dtx}
\end{document}
%</driver>
%
% \fi
% 
% \CheckSum{2563}
%% \CharacterTable
%%  {Upper-case    \A\B\C\D\E\F\G\H\I\J\K\L\M\N\O\P\Q\R\S\T\U\V\W\X\Y\Z
%%   Lower-case    \a\b\c\d\e\f\g\h\i\j\k\l\m\n\o\p\q\r\s\t\u\v\w\x\y\z
%%   Digits        \0\1\2\3\4\5\6\7\8\9
%%   Exclamation   \!     Double quote  \"     Hash (number) \#
%%   Dollar        \$     Percent       \%     Ampersand     \&
%%   Acute accent  \'     Left paren    \(     Right paren   \)
%%   Asterisk      \*     Plus          \+     Comma         \,
%%   Minus         \-     Point         \.     Solidus       \/
%%   Colon         \:     Semicolon     \;     Less than     \<
%%   Equals        \=     Greater than  \>     Question mark \?
%%   Commercial at \@     Left bracket  \[     Backslash     \\
%%   Right bracket \]     Circumflex    \^     Underscore    \_
%%   Grave accent  \`     Left brace    \{     Vertical bar  \|
%%   Right brace   \}     Tilde         \~}
%
% \tableofcontents
%
% \RecordChanges
% \StopEventually{\PrintIndex\PrintChanges}
%
% \section{The \starTeX{} format}
%    \begin{macrocode}
%<*code>
%    \end{macrocode}
% This file contains the documented source code of \starTeX{}, a
% simplified and more robust \TeX{} format intended for students
% writing their first report or essay. For more information on
% \starTeX{}, see the other files accompanying this.
% 
% The main guidelines for this implementation are:
% \begin{itemize}
% \item \starTeX{} is implemented on top of plain \TeX{}.
%
% \item All names defined in the \starTeX{} implementation contain
%   at least one uppercase letter, like \cs{Cdef} or \cs{NewEnvir}. 
%   This makes it easier to distinguish
%   them from \TeX's internal names and the names defined in plain
%   \TeX{}, nearly all of which consist of lowercase letters only.
%
% \item Most \TeX{} macro packages tend to use a rather terse
% programming style, like
% \begin{quote}
% \verb*:\advance\var\@ne:
% \end{quote}
% This improves processing speed and reduces storage, but make the
% code more difficult to read. Since this is a package which aims to
% be easily understandable and adaptable, I will use a more verbose style:
% \begin{quote}
% \verb*:\advance \var by 1:
% \end{quote}
% \end{itemize}
% To avoid possible confusion, \starTeX{} commands will be called
% \emph{commands} in this document, while \TeX{} commands will be
% called \emph{macros}.
%
% \subsection{States}
% \begin{macro}{\State}
% To avoid improper nesting, such as use of \stcs{psfig} within a
% \stcs{table}, we introduce a global state variable. Its values are
% given in table~\ref{tab:states}.
% \begin{table}
%   \begin{center}
%     \begin{tabular}{|c|l|}
%       \hline
%       \textbf{Value}& \textbf{State}\\
%       \hline\hline
%       \texttt{x}& Ordinary body text\\
%       \texttt{p}& Inside a \stcs{psfig} caption\\
%       \texttt{t}& Inside a \stcs{table} caption\\
%       \texttt{r}& Inside a \stcs{table} row\\
%       \hline
%     \end{tabular}
%   \end{center}
%   \caption{The possible values of \cs{State}}\label{tab:states}
% \end{table}
% The initial state is body text:
%    \begin{macrocode}
\let \State = x
%    \end{macrocode}
% \end{macro}
%
% \subsection{Command handling}
% \begin{macro}{\Command}
% \starTeX{} uses the syntax \stcs{\xx} for its commands rather than
% the usual \cs{\xx}. This is easily implemented by making "<" an
% active character calling the macro \cs{Command}. The command name is
% the text between the "<" and a subsequent ">". 
%
% \cs{Command} first checks whether the command immediately
% follows a "<define>" in which case it is being defined;
% \cs{DefineCmd} handles that. Otherwise, \cs{Command} converts the
% command name into lowercase and 
% checks whether it has been defined. If so, it is called;
% otherwise an error message is produced and the command is ignored.
%    \begin{macrocode}
\def \Command #1>{\ifDefining
    \def \Next {\lowercase{\DefineCmd{#1}}}%
  \else
    \lowercase{\expandafter\ifx \csname >#1\endcsname}\relax
      \Error{Unknown command <#1> ignored.}{}%
      \let \Next = \relax
    \else
      \def \Next {\lowercase{\csname >#1\endcsname}}%
    \fi
  \fi \Next }
%    \end{macrocode}
% \end{macro}
% The "<" character must be bound to \cs{Command}:
%    \begin{macrocode}
{\catcode `\< = \active  \global\let < = \Command }
%    \end{macrocode}
%
% \subsubsection{Internal form of command names}
% All \starTeX{} commands are represented by a \TeX{} macro whose name
% is constructed by \cs{csname}. To avoid confusion with predefined
% \TeX{} macros, an initial ">"\footnote{The character \texttt{>} was
% chosen as it cannot possibly occur in a command name.}
% is inserted; for example, the name of \stcs{h1} is
% represented by "\csname >h1\endcsname".
%
% \begin{macro}{\Cdef}
% To simplify the declaration of \starTeX{} commands, the macro
% \cs{Cdef} is defined. It takes two parameters: "#1" is the command name
% (with no uppercase letters and no angle brackets), and
% "#2" is the command definition. 
%    \begin{macrocode}
\def \Cdef #1#2{\expandafter\gdef \csname >#1\endcsname{#2}}
%    \end{macrocode}
% \end{macro}
% \begin{macro}{\Ccall}
% In case we need to call \starTeX{} commands from \TeX{} code, a
% \cs{Ccall} macro is introduced.
%    \begin{macrocode}
\def \Ccall #1{\csname >#1\endcsname}
%    \end{macrocode}
% \end{macro}
%
% \subsubsection{User-defined commands}
% \begin{macro}{<define>}
% \begin{macro}{\ifDefining}
% The implementation of \stcs{define} is a little sneaky; it just sets
% a flag \cs{ifDefining}. When the following command name is
% found, the \cs{Command} routine checks the state of the flag and
% calls \cs{DefineCmd} when the flag is raised.\footnote{This means that
%   the user can insert text between the \stcs{define} and the command
%   name, and this will be typeset as normal text. There is a
%   slight chance that this might confuse some users, but I prefer this
%   solution to using much more complicated code.}
%    \begin{macrocode}
\newif \ifDefining
\Cdef {define}{\Definingtrue}
%    \end{macrocode}
% \end{macro}
% \end{macro}
%
% \begin{macro}{\DefineCmd}
% The command \cs{DefineCmd} is used to define a new user command;
% "#1" is the new command name (already translated into lowercase) and
% "#2" is the definition.
% The main job of \cs{DefineCmd} is to check whether the definition is
% legal. There are two reasons why it may not be legal:
% \begin{itemize}
% \item The definition may be nested, as in
% \begin{quote}
%   "<define><comm-1>...<define><comm-2>..."
% \end{quote}
%   This is checked using a counter \cs{CallLevel} which is
%   incremented whenever a user-defined command is called, and
%   decremented on return; see the definition of \cs{Call} and
%   \cs{Return} below.
%
% \item The command may already be defined.
% \end{itemize}
% If the definition is legal, \cs{FetchDef} is called to perform the
% actual defining.
%    \begin{macrocode}
\def \DefineCmd #1{\Definingfalse
  \ifnum \CallLevel>0
    \Error{Nested definitions are not allowed;}%
      {the definition of <#1> is ignored.}\let \Next = \relax
  \else
    \expandafter\ifx \csname>#1\endcsname\relax
      \def \Next {\begingroup \catcode`\^^M = 12 \FetchDef{#1}}%
    \else
      \Error{Command <#1> already defined;}{this definition ignored.}%
      \def \Next {\begingroup \catcode`\^^M = 12 \IgnoreDef{#1}}%
    \fi
  \fi \Next }
%    \end{macrocode}
% \end{macro}
% \begin{macro}{\FetchDef}
% \cs{FetchDef} defines the macro by using everything up to the
% end-of-line mark as the definition. This implies modifying the
% \cs{catcode} of the end-of-line character ("^^M"). Also note
% that calls on \cs{Call} and \cs{Return} are inserted into the
% definition. 
%    \begin{macrocode}
\begingroup \catcode`\^^M = 12
  \gdef \FetchDef #1#2^^M{\expandafter%
    \gdef\csname>#1\endcsname {\Call #2\Return }\endgroup }%
%    \end{macrocode}
% \end{macro}
% \begin{macro}{\IgnoreDef}
% I also define \cs{IgnoreDef} which is quite similar to
% \cs{FetchDef}, but no command is 
% defined. It is used to ignore the rest of the line in case
% the user tries an illegal definition.
%    \begin{macrocode}
  \gdef \IgnoreDef #1#2^^M{\endgroup }%
\endgroup
%    \end{macrocode}
% \end{macro}
% \begin{macro}{\CallLevel}
% Finally, we must declare the user command level counter
% \cs{CallLevel}: 
%    \begin{macrocode}
\newcount \CallLevel
%    \end{macrocode}
% \end{macro}
% \begin{macro}{\Call}
% \begin{macro}{\Return}
% We must also define the two commands used for incrementing and
% decrementing the counter:
%    \begin{macrocode}
\def \Call   {\global\advance \CallLevel by  1 }
\def \Return {\global\advance \CallLevel by -1 }
%    \end{macrocode}
% \end{macro}
% \end{macro}
%
% \subsubsection{Catcode modifications}
% \starTeX{} uses only one special character: "<", which must be
% active. All the other special characters of \TeX{} and \LaTeX{} are
% assigned suitable \cs{catcode}s turning them into ordinary
% characters.
%
% \begin{macro}{\SpecialCatCodes}
% \cs{SpecialCatCodes} sets the \cs{catcode}s to the \starTeX{}
% values: 
%    \begin{macrocode}
\def \SpecialCatCodes {%
  \catcode `\\ = 12  \catcode `\{ = 12  \catcode `\} = 12
  \catcode `\$ = 12  \catcode `\& = 12  \catcode `\# = 12
  \catcode `\^ = 12  \catcode `\_ = 12  \catcode `\~ = 12
  \catcode `\% = 12  \catcode `\< = \active }
%    \end{macrocode}
% \end{macro}
% \begin{macro}{\StandardCatCodes}
% \cs{StandardCatCodes} restores the \cs{catcode}s to their normal
% values. This is necessary when reading for instance style files.
%    \begin{macrocode}
\def \StandardCatCodes {%
  \catcode `\\ = 0  \catcode `\{ = 1  \catcode `\} = 2
  \catcode `\$ = 3  \catcode `\& = 4  \catcode `\# = 6
  \catcode `\^ = 7  \catcode `\_ = 8  \catcode `\~ = \active
  \catcode `\% = 14 \catcode `\< = 12 }
%    \end{macrocode}
% \end{macro}
%
% \subsubsection{Environments}
% A \starTeX{} environment is a piece of text enclosed in a
% \stcs{\xx}\dots\stcs{/\xx} pair. 
%
% \begin{macro}{\NewEnvir}
% \cs{NewEnvir} is used to start a new environment. It has three
% parameters: "#1" is the name of the environment, "#2" is
% the \TeX{} commands used to start the new environment (usually just
% a \cs{begingroup}), and "#3" is
% the \TeX{} commands used to terminate the environment (usually just
% a \cs{endgroup}). 
%
% \begin{macro}{\PrevEnv}
% \begin{macro}{\PrevEnvLine}
% \begin{macro}{\CurEnv}
% \begin{macro}{\CurEnvExit}
% \begin{macro}{\CurEnvLine}
% \cs{NewEnvir} saves information about the current environment's name
% (in \cs{CurEnv}), the line on which it starts (in \cs{CurEnvLine}) and
% which command is used to exit it (in \cs{CurEnvExit}). It also keeps
% tracks of the outer environment's name (in \cs{PrevEnv}) and start line
% (in \cs{PrevEnvLine}) for better error reporting and error recovery.
%    \begin{macrocode}
\def \NewEnvir #1#2#3{#2\relax
  \let \PrevEnv = \CurEnv  \PrevEnvLine = \CurEnvLine
  \def \CurEnv {#1}\def \CurEnvExit {#3}\CurEnvLine = \inputlineno }
%    \end{macrocode}
% We need default definitions of \cs{CurEnv} and \cs{CurEnvExit}: 
%    \begin{macrocode}
\def \CurEnv {}\def \CurEnvExit {\relax}
%    \end{macrocode}
% We also need to declare the two line counters:
%    \begin{macrocode}
\newcount \CurEnvLine  \newcount \PrevEnvLine
%    \end{macrocode}
% \end{macro}
% \end{macro}
% \end{macro}
% \end{macro}
% \end{macro}
% \end{macro}
% 
% \begin{macro}{\EndEnvir}
% \cs{EndEnvir} is used to terminate an environment. It check that the
% correct environment is terminated, and tries to correct user
% mistakes. It checks for the following situations:
% \begin{itemize}
% \item If the name of the environment to be terminated is the same as
% that of the current environment, everything is OK, and we can safely
% leave the current environment.
%
% \item If the name of the terminated environment is not the name of
% the current one, but matches the name of the outer environment,
% we assume that the user has forgotten a \stcs{/\xx} command.
% An error message is given, and both the current environment and the
% outer one are terminated.
%
% \item If the name of the terminated environment matches neither the
% current nor the outer environment, we assume that the user has just
% misspelled the command. The best thing we can do in this case is to
% give an error message and ignore the command. If the user had
% intended to terminate the current environment, we get erroneous
% processing of the following text, but the situation will normalize
% when the outer environment is terminated. 
% \end{itemize}
%    \begin{macrocode}
\def \EndEnvir #1{%
  \ifTextEqual{#1}{\CurEnv}\let \Next = \CurEnvExit
  \else \EnvirError{#1}\fi
  \Next }
%    \end{macrocode}
% \end{macro}
% \begin{macro}{\EnvirError}
% \cs{EnvirError} is an auxiliary command giving a proper error
% message and placing---in \cs{Next}---the best command to recover
% from the error.
%    \begin{macrocode}
\def \EnvirError #1{\ifTextEqual{#1}{\PrevEnv}%
    \Error{<\CurEnv> on line \the\CurEnvLine\space terminated by
      </#1>.}{An extra </\CurEnv> has been inserted.}%
    \def \Next {\CurEnvExit \CurEnvExit }%
  \else
    \Error{<\CurEnv> on line \the\CurEnvLine\space terminated by 
      </#1>.}{The </#1> will be ignored.}%
    \let \Next = \relax
  \fi }
%    \end{macrocode}
% \end{macro}
%
% \subsection{Document styles}
% \begin{macro}{<style>}
% The command \stcs{style} is called to read a document style. It
% calls \cs{ReadStyle} to do the actual work.
%    \begin{macrocode}
\Cdef {style}{\IfNextChar{[}{\ReadStyle}%
  {\Error{No style name given;}{the syntax is: <style>[style name].}}}
%    \end{macrocode}
% \end{macro}
% \begin{macro}{\ReadStyle}
% \begin{macro}{\StyleLine}
% \cs{ReadStyle} reads the style file (with extension
% \texttt{.stx}) using the standard \TeX{} \cs{catcode}s. Afterwards, the
% \stcs{style} command is redefined, since \stcs{style} should be
% called once only. Also, the current line number is saved (in
% \cs{StyleLine}) in case it is needed later in an error message.
%    \begin{macrocode}
\def \ReadStyle [#1]{\IfFileExists{#1.stx}%
    {\edef \StyleLine {\the\inputlineno}%
     \Cdef {style}{\Error{Command <style> already used on line 
                     \StyleLine;}{this use of <style> is ignored.}}%
     \StandardCatCodes \input #1.stx
     \SpecialCatCodes }%
    {\Error{Style file `#1.stx' could not be found;}%
           {style definition ignored.}}}
%    \end{macrocode}
% \end{macro}
% \end{macro}
%
% \subsection{Fonts}
% \starTeX{} allows the user a limited amount of font selection:
% \begin{itemize}
% \item use of \textbf{bold text} using \stcs{b}\dots\stcs{/b},
% 
% \item use of \textit{italic text} using \stcs{i}\dots\stcs{/i}, and
%
% \item use of \texttt{typewriter text} using \stcs{tt}\dots\stcs{/tt}.
% \end{itemize}
% All other font changes, including font size changes, are done by the
% document style.
%
% Since the font change commands mentioned above may be combined,
% \starTeX{} uses an internal font naming scheme in which a font name
% consists of five parts, as shown in table\vref{tab:fontname}. For
% example, the font ``Standard bold roman italic normal-size'' is
% represented internally as "F-RSBIN".
%
% \begin{table}
%   \begin{center}
%     \begin{tabular}{|l|l|l|l|l|}
%       \hline
%       \textbf{Kind}& \textbf{Alternate}& \textbf{Bold}& 
%	  \textbf{Italic}& \textbf{Size}\\
%       \hline\hline
%       "R" (Roman)& "S" (Standard)& "M" (Medium)& "U" (Upright)&
%         "X" (Extra large)\\
%       "T" (Teletype)& "A" (Alternate)& "B" (Bold)& "I" (Italic)&
%         "L" (Large)\\ 
%       &&&& "N" (Normal)\\
%       &&&& "S" (Small)\\
%       \hline
%     \end{tabular}
%   \end{center}
%   \caption{The five parts of the internal \starTeX{} text font name}
%   \label{tab:fontname}
% \end{table}
%
% \begin{table}
%   \begin{center}
%     \begin{tabular}{|l|l|l|}
%       \hline
%       \textbf{Size}& \textbf{Position}& \textbf{Family}\\
%       \hline\hline
%       "X" (Extra large)& "N" (Normal)& "0" (Roman math)\\
%       "L" (Large)& "S" (Script)& "1" (Italic math)\\
%       "N" (Normal)& "X" (Scriptscript)& "2" (Math symbols)\\
%       "S" (Small)& & "3" (Large math symbols)\\
%       & & "4" (Bold math)\\
%       & & "5" (Extended italic math)\\
%       & & "6" (AMS symbols B)\\
%       \hline
%     \end{tabular}
%   \end{center}
%   \caption{The three parts of the internal \starTeX{} math font
%   name}\label{tab:mathfont}
% \end{table}
%
% \subsubsection{Font definitions}
% \begin{macro}{\FontDef}
% The macro \cs{FontDef} is used to define a \starTeX{} font pair. To be
% more precise, it defines which \TeX{} font specifications to use for
% each internal \starTeX{} font name. "#1" gives the font kind, "#2"
% the boldness, "#3" whether the font is italic or not, and "#4" the font
% size. Two fonts are defined for each
% internal \starTeX{} name: the standard Cork encoded font (in "#5")
% and the additional companion font (in "#6"). 
%    \begin{macrocode}
\def \FontDef #1#2#3#4#5#6{%
  \expandafter\def \csname F-#1S#2#3#4\endcsname {#5}%
  \expandafter\def \csname F-#1A#2#3#4\endcsname {#6}}
%    \end{macrocode}
% \end{macro}
% \begin{macro}{\MathFontDef}
% Similarly, \cs{MathFontDef} is used to define a math font. "#1" is
% the size, "#2" is the family, and "#3"--"#5" are the three fonts to
% be used as math text font, script font and scriptscript font.
%    \begin{macrocode}
\def \MathFontDef #1#2#3#4#5{%
  \expandafter\def \csname M-#1N#2\endcsname {#3}%
  \expandafter\def \csname M-#1S#2\endcsname {#4}%
  \expandafter\def \csname M-#1X#2\endcsname {#5}}
%    \end{macrocode}
% \end{macro}
%
% \subsubsection{Predefined fonts}
% The following fonts are predefined, which means that they are the
% fonts the user will get unless the document style dictates
% otherwise.
% 
% The fonts used are the so-called \emph{EC fonts}.
% Since there are no bold typewriter characters in this set,
% ordinary typewriter characters are used instead.
%
% Also defined are various size dependant values:
% \begin{description}
% \item[\cs{LineSkip}\ParX{}] line skip, and 
%
% \item[\cs{CodeSkip}\ParX{}] line skip in code listings.
% \end{description}
%
% \paragraph{Normal size fonts}
% An 11 point (actually 10.95 pt) font is the standard text font.
% \changes{1.02}{1998/01/12}{Changed from `dc' to `ec' fonts.}
%    \begin{macrocode}
\def \XIpt { at 10.95pt}
\FontDef{R}{M}{U}{N}{ecrm1095}{tcrm1095}  
\FontDef{R}{M}{I}{N}{ecti1095}{tcti1095}
\FontDef{R}{B}{U}{N}{ecbx1095}{tcbx1095}
\FontDef{R}{B}{I}{N}{ecbi1095}{tcbi1095}
\FontDef{T}{M}{U}{N}{ectt1095}{tctt1095}
\FontDef{T}{M}{I}{N}{ecit1095}{tcit1095}
\FontDef{T}{B}{U}{N}{ectt1095}{tctt1095}
\FontDef{T}{B}{I}{N}{ecit1095}{tcit1095}
\MathFontDef{N}{0}{cmr10\XIpt}{cmr8}{cmr6}
\MathFontDef{N}{1}{cmmi10\XIpt}{cmmi8}{cmmi6}
\MathFontDef{N}{2}{cmsy10\XIpt}{cmsy8}{cmsy6}
\MathFontDef{N}{3}{cmex10\XIpt}{cmex10\XIpt}{cmex10\XIpt}
\MathFontDef{N}{4}{cmbx10\XIpt}{cmbx8}{cmbx6}
\MathFontDef{N}{5}{ecti1095}{ecti0800}{ecti0600}
\MathFontDef{N}{6}{msbm10\XIpt}{msbm8}{msbm6}
\def \LineSkipN {12pt}  \def \CodeSkipN {11pt plus 0.0pt minus 0.1pt}
%    \end{macrocode}
%
% \paragraph{Large fonts}
% The slightly larger characters are 14.4 pt.
%    \begin{macrocode}
\def \XIVpt { at 14.4pt}
\FontDef{R}{M}{U}{L}{ecrm1440}{tcrm1440}  
\FontDef{R}{M}{I}{L}{ecti1440}{tcti1440}
\FontDef{R}{B}{U}{L}{ecbx1440}{tcbx1440}
\FontDef{R}{B}{I}{L}{ecbi1440}{tcbi1440}
\FontDef{T}{M}{U}{L}{ectt1440}{tctt1440}
\FontDef{T}{M}{I}{L}{ecit1440}{tcit1440}
\FontDef{T}{B}{U}{L}{ectt1440}{tctt1440}
\FontDef{T}{B}{I}{L}{ecit1440}{tcit1440}
\MathFontDef{L}{0}{cmr12\XIVpt}{cmr10}{cmr8}
\MathFontDef{L}{1}{cmmi12\XIVpt}{cmmi10}{cmmi8}
\MathFontDef{L}{2}{cmsy10\XIVpt}{cmsy10}{cmsy8}
\MathFontDef{L}{3}{cmex10\XIVpt}{cmex10\XIVpt}{cmex10\XIVpt}
\MathFontDef{L}{4}{cmbx12\XIVpt}{cmbx10}{cmbx8}
\MathFontDef{L}{5}{ecti1440}{ecti1000}{ecti0800}
\MathFontDef{L}{6}{msbm10\XIVpt}{msbm10}{msbm8}
\def \LineSkipL {15pt}
%    \end{macrocode}
%
% \paragraph{The extra large fonts}
% The largest fonts are 17.28~pt.
%    \begin{macrocode}
\def \XVIIpt { at 17.28pt}
\FontDef{R}{M}{U}{X}{ecrm1728}{tcrm1728}  
\FontDef{R}{M}{I}{X}{ecti1728}{tcti1728}
\FontDef{R}{B}{U}{X}{ecbx1728}{tcbx1728}
\FontDef{R}{B}{I}{X}{ecbi1728}{tcbi1728}
\FontDef{T}{M}{U}{X}{ectt1728}{tctt1728}
\FontDef{T}{M}{I}{X}{ecit1728}{tcit1728}
\FontDef{T}{B}{U}{X}{ectt1728}{tctt1728}
\FontDef{T}{B}{I}{X}{ecit1728}{tcit1728}
\MathFontDef{X}{0}{cmr17}{cmr12\XIVpt}{cmr9}
\MathFontDef{X}{1}{cmmi12\XVIIpt}{cmmi12}{cmmi9}
\MathFontDef{X}{2}{cmsy10\XVIIpt}{cmsy10 at 12pt}{cmsy9}
\MathFontDef{X}{3}{cmex10\XVIIpt}{cmex10\XVIIpt}{cmex10\XVIIpt}
\MathFontDef{X}{4}{cmbx12\XVIIpt}{cmbx12}{cmbx9}
\MathFontDef{X}{5}{ecti1728}{ecti1200}{ecti0900}
\MathFontDef{X}{6}{msbm10\XVIIpt}{msbm10 at 12pt}{msbm9}
\def \LineSkipX {17pt}
%    \end{macrocode}
%
% \paragraph{Small fonts}
% A smaller font for footnotes and other special text is 10~pt.
%    \begin{macrocode}
\FontDef{R}{M}{U}{S}{ecrm1000}{tcrm1000}  
\FontDef{R}{M}{I}{S}{ecti1000}{tcti1000}
\FontDef{R}{B}{U}{S}{ecbx1000}{tcbx1000}
\FontDef{R}{B}{I}{S}{ecbi1000}{tcbi1000}
\FontDef{T}{M}{U}{S}{ectt1000}{tctt1000}
\FontDef{T}{M}{I}{S}{ecit1000}{tcit1000}
\FontDef{T}{B}{U}{S}{ectt1000}{tctt1000}
\FontDef{T}{B}{I}{S}{ecit1000}{tcit1000}
\MathFontDef{S}{0}{cmr10}{cmr7}{cmr5}
\MathFontDef{S}{1}{cmmi10}{cmmi7}{cmmi5}
\MathFontDef{S}{2}{cmsy10}{cmsy7}{cmsy5}
\MathFontDef{S}{3}{cmex10}{cmex10}{cmex10}
\MathFontDef{S}{4}{cmbx10}{cmbx7}{cmbx5}
\MathFontDef{S}{5}{ecti1000}{ecti0700}{ecti0500}
\MathFontDef{S}{6}{msbm10}{msbm7}{msbm5}
\def \LineSkipS {11pt}
%    \end{macrocode}
%
% \subsubsection{Font selection}
% \begin{macro}{\FontAlt}
% \begin{macro}{\FontBold}
% \begin{macro}{\FontItal}
% \begin{macro}{\FontKind}
% \begin{macro}{\FontSize}
% Selecting a \starTeX{} font is a two-step procedure:
% \begin{enumerate}
% \item Define \cs{FontKind}, \cs{FontAlt}, \cs{FontBold},
%   \cs{FontItal} and 
%   \cs{FontSize} to contain the correct letters, as shown in
%   table\vref{tab:fontname}. 
%
% \item Call \cs{SelectFont}.
% \end{enumerate}
% (The reason for this procedure is to avoid loading unused fonts.) 
% \end{macro}
% \end{macro}
% \end{macro}
% \end{macro}
% \end{macro}
%
% \begin{macro}{\SelectFont}
% As mentioned, "\SelectFont" selects a font according to the
% specifications given.
%    \begin{macrocode}
\def \SelectFont {%
  \edef \ThisFont {F-\FontKind\FontAlt\FontBold\FontItal\FontSize}%
  \font \CurFont = \csname\ThisFont\endcsname \CurFont }
%    \end{macrocode}
% \end{macro}
%
% \begin{macro}{\ResetFont}
% \cs{ResetFont} provides standard settings for the various \starTeX{}
% font parameters.
%    \begin{macrocode}
\def \ResetFont {%
  \def \FontKind{R}\def \FontAlt{S}\def \FontBold{M}%
  \def \FontItal{U}\def \FontSize{N}}
%    \end{macrocode}
% \end{macro}
% \begin{macro}{\CheckItCorr}
% \cs{CheckItCorr} is called after every group of italic text. Unless
% the following character is a low character like `\texttt{.}' or
% `\texttt{,}', an italic correction command \cs{/} is inserted.
%    \begin{macrocode}
\def \CheckItCorr {%
  \IfNextCharTwo{.}{,}{}{\/}}
%    \end{macrocode}
% \end{macro}
% 
% \subsubsection{Font size selection}
% \begin{macro}{\SetSize}
% \cs{SetSize} is used to set various parameters connected with a size
% change. It also contains a call on \cs{SelectFont}.
%    \begin{macrocode}
\def \SetSize #1{\def \FontSize{#1}\SelectFont
  \baselineskip = \csname LineSkip#1\endcsname
  \ParIndent = \StdParIndent }
%    \end{macrocode}
% \end{macro}
%
% \subsubsection{\starTeX{} commands for font changes}
% \begin{macro}{<b>}
% \begin{macro}{</b>}
% \begin{macro}{<i>}
% \begin{macro}{</i>}
% \begin{macro}{<tt>}
% \begin{macro}{</tt>}
% The following commands are used to indicate bold, italic, or
% typewriter text. 
%    \begin{macrocode}
\Cdef {b}{\NewEnvir{b}{\begingroup}{\endgroup}%
  \def\FontBold{B}\SelectFont}
\Cdef {/b}{\EndEnvir{b}}
\Cdef {i}{\NewEnvir{i}{\begingroup}{\endgroup\CheckItCorr}%
  \def\FontItal{I}\SelectFont}
\Cdef {/i}{\EndEnvir{i}}
\Cdef {tt}{\NewEnvir{tt}{\begingroup}{\endgroup}%
  \def\FontKind{T}\SelectFont}
\Cdef {/tt}{\EndEnvir{tt}}
%    \end{macrocode}
% \end{macro}
% \end{macro}
% \end{macro}
% \end{macro}
% \end{macro}
% \end{macro}

% \subsection{Body text}
% A \starTeX{} file is enclosed in a \stenv{body} environment. The main
% reason for this is to be able to detect the end of the document and
% terminate properly.\footnote{An \cs{everyeof}
% command---had it existed---would have made the \stenv{body} environment
% superfluous, thus removing yet another possible error source. 
% ($\varepsilon$-\TeX\cite{etex} does provide such a primitive, but
% I have chosen not to apply it as use of $\varepsilon$-\TeX{} is not
% yet very widespread.)} It is also used to print an introductory
% message.
%
% \begin{macro}{<body>}
% \stcs{body} just redefines itself to produce an error message if
% used a second time; this is to ensure that it is only used once.
%    \begin{macrocode}
\Cdef {body}{\NewEnvir{body}{\begingroup}{\endgroup}%
  \message{^^JThis is StarTeX, version \CodeVersion^^J}%
  \global\everypar = {\NewPar }%
  \xdef \BodyLine {\the\inputlineno}%
  \Cdef {body}{\Error{Command <body> already used on line \BodyLine;}%
    {this use of <body> was ignored.}}}
%    \end{macrocode}
% \end{macro}
% \begin{macro}{\BodyError}
% If the user has forgotten the \stenv{body} environment, the first
% text paragraph will trigger the \cs{BodyError} macro.
%    \begin{macrocode}
\def \BodyError {\message{^^JThis is StarTeX, version \CodeVersion^^J}%
  \Error{A <body>...</body> environment should surround
    the whole document.}{A missing <body> was inserted.}%
  \global\everypar = {\NewPar }}
%    \end{macrocode}
% \end{macro}
% \begin{macro}{</body>}
% \stcs{/body} is used to terminate the document. This involves two things:
% \begin{enumerate}
% \item the current page must be terminated, and
%
% \item the \texttt{.aux} file must be check to determine whether any
%   references have changed.
%   The \texttt{.aux} file must be checked if both
%   \begin{itemize}
%   \item an \texttt{.aux} file has been read (\cs{ifAuxRead} is true),
%     and 
%
%   \item a new \texttt{.aux} file has been created (\cs{ifAuxOpen} is
%     true). 
%   \end{itemize}
%   If we already know that a rerun is necessary (because \cs{ifRerun}
%   is true), there is no need to check the \texttt{.aux} file.
% \end{enumerate}
%    \begin{macrocode}
\Cdef {/body}{\EndEnvir{body}\endgraf\vfill\supereject
  \let \Next = \CheckAux
  \ifRerun \let \Next = \relax \fi
  \ifAuxRead \else \let \Next = \relax \fi
  \ifAuxOpen \else \let \Next = \relax \fi
  \Next
%    \end{macrocode}
% Give a warning if rerunning \starTeX{} is necessary.
%    \begin{macrocode}
  \ifRerun \Warning{Cross-references are not correct;}%
    {please run StarTeX again.}\fi
%    \end{macrocode}
% And now, we can finally terminate the run.
%    \begin{macrocode}
  \end}
%    \end{macrocode}
% \end{macro}
%
% \subsubsection{Paragraph breaking}
% \begin{macro}{<p>}
% The \starTeX{} command \stcs{p} is used to separate paragraphs:
%    \begin{macrocode}
\Cdef {p}{\endgraf}
%    \end{macrocode}
% \end{macro}
% Blank lines are just ignored:
%    \begin{macrocode}
\def \par {}
%    \end{macrocode}
% \starTeX{} uses parameters which allow rather sloppy paragraph
% setting: 
%    \begin{macrocode}
\pretolerance = 2500 \tolerance = 9999  \hbadness = 10000
\emergencystretch = 3cm
%    \end{macrocode}
% The reasons for this are:
% \begin{itemize}
% \item If a good set of breaks is possible, \TeX{} will still choose
%   that. 
%
% \item If no good breaks are possible, it is better to have a sloppily
%   set paragraph than problem lines sticking into the margin. The
%   potential users of \starTeX{} tend to just ignore messages about
%   overfull boxes.
% \end{itemize}
% Note, however, that \cs{hfuzz} has not been increased, as such a
% change is immediately visible and gives a poor impression.
%
% \subsubsection{Paragraph indentation}
% The paragraph handling mechanism must be able to let some paragraphs
% be indented and others not; for instance, the first paragraph
% following a title should not be indented, while the subsequent
% ones should be.
%
% In \starTeX{} this is solved by disabling the original
% \cs{parindent} and instead insert a call on \cs{NewPar} at
% the start of every paragraph. However, \cs{everypar} is initialized
% to \cs{BodyError} to detect a missing \stenv{body} environment; the
% \stcs{body} command will set \cs{everypar} correctly.
%    \begin{macrocode}
\parindent = 0pt \everypar = {\BodyError }
%    \end{macrocode}
% \begin{macro}{\NewPar}
% This \cs{NewPar} inserts an indentation \cs{ParIndent} wide if
% \cs{ifIndent} is true.\footnote{We cannot use the standard
% \cs{parindent} to this, as that is inserted when the
% paragraph is started, in other words before \cs{NewPar} is
% called.} It also sets \cs{ifIndent} to true, ensuring the
% indentation of subsequent paragraphs. \cs{NewPar} sets \cs{parskip}
% to its standard current value \cs{CurParSkip}.
%    \begin{macrocode}
\def \NewPar {\ifIndent \kern \ParIndent \fi  \Indenttrue 
  \global\parskip = \CurParSkip }
%    \end{macrocode}
% \end{macro}
% (In some environments, like \stenv{list}, no paragraphs are
% indented. By setting \cs{ParIndent} to 0~pt, the whole paragraph
% indentation mechanism is disabled.)
%
% \begin{macro}{\ParIndent}
% \begin{macro}{\ifIndent}
% \begin{macro}{\StdParIndent}
% We must declare \cs{ParIndent} and \cs{ifIndent}, and also
% \cs{StdParIndent} which holds the value to which \cs{ParIndent} is
% set when it is not 0:
%    \begin{macrocode}
\newskip \ParIndent  \def \StdParIndent {1em}
\newif \ifIndent
%    \end{macrocode}
% \end{macro}
% \end{macro}
% \end{macro}
%
% \subsubsection{Paragraph separation}
% Sometimes it is necessary to omit the paragraph
% separation (the vertical space inserted between every paragraph),
% for instance after having inserted some special vertical space. This
% is done by setting \cs{parskip} to 0~pt. The following paragraph
% will reset this to \cs{CurParSkip} so that subsequent paragraphs
% will have the standard separation; see the definition of \cs{NewPar}
% above. 
%
% \begin{macro}{\CurParSkip}
% \cs{CurParSkip} must be defined:
%    \begin{macrocode}
\newskip \CurParSkip
%    \end{macrocode}
% \end{macro}
%
% \begin{macro}{\AddVspace}
% Several insertions of vertical space (with \cs{vskip}) should not
% just accumulate; this will lead to excessive vertical spacing.
% Instead, \cs{AddVspace} should be used; it inserts the maximum of
% the last space (if any) and the current one, and it also sets
% \cs{parskip} to 0~pt so that the automatic paragraph separation will
% not interfere.
%    \begin{macrocode}
\def \AddVspace #1{\ifvmode \else \endgraf \fi
  \skip1 = #1\relax
  \ifdim \lastskip < \skip1 \relax
    \ifdim \lastskip > 0pt \vskip -\lastskip \fi
    \vskip \skip1
  \fi \parskip = 0pt \relax }
%    \end{macrocode}
% \end{macro}
% 
% \subsubsection{The document head}
% This section implements the various commands used at the start of
% the document.
%
% \begin{macro}{<title>}
% \begin{macro}{</title>}
% \stenv{title} is used to set the document title. It is centered, and
% uses the biggest fonts available.
%    \begin{macrocode}
\Cdef {title}{\AddVspace{30pt plus 10pt}
  \NewEnvir{title}{\begingroup}{\endgraf\endgroup}
  \leftskip = 2cm plus 1fill  \rightskip = \leftskip
  \ParIndent = 0pt  \CurParSkip = 0pt  
  \ResetFont \SetSize{X}}
\Cdef {/title}{\EndEnvir{title}%
  \AddVspace{20pt plus 4pt}}
%    \end{macrocode}
% \end{macro}
% \end{macro}
% \begin{macro}{<author>}
% \begin{macro}{</author>}
% \stenv{author} is used for the author's name. It is set like
% \stenv{title}, but uses a smaller font.
%    \begin{macrocode}
\Cdef {author}{\AddVspace{10pt plus 3pt}
  \NewEnvir{author}{\begingroup}{\endgraf\endgroup}
  \leftskip = 2cm plus 1fill  \rightskip = \leftskip
  \ParIndent = 0pt  \CurParSkip = 0pt
  \ResetFont \SetSize{L}}
\Cdef {/author}{\EndEnvir{author}%
  \AddVspace{20pt plus 4pt}}
%    \end{macrocode}
% \end{macro}
% \end{macro}
% \begin{macro}{<info>}
% \begin{macro}{</info>}
% \stenv{info} is used for additional information, like the date. It
% is centered, but uses the normal text font.
%    \begin{macrocode}
\Cdef {info}{\AddVspace{10pt plus 3pt}
  \NewEnvir{info}{\begingroup}{\endgraf\endgroup}
  \leftskip = 2cm plus 1fill  \rightskip = \leftskip
  \ParIndent = 0pt  \CurParSkip = 0pt
  \ResetFont \SetSize{N}}
\Cdef {/info}{\EndEnvir{info}%
  \AddVspace{20pt plus 4pt}}
%    \end{macrocode}
% \end{macro}
% \end{macro}
% \begin{macro}{<abstract>}
% \begin{macro}{</abstract>}
% \stenv{abstract} is used for the short abstract usually given with
% each document. It is set with indented margins, and the type is
% smaller than the text type. The word `Abstract' is set in boldface.
%    \begin{macrocode}
\Cdef {abstract}{\AddVspace{10pt plus 3pt}
  \NewEnvir{abstract}{\begingroup}{\endgraf\endgroup}
  \ResetFont \def \FontBold{B} \SetSize{S}  \CurParSkip = 0pt  
  \leftskip = 2cm  \rightskip = \leftskip  \Indentfalse
  \centerline{\AbstractName}
  \ResetFont \SetSize{S}\Indentfalse }
\Cdef {/abstract}{\EndEnvir{abstract}%
  \AddVspace{10pt plus 2pt}}
%    \end{macrocode}
% \end{macro}
% \end{macro}
%
% \subsubsection{Sectioning commands}
% To keep track of the section numbers, we need four counters:
%    \begin{macrocode}
\newcount \SectI    \newcount \SectII
\newcount \SectIII  \newcount \SectIV
%    \end{macrocode}
% \begin{macro}{\Heading}
% Since most of the work of a sectioning command is the same
% irrespective of its level, the common code has been place in
% \cs{Heading} which has four parameters: "#1" is the vertical space
% before the section, "#2" is the font size (\texttt{N}, \texttt{L} or
% \texttt{X}), "#3" tells whether to use bold type or not (\texttt{B}
% or \texttt{M}), and "#4" is the \TeX{} command to produce the
% section number. 
%    \begin{macrocode}
\def \Heading #1#2#3#4{\AddVspace{#1}
  \def \FontBold {#3} \SetSize{#2}
  \setbox0 = \hbox{#4\kern 0.5\baselineskip}
  \hangindent = \wd0  \hangafter = 1  \raggedright
  \ParIndent = 0pt  \CurParSkip = 0pt  \leavevmode  \box0 }
%    \end{macrocode}
% \end{macro}
%
% \begin{macro}{<h1>}
% \stcs{H1} must update the counters and add vertical space. The rest
% of the work is done by \cs{Heading}.
%    \begin{macrocode}
\Cdef {h1}{\endgraf
  \global\advance \SectI by 1 
  \global\SectII = 0  \global\SectIII = 0  \global\SectIV = 0
  \edef \CurDef {\SectIim}
  \NewEnvir{h1}{\begingroup}{\endgroup}
  \Heading{24pt plus 12pt minus 3pt}{X}{B}{\SectIim}}
%    \end{macrocode}
% \end{macro}
% \begin{macro}{</h1>}
% \stcs{/H1} terminates the heading. It adds some vertical space
% (protcted with a top \cs{penalty} so that a page break will never
% occur) and sets \cs{Indentfalse} to prevent indentation of the
% following line.
%    \begin{macrocode}
\Cdef {/h1}{\endgraf
  \nobreak\vskip 6pt plus 1.5pt
  \EndEnvir{h1}  \Indentfalse }
%    \end{macrocode}
% \end{macro}
%
% \begin{macro}{<h2>}
% \begin{macro}{</h2>}
% \stcs{H2} and \stcs{/h2} work more or less like \stcs{h1} and
% \stcs{/h1}: 
%    \begin{macrocode}
\Cdef {h2}{\endgraf
  \global\advance \SectII by 1
  \global\SectIII = 0  \global\SectIV = 0
  \edef \CurDef {\SectIIim}
  \NewEnvir{h2}{\begingroup}{\endgroup}
  \Heading{14pt plus 7pt minus 2pt}{L}{B}{\SectIIim}}
\Cdef {/h2}{\endgraf
  \nobreak\vskip 4pt plus 1pt
  \EndEnvir{h2}\Indentfalse }
%    \end{macrocode}
% \end{macro}
% \end{macro}
% \begin{macro}{<h3>}
% \begin{macro}{</h3>}
% \stcs{H3} and \stcs{/h3} work more or less like the two preceding
% ones: 
%    \begin{macrocode}
\Cdef {h3}{\endgraf
  \global\advance \SectIII by 1  \global\SectIV = 0
  \edef \CurDef {\SectIIIim}
  \NewEnvir{h3}{\begingroup}{\endgroup}
  \Heading{10pt plus 5pt minus 1pt}{N}{B}{\SectIIIim}}
\Cdef {/h3}{\endgraf
  \nobreak\vskip 2pt plus 1pt
  \EndEnvir{h3}\Indentfalse }
%    \end{macrocode}
% \end{macro}
% \end{macro}
% \begin{macro}{<h4>}
% \begin{macro}{</h4>}
% \stcs{H4} and \stcs{/h4} are slightly different from the previous
% ones in that they use a run-in title (the text continues on the same
% line as the title): 
%    \begin{macrocode}
\Cdef {h4}{\AddVspace{8pt plus 4pt minus 1pt}
  \global\advance \SectIV by 1
  \edef \CurDef {\SectIVim}
  \NewEnvir{h4}{\begingroup}{\endgroup}
  \def \FontBold{B}\SelectFont \Indentfalse \SectIVim }
\Cdef {/h4}{\kern 0.5\baselineskip
  \EndEnvir{h4}}
%    \end{macrocode}
% \end{macro}
% \end{macro}
%
% \subsubsection{Lists}
% \begin{macro}{<list>}
% \begin{macro}{</list>}
% The \stenv{list} environment increases the left margin and sets the
% item counter to 0; everything
% else is handled by the various kinds of \cs{item}.
%    \begin{macrocode}
\Cdef {list}{\AddVspace{\ListSkip}
  \NewEnvir{list}{\begingroup}{\endgroup}
  \advance \leftskip by \ListIndent
  \Indentfalse  \CurParSkip = \ListSkip  \ParIndent = 0pt
  \ItemCount = 0 }
\Cdef {/list}{\endgraf
  \EndEnvir{list}  \Indentfalse }
%    \end{macrocode}
% \end{macro}
% \end{macro}
%
% \begin{macro}{<item>}
% \stcs{item} inserts a bulleted item.
%    \begin{macrocode}
\Cdef {item}{\endgraf  \Indentfalse
  \leavevmode\llap{\BulletItemFormat}\ignorespaces }
%    \end{macrocode}
% \end{macro}
%
% \begin{macro}{<numitem>}
% \stcs{numitem} gives a numbered item.
%    \begin{macrocode}
\Cdef {numitem}{\endgraf
  \advance \ItemCount by 1  \edef \CurDef {\the\ItemCount}
  \Indentfalse
  \leavevmode\llap{\NumItemFormat{\ItemCount}}\ignorespaces }
%    \end{macrocode}
% \end{macro}
% \begin{macro}{\ItemCount}
% The item counter \cs{ItemCount} must also be declared:
%    \begin{macrocode}
\newcount \ItemCount
%    \end{macrocode}
% \end{macro}
%
% \begin{macro}{<textitem>}
% \begin{macro}{</textitem>}
% The \stenv{textitem} environment starts another item with a bold
% text. 
%    \begin{macrocode}
\Cdef {textitem}{\endgraf
  \NewEnvir{textitem}{\begingroup}{\endgroup}
  \def \FontBold {B}\SelectFont \leavevmode \kern -\ListIndent
  \ignorespaces }
\Cdef {/textitem}{\unskip
  \EndEnvir{textitem}%
  \hskip 1em  \ignorespaces }
%    \end{macrocode}
% \end{macro}
% \end{macro}
%
% \subsubsection{Displays}
% \begin{macro}{<display>}
% \begin{macro}{</display>}
% A \stenv{display} environment is implemented by increasing both
% margins. 
%    \begin{macrocode}
\Cdef {display}{\AddVspace{\DisplayPreSkip}
  \NewEnvir{display}{\begingroup}{\endgroup}%
  \advance \leftskip  by \DisplayIndent
  \advance \rightskip by \DisplayIndent
  \CurParSkip = \DisplayParSkip \ParIndent = 0pt \relax }
\Cdef {/display}{\AddVspace{\DisplayPostSkip}
  \EndEnvir{display}}
%    \end{macrocode}
% \end{macro}
% \end{macro}
%
% \subsubsection{Code}
% Handling code like computer programs is quite easy. It involves
% \begin{itemize}
% \item changing "<"'s \cs{catcode} into that of an ordinary
% character,
%
% \item making space and end-of-lines into active characters with a
% suitable definition, and
%
% \item selecting a fixed-width font and a suitable leading.
% \end{itemize}
% It is necessary to remember whether the code followed a "<p>".
% If so, some vertical space will be added before and after the code.
%    \begin{macrocode}
\newif \ifCodePar
%    \end{macrocode}
% \begin{macro}{\CodeSetup}
% All the initialization is handled by the \cs{CodeSetup} macro:
%    \begin{macrocode}
\def \CodeSetup {\ifvmode \CodePartrue \AddVspace{\DisplayPreSkip}
    \else \CodeParfalse \fi
  \def\FontKind{T}\SelectFont
  \baselineskip = \CodeSkipN  \CurParSkip = 0pt  \ParIndent = 0pt
  \catcode`\< = 12  \frenchspacing   \obeylines  \obeyspaces }
%    \end{macrocode}
% \end{macro}
% \begin{macro}{\obeylines}
% Plain \TeX's definition of \cs{obeylines} must be redefined since
% \cs{par} has been changed:
%    \begin{macrocode}
\begingroup
  \catcode`\^^M = \active %
  \gdef\obeylines{\catcode`\^^M=\active \def^^M{\endgraf\leavevmode}}%
\endgroup
%    \end{macrocode}
% \end{macro}
% \begin{macro}{\obeyspaces}
% The appearance of a space when \cs{obeyspaces} is
% in effect must also be modified. By inserting a \cs{leavevmode}, we
% will not lose spaces at the beginning of a line.
%    \begin{macrocode}
\begingroup
  \obeyspaces\gdef {\leavevmode\space}%
\endgroup
%    \end{macrocode}
% \end{macro}
% \begin{macro}{\CodeFinish}
% After the code, \cs{CodeFinish} is called to terminate the mode.
% It adds the needed vertical space.
%    \begin{macrocode}
\def \CodeFinish {\ifCodePar 
    \def \Next {\endgraf  \vskip -\baselineskip  \vskip \DisplayPostSkip
      \global\Indentfalse }%
  \else
    \let \Next = \relax 
  \fi \Next }
%    \end{macrocode}
% \end{macro}
% \begin{macro}{<code>}
% Time has come to write the user interface. The \stenv{code}
% environment reads the code as parameter to the command \cs{ReadCode}.
%    \begin{macrocode}
\Cdef {code}{\NewEnvir{code}{\begingroup}{\endgroup}%
  \CodeSetup \ReadCode }
%    \end{macrocode}
% \end{macro}
% \begin{macro}{\ReadCode}
% \begin{macro}{</code>}
% \cs{ReadCode} just inserts the code (which is parameter "#1")
% and terminates the environment.
%    \begin{macrocode}
\begingroup
  \catcode `\< = 12
  \gdef \ReadCode #1</code>{#1\CodeFinish\EndEnvir{code}}%
\endgroup
%    \end{macrocode}
% \end{macro}
% \end{macro}
% \begin{macro}{<codefile>}
% \stcs{codefile} checks for correct syntax. If the file name is given
% properly, \cs{ReadCodeFile} is called to perform the actual reading.
%    \begin{macrocode}
\Cdef {codefile}{\IfNextChar{[}%
    {\ReadCodeFile}%
    {\Error{No code file name given;}%
      {the syntax is <codefile>[file name].}}}
%    \end{macrocode}
% \end{macro}
% \begin{macro}{\ReadCodeFile}
% \cs{ReadCodeFile} just reads the specified file
%    \begin{macrocode}
\def \ReadCodeFile [#1]{\endgraf \begingroup 
  \CodeSetup 
  \IfFileExists{#1}{\input #1}%
    {\Error{Code file `#1' cound not be found.}{}}
  \CodeFinish \endgroup }
%    \end{macrocode}
% \end{macro}
% 
% \subsubsection{Footnotes}
% \begin{macro}{<footnote>}
% \begin{macro}{</footnote>}
% Implementing footnotes is easy, as the mechanism is ready in plain
% \TeX\@. 
%    \begin{macrocode}
\Cdef {footnote}{\NewEnvir{footnote}%
    {\global\advance \FootnoteCnt by 1
    \footnote{\FootnoteIm{\FootnoteCnt}}\bgroup
      \edef \CurDef {\the\FootnoteCnt}}%
    {\egroup}%
  \ResetFont \SetSize{S}}
\Cdef {/footnote}{\EndEnvir{footnote}}
%    \end{macrocode}
% \end{macro}
% \end{macro}
% \begin{macro}{\FootnoteCnt}
% We must remember to declare the footnote counter:
%    \begin{macrocode}
\newcount \FootnoteCnt
%    \end{macrocode}
% \end{macro}
% 
% \subsubsection{Page headers and footers}
% By default, \starTeX{} has the same headers and footer as \TeX\@.
% The footer must, however, be redefined so that it always uses the
% same font as the body text.
%    \begin{macrocode}
\footline = {\ResetFont\SelectFont \hfil \folio \hfil}
%    \end{macrocode}
%
% \subsection{Customization}
% This section contains definitions that are likely to be changed in
% local adaptions.
%
% \subsubsection{Sizes}
% \begin{macro}{\DisplayIndent}
% \begin{macro}{\DisplayParSkip}
% \begin{macro}{\DisplayPostSkip}
% \begin{macro}{\DisplayPreskip}
% The following four sizes determine the appearance of displays. 
%    \begin{macrocode}
\def \DisplayIndent    {\ListIndent}
\def \DisplayParSkip   {\ListParSkip}
\def \DisplayPostSkip  {5pt plus 2pt minus 1pt\relax}
\def \DisplayPreSkip   {5pt plus 2pt minus 1pt\relax}
%    \end{macrocode}
% \end{macro}
% \end{macro}
% \end{macro}
% \end{macro}
%
% \begin{macro}{\ListIndent}
% \begin{macro}{\ListParSkip}
% \begin{macro}{\ListSkip}
% The next three are used when setting lists.
%    \begin{macrocode}
\def \ListIndent       {25pt\relax}
\def \ListParSkip      {\ListSkip}
\def \ListSkip         {10pt plus 2pt minus 1pt\relax}
%    \end{macrocode}
% \end{macro}
% \end{macro}
% \end{macro}
% 
% \subsubsection{Formats}
% \begin{macro}{\SectIim}
% \begin{macro}{\SectIIim}
% \begin{macro}{\SectIIIim}
% \begin{macro}{\SectIVim}
% \cs{SetcIim}, \cs{SectIIim}, \cs{SectIIIim} and \cs{SectIVim} define
% what the section number at various levels should look
% like.
%    \begin{macrocode}
\def \SectIim {\the\SectI}
\def \SectIIim {\SectIim.\the\SectII}
\def \SectIIIim {\SectIIim.\the\SectIII}
\def \SectIVim {\SectIIIim.\the\SectIV}
%    \end{macrocode}
% \end{macro}
% \end{macro}
% \end{macro}
% \end{macro}
%
% \begin{macro}{\FootnoteIm}
% \cs{FootnoteIm} defines the appearance of the footnote mark; "#1" is
% the footnote number.
%    \begin{macrocode}
\def \FootnoteIm #1{$^{\the #1}$}
%    \end{macrocode}
% \end{macro}
%
% \begin{macro}{\FigIm}
% \begin{macro}{\TabIm}
% \cs{FigIm} and \cs{TabIm} define the appearance of the figure and
% table counters, respectively.
%    \begin{macrocode}
\def \FigIm {\the\FigCnt}
\def \TabIm {\the\TabCnt}
%    \end{macrocode}
% \end{macro}
% \end{macro}
%
% \begin{macro}{\BulletItemFormat}
% \begin{macro}{\NumItemFormat}
% The two commands \cs{BulletItemFormat} and \cs{NumItemFormat} define
% the appearance of the marks used by \stcs{item} and \stcs{numitem},
% respectively. 
%    \begin{macrocode}
\def \BulletItemFormat {$\bullet$\kern 6pt\relax}
\def \NumItemFormat  #1{\the#1.\kern 4pt\relax}
%    \end{macrocode}
% \end{macro}
% \end{macro}
%
% \subsubsection{Character set}
% This version of \starTeX{} uses \textsc{iso}~8859-1 as its standard
% character set. (Users of other character sets will have to modify
% this.) 
%
% \begin{macro}{\E}
% \cs{E} is used to redefine the encoding of a character; this is
% done by making the character active. "#1" is the character's number,
% and "#2" is the \TeX{} code to which it is to be defined.
% The implementation of \cs{E} uses the standard technique of
% defining a \cs{uccode} and then calling \cs{uppercase} to
% insert the correct character where it is defined.)
%    \begin{macrocode}
\def \E #1#2{\catcode#1 = \active 
  \begingroup \uccode`\~ = #1\uppercase{\endgroup \def ~{#2}}}
%    \end{macrocode}
% \end{macro}
% There are five different kinds of reencoding:
% \begin{itemize}
% \item Some characters are reencoded to a different character in the
%   same font using the \cs{char} primitive; examples are
%   161~({\StdECR \symbol{189}}) and 163~({\StdECR \symbol{191}}).
% \end{itemize}
%    \begin{macrocode}
\E{161}{\char189 } \E{163}{\char191 } \E{167}{\char159 }
\E{171}{\char19 }  \E{184}{\char11 }  \E{187}{\char20 }
\E{191}{\char190 } \E{223}{\char255 } \E{255}{\char184 }
%    \end{macrocode}
%
% \begin{macro}{\CC}
% \begin{itemize}
% \item Other characters can be found in the \emph{Companion font},
%   known to \starTeX{} as the Alternate font. The macro \cs{CC} is
%   used for this; its parameter "#1" is the character's position in the
%   companion font. Examples are 162~({\CompECR \symbol{162}}) and
%   164~({\CompECR \symbol{164}}).
% \end{itemize}
%    \begin{macrocode}
\def \CC #1{{\def\FontAlt{A}\SelectFont \char#1}}
\E{162}{\CC{162}} \E{164}{\CC{164}} \E{165}{\CC{165}}
\E{166}{\CC{166}} \E{168}{\CC{168}} \E{169}{\CC{169}}
\E{170}{\CC{170}} \E{172}{\CC{172}} \E{174}{\CC{174}}
\E{175}{\CC{175}} \E{176}{\CC{176}} \E{180}{\CC{180}}
\E{182}{\CC{182}} \E{186}{\CC{186}} 
%    \end{macrocode}
% \end{macro}
%
% \begin{macro}{\MC}
% \begin{itemize}
% \item A few characters are math characters, like 185~({\CompECR
%   \symbol{185}}) and 246~({\CompECR \symbol{246}}). They are defined using
%   the macro \cs{MC} which has two parameters: "#1" is the character's
%   position in the companion font to be used in text mode, and
%   "#2" is some \TeX{} code to be used in math mode.
% \end{itemize}
%    \begin{macrocode}
\def \MC #1#2{\ifmmode #2 \else \CC{#1}\fi }
\E{177}{\MC{177}{\pm}}       \E{178}{\MC{178}{^2{}}} 
\E{179}{\MC{179}{^3{}}}      \E{181}{\MC{181}{\mu}}
\E{183}{\MC{183}{\cdot}}     \E{185}{\MC{185}{^1{}}}
\E{188}{\MC{188}{{1\over4}}} \E{189}{\MC{189}{{1\over2}}}
\E{190}{\MC{190}{{3\over4}}} \E{215}{\MC{214}{\times}}
\E{247}{\MC{246}{\div}}
%    \end{macrocode}
% \end{macro}
%
% \begin{itemize}
% \item Character 160 is a ``non-break space'' which is treated just
%   like a \textasciitilde{} in \TeX.
% \end{itemize}
%    \begin{macrocode}
\E{160}{\nobreak\ }
%    \end{macrocode}
%
% \begin{itemize}
% \item Character 173 is a ``soft hyphen'' which is treated like "\-"
%   in \TeX.
% \end{itemize}
%    \begin{macrocode}
\E{173}{\-}
%    \end{macrocode}
%
% \subsubsection{Language-specific definitions}
% \begin{macro}{\Abstractname}
% \begin{macro}{\Figurename}
% \begin{macro}{\Tablename}
% These three names vary from one language to another.
%    \begin{macrocode}
\def \AbstractName {Abstract}
\def \FigureName   {Figure}
\def \TableName    {Table}
%    \end{macrocode}
% \end{macro}
% \end{macro}
% \end{macro}
% \begin{macro}{\TimeSep}
% Not all languages use a ``\texttt{:}'' to separate the hours,
% minutes, and seconds.
%    \begin{macrocode}
\def \TimeSep      {:}
%    \end{macrocode}
% \end{macro}
% \begin{macro}{<today>}
% \begin{macro}{\Month}
% \begin{macro}{\Th}
% The \stcs{today} command should be redefined for each language.
% To simplify the code, two auxiliary macros are defined: \cs{Month}
% gives the name of the month, and \cs{Th} gives the correct suffix
% for the day of the month.
%    \begin{macrocode}
\Cdef {today}{\the\day\Th{\day} \Month\space\the\year}
\def \Month {\ifcase \month \or January\or February\or March\or
  April\or May\or June\or July\or August\or September\or
  October\or November\or December\fi }
\def \Th #1{%
  \ifnum #1=1 st\else\ifnum #1=21 st\else\ifnum #1=31 st\else
  \ifnum #1=2 nd\else\ifnum #1=22 nd\else
  \ifnum #1=3 rd\else\ifnum #1=23 rd\else th\fi\fi\fi\fi\fi\fi\fi }
%    \end{macrocode}
% \end{macro}
% \end{macro}
% \end{macro}
% \begin{macro}{<now>}
% The \stcs{now} command will quite probably remain unchanged for most
% languages (except for the \cs{TimeSep}; see above), but is given
% here just the same. 
%    \begin{macrocode}
\Cdef {now}{\Minutes = \time  \Hours = \Minutes
  \divide \Hours by 60  \Htemp = \Hours  \multiply \Htemp by -60
  \advance \Minutes by \Htemp
  \the\Hours \TimeSep \ifnum \Minutes > 9 \else 0\fi \the\Minutes }
%    \end{macrocode}
% The three counters used by \stcs{now} must be defined.
%    \begin{macrocode}
\newcount \Minutes  \newcount \Hours  \newcount \Htemp
%    \end{macrocode}
% \end{macro}
%
% \subsection{Math mode}
% A few modifications to the \TeX{} math mode are necessary.
% 
% The "<" character must be active also in math mode.
%    \begin{macrocode}
\mathcode`< = "8000
%    \end{macrocode}
% \begin{macro}{<math>}
% \begin{macro}{</math>}
% The inline math mode "$"\dots"$" is called \stenv{math} in \starTeX.
%    \begin{macrocode}
\Cdef {math}{\MathFonts \NewEnvir{math}{$}{$}}
\Cdef {/math}{\EndEnvir{math}}
%    \end{macrocode}
% \end{macro}
% \end{macro}
% \begin{macro}{<displaymath>}
% \begin{macro}{</displaymath>}
% The display math environment "$$"\dots"$$" is also available in
% \starTeX{} using the notation \stenv{displaymath}. 
%    \begin{macrocode}
\Cdef {displaymath}{\endgraf \MathFonts \NewEnvir{displaymath}{$$}{$$}}
\Cdef {/displaymath}{\EndEnvir{displaymath}}
%    \end{macrocode}
% \end{macro}
% \end{macro}
%
% \subsubsection{Math fonts}
% \begin{macro}{\MSetFont}
% The \cs{MSetFont} macro is used to simplify the definition of math
% fonts. It takes two parameters: "#1" is the math font element being
% defined (i.e., \cs{textfont0}), and "#2" is the font name.
%    \begin{macrocode}
\def \MSetFont #1#2{\font \NewMFont = \csname #2\endcsname
  #1 = \NewMFont }
%    \end{macrocode}
% \end{macro}
% \begin{macro}{\MathFonts}
% The \cs{MathFonts} macro is called whenever math mode is entered. It
% will define all the required math fonts. To avoid unnecessary work,
% new fonts will only be assigned if the font size has changed since
% the last time.
%    \begin{macrocode}
\def \MathFonts {\if \FontSize \LastMathSize \else
    \MSetFont{\textfont0}{M-\FontSize N0}%
    \MSetFont{\scriptfont0}{M-\FontSize S0}%
    \MSetFont{\scriptscriptfont0}{M-\FontSize X0}%
    \MSetFont{\textfont1}{M-\FontSize N1}%
    \MSetFont{\scriptfont1}{M-\FontSize S1}%
    \MSetFont{\scriptscriptfont1}{M-\FontSize X1}%
    \MSetFont{\textfont2}{M-\FontSize N2}%
    \MSetFont{\scriptfont2}{M-\FontSize S2}%
    \MSetFont{\scriptscriptfont2}{M-\FontSize X2}%
    \MSetFont{\textfont3}{M-\FontSize N3}%
    \MSetFont{\scriptfont3}{M-\FontSize S3}%
    \MSetFont{\scriptscriptfont3}{M-\FontSize X3}%
    \MSetFont{\textfont4}{M-\FontSize N4}%
    \MSetFont{\scriptfont4}{M-\FontSize S4}%
    \MSetFont{\scriptscriptfont4}{M-\FontSize X4}%
    \MSetFont{\textfont5}{M-\FontSize N5}%
    \MSetFont{\scriptfont5}{M-\FontSize S5}%
    \MSetFont{\scriptscriptfont5}{M-\FontSize X5}%
    \MSetFont{\textfont6}{M-\FontSize N6}%
    \MSetFont{\scriptfont6}{M-\FontSize S6}%
    \MSetFont{\scriptscriptfont6}{M-\FontSize X6}%
    \let \LastMathSize = \FontSize
  \fi }
%    \end{macrocode}
% \end{macro}
% \begin{macro}{\LastMathSize}
% \cs{LastMathSize} contains the last math size used. It
% must be initialized to an unused value so that the math fonts are
% always loaded the first time math mode is entered.
%    \begin{macrocode}
\def \LastMathSize {?}
%    \end{macrocode}
% \end{macro}
%
% \subsubsection{Letters}
% The standard 26 letters are already defined, but we must define the
% others, like `\'e' and `\AA'. These are taken from the \textsc{ecti}
% font (number~5 in \starTeX). 
%    \begin{macrocode}
\def \Df #1{\mathcode"#1 = "05#1 \relax }
\Df{C0}\Df{C1}\Df{C2}\Df{C3}\Df{C4}\Df{C5}\Df{C6}\Df{C7}
\Df{C8}\Df{C9}\Df{CA}\Df{CB}\Df{CC}\Df{CD}\Df{CE}\Df{CF}
\Df{D0}\Df{D1}\Df{D2}\Df{D3}\Df{D4}\Df{D5}\Df{D6}
\Df{D8}\Df{D9}\Df{DA}\Df{DB}\Df{DC}\Df{DD}\Df{DE}\Df{DF}
\Df{E0}\Df{E1}\Df{E2}\Df{E3}\Df{E4}\Df{E5}\Df{E6}\Df{E7}
\Df{E8}\Df{E9}\Df{EA}\Df{EB}\Df{EC}\Df{ED}\Df{EE}\Df{EF}
\Df{F0}\Df{F1}\Df{F2}\Df{F3}\Df{F4}\Df{F5}\Df{F6}
\Df{F8}\Df{F9}\Df{FA}\Df{FB}\Df{FC}\Df{FD}\Df{FE}\Df{FF}
%    \end{macrocode}
%
% \subsubsection{Bold letters}
% \begin{macro}{<bolda>}
% \begin{macro}{<bolducA>}
% The bold letters \stcs{bolda}\dots\stcs{boldz} and
% \stcs{bolducA}\dots\stcs{bolducZ} are intended for use i math mode.
% They are found on the \textsc{cmbx} font (number~4 in \starTeX).
% \changes{1.04}{1999/03/10}{Added \stcs{bolda}, \stcs{bolducA}, and
%   related commands.}
%    \begin{macrocode}
\def \Df #1#2#3{\Cdef{bold#1}{\MSy{\mathchar"04#2}}%
  \Cdef{bolduc#1}{\MSy{\mathchar"04#3}}}
\Df{a}{61}{41}\Df{b}{62}{42}\Df{c}{63}{43}\Df{d}{64}{44}
\Df{e}{65}{45}\Df{f}{66}{46}\Df{g}{67}{47}\Df{h}{68}{48}
\Df{i}{69}{49}\Df{j}{6A}{4A}\Df{k}{6B}{4B}\Df{l}{6C}{4C}
\Df{m}{6D}{4D}\Df{n}{6E}{4E}\Df{o}{6F}{4F}\Df{p}{70}{50}
\Df{q}{71}{51}\Df{r}{72}{52}\Df{s}{73}{53}\Df{t}{74}{54}
\Df{u}{75}{55}\Df{v}{76}{56}\Df{w}{77}{57}\Df{x}{78}{58}
\Df{y}{79}{59}\Df{z}{7A}{5A}
%    \end{macrocode}
% \end{macro}
% \end{macro}
%
% \subsubsection{Operators}
% \begin{macro}{\MSy}
% \cs{MSy} is used to insert a mathematical symbol. It will expand to
% "#1", and insert a math mode environment ("$"\dots"$") if required.
%    \begin{macrocode}
\def \MSy #1{\ifmmode #1\else $#1$\fi }
%    \end{macrocode}
% \end{macro}
% \begin{macro}{\MOp}
% \cs{MOp} is also used to insert the mathematical symbol "#2", but
% gives an error message if not in a math environment. Parameter "#1"
% is the \starTeX{} command name, which is used in the error message.
%    \begin{macrocode}
\def \MOp #1#2{\ifmmode \def \MNext {#2}\else
  \Error{Command <#1> is only allowed in math mode;}%
    {command ignored.}\let \MNext = \relax \fi 
  \MNext }
%    \end{macrocode}
% \end{macro}
% \begin{macro}{\Mdef}
% \cs{Mdef} is used to define a \starTeX{} mathematical symbol
% command whenever the \starTeX{} command is
% identical to the name of the 
% corresponding \TeX{} macro; parameter "#1" is the name of both.
%    \begin{macrocode}
\def \Mdef #1{\Cdef{#1}{\MSy{\csname #1\endcsname}}}
%    \end{macrocode}
% \end{macro}
% \begin{macro}{\Odef}
% \cs{Odef} is similar to \cs{Mdef} but is for definitions based on
% \cs{MOp} rather than \cs{MSy}.
%    \begin{macrocode}
\def \Odef #1{\Cdef{#1}{\MOp{#1}{\csname #1\endcsname}}}
%    \end{macrocode}
% \end{macro}
% These commands are now used to define all the standard unary and
% binary operators. 
%    \begin{macrocode}
\Mdef{amalg} \Mdef{bigcirc} \Mdef{bigtriangleup} 
\Mdef{bigtriangledown} \Mdef{bullet} \Mdef{cap} \Mdef{circ}
\Mdef{cup} \Mdef{dagger} \Mdef{ddagger} \Mdef{diamond} \Mdef{mp}
\Mdef{odot} \Mdef{ominus} \Mdef{oplus} \Mdef{oslash} \Mdef{otimes}
\Mdef{setminus} \Mdef{sqcap} \Mdef{sqcup} \Mdef{star}
\Mdef{triangleleft} \Mdef{triangleright} \Mdef{uplus} \Mdef{vee}
\Mdef{wedge} \Mdef{wr}
%    \end{macrocode}
%
% \subsubsection{Relations}
% These are the various relations available.
% \changes{1.01}{1996/11/29}{Added forgotten operator \stcs{propto}.}
%    \begin{macrocode}
\Mdef{approx} \Mdef{asymp} \Mdef{bowtie} \Mdef{cong} \Mdef{doteq}
\Mdef{dashv} \Mdef{equiv} \Mdef{frown} \Mdef{geq} \Mdef{gg} \Mdef{in}
\Mdef{leq} \Mdef{ll} \Mdef{mid} \Mdef{models} \Mdef{neq} \Mdef{ni}
\Mdef{notin} \Mdef{parallel} \Mdef{perp} \Mdef{prec} \Mdef{preceq} 
\Mdef{propto} \Mdef{sim} \Mdef{simeq} \Mdef{smile} \Mdef{sqsubseteq} 
\Mdef{sqsupseteq} \Mdef{subset} \Mdef{subseteq} \Mdef{supset} 
\Mdef{supseteq} \Mdef{succ} \Mdef{succeq} \Mdef{vdash}
%    \end{macrocode}
% \begin{macro}{<gt>}
% \begin{macro}{<lt>}
% The two symbols \stcs{gt} and \stcs{lt} must be defined; the former
% is not really necessary, but is included for symmetry reasons. Note
% also that the definition of \stcs{lt} is slightly tricky since "<"
% is an active symbol in math mode.
%    \begin{macrocode}
 \Cdef{gt}{>}
 \Cdef{lt}{\ifmmode \mathchar"313C \else <\fi}
%    \end{macrocode}
% \end{macro}
% \end{macro}
% 
% \subsubsection{Delimitiers}
% In addition to (), [], $\{\}$ and $\vert$, the following delimiters
% are defined: 
%    \begin{macrocode}
\Mdef{langle} \Mdef{lceil} \Mdef{lfloor} \Mdef{rangle} 
\Mdef{rceil} \Mdef{rfloor}
%    \end{macrocode}
%
% \subsubsection{Arrows}
% All the standard arrows are included.
%    \begin{macrocode}
\Mdef{leftarrow} \Mdef{rightarrow} \Mdef{leftrightarrow}
\Mdef{longleftarrow} \Mdef{longrightarrow} \Mdef{longleftrightarrow}
\Mdef{uparrow} \Mdef{downarrow}
\Mdef{updownarrow} \Mdef{nearrow} \Mdef{nwarrow} \Mdef{searrow}
\Mdef{swarrow} \Mdef{mapsto} \Mdef{longmapsto} \Mdef{hookleftarrow}
\Mdef{hookrightarrow} \Mdef{leftharpoonup} \Mdef{rightharpoonup}
\Mdef{leftharpoondown} \Mdef{rightharpoondown} \Mdef{rightleftharpoons}
%    \end{macrocode}
% In \TeX, some arrows use uppercase letters to indicate double
% arrows. I use the prefix "double" instead.
%    \begin{macrocode}
\Cdef{doubleleftarrow}{\MSy{\Leftarrow}}
\Cdef{doublerightarrow}{\MSy{\Rightarrow}}
\Cdef{doubleleftrightarrow}{\MSy{\Leftrightarrow}}
\Cdef{doublelongleftarrow}{\MSy{\Longleftarrow}}
\Cdef{doublelongrightarrow}{\MSy{\Longrightarrow}}
\Cdef{doublelongleftrightarrow}{\MSy{\Longleftrightarrow}}
\Cdef{doubleuparrow}{\MSy{\Uparrow}}
\Cdef{doubledownarrow}{\MSy{\Downarrow}}
\Cdef{doubleupdownarrow}{\MSy{\Updownarrow}}
%    \end{macrocode}
%
% \subsubsection{Big symbols}
% These are the big symbols.
%    \begin{macrocode}
\Odef{bigcap} \Odef{bigcup} \Odef{bigodot} \Odef{bigoplus}
\Odef{bigotimes} \Odef{bigsqcup} \Odef{biguplus} \Odef{bigvee}
\Odef{bigwedge} \Odef{int} \Odef{oint} \Odef{prod} \Odef{sum}
%    \end{macrocode}
% 
% \subsubsection{Function names}
% Some names are defined as commands to ensure that they will be
% printed in the correct font.
% \changes{1.01}{1996/11/29}{Name of \cs{sup} operator changed to
% \stcs{sup-op} because of name clash with the \stcs{sup} command used
% for indices. (Detected by Marco Paggi,
% \textsf{aampaggi@input.lognet.it}.)} 
%    \begin{macrocode}
\Mdef{arccos} \Mdef{arcsin} \Mdef{arctan} \Mdef{arg} \Mdef{cos}
\Mdef{cosh} \Mdef{cot} \Mdef{coth} \Mdef{csc} \Mdef{deg} \Mdef{det}
\Mdef{dim} \Mdef{exp} \Mdef{gcd} \Mdef{hom} \Mdef{inf} \Mdef{ker}
\Mdef{lg} \Mdef{lim} \Mdef{liminf} \Mdef{limsup} \Mdef{ln}
\Mdef{log} \Mdef{max} \Mdef{min} \Mdef{sec} \Mdef{sin} \Mdef{sinh}
\Cdef{sup-op}{\MSy{\sup}} \Mdef{tan} \Mdef{tanh}
%    \end{macrocode}
% \stcs{pr} must be given a special definition as its name in plain \TeX{}
% has an uppercase letter.
%    \begin{macrocode}
\Cdef{pr}{\MSy{\Pr}}
%    \end{macrocode}
%
% \subsubsection{Greek letters}
% The lowercase Greek letters can be found just using the plain \TeX{}
% name.
%    \begin{macrocode}
\Mdef{alpha} \Mdef{beta} \Mdef{gamma} \Mdef{delta} \Mdef{epsilon}
\Mdef{varepsilon} \Mdef{zeta} \Mdef{eta} \Mdef{theta} \Mdef{vartheta}
\Mdef{iota} \Mdef{kappa} \Mdef{lambda} \Mdef{mu} \Mdef{nu} \Mdef{xi}
\Cdef {omicron}{\MSy{o}} \Mdef{pi} \Mdef{varpi} \Mdef{rho} \Mdef{varrho}
\Mdef{sigma} \Mdef{varsigma} \Mdef{tau} \Mdef{upsilon} \Mdef{phi}
\Mdef{varphi} \Mdef{chi} \Mdef{psi} \Mdef{omega}
%    \end{macrocode}
% The uppercase Greek letters can be found in the \textsc{cmr}
% font, which is font no 0 in \starTeX.
% Those that also exist in the Latin alphabet, can be found in their
% normal position, while the special Greek letters are placed at the
% beginning of the font table.
%    \begin{macrocode}
\def \Df #1#2{\Cdef{#1}{\MSy{\mathchar"#2}}}
\Df{ucalpha}{0041} \Df{ucbeta}{0042}    \Df{ucgamma}{0000}
\Df{ucdelta}{0001} \Df{ucepsilon}{0045} \Df{uczeta}{005A}
\Df{uceta}{0048}   \Df{uctheta}{0002}   \Df{uciota}{0049}
\Df{uckappa}{004B} \Df{uclambda}{0003}  \Df{ucmu}{004D}
\Df{ucnu}{004E}    \Df{ucxi}{0004}      \Df{ucomicron}{004F}
\Df{ucpi}{0005}    \Df{ucrho}{0050}     \Df{ucsigma}{0006}
\Df{uctau}{0054}   \Df{ucupsilon}{0007} \Df{ucphi}{0008}
\Df{ucchi}{0058}   \Df{ucpsi}{0009}     \Df{ucomega}{000a}
%    \end{macrocode}
%
% \subsubsection{Blackboard set letters}
% The socalled ``blackboard bold'' letters $\Bbb{C}$, $\Bbb{N}$,
% $\Bbb{R}$ and $\Bbb{Z}$ are used to denote the standard sets.
% \changes{1.03}{1998/04/30}{Added the various set names. (Suggested
% by Christian Sommerfeldt \O{}ien, \textsf{christso@ifi.uio.no}.)}
%    \begin{macrocode}
\def \Df #1#2{\Cdef{#1}{\MSy{\mathchar"06#2}}}
\Df{cset}{43} \Df{nset}{4E} \Df{rset}{52} \Df{zset}{5A}
%    \end{macrocode}
%
% \subsubsection{Calligraphic letters}
% \begin{macro}{cala}
% The calligraphic letters can be found in font~2.
% \changes{1.04}{1999/03/10}{Fixed a bug in the \stcs{calA} and
%   related commands.}
%    \begin{macrocode}
\def \Df #1#2{\Cdef{cal#1}{\MSy{\mathchar"02#2}}}
\Df{a}{41} \Df{b}{42} \Df{c}{43} \Df{d}{44} \Df{e}{45}
\Df{f}{46} \Df{g}{47} \Df{h}{48} \Df{i}{49} \Df{j}{4A}
\Df{k}{4B} \Df{l}{4C} \Df{m}{4D} \Df{n}{4E} \Df{o}{4F}
\Df{p}{50} \Df{q}{51} \Df{r}{52} \Df{s}{53} \Df{t}{54}
\Df{u}{55} \Df{v}{56} \Df{w}{57} \Df{x}{58} \Df{y}{59}
\Df{z}{5A}
%    \end{macrocode}
% \end{macro}
% 
% \subsubsection{Other symbols}
% A few symbols are left.
%    \begin{macrocode}
\Mdef{aleph} \Mdef{angle} \Mdef{bot} \Mdef{ell} \Mdef{emptyset}
\Mdef{exists} \Mdef{forall} \Mdef{hbar} \Mdef{nabla} \Mdef{neg}
\Odef{not} \Mdef{partial} \Mdef{surd} \Mdef{top} \Mdef{wp}
\Cdef{infinity}{\MSy{\infty}} \Cdef{im}{\MSy{\Im}} \Cdef{re}{\MSy{\Re}}
\Cdef{,}{{,}} 
\Cdef{:}{\MSy{\vdots}} 
\Cdef{:::}{\MSy{\cdots}}
%    \end{macrocode}
% \begin{macro}{<...>}
% The definition of \stcs{...} has a minute space to allow \starTeX{}
% to split the line.
%    \begin{macrocode}
\Cdef{...}{\MSy{\ldots}\hskip 0.001pt \relax }
%    \end{macrocode}
% \end{macro}
%
% \subsubsection{Subscripts and superscripts}
% \begin{macro}{<sub>}
% \begin{macro}{</sub>}
% \begin{macro}{<sup>}
% \begin{macro}{</sup>}
% The sub- and superscripts in equations are accessible through
% \stenv{sub} and \stenv{sup}.
%    \begin{macrocode}
\Cdef {sub}{\MOp{sub}{\NewEnvir{sub}{_\bgroup}{\egroup}}}
\Cdef {/sub}{\MOp{/sub}{\EndEnvir{sub}}}
\Cdef {sup}{\MOp{sup}{\NewEnvir{sup}{^\bgroup}{\egroup}}}
\Cdef {/sup}{\MOp{/sup}{\EndEnvir{sup}}}
%    \end{macrocode}
% \end{macro}
% \end{macro}
% \end{macro}
% \end{macro}
%
% \subsubsection{Fractions}
% \begin{macro}{<frac>}
% \begin{macro}{<over>}
% \begin{macro}{</frac>}
% Fractions are set with in a
% \stcs{frac}\dots\stcs{over}\dots\stcs{/frac} environment.
%    \begin{macrocode}
\Cdef {frac}{\MOp{frac}{\NewEnvir{frac}{\bgroup}{\egroup}%
    \NOver = 0\relax}}
\Cdef {over}{\MOp{over}{\FracOver}}
\Cdef {/frac}{\MOp{/frac}{\ifnum \NOver = 0
    \Error{No <over> in the <frac>...<over>...</frac> environment.}{}\fi
    \EndEnvir{frac}}}
\def \FracOver {\ifTextEqual{\CurEnv}{frac}%
    \ifnum \NOver = 0 \over
    \else \Error{Only one <over> may occur in each <frac>...</frac>
        environment.}{}%
    \fi  \advance \NOver by 1
  \else \Error{<over> only allowed in a 
    <frac>...<over>...</frac> environment.}{}%
  \fi }
%    \end{macrocode}
% \end{macro}
% \end{macro}
% \end{macro}
% To check for legal use of \stcs{over}, a counter named \cs{NOver} is
% used.
%    \begin{macrocode}
\newcount \NOver
%    \end{macrocode}
%
% \subsubsection{Roots}
% \begin{macro}{<sqrt>}
% \begin{macro}{</sqrt>}
% The \stenv{sqrt} environment is just an interface to the \cs{sqrt}
% macro in plain \TeX.
%    \begin{macrocode}
\Cdef {sqrt}{\MOp{sqrt}{\NewEnvir{sqrt}{\sqrt\bgroup}{\egroup}}}
\Cdef {/sqrt}{\MOp{/sqrt}{\EndEnvir{sqrt}}}
%    \end{macrocode}
% \end{macro}
% \end{macro}
%
% \subsection{PostScript figures}
% \begin{macro}{<psfig>}
% \begin{macro}{</psfig>}
% \starTeX{} allows the inclusion of \ps{} figures through the
% \stenv{psfig} environment. It checks if the command is allowed in the
% current state, and, if it is, \cs{PSfig} is called; if not,
% \cs{PSfigError} is called instead.
%    \begin{macrocode}
\Cdef {psfig}{\if \State x \let \Next = \PSfig \else 
  \let \Next = \PSfigError \fi
  \Next }
\Cdef {/psfig}{\AddVspace{6pt plus 1pt}
  \ifx \PSfile \relax \else 
    \centerline{\epsfbox{\PSfile}}%
    \AddVspace{10pt plus 2pt minus 1pt}
  \fi
  \EndEnvir{psfig}}
%    \end{macrocode}
% \end{macro}
% \end{macro}
% The actual inclusion of the \ps{} file in handled by \cs{epsfbox}
% from the file \texttt{epsf.tex} which is provided with the
% \textsc{dvips} program.
%    \begin{macrocode}
\input epsf.tex
%    \end{macrocode}
% \begin{macro}{\PSfig}
% \cs{PSfig} checks whether the user has remembered to provide the
% file name; if so, \cs{PSfetch} is called to check the \ps{} file.
%    \begin{macrocode}
\def \PSfig{\NewEnvir{psfig}{\topinsert}{\endinsert}
  \let \PSfile = \relax \let \State = p
  \IfNextChar{[}%
    {\PSfetch}%
    {\Error{No file name for PostScript figure;}%
      {the syntax is: <psfig>[file name]caption text</psfig>.}%
     \PScaption}}
%    \end{macrocode}
% \end{macro}
% \begin{macro}{\PSfigError}
% \cs{PSfigError} gives an error message and starts a dummy
% environment. 
%    \begin{macrocode}
\def \PSfigError {\endgraf
  \Error{Calls on <psfig> not allowed inside <psfig> or <table>;}%
     {the command was ignored.}%
  \NewEnvir{psfig}{\begingroup}{\endgroup}}
%    \end{macrocode}
% \end{macro}
% \begin{macro}{\PSfetch}
% \cs{PSfetch} checks that the \ps{} exists, and notes its name in
% \cs{PSfile}. 
%    \begin{macrocode}
\def \PSfetch [#1]{\IfFileExists{#1}%
    {\gdef \PSfile {#1}}%
    {\Error{PostScript file `#1' could not be found.}{}}%
    \PScaption }
%    \end{macrocode}
% \end{macro}
% \begin{macro}{\epsfsize}
% \cs{epsfsize} is used by \cs{epsfbox} to scale the \ps{} figure.
% We want figures to fill either 80\% of the text width or 40\% of the
% text height. This is done by comparing
% $\frac{0.8\mbox{\cs{hsize}}}{\mathtt{\#1}}$ and
% $\frac{0.4\mbox{\cs{vsize}}}{\mathtt{\#2}}$, where "#1" and "#2"
% are the horizontal and vertical original sizes of the the figure.
%    \begin{macrocode}
\def \epsfsize #1#2{0pt \dimen1 = 0.8\hsize  \dimen2 = 0.4\vsize
  \floatdiv{\dimen2}{#1}{#2}%
  \ifdim \dimen1 < \divres \epsfxsize = 0.8\hsize
  \else \epsfysize = 0.4\vsize \fi }
%    \end{macrocode}
% \end{macro}
% \begin{macro}{\PScaption}
% The formatting of the caption is done by \cs{PScaption}:
%    \begin{macrocode}
\def \PScaption {\global\advance \FigCnt by 1
  \edef \CurDef {\the\FigCnt}
  \leftskip = 0.1\hsize  \rightskip = \leftskip
  \ResetFont \SetSize{S}\Indentfalse
  \FigureName~\FigIm: \ignorespaces }
%    \end{macrocode}
% \end{macro}
% \cs{PScaption} needs a counter to count the figures:
%    \begin{macrocode}
\newcount \FigCnt
%    \end{macrocode}
%
% \subsection{Tables}
% \begin{macro}{<table>}
% \begin{macro}{</table>}
% Tables are created with the \stenv{table} environment. It checks whether
% the command is allowed in the current state and calls
% \cs{Table} if it is; otherwise it calls \cs{TableError}.
%    \begin{macrocode}
\Cdef {table}{\if \State x\let \Next = \Table \else
  \let \Next = \TableError \fi
  \Next }
\Cdef {/table}{\EndEnvir{table}}
%    \end{macrocode}
% \end{macro}
% \end{macro}
% The table is temporarily placed in a box so that it may be centered
% when output.
%    \begin{macrocode}
\newbox \TableBox
%    \end{macrocode}
% \begin{macro}{\Table}
% \cs{Table} starts the table environment.
%    \begin{macrocode}
\def \Table{\topinsert
  \NewEnvir{table}{\TableCaption}{\LastRow\endinsert}%
  \let \State = t}
%    \end{macrocode}
% \end{macro}
% \begin{macro}{\TableCaption}
% The formatting of the table caption is done by \cs{TableCaption}:
%    \begin{macrocode}
\def \TableCaption {\global\advance \TabCnt by 1
  \edef \CurDef {\the\TabCnt}
  \leftskip = 0.1\hsize  \rightskip = \leftskip
  \ResetFont \SetSize{S}\Indentfalse
  \TableName~\TabIm: \ignorespaces }
%    \end{macrocode}
% \end{macro}
% \cs{TableCaption} needs a counter to count the tables:
%    \begin{macrocode}
\newcount \TabCnt
%    \end{macrocode}
% \begin{macro}{\TableError}
% \cs{TableError} gives an error message and starts a dummy
% environment. 
%    \begin{macrocode}
\def \TableError {\endgraf
  \Error{Calls on <table> not allowed inside <psfig>
    or <table>;}{the command was ignored.}%
  \NewEnvir{table}{\begingroup}{\endgroup}}
%    \end{macrocode}
% \end{macro}
%
% \subsubsection{Table rows}
% \begin{macro}{<row>}
% \begin{macro}{\Row}
% The command \stcs{row} is the interface to \cs{Row} which is used to
% start another table row. Its actions depend on the current state.
%    \begin{macrocode}
\Cdef {row}{\Row}
\def \Row {\if \State t\FirstRow\NewRow \else
   \if \State r\EndRow\NewRow   \else
   \Error{The <row> command is only allowed inside a
     <table>.}{The command was ignored.}\fi\fi }
%    \end{macrocode}
% \end{macro}
% \end{macro}
% \begin{macro}{\FirstRow}
% \cs{FirstRow} is called prior to the first table row.
%    \begin{macrocode}
\def \FirstRow {\endgraf \ResetFont \SelectFont \let \State = r%
  \setbox\TableBox = \hbox\bgroup \vbox\bgroup \offinterlineskip 
    \halign\bgroup \vrule ##\strut&& 
      \kern 6pt \hfil ##\unskip \hfil \kern 6pt \vrule \cr
      \noalign{\hrule}}
%    \end{macrocode}
% \end{macro}
% \begin{macro}{\NewRow}
% \cs{NewRow} is used to start another table row.
%    \begin{macrocode}
\def \NewRow {&}
%    \end{macrocode}
% \end{macro}
% \begin{macro}{\EndRow}
% The command \cs{EndRow} is used to terminate the current table row.
%    \begin{macrocode}
\def \EndRow {\cr \noalign{\hrule}}
%    \end{macrocode}
% \end{macro}
% \begin{macro}{\LastRow}
% The command \cs{LastRow} is used to terminate the last table row.
%    \begin{macrocode}
\def \LastRow {\if \State r\EndRow\PrintTable\fi}
%    \end{macrocode}
% \end{macro}
% \begin{macro}{\PrintTable}
% \cs{PrintTable} is the final command used when processing a table.
% It terminates the table structure and outputs it.
%    \begin{macrocode}
\def \PrintTable {\egroup\egroup\egroup
  \AddVspace{6pt plus 2pt minus 1pt}
  \centerline{\box\TableBox}%
  \AddVspace{10pt plus 4pt minus 2pt}} 
%    \end{macrocode}
% \end{macro}
%
% \subsubsection{Table columns}
% \begin{macro}{<col>}
% \stcs{col} is used to start another table column.
%    \begin{macrocode}
\Cdef {col}{\if \State r&\else
  \Error{The <col> command is only allowed after a <row>
    inside a <table>.}{The command was ignored.}\fi}
%    \end{macrocode}
% \end{macro}
% 
% \subsection{Cross-references}
% This section describes the various commands used for
% cross-referencing.
%
% The basic idea is the same as the one used in \LaTeX: an extra file (with
% extension \texttt{.aux}) contains all the labels defined in the
% document. There are two reasons for this:
% \begin{itemize}
% \item Using a separate file makes it possible to have forward
%   references. 
% \item When processing a document, the page number of each label is
%   not known until the page is actually written to the \textsc{dvi}
%   file. The \cs{write} command is, consequently, delayed until that stage
%   of the processing, so using a file will always give us the
%   correct page number.
% \end{itemize}
% \begin{macro}{\ifAuxOpen}
% The \texttt{.aux} file is only created when there are \stcs{label}
% commands in the document. The first \stcs{label} creates the file,
% and the flag \cs{ifAuxOpen} indicates whether this has been done.
%    \begin{macrocode}
\newif \ifAuxOpen
%    \end{macrocode}
% \end{macro}
%
% \subsubsection{Defining labels}
% \begin{macro}{<label>}
% \stcs{label} is used to define another cross-reference label. The
% value of the label is the current value of \cs{CurDef} which is set
% by \stcs{h1}, \stcs{table}, etc.
%
% \stcs{label} first checks the syntax, and then alters the
% \cs{catcode} of "<" in case there are  any "<"s in the label.
% Finally, it calls \cs{NewLabel} to perform the actual definition.
%    \begin{macrocode}
\Cdef {label}{\IfNextChar{[}%
    {\begingroup \catcode`\< = 12 \NewLabel}%
    {\Error{No label given;}%
      {the syntax is <label>[your label].}}}
%    \end{macrocode}
% \end{macro}
% \begin{macro}{\NewLabel}
% \cs{NewLabel} first reads the old ".aux" file (unless it has already
% been read) and opens a new ".aux" file (unless that has been done).
%    \begin{macrocode}
\def \NewLabel [#1]{\endgroup
  \ifAuxRead \else \ReadAuxFile \fi
  \ifAuxOpen \else 
    \immediate\openout \NewAuxFile = \jobname.aux
    \global\AuxOpentrue \fi
%    \end{macrocode}
% Then, it checks whether the label has been previously defined, and
% prints an error message if it has been. Otherwise, it notes the
% definition and writes the necessary data to the ".aux"
% file.\footnote{Note the trick of redefining \cs{the} so that
%   \cs{pageno} is not expanded until the page is completed. This was
%   found in \cite[Exercise~21.10]{texbook}.}
%    \begin{macrocode}
  \expandafter\ifx \csname L]#1\endcsname \relax
    \expandafter\edef \csname L]#1\endcsname{\the\inputlineno}%
    {\let \the = 0\relax
      \edef \WriteCurDef {\write \NewAuxFile 
        {\string\LabelDef 
          ]#1]\the\pageno]\CurDef E-o-LabelDef\string\relax}}%
      \WriteCurDef }%
  \else
    \Error{Label `#1' already defined on line \csname L]#1\endcsname;}%
      {this definition is ignored.}%
  \fi }
%    \end{macrocode}
% \end{macro}
% Note that when a label \emph{\texttt{xxx}} is defined, a macro named
% \cs{L]\emph{xxx}} is defined, and its definition is the line number.
% This is used to give a better error message in case there are
% multiple definitions of that label.
%
% The new ".aux" file must be declared.
%    \begin{macrocode}
\newwrite \NewAuxFile
%    \end{macrocode}
% We must also define an initial value for \cs{CurDef} in case the
% user calls \stcs{label} prior to the first \stcs{h\emph{x}}.
%    \begin{macrocode}
\edef \CurDef {0}
%    \end{macrocode}
%
% \subsubsection{Reading the cross-reference label file}
% When the user first references a label (using a \stcs{ref} command),
% the \texttt{.aux} file is read to find all the labels that were
% defined during the previous run. Each label will be represented by
% to macros:
% \begin{description}
% \item[\texttt{R]\emph{xxx}}] gives the reference, i.e.~``2.1'' if
%   the label was defined in Section~2.1 of the document.
% \item[\texttt{P]\emph{xxx}}] gives the page number where the label
%   was placed.
% \end{description}
% \begin{macro}{\ReadAuxFile}
% \cs{ReadAuxFile} reads the ".aux" file produced by the previous run.
% It  redefines the \cs{catcode} of "<" in case that characters occurs
% in a label.
%    \begin{macrocode}
\def \ReadAuxFile {\IfFileExists{\jobname.aux}%
    {\begingroup
       \def \LabelDef {\begingroup
          \catcode `\\ = 12\relax \FetchLabel }%
       \catcode `\< = 12 \catcode `\\ = 0\relax
       \input \jobname.aux
       \global\AuxReadtrue
     \endgroup }%
    {\global\Reruntrue}}
%    \end{macrocode}
% \end{macro}
% \begin{macro}{\LabelDef}
% A \starTeX{} ".aux" file consists of a sequence of calls on
% \cs{LabelDef}. \cs{LabelDef} just alters the \cs{catcode} of "\" (in
% case there are "\"s in the label), and calls \cs{FetchLabel} to do
% the actual definition.
%    \begin{macrocode}
%    \end{macrocode}
% \end{macro}
% \begin{macro}{\FetchLabel}
% \cs{FetchLabel} has three parameters.
% The label is parameter "#1"; it is enclosed by a set
% of "]"s.\footnote{We can be certain that no \texttt{]} occurs in the
% label.} Parameter "#2" is the referenced page number; it is also
% terminated by a "]". The label reference is parameter "#3"; it is
% terminated by the text ``\texttt{E-o-LabelDef}''.
%    \begin{macrocode}
\def \FetchLabel ]#1]#2]#3E-o-LabelDef{%
    \expandafter\gdef \csname R]#1\endcsname {#3}%
    \expandafter\gdef \csname P]#1\endcsname {#2}%
  \endgroup }
%    \end{macrocode}
% \end{macro}
% The flag \cs{ifAuxRead} tells whether the ".aux" file from the
% previous run has been read.
%    \begin{macrocode}
\newif \ifAuxRead
%    \end{macrocode}
% The flag \cs{ifRerun} is set whenever it is detected that a new
% label has been defined or the definition of a label has been
% modified. This implies that it is necessary to run \starTeX{}
% again.\footnote{It would be nice to be able to 
% do this automatically, but that is not possible with the present
% \TeX.} 
%    \begin{macrocode}
\newif \ifRerun
%    \end{macrocode}
% 
% \subsubsection{Checking the \texttt{.aux} file}
% \begin{macro}{\CheckAux}
% At the end of each run, the \texttt{.aux} file is checked by
% \cs{CheckAux} to see if
% any label were modified during the run. If so, the flag \cs{ifRerun}
% is set.
%    \begin{macrocode}
\def \CheckAux {\immediate\closeout \NewAuxFile
  \begingroup
    \def \LabelDef {\begingroup
        \catcode `\\ = 12 \relax  \CheckLabel }%
    \catcode `\< = 12  \catcode `\\ = 0\relax
    \input \jobname.aux
  \endgroup }
%    \end{macrocode}
% \end{macro}
% \begin{macro}{\CheckLabel}
% \cs{CheckLabel} does the actual checking by comparing the new
% definition with the old one.
%    \begin{macrocode}
\def \CheckLabel ]#1]#2]#3E-o-LabelDef{%
  \ifTextEqual{#2++#3}%
    {\csname P]#1\endcsname++\csname R]#1\endcsname}%
  \else
    \global\Reruntrue
  \fi \endgroup }
%    \end{macrocode}
% \end{macro}
%
% \subsubsection{Referencing labels}
% \begin{macro}{<ref>}
% \stcs{ref} is used to reference a label. It checks the syntax, and
% calls \cs{GiveRef} to produce the actual reference.
%    \begin{macrocode}
\Cdef {ref}{\IfNextChar{[}%
  {\begingroup \catcode `\< = 12 \relax \GiveRef }%
  {\Error{No label referenced;}{the syntax is <ref>[your label].}}}
%    \end{macrocode}
% \end{macro}
% \begin{macro}{\GiveRef}
% \cs{GiveRef} checks whether the reference has been defined, and, if
% it is defined, typesets the reference in the format specified by
% \cs{RefFormat}. 
%    \begin{macrocode}
\def \GiveRef [#1]{\ifAuxRead \else \ReadAuxFile \fi
  \expandafter\ifx \csname R]#1\endcsname \relax
    [label #1]%
    \Error{Label `#1' not defined.}{}\global\Reruntrue
  \else 
    \RefFormat{\csname R]#1\endcsname}{\csname P]#1\endcsname}%
  \fi \endgroup }
%    \end{macrocode}
% \end{macro}
% \begin{macro}{\RefFormat}
% The macro \cs{RefFormat} gives the appearance of the reference. The
% default is a number followed by ``on page
% \emph{x}'' unless the reference is to the same page.\footnote{The
%   page test used will not always give the correct result; if the
%   reference is very near the top or the bottom of a page,
%   \cs{pageno} will sometimes be 1 off. This is not a serious problem,
%   as the referenced object in that case will be very close to the
%   reference.
%
%   This problem was solved in the \texttt{varioref}\cite{varioref} package in
%   \LaTeX, but the solution is somewhat complicated,
%   so I have chosen to stick with the simple solution even if it is
%   not optimal.} 
%    \begin{macrocode}
\def \RefFormat #1#2{#1\ifTextEqual{#2}{\the\pageno}\else 
    \space on page~#2\fi }
%    \end{macrocode}
% \end{macro}
% 
% \subsection{Miscellaneous other commands}
%
% \subsubsection{Symbols and logos}
% \begin{macro}{<>}
% \begin{macro}{<->}
% \begin{macro}{<-->}
% \begin{macro}{<--->}
% \begin{macro}{<``>}
% \begin{macro}{<''>}
% \begin{macro}{<tex>}
% \begin{macro}{<latex>}
% \begin{macro}{<startex>}
% Here is the definition of some symbols that the user needs for
% quality typesetting.
%    \begin{macrocode}
{\catcode`\~ = 12 \Cdef {~}{\nobreak\ }}
\Cdef {}{\null}
\Cdef {-}{\-} \Cdef {--}{--} \Cdef {---}{---}
\Cdef {``}{``} \Cdef {''}{''}
\Cdef {tex}{\TeX} \Cdef {latex}{La\TeX} \Cdef {startex}{Star\TeX}
%    \end{macrocode}
% \end{macro}
% \end{macro}
% \end{macro}
% \end{macro}
% \end{macro}
% \end{macro}
% \end{macro}
% \end{macro}
% \end{macro}
%
% \subsubsection{Error handling}
% \begin{macro}{\Error}
% This macro gives an error message with one or two lines of text.
% (The second line is omitted if it is empty.)
%    \begin{macrocode}
\def \Error #1#2{%
  \message{^^J** StarTeX error detected on line \the\inputlineno:^^J}%
  \message{\space\space\space #1^^J}%
  \ifTextEqual{#2}{}\else \message{\space\space\space #2^^J}\fi}
%    \end{macrocode}
% \end{macro}
% \begin{macro}{\Warning}
% This macro gives a warning with one or two lines of text.
% (The second line is omitted if it is empty.)
%    \begin{macrocode}
\def \Warning #1#2{%
  \message{^^J** StarTeX warning:^^J}%
  \message{\space\space\space #1^^J}%
  \ifTextEqual{#2}{}\else \message{\space\space\space #2^^J}\fi}
%    \end{macrocode}
% \end{macro}
% We need to modify the \cs{newlinechar} in order to produce a line
% feed when we want it. The \textsc{lf} character (\textsc{ascii} code
% 10 = \texttt{\symbol{94}\symbol{94}J}) seems a good choice.
%    \begin{macrocode}
\newlinechar = `^^J
%    \end{macrocode}
%
% \subsubsection{Comments}
% \begin{macro}{<comment>}
% \begin{macro}{</comment>}
% The environment \stenv{comment} is defined to make it possible for
% the user to add comments to his or her \starTeX{} code. It uses
% \cs{SkipComment} to fetch the comment text.
%    \begin{macrocode}
\Cdef {comment}{\NewEnvir{comment}{\begingroup}{\endgroup}%
  \catcode `\< = 12 \SkipComment }
%    \end{macrocode}
% \end{macro}
% \begin{macro}{\SkipComment}
% \cs{SkipComment} works just like \cs{ReadCode} for a \stenv{code}
% environment, except that the code is simply ignored.
%    \begin{macrocode}
{\catcode `\< = 12
  \gdef \SkipComment #1</comment>{\EndEnvir{comment}}}
%    \end{macrocode}
% \end{macro}
% \end{macro}
%
% \subsubsection{Text handling}
% There is very little text handling in \starTeX, but a few
% macros are necessary.
% 
% \begin{macro}{\ifTextEqual}
% The macro \cs{ifTextEqual}\footnote{The name \cs{ifTextEqual} starts
% with a lowercase \cs{if}- to show its relationship to the various
% \cs{if}\dots{} macros of primitive \TeX\@. \cs{ifTextEqual} is
% used just like these, which means that it must have a corresponding
% \cs{fi}. It may also have an \cs{else} part. Note, however, that
% calls on \cs{ifTextEqual} may not occur inside other \cs{if}s.}
% tests if the two texts "#1" and "#2" are equal. 
% This is achieved by letting the \TeX{} primitive \cs{ifx} do the
% job.  
%    \begin{macrocode}
\def \ifTextEqual #1#2{\edef \TmpA {#1}\edef \TmpB {#2}%
  \ifx \TmpA\TmpB }
%    \end{macrocode}
% \end{macro}
%
% \begin{macro}{\IfNextChar}
% Sometimes we need to check the next input character, for instance to
% see if it is a "[". The macro \cs{IfNextChar} expands into "#2" if
% the next input character is "#1"; otherwise it expands into "#3". 
%
% The \TeX{} primitive \cs{futurelet} is used to implement
% \cs{IfNextChar}; it was designed just for this.
% Note that the call on \cs{IfNextChar} must be at the very end of the
% macro in which it occurs.
%    \begin{macrocode}
\def \IfNextChar #1#2#3{\def \TestChar {#1}%
  \def \AltA {#2}\def \AltB {#3}%
  \futurelet \NextChar \TestNextChar }
\def \TestNextChar {%
  \if \NextChar \TestChar \let\Next=\AltA \else \let\Next=\AltB \fi
  \Next }
%    \end{macrocode}
% \end{macro}
% \begin{macro}{\IfNextCharTwo}
% A similar macro testing whether the following character is one of
% two possible alternatives, is also necessary. It takes four
% parameters: "#1" and "#2" are the two characters to check for, and
% the result of \cs{IfNextCharTwo} is "#3" if the next character is
% either of the two alternatives, or else "#4".
%    \begin{macrocode}
\def \IfNextCharTwo #1#2#3#4{\def \TestChar {#1}\def \TestCharX {#2}%
  \def \AltA {#3}\def \AltB {#4}%
  \futurelet \NextChar \TestNextCharTwo }
\def \TestNextCharTwo {%
  \if \NextChar \TestChar  \let \Next = \AltA \else 
  \if \NextChar \TestCharX \let \Next = \AltA \else
    \let \Next = \AltB 
  \fi \fi
  \Next }
%    \end{macrocode}
% \end{macro}
%
% \subsubsection{Numerical operations}
% \begin{macro}{\floatdiv}
% \TeX{} has no floating-point division, so a routine like
% \cs{floatdiv} is necessary. It computes
% $\frac{\mbox{\texttt{\scriptsize  \#1}}\times
% \mbox{\texttt{\scriptsize  \#2}}} 
% {\mbox{\texttt{\scriptsize  \#3}}}$ and places the result in
% \cs{divres}.\footnote{The algorithm used is taken from the code in
% \texttt{epsf.tex}.} All parameters must be \cs{dimen} registers.
%    \begin{macrocode}
\def \floatdiv #1#2#3{\divtemp = #1\divide \divtemp by #3%
  \divres = #2\multiply \divres by \divtemp
  \multiply \divtemp by #3%
  \divrem = #1\advance \divrem by -\divtemp
  \divtemp = #2%
  \loop
    \multiply \divrem by 2  \divide \divtemp by 2
  \ifnum \divtemp > 0
    \ifnum \divrem < #3\else
      \advance \divrem by -#3\advance \divres by \divtemp
    \fi
  \repeat }
%    \end{macrocode}
% \end{macro}
% The necessary registers must be declared:
%    \begin{macrocode}
\newdimen \divrem   \newdimen \divres   \newdimen \divtemp
%    \end{macrocode}
% 
% \subsubsection{File operations}
% \begin{macro}{\ifFileExists}
% To provide better error messages, \starTeX{} tests whether a file
% exists before any attempt is made to read it. \cs{ifFileExists}
% first checks if file "#1" exists;\footnote{This checking is done by
% using \cs{ifeof} which means that empty files will be regarded as
% non-existant; this should not really be a problem.} if it does, the
% file is closed and "#2" is executed, otherwise "#3" is executed. 
%    \begin{macrocode}
\def \IfFileExists #1#2#3{\openin\TestFile = #1
  \ifeof\TestFile \def \Next {#3}\else
  \closein\TestFile \def \Next {#2}\fi
  \Next }
%    \end{macrocode}
% \end{macro}
% \begin{macro}{\TestFile}
% We must also define the input file we use for testing.
%    \begin{macrocode}
\newread \TestFile
%    \end{macrocode}
% \end{macro}
%
% \subsubsection{Format creation}
% \begin{macro}{<make-new-format>}
% \stcs{make-new-format} is used by the system to make a new format; it
% is not intended to be called by the user.
%    \begin{macrocode}
\Cdef {make-new-format}{\dump}
%    \end{macrocode}
% \end{macro}
%
% \begin{macro}{<trace-on>}
% \begin{macro}{<trace-off>}
% \stcs{trace-on} is used during debugging to trace the
% expansion of commands, and \stcs{trace-off} turns this tracing off. 
% Using tracing requires insight into both the inner
% mechanisms of \starTeX{} as well as \TeX{} itself.
%    \begin{macrocode}
\Cdef {trace-on}{\tracingcommands = 1 \tracingmacros = 1
  \tracingrestores = 1 \tracingoutput = 1 \relax }
\Cdef {trace-off}{\tracingcommands = 0 \tracingmacros = 0
  \tracingrestores = 0 \tracingoutput = 0 \relax }
%    \end{macrocode}
% \end{macro}
% \end{macro}
%
% \subsection{Final initialization}
% In the final initialization, we need to load the hyphenation
% patterns defined at this site:
%    \begin{macrocode}
\input startex.lan
%    \end{macrocode}
% We select the default text font:
%    \begin{macrocode}
\ResetFont \SetSize{N}
%    \end{macrocode}
% Our very last initialization is to modify the \cs{catcode}s:
%    \begin{macrocode}
\endinput \SpecialCatCodes
%</code>
%    \end{macrocode}
%
% \section{Language definitions}
% This section describes the file "startex.lan" that defines which
% hyphenation patterns are known to \starTeX\@. It should be adapted
% to the needs of each local installation.
%    \begin{macrocode}
%<*languages>
%    \end{macrocode}
%
% \begin{macro}{\DefLang}
% The macro \cs{DefLang} is used to defined a new language. It has
% three parameters: "#1" is the internal number of the language, "#2"
% is the internal name of the language, and "#3" is the name of the
% file from which to read hyphenation patterns for the language. 
%    \begin{macrocode}
\def \DefLang #1#2#3{\expandafter\def \csname L-#2\endcsname {#1}%
  \language = #1 \input #3}%
%    \end{macrocode}
% \end{macro}
% Only one language (Norwegian) is predefined here, as this file will
% have to be adapted to local needs anyhow.
%    \begin{macrocode}
\DefLang{1}{norsk}{nohyph2.tex}%
%    \end{macrocode}
% \texttt{american} (American English)
% has already been defined as language~0 by plain \TeX.
%    \begin{macrocode}
\expandafter\def \csname L-american\endcsname {0}%
%    \end{macrocode}
% We want American English as our default:
%    \begin{macrocode}
\language = 0
%</languages>
%    \end{macrocode}
%
% \section{Sample document styles}
% Included are also four document styles:
% \begin{description}
% \item[article] is used for ordinary articles. As this
%   is the default definition of \starTeX, this style is empty. 
%
% \item[a4-article] is just an adaption of \emph{article} to fit
% A4~paper. 
%
% \item[ifi-article] is the style used for articles at Ifi
% (``Institutt for informatikk'', the Department of Informatics at the
% Univeristy of Oslo). It uses Concorde as the main text font,
% Palatino for its math, and A4 paper. (If you want to use this style,
% you must have access to a Concorde font; the Palatino Math is
% available at \textsc{ctan}.)
%
% \item[ifi-artikkel] is \emph{ifi-article} adapted to Norwegian text.
% \end{description}
% Each style file contains the necessary redefinitions. 
%
% \subsection{Paper size}
% Every style except \emph{article} uses A4 paper. The margins are set
% to 2.5~cm.
%    \begin{macrocode}
%<*a4-article|ifi-article|ifi-artikkel>
\hoffset = -1in  \advance \hoffset by 2.5cm  \hsize = 16cm
\voffset = -1in  \advance \voffset by 2.5cm  \vsize = 24.7cm
%</a4-article|ifi-article|ifi-artikkel>
%    \end{macrocode}
% \subsection{Language adaption}
% The \emph{ifi-artikkel} is for text in Norwegian. We must select the
% correct language and adjust \cs{lefthyphenmin} and
% \cs{righthyphenmin}. 
%    \begin{macrocode}
%<*ifi-artikkel>
\language = \csname L-norsk\endcsname
\lefthyphenmin = 1 \righthyphenmin = 2
%    \end{macrocode}
% \begin{macro}{\AbstractName}
% \begin{macro}{\FigureName}
% \begin{macro}{\TableName}
% \begin{macro}{\TimeSep}
% Some names must also be redefined.
%    \begin{macrocode}
\def \AbstractName {Sammendrag}%
\def \FigureName   {Figur}%
\def \TableName    {Tabell}%
\def \TimeSep      {.}%
%    \end{macrocode}
% \end{macro}
% \end{macro}
% \end{macro}
% \end{macro}
% \begin{macro}{\RefFormat}
% The format for references must be translated into Norwegian.
%    \begin{macrocode}
\def \RefFormat #1#2{#1\ifTextEqual{#2}{\the\pageno}\else
  \space p\aa{} side~#2\fi }%
%    \end{macrocode}
% \end{macro}
% \begin{macro}{<today>}
% \begin{macro}{\Month}
% Finally, the date command \stcs{today} must be redefined.
% \begin{macrocode}
\Cdef {today}{\the\day.\space \Month\space \the\year }%
\def \Month {\ifcase \month \or januar\or februar\or mars\or
  april\or mai\or juni\or juli\or august\or september\or
  oktober\or november\or desember\fi }%
%</ifi-artikkel>
%    \end{macrocode}
% \end{macro}
% \end{macro}
% 
% \subsection{Text font}
% The two \emph{ifi-} styles use Concorde\footnote{In the Karl Berry
% font name scheme \cite{fontname}, Concorde is known as \emph{poc-}.} as
% the main body text 
% font; it is the official font of the University of Oslo. Since
% Concorde is a wider font than \textsc{ecr}, and because it has a
% larger x-height, a smaller font size is used. The EC 
% Companion font is kept for the special symbols. 
%    \begin{macrocode}
%<*ifi-article|ifi-artikkel>
\FontDef{R}{M}{U}{N}{pocr8t at 9.8pt}{tcrm1095}%
\FontDef{R}{M}{I}{N}{pocri8t at 9.8pt}{tcti1095}%
\FontDef{R}{B}{U}{N}{pocb8t at 9.8pt}{tcbx1095}%
\FontDef{R}{B}{I}{N}{pocbi8t at 9.8pt}{tcbi1095}%
\FontDef{T}{M}{U}{N}{pcrr8t at 9.8pt}{tctt1095}%
\FontDef{T}{M}{I}{N}{pcrro8t at 9.8pt}{tcit1095}%
\FontDef{T}{B}{U}{N}{pcrb8t at 9.8pt}{tctt1095}%
\FontDef{T}{B}{I}{N}{pcrbo8t at 9.8pt}{tcit1095}%
\FontDef{R}{M}{U}{L}{pocr8t at 12.9pt}{tcrm1440}%
\FontDef{R}{M}{I}{L}{pocri8t at 12.9pt}{tcti1440}%
\FontDef{R}{B}{U}{L}{pocb8t at 12.9pt}{tcbx1440}%
\FontDef{R}{B}{I}{L}{pocbi8t at 12.9pt}{tcbi1440}%
\FontDef{T}{M}{U}{L}{pcrr8t at 12.9pt}{tctt1440}%
\FontDef{T}{M}{I}{L}{pcrro8t at 12.9pt}{tcit1440}%
\FontDef{T}{B}{U}{L}{pcrb8t at 12.9pt}{tctt1440}%
\FontDef{T}{B}{I}{L}{pcrbo8t at 12.9pt}{tcit1440}%
\FontDef{R}{M}{U}{X}{pocr8t at 15.5pt}{tcrm1728}%
\FontDef{R}{M}{I}{X}{pocri8t at 15.5pt}{tcti1728}%
\FontDef{R}{B}{U}{X}{pocb8t at 15.5pt}{tcbx1728}%
\FontDef{R}{B}{I}{X}{pocbi8t at 15.5pt}{tcbi1728}%
\FontDef{T}{M}{U}{X}{pcrr8t at 15.5pt}{tctt1728}%
\FontDef{T}{M}{I}{X}{pcrro8t at 15.5pt}{tcit1728}%
\FontDef{T}{B}{U}{X}{pcrb8t at 15.5pt}{tctt1728}%
\FontDef{T}{B}{I}{X}{pcrro8t at 15.5pt}{tcit1728}%
\FontDef{R}{M}{U}{S}{pocr8t at 8.9pt}{tcrm1000}%
\FontDef{R}{M}{I}{S}{pocri8t at 8.9pt}{tcti1000}%
\FontDef{R}{B}{U}{S}{pocb8t at 8.9pt}{tcbx1000}%
\FontDef{R}{B}{I}{S}{pocbi8t at 8.9pt}{tcbi1000}%
\FontDef{T}{M}{U}{S}{pcrr8t at 8.9pt}{tctt1000}%
\FontDef{T}{M}{I}{S}{pcrro8t at 8.9pt}{tcit1000}%
\FontDef{T}{B}{U}{S}{pcrb8t at 8.9pt}{tctt1000}%
\FontDef{T}{B}{I}{S}{pcrbo8t at 8.9pt}{tcit1000}%
%    \end{macrocode}
% The ``Math Palatino'' font is used for math.
% \changes{1.04}{1999/02/25}{Changed math font to ``Palatino Math''
%   for the Ifi document styles.}
%    \begin{macrocode}
\def \XIIpt  { at 12pt}
\def \Xpt    { at 10pt}
\def \IXpt   { at 9pt}
\def \VIIIpt { at 8pt}
\def \VIIpt  { at 7pt}
\def \VIpt   { at 6pt}
\def \Vpt    { at 5pt}
\MathFontDef{N}{0}{zppler7t\XIpt}{zppler7t\VIIIpt}{zppler7t\VIpt}
\MathFontDef{N}{1}{zppler7m\XIpt}{zppler7m\VIIIpt}{zppler7m\VIpt}
\MathFontDef{N}{2}{zppler7y\XIpt}{zppler7y\VIIIpt}{zppler7y\VIpt}
\MathFontDef{N}{3}{zppler7v\XIpt}{zppler7v\XIpt}{zppler7v\XIpt}
\MathFontDef{N}{4}{zppleb7t\XIpt}{zppleb7t\VIIIpt}{zppleb7t\VIpt}
\MathFontDef{N}{5}{pplri8t\XIpt}{pplri8t\VIIIpt}{pplri8t\VIpt}
\MathFontDef{L}{0}{zppler7t\XIVpt}{zppler7t\Xpt}{zppler7t\VIIIpt}
\MathFontDef{L}{1}{zppler7m\XIVpt}{zppler7m\Xpt}{zppler7m\VIIIpt}
\MathFontDef{L}{2}{zppler7y\XIVpt}{zppler7y\Xpt}{zppler7y\VIIIpt}
\MathFontDef{L}{3}{zppler7v\XIVpt}{zppler7v\XIVpt}{zppler7v\XIVpt}
\MathFontDef{L}{4}{zppleb7t\XIVpt}{zppleb7t\Xpt}{zppleb7t\VIIIpt}
\MathFontDef{L}{5}{pplri8t\XIVpt}{pplri8t\Xpt}{pplri8t\VIIIpt}
\MathFontDef{X}{0}{zppler7t\XVIIpt}{zppler7t\XIIpt}{zppler7t\IXpt}
\MathFontDef{X}{1}{zppler7m\XVIIpt}{zppler7m\XIIpt}{zppler7m\IXpt}
\MathFontDef{X}{2}{zppler7y\XVIIpt}{zppler7y\XIIpt}{zppler7y\IXpt}
\MathFontDef{X}{3}{zppler7v\XVIIpt}{zppler7v\XVIIpt}{zppler7v\XVIIpt}
\MathFontDef{X}{4}{zppleb7t\XVIIpt}{zppleb7t\XIIpt}{zppleb7t\IXpt}
\MathFontDef{X}{5}{pplri8t\XVIIpt}{pplri8t\XIIpt}{pplri8t\IXpt}
\MathFontDef{S}{0}{zppler7t\Xpt}{zppler7t\VIIpt}{zppler7t\Vpt}
\MathFontDef{S}{1}{zppler7m\Xpt}{zppler7m\VIIpt}{zppler7m\Vpt}
\MathFontDef{S}{2}{zppler7y\Xpt}{zppler7y\VIIpt}{zppler7y\Vpt}
\MathFontDef{S}{3}{zppler7v\Xpt}{zppler7v\Xpt}{zppler7v\Xpt}
\MathFontDef{S}{4}{zppleb7t\Xpt}{zppleb7t\VIIpt}{zppleb7t\Vpt}
\MathFontDef{S}{5}{pplri8t\Xpt}{pplri8t\VIIpt}{pplri8t\Vpt}
%    \end{macrocode}
% We must select the newly selected standard font.
%    \begin{macrocode}
\SelectFont
%</ifi-article|ifi-artikkel>
%    \end{macrocode}
%
% \bibliography{startex}
% \bibliographystyle{plain}
%
% \clearpage 
% \Finale
\endinput