% \iffalse meta-comment
%
% Copyright 1993 1994 1995 1996 1997 1998 1999
% The LaTeX3 Project and any individual authors listed elsewhere
% in this file.
%
% This file is part of the LaTeX2e system.
% ----------------------------------------
%
% It may be distributed under the terms of the LaTeX Project Public
% License, as described in lppl.txt in the base LaTeX distribution.
% Either version 1.0 or, at your option, any later version.
%
% \fi
%
% \iffalse
%%% From File: ltspace.dtx
%<*driver>
% \fi
\ProvidesFile{ltspace.dtx}
[1998/08/17 v1.2w LaTeX Kernel (spacing)]
% \iffalse
\documentclass{ltxdoc}
\GetFileInfo{ltspace.dtx}
\title{\filename}
\date{\filedate}
\author{%
Johannes Braams\and
David Carlisle\and
Alan Jeffrey\and
Leslie Lamport\and
Frank Mittelbach\and
Chris Rowley\and
Rainer Sch\"opf}
\begin{document}
\maketitle
\DocInput{\filename}
\end{document}
%</driver>
% \fi
%
% \CheckSum{424}
%
% \changes{v1.1a}{1994/05/16}{(ASAJ) Split from ltinit.dtx.}
% \changes{v1.2f}{1995/05/25}{Macros moved to ltlists.dtx}
% \changes{v1.2n}{1996/04/22}{Documentation Improvements}
% \changes{v1.2o}{1996/06/22}{Documentation of problems added}
% \changes{v1.2q}{1996/07/27}{Further documentation of problems}
% \changes{v1.2r}{1996/07/27}{Correct documentation of problems}
% \changes{v1.2w}{1998/08/17}{Documentation fixes.}
%
%
% \section{Spacing}
%
% This section deals with spacing, and line- and page-breaking.
%
% \subsection{User Commands}
%
% \DescribeMacro\nopagebreak\oarg{i} : \meta{i} = 0,...,4.
%
% Default argument = 4. Puts a penalty into the vertical list
% output as follows:\\
% 0 : penalty = 0\\
% 1 : penalty = |\@lowpenalty|\\
% 2 : penalty = |\@medpenalty|\\
% 3 : penalty = |\@highpenalty|\\
% 4 : penalty = 10000
%
% \DescribeMacro\pagebreak\oarg{i} :
% same as \nopagebreak except negatives of its penalty
%
% \DescribeMacro\linebreak\oarg{i} : analog of the above
%
% \DescribeMacro\nolinebreak\oarg{i} : analog of the above
%
% \DescribeMacro\samepage :
% inhibits page breaking most places by setting the
% following penalties to 10000:\\
% |\interlinepenalty|\\
% |\postdisplaypenalty|\\
% |\interdisplaylinepenalty|\\
% |\@beginparpenalty|\\
% |\@endparpenalty|\\
% |\@itempenalty|\\
% |\@secpenalty|\\
% |\interfootnotelinepenalty|
%
% \DescribeMacro{\\} : initially defined to be |\newline|
%
% |\\|\oarg{length} : initially defined to be
% |\vspace|\marg{length}|\newline|\\
% Note: |\\*| adds a |\vadjust{\penalty 10000}|
%
% OBSOLETE COMMANDS (which never made it into the manual):
%
% |\obeycr| : defines <CR> == |\\\relax|\\
% |\restorecr| : restores <CR> to its usual meaning.
%
%
% \StopEventually{}
%
% \subsection{Chris' comments}
%
% There are several aspects of the handling of space in horizontal
% mode that are inconsistent or do not work well in some cases.
% These are largely concerned with ignoring the effect of space
% tokens that would otherwise typeset an inter-word space.
%
% Negating the effect of such space tokens is achieved by two
% mechanisms:
% \begin{itemize}
% \item |\unskip| is used to remove the glue just added by
% a space that has already had its effect; it is sometimes
% invoked after an |\ifdim| test on |\lastskip| (see below);
% \item |\ignorespaces| is used to ignore space-tokens yet to come.
% \end{itemize}
%
% The test done on |\lastskip| is sometimes for equality with zero and
% sometimes for being positive. Recall also that the test is only on
% the natural length of the glue and that no glue cannot be
% distinguished from glue whose natural length is zero: to summarise,
% a pretty awful test. It is not clear why these tests are not all
% the same; I think that they should all be for equality. One place
% where |\unskip| is often used is just before a |\par| (which itself
% internally does an |\unskip|) and one bit of code (in |\@item|) even
% has two |\unskips| before a |\par|. These uses may be fossil code
% but if they are necessary, maybe |\@killglue| would be even safer.
%
% Such removal of glue by |\unskip| may sometimes have the wrong result,
% removing not the glue from a space-token but other explicit glue;
% this is sometimes not what is intended.
%
% A common way to prevent such removal is to add an |\hskip\z@| after
% the glue that should not be removed. This protects that glue
% against one |\unskip| with no test but not against more than one.
% It does work for `tested |\unskip|s'. This is used
% by |\hspace*| but not by |\hspace|; this is inconsistent as the star
% is supposed to prevent removal only at the beginning of a line, not
% at the end, or in a tabular, etc.
%
% If this reason for removing glue were the only consideration then a
% tested-|\unskip| and protection by |\hskip\z@| would suffice but
% would need to be consistently implemented.
%
% However, the class of invisibles, commands and environments tries to
% be even cleverer: one of these tries to leave only one inter-word
% space whenever there is one before it and one after it; and it does
% this quite well.
%
% But problems can arise when there is not a space-token on
% both sides of it; in particular, when an invisible appears at the
% beginning or end of a piece of text the method still leaves one space
% token whereas usually in these cases it should leave none.
%
% Also, the current rules do not work well when more than one such
% command appears consecutively, separated by space-tokens; it leaves
% glue between every other invisible.
%
% There is also a question about what these commands should do when
% they occur next to spaces that do not come from space tokens but,
% for example, from |\hspace|. Should they still produce `just one
% space'? If so, which one? It is good to note that the manual
% is sufficiently cautious about invisibles that we are not obliged to
% make anything work.
%
% Another interesting side-road to explore is whether the space-tokens
% either side of an |\hspace{...}| should be ignored.
%
% One alternative to the current algorithm that is often suggested is
% that all glue around the invisible should be consolidated into a
% space after it (usually without stating how much glue should be put
% there). The command |\nolinebreak| is implemented this way (and
% |\linebreak| should also be). This does not work correctly for the
% following common case:
% \begin{verbatim}
% ... some text
% \index{some-word}
% some-word and more text.
% \end{verbatim}
% This is optimal coding since it is normal to index a word that gets
% split across a page-break on its starting page. This would, on the
% other hand, fix another common (and documented) failure of the
% current system: when the invisible is the last thing in a paragraph
% the space before it is not removed and, worse, it is also hidden
% from the paragraph-ending mechanism so that an `empty' line can be
% created at the end of the paragraph.
%
% Another deficiency (I think) of the current system is that the
% following is treated as having the |\index| command between the
% paragraphs, which is probably not what the author intended (since
% there is no empty line after it).
% \begin{verbatim}
%
% \index{beginnings}
% Beginnings of paragraphs ...
% \end{verbatim}
%
% I know of no algorithm that will handle satisfactorily even
% all the most common cases; note that it could be that the best
% algorithm may be different for different invisibles since,
% for example, the common uses and expected behaviour of
% |\index|, |\marginpar|, |\linebreak|, |\pagebreak| and
% |\vspace| are somewhat different. [For example, is
% |\vspace| ever used in the middle of a paragraph?]
%
% One method that can (and is) used to make invisible commands produce
% no space when used at the beginning of text is to put in some glue
% that is nearly enough the same as no glue or glue of zero length in
% all respects except for the precise test for not being exactly equal
% to zero; examples of such glue are |\hskip 1sp| and, possibly better
% but more complex, |\hskip -1sp \hskip 1sp|. However, this only works
% when it is known that user-supplied text is about to start.
%
% Some similar concerns apply to the handling of space and penalties
% in vertical mode; there is an extra hurdle here as |\unskip| does
% not work on the main vertical list. The complexity of the tests done
% by |\addvspace| have never been explained.
%
% The implementation of space hacks etc for vertical mode is another
% major area that needs further attention; my earlier experiments
% did not produce much improvement over the current unsatisfactory
% situation.
%
% One particular problem is what happens when the following very
% natural coding is used (part of the problem here is that this looks
% like an hmode problem, but it is not):
% \begin{verbatim}
% ... end of text.
%
% \begin{enumerate}
% \item \label{item:xxx} Item text.
% \end{enumerate}
% \end{verbatim}
%
% \subsection{Some immediate actions}
%
% \begin{itemize}
% \item Fix bug in |\linebreak|.
% \item Fix bug in |\\*|.
% \item Reimplement |\\|, etc, removing extra |\vadjust|s and getting
% better error trapping (this seems to involve a lot more tokens).
% \item Investigate whether |\\|, etc need to be errors in vmode; I
% think that they could be noops (maybe with a warning).
% \item Make all(?) |\unskips| include test for zero skip (rather than
% other tests or no test).
% \item Consider replacing |\hskip 1sp| by something better (here
% called an `infinitesimal' skip).
% \item Look at all |\hskip\z@| (or similar) to see if they should be
% changed to an `infinitesimal' skip.
% \item Resolve the inconsistency between |\hspace| and |\hspace*|.
% \item Remove unnecessary |\unskips|.
% \item Investigate and rationalise the `newline' code.
% \item Find better algorithms for all sorts of things or, easier(?),
% fix \TeX{} itself.
% \end{itemize}
%
% \subsection{The code}
%
% \begin{macrocode}
%<*2ekernel>
\message{spacing,}
% \end{macrocode}
%
% \begin{macro}{\pagebreak}
% \begin{macro}{\nopagebreak}
% \changes{v1.2h}{1995/07/05}{Reimplemented both using \cs{@no@pgbk}}
% \changes{v1.2j}{1995/10/16}{(DPC) Use \cs{@testopt} /1911}
% \begin{macrocode}
\def\pagebreak{\@testopt{\@no@pgbk-}4}
\def\nopagebreak{\@testopt\@no@pgbk4}
% \end{macrocode}
% \end{macro}
% \end{macro}
%
%
% \begin{macro}{\@no@pgbk}
% \changes{v1.2h}{1995/07/05}{Macro replaces \cs{@pgbk}
% and \cs{@nopgbk}}
% \begin{macrocode}
\def\@no@pgbk #1[#2]{%
\ifvmode
\penalty #1\@getpen{#2}%
\else
\@bsphack
\vadjust{\penalty #1\@getpen{#2}}%
\@esphack
\fi}
% \end{macrocode}
% \end{macro}
%
% \begin{macro}{\linebreak}
% \begin{macro}{\nolinebreak}
% \changes{v1.2u}{1996/10/29}{Reimplemented both using \cs{@no@lnbk}}
% \changes{v1.2j}{1995/10/16}{(DPC) Use \cs{@testopt} /1911}
% \begin{macrocode}
\def\linebreak{\@testopt{\@no@lnbk-}4}
\def\nolinebreak{\@testopt\@no@lnbk4}
% \end{macrocode}
% \end{macro}
% \end{macro}
% \begin{macro}{\@no@lnbk}
% \changes{v1.2u}{1996/10/29}{Macro replaces \cs{@lnbk}
% and \cs{@nolnbk}}
% \begin{macrocode}
\def\@no@lnbk #1[#2]{%
\ifvmode
\@nolnerr
\else
\@tempskipa\lastskip
\unskip
\penalty #1\@getpen{#2}%
\ifdim\@tempskipa>\z@
\hskip\@tempskipa
\ignorespaces
\fi
\fi}
% \end{macrocode}
% \end{macro}
%
% \begin{macro}{\samepage}
% \begin{macrocode}
\def\samepage{\interlinepenalty\@M
\postdisplaypenalty\@M
\interdisplaylinepenalty\@M
\@beginparpenalty\@M
\@endparpenalty\@M
\@itempenalty\@M
\@secpenalty\@M
\interfootnotelinepenalty\@M}
% \end{macrocode}
% \end{macro}
%
%
% \begin{macro}{\\}
% \changes{v1.2a}{1994/11/11}{(DPC) Make robust}
% \changes{v1.2d}{1994/11/14}{(DPC) Macro modified}
%
% \changes{v1.2u}{1996/10/29}{Corrected and rationalised code}
% The purpose of the new code is to fix a few bugs; however, it also
% attempts to optimize the following, in order of priority:
% \begin{enumerate}
% \item efficient execution of plain |\\|;
% \item efficient execution of |\\[...]|;
% \item memory use;
% \item name-space use.
% \end{enumerate}
% The changes should make no difference to the typeset output.
% It appears to be safe to use |\reserved@e| and |\reserved@f| here
% (other reserved macros are somewhat disastrous).
%
% These changes made |\newline| even less robust than it had been,
% so now it is explicitly robust, like |\\|.
% \begin{macro}{\@normalcr}
% The internal definition of the `normal' definition of |\\|.
% \begin{macrocode}
\DeclareRobustCommand\\{%
\let \reserved@e \relax
\let \reserved@f \relax
\@ifstar{\let \reserved@e \vadjust \let \reserved@f \nobreak
\@xnewline}%
\@xnewline}
\expandafter\let\expandafter\@normalcr
\csname\expandafter\@gobble\string\\ \endcsname
% \end{macrocode}
% \end{macro}
% \end{macro}
% \begin{macro}{\newline}
% A simple form of the `normal' definition of |\\|.
% \changes{v1.2v}{1997/05/07}{Made completely robust.}
% \begin{macrocode}
\DeclareRobustCommand\newline{\@normalcr\relax}
% \end{macrocode}
% \end{macro}
%
% \begin{macro}{\@xnewline}
% \begin{macrocode}
\def\@xnewline{\@ifnextchar[% ] bracket matching
\@newline
{\@gnewline\relax}}
% \end{macrocode}
% \end{macro}
%
% \begin{macro}{\@newline}
% \begin{macrocode}
\def\@newline[#1]{\let \reserved@e \vadjust
\@gnewline {\vskip #1}}
% \end{macrocode}
% \end{macro}
%
% \begin{macro}{\@gnewline}
% \changes{v1.2u}{1996/10/29}{Added macro}
% The |\nobreak| added to prevent null lines when |\\|
% ends an overfull line. Change made 24 May 89 as suggested by
% Frank Mittelbach and Rainer Sch\"opf
% \changes{v1.2h}{1995/07/05}{Use \cs{break}}
% \begin{macrocode}
\def\@gnewline #1{%
\ifvmode
\@nolnerr
\else
\unskip \reserved@e {\reserved@f#1}\nobreak \hfil \break
\fi}
% \end{macrocode}
% \end{macro}
%
%
% \begin{macro}{\@getpen}
% \begin{macrocode}
\def\@getpen#1{\ifcase #1 \z@ \or \@lowpenalty\or
\@medpenalty \or \@highpenalty
\else \@M \fi}
% \end{macrocode}
% \end{macro}
%
%
% \begin{macro}{\if@nobreak}
% \changes{v1.2p}{1996/07/26}{put \cs{global} inside definition}
% Switch used to avoid page breaks caused by |\label| after a
% section heading, etc. It should be \textbf{GLOBALLY} set true
% after the |\nobreak| and \textbf{globally} set false by the
% next invocation of |\everypar|.
%
% Commands that reset |\everypar| should globally set it false if
% appropriate.
% \begin{macrocode}
\def\@nobreakfalse{\global\let\if@nobreak\iffalse}
\def\@nobreaktrue {\global\let\if@nobreak\iftrue}
\@nobreakfalse
% \end{macrocode}
% \end{macro}
%
%
%
% \begin{macro}{\@savsk}
% \begin{macro}{\@savsf}
% Registers used to save the space factor and last skip.
% \begin{macrocode}
\newdimen\@savsk
\newcount\@savsf
% \end{macrocode}
% \end{macro}
% \end{macro}
%
%
%
%
% \begin{macro}{\@bsphack}
% \changes{LaTeX2e}{1993/12/08}
% {Command reimplemented; late birthday present for Chris}
% \changes{LaTeX2e}{1993/12/08}{Command reimplemented}
% \changes{LaTeX2e}{1993/12/16}{Corrected optimisation :-)}
% |\@bsphack| and |\@esphack|
% used by macros such as |\index| and
% |\begin{@float}| \ldots |\end{@float}|
% that want to be invisible --- i.e.,
% not leave any extra space when used in the middle of text. Such
% a macro should begin with |\@bsphack| and end with |\@esphack|
% The macro in question should not create any text, nor change the
% mode.
%
% Before giving the current definition we give an extended definition
% that is currently not used (because it doesnt work as advertised:-)
%
% These are generalised hacks which attempt to do sensible things
% when `invisible commands' appear in vmode too.
%
% They need to cope with space in both hmode (plus spacefactor) and
% vmode, and also cope with breaks etc. In vmode this means
% ensuring that any following |\addvspace|, etc sees the correct
% glue in |\lastskip|.
%
% In fact, these improved versions should be used for other cases
% of `whatsits, thingies etc' which should be invisible. They are
% only for commands, not environments (see notes on |\@Esphack|).
%
% BTW, anyone know why the standard hacks are surrounded by
% |\ifmmode\else| rather than simply |\ifhmode|?
%
% And are there any cases where saving the spacefactor is
% essential? I have some extensions where it is, but it does not
% appear to be so in the standard uses.
%\begin{verbatim}
%\def \@bsphack{%
% \relax \ifvmode
% \@savsk \lastskip
% \ifdim \lastskip=\z@
% \else
% \vskip -\lastskip
% \fi
% \else
% \ifhmode
% \@savsk \lastskip
% \@savsf \spacefactor
% \fi
% \fi
%}
%\end{verbatim}
% I think that, in vmode, it is the safest to put
% in a |\nobreak| immediately after such things since writes,
% inserts etc followed by glue give valid breakpoints and, in
% general, it is possible to create breaks but impossible to
% destroy them.
%\begin{verbatim}
%\def \@esphack{%
% \relax \ifvmode
% \nobreak
% \ifdim \@savsk=\z@
% \else
% \vskip\@savsk
% \fi
% \else
% \ifhmode
% \spacefactor \@savsf
% \ifdim \@savsk>\z@
% \ignorespaces
% \fi
% \fi
% \fi
%}
%\end{verbatim}
% For the moment we are going to ignore the vertical versions until
% they are correct.
% \changes{LaTeX2e}{1993/12/19}{There seem to be problems with selfmade
% birthday presents}
% \begin{macrocode}
\def\@bsphack{%
\relax
\ifhmode
\@savsk\lastskip
\@savsf\spacefactor
\fi}
% \end{macrocode}
% \end{macro}
%
% \begin{macro}{\@esphack}
% Companion to |\@bsphack|.
% \begin{macrocode}
\def\@esphack{%
\relax
\ifhmode
\spacefactor\@savsf
\ifdim\@savsk>\z@
\ignorespaces
\fi
\fi}
% \end{macrocode}
% \end{macro}
%
% \begin{macro}{\@Esphack}
% A variant of |\@esphack| that sets the |@ignore| switch to
% true (as |\@esphack| used to do previously).
% This is currently used only for floats and similar environments.
% \changes{v1.2s}{1996/08/02}{Remove \cs{global} before \cs{@ignore...}}
% \begin{macrocode}
\def\@Esphack{%
\relax
\ifhmode
\spacefactor\@savsf
\ifdim\@savsk>\z@
\@ignoretrue
\ignorespaces
\fi
\fi}
% \end{macrocode}
% \end{macro}
%
%
% \begin{macro}{\@vbsphack}
% \changes{LaTeX2e}{1993/12/08}{Command added}
% Another variant which is useful for invisible things which should
% not live in vmode (this is how some people feel about marginals).
%
% If it occurs in vmode then it enters hmode and ensures that
% |\@savsk| is nonzero so that the |\ignorespaces| is put in later.
% It is not used at present.
% \changes{v1.2f}{1995/05/25}{(CAR) not used so `removed'.}
%\begin{verbatim}
% \def \@vbsphack{ %
% \relax \ifvmode
% \leavevmode
% \@savsk 1sp
% \@savsf \spacefactor
% \else
% \ifhmode
% \@savsk \lastskip
% \@savsf \spacefactor
% \fi
% \fi
% }
%\end{verbatim}
% \end{macro}
%
%
% \subsection{Vertical spacing}
%
%
% \LaTeX\ supports the plain \TeX\ commands
% |\smallskip|, |\medskip| and |\bigskip|.
% However, it redefines them using |\vspace| instead of |\vskip|.
%
% Extra vertical space is added by the command
% |\addvspace|\marg{skip},
% which adds a vertical skip of \meta{skip} to the document.
% The sequence\\
% |\addvspace|\marg{s1} |\addvspace|\marg{s2}
% is equivalent to\\
% |\addvspace|\marg{maximum of s1, s2}.
%
% |\addvspace| should be used only in vertical mode, and gives an
% error if it's not. The |\addvspace| command does \emph{not} add
% vertical space if |@minipage| is true. The minipage environment uses
% this to inhibit the addition of extra vertical space at the beginning.
%
% Penalties are put into the vertical list with the
% |\addpenalty|\marg{penalty}
% command. It works properly when |\addpenalty| and |\addvspace|
% commands are mixed.
%
% The |@nobreak| switch is set true used when in vertical mode and no
% page break should occur. (Right now, it is used only by the section
% heading commands to inhibit page breaking after a heading.)
%
%
%\begin{verbatim}
% \addvspace{SKIP} ==
% BEGIN
% if vmode
% then if @minipage
% else if \lastskip =0
% then \vskip SKIP
% else if \lastskip < SKIP
% then \vskip -\lastskip
% \vskip SKIP
% else if SKIP < 0 and \lastskip >= 0
% then \vskip -\lastskip
% \vskip \lastskip + SKIP
% fi fi fi fi
% else useful error message (CAR).
% fi
% END
%\end{verbatim}
%
% \begin{macro}{\@xaddvskip}
% Internal macro for |\vspace| handling the case that space has
% previously been added.
% \begin{macrocode}
\def\@xaddvskip{%
\ifdim\lastskip<\@tempskipb
\vskip-\lastskip
\vskip\@tempskipb
\else
\ifdim\@tempskipb<\z@
\ifdim\lastskip<\z@
\else
\advance\@tempskipb\lastskip
\vskip-\lastskip
\vskip \@tempskipb
\fi
\fi
\fi}
% \end{macrocode}
% \end{macro}
%
%
% \begin{macro}{\addvspace}
% \changes{v1.2b}{1994/11/12}{Corrected error message}
% \changes{v1.2c}{1994/11/13}{Recorrected error message}
% Add vertical space taking into account space already added, as
% described above.
% \begin{macrocode}
\def\addvspace#1{%
\ifvmode
\if@minipage\else
\ifdim \lastskip =\z@
\vskip #1\relax
\else
\@tempskipb#1\relax
\@xaddvskip
\fi
\fi
\else
\@noitemerr
\fi}
% \end{macrocode}
% \end{macro}
%
% \begin{macro}{\addpenalty}
% \changes{v1.2b}{1994/11/12}{Corrected error message}
% \changes{v1.2c}{1994/11/13}{Recorrected error message}
% \begin{macrocode}
\def\addpenalty#1{%
\ifvmode
\if@minipage
\else
\if@nobreak
\else
\ifdim\lastskip=\z@
\penalty#1\relax
\else
\@tempskipb\lastskip
\vskip -\lastskip
\penalty#1%
\vskip\@tempskipb
\fi
\fi
\fi
\else
\@noitemerr
\fi}
% \end{macrocode}
% \end{macro}
%
% \begin{macro}{\vspace}
% \changes{v1.2m}{1996/01/20}{Made robust}
% \begin{macro}{\@vspace}
% \begin{macro}{\@vspacer}
% \changes{v1.2f}{1995/05/25}
% {(CAR) macros modified to be more efficient}
% \changes{v1.2f}{1995/05/25}{(CAR) \cs{@restorepar} added to avoid
% possible infinite tail recursion caused by a typo in the argument.}
% The new code for these commands depends on the following facts:
% \begin{itemize}
% \item The value of prevdepth is changed only when a box or rule
% is created and added to a vertical list;
% \item The value of prevdepth is used only when a box is created
% and added to a vertical list;
% \item The value of prevdepth is always local to the building of
% one vertical list.
% \end{itemize}
% \begin{macrocode}
\DeclareRobustCommand\vspace{\@ifstar\@vspacer\@vspace}
\def\@vspace #1{%
\ifvmode
\vskip #1
\vskip\z@skip
\else
\@bsphack
\vadjust{\@restorepar
\vskip #1
\vskip\z@skip
}%
\@esphack
\fi}
% \end{macrocode}
%
% \begin{macrocode}
\def\@vspacer#1{%
\ifvmode
\dimen@\prevdepth
\hrule \@height\z@
\nobreak
\vskip #1
\vskip\z@skip
\prevdepth\dimen@
\else
\@bsphack
\vadjust{\@restorepar
\hrule \@height\z@
\nobreak
\vskip #1
\vskip\z@skip}%
\@esphack
\fi}
% \end{macrocode}
% \end{macro}
% \end{macro}
% \end{macro}
%
% \begin{macro}{\smallskip}
% \begin{macro}{\medskip}
% \begin{macro}{\bigskip}
% \begin{macrocode}
\def\smallskip{\vspace\smallskipamount}
\def\medskip{\vspace\medskipamount}
\def\bigskip{\vspace\bigskipamount}
% \end{macrocode}
% \end{macro}
% \end{macro}
% \end{macro}
%
% \begin{macro}{\smallskipamount}
% \begin{macro}{\medskipamount}
% \begin{macro}{\bigskipamount}
% \begin{macrocode}
\newskip\smallskipamount \smallskipamount=3pt plus 1pt minus 1pt
\newskip\medskipamount \medskipamount =6pt plus 2pt minus 2pt
\newskip\bigskipamount \bigskipamount =12pt plus 4pt minus 4pt
% \end{macrocode}
% \end{macro}
% \end{macro}
% \end{macro}
%
%
%
% \subsection{Horizontal space}
%
% \begin{macro}{\nobreakspace}
% \changes{v1.2k}{1995/12/04}{(Macro added}
% \begin{macro}{\@xobeysp}
% \changes{v1.2t}{1996/09/28}{Moved from ltmiscen.dtx and redefined to
% use \cs{nobreakspace }}
%
% This is a robust command that produces a horizontal space at
% which, in paragraph-mode, a line-break is not possible. We then
% define an active |~| to expand to it since this is the documented
% behaviour of |~|. One reason for introducing this is that some
% 8-bit input encodings have a slot for such a space and we do not
% want to use active characters as the \LaTeX{} internal commands.
%
% The braces in the definition of |~| are needed to ensure that a
% following space is preserved when reading to/from internal files.
% \changes{v1.2l}{1995/12/04}{(braces added to definition of tilde}
%
% We need to keep \cs{@xobeysp} as it is widely used; so here it is
% let to the non-robust command \cs{nobreakspace }.
%
% \begin{macrocode}
\DeclareRobustCommand{\nobreakspace}{%
\leavevmode\nobreak\ }
\catcode `\~=13
\def~{\nobreakspace{}}
\expandafter\let\expandafter\@xobeysp\csname nobreakspace \endcsname
% \end{macrocode}
% \end{macro}
% \end{macro}
%
% \begin{macrocode}
% \end{macrocode}
%
%
% \begin{macro}{\,}
% Used in paragraph mode produces a |\thinspace|. It has the
% ordinary definition in math mode. Useful for quotes inside quotes,
% as in |``\,`Foo', he said.''|
% \changes{v1.0o}{1994/05/11}{Use \cs{DeclareRobustCommand}. ASAJ.}
% \begin{macrocode}
\DeclareRobustCommand{\,}{%
\relax\ifmmode\mskip\thinmuskip\else\thinspace\fi
}
% \end{macrocode}
% \end{macro}
%
% \begin{macro}{\@}
% Placed before a '.', makes it a sentence-ending period. Does the
% right thing for other punctuation marks as well. Does this by
% setting spacefactor to 1000.
% \begin{macrocode}
\def\@{\spacefactor\@m}
% \end{macrocode}
% \end{macro}
%
%
% \begin{macro}{\hspace}
% \changes{v1.0o}{1994/05/11}{Use \cs{DeclareRobustCommand}. ASAJ.}
% \begin{macrocode}
\DeclareRobustCommand\hspace{\@ifstar\@hspacer\@hspace}
% \end{macrocode}
% \end{macro}
%
% \begin{macro}{\@hspace}
% \changes{LaTeX2e}{1993/08/05}
% {(RmS) Removed superfluous \cs{leavevmode} in \cs{@hspace} and
% \cs{@hspacer}, as suggested by CAR.}
% \begin{macrocode}
\def\@hspace#1{\hskip #1\relax}
% \end{macrocode}
% \end{macro}
%
%
% \begin{macro}{\@hspacer}
% extra |\hskip 0pt| added 1985/17/12 to guard
% against a following |\unskip|
% |\relax| added 13 Oct 88 for usual \TeX\ lossage
% replaced both changes by |\hskip\z@skip| 27 Nov 91
% \begin{macrocode}
\def\@hspacer#1{\vrule \@width\z@\nobreak
\hskip #1\hskip \z@skip}
% \end{macrocode}
% \end{macro}
%
%
%
% \begin{macro}{\fill}
% \begin{macrocode}
\newskip\fill
\fill = 0pt plus 1fill
% \end{macrocode}
% \end{macro}
%
%
%
% \begin{macro}{\stretch}
% \begin{macrocode}
\def\stretch#1{\z@ \@plus #1fill\relax}
% \end{macrocode}
% \end{macro}
%
%
%
%
% \begin{macro}{\thinspace}
% \begin{macro}{\negthinspace}
% \begin{macro}{\enspace}
% \begin{macrocode}
\def\thinspace{\kern .16667em }
\def\negthinspace{\kern-.16667em }
\def\enspace{\kern.5em }
% \end{macrocode}
% \end{macro}
% \end{macro}
% \end{macro}
%
% \begin{macro}{\enskip}
% \begin{macro}{\quad}
% \begin{macro}{\qquad}
% \begin{macrocode}
\def\enskip{\hskip.5em\relax}
\def\quad{\hskip1em\relax}
\def\qquad{\hskip2em\relax}
% \end{macrocode}
% \end{macro}
% \end{macro}
% \end{macro}
%
% \begin{macro}{\obeycr}
% \begin{macro}{\restorecr}
% The following definitions will probably get deleted or moved to
% compatibility mode soon.
%
% \changes{v1.2g}{1995/06/11}
% {(CAR) \cs{relax} added to stop silent eating of *.}
% \begin{macrocode}
{\catcode`\^^M=13 \gdef\obeycr{\catcode`\^^M13 \def^^M{\\\relax}%
\@gobblecr}%
{\catcode`\^^M=13 \gdef\@gobblecr{\@ifnextchar
\@gobble\ignorespaces}}
\gdef\restorecr{\catcode`\^^M5 }}
% \end{macrocode}
% \end{macro}
% \end{macro}
%
% \begin{macrocode}
%</2ekernel>
% \end{macrocode}
%
% \Finale
\endinput
|
