diff options
author | Karl Berry <karl@freefriends.org> | 2006-01-11 23:51:26 +0000 |
---|---|---|
committer | Karl Berry <karl@freefriends.org> | 2006-01-11 23:51:26 +0000 |
commit | 2e6fee08d3bc60b67ac039b5269615cb6760529f (patch) | |
tree | 9556b835f4a300511f771f08faf4f2f3f5a5beb0 /Master/texmf-dist/source | |
parent | 952224f90ec08c24c0bcb94779679ddc2798a450 (diff) |
trunk/Master/texmf-dist/source/latex/booktabs
git-svn-id: svn://tug.org/texlive/trunk@171 c570f23f-e606-0410-a88d-b1316a301751
Diffstat (limited to 'Master/texmf-dist/source')
-rw-r--r-- | Master/texmf-dist/source/latex/booktabs/booktabs.dtx | 1050 | ||||
-rw-r--r-- | Master/texmf-dist/source/latex/booktabs/booktabs.ins | 86 |
2 files changed, 1136 insertions, 0 deletions
diff --git a/Master/texmf-dist/source/latex/booktabs/booktabs.dtx b/Master/texmf-dist/source/latex/booktabs/booktabs.dtx new file mode 100644 index 00000000000..702c13d4067 --- /dev/null +++ b/Master/texmf-dist/source/latex/booktabs/booktabs.dtx @@ -0,0 +1,1050 @@ +% \iffalse (meta-comment) +% ----------------------------------------------------------------- +% File: booktabs.dtx +% Author: Simon Fear +% Maintainer: +% +% This file is part of the booktabs package for publication +% quality tables for LaTeX +% +% Copyright (C) 1995--2005 Simon Fear +% +% This program is free software; you can redistribute it and/or +% modify it under the terms of the GNU General Public License as +% published by the Free Software Foundation; either version 2 of +% the License, or (at your option) any later version. +% +% This program is distributed in the hope that it will be useful, +% but WITHOUT ANY WARRANTY; without even the implied warranty of +% MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the +% GNU General Public License for more details. +% +% You should have received a copy of the GNU General Public +% License along with this program; if not, write to the Free +% Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, +% MA 02111-1307 USA +% +% ----------------------------------------------------------------- +% This file contains the documentations and source code for the +% booktabs package for use with LaTeX2e. See the file 'README' +% for a list of all the files as well as directions for the +% installation of this package. +% ----------------------------------------------------------------- +% \fi +% +% \iffalse +%<*driver> +\ProvidesFile{booktabs.dtx} +%</driver> +%<package>\NeedsTeXFormat{LaTeX2e}[1999/12/01] +%<package>\ProvidesPackage{booktabs} +%<*package> + [2005/03/16 v1.6180 publication quality tables] +%</package> +% +%<*driver> +\documentclass{ltxdoc} + \usepackage{booktabs}[2005/03/16] + \EnableCrossrefs + \CodelineIndex + \RecordChanges + \setlength\hfuzz{15pt} + \hbadness=7000 + \MakeShortVerb{\=} + \MakeShortVerb{\"} +\begin{document} + \DocInput{booktabs.dtx} + \PrintChanges + \PrintIndex +\end{document} +%</driver> +% \fi +% +% \CheckSum{484} +% \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 \~} +% +% +% +% \DeleteShortVerb{\|} +% \GetFileInfo{booktabs.dtx} +% +% \title{Publication quality tables in \LaTeX% +% \thanks{This file has version number \fileversion, +% (converging to phi, the golden ratio), +% last revised \filedate.}} +% \author{Simon Fear\\300A route de Meyrin\\Meyrin\\Switzerland} +% \date{Printed \today} +% +% \maketitle +% +% +% \begin{abstract} +% This article describes some additional commands to enhance +% the quality of tables in \LaTeX. Guidelines are given as +% to what constitutes a good table in this context. The 2000 +% release (Version 1.61) of the =booktabs= package, described +% herein, adds some enhancements to the 1995 release (Version 1.00), +% most notably =longtable= compatibility. +% The latest releases (Versions v1.618 and v1.6180) are only patches\footnote{By Danie Els +% (\texttt{dnjels@sun.ac.za}) in the absence of the author.} to correct +% a problem with the vertical spacing of \cmd{\cmidrule} and the bug in the \cmd{\cmidrule} +% trim command. It add support for the =colortbl= package as well. +% \end{abstract} +% +% \section{Introduction} +% +% The routines described below are to enable the easy production +% of tables such as should appear in published scientific books and +% journals. What distinguishes these from plain +% \LaTeX{} tables is the default use of additional space above and below rules, +% and rules of varying `thickness'. What further distinguishes them from +% the tables many people {\em do} produce using \LaTeX{} is the +% absence of vertical rules and double rules. +% +% I must draw a clear +% distinction between what I call here a {\em formal table}, +% which is a set of values in labelled columns, as distinct from +% what I will call a {\em tableau}. The latter is the kind of thing +% illustrated in the \LaTeX{} manual, and increasingly +% common as the output of +% many database management systems; it will probably have +% icons in abundance, and no doubt use colour too. +% The layout of such a {\em tableau} is determined +% (hopefully) as a one-off, given a jumble of material the +% designer is trying to combine into a meaningful configuration. +% But the layout of a {\em table} has been established over +% centuries of experience and should only be altered in +% extraordinary circumstances. +% +% By way of illustration, consider this tableau from the \LaTeX{} +% manual (p. 64 old edition): +% +% \begin{center} +% \begin{tabular}{||l|lr||} \hline +% gnats & gram & \$13.65 \\ \cline{2-3} +% & each & .01 \\ \hline +% gnu & stuffed & 92.50 \\ \cline{1-1} \cline{3-3} +% emu & & 33.33 \\ \hline +% armadillo & frozen & 8.99 \\ \hline +% \end{tabular} +% \end{center} +% +% This is a hotch-potch of information that is probably +% reasonably clearly presented as is (but is the emu stuffed or +% not?). However, as a published table, this should +% much rather appear along the lines suggested +% further down the page in the manual: +% +% \begin{center} +% \begin{tabular}{@{}llr@{}} \toprule +% \multicolumn{2}{c}{Item} \\ \cmidrule(r){1-2} +% Animal & Description & Price (\$)\\ \midrule +% Gnat & per gram & 13.65 \\ +% & each & 0.01 \\ +% Gnu & stuffed & 92.50 \\ +% Emu & stuffed & 33.33 \\ +% Armadillo & frozen & 8.99 \\ \bottomrule +% \end{tabular} +% \end{center} +% +% It takes much less work to lay +% this out, as a formal table; we don't have to work out a new +% layout for everything we do. Moreover, we can be almost certain +% that the data cannot be misread, because the reader does not +% have to learn how to read some novel presentation. +% +% The above table cannot be produced in pure \LaTeX, +% unfortunately. It can be laid out as it should be, but despite +% your best efforts, using plain =\hline= commands produces +% +% \begin{center} +% \begin{tabular}{@{}llr@{}} \hline +% \multicolumn{2}{c}{Item} \\ \cline{1-2} +% Animal & Description & Price (\$)\\ \hline +% Gnat & per gram & 13.65 \\ +% & each & 0.01 \\ +% Gnu & stuffed & 92.50 \\ +% Emu & stuffed & 33.33 \\ +% Armadillo & frozen & 8.99 \\ \hline +% \end{tabular} +% \end{center} +% +% Note (if it is not already obvious) that there is not enough +% space between the top line and the capital I of `Item', and so +% on for all the lines: contrast with the previous version. Also, in +% the first version +% the top and bottom rules (ie lines) are +% heavier than the middle rule, which is turn heavier than the +% subrule underneath `Item'. Of course you \emph{could} redefine +% =\doublerulesep= and then use =\hline\hline= to get something +% near the same effect, and you could use struts to improve +% the spacing. But you should not have to +% think of such things. The =booktabs= style defines its commands so +% that such things are taken care of automatically. +% +% In general, I would say that +% this package is of no interest to those looking for +% an alternative to Pic\TeX{} to conjure up fancy tableaux. +% Rather, it is a style guide for authors of scientific papers +% and books as regards table layout. It is not going too far to +% say that if you cannot +% create a table using the commands in this package, you should +% redesign it. +% +% \subsection{A note on terminology\label{notation}} +% +% In British typesetting, a `line' is always called a `rule'. Perhaps +% confusingly (for historic reasons in fact), the `thickness' of +% rule is often referred to as is its `width' (whereas just about +% everyone else would call this `depth' or `height', if they were thinking +% of a horizontal rule). A `thick black line' is called a `heavy +% rule'. I have used this terminology in most of the new commands +% below. If nothing else it avoids confusion with =\hline=. +% +% \section{The layout of formal tables} +% +% You will not go far wrong if you remember two simple +% guidelines at all times: +% \begin{enumerate} +% \item Never, ever use vertical rules. +% \item Never use double rules. +% \end{enumerate} +% These guidelines may seem extreme but I have +% never found a good argument in favour of breaking them. For +% example, if you feel that the information in the left half of +% a table is so different from that on the right that it needs +% to be separated by a vertical line, then you should use two +% tables instead. Not everyone follows the second guideline: +% I have worked for a publisher who insisted on a +% double light rule above a row of totals. But this would not +% have been my choice. +% +% There are three further guidelines worth mentioning here as they +% are generally not known outside the circle of professional +% typesetters and subeditors: +% \begin{enumerate}\setcounter{enumi}{2} +% \item Put the units in the column heading (not in the body of +% the table). +% \item Always precede a decimal point by a digit; thus 0.1 +% {\em not} just .1. +% \item Do not use `ditto' signs or any other such convention to +% repeat a previous value. In many circumstances a blank +% will serve just as well. If it won't, then repeat the value. +% \end{enumerate} +% +% Whether or not you wish to follow the minor niceties, +% if you use only the following commands in your formal tables +% your reader will be grateful. I stress that +% the guidelines are not just to +% keep the pedantic happy. The principal is that enforced structure of +% presentation +% enforces structured thought in the first instance. +% +% \section{Use of the new commands} +% +% \DescribeMacro\toprule +% \DescribeMacro\midrule +% \DescribeMacro\bottomrule +% In the simplest of cases a table begins with a =\toprule=, has +% a single row of column headings, then a dividing rule called +% here a =\midrule=; after the columns of data we finish off with +% a =\bottomrule=. Most book publishers set the =\toprule= and +% =\bottomrule= heavier (ie thicker, or darker; see section \ref{notation}) +% than the intermediate =\midrule=. However, when tables appear +% in very small typesizes it is sometimes impossible to make this +% distinction, and moreover quite a few journals routinely use all rules of +% the same heaviness. +% +% The rule commands here all take a default +% which may be reset within the document (preferably, but not +% necessarily, in the preamble). For the top and bottom rules this +% default +% is =\heavyrulewidth= and for midrules it is =\lightrulewidth= +% (fully described below). In very rare cases where you need to do +% something special, you may use the optional arguments to the rule +% commands which have formal syntax as follows: +% \begin{quote} +% "\toprule["\meta{wd}"]" \\ +% "\midrule["\meta{wd}"]" \\ +% "\bottomrule["\meta{wd}"]" +% \end{quote} +% where \meta{wd} is a \TeX dimension (for example 1pt, .5em, +% etc.). +% +% All the rule commands described here go after the closing +% =\\= of the preceding row (except =\toprule=, which +% comes right after the =\tabular{}= command); in other words, +% exactly where plain \LaTeX{} allows =\hline= or =\cline=. +% +% \DescribeMacro\cmidrule +% Frequently we need a sub-rule to extend over only some of the columns, +% for which we need a =\cmidrule= (the analogue of \LaTeX's =\cline= +% command). Generally, this rule should not come to the full width of +% the columns, and this is especially the case when we need to begin +% a =\cmidrule= straight after the end of another one (\LaTeX's +% =\cline=s crash into each other here if you are not extra careful with +% =\extracolsep=). Thus, you will generally want to use the optional +% `trimming' commands. +% +% The trimming commands, if used at all, go in +% parentheses (like this), with no spaces separating them. The possible +% specifications are =r=, =r{=\meta{wd}=}=, =l= and +% =l{=\meta{wd}=}=, or any combination of these, where \meta{wd} is a +% dimension, and =r= and =l= indicate whether the right and/or left ends +% of the rule should be trimmed. The form without explicit argument is +% equivalent to =r{\cmidrulekern}=, where =\cmidrulekern= defaults to +% $0.5$ em, but can be set by the user in the preamble.\footnote{User +% feedback suggested the Version 1.00 default, 0.25 em, was too +% small. Sorry for any loss of backward compatibility. Remember that +% you +% can easily set \texttt{\textbackslash cmidrulekern} in the preamble, or just use +% \texttt{(r\{.25em\})} to recover the original behaviour.} +% +% Here's an illustrative example: =(lr{.75em})= gives you a default +% left trim and exactly $0.75$ em right trim. Equally valid here is +% =(r{.75em}l)=.\footnote{As a matter of fact, \texttt{(lrrlr\{.75em\})} +% does the same thing: only the +% last encountered left and the last encountered right specification +% are applied.} +% +% The full syntax of the command is +% \begin{quote} +% "\cmidrule["\meta{wd}"]("\meta{trim}"){"$a$--$b$"}" +% \end{quote} +% where \meta{wd} is an optional rule width command, in square +% brackets [like this] (the +% default here is =\cmidrulewidth=), and the last argument, \emph{which +% is not optional}, gives the column numbers to be spanned. +% +% An example of the commands in use is given by the code used to +% produce the example table above: +% \begin{verbatim} +% \begin{tabular}{@{}llr@{}} \toprule +% \multicolumn{2}{c}{Item} \\ \cmidrule(r){1-2} +% Animal & Description & Price (\$)\\ \midrule +% Gnat & per gram & 13.65 \\ +% & each & 0.01 \\ +% Gnu & stuffed & 92.50 \\ +% Emu & stuffed & 33.33 \\ +% Armadillo & frozen & 8.99 \\ \bottomrule +% \end{tabular} +% \end{verbatim} +% +% \DescribeMacro\addlinespace +% Occasionally we want to put an extra space between certain rows +% of a table; for example, before the last row, if this is a +% total. This is simply a matter of inserting +% \begin{quote} +% "\addlinespace["\meta{wd}"]" +% \end{quote} +% after the =\\= alignment marker. Between ordinary rows of text, the +% effect is identical to the ordinary \LaTeX\ usage +% =\\[\defaultaddspace]=, which I find rather clumsy, and it is better +% than =\\ \\=, which inserts too much space. Also, +% =\addlinespace= can be used before, after, or between rules if you +% want to control the exact amount of space to be inserted. The default +% space before or after an adjacent rule is replaced by exactly +% =\defaultaddspace= or the amount of +% space specified in the optional argument.\footnote{This is a change +% from version 1.00, where the space was sometimes \emph{in addition +% to} default rule space.} +% +% \section{Abuse of the new commands} +% +% Let's face it, nobody can leave well alone, so here are some +% guidelines and extra commands. +% +% The new rule commands are not guaranteed to work with =\hline= +% or =\cline=, although these remain available and unchanged. I +% cannot foresee any reason to want to mix them. +% +% More importantly the rules generated by the new commands are in +% no way guaranteed to connect with verticals generated by ={|}= +% characters in the preamble. This is a feature (see above). You +% should not use vertical rules in tables, end of story. +% +% \DescribeMacro\morecmidrules +% If you just cannot stop yourself from using a double rule, +% even a construction as bizarre as =\toprule\bottomrule\midrule= +% will work without generating an error message (just as you can +% double =\hline=). These rules will be separated by the ordinary +% \LaTeX\ separator =\doublerulesep=. However if your perversion +% is to want double =\cmidrule=s you will need the extra command +% =\morecmidrules= to do so properly, because normally two +% =\cmidrule=s in a row is a sane construction +% calling for two rules on the same `rule row'. Thus in +% \begin{quote} +% "\cmidrule{1-2}\cmidrule{1-2}" +% \end{quote} +% the second command writes a rule that just overwrites the first +% one; I suppose you wanted +% \begin{quote} +% "\cmidrule{1-2}\morecmidrules\cmidrule{1-2}" +% \end{quote} +% which gives you a double rule between columns one and two, +% separated by =\cmidrulesep= (note: since a =\cmidrule= is +% generally very light, the ordinary =\doublerulesep= is probably +% too much space). Finish off a whole row of rules before giving the +% =\morecmidrules= command. Note that =\morecmidrules= has no +% effect whatsoever if it does not immediately follow a +% =\cmidrule= (ie it is not a general space-generating command). +% +% \DescribeMacro\specialrule +% If you find some extraordinary need to specify exactly $0.5$ em, say, between two +% rules, you could use a construction such as =\midrule = +% =\addlinespace[.5em] \midrule=. In a rare fit of tolerance, though, +% I have also +% provided the command +% \begin{quote} +% "\specialrule{"\meta{wd}"}{"\meta{abovespace}"}{"\meta{belowspace}"}" +% \end{quote} +% where all three arguments are mandatory (I couldn't be bothered to +% program in defaults). If you use this frequently, you have +% misunderstood the purpose and content of the guidelines given above. A +% preceeding rule does not add its default space below, and a following +% rule adds no space above itself, so you get \emph{exactly} the space +% specified in the arguments.\footnote{This is a change from Version 1.00, +% which rather liked to add an extra \texttt{\textbackslash doublerulesep} space whenever it +% could.} +% +% +% \section{Booktabs and longtables} +% +% If you have both =booktabs= and =longtable= packages loaded, the +% =booktabs= rule commands can now all be used exactly as described +% above, within a longtable. +% +% There is an addition worth noting: within a =longtable=, you can use +% the optional left and right trimming commands, which normally only +% work for =\cmidrule=s, with =\toprule=, =\midrule= and =\bottomrule= +% (and if you must, also with =\specialrule=). Users who hacked the +% previous release for longtable compatability\footnote{Jim Service was +% the first} seemed to like all the rules to be right trimmed 0.5 em. I +% think you can do the same by making =@{}= be the last column +% specifier. Still, after working out the rest of the code, it was easy +% to add parsing for the optional arguments, so I did. (I didn't go the +% whole way and allow the optional trimming \emph{outside} a +% =longtable=; this would be a huge amount of work. If you must have +% trimmed rules, make all your tables be =longtable=s!) +% +% A somewhat technical note: within a =longtable=, =\hline= and +% =\hline\hline= both produce a \emph{double} rule (to allow for page +% breaks occurring at that point). But the =booktabs= rules do +% \emph{not}. Longtable's automatic doubling of =\hline= is +% questionable, even according to the documentation within that +% package. But doubled =booktabs= rules make almost no sense at all. In +% the unfortunate event that a =booktabs= rule should occur at a page +% break, then you will have to make the necessary adjustments by +% hand. (In general, this will mean deleting the offending rule.) +% +% +% \section{Booktabs and and the colortbl package} +% +% =Booktabs= is now compatible with the =colortbl= package.\footnote{Since v1.6180} +% The \cmd{\arrayrulecolor} command will result in coloured +% rules if the =colortbl= package is loaded. +% +% +% \section{Technical summary of commands} +% +% The new rule commands are valid inside the standard =tabular= (and +% =array=) environment, in the modified =tabular= and =array= of +% =\usepackage{array}=, and within both standard tables and longtables +% after =\usepackage{longtable}=. +% +% The commands follow the standard placement syntax of =\hline=. +% There can be space (including carriage-return, +% but not two carriage-returns) between successive +% rule commands.\footnote{A welcome change from Version 1.00, where space between +% rule commands generated a very baffling error message.} +% +% In what amounts to quite a big change from former releases, within +% the macro code I now define three classes of rules. (But we don't need these +% definitions within ordinary use, so I haven't even mentioned them above.) +% A class 1 rule (otherwise called a `normal' rule) is any of +% =\toprule=, =\midrule=, =\bottomrule=, or =\cmidrule=. The class 2 +% rules are =\specialrule= and =\addlinespace=. Finally, a class 0 rule +% is none of the preceeding --- or in other words, not a rule at +% all.\footnote{Except that \texttt{\textbackslash hline} and +% \texttt{\textbackslash cline} are class 0. Still, there +% is no reason to lose sleep over this, since one would not want to mix +% the two rule-drawing systems.} Note that =\addlinespace= counts as a +% class 2 rule, not as class 0 text. +% +% In the following, we first describe each command in `normal use', meaning +% that the rule is being used between two lines of text (or more +% technically, is preceded and followed by a class 0 rule). After that, +% we will look at the exceptions. +% \begin{quote} +% "\toprule["\meta{wd}"]" +% \end{quote} +% A rule of width \meta{wd} (default =\heavyrulewidth=) with +% =\abovetopsep= space above and +% =\belowrulesep= extra vertical space inserted below it. By default, +% =\abovetopsep= is zero, which seems sensible for a rule designed to +% go at the top. However, if your +% tables have captions, it can make sense to use =\abovetopsep= to +% insert a reasonable amount of space between caption and table, +% rather than remember to use a =\vspace{}= command in the float. +% \begin{quote} +% "\midrule["\meta{wd}"]" +% \end{quote} +% A \meta{wd} (default =\lightrulewidth=) rule with +% =\aboverulesep= space above it and +% with =\belowrulesep= space below it. +% \begin{quote} +% "\bottomrule["\meta{wd}"]" +% \end{quote} +% A \meta{wd} (default =\heavyrulewidth=) rule with =\aboverulesep= +% space above it and with =\belowbottomsep= space below it. By default +% =\belowbottomsep= is zero\footnote{This is a change from Version 1.00, +% where there was always a \texttt{\textbackslash belowrulesep}}. +% There is a frequent and legitimate reason +% you might want space below a bottom rule: namely, when there's a table +% footnote.\footnote{But don't use footnotes, Donald.} +% If you don't override the default you could use +% =\bottomrule \addlinespace[\belowrulesep]= +% or you could put a suitably sized strut +% into the footnote text.\footnote{I don't like either of these. Sort it out in +% Version 1.618?} But the default has to be zero, so that it behaves +% sensibly in a =longtable= footer. +% \begin{quote} +% "\cmidrule["\meta{wd}"]("\meta{trim}"){"$a$--$b$"}" +% \end{quote} +% A \meta{wd} (default =\cmidrulewidth=) rule with +% =\aboverulesep= space above it (unless following another +% =\cmidrule=, in which case it is on the same vertical alignment; or +% if following =\morecmidrules=, separated from a previous =\cmidrule= +% by =\cmidrulesep=). A =\cmidrule= has +% =\belowrulesep= below it (unless followed by another +% =\cmidrule=, in which case the following rule is on the same +% vertical alignment; or if followed by =\morecmdirules=, when there +% will be +% =\cmidrulesep= below it). +% +% The =\cmidrule= spans columns $a$ to $b$ as specified in the mandatory +% argument. The optional argument \meta{trim}, which goes in parentheses +% if at all, can contain any sequence of the tokens =r=, =l= and +% ={=\meta{wd}=}=, with the latter setting the kerning to be applied to +% right or left sides as specified by the immediately preceding token. +% (There's currently no error checking done here, so be careful to get +% the syntax right.) +% \begin{quote} +% "\morecmidrules" +% \end{quote} +% Instructs \LaTeX\ to begin a new row of =\cmidrules=, +% separated from the last by =\cmidrulesep=. Has no meaning +% in any other context. +% \begin{quote} +% "\specialrule{"\meta{wd}"}{"\meta{abovespace}"}{"\meta{belowspace}"}" +% \end{quote} +% A \meta{wd} rule (note: here this is a mandatory argument) with +% \meta{abovespace} above it and \meta{belowspace} below it. +% \begin{quote} +% "\addlinespace["\meta{wd}"]" +% \end{quote} +% Technically this has the same effect as +% =\specialrule{0pt}{0pt}{=\meta{wd}=}=, i.e. a zero-width rule with no +% space above and with \meta{wd} (default =\defaultaddspace=) space +% below. This command was primarily designed to add space between rows +% in the body of the table, but it may also be used to specify an exact +% amount of space above or below a class 1 rule. +% +% \vspace{1em} +% Now we come to the exceptions to the above. We have already seen in +% the definitions that the type 2 rules are preceded and followed by +% exactly the amount of space specified by the arguments. That is, a +% type 2 rule suppresses the space that would normally be generated by a +% previous type 1 rule (e.g. =\belowrulesep= after a =\toprule=) and +% replaces it by the argument of the type 2 rule. Similarly, in the +% combination \{type 2 rule\}\{type 1 rule\}, the ordinary space above +% the type 1 rule (e.g. =\aboverulesep=) is suppressed. But in the +% combination \{type 2 rule\}\{type 2 rule\}, no space is suppressed: +% the rules will be separated by both the first rule's +% ={=\meta{belowspace}=}= and the second rule's ={=\meta{abovespace}=}= +% arguments. Last but not least, the combination \{type 1 rule\}\{type 1 +% rule\} will always give rules separated by =\doublerulesep=, +% suppressing all normal space generated between the rules (but +% retaining normal space above the first and below the second). +% +% As an exception to this last exception, `type 1 rule' excludes +% =\cmidrule=. Such rules combine with other =\cmidrule=s and +% =\morecmidrules= in normal use as described above. I don't know and +% I don't care +% care what the combination =\toprule\cmidrule{1-2}\midrule= would +% produce. I can see no excuse for such usage. +% +%\vspace{1em} +% The default dimensions are defined at the beginning of the macro +% description section (Section \ref{code}). The user can change +% these defaults in the preamble, or outside a +% tabular environment, by simply +% inserting a command in exactly the same format as in Section \ref{code}; the +% redefinition will stay in effect for the rest of the document +% or until redefined again. \emph{Inside a table} you would have to make the +% assignment globally in a =noalign= group: +% e.g. \verb|\noalign\{\global\abovetopsep=1em\toprule}|. I hope +% you never have to do that. +% +% \section{Acknowledgments} +% Hugely indebted of course to DEK and Lamport; the optional +% argument and =\cmidrule= +% stuff especially was stolen from =latex.sty=. +% The documentation driver stuff is stolen from the +% tools package description =dcolumn.dtx= by David Carlisle. +% +% For beta testing and encouragement ... +% +% \StopEventually +% +% \section{The code\label{code}} +% +% The current version is defined at the top of the file looking +% something like this +% \begin{macrocode} +%<*package> +%\NeedsTeXFormat{LaTeX2e} +%\ProvidesPackage{booktabs} +% [\filedate\space version\fileversion] +% \end{macrocode} +% +% First we set up the new dimensions described above: +% \begin{macrocode} +\newdimen\heavyrulewidth +\newdimen\lightrulewidth +\newdimen\cmidrulewidth +\newdimen\belowrulesep +\newdimen\belowbottomsep +\newdimen\aboverulesep +\newdimen\abovetopsep +\newdimen\cmidrulesep +\newdimen\cmidrulekern +\newdimen\defaultaddspace +\heavyrulewidth=.08em +\lightrulewidth=.05em +\cmidrulewidth=.03em +\belowrulesep=.65ex +\belowbottomsep=0pt +\aboverulesep=.4ex +\abovetopsep=0pt +\cmidrulesep=\doublerulesep +\cmidrulekern=.5em +\defaultaddspace=.5em +% \end{macrocode} +% And some internal counters of no interest to the end user: +% \changes{v1.618}{2003/03/28}{Remove \cs{@cmidrulewidth} } +% \begin{macrocode} +\newcount\@cmidla +\newcount\@cmidlb +\newdimen\@aboverulesep +\newdimen\@belowrulesep +\newcount\@thisruleclass +\newcount\@lastruleclass +\@lastruleclass=0 +\newdimen\@thisrulewidth +% \end{macrocode} +% which will be described as needed below. +% +% \begin{macro}{\futurenonspacelet} +% Next we define a very useful macro (more-or-less straight from the +% \TeX book's Dirty Tricks chapter; documented there). Use +% =\futurenonspacelet= instead of =\futurelet= when looking for the next +% (non-space) token after a macro that has an argument. (After a macro +% without an argument, space is ignored anyway, so =\futurenonspacelet= +% wouldn't be needed.) This hack allows users to type white +% space between successive rule commands (which did not work in Version +% 1.00). +% \begin{macrocode} +\def\futurenonspacelet#1{\def\@BTcs{#1}% + \afterassignment\@BTfnslone\let\nexttoken= } +\def\@BTfnslone{\expandafter\futurelet\@BTcs\@BTfnsltwo} +\def\@BTfnsltwo{\expandafter\ifx\@BTcs\@sptoken\let\next=\@BTfnslthree + \else\let\next=\nexttoken\fi \next} +\def\@BTfnslthree{\afterassignment\@BTfnslone\let\next= } +% \end{macrocode} +% \end{macro} +% +% \subsection{Full width rules} +% +% When we are not in a =longtable= environment, we can simply +% implement the full width rules as a =\hrule= in a =\noalign{}= +% group. But within a =longtable=, the rule has to be drawn like a +% =\cmidrule{1-\LT@cols}= (the rationale for this is explained in the +% =longtable= documentation). +% +% In order to allow for both, all the rule macros have to +% open a =\noalign= group immediately, while they work out whether +% they have been called within a =longtable=; if you don't do this, +% \TeX's underlying =\halign= process gets hiccups. +% I use \LaTeX's +% dirty trick (\verb?\ifnum=0`}?) to fool the parser that the bracket +% count is OK. The bracket really gets closed after all the skipping +% at the end of the =\@BTendrule= macro. +% +% The class 1 rules, and =\specialrule=, really only differ in the +% defaults for space above and +% below, and the width, passed to a common routine, =\@BTrule=, +% described below. The spaces, =\@aboverulesep= and =\@belowrulesep=, +% are set within the =\noalign= group, so are inherited by +% =\@BTrule=. Similarly, =\@BTrule= knows as much as it needs to +% about the routine that called it by examining the inherited +% =\@thisruleclass=. The optional width argument is parsed by +% =\@BTrule= after being set to default if absent. +% +% +% \begin{macro}{\toprule} +% \begin{macro}{\midrule} +% \begin{macro}{\bottomrule} +% \begin{macro}{\specialrule} +% \begin{macrocode} +\def\toprule{\noalign{\ifnum0=`}\fi + \@aboverulesep=\abovetopsep + \global\@belowrulesep=\aboverulesep %global cos for use in the next noalign + \global\@thisruleclass=\@ne + \@ifnextchar[{\@BTrule}{\@BTrule[\heavyrulewidth]}} +\def\midrule{\noalign{\ifnum0=`}\fi + \@aboverulesep=\aboverulesep + \global\@belowrulesep=\belowrulesep + \global\@thisruleclass=\@ne + \@ifnextchar[{\@BTrule}{\@BTrule[\lightrulewidth]}} +\def\bottomrule{\noalign{\ifnum0=`}\fi + \@aboverulesep=\aboverulesep + \global\@belowrulesep=\belowbottomsep + \global\@thisruleclass=\@ne + \@ifnextchar[{\@BTrule}{\@BTrule[\heavyrulewidth]}} +\def\specialrule#1#2#3{\noalign{\ifnum0=`}\fi + \@aboverulesep=#2\global\@belowrulesep=#3\global\@thisruleclass=\tw@ + \@BTrule[#1]} +% \end{macrocode} +% \end{macro} +% \end{macro} +% \end{macro} +% \end{macro} +% \begin{macro}{\addlinespace} +% An =\addlinespace= is essentially a zero-width rule with +% zero space above and argument (or default) space below. But because +% the rule is not actually drawn, but is just a =\vskip=, there is no +% need to check if we're in a longtable, so we don't need to call +% =\@BTrule= as for `real' rules. But we do share the =\@BTendrule= +% lookahead and flagsetting code (described below), and the =\vskip= is done there. +% \begin{macrocode} +\def\addlinespace{\noalign{\ifnum0=`}\fi + \@ifnextchar[{\@addspace}{\@addspace[\defaultaddspace]}} +\def\@addspace[#1]{\global\@belowrulesep=#1\global\@thisruleclass=\tw@ + \futurelet\@tempa\@BTendrule} +% \end{macrocode} +% \end{macro} +% +% +% \begin{macro}{\@BTrule} +% All the rules (except =\addlinespace=) share this code. +% \begin{macrocode} +\def\@BTrule[#1]{% + \global\@thisrulewidth=#1\relax +% \end{macrocode} +% Save the width argument (if the user didn't give one, then the +% calling routine will have called =\@BTrule= with the default) in a global +% variable for later use when drawing the rule. +% \begin{macrocode} + \ifnum\@thisruleclass=\tw@\vskip\@aboverulesep\else +% \end{macrocode} +% Specialrules always insert specified space above. (Note: addlinespaces +% don't come here). +% \begin{macrocode} + \ifnum\@lastruleclass=\z@\vskip\@aboverulesep\else + \ifnum\@lastruleclass=\@ne\vskip\doublerulesep\fi\fi\fi +% \end{macrocode} +% After text (last rule class 0), precede the rule by =\aboverulesep=; +% but if immediately after a +% previous rule, insert a =\doublerulesep=. +% +% Now we work out, by a very nasty hack, if we're within a +% =longtable=. It's easy if =\longtable= isn't even defined: then we +% can't be. But it is not enough just to check if =longtable= is +% loaded --- we might be within an ordinary table rather than a +% longtable. So we look to see if =\hline= has been re-defined +% from its \LaTeX\ definition to be the same as +% =\LT@hline=. (Longtable currently does this redefinition when it opens a +% =longtable= environment, but not globally, so it is +% cleared it when the environment closes.) Another +% package could potentially do this! And =longtable= might change the +% way it implements this! So, it is not entirely safe, but I have +% found no better way so far. +% +% We set up =\@BTswitch= to call =\@BTnormal= or =\@BLTrule=, as +% appropriate, then call it. +% \begin{macrocode} + \ifx\longtable\undefined + \let\@BTswitch\@BTnormal + \else\ifx\hline\LT@hline + \let\@BTswitch\@BLTrule + \else + \let\@BTswitch\@BTnormal + \fi\fi + \@BTswitch} +% \end{macrocode} +% \end{macro} +% +% +% \begin{macro}{\CT@arc@} +% \changes{v1.6180}{2005/03/16}{add colortbl command for color support} +% This is support for the colortbl package for colored +% rules. \cmd{\CT@arc@} hold the \cmd{\arrayrulecolor} setting. +% \begin{macrocode} +\AtBeginDocument{% + \providecommand*\CT@arc@{}}%% colortbl support +% \end{macrocode} +% \end{macro} +% +% +% \begin{macro}{\@BTnormal} +% \changes{v1.6180}{2005/03/16}{add colortbl \cmd{\CT@arc@} command for color support} +% This is when we're \emph{not} within a longtable. We are already in a +% =\noalign= group, all we need do is draw an =\hrule= and gobble any +% trailing spaces, then call the closing routine with =\@tempa= set +% equal to the next token in the document. +% +% \begin{macrocode} +\def\@BTnormal{% + {\CT@arc@\hrule\@height\@thisrulewidth}% + \futurenonspacelet\@tempa\@BTendrule} +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\@BLTrule} +% This is for full width rule within a =longtable=. First we check if +% a kerning argument has been used; if so let =\@@BLTrule= read it, +% else call =\@@BLTrule= with an empty string: +% \begin{macrocode} +\def\@BLTrule{\@ifnextchar({\@@BLTrule}{\@@BLTrule()}} +% \end{macrocode} +% \end{macro} +% \begin{macro}{\@@BLTrule} +% \begin{macrocode} +\def\@@BLTrule(#1){\@setrulekerning{#1}% +\global\@cmidlb\LT@cols +% \end{macrocode} +% The =\@setrulekerning= routine parses the kerning argument tokens +% and sets global kerning widths accordingly (or to defaults, if user +% hasn't set them explicitly). The global assignment to =\@cmidlb= +% sets up the column count for the =\@cmidruleb= macro, which is +% shared with =cmidrules=. +% \begin{macrocode} +\ifnum0=`{\fi}% +% \end{macrocode} +% Close the currently open =\noalign= group. Within a =longtable=, +% rules are all to be drawn as leaders within a text box that is +% =\LT@cols= columns wide. +% \begin{macrocode} +\@cmidruleb +% \end{macrocode} +% Draw the rule. We share the =\@cmidruleb= code with ordinary =\cmidrule=s. +% \begin{macrocode} +\noalign{\ifnum0=`}\fi +% \end{macrocode} +% We have to open a new noalign immediately else \TeX will start a new +% text box where we don't want one. Then, after gobbling any +% unwanted white space, we call the closing routine. +% \begin{macrocode} +\futurenonspacelet\@tempa\@BTendrule} +% \end{macrocode} +% \end{macro} +% \begin{macro}{\@BTendrule} +% We look one step ahead (token is in =\@tempa=) to see if another +% rule follows (shame on user!). If so, we set =\@lastruleclass= equal +% to =\@thisruleclass= (thus setting it up for the following rule). If +% there isn't a following rule, we clear =\@lastruleclass= (ie set it +% to zero), which isn't technically true since we have just drawn a +% rule, but sets it up correctly for the next rule encountered, which +% must be following some intervening text. +% \begin{macrocode} +\def\@BTendrule{\ifx\@tempa\toprule\global\@lastruleclass=\@thisruleclass + \else\ifx\@tempa\midrule\global\@lastruleclass=\@thisruleclass + \else\ifx\@tempa\bottomrule\global\@lastruleclass=\@thisruleclass + \else\ifx\@tempa\cmidrule\global\@lastruleclass=\@thisruleclass + \else\ifx\@tempa\specialrule\global\@lastruleclass=\@thisruleclass + \else\ifx\@tempa\addlinespace\global\@lastruleclass=\@thisruleclass + \else\global\@lastruleclass=\z@\fi\fi\fi\fi\fi\fi + \ifnum\@lastruleclass=\@ne\relax\else\vskip\@belowrulesep\fi + \ifnum0=`{\fi}} +% \end{macrocode} +% \end{macro} +% +% +% \subsection{Special subrules} +% +% \begin{macro}{\@setrulekerning} +% The following code parses the trimming arguments (if there are any) +% for =\cmidrule= or a =\BLTrule=. The rule +% will be trimmed left and right by =\cmrkern@l= and =\cmrkern@l=, +% which are zero by default, set to =\cmidrulekern= by the plain +% =(lr)= arguments, or user set as in =(r{.5em})=. We parse token by +% token through the arguments. The +% tokens =r= and =l= cause =\cmrkern@r= or =\cmrkern@l= to be set to +% =\cmidrulekern=. There is no lookahead to see if a width is the +% next token; this strategy is efficient for the plain commands, +% while inefficient for the qualified commands, but more importantly +% it is much easier to program. Tokens =r= and =l= also set +% =\cmrswitch= so that if the next token turns out to be +% ={=\meta{wd}=}= then the kerning will be done on the side +% currently specified. I have been too lazy to program an error +% message should one encounter tokens other than =r=, =l= or +% ={=\meta{wd}=}=. +% \changes{v1.6180}{2005/03/16}{Refine option testing in \cmd{\@setrulekerning}} +% \begin{macrocode} +\def\@setrulekerning#1{% + \global\let\cmrkern@l\z@ + \global\let\cmrkern@r\z@ + \@tfor\@tempa :=#1\do + {\def\@tempb{r}% + \ifx\@tempa\@tempb + \global\let\cmrkern@r\cmidrulekern + \def\cmrsideswitch{\cmrkern@r}% + \else + \def\@tempb{l}% + \ifx\@tempa\@tempb + \global\let\cmrkern@l\cmidrulekern + \def\cmrsideswitch{\cmrkern@l}% + \else + \global\expandafter\let\cmrsideswitch\@tempa + \fi + \fi}} +% \end{macrocode} +% \end{macro} +% \begin{macro}{\cmidrule} +% \begin{macro}{\@cmidrule} +% \begin{macro}{\@@cmidrule} +% \begin{macro}{\@@@cmidrule} +% The =\cmidrule= re-uses =\@lastruleclass= in an entirely different +% way from the full width rules. (Maybe I should have used a +% different flag; it seemed efficient at the time \ldots). This +% is (left) set to one if you are in the middle of a row of +% =\cmidrules=, or starting a new one (with =\morecmidrules=). +% Otherwise, when =\@lastruleclass= is zero, we precede the +% rule with =\aboverulesep=. +% \begin{macrocode} +\def\cmidrule{\noalign{\ifnum0=`}\fi + \@ifnextchar[{\@cmidrule}{\@cmidrule[\cmidrulewidth]}} +\def\@cmidrule[#1]{\@ifnextchar({\@@cmidrule[#1]}{\@@cmidrule[#1]()}} +\def\@@cmidrule[#1](#2)#3{\@@@cmidrule[#3]{#1}{#2}} +% \end{macrocode} +% The above is fiddling around to set defaults for missing optional +% arguments. We also pass to =\@@@cmidrule= in a different order, +% namely =[a-b]=={width required}= ={kerning commands}= (this being +% the order in which the arguments are actually processed): +% \begin{macrocode} +\def\@@@cmidrule[#1-#2]#3#4{\global\@cmidla#1\relax + \global\advance\@cmidla\m@ne + \ifnum\@cmidla>0\global\let\@gtempa\@cmidrulea\else + \global\let\@gtempa\@cmidruleb\fi + \global\@cmidlb#2\relax + \global\advance\@cmidlb-\@cmidla +% \end{macrocode} +% This has set up a switch (=\@gtempa=) to call the relevant routine, +% =\@cmidrulea= or =\@cmidruleb=, depending on whether we start +% from column one or not. +% \begin{macrocode} + \global\@thisrulewidth=#3 +% \end{macrocode} +% That is, set per default or given argument. Then parse any trimming +% arguments to set, globally, =\cmrkern@r= and =\cmrkern@l= accordingly: +% \begin{macrocode} + \@setrulekerning{#4} +% \end{macrocode} +% Now insert space above if needed, close the =\noalign=, +% then switch to appropriate rule drawing routine as determined +% above (=\let= to =\@gtempa=): +% \begin{macrocode} + \ifnum\@lastruleclass=\z@\vskip \aboverulesep\fi + \ifnum0=`{\fi}\@gtempa +% \end{macrocode} +% Having now drawn the rule, +% open another =\noalign=, and call the closing routine: +% \begin{macrocode} + \noalign{\ifnum0=`}\fi\futurenonspacelet\@tempa\@xcmidrule} +% \end{macrocode} +% \end{macro} +% \end{macro} +% \end{macro} +% \end{macro} +% +% \begin{macro}{\@xcmidrule} +% In this closing routine, see if another =\cmidrule= follows; +% if so, backspace +% vertical so it will line up with the one you just drew, +% and setting =\@lastruleclass= to 1 will suppress +% adding space above the next. If a =\morecimdrules= follows, +% we add (positive) =\cmidrulesep= (and again set =\@lastruleclass= +% to one). Otherwise this is the last rule of the current group +% and we can just add =\belowrulesep=. Finally, we close the =\noalign=. +% \changes{v1.618}{2003/03/28}{Change to \cmd{\@xcmidrule}: +% replace \cmd{\@cmidrulewidth} with \cmd{\@thisrulewidth }} +% \begin{macrocode} +\def\@xcmidrule{% + \ifx\@tempa\cmidrule + \vskip-\@thisrulewidth + \global\@lastruleclass=\@ne + \else \ifx\@tempa\morecmidrules + \vskip \cmidrulesep + \global\@lastruleclass=\@ne\else + \vskip \belowrulesep + \global\@lastruleclass=\z@ + \fi\fi + \ifnum0=`{\fi}} +% \end{macrocode} +% \end{macro} +% \begin{macro}{\@cmidrulea} +% This code (called below) actually draws the rules. They are drawn +% as boxes in text, rather than in a =\noalign= group, which permits +% the left and right kerning. +% \changes{v1.6180}{2005/03/16}{add colortbl \cmd{\CT@arc@} command for color support} +% \begin{macrocode} +\def\@cmidrulea{% + \multispan\@cmidla&\multispan\@cmidlb + \unskip\hskip\cmrkern@l% + {\CT@arc@\leaders\hrule \@height\@thisrulewidth\hfill}% + \hskip\cmrkern@r\cr}% +% \end{macrocode} +% \end{macro} +% \begin{macro}{\@cmidruleb} +% \changes{v1.6180}{2005/03/16}{add colortbl \cmd{\CT@arc@} command for color support} +% \begin{macrocode} +\def\@cmidruleb{% + \multispan\@cmidlb + \unskip\hskip \cmrkern@l% + {\CT@arc@\leaders\hrule \@height\@thisrulewidth\hfill}% + \hskip\cmrkern@r\cr}% +% \end{macrocode} +% \end{macro} +% +% \begin{macro}{\morecmidrules} +% This is really a dummy command; all the work is done above +% within the =\cmidrule= routine. We look one step ahead there +% to see if a =\morecmidrules= follows the current =\cmidrule=, +% and if so set the flag. Otherwise, =\morecmidrules= itself +% does nothing. +% \begin{macrocode} +\def\morecmidrules{\noalign{\relax}} +% \end{macrocode} +% \end{macro} +% +% \begin{macrocode} +%</package> +% \end{macrocode} +% +% \Finale +% diff --git a/Master/texmf-dist/source/latex/booktabs/booktabs.ins b/Master/texmf-dist/source/latex/booktabs/booktabs.ins new file mode 100644 index 00000000000..93bfea5ff9c --- /dev/null +++ b/Master/texmf-dist/source/latex/booktabs/booktabs.ins @@ -0,0 +1,86 @@ +% ----------------------------------------------------------------- +% File: booktabs.ins +% Author: Simon Fear +% Maintainer: +% +% This file is part of the booktabs package for publication +% quality tables for LaTeX +% +% Copyright (C) 1995--2005 Simon Fear +% +% This program is free software; you can redistribute it and/or +% modify it under the terms of the GNU General Public License as +% published by the Free Software Foundation; either version 2 of +% the License, or (at your option) any later version. +% +% This program is distributed in the hope that it will be useful, +% but WITHOUT ANY WARRANTY; without even the implied warranty of +% MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the +% GNU General Public License for more details. +% +% You should have received a copy of the GNU General Public +% License along with this program; if not, write to the Free +% Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, +% MA 02111-1307 USA +% +% ----------------------------------------------------------------- +% This file will generate fast loadable files and documentation +% driver files from the doc files in this package when run through +% LaTeX or TeX. +% +% The file README' contains a list of all the files as well +% directions for installation of this package. +% ----------------------------------------------------------------- + +\def\batchfile{booktabs.ins} +\def\filedate{2005/03/16} +\input docstrip.tex +\keepsilent + +\preamble + +----------------------------------------------------------------- +Author: Simon Fear +Maintainer: + +This file is part of the booktabs package for publication +quality tables for LaTeX + + Copyright (C) 1995--2005 Simon Fear + +This program is free software; you can redistribute it and/or +modify it under the terms of the GNU General Public License as +published by the Free Software Foundation; either version 2 of +the License, or (at your option) any later version. + +This program is distributed in the hope that it will be useful, +but WITHOUT ANY WARRANTY; without even the implied warranty of +MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the +GNU General Public License for more details. + +You should have received a copy of the GNU General Public +License along with this program; if not, write to the Free +Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, +MA 02111-1307 USA +----------------------------------------------------------------- +\endpreamble + +\generate{% + \askforoverwritefalse + \file{booktabs.sty}{\from{booktabs.dtx}{package}}% +} + +{\obeyspaces% +\Msg{**********************************************************}% +\Msg{* *}% +\Msg{* To finish the installation you have to move the *}% +\Msg{* following file into a directory searched by TeX: *}% +\Msg{* *}% +\Msg{* booktabs.sty *}% +\Msg{* *}% +\Msg{* To produce the documentation run the file booktabs.dtx *}% +\Msg{* through LaTeX. *}% +\Msg{* *}% +\Msg{* Happy TeXing! *}% +\Msg{**********************************************************}} +\endbatchfile |