summaryrefslogtreecommitdiff
path: root/Master/texmf-dist
diff options
context:
space:
mode:
authorKarl Berry <karl@freefriends.org>2015-09-28 21:17:02 +0000
committerKarl Berry <karl@freefriends.org>2015-09-28 21:17:02 +0000
commit3b599d9910b4fe668c87ddd32a48d48f8a377560 (patch)
tree1eaf15a6fc5cc826af5507e4c768df4ff428f4a4 /Master/texmf-dist
parent22067fe830ec3090316787abecc6644540967b54 (diff)
fcolumn (28sep15)
git-svn-id: svn://tug.org/texlive/trunk@38489 c570f23f-e606-0410-a88d-b1316a301751
Diffstat (limited to 'Master/texmf-dist')
-rw-r--r--Master/texmf-dist/doc/latex/fcolumn/README41
-rw-r--r--Master/texmf-dist/doc/latex/fcolumn/fcolumn.pdfbin212093 -> 247757 bytes
-rw-r--r--Master/texmf-dist/source/latex/fcolumn/fcolumn.dtx788
-rw-r--r--Master/texmf-dist/source/latex/fcolumn/fcolumn.ins55
-rw-r--r--Master/texmf-dist/tex/latex/fcolumn/fcolumn.sty263
5 files changed, 701 insertions, 446 deletions
diff --git a/Master/texmf-dist/doc/latex/fcolumn/README b/Master/texmf-dist/doc/latex/fcolumn/README
index 27e64940f25..b9fbe8c82a0 100644
--- a/Master/texmf-dist/doc/latex/fcolumn/README
+++ b/Master/texmf-dist/doc/latex/fcolumn/README
@@ -11,9 +11,9 @@ version of a generic column F. The generic version expects three
arguments: |#1| is the separator, |#2| is the decimal mark, and |#3|
the coding used for grouping digits of the integer part and decimal
part. For example the f-column in the current version of the package
-is using 3.2 as #3, meaning that numbers are typeset with 2 decimal
-digits and grouping is done on 3 digits. The decimal separator is
-``,'' and the grouping character is ``.'', thereby coding for the
+is using 3,2 as #3, meaning that numbers are typeset with 2 decimal
+digits and grouping is done on 3 digits. The grouping character is
+``.'' and the decimal separator is ``,'', thereby coding for the
continental European standard. People in the Anglo-saxon world would
rather code |\newcolumntype{f}{F{,}{.}{3,2}}|.
@@ -29,20 +29,35 @@ cash 145,85 profit 27.542,62
---------- ----------
227.542,62 227.542,62
-The core part of the table is input as
-\begin{table}
+The core part of this table is input as
\begin{tabular}{lflf}
-house & 20000000 & equity capital & 5000000\\
-bank account & -60323 & mortgage & 15000000\\
-savings & 2800000\\
+house & 20000000 & equity capital & 5000000\\
+bank account & -60323 & mortgage & 15000000\\
+savings & 2800000 \\
cash & 14585 & profit & 2754262\\
\sumline
\end{tabular}
-\end{table}
The column specifier ``f'' (for ``finance'') is able to do the
typesetting. It constructs the sum line, typesets the numbers,
-calculates the totals, and checks whether the two columns are in
-balance; if not, the user is warned via a message. A special command
-\resetsumline exists which alllows one to restart a table, so that
-multiple tables in one tabular environment are possible.
+calculates the totals, makes rules of the correct widths, and
+checks whether the two columns are in balance; if not, the user
+is warned via a message. A special command \resetsumline exists
+which allows one to restart a table, so that multiple tables
+in one tabular environment are possible.
+
+To install:
+-----------
+Run the file fcolumn.ins through LaTeX (in a directory also
+containing fcolumn.dtx) and move the file fcolumn.sty to a place
+searched by TeX.
+
+To produce the documentation:
+-----------------------------
+Run the file fcolumn.dtx through LaTeX. Update the glossary
+and index by makeindex:
+
+ makeindex -s gglo.ist -o fcolumn.gls fcolumn.glo
+ makeindex -s gglo.ist -o fcolumn.ind fcolumn.idx
+
+and then LaTeX again; presto!
diff --git a/Master/texmf-dist/doc/latex/fcolumn/fcolumn.pdf b/Master/texmf-dist/doc/latex/fcolumn/fcolumn.pdf
index 33c18d4202e..e5a127dbebe 100644
--- a/Master/texmf-dist/doc/latex/fcolumn/fcolumn.pdf
+++ b/Master/texmf-dist/doc/latex/fcolumn/fcolumn.pdf
Binary files differ
diff --git a/Master/texmf-dist/source/latex/fcolumn/fcolumn.dtx b/Master/texmf-dist/source/latex/fcolumn/fcolumn.dtx
index 3e83646dfba..182e45670c0 100644
--- a/Master/texmf-dist/source/latex/fcolumn/fcolumn.dtx
+++ b/Master/texmf-dist/source/latex/fcolumn/fcolumn.dtx
@@ -1,60 +1,22 @@
% \iffalse meta-comment
-%
% Copyright (C) 2015 by Edgar Olthof
%
% This file may be distributed and/or modified under the
% conditions of the LaTeX Project Public License, either
-% version 1.2 of this license or (at your option) any later
+% version 1.3 of this license or (at your option) any later
% version. The latest version of this license is in
%
% http://www.latex-project.org/lppl.txt
%
-% and version 1.2 or later is part of all distributions of
+% and version 1.3 or later is part of all distributions of
% LaTeX version 1999/12/01 or later.
%
% \fi
% \iffalse
%<package>\NeedsTeXFormat{LaTeX2e}
%<package>\ProvidesPackage{fcolumn}
-%<package> [2015/03/07 v1.0 automatic currency package]
+%<package> [2015/09/20 v1.1.1 Automatic Currency Package (EO)]
%<package>\RequirePackage{array}
-%<*batchfile>
-\begingroup
-%%
-%% Copyright (C) 2015 by Edgar Olthof
-%%
-%% This file may be distributed and/or modified under the
-%% conditions of the LaTeX Project Public License, either
-%% version 1.2 of this license or (at your option) any later
-%% version. The latest version of this license is in
-%%
-%% http://www.latex-project.org/lppl.txt
-%%
-%% and version 1.2 or later is part of all distributions of
-%% LaTeX version 1999/12/01 or later.
-%%
-\input docstrip.tex
-\keepsilent
-\preamble
-
-This is a generated file.
-
-Copyright (C) 2015 by Edgar Olthof
-
-This file may be distributed and/or modified under the
-conditions of the LaTeX Project Public License, either
-version 1.2 of this license or (at your option) any later
-version. The latest version of this license is in
-
- http://www.latex-project.org/lppl.txt
-
-and version 1.2 or later is part of all distributions of
-LaTeX version 1999/12/01 or later.
-\endpreamble
-\askforoverwritefalse
-\generate{\file{fcolumn.sty}{\from{fcolumn.dtx}{package}}}
-\endgroup
-%</batchfile>
%<*driver>
\documentclass{ltxdoc}
\usepackage{fcolumn}
@@ -68,9 +30,7 @@ LaTeX version 1999/12/01 or later.
\end{document}
%</driver>
% \fi
-%
-% \CheckSum{0}
-%
+% \CheckSum{618}
% \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
@@ -89,95 +49,233 @@ LaTeX version 1999/12/01 or later.
%
% \changes{v0.1}{2015/02/06}{First working version}
% \changes{v1.0}{2015/03/07}{Three-argument version is working properly}
+% \changes{v1.1}{2015/09/14}{Automatic checking of column balance performed when number of
+% F-columns is even (behaviour can be overridden). Empty entries are now recognised and
+% correctly treated as such, except for the one ended by the double backslash. Not serious;
+% workaround possible. Furthermore optimisation of code: minimised the number of private
+% counts and resetting of column counter done in a nicer way.}
+% \changes{v1.1.1}{2015/09/27}{Installation procedure changed from .ins-in-.dtx to separate
+% .ins and .dtx after discussion with Karl Berry as well as some minor code improvements.}
% \GetFileInfo{fcolumn.sty}
-% \DoNotIndex{\fi,\let,\@ne}
+% \DoNotIndex{\@arstrut,\@arstrutbox,\@depth,\@firstampfalse,\@firstamptrue,\@gobble,\@halignto}
+% \DoNotIndex{\@height,\@mkpream,\@ne,\@nextchar,\@preamble,\@sharp,\@tempdima,\@tempswafalse}
+% \DoNotIndex{\@tempswatrue,\@\temptokena,\@tfor,\@whilesw,\@width,\advance,\arraystretch,\baselineskip}
+% \DoNotIndex{\begingroup,\bgroup,\count,\count@,\cr,\csname,\def,\divide,\do,\dp,\edef,\egroup}
+% \DoNotIndex{\else,\endcsname,\endgroup,\expandafter,\extrarowheight,\fi,\gdef,\halignto,\hbox}
+% \DoNotIndex{\hfil,\ht,\if,\if@firstamp,\if@tempswa,\ifcase,\lineskip,\loop,\m@ne,\m@th,\message}
+% \DoNotIndex{\multiply,\or,\par,\prepnext@tok,\protect,\repeat,\setbox,\space,\string,\strutbox}
+% \DoNotIndex{\tabskip,\tabularnewline,\the,\the@toks,\thr@@,\tw@,\unhbox,\vbox,\vcenter,\vrule}
+% \DoNotIndex{\vskip,\vtop,\wd,\xdef,\z@,\z@skip}
% \title{The \textsf{fcolumn} package\thanks{This file has version
% number \fileversion, last revised \filedate.}}
-% \author{Edgar Olthof}
+% \author{Edgar Olthof\\
+% \texttt{edgar <dot> olthof <at> inter <dot> nl <dot> net}}
% \date{Printed \today}
% \maketitle
% \begin{abstract}
% In financial reports, text and currency amounts are regularly put in
% one table, e.g., a year balance or a profit-and-loss overview.
-% This package provides the settings for automatically typesetting
+% This package provides the settings for automatically typesetting and checking
% such columns, including the sum line (preceded by a rule of the correct
% width), using the specifier |f|.
+% \end{abstract}
%
-% The column specifier |f| itself is rather simple. It is the predefined version
-% of a generic column |F|. The generic version expects three arguments:
-% |#1| is the separator, |#2| is the decimal mark, and |#3| the coding used for
+% \section{Introduction}\label{intro}
+% The package |fcolumn| provides the macros for an extra tabular specifier that makes creating
+% financial tables easy. The column specifier |f| itself is rather simple. It is the
+% predefined version of a generic column |F|. The generic version expects three arguments:
+% |#1| is the group separator, |#2| is the decimal mark, and |#3| the coding used for
% grouping digits of the integer part and decimal part.
%
-% The f-column in the current version of the package is defined as
-% |\newcolumntype{f}{F{.}{,}{3,2}}| (which is the continental European standard, implemented here)
-% and people in the Anglo-saxon world would rather code |\newcolumntype{f}{F{,}{.}{3,2}}|.
+% The f-column in the current version of the package is defined for the continental European
+% standard: |\newcolumntype{f}{F{.}{,}{3,2}}|. This means that a number like 1234567 will be
+% typeset as $12{.}345{,}67$. People in the Anglo-saxon world would rather
+% code |\newcolumntype{f}{F{,}{.}{3,2}}|, yielding $12{,}345{.}67$ for the number given above.
% The default value for |#3| is 3,2, indicating that grouping of the integer part
% is by three digits and that the decimal part consists of two digits.
% If however, in your country or company grouping is done with a thinspace every
% four digits and there are three digits after the decimal mark---that happens to
% be a |\cdot|---, then simply specify |\newcolumntype{f}{F{\,}{\cdot}{4,3}}| in
-% that case. By default two digits are used for the decimal part, so if you really
+% that case.
+%
+% By default two digits are used for the decimal part, so if you really
% want no decimal digits (in that case of course also skipping the decimal mark)
% you have to explicitly specify |x,0|.
-% \end{abstract}
-%
-% \section{Introduction}
-% To show where and how the f-column is used, let's look at a typical financial table.
%
+% This package requires and loads the |array| package~\cite{array}.
+% To show where and how the F-column is used, let's look at some typical financial information
+% as shown in Table~\ref{tab:ex1}
% \begin{table}[htb]
+% \caption{Example Table}
+% \label{tab:ex1}
% \begin{tabular}{@{}lflf@{}}
% \multicolumn4c{\bfseries Balance sheet}\\
-% \multicolumn1{@{}l}{properties}&\multicolumn1r{31 dec 2014}
-% &\multicolumn1l{debts}&\multicolumn1{l@{}}{31 dec 2014}\\
-% \hline
+% \toprule
+% properties&\multicolumn1r{31 dec 2014}&
+% debts&\multicolumn1{l@{}}{31 dec 2014}\\
+% \midrule
% house & 20000000 & equity capital& 5000000\\
% bank account& -60323 & mortgage & 15000000\\
-% savings& 2800000 & & \leeg \\
+% savings& 2800000 \\
% cash& 14585 & profit & 2754262\\
% \sumline
-% \check12
-% \hline
+% \bottomrule
% \end{tabular}
% \end{table}
-%
-% This table was entered as:
+% and how this is entered in \LaTeX\ (Table~\ref{tab:ex2}).
+% \begin{table}[htb]
+% \caption{Verbatim version of Example Table}
+% \label{tab:ex2}
% \begin{verbatim}
% \begin{table}[htb]
+% \caption{Example Table}
+% \label{tab:ex1}
% \begin{tabular}{@{}lflf@{}}
% \multicolumn4c{\bfseries Balance sheet}\\
-% \multicolumn1{@{}l}{properties}&\multicolumn1r{31 dec 2014}
-% &\multicolumn1l{debts}&\multicolumn1{l@{}}{31 dec 2014}\\
-% \hline
+% \toprule
+% properties&\multicolumn1r{31 dec 2014}&
+% debts&\multicolumn1{l@{}}{31 dec 2014}\\
+% \midrule
% house & 20000000 & equity capital& 5000000\\
% bank account& -60323 & mortgage & 15000000\\
-% savings& 2800000 & & \leeg \\
+% savings& 2800000 \\
% cash& 14585 & profit & 2754262\\
% \sumline
-% \check12
-% \hline
+% \bottomrule
% \end{tabular}
% \end{table}
% \end{verbatim}
+% \end{table}
% All the work was done by the column specifier ``f'' (for ``finance''). In this case
-% it constructs the |\sumline|, typesets the numbers, calculates the totals,
-% and checks whether the two columns are in balance; if not, the user is warned
-% via a |\message|. Of course for nice settings one should use the
-% |booktabs| package, but that is not the point here.
+% it constructs the |\sumline|, typesets the numbers, calculates the totals, determines
+% the widths of the sumrules, and checks whether the two columns are in balance; if not,
+% the user is warned via a |\message|. Of course for nice settings the
+% |booktabs| package~\cite{booktabs} was used, but that is not the point here.
%
-% This package is heavily inspired by the |dcolumn| package by David Carlisle,
+% This package is heavily inspired by the |dcolumn| package by David Carlisle~\cite{dcolumn},
% some constructions are more or less copied from that package.
%
-% There is at least one things to improve in this version:
-% An empty entry in an f-column should result in an empty spot in the table.
-% Now an empty entry is not correct and is repaired by \LaTeX\ to be equal to
-% 0, printing a number in the table. This can be prevented by explicitly specifying
-% |\leeg|, like in the table above, but that should be automated.
-%
+% \section{Commands}
+% The user only needs to know six commands or constructions. These six are given here.
+% \begin{macro}{F}
+% In the tabular the column specifier |F| can be given with arguments, or the predefined
+% version |f|, where the three arguments of |F| are |{.}|, |{,}|, and |3,2|. If you
+% want |g| to be your own definition like the curious one given in Section~\ref{intro}, then
+% specify |\newcolumntype{g}{F{\,}{\cdot}{4,3}}| prior to using |g| in a tabular.
+%
+% Entries in an F-column are from that moment on, treated as (integer) numbers unless
+% explicitly escaped by |\leeg|, see below. The numbers are typeset according to the template
+% the user gives with his/her |F|-column.
+% \end{macro}
+% \begin{macro}{\sumline}
+% The numbers in an |F|-column are typeset as a financial amount, but the real benefit comes
+% with the |\sumline|. It does three things:
+% \begin{compactitem}
+% \item It calculates the total of the column so far and the maximum width encountered
+% so far, including the width of the total;
+% \item It generates a rule with width calculated in the first item;
+% \item It checks the columns that are supposed to balance whether or not they actually do.
+% If so, nothing happens. If not, a |\message| is given that column $i$ and $j$ do not balance,
+% where $i$ and $j$ are the relevant columns. This is only done if the total number of
+% |F|-columns is even, e.g., if there are six |F|-columns, then~1 is checked against~4, 2 against~5,
+% and 3 against~6. If the number of |F|-columns is odd then anything could be possible in that table
+% and nothing is assumed about structure within the table. This behaviour can be overridden, see below.
+% \end{compactitem}
+% \end{macro}
+% \begin{macro}{\resetsumline}
+% Suppose you want to typeset one tabular with the profit-and-loss of many projects individually.
+% The layout of those tabulars is the same and it were nice if all columns were aligned. This can
+% be done by making it one big tabular with a fresh start for each project. The macro |\resetsumlines|
+% is used for that: it resets all totals and all column widths, see for example Table~\ref{tab:ex3}.
+% \begin{table}[htb]
+% \caption{Example: multiple projects}
+% \label{tab:ex3}
+% \begin{tabular}{@{}lfflff@{}}
+% \multicolumn6c{\bfseries Project~1}\\
+% \toprule
+% expense&\multicolumn1r{actual}&\multicolumn1r{budget}&
+% income&\multicolumn1r{actual}&\multicolumn1{r@{}}{budget}\\
+% \midrule
+% food & 45000 & 50000 & tickets& 100000 & 100000\\
+% drinks & 35000 & 40000 \\
+% music & 18000 & 10000 \\
+% profit & 2000 \\
+% \sumline
+% \resetsumline
+% \multicolumn6c{\bfseries Project~2}\\
+% \toprule
+% expense&\multicolumn1r{actual}&\multicolumn1r{budget}&
+% income&\multicolumn1{r}{actual}&\multicolumn1{r@{}}{budget}\\
+% \midrule
+% food & 25000 & 30000 & tickets& 40000 & 45000\\
+% drinks & 10000 & 8000 \\
+% music & 8000 & 7000 & loss & 3000 \\
+% \sumline
+% \bottomrule
+% \end{tabular}
+% \end{table}
+% Note that the rules in the first and third F-columns of project~1 cover $1{.}000{,}00$ whereas
+% in project~2 those rules are shorter since they only cover $430{,}00$; still the columns are
+% aligned. The verbatim way of setting up Table~\ref{tab:ex3} is given in Table~\ref{tab:ex4}.
+% \begin{table}[htb]
+% \caption{Verbatim version of Table~\ref{tab:ex3}}
+% \label{tab:ex4}
+% \begin{verbatim}
+% \begin{table}[htb]
+% \caption{Example: multiple projects}
+% \label{tab:ex3}
+% \begin{tabular}{@{}lfflff@{}}
+% \multicolumn6c{\bfseries Project~1}\\
+% \toprule
+% expense&\multicolumn1r{actual}&\multicolumn1r{budget}&
+% income&\multicolumn1r{actual}&\multicolumn1{r@{}}{budget}\\
+% \midrule
+% food & 45000 & 50000 & tickets& 100000 & 100000\\
+% drinks & 35000 & 40000 \\
+% music & 18000 & 10000 \\
+% profit & 2000 \\
+% \sumline
+% \resetsumline
+% \multicolumn6c{\bfseries Project~2}\\
+% \toprule
+% expense&\multicolumn1r{actual}&\multicolumn1r{budget}&
+% income&\multicolumn1r{actual}&\multicolumn1{r@{}}{budget}\\
+% \midrule
+% food & 25000 & 30000 & tickets& 40000 & 45000\\
+% drinks & 10000 & 8000 \\
+% music & 8000 & 7000 & loss & 3000 \\
+% \sumline
+% \bottomrule
+% \end{tabular}
+% \end{table}
+% \end{verbatim}
+% \end{table}
+% \end{macro}
+% \begin{macro}{\leeg}
+% If an |F|-column should be empty then simply leave it empty. If however it should not be empty
+% but the entry should be treated as text---even it is a number---this can be done with |\leeg|.
+% It expects an argument and this argument is typeset in the column, affecting the maximum column
+% width so far. The common case is where |p.m.| (\textit{pro memoria}) is entered. An empty |F|-column
+% followed by |\\| does not work: either remove the |&| or specify |\leeg{}|.
+% \end{macro}
+% \begin{macro}{\checkfcolumns}
+% The automatic column balance check can also be done manually. If |F|-columns~1 and~4 should balance
+% and you want them to be checked, then simply say |\checkfcolumns14|. With more than nine |F|-columns
+% you may be forced to say something like |\checkfcolumns{10}{12}|.
+% \end{macro}
+% \begin{macro}{\strictaccounting}
+% In the rare occasion that a negative number occurs in a financial table, the sign of that number
+% can be an explicit minus sign ($-$) or the number is coloured red, or it is typeset between parentheses,
+% and there may be even other ways. By default (for aesthetic reasons) |fcolumn| typesets it with
+% a minus~sign, but strict accounting prescibes that the number should be put between parentheses.
+% The latter can be accomplished by setting |\strictaccountingtrue|.
+% \end{macro}
+% \StopEventually{\PrintChanges\PrintIndex}
% \section{The macros}
% \begin{macro}{column F}
% \begin{macro}{column f}
-%
-% Note that the definition of the column type |f| does not use private macros
+% The column specifier |F| is the generic one, and |f| is the default (European) one
+% for easy use. Note that the definition of the column type |f| does not use private macros
% (no |@|), so overriding its definition is easy for a user.
% \begin{macrocode}
\newcolumntype{F}[3]{>{\b@fi{#1}{#2}{#3}}r<{\e@fi}}
@@ -186,18 +284,20 @@ LaTeX version 1999/12/01 or later.
% \end{macro}
% \end{macro}
% \begin{macro}{\FCsc@l}
+% \begin{macro}{\FCtc@l}
% Two \meta{count}s are defined, that both start at zero: the \meta{count}
-% |\FCsc@l|, that keeps track at which f-column the tabular is working on and
-% the \meta{count} |\FCtc@l|, that records the number of f-columns that were
+% |\FCsc@l|, that keeps track at which F-column the tabular is working on and
+% the \meta{count} |\FCtc@l|, that records the number of F-columns that were
% encountered so far. Later in the package the code can be found for generating
-% a new \meta{count} and a new \meta{dimen} if the number of requested f-columns
-% is larger than currently available. This is of course the case when an f-column
+% a new \meta{count} and a new \meta{dimen} if the number of requested F-columns
+% is larger than currently available. This is of course the case when an F-column
% is used for the first time.
% \begin{macrocode}
-\newcount\FCsc@l \FCsc@l0
-\newcount\FCtc@l \FCtc@l0
+\newcount\FCsc@l \FCsc@l=0
+\newcount\FCtc@l \FCtc@l=0
% \end{macrocode}
% \end{macro}
+% \end{macro}
% \begin{macro}{\geldm@cro}
% The macro |\geldm@cro| takes a number and by default interprets this as
% an amount expressed in cents (dollar cents, euro cents, centen,
@@ -207,68 +307,68 @@ LaTeX version 1999/12/01 or later.
% As explained, this can be changed. It uses a private boolean |\withs@p|
% and an accessable, i.e., without |@| boolean: |\strictaccounting|. The latter
% is used to typeset negative numbers between parentheses. By default it doesn't
-% do this: a minus sign is used. It also uses two private \meta{count}s. I'm still
-% figuring out whether this is needed; everything is done local within a column
-% of an |\ialign|, so using |\count0|, |\count1|, etc.\ might also work: it would
-% safe two \meta{count}s.
+% do this: a minus sign is used.
% \begin{macrocode}
-\newcount\g@lda
-\newcount\g@ldb
\newif\ifwiths@p
\newif\ifstrictaccounting \strictaccountingfalse
% \end{macrocode}
% Actually |\geldm@cro| is only a wrapper around |\g@ldm@cro|.
% \begin{macrocode}
\def\geldm@cro#1#2{\withs@pfalse
- \afterassignment\g@ldm@cro\count@#2\relax{#1}}
+\afterassignment\g@ldm@cro\count@#2\relax{#1}}
% \end{macrocode}
% \end{macro}
% \begin{macro}{\g@ldm@cro}
% This macro starts by looking at the sign of |#2|: if it is negative, it prints the correct
-% indicator (a parenthesis or a minus sign), assigns the absolute value of |#2| to |\g@lda|
-% and goes on. Note that |\geldm@cro| and therefore |\g@ldm@cro| are used within |$|s, so
-% it is really a minus sign that is printed, not a hyphen.
+% indicator (a parenthesis or a minus sign), assigns the absolute value of |#2| to |\count2|
+% and goes on. Note that |\geldm@cro| and therefore |\g@ldm@cro| are always used within |$|s, so
+% it is really a minus sign that is printed, not a hyphen. All calculations are done with
+% |\count0|, |\count1|, etc.\ i.e., without fcolumn-specific \meta{count}s because it is all
+% done locally. Leaving the tabular environment will restore their values.
% \begin{macrocode}
\def\g@ldm@cro#1\relax#2{%
- \ifnum#2<0 \ifstrictaccounting(\else-\fi\g@lda=-#2 \else\g@lda=#2 \fi
+\ifnum#2<0 \ifstrictaccounting(\else-\fi\count2=-#2 \else\count2=#2 \fi
% \end{macrocode}
% Calculate the entire currency units: this is the result of $x/a$ as integer
% division, with $a=10^n$ and $n$ the part of |#1| after the separator (if any).
% Here the first character of |#1| is discarded, so the separator in |#1| is not strict:
% you could also specify |3.2| instead of |3,2| (or even |3p2|).
% \begin{macrocode}
- \ifx\relax#1\relax\g@ldb=2 \else \g@ldb\@gobble#1\relax\fi
- \loop
- \ifnum\g@ldb>0 \divide\g@lda by 10 \advance\g@ldb by \m@ne
- \repeat
+\count3=\ifx\relax#1\relax2 \else \@gobble#1\relax\fi
+\count4=\count3
+\loop
+ \ifnum\count3>0 \divide\count2 by 10 \advance\count3 by \m@ne
+\repeat
% \end{macrocode}
-% The value in |\g@lda| is then output by |\g@ldens| using the separation given.
+% The value in |\count2| is then output by |\g@ldens| using the separation given.
% \begin{macrocode}
- \g@ldens{\the\count@}%
+\g@ldens{\the\count@}%
% \end{macrocode}
% If there is a decimal part\dots
% \begin{macrocode}
- \ifx\relax#1\relax\g@ldb=2 \else \g@ldb\@gobble#1\relax\fi
- \ifnum\g@ldb>0\decim@lmark
+\ifnum\count4>0\decim@lmark
% \end{macrocode}
% Next the decimal part is dealt with. Now $x\bmod a$ is calculated in the
-% usual way: $x-(x/a)*a$. The minus sign necessary for this calculation is
-% introduced in the next line by changing the comparison to |>| instead of |<|.
-% \begin{macrocode}
- \ifnum#2>0 \g@lda=-#2\else\g@lda=#2 \fi
- \loop
- \ifnum\g@ldb>0 \divide\g@lda by 10 \advance\g@ldb by \m@ne
- \repeat
- \ifx\relax#1\relax\g@ldb=2 \else \g@ldb\@gobble#1\relax\fi
- \loop
- \ifnum\g@ldb>0 \multiply\g@lda by 10 \advance\g@ldb by \m@ne
- \repeat
- \ifnum#2>0 \advance\g@lda by #2
- \else \advance\g@lda by -#2
- \fi
- \ifx\relax#1\relax\g@ldb=2 \else \g@ldb\@gobble#1\relax\fi
- \zerop@d{\number\g@ldb}{\number\g@lda}%
+% usual way: $x-(x/a)*a$ with integer division. The minus sign necessary
+% for this calculation is introduced in the next line by changing the
+% comparison from |<| to |>|.
+% \begin{macrocode}
+ \ifnum#2>0 \count2=-#2\else\count2=#2 \fi
+ \count3=\count4
+ \loop
+ \ifnum\count3>0 \divide\count2 by 10 \advance\count3 by \m@ne
+ \repeat
+% \end{macrocode}
+% The value of |\count3| is~0 now. Now counting up again, this saved an assignment.
+% \begin{macrocode}
+ \loop
+ \ifnum\count3<\count4 \multiply\count2 by 10 \advance\count3 by \@ne
+ \repeat
+ \ifnum#2>0 \advance\count2 by #2
+ \else \advance\count2 by -#2
\fi
+ \zerop@d{\number\count3}{\number\count2}%
+\fi
% \end{macrocode}
% If the negative number is indicated by putting it between parentheses, then the
% closing parenthesis should stick out of the column, otherwise the alignment of
@@ -277,53 +377,52 @@ LaTeX version 1999/12/01 or later.
% may even stick out of the table. I don't like this, therefore I chose to put
% |\strictaccountingfalse|. Change if you like.
% \begin{macrocode}
- \ifnum#2<0 \ifstrictaccounting\rlap{)}\fi\fi}
+\ifnum#2<0 \ifstrictaccounting\rlap{)}\fi\fi}
% \end{macrocode}
% \end{macro}
% \begin{macro}{\g@ldens}
% Here the whole currency units are dealt with. The macro |\g@ldens| is used
% recursively, therefore the double braces; this allows to use |\count0| locally.
% Tail recursion is not possible here (at least I don't know of a way), but that
-% is not very important, as the largest number ($2^{31}-1$) will only cause a
+% is not very important, as the largest number (which is $2^{31}-1$) will only cause a
% threefold recursion. This also implies that the largest amount this package
% can deal with is 2.147.483.647 (using |x.0|). For most people this is probably
% more than enough if the currency is euros or dollars. If not, then it is
% more likely that you are spending your time on the beach than studying this package.
% And otherwise make clear that you use a currency unit of k\$.
% \begin{macrocode}
-\def\g@ldens#1{{\g@ldb=\g@lda \count0=#1
+\def\g@ldens#1{{\count3=\count2 \count0=#1
% \end{macrocode}
% First divide by $10^n$, where $n$ is |#1|.
% \begin{macrocode}
- \loop
- \ifnum\count0>0
- \divide\g@lda by 10
- \advance\count0 by \m@ne
- \repeat
+\loop
+ \ifnum\count0>0
+ \divide\count2 by 10
+ \advance\count0 by \m@ne
+\repeat
% \end{macrocode}
% Here is the recursive part,
% \begin{macrocode}
- \ifnum\g@lda>0 \g@ldens{#1}\fi
+\ifnum\count2>0 \g@ldens{#1}\fi
% \end{macrocode}
% and then reconstruct the rest of the number.
% \begin{macrocode}
- \count0=#1
- \loop
- \ifnum\count0>0
- \multiply\g@lda by 10
- \advance\count0 by \m@ne
- \repeat
- \g@lda=-\g@lda
- \advance\g@lda by \g@ldb \du@zendprint{#1}}}
+\count0=#1
+\loop
+ \ifnum\count0>0
+ \multiply\count2 by 10
+ \advance\count0 by\m@ne
+\repeat
+\count2=-\count2
+\advance\count2 by \count3 \du@zendprint{#1}}}
% \end{macrocode}
% \end{macro}
% \begin{macro}{\du@zendprint}
% The macro |\du@zendprint| takes care for correctly printing the separator
% and possible trailing zeros.
% \begin{macrocode}
-\def\du@zendprint#1{\ifwiths@p\sep@rator\zerop@d{#1}{\number\g@lda}%
- \else\zerop@d{1}{\number\g@lda}\fi
- \global\withs@ptrue}
+\def\du@zendprint#1{\ifwiths@p\sep@rator\zerop@d{#1}{\number\count2}%
+ \else\zerop@d{1}{\number\count2}\fi \global\withs@ptrue}
% \end{macrocode}
% \end{macro}
% \begin{macro}{\zerop@d}
@@ -343,44 +442,31 @@ LaTeX version 1999/12/01 or later.
% First determine the number of digits of |#2| (expressed in the decimal
% system). This number is in |\count0| and is at least~1.
% \begin{macrocode}
- \loop
- \divide \count1 by 10
- \ifnum\count1>0 \advance\count0 by 1
- \repeat
+\loop
+ \divide \count1 by 10
+ \ifnum\count1>0 \advance\count0 by\@ne
+\repeat
% \end{macrocode}
% The number of zeros to be padded is $\max(0,|#1-\count0|)$ (the second
% argument can be negative), so a simple loop suffices.
% \begin{macrocode}
- \loop
- \ifnum\count0<#1 0\advance\count0 by 1
- \repeat
- \number#2}}
+\loop
+ \ifnum\count0<#1\relax 0\advance\count0 by\@ne
+\repeat
+\number#2}}
% \end{macrocode}
% \end{macro}
% \begin{macro}{\zetg@ld}
% This macro takes care for several things: it increases the total for a given
-% f-column, it records the largest width of the entries in that column and it
+% F-column, it records the largest width of the entries in that column and it
% typesets |#1| via |\geldm@cro|.
% \begin{macrocode}
\def\zetg@ld#1#2{\global\advance\csname
- FCtot@\romannumeral\FCsc@l\endcsname by #1
- \setbox0=\hbox{$\geldm@cro{#1}{#2}$}%
- \ifdim\wd0>\csname FCwd@\romannumeral\FCsc@l\endcsname
- \global\csname FCwd@\romannumeral\FCsc@l\endcsname=\wd0
- \fi\unhbox0}
-% \end{macrocode}
-% \end{macro}
-% \begin{macro}{\check}
-% The macro |\check| provides a way to the user to check that the appropriate
-% columns are balanced (as it should in a balance). Arguments |#1| and |#2| are
-% the column numbers to compare. It is the responsibility of the user to provide
-% the correct numbers here, otherwise bogus output is generated. But the output is
-% only to screen and the transcript file; it doesn't change the appearance of your
-% document, so you can safely ignore the result then, repair and go on.
-% \begin{macrocode}
-\def\check#1#2{\noalign{\ifnum\csname FCtot@\romannumeral#1\endcsname=
- \csname FCtot@\romannumeral#2\endcsname\else
- \message{Columns #1 and #2 do not balance!}\fi}}
+FCtot@\romannumeral\FCsc@l\endcsname by #1
+\setbox0=\hbox{$\geldm@cro{#1}{#2}$}%
+\ifdim\wd0>\csname FCwd@\romannumeral\FCsc@l\endcsname
+ \global\csname FCwd@\romannumeral\FCsc@l\endcsname=\wd0
+\fi\unhbox0}
% \end{macrocode}
% \end{macro}
% \begin{macro}{\b@fi}
@@ -392,84 +478,107 @@ LaTeX version 1999/12/01 or later.
% The plain \LaTeX\ command |\@empty| is used for that. The separator and decimal mark
% are within a math environment, so you can indeed specify |\,| instead of |\thinspace|,
% but there is an extra brace around, so it doesn't affect
-% the spacing between the digits (trick copied from |dcolumn|).
+% the spacing between the digits (trick copied from |dcolumn|, Ref.~\cite{dcolumn}).
% \begin{macrocode}
\newcount\bedr@g
\def\b@fi#1#2#3{\def\sep@rator{{#1}}\def\decim@lmark{{#2}}%
- \let\ignorespaces=\@empty \def\sp@l{#3}\global\advance\FCsc@l by \@ne
+\def\sp@l{#3}\let\ignorespaces=\@empty \let\unskip=\@empty
+\global\advance\FCsc@l by \@ne
% \end{macrocode}
-% The value specified by the user is then captured by |\bedr@g|.
+% The value specified by the user is then captured by |\bedr@g| and this is done in a
+% special way: |\bedr@g| is assigned globally within |\box0|. Why? To later check
+% the width of this box, see |\e@fi|.
% \begin{macrocode}
- \bedr@g=}
+\setbox0=\hbox\bgroup\global\bedr@g=}
% \end{macrocode}
% \end{macro}
% \begin{macro}{\e@fi}
-% Once captured, the rest is handled by |\e@fi|. The number is then
-% fed to |\zetg@ld| for the current column.
-% \begin{macrocode}
-\def\e@fi{\relax \zetg@ld{\number\bedr@g}{\sp@l}}
+% If the user enters a number, say 10, \LaTeX\ will read
+% |\setbox| |0| |=| |\hbox| |\bgroup| |\global| |\bedr@g| |=| |10| |123| |\relax| |\egroup|, which
+% is perfectly valid. The count |\bedr@g| is~10 now and |\box0| is non-empty, in fact containing
+% the number 123 (although any number would be good: it is discarded anyway). If however nothing
+% was entered by the user, \LaTeX\ will read |\setbox| |0| |=| |\hbox| |\bgroup| |\global| |\bedr@g|
+% |=| |123| |\relax| |\egroup|, which is also perfectly valid, but now |\bedr@g| is~123 and |\box0|
+% is empty. So the width of |\box0| can be used to discriminate without actually using its contents
+% in the horizontal list. Later, with the code for |\leeg| a related trick will be used. The space
+% at the beginning of |\e@fi| is important, otherwise the 123 might be concatenated with the user
+% entry, leading to wrong numbers.
+% \begin{macrocode}
+\def\e@fi{ 123\relax\egroup\ifdim\wd0>\z@
+ \zetg@ld{\number\bedr@g}{\sp@l}%
+\fi}
% \end{macrocode}
+% Please note that this trick does not work for an empty F-column that is ended by |\\|. The reason
+% is that the first thing |\\| does---even before providing the |\cr| ---is to issue a |\relax|.
+% This stops the assignment to |\bedr@g| and causes a \TeX\ error, complaining about a missing number.
+% This is then repaired by \TeX\ by inserting a zero, so it doesn't affect the total of that column,
+% but in particular this zero is then typeset, which it is not supposed to. I decided to leave it
+% this way because in practice this is never a problem: simply discard the last |&| and the column
+% before the offending column will now be ended by |\\|. Iterate for this line until a filled F-column
+% or a non-F-column is encountered. This is assuming that you did not specify \verb+|+s in the tabular
+% description: for them to work for the whole table you should have all |&|s in place (this is true in
+% general, not only for |fcolumn|). On the other hand you should never, ever use vertical rules in
+% tables, according to Ref.~\cite{booktabs}. And if you really must, then fill the empty F-columns
+% with |\leeg{}|.
% \end{macro}
%
% Here are adaptations to existing macros.
-% \begin{macro}{\@xarraycr}
-% Among all its tasks, now |\\| also resets the column number. This
-% is a transparant change, i.e., loading this package without using
-% the ``f'' column specifier does
-% not change the effect of package |array.sty| in any way.
-% This resetting has to be done after the |\cr|, so it is stored
-% in a |\noalign|.
-% \begin{macrocode}
-\def\@xarraycr{\@ifnextchar [%
- \@argarraycr {\ifnum 0=`{}\fi \cr\noalign{\global\FCsc@l=0 }}}
-% \end{macrocode}
-% \end{macro}
% \begin{macro}{\@array}
% The definition of |\@array| had to be extended slightly because
% it should also include |\@mksumline| (acting on the same |#2|
-% as |\@mkpream| gets). Also this change is transparant: it only adds
+% as |\@mkpream| gets). This change is transparant: it only adds
% functionality and if you don't use that, you won't notice the difference.
+% It starts by just copying the original definition from the |array| package~\cite{array}.
% \begin{macrocode}
\def\@array[#1]#2{%
- \@tempdima \ht \strutbox
- \advance \@tempdima by\extrarowheight
- \setbox \@arstrutbox \hbox{\vrule
- \@height \arraystretch \@tempdima
- \@depth \arraystretch \dp \strutbox
- \@width \z@}%
- \begingroup
- \@mkpream{#2}%
- \xdef\@preamble{\noexpand \ialign \@halignto
- \bgroup \@arstrut \@preamble
- \tabskip \z@ \cr}%
+\@tempdima \ht \strutbox
+\advance \@tempdima by\extrarowheight
+\setbox \@arstrutbox \hbox{\vrule
+ \@height \arraystretch \@tempdima
+ \@depth \arraystretch \dp \strutbox
+ \@width \z@}%
+% \end{macrocode}
+% Here comes the first change: after each |\\| (or |\cr| for that matter) the \meta{count}
+% |\FCsc@l| should be reset. This is easiest done with |\everycr|, but |\everycr| is
+% put to |{}| by |\ialign|, so that definition should change. The resetting should
+% be done globally.
+% \begin{macrocode}
+\def\ialign{\everycr{\noalign{\global\FCsc@l=0 }}%
+ \tabskip\z@skip\halign}
+% \end{macrocode}
+% Then the definition is picked up again.
+% \begin{macrocode}
+\begingroup
+\@mkpream{#2}%
+\xdef\@preamble{\noexpand \ialign \@halignto
+\bgroup\@arstrut\@preamble\tabskip\z@\cr}%
+\endgroup
% \end{macrocode}
% The combination |\endgroup| followed by |\begingroup| seems redundant, but that is
% not the case: the |\endgroup| restores everything that was not |\global|. With the
% following |\begingroup| it is ensured that |\@mksumline| experiences the same
% settings as |\@mkpream| did.
% \begin{macrocode}
- \endgroup
- \begingroup
- \@mksumline{#2}%
- \endgroup
+\begingroup
+\@mksumline{#2}%
+\endgroup
% \end{macrocode}
% As a side product of |\@mksumline| also the \meta{count}s for the totals and
% \meta{dimen}s for the widths of the colums are created. The columns should start
-% fresh, i.e., totals are 0 and widths are
-% 0\thinspace pt.
+% fresh, i.e., totals are~0 and widths are 0\thinspace pt.
% \begin{macrocode}
- \res@tsumline
+\res@tsumline
% \end{macrocode}
% From here on it is just the old definition of |array.sty|.
% \begin{macrocode}
- \@arrayleft
- \if #1t\vtop \else \if#1b\vbox \else \vcenter \fi \fi
- \bgroup
- \let \@sharp ##\let \protect \relax
- \lineskip \z@
- \baselineskip \z@
- \m@th
- \let\\\@arraycr \let\tabularnewline\\\let\par\@empty \@preamble}
+\@arrayleft
+\if #1t\vtop \else \if#1b\vbox \else \vcenter \fi \fi
+\bgroup
+\let \@sharp ##\let \protect \relax
+\lineskip \z@
+\baselineskip \z@
+\m@th
+\let\\\@arraycr \let\tabularnewline\\\let\par\@empty \@preamble}
% \end{macrocode}
% Because |\@array| was changed here and it is this version that should be used,
% |\@@array| should be |\let| equal to |\@array| again.
@@ -484,20 +593,22 @@ LaTeX version 1999/12/01 or later.
% Furthermore most entries will simply add nothing to |\sumline|, e.g., |@|, |!|,
% and \verb+|+ can be fully ignored. Ampersands are only inserted by |c|, |l|, |r|,
% |p|, |m|, and |b|. So a specifier like |@{}lflf@{}| will yield the sumline
-% |&\a&&\a\\|, where |\a| is a macro that prints the desired result of the column.
-% Had the specifier been \verb+l|f||@{ }l|f+, then the same sumline must
+% |&\a&&\a\\|, (where |\a| is a macro that prints the desired result of the column,
+% see later). Had the specifier been \verb+l|f||@{ }l|f+, then the same sumline must
% be constructed: all difficulties are already picked up and solved in the creation
% of the preamble.
%
% In reality the sumline must be constructed from the expanded form of the specifier,
-% so |@{}lf@{}| will expand as |@{}l>{\b@fi#1#2#2}r<{\e@fi}@{}|. The
+% so |@{}lf@{}| will expand as |@{}l>{\b@fi{.}{,}{3,2}}r<{\e@fi}@{}|. The
% rules for constructing the sumline are now very simple:
% \begin{compactitem}
% \item add an ampersand when |c|, |l|, |r|, |p|, |m|, or |b| is found, unless it is
% the first one (this is the same as in the preamble);
% \item add a |\a| when |<{\e@fi}| is found;
-% \item ignore everything else.
+% \item ignore everything else;
+% \item close with a |\\|.
% \end{compactitem}
+% (In reality also the column check is inserted just before the |\\|, see |\aut@check|.)
% To discriminate, a special version of |\testpach| could be written, but that is
% not necessary: |\testpach| can do all the work, although much of it will be discarded.
% Here speed is sacrificed for space and this can be afforded because the creation
@@ -505,41 +616,43 @@ LaTeX version 1999/12/01 or later.
%
% The start is copied from |\@mkpream|.
% \begin{macrocode}
-\def\@mksumline#1{\gdef\sumline{}
- \@lastchclass 4 \@firstamptrue
+\def\@mksumline#1{\gdef\sumline{}\@lastchclass 4 \@firstamptrue
% \end{macrocode}
% At first the column number is reset and the actual code for what was called |\a|
% above is made inactive.
% \begin{macrocode}
- \global\FCsc@l=\z@
- \let\prr@sult=\relax
+\global\FCsc@l=0
+\let\prr@sult=\relax
% \end{macrocode}
% Then |\@mkpream| is picked up again.
% \begin{macrocode}
- \@temptokena{#1}
- \@tempswatrue
- \@whilesw\if@tempswa\fi{\@tempswafalse\the\NC@list}%
- \count@\m@ne
- \let\the@toks\relax
- \prepnext@tok
+\@temptokena{#1}
+\@tempswatrue
+\@whilesw\if@tempswa\fi{\@tempswafalse\the\NC@list}%
+\count0\m@ne
+\let\the@toks\relax
+\prepnext@tok
% \end{macrocode}
% Next is the loop over all tokens in the expanded form of the specifier. The change
% with respect to |\@mkpream| is that the body of the loop is now only dealing
-% with classes 0, 2, and 10. What to do in those cases is of course different from
+% with F-classes~0, 2, and 10. What to do in those cases is of course different from
% what to do when constructing the preamble, so special definitions are created, see
% below.
% \begin{macrocode}
- \expandafter \@tfor \expandafter \@nextchar
- \expandafter :\expandafter =\the\@temptokena \do
- {\@testpach
- \ifcase \@chclass \@classfz
+\expandafter \@tfor \expandafter \@nextchar
+ \expandafter :\expandafter =\the\@temptokena \do
+ {\@testpach
+ \ifcase \@chclass \@classfz
\or \or \@classfii \or
\or \or \or \or \or \or \or \@classfx \fi
- \@lastchclass\@chclass}%
+ \@lastchclass\@chclass}%
% \end{macrocode}
-% And the macro is finished by appending the |\\| to the sumline.
+% And the macro is finished by applying the |\aut@check| and appending the |\\| to the sumline.
+% Note that the |\aut@check| is performed {\itshape in\/} the last column, but since it does not
+% put anything in the horizontal list---it only writes to screen and transcript file---, this
+% is harmless.
% \begin{macrocode}
- \xdef\sumline{\sumline\noexpand\\}}
+\xdef\sumline{\sumline\noexpand\aut@check\noexpand\\}}
% \end{macrocode}
% \end{macro}
% \begin{macro}{\@addtosumline}
@@ -550,26 +663,26 @@ LaTeX version 1999/12/01 or later.
% \end{macrocode}
% \end{macro}
% \begin{macro}{\@classfx}
-% Class~10 for the sumline creation is a stripped down version of |\@classx|: add an ampersand
+% Class~f10 for the sumline creation is a stripped down version of |\@classx|: add an ampersand
% unless it is the first. It deals with the specifiers |b|, |m|, |p|, |c|, |l|, and |r|.
% \begin{macrocode}
\def\@classfx{\if@firstamp \@firstampfalse \else \@addtosumline &\fi}
% \end{macrocode}
% \end{macro}
% \begin{macro}{\@classfz}
-% Class~0 is applicable for specifiers |c|, |l|, and |r| and if the arguments of |p|, |m|,
+% Class~f0 is applicable for specifiers |c|, |l|, and |r|, and if the arguments of |p|, |m|,
% or |b| are given. The latter three cases, with |\@chnum| is 0, 1, or~2 should be ignored
-% and the first three cases are now similar to class~10.
+% and the first three cases are now similar to class~f10.
% \begin{macrocode}
\def\@classfz{\ifnum\@chnum<\thr@@ \@classfx\fi}
% \end{macrocode}
% \end{macro}
% \begin{macro}{\@classfii}
-% Here comes the nice and nasty part. Class~2 is applicable if a |<| is specified. This is
+% Here comes the nice and nasty part. Class~f2 is applicable if a~|<| is specified. This is
% tested by checking |\@lastchclass|, which should be equal to~8. Then it is checked that
% the argument to |<| is indeed |\e@fi|. This check is rather clumsy but this was the
% first way, after many attempts, that worked. It is necessary because the usage of |<|
-% is not restricted to |\e@fi|: the user may have specified other \TeX-code using |<|.
+% is not restricted to |\e@fi|: the user may have specified other \LaTeX-code using~|<|.
% \begin{macrocode}
\def\@classfii{\ifnum\@lastchclass=8
\edef\t@stm{\expandafter\string\@nextchar}
@@ -580,26 +693,32 @@ LaTeX version 1999/12/01 or later.
% \begin{macrocode}
\@addtosumline{\prr@sult}
% \end{macrocode}
-% But we're not done yet: in the following lines of code the appropriate \meta{count}s
-% and \meta{dimen}s are created, if necessary.
+% But we're not done yet: in the following lines of code the appropriate \meta{count}s and
+% \meta{dimen}s are created, if necessary. Note that |\FCsc@l| was set to~0 in the beginning
+% of |\@mksumline|, so it is well-defined when |\@classfii| is used.
% \begin{macrocode}
- \global\advance\FCsc@l by \@ne
+ \global\advance\FCsc@l by \@ne
\ifnum\FCsc@l>\FCtc@l
% \end{macrocode}
-% Apparently the number of requested columns is larger than the currently available number,
-% so a new one should be created. What is checked here is its existence: if it exists, it is not
-% created by |fcolumn| so it may be something other than a \meta{count}, in which things go haywire.
-% Therefore it is assumed that only |fcolumn| is allowed to generate a \meta{count} called
-% |\FCtot@<some romannumeral>|.
+% Apparently the number of requested columns is larger than the currently available number of
+% relevant \meta{count}s and \meta{dimen}s, so new ones should be created. What is checked
+% here is merely the existence of |\FCtot@<some romannumeral>|. If it already exists---although
+% it may not even be a \meta{count}; that is not checked---it is not created by |fcolumn| and a
+% warning is given. In case it is a \meta{count} you're just lucky, although any change to this
+% \meta{count} is global anyway, so things will be overwritten. In the case it is not a
+% \meta{count}, things will go haywire and you'll soon find out. The remedy then is to rename
+% your \meta{count} prior to |fcolumn| to avoid this name clash.
% \begin{macrocode}
\expandafter\ifx\csname FCtot@\romannumeral\FCsc@l\endcsname\relax
\expandafter\newcount\csname FCtot@\romannumeral\FCsc@l\endcsname
\else
- \message{\noexpand\FCtot@\romannumeral\FCsc@l already exists.}
+ \message{^^JWarning: FCtot@\romannumeral\FCsc@l \space is already
+ defined and it may not even be a <count>. I'll proceed,
+ but with fingers crossed. }
\fi
% \end{macrocode}
-% And the same is applicable for the \meta{dimen}: it is assumed that only |fcolumn| is allowed to
-% generate a \meta{dimen} called |\FCwd@<some romannumeral>|.
+% And the same is applicable for the \meta{dimen}: in case of a name clash you have to
+% rename your \meta{dimen} prior to |fcolumn|.
% \begin{macrocode}
\expandafter\ifx\csname FCwd@\romannumeral\FCsc@l\endcsname\relax
\expandafter\newdimen\csname FCwd@\romannumeral\FCsc@l\endcsname
@@ -608,65 +727,115 @@ LaTeX version 1999/12/01 or later.
% \begin{macrocode}
\global\FCtc@l=\FCsc@l
\else
- \message{\noexpand\FCwd@\romannumeral\FCsc@l already exists.}
+ \message{^^JWarning: FCwd@\romannumeral\FCsc@l \space is already
+ defined and it may not even be a <dimen>. I'll proceed,
+ but with fingers crossed. }
\fi
\fi
\fi
\fi}
% \end{macrocode}
-% Once created they are not initialised here because that is done later in one go.
+% Once created it is not necessary to initialise them here because that is done later in one go.
% \end{macro}
% \begin{macro}{\leeg}
-% The macro |\prr@sult| (see below) actually puts the information together.
-% It starts with switching off the default behaviour of the f-column by
-% providing the number 0 to comfort the assignment in |\b@fi|, although
-% any number would be good: it
-% is discarded anyway. This is achieved by |\let|ting |\e@fi| to be |\relax|.
-% This |\let| is only local to the current column. Since the user may from time to
-% time also need an empty entry in the table, this definition is without at-sign.
-% \begin{macrocode}
-\def\leeg{0\relax \let\e@fi=\relax}
-% \end{macrocode}
-% A better solution were to check for a number. If present then do the
-% default action, otherwise generate an empty entry.
+% This macro is used to overrule the default behaviour of the pair |\b@fi| and |\e@fi|. It
+% starts with ending the assignment to |\bedr@g| in the same way that |\e@fi| would normally
+% do. Then the effect of |\e@fi| (that is still in the preamble) is annihilated by |\let|ting
+% it to be |\relax|. This |\let| is only local to the current column. Then the argument to
+% |\leeg| is treated in a similar way as |\e@fi| would do with a typeset number.
+%
+% Since the user may from time to time also need a column entry other than a number in the
+% table, e.g., |\leeg{p.m.}|, this definition is without at-sign.
+% \begin{macrocode}
+\def\leeg#1{ 234\relax\egroup \let\e@fi=\relax \setbox0=\hbox{#1}%
+\ifdim\wd0>\csname FCwd@\romannumeral\FCsc@l\endcsname
+ \global\csname FCwd@\romannumeral\FCsc@l\endcsname=\wd0
+\fi\unhbox0}
+% \end{macrocode}
+% Note that anything may be given as argument to |\leeg|, so in principle it can also be used
+% to cheat: |\leeg{0,03}| will insert |0,03| in the table but it doesn't increase the totals
+% of that column by~3 (assuming |3,2| coding for the separations). But you won't cheat, won't you?
+% It may affect the width, so be careful: don't insert the unabridged version of Romeo and Julia here.
% \end{macro}
% \begin{macro}{\prr@sult}
-% And here it is: the end result. First annihilate the effect of |\e@fi|.
+% The macro |\prr@sult| actually puts the information together. It starts like |\leeg|.
% \begin{macrocode}
-\def\prr@sult{\leeg
+\def\prr@sult{ 345\relax\egroup \let\e@fi=\relax
% \end{macrocode}
% Then the information for the last line is computed. It is not sufficient to calculate
% the width of the result (in points) to use that as the width of the rule separating
-% the individual entries and the result. It may be that the sum is larger
-% (in points) than any of the entries, e.g., when the result of $6+6$ is typeset. The
-% width of the rule should be equal to the width of |\hbox{$12$}| then. On the other
-% hand the width of the rule when summing $24$ and $-24$ should be that of |\hbox{$-24$}|
-% (or |\hbox{$(24$}|, see above), not the width of the result |\hbox{$0$}|. Therefore the
+% the individual entries and the result. It may be that the sum is larger (in points) than
+% any of the entries, e.g., when the result of $600+600$ is typeset. The width of the rule
+% should be equal to the width of |\hbox{$12{,}00$}| then (using specifier |3,2|). On the other
+% hand the width of the rule when summing $2400$ and $-2400$ should be that of |\hbox{$-24{,}00$}|
+% (or |\hbox{$(24{,}00$}|, see above), not the width of the result |\hbox{$0{,}00$}|. Therefore the
% maximum of all entry widths, including the result, was calculated.
% \begin{macrocode}
- \setbox0=\hbox{$\geldm@cro{\number\csname
- FCtot@\romannumeral\FCsc@l\endcsname}{\sp@l}$}%
- \ifdim\wd0>\csname FCwd@\romannumeral\FCsc@l\endcsname
- \global\csname FCwd@\romannumeral\FCsc@l\endcsname=\wd0
- \fi
- \vbox{\hbox to \csname
- FCwd@\romannumeral\FCsc@l\endcsname{\hrulefill}\vskip2pt
- \hbox to \csname
- FCwd@\romannumeral\FCsc@l\endcsname{\hfil\unhbox0}}}
+\setbox0=\hbox{$\geldm@cro{\number\csname
+FCtot@\romannumeral\FCsc@l\endcsname}{\sp@l}$}%
+\ifdim\wd0>\csname FCwd@\romannumeral\FCsc@l\endcsname
+ \global\csname FCwd@\romannumeral\FCsc@l\endcsname=\wd0
+\fi
+\vbox{\hrule width \csname FCwd@\romannumeral\FCsc@l\endcsname
+\vskip2pt
+\hbox to \csname FCwd@\romannumeral\FCsc@l\endcsname{\hfil\unhbox0}}}
+% \end{macrocode}
+% \end{macro}
+% \begin{macro}{\aut@check}
+% If the number of F-columns is even, it is assumed that they are part of two sets of columns of
+% which each column of the first set should balance the appropriate column of the second set. If
+% on the other hand the number of columns is odd, then at least one column has nothing to balance
+% against and no checking occurs. It is correct to check for oddness of |\FCsc@l| since this
+% |\aut@check| is only performed in the last column of the tabular: the value of |\FCsc@l| now
+% equals the number of columns used in the current tabular (and may differ from |\FCtc@l|).
+%
+% The output is only to screen and the transcript file; it doesn't change the appearance of your
+% document, so in case the assumption is wrong you can safely ignore the result and go on.
+% The \meta{count}s 0 and~1 are used here and this can be done because any content of those
+% \meta{count}s from previous calculations has become irrelevant at this moment.
+% \begin{macrocode}
+\def\aut@check{\ifodd\FCsc@l\else \count0=\@ne \count1=\FCsc@l
+ \divide\count1 by \tw@ \advance\count1 by \@ne
+ \loop
+ \ifnum\csname FCtot@\romannumeral\count0\endcsname=
+ \csname FCtot@\romannumeral\count1\endcsname\else
+ \message{^^JWarning: F-columns \number\count0 \space
+ and \number\count1 \space do not balance! }%
+ \fi
+ \ifnum\count1=\FCsc@l\else
+ \advance\count0 by\@ne \advance\count1 by\@ne
+ \repeat
+\fi }
+% \end{macrocode}
+% \end{macro}
+% \begin{macro}{\checkfcolumns}
+% But the assumptions for |\aut@check| may be wrong, therefore manual
+% control on this checking is also made possible here. The macro |\checkfcolumns|
+% provides a way to the user to check that the appropriate
+% columns are balanced (as it should in a balance). Arguments |#1| and |#2| are
+% the column numbers to compare. It is the responsibility of the user to provide
+% the correct numbers here, otherwise bogus output is generated.
+% \begin{macrocode}
+\def\checkfcolumns#1#2{\noalign{\ifnum\csname FCtot@\romannumeral#1
+\endcsname=\csname FCtot@\romannumeral#2\endcsname\else
+ \message{^^JWarning: F-columns #1 and #2 do not balance! }%
+\fi}}
% \end{macrocode}
% \end{macro}
% \begin{macro}{\res@tsumline}
% Since all changes to the totals and widths of the columns are global,
% they have to be reset actively at the start of a tabular or array.
% That is an action by itself, but it may occur more often, on request
-% of the user, therefore a special macro is defined. The count
-% |\FCsc@l| can be used as a temporary register because its actual
-% value is not important anymore.
+% of the user, therefore a special macro is defined. A side effect of
+% this macro is that |\FCsc@l| is reset to~0. This is an advantage:
+% it should be zero at the beginning of a line in the table (for other
+% lines this is done by the |\\|).
% \begin{macrocode}
-\def\res@tsumline{\FCsc@l\FCtc@l \loop\ifnum\FCsc@l>0
- \global\csname FCtot@\romannumeral\FCsc@l\endcsname=0
- \global\csname FCwd@\romannumeral\FCsc@l\endcsname=0pt
- \advance\FCsc@l by \m@ne \repeat}
+\def\res@tsumline{\FCsc@l=\FCtc@l\loop\ifnum\FCsc@l>0
+ \global\csname FCtot@\romannumeral\FCsc@l\endcsname=0
+ \global\csname FCwd@\romannumeral\FCsc@l\endcsname=\z@
+ \advance\FCsc@l by \m@ne
+\repeat}
% \end{macrocode}
% \end{macro}
% \begin{macro}{\resetsumline}
@@ -676,19 +845,18 @@ LaTeX version 1999/12/01 or later.
% \end{macrocode}
% \end{macro}
% That's it!
-% \StopEventually{\PrintIndex}
-% \Finale
-%\typeout{****************************************************}
-%\typeout{*}
-%\typeout{* To finish the installation you have to move the}
-%\typeout{* following file into a directory searched by TeX:}
-%\typeout{*}
-%\typeout{*\space\space\space fcolumn.sty}
-%\typeout{*}
-%\typeout{* To produce the documentation run the file}
-%\typeout{* fcolumn.dtx through LaTeX.}
-%\typeout{*}
-%\typeout{* Happy TeXing!}
-%\typeout{*}
-%\typeout{****************************************************}
+%\section*{Acknowledgement}
+% Thanks to Karl Berry for valuable comments regarding the consistency of
+% the installation procedure of this version.
+%\bibliographystyle{plain}
+%\begin{thebibliography}{9}
+%\itemsep0pt
+%\parsep0pt
+%\parskip0pt
+%\bibitem{array} Frank Mittelbach and David Carlisle.
+% A new implementation of \LaTeX's |tabular| and |array| environment.
+%\bibitem{booktabs} Simon Fear. Publication quality tables in \LaTeX.
+%\bibitem{dcolumn} David Carlisle. The |dcolumn| package.
+%\end{thebibliography}
+%\Finale
\endinput
diff --git a/Master/texmf-dist/source/latex/fcolumn/fcolumn.ins b/Master/texmf-dist/source/latex/fcolumn/fcolumn.ins
new file mode 100644
index 00000000000..78380b6db39
--- /dev/null
+++ b/Master/texmf-dist/source/latex/fcolumn/fcolumn.ins
@@ -0,0 +1,55 @@
+%%
+%% Copyright (C) 2015 by Edgar Olthof
+%%
+%% This file may be distributed and/or modified under the
+%% conditions of the LaTeX Project Public License, either
+%% version 1.3 of this license or (at your option) any later
+%% version. The latest version of this license is in
+%%
+%% http://www.latex-project.org/lppl.txt
+%%
+%% and version 1.3 or later is part of all distributions of
+%% LaTeX version 1999/12/01 or later.
+%%
+\input docstrip.tex
+\keepsilent
+\preamble
+
+This is a generated file.
+
+Copyright (C) 2015 by Edgar Olthof
+
+This file may be distributed and/or modified under the
+conditions of the LaTeX Project Public License, either
+version 1.3 of this license or (at your option) any later
+version. The latest version of this license is in
+
+ http://www.latex-project.org/lppl.txt
+
+and version 1.3 or later is part of all distributions of
+LaTeX version 1999/12/01 or later.
+\endpreamble
+\askforoverwritefalse
+\generate{\file{fcolumn.sty}{\from{fcolumn.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{* fcolumn.sty *}
+\Msg{* *}
+\Msg{* To produce the documentation, run the file *}
+\Msg{* fcolumn.dtx through LaTeX. Make sure to update *}
+\Msg{* the glossary and index by makeindex: *}
+\Msg{* *}
+\Msg{* makeindex -s gglo.ist -o fcolumn.gls fcolumn.glo *}
+\Msg{* makeindex -s gglo.ist -o fcolumn.ind fcolumn.idx *}
+\Msg{* *}
+\Msg{* and then LaTeX again; presto! *}
+\Msg{* *}
+\Msg{* Happy TeXing! *}
+\Msg{* *}
+\Msg{******************************************************}
+\endbatchfile
diff --git a/Master/texmf-dist/tex/latex/fcolumn/fcolumn.sty b/Master/texmf-dist/tex/latex/fcolumn/fcolumn.sty
index f0c7225d080..88a33413bce 100644
--- a/Master/texmf-dist/tex/latex/fcolumn/fcolumn.sty
+++ b/Master/texmf-dist/tex/latex/fcolumn/fcolumn.sty
@@ -12,138 +12,132 @@
%%
%% This file may be distributed and/or modified under the
%% conditions of the LaTeX Project Public License, either
-%% version 1.2 of this license or (at your option) any later
+%% version 1.3 of this license or (at your option) any later
%% version. The latest version of this license is in
%%
%% http://www.latex-project.org/lppl.txt
%%
-%% and version 1.2 or later is part of all distributions of
+%% and version 1.3 or later is part of all distributions of
%% LaTeX version 1999/12/01 or later.
\NeedsTeXFormat{LaTeX2e}
\ProvidesPackage{fcolumn}
- [2015/03/07 v1.0 automatic currency package]
+ [2015/09/20 v1.1.1 Automatic Currency Package (EO)]
\RequirePackage{array}
\newcolumntype{F}[3]{>{\b@fi{#1}{#2}{#3}}r<{\e@fi}}
\newcolumntype{f}{F{.}{,}{3,2}}
-\newcount\FCsc@l \FCsc@l0
-\newcount\FCtc@l \FCtc@l0
-\newcount\g@lda
-\newcount\g@ldb
+\newcount\FCsc@l \FCsc@l=0
+\newcount\FCtc@l \FCtc@l=0
\newif\ifwiths@p
\newif\ifstrictaccounting \strictaccountingfalse
\def\geldm@cro#1#2{\withs@pfalse
- \afterassignment\g@ldm@cro\count@#2\relax{#1}}
+\afterassignment\g@ldm@cro\count@#2\relax{#1}}
\def\g@ldm@cro#1\relax#2{%
- \ifnum#2<0 \ifstrictaccounting(\else-\fi\g@lda=-#2 \else\g@lda=#2 \fi
- \ifx\relax#1\relax\g@ldb=2 \else \g@ldb\@gobble#1\relax\fi
+\ifnum#2<0 \ifstrictaccounting(\else-\fi\count2=-#2 \else\count2=#2 \fi
+\count3=\ifx\relax#1\relax2 \else \@gobble#1\relax\fi
+\count4=\count3
+\loop
+ \ifnum\count3>0 \divide\count2 by 10 \advance\count3 by \m@ne
+\repeat
+\g@ldens{\the\count@}%
+\ifnum\count4>0\decim@lmark
+ \ifnum#2>0 \count2=-#2\else\count2=#2 \fi
+ \count3=\count4
\loop
- \ifnum\g@ldb>0 \divide\g@lda by 10 \advance\g@ldb by \m@ne
+ \ifnum\count3>0 \divide\count2 by 10 \advance\count3 by \m@ne
\repeat
- \g@ldens{\the\count@}%
- \ifx\relax#1\relax\g@ldb=2 \else \g@ldb\@gobble#1\relax\fi
- \ifnum\g@ldb>0\decim@lmark
- \ifnum#2>0 \g@lda=-#2\else\g@lda=#2 \fi
- \loop
- \ifnum\g@ldb>0 \divide\g@lda by 10 \advance\g@ldb by \m@ne
- \repeat
- \ifx\relax#1\relax\g@ldb=2 \else \g@ldb\@gobble#1\relax\fi
- \loop
- \ifnum\g@ldb>0 \multiply\g@lda by 10 \advance\g@ldb by \m@ne
- \repeat
- \ifnum#2>0 \advance\g@lda by #2
- \else \advance\g@lda by -#2
- \fi
- \ifx\relax#1\relax\g@ldb=2 \else \g@ldb\@gobble#1\relax\fi
- \zerop@d{\number\g@ldb}{\number\g@lda}%
- \fi
- \ifnum#2<0 \ifstrictaccounting\rlap{)}\fi\fi}
-\def\g@ldens#1{{\g@ldb=\g@lda \count0=#1
- \loop
- \ifnum\count0>0
- \divide\g@lda by 10
- \advance\count0 by \m@ne
- \repeat
- \ifnum\g@lda>0 \g@ldens{#1}\fi
- \count0=#1
\loop
- \ifnum\count0>0
- \multiply\g@lda by 10
- \advance\count0 by \m@ne
+ \ifnum\count3<\count4 \multiply\count2 by 10 \advance\count3 by \@ne
\repeat
- \g@lda=-\g@lda
- \advance\g@lda by \g@ldb \du@zendprint{#1}}}
-\def\du@zendprint#1{\ifwiths@p\sep@rator\zerop@d{#1}{\number\g@lda}%
- \else\zerop@d{1}{\number\g@lda}\fi
- \global\withs@ptrue}
+ \ifnum#2>0 \advance\count2 by #2
+ \else \advance\count2 by -#2
+ \fi
+ \zerop@d{\number\count3}{\number\count2}%
+\fi
+\ifnum#2<0 \ifstrictaccounting\rlap{)}\fi\fi}
+\def\g@ldens#1{{\count3=\count2 \count0=#1
+\loop
+ \ifnum\count0>0
+ \divide\count2 by 10
+ \advance\count0 by \m@ne
+\repeat
+\ifnum\count2>0 \g@ldens{#1}\fi
+\count0=#1
+\loop
+ \ifnum\count0>0
+ \multiply\count2 by 10
+ \advance\count0 by\m@ne
+\repeat
+\count2=-\count2
+\advance\count2 by \count3 \du@zendprint{#1}}}
+\def\du@zendprint#1{\ifwiths@p\sep@rator\zerop@d{#1}{\number\count2}%
+ \else\zerop@d{1}{\number\count2}\fi \global\withs@ptrue}
\def\zerop@d#1#2{{\count0=1 \count1=#2
- \loop
- \divide \count1 by 10
- \ifnum\count1>0 \advance\count0 by 1
- \repeat
- \loop
- \ifnum\count0<#1 0\advance\count0 by 1
- \repeat
- \number#2}}
+\loop
+ \divide \count1 by 10
+ \ifnum\count1>0 \advance\count0 by\@ne
+\repeat
+\loop
+ \ifnum\count0<#1\relax 0\advance\count0 by\@ne
+\repeat
+\number#2}}
\def\zetg@ld#1#2{\global\advance\csname
- FCtot@\romannumeral\FCsc@l\endcsname by #1
- \setbox0=\hbox{$\geldm@cro{#1}{#2}$}%
- \ifdim\wd0>\csname FCwd@\romannumeral\FCsc@l\endcsname
- \global\csname FCwd@\romannumeral\FCsc@l\endcsname=\wd0
- \fi\unhbox0}
-\def\check#1#2{\noalign{\ifnum\csname FCtot@\romannumeral#1\endcsname=
- \csname FCtot@\romannumeral#2\endcsname\else
- \message{Columns #1 and #2 do not balance!}\fi}}
+FCtot@\romannumeral\FCsc@l\endcsname by #1
+\setbox0=\hbox{$\geldm@cro{#1}{#2}$}%
+\ifdim\wd0>\csname FCwd@\romannumeral\FCsc@l\endcsname
+ \global\csname FCwd@\romannumeral\FCsc@l\endcsname=\wd0
+\fi\unhbox0}
\newcount\bedr@g
\def\b@fi#1#2#3{\def\sep@rator{{#1}}\def\decim@lmark{{#2}}%
- \let\ignorespaces=\@empty \def\sp@l{#3}\global\advance\FCsc@l by \@ne
- \bedr@g=}
-\def\e@fi{\relax \zetg@ld{\number\bedr@g}{\sp@l}}
-\def\@xarraycr{\@ifnextchar [%
- \@argarraycr {\ifnum 0=`{}\fi \cr\noalign{\global\FCsc@l=0 }}}
+\def\sp@l{#3}\let\ignorespaces=\@empty \let\unskip=\@empty
+\global\advance\FCsc@l by \@ne
+\setbox0=\hbox\bgroup\global\bedr@g=}
+\def\e@fi{ 123\relax\egroup\ifdim\wd0>\z@
+ \zetg@ld{\number\bedr@g}{\sp@l}%
+\fi}
\def\@array[#1]#2{%
- \@tempdima \ht \strutbox
- \advance \@tempdima by\extrarowheight
- \setbox \@arstrutbox \hbox{\vrule
- \@height \arraystretch \@tempdima
- \@depth \arraystretch \dp \strutbox
- \@width \z@}%
- \begingroup
- \@mkpream{#2}%
- \xdef\@preamble{\noexpand \ialign \@halignto
- \bgroup \@arstrut \@preamble
- \tabskip \z@ \cr}%
- \endgroup
- \begingroup
- \@mksumline{#2}%
- \endgroup
- \res@tsumline
- \@arrayleft
- \if #1t\vtop \else \if#1b\vbox \else \vcenter \fi \fi
- \bgroup
- \let \@sharp ##\let \protect \relax
- \lineskip \z@
- \baselineskip \z@
- \m@th
- \let\\\@arraycr \let\tabularnewline\\\let\par\@empty \@preamble}
+\@tempdima \ht \strutbox
+\advance \@tempdima by\extrarowheight
+\setbox \@arstrutbox \hbox{\vrule
+ \@height \arraystretch \@tempdima
+ \@depth \arraystretch \dp \strutbox
+ \@width \z@}%
+\def\ialign{\everycr{\noalign{\global\FCsc@l=0 }}%
+ \tabskip\z@skip\halign}
+\begingroup
+\@mkpream{#2}%
+\xdef\@preamble{\noexpand \ialign \@halignto
+\bgroup\@arstrut\@preamble\tabskip\z@\cr}%
+\endgroup
+\begingroup
+\@mksumline{#2}%
+\endgroup
+\res@tsumline
+\@arrayleft
+\if #1t\vtop \else \if#1b\vbox \else \vcenter \fi \fi
+\bgroup
+\let \@sharp ##\let \protect \relax
+\lineskip \z@
+\baselineskip \z@
+\m@th
+\let\\\@arraycr \let\tabularnewline\\\let\par\@empty \@preamble}
\let\@@array=\@array
-\def\@mksumline#1{\gdef\sumline{}
- \@lastchclass 4 \@firstamptrue
- \global\FCsc@l=\z@
- \let\prr@sult=\relax
- \@temptokena{#1}
- \@tempswatrue
- \@whilesw\if@tempswa\fi{\@tempswafalse\the\NC@list}%
- \count@\m@ne
- \let\the@toks\relax
- \prepnext@tok
- \expandafter \@tfor \expandafter \@nextchar
- \expandafter :\expandafter =\the\@temptokena \do
- {\@testpach
- \ifcase \@chclass \@classfz
+\def\@mksumline#1{\gdef\sumline{}\@lastchclass 4 \@firstamptrue
+\global\FCsc@l=0
+\let\prr@sult=\relax
+\@temptokena{#1}
+\@tempswatrue
+\@whilesw\if@tempswa\fi{\@tempswafalse\the\NC@list}%
+\count0\m@ne
+\let\the@toks\relax
+\prepnext@tok
+\expandafter \@tfor \expandafter \@nextchar
+ \expandafter :\expandafter =\the\@temptokena \do
+ {\@testpach
+ \ifcase \@chclass \@classfz
\or \or \@classfii \or
\or \or \or \or \or \or \or \@classfx \fi
- \@lastchclass\@chclass}%
- \xdef\sumline{\sumline\noexpand\\}}
+ \@lastchclass\@chclass}%
+\xdef\sumline{\sumline\noexpand\aut@check\noexpand\\}}
\def\@addtosumline#1{\xdef\sumline{\sumline #1}}
\def\@classfx{\if@firstamp \@firstampfalse \else \@addtosumline &\fi}
\def\@classfz{\ifnum\@chnum<\thr@@ \@classfx\fi}
@@ -157,32 +151,55 @@
\expandafter\ifx\csname FCtot@\romannumeral\FCsc@l\endcsname\relax
\expandafter\newcount\csname FCtot@\romannumeral\FCsc@l\endcsname
\else
- \message{\noexpand\FCtot@\romannumeral\FCsc@l already exists.}
+ \message{^^JWarning: FCtot@\romannumeral\FCsc@l \space is already
+ defined and it may not even be a <count>. I'll proceed,
+ but with fingers crossed. }
\fi
\expandafter\ifx\csname FCwd@\romannumeral\FCsc@l\endcsname\relax
\expandafter\newdimen\csname FCwd@\romannumeral\FCsc@l\endcsname
\global\FCtc@l=\FCsc@l
\else
- \message{\noexpand\FCwd@\romannumeral\FCsc@l already exists.}
+ \message{^^JWarning: FCwd@\romannumeral\FCsc@l \space is already
+ defined and it may not even be a <dimen>. I'll proceed,
+ but with fingers crossed. }
\fi
\fi
\fi
\fi}
-\def\leeg{0\relax \let\e@fi=\relax}
-\def\prr@sult{\leeg
- \setbox0=\hbox{$\geldm@cro{\number\csname
- FCtot@\romannumeral\FCsc@l\endcsname}{\sp@l}$}%
- \ifdim\wd0>\csname FCwd@\romannumeral\FCsc@l\endcsname
- \global\csname FCwd@\romannumeral\FCsc@l\endcsname=\wd0
- \fi
- \vbox{\hbox to \csname
- FCwd@\romannumeral\FCsc@l\endcsname{\hrulefill}\vskip2pt
- \hbox to \csname
- FCwd@\romannumeral\FCsc@l\endcsname{\hfil\unhbox0}}}
-\def\res@tsumline{\FCsc@l\FCtc@l \loop\ifnum\FCsc@l>0
- \global\csname FCtot@\romannumeral\FCsc@l\endcsname=0
- \global\csname FCwd@\romannumeral\FCsc@l\endcsname=0pt
- \advance\FCsc@l by \m@ne \repeat}
+\def\leeg#1{ 234\relax\egroup \let\e@fi=\relax \setbox0=\hbox{#1}%
+\ifdim\wd0>\csname FCwd@\romannumeral\FCsc@l\endcsname
+ \global\csname FCwd@\romannumeral\FCsc@l\endcsname=\wd0
+\fi\unhbox0}
+\def\prr@sult{ 345\relax\egroup \let\e@fi=\relax
+\setbox0=\hbox{$\geldm@cro{\number\csname
+FCtot@\romannumeral\FCsc@l\endcsname}{\sp@l}$}%
+\ifdim\wd0>\csname FCwd@\romannumeral\FCsc@l\endcsname
+ \global\csname FCwd@\romannumeral\FCsc@l\endcsname=\wd0
+\fi
+\vbox{\hrule width \csname FCwd@\romannumeral\FCsc@l\endcsname
+\vskip2pt
+\hbox to \csname FCwd@\romannumeral\FCsc@l\endcsname{\hfil\unhbox0}}}
+\def\aut@check{\ifodd\FCsc@l\else \count0=\@ne \count1=\FCsc@l
+ \divide\count1 by \tw@ \advance\count1 by \@ne
+ \loop
+ \ifnum\csname FCtot@\romannumeral\count0\endcsname=
+ \csname FCtot@\romannumeral\count1\endcsname\else
+ \message{^^JWarning: F-columns \number\count0 \space
+ and \number\count1 \space do not balance! }%
+ \fi
+ \ifnum\count1=\FCsc@l\else
+ \advance\count0 by\@ne \advance\count1 by\@ne
+ \repeat
+\fi }
+\def\checkfcolumns#1#2{\noalign{\ifnum\csname FCtot@\romannumeral#1
+\endcsname=\csname FCtot@\romannumeral#2\endcsname\else
+ \message{^^JWarning: F-columns #1 and #2 do not balance! }%
+\fi}}
+\def\res@tsumline{\FCsc@l=\FCtc@l\loop\ifnum\FCsc@l>0
+ \global\csname FCtot@\romannumeral\FCsc@l\endcsname=0
+ \global\csname FCwd@\romannumeral\FCsc@l\endcsname=\z@
+ \advance\FCsc@l by \m@ne
+\repeat}
\def\resetsumline{\noalign{\res@tsumline}}
\endinput
%%