summaryrefslogtreecommitdiff
path: root/macros
diff options
context:
space:
mode:
authorNorbert Preining <norbert@preining.info>2023-07-26 03:00:50 +0000
committerNorbert Preining <norbert@preining.info>2023-07-26 03:00:50 +0000
commitcd61ef4a2d1994eb5729fcec4bfcf1fb1682f21d (patch)
tree9398dfb1919cec07283c0a559edf17de12e61924 /macros
parentbe53973de38a7aa9a3a0a8c70fd92bdf893c1a35 (diff)
CTAN sync 202307260300
Diffstat (limited to 'macros')
-rw-r--r--macros/latex/contrib/fcolumn/README10
-rw-r--r--macros/latex/contrib/fcolumn/fcolumn.dtx2075
-rw-r--r--macros/latex/contrib/fcolumn/fcolumn.ins16
-rw-r--r--macros/latex/contrib/fcolumn/fcolumn.pdfbin319237 -> 517071 bytes
-rw-r--r--macros/latex/contrib/fcolumn/makefile9
-rw-r--r--macros/latex/contrib/siunitx/CHANGELOG.md10
-rw-r--r--macros/latex/contrib/siunitx/siunitx-code.pdfbin642405 -> 643576 bytes
-rw-r--r--macros/latex/contrib/siunitx/siunitx-compound.dtx7
-rw-r--r--macros/latex/contrib/siunitx/siunitx-number.dtx39
-rw-r--r--macros/latex/contrib/siunitx/siunitx-print.dtx32
-rw-r--r--macros/latex/contrib/siunitx/siunitx.dtx2
-rw-r--r--macros/latex/contrib/siunitx/siunitx.pdfbin691136 -> 691136 bytes
12 files changed, 1130 insertions, 1070 deletions
diff --git a/macros/latex/contrib/fcolumn/README b/macros/latex/contrib/fcolumn/README
index 2e9980d6a2..a195ad4949 100644
--- a/macros/latex/contrib/fcolumn/README
+++ b/macros/latex/contrib/fcolumn/README
@@ -16,7 +16,7 @@ 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}{}}|.
+rather code |\newcolumntype{f}{F,.{3,2}{}}|.
To show where and how the f-column is used, let's look at a typical
financial table.
@@ -40,7 +40,7 @@ cash & 145,85 & profit & 27542,62 \\
\end{tabular}
The column specifier ``f'' (for ``finance'') is able to do the
-typesetting. It constructs the sum line, typesets the numbers,
+typesetting. It constructs the sumline, typesets the numbers,
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
@@ -53,9 +53,9 @@ fcolumn.pdf.
To install:
-----------
-Run the file fcolumn.ins through (La)TeX (in a directory also
-containing fcolumn.dtx) or type ``make install'' and move the
-resultant file fcolumn.sty to a place searched by TeX.
+Type ``make install'' or run the file fcolumn.ins through
+(La)TeX (in a directory also containing fcolumn.dtx) and move
+the resultant file fcolumn.sty to a place searched by TeX.
To generate the documentation:
------------------------------
diff --git a/macros/latex/contrib/fcolumn/fcolumn.dtx b/macros/latex/contrib/fcolumn/fcolumn.dtx
index 0aef28889c..53dbfcbabe 100644
--- a/macros/latex/contrib/fcolumn/fcolumn.dtx
+++ b/macros/latex/contrib/fcolumn/fcolumn.dtx
@@ -1,5 +1,5 @@
% \iffalse meta-comment
-% Copyright (C) 2015--2022 by Edgar Olthof
+% Copyright (C) 2015--2023 by Edgar Olthof
%
% This file may be distributed and/or modified under the
% conditions of the LaTeX Project Public License, either
@@ -14,7 +14,7 @@
% \fi
% \iffalse
%<package>\ProvidesPackage{fcolumn}
-%<package> [2022/02/01 v1.4.1 Automatic Currency Package (EHTO)]
+%<package> [2023/07/25 v1.4.2 Automatic Currency Package (EHTO)]
%<package>\NeedsTeXFormat{LaTeX2e}
%<package>\RequirePackage{array}[v2.4k]
%<*driver>
@@ -24,245 +24,278 @@
\usepackage{fcolumn}
\usepackage{booktabs}
\usepackage[tableposition=top,aboveskip=0pt]{caption}
-\DeclareCaptionStyle{normal}{labelfont={rm,bf,small},textfont={rm,small},%
+\DeclareCaptionStyle{normal}{labelfont={rm,bf,small},textfont={rm,small},
indention=0pt,labelsep=colon,justification=raggedright}
\captionsetup[table]{style=normal}
\captionsetup{singlelinecheck=false}
\usepackage{paralist}
\usepackage{url}
-\usepackage[hyperfootnotes=false,hidelinks]{hyperref}
\usepackage{eurosym}
\usepackage{color}
+\usepackage[hang,symbol*]{footmisc}
+\usepackage[hidelinks]{hyperref}
\EnableCrossrefs
\CodelineIndex
\begin{document}
\DocInput{fcolumn.dtx}
\end{document}
%</driver>
-% \fi
-% \RecordChanges
-% \CheckSum{1200}
-% \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 \~}
+%\fi
+%\RecordChanges
+%\CheckSum{1224}
+%\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 \~}
%
-% \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.}
-% \changes{v1.1.2}{2015/09/29}{Some inconsistencies between explanatory text and actual
-% code removed.}
-% \changes{v1.2}{2019/03/06}{Input parsing changed after comment from Frank
-% Mittelbach. He (Frank) also gave various suggestions for improving robustness
-% or user friendlyness of this package. This version is only backwards compatible
-% when zero decimal digits were and are specified as modifier.}
-% \changes{v1.3}{2021/01/03}{Christian Hoff requested the possibility to provide extra formatting
-% information to a column, e.g., colouring. That was implemented by having an extra parameter
-% to the generic F-column. The current solution is not very robust, as font and/or size change
-% in math environment are very tricky, but providing colour information works, see the example
-% in the main text. This version is now compatible with package longtable. Version~1.3 is
-% backwards compatible to~1.2: it only adds functionality.}
-% \changes{v1.4}{2021/11/21}{More robust version on the extra formatting information: alternative
-% formatting after the comma (if any), instead of additional information. Corrected a few typos.}
-% \changes{v1.4.1}{2022/01/31}{Definition needed for longtable only if that package is loaded.
-% General update, streamlining code. Excess digits silently ignored when all zero.}
-% \GetFileInfo{fcolumn.sty}
-% \DoNotIndex{\@@endpbox,\@@startpbox,\@acol,\@arstrut,\@arstrutbox,\@classiv,\@classz,\@depth}
-% \DoNotIndex{\@firstampfalse,\@firstamptrue,\@gobble,\@halignto,\@height,\@mkpream,\@ne,\@nextchar}
-% \DoNotIndex{\@preamble,\@sharp,\@tempdima,\@tempswafalse,\@tempswatrue,\@\temptokena,\@tfor}
-% \DoNotIndex{\@whilesw,\@width,\advance,\arraystretch,\baselineskip,\begingroup,\bgroup,\count}
-% \DoNotIndex{\count@,\cr,\csname,\def,\divide,\do,\dp,\edef,\egroup,\hfil,\hline,\ht,\if,\if@firstamp}
-% \DoNotIndex{\if@tempswa,\ifcase,\kill,\lineskip,\LT@@hl,\LT@@tabarray,\LT@bchunk,\LT@caption,\LT@endpbox}
-% \DoNotIndex{\LT@hline,\LT@kill,\LT@LL@FM@cr,\LT@make@row,\LT@mcol,\LT@no@pgbk,\LT@nofcols,\LT@rows}
-% \DoNotIndex{\LT@setprevdepth,\LT@startpbox,\LT@tabularcr,\LTleft,\LTright,\loop,\m@ne,\m@th,\message}
-% \DoNotIndex{\multiply,\or,\par,\prepnext@tok,\protect,\refstepcounter,\repeat,\setbox,\space,\string}
-% \DoNotIndex{\strutbox,\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\\
-% \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 and checking
-% such columns, including the sum line (preceded by a rule of the correct
-% width), using the specifier |f|.
-% \end{abstract}
+%\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.}
+%\changes{v1.1.2}{2015/09/29}{Some inconsistencies between explanatory text and actual
+% code removed.}
+%\changes{v1.2}{2019/03/06}{Input parsing changed after comment from Frank
+% Mittelbach. He (Frank) also gave various suggestions for improving robustness
+% or user friendliness of this package. This version is only backwards compatible
+% when zero decimal digits were and are specified as modifier.}
+%\changes{v1.3}{2021/01/03}{Christian Hoff requested the possibility to provide extra formatting
+% information to a column, e.g., colouring. That was implemented by having an extra parameter
+% to the generic F-column. The current solution is not very robust, as font and/or size change
+% in math environment are very tricky, but providing colour information works, see the example
+% in the main text. This version is now compatible with package longtable and backwards
+% compatible to fcolumn v1.2: it only adds functionality.}
+%\changes{v1.4}{2021/11/21}{More robust version on the extra formatting information: alternative
+% formatting after the comma (if any), instead of additional information. Corrected a few typos.}
+%\changes{v1.4.1}{2022/02/01}{Definition needed for longtable only if that package is loaded.
+% General update, streamlining code. Excess digits silently ignored when all zero.}
+%\changes{v1.4.2}{2023/07/25}{Better documentation. Removed redundant endgroup/begingroup
+% pairs and other small improvements. Unbalanced columns are reported by stating their
+% difference. Implementation of option ``red''. No changes in user experience.}
+%\GetFileInfo{fcolumn.sty}
+%\DoNotIndex{\@@endpbox,\@@startpbox,\@acol,\@arstrut,\@arstrutbox,\@classiv,\@classz,\@depth}
+%\DoNotIndex{\@endpbox,\@firstampfalse,\@firstamptrue,\@gobble,\@halignto,\@height,\@mkpream}
+%\DoNotIndex{\@ne,\@nextchar,\@preamble,\@sharp,\@startpbox,\@tabacol,\@tabarray,\@tabclassiv}
+%\DoNotIndex{\@tabclassz,\@tempdima,\@tempswafalse,\@tempswatrue,\@\temptokena,\@tfor,\@whilesw}
+%\DoNotIndex{\@width,\advance,\arraystretch,\baselineskip,\bgroup,\break,\count,\count@,\cr}
+%\DoNotIndex{\def,\divide,\do,\dp,\edef,\else,\egroup,\fi,\global,\hfil,\hline,\ht,\hsize}
+%\DoNotIndex{\if,\if@firstamp,\if@tempswa,\ifcase,\ifdim,\ifnum,\ifodd,\ifx,\kill,\lineskip}
+%\DoNotIndex{\LT@@hl,\LT@@tabarray,\LT@bchunk,\LT@caption,\LT@endpbox,\LT@hline,\LT@kill}
+%\DoNotIndex{\LT@LL@FM@cr,\LT@make@row,\LT@mcol,\LT@no@pgbk,\LT@nofcols,\LT@rows,\LT@setprevdepth}
+%\DoNotIndex{\LT@startpbox,\LT@tabularcr,\LTleft,\LTright,\let,\loop,\m@ne,\m@th,\message}
+%\DoNotIndex{\multiply,\newpage,\nopagebreak,\or,\pagebreak,\par,\prepnext@tok,\protect}
+%\DoNotIndex{\refstepcounter,\relax,\repeat,\rlap,\setbox,\space,\stepcounter,\string}
+%\DoNotIndex{\strutbox,\tabcolsep,\tabskip,\tabularnewline,\the,\the@toks,\thr@@,\tw@}
+%\DoNotIndex{\unhbox,\vbox,\vcenter,\vrule,\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\\
+%\texttt{edgar <dot> olthof <at> inter <dot> nl <dot> net}}
+%\date{Printed \today}
+%\maketitle
+%\begin{abstract}
+%\noindent 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 and checking
+%such columns, including the sum line (preceded by a rule of the correct
+%width), using the specifier |f|.
+%\end{abstract}
%
-% \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 four arguments: 1)~grouping
-% character of the integer part on output, 2)~decimal mark used on output,
-% 3)~compact additional information on input/output characteristics, and
-% 4)~anything, but primarily used for providing formatting information, see below.
+%\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 four arguments: 1)~grouping
+%character of the integer part on output, 2)~decimal mark used on output,
+%3)~compact additional information on input/output characteristics, and
+%4)~anything, but primarily used for providing formatting information, see below.
%
-% 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 12345,67 will be typeset as $12{.}345{,}67$.
-% People in the Anglo-Saxon world would rather code
-% |\newcolumntype{f}{F{,}{.}{3,2}{}}| for the same input, yielding
-% $12{,}345{.}67$ as output for the number given above. The default
-% value for |#3| is |3,2|, indicating that grouping of the integer
-% part is by three digits, that a comma is used in the \TeX-source to
-% indicate the decimal mark, and that the decimal part consists of
-% two digits. However, if in your country or company grouping is done
-% with a thinspace every four digits, that the decimal mark in the
-% source should be the character |p|, and there are three digits after
-% the decimal mark---that happens to be a |\cdot|---, then simply specify
-% |\newcolumntype{f}{F{\,}{\cdot}{4p3}{}}| in that case. The input could
-% be 123456p78 then, yielding $12{\,}3456{\cdot}780$ as output.
+%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 12345,67 will be typeset as $12{.}345{,}67$.
+%People in the Anglo-Saxon world would rather code
+%|\newcolumntype{f}{F,.{3,2}{}}| for the same input, yielding
+%$12{,}345{.}67$ as output for the number given above. The default
+%value for |#3| is |3,2|, indicating that grouping of the integer
+%part is by three digits, that a comma is used in the \TeX-source to
+%indicate the decimal mark, and that the decimal part consists of
+%two digits. However, if in your country or company grouping is done
+%with a thinspace every four digits, that the decimal mark in the
+%source should be the character |p|, and there are three digits after
+%the decimal mark---that happens to be a |\cdot|---, then simply specify
+%|\newcolumntype{f}{F{\,}{\cdot}{4p3}{}}| in that case. The input could
+%be 123456p78 then, yielding $12{\,}3456{\cdot}780$ as output.
%
-% 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|. If you want no grouping
-% character, specify |0,x|.
+%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|. If you want no grouping
+%character, specify~|0,x|.
%
-% As the fourth parameter you can insert anything just before the typesetting
-% of an amount in a column takes place. Its purpose is to add additional
-% formatting information, e.g., |\color{red}| to have the contents of a column
-% coloured red, but it can be misused, so use with care. And it can't do all!
+%As the fourth parameter you can insert anything just before the typesetting
+%of an amount in a column takes place. Its purpose is to add additional
+%formatting information, e.g., |\color{blue}| to have the contents of a column
+%coloured blue, but it can be misused, so use with care. And it can't do all!
%
-% 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}\\
-% \toprule
-% properties & \leeg{31 dec 2014} & debts & \leeg{31 dec 2014}\\
-% \midrule
-% house & 200000 & equity capital & 50000 \\
-% bank account & -603,23 & mortgage & 150000 \\
-% savings & 28000 \\
-% cash & 145,85 & profit & 27542,62 \\
-% \sumline
-% \bottomrule
-% \end{tabular}
-% \end{table}
-% and how this is entered in \LaTeX\ (Table~\ref{tab:ex2}).
-% \begin{table}[htb]
-% \caption{Verbatim version of Example Table~\ref{tab:ex1}.}
-% \label{tab:ex2}
-% \begin{verbatim}
+%\floatsep=5pt plus 2pt minus 3pt
+%\textfloatsep=5pt plus 2pt minus 3pt
+%\renewcommand{\bottomfraction}{0.9}
+%\renewcommand{\textfraction}{0.1}
%\begin{table}[htb]
-%\caption{Example Table.}
-%\label{tab:ex1}
+%\caption{Example Table.} \label{tab:ex1}
%\begin{tabular}{@{}lflf@{}}
-%\multicolumn4c{\bfseries Balance sheet}\\
+%\multicolumn4{@{}c@{}}{\bfseries Balance sheet}\\
%\toprule
%properties & \leeg{31 dec 2014} & debts & \leeg{31 dec 2014}\\
%\midrule
-%house & 200000 & equity capital & 50000 \\
-%bank account & -603,23 & mortgage & 150000 \\
-%savings & 28000 \\
-%cash & 145,85 & profit & 27542,62 \\
+%house & 200000 & equity capital & 50000 \\
+%bank account & -603,23 & mortgage & 150000 \\
+%savings & 28000 \\
+%cash & 145,85 & profit & 27542,62 \\
+%\sumline
+%\bottomrule
+%\end{tabular}
+%\end{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}
+%and how this is entered in \LaTeX\ (Table~\ref{tab:ex2}).
+%\begin{table}[htb]
+%\caption{Verbatim version of Example Table~\ref{tab:ex1}.}
+%\label{tab:ex2} \topsep0pt
+%\begin{verbatim}
+%\begin{table}[htb]
+%\caption{Example Table.} \label{tab:ex1}
+%\begin{tabular}{@{}lflf@{}}
+%\multicolumn4{@{}c@{}}{\bfseries Balance sheet}\\
+%\toprule
+%properties & \leeg{31 dec 2014} & debts & \leeg{31 dec 2014}\\
+%\midrule
+%house & 200000 & equity capital & 50000 \\
+%bank account & -603,23 & mortgage & 150000 \\
+%savings & 28000 \\
+%cash & 145,85 & profit & 27542,62 \\
%\sumline
%\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, determines the widths of the sumrules, and checks
-% whether the two columns are in balance; if not, the user is warned via a
-% |\PackageWarning|. Of course for nice settings the |booktabs|
-% package~\cite{booktabs} was used, but that is not the point here.
+%\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, determines the widths of the sumrules, and checks
+%whether the two columns are in balance; if not, the user is warned via a
+%|\PackageWarning|. 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~\cite{dcolumn}; some constructions are more or less
-% copied from that package. From version~1.3 onwards it incorporates the idea
-% of Christian Hoff of providing additional (formatting) information per column.
-% A rather contrived example is given in Table~\ref{tab:ex3}, combining colour and fonts.
-% \newcolumntype{q}[1]{F{.}{,}{3,2}{#1}}
-% \begin{table}[htb]
-% \caption{Example Table with column formatting.}
-% \label{tab:ex3}
-% \begin{tabular}{@{}lq{\color{red}\mathsf,\mathbf}lq{\color{green}}@{}}
-% \multicolumn4c{\bfseries Balance sheet}\\
-% \toprule
-% properties & \leeg{31 dec 2014} & debts & \leeg{31 dec 2014}\\
-% \midrule
-% house & 200000 & equity capital & 50000 \\
-% bank account & -603,23 & mortgage & 150000 \\
-% savings & 28000 \\
-% cash & 145,85 & profit & 27542,62 \\
-% \sumline
-% \bottomrule
-% \end{tabular}
-% \end{table}
-% How this is entered in \LaTeX\ is shown in Table~\ref{tab:ex4}.
-% \begin{table}[htb]
-% \caption{Almost verbatim version of Example Table~\ref{tab:ex3}.}
-% \label{tab:ex4}
-% \begin{verbatim}
-%\newcolumntype{q}[1]{F{.}{,}{3,2}{#1}}
+%This package is heavily inspired by the |dcolumn| package by
+%David Carlisle~\cite{dcolumn}; some constructions are more or less
+%copied from that package. From version~1.3 onwards it incorporates the idea
+%of Christian Hoff of providing additional (formatting) information per column.
+%A rather contrived example is given in Table~\ref{tab:ex3}, combining colour and fonts.
+%\newcolumntype{q}[1]{F.,{3,2}{#1}}
+%\begin{table}[htb]
+%\caption{Example Table with column formatting.} \label{tab:ex3}
+%\begin{tabular}{@{}lq{\color{blue}\mathsf,\mathbf}lq{\color{magenta}}@{}}
+%\multicolumn4{@{}c@{}}{\bfseries Balance sheet}\\
+%\toprule
+%properties & \leeg{31 dec 2014} & debts & \leeg{31 dec 2014}\\
+%\midrule
+%house & 200000 & equity capital & 50000 \\
+%bank account & -603,23 & mortgage & 150000 \\
+%savings & 28000 \\
+%cash & 145,85 & profit & 27542,62 \\
+%\sumline
+%\bottomrule
+%\end{tabular}
+%\end{table}
+%How this is entered in \LaTeX\ is shown in Table~\ref{tab:ex4}. The font changing commands
+%like |\mathsf| and |\mathbf| act on an argument, hence require braces,
+%but these are already provided internally for this purpose. For that reason this
+%type of commands must be given last, without braces (and if you don't specify a font
+%changing command, these extra internal braces are just redundant). The fourth argument to the
+%\begin{table}[htb]
+%\caption{Almost verbatim version of Example Table~\ref{tab:ex3}.}
+%\label{tab:ex4} \topsep0pt
+%\begin{verbatim}
+%\newcolumntype{q}[1]{F.,{3,2}{#1}}
%\begin{table}[htb]
-%\caption{Example Table with column formatting.}
-%\label{tab:ex3}
-%\begin{tabular}{@{}lq{\color{red}\mathsf,\mathbf}lq{\color{green}}@{}}
-%\multicolumn4c{\bfseries Balance sheet}\\
+%\caption{Example Table with column formatting.} \label{tab:ex3}
+%\begin{tabular}{@{}lq{\color{blue}\mathsf,\mathbf}lq{\color{magenta}}@{}}
+%\multicolumn4{@{}c@{}}{\bfseries Balance sheet}\\
%...
%... (Same financial contents as in Table 1.)
%...
%\end{tabular}
%\end{table}
-% \end{verbatim}
-% \end{table}
-% The font changing commands like |\mathsf| and |\mathbf| act on an argument, hence require
-% braces, but these are already provided internally for this purpose. For that reason this
-% type of commands must be given last, without braces (and if you don't specify a font
-% changing command, these extra internal braces are just redundant). The fourth argument to the
-% new columntype may consist of two parts, separated by a comma. In that case, the part to
-% the left of the comma is applied to the data entered by the user and the right part
-% is applied to the result. The example in Table~\ref{tab:ex3} shows this: the bold font
-% is only used in the |\sumline| and there is no colour specification, so that's back to
-% the default (black). If you want formatting for the whole column, like green colour as
-% in the last column of Table~\ref{tab:ex3}, leave out the comma.
+%\end{verbatim}
+%\end{table}
+%new columntype may consist of two parts, separated by a comma. In that case, the part to
+%the left of the comma is applied to the data entered by the user and the right part, if
+%non-empty, is the replacement formatting for the result. The example in Table~\ref{tab:ex3}
+%shows this: the bold font is only used in the |\sumline| and there is no colour specification,
+%so that's back to the default (black). If you want formatting for the whole column, like magenta
+%in the case of the last column of Table~\ref{tab:ex3}, leave out the comma.
+%Font size changes, e.g., ``|\huge|'' in ``|\huge\mathbf|''
+%as parameter to column type |q| are ignored by \LaTeX, since the formatting
+%information is used in math environment, which has its own way of handling this.
+%This isn't bad, as size changes in one column, without overall changes to the
+%table look terrible. If you want something huge, make a |\huge| table.
%
-% Note that changes in font size, e.g., ``|\huge|'' in ``|\huge\color{red}\mathbf|''
-% as parameter to column type |q| are ignored by \LaTeX, since the formatting
-% information is used in math environment, which has its own way of handling this.
-% This isn't bad, as size changes in one column, without overall changes to the
-% table look terrible. If you want something huge, make a |\huge| table.
-%
-% This package now also works with |longtable|~\cite{longtable}, provided
-% |longtable| is loaded before |fcolumn|; it checks for that. For a change,
-% the raw formatting of the multipage table is shown first, in Table~\ref{tab:ex5}.
-% For more information on how to handle |\endhead| and related commands, see the
-% documentation of |longtable|~\cite{longtable}.
-% \begin{table}[htb]
-% \caption{Almost verbatim version of Example Table~\ref{tab:ex6}.}
-% \label{tab:ex5}
-% \begin{verbatim}
%\begin{longtable}[l]{@{}lflf@{}}
-%\caption{\label{tab:ex6}Table showing compatibility of fcolumn and
+%\caption{\label{tab:ex5}Table showing compatibility of fcolumn and
+% longtable.}\\
+%\multicolumn4{@{}c@{}}{\bfseries Balance sheet}\\
+%\toprule
+%properties & \leeg{31 dec 2014} & debts & \leeg{31 dec 2014}\\
+%\midrule
+%\endfirsthead
+%\caption[]{\textit{(continued from previous page)\/}}\\
+%\toprule
+%properties & \leeg{31 dec 2014} & debts & \leeg{31 dec 2014}\\
+%\midrule
+%\endhead
+%\bottomrule
+%\multicolumn4{@{}r@{}}{\small\textit{(Table continues on next page)\/}}\\
+%\endfoot
+%\bottomrule
+%\endlastfoot
+%house & 200000 & equity capital & 50000 \\
+%bank account & -603,23 & mortgage & 150000 \\
+%savings 1 & 7000 \\
+%savings 2 & 7000 \\
+%savings 3 & 7000 \\
+%savings 4 & 7000 \\
+%cash & 145,85 & profit & 27542,62 \\
+%\sumline
+%\end{longtable}
+%As is demonstrated in the preceding page break, the package |fcolumn| now also works
+%with |longtable|~\cite{longtable}, provided |longtable| is loaded
+%before~|fcolumn|; it checks
+%for that. The raw formatting of the multipage table
+%is shown in Table~\ref{tab:ex6}.
+%\begin{table}[htb]
+%\caption{Almost verbatim version of Example Table~\ref{tab:ex5}.}
+%\label{tab:ex6} \topsep0pt
+%\begin{verbatim}
+%\begin{longtable}[l]{@{}lflf@{}}
+%\caption{\label{tab:ex5}Table showing compatibility of fcolumn and
% longtable.}\\
-%\multicolumn4c{\bfseries Balance sheet}\\
+%\multicolumn4{@{}c@{}}{\bfseries Balance sheet}\\
%\toprule
%properties & \leeg{31 dec 2014} & debts & \leeg{31 dec 2014}\\
%\midrule
@@ -273,307 +306,306 @@ indention=0pt,labelsep=colon,justification=raggedright}
%\midrule
%\endhead
%\bottomrule
-%\multicolumn4{r@{}}{\small\textit{(Table continues on next page)\/}}\\
+%\multicolumn4{@{}r@{}}{\small\textit{(Table continues on next page)\/}}\\
%\endfoot
%\bottomrule
%\endlastfoot
-%house & 200000 & equity capital & 50000 \\
+%house & 200000 & equity capital & 50000 \\
%...
%... (Somewhat altered financial contents compared to
%... Table 1 to demonstrate the page break.)
%...
%\sumline
%\end{longtable}
-% \end{verbatim}
-% \end{table}
+%\end{verbatim}
+%\end{table}
+%For more information on how to handle |\endhead| and its ilk, see the
+%documentation of |longtable|~\cite{longtable}. Not shown here are the new |fcolumn|
+%formatting possibilities (like new fonts and/or colours), but it has been
+%checked they do work in combination with |longtable|. |fcolumn| also works happily
+%together with |supertabular|~\cite{supertabular}, so you can choose which package you
+%want if you need this. There is no race condition with |supertabular|: you can load
+%it before or after |fcolumn|.
%
-% \noindent And here is the result (it's ugly, but it had to be this long to
-% demonstrate the page break). Not shown here are the new |fcolumn|
-% formatting possibilities (like new fonts and/or colours), but it has been
-% checked they do work in combination with |longtable|.
-% \begin{longtable}[l]{@{}lflf@{}}
-% \caption{\label{tab:ex6}Table showing compatibility of fcolumn and
-% longtable.}\\
-% \multicolumn4c{\bfseries Balance sheet}\\
-% \toprule
-% properties & \leeg{31 dec 2014} & debts & \leeg{31 dec 2014}\\
-% \midrule
-% \endfirsthead
-% \caption[]{\textit{(continued from previous page)\/}}\\
-% \toprule
-% properties & \leeg{31 dec 2014} & debts & \leeg{31 dec 2014}\\
-% \midrule
-% \endhead
-% \bottomrule
-% \multicolumn4{r@{}}{\small\textit{(Table continues on next page)\/}}\\
-% \endfoot
-% \bottomrule
-% \endlastfoot
-% house & 200000 & equity capital & 50000 \\
-% bank account & -603,23 & mortgage & 150000 \\
-% savings 1 & 4000 \\
-% savings 2 & 4000 \\
-% savings 3 & 4000 \\
-% savings 4 & 4000 \\
-% savings 5 & 4000 \\
-% savings 6 & 4000 \\
-% savings 7 & 4000 \\
-% cash & 145,85 & profit & 27542,62 \\
-% \sumline
-% \end{longtable}
-% \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 four arguments of |F| are |{.}|,
-% |{,}|, |{3,2}|, and |{}|. If you want |g| to be your own definition like the
-% curious one given in Section~\ref{intro}, then specify
-% |\newcolumntype{g}{F{\,}{\cdot}{4p3}{}}| prior to using |g| in a tabular.
-%
-% Entries in an F-column are, from that moment on, treated as numbers unless
-% explicitly escaped by |\leeg|, see below. The numbers are typeset
-% according to the template the user gives with his/her |F|-column.
-% The ``middle'' character of |#3| is an important switch: it does more
-% than just setting the input decimal mark. By default the input grouping
-% character is the dot, except when the dot is specified as input decimal
-% mark; in that case the comma is acting as input grouping character. With
-% this convention continental Europe and the Anglo-Saxon part of the world
-% is served. And using input grouping markers is optional anyway.
-% \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{compactenum}[1)]
-% \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 |\PackageWarning| 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{compactenum}
-% By default the vertical separation between the rule and the total is
-% 2\thinspace pt, but this can be changed by the optional argument to
-% |\sumline|. Give, e.g., |\sumline[10pt]|, in case you want this spacing
-% to be 10\thinspace pt. And you may even give two options, like in
-% |\sumline[10pt][5pt]|, in which the second option is the extra space below
-% the summary row. In fact that second option is parsed to |\\|, that is
-% implicit in |\sumline|.
-% \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
-% |\resetsumline| is used for that: it resets all totals and all column
-% widths, see for example Table~\ref{tab:ex7}.
-% \begin{table}[htb]
-% \caption{Example: multiple projects.}
-% \label{tab:ex7}
-% \begin{tabular}{@{}lfflff@{}}
-% \multicolumn6c{\bfseries Project~1}\\
-% \toprule
-% expense & \leeg{actual} & \leeg{budget} &
-% income & \leeg{actual} & \leeg{budget} \\
-% \midrule
-% food & 450,20 & 500 & tickets & 1200 & 1000 \\
-% drinks & 547,5 & 400 \\
-% music & 180 & 100 \\
-% profit & 22,3 \\
-% \sumline
-% \resetsumline
-% \multicolumn6c{\bfseries Project~2}\\
-% \toprule
-% expense & \leeg{actual} & \leeg{budget} &
-% income & \leeg{actual} & \leeg{budget} \\
-% \midrule
-% food & 250 & 300 & tickets & 400 & 450 \\
-% drinks & 100 & 80 \\
-% music & 80 & 70 & loss & 30 \\
-% \sumline
-% \bottomrule
-% \end{tabular}
-% \end{table}
-% Note that the rules in the first and third |F|-columns of project~1 cover
-% $1{.}200{,}00$ whereas in project~2 those rules are narrower since they
-% only cover $430{,}00$; still the columns are aligned. The verbatim way
-% of setting up Table~\ref{tab:ex7} is given in Table~\ref{tab:ex8}.
-% \begin{table}[htb]
-% \caption{Verbatim version of Table~\ref{tab:ex7}.}
-% \label{tab:ex8}
-% \begin{verbatim}
+%Unlike in the package |dcolumn|~\cite{dcolumn}, the alignment on the decimal mark in
+%|fcolumn| is achieved by right aligning numbers that all have the same number of
+%digits to the right of this mark, in combination with the fact that all decimal glyphs
+%have the same width\footnote{No common font is known in which this is not the case, but
+%exceptions may exist.}. This width is font specific, so changing fonts within a table
+%may ruin the alignment, see e.g., the first |F|-column in Table~\ref{tab:ex3}.
+%\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 four arguments of |F| are |{.}|,
+%|{,}|, |{3,2}|, and |{}|. If you want |g| to be your own definition like the
+%curious one given in Section~\ref{intro}, then specify
+%|\newcolumntype{g}{F{\,}{\cdot}{4p3}{}}| prior to using |g| in a tabular.
+%
+%Entries in an |F|-column are, from that moment on, treated as numbers unless
+%explicitly escaped by |\leeg|, see below. The numbers are typeset
+%according to the template the user gives with his/her |F|-column.
+%The ``middle'' character of |#3| is an important switch: it does more
+%than just setting the input decimal mark. By default the input grouping
+%character is the dot, except when the dot is specified as input decimal
+%mark; in that case the comma is acting as input grouping character. With
+%this convention continental Europe and the Anglo-Saxon part of the world
+%are served. This is hard coded, but using input grouping markers is optional anyway.
+%\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{compactenum}[1)]
+%\item It calculates and shows 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. Otherwise a |\PackageWarning| 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{compactenum}
+%By default the vertical separation between the rule and the total is
+%2\thinspace pt, but this can be changed by the optional argument to
+%|\sumline|. Give, e.g., |\sumline[10pt]|, in case you want this spacing
+%to be 10\thinspace pt. And you may even give two options, like in
+%|\sumline[10pt][5pt]|, in which the second option is the extra space below
+%the summary row. In fact that second option is parsed to |\\|, that is
+%implicit in |\sumline|.
+%\end{macro}
+%\begin{macro}{\resetsumline}
+%Suppose you want to typeset one tabular with the profit-and-loss of many
+%projects individually. In case the layout of those tabulars is similar 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 |\resetsumline|
+%is used for that: it resets all totals and all column widths, see for example
+%Table~\ref{tab:ex7}.
%\begin{table}[htb]
-%\caption{Example: multiple projects.}
-%\label{tab:ex3}
+%\caption{Example: multiple projects.} \label{tab:ex7}
%\begin{tabular}{@{}lfflff@{}}
-%\multicolumn6c{\bfseries Project~1}\\
+%\multicolumn6{@{}c@{}}{\bfseries Project~1}\\
+%\toprule
+%expense & \leeg{actual} & \leeg{budget} &
+%income & \leeg{actual} & \leeg{budget} \\
+%\midrule
+%food & 450,20 & 500 & tickets & 1200 & 1000 \\
+%drinks & 547,5 & 400 \\
+%music & 180 & 100 \\
+%profit & 22,3 \\
+%\sumline[2pt][10pt]
+%\resetsumline
+%\multicolumn6{@{}c@{}}{\bfseries Project~2}\\
%\toprule
%expense & \leeg{actual} & \leeg{budget} &
%income & \leeg{actual} & \leeg{budget} \\
%\midrule
-%food & 450,2 & 500 & tickets & 1200 & 1000 \\
-%drinks & 547,5 & 400 \\
-%music & 180 & 100 \\
-%profit & 22,3 \\
+%food & 250 & 300 & tickets & 400 & 450 \\
+%drinks & 100 & 80 \\
+%music & 80 & 70 & loss & 30 \\
%\sumline
+%\bottomrule
+%\end{tabular}
+%\end{table}
+%Note that the rules in the first and third |F|-columns of Project~1 cover
+%$1{.}200{,}00$ whereas in Project~2 those rules are narrower since they
+%only cover $430{,}00$; still the columns are aligned (similar story for
+%|F|-columns two and four). The verbatim way of setting up Table~\ref{tab:ex7}
+%is given in Table~\ref{tab:ex8}.
+%\begin{table}[htb]
+%\caption{Verbatim version of Table~\ref{tab:ex7}.}
+%\label{tab:ex8} \topsep0pt
+%\begin{verbatim}
+%\begin{table}[htb]
+%\caption{Example: multiple projects.} \label{tab:ex7}
+%\begin{tabular}{@{}lfflff@{}}
+%\multicolumn6{@{}c@{}}{\bfseries Project~1}\\
+%expense & \leeg{actual} & \leeg{budget} &
+%income & \leeg{actual} & \leeg{budget} \\
+%\midrule
+%food & 450,20 & 500 & tickets & 1200 & 1000 \\
+%drinks & 547,5 & 400 \\
+%music & 180 & 100 \\
+%profit & 22,3 \\
+%\sumline[2pt][10pt]
%\resetsumline
-%\multicolumn6c{\bfseries Project~2}\\
+%\multicolumn6{@{}c@{}}{\bfseries Project~2}\\
%\toprule
%expense & \leeg{actual} & \leeg{budget} &
-%income & \leeg{actual} & \leeg{budget} \\
+%income & \leeg{actual} & \leeg{budget} \\
%\midrule
-%food & 250 & 300 & tickets & 400 & 450 \\
-%drinks & 100 & 80 \\
-%music & 80 & 70 & loss & 30 \\
+%food & 250 & 300 & tickets & 400 & 450 \\
+%drinks & 100 & 80 \\
+%music & 80 & 70 & loss & 30 \\
%\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
-% if it is a number---, this can be done with |\leeg|. It expects an
-% argument and this argument is typeset in the column. The common case is
-% where |p.m.| (\textit{pro memoria}) is entered. In contrast to v1.1.2
-% of this package, now even an empty |F|-column followed by |\\| is allowed.
-% \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}|. If
-% |\checkfcolumns| is used, the automatic check is disabled. Multiple
-% |\checkfcolumns|s are supported; if |F|-columns~1, 2, and~3 should
-% balance, you specify |\checkfcolumns12| and |\checkfcolumns23|.
-% There is no explicit command to disable all checking, but
-% |\checkfcolumns11| obviously does the trick.
-% \end{macro}
-% \begin{macro}{\ifstrict@ccounting}
-% 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 |\strict@ccountingtrue|, but since this contains a non-letter,
-% it is also possible to invoke |fcolumn| with the option |strict|, i.e.,
-% |\usepackage[strict]{fcolumn}|, which sets this flag.
-% \end{macro}
-% \StopEventually{\PrintChanges\PrintIndex}
-% \section{The macros}
-% Here follows the actual code.
-% \subsection{Option}
-% \begin{macro}{option strict}
-% There is one option. If set, strict accounting rules are used in display.
-% \begin{macrocode}
+%\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
+%if it is a number---, this can be done with |\leeg|. It expects an
+%argument and this argument is typeset in the column. The common case is
+%where |p.m.| (\textit{pro memoria}) is entered. In contrast to v1.1.2
+%of this package, now even an empty |F|-column followed by |\\| is allowed.
+%\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}|. If
+%|\checkfcolumns| is used, the automatic check is disabled. Multiple
+%|\checkfcolumns|s are supported; if |F|-columns~1, 2, and~3 should
+%balance, you specify |\checkfcolumns12| and |\checkfcolumns23|.
+%There is no explicit command to disable all checking, but
+%|\checkfcolumns11| obviously does the trick at the expense of
+%some trivial calculations.
+%\end{macro}
+%\begin{macro}{\ifstrict@ccounting}
+%\begin{macro}{\ifminusr@d}
+%In the rare occasion that a negative number occurs in a financial table,
+%the sign of that number can be an explicit minus sign, i.e.,~`$-$' 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 prescribes that the
+%number should be put between parentheses. The latter can be accomplished
+%by setting |\strict@ccountingtrue|, but since this contains a non-letter,
+%it is also possible to invoke |fcolumn| with the option |strict|, i.e.,
+%|\usepackage[strict]{fcolumn}|, which sets this flag. If you want the
+%negative numbers to be coloured red, use the option |red|, i.e.,
+%|\usepackage[red]{fcolumn}|, but note that |strict| and |red|
+%are mutually exclusive: if both are used, |red| is reset because
+%|strict| is strict.
+%\end{macro}
+%\end{macro}
+%\StopEventually{\PrintChanges\PrintIndex}
+%\section{The macros}
+%Here follows the actual code.
+%\subsection{Options}
+%\begin{macro}{option red}
+%\begin{macro}{option strict}
+%There are two options. If |red| is set, negative numbers are displayed in red.
+%If |strict| is set, strict accounting rules are used in display and this overrules
+%a possibly set |red|. Depending on the two booleans associated with these options
+%the |color| package~\cite{color} will be loaded or not.
+% \begin{macrocode}
+\newif\ifminusr@d \minusr@dfalse
\newif\ifstrict@ccounting \strict@ccountingfalse
+\DeclareOption{red}{\minusr@dtrue}
\DeclareOption{strict}{\strict@ccountingtrue}
-\ProcessOptions
+\ProcessOptions \ifminusr@d \ifstrict@ccounting \minusr@dfalse
+\PackageWarningNoLine{fcolumn}{Option `red' is reset due to use
+ of `strict'}\else\usepackage{color}\fi\fi
% \end{macrocode}
-% \end{macro}
+%\end{macro}
+%\end{macro}
%\subsection{Definitions}
-% \begin{macro}{column F}
-% \begin{macro}{column f}
-% The column specifier |F| is the generic one, and |f| is the default
-% (continental 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{macro}{column F}
+%\begin{macro}{column f}
+%The column specifier |F| is the generic one, and |f| is the default
+%(continental 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}[4]{>{\b@fi{#1}{#2}{#3}{#4}}r<{\e@fi}}
-\newcolumntype{f}{F{.}{,}{3,2}{}}
-% \end{macrocode}
-% \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 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 is used for the first time.
+\newcolumntype{f}{F.,{3,2}{}}
+% \end{macrocode}
+%\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 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 is used for the first time.
% \begin{macrocode}
\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, Pfennige,
-% centimes, kopecks, groszy) and typesets it as the amount in entire
-% currency units (dollars, euros, guldens, Marke, francs, rubles, z\l oty)
-% with comma as decimal mark and the dot as grouping character
-% (thousand separator if the first part of |#1| is~3).
-% As explained, this can be changed. It uses two private booleans:
-% |\withs@p| and |\strict@ccounting|. The latter is used to typeset
-% negative numbers between parentheses. By default it doesn't do this:
-% a minus sign is used.
+%\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, Pfennige,
+%centimes, kopecks, groszy) and typesets it as the amount in entire
+%currency units (dollars, euros, guldens, Marke, francs, rubles, z\l oty)
+%with comma as decimal mark and the dot as grouping character
+%(thousand separator if the first part of |#1| is~3).
+%As explained, this can be changed. It uses three private booleans:
+%|\ifwiths@p|, |\ifstrict@ccounting|, and |\ifminusr@d|. The latter two
+%are used to typeset negative numbers in a special way. By default it
+%doesn't do this: a minus sign is used.
% \begin{macrocode}
\newif\ifwiths@p
% \end{macrocode}
-% Actually |\geldm@cro| is only a wrapper around |\g@ldm@cro|.
+%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@#1\relax{#2}}
% \end{macrocode}
-% \end{macro}
-% \begin{macro}{\g@ldm@cro}
-% After setting the environment for formatting, 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 |\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 |F|-column-specific
-% \meta{count}s because it is all done locally. Leaving the tabular environment will
-% restore their values. This is also true for the effect of |\FCform@t|, so that
-% formatting information is local to this column. The reason for inserting the |{|
-% between |\FCform@t| and |\ifnum| (and the accompanying |}| just before finishing
-% this definition) is to facilitate the possible use of |\mathbf| or any other font
-% changing command as the last item in |\FCform@t|.
-% \begin{macrocode}
-\def\g@ldm@cro#1\relax#2{\FCform@t{\ifnum#2<0 \ifstrict@ccounting
- (\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}
-\count4=\ifx\relax#1\relax 2 \else \@gobble#1\relax\fi
+%\end{macro}
+%\begin{macro}{\g@ldm@cro}
+%After setting the environment for formatting, 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 |\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 |F|-column-specific
+%\meta{count}s because it is all done locally. Leaving the tabular environment will
+%restore their values. This is also true for the effect of |\FCform@t|, so that
+%formatting information is local to this column. The reason for inserting the |{|
+%between |\FCform@t| and |\ifnum| (and the accompanying |}| just before finishing
+%this definition) is to facilitate the possible use of |\mathbf| or any other font
+%changing command as the last item in |\FCform@t|.
+%
+%When |\FCform@t| includes a colour change, this is overruled for negative numbers
+%when option |red| is set. For that reason it is unwise to use a red colour for the
+%whole column in combination with the option |red|: sign information is obscured then.
+% \begin{macrocode}
+\def\g@ldm@cro#1\relax#2{\ifstrict@ccounting\def\bm@nus{(}
+ \def\em@nus{)}\else\def\em@nus{}\ifminusr@d\def\bm@nus{\color{red}}
+ \else\def\bm@nus{-}\fi\fi\FCform@t{\ifnum#2<0 \bm@nus\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}
+\count4=\ifx\relax#1\relax 2 \else\@gobble#1\relax\fi
\count3=0
\loop\ifnum\count3<\count4
\divide\count2 by 10 \advance\count3 by \@ne
\repeat
% \end{macrocode}
-% Note that |\count3| now equals |\count4|: this going up-and-down will be used more often,
-% it saves several assignments. The value in |\count2| is then output by |\g@ldens| using
-% the separation given (and stored in |\count@|).
+%Note that |\count3| now equals |\count4|: this going up-and-down will be used more often,
+%it saves several assignments. The value in |\count2| is then output by |\g@ldens| using
+%the separation given (and stored in |\count@|).
% \begin{macrocode}
\g@ldens{\the\count@}%
% \end{macrocode}
-% If there is a decimal part\dots
+%If there is a decimal part\dots
% \begin{macrocode}
-\ifnum\count3>0\decim@lmark
+\ifnum\count3>0 {\decim@lmark}
% \end{macrocode}
-% Next the decimal part is dealt with. Now $x\bmod a$ is calculated in the
-% usual way: $x-(x/a)*a$ with integer division. The minus sign necessary
-% for this calculation is introduced in the next line by changing the
-% comparison from |<| to |>|.
+%Next the decimal part is dealt with. Now $x\bmod a$ is calculated in the
+%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
\loop\ifnum\count3>0
\divide\count2 by 10 \advance\count3 by \m@ne
\repeat
% \end{macrocode}
-% The value of |\count3| is now~0, so counting up again.
+%The value of |\count3| is now~0, so counting up again.
% \begin{macrocode}
\loop\ifnum\count3<\count4
\multiply\count2 by 10 \advance\count3 by \@ne
@@ -584,62 +616,52 @@ indention=0pt,labelsep=colon,justification=raggedright}
\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 this entry in the column is wrong. This is done by an
-% |\rlap| and therefore does not influence the column width. For the last
-% column this means that this parenthesis may even stick out of the table.
-% I don't like this, therefore I chose to put |\strict@ccountingfalse|.
-% Change if you like, by setting the option |strict|.
-%
-% If overflow was detected, an exclamation mark is output to the right of
-% the value that caused this. This of course ruins the appearance of the
-% table, but in this case that serves a clear goal: there's something wrong
-% and you should know.
+%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 this entry in the column is wrong. This is done by an
+%|\rlap| and therefore does not influence the column width. For the last
+%column this means that this parenthesis may even stick out of the table.
+%I don't like this, therefore I chose to put |\strict@ccountingfalse|.
+%Change if you like, by setting the option |strict|.
+%
+%If overflow was detected, an exclamation mark is output to the right of
+%the value that caused this. This of course ruins the appearance of the
+%table, but in this case that serves a clear goal: there's something wrong
+%and you should know.
% \begin{macrocode}
-\ifx\FCs@gn\m@ne \ifnum#2<0 \ifstrict@ccounting
- \rlap{)~!}\else\rlap{~!}\fi\else\ifstrict@ccounting
- \rlap{\phantom{)}~!}\else\rlap{~!}\fi\fi
-\else \ifnum#2<0 \ifstrict@ccounting\rlap{)}\fi\fi
-\fi}}
+\ifx\FCs@gn\m@ne\ifnum#2<0 \rlap{\em@nus~!}
+ \else\rlap{\phantom{\em@nus}~!}\fi
+\else\ifnum#2<0 \rlap{\em@nus}\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. This also implies that tail recursion is not
-% possible here, but that is not very important, as the largest number
-% (which is $2^{31}-1$) will only cause a threefold recursion using the
-% default |3,2| (ninefold when using~|1,0|, but who does that?). The
-% largest amount this package can deal with is therefore 2.147.483.647
-% (using |3,0|). For most people this is probably more than enough if
-% the currency is euros or dollars. And otherwise make clear that you
-% use a currency unit of~k\EUR{} (or even M\EUR{} for the very rich).
-% The author is thinking of ways to use two counters for each number.
-% The maximum then becomes $2^{63}-1$. Even expressed in cents this
-% would lead to a maximum of slightly more than 92.2\thinspace P\EUR{};
-% about 100 times the current world economy~\cite{world_economy}. Yet
-% another method is to use Heiko Oberdiek's package |bigintcalc|:
-% then only memory restrictions apply. This, however, would require
-% a major rewrite of |fcolumn|.
+%\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. This also implies that tail recursion is not
+%possible here, but that is not very important, as the largest number
+%(which is $2^{31}-1$) will only cause a threefold recursion using the
+%default |3,2| (ninefold when using~|1,0|, but who does that?). The
+%largest amount this package can deal with is therefore 2.147.483.647
+%(using |3,0|). For most people this is probably more than enough if
+%the currency is euros or dollars. And otherwise clearly state that you
+%use a currency unit of~k\EUR{} (or even M\EUR{} for the very rich).
%
-% There is no straightforward interpretation of |#1| being zero or
-% negative, therefore this is used as an indicator that no grouping
-% character should be used.
+%There is no obvious interpretation of |#1| being zero or negative,
+%therefore this is used as an indicator that no grouping character should be used.
% \begin{macrocode}
\def\g@ldens#1{{\count3=\count2 \count0=#1
% \end{macrocode}
-% First divide by $10^n$, where $n$ is |#1|.
+%First divide by $10^n$, where $n$ is |#1|.
% \begin{macrocode}
\ifnum\count0<1 \count0=3 \fi
\loop \ifnum\count0>0 \divide\count2 by 10 \advance\count0 by \m@ne
\repeat
% \end{macrocode}
-% Here is the recursive part,
+%Here is the recursive part,
% \begin{macrocode}
\ifnum\count2>0 \g@ldens{#1}\fi
% \end{macrocode}
-% and then reconstruct the rest of the number.
+%and then reconstruct the rest of the number.
% \begin{macrocode}
\count0=#1
\ifnum\count0<1 \count0=3 \fi
@@ -648,38 +670,38 @@ indention=0pt,labelsep=colon,justification=raggedright}
\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. The former, however, is only done if |#1|
-% is larger than zero.
+%\end{macro}
+%\begin{macro}{\du@zendprint}
+%The macro |\du@zendprint| takes care for correctly printing the separator
+%and possible trailing zeros. The former, however, is only done if |#1|
+%is larger than zero.
% \begin{macrocode}
-\def\du@zendprint#1{\ifwiths@p\ifnum#1>0 \sep@rator\fi
+\def\du@zendprint#1{\ifwiths@p\ifnum#1>0 {\sep@rator}\fi
\zerop@d{#1}{\number\count2}%
\else\zerop@d1{\number\count2}\fi\global\withs@ptrue}
% \end{macrocode}
-% \end{macro}
-% \begin{macro}{\zerop@d}
-% The macro |\zerop@d| uses at least |#1| digits for printing the
-% number |#2|, padding with zeros when necessary. Note: |#1| being
-% zero or negative is a flag that it should be interpreted as 3.
-% A bit ugly, but it works, since the related code knows about this.
+%\end{macro}
+%\begin{macro}{\zerop@d}
+%The macro |\zerop@d| uses at least |#1| digits for printing the
+%number |#2|, padding with zeros when necessary. Note: |#1| being
+%zero or negative is a flag that it should be interpreted as 3.
+%A bit ugly, but it works, since the related code knows about this.
%
-% It is done within an extra pair of braces, so that |\count0|
-% and |\count1| can be used without disturbing their values in other macros.
+%It is done within an extra pair of braces, so that |\count0|
+%and |\count1| can be used without disturbing their values in other macros.
% \begin{macrocode}
\def\zerop@d#1#2{{\count0=1 \count1=#2
% \end{macrocode}
-% First determine the number of digits of |#2| (expressed in the decimal
-% system). This number is in |\count0| and is at least~1.
+%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 \@ne
\repeat
% \end{macrocode}
-% If |#1| is positive, the number of zeros to be padded is
-% $\max(0,|#1-\count0|)$ (the second argument can be negative), so a
-% simple loop suffices. If it is zero or negative, this is a signal
-% that it should be interpreted as 3 (and no separator will be output).
+%If |#1| is positive, the number of zeros to be padded is
+%$\max(0,|#1-\count0|)$ (the second argument can be negative), so a
+%simple loop suffices. If it is zero or negative, this is a signal
+%that it should be interpreted as 3 (and no separator will be output).
% \begin{macrocode}
\ifnum#1>0
\loop \ifnum\count0<#1\relax 0\advance\count0 by \@ne
@@ -690,21 +712,21 @@ indention=0pt,labelsep=colon,justification=raggedright}
\repeat
\fi\number#2}}
% \end{macrocode}
-% \end{macro}
-% \begin{macro}{\zetg@ld}
-% This macro takes care for several things: it increases the subtotal for a
-% given F-column, it checks whether or not that subtotal has overflown, it
-% records the largest width of the entries in that column and it typesets
-% |#1| via |\geldm@cro|.
+%\end{macro}
+%\begin{macro}{\zetg@ld}
+%This macro takes care for several things: it increases the subtotal for a
+%given |F|-column, it checks whether or not that subtotal has overflown, 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{\count0=#2\relax \let\FCs@gn=\@ne
% \end{macrocode}
-% First it checks whether there is a risk of overflow in this step. If
-% $A$ and $B$ are two \TeX-registers and $B$ is to be added to $A$,
-% overflow will not occur if one is (or both are) zero or if $A$ and~$B$
-% have different signs; otherwise be careful. Note that \TeX\ does not
-% check for overflow when performing an |\advance| (done in section~1238
-% of Ref.~\cite{CandTB}), in contrast to |\multiply|, see section~105.
+%First it checks whether there is a risk of overflow in this step. If
+%$A$ and $B$ are two \TeX-registers and $B$ is to be added to $A$,
+%overflow cannot occur if 1)~one is (or both are) zero or 2)~if $A$ and~$B$
+%have different signs; otherwise be careful. Note that \TeX\ does not
+%check for overflow when performing an |\advance| (done in section~1238
+%of Ref.~\cite{CandTB}), in contrast to |\multiply|, see section~105.
% \begin{macrocode}
\ifnum\count0<0
\ifnum\csname FCtot@\romannumeral\FCsc@l\endcsname<0
@@ -719,469 +741,483 @@ indention=0pt,labelsep=colon,justification=raggedright}
\global\advance\csname FCtot@\romannumeral\FCsc@l\endcsname by \count0
\ifx\FCs@gn\m@ne
% \end{macrocode}
-% They have the same sign, hence risk of overflow. Record the sign
-% of |\count0| (and of the original total of this column; we just established
-% they were the same) in |\FCs@gn|. Table~\ref{tab:ex9}
-% \begin{table}[htb]
-% \def\PackageError#1#2#3{\PackageWarning{#1}{#2: #3}}
-% \caption{Examples on overflow.}
-% \label{tab:ex9}
-% \begin{tabular}{@{}lfff@{}}
-% \multicolumn4c{\bfseries Projects}\\
-% \toprule
-% income & \leeg{31 dec 2014} & \leeg{31 dec 2015} & \leeg{31 dec 2016} \\
-% \midrule
-% item~1 & 20000000 & 20000000 &20000000 \\
-% item~2 & 10000000 & 2000000 &-1500000 \\
-% item~3 & 5000000 & -1500000 & 2000000 \\
-% \checkfcolumns12
-% \checkfcolumns23
-% \sumline
-% \bottomrule
-% \end{tabular}
-% \end{table}
-% shows what can go wrong if the numbers are too large: in the left
-% |F|-column the sumline is incorrect and the number that caused
-% the overflow is indicated by an exclamation mark. In the middle
-% |F|-column, overflow occurs twice and because this is once positive,
-% once negative here, cancellation of errors occurs and the sumline
-% is correct in the end. Nevertheless, it is advised to swap the two
-% items that caused the overflow, as shown in the right |F|-column.
+%They have the same sign, hence risk of overflow. Record the sign
+%of |\count0| (and of the original total of this column; we just established
+%they were the same) in |\FCs@gn|. Table~\ref{tab:ex9}
+%\begin{table}[htb]
+%\def\PackageError#1#2#3{\PackageWarning{#1}{#2: #3}}
+%\caption{Examples on overflow.}
+%\label{tab:ex9}
+%\begin{tabular}{@{}lfff@{}}
+%\multicolumn4{@{}c@{}}{\bfseries Projects}\\
+%\toprule
+%income & \leeg{31 dec 2014} & \leeg{31 dec 2015} & \leeg{31 dec 2016} \\
+%\midrule
+%item~1 & 20000000 & 20000000 & 20000000 \\
+%item~2 & 10000000 & 2000000 & -1500000 \\
+%item~3 & 5000000 & -1500000 & 2000000 \\
+%\checkfcolumns23
+%\sumline
+%\bottomrule
+%\end{tabular}
+%\end{table}
+%shows what can go wrong if the numbers are too large: in the left
+%|F|-column the sumline is incorrect and the number that caused
+%the overflow is indicated by an exclamation mark. In the middle
+%|F|-column, overflow occurs twice and because this is once positive,
+%once negative here, cancellation of errors occurs and the sumline
+%is correct in the end. Nevertheless, it is advised to swap the two
+%items that caused the overflow, as shown in the right |F|-column.
%
-% Check that the sign of the updated column total is still correct.
-% If so, |\FCs@gn| is reset (to |\@ne|) at the end of this chunk.
+%Check that the sign of the updated column total is still correct.
+%If so, |\FCs@gn| is reset (to |\@ne|) at the end of this chunk.
% \begin{macrocode}
\ifnum\count0>0 \let\FCs@gn\@ne \fi
\count0=\csname FCtot@\romannumeral\FCsc@l\endcsname
\ifnum\FCs@gn<0 \count0=-\count0 \fi
\ifnum\count0<0
\let\FCs@gn=\m@ne
- \PackageError{fcolumn}{Register overflow}{Overflow occurred
- in fcolumn \number\FCsc@l.\MessageBreak You can press <enter>
- now and I'll proceed, but check\MessageBreak your table. The
- offending entry is indicated with an\MessageBreak exclamation
- mark in the output.}%
+ \PackageError{fcolumn}{Register overflow}{Overflow occurred in
+ fcolumn \number\FCsc@l\space near or at line \the\inputlineno.
+ You can\MessageBreak press <enter> now and I'll proceed, but check
+ your table.\MessageBreak The offending entry is indicated with an
+ exclamation mark\MessageBreak in the output.}%
\else\let\FCs@gn=\@ne
\fi
\fi
% \end{macrocode}
-% The value of |\FCs@gn| is used in |\geldm@cro| below.
+%The value of |\FCs@gn| is used in |\geldm@cro| below.
% \begin{macrocode}
\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}
+%\end{macro}
%
-% The \meta{count}s |\FC@l| and |\FC@r| capture the parts to the left and to the
-% right of the decimal mark, respectively.
+%The \meta{count}s |\FC@l| and |\FC@r| capture the parts to the left and to the
+%right of the decimal mark, respectively.
% \begin{macrocode}
\newcount\FC@l \newcount\FC@r
% \end{macrocode}
-% Some auxiliary definitions for capturing compacted information.
+%Some auxiliary definitions for capturing compacted information.
% \begin{macrocode}
\def\setucc@de#1#2\relax{\uccode`\~=`#1 }
\def\assignform@t#1,#2,#3\assignform@t{\def\FCform@t{#1}%
\def\FCform@tt{#2}\ifx\FCform@tt\@empty \def\FCform@tt{#1}\fi}
% \end{macrocode}
-% \begin{macro}{\m@thcodeswitch}
-% As will be shown below, once the first digit, or sign, or decimal mark, or
-% grouping character is scanned, the decimal digits should loose their
-% activeness. That is done here for the digits in a rather blunt way, by putting their
-% |\mathcode|s to zero if |#1| equals~0, since the actual |\mathcode| is not important---as
-% long as it is not |"8000|---because the digits are not used for typesetting.
-% (And even if they were; it's inside |\box0|, whose contents will be discarded.)
-% When the |$| in |\e@fi| is encountered, the digits get back their original
-% |\mathcode|s so that the actual typesetting in |\zetg@ld| is correct again.
-% With a non-zero |#1| the activeness is switched on.
+%\begin{macro}{\m@thcodeswitch}
+%As will be shown below, once the first digit, or sign, or decimal mark, or
+%grouping character is scanned, the decimal digits should loose their activeness.
+%That is done here for the digits in a rather blunt way, by putting their
+%|\mathcode|s to zero if |#1| equals~0, since the actual |\mathcode| is not
+%important---as long as it is not |"8000|---because the digits are not used for
+%typesetting. (And even if they were; it's inside |\box0|, whose contents will be
+%discarded.) When the |$| in |\e@fi| is encountered, the digits get back their
+%original |\mathcode|s so that the actual typesetting in |\zetg@ld| is correct again.
+%With a non-zero |#1| the activeness is switched on.
% \begin{macrocode}
\def\m@thcodeswitch#1{\count0=10 \loop\ifnum\count0>0
\advance\count0 by \m@ne\mathcode\expandafter`\the\count0=
\ifnum#1=0 0 \else "8000 \fi\repeat}
% \end{macrocode}
-% \end{macro}
-% \def\<#1>{\leavevmode\hbox{$\langle$#1\/$\rangle$}}
-% \begin{macro}{\b@fi}
-% The macro |\b@fi| provides the beginning of the financial column. It
-% will be inserted in the column to capture the number entered by the user.
-% 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|, Ref.~\cite{dcolumn}).
-% \begin{macrocode}
-\def\b@fi#1#2#3#4{%
-% \end{macrocode}
-% An intermediate macro |\sep@xt| to extract the first character
-% of |#1|, which in most cases will be the only character.
-% \begin{macrocode}
-\def\sep@xt##1##2\end{\def\sep@rator{{##1}}}\sep@xt#1\end
-\def\decim@lmark{{#2}}\def\sp@l{#3}\assignform@t#4,,\assignform@t
-\global\advance\FCsc@l by \@ne\global\FC@l=0 \global\FC@r=1
-% \end{macrocode}
-% The value specified by the user is then captured by |\FC@l| and
-% this is done in a special way: |\FC@l| is assigned globally within
-% |\box0|. Why? To use it as scribbling paper to examine what the
-% user entered, without dumping it into the horizontal list.
+%\end{macro}
+%\def\<#1>{\leavevmode\hbox{$\langle$#1\/$\rangle$}}
+%\begin{macro}{\b@fi}
+%The macro |\b@fi| provides the beginning of the financial column. It
+%will be inserted in the column to capture the number entered by the user.
+%The separator and decimal mark are within a math environment, so you
+%can indeed specify |\,| instead of |\thinspace|, but there are extra
+%braces around them in the code where they are used, so it doesn't
+%affect the spacing between the digits (trick copied from |dcolumn|,
+%Ref.~\cite{dcolumn}).
+% \begin{macrocode}
+\def\b@fi#1#2#3#4{\sep@xt#1\sep@xt\def\decim@lmark{#2}\def\sp@l{#3}%
+\assignform@t#4,,\assignform@t\global\advance\FCsc@l by \@ne
+\global\FC@l=0 \global\FC@r=1
+% \end{macrocode}
+%The value specified by the user is then captured by |\FC@l| and
+%this is done in a special way: |\FC@l| is assigned globally within
+%|\box0|. Why? To use it as scribbling paper to examine what the
+%user entered, without dumping it into the horizontal list.
%
-% There are four parts to an |F|-column entry, all parts optional,
-% making 16 combinations. The sequence is
-% (in the Backus--Naur notation of Ref.~\cite{CandTA}):
-% \<sign> \<integer constant> \<decimal mark>
-% \<integer constant>. Here \<sign> is a plus or minus character with
-% category code~12, \<integer constant> is a sequence of zero
-% or more (decimal) \<digit>s, and \<decimal mark> is the middel
-% part of |#3|, i.e., the comma in |3,2| or the period in |3.2|. If
-% the \<decimal mark> is absent with no space characters between the
-% two \<integer constant> terms, these merge, making four redundant
-% entries. One of the combinations is \<empty>, a sequence of characters
-% outside the ones that are made active, e.g., plain text or nothing at all:
-% this is the only combination that doesn't put anything in an |F|-column---and
-% was the most difficult part to handle.
+%There are four parts to an |F|-column entry, all parts optional,
+%making 16 combinations. The sequence is
+%(in the Backus--Naur notation of Ref.~\cite{CandTA}):
+%\<sign> \<integer constant> \<decimal mark>
+%\<integer constant>. Here \<sign> is a plus or minus character with
+%category code~12, \<integer constant> is a sequence of zero
+%or more (decimal) \<digit>s, and \<decimal mark> is the middel
+%part of |#3|, i.e., the comma in |3,2| or the period in |3.2|. If
+%the \<decimal mark> is absent with no space characters between the
+%two \<integer constant> terms, these merge, making four redundant
+%entries. One of the combinations is \<empty>, a sequence of characters
+%outside the ones that are made active, e.g., plain text or nothing at all:
+%this is the only combination that doesn't put anything in an |F|-column---and
+%was the most difficult part to handle.
%
-% The minus sign must be captured separately, because in an entry like
-% |-0,07| the 7~cents are negative, but this cannot be seen from the
-% part to the left of the decimal mark, since $-0$ is~$0$ in \TeX\ (in
-% fact in most computer languages, but not in~MIX~\cite{TAOCP1}), so
-% |\ifnum-0<0| yields |false|. |\FCs@gn| is a general purpose flag. Its
-% first use is to capture the sign.
+%The minus sign must be captured separately, because in an entry like
+%|-0,07| the 7~cents are negative, but this cannot be seen from the
+%part to the left of the decimal mark, since $-0$ is~$0$ in \TeX\ (in
+%fact in most computer languages, but not in~MIX~\cite{TAOCP1}), so
+%|\ifnum-0<0| yields |false|. |\FCs@gn| is a general purpose flag. Its
+%first use is to capture the sign.
% \begin{macrocode}
\let\FCs@gn=\@ne\relax \setbox0\hbox\bgroup$
% \end{macrocode}
-% Do the scan inside a box and inside math mode. Start with defining all
-% characters that may appear as the first one in an |F|-column as active;
-% digits first. This needs a complicated |\edef| with accompanying |\noexpand|
-% because the starting digit for |\FC@l| must be packed inside each definition.
+%Do the scan inside a box and inside math mode. Start with defining all
+%characters that may appear as the first one in an |F|-column as active;
+%digits first. This needs a complicated |\edef| with accompanying |\noexpand|
+%because the starting digit for |\FC@l| must be packed inside each definition.
% \begin{macrocode}
\count@=10 \loop\ifnum\count@>0 \advance\count@ by \m@ne
\uccode`\~=\expandafter`\the\count@ \uppercase{\edef~}{\noexpand
\m@thcodeswitch0 \global\FC@l=\the\count@}\repeat
% \end{macrocode}
-% For the input decimal mark something extra is needed: if it is the
-% first character in an |F|-column (like in |,07|), it should also
-% restore the |\mathcode|s of the digits. Checking whether or not it
-% is the first is easy, since in that case the |\mathcode|s of the
-% decimal digits is still |"8000|. The assignment to |\FC@r| starts
-% with~1, so that appended digits get captured correctly, even if they
-% start with~0; postprocessing of |\FC@r| is done in |\e@fi|.
-% The input decimal mark switches itself off as active character,
-% so at most one input decimal mark is allowed (N.B.:
-% this makes sense).
-% \begin{macrocode}
-\def\deactdecm@rk##1##2\relax{\mathcode`##1=0 }%
+%For the input decimal mark something extra is needed: if it is the
+%first character in an |F|-column (like in |,07|), it should also
+%restore the |\mathcode|s of the digits. Checking whether or not it
+%is the first is easy, since in that case the |\mathcode|s of the
+%decimal digits are still |"8000|. The assignment to |\FC@r| starts
+%with~1 (with no space following), so that appended digits get captured
+%correctly, even if they start with~0; postprocessing of |\FC@r| is done
+%in |\e@fi|. The input decimal mark switches itself off as active
+%character, so at most one input decimal mark is allowed (N.B.:
+%this makes sense).
+% \begin{macrocode}
\afterassignment\setucc@de\count@#3\relax
\uppercase{\def~}{\ifnum\mathcode`\0="8000 \m@thcodeswitch0 \fi
\afterassignment\deactdecm@rk\count@#3\relax \global\FC@r=1}%
% \end{macrocode}
-% The input grouping character effectively expands to ``nothing, i.e.,
-% ignore'' in a complicated way: it ignores the character and resumes
-% scanning the number. The test prior to that action is needed if
-% the grouping character is the first character encountered in the
-% |F|-column. Which part to continue with depends on whether or not an
-% input decimal mark was encountered; that can be checked by looking
-% at its |\mathcode|.
-%
-% The input grouping character is the dot ``|.|'', except when that character
-% was already chosen as input decimal mark. In that case, the grouping
-% character will be the comma. This is easy to check because the |\uccode|
-% of |`\~| is still preserved.
+%The input grouping character effectively expands to ``nothing, i.e.,
+%ignore'' in a complicated way: it ignores the character and resumes
+%scanning the number. The test prior to that action is needed if
+%the grouping character is the first character encountered in the
+%|F|-column. Which part to continue with depends on whether or not an
+%input decimal mark was encountered; that can be checked by looking
+%at its |\mathcode|.
+%
+%The input grouping character is the dot ``|.|'', except when that character
+%was already chosen as input decimal mark. In that case, the grouping
+%character will be the comma. This is easy to check because the |\uccode|
+%of |`\~| is still preserved.
% \begin{macrocode}
-\def\d@cm##1##2{\count@=\mathcode`##1 }
\ifnum\uccode`\~=`. \uccode`\~=`,\relax\else \uccode`\~=`.\relax\fi
\uppercase{\def~}{\ifnum\mathcode`\0="8000 \m@thcodeswitch0 \fi
\afterassignment\d@cm\count@#3\relax
% \end{macrocode}
-% The |\expandafter| below is necessary because the global assignment
-% should act after the |\fi|.
+%The |\expandafter| below is necessary because the global assignment
+%should act after the |\fi|.
% \begin{macrocode}
\ifnum\count@=\mathcode`- \expandafter\global\FC@l=\the\FC@l
\else \expandafter\global\FC@r=\the\FC@r\fi}%
% \end{macrocode}
-% The signs are relatively simple: record the sign, restore |\mathcode|s if
-% needed (it should be: a minus sign between digits screws up everything),
-% and start scanning the number.
+%The signs are relatively simple: record the sign, restore |\mathcode|s if
+%needed (it should be: a minus sign between digits screws up everything),
+%and start scanning the number.
% \begin{macrocode}
\uccode`\~=`+\relax \uppercase{\def~}{\ifnum\mathcode`\0="8000
\m@thcodeswitch0 \fi\global\FC@l=0}
\uccode`\~=`-\relax \uppercase{\def~}{\ifnum\mathcode`\0="8000
\m@thcodeswitch0 \fi\global\let\FCs@gn\m@ne \global\FC@l=0}%
% \end{macrocode}
-% Now actually activate all these codes.
+%Now actually activate all these codes.
% \begin{macrocode}
\mathcode`-="8000 \mathcode`+="8000 \mathcode`.="8000
% \end{macrocode}
-% These three remain active until the |$| in |\e@fi| is encountered.
-% The following ones will, except in the \<empty> case, have their activeness
-% turned off at some time.
-% \begin{macrocode}
-\m@thcodeswitch1 \def\actdecm@rk##1##2\relax{%
-\ifx##1.\relax \mathcode`,="8000 \else \mathcode`##1="8000 \fi}%
-\afterassignment\actdecm@rk\count@#3\relax}
-% \end{macrocode}
-% \end{macro}
-% \begin{macro}{\e@fi}
-% If the digits are still active then either nothing was entered or only
-% characters that did not deactivate the digits were entered. In
-% both cases the output should be \<empty>. To flag this outside
-% the group that started with the opening |$| of |\b@fi|, |\FC@r| is set
-% globally to a negative value. This doesn't harm, because it didn't
-% contain relevant information anyway. Outside the group, the sign
-% of |\FC@r| can then be tested. This is a slight misuse of this
-% |\count|, but now it's documented. In effect, |\FC@r| can only be~|-1|,
-% |1|, or at least~|10|, so the comparison |\ifnum\FC@r>0| does not miss~0.
+%These three remain active until the |$| in |\e@fi| is encountered.
+%The following ones will, except in the \<empty> case, have their activeness
+%turned off at some time.
+% \begin{macrocode}
+\m@thcodeswitch1 \afterassignment\actdecm@rk\count@#3\relax}
+% \end{macrocode}
+%\end{macro}
+%\begin{macro}{\sep@xt}
+%\begin{macro}{\actdecm@rk}
+%\begin{macro}{\deactdecm@rk}
+%\begin{macro}{\d@cm}
+%Again a few small macros to do important work. At first |\sep@xt| to extract
+%the first character of |#1|, which in most cases will be the only character.
+%Then |\actdecm@rk| and |\deactdecm@rk| to activate and deactive respectively the
+%decimal mark. Finally, |\d@cm| captures a |\mathcode| for further investigation.
+% \begin{macrocode}
+\def\sep@xt#1#2\sep@xt{\def\sep@rator{#1}}
+\def\actdecm@rk#1#2\relax{\ifx#1.\relax \mathcode`,="8000
+ \else \mathcode`#1="8000 \fi}
+\def\deactdecm@rk#1#2\relax{\mathcode`#1=0 }%
+\def\d@cm#1#2{\count@=\mathcode`#1 }
+% \end{macrocode}
+%\end{macro}
+%\end{macro}
+%\end{macro}
+%\end{macro}
+%\begin{macro}{\e@fi}
+%If the digits are still active then either nothing, or only
+%characters that did not deactivate the digits were entered. In
+%both cases the output should be \<empty>. To flag this situation outside
+%the group that started with the opening |$| of |\b@fi|, |\FC@r| is set
+%globally to a negative value. This doesn't harm, because it didn't
+%contain relevant information anyway. Outside the group, the sign
+%of |\FC@r| can then be tested. This is a slight misuse of this
+%|\count|, but now it's documented. In effect, |\FC@r| can only be~$-1$,
+%1, or at least~10, so the comparison |\ifnum\FC@r>0| does not miss~0.
% \begin{macrocode}
\def\e@fi{\ifnum\mathcode`\0="8000 \global\FC@r=\m@ne\fi$\egroup
\ifnum\FC@r>0
% \end{macrocode}
-% If there was no decimal mark or if there was a decimal mark but no
-% decimal part, |\FC@r| will still be~1, which doesn't parse well with
-% |\secd@xt|, so a zero is appended, i.e., yielding~|10|.
+%If there was no decimal mark or if there was a decimal mark but no
+%decimal part, |\FC@r| will still be~1, which doesn't parse well with
+%|\secd@xt|, so a zero is appended, i.e., yielding~|10|.
% \begin{macrocode}
\ifnum\FC@r=1 \FC@r=10 \fi
% \end{macrocode}
-% Next is a loop for bringing the decimal part in the correct way to
-% the integer part. The loop is performed the number of decimal
-% digits to be printed (the~2 in |3,2| of the default setting).
-% \begin{macrocode}
- \def\i@ts##1##2{\count0=##2}
- \afterassignment\i@ts\count@\sp@l
- \loop\ifnum\count0>0 \multiply\FC@l by 10
- \expandafter\secd@xt\number\FC@r\secd@xt \advance\count0 by \m@ne
- \repeat
-% \end{macrocode}
-% This also means that if you provided more decimal digits than this, the excess
-% digit(s) will not be handled. This is truncation, not rounding! If all
-% truncated digits are zero, this truncation is a no-op and they are silently ignored,
-% see the example Table~\ref{tab:ex10},
-% \begin{table}[htb]
-% \newcolumntype{d}{D{,}{,}{-1}}
-% \caption{Truncating excess digits.}
-% \label{tab:ex10}
-% \begin{tabular}{@{}ldfl@{}}
-% \toprule
-% composer & \multicolumn1c{raw entry} & \leeg{debt} & remark \\
-% \midrule
-% Berg & 123,450 & 123,450 & ``0'' silently ignored \\
-% Eisler & 234,567 & 234,567 & ``7'' ignored with warning \\
-% Sch\"onberg & 345,6709 & 345,6709 & ``09'' ignored with warning \\
-% Webern & 2,3456 & 2,3456 & ``56'' ignored with warning \\
-% \sumline
-% \bottomrule
-% \end{tabular}
-% \end{table}
-% that was created with |@{}ldfl@{}| (|d| for centering on the decimal
-% mark~\cite{dcolumn}) as tabular key. If, however, at least one of them is
-% not zero, a |\PackageWarning| will be given, showing the excess digit(s).
+%Next is a loop for bringing the decimal part in the correct way to
+%the integer part. This loop is performed the number of decimal
+%digits to be printed (the~2 in |3,2| of the default setting). The higher
+%this number is, e.g., 6~when |3,6| is chosen as |#3|---but who does that?---,
+%the lower the maximum initial value for |\FC@l| can be, before a low-level
+%\TeX\ arithmetic overflow will occur, so don't overdo it.
+% \begin{macrocode}
+\afterassignment\i@ts\count@\sp@l
+ \loop\ifnum\count0>0 \multiply\FC@l by 10 \expandafter\secd@xt
+ \number\FC@r\secd@xt \advance\count0 by \m@ne \repeat
+% \end{macrocode}
+%This also means that if more decimal digits than this are provided, the excess
+%digit(s) will not be handled. This is truncation, not rounding! If all
+%truncated digits are zero, this truncation is exact and they are silently ignored,
+%see the example Table~\ref{tab:ex10},
+%\begin{table}[htb]
+%\newcolumntype{d}{D{,}{,}{-1}}
+%\caption{Truncating excess digits.}
+%\label{tab:ex10}
+%\begin{tabular}{@{}ldfl@{}}
+%\toprule
+%composer & \multicolumn1c{raw entry} & \leeg{debt} & remark \\
+%\midrule
+%Berg & 123,450 & 123,450 & silently ignoring digit ``0'' \\
+%Eisler & 234,563 & 234,563 & warning: digit ``3'' ignored \\
+%Sch\"onberg & 345,6704 & 345,6704 & warning: digits ``04'' ignored \\
+%Webern & 2,3456 & 2,3456 & warning: digits ``56'' ignored, i.e.,\\
+% & & & \quad without rounding this entry to 2,35 \\
+%\sumline
+%\bottomrule
+%\end{tabular}
+%\end{table}
+%that was created with |@{}ldfl@{}| (|d| for centering on the decimal
+%mark~\cite{dcolumn}) as tabular key. If, however, at least one of them is
+%not zero, a |\PackageWarning| will be given, showing the discarded digit(s).
+%There is nothing magical about the constant~19 below: it is simply the
+%concatenation of~|1| and the largest decimal digit. The smallest next value
+%that |\FC@r| can have is~100, so all values from~19 up to and including~99
+%would have worked here.
% \begin{macrocode}
\ifnum\expandafter\@gobble\number\FC@r>0
- \def\tw@l##1##2\relax{##2}
\PackageWarning{fcolumn}{Excess digit\ifnum\FC@r>19 s\fi\space
``\expandafter\tw@l\number\FC@r\relax'' in decimal part
\MessageBreak ignored near or}
\fi
% \end{macrocode}
-% Don't forget to correct for the sign (once this is done, |\FCs@gn|
-% is free again and can and will be used for other purposes).
-% Then output the result.
+%Don't forget to correct for the sign (once this is done, |\FCs@gn|
+%is free again and can and will be used for other purposes).
+%Then output the result.
% \begin{macrocode}
- \ifx\FCs@gn\m@ne\relax \FC@l=-\FC@l \fi
- \zetg@ld{\sp@l}{\FC@l}%
+ \ifx\FCs@gn\m@ne\relax\FC@l=-\FC@l\fi\zetg@ld{\sp@l}{\FC@l}%
\fi}
% \end{macrocode}
-% \end{macro}
-% \begin{macro}{\secd@xt}
-% The second digit from the left is needed from a string of characters
-% representing a decimal number. The macro |\secd@xt| extracts that digit,
-% provided that the number has at least two digits, but that is guaranteed by
-% |\e@fi|. That second digit is then added to |\FC@l|.
-% A new number is assigned to |\FC@r|, that consists
-% of the digits of |#1#3|. If |#3| was empty, 10 is assigned. In this way
-% |\FC@r| is prepared for insertion in the next invocation of |\secd@xt|.
-% In iterating: 1234 yields~134, yields~14, yields~10, stays~10, etc.
-% \begin{macrocode}
-\def\secd@xt#1#2#3\secd@xt{\advance\FC@l by #2
-\FC@r=#1#3 \ifnum\FC@r=1 \FC@r=10 \fi}
-% \end{macrocode}
-% \end{macro}
-% \subsection{Adaptations to existing macros}
-% \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).
-% 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 v2.4k of the |array| package~\cite{array},
-% compacted.
+%\end{macro}
+%\begin{macro}{\i@ts}
+%\begin{macro}{\tw@l}
+%Macro |\e@fi| uses two very tiny macros, given here. In previous versions
+%these were defined inside |\e@fi|, so they also got undefined when |\e@fi|
+%ended. The current solution makes execution a bit faster.
+% \begin{macrocode}
+\def\i@ts#1#2{\count0=#2} \def\tw@l#1#2\relax{#2}
+% \end{macrocode}
+%\end{macro}
+%\end{macro}
+%\begin{macro}{\secd@xt}
+%The second digit from the left is needed from a string of characters
+%representing a decimal number (that should be at least 10 and start with
+%a~|1|, but that is guaranteed by |\e@fi|). The macro |\secd@xt| extracts
+%that digit, which is then added to |\FC@l|. A new number is assigned to
+%|\FC@r|, that consists of the digits of |1#2|, unless |#2| was empty; in
+%that case~10 is assigned. In this way |\FC@r| is prepared for insertion
+%in the next invocation of |\secd@xt|. In iterating: 1234 yields~134,
+%yields~14, yields~10, stays~10, etc.
+% \begin{macrocode}
+\def\secd@xt1#1#2\secd@xt{\advance\FC@l by #1
+ \FC@r=1#2 \ifnum\FC@r=1 \FC@r=10 \fi}
+% \end{macrocode}
+%\end{macro}
+%\subsection{Adaptations to existing macros}
+%\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).
+%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 v2.4k (or later) of the |array| package~\cite{array},
+%compacted.
% \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@}%
% \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.
+%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.
+%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| in the previous chunk followed by |\begingroup|
-% in the next chunk 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}
-\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.
+%Before ending the |\begingroup|, the sumline is created.
+%As a side product of |\@mksumline| also the \meta{count}s for the totals
+%and \meta{dimen}s for the widths of the columns are created. All columns
+%should start fresh, i.e., totals are~0 and widths are 0\thinspace pt.
% \begin{macrocode}
-\res@tsumline
+\@mksumline{#2}\endgroup\res@tsumline
% \end{macrocode}
-% From here on it is just the old definition of |array.sty|.
+%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}
% \end{macrocode}
-% Because |\@array| was changed here and it is this version that should
-% be used, |\@@array| should be |\let| equal to |\@array| again.
+%Because |\@array| was changed here and it is this version that should
+%be used, |\@@array| should be |\let| equal to |\@array| again.
% \begin{macrocode}
\let\@@array=\@array
% \end{macrocode}
-% \end{macro}
-% Much of the techniques here are repeated in |\LT@array|.
-% \subsection{The sumline, close to a postamble}
-% \begin{macro}{\@mksumline}
-% The construction of the sumline is much easier than that of the preamble
-% for several reasons. It may be safely assumed that the preamble specifier
-% is grammatically correct because it has already been screened by |\@mkpream|.
-% Furthermore, most entries will simply add nothing to |\s@ml@ne|, e.g., |@|, |!|,
-% and \verb+|+ can be fully ignored. Ampersands are only inserted by |c|, |l|, |r|,
-% |p|, |m|, and |b|. So, if |\a| is a macro that prints the desired result of the
-% column (see below), then a specifier like |@{}lflf@{}| will yield the sumline
-% |&\a&&\a\\|. 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.
+%\end{macro}
+%Much of the techniques here are repeated in |\LT@array|, see Section~\ref{multipage}.
+%\subsection{The sumline, close to a postamble}
+%\begin{macro}{\@mksumline}
+%The construction of the sumline is much easier than that of the preamble
+%for several reasons. It may be safely assumed that the preamble specifier
+%is grammatically correct because it has already been screened by |\@mkpream|.
+%Furthermore, most entries will simply add nothing to |\s@ml@ne|, e.g., |@|, |!|,
+%and \verb+|+ can be fully ignored. Ampersands are only inserted by |c|, |l|, |r|,
+%|p|, |m|, and |b|. So, if |\a| is a macro that prints the desired result of the
+%column (see below), then a specifier like |@{}lflf@{}| will yield the sumline
+%|&\a&&\a\\|. 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{.}{,}{3,2}{}}r<{\e@fi}@{}|. The rules for constructing the
-% sumline are now very simple:
-% \begin{compactitem}[$\bullet$]
-% \item add an ampersand when |c|, |l|, |r|, |p|, |m|, or |b| is found,
-% unless it is the first one (this is the same as in the preamble);
-% \item add a |\a| when |<{\e@fi}| is found;
-% \item ignore everything else;
-% \item close with a |\\|.
-% \end{compactitem}
-% (For completeness' sake it should be mentioned that prior to the |\\| also the column
-% check is inserted, see |\aut@check|.)
-% To discriminate, a special version of |\@testpach|~\cite{array}
-% could be written, but that is not necessary: |\@testpach| can do all the work,
-% although much of it will be discarded. Here speed is sacrificed for space
-% and this can be afforded because the creation of the sumline is done only
-% once per |\tabular| or |\longtable|.
-%
-% The start is copied from |\@mkpream|.
+%In reality the sumline must be constructed from the expanded form of the
+%specifier, so |@{}lf@{}| will expand as
+%|@{}l>{\b@fi.,{3,2}{}}r<{\e@fi}@{}|. The rules for constructing the
+%sumline are now very simple:
+%\begin{compactitem}[$\bullet$]
+%\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 close with a |\\|.
+%\end{compactitem}
+%(For completeness' sake it should be mentioned that prior to the |\\| also the column
+%check is inserted, see |\aut@check|.)
+%To discriminate, a special version of |\@testpach|~\cite{array}
+%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 of the sumline is done only
+%once per |tabular| or |longtable|.
+%
+%The start is copied from |\@mkpream|.
% \begin{macrocode}
\def\@mksumline#1{\gdef\s@ml@ne{}\@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.
+%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=0
-\let\prr@sult=\relax
+\global\FCsc@l=0 \let\prr@sult=\relax
% \end{macrocode}
-% Then |\@mkpream| is picked up again.
+%Then |\@mkpream| is picked up again.
% \begin{macrocode}
\@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 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.
+%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 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\or\or\@classfii\or
\or\or\or\or\or\or\or\@classfx\fi\@lastchclass\@chclass}%
% \end{macrocode}
-% 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.
+%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\s@ml@ne{\s@ml@ne\noexpand\aut@check\noexpand\\}}
% \end{macrocode}
-% \end{macro}
-% \begin{macro}{\@addtosumline}
-% Macro |\@addtosumline|, as its name already suggests, adds something to the
-% sumline, like its counterpart |\@addtopreamble| did to the preamble.
+%\end{macro}
+%\begin{macro}{\@addtosumline}
+%Macro |\@addtosumline|, as its name already suggests, adds something to the
+%sumline, like its counterpart |\@addtopreamble| did to the preamble.
% \begin{macrocode}
\def\@addtosumline#1{\xdef\s@ml@ne{\s@ml@ne #1}}
% \end{macrocode}
-% \end{macro}
-% \begin{macro}{\@classfx}
-% 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|.
+%\end{macro}
+%\begin{macro}{\@classfx}
+%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~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~f10.
+%\end{macro}
+%\begin{macro}{\@classfz}
+%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~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~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
-% \LaTeX-code using~|<|.
+%\begin{macro}{\@classfii}
+%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
+%\LaTeX-code using~|<|.
% \begin{macrocode}
\def\@classfii{\ifnum\@lastchclass=8
\edef\t@stm{\expandafter\string\@nextchar}
- \edef\t@stn{\string\e@fi}
- \ifx\t@stm\t@stn
+ \edef\t@stn{\string\e@fi} \ifx\t@stm\t@stn
% \end{macrocode}
-% If both tests yield |true|, i.e., we encountered a |<{\e@fi}| where we expect
-% one to find, then add the macro to typeset everything.
+%If both tests yield |true|, i.e., we encountered a |<{\e@fi}| where we expect
+%one to find, then add the macro to typeset everything.
% \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. 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
- \ifnum\FCsc@l>\FCtc@l
-% \end{macrocode}
-% 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 cannot be checked---it is not created by
-% |fcolumn| and an error is given. In case it is a \meta{count} you're just
-% lucky, and you could ignore that error, 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.
+%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 \ifnum\FCsc@l>\FCtc@l
+% \end{macrocode}
+%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 cannot be checked---it is not created by
+%|fcolumn| and an error is given. In case it is a \meta{count} you're just
+%lucky, and you could ignore that error, 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
@@ -1193,14 +1229,14 @@ indention=0pt,labelsep=colon,justification=raggedright}
will get overwritten.}%
\fi
% \end{macrocode}
-% 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|.
+%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
% \end{macrocode}
-% If the creation was successful, the \meta{count} |\FCtc@l| should be
-% increased.
+%If the creation was successful, the \meta{count} |\FCtc@l| should be
+%increased.
% \begin{macrocode}
\global\FCtc@l=\FCsc@l
\else
@@ -1214,40 +1250,41 @@ indention=0pt,labelsep=colon,justification=raggedright}
\fi
\fi}
% \end{macrocode}
-% Once created, it is not necessary to initialise them here because that
-% is done later in one go.
-% \end{macro}
-% \begin{macro}{\sumline}
-% The command for the sumline has one optional argument: the separation
-% between the rule and the total. By default this is 2\thinspace pt, but
-% the user may specify |\sumline[10pt]| if that separation needs to be
-% 10\thinspace pt. The assignment needs to be global, because it is done
-% in the first column of the tabular, but is valid for the whole line.
+%Once created, it is not necessary to initialise them here because that
+%is done later in one go.
+%\end{macro}
+%\begin{macro}{\sumline}
+%The command for the sumline has one optional argument: the separation
+%between the rule and the total. By default this is 2\thinspace pt, but
+%the user may specify |\sumline[10pt]| if that separation needs to be
+%10\thinspace pt. The assignment needs to be global, because it is done
+%in the first column of the tabular, but is valid for the whole line.
% \begin{macrocode}
\newdimen\s@mlinesep
\def\sumline{\@ifnextchar[\s@mline{\s@mline[2pt]}}
\def\s@mline[#1]{\global\s@mlinesep=#1 \s@ml@ne}
% \end{macrocode}
-% In the introduction it was stated that |\sumline| has two options, but in
-% reality that second option is the option to |\\| that is issued by |\s@ml@ne|.
-% \end{macro}
-% \begin{macro}{\prr@sult}
-% The macro |\prr@sult| actually puts the information together. It starts
-% like |\leeg|.
+%In the introduction it was stated that |\sumline| has two options, but in
+%reality that second option is the option to |\\| that is issued by |\s@ml@ne|.
+%\end{macro}
+%\begin{macro}{\prr@sult}
+%The macro |\prr@sult| actually puts the information together. It starts
+%like |\leeg|.
% \begin{macrocode}
\def\prr@sult{$\egroup \let\e@fi=\relax \let\FCform@t=\FCform@tt
% \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 happen
-% that the sum is wider (in points) than any of the entries, e.g., when the
-% result of $6+6$ (using specifier~|3,2|) is typeset. The width of the rule
-% should be equal to the width of |\hbox{$12{,}00$}| then. On the other
-% hand the width of the rule when summing $24$ and $-24$ 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}
+%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 happen
+%that the sum is wider (in points) than any of the entries, e.g., when the
+%result of $6+6$ (using specifier~|3,2|) is typeset. The width of the rule
+%should be equal to the width of |\hbox{$12{,}00$}| then. On the other
+%hand the width of the rule when summing $24$ and $-24$ 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. This excludes the extension to the
+%right in case parentheses are used, again for aesthetic reasons.
+% \begin{macrocode}
\setbox0=\hbox{$\geldm@cro{\sp@l}{\number\csname
FCtot@\romannumeral\FCsc@l\endcsname}$}%
\ifdim\wd0>\csname FCwd@\romannumeral\FCsc@l\endcsname
@@ -1257,144 +1294,149 @@ FCtot@\romannumeral\FCsc@l\endcsname}$}%
\vskip\s@mlinesep
\hbox to \csname FCwd@\romannumeral\FCsc@l\endcsname{\hfil\unhbox0}}}
% \end{macrocode}
-% \end{macro}
-% \subsection{Other checks}
-% \begin{macro}{\leeg}
-% This macro is used to overrule the default behaviour of the pair |\b@fi|
-% and |\e@fi|. It starts with ending the groups 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.
+%\end{macro}
+%\subsection{Other checks}
+%\begin{macro}{\leeg}
+%This macro is used to overrule the default behaviour of the pair |\b@fi|
+%and |\e@fi|. It starts with ending the groups 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 processed in the normal way for a right
+%aligned column.
%
-% 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. By defining |\leeg| in this way, instead of |\multicolumn1r{}|
-% (which contains |\omit|), the default spacing in the column is retained.
-% It doesn't alter the width of the sumrule, but has its normal effect on
-% the column width, so be careful: don't insert the unabridged version of
-% Romeo and Juliet~\cite{RJ} here. It is not typeset in math mode, nor
-% does it use the extra (formatting) information of |#4| of the fcolumn,
-% so you're completely free here.
+%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. By defining |\leeg| in this way, instead of |\multicolumn1r{}|
+%(which contains |\omit|), the default spacing in the column is retained.
+%It doesn't alter the width of the sumrule, but has its normal effect on
+%the column width, so be careful: don't insert the unabridged version of
+%Romeo and Juliet~\cite{RJ} here. It is not typeset in math mode, nor
+%does it use the extra (formatting) information of |#4| of the fcolumn,
+%so you're completely free here.
% \begin{macrocode}
\def\leeg#1{$\egroup \let\e@fi=\relax #1}
% \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 the text~|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?
-% \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. 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 |\\|).
+%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 the text~|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?
+%\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. 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=\z@
\advance\FCsc@l by \m@ne\repeat}
% \end{macrocode}
-% \end{macro}
-% \begin{macro}{\resetsumline}
-% To reset a sumline within a table, it should be done within a |\noalign|.
+%\end{macro}
+%\begin{macro}{\resetsumline}
+%To reset a sumline within a table, it should be done within a |\noalign|.
% \begin{macrocode}
\def\resetsumline{\noalign{\res@tsumline}}
% \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|).
+%\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 be less than |\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.
+%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.
%
-% If the list |\FC@chklist| is empty, the list for the automatic check is
-% generated (which will remain empty if |\FCsc@l| is odd).
+%If the list |\FC@chklist| is empty, the list for the automatic check is
+%generated (which will remain empty if |\FCsc@l| is odd).
% \begin{macrocode}
\def\FC@chklist{}
\def\aut@check{\ifx\@empty\FC@chklist\relax
\ifodd\FCsc@l\else
- \count0=\@ne \count1=\FCsc@l
- \divide\count1 by \tw@
- \loop\ifnum\count1<\FCsc@l
- \advance\count1 by \@ne
+ \count0=\@ne \count1=\FCsc@l \divide\count1 by \tw@
+ \loop\ifnum\count1<\FCsc@l \advance\count1 by \@ne
\xdef\FC@chklist{\FC@chklist\number\count0,\number\count1;}%
- \advance\count0 by\@ne
- \repeat
+ \advance\count0 by\@ne \repeat
\fi
\fi
% \end{macrocode}
-% Then this list is peeled off and processed.
-% \begin{macrocode}
-\loop
-\ifx\FC@chklist\@empty\let\FCs@gn=\@ne\else\let\FCs@gn=\m@ne\fi
-\ifx\FCs@gn\m@ne
- \expandafter\fre@t\FC@chklist\end
- \ifnum\csname FCtot@\romannumeral\count0\endcsname=
- \csname FCtot@\romannumeral\count1\endcsname\else
- \PackageWarning{fcolumn}{F-columns \number\count0 \space and
- \number\count1 \space do not balance due to \string\sumline
- \MessageBreak near or}%
+%Then this list is peeled off and processed.
+%The comparison is done between the internal representation of the totals,
+%so this only makes sense when the decimal part, i.e., the |2| of |3,2| in
+%the default value for |#3| of~|f| is the same for both columns. That is
+%almost always the case due to the nature of this type of tabulars. The
+%\meta{count} |\FC@l| is free to be used for the calculations needed here.
+% \begin{macrocode}
+\loop\ifx\FC@chklist\@empty\else
+ \expandafter\fre@t\FC@chklist\fre@t
+ \FC@l=\csname FCtot@\romannumeral\count0\endcsname
+ \advance\FC@l by -\csname FCtot@\romannumeral\count1\endcsname
+ \ifnum\FC@l=0 \else \ifnum\FC@l<0 \FC@l=-\FC@l \fi
+ \PackageWarning{fcolumn}{Representations of F-columns \number\count0
+ \space and \number\count1 \space differ by\MessageBreak
+ \number\FC@l\space due to \string\sumline\space near or}%
\fi
\repeat}
% \end{macrocode}
-% When |\aut@check| is finished, |\FC@chklist| is empty again, i.e., well
-% prepared for the next time it is used. This also means that the default
-% behaviour kicks in again: if that's not what you want, you should
-% specify the appropriate |\checkfcolumns| lines again.
-% \end{macro}
-% \begin{macro}{\fre@t}
-% This function eats the first two numbers off |\FC@chklist|.
-% \begin{macrocode}
-\def\fre@t#1,#2;#3\end{\count0=#1 \count1=#2 \xdef\FC@chklist{#3}}
-% \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 |F|-column numbers to compare. It is the responsibility of
-% the user to provide the correct numbers here, otherwise bogus output
-% is generated. If this manual check is inserted, the automatic check
-% will not be performed.
+%When |\aut@check| is finished, |\FC@chklist| is empty again, i.e., well
+%prepared for the next time it is used. This also means that the default
+%behaviour kicks in again: if that's not what you want, you should
+%specify the appropriate |\checkfcolumns| lines again.
+%\end{macro}
+%\begin{macro}{\fre@t}
+%This function eats the first two numbers off |\FC@chklist|.
+% \begin{macrocode}
+\def\fre@t#1,#2;#3\fre@t{\count0=#1 \count1=#2 \xdef\FC@chklist{#3}}
+% \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 |F|-column numbers to compare. It is the responsibility of
+%the user to provide the correct numbers here, otherwise bogus output
+%is generated. If this manual check is inserted, the automatic check
+%will not be performed.
% \begin{macrocode}
\def\checkfcolumns#1#2{\noalign{\xdef\FC@chklist{\FC@chklist #1,#2;}}}
% \end{macrocode}
-% \end{macro}
-% \subsection{Support for package longtable}
-% Package |longtable| is used for tables that may span multiple pages. The
-% packages |fcolumn| and |longtable| work together provided |longtable| is
-% loaded first, so that |fcolumn| can adapt one definition of |longtable|. If
-% |longtable| is not loaded, it's a waste of memory having that definition, so
-% it is only defined when necessary. For that reason the user is warned (not
-% loading |longtable| is obviously not an error if you don't use it) if
-% |fcolumn| is loaded without prior loading of |longtable|.
+%\end{macro}
+%\subsection{Support for multipage tables}\label{multipage}
+%Packages |longtable|~\cite{longtable} and |supertabular|~\cite{supertabular}
+%can be used for tables that span multiple pages. Package |supertabular| works
+%with |fcolumn| out of the box (no changes needed), but may lead to different
+%column widths on individual pages. If you don't want that, use |longtable|.
+%The packages |fcolumn| and |longtable| also work together provided |longtable|
+%is loaded first, so that |fcolumn| can adapt one definition of |longtable|.
+%It's a long definition, so if |longtable| is not loaded, it's a waste of memory
+%having it; that's why it is only defined when necessary. For that reason the
+%user is politely warned (not loading |longtable| is obviously not an error if
+%you don't use it) if |fcolumn| is loaded without prior loading of |longtable|.
% \begin{macrocode}
\ifx\longtable\@undefined
- \PackageWarningNoLine{fcolumn}{fcolumn is loaded without longtable.
- That's OK, but\MessageBreak if you want to load longtable as well,
- make sure \MessageBreak it is done before loading fcolumn}\else
+ \PackageWarningNoLine{fcolumn}{fcolumn is loaded without package
+ longtable.\MessageBreak That's perfectly OK, but if you want to
+ load\MessageBreak longtable as well, make sure it is done before
+ \MessageBreak loading fcolumn}\else
% \end{macrocode}
-% \begin{macro}{\LT@array}
-% And here is the only definition of |longtable| that needs to be extended
-% to make |fcolumn| work with that package. The lines are compacted a bit
-% w.r.t.\ the original |longtable| code; if you want to study the
-% code, have a look at the documentation~\cite{longtable}.
+%\begin{macro}{\LT@array}
+%And here is the only definition of |longtable| that needs to be extended
+%to make |fcolumn| work with that package. The lines are compacted a bit
+%w.r.t.\ the original |longtable| code; if you want to study the
+%code, have a look at the documentation~\cite{longtable}.
% \begin{macrocode}
\def\LT@array[#1]#2{\refstepcounter{table}\stepcounter{LT@tables}\if
l#1 \LTleft\z@\LTright\fill\else\if r#1 \LTleft\fill\LTright\z@\else
@@ -1415,46 +1457,51 @@ FCtot@\romannumeral\FCsc@l\endcsname}$}%
\begingroup\@mkpream{#2}\xdef\LT@bchunk{\global\advance\c@LT@chunks
\@ne\global\LT@rows\z@\setbox\z@\vbox\bgroup\LT@setprevdepth\tabskip
\LTleft\noexpand\halign to\hsize\bgroup\tabskip\z@\@arstrut\@preamble
- \tabskip\LTright\cr}\endgroup
+ \tabskip\LTright\cr}
% \end{macrocode}
-% Until this line it was just the code for |\LT@array| from package~|longtable|.
-% The two lines of the next chunk are new to |\LT@array|. Their
-% purpose is the same as in |\@array| above.
+%Until this line it was just the code for |\LT@array| from package~|longtable|.
+%The two lines of the next chunk are new to |\LT@array|. Their
+%purpose is the same as in |\@array| above.
% \begin{macrocode}
- \begingroup\@mksumline{#2}\endgroup\res@tsumline
+ \@mksumline{#2}\endgroup\res@tsumline
\everycr{\noalign{\global\FCsc@l=0 }}%
% \end{macrocode}
-% From here on |\LT@array| is picked up again, ending with the |\fi| that belongs
-% to the |\ifx| that started this definition.
+%From here on |\LT@array| is picked up again, ending with the |\fi| that belongs
+%to the |\ifx| that started this definition.
% \begin{macrocode}
\expandafter\LT@nofcols\LT@bchunk&\LT@nofcols\LT@make@row\m@th\let
\par\@empty\lineskip\z@\baselineskip\z@\LT@bchunk}
\fi
% \end{macrocode}
% \end{macro}
-% That's it!
-% \section*{Acknowledgement}
-% Thanks to Karl Berry for valuable comments regarding the consistency of the
-% installation procedure of this version. Frank Mittelbach gave various useful
-% suggestions for improving the input parsing as well as hints to make the package
-% more \LaTeX-like. He also challenged me to make |fcolumn| compatible with
-% |longtable|. Christian Hoff's request on column formatting triggered many
-% happy hours of coding.
-% \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. The |booktabs| package. Publication quality tables in \LaTeX.
-% \bibitem{dcolumn} David Carlisle. The |dcolumn| package.
-% \bibitem{longtable} David Carlisle. The |longtable| package.
-% \bibitem{world_economy} According to the IMF \url{www.imf.org}.
-% \bibitem{CandTB} Donald Knuth, Computers \& Typesetting/B, \TeX: the program.
-% \bibitem{CandTA} Donald Knuth, Computers \& Typesetting/A, The \TeX book.
-% \bibitem{TAOCP1} Donald Knuth, The Art of Computers Programming, volume~1.
-% \bibitem{RJ} William Shakespeare, Romeo and Juliet, a tragedy (1597).
-% \end{thebibliography}
-% \Finale
+%That's it!
+%
+%\section*{Acknowledgement}
+%Thanks to Karl Berry for valuable comments regarding the consistency of the
+%installation procedure of this version. Frank Mittelbach gave various useful
+%suggestions for improving the input parsing as well as hints to make the package
+%more \LaTeX-like. He also challenged me to make |fcolumn| compatible with
+%|longtable|. Christian Hoff's request on column formatting triggered many
+%happy hours of coding.
+%\bibliographystyle{plain}
+%\begin{thebibliography}{10}
+%\itemsep0pt \parsep0pt \parskip0pt
+%\bibitem{array} Frank Mittelbach and David Carlisle.
+% A new implementation of \LaTeX's |tabular| and |array| environment.
+%\bibitem{booktabs} Simon Fear. The |booktabs| package. Publication quality tables in \LaTeX.
+%\bibitem{dcolumn} David Carlisle. The |dcolumn| package.
+%\bibitem{longtable} David Carlisle. The |longtable| package.
+%\bibitem{supertabular} Theo
+% Jurriens \iffalse I checked with Theo that `Jurriens' is correct; not `Jurri\"ens'. \fi
+% and Johannes Braams. The |supertabular| package.
+%\bibitem{color} David Carlisle. The |color| package.
+%\bibitem{CandTB} Donald Knuth, {\it Computers \& Typesetting/B, ``\TeX: the program,''\/}
+% Addison-Wesley, Reading (1991).
+%\bibitem{CandTA} Donald Knuth, {\it Computers \& Typesetting/A, ``The \TeX book,''\/}
+% Addison-Wesley, Reading (1991).
+%\bibitem{TAOCP1} Donald Knuth, {\it The Art of Computers Programming, volume~1,
+% ``Fundamental Algorithms,''\/} Addison-Wesley, Reading (1997).
+%\bibitem{RJ} William Shakespeare, Romeo and Juliet, a tragedy (1597).
+%\end{thebibliography}
+%\Finale
\endinput
diff --git a/macros/latex/contrib/fcolumn/fcolumn.ins b/macros/latex/contrib/fcolumn/fcolumn.ins
index d4074c1507..a759d441c6 100644
--- a/macros/latex/contrib/fcolumn/fcolumn.ins
+++ b/macros/latex/contrib/fcolumn/fcolumn.ins
@@ -15,9 +15,7 @@
\keepsilent
\preamble
-This is a generated file.
-
-Copyright (C) 2015--2022 by Edgar Olthof
+Copyright (C) 2015--2023 by Edgar Olthof
This file may be distributed and/or modified under the
conditions of the LaTeX Project Public License, either
@@ -32,26 +30,24 @@ LaTeX version 1999/12/01 or later.
\askforoverwritefalse
\usedir{tex/latex/tools}
\generate{\file{fcolumn.sty}{\from{fcolumn.dtx}{package}}}
-
\obeyspaces
\Msg{*******************************************************}
\Msg{* *}
-\Msg{* To finish the installation you have to move the *}
-\Msg{* following file *}
+\Msg{* To finish the installation you should move the file *}
\Msg{* *}
-\Msg{* fcolumn.sty *}
+\Msg{* fcolumn.sty *}
\Msg{* *}
\Msg{* into a directory searched by TeX. *}
\Msg{* To generate the documentation, type ``make'' or *}
\Msg{* run the file fcolumn.dtx through LaTeX. Make sure *}
\Msg{* the glossary and index are updated by makeindex: *}
\Msg{* *}
-\Msg{* makeindex -s gglo.ist -o fcolumn.gls fcolumn.glo *}
-\Msg{* makeindex -s gind.ist -o fcolumn.ind fcolumn.idx *}
+\Msg{* makeindex -s gglo.ist -o fcolumn.gls fcolumn.glo *}
+\Msg{* makeindex -s gind.ist -o fcolumn.ind fcolumn.idx *}
\Msg{* *}
\Msg{* and then LaTeX again; presto! *}
\Msg{* *}
-\Msg{* Happy TeXing! *}
+\Msg{* Happy TeXing! *}
\Msg{* *}
\Msg{*******************************************************}
\endbatchfile
diff --git a/macros/latex/contrib/fcolumn/fcolumn.pdf b/macros/latex/contrib/fcolumn/fcolumn.pdf
index e6755b2495..e3bdff4822 100644
--- a/macros/latex/contrib/fcolumn/fcolumn.pdf
+++ b/macros/latex/contrib/fcolumn/fcolumn.pdf
Binary files differ
diff --git a/macros/latex/contrib/fcolumn/makefile b/macros/latex/contrib/fcolumn/makefile
index 6a05537aa6..58b140ea85 100644
--- a/macros/latex/contrib/fcolumn/makefile
+++ b/macros/latex/contrib/fcolumn/makefile
@@ -1,4 +1,4 @@
-.SUFFIXES: .dtx .glo .gls .idx .ind .ins .pdf .sty
+.SUFFIXES: .dtx .glo .gls .idx .ind .ins .pdf .sty .tex
DIST=fcolumn.dtx fcolumn.ins fcolumn.pdf makefile README
.ins.sty:
@@ -19,13 +19,16 @@ DIST=fcolumn.dtx fcolumn.ins fcolumn.pdf makefile README
.dtx.pdf:
pdflatex $<
-all: fcolumn.sty fcolumn.pdf
+.tex.pdf:
+ pdflatex $<
+
+all: fcolumn.pdf
install:
pdftex fcolumn.ins
dist: $(DIST)
- mkdir FCv1.4.1temp; cp $(DIST) FCv1.4.1temp; tar -czf fcolumn.tgz FCv1.4.1temp; rm -r FCv1.4.1temp
+ mkdir FCv1.4.2temp ; cp $(DIST) FCv1.4.2temp/ ; tar -czf fcolumn.tgz FCv1.4.2temp/ ; rm -r FCv1.4.2temp/
fcolumn.pdf: fcolumn.dtx fcolumn.gls fcolumn.ind fcolumn.sty
fcolumn.glo fcolumn.idx: fcolumn.dtx
diff --git a/macros/latex/contrib/siunitx/CHANGELOG.md b/macros/latex/contrib/siunitx/CHANGELOG.md
index a471996dcf..cbd19e3d0d 100644
--- a/macros/latex/contrib/siunitx/CHANGELOG.md
+++ b/macros/latex/contrib/siunitx/CHANGELOG.md
@@ -7,6 +7,13 @@ Changelog](https://keepachangelog.com/en/1.0.0/), and this project adheres to
## [Unreleased]
+## [v3.2.8] - 2023-07-25
+
+### Fixed
+- Suppression of leading `1` when `per-mode = symbol` is used with
+ non-repeated units ranges (see issue
+ [\#677](https://github.com/josephwright/siunitx/issues/677))
+
## [v3.2.7] - 2023-07-20
### Fixed
@@ -1914,7 +1921,8 @@ Changelog](https://keepachangelog.com/en/1.0.0/), and this project adheres to
### Added
- First public testing release (as `si`)
-[Unreleased]: https://github.com/josephwright/siunitx/compare/v3.2.7...HEAD
+[Unreleased]: https://github.com/josephwright/siunitx/compare/v3.2.8...HEAD
+[v3.2.8]: https://github.com/josephwright/siunitx/compare/v3.2.7...v3.2.8
[v3.2.7]: https://github.com/josephwright/siunitx/compare/v3.2.6...v3.2.7
[v3.2.6]: https://github.com/josephwright/siunitx/compare/v3.2.5...v3.2.6
[v3.2.5]: https://github.com/josephwright/siunitx/compare/v3.2.4...v3.2.5
diff --git a/macros/latex/contrib/siunitx/siunitx-code.pdf b/macros/latex/contrib/siunitx/siunitx-code.pdf
index ad5229ba63..bdd6dfe01a 100644
--- a/macros/latex/contrib/siunitx/siunitx-code.pdf
+++ b/macros/latex/contrib/siunitx/siunitx-code.pdf
Binary files differ
diff --git a/macros/latex/contrib/siunitx/siunitx-compound.dtx b/macros/latex/contrib/siunitx/siunitx-compound.dtx
index c4d2248845..51701c6915 100644
--- a/macros/latex/contrib/siunitx/siunitx-compound.dtx
+++ b/macros/latex/contrib/siunitx/siunitx-compound.dtx
@@ -337,7 +337,6 @@
% \l_@@_exp_bracket_bool ,
% \l_@@_exp_combine_bool ,
% \l_@@_unit_bracket_bool ,
-% \l_@@_unit_repeat_bool ,
% \l_@@_unit_repeat_bool
% }
% List options.
@@ -667,7 +666,9 @@
%
% \begin{macro}{\siunitx_compound_quantity:nn}
% For quantities, life is more complex as there are interactions between the
-% options for exponents and units.
+% options for exponents and units. The \cs{c_empty_tl} suppresses any
+% leading \texttt{1} if the unit starts \cs{per}; the \pkg{siunitx-print}
+% module will skip any printing in this case.
% \begin{macrocode}
\cs_new_protected:Npn \siunitx_compound_quantity:nn #1#2
{
@@ -698,7 +699,7 @@
\siunitx_print_number:V \l_@@_bracket_close_tl
}
{ \@@_print:N \siunitx_print_number:x }
- \@@_print_quantity:n { }
+ \@@_print_quantity:n { \c_empty_tl }
}
\group_end:
}
diff --git a/macros/latex/contrib/siunitx/siunitx-number.dtx b/macros/latex/contrib/siunitx/siunitx-number.dtx
index 0b51a9cd1c..8bc0a2dd71 100644
--- a/macros/latex/contrib/siunitx/siunitx-number.dtx
+++ b/macros/latex/contrib/siunitx/siunitx-number.dtx
@@ -2619,7 +2619,8 @@
% }
% \begin{macro}[EXP]{\@@_round_auxiv:nnN, \@@_round_auxv:nnN}
% \begin{macro}[EXP]{\@@_round_auxvi:nN}
-% \begin{macro}[EXP]{\@@_round_auxvii:nnN, \@@_round_auxviii:nnN}
+% \begin{macro}[EXP]{\@@_round_auxvii:nnTF}
+% \begin{macro}[EXP]{\@@_round_auxviii:nnN, \@@_round_auxix:nnN}
% \begin{macro}[EXP]{\@@_round_final_integer:nnw, \@@_round_final_decimal:nnw}
% \begin{macro}[EXP]{\@@_round_final_signicant:n}
% \begin{macro}[EXP]{\@@_round_final_signicant:N}
@@ -2731,37 +2732,32 @@
\@@_round_auxvi:nnN
{#1} { } #2 \q_recursion_tail \q_recursion_stop
}
- \bool_lazy_or:nnTF
- { \int_compare_p:nNn { 0 \tl_head:n {#1} } < 5 }
- {
- \bool_lazy_all_p:n
- {
- { \l_@@_round_half_even_bool }
- { ! \int_if_odd_p:n {#3} }
- { \@@_round_if_half_p:n {#1} }
- }
- }
+ \@@_round_auxvii:nnTF {#1} {#3}
{ \@@_round_final_decimal:nnw }
- { \@@_round_auxvii:nnN }
+ { \@@_round_auxviii:nnN }
{#2} { } #3
}
\cs_new:Npn \@@_round_auxvi:nnN #1#2#3
{
\quark_if_recursion_tail_stop_do:Nn #3
{ { 0 } { } }
+ \@@_round_auxvii:nnTF {#1} {#3}
+ { \@@_round_final_integer:nnw }
+ { \@@_round_auxix:nnN }
+ { } {#2} #3
+ }
+\cs_new:Npn \@@_round_auxvii:nnTF #1#2
+ {
\bool_lazy_or:nnTF
{ \int_compare_p:nNn { 0 \tl_head:n {#1} } < 5 }
{
\bool_lazy_all_p:n
{
{ \l_@@_round_half_even_bool }
- { ! \int_if_odd_p:n {#3} }
+ { ! \int_if_odd_p:n {#2} }
{ \@@_round_if_half_p:n {#1} }
}
}
- { \@@_round_final_integer:nnw }
- { \@@_round_auxviii:nnN }
- { } {#2} #3
}
% \end{macrocode}
% The main rounding routines. These are only every called when there is
@@ -2771,7 +2767,7 @@
% where a zero is rounded up: that automatically means a need to trim
% the other end.
% \begin{macrocode}
-\cs_new:Npn \@@_round_auxvii:nnN #1#2#3
+\cs_new:Npn \@@_round_auxviii:nnN #1#2#3
{
\quark_if_recursion_tail_stop_do:Nn #3
{
@@ -2782,12 +2778,12 @@
{ \@@_round_truncate:n {#2} }
}
{
- \@@_round_auxviii:nnN {#2} { } #1
+ \@@_round_auxix:nnN {#2} { } #1
\q_recursion_tail \q_recursion_stop
}
}
\int_compare:nNnTF {#3} = 9
- { \@@_round_auxvii:nnN {#1} { 0 #2 } }
+ { \@@_round_auxviii:nnN {#1} { 0 #2 } }
{
\int_compare:nNnTF {#3} = 0
{
@@ -2810,7 +2806,7 @@
}
}
}
-\cs_new:Npn \@@_round_auxviii:nnN #1#2#3
+\cs_new:Npn \@@_round_auxix:nnN #1#2#3
{
\quark_if_recursion_tail_stop_do:Nn #3
{
@@ -2830,7 +2826,7 @@
}
}
\int_compare:nNnTF {#3} = 9
- { \@@_round_auxviii:nnN {#1} { 0 #2 } }
+ { \@@_round_auxix:nnN {#1} { 0 #2 } }
{
\@@_round_final:fn
{ \int_eval:n { #3 + 1 } }
@@ -2985,6 +2981,7 @@
% \end{macro}
% \end{macro}
% \end{macro}
+% \end{macro}
%
% \begin{macro}[EXP]{\@@_round_if_half_p:n}
% \begin{macro}[EXP]{\@@_round_if_half:N}
diff --git a/macros/latex/contrib/siunitx/siunitx-print.dtx b/macros/latex/contrib/siunitx/siunitx-print.dtx
index 0e8277cd93..8654060651 100644
--- a/macros/latex/contrib/siunitx/siunitx-print.dtx
+++ b/macros/latex/contrib/siunitx/siunitx-print.dtx
@@ -106,7 +106,8 @@
% roman font, and |^| and |_| will typically be used to indicate super- and
% subscripts, respectively. These elements will be correctly handled when
% printing for example using |\mathsf| in math mode, or using only text
-% fonts.
+% fonts. No printing takes place if the \cs{material} is entirely empty
+% after a single expansion.
% \end{function}
%
% \begin{function}
@@ -120,7 +121,8 @@
% with a fixed text or math mode output. The printing does \emph{not} set
% color (which is managed on a |unit|/|number| basis), but otherwise sets
% the font as described above. The |match| function uses either the prevailing
-% math or text mode.
+% math or text mode. No printing takes place if the \cs{material} is entirely
+% empty after a single expansion.
% \end{function}
%
% \subsection{Key--value options}
@@ -449,7 +451,10 @@
% }
% \begin{macro}{\@@_aux:nn}
% The main printing function doesn't actually need to do very much: just set
-% the color and select the correct sub-function.
+% the color and select the correct sub-function. The \cs{tl_if_empty:oF} is
+% needed to allow a leading \texttt{1} to be omitted in the case of a unit
+% starting \cs{per} in cases where we can't simply use an entirely empty
+% value: see the \pkg{siunitx-compound} submodule.
% \begin{macrocode}
\cs_new_protected:Npn \siunitx_print_number:n #1
{ \@@_aux:nn { number } {#1} }
@@ -459,17 +464,20 @@
\cs_generate_variant:Nn \siunitx_print_unit:n { V , x }
\cs_new_protected:Npn \@@_aux:nn #1#2
{
- \tl_if_empty:cTF { l_@@_ #1 _color_tl }
- { \use:n }
- { \exp_args:Nv \textcolor { l_@@_ #1 _color_tl } }
- {
- \use:c
+ \tl_if_empty:oF {#2}
+ {
+ \tl_if_empty:cTF { l_@@_ #1 _color_tl }
+ { \use:n }
+ { \exp_args:Nv \textcolor { l_@@_ #1 _color_tl } }
{
- siunitx_print_
- \tl_use:c { l_@@_ #1 _mode_tl } :n
+ \use:c
+ {
+ siunitx_print_
+ \tl_use:c { l_@@_ #1 _mode_tl } :n
+ }
+ {#2}
}
- {#2}
- }
+ }
}
% \end{macrocode}
% \end{macro}
diff --git a/macros/latex/contrib/siunitx/siunitx.dtx b/macros/latex/contrib/siunitx/siunitx.dtx
index 31b3e79181..75b83bf182 100644
--- a/macros/latex/contrib/siunitx/siunitx.dtx
+++ b/macros/latex/contrib/siunitx/siunitx.dtx
@@ -122,7 +122,7 @@
%
% Identify the package and give the over all version information.
% \begin{macrocode}
-\ProvidesExplPackage {siunitx} {2023-07-20} {3.2.7}
+\ProvidesExplPackage {siunitx} {2023-07-25} {3.2.8}
{A comprehensive (SI) units package}
% \end{macrocode}
%
diff --git a/macros/latex/contrib/siunitx/siunitx.pdf b/macros/latex/contrib/siunitx/siunitx.pdf
index c89c21174a..3d9dd8b90c 100644
--- a/macros/latex/contrib/siunitx/siunitx.pdf
+++ b/macros/latex/contrib/siunitx/siunitx.pdf
Binary files differ