fcolumn (12mar15)

git-svn-id: svn://tug.org/texlive/trunk@36499 c570f23f-e606-0410-a88d-b1316a301751

7 files changed, 933 insertions, 1 deletions

diff --git a/Master/texmf-dist/doc/latex/fcolumn/README b/Master/texmf-dist/doc/latex/fcolumn/README
new file mode 100644
index 00000000000..27e64940f25
--- /dev/null
+++ b/Master/texmf-dist/doc/latex/fcolumn/README
@@ -0,0 +1,48 @@
+The LaTeX-package ``fcolumn''
+
+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
+such columns, including the sum line (preceded by a rule of the correct
+width), using the specifier ``f''.
+
+The column specifier f itself is rather simple.  It is the predefined
+version of a generic column F.  The generic version expects three 
+arguments: |#1| is the separator, |#2| is the decimal mark, and |#3| 
+the coding used for grouping digits of the integer part and decimal
+part.  For example the f-column in the current version of the package
+is using 3.2 as #3, meaning that numbers are typeset with 2 decimal 
+digits and grouping is done on 3 digits.  The decimal separator is 
+``,'' and the grouping character is ``.'', thereby coding for the 
+continental European standard.  People in the Anglo-saxon world would
+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.
+
+properties    31 dec 2014   debts         31 dec 2014
+-----------------------------------------------------
+house          200.000,00   equity capital  50.000,00
+bank account      -603,23   mortgage       150.000,00
+savings         28.000,00                           
+cash               145,85   profit          27.542,62
+               ----------                  ----------
+               227.542,62                  227.542,62
+
+The core part of the table is input as 
+\begin{table}
+\begin{tabular}{lflf}
+house    &      20000000 &  equity capital &  5000000\\
+bank account &    -60323 &  mortgage     &   15000000\\
+savings     &    2800000\\
+cash         &     14585 &  profit         &  2754262\\
+\sumline
+\end{tabular}
+\end{table}
+
+The column specifier ``f'' (for ``finance'') is able to do the 
+typesetting.  It constructs the sum line, typesets the numbers,
+calculates the totals, and checks whether the two columns are in
+balance; if not, the user is warned via a message. A special command
+\resetsumline exists which alllows one to restart a table, so that
+multiple tables in one tabular environment are possible.
diff --git a/Master/texmf-dist/doc/latex/fcolumn/fcolumn.pdf b/Master/texmf-dist/doc/latex/fcolumn/fcolumn.pdf
new file mode 100644
index 00000000000..33c18d4202e
--- /dev/null
+++ b/Master/texmf-dist/doc/latex/fcolumn/fcolumn.pdf
diff --git a/Master/texmf-dist/source/latex/fcolumn/fcolumn.dtx b/Master/texmf-dist/source/latex/fcolumn/fcolumn.dtx
new file mode 100644
index 00000000000..3e83646dfba
--- /dev/null
+++ b/Master/texmf-dist/source/latex/fcolumn/fcolumn.dtx
@@ -0,0 +1,694 @@
+% \iffalse meta-comment
+%
+% Copyright (C) 2015 by Edgar Olthof
+%
+% This file may be distributed and/or modified under the
+% conditions of the LaTeX Project Public License, either
+% version 1.2 of this license or (at your option) any later
+% version.  The latest version of this license is in
+%
+%    http://www.latex-project.org/lppl.txt
+%
+% and version 1.2 or later is part of all distributions of
+% LaTeX version 1999/12/01 or later.
+%
+% \fi
+% \iffalse
+%<package>\NeedsTeXFormat{LaTeX2e}
+%<package>\ProvidesPackage{fcolumn}
+%<package>          [2015/03/07 v1.0 automatic currency package]
+%<package>\RequirePackage{array}
+%<*batchfile>
+\begingroup
+%%
+%% Copyright (C) 2015 by Edgar Olthof
+%%
+%% This file may be distributed and/or modified under the
+%% conditions of the LaTeX Project Public License, either
+%% version 1.2 of this license or (at your option) any later
+%% version.  The latest version of this license is in
+%%
+%%    http://www.latex-project.org/lppl.txt
+%%
+%% and version 1.2 or later is part of all distributions of
+%% LaTeX version 1999/12/01 or later.
+%%
+\input docstrip.tex
+\keepsilent
+\preamble
+
+This is a generated file.
+
+Copyright (C) 2015 by Edgar Olthof
+
+This file may be distributed and/or modified under the
+conditions of the LaTeX Project Public License, either
+version 1.2 of this license or (at your option) any later
+version.  The latest version of this license is in
+   
+   http://www.latex-project.org/lppl.txt
+
+and version 1.2 or later is part of all distributions of
+LaTeX version 1999/12/01 or later.
+\endpreamble
+\askforoverwritefalse
+\generate{\file{fcolumn.sty}{\from{fcolumn.dtx}{package}}}
+\endgroup
+%</batchfile>
+%<*driver>
+\documentclass{ltxdoc}
+\usepackage{fcolumn}
+\usepackage{booktabs}
+\usepackage{paralist}
+\EnableCrossrefs
+\CodelineIndex
+\RecordChanges
+\begin{document}
+   \DocInput{fcolumn.dtx}
+\end{document}
+%</driver>
+% \fi
+%
+% \CheckSum{0}
+%
+% \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}
+% \GetFileInfo{fcolumn.sty}
+% \DoNotIndex{\fi,\let,\@ne}
+% \title{The \textsf{fcolumn} package\thanks{This file has version
+% number \fileversion, last revised \filedate.}}
+% \author{Edgar Olthof}
+% \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
+% such columns, including the sum line (preceded by a rule of the correct
+% width), using the specifier |f|.
+%
+% The column specifier |f| itself is rather simple.  It is the predefined version
+% of a generic column |F|.  The generic version expects three arguments:
+% |#1| is the separator, |#2| is the decimal mark, and |#3| the coding used for
+% grouping digits of the integer part and decimal part.
+%
+% The f-column in the current version of the package is defined as
+% |\newcolumntype{f}{F{.}{,}{3,2}}| (which is the continental European standard, implemented here)
+% and people in the Anglo-saxon world would rather code |\newcolumntype{f}{F{,}{.}{3,2}}|.
+% The default value for |#3| is 3,2, indicating that grouping of the integer part
+% is by three digits and that the decimal part consists of two digits. 
+% If however, in your country or company grouping is done with a thinspace every 
+% four digits and there are three digits after the decimal mark---that happens to 
+% be a |\cdot|---, then simply specify |\newcolumntype{f}{F{\,}{\cdot}{4,3}}| in 
+% that case.  By default two digits are used for the decimal part, so if you really
+% want no decimal digits (in that case of course also skipping the decimal mark) 
+% you have to explicitly specify |x,0|.
+% \end{abstract}
+%
+% \section{Introduction}
+% To show where and how the f-column is used, let's look at a typical financial table.
+%
+% \begin{table}[htb]
+% \begin{tabular}{@{}lflf@{}}
+% \multicolumn4c{\bfseries Balance sheet}\\
+% \multicolumn1{@{}l}{properties}&\multicolumn1r{31 dec 2014}
+% &\multicolumn1l{debts}&\multicolumn1{l@{}}{31 dec 2014}\\
+% \hline
+% house &     20000000 & equity capital&  5000000\\
+% bank account& -60323 & mortgage &      15000000\\
+% savings&     2800000 &         & \leeg         \\
+% cash&          14585 & profit  &        2754262\\
+% \sumline
+% \check12
+% \hline
+% \end{tabular}
+% \end{table}
+%
+% This table was entered as:
+% \begin{verbatim}
+% \begin{table}[htb]
+% \begin{tabular}{@{}lflf@{}}
+% \multicolumn4c{\bfseries Balance sheet}\\
+% \multicolumn1{@{}l}{properties}&\multicolumn1r{31 dec 2014}
+% &\multicolumn1l{debts}&\multicolumn1{l@{}}{31 dec 2014}\\
+% \hline
+% house &     20000000 & equity capital&  5000000\\
+% bank account& -60323 & mortgage &      15000000\\
+% savings&     2800000 &         & \leeg         \\
+% cash&          14585 & profit  &        2754262\\
+% \sumline
+% \check12
+% \hline
+% \end{tabular}
+% \end{table}
+% \end{verbatim}
+% All the work was done by the column specifier ``f'' (for ``finance'').  In this case
+% it constructs the |\sumline|, typesets the numbers, calculates the totals,
+% and checks whether the two columns are in balance; if not, the user is warned
+% via a |\message|.  Of course for nice settings one should use the
+% |booktabs| package, but that is not the point here.
+%
+% This package is heavily inspired by the |dcolumn| package by David Carlisle,
+% some constructions are more or less copied from that package.
+%
+% There is at least one things to improve in this version:
+% An empty entry in an f-column should result in an empty spot in the table.
+% Now an empty entry is not correct and is repaired by \LaTeX\ to be equal to
+% 0, printing a number in the table.  This can be prevented by explicitly specifying 
+% |\leeg|, like in the table above, but that should be automated.
+%
+% \section{The macros}
+% \begin{macro}{column F}
+% \begin{macro}{column f}
+%
+% Note that the definition of the column type |f| does not use private macros 
+% (no |@|), so overriding its definition is easy for a user.
+%    \begin{macrocode}
+\newcolumntype{F}[3]{>{\b@fi{#1}{#2}{#3}}r<{\e@fi}}
+\newcolumntype{f}{F{.}{,}{3,2}}
+%    \end{macrocode}
+% \end{macro}
+% \end{macro}
+% \begin{macro}{\FCsc@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@l0
+\newcount\FCtc@l \FCtc@l0
+%    \end{macrocode}
+% \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, kopecks, groszy) and typesets it as the amount in entire 
+% currency units (dollars, euros, guldens, Marke, ruble, z\l oty)
+% with comma as decimal separator and the dot as thousand separator.
+% As explained, this can be changed. It uses a private boolean |\withs@p|
+% and an accessable, i.e., without |@| boolean: |\strictaccounting|.  The latter  
+% is used to typeset negative numbers between parentheses.  By default it doesn't
+% do this: a minus sign is used.  It also uses two private \meta{count}s. I'm still
+% figuring out whether this is needed; everything is done local within a column
+% of an |\ialign|, so using |\count0|, |\count1|, etc.\ might also work: it would
+% safe two \meta{count}s.
+%    \begin{macrocode}
+\newcount\g@lda
+\newcount\g@ldb
+\newif\ifwiths@p
+\newif\ifstrictaccounting \strictaccountingfalse
+%    \end{macrocode}
+% Actually |\geldm@cro| is only a wrapper around |\g@ldm@cro|.
+%    \begin{macrocode}
+\def\geldm@cro#1#2{\withs@pfalse 
+ \afterassignment\g@ldm@cro\count@#2\relax{#1}}
+%    \end{macrocode}
+% \end{macro}
+% \begin{macro}{\g@ldm@cro}
+% This macro starts by looking at the sign of |#2|: if it is negative, it prints the correct
+% indicator (a parenthesis or a minus sign), assigns the absolute value of |#2| to |\g@lda| 
+% and goes on.  Note that |\geldm@cro| and therefore |\g@ldm@cro| are used within |$|s, so 
+% it is really a minus sign that is printed, not a hyphen.
+%    \begin{macrocode}
+\def\g@ldm@cro#1\relax#2{%
+ \ifnum#2<0 \ifstrictaccounting(\else-\fi\g@lda=-#2 \else\g@lda=#2 \fi
+%    \end{macrocode}
+% Calculate the entire currency units: this is the result of $x/a$ as integer
+% division, with $a=10^n$ and $n$ the part of |#1| after the separator (if any).
+% Here the first character of |#1| is discarded, so the separator in |#1| is not strict:
+% you could also specify |3.2| instead of |3,2| (or even |3p2|).
+%   \begin{macrocode}
+ \ifx\relax#1\relax\g@ldb=2 \else \g@ldb\@gobble#1\relax\fi
+ \loop
+  \ifnum\g@ldb>0 \divide\g@lda by 10 \advance\g@ldb by \m@ne
+ \repeat
+%    \end{macrocode}
+% The value in |\g@lda| is then output by |\g@ldens| using the separation given.
+%    \begin{macrocode}
+ \g@ldens{\the\count@}%
+%    \end{macrocode}
+% If there is a decimal part\dots
+%    \begin{macrocode}
+ \ifx\relax#1\relax\g@ldb=2 \else \g@ldb\@gobble#1\relax\fi
+ \ifnum\g@ldb>0\decim@lmark
+%    \end{macrocode}
+% Next the decimal part is dealt with. Now $x\bmod a$ is calculated in the 
+% usual way: $x-(x/a)*a$.  The minus sign necessary for this calculation is 
+% introduced in the next line by changing the comparison to |>| instead of |<|.
+%    \begin{macrocode}
+  \ifnum#2>0 \g@lda=-#2\else\g@lda=#2 \fi
+  \loop
+   \ifnum\g@ldb>0 \divide\g@lda by 10 \advance\g@ldb by \m@ne
+  \repeat
+  \ifx\relax#1\relax\g@ldb=2 \else \g@ldb\@gobble#1\relax\fi
+  \loop
+   \ifnum\g@ldb>0 \multiply\g@lda by 10 \advance\g@ldb by \m@ne
+  \repeat
+  \ifnum#2>0 \advance\g@lda by #2
+  \else \advance\g@lda by -#2
+  \fi
+  \ifx\relax#1\relax\g@ldb=2 \else \g@ldb\@gobble#1\relax\fi
+  \zerop@d{\number\g@ldb}{\number\g@lda}%
+ \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 
+% |\strictaccountingfalse|.  Change if you like.
+%    \begin{macrocode}
+ \ifnum#2<0 \ifstrictaccounting\rlap{)}\fi\fi}
+%    \end{macrocode}
+% \end{macro}
+% \begin{macro}{\g@ldens}
+% Here the whole currency units are dealt with.  The macro |\g@ldens| is used
+% recursively, therefore the double braces; this allows to use |\count0| locally.
+% Tail recursion is not possible here (at least I don't know of a way), but that 
+% is not very important, as the largest number ($2^{31}-1$) will only cause a 
+% threefold recursion.  This also implies that the largest amount this package
+% can deal with is 2.147.483.647 (using |x.0|). For most people this is probably
+% more than enough if the currency is euros or dollars.  If not, then it is 
+% more likely that you are spending your time on the beach than studying this package.
+% And otherwise make clear that you use a currency unit of k\$.
+%    \begin{macrocode}
+\def\g@ldens#1{{\g@ldb=\g@lda \count0=#1
+%    \end{macrocode}
+% First divide by $10^n$, where $n$ is |#1|.
+%    \begin{macrocode}
+ \loop
+  \ifnum\count0>0
+  \divide\g@lda by 10
+  \advance\count0 by \m@ne
+ \repeat
+%    \end{macrocode}
+% Here is the recursive part,
+%    \begin{macrocode}
+ \ifnum\g@lda>0 \g@ldens{#1}\fi
+%    \end{macrocode}
+% and then reconstruct the rest of the number.
+%    \begin{macrocode}
+ \count0=#1
+ \loop
+  \ifnum\count0>0
+  \multiply\g@lda by 10
+  \advance\count0 by \m@ne
+ \repeat
+ \g@lda=-\g@lda
+ \advance\g@lda by \g@ldb \du@zendprint{#1}}}
+%    \end{macrocode}
+% \end{macro}
+% \begin{macro}{\du@zendprint}
+% The macro |\du@zendprint| takes care for correctly printing the separator
+% and possible trailing zeros.
+%    \begin{macrocode}
+\def\du@zendprint#1{\ifwiths@p\sep@rator\zerop@d{#1}{\number\g@lda}%
+ \else\zerop@d{1}{\number\g@lda}\fi
+ \global\withs@ptrue}
+%    \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.
+% Initially this was a nice macro using tail recursion until it was
+% found that the running time of that macro was proportional to $n^2$,
+% where $n$ is roughly the number of zeros to be padded. The worst
+% case situation is when printing ``0''.  The running time of the current version
+% is only linear in $n$.
+%
+% 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.
+%    \begin{macrocode}
+ \loop
+  \divide \count1 by 10
+  \ifnum\count1>0 \advance\count0 by 1
+ \repeat
+%    \end{macrocode}
+% The number of zeros to be padded is $\max(0,|#1-\count0|)$ (the second
+% argument can be negative), so a simple loop suffices.
+%    \begin{macrocode}
+ \loop
+  \ifnum\count0<#1 0\advance\count0 by 1
+ \repeat 
+ \number#2}}
+%    \end{macrocode}
+% \end{macro}
+% \begin{macro}{\zetg@ld}
+% This macro takes care for several things: it increases the total for a given
+% f-column, it records the largest width of the entries in that column and it
+% typesets |#1| via |\geldm@cro|.
+%    \begin{macrocode}
+\def\zetg@ld#1#2{\global\advance\csname 
+ FCtot@\romannumeral\FCsc@l\endcsname by #1
+ \setbox0=\hbox{$\geldm@cro{#1}{#2}$}%
+ \ifdim\wd0>\csname FCwd@\romannumeral\FCsc@l\endcsname
+  \global\csname FCwd@\romannumeral\FCsc@l\endcsname=\wd0
+ \fi\unhbox0}
+%    \end{macrocode}
+% \end{macro}
+% \begin{macro}{\check}
+% The macro |\check| provides a way to the user to check that the  appropriate 
+% columns are balanced (as it should in a balance).  Arguments |#1| and |#2| are
+% the column numbers to compare.  It is the responsibility of the user to provide
+% the correct numbers here, otherwise bogus output is generated.  But the output is
+% only to screen and the transcript file; it doesn't change the appearance of your
+% document, so you can safely ignore the result then, repair and go on.
+%    \begin{macrocode}
+\def\check#1#2{\noalign{\ifnum\csname FCtot@\romannumeral#1\endcsname=
+ \csname FCtot@\romannumeral#2\endcsname\else
+ \message{Columns #1 and #2 do not balance!}\fi}}
+%    \end{macrocode}
+% \end{macro}
+% \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 |\let| is only local to the column and is necessary because
+% |\ignorespaces| is automatically inserted by the preamble-generating macro |\@mkpream|.
+% Its effect should be annihilated, otherwise the assignment to |\bedr@g| goes wrong.
+% The plain \LaTeX\ command |\@empty| is used for that.  The separator and decimal mark
+% are within a math environment, so you can indeed specify |\,| instead of |\thinspace|,
+% but there is an extra brace around, so it doesn't affect
+% the spacing between the digits (trick copied from |dcolumn|).
+%    \begin{macrocode}
+\newcount\bedr@g
+\def\b@fi#1#2#3{\def\sep@rator{{#1}}\def\decim@lmark{{#2}}%
+ \let\ignorespaces=\@empty \def\sp@l{#3}\global\advance\FCsc@l by \@ne 
+%    \end{macrocode}
+% The value specified by the user is then captured by |\bedr@g|.
+%    \begin{macrocode}
+ \bedr@g=}
+%    \end{macrocode}
+% \end{macro}
+% \begin{macro}{\e@fi}
+% Once captured, the rest is handled by |\e@fi|.  The number is then 
+% fed to |\zetg@ld| for the current column.
+%    \begin{macrocode}
+\def\e@fi{\relax \zetg@ld{\number\bedr@g}{\sp@l}}
+%    \end{macrocode}
+% \end{macro}
+%
+% Here are adaptations to existing macros.
+% \begin{macro}{\@xarraycr}
+% Among all its tasks, now |\\| also resets the column number.  This 
+% is a transparant change, i.e., loading this package without using 
+% the ``f'' column specifier does
+% not change the effect of package |array.sty| in any way.
+% This resetting has to be done after the |\cr|, so it is stored
+% in a |\noalign|.
+%    \begin{macrocode}
+\def\@xarraycr{\@ifnextchar [%
+ \@argarraycr {\ifnum 0=`{}\fi \cr\noalign{\global\FCsc@l=0 }}}
+%    \end{macrocode}
+% \end{macro}
+% \begin{macro}{\@array}
+% The definition of |\@array| had to be extended slightly because
+% it should also include |\@mksumline| (acting on the same |#2|
+% as |\@mkpream| gets).  Also this change is transparant: it only adds 
+% functionality and if you don't use that, you won't notice the difference.
+%    \begin{macrocode}
+\def\@array[#1]#2{%
+ \@tempdima \ht \strutbox
+ \advance \@tempdima by\extrarowheight
+ \setbox \@arstrutbox \hbox{\vrule
+            \@height \arraystretch \@tempdima
+            \@depth \arraystretch \dp \strutbox
+            \@width \z@}%
+ \begingroup
+ \@mkpream{#2}%
+ \xdef\@preamble{\noexpand \ialign \@halignto
+                 \bgroup \@arstrut \@preamble
+                         \tabskip \z@ \cr}%
+%    \end{macrocode}
+% The combination |\endgroup| followed by |\begingroup| seems redundant, but that is
+% not the case: the |\endgroup| restores everything that was not |\global|.  With the
+% following |\begingroup| it is ensured that |\@mksumline| experiences the same
+% settings as |\@mkpream| did.
+%    \begin{macrocode}
+ \endgroup 
+ \begingroup
+ \@mksumline{#2}%
+ \endgroup
+%    \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.
+%    \begin{macrocode}
+ \res@tsumline
+%    \end{macrocode}
+% From here on it is just the old definition of |array.sty|.
+%    \begin{macrocode}
+ \@arrayleft
+ \if #1t\vtop \else \if#1b\vbox \else \vcenter \fi \fi
+ \bgroup
+ \let \@sharp ##\let \protect \relax
+ \lineskip \z@
+ \baselineskip \z@
+ \m@th
+ \let\\\@arraycr \let\tabularnewline\\\let\par\@empty \@preamble} 
+%    \end{macrocode}
+% 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}
+% \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 |\sumline|, e.g., |@|, |!|,
+% and \verb+|+ can be fully ignored. Ampersands are only inserted by |c|, |l|, |r|,  
+% |p|, |m|, and |b|.  So a specifier like |@{}lflf@{}| will yield the sumline
+% |&\a&&\a\\|, where |\a| is a macro that prints the desired result of the column.
+% Had the specifier been \verb+l|f||@{   }l|f+, then the same sumline must
+% be constructed: all difficulties are already picked up and solved in the creation
+% of the preamble.
+%
+% In reality the sumline must be constructed from the expanded form of the specifier,
+% so |@{}lf@{}| will expand as |@{}l>{\b@fi#1#2#2}r<{\e@fi}@{}|. The
+% rules for constructing the sumline are now very simple:
+% \begin{compactitem}
+% \item add an ampersand when |c|, |l|, |r|, |p|, |m|, or |b| is found, unless it is
+% the first one (this is the same as in the preamble);
+% \item add a |\a| when |<{\e@fi}| is found;
+% \item ignore everything else.
+% \end{compactitem}
+% To discriminate, a special version of |\testpach| could be written, but that is 
+% not necessary: |\testpach| can do all the work, although much of it will be discarded.
+% Here speed is sacrificed for space and this can be afforded because the creation
+% of the sumline is only done once per |\tabular|.
+% 
+% The start is copied from |\@mkpream|.
+%    \begin{macrocode}
+\def\@mksumline#1{\gdef\sumline{}
+ \@lastchclass 4 \@firstamptrue
+%    \end{macrocode}
+% At first the column number is reset and the actual code for what was called |\a| 
+% above is made inactive.
+%    \begin{macrocode}
+ \global\FCsc@l=\z@
+ \let\prr@sult=\relax
+%    \end{macrocode}
+% Then |\@mkpream| is picked up again.
+%    \begin{macrocode}
+ \@temptokena{#1}
+ \@tempswatrue
+ \@whilesw\if@tempswa\fi{\@tempswafalse\the\NC@list}%
+ \count@\m@ne
+ \let\the@toks\relax
+ \prepnext@tok
+%    \end{macrocode}
+% Next is the loop over all tokens in the expanded form of the specifier.  The change
+% with respect to |\@mkpream| is that the body of the loop is now only dealing
+% with classes 0, 2, and 10.  What to do in those cases is of course different from
+% 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 appending the |\\| to the sumline.
+%    \begin{macrocode}
+ \xdef\sumline{\sumline\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.
+%    \begin{macrocode}
+\def\@addtosumline#1{\xdef\sumline{\sumline #1}}
+%    \end{macrocode}
+% \end{macro}
+% \begin{macro}{\@classfx}
+% Class~10 for the sumline creation is a stripped down version of |\@classx|: add an ampersand
+% unless it is the first.  It deals with the specifiers |b|, |m|, |p|, |c|, |l|, and |r|.
+%    \begin{macrocode}
+\def\@classfx{\if@firstamp \@firstampfalse \else \@addtosumline &\fi} 
+%    \end{macrocode}
+% \end{macro}
+% \begin{macro}{\@classfz}
+% Class~0 is applicable for specifiers |c|, |l|, and |r| and if the arguments of |p|, |m|,
+% or |b| are given.  The latter three cases, with |\@chnum| is 0, 1, or~2 should be ignored
+% and the first three cases are now similar to class~10.
+%    \begin{macrocode}
+\def\@classfz{\ifnum\@chnum<\thr@@ \@classfx\fi}
+%    \end{macrocode}
+% \end{macro}
+% \begin{macro}{\@classfii}
+% Here comes the nice and nasty part. Class~2 is applicable if a |<| is specified.  This is
+% tested by checking |\@lastchclass|, which should be equal to~8.  Then it is checked that
+% the argument to |<| is indeed |\e@fi|.  This check is rather clumsy but this was the
+% first way, after many attempts, that worked.  It is necessary because the usage of |<|
+% is not restricted to |\e@fi|: the user may have specified other \TeX-code using |<|.
+%    \begin{macrocode}
+\def\@classfii{\ifnum\@lastchclass=8 
+ \edef\t@stm{\expandafter\string\@nextchar}
+ \edef\t@stn{\string\e@fi}
+ \ifx\t@stm\t@stn
+%    \end{macrocode}
+% If both tests yield |true| 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.
+%    \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,
+% so a new one should be created.  What is checked here is its existence: if it exists, it is not
+% created by |fcolumn| so it may be something other than a \meta{count}, in which things go haywire.
+% Therefore it is assumed that only |fcolumn| is allowed to generate a \meta{count} called 
+% |\FCtot@<some romannumeral>|.
+%    \begin{macrocode}
+   \expandafter\ifx\csname FCtot@\romannumeral\FCsc@l\endcsname\relax
+    \expandafter\newcount\csname FCtot@\romannumeral\FCsc@l\endcsname
+   \else
+    \message{\noexpand\FCtot@\romannumeral\FCsc@l already exists.}
+   \fi
+%    \end{macrocode}
+% And the same is applicable for the \meta{dimen}: it is assumed that only |fcolumn| is allowed to
+% generate a \meta{dimen} called |\FCwd@<some romannumeral>|.
+%    \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.
+%    \begin{macrocode}
+    \global\FCtc@l=\FCsc@l
+   \else
+    \message{\noexpand\FCwd@\romannumeral\FCsc@l already exists.}
+   \fi
+  \fi
+ \fi
+\fi}
+%    \end{macrocode}
+% Once created they are not initialised here because that is done later in one go.
+% \end{macro}
+% \begin{macro}{\leeg}
+% The macro |\prr@sult| (see below) actually puts the information together.
+% It starts with switching off the default behaviour of the f-column by 
+% providing the number 0 to comfort the assignment in |\b@fi|, although
+% any number would be good: it
+% is discarded anyway.  This is achieved by |\let|ting |\e@fi| to be |\relax|.
+% This |\let| is only local to the current column.  Since the user may from time to
+% time also need an empty entry in the table, this definition is without at-sign.
+%    \begin{macrocode}
+\def\leeg{0\relax \let\e@fi=\relax}
+%    \end{macrocode}
+% A better solution were to check for a number.  If present then do the 
+% default action, otherwise generate an empty entry.
+% \end{macro}
+% \begin{macro}{\prr@sult}
+% And here it is: the end result.  First annihilate the effect of |\e@fi|.
+%    \begin{macrocode}
+\def\prr@sult{\leeg
+%    \end{macrocode}
+% Then the information for the last line is computed.  It is not sufficient to calculate
+% the width of the result (in points) to use that as the width of the rule separating 
+% the individual entries and the result.  It may be that the sum is larger
+% (in points) than any of the entries, e.g., when the result of $6+6$ is typeset.  The
+% width of the rule should be equal to the width of |\hbox{$12$}| then.  On the other
+% hand the width of the rule when summing $24$ and $-24$ should be that of |\hbox{$-24$}|
+% (or |\hbox{$(24$}|, see above), not the width of the result |\hbox{$0$}|.  Therefore the 
+% maximum of all entry widths, including the result, was calculated.
+%    \begin{macrocode}
+ \setbox0=\hbox{$\geldm@cro{\number\csname
+ FCtot@\romannumeral\FCsc@l\endcsname}{\sp@l}$}%
+ \ifdim\wd0>\csname FCwd@\romannumeral\FCsc@l\endcsname
+   \global\csname FCwd@\romannumeral\FCsc@l\endcsname=\wd0
+ \fi
+ \vbox{\hbox to \csname
+  FCwd@\romannumeral\FCsc@l\endcsname{\hrulefill}\vskip2pt
+  \hbox to \csname 
+  FCwd@\romannumeral\FCsc@l\endcsname{\hfil\unhbox0}}}
+%    \end{macrocode}
+% \end{macro}
+% \begin{macro}{\res@tsumline}
+% Since all changes to the totals and widths of the columns are global,
+% they have to be reset actively at the start of a tabular or array.
+% That is an action by itself, but it may occur more often, on request
+% of the user, therefore a special macro is defined.  The count 
+% |\FCsc@l| can be used as a temporary register because its actual
+% value is not important anymore.
+%    \begin{macrocode}
+\def\res@tsumline{\FCsc@l\FCtc@l \loop\ifnum\FCsc@l>0
+   \global\csname FCtot@\romannumeral\FCsc@l\endcsname=0
+   \global\csname FCwd@\romannumeral\FCsc@l\endcsname=0pt
+   \advance\FCsc@l by \m@ne \repeat}
+%    \end{macrocode}
+% \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}
+% That's it!
+% \StopEventually{\PrintIndex}
+% \Finale
+%\typeout{****************************************************}
+%\typeout{*}
+%\typeout{* To finish the installation you have to move the}
+%\typeout{* following file into a directory searched by TeX:}
+%\typeout{*}
+%\typeout{*\space\space\space fcolumn.sty}
+%\typeout{*}
+%\typeout{* To produce the documentation run the file}
+%\typeout{* fcolumn.dtx through LaTeX.}
+%\typeout{*}
+%\typeout{* Happy TeXing!}
+%\typeout{*}
+%\typeout{****************************************************}
+\endinput
diff --git a/Master/texmf-dist/tex/latex/fcolumn/fcolumn.sty b/Master/texmf-dist/tex/latex/fcolumn/fcolumn.sty
new file mode 100644
index 00000000000..f0c7225d080
--- /dev/null
+++ b/Master/texmf-dist/tex/latex/fcolumn/fcolumn.sty
@@ -0,0 +1,189 @@
+%%
+%% This is file `fcolumn.sty',
+%% generated with the docstrip utility.
+%%
+%% The original source files were:
+%%
+%% fcolumn.dtx  (with options: `package')
+%% 
+%% This is a generated file.
+%% 
+%% Copyright (C) 2015 by Edgar Olthof
+%% 
+%% This file may be distributed and/or modified under the
+%% conditions of the LaTeX Project Public License, either
+%% version 1.2 of this license or (at your option) any later
+%% version.  The latest version of this license is in
+%% 
+%%    http://www.latex-project.org/lppl.txt
+%% 
+%% and version 1.2 or later is part of all distributions of
+%% LaTeX version 1999/12/01 or later.
+\NeedsTeXFormat{LaTeX2e}
+\ProvidesPackage{fcolumn}
+          [2015/03/07 v1.0 automatic currency package]
+\RequirePackage{array}
+\newcolumntype{F}[3]{>{\b@fi{#1}{#2}{#3}}r<{\e@fi}}
+\newcolumntype{f}{F{.}{,}{3,2}}
+\newcount\FCsc@l \FCsc@l0
+\newcount\FCtc@l \FCtc@l0
+\newcount\g@lda
+\newcount\g@ldb
+\newif\ifwiths@p
+\newif\ifstrictaccounting \strictaccountingfalse
+\def\geldm@cro#1#2{\withs@pfalse
+ \afterassignment\g@ldm@cro\count@#2\relax{#1}}
+\def\g@ldm@cro#1\relax#2{%
+ \ifnum#2<0 \ifstrictaccounting(\else-\fi\g@lda=-#2 \else\g@lda=#2 \fi
+ \ifx\relax#1\relax\g@ldb=2 \else \g@ldb\@gobble#1\relax\fi
+ \loop
+  \ifnum\g@ldb>0 \divide\g@lda by 10 \advance\g@ldb by \m@ne
+ \repeat
+ \g@ldens{\the\count@}%
+ \ifx\relax#1\relax\g@ldb=2 \else \g@ldb\@gobble#1\relax\fi
+ \ifnum\g@ldb>0\decim@lmark
+  \ifnum#2>0 \g@lda=-#2\else\g@lda=#2 \fi
+  \loop
+   \ifnum\g@ldb>0 \divide\g@lda by 10 \advance\g@ldb by \m@ne
+  \repeat
+  \ifx\relax#1\relax\g@ldb=2 \else \g@ldb\@gobble#1\relax\fi
+  \loop
+   \ifnum\g@ldb>0 \multiply\g@lda by 10 \advance\g@ldb by \m@ne
+  \repeat
+  \ifnum#2>0 \advance\g@lda by #2
+  \else \advance\g@lda by -#2
+  \fi
+  \ifx\relax#1\relax\g@ldb=2 \else \g@ldb\@gobble#1\relax\fi
+  \zerop@d{\number\g@ldb}{\number\g@lda}%
+ \fi
+ \ifnum#2<0 \ifstrictaccounting\rlap{)}\fi\fi}
+\def\g@ldens#1{{\g@ldb=\g@lda \count0=#1
+ \loop
+  \ifnum\count0>0
+  \divide\g@lda by 10
+  \advance\count0 by \m@ne
+ \repeat
+ \ifnum\g@lda>0 \g@ldens{#1}\fi
+ \count0=#1
+ \loop
+  \ifnum\count0>0
+  \multiply\g@lda by 10
+  \advance\count0 by \m@ne
+ \repeat
+ \g@lda=-\g@lda
+ \advance\g@lda by \g@ldb \du@zendprint{#1}}}
+\def\du@zendprint#1{\ifwiths@p\sep@rator\zerop@d{#1}{\number\g@lda}%
+ \else\zerop@d{1}{\number\g@lda}\fi
+ \global\withs@ptrue}
+\def\zerop@d#1#2{{\count0=1 \count1=#2
+ \loop
+  \divide \count1 by 10
+  \ifnum\count1>0 \advance\count0 by 1
+ \repeat
+ \loop
+  \ifnum\count0<#1 0\advance\count0 by 1
+ \repeat
+ \number#2}}
+\def\zetg@ld#1#2{\global\advance\csname
+ FCtot@\romannumeral\FCsc@l\endcsname by #1
+ \setbox0=\hbox{$\geldm@cro{#1}{#2}$}%
+ \ifdim\wd0>\csname FCwd@\romannumeral\FCsc@l\endcsname
+  \global\csname FCwd@\romannumeral\FCsc@l\endcsname=\wd0
+ \fi\unhbox0}
+\def\check#1#2{\noalign{\ifnum\csname FCtot@\romannumeral#1\endcsname=
+ \csname FCtot@\romannumeral#2\endcsname\else
+ \message{Columns #1 and #2 do not balance!}\fi}}
+\newcount\bedr@g
+\def\b@fi#1#2#3{\def\sep@rator{{#1}}\def\decim@lmark{{#2}}%
+ \let\ignorespaces=\@empty \def\sp@l{#3}\global\advance\FCsc@l by \@ne
+ \bedr@g=}
+\def\e@fi{\relax \zetg@ld{\number\bedr@g}{\sp@l}}
+\def\@xarraycr{\@ifnextchar [%
+ \@argarraycr {\ifnum 0=`{}\fi \cr\noalign{\global\FCsc@l=0 }}}
+\def\@array[#1]#2{%
+ \@tempdima \ht \strutbox
+ \advance \@tempdima by\extrarowheight
+ \setbox \@arstrutbox \hbox{\vrule
+            \@height \arraystretch \@tempdima
+            \@depth \arraystretch \dp \strutbox
+            \@width \z@}%
+ \begingroup
+ \@mkpream{#2}%
+ \xdef\@preamble{\noexpand \ialign \@halignto
+                 \bgroup \@arstrut \@preamble
+                         \tabskip \z@ \cr}%
+ \endgroup
+ \begingroup
+ \@mksumline{#2}%
+ \endgroup
+ \res@tsumline
+ \@arrayleft
+ \if #1t\vtop \else \if#1b\vbox \else \vcenter \fi \fi
+ \bgroup
+ \let \@sharp ##\let \protect \relax
+ \lineskip \z@
+ \baselineskip \z@
+ \m@th
+ \let\\\@arraycr \let\tabularnewline\\\let\par\@empty \@preamble}
+\let\@@array=\@array
+\def\@mksumline#1{\gdef\sumline{}
+ \@lastchclass 4 \@firstamptrue
+ \global\FCsc@l=\z@
+ \let\prr@sult=\relax
+ \@temptokena{#1}
+ \@tempswatrue
+ \@whilesw\if@tempswa\fi{\@tempswafalse\the\NC@list}%
+ \count@\m@ne
+ \let\the@toks\relax
+ \prepnext@tok
+ \expandafter \@tfor \expandafter \@nextchar
+  \expandafter :\expandafter =\the\@temptokena \do
+  {\@testpach
+  \ifcase \@chclass \@classfz
+  \or \or \@classfii \or
+  \or \or \or \or \or \or \or \@classfx \fi
+  \@lastchclass\@chclass}%
+ \xdef\sumline{\sumline\noexpand\\}}
+\def\@addtosumline#1{\xdef\sumline{\sumline #1}}
+\def\@classfx{\if@firstamp \@firstampfalse \else \@addtosumline &\fi}
+\def\@classfz{\ifnum\@chnum<\thr@@ \@classfx\fi}
+\def\@classfii{\ifnum\@lastchclass=8
+ \edef\t@stm{\expandafter\string\@nextchar}
+ \edef\t@stn{\string\e@fi}
+ \ifx\t@stm\t@stn
+  \@addtosumline{\prr@sult}
+  \global\advance\FCsc@l by \@ne
+  \ifnum\FCsc@l>\FCtc@l
+   \expandafter\ifx\csname FCtot@\romannumeral\FCsc@l\endcsname\relax
+    \expandafter\newcount\csname FCtot@\romannumeral\FCsc@l\endcsname
+   \else
+    \message{\noexpand\FCtot@\romannumeral\FCsc@l already exists.}
+   \fi
+   \expandafter\ifx\csname FCwd@\romannumeral\FCsc@l\endcsname\relax
+    \expandafter\newdimen\csname FCwd@\romannumeral\FCsc@l\endcsname
+    \global\FCtc@l=\FCsc@l
+   \else
+    \message{\noexpand\FCwd@\romannumeral\FCsc@l already exists.}
+   \fi
+  \fi
+ \fi
+\fi}
+\def\leeg{0\relax \let\e@fi=\relax}
+\def\prr@sult{\leeg
+ \setbox0=\hbox{$\geldm@cro{\number\csname
+ FCtot@\romannumeral\FCsc@l\endcsname}{\sp@l}$}%
+ \ifdim\wd0>\csname FCwd@\romannumeral\FCsc@l\endcsname
+   \global\csname FCwd@\romannumeral\FCsc@l\endcsname=\wd0
+ \fi
+ \vbox{\hbox to \csname
+  FCwd@\romannumeral\FCsc@l\endcsname{\hrulefill}\vskip2pt
+  \hbox to \csname
+  FCwd@\romannumeral\FCsc@l\endcsname{\hfil\unhbox0}}}
+\def\res@tsumline{\FCsc@l\FCtc@l \loop\ifnum\FCsc@l>0
+   \global\csname FCtot@\romannumeral\FCsc@l\endcsname=0
+   \global\csname FCwd@\romannumeral\FCsc@l\endcsname=0pt
+   \advance\FCsc@l by \m@ne \repeat}
+\def\resetsumline{\noalign{\res@tsumline}}
+\endinput
+%%
+%% End of file `fcolumn.sty'.
diff --git a/Master/tlpkg/bin/tlpkg-ctan-check b/Master/tlpkg/bin/tlpkg-ctan-check
index aaeaf6bf911..a21c71e7a5e 100755
--- a/Master/tlpkg/bin/tlpkg-ctan-check
+++ b/Master/tlpkg/bin/tlpkg-ctan-check
@@ -207,7 +207,7 @@ my @TLP_working = qw(
     fancybox fancyhdr fancyhdr-it fancylabel fancynum fancypar
     fancyref fancyslides fancytabs fancytooltips fancyvrb fandol
     FAQ-en fast-diagram fbb fbithesis fbs
-    fc fcltxdoc fdsymbol featpost fenixpar
+    fc fcltxdoc fcolumn fdsymbol featpost fenixpar
     fetamont feupphdteses feyn feynmf feynmp-auto fge
     fifinddo-info fifo-stack fig4latex figbas figbib figflow figsize
     filecontents filedate filehook fileinfo filemod
diff --git a/Master/tlpkg/tlpsrc/collection-latexextra.tlpsrc b/Master/tlpkg/tlpsrc/collection-latexextra.tlpsrc
index fece6363304..2bde6d9ffbd 100644
--- a/Master/tlpkg/tlpsrc/collection-latexextra.tlpsrc
+++ b/Master/tlpkg/tlpsrc/collection-latexextra.tlpsrc
@@ -302,6 +302,7 @@ depend fancypar
 depend fancyslides
 depend fancytabs
 depend fancytooltips
+depend fcolumn
 depend fifo-stack
 depend figsize
 depend filecontents
diff --git a/Master/tlpkg/tlpsrc/fcolumn.tlpsrc b/Master/tlpkg/tlpsrc/fcolumn.tlpsrc
new file mode 100644
index 00000000000..e69de29bb2d
--- /dev/null
+++ b/Master/tlpkg/tlpsrc/fcolumn.tlpsrc