%\iffalse
% csvtools.dtx generated using makedtx version 0.91b (c) Nicola Talbot
% Command line args:
%   -src "csvtools\.sty\Z=>csvtools.sty"
%   -src "csvpie\.sty\Z=>csvpie.sty"
%   -src "csvsort\.sty\Z=>csvsort.sty"
%   -doc "manual.tex"
%   -author "Nicola Talbot"
%   -dir "source"
%   csvtools
% Created on 2007/7/3 10:24
%\fi
%\iffalse
%<*package>
%% \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         \~}
%</package>
%\fi
% \iffalse
% Doc-Source file to use with LaTeX2e
% Copyright (C) 2007 Nicola Talbot, all rights reserved.
% \fi
% \iffalse
%<*driver>
\documentclass{ltxdoc}
\usepackage[colorlinks,
           bookmarks,
           hyperindex=false,
           pdfauthor={Nicola Talbot},
           pdftitle={csvtools: A LaTeX2e Package Providing Access to Data Saved in a CSV File},
           pdfkeywords={LaTeX,package,csv}]{hyperref}
\usepackage{csvpie}
\usepackage{colortbl}


\renewcommand{\usage}[1]{\hyperpage{#1}}
\renewcommand{\main}[1]{\hyperpage{#1}}
\newcommand{\see}[2]{\emph{see} #1}
\makeatletter
\def\index@prologue{\section*{Index}}
\makeatother
\RecordChanges
\PageIndex
\CheckSum{1878}
\OnlyDescription
\newcommand{\stynamefmt}[1]{\textsf{#1}}
\newcommand{\envname}[1]{\textsf{#1}}


\newcommand{\cmdname}[1]{\texttt{\symbol{92}#1}}

\iffalse
Doc-Source file to use with LaTeX2e
Copyright (C) 2003 Nicola Talbot, all rights reserved.
\fi


\newcounter{example}
\newenvironment{example}[2]{\refstepcounter{example}\label{#2}%
\subsection*{Example \theexample\ (#1)}%
\addcontentsline{loe}{section}{\protect\numberline{\theexample}#1}}{\par\centerline{\rule{2in}{1pt}}\par}

\begin{document}
\DocInput{csvtools.dtx}
\end{document}
%</driver>
%\fi
% \DeleteShortVerb{\|}
% \MakeShortVerb{"}
% \title{csvtools v1.24 : A \LaTeXe\ Package Providing Access to Data Saved in a CSV File}
% \author{Nicola Talbot}
% \date{3rd July 2007}
% \maketitle
% \tableofcontents
%
%\section*{List of Examples}
%
%\makeatletter\@starttoc{loe}\makeatother
%
% \section{Introduction}
% The \stynamefmt{csvtools} package allows you to repeatedly perform
% a set of \LaTeX\ commands on data in each row
% of a comma separated variable (CSV) file.
% This can be used for mail merging, generating
% tables etc.
%
%\changes{1.2}{2006 Sept 1}{separator can now be changed}%
%As from version 1.2, you can specify a different separator.
%To change the separator, use the command:\\[10pt]
%\DescribeMacro{\setcsvseparator}\cmdname{setcsvseparator}\{\meta{separator}\}\\[10pt]
%For example, if your data is separated by colons instead of
%commas, do:
%\begin{verbatim}
%\setcsvseparator{:}
%\end{verbatim}
%\changes{1.2}{2006 Sept 1}{entries can be delimited with double quotes}%
%If your separator occurs within an entry, the entry must be
%enclosed in double quotes, for example:
%\begin{verbatim}
%Name,Address,Telephone
%A.N. Other,"1 The Street,The Town",0123456789
%\end{verbatim}
%Be careful of \TeX\ special characters occuring within a CSV
%file, for example:
%\begin{verbatim}
%Name,Address,Telephone
%Jack \& Jill,"2 The Street,The Town",0123456789
%\end{verbatim}
%
% \section{Mail Merging and Similar Applications}
%
%\DescribeMacro{\applyCSVfile}\noindent
% \cmdname{applyCSVfile}[\meta{n}]\{\meta{filename}\}\{\meta{text}\}\\
% \cmdname{applyCSVfile*}[\meta{n}]\{\meta{filename}\}\{\meta{text}\}\\[10pt]
% Letters can be generated using data given in
% each line from \meta{filename}.
% If the CSV file contains a header row, the
% unstarred version of \cmdname{applyCSVfile} should
% be used, otherwise the starred version \cmdname{applyCSVfile*}
% should be used.  The optional argument \meta{n}
% specifies on which line the actual data (not header line)
% starts.  The unstarred version defaults to line 2
% (the header row is always assumed to be on line 1)
% and the starred version defaults to 1.
%
% With the unstarred version, the entries in the header row
% are used to generate commands of the form
% \DescribeMacro{\insert...}\cmdname{insert}\meta{identifier}\footnote{See Note~\ref{itm:insert} in Section~\ref{sec:drawbacks}}
% to access corresponding elements
% in the row currently being processed.  For example,
% suppose the first line of the CSV file looks like:
%\begin{verbatim}
%Name,Address,Time,Date
%\end{verbatim}
% then the commands \cmdname{insertName}, \cmdname{insertAddress},
% \cmdname{insertTime} and \cmdname{insertDate} are
% created, allowing you to use the entries in the first,
% second, third and fourth columns of the current row.
% If the header text contains non-alphabetical characters,
% e.g.\ \texttt{Full Name}, then you will need to use
% \DescribeMacro{\insertbyname}\cmdname{insertbyname}\texttt{\{}\meta{text}\texttt{\}},
% e.g.\ \verb'\insertbyname{Full Name}'.
%
%Alternatively, you can use the
%\DescribeMacro{\field}\cmdname{field}\{\meta{col}\} command, where
%\meta{col} is the column number of the entry, so \verb"\field{1}"
%indicates the first entry in the current row and \verb"\field{2}"
%indicates the second entry in the current row.
%
% \begin{example}{Mail Merging}{ex:mail}
% Suppose there is a file called \texttt{details.csv} that has the
% following contents:
%\begin{verbatim}
%Name,Address,Time,Date
%Miss A. Person,1 The Road\\The Town\\AB1 2XY,15.00,4th May 2004
%Mr A. N. Other,2 The Road\\The Town\\AB1 2XY,15.30,11th May 2004
%\end{verbatim}
%then the following code can be used to generate a letter for each
%person in the CSV file\footnote{Remeber to use a letter type of
%class file}:
%\begin{verbatim}
%\applyCSVfile{details.csv}{%
%\begin{letter}{\insertName\\\insertAddress}
%\opening{Dear \insertName}
%
%You are invited to an interview at \insertTime\ on the \insertDate.
%
%\closing{Yours Sincerely}
%\end{letter}}
%\end{verbatim}
% Note that you could also use \verb'\insertbyname{Name}' etc instead
% of \cmdname{insertName} etc.  Also note that you need to specify the
% file extension when specifying the filename.
%\end{example}
%
% \begin{example}{Multiple Figures}{ex:ps}
% Suppose \texttt{sample3.csv} looks like:
%\begin{verbatim}
%File,Caption
%circle.ps,A Circle
%rectangle.ps,A Rectangle
%triangle.ps,A Triangle
%\end{verbatim}
% Assuming that the files \texttt{circle.ps}, \texttt{rectangle.ps}
% and \texttt{triangle.ps} exist, then the following code will
% generate a figure for each graphics file\footnote{The
% \stynamefmt{graphics} or \stynamefmt{graphicx} package will be
%needed.}:
%\begin{verbatim}
%\applyCSVfile{sample3.csv}{
%\begin{figure}
%\centerline{\includegraphics{\insertFile}}
%\caption{\insertCaption}
%\end{figure}}
%\end{verbatim}
% Note that in this example, you can't use \verb'\insertbyname{File}'.
% (See Note~\ref{itm:psbyname} in Section~\ref{sec:drawbacks}.)
%\end{example}
%
% \begin{example}{Mail Merging using \cmdname{field}}{ex:field}
% Suppose there is a file called \texttt{details.csv} that has the
% following contents:
%\begin{verbatim}
%Miss A. Person,1 The Road\\The Town\\AB1 2XY,15.00,4th May 2004
%Mr A. N. Other,2 The Road\\The Town\\AB1 2XY,15.30,11th May 2004
%\end{verbatim}
%In this case the data has no header file, so the starred version
%of \cmdname{applyCSVfile} must be used. Since there is no header
%file, you must use \cmdname{field} to access the entries:
%\begin{verbatim}
%\applyCSVfile*{details.csv}{%
%\begin{letter}{\field{1}\\\field{2}}
%\opening{Dear \field{1}}
%
%You are invited to an interview at \field{3}\ on the \field{4}.
%
%\closing{Yours Sincerely}
%\end{letter}}
%\end{verbatim}
%\end{example}
%
% \section{Converting data in a CSV file into a tabular environment}
% \DescribeMacro{\CSVtotabular}
% \cmdname{CSVtotabular}\{\meta{filename}\}\{\meta{col-align}\}\{\meta{first}\}\{\meta{middle}\}\{\meta{last}\}\\[10pt]
% \meta{filename} is the name of the CSV file which must have a header row on line~1,
% \meta{col-align} is the column alignment argument that gets passed
% to the \envname{tabular} environment, \meta{first}
% is the code for the first line, \meta{middle} is the code
% for the middle lines and \meta{last} is the code for the last line.
% This is best demonstrated with an example.
%
% \begin{example}{Aligning Data from a CSV file}{ex:tab1}
% Suppose the file \texttt{sample.csv} looks like:
%\begin{verbatim}
%Name,Assignment 1,Assignment 2,Total
%A. Smith,80,70,150
%B. Jones,60,80,140
%J. Doe,85,75,160
%,75,75,150
%\end{verbatim}
% then the following code can be used to align the data:
%\begin{verbatim}
%\CSVtotabular{sample.csv}{lccc}{%
%\bfseries Name &
%\bfseries Assignment 1&
%\bfseries Assignment 2&
%\bfseries Total\\}{%
%\insertName &
%\insertbyname{Assignment 1} &
%\insertbyname{Assignment 2} &
%\insertTotal\\}{%
% &
%\insertbyname{Assignment 1} &
%\insertbyname{Assignment 2} &
%\insertTotal}
%\end{verbatim}
% The result of this code is shown in
% Table~\ref{tab:ex1}\footnote{Note that \cmdname{CSVtotabular} only
% puts the data in a \envname{tabular} environment not in a table}.
%\begin{table}
%\caption{Example~\ref{ex:tab1}}
%\label{tab:ex1}
%\vspace{10pt}
%\centerline{%
%\begin{tabular}{lccc}
%\bfseries Name &
%\bfseries Assignment 1 &
%\bfseries Assignment 2 &
%\bfseries Total\\
%A. Smith&80&70&150\\
%B. Jones&60&80&140\\
%J. Doe&85&75&160\\
% &75&75&150
%\end{tabular}}
%\end{table}
%\end{example}
%
% \vspace{10pt}
% \DescribeMacro{\ifnextrowlast}
% \cmdname{ifnextrowlast}\{\meta{last-code}\}\{\meta{not-last-code}\}\\[10pt]
% The command \cmdname{ifnextrowlast} can be used to vary what happens
% on the last but one row.  The following example illustrates
% this by placing \cmdname{hline}\cmdname{hline} after the penultimate row.
%
% \begin{example}{Adding Lines}{ex:tab2}
%\begin{verbatim}
%\CSVtotabular{sample.csv}{|l|ccc|}{%
%\hline\bfseries Name &
%\bfseries Assignment 1&
%\bfseries Assignment 2&
%\bfseries Total\\\hline\hline}{%
%\insertName &
%\insertbyname{Assignment 1} &
%\insertbyname{Assignment 2} &
%\insertTotal
%\ifnextrowlast{\\\hline\hline}{\\}}{%
% &
%\insertbyname{Assignment 1} &
%\insertbyname{Assignment 2} &
%\insertTotal\\\hline}
%\end{verbatim}
%This result of this code is shown in Table~\ref{tab:ex2}.\begin{table}
%\caption{Example~\ref{ex:tab2}}
%\label{tab:ex2}
%\vspace{10pt}
%\begin{center}
%\begin{tabular}{|l|ccc|}
%\hline\bfseries Name &
%\bfseries Assignment 1 &
%\bfseries Assignment 2 &
%\bfseries Total\\\hline\hline
%A. Smith&80&70&150\\
%B. Jones&60&80&140\\
%J. Doe&85&75&160\\\hline\hline
% &75&75&150\\\hline
%\end{tabular}
%\end{center}
%\end{table}
%\end{example}
%
% \begin{example}{Added Complexity}{ex:tab3}
% In this example, \cmdname{multicolumn} is used to override
% the column specifier for the first column in the
% last row.
%\begin{verbatim}
%\CSVtotabular{sample2.csv}{|l|ccc|}{%
%\hline\bfseries Name &
%\bfseries Assignment 1 &
%\bfseries Assignment 2 &
%\bfseries Total\\\hline\hline
%}{%
%\insertName &
%\insertbyname{Assignment 1} &
%\insertbyname{Assignment 2} &
%\insertTotal
%\ifnextrowlast{\\\hline\multicolumn{1}{l|}{}}{\\}
%}{%
% &
%\insertbyname{Assignment 1} &
%\insertbyname{Assignment 2} &
%\insertTotal\\\cline{2-4}
%}
%\end{verbatim}
% Notice that instead of placing \verb'\multicolumn{1}{l|}{}'
% at the start of the final argument, it is instead
% placed in the first argument to \cmdname{ifnextrowlast}\footnote{See
% Note~\ref{itm:noalign} in Section~\ref{sec:drawbacks}}.
%The result of this code is shown in Table~\ref{tab:ex3}.\begin{table}
%\caption{Example~\ref{ex:tab3}}
%\label{tab:ex3}
%\vspace{10pt}
%\begin{center}
%\begin{tabular}{|l|ccc|}
%\hline\bfseries Name &
%\bfseries Assignment 1 &
%\bfseries Assignment 2 &
%\bfseries Total\\\hline\hline
%A. Smith&80&70&150\\
%B. Jones&60&80&140\\
%J. Doe&85&75&160\\\hline
% \multicolumn{1}{l|}{}&75&75&150\\\cline{2-4}
%\end{tabular}
%\end{center}
%\end{table}
%\end{example}
%
% \section{Converting CSV file into longtable environment}
%\DescribeMacro{\CSVtolongtable}
%The command \cmdname{CSVtolongtable} works in the same way as \cmdname{CSVtotabular} but
% creates a \envname{longtable} environment instead of a \envname{tabular} environment.
%
%\begin{example}{Using a longtable environment}{ex:long}
%Suppose the CSV file in the previous example, contains, say, 100 entries.
%This will no longer fit onto one page, so it would be better to use
%\envname{CSVtolongtable} instead. For example:
%\begin{verbatim}
%\CSVtolongtable{sample.csv}{|l|ccc|}{%
%\caption{Student Marks}\label{tab:students}\\
%\hline
%\bfseries Name &
%\bfseries Assignment 1 &
%\bfseries Assignment 2 &
%\bfseries Total\\\hline
%\endfirsthead
%\caption[]{Student Marks}\\
%\hline
%\bfseries Name &
%\bfseries Assignment 1 &
%\bfseries Assignment 2 &
%\bfseries Total\\\hline
%\endhead
%\hline
%\multicolumn{3}{r}{\em Continued on next page}
%\endfoot
%\hline
%\endlastfoot}{%
%\insertName &
%\insertbyname{Assignment 1} &
%\insertbyname{Assignment 2} &
%\insertTotal
%\ifnextrowlast{\\\hline\hline}{\\}}{%
% & \insertbyname{Assignment 1} &
% \insertbyname{Assignment 2} &
%\insertTotal\\}
%\end{verbatim}
%\end{example}
%
% \section{Associated Counters}
%
% Within the \cmdname{CSVtotabular}, \cmdname{CSVtolongtable} and
%\cmdname{applyCSVfile} commands, there are two
% counters, \DescribeMacro{csvlinenum}\DescribeMacro{csvrownumber}%
% \texttt{csvlinenum} and \texttt{csvrownumber}.
% The former, \texttt{csvlinenum}, is the current line number in the CSV
% file, whereas the latter, \texttt{csvrownumber}, is the current
% data row. Of the two counters, \texttt{csvrownumber} is likely to be
% the most useful.
%
% \begin{example}{Stripy Table}{ex:stripy}
%David~Carlisle's \stynamefmt{colortbl} package defines the command
%\cmdname{rowcolor} which enables you to specify the row colour.
%Suppose you want a stripy table\footnote{This is designed as an
%example of how to use the package, not incouragement to produce
%garish tables!},
% this can be achieved as follows:
%\begin{verbatim}
%\CSVtotabular{sample2.csv}{lccc}{%
%\rowcolor{green}\bfseries Name &
%\bfseries Assignment 1 &
%\bfseries Assignment 2 &
%\bfseries Total\\\rowcolor{blue}
%}{%
%\insertName &
%\insertbyname{Assignment 1} &
%\insertbyname{Assignment 2} &
%\insertTotal
%\ifthenelse{\isodd{\value{csvrownumber}}}{%
%\\\rowcolor{green}}{\\\rowcolor{blue}}
%}{%
% &
%\insertbyname{Assignment 1} &
%\insertbyname{Assignment 2} &
%\insertTotal
%}
%\end{verbatim}
%The resulting table is illustrated in Table~\ref{tab:stripy}.
%\begin{table}[htb]
%\caption{Stripy Table Example}
%\label{tab:stripy}
%\vspace{10pt}
%\setcounter{csvrownumber}{0}
%\begin{tabular}{lccc}
%\rowcolor{green}\bfseries Name &
%\bfseries Assignment 1 &
%\bfseries Assignment 2 &
%\bfseries Total\\\rowcolor{blue}
%\refstepcounter{csvrownumber}A. Smith &
%80 &
%70 &
%150
%\ifthenelse{\isodd{\value{csvrownumber}}}{\\\rowcolor{green}}{\\\rowcolor{blue}}
%\refstepcounter{csvrownumber}B. Jones &
%60 &
%80 &
%140
%\ifthenelse{\isodd{\value{csvrownumber}}}{\\\rowcolor{green}}{\\\rowcolor{blue}}
%\refstepcounter{csvrownumber}J. Doe &
%85 &
%75 &
%160
%\ifthenelse{\isodd{\value{csvrownumber}}}{\\\rowcolor{green}}{\\\rowcolor{blue}}
%\refstepcounter{csvrownumber} &
%75 &
%75 &
%150
%\end{tabular}
%\end{table}
%\end{example}
%
%\begin{example}{More Mail Merging}{ex:mailref}
%This is an example of mail merging where the letter reference
%is generated from the value of \texttt{csvrownumber}.  The CSV file
%is as used in Example~\ref{ex:mail} on page~\pageref{ex:mail}.\begin{verbatim}
%\applyCSVfile{details.csv}{%
%\begin{letter}{\insertName\\\insertAddress}
%\opening{Dear \insertName}
%
%\textbf{Ref : } interview.\thecsvrownumber
%
%You are invited to an interview at \insertTime\ on the \insertDate.
%
%\closing{Yours Sincerely}
%\end{letter}}
%\end{verbatim}
%\end{example}
%
% \section{Cross-Referencing}
%
% Labels can be generated using the standard
% \cmdname{label} command, but you will need some way
% to make each label unique.  Example~\ref{ex:ref1}
% does this by using \cmdname{thecsvrownumber},
% whereas Example~\ref{ex:ref2} uses \cmdname{insert}\meta{identifier}.
%
% \begin{example}{Labelling within \cmdname{applyCSVfile}}{ex:ref1}
% Example~\ref{ex:ps} on page~\pageref{ex:ps} can be modified to  label each figure:
%\begin{verbatim}
%\applyCSVfile{sample3.csv}{
%\begin{figure}
%\centerline{\includegraphics{\insertFile}}
%\caption{\insertCaption}
%\label{fig:pic\thecsvrownumber}
%\end{figure}}
%\end{verbatim}
% This example uses
% \verb'\label{fig:pic\thecsvrownumber}', so the first figure
%generated by this \cmdname{applyCSVfile} command will have the
%label \texttt{fig:pic1}, the second \texttt{fig:pic2} etc.
%\end{example}
%
% \begin{example}{Labelling within \cmdname{applyCSVfile}}{ex:ref2}
% Modifying the previous example, we now have:
%\begin{verbatim}
%\applyCSVfile{sample3.csv}{
%\begin{figure}
%\centerline{\includegraphics{\insertFile}}
%\caption{\insertCaption}
%\label{fig:\insertFile}
%\end{figure}}
%\end{verbatim}
% The labels for each figure are now:
% \texttt{fig:circle.ps}, \texttt{fig:rectangle.ps}
% and \texttt{fig:triangle.ps}, respectively.
%\end{example}
%
%\begin{example}{Labelling within \cmdname{CSVtotabular}}{ex:timetogrowth}
% This example is slightly more complicated.
% The CSV file, \texttt{data.csv} looks like:
%\begin{verbatim}
%Incubation Temperature,Incubation Time,Time to Growth
%40,120,40
%40,90,60
%35,180,20
%\end{verbatim}
%The following code generates a table using the data
% with an additional column that generates the experiment
% number. (See note~\ref{itm:csvrownumber}.)
%\begin{verbatim}
%\begin{table}
%\caption{Time to Growth Experiments}
%\label{tab:exp}
%\vspace{10pt}
%\centering
%\CSVtotabular{data.csv}{cccc}{%
% % Header Row
%\bfseries Experiment &
%\bfseries \begin{tabular}{c}Incubation\\Temperature\end{tabular} &
%\bfseries \begin{tabular}{c}Incubation\\Time\end{tabular} &
%\bfseries \begin{tabular}{c}Time\\to\\Growth\end{tabular}\\}{%
% % Middle Rows
%\label{exp:\insertbyname{Incubation Temperature}:\insertbyname{Incubation Time}}
%\thecsvrownumber &
%\insertbyname{Incubation Temperature} &
%\insertbyname{Incubation Time} &
%\insertbyname{Time to Growth} \\}{%
% % Final Row
%\label{exp:\insertbyname{Incubation Temperature}:\insertbyname{Incubation Time}}
%\thecsvrownumber &
%\insertbyname{Incubation Temperature} &
%\insertbyname{Incubation Time} &
%\insertbyname{Time to Growth}}
%\par
%\end{table}
%
%It can be seen from Table~\ref{tab:exp}, that
%Experiment~\ref{exp:35:180} had the shortest time to growth.
%\end{verbatim}
% In this example, each experiment has the corresponding label
% \texttt{exp:}\meta{Incubation Temperature}\texttt{:}\meta{Incubation Time}
% so the first experiment has label \texttt{exp:40:120}, the
% second experiment has the label \texttt{exp:40:90} and the
% third experiment has the label \texttt{exp:35:180}.
%
% Table~\ref{tab:timetogrowth} shows the resulting table for
% this example.
% \begin{table}
%\caption{Time to Growth Experiments}
%\label{tab:timetogrowth}
%\vspace{10pt}
%\centering
%\begin{tabular}{cccc}
%\bfseries Experiment &
%\bfseries \begin{tabular}{c}Incubation\\Temperature\end{tabular} &
%\bfseries \begin{tabular}{c}Incubation\\Time\end{tabular} &
%\bfseries \begin{tabular}{c}Time\\to\\Growth\end{tabular}\\
% 1 & 40 & 120 & 40\\
% 2 & 40 & 90 & 60\\
% 3 & 35 & 180 & 20
%\end{tabular}
%\par
% \end{table}
%\end{example}
%
% The following example is more refined in that it
% takes advantage of the fact that the time to growth data consists
% of integers only, so the experiment with the maximum growth can be
% determined by \LaTeX.
%
% \begin{example}{Labelling within \cmdname{CSVtotabular}}{ex:tablabel}
%\begin{verbatim}
%\newcounter{maxgrowth}
%\newcounter{incT} % incubation temperature
%\newcounter{inct} % incubation time
%
%\begin{table}
%\caption{Time to Growth Experiments}
%\label{tab:exp}
%\vspace{10pt}
%\centering
%\CSVtotabular{data.csv}{cccc}{%
% % Header row
%\bfseries Experiment &
%\bfseries \begin{tabular}{c}Incubation\\Temperature\end{tabular} &
%\bfseries \begin{tabular}{c}Incubation\\Time\end{tabular} &
%\bfseries \begin{tabular}{c}Time\\to\\Growth\end{tabular}\\}{%
% % Middle rows
%\label{exp:\insertbyname{Incubation Temperature}:\insertbyname{Incubation Time}}
%\thecsvrownumber &
%\insertbyname{Incubation Temperature} &
%\insertbyname{Incubation Time} &
%\insertbyname{Time to Growth}%
%\ifthenelse{\value{maxgrowth}<\insertbyname{Time to Growth}}{%
%\setcounter{maxgrowth}{\insertbyname{Time to Growth}}%
%\setcounter{incT}{\insertbyname{Incubation Temperature}}%
%\setcounter{inct}{\insertbyname{Incubation Time}}}{}%
%\\}{%
% % Last row
%\label{exp:\insertbyname{Incubation Temperature}:\insertbyname{Incubation Time}}
%\thecsvrownumber &
%\insertbyname{Incubation Temperature} &
%\insertbyname{Incubation Time} &
%\insertbyname{Time to Growth}%
%\ifthenelse{\value{maxgrowth}<\insertbyname{Time to Growth}}{%
%\setcounter{maxgrowth}{\insertbyname{Time to Growth}}%
%\setcounter{incT}{\insertbyname{Incubation Temperature}}%
%\setcounter{inct}{\insertbyname{Incubation Time}}}{}%
%}
%\par
%\end{table}
%
%As can be seen from Table~\ref{tab:exp},
%Experiment~\ref{exp:\theincT:\theinct}
%had the maximum time to growth, with
%incubation time \theinct,
%incubation temperature \theincT\ and
%time to growth, \themaxgrowth.
%\end{verbatim}
% \end{example}
%
% \section{Saving Entries}
% Entries can be saved using the command:\\[10pt]
% \DescribeMacro{\csvSaveEntry}
%\changes{1.22}{2007 January 3}{added final optional argument to 'csvSaveEntry}
% \cmdname{csvSaveEntry}[\meta{counter}]\{\meta{identifier}\}[\meta{empty text}]\\[10pt]
% where \meta{counter} is a \LaTeX\ counter, by default
%\texttt{csvrownumber},
% and \meta{identifier} is the header entry.  The entry
% can then be used with the command:\\[10pt]
% \DescribeMacro{\csvGetEntry}
% \cmdname{csvGetEntry}\{\meta{counter}\}\{\meta{identifier}\}\\[10pt]
%The final optional argument \meta{empty text} to
%\cmdname{csvSaveEntry} is the text to use if the entry is blank.
%For example, \verb|\csvSaveEntry{Time}[MISSING DATA]| will
%print MISSING DATA if the Time field is blank.
%
%The following example illustrates the use of these commands.
%
% \begin{example}{Saving Entries}{ex:index}
% This example illustrates how you can use one CSV
% file to access data in other CSV files.
% This example has several CSV files:
%\\[10pt]
% File \texttt{index.csv}:
%\begin{verbatim}
%File,Temperature,NaCl,pH
%exp25a.csv,25,4.7,0.5
%exp25b.csv,25,4.8,1.5
%exp30a.csv,30,5.12,4.5
%\end{verbatim}
% File \texttt{exp25a.csv}:
%\begin{verbatim}
%Time,Logcount
%0,3.75
%23,3.9
%45,4.0
%\end{verbatim}
% File \texttt{exp25b.csv}:
%\begin{verbatim}
%Time,Logcount
%0,3.6
%60,3.8
%120,4.0
%\end{verbatim}
% File \texttt{exp30a.csv}:
%\begin{verbatim}
%Time,Logcount
%0,3.73
%23,3.67
%60,4.9
%\end{verbatim}
%
% It is not possible to nest \cmdname{CSVtotabular},
%\cmdname{CSVtolongtable} and \cmdname{applyCSVfile},
% so if you need to go through \texttt{index.csv} and use each file
% named in there, you can first go through \texttt{index.csv}
%storing the information using \cmdname{csvSaveEntry} as follows:
%\begin{verbatim}
%\newcounter{maxexperiments}
%\applyCSVfile{sample5.csv}{%
%\stepcounter{maxexperiments}
%\csvSaveEntry{File}
%\csvSaveEntry{Temperature}
%\csvSaveEntry{NaCl}
%\csvSaveEntry{pH}
%}
%\end{verbatim}
% The counter \texttt{maxexperiments} simply counts the number of
% entries in \texttt{index.csv}.
% The entries can now be used to generate a table for each
% file listed in \texttt{index.csv} (the \cmdname{whiledo} command
%is defined in the \stynamefmt{ifthen} package):
%\begin{verbatim}
%\newcounter{experiment}
%\whiledo{\value{experiment}<\value{maxexperiments}}{%
%\stepcounter{experiment}
%\begin{table}
%\caption{Temperature = \protect\csvGetEntry{experiment}{Temperature},
%NaCl = \protect\csvGetEntry{experiment}{NaCl},
%pH = \protect\csvGetEntry{experiment}{pH}}
%\vspace{10pt}
%\centering
%\CSVtotabular{\csvGetEntry{experiment}{File}}{ll}{%
%Time & Log Count\\}{%
%\insertTime & \insertLogcount\\}{%
%\insertTime & \insertLogcount}
%
%\end{table}
%}
%\end{verbatim}
% Note that \cmdname{csvGetEntry} needs to be \cmdname{protect}ed
%within the \cmdname{caption} command.
%
% This example can be modified if, say, you only want the
% tables where the temperature is 25:
%\begin{verbatim}
%\setcounter{experiment}{0}
%\whiledo{\value{experiment}<\value{maxexperiments}}{%
%\stepcounter{experiment}
%\ifthenelse{\equal{\csvGetEntry{experiment}{Temperature}}{25}}{%
%\begin{table}
%\caption{Temperature = \protect\csvGetEntry{experiment}{Temperature},
%NaCl = \protect\csvGetEntry{experiment}{NaCl},
%pH = \protect\csvGetEntry{experiment}{pH}}
%\vspace{10pt}
%\centering
%\CSVtotabular{\csvGetEntry{experiment}{File}}{ll}{%
%Time & Log Count\\}{%
%\insertTime & \insertLogcount\\}{%
%\insertTime & \insertLogcount}\par
%\end{table}}{}
%}
%\end{verbatim}
% \end{example}
%
%\section{Pie Charts (csvpie.sty)}
%
%\changes{1.2}{2006 Sept 1}{csvpie.sty added}
%If you want to create a pie chart from data stored in a CSV file,
%you can use the \stynamefmt{csvpie} package, distributed with the
%\stynamefmt{csvtools} package. A basic pie chart can be created
%using the command:\\[10pt]
%\DescribeMacro{\csvpiechart}\noindent
%\cmdname{csvpiechart}[\meta{options}]\{\meta{variable}\}\{\meta{filename}\}\\[10pt]
%where \meta{filename} is the name of the CSV file containing the
%data, and \meta{variable} is the command indicating the entry that
%contains the value for the given segment.
%The starred version of \cmdname{csvpiechart} should be used if
%the CSV file has no header row.
%
%The pie charts have ``inner'' labels on the segment, and
%``outer'' labels outside the chart. The labels are given by the
%commands \DescribeMacro{\csvpieinnerlabel}\cmdname{csvpieinnerlabel}
%and \DescribeMacro{\csvpieouterlabel}\cmdname{csvpieouterlabel}.
%The default definitions are:
%\begin{verbatim}
%\newcommand{\csvpieouterlabel}{\field{1}}
%\newcommand{\csvpieinnerlabel}{\field{2}\%}
%\end{verbatim}
%This assumes that the second column contains the data, and the
%first column contains a description, but can be redefined
%as necessary.
%
%The pie chart display can be modified using the optional
%argument to \cmdname{csvpiechart}.
%This argument should be a \meta{key}=\meta{value} list.
%The available keys are as follows:
%\begin{description}
%\item[start] This should be an integer specifying the starting angle
%of the first segment. This is 0 by default.
%
%\item[total] This should be an integer specifying the sum of all
%the segment values. This is 100 by default.
%
%\item[radius] This should be a length specifying the radius of the
%pie chart. (Default: 2cm)
%
%\item[inner] This should be a fraction specifying the relative distance
%along the radius to start the inner label. (Default: 0.25)
%
%\item[outer] This should be a fraction specifying the relative distance
%along the radius to start the outer label. (Default: 1.25)
%
%\item[cutaway] This should be a comma-separated list of numbers
%corresponding to the segments that should be cut away from the
%rest of the pie chart. Since the value may contain commas, the value
%should always be enclosed in braces. Ranges may also be used.
%If a range is used, all the segments in the given range are
%kept together, so, for example, \verb"cutaway={1,2}" will separate
%the first two segments from the pie chart, and the two segments
%will also be separated from each other, whereas \verb"cutaway={1-2}"
%will separate the first two segments from the pie chart, but will
%keep the two segments together.
%
%\item[offset] This should be a fraction specifying the
%relative distance along the radius to shift the cut away
%segments. (Default: 0.1)
%
%\item[firstrow] This should be the number of the first row
%containing the actual data. This is equivalent to the optional
%argument of \cmdname{applyCSVfile} or \cmdname{applyCSVfile*}.
%\end{description}
%
%Note that \TeX\ performs integer arithmetic. Although the CSV
%file may contain decimal numbers, rounding will occur when
%constructing the pie charts.
%
%The colours for the pie chart segments can be set using the
%command\DescribeMacro{\csvpiesegmentcol}:\\[10pt]
%\cmdname{csvpiesegmentcol}\{\meta{n}\}\{\meta{colour}\}\\[10pt]
%where \meta{n} is the segment number, and \meta{colour} is a
%defined colour name. For example, if you want the first segment
%in the pie chart to be yellow, do:
%\begin{verbatim}
%\csvpiesegmentcol{1}{yellow}
%\end{verbatim}
%There are 8 predefined segment colours, if your pie chart has
%more than 8 segments, you will need to specify the remainder.
%
%You can obtain the colour name for a given segment
%using\DescribeMacro{\csvpiesegcolname}:\\[10pt]
%\cmdname{csvpiesegcolname}\{\meta{n}\}\\[10pt]
%where \meta{n} is the segment number. The \cmdname{csvpiechart}
%command uses \cmdname{applyCSVfile}, so the \texttt{csvrownumber}
%counter can be used. This means that you can change the text
%colour of the outer label to match the segment. For example:
%\begin{verbatim}
%\renewcommand{\csvpieouterlabel}{%
%\color{\csvpiesegcolname{\value{csvrownumber}}}\field{2}}
%\end{verbatim}
%Note that \cmdname{value} must be used since \meta{n} has to be
%a number.
%
%If you want grey pie charts, either use the \texttt{monochrome}
%package option:
%\begin{verbatim}
%\usepackage[monochrome]{csvpie}
%\end{verbatim}
%or use the command \DescribeMacro{\colorpiechartfalse}\cmdname{colorpiechartfalse} prior to using \cmdname{csvpiechart}.
%To switch back to colour pie charts, use
%\DescribeMacro{\colorpiecharttrue}\cmdname{colorpiecharttrue}.
%
%\begin{example}{A Pie Chart}{ex:piechart}
%Given a CSV file
%(called \texttt{fruit.csv}) containing:
%\begin{verbatim}
%Name,Value
%Apples,20
%Pears,15
%"lemons,limes",30.5
%Peaches,24.5
%Cherries,10
%\end{verbatim}
%Then the value for each segment is given by the second column, so
%\meta{variable} should be \verb"\field{2}" or \verb|\insertValue|.
%The pie charts shown in Figure~\ref{fig:piechart} can be created
%using:
%\begin{verbatim}
% % Change the way the labels are displayed
%\renewcommand{\csvpieinnerlabel}{\sffamily\insertValue\%}
%\renewcommand{\csvpieoutlabel}{%
%\color{\csvpiesegcolname{\value{csvrownumber}}}\sffamily\insertName}
%
%\begin{figure}
%\begin{center}
%\begin{tabular}{cc}
%\csvpiechart[start=45,cutaway={1,2}]{\insertValue}{fruit.csv} &
%\csvpiechart[start=45,cutaway={1-2}]{\insertValue}{fruit.csv} \\
%(a) & (b)
%\end{tabular}
%\end{center}
%\caption{Pie Chart Example (a) cutaway=\{1,2\} (b) cutaway=\{1-2\}}
%\end{figure}
%\end{verbatim}
%The inner and outer labels have been
%redefined to use a sans-serif font, and the outer label is in
%the same colour as its corresponding segment.
%Both pie charts have a starting angle of $45^\circ$, and theyhave the first two segments cutaway, but in (a)
%the first two segments are separated from each other, whereas in
%(b), the first two segments are joined, although separated
%from the rest of the pie chart.
%
%If the CSV file has no header row, the starred version should be
%used, e.g.:
%\begin{verbatim}
%\csvpiechart*[cutaway={1-2}]{\field{2}}{fruit.csv}
%\end{verbatim}
%
%\begin{figure}[htbp]
%\begin{center}
%\begin{tabular}{cc}
%\csvpiesetkeys{start=45,cutaway={1,2}}%
%\edef\csvstartang{\thecsvstartangle}\relax
%\setcounter{csvrownumber}{0}\relax
%\stepcounter{csvrownumber}\relax
%\csvsetsegmentparams{\thecsvrownumber}{20}\relax
%\stepcounter{csvrownumber}\relax
%\csvsetsegmentparams{\thecsvrownumber}{15}\relax
%\stepcounter{csvrownumber}\relax
%\csvsetsegmentparams{\thecsvrownumber}{30.5}\relax
%\stepcounter{csvrownumber}\relax
%\csvsetsegmentparams{\thecsvrownumber}{24.5}\relax
%\stepcounter{csvrownumber}\relax
%\csvsetsegmentparams{\thecsvrownumber}{10}\relax
%\setcounter{csvstartangle}{\csvstartang}\relax
%\ifthenelse{\equal{\csvpiecutaways}{}}{}{\csvcomputeoffsets}\relax
%\begin{tikzpicture}
%\setcounter{csvrownumber}{0}\relax
%\stepcounter{csvrownumber}\relax
%\renewcommand*{\csvpieinnerlabel}{\sffamily20\%}
%\renewcommand*{\csvpieouterlabel}{\color{\csvpiesegcolname{\value{csvrownumber}}}\sffamily
%Apples}
%\csvpiesegment{20}
%\stepcounter{csvrownumber}\relax
%\renewcommand*{\csvpieinnerlabel}{\sffamily15\%}
%\renewcommand*{\csvpieouterlabel}{\color{\csvpiesegcolname{\value{csvrownumber}}}\sffamily
%Pears}
%\csvpiesegment{15}
%\stepcounter{csvrownumber}\relax
%\renewcommand*{\csvpieinnerlabel}{\sffamily30.5\%}
%\renewcommand*{\csvpieouterlabel}{\color{\csvpiesegcolname{\value{csvrownumber}}}\sffamily
%lemons,limes}
%\csvpiesegment{30.5}
%\stepcounter{csvrownumber}\relax
%\renewcommand*{\csvpieinnerlabel}{\sffamily24.5\%}
%\renewcommand*{\csvpieouterlabel}{\color{\csvpiesegcolname{\value{csvrownumber}}}\sffamily
%Peaches}
%\csvpiesegment{24.5}
%\stepcounter{csvrownumber}\relax
%\renewcommand*{\csvpieinnerlabel}{\sffamily10\%}
%\renewcommand*{\csvpieouterlabel}{\color{\csvpiesegcolname{\value{csvrownumber}}}\sffamily
%Cherries}
%\csvpiesegment{10}
%\end{tikzpicture}
% &
%\csvpiesetkeys{start=45,cutaway={1-2}}%
%\edef\csvstartang{\thecsvstartangle}\relax
%\setcounter{csvrownumber}{0}\relax
%\stepcounter{csvrownumber}\relax
%\csvsetsegmentparams{\thecsvrownumber}{20}\relax
%\stepcounter{csvrownumber}\relax
%\csvsetsegmentparams{\thecsvrownumber}{15}\relax
%\stepcounter{csvrownumber}\relax
%\csvsetsegmentparams{\thecsvrownumber}{30.5}\relax
%\stepcounter{csvrownumber}\relax
%\csvsetsegmentparams{\thecsvrownumber}{24.5}\relax
%\stepcounter{csvrownumber}\relax
%\csvsetsegmentparams{\thecsvrownumber}{10}\relax
%\setcounter{csvstartangle}{\csvstartang}\relax
%\ifthenelse{\equal{\csvpiecutaways}{}}{}{\csvcomputeoffsets}\relax
%\begin{tikzpicture}
%\setcounter{csvrownumber}{0}\relax
%\stepcounter{csvrownumber}\relax
%\renewcommand*{\csvpieinnerlabel}{\sffamily20\%}
%\renewcommand*{\csvpieouterlabel}{\color{\csvpiesegcolname{\value{csvrownumber}}}\sffamily
%Apples}
%\csvpiesegment{20}
%\stepcounter{csvrownumber}\relax
%\renewcommand*{\csvpieinnerlabel}{\sffamily15\%}
%\renewcommand*{\csvpieouterlabel}{\color{\csvpiesegcolname{\value{csvrownumber}}}\sffamily
%Pears}
%\csvpiesegment{15}
%\stepcounter{csvrownumber}\relax
%\renewcommand*{\csvpieinnerlabel}{\sffamily30.5\%}
%\renewcommand*{\csvpieouterlabel}{\color{\csvpiesegcolname{\value{csvrownumber}}}\sffamily
%lemons,limes}
%\csvpiesegment{30.5}
%\stepcounter{csvrownumber}\relax
%\renewcommand*{\csvpieinnerlabel}{\sffamily24.5\%}
%\renewcommand*{\csvpieouterlabel}{\color{\csvpiesegcolname{\value{csvrownumber}}}\sffamily
%Peaches}
%\csvpiesegment{24.5}
%\stepcounter{csvrownumber}\relax
%\renewcommand*{\csvpieinnerlabel}{\sffamily10\%}
%\renewcommand*{\csvpieouterlabel}{\color{\csvpiesegcolname{\value{csvrownumber}}}\sffamily
%Cherries}
%\csvpiesegment{10}
%\end{tikzpicture}
% \\
%(a) & (b)
%\end{tabular}
%\end{center}
%\caption{Pie Chart Example (a) cutaway=\{1,2\}
%(b) cutaway=\{1-2\}}
%\label{fig:piechart}
%\end{figure}
%\end{example}
%
%\section{Sorting Data (csvsort.sty)}
%
%The \stynamefmt{csvsort} package (which forms part of the
%\stynamefmt{csvtools} bundle) provides analogous commands to
%those provided by \stynamefmt{csvtools}, but the data is first
%sorted. The \stynamefmt{csvsort} package needs to be loaded
%separately in order to access the necessary commands. The package
%options should be a list of key=value pairs, where the available
%keys are:
%\begin{description}
%\item[verbose] Verbose mode. This is a boolean key. If set, the
%comparisons performed by the insertion sort code are printed to
%the screen. (Default: \texttt{verbose=true}.)
%
%\item[sort] This key specifies how to sort the data. It may take
%one of the following values:
%\begin{itemize}
%\item \texttt{alphabetical ascending} (or just \texttt{alphabetical})
%\item \texttt{alphabetical descending}
%\item \texttt{numerical ascending} (or just \texttt{numerical})
%\item \texttt{numerical descending}
%\end{itemize}
%(Default: \texttt{sort=alphabetical ascending})
%
%\item[variable] The sort variable. (Default: \verb|sort=\field{1}|)
%
%\item[sfirstdataline] The line on which the data starts in a data
%file without a header row. (Default: \texttt{sfirstdataline=1}.)
%
%\item[firstdataline] The line on which the data starts in a data
%file with a header row. (Default: \texttt{firstdataline=2}.)
%\end{description}
%
%Note that the \stynamefmt{csvsort} package requires
%the \stynamefmt{xfor} package and \'Eamonn McManus'
%\texttt{compare.tex} file. The \stynamefmt{csvsort} package uses
%an insertion sort method to sort the data, so large amounts of data
%may slow processing time. The following commands are provided
%by \stynamefmt{csvsort}:
%
%\DescribeMacro{\sortapplyCSVfile}\noindent
% \cmdname{sortapplyCSVfile}\oarg{options}\marg{filename}\marg{text}\par
% \cmdname{sortapplyCSVfile*}\oarg{options}\marg{filename}\marg{text}\\[10pt]
%These commands are analogous to \cmdname{applyCSVfile} and
%\cmdname{applyCSVfile*}, except that the data is first sorted.
%The optional argument is a key=value list. The keys are the same
%as those used in the package options, described above. These options
%only apply to the given instance of the command, whereas the
%package options apply to all \stynamefmt{csvsort} commands, unless
%overridden in \meta{options}. Example, suppose you have a file
%called \texttt{unsorted.csv} which looks like:
%\begin{verbatim}
%First Name,Surname,Age
%Zephram,Lang,60
%Fred,Lang,10
%Barney,Langley,25
%Jane,Brown,5
%Adam,Smith,24
%Bert,Jones,18
%\end{verbatim}
%Then
%\begin{verbatim}
%\sortapplyCSVfile[sort=alphabetical,variable=\insertSurname]{unsorted.csv}{%
%\insertSurname, \insertbyname{First Name}. Age: \insertAge\par}
%\end{verbatim}
%will produce the following output:\par
%Brown, Jane. Age:5\par
%Jones, Bart. Age:18\par
%Lang, Zephram. Age:60\par
%Lang, Fred. Age:10\par
%Langley, Barney. Age:25\par
%Smith, Adam. Age: 24\par
%Note that the data has only been sorted according to the surname.
%To sort first by surname, then by first name, you can
%do something like:
%\begin{verbatim}
%\sortapplyCSVfile[sort=alphabetical,
%variable={\insertSurname,\insertbyname{First Name}}]{unsorted.csv}{%
%\insertSurname, \insertbyname{First Name}. Age: \insertAge\par}
%\end{verbatim}
%As with \cmdname{applyCSVfile}, you must use \cmdname{field} if
%you use the starred version:
%\begin{verbatim}
%\sortapplyCSVfile*[sort=alphabetical,
%variable={\field{2},\field{1}}]{unsorted.csv}{%
%\field{2}, \field{1}. Age: \field{3}\par}
%\end{verbatim}
%
%The commands:\\[10pt]\DescribeMacro{\sortCSVtotabular}
%\cmdname{sortCSVtotabular}\oarg{options}\marg{filename}\marg{col-spec}\marg{first row}\marg{all but last row}\marg{last row}\\
%\DescribeMacro{\sortCSVtolongtable}
%\cmdname{sortCSVtolongtable}\oarg{options}\marg{filename}\marg{col-spec}\marg{first row}\marg{all but last row}\marg{last row}\\
%Are analogous to \cmdname{CSVtotabular} and \cmdname{CSVtolongtable},
%where, again, \meta{options} is a list of key=value pairs, the same
%as \cmdname{sortapplyCSVfile}. Using the same example data as above,
%the following command will sort the data according to age (in
%numerical order) and place in a tabular environment:
%\begin{verbatim}
%\sortCSVtotabular[sort=numerical,variable=\insertAge]{unsorted.csv}{llr}{%
%\bfseries Surname & \bfseries First Name & \bfseries Age\\}{%
%\insertSurname & \insertbyname{First Name} & \insertAge\\}{%
%\insertSurname & \insertbyname{First Name} & \insertAge}
%\end{verbatim}
%
%Note that the counter \texttt{csvlinenum} has no meaning in the
%commands provided by the \stynamefmt{csvsort} package. The
%\texttt{csvrownumber} counter corresponds to the sorted data row.
%
%\section{The csvtools.pl Perl Script}
%
% Suppose you have several large CSV files, and you have included
%the information into your document using \cmdname{applyCSVfile},
%\cmdname{CSVtolongtable}, \cmdname{CSVtotabular} or
%\cmdname{csvpiechart}, which has
%made life so much easier for you, but you are now required by a
%journal to submit your source code in a single \texttt{.tex} file.
% They don't want all your CSV files, so what do you do?  If you
%have Perl installed on your system you can use the
%\texttt{csvtools.pl} Perl script.  This has the following syntax:\\[5pt]
% \texttt{csvtools.pl} \meta{in-file} \meta{out-file}\\[5pt]
% where \meta{in-file} is the name of your file that contains the
%\cmdname{applyCSVfile}, \cmdname{CSVtotabular} etc
% commands, and \meta{out-file} is a new file which will be created by \texttt{csvtools.pl}.  This new
% file will be the same as \meta{in-file} except that all
%occurances of \cmdname{applyCSVfile}, \cmdname{CSVtolongtable},
%\cmdname{CSVtotabular} and \cmdname{csvpiechart} will be replaced
%by the relevant data extracted from the named CSV files.
%
%\begin{example}{csvtools.pl --- Aligning Data}{ex:perl}
% Suppose the file \texttt{mydoc.tex} contains the code given in
%Example~\ref{ex:tab1}, with the associated CSV file
%\texttt{sample.csv} also given in that example.  Then if you do:
%\begin{verbatim}
%csvtools.pl mydoc.tex mydocnew.tex
%\end{verbatim}
%the file \texttt{mydocnew.tex} will be created which will be
%identical to \texttt{mydoc.tex} except the lines
% containing the code \verb'\CSVtotabular{sample.csv}{lccc}{'\ldots\verb/}{/\dots\verb!}{!\ldots\verb+}+ will be replaced
%with the lines:
%\begin{verbatim}
% % \CSVtotabular{sample.csv}... converted using csvtools.pl
% %>> START INSERT
% \begin{tabular}{lccc}
% \bfseries Name &
% \bfseries Assignment 1 &
% \bfseries Assignment 2 &
% \bfseries Total\\
% A. Smith&80&70&150\\
% B. Jones&60&80&140\\
% J. Doe&85&75&160\\
%  &75&75&150
% \end{tabular}%<< END INSERT
%\end{verbatim}
%\end{example}
%
%Similarly, \texttt{csvtools.pl} will substitute all occurrances
%of \cmdname{CSVtolongtable}, \cmdname{applyCSVfile}
%and \cmdname{csvpiechart}.
%
%\subsection{Notes}
%\begin{enumerate}
%\item If perl is located in a directory other than \texttt{/usr/bin/}
%you will need to edit the first line of \texttt{csvtools.pl} as
%appropriate. You can find the location using the command:
%\begin{verbatim}
%which perl
%\end{verbatim}
%
%\item If you can't directly execute a Perl script,
% you can do:\\[5pt]
% \texttt{perl csvtools.pl} \meta{in-file} \meta{out-file}
%
%\item You must first \LaTeX\ your document before using
%\texttt{csvtools.pl} as it checks the log file for any counters
%that have been defined.
%
%\item \texttt{csvtools.pl} only knows about a very limited set
%of \LaTeX\ commands.  It should be able to understand:
%\begin{verbatim}
%\CSVtotabular{\csvGetEntry{experiment}{File}}{ll}{...
%\end{verbatim}
%(see Example~\ref{ex:index}), but it won't be able to understand,
%say,
%\begin{verbatim}
%\newcommand{\filename}{\csvGetEntry{experiment}{File}}
%\CSVtotabular{\filename}{ll}{...
%\end{verbatim}
%It can pick up on \cmdname{addtocounter}, \cmdname{stepcounter},
%\cmdname{refstepcounter} and \cmdname{setcounter} but only if
%they are used explicitly in the named \texttt{.tex} file. (It
%ignores any files that have been included using
%\cmdname{input}, \cmdname{include} etc.)
%
%\item This Perl script has only been tested under Linux, but it
%ought to work under other systems.
%\end{enumerate}
%
% \section{Bugs/Drawbacks/``Features''}\label{sec:drawbacks}
% \begin{enumerate}
% \item\label{itm:insert}
% The package doesn't check to see whether
%\cmdname{insert}\meta{identifier} exists, otherwise you would not
%be able to use multiple CSV files with the same headers, as in
%Example~\ref{ex:index}.  Therefore it is recommended that
% you check to make sure that the command does not already exist.
% For example, the \TeX\ commands \cmdname{insert} and
%\cmdname{insertpenalties} already exist, so a blank header or a
%header named \texttt{penalties} would cause problems. (These two
%will now cause an error as from version 1.1, but it's something
%bear in mind.)
%
% \item Note also that \cmdname{insertbyname} doesn't check
% if you've given a valid label, so if no text appears,
% check you've spelt it correctly, checking punctuation, spaces and case.
%
% \item\label{itm:psbyname}
% Note that in Example~\ref{ex:ps}, replacing line~3 with:
%\begin{verbatim}
%\centerline{\includegraphics{\insertbyname{File}}}
%\end{verbatim}
% will cause an error, as \verb/\insertbyname{File}/ doesn't get
% fully expanded by the time it gets passed to
% \cmdname{includegraphics}, and will prevent
%\cmdname{includegraphics} from
% finding the file.  It is possible to get around this using
% \TeX's \cmdname{edef} command:
%\begin{verbatim}
%\edef\psfilename{\insertbyname{File}}
%\centerline{\includegraphics{\psfilename}}
%\end{verbatim}
%
% \item\label{itm:noalign} You can't have commands like
%\cmdname{hline}, \cmdname{cline} and \cmdname{multicolumn} in the
%first column of the \meta{middle} or \meta{last} code of
%\cmdname{CSVtotabular} or \cmdname{CSVtolongtable}.  If you do,
%it will generate a \verb/misplaced \noalign/ error, instead you
%need to put it at the end of the \meta{first} or \meta{middle}
%code.  (See Example~\ref{ex:tab3}.)
%
% \item You can't have nested \cmdname{applyCSVfile},
%\cmdname{CSVtolongtable} and \cmdname{CSVtotabular}
% commands. Nor can you have \cmdname{csvpiechart} within
%one of these commands (See Example~\ref{ex:index})
%
% \item If the CSV file has a header row, it must be on the first
%line.
%
% \item It is possible for \TeX\ to run out of memory if you use
% \cmdname{csvSaveEntry} on a large file.
%
% \item\label{itm:csvrownumber} In version 1.0, there was an
%inconsistency with \texttt{csvrownumber} within
%\cmdname{applyCSVfile} and \cmdname{CSVtotabular}.  In the
%former it excluded the header row, whereas the latter
%included it.  This has been changed in version 1.1 so that within
%\cmdname{applyCSVfile}, \cmdname{CSVtotabular} and
%\cmdname{CSVtolongtable}, \texttt{csvrownumber} refers to the
%data row (excluding header row.)  I hope this doesn't cause
%problems, but it makes more sense that they should be
%consistent.  So if you have no blank lines in your CSV file,
%\texttt{csvrownumber} should always be 1 more than
%\texttt{csvlinenumber}.
%
%\end{enumerate}
%
% \section{Contact Details}
%
% Dr Nicola Talbot\\
% School of Computing Sciences\\
% University of East Anglia\\
% Norwich.  NR4 7TJ. England.
% \\[10pt]
% \url{http://theoval.cmp.uea.ac.uk/~nlct/}
%
%\StopEventually{\PrintIndex\addcontentsline{toc}{section}{Index}}
%
%
%
%\section{The Code}
%\iffalse
%    \begin{macrocode}
%<*csvtools.sty>
%    \end{macrocode}
%\fi
%\subsection{csvtools.sty}
% Declare package
%    \begin{macrocode}
\NeedsTeXFormat{LaTeX2e}
\ProvidesPackage{csvtools}[2007/07/03 v1.24 (NLCT)]
%    \end{macrocode}
% Required packages
%    \begin{macrocode}
\RequirePackage{ifthen}
\RequirePackage{longtable}
%    \end{macrocode}
% Define command to set the separator character.
%    \begin{macrocode}
\newcommand*{\@csvseparator}{,}
\newcommand*{\setcsvseparator}[1]{%
\renewcommand*{\@csvseparator}{#1}%
\construct@lopoff{#1}%
\@construct@qlopoff{#1}%
\@construct@lopoff{#1}}
%    \end{macrocode}
% Define command to remove first element from list. CSV files
% usually use double quote characters to enclose entries containing
% a comma. The first argument is the control sequence containing
% the list, the second argument is the control sequence to
% contain the first item in the list.
%    \begin{macrocode}
\newcommand*{\extractentry}[2]{%
\expandafter\csvlopoff#1\to{#1}{#2}}
%    \end{macrocode}
% The definitions are constructed on the fly to allow for
% different separators. The argument is the separator.
%\changes{1.23}{2007 May 23}{'csvlopoff changed to a long command}
%    \begin{macrocode}
\edef\construct@lopoff#1{%
\noexpand\long\noexpand\def\noexpand\csvlopoff#1##1##2\noexpand\to##3##4{%
\noexpand\ifx"##1\noexpand\relax
\noexpand\@csv@qlopoff#1##1##2\noexpand\to##3##4\relax
\noexpand\else
\noexpand\@csv@lopoff#1##1##2\noexpand\to##3##4\relax
\noexpand\fi
}}
%    \end{macrocode}
% Command to contruct control sequence to be used when the
% entry is surrounded by double quotes.
%\changes{1.23}{2007 May 23}{'@csv@qlopoff changed to a long command}
%    \begin{macrocode}
\edef\@construct@qlopoff#1{%
\noexpand\long\noexpand\def\noexpand\@csv@qlopoff#1"##1"#1##2\noexpand\to##3##4{%
\noexpand\def##4{##1}\noexpand\def##3{#1##2}}}
%    \end{macrocode}
% Command to construct control sequence to be used when the
% entry isn't surrounded by double quotes.
%\changes{1.23}{2007 May 23}{'@csv@lopoff changed to a long command}
%    \begin{macrocode}
\edef\@construct@lopoff#1{%
\noexpand\long\noexpand\def\noexpand\@csv@lopoff#1##1#1##2\noexpand\to##3##4{%
\noexpand\def##4{##1}\noexpand\def##3{#1##2}}}
%    \end{macrocode}
% Set the default separator:
%    \begin{macrocode}
\setcsvseparator{,}
%    \end{macrocode}
% Define some variables
%    \begin{macrocode}
\newread\csvfile
\newif\ifmore
\newcount\c@field
\newcount\c@maxfields
\newcounter{csvlinenum}
\newcount\maxlines
\newcount\csvlastbutone
\newcounter{csvrownumber}
\toksdef\csv@tb=2
%    \end{macrocode}
% Define commands to access element in current row.
% Access element by column number:
%    \begin{macrocode}
\newcommand{\field}[1]{\csname @field\romannumeral#1\endcsname}
%    \end{macrocode}
% Set field (header) label, the first argument is the index, second argument is the header for that column.)
%    \begin{macrocode}
\newcommand{\setcsvfieldlabel}[2]{%
\expandafter\def\csname @fieldlabel\romannumeral#1\endcsname{#2}}
%    \end{macrocode}
% Access element by header:
%    \begin{macrocode}
\newcommand{\insertbyname}[1]{\csname insert#1\endcsname}
%    \end{macrocode}
% Define command to trim trailing space (caused by EOL) in argument.
% Argument should be a command name containing the current line.
%    \begin{macrocode}
\newcommand{\trim}[1]{\def\@trmstr{}%
\def\csv@trmarg{#1}%
\if\par\csv@trmarg
\else
\expandafter\@trim#1\end
\let#1\@trmstr
\fi
}
\long\def\@trim#1 \end{\def\@trmstr{#1}}
%    \end{macrocode}
% Define "\applyCSVfile"
%    \begin{macrocode}
\newcommand{\applyCSVfile}{\@ifstar\@sapplyCSVfile\@applyCSVfile}
%    \end{macrocode}
% Starred version of "\applyCSVfile":
%    \begin{macrocode}
\newcommand{\@sapplyCSVfile}[3][1]{%
\IfFileExists{#2}{%
\openin\csvfile=#2
\ifeof\csvfile\morefalse\else\moretrue\fi
\ifmore
{\c@csvlinenum=1\relax
\global\c@csvrownumber=0\relax
\loop
\read\csvfile to\csvline
\advance\c@csvlinenum by 1\relax
\ifnum\c@csvlinenum>#1\relax
\trim{\csvline}%
\csv@tb=\expandafter{\csvline}%
\edef\@csvlin@{\@csvseparator\the\csv@tb\@csvseparator}%
\c@field = 0\relax
\whiledo{\not\equal{\@csvlin@}{\@csvseparator\par\@csvseparator} \and
\not\equal{\@csvlin@}{\@csvseparator\@csvseparator} \and
\not\equal{\@csvlin@}{\@csvseparator}}{%
\extractentry{\@csvlin@}{\param}%
\csv@tb=\expandafter{\param}%
\advance\c@field by 1\relax
\expandafter\xdef\csname @field\romannumeral\c@field\endcsname{%
\the\csv@tb}%
}%
\ifthenelse{\not\equal{\csvline}{\par} \and
\not\equal{\csvline}{}}{%
\refstepcounter{csvrownumber}\relax
#3}{}%
\fi
\ifeof\csvfile\morefalse\else\moretrue\fi
\ifmore
\repeat}%
\closein\csvfile
\fi
}{\PackageError{csvtools}{Can't find file '#2'}{}}%
}
%    \end{macrocode}
% Unstarred version:
%    \begin{macrocode}
\newcommand{\@applyCSVfile}[3][2]{%
\ifnum#1<2\relax
\PackageError{csvtools}{Header line required}{The optional
argument to `\string\applyCSVfile' needs to be > 1.  The header
line should be on line 1}%
\else
\IfFileExists{#2}{%
\openin\csvfile=#2
\ifeof\csvfile\morefalse\else\moretrue\fi
\ifmore
\global\c@csvlinenum=1\relax
\global\c@csvrownumber=0\relax
\global\c@maxfields=0\relax
{\loop
\read\csvfile to\csvline
\trim{\csvline}%
\ifnum\c@csvlinenum=1\relax
\ifthenelse{\not\equal{\csvline}{\par} \and
\not\equal{\csvline}{}}{%
\csv@tb=\expandafter{\csvline}\relax
\edef\@csvlin@{\@csvseparator\the\csv@tb\@csvseparator}%
\c@field=0\relax
\whiledo{\not\equal{\@csvlin@}{\@csvseparator\par\@csvseparator}
\and \not\equal{\@csvlin@}{\@csvseparator\@csvseparator}
\and \not\equal{\@csvlin@}{\@csvseparator}}{%
\extractentry{\@csvlin@}{\param}%
\advance\c@field by 1\relax
\csv@tb=\expandafter{\param}\relax
\expandafter
\xdef\csname @fieldlabel\romannumeral\c@field\endcsname{\the\csv@tb}%
}%
\c@maxfields=\c@field
}{%
\PackageError{csvtools}{Header line missing in file #2}{Header
line required on line 1}}%
\fi
\c@field=0\relax
\whiledo{\c@field < \c@maxfields}{%
\advance\c@field by 1\relax
\expandafter\let\csname @field\romannumeral\c@field\endcsname\relax%
\edef\@fieldlabel{%
\csname @fieldlabel\romannumeral\c@field\endcsname}%
\ifthenelse{\equal{\@fieldlabel}{}
\TE@or \equal{\@fieldlabel}{penalties}
\TE@or \equal{\@fieldlabel}{byname}}{\PackageError{csvtools}{%
\string\insert\@fieldlabel \space already defined}{%
You can't have the label '\@fieldlabel' in your header row.}}{}%
\expandafter\let\csname insert\@fieldlabel\endcsname\relax%
}%
\advance\c@csvlinenum by 1\relax
\ifnum\c@csvlinenum>#1\relax
\csv@tb=\expandafter{\csvline}\relax
\edef\@csvlin@{\@csvseparator\the\csv@tb\@csvseparator}%
\c@field=0\relax
\whiledo{\not\equal{\@csvlin@}{\@csvseparator\par\@csvseparator}
\and \not\equal{\@csvlin@}{\@csvseparator\@csvseparator}
\and \not\equal{\@csvlin@}{\@csvseparator}}{%
\extractentry{\@csvlin@}{\param}%
\csv@tb=\expandafter{\param}\relax
\advance\c@field by 1\relax
\expandafter
\xdef\csname @field\romannumeral\c@field\endcsname{\the\csv@tb}%
\edef\@fieldlabel{%
\csname @fieldlabel\romannumeral\c@field\endcsname}%
\ifthenelse{\equal{\@fieldlabel}{}
\TE@or \equal{\@fieldlabel}{penalties}
\TE@or \equal{\@fieldlabel}{byname}}{\PackageError{csvtools}{%
\string\insert\@fieldlabel \space already defined}{%
You can't have the label '\@fieldlabel' in your header row.}}{}%
\expandafter\xdef\csname insert\@fieldlabel\endcsname{\the\csv@tb}%
}%
\ifthenelse{\not\equal{\csvline}{\par}
\and \not\equal{\csvline}{}}{%
\global\advance\c@csvrownumber by 1\relax
#3}{}%
\fi
\ifeof\csvfile\morefalse\else\moretrue\fi
\ifmore
\repeat}%
\closein\csvfile
\fi
}{\PackageError{csvtools}{Can't find file '#2'}{}}%
\fi
}
%    \end{macrocode}
% Define "\CSVtotabular". Inserts contents of csv file into a tabular
% environment.
%    \begin{macrocode}
\newcommand{\CSVtotabular}[5]{%
\openin\csvfile=#1
\c@csvlinenum=0\relax
\loop
\advance\c@csvlinenum by 1\relax
\read\csvfile to\csvline
\expandafter
\xdef\csname @csvline\romannumeral\c@csvlinenum\endcsname{\csvline}%
\ifthenelse{\not\equal{\csvline}{\par}}{\trim{\csvline}}{}%
\ifnum\c@csvlinenum=1\relax
\csv@tb=\expandafter{\csvline}
\edef\@csvlin@{\@csvseparator\the\csv@tb\@csvseparator}%
\c@field = 0\relax
\whiledo{\not\equal{\@csvlin@}{\@csvseparator\par\@csvseparator}%
\and \not\equal{\@csvlin@}{\@csvseparator\@csvseparator}%
\and \not\equal{\@csvlin@}{\@csvseparator}}{%
\extractentry{\@csvlin@}{\param}%
\csv@tb=\expandafter{\param}%
\advance\c@field by 1\relax
\expandafter\xdef\csname insert\the\csv@tb\endcsname{%
\noexpand\field{\the\c@field}}%
}%
\fi
\csv@tb=\expandafter{\csvline}%
\edef\@csvlin@{\@csvseparator\the\csv@tb\@csvseparator}%
\c@field = 0\relax
\whiledo{\not\equal{\@csvlin@}{\@csvseparator\par\@csvseparator}
\and \not\equal{\@csvlin@}{\@csvseparator\@csvseparator}
\and \not\equal{\@csvlin@}{\@csvseparator}}{%
\extractentry{\@csvlin@}{\param}%
\csv@tb=\expandafter{\param}
\advance\c@field by 1\relax
\expandafter
\xdef\csname @l\romannumeral\c@csvlinenum @field\romannumeral\c@field\endcsname{\the\csv@tb}%
}%
\ifeof\csvfile\morefalse\else\moretrue\fi
\ifmore
\repeat
\closein\csvfile
{%
\def\field##1{%
\csname @l\romannumeral\c@csvlinenum @field\romannumeral##1\endcsname}%
\def\@r@wh{\begin{tabular}{#2}#3}%
\def\@r@w{#4}%
\def\@r@wl{#5\end{tabular}}%
\maxlines=\c@csvlinenum
\advance\maxlines by -1\relax
\csvlastbutone=\maxlines
\advance\csvlastbutone by -1\relax
\c@csvlinenum=1\relax
\setcounter{csvrownumber}{0}\relax
\whiledo{\not{\c@csvlinenum>\maxlines}}{%
\ifthenelse{\expandafter\equal{%
\csname @csvline\romannumeral\c@csvlinenum\endcsname}{\par}}%
{\relax}{%
\ifnum\c@csvlinenum=1\relax
\@r@wh
\else
\refstepcounter{csvrownumber}%
\ifnum\c@csvlinenum=\maxlines\@r@wl\else\@r@w\fi
\fi}%
\global\advance\c@csvlinenum by 1\relax
}\relax
}}
%    \end{macrocode}
% Define "\CSVtolongtable". Similar to the above, but uses a
% longtable environment.
%    \begin{macrocode}
\newcommand{\CSVtolongtable}[5]{%
\openin\csvfile=#1
\c@csvlinenum=0\relax
\loop
\advance\c@csvlinenum by 1\relax
\read\csvfile to\csvline
\expandafter
\xdef\csname @csvline\romannumeral\c@csvlinenum\endcsname{\csvline}%
\ifthenelse{\not\equal{\csvline}{\par}}{\trim{\csvline}}{}%
\ifnum\c@csvlinenum=1\relax
\csv@tb=\expandafter{\csvline}%
\edef\@csvlin@{\@csvseparator\the\csv@tb\@csvseparator}%
\c@field = 0\relax
\whiledo{\not\equal{\@csvlin@}{\@csvseparator\par\@csvseparator}
\and \not\equal{\@csvlin@}{\@csvseparator\@csvseparator}
\and \not\equal{\@csvlin@}{\@csvseparator}}{%
\extractentry{\@csvlin@}{\param}%
\csv@tb=\expandafter{\param}%
\advance\c@field by 1\relax
\expandafter
\xdef\csname insert\the\csv@tb\endcsname{%
\noexpand\field{\the\c@field}}%
}%
\fi
\csv@tb=\expandafter{\csvline}%
\edef\@csvlin@{\@csvseparator\the\csv@tb\@csvseparator}%
\c@field = 0\relax
\whiledo{\not\equal{\@csvlin@}{\@csvseparator\par\@csvseparator}
\and \not\equal{\@csvlin@}{\@csvseparator\@csvseparator}
\and \not\equal{\@csvlin@}{\@csvseparator}}{%
\extractentry{\@csvlin@}{\param}%
\csv@tb=\expandafter{\param}%
\advance\c@field by 1\relax
\expandafter
\xdef\csname @l\romannumeral\c@csvlinenum @field\romannumeral\c@field\endcsname{\the\csv@tb}%
}%
\ifeof\csvfile\morefalse\else\moretrue\fi
\ifmore
\repeat
\closein\csvfile
{\def\field##1{%
\csname @l\romannumeral\c@csvlinenum @field\romannumeral##1\endcsname}%
\def\@r@wh{\begin{longtable}{#2}#3}%
\def\@r@w{#4}%
\def\@r@wl{#5\end{longtable}}%
\maxlines=\c@csvlinenum
\advance\maxlines by -1\relax
\csvlastbutone=\maxlines
\advance\csvlastbutone by -1\relax
\c@csvlinenum=1\relax
\setcounter{csvrownumber}{0}\relax
\whiledo{\not{\c@csvlinenum>\maxlines}}{%
\ifthenelse{\expandafter\equal{%
\csname @csvline\romannumeral\c@csvlinenum\endcsname}{\par}}%
{\relax}{%
\ifnum\c@csvlinenum=1\relax
\@r@wh
\else
\refstepcounter{csvrownumber}%
\ifnum\c@csvlinenum=\maxlines\@r@wl\else\@r@w\fi
\fi}%
\global\advance\c@csvlinenum by 1\relax
}\relax
}}
%    \end{macrocode}
% Define a command to determine if on the penultimate row
%    \begin{macrocode}
\newcommand{\ifnextrowlast}[2]{%
\ifnum\c@csvlinenum=\csvlastbutone#1\else#2\fi}
%    \end{macrocode}
% Define means to store entry for later use
%    \begin{macrocode}
\newcommand{\csvSaveEntry}[2][csvrownumber]{%
\@ifnextchar[{\@csvSaveEntry{#1}{#2}}{%
\@csvSaveEntry{#1}{#2}[\insertbyname{#2}]}%
}
%    \end{macrocode}
%    \begin{macrocode}
\def\@csvSaveEntry#1#2[#3]{%
\edef\@csv@tmp{\insertbyname{#2}}%
\if\relax\@csv@tmp\relax
\edef\@entry{#3}%
\else
\edef\@entry{\insertbyname{#2}}%
\fi
\expandafter
\xdef\csname #2\romannumeral\value{#1}\endcsname{\@entry}}
%    \end{macrocode}
% Define means to access stored entry.
%    \begin{macrocode}
\newcommand{\csvGetEntry}[2]{%
\csname #2\romannumeral\value{#1}\endcsname}
%    \end{macrocode}
%\iffalse
%    \begin{macrocode}
%</csvtools.sty>
%    \end{macrocode}
%\fi
%\iffalse
%    \begin{macrocode}
%<*csvpie.sty>
%    \end{macrocode}
%\fi
%\subsection{csvpie.sty}
% Declare package.
%    \begin{macrocode}
\NeedsTeXFormat{LaTeX2e}
\ProvidesPackage{csvpie}[2006/09/01 v1.0]
%    \end{macrocode}
% Declare options.
%    \begin{macrocode}
\newif\ifcolorpiechart
\colorpiecharttrue
\DeclareOption{color}{\colorpiecharttrue}
\DeclareOption{monochrome}{\colorpiechartfalse}
%    \end{macrocode}
% Process options
%    \begin{macrocode}
\ProcessOptions
%    \end{macrocode}
% Specify required packages
%    \begin{macrocode}
\RequirePackage{csvtools}
\RequirePackage{tikz}
\RequirePackage{calc}
\RequirePackage{keyval}
%    \end{macrocode}
% Define some variables
%    \begin{macrocode}
\newlength\csvradius
\newlength\csvinner
\newlength\csvouter
\newlength\csvcutawayoffset
\newcounter{csvstartangle}
\newcounter{csvendangle}
\newcounter{csvangle}
\newcounter{csvpietotal}
%    \end{macrocode}
% Set default values.
%    \begin{macrocode}
\newcommand*{\csvpiedefaults}{radius=2cm,inner=0.25,outer=1.25,%
cutaway={},total=100,start=0,offset=0.1}
%    \end{macrocode}
% Define command to specify inner and outer labels.
% The default puts the entry in column 1 (usually the label)
% in the outer area, and the entry in column 2 (usually the
% value) in the inner area. These can be redefined by the user
% to use other variables, or change the format etc.
%    \begin{macrocode}
\newcommand{\csvpieouterlabel}{\field{1}}
\newcommand{\csvpieinnerlabel}{\field{2}\%}
%    \end{macrocode}
% Define command to store list of row numbers corresponding to
% cut away segments. (Segments that have been offset from the
% pie chart.) By default, none of the segments should be cut away.
%    \begin{macrocode}
\newcommand*{\csvpiecutaways}{}
%    \end{macrocode}
% Define a command to assign colour name "#2" to segment number "#1"
%    \begin{macrocode}
\newcommand*{\csvpiesegmentcol}[2]{%
\expandafter\def\csname csvpie@segcol\romannumeral#1\endcsname{#2}}
%    \end{macrocode}
% Define command to get colour name assigned to segment number "#1"
%    \begin{macrocode}
\newcommand*{\csvpiesegcolname}[1]{%
\csname csvpie@segcol\romannumeral#1\endcsname}
%    \end{macrocode}
% Set default colours. These can be changed by the user. More colours
% can be added in a similar manner if there are more than 8 segments.
%    \begin{macrocode}
\ifcolorpiechart
\csvpiesegmentcol{1}{red}
\csvpiesegmentcol{2}{green}
\csvpiesegmentcol{3}{blue}
\csvpiesegmentcol{4}{yellow}
\csvpiesegmentcol{5}{magenta}
\csvpiesegmentcol{6}{cyan}
\csvpiesegmentcol{7}{orange}
\csvpiesegmentcol{8}{white}
\else
\csvpiesegmentcol{1}{black!15}
\csvpiesegmentcol{2}{black!25}
\csvpiesegmentcol{3}{black!35}
\csvpiesegmentcol{4}{black!45}
\csvpiesegmentcol{5}{black!55}
\csvpiesegmentcol{6}{black!65}
\csvpiesegmentcol{7}{black!75}
\csvpiesegmentcol{8}{black!85}
\fi
%    \end{macrocode}
% Define keys for "\csvpiechart" optional argument
%    \begin{macrocode}
\define@key{csvpie}{start}{\setcounter{csvstartangle}{#1}}
\define@key{csvpie}{total}{\setcounter{csvpietotal}{#1}}
\define@key{csvpie}{radius}{\setlength{\csvradius}{#1}}
\define@key{csvpie}{inner}{\setlength{\csvinner}{#1\csvradius}}
\define@key{csvpie}{outer}{\setlength{\csvouter}{#1\csvradius}}
\define@key{csvpie}{offset}{\setlength{\csvcutawayoffset}{#1\csvradius}}
\define@key{csvpie}{cutaway}{\renewcommand*{\csvpiecutaways}{#1}}
\providecommand*{\csv@startrow}{1}
\define@key{csvpie}{firstrow}{\renewcommand*{\csv@startrow}{#1}}
%    \end{macrocode}
% Define "\csvpiechart". The starred version uses "\applyCSVfile*",
% the unstarred version uses "\applyCSVfile".
%    \begin{macrocode}
\newcommand{\csvpiechart}{%
\@ifstar{\def\@@pieapplyCSVfile{\applyCSVfile*}%
\def\csv@startrow{1}%
\@csvpiechart}{%
\def\@@pieapplyCSVfile{\applyCSVfile}%
\def\csv@startrow{2}%
\@csvpiechart}%
}
%    \end{macrocode}
% Set the keys for pie chart:
%    \begin{macrocode}
\newcommand*{\csvpiesetkeys}[1]{%
\edef\csv@piesk{\noexpand\setkeys{csvpie}{\csvpiedefaults,#1}}%
\csv@piesk}
%    \end{macrocode}
% The main body of "\csvpiechart".
%    \begin{macrocode}
\newcommand{\@csvpiechart}[3][]{%
{\csvpiesetkeys{#1}%
\edef\csv@startang{\number\c@csvstartangle}%
\csv@computeangles{#2}{#3}%
\setcounter{csvstartangle}{\csv@startang}%
\begin{tikzpicture}%
\@@pieapplyCSVfile[\csv@startrow]{#3}{%
\csvpiesegment{#2}}%
\end{tikzpicture}%
}}
%    \end{macrocode}
% Do individual segment in pie chart.
%    \begin{macrocode}
\newcommand*{\csvpiesegment}[1]{%
\setcounter{csvstartangle}{%
\csname csv@sang@\romannumeral\c@csvrownumber\endcsname}%
\setcounter{csvangle}{%
\csname csv@angle@\romannumeral\c@csvrownumber\endcsname}%
\setcounter{csvendangle}{\value{csvangle} + \value{csvstartangle}}%
\setcounter{csvangle}{%
\csname csv@cut@angle\romannumeral\c@csvrownumber\endcsname}%
\ifthenelse{\value{csvangle}>180}{\addtocounter{csvangle}{-360}}{}%
\edef\@csv@shift{(\number\c@csvangle:%
\csname csv@cut@len\romannumeral\c@csvrownumber\endcsname)}%
\setcounter{csvangle}{%
\csname csv@angle@\romannumeral\c@csvrownumber\endcsname/2
+\value{csvstartangle}}%
\begin{scope}[shift={\@csv@shift}]%
\fill[color=\csvpiesegcolname\c@csvrownumber] (0,0) --
(\thecsvstartangle:\csvradius)
arc (\thecsvstartangle:\thecsvendangle:\csvradius) -- cycle;
% if 90 < csvangle < 270, the text will look upside-down, so
% adjust accordingly. Reusing csvstartangle, to save defining
% a new counter, as it's no longer required
\ifthenelse{\(\value{csvangle}>90 \and \value{csvangle}<270\)
\TE@or \value{csvangle}<-90}{%
\setcounter{csvstartangle}{\value{csvangle}-180}%
\draw (\thecsvangle:\csvinner)
node[left,rotate=\thecsvstartangle]{\csvpieinnerlabel};
\draw (\thecsvangle:\csvouter)
node[left,rotate=\thecsvstartangle]{\csvpieouterlabel};
}{%
\draw (\thecsvangle:\csvinner)
node[right,rotate=\thecsvangle]{\csvpieinnerlabel};
\draw (\thecsvangle:\csvouter)
node[right,rotate=\thecsvangle]{\csvpieouterlabel};
}%
\end{scope}%
}
%    \end{macrocode}
% Compute the angles for each segment. First argument is
% variable, second argument is the name
% of the CSV file.
%    \begin{macrocode}
\newcommand*{\csv@computeangles}[2]{%
\@@pieapplyCSVfile[\csv@startrow]{#2}{%
\csvsetsegmentparams{\c@csvrownumber}{#1}%
}%
\ifthenelse{\equal{\csvpiecutaways}{}}{}{\csvcomputeoffsets}}
%    \end{macrocode}
% Compute offset angles for cutaway segments.
% "\csv@row" should either be a single number (e.g.\ "2")
% or a number range (e.g.\ "1-2")
%    \begin{macrocode}
\newcommand*{\csvcomputeoffsets}{%
\@for\csv@row:=\csvpiecutaways\do{%
\expandafter\@csv@set@off\csv@row-\relax
}}
%    \end{macrocode}
% Set the offset angle.
%    \begin{macrocode}
\def\@csv@set@off#1-#2\relax{%
\ifthenelse{\equal{#2}{}}{%
\@@csv@set@off{#1}}{%
\@@csv@set@offr#1-#2\relax}%
}
%    \end{macrocode}
% Set offset for individual segment:
%    \begin{macrocode}
\newcommand*{\@@csv@set@off}[1]{%
\setcounter{csvangle}{%
\csname csv@angle@\romannumeral#1\endcsname/2
+ \csname csv@sang@\romannumeral#1\endcsname}%
\expandafter\xdef\csname csv@cut@angle\romannumeral#1\endcsname{%
\number\c@csvangle}%
\expandafter
\gdef\csname csv@cut@len\romannumeral\csv@row\endcsname{%
\csvcutawayoffset}%
}
%    \end{macrocode}
% Set offset for range of segments:
%    \begin{macrocode}
\newcount\@csv@seg
\def\@@csv@set@offr#1-#2-\relax{%
\ifnum#1>#2\relax
\PackageError{csvpie}{Segment ranges must go in ascending
order}{Try #2-#1 instead of #1-#2}%
\else
\setcounter{csvangle}{0}%
\@csv@seg=#1\relax
\whiledo{\not\(\@csv@seg > #2\)}{%
\addtocounter{csvangle}{%
\csname csv@angle@\romannumeral\@csv@seg\endcsname}%
\advance\@csv@seg by 1}%
\setcounter{csvangle}{\value{csvangle}/2
+ \csname csv@sang@\romannumeral#1\endcsname}%
\@csv@seg=#1\relax
\whiledo{\not\(\@csv@seg > #2\)}{%
\expandafter\xdef\csname csv@cut@angle\romannumeral\@csv@seg\endcsname{%
\number\c@csvangle}%
\expandafter
\gdef\csname csv@cut@len\romannumeral\@csv@seg\endcsname{%
\csvcutawayoffset}%
\advance\@csv@seg by 1}%
\fi
}
%    \end{macrocode}
% Set the relevent variables required for "\csvpiesegment":
% The first argument is the segment number, the second argument
% is the variable. The "csvstartangle" counter is updated.
% This command should be called sequentially, and is provided
% for the convenience of \texttt{csvtools.pl}.
%    \begin{macrocode}
\newcommand*{\csvsetsegmentparams}[2]{%
\ifthenelse{\value{csvstartangle}>180}{%
\addtocounter{csvstartangle}{-360}}{}%
\ifthenelse{\value{csvstartangle}<-180}{%
\addtocounter{csvstartangle}{360}}{}%
\expandafter
\xdef\csname csv@sang@\romannumeral#1\endcsname{%
\number\c@csvstartangle}%
\setcounter{csvangle}{360*\real{#2}/\value{csvpietotal}}%
\addtocounter{csvstartangle}{\value{csvangle}}%
\expandafter
\xdef\csname csv@angle@\romannumeral#1\endcsname{%
\number\c@csvangle}%
\expandafter
\gdef\csname csv@cut@angle\romannumeral#1\endcsname{0}%
\expandafter
\gdef\csname csv@cut@len\romannumeral#1\endcsname{0cm}%
}
%    \end{macrocode}
%\iffalse
%    \begin{macrocode}
%</csvpie.sty>
%    \end{macrocode}
%\fi
%\iffalse
%    \begin{macrocode}
%<*csvsort.sty>
%    \end{macrocode}
%\fi
%\subsection{csvsort.sty}
% Declare package:
%    \begin{macrocode}
\NeedsTeXFormat{LaTeX2e}
\ProvidesPackage{csvsort}[2007/07/03 v1.0 (NLCT)]
%    \end{macrocode}
% If \cmdname{compare} hasn't been defined, use compare.tex
%    \begin{macrocode}
\@ifundefined{compare}{\input{compare}}{}
%    \end{macrocode}
% Define comparison commands. Each of the comparison commands must
% set the boolean \cmdname{if@csv@ishigher}
%    \begin{macrocode}
\newif\if@csv@ishigher
%    \end{macrocode}
% Alphabetical ascending comparison:
%    \begin{macrocode}
\newcommand{\csv@alphaascendingcompare}[2]{%
\@csv@extractkey#1{\@csv@keya}%
\@csv@extractkey#2{\@csv@keyb}%
\csvsort@verbose{Comparing \@csv@keya\space and \@csv@keyb}%
\edef\@csvsrt@result{\noexpand\compare{\@csv@keya}{\@csv@keyb}}%
\ifnum\@csvsrt@result=1\relax
  \@csv@ishighertrue
\else
  \@csv@ishigherfalse
\fi}
%    \end{macrocode}
% Alphabetical descending comparison:
%    \begin{macrocode}
\newcommand{\csv@alphadescendingcompare}[2]{%
\@csv@extractkey#1{\@csv@keya}%
\@csv@extractkey#2{\@csv@keyb}%
\csvsort@verbose{Comparing \@csv@keya\space and \@csv@keyb}%
\edef\@csvsrt@result{\noexpand\compare{\@csv@keya}{\@csv@keyb}}%
\ifnum\@csvsrt@result=-1\relax
  \@csv@ishighertrue
\else
  \@csv@ishigherfalse
\fi}
%    \end{macrocode}
% Numerical ascending comparison:
%    \begin{macrocode}
\newcommand{\csv@numericascendingcompare}[2]{%
\@csv@extractkey#1{\@csv@keya}%
\@csv@extractkey#2{\@csv@keyb}%
\if\relax\@csv@keya\relax\def\@csv@keya{0}\fi
\if\relax\@csv@keyb\relax\def\@csv@keyb{0}\fi
\csvsort@verbose{Comparing \number\@csv@keya\space and \number\@csv@keyb}%
\ifnum\@csv@keya>\@csv@keyb\relax
  \@csv@ishighertrue
\else
  \@csv@ishigherfalse
\fi}
%    \end{macrocode}
% Numerical descending comparison:
%    \begin{macrocode}
\newcommand{\csv@numericdescendingcompare}[2]{%
\@csv@extractkey#1{\@csv@keya}%
\@csv@extractkey#2{\@csv@keyb}%
\if\relax\@csv@keya\relax\def\@csv@keya{0}\fi
\if\relax\@csv@keyb\relax\def\@csv@keyb{0}\fi
\csvsort@verbose{Comparing \number\@csv@keya\space and \number\@csv@keyb}%
\ifnum\@csv@keya<\@csv@keyb\relax
  \@csv@ishighertrue
\else
  \@csv@ishigherfalse
\fi}
%    \end{macrocode}
% The sort key needs to be extracted from the data (3rd argument
% must be a control sequence):
%    \begin{macrocode}
\def\@csv@extractkey#1#2#3{\def#3{#1}}
%    \end{macrocode}
% Conversely, extract the data and ignore the sort key:
%    \begin{macrocode}
\def\@csv@extractdata#1#2#3{\def#3{#2}}
%    \end{macrocode}
% The command \cmdname{csv@compare} should be set to the
% appropriate comparison command (\cmdname{csv@alphaascendingcompare}
% by default.)
%    \begin{macrocode}
\let\csv@compare\csv@alphaascendingcompare
%    \end{macrocode}
% The package options are a comma-separated list of key=value
% pairs, so need the \stynamefmt{xkeyval} package.
%    \begin{macrocode}
\RequirePackage{xkeyval}
%    \end{macrocode}
% Package options. Verbose mode:
%    \begin{macrocode}
\define@boolkey{csvsort.sty}[csvsrt]{verbose}[true]{}
%    \end{macrocode}
% Default is verbose:
%    \begin{macrocode}
\csvsrtverbosetrue
%    \end{macrocode}
% Define command to display message if verbose option is set:
%    \begin{macrocode}
\newcommand{\csvsort@verbose}[1]{%
\ifcsvsrtverbose\typeout{#1}\fi}
%    \end{macrocode}
% Sort type (can be either numerical or alphabetical, ascending
% or descending):
%    \begin{macrocode}
\define@choicekey{csvsort.sty}{sort}[\val\nr]{alphabetical,numerical,%
alphabetical ascending,numerical ascending,%
alphabetical descending,numerical descending,%
alphabeticalascending,alphabeticaldescending,%
numericalascending,numericaldescending}{%
\ifcase\nr
 % alphabetical (ascending)
   \let\csv@compare\csv@alphaascendingcompare
 \or
 % numerical (ascending)
   \let\csv@compare\csv@numericascendingcompare
 \or
 % alphabetical ascending
   \let\csv@compare\csv@alphaascendingcompare
 \or
 % numerical ascending
   \let\csv@compare\csv@numericascendingcompare
 \or
 % alphabetical descending
   \let\csv@compare\csv@alphadescendingcompare
 \or
 % numerical descending
   \let\csv@compare\csv@numericdescendingcompare
 \or
 % alphabeticalascending
   \let\csv@compare\csv@alphaascendingcompare
 \or
 % alphabetical descending
   \let\csv@compare\csv@alphadescendingcompare
 \or
 % numericalascending
   \let\csv@compare\csv@numericascendingcompare
 \or
 % numericaldescending
   \let\csv@compare\csv@numericdescendingcompare
\fi}
%    \end{macrocode}
% The variable used to sort is given by:
%    \begin{macrocode}
\newcommand{\csv@sortvariable}{\field{1}}
%    \end{macrocode}
% A different variable can be set using the variable package option:
%    \begin{macrocode}
\define@key{csvsort.sty}{variable}{%
\renewcommand{\csv@sortvariable}{#1}}
%    \end{macrocode}
% The first data line in CSV files with no headers is given by
%    \begin{macrocode}
\newcounter{sfirstdataline}
%    \end{macrocode}
% and defaults to line 1:
%    \begin{macrocode}
\setcounter{sfirstdataline}{1}
%    \end{macrocode}
% This value can be set using the "sfirstdataline" option:
%    \begin{macrocode}
\define@key{csvsort.sty}{sfirstdataline}{%
\setcounter{sfirstdataline}{#1}}
%    \end{macrocode}
% The first data line in CSV files with headers is given by
%    \begin{macrocode}
\newcounter{firstdataline}
%    \end{macrocode}
% and defaults to line 2:
%    \begin{macrocode}
\setcounter{firstdataline}{2}
%    \end{macrocode}
% This value can be set using the "firstdataline" option:
%    \begin{macrocode}
\define@key{csvsort.sty}{firstdataline}{%
\setcounter{firstdataline}{#1}}
%    \end{macrocode}
% Process package options:
%    \begin{macrocode}
\ProcessOptionsX
%    \end{macrocode}
% Required packages:
%    \begin{macrocode}
\RequirePackage{csvtools}
\RequirePackage{xfor}
%    \end{macrocode}
% Insertion sort macro. The argument is a comma-separated list
% in the form \marg{sort-key1}\marg{data1},\ldots,%
%\marg{sort-keyN}\marg{dataN}. The list is sorted according to
% \cmdname{csv@compare}
%\begin{macro}{\csv@insertionsort}
%    \begin{macrocode}
\newcommand{\csv@insertionsort}[1]{%
\let\@csv@sortedlist\relax
\@for\@csv@listelement:=#1\do{%
\if\@csv@sortedlist\relax
 \expandafter\toks@\expandafter{\@csv@listelement}%
 \edef\@csv@sortedlist{\the\toks@}%
\else
 \expandafter\@csv@insert@into\expandafter
  {\@csv@listelement}{\@csv@sortedlist}%
\fi}\let#1\@csv@sortedlist}
%    \end{macrocode}
%\end{macro}
% Insert element (first argument) into a sorted list (second argument)
% Each element must be in the form \marg{sort-key}\marg{data}, where
% \cmdname{csv@compare} sorts according to \meta{sort-key}. The
% insertion code ignores \meta{data}.
%\begin{macro}{\@csv@insertinto}
%    \begin{macrocode}
\newcommand\@csv@insert@into[2]{%
\let\@csv@tmplist\relax
\@for\@sort@list@element:=#2\do{%
\expandafter\csv@compare\expandafter{\@sort@list@element}{#1}%
\if@csv@ishigher
  \toks@{#1}%
  \if\relax\@csv@tmplist
    \edef\@csv@tmplist{\the\toks@,\@sort@list@element}%
  \else
    \edef\@csv@tmplist{\@csv@tmplist,\the\toks@,\@sort@list@element}%
  \fi
  \@endfortrue
\else
  \if\relax\@csv@tmplist
    \edef\@csv@tmplist{\@sort@list@element}%
  \else
    \edef\@csv@tmplist{\@csv@tmplist,\@sort@list@element}%
  \fi
\fi
}%
\if@endfor
 \ifx\@forremainder\@empty
 \else
   \edef\@csv@tmplist{\@csv@tmplist,\@forremainder}%
 \fi
\else
 \toks@{#1}%
 \if\relax\@csv@tmplist
   \edef\@csv@tmplist{\the\toks@}%
 \else
   \edef\@csv@tmplist{\@csv@tmplist,\the\toks@}%
 \fi
\fi
\@endforfalse
\let#2\@csv@tmplist
}
%    \end{macrocode}
%\end{macro}
% Define \cmdname{sortapplyCSVfile}. This is akin to \stynamefmt{csvtools}'
% \cmdname{applyCSVfile}, except it stores each data row in a list,
% sorts the list, and then iterates through the sorted list. As
% with, \cmdname{applyCSVfile}, the starred version should be used
% for a file that doesn't contain a header row, and the un-starred
% version should be used for a file that has a header row. Both
% forms take an optional argument, which should be a key=value list,
% the same as the package options, but they only apply to the
% current instance of \cmdname{sortapplyCSVfile}.
%\begin{macro}{\sortapplyCSVfile}
%    \begin{macrocode}
\newcommand{\sortapplyCSVfile}{%
\@ifstar\@ssortapplyCSVfile\@sortapplyCSVfile}
%    \end{macrocode}
%\end{macro}
% Starred version (see above):
%\begin{macro}{\@ssortapplyCSVfile}
%    \begin{macrocode}
\newcommand{\@ssortapplyCSVfile}[3][]{%
\bgroup
\setkeys{csvsort.sty}{#1}%
\def\@csv@list{}%
\@sapplyCSVfile[\c@sfirstdataline]{#2}{%
\edef\csv@key{\csv@sortvariable}%
\if\relax\@csv@list\relax
\protected@xdef\@csv@list{{\csv@key}{\csvline}}%
\else
\protected@xdef\@csv@list{\@csv@list,{\csv@key}{\csvline}}%
\fi
}%
\@ssortapplyCSVdata{\@csv@list}{#3}%
\egroup}
%    \end{macrocode}
%\end{macro}
% Unstarred version (see above):
%\begin{macro}{\@sortapplyCSVfile}
%    \begin{macrocode}
\newcommand{\@sortapplyCSVfile}[3][]{%
\bgroup
\setkeys{csvsort.sty}{#1}%
\def\@csv@list{}%
\@applyCSVfile[\c@firstdataline]{#2}{%
\edef\csv@key{\csv@sortvariable}%
\if\relax\@csv@list\relax
\protected@xdef\@csv@list{{\csv@key}{\csvline}}%
\else
\protected@xdef\@csv@list{\@csv@list,{\csv@key}{\csvline}}%
\fi
}%
\@sortapplyCSVdata{\@csv@list}{#3}%
\egroup}
%    \end{macrocode}
%\end{macro}
% \cmdname{sortapplyCSVdata}\oarg{sort type}\marg{cmd}\marg{text}
% will sort the data given by \meta{cmd}, which should contain
% a comma separated
% list of the form \marg{sort1}\marg{data1},\ldots\marg{sortN}\marg{dataN}
% and the applies \meta{text} to each data element of the list.
% The sort ordering is specified by \meta{sort type}, which should
% have the same values as the sort key used by
% \cmdname{sortapplyCSVfile}.
%  The starred version doesn't permit
% \cmdname{insert}\meta{label} commands. First argument is the list,
% the second argument specifies what to do for each data element.
% (Note, these commands are primarily provided for the
% benefit of \texttt{csvtools.pl}, which is why they're not
% documented in the main part of the \stynamefmt{csvtools} manual.)
%\begin{macro}{\sortapplyCSVdata}
%    \begin{macrocode}
\newcommand{\sortapplyCSVdata}{%
\@ifstar\@ssortapplyCSVdata\@sortapplyCSVdata}
%    \end{macrocode}
%\end{macro}
% Starred version:
%\begin{macro}{\@ssortapplyCSVdata}
%    \begin{macrocode}
\newcommand{\@ssortapplyCSVdata}[3][\relax]{%
\bgroup
\if\relax#1\relax\else\setkeys{csvsort.sty}{sort=#1}\fi
\csv@insertionsort{#2}%
\c@csvrownumber=0\relax
\@for\@csv@element:=#2\do{%
\expandafter\@csv@extractdata\@csv@element\csvline
\expandafter\toks@\expandafter{\csvline}%
\edef\@csvlin@{\@csvseparator\the\toks@\@csvseparator}%
\long\edef\@csvlin@{\@csvseparator\csvline\@csvseparator}%
\c@field=0\relax
\whiledo{\not\equal{\@csvlin@}{\@csvseparator}}{%
\extractentry{\@csvlin@}{\param}%
\expandafter\toks@\expandafter{\param}%
\advance\c@field by 1\relax
\expandafter\xdef\csname @field\romannumeral\c@field\endcsname{%
\the\toks@}%
}%
\ifthenelse{\not\equal{\csvline}{}}{%
\refstepcounter{csvrownumber}%
#3}{}%
}%
\egroup
}
%    \end{macrocode}
%\end{macro}
% Unstarred version (field labels will need to be set (using
% \cmdname{setcsvfieldlabel}) before using this macro.)
%\begin{macro}{\@sortapplyCSVdata}
%    \begin{macrocode}
\newcommand{\@sortapplyCSVdata}[3][\relax]{%
\bgroup
\if\relax#1\relax\else\setkeys{csvsort.sty}{sort=#1}\fi
\csv@insertionsort{#2}%
\c@csvrownumber=0\relax
\@for\@csv@element:=#2\do{%
\expandafter\@csv@extractdata\@csv@element\csvline
\expandafter\toks@\expandafter{\csvline}%
\edef\@csvlin@{\@csvseparator\the\toks@\@csvseparator}%
\long\edef\@csvlin@{\@csvseparator\csvline\@csvseparator}%
\c@field=0\relax
\whiledo{\not\equal{\@csvlin@}{\@csvseparator}}{%
\extractentry{\@csvlin@}{\param}%
\expandafter\toks@\expandafter{\param}%
\advance\c@field by 1\relax
\expandafter\xdef\csname @field\romannumeral\c@field\endcsname{%
\the\toks@}%
\edef\@fieldlabel{%
\csname @fieldlabel\romannumeral\c@field\endcsname}%
\expandafter\xdef\csname insert\@fieldlabel\endcsname{\the\toks@}%
}%
\ifthenelse{\not\equal{\csvline}{}}{%
\refstepcounter{csvrownumber}%
#3}{}%
}%
\egroup
}
%    \end{macrocode}
%\end{macro}
% Define \cmdname{sortCSVtotabular}. This is akin to \stynamefmt{csvtools}'
% \cmdname{CSVtotabular}, but sorts the data. Takes an optional
% argument that is a key=value list which is that same as the
% package options, but only applies to this instance of
%\cmdname{sortCSVtotabular}. This command first loads all the data
% into a list, sorts the list, and then puts it in a tabular
% environment. The total number of data rows is stored in
% \cmdname{maxlines}.
%\begin{macro}{\sortCSVtotabular}
%    \begin{macrocode}
\newcommand{\sortCSVtotabular}[6][]{%
\@sortCSVtotabular[#1]{#2}{#3}{#4}{#5}{#6}{tabular}}
%    \end{macrocode}
%\end{macro}
% As above, but use longtable environment instead:
%\begin{macro}{\sortCSVtolongtable}
%    \begin{macrocode}
\newcommand{\sortCSVtolongtable}[6][]{%
\@sortCSVtotabular[#1]{#2}{#3}{#4}{#5}{#6}{longtable}}
%    \end{macrocode}
%\end{macro}
% The final argument specifies which tabular-like environment
% to use:
%\begin{macro}{\@sortCSVtotabular}
%    \begin{macrocode}
\newcommand{\@sortCSVtotabular}[7][]{%
\bgroup
\setkeys{csvsort.sty}{#1}%
%    \end{macrocode}
% Read all data into \cmdname{@csv@list}:
%    \begin{macrocode}
\def\@csv@list{}%
\@applyCSVfile[\c@firstdataline]{#2}{%
\edef\csv@key{\csv@sortvariable}%
\if\relax\@csv@list\relax
\protected@xdef\@csv@list{{\csv@key}{\csvline}}%
\else
\protected@xdef\@csv@list{\@csv@list,{\csv@key}{\csvline}}%
\fi
}%
%    \end{macrocode}
% Sort data, and put in required tabular-like environment
%    \begin{macrocode}
\sortCSVdatatotabular{\c@csvrownumber}{\@csv@list}{#3}{#4}{#5}{#6}{#7}%
\egroup
}
%    \end{macrocode}
%\end{macro}
% \cmdname{sortCSVdatatotabular}\marg{n}\marg{list}\marg{col-align}\marg{header-row}\marg{all but last rows}\marg{last row}\marg{env-name}\par
% This command sorts the data specified by \meta{list} (which has
% \meta{n} elements) and puts in a tabular-like environment called
% \meta{env-name}. This command is mainly provided for the benefit
% of \texttt{csvtools.pl}, which is why it is not documented in the
% main part of the \stynamefmt{csvtools} manual. Field names must be
% set prior to use (using \cmdname{setcsvfieldlabel}.)
%\begin{macro}{\sortCSVdatatotabular}
%    \begin{macrocode}
\newcommand{\sortCSVdatatotabular}[7]{%
%    \end{macrocode}
% Set \cmdname{maxlines} (total number of data rows, not including
% header row).
%    \begin{macrocode}
\maxlines=#1\relax
%    \end{macrocode}
% Sort the data:
%    \begin{macrocode}
\csv@insertionsort{#2}%
%    \end{macrocode}
% Set \cmdname{csvlastbutone} to \cmdname{maxlines}$-1$:
%    \begin{macrocode}
\csvlastbutone=\maxlines
\advance\csvlastbutone by -1\relax
%    \end{macrocode}
% Reset \cmdname{c@csvrownumber}:
%    \begin{macrocode}
\c@csvrownumber=0\relax
\c@csvlinenum=0\relax
%    \end{macrocode}
% Initialise first row, last row and middle rows (in case not enough
% data)
%    \begin{macrocode}
\def\@r@wh{#4}\def\@r@w{}\def\@r@wl{}%
%    \end{macrocode}
% Iterate through each element of sorted list
%    \begin{macrocode}
\@for\@csv@element:=#2\do{%
%    \end{macrocode}
% Extract data from the current list element (ignoring sort key)
% and store in \cmdname{csvline}:
%    \begin{macrocode}
\expandafter\@csv@extractdata\@csv@element\csvline
%    \end{macrocode}
% Extract each field of \cmdname{csvline}:
%    \begin{macrocode}
\expandafter\toks@\expandafter{\csvline}%
\edef\@csvlin@{\@csvseparator\the\toks@\@csvseparator}%
\long\edef\@csvlin@{\@csvseparator\csvline\@csvseparator}%
\c@field=0\relax
\whiledo{\not\equal{\@csvlin@}{\@csvseparator}}{%
\extractentry{\@csvlin@}{\param}%
%    \end{macrocode}
% Store field in \cmdname{toks@}
%    \begin{macrocode}
\expandafter\toks@\expandafter{\param}%
%    \end{macrocode}
% Increment field counter
%    \begin{macrocode}
\advance\c@field by 1\relax
%    \end{macrocode}
% Define \cmdname{field}\meta{n} and associated
% \cmdname{insert}\meta{label}
%    \begin{macrocode}
\expandafter\xdef\csname @field\romannumeral\c@field\endcsname{%
\the\toks@}%
\edef\@fieldlabel{%
\csname @fieldlabel\romannumeral\c@field\endcsname}%
\expandafter\xdef\csname insert\@fieldlabel\endcsname{\the\toks@}%
}%
%    \end{macrocode}
% Increment row counter.
%    \begin{macrocode}
\refstepcounter{csvrownumber}%
\refstepcounter{csvlinenum}%
%    \end{macrocode}
% Check to see if this is the first row, last row, of somewhere in
% between
%    \begin{macrocode}
\ifnum\c@csvrownumber=\maxlines
 \protected@edef\@r@wl{#6}%
\else
 \expandafter\toks@\expandafter{\@r@w}
 \protected@edef\@r@w{\the\toks@#5}%
\fi
}%
%    \end{macrocode}
% Now do tabular environment:
%    \begin{macrocode}
\begin{#7}{#3}%
\@r@wh
\@r@w
\@r@wl
\end{#7}%
}
%    \end{macrocode}
%\end{macro}
%\iffalse
%    \begin{macrocode}
%</csvsort.sty>
%    \end{macrocode}
%\fi
%\Finale
\endinput

You are viewing proxied material from ftp.icm.edu.pl. The copyright of proxied material belongs to its original authors. Any comments or complaints in relation to proxied material should be directed to the original authors of the content concerned. Please see the disclaimer for more details.