path: root/Master/texmf-dist/source/latex/csvtools/csvtools.dtx

%\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