csvsimple filtering problem

Question

The following code works as I would expect.

\documentclass{scrartcl}
\usepackage{csvsimple,filecontents}

\begin{document} 

\begin{filecontents*}{chunk.csv}
AAAA,aaaaa
BBBB,bbbbb
CCCC,ccccc
DDDD,ddddd
EEEE,eeeee
FFFFF,fffffff
GGGGG,ggggg
HHHHHH,hhhhh
\end{filecontents*}

\csvreader[no head,before filter= \ifthenelse{\thecsvrow<4}{\csvfilteraccept}{\csvfilterreject}]{chunk.csv}{}{%
\begin{itemize} 
\item \csvcoli , \csvcolii
\end{itemize}}

\end{document} 

When I change the filter condition from "\thecsvrow<4" to "\thecsvrow=4" or "\thecsvrow>4" the code breaks. Pointers as to how to make this work would be greatly appreciated.

Answer 1

The per-item testing should be performed when setting the item, rather than in the header, since the filter are set globally. From the csvsimple documentation:

\csvfilterreject

All following data lines will be ignored. This command overwrites all previous filter settings.

\documentclass{article}
\usepackage{csvsimple,filecontents}% http://ctan.org/pkg/{csvsimple,filecontents}

\begin{document} 

\begin{filecontents*}{chunk.csv}
AAAA,aaaaa
BBBB,bbbbb
CCCC,ccccc
DDDD,ddddd
EEEE,eeeee
FFFFF,fffffff
GGGGG,ggggg
HHHHHH,hhhhh
\end{filecontents*}

\begin{itemize}
\csvreader[no head]{chunk.csv}{}{\ifnum\value{csvrow}<4\relax\item \csvcoli, \csvcolii\fi}
\end{itemize}

\hrulefill

\begin{itemize}
\csvreader[no head]{chunk.csv}{}{\ifnum\value{csvrow}=4\relax\item \csvcoli, \csvcolii\fi}
\end{itemize}

\hrulefill

\begin{itemize}
\csvreader[no head]{chunk.csv}{}{\ifnum\value{csvrow}>4\relax\item \csvcoli, \csvcolii\fi}
\end{itemize}

\end{document}

Low-level conditionals are used to (not) print each item, testing the value of csvrow.

---

Here is an alternative option using datatool. The interface is very similar, and the transition should therefore not be a problem:

\documentclass{article}
\usepackage{datatool,filecontents}% http://ctan.org/pkg/{datatool,filecontents}

\begin{document} 

\begin{filecontents*}{chunk.csv}
AAAA,aaaaa
BBBB,bbbbb
CCCC,ccccc
DDDD,ddddd
EEEE,eeeee
FFFFF,fffffff
GGGGG,ggggg
HHHHHH,hhhhh
\end{filecontents*}

\DTLloaddb[noheader,keys={first,last}]{chunk}{chunk.csv}
\begin{itemize}
  \DTLforeach{chunk}{\First=first,\Last=last}{\ifnum\value{DTLrowi}<4\relax\item \First, \Last\fi}
\end{itemize}

\hrulefill

\begin{itemize}
  \DTLforeach{chunk}{\First=first,\Last=last}{\ifnum\value{DTLrowi}=4\relax\item \First, \Last\fi}
\end{itemize}

\hrulefill

\begin{itemize}
  \DTLforeach{chunk}{\First=first,\Last=last}{\ifnum\value{DTLrowi}>4\relax\item \First, \Last\fi}
\end{itemize}

\end{document}

Low-level conditionals are used to (not) print each item, testing the value of DTLrowi.

Answer 2

The csvrow counter is used for accepted input lines (after filtering). If you want to count line numbers of the input file to build a filter on it, you should use the counter csvinputline:

\documentclass{article}
\usepackage{csvsimple,filecontents}

\begin{document}

\begin{filecontents*}{chunk.csv}
AAAA,aaaaa
BBBB,bbbbb
CCCC,ccccc
DDDD,ddddd
EEEE,eeeee
FFFFF,fffffff
GGGGG,ggggg
HHHHHH,hhhhh
\end{filecontents*}

\begin{itemize}
\csvreader[no head,filter={\value{csvinputline}<4}]{chunk.csv}{}{\item \csvcoli, \csvcolii}
\end{itemize}

\hrulefill

\begin{itemize}
\csvreader[no head,filter={\value{csvinputline}=4}]{chunk.csv}{}{\item \csvcoli, \csvcolii}
\end{itemize}

\hrulefill

\begin{itemize}
\csvreader[no head,filter={\value{csvinputline}>4}]{chunk.csv}{}{\item \csvcoli, \csvcolii}
\end{itemize}

\end{document}

The resulting output is the same as in the answer of @Werner.