I want to draw a special chart, for which the data are given as shown in the minimal example: The first row shows years, the first column Companies and at the cross-points you find a string of data. Using the readarray package, I can extract each cell. Furthermore, I also want to break down the sting of data. For such a task, the listofitems comes along the readarray package. However, I fail to do so. The (not) working examples leads to
Paragraph ended before \dataA was complete.
Using \numberlen, which is supposed to show the number of list entries, I got 2 -- not 4.
Does anybody have a hint on a solution?
\documentclass{article}
\usepackage{readarray}
\readarraysepchar{;}
\setsepchar{,}
\begin{document}
\def\dataF{%
;2014;2015
A;36,3,29,4;35,6,27,8
B;19,4,35,7;23,10,33,7}
\readarray*\dataF\dataA[2,2]
\begin{tabular}{c|c}
\dataA[1,1] & \dataA[1,2] \\
\dataA[2,1] & \dataA[2,2] \\
\end{tabular}
\readlist*\numbers{\dataA[2,2]}
%\def\numbersF{\Daten[2,2]}
%\readlist*\numbers{\numbersF} % fails as well
\begin{tabular}{c|c}
\dataA[1,1] & \dataA[1,2] \\
\dataA[2,1] & \dataA[2,2] \\
\dataA[2,1] & \number[1] \\
\end{tabular}
\end{document}

\numberis certainly wrong: it is a TeX primitive. – egreg Sep 26 '19 at 14:55