It’s exam time, and I need to script the reading of content from a number of Excel files, put the information into a DataFrame, compute grades, and write the result to an Excel file.
Last time I did this, I used package ExcelFiles + DataFrames (in 2020). It seems like ExcelFiles was last updated in 2019.
I think most people use XSLX.jl, although that would have been true in 2020 as well so maybe there was a specific reason why you didn’t choose it then?
In any case if you’re essentially doing the same thing you did 2 years ago and it worked for you at that point you might not want to chang anything - just re-instantiate the manifest from when you last did this and work with the versions of packages that worked for you last time!
I have to modify it… the way to compute grades has changed :-o. I don’t remember why I chose ExcelFiles last time, but will check XSLX.
The package XLSX seems to be still in an early state. In my case a number written in the form of -5,01085245571595E-09 caused readxlsx() to crash with the missleading error message ERROR: Styles not found for this workbook.
What about Pandas.jl?
Remark:
It might be that the issue is caused by the digital delimiter char,
which is on my machine the comma ,.
Followup:
After I have tried the alternatives: Taro.jl, PyCall-pandas, Pandas.jl I have to admit,
the others are also failing with my Excel-File.
I use XLSX successfully, and OdsIO for LibreOffice. I only use North American decimal characters though.
The error reported in this issue arises not because of the written form of the number but because the file being read does not comply with ECMA-376 standard.
Recent versions of XLSX.jl now also read ISO/IEC 29500 (strict) Excel files, too, so this file should now read without trouble.