Corrupted file because of table header row attribute not present in XML

[sertal70]

Hi all,
we are using exceljs in a project aimed to produce reports in excel format starting from an empty excel template.

As we started to code, we run into an issue that can be synthesized in the simple following code fragment:

  const workbook = new Excel.Workbook();
  await workbook.xlsx.readFile('./template.xlsx');
  await workbook.xlsx.writeFile('report.xlsx');

The issue is: if the template.xlsx contains an Excel table with headers, the resulting report.xlsx file will be corrupted. As soon as we try to open the output file in Excel, a message is displayed telling us that a portion of the file is corrupted. Then Excel try to recover it and actually it succeeds.

Diving into the library code, we observed that the parseOpen() function in /lib/xlsx/xform/table/table-xform.js relies on the presence of the table attribute headerRowCount in the XML file. We found that the attribute is missing, even if the table in Excel has been defined "with headers". By forcing the attribute presence in the parseOpen() function, everything works fine.

Note also that the original template is loaded without any issue in Excel: Excel correctly identifies the table and its header.

The question is: who is wrong with the headerRowCount attribute ? Excel or the library? In other words: is the library expecting the attribute to be in the wrong place? Or maybe it was the right place but it has been moved in one of the last release of Excel?

Anyone having the same issue?

[c4lyp5o]

i, too, have the same problem. it seems by loading the file like this:

const workbook = new Excel.Workbook(); await workbook.xlsx.readFile(filename, { ignoreNodes: [ 'tableParts', ], });

and saving anything commited to the any cells will not corrupt the file but it will strip all the table properties as it is ignored. this is the only way that i have found.