[Corruption making unusable] Saving overwrites cell above

[DesWurstes]

Example file:

4/na,136,tani wferesi,wferah Al tani,Swfeate An-tani,wfere An-tani,аур нА-сиа,PLKGT,,"25.5,33.5,43.5,101.5,102.5,113.5
"
5/ma,150,ß,wferah Midah,Swfeate Madh,wfere aida,ура льаида,3WRU7W,"mzuw,mwfe,hjbfaes,iw'dd,fa,idah,ida,nu'da",107.5
83/mfaffibn,34,wefmk wferesi,,,,,,,
id,ln,tr,en,fr,de,ru,pid,tags,notes

• Modify a cell in the 0th row then save; extra columns are added
• Modify a cell in 2nd, 3rd, etc.. then save; the cell just above is modified
• Delete a whole row. Either 1) the uppermost row we deleted comes back 2) more columns are added

[jjuback]

@DesWurstes Thank you for providing a test case so I have something to work with. In your example, shouldn't the last line be the header row?

id,ln,tr,en,fr,de,ru,pid,tags,notes

At any rate, it looks like the first issue is caused by the multiline cell at the end of the first data row not having any content on the second line. If I precede the lone quote mark with a single character, modifying a cell works as expected. I use regular expressions to parse line/column breaks, and this appears to be an edge case that is not accounted for. I suspect that it causes the other issues as well. Still investigating.

[DesWurstes]

Yes, in a conventional file that line would have been the 0th row. Many thanks for your work on this project.