Small description
When I load a CSV file with unix (\n) line endings into vd and save it, it's written over with DOS (\r\n) line endings, and I have to use a tool like dos2unix after every save to normalize it.
Expected result
I think it should either preserve the line ending format of the original file that was loaded.
Additional context
Happens with v1.3, v1.5, and v2.-1dev.
Hi @dbandstra!
Yeah, his is happening DOS line endings are the default for the Python csv writer.
VisiData passes a few options to the csv writer. Something I can do for now is add lineterminator to those. This would mean that you could configure which lineterminator you wanted for a particular session, by passing the command line argument --csv-lineterminator $'\n' or set a new default for all sessions in your ~/.visidatarc with options.csv_lineterminator = '\n'.
I think that \r\n is the standard for CSV:
https://tools.ietf.org/html/rfc4180
In this way you could manage carriage return inside cells (\n).
Most helpful comment
I think that \r\n is the standard for CSV:
https://tools.ietf.org/html/rfc4180
In this way you could manage carriage return inside cells (\n).