When processing a TSV file with text in the third column (and using --columns 2), I noticed issues with double quotes causing problems that smashed several lines together in the output (along with the first and second column of metadata. The solution I believe is to add the parameter "quoting=csv.QUOTE_NONE" in the call to csv.reader (in split.py) See this StackOverflow article for more details about quotes and csv reader:
https://stackoverflow.com/questions/26920849/how-to-handle-tsv-with-uneven-quotes
When processing a TSV file with text in the third column (and using --columns 2), I noticed issues with double quotes causing problems that smashed several lines together in the output (along with the first and second column of metadata. The solution I believe is to add the parameter "quoting=csv.QUOTE_NONE" in the call to csv.reader (in split.py) See this StackOverflow article for more details about quotes and csv reader: https://stackoverflow.com/questions/26920849/how-to-handle-tsv-with-uneven-quotes