Feature Request: CSV diffing that ignores useless information
johnpyp opened this issue · comments
johnpyp commented
CSVs are notoriously inconsistent at giving you the same information. A few examples of things difftastic could help with that would be great:
- Don't care about header ordering - various csv writers will take in objects that are unordered, and non-deterministically choose a random order
- Ignore extraneous whitespace in places that it doesn't matter like extra newline (maybe some whitespace between fields that are quotes, I'm not sure)
- Inconsistent quotes where unnecessary (
"Asdf"
andAsdf
should be considered equal)
Wilfred Hughes commented
Thanks for the suggestion. I think this is out of scope for difftastic: CSV is a somewhat ambiguous format, and difftastic does things like treating comment and strings separately that don't make sense in CSV.