zeroshade commented on issue #34334:
URL: https://github.com/apache/arrow/issues/34334#issuecomment-1443967659
That's definitely a good question. So in theory the only real use for
putting `{` or `[` to mark the field would be for the auto detection (and that
should be configurable), alternately a user could essentially pipe the records
through a channel and then process a column if they know it's a
list/json/whatever and expand that column themselves....
If you want to process a list type column via csv you'll have to essentially
have every row parse that value as if it were its own one-line CSV
basically.... and then convert each value to the correct type. I mean, it'd
definitely be convenient for users if you want to implement that but I'd
definitely say you should make a separate issue/PR for that from this. Maybe
even gather feedback. I'll take a look at how the C++ library handles (or
doesn't handle) list types for CSV
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
