scovich commented on issue #7299:
URL: https://github.com/apache/arrow-rs/issues/7299#issuecomment-3450889623

   > Just realized, if metadata parsing is going to have an optimization to 
skip some row groups that [@etseidl](https://github.com/etseidl) mentioned in 
[here](https://github.com/apache/arrow-rs/pull/7307#issuecomment-3443712138), 
and thus we wouldn't know about row group first row positions, it'd be better 
to know whether we'll use row number virtual column before we parse the 
metadata, to be able to fallback to parsing without optimization. So that goes 
in favour of having this as a `ArrowReaderOptions`.
   
   The idea being, to collect enough information while parsing metadata to give 
accurate row numbers even when some row groups and/or pages get skipped? That 
_is_ a bit annoying, if requesting row numbers requires parsing significantly 
more metadata just to grab the rowcount for each skipped item...


-- 
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]

Reply via email to