Github user sameeragarwal commented on the pull request:
https://github.com/apache/spark/pull/11649#issuecomment-195665782
I do see the benefits of propagating constraints in the physical plan but
if I understand correctly, wouldn't we have problems with joins?
For e.g., if a join has a condition on (`t1.a === t2.a && t1.a == 5`), just
inferring the constraints/filters during physical planning wouldn't give us the
liberty of pushing the additional `isNotNull(t1.a)`, `isNotNull(t2.a)` and
`t2.a === 5` filters all the way to the leaf nodes on either side of the join
in the query plan (there'd be no `PredicatePushdown` rules).
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at [email protected] or file a JIRA ticket
with INFRA.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
