Github user viirya commented on the issue:
https://github.com/apache/spark/pull/14912
@nsyca Thanks for your detailed comment. I would like to leave the decision
of predicate transformation to later PRs, as this PR is not motivated by this.
I think to simplify a predicate such as `(a > 10 || b > 2) && (a > 10 || c
== 3)` to `(a > 10) || (b > 2 && c == 3)`, is to eliminate redundant filtering
expressions needed to run in `Filter` in execution time.
As I said in before comment, my first opinion is not to complicate the
predicate handling too much. We can keep a form of predicate which benefits
predicate pushdown most, I guess the form should be CNF. We can also keep the
simplified form of predicate which is better for execution in `Filter`.
---
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]
