tanelk commented on a change in pull request #29092:
URL: https://github.com/apache/spark/pull/29092#discussion_r501167719
##########
File path:
sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/optimizer/Optimizer.scala
##########
@@ -1854,3 +1855,37 @@ object OptimizeLimitZero extends Rule[LogicalPlan] {
empty(ll)
}
}
+
+/**
+ * Infers filters from [[Generate]], such that rows that would have been
removed
+ * by this [[Generate]] can be removed earlier - before joins and in data
sources.
+ */
+object InferFiltersFromGenerate extends Rule[LogicalPlan] {
+ def apply(plan: LogicalPlan): LogicalPlan = plan transformUp {
+ case generate @ Generate(e: ExplodeBase, _, false, _, _, _)
+ if e.deterministic && !e.child.foldable =>
Review comment:
Seeing the failed results there is another obvious issue with allowing
foldable inferred filters - the idempotence is broken. I would very much like
to keep this rule idempotent.
Should I try to include foldable inferred filters while keeping it
idempotent or shall I revert the last change?
----------------------------------------------------------------
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.
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
