yeshengm commented on a change in pull request #25164: [SPARK-28375][SQL] Prevent the PullupCorrelatedPredicates optimizer rule from removing predicates if run multiple times URL: https://github.com/apache/spark/pull/25164#discussion_r304765685
########## File path: sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/optimizer/subquery.scala ########## @@ -275,13 +275,16 @@ object PullupCorrelatedPredicates extends Rule[LogicalPlan] with PredicateHelper plan transformExpressions { case ScalarSubquery(sub, children, exprId) if children.nonEmpty => val (newPlan, newCond) = pullOutCorrelatedPredicates(sub, outerPlans) - ScalarSubquery(newPlan, newCond, exprId) + val conds = newCond ++ children.filter(_.isInstanceOf[Predicate]) Review comment: I doubt it... Because the logic for checking `OuterReferences` and the logic for actually pulling up predicates are slightly different. With that being said, even though `l.children` is non-empty, it does not necessarily mean that `newCond` is non-empty. The most natural way I can think of is that we combine these two rule `PullupCorrelatedPredicates` and `RewriteSubqueryPredicates`, `RewriteSubqueryPredicates` completely removes those hacky list subqueries. I don't think the plan can change if we apply these two rules in a single `Once` batch. ---------------------------------------------------------------- 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: us...@infra.apache.org With regards, Apache Git Services --------------------------------------------------------------------- To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org For additional commands, e-mail: reviews-h...@spark.apache.org