Github user cloud-fan commented on a diff in the pull request: https://github.com/apache/spark/pull/22326#discussion_r220109700 --- Diff: sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/optimizer/Optimizer.scala --- @@ -1234,6 +1237,59 @@ object PushPredicateThroughJoin extends Rule[LogicalPlan] with PredicateHelper { } } +/** + * Correctly handle PythonUDF which need access both side of join side by changing the new join + * type to Cross. + */ +object HandlePythonUDFInJoinCondition extends Rule[LogicalPlan] with PredicateHelper { + override def apply(plan: LogicalPlan): LogicalPlan = plan.resolveOperatorsUp { + case j @ Join(_, _, joinType, condition) + if condition.map(splitConjunctivePredicates).getOrElse(Nil).exists( + _.collectFirst { case udf: PythonUDF => udf }.isDefined) => + if (!joinType.isInstanceOf[InnerLike] && joinType != LeftSemi) { + // The current strategy only support InnerLike and LeftSemi join because other type + // can not simply be resolved by adding a Cross join. If we pass the plan here, it'll --- End diff -- it's not about cross join. It breaks SQL semantic if we run the join condition as a filter after join, for non-inner joins.
--- --------------------------------------------------------------------- To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org For additional commands, e-mail: reviews-h...@spark.apache.org