cloud-fan commented on a change in pull request #35149:
URL: https://github.com/apache/spark/pull/35149#discussion_r781000668
##########
File path:
sql/core/src/main/scala/org/apache/spark/sql/execution/adaptive/AQEPropagateEmptyRelation.scala
##########
@@ -32,14 +33,24 @@ import
org.apache.spark.sql.execution.joins.HashedRelationWithAllNullKeys
*/
object AQEPropagateEmptyRelation extends PropagateEmptyRelationBase {
override protected def isEmpty(plan: LogicalPlan): Boolean =
- super.isEmpty(plan) || getRowCount(plan).contains(0)
+ super.isEmpty(plan) || getEstimatedRowCount(plan).contains(0)
override protected def nonEmpty(plan: LogicalPlan): Boolean =
- super.nonEmpty(plan) || getRowCount(plan).exists(_ > 0)
+ super.nonEmpty(plan) || getEstimatedRowCount(plan).exists(_ > 0)
- private def getRowCount(plan: LogicalPlan): Option[BigInt] = plan match {
+ private def getEstimatedRowCount(plan: LogicalPlan): Option[BigInt] = plan
match {
Review comment:
let's document the assumptions: `0` means the plan must produce 0 row.
Positive value means an estimated row count which can be over-estimated.
--
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.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
