cloud-fan commented on a change in pull request #27232:
[SPARK-30525][SQL]HiveTableScanExec do not need to prune partitions again after
pushing down to SessionCatalog for partition pruning
URL: https://github.com/apache/spark/pull/27232#discussion_r373916523
##########
File path:
sql/hive/src/main/scala/org/apache/spark/sql/hive/execution/HiveTableScanExec.scala
##########
@@ -162,11 +162,29 @@ case class HiveTableScanExec(
}
}
+ @transient lazy val prunedPartitions: Seq[HivePartition] = {
+ if (relation.prunedPartitions.nonEmpty) {
+ val hivePartitions =
+ relation.prunedPartitions.get.map(HiveClientImpl.toHivePartition(_,
hiveQlTable))
+ if (partitionPruningPred.forall(!ExecSubqueryExpression.hasSubquery(_)))
{
+ hivePartitions
+ } else {
+ prunePartitions(hivePartitions)
+ }
+ } else {
+ if (sparkSession.sessionState.conf.metastorePartitionPruning) {
+ rawPartitions
+ } else {
+ prunePartitions(rawPartitions)
+ }
+ }
+ }
+
// exposed for tests
- @transient lazy val rawPartitions = {
+ @transient lazy val rawPartitions: Seq[HivePartition] = {
Review comment:
when we call `rawPartitions`, the `relation.prunedPartitions` must be empty.
We can remove `relation.prunedPartitions.getOrElse` below.
----------------------------------------------------------------
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]
With regards,
Apache Git Services
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
