cloud-fan commented on a change in pull request #34444:
URL: https://github.com/apache/spark/pull/34444#discussion_r740031392
##########
File path:
sql/core/src/main/scala/org/apache/spark/sql/execution/joins/ShuffledHashJoinExec.scala
##########
@@ -332,6 +327,266 @@ case class ShuffledHashJoinExec(
HashedRelationInfo(relationTerm, keyIsUnique = false, isEmpty = false)
}
+ override def doProduce(ctx: CodegenContext): String = {
+ // Specialize `doProduce` code for full outer join, because full outer
join needs to
+ // iterate streamed and build side separately.
+ if (joinType != FullOuter) {
+ return super.doProduce(ctx)
+ }
+
+ val HashedRelationInfo(relationTerm, _, _) = prepareRelation(ctx)
+
+ // Inline mutable state since not many join operations in a task
+ val keyIsUnique = ctx.addMutableState("boolean", "keyIsUnique",
+ v => s"$v = $relationTerm.keyIsUnique();", forceInline = true)
Review comment:
`prepareRelation` returns `keyIsUnique`, so it's a consistent before we
generating the code. We don't need to get it at runtime in the generated code.
##########
File path:
sql/core/src/main/scala/org/apache/spark/sql/execution/joins/ShuffledHashJoinExec.scala
##########
@@ -332,6 +327,266 @@ case class ShuffledHashJoinExec(
HashedRelationInfo(relationTerm, keyIsUnique = false, isEmpty = false)
}
+ override def doProduce(ctx: CodegenContext): String = {
+ // Specialize `doProduce` code for full outer join, because full outer
join needs to
+ // iterate streamed and build side separately.
+ if (joinType != FullOuter) {
+ return super.doProduce(ctx)
+ }
+
+ val HashedRelationInfo(relationTerm, _, _) = prepareRelation(ctx)
+
+ // Inline mutable state since not many join operations in a task
+ val keyIsUnique = ctx.addMutableState("boolean", "keyIsUnique",
+ v => s"$v = $relationTerm.keyIsUnique();", forceInline = true)
Review comment:
`prepareRelation` returns `keyIsUnique`, so it's a constant before we
generating the code. We don't need to get it at runtime in the generated code.
--
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]
