allisonwang-db commented on a change in pull request #32303:
URL: https://github.com/apache/spark/pull/32303#discussion_r647142502
##########
File path:
sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/parser/AstBuilder.scala
##########
@@ -871,7 +871,13 @@ class AstBuilder extends SqlBaseBaseVisitor[AnyRef] with
SQLConfHelper with Logg
override def visitFromClause(ctx: FromClauseContext): LogicalPlan =
withOrigin(ctx) {
val from = ctx.relation.asScala.foldLeft(null: LogicalPlan) { (left,
relation) =>
val right = plan(relation.relationPrimary)
- val join = right.optionalMap(left)(Join(_, _, Inner, None,
JoinHint.NONE))
+ val join = right.optionalMap(left) { (left, right) =>
+ if (relation.LATERAL != null) {
+ LateralJoin(left, LateralSubquery(right), Inner, None)
Review comment:
Actually, does it make sense to have join hints for lateral joins? A
lateral join is essentially a nested loop join. Ideally, the evaluation logic
should be for each row in the left, plug the outer query attribute values into
the outer references and evaluate the subquery. So it should only be planned as
a (correlated) nested loop join. But since Spark doesn't support such
execution, it first decorrelates the subquery and then rewrites the lateral
join as a normal join. This seems to be an implementation detail and it doesn't
make sense to add join hints here.
--
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
---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org
For additional commands, e-mail: reviews-h...@spark.apache.org
