yadavay-amzn commented on code in PR #56984:
URL: https://github.com/apache/spark/pull/56984#discussion_r3521430894
##########
sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/analysis/CheckAnalysis.scala:
##########
@@ -653,6 +653,7 @@ trait CheckAnalysis extends LookupCatalog with
QueryErrorsBase with PlanToString
case s: StructType
if s.find(_.name == "end").map(_.dataType) ==
Some(TimestampType) =>
Review Comment:
Widened the window-struct guard to accept a nanosecond end field
(AnyTimestampNanoType) alongside TimestampType. It is unreachable today
(TimeWindow rejects nanos until SPARK-57829 lands), so no behavior change now,
but it prevents a window-over-nanos struct from falling through to the reject
case once window-over-nanos is enabled. The getLong(0)-on-struct in the exec
stats path is a separate pre-existing issue for all struct event-times, so I
left it out of scope here.
##########
sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/analysis/UnsupportedOperationChecker.scala:
##########
@@ -670,6 +690,21 @@ object UnsupportedOperationChecker extends Logging {
}
private def checkForStreamStreamJoinWatermark(join: Join): Unit = {
Review Comment:
Removed - the hoisted check fires first for every stream-stream join (incl
InnerLike), so this block was dead. The inner-join negative test still passes.
--
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]
