sunchao commented on code in PR #5972:
URL: https://github.com/apache/datafusion-comet/pull/5972#discussion_r4178707336
##########
spark/src/main/scala/org/apache/comet/rules/CometExecRule.scala:
##########
@@ -814,6 +825,249 @@ case class CometExecRule(session: SparkSession)
}
}
+ /** Keep ANSI next_day in Spark's row pipeline where consumers can skip or
catch its errors. */
+ private def preserveNextDayEvaluationMasks(plan: SparkPlan): SparkPlan = {
+ def nextDayName(expr: Expression): Option[String] = expr.collectFirst {
+ // Both the native kernel and dispatcher can throw on rows that Spark
skips.
+ case nextDay: NextDay if nextDay.failOnError => "next_day"
Review Comment:
Fixed in `8f033a6952`: valid literal weekdays (including Monday) and literal
NULL do not enroll in the ANSI error mask. The new native-admission fixture
exercises the Monday-under-LIMIT case beside eager parent expressions, sorting
and unfiltered MAX. Nine next_day SQL cases and six focused planner regressions
pass on Spark 4.1.3; the shared nine-test unbase64 mask suite also passes after
rebasing onto #5533.
--
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]
