Github user attilapiros commented on a diff in the pull request:
https://github.com/apache/spark/pull/20069#discussion_r158885112
--- Diff:
sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/expressions/predicates.scala
---
@@ -47,6 +48,15 @@ trait Predicate extends Expression {
trait PredicateHelper {
+
+ protected def partitionByDeterminism(predicates: Seq[Expression]) = {
+ if (SQLConf.get.outOfOrderPredicateEvaluationEnabled) {
+ predicates.partition(_.deterministic)
+ } else {
+ predicates.span(_.deterministic)
+ }
+ }
+
--- End diff --
Is not this "if" in the other way around here? As Seq#partition keeps the
relative orders and Seq#span does not.
After reversing this "if" the "outOfOrderPredicateEvaluationEnabled"
default value (true) will be consistent with the old behavior, before your PR
(I mean using the Seq#span), but in the unit tests the relevant "if" conditions
should be changed too.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org
For additional commands, e-mail: reviews-h...@spark.apache.org
