Github user marmbrus commented on a diff in the pull request:
https://github.com/apache/spark/pull/1785#discussion_r15831030
--- Diff:
sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/analysis/HiveTypeCoercion.scala
---
@@ -336,28 +338,33 @@ trait HiveTypeCoercion {
}
/**
- * Ensures that NullType gets casted to some other types under certain
circumstances.
+ * Coerces the type of different branches of a CASE WHEN statement to a
common type.
*/
- object CastNulls extends Rule[LogicalPlan] {
+ object CaseWhenCoercion extends Rule[LogicalPlan] with TypeWidening {
def apply(plan: LogicalPlan): LogicalPlan = plan
transformAllExpressions {
- case cw @ CaseWhen(branches) =>
+ case cw @ CaseWhen(branches) if !cw.resolved &&
!branches.exists(!_.resolved) =>
val valueTypes = branches.sliding(2, 2).map {
- case Seq(_, value) if value.resolved => Some(value.dataType)
- case Seq(elseVal) if elseVal.resolved => Some(elseVal.dataType)
- case _ => None
+ case Seq(_, value) => value.dataType
+ case Seq(elseVal) => elseVal.dataType
}.toSeq
- if (valueTypes.distinct.size == 2 && valueTypes.exists(_ ==
Some(NullType))) {
- val otherType = valueTypes.filterNot(_ == Some(NullType))(0).get
+
+ logDebug(s"Input values for null casting
${valueTypes.mkString(",")}")
--- End diff --
Its essentially free and could be useful to turn on if we ever have
problems with this rule again.
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at [email protected] or file a JIRA ticket
with INFRA.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
