dongjoon-hyun commented on a change in pull request #23726: [SPARK-26812][SQL]
Report correct nullability for complex datatypes in Union
URL: https://github.com/apache/spark/pull/23726#discussion_r255319845
##########
File path:
sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/plans/logical/basicLogicalOperators.scala
##########
@@ -230,9 +231,19 @@ case class Union(children: Seq[LogicalPlan]) extends
LogicalPlan {
}
// updating nullability to make all the children consistent
- override def output: Seq[Attribute] =
- children.map(_.output).transpose.map(attrs =>
- attrs.head.withNullability(attrs.exists(_.nullable)))
+ override def output: Seq[Attribute] = {
+ children.map(_.output).transpose.map { attrs =>
+ val firstAttr = attrs.head
+ val outAttr = try {
+ firstAttr.withDataType(attrs.map(_.dataType).reduce(StructType.merge))
+ } catch {
+ // This should never happen, since `output` is called only when the
plan is `resolved`, so
Review comment:
I understand that this is a fallback logic. But, if this should never
happen, we should use more explicit ways. We need to find bugs eagerly. For
example, `assert()`, `logWarning`, or `not using try` at all.
----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
[email protected]
With regards,
Apache Git Services
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
