Github user viirya commented on a diff in the pull request:
https://github.com/apache/spark/pull/11298#discussion_r53759608
--- Diff:
sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/analysis/Analyzer.scala
---
@@ -607,14 +607,21 @@ class Analyzer(
def apply(plan: LogicalPlan): LogicalPlan = plan resolveOperators {
// Skip sort with aggregate. This will be handled in
ResolveAggregateFunctions
case sa @ Sort(_, _, child: Aggregate) => sa
-
+ case sa @ Sort(_, _, _)
+ if sa.order.exists(so =>
ResolveAggregateFunctions.containsAggregate(so.child)) => sa
case s @ Sort(order, _, child) if !s.resolved && child.resolved =>
try {
val newOrder = order.map(resolveExpressionRecursively(_,
child).asInstanceOf[SortOrder])
val requiredAttrs = AttributeSet(newOrder).filter(_.resolved)
val missingAttrs = requiredAttrs -- child.outputSet
- if (missingAttrs.nonEmpty) {
- // Add missing attributes and then project them away after the
sort.
+ // resolve the unresolved functions for knowing if they are
aggregate functions
+ val evaluatedSort =
+ ResolveFunctions.resolveFunctions(Sort(newOrder, s.global,
child)).asInstanceOf[Sort]
+ if (evaluatedSort.order.exists(so =>
+ ResolveAggregateFunctions.containsAggregate(so.child))) {
--- End diff --
Can we move this part of resolving functions of Sort as another pattern
matching case? So this original pattern matching case can be untouched? I think
it will more clear.
---
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]
