Github user davies commented on a diff in the pull request:
https://github.com/apache/spark/pull/11153#discussion_r52646199
--- Diff:
sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/analysis/Analyzer.scala
---
@@ -572,98 +572,64 @@ class Analyzer(
// Skip sort with aggregate. This will be handled in
ResolveAggregateFunctions
case sa @ Sort(_, _, child: Aggregate) => sa
- case s @ Sort(_, _, child) if !s.resolved && child.resolved =>
- val (newOrdering, missingResolvableAttrs) =
collectResolvableMissingAttrs(s.order, child)
-
- if (missingResolvableAttrs.isEmpty) {
- val unresolvableAttrs = s.order.filterNot(_.resolved)
- logDebug(s"Failed to find $unresolvableAttrs in
${child.output.mkString(", ")}")
- s // Nothing we can do here. Return original plan.
- } else {
- // Add the missing attributes into projectList of Project/Window
or
- // aggregateExpressions of Aggregate, if they are in the
inputSet
- // but not in the outputSet of the plan.
- val newChild = child transformUp {
- case p: Project =>
- p.copy(projectList = p.projectList ++
- missingResolvableAttrs.filter((p.inputSet --
p.outputSet).contains))
- case w: Window =>
- w.copy(projectList = w.projectList ++
- missingResolvableAttrs.filter((w.inputSet --
w.outputSet).contains))
- case a: Aggregate =>
- val resolvableAttrs =
missingResolvableAttrs.filter(a.groupingExpressions.contains)
- val notResolvedAttrs =
resolvableAttrs.filterNot(a.aggregateExpressions.contains)
- val newAggregateExpressions = a.aggregateExpressions ++
notResolvedAttrs
- a.copy(aggregateExpressions = newAggregateExpressions)
- case o => o
- }
-
+ case s @ Sort(order, _, child) if !s.resolved && child.resolved =>
+ 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.
Project(child.output,
- Sort(newOrdering, s.global, newChild))
+ Sort(newOrder, s.global, addMissingAttr(child, missingAttrs)))
+ } else if (newOrder != order) {
+ s.copy(order = newOrder)
+ } else {
+ s
}
}
/**
- * Traverse the tree until resolving the sorting attributes
- * Return all the resolvable missing sorting attributes
- */
- @tailrec
- private def collectResolvableMissingAttrs(
- ordering: Seq[SortOrder],
- plan: LogicalPlan): (Seq[SortOrder], Seq[Attribute]) = {
+ * Add the missing attributes into projectList of Project/Window or
aggregateExpressions of
+ * Aggregate.
+ */
+ private def addMissingAttr(plan: LogicalPlan, missingAttrs:
AttributeSet): LogicalPlan = {
--- End diff --
Because it can not go over JOIN, it's very uncommon to have thousands of
unary nodes in practice.
---
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]
