ostronaut commented on code in PR #49144:
URL: https://github.com/apache/spark/pull/49144#discussion_r1904196842
##########
sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/optimizer/InsertMapSortExpression.scala:
##########
@@ -77,15 +71,56 @@ object InsertMapSortInGroupingExpressions extends
Rule[LogicalPlan] {
}.asInstanceOf[NamedExpression]
}
val newChild = Project(child.output ++ exprToMapSort.values, child)
- val newAgg = Aggregate(newGroupingKeys, newAggregateExprs, newChild)
+ val newAgg = Aggregate(newGroupingKeys, newAggregateExprs, newChild,
hint)
newAgg -> agg.output.zip(newAgg.output)
}
}
+}
+
+/**
+ * Adds [[MapSort]] to [[RepartitionByExpression]] expressions containing map
columns,
+ * as the key/value pairs need to be in the correct order before
repartitioning:
+ *
+ * SELECT * FROM TABLE DISTRIBUTE BY map_column =>
+ * SELECT * FROM TABLE DISTRIBUTE BY map_sort(map_column)
+ */
+object InsertMapSortInRepartitionExpressions extends Rule[LogicalPlan] {
Review Comment:
initially i wanted to do the same, but logic for
`InsertMapSortInGroupingExpressions` and
`InsertMapSortInRepartitionExpressions` is quite different: Grouping produces a
new output after applying the changes, while Repartition only updates existing
`RepartitionByExpression` by replacing `partitionExpressions`.
Also, there is a dependency between `InsertMapSortInGroupingExpressions` and
`PullOutGroupingExpressions`, as mentioned in [this
comment](https://github.com/apache/spark/blob/master/sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/optimizer/Optimizer.scala#L320).
For those reasons i decided to split them into separate Rules. But if you
think performance saving from reduced traverse will be significant, we can
combine those.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org
For additional commands, e-mail: reviews-h...@spark.apache.org
