Github user rxin commented on a diff in the pull request:
https://github.com/apache/spark/pull/11815#discussion_r56618306
--- Diff:
sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/analysis/Analyzer.scala
---
@@ -618,9 +622,33 @@ class Analyzer(
* clause. This rule detects such queries and adds the required
attributes to the original
* projection, so that they will be available during sorting. Another
projection is added to
* remove these attributes after sorting.
+ *
+ * This rule also resolves the position number in Sort references. This
support is introduced
+ * in Spark 2.0. Before Spark 2.0, the integers in Order By has no
effect on output sorting.
*/
object ResolveSortReferences extends Rule[LogicalPlan] {
def apply(plan: LogicalPlan): LogicalPlan = plan resolveOperators {
+ // Replace the index with the related attribute for ORDER BY
+ // which is a 1-base position of the projection list.
+ case s @ Sort(orders, global, child)
+ if child.resolved && orders.exists(o =>
IntegerIndex.unapply(o.child).nonEmpty) =>
+ val newOrders = orders map {
+ case s @ SortOrder(IntegerIndex(index), direction) =>
+ if (index > 0 && index <= child.output.size &&
conf.orderByOrdinal) {
+ SortOrder(child.output(index - 1), direction)
+ } else if (!conf.orderByOrdinal) {
+ failAnalysis(
+ "Integer in the Order/Sort By clause is not allowed " +
--- End diff --
i'd say if the flag is set to false, we should just ignore the foldable
constants.
---
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]
