mihaibudiu commented on code in PR #4495:
URL: https://github.com/apache/calcite/pull/4495#discussion_r2264172162
##########
core/src/main/java/org/apache/calcite/tools/RelBuilder.java:
##########
@@ -2819,28 +2819,56 @@ private RelBuilder aggregate_(ImmutableBitSet groupSet,
* GROUP_ID returns wrong result</a> and
* <a
href="https://issues.apache.org/jira/browse/CALCITE-4748";>[CALCITE-4748]
* If there are duplicate GROUPING SETS, Calcite should return duplicate
- * rows</a>.
+ * rows</a> and
+ * <a
href="https://issues.apache.org/jira/browse/CALCITE-7126";>[CALCITE-7126]
+ * The calculation result of grouping function is wrong</a>.
*/
private RelBuilder rewriteAggregateWithDuplicateGroupSets(
ImmutableBitSet groupSet,
ImmutableSortedMultiset<ImmutableBitSet> groupSets,
List<AggCallPlus> aggregateCalls) {
+ List<AggregateCall> calls =
+ aggregateCalls.stream()
+ .map(AggCallPlus::aggregateCall)
+ .collect(Collectors.toList());
+ return rewriteAggregateWithDuplicateGroupSetsImpl(groupSet, groupSets,
calls);
+ }
+
+ private RelBuilder rewriteAggregateWithDuplicateGroupSetsImpl(
+ ImmutableBitSet groupSet,
+ ImmutableSortedMultiset<ImmutableBitSet> groupSets,
+ List<AggregateCall> aggregateCalls) {
+ final int groupCount = groupSet.cardinality();
+ // No need to handle the case without GROUP BY in aggregate.
+ if (groupCount > 0) {
+ aggregateCalls = aggregateCalls.stream()
+ .map(c -> c.withAllowChangeNullable(false))
Review Comment:
why do you need to store the flag in the aggregate call itself?
Can't you just keep a map from call to flag in the builder?
--
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: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
