snuyanzin commented on code in PR #20850:
URL: https://github.com/apache/flink/pull/20850#discussion_r993941510
##########
flink-table/flink-table-planner/src/test/resources/org/apache/flink/table/planner/catalog/JavaCatalogTableTest.xml:
##########
@@ -76,12 +75,11 @@ LogicalProject(EXPR$0=[$0])
</Resource>
<Resource name="optimized exec plan">
<![CDATA[
-GroupWindowAggregate(window=[TumblingGroupWindow('w, proctime, 600000)],
select=[COUNT($f3) AS EXPR$0])
+GroupWindowAggregate(window=[TumblingGroupWindow('w, proctime, 600000)],
select=[COUNT(*) AS EXPR$0])
Review Comment:
After debugging I found a reason for that. It is invoked by the change
within [1] . To be more precise by optimization within these lines in
_org.apache.calcite.tools.RelBuilder_ (line number 3402 - 3406)
```java
if (aggFunction instanceof SqlCountAggFunction && !distinct) {
args = args.stream()
.filter(r::fieldIsNullable)
.collect(Util.toImmutableList());
}
```
I did a double check for that: copied a class to Flink repo and commented
only this line and then no need to change rules (this, two below where you also
commented, and some more, full changeset [2]), ci for that is also green except
one non related Hive issue [3]
[1]
https://github.com/apache/calcite/commit/a680abe0528faacf8d284419c76c36bbc1a65aa3
[2]
https://github.com/snuyanzin/flink/commit/481543210509f682ac89ef3ce42d2f324937bbc4
[3]
https://dev.azure.com/snuyanzin/flink/_build/results?buildId=1450&view=results
--
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]
