gianm commented on a change in pull request #9545: SQL support for joins on
subqueries.
URL: https://github.com/apache/druid/pull/9545#discussion_r395941085
##########
File path:
sql/src/test/java/org/apache/druid/sql/calcite/CalciteParameterQueryTest.java
##########
@@ -602,14 +602,13 @@ public void testWrongTypeParameter() throws Exception
and(
bound("l1", "3", null, true, false, null,
StringComparators.NUMERIC),
selector("f1", useDefault ? "0.0" : null, null)
-
)
)
.aggregators(aggregators(new CountAggregatorFactory("a0")))
.context(TIMESERIES_CONTEXT_DEFAULT)
.build()
) : ImmutableList.of(),
- useDefault ? ImmutableList.of() : ImmutableList.of(new Object[]{0L}),
+ ImmutableList.of(),
Review comment:
Your analysis sounds right to me. I tracked this down to the removal of
AggregateMergeRule, which I did because it was causing one of the new tests to
fail (testDoubleNestedGroupBy2). Without that rule loaded, this query gets
executed by a BindableAggregate, which doesn't seem to properly generate the
expected row.
Since this is a situation where two bugs are conflicting, I think it might
take some time to unwind. I'd rather look into that in a future patch. My
preference would be to do this:
1. Identify and fix the bug with AggregateMergeRule, and re-enable it. That
should get us back to normal on this test.
2. After that, stop using Bindable completely, and instead use native Druid
queries for everything (even queries on literal values). We can do this by
adding a few more capabilities to native Druid queries. Then we won't have to
worry about it.
----------------------------------------------------------------
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.
For queries about this service, please contact Infrastructure at:
[email protected]
With regards,
Apache Git Services
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
