You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@calcite.apache.org by "Yingyu (Jira)" <ji...@apache.org> on 2021/06/03 17:39:00 UTC
[jira] [Updated] (CALCITE-4634) Improve
AggregateProjectPullUpConstantsRule to remove all constant keys
[ https://issues.apache.org/jira/browse/CALCITE-4634?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Yingyu updated CALCITE-4634:
----------------------------
Description:
*AggregateProjectPullUpConstantsRule* removes constant keys from an Aggregate. However, this rules never removes the last column, because Aggregate([]) returns 1 row even if its input is empty.
The problem is if the remaining last column is a constant, it will fail on *SQL Server* with this error:
"Each GROUP BY expression must contain at least one column that is not an outer reference."
To improve on this we can improve the AggregateProjectPullUpConstantsRule to:
- Remove all constant keys from an Aggregate
- When no GROUP BY column is left, add a filter to the Aggregate([]) so that it returns empty dataset if its input is empty.
E.g.
select 'abc' from table_name group by ('abc')
will be translated to:
select 'abc' from table_name group by () having count(*) > 0
was:
AggregateProjectPullUpConstantsRule removes constant keys from an Aggregate. However, this rules never removes the last column, because Aggregate([]) returns 1 row even if its input is empty.
The problem is if the remaining last column is a constant, it will fail on SQL Server with this error:
"Each GROUP BY expression must contain at least one column that is not an outer reference."
To improve on this we can improve the AggregateProjectPullUpConstantsRule to:
- Remove all constant keys from an Aggregate
- When no GROUP BY column is left, add a filter to the Aggregate([]) so that it returns empty dataset if its input is empty.
E.g.
select 'abc' from table_name group by ('abc')
will be translated to:
select 'abc' from table_name group by () having count(*) > 0
> Improve AggregateProjectPullUpConstantsRule to remove all constant keys
> -----------------------------------------------------------------------
>
> Key: CALCITE-4634
> URL: https://issues.apache.org/jira/browse/CALCITE-4634
> Project: Calcite
> Issue Type: Improvement
> Components: core
> Affects Versions: 1.16.0
> Reporter: Yingyu
> Priority: Major
>
> *AggregateProjectPullUpConstantsRule* removes constant keys from an Aggregate. However, this rules never removes the last column, because Aggregate([]) returns 1 row even if its input is empty.
> The problem is if the remaining last column is a constant, it will fail on *SQL Server* with this error:
> "Each GROUP BY expression must contain at least one column that is not an outer reference."
>
> To improve on this we can improve the AggregateProjectPullUpConstantsRule to:
> - Remove all constant keys from an Aggregate
> - When no GROUP BY column is left, add a filter to the Aggregate([]) so that it returns empty dataset if its input is empty.
>
> E.g.
> select 'abc' from table_name group by ('abc')
> will be translated to:
> select 'abc' from table_name group by () having count(*) > 0
--
This message was sent by Atlassian Jira
(v8.3.4#803005)