[
https://issues.apache.org/jira/browse/FLINK-18445?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17759576#comment-17759576
]
lincoln lee commented on FLINK-18445:
-------------------------------------
This can be an useful optimization for lookup joins, will make it in 1.19.
> Short circuit join condition for lookup join
> --------------------------------------------
>
> Key: FLINK-18445
> URL: https://issues.apache.org/jira/browse/FLINK-18445
> Project: Flink
> Issue Type: Improvement
> Components: Table SQL / Planner
> Reporter: Rui Li
> Assignee: lincoln lee
> Priority: Not a Priority
> Labels: auto-deprioritized-major, auto-deprioritized-minor
>
> Consider the following query:
> {code}
> select *
> from probe
> left join
> build for system_time as of probe.ts
> on probe.key=build.key and probe.col is not null
> {code}
> In current implementation, we lookup each probe.key in build to decide
> whether a match is found. A possible optimization is to skip the lookup for
> rows whose {{col}} is null.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
