You are viewing a plain text version of this content. The canonical link for it is here.

Posted to issues@flink.apache.org by "Hequn Cheng (JIRA)" <ji...@apache.org> on 2018/10/11 03:59:00 UTC

[jira] [Commented] (FLINK-7999) Variable Join Window Boundaries

    [ https://issues.apache.org/jira/browse/FLINK-7999?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16645931#comment-16645931 ] 

Hequn Cheng commented on FLINK-7999:
------------------------------------

I think [FLINK-8478|https://issues.apache.org/jira/browse/FLINK-8478] and [FLINK-5725|https://issues.apache.org/jira/browse/FLINK-5725] both can solve the problem. I will close this task.

> Variable Join Window Boundaries
> -------------------------------
>
>                 Key: FLINK-7999
>                 URL: https://issues.apache.org/jira/browse/FLINK-7999
>             Project: Flink
>          Issue Type: New Feature
>          Components: Table API &amp; SQL
>            Reporter: Seth Wiesman
>            Priority: Major
>
> Allow window joins with variable length based on row attributes. 
> Consider a two streams joined on an id, where one has start and end dates, it would be useful to be able to join each row during is live durations. Today this can be expressed in the datastream api using a CoProcessFunction. 
>  left.id = right.id AND (left.time > right.start and left.time < right.end)



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)