[
https://issues.apache.org/jira/browse/FLINK-5315?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16568255#comment-16568255
]
Rong Rong commented on FLINK-5315:
----------------------------------
You are right, in the actual translation {{distinct}} is translated to a all
field {{GROUP GY}} in the {{select}} statement. so yes. I might have chosen the
wright word syntactically they are kind of similar: like I mentioned before.
{{select('a, 'b).distinct}} should considered to be {{select.distinct('a, 'b)}}
syntactically since it is aligned with the SQL expression {{SELECT DISTINCT a,
b FROM table}}.
But I see your point, making the {{distinct}} directly stick with the
aggregation expression/function makes it clear what it is modified with.
Do you think we should aligned the syntax? e.g. (1) making {{select.distinct('a
,'b)}} or (2) having {{udagg('a, 'b).distinct}}.
The reason why I brought this up is I am in the SQL domain most of the time,
and {{SELECT DISTINCT a, b FROM table}} and {{SELECT COUNT(DISTINCT a),
SUM(DISTINCT b) FROM table}} are aligned and make perfect sense to me.
> Support distinct aggregations in table api
> ------------------------------------------
>
> Key: FLINK-5315
> URL: https://issues.apache.org/jira/browse/FLINK-5315
> Project: Flink
> Issue Type: Sub-task
> Components: Table API & SQL
> Reporter: Kurt Young
> Assignee: Rong Rong
> Priority: Major
>
> Support distinct aggregations in Table API in the following format:
> For Expressions:
> {code:scala}
> 'a.count.distinct // Expressions distinct modifier
> {code}
> For User-defined Function:
> {code:scala}
> singleArgUdaggFunc.distinct('a) // FunctionCall distinct modifier
> multiArgUdaggFunc.distinct('a, 'b) // FunctionCall distinct modifier
> {code}
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
