[
https://issues.apache.org/jira/browse/FLINK-18674?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17164244#comment-17164244
]
Timo Walther commented on FLINK-18674:
--------------------------------------
TBH I already expected that this topic will come up. Because sometimes users
might have a bigger preprocessing pipeline before they emit data to the SQL
engine. First of all we should clarify what are the use cases for users and
would FLIP-27 help in such cases. If not, we can think about adding a special
ScanRuntimeProvider for advanced users.
> Support to bridge Transformation (DataStream) with FLIP-95 interface?
> ---------------------------------------------------------------------
>
> Key: FLINK-18674
> URL: https://issues.apache.org/jira/browse/FLINK-18674
> Project: Flink
> Issue Type: Sub-task
> Components: Table SQL / API
> Reporter: Jark Wu
> Priority: Major
>
> A user complained in the user ML [1] that the old connector loigic is hard to
> migrate to the FLIP-95 interfaces, because they heavily used DataStream in
> the TableSource/TableSink and it is not possible to replace with new
> interface right now.
> This issue can be used to collect the user requirements around
> DataStream/Transformation + FLIP-95. We can also evaluate/discuss whether and
> how to support it.
> [1]: http://apache-flink.147419.n8.nabble.com/1-11Flink-SQL-API-tp5261.html
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
