[
https://issues.apache.org/jira/browse/FLINK-4679?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15843627#comment-15843627
]
Haohui Mai commented on FLINK-4679:
-----------------------------------
There are two aspects here: (1) translating the over clause using the table
APIs, and (2) reusing the code from various jiras mentioned above to implement
a cleaner version of the table APIs. It looks to me that they are separate
issues.
Is it a good idea to teach the SQL layer to recognize {{OVER}} + {{PARTITION
BY}} and to directly translate it to the {{tumble}} call in the table APIs?
That way the evolution of the table APIs can be transparent to the SQL layer.
> Add TumbleRow row-windows to Table API
> --------------------------------------
>
> Key: FLINK-4679
> URL: https://issues.apache.org/jira/browse/FLINK-4679
> Project: Flink
> Issue Type: Sub-task
> Components: Table API & SQL
> Affects Versions: 1.2.0
> Reporter: Fabian Hueske
> Assignee: Jark Wu
>
> Add TumbleRow row-windows for streaming tables as described in
> [FLIP-11|https://cwiki.apache.org/confluence/display/FLINK/FLIP-11%3A+Table+API+Stream+Aggregations].
>
> TumbleRow row windows are a syntactic shortcut for a special type of SQL OVER
> windows, i.e., windows of the from
> {code}
> SELECT STREAM rowtime,
> productId,
> units,
> SUM(units) OVER (PARTITION BY FLOOR(rowtime() TO HOUR)) AS
> unitsSinceTopOfHour
> FROM Orders;
> {code}
> i.e., OVER windows which are partitioned by the time attribute.
> This issue is about extending the Table API with the TumbleRow shortcut. It
> should reuse the OVER window translation and runtime code implemented by
> FLINK-5653, FLINK-5654, FLINK-5655, FLINK-5656, FLINK-5657, and FLINK-5658.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
