You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@flink.apache.org by "Jark Wu (Jira)" <ji...@apache.org> on 2020/10/14 09:51:00 UTC

[jira] [Assigned] (FLINK-19639) Support nested push down for SupportsProjectionPushDown interface in planner

     [ https://issues.apache.org/jira/browse/FLINK-19639?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Jark Wu reassigned FLINK-19639:
-------------------------------

    Assignee: Shengkai Fang

> Support nested push down for SupportsProjectionPushDown interface in planner
> ----------------------------------------------------------------------------
>
>                 Key: FLINK-19639
>                 URL: https://issues.apache.org/jira/browse/FLINK-19639
>             Project: Flink
>          Issue Type: Sub-task
>          Components: Table SQL / Planner
>            Reporter: Shengkai Fang
>            Assignee: Shengkai Fang
>            Priority: Major
>
> Support to push nested projection into {{TableSourceScan}}. But it may cause name conflicts in some situaion. For example, we create the table with ddl
> {code:sql}
> CREATE TABLE NestedTable (
>      nest1 ROW<a INT>,
>      nest2 ROW<a INT>
> )
> {code}
> and with query
> {code:sql}
> SELECT nest1.a, nest2.a from NestedTable
> {code}
> and we will get 2 {{`a`}} in the new schema when pushing projection. 
>  Here we useĀ {{'_'}} to concatenate the names of all levels as the name of the nested fields. In this example, we will get fields {{nest1_a, nest2_a}} in the new schema.



--
This message was sent by Atlassian Jira
(v8.3.4#803005)