[
https://issues.apache.org/jira/browse/FLINK-33490?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17789494#comment-17789494
]
lincoln lee commented on FLINK-33490:
-------------------------------------
Here're my two cents: from my previous reference to the sql standard, it
forbids two elements at different positions from having equivalent names,
especially when created through a sql schema statement, but I didn't find any
specific recommendations for the query part, this is also consistent with
testing provided by the [~xuyangzhong], these rdbms are indeed allowed
duplicate columns in select statement(select * from t1 l, t1 r).
So, for the create view statements involved in this issue, I would prefer to
keep the behavior consistent with all the create statements (create
[materialized] view / table [as query]) and follow the standard.
I'm fine with initiating a FLIP or not.
> Validate the name conflicts when creating view
> ----------------------------------------------
>
> Key: FLINK-33490
> URL: https://issues.apache.org/jira/browse/FLINK-33490
> Project: Flink
> Issue Type: Bug
> Components: Table SQL / Planner
> Affects Versions: 1.19.0
> Reporter: Shengkai Fang
> Assignee: xuyang
> Priority: Major
> Labels: pull-request-available
>
> We should forbid
> ```
> CREATE VIEW id_view AS
> SELECT id, uid AS id FROM id_table
> ```
> As the SQL standards states,
> If <regular view specification> is specified, then:
> i) If any two columns in the table specified by the <query expression> have
> equivalent <column name>s, or if any column of that table has an
> implementation-dependent name, then a <view column list> shall be specified.
> ii) Equivalent <column name>s shall not be specified more than once in the
> <view column list>.
> Many databases also throw exception when view name conflicts, e.g. mysql,
> postgres.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
