[
https://issues.apache.org/jira/browse/FLINK-3640?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15232194#comment-15232194
]
ASF GitHub Bot commented on FLINK-3640:
---------------------------------------
Github user fhueske commented on the pull request:
https://github.com/apache/flink/pull/1862#issuecomment-207438727
Thanks for the PR. I had a few minor comments but otherwise it looks really
good.
There are a few follow up issues, IMO:
- Check if we somehow can get around the `EnumerableToLogicalTableScan`.
Maybe the Calcite community can help. I will open a JIRA for this once the PR
is merged.
- Check how we can exclude unsupported SQL features such as outer joins,
intersection, etc. Also here, the Calcite community should be able to help. I
will open a JIRA for this once the PR is merged.
- Refactor `TranslationContext` and `TableEnvironment` to prevent that the
same planner is used several times. I'll start a discussion about this soon.
> Add support for SQL queries in DataSet programs
> -----------------------------------------------
>
> Key: FLINK-3640
> URL: https://issues.apache.org/jira/browse/FLINK-3640
> Project: Flink
> Issue Type: New Feature
> Components: Table API
> Affects Versions: 1.1.0
> Reporter: Vasia Kalavri
> Assignee: Vasia Kalavri
>
> This issue covers the task of supporting SQL queries embedded in DataSet
> programs. In this mode, the input and output of a SQL query is a Table. For
> this issue, we need to make the following additions to the Table API:
> - add a {{tEnv.sql(query: String): Table}} method for converting a query
> result into a Table
> - integrate Calcite's SQL parser into the batch Table API translation process.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
