[
https://issues.apache.org/jira/browse/FLINK-22043?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17312086#comment-17312086
]
Yun Tang edited comment on FLINK-22043 at 3/31/21, 7:23 AM:
------------------------------------------------------------
[~hackergin] Actually, I have no idea of why we only need to support this in
SQL, what will happen if users use thrift in their UDF code when using SQL?
In general, I prefer to support thrift in general case unless we really have
some difficulties. Have you ever considered your current implementation to
support general case instead of only supporting SQL? I think extending it to
support general case should not be something complex.
was (Author: yunta):
[~hackergin] Actually, I have no idea of why we only need to support this in
SQL, what will happen if users use thrift in their UDF code when using SQL?
In general, I prefer to support thrift in general case especially we really
have some difficulties. Have you ever considered your current implementation to
support general case instead of only supporting SQL? I think extending it to
support general case should not be something complex.
> Introduce thrift format to (de)serialize thrift message only with table
> -------------------------------------------------------------------------
>
> Key: FLINK-22043
> URL: https://issues.apache.org/jira/browse/FLINK-22043
> Project: Flink
> Issue Type: New Feature
> Components: Formats (JSON, Avro, Parquet, ORC, SequenceFile)
> Reporter: jinfeng
> Priority: Major
>
> We have implement thrift format in our company. I found that there was a mail
> discuss
> http://apache-flink-mailing-list-archive.1008284.n3.nabble.com/thrift-support-td43257.html
> , but no one seems to have time to implement it.
> [~yuyang08] has done most of the things , I can help implement it based on
> https://github.com/apache/flink/pull/8067
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
