[
https://issues.apache.org/jira/browse/FLINK-19869?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Martijn Visser updated FLINK-19869:
-----------------------------------
Summary: Flink SQL JDBC connector for Postgres can't handle UUID datatype
(was: Support java.util.UUID as RAW type in PostgresCatalog and handle it in
PostgresRowConverter)
> Flink SQL JDBC connector for Postgres can't handle UUID datatype
> ----------------------------------------------------------------
>
> Key: FLINK-19869
> URL: https://issues.apache.org/jira/browse/FLINK-19869
> Project: Flink
> Issue Type: Improvement
> Components: Connectors / JDBC
> Affects Versions: 1.11.2
> Reporter: Jingwei Zhang
> Priority: Minor
>
> Problem:
> UUID is not a SQL standard type. But it is supported by Postgres as an
> internal type which postgres client will help to convert to java.util.UUID
> when queried. However, it seems flink doesn't support it as a sql extension
> in PostgresCatalog. There is no uuid serializer provided for it, either. So
> if I want to filter a result set and put the filter result back to a table
> with same schema with uuid column type in it. I will have problem.
> Proposal:
> Handle UUID in postgres as a SQL extension. Use the DataTypes.RAW to wrap its
> original class info and its serializer so as to expose it as a RAW
> JDBCType(only in postgres) in PostgresCatalog.
> And meanwhile, for PostgresRowConverter, we could use BinaryRawValueData to
> handle those SQL extension for byte serialization/deserialization.
--
This message was sent by Atlassian Jira
(v8.20.7#820007)
