Jaskaran Singh Kukreja created FLINK-39935:
----------------------------------------------
Summary: Allow connectors to override lineage dataset name in
TableLineageDatasetImpl
Key: FLINK-39935
URL: https://issues.apache.org/jira/browse/FLINK-39935
Project: Flink
Issue Type: Improvement
Components: Table SQL / Planner
Reporter: Jaskaran Singh Kukreja
Currently, {{TableLineageDatasetImpl}} always uses
{{contextResolvedTable.getIdentifier().asSummaryString()}} (i.e.,
{{{}catalog-name.db.table{}}}) as the lineage dataset name.
In some cases, connectors can provide a better catalog identifier for example,
Paimon's JDBC metastore exposes a {{catalog-key
(https://paimon.apache.org/docs/1.1/concepts/catalog/#jdbc-catalog)}} which is
more consistent and stable than the logical catalog name registered in Flink.
Connectors should be able to override the dataset name via their
{{LineageVertexProvider}} to use these identifiers.
This is very similar to how today {{TableLineageDatasetImpl}} allows
over-riding namespace, code ref:
https://github.com/apache/flink/blob/master/flink-table/flink-table-planner/src/main/java/org/apache/flink/table/planner/lineage/TableLineageDatasetImpl.java#L46
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
