SEZ9 commented on issue #13296:
URL: https://github.com/apache/seatunnel/issues/13296#issuecomment-6962819302

   Thanks for the detailed report. The per-DDL result table, the job config and 
the stack trace make this easy to follow.
   
   I have not verified this against the code yet. As the report describes it, 
`CustomAlterTableParserListener` passes column names on as written and 
`AlterTableSchemaEventHandler` then compares them exactly. One possible 
direction is to resolve referenced column names case-insensitively against the 
existing table schema and keep the existing column's name, as you describe 
Debezium doing. It is still open whether that normalisation belongs in the 
MySQL-CDC resolver or in `AlterTableSchemaEventHandler`.
   
   Since you indicated you are willing to submit a PR, it would be very 
welcome. Unit tests covering each of the six DDL cases in your table would be 
valuable. The cases that raise no error matter most:
   
   - `DROP COLUMN Name` and `MODIFY COLUMN NAME VARCHAR(128)`, which are not 
applied.
   - `CHANGE COLUMN name NAME VARCHAR(128)`, where the schema's column name 
diverges from the one in the change records.
   
   <!-- streview-comment:2420 -->


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: [email protected]

For queries about this service, please contact Infrastructure at:
[email protected]

Reply via email to