SEZ9 commented on issue #12589: URL: https://github.com/apache/seatunnel/issues/12589#issuecomment-6159884827
Thanks, @KaustAbhinand — appreciated. To keep the evidence in one place, please post your findings on the PR mentioned above rather than here. The two things this issue is about, and what would be most useful to verify against that PR: 1. Type mapping during table_path discovery: with DuckDB columns declared using the TIMESTAMP_S/MS/NS aliases, confirm the discovered schema reports a timestamp type rather than STRING, and that fractional precision and NULL values round-trip correctly for each alias. 2. Unzoned timestamp values: confirm the read value is not shifted by the JVM offset via the UTC-Calendar getTimestamp fallback. Running the same input with the JVM time zone set to Asia/Shanghai, UTC and America/Los_Angeles, and including the raw DuckDB value alongside the value SeaTunnel emits for each zone, would be very helpful. Including the exact DDL/insert statements and the job config you used will make the result easy for others to reproduce. This issue stays open until that PR is reviewed and merged. <!-- streview-comment:1740 --> -- 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]
