davidzollo commented on PR #10099: URL: https://github.com/apache/seatunnel/pull/10099#issuecomment-6394074469
The review fixes and conflict resolution are now on the current branch, synchronized with upstream dev (`d676f91b3d6aaa2cca693816697a211e4d384805`). The bulk-copy executor now snapshots the selected destination columns’ JDBC metadata before closing the query resources, preserves the actual scale, maps input ordinals to destination column names, closes each bulk-copy object, and leaves connection transactions to the existing JDBC writer. SQLExceptions and buffered rows survive the existing reconnect/retry path. Table-only names retain SQL Server default-schema resolution. Microsoft driver types are loaded through the installed driver’s classloader without a new packaged dependency. The opt-in path rejects incompatible XA, upsert, custom-query and PostgreSQL-copy settings before writer creation. Added executor/configuration/builder coverage and updated the existing SQL Server E2E to reconcile all selected values from the actual bulk-copy sink, including an omitted leading destination column with a default. EN/ZH documentation describes the insert-only and at-least-once contract. Scoped JDBC and JDBC E2E part 3 Spotless passed, and an independent maintainer reviewed the final patch. No local functional tests were run; current-head GitHub CI and the real SQL Server E2E still need to pass. The PR remains a draft. -- 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]
