SEZ9 commented on PR #12079: URL: https://github.com/apache/seatunnel/pull/12079#issuecomment-5825936668
Thanks for the follow-up on `3df75448c`. I haven't been able to verify the current head myself yet, so rather than closing anything on description alone I'd like to confirm a few points against the diff: 1. **F8 (redundant helpers / unused `getOutputId()`)** — could you point me at the hunk in `3df75448c` where `validateTransform` switches to reading `outputId`/`inputIds` from the scheduled transform, and confirm the old `getTransformInputIds`/`getTransformOutputId` statics are removed? 2. **F7 (docs)** — please link the `incompatible-changes.md` entries (en and zh) added around `d8045271fcf6` so I can check they still describe the current fail-fast behavior. 3. **F5 (duplicated scheduler)** — please point to where both `MultipleTableJobConfigParser#parseTransforms` and `DryRunConnectValidator` call into the shared `TransformDependencyScheduler`, so I can confirm there is no remaining copy. 4. **F1 (duplicate output IDs / `DEFAULT_ID`)** — how does the shared scheduler handle multiple transforms that all omit `plugin_output`, and which test covers it? I want to see that `unresolvedInputCount` can't be double-decremented. 5. **F2 / F4 (legacy `findLast` fallback, multi-transform compatibility)** — a short summary of the current fallback behavior for multi-transform jobs whose last transform omits `plugin_input`, and confirmation the incompatible-changes entry describes exactly that. 6. **F3 (`getTransformInputIds` vs `getInputIds`)** — a one-line confirmation that validation and scheduling now share the same `getInputIds()` path. 7. **F6 (single-transform self-cycle)** — confirm a self-referencing single transform is now rejected in dry-run, matching the runtime parser, and point to the test. Once I can see those in the diff I'm happy to approve. <!-- streview-comment:1301 --> -- 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]
