SEZ9 commented on PR #12079: URL: https://github.com/apache/seatunnel/pull/12079#issuecomment-5882465438
Thanks for the detailed write-ups on F6, F7 and F8. **F6** — Having `DryRunConnectValidator.scheduleTransforms` delegate straight to `TransformDependencyScheduler.scheduleTransforms` and only translate `JobDefineCheckException` into `ConfigCheckException` is the shape I was hoping for, and the two tests you named (`ConfigParserUtilTest.testSingleExplicitSelfCycleCannotUseLegacyFallback` and `SeaTunnelConfValidateCommandTest.testConnectDryRunRejectsSingleExplicitSelfCycle`) sound like the right coverage. One request: please have the dry-run test also assert on the `self -> self` message text, not just the exception type, so a future wrapper regression that drops the cycle detail is caught. **F7 / F8** — Understood that the `### Transform Dependency Resolution` entry has been in place since the `d796232d2954` round and that the static `getOutputId` helper is gone. I'll verify both against the diff at `363826c6c520` before closing them out. Before I do that pass, two things: 1. Your note refers to "these five", but I only see write-ups for F6–F8 here. Could you post a short status at `363826c6c520` for **F1** (duplicate default output IDs double-decrementing `unresolvedInputCount`), **F2** (terminal `findLast` fallback narrowed to single-transform jobs), **F3** (`getTransformInputIds` vs `getInputIds` divergence), **F4** (multi-transform jobs that previously passed via the implicit last-schema fallback), and **F5** (scheduler duplicated between `DryRunConnectValidator` and `MultipleTableJobConfigParser`)? If F3 and F5 are covered by the shared `TransformDependencyScheduler`, please say so explicitly rather than leaving me to infer it. 2. For F1, please point me to the test that exercises two or more transforms both omitting `plugin_output` alongside a multi-input dependent, so I can confirm the out-of-order scheduling is gone. Once those are in I'll go through the diff and follow up. <!-- streview-comment:1389 --> -- 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]
