DanielLeens commented on PR #12079: URL: https://github.com/apache/seatunnel/pull/12079#issuecomment-5881932105
Reposting per-item so nothing depends on comment length/rendering, per your ask. **F3 (scheduling vs. validation ID source)** — one path. `TransformDependencyScheduler.ScheduledTransform`'s private constructor resolves both fields exactly once from `ReadonlyConfig`: `outputId` via `PLUGIN_OUTPUT` (default `DEFAULT_ID`), and `inputIds` via `ConfigParserUtil.getInputIds(readonlyConfig)` (`TransformDependencyScheduler.java:229-231` on `363826c`). `DryRunConnectValidator.validateTransform` never re-resolves these from `ReadonlyConfig` — it reads the already-computed `scheduledTransform.getOutputId()` / `getInputIds()` (`DryRunConnectValidator.java:240,242`). The only other call site of `ConfigParserUtil.getInputIds` left in that file is unrelated: `validateSink` (`:305`), which resolves a sink's own `plugin_input`, not a transform's — sinks aren't part of the transform scheduling graph. So scheduling and validation can't diverge for transforms; there's exactly one computation. -- 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]
