bgeng777 opened a new pull request, #29509: URL: https://github.com/apache/flink/pull/29509
## What is the purpose of the change Fixes [FLINK-32040](https://issues.apache.org/jira/browse/FLINK-32040). Calling `with_idleness()` after `with_timestamp_assigner()` discards the Python timestamp assigner. Event-time windows can then fail with `Record has Java Long.MIN_VALUE timestamp`. `with_watermark_alignment()` loses the assigner for the same reason. This change preserves the Python timestamp assigner when either method creates a new watermark strategy. ## Brief change log - Retain the timestamp assigner in `with_idleness()` and both overloads of `with_watermark_alignment()`. - Add tests for assigner preservation, configuring the assigner after enrichment, and independence of the original strategy. - Extend the existing event-time tumbling-window test with idleness detection. The shared test runs in process and thread modes. ## Verifying this change - Before the fix, the new preservation tests failed and the process-mode window test reproduced the `Long.MIN_VALUE` error from the issue. - After the fix, all 8 watermark-strategy tests and the process/thread window tests passed: **10 passed**. - Flake8 and `git diff --check` passed. - Local verification used the current Python sources with existing Flink 2.3-SNAPSHOT Java build artifacts, Python 3.11 and JDK 21. Full `./mvnw clean verify` has not been run; validation against a fresh master build remains for CI. The alignment tests verify strategy construction and preservation of the Python assigner; they do not claim end-to-end source watermark-alignment coverage. ## Does this pull request potentially affect one of the following parts: - Dependencies (does it add or upgrade a dependency): no - The public API, i.e., is any changed class annotated with `@Public(Evolving)`: no; no API signatures or stability annotations change - The serializers: no - The runtime per-record code paths (performance sensitive): no; the change is limited to strategy construction - Anything that affects deployment or recovery: JobManager (and its components), Checkpointing, Kubernetes/Yarn, ZooKeeper: no - The S3 file system connector: no ## Documentation - Does this pull request introduce a new feature? no - If yes, how is the feature documented? not applicable --- ##### Was generative AI tooling used to co-author this PR? - [X] Yes (please specify the tool below) Generated-by: Codex (GPT-6) -- 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]
