VladaZakharova commented on PR #75968: URL: https://github.com/apache/airflow/pull/75968#issuecomment-6763938042
> The fix does what the description says in both modes, but defaulting `fail_on_empty_predictions` to `True` changes behaviour for existing tasks, so it needs a changelog note. > > I checked the assumptions against `google-cloud-aiplatform` 1.164.0, the provider's floor: `wait_for_completion()` leaves the final stats on the resource, and the trigger's `to_dict` emits string counters under `completion_stats` and omits the key when stats are unset. CI is green. I did not run this against a real Vertex AI job. > > ### Smaller observations > * Two comments inline: the changelog note (minor) and a log line (nit). > * nit: the new `context` and `mock_ti` mocks in `test_vertex_ai.py` (for example line 2169) have no spec. `.github/instructions/code-review.instructions.md`: "**Flag any `mock.Mock()` or `mock.MagicMock()` without `spec` or `autospec`.** Unspec'd mocks silently accept any attribute access, hiding real bugs." The surrounding tests in that file do the same, so this is a note, not a request. > > > _This review was drafted by an AI-assisted tool and > > confirmed by an Apache Airflow maintainer. The findings > > below are observations, not blockers; an Apache Airflow > > maintainer — a real person — will take the next look at the > > PR. If you think a finding is mis-applied, please reply on > > the PR and a maintainer will weigh in._ > > _More on how Apache Airflow handles maintainer review:_ > > [Contributing guide](https://github.com/apache/airflow/tree/main/contributing-docs). Thanks for checking against 1.164.0. Added the changelog note (your wording) under the Changelog header, moved the success log below the completion-stats check, and gave the mocks in the new tests a spec. The commit message now describes the default change and the opt-out. -- 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]
