hussein-awala opened a new pull request, #75898: URL: https://github.com/apache/airflow/pull/75898
Reuse the identical shared lazy-cache implementation and retain `airflow.providers_manager.LazyDictWithCache` as an importable alias. New instances now have `__module__` and a pickle path of `airflow._shared.providers_discovery.providers_discovery`; old pickle globals still resolve through the alias. Net production reduction: 43 lines. Validation: 42 core, 14 shared-library and 30 SDK tests passed, including old/new pickle-global resolution and new-instance round-trip checks. A separate two-case compatibility proof fails on the unchanged base and passes on the change. | Path | Base | Change | Delta | | --- | ---: | ---: | ---: | | Cache hit | 67.34 ns | 66.56 ns | -1.15% | | First resolution | 390.16 ns | 394.81 ns | +1.19% | | Dict control | 28.36 ns | 28.52 ns | +0.56% | Medians are from three alternating base/head pairs. The implementation body is unchanged; small bidirectional timing movements support no speed claim. Required fast/manual hooks passed. Combined validation of 34 cleanup changes at `7f1dcb0d02`: core DB groups had 12,077 passed, 78 skipped/xfail entries and one timeout; nonDB had 5,525 passed, 18 failed, 11 skipped and 2 xfailed. All 19 failed cases passed serial retries; the original broad failures remain recorded and their resource/order cause is unproven. All 116 selected documentation builds succeeded across the broad run and serial retries after initial OOM/Chromium failures. --- ##### Was generative AI tooling used to co-author this PR? - [x] Yes — Codex (GPT-6) Generated-by: Codex (GPT-6) following [the guidelines](https://github.com/apache/airflow/blob/main/contributing-docs/05_pull_requests.rst#gen-ai-assisted-contributions) --- Drafted-by: Codex (GPT-6) (no human review before posting) -- 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]
