aglinxinyuan commented on PR #7061: URL: https://github.com/apache/texera/pull/7061#issuecomment-5125438803
Rebased this onto main and resolved the conflict — no action needed on your end. #6988 landed this morning and added a `test_start_worker_handler.py` at the same path, so this was an add/add collision rather than a real overlap in intent. I took your file as the base: it subsumes what #6988 asserted about the handler and pins four things that one didn't — branch precedence (`is_source` wins even when reader threads exist), the empty marker scope, the `EmptyRequest` checked through `get_one_of` the way `AsyncRpcServer` unwraps it, and the markers landing on the data sub-queue rather than the control one. The one thing I carried over from #6988 is its two non-source cases against a real `StateManager` — `READY`/0 with reader threads and `PAUSED`/1 echoed untouched. Your real-collaborator tests only covered the source path, and by your own argument in the description a mocked `get_state_with_version` can't catch a hoisted snapshot, so it seemed worth having that guard on both paths. `_build_real_handler` takes an optional `input_manager` now to allow it. Also refreshed the title and description, which still described this as the first test for the handler. 23 cases in the file, handlers package at 69 passing, `black`/`ruff` clean locally. -- 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]
