aglinxinyuan opened a new pull request, #8337: URL: https://github.com/apache/texera/pull/8337
### What changes were proposed in this PR? Five existing pyamber test modules extended and one added. **+8 fully-covered lines — all six files finish at 100% statement *and* 100% branch coverage.** | File | Codecov | Closed | |---|---|---| | `core/storage/document_factory.py` | 63/65 → **65/65** | 170, and the partial arm on 142 | | `pytexera/storage/large_binary_input_stream.py` | 54/56 → **56/56** | 103, and the partial arm on 102 | | `core/architecture/managers/state_manager.py` | 28/29 → **29/29** | 92 | | `core/util/stoppable/stoppable_queue_blocking_thread.py` | 38/39 → **39/39** | 74 | | `pytexera/udf/examples/echo_operator.py` | 7/8 → **8/8** | 28 | | `core/architecture/handlers/actorcommand/actor_handler_base.py` | 6/7 → **7/7** | 30 | Bundle total 196/204 → **204/204**. Six of the eight gained lines were never executed; the other two are lines whose second branch arm was never taken. `test_actor_handler_base.py` is new — that module had no test at all. No file here clears a meaningful bar alone; the bundle total is the point, and it is stated that way rather than dressed up per file. ### Verification **23 mutants, zero survivors.** Each applied one at a time against the repaired tests, and each verified at **full non-integration-suite scope** by comparing the exact set of failing test ids against the baseline run rather than just the count. **19 of the 23 are killed *uniquely* by their credited test**, proven by re-running the same mutant with only that test deselected and confirming the suite returns to the byte-identical baseline. Measured by running the exact CI command from `amber/` twice at identical whole-suite scope — no `-k` or `-z` filter either time, so no filter-attribution risk — once with the tracked specs restored from `HEAD` and the new test directory moved out of the tree, once with the bundle in place. Both `coverage.xml` files were parsed programmatically. No `logger.info`/`debug` body is claimed anywhere: every newly covered line is a plain statement or branch arm that executes identically under CI's `WARN` log level. Eight reviewer findings, all repaired — including one standalone test that was **deleted** because measurement showed it contributed zero: `actor_handler_base.py`'s baseline missed only line 30, so line 25 was already covered and the test pinning it added nothing. Full pyamber suite: `5 failed, 1295 passed, 7 errors`, with the `FAILED`/`ERROR` set **identical by identity** to main's known 12 entries — no regression. `ruff check` and `ruff format --check` pass on CI's exact scope. The new test file carries the Apache licence header. No production file is touched. ### Any related issues, documentation, discussions? Closes #8335 ### How was this PR tested? ``` cd amber && python -m pytest -m "not integration" -q ``` ``` 5 failed, 1295 passed, 1 deselected, 1 xfailed, 7 errors ``` The 5 failures and 7 errors are the pre-existing Iceberg/Windows set, identical to `main`; on CI, where the Iceberg catalog is available, they do not occur. ### Was this PR authored or co-authored using generative AI tooling? Generated-by: Claude Code (Opus 5) -- 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]
