test(operator-trend): remove 2 fully-redundant recovery-tier mirror test files#116
Conversation
…est files Verified via coverage analysis: both files have zero sole-covered production lines across all of src/ — every line they exercise is covered by the golden net + remaining suite. The other 10 mirror files in the family carry unique coverage and are retained.
There was a problem hiding this comment.
💡 Codex Review
https://github.com/saagpatel/GithubRepoAuditor/blob/6062830e5f254cc81ed6f117dfb70aca13908faa/.coverage#L1
Remove the generated coverage database
This commit adds .coverage, which is coverage.py's generated SQLite data file; inspecting its file table shows author-local absolute paths such as /Users/d/Projects/GHRA-mirror/src/.... Keeping this tracked means fresh checkouts inherit stale, machine-specific coverage data, and any coverage-enabled test run will rewrite the tracked binary and leave developers with a dirty worktree unless they delete or revert it.
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
What
Deletes 2 of the 12 recovery-tier mirror test files (~448 lines):
test_operator_trend_closure_forecast_reset_reentry_rebuild_reentry_restore.pytest_operator_trend_closure_forecast_reset_reentry_rebuild_reentry_restore_rerererestore_recovery.pyWhy these 2 (and only these 2)
Coverage analysis of all 12 mirror files: only these 2 have zero sole-covered production lines across all of
src/— every line they exercise is also covered by the golden net + the rest of the suite. The other 10 mirror files carry 59 unique production lines between them and are retained.Verification
uv run pytest -q→ 2535 passed, 2 skippeduv run ruff check src/ tests/→ cleanThis is the safe, verified slice of the mirror-test cleanup; wholesale deletion was rejected because 10/12 files provide unique coverage.