You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
These are the modules where the recent data bugs were. For a storage engine they are low. The workflow comment points to coverage-thresholds.toml, but that file does not exist, so the values are hard-coded in the YAML.
Branch coverage runs nightly, but it has no gate.
Ask
Raise the line gates for persistent_facts, btree_v6, wal.rs, packed_pages and index.rs in steps, aiming at 90%.
Add branch-coverage gates for the same modules.
Create coverage-thresholds.toml (or fix the comment) so there is one source of truth.
For each uncovered error path in these modules, either add a test or mark it unreachable with a reason.
Done on v3: #522 made the gates real, and #584 brought the storage files to 90%+ branch coverage (btree.rs 95.7%, persistent_facts.rs 91.7%, index.rs 100%), with floors in coverage-thresholds.toml.
Activity
test(cache): reach the concurrent-miss path deterministically (#396)
test(cache): skip the threaded concurrent-miss test on WASI (#396)
ci(coverage): cache.rs line floor from the CI measurement (#396)
docs: WASI test count from CI (#396)
Merge pull request #522 from project-minigraf/ci/396-coverage-gates
ci(coverage): graph/storage.rs branch floor from the lowest run (#396)
Merge pull request #523 from project-minigraf/ci/396-coverage-gates-v2
release: v2.0.5
Merge pull request #584 from project-minigraf/test/396-branch-coverage
adityamukho commentedon Oct 14, 2026
Done on v3: #522 made the gates real, and #584 brought the storage files to 90%+ branch coverage (btree.rs 95.7%, persistent_facts.rs 91.7%, index.rs 100%), with floors in coverage-thresholds.toml.