Skip to content

Test coverage

Reports how much of your code your tests run, including only the lines you changed. lumioguard CC never runs your tests. It reads the LCOV report your test tool already writes.

Rule ID Measures
Line coverage coverage.line Share of executable lines the tests ran, per file
Branch coverage coverage.branch Share of branches the tests ran, per file
Changed-line coverage coverage.changed_line The same, for lines added since the comparison commit
Changed-branch coverage coverage.changed_branch The same, for branches on added lines

All four are reported only. They never block.

Turn it on

  1. Make your test tool write an LCOV report:

    Language Tool
    JavaScript and TypeScript Jest, Vitest, c8 or nyc with the lcov reporter
    Python coverage lcov from coverage.py
    Java A JaCoCo report converted to LCOV

  2. Point the configuration at it:

    "coverage": { "lcovFile": "coverage/lcov.info", "required": false }

  3. Run your tests, then run lumioguard-cc check --base main to include changed-line coverage.

How it works

  • Line and branch coverage come from the report's DA and BRDA records, so your test tool decides what counts as executable. Branches marked - were never reached and are left out.
  • Changed-line and changed-branch coverage count only lines added since the Git merge base, plus every line of untracked files. They need --base, because a stored baseline has no source code.
  • A file with nothing to cover gets the status not_applicable and no value. It never shows as 100 %.
  • Report paths are matched relative to the report first, then to the project root.

When the report cannot be used

Diagnostic Cause
coverage.report_unavailable The report file is missing
coverage.report_stale A source file changed after the report was written
coverage.no_sources_mapped No path in the report matches an analyzed file

With "required": false these are warnings. With "required": true they make the result incomplete (exit code 2). Use required only in CI, straight after the tests run.

Note

Coverage shows which code ran during the tests, not whether the tests checked the right things. The staleness check compares file times only.