reactivando / use-pubsub-js / 28139058195

100%

master: 80%

push

github

test(pubsub): cover on() cleanup with no signal (mutation-audit finding) (#94)

Ran a one-off Stryker mutation audit (87% score, 224 killed) to find tests that
pass without actually catching bugs. Most survivors are equivalent mutants —
empty-channel pruning (memory-only, unobservable) and React dependency-array
mutations (the known near-unkillable class). One was a real robustness gap:

  on(token, handler, {})            // options present, signal undefined
  // cleanup did `options?.signal?.removeEventListener` — dropping the second
  // `?.` would throw; no test exercised options-without-signal.

Added tests for on() with options-but-no-signal and on() with no options at all
(cleanup must not throw). Mutation-verified: the new test fails if the `?.` is
dropped. The Stryker tooling was used as a one-off audit and not added as a
permanent dependency (kept the lib's devDeps lean); findings recorded in memory.

148 → 150 tests, 100% coverage, lint clean.

Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

32 of 32 branches covered (100.0%)

Branch coverage included in aggregate %.

214 of 214 relevant lines covered (100.0%)

182.48 hits per line