visgl / deck.gl / 24096670908

82%

push

github

chore(test-utils): Separate tape tests from vitest and clean up CI (#10061)

* feat(test-utils): Add vitest entry point with tape backward compat

- Add @deck.gl/test-utils/vitest entry point with vi.spyOn as default spy factory
- Add @deck.gl/test-utils/tape entry point with probe.gl makeSpy default (with deprecation warning)
- Extract SpyFactory abstraction in lifecycle-test.ts for spy provider flexibility
- Add tape-compat and vitest-entry smoke tests to verify both paths work
- Update package.json exports to support both entry points
- Add vitest 4.0.18 as optional peer dependency

Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>

* ci: Add tape-compat and vitest-smoke tests to CI workflow

Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>

* fix(examples): Disable import/namespace rule for vite config

The eslint-plugin-import namespace rule fails parsing vite module internals.

Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>

* fix(test-utils): Re-export from tape.ts for backward compat

The index.ts was exporting directly from lifecycle-test.ts which requires
createSpy, breaking existing tests. Now exports from tape.ts which provides
a default spy factory.

Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>

* fix(test-utils): Restore JSDoc comments and add called to Spy type

- Restore JSDoc comments that were stripped from lifecycle-test.ts
- Add 'called' property to Spy type for TypeScript compatibility with probe.gl spies

Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>

* fix(test-utils): Preserve original file formatting

Restore original lifecycle-test.ts and reapply minimal SpyFactory
abstraction changes to avoid unintended formatting modifications.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* refactor(test-utils): Add resetSpy injection for framework-agnostic spy cleanup

- Remove vitest-specific restoreSpy from core lifecycle-test.ts
- Add ResetSpy type and resetSpy option to TestLayerO... (continued)

3171 of 3773 branches covered (84.04%)

Branch coverage included in aggregate %.

14484 of 17907 relevant lines covered (80.88%)

26616.05 hits per line