Skip to content

JS_TEST_SUITE_REVIEW.md

Latest commit

 

History

History
91 lines (67 loc) · 3.73 KB

JS_TEST_SUITE_REVIEW.md

File metadata and controls

91 lines (67 loc) · 3.73 KB

JS Test Suite and Coverage Review

Date: 2026-02-24

Scope reviewed

  • Package: packages/buckaroo-js-core
  • Unit test runner/configuration: Jest (package.json, jest.config.ts)
  • Browser/integration test inventory: Playwright specs under pw-tests/
  • Coverage source: pnpm test -- --coverage

Current test suite snapshot

  • Jest is configured with testMatch as:
    • !**/*.spec.ts
    • **/*.test.ts
  • Existing test files in src/:
    • OperationsList.test.tsx
    • Operations.test.tsx
    • DFViewerParts/gridUtils.test.ts
    • DFViewerParts/SmartRowCache.test.ts
  • Result: only 2 of 4 unit test files currently match Jest config (the .test.ts files), while the two .test.tsx files are silently excluded.
  • Playwright has 23 *.spec.ts files under pw-tests/, giving broad E2E/integration surface coverage.

Coverage results (from Jest run)

Overall (Jest unit tests):

  • Statements: 54.85%
  • Branches: 29.83%
  • Functions: 37.01%
  • Lines: 53.89%

Notable by-module highlights:

  • Strong:
    • SmartRowCache.ts: ~91.5% statements / ~81.25% branches
    • DFWhole.ts and baked_data/colorMap.ts: 100%
  • Weak / high-risk:
    • DFViewerInfinite.tsx: ~12.66% statements
    • ChartCell.tsx: ~21.81% statements
    • HistogramCell.tsx: ~17.94% statements
    • Styler.tsx: ~11.26% statements
    • Displayer.ts: ~37.31% statements
    • useColorScheme.ts: ~33.33% statements

Key findings

  1. Jest glob likely excludes intended React tests (.test.tsx).

    • This is the highest-value immediate fix because coverage and confidence are currently understated for component behavior already having test files.

  2. Coverage is concentrated in utility/cache logic, not rendering-heavy components.

    • SmartRowCache and gridUtils are tested reasonably well; UI-heavy modules that drive user behavior are mostly untested at unit level.

  3. Branch coverage is the main gap (29.83%).

    • Suggests limited exercise of error paths, conditional rendering, and edge-case prop/state combinations.

  4. E2E inventory is large, but unit coverage is narrow.

    • Many Playwright specs exist; however, they do not replace fast deterministic unit-level tests for view/model boundary logic.

Suggested improvement plan (for follow-up implementation)

Priority 0 (quick wins)

  1. Update Jest testMatch to include .test.tsx (and optionally .test.jsx/.test.js if desired).
  2. Add/enable coverage thresholds in Jest (start pragmatic, raise gradually), e.g.:
    • global lines/stmts ≥ 60
    • branch ≥ 40
  3. Add test:coverage script for CI consistency.

Priority 1 (high-impact module tests)

  1. Add focused RTL tests for:
    • DFViewerInfinite.tsx (loading transitions, pagination/infinite callbacks, empty/error states)
    • HistogramCell.tsx and ChartCell.tsx (data-shape handling, fallback rendering)
    • Styler.tsx (style mapping edge cases)
  2. Add branch-focused tests to Displayer.ts for display-mode switching and malformed payload handling.

Priority 2 (suite structure + maintainability)

  1. Split Jest projects/config by test type (pure unit vs component DOM tests) to keep feedback fast.
  2. Reduce console noise in tests (console.log in core classes) with explicit debug flags/mocks to improve signal.
  3. Map Playwright specs to explicit risk areas (smoke/regression/visual/theme) and run tiers in CI (smoke on PR, full nightly).

Suggested PR checklist for implementation agent

  • Fix Jest testMatch glob(s) and verify .test.tsx suites execute.
  • Add test:coverage script + initial thresholds.
  • Add at least one new test file targeting one low-coverage module (DFViewerInfinite.tsx preferred).
  • Re-run coverage and compare deltas.
  • Document CI strategy for Jest vs Playwright lanes.