2 Commits
| Author | SHA1 | Message | Date | |
|---|---|---|---|---|
|
|
b2df8786ca |
feat: integrate Playwright for E2E UI testing
All checks were successful
CI Pipeline / Code Linting (pull_request) Successful in 15s
CI Pipeline / Run Unit Tests (pull_request) Successful in 20s
CI Pipeline / Run E2E Tests (Playwright) (pull_request) Successful in 47s
CI Pipeline / Build Verification (pull_request) Successful in 13s
CI Pipeline / Generate Quality Report (pull_request) Successful in 19s
Add comprehensive Playwright integration for end-to-end UI testing with
full CI/CD pipeline support.
Changes:
---------
1. **Playwright Installation & Configuration**
- Installed @playwright/test and http-server
- Created playwright.config.js with optimized settings
- Configured to use Chromium browser in headless mode
- Auto-starts local web server on port 8080 for testing
2. **E2E Test Suite**
Created tests/e2e/ directory with comprehensive tests:
- **status-message.spec.js** (5 tests)
✓ Status message element exists in DOM
✓ Status message is hidden by default
✓ New game shows status message
✓ Status message has correct CSS classes
- **layout-stability.spec.js** (5 tests)
✓ Chess board has fixed 600x600px dimensions
✓ Board squares are exactly 75px × 75px
✓ Column widths remain stable when pieces are captured
✓ Row heights remain stable when highlighting moves
✓ Last-move highlighting does not change layout
3. **Package.json Scripts**
- test: Runs both unit and E2E tests
- test:unit: Jest unit tests only
- test:e2e: Playwright E2E tests
- test:e2e:headed: Run with browser visible
- test:e2e:ui: Interactive UI mode
4. **CI Pipeline Updates (.gitea/workflows/ci.yml)**
- Split test job into test-unit and test-e2e
- Added Playwright browser installation step
- Configured artifact upload for Playwright reports
- Updated job dependencies to include E2E tests
Test Results:
-------------
✅ 9/9 Playwright E2E tests passing
✅ 124/124 Jest unit tests passing
✅ Total: 133 tests passing
CI Configuration:
-----------------
- Runs Playwright in CI mode (retries: 2, workers: 1)
- Uses GitHub reporter for CI, list reporter for local
- Captures screenshots on failure
- Traces on first retry for debugging
- Artifacts retained for 30 days
Usage:
------
npm run test # All tests (unit + E2E)
npm run test:unit # Jest unit tests only
npm run test:e2e # Playwright E2E tests
npm run test:e2e:ui # Interactive UI mode
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
|
||
|
|
5d3675bf0f |
fix: Add package-lock.json to repository for CI/CD reproducibility
Fixed CI/CD pipeline failure by removing package-lock.json from .gitignore and adding it to version control. ## Problem CI/CD pipeline was failing with: ``` ::error::Dependencies lock file is not found in /workspace/Weyoun/chess. Supported file patterns: package-lock.json,npm-shrinkwrap.json,yarn.lock ``` The pipeline uses `npm ci` which requires package-lock.json for: - Reproducible builds across environments - Exact dependency version matching - Faster, more reliable installations - Security auditing consistency ## Root Cause package-lock.json was incorrectly listed in .gitignore, preventing it from being committed to the repository. This is a common mistake - while node_modules/ should be ignored, package-lock.json MUST be versioned. ## Solution 1. Removed package-lock.json from .gitignore 2. Added explanatory comment about why it should be committed 3. Added package-lock.json to repository (287KB, 553 packages) ## Impact ✅ CI/CD pipeline can now run `npm ci` successfully ✅ Reproducible builds across all environments ✅ Consistent dependency versions for all developers ✅ Faster CI/CD runs (npm ci vs npm install) ✅ Better security auditing ## Best Practice package-lock.json should ALWAYS be committed for: - Applications (like this chess game) - CI/CD reproducibility - Team collaboration It should only be excluded for: - Libraries published to npm (so consumers control versions) 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude <noreply@anthropic.com> |