perf(javascript): switch test environment from jsdom to happy-dom

[austingreco]

What type of PR is this? (check all applicable)

• Refactor
• Feature
• Bug Fix
• Optimization
• Documentation Update

What changed?

• Replaced jsdom with happy-dom as the test DOM environment in vitest config
• Updated one test assertion that relied on jsdom's CSS color normalization (rgb(255, 0, 0) → red)
• Updated ResizeObserver mock comment (happy-dom provides it natively, but we keep the mock to prevent real observation side effects)

Why?
happy-dom is a lighter-weight DOM implementation that provides faster test execution. Local benchmarks:

Metric	jsdom	happy-dom	Change
Total duration	31.46s	27.57s	-12%
Table integration tests	31.19s	23.30s	-25%
Test execution time	51.03s	44.82s	-12%

CI environments (slower hardware) tend to amplify these differences further.

How did you test it?

• Ran full test suite locally: 865 tests, 105 files, all passing
• yarn vitest --run

Potential risks

• happy-dom has minor behavioral differences from jsdom (e.g., doesn't normalize CSS color values to rgb). One test was updated to account for this. Other tests may surface similar minor differences if they rely on jsdom-specific behavior.

Release notes

• N/A (test infrastructure only, no user-facing changes)

Documentation Changes

• N/A

[github-actions]

📊 Combined Coverage Report

Component	Coverage	Baseline	New Code	Status
🐍 Python	81.81%	≥70%	≥90% ✅	✅
🐹 Go	63.6%	≥60%	≥90% ✅	✅
📜 JavaScript	N/A% (no coverage data)	≥78.74%	≥90% ✅	⏭️

---

Coverage Policy

Baseline (Existing Code):

• Python: ≥70% (current coverage)
• Go: ≥60% (current coverage)
• JavaScript: ≥78.74% (current coverage)

New/Changed Code: ≥90% STRICTLY ENFORCED

Long-term Goal: Maintain and improve coverage baselines

---

📝 Coverage Guidelines

• All new code must include tests
• Coverage should not decrease from the base branch
• Use # pragma: no cover sparingly for truly untestable code
• Aim for meaningful tests, not just coverage numbers

🔗 Quick Links

• Python Coverage Details
• Go Coverage Details
• Coverage Artifacts

[github-actions]

JavaScript Coverage Report

Total Coverage: 79.18%

Coverage Policy:

• Baseline (existing code): ≥78.74% (current coverage)
• New/changed code: ≥90% ✅ STRICTLY ENFORCED

Coverage Details

$(cat coverage/coverage-summary.json 2>/dev/null | node -p "const c=require('fs').readFileSync(0,'utf8'); JSON.stringify(JSON.parse(c).total, null, 2)" || echo "Coverage summary not available")

View detailed HTML report in artifacts

[craigmarker]

How are you weighing the cons of adopting the happy-dom environment? My concern is primarily about less realistic/less established. What do you thiink?

[CLAassistant]

All committers have signed the CLA.