 Daniel Volz
|
98939877db
|
feat: expand Playwright E2E coverage (#155)
* feat: comprehensive Playwright E2E test rewrite
Rewrite all E2E tests with correct CSS selectors, add new spec files,
and implement robust auth handling to work within backend rate limits.
Changes:
- Rewrite fixtures/index.ts with JWT-based /auth/me mock to avoid
10 req/min rate limit on /auth/me during test runs
- Rewrite auth.setup.ts with offline JWT validity check to reuse
existing auth state across runs (saves login rate-limit budget)
- Rewrite auth.spec.ts (6 tests) - login page, fields, submit,
redirect guard, invalid credentials, login/register toggle
- Rewrite dashboard.spec.ts (8 tests) - header, nav tabs,
navigation, overview/schedules sections, days selector, redirect
- Rewrite medications.spec.ts (8 tests) - form fields, stock
inventory, package type toggle, intake schedule, save/cancel,
unsaved changes guard
- Rewrite settings.spec.ts (12 tests) - language, notification
matrix, thresholds, calculation mode, toggle switch, export/import,
user menu navigation
- Create planner.spec.ts (9 tests) - form, date inputs, calculate,
reset, checkbox, submit, tab state, eyebrow heading
- Create schedule.spec.ts (12 tests) - timeline, days selector,
past/future toggles, day blocks, today highlight, collapse/expand,
overview table, share button
- Update playwright.config.ts: remove mobile projects, enable
webServer section for CI
- Add .github/workflows/e2e.yml CI workflow for Playwright tests
Total: 57 E2E tests across 6 spec files, all passing consistently
across 5+ consecutive runs without backend restart.
Closes #154
* feat: add comprehensive E2E data tests with medication CRUD, dashboard, planner, schedule
Add 48 new Playwright E2E tests covering real medication data scenarios:
- medication-crud: 14 tests for create/edit/delete/list via UI form
- dashboard-data: 13 tests for overview table, timeline, dose tracking
- planner-data: 9 tests for demand calculator with results/status chips
- schedule-data: 11 tests for timeline, collapse/expand, dose mark/undo
Infrastructure improvements:
- Add API helpers (createMedicationViaAPI, deleteMedicationViaAPI,
deleteAllMedicationsViaAPI) with retry logic for rate-limit resilience
- Configure chromium-data project for serial execution with retry:1
- Add /auth/me mock to avoid rate-limit exhaustion on auth endpoint
- Increase navigateTo reliability with networkidle waits
- Increase auth token validity threshold from 2 to 10 minutes
- Make backend rate limit configurable via RATE_LIMIT_MAX env var
- Set RATE_LIMIT_MAX=300 in dev docker-compose for E2E test support
Total suite: 57 empty-state + 48 data tests = 105 tests (chromium)
* test: add E2E tests for medication editing, stock status, and share schedule
- medication-edit.spec.ts: 10 tests covering generic name, notes,
taken-by add/remove, expiry date, refill, intake schedule editing,
adding intake rows, reminder toggle, and package type changes
- stock-status.spec.ts: 12 tests verifying dashboard shows correct
status chips (High/Normal/Warning/Danger) for different stock levels,
overview table, reorder card, detail modal, and planner integration
- share-schedule.spec.ts: 10 tests for taken-by badges, share button,
share dialog, link generation, shared schedule page navigation,
dose tracking on shared page, and notes display
- fixtures/index.ts: add createShareTokenViaAPI, updateSettingsViaAPI
helpers; expand createMedicationViaAPI with takenBy, notes, expiryDate
- playwright.config.ts: update testMatch/testIgnore for new test files
- docker-compose.dev.yml: increase RATE_LIMIT_MAX to 1000 for E2E tests
* docs: refine release-manager instructions for CLI safety and commit-linked release notes
* fix: resolve PR155 CI failures for frontend lint and e2e proxy
* fix: stabilize auth-related e2e checks in CI
|
2026-02-12 20:06:11 +01:00 |
|
dependabot[bot]