Testing¶

Running Tests¶

Frontend Unit Tests (Vitest)¶

# Run all frontend unit tests
make test-frontend

# Run with watch mode
cd frontend && npm test

# Run specific test file
cd frontend && npm test src/components/widgets/WidgetRenderer.test.tsx

Frontend tests use Vitest with React Testing Library. All tests should use data-testid attributes from frontend/src/test-ids.ts for element selection (never rely on text content due to i18n).

Backend Tests¶

# Run all backend tests (unit + integration)
make test-backend

# Run with coverage report
make test-coverage

# Run specific test suites
make test-reports     # Report/analytics tests
make test-tags        # Tag system tests
make test-import      # CSV import tests
make test-budgets     # Budget tests

E2E Tests (Playwright)¶

E2E tests require the development servers to be running.

# Install Playwright browsers (first time only)
make test-e2e-install

# Run E2E tests (headless)
make test-e2e

# Run E2E tests with visible browser
make test-e2e-headed

# Run E2E tests in debug mode
make test-e2e-debug

# Run specific E2E test suites
make test-e2e-import  # Import workflow tests
make test-e2e-full    # Full workflow tests

All Tests¶

# Run both unit and E2E tests
make test-all

i18n Translation Tests¶

Translation completeness is verified automatically:

# Run i18n unit tests (checks all locales have matching keys)
cd frontend && npm test src/test/i18n.test.ts

The i18n test ensures: - All keys in en-US.json exist in every other locale - No extra/orphaned keys in non-English locales - All translated strings are actually different from English (catches copy-paste errors)

Frontend Linting¶

make lint-frontend

Test Data¶

Quick Samples¶

The /samples/ directory contains sample CSV files for database seeding:

bofa.csv - Bank of America checking format
amex.csv - American Express format

These are imported when running make db-seed.

Anonymizing Real Data¶

To test with realistic data without exposing sensitive information, use the anonymization tool.

Setup¶

# 1. Put your real bank CSVs in data/raw/ (gitignored)
cp ~/Downloads/amex_statement.csv data/raw/
cp ~/Downloads/bofa_checking.csv data/raw/

# 2. Check what needs processing
make anonymize-status

# 3. Anonymize all new/changed files
make anonymize

# 4. Find scrubbed files in data/anonymized/
ls data/anonymized/

How Anonymization Works¶

Merchants are consistently tokenized: "AMAZON" → "Acme Store" (same fake name everywhere)
Account numbers, card member names, and reference IDs are replaced
Amounts and dates are preserved for realistic testing
A manifest tracks file hashes so unchanged files are skipped on re-runs

Commands¶

make anonymize         # Process new/changed files
make anonymize-status  # Show pending/processed files
make anonymize-force   # Reprocess all files

The anonymized files in data/anonymized/ are safe to commit and share.

Coverage¶

Test coverage is tracked via Codecov.

To generate a local coverage report:

make test-coverage
# Report saved to backend/htmlcov/
open backend/htmlcov/index.html

CI/CD¶

Tests run automatically on every push and pull request via GitHub Actions:

CI Workflow (.github/workflows/ci.yaml): Runs backend tests with coverage
Coverage results are uploaded to Codecov
PRs show coverage diff in comments