Skip to main content

Testing Strategy

The Kingisoovitaja system employs a multi-layered testing strategy to ensure reliability, accuracy, and performance.

Testing Pyramid​

Unit Tests​

Coverage Target: >80%​

Focus areas:

  • Context extraction logic
  • Search query generation
  • Diversity algorithms
  • Validation functions
  • Utility functions

Example:

describe('Context Extraction', () => {
  it('extracts product type from query', () => {
    const context = extractContext('näita raamatuid');
    
    expect(context.productType).toBe('Raamat');
    expect(context.intent).toBe('product_search');
    expect(context.language).toBe('et');
  });
  
  it('handles budget constraints', () => {
    const context = extractContext('kuni 30 eurot');
    
    expect(context.budget?.max).toBe(30);
    expect(context.budget?.min).toBeUndefined();
  });
});

Integration Tests​

Coverage Target: >70%​

Focus areas:

  • API route handling
  • Database queries
  • Pipeline orchestration
  • Context preservation
  • Search flow

Example:

describe('Search Pipeline', () => {
  it('returns products matching context', async () => {
    const context = {
      productType: 'Raamat',
      category: 'Ilukirjandus',
      budget: { max: 30 }
    };
    
    const results = await searchOrchestrator.orchestrate(context);
    
    expect(results.products).toHaveLength(3);
    expect(results.products[0].price).toBeLessThanOrEqual(30);
    expect(results.searchDurationMs).toBeLessThan(1500);
  });
});

End-to-End Tests​

Coverage Target: Critical flows 100%​

Critical flows:

  1. Simple product search
  2. Show more pagination
  3. Budget refinement
  4. Author search
  5. Multi-language switching
  6. Error recovery

Example:

describe('User Flow: Show More', () => {
  it('maintains context across requests', async () => {
    // Turn 1: Initial search
    await userTypes('raamatud');
    await waitFor(() => screen.getAllByTestId('product-card'));
    
    const firstProducts = getProductIds();
    expect(firstProducts).toHaveLength(3);
    
    // Turn 2: Show more
    await userTypes('näita rohkem');
    await waitFor(() => screen.getAllByTestId('product-card'));
    
    const allProducts = getProductIds();
    expect(allProducts).toHaveLength(6);
    
    // Verify no duplicates
    const unique = new Set(allProducts);
    expect(unique.size).toBe(6);
  });
});

Testing Tools​

Frontend​

  • Jest - Unit testing framework
  • React Testing Library - Component testing
  • MSW - API mocking
  • Playwright - E2E testing

Backend​

  • Vitest - Fast unit testing
  • Supertest - API testing
  • Test Containers - Database testing

Test Execution​

Performance Testing​

Load Testing​

describe('Load Tests', () => {
  it('handles 100 concurrent requests', async () => {
    const requests = Array(100).fill(null).map(() =>
      fetch('/api/chat', { method: 'POST', body: testPayload })
    );
    
    const start = Date.now();
    const results = await Promise.all(requests);
    const duration = Date.now() - start;
    
    expect(results.every(r => r.ok)).toBe(true);
    expect(duration).toBeLessThan(5000);  // Parallel processing
  });
});

Benchmark Testing​

describe('Benchmarks', () => {
  it('context extraction completes in <1s', async () => {
    const start = performance.now();
    await extractContext('näita raamatuid');
    const duration = performance.now() - start;
    
    expect(duration).toBeLessThan(1000);
  });
  
  it('diversity selection completes in <20ms', () => {
    const start = performance.now();
    selectDiverseProducts(candidates, 3);
    const duration = performance.now() - start;
    
    expect(duration).toBeLessThan(20);
  });
});

Test Data Management​

Mock Products​

const MOCK_PRODUCTS = {
  books: [
    {
      id: 'test-1',
      title: 'Harry Potter',
      authors: 'J.K. Rowling',
      category: 'Ilukirjandus',
      price: 25.00
    }
  ],
  
  giftCards: [
    {
      id: 'test-gc-1',
      title: 'Kinkekaart 20€',
      category: 'Kinkekaardid',
      price: 20.00
    }
  ]
};

Test Conversations​

const TEST_CONVERSATIONS = {
  simpleSearch: [
    { role: 'user', content: 'raamatud' },
    { role: 'assistant', content: '[Products shown]' }
  ],
  
  showMore: [
    { role: 'user', content: 'raamatud' },
    { role: 'assistant', content: '[Products A,B,C]' },
    { role: 'user', content: 'näita rohkem' },
    { role: 'assistant', content: '[Products D,E,F]' }
  ]
};

Continuous Integration​

# .github/workflows/test.yml
name: Test Pipeline

on: [push, pull_request]

jobs:
  test:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v3
      - uses: actions/setup-node@v3
      
      - name: Install dependencies
        run: npm ci
      
      - name: Run linter
        run: npm run lint
      
      - name: Type check
        run: npm run type-check
      
      - name: Unit tests
        run: npm test
      
      - name: Integration tests
        run: npm run test:integration
      
      - name: E2E tests
        run: npm run test:e2e
      
      - name: Build
        run: npm run build

Quality Gates​

Pre-commit​

  • Linting passes
  • Type checking passes
  • Unit tests pass
  • No TODO/FIXME in committed code

Pre-merge​

  • All tests pass
  • Coverage threshold met
  • Performance benchmarks pass
  • Code review approved

Pre-deploy​

  • E2E tests pass
  • Load tests pass
  • No critical errors in logs
  • Smoke tests on staging pass

Test Coverage​

Current Coverage:

Overall: 82%
─────────────────
Frontend: 85%
  - Components: 90%
  - Hooks: 88%
  - Utils: 92%

Backend: 78%
  - API Routes: 75%
  - Orchestrators: 82%
  - Services: 85%
  - Utils: 90%

Critical Paths: 95%