testing-implementation（测试实施）

Testing Implementation Skill

Domain-specific guidance for writing comprehensive tests, test automation, and quality assurance.

When To Use This Skill

Load this Skill when task has tags:

• testing, test, qa, quality, coverage
• unit-test, integration-test, e2e, jest, junit

Validation Commands

Run Tests

# Gradle (Kotlin/Java)
./gradlew test
./gradlew test --tests "*UserServiceTest*"
./gradlew test --tests "UserServiceTest.shouldCreateUser"

# NPM (JavaScript/TypeScript)
npm test
npm test -- --coverage
npm test -- UserService.test.ts
npm test -- -t "should create user"

# Python
pytest
pytest tests/test_user_service.py
pytest tests/test_user_service.py::test_create_user
pytest --cov=src tests/

Check Coverage

# Gradle
./gradlew test jacocoTestReport
# Report: build/reports/jacoco/test/html/index.html

# NPM
npm test -- --coverage
# Report: coverage/lcov-report/index.html

# Python
pytest --cov=src --cov-report=html tests/
# Report: htmlcov/index.html

Success Criteria (Before Completing Task)

✅ ALL tests MUST pass (0 failures) ✅ Coverage goals met (specified in task, typically 80%+) ✅ No flaky tests (run multiple times to verify) ✅ Test execution time acceptable (< 5min for unit tests) ✅ All edge cases covered

Common Testing Tasks

Unit Tests

• Test individual functions/methods in isolation
• Mock external dependencies
• Focus on business logic
• Fast execution (milliseconds)

Integration Tests

• Test components working together
• Real database (in-memory)
• Real services
• Test actual integration points

E2E Tests

• Test full user workflows
• Simulated user interactions
• Real or simulated backend
• Validates end-to-end functionality

Security Tests

• SQL injection attempts
• XSS attacks
• CSRF protection
• Authentication/authorization

Performance Tests

• Response time under load
• Concurrent request handling
• Memory usage
• Query performance

Testing Principles

Test Types and When to Use

Unit Tests (70% of tests):

• Pure functions with no side effects ✅
• Business logic calculations ✅
• Validation logic ✅
• Data transformations ✅
• Mock external dependencies

Integration Tests (20% of tests):

• API endpoints end-to-end ✅
• Database operations ✅
• Service layer with repositories ✅
• Real infrastructure (in-memory DB)

E2E Tests (10% of tests):

• Critical user workflows ✅
• Authentication flows ✅
• Checkout/payment processes ✅
• Slower, more fragile

Arrange-Act-Assert Pattern

@Test
fun `should calculate total with tax`() {
    // Arrange - Set up test data
    val items = listOf(
        Item(price = 10.0),
        Item(price = 20.0)
    )
    val taxRate = 0.1

    // Act - Execute the function being tested
    val total = calculateTotal(items, taxRate)

    // Assert - Verify the result
    assertEquals(33.0, total)
}

Test Edge Cases

@Test
fun `should handle empty list`() {
    val result = calculateTotal(emptyList(), 0.1)
    assertEquals(0.0, result)
}

@Test
fun `should handle zero tax rate`() {
    val items = listOf(Item(price = 10.0))
    val result = calculateTotal(items, 0.0)
    assertEquals(10.0, result)
}

@Test
fun `should handle negative prices`() {
    val items = listOf(Item(price = -10.0))
    assertThrows<IllegalArgumentException> {
        calculateTotal(items, 0.1)
    }
}

Test Error Conditions

@Test
fun `should throw when user not found`() {
    val nonExistentId = UUID.randomUUID()

    assertThrows<NotFoundException> {
        userService.getUserById(nonExistentId)
    }
}

@Test
fun `should return error response for invalid email`() {
    val response = api.createUser(email = "invalid-email")

    assertEquals(400, response.statusCode)
    assertTrue(response.body.contains("Invalid email"))
}

Common Blocker Scenarios

Blocker 1: Implementation Has Bugs

Issue: Tests fail because code being tested has bugs

What to try:

• Debug the implementation code
• Add logging to understand behavior
• Simplify test to isolate issue
• Check if bug is in test or implementation

If blocked: Report to orchestrator - implementation needs fixing by Senior Engineer

Blocker 2: Missing Test Infrastructure

Issue: No test database, mock servers, or fixtures available

What to try:

• Check for existing test setup in codebase
• Look for test configuration files
• Check documentation for test setup
• Use H2 in-memory database for SQLite/PostgreSQL

If blocked: Report to orchestrator - test infrastructure needs provisioning

Blocker 3: Flaky Existing Tests

Issue: Existing tests fail randomly, making new test validation impossible

What to try:

• Isolate new tests in separate test class
• Run only new tests: ./gradlew test --tests "NewTestClass"
• Document flaky test issue

If blocked: Report to orchestrator - flaky tests need fixing first

Blocker 4: Unclear Test Requirements

Issue: Don't know what behavior to test or what's expected

What to try:

• Review requirements section in task
• Check acceptance criteria
• Look at existing similar tests
• Check API documentation

If blocked: Report to orchestrator - need clarification on expected behavior

Blocker 5: Can't Reproduce Bug

Issue: Bug report unclear, can't write test that reproduces issue

What to try:

• Follow reproduction steps exactly
• Check environment differences
• Add logging to understand actual behavior
• Test in different configurations

If blocked: Report to orchestrator - need clearer reproduction steps

Blocker Report Format

⚠️ BLOCKED - Requires Senior Engineer

Issue: [Specific problem - implementation bug, missing fixtures, unclear requirements]

Attempted Fixes:
- [What you tried #1]
- [What you tried #2]
- [Why attempts didn't work]

Root Cause (if known): [Your analysis]

Partial Progress: [What tests you DID complete]

Context for Senior Engineer:
- Test output: [Test failures]
- Code being tested: [File and method]
- Test code: [Your test code]

Requires: [What needs to happen]

Test Patterns

Mock External Dependencies (Unit Tests)

@Test
fun `should fetch user from API`() {
    // Arrange - Mock external API
    val mockApi = mockk<UserApi>()
    every { mockApi.getUser(any()) } returns User(id = "123", name = "John")

    val service = UserService(mockApi)

    // Act
    val user = service.getUserById("123")

    // Assert
    assertEquals("John", user.name)
    verify { mockApi.getUser("123") }
}

Use Real Database (Integration Tests)

@SpringBootTest
@Transactional  // Auto-rollback after each test
class UserRepositoryTest {

    @Autowired
    private lateinit var userRepository: UserRepository

    @Test
    fun `should save and retrieve user`() {
        // Arrange
        val user = User(email = " [email protected] ", name = "Test")

        // Act
        val saved = userRepository.save(user)
        val retrieved = userRepository.findById(saved.id)

        // Assert
        assertNotNull(retrieved)
        assertEquals(" [email protected] ", retrieved?.email)
    }
}

Test Async Operations

test('should fetch data asynchronously', async () => {
  // Arrange
  const api = new UserApi();

  // Act
  const user = await api.getUser('123');

  // Assert
  expect(user.name).toBe('John');
});

Test Error Handling

@Test
fun `should handle network error gracefully`() {
    // Arrange - Mock to throw exception
    val mockApi = mockk<UserApi>()
    every { mockApi.getUser(any()) } throws NetworkException("Connection failed")

    val service = UserService(mockApi)

    // Act & Assert
    assertThrows<ServiceException> {
        service.getUserById("123")
    }
}

Coverage Targets

Good Coverage:

• Business logic: 90%+
• Service layer: 85%+
• Controllers/APIs: 80%+
• Utilities: 90%+

Lower Coverage OK:

• Configuration classes: 50%
• DTOs/Entities: 30%
• Main/startup code: Varies

Focus on:

• Critical paths (authentication, payment)
• Complex business logic
• Edge cases and error handling

What to Test vs What to Skip

✅ DO Test

• Business logic and calculations
• API request/response handling
• Database operations
• Error handling
• Edge cases (null, empty, invalid input)
• Security (injection, XSS, auth)
• State transitions
• Conditional logic

❌ DON'T Test

• Third-party library internals
• Framework code (Spring, React)
• Getters/setters with no logic
• Private methods (test via public interface)
• Configuration files (unless logic)

Common Patterns to Follow

1. Arrange-Act-Assert structure
2. One assertion per test (or closely related assertions)
3. Descriptive test names (shouldCreateUserWhenValidData)
4. Independent tests (no shared state between tests)
5. Fast unit tests (< 1 second each)
6. Test edge cases (null, empty, boundary values)
7. Clean up after tests (transactions, file deletion)

What NOT to Do

❌ Don't skip edge case testing ❌ Don't write tests that depend on order ❌ Don't test implementation details ❌ Don't create flaky tests (timing-dependent) ❌ Don't mock everything (use real infrastructure where appropriate) ❌ Don't skip test cleanup (database, files) ❌ Don't mark task complete with failing tests

Focus Areas

When reading task sections, prioritize:

• requirements - What needs testing
• testing-strategy - Test approach
• acceptance-criteria - Success conditions
• implementation - Code to test

Remember

• Test edge cases - null, empty, invalid, boundary values
• Use real infrastructure - in-memory database for integration tests
• Fast feedback - run tests incrementally during development
• Clear test names - describe what is being tested
• Independent tests - no shared state
• Report blockers promptly - implementation bugs, missing infrastructure
• Coverage goals matter - aim for 80%+ on business logic
• Validation is mandatory - ALL tests must pass before completion

Additional Resources

For deeper patterns and examples, see:

• PATTERNS.md - Advanced testing patterns, mocking strategies (load if needed)
• BLOCKERS.md - Detailed testing-specific blockers (load if stuck)
• examples.md - Complete test examples (load if uncertain)