Langchain4j Testing Strategies

LangChain4J Testing Strategies

When to Use This Skill

Use this skill when:

• Building AI-powered applications with LangChain4J
• Writing unit tests for AI services and guardrails
• Setting up integration tests with real LLM models
• Creating mock-based tests for faster test execution
• Using Testcontainers for isolated testing environments
• Testing RAG (Retrieval-Augmented Generation) systems
• Validating tool execution and function calling
• Testing streaming responses and async operations
• Setting up end-to-end tests for AI workflows
• Implementing performance and load testing

Instructions

To test LangChain4J applications effectively, follow these key strategies:

1. Start with Unit Testing

Use mock models for fast, isolated testing of business logic. See references/unit-testing.md for detailed examples.

// Example: Mock ChatModel for unit tests
ChatModel mockModel = mock(ChatModel.class);
when(mockModel.generate(any(String.class)))
    .thenReturn(Response.from(AiMessage.from("Mocked response")));

var service = AiServices.builder(AiService.class)
        .chatModel(mockModel)
        .build();

2. Configure Testing Dependencies

Setup proper Maven/Gradle dependencies for testing. See references/testing-dependencies.md for complete configuration.

Key dependencies:

• langchain4j-test - Testing utilities and guardrail assertions
• testcontainers - Integration testing with containerized services
• mockito - Mock external dependencies
• assertj - Better assertions

3. Implement Integration Tests

Test with real services using Testcontainers. See references/integration-testing.md for container setup examples.

@Testcontainers
class OllamaIntegrationTest {
    @Container
    static GenericContainer<?> ollama = new GenericContainer<>(
        DockerImageName.parse("ollama/ollama:latest")
    ).withExposedPorts(11434);

    @Test
    void shouldGenerateResponse() {
        ChatModel model = OllamaChatModel.builder()
                .baseUrl(ollama.getEndpoint())
                .build();
        String response = model.generate("Test query");
        assertNotNull(response);
    }
}

4. Test Advanced Features

For streaming responses, memory management, and complex workflows, refer to references/advanced-testing.md.

5. Apply Testing Workflows

Follow testing pyramid patterns and best practices from references/workflow-patterns.md.

• 70% Unit Tests: Fast, isolated business logic testing
• 20% Integration Tests: Real service interactions
• 10% End-to-End Tests: Complete user workflows

Examples

Basic Unit Test

@Test
void shouldProcessQueryWithMock() {
    ChatModel mockModel = mock(ChatModel.class);
    when(mockModel.generate(any(String.class)))
        .thenReturn(Response.from(AiMessage.from("Test response")));

    var service = AiServices.builder(AiService.class)
            .chatModel(mockModel)
            .build();

    String result = service.chat("What is Java?");
    assertEquals("Test response", result);
}

Integration Test with Testcontainers

@Testcontainers
class RAGIntegrationTest {
    @Container
    static GenericContainer<?> ollama = new GenericContainer<>(
        DockerImageName.parse("ollama/ollama:latest")
    );

    @Test
    void shouldCompleteRAGWorkflow() {
        // Setup models and stores
        var chatModel = OllamaChatModel.builder()
                .baseUrl(ollama.getEndpoint())
                .build();

        var embeddingModel = OllamaEmbeddingModel.builder()
                .baseUrl(ollama.getEndpoint())
                .build();

        var store = new InMemoryEmbeddingStore<>();
        var retriever = EmbeddingStoreContentRetriever.builder()
                .chatModel(chatModel)
                .embeddingStore(store)
                .embeddingModel(embeddingModel)
                .build();

        // Test complete workflow
        var assistant = AiServices.builder(RagAssistant.class)
                .chatLanguageModel(chatModel)
                .contentRetriever(retriever)
                .build();

        String response = assistant.chat("What is Spring Boot?");
        assertNotNull(response);
        assertTrue(response.contains("Spring"));
    }
}

Best Practices

Test Isolation

• Each test must be independent
• Use @BeforeEach and @AfterEach for setup/teardown
• Avoid sharing state between tests

Mock External Dependencies

• Never call real APIs in unit tests
• Use mocks for ChatModel, EmbeddingModel, and external services
• Test error handling scenarios

Performance Considerations

• Unit tests should run in < 50ms
• Integration tests should use container reuse
• Include timeout assertions for slow operations

Quality Assertions

• Test both success and error scenarios
• Validate response coherence and relevance
• Include edge case testing (empty inputs, large payloads)

Reference Documentation

For comprehensive testing guides and API references, see the included reference documents:

• Testing Dependencies - Maven/Gradle configuration and setup
• Unit Testing - Mock models, guardrails, and individual components
• Integration Testing - Testcontainers and real service testing
• Advanced Testing - Streaming, memory, and error handling
• Workflow Patterns - Test pyramid and best practices

Common Patterns

Mock Strategy

// For fast unit tests
ChatModel mockModel = mock(ChatModel.class);
when(mockModel.generate(anyString())).thenReturn(Response.from(AiMessage.from("Mocked")));

// For specific responses
when(mockModel.generate(eq("Hello"))).thenReturn(Response.from(AiMessage.from("Hi")));
when(mockModel.generate(contains("Java"))).thenReturn(Response.from(AiMessage.from("Java response")));

Test Configuration

// Use test-specific profiles
@TestPropertySource(properties = {
    "langchain4j.ollama.base-url=http://localhost:11434"
})
class TestConfig {
    // Test with isolated configuration
}

Assertion Helpers

// Custom assertions for AI responses
assertThat(response).isNotNull().isNotEmpty();
assertThat(response).containsAll(expectedKeywords);
assertThat(response).doesNotContain("error");

Performance Requirements

• Unit Tests: < 50ms per test
• Integration Tests: Use container reuse for faster startup
• Timeout Tests: Include @Timeout for external service calls
• Memory Management: Test conversation window limits and cleanup

Security Considerations

• Never use real API keys in tests
• Mock external API calls completely
• Test prompt injection detection
• Validate output sanitization

Testing Pyramid Implementation

70% Unit Tests
  ├─ Business logic validation
  ├─ Guardrail testing
  ├─ Mock tool execution
  └─ Edge case handling

20% Integration Tests
  ├─ Testcontainers with Ollama
  ├─ Vector store testing
  ├─ RAG workflow validation
  └─ Performance benchmarking

10% End-to-End Tests
  ├─ Complete user journeys
  ├─ Real model interactions
  └─ Performance under load

Related Skills

• spring-boot-test-patterns
• unit-test-service-layer
• unit-test-boundary-conditions

References

• Testing Dependencies
• Unit Testing
• Integration Testing
• Advanced Testing
• Workflow Patterns