Phase 1: Setup Testcontainers Infrastructure for Vector Search Tests

[robfrank]

Overview

Create Testcontainers infrastructure to replace Mockito mocks with real production code for better test coverage and lower maintenance.

Priority: HIGH
Estimated Time: Week 1

Context

After analyzing both approaches (see PHASE1_COMPARISON.md), we chose Testcontainers + Real Production Code instead of in-memory stubs because:

• ✅ Lower implementation cost (5h vs 10h)
• ✅ Better bug detection (catches SQL errors, schema issues, vector index config)
• ✅ Less code to maintain (~50 lines vs ~200 lines)
• ✅ Reuses production code (ContentPersistenceAdapter, ArcadeContentRepository)

Tasks

1.1 Create ArcadeDB Testcontainer

File: src/test/java/it/robfrank/linklift/testcontainers/ArcadeDbContainer.java

Create a reusable Testcontainer for ArcadeDB that handles:

• Database lifecycle management
• Schema initialization (Content vertex type, vector index)
• Wait strategy for container readiness
• Helper methods for database creation and cleanup

Key Features:

• Vector index: LSM_TREE METRIC COSINE DIMENSIONS 384
• Exposes methods: getHttpUrl(), getRootPassword(), createDatabase(), cleanDatabase()

1.2 Create Base Test Class

File: src/test/java/it/robfrank/linklift/testcontainers/ArcadeDbTestBase.java

Abstract base class that:

• Manages shared Testcontainer lifecycle (@Container static field)
• Sets up real ContentPersistenceAdapter and ArcadeContentRepository in @BeforeEach
• Cleans database between tests in @AfterEach
• Provides repository field for test classes

1.3 Create FakeEmbeddingGenerator

File: src/test/java/it/robfrank/linklift/adapter/out/ai/FakeEmbeddingGenerator.java

Deterministic fake implementation to replace mocked EmbeddingGenerator:

• Generates consistent 384-dimensional embeddings based on text hash
• Provides test helpers: throwOnNextCall(), clearCache(), getCacheSize()
• No randomness - same text always produces same embedding

1.4 Update Dependencies

File: pom.xml

Add dependencies:

<dependency>
    <groupId>org.testcontainers</groupId>
    <artifactId>testcontainers</artifactId>
    <version>1.19.3</version>
    <scope>test</scope>
</dependency>
<dependency>
    <groupId>org.testcontainers</groupId>
    <artifactId>junit-jupiter</artifactId>
    <version>1.19.3</version>
    <scope>test</scope>
</dependency>
<dependency>
    <groupId>org.awaitility</groupId>
    <artifactId>awaitility</artifactId>
    <version>4.2.0</version>
    <scope>test</scope>
</dependency>

1.5 Verify Infrastructure

• Container starts successfully
• Schema initializes correctly
• Vector index is created with correct configuration
• Database cleanup works between tests
• Create test resource directory for WireMock (future use)

Success Criteria

• ArcadeDbContainer starts and initializes schema
• ArcadeDbTestBase provides working repository field
• FakeEmbeddingGenerator produces deterministic embeddings
• All dependencies added to pom.xml
• Infrastructure verified with simple test

Related

• See PHASE1_COMPARISON.md for detailed analysis
• See TEST_REFACTORING_PLAN.md for full plan
• Blocks: Phase 2 (Refactor Service Tests)