android-unit-testing

Android Unit Testing Excellence 🧪🛡️

The ultimate guide to writing robust, reliable, and maintainable unit tests for Android applications using modern tools like MockK, Turbine, and Coroutine Test.

⚡ When to Use

• Business Logic Verification: Testing Use Cases or ViewModels.
• Data Layer Testing: Testing Repositories or Data Sources (with Fakes).
• Reactive Stream Testing: Verifying emitted values from Flow and StateFlow.
• Logic Refactoring: Ensuring no regressions during code changes.
• TDD (Test-Driven Development): Writing tests before implementation.

🏗️ Core Tools

1. JUnit 5 (Modern Standard)

• Use @Test, @BeforeEach, and @DisplayName for clear tests.
• Rule: Prefer JUnit 5 for new projects, but JUnit 4 is common in legacy.

2. MockK (Power Mocking)

• Rule: Use MockK for Kotlin-first mocking.
• Shorthand: every { ... } returns ... and coEvery { ... } returns ... for suspend functions.

  val repository = mockk<UserRepository>()
  coEvery { repository.getUser("123") } returns User("John")
  

3. Turbine (Flow Testing)

• Rule: Never use collect or first() in tests manually. Use Turbine.

  flow.test {
      assertEquals(1, awaitItem())
      assertEquals(2, awaitItem())
      awaitComplete()
  }
  

🚀 Coroutine Testing (Essentials)

Use runTest

• runTest skips delay() and ensures tests run fast and predictably.

Inject Dispatchers

• Rule: NEVER hardcode Dispatchers.IO or Main. Inject them.
• In Tests: Use StandardTestDispatcher or UnconfinedTestDispatcher.

🧪 Testing Patterns

1. ViewModels

• Standard: Test the Initial State → Trigger Action → Verify Final State and/or Side Effects.
• Note: Ensure Dispatchers.setMain(testDispatcher) is called before and Dispatchers.resetMain() after tests.

2. Repositories

• Standard: Test that data is correctly mapped and error handling is resilient.
• Fakes: Use FakeLocalSource or FakeRemoteSource for complex integration tests.