Why Mocking services in tests in Angular? - Purpose & Use Cases

The Big Idea

Discover how mocking services can turn your slow, flaky tests into lightning-fast, reliable checks!

The Scenario

Imagine testing an Angular component that depends on a service fetching data from the internet. You try to run tests, but every test waits for real network calls, making tests slow and flaky.

The Problem

Manually calling real services in tests is slow, unreliable, and can fail if the network or backend is down. It also makes tests hard to run anywhere and slows down development.

The Solution

Mocking services means replacing real services with fake ones that return fixed data instantly. This makes tests fast, reliable, and easy to control.

Before vs After

✗ Before

service.getData().subscribe(data => expect(data).toBeTruthy());

✓ After

mockService.getData.and.returnValue(of(mockData));

What It Enables

Mocking services lets you test components in isolation, ensuring your tests are fast, stable, and focused only on your code.

Real Life Example

When testing a user profile component, you mock the user service to return a fake user instantly instead of waiting for a real server response.

Key Takeaways

Manual service calls in tests cause slow and unreliable tests.

Mocking replaces real services with fake ones for fast, stable tests.

This helps test components independently and confidently.

Practice

1. What is the main purpose of mocking services in Angular tests?

easy

A. To automatically generate service code

B. To speed up the Angular application in production

C. To add new features to the service during testing

D. To replace real services with fake ones for isolated testing

Solution

1. Step 1: Understand the role of mocking

   Mocking replaces real dependencies with controlled fake versions to isolate the component under test.

2. Step 2: Identify the testing benefit

   This isolation helps tests run faster and more reliably without depending on real service behavior.

3. Final Answer:

   To replace real services with fake ones for isolated testing -> Option D

4. Quick Check:

   Mocking = Replace real with fake [OK]

Hint: Mocking means replacing real services with fakes in tests [OK]

Common Mistakes:

• Thinking mocking speeds up production app
• Confusing mocking with adding features
• Believing mocking auto-generates code

2. Which syntax correctly provides a mock service using useClass in Angular test setup?

easy

A. providers: [{ provide: RealService, useClass: MockService }]

B. providers: [{ useClass: RealService, provide: MockService }]

C. providers: [{ provide: MockService, useClass: RealService }]

D. providers: [{ useValue: MockService, provide: RealService }]

Solution

1. Step 1: Recall Angular provider syntax

   Angular expects an object with 'provide' as the token and 'useClass' as the mock class.

2. Step 2: Match correct order and keys

   The correct order is 'provide' first, then 'useClass' with the mock class.

3. Final Answer:

   providers: [{ provide: RealService, useClass: MockService }] -> Option A

4. Quick Check:

   Provide token, then useClass mock [OK]

Hint: Remember: provide token first, then useClass mock class [OK]

Common Mistakes:

• Swapping provide and useClass keys
• Using useValue instead of useClass incorrectly
• Providing mock as token instead of real service

3. Given this Angular test setup, what will component.getData() return?

class MockService {
  fetch() { return 'mocked data'; }
}

TestBed.configureTestingModule({
  providers: [{ provide: RealService, useClass: MockService }]
});

const service = TestBed.inject(RealService);
const component = new MyComponent(service);

component.getData = function() { return this.service.fetch(); };

medium

A. undefined

B. 'real data'

C. 'mocked data'

D. Throws runtime error

Solution

1. Step 1: Identify the injected service

   The test replaces RealService with MockService using useClass, so service is an instance of MockService.

2. Step 2: Trace method call in component

   component.getData calls service.fetch(), which returns 'mocked data' from MockService.

3. Final Answer:

   'mocked data' -> Option C

4. Quick Check:

   MockService fetch() returns 'mocked data' [OK]

Hint: Injected service is mock, so method returns mock's value [OK]

Common Mistakes:

• Assuming real service is used
• Expecting undefined instead of mock return
• Thinking method throws error without real service

4. What is wrong with this Angular test provider setup?

providers: [{ provide: RealService, useValue: MockService }]

medium

A. Missing import for RealService

B. useValue expects an instance, not a class reference

C. useValue cannot be used in providers

D. provide should be MockService, not RealService

Solution

1. Step 1: Understand useValue usage

   useValue expects an actual instance or object, not a class reference.

2. Step 2: Identify the mistake

   MockService is a class, but useValue is given the class itself, not an instance like new MockService().

3. Final Answer:

   useValue expects an instance, not a class reference -> Option B

4. Quick Check:

   useValue needs instance, not class [OK]

Hint: useValue needs instance (new), not class name [OK]

Common Mistakes:

• Passing class instead of instance to useValue
• Confusing provide token with mock class
• Assuming useValue can't be used in providers

5. You want to mock a service method that returns different values on consecutive calls in Angular tests. Which approach correctly achieves this?

hard

A. Create a mock class with a method using a call count variable to return different values

B. Use useValue with a plain object having the method returning a fixed value

C. Use useClass with the real service and override the method in the test

D. Inject the real service and spy on the method without mocking

Solution

1. Step 1: Understand requirement for different returns

   Returning different values on consecutive calls requires state tracking inside the mock method.

2. Step 2: Choose correct mocking approach

   A mock class with a call count variable can track calls and return different values accordingly.

3. Step 3: Evaluate other options

   useValue with fixed return can't vary returns; overriding real service method is complex; spying alone doesn't mock service.

4. Final Answer:

   Create a mock class with a method using a call count variable to return different values -> Option A

5. Quick Check:

   Mock class with state tracks calls for varied returns [OK]

Hint: Use mock class with call count to vary method returns [OK]

Common Mistakes:

• Using fixed return object for varying outputs
• Overriding real service instead of mocking
• Relying only on spies without mocks