Choosing the right test double in JUnit - Practice Problems & Coding Challenges

Challenge - 5 Problems

🧠 Conceptual

intermediate

Understanding Test Doubles: Which to Use?

You want to isolate a class under test by replacing its dependency with a test double that records interactions but does not implement any real logic. Which test double should you use?

A Stub

B Fake

C Spy

D Mock

💡 Hint

Think about which test double lets you verify method calls after the test runs.

A Spy records interactions with the dependency and allows verification of method calls without implementing full logic. Mocks also verify calls but are usually pre-programmed with expectations. Stubs provide canned responses, and Fakes have working implementations.

❓ Predict Output

intermediate

Output of JUnit Test Using a Stub

What is the output of the following JUnit test when run?

JUnit

import static org.junit.jupiter.api.Assertions.assertTrue;
import org.junit.jupiter.api.Test;

public class PaymentProcessorTest {
    interface PaymentGateway {
        boolean charge(double amount);
    }

    static class PaymentGatewayStub implements PaymentGateway {
        public boolean charge(double amount) {
            return true; // always succeeds
        }
    }

    static class PaymentProcessor {
        private PaymentGateway gateway;

        public PaymentProcessor(PaymentGateway gateway) {
            this.gateway = gateway;
        }

        public boolean process(double amount) {
            return gateway.charge(amount);
        }
    }

    @Test
    void testProcessPayment() {
        PaymentGateway stub = new PaymentGatewayStub();
        PaymentProcessor processor = new PaymentProcessor(stub);
        boolean result = processor.process(100.0);
        assertTrue(result);
    }
}

A Test does not compile

B Test fails due to NullPointerException

C Test fails due to assertion error

D Test passes successfully

💡 Hint

The stub always returns true for charge, so process should succeed.

The stub implementation of PaymentGateway always returns true for charge, so the process method returns true, making the assertion pass.

❓ assertion

advanced

Correct Assertion for Mock Verification

Given a mock object mockService in JUnit with Mockito, which assertion correctly verifies that process() was called exactly once?

JUnit

mockService.process();

A verify(mockService, times(1)).process();

B assertEquals(1, mockService.process());

C verify(mockService).process(1);

D assertTrue(mockService.process.calledOnce());

💡 Hint

Mockito uses verify() with times() to check call counts.

Option A uses Mockito's verify method with times(1) to check that process() was called once. Other options are invalid syntax or incorrect usage.

🔧 Debug

advanced

Debugging a Test Failure with a Fake

A test using a fake database implementation fails because data is not saved as expected. Which is the most likely cause?

A The mock was not verified

B The fake does not implement the save method correctly

C The stub returns incorrect data

D The spy did not record interactions

💡 Hint

Fakes have working implementations; if data is not saved, the implementation may be wrong.

Fakes simulate real behavior. If data is not saved, the fake's save method likely has a bug or incomplete implementation.

❓ framework

expert

Choosing the Right Test Double in Complex JUnit Tests

You have a service that depends on a slow external API. You want to test your service's logic without calling the real API, but also want to verify that your service calls the API with correct parameters. Which test double combination is best?

A Use a stub to simulate API responses and a spy to verify method calls

B Use a fake with a working API implementation only

C Use a mock to simulate API and verify calls, no stub needed

D Use a spy only to simulate and verify API calls

💡 Hint

Consider separating behavior simulation and interaction verification.

A stub provides canned responses to simulate the API, while a spy wraps the stub to record and verify calls. This combination isolates logic and verifies interactions effectively.