I've recently started using Mockito 3 + Junit 5 + Spring 5 and I'm writing some example tests in order to understand how Mockito works. I have a question about inner calls. So, I have a spring component A in which is injected some DAO component someObjectDAO. The A class:
#Component("aClass")
public class A {
#Autowired
private ObjectDAO someObjectDAO;
public Long countRecords() {
ObjectSearchCriteria search = new ObjectSearchCriteria();
return someObjectDAO.count(search);
}
}
I want to test A's countRecords method. I've mocked and injected someObjectDAO like this:
#ExtendWith(MockitoExtension.class)
#ContextConfiguration("contextConfFileSomewhere")
public class ATest {
#Mock
ObjectDAO someObjectDAOMock;
#InjectMocks
A aComponent;
#Test
void testCount() {
ObjectSearchCriteria search = Mockito.mock(ObjectSearchCriteria.class);
Mockito.when(someObjectDAOMock.count(search)).thenReturn(1L);
Assertion.assertEquals(1L, aComponent.countRecords());
}
}
But this way is incorrect, in fact PotentialStubbingProblem is raised.
org.mockito.exceptions.misusing.PotentialStubbingProblem:
Strict stubbing argument mismatch. Please check:
- this invocation of 'count' method:
someObjectDAO.count(
com.example.java.ObjectSearchCriteria#45cc6b13
);
-> at com.example.java.A.countRecords()
- has following stubbing(s) with different arguments:
1. someObjectDAOMock.count(
Mock for ObjectSearchCriteria, hashCode: 204078646
);
Typically, stubbing argument mismatch indicates user mistake when writing tests.
Mockito fails early so that you can debug potential problem easily.
However, there are legit scenarios when this exception generates false negative signal:
- stubbing the same method multiple times using 'given().will()' or 'when().then()' API
Please use 'will().given()' or 'doReturn().when()' API for stubbing.
- stubbed method is intentionally invoked with different arguments by code under test
Please use default or 'silent' JUnit Rule (equivalent of Strictness.LENIENT).
For more information see javadoc for PotentialStubbingProblem class.
If I understand correctly the exception indicates that I'm passing an object different from the actual object used in the code under test, right?
So, how can I mock a inner method which uses a local variable as parameter?
You want to mock the call to ObjectDAO.count which has parameters. Instead of passing an instance of the expected parameter as an argument, you should use an argument matcher:
Mockito.when(someObjectDAOMock.count(Mockito.any(ObjectSearchCriteria.class)))
.thenReturn(1L);
Edit: You should probably never even want to "mock a local variable". Your goal is to test the system under test (countRecords method) without knowing implementation details. All you can do is mock the dependencies.
My Code is as below,
#RunWith(MockitoJUnitRunner.class)
public class MyClass {
private static final String code ="Test";
#Mock
private MyClassDAO dao;
#InjectMocks
private MyClassService Service = new MyClassServiceImpl();
#Test
public void testDoSearch() throws Exception {
final String METHOD_NAME = logger.getName().concat(".testDoSearchEcRcfInspections()");
CriteriaDTO dto = new CriteriaDTO();
dto.setCode(code);
inspectionService.searchEcRcfInspections(dto);
List<SearchCriteriaDTO> summaryList = new ArrayList<SearchCriteriaDTO>();
inspectionsSummaryList.add(dto);
when(dao.doSearch(dto)).thenReturn(inspectionsSummaryList);//got error in this line
verify(dao).doSearchInspections(dto);
}
}
I am getting below exception
org.mockito.exceptions.misusing.UnnecessaryStubbingException:
Unnecessary stubbings detected in test class: Test
Clean & maintainable test code requires zero unnecessary code.
Following stubbings are unnecessary (click to navigate to relevant line of code):
1. -> at service.Test.testDoSearch(Test.java:72)
Please remove unnecessary stubbings or use 'silent' option. More info: javadoc for UnnecessaryStubbingException class.
at org.mockito.internal.exceptions.Reporter.formatUnncessaryStubbingException(Reporter.java:838)
at org.mockito.internal.junit.UnnecessaryStubbingsReporter.validateUnusedStubs(UnnecessaryStubbingsReporter.java:34)
at org.mockito.internal.runners.StrictRunner.run(StrictRunner.java:49)
at org.mockito.junit.MockitoJUnitRunner.run(MockitoJUnitRunner.java:103)
at org.eclipse.jdt.internal.junit4.runner.JUnit4TestReference.run(JUnit4TestReference.java:86)
at org.eclipse.jdt.internal.junit.runner.TestExecution.run(TestExecution.java:38)
at org.eclipse.jdt.internal.junit.runner.RemoteTestRunner.runTests(RemoteTestRunner.java:459)
at org.eclipse.jdt.internal.junit.runner.RemoteTestRunner.runTests(RemoteTestRunner.java:675)
at org.eclipse.jdt.internal.junit.runner.RemoteTestRunner.run(RemoteTestRunner.java:382)
at org.eclipse.jdt.internal.junit.runner.RemoteTestRunner.main(RemoteTestRunner.java:192)
Please help me how to resolve
At first you should check your test logic. Usually there are 3 cases. First, you are mocking the wrong method (you made a typo or someone changed tested code so that mocked method is no longer used). Second, your test is failing before this method is called. Third, your logic falls in wrong if/switch branch somewhere in the code so that mocked method is not called.
If this is the first case you always want to change the mocked method for the one used in the code. With the second and the third it depends. Usually you should just delete this mock if it has no use. But sometimes there are certain cases in parametrized tests, which should take this different path or fail earlier. Then you can split this test into two or more separate ones but that's not always good looking. 3 test methods with possibly 3 arguments providers can make your test look unreadable. In that case for JUnit 4 you silent this exception with either
#RunWith(MockitoJUnitRunner.Silent.class)
annotation or if you are using rule approach
#Rule
public MockitoRule rule = MockitoJUnit.rule().strictness(Strictness.LENIENT);
or (the same behaviour)
#Rule
public MockitoRule rule = MockitoJUnit.rule().silent();
For JUnit 5 tests you can silence this exception using this annotation provided in mockito-junit-jupiter package:
#ExtendWith(MockitoExtension.class)
#MockitoSettings(strictness = Strictness.LENIENT)
class JUnit5MockitoTest {
}
Replace #RunWith(MockitoJUnitRunner.class) with #RunWith(MockitoJUnitRunner.Silent.class).
For me neither the #Rule nor the #RunWith(MockitoJUnitRunner.Silent.class) suggestions worked. It was a legacy project where we upgraded to mockito-core 2.23.0.
We could get rid of the UnnecessaryStubbingException by using:
Mockito.lenient().when(mockedService.getUserById(any())).thenReturn(new User());
instead of:
when(mockedService.getUserById(any())).thenReturn(new User());
Needless to say that you should rather look at the test code, but we needed to get the stuff compiled and the tests running first of all ;)
Silent is not a solution. You need fix your mock in your test. See official documentation here.
Unnecessary stubs are stubbed method calls that were never realized during test execution (see also MockitoHint), example:
//code under test:
...
String result = translator.translate("one")
...
//test:
...
when(translator.translate("one")).thenReturn("jeden"); // <- stubbing realized during code execution
when(translator.translate("two")).thenReturn("dwa"); // <- stubbing never realized
...
Notice that one of the stubbed methods were never realized in the code under test, during test execution. The stray stubbing might be an oversight of the developer, the artifact of copy-paste or the effect not understanding the test/code. Either way, the developer ends up with unnecessary test code. In order to keep the codebase clean & maintainable it is necessary to remove unnecessary code. Otherwise tests are harder to read and reason about.
To find out more about detecting unused stubbings see MockitoHint.
when(dao.doSearch(dto)).thenReturn(inspectionsSummaryList);//got error in this line
verify(dao).doSearchInspections(dto);
The when here configures your mock to do something. However, you donot use this mock in any way anymore after this line (apart from doing a verify). Mockito warns you that the when line therefore is pointless. Perhaps you made a logic error?
Replace
#RunWith(MockitoJUnitRunner.class)
with
#RunWith(MockitoJUnitRunner.Silent.class)
or remove #RunWith(MockitoJUnitRunner.class)
or just comment out the unwanted mocking calls (shown as unauthorised stubbing).
This was already pointed out in this comment, but I think that's too easy to overlook: You may run into an UnnecessaryStubbingException if you simply convert a JUnit 4 test class to a JUnit 5 test class by replacing an existing #Before with #BeforeEach, and if you perform some stubbing in that setup method that is not realized by at least one of the test cases.
This Mockito thread has more information on that, basically there is a subtle difference in the test execution between #Before and #BeforeEach. With #Before, it was sufficient if any test case realized the stubbings, with #BeforeEach, all cases would have to.
If you don't want to break up the setup of #BeforeEach into many small bits (as the comment cited above rightly points out), there's another option still instead of activating the lenient mode for the whole test class: you can merely make those stubbings in the #BeforeEach method lenient individually using lenient().
As others pointed out it is usually the simplest to remove the line that is unnecessarily stubbing a method call.
In my case it was in a #BeforeEach and it was relevant most of the time. In the only test where that method was not used I reset the mock, e.g.:
myMock.reset()
Hope this helps others with the same problem.
(Note that if there are multiple mocked calls on the same mock this could be inconvenient as well since you'll have to mock all the other methods except the one that isn't called.)
Looking at a part of your stack trace it looks like you are stubbing the dao.doSearch() elsewhere. More like repeatedly creating the stubs of the same method.
Following stubbings are unnecessary (click to navigate to relevant line of code):
1. -> at service.Test.testDoSearch(Test.java:72)
Please remove unnecessary stubbings or use 'silent' option. More info: javadoc for UnnecessaryStubbingException class.
Consider the below Test Class for example:
#RunWith(MockitoJUnitRunner.class)
public class SomeTest {
#Mock
Service1 svc1Mock1;
#Mock
Service2 svc2Mock2;
#InjectMock
TestClass class;
//Assume you have many dependencies and you want to set up all the stubs
//in one place assuming that all your tests need these stubs.
//I know that any initialization code for the test can/should be in a
//#Before method. Lets assume there is another method just to create
//your stubs.
public void setUpRequiredStubs() {
when(svc1Mock1.someMethod(any(), any())).thenReturn(something));
when(svc2Mock2.someOtherMethod(any())).thenReturn(somethingElse);
}
#Test
public void methodUnderTest_StateUnderTest_ExpectedBehavior() {
// You forget that you defined the stub for svcMock1.someMethod or
//thought you could redefine it. Well you cannot. That's going to be
//a problem and would throw your UnnecessaryStubbingException.
when(svc1Mock1.someMethod(any(),any())).thenReturn(anyThing);//ERROR!
setUpRequiredStubs();
}
}
I would rather considering refactoring your tests to stub where necessary.
Well, In my case Mockito error was telling me to call the actual method after the when or whenever stub. Since we were not invoking the conditions that we just mocked, Mockito was reporting that as unnecessary stubs or code.
Here is what it was like when the error was coming :
#Test
fun `should return error when item list is empty for getStockAvailability`() {
doAnswer(
Answer<Void> { invocation ->
val callback =
invocation.arguments[1] as GetStockApiCallback<StockResultViewState.Idle, StockResultViewState.Error>
callback.onApiCallError(stockResultViewStateError)
null
}
).whenever(stockViewModelTest)
.getStockAvailability(listOf(), getStocksApiCallBack)
}
then I just called the actual method mentioned in when statement to mock the method.
changes done is as below
stockViewModelTest.getStockAvailability(listOf(), getStocksApiCallBack)
#Test
fun `should return error when item list is empty for getStockAvailability`() {
doAnswer(
Answer<Void> { invocation ->
val callback =
invocation.arguments[1] as GetStockApiCallback<StockResultViewState.Idle, StockResultViewState.Error>
callback.onApiCallError(stockResultViewStateError)
null
}
).whenever(stockViewModelTest)
.getStockAvailability(listOf(), getStocksApiCallBack)
//called the actual method here
stockViewModelTest.getStockAvailability(listOf(), getStocksApiCallBack)
}
it's working now.
If you're using this style instead:
#Rule
public MockitoRule rule = MockitoJUnit.rule().strictness(Strictness.STRICT_STUBS);
replace it with:
#Rule
public MockitoRule rule = MockitoJUnit.rule().silent();
I had UnnecessaryStubbingException when I tried to use the when methods on a Spy object.
Mockito.lenient() silenced the exception but the test results were not correct.
In case of Spy objects, one has to call the methods directly.
#ExtendWith(MockitoExtension.class)
#RunWith(JUnitPlatform.class)
class ArithmTest {
#Spy
private Arithm arithm;
#Test
void testAddition() {
int res = arithm.add(2, 5);
// doReturn(7).when(arithm).add(2, 5);
assertEquals(res, 7);
}
}
In case of a large project, it's difficult to fix each of these exceptions. At the same time, using Silent is not advised. I have written a script to remove all the unnecessary stubbings given a list of them.
https://gist.github.com/cueo/da1ca49e92679ac49f808c7ef594e75b
We just need to copy-paste the mvn output and write the list of these exceptions using regex and let the script take care of the rest.
If you use any() when mocking, you have to relpace #RunWith(MockitoJUnitRunner.class) with
#RunWith(MockitoJUnitRunner.Silent.class).
I am using eclipse to run some unit tests on a class and I am using Mockito so I don't have to connect to the database. I have used anyString() in other tests which works but it isn't working in this test. If I change it from anyString() to "" the error disappears and the test passes.
My test is:
#Test
public void test_GetUserByUsername_CallsCreateEntityManager_WhenAddUserMethodIsCalled() {
//Arrange
EntityManagerFactory mockEntityManagerFactory = mock(EntityManagerFactory.class);
EntityManager mockEntityManager= mock(EntityManager.class);
UserRepositoryImplementation userRepositoryImplementation = new UserRepositoryImplementation();
userRepositoryImplementation.setEntityManagerFactory(mockEntityManagerFactory);
when(mockEntityManagerFactory.createEntityManager()).thenReturn(mockEntityManager);
//Act
userRepositoryImplementation.getUserByUsername(anyString());
//Assert
verify(mockEntityManagerFactory, times(1)).createEntityManager();
}
Can anyone explain why I am getting the error and what I can do to resolve it?
userRepositoryImplementation.getUserByUsername(anyString());
This is not the right use of anyString().
It can be used for stubbing or for verifying. But not for actual call of method.
From documentation:
Allow flexible verification or stubbing.
If you want a random string when test is running try to use RandomStringUtils or any other similar library.
userRepositoryImplementation.getUserByUsername(RandomStringUtils.random(length));
You can use Matchers, like anyString() for mocking up (stubbing) an object. I.e. inside a when() invocation. Your invocation is an actual invocation:
//Act
userRepositoryImplementation.getUserByUsername(anyString());
So that's right: for testing you have to add some real input, like "", "salala" or null.
I am trying to specific String value in mockito using when() method like
#Mock Individual indProvider;
when(indProvider.asProvider().getProviderId()).thenReturn("795316051750");
But I am getting exception
org.mockito.exceptions.misusing.InvalidUseOfMatchersException:
Invalid use of argument matchers!
0 matchers expected, 1 recorded.
This exception may occur if matchers are combined with raw values:
//incorrect:
someMethod(anyObject(), "raw String");
When using matchers, all arguments have to be provided by matchers.
For example:
//correct:
someMethod(anyObject(), eq("String by matcher"));
For more info see javadoc for Matchers class.
How can I return specific String in such case?
There's probably somewhere in the code a mockito Matcher used outside of stubbed invocation or verified invocation.
If your code base use JUnit I recommend the use of the mockito runner, it will validate the mockito use for each test method, which means any issue can be pinpointed to the test incorrectly written.
#RunWith(MockitoJUnitRunner.class)
public class TheTest {
#Test public void shoudl_perform_something() { ... }
}
Or with a mockito rule (since 1.10.17)
public class TheTest {
#Rule MockitoRule mockitoRule = MockitoJUnit.rule();
#Test public void should_perform_something() { ... }
}
You don't have to use have #Mocks in your test to benefit from this framework validation.
And anyway both are the preferred ways to instantiate mocks compared to MockitoAnnotations.initMocks(...)
Also which version of mockito are you using ? I believe Mockito has improved some messages with the location of the bad code.
Note that this code cannot work properly, because indProvider.asProvider() is not stubbed so mockito will use the default value which is null. You have to do this either way (deep stubs is here for legacy code, a mock returning a mock is usually considered a code smell and it shows the tested code breaks the Law of Demeter) :
#Mock(answer = RETURNS_DEEP_STUBS) Individual indProvider;
// ...
when(indProvider.asProvider().getProviderId()).thenReturn("795316051750");
or as answered by #przemek, it's more verbose and shows the same weakness in the tested code.
#Mock(answer = RETURNS_DEEP_STUBS) Individual indProvider;
#Mock Provider provider;
// ...
when(indProvider.asProvider()).thenReturn(provider);
when(provider.getProviderId()).thenReturn("795316051750");
I advise you to read this book (Growing Object Oriented Software Guided by Tests), it's one of the very best book you can read about writing softwares guided by tests. I find it even more important than infamous Effective Java book.
You should use multiple when() interceptions:
#Mock Individual indProvider;
#Mock Provider provider;
when(indProvider.asProvider()).thenReturn(provider);
when(provider.getProviderId()).thenReturn("795316051750");
You can solve it like this:
Mockito.when(indProvider.asProvider().getProviderId())
.thenAnswer(t -> "795316051750");
I have a strange problem with EasyMock
This is the call I'm making which throws an IllegalStateException : Matcher expected as expected
expect(this.mock.expectedOperation(gt(0l), MyClass.class)).andReturn(createClassObject());
If I replace the above call with
expect(this.mock.expectedOperation(gt(0l), createClass(MyClass.class))).andReturn(createClassObject());
#SuppressWarnings("unchecked")
public static <T> Class<T> createClass(Class<T> clazz)
{
return (Class<T>) EasyMock.anyObject();
}
Most times i don't get an error, but sometimes I do. It stays IllegalStateException : Matcher expected ..
Sometimes I get the IllegalStateException : 2 Matchers expected 1 recorder error for doing this:
MyClass object = createClassObject();
expect(this.mock.expectedOperation(anyLong(), anyLong()).andReturn(object);
public MyClass createClassObject() {
// Actually sets properties and then returns
return new MyClass();
}
But it runs when I do this:
expect(this.mock.expectedOperation(anyLong(), anyLong()).andReturn(createClassObject());
In the above example, sometimes the former runs and latter fails.
Sometimes this fails :
MyClass object = createClassObject();
expect(this.mock.expectedOperation(1, MyClass.class)).andReturn(object);
I have quadruple checked the reset, replay, verify calls. These tests run sometimes and sometimes fail.
If i run my unit test, it randomly fails at least once in one of the above listed situations. Why? How do I get it running?
EDIT : I'm using EasyMock version 3.1 and
MockedClass mock = EasyMock.createMock(MockedClass.class);
Found the problem. We can't use the gt(0) etc methods to pass parameters into the unit being tested.
In another test by my mistake had used :
service.performOperation(1, gt(0l));
service was not a mock, but the unit i was testing.
On checking the documentation I saw that gt(0l) returns 0, which was causing this test to pass by others to fail. (Dunno why?) Since tests are executed randomly, random conditions were failing.
Using the reset(mocks..) at the beginning of the tests did nothing to help.