How do I mock a method of a class' member field? - java

I'm using Java 6 and Mockito 1.8.5. I want to mock a class' member field's method, but I can't figure out how. I have these classes ...
public class CacheService implements CacheCallback {
private final Cache cache;
...
public static CacheService getInstance() {
return INSTANCE;
}
private CacheService() {
cache = new DefaultCacheImpl();
}
public boolean saveNodes(final Map<Long, XmlNode> nodeMap) {
...
cache.saveNodes(nodeMap);
}
...
}
public class DefaultCacheImpl implements Cache {
...
public void saveNodes(Map<Long, XmlNode> xmlNodes) {
dao.updateDB(xmlNodes);
}
...
}
I can't figure out how to mock the "cache" member field's method "saveNodes". I'm mocking the method below, but because there is no setter in the CacheService class for the field, I can't figure out how to inject my mock ..
public class PopulateCacheServiceImpl extends RemoteServiceServlet implements PopulateCacheService {
...
public Boolean initCache() {
boolean ret = false;
try {
setupMocks();
CacheService.getInstance().startCache();
PopulateCache.addTestEntriesToCache();
ret = true;
} catch (Exception e) {
e.printStackTrace(System.err);
ret = false;
} // try
return ret;
} // initCache
private void setupMocks() {
DefaultCacheImpl cache = mock(DefaultCacheImpl.class);
doAnswer(new Answer<Object>() {
public Object answer(InvocationOnMock invocation) throws Throwable {
return null;
}
}).when(cache).saveNodes(Matchers.anyMap());
} // setupMocks
}
Are there any other ways to do this with Mockito? Thanks, - Dave

The problem is in this line:
cache = new DefaultCacheImpl();
If you construct a cache object inside your CacheService, they are tightly coupled. You can not use the CacheService with another cache implementation.
Instead, pass the cache implementation to the constructor:
public CacheService(Cache cacheImpl) {
this.cache = cacheImpl;
}
This allows the CacheService to use any Cache implementation.

What about making two constructors? The one you have would stay there. Another one would let you pass in the Cache implementation and allow you to test the class. The new constructor can have protected access to limit which classes can use it.

If you can change the source, decopule those classes. Get rid of cache = new DefaultCacheImpl(); from constructor as Sjoerd suggested.
If you can't - use PowerMock to mock the constructor of DefaultCacheImpl. I must say that this is really ugly solution (the only uglier is mocking static initialization code).
Note:
Your code is an answer to popular question "Why do I need dependency injection for?". I think people were looking at code like this when they invented DI.

Related

Unable to mock method call response in mockito [duplicate]

I've written a factory to produce java.sql.Connection objects:
public class MySQLDatabaseConnectionFactory implements DatabaseConnectionFactory {
#Override public Connection getConnection() {
try {
return DriverManager.getConnection(...);
} catch (SQLException e) {
throw new RuntimeException(e);
}
}
}
I'd like to validate the parameters passed to DriverManager.getConnection, but I don't know how to mock a static method. I'm using JUnit 4 and Mockito for my test cases. Is there a good way to mock/verify this specific use-case?
Use PowerMockito on top of Mockito.
Example code:
#RunWith(PowerMockRunner.class)
#PrepareForTest(DriverManager.class)
public class Mocker {
#Test
public void shouldVerifyParameters() throws Exception {
//given
PowerMockito.mockStatic(DriverManager.class);
BDDMockito.given(DriverManager.getConnection(...)).willReturn(...);
//when
sut.execute(); // System Under Test (sut)
//then
PowerMockito.verifyStatic();
DriverManager.getConnection(...);
}
More information:
Why doesn't Mockito mock static methods?
Mocking of static methods in Mockito is possible since Mockito 3.4.0.
For more details see:
https://github.com/mockito/mockito/releases/tag/v3.4.0
https://github.com/mockito/mockito/issues/1013
https://javadoc.io/doc/org.mockito/mockito-core/latest/org/mockito/Mockito.html#static_mocks
assertEquals("foo", Foo.method());
try (MockedStatic mocked = mockStatic(Foo.class)) {
mocked.when(Foo::method).thenReturn("bar");
assertEquals("bar", Foo.method());
mocked.verify(Foo::method);
}
assertEquals("foo", Foo.method());
In your case, something like this:
#Test
public void testStaticMockWithVerification() throws SQLException {
try (MockedStatic<DriverManager> dummy = Mockito.mockStatic(DriverManager.class)) {
DatabaseConnectionFactory factory = new MySQLDatabaseConnectionFactory();
dummy.when(() -> DriverManager.getConnection("arg1", "arg2", "arg3"))
.thenReturn(new Connection() {/*...*/});
factory.getConnection();
dummy.verify(() -> DriverManager.getConnection(eq("arg1"), eq("arg2"), eq("arg3")));
}
}
NOTE: mocking STATIC METHODS requires mockito-inline dependency instead of mockito-core.
For JUnit5 also add this:
<dependency>
<groupId>org.mockito</groupId>
<artifactId>mockito-junit-jupiter</artifactId>
<version>${mockito.version}</version>
<scope>test</scope>
</dependency>
The typical strategy for dodging static methods that you have no way of avoiding using, is by creating wrapped objects and using the wrapper objects instead.
The wrapper objects become facades to the real static classes, and you do not test those.
A wrapper object could be something like
public class Slf4jMdcWrapper {
public static final Slf4jMdcWrapper SINGLETON = new Slf4jMdcWrapper();
public String myApisToTheSaticMethodsInSlf4jMdcStaticUtilityClass() {
return MDC.getWhateverIWant();
}
}
Finally, your class under test can use this singleton object by, for example,
having a default constructor for real life use:
public class SomeClassUnderTest {
final Slf4jMdcWrapper myMockableObject;
/** constructor used by CDI or whatever real life use case */
public myClassUnderTestContructor() {
this.myMockableObject = Slf4jMdcWrapper.SINGLETON;
}
/** constructor used in tests*/
myClassUnderTestContructor(Slf4jMdcWrapper myMock) {
this.myMockableObject = myMock;
}
}
And here you have a class that can easily be tested, because you do not directly use a class with static methods.
If you are using CDI and can make use of the #Inject annotation then it is even easier.
Just make your Wrapper bean #ApplicationScoped, get that thing injected as a collaborator (you do not even need messy constructors for testing), and go on with the mocking.
I had a similar issue. The accepted answer did not work for me, until I made the change: #PrepareForTest(TheClassThatContainsStaticMethod.class), according to PowerMock's documentation for mockStatic.
And I don't have to use BDDMockito.
My class:
public class SmokeRouteBuilder {
public static String smokeMessageId() {
try {
return InetAddress.getLocalHost().getHostAddress();
} catch (UnknownHostException e) {
log.error("Exception occurred while fetching localhost address", e);
return UUID.randomUUID().toString();
}
}
}
My test class:
#RunWith(PowerMockRunner.class)
#PrepareForTest(SmokeRouteBuilder.class)
public class SmokeRouteBuilderTest {
#Test
public void testSmokeMessageId_exception() throws UnknownHostException {
UUID id = UUID.randomUUID();
mockStatic(InetAddress.class);
mockStatic(UUID.class);
when(InetAddress.getLocalHost()).thenThrow(UnknownHostException.class);
when(UUID.randomUUID()).thenReturn(id);
assertEquals(id.toString(), SmokeRouteBuilder.smokeMessageId());
}
}
As mentioned before you can not mock static methods with mockito.
If changing your testing framework is not an option you can do the following:
Create an interface for DriverManager, mock this interface, inject it via some kind of dependency injection and verify on that mock.
For those who use JUnit 5, Powermock is not an option. You'll require the following dependencies to successfully mock a static method with just Mockito.
testCompile group: 'org.mockito', name: 'mockito-core', version: '3.6.0'
testCompile group: 'org.mockito', name: 'mockito-junit-jupiter', version: '3.6.0'
testCompile group: 'org.mockito', name: 'mockito-inline', version: '3.6.0'
mockito-junit-jupiter add supports for JUnit 5.
And support for mocking static methods is provided by mockito-inline dependency.
Example:
#Test
void returnUtilTest() {
assertEquals("foo", UtilClass.staticMethod("foo"));
try (MockedStatic<UtilClass> classMock = mockStatic(UtilClass.class)) {
classMock.when(() -> UtilClass.staticMethod("foo")).thenReturn("bar");
assertEquals("bar", UtilClass.staticMethod("foo"));
}
assertEquals("foo", UtilClass.staticMethod("foo"));
}
The try-with-resource block is used to make the static mock remains temporary, so it's mocked only within that scope.
When not using a try block, make sure to close the mock, once you are done with the assertions.
MockedStatic<UtilClass> classMock = mockStatic(UtilClass.class)
classMock.when(() -> UtilClass.staticMethod("foo")).thenReturn("bar");
assertEquals("bar", UtilClass.staticMethod("foo"));
classMock.close();
Mocking void methods:
When mockStatic is called on a class, all the static void methods in that class automatically get mocked to doNothing().
Observation : When you call static method within a static entity, you need to change the class in #PrepareForTest.
For e.g. :
securityAlgo = MessageDigest.getInstance(SECURITY_ALGORITHM);
For the above code if you need to mock MessageDigest class, use
#PrepareForTest(MessageDigest.class)
While if you have something like below :
public class CustomObjectRule {
object = DatatypeConverter.printHexBinary(MessageDigest.getInstance(SECURITY_ALGORITHM)
.digest(message.getBytes(ENCODING)));
}
then, you'd need to prepare the class this code resides in.
#PrepareForTest(CustomObjectRule.class)
And then mock the method :
PowerMockito.mockStatic(MessageDigest.class);
PowerMockito.when(MessageDigest.getInstance(Mockito.anyString()))
.thenThrow(new RuntimeException());
You can do it with a little bit of refactoring:
public class MySQLDatabaseConnectionFactory implements DatabaseConnectionFactory {
#Override public Connection getConnection() {
try {
return _getConnection(...some params...);
} catch (SQLException e) {
throw new RuntimeException(e);
}
}
//method to forward parameters, enabling mocking, extension, etc
Connection _getConnection(...some params...) throws SQLException {
return DriverManager.getConnection(...some params...);
}
}
Then you can extend your class MySQLDatabaseConnectionFactory to return a mocked connection, do assertions on the parameters, etc.
The extended class can reside within the test case, if it's located in the same package (which I encourage you to do)
public class MockedConnectionFactory extends MySQLDatabaseConnectionFactory {
Connection _getConnection(...some params...) throws SQLException {
if (some param != something) throw new InvalidParameterException();
//consider mocking some methods with when(yourMock.something()).thenReturn(value)
return Mockito.mock(Connection.class);
}
}
I also wrote a combination of Mockito and AspectJ: https://github.com/iirekm/varia/tree/develop/ajmock
Your example becomes:
when(() -> DriverManager.getConnection(...)).thenReturn(...);
Mockito cannot capture static methods, but since Mockito 2.14.0 you can simulate it by creating invocation instances of static methods.
Example (extracted from their tests):
public class StaticMockingExperimentTest extends TestBase {
Foo mock = Mockito.mock(Foo.class);
MockHandler handler = Mockito.mockingDetails(mock).getMockHandler();
Method staticMethod;
InvocationFactory.RealMethodBehavior realMethod = new InvocationFactory.RealMethodBehavior() {
#Override
public Object call() throws Throwable {
return null;
}
};
#Before
public void before() throws Throwable {
staticMethod = Foo.class.getDeclaredMethod("staticMethod", String.class);
}
#Test
public void verify_static_method() throws Throwable {
//register staticMethod call on mock
Invocation invocation = Mockito.framework().getInvocationFactory().createInvocation(mock, withSettings().build(Foo.class), staticMethod, realMethod,
"some arg");
handler.handle(invocation);
//verify staticMethod on mock
//Mockito cannot capture static methods so we will simulate this scenario in 3 steps:
//1. Call standard 'verify' method. Internally, it will add verificationMode to the thread local state.
// Effectively, we indicate to Mockito that right now we are about to verify a method call on this mock.
verify(mock);
//2. Create the invocation instance using the new public API
// Mockito cannot capture static methods but we can create an invocation instance of that static invocation
Invocation verification = Mockito.framework().getInvocationFactory().createInvocation(mock, withSettings().build(Foo.class), staticMethod, realMethod,
"some arg");
//3. Make Mockito handle the static method invocation
// Mockito will find verification mode in thread local state and will try verify the invocation
handler.handle(verification);
//verify zero times, method with different argument
verify(mock, times(0));
Invocation differentArg = Mockito.framework().getInvocationFactory().createInvocation(mock, withSettings().build(Foo.class), staticMethod, realMethod,
"different arg");
handler.handle(differentArg);
}
#Test
public void stubbing_static_method() throws Throwable {
//register staticMethod call on mock
Invocation invocation = Mockito.framework().getInvocationFactory().createInvocation(mock, withSettings().build(Foo.class), staticMethod, realMethod,
"foo");
handler.handle(invocation);
//register stubbing
when(null).thenReturn("hey");
//validate stubbed return value
assertEquals("hey", handler.handle(invocation));
assertEquals("hey", handler.handle(invocation));
//default null value is returned if invoked with different argument
Invocation differentArg = Mockito.framework().getInvocationFactory().createInvocation(mock, withSettings().build(Foo.class), staticMethod, realMethod,
"different arg");
assertEquals(null, handler.handle(differentArg));
}
static class Foo {
private final String arg;
public Foo(String arg) {
this.arg = arg;
}
public static String staticMethod(String arg) {
return "";
}
#Override
public String toString() {
return "foo:" + arg;
}
}
}
Their goal is not to directly support static mocking, but to improve its public APIs so that other libraries, like Powermockito, don't have to rely on internal APIs or directly have to duplicate some Mockito code. (source)
Disclaimer: Mockito team thinks that the road to hell is paved with static methods. However, Mockito's job is not to protect your code from static methods. If you don’t like your team doing static mocking, stop using Powermockito in your organization. Mockito needs to evolve as a toolkit with an opinionated vision on how Java tests should be written (e.g. don't mock statics!!!). However, Mockito is not dogmatic. We don't want to block unrecommended use cases like static mocking. It's just not our job.
To mock static method you should use a Powermock look at:
https://github.com/powermock/powermock/wiki/MockStatic.
Mockito doesn't provide this functionality.
You can read nice a article about mockito:
http://refcardz.dzone.com/refcardz/mockito
I found one solution in Mockito. This feature comes with a version only from 3.4.0
https://asolntsev.github.io/en/2020/07/11/mockito-static-methods/
dependency
In your build.gradle replace mockito-core:3.3.3 by mockito-inline:3.4.0:
testImplementation('org.mockito:mockito-inline:3.4.0')
what are we going to mock
class Buddy
{
static String name()
{
return "John";
}
}
Mock the static method
#Test
void lookMomICanMockStaticMethods()
{
assertThat(Buddy.name()).isEqualTo("John");
try (MockedStatic<Buddy> theMock = Mockito.mockStatic(Buddy.class))
{
theMock.when(Buddy::name).thenReturn("Rafael");
assertThat(Buddy.name()).isEqualTo("Rafael");
}
assertThat(Buddy.name()).isEqualTo("John");
}
I think this could help us.
Since that method is static, it already has everything you need to use it, so it defeats the purpose of mocking.
Mocking the static methods is considered to be a bad practice.
If you try to do that, it means there is something wrong with the way you want to perform testing.
Of course you can use PowerMockito or any other framework capable of doing that, but try to rethink your approach.
For example: try to mock/provide the objects, which that static method consumes instead.
When you try to mock the static method, you have to write the test inside the try block. because It's important to note that scoped mocks must be closed by the entity that activates the mock.
try (MockedStatic<Tester> tester = Mockito.mockStatic(Tester.class)) {
tester.when(() -> Tester.testStatic("Testing..")).thenReturn(mock(ReturnObject.class));
//Here you have to write the test cases
}
In the above example, we have to mock the Tester Class testStatic Method with input param as "Testing...". Here, this method will return a ReturnObject class type object. Hence we write mockito when chain like above.
Don't forget to add below dependency in your Gradle/maven
testImplementation 'org.mockito:mockito-inline:4.3.1'
Use JMockit framework. It worked for me. You don't have to write statements for mocking DBConenction.getConnection() method. Just the below code is enough.
#Mock below is mockit.Mock package
Connection jdbcConnection = Mockito.mock(Connection.class);
MockUp<DBConnection> mockUp = new MockUp<DBConnection>() {
DBConnection singleton = new DBConnection();
#Mock
public DBConnection getInstance() {
return singleton;
}
#Mock
public Connection getConnection() {
return jdbcConnection;
}
};
There is an easy solution by using java FunctionalInterface and then add that interface as dependency for the class you are trying to unit test.
For mocking static functions i was able to do it that way:
create a wrapper function in some helper class/object. (using a name variant might be beneficial for keeping things separated and maintainable.)
use this wrapper in your codes. (Yes, codes need to be realized with testing in mind.)
mock the wrapper function.
wrapper code snippet (not really functional, just for illustration)
class myWrapperClass ...
def myWrapperFunction (...) {
return theOriginalFunction (...)
}
of course having multiple such functions accumulated in a single wrapper class might be beneficial in terms of code reuse.
Here I share my mockito MockStatic solution based on an extension as promised in my answer to leokom's solution.
So, why does Mockito choose try-with-resources? Well, simply because they want to keep a tidy ship. That is good programming after all. Try-with-resources allows construction with guaranteed calling of the close method. But in JUnit we already have that in BeforeEach and AfterEach. And one can easily add these for a generic purpose to each test class using an Extension that implements BeforeEachCallback and AfterEachCallback.
So much for the theory. Let's make a static mock for
Instant.now()
I started with an annotation to be able to mark the fields in my test class that I want to use as static mocks.
#Retention(RetentionPolicy.RUNTIME)
#Target(ElementType.FIELD)
public #interface StaticMock {
}
This allows me to create a field in my test class for static mocking that I can easily find in my Extension class.
#StaticMock
private MockedStatic<Instant> staticInstantMock;
I added the Extension I created to my test class. You have two options.
Create an Extension for this purpose and add it to the class next to MockitoExtension, which you also need.
Create an Extension and have it inherit from MockitoExtension. Now you can replace MockitoExtension on your test class.
I used the latter of the two.
#ExtendWith({CompanyMockitoExtension.class})
class MyExtendedTestClass {
Now we need something to be returned for the static when it's called:
#Mock
private Instant now;
staticInstantMock.when(Instant::now).thenReturn(now);
The whole test class:
#ExtendWith({CompanyMockitoExtension.class})
class MyExtendedTestClass {
#StaticMock
private MockedStatic<Instant> staticInstantMock;
#Mock
private Instant now;
#Test
void myTestMethod() {
staticInstantMock.when(Instant::now).thenReturn(now);
assertThat(Instant::now).isSameAs(now); // This would normally happen in the class you are testing...
}
}
Now let's take a look at the Extension class.
import static org.mockito.Mockito.mockStatic;
import java.lang.reflect.Field;
import java.lang.reflect.Method;
import java.util.List;
import org.junit.jupiter.api.extension.ExtensionContext;
import org.mockito.MockedStatic;
import org.mockito.Mockito;
import org.mockito.junit.jupiter.MockitoExtension;
public class CompanyMockitoExtension extends MockitoExtension {
#Override
public void beforeEach(ExtensionContext context) {
super.beforeEach(context); // Don't forget to call the super!!
if (context.getTestInstance().isEmpty()) { // Just to be sure...
return;
}
// Get the unit test instance
Object testSubject = context.getTestInstance().get();
initializeStaticMocks(testSubject);
}
private void initializeStaticMocks(Object testSubject) {
// Find all fields that I want to static mock
List<Field> staticMockFields = ReflectionHelper.getFieldsWithAnnotation(testSubject, StaticMock.class);
staticMockFields.forEach(field -> initializeStaticMock(field, testSubject));
}
private void initializeStaticMock(Field field, Object testSubject) {
// Get the type of the static mock. It is within the generic MockedStatic<> class type.
Class<?> typeForStaticMock = (Class<?>) ReflectionHelper.getTypesForGeneric(field)[0];
try {
// Now set the field with the mockStatic method of Mockito.
field.setAccessible(true);
field.set(testSubject, mockStatic(typeForStaticMock));
} catch (IllegalAccessException e) {
throw new RuntimeException("Failed to instantiate Static Mock with type: " + typeForStaticMock.getName());
}
}
#Override
public void afterEach(ExtensionContext context) {
super.afterEach(context); // Again, do not forget to call the super.
if (context.getTestInstance().isEmpty()) {
return;
}
Object testSubject = context.getTestInstance().get();
closeStaticMocks(testSubject); // Close all static mocks.
}
private void closeStaticMocks(Object testSubject) {
// Again find all fields we annotated
List<Field> staticMockFields = ReflectionHelper.getFieldsWithAnnotation(testSubject, StaticMock.class);
staticMockFields.forEach(field -> closeStaticMock(field, testSubject));
}
private void closeStaticMock(Field field, Object testSubject) {
// Get the instance and simply call close.
MockedStatic<?> mockedStaticInstance = ReflectionHelper.getFieldInstance(field, testSubject, MockedStatic.class);
mockedStaticInstance.close();
}
}
The nice thing about this extension is that you can add additional mocking stuff. I added verification of no more interactions on all mocks in the AfterEach. This is now automatic when we use this extension. I also added similar behavior for construction mocking as for the static mocking.
As you see, I made my own reflection helper class. I know there are some standard reflection helper classes and those might be better. Here is mine for this purpose.
public class ReflectionHelper {
public static List<Field> getFieldsWithAnnotation(
Object testSubject,
Class<? extends Annotation> annotationType
) {
Class<?> testSubjectClass = testSubject.getClass();
return Arrays.stream(testSubjectClass.getDeclaredFields())
.filter(field -> field.isAnnotationPresent(annotationType))
.collect(toUnmodifiableList());
}
public static List<Field> getCollectionFields(Object testSubject) {
Class<?> testSubjectClass = testSubject.getClass();
return Arrays.stream(testSubjectClass.getDeclaredFields())
.filter(field -> Collection.class.isAssignableFrom(field.getType()))
.collect(toUnmodifiableList());
}
#SuppressWarnings("unchecked")
public static <T> T getFieldInstance(Field field, Object testSubject, Class<T> type) {
return (T) getFieldInstance(field, testSubject);
}
public static Object getFieldInstance(Field field, Object testSubject) {
try {
boolean isStatic = isStatic(field.getModifiers());
Object context = isStatic ? null : testSubject;
field.setAccessible(true);
return field.get(context);
} catch (IllegalAccessException e) {
throw new RuntimeException("Failed to get instance of field.");
}
}
public static Type[] getTypesForGeneric(Field field) {
ParameterizedType parameterizedType = (ParameterizedType) field.getGenericType();
return parameterizedType.getActualTypeArguments();
}
}
Refactor it a little bit:
public class MySQLDatabaseConnectionFactory implements DatabaseConnectionFactory {
ConnectionSupplier connectionSupplier = () -> SupplierDriverManager.getConnection();
public void setConnSupplier(ConnectionSupplier supplier) {
this.connectionSupplier = supplier;
}
#Override
public Connection getConnection() {
try {
return connectionSupplier.conn();
} catch (SQLException e) {
throw new RuntimeException(e);
}
}
#FunctionalInterface
interface ConnectionSupplier {
Connection conn();
}
}
Then you can use the mockito:
MySQLDatabaseConnectionFactory.ConnectionSupplier connectionSupplier = mock(MySQLDatabaseConnectionFactory.ConnectionSupplier.class);
when(connectionSupplier.conn()).thenReturn(yourMockObject);
yourConnectionFactory.setConnSupplier(connectionSupplier);

Block access to specifc attribute inside own class

I'm using a public java API to build my multi thread application, but I don't know if some methods of this API are thread-safe [indeed I really think they aren't].
I was looking for the most clean and straight way to implement this safety on my code and decided to use lambda expression.
I'm planning to use something like this:
public class Myclass {
private NotThreadSafe obj;
public Myclass() {
this.obj = new NotThreadSafe();
}
private synchronized void useNonThreadSafeObject(FunctionalInterface r) {
r.exec(this.obj);
}
private void method1() {
this.useNonThreadSafeObject((NotThreadSafe p)->{/*do stuff*/};);
}
private void method2() {
this.obj.someMethod();
}
/* any other methods */
}
I want to make sure that no method in Myclass will call directly to 'obj'
so all methods must use obj as method1 do, avoiding directly call as method2
of course java doens't have a modifier to do it, but I was wondering if was an annotation or other way to check it at compiler time...
----[I wasn't clear at first time, trying to explain more]----
You can simply split the class into two:
class Guard {
private final NotThreadSafe obj;
public Guard() {
this.obj = new NotThreadSafe();
}
synchronized void useNonThreadSafeObject(FunctionalInterface r) {
r.exec(this.obj);
}
}
public class Myclass {
private Guard guard=new Guard();
private void method1() {
guard.useNonThreadSafeObject(p -> {/*do stuff with p*/});
}
private void method2() {
guard.useNonThreadSafeObject(p->{/*do other stuff with p*/});
}
/* any other methods */
}
This is the best isolation you can get as no direct access to obj is possible from MyClass. But note that this is not bullet-proof, it’s still possible to create a custom implementation of FunctionalInterface which will keep the received NotThreadSafe instance and store it somewhere in the heap for accessing it at a later time without the Guard. However, that’s a bigger effort and unlikely to happen accidentally…

How do you create and later access an application level resource?

Edit: I am trying to create a shared database connection pool for all sessions of a web application. A different post said the best way to create a servlet context object was by having the init listener create it. I am however unclear on how to make this object available for use by my servlet.
Another way you could do this is use static initialization:
public class SomeClass {
private static final Object[] CONTENT;
static {
CONTENT = new Object[SomeOtherClass.getContentSize()]; // To show you can access runtime variables
}
}
This will initialize the CONTENT array once the class is loaded using the ClassLoader.
One solution is using a private holder class:
public class SomeClass {
private static class ResourceHolder {
private static final Resource INSTANCE = new Resource();
}
public static Resource getInstance() {
return ResourceHolder.INSTANCE;
}
}
the instance will be initialized when SomeClass.getInstance() is called the first time.
The simplest lazy initialisation is to use an enum with one instance.
enum Singleton {
INSTANCE; // lazy initialised
}
The added problem is you want initialisation values. To handle this you can nest the class.
enum Utility {;
static MyType val;
static OtherType val2;
enum Holder {
INSTANCE;
Holder() {
// uses val and val2
}
}
public static Holder getInstance(MyType val, OtherType val2) {
Utility.val = val;
Utility.val2 = val2;
return Holder.INSTANCE; // only created the first time.
}
}
Note: this is thread safe as static block initialisation is safe.
Something like:
public static abstract class Lazy<T> {
private T t = null;
public synchronized T get() {
if (t == null) {
t = create();
}
return t;
}
protected abstract T create();
}
public static final Lazy<List<String>> lazyList = new Lazy<List<String>>(){
#Override
protected List<String> create() {
return new ArrayList<String>();
}
};
I'll caution you up front, what you're describing has a bit of code smell, and I suspect you'll do better to avoid this pattern entirely. A static resource that depends on external runtime state breaks all sorts of best practices about variable scope.
What you're describing, however, would best be implemented by either a Supplier or a Future, depending on the work involved in successfully constructing the object you need. The difference is somewhat pedantic, but you'd generally use a Future to hold a reference that will take a long time to compute, while a Supplier generally returns quickly. Future also has some nice hook-ins with Java's concurrency utilities, but by the sound of it you don't need that.
You'd use a Supplier like so:
public class GlobalState {
public static final Supplier<LazyData> MY_DATA = Suppliers.memoize(
new Supplier<LazyData>() {
public LazyData get() {
// do whatever you need to construct your object, only gets executed once needed
}
});
...
}
Suppliers.memoize() will cache the result of the first call to the underlying Supplier in a threadsafe way, so simply wrapping the Supplier you define with this call prevents duplicate processing.

Think about Java constructor

In this example:
class A {
public A() {
// pre-init1
// post-init1
}
}
class B extends A {
public B() {
super();
// init2
}
}
I want to let init2 before init1, cuz super() must be occurred at the very beginning, so the only way is to add another init method:
class A {
public A() {
init();
}
protected void init() {
// pre-init1
// post-init1
}
}
class B extends A {
public B() {
super();
}
protected void init() {
// init2
super.init();
}
}
Can I get rid of init() method?
Or, I have to make final fields non-final.
Or, is there any way to let A do post-init1 after init2, but not introduce init() method?
EDIT
Here the code from practice, well I think I need this special init() for the special case,
This is a base support class for Spring JUnit test, for some reasons I can't use the SpringJUnit4Runner from spring-test, so I created my own,
// wire the bean on demand.
public static <T> T selfWire(T bean) {
if (bean == null)
throw new NullPointerException("bean");
ApplicationContext context = buildAnnotationDescribedContext(bean.getClass());
AutowireCapableBeanFactory beanFactory = context.getAutowireCapableBeanFactory();
beanFactory.autowireBean(bean);
if (bean instanceof ApplicationContextAware) {
((ApplicationContextAware) bean).setApplicationContext(context);
}
if (bean instanceof InitializingBean) {
try {
((InitializingBean) bean).afterPropertiesSet();
} catch (RuntimeException e) {
throw e;
} catch (Exception e) {
throw new RuntimeException("Failed to initialize bean", e);
}
}
return bean;
}
#Import(TestContext.class)
public abstract class WiredTestCase
extends Assert
implements InitializingBean {
// ...
public WiredTestCase() {
init();
ApplicationContextBuilder.selfWire(this);
logger.debug("WiredTestCase wired");
}
protected void init() {
}
#Overrdie
public void afterPropertiesSet() {
}
}
#Import({ TestDaoConfig.class })
public class WiredDaoTestCase
extends WiredTestCase {
public WiredDaoTestCase() {
// init... moved to init()
}
protected void init() {
// Collect entity classes from #Using annotation
// and config the session factory.
}
}
#Using(IcsfIdentityUnit.class)
#ImportSamples(R_ACLSamples.class)
public class R_AuthorityTest
extends WiredDaoTestCase {
#Inject
R_Authority authority;
#Inject
ScannedResourceRegistry registry;
#Overrdie
public void afterPropertiesSet() {
// Do a lot of reflection discoveries.
// ...
super.afterPropertiesSet();
}
#Test
public void testXxx() { ... }
// ...
}
The code is very long, but the idea is simple, in R_AuthorityTest there are DAO beans to be injected, which depends on SessionFactory, and the session factory is configured in WiredDaoTestCase which is the base class of R_AuthorityTest. Despite of final fields, I have to initialize the session factory before WiredTestCase(). I can't initialize them just in static constructor, because I build the persistence unit on the fly from annotations on this.getClass(). So, personally, I think sometime it's reasonable to do some pre-init before super constructor, and maybe init method is the only way in this case?
Even in the second example, you'd need to make B.init() call super.init(), otherwise the logic of init1 wouldn't be executed at all.
I would try not to use this init approach - typically calling virtual methods in a constructor is a really bad idea. You haven't really explained why you need init2 to occur before init1 though... I suspect there's a better design available, but it's hard to suggest a way forward as we don't know what you're trying to do. For example, giving your superclass constructor take some parameters may well be the way forward - but we can't really say at the moment.
If you could give a more representative example (including the final fields you mention later) we could probably help you more.
A superclass constructor should be called first - you cannot have any statements before it, which makes sense, since it is necessary to instantiate the object before initializing it.
Using a separate method as you do is an acceptable work-around for this issue, if you cannot eliminate the need for it by redesigning your application.
You mentioned final members, so I suggest you want to initialize them in a special order...
But if we can't see the exact problem we can't give you fair answer.
Anyway, I just would like to point out that final member can get assigned only at two places (as far as I know).
at the place where you declare them,
or in a contructor context.
Any other attempt to assign a value to a final member will be compiler error.
I really would like to understand the origin of your questiuon and help. Could you provide more details?

How can I implement an abstract singleton class in Java?

Here is my sample abstract singleton class:
public abstract class A {
protected static A instance;
public static A getInstance() {
return instance;
}
//...rest of my abstract methods...
}
And here is the concrete implementation:
public class B extends A {
private B() { }
static {
instance = new B();
}
//...implementations of my abstract methods...
}
Unfortunately I can't get the static code in class B to execute, so the instance variable never gets set. I have tried this:
Class c = B.class;
A.getInstance() - returns null;
and this
ClassLoader.getSystemClassLoader().loadClass("B");
A.getInstance() - return null;
Running both these in the eclipse debugger the static code never gets executed. The only way I could find to get the static code executed is to change the accessibility on B's constructor to public, and to call it.
I'm using sun-java6-jre on Ubuntu 32bit to run these tests.
Abstract Singleton? Doesn't sound viable to me. The Singleton pattern requires a private constructor and this already makes subclassing impossible. You'll need to rethink your design. The Abstract Factory pattern may be more suitable for the particular purpose.
You are trying to get an abstract class play two very different roles:
the abstract factory role for a
(singleton) service that can have
multiple substitutable
implementations,
the service
interface role,
and on top of that you also want the service to be singleton and enforce 'singletoness' on the entire family of classes, for some reason you aren't considering caching the service instances.
Somebody (I would) will say it smells very bad, for multiple reasons
it violates separation of concerns,
singletons make unit testing impossible",
etc.
Somebody else will say it's ok-ish, it doesn't need a lot of different infrastructure and has kind of fluent-ish interface that you see in some very common third party (legacy) Java API.
The bad part is demanding the children to select what implementation should the parent factory method return.
That responsibility should be pushed up and centralised into the abstract superclass. Otherwise you are mixing together patterns that are used in very different contexts, Abstract Factory (parent decide what family of classes clients are going to get) and Factory Method (children factories select what the clients will get).
Factory Method is also not practically possible because you can't override static methods, nor constructors.
There are some (ugly) ways to achieve your objective though:
public abstract class A{
public static A getInstance(...){
if (...)
return B.getInstance();
return C.getInstance();
}
public abstract void doSomething();
public abstract void doSomethingElse();
}
public class B extends A{
private static B instance=new B();
private B(){
}
public static B getInstance(){
return instance;
}
public void doSomething(){
...
}
...
}
//do similarly for class C
The parent could also use reflection, cache instances, etc.
A more test and extension friendly solution is simply standard separation of concerns. The children aren't going to be singleton anymore per se, but you package them into some internal package that you will document as "private" and a public abstract parent in an external package will handle caching or pooling of children instances, enforcing whatever instantiation policy is required on these classes.
A.getInstance() will never call a derived instance since it's statically bound.
I would separate the creation of the object from the actual object itself and create an appropriate factory returning a particular class type. It's not clear how you'd parameterise that, given your example code - is it parameterised via some argument, or is the class selection static ?
You may want to rethink the singleton, btw. It's a common antipattern and makes testing (in particular) a pain, since classes under test will provide their own instance of that class as a singleton. You can't provide a dummy implementation nor (easily) create a new instance for each test.
In addition to problems others have pointed out, having the instance field in A means that you can only have one singleton in the entire VM. If you also have:
public class C extends A {
private C() { }
static {
instance = new C();
}
//...implementations of my abstract methods...
}
... then whichever of B or C gets loaded last will win, and the other's singleton instance will be lost.
This is just a bad way to do things.
Singletons are kind of yucky. Abstract insists on inheritance which you more often than not want to avoid if possible. Overall I'd rethink if what you are trying to do is the simplest possible way, and if so, then be sure to use a factory and not a singleton (singletons are notoriously hard to substitute in unit tests whereas factories can be told to substitute test instances easily).
Once you start looking into implementing it as a factory the abstract thing will sort itself out (either it will clearly be necessary or it may factor out easily in place of an interface).
I found a better way to use Singleton in abstract class, which use a static Map to maintain the instance of subclass.
public abstract class AbstractSingleton {
private static Map<String, AbstractSingleton> registryMap = new HashMap<String, AbstractSingleton>();
AbstractSingleton() throws SingletonException {
String clazzName = this.getClass().getName();
if (registryMap.containsKey(clazzName)) {
throw new SingletonException("Cannot construct instance for class " + clazzName + ", since an instance already exists!");
} else {
synchronized (registryMap) {
if (registryMap.containsKey(clazzName)) {
throw new SingletonException("Cannot construct instance for class " + clazzName + ", since an instance already exists!");
} else {
registryMap.put(clazzName, this);
}
}
}
}
#SuppressWarnings("unchecked")
public static <T extends AbstractSingleton> T getInstance(final Class<T> clazz) throws InstantiationException, IllegalAccessException {
String clazzName = clazz.getName();
if (!registryMap.containsKey(clazzName)) {
synchronized (registryMap) {
if (!registryMap.containsKey(clazzName)) {
T instance = clazz.newInstance();
return instance;
}
}
}
return (T) registryMap.get(clazzName);
}
public static AbstractSingleton getInstance(final String clazzName)
throws ClassNotFoundException, InstantiationException, IllegalAccessException {
if (!registryMap.containsKey(clazzName)) {
Class<? extends AbstractSingleton> clazz = Class.forName(clazzName).asSubclass(AbstractSingleton.class);
synchronized (registryMap) {
if (!registryMap.containsKey(clazzName)) {
AbstractSingleton instance = clazz.newInstance();
return instance;
}
}
}
return registryMap.get(clazzName);
}
#SuppressWarnings("unchecked")
public static <T extends AbstractSingleton> T getInstance(final Class<T> clazz, Class<?>[] parameterTypes, Object[] initargs)
throws SecurityException, NoSuchMethodException, IllegalArgumentException,
InvocationTargetException, InstantiationException, IllegalAccessException {
String clazzName = clazz.getName();
if (!registryMap.containsKey(clazzName)) {
synchronized (registryMap) {
if (!registryMap.containsKey(clazzName)) {
Constructor<T> constructor = clazz.getConstructor(parameterTypes);
T instance = constructor.newInstance(initargs);
return instance;
}
}
}
return (T) registryMap.get(clazzName);
}
static class SingletonException extends Exception {
private static final long serialVersionUID = -8633183690442262445L;
private SingletonException(String message) {
super(message);
}
}
}
From: https://www.cnblogs.com/wang9192/p/3975748.html

Categories