I don't understand the below behavior:
I have a method:
#Transactional
public void processRejection(final Path path) {
try {
//some code here
} catch (final Exception e) {
this.handleException(e));
}
}
which calls the below which does a saves an entity which doesn't yet exists in the database:
void handleException(final Throwable e) {
this.filesMonitoringJpaManager.save(someEntityHere);
throw new Exception(...)
}
Now the strange is when I comment the throw new Exception(...) the save works, but when I uncomment throw new Exception(...) then the save doesn't work and I have no clue why?
What strange behavior is that from JPA or Hibernate? Is it something about Java Exception mechanism which I don't understand?
#Transactional is meant to roll back when something goes wrong (an exception is thrown). You're saving an entity in the catch block, but you're rethrowing an exception causing the transactional method to roll back.
But you can specify an exception, that will not cause rollback:
#Transactional(noRollbackFor = {MyException.class})
public void processRejection(final Path path) {
try {
//somecode here whatever
} catch (final Exception e) {
this.handleException(e));
}
}
void handleException(final Throwable e) {
this.filesMonitoringJpaManager.save(someEntityHere);
throw new MyException(...)
}
This works for org.springframework.transaction.annotation.Transactional. If you're using javax.transaction.Transactional, then you can achieve it by using dontRollbackOn property.
I always tested my exceptions with annotations.
#Test (expected = Exception.class)
public void test1() {
methodToTest() // throws an exception
}
I finally switched to Java 8, and I came across lambda Expressions. Now there is another option to get the desired result.
#Test
public void test2() {
assertThrows(Exception.class, () -> methodToTest());
}
public static <X extends Throwable> Throwable assertThrows(
final Class<X> exceptionClass, final Runnable block) {
try {
block.run();
} catch(Throwable ex) {
if (exceptionClass.isInstance(ex))
return ex;
}
fail("Failed to throw expected exception");
return null;
}
I understand that with the second version you can check for single methods more precisely, and you don't have to worry about other methods within a single test that could throw the expected exception as well. Furthermore, with an "assertThrows" method, all tests can have the same structure, because it all comes down to a call for an assertion.
Besides those two points, are there any pro arguments for the new way? For me, it feels like its still superior to go with the annotations, as long as I am only testing a single method within a single test.
You missed a third way, the ExpectedException jUnit rule:
public class SimpleExpectedExceptionTest {
#Rule
public ExpectedException thrown= ExpectedException.none();
#Test
public void myMethod_throws_no_exception_when_passed_greeting() {
fixture.myMethod("hello");
}
#Test
public void myMethod_throws_MyException_when_passed_farewell() {
thrown.expect(MyException.class);
fixture.myMethod("goodbye");
}
}
I find this clearer than the #Test (expected = ...) version, since the expectation goes closer to the method call.
There is also the plain old Java version, which we used to make do with:
try {
fixture.myMethod("should throw");
fail("Expected an exception");
} catch (MyException e) {
// expected
}
Which of the many is "better" depends entirely on context. Don't adopt one universally. Pick the one that gives you the clearest test in a given situation.
When you begin coding non-test code in a lambda-centric style, it is likely that you'll find yourself wanting to use the lambda-centric assertThrows().
If all you want to test is that an exception is thrown, the first syntax is better. It is standard, it is concise, and it prevents you from writing the same ugly try-catch over and over.
You could have a test slightly more complicated where you want to assert that an exception is thrown and some method is not called. In this case, manually catching the exception is reasonable.
#Test
public void test1() {
DBClient dbClient = spy(new DBClient());
try {
new Server().doRequest(new InvalidRequest(), dbClient);
fail("should have thrown");
} catch (InvalidRequestException e) {
verify(dbClient, times(0)).query(any(Query.class));
}
}
Regarding the use on lambdas more specifically, it's up to you. Just note that Runnable cannot throw checked exceptions so you would need something like
#FunctionalInterface
public interface ThrowingRunnable<E extends Exception> {
void run() throws E;
}
I don't see a problem with either approach--there is no right or wrong in matters of style; just use the one that best suits each situation. I suggest that assertThrows should also check for thrown exceptions that aren't of the expected type, and, as #Dici suggests, that a functional interface that allows checked exceptions be used:
public static <X extends Throwable> Throwable assertThrows(
final Class<X> exceptionClass, final CheckedRunnable block) {
try {
block.run();
} catch(Throwable ex) {
if (exceptionClass.isInstance(ex)) {
return ex;
} else {
throw new AssertionError("Unexpected exception was thrown", ex);
}
}
fail("Failed to throw expected exception");
}
#FunctionalInterface
public interface CheckedRunnable<R extends RuntimeException> {
void run () throws Exception;
}
Our test code uses assert.
So is there any functions in assert class that can expect that a void function throws an exception?
#Test(expected=MyCustomException.class)
public void testBlahBlah_BlahBlah(){
voidMethodThatThrowsException();
}
This annotation of JUnit is what you are looking for.
Two ways I accomplish this:
First way is if I want to make sure an exception is thrown but not fail the test for it and I want to continue asserting.
try {
myObject.methodThatShouldThrowAnException("baddata");
assert.fail("Exception should have been thrown");
catch (SomeException e) { }
assert.equals(anotherMethod(), "hello");
The second way is if I don't want to do additional asserts (This is JUnit, could vary for the test framework you are using...TestNG has similar syntax):
#Test(expected = SomeException.class)
public void myTest() {
myObject.methodThatShouldThrowAnException("baddata");
}
Most testing frameworks (notably JUnit) will allow you to specify an exception class, so that the test will fail if an exception is not thrown.
If you can't use this method, then you can just wrap it in a try/catch and do the assertion in the catch block.
Ex (using JUnit):
#Test(expected=SomeException.class)
public void testSomeFunction_expectedSomeException(){
MyClass clazz = new MyClass();
clazz.someMethodThatThrowsTheException();
fail("shouldn't have gotten here...");
}
Or:
//#Test -- this annotation may or may not be supported by your framework
public void testSomeFunction_expectedSomeException(){
MyClass clazz = new MyClass();
try {
clazz.someMethodThatThrowsTheException();
fail("shouldn't have gotten here...");
}
catch (SomeException ex) {
// assert...
}
catch (Exception ex) {
fail("Some other exception occurred, also not good");
}
}
Suppose I have a class and a method
class A {
void foo() throws Exception() {
...
}
}
Now I would like to call foo for each instance of A delivered by a stream like:
void bar() throws Exception {
Stream<A> as = ...
as.forEach(a -> a.foo());
}
Question: How do I properly handle the exception? The code does not compile on my machine because I do not handle the possible exceptions that can be thrown by foo(). The throws Exception of bar seems to be useless here. Why is that?
You need to wrap your method call into another one, where you do not throw checked exceptions. You can still throw anything that is a subclass of RuntimeException.
A normal wrapping idiom is something like:
private void safeFoo(final A a) {
try {
a.foo();
} catch (Exception ex) {
throw new RuntimeException(ex);
}
}
(Supertype exception Exception is only used as example, never try to catch it yourself)
Then you can call it with: as.forEach(this::safeFoo).
If all you want is to invoke foo, and you prefer to propagate the exception as is (without wrapping), you can also just use Java's for loop instead (after turning the Stream into an Iterable with some trickery):
for (A a : (Iterable<A>) as::iterator) {
a.foo();
}
This is, at least, what I do in my JUnit tests, where I don't want to go through the trouble of wrapping my checked exceptions (and in fact prefer my tests to throw the unwrapped original ones)
This question may be a little old, but because I think the "right" answer here is only one way which can lead to some issues hidden Issues later in your code. Even if there is a little Controversy, Checked Exceptions exist for a reason.
The most elegant way in my opinion can you find was given by Misha here Aggregate runtime exceptions in Java 8 streams
by just performing the actions in "futures". So you can run all the working parts and collect not working Exceptions as a single one. Otherwise you could collect them all in a List and process them later.
A similar approach comes from Benji Weber. He suggests to create an own type to collect working and not working parts.
Depending on what you really want to achieve a simple mapping between the input values and Output Values occurred Exceptions may also work for you.
If you don't like any of these ways consider using (depending on the Original Exception) at least an own exception.
You might want to do one of the following:
propagate checked exception,
wrap it and propagate unchecked exception, or
catch the exception and stop propagation.
Several libraries let you do that easily. Example below is written using my NoException library.
// Propagate checked exception
as.forEach(Exceptions.sneak().consumer(A::foo));
// Wrap and propagate unchecked exception
as.forEach(Exceptions.wrap().consumer(A::foo));
as.forEach(Exceptions.wrap(MyUncheckedException::new).consumer(A::foo));
// Catch the exception and stop propagation (using logging handler for example)
as.forEach(Exceptions.log().consumer(Exceptions.sneak().consumer(A::foo)));
I suggest to use Google Guava Throwables class
propagate(Throwable throwable)
Propagates throwable as-is if it is an
instance of RuntimeException or Error, or else as a last resort, wraps
it in a RuntimeException and then propagates.**
void bar() {
Stream<A> as = ...
as.forEach(a -> {
try {
a.foo()
} catch(Exception e) {
throw Throwables.propagate(e);
}
});
}
UPDATE:
Now that it is deprecated use:
void bar() {
Stream<A> as = ...
as.forEach(a -> {
try {
a.foo()
} catch(Exception e) {
Throwables.throwIfUnchecked(e);
throw new RuntimeException(e);
}
});
}
You can wrap and unwrap exceptions this way.
class A {
void foo() throws Exception {
throw new Exception();
}
};
interface Task {
void run() throws Exception;
}
static class TaskException extends RuntimeException {
private static final long serialVersionUID = 1L;
public TaskException(Exception e) {
super(e);
}
}
void bar() throws Exception {
Stream<A> as = Stream.generate(()->new A());
try {
as.forEach(a -> wrapException(() -> a.foo())); // or a::foo instead of () -> a.foo()
} catch (TaskException e) {
throw (Exception)e.getCause();
}
}
static void wrapException(Task task) {
try {
task.run();
} catch (Exception e) {
throw new TaskException(e);
}
}
More readable way:
class A {
void foo() throws MyException() {
...
}
}
Just hide it in a RuntimeException to get it past forEach()
void bar() throws MyException {
Stream<A> as = ...
try {
as.forEach(a -> {
try {
a.foo();
} catch(MyException e) {
throw new RuntimeException(e);
}
});
} catch(RuntimeException e) {
throw (MyException) e.getCause();
}
}
Although at this point I won't hold against someone if they say skip the streams and go with a for loop, unless:
you're not creating your stream using Collection.stream(), i.e. not straight forward translation to a for loop.
you're trying to use parallelstream()
Is it possible to construct a snippet of code in Java that would make a hypothetical java.lang.ChuckNorrisException uncatchable?
Thoughts that came to mind are using for example interceptors or aspect-oriented programming.
I haven't tried this, so I don't know if the JVM would restrict something like this, but maybe you could compile code which throws ChuckNorrisException, but at runtime provide a class definition of ChuckNorrisException which does not extend Throwable.
UPDATE:
It doesn't work. It generates a verifier error:
Exception in thread "main" java.lang.VerifyError: (class: TestThrow, method: ma\
in signature: ([Ljava/lang/String;)V) Can only throw Throwable objects
Could not find the main class: TestThrow. Program will exit.
UPDATE 2:
Actually, you can get this to work if you disable the byte code verifier! (-Xverify:none)
UPDATE 3:
For those following from home, here is the full script:
Create the following classes:
public class ChuckNorrisException
extends RuntimeException // <- Comment out this line on second compilation
{
public ChuckNorrisException() { }
}
public class TestVillain {
public static void main(String[] args) {
try {
throw new ChuckNorrisException();
}
catch(Throwable t) {
System.out.println("Gotcha!");
}
finally {
System.out.println("The end.");
}
}
}
Compile classes:
javac -cp . TestVillain.java ChuckNorrisException.java
Run:
java -cp . TestVillain
Gotcha!
The end.
Comment out "extends RuntimeException" and recompile ChuckNorrisException.java only :
javac -cp . ChuckNorrisException.java
Run:
java -cp . TestVillain
Exception in thread "main" java.lang.VerifyError: (class: TestVillain, method: main signature: ([Ljava/lang/String;)V) Can only throw Throwable objects
Could not find the main class: TestVillain. Program will exit.
Run without verification:
java -Xverify:none -cp . TestVillain
The end.
Exception in thread "main"
After having pondered this, I have successfully created an uncatchable exception. I chose to name it JulesWinnfield, however, rather than Chuck, because it is one mushroom-cloud-laying-mother-exception. Furthermore, it might not be exactly what you had in mind, but it certainly can't be caught. Observe:
public static class JulesWinnfield extends Exception
{
JulesWinnfield()
{
System.err.println("Say 'What' again! I dare you! I double dare you!");
System.exit(25-17); // And you shall know I am the LORD
}
}
public static void main(String[] args)
{
try
{
throw new JulesWinnfield();
}
catch(JulesWinnfield jw)
{
System.out.println("There's a word for that Jules - a bum");
}
}
Et voila! Uncaught exception.
Output:
run:
Say 'What' again! I dare you! I double dare you!
Java Result: 8
BUILD SUCCESSFUL (total time: 0 seconds)
When I have a little more time, I'll see if I can't come up with something else, as well.
Also, check this out:
public static class JulesWinnfield extends Exception
{
JulesWinnfield() throws JulesWinnfield, VincentVega
{
throw new VincentVega();
}
}
public static class VincentVega extends Exception
{
VincentVega() throws JulesWinnfield, VincentVega
{
throw new JulesWinnfield();
}
}
public static void main(String[] args) throws VincentVega
{
try
{
throw new JulesWinnfield();
}
catch(JulesWinnfield jw)
{
}
catch(VincentVega vv)
{
}
}
Causes a stack overflow - again, exceptions remain uncaught.
With such an exception it would obviously be mandatory to use a System.exit(Integer.MIN_VALUE); from the constructor because this is what would happen if you threw such an exception ;)
Any code can catch Throwable. So no, whatever exception you create is going to be a subclass of Throwable and will be subject to being caught.
public class ChuckNorrisException extends Exception {
public ChuckNorrisException() {
System.exit(1);
}
}
(Granted, technically this exception is never actually thrown, but a proper ChuckNorrisException can't be thrown -- it throws you first.)
Any exception you throw has to extend Throwable, so it can be always caught. So answer is no.
If you want to make it difficult to handle, you can override methods getCause(), getMessage(), getStackTrace(), toString() to throw another java.lang.ChuckNorrisException.
My answer is based on #jtahlborn's idea, but it's a fully working Java program, that can be packaged into a JAR file and even deployed to your favorite application server as a part of a web application.
First of all, let's define ChuckNorrisException class so that it doesn't crash JVM from the beginning (Chuck really loves crashing JVMs BTW :)
package chuck;
import java.io.PrintStream;
import java.io.PrintWriter;
public class ChuckNorrisException extends Exception {
public ChuckNorrisException() {
}
#Override
public Throwable getCause() {
return null;
}
#Override
public String getMessage() {
return toString();
}
#Override
public void printStackTrace(PrintWriter s) {
super.printStackTrace(s);
}
#Override
public void printStackTrace(PrintStream s) {
super.printStackTrace(s);
}
}
Now goes Expendables class to construct it:
package chuck;
import javassist.*;
public class Expendables {
private static Class clz;
public static ChuckNorrisException getChuck() {
try {
if (clz == null) {
ClassPool pool = ClassPool.getDefault();
CtClass cc = pool.get("chuck.ChuckNorrisException");
cc.setSuperclass(pool.get("java.lang.Object"));
clz = cc.toClass();
}
return (ChuckNorrisException)clz.newInstance();
} catch (Exception ex) {
throw new RuntimeException(ex);
}
}
}
And finally the Main class to kick some butt:
package chuck;
public class Main {
public void roundhouseKick() throws Exception {
throw Expendables.getChuck();
}
public void foo() {
try {
roundhouseKick();
} catch (Throwable ex) {
System.out.println("Caught " + ex.toString());
}
}
public static void main(String[] args) {
try {
System.out.println("before");
new Main().foo();
System.out.println("after");
} finally {
System.out.println("finally");
}
}
}
Compile and run it with following command:
java -Xverify:none -cp .:<path_to_javassist-3.9.0.GA.jar> chuck.Main
You will get following output:
before
finally
No surprise - it's a roundhouse kick after all :)
In the constructor you could start a thread which repeatedly calls originalThread.stop (ChuckNorisException.this)
The thread could catch the exception repeatedly but would keep throwing it until it dies.
No. All exceptions in Java must subclass java.lang.Throwable, and although it may not be good practice, you can catch every type of exception like so:
try {
//Stuff
} catch ( Throwable T ){
//Doesn't matter what it was, I caught it.
}
See the java.lang.Throwable documentation for more information.
If you're trying to avoid checked exceptions (ones that must be explicitly handled) then you will want to subclass Error, or RuntimeException.
Actually the accepted answer is not so nice because Java needs to be run without verification, i.e. the code would not work under normal circumstances.
AspectJ to the rescue for the real solution!
Exception class:
package de.scrum_master.app;
public class ChuckNorrisException extends RuntimeException {
public ChuckNorrisException(String message) {
super(message);
}
}
Aspect:
package de.scrum_master.aspect;
import de.scrum_master.app.ChuckNorrisException;
public aspect ChuckNorrisAspect {
before(ChuckNorrisException chuck) : handler(*) && args(chuck) {
System.out.println("Somebody is trying to catch Chuck Norris - LOL!");
throw chuck;
}
}
Sample application:
package de.scrum_master.app;
public class Application {
public static void main(String[] args) {
catchAllMethod();
}
private static void catchAllMethod() {
try {
exceptionThrowingMethod();
}
catch (Throwable t) {
System.out.println("Gotcha, " + t.getClass().getSimpleName() + "!");
}
}
private static void exceptionThrowingMethod() {
throw new ChuckNorrisException("Catch me if you can!");
}
}
Output:
Somebody is trying to catch Chuck Norris - LOL!
Exception in thread "main" de.scrum_master.app.ChuckNorrisException: Catch me if you can!
at de.scrum_master.app.Application.exceptionThrowingMethod(Application.java:18)
at de.scrum_master.app.Application.catchAllMethod(Application.java:10)
at de.scrum_master.app.Application.main(Application.java:5)
A variant on the theme is the surprising fact that you can throw undeclared checked exceptions from Java code. Since it is not declared in the methods signature, the compiler won't let you catch the exception itself, though you can catch it as java.lang.Exception.
Here's a helper class that lets you throw anything, declared or not:
public class SneakyThrow {
public static RuntimeException sneak(Throwable t) {
throw SneakyThrow.<RuntimeException> throwGivenThrowable(t);
}
private static <T extends Throwable> RuntimeException throwGivenThrowable(Throwable t) throws T {
throw (T) t;
}
}
Now throw SneakyThrow.sneak(new ChuckNorrisException()); does throw a ChuckNorrisException, but the compiler complains in
try {
throw SneakyThrow.sneak(new ChuckNorrisException());
} catch (ChuckNorrisException e) {
}
about catching an exception that is not thrown if ChuckNorrisException is a checked exception.
The only ChuckNorrisExceptions in Java should be OutOfMemoryError and StackOverflowError.
You can actually "catch" them in the means that a catch(OutOfMemoryError ex) will execute in case the exception is thrown, but that block will automatically rethrow the exception to the caller.
I don't think that public class ChuckNorrisError extends Error does the trick but you could give it a try. I found no documentation about extending Error
Is it possible to construct a snippet of code in java that would make a hypothetical java.lang.ChuckNorrisException uncatchable?
Yes, and here's the answer: Design your java.lang.ChuckNorrisException such that it is not an instance of java.lang.Throwable. Why? An unthrowable object is uncatchable by definition because you can never catch something that can never be thrown.
You can keep ChuckNorris internal or private and encapsulate him or swollow him...
try { doChuckAction(); } catch(ChuckNorrisException cne) { /*do something else*/ }
Two fundamental problems with exception handling in Java are that it uses the type of an exception to indicate whether action should be taken based upon it, and that anything which takes action based upon an exception (i.e. "catch"es it) is presumed to resolve the underlying condition. It would be useful to have a means by which an exception object could decide which handlers should execute, and whether the handlers that have executed so far have cleaned things up enough for the present method to satisfy its exit conditions. While this could be used to make "uncatchable" exceptions, two bigger uses would be to (1) make exceptions which will only be considered handled when they're caught by code that actually knows how to deal with them, and (2) allow for sensible handling of exceptions which occur in a finally block (if a FooException during a finally block during the unwinding of a BarException, both exceptions should propagate up the call stack; both should be catchable, but unwinding should continue until both have been caught). Unfortunately, I don't think there would be any way to make existing exception-handling code work that way without breaking things.
It is easily possible to simulate a uncaught exception on the current thread. This will trigger the regular behavior of an uncaught exception, and thus gets the job done semantically. It will, however, not necessarily stop the current thread's execution, as no exception is actually thrown.
Throwable exception = /* ... */;
Thread currentThread = Thread.currentThread();
Thread.UncaughtExceptionHandler uncaughtExceptionHandler =
currentThread.getUncaughtExceptionHandler();
uncaughtExceptionHandler.uncaughtException(currentThread, exception);
// May be reachable, depending on the uncaught exception handler.
This is actually useful in (very rare) situations, for example when proper Error handling is required, but the method is invoked from a framework catching (and discarding) any Throwable.
Call System.exit(1) in the finalize, and just throw a copy of the exception from all the other methods, so that the program will exit.