/*
* Licensed to the Apache Software Foundation (ASF) under one or more
* contributor license agreements. See the NOTICE file distributed with
* this work for additional information regarding copyright ownership.
* The ASF licenses this file to You under the Apache License, Version 2.0
* (the "License"); you may not use this file except in compliance with
* the License. You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
package org.apache.commons.lang3.concurrent;
import java.util.concurrent.ConcurrentMap;
import java.util.concurrent.ExecutionException;
import java.util.concurrent.Future;
import java.util.concurrent.TimeUnit;
/**
* <p>
* An utility class providing functionality related to the {@code
* java.util.concurrent} package.
* </p>
*
* @since 3.0
* @version $Id: ConcurrentUtils.java 1088899 2011-04-05 05:31:27Z bayard $
*/
public class ConcurrentUtils {
/**
* Private constructor so that no instances can be created. This class
* contains only static utility methods.
*/
private ConcurrentUtils() {
}
/**
* Inspects the cause of the specified {@code ExecutionException} and
* creates a {@code ConcurrentException} with the checked cause if
* necessary. This method performs the following checks on the cause of the
* passed in exception:
* <ul>
* <li>If the passed in exception is <b>null</b> or the cause is
* <b>null</b>, this method returns <b>null</b>.</li>
* <li>If the cause is a runtime exception, it is directly thrown.</li>
* <li>If the cause is an error, it is directly thrown, too.</li>
* <li>In any other case the cause is a checked exception. The method then
* creates a {@link ConcurrentException}, initializes it with the cause, and
* returns it.</li>
* </ul>
*
* @param ex the exception to be processed
* @return a {@code ConcurrentException} with the checked cause
*/
public static ConcurrentException extractCause(ExecutionException ex) {
if (ex == null || ex.getCause() == null) {
return null;
}
throwCause(ex);
return new ConcurrentException(ex.getMessage(), ex.getCause());
}
/**
* Inspects the cause of the specified {@code ExecutionException} and
* creates a {@code ConcurrentRuntimeException} with the checked cause if
* necessary. This method works exactly like
* {@link #extractCause(ExecutionException)}. The only difference is that
* the cause of the specified {@code ExecutionException} is extracted as a
* runtime exception. This is an alternative for client code that does not
* want to deal with checked exceptions.
*
* @param ex the exception to be processed
* @return a {@code ConcurrentRuntimeException} with the checked cause
*/
public static ConcurrentRuntimeException extractCauseUnchecked(
ExecutionException ex) {
if (ex == null || ex.getCause() == null) {
return null;
}
throwCause(ex);
return new ConcurrentRuntimeException(ex.getMessage(), ex.getCause());
}
/**
* Handles the specified {@code ExecutionException}. This method calls
* {@link #extractCause(ExecutionException)} for obtaining the cause of the
* exception - which might already cause an unchecked exception or an error
* being thrown. If the cause is a checked exception however, it is wrapped
* in a {@code ConcurrentException}, which is thrown. If the passed in
* exception is <b>null</b> or has no cause, the method simply returns
* without throwing an exception.
*
* @param ex the exception to be handled
* @throws ConcurrentException if the cause of the {@code
* ExecutionException} is a checked exception
*/
public static void handleCause(ExecutionException ex)
throws ConcurrentException {
ConcurrentException cex = extractCause(ex);
if (cex != null) {
throw cex;
}
}
/**
* Handles the specified {@code ExecutionException} and transforms it into a
* runtime exception. This method works exactly like
* {@link #handleCause(ExecutionException)}, but instead of a
* {@link ConcurrentException} it throws a
* {@link ConcurrentRuntimeException}. This is an alternative for client
* code that does not want to deal with checked exceptions.
*
* @param ex the exception to be handled
* @throws ConcurrentRuntimeException if the cause of the {@code
* ExecutionException} is a checked exception; this exception is then
* wrapped in the thrown runtime exception
*/
public static void handleCauseUnchecked(ExecutionException ex) {
ConcurrentRuntimeException crex = extractCauseUnchecked(ex);
if (crex != null) {
throw crex;
}
}
/**
* Tests whether the specified {@code Throwable} is a checked exception. If
* not, an exception is thrown.
*
* @param ex the {@code Throwable} to check
* @return a flag whether the passed in exception is a checked exception
* @throws IllegalArgumentException if the {@code Throwable} is not a
* checked exception
*/
static Throwable checkedException(Throwable ex) {
if (ex != null && !(ex instanceof RuntimeException)
&& !(ex instanceof Error)) {
return ex;
} else {
throw new IllegalArgumentException("Not a checked exception: " + ex);
}
}
/**
* Tests whether the cause of the specified {@code ExecutionException}
* should be thrown and does it if necessary.
*
* @param ex the exception in question
*/
private static void throwCause(ExecutionException ex) {
if (ex.getCause() instanceof RuntimeException) {
throw (RuntimeException) ex.getCause();
}
if (ex.getCause() instanceof Error) {
throw (Error) ex.getCause();
}
}
//-----------------------------------------------------------------------
/**
* Invokes the specified {@code ConcurrentInitializer} and returns the
* object produced by the initializer. This method just invokes the {@code
* get()} method of the given {@code ConcurrentInitializer}. It is
* <b>null</b>-safe: if the argument is <b>null</b>, result is also
* <b>null</b>.
*
* @param <T> the type of the object produced by the initializer
* @param initializer the {@code ConcurrentInitializer} to be invoked
* @return the object managed by the {@code ConcurrentInitializer}
* @throws ConcurrentException if the {@code ConcurrentInitializer} throws
* an exception
*/
public static <T> T initialize(ConcurrentInitializer<T> initializer)
throws ConcurrentException {
return (initializer != null) ? initializer.get() : null;
}
/**
* Invokes the specified {@code ConcurrentInitializer} and transforms
* occurring exceptions to runtime exceptions. This method works like
* {@link #initialize(ConcurrentInitializer)}, but if the {@code
* ConcurrentInitializer} throws a {@link ConcurrentException}, it is
* caught, and the cause is wrapped in a {@link ConcurrentRuntimeException}.
* So client code does not have to deal with checked exceptions.
*
* @param <T> the type of the object produced by the initializer
* @param initializer the {@code ConcurrentInitializer} to be invoked
* @return the object managed by the {@code ConcurrentInitializer}
* @throws ConcurrentRuntimeException if the initializer throws an exception
*/
public static <T> T initializeUnchecked(ConcurrentInitializer<T> initializer) {
try {
return initialize(initializer);
} catch (ConcurrentException cex) {
throw new ConcurrentRuntimeException(cex.getCause());
}
}
//-----------------------------------------------------------------------
/**
* <p>
* Puts a value in the specified {@code ConcurrentMap} if the key is not yet
* present. This method works similar to the {@code putIfAbsent()} method of
* the {@code ConcurrentMap} interface, but the value returned is different.
* Basically, this method is equivalent to the following code fragment:
*
* <pre>
* if (!map.containsKey(key)) {
* map.put(key, value);
* return value;
* } else {
* return map.get(key);
* }
* </pre>
*
* except that the action is performed atomically. So this method always
* returns the value which is stored in the map.
* </p>
* <p>
* This method is <b>null</b>-safe: It accepts a <b>null</b> map as input
* without throwing an exception. In this case the return value is
* <b>null</b>, too.
* </p>
*
* @param <K> the type of the keys of the map
* @param <V> the type of the values of the map
* @param map the map to be modified
* @param key the key of the value to be added
* @param value the value to be added
* @return the value stored in the map after this operation
*/
public static <K, V> V putIfAbsent(ConcurrentMap<K, V> map, K key, V value) {
if (map == null) {
return null;
}
V result = map.putIfAbsent(key, value);
return (result != null) ? result : value;
}
/**
* Checks if a concurrent map contains a key and creates a corresponding
* value if not. This method first checks the presence of the key in the
* given map. If it is already contained, its value is returned. Otherwise
* the {@code get()} method of the passed in {@link ConcurrentInitializer}
* is called. With the resulting object
* {@link #putIfAbsent(ConcurrentMap, Object, Object)} is called. This
* handles the case that in the meantime another thread has added the key to
* the map. Both the map and the initializer can be <b>null</b>; in this
* case this method simply returns <b>null</b>.
*
* @param <K> the type of the keys of the map
* @param <V> the type of the values of the map
* @param map the map to be modified
* @param key the key of the value to be added
* @param init the {@link ConcurrentInitializer} for creating the value
* @return the value stored in the map after this operation; this may or may
* not be the object created by the {@link ConcurrentInitializer}
* @throws ConcurrentException if the initializer throws an exception
*/
public static <K, V> V createIfAbsent(ConcurrentMap<K, V> map, K key,
ConcurrentInitializer<V> init) throws ConcurrentException {
if (map == null || init == null) {
return null;
}
V value = map.get(key);
if (value == null) {
return putIfAbsent(map, key, init.get());
}
return value;
}
/**
* Checks if a concurrent map contains a key and creates a corresponding
* value if not, suppressing checked exceptions. This method calls
* {@code createIfAbsent()}. If a {@link ConcurrentException} is thrown, it
* is caught and re-thrown as a {@link ConcurrentRuntimeException}.
*
* @param <K> the type of the keys of the map
* @param <V> the type of the values of the map
* @param map the map to be modified
* @param key the key of the value to be added
* @param init the {@link ConcurrentInitializer} for creating the value
* @return the value stored in the map after this operation; this may or may
* not be the object created by the {@link ConcurrentInitializer}
* @throws ConcurrentRuntimeException if the initializer throws an exception
*/
public static <K, V> V createIfAbsentUnchecked(ConcurrentMap<K, V> map,
K key, ConcurrentInitializer<V> init) {
try {
return createIfAbsent(map, key, init);
} catch (ConcurrentException cex) {
throw new ConcurrentRuntimeException(cex.getCause());
}
}
//-----------------------------------------------------------------------
/**
* <p>
* Gets an implementation of <code>Future</code> that is immediately done
* and returns the specified constant value.
* </p>
* <p>
* This can be useful to return a simple constant immediately from the
* concurrent processing, perhaps as part of avoiding nulls.
* A constant future can also be useful in testing.
* </p>
*
* @param <T> the type of the value used by this {@code Future} object
* @param value the constant value to return, may be null
* @return an instance of Future that will return the value, never null
*/
public static <T> Future<T> constantFuture(T value) {
return new ConstantFuture<T>(value);
}
/**
* A specialized {@code Future} implementation which wraps a constant value.
* @param <T> the type of the value wrapped by this class
*/
static final class ConstantFuture<T> implements Future<T> {
/** The constant value. */
private final T value;
/**
* Creates a new instance of {@code ConstantFuture} and initializes it
* with the constant value.
*
* @param value the value (may be <b>null</b>)
*/
ConstantFuture(T value) {
this.value = value;
}
/**
* {@inheritDoc} This implementation always returns <b>true</b> because
* the constant object managed by this {@code Future} implementation is
* always available.
*/
public boolean isDone() {
return true;
}
/**
* {@inheritDoc} This implementation just returns the constant value.
*/
public T get() {
return value;
}
/**
* {@inheritDoc} This implementation just returns the constant value; it
* does not block, therefore the timeout has no meaning.
*/
public T get(long timeout, TimeUnit unit) {
return value;
}
/**
* {@inheritDoc} This implementation always returns <b>false</b>; there
* is no background process which could be cancelled.
*/
public boolean isCancelled() {
return false;
}
/**
* {@inheritDoc} The cancel operation is not supported. This
* implementation always returns <b>false</b>.
*/
public boolean cancel(boolean mayInterruptIfRunning) {
return false;
}
}
}
|