org.springframework.batch.core.scope.context.SynchronizationManagerSupport Maven / Gradle / Ivy
/*
* Copyright 2013-2018 the original author or authors.
*
* Licensed 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
*
* https://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.springframework.batch.core.scope.context;
import java.util.Map;
import java.util.Stack;
import java.util.concurrent.ConcurrentHashMap;
import java.util.concurrent.atomic.AtomicInteger;
import org.springframework.batch.core.jsr.configuration.support.BatchPropertyContext;
import org.springframework.lang.Nullable;
/**
* Central convenience class for framework use in managing the scope
* context.
*
* @author Dave Syer
* @author Jimmy Praet
* @author Mahmoud Ben Hassine
* @since 3.0
*/
public abstract class SynchronizationManagerSupport {
/*
* We have to deal with single and multi-threaded execution, with a single
* and with multiple step execution instances. That's 2x2 = 4 scenarios.
*/
/**
* Storage for the current execution; has to be ThreadLocal because it
* is needed to locate a context in components that are not part of a
* step/job (like when re-hydrating a scoped proxy). Doesn't use
* InheritableThreadLocal because there are side effects if a step is trying
* to run multiple child steps (e.g. with partitioning). The Stack is used
* to cover the single threaded case, so that the API is the same as
* multi-threaded.
*/
private final ThreadLocal> executionHolder = new ThreadLocal<>();
/**
* Reference counter for each execution: how many threads are using the
* same one?
*/
private final Map counts = new ConcurrentHashMap<>();
/**
* Simple map from a running execution to the associated context.
*/
private final Map contexts = new ConcurrentHashMap<>();
/**
* Getter for the current context if there is one, otherwise returns {@code null}.
*
* @return the current context or {@code null} if there is none (if one
* has not been registered for this thread).
*/
@Nullable
public C getContext() {
if (getCurrent().isEmpty()) {
return null;
}
synchronized (contexts) {
return contexts.get(getCurrent().peek());
}
}
/**
* Register a context with the current thread - always put a matching {@link #close()} call in a finally block to
* ensure that the correct
* context is available in the enclosing block.
*
* @param execution the execution to register
* @return a new context or the current one if it has the same
* execution
*/
@Nullable
public C register(@Nullable E execution) {
if (execution == null) {
return null;
}
getCurrent().push(execution);
C context;
synchronized (contexts) {
context = contexts.get(execution);
if (context == null) {
context = createNewContext(execution, null);
contexts.put(execution, context);
}
}
increment();
return context;
}
/**
* Register a context with the current thread - always put a matching {@link #close()} call in a finally block to
* ensure that the correct
* context is available in the enclosing block.
*
* @param execution the execution to register
* @param propertyContext instance of {@link BatchPropertyContext} to be registered with this thread.
* @return a new context or the current one if it has the same
* execution
*/
@Nullable
public C register(@Nullable E execution, @Nullable BatchPropertyContext propertyContext) {
if (execution == null) {
return null;
}
getCurrent().push(execution);
C context;
synchronized (contexts) {
context = contexts.get(execution);
if (context == null) {
context = createNewContext(execution, propertyContext);
contexts.put(execution, context);
}
}
increment();
return context;
}
/**
* Method for unregistering the current context - should always and only be
* used by in conjunction with a matching {@link #register(Object)} to ensure that {@link #getContext()} always returns
* the correct value.
* Does not call close on the context - that is left up to the caller
* because he has a reference to the context (having registered it) and only
* he has knowledge of when the execution actually ended.
*/
public void close() {
C oldSession = getContext();
if (oldSession == null) {
return;
}
decrement();
}
private void decrement() {
E current = getCurrent().pop();
if (current != null) {
int remaining = counts.get(current).decrementAndGet();
if (remaining <= 0) {
synchronized (contexts) {
contexts.remove(current);
counts.remove(current);
}
}
}
}
public void increment() {
E current = getCurrent().peek();
if (current != null) {
AtomicInteger count;
synchronized (counts) {
count = counts.get(current);
if (count == null) {
count = new AtomicInteger();
counts.put(current, count);
}
}
count.incrementAndGet();
}
}
public Stack getCurrent() {
if (executionHolder.get() == null) {
executionHolder.set(new Stack<>());
}
return executionHolder.get();
}
/**
* A convenient "deep" close operation. Call this instead of {@link #close()} if the execution for the current
* context is ending.
* Delegates to {@link #close(Object)} and then ensures that {@link #close()} is also called in a finally block.
*/
public void release() {
C context = getContext();
try {
if (context != null) {
close(context);
}
} finally {
close();
}
}
protected abstract void close(C context);
protected abstract C createNewContext(E execution, @Nullable BatchPropertyContext propertyContext);
}