Java tutorial
/* * Copyright (c) 2000, 2013, Oracle and/or its affiliates. All rights reserved. * DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER. * * This code is free software; you can redistribute it and/or modify it * under the terms of the GNU General Public License version 2 only, as * published by the Free Software Foundation. Oracle designates this * particular file as subject to the "Classpath" exception as provided * by Oracle in the LICENSE file that accompanied this code. * * This code is distributed in the hope that it will be useful, but WITHOUT * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License * version 2 for more details (a copy is included in the LICENSE file that * accompanied this code). * * You should have received a copy of the GNU General Public License version * 2 along with this work; if not, write to the Free Software Foundation, * Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA. * * Please contact Oracle, 500 Oracle Parkway, Redwood Shores, CA 94065 USA * or visit www.oracle.com if you need additional information or have any * questions. */ package java.beans; /** * The PersistenceDelegate class takes the responsibility * for expressing the state of an instance of a given class * in terms of the methods in the class's public API. Instead * of associating the responsibility of persistence with * the class itself as is done, for example, by the * {@code readObject} and {@code writeObject} * methods used by the {@code ObjectOutputStream}, streams like * the {@code XMLEncoder} which * use this delegation model can have their behavior controlled * independently of the classes themselves. Normally, the class * is the best place to put such information and conventions * can easily be expressed in this delegation scheme to do just that. * Sometimes however, it is the case that a minor problem * in a single class prevents an entire object graph from * being written and this can leave the application * developer with no recourse but to attempt to shadow * the problematic classes locally or use alternative * persistence techniques. In situations like these, the * delegation model gives a relatively clean mechanism for * the application developer to intervene in all parts of the * serialization process without requiring that modifications * be made to the implementation of classes which are not part * of the application itself. * <p> * In addition to using a delegation model, this persistence * scheme differs from traditional serialization schemes * in requiring an analog of the {@code writeObject} * method without a corresponding {@code readObject} * method. The {@code writeObject} analog encodes each * instance in terms of its public API and there is no need to * define a {@code readObject} analog * since the procedure for reading the serialized form * is defined by the semantics of method invocation as laid * out in the Java Language Specification. * Breaking the dependency between {@code writeObject} * and {@code readObject} implementations, which may * change from version to version, is the key factor * in making the archives produced by this technique immune * to changes in the private implementations of the classes * to which they refer. * <p> * A persistence delegate, may take control of all * aspects of the persistence of an object including: * <ul> * <li> * Deciding whether or not an instance can be mutated * into another instance of the same class. * <li> * Instantiating the object, either by calling a * public constructor or a public factory method. * <li> * Performing the initialization of the object. * </ul> * @see XMLEncoder * * @since 1.4 * * @author Philip Milne */ public abstract class PersistenceDelegate { /** * The {@code writeObject} is a single entry point to the persistence * and is used by an {@code Encoder} in the traditional * mode of delegation. Although this method is not final, * it should not need to be subclassed under normal circumstances. * <p> * This implementation first checks to see if the stream * has already encountered this object. Next the * {@code mutatesTo} method is called to see if * that candidate returned from the stream can * be mutated into an accurate copy of {@code oldInstance}. * If it can, the {@code initialize} method is called to * perform the initialization. If not, the candidate is removed * from the stream, and the {@code instantiate} method * is called to create a new candidate for this object. * * @param oldInstance The instance that will be created by this expression. * @param out The stream to which this expression will be written. * * @throws NullPointerException if {@code out} is {@code null} */ public void writeObject(Object oldInstance, Encoder out) { Object newInstance = out.get(oldInstance); if (!mutatesTo(oldInstance, newInstance)) { out.remove(oldInstance); out.writeExpression(instantiate(oldInstance, out)); } else { initialize(oldInstance.getClass(), oldInstance, newInstance, out); } } /** * Returns true if an <em>equivalent</em> copy of {@code oldInstance} may be * created by applying a series of statements to {@code newInstance}. * In the specification of this method, we mean by equivalent that the modified instance * is indistinguishable from {@code oldInstance} in the behavior * of the relevant methods in its public API. [Note: we use the * phrase <em>relevant</em> methods rather than <em>all</em> methods * here only because, to be strictly correct, methods like {@code hashCode} * and {@code toString} prevent most classes from producing truly * indistinguishable copies of their instances]. * <p> * The default behavior returns {@code true} * if the classes of the two instances are the same. * * @param oldInstance The instance to be copied. * @param newInstance The instance that is to be modified. * @return True if an equivalent copy of {@code newInstance} may be * created by applying a series of mutations to {@code oldInstance}. */ protected boolean mutatesTo(Object oldInstance, Object newInstance) { return (newInstance != null && oldInstance != null && oldInstance.getClass() == newInstance.getClass()); } /** * Returns an expression whose value is {@code oldInstance}. * This method is used to characterize the constructor * or factory method that should be used to create the given object. * For example, the {@code instantiate} method of the persistence * delegate for the {@code Field} class could be defined as follows: * <pre> * Field f = (Field)oldInstance; * return new Expression(f, f.getDeclaringClass(), "getField", new Object[]{f.getName()}); * </pre> * Note that we declare the value of the returned expression so that * the value of the expression (as returned by {@code getValue}) * will be identical to {@code oldInstance}. * * @param oldInstance The instance that will be created by this expression. * @param out The stream to which this expression will be written. * @return An expression whose value is {@code oldInstance}. * * @throws NullPointerException if {@code out} is {@code null} * and this value is used in the method */ protected abstract Expression instantiate(Object oldInstance, Encoder out); /** * Produce a series of statements with side effects on {@code newInstance} * so that the new instance becomes <em>equivalent</em> to {@code oldInstance}. * In the specification of this method, we mean by equivalent that, after the method * returns, the modified instance is indistinguishable from * {@code newInstance} in the behavior of all methods in its * public API. * <p> * The implementation typically achieves this goal by producing a series of * "what happened" statements involving the {@code oldInstance} * and its publicly available state. These statements are sent * to the output stream using its {@code writeExpression} * method which returns an expression involving elements in * a cloned environment simulating the state of an input stream during * reading. Each statement returned will have had all instances * the old environment replaced with objects which exist in the new * one. In particular, references to the target of these statements, * which start out as references to {@code oldInstance} are returned * as references to the {@code newInstance} instead. * Executing these statements effects an incremental * alignment of the state of the two objects as a series of * modifications to the objects in the new environment. * By the time the initialize method returns it should be impossible * to tell the two instances apart by using their public APIs. * Most importantly, the sequence of steps that were used to make * these objects appear equivalent will have been recorded * by the output stream and will form the actual output when * the stream is flushed. * <p> * The default implementation, calls the {@code initialize} * method of the type's superclass. * * @param type the type of the instances * @param oldInstance The instance to be copied. * @param newInstance The instance that is to be modified. * @param out The stream to which any initialization statements should be written. * * @throws NullPointerException if {@code out} is {@code null} */ protected void initialize(Class<?> type, Object oldInstance, Object newInstance, Encoder out) { Class<?> superType = type.getSuperclass(); PersistenceDelegate info = out.getPersistenceDelegate(superType); info.initialize(superType, oldInstance, newInstance, out); } }