/*******************************************************************************
* Copyright (c) 1998, 2015 Oracle and/or its affiliates. All rights reserved.
* This program and the accompanying materials are made available under the
* terms of the Eclipse Public License v1.0 and Eclipse Distribution License v. 1.0
* which accompanies this distribution.
* The Eclipse Public License is available at http://www.eclipse.org/legal/epl-v10.html
* and the Eclipse Distribution License is available at
* http://www.eclipse.org/org/documents/edl-v10.php.
*
* Contributors:
* Gordon Yorke - Part of the Cache Interceptor feature. (ER 219683)
******************************************************************************/
package org.eclipse.persistence.internal.identitymaps;
import java.util.Enumeration;
import java.util.HashMap;
import java.util.Map;
import org.eclipse.persistence.descriptors.ClassDescriptor;
import org.eclipse.persistence.exceptions.QueryException;
import org.eclipse.persistence.indirection.ValueHolderInterface;
import org.eclipse.persistence.internal.sessions.AbstractSession;
import org.eclipse.persistence.mappings.ForeignReferenceMapping;
/**
* <p><b>Purpose</b>: Provides the interface for IdentityMap interaction.
* <p><b>Responsibilities</b>:
* Provides access to all of the public interface for the EclipseLink IdentityMaps.
* This interface can be used if implementing custom identity maps that are radically
* different than the stock IdentityMaps otherwise simply extending the appropriate
* IdentityMap implementation is the best approach.
*
* @see CacheKey
* @since EclipseLink 1.0M7
*/
public interface IdentityMap extends Cloneable{
/**
* Acquire a deferred lock on the object.
* This is used while reading if the object has relationships without indirection.
* This first thread will get an active lock.
* Other threads will get deferred locks, all threads will wait until all other threads are complete before releasing their locks.
*/
public CacheKey acquireDeferredLock(Object primaryKey, boolean isCacheCheckComplete);
/**
* Acquire an active lock on the object.
* This is used by reading (when using indirection or no relationships) and by merge.
*/
public CacheKey acquireLock(Object primaryKey, boolean forMerge, boolean isCacheCheckComplete);
/**
* Acquire an active lock on the object, if not already locked.
* This is used by merge for missing existing objects.
*/
public CacheKey acquireLockNoWait(Object primaryKey, boolean forMerge);
/**
* Acquire an active lock on the object, if not already locked.
* This is used by merge for missing existing objects.
*/
public CacheKey acquireLockWithWait(Object primaryKey, boolean forMerge, int wait);
/**
* Acquire a read lock on the object.
* This is used by UnitOfWork cloning.
* This will allow multiple users to read the same object but prevent writes to the object while the read lock is held.
*/
public CacheKey acquireReadLockOnCacheKey(Object primaryKey);
/**
* Acquire a read lock on the object, if not already locked.
* This is used by UnitOfWork cloning.
* This will allow multiple users to read the same object but prevent writes to the object while the read lock is held.
*/
public CacheKey acquireReadLockOnCacheKeyNoWait(Object primaryKey);
/**
* Clone the map and all of the CacheKeys.
* This is used by UnitOfWork commitAndResumeOnFailure to avoid corrupting the cache during a failed commit.
*/
public Object clone();
/**
* Add all locked CacheKeys to the map grouped by thread.
* Used to print all the locks in the identity map.
*/
public void collectLocks(HashMap threadList);
/**
* Return true if an CacheKey with the primary key is in the map.
* User API.
* @param primaryKey is the primary key for the object to search for.
*/
public boolean containsKey(Object primaryKey);
/**
* Allow for the cache to be iterated on.
*/
public Enumeration elements();
/**
* Return the object cached in the identity map or null if it could not be found.
* User API.
*/
public Object get(Object primaryKey);
/**
* ADVANCED:
* Using a list of Entity PK this method will attempt to bulk load the entire list from the cache.
* In certain circumstances this can have large performance improvements over loading each item individually.
* @param pkList List of Entity PKs to extract from the cache
* @param ClassDescriptor Descriptor type to be retrieved.
* @return Map of Entity PKs associated to the Entities that were retrieved
* @throws QueryException
*/
public Map<Object, Object> getAllFromIdentityMapWithEntityPK(Object[] pkList, ClassDescriptor descriptor, AbstractSession session);
/**
* ADVANCED:
* Using a list of Entity PK this method will attempt to bulk load the entire list from the cache.
* In certain circumstances this can have large performance improvements over loading each item individually.
* @param pkList List of Entity PKs to extract from the cache
* @param ClassDescriptor Descriptor type to be retrieved.
* @return Map of Entity PKs associated to the Entities that were retrieved
* @throws QueryException
*/
public Map<Object, CacheKey> getAllCacheKeysFromIdentityMapWithEntityPK(Object[] pkList, ClassDescriptor descriptor, AbstractSession session);
/**
* Get the cache key (with object) for the primary key.
*/
public CacheKey getCacheKey(Object primaryKey, boolean forMerge);
/**
* Get the cache key (with object) for the primary key in order to acquire a lock.
*/
public CacheKey getCacheKeyForLock(Object primaryKey);
/**
* Return the class that this is the map for.
*/
public Class getDescriptorClass();
/**
* Return the descriptor that this is the map for.
*/
public ClassDescriptor getDescriptor();
/**
* @return The maxSize for the IdentityMap (NOTE: some subclasses may use this differently).
*/
public int getMaxSize();
/**
* Return the number of CacheKeys in the IdentityMap.
* This may contain weak referenced objects that have been garbage collected.
*/
public int getSize();
/**
* Return the number of actual objects of type myClass in the IdentityMap.
* Recurse = true will include subclasses of myClass in the count.
*/
public int getSize(Class myClass, boolean recurse);
/**
* Get the wrapper object from the cache key associated with the given primary key,
* this is used for EJB2.
*/
public Object getWrapper(Object primaryKey);
/**
* Get the write lock value from the cache key associated to the primarykey.
* User API.
*/
public Object getWriteLockValue(Object primaryKey);
/**
* Allow for the CacheKeys to be iterated on.
* Read locks should be checked
*/
public Enumeration keys();
/**
* Allow for the CacheKeys to be iterated on.
* @param checkReadLocks - true if readLocks should be checked, false otherwise.
*/
public Enumeration keys(boolean checkReadLocks);
/**
* Notify the cache that a lazy relationship has been triggered in the object
* and the cache may need to be updated
*/
public void lazyRelationshipLoaded(Object rootEntity, ValueHolderInterface valueHolder, ForeignReferenceMapping mapping);
/**
* Store the object in the cache at its primary key.
* This is used by InsertObjectQuery, typically into the UnitOfWork identity map.
* Merge and reads do not use put, but acquireLock.
* Also an advanced (very) user API.
* @param primaryKey is the primary key for the object.
* @param object is the domain object to cache.
* @param writeLockValue is the current write lock value of object, if null the version is ignored.
*/
public CacheKey put(Object primaryKey, Object object, Object writeLockValue, long readTime);
/**
* This method may be called during initialize all identity maps. It allows the identity map
* or interceptor the opportunity to release any resources before being thrown away.
*/
public void release();
/**
* Remove the CacheKey with the primaryKey from the map.
* This is used by DeleteObjectQuery and merge.
* This is also an advanced (very) user API.
*/
public Object remove(Object primaryKey, Object object);
/**
* Remove the CacheKey from the map.
*/
public Object remove(CacheKey cacheKey);
/**
* This method will be used to update the max cache size, any objects exceeding the max cache size will
* be remove from the cache. Please note that this does not remove the object from the identityMap, except in
* the case of the CacheIdentityMap.
*/
public void updateMaxSize(int maxSize);
/**
* Set the descriptor that this is the map for.
*/
public void setDescriptor(ClassDescriptor descriptor);
/**
* Update the wrapper object in the CacheKey associated with the given primaryKey,
* this is used for EJB2.
*/
public void setWrapper(Object primaryKey, Object wrapper);
/**
* Update the write lock value of the CacheKey associated with the given primaryKey.
* This is used by UpdateObjectQuery, and is also an advanced (very) user API.
*/
public void setWriteLockValue(Object primaryKey, Object writeLockValue);
public String toString();
}