jaroslav@557: /*
jaroslav@557: * Copyright (c) 2000, 2010, Oracle and/or its affiliates. All rights reserved.
jaroslav@557: * DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER.
jaroslav@557: *
jaroslav@557: * This code is free software; you can redistribute it and/or modify it
jaroslav@557: * under the terms of the GNU General Public License version 2 only, as
jaroslav@557: * published by the Free Software Foundation. Oracle designates this
jaroslav@557: * particular file as subject to the "Classpath" exception as provided
jaroslav@557: * by Oracle in the LICENSE file that accompanied this code.
jaroslav@557: *
jaroslav@557: * This code is distributed in the hope that it will be useful, but WITHOUT
jaroslav@557: * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
jaroslav@557: * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
jaroslav@557: * version 2 for more details (a copy is included in the LICENSE file that
jaroslav@557: * accompanied this code).
jaroslav@557: *
jaroslav@557: * You should have received a copy of the GNU General Public License version
jaroslav@557: * 2 along with this work; if not, write to the Free Software Foundation,
jaroslav@557: * Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA.
jaroslav@557: *
jaroslav@557: * Please contact Oracle, 500 Oracle Parkway, Redwood Shores, CA 94065 USA
jaroslav@557: * or visit www.oracle.com if you need additional information or have any
jaroslav@557: * questions.
jaroslav@557: */
jaroslav@557:
jaroslav@557: package java.util;
jaroslav@557: import java.io.*;
jaroslav@557:
jaroslav@557: /**
jaroslav@557: *
Hash table and linked list implementation of the Map interface,
jaroslav@557: * with predictable iteration order. This implementation differs from
jaroslav@557: * HashMap in that it maintains a doubly-linked list running through
jaroslav@557: * all of its entries. This linked list defines the iteration ordering,
jaroslav@557: * which is normally the order in which keys were inserted into the map
jaroslav@557: * (insertion-order). Note that insertion order is not affected
jaroslav@557: * if a key is re-inserted into the map. (A key k is
jaroslav@557: * reinserted into a map m if m.put(k, v) is invoked when
jaroslav@557: * m.containsKey(k) would return true immediately prior to
jaroslav@557: * the invocation.)
jaroslav@557: *
jaroslav@557: *
This implementation spares its clients from the unspecified, generally
jaroslav@557: * chaotic ordering provided by {@link HashMap} (and {@link Hashtable}),
jaroslav@557: * without incurring the increased cost associated with {@link TreeMap}. It
jaroslav@557: * can be used to produce a copy of a map that has the same order as the
jaroslav@557: * original, regardless of the original map's implementation:
jaroslav@557: *
jaroslav@557: * void foo(Map m) {
jaroslav@557: * Map copy = new LinkedHashMap(m);
jaroslav@557: * ...
jaroslav@557: * }
jaroslav@557: *
jaroslav@557: * This technique is particularly useful if a module takes a map on input,
jaroslav@557: * copies it, and later returns results whose order is determined by that of
jaroslav@557: * the copy. (Clients generally appreciate having things returned in the same
jaroslav@557: * order they were presented.)
jaroslav@557: *
jaroslav@557: * A special {@link #LinkedHashMap(int,float,boolean) constructor} is
jaroslav@557: * provided to create a linked hash map whose order of iteration is the order
jaroslav@557: * in which its entries were last accessed, from least-recently accessed to
jaroslav@557: * most-recently (access-order). This kind of map is well-suited to
jaroslav@557: * building LRU caches. Invoking the put or get method
jaroslav@557: * results in an access to the corresponding entry (assuming it exists after
jaroslav@557: * the invocation completes). The putAll method generates one entry
jaroslav@557: * access for each mapping in the specified map, in the order that key-value
jaroslav@557: * mappings are provided by the specified map's entry set iterator. No
jaroslav@557: * other methods generate entry accesses. In particular, operations on
jaroslav@557: * collection-views do not affect the order of iteration of the backing
jaroslav@557: * map.
jaroslav@557: *
jaroslav@557: *
The {@link #removeEldestEntry(Map.Entry)} method may be overridden to
jaroslav@557: * impose a policy for removing stale mappings automatically when new mappings
jaroslav@557: * are added to the map.
jaroslav@557: *
jaroslav@557: *
This class provides all of the optional Map operations, and
jaroslav@557: * permits null elements. Like HashMap, it provides constant-time
jaroslav@557: * performance for the basic operations (add, contains and
jaroslav@557: * remove), assuming the hash function disperses elements
jaroslav@557: * properly among the buckets. Performance is likely to be just slightly
jaroslav@557: * below that of HashMap, due to the added expense of maintaining the
jaroslav@557: * linked list, with one exception: Iteration over the collection-views
jaroslav@557: * of a LinkedHashMap requires time proportional to the size
jaroslav@557: * of the map, regardless of its capacity. Iteration over a HashMap
jaroslav@557: * is likely to be more expensive, requiring time proportional to its
jaroslav@557: * capacity.
jaroslav@557: *
jaroslav@557: *
A linked hash map has two parameters that affect its performance:
jaroslav@557: * initial capacity and load factor. They are defined precisely
jaroslav@557: * as for HashMap. Note, however, that the penalty for choosing an
jaroslav@557: * excessively high value for initial capacity is less severe for this class
jaroslav@557: * than for HashMap, as iteration times for this class are unaffected
jaroslav@557: * by capacity.
jaroslav@557: *
jaroslav@557: *
Note that this implementation is not synchronized.
jaroslav@557: * If multiple threads access a linked hash map concurrently, and at least
jaroslav@557: * one of the threads modifies the map structurally, it must be
jaroslav@557: * synchronized externally. This is typically accomplished by
jaroslav@557: * synchronizing on some object that naturally encapsulates the map.
jaroslav@557: *
jaroslav@557: * If no such object exists, the map should be "wrapped" using the
jaroslav@557: * {@link Collections#synchronizedMap Collections.synchronizedMap}
jaroslav@557: * method. This is best done at creation time, to prevent accidental
jaroslav@557: * unsynchronized access to the map:
jaroslav@557: * Map m = Collections.synchronizedMap(new LinkedHashMap(...));
jaroslav@557: *
jaroslav@557: * A structural modification is any operation that adds or deletes one or more
jaroslav@557: * mappings or, in the case of access-ordered linked hash maps, affects
jaroslav@557: * iteration order. In insertion-ordered linked hash maps, merely changing
jaroslav@557: * the value associated with a key that is already contained in the map is not
jaroslav@557: * a structural modification. In access-ordered linked hash maps,
jaroslav@557: * merely querying the map with get is a structural
jaroslav@557: * modification.)
jaroslav@557: *
jaroslav@557: * The iterators returned by the iterator method of the collections
jaroslav@557: * returned by all of this class's collection view methods are
jaroslav@557: * fail-fast: if the map is structurally modified at any time after
jaroslav@557: * the iterator is created, in any way except through the iterator's own
jaroslav@557: * remove method, the iterator will throw a {@link
jaroslav@557: * ConcurrentModificationException}. Thus, in the face of concurrent
jaroslav@557: * modification, the iterator fails quickly and cleanly, rather than risking
jaroslav@557: * arbitrary, non-deterministic behavior at an undetermined time in the future.
jaroslav@557: *
jaroslav@557: *
Note that the fail-fast behavior of an iterator cannot be guaranteed
jaroslav@557: * as it is, generally speaking, impossible to make any hard guarantees in the
jaroslav@557: * presence of unsynchronized concurrent modification. Fail-fast iterators
jaroslav@557: * throw ConcurrentModificationException on a best-effort basis.
jaroslav@557: * Therefore, it would be wrong to write a program that depended on this
jaroslav@557: * exception for its correctness: the fail-fast behavior of iterators
jaroslav@557: * should be used only to detect bugs.
jaroslav@557: *
jaroslav@557: *
This class is a member of the
jaroslav@557: *
jaroslav@557: * Java Collections Framework.
jaroslav@557: *
jaroslav@557: * @param the type of keys maintained by this map
jaroslav@557: * @param the type of mapped values
jaroslav@557: *
jaroslav@557: * @author Josh Bloch
jaroslav@557: * @see Object#hashCode()
jaroslav@557: * @see Collection
jaroslav@557: * @see Map
jaroslav@557: * @see HashMap
jaroslav@557: * @see TreeMap
jaroslav@557: * @see Hashtable
jaroslav@557: * @since 1.4
jaroslav@557: */
jaroslav@557:
jaroslav@557: public class LinkedHashMap
jaroslav@557: extends HashMap
jaroslav@557: implements Map
jaroslav@557: {
jaroslav@557:
jaroslav@557: private static final long serialVersionUID = 3801124242820219131L;
jaroslav@557:
jaroslav@557: /**
jaroslav@557: * The head of the doubly linked list.
jaroslav@557: */
jaroslav@557: private transient Entry header;
jaroslav@557:
jaroslav@557: /**
jaroslav@557: * The iteration ordering method for this linked hash map: true
jaroslav@557: * for access-order, false for insertion-order.
jaroslav@557: *
jaroslav@557: * @serial
jaroslav@557: */
jaroslav@557: private final boolean accessOrder;
jaroslav@557:
jaroslav@557: /**
jaroslav@557: * Constructs an empty insertion-ordered LinkedHashMap instance
jaroslav@557: * with the specified initial capacity and load factor.
jaroslav@557: *
jaroslav@557: * @param initialCapacity the initial capacity
jaroslav@557: * @param loadFactor the load factor
jaroslav@557: * @throws IllegalArgumentException if the initial capacity is negative
jaroslav@557: * or the load factor is nonpositive
jaroslav@557: */
jaroslav@557: public LinkedHashMap(int initialCapacity, float loadFactor) {
jaroslav@557: super(initialCapacity, loadFactor);
jaroslav@557: accessOrder = false;
jaroslav@557: }
jaroslav@557:
jaroslav@557: /**
jaroslav@557: * Constructs an empty insertion-ordered LinkedHashMap instance
jaroslav@557: * with the specified initial capacity and a default load factor (0.75).
jaroslav@557: *
jaroslav@557: * @param initialCapacity the initial capacity
jaroslav@557: * @throws IllegalArgumentException if the initial capacity is negative
jaroslav@557: */
jaroslav@557: public LinkedHashMap(int initialCapacity) {
jaroslav@557: super(initialCapacity);
jaroslav@557: accessOrder = false;
jaroslav@557: }
jaroslav@557:
jaroslav@557: /**
jaroslav@557: * Constructs an empty insertion-ordered LinkedHashMap instance
jaroslav@557: * with the default initial capacity (16) and load factor (0.75).
jaroslav@557: */
jaroslav@557: public LinkedHashMap() {
jaroslav@557: super();
jaroslav@557: accessOrder = false;
jaroslav@557: }
jaroslav@557:
jaroslav@557: /**
jaroslav@557: * Constructs an insertion-ordered LinkedHashMap instance with
jaroslav@557: * the same mappings as the specified map. The LinkedHashMap
jaroslav@557: * instance is created with a default load factor (0.75) and an initial
jaroslav@557: * capacity sufficient to hold the mappings in the specified map.
jaroslav@557: *
jaroslav@557: * @param m the map whose mappings are to be placed in this map
jaroslav@557: * @throws NullPointerException if the specified map is null
jaroslav@557: */
jaroslav@557: public LinkedHashMap(Map extends K, ? extends V> m) {
jaroslav@557: super(m);
jaroslav@557: accessOrder = false;
jaroslav@557: }
jaroslav@557:
jaroslav@557: /**
jaroslav@557: * Constructs an empty LinkedHashMap instance with the
jaroslav@557: * specified initial capacity, load factor and ordering mode.
jaroslav@557: *
jaroslav@557: * @param initialCapacity the initial capacity
jaroslav@557: * @param loadFactor the load factor
jaroslav@557: * @param accessOrder the ordering mode - true for
jaroslav@557: * access-order, false for insertion-order
jaroslav@557: * @throws IllegalArgumentException if the initial capacity is negative
jaroslav@557: * or the load factor is nonpositive
jaroslav@557: */
jaroslav@557: public LinkedHashMap(int initialCapacity,
jaroslav@557: float loadFactor,
jaroslav@557: boolean accessOrder) {
jaroslav@557: super(initialCapacity, loadFactor);
jaroslav@557: this.accessOrder = accessOrder;
jaroslav@557: }
jaroslav@557:
jaroslav@557: /**
jaroslav@557: * Called by superclass constructors and pseudoconstructors (clone,
jaroslav@557: * readObject) before any entries are inserted into the map. Initializes
jaroslav@557: * the chain.
jaroslav@557: */
jaroslav@557: void init() {
jaroslav@557: header = new Entry<>(-1, null, null, null);
jaroslav@557: header.before = header.after = header;
jaroslav@557: }
jaroslav@557:
jaroslav@557: /**
jaroslav@557: * Transfers all entries to new table array. This method is called
jaroslav@557: * by superclass resize. It is overridden for performance, as it is
jaroslav@557: * faster to iterate using our linked list.
jaroslav@557: */
jaroslav@557: void transfer(HashMap.Entry[] newTable) {
jaroslav@557: int newCapacity = newTable.length;
jaroslav@557: for (Entry e = header.after; e != header; e = e.after) {
jaroslav@557: int index = indexFor(e.hash, newCapacity);
jaroslav@557: e.next = newTable[index];
jaroslav@557: newTable[index] = e;
jaroslav@557: }
jaroslav@557: }
jaroslav@557:
jaroslav@557:
jaroslav@557: /**
jaroslav@557: * Returns true if this map maps one or more keys to the
jaroslav@557: * specified value.
jaroslav@557: *
jaroslav@557: * @param value value whose presence in this map is to be tested
jaroslav@557: * @return true if this map maps one or more keys to the
jaroslav@557: * specified value
jaroslav@557: */
jaroslav@557: public boolean containsValue(Object value) {
jaroslav@557: // Overridden to take advantage of faster iterator
jaroslav@557: if (value==null) {
jaroslav@557: for (Entry e = header.after; e != header; e = e.after)
jaroslav@557: if (e.value==null)
jaroslav@557: return true;
jaroslav@557: } else {
jaroslav@557: for (Entry e = header.after; e != header; e = e.after)
jaroslav@557: if (value.equals(e.value))
jaroslav@557: return true;
jaroslav@557: }
jaroslav@557: return false;
jaroslav@557: }
jaroslav@557:
jaroslav@557: /**
jaroslav@557: * Returns the value to which the specified key is mapped,
jaroslav@557: * or {@code null} if this map contains no mapping for the key.
jaroslav@557: *
jaroslav@557: * More formally, if this map contains a mapping from a key
jaroslav@557: * {@code k} to a value {@code v} such that {@code (key==null ? k==null :
jaroslav@557: * key.equals(k))}, then this method returns {@code v}; otherwise
jaroslav@557: * it returns {@code null}. (There can be at most one such mapping.)
jaroslav@557: *
jaroslav@557: *
A return value of {@code null} does not necessarily
jaroslav@557: * indicate that the map contains no mapping for the key; it's also
jaroslav@557: * possible that the map explicitly maps the key to {@code null}.
jaroslav@557: * The {@link #containsKey containsKey} operation may be used to
jaroslav@557: * distinguish these two cases.
jaroslav@557: */
jaroslav@557: public V get(Object key) {
jaroslav@557: Entry e = (Entry)getEntry(key);
jaroslav@557: if (e == null)
jaroslav@557: return null;
jaroslav@557: e.recordAccess(this);
jaroslav@557: return e.value;
jaroslav@557: }
jaroslav@557:
jaroslav@557: /**
jaroslav@557: * Removes all of the mappings from this map.
jaroslav@557: * The map will be empty after this call returns.
jaroslav@557: */
jaroslav@557: public void clear() {
jaroslav@557: super.clear();
jaroslav@557: header.before = header.after = header;
jaroslav@557: }
jaroslav@557:
jaroslav@557: /**
jaroslav@557: * LinkedHashMap entry.
jaroslav@557: */
jaroslav@557: private static class Entry extends HashMap.Entry {
jaroslav@557: // These fields comprise the doubly linked list used for iteration.
jaroslav@557: Entry before, after;
jaroslav@557:
jaroslav@557: Entry(int hash, K key, V value, HashMap.Entry next) {
jaroslav@557: super(hash, key, value, next);
jaroslav@557: }
jaroslav@557:
jaroslav@557: /**
jaroslav@557: * Removes this entry from the linked list.
jaroslav@557: */
jaroslav@557: private void remove() {
jaroslav@557: before.after = after;
jaroslav@557: after.before = before;
jaroslav@557: }
jaroslav@557:
jaroslav@557: /**
jaroslav@557: * Inserts this entry before the specified existing entry in the list.
jaroslav@557: */
jaroslav@557: private void addBefore(Entry existingEntry) {
jaroslav@557: after = existingEntry;
jaroslav@557: before = existingEntry.before;
jaroslav@557: before.after = this;
jaroslav@557: after.before = this;
jaroslav@557: }
jaroslav@557:
jaroslav@557: /**
jaroslav@557: * This method is invoked by the superclass whenever the value
jaroslav@557: * of a pre-existing entry is read by Map.get or modified by Map.set.
jaroslav@557: * If the enclosing Map is access-ordered, it moves the entry
jaroslav@557: * to the end of the list; otherwise, it does nothing.
jaroslav@557: */
jaroslav@557: void recordAccess(HashMap m) {
jaroslav@557: LinkedHashMap lm = (LinkedHashMap)m;
jaroslav@557: if (lm.accessOrder) {
jaroslav@557: lm.modCount++;
jaroslav@557: remove();
jaroslav@557: addBefore(lm.header);
jaroslav@557: }
jaroslav@557: }
jaroslav@557:
jaroslav@557: void recordRemoval(HashMap m) {
jaroslav@557: remove();
jaroslav@557: }
jaroslav@557: }
jaroslav@557:
jaroslav@557: private abstract class LinkedHashIterator implements Iterator {
jaroslav@557: Entry nextEntry = header.after;
jaroslav@557: Entry lastReturned = null;
jaroslav@557:
jaroslav@557: /**
jaroslav@557: * The modCount value that the iterator believes that the backing
jaroslav@557: * List should have. If this expectation is violated, the iterator
jaroslav@557: * has detected concurrent modification.
jaroslav@557: */
jaroslav@557: int expectedModCount = modCount;
jaroslav@557:
jaroslav@557: public boolean hasNext() {
jaroslav@557: return nextEntry != header;
jaroslav@557: }
jaroslav@557:
jaroslav@557: public void remove() {
jaroslav@557: if (lastReturned == null)
jaroslav@557: throw new IllegalStateException();
jaroslav@557: if (modCount != expectedModCount)
jaroslav@557: throw new ConcurrentModificationException();
jaroslav@557:
jaroslav@557: LinkedHashMap.this.remove(lastReturned.key);
jaroslav@557: lastReturned = null;
jaroslav@557: expectedModCount = modCount;
jaroslav@557: }
jaroslav@557:
jaroslav@557: Entry nextEntry() {
jaroslav@557: if (modCount != expectedModCount)
jaroslav@557: throw new ConcurrentModificationException();
jaroslav@557: if (nextEntry == header)
jaroslav@557: throw new NoSuchElementException();
jaroslav@557:
jaroslav@557: Entry e = lastReturned = nextEntry;
jaroslav@557: nextEntry = e.after;
jaroslav@557: return e;
jaroslav@557: }
jaroslav@557: }
jaroslav@557:
jaroslav@557: private class KeyIterator extends LinkedHashIterator {
jaroslav@557: public K next() { return nextEntry().getKey(); }
jaroslav@557: }
jaroslav@557:
jaroslav@557: private class ValueIterator extends LinkedHashIterator {
jaroslav@557: public V next() { return nextEntry().value; }
jaroslav@557: }
jaroslav@557:
jaroslav@557: private class EntryIterator extends LinkedHashIterator> {
jaroslav@557: public Map.Entry next() { return nextEntry(); }
jaroslav@557: }
jaroslav@557:
jaroslav@557: // These Overrides alter the behavior of superclass view iterator() methods
jaroslav@557: Iterator newKeyIterator() { return new KeyIterator(); }
jaroslav@557: Iterator newValueIterator() { return new ValueIterator(); }
jaroslav@557: Iterator> newEntryIterator() { return new EntryIterator(); }
jaroslav@557:
jaroslav@557: /**
jaroslav@557: * This override alters behavior of superclass put method. It causes newly
jaroslav@557: * allocated entry to get inserted at the end of the linked list and
jaroslav@557: * removes the eldest entry if appropriate.
jaroslav@557: */
jaroslav@557: void addEntry(int hash, K key, V value, int bucketIndex) {
jaroslav@557: createEntry(hash, key, value, bucketIndex);
jaroslav@557:
jaroslav@557: // Remove eldest entry if instructed, else grow capacity if appropriate
jaroslav@557: Entry eldest = header.after;
jaroslav@557: if (removeEldestEntry(eldest)) {
jaroslav@557: removeEntryForKey(eldest.key);
jaroslav@557: } else {
jaroslav@557: if (size >= threshold)
jaroslav@557: resize(2 * table.length);
jaroslav@557: }
jaroslav@557: }
jaroslav@557:
jaroslav@557: /**
jaroslav@557: * This override differs from addEntry in that it doesn't resize the
jaroslav@557: * table or remove the eldest entry.
jaroslav@557: */
jaroslav@557: void createEntry(int hash, K key, V value, int bucketIndex) {
jaroslav@557: HashMap.Entry old = table[bucketIndex];
jaroslav@557: Entry e = new Entry<>(hash, key, value, old);
jaroslav@557: table[bucketIndex] = e;
jaroslav@557: e.addBefore(header);
jaroslav@557: size++;
jaroslav@557: }
jaroslav@557:
jaroslav@557: /**
jaroslav@557: * Returns true if this map should remove its eldest entry.
jaroslav@557: * This method is invoked by put and putAll after
jaroslav@557: * inserting a new entry into the map. It provides the implementor
jaroslav@557: * with the opportunity to remove the eldest entry each time a new one
jaroslav@557: * is added. This is useful if the map represents a cache: it allows
jaroslav@557: * the map to reduce memory consumption by deleting stale entries.
jaroslav@557: *
jaroslav@557: * Sample use: this override will allow the map to grow up to 100
jaroslav@557: * entries and then delete the eldest entry each time a new entry is
jaroslav@557: * added, maintaining a steady state of 100 entries.
jaroslav@557: *
jaroslav@557: * private static final int MAX_ENTRIES = 100;
jaroslav@557: *
jaroslav@557: * protected boolean removeEldestEntry(Map.Entry eldest) {
jaroslav@557: * return size() > MAX_ENTRIES;
jaroslav@557: * }
jaroslav@557: *
jaroslav@557: *
jaroslav@557: * This method typically does not modify the map in any way,
jaroslav@557: * instead allowing the map to modify itself as directed by its
jaroslav@557: * return value. It is permitted for this method to modify
jaroslav@557: * the map directly, but if it does so, it must return
jaroslav@557: * false (indicating that the map should not attempt any
jaroslav@557: * further modification). The effects of returning true
jaroslav@557: * after modifying the map from within this method are unspecified.
jaroslav@557: *
jaroslav@557: *
This implementation merely returns false (so that this
jaroslav@557: * map acts like a normal map - the eldest element is never removed).
jaroslav@557: *
jaroslav@557: * @param eldest The least recently inserted entry in the map, or if
jaroslav@557: * this is an access-ordered map, the least recently accessed
jaroslav@557: * entry. This is the entry that will be removed it this
jaroslav@557: * method returns true. If the map was empty prior
jaroslav@557: * to the put or putAll invocation resulting
jaroslav@557: * in this invocation, this will be the entry that was just
jaroslav@557: * inserted; in other words, if the map contains a single
jaroslav@557: * entry, the eldest entry is also the newest.
jaroslav@557: * @return true if the eldest entry should be removed
jaroslav@557: * from the map; false if it should be retained.
jaroslav@557: */
jaroslav@557: protected boolean removeEldestEntry(Map.Entry eldest) {
jaroslav@557: return false;
jaroslav@557: }
jaroslav@557: }