1291 lines
49 KiB
Java
1291 lines
49 KiB
Java
/*
|
|
* Copyright (c) 1997, 2023, 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.util;
|
|
|
|
import java.util.function.Consumer;
|
|
import java.util.function.BiConsumer;
|
|
import java.util.function.BiFunction;
|
|
import java.io.IOException;
|
|
import java.util.function.Function;
|
|
|
|
// Android-added: Note about spliterator order b/33945212 in Android N
|
|
/**
|
|
* <p>Hash table and linked list implementation of the {@code Map} interface,
|
|
* with well-defined encounter order. This implementation differs from
|
|
* {@code HashMap} in that it maintains a doubly-linked list running through all of
|
|
* its entries. This linked list defines the encounter order (the order of iteration),
|
|
* which is normally the order in which keys were inserted into the map
|
|
* (<i>insertion-order</i>). The least recently inserted entry (the eldest) is
|
|
* first, and the youngest entry is last. Note that encounter order is not affected
|
|
* if a key is <i>re-inserted</i> into the map with the {@code put} method. (A key
|
|
* {@code k} is reinserted into a map {@code m} if {@code m.put(k, v)} is invoked when
|
|
* {@code m.containsKey(k)} would return {@code true} immediately prior to
|
|
* the invocation.) The reverse-ordered view of this map is in the opposite order, with
|
|
* the youngest entry appearing first and the eldest entry appearing last.
|
|
* The encounter order of entries already in the map can be changed by using
|
|
* the {@link #putFirst putFirst} and {@link #putLast putLast} methods.
|
|
*
|
|
* <p>This implementation spares its clients from the unspecified, generally
|
|
* chaotic ordering provided by {@link HashMap} (and {@link Hashtable}),
|
|
* without incurring the increased cost associated with {@link TreeMap}. It
|
|
* can be used to produce a copy of a map that has the same order as the
|
|
* original, regardless of the original map's implementation:
|
|
* <pre>{@code
|
|
* void foo(Map<String, Integer> m) {
|
|
* Map<String, Integer> copy = new LinkedHashMap<>(m);
|
|
* ...
|
|
* }
|
|
* }</pre>
|
|
* This technique is particularly useful if a module takes a map on input,
|
|
* copies it, and later returns results whose order is determined by that of
|
|
* the copy. (Clients generally appreciate having things returned in the same
|
|
* order they were presented.)
|
|
*
|
|
* <p>A special {@link #LinkedHashMap(int,float,boolean) constructor} is
|
|
* provided to create a linked hash map whose encounter order is the order
|
|
* in which its entries were last accessed, from least-recently accessed to
|
|
* most-recently (<i>access-order</i>). This kind of map is well-suited to
|
|
* building LRU caches. Invoking the {@code put}, {@code putIfAbsent},
|
|
* {@code get}, {@code getOrDefault}, {@code compute}, {@code computeIfAbsent},
|
|
* {@code computeIfPresent}, or {@code merge} methods results
|
|
* in an access to the corresponding entry (assuming it exists after the
|
|
* invocation completes). The {@code replace} methods only result in an access
|
|
* of the entry if the value is replaced. The {@code putAll} method generates one
|
|
* entry access for each mapping in the specified map, in the order that
|
|
* key-value mappings are provided by the specified map's entry set iterator.
|
|
* <i>No other methods generate entry accesses.</i> Invoking these methods on the
|
|
* reversed view generates accesses to entries on the backing map. Note that in the
|
|
* reversed view, an access to an entry moves it first in encounter order.
|
|
* Explicit-positioning methods such as {@code putFirst} or {@code lastEntry}, whether on
|
|
* the map or on its reverse-ordered view, perform the positioning operation and
|
|
* do not generate entry accesses. Operations on the {@code keySet}, {@code values},
|
|
* and {@code entrySet} views or on their sequenced counterparts do <i>not</i> affect
|
|
* the encounter order of the backing map.
|
|
*
|
|
* <p>The {@link #removeEldestEntry(Map.Entry)} method may be overridden to
|
|
* impose a policy for removing stale mappings automatically when new mappings
|
|
* are added to the map. Alternatively, since the "eldest" entry is the first
|
|
* entry in encounter order, programs can inspect and remove stale mappings through
|
|
* use of the {@link #firstEntry firstEntry} and {@link #pollFirstEntry pollFirstEntry}
|
|
* methods.
|
|
*
|
|
* <p>This class provides all of the optional {@code Map} and {@code SequencedMap} operations,
|
|
* and it permits null elements. Like {@code HashMap}, it provides constant-time
|
|
* performance for the basic operations ({@code add}, {@code contains} and
|
|
* {@code remove}), assuming the hash function disperses elements
|
|
* properly among the buckets. Performance is likely to be just slightly
|
|
* below that of {@code HashMap}, due to the added expense of maintaining the
|
|
* linked list, with one exception: Iteration over the collection-views
|
|
* of a {@code LinkedHashMap} requires time proportional to the <i>size</i>
|
|
* of the map, regardless of its capacity. Iteration over a {@code HashMap}
|
|
* is likely to be more expensive, requiring time proportional to its
|
|
* <i>capacity</i>.
|
|
*
|
|
* <p>A linked hash map has two parameters that affect its performance:
|
|
* <i>initial capacity</i> and <i>load factor</i>. They are defined precisely
|
|
* as for {@code HashMap}. Note, however, that the penalty for choosing an
|
|
* excessively high value for initial capacity is less severe for this class
|
|
* than for {@code HashMap}, as iteration times for this class are unaffected
|
|
* by capacity.
|
|
*
|
|
* <p><strong>Note that this implementation is not synchronized.</strong>
|
|
* If multiple threads access a linked hash map concurrently, and at least
|
|
* one of the threads modifies the map structurally, it <em>must</em> be
|
|
* synchronized externally. This is typically accomplished by
|
|
* synchronizing on some object that naturally encapsulates the map.
|
|
*
|
|
* If no such object exists, the map should be "wrapped" using the
|
|
* {@link Collections#synchronizedMap Collections.synchronizedMap}
|
|
* method. This is best done at creation time, to prevent accidental
|
|
* unsynchronized access to the map:<pre>
|
|
* Map m = Collections.synchronizedMap(new LinkedHashMap(...));</pre>
|
|
*
|
|
* A structural modification is any operation that adds or deletes one or more
|
|
* mappings or, in the case of access-ordered linked hash maps, affects
|
|
* iteration order. In insertion-ordered linked hash maps, merely changing
|
|
* the value associated with a key that is already contained in the map is not
|
|
* a structural modification. <strong>In access-ordered linked hash maps,
|
|
* merely querying the map with {@code get} is a structural modification.
|
|
* </strong>)
|
|
*
|
|
* <p>The iterators returned by the {@code iterator} method of the collections
|
|
* returned by all of this class's collection view methods are
|
|
* <em>fail-fast</em>: if the map is structurally modified at any time after
|
|
* the iterator is created, in any way except through the iterator's own
|
|
* {@code remove} method, the iterator will throw a {@link
|
|
* ConcurrentModificationException}. Thus, in the face of concurrent
|
|
* modification, the iterator fails quickly and cleanly, rather than risking
|
|
* arbitrary, non-deterministic behavior at an undetermined time in the future.
|
|
*
|
|
* <p>Note that the fail-fast behavior of an iterator cannot be guaranteed
|
|
* as it is, generally speaking, impossible to make any hard guarantees in the
|
|
* presence of unsynchronized concurrent modification. Fail-fast iterators
|
|
* throw {@code ConcurrentModificationException} on a best-effort basis.
|
|
* Therefore, it would be wrong to write a program that depended on this
|
|
* exception for its correctness: <i>the fail-fast behavior of iterators
|
|
* should be used only to detect bugs.</i>
|
|
*
|
|
* <p>The spliterators returned by the spliterator method of the collections
|
|
* returned by all of this class's collection view methods are
|
|
* <em><a href="Spliterator.html#binding">late-binding</a></em>,
|
|
* <em>fail-fast</em>, and additionally report {@link Spliterator#ORDERED}.
|
|
* <em>Note</em>: The implementation of these spliterators in Android Nougat
|
|
* (API levels 24 and 25) uses the wrong order (inconsistent with the
|
|
* iterators, which use the correct order), despite reporting
|
|
* {@link Spliterator#ORDERED}. You may use the following code fragments
|
|
* to obtain a correctly ordered Spliterator on API level 24 and 25:
|
|
* <ul>
|
|
* <li>For a Collection view {@code c = lhm.keySet()},
|
|
* {@code c = lhm.entrySet()} or {@code c = lhm.values()}, use
|
|
* {@code java.util.Spliterators.spliterator(c, c.spliterator().characteristics())}
|
|
* instead of {@code c.spliterator()}.
|
|
* <li>Instead of {@code c.stream()} or {@code c.parallelStream()}, use
|
|
* {@code java.util.stream.StreamSupport.stream(spliterator, false)}
|
|
* to construct a (nonparallel) {@link java.util.stream.Stream} from
|
|
* such a {@code Spliterator}.
|
|
* </ul>
|
|
* Note that these workarounds are only suggested where {@code lhm} is a
|
|
* {@code LinkedHashMap}.
|
|
*
|
|
* <p>This class is a member of the
|
|
* <a href="{@docRoot}/java.base/java/util/package-summary.html#CollectionsFramework">
|
|
* Java Collections Framework</a>.
|
|
*
|
|
* @implNote
|
|
* The spliterators returned by the spliterator method of the collections
|
|
* returned by all of this class's collection view methods are created from
|
|
* the iterators of the corresponding collections.
|
|
*
|
|
* @param <K> the type of keys maintained by this map
|
|
* @param <V> the type of mapped values
|
|
*
|
|
* @author Josh Bloch
|
|
* @see Object#hashCode()
|
|
* @see Collection
|
|
* @see Map
|
|
* @see HashMap
|
|
* @see TreeMap
|
|
* @see Hashtable
|
|
* @since 1.4
|
|
*/
|
|
public class LinkedHashMap<K,V>
|
|
extends HashMap<K,V>
|
|
implements SequencedMap<K,V>, Map<K, V>
|
|
{
|
|
|
|
/*
|
|
* Implementation note. A previous version of this class was
|
|
* internally structured a little differently. Because superclass
|
|
* HashMap now uses trees for some of its nodes, class
|
|
* LinkedHashMap.Entry is now treated as intermediary node class
|
|
* that can also be converted to tree form. The name of this
|
|
* class, LinkedHashMap.Entry, is confusing in several ways in its
|
|
* current context, but cannot be changed. Otherwise, even though
|
|
* it is not exported outside this package, some existing source
|
|
* code is known to have relied on a symbol resolution corner case
|
|
* rule in calls to removeEldestEntry that suppressed compilation
|
|
* errors due to ambiguous usages. So, we keep the name to
|
|
* preserve unmodified compilability.
|
|
*
|
|
* The changes in node classes also require using two fields
|
|
* (head, tail) rather than a pointer to a header node to maintain
|
|
* the doubly-linked before/after list. This class also
|
|
* previously used a different style of callback methods upon
|
|
* access, insertion, and removal.
|
|
*/
|
|
|
|
/**
|
|
* HashMap.Node subclass for normal LinkedHashMap entries.
|
|
*/
|
|
static class Entry<K,V> extends HashMap.Node<K,V> {
|
|
Entry<K,V> before, after;
|
|
Entry(int hash, K key, V value, Node<K,V> next) {
|
|
super(hash, key, value, next);
|
|
}
|
|
}
|
|
|
|
@java.io.Serial
|
|
private static final long serialVersionUID = 3801124242820219131L;
|
|
|
|
/**
|
|
* The head (eldest) of the doubly linked list.
|
|
*/
|
|
transient LinkedHashMap.Entry<K,V> head;
|
|
|
|
/**
|
|
* The tail (youngest) of the doubly linked list.
|
|
*/
|
|
transient LinkedHashMap.Entry<K,V> tail;
|
|
|
|
/**
|
|
* The iteration ordering method for this linked hash map: {@code true}
|
|
* for access-order, {@code false} for insertion-order.
|
|
*
|
|
* @serial
|
|
*/
|
|
final boolean accessOrder;
|
|
|
|
// internal utilities
|
|
|
|
// link at the end of list
|
|
private void linkNodeAtEnd(LinkedHashMap.Entry<K,V> p) {
|
|
if (putMode == PUT_FIRST) {
|
|
LinkedHashMap.Entry<K,V> first = head;
|
|
head = p;
|
|
if (first == null)
|
|
tail = p;
|
|
else {
|
|
p.after = first;
|
|
first.before = p;
|
|
}
|
|
} else {
|
|
LinkedHashMap.Entry<K,V> last = tail;
|
|
tail = p;
|
|
if (last == null)
|
|
head = p;
|
|
else {
|
|
p.before = last;
|
|
last.after = p;
|
|
}
|
|
}
|
|
}
|
|
|
|
// apply src's links to dst
|
|
private void transferLinks(LinkedHashMap.Entry<K,V> src,
|
|
LinkedHashMap.Entry<K,V> dst) {
|
|
LinkedHashMap.Entry<K,V> b = dst.before = src.before;
|
|
LinkedHashMap.Entry<K,V> a = dst.after = src.after;
|
|
if (b == null)
|
|
head = dst;
|
|
else
|
|
b.after = dst;
|
|
if (a == null)
|
|
tail = dst;
|
|
else
|
|
a.before = dst;
|
|
}
|
|
|
|
// overrides of HashMap hook methods
|
|
|
|
void reinitialize() {
|
|
super.reinitialize();
|
|
head = tail = null;
|
|
}
|
|
|
|
Node<K,V> newNode(int hash, K key, V value, Node<K,V> e) {
|
|
LinkedHashMap.Entry<K,V> p =
|
|
new LinkedHashMap.Entry<>(hash, key, value, e);
|
|
linkNodeAtEnd(p);
|
|
return p;
|
|
}
|
|
|
|
Node<K,V> replacementNode(Node<K,V> p, Node<K,V> next) {
|
|
LinkedHashMap.Entry<K,V> q = (LinkedHashMap.Entry<K,V>)p;
|
|
LinkedHashMap.Entry<K,V> t =
|
|
new LinkedHashMap.Entry<>(q.hash, q.key, q.value, next);
|
|
transferLinks(q, t);
|
|
return t;
|
|
}
|
|
|
|
TreeNode<K,V> newTreeNode(int hash, K key, V value, Node<K,V> next) {
|
|
TreeNode<K,V> p = new TreeNode<>(hash, key, value, next);
|
|
linkNodeAtEnd(p);
|
|
return p;
|
|
}
|
|
|
|
TreeNode<K,V> replacementTreeNode(Node<K,V> p, Node<K,V> next) {
|
|
LinkedHashMap.Entry<K,V> q = (LinkedHashMap.Entry<K,V>)p;
|
|
TreeNode<K,V> t = new TreeNode<>(q.hash, q.key, q.value, next);
|
|
transferLinks(q, t);
|
|
return t;
|
|
}
|
|
|
|
void afterNodeRemoval(Node<K,V> e) { // unlink
|
|
LinkedHashMap.Entry<K,V> p =
|
|
(LinkedHashMap.Entry<K,V>)e, b = p.before, a = p.after;
|
|
p.before = p.after = null;
|
|
if (b == null)
|
|
head = a;
|
|
else
|
|
b.after = a;
|
|
if (a == null)
|
|
tail = b;
|
|
else
|
|
a.before = b;
|
|
}
|
|
|
|
void afterNodeInsertion(boolean evict) { // possibly remove eldest
|
|
LinkedHashMap.Entry<K,V> first;
|
|
if (evict && (first = head) != null && removeEldestEntry(first)) {
|
|
K key = first.key;
|
|
removeNode(hash(key), key, null, false, true);
|
|
}
|
|
}
|
|
|
|
static final int PUT_NORM = 0;
|
|
static final int PUT_FIRST = 1;
|
|
static final int PUT_LAST = 2;
|
|
transient int putMode = PUT_NORM;
|
|
|
|
// Called after update, but not after insertion
|
|
void afterNodeAccess(Node<K,V> e) {
|
|
LinkedHashMap.Entry<K,V> last;
|
|
LinkedHashMap.Entry<K,V> first;
|
|
if ((putMode == PUT_LAST || (putMode == PUT_NORM && accessOrder)) && (last = tail) != e) {
|
|
// move node to last
|
|
LinkedHashMap.Entry<K,V> p =
|
|
(LinkedHashMap.Entry<K,V>)e, b = p.before, a = p.after;
|
|
p.after = null;
|
|
if (b == null)
|
|
head = a;
|
|
else
|
|
b.after = a;
|
|
if (a != null)
|
|
a.before = b;
|
|
else
|
|
last = b;
|
|
if (last == null)
|
|
head = p;
|
|
else {
|
|
p.before = last;
|
|
last.after = p;
|
|
}
|
|
tail = p;
|
|
++modCount;
|
|
} else if (putMode == PUT_FIRST && (first = head) != e) {
|
|
// move node to first
|
|
LinkedHashMap.Entry<K,V> p =
|
|
(LinkedHashMap.Entry<K,V>)e, b = p.before, a = p.after;
|
|
p.before = null;
|
|
if (a == null)
|
|
tail = b;
|
|
else
|
|
a.before = b;
|
|
if (b != null)
|
|
b.after = a;
|
|
else
|
|
first = a;
|
|
if (first == null)
|
|
tail = p;
|
|
else {
|
|
p.after = first;
|
|
first.before = p;
|
|
}
|
|
head = p;
|
|
++modCount;
|
|
}
|
|
}
|
|
|
|
/**
|
|
* {@inheritDoc}
|
|
* <p>
|
|
* If this map already contains a mapping for this key, the mapping is relocated if necessary
|
|
* so that it is first in encounter order.
|
|
*
|
|
* @since 21
|
|
*/
|
|
public V putFirst(K k, V v) {
|
|
try {
|
|
putMode = PUT_FIRST;
|
|
return this.put(k, v);
|
|
} finally {
|
|
putMode = PUT_NORM;
|
|
}
|
|
}
|
|
|
|
/**
|
|
* {@inheritDoc}
|
|
* <p>
|
|
* If this map already contains a mapping for this key, the mapping is relocated if necessary
|
|
* so that it is last in encounter order.
|
|
*
|
|
* @since 21
|
|
*/
|
|
public V putLast(K k, V v) {
|
|
try {
|
|
putMode = PUT_LAST;
|
|
return this.put(k, v);
|
|
} finally {
|
|
putMode = PUT_NORM;
|
|
}
|
|
}
|
|
|
|
void internalWriteEntries(java.io.ObjectOutputStream s) throws IOException {
|
|
for (LinkedHashMap.Entry<K,V> e = head; e != null; e = e.after) {
|
|
s.writeObject(e.key);
|
|
s.writeObject(e.value);
|
|
}
|
|
}
|
|
|
|
/**
|
|
* Constructs an empty insertion-ordered {@code LinkedHashMap} instance
|
|
* with the specified initial capacity and load factor.
|
|
*
|
|
* @apiNote
|
|
* To create a {@code LinkedHashMap} with an initial capacity that accommodates
|
|
* an expected number of mappings, use {@link #newLinkedHashMap(int) newLinkedHashMap}.
|
|
*
|
|
* @param initialCapacity the initial capacity
|
|
* @param loadFactor the load factor
|
|
* @throws IllegalArgumentException if the initial capacity is negative
|
|
* or the load factor is nonpositive
|
|
*/
|
|
public LinkedHashMap(int initialCapacity, float loadFactor) {
|
|
super(initialCapacity, loadFactor);
|
|
accessOrder = false;
|
|
}
|
|
|
|
/**
|
|
* Constructs an empty insertion-ordered {@code LinkedHashMap} instance
|
|
* with the specified initial capacity and a default load factor (0.75).
|
|
*
|
|
* @apiNote
|
|
* To create a {@code LinkedHashMap} with an initial capacity that accommodates
|
|
* an expected number of mappings, use {@link #newLinkedHashMap(int) newLinkedHashMap}.
|
|
*
|
|
* @param initialCapacity the initial capacity
|
|
* @throws IllegalArgumentException if the initial capacity is negative
|
|
*/
|
|
public LinkedHashMap(int initialCapacity) {
|
|
super(initialCapacity);
|
|
accessOrder = false;
|
|
}
|
|
|
|
/**
|
|
* Constructs an empty insertion-ordered {@code LinkedHashMap} instance
|
|
* with the default initial capacity (16) and load factor (0.75).
|
|
*/
|
|
public LinkedHashMap() {
|
|
super();
|
|
accessOrder = false;
|
|
}
|
|
|
|
/**
|
|
* Constructs an insertion-ordered {@code LinkedHashMap} instance with
|
|
* the same mappings as the specified map. The {@code LinkedHashMap}
|
|
* instance is created with a default load factor (0.75) and an initial
|
|
* capacity sufficient to hold the mappings in the specified map.
|
|
*
|
|
* @param m the map whose mappings are to be placed in this map
|
|
* @throws NullPointerException if the specified map is null
|
|
*/
|
|
public LinkedHashMap(Map<? extends K, ? extends V> m) {
|
|
super();
|
|
accessOrder = false;
|
|
putMapEntries(m, false);
|
|
}
|
|
|
|
/**
|
|
* Constructs an empty {@code LinkedHashMap} instance with the
|
|
* specified initial capacity, load factor and ordering mode.
|
|
*
|
|
* @param initialCapacity the initial capacity
|
|
* @param loadFactor the load factor
|
|
* @param accessOrder the ordering mode - {@code true} for
|
|
* access-order, {@code false} for insertion-order
|
|
* @throws IllegalArgumentException if the initial capacity is negative
|
|
* or the load factor is nonpositive
|
|
*/
|
|
public LinkedHashMap(int initialCapacity,
|
|
float loadFactor,
|
|
boolean accessOrder) {
|
|
super(initialCapacity, loadFactor);
|
|
this.accessOrder = accessOrder;
|
|
}
|
|
|
|
|
|
/**
|
|
* Returns {@code true} if this map maps one or more keys to the
|
|
* specified value.
|
|
*
|
|
* @param value value whose presence in this map is to be tested
|
|
* @return {@code true} if this map maps one or more keys to the
|
|
* specified value
|
|
*/
|
|
public boolean containsValue(Object value) {
|
|
for (LinkedHashMap.Entry<K,V> e = head; e != null; e = e.after) {
|
|
V v = e.value;
|
|
if (v == value || (value != null && value.equals(v)))
|
|
return true;
|
|
}
|
|
return false;
|
|
}
|
|
|
|
/**
|
|
* Returns the value to which the specified key is mapped,
|
|
* or {@code null} if this map contains no mapping for the key.
|
|
*
|
|
* <p>More formally, if this map contains a mapping from a key
|
|
* {@code k} to a value {@code v} such that {@code (key==null ? k==null :
|
|
* key.equals(k))}, then this method returns {@code v}; otherwise
|
|
* it returns {@code null}. (There can be at most one such mapping.)
|
|
*
|
|
* <p>A return value of {@code null} does not <i>necessarily</i>
|
|
* indicate that the map contains no mapping for the key; it's also
|
|
* possible that the map explicitly maps the key to {@code null}.
|
|
* The {@link #containsKey containsKey} operation may be used to
|
|
* distinguish these two cases.
|
|
*/
|
|
public V get(Object key) {
|
|
Node<K,V> e;
|
|
if ((e = getNode(key)) == null)
|
|
return null;
|
|
if (accessOrder)
|
|
afterNodeAccess(e);
|
|
return e.value;
|
|
}
|
|
|
|
/**
|
|
* {@inheritDoc}
|
|
*/
|
|
public V getOrDefault(Object key, V defaultValue) {
|
|
Node<K,V> e;
|
|
if ((e = getNode(key)) == null)
|
|
return defaultValue;
|
|
if (accessOrder)
|
|
afterNodeAccess(e);
|
|
return e.value;
|
|
}
|
|
|
|
/**
|
|
* {@inheritDoc}
|
|
*/
|
|
public void clear() {
|
|
super.clear();
|
|
head = tail = null;
|
|
}
|
|
|
|
// Android-added: eldest(), for internal use in LRU caches
|
|
/**
|
|
* Returns the eldest entry in the map, or {@code null} if the map is empty.
|
|
*
|
|
* @return eldest entry in the map, or {@code null} if the map is empty
|
|
*
|
|
* @hide
|
|
*/
|
|
public Map.Entry<K, V> eldest() {
|
|
return head;
|
|
}
|
|
|
|
/**
|
|
* Returns {@code true} if this map should remove its eldest entry.
|
|
* This method is invoked by {@code put} and {@code putAll} after
|
|
* inserting a new entry into the map. It provides the implementor
|
|
* with the opportunity to remove the eldest entry each time a new one
|
|
* is added. This is useful if the map represents a cache: it allows
|
|
* the map to reduce memory consumption by deleting stale entries.
|
|
*
|
|
* <p>Sample use: this override will allow the map to grow up to 100
|
|
* entries and then delete the eldest entry each time a new entry is
|
|
* added, maintaining a steady state of 100 entries.
|
|
* <pre>
|
|
* private static final int MAX_ENTRIES = 100;
|
|
*
|
|
* protected boolean removeEldestEntry(Map.Entry eldest) {
|
|
* return size() > MAX_ENTRIES;
|
|
* }
|
|
* </pre>
|
|
*
|
|
* <p>This method typically does not modify the map in any way,
|
|
* instead allowing the map to modify itself as directed by its
|
|
* return value. It <i>is</i> permitted for this method to modify
|
|
* the map directly, but if it does so, it <i>must</i> return
|
|
* {@code false} (indicating that the map should not attempt any
|
|
* further modification). The effects of returning {@code true}
|
|
* after modifying the map from within this method are unspecified.
|
|
*
|
|
* <p>This implementation merely returns {@code false} (so that this
|
|
* map acts like a normal map - the eldest element is never removed).
|
|
*
|
|
* @param eldest The least recently inserted entry in the map, or if
|
|
* this is an access-ordered map, the least recently accessed
|
|
* entry. This is the entry that will be removed if this
|
|
* method returns {@code true}. If the map was empty prior
|
|
* to the {@code put} or {@code putAll} invocation resulting
|
|
* in this invocation, this will be the entry that was just
|
|
* inserted; in other words, if the map contains a single
|
|
* entry, the eldest entry is also the newest.
|
|
* @return {@code true} if the eldest entry should be removed
|
|
* from the map; {@code false} if it should be retained.
|
|
*/
|
|
protected boolean removeEldestEntry(Map.Entry<K,V> eldest) {
|
|
return false;
|
|
}
|
|
|
|
/**
|
|
* Returns a {@link Set} view of the keys contained in this map. The encounter
|
|
* order of the keys in the view matches the encounter order of mappings of
|
|
* this map. The set is backed by the map, so changes to the map are
|
|
* reflected in the set, and vice-versa. If the map is modified
|
|
* while an iteration over the set is in progress (except through
|
|
* the iterator's own {@code remove} operation), the results of
|
|
* the iteration are undefined. The set supports element removal,
|
|
* which removes the corresponding mapping from the map, via the
|
|
* {@code Iterator.remove}, {@code Set.remove},
|
|
* {@code removeAll}, {@code retainAll}, and {@code clear}
|
|
* operations. It does not support the {@code add} or {@code addAll}
|
|
* operations.
|
|
* Its {@link Spliterator} typically provides faster sequential
|
|
* performance but much poorer parallel performance than that of
|
|
* {@code HashMap}.
|
|
*
|
|
* @return a set view of the keys contained in this map
|
|
*/
|
|
public Set<K> keySet() {
|
|
return sequencedKeySet();
|
|
}
|
|
|
|
/**
|
|
* {@inheritDoc}
|
|
* <p>
|
|
* The returned view has the same characteristics as specified for the view
|
|
* returned by the {@link #keySet keySet} method.
|
|
*
|
|
* @return {@inheritDoc}
|
|
* @since 21
|
|
*/
|
|
public SequencedSet<K> sequencedKeySet() {
|
|
Set<K> ks = keySet;
|
|
if (ks == null) {
|
|
SequencedSet<K> sks = new LinkedKeySet(false);
|
|
keySet = sks;
|
|
return sks;
|
|
} else {
|
|
// The cast should never fail, since the only assignment of non-null to keySet is
|
|
// above, and assignments in AbstractMap and HashMap are in overridden methods.
|
|
return (SequencedSet<K>) ks;
|
|
}
|
|
}
|
|
|
|
static <K1,V1> Node<K1,V1> nsee(Node<K1,V1> node) {
|
|
if (node == null)
|
|
throw new NoSuchElementException();
|
|
else
|
|
return node;
|
|
}
|
|
|
|
final <T> T[] keysToArray(T[] a) {
|
|
return keysToArray(a, false);
|
|
}
|
|
|
|
final <T> T[] keysToArray(T[] a, boolean reversed) {
|
|
Object[] r = a;
|
|
int idx = 0;
|
|
if (reversed) {
|
|
for (LinkedHashMap.Entry<K,V> e = tail; e != null; e = e.before) {
|
|
r[idx++] = e.key;
|
|
}
|
|
} else {
|
|
for (LinkedHashMap.Entry<K,V> e = head; e != null; e = e.after) {
|
|
r[idx++] = e.key;
|
|
}
|
|
}
|
|
return a;
|
|
}
|
|
|
|
final <T> T[] valuesToArray(T[] a, boolean reversed) {
|
|
Object[] r = a;
|
|
int idx = 0;
|
|
if (reversed) {
|
|
for (LinkedHashMap.Entry<K,V> e = tail; e != null; e = e.before) {
|
|
r[idx++] = e.value;
|
|
}
|
|
} else {
|
|
for (LinkedHashMap.Entry<K,V> e = head; e != null; e = e.after) {
|
|
r[idx++] = e.value;
|
|
}
|
|
}
|
|
return a;
|
|
}
|
|
|
|
final class LinkedKeySet extends AbstractSet<K> implements SequencedSet<K> {
|
|
final boolean reversed;
|
|
LinkedKeySet(boolean reversed) { this.reversed = reversed; }
|
|
public final int size() { return size; }
|
|
public final void clear() { LinkedHashMap.this.clear(); }
|
|
public final Iterator<K> iterator() {
|
|
return new LinkedKeyIterator(reversed);
|
|
}
|
|
public final boolean contains(Object o) { return containsKey(o); }
|
|
public final boolean remove(Object key) {
|
|
return removeNode(hash(key), key, null, false, true) != null;
|
|
}
|
|
public final Spliterator<K> spliterator() {
|
|
return Spliterators.spliterator(this, Spliterator.SIZED |
|
|
Spliterator.ORDERED |
|
|
Spliterator.DISTINCT);
|
|
}
|
|
|
|
public Object[] toArray() {
|
|
return keysToArray(new Object[size], reversed);
|
|
}
|
|
|
|
public <T> T[] toArray(T[] a) {
|
|
return keysToArray(prepareArray(a), reversed);
|
|
}
|
|
|
|
public final void forEach(Consumer<? super K> action) {
|
|
if (action == null)
|
|
throw new NullPointerException();
|
|
int mc = modCount;
|
|
if (reversed) {
|
|
// Android-changed: Detect changes to modCount early.
|
|
for (LinkedHashMap.Entry<K,V> e = tail; e != null && modCount == mc; e = e.before)
|
|
action.accept(e.key);
|
|
} else {
|
|
// Android-changed: Detect changes to modCount early.
|
|
for (LinkedHashMap.Entry<K,V> e = head; (e != null && modCount == mc); e = e.after)
|
|
action.accept(e.key);
|
|
}
|
|
if (modCount != mc)
|
|
throw new ConcurrentModificationException();
|
|
}
|
|
public final void addFirst(K k) { throw new UnsupportedOperationException(); }
|
|
public final void addLast(K k) { throw new UnsupportedOperationException(); }
|
|
public final K getFirst() { return nsee(reversed ? tail : head).key; }
|
|
public final K getLast() { return nsee(reversed ? head : tail).key; }
|
|
public final K removeFirst() {
|
|
var node = nsee(reversed ? tail : head);
|
|
removeNode(node.hash, node.key, null, false, false);
|
|
return node.key;
|
|
}
|
|
public final K removeLast() {
|
|
var node = nsee(reversed ? head : tail);
|
|
removeNode(node.hash, node.key, null, false, false);
|
|
return node.key;
|
|
}
|
|
public SequencedSet<K> reversed() {
|
|
if (reversed) {
|
|
return LinkedHashMap.this.sequencedKeySet();
|
|
} else {
|
|
return new LinkedKeySet(true);
|
|
}
|
|
}
|
|
}
|
|
|
|
/**
|
|
* Returns a {@link Collection} view of the values contained in this map. The
|
|
* encounter order of values in the view matches the encounter order of entries in
|
|
* this map. The collection is backed by the map, so changes to the map are
|
|
* reflected in the collection, and vice-versa. If the map is
|
|
* modified while an iteration over the collection is in progress
|
|
* (except through the iterator's own {@code remove} operation),
|
|
* the results of the iteration are undefined. The collection
|
|
* supports element removal, which removes the corresponding
|
|
* mapping from the map, via the {@code Iterator.remove},
|
|
* {@code Collection.remove}, {@code removeAll},
|
|
* {@code retainAll} and {@code clear} operations. It does not
|
|
* support the {@code add} or {@code addAll} operations.
|
|
* Its {@link Spliterator} typically provides faster sequential
|
|
* performance but much poorer parallel performance than that of
|
|
* {@code HashMap}.
|
|
*
|
|
* @return a view of the values contained in this map
|
|
*/
|
|
public Collection<V> values() {
|
|
return sequencedValues();
|
|
}
|
|
|
|
/**
|
|
* {@inheritDoc}
|
|
* <p>
|
|
* The returned view has the same characteristics as specified for the view
|
|
* returned by the {@link #values values} method.
|
|
*
|
|
* @return {@inheritDoc}
|
|
* @since 21
|
|
*/
|
|
public SequencedCollection<V> sequencedValues() {
|
|
Collection<V> vs = values;
|
|
if (vs == null) {
|
|
SequencedCollection<V> svs = new LinkedValues(false);
|
|
values = svs;
|
|
return svs;
|
|
} else {
|
|
// The cast should never fail, since the only assignment of non-null to values is
|
|
// above, and assignments in AbstractMap and HashMap are in overridden methods.
|
|
return (SequencedCollection<V>) vs;
|
|
}
|
|
}
|
|
|
|
final class LinkedValues extends AbstractCollection<V> implements SequencedCollection<V> {
|
|
final boolean reversed;
|
|
LinkedValues(boolean reversed) { this.reversed = reversed; }
|
|
public final int size() { return size; }
|
|
public final void clear() { LinkedHashMap.this.clear(); }
|
|
public final Iterator<V> iterator() {
|
|
return new LinkedValueIterator(reversed);
|
|
}
|
|
public final boolean contains(Object o) { return containsValue(o); }
|
|
public final Spliterator<V> spliterator() {
|
|
return Spliterators.spliterator(this, Spliterator.SIZED |
|
|
Spliterator.ORDERED);
|
|
}
|
|
|
|
public Object[] toArray() {
|
|
return valuesToArray(new Object[size], reversed);
|
|
}
|
|
|
|
public <T> T[] toArray(T[] a) {
|
|
return valuesToArray(prepareArray(a), reversed);
|
|
}
|
|
|
|
public final void forEach(Consumer<? super V> action) {
|
|
if (action == null)
|
|
throw new NullPointerException();
|
|
int mc = modCount;
|
|
if (reversed) {
|
|
// Android-changed: Detect changes to modCount early.
|
|
for (LinkedHashMap.Entry<K,V> e = tail; e != null && modCount == mc; e = e.before)
|
|
action.accept(e.value);
|
|
} else {
|
|
// Android-changed: Detect changes to modCount early.
|
|
for (LinkedHashMap.Entry<K,V> e = head; (e != null && modCount == mc); e = e.after)
|
|
action.accept(e.value);
|
|
}
|
|
if (modCount != mc)
|
|
throw new ConcurrentModificationException();
|
|
}
|
|
public final void addFirst(V v) { throw new UnsupportedOperationException(); }
|
|
public final void addLast(V v) { throw new UnsupportedOperationException(); }
|
|
public final V getFirst() { return nsee(reversed ? tail : head).value; }
|
|
public final V getLast() { return nsee(reversed ? head : tail).value; }
|
|
public final V removeFirst() {
|
|
var node = nsee(reversed ? tail : head);
|
|
removeNode(node.hash, node.key, null, false, false);
|
|
return node.value;
|
|
}
|
|
public final V removeLast() {
|
|
var node = nsee(reversed ? head : tail);
|
|
removeNode(node.hash, node.key, null, false, false);
|
|
return node.value;
|
|
}
|
|
public SequencedCollection<V> reversed() {
|
|
if (reversed) {
|
|
return LinkedHashMap.this.sequencedValues();
|
|
} else {
|
|
return new LinkedValues(true);
|
|
}
|
|
}
|
|
}
|
|
|
|
/**
|
|
* Returns a {@link Set} view of the mappings contained in this map. The encounter
|
|
* order of the view matches the encounter order of entries of this map.
|
|
* The set is backed by the map, so changes to the map are
|
|
* reflected in the set, and vice-versa. If the map is modified
|
|
* while an iteration over the set is in progress (except through
|
|
* the iterator's own {@code remove} operation, or through the
|
|
* {@code setValue} operation on a map entry returned by the
|
|
* iterator) the results of the iteration are undefined. The set
|
|
* supports element removal, which removes the corresponding
|
|
* mapping from the map, via the {@code Iterator.remove},
|
|
* {@code Set.remove}, {@code removeAll}, {@code retainAll} and
|
|
* {@code clear} operations. It does not support the
|
|
* {@code add} or {@code addAll} operations.
|
|
* Its {@link Spliterator} typically provides faster sequential
|
|
* performance but much poorer parallel performance than that of
|
|
* {@code HashMap}.
|
|
*
|
|
* @return a set view of the mappings contained in this map
|
|
*/
|
|
public Set<Map.Entry<K,V>> entrySet() {
|
|
return sequencedEntrySet();
|
|
}
|
|
|
|
/**
|
|
* {@inheritDoc}
|
|
* <p>
|
|
* The returned view has the same characteristics as specified for the view
|
|
* returned by the {@link #entrySet entrySet} method.
|
|
*
|
|
* @return {@inheritDoc}
|
|
* @since 21
|
|
*/
|
|
public SequencedSet<Map.Entry<K, V>> sequencedEntrySet() {
|
|
Set<Map.Entry<K, V>> es = entrySet;
|
|
if (es == null) {
|
|
SequencedSet<Map.Entry<K, V>> ses = new LinkedEntrySet(false);
|
|
entrySet = ses;
|
|
return ses;
|
|
} else {
|
|
// The cast should never fail, since the only assignment of non-null to entrySet is
|
|
// above, and assignments in HashMap are in overridden methods.
|
|
return (SequencedSet<Map.Entry<K, V>>) es;
|
|
}
|
|
}
|
|
|
|
final class LinkedEntrySet extends AbstractSet<Map.Entry<K,V>>
|
|
implements SequencedSet<Map.Entry<K,V>> {
|
|
final boolean reversed;
|
|
LinkedEntrySet(boolean reversed) { this.reversed = reversed; }
|
|
public final int size() { return size; }
|
|
public final void clear() { LinkedHashMap.this.clear(); }
|
|
public final Iterator<Map.Entry<K,V>> iterator() {
|
|
return new LinkedEntryIterator(reversed);
|
|
}
|
|
public final boolean contains(Object o) {
|
|
if (!(o instanceof Map.Entry<?, ?> e))
|
|
return false;
|
|
Object key = e.getKey();
|
|
Node<K,V> candidate = getNode(key);
|
|
return candidate != null && candidate.equals(e);
|
|
}
|
|
public final boolean remove(Object o) {
|
|
if (o instanceof Map.Entry<?, ?> e) {
|
|
Object key = e.getKey();
|
|
Object value = e.getValue();
|
|
return removeNode(hash(key), key, value, true, true) != null;
|
|
}
|
|
return false;
|
|
}
|
|
public final Spliterator<Map.Entry<K,V>> spliterator() {
|
|
return Spliterators.spliterator(this, Spliterator.SIZED |
|
|
Spliterator.ORDERED |
|
|
Spliterator.DISTINCT);
|
|
}
|
|
public final void forEach(Consumer<? super Map.Entry<K,V>> action) {
|
|
if (action == null)
|
|
throw new NullPointerException();
|
|
int mc = modCount;
|
|
if (reversed) {
|
|
// Android-changed: Detect changes to modCount early.
|
|
for (LinkedHashMap.Entry<K,V> e = tail; e != null && mc == modCount; e = e.before)
|
|
action.accept(e);
|
|
} else {
|
|
// Android-changed: Detect changes to modCount early.
|
|
for (LinkedHashMap.Entry<K,V> e = head; (e != null && mc == modCount); e = e.after)
|
|
action.accept(e);
|
|
}
|
|
if (modCount != mc)
|
|
throw new ConcurrentModificationException();
|
|
}
|
|
final Node<K,V> nsee(Node<K,V> e) {
|
|
if (e == null)
|
|
throw new NoSuchElementException();
|
|
else
|
|
return e;
|
|
}
|
|
public final void addFirst(Map.Entry<K,V> e) { throw new UnsupportedOperationException(); }
|
|
public final void addLast(Map.Entry<K,V> e) { throw new UnsupportedOperationException(); }
|
|
public final Map.Entry<K,V> getFirst() { return nsee(reversed ? tail : head); }
|
|
public final Map.Entry<K,V> getLast() { return nsee(reversed ? head : tail); }
|
|
public final Map.Entry<K,V> removeFirst() {
|
|
var node = nsee(reversed ? tail : head);
|
|
removeNode(node.hash, node.key, null, false, false);
|
|
return node;
|
|
}
|
|
public final Map.Entry<K,V> removeLast() {
|
|
var node = nsee(reversed ? head : tail);
|
|
removeNode(node.hash, node.key, null, false, false);
|
|
return node;
|
|
}
|
|
public SequencedSet<Map.Entry<K,V>> reversed() {
|
|
if (reversed) {
|
|
return LinkedHashMap.this.sequencedEntrySet();
|
|
} else {
|
|
return new LinkedEntrySet(true);
|
|
}
|
|
}
|
|
}
|
|
|
|
// Map overrides
|
|
|
|
public void forEach(BiConsumer<? super K, ? super V> action) {
|
|
if (action == null)
|
|
throw new NullPointerException();
|
|
int mc = modCount;
|
|
// Android-changed: Detect changes to modCount early.
|
|
for (LinkedHashMap.Entry<K,V> e = head; modCount == mc && e != null; e = e.after)
|
|
action.accept(e.key, e.value);
|
|
if (modCount != mc)
|
|
throw new ConcurrentModificationException();
|
|
}
|
|
|
|
public void replaceAll(BiFunction<? super K, ? super V, ? extends V> function) {
|
|
if (function == null)
|
|
throw new NullPointerException();
|
|
int mc = modCount;
|
|
// Android-changed: Detect changes to modCount early.
|
|
for (LinkedHashMap.Entry<K,V> e = head; modCount == mc && e != null; e = e.after)
|
|
e.value = function.apply(e.key, e.value);
|
|
if (modCount != mc)
|
|
throw new ConcurrentModificationException();
|
|
}
|
|
|
|
// Iterators
|
|
|
|
abstract class LinkedHashIterator {
|
|
LinkedHashMap.Entry<K,V> next;
|
|
LinkedHashMap.Entry<K,V> current;
|
|
int expectedModCount;
|
|
boolean reversed;
|
|
|
|
LinkedHashIterator(boolean reversed) {
|
|
this.reversed = reversed;
|
|
next = reversed ? tail : head;
|
|
expectedModCount = modCount;
|
|
current = null;
|
|
}
|
|
|
|
public final boolean hasNext() {
|
|
return next != null;
|
|
}
|
|
|
|
final LinkedHashMap.Entry<K,V> nextNode() {
|
|
LinkedHashMap.Entry<K,V> e = next;
|
|
if (modCount != expectedModCount)
|
|
throw new ConcurrentModificationException();
|
|
if (e == null)
|
|
throw new NoSuchElementException();
|
|
current = e;
|
|
next = reversed ? e.before : e.after;
|
|
return e;
|
|
}
|
|
|
|
public final void remove() {
|
|
Node<K,V> p = current;
|
|
if (p == null)
|
|
throw new IllegalStateException();
|
|
if (modCount != expectedModCount)
|
|
throw new ConcurrentModificationException();
|
|
current = null;
|
|
removeNode(p.hash, p.key, null, false, false);
|
|
expectedModCount = modCount;
|
|
}
|
|
}
|
|
|
|
final class LinkedKeyIterator extends LinkedHashIterator
|
|
implements Iterator<K> {
|
|
LinkedKeyIterator(boolean reversed) { super(reversed); }
|
|
public final K next() { return nextNode().getKey(); }
|
|
}
|
|
|
|
final class LinkedValueIterator extends LinkedHashIterator
|
|
implements Iterator<V> {
|
|
LinkedValueIterator(boolean reversed) { super(reversed); }
|
|
public final V next() { return nextNode().value; }
|
|
}
|
|
|
|
final class LinkedEntryIterator extends LinkedHashIterator
|
|
implements Iterator<Map.Entry<K,V>> {
|
|
LinkedEntryIterator(boolean reversed) { super(reversed); }
|
|
public final Map.Entry<K,V> next() { return nextNode(); }
|
|
}
|
|
|
|
/**
|
|
* Creates a new, empty, insertion-ordered LinkedHashMap suitable for the expected number of mappings.
|
|
* The returned map uses the default load factor of 0.75, and its initial capacity is
|
|
* generally large enough so that the expected number of mappings can be added
|
|
* without resizing the map.
|
|
*
|
|
* @param numMappings the expected number of mappings
|
|
* @param <K> the type of keys maintained by the new map
|
|
* @param <V> the type of mapped values
|
|
* @return the newly created map
|
|
* @throws IllegalArgumentException if numMappings is negative
|
|
* @since 19
|
|
*/
|
|
public static <K, V> LinkedHashMap<K, V> newLinkedHashMap(int numMappings) {
|
|
if (numMappings < 0) {
|
|
throw new IllegalArgumentException("Negative number of mappings: " + numMappings);
|
|
}
|
|
return new LinkedHashMap<>(HashMap.calculateHashMapCapacity(numMappings));
|
|
}
|
|
|
|
// Reversed View
|
|
|
|
/**
|
|
* {@inheritDoc}
|
|
* <p>
|
|
* Modifications to the reversed view and its map views are permitted and will be
|
|
* propagated to this map. In addition, modifications to this map will be visible
|
|
* in the reversed view and its map views.
|
|
*
|
|
* @return {@inheritDoc}
|
|
* @since 21
|
|
*/
|
|
public SequencedMap<K, V> reversed() {
|
|
return new ReversedLinkedHashMapView<>(this);
|
|
}
|
|
|
|
static class ReversedLinkedHashMapView<K, V> extends AbstractMap<K, V>
|
|
implements SequencedMap<K, V> {
|
|
final LinkedHashMap<K, V> base;
|
|
|
|
ReversedLinkedHashMapView(LinkedHashMap<K, V> lhm) {
|
|
base = lhm;
|
|
}
|
|
|
|
// Object
|
|
// inherit toString() from AbstractMap; it depends on entrySet()
|
|
|
|
public boolean equals(Object o) {
|
|
return base.equals(o);
|
|
}
|
|
|
|
public int hashCode() {
|
|
return base.hashCode();
|
|
}
|
|
|
|
// Map
|
|
|
|
public int size() {
|
|
return base.size();
|
|
}
|
|
|
|
public boolean isEmpty() {
|
|
return base.isEmpty();
|
|
}
|
|
|
|
public boolean containsKey(Object key) {
|
|
return base.containsKey(key);
|
|
}
|
|
|
|
public boolean containsValue(Object value) {
|
|
return base.containsValue(value);
|
|
}
|
|
|
|
public V get(Object key) {
|
|
return base.get(key);
|
|
}
|
|
|
|
public V put(K key, V value) {
|
|
return base.put(key, value);
|
|
}
|
|
|
|
public V remove(Object key) {
|
|
return base.remove(key);
|
|
}
|
|
|
|
public void putAll(Map<? extends K, ? extends V> m) {
|
|
base.putAll(m);
|
|
}
|
|
|
|
public void clear() {
|
|
base.clear();
|
|
}
|
|
|
|
public Set<K> keySet() {
|
|
return base.sequencedKeySet().reversed();
|
|
}
|
|
|
|
public Collection<V> values() {
|
|
return base.sequencedValues().reversed();
|
|
}
|
|
|
|
public Set<Entry<K, V>> entrySet() {
|
|
return base.sequencedEntrySet().reversed();
|
|
}
|
|
|
|
public V getOrDefault(Object key, V defaultValue) {
|
|
return base.getOrDefault(key, defaultValue);
|
|
}
|
|
|
|
public void forEach(BiConsumer<? super K, ? super V> action) {
|
|
if (action == null)
|
|
throw new NullPointerException();
|
|
int mc = base.modCount;
|
|
for (LinkedHashMap.Entry<K,V> e = base.tail; e != null; e = e.before)
|
|
action.accept(e.key, e.value);
|
|
if (base.modCount != mc)
|
|
throw new ConcurrentModificationException();
|
|
}
|
|
|
|
public void replaceAll(BiFunction<? super K, ? super V, ? extends V> function) {
|
|
if (function == null)
|
|
throw new NullPointerException();
|
|
int mc = base.modCount;
|
|
for (LinkedHashMap.Entry<K,V> e = base.tail; e != null; e = e.before)
|
|
e.value = function.apply(e.key, e.value);
|
|
if (base.modCount != mc)
|
|
throw new ConcurrentModificationException();
|
|
}
|
|
|
|
public V putIfAbsent(K key, V value) {
|
|
return base.putIfAbsent(key, value);
|
|
}
|
|
|
|
public boolean remove(Object key, Object value) {
|
|
return base.remove(key, value);
|
|
}
|
|
|
|
public boolean replace(K key, V oldValue, V newValue) {
|
|
return base.replace(key, oldValue, newValue);
|
|
}
|
|
|
|
public V replace(K key, V value) {
|
|
return base.replace(key, value);
|
|
}
|
|
|
|
public V computeIfAbsent(K key, Function<? super K, ? extends V> mappingFunction) {
|
|
return base.computeIfAbsent(key, mappingFunction);
|
|
}
|
|
|
|
public V computeIfPresent(K key, BiFunction<? super K, ? super V, ? extends V> remappingFunction) {
|
|
return base.computeIfPresent(key, remappingFunction);
|
|
}
|
|
|
|
public V compute(K key, BiFunction<? super K, ? super V, ? extends V> remappingFunction) {
|
|
return base.compute(key, remappingFunction);
|
|
}
|
|
|
|
public V merge(K key, V value, BiFunction<? super V, ? super V, ? extends V> remappingFunction) {
|
|
return base.merge(key, value, remappingFunction);
|
|
}
|
|
|
|
// SequencedMap
|
|
|
|
public SequencedMap<K, V> reversed() {
|
|
return base;
|
|
}
|
|
|
|
public Entry<K, V> firstEntry() {
|
|
return base.lastEntry();
|
|
}
|
|
|
|
public Entry<K, V> lastEntry() {
|
|
return base.firstEntry();
|
|
}
|
|
|
|
public Entry<K, V> pollFirstEntry() {
|
|
return base.pollLastEntry();
|
|
}
|
|
|
|
public Entry<K, V> pollLastEntry() {
|
|
return base.pollFirstEntry();
|
|
}
|
|
|
|
public V putFirst(K k, V v) {
|
|
return base.putLast(k, v);
|
|
}
|
|
|
|
public V putLast(K k, V v) {
|
|
return base.putFirst(k, v);
|
|
}
|
|
}
|
|
}
|