2
* Written by Doug Lea with assistance from members of JCP JSR-166
3
* Expert Group and released to the public domain, as explained at
4
* http://creativecommons.org/licenses/publicdomain
8
* Repackaged for use in Ehcache by Chris Dennis. The only changes
9
* to this version where to override the AbstractCollection toArray
10
* implementation in KeySet.toArray(), Values.toArray(), and
11
* EntrySet.toArray() to ensure correct operation when using the 1.5
12
* version of AbstractCollection.
15
package net.sf.ehcache.store.chm;
17
import java.io.IOException;
18
import java.io.Serializable;
19
import java.util.AbstractCollection;
20
import java.util.AbstractMap;
21
import java.util.AbstractSet;
22
import java.util.ArrayList;
23
import java.util.Collection;
24
import java.util.ConcurrentModificationException;
25
import java.util.Enumeration;
26
import java.util.HashMap;
27
import java.util.Hashtable;
28
import java.util.Iterator;
30
import java.util.NoSuchElementException;
32
import java.util.concurrent.ConcurrentMap;
33
import java.util.concurrent.locks.ReentrantReadWriteLock;
35
import net.sf.ehcache.util.FindBugsSuppressWarnings;
38
* A hash table supporting full concurrency of retrievals and
39
* adjustable expected concurrency for updates. This class obeys the
40
* same functional specification as {@link java.util.Hashtable}, and
41
* includes versions of methods corresponding to each method of
42
* <tt>Hashtable</tt>. However, even though all operations are
43
* thread-safe, retrieval operations do <em>not</em> entail locking,
44
* and there is <em>not</em> any support for locking the entire table
45
* in a way that prevents all access. This class is fully
46
* interoperable with <tt>Hashtable</tt> in programs that rely on its
47
* thread safety but not on its synchronization details.
49
* <p> Retrieval operations (including <tt>get</tt>) generally do not
50
* block, so may overlap with update operations (including
51
* <tt>put</tt> and <tt>remove</tt>). Retrievals reflect the results
52
* of the most recently <em>completed</em> update operations holding
53
* upon their onset. For aggregate operations such as <tt>putAll</tt>
54
* and <tt>clear</tt>, concurrent retrievals may reflect insertion or
55
* removal of only some entries. Similarly, Iterators and
56
* Enumerations return elements reflecting the state of the hash table
57
* at some point at or since the creation of the iterator/enumeration.
58
* They do <em>not</em> throw {@link ConcurrentModificationException}.
59
* However, iterators are designed to be used by only one thread at a time.
61
* <p> The allowed concurrency among update operations is guided by
62
* the optional <tt>concurrencyLevel</tt> constructor argument
63
* (default <tt>16</tt>), which is used as a hint for internal sizing. The
64
* table is internally partitioned to try to permit the indicated
65
* number of concurrent updates without contention. Because placement
66
* in hash tables is essentially random, the actual concurrency will
67
* vary. Ideally, you should choose a value to accommodate as many
68
* threads as will ever concurrently modify the table. Using a
69
* significantly higher value than you need can waste space and time,
70
* and a significantly lower value can lead to thread contention. But
71
* overestimates and underestimates within an order of magnitude do
72
* not usually have much noticeable impact. A value of one is
73
* appropriate when it is known that only one thread will modify and
74
* all others will only read. Also, resizing this or any other kind of
75
* hash table is a relatively slow operation, so, when possible, it is
76
* a good idea to provide estimates of expected table sizes in
79
* <p>This class and its views and iterators implement all of the
80
* <em>optional</em> methods of the {@link Map} and {@link Iterator}
83
* <p> Like {@link Hashtable} but unlike {@link HashMap}, this class
84
* does <em>not</em> allow <tt>null</tt> to be used as a key or value.
86
* <p>This class is a member of the
87
* <a href="{@docRoot}/../technotes/guides/collections/index.html">
88
* Java Collections Framework</a>.
92
* @param <K> the type of keys maintained by this map
93
* @param <V> the type of mapped values
95
class ConcurrentHashMap<K, V> extends AbstractMap<K, V>
96
implements ConcurrentMap<K, V>, Serializable {
97
private static final long serialVersionUID = 7249069246763182397L;
100
* The basic strategy is to subdivide the table among Segments,
101
* each of which itself is a concurrently readable hash table.
104
/* ---------------- Constants -------------- */
107
* The default initial capacity for this table,
108
* used when not otherwise specified in a constructor.
110
static final int DEFAULT_INITIAL_CAPACITY = 16;
113
* The default load factor for this table, used when not
114
* otherwise specified in a constructor.
116
static final float DEFAULT_LOAD_FACTOR = 0.75f;
119
* The default concurrency level for this table, used when not
120
* otherwise specified in a constructor.
122
static final int DEFAULT_CONCURRENCY_LEVEL = 16;
125
* The maximum capacity, used if a higher value is implicitly
126
* specified by either of the constructors with arguments. MUST
127
* be a power of two <= 1<<30 to ensure that entries are indexable
130
static final int MAXIMUM_CAPACITY = 1 << 30;
133
* The maximum number of segments to allow; used to bound
134
* constructor arguments.
136
static final int MAX_SEGMENTS = 1 << 16; // slightly conservative
139
* Number of unsynchronized retries in size and containsValue
140
* methods before resorting to locking. This is used to avoid
141
* unbounded retries if tables undergo continuous modification
142
* which would make it impossible to obtain an accurate result.
144
static final int RETRIES_BEFORE_LOCK = 2;
146
/* ---------------- Fields -------------- */
149
* Mask value for indexing into segments. The upper bits of a
150
* key's hash code are used to choose the segment.
152
final int segmentMask;
155
* Shift value for indexing within segments.
157
final int segmentShift;
160
* The segments, each of which is a specialized hash table
162
final Segment<K,V>[] segments;
164
transient Set<K> keySet;
165
transient Set<Map.Entry<K,V>> entrySet;
166
transient Collection<V> values;
168
/* ---------------- Small Utilities -------------- */
171
* Applies a supplemental hash function to a given hashCode, which
172
* defends against poor quality hash functions. This is critical
173
* because ConcurrentHashMap uses power-of-two length hash tables,
174
* that otherwise encounter collisions for hashCodes that do not
175
* differ in lower or upper bits.
177
protected static int hash(int h) {
178
// Spread bits to regularize both segment and index locations,
179
// using variant of single-word Wang/Jenkins hash.
180
h += (h << 15) ^ 0xffffcd7d;
184
h += (h << 2) + (h << 14);
185
return h ^ (h >>> 16);
189
* Returns the segment that should be used for key with given hash
190
* @param hash the hash code for the key
191
* @return the segment
193
final Segment<K,V> segmentFor(int hash) {
194
return segments[(hash >>> segmentShift) & segmentMask];
197
/* ---------------- Inner Classes -------------- */
200
* ConcurrentHashMap list entry. Note that this is never exported
201
* out as a user-visible Map.Entry.
203
* Because the value field is volatile, not final, it is legal wrt
204
* the Java Memory Model for an unsynchronized reader to see null
205
* instead of initial value when read via a data race. Although a
206
* reordering leading to this is not likely to ever actually
207
* occur, the Segment.readValueUnderLock method is used as a
208
* backup in case a null (pre-initialized) value is ever seen in
209
* an unsynchronized access method.
211
static class HashEntry<K,V> {
215
final HashEntry<K,V> next;
217
HashEntry(K key, int hash, HashEntry<K,V> next, V value) {
224
@SuppressWarnings("unchecked")
225
static final <K,V> HashEntry<K,V>[] newArray(int i) {
226
return new HashEntry[i];
230
protected Segment<K, V> createSegment(int initialCapacity, float lf) {
231
return new Segment<K, V>(initialCapacity, lf);
235
* Segments are specialized versions of hash tables. This
236
* subclasses from ReentrantLock opportunistically, just to
237
* simplify some locking and avoid separate construction.
239
static class Segment<K,V> extends ReentrantReadWriteLock implements Serializable {
241
* Segments maintain a table of entry lists that are ALWAYS
242
* kept in a consistent state, so can be read without locking.
243
* Next fields of nodes are immutable (final). All list
244
* additions are performed at the front of each bin. This
245
* makes it easy to check changes, and also fast to traverse.
246
* When nodes would otherwise be changed, new nodes are
247
* created to replace them. This works well for hash tables
248
* since the bin lists tend to be short. (The average length
249
* is less than two for the default load factor threshold.)
251
* Read operations can thus proceed without locking, but rely
252
* on selected uses of volatiles to ensure that completed
253
* write operations performed by other threads are
254
* noticed. For most purposes, the "count" field, tracking the
255
* number of elements, serves as that volatile variable
256
* ensuring visibility. This is convenient because this field
257
* needs to be read in many read operations anyway:
259
* - All (unsynchronized) read operations must first read the
260
* "count" field, and should not look at table entries if
263
* - All (synchronized) write operations should write to
264
* the "count" field after structurally changing any bin.
265
* The operations must not take any action that could even
266
* momentarily cause a concurrent read operation to see
267
* inconsistent data. This is made easier by the nature of
268
* the read operations in Map. For example, no operation
269
* can reveal that the table has grown but the threshold
270
* has not yet been updated, so there are no atomicity
271
* requirements for this with respect to reads.
273
* As a guide, all critical volatile reads and writes to the
274
* count field are marked in code comments.
277
private static final long serialVersionUID = 2249069246763182397L;
280
* The number of elements in this segment's region.
282
transient volatile int count;
285
* Number of updates that alter the size of the table. This is
286
* used during bulk-read methods to make sure they see a
287
* consistent snapshot: If modCounts change during a traversal
288
* of segments computing size or checking containsValue, then
289
* we might have an inconsistent view of state so (usually)
292
transient int modCount;
295
* The table is rehashed when its size exceeds this threshold.
296
* (The value of this field is always <tt>(int)(capacity *
299
transient int threshold;
302
* The per-segment table.
304
transient volatile HashEntry<K,V>[] table;
307
* The load factor for the hash table. Even though this value
308
* is same for all segments, it is replicated to avoid needing
309
* links to outer object.
312
final float loadFactor;
314
Segment(int initialCapacity, float lf) {
316
setTable(HashEntry.<K,V>newArray(initialCapacity));
319
@SuppressWarnings("unchecked")
320
static final <K,V> Segment<K,V>[] newArray(int i) {
321
return new Segment[i];
325
* Sets table to new HashEntry array.
326
* Call only while holding lock or in constructor.
328
void setTable(HashEntry<K,V>[] newTable) {
329
threshold = (int)(newTable.length * loadFactor);
334
* Returns properly casted first entry of bin for given hash.
336
HashEntry<K,V> getFirst(int hash) {
337
HashEntry<K,V>[] tab = table;
338
return tab[hash & (tab.length - 1)];
341
/* Specialized implementations of map methods */
343
V get(Object key, int hash) {
346
if (count != 0) { // read-volatile
347
HashEntry<K,V> e = getFirst(hash);
349
if (e.hash == hash && key.equals(e.key)) {
361
boolean containsKey(Object key, int hash) {
364
if (count != 0) { // read-volatile
365
HashEntry<K,V> e = getFirst(hash);
367
if (e.hash == hash && key.equals(e.key))
378
boolean containsValue(Object value) {
381
if (count != 0) { // read-volatile
382
HashEntry<K,V>[] tab = table;
383
int len = tab.length;
384
for (int i = 0 ; i < len; i++) {
385
for (HashEntry<K,V> e = tab[i]; e != null; e = e.next) {
398
boolean replace(K key, int hash, V oldValue, V newValue) {
401
HashEntry<K,V> e = getFirst(hash);
402
while (e != null && (e.hash != hash || !key.equals(e.key)))
405
boolean replaced = false;
406
if (e != null && oldValue.equals(e.value)) {
412
writeLock().unlock();
416
V replace(K key, int hash, V newValue) {
419
HashEntry<K,V> e = getFirst(hash);
420
while (e != null && (e.hash != hash || !key.equals(e.key)))
430
writeLock().unlock();
435
V put(K key, int hash, V value, boolean onlyIfAbsent) {
439
if (c++ > threshold) // ensure capacity
441
HashEntry<K,V>[] tab = table;
442
int index = hash & (tab.length - 1);
443
HashEntry<K,V> first = tab[index];
444
HashEntry<K,V> e = first;
445
while (e != null && (e.hash != hash || !key.equals(e.key)))
457
tab[index] = new HashEntry<K,V>(key, hash, first, value);
458
count = c; // write-volatile
462
writeLock().unlock();
467
HashEntry<K,V>[] oldTable = table;
468
int oldCapacity = oldTable.length;
469
if (oldCapacity >= MAXIMUM_CAPACITY)
473
* Reclassify nodes in each list to new Map. Because we are
474
* using power-of-two expansion, the elements from each bin
475
* must either stay at same index, or move with a power of two
476
* offset. We eliminate unnecessary node creation by catching
477
* cases where old nodes can be reused because their next
478
* fields won't change. Statistically, at the default
479
* threshold, only about one-sixth of them need cloning when
480
* a table doubles. The nodes they replace will be garbage
481
* collectable as soon as they are no longer referenced by any
482
* reader thread that may be in the midst of traversing table
486
HashEntry<K,V>[] newTable = HashEntry.newArray(oldCapacity<<1);
487
threshold = (int)(newTable.length * loadFactor);
488
int sizeMask = newTable.length - 1;
489
for (int i = 0; i < oldCapacity ; i++) {
490
// We need to guarantee that any existing reads of old Map can
491
// proceed. So we cannot yet null out each bin.
492
HashEntry<K,V> e = oldTable[i];
495
HashEntry<K,V> next = e.next;
496
int idx = e.hash & sizeMask;
498
// Single node on list
503
// Reuse trailing consecutive sequence at same slot
504
HashEntry<K,V> lastRun = e;
506
for (HashEntry<K,V> last = next;
509
int k = last.hash & sizeMask;
515
newTable[lastIdx] = lastRun;
517
// Clone all remaining nodes
518
for (HashEntry<K,V> p = e; p != lastRun; p = p.next) {
519
int k = p.hash & sizeMask;
520
HashEntry<K,V> n = newTable[k];
521
newTable[k] = relinkHashEntry(p, n);
530
* Remove; match on key only if value null, else match both.
532
V remove(Object key, int hash, Object value) {
536
HashEntry<K,V>[] tab = table;
537
int index = hash & (tab.length - 1);
538
HashEntry<K,V> first = tab[index];
539
HashEntry<K,V> e = first;
540
while (e != null && (e.hash != hash || !key.equals(e.key)))
546
if (value == null || value.equals(v)) {
548
// All entries following removed node can stay
549
// in list, but all preceding ones need to be
552
HashEntry<K,V> newFirst = e.next;
553
for (HashEntry<K,V> p = first; p != e; p = p.next)
554
newFirst = relinkHashEntry(p, newFirst);
555
tab[index] = newFirst;
556
count = c; // write-volatile
561
writeLock().unlock();
569
HashEntry<K,V>[] tab = table;
570
for (int i = 0; i < tab.length ; i++)
573
count = 0; // write-volatile
576
writeLock().unlock();
580
protected HashEntry<K, V> relinkHashEntry(HashEntry<K,V> e, HashEntry<K,V> next) {
581
return new HashEntry<K,V>(e.key, e.hash, next, e.value);
587
/* ---------------- Public operations -------------- */
590
* Creates a new, empty map with the specified initial
591
* capacity, load factor and concurrency level.
593
* @param initialCapacity the initial capacity. The implementation
594
* performs internal sizing to accommodate this many elements.
595
* @param loadFactor the load factor threshold, used to control resizing.
596
* Resizing may be performed when the average number of elements per
597
* bin exceeds this threshold.
598
* @param concurrencyLevel the estimated number of concurrently
599
* updating threads. The implementation performs internal sizing
600
* to try to accommodate this many threads.
601
* @throws IllegalArgumentException if the initial capacity is
602
* negative or the load factor or concurrencyLevel are
605
public ConcurrentHashMap(int initialCapacity,
606
float loadFactor, int concurrencyLevel) {
607
if (!(loadFactor > 0) || initialCapacity < 0 || concurrencyLevel <= 0)
608
throw new IllegalArgumentException();
610
if (concurrencyLevel > MAX_SEGMENTS)
611
concurrencyLevel = MAX_SEGMENTS;
613
// Find power-of-two sizes best matching arguments
616
while (ssize < concurrencyLevel) {
620
segmentShift = 32 - sshift;
621
segmentMask = ssize - 1;
622
this.segments = Segment.newArray(ssize);
624
if (initialCapacity > MAXIMUM_CAPACITY)
625
initialCapacity = MAXIMUM_CAPACITY;
626
int c = initialCapacity / ssize;
627
if (c * ssize < initialCapacity)
633
for (int i = 0; i < this.segments.length; ++i)
634
this.segments[i] = createSegment(cap, loadFactor);
638
* Creates a new, empty map with the specified initial capacity,
639
* and with default load factor (0.75) and concurrencyLevel (16).
641
* @param initialCapacity the initial capacity. The implementation
642
* performs internal sizing to accommodate this many elements.
643
* @throws IllegalArgumentException if the initial capacity of
644
* elements is negative.
646
public ConcurrentHashMap(int initialCapacity) {
647
this(initialCapacity, DEFAULT_LOAD_FACTOR, DEFAULT_CONCURRENCY_LEVEL);
651
* Creates a new, empty map with a default initial capacity (16),
652
* load factor (0.75) and concurrencyLevel (16).
654
public ConcurrentHashMap() {
655
this(DEFAULT_INITIAL_CAPACITY, DEFAULT_LOAD_FACTOR, DEFAULT_CONCURRENCY_LEVEL);
659
* Creates a new map with the same mappings as the given map.
660
* The map is created with a capacity of 1.5 times the number
661
* of mappings in the given map or 16 (whichever is greater),
662
* and a default load factor (0.75) and concurrencyLevel (16).
666
public ConcurrentHashMap(Map<? extends K, ? extends V> m) {
667
this(Math.max((int) (m.size() / DEFAULT_LOAD_FACTOR) + 1,
668
DEFAULT_INITIAL_CAPACITY),
669
DEFAULT_LOAD_FACTOR, DEFAULT_CONCURRENCY_LEVEL);
674
* Returns <tt>true</tt> if this map contains no key-value mappings.
676
* @return <tt>true</tt> if this map contains no key-value mappings
678
public boolean isEmpty() {
679
final Segment<K,V>[] segments = this.segments;
681
* We keep track of per-segment modCounts to avoid ABA
682
* problems in which an element in one segment was added and
683
* in another removed during traversal, in which case the
684
* table was never actually empty at any point. Note the
685
* similar use of modCounts in the size() and containsValue()
686
* methods, which are the only other methods also susceptible
689
int[] mc = new int[segments.length];
691
for (int i = 0; i < segments.length; ++i) {
692
if (segments[i].count != 0)
695
mcsum += mc[i] = segments[i].modCount;
697
// If mcsum happens to be zero, then we know we got a snapshot
698
// before any modifications at all were made. This is
699
// probably common enough to bother tracking.
701
for (int i = 0; i < segments.length; ++i) {
702
if (segments[i].count != 0 ||
703
mc[i] != segments[i].modCount)
711
* Returns the number of key-value mappings in this map. If the
712
* map contains more than <tt>Integer.MAX_VALUE</tt> elements, returns
713
* <tt>Integer.MAX_VALUE</tt>.
715
* @return the number of key-value mappings in this map
717
@FindBugsSuppressWarnings("UL_UNRELEASED_LOCK")
719
final Segment<K,V>[] segments = this.segments;
722
int[] mc = new int[segments.length];
723
// Try a few times to get accurate count. On failure due to
724
// continuous async changes in table, resort to locking.
725
for (int k = 0; k < RETRIES_BEFORE_LOCK; ++k) {
729
for (int i = 0; i < segments.length; ++i) {
730
sum += segments[i].count;
731
mcsum += mc[i] = segments[i].modCount;
734
for (int i = 0; i < segments.length; ++i) {
735
check += segments[i].count;
736
if (mc[i] != segments[i].modCount) {
737
check = -1; // force retry
745
if (check != sum) { // Resort to locking all segments
747
for (int i = 0; i < segments.length; ++i)
748
segments[i].readLock().lock();
750
for (int i = 0; i < segments.length; ++i)
751
sum += segments[i].count;
753
for (int i = 0; i < segments.length; ++i)
754
segments[i].readLock().unlock();
757
if (sum > Integer.MAX_VALUE)
758
return Integer.MAX_VALUE;
764
* Returns the value to which the specified key is mapped,
765
* or {@code null} if this map contains no mapping for the key.
767
* <p>More formally, if this map contains a mapping from a key
768
* {@code k} to a value {@code v} such that {@code key.equals(k)},
769
* then this method returns {@code v}; otherwise it returns
770
* {@code null}. (There can be at most one such mapping.)
772
* @throws NullPointerException if the specified key is null
774
public V get(Object key) {
775
int hash = hash(key.hashCode());
776
return segmentFor(hash).get(key, hash);
780
* Tests if the specified object is a key in this table.
782
* @param key possible key
783
* @return <tt>true</tt> if and only if the specified object
784
* is a key in this table, as determined by the
785
* <tt>equals</tt> method; <tt>false</tt> otherwise.
786
* @throws NullPointerException if the specified key is null
788
public boolean containsKey(Object key) {
789
int hash = hash(key.hashCode());
790
return segmentFor(hash).containsKey(key, hash);
794
* Returns <tt>true</tt> if this map maps one or more keys to the
795
* specified value. Note: This method requires a full internal
796
* traversal of the hash table, and so is much slower than
797
* method <tt>containsKey</tt>.
799
* @param value value whose presence in this map is to be tested
800
* @return <tt>true</tt> if this map maps one or more keys to the
802
* @throws NullPointerException if the specified value is null
804
@FindBugsSuppressWarnings("UL_UNRELEASED_LOCK")
805
public boolean containsValue(Object value) {
807
throw new NullPointerException();
809
// See explanation of modCount use above
811
final Segment<K,V>[] segments = this.segments;
812
int[] mc = new int[segments.length];
814
// Try a few times without locking
815
for (int k = 0; k < RETRIES_BEFORE_LOCK; ++k) {
818
for (int i = 0; i < segments.length; ++i) {
819
int c = segments[i].count;
820
mcsum += mc[i] = segments[i].modCount;
821
if (segments[i].containsValue(value))
824
boolean cleanSweep = true;
826
for (int i = 0; i < segments.length; ++i) {
827
int c = segments[i].count;
828
if (mc[i] != segments[i].modCount) {
838
// Resort to locking all segments
839
for (int i = 0; i < segments.length; ++i)
840
segments[i].readLock().lock();
842
for (int i = 0; i < segments.length; ++i) {
843
if (segments[i].containsValue(value)) {
848
for (int i = 0; i < segments.length; ++i)
849
segments[i].readLock().unlock();
855
* Legacy method testing if some key maps into the specified value
856
* in this table. This method is identical in functionality to
857
* {@link #containsValue}, and exists solely to ensure
858
* full compatibility with class {@link java.util.Hashtable},
859
* which supported this method prior to introduction of the
860
* Java Collections framework.
862
* @param value a value to search for
863
* @return <tt>true</tt> if and only if some key maps to the
864
* <tt>value</tt> argument in this table as
865
* determined by the <tt>equals</tt> method;
866
* <tt>false</tt> otherwise
867
* @throws NullPointerException if the specified value is null
869
public boolean contains(Object value) {
870
return containsValue(value);
874
* Maps the specified key to the specified value in this table.
875
* Neither the key nor the value can be null.
877
* <p> The value can be retrieved by calling the <tt>get</tt> method
878
* with a key that is equal to the original key.
880
* @param key key with which the specified value is to be associated
881
* @param value value to be associated with the specified key
882
* @return the previous value associated with <tt>key</tt>, or
883
* <tt>null</tt> if there was no mapping for <tt>key</tt>
884
* @throws NullPointerException if the specified key or value is null
886
public V put(K key, V value) {
888
throw new NullPointerException();
889
int hash = hash(key.hashCode());
890
return segmentFor(hash).put(key, hash, value, false);
896
* @return the previous value associated with the specified key,
897
* or <tt>null</tt> if there was no mapping for the key
898
* @throws NullPointerException if the specified key or value is null
900
public V putIfAbsent(K key, V value) {
902
throw new NullPointerException();
903
int hash = hash(key.hashCode());
904
return segmentFor(hash).put(key, hash, value, true);
908
* Copies all of the mappings from the specified map to this one.
909
* These mappings replace any mappings that this map had for any of the
910
* keys currently in the specified map.
912
* @param m mappings to be stored in this map
914
public void putAll(Map<? extends K, ? extends V> m) {
915
for (Map.Entry<? extends K, ? extends V> e : m.entrySet())
916
put(e.getKey(), e.getValue());
920
* Removes the key (and its corresponding value) from this map.
921
* This method does nothing if the key is not in the map.
923
* @param key the key that needs to be removed
924
* @return the previous value associated with <tt>key</tt>, or
925
* <tt>null</tt> if there was no mapping for <tt>key</tt>
926
* @throws NullPointerException if the specified key is null
928
public V remove(Object key) {
929
int hash = hash(key.hashCode());
930
return segmentFor(hash).remove(key, hash, null);
936
* @throws NullPointerException if the specified key is null
938
public boolean remove(Object key, Object value) {
939
int hash = hash(key.hashCode());
942
return segmentFor(hash).remove(key, hash, value) != null;
948
* @throws NullPointerException if any of the arguments are null
950
public boolean replace(K key, V oldValue, V newValue) {
951
if (oldValue == null || newValue == null)
952
throw new NullPointerException();
953
int hash = hash(key.hashCode());
954
return segmentFor(hash).replace(key, hash, oldValue, newValue);
960
* @return the previous value associated with the specified key,
961
* or <tt>null</tt> if there was no mapping for the key
962
* @throws NullPointerException if the specified key or value is null
964
public V replace(K key, V value) {
966
throw new NullPointerException();
967
int hash = hash(key.hashCode());
968
return segmentFor(hash).replace(key, hash, value);
972
* Removes all of the mappings from this map.
974
public void clear() {
975
for (int i = 0; i < segments.length; ++i)
980
* Returns a {@link Set} view of the keys contained in this map.
981
* The set is backed by the map, so changes to the map are
982
* reflected in the set, and vice-versa. The set supports element
983
* removal, which removes the corresponding mapping from this map,
984
* via the <tt>Iterator.remove</tt>, <tt>Set.remove</tt>,
985
* <tt>removeAll</tt>, <tt>retainAll</tt>, and <tt>clear</tt>
986
* operations. It does not support the <tt>add</tt> or
987
* <tt>addAll</tt> operations.
989
* <p>The view's <tt>iterator</tt> is a "weakly consistent" iterator
990
* that will never throw {@link ConcurrentModificationException},
991
* and guarantees to traverse elements as they existed upon
992
* construction of the iterator, and may (but is not guaranteed to)
993
* reflect any modifications subsequent to construction.
995
public Set<K> keySet() {
997
return (ks != null) ? ks : (keySet = new KeySet());
1001
* Returns a {@link Collection} view of the values contained in this map.
1002
* The collection is backed by the map, so changes to the map are
1003
* reflected in the collection, and vice-versa. The collection
1004
* supports element removal, which removes the corresponding
1005
* mapping from this map, via the <tt>Iterator.remove</tt>,
1006
* <tt>Collection.remove</tt>, <tt>removeAll</tt>,
1007
* <tt>retainAll</tt>, and <tt>clear</tt> operations. It does not
1008
* support the <tt>add</tt> or <tt>addAll</tt> operations.
1010
* <p>The view's <tt>iterator</tt> is a "weakly consistent" iterator
1011
* that will never throw {@link ConcurrentModificationException},
1012
* and guarantees to traverse elements as they existed upon
1013
* construction of the iterator, and may (but is not guaranteed to)
1014
* reflect any modifications subsequent to construction.
1016
public Collection<V> values() {
1017
Collection<V> vs = values;
1018
return (vs != null) ? vs : (values = new Values());
1022
* Returns a {@link Set} view of the mappings contained in this map.
1023
* The set is backed by the map, so changes to the map are
1024
* reflected in the set, and vice-versa. The set supports element
1025
* removal, which removes the corresponding mapping from the map,
1026
* via the <tt>Iterator.remove</tt>, <tt>Set.remove</tt>,
1027
* <tt>removeAll</tt>, <tt>retainAll</tt>, and <tt>clear</tt>
1028
* operations. It does not support the <tt>add</tt> or
1029
* <tt>addAll</tt> operations.
1031
* <p>The view's <tt>iterator</tt> is a "weakly consistent" iterator
1032
* that will never throw {@link ConcurrentModificationException},
1033
* and guarantees to traverse elements as they existed upon
1034
* construction of the iterator, and may (but is not guaranteed to)
1035
* reflect any modifications subsequent to construction.
1037
public Set<Map.Entry<K,V>> entrySet() {
1038
Set<Map.Entry<K,V>> es = entrySet;
1039
return (es != null) ? es : (entrySet = new EntrySet());
1043
* Returns an enumeration of the keys in this table.
1045
* @return an enumeration of the keys in this table
1048
public Enumeration<K> keys() {
1049
return new KeyIterator();
1053
* Returns an enumeration of the values in this table.
1055
* @return an enumeration of the values in this table
1058
public Enumeration<V> elements() {
1059
return new ValueIterator();
1062
/* ---------------- Iterator Support -------------- */
1064
abstract class HashIterator {
1065
int nextSegmentIndex;
1067
HashEntry<K,V>[] currentTable;
1068
HashEntry<K, V> nextEntry;
1069
HashEntry<K, V> lastReturned;
1072
nextSegmentIndex = segments.length - 1;
1073
nextTableIndex = -1;
1077
public boolean hasMoreElements() { return hasNext(); }
1079
final void advance() {
1080
if (nextEntry != null && (nextEntry = nextEntry.next) != null)
1083
while (nextTableIndex >= 0) {
1084
if ( (nextEntry = currentTable[nextTableIndex--]) != null)
1088
while (nextSegmentIndex >= 0) {
1089
Segment<K,V> seg = segments[nextSegmentIndex--];
1090
if (seg.count != 0) {
1091
currentTable = seg.table;
1092
for (int j = currentTable.length - 1; j >= 0; --j) {
1093
if ( (nextEntry = currentTable[j]) != null) {
1094
nextTableIndex = j - 1;
1102
public boolean hasNext() { return nextEntry != null; }
1104
HashEntry<K,V> nextEntry() {
1105
if (nextEntry == null)
1106
throw new NoSuchElementException();
1107
lastReturned = nextEntry;
1109
return lastReturned;
1112
public void remove() {
1113
if (lastReturned == null)
1114
throw new IllegalStateException();
1115
ConcurrentHashMap.this.remove(lastReturned.key);
1116
lastReturned = null;
1120
final class KeyIterator
1121
extends HashIterator
1122
implements Iterator<K>, Enumeration<K>
1124
public K next() { return super.nextEntry().key; }
1125
public K nextElement() { return super.nextEntry().key; }
1128
final class ValueIterator
1129
extends HashIterator
1130
implements Iterator<V>, Enumeration<V>
1132
public V next() { return super.nextEntry().value; }
1133
public V nextElement() { return super.nextEntry().value; }
1137
* Custom Entry class used by EntryIterator.next(), that relays
1138
* setValue changes to the underlying map.
1140
final class WriteThroughEntry
1141
extends SimpleEntry<K,V>
1143
WriteThroughEntry(K k, V v) {
1148
* Set our entry's value and write through to the map. The
1149
* value to return is somewhat arbitrary here. Since a
1150
* WriteThroughEntry does not necessarily track asynchronous
1151
* changes, the most recent "previous" value could be
1152
* different from what we return (or could even have been
1153
* removed in which case the put will re-establish). We do not
1154
* and cannot guarantee more.
1156
public V setValue(V value) {
1157
if (value == null) throw new NullPointerException();
1158
V v = super.setValue(value);
1159
ConcurrentHashMap.this.put(getKey(), value);
1164
final class EntryIterator
1165
extends HashIterator
1166
implements Iterator<Entry<K,V>>
1168
public Map.Entry<K,V> next() {
1169
HashEntry<K,V> e = super.nextEntry();
1170
return new WriteThroughEntry(e.key, e.value);
1174
final class KeySet extends AbstractSet<K> {
1175
public Iterator<K> iterator() {
1176
return new KeyIterator();
1179
return ConcurrentHashMap.this.size();
1181
public boolean isEmpty() {
1182
return ConcurrentHashMap.this.isEmpty();
1184
public boolean contains(Object o) {
1185
return ConcurrentHashMap.this.containsKey(o);
1187
public boolean remove(Object o) {
1188
return ConcurrentHashMap.this.remove(o) != null;
1190
public void clear() {
1191
ConcurrentHashMap.this.clear();
1193
public Object[] toArray() {
1194
Collection<K> c = new ArrayList<K>();
1195
for (Iterator<K> i = iterator(); i.hasNext(); )
1199
public <T> T[] toArray(T[] a) {
1200
Collection<K> c = new ArrayList<K>();
1201
for (Iterator<K> i = iterator(); i.hasNext(); )
1203
return c.toArray(a);
1207
final class Values extends AbstractCollection<V> {
1208
public Iterator<V> iterator() {
1209
return new ValueIterator();
1212
return ConcurrentHashMap.this.size();
1214
public boolean isEmpty() {
1215
return ConcurrentHashMap.this.isEmpty();
1217
public boolean contains(Object o) {
1218
return ConcurrentHashMap.this.containsValue(o);
1220
public void clear() {
1221
ConcurrentHashMap.this.clear();
1223
public Object[] toArray() {
1224
Collection<V> c = new ArrayList<V>();
1225
for (Iterator<V> i = iterator(); i.hasNext(); )
1229
public <T> T[] toArray(T[] a) {
1230
Collection<V> c = new ArrayList<V>();
1231
for (Iterator<V> i = iterator(); i.hasNext(); )
1233
return c.toArray(a);
1237
final class EntrySet extends AbstractSet<Map.Entry<K,V>> {
1238
public Iterator<Map.Entry<K,V>> iterator() {
1239
return new EntryIterator();
1241
public boolean contains(Object o) {
1242
if (!(o instanceof Map.Entry))
1244
Map.Entry<?,?> e = (Map.Entry<?,?>)o;
1245
V v = ConcurrentHashMap.this.get(e.getKey());
1246
return v != null && v.equals(e.getValue());
1248
public boolean remove(Object o) {
1249
if (!(o instanceof Map.Entry))
1251
Map.Entry<?,?> e = (Map.Entry<?,?>)o;
1252
return ConcurrentHashMap.this.remove(e.getKey(), e.getValue());
1255
return ConcurrentHashMap.this.size();
1257
public boolean isEmpty() {
1258
return ConcurrentHashMap.this.isEmpty();
1260
public void clear() {
1261
ConcurrentHashMap.this.clear();
1263
public Object[] toArray() {
1264
Collection<Map.Entry<K,V>> c = new ArrayList<Map.Entry<K,V>>();
1265
for (Iterator<Map.Entry<K,V>> i = iterator(); i.hasNext(); )
1269
public <T> T[] toArray(T[] a) {
1270
Collection<Map.Entry<K,V>> c = new ArrayList<Map.Entry<K,V>>();
1271
for (Iterator<Map.Entry<K,V>> i = iterator(); i.hasNext(); )
1273
return c.toArray(a);
1278
* This duplicates java.util.AbstractMap.SimpleEntry until this class
1279
* is made accessible.
1281
static class SimpleEntry<K,V> implements Entry<K,V> {
1285
public SimpleEntry(K key, V value) {
1290
public SimpleEntry(Entry<K,V> e) {
1291
this.key = e.getKey();
1292
this.value = e.getValue();
1299
public V getValue() {
1303
public V setValue(V value) {
1304
V oldValue = this.value;
1309
public boolean equals(Object o) {
1310
if (!(o instanceof Map.Entry))
1312
Map.Entry e = (Map.Entry)o;
1313
return eq(key, e.getKey()) && eq(value, e.getValue());
1316
public int hashCode() {
1317
return ((key == null) ? 0 : key.hashCode()) ^
1318
((value == null) ? 0 : value.hashCode());
1321
public String toString() {
1322
return key + "=" + value;
1325
static boolean eq(Object o1, Object o2) {
1326
return (o1 == null ? o2 == null : o1.equals(o2));
1330
/* ---------------- Serialization Support -------------- */
1333
* Save the state of the <tt>ConcurrentHashMap</tt> instance to a
1334
* stream (i.e., serialize it).
1335
* @param s the stream
1337
* the key (Object) and value (Object)
1338
* for each key-value mapping, followed by a null pair.
1339
* The key-value mappings are emitted in no particular order.
1341
private void writeObject(java.io.ObjectOutputStream s) throws IOException {
1342
s.defaultWriteObject();
1344
for (int k = 0; k < segments.length; ++k) {
1345
Segment<K,V> seg = segments[k];
1346
seg.readLock().lock();
1348
HashEntry<K,V>[] tab = seg.table;
1349
for (int i = 0; i < tab.length; ++i) {
1350
for (HashEntry<K,V> e = tab[i]; e != null; e = e.next) {
1351
s.writeObject(e.key);
1352
s.writeObject(e.value);
1356
seg.readLock().unlock();
1359
s.writeObject(null);
1360
s.writeObject(null);
1364
* Reconstitute the <tt>ConcurrentHashMap</tt> instance from a
1365
* stream (i.e., deserialize it).
1366
* @param s the stream
1368
private void readObject(java.io.ObjectInputStream s)
1369
throws IOException, ClassNotFoundException {
1370
s.defaultReadObject();
1372
// Initialize each segment to be minimally sized, and let grow.
1373
for (int i = 0; i < segments.length; ++i) {
1374
segments[i].setTable(new HashEntry[1]);
1377
// Read the keys and values, and put the mappings in the table
1379
K key = (K) s.readObject();
1380
V value = (V) s.readObject();
added
removed
src/main/java/net/sf/ehcache/store/chm/ConcurrentHashMap.java
