ViewVC Help

View File | Revision Log | Show Annotations | Download File | Root Listing

root/jsr166/jsr166/src/main/java/util/concurrent/ConcurrentHashMap.java

Comparing jsr166/src/main/java/util/concurrent/ConcurrentHashMap.java (file contents):
Revision 1.110 by jsr166, Wed Apr 27 14:06:30 2011 UTC vs.
Revision 1.222 by dl, Mon Jun 17 18:53:58 2013 UTC

#	Line 5 | Line 5
5		*/
6
7		package java.util.concurrent;
8	<	import java.util.concurrent.locks.*;
9	<	import java.util.*;
8	>
9	>	import java.io.ObjectStreamField;
10		import java.io.Serializable;
11	<	import java.io.IOException;
12	<	import java.io.ObjectInputStream;
13	<	import java.io.ObjectOutputStream;
11	>	import java.lang.reflect.ParameterizedType;
12	>	import java.lang.reflect.Type;
13	>	import java.util.Arrays;
14	>	import java.util.Collection;
15	>	import java.util.Comparator;
16	>	import java.util.ConcurrentModificationException;
17	>	import java.util.Enumeration;
18	>	import java.util.HashMap;
19	>	import java.util.Hashtable;
20	>	import java.util.Iterator;
21	>	import java.util.Map;
22	>	import java.util.NoSuchElementException;
23	>	import java.util.Set;
24	>	import java.util.Spliterator;
25	>	import java.util.concurrent.ConcurrentMap;
26	>	import java.util.concurrent.ForkJoinPool;
27	>	import java.util.concurrent.atomic.AtomicReference;
28	>	import java.util.concurrent.locks.LockSupport;
29	>	import java.util.concurrent.locks.ReentrantLock;
30	>	import java.util.function.BiConsumer;
31	>	import java.util.function.BiFunction;
32	>	import java.util.function.BinaryOperator;
33	>	import java.util.function.Consumer;
34	>	import java.util.function.DoubleBinaryOperator;
35	>	import java.util.function.Function;
36	>	import java.util.function.IntBinaryOperator;
37	>	import java.util.function.LongBinaryOperator;
38	>	import java.util.function.ToDoubleBiFunction;
39	>	import java.util.function.ToDoubleFunction;
40	>	import java.util.function.ToIntBiFunction;
41	>	import java.util.function.ToIntFunction;
42	>	import java.util.function.ToLongBiFunction;
43	>	import java.util.function.ToLongFunction;
44	>	import java.util.stream.Stream;
45
46		/**
47		* A hash table supporting full concurrency of retrievals and
48	<	* adjustable expected concurrency for updates. This class obeys the
48	>	* high expected concurrency for updates. This class obeys the
49		* same functional specification as {@link java.util.Hashtable}, and
50		* includes versions of methods corresponding to each method of
51	<	* <tt>Hashtable</tt>. However, even though all operations are
51	>	* {@code Hashtable}. However, even though all operations are
52		* thread-safe, retrieval operations do <em>not</em> entail locking,
53		* and there is <em>not</em> any support for locking the entire table
54		* in a way that prevents all access.  This class is fully
55	<	* interoperable with <tt>Hashtable</tt> in programs that rely on its
55	>	* interoperable with {@code Hashtable} in programs that rely on its
56		* thread safety but not on its synchronization details.
57		*
58	<	* <p> Retrieval operations (including <tt>get</tt>) generally do not
59	<	* block, so may overlap with update operations (including
60	<	* <tt>put</tt> and <tt>remove</tt>). Retrievals reflect the results
61	<	* of the most recently <em>completed</em> update operations holding
62	<	* upon their onset.  For aggregate operations such as <tt>putAll</tt>
63	<	* and <tt>clear</tt>, concurrent retrievals may reflect insertion or
64	<	* removal of only some entries.  Similarly, Iterators and
65	<	* Enumerations return elements reflecting the state of the hash table
66	<	* at some point at or since the creation of the iterator/enumeration.
67	<	* They do <em>not</em> throw {@link ConcurrentModificationException}.
68	<	* However, iterators are designed to be used by only one thread at a time.
69	<	*
70	<	* <p> The allowed concurrency among update operations is guided by
71	<	* the optional <tt>concurrencyLevel</tt> constructor argument
72	<	* (default <tt>16</tt>), which is used as a hint for internal sizing.  The
73	<	* table is internally partitioned to try to permit the indicated
74	<	* number of concurrent updates without contention. Because placement
75	<	* in hash tables is essentially random, the actual concurrency will
76	<	* vary.  Ideally, you should choose a value to accommodate as many
77	<	* threads as will ever concurrently modify the table. Using a
78	<	* significantly higher value than you need can waste space and time,
79	<	* and a significantly lower value can lead to thread contention. But
80	<	* overestimates and underestimates within an order of magnitude do
81	<	* not usually have much noticeable impact. A value of one is
82	<	* appropriate when it is known that only one thread will modify and
83	<	* all others will only read. Also, resizing this or any other kind of
84	<	* hash table is a relatively slow operation, so, when possible, it is
85	<	* a good idea to provide estimates of expected table sizes in
86	<	* constructors.
58	>	* <p>Retrieval operations (including {@code get}) generally do not
59	>	* block, so may overlap with update operations (including {@code put}
60	>	* and {@code remove}). Retrievals reflect the results of the most
61	>	* recently <em>completed</em> update operations holding upon their
62	>	* onset. (More formally, an update operation for a given key bears a
63	>	* <em>happens-before</em> relation with any (non-null) retrieval for
64	>	* that key reporting the updated value.)  For aggregate operations
65	>	* such as {@code putAll} and {@code clear}, concurrent retrievals may
66	>	* reflect insertion or removal of only some entries.  Similarly,
67	>	* Iterators and Enumerations return elements reflecting the state of
68	>	* the hash table at some point at or since the creation of the
69	>	* iterator/enumeration.  They do <em>not</em> throw {@link
70	>	* ConcurrentModificationException}.  However, iterators are designed
71	>	* to be used by only one thread at a time.  Bear in mind that the
72	>	* results of aggregate status methods including {@code size}, {@code
73	>	* isEmpty}, and {@code containsValue} are typically useful only when
74	>	* a map is not undergoing concurrent updates in other threads.
75	>	* Otherwise the results of these methods reflect transient states
76	>	* that may be adequate for monitoring or estimation purposes, but not
77	>	* for program control.
78	>	*
79	>	* <p>The table is dynamically expanded when there are too many
80	>	* collisions (i.e., keys that have distinct hash codes but fall into
81	>	* the same slot modulo the table size), with the expected average
82	>	* effect of maintaining roughly two bins per mapping (corresponding
83	>	* to a 0.75 load factor threshold for resizing). There may be much
84	>	* variance around this average as mappings are added and removed, but
85	>	* overall, this maintains a commonly accepted time/space tradeoff for
86	>	* hash tables.  However, resizing this or any other kind of hash
87	>	* table may be a relatively slow operation. When possible, it is a
88	>	* good idea to provide a size estimate as an optional {@code
89	>	* initialCapacity} constructor argument. An additional optional
90	>	* {@code loadFactor} constructor argument provides a further means of
91	>	* customizing initial table capacity by specifying the table density
92	>	* to be used in calculating the amount of space to allocate for the
93	>	* given number of elements.  Also, for compatibility with previous
94	>	* versions of this class, constructors may optionally specify an
95	>	* expected {@code concurrencyLevel} as an additional hint for
96	>	* internal sizing.  Note that using many keys with exactly the same
97	>	* {@code hashCode()} is a sure way to slow down performance of any
98	>	* hash table. To ameliorate impact, when keys are {@link Comparable},
99	>	* this class may use comparison order among keys to help break ties.
100	>	*
101	>	* <p>A {@link Set} projection of a ConcurrentHashMap may be created
102	>	* (using {@link #newKeySet()} or {@link #newKeySet(int)}), or viewed
103	>	* (using {@link #keySet(Object)} when only keys are of interest, and the
104	>	* mapped values are (perhaps transiently) not used or all take the
105	>	* same mapping value.
106	>	*
107	>	* <p>A ConcurrentHashMap can be used as scalable frequency map (a
108	>	* form of histogram or multiset) by using {@link
109	>	* java.util.concurrent.atomic.LongAdder} values and initializing via
110	>	* {@link #computeIfAbsent computeIfAbsent}. For example, to add a count
111	>	* to a {@code ConcurrentHashMap<String,LongAdder> freqs}, you can use
112	>	* {@code freqs.computeIfAbsent(k -> new LongAdder()).increment();}
113		*
114		* <p>This class and its views and iterators implement all of the
115		* <em>optional</em> methods of the {@link Map} and {@link Iterator}
116		* interfaces.
117		*
118	<	* <p> Like {@link Hashtable} but unlike {@link HashMap}, this class
119	<	* does <em>not</em> allow <tt>null</tt> to be used as a key or value.
118	>	* <p>Like {@link Hashtable} but unlike {@link HashMap}, this class
119	>	* does <em>not</em> allow {@code null} to be used as a key or value.
120	>	*
121	>	* <p>ConcurrentHashMaps support a set of sequential and parallel bulk
122	>	* operations that, unlike most {@link Stream} methods, are designed
123	>	* to be safely, and often sensibly, applied even with maps that are
124	>	* being concurrently updated by other threads; for example, when
125	>	* computing a snapshot summary of the values in a shared registry.
126	>	* There are three kinds of operation, each with four forms, accepting
127	>	* functions with Keys, Values, Entries, and (Key, Value) arguments
128	>	* and/or return values. Because the elements of a ConcurrentHashMap
129	>	* are not ordered in any particular way, and may be processed in
130	>	* different orders in different parallel executions, the correctness
131	>	* of supplied functions should not depend on any ordering, or on any
132	>	* other objects or values that may transiently change while
133	>	* computation is in progress; and except for forEach actions, should
134	>	* ideally be side-effect-free. Bulk operations on {@link Map.Entry}
135	>	* objects do not support method {@code setValue}.
136	>	*
137	>	* <ul>
138	>	* <li> forEach: Perform a given action on each element.
139	>	* A variant form applies a given transformation on each element
140	>	* before performing the action.</li>
141	>	*
142	>	* <li> search: Return the first available non-null result of
143	>	* applying a given function on each element; skipping further
144	>	* search when a result is found.</li>
145	>	*
146	>	* <li> reduce: Accumulate each element.  The supplied reduction
147	>	* function cannot rely on ordering (more formally, it should be
148	>	* both associative and commutative).  There are five variants:
149	>	*
150	>	* <ul>
151	>	*
152	>	* <li> Plain reductions. (There is not a form of this method for
153	>	* (key, value) function arguments since there is no corresponding
154	>	* return type.)</li>
155	>	*
156	>	* <li> Mapped reductions that accumulate the results of a given
157	>	* function applied to each element.</li>
158	>	*
159	>	* <li> Reductions to scalar doubles, longs, and ints, using a
160	>	* given basis value.</li>
161	>	*
162	>	* </ul>
163	>	* </li>
164	>	* </ul>
165	>	*
166	>	* <p>These bulk operations accept a {@code parallelismThreshold}
167	>	* argument. Methods proceed sequentially if the current map size is
168	>	* estimated to be less than the given threshold. Using a value of
169	>	* {@code Long.MAX_VALUE} suppresses all parallelism.  Using a value
170	>	* of {@code 1} results in maximal parallelism by partitioning into
171	>	* enough subtasks to fully utilize the {@link
172	>	* ForkJoinPool#commonPool()} that is used for all parallel
173	>	* computations. Normally, you would initially choose one of these
174	>	* extreme values, and then measure performance of using in-between
175	>	* values that trade off overhead versus throughput.
176	>	*
177	>	* <p>The concurrency properties of bulk operations follow
178	>	* from those of ConcurrentHashMap: Any non-null result returned
179	>	* from {@code get(key)} and related access methods bears a
180	>	* happens-before relation with the associated insertion or
181	>	* update.  The result of any bulk operation reflects the
182	>	* composition of these per-element relations (but is not
183	>	* necessarily atomic with respect to the map as a whole unless it
184	>	* is somehow known to be quiescent).  Conversely, because keys
185	>	* and values in the map are never null, null serves as a reliable
186	>	* atomic indicator of the current lack of any result.  To
187	>	* maintain this property, null serves as an implicit basis for
188	>	* all non-scalar reduction operations. For the double, long, and
189	>	* int versions, the basis should be one that, when combined with
190	>	* any other value, returns that other value (more formally, it
191	>	* should be the identity element for the reduction). Most common
192	>	* reductions have these properties; for example, computing a sum
193	>	* with basis 0 or a minimum with basis MAX_VALUE.
194	>	*
195	>	* <p>Search and transformation functions provided as arguments
196	>	* should similarly return null to indicate the lack of any result
197	>	* (in which case it is not used). In the case of mapped
198	>	* reductions, this also enables transformations to serve as
199	>	* filters, returning null (or, in the case of primitive
200	>	* specializations, the identity basis) if the element should not
201	>	* be combined. You can create compound transformations and
202	>	* filterings by composing them yourself under this "null means
203	>	* there is nothing there now" rule before using them in search or
204	>	* reduce operations.
205	>	*
206	>	* <p>Methods accepting and/or returning Entry arguments maintain
207	>	* key-value associations. They may be useful for example when
208	>	* finding the key for the greatest value. Note that "plain" Entry
209	>	* arguments can be supplied using {@code new
210	>	* AbstractMap.SimpleEntry(k,v)}.
211	>	*
212	>	* <p>Bulk operations may complete abruptly, throwing an
213	>	* exception encountered in the application of a supplied
214	>	* function. Bear in mind when handling such exceptions that other
215	>	* concurrently executing functions could also have thrown
216	>	* exceptions, or would have done so if the first exception had
217	>	* not occurred.
218	>	*
219	>	* <p>Speedups for parallel compared to sequential forms are common
220	>	* but not guaranteed.  Parallel operations involving brief functions
221	>	* on small maps may execute more slowly than sequential forms if the
222	>	* underlying work to parallelize the computation is more expensive
223	>	* than the computation itself.  Similarly, parallelization may not
224	>	* lead to much actual parallelism if all processors are busy
225	>	* performing unrelated tasks.
226	>	*
227	>	* <p>All arguments to all task methods must be non-null.
228		*
229		* <p>This class is a member of the
230		* <a href="{@docRoot}/../technotes/guides/collections/index.html">
#	Line 70 | Line 235 | import java.io.ObjectOutputStream;
235		* @param <K> the type of keys maintained by this map
236		* @param <V> the type of mapped values
237		*/
238	<	public class ConcurrentHashMap<K, V> extends AbstractMap<K, V>
74	<	implements ConcurrentMap<K, V>, Serializable {
238	>	public class ConcurrentHashMap<K,V> implements ConcurrentMap<K,V>, Serializable {
239		private static final long serialVersionUID = 7249069246763182397L;
240
241		/*
242	<	* The basic strategy is to subdivide the table among Segments,
243	<	* each of which itself is a concurrently readable hash table.  To
244	<	* reduce footprint, all but one segments are constructed only
245	<	* when first needed (see ensureSegment). To maintain visibility
246	<	* in the presence of lazy construction, accesses to segments as
247	<	* well as elements of segment's table must use volatile access,
248	<	* which is done via Unsafe within methods segmentAt etc
249	<	* below. These provide the functionality of AtomicReferenceArrays
250	<	* but reduce the levels of indirection. Additionally,
251	<	* volatile-writes of table elements and entry "next" fields
252	<	* within locked operations use the cheaper "lazySet" forms of
253	<	* writes (via putOrderedObject) because these writes are always
254	<	* followed by lock releases that maintain sequential consistency
255	<	* of table updates.
256	<	*
257	<	* Historical note: The previous version of this class relied
258	<	* heavily on "final" fields, which avoided some volatile reads at
259	<	* the expense of a large initial footprint.  Some remnants of
260	<	* that design (including forced construction of segment 0) exist
261	<	* to ensure serialization compatibility.
242	>	* Overview:
243	>	*
244	>	* The primary design goal of this hash table is to maintain
245	>	* concurrent readability (typically method get(), but also
246	>	* iterators and related methods) while minimizing update
247	>	* contention. Secondary goals are to keep space consumption about
248	>	* the same or better than java.util.HashMap, and to support high
249	>	* initial insertion rates on an empty table by many threads.
250	>	*
251	>	* This map usually acts as a binned (bucketed) hash table.
252	>	* Each key-value mapping is held in a Node.  Most nodes are
253	>	* instances of the basic Node class with hash, key, value, and
254	>	* next fields. However, various subclasses exist: TreeNodes are
255	>	* arranged in balanced trees, not lists.  TreeBins hold the roots
256	>	* of sets of TreeNodes. ForwardingNodes are placed at the heads
257	>	* of bins during resizing. ReservationNodes are used as
258	>	* placeholders while establishing values in computeIfAbsent and
259	>	* related methods.  The three type TreeBin, ForwardingNode, and
260	>	* ReservationNode do not hold normal user keys, values, or
261	>	* hashes, and are readily distinguishable during search etc
262	>	* because they have negative hash fields and null key and value
263	>	* fields. (These special nodes are either uncommon or transient,
264	>	* so the impact of carrying around some unused fields is
265	>	* insignficant.)
266	>	*
267	>	* The table is lazily initialized to a power-of-two size upon the
268	>	* first insertion.  Each bin in the table normally contains a
269	>	* list of Nodes (most often, the list has only zero or one Node).
270	>	* Table accesses require volatile/atomic reads, writes, and
271	>	* CASes.  Because there is no other way to arrange this without
272	>	* adding further indirections, we use intrinsics
273	>	* (sun.misc.Unsafe) operations.
274	>	*
275	>	* We use the top (sign) bit of Node hash fields for control
276	>	* purposes -- it is available anyway because of addressing
277	>	* constraints.  Nodes with negative hash fields are specially
278	>	* handled or ignored in map methods.
279	>	*
280	>	* Insertion (via put or its variants) of the first node in an
281	>	* empty bin is performed by just CASing it to the bin.  This is
282	>	* by far the most common case for put operations under most
283	>	* key/hash distributions.  Other update operations (insert,
284	>	* delete, and replace) require locks.  We do not want to waste
285	>	* the space required to associate a distinct lock object with
286	>	* each bin, so instead use the first node of a bin list itself as
287	>	* a lock. Locking support for these locks relies on builtin
288	>	* "synchronized" monitors.
289	>	*
290	>	* Using the first node of a list as a lock does not by itself
291	>	* suffice though: When a node is locked, any update must first
292	>	* validate that it is still the first node after locking it, and
293	>	* retry if not. Because new nodes are always appended to lists,
294	>	* once a node is first in a bin, it remains first until deleted
295	>	* or the bin becomes invalidated (upon resizing).
296	>	*
297	>	* The main disadvantage of per-bin locks is that other update
298	>	* operations on other nodes in a bin list protected by the same
299	>	* lock can stall, for example when user equals() or mapping
300	>	* functions take a long time.  However, statistically, under
301	>	* random hash codes, this is not a common problem.  Ideally, the
302	>	* frequency of nodes in bins follows a Poisson distribution
303	>	* (http://en.wikipedia.org/wiki/Poisson_distribution) with a
304	>	* parameter of about 0.5 on average, given the resizing threshold
305	>	* of 0.75, although with a large variance because of resizing
306	>	* granularity. Ignoring variance, the expected occurrences of
307	>	* list size k are (exp(-0.5) * pow(0.5, k) / factorial(k)). The
308	>	* first values are:
309	>	*
310	>	* 0:    0.60653066
311	>	* 1:    0.30326533
312	>	* 2:    0.07581633
313	>	* 3:    0.01263606
314	>	* 4:    0.00157952
315	>	* 5:    0.00015795
316	>	* 6:    0.00001316
317	>	* 7:    0.00000094
318	>	* 8:    0.00000006
319	>	* more: less than 1 in ten million
320	>	*
321	>	* Lock contention probability for two threads accessing distinct
322	>	* elements is roughly 1 / (8 * #elements) under random hashes.
323	>	*
324	>	* Actual hash code distributions encountered in practice
325	>	* sometimes deviate significantly from uniform randomness.  This
326	>	* includes the case when N > (1<<30), so some keys MUST collide.
327	>	* Similarly for dumb or hostile usages in which multiple keys are
328	>	* designed to have identical hash codes or ones that differs only
329	>	* in high bits. So we use a secondary strategy that applies when
330	>	* the number of nodes in a bin exceeds a threshold. These
331	>	* TreeBins use a balanced tree to hold nodes (a specialized form
332	>	* of red-black trees), bounding search time to O(log N).  Each
333	>	* search step in a TreeBin is at least twice as slow as in a
334	>	* regular list, but given that N cannot exceed (1<<64) (before
335	>	* running out of addresses) this bounds search steps, lock hold
336	>	* times, etc, to reasonable constants (roughly 100 nodes
337	>	* inspected per operation worst case) so long as keys are
338	>	* Comparable (which is very common -- String, Long, etc).
339	>	* TreeBin nodes (TreeNodes) also maintain the same "next"
340	>	* traversal pointers as regular nodes, so can be traversed in
341	>	* iterators in the same way.
342	>	*
343	>	* The table is resized when occupancy exceeds a percentage
344	>	* threshold (nominally, 0.75, but see below).  Any thread
345	>	* noticing an overfull bin may assist in resizing after the
346	>	* initiating thread allocates and sets up the replacement
347	>	* array. However, rather than stalling, these other threads may
348	>	* proceed with insertions etc.  The use of TreeBins shields us
349	>	* from the worst case effects of overfilling while resizes are in
350	>	* progress.  Resizing proceeds by transferring bins, one by one,
351	>	* from the table to the next table. To enable concurrency, the
352	>	* next table must be (incrementally) prefilled with place-holders
353	>	* serving as reverse forwarders to the old table.  Because we are
354	>	* using power-of-two expansion, the elements from each bin must
355	>	* either stay at same index, or move with a power of two
356	>	* offset. We eliminate unnecessary node creation by catching
357	>	* cases where old nodes can be reused because their next fields
358	>	* won't change.  On average, only about one-sixth of them need
359	>	* cloning when a table doubles. The nodes they replace will be
360	>	* garbage collectable as soon as they are no longer referenced by
361	>	* any reader thread that may be in the midst of concurrently
362	>	* traversing table.  Upon transfer, the old table bin contains
363	>	* only a special forwarding node (with hash field "MOVED") that
364	>	* contains the next table as its key. On encountering a
365	>	* forwarding node, access and update operations restart, using
366	>	* the new table.
367	>	*
368	>	* Each bin transfer requires its bin lock, which can stall
369	>	* waiting for locks while resizing. However, because other
370	>	* threads can join in and help resize rather than contend for
371	>	* locks, average aggregate waits become shorter as resizing
372	>	* progresses.  The transfer operation must also ensure that all
373	>	* accessible bins in both the old and new table are usable by any
374	>	* traversal.  This is arranged by proceeding from the last bin
375	>	* (table.length - 1) up towards the first.  Upon seeing a
376	>	* forwarding node, traversals (see class Traverser) arrange to
377	>	* move to the new table without revisiting nodes.  However, to
378	>	* ensure that no intervening nodes are skipped, bin splitting can
379	>	* only begin after the associated reverse-forwarders are in
380	>	* place.
381	>	*
382	>	* The traversal scheme also applies to partial traversals of
383	>	* ranges of bins (via an alternate Traverser constructor)
384	>	* to support partitioned aggregate operations.  Also, read-only
385	>	* operations give up if ever forwarded to a null table, which
386	>	* provides support for shutdown-style clearing, which is also not
387	>	* currently implemented.
388	>	*
389	>	* Lazy table initialization minimizes footprint until first use,
390	>	* and also avoids resizings when the first operation is from a
391	>	* putAll, constructor with map argument, or deserialization.
392	>	* These cases attempt to override the initial capacity settings,
393	>	* but harmlessly fail to take effect in cases of races.
394	>	*
395	>	* The element count is maintained using a specialization of
396	>	* LongAdder. We need to incorporate a specialization rather than
397	>	* just use a LongAdder in order to access implicit
398	>	* contention-sensing that leads to creation of multiple
399	>	* CounterCells.  The counter mechanics avoid contention on
400	>	* updates but can encounter cache thrashing if read too
401	>	* frequently during concurrent access. To avoid reading so often,
402	>	* resizing under contention is attempted only upon adding to a
403	>	* bin already holding two or more nodes. Under uniform hash
404	>	* distributions, the probability of this occurring at threshold
405	>	* is around 13%, meaning that only about 1 in 8 puts check
406	>	* threshold (and after resizing, many fewer do so).
407	>	*
408	>	* TreeBins use a special form of comparison for search and
409	>	* related operations (which is the main reason we cannot use
410	>	* existing collections such as TreeMaps). TreeBins contain
411	>	* Comparable elements, but may contain others, as well as
412	>	* elements that are Comparable but not necessarily Comparable
413	>	* for the same T, so we cannot invoke compareTo among them. To
414	>	* handle this, the tree is ordered primarily by hash value, then
415	>	* by Comparable.compareTo order if applicable.  On lookup at a
416	>	* node, if elements are not comparable or compare as 0 then both
417	>	* left and right children may need to be searched in the case of
418	>	* tied hash values. (This corresponds to the full list search
419	>	* that would be necessary if all elements were non-Comparable and
420	>	* had tied hashes.)  The red-black balancing code is updated from
421	>	* pre-jdk-collections
422	>	* (http://gee.cs.oswego.edu/dl/classes/collections/RBCell.java)
423	>	* based in turn on Cormen, Leiserson, and Rivest "Introduction to
424	>	* Algorithms" (CLR).
425	>	*
426	>	* TreeBins also require an additional locking mechanism.  While
427	>	* list traversal is always possible by readers evern during
428	>	* updates, tree traversal is not, mainly beause of tree-rotations
429	>	* that may change the root node and/or its linkages.  TreeBins
430	>	* include a simple read-write lock mechanism parasitic on the
431	>	* main bin-synchronization strategy: Structural adjustments
432	>	* associated with an insertion or removal are already bin-locked
433	>	* (and so cannot conflict with other writers) but must wait for
434	>	* ongoing readers to finish. Since there can be only one such
435	>	* waiter, we use a simple scheme using a single "waiter" field to
436	>	* block writers.  However, readers need never block.  If the root
437	>	* lock is held, they proceed along the slow traversal path (via
438	>	* next-pointers) until the lock becomes available or the list is
439	>	* exhausted, whichever comes first. These cases are not fast, but
440	>	* maximize aggregate expected throughput.
441	>	*
442	>	* Maintaining API and serialization compatibility with previous
443	>	* versions of this class introduces several oddities. Mainly: We
444	>	* leave untouched but unused constructor arguments refering to
445	>	* concurrencyLevel. We accept a loadFactor constructor argument,
446	>	* but apply it only to initial table capacity (which is the only
447	>	* time that we can guarantee to honor it.) We also declare an
448	>	* unused "Segment" class that is instantiated in minimal form
449	>	* only when serializing.
450	>	*
451	>	* This file is organized to make things a little easier to follow
452	>	* while reading than they might otherwise:. First the main static
453	>	* declarations and utilities, then fields, then main public
454	>	* methods (with a few factorings of multiple public methods into
455	>	* internal ones), then sizing methods, trees, traversers, and
456	>	* bulk operations.
457		*/
458
459		/* ---------------- Constants -------------- */
460
461		/**
462	<	* The default initial capacity for this table,
463	<	* used when not otherwise specified in a constructor.
462	>	* The largest possible table capacity.  This value must be
463	>	* exactly 1<<30 to stay within Java array allocation and indexing
464	>	* bounds for power of two table sizes, and is further required
465	>	* because the top two bits of 32bit hash fields are used for
466	>	* control purposes.
467		*/
468	<	static final int DEFAULT_INITIAL_CAPACITY = 16;
468	>	private static final int MAXIMUM_CAPACITY = 1 << 30;
469
470		/**
471	<	* The default load factor for this table, used when not
472	<	* otherwise specified in a constructor.
471	>	* The default initial table capacity.  Must be a power of 2
472	>	* (i.e., at least 1) and at most MAXIMUM_CAPACITY.
473		*/
474	<	static final float DEFAULT_LOAD_FACTOR = 0.75f;
474	>	private static final int DEFAULT_CAPACITY = 16;
475
476		/**
477	<	* The default concurrency level for this table, used when not
478	<	* otherwise specified in a constructor.
477	>	* The largest possible (non-power of two) array size.
478	>	* Needed by toArray and related methods.
479		*/
480	<	static final int DEFAULT_CONCURRENCY_LEVEL = 16;
480	>	static final int MAX_ARRAY_SIZE = Integer.MAX_VALUE - 8;
481
482		/**
483	<	* The maximum capacity, used if a higher value is implicitly
484	<	* specified by either of the constructors with arguments.  MUST
123	<	* be a power of two <= 1<<30 to ensure that entries are indexable
124	<	* using ints.
483	>	* The default concurrency level for this table. Unused but
484	>	* defined for compatibility with previous versions of this class.
485		*/
486	<	static final int MAXIMUM_CAPACITY = 1 << 30;
486	>	private static final int DEFAULT_CONCURRENCY_LEVEL = 16;
487
488		/**
489	<	* The minimum capacity for per-segment tables.  Must be a power
490	<	* of two, at least two to avoid immediate resizing on next use
491	<	* after lazy construction.
489	>	* The load factor for this table. Overrides of this value in
490	>	* constructors affect only the initial table capacity.  The
491	>	* actual floating point value isn't normally used -- it is
492	>	* simpler to use expressions such as {@code n - (n >>> 2)} for
493	>	* the associated resizing threshold.
494		*/
495	<	static final int MIN_SEGMENT_TABLE_CAPACITY = 2;
495	>	private static final float LOAD_FACTOR = 0.75f;
496
497		/**
498	<	* The maximum number of segments to allow; used to bound
499	<	* constructor arguments. Must be power of two less than 1 << 24.
498	>	* The bin count threshold for using a tree rather than list for a
499	>	* bin.  Bins are converted to trees when adding an element to a
500	>	* bin with at least this many nodes. The value should be at least
501	>	* 8 to mesh with assumptions in tree removal about conversion
502	>	* back to plain bins upon shrinkage.
503		*/
504	<	static final int MAX_SEGMENTS = 1 << 16; // slightly conservative
504	>	static final int TREEIFY_THRESHOLD = 8;
505
506		/**
507	<	* Number of unsynchronized retries in size and containsValue
508	<	* methods before resorting to locking. This is used to avoid
509	<	* unbounded retries if tables undergo continuous modification
145	<	* which would make it impossible to obtain an accurate result.
507	>	* The bin count threshold for untreeifying a (split) bin during a
508	>	* resize operation. Should be less than TREEIFY_THRESHOLD, and at
509	>	* most 6 to mesh with shrinkage detection under removal.
510		*/
511	<	static final int RETRIES_BEFORE_LOCK = 2;
148	<
149	<	/* ---------------- Fields -------------- */
511	>	static final int UNTREEIFY_THRESHOLD = 6;
512
513		/**
514	<	* Mask value for indexing into segments. The upper bits of a
515	<	* key's hash code are used to choose the segment.
514	>	* The smallest table capacity for which bins may be treeified.
515	>	* (Otherwise the table is resized if too many nodes in a bin.)
516	>	* Should be at least 4 * TREEIFY_THRESHOLD to avoid conflicts
517	>	* between resizing and treeification thresholds.
518		*/
519	<	final int segmentMask;
519	>	static final int MIN_TREEIFY_CAPACITY = 64;
520
521		/**
522	<	* Shift value for indexing within segments.
522	>	* Minimum number of rebinnings per transfer step. Ranges are
523	>	* subdivided to allow multiple resizer threads.  This value
524	>	* serves as a lower bound to avoid resizers encountering
525	>	* excessive memory contention.  The value should be at least
526	>	* DEFAULT_CAPACITY.
527		*/
528	<	final int segmentShift;
528	>	private static final int MIN_TRANSFER_STRIDE = 16;
529
530	<	/**
531	<	* The segments, each of which is a specialized hash table.
530	>	/*
531	>	* Encodings for Node hash fields. See above for explanation.
532		*/
533	<	final Segment<K,V>[] segments;
534	<
535	<	transient Set<K> keySet;
536	<	transient Set<Map.Entry<K,V>> entrySet;
537	<	transient Collection<V> values;
538	<
539	<	/**
540	<	* ConcurrentHashMap list entry. Note that this is never exported
541	<	* out as a user-visible Map.Entry.
533	>	static final int MOVED     = 0x8fffffff; // (-1) hash for forwarding nodes
534	>	static final int TREEBIN   = 0x80000000; // hash for heads of treea
535	>	static final int RESERVED  = 0x80000001; // hash for transient reservations
536	>	static final int HASH_BITS = 0x7fffffff; // usable bits of normal node hash
537	>
538	>	/** Number of CPUS, to place bounds on some sizings */
539	>	static final int NCPU = Runtime.getRuntime().availableProcessors();
540	>
541	>	/** For serialization compatibility. */
542	>	private static final ObjectStreamField[] serialPersistentFields = {
543	>	new ObjectStreamField("segments", Segment[].class),
544	>	new ObjectStreamField("segmentMask", Integer.TYPE),
545	>	new ObjectStreamField("segmentShift", Integer.TYPE)
546	>	};
547	>
548	>	/* ---------------- Nodes -------------- */
549	>
550	>	/**
551	>	* Key-value entry.  This class is never exported out as a
552	>	* user-mutable Map.Entry (i.e., one supporting setValue; see
553	>	* MapEntry below), but can be used for read-only traversals used
554	>	* in bulk tasks.  Subclasses of Node with a hash field of MOVED are special,
555	>	* and contain null keys and values (but are never exported).
556	>	* Otherwise, keys and vals are never null.
557		*/
558	<	static final class HashEntry<K,V> {
558	>	static class Node<K,V> implements Map.Entry<K,V> {
559		final int hash;
560		final K key;
561	<	volatile V value;
562	<	volatile HashEntry<K,V> next;
561	>	volatile V val;
562	>	Node<K,V> next;
563
564	<	HashEntry(int hash, K key, V value, HashEntry<K,V> next) {
564	>	Node(int hash, K key, V val, Node<K,V> next) {
565		this.hash = hash;
566		this.key = key;
567	<	this.value = value;
567	>	this.val = val;
568		this.next = next;
569		}
570
571	<	/**
572	<	* Sets next field with volatile write semantics.  (See above
573	<	* about use of putOrderedObject.)
574	<	*/
575	<	final void setNext(HashEntry<K,V> n) {
576	<	UNSAFE.putOrderedObject(this, nextOffset, n);
571	>	public final K getKey()       { return key; }
572	>	public final V getValue()     { return val; }
573	>	public final int hashCode()   { return key.hashCode() ^ val.hashCode(); }
574	>	public final String toString(){ return key + "=" + val; }
575	>	public final V setValue(V value) {
576	>	throw new UnsupportedOperationException();
577	>	}
578	>
579	>	public final boolean equals(Object o) {
580	>	Object k, v, u; Map.Entry<?,?> e;
581	>	return ((o instanceof Map.Entry) &&
582	>	(k = (e = (Map.Entry<?,?>)o).getKey()) != null &&
583	>	(v = e.getValue()) != null &&
584	>	(k == key || k.equals(key)) &&
585	>	(v == (u = val) || v.equals(u)));
586	>	}
587	>
588	>	Node<K,V> find(int h, Object k) {
589	>	Node<K,V> e = this;
590	>	if (k != null) {
591	>	do {
592	>	K ek;
593	>	if (e.hash == h &&
594	>	((ek = e.key) == k || (ek != null && k.equals(ek))))
595	>	return e;
596	>	} while ((e = e.next) != null);
597	>	}
598	>	return null;
599		}
600	+	}
601
602	<	// Unsafe mechanics
603	<	static final sun.misc.Unsafe UNSAFE;
604	<	static final long nextOffset;
605	<	static {
606	<	try {
607	<	UNSAFE = sun.misc.Unsafe.getUnsafe();
608	<	Class k = HashEntry.class;
609	<	nextOffset = UNSAFE.objectFieldOffset
610	<	(k.getDeclaredField("next"));
611	<	} catch (Exception e) {
612	<	throw new Error(e);
602	>	/* ---------------- Static utilities -------------- */
603	>
604	>	/**
605	>	* Spreads (XORs) higher bits of hash to lower and also forces top
606	>	* bit to 0. Because the table uses power-of-two masking, sets of
607	>	* hashes that vary only in bits above the current mask will
608	>	* always collide. (Among known examples are sets of Float keys
609	>	* holding consecutive whole numbers in small tables.)  So we
610	>	* apply a transform that spreads the impact of higher bits
611	>	* downward. There is a tradeoff between speed, utility, and
612	>	* quality of bit-spreading. Because many common sets of hashes
613	>	* are already reasonably distributed (so don't benefit from
614	>	* spreading), and because we use trees to handle large sets of
615	>	* collisions in bins, we just XOR some shifted bits in the
616	>	* cheapest possible way to reduce systematic lossage, as well as
617	>	* to incorporate impact of the highest bits that would otherwise
618	>	* never be used in index calculations because of table bounds.
619	>	*/
620	>	static final int spread(int h) {
621	>	return (h ^ (h >>> 16)) & HASH_BITS;
622	>	}
623	>
624	>	/**
625	>	* Returns a power of two table size for the given desired capacity.
626	>	* See Hackers Delight, sec 3.2
627	>	*/
628	>	private static final int tableSizeFor(int c) {
629	>	int n = c - 1;
630	>	n |= n >>> 1;
631	>	n |= n >>> 2;
632	>	n |= n >>> 4;
633	>	n |= n >>> 8;
634	>	n |= n >>> 16;
635	>	return (n < 0) ? 1 : (n >= MAXIMUM_CAPACITY) ? MAXIMUM_CAPACITY : n + 1;
636	>	}
637	>
638	>	/**
639	>	* Returns x's Class if it is of the form "class C implements
640	>	* Comparable<C>", else null.
641	>	*/
642	>	static Class<?> comparableClassFor(Object x) {
643	>	if (x instanceof Comparable) {
644	>	Class<?> c; Type[] ts, as; Type t; ParameterizedType p;
645	>	if ((c = x.getClass()) == String.class) // bypass checks
646	>	return c;
647	>	if ((ts = c.getGenericInterfaces()) != null) {
648	>	for (int i = 0; i < ts.length; ++i) {
649	>	if (((t = ts[i]) instanceof ParameterizedType) &&
650	>	((p = (ParameterizedType)t).getRawType() ==
651	>	Comparable.class) &&
652	>	(as = p.getActualTypeArguments()) != null &&
653	>	as.length == 1 && as[0] == c) // type arg is c
654	>	return c;
655	>	}
656		}
657		}
658	+	return null;
659		}
660
661		/**
662	<	* Gets the ith element of given table (if nonnull) with volatile
663	<	* read semantics. Note: This is manually integrated into a few
664	<	* performance-sensitive methods to reduce call overhead.
662	>	* Returns k.compareTo(x) if x matches kc (k's screened comparable
663	>	* class), else 0.
664	>	*/
665	>	@SuppressWarnings({"rawtypes","unchecked"}) // for cast to Comparable
666	>	static int compareComparables(Class<?> kc, Object k, Object x) {
667	>	return (x == null || x.getClass() != kc ? 0 :
668	>	((Comparable)k).compareTo(x));
669	>	}
670	>
671	>	/* ---------------- Table element access -------------- */
672	>
673	>	/*
674	>	* Volatile access methods are used for table elements as well as
675	>	* elements of in-progress next table while resizing.  All uses of
676	>	* the tab arguments must be null checked by callers.  All callers
677	>	* also paranoically precheck that tab's length is not zero (or an
678	>	* equivalent check), thus ensuring that any index argument taking
679	>	* the form of a hash value anded with (length - 1) is a valid
680	>	* index.  Note that, to be correct wrt arbitrary concurrency
681	>	* errors by users, these checks must operate on local variables,
682	>	* which accounts for some odd-looking inline assignments below.
683	>	* Note that calls to setTabAt always occur within locked regions,
684	>	* and so do not need full volatile semantics, but still require
685	>	* ordering to maintain concurrent readability.
686		*/
687	+
688		@SuppressWarnings("unchecked")
689	<	static final <K,V> HashEntry<K,V> entryAt(HashEntry<K,V>[] tab, int i) {
690	<	return (tab == null) ? null :
691	<	(HashEntry<K,V>) UNSAFE.getObjectVolatile
692	<	(tab, ((long)i << TSHIFT) + TBASE);
689	>	static final <K,V> Node<K,V> tabAt(Node<K,V>[] tab, int i) {
690	>	return (Node<K,V>)U.getObjectVolatile(tab, ((long)i << ASHIFT) + ABASE);
691	>	}
692	>
693	>	static final <K,V> boolean casTabAt(Node<K,V>[] tab, int i,
694	>	Node<K,V> c, Node<K,V> v) {
695	>	return U.compareAndSwapObject(tab, ((long)i << ASHIFT) + ABASE, c, v);
696		}
697
698	+	static final <K,V> void setTabAt(Node<K,V>[] tab, int i, Node<K,V> v) {
699	+	U.putOrderedObject(tab, ((long)i << ASHIFT) + ABASE, v);
700	+	}
701	+
702	+	/* ---------------- Fields -------------- */
703	+
704	+	/**
705	+	* The array of bins. Lazily initialized upon first insertion.
706	+	* Size is always a power of two. Accessed directly by iterators.
707	+	*/
708	+	transient volatile Node<K,V>[] table;
709	+
710		/**
711	<	* Sets the ith element of given table, with volatile write
225	<	* semantics. (See above about use of putOrderedObject.)
711	>	* The next table to use; non-null only while resizing.
712		*/
713	<	static final <K,V> void setEntryAt(HashEntry<K,V>[] tab, int i,
714	<	HashEntry<K,V> e) {
715	<	UNSAFE.putOrderedObject(tab, ((long)i << TSHIFT) + TBASE, e);
713	>	private transient volatile Node<K,V>[] nextTable;
714	>
715	>	/**
716	>	* Base counter value, used mainly when there is no contention,
717	>	* but also as a fallback during table initialization
718	>	* races. Updated via CAS.
719	>	*/
720	>	private transient volatile long baseCount;
721	>
722	>	/**
723	>	* Table initialization and resizing control.  When negative, the
724	>	* table is being initialized or resized: -1 for initialization,
725	>	* else -(1 + the number of active resizing threads).  Otherwise,
726	>	* when table is null, holds the initial table size to use upon
727	>	* creation, or 0 for default. After initialization, holds the
728	>	* next element count value upon which to resize the table.
729	>	*/
730	>	private transient volatile int sizeCtl;
731	>
732	>	/**
733	>	* The next table index (plus one) to split while resizing.
734	>	*/
735	>	private transient volatile int transferIndex;
736	>
737	>	/**
738	>	* The least available table index to split while resizing.
739	>	*/
740	>	private transient volatile int transferOrigin;
741	>
742	>	/**
743	>	* Spinlock (locked via CAS) used when resizing and/or creating CounterCells.
744	>	*/
745	>	private transient volatile int cellsBusy;
746	>
747	>	/**
748	>	* Table of counter cells. When non-null, size is a power of 2.
749	>	*/
750	>	private transient volatile CounterCell[] counterCells;
751	>
752	>	// views
753	>	private transient KeySetView<K,V> keySet;
754	>	private transient ValuesView<K,V> values;
755	>	private transient EntrySetView<K,V> entrySet;
756	>
757	>
758	>	/* ---------------- Public operations -------------- */
759	>
760	>	/**
761	>	* Creates a new, empty map with the default initial table size (16).
762	>	*/
763	>	public ConcurrentHashMap() {
764		}
765
766		/**
767	<	* Applies a supplemental hash function to a given hashCode, which
768	<	* defends against poor quality hash functions.  This is critical
769	<	* because ConcurrentHashMap uses power-of-two length hash tables,
770	<	* that otherwise encounter collisions for hashCodes that do not
771	<	* differ in lower or upper bits.
767	>	* Creates a new, empty map with an initial table size
768	>	* accommodating the specified number of elements without the need
769	>	* to dynamically resize.
770	>	*
771	>	* @param initialCapacity The implementation performs internal
772	>	* sizing to accommodate this many elements.
773	>	* @throws IllegalArgumentException if the initial capacity of
774	>	* elements is negative
775		*/
776	<	private static int hash(int h) {
777	<	// Spread bits to regularize both segment and index locations,
778	<	// using variant of single-word Wang/Jenkins hash.
779	<	h += (h <<  15) ^ 0xffffcd7d;
780	<	h ^= (h >>> 10);
781	<	h += (h <<   3);
782	<	h ^= (h >>>  6);
246	<	h += (h <<   2) + (h << 14);
247	<	return h ^ (h >>> 16);
776	>	public ConcurrentHashMap(int initialCapacity) {
777	>	if (initialCapacity < 0)
778	>	throw new IllegalArgumentException();
779	>	int cap = ((initialCapacity >= (MAXIMUM_CAPACITY >>> 1)) ?
780	>	MAXIMUM_CAPACITY :
781	>	tableSizeFor(initialCapacity + (initialCapacity >>> 1) + 1));
782	>	this.sizeCtl = cap;
783		}
784
785		/**
786	<	* Segments are specialized versions of hash tables.  This
787	<	* subclasses from ReentrantLock opportunistically, just to
788	<	* simplify some locking and avoid separate construction.
786	>	* Creates a new map with the same mappings as the given map.
787	>	*
788	>	* @param m the map
789		*/
790	<	static final class Segment<K,V> extends ReentrantLock implements Serializable {
791	<	/*
792	<	* Segments maintain a table of entry lists that are always
793	<	* kept in a consistent state, so can be read (via volatile
259	<	* reads of segments and tables) without locking.  This
260	<	* requires replicating nodes when necessary during table
261	<	* resizing, so the old lists can be traversed by readers
262	<	* still using old version of table.
263	<	*
264	<	* This class defines only mutative methods requiring locking.
265	<	* Except as noted, the methods of this class perform the
266	<	* per-segment versions of ConcurrentHashMap methods.  (Other
267	<	* methods are integrated directly into ConcurrentHashMap
268	<	* methods.) These mutative methods use a form of controlled
269	<	* spinning on contention via methods scanAndLock and
270	<	* scanAndLockForPut. These intersperse tryLocks with
271	<	* traversals to locate nodes.  The main benefit is to absorb
272	<	* cache misses (which are very common for hash tables) while
273	<	* obtaining locks so that traversal is faster once
274	<	* acquired. We do not actually use the found nodes since they
275	<	* must be re-acquired under lock anyway to ensure sequential
276	<	* consistency of updates (and in any case may be undetectably
277	<	* stale), but they will normally be much faster to re-locate.
278	<	* Also, scanAndLockForPut speculatively creates a fresh node
279	<	* to use in put if no node is found.
280	<	*/
790	>	public ConcurrentHashMap(Map<? extends K, ? extends V> m) {
791	>	this.sizeCtl = DEFAULT_CAPACITY;
792	>	putAll(m);
793	>	}
794
795	<	private static final long serialVersionUID = 2249069246763182397L;
795	>	/**
796	>	* Creates a new, empty map with an initial table size based on
797	>	* the given number of elements ({@code initialCapacity}) and
798	>	* initial table density ({@code loadFactor}).
799	>	*
800	>	* @param initialCapacity the initial capacity. The implementation
801	>	* performs internal sizing to accommodate this many elements,
802	>	* given the specified load factor.
803	>	* @param loadFactor the load factor (table density) for
804	>	* establishing the initial table size
805	>	* @throws IllegalArgumentException if the initial capacity of
806	>	* elements is negative or the load factor is nonpositive
807	>	*
808	>	* @since 1.6
809	>	*/
810	>	public ConcurrentHashMap(int initialCapacity, float loadFactor) {
811	>	this(initialCapacity, loadFactor, 1);
812	>	}
813
814	<	/**
815	<	* The maximum number of times to tryLock in a prescan before
816	<	* possibly blocking on acquire in preparation for a locked
817	<	* segment operation. On multiprocessors, using a bounded
818	<	* number of retries maintains cache acquired while locating
819	<	* nodes.
820	<	*/
821	<	static final int MAX_SCAN_RETRIES =
822	<	Runtime.getRuntime().availableProcessors() > 1 ? 64 : 1;
814	>	/**
815	>	* Creates a new, empty map with an initial table size based on
816	>	* the given number of elements ({@code initialCapacity}), table
817	>	* density ({@code loadFactor}), and number of concurrently
818	>	* updating threads ({@code concurrencyLevel}).
819	>	*
820	>	* @param initialCapacity the initial capacity. The implementation
821	>	* performs internal sizing to accommodate this many elements,
822	>	* given the specified load factor.
823	>	* @param loadFactor the load factor (table density) for
824	>	* establishing the initial table size
825	>	* @param concurrencyLevel the estimated number of concurrently
826	>	* updating threads. The implementation may use this value as
827	>	* a sizing hint.
828	>	* @throws IllegalArgumentException if the initial capacity is
829	>	* negative or the load factor or concurrencyLevel are
830	>	* nonpositive
831	>	*/
832	>	public ConcurrentHashMap(int initialCapacity,
833	>	float loadFactor, int concurrencyLevel) {
834	>	if (!(loadFactor > 0.0f) || initialCapacity < 0 || concurrencyLevel <= 0)
835	>	throw new IllegalArgumentException();
836	>	if (initialCapacity < concurrencyLevel)   // Use at least as many bins
837	>	initialCapacity = concurrencyLevel;   // as estimated threads
838	>	long size = (long)(1.0 + (long)initialCapacity / loadFactor);
839	>	int cap = (size >= (long)MAXIMUM_CAPACITY) ?
840	>	MAXIMUM_CAPACITY : tableSizeFor((int)size);
841	>	this.sizeCtl = cap;
842	>	}
843
844	<	/**
295	<	* The per-segment table. Elements are accessed via
296	<	* entryAt/setEntryAt providing volatile semantics.
297	<	*/
298	<	transient volatile HashEntry<K,V>[] table;
844	>	// Original (since JDK1.2) Map methods
845
846	<	/**
847	<	* The number of elements. Accessed only either within locks
848	<	* or among other volatile reads that maintain visibility.
849	<	*/
850	<	transient int count;
846	>	/**
847	>	* {@inheritDoc}
848	>	*/
849	>	public int size() {
850	>	long n = sumCount();
851	>	return ((n < 0L) ? 0 :
852	>	(n > (long)Integer.MAX_VALUE) ? Integer.MAX_VALUE :
853	>	(int)n);
854	>	}
855
856	<	/**
857	<	* The total number of mutative operations in this segment.
858	<	* Even though this may overflows 32 bits, it provides
859	<	* sufficient accuracy for stability checks in CHM isEmpty()
860	<	* and size() methods.  Accessed only either within locks or
861	<	* among other volatile reads that maintain visibility.
312	<	*/
313	<	transient int modCount;
856	>	/**
857	>	* {@inheritDoc}
858	>	*/
859	>	public boolean isEmpty() {
860	>	return sumCount() <= 0L; // ignore transient negative values
861	>	}
862
863	<	/**
864	<	* The table is rehashed when its size exceeds this threshold.
865	<	* (The value of this field is always <tt>(int)(capacity *
866	<	* loadFactor)</tt>.)
867	<	*/
868	<	transient int threshold;
863	>	/**
864	>	* Returns the value to which the specified key is mapped,
865	>	* or {@code null} if this map contains no mapping for the key.
866	>	*
867	>	* <p>More formally, if this map contains a mapping from a key
868	>	* {@code k} to a value {@code v} such that {@code key.equals(k)},
869	>	* then this method returns {@code v}; otherwise it returns
870	>	* {@code null}.  (There can be at most one such mapping.)
871	>	*
872	>	* @throws NullPointerException if the specified key is null
873	>	*/
874	>	public V get(Object key) {
875	>	Node<K,V>[] tab; Node<K,V> e, p; int n, eh; K ek;
876	>	int h = spread(key.hashCode());
877	>	if ((tab = table) != null && (n = tab.length) > 0 &&
878	>	(e = tabAt(tab, (n - 1) & h)) != null) {
879	>	if ((eh = e.hash) == h) {
880	>	if ((ek = e.key) == key || (ek != null && key.equals(ek)))
881	>	return e.val;
882	>	}
883	>	else if (eh < 0)
884	>	return (p = e.find(h, key)) != null ? p.val : null;
885	>	while ((e = e.next) != null) {
886	>	if (e.hash == h &&
887	>	((ek = e.key) == key || (ek != null && key.equals(ek))))
888	>	return e.val;
889	>	}
890	>	}
891	>	return null;
892	>	}
893
894	<	/**
895	<	* The load factor for the hash table.  Even though this value
896	<	* is same for all segments, it is replicated to avoid needing
897	<	* links to outer object.
898	<	* @serial
899	<	*/
900	<	final float loadFactor;
894	>	/**
895	>	* Tests if the specified object is a key in this table.
896	>	*
897	>	* @param  key possible key
898	>	* @return {@code true} if and only if the specified object
899	>	*         is a key in this table, as determined by the
900	>	*         {@code equals} method; {@code false} otherwise
901	>	* @throws NullPointerException if the specified key is null
902	>	*/
903	>	public boolean containsKey(Object key) {
904	>	return get(key) != null;
905	>	}
906
907	<	Segment(float lf, int threshold, HashEntry<K,V>[] tab) {
908	<	this.loadFactor = lf;
909	<	this.threshold = threshold;
910	<	this.table = tab;
907	>	/**
908	>	* Returns {@code true} if this map maps one or more keys to the
909	>	* specified value. Note: This method may require a full traversal
910	>	* of the map, and is much slower than method {@code containsKey}.
911	>	*
912	>	* @param value value whose presence in this map is to be tested
913	>	* @return {@code true} if this map maps one or more keys to the
914	>	*         specified value
915	>	* @throws NullPointerException if the specified value is null
916	>	*/
917	>	public boolean containsValue(Object value) {
918	>	if (value == null)
919	>	throw new NullPointerException();
920	>	Node<K,V>[] t;
921	>	if ((t = table) != null) {
922	>	Traverser<K,V> it = new Traverser<K,V>(t, t.length, 0, t.length);
923	>	for (Node<K,V> p; (p = it.advance()) != null; ) {
924	>	V v;
925	>	if ((v = p.val) == value || (v != null && value.equals(v)))
926	>	return true;
927	>	}
928		}
929	+	return false;
930	+	}
931
932	<	final V put(K key, int hash, V value, boolean onlyIfAbsent) {
933	<	HashEntry<K,V> node = tryLock() ? null :
934	<	scanAndLockForPut(key, hash, value);
935	<	V oldValue;
936	<	try {
937	<	HashEntry<K,V>[] tab = table;
938	<	int index = (tab.length - 1) & hash;
939	<	HashEntry<K,V> first = entryAt(tab, index);
940	<	for (HashEntry<K,V> e = first;;) {
941	<	if (e != null) {
942	<	K k;
943	<	if ((k = e.key) == key ||
944	<	(e.hash == hash && key.equals(k))) {
945	<	oldValue = e.value;
946	<	if (!onlyIfAbsent) {
947	<	e.value = value;
948	<	++modCount;
932	>	/**
933	>	* Maps the specified key to the specified value in this table.
934	>	* Neither the key nor the value can be null.
935	>	*
936	>	* <p>The value can be retrieved by calling the {@code get} method
937	>	* with a key that is equal to the original key.
938	>	*
939	>	* @param key key with which the specified value is to be associated
940	>	* @param value value to be associated with the specified key
941	>	* @return the previous value associated with {@code key}, or
942	>	*         {@code null} if there was no mapping for {@code key}
943	>	* @throws NullPointerException if the specified key or value is null
944	>	*/
945	>	public V put(K key, V value) {
946	>	return putVal(key, value, false);
947	>	}
948	>
949	>	/** Implementation for put and putIfAbsent */
950	>	final V putVal(K key, V value, boolean onlyIfAbsent) {
951	>	if (key == null || value == null) throw new NullPointerException();
952	>	int hash = spread(key.hashCode());
953	>	int binCount = 0;
954	>	for (Node<K,V>[] tab = table;;) {
955	>	Node<K,V> f; int n, i, fh;
956	>	if (tab == null || (n = tab.length) == 0)
957	>	tab = initTable();
958	>	else if ((f = tabAt(tab, i = (n - 1) & hash)) == null) {
959	>	if (casTabAt(tab, i, null,
960	>	new Node<K,V>(hash, key, value, null)))
961	>	break;                   // no lock when adding to empty bin
962	>	}
963	>	else if ((fh = f.hash) == MOVED)
964	>	tab = helpTransfer(tab, f);
965	>	else {
966	>	V oldVal = null;
967	>	synchronized (f) {
968	>	if (tabAt(tab, i) == f) {
969	>	if (fh >= 0) {
970	>	binCount = 1;
971	>	for (Node<K,V> e = f;; ++binCount) {
972	>	K ek;
973	>	if (e.hash == hash &&
974	>	((ek = e.key) == key ||
975	>	(ek != null && key.equals(ek)))) {
976	>	oldVal = e.val;
977	>	if (!onlyIfAbsent)
978	>	e.val = value;
979	>	break;
980	>	}
981	>	Node<K,V> pred = e;
982	>	if ((e = e.next) == null) {
983	>	pred.next = new Node<K,V>(hash, key,
984	>	value, null);
985	>	break;
986	>	}
987	>	}
988	>	}
989	>	else if (f instanceof TreeBin) {
990	>	Node<K,V> p;
991	>	binCount = 2;
992	>	if ((p = ((TreeBin<K,V>)f).putTreeVal(hash, key,
993	>	value)) != null) {
994	>	oldVal = p.val;
995	>	if (!onlyIfAbsent)
996	>	p.val = value;
997		}
354	–	break;
998		}
356	–	e = e.next;
357	–	}
358	–	else {
359	–	if (node != null)
360	–	node.setNext(first);
361	–	else
362	–	node = new HashEntry<K,V>(hash, key, value, first);
363	–	int c = count + 1;
364	–	if (c > threshold && tab.length < MAXIMUM_CAPACITY)
365	–	rehash(node);
366	–	else
367	–	setEntryAt(tab, index, node);
368	–	++modCount;
369	–	count = c;
370	–	oldValue = null;
371	–	break;
372	–	}
373	–	}
374	–	} finally {
375	–	unlock();
376	–	}
377	–	return oldValue;
378	–	}
379	–
380	–	/**
381	–	* Doubles size of table and repacks entries, also adding the
382	–	* given node to new table
383	–	*/
384	–	@SuppressWarnings("unchecked")
385	–	private void rehash(HashEntry<K,V> node) {
386	–	/*
387	–	* Reclassify nodes in each list to new table.  Because we
388	–	* are using power-of-two expansion, the elements from
389	–	* each bin must either stay at same index, or move with a
390	–	* power of two offset. We eliminate unnecessary node
391	–	* creation by catching cases where old nodes can be
392	–	* reused because their next fields won't change.
393	–	* Statistically, at the default threshold, only about
394	–	* one-sixth of them need cloning when a table
395	–	* doubles. The nodes they replace will be garbage
396	–	* collectable as soon as they are no longer referenced by
397	–	* any reader thread that may be in the midst of
398	–	* concurrently traversing table. Entry accesses use plain
399	–	* array indexing because they are followed by volatile
400	–	* table write.
401	–	*/
402	–	HashEntry<K,V>[] oldTable = table;
403	–	int oldCapacity = oldTable.length;
404	–	int newCapacity = oldCapacity << 1;
405	–	threshold = (int)(newCapacity * loadFactor);
406	–	HashEntry<K,V>[] newTable =
407	–	(HashEntry<K,V>[]) new HashEntry[newCapacity];
408	–	int sizeMask = newCapacity - 1;
409	–	for (int i = 0; i < oldCapacity ; i++) {
410	–	HashEntry<K,V> e = oldTable[i];
411	–	if (e != null) {
412	–	HashEntry<K,V> next = e.next;
413	–	int idx = e.hash & sizeMask;
414	–	if (next == null)   //  Single node on list
415	–	newTable[idx] = e;
416	–	else { // Reuse consecutive sequence at same slot
417	–	HashEntry<K,V> lastRun = e;
418	–	int lastIdx = idx;
419	–	for (HashEntry<K,V> last = next;
420	–	last != null;
421	–	last = last.next) {
422	–	int k = last.hash & sizeMask;
423	–	if (k != lastIdx) {
424	–	lastIdx = k;
425	–	lastRun = last;
426	–	}
427	–	}
428	–	newTable[lastIdx] = lastRun;
429	–	// Clone remaining nodes
430	–	for (HashEntry<K,V> p = e; p != lastRun; p = p.next) {
431	–	V v = p.value;
432	–	int h = p.hash;
433	–	int k = h & sizeMask;
434	–	HashEntry<K,V> n = newTable[k];
435	–	newTable[k] = new HashEntry<K,V>(h, p.key, v, n);
436	–	}
437	–	}
438	–	}
439	–	}
440	–	int nodeIndex = node.hash & sizeMask; // add the new node
441	–	node.setNext(newTable[nodeIndex]);
442	–	newTable[nodeIndex] = node;
443	–	table = newTable;
444	–	}
445	–
446	–	/**
447	–	* Scans for a node containing given key while trying to
448	–	* acquire lock, creating and returning one if not found. Upon
449	–	* return, guarantees that lock is held. Unlike in most
450	–	* methods, calls to method equals are not screened: Since
451	–	* traversal speed doesn't matter, we might as well help warm
452	–	* up the associated code and accesses as well.
453	–	*
454	–	* @return a new node if key not found, else null
455	–	*/
456	–	private HashEntry<K,V> scanAndLockForPut(K key, int hash, V value) {
457	–	HashEntry<K,V> first = entryForHash(this, hash);
458	–	HashEntry<K,V> e = first;
459	–	HashEntry<K,V> node = null;
460	–	int retries = -1; // negative while locating node
461	–	while (!tryLock()) {
462	–	HashEntry<K,V> f; // to recheck first below
463	–	if (retries < 0) {
464	–	if (e == null) {
465	–	if (node == null) // speculatively create node
466	–	node = new HashEntry<K,V>(hash, key, value, null);
467	–	retries = 0;
999		}
469	–	else if (key.equals(e.key))
470	–	retries = 0;
471	–	else
472	–	e = e.next;
1000		}
1001	<	else if (++retries > MAX_SCAN_RETRIES) {
1002	<	lock();
1001	>	if (binCount != 0) {
1002	>	if (binCount >= TREEIFY_THRESHOLD)
1003	>	treeifyBin(tab, i);
1004	>	if (oldVal != null)
1005	>	return oldVal;
1006		break;
1007		}
478	–	else if ((retries & 1) == 0 &&
479	–	(f = entryForHash(this, hash)) != first) {
480	–	e = first = f; // re-traverse if entry changed
481	–	retries = -1;
482	–	}
483	–	}
484	–	return node;
485	–	}
486	–
487	–	/**
488	–	* Scans for a node containing the given key while trying to
489	–	* acquire lock for a remove or replace operation. Upon
490	–	* return, guarantees that lock is held.  Note that we must
491	–	* lock even if the key is not found, to ensure sequential
492	–	* consistency of updates.
493	–	*/
494	–	private void scanAndLock(Object key, int hash) {
495	–	// similar to but simpler than scanAndLockForPut
496	–	HashEntry<K,V> first = entryForHash(this, hash);
497	–	HashEntry<K,V> e = first;
498	–	int retries = -1;
499	–	while (!tryLock()) {
500	–	HashEntry<K,V> f;
501	–	if (retries < 0) {
502	–	if (e == null || key.equals(e.key))
503	–	retries = 0;
504	–	else
505	–	e = e.next;
506	–	}
507	–	else if (++retries > MAX_SCAN_RETRIES) {
508	–	lock();
509	–	break;
510	–	}
511	–	else if ((retries & 1) == 0 &&
512	–	(f = entryForHash(this, hash)) != first) {
513	–	e = first = f;
514	–	retries = -1;
515	–	}
1008		}
1009		}
1010	+	addCount(1L, binCount);
1011	+	return null;
1012	+	}
1013
1014	<	/**
1015	<	* Remove; match on key only if value null, else match both.
1016	<	*/
1017	<	final V remove(Object key, int hash, Object value) {
1018	<	if (!tryLock())
1019	<	scanAndLock(key, hash);
1020	<	V oldValue = null;
1021	<	try {
1022	<	HashEntry<K,V>[] tab = table;
1023	<	int index = (tab.length - 1) & hash;
1024	<	HashEntry<K,V> e = entryAt(tab, index);
1025	<	HashEntry<K,V> pred = null;
1026	<	while (e != null) {
1027	<	K k;
1028	<	HashEntry<K,V> next = e.next;
1029	<	if ((k = e.key) == key ||
1030	<	(e.hash == hash && key.equals(k))) {
1031	<	V v = e.value;
1032	<	if (value == null || value == v || value.equals(v)) {
1033	<	if (pred == null)
1034	<	setEntryAt(tab, index, next);
1035	<	else
1036	<	pred.setNext(next);
1037	<	++modCount;
1038	<	--count;
1039	<	oldValue = v;
1014	>	/**
1015	>	* Copies all of the mappings from the specified map to this one.
1016	>	* These mappings replace any mappings that this map had for any of the
1017	>	* keys currently in the specified map.
1018	>	*
1019	>	* @param m mappings to be stored in this map
1020	>	*/
1021	>	public void putAll(Map<? extends K, ? extends V> m) {
1022	>	tryPresize(m.size());
1023	>	for (Map.Entry<? extends K, ? extends V> e : m.entrySet())
1024	>	putVal(e.getKey(), e.getValue(), false);
1025	>	}
1026	>
1027	>	/**
1028	>	* Removes the key (and its corresponding value) from this map.
1029	>	* This method does nothing if the key is not in the map.
1030	>	*
1031	>	* @param  key the key that needs to be removed
1032	>	* @return the previous value associated with {@code key}, or
1033	>	*         {@code null} if there was no mapping for {@code key}
1034	>	* @throws NullPointerException if the specified key is null
1035	>	*/
1036	>	public V remove(Object key) {
1037	>	return replaceNode(key, null, null);
1038	>	}
1039	>
1040	>	/**
1041	>	* Implementation for the four public remove/replace methods:
1042	>	* Replaces node value with v, conditional upon match of cv if
1043	>	* non-null.  If resulting value is null, delete.
1044	>	*/
1045	>	final V replaceNode(Object key, V value, Object cv) {
1046	>	int hash = spread(key.hashCode());
1047	>	for (Node<K,V>[] tab = table;;) {
1048	>	Node<K,V> f; int n, i, fh;
1049	>	if (tab == null || (n = tab.length) == 0 ||
1050	>	(f = tabAt(tab, i = (n - 1) & hash)) == null)
1051	>	break;
1052	>	else if ((fh = f.hash) == MOVED)
1053	>	tab = helpTransfer(tab, f);
1054	>	else {
1055	>	V oldVal = null;
1056	>	boolean validated = false;
1057	>	synchronized (f) {
1058	>	if (tabAt(tab, i) == f) {
1059	>	if (fh >= 0) {
1060	>	validated = true;
1061	>	for (Node<K,V> e = f, pred = null;;) {
1062	>	K ek;
1063	>	if (e.hash == hash &&
1064	>	((ek = e.key) == key ||
1065	>	(ek != null && key.equals(ek)))) {
1066	>	V ev = e.val;
1067	>	if (cv == null || cv == ev ||
1068	>	(ev != null && cv.equals(ev))) {
1069	>	oldVal = ev;
1070	>	if (value != null)
1071	>	e.val = value;
1072	>	else if (pred != null)
1073	>	pred.next = e.next;
1074	>	else
1075	>	setTabAt(tab, i, e.next);
1076	>	}
1077	>	break;
1078	>	}
1079	>	pred = e;
1080	>	if ((e = e.next) == null)
1081	>	break;
1082	>	}
1083	>	}
1084	>	else if (f instanceof TreeBin) {
1085	>	validated = true;
1086	>	TreeBin<K,V> t = (TreeBin<K,V>)f;
1087	>	TreeNode<K,V> r, p;
1088	>	if ((r = t.root) != null &&
1089	>	(p = r.findTreeNode(hash, key, null)) != null) {
1090	>	V pv = p.val;
1091	>	if (cv == null || cv == pv ||
1092	>	(pv != null && cv.equals(pv))) {
1093	>	oldVal = pv;
1094	>	if (value != null)
1095	>	p.val = value;
1096	>	else if (t.removeTreeNode(p))
1097	>	setTabAt(tab, i, untreeify(t.first));
1098	>	}
1099	>	}
1100		}
546	–	break;
1101		}
548	–	pred = e;
549	–	e = next;
1102		}
1103	<	} finally {
1104	<	unlock();
1105	<	}
1106	<	return oldValue;
1107	<	}
556	<
557	<	final boolean replace(K key, int hash, V oldValue, V newValue) {
558	<	if (!tryLock())
559	<	scanAndLock(key, hash);
560	<	boolean replaced = false;
561	<	try {
562	<	HashEntry<K,V> e;
563	<	for (e = entryForHash(this, hash); e != null; e = e.next) {
564	<	K k;
565	<	if ((k = e.key) == key ||
566	<	(e.hash == hash && key.equals(k))) {
567	<	if (oldValue.equals(e.value)) {
568	<	e.value = newValue;
569	<	++modCount;
570	<	replaced = true;
571	<	}
572	<	break;
1103	>	if (validated) {
1104	>	if (oldVal != null) {
1105	>	if (value == null)
1106	>	addCount(-1L, -1);
1107	>	return oldVal;
1108		}
1109	+	break;
1110		}
575	–	} finally {
576	–	unlock();
1111		}
578	–	return replaced;
1112		}
1113	+	return null;
1114	+	}
1115
1116	<	final V replace(K key, int hash, V value) {
1117	<	if (!tryLock())
1118	<	scanAndLock(key, hash);
1119	<	V oldValue = null;
1120	<	try {
1121	<	HashEntry<K,V> e;
1122	<	for (e = entryForHash(this, hash); e != null; e = e.next) {
1123	<	K k;
1124	<	if ((k = e.key) == key ||
1125	<	(e.hash == hash && key.equals(k))) {
1126	<	oldValue = e.value;
1127	<	e.value = value;
1128	<	++modCount;
1129	<	break;
1116	>	/**
1117	>	* Removes all of the mappings from this map.
1118	>	*/
1119	>	public void clear() {
1120	>	long delta = 0L; // negative number of deletions
1121	>	int i = 0;
1122	>	Node<K,V>[] tab = table;
1123	>	while (tab != null && i < tab.length) {
1124	>	int fh;
1125	>	Node<K,V> f = tabAt(tab, i);
1126	>	if (f == null)
1127	>	++i;
1128	>	else if ((fh = f.hash) == MOVED) {
1129	>	tab = helpTransfer(tab, f);
1130	>	i = 0; // restart
1131	>	}
1132	>	else {
1133	>	synchronized (f) {
1134	>	if (tabAt(tab, i) == f) {
1135	>	Node<K,V> p = (fh >= 0 ? f :
1136	>	(f instanceof TreeBin) ?
1137	>	((TreeBin<K,V>)f).first : null);
1138	>	while (p != null) {
1139	>	--delta;
1140	>	p = p.next;
1141	>	}
1142	>	setTabAt(tab, i++, null);
1143		}
1144		}
597	–	} finally {
598	–	unlock();
1145		}
600	–	return oldValue;
1146		}
1147	+	if (delta != 0L)
1148	+	addCount(delta, -1);
1149	+	}
1150
1151	<	final void clear() {
1152	<	lock();
1153	<	try {
1154	<	HashEntry<K,V>[] tab = table;
1155	<	for (int i = 0; i < tab.length ; i++)
1156	<	setEntryAt(tab, i, null);
1157	<	++modCount;
1158	<	count = 0;
1159	<	} finally {
1160	<	unlock();
1161	<	}
1162	<	}
1151	>	/**
1152	>	* Returns a {@link Set} view of the keys contained in this map.
1153	>	* The set is backed by the map, so changes to the map are
1154	>	* reflected in the set, and vice-versa. The set supports element
1155	>	* removal, which removes the corresponding mapping from this map,
1156	>	* via the {@code Iterator.remove}, {@code Set.remove},
1157	>	* {@code removeAll}, {@code retainAll}, and {@code clear}
1158	>	* operations.  It does not support the {@code add} or
1159	>	* {@code addAll} operations.
1160	>	*
1161	>	* <p>The view's {@code iterator} is a "weakly consistent" iterator
1162	>	* that will never throw {@link ConcurrentModificationException},
1163	>	* and guarantees to traverse elements as they existed upon
1164	>	* construction of the iterator, and may (but is not guaranteed to)
1165	>	* reflect any modifications subsequent to construction.
1166	>	*
1167	>	* @return the set view
1168	>	*/
1169	>	public KeySetView<K,V> keySet() {
1170	>	KeySetView<K,V> ks;
1171	>	return (ks = keySet) != null ? ks : (keySet = new KeySetView<K,V>(this, null));
1172	>	}
1173	>
1174	>	/**
1175	>	* Returns a {@link Collection} view of the values contained in this map.
1176	>	* The collection is backed by the map, so changes to the map are
1177	>	* reflected in the collection, and vice-versa.  The collection
1178	>	* supports element removal, which removes the corresponding
1179	>	* mapping from this map, via the {@code Iterator.remove},
1180	>	* {@code Collection.remove}, {@code removeAll},
1181	>	* {@code retainAll}, and {@code clear} operations.  It does not
1182	>	* support the {@code add} or {@code addAll} operations.
1183	>	*
1184	>	* <p>The view's {@code iterator} is a "weakly consistent" iterator
1185	>	* that will never throw {@link ConcurrentModificationException},
1186	>	* and guarantees to traverse elements as they existed upon
1187	>	* construction of the iterator, and may (but is not guaranteed to)
1188	>	* reflect any modifications subsequent to construction.
1189	>	*
1190	>	* @return the collection view
1191	>	*/
1192	>	public Collection<V> values() {
1193	>	ValuesView<K,V> vs;
1194	>	return (vs = values) != null ? vs : (values = new ValuesView<K,V>(this));
1195		}
1196
1197	<	// Accessing segments
1197	>	/**
1198	>	* Returns a {@link Set} view of the mappings contained in this map.
1199	>	* The set is backed by the map, so changes to the map are
1200	>	* reflected in the set, and vice-versa.  The set supports element
1201	>	* removal, which removes the corresponding mapping from the map,
1202	>	* via the {@code Iterator.remove}, {@code Set.remove},
1203	>	* {@code removeAll}, {@code retainAll}, and {@code clear}
1204	>	* operations.
1205	>	*
1206	>	* <p>The view's {@code iterator} is a "weakly consistent" iterator
1207	>	* that will never throw {@link ConcurrentModificationException},
1208	>	* and guarantees to traverse elements as they existed upon
1209	>	* construction of the iterator, and may (but is not guaranteed to)
1210	>	* reflect any modifications subsequent to construction.
1211	>	*
1212	>	* @return the set view
1213	>	*/
1214	>	public Set<Map.Entry<K,V>> entrySet() {
1215	>	EntrySetView<K,V> es;
1216	>	return (es = entrySet) != null ? es : (entrySet = new EntrySetView<K,V>(this));
1217	>	}
1218
1219		/**
1220	<	* Gets the jth element of given segment array (if nonnull) with
1221	<	* volatile element access semantics via Unsafe. (The null check
1222	<	* can trigger harmlessly only during deserialization.) Note:
1223	<	* because each element of segments array is set only once (using
1224	<	* fully ordered writes), some performance-sensitive methods rely
625	<	* on this method only as a recheck upon null reads.
1220	>	* Returns the hash code value for this {@link Map}, i.e.,
1221	>	* the sum of, for each key-value pair in the map,
1222	>	* {@code key.hashCode() ^ value.hashCode()}.
1223	>	*
1224	>	* @return the hash code value for this map
1225		*/
1226	<	@SuppressWarnings("unchecked")
1227	<	static final <K,V> Segment<K,V> segmentAt(Segment<K,V>[] ss, int j) {
1228	<	long u = (j << SSHIFT) + SBASE;
1229	<	return ss == null ? null :
1230	<	(Segment<K,V>) UNSAFE.getObjectVolatile(ss, u);
1226	>	public int hashCode() {
1227	>	int h = 0;
1228	>	Node<K,V>[] t;
1229	>	if ((t = table) != null) {
1230	>	Traverser<K,V> it = new Traverser<K,V>(t, t.length, 0, t.length);
1231	>	for (Node<K,V> p; (p = it.advance()) != null; )
1232	>	h += p.key.hashCode() ^ p.val.hashCode();
1233	>	}
1234	>	return h;
1235		}
1236
1237		/**
1238	<	* Returns the segment for the given index, creating it and
1239	<	* recording in segment table (via CAS) if not already present.
1238	>	* Returns a string representation of this map.  The string
1239	>	* representation consists of a list of key-value mappings (in no
1240	>	* particular order) enclosed in braces ("{@code {}}").  Adjacent
1241	>	* mappings are separated by the characters {@code ", "} (comma
1242	>	* and space).  Each key-value mapping is rendered as the key
1243	>	* followed by an equals sign ("{@code =}") followed by the
1244	>	* associated value.
1245		*
1246	<	* @param k the index
639	<	* @return the segment
1246	>	* @return a string representation of this map
1247		*/
1248	<	@SuppressWarnings("unchecked")
1249	<	private Segment<K,V> ensureSegment(int k) {
1250	<	final Segment<K,V>[] ss = this.segments;
1251	<	long u = (k << SSHIFT) + SBASE; // raw offset
1252	<	Segment<K,V> seg;
1253	<	if ((seg = (Segment<K,V>)UNSAFE.getObjectVolatile(ss, u)) == null) {
1254	<	Segment<K,V> proto = ss[0]; // use segment 0 as prototype
1255	<	int cap = proto.table.length;
1256	<	float lf = proto.loadFactor;
1257	<	int threshold = (int)(cap * lf);
1258	<	HashEntry<K,V>[] tab = (HashEntry<K,V>[])new HashEntry[cap];
1259	<	if ((seg = (Segment<K,V>)UNSAFE.getObjectVolatile(ss, u))
1260	<	== null) { // recheck
1261	<	Segment<K,V> s = new Segment<K,V>(lf, threshold, tab);
1262	<	while ((seg = (Segment<K,V>)UNSAFE.getObjectVolatile(ss, u))
1263	<	== null) {
1264	<	if (UNSAFE.compareAndSwapObject(ss, u, null, seg = s))
658	<	break;
659	<	}
1248	>	public String toString() {
1249	>	Node<K,V>[] t;
1250	>	int f = (t = table) == null ? 0 : t.length;
1251	>	Traverser<K,V> it = new Traverser<K,V>(t, f, 0, f);
1252	>	StringBuilder sb = new StringBuilder();
1253	>	sb.append('{');
1254	>	Node<K,V> p;
1255	>	if ((p = it.advance()) != null) {
1256	>	for (;;) {
1257	>	K k = p.key;
1258	>	V v = p.val;
1259	>	sb.append(k == this ? "(this Map)" : k);
1260	>	sb.append('=');
1261	>	sb.append(v == this ? "(this Map)" : v);
1262	>	if ((p = it.advance()) == null)
1263	>	break;
1264	>	sb.append(',').append(' ');
1265		}
1266		}
1267	<	return seg;
1267	>	return sb.append('}').toString();
1268		}
1269
665	–	// Hash-based segment and entry accesses
666	–
1270		/**
1271	<	* Gets the segment for the given hash code.
1271	>	* Compares the specified object with this map for equality.
1272	>	* Returns {@code true} if the given object is a map with the same
1273	>	* mappings as this map.  This operation may return misleading
1274	>	* results if either map is concurrently modified during execution
1275	>	* of this method.
1276	>	*
1277	>	* @param o object to be compared for equality with this map
1278	>	* @return {@code true} if the specified object is equal to this map
1279		*/
1280	<	@SuppressWarnings("unchecked")
1281	<	private Segment<K,V> segmentForHash(int h) {
1282	<	long u = (((h >>> segmentShift) & segmentMask) << SSHIFT) + SBASE;
1283	<	return (Segment<K,V>) UNSAFE.getObjectVolatile(segments, u);
1280	>	public boolean equals(Object o) {
1281	>	if (o != this) {
1282	>	if (!(o instanceof Map))
1283	>	return false;
1284	>	Map<?,?> m = (Map<?,?>) o;
1285	>	Node<K,V>[] t;
1286	>	int f = (t = table) == null ? 0 : t.length;
1287	>	Traverser<K,V> it = new Traverser<K,V>(t, f, 0, f);
1288	>	for (Node<K,V> p; (p = it.advance()) != null; ) {
1289	>	V val = p.val;
1290	>	Object v = m.get(p.key);
1291	>	if (v == null || (v != val && !v.equals(val)))
1292	>	return false;
1293	>	}
1294	>	for (Map.Entry<?,?> e : m.entrySet()) {
1295	>	Object mk, mv, v;
1296	>	if ((mk = e.getKey()) == null ||
1297	>	(mv = e.getValue()) == null ||
1298	>	(v = get(mk)) == null ||
1299	>	(mv != v && !mv.equals(v)))
1300	>	return false;
1301	>	}
1302	>	}
1303	>	return true;
1304		}
1305
1306		/**
1307	<	* Gets the table entry for the given segment and hash code.
1307	>	* Stripped-down version of helper class used in previous version,
1308	>	* declared for the sake of serialization compatibility
1309		*/
1310	<	@SuppressWarnings("unchecked")
1311	<	static final <K,V> HashEntry<K,V> entryForHash(Segment<K,V> seg, int h) {
1312	<	HashEntry<K,V>[] tab;
1313	<	return (seg == null || (tab = seg.table) == null) ? null :
683	<	(HashEntry<K,V>) UNSAFE.getObjectVolatile
684	<	(tab, ((long)(((tab.length - 1) & h)) << TSHIFT) + TBASE);
1310	>	static class Segment<K,V> extends ReentrantLock implements Serializable {
1311	>	private static final long serialVersionUID = 2249069246763182397L;
1312	>	final float loadFactor;
1313	>	Segment(float lf) { this.loadFactor = lf; }
1314		}
1315
687	–	/* ---------------- Public operations -------------- */
688	–
1316		/**
1317	<	* Creates a new, empty map with the specified initial
1318	<	* capacity, load factor and concurrency level.
1319	<	*
1320	<	* @param initialCapacity the initial capacity. The implementation
1321	<	* performs internal sizing to accommodate this many elements.
1322	<	* @param loadFactor  the load factor threshold, used to control resizing.
1323	<	* Resizing may be performed when the average number of elements per
697	<	* bin exceeds this threshold.
698	<	* @param concurrencyLevel the estimated number of concurrently
699	<	* updating threads. The implementation performs internal sizing
700	<	* to try to accommodate this many threads.
701	<	* @throws IllegalArgumentException if the initial capacity is
702	<	* negative or the load factor or concurrencyLevel are
703	<	* nonpositive.
1317	>	* Saves the state of the {@code ConcurrentHashMap} instance to a
1318	>	* stream (i.e., serializes it).
1319	>	* @param s the stream
1320	>	* @serialData
1321	>	* the key (Object) and value (Object)
1322	>	* for each key-value mapping, followed by a null pair.
1323	>	* The key-value mappings are emitted in no particular order.
1324		*/
1325	<	@SuppressWarnings("unchecked")
1326	<	public ConcurrentHashMap(int initialCapacity,
1327	<	float loadFactor, int concurrencyLevel) {
1328	<	if (!(loadFactor > 0) || initialCapacity < 0 || concurrencyLevel <= 0)
709	<	throw new IllegalArgumentException();
710	<	if (concurrencyLevel > MAX_SEGMENTS)
711	<	concurrencyLevel = MAX_SEGMENTS;
712	<	// Find power-of-two sizes best matching arguments
1325	>	private void writeObject(java.io.ObjectOutputStream s)
1326	>	throws java.io.IOException {
1327	>	// For serialization compatibility
1328	>	// Emulate segment calculation from previous version of this class
1329		int sshift = 0;
1330		int ssize = 1;
1331	<	while (ssize < concurrencyLevel) {
1331	>	while (ssize < DEFAULT_CONCURRENCY_LEVEL) {
1332		++sshift;
1333		ssize <<= 1;
1334		}
1335	<	this.segmentShift = 32 - sshift;
1336	<	this.segmentMask = ssize - 1;
1337	<	if (initialCapacity > MAXIMUM_CAPACITY)
1338	<	initialCapacity = MAXIMUM_CAPACITY;
1339	<	int c = initialCapacity / ssize;
1340	<	if (c * ssize < initialCapacity)
1341	<	++c;
1342	<	int cap = MIN_SEGMENT_TABLE_CAPACITY;
1343	<	while (cap < c)
1344	<	cap <<= 1;
1345	<	// create segments and segments[0]
1346	<	Segment<K,V> s0 =
1347	<	new Segment<K,V>(loadFactor, (int)(cap * loadFactor),
1348	<	(HashEntry<K,V>[])new HashEntry[cap]);
1349	<	Segment<K,V>[] ss = (Segment<K,V>[])new Segment[ssize];
1350	<	UNSAFE.putOrderedObject(ss, SBASE, s0); // ordered write of segments[0]
1351	<	this.segments = ss;
1335	>	int segmentShift = 32 - sshift;
1336	>	int segmentMask = ssize - 1;
1337	>	@SuppressWarnings("unchecked") Segment<K,V>[] segments = (Segment<K,V>[])
1338	>	new Segment<?,?>[DEFAULT_CONCURRENCY_LEVEL];
1339	>	for (int i = 0; i < segments.length; ++i)
1340	>	segments[i] = new Segment<K,V>(LOAD_FACTOR);
1341	>	s.putFields().put("segments", segments);
1342	>	s.putFields().put("segmentShift", segmentShift);
1343	>	s.putFields().put("segmentMask", segmentMask);
1344	>	s.writeFields();
1345	>
1346	>	Node<K,V>[] t;
1347	>	if ((t = table) != null) {
1348	>	Traverser<K,V> it = new Traverser<K,V>(t, t.length, 0, t.length);
1349	>	for (Node<K,V> p; (p = it.advance()) != null; ) {
1350	>	s.writeObject(p.key);
1351	>	s.writeObject(p.val);
1352	>	}
1353	>	}
1354	>	s.writeObject(null);
1355	>	s.writeObject(null);
1356	>	segments = null; // throw away
1357		}
1358
1359		/**
1360	<	* Creates a new, empty map with the specified initial capacity
1361	<	* and load factor and with the default concurrencyLevel (16).
1362	<	*
1363	<	* @param initialCapacity The implementation performs internal
1364	<	* sizing to accommodate this many elements.
1365	<	* @param loadFactor  the load factor threshold, used to control resizing.
1366	<	* Resizing may be performed when the average number of elements per
1367	<	* bin exceeds this threshold.
1368	<	* @throws IllegalArgumentException if the initial capacity of
1369	<	* elements is negative or the load factor is nonpositive
1360	>	* Reconstitutes the instance from a stream (that is, deserializes it).
1361	>	* @param s the stream
1362	>	*/
1363	>	private void readObject(java.io.ObjectInputStream s)
1364	>	throws java.io.IOException, ClassNotFoundException {
1365	>	/*
1366	>	* To improve performance in typical cases, we create nodes
1367	>	* while reading, then place in table once size is known.
1368	>	* However, we must also validate uniqueness and deal with
1369	>	* overpopulated bins while doing so, which requires
1370	>	* specialized versions of putVal mechanics.
1371	>	*/
1372	>	sizeCtl = -1; // force exclusion for table construction
1373	>	s.defaultReadObject();
1374	>	long size = 0L;
1375	>	Node<K,V> p = null;
1376	>	for (;;) {
1377	>	@SuppressWarnings("unchecked") K k = (K) s.readObject();
1378	>	@SuppressWarnings("unchecked") V v = (V) s.readObject();
1379	>	if (k != null && v != null) {
1380	>	p = new Node<K,V>(spread(k.hashCode()), k, v, p);
1381	>	++size;
1382	>	}
1383	>	else
1384	>	break;
1385	>	}
1386	>	if (size == 0L)
1387	>	sizeCtl = 0;
1388	>	else {
1389	>	int n;
1390	>	if (size >= (long)(MAXIMUM_CAPACITY >>> 1))
1391	>	n = MAXIMUM_CAPACITY;
1392	>	else {
1393	>	int sz = (int)size;
1394	>	n = tableSizeFor(sz + (sz >>> 1) + 1);
1395	>	}
1396	>	@SuppressWarnings({"rawtypes","unchecked"})
1397	>	Node<K,V>[] tab = (Node<K,V>[])new Node[n];
1398	>	int mask = n - 1;
1399	>	long added = 0L;
1400	>	while (p != null) {
1401	>	boolean insertAtFront;
1402	>	Node<K,V> next = p.next, first;
1403	>	int h = p.hash, j = h & mask;
1404	>	if ((first = tabAt(tab, j)) == null)
1405	>	insertAtFront = true;
1406	>	else {
1407	>	K k = p.key;
1408	>	if (first.hash < 0) {
1409	>	TreeBin<K,V> t = (TreeBin<K,V>)first;
1410	>	if (t.putTreeVal(h, k, p.val) == null)
1411	>	++added;
1412	>	insertAtFront = false;
1413	>	}
1414	>	else {
1415	>	int binCount = 0;
1416	>	insertAtFront = true;
1417	>	Node<K,V> q; K qk;
1418	>	for (q = first; q != null; q = q.next) {
1419	>	if (q.hash == h &&
1420	>	((qk = q.key) == k ||
1421	>	(qk != null && k.equals(qk)))) {
1422	>	insertAtFront = false;
1423	>	break;
1424	>	}
1425	>	++binCount;
1426	>	}
1427	>	if (insertAtFront && binCount >= TREEIFY_THRESHOLD) {
1428	>	insertAtFront = false;
1429	>	++added;
1430	>	p.next = first;
1431	>	TreeNode<K,V> hd = null, tl = null;
1432	>	for (q = p; q != null; q = q.next) {
1433	>	TreeNode<K,V> t = new TreeNode<K,V>
1434	>	(q.hash, q.key, q.val, null, null);
1435	>	if ((t.prev = tl) == null)
1436	>	hd = t;
1437	>	else
1438	>	tl.next = t;
1439	>	tl = t;
1440	>	}
1441	>	setTabAt(tab, j, new TreeBin<K,V>(hd));
1442	>	}
1443	>	}
1444	>	}
1445	>	if (insertAtFront) {
1446	>	++added;
1447	>	p.next = first;
1448	>	setTabAt(tab, j, p);
1449	>	}
1450	>	p = next;
1451	>	}
1452	>	table = tab;
1453	>	sizeCtl = n - (n >>> 2);
1454	>	baseCount = added;
1455	>	}
1456	>	}
1457	>
1458	>	// ConcurrentMap methods
1459	>
1460	>	/**
1461	>	* {@inheritDoc}
1462		*
1463	<	* @since 1.6
1463	>	* @return the previous value associated with the specified key,
1464	>	*         or {@code null} if there was no mapping for the key
1465	>	* @throws NullPointerException if the specified key or value is null
1466		*/
1467	<	public ConcurrentHashMap(int initialCapacity, float loadFactor) {
1468	<	this(initialCapacity, loadFactor, DEFAULT_CONCURRENCY_LEVEL);
1467	>	public V putIfAbsent(K key, V value) {
1468	>	return putVal(key, value, true);
1469		}
1470
1471		/**
1472	<	* Creates a new, empty map with the specified initial capacity,
758	<	* and with default load factor (0.75) and concurrencyLevel (16).
1472	>	* {@inheritDoc}
1473		*
1474	<	* @param initialCapacity the initial capacity. The implementation
761	<	* performs internal sizing to accommodate this many elements.
762	<	* @throws IllegalArgumentException if the initial capacity of
763	<	* elements is negative.
1474	>	* @throws NullPointerException if the specified key is null
1475		*/
1476	<	public ConcurrentHashMap(int initialCapacity) {
1477	<	this(initialCapacity, DEFAULT_LOAD_FACTOR, DEFAULT_CONCURRENCY_LEVEL);
1476	>	public boolean remove(Object key, Object value) {
1477	>	if (key == null)
1478	>	throw new NullPointerException();
1479	>	return value != null && replaceNode(key, null, value) != null;
1480		}
1481
1482		/**
1483	<	* Creates a new, empty map with a default initial capacity (16),
1484	<	* load factor (0.75) and concurrencyLevel (16).
1483	>	* {@inheritDoc}
1484	>	*
1485	>	* @throws NullPointerException if any of the arguments are null
1486		*/
1487	<	public ConcurrentHashMap() {
1488	<	this(DEFAULT_INITIAL_CAPACITY, DEFAULT_LOAD_FACTOR, DEFAULT_CONCURRENCY_LEVEL);
1487	>	public boolean replace(K key, V oldValue, V newValue) {
1488	>	if (key == null || oldValue == null || newValue == null)
1489	>	throw new NullPointerException();
1490	>	return replaceNode(key, newValue, oldValue) != null;
1491		}
1492
1493		/**
1494	<	* Creates a new map with the same mappings as the given map.
779	<	* The map is created with a capacity of 1.5 times the number
780	<	* of mappings in the given map or 16 (whichever is greater),
781	<	* and a default load factor (0.75) and concurrencyLevel (16).
1494	>	* {@inheritDoc}
1495		*
1496	<	* @param m the map
1496	>	* @return the previous value associated with the specified key,
1497	>	*         or {@code null} if there was no mapping for the key
1498	>	* @throws NullPointerException if the specified key or value is null
1499		*/
1500	<	public ConcurrentHashMap(Map<? extends K, ? extends V> m) {
1501	<	this(Math.max((int) (m.size() / DEFAULT_LOAD_FACTOR) + 1,
1502	<	DEFAULT_INITIAL_CAPACITY),
1503	<	DEFAULT_LOAD_FACTOR, DEFAULT_CONCURRENCY_LEVEL);
789	<	putAll(m);
1500	>	public V replace(K key, V value) {
1501	>	if (key == null || value == null)
1502	>	throw new NullPointerException();
1503	>	return replaceNode(key, value, null);
1504		}
1505
1506	+	// Overrides of JDK8+ Map extension method defaults
1507	+
1508		/**
1509	<	* Returns <tt>true</tt> if this map contains no key-value mappings.
1509	>	* Returns the value to which the specified key is mapped, or the
1510	>	* given default value if this map contains no mapping for the
1511	>	* key.
1512		*
1513	<	* @return <tt>true</tt> if this map contains no key-value mappings
1513	>	* @param key the key whose associated value is to be returned
1514	>	* @param defaultValue the value to return if this map contains
1515	>	* no mapping for the given key
1516	>	* @return the mapping for the key, if present; else the default value
1517	>	* @throws NullPointerException if the specified key is null
1518		*/
1519	<	public boolean isEmpty() {
1520	<	/*
1521	<	* Sum per-segment modCounts to avoid mis-reporting when
1522	<	* elements are concurrently added and removed in one segment
1523	<	* while checking another, in which case the table was never
1524	<	* actually empty at any point. (The sum ensures accuracy up
1525	<	* through at least 1<<31 per-segment modifications before
1526	<	* recheck.)  Methods size() and containsValue() use similar
1527	<	* constructions for stability checks.
1528	<	*/
1529	<	long sum = 0L;
1530	<	final Segment<K,V>[] segments = this.segments;
809	<	for (int j = 0; j < segments.length; ++j) {
810	<	Segment<K,V> seg = segmentAt(segments, j);
811	<	if (seg != null) {
812	<	if (seg.count != 0)
813	<	return false;
814	<	sum += seg.modCount;
1519	>	public V getOrDefault(Object key, V defaultValue) {
1520	>	V v;
1521	>	return (v = get(key)) == null ? defaultValue : v;
1522	>	}
1523	>
1524	>	public void forEach(BiConsumer<? super K, ? super V> action) {
1525	>	if (action == null) throw new NullPointerException();
1526	>	Node<K,V>[] t;
1527	>	if ((t = table) != null) {
1528	>	Traverser<K,V> it = new Traverser<K,V>(t, t.length, 0, t.length);
1529	>	for (Node<K,V> p; (p = it.advance()) != null; ) {
1530	>	action.accept(p.key, p.val);
1531		}
1532		}
1533	<	if (sum != 0L) { // recheck unless no modifications
1534	<	for (int j = 0; j < segments.length; ++j) {
1535	<	Segment<K,V> seg = segmentAt(segments, j);
1536	<	if (seg != null) {
1537	<	if (seg.count != 0)
1538	<	return false;
1539	<	sum -= seg.modCount;
1533	>	}
1534	>
1535	>	public void replaceAll(BiFunction<? super K, ? super V, ? extends V> function) {
1536	>	if (function == null) throw new NullPointerException();
1537	>	Node<K,V>[] t;
1538	>	if ((t = table) != null) {
1539	>	Traverser<K,V> it = new Traverser<K,V>(t, t.length, 0, t.length);
1540	>	for (Node<K,V> p; (p = it.advance()) != null; ) {
1541	>	V oldValue = p.val;
1542	>	for (K key = p.key;;) {
1543	>	V newValue = function.apply(key, oldValue);
1544	>	if (newValue == null)
1545	>	throw new NullPointerException();
1546	>	if (replaceNode(key, newValue, oldValue) != null ||
1547	>	(oldValue = get(key)) == null)
1548	>	break;
1549		}
1550		}
826	–	if (sum != 0L)
827	–	return false;
1551		}
829	–	return true;
1552		}
1553
1554		/**
1555	<	* Returns the number of key-value mappings in this map.  If the
1556	<	* map contains more than <tt>Integer.MAX_VALUE</tt> elements, returns
1557	<	* <tt>Integer.MAX_VALUE</tt>.
1555	>	* If the specified key is not already associated with a value,
1556	>	* attempts to compute its value using the given mapping function
1557	>	* and enters it into this map unless {@code null}.  The entire
1558	>	* method invocation is performed atomically, so the function is
1559	>	* applied at most once per key.  Some attempted update operations
1560	>	* on this map by other threads may be blocked while computation
1561	>	* is in progress, so the computation should be short and simple,
1562	>	* and must not attempt to update any other mappings of this map.
1563		*
1564	<	* @return the number of key-value mappings in this map
1564	>	* @param key key with which the specified value is to be associated
1565	>	* @param mappingFunction the function to compute a value
1566	>	* @return the current (existing or computed) value associated with
1567	>	*         the specified key, or null if the computed value is null
1568	>	* @throws NullPointerException if the specified key or mappingFunction
1569	>	*         is null
1570	>	* @throws IllegalStateException if the computation detectably
1571	>	*         attempts a recursive update to this map that would
1572	>	*         otherwise never complete
1573	>	* @throws RuntimeException or Error if the mappingFunction does so,
1574	>	*         in which case the mapping is left unestablished
1575		*/
1576	<	public int size() {
1577	<	// Try a few times to get accurate count. On failure due to
1578	<	// continuous async changes in table, resort to locking.
1579	<	final Segment<K,V>[] segments = this.segments;
1580	<	int size;
1581	<	boolean overflow; // true if size overflows 32 bits
1582	<	long sum;         // sum of modCounts
1583	<	long last = 0L;   // previous sum
1584	<	int retries = -1; // first iteration isn't retry
1585	<	try {
1586	<	for (;;) {
1587	<	if (retries++ == RETRIES_BEFORE_LOCK) {
1588	<	for (int j = 0; j < segments.length; ++j)
1589	<	ensureSegment(j).lock(); // force creation
1590	<	}
1591	<	sum = 0L;
1592	<	size = 0;
1593	<	overflow = false;
1594	<	for (int j = 0; j < segments.length; ++j) {
1595	<	Segment<K,V> seg = segmentAt(segments, j);
1596	<	if (seg != null) {
1597	<	sum += seg.modCount;
861	<	int c = seg.count;
862	<	if (c < 0 || (size += c) < 0)
863	<	overflow = true;
1576	>	public V computeIfAbsent(K key, Function<? super K, ? extends V> mappingFunction) {
1577	>	if (key == null || mappingFunction == null)
1578	>	throw new NullPointerException();
1579	>	int h = spread(key.hashCode());
1580	>	V val = null;
1581	>	int binCount = 0;
1582	>	for (Node<K,V>[] tab = table;;) {
1583	>	Node<K,V> f; int n, i, fh;
1584	>	if (tab == null || (n = tab.length) == 0)
1585	>	tab = initTable();
1586	>	else if ((f = tabAt(tab, i = (n - 1) & h)) == null) {
1587	>	Node<K,V> r = new ReservationNode<K,V>();
1588	>	synchronized (r) {
1589	>	if (casTabAt(tab, i, null, r)) {
1590	>	binCount = 1;
1591	>	Node<K,V> node = null;
1592	>	try {
1593	>	if ((val = mappingFunction.apply(key)) != null)
1594	>	node = new Node<K,V>(h, key, val, null);
1595	>	} finally {
1596	>	setTabAt(tab, i, node);
1597	>	}
1598		}
1599		}
1600	<	if (sum == last)
1600	>	if (binCount != 0)
1601		break;
868	–	last = sum;
1602		}
1603	<	} finally {
1604	<	if (retries > RETRIES_BEFORE_LOCK) {
1605	<	for (int j = 0; j < segments.length; ++j)
1606	<	segmentAt(segments, j).unlock();
1603	>	else if ((fh = f.hash) == MOVED)
1604	>	tab = helpTransfer(tab, f);
1605	>	else {
1606	>	boolean added = false;
1607	>	synchronized (f) {
1608	>	if (tabAt(tab, i) == f) {
1609	>	if (fh >= 0) {
1610	>	binCount = 1;
1611	>	for (Node<K,V> e = f;; ++binCount) {
1612	>	K ek; V ev;
1613	>	if (e.hash == h &&
1614	>	((ek = e.key) == key ||
1615	>	(ek != null && key.equals(ek)))) {
1616	>	val = e.val;
1617	>	break;
1618	>	}
1619	>	Node<K,V> pred = e;
1620	>	if ((e = e.next) == null) {
1621	>	if ((val = mappingFunction.apply(key)) != null) {
1622	>	added = true;
1623	>	pred.next = new Node<K,V>(h, key, val, null);
1624	>	}
1625	>	break;
1626	>	}
1627	>	}
1628	>	}
1629	>	else if (f instanceof TreeBin) {
1630	>	binCount = 2;
1631	>	TreeBin<K,V> t = (TreeBin<K,V>)f;
1632	>	TreeNode<K,V> r, p;
1633	>	if ((r = t.root) != null &&
1634	>	(p = r.findTreeNode(h, key, null)) != null)
1635	>	val = p.val;
1636	>	else if ((val = mappingFunction.apply(key)) != null) {
1637	>	added = true;
1638	>	t.putTreeVal(h, key, val);
1639	>	}
1640	>	}
1641	>	}
1642	>	}
1643	>	if (binCount != 0) {
1644	>	if (binCount >= TREEIFY_THRESHOLD)
1645	>	treeifyBin(tab, i);
1646	>	if (!added)
1647	>	return val;
1648	>	break;
1649	>	}
1650		}
1651		}
1652	<	return overflow ? Integer.MAX_VALUE : size;
1652	>	if (val != null)
1653	>	addCount(1L, binCount);
1654	>	return val;
1655		}
1656
1657		/**
1658	<	* Returns the value to which the specified key is mapped,
1659	<	* or {@code null} if this map contains no mapping for the key.
1660	<	*
1661	<	* <p>More formally, if this map contains a mapping from a key
1662	<	* {@code k} to a value {@code v} such that {@code key.equals(k)},
1663	<	* then this method returns {@code v}; otherwise it returns
1664	<	* {@code null}.  (There can be at most one such mapping.)
1658	>	* If the value for the specified key is present, attempts to
1659	>	* compute a new mapping given the key and its current mapped
1660	>	* value.  The entire method invocation is performed atomically.
1661	>	* Some attempted update operations on this map by other threads
1662	>	* may be blocked while computation is in progress, so the
1663	>	* computation should be short and simple, and must not attempt to
1664	>	* update any other mappings of this map.
1665		*
1666	<	* @throws NullPointerException if the specified key is null
1666	>	* @param key key with which a value may be associated
1667	>	* @param remappingFunction the function to compute a value
1668	>	* @return the new value associated with the specified key, or null if none
1669	>	* @throws NullPointerException if the specified key or remappingFunction
1670	>	*         is null
1671	>	* @throws IllegalStateException if the computation detectably
1672	>	*         attempts a recursive update to this map that would
1673	>	*         otherwise never complete
1674	>	* @throws RuntimeException or Error if the remappingFunction does so,
1675	>	*         in which case the mapping is unchanged
1676		*/
1677	<	public V get(Object key) {
1678	<	Segment<K,V> s; // manually integrate access methods to reduce overhead
1679	<	HashEntry<K,V>[] tab;
1680	<	int h = hash(key.hashCode());
1681	<	long u = (((h >>> segmentShift) & segmentMask) << SSHIFT) + SBASE;
1682	<	if ((s = (Segment<K,V>)UNSAFE.getObjectVolatile(segments, u)) != null &&
1683	<	(tab = s.table) != null) {
1684	<	for (HashEntry<K,V> e = (HashEntry<K,V>) UNSAFE.getObjectVolatile
1685	<	(tab, ((long)(((tab.length - 1) & h)) << TSHIFT) + TBASE);
1686	<	e != null; e = e.next) {
1687	<	K k;
1688	<	if ((k = e.key) == key || (e.hash == h && key.equals(k)))
1689	<	return e.value;
1677	>	public V computeIfPresent(K key, BiFunction<? super K, ? super V, ? extends V> remappingFunction) {
1678	>	if (key == null || remappingFunction == null)
1679	>	throw new NullPointerException();
1680	>	int h = spread(key.hashCode());
1681	>	V val = null;
1682	>	int delta = 0;
1683	>	int binCount = 0;
1684	>	for (Node<K,V>[] tab = table;;) {
1685	>	Node<K,V> f; int n, i, fh;
1686	>	if (tab == null || (n = tab.length) == 0)
1687	>	tab = initTable();
1688	>	else if ((f = tabAt(tab, i = (n - 1) & h)) == null)
1689	>	break;
1690	>	else if ((fh = f.hash) == MOVED)
1691	>	tab = helpTransfer(tab, f);
1692	>	else {
1693	>	synchronized (f) {
1694	>	if (tabAt(tab, i) == f) {
1695	>	if (fh >= 0) {
1696	>	binCount = 1;
1697	>	for (Node<K,V> e = f, pred = null;; ++binCount) {
1698	>	K ek;
1699	>	if (e.hash == h &&
1700	>	((ek = e.key) == key ||
1701	>	(ek != null && key.equals(ek)))) {
1702	>	val = remappingFunction.apply(key, e.val);
1703	>	if (val != null)
1704	>	e.val = val;
1705	>	else {
1706	>	delta = -1;
1707	>	Node<K,V> en = e.next;
1708	>	if (pred != null)
1709	>	pred.next = en;
1710	>	else
1711	>	setTabAt(tab, i, en);
1712	>	}
1713	>	break;
1714	>	}
1715	>	pred = e;
1716	>	if ((e = e.next) == null)
1717	>	break;
1718	>	}
1719	>	}
1720	>	else if (f instanceof TreeBin) {
1721	>	binCount = 2;
1722	>	TreeBin<K,V> t = (TreeBin<K,V>)f;
1723	>	TreeNode<K,V> r, p;
1724	>	if ((r = t.root) != null &&
1725	>	(p = r.findTreeNode(h, key, null)) != null) {
1726	>	val = remappingFunction.apply(key, p.val);
1727	>	if (val != null)
1728	>	p.val = val;
1729	>	else {
1730	>	delta = -1;
1731	>	if (t.removeTreeNode(p))
1732	>	setTabAt(tab, i, untreeify(t.first));
1733	>	}
1734	>	}
1735	>	}
1736	>	}
1737	>	}
1738	>	if (binCount != 0)
1739	>	break;
1740		}
1741		}
1742	<	return null;
1742	>	if (delta != 0)
1743	>	addCount((long)delta, binCount);
1744	>	return val;
1745		}
1746
1747		/**
1748	<	* Tests if the specified object is a key in this table.
1748	>	* Attempts to compute a mapping for the specified key and its
1749	>	* current mapped value (or {@code null} if there is no current
1750	>	* mapping). The entire method invocation is performed atomically.
1751	>	* Some attempted update operations on this map by other threads
1752	>	* may be blocked while computation is in progress, so the
1753	>	* computation should be short and simple, and must not attempt to
1754	>	* update any other mappings of this Map.
1755		*
1756	<	* @param  key   possible key
1757	<	* @return <tt>true</tt> if and only if the specified object
1758	<	*         is a key in this table, as determined by the
1759	<	*         <tt>equals</tt> method; <tt>false</tt> otherwise.
1760	<	* @throws NullPointerException if the specified key is null
1761	<	*/
1762	<	@SuppressWarnings("unchecked")
1763	<	public boolean containsKey(Object key) {
1764	<	Segment<K,V> s; // same as get() except no need for volatile value read
1765	<	HashEntry<K,V>[] tab;
1766	<	int h = hash(key.hashCode());
1767	<	long u = (((h >>> segmentShift) & segmentMask) << SSHIFT) + SBASE;
1768	<	if ((s = (Segment<K,V>)UNSAFE.getObjectVolatile(segments, u)) != null &&
1769	<	(tab = s.table) != null) {
1770	<	for (HashEntry<K,V> e = (HashEntry<K,V>) UNSAFE.getObjectVolatile
1771	<	(tab, ((long)(((tab.length - 1) & h)) << TSHIFT) + TBASE);
1772	<	e != null; e = e.next) {
1773	<	K k;
1774	<	if ((k = e.key) == key || (e.hash == h && key.equals(k)))
1775	<	return true;
1756	>	* @param key key with which the specified value is to be associated
1757	>	* @param remappingFunction the function to compute a value
1758	>	* @return the new value associated with the specified key, or null if none
1759	>	* @throws NullPointerException if the specified key or remappingFunction
1760	>	*         is null
1761	>	* @throws IllegalStateException if the computation detectably
1762	>	*         attempts a recursive update to this map that would
1763	>	*         otherwise never complete
1764	>	* @throws RuntimeException or Error if the remappingFunction does so,
1765	>	*         in which case the mapping is unchanged
1766	>	*/
1767	>	public V compute(K key,
1768	>	BiFunction<? super K, ? super V, ? extends V> remappingFunction) {
1769	>	if (key == null || remappingFunction == null)
1770	>	throw new NullPointerException();
1771	>	int h = spread(key.hashCode());
1772	>	V val = null;
1773	>	int delta = 0;
1774	>	int binCount = 0;
1775	>	for (Node<K,V>[] tab = table;;) {
1776	>	Node<K,V> f; int n, i, fh;
1777	>	if (tab == null || (n = tab.length) == 0)
1778	>	tab = initTable();
1779	>	else if ((f = tabAt(tab, i = (n - 1) & h)) == null) {
1780	>	Node<K,V> r = new ReservationNode<K,V>();
1781	>	synchronized (r) {
1782	>	if (casTabAt(tab, i, null, r)) {
1783	>	binCount = 1;
1784	>	Node<K,V> node = null;
1785	>	try {
1786	>	if ((val = remappingFunction.apply(key, null)) != null) {
1787	>	delta = 1;
1788	>	node = new Node<K,V>(h, key, val, null);
1789	>	}
1790	>	} finally {
1791	>	setTabAt(tab, i, node);
1792	>	}
1793	>	}
1794	>	}
1795	>	if (binCount != 0)
1796	>	break;
1797	>	}
1798	>	else if ((fh = f.hash) == MOVED)
1799	>	tab = helpTransfer(tab, f);
1800	>	else {
1801	>	synchronized (f) {
1802	>	if (tabAt(tab, i) == f) {
1803	>	if (fh >= 0) {
1804	>	binCount = 1;
1805	>	for (Node<K,V> e = f, pred = null;; ++binCount) {
1806	>	K ek;
1807	>	if (e.hash == h &&
1808	>	((ek = e.key) == key ||
1809	>	(ek != null && key.equals(ek)))) {
1810	>	val = remappingFunction.apply(key, e.val);
1811	>	if (val != null)
1812	>	e.val = val;
1813	>	else {
1814	>	delta = -1;
1815	>	Node<K,V> en = e.next;
1816	>	if (pred != null)
1817	>	pred.next = en;
1818	>	else
1819	>	setTabAt(tab, i, en);
1820	>	}
1821	>	break;
1822	>	}
1823	>	pred = e;
1824	>	if ((e = e.next) == null) {
1825	>	val = remappingFunction.apply(key, null);
1826	>	if (val != null) {
1827	>	delta = 1;
1828	>	pred.next =
1829	>	new Node<K,V>(h, key, val, null);
1830	>	}
1831	>	break;
1832	>	}
1833	>	}
1834	>	}
1835	>	else if (f instanceof TreeBin) {
1836	>	binCount = 1;
1837	>	TreeBin<K,V> t = (TreeBin<K,V>)f;
1838	>	TreeNode<K,V> r, p;
1839	>	if ((r = t.root) != null)
1840	>	p = r.findTreeNode(h, key, null);
1841	>	else
1842	>	p = null;
1843	>	V pv = (p == null) ? null : p.val;
1844	>	val = remappingFunction.apply(key, pv);
1845	>	if (val != null) {
1846	>	if (p != null)
1847	>	p.val = val;
1848	>	else {
1849	>	delta = 1;
1850	>	t.putTreeVal(h, key, val);
1851	>	}
1852	>	}
1853	>	else if (p != null) {
1854	>	delta = -1;
1855	>	if (t.removeTreeNode(p))
1856	>	setTabAt(tab, i, untreeify(t.first));
1857	>	}
1858	>	}
1859	>	}
1860	>	}
1861	>	if (binCount != 0) {
1862	>	if (binCount >= TREEIFY_THRESHOLD)
1863	>	treeifyBin(tab, i);
1864	>	break;
1865	>	}
1866		}
1867		}
1868	<	return false;
1868	>	if (delta != 0)
1869	>	addCount((long)delta, binCount);
1870	>	return val;
1871		}
1872
1873		/**
1874	<	* Returns <tt>true</tt> if this map maps one or more keys to the
1875	<	* specified value. Note: This method requires a full internal
1876	<	* traversal of the hash table, and so is much slower than
1877	<	* method <tt>containsKey</tt>.
1874	>	* If the specified key is not already associated with a
1875	>	* (non-null) value, associates it with the given value.
1876	>	* Otherwise, replaces the value with the results of the given
1877	>	* remapping function, or removes if {@code null}. The entire
1878	>	* method invocation is performed atomically.  Some attempted
1879	>	* update operations on this map by other threads may be blocked
1880	>	* while computation is in progress, so the computation should be
1881	>	* short and simple, and must not attempt to update any other
1882	>	* mappings of this Map.
1883		*
1884	<	* @param value value whose presence in this map is to be tested
1885	<	* @return <tt>true</tt> if this map maps one or more keys to the
1886	<	*         specified value
1887	<	* @throws NullPointerException if the specified value is null
1884	>	* @param key key with which the specified value is to be associated
1885	>	* @param value the value to use if absent
1886	>	* @param remappingFunction the function to recompute a value if present
1887	>	* @return the new value associated with the specified key, or null if none
1888	>	* @throws NullPointerException if the specified key or the
1889	>	*         remappingFunction is null
1890	>	* @throws RuntimeException or Error if the remappingFunction does so,
1891	>	*         in which case the mapping is unchanged
1892		*/
1893	<	public boolean containsValue(Object value) {
1894	<	// Same idea as size()
949	<	if (value == null)
1893	>	public V merge(K key, V value, BiFunction<? super V, ? super V, ? extends V> remappingFunction) {
1894	>	if (key == null || value == null || remappingFunction == null)
1895		throw new NullPointerException();
1896	<	final Segment<K,V>[] segments = this.segments;
1897	<	boolean found = false;
1898	<	long last = 0L;   // previous sum
1899	<	int retries = -1;
1900	<	try {
1901	<	outer: for (;;) {
1902	<	if (retries++ == RETRIES_BEFORE_LOCK) {
1903	<	for (int j = 0; j < segments.length; ++j)
1904	<	ensureSegment(j).lock(); // force creation
1905	<	}
1906	<	long sum = 0L;
1907	<	for (int j = 0; j < segments.length; ++j) {
1908	<	HashEntry<K,V>[] tab;
1909	<	Segment<K,V> seg = segmentAt(segments, j);
1910	<	if (seg != null && (tab = seg.table) != null) {
1911	<	for (int i = 0 ; i < tab.length; i++) {
1912	<	HashEntry<K,V> e;
1913	<	for (e = entryAt(tab, i); e != null; e = e.next) {
1914	<	V v = e.value;
1915	<	if (v != null && value.equals(v)) {
1916	<	found = true;
1917	<	break outer;
1896	>	int h = spread(key.hashCode());
1897	>	V val = null;
1898	>	int delta = 0;
1899	>	int binCount = 0;
1900	>	for (Node<K,V>[] tab = table;;) {
1901	>	Node<K,V> f; int n, i, fh;
1902	>	if (tab == null || (n = tab.length) == 0)
1903	>	tab = initTable();
1904	>	else if ((f = tabAt(tab, i = (n - 1) & h)) == null) {
1905	>	if (casTabAt(tab, i, null, new Node<K,V>(h, key, value, null))) {
1906	>	delta = 1;
1907	>	val = value;
1908	>	break;
1909	>	}
1910	>	}
1911	>	else if ((fh = f.hash) == MOVED)
1912	>	tab = helpTransfer(tab, f);
1913	>	else {
1914	>	synchronized (f) {
1915	>	if (tabAt(tab, i) == f) {
1916	>	if (fh >= 0) {
1917	>	binCount = 1;
1918	>	for (Node<K,V> e = f, pred = null;; ++binCount) {
1919	>	K ek;
1920	>	if (e.hash == h &&
1921	>	((ek = e.key) == key ||
1922	>	(ek != null && key.equals(ek)))) {
1923	>	val = remappingFunction.apply(e.val, value);
1924	>	if (val != null)
1925	>	e.val = val;
1926	>	else {
1927	>	delta = -1;
1928	>	Node<K,V> en = e.next;
1929	>	if (pred != null)
1930	>	pred.next = en;
1931	>	else
1932	>	setTabAt(tab, i, en);
1933	>	}
1934	>	break;
1935	>	}
1936	>	pred = e;
1937	>	if ((e = e.next) == null) {
1938	>	delta = 1;
1939	>	val = value;
1940	>	pred.next =
1941	>	new Node<K,V>(h, key, val, null);
1942	>	break;
1943		}
1944		}
1945		}
1946	<	sum += seg.modCount;
1946	>	else if (f instanceof TreeBin) {
1947	>	binCount = 2;
1948	>	TreeBin<K,V> t = (TreeBin<K,V>)f;
1949	>	TreeNode<K,V> r = t.root;
1950	>	TreeNode<K,V> p = (r == null) ? null :
1951	>	r.findTreeNode(h, key, null);
1952	>	val = (p == null) ? value :
1953	>	remappingFunction.apply(p.val, value);
1954	>	if (val != null) {
1955	>	if (p != null)
1956	>	p.val = val;
1957	>	else {
1958	>	delta = 1;
1959	>	t.putTreeVal(h, key, val);
1960	>	}
1961	>	}
1962	>	else if (p != null) {
1963	>	delta = -1;
1964	>	if (t.removeTreeNode(p))
1965	>	setTabAt(tab, i, untreeify(t.first));
1966	>	}
1967	>	}
1968		}
1969		}
1970	<	if (retries > 0 && sum == last)
1970	>	if (binCount != 0) {
1971	>	if (binCount >= TREEIFY_THRESHOLD)
1972	>	treeifyBin(tab, i);
1973		break;
1974	<	last = sum;
982	<	}
983	<	} finally {
984	<	if (retries > RETRIES_BEFORE_LOCK) {
985	<	for (int j = 0; j < segments.length; ++j)
986	<	segmentAt(segments, j).unlock();
1974	>	}
1975		}
1976		}
1977	<	return found;
1977	>	if (delta != 0)
1978	>	addCount((long)delta, binCount);
1979	>	return val;
1980		}
1981
1982	+	// Hashtable legacy methods
1983	+
1984		/**
1985		* Legacy method testing if some key maps into the specified value
1986		* in this table.  This method is identical in functionality to
1987	<	* {@link #containsValue}, and exists solely to ensure
1987	>	* {@link #containsValue(Object)}, and exists solely to ensure
1988		* full compatibility with class {@link java.util.Hashtable},
1989		* which supported this method prior to introduction of the
1990		* Java Collections framework.
1991		*
1992		* @param  value a value to search for
1993	<	* @return <tt>true</tt> if and only if some key maps to the
1994	<	*         <tt>value</tt> argument in this table as
1995	<	*         determined by the <tt>equals</tt> method;
1996	<	*         <tt>false</tt> otherwise
1993	>	* @return {@code true} if and only if some key maps to the
1994	>	*         {@code value} argument in this table as
1995	>	*         determined by the {@code equals} method;
1996	>	*         {@code false} otherwise
1997		* @throws NullPointerException if the specified value is null
1998		*/
1999	<	public boolean contains(Object value) {
1999	>	@Deprecated public boolean contains(Object value) {
2000		return containsValue(value);
2001		}
2002
2003		/**
2004	<	* Maps the specified key to the specified value in this table.
1013	<	* Neither the key nor the value can be null.
1014	<	*
1015	<	* <p> The value can be retrieved by calling the <tt>get</tt> method
1016	<	* with a key that is equal to the original key.
2004	>	* Returns an enumeration of the keys in this table.
2005		*
2006	<	* @param key key with which the specified value is to be associated
2007	<	* @param value value to be associated with the specified key
1020	<	* @return the previous value associated with <tt>key</tt>, or
1021	<	*         <tt>null</tt> if there was no mapping for <tt>key</tt>
1022	<	* @throws NullPointerException if the specified key or value is null
2006	>	* @return an enumeration of the keys in this table
2007	>	* @see #keySet()
2008		*/
2009	<	@SuppressWarnings("unchecked")
2010	<	public V put(K key, V value) {
2011	<	Segment<K,V> s;
2012	<	if (value == null)
1028	<	throw new NullPointerException();
1029	<	int hash = hash(key.hashCode());
1030	<	int j = (hash >>> segmentShift) & segmentMask;
1031	<	if ((s = (Segment<K,V>)UNSAFE.getObject          // nonvolatile; recheck
1032	<	(segments, (j << SSHIFT) + SBASE)) == null) //  in ensureSegment
1033	<	s = ensureSegment(j);
1034	<	return s.put(key, hash, value, false);
2009	>	public Enumeration<K> keys() {
2010	>	Node<K,V>[] t;
2011	>	int f = (t = table) == null ? 0 : t.length;
2012	>	return new KeyIterator<K,V>(t, f, 0, f, this);
2013		}
2014
2015		/**
2016	<	* {@inheritDoc}
2016	>	* Returns an enumeration of the values in this table.
2017		*
2018	<	* @return the previous value associated with the specified key,
2019	<	*         or <tt>null</tt> if there was no mapping for the key
1042	<	* @throws NullPointerException if the specified key or value is null
2018	>	* @return an enumeration of the values in this table
2019	>	* @see #values()
2020		*/
2021	<	@SuppressWarnings("unchecked")
2022	<	public V putIfAbsent(K key, V value) {
2023	<	Segment<K,V> s;
2024	<	if (value == null)
1048	<	throw new NullPointerException();
1049	<	int hash = hash(key.hashCode());
1050	<	int j = (hash >>> segmentShift) & segmentMask;
1051	<	if ((s = (Segment<K,V>)UNSAFE.getObject
1052	<	(segments, (j << SSHIFT) + SBASE)) == null)
1053	<	s = ensureSegment(j);
1054	<	return s.put(key, hash, value, true);
2021	>	public Enumeration<V> elements() {
2022	>	Node<K,V>[] t;
2023	>	int f = (t = table) == null ? 0 : t.length;
2024	>	return new ValueIterator<K,V>(t, f, 0, f, this);
2025		}
2026
2027	+	// ConcurrentHashMap-only methods
2028	+
2029		/**
2030	<	* Copies all of the mappings from the specified map to this one.
2031	<	* These mappings replace any mappings that this map had for any of the
2032	<	* keys currently in the specified map.
2030	>	* Returns the number of mappings. This method should be used
2031	>	* instead of {@link #size} because a ConcurrentHashMap may
2032	>	* contain more mappings than can be represented as an int. The
2033	>	* value returned is an estimate; the actual count may differ if
2034	>	* there are concurrent insertions or removals.
2035		*
2036	<	* @param m mappings to be stored in this map
2036	>	* @return the number of mappings
2037	>	* @since 1.8
2038		*/
2039	<	public void putAll(Map<? extends K, ? extends V> m) {
2040	<	for (Map.Entry<? extends K, ? extends V> e : m.entrySet())
2041	<	put(e.getKey(), e.getValue());
2039	>	public long mappingCount() {
2040	>	long n = sumCount();
2041	>	return (n < 0L) ? 0L : n; // ignore transient negative values
2042		}
2043
2044		/**
2045	<	* Removes the key (and its corresponding value) from this map.
2046	<	* This method does nothing if the key is not in the map.
2045	>	* Creates a new {@link Set} backed by a ConcurrentHashMap
2046	>	* from the given type to {@code Boolean.TRUE}.
2047		*
2048	<	* @param  key the key that needs to be removed
2049	<	* @return the previous value associated with <tt>key</tt>, or
1075	<	*         <tt>null</tt> if there was no mapping for <tt>key</tt>
1076	<	* @throws NullPointerException if the specified key is null
2048	>	* @return the new set
2049	>	* @since 1.8
2050		*/
2051	<	public V remove(Object key) {
2052	<	int hash = hash(key.hashCode());
2053	<	Segment<K,V> s = segmentForHash(hash);
1081	<	return s == null ? null : s.remove(key, hash, null);
2051	>	public static <K> KeySetView<K,Boolean> newKeySet() {
2052	>	return new KeySetView<K,Boolean>
2053	>	(new ConcurrentHashMap<K,Boolean>(), Boolean.TRUE);
2054		}
2055
2056		/**
2057	<	* {@inheritDoc}
2057	>	* Creates a new {@link Set} backed by a ConcurrentHashMap
2058	>	* from the given type to {@code Boolean.TRUE}.
2059		*
2060	<	* @throws NullPointerException if the specified key is null
2060	>	* @param initialCapacity The implementation performs internal
2061	>	* sizing to accommodate this many elements.
2062	>	* @throws IllegalArgumentException if the initial capacity of
2063	>	* elements is negative
2064	>	* @return the new set
2065	>	* @since 1.8
2066		*/
2067	<	public boolean remove(Object key, Object value) {
2068	<	int hash = hash(key.hashCode());
2069	<	Segment<K,V> s;
1092	<	return value != null && (s = segmentForHash(hash)) != null &&
1093	<	s.remove(key, hash, value) != null;
2067	>	public static <K> KeySetView<K,Boolean> newKeySet(int initialCapacity) {
2068	>	return new KeySetView<K,Boolean>
2069	>	(new ConcurrentHashMap<K,Boolean>(initialCapacity), Boolean.TRUE);
2070		}
2071
2072		/**
2073	<	* {@inheritDoc}
2073	>	* Returns a {@link Set} view of the keys in this map, using the
2074	>	* given common mapped value for any additions (i.e., {@link
2075	>	* Collection#add} and {@link Collection#addAll(Collection)}).
2076	>	* This is of course only appropriate if it is acceptable to use
2077	>	* the same value for all additions from this view.
2078		*
2079	<	* @throws NullPointerException if any of the arguments are null
2079	>	* @param mappedValue the mapped value to use for any additions
2080	>	* @return the set view
2081	>	* @throws NullPointerException if the mappedValue is null
2082		*/
2083	<	public boolean replace(K key, V oldValue, V newValue) {
2084	<	int hash = hash(key.hashCode());
1103	<	if (oldValue == null || newValue == null)
2083	>	public KeySetView<K,V> keySet(V mappedValue) {
2084	>	if (mappedValue == null)
2085		throw new NullPointerException();
2086	<	Segment<K,V> s = segmentForHash(hash);
1106	<	return s != null && s.replace(key, hash, oldValue, newValue);
2086	>	return new KeySetView<K,V>(this, mappedValue);
2087		}
2088
2089	+	/* ---------------- Special Nodes -------------- */
2090	+
2091		/**
2092	<	* {@inheritDoc}
1111	<	*
1112	<	* @return the previous value associated with the specified key,
1113	<	*         or <tt>null</tt> if there was no mapping for the key
1114	<	* @throws NullPointerException if the specified key or value is null
2092	>	* A node inserted at head of bins during transfer operations.
2093		*/
2094	<	public V replace(K key, V value) {
2095	<	int hash = hash(key.hashCode());
2096	<	if (value == null)
2097	<	throw new NullPointerException();
2098	<	Segment<K,V> s = segmentForHash(hash);
2099	<	return s == null ? null : s.replace(key, hash, value);
2094	>	static final class ForwardingNode<K,V> extends Node<K,V> {
2095	>	final Node<K,V>[] nextTable;
2096	>	ForwardingNode(Node<K,V>[] tab) {
2097	>	super(MOVED, null, null, null);
2098	>	this.nextTable = tab;
2099	>	}
2100	>
2101	>	Node<K,V> find(int h, Object k) {
2102	>	Node<K,V> e; int n;
2103	>	Node<K,V>[] tab = nextTable;
2104	>	if (k != null && tab != null && (n = tab.length) > 0 &&
2105	>	(e = tabAt(tab, (n - 1) & h)) != null) {
2106	>	do {
2107	>	int eh; K ek;
2108	>	if ((eh = e.hash) == h &&
2109	>	((ek = e.key) == k || (ek != null && k.equals(ek))))
2110	>	return e;
2111	>	if (eh < 0)
2112	>	return e.find(h, k);
2113	>	} while ((e = e.next) != null);
2114	>	}
2115	>	return null;
2116	>	}
2117		}
2118
2119		/**
2120	<	* Removes all of the mappings from this map.
2120	>	* A place-holder node used in computeIfAbsent and compute
2121		*/
2122	<	public void clear() {
2123	<	final Segment<K,V>[] segments = this.segments;
2124	<	for (int j = 0; j < segments.length; ++j) {
2125	<	Segment<K,V> s = segmentAt(segments, j);
2126	<	if (s != null)
2127	<	s.clear();
2122	>	static final class ReservationNode<K,V> extends Node<K,V> {
2123	>	ReservationNode() {
2124	>	super(RESERVED, null, null, null);
2125	>	}
2126	>
2127	>	Node<K,V> find(int h, Object k) {
2128	>	return null;
2129		}
2130		}
2131
2132	+	/* ---------------- Table Initialization and Resizing -------------- */
2133	+
2134		/**
2135	<	* Returns a {@link Set} view of the keys contained in this map.
1138	<	* The set is backed by the map, so changes to the map are
1139	<	* reflected in the set, and vice-versa.  The set supports element
1140	<	* removal, which removes the corresponding mapping from this map,
1141	<	* via the <tt>Iterator.remove</tt>, <tt>Set.remove</tt>,
1142	<	* <tt>removeAll</tt>, <tt>retainAll</tt>, and <tt>clear</tt>
1143	<	* operations.  It does not support the <tt>add</tt> or
1144	<	* <tt>addAll</tt> operations.
1145	<	*
1146	<	* <p>The view's <tt>iterator</tt> is a "weakly consistent" iterator
1147	<	* that will never throw {@link ConcurrentModificationException},
1148	<	* and guarantees to traverse elements as they existed upon
1149	<	* construction of the iterator, and may (but is not guaranteed to)
1150	<	* reflect any modifications subsequent to construction.
2135	>	* Initializes table, using the size recorded in sizeCtl.
2136		*/
2137	<	public Set<K> keySet() {
2138	<	Set<K> ks = keySet;
2139	<	return (ks != null) ? ks : (keySet = new KeySet());
2137	>	private final Node<K,V>[] initTable() {
2138	>	Node<K,V>[] tab; int sc;
2139	>	while ((tab = table) == null || tab.length == 0) {
2140	>	if ((sc = sizeCtl) < 0)
2141	>	Thread.yield(); // lost initialization race; just spin
2142	>	else if (U.compareAndSwapInt(this, SIZECTL, sc, -1)) {
2143	>	try {
2144	>	if ((tab = table) == null || tab.length == 0) {
2145	>	int n = (sc > 0) ? sc : DEFAULT_CAPACITY;
2146	>	@SuppressWarnings({"rawtypes","unchecked"})
2147	>	Node<K,V>[] nt = (Node<K,V>[])new Node[n];
2148	>	table = tab = nt;
2149	>	sc = n - (n >>> 2);
2150	>	}
2151	>	} finally {
2152	>	sizeCtl = sc;
2153	>	}
2154	>	break;
2155	>	}
2156	>	}
2157	>	return tab;
2158		}
2159
2160		/**
2161	<	* Returns a {@link Collection} view of the values contained in this map.
2162	<	* The collection is backed by the map, so changes to the map are
2163	<	* reflected in the collection, and vice-versa.  The collection
2164	<	* supports element removal, which removes the corresponding
2165	<	* mapping from this map, via the <tt>Iterator.remove</tt>,
1163	<	* <tt>Collection.remove</tt>, <tt>removeAll</tt>,
1164	<	* <tt>retainAll</tt>, and <tt>clear</tt> operations.  It does not
1165	<	* support the <tt>add</tt> or <tt>addAll</tt> operations.
2161	>	* Adds to count, and if table is too small and not already
2162	>	* resizing, initiates transfer. If already resizing, helps
2163	>	* perform transfer if work is available.  Rechecks occupancy
2164	>	* after a transfer to see if another resize is already needed
2165	>	* because resizings are lagging additions.
2166		*
2167	<	* <p>The view's <tt>iterator</tt> is a "weakly consistent" iterator
2168	<	* that will never throw {@link ConcurrentModificationException},
1169	<	* and guarantees to traverse elements as they existed upon
1170	<	* construction of the iterator, and may (but is not guaranteed to)
1171	<	* reflect any modifications subsequent to construction.
2167	>	* @param x the count to add
2168	>	* @param check if <0, don't check resize, if <= 1 only check if uncontended
2169		*/
2170	<	public Collection<V> values() {
2171	<	Collection<V> vs = values;
2172	<	return (vs != null) ? vs : (values = new Values());
2170	>	private final void addCount(long x, int check) {
2171	>	CounterCell[] as; long b, s;
2172	>	if ((as = counterCells) != null ||
2173	>	!U.compareAndSwapLong(this, BASECOUNT, b = baseCount, s = b + x)) {
2174	>	CounterCell a; long v; int m;
2175	>	boolean uncontended = true;
2176	>	if (as == null || (m = as.length - 1) < 0 ||
2177	>	(a = as[ThreadLocalRandom.getProbe() & m]) == null ||
2178	>	!(uncontended =
2179	>	U.compareAndSwapLong(a, CELLVALUE, v = a.value, v + x))) {
2180	>	fullAddCount(x, uncontended);
2181	>	return;
2182	>	}
2183	>	if (check <= 1)
2184	>	return;
2185	>	s = sumCount();
2186	>	}
2187	>	if (check >= 0) {
2188	>	Node<K,V>[] tab, nt; int sc;
2189	>	while (s >= (long)(sc = sizeCtl) && (tab = table) != null &&
2190	>	tab.length < MAXIMUM_CAPACITY) {
2191	>	if (sc < 0) {
2192	>	if (sc == -1 || transferIndex <= transferOrigin ||
2193	>	(nt = nextTable) == null)
2194	>	break;
2195	>	if (U.compareAndSwapInt(this, SIZECTL, sc, sc - 1))
2196	>	transfer(tab, nt);
2197	>	}
2198	>	else if (U.compareAndSwapInt(this, SIZECTL, sc, -2))
2199	>	transfer(tab, null);
2200	>	s = sumCount();
2201	>	}
2202	>	}
2203		}
2204
2205		/**
2206	<	* Returns a {@link Set} view of the mappings contained in this map.
1180	<	* The set is backed by the map, so changes to the map are
1181	<	* reflected in the set, and vice-versa.  The set supports element
1182	<	* removal, which removes the corresponding mapping from the map,
1183	<	* via the <tt>Iterator.remove</tt>, <tt>Set.remove</tt>,
1184	<	* <tt>removeAll</tt>, <tt>retainAll</tt>, and <tt>clear</tt>
1185	<	* operations.  It does not support the <tt>add</tt> or
1186	<	* <tt>addAll</tt> operations.
1187	<	*
1188	<	* <p>The view's <tt>iterator</tt> is a "weakly consistent" iterator
1189	<	* that will never throw {@link ConcurrentModificationException},
1190	<	* and guarantees to traverse elements as they existed upon
1191	<	* construction of the iterator, and may (but is not guaranteed to)
1192	<	* reflect any modifications subsequent to construction.
2206	>	* Helps transfer if a resize is in progress.
2207		*/
2208	<	public Set<Map.Entry<K,V>> entrySet() {
2209	<	Set<Map.Entry<K,V>> es = entrySet;
2210	<	return (es != null) ? es : (entrySet = new EntrySet());
2208	>	final Node<K,V>[] helpTransfer(Node<K,V>[] tab, Node<K,V> f) {
2209	>	Node<K,V>[] nextTab; int sc;
2210	>	if ((f instanceof ForwardingNode) &&
2211	>	(nextTab = ((ForwardingNode<K,V>)f).nextTable) != null) {
2212	>	if (nextTab == nextTable && tab == table &&
2213	>	transferIndex > transferOrigin && (sc = sizeCtl) < -1 &&
2214	>	U.compareAndSwapInt(this, SIZECTL, sc, sc - 1))
2215	>	transfer(tab, nextTab);
2216	>	return nextTab;
2217	>	}
2218	>	return table;
2219		}
2220
2221		/**
2222	<	* Returns an enumeration of the keys in this table.
2222	>	* Tries to presize table to accommodate the given number of elements.
2223		*
2224	<	* @return an enumeration of the keys in this table
1203	<	* @see #keySet()
2224	>	* @param size number of elements (doesn't need to be perfectly accurate)
2225		*/
2226	<	public Enumeration<K> keys() {
2227	<	return new KeyIterator();
2226	>	private final void tryPresize(int size) {
2227	>	int c = (size >= (MAXIMUM_CAPACITY >>> 1)) ? MAXIMUM_CAPACITY :
2228	>	tableSizeFor(size + (size >>> 1) + 1);
2229	>	int sc;
2230	>	while ((sc = sizeCtl) >= 0) {
2231	>	Node<K,V>[] tab = table; int n;
2232	>	if (tab == null || (n = tab.length) == 0) {
2233	>	n = (sc > c) ? sc : c;
2234	>	if (U.compareAndSwapInt(this, SIZECTL, sc, -1)) {
2235	>	try {
2236	>	if (table == tab) {
2237	>	@SuppressWarnings({"rawtypes","unchecked"})
2238	>	Node<K,V>[] nt = (Node<K,V>[])new Node[n];
2239	>	table = nt;
2240	>	sc = n - (n >>> 2);
2241	>	}
2242	>	} finally {
2243	>	sizeCtl = sc;
2244	>	}
2245	>	}
2246	>	}
2247	>	else if (c <= sc || n >= MAXIMUM_CAPACITY)
2248	>	break;
2249	>	else if (tab == table &&
2250	>	U.compareAndSwapInt(this, SIZECTL, sc, -2))
2251	>	transfer(tab, null);
2252	>	}
2253		}
2254
2255		/**
2256	<	* Returns an enumeration of the values in this table.
2257	<	*
1212	<	* @return an enumeration of the values in this table
1213	<	* @see #values()
2256	>	* Moves and/or copies the nodes in each bin to new table. See
2257	>	* above for explanation.
2258		*/
2259	<	public Enumeration<V> elements() {
2260	<	return new ValueIterator();
2259	>	private final void transfer(Node<K,V>[] tab, Node<K,V>[] nextTab) {
2260	>	int n = tab.length, stride;
2261	>	if ((stride = (NCPU > 1) ? (n >>> 3) / NCPU : n) < MIN_TRANSFER_STRIDE)
2262	>	stride = MIN_TRANSFER_STRIDE; // subdivide range
2263	>	if (nextTab == null) {            // initiating
2264	>	try {
2265	>	@SuppressWarnings({"rawtypes","unchecked"})
2266	>	Node<K,V>[] nt = (Node<K,V>[])new Node[n << 1];
2267	>	nextTab = nt;
2268	>	} catch (Throwable ex) {      // try to cope with OOME
2269	>	sizeCtl = Integer.MAX_VALUE;
2270	>	return;
2271	>	}
2272	>	nextTable = nextTab;
2273	>	transferOrigin = n;
2274	>	transferIndex = n;
2275	>	ForwardingNode<K,V> rev = new ForwardingNode<K,V>(tab);
2276	>	for (int k = n; k > 0;) {    // progressively reveal ready slots
2277	>	int nextk = (k > stride) ? k - stride : 0;
2278	>	for (int m = nextk; m < k; ++m)
2279	>	nextTab[m] = rev;
2280	>	for (int m = n + nextk; m < n + k; ++m)
2281	>	nextTab[m] = rev;
2282	>	U.putOrderedInt(this, TRANSFERORIGIN, k = nextk);
2283	>	}
2284	>	}
2285	>	int nextn = nextTab.length;
2286	>	ForwardingNode<K,V> fwd = new ForwardingNode<K,V>(nextTab);
2287	>	boolean advance = true;
2288	>	for (int i = 0, bound = 0;;) {
2289	>	int nextIndex, nextBound, fh; Node<K,V> f;
2290	>	while (advance) {
2291	>	if (--i >= bound)
2292	>	advance = false;
2293	>	else if ((nextIndex = transferIndex) <= transferOrigin) {
2294	>	i = -1;
2295	>	advance = false;
2296	>	}
2297	>	else if (U.compareAndSwapInt
2298	>	(this, TRANSFERINDEX, nextIndex,
2299	>	nextBound = (nextIndex > stride ?
2300	>	nextIndex - stride : 0))) {
2301	>	bound = nextBound;
2302	>	i = nextIndex - 1;
2303	>	advance = false;
2304	>	}
2305	>	}
2306	>	if (i < 0 || i >= n || i + n >= nextn) {
2307	>	for (int sc;;) {
2308	>	if (U.compareAndSwapInt(this, SIZECTL, sc = sizeCtl, ++sc)) {
2309	>	if (sc == -1) {
2310	>	nextTable = null;
2311	>	table = nextTab;
2312	>	sizeCtl = (n << 1) - (n >>> 1);
2313	>	}
2314	>	return;
2315	>	}
2316	>	}
2317	>	}
2318	>	else if ((f = tabAt(tab, i)) == null) {
2319	>	if (casTabAt(tab, i, null, fwd)) {
2320	>	setTabAt(nextTab, i, null);
2321	>	setTabAt(nextTab, i + n, null);
2322	>	advance = true;
2323	>	}
2324	>	}
2325	>	else if ((fh = f.hash) == MOVED)
2326	>	advance = true; // already processed
2327	>	else {
2328	>	synchronized (f) {
2329	>	if (tabAt(tab, i) == f) {
2330	>	Node<K,V> ln, hn;
2331	>	if (fh >= 0) {
2332	>	int runBit = fh & n;
2333	>	Node<K,V> lastRun = f;
2334	>	for (Node<K,V> p = f.next; p != null; p = p.next) {
2335	>	int b = p.hash & n;
2336	>	if (b != runBit) {
2337	>	runBit = b;
2338	>	lastRun = p;
2339	>	}
2340	>	}
2341	>	if (runBit == 0) {
2342	>	ln = lastRun;
2343	>	hn = null;
2344	>	}
2345	>	else {
2346	>	hn = lastRun;
2347	>	ln = null;
2348	>	}
2349	>	for (Node<K,V> p = f; p != lastRun; p = p.next) {
2350	>	int ph = p.hash; K pk = p.key; V pv = p.val;
2351	>	if ((ph & n) == 0)
2352	>	ln = new Node<K,V>(ph, pk, pv, ln);
2353	>	else
2354	>	hn = new Node<K,V>(ph, pk, pv, hn);
2355	>	}
2356	>	}
2357	>	else if (f instanceof TreeBin) {
2358	>	TreeBin<K,V> t = (TreeBin<K,V>)f;
2359	>	TreeNode<K,V> lo = null, loTail = null;
2360	>	TreeNode<K,V> hi = null, hiTail = null;
2361	>	int lc = 0, hc = 0;
2362	>	for (Node<K,V> e = t.first; e != null; e = e.next) {
2363	>	int h = e.hash;
2364	>	TreeNode<K,V> p = new TreeNode<K,V>
2365	>	(h, e.key, e.val, null, null);
2366	>	if ((h & n) == 0) {
2367	>	if ((p.prev = loTail) == null)
2368	>	lo = p;
2369	>	else
2370	>	loTail.next = p;
2371	>	loTail = p;
2372	>	++lc;
2373	>	}
2374	>	else {
2375	>	if ((p.prev = hiTail) == null)
2376	>	hi = p;
2377	>	else
2378	>	hiTail.next = p;
2379	>	hiTail = p;
2380	>	++hc;
2381	>	}
2382	>	}
2383	>	ln = (lc <= UNTREEIFY_THRESHOLD ?  untreeify(lo) :
2384	>	(hc != 0) ? new TreeBin<K,V>(lo) : t);
2385	>	hn = (hc <= UNTREEIFY_THRESHOLD ? untreeify(hi) :
2386	>	(lc != 0) ? new TreeBin<K,V>(hi) : t);
2387	>	}
2388	>	else
2389	>	ln = hn = null;
2390	>	setTabAt(nextTab, i, ln);
2391	>	setTabAt(nextTab, i + n, hn);
2392	>	setTabAt(tab, i, fwd);
2393	>	advance = true;
2394	>	}
2395	>	}
2396	>	}
2397	>	}
2398	>	}
2399	>
2400	>	/* ---------------- Counter support -------------- */
2401	>
2402	>	/**
2403	>	* A padded cell for distributing counts.  Adapted from LongAdder
2404	>	* and Striped64.  See their internal docs for explanation.
2405	>	*/
2406	>	@sun.misc.Contended static final class CounterCell {
2407	>	volatile long value;
2408	>	CounterCell(long x) { value = x; }
2409	>	}
2410	>
2411	>	final long sumCount() {
2412	>	CounterCell[] as = counterCells; CounterCell a;
2413	>	long sum = baseCount;
2414	>	if (as != null) {
2415	>	for (int i = 0; i < as.length; ++i) {
2416	>	if ((a = as[i]) != null)
2417	>	sum += a.value;
2418	>	}
2419	>	}
2420	>	return sum;
2421	>	}
2422	>
2423	>	// See LongAdder version for explanation
2424	>	private final void fullAddCount(long x, boolean wasUncontended) {
2425	>	int h;
2426	>	if ((h = ThreadLocalRandom.getProbe()) == 0) {
2427	>	ThreadLocalRandom.localInit();      // force initialization
2428	>	h = ThreadLocalRandom.getProbe();
2429	>	wasUncontended = true;
2430	>	}
2431	>	boolean collide = false;                // True if last slot nonempty
2432	>	for (;;) {
2433	>	CounterCell[] as; CounterCell a; int n; long v;
2434	>	if ((as = counterCells) != null && (n = as.length) > 0) {
2435	>	if ((a = as[(n - 1) & h]) == null) {
2436	>	if (cellsBusy == 0) {            // Try to attach new Cell
2437	>	CounterCell r = new CounterCell(x); // Optimistic create
2438	>	if (cellsBusy == 0 &&
2439	>	U.compareAndSwapInt(this, CELLSBUSY, 0, 1)) {
2440	>	boolean created = false;
2441	>	try {               // Recheck under lock
2442	>	CounterCell[] rs; int m, j;
2443	>	if ((rs = counterCells) != null &&
2444	>	(m = rs.length) > 0 &&
2445	>	rs[j = (m - 1) & h] == null) {
2446	>	rs[j] = r;
2447	>	created = true;
2448	>	}
2449	>	} finally {
2450	>	cellsBusy = 0;
2451	>	}
2452	>	if (created)
2453	>	break;
2454	>	continue;           // Slot is now non-empty
2455	>	}
2456	>	}
2457	>	collide = false;
2458	>	}
2459	>	else if (!wasUncontended)       // CAS already known to fail
2460	>	wasUncontended = true;      // Continue after rehash
2461	>	else if (U.compareAndSwapLong(a, CELLVALUE, v = a.value, v + x))
2462	>	break;
2463	>	else if (counterCells != as || n >= NCPU)
2464	>	collide = false;            // At max size or stale
2465	>	else if (!collide)
2466	>	collide = true;
2467	>	else if (cellsBusy == 0 &&
2468	>	U.compareAndSwapInt(this, CELLSBUSY, 0, 1)) {
2469	>	try {
2470	>	if (counterCells == as) {// Expand table unless stale
2471	>	CounterCell[] rs = new CounterCell[n << 1];
2472	>	for (int i = 0; i < n; ++i)
2473	>	rs[i] = as[i];
2474	>	counterCells = rs;
2475	>	}
2476	>	} finally {
2477	>	cellsBusy = 0;
2478	>	}
2479	>	collide = false;
2480	>	continue;                   // Retry with expanded table
2481	>	}
2482	>	h = ThreadLocalRandom.advanceProbe(h);
2483	>	}
2484	>	else if (cellsBusy == 0 && counterCells == as &&
2485	>	U.compareAndSwapInt(this, CELLSBUSY, 0, 1)) {
2486	>	boolean init = false;
2487	>	try {                           // Initialize table
2488	>	if (counterCells == as) {
2489	>	CounterCell[] rs = new CounterCell[2];
2490	>	rs[h & 1] = new CounterCell(x);
2491	>	counterCells = rs;
2492	>	init = true;
2493	>	}
2494	>	} finally {
2495	>	cellsBusy = 0;
2496	>	}
2497	>	if (init)
2498	>	break;
2499	>	}
2500	>	else if (U.compareAndSwapLong(this, BASECOUNT, v = baseCount, v + x))
2501	>	break;                          // Fall back on using base
2502	>	}
2503		}
2504
2505	<	/* ---------------- Iterator Support -------------- */
2505	>	/* ---------------- Conversion from/to TreeBins -------------- */
2506
2507	<	abstract class HashIterator {
2508	<	int nextSegmentIndex;
2509	<	int nextTableIndex;
2510	<	HashEntry<K,V>[] currentTable;
2511	<	HashEntry<K, V> nextEntry;
2512	<	HashEntry<K, V> lastReturned;
2507	>	/**
2508	>	* Replaces all linked nodes in bin at given index unless table is
2509	>	* too small, in which case resizes instead.
2510	>	*/
2511	>	private final void treeifyBin(Node<K,V>[] tab, int index) {
2512	>	Node<K,V> b; int n, sc;
2513	>	if (tab != null) {
2514	>	if ((n = tab.length) < MIN_TREEIFY_CAPACITY &&
2515	>	tab == table && (sc = sizeCtl) >= 0 &&
2516	>	U.compareAndSwapInt(this, SIZECTL, sc, -2))
2517	>	transfer(tab, null);
2518	>	else if ((b = tabAt(tab, index)) != null) {
2519	>	synchronized(b) {
2520	>	if (tabAt(tab, index) == b) {
2521	>	TreeNode<K,V> hd = null, tl = null;
2522	>	for (Node<K,V> e = b; e != null; e = e.next) {
2523	>	TreeNode<K,V> p =
2524	>	new TreeNode<K,V>(e.hash, e.key, e.val,
2525	>	null, null);
2526	>	if ((p.prev = tl) == null)
2527	>	hd = p;
2528	>	else
2529	>	tl.next = p;
2530	>	tl = p;
2531	>	}
2532	>	setTabAt(tab, index, new TreeBin<K,V>(hd));
2533	>	}
2534	>	}
2535	>	}
2536	>	}
2537	>	}
2538
2539	<	HashIterator() {
2540	<	nextSegmentIndex = segments.length - 1;
2541	<	nextTableIndex = -1;
2542	<	advance();
2539	>	/**
2540	>	* Returns a list on non-TreeNodes replacing those in given list
2541	>	*/
2542	>	static <K,V> Node<K,V> untreeify(Node<K,V> b) {
2543	>	Node<K,V> hd = null, tl = null;
2544	>	for (Node<K,V> q = b; q != null; q = q.next) {
2545	>	Node<K,V> p = new Node<K,V>(q.hash, q.key, q.val, null);
2546	>	if (tl == null)
2547	>	hd = p;
2548	>	else
2549	>	tl.next = p;
2550	>	tl = p;
2551	>	}
2552	>	return hd;
2553	>	}
2554	>
2555	>	/* ---------------- TreeNodes -------------- */
2556	>
2557	>	/**
2558	>	* Nodes for use in TreeBins
2559	>	*/
2560	>	static final class TreeNode<K,V> extends Node<K,V> {
2561	>	TreeNode<K,V> parent;  // red-black tree links
2562	>	TreeNode<K,V> left;
2563	>	TreeNode<K,V> right;
2564	>	TreeNode<K,V> prev;    // needed to unlink next upon deletion
2565	>	boolean red;
2566	>
2567	>	TreeNode(int hash, K key, V val, Node<K,V> next,
2568	>	TreeNode<K,V> parent) {
2569	>	super(hash, key, val, next);
2570	>	this.parent = parent;
2571	>	}
2572	>
2573	>	Node<K,V> find(int h, Object k) {
2574	>	return findTreeNode(h, k, null);
2575	>	}
2576	>
2577	>	/**
2578	>	* Returns the TreeNode (or null if not found) for the given key
2579	>	* starting at given root.
2580	>	*/
2581	>	final TreeNode<K,V> findTreeNode(int h, Object k, Class<?> kc) {
2582	>	if (k == null)
2583	>	return null;
2584	>	TreeNode<K,V> p = this;
2585	>	do  {
2586	>	int ph, dir; K pk; TreeNode<K,V> q;
2587	>	TreeNode<K,V> pl = p.left, pr = p.right;
2588	>	if ((ph = p.hash) > h)
2589	>	p = pl;
2590	>	else if (ph < h)
2591	>	p = pr;
2592	>	else if ((pk = p.key) == k || (pk != null && k.equals(pk)))
2593	>	return p;
2594	>	else if (pl == null && pr == null)
2595	>	break;
2596	>	else if ((kc != null || (kc = comparableClassFor(k)) != null) &&
2597	>	(dir = compareComparables(kc, k, pk)) != 0)
2598	>	p = (dir < 0) ? pl : pr;
2599	>	else if (pl == null)
2600	>	p = pr;
2601	>	else if (pr == null || (q = pr.findTreeNode(h, k, kc)) == null)
2602	>	p = pl;
2603	>	else
2604	>	return q;
2605	>	} while (p != null);
2606	>	return null;
2607	>	}
2608	>	}
2609	>
2610	>	/* ---------------- TreeBins -------------- */
2611	>
2612	>	/**
2613	>	* TreeNodes used at the heads of bins. TreeBins do not hold user
2614	>	* keys or values, but instead point to list of TreeNodes and
2615	>	* their root. They also maintain a parasitic read-write lock
2616	>	* forcing writers (who hold bin lock) to wait for readers (who do
2617	>	* not) to complete before tree restructuring operations.
2618	>	*/
2619	>	static final class TreeBin<K,V> extends Node<K,V> {
2620	>	TreeNode<K,V> root;
2621	>	volatile TreeNode<K,V> first;
2622	>	volatile Thread waiter;
2623	>	static final int WRITER = 1; // values for lockState
2624	>	static final int WAITER = 2;
2625	>	static final int READER = 4;
2626	>	volatile int lockState;
2627	>
2628	>	/**
2629	>	* Creates bin with initial set of nodes headed by b.
2630	>	*/
2631	>	TreeBin(TreeNode<K,V> b) {
2632	>	super(TREEBIN, null, null, null);
2633	>	first = b;
2634	>	TreeNode<K,V> r = null;
2635	>	for (TreeNode<K,V> x = b, next; x != null; x = next) {
2636	>	next = (TreeNode<K,V>)x.next;
2637	>	x.left = x.right = null;
2638	>	if (r == null) {
2639	>	x.parent = null;
2640	>	x.red = false;
2641	>	r = x;
2642	>	}
2643	>	else {
2644	>	Object key = x.key;
2645	>	int hash = x.hash;
2646	>	Class<?> kc = null;
2647	>	for (TreeNode<K,V> p = r;;) {
2648	>	int dir, ph;
2649	>	if ((ph = p.hash) > hash)
2650	>	dir = -1;
2651	>	else if (ph < hash)
2652	>	dir = 1;
2653	>	else if ((kc != null ||
2654	>	(kc = comparableClassFor(key)) != null))
2655	>	dir = compareComparables(kc, key, p.key);
2656	>	else
2657	>	dir = 0;
2658	>	TreeNode<K,V> xp = p;
2659	>	if ((p = (dir <= 0) ? p.left : p.right) == null) {
2660	>	x.parent = xp;
2661	>	if (dir <= 0)
2662	>	xp.left = x;
2663	>	else
2664	>	xp.right = x;
2665	>	r = balanceInsertion(r, x);
2666	>	break;
2667	>	}
2668	>	}
2669	>	}
2670	>	}
2671	>	root = r;
2672		}
2673
2674		/**
2675	<	* Sets nextEntry to first node of next non-empty table
1236	<	* (in backwards order, to simplify checks).
2675	>	* Acquires write lock for tree restructuring
2676		*/
2677	<	final void advance() {
2677	>	private final void lockRoot() {
2678	>	if (!U.compareAndSwapInt(this, LOCKSTATE, 0, WRITER))
2679	>	contendedLock(); // offload to separate method
2680	>	}
2681	>
2682	>	/**
2683	>	* Releases write lock for tree restructuring
2684	>	*/
2685	>	private final void unlockRoot() {
2686	>	lockState = 0;
2687	>	}
2688	>
2689	>	/**
2690	>	* Possibly blocks awaiting root lock
2691	>	*/
2692	>	private final void contendedLock() {
2693	>	boolean waiting = false;
2694	>	for (int s;;) {
2695	>	if (((s = lockState) & WRITER) == 0) {
2696	>	if (U.compareAndSwapInt(this, LOCKSTATE, s, WRITER)) {
2697	>	if (waiting)
2698	>	waiter = null;
2699	>	return;
2700	>	}
2701	>	}
2702	>	else if ((s | WAITER) == 0) {
2703	>	if (U.compareAndSwapInt(this, LOCKSTATE, s, s | WAITER)) {
2704	>	waiting = true;
2705	>	waiter = Thread.currentThread();
2706	>	}
2707	>	}
2708	>	else if (waiting)
2709	>	LockSupport.park(this);
2710	>	}
2711	>	}
2712	>
2713	>	/**
2714	>	* Returns matching node or null if none. Tries to search
2715	>	* using tree compareisons from root, but continues linear
2716	>	* search when lock not available.
2717	>	*/
2718	>	final Node<K,V> find(int h, Object k) {
2719	>	if (k != null) {
2720	>	for (Node<K,V> e = first; e != null; e = e.next) {
2721	>	int s; K ek;
2722	>	if (((s = lockState) & (WAITER|WRITER)) != 0) {
2723	>	if (e.hash == h &&
2724	>	((ek = e.key) == k || (ek != null && k.equals(ek))))
2725	>	return e;
2726	>	}
2727	>	else if (U.compareAndSwapInt(this, LOCKSTATE, s,
2728	>	s + READER)) {
2729	>	TreeNode<K,V> r, p;
2730	>	try {
2731	>	p = ((r = root) == null ? null :
2732	>	r.findTreeNode(h, k, null));
2733	>	} finally {
2734	>	Thread w;
2735	>	if (U.getAndAddInt(this, LOCKSTATE, -READER) ==
2736	>	(READER|WAITER) && (w = waiter) != null)
2737	>	LockSupport.unpark(w);
2738	>	}
2739	>	return p;
2740	>	}
2741	>	}
2742	>	}
2743	>	return null;
2744	>	}
2745	>
2746	>	/**
2747	>	* Finds or adds a node.
2748	>	* @return null if added
2749	>	*/
2750	>	final TreeNode<K,V> putTreeVal(int h, K k, V v) {
2751	>	TreeNode<K,V> p;
2752	>	if ((p = root) == null) {
2753	>	first = root = new TreeNode<K,V>(h, k, v, null, null);
2754	>	return null;
2755	>	}
2756	>	Class<?> kc = null;
2757		for (;;) {
2758	<	if (nextTableIndex >= 0) {
2759	<	if ((nextEntry = entryAt(currentTable,
2760	<	nextTableIndex--)) != null)
2761	<	break;
2758	>	int dir, ph; K pk; TreeNode<K,V> q, pr;
2759	>	if ((ph = p.hash) > h)
2760	>	dir = -1;
2761	>	else if (ph < h)
2762	>	dir = 1;
2763	>	else if ((pk = p.key) == k || (pk != null && k.equals(pk)))
2764	>	return p;
2765	>	else if ((kc == null &&
2766	>	(kc = comparableClassFor(k)) == null) ||
2767	>	(dir = compareComparables(kc, k, pk)) == 0) {
2768	>	if (p.left == null)
2769	>	dir = 1;
2770	>	else if ((pr = p.right) == null ||
2771	>	(q = pr.findTreeNode(h, k, kc)) == null)
2772	>	dir = -1;
2773	>	else
2774	>	return q;
2775		}
2776	<	else if (nextSegmentIndex >= 0) {
2777	<	Segment<K,V> seg = segmentAt(segments, nextSegmentIndex--);
2778	<	if (seg != null && (currentTable = seg.table) != null)
2779	<	nextTableIndex = currentTable.length - 1;
2776	>	TreeNode<K,V> xp = p;
2777	>	if ((p = (dir < 0) ? p.left : p.right) == null) {
2778	>	TreeNode<K,V> x, f = first;
2779	>	first = x = new TreeNode<K,V>(h, k, v, f, xp);
2780	>	if (f != null)
2781	>	f.prev = x;
2782	>	if (dir < 0)
2783	>	xp.left = x;
2784	>	else
2785	>	xp.right = x;
2786	>	if (!xp.red)
2787	>	x.red = true;
2788	>	else {
2789	>	lockRoot();
2790	>	try {
2791	>	root = balanceInsertion(root, x);
2792	>	} finally {
2793	>	unlockRoot();
2794	>	}
2795	>	}
2796	>	//                    assert checkInvariants(root);
2797	>	return null;
2798	>	}
2799	>	}
2800	>	}
2801	>
2802	>	/**
2803	>	* Removes the given node, that must be present before this
2804	>	* call.  This is messier than typical red-black deletion code
2805	>	* because we cannot swap the contents of an interior node
2806	>	* with a leaf successor that is pinned by "next" pointers
2807	>	* that are accessible independently of lock. So instead we
2808	>	* swap the tree linkages.
2809	>	*
2810	>	* @return true if now too small so should be untreeified.
2811	>	*/
2812	>	final boolean removeTreeNode(TreeNode<K,V> p) {
2813	>	TreeNode<K,V> next = (TreeNode<K,V>)p.next;
2814	>	TreeNode<K,V> pred = p.prev;  // unlink traversal pointers
2815	>	TreeNode<K,V> r, rl;
2816	>	if (pred == null)
2817	>	first = next;
2818	>	else
2819	>	pred.next = next;
2820	>	if (next != null)
2821	>	next.prev = pred;
2822	>	if (first == null) {
2823	>	root = null;
2824	>	return true;
2825	>	}
2826	>	if ((r = root) == null || r.right == null ||
2827	>	(rl = r.left) == null || rl.left == null)
2828	>	return true;
2829	>	lockRoot();
2830	>	try {
2831	>	TreeNode<K,V> replacement;
2832	>	TreeNode<K,V> pl = p.left;
2833	>	TreeNode<K,V> pr = p.right;
2834	>	if (pl != null && pr != null) {
2835	>	TreeNode<K,V> s = pr, sl;
2836	>	while ((sl = s.left) != null) // find successor
2837	>	s = sl;
2838	>	boolean c = s.red; s.red = p.red; p.red = c; // swap colors
2839	>	TreeNode<K,V> sr = s.right;
2840	>	TreeNode<K,V> pp = p.parent;
2841	>	if (s == pr) { // p was s's direct parent
2842	>	p.parent = s;
2843	>	s.right = p;
2844	>	}
2845	>	else {
2846	>	TreeNode<K,V> sp = s.parent;
2847	>	if ((p.parent = sp) != null) {
2848	>	if (s == sp.left)
2849	>	sp.left = p;
2850	>	else
2851	>	sp.right = p;
2852	>	}
2853	>	if ((s.right = pr) != null)
2854	>	pr.parent = s;
2855	>	}
2856	>	p.left = null;
2857	>	if ((p.right = sr) != null)
2858	>	sr.parent = p;
2859	>	if ((s.left = pl) != null)
2860	>	pl.parent = s;
2861	>	if ((s.parent = pp) == null)
2862	>	r = s;
2863	>	else if (p == pp.left)
2864	>	pp.left = s;
2865	>	else
2866	>	pp.right = s;
2867	>	if (sr != null)
2868	>	replacement = sr;
2869	>	else
2870	>	replacement = p;
2871		}
2872	+	else if (pl != null)
2873	+	replacement = pl;
2874	+	else if (pr != null)
2875	+	replacement = pr;
2876		else
2877	<	break;
2877	>	replacement = p;
2878	>	if (replacement != p) {
2879	>	TreeNode<K,V> pp = replacement.parent = p.parent;
2880	>	if (pp == null)
2881	>	r = replacement;
2882	>	else if (p == pp.left)
2883	>	pp.left = replacement;
2884	>	else
2885	>	pp.right = replacement;
2886	>	p.left = p.right = p.parent = null;
2887	>	}
2888	>
2889	>	root = (p.red) ? r : balanceDeletion(r, replacement);
2890	>
2891	>	if (p == replacement) {  // detach pointers
2892	>	TreeNode<K,V> pp;
2893	>	if ((pp = p.parent) != null) {
2894	>	if (p == pp.left)
2895	>	pp.left = null;
2896	>	else if (p == pp.right)
2897	>	pp.right = null;
2898	>	p.parent = null;
2899	>	}
2900	>	}
2901	>	} finally {
2902	>	unlockRoot();
2903		}
2904	+	//            assert checkInvariants(root);
2905	+	return false;
2906		}
2907
2908	<	final HashEntry<K,V> nextEntry() {
2909	<	HashEntry<K,V> e = nextEntry;
2910	<	if (e == null)
2911	<	throw new NoSuchElementException();
2912	<	lastReturned = e; // cannot assign until after null check
2913	<	if ((nextEntry = e.next) == null)
2914	<	advance();
2915	<	return e;
2908	>	/* ------------------------------------------------------------ */
2909	>	// Red-black tree methods, all adapted from CLR
2910	>
2911	>	static <K,V> TreeNode<K,V> rotateLeft(TreeNode<K,V> root,
2912	>	TreeNode<K,V> p) {
2913	>	if (p != null) {
2914	>	TreeNode<K,V> r = p.right, pp, rl;
2915	>	if ((rl = p.right = r.left) != null)
2916	>	rl.parent = p;
2917	>	if ((pp = r.parent = p.parent) == null)
2918	>	(root = r).red = false;
2919	>	else if (pp.left == p)
2920	>	pp.left = r;
2921	>	else
2922	>	pp.right = r;
2923	>	r.left = p;
2924	>	p.parent = r;
2925	>	}
2926	>	return root;
2927		}
2928
2929	<	public final boolean hasNext() { return nextEntry != null; }
2930	<	public final boolean hasMoreElements() { return nextEntry != null; }
2929	>	static <K,V> TreeNode<K,V> rotateRight(TreeNode<K,V> root,
2930	>	TreeNode<K,V> p) {
2931	>	if (p != null) {
2932	>	TreeNode<K,V> l = p.left, pp, lr;
2933	>	if ((lr = p.left = l.right) != null)
2934	>	lr.parent = p;
2935	>	if ((pp = l.parent = p.parent) == null)
2936	>	(root = l).red = false;
2937	>	else if (pp.right == p)
2938	>	pp.right = l;
2939	>	else
2940	>	pp.left = l;
2941	>	l.right = p;
2942	>	p.parent = l;
2943	>	}
2944	>	return root;
2945	>	}
2946	>
2947	>	static <K,V> TreeNode<K,V> balanceInsertion(TreeNode<K,V> root,
2948	>	TreeNode<K,V> x) {
2949	>	x.red = true;
2950	>	for (TreeNode<K,V> xp, xpp, xppl, xppr;;) {
2951	>	if ((xp = x.parent) == null) {
2952	>	x.red = false;
2953	>	return x;
2954	>	}
2955	>	else if (!xp.red || (xpp = xp.parent) == null)
2956	>	return root;
2957	>	if (xp == (xppl = xpp.left)) {
2958	>	if ((xppr = xpp.right) != null && xppr.red) {
2959	>	xppr.red = false;
2960	>	xp.red = false;
2961	>	xpp.red = true;
2962	>	x = xpp;
2963	>	}
2964	>	else {
2965	>	if (x == xp.right) {
2966	>	root = rotateLeft(root, x = xp);
2967	>	xpp = (xp = x.parent) == null ? null : xp.parent;
2968	>	}
2969	>	if (xp != null) {
2970	>	xp.red = false;
2971	>	if (xpp != null) {
2972	>	xpp.red = true;
2973	>	root = rotateRight(root, xpp);
2974	>	}
2975	>	}
2976	>	}
2977	>	}
2978	>	else {
2979	>	if (xppl != null && xppl.red) {
2980	>	xppl.red = false;
2981	>	xp.red = false;
2982	>	xpp.red = true;
2983	>	x = xpp;
2984	>	}
2985	>	else {
2986	>	if (x == xp.left) {
2987	>	root = rotateRight(root, x = xp);
2988	>	xpp = (xp = x.parent) == null ? null : xp.parent;
2989	>	}
2990	>	if (xp != null) {
2991	>	xp.red = false;
2992	>	if (xpp != null) {
2993	>	xpp.red = true;
2994	>	root = rotateLeft(root, xpp);
2995	>	}
2996	>	}
2997	>	}
2998	>	}
2999	>	}
3000	>	}
3001	>
3002	>	static <K,V> TreeNode<K,V> balanceDeletion(TreeNode<K,V> root,
3003	>	TreeNode<K,V> x) {
3004	>	for (TreeNode<K,V> xp, xpl, xpr;;)  {
3005	>	if (x == null || x == root)
3006	>	return root;
3007	>	else if ((xp = x.parent) == null) {
3008	>	x.red = false;
3009	>	return x;
3010	>	}
3011	>	else if (x.red) {
3012	>	x.red = false;
3013	>	return root;
3014	>	}
3015	>	else if ((xpl = xp.left) == x) {
3016	>	if ((xpr = xp.right) != null && xpr.red) {
3017	>	xpr.red = false;
3018	>	xp.red = true;
3019	>	root = rotateLeft(root, xp);
3020	>	xpr = (xp = x.parent) == null ? null : xp.right;
3021	>	}
3022	>	if (xpr == null)
3023	>	x = xp;
3024	>	else {
3025	>	TreeNode<K,V> sl = xpr.left, sr = xpr.right;
3026	>	if ((sr == null || !sr.red) &&
3027	>	(sl == null || !sl.red)) {
3028	>	xpr.red = true;
3029	>	x = xp;
3030	>	}
3031	>	else {
3032	>	if (sr == null || !sr.red) {
3033	>	if (sl != null)
3034	>	sl.red = false;
3035	>	xpr.red = true;
3036	>	root = rotateRight(root, xpr);
3037	>	xpr = (xp = x.parent) == null ?
3038	>	null : xp.right;
3039	>	}
3040	>	if (xpr != null) {
3041	>	xpr.red = (xp == null) ? false : xp.red;
3042	>	if ((sr = xpr.right) != null)
3043	>	sr.red = false;
3044	>	}
3045	>	if (xp != null) {
3046	>	xp.red = false;
3047	>	root = rotateLeft(root, xp);
3048	>	}
3049	>	x = root;
3050	>	}
3051	>	}
3052	>	}
3053	>	else { // symmetric
3054	>	if (xpl != null && xpl.red) {
3055	>	xpl.red = false;
3056	>	xp.red = true;
3057	>	root = rotateRight(root, xp);
3058	>	xpl = (xp = x.parent) == null ? null : xp.left;
3059	>	}
3060	>	if (xpl == null)
3061	>	x = xp;
3062	>	else {
3063	>	TreeNode<K,V> sl = xpl.left, sr = xpl.right;
3064	>	if ((sl == null || !sl.red) &&
3065	>	(sr == null || !sr.red)) {
3066	>	xpl.red = true;
3067	>	x = xp;
3068	>	}
3069	>	else {
3070	>	if (sl == null || !sl.red) {
3071	>	if (sr != null)
3072	>	sr.red = false;
3073	>	xpl.red = true;
3074	>	root = rotateLeft(root, xpl);
3075	>	xpl = (xp = x.parent) == null ?
3076	>	null : xp.left;
3077	>	}
3078	>	if (xpl != null) {
3079	>	xpl.red = (xp == null) ? false : xp.red;
3080	>	if ((sl = xpl.left) != null)
3081	>	sl.red = false;
3082	>	}
3083	>	if (xp != null) {
3084	>	xp.red = false;
3085	>	root = rotateRight(root, xp);
3086	>	}
3087	>	x = root;
3088	>	}
3089	>	}
3090	>	}
3091	>	}
3092	>	}
3093	>
3094	>	/**
3095	>	* Recursive invariant check
3096	>	*/
3097	>	static <K,V> boolean checkTreeNode(TreeNode<K,V> t) {
3098	>	TreeNode<K,V> tp = t.parent, tl = t.left, tr = t.right,
3099	>	tb = t.prev, tn = (TreeNode<K,V>)t.next;
3100	>	if (tb != null && tb.next != t)
3101	>	return false;
3102	>	if (tn != null && tn.prev != t)
3103	>	return false;
3104	>	if (tp != null && t != tp.left && t != tp.right)
3105	>	return false;
3106	>	if (tl != null && (tl.parent != t || tl.hash > t.hash))
3107	>	return false;
3108	>	if (tr != null && (tr.parent != t || tr.hash < t.hash))
3109	>	return false;
3110	>	if (t.red && tl != null && tl.red && tr != null && tr.red)
3111	>	return false;
3112	>	if (tl != null && !checkTreeNode(tl))
3113	>	return false;
3114	>	if (tr != null && !checkTreeNode(tr))
3115	>	return false;
3116	>	return true;
3117	>	}
3118	>
3119	>	private static final sun.misc.Unsafe U;
3120	>	private static final long LOCKSTATE;
3121	>	static {
3122	>	try {
3123	>	U = sun.misc.Unsafe.getUnsafe();
3124	>	Class<?> k = TreeBin.class;
3125	>	LOCKSTATE = U.objectFieldOffset
3126	>	(k.getDeclaredField("lockState"));
3127	>	} catch (Exception e) {
3128	>	throw new Error(e);
3129	>	}
3130	>	}
3131	>	}
3132	>
3133	>	/* ----------------Table Traversal -------------- */
3134	>
3135	>	/**
3136	>	* Encapsulates traversal for methods such as containsValue; also
3137	>	* serves as a base class for other iterators and spliterators.
3138	>	*
3139	>	* Method advance visits once each still-valid node that was
3140	>	* reachable upon iterator construction. It might miss some that
3141	>	* were added to a bin after the bin was visited, which is OK wrt
3142	>	* consistency guarantees. Maintaining this property in the face
3143	>	* of possible ongoing resizes requires a fair amount of
3144	>	* bookkeeping state that is difficult to optimize away amidst
3145	>	* volatile accesses.  Even so, traversal maintains reasonable
3146	>	* throughput.
3147	>	*
3148	>	* Normally, iteration proceeds bin-by-bin traversing lists.
3149	>	* However, if the table has been resized, then all future steps
3150	>	* must traverse both the bin at the current index as well as at
3151	>	* (index + baseSize); and so on for further resizings. To
3152	>	* paranoically cope with potential sharing by users of iterators
3153	>	* across threads, iteration terminates if a bounds checks fails
3154	>	* for a table read.
3155	>	*/
3156	>	static class Traverser<K,V> {
3157	>	Node<K,V>[] tab;        // current table; updated if resized
3158	>	Node<K,V> next;         // the next entry to use
3159	>	int index;              // index of bin to use next
3160	>	int baseIndex;          // current index of initial table
3161	>	int baseLimit;          // index bound for initial table
3162	>	final int baseSize;     // initial table size
3163	>
3164	>	Traverser(Node<K,V>[] tab, int size, int index, int limit) {
3165	>	this.tab = tab;
3166	>	this.baseSize = size;
3167	>	this.baseIndex = this.index = index;
3168	>	this.baseLimit = limit;
3169	>	this.next = null;
3170	>	}
3171	>
3172	>	/**
3173	>	* Advances if possible, returning next valid node, or null if none.
3174	>	*/
3175	>	final Node<K,V> advance() {
3176	>	Node<K,V> e;
3177	>	if ((e = next) != null)
3178	>	e = e.next;
3179	>	for (;;) {
3180	>	Node<K,V>[] t; int i, n; K ek;  // must use locals in checks
3181	>	if (e != null)
3182	>	return next = e;
3183	>	if (baseIndex >= baseLimit || (t = tab) == null ||
3184	>	(n = t.length) <= (i = index) || i < 0)
3185	>	return next = null;
3186	>	if ((e = tabAt(t, index)) != null && e.key == null) {
3187	>	if (e instanceof ForwardingNode) {
3188	>	tab = ((ForwardingNode<K,V>)e).nextTable;
3189	>	e = null;
3190	>	continue;
3191	>	}
3192	>	else if (e instanceof TreeBin)
3193	>	e = ((TreeBin<K,V>)e).first;
3194	>	else
3195	>	e = null;
3196	>	}
3197	>	if ((index += baseSize) >= n)
3198	>	index = ++baseIndex;    // visit upper slots if present
3199	>	}
3200	>	}
3201	>	}
3202	>
3203	>	/**
3204	>	* Base of key, value, and entry Iterators. Adds fields to
3205	>	* Traverser to support iterator.remove
3206	>	*/
3207	>	static class BaseIterator<K,V> extends Traverser<K,V> {
3208	>	final ConcurrentHashMap<K,V> map;
3209	>	Node<K,V> lastReturned;
3210	>	BaseIterator(Node<K,V>[] tab, int size, int index, int limit,
3211	>	ConcurrentHashMap<K,V> map) {
3212	>	super(tab, size, index, limit);
3213	>	this.map = map;
3214	>	advance();
3215	>	}
3216	>
3217	>	public final boolean hasNext() { return next != null; }
3218	>	public final boolean hasMoreElements() { return next != null; }
3219
3220		public final void remove() {
3221	<	if (lastReturned == null)
3221	>	Node<K,V> p;
3222	>	if ((p = lastReturned) == null)
3223		throw new IllegalStateException();
1271	–	ConcurrentHashMap.this.remove(lastReturned.key);
3224		lastReturned = null;
3225	+	map.replaceNode(p.key, null, null);
3226		}
3227		}
3228
3229	<	final class KeyIterator
3230	<	extends HashIterator
3231	<	implements Iterator<K>, Enumeration<K>
3232	<	{
3233	<	public final K next()        { return super.nextEntry().key; }
3234	<	public final K nextElement() { return super.nextEntry().key; }
3235	<	}
3236	<
3237	<	final class ValueIterator
3238	<	extends HashIterator
3239	<	implements Iterator<V>, Enumeration<V>
3240	<	{
3241	<	public final V next()        { return super.nextEntry().value; }
3242	<	public final V nextElement() { return super.nextEntry().value; }
3229	>	static final class KeyIterator<K,V> extends BaseIterator<K,V>
3230	>	implements Iterator<K>, Enumeration<K> {
3231	>	KeyIterator(Node<K,V>[] tab, int index, int size, int limit,
3232	>	ConcurrentHashMap<K,V> map) {
3233	>	super(tab, index, size, limit, map);
3234	>	}
3235	>
3236	>	public final K next() {
3237	>	Node<K,V> p;
3238	>	if ((p = next) == null)
3239	>	throw new NoSuchElementException();
3240	>	K k = p.key;
3241	>	lastReturned = p;
3242	>	advance();
3243	>	return k;
3244	>	}
3245	>
3246	>	public final K nextElement() { return next(); }
3247	>	}
3248	>
3249	>	static final class ValueIterator<K,V> extends BaseIterator<K,V>
3250	>	implements Iterator<V>, Enumeration<V> {
3251	>	ValueIterator(Node<K,V>[] tab, int index, int size, int limit,
3252	>	ConcurrentHashMap<K,V> map) {
3253	>	super(tab, index, size, limit, map);
3254	>	}
3255	>
3256	>	public final V next() {
3257	>	Node<K,V> p;
3258	>	if ((p = next) == null)
3259	>	throw new NoSuchElementException();
3260	>	V v = p.val;
3261	>	lastReturned = p;
3262	>	advance();
3263	>	return v;
3264	>	}
3265	>
3266	>	public final V nextElement() { return next(); }
3267	>	}
3268	>
3269	>	static final class EntryIterator<K,V> extends BaseIterator<K,V>
3270	>	implements Iterator<Map.Entry<K,V>> {
3271	>	EntryIterator(Node<K,V>[] tab, int index, int size, int limit,
3272	>	ConcurrentHashMap<K,V> map) {
3273	>	super(tab, index, size, limit, map);
3274	>	}
3275	>
3276	>	public final Map.Entry<K,V> next() {
3277	>	Node<K,V> p;
3278	>	if ((p = next) == null)
3279	>	throw new NoSuchElementException();
3280	>	K k = p.key;
3281	>	V v = p.val;
3282	>	lastReturned = p;
3283	>	advance();
3284	>	return new MapEntry<K,V>(k, v, map);
3285	>	}
3286		}
3287
3288		/**
3289	<	* Custom Entry class used by EntryIterator.next(), that relays
3290	<	* setValue changes to the underlying map.
3291	<	*/
3292	<	final class WriteThroughEntry
3293	<	extends AbstractMap.SimpleEntry<K,V>
3294	<	{
3295	<	WriteThroughEntry(K k, V v) {
3296	<	super(k,v);
3289	>	* Exported Entry for EntryIterator
3290	>	*/
3291	>	static final class MapEntry<K,V> implements Map.Entry<K,V> {
3292	>	final K key; // non-null
3293	>	V val;       // non-null
3294	>	final ConcurrentHashMap<K,V> map;
3295	>	MapEntry(K key, V val, ConcurrentHashMap<K,V> map) {
3296	>	this.key = key;
3297	>	this.val = val;
3298	>	this.map = map;
3299	>	}
3300	>	public K getKey()        { return key; }
3301	>	public V getValue()      { return val; }
3302	>	public int hashCode()    { return key.hashCode() ^ val.hashCode(); }
3303	>	public String toString() { return key + "=" + val; }
3304	>
3305	>	public boolean equals(Object o) {
3306	>	Object k, v; Map.Entry<?,?> e;
3307	>	return ((o instanceof Map.Entry) &&
3308	>	(k = (e = (Map.Entry<?,?>)o).getKey()) != null &&
3309	>	(v = e.getValue()) != null &&
3310	>	(k == key || k.equals(key)) &&
3311	>	(v == val || v.equals(val)));
3312		}
3313
3314		/**
3315		* Sets our entry's value and writes through to the map. The
3316	<	* value to return is somewhat arbitrary here. Since a
3317	<	* WriteThroughEntry does not necessarily track asynchronous
3318	<	* changes, the most recent "previous" value could be
3319	<	* different from what we return (or could even have been
3320	<	* removed in which case the put will re-establish). We do not
1310	<	* and cannot guarantee more.
3316	>	* value to return is somewhat arbitrary here. Since we do not
3317	>	* necessarily track asynchronous changes, the most recent
3318	>	* "previous" value could be different from what we return (or
3319	>	* could even have been removed, in which case the put will
3320	>	* re-establish). We do not and cannot guarantee more.
3321		*/
3322		public V setValue(V value) {
3323		if (value == null) throw new NullPointerException();
3324	<	V v = super.setValue(value);
3325	<	ConcurrentHashMap.this.put(getKey(), value);
3324	>	V v = val;
3325	>	val = value;
3326	>	map.put(key, value);
3327		return v;
3328		}
3329		}
3330
3331	<	final class EntryIterator
3332	<	extends HashIterator
3333	<	implements Iterator<Entry<K,V>>
3334	<	{
3335	<	public Map.Entry<K,V> next() {
3336	<	HashEntry<K,V> e = super.nextEntry();
3337	<	return new WriteThroughEntry(e.key, e.value);
3331	>	static final class KeySpliterator<K,V> extends Traverser<K,V>
3332	>	implements Spliterator<K> {
3333	>	long est;               // size estimate
3334	>	KeySpliterator(Node<K,V>[] tab, int size, int index, int limit,
3335	>	long est) {
3336	>	super(tab, size, index, limit);
3337	>	this.est = est;
3338	>	}
3339	>
3340	>	public Spliterator<K> trySplit() {
3341	>	int i, f, h;
3342	>	return (h = ((i = baseIndex) + (f = baseLimit)) >>> 1) <= i ? null :
3343	>	new KeySpliterator<K,V>(tab, baseSize, baseLimit = h,
3344	>	f, est >>>= 1);
3345	>	}
3346	>
3347	>	public void forEachRemaining(Consumer<? super K> action) {
3348	>	if (action == null) throw new NullPointerException();
3349	>	for (Node<K,V> p; (p = advance()) != null;)
3350	>	action.accept(p.key);
3351	>	}
3352	>
3353	>	public boolean tryAdvance(Consumer<? super K> action) {
3354	>	if (action == null) throw new NullPointerException();
3355	>	Node<K,V> p;
3356	>	if ((p = advance()) == null)
3357	>	return false;
3358	>	action.accept(p.key);
3359	>	return true;
3360	>	}
3361	>
3362	>	public long estimateSize() { return est; }
3363	>
3364	>	public int characteristics() {
3365	>	return Spliterator.DISTINCT | Spliterator.CONCURRENT |
3366	>	Spliterator.NONNULL;
3367		}
3368		}
3369
3370	<	final class KeySet extends AbstractSet<K> {
3371	<	public Iterator<K> iterator() {
3372	<	return new KeyIterator();
3370	>	static final class ValueSpliterator<K,V> extends Traverser<K,V>
3371	>	implements Spliterator<V> {
3372	>	long est;               // size estimate
3373	>	ValueSpliterator(Node<K,V>[] tab, int size, int index, int limit,
3374	>	long est) {
3375	>	super(tab, size, index, limit);
3376	>	this.est = est;
3377		}
3378	<	public int size() {
3379	<	return ConcurrentHashMap.this.size();
3378	>
3379	>	public Spliterator<V> trySplit() {
3380	>	int i, f, h;
3381	>	return (h = ((i = baseIndex) + (f = baseLimit)) >>> 1) <= i ? null :
3382	>	new ValueSpliterator<K,V>(tab, baseSize, baseLimit = h,
3383	>	f, est >>>= 1);
3384		}
3385	<	public boolean isEmpty() {
3386	<	return ConcurrentHashMap.this.isEmpty();
3385	>
3386	>	public void forEachRemaining(Consumer<? super V> action) {
3387	>	if (action == null) throw new NullPointerException();
3388	>	for (Node<K,V> p; (p = advance()) != null;)
3389	>	action.accept(p.val);
3390		}
3391	<	public boolean contains(Object o) {
3392	<	return ConcurrentHashMap.this.containsKey(o);
3391	>
3392	>	public boolean tryAdvance(Consumer<? super V> action) {
3393	>	if (action == null) throw new NullPointerException();
3394	>	Node<K,V> p;
3395	>	if ((p = advance()) == null)
3396	>	return false;
3397	>	action.accept(p.val);
3398	>	return true;
3399		}
3400	<	public boolean remove(Object o) {
3401	<	return ConcurrentHashMap.this.remove(o) != null;
3400	>
3401	>	public long estimateSize() { return est; }
3402	>
3403	>	public int characteristics() {
3404	>	return Spliterator.CONCURRENT | Spliterator.NONNULL;
3405	>	}
3406	>	}
3407	>
3408	>	static final class EntrySpliterator<K,V> extends Traverser<K,V>
3409	>	implements Spliterator<Map.Entry<K,V>> {
3410	>	final ConcurrentHashMap<K,V> map; // To export MapEntry
3411	>	long est;               // size estimate
3412	>	EntrySpliterator(Node<K,V>[] tab, int size, int index, int limit,
3413	>	long est, ConcurrentHashMap<K,V> map) {
3414	>	super(tab, size, index, limit);
3415	>	this.map = map;
3416	>	this.est = est;
3417	>	}
3418	>
3419	>	public Spliterator<Map.Entry<K,V>> trySplit() {
3420	>	int i, f, h;
3421	>	return (h = ((i = baseIndex) + (f = baseLimit)) >>> 1) <= i ? null :
3422	>	new EntrySpliterator<K,V>(tab, baseSize, baseLimit = h,
3423	>	f, est >>>= 1, map);
3424		}
3425	<	public void clear() {
3426	<	ConcurrentHashMap.this.clear();
3425	>
3426	>	public void forEachRemaining(Consumer<? super Map.Entry<K,V>> action) {
3427	>	if (action == null) throw new NullPointerException();
3428	>	for (Node<K,V> p; (p = advance()) != null; )
3429	>	action.accept(new MapEntry<K,V>(p.key, p.val, map));
3430	>	}
3431	>
3432	>	public boolean tryAdvance(Consumer<? super Map.Entry<K,V>> action) {
3433	>	if (action == null) throw new NullPointerException();
3434	>	Node<K,V> p;
3435	>	if ((p = advance()) == null)
3436	>	return false;
3437	>	action.accept(new MapEntry<K,V>(p.key, p.val, map));
3438	>	return true;
3439		}
3440	+
3441	+	public long estimateSize() { return est; }
3442	+
3443	+	public int characteristics() {
3444	+	return Spliterator.DISTINCT | Spliterator.CONCURRENT |
3445	+	Spliterator.NONNULL;
3446	+	}
3447	+	}
3448	+
3449	+	// Parallel bulk operations
3450	+
3451	+	/**
3452	+	* Computes initial batch value for bulk tasks. The returned value
3453	+	* is approximately exp2 of the number of times (minus one) to
3454	+	* split task by two before executing leaf action. This value is
3455	+	* faster to compute and more convenient to use as a guide to
3456	+	* splitting than is the depth, since it is used while dividing by
3457	+	* two anyway.
3458	+	*/
3459	+	final int batchFor(long b) {
3460	+	long n;
3461	+	if (b == Long.MAX_VALUE || (n = sumCount()) <= 1L || n < b)
3462	+	return 0;
3463	+	int sp = ForkJoinPool.getCommonPoolParallelism() << 2; // slack of 4
3464	+	return (b <= 0L || (n /= b) >= sp) ? sp : (int)n;
3465	+	}
3466	+
3467	+	/**
3468	+	* Performs the given action for each (key, value).
3469	+	*
3470	+	* @param parallelismThreshold the (estimated) number of elements
3471	+	* needed for this operation to be executed in parallel
3472	+	* @param action the action
3473	+	* @since 1.8
3474	+	*/
3475	+	public void forEach(long parallelismThreshold,
3476	+	BiConsumer<? super K,? super V> action) {
3477	+	if (action == null) throw new NullPointerException();
3478	+	new ForEachMappingTask<K,V>
3479	+	(null, batchFor(parallelismThreshold), 0, 0, table,
3480	+	action).invoke();
3481	+	}
3482	+
3483	+	/**
3484	+	* Performs the given action for each non-null transformation
3485	+	* of each (key, value).
3486	+	*
3487	+	* @param parallelismThreshold the (estimated) number of elements
3488	+	* needed for this operation to be executed in parallel
3489	+	* @param transformer a function returning the transformation
3490	+	* for an element, or null if there is no transformation (in
3491	+	* which case the action is not applied)
3492	+	* @param action the action
3493	+	* @since 1.8
3494	+	*/
3495	+	public <U> void forEach(long parallelismThreshold,
3496	+	BiFunction<? super K, ? super V, ? extends U> transformer,
3497	+	Consumer<? super U> action) {
3498	+	if (transformer == null || action == null)
3499	+	throw new NullPointerException();
3500	+	new ForEachTransformedMappingTask<K,V,U>
3501	+	(null, batchFor(parallelismThreshold), 0, 0, table,
3502	+	transformer, action).invoke();
3503	+	}
3504	+
3505	+	/**
3506	+	* Returns a non-null result from applying the given search
3507	+	* function on each (key, value), or null if none.  Upon
3508	+	* success, further element processing is suppressed and the
3509	+	* results of any other parallel invocations of the search
3510	+	* function are ignored.
3511	+	*
3512	+	* @param parallelismThreshold the (estimated) number of elements
3513	+	* needed for this operation to be executed in parallel
3514	+	* @param searchFunction a function returning a non-null
3515	+	* result on success, else null
3516	+	* @return a non-null result from applying the given search
3517	+	* function on each (key, value), or null if none
3518	+	* @since 1.8
3519	+	*/
3520	+	public <U> U search(long parallelismThreshold,
3521	+	BiFunction<? super K, ? super V, ? extends U> searchFunction) {
3522	+	if (searchFunction == null) throw new NullPointerException();
3523	+	return new SearchMappingsTask<K,V,U>
3524	+	(null, batchFor(parallelismThreshold), 0, 0, table,
3525	+	searchFunction, new AtomicReference<U>()).invoke();
3526	+	}
3527	+
3528	+	/**
3529	+	* Returns the result of accumulating the given transformation
3530	+	* of all (key, value) pairs using the given reducer to
3531	+	* combine values, or null if none.
3532	+	*
3533	+	* @param parallelismThreshold the (estimated) number of elements
3534	+	* needed for this operation to be executed in parallel
3535	+	* @param transformer a function returning the transformation
3536	+	* for an element, or null if there is no transformation (in
3537	+	* which case it is not combined)
3538	+	* @param reducer a commutative associative combining function
3539	+	* @return the result of accumulating the given transformation
3540	+	* of all (key, value) pairs
3541	+	* @since 1.8
3542	+	*/
3543	+	public <U> U reduce(long parallelismThreshold,
3544	+	BiFunction<? super K, ? super V, ? extends U> transformer,
3545	+	BiFunction<? super U, ? super U, ? extends U> reducer) {
3546	+	if (transformer == null || reducer == null)
3547	+	throw new NullPointerException();
3548	+	return new MapReduceMappingsTask<K,V,U>
3549	+	(null, batchFor(parallelismThreshold), 0, 0, table,
3550	+	null, transformer, reducer).invoke();
3551	+	}
3552	+
3553	+	/**
3554	+	* Returns the result of accumulating the given transformation
3555	+	* of all (key, value) pairs using the given reducer to
3556	+	* combine values, and the given basis as an identity value.
3557	+	*
3558	+	* @param parallelismThreshold the (estimated) number of elements
3559	+	* needed for this operation to be executed in parallel
3560	+	* @param transformer a function returning the transformation
3561	+	* for an element
3562	+	* @param basis the identity (initial default value) for the reduction
3563	+	* @param reducer a commutative associative combining function
3564	+	* @return the result of accumulating the given transformation
3565	+	* of all (key, value) pairs
3566	+	* @since 1.8
3567	+	*/
3568	+	public double reduceToDoubleIn(long parallelismThreshold,
3569	+	ToDoubleBiFunction<? super K, ? super V> transformer,
3570	+	double basis,
3571	+	DoubleBinaryOperator reducer) {
3572	+	if (transformer == null || reducer == null)
3573	+	throw new NullPointerException();
3574	+	return new MapReduceMappingsToDoubleTask<K,V>
3575	+	(null, batchFor(parallelismThreshold), 0, 0, table,
3576	+	null, transformer, basis, reducer).invoke();
3577	+	}
3578	+
3579	+	/**
3580	+	* Returns the result of accumulating the given transformation
3581	+	* of all (key, value) pairs using the given reducer to
3582	+	* combine values, and the given basis as an identity value.
3583	+	*
3584	+	* @param parallelismThreshold the (estimated) number of elements
3585	+	* needed for this operation to be executed in parallel
3586	+	* @param transformer a function returning the transformation
3587	+	* for an element
3588	+	* @param basis the identity (initial default value) for the reduction
3589	+	* @param reducer a commutative associative combining function
3590	+	* @return the result of accumulating the given transformation
3591	+	* of all (key, value) pairs
3592	+	* @since 1.8
3593	+	*/
3594	+	public long reduceToLong(long parallelismThreshold,
3595	+	ToLongBiFunction<? super K, ? super V> transformer,
3596	+	long basis,
3597	+	LongBinaryOperator reducer) {
3598	+	if (transformer == null || reducer == null)
3599	+	throw new NullPointerException();
3600	+	return new MapReduceMappingsToLongTask<K,V>
3601	+	(null, batchFor(parallelismThreshold), 0, 0, table,
3602	+	null, transformer, basis, reducer).invoke();
3603	+	}
3604	+
3605	+	/**
3606	+	* Returns the result of accumulating the given transformation
3607	+	* of all (key, value) pairs using the given reducer to
3608	+	* combine values, and the given basis as an identity value.
3609	+	*
3610	+	* @param parallelismThreshold the (estimated) number of elements
3611	+	* needed for this operation to be executed in parallel
3612	+	* @param transformer a function returning the transformation
3613	+	* for an element
3614	+	* @param basis the identity (initial default value) for the reduction
3615	+	* @param reducer a commutative associative combining function
3616	+	* @return the result of accumulating the given transformation
3617	+	* of all (key, value) pairs
3618	+	* @since 1.8
3619	+	*/
3620	+	public int reduceToInt(long parallelismThreshold,
3621	+	ToIntBiFunction<? super K, ? super V> transformer,
3622	+	int basis,
3623	+	IntBinaryOperator reducer) {
3624	+	if (transformer == null || reducer == null)
3625	+	throw new NullPointerException();
3626	+	return new MapReduceMappingsToIntTask<K,V>
3627	+	(null, batchFor(parallelismThreshold), 0, 0, table,
3628	+	null, transformer, basis, reducer).invoke();
3629	+	}
3630	+
3631	+	/**
3632	+	* Performs the given action for each key.
3633	+	*
3634	+	* @param parallelismThreshold the (estimated) number of elements
3635	+	* needed for this operation to be executed in parallel
3636	+	* @param action the action
3637	+	* @since 1.8
3638	+	*/
3639	+	public void forEachKey(long parallelismThreshold,
3640	+	Consumer<? super K> action) {
3641	+	if (action == null) throw new NullPointerException();
3642	+	new ForEachKeyTask<K,V>
3643	+	(null, batchFor(parallelismThreshold), 0, 0, table,
3644	+	action).invoke();
3645	+	}
3646	+
3647	+	/**
3648	+	* Performs the given action for each non-null transformation
3649	+	* of each key.
3650	+	*
3651	+	* @param parallelismThreshold the (estimated) number of elements
3652	+	* needed for this operation to be executed in parallel
3653	+	* @param transformer a function returning the transformation
3654	+	* for an element, or null if there is no transformation (in
3655	+	* which case the action is not applied)
3656	+	* @param action the action
3657	+	* @since 1.8
3658	+	*/
3659	+	public <U> void forEachKey(long parallelismThreshold,
3660	+	Function<? super K, ? extends U> transformer,
3661	+	Consumer<? super U> action) {
3662	+	if (transformer == null || action == null)
3663	+	throw new NullPointerException();
3664	+	new ForEachTransformedKeyTask<K,V,U>
3665	+	(null, batchFor(parallelismThreshold), 0, 0, table,
3666	+	transformer, action).invoke();
3667	+	}
3668	+
3669	+	/**
3670	+	* Returns a non-null result from applying the given search
3671	+	* function on each key, or null if none. Upon success,
3672	+	* further element processing is suppressed and the results of
3673	+	* any other parallel invocations of the search function are
3674	+	* ignored.
3675	+	*
3676	+	* @param parallelismThreshold the (estimated) number of elements
3677	+	* needed for this operation to be executed in parallel
3678	+	* @param searchFunction a function returning a non-null
3679	+	* result on success, else null
3680	+	* @return a non-null result from applying the given search
3681	+	* function on each key, or null if none
3682	+	* @since 1.8
3683	+	*/
3684	+	public <U> U searchKeys(long parallelismThreshold,
3685	+	Function<? super K, ? extends U> searchFunction) {
3686	+	if (searchFunction == null) throw new NullPointerException();
3687	+	return new SearchKeysTask<K,V,U>
3688	+	(null, batchFor(parallelismThreshold), 0, 0, table,
3689	+	searchFunction, new AtomicReference<U>()).invoke();
3690	+	}
3691	+
3692	+	/**
3693	+	* Returns the result of accumulating all keys using the given
3694	+	* reducer to combine values, or null if none.
3695	+	*
3696	+	* @param parallelismThreshold the (estimated) number of elements
3697	+	* needed for this operation to be executed in parallel
3698	+	* @param reducer a commutative associative combining function
3699	+	* @return the result of accumulating all keys using the given
3700	+	* reducer to combine values, or null if none
3701	+	* @since 1.8
3702	+	*/
3703	+	public K reduceKeys(long parallelismThreshold,
3704	+	BiFunction<? super K, ? super K, ? extends K> reducer) {
3705	+	if (reducer == null) throw new NullPointerException();
3706	+	return new ReduceKeysTask<K,V>
3707	+	(null, batchFor(parallelismThreshold), 0, 0, table,
3708	+	null, reducer).invoke();
3709	+	}
3710	+
3711	+	/**
3712	+	* Returns the result of accumulating the given transformation
3713	+	* of all keys using the given reducer to combine values, or
3714	+	* null if none.
3715	+	*
3716	+	* @param parallelismThreshold the (estimated) number of elements
3717	+	* needed for this operation to be executed in parallel
3718	+	* @param transformer a function returning the transformation
3719	+	* for an element, or null if there is no transformation (in
3720	+	* which case it is not combined)
3721	+	* @param reducer a commutative associative combining function
3722	+	* @return the result of accumulating the given transformation
3723	+	* of all keys
3724	+	* @since 1.8
3725	+	*/
3726	+	public <U> U reduceKeys(long parallelismThreshold,
3727	+	Function<? super K, ? extends U> transformer,
3728	+	BiFunction<? super U, ? super U, ? extends U> reducer) {
3729	+	if (transformer == null || reducer == null)
3730	+	throw new NullPointerException();
3731	+	return new MapReduceKeysTask<K,V,U>
3732	+	(null, batchFor(parallelismThreshold), 0, 0, table,
3733	+	null, transformer, reducer).invoke();
3734	+	}
3735	+
3736	+	/**
3737	+	* Returns the result of accumulating the given transformation
3738	+	* of all keys using the given reducer to combine values, and
3739	+	* the given basis as an identity value.
3740	+	*
3741	+	* @param parallelismThreshold the (estimated) number of elements
3742	+	* needed for this operation to be executed in parallel
3743	+	* @param transformer a function returning the transformation
3744	+	* for an element
3745	+	* @param basis the identity (initial default value) for the reduction
3746	+	* @param reducer a commutative associative combining function
3747	+	* @return the result of accumulating the given transformation
3748	+	* of all keys
3749	+	* @since 1.8
3750	+	*/
3751	+	public double reduceKeysToDouble(long parallelismThreshold,
3752	+	ToDoubleFunction<? super K> transformer,
3753	+	double basis,
3754	+	DoubleBinaryOperator reducer) {
3755	+	if (transformer == null || reducer == null)
3756	+	throw new NullPointerException();
3757	+	return new MapReduceKeysToDoubleTask<K,V>
3758	+	(null, batchFor(parallelismThreshold), 0, 0, table,
3759	+	null, transformer, basis, reducer).invoke();
3760	+	}
3761	+
3762	+	/**
3763	+	* Returns the result of accumulating the given transformation
3764	+	* of all keys using the given reducer to combine values, and
3765	+	* the given basis as an identity value.
3766	+	*
3767	+	* @param parallelismThreshold the (estimated) number of elements
3768	+	* needed for this operation to be executed in parallel
3769	+	* @param transformer a function returning the transformation
3770	+	* for an element
3771	+	* @param basis the identity (initial default value) for the reduction
3772	+	* @param reducer a commutative associative combining function
3773	+	* @return the result of accumulating the given transformation
3774	+	* of all keys
3775	+	* @since 1.8
3776	+	*/
3777	+	public long reduceKeysToLong(long parallelismThreshold,
3778	+	ToLongFunction<? super K> transformer,
3779	+	long basis,
3780	+	LongBinaryOperator reducer) {
3781	+	if (transformer == null || reducer == null)
3782	+	throw new NullPointerException();
3783	+	return new MapReduceKeysToLongTask<K,V>
3784	+	(null, batchFor(parallelismThreshold), 0, 0, table,
3785	+	null, transformer, basis, reducer).invoke();
3786	+	}
3787	+
3788	+	/**
3789	+	* Returns the result of accumulating the given transformation
3790	+	* of all keys using the given reducer to combine values, and
3791	+	* the given basis as an identity value.
3792	+	*
3793	+	* @param parallelismThreshold the (estimated) number of elements
3794	+	* needed for this operation to be executed in parallel
3795	+	* @param transformer a function returning the transformation
3796	+	* for an element
3797	+	* @param basis the identity (initial default value) for the reduction
3798	+	* @param reducer a commutative associative combining function
3799	+	* @return the result of accumulating the given transformation
3800	+	* of all keys
3801	+	* @since 1.8
3802	+	*/
3803	+	public int reduceKeysToInt(long parallelismThreshold,
3804	+	ToIntFunction<? super K> transformer,
3805	+	int basis,
3806	+	IntBinaryOperator reducer) {
3807	+	if (transformer == null || reducer == null)
3808	+	throw new NullPointerException();
3809	+	return new MapReduceKeysToIntTask<K,V>
3810	+	(null, batchFor(parallelismThreshold), 0, 0, table,
3811	+	null, transformer, basis, reducer).invoke();
3812	+	}
3813	+
3814	+	/**
3815	+	* Performs the given action for each value.
3816	+	*
3817	+	* @param parallelismThreshold the (estimated) number of elements
3818	+	* needed for this operation to be executed in parallel
3819	+	* @param action the action
3820	+	* @since 1.8
3821	+	*/
3822	+	public void forEachValue(long parallelismThreshold,
3823	+	Consumer<? super V> action) {
3824	+	if (action == null)
3825	+	throw new NullPointerException();
3826	+	new ForEachValueTask<K,V>
3827	+	(null, batchFor(parallelismThreshold), 0, 0, table,
3828	+	action).invoke();
3829	+	}
3830	+
3831	+	/**
3832	+	* Performs the given action for each non-null transformation
3833	+	* of each value.
3834	+	*
3835	+	* @param parallelismThreshold the (estimated) number of elements
3836	+	* needed for this operation to be executed in parallel
3837	+	* @param transformer a function returning the transformation
3838	+	* for an element, or null if there is no transformation (in
3839	+	* which case the action is not applied)
3840	+	* @param action the action
3841	+	* @since 1.8
3842	+	*/
3843	+	public <U> void forEachValue(long parallelismThreshold,
3844	+	Function<? super V, ? extends U> transformer,
3845	+	Consumer<? super U> action) {
3846	+	if (transformer == null || action == null)
3847	+	throw new NullPointerException();
3848	+	new ForEachTransformedValueTask<K,V,U>
3849	+	(null, batchFor(parallelismThreshold), 0, 0, table,
3850	+	transformer, action).invoke();
3851	+	}
3852	+
3853	+	/**
3854	+	* Returns a non-null result from applying the given search
3855	+	* function on each value, or null if none.  Upon success,
3856	+	* further element processing is suppressed and the results of
3857	+	* any other parallel invocations of the search function are
3858	+	* ignored.
3859	+	*
3860	+	* @param parallelismThreshold the (estimated) number of elements
3861	+	* needed for this operation to be executed in parallel
3862	+	* @param searchFunction a function returning a non-null
3863	+	* result on success, else null
3864	+	* @return a non-null result from applying the given search
3865	+	* function on each value, or null if none
3866	+	* @since 1.8
3867	+	*/
3868	+	public <U> U searchValues(long parallelismThreshold,
3869	+	Function<? super V, ? extends U> searchFunction) {
3870	+	if (searchFunction == null) throw new NullPointerException();
3871	+	return new SearchValuesTask<K,V,U>
3872	+	(null, batchFor(parallelismThreshold), 0, 0, table,
3873	+	searchFunction, new AtomicReference<U>()).invoke();
3874	+	}
3875	+
3876	+	/**
3877	+	* Returns the result of accumulating all values using the
3878	+	* given reducer to combine values, or null if none.
3879	+	*
3880	+	* @param parallelismThreshold the (estimated) number of elements
3881	+	* needed for this operation to be executed in parallel
3882	+	* @param reducer a commutative associative combining function
3883	+	* @return the result of accumulating all values
3884	+	* @since 1.8
3885	+	*/
3886	+	public V reduceValues(long parallelismThreshold,
3887	+	BiFunction<? super V, ? super V, ? extends V> reducer) {
3888	+	if (reducer == null) throw new NullPointerException();
3889	+	return new ReduceValuesTask<K,V>
3890	+	(null, batchFor(parallelismThreshold), 0, 0, table,
3891	+	null, reducer).invoke();
3892	+	}
3893	+
3894	+	/**
3895	+	* Returns the result of accumulating the given transformation
3896	+	* of all values using the given reducer to combine values, or
3897	+	* null if none.
3898	+	*
3899	+	* @param parallelismThreshold the (estimated) number of elements
3900	+	* needed for this operation to be executed in parallel
3901	+	* @param transformer a function returning the transformation
3902	+	* for an element, or null if there is no transformation (in
3903	+	* which case it is not combined)
3904	+	* @param reducer a commutative associative combining function
3905	+	* @return the result of accumulating the given transformation
3906	+	* of all values
3907	+	* @since 1.8
3908	+	*/
3909	+	public <U> U reduceValues(long parallelismThreshold,
3910	+	Function<? super V, ? extends U> transformer,
3911	+	BiFunction<? super U, ? super U, ? extends U> reducer) {
3912	+	if (transformer == null || reducer == null)
3913	+	throw new NullPointerException();
3914	+	return new MapReduceValuesTask<K,V,U>
3915	+	(null, batchFor(parallelismThreshold), 0, 0, table,
3916	+	null, transformer, reducer).invoke();
3917	+	}
3918	+
3919	+	/**
3920	+	* Returns the result of accumulating the given transformation
3921	+	* of all values using the given reducer to combine values,
3922	+	* and the given basis as an identity value.
3923	+	*
3924	+	* @param parallelismThreshold the (estimated) number of elements
3925	+	* needed for this operation to be executed in parallel
3926	+	* @param transformer a function returning the transformation
3927	+	* for an element
3928	+	* @param basis the identity (initial default value) for the reduction
3929	+	* @param reducer a commutative associative combining function
3930	+	* @return the result of accumulating the given transformation
3931	+	* of all values
3932	+	* @since 1.8
3933	+	*/
3934	+	public double reduceValuesToDouble(long parallelismThreshold,
3935	+	ToDoubleFunction<? super V> transformer,
3936	+	double basis,
3937	+	DoubleBinaryOperator reducer) {
3938	+	if (transformer == null || reducer == null)
3939	+	throw new NullPointerException();
3940	+	return new MapReduceValuesToDoubleTask<K,V>
3941	+	(null, batchFor(parallelismThreshold), 0, 0, table,
3942	+	null, transformer, basis, reducer).invoke();
3943	+	}
3944	+
3945	+	/**
3946	+	* Returns the result of accumulating the given transformation
3947	+	* of all values using the given reducer to combine values,
3948	+	* and the given basis as an identity value.
3949	+	*
3950	+	* @param parallelismThreshold the (estimated) number of elements
3951	+	* needed for this operation to be executed in parallel
3952	+	* @param transformer a function returning the transformation
3953	+	* for an element
3954	+	* @param basis the identity (initial default value) for the reduction
3955	+	* @param reducer a commutative associative combining function
3956	+	* @return the result of accumulating the given transformation
3957	+	* of all values
3958	+	* @since 1.8
3959	+	*/
3960	+	public long reduceValuesToLong(long parallelismThreshold,
3961	+	ToLongFunction<? super V> transformer,
3962	+	long basis,
3963	+	LongBinaryOperator reducer) {
3964	+	if (transformer == null || reducer == null)
3965	+	throw new NullPointerException();
3966	+	return new MapReduceValuesToLongTask<K,V>
3967	+	(null, batchFor(parallelismThreshold), 0, 0, table,
3968	+	null, transformer, basis, reducer).invoke();
3969	+	}
3970	+
3971	+	/**
3972	+	* Returns the result of accumulating the given transformation
3973	+	* of all values using the given reducer to combine values,
3974	+	* and the given basis as an identity value.
3975	+	*
3976	+	* @param parallelismThreshold the (estimated) number of elements
3977	+	* needed for this operation to be executed in parallel
3978	+	* @param transformer a function returning the transformation
3979	+	* for an element
3980	+	* @param basis the identity (initial default value) for the reduction
3981	+	* @param reducer a commutative associative combining function
3982	+	* @return the result of accumulating the given transformation
3983	+	* of all values
3984	+	* @since 1.8
3985	+	*/
3986	+	public int reduceValuesToInt(long parallelismThreshold,
3987	+	ToIntFunction<? super V> transformer,
3988	+	int basis,
3989	+	IntBinaryOperator reducer) {
3990	+	if (transformer == null || reducer == null)
3991	+	throw new NullPointerException();
3992	+	return new MapReduceValuesToIntTask<K,V>
3993	+	(null, batchFor(parallelismThreshold), 0, 0, table,
3994	+	null, transformer, basis, reducer).invoke();
3995	+	}
3996	+
3997	+	/**
3998	+	* Performs the given action for each entry.
3999	+	*
4000	+	* @param parallelismThreshold the (estimated) number of elements
4001	+	* needed for this operation to be executed in parallel
4002	+	* @param action the action
4003	+	* @since 1.8
4004	+	*/
4005	+	public void forEachEntry(long parallelismThreshold,
4006	+	Consumer<? super Map.Entry<K,V>> action) {
4007	+	if (action == null) throw new NullPointerException();
4008	+	new ForEachEntryTask<K,V>(null, batchFor(parallelismThreshold), 0, 0, table,
4009	+	action).invoke();
4010	+	}
4011	+
4012	+	/**
4013	+	* Performs the given action for each non-null transformation
4014	+	* of each entry.
4015	+	*
4016	+	* @param parallelismThreshold the (estimated) number of elements
4017	+	* needed for this operation to be executed in parallel
4018	+	* @param transformer a function returning the transformation
4019	+	* for an element, or null if there is no transformation (in
4020	+	* which case the action is not applied)
4021	+	* @param action the action
4022	+	* @since 1.8
4023	+	*/
4024	+	public <U> void forEachEntry(long parallelismThreshold,
4025	+	Function<Map.Entry<K,V>, ? extends U> transformer,
4026	+	Consumer<? super U> action) {
4027	+	if (transformer == null || action == null)
4028	+	throw new NullPointerException();
4029	+	new ForEachTransformedEntryTask<K,V,U>
4030	+	(null, batchFor(parallelismThreshold), 0, 0, table,
4031	+	transformer, action).invoke();
4032	+	}
4033	+
4034	+	/**
4035	+	* Returns a non-null result from applying the given search
4036	+	* function on each entry, or null if none.  Upon success,
4037	+	* further element processing is suppressed and the results of
4038	+	* any other parallel invocations of the search function are
4039	+	* ignored.
4040	+	*
4041	+	* @param parallelismThreshold the (estimated) number of elements
4042	+	* needed for this operation to be executed in parallel
4043	+	* @param searchFunction a function returning a non-null
4044	+	* result on success, else null
4045	+	* @return a non-null result from applying the given search
4046	+	* function on each entry, or null if none
4047	+	* @since 1.8
4048	+	*/
4049	+	public <U> U searchEntries(long parallelismThreshold,
4050	+	Function<Map.Entry<K,V>, ? extends U> searchFunction) {
4051	+	if (searchFunction == null) throw new NullPointerException();
4052	+	return new SearchEntriesTask<K,V,U>
4053	+	(null, batchFor(parallelismThreshold), 0, 0, table,
4054	+	searchFunction, new AtomicReference<U>()).invoke();
4055	+	}
4056	+
4057	+	/**
4058	+	* Returns the result of accumulating all entries using the
4059	+	* given reducer to combine values, or null if none.
4060	+	*
4061	+	* @param parallelismThreshold the (estimated) number of elements
4062	+	* needed for this operation to be executed in parallel
4063	+	* @param reducer a commutative associative combining function
4064	+	* @return the result of accumulating all entries
4065	+	* @since 1.8
4066	+	*/
4067	+	public Map.Entry<K,V> reduceEntries(long parallelismThreshold,
4068	+	BiFunction<Map.Entry<K,V>, Map.Entry<K,V>, ? extends Map.Entry<K,V>> reducer) {
4069	+	if (reducer == null) throw new NullPointerException();
4070	+	return new ReduceEntriesTask<K,V>
4071	+	(null, batchFor(parallelismThreshold), 0, 0, table,
4072	+	null, reducer).invoke();
4073		}
4074
4075	<	final class Values extends AbstractCollection<V> {
4076	<	public Iterator<V> iterator() {
4077	<	return new ValueIterator();
4075	>	/**
4076	>	* Returns the result of accumulating the given transformation
4077	>	* of all entries using the given reducer to combine values,
4078	>	* or null if none.
4079	>	*
4080	>	* @param parallelismThreshold the (estimated) number of elements
4081	>	* needed for this operation to be executed in parallel
4082	>	* @param transformer a function returning the transformation
4083	>	* for an element, or null if there is no transformation (in
4084	>	* which case it is not combined)
4085	>	* @param reducer a commutative associative combining function
4086	>	* @return the result of accumulating the given transformation
4087	>	* of all entries
4088	>	* @since 1.8
4089	>	*/
4090	>	public <U> U reduceEntries(long parallelismThreshold,
4091	>	Function<Map.Entry<K,V>, ? extends U> transformer,
4092	>	BiFunction<? super U, ? super U, ? extends U> reducer) {
4093	>	if (transformer == null || reducer == null)
4094	>	throw new NullPointerException();
4095	>	return new MapReduceEntriesTask<K,V,U>
4096	>	(null, batchFor(parallelismThreshold), 0, 0, table,
4097	>	null, transformer, reducer).invoke();
4098	>	}
4099	>
4100	>	/**
4101	>	* Returns the result of accumulating the given transformation
4102	>	* of all entries using the given reducer to combine values,
4103	>	* and the given basis as an identity value.
4104	>	*
4105	>	* @param parallelismThreshold the (estimated) number of elements
4106	>	* needed for this operation to be executed in parallel
4107	>	* @param transformer a function returning the transformation
4108	>	* for an element
4109	>	* @param basis the identity (initial default value) for the reduction
4110	>	* @param reducer a commutative associative combining function
4111	>	* @return the result of accumulating the given transformation
4112	>	* of all entries
4113	>	* @since 1.8
4114	>	*/
4115	>	public double reduceEntriesToDouble(long parallelismThreshold,
4116	>	ToDoubleFunction<Map.Entry<K,V>> transformer,
4117	>	double basis,
4118	>	DoubleBinaryOperator reducer) {
4119	>	if (transformer == null || reducer == null)
4120	>	throw new NullPointerException();
4121	>	return new MapReduceEntriesToDoubleTask<K,V>
4122	>	(null, batchFor(parallelismThreshold), 0, 0, table,
4123	>	null, transformer, basis, reducer).invoke();
4124	>	}
4125	>
4126	>	/**
4127	>	* Returns the result of accumulating the given transformation
4128	>	* of all entries using the given reducer to combine values,
4129	>	* and the given basis as an identity value.
4130	>	*
4131	>	* @param parallelismThreshold the (estimated) number of elements
4132	>	* needed for this operation to be executed in parallel
4133	>	* @param transformer a function returning the transformation
4134	>	* for an element
4135	>	* @param basis the identity (initial default value) for the reduction
4136	>	* @param reducer a commutative associative combining function
4137	>	* @return the result of accumulating the given transformation
4138	>	* of all entries
4139	>	* @since 1.8
4140	>	*/
4141	>	public long reduceEntriesToLong(long parallelismThreshold,
4142	>	ToLongFunction<Map.Entry<K,V>> transformer,
4143	>	long basis,
4144	>	LongBinaryOperator reducer) {
4145	>	if (transformer == null || reducer == null)
4146	>	throw new NullPointerException();
4147	>	return new MapReduceEntriesToLongTask<K,V>
4148	>	(null, batchFor(parallelismThreshold), 0, 0, table,
4149	>	null, transformer, basis, reducer).invoke();
4150	>	}
4151	>
4152	>	/**
4153	>	* Returns the result of accumulating the given transformation
4154	>	* of all entries using the given reducer to combine values,
4155	>	* and the given basis as an identity value.
4156	>	*
4157	>	* @param parallelismThreshold the (estimated) number of elements
4158	>	* needed for this operation to be executed in parallel
4159	>	* @param transformer a function returning the transformation
4160	>	* for an element
4161	>	* @param basis the identity (initial default value) for the reduction
4162	>	* @param reducer a commutative associative combining function
4163	>	* @return the result of accumulating the given transformation
4164	>	* of all entries
4165	>	* @since 1.8
4166	>	*/
4167	>	public int reduceEntriesToInt(long parallelismThreshold,
4168	>	ToIntFunction<Map.Entry<K,V>> transformer,
4169	>	int basis,
4170	>	IntBinaryOperator reducer) {
4171	>	if (transformer == null || reducer == null)
4172	>	throw new NullPointerException();
4173	>	return new MapReduceEntriesToIntTask<K,V>
4174	>	(null, batchFor(parallelismThreshold), 0, 0, table,
4175	>	null, transformer, basis, reducer).invoke();
4176	>	}
4177	>
4178	>
4179	>	/* ----------------Views -------------- */
4180	>
4181	>	/**
4182	>	* Base class for views.
4183	>	*/
4184	>	abstract static class CollectionView<K,V,E>
4185	>	implements Collection<E>, java.io.Serializable {
4186	>	private static final long serialVersionUID = 7249069246763182397L;
4187	>	final ConcurrentHashMap<K,V> map;
4188	>	CollectionView(ConcurrentHashMap<K,V> map)  { this.map = map; }
4189	>
4190	>	/**
4191	>	* Returns the map backing this view.
4192	>	*
4193	>	* @return the map backing this view
4194	>	*/
4195	>	public ConcurrentHashMap<K,V> getMap() { return map; }
4196	>
4197	>	/**
4198	>	* Removes all of the elements from this view, by removing all
4199	>	* the mappings from the map backing this view.
4200	>	*/
4201	>	public final void clear()      { map.clear(); }
4202	>	public final int size()        { return map.size(); }
4203	>	public final boolean isEmpty() { return map.isEmpty(); }
4204	>
4205	>	// implementations below rely on concrete classes supplying these
4206	>	// abstract methods
4207	>	/**
4208	>	* Returns a "weakly consistent" iterator that will never
4209	>	* throw {@link ConcurrentModificationException}, and
4210	>	* guarantees to traverse elements as they existed upon
4211	>	* construction of the iterator, and may (but is not
4212	>	* guaranteed to) reflect any modifications subsequent to
4213	>	* construction.
4214	>	*/
4215	>	public abstract Iterator<E> iterator();
4216	>	public abstract boolean contains(Object o);
4217	>	public abstract boolean remove(Object o);
4218	>
4219	>	private static final String oomeMsg = "Required array size too large";
4220	>
4221	>	public final Object[] toArray() {
4222	>	long sz = map.mappingCount();
4223	>	if (sz > MAX_ARRAY_SIZE)
4224	>	throw new OutOfMemoryError(oomeMsg);
4225	>	int n = (int)sz;
4226	>	Object[] r = new Object[n];
4227	>	int i = 0;
4228	>	for (E e : this) {
4229	>	if (i == n) {
4230	>	if (n >= MAX_ARRAY_SIZE)
4231	>	throw new OutOfMemoryError(oomeMsg);
4232	>	if (n >= MAX_ARRAY_SIZE - (MAX_ARRAY_SIZE >>> 1) - 1)
4233	>	n = MAX_ARRAY_SIZE;
4234	>	else
4235	>	n += (n >>> 1) + 1;
4236	>	r = Arrays.copyOf(r, n);
4237	>	}
4238	>	r[i++] = e;
4239	>	}
4240	>	return (i == n) ? r : Arrays.copyOf(r, i);
4241	>	}
4242	>
4243	>	@SuppressWarnings("unchecked")
4244	>	public final <T> T[] toArray(T[] a) {
4245	>	long sz = map.mappingCount();
4246	>	if (sz > MAX_ARRAY_SIZE)
4247	>	throw new OutOfMemoryError(oomeMsg);
4248	>	int m = (int)sz;
4249	>	T[] r = (a.length >= m) ? a :
4250	>	(T[])java.lang.reflect.Array
4251	>	.newInstance(a.getClass().getComponentType(), m);
4252	>	int n = r.length;
4253	>	int i = 0;
4254	>	for (E e : this) {
4255	>	if (i == n) {
4256	>	if (n >= MAX_ARRAY_SIZE)
4257	>	throw new OutOfMemoryError(oomeMsg);
4258	>	if (n >= MAX_ARRAY_SIZE - (MAX_ARRAY_SIZE >>> 1) - 1)
4259	>	n = MAX_ARRAY_SIZE;
4260	>	else
4261	>	n += (n >>> 1) + 1;
4262	>	r = Arrays.copyOf(r, n);
4263	>	}
4264	>	r[i++] = (T)e;
4265	>	}
4266	>	if (a == r && i < n) {
4267	>	r[i] = null; // null-terminate
4268	>	return r;
4269	>	}
4270	>	return (i == n) ? r : Arrays.copyOf(r, i);
4271		}
4272	<	public int size() {
4273	<	return ConcurrentHashMap.this.size();
4272	>
4273	>	/**
4274	>	* Returns a string representation of this collection.
4275	>	* The string representation consists of the string representations
4276	>	* of the collection's elements in the order they are returned by
4277	>	* its iterator, enclosed in square brackets ({@code "[]"}).
4278	>	* Adjacent elements are separated by the characters {@code ", "}
4279	>	* (comma and space).  Elements are converted to strings as by
4280	>	* {@link String#valueOf(Object)}.
4281	>	*
4282	>	* @return a string representation of this collection
4283	>	*/
4284	>	public final String toString() {
4285	>	StringBuilder sb = new StringBuilder();
4286	>	sb.append('[');
4287	>	Iterator<E> it = iterator();
4288	>	if (it.hasNext()) {
4289	>	for (;;) {
4290	>	Object e = it.next();
4291	>	sb.append(e == this ? "(this Collection)" : e);
4292	>	if (!it.hasNext())
4293	>	break;
4294	>	sb.append(',').append(' ');
4295	>	}
4296	>	}
4297	>	return sb.append(']').toString();
4298		}
4299	<	public boolean isEmpty() {
4300	<	return ConcurrentHashMap.this.isEmpty();
4299	>
4300	>	public final boolean containsAll(Collection<?> c) {
4301	>	if (c != this) {
4302	>	for (Object e : c) {
4303	>	if (e == null || !contains(e))
4304	>	return false;
4305	>	}
4306	>	}
4307	>	return true;
4308		}
4309	<	public boolean contains(Object o) {
4310	<	return ConcurrentHashMap.this.containsValue(o);
4309	>
4310	>	public final boolean removeAll(Collection<?> c) {
4311	>	boolean modified = false;
4312	>	for (Iterator<E> it = iterator(); it.hasNext();) {
4313	>	if (c.contains(it.next())) {
4314	>	it.remove();
4315	>	modified = true;
4316	>	}
4317	>	}
4318	>	return modified;
4319	>	}
4320	>
4321	>	public final boolean retainAll(Collection<?> c) {
4322	>	boolean modified = false;
4323	>	for (Iterator<E> it = iterator(); it.hasNext();) {
4324	>	if (!c.contains(it.next())) {
4325	>	it.remove();
4326	>	modified = true;
4327	>	}
4328	>	}
4329	>	return modified;
4330		}
4331	<	public void clear() {
4332	<	ConcurrentHashMap.this.clear();
4331	>
4332	>	}
4333	>
4334	>	/**
4335	>	* A view of a ConcurrentHashMap as a {@link Set} of keys, in
4336	>	* which additions may optionally be enabled by mapping to a
4337	>	* common value.  This class cannot be directly instantiated.
4338	>	* See {@link #keySet() keySet()},
4339	>	* {@link #keySet(Object) keySet(V)},
4340	>	* {@link #newKeySet() newKeySet()},
4341	>	* {@link #newKeySet(int) newKeySet(int)}.
4342	>	*
4343	>	* @since 1.8
4344	>	*/
4345	>	public static class KeySetView<K,V> extends CollectionView<K,V,K>
4346	>	implements Set<K>, java.io.Serializable {
4347	>	private static final long serialVersionUID = 7249069246763182397L;
4348	>	private final V value;
4349	>	KeySetView(ConcurrentHashMap<K,V> map, V value) {  // non-public
4350	>	super(map);
4351	>	this.value = value;
4352	>	}
4353	>
4354	>	/**
4355	>	* Returns the default mapped value for additions,
4356	>	* or {@code null} if additions are not supported.
4357	>	*
4358	>	* @return the default mapped value for additions, or {@code null}
4359	>	* if not supported
4360	>	*/
4361	>	public V getMappedValue() { return value; }
4362	>
4363	>	/**
4364	>	* {@inheritDoc}
4365	>	* @throws NullPointerException if the specified key is null
4366	>	*/
4367	>	public boolean contains(Object o) { return map.containsKey(o); }
4368	>
4369	>	/**
4370	>	* Removes the key from this map view, by removing the key (and its
4371	>	* corresponding value) from the backing map.  This method does
4372	>	* nothing if the key is not in the map.
4373	>	*
4374	>	* @param  o the key to be removed from the backing map
4375	>	* @return {@code true} if the backing map contained the specified key
4376	>	* @throws NullPointerException if the specified key is null
4377	>	*/
4378	>	public boolean remove(Object o) { return map.remove(o) != null; }
4379	>
4380	>	/**
4381	>	* @return an iterator over the keys of the backing map
4382	>	*/
4383	>	public Iterator<K> iterator() {
4384	>	Node<K,V>[] t;
4385	>	ConcurrentHashMap<K,V> m = map;
4386	>	int f = (t = m.table) == null ? 0 : t.length;
4387	>	return new KeyIterator<K,V>(t, f, 0, f, m);
4388	>	}
4389	>
4390	>	/**
4391	>	* Adds the specified key to this set view by mapping the key to
4392	>	* the default mapped value in the backing map, if defined.
4393	>	*
4394	>	* @param e key to be added
4395	>	* @return {@code true} if this set changed as a result of the call
4396	>	* @throws NullPointerException if the specified key is null
4397	>	* @throws UnsupportedOperationException if no default mapped value
4398	>	* for additions was provided
4399	>	*/
4400	>	public boolean add(K e) {
4401	>	V v;
4402	>	if ((v = value) == null)
4403	>	throw new UnsupportedOperationException();
4404	>	return map.putVal(e, v, true) == null;
4405	>	}
4406	>
4407	>	/**
4408	>	* Adds all of the elements in the specified collection to this set,
4409	>	* as if by calling {@link #add} on each one.
4410	>	*
4411	>	* @param c the elements to be inserted into this set
4412	>	* @return {@code true} if this set changed as a result of the call
4413	>	* @throws NullPointerException if the collection or any of its
4414	>	* elements are {@code null}
4415	>	* @throws UnsupportedOperationException if no default mapped value
4416	>	* for additions was provided
4417	>	*/
4418	>	public boolean addAll(Collection<? extends K> c) {
4419	>	boolean added = false;
4420	>	V v;
4421	>	if ((v = value) == null)
4422	>	throw new UnsupportedOperationException();
4423	>	for (K e : c) {
4424	>	if (map.putVal(e, v, true) == null)
4425	>	added = true;
4426	>	}
4427	>	return added;
4428	>	}
4429	>
4430	>	public int hashCode() {
4431	>	int h = 0;
4432	>	for (K e : this)
4433	>	h += e.hashCode();
4434	>	return h;
4435	>	}
4436	>
4437	>	public boolean equals(Object o) {
4438	>	Set<?> c;
4439	>	return ((o instanceof Set) &&
4440	>	((c = (Set<?>)o) == this ||
4441	>	(containsAll(c) && c.containsAll(this))));
4442	>	}
4443	>
4444	>	public Spliterator<K> spliterator() {
4445	>	Node<K,V>[] t;
4446	>	ConcurrentHashMap<K,V> m = map;
4447	>	long n = m.sumCount();
4448	>	int f = (t = m.table) == null ? 0 : t.length;
4449	>	return new KeySpliterator<K,V>(t, f, 0, f, n < 0L ? 0L : n);
4450	>	}
4451	>
4452	>	public void forEach(Consumer<? super K> action) {
4453	>	if (action == null) throw new NullPointerException();
4454	>	Node<K,V>[] t;
4455	>	if ((t = map.table) != null) {
4456	>	Traverser<K,V> it = new Traverser<K,V>(t, t.length, 0, t.length);
4457	>	for (Node<K,V> p; (p = it.advance()) != null; )
4458	>	action.accept(p.key);
4459	>	}
4460		}
4461		}
4462
4463	<	final class EntrySet extends AbstractSet<Map.Entry<K,V>> {
4464	<	public Iterator<Map.Entry<K,V>> iterator() {
4465	<	return new EntryIterator();
4463	>	/**
4464	>	* A view of a ConcurrentHashMap as a {@link Collection} of
4465	>	* values, in which additions are disabled. This class cannot be
4466	>	* directly instantiated. See {@link #values()}.
4467	>	*/
4468	>	static final class ValuesView<K,V> extends CollectionView<K,V,V>
4469	>	implements Collection<V>, java.io.Serializable {
4470	>	private static final long serialVersionUID = 2249069246763182397L;
4471	>	ValuesView(ConcurrentHashMap<K,V> map) { super(map); }
4472	>	public final boolean contains(Object o) {
4473	>	return map.containsValue(o);
4474	>	}
4475	>
4476	>	public final boolean remove(Object o) {
4477	>	if (o != null) {
4478	>	for (Iterator<V> it = iterator(); it.hasNext();) {
4479	>	if (o.equals(it.next())) {
4480	>	it.remove();
4481	>	return true;
4482	>	}
4483	>	}
4484	>	}
4485	>	return false;
4486		}
4487	+
4488	+	public final Iterator<V> iterator() {
4489	+	ConcurrentHashMap<K,V> m = map;
4490	+	Node<K,V>[] t;
4491	+	int f = (t = m.table) == null ? 0 : t.length;
4492	+	return new ValueIterator<K,V>(t, f, 0, f, m);
4493	+	}
4494	+
4495	+	public final boolean add(V e) {
4496	+	throw new UnsupportedOperationException();
4497	+	}
4498	+	public final boolean addAll(Collection<? extends V> c) {
4499	+	throw new UnsupportedOperationException();
4500	+	}
4501	+
4502	+	public Spliterator<V> spliterator() {
4503	+	Node<K,V>[] t;
4504	+	ConcurrentHashMap<K,V> m = map;
4505	+	long n = m.sumCount();
4506	+	int f = (t = m.table) == null ? 0 : t.length;
4507	+	return new ValueSpliterator<K,V>(t, f, 0, f, n < 0L ? 0L : n);
4508	+	}
4509	+
4510	+	public void forEach(Consumer<? super V> action) {
4511	+	if (action == null) throw new NullPointerException();
4512	+	Node<K,V>[] t;
4513	+	if ((t = map.table) != null) {
4514	+	Traverser<K,V> it = new Traverser<K,V>(t, t.length, 0, t.length);
4515	+	for (Node<K,V> p; (p = it.advance()) != null; )
4516	+	action.accept(p.val);
4517	+	}
4518	+	}
4519	+	}
4520	+
4521	+	/**
4522	+	* A view of a ConcurrentHashMap as a {@link Set} of (key, value)
4523	+	* entries.  This class cannot be directly instantiated. See
4524	+	* {@link #entrySet()}.
4525	+	*/
4526	+	static final class EntrySetView<K,V> extends CollectionView<K,V,Map.Entry<K,V>>
4527	+	implements Set<Map.Entry<K,V>>, java.io.Serializable {
4528	+	private static final long serialVersionUID = 2249069246763182397L;
4529	+	EntrySetView(ConcurrentHashMap<K,V> map) { super(map); }
4530	+
4531		public boolean contains(Object o) {
4532	<	if (!(o instanceof Map.Entry))
4533	<	return false;
4534	<	Map.Entry<?,?> e = (Map.Entry<?,?>)o;
4535	<	V v = ConcurrentHashMap.this.get(e.getKey());
4536	<	return v != null && v.equals(e.getValue());
4532	>	Object k, v, r; Map.Entry<?,?> e;
4533	>	return ((o instanceof Map.Entry) &&
4534	>	(k = (e = (Map.Entry<?,?>)o).getKey()) != null &&
4535	>	(r = map.get(k)) != null &&
4536	>	(v = e.getValue()) != null &&
4537	>	(v == r || v.equals(r)));
4538		}
4539	+
4540		public boolean remove(Object o) {
4541	<	if (!(o instanceof Map.Entry))
4542	<	return false;
4543	<	Map.Entry<?,?> e = (Map.Entry<?,?>)o;
4544	<	return ConcurrentHashMap.this.remove(e.getKey(), e.getValue());
4541	>	Object k, v; Map.Entry<?,?> e;
4542	>	return ((o instanceof Map.Entry) &&
4543	>	(k = (e = (Map.Entry<?,?>)o).getKey()) != null &&
4544	>	(v = e.getValue()) != null &&
4545	>	map.remove(k, v));
4546	>	}
4547	>
4548	>	/**
4549	>	* @return an iterator over the entries of the backing map
4550	>	*/
4551	>	public Iterator<Map.Entry<K,V>> iterator() {
4552	>	ConcurrentHashMap<K,V> m = map;
4553	>	Node<K,V>[] t;
4554	>	int f = (t = m.table) == null ? 0 : t.length;
4555	>	return new EntryIterator<K,V>(t, f, 0, f, m);
4556	>	}
4557	>
4558	>	public boolean add(Entry<K,V> e) {
4559	>	return map.putVal(e.getKey(), e.getValue(), false) == null;
4560		}
4561	<	public int size() {
4562	<	return ConcurrentHashMap.this.size();
4561	>
4562	>	public boolean addAll(Collection<? extends Entry<K,V>> c) {
4563	>	boolean added = false;
4564	>	for (Entry<K,V> e : c) {
4565	>	if (add(e))
4566	>	added = true;
4567	>	}
4568	>	return added;
4569		}
4570	<	public boolean isEmpty() {
4571	<	return ConcurrentHashMap.this.isEmpty();
4570	>
4571	>	public final int hashCode() {
4572	>	int h = 0;
4573	>	Node<K,V>[] t;
4574	>	if ((t = map.table) != null) {
4575	>	Traverser<K,V> it = new Traverser<K,V>(t, t.length, 0, t.length);
4576	>	for (Node<K,V> p; (p = it.advance()) != null; ) {
4577	>	h += p.hashCode();
4578	>	}
4579	>	}
4580	>	return h;
4581		}
4582	<	public void clear() {
4583	<	ConcurrentHashMap.this.clear();
4582	>
4583	>	public final boolean equals(Object o) {
4584	>	Set<?> c;
4585	>	return ((o instanceof Set) &&
4586	>	((c = (Set<?>)o) == this ||
4587	>	(containsAll(c) && c.containsAll(this))));
4588	>	}
4589	>
4590	>	public Spliterator<Map.Entry<K,V>> spliterator() {
4591	>	Node<K,V>[] t;
4592	>	ConcurrentHashMap<K,V> m = map;
4593	>	long n = m.sumCount();
4594	>	int f = (t = m.table) == null ? 0 : t.length;
4595	>	return new EntrySpliterator<K,V>(t, f, 0, f, n < 0L ? 0L : n, m);
4596	>	}
4597	>
4598	>	public void forEach(Consumer<? super Map.Entry<K,V>> action) {
4599	>	if (action == null) throw new NullPointerException();
4600	>	Node<K,V>[] t;
4601	>	if ((t = map.table) != null) {
4602	>	Traverser<K,V> it = new Traverser<K,V>(t, t.length, 0, t.length);
4603	>	for (Node<K,V> p; (p = it.advance()) != null; )
4604	>	action.accept(new MapEntry<K,V>(p.key, p.val, map));
4605	>	}
4606		}
4607	+
4608		}
4609
4610	<	/* ---------------- Serialization Support -------------- */
4610	>	// -------------------------------------------------------
4611
4612		/**
4613	<	* Saves the state of the <tt>ConcurrentHashMap</tt> instance to a
4614	<	* stream (i.e., serializes it).
1402	<	* @param s the stream
1403	<	* @serialData
1404	<	* the key (Object) and value (Object)
1405	<	* for each key-value mapping, followed by a null pair.
1406	<	* The key-value mappings are emitted in no particular order.
4613	>	* Base class for bulk tasks. Repeats some fields and code from
4614	>	* class Traverser, because we need to subclass CountedCompleter.
4615		*/
4616	<	private void writeObject(java.io.ObjectOutputStream s) throws IOException {
4617	<	// force all segments for serialization compatibility
4618	<	for (int k = 0; k < segments.length; ++k)
4619	<	ensureSegment(k);
4620	<	s.defaultWriteObject();
4621	<
4622	<	final Segment<K,V>[] segments = this.segments;
4623	<	for (int k = 0; k < segments.length; ++k) {
4624	<	Segment<K,V> seg = segmentAt(segments, k);
4625	<	seg.lock();
4626	<	try {
4627	<	HashEntry<K,V>[] tab = seg.table;
4628	<	for (int i = 0; i < tab.length; ++i) {
4629	<	HashEntry<K,V> e;
4630	<	for (e = entryAt(tab, i); e != null; e = e.next) {
4631	<	s.writeObject(e.key);
4632	<	s.writeObject(e.value);
4616	>	abstract static class BulkTask<K,V,R> extends CountedCompleter<R> {
4617	>	Node<K,V>[] tab;        // same as Traverser
4618	>	Node<K,V> next;
4619	>	int index;
4620	>	int baseIndex;
4621	>	int baseLimit;
4622	>	final int baseSize;
4623	>	int batch;              // split control
4624	>
4625	>	BulkTask(BulkTask<K,V,?> par, int b, int i, int f, Node<K,V>[] t) {
4626	>	super(par);
4627	>	this.batch = b;
4628	>	this.index = this.baseIndex = i;
4629	>	if ((this.tab = t) == null)
4630	>	this.baseSize = this.baseLimit = 0;
4631	>	else if (par == null)
4632	>	this.baseSize = this.baseLimit = t.length;
4633	>	else {
4634	>	this.baseLimit = f;
4635	>	this.baseSize = par.baseSize;
4636	>	}
4637	>	}
4638	>
4639	>	/**
4640	>	* Same as Traverser version
4641	>	*/
4642	>	final Node<K,V> advance() {
4643	>	Node<K,V> e;
4644	>	if ((e = next) != null)
4645	>	e = e.next;
4646	>	for (;;) {
4647	>	Node<K,V>[] t; int i, n; K ek;  // must use locals in checks
4648	>	if (e != null)
4649	>	return next = e;
4650	>	if (baseIndex >= baseLimit || (t = tab) == null ||
4651	>	(n = t.length) <= (i = index) || i < 0)
4652	>	return next = null;
4653	>	if ((e = tabAt(t, index)) != null && e.key == null) {
4654	>	if (e instanceof ForwardingNode) {
4655	>	tab = ((ForwardingNode<K,V>)e).nextTable;
4656	>	e = null;
4657	>	continue;
4658		}
4659	+	else if (e instanceof TreeBin)
4660	+	e = ((TreeBin<K,V>)e).first;
4661	+	else
4662	+	e = null;
4663		}
4664	<	} finally {
4665	<	seg.unlock();
4664	>	if ((index += baseSize) >= n)
4665	>	index = ++baseIndex;    // visit upper slots if present
4666		}
4667		}
1431	–	s.writeObject(null);
1432	–	s.writeObject(null);
4668		}
4669
4670	<	/**
4671	<	* Reconstitutes the <tt>ConcurrentHashMap</tt> instance from a
4672	<	* stream (i.e., deserializes it).
4673	<	* @param s the stream
4674	<	*/
4675	<	@SuppressWarnings("unchecked")
4676	<	private void readObject(java.io.ObjectInputStream s)
4677	<	throws IOException, ClassNotFoundException {
4678	<	s.defaultReadObject();
4670	>	/*
4671	>	* Task classes. Coded in a regular but ugly format/style to
4672	>	* simplify checks that each variant differs in the right way from
4673	>	* others. The null screenings exist because compilers cannot tell
4674	>	* that we've already null-checked task arguments, so we force
4675	>	* simplest hoisted bypass to help avoid convoluted traps.
4676	>	*/
4677	>	@SuppressWarnings("serial")
4678	>	static final class ForEachKeyTask<K,V>
4679	>	extends BulkTask<K,V,Void> {
4680	>	final Consumer<? super K> action;
4681	>	ForEachKeyTask
4682	>	(BulkTask<K,V,?> p, int b, int i, int f, Node<K,V>[] t,
4683	>	Consumer<? super K> action) {
4684	>	super(p, b, i, f, t);
4685	>	this.action = action;
4686	>	}
4687	>	public final void compute() {
4688	>	final Consumer<? super K> action;
4689	>	if ((action = this.action) != null) {
4690	>	for (int i = baseIndex, f, h; batch > 0 &&
4691	>	(h = ((f = baseLimit) + i) >>> 1) > i;) {
4692	>	addToPendingCount(1);
4693	>	new ForEachKeyTask<K,V>
4694	>	(this, batch >>>= 1, baseLimit = h, f, tab,
4695	>	action).fork();
4696	>	}
4697	>	for (Node<K,V> p; (p = advance()) != null;)
4698	>	action.accept(p.key);
4699	>	propagateCompletion();
4700	>	}
4701	>	}
4702	>	}
4703
4704	<	// Re-initialize segments to be minimally sized, and let grow.
4705	<	int cap = MIN_SEGMENT_TABLE_CAPACITY;
4706	<	final Segment<K,V>[] segments = this.segments;
4707	<	for (int k = 0; k < segments.length; ++k) {
4708	<	Segment<K,V> seg = segments[k];
4709	<	if (seg != null) {
4710	<	seg.threshold = (int)(cap * seg.loadFactor);
4711	<	seg.table = (HashEntry<K,V>[]) new HashEntry[cap];
4704	>	@SuppressWarnings("serial")
4705	>	static final class ForEachValueTask<K,V>
4706	>	extends BulkTask<K,V,Void> {
4707	>	final Consumer<? super V> action;
4708	>	ForEachValueTask
4709	>	(BulkTask<K,V,?> p, int b, int i, int f, Node<K,V>[] t,
4710	>	Consumer<? super V> action) {
4711	>	super(p, b, i, f, t);
4712	>	this.action = action;
4713	>	}
4714	>	public final void compute() {
4715	>	final Consumer<? super V> action;
4716	>	if ((action = this.action) != null) {
4717	>	for (int i = baseIndex, f, h; batch > 0 &&
4718	>	(h = ((f = baseLimit) + i) >>> 1) > i;) {
4719	>	addToPendingCount(1);
4720	>	new ForEachValueTask<K,V>
4721	>	(this, batch >>>= 1, baseLimit = h, f, tab,
4722	>	action).fork();
4723	>	}
4724	>	for (Node<K,V> p; (p = advance()) != null;)
4725	>	action.accept(p.val);
4726	>	propagateCompletion();
4727		}
4728		}
4729	+	}
4730
4731	<	// Read the keys and values, and put the mappings in the table
4732	<	for (;;) {
4733	<	K key = (K) s.readObject();
4734	<	V value = (V) s.readObject();
4735	<	if (key == null)
4736	<	break;
4737	<	put(key, value);
4731	>	@SuppressWarnings("serial")
4732	>	static final class ForEachEntryTask<K,V>
4733	>	extends BulkTask<K,V,Void> {
4734	>	final Consumer<? super Entry<K,V>> action;
4735	>	ForEachEntryTask
4736	>	(BulkTask<K,V,?> p, int b, int i, int f, Node<K,V>[] t,
4737	>	Consumer<? super Entry<K,V>> action) {
4738	>	super(p, b, i, f, t);
4739	>	this.action = action;
4740	>	}
4741	>	public final void compute() {
4742	>	final Consumer<? super Entry<K,V>> action;
4743	>	if ((action = this.action) != null) {
4744	>	for (int i = baseIndex, f, h; batch > 0 &&
4745	>	(h = ((f = baseLimit) + i) >>> 1) > i;) {
4746	>	addToPendingCount(1);
4747	>	new ForEachEntryTask<K,V>
4748	>	(this, batch >>>= 1, baseLimit = h, f, tab,
4749	>	action).fork();
4750	>	}
4751	>	for (Node<K,V> p; (p = advance()) != null; )
4752	>	action.accept(p);
4753	>	propagateCompletion();
4754	>	}
4755	>	}
4756	>	}
4757	>
4758	>	@SuppressWarnings("serial")
4759	>	static final class ForEachMappingTask<K,V>
4760	>	extends BulkTask<K,V,Void> {
4761	>	final BiConsumer<? super K, ? super V> action;
4762	>	ForEachMappingTask
4763	>	(BulkTask<K,V,?> p, int b, int i, int f, Node<K,V>[] t,
4764	>	BiConsumer<? super K,? super V> action) {
4765	>	super(p, b, i, f, t);
4766	>	this.action = action;
4767	>	}
4768	>	public final void compute() {
4769	>	final BiConsumer<? super K, ? super V> action;
4770	>	if ((action = this.action) != null) {
4771	>	for (int i = baseIndex, f, h; batch > 0 &&
4772	>	(h = ((f = baseLimit) + i) >>> 1) > i;) {
4773	>	addToPendingCount(1);
4774	>	new ForEachMappingTask<K,V>
4775	>	(this, batch >>>= 1, baseLimit = h, f, tab,
4776	>	action).fork();
4777	>	}
4778	>	for (Node<K,V> p; (p = advance()) != null; )
4779	>	action.accept(p.key, p.val);
4780	>	propagateCompletion();
4781	>	}
4782	>	}
4783	>	}
4784	>
4785	>	@SuppressWarnings("serial")
4786	>	static final class ForEachTransformedKeyTask<K,V,U>
4787	>	extends BulkTask<K,V,Void> {
4788	>	final Function<? super K, ? extends U> transformer;
4789	>	final Consumer<? super U> action;
4790	>	ForEachTransformedKeyTask
4791	>	(BulkTask<K,V,?> p, int b, int i, int f, Node<K,V>[] t,
4792	>	Function<? super K, ? extends U> transformer, Consumer<? super U> action) {
4793	>	super(p, b, i, f, t);
4794	>	this.transformer = transformer; this.action = action;
4795	>	}
4796	>	public final void compute() {
4797	>	final Function<? super K, ? extends U> transformer;
4798	>	final Consumer<? super U> action;
4799	>	if ((transformer = this.transformer) != null &&
4800	>	(action = this.action) != null) {
4801	>	for (int i = baseIndex, f, h; batch > 0 &&
4802	>	(h = ((f = baseLimit) + i) >>> 1) > i;) {
4803	>	addToPendingCount(1);
4804	>	new ForEachTransformedKeyTask<K,V,U>
4805	>	(this, batch >>>= 1, baseLimit = h, f, tab,
4806	>	transformer, action).fork();
4807	>	}
4808	>	for (Node<K,V> p; (p = advance()) != null; ) {
4809	>	U u;
4810	>	if ((u = transformer.apply(p.key)) != null)
4811	>	action.accept(u);
4812	>	}
4813	>	propagateCompletion();
4814	>	}
4815	>	}
4816	>	}
4817	>
4818	>	@SuppressWarnings("serial")
4819	>	static final class ForEachTransformedValueTask<K,V,U>
4820	>	extends BulkTask<K,V,Void> {
4821	>	final Function<? super V, ? extends U> transformer;
4822	>	final Consumer<? super U> action;
4823	>	ForEachTransformedValueTask
4824	>	(BulkTask<K,V,?> p, int b, int i, int f, Node<K,V>[] t,
4825	>	Function<? super V, ? extends U> transformer, Consumer<? super U> action) {
4826	>	super(p, b, i, f, t);
4827	>	this.transformer = transformer; this.action = action;
4828	>	}
4829	>	public final void compute() {
4830	>	final Function<? super V, ? extends U> transformer;
4831	>	final Consumer<? super U> action;
4832	>	if ((transformer = this.transformer) != null &&
4833	>	(action = this.action) != null) {
4834	>	for (int i = baseIndex, f, h; batch > 0 &&
4835	>	(h = ((f = baseLimit) + i) >>> 1) > i;) {
4836	>	addToPendingCount(1);
4837	>	new ForEachTransformedValueTask<K,V,U>
4838	>	(this, batch >>>= 1, baseLimit = h, f, tab,
4839	>	transformer, action).fork();
4840	>	}
4841	>	for (Node<K,V> p; (p = advance()) != null; ) {
4842	>	U u;
4843	>	if ((u = transformer.apply(p.val)) != null)
4844	>	action.accept(u);
4845	>	}
4846	>	propagateCompletion();
4847	>	}
4848	>	}
4849	>	}
4850	>
4851	>	@SuppressWarnings("serial")
4852	>	static final class ForEachTransformedEntryTask<K,V,U>
4853	>	extends BulkTask<K,V,Void> {
4854	>	final Function<Map.Entry<K,V>, ? extends U> transformer;
4855	>	final Consumer<? super U> action;
4856	>	ForEachTransformedEntryTask
4857	>	(BulkTask<K,V,?> p, int b, int i, int f, Node<K,V>[] t,
4858	>	Function<Map.Entry<K,V>, ? extends U> transformer, Consumer<? super U> action) {
4859	>	super(p, b, i, f, t);
4860	>	this.transformer = transformer; this.action = action;
4861	>	}
4862	>	public final void compute() {
4863	>	final Function<Map.Entry<K,V>, ? extends U> transformer;
4864	>	final Consumer<? super U> action;
4865	>	if ((transformer = this.transformer) != null &&
4866	>	(action = this.action) != null) {
4867	>	for (int i = baseIndex, f, h; batch > 0 &&
4868	>	(h = ((f = baseLimit) + i) >>> 1) > i;) {
4869	>	addToPendingCount(1);
4870	>	new ForEachTransformedEntryTask<K,V,U>
4871	>	(this, batch >>>= 1, baseLimit = h, f, tab,
4872	>	transformer, action).fork();
4873	>	}
4874	>	for (Node<K,V> p; (p = advance()) != null; ) {
4875	>	U u;
4876	>	if ((u = transformer.apply(p)) != null)
4877	>	action.accept(u);
4878	>	}
4879	>	propagateCompletion();
4880	>	}
4881	>	}
4882	>	}
4883	>
4884	>	@SuppressWarnings("serial")
4885	>	static final class ForEachTransformedMappingTask<K,V,U>
4886	>	extends BulkTask<K,V,Void> {
4887	>	final BiFunction<? super K, ? super V, ? extends U> transformer;
4888	>	final Consumer<? super U> action;
4889	>	ForEachTransformedMappingTask
4890	>	(BulkTask<K,V,?> p, int b, int i, int f, Node<K,V>[] t,
4891	>	BiFunction<? super K, ? super V, ? extends U> transformer,
4892	>	Consumer<? super U> action) {
4893	>	super(p, b, i, f, t);
4894	>	this.transformer = transformer; this.action = action;
4895	>	}
4896	>	public final void compute() {
4897	>	final BiFunction<? super K, ? super V, ? extends U> transformer;
4898	>	final Consumer<? super U> action;
4899	>	if ((transformer = this.transformer) != null &&
4900	>	(action = this.action) != null) {
4901	>	for (int i = baseIndex, f, h; batch > 0 &&
4902	>	(h = ((f = baseLimit) + i) >>> 1) > i;) {
4903	>	addToPendingCount(1);
4904	>	new ForEachTransformedMappingTask<K,V,U>
4905	>	(this, batch >>>= 1, baseLimit = h, f, tab,
4906	>	transformer, action).fork();
4907	>	}
4908	>	for (Node<K,V> p; (p = advance()) != null; ) {
4909	>	U u;
4910	>	if ((u = transformer.apply(p.key, p.val)) != null)
4911	>	action.accept(u);
4912	>	}
4913	>	propagateCompletion();
4914	>	}
4915	>	}
4916	>	}
4917	>
4918	>	@SuppressWarnings("serial")
4919	>	static final class SearchKeysTask<K,V,U>
4920	>	extends BulkTask<K,V,U> {
4921	>	final Function<? super K, ? extends U> searchFunction;
4922	>	final AtomicReference<U> result;
4923	>	SearchKeysTask
4924	>	(BulkTask<K,V,?> p, int b, int i, int f, Node<K,V>[] t,
4925	>	Function<? super K, ? extends U> searchFunction,
4926	>	AtomicReference<U> result) {
4927	>	super(p, b, i, f, t);
4928	>	this.searchFunction = searchFunction; this.result = result;
4929	>	}
4930	>	public final U getRawResult() { return result.get(); }
4931	>	public final void compute() {
4932	>	final Function<? super K, ? extends U> searchFunction;
4933	>	final AtomicReference<U> result;
4934	>	if ((searchFunction = this.searchFunction) != null &&
4935	>	(result = this.result) != null) {
4936	>	for (int i = baseIndex, f, h; batch > 0 &&
4937	>	(h = ((f = baseLimit) + i) >>> 1) > i;) {
4938	>	if (result.get() != null)
4939	>	return;
4940	>	addToPendingCount(1);
4941	>	new SearchKeysTask<K,V,U>
4942	>	(this, batch >>>= 1, baseLimit = h, f, tab,
4943	>	searchFunction, result).fork();
4944	>	}
4945	>	while (result.get() == null) {
4946	>	U u;
4947	>	Node<K,V> p;
4948	>	if ((p = advance()) == null) {
4949	>	propagateCompletion();
4950	>	break;
4951	>	}
4952	>	if ((u = searchFunction.apply(p.key)) != null) {
4953	>	if (result.compareAndSet(null, u))
4954	>	quietlyCompleteRoot();
4955	>	break;
4956	>	}
4957	>	}
4958	>	}
4959	>	}
4960	>	}
4961	>
4962	>	@SuppressWarnings("serial")
4963	>	static final class SearchValuesTask<K,V,U>
4964	>	extends BulkTask<K,V,U> {
4965	>	final Function<? super V, ? extends U> searchFunction;
4966	>	final AtomicReference<U> result;
4967	>	SearchValuesTask
4968	>	(BulkTask<K,V,?> p, int b, int i, int f, Node<K,V>[] t,
4969	>	Function<? super V, ? extends U> searchFunction,
4970	>	AtomicReference<U> result) {
4971	>	super(p, b, i, f, t);
4972	>	this.searchFunction = searchFunction; this.result = result;
4973	>	}
4974	>	public final U getRawResult() { return result.get(); }
4975	>	public final void compute() {
4976	>	final Function<? super V, ? extends U> searchFunction;
4977	>	final AtomicReference<U> result;
4978	>	if ((searchFunction = this.searchFunction) != null &&
4979	>	(result = this.result) != null) {
4980	>	for (int i = baseIndex, f, h; batch > 0 &&
4981	>	(h = ((f = baseLimit) + i) >>> 1) > i;) {
4982	>	if (result.get() != null)
4983	>	return;
4984	>	addToPendingCount(1);
4985	>	new SearchValuesTask<K,V,U>
4986	>	(this, batch >>>= 1, baseLimit = h, f, tab,
4987	>	searchFunction, result).fork();
4988	>	}
4989	>	while (result.get() == null) {
4990	>	U u;
4991	>	Node<K,V> p;
4992	>	if ((p = advance()) == null) {
4993	>	propagateCompletion();
4994	>	break;
4995	>	}
4996	>	if ((u = searchFunction.apply(p.val)) != null) {
4997	>	if (result.compareAndSet(null, u))
4998	>	quietlyCompleteRoot();
4999	>	break;
5000	>	}
5001	>	}
5002	>	}
5003	>	}
5004	>	}
5005	>
5006	>	@SuppressWarnings("serial")
5007	>	static final class SearchEntriesTask<K,V,U>
5008	>	extends BulkTask<K,V,U> {
5009	>	final Function<Entry<K,V>, ? extends U> searchFunction;
5010	>	final AtomicReference<U> result;
5011	>	SearchEntriesTask
5012	>	(BulkTask<K,V,?> p, int b, int i, int f, Node<K,V>[] t,
5013	>	Function<Entry<K,V>, ? extends U> searchFunction,
5014	>	AtomicReference<U> result) {
5015	>	super(p, b, i, f, t);
5016	>	this.searchFunction = searchFunction; this.result = result;
5017	>	}
5018	>	public final U getRawResult() { return result.get(); }
5019	>	public final void compute() {
5020	>	final Function<Entry<K,V>, ? extends U> searchFunction;
5021	>	final AtomicReference<U> result;
5022	>	if ((searchFunction = this.searchFunction) != null &&
5023	>	(result = this.result) != null) {
5024	>	for (int i = baseIndex, f, h; batch > 0 &&
5025	>	(h = ((f = baseLimit) + i) >>> 1) > i;) {
5026	>	if (result.get() != null)
5027	>	return;
5028	>	addToPendingCount(1);
5029	>	new SearchEntriesTask<K,V,U>
5030	>	(this, batch >>>= 1, baseLimit = h, f, tab,
5031	>	searchFunction, result).fork();
5032	>	}
5033	>	while (result.get() == null) {
5034	>	U u;
5035	>	Node<K,V> p;
5036	>	if ((p = advance()) == null) {
5037	>	propagateCompletion();
5038	>	break;
5039	>	}
5040	>	if ((u = searchFunction.apply(p)) != null) {
5041	>	if (result.compareAndSet(null, u))
5042	>	quietlyCompleteRoot();
5043	>	return;
5044	>	}
5045	>	}
5046	>	}
5047	>	}
5048	>	}
5049	>
5050	>	@SuppressWarnings("serial")
5051	>	static final class SearchMappingsTask<K,V,U>
5052	>	extends BulkTask<K,V,U> {
5053	>	final BiFunction<? super K, ? super V, ? extends U> searchFunction;
5054	>	final AtomicReference<U> result;
5055	>	SearchMappingsTask
5056	>	(BulkTask<K,V,?> p, int b, int i, int f, Node<K,V>[] t,
5057	>	BiFunction<? super K, ? super V, ? extends U> searchFunction,
5058	>	AtomicReference<U> result) {
5059	>	super(p, b, i, f, t);
5060	>	this.searchFunction = searchFunction; this.result = result;
5061	>	}
5062	>	public final U getRawResult() { return result.get(); }
5063	>	public final void compute() {
5064	>	final BiFunction<? super K, ? super V, ? extends U> searchFunction;
5065	>	final AtomicReference<U> result;
5066	>	if ((searchFunction = this.searchFunction) != null &&
5067	>	(result = this.result) != null) {
5068	>	for (int i = baseIndex, f, h; batch > 0 &&
5069	>	(h = ((f = baseLimit) + i) >>> 1) > i;) {
5070	>	if (result.get() != null)
5071	>	return;
5072	>	addToPendingCount(1);
5073	>	new SearchMappingsTask<K,V,U>
5074	>	(this, batch >>>= 1, baseLimit = h, f, tab,
5075	>	searchFunction, result).fork();
5076	>	}
5077	>	while (result.get() == null) {
5078	>	U u;
5079	>	Node<K,V> p;
5080	>	if ((p = advance()) == null) {
5081	>	propagateCompletion();
5082	>	break;
5083	>	}
5084	>	if ((u = searchFunction.apply(p.key, p.val)) != null) {
5085	>	if (result.compareAndSet(null, u))
5086	>	quietlyCompleteRoot();
5087	>	break;
5088	>	}
5089	>	}
5090	>	}
5091	>	}
5092	>	}
5093	>
5094	>	@SuppressWarnings("serial")
5095	>	static final class ReduceKeysTask<K,V>
5096	>	extends BulkTask<K,V,K> {
5097	>	final BiFunction<? super K, ? super K, ? extends K> reducer;
5098	>	K result;
5099	>	ReduceKeysTask<K,V> rights, nextRight;
5100	>	ReduceKeysTask
5101	>	(BulkTask<K,V,?> p, int b, int i, int f, Node<K,V>[] t,
5102	>	ReduceKeysTask<K,V> nextRight,
5103	>	BiFunction<? super K, ? super K, ? extends K> reducer) {
5104	>	super(p, b, i, f, t); this.nextRight = nextRight;
5105	>	this.reducer = reducer;
5106	>	}
5107	>	public final K getRawResult() { return result; }
5108	>	public final void compute() {
5109	>	final BiFunction<? super K, ? super K, ? extends K> reducer;
5110	>	if ((reducer = this.reducer) != null) {
5111	>	for (int i = baseIndex, f, h; batch > 0 &&
5112	>	(h = ((f = baseLimit) + i) >>> 1) > i;) {
5113	>	addToPendingCount(1);
5114	>	(rights = new ReduceKeysTask<K,V>
5115	>	(this, batch >>>= 1, baseLimit = h, f, tab,
5116	>	rights, reducer)).fork();
5117	>	}
5118	>	K r = null;
5119	>	for (Node<K,V> p; (p = advance()) != null; ) {
5120	>	K u = p.key;
5121	>	r = (r == null) ? u : u == null ? r : reducer.apply(r, u);
5122	>	}
5123	>	result = r;
5124	>	CountedCompleter<?> c;
5125	>	for (c = firstComplete(); c != null; c = c.nextComplete()) {
5126	>	@SuppressWarnings("unchecked") ReduceKeysTask<K,V>
5127	>	t = (ReduceKeysTask<K,V>)c,
5128	>	s = t.rights;
5129	>	while (s != null) {
5130	>	K tr, sr;
5131	>	if ((sr = s.result) != null)
5132	>	t.result = (((tr = t.result) == null) ? sr :
5133	>	reducer.apply(tr, sr));
5134	>	s = t.rights = s.nextRight;
5135	>	}
5136	>	}
5137	>	}
5138	>	}
5139	>	}
5140	>
5141	>	@SuppressWarnings("serial")
5142	>	static final class ReduceValuesTask<K,V>
5143	>	extends BulkTask<K,V,V> {
5144	>	final BiFunction<? super V, ? super V, ? extends V> reducer;
5145	>	V result;
5146	>	ReduceValuesTask<K,V> rights, nextRight;
5147	>	ReduceValuesTask
5148	>	(BulkTask<K,V,?> p, int b, int i, int f, Node<K,V>[] t,
5149	>	ReduceValuesTask<K,V> nextRight,
5150	>	BiFunction<? super V, ? super V, ? extends V> reducer) {
5151	>	super(p, b, i, f, t); this.nextRight = nextRight;
5152	>	this.reducer = reducer;
5153	>	}
5154	>	public final V getRawResult() { return result; }
5155	>	public final void compute() {
5156	>	final BiFunction<? super V, ? super V, ? extends V> reducer;
5157	>	if ((reducer = this.reducer) != null) {
5158	>	for (int i = baseIndex, f, h; batch > 0 &&
5159	>	(h = ((f = baseLimit) + i) >>> 1) > i;) {
5160	>	addToPendingCount(1);
5161	>	(rights = new ReduceValuesTask<K,V>
5162	>	(this, batch >>>= 1, baseLimit = h, f, tab,
5163	>	rights, reducer)).fork();
5164	>	}
5165	>	V r = null;
5166	>	for (Node<K,V> p; (p = advance()) != null; ) {
5167	>	V v = p.val;
5168	>	r = (r == null) ? v : reducer.apply(r, v);
5169	>	}
5170	>	result = r;
5171	>	CountedCompleter<?> c;
5172	>	for (c = firstComplete(); c != null; c = c.nextComplete()) {
5173	>	@SuppressWarnings("unchecked") ReduceValuesTask<K,V>
5174	>	t = (ReduceValuesTask<K,V>)c,
5175	>	s = t.rights;
5176	>	while (s != null) {
5177	>	V tr, sr;
5178	>	if ((sr = s.result) != null)
5179	>	t.result = (((tr = t.result) == null) ? sr :
5180	>	reducer.apply(tr, sr));
5181	>	s = t.rights = s.nextRight;
5182	>	}
5183	>	}
5184	>	}
5185	>	}
5186	>	}
5187	>
5188	>	@SuppressWarnings("serial")
5189	>	static final class ReduceEntriesTask<K,V>
5190	>	extends BulkTask<K,V,Map.Entry<K,V>> {
5191	>	final BiFunction<Map.Entry<K,V>, Map.Entry<K,V>, ? extends Map.Entry<K,V>> reducer;
5192	>	Map.Entry<K,V> result;
5193	>	ReduceEntriesTask<K,V> rights, nextRight;
5194	>	ReduceEntriesTask
5195	>	(BulkTask<K,V,?> p, int b, int i, int f, Node<K,V>[] t,
5196	>	ReduceEntriesTask<K,V> nextRight,
5197	>	BiFunction<Entry<K,V>, Map.Entry<K,V>, ? extends Map.Entry<K,V>> reducer) {
5198	>	super(p, b, i, f, t); this.nextRight = nextRight;
5199	>	this.reducer = reducer;
5200	>	}
5201	>	public final Map.Entry<K,V> getRawResult() { return result; }
5202	>	public final void compute() {
5203	>	final BiFunction<Map.Entry<K,V>, Map.Entry<K,V>, ? extends Map.Entry<K,V>> reducer;
5204	>	if ((reducer = this.reducer) != null) {
5205	>	for (int i = baseIndex, f, h; batch > 0 &&
5206	>	(h = ((f = baseLimit) + i) >>> 1) > i;) {
5207	>	addToPendingCount(1);
5208	>	(rights = new ReduceEntriesTask<K,V>
5209	>	(this, batch >>>= 1, baseLimit = h, f, tab,
5210	>	rights, reducer)).fork();
5211	>	}
5212	>	Map.Entry<K,V> r = null;
5213	>	for (Node<K,V> p; (p = advance()) != null; )
5214	>	r = (r == null) ? p : reducer.apply(r, p);
5215	>	result = r;
5216	>	CountedCompleter<?> c;
5217	>	for (c = firstComplete(); c != null; c = c.nextComplete()) {
5218	>	@SuppressWarnings("unchecked") ReduceEntriesTask<K,V>
5219	>	t = (ReduceEntriesTask<K,V>)c,
5220	>	s = t.rights;
5221	>	while (s != null) {
5222	>	Map.Entry<K,V> tr, sr;
5223	>	if ((sr = s.result) != null)
5224	>	t.result = (((tr = t.result) == null) ? sr :
5225	>	reducer.apply(tr, sr));
5226	>	s = t.rights = s.nextRight;
5227	>	}
5228	>	}
5229	>	}
5230	>	}
5231	>	}
5232	>
5233	>	@SuppressWarnings("serial")
5234	>	static final class MapReduceKeysTask<K,V,U>
5235	>	extends BulkTask<K,V,U> {
5236	>	final Function<? super K, ? extends U> transformer;
5237	>	final BiFunction<? super U, ? super U, ? extends U> reducer;
5238	>	U result;
5239	>	MapReduceKeysTask<K,V,U> rights, nextRight;
5240	>	MapReduceKeysTask
5241	>	(BulkTask<K,V,?> p, int b, int i, int f, Node<K,V>[] t,
5242	>	MapReduceKeysTask<K,V,U> nextRight,
5243	>	Function<? super K, ? extends U> transformer,
5244	>	BiFunction<? super U, ? super U, ? extends U> reducer) {
5245	>	super(p, b, i, f, t); this.nextRight = nextRight;
5246	>	this.transformer = transformer;
5247	>	this.reducer = reducer;
5248	>	}
5249	>	public final U getRawResult() { return result; }
5250	>	public final void compute() {
5251	>	final Function<? super K, ? extends U> transformer;
5252	>	final BiFunction<? super U, ? super U, ? extends U> reducer;
5253	>	if ((transformer = this.transformer) != null &&
5254	>	(reducer = this.reducer) != null) {
5255	>	for (int i = baseIndex, f, h; batch > 0 &&
5256	>	(h = ((f = baseLimit) + i) >>> 1) > i;) {
5257	>	addToPendingCount(1);
5258	>	(rights = new MapReduceKeysTask<K,V,U>
5259	>	(this, batch >>>= 1, baseLimit = h, f, tab,
5260	>	rights, transformer, reducer)).fork();
5261	>	}
5262	>	U r = null;
5263	>	for (Node<K,V> p; (p = advance()) != null; ) {
5264	>	U u;
5265	>	if ((u = transformer.apply(p.key)) != null)
5266	>	r = (r == null) ? u : reducer.apply(r, u);
5267	>	}
5268	>	result = r;
5269	>	CountedCompleter<?> c;
5270	>	for (c = firstComplete(); c != null; c = c.nextComplete()) {
5271	>	@SuppressWarnings("unchecked") MapReduceKeysTask<K,V,U>
5272	>	t = (MapReduceKeysTask<K,V,U>)c,
5273	>	s = t.rights;
5274	>	while (s != null) {
5275	>	U tr, sr;
5276	>	if ((sr = s.result) != null)
5277	>	t.result = (((tr = t.result) == null) ? sr :
5278	>	reducer.apply(tr, sr));
5279	>	s = t.rights = s.nextRight;
5280	>	}
5281	>	}
5282	>	}
5283	>	}
5284	>	}
5285	>
5286	>	@SuppressWarnings("serial")
5287	>	static final class MapReduceValuesTask<K,V,U>
5288	>	extends BulkTask<K,V,U> {
5289	>	final Function<? super V, ? extends U> transformer;
5290	>	final BiFunction<? super U, ? super U, ? extends U> reducer;
5291	>	U result;
5292	>	MapReduceValuesTask<K,V,U> rights, nextRight;
5293	>	MapReduceValuesTask
5294	>	(BulkTask<K,V,?> p, int b, int i, int f, Node<K,V>[] t,
5295	>	MapReduceValuesTask<K,V,U> nextRight,
5296	>	Function<? super V, ? extends U> transformer,
5297	>	BiFunction<? super U, ? super U, ? extends U> reducer) {
5298	>	super(p, b, i, f, t); this.nextRight = nextRight;
5299	>	this.transformer = transformer;
5300	>	this.reducer = reducer;
5301	>	}
5302	>	public final U getRawResult() { return result; }
5303	>	public final void compute() {
5304	>	final Function<? super V, ? extends U> transformer;
5305	>	final BiFunction<? super U, ? super U, ? extends U> reducer;
5306	>	if ((transformer = this.transformer) != null &&
5307	>	(reducer = this.reducer) != null) {
5308	>	for (int i = baseIndex, f, h; batch > 0 &&
5309	>	(h = ((f = baseLimit) + i) >>> 1) > i;) {
5310	>	addToPendingCount(1);
5311	>	(rights = new MapReduceValuesTask<K,V,U>
5312	>	(this, batch >>>= 1, baseLimit = h, f, tab,
5313	>	rights, transformer, reducer)).fork();
5314	>	}
5315	>	U r = null;
5316	>	for (Node<K,V> p; (p = advance()) != null; ) {
5317	>	U u;
5318	>	if ((u = transformer.apply(p.val)) != null)
5319	>	r = (r == null) ? u : reducer.apply(r, u);
5320	>	}
5321	>	result = r;
5322	>	CountedCompleter<?> c;
5323	>	for (c = firstComplete(); c != null; c = c.nextComplete()) {
5324	>	@SuppressWarnings("unchecked") MapReduceValuesTask<K,V,U>
5325	>	t = (MapReduceValuesTask<K,V,U>)c,
5326	>	s = t.rights;
5327	>	while (s != null) {
5328	>	U tr, sr;
5329	>	if ((sr = s.result) != null)
5330	>	t.result = (((tr = t.result) == null) ? sr :
5331	>	reducer.apply(tr, sr));
5332	>	s = t.rights = s.nextRight;
5333	>	}
5334	>	}
5335	>	}
5336	>	}
5337	>	}
5338	>
5339	>	@SuppressWarnings("serial")
5340	>	static final class MapReduceEntriesTask<K,V,U>
5341	>	extends BulkTask<K,V,U> {
5342	>	final Function<Map.Entry<K,V>, ? extends U> transformer;
5343	>	final BiFunction<? super U, ? super U, ? extends U> reducer;
5344	>	U result;
5345	>	MapReduceEntriesTask<K,V,U> rights, nextRight;
5346	>	MapReduceEntriesTask
5347	>	(BulkTask<K,V,?> p, int b, int i, int f, Node<K,V>[] t,
5348	>	MapReduceEntriesTask<K,V,U> nextRight,
5349	>	Function<Map.Entry<K,V>, ? extends U> transformer,
5350	>	BiFunction<? super U, ? super U, ? extends U> reducer) {
5351	>	super(p, b, i, f, t); this.nextRight = nextRight;
5352	>	this.transformer = transformer;
5353	>	this.reducer = reducer;
5354	>	}
5355	>	public final U getRawResult() { return result; }
5356	>	public final void compute() {
5357	>	final Function<Map.Entry<K,V>, ? extends U> transformer;
5358	>	final BiFunction<? super U, ? super U, ? extends U> reducer;
5359	>	if ((transformer = this.transformer) != null &&
5360	>	(reducer = this.reducer) != null) {
5361	>	for (int i = baseIndex, f, h; batch > 0 &&
5362	>	(h = ((f = baseLimit) + i) >>> 1) > i;) {
5363	>	addToPendingCount(1);
5364	>	(rights = new MapReduceEntriesTask<K,V,U>
5365	>	(this, batch >>>= 1, baseLimit = h, f, tab,
5366	>	rights, transformer, reducer)).fork();
5367	>	}
5368	>	U r = null;
5369	>	for (Node<K,V> p; (p = advance()) != null; ) {
5370	>	U u;
5371	>	if ((u = transformer.apply(p)) != null)
5372	>	r = (r == null) ? u : reducer.apply(r, u);
5373	>	}
5374	>	result = r;
5375	>	CountedCompleter<?> c;
5376	>	for (c = firstComplete(); c != null; c = c.nextComplete()) {
5377	>	@SuppressWarnings("unchecked") MapReduceEntriesTask<K,V,U>
5378	>	t = (MapReduceEntriesTask<K,V,U>)c,
5379	>	s = t.rights;
5380	>	while (s != null) {
5381	>	U tr, sr;
5382	>	if ((sr = s.result) != null)
5383	>	t.result = (((tr = t.result) == null) ? sr :
5384	>	reducer.apply(tr, sr));
5385	>	s = t.rights = s.nextRight;
5386	>	}
5387	>	}
5388	>	}
5389	>	}
5390	>	}
5391	>
5392	>	@SuppressWarnings("serial")
5393	>	static final class MapReduceMappingsTask<K,V,U>
5394	>	extends BulkTask<K,V,U> {
5395	>	final BiFunction<? super K, ? super V, ? extends U> transformer;
5396	>	final BiFunction<? super U, ? super U, ? extends U> reducer;
5397	>	U result;
5398	>	MapReduceMappingsTask<K,V,U> rights, nextRight;
5399	>	MapReduceMappingsTask
5400	>	(BulkTask<K,V,?> p, int b, int i, int f, Node<K,V>[] t,
5401	>	MapReduceMappingsTask<K,V,U> nextRight,
5402	>	BiFunction<? super K, ? super V, ? extends U> transformer,
5403	>	BiFunction<? super U, ? super U, ? extends U> reducer) {
5404	>	super(p, b, i, f, t); this.nextRight = nextRight;
5405	>	this.transformer = transformer;
5406	>	this.reducer = reducer;
5407	>	}
5408	>	public final U getRawResult() { return result; }
5409	>	public final void compute() {
5410	>	final BiFunction<? super K, ? super V, ? extends U> transformer;
5411	>	final BiFunction<? super U, ? super U, ? extends U> reducer;
5412	>	if ((transformer = this.transformer) != null &&
5413	>	(reducer = this.reducer) != null) {
5414	>	for (int i = baseIndex, f, h; batch > 0 &&
5415	>	(h = ((f = baseLimit) + i) >>> 1) > i;) {
5416	>	addToPendingCount(1);
5417	>	(rights = new MapReduceMappingsTask<K,V,U>
5418	>	(this, batch >>>= 1, baseLimit = h, f, tab,
5419	>	rights, transformer, reducer)).fork();
5420	>	}
5421	>	U r = null;
5422	>	for (Node<K,V> p; (p = advance()) != null; ) {
5423	>	U u;
5424	>	if ((u = transformer.apply(p.key, p.val)) != null)
5425	>	r = (r == null) ? u : reducer.apply(r, u);
5426	>	}
5427	>	result = r;
5428	>	CountedCompleter<?> c;
5429	>	for (c = firstComplete(); c != null; c = c.nextComplete()) {
5430	>	@SuppressWarnings("unchecked") MapReduceMappingsTask<K,V,U>
5431	>	t = (MapReduceMappingsTask<K,V,U>)c,
5432	>	s = t.rights;
5433	>	while (s != null) {
5434	>	U tr, sr;
5435	>	if ((sr = s.result) != null)
5436	>	t.result = (((tr = t.result) == null) ? sr :
5437	>	reducer.apply(tr, sr));
5438	>	s = t.rights = s.nextRight;
5439	>	}
5440	>	}
5441	>	}
5442	>	}
5443	>	}
5444	>
5445	>	@SuppressWarnings("serial")
5446	>	static final class MapReduceKeysToDoubleTask<K,V>
5447	>	extends BulkTask<K,V,Double> {
5448	>	final ToDoubleFunction<? super K> transformer;
5449	>	final DoubleBinaryOperator reducer;
5450	>	final double basis;
5451	>	double result;
5452	>	MapReduceKeysToDoubleTask<K,V> rights, nextRight;
5453	>	MapReduceKeysToDoubleTask
5454	>	(BulkTask<K,V,?> p, int b, int i, int f, Node<K,V>[] t,
5455	>	MapReduceKeysToDoubleTask<K,V> nextRight,
5456	>	ToDoubleFunction<? super K> transformer,
5457	>	double basis,
5458	>	DoubleBinaryOperator reducer) {
5459	>	super(p, b, i, f, t); this.nextRight = nextRight;
5460	>	this.transformer = transformer;
5461	>	this.basis = basis; this.reducer = reducer;
5462	>	}
5463	>	public final Double getRawResult() { return result; }
5464	>	public final void compute() {
5465	>	final ToDoubleFunction<? super K> transformer;
5466	>	final DoubleBinaryOperator reducer;
5467	>	if ((transformer = this.transformer) != null &&
5468	>	(reducer = this.reducer) != null) {
5469	>	double r = this.basis;
5470	>	for (int i = baseIndex, f, h; batch > 0 &&
5471	>	(h = ((f = baseLimit) + i) >>> 1) > i;) {
5472	>	addToPendingCount(1);
5473	>	(rights = new MapReduceKeysToDoubleTask<K,V>
5474	>	(this, batch >>>= 1, baseLimit = h, f, tab,
5475	>	rights, transformer, r, reducer)).fork();
5476	>	}
5477	>	for (Node<K,V> p; (p = advance()) != null; )
5478	>	r = reducer.applyAsDouble(r, transformer.applyAsDouble(p.key));
5479	>	result = r;
5480	>	CountedCompleter<?> c;
5481	>	for (c = firstComplete(); c != null; c = c.nextComplete()) {
5482	>	@SuppressWarnings("unchecked") MapReduceKeysToDoubleTask<K,V>
5483	>	t = (MapReduceKeysToDoubleTask<K,V>)c,
5484	>	s = t.rights;
5485	>	while (s != null) {
5486	>	t.result = reducer.applyAsDouble(t.result, s.result);
5487	>	s = t.rights = s.nextRight;
5488	>	}
5489	>	}
5490	>	}
5491	>	}
5492	>	}
5493	>
5494	>	@SuppressWarnings("serial")
5495	>	static final class MapReduceValuesToDoubleTask<K,V>
5496	>	extends BulkTask<K,V,Double> {
5497	>	final ToDoubleFunction<? super V> transformer;
5498	>	final DoubleBinaryOperator reducer;
5499	>	final double basis;
5500	>	double result;
5501	>	MapReduceValuesToDoubleTask<K,V> rights, nextRight;
5502	>	MapReduceValuesToDoubleTask
5503	>	(BulkTask<K,V,?> p, int b, int i, int f, Node<K,V>[] t,
5504	>	MapReduceValuesToDoubleTask<K,V> nextRight,
5505	>	ToDoubleFunction<? super V> transformer,
5506	>	double basis,
5507	>	DoubleBinaryOperator reducer) {
5508	>	super(p, b, i, f, t); this.nextRight = nextRight;
5509	>	this.transformer = transformer;
5510	>	this.basis = basis; this.reducer = reducer;
5511	>	}
5512	>	public final Double getRawResult() { return result; }
5513	>	public final void compute() {
5514	>	final ToDoubleFunction<? super V> transformer;
5515	>	final DoubleBinaryOperator reducer;
5516	>	if ((transformer = this.transformer) != null &&
5517	>	(reducer = this.reducer) != null) {
5518	>	double r = this.basis;
5519	>	for (int i = baseIndex, f, h; batch > 0 &&
5520	>	(h = ((f = baseLimit) + i) >>> 1) > i;) {
5521	>	addToPendingCount(1);
5522	>	(rights = new MapReduceValuesToDoubleTask<K,V>
5523	>	(this, batch >>>= 1, baseLimit = h, f, tab,
5524	>	rights, transformer, r, reducer)).fork();
5525	>	}
5526	>	for (Node<K,V> p; (p = advance()) != null; )
5527	>	r = reducer.applyAsDouble(r, transformer.applyAsDouble(p.val));
5528	>	result = r;
5529	>	CountedCompleter<?> c;
5530	>	for (c = firstComplete(); c != null; c = c.nextComplete()) {
5531	>	@SuppressWarnings("unchecked") MapReduceValuesToDoubleTask<K,V>
5532	>	t = (MapReduceValuesToDoubleTask<K,V>)c,
5533	>	s = t.rights;
5534	>	while (s != null) {
5535	>	t.result = reducer.applyAsDouble(t.result, s.result);
5536	>	s = t.rights = s.nextRight;
5537	>	}
5538	>	}
5539	>	}
5540	>	}
5541	>	}
5542	>
5543	>	@SuppressWarnings("serial")
5544	>	static final class MapReduceEntriesToDoubleTask<K,V>
5545	>	extends BulkTask<K,V,Double> {
5546	>	final ToDoubleFunction<Map.Entry<K,V>> transformer;
5547	>	final DoubleBinaryOperator reducer;
5548	>	final double basis;
5549	>	double result;
5550	>	MapReduceEntriesToDoubleTask<K,V> rights, nextRight;
5551	>	MapReduceEntriesToDoubleTask
5552	>	(BulkTask<K,V,?> p, int b, int i, int f, Node<K,V>[] t,
5553	>	MapReduceEntriesToDoubleTask<K,V> nextRight,
5554	>	ToDoubleFunction<Map.Entry<K,V>> transformer,
5555	>	double basis,
5556	>	DoubleBinaryOperator reducer) {
5557	>	super(p, b, i, f, t); this.nextRight = nextRight;
5558	>	this.transformer = transformer;
5559	>	this.basis = basis; this.reducer = reducer;
5560	>	}
5561	>	public final Double getRawResult() { return result; }
5562	>	public final void compute() {
5563	>	final ToDoubleFunction<Map.Entry<K,V>> transformer;
5564	>	final DoubleBinaryOperator reducer;
5565	>	if ((transformer = this.transformer) != null &&
5566	>	(reducer = this.reducer) != null) {
5567	>	double r = this.basis;
5568	>	for (int i = baseIndex, f, h; batch > 0 &&
5569	>	(h = ((f = baseLimit) + i) >>> 1) > i;) {
5570	>	addToPendingCount(1);
5571	>	(rights = new MapReduceEntriesToDoubleTask<K,V>
5572	>	(this, batch >>>= 1, baseLimit = h, f, tab,
5573	>	rights, transformer, r, reducer)).fork();
5574	>	}
5575	>	for (Node<K,V> p; (p = advance()) != null; )
5576	>	r = reducer.applyAsDouble(r, transformer.applyAsDouble(p));
5577	>	result = r;
5578	>	CountedCompleter<?> c;
5579	>	for (c = firstComplete(); c != null; c = c.nextComplete()) {
5580	>	@SuppressWarnings("unchecked") MapReduceEntriesToDoubleTask<K,V>
5581	>	t = (MapReduceEntriesToDoubleTask<K,V>)c,
5582	>	s = t.rights;
5583	>	while (s != null) {
5584	>	t.result = reducer.applyAsDouble(t.result, s.result);
5585	>	s = t.rights = s.nextRight;
5586	>	}
5587	>	}
5588	>	}
5589	>	}
5590	>	}
5591	>
5592	>	@SuppressWarnings("serial")
5593	>	static final class MapReduceMappingsToDoubleTask<K,V>
5594	>	extends BulkTask<K,V,Double> {
5595	>	final ToDoubleBiFunction<? super K, ? super V> transformer;
5596	>	final DoubleBinaryOperator reducer;
5597	>	final double basis;
5598	>	double result;
5599	>	MapReduceMappingsToDoubleTask<K,V> rights, nextRight;
5600	>	MapReduceMappingsToDoubleTask
5601	>	(BulkTask<K,V,?> p, int b, int i, int f, Node<K,V>[] t,
5602	>	MapReduceMappingsToDoubleTask<K,V> nextRight,
5603	>	ToDoubleBiFunction<? super K, ? super V> transformer,
5604	>	double basis,
5605	>	DoubleBinaryOperator reducer) {
5606	>	super(p, b, i, f, t); this.nextRight = nextRight;
5607	>	this.transformer = transformer;
5608	>	this.basis = basis; this.reducer = reducer;
5609	>	}
5610	>	public final Double getRawResult() { return result; }
5611	>	public final void compute() {
5612	>	final ToDoubleBiFunction<? super K, ? super V> transformer;
5613	>	final DoubleBinaryOperator reducer;
5614	>	if ((transformer = this.transformer) != null &&
5615	>	(reducer = this.reducer) != null) {
5616	>	double r = this.basis;
5617	>	for (int i = baseIndex, f, h; batch > 0 &&
5618	>	(h = ((f = baseLimit) + i) >>> 1) > i;) {
5619	>	addToPendingCount(1);
5620	>	(rights = new MapReduceMappingsToDoubleTask<K,V>
5621	>	(this, batch >>>= 1, baseLimit = h, f, tab,
5622	>	rights, transformer, r, reducer)).fork();
5623	>	}
5624	>	for (Node<K,V> p; (p = advance()) != null; )
5625	>	r = reducer.applyAsDouble(r, transformer.applyAsDouble(p.key, p.val));
5626	>	result = r;
5627	>	CountedCompleter<?> c;
5628	>	for (c = firstComplete(); c != null; c = c.nextComplete()) {
5629	>	@SuppressWarnings("unchecked") MapReduceMappingsToDoubleTask<K,V>
5630	>	t = (MapReduceMappingsToDoubleTask<K,V>)c,
5631	>	s = t.rights;
5632	>	while (s != null) {
5633	>	t.result = reducer.applyAsDouble(t.result, s.result);
5634	>	s = t.rights = s.nextRight;
5635	>	}
5636	>	}
5637	>	}
5638	>	}
5639	>	}
5640	>
5641	>	@SuppressWarnings("serial")
5642	>	static final class MapReduceKeysToLongTask<K,V>
5643	>	extends BulkTask<K,V,Long> {
5644	>	final ToLongFunction<? super K> transformer;
5645	>	final LongBinaryOperator reducer;
5646	>	final long basis;
5647	>	long result;
5648	>	MapReduceKeysToLongTask<K,V> rights, nextRight;
5649	>	MapReduceKeysToLongTask
5650	>	(BulkTask<K,V,?> p, int b, int i, int f, Node<K,V>[] t,
5651	>	MapReduceKeysToLongTask<K,V> nextRight,
5652	>	ToLongFunction<? super K> transformer,
5653	>	long basis,
5654	>	LongBinaryOperator reducer) {
5655	>	super(p, b, i, f, t); this.nextRight = nextRight;
5656	>	this.transformer = transformer;
5657	>	this.basis = basis; this.reducer = reducer;
5658	>	}
5659	>	public final Long getRawResult() { return result; }
5660	>	public final void compute() {
5661	>	final ToLongFunction<? super K> transformer;
5662	>	final LongBinaryOperator reducer;
5663	>	if ((transformer = this.transformer) != null &&
5664	>	(reducer = this.reducer) != null) {
5665	>	long r = this.basis;
5666	>	for (int i = baseIndex, f, h; batch > 0 &&
5667	>	(h = ((f = baseLimit) + i) >>> 1) > i;) {
5668	>	addToPendingCount(1);
5669	>	(rights = new MapReduceKeysToLongTask<K,V>
5670	>	(this, batch >>>= 1, baseLimit = h, f, tab,
5671	>	rights, transformer, r, reducer)).fork();
5672	>	}
5673	>	for (Node<K,V> p; (p = advance()) != null; )
5674	>	r = reducer.applyAsLong(r, transformer.applyAsLong(p.key));
5675	>	result = r;
5676	>	CountedCompleter<?> c;
5677	>	for (c = firstComplete(); c != null; c = c.nextComplete()) {
5678	>	@SuppressWarnings("unchecked") MapReduceKeysToLongTask<K,V>
5679	>	t = (MapReduceKeysToLongTask<K,V>)c,
5680	>	s = t.rights;
5681	>	while (s != null) {
5682	>	t.result = reducer.applyAsLong(t.result, s.result);
5683	>	s = t.rights = s.nextRight;
5684	>	}
5685	>	}
5686	>	}
5687	>	}
5688	>	}
5689	>
5690	>	@SuppressWarnings("serial")
5691	>	static final class MapReduceValuesToLongTask<K,V>
5692	>	extends BulkTask<K,V,Long> {
5693	>	final ToLongFunction<? super V> transformer;
5694	>	final LongBinaryOperator reducer;
5695	>	final long basis;
5696	>	long result;
5697	>	MapReduceValuesToLongTask<K,V> rights, nextRight;
5698	>	MapReduceValuesToLongTask
5699	>	(BulkTask<K,V,?> p, int b, int i, int f, Node<K,V>[] t,
5700	>	MapReduceValuesToLongTask<K,V> nextRight,
5701	>	ToLongFunction<? super V> transformer,
5702	>	long basis,
5703	>	LongBinaryOperator reducer) {
5704	>	super(p, b, i, f, t); this.nextRight = nextRight;
5705	>	this.transformer = transformer;
5706	>	this.basis = basis; this.reducer = reducer;
5707	>	}
5708	>	public final Long getRawResult() { return result; }
5709	>	public final void compute() {
5710	>	final ToLongFunction<? super V> transformer;
5711	>	final LongBinaryOperator reducer;
5712	>	if ((transformer = this.transformer) != null &&
5713	>	(reducer = this.reducer) != null) {
5714	>	long r = this.basis;
5715	>	for (int i = baseIndex, f, h; batch > 0 &&
5716	>	(h = ((f = baseLimit) + i) >>> 1) > i;) {
5717	>	addToPendingCount(1);
5718	>	(rights = new MapReduceValuesToLongTask<K,V>
5719	>	(this, batch >>>= 1, baseLimit = h, f, tab,
5720	>	rights, transformer, r, reducer)).fork();
5721	>	}
5722	>	for (Node<K,V> p; (p = advance()) != null; )
5723	>	r = reducer.applyAsLong(r, transformer.applyAsLong(p.val));
5724	>	result = r;
5725	>	CountedCompleter<?> c;
5726	>	for (c = firstComplete(); c != null; c = c.nextComplete()) {
5727	>	@SuppressWarnings("unchecked") MapReduceValuesToLongTask<K,V>
5728	>	t = (MapReduceValuesToLongTask<K,V>)c,
5729	>	s = t.rights;
5730	>	while (s != null) {
5731	>	t.result = reducer.applyAsLong(t.result, s.result);
5732	>	s = t.rights = s.nextRight;
5733	>	}
5734	>	}
5735	>	}
5736	>	}
5737	>	}
5738	>
5739	>	@SuppressWarnings("serial")
5740	>	static final class MapReduceEntriesToLongTask<K,V>
5741	>	extends BulkTask<K,V,Long> {
5742	>	final ToLongFunction<Map.Entry<K,V>> transformer;
5743	>	final LongBinaryOperator reducer;
5744	>	final long basis;
5745	>	long result;
5746	>	MapReduceEntriesToLongTask<K,V> rights, nextRight;
5747	>	MapReduceEntriesToLongTask
5748	>	(BulkTask<K,V,?> p, int b, int i, int f, Node<K,V>[] t,
5749	>	MapReduceEntriesToLongTask<K,V> nextRight,
5750	>	ToLongFunction<Map.Entry<K,V>> transformer,
5751	>	long basis,
5752	>	LongBinaryOperator reducer) {
5753	>	super(p, b, i, f, t); this.nextRight = nextRight;
5754	>	this.transformer = transformer;
5755	>	this.basis = basis; this.reducer = reducer;
5756	>	}
5757	>	public final Long getRawResult() { return result; }
5758	>	public final void compute() {
5759	>	final ToLongFunction<Map.Entry<K,V>> transformer;
5760	>	final LongBinaryOperator reducer;
5761	>	if ((transformer = this.transformer) != null &&
5762	>	(reducer = this.reducer) != null) {
5763	>	long r = this.basis;
5764	>	for (int i = baseIndex, f, h; batch > 0 &&
5765	>	(h = ((f = baseLimit) + i) >>> 1) > i;) {
5766	>	addToPendingCount(1);
5767	>	(rights = new MapReduceEntriesToLongTask<K,V>
5768	>	(this, batch >>>= 1, baseLimit = h, f, tab,
5769	>	rights, transformer, r, reducer)).fork();
5770	>	}
5771	>	for (Node<K,V> p; (p = advance()) != null; )
5772	>	r = reducer.applyAsLong(r, transformer.applyAsLong(p));
5773	>	result = r;
5774	>	CountedCompleter<?> c;
5775	>	for (c = firstComplete(); c != null; c = c.nextComplete()) {
5776	>	@SuppressWarnings("unchecked") MapReduceEntriesToLongTask<K,V>
5777	>	t = (MapReduceEntriesToLongTask<K,V>)c,
5778	>	s = t.rights;
5779	>	while (s != null) {
5780	>	t.result = reducer.applyAsLong(t.result, s.result);
5781	>	s = t.rights = s.nextRight;
5782	>	}
5783	>	}
5784	>	}
5785	>	}
5786	>	}
5787	>
5788	>	@SuppressWarnings("serial")
5789	>	static final class MapReduceMappingsToLongTask<K,V>
5790	>	extends BulkTask<K,V,Long> {
5791	>	final ToLongBiFunction<? super K, ? super V> transformer;
5792	>	final LongBinaryOperator reducer;
5793	>	final long basis;
5794	>	long result;
5795	>	MapReduceMappingsToLongTask<K,V> rights, nextRight;
5796	>	MapReduceMappingsToLongTask
5797	>	(BulkTask<K,V,?> p, int b, int i, int f, Node<K,V>[] t,
5798	>	MapReduceMappingsToLongTask<K,V> nextRight,
5799	>	ToLongBiFunction<? super K, ? super V> transformer,
5800	>	long basis,
5801	>	LongBinaryOperator reducer) {
5802	>	super(p, b, i, f, t); this.nextRight = nextRight;
5803	>	this.transformer = transformer;
5804	>	this.basis = basis; this.reducer = reducer;
5805	>	}
5806	>	public final Long getRawResult() { return result; }
5807	>	public final void compute() {
5808	>	final ToLongBiFunction<? super K, ? super V> transformer;
5809	>	final LongBinaryOperator reducer;
5810	>	if ((transformer = this.transformer) != null &&
5811	>	(reducer = this.reducer) != null) {
5812	>	long r = this.basis;
5813	>	for (int i = baseIndex, f, h; batch > 0 &&
5814	>	(h = ((f = baseLimit) + i) >>> 1) > i;) {
5815	>	addToPendingCount(1);
5816	>	(rights = new MapReduceMappingsToLongTask<K,V>
5817	>	(this, batch >>>= 1, baseLimit = h, f, tab,
5818	>	rights, transformer, r, reducer)).fork();
5819	>	}
5820	>	for (Node<K,V> p; (p = advance()) != null; )
5821	>	r = reducer.applyAsLong(r, transformer.applyAsLong(p.key, p.val));
5822	>	result = r;
5823	>	CountedCompleter<?> c;
5824	>	for (c = firstComplete(); c != null; c = c.nextComplete()) {
5825	>	@SuppressWarnings("unchecked") MapReduceMappingsToLongTask<K,V>
5826	>	t = (MapReduceMappingsToLongTask<K,V>)c,
5827	>	s = t.rights;
5828	>	while (s != null) {
5829	>	t.result = reducer.applyAsLong(t.result, s.result);
5830	>	s = t.rights = s.nextRight;
5831	>	}
5832	>	}
5833	>	}
5834	>	}
5835	>	}
5836	>
5837	>	@SuppressWarnings("serial")
5838	>	static final class MapReduceKeysToIntTask<K,V>
5839	>	extends BulkTask<K,V,Integer> {
5840	>	final ToIntFunction<? super K> transformer;
5841	>	final IntBinaryOperator reducer;
5842	>	final int basis;
5843	>	int result;
5844	>	MapReduceKeysToIntTask<K,V> rights, nextRight;
5845	>	MapReduceKeysToIntTask
5846	>	(BulkTask<K,V,?> p, int b, int i, int f, Node<K,V>[] t,
5847	>	MapReduceKeysToIntTask<K,V> nextRight,
5848	>	ToIntFunction<? super K> transformer,
5849	>	int basis,
5850	>	IntBinaryOperator reducer) {
5851	>	super(p, b, i, f, t); this.nextRight = nextRight;
5852	>	this.transformer = transformer;
5853	>	this.basis = basis; this.reducer = reducer;
5854	>	}
5855	>	public final Integer getRawResult() { return result; }
5856	>	public final void compute() {
5857	>	final ToIntFunction<? super K> transformer;
5858	>	final IntBinaryOperator reducer;
5859	>	if ((transformer = this.transformer) != null &&
5860	>	(reducer = this.reducer) != null) {
5861	>	int r = this.basis;
5862	>	for (int i = baseIndex, f, h; batch > 0 &&
5863	>	(h = ((f = baseLimit) + i) >>> 1) > i;) {
5864	>	addToPendingCount(1);
5865	>	(rights = new MapReduceKeysToIntTask<K,V>
5866	>	(this, batch >>>= 1, baseLimit = h, f, tab,
5867	>	rights, transformer, r, reducer)).fork();
5868	>	}
5869	>	for (Node<K,V> p; (p = advance()) != null; )
5870	>	r = reducer.applyAsInt(r, transformer.applyAsInt(p.key));
5871	>	result = r;
5872	>	CountedCompleter<?> c;
5873	>	for (c = firstComplete(); c != null; c = c.nextComplete()) {
5874	>	@SuppressWarnings("unchecked") MapReduceKeysToIntTask<K,V>
5875	>	t = (MapReduceKeysToIntTask<K,V>)c,
5876	>	s = t.rights;
5877	>	while (s != null) {
5878	>	t.result = reducer.applyAsInt(t.result, s.result);
5879	>	s = t.rights = s.nextRight;
5880	>	}
5881	>	}
5882	>	}
5883	>	}
5884	>	}
5885	>
5886	>	@SuppressWarnings("serial")
5887	>	static final class MapReduceValuesToIntTask<K,V>
5888	>	extends BulkTask<K,V,Integer> {
5889	>	final ToIntFunction<? super V> transformer;
5890	>	final IntBinaryOperator reducer;
5891	>	final int basis;
5892	>	int result;
5893	>	MapReduceValuesToIntTask<K,V> rights, nextRight;
5894	>	MapReduceValuesToIntTask
5895	>	(BulkTask<K,V,?> p, int b, int i, int f, Node<K,V>[] t,
5896	>	MapReduceValuesToIntTask<K,V> nextRight,
5897	>	ToIntFunction<? super V> transformer,
5898	>	int basis,
5899	>	IntBinaryOperator reducer) {
5900	>	super(p, b, i, f, t); this.nextRight = nextRight;
5901	>	this.transformer = transformer;
5902	>	this.basis = basis; this.reducer = reducer;
5903	>	}
5904	>	public final Integer getRawResult() { return result; }
5905	>	public final void compute() {
5906	>	final ToIntFunction<? super V> transformer;
5907	>	final IntBinaryOperator reducer;
5908	>	if ((transformer = this.transformer) != null &&
5909	>	(reducer = this.reducer) != null) {
5910	>	int r = this.basis;
5911	>	for (int i = baseIndex, f, h; batch > 0 &&
5912	>	(h = ((f = baseLimit) + i) >>> 1) > i;) {
5913	>	addToPendingCount(1);
5914	>	(rights = new MapReduceValuesToIntTask<K,V>
5915	>	(this, batch >>>= 1, baseLimit = h, f, tab,
5916	>	rights, transformer, r, reducer)).fork();
5917	>	}
5918	>	for (Node<K,V> p; (p = advance()) != null; )
5919	>	r = reducer.applyAsInt(r, transformer.applyAsInt(p.val));
5920	>	result = r;
5921	>	CountedCompleter<?> c;
5922	>	for (c = firstComplete(); c != null; c = c.nextComplete()) {
5923	>	@SuppressWarnings("unchecked") MapReduceValuesToIntTask<K,V>
5924	>	t = (MapReduceValuesToIntTask<K,V>)c,
5925	>	s = t.rights;
5926	>	while (s != null) {
5927	>	t.result = reducer.applyAsInt(t.result, s.result);
5928	>	s = t.rights = s.nextRight;
5929	>	}
5930	>	}
5931	>	}
5932	>	}
5933	>	}
5934	>
5935	>	@SuppressWarnings("serial")
5936	>	static final class MapReduceEntriesToIntTask<K,V>
5937	>	extends BulkTask<K,V,Integer> {
5938	>	final ToIntFunction<Map.Entry<K,V>> transformer;
5939	>	final IntBinaryOperator reducer;
5940	>	final int basis;
5941	>	int result;
5942	>	MapReduceEntriesToIntTask<K,V> rights, nextRight;
5943	>	MapReduceEntriesToIntTask
5944	>	(BulkTask<K,V,?> p, int b, int i, int f, Node<K,V>[] t,
5945	>	MapReduceEntriesToIntTask<K,V> nextRight,
5946	>	ToIntFunction<Map.Entry<K,V>> transformer,
5947	>	int basis,
5948	>	IntBinaryOperator reducer) {
5949	>	super(p, b, i, f, t); this.nextRight = nextRight;
5950	>	this.transformer = transformer;
5951	>	this.basis = basis; this.reducer = reducer;
5952	>	}
5953	>	public final Integer getRawResult() { return result; }
5954	>	public final void compute() {
5955	>	final ToIntFunction<Map.Entry<K,V>> transformer;
5956	>	final IntBinaryOperator reducer;
5957	>	if ((transformer = this.transformer) != null &&
5958	>	(reducer = this.reducer) != null) {
5959	>	int r = this.basis;
5960	>	for (int i = baseIndex, f, h; batch > 0 &&
5961	>	(h = ((f = baseLimit) + i) >>> 1) > i;) {
5962	>	addToPendingCount(1);
5963	>	(rights = new MapReduceEntriesToIntTask<K,V>
5964	>	(this, batch >>>= 1, baseLimit = h, f, tab,
5965	>	rights, transformer, r, reducer)).fork();
5966	>	}
5967	>	for (Node<K,V> p; (p = advance()) != null; )
5968	>	r = reducer.applyAsInt(r, transformer.applyAsInt(p));
5969	>	result = r;
5970	>	CountedCompleter<?> c;
5971	>	for (c = firstComplete(); c != null; c = c.nextComplete()) {
5972	>	@SuppressWarnings("unchecked") MapReduceEntriesToIntTask<K,V>
5973	>	t = (MapReduceEntriesToIntTask<K,V>)c,
5974	>	s = t.rights;
5975	>	while (s != null) {
5976	>	t.result = reducer.applyAsInt(t.result, s.result);
5977	>	s = t.rights = s.nextRight;
5978	>	}
5979	>	}
5980	>	}
5981	>	}
5982	>	}
5983	>
5984	>	@SuppressWarnings("serial")
5985	>	static final class MapReduceMappingsToIntTask<K,V>
5986	>	extends BulkTask<K,V,Integer> {
5987	>	final ToIntBiFunction<? super K, ? super V> transformer;
5988	>	final IntBinaryOperator reducer;
5989	>	final int basis;
5990	>	int result;
5991	>	MapReduceMappingsToIntTask<K,V> rights, nextRight;
5992	>	MapReduceMappingsToIntTask
5993	>	(BulkTask<K,V,?> p, int b, int i, int f, Node<K,V>[] t,
5994	>	MapReduceMappingsToIntTask<K,V> nextRight,
5995	>	ToIntBiFunction<? super K, ? super V> transformer,
5996	>	int basis,
5997	>	IntBinaryOperator reducer) {
5998	>	super(p, b, i, f, t); this.nextRight = nextRight;
5999	>	this.transformer = transformer;
6000	>	this.basis = basis; this.reducer = reducer;
6001	>	}
6002	>	public final Integer getRawResult() { return result; }
6003	>	public final void compute() {
6004	>	final ToIntBiFunction<? super K, ? super V> transformer;
6005	>	final IntBinaryOperator reducer;
6006	>	if ((transformer = this.transformer) != null &&
6007	>	(reducer = this.reducer) != null) {
6008	>	int r = this.basis;
6009	>	for (int i = baseIndex, f, h; batch > 0 &&
6010	>	(h = ((f = baseLimit) + i) >>> 1) > i;) {
6011	>	addToPendingCount(1);
6012	>	(rights = new MapReduceMappingsToIntTask<K,V>
6013	>	(this, batch >>>= 1, baseLimit = h, f, tab,
6014	>	rights, transformer, r, reducer)).fork();
6015	>	}
6016	>	for (Node<K,V> p; (p = advance()) != null; )
6017	>	r = reducer.applyAsInt(r, transformer.applyAsInt(p.key, p.val));
6018	>	result = r;
6019	>	CountedCompleter<?> c;
6020	>	for (c = firstComplete(); c != null; c = c.nextComplete()) {
6021	>	@SuppressWarnings("unchecked") MapReduceMappingsToIntTask<K,V>
6022	>	t = (MapReduceMappingsToIntTask<K,V>)c,
6023	>	s = t.rights;
6024	>	while (s != null) {
6025	>	t.result = reducer.applyAsInt(t.result, s.result);
6026	>	s = t.rights = s.nextRight;
6027	>	}
6028	>	}
6029	>	}
6030		}
6031		}
6032
6033		// Unsafe mechanics
6034	<	private static final sun.misc.Unsafe UNSAFE;
6035	<	private static final long SBASE;
6036	<	private static final int SSHIFT;
6037	<	private static final long TBASE;
6038	<	private static final int TSHIFT;
6034	>	private static final sun.misc.Unsafe U;
6035	>	private static final long SIZECTL;
6036	>	private static final long TRANSFERINDEX;
6037	>	private static final long TRANSFERORIGIN;
6038	>	private static final long BASECOUNT;
6039	>	private static final long CELLSBUSY;
6040	>	private static final long CELLVALUE;
6041	>	private static final long ABASE;
6042	>	private static final int ASHIFT;
6043
6044		static {
1474	–	int ss, ts;
6045		try {
6046	<	UNSAFE = sun.misc.Unsafe.getUnsafe();
6047	<	Class tc = HashEntry[].class;
6048	<	Class sc = Segment[].class;
6049	<	TBASE = UNSAFE.arrayBaseOffset(tc);
6050	<	SBASE = UNSAFE.arrayBaseOffset(sc);
6051	<	ts = UNSAFE.arrayIndexScale(tc);
6052	<	ss = UNSAFE.arrayIndexScale(sc);
6046	>	U = sun.misc.Unsafe.getUnsafe();
6047	>	Class<?> k = ConcurrentHashMap.class;
6048	>	SIZECTL = U.objectFieldOffset
6049	>	(k.getDeclaredField("sizeCtl"));
6050	>	TRANSFERINDEX = U.objectFieldOffset
6051	>	(k.getDeclaredField("transferIndex"));
6052	>	TRANSFERORIGIN = U.objectFieldOffset
6053	>	(k.getDeclaredField("transferOrigin"));
6054	>	BASECOUNT = U.objectFieldOffset
6055	>	(k.getDeclaredField("baseCount"));
6056	>	CELLSBUSY = U.objectFieldOffset
6057	>	(k.getDeclaredField("cellsBusy"));
6058	>	Class<?> ck = CounterCell.class;
6059	>	CELLVALUE = U.objectFieldOffset
6060	>	(ck.getDeclaredField("value"));
6061	>	Class<?> sc = Node[].class;
6062	>	ABASE = U.arrayBaseOffset(sc);
6063	>	int scale = U.arrayIndexScale(sc);
6064	>	if ((scale & (scale - 1)) != 0)
6065	>	throw new Error("data type scale not a power of two");
6066	>	ASHIFT = 31 - Integer.numberOfLeadingZeros(scale);
6067		} catch (Exception e) {
6068		throw new Error(e);
6069		}
1486	–	if ((ss & (ss-1)) != 0 || (ts & (ts-1)) != 0)
1487	–	throw new Error("data type scale not a power of two");
1488	–	SSHIFT = 31 - Integer.numberOfLeadingZeros(ss);
1489	–	TSHIFT = 31 - Integer.numberOfLeadingZeros(ts);
6070		}
1491	–
6071		}

Diff Legend

–	Removed lines
+	Added lines
<	Changed lines
>	Changed lines