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 1.83 by jsr166, Mon Aug 22 03:42:10 2005 UTC vs.
Revision 1.155 by dl, Thu Dec 27 20:47:47 2012 UTC

#	Line 1 | Line 1
1		/*
2		* Written by Doug Lea with assistance from members of JCP JSR-166
3		* Expert Group and released to the public domain, as explained at
4	<	* http://creativecommons.org/licenses/publicdomain
4	>	* http://creativecommons.org/publicdomain/zero/1.0/
5		*/
6
7		package java.util.concurrent;
8	<	import java.util.concurrent.locks.*;
9	<	import java.util.*;
8	>	import java.util.concurrent.ForkJoinPool;
9	>	import java.util.concurrent.CountedCompleter;
10	>	import java.util.function.*;
11	>	import java.util.stream.Spliterator;
12	>	import java.util.stream.Stream;
13	>	import java.util.stream.Streams;
14	>
15	>	import java.util.Comparator;
16	>	import java.util.Arrays;
17	>	import java.util.Map;
18	>	import java.util.Set;
19	>	import java.util.Collection;
20	>	import java.util.AbstractMap;
21	>	import java.util.AbstractSet;
22	>	import java.util.AbstractCollection;
23	>	import java.util.Hashtable;
24	>	import java.util.HashMap;
25	>	import java.util.Iterator;
26	>	import java.util.Enumeration;
27	>	import java.util.ConcurrentModificationException;
28	>	import java.util.NoSuchElementException;
29	>	import java.util.concurrent.ConcurrentMap;
30	>	import java.util.concurrent.locks.AbstractQueuedSynchronizer;
31	>	import java.util.concurrent.atomic.AtomicInteger;
32	>	import java.util.concurrent.atomic.AtomicReference;
33		import java.io.Serializable;
11	–	import java.io.IOException;
12	–	import java.io.ObjectInputStream;
13	–	import java.io.ObjectOutputStream;
34
35		/**
36		* A hash table supporting full concurrency of retrievals and
37	<	* adjustable expected concurrency for updates. This class obeys the
37	>	* high expected concurrency for updates. This class obeys the
38		* same functional specification as {@link java.util.Hashtable}, and
39		* includes versions of methods corresponding to each method of
40	<	* <tt>Hashtable</tt>. However, even though all operations are
40	>	* {@code Hashtable}. However, even though all operations are
41		* thread-safe, retrieval operations do <em>not</em> entail locking,
42		* and there is <em>not</em> any support for locking the entire table
43		* in a way that prevents all access.  This class is fully
44	<	* interoperable with <tt>Hashtable</tt> in programs that rely on its
44	>	* interoperable with {@code Hashtable} in programs that rely on its
45		* thread safety but not on its synchronization details.
46		*
47	<	* <p> Retrieval operations (including <tt>get</tt>) generally do not
48	<	* block, so may overlap with update operations (including
49	<	* <tt>put</tt> and <tt>remove</tt>). Retrievals reflect the results
50	<	* of the most recently <em>completed</em> update operations holding
51	<	* upon their onset.  For aggregate operations such as <tt>putAll</tt>
52	<	* and <tt>clear</tt>, concurrent retrievals may reflect insertion or
53	<	* removal of only some entries.  Similarly, Iterators and
54	<	* Enumerations return elements reflecting the state of the hash table
55	<	* at some point at or since the creation of the iterator/enumeration.
56	<	* They do <em>not</em> throw {@link ConcurrentModificationException}.
57	<	* However, iterators are designed to be used by only one thread at a time.
58	<	*
59	<	* <p> The allowed concurrency among update operations is guided by
60	<	* the optional <tt>concurrencyLevel</tt> constructor argument
61	<	* (default <tt>16</tt>), which is used as a hint for internal sizing.  The
62	<	* table is internally partitioned to try to permit the indicated
63	<	* number of concurrent updates without contention. Because placement
64	<	* in hash tables is essentially random, the actual concurrency will
65	<	* vary.  Ideally, you should choose a value to accommodate as many
66	<	* threads as will ever concurrently modify the table. Using a
67	<	* significantly higher value than you need can waste space and time,
68	<	* and a significantly lower value can lead to thread contention. But
69	<	* overestimates and underestimates within an order of magnitude do
70	<	* not usually have much noticeable impact. A value of one is
71	<	* appropriate when it is known that only one thread will modify and
72	<	* all others will only read. Also, resizing this or any other kind of
73	<	* hash table is a relatively slow operation, so, when possible, it is
74	<	* a good idea to provide estimates of expected table sizes in
75	<	* constructors.
47	>	* <p>Retrieval operations (including {@code get}) generally do not
48	>	* block, so may overlap with update operations (including {@code put}
49	>	* and {@code remove}). Retrievals reflect the results of the most
50	>	* recently <em>completed</em> update operations holding upon their
51	>	* onset. (More formally, an update operation for a given key bears a
52	>	* <em>happens-before</em> relation with any (non-null) retrieval for
53	>	* that key reporting the updated value.)  For aggregate operations
54	>	* such as {@code putAll} and {@code clear}, concurrent retrievals may
55	>	* reflect insertion or removal of only some entries.  Similarly,
56	>	* Iterators and Enumerations return elements reflecting the state of
57	>	* the hash table at some point at or since the creation of the
58	>	* iterator/enumeration.  They do <em>not</em> throw {@link
59	>	* ConcurrentModificationException}.  However, iterators are designed
60	>	* to be used by only one thread at a time.  Bear in mind that the
61	>	* results of aggregate status methods including {@code size}, {@code
62	>	* isEmpty}, and {@code containsValue} are typically useful only when
63	>	* a map is not undergoing concurrent updates in other threads.
64	>	* Otherwise the results of these methods reflect transient states
65	>	* that may be adequate for monitoring or estimation purposes, but not
66	>	* for program control.
67	>	*
68	>	* <p>The table is dynamically expanded when there are too many
69	>	* collisions (i.e., keys that have distinct hash codes but fall into
70	>	* the same slot modulo the table size), with the expected average
71	>	* effect of maintaining roughly two bins per mapping (corresponding
72	>	* to a 0.75 load factor threshold for resizing). There may be much
73	>	* variance around this average as mappings are added and removed, but
74	>	* overall, this maintains a commonly accepted time/space tradeoff for
75	>	* hash tables.  However, resizing this or any other kind of hash
76	>	* table may be a relatively slow operation. When possible, it is a
77	>	* good idea to provide a size estimate as an optional {@code
78	>	* initialCapacity} constructor argument. An additional optional
79	>	* {@code loadFactor} constructor argument provides a further means of
80	>	* customizing initial table capacity by specifying the table density
81	>	* to be used in calculating the amount of space to allocate for the
82	>	* given number of elements.  Also, for compatibility with previous
83	>	* versions of this class, constructors may optionally specify an
84	>	* expected {@code concurrencyLevel} as an additional hint for
85	>	* internal sizing.  Note that using many keys with exactly the same
86	>	* {@code hashCode()} is a sure way to slow down performance of any
87	>	* hash table.
88	>	*
89	>	* <p>A {@link Set} projection of a ConcurrentHashMap may be created
90	>	* (using {@link #newKeySet()} or {@link #newKeySet(int)}), or viewed
91	>	* (using {@link #keySet(Object)} when only keys are of interest, and the
92	>	* mapped values are (perhaps transiently) not used or all take the
93	>	* same mapping value.
94	>	*
95	>	* <p>A ConcurrentHashMap can be used as scalable frequency map (a
96	>	* form of histogram or multiset) by using {@link
97	>	* java.util.concurrent.atomic.LongAdder} values and initializing via
98	>	* {@link #computeIfAbsent}. For example, to add a count to a {@code
99	>	* ConcurrentHashMap<String,LongAdder> freqs}, you can use {@code
100	>	* freqs.computeIfAbsent(k -> new LongAdder()).increment();}
101		*
102		* <p>This class and its views and iterators implement all of the
103		* <em>optional</em> methods of the {@link Map} and {@link Iterator}
104		* interfaces.
105		*
106	<	* <p> Like {@link Hashtable} but unlike {@link HashMap}, this class
107	<	* does <em>not</em> allow <tt>null</tt> to be used as a key or value.
106	>	* <p>Like {@link Hashtable} but unlike {@link HashMap}, this class
107	>	* does <em>not</em> allow {@code null} to be used as a key or value.
108	>	*
109	>	* <p>ConcurrentHashMaps support sequential and parallel operations
110	>	* bulk operations. (Parallel forms use the {@link
111	>	* ForkJoinPool#commonPool()}). Tasks that may be used in other
112	>	* contexts are available in class {@link ForkJoinTasks}. These
113	>	* operations are designed to be safely, and often sensibly, applied
114	>	* even with maps that are being concurrently updated by other
115	>	* threads; for example, when computing a snapshot summary of the
116	>	* values in a shared registry.  There are three kinds of operation,
117	>	* each with four forms, accepting functions with Keys, Values,
118	>	* Entries, and (Key, Value) arguments and/or return values. Because
119	>	* the elements of a ConcurrentHashMap are not ordered in any
120	>	* particular way, and may be processed in different orders in
121	>	* different parallel executions, the correctness of supplied
122	>	* functions should not depend on any ordering, or on any other
123	>	* objects or values that may transiently change while computation is
124	>	* in progress; and except for forEach actions, should ideally be
125	>	* side-effect-free.
126	>	*
127	>	* <ul>
128	>	* <li> forEach: Perform a given action on each element.
129	>	* A variant form applies a given transformation on each element
130	>	* before performing the action.</li>
131	>	*
132	>	* <li> search: Return the first available non-null result of
133	>	* applying a given function on each element; skipping further
134	>	* search when a result is found.</li>
135	>	*
136	>	* <li> reduce: Accumulate each element.  The supplied reduction
137	>	* function cannot rely on ordering (more formally, it should be
138	>	* both associative and commutative).  There are five variants:
139	>	*
140	>	* <ul>
141	>	*
142	>	* <li> Plain reductions. (There is not a form of this method for
143	>	* (key, value) function arguments since there is no corresponding
144	>	* return type.)</li>
145	>	*
146	>	* <li> Mapped reductions that accumulate the results of a given
147	>	* function applied to each element.</li>
148	>	*
149	>	* <li> Reductions to scalar doubles, longs, and ints, using a
150	>	* given basis value.</li>
151	>	*
152	>	* </li>
153	>	* </ul>
154	>	* </ul>
155	>	*
156	>	* <p>The concurrency properties of bulk operations follow
157	>	* from those of ConcurrentHashMap: Any non-null result returned
158	>	* from {@code get(key)} and related access methods bears a
159	>	* happens-before relation with the associated insertion or
160	>	* update.  The result of any bulk operation reflects the
161	>	* composition of these per-element relations (but is not
162	>	* necessarily atomic with respect to the map as a whole unless it
163	>	* is somehow known to be quiescent).  Conversely, because keys
164	>	* and values in the map are never null, null serves as a reliable
165	>	* atomic indicator of the current lack of any result.  To
166	>	* maintain this property, null serves as an implicit basis for
167	>	* all non-scalar reduction operations. For the double, long, and
168	>	* int versions, the basis should be one that, when combined with
169	>	* any other value, returns that other value (more formally, it
170	>	* should be the identity element for the reduction). Most common
171	>	* reductions have these properties; for example, computing a sum
172	>	* with basis 0 or a minimum with basis MAX_VALUE.
173	>	*
174	>	* <p>Search and transformation functions provided as arguments
175	>	* should similarly return null to indicate the lack of any result
176	>	* (in which case it is not used). In the case of mapped
177	>	* reductions, this also enables transformations to serve as
178	>	* filters, returning null (or, in the case of primitive
179	>	* specializations, the identity basis) if the element should not
180	>	* be combined. You can create compound transformations and
181	>	* filterings by composing them yourself under this "null means
182	>	* there is nothing there now" rule before using them in search or
183	>	* reduce operations.
184	>	*
185	>	* <p>Methods accepting and/or returning Entry arguments maintain
186	>	* key-value associations. They may be useful for example when
187	>	* finding the key for the greatest value. Note that "plain" Entry
188	>	* arguments can be supplied using {@code new
189	>	* AbstractMap.SimpleEntry(k,v)}.
190	>	*
191	>	* <p>Bulk operations may complete abruptly, throwing an
192	>	* exception encountered in the application of a supplied
193	>	* function. Bear in mind when handling such exceptions that other
194	>	* concurrently executing functions could also have thrown
195	>	* exceptions, or would have done so if the first exception had
196	>	* not occurred.
197	>	*
198	>	* <p>Speedups for parallel compared to sequential forms are common
199	>	* but not guaranteed.  Parallel operations involving brief functions
200	>	* on small maps may execute more slowly than sequential forms if the
201	>	* underlying work to parallelize the computation is more expensive
202	>	* than the computation itself.  Similarly, parallelization may not
203	>	* lead to much actual parallelism if all processors are busy
204	>	* performing unrelated tasks.
205	>	*
206	>	* <p>All arguments to all task methods must be non-null.
207		*
208		* <p>This class is a member of the
209	<	* <a href="{@docRoot}/../guide/collections/index.html">
209	>	* <a href="{@docRoot}/../technotes/guides/collections/index.html">
210		* Java Collections Framework</a>.
211		*
212		* @since 1.5
#	Line 70 | Line 214 | import java.io.ObjectOutputStream;
214		* @param <K> the type of keys maintained by this map
215		* @param <V> the type of mapped values
216		*/
217	<	public class ConcurrentHashMap<K, V> extends AbstractMap<K, V>
218	<	implements ConcurrentMap<K, V>, Serializable {
217	>	public class ConcurrentHashMap<K, V>
218	>	implements ConcurrentMap<K, V>, Serializable {
219		private static final long serialVersionUID = 7249069246763182397L;
220
221		/*
222	<	* The basic strategy is to subdivide the table among Segments,
223	<	* each of which itself is a concurrently readable hash table.
222	>	* Overview:
223	>	*
224	>	* The primary design goal of this hash table is to maintain
225	>	* concurrent readability (typically method get(), but also
226	>	* iterators and related methods) while minimizing update
227	>	* contention. Secondary goals are to keep space consumption about
228	>	* the same or better than java.util.HashMap, and to support high
229	>	* initial insertion rates on an empty table by many threads.
230	>	*
231	>	* Each key-value mapping is held in a Node.  Because Node key
232	>	* fields can contain special values, they are defined using plain
233	>	* Object types (not type "K"). This leads to a lot of explicit
234	>	* casting (and many explicit warning suppressions to tell
235	>	* compilers not to complain about it). It also allows some of the
236	>	* public methods to be factored into a smaller number of internal
237	>	* methods (although sadly not so for the five variants of
238	>	* put-related operations). The validation-based approach
239	>	* explained below leads to a lot of code sprawl because
240	>	* retry-control precludes factoring into smaller methods.
241	>	*
242	>	* The table is lazily initialized to a power-of-two size upon the
243	>	* first insertion.  Each bin in the table normally contains a
244	>	* list of Nodes (most often, the list has only zero or one Node).
245	>	* Table accesses require volatile/atomic reads, writes, and
246	>	* CASes.  Because there is no other way to arrange this without
247	>	* adding further indirections, we use intrinsics
248	>	* (sun.misc.Unsafe) operations.  The lists of nodes within bins
249	>	* are always accurately traversable under volatile reads, so long
250	>	* as lookups check hash code and non-nullness of value before
251	>	* checking key equality.
252	>	*
253	>	* We use the top (sign) bit of Node hash fields for control
254	>	* purposes -- it is available anyway because of addressing
255	>	* constraints.  Nodes with negative hash fields are forwarding
256	>	* nodes to either TreeBins or resized tables.  The lower 31 bits
257	>	* of each normal Node's hash field contain a transformation of
258	>	* the key's hash code.
259	>	*
260	>	* Insertion (via put or its variants) of the first node in an
261	>	* empty bin is performed by just CASing it to the bin.  This is
262	>	* by far the most common case for put operations under most
263	>	* key/hash distributions.  Other update operations (insert,
264	>	* delete, and replace) require locks.  We do not want to waste
265	>	* the space required to associate a distinct lock object with
266	>	* each bin, so instead use the first node of a bin list itself as
267	>	* a lock. Locking support for these locks relies on builtin
268	>	* "synchronized" monitors.
269	>	*
270	>	* Using the first node of a list as a lock does not by itself
271	>	* suffice though: When a node is locked, any update must first
272	>	* validate that it is still the first node after locking it, and
273	>	* retry if not. Because new nodes are always appended to lists,
274	>	* once a node is first in a bin, it remains first until deleted
275	>	* or the bin becomes invalidated (upon resizing).  However,
276	>	* operations that only conditionally update may inspect nodes
277	>	* until the point of update. This is a converse of sorts to the
278	>	* lazy locking technique described by Herlihy & Shavit.
279	>	*
280	>	* The main disadvantage of per-bin locks is that other update
281	>	* operations on other nodes in a bin list protected by the same
282	>	* lock can stall, for example when user equals() or mapping
283	>	* functions take a long time.  However, statistically, under
284	>	* random hash codes, this is not a common problem.  Ideally, the
285	>	* frequency of nodes in bins follows a Poisson distribution
286	>	* (http://en.wikipedia.org/wiki/Poisson_distribution) with a
287	>	* parameter of about 0.5 on average, given the resizing threshold
288	>	* of 0.75, although with a large variance because of resizing
289	>	* granularity. Ignoring variance, the expected occurrences of
290	>	* list size k are (exp(-0.5) * pow(0.5, k) / factorial(k)). The
291	>	* first values are:
292	>	*
293	>	* 0:    0.60653066
294	>	* 1:    0.30326533
295	>	* 2:    0.07581633
296	>	* 3:    0.01263606
297	>	* 4:    0.00157952
298	>	* 5:    0.00015795
299	>	* 6:    0.00001316
300	>	* 7:    0.00000094
301	>	* 8:    0.00000006
302	>	* more: less than 1 in ten million
303	>	*
304	>	* Lock contention probability for two threads accessing distinct
305	>	* elements is roughly 1 / (8 * #elements) under random hashes.
306	>	*
307	>	* Actual hash code distributions encountered in practice
308	>	* sometimes deviate significantly from uniform randomness.  This
309	>	* includes the case when N > (1<<30), so some keys MUST collide.
310	>	* Similarly for dumb or hostile usages in which multiple keys are
311	>	* designed to have identical hash codes. Also, although we guard
312	>	* against the worst effects of this (see method spread), sets of
313	>	* hashes may differ only in bits that do not impact their bin
314	>	* index for a given power-of-two mask.  So we use a secondary
315	>	* strategy that applies when the number of nodes in a bin exceeds
316	>	* a threshold, and at least one of the keys implements
317	>	* Comparable.  These TreeBins use a balanced tree to hold nodes
318	>	* (a specialized form of red-black trees), bounding search time
319	>	* to O(log N).  Each search step in a TreeBin is around twice as
320	>	* slow as in a regular list, but given that N cannot exceed
321	>	* (1<<64) (before running out of addresses) this bounds search
322	>	* steps, lock hold times, etc, to reasonable constants (roughly
323	>	* 100 nodes inspected per operation worst case) so long as keys
324	>	* are Comparable (which is very common -- String, Long, etc).
325	>	* TreeBin nodes (TreeNodes) also maintain the same "next"
326	>	* traversal pointers as regular nodes, so can be traversed in
327	>	* iterators in the same way.
328	>	*
329	>	* The table is resized when occupancy exceeds a percentage
330	>	* threshold (nominally, 0.75, but see below).  Any thread
331	>	* noticing an overfull bin may assist in resizing after the
332	>	* initiating thread allocates and sets up the replacement
333	>	* array. However, rather than stalling, these other threads may
334	>	* proceed with insertions etc.  The use of TreeBins shields us
335	>	* from the worst case effects of overfilling while resizes are in
336	>	* progress.  Resizing proceeds by transferring bins, one by one,
337	>	* from the table to the next table. To enable concurrency, the
338	>	* next table must be (incrementally) prefilled with place-holders
339	>	* serving as reverse forwarders to the old table.  Because we are
340	>	* using power-of-two expansion, the elements from each bin must
341	>	* either stay at same index, or move with a power of two
342	>	* offset. We eliminate unnecessary node creation by catching
343	>	* cases where old nodes can be reused because their next fields
344	>	* won't change.  On average, only about one-sixth of them need
345	>	* cloning when a table doubles. The nodes they replace will be
346	>	* garbage collectable as soon as they are no longer referenced by
347	>	* any reader thread that may be in the midst of concurrently
348	>	* traversing table.  Upon transfer, the old table bin contains
349	>	* only a special forwarding node (with hash field "MOVED") that
350	>	* contains the next table as its key. On encountering a
351	>	* forwarding node, access and update operations restart, using
352	>	* the new table.
353	>	*
354	>	* Each bin transfer requires its bin lock, which can stall
355	>	* waiting for locks while resizing. However, because other
356	>	* threads can join in and help resize rather than contend for
357	>	* locks, average aggregate waits become shorter as resizing
358	>	* progresses.  The transfer operation must also ensure that all
359	>	* accessible bins in both the old and new table are usable by any
360	>	* traversal.  This is arranged by proceeding from the last bin
361	>	* (table.length - 1) up towards the first.  Upon seeing a
362	>	* forwarding node, traversals (see class Traverser) arrange to
363	>	* move to the new table without revisiting nodes.  However, to
364	>	* ensure that no intervening nodes are skipped, bin splitting can
365	>	* only begin after the associated reverse-forwarders are in
366	>	* place.
367	>	*
368	>	* The traversal scheme also applies to partial traversals of
369	>	* ranges of bins (via an alternate Traverser constructor)
370	>	* to support partitioned aggregate operations.  Also, read-only
371	>	* operations give up if ever forwarded to a null table, which
372	>	* provides support for shutdown-style clearing, which is also not
373	>	* currently implemented.
374	>	*
375	>	* Lazy table initialization minimizes footprint until first use,
376	>	* and also avoids resizings when the first operation is from a
377	>	* putAll, constructor with map argument, or deserialization.
378	>	* These cases attempt to override the initial capacity settings,
379	>	* but harmlessly fail to take effect in cases of races.
380	>	*
381	>	* The element count is maintained using a specialization of
382	>	* LongAdder. We need to incorporate a specialization rather than
383	>	* just use a LongAdder in order to access implicit
384	>	* contention-sensing that leads to creation of multiple
385	>	* Cells.  The counter mechanics avoid contention on
386	>	* updates but can encounter cache thrashing if read too
387	>	* frequently during concurrent access. To avoid reading so often,
388	>	* resizing under contention is attempted only upon adding to a
389	>	* bin already holding two or more nodes. Under uniform hash
390	>	* distributions, the probability of this occurring at threshold
391	>	* is around 13%, meaning that only about 1 in 8 puts check
392	>	* threshold (and after resizing, many fewer do so). The bulk
393	>	* putAll operation further reduces contention by only committing
394	>	* count updates upon these size checks.
395	>	*
396	>	* Maintaining API and serialization compatibility with previous
397	>	* versions of this class introduces several oddities. Mainly: We
398	>	* leave untouched but unused constructor arguments refering to
399	>	* concurrencyLevel. We accept a loadFactor constructor argument,
400	>	* but apply it only to initial table capacity (which is the only
401	>	* time that we can guarantee to honor it.) We also declare an
402	>	* unused "Segment" class that is instantiated in minimal form
403	>	* only when serializing.
404		*/
405
406		/* ---------------- Constants -------------- */
407
408		/**
409	<	* The default initial capacity for this table,
410	<	* used when not otherwise specified in a constructor.
409	>	* The largest possible table capacity.  This value must be
410	>	* exactly 1<<30 to stay within Java array allocation and indexing
411	>	* bounds for power of two table sizes, and is further required
412	>	* because the top two bits of 32bit hash fields are used for
413	>	* control purposes.
414		*/
415	<	static final int DEFAULT_INITIAL_CAPACITY = 16;
415	>	private static final int MAXIMUM_CAPACITY = 1 << 30;
416
417		/**
418	<	* The default load factor for this table, used when not
419	<	* otherwise specified in a constructor.
418	>	* The default initial table capacity.  Must be a power of 2
419	>	* (i.e., at least 1) and at most MAXIMUM_CAPACITY.
420		*/
421	<	static final float DEFAULT_LOAD_FACTOR = 0.75f;
421	>	private static final int DEFAULT_CAPACITY = 16;
422
423		/**
424	<	* The default concurrency level for this table, used when not
425	<	* otherwise specified in a constructor.
424	>	* The largest possible (non-power of two) array size.
425	>	* Needed by toArray and related methods.
426		*/
427	<	static final int DEFAULT_CONCURRENCY_LEVEL = 16;
427	>	static final int MAX_ARRAY_SIZE = Integer.MAX_VALUE - 8;
428
429		/**
430	<	* The maximum capacity, used if a higher value is implicitly
431	<	* specified by either of the constructors with arguments.  MUST
105	<	* be a power of two <= 1<<30 to ensure that entries are indexable
106	<	* using ints.
430	>	* The default concurrency level for this table. Unused but
431	>	* defined for compatibility with previous versions of this class.
432		*/
433	<	static final int MAXIMUM_CAPACITY = 1 << 30;
433	>	private static final int DEFAULT_CONCURRENCY_LEVEL = 16;
434
435		/**
436	<	* The maximum number of segments to allow; used to bound
437	<	* constructor arguments.
436	>	* The load factor for this table. Overrides of this value in
437	>	* constructors affect only the initial table capacity.  The
438	>	* actual floating point value isn't normally used -- it is
439	>	* simpler to use expressions such as {@code n - (n >>> 2)} for
440	>	* the associated resizing threshold.
441		*/
442	<	static final int MAX_SEGMENTS = 1 << 16; // slightly conservative
442	>	private static final float LOAD_FACTOR = 0.75f;
443
444		/**
445	<	* Number of unsynchronized retries in size and containsValue
446	<	* methods before resorting to locking. This is used to avoid
447	<	* unbounded retries if tables undergo continuous modification
120	<	* which would make it impossible to obtain an accurate result.
445	>	* The bin count threshold for using a tree rather than list for a
446	>	* bin.  The value reflects the approximate break-even point for
447	>	* using tree-based operations.
448		*/
449	<	static final int RETRIES_BEFORE_LOCK = 2;
449	>	private static final int TREE_THRESHOLD = 8;
450	>
451	>	/**
452	>	* Minimum number of rebinnings per transfer step. Ranges are
453	>	* subdivided to allow multiple resizer threads.  This value
454	>	* serves as a lower bound to avoid resizers encountering
455	>	* excessive memory contention.  The value should be at least
456	>	* DEFAULT_CAPACITY.
457	>	*/
458	>	private static final int MIN_TRANSFER_STRIDE = 16;
459	>
460	>	/*
461	>	* Encodings for Node hash fields. See above for explanation.
462	>	*/
463	>	static final int MOVED     = 0x80000000; // hash field for forwarding nodes
464	>	static final int HASH_BITS = 0x7fffffff; // usable bits of normal node hash
465	>
466	>	/** Number of CPUS, to place bounds on some sizings */
467	>	static final int NCPU = Runtime.getRuntime().availableProcessors();
468	>
469	>	/* ---------------- Counters -------------- */
470	>
471	>	// Adapted from LongAdder and Striped64.
472	>	// See their internal docs for explanation.
473	>
474	>	// A padded cell for distributing counts
475	>	static final class Cell {
476	>	volatile long p0, p1, p2, p3, p4, p5, p6;
477	>	volatile long value;
478	>	volatile long q0, q1, q2, q3, q4, q5, q6;
479	>	Cell(long x) { value = x; }
480	>	}
481	>
482	>	/**
483	>	* Holder for the thread-local hash code determining which
484	>	* Cell to use. The code is initialized via the
485	>	* cellHashCodeGenerator, but may be moved upon collisions.
486	>	*/
487	>	static final class CellHashCode {
488	>	int code;
489	>	}
490	>
491	>	/**
492	>	* Generates initial value for per-thread CellHashCodes
493	>	*/
494	>	static final AtomicInteger cellHashCodeGenerator = new AtomicInteger();
495	>
496	>	/**
497	>	* Increment for cellHashCodeGenerator. See class ThreadLocal
498	>	* for explanation.
499	>	*/
500	>	static final int SEED_INCREMENT = 0x61c88647;
501	>
502	>	/**
503	>	* Per-thread counter hash codes. Shared across all instances
504	>	*/
505	>	static final ThreadLocal<CellHashCode> threadCellHashCode =
506	>	new ThreadLocal<CellHashCode>();
507
508		/* ---------------- Fields -------------- */
509
510		/**
511	<	* Mask value for indexing into segments. The upper bits of a
512	<	* key's hash code are used to choose the segment.
511	>	* The array of bins. Lazily initialized upon first insertion.
512	>	* Size is always a power of two. Accessed directly by iterators.
513		*/
514	<	final int segmentMask;
514	>	transient volatile Node<V>[] table;
515
516		/**
517	<	* Shift value for indexing within segments.
517	>	* The next table to use; non-null only while resizing.
518		*/
519	<	final int segmentShift;
519	>	private transient volatile Node<V>[] nextTable;
520
521		/**
522	<	* The segments, each of which is a specialized hash table
522	>	* Base counter value, used mainly when there is no contention,
523	>	* but also as a fallback during table initialization
524	>	* races. Updated via CAS.
525		*/
526	<	final Segment<K,V>[] segments;
526	>	private transient volatile long baseCount;
527
528	<	transient Set<K> keySet;
529	<	transient Set<Map.Entry<K,V>> entrySet;
530	<	transient Collection<V> values;
528	>	/**
529	>	* Table initialization and resizing control.  When negative, the
530	>	* table is being initialized or resized: -1 for initialization,
531	>	* else -(1 + the number of active resizing threads).  Otherwise,
532	>	* when table is null, holds the initial table size to use upon
533	>	* creation, or 0 for default. After initialization, holds the
534	>	* next element count value upon which to resize the table.
535	>	*/
536	>	private transient volatile int sizeCtl;
537
538	<	/* ---------------- Small Utilities -------------- */
538	>	/**
539	>	* The next table index (plus one) to split while resizing.
540	>	*/
541	>	private transient volatile int transferIndex;
542
543		/**
544	<	* Returns a hash code for non-null Object x.
150	<	* Uses the same hash code spreader as most other java.util hash tables.
151	<	* @param x the object serving as a key
152	<	* @return the hash code
544	>	* The least available table index to split while resizing.
545		*/
546	<	static int hash(Object x) {
547	<	int h = x.hashCode();
548	<	h += ~(h << 9);
549	<	h ^=  (h >>> 14);
550	<	h +=  (h << 4);
551	<	h ^=  (h >>> 10);
160	<	return h;
161	<	}
546	>	private transient volatile int transferOrigin;
547	>
548	>	/**
549	>	* Spinlock (locked via CAS) used when resizing and/or creating Cells.
550	>	*/
551	>	private transient volatile int cellsBusy;
552
553		/**
554	<	* Returns the segment that should be used for key with given hash
165	<	* @param hash the hash code for the key
166	<	* @return the segment
554	>	* Table of counter cells. When non-null, size is a power of 2.
555		*/
556	<	final Segment<K,V> segmentFor(int hash) {
557	<	return segments[(hash >>> segmentShift) & segmentMask];
556	>	private transient volatile Cell[] counterCells;
557	>
558	>	// views
559	>	private transient KeySetView<K,V> keySet;
560	>	private transient ValuesView<K,V> values;
561	>	private transient EntrySetView<K,V> entrySet;
562	>
563	>	/** For serialization compatibility. Null unless serialized; see below */
564	>	private Segment<K,V>[] segments;
565	>
566	>	/* ---------------- Table element access -------------- */
567	>
568	>	/*
569	>	* Volatile access methods are used for table elements as well as
570	>	* elements of in-progress next table while resizing.  Uses are
571	>	* null checked by callers, and implicitly bounds-checked, relying
572	>	* on the invariants that tab arrays have non-zero size, and all
573	>	* indices are masked with (tab.length - 1) which is never
574	>	* negative and always less than length. Note that, to be correct
575	>	* wrt arbitrary concurrency errors by users, bounds checks must
576	>	* operate on local variables, which accounts for some odd-looking
577	>	* inline assignments below.
578	>	*/
579	>
580	>	@SuppressWarnings("unchecked") static final <V> Node<V> tabAt
581	>	(Node<V>[] tab, int i) { // used by Traverser
582	>	return (Node<V>)U.getObjectVolatile(tab, ((long)i << ASHIFT) + ABASE);
583	>	}
584	>
585	>	private static final <V> boolean casTabAt
586	>	(Node<V>[] tab, int i, Node<V> c, Node<V> v) {
587	>	return U.compareAndSwapObject(tab, ((long)i << ASHIFT) + ABASE, c, v);
588	>	}
589	>
590	>	private static final <V> void setTabAt
591	>	(Node<V>[] tab, int i, Node<V> v) {
592	>	U.putObjectVolatile(tab, ((long)i << ASHIFT) + ABASE, v);
593		}
594
595	<	/* ---------------- Inner Classes -------------- */
595	>	/* ---------------- Nodes -------------- */
596
597		/**
598	<	* ConcurrentHashMap list entry. Note that this is never exported
599	<	* out as a user-visible Map.Entry.
600	<	*
601	<	* Because the value field is volatile, not final, it is legal wrt
602	<	* the Java Memory Model for an unsynchronized reader to see null
603	<	* instead of initial value when read via a data race.  Although a
604	<	* reordering leading to this is not likely to ever actually
605	<	* occur, the Segment.readValueUnderLock method is used as a
183	<	* backup in case a null (pre-initialized) value is ever seen in
184	<	* an unsynchronized access method.
598	>	* Key-value entry. Note that this is never exported out as a
599	>	* user-visible Map.Entry (see MapEntry below). Nodes with a hash
600	>	* field of MOVED are special, and do not contain user keys or
601	>	* values.  Otherwise, keys are never null, and null val fields
602	>	* indicate that a node is in the process of being deleted or
603	>	* created. For purposes of read-only access, a key may be read
604	>	* before a val, but can only be used after checking val to be
605	>	* non-null.
606		*/
607	<	static final class HashEntry<K,V> {
187	<	final K key;
607	>	static class Node<V> {
608		final int hash;
609	<	volatile V value;
610	<	final HashEntry<K,V> next;
609	>	final Object key;
610	>	volatile V val;
611	>	volatile Node<V> next;
612
613	<	HashEntry(K key, int hash, HashEntry<K,V> next, V value) {
193	<	this.key = key;
613	>	Node(int hash, Object key, V val, Node<V> next) {
614		this.hash = hash;
615	+	this.key = key;
616	+	this.val = val;
617		this.next = next;
196	–	this.value = value;
618		}
198	–
199	–	@SuppressWarnings("unchecked")
200	–	static final <K,V> HashEntry<K,V>[] newArray(int i) {
201	–	return new HashEntry[i];
202	–	}
619		}
620
621	+	/* ---------------- TreeBins -------------- */
622	+
623		/**
624	<	* Segments are specialized versions of hash tables.  This
625	<	* subclasses from ReentrantLock opportunistically, just to
626	<	* simplify some locking and avoid separate construction.
627	<	*/
628	<	static final class Segment<K,V> extends ReentrantLock implements Serializable {
629	<	/*
630	<	* Segments maintain a table of entry lists that are ALWAYS
631	<	* kept in a consistent state, so can be read without locking.
632	<	* Next fields of nodes are immutable (final).  All list
633	<	* additions are performed at the front of each bin. This
634	<	* makes it easy to check changes, and also fast to traverse.
635	<	* When nodes would otherwise be changed, new nodes are
636	<	* created to replace them. This works well for hash tables
637	<	* since the bin lists tend to be short. (The average length
220	<	* is less than two for the default load factor threshold.)
221	<	*
222	<	* Read operations can thus proceed without locking, but rely
223	<	* on selected uses of volatiles to ensure that completed
224	<	* write operations performed by other threads are
225	<	* noticed. For most purposes, the "count" field, tracking the
226	<	* number of elements, serves as that volatile variable
227	<	* ensuring visibility.  This is convenient because this field
228	<	* needs to be read in many read operations anyway:
229	<	*
230	<	*   - All (unsynchronized) read operations must first read the
231	<	*     "count" field, and should not look at table entries if
232	<	*     it is 0.
233	<	*
234	<	*   - All (synchronized) write operations should write to
235	<	*     the "count" field after structurally changing any bin.
236	<	*     The operations must not take any action that could even
237	<	*     momentarily cause a concurrent read operation to see
238	<	*     inconsistent data. This is made easier by the nature of
239	<	*     the read operations in Map. For example, no operation
240	<	*     can reveal that the table has grown but the threshold
241	<	*     has not yet been updated, so there are no atomicity
242	<	*     requirements for this with respect to reads.
243	<	*
244	<	* As a guide, all critical volatile reads and writes to the
245	<	* count field are marked in code comments.
246	<	*/
624	>	* Nodes for use in TreeBins
625	>	*/
626	>	static final class TreeNode<V> extends Node<V> {
627	>	TreeNode<V> parent;  // red-black tree links
628	>	TreeNode<V> left;
629	>	TreeNode<V> right;
630	>	TreeNode<V> prev;    // needed to unlink next upon deletion
631	>	boolean red;
632	>
633	>	TreeNode(int hash, Object key, V val, Node<V> next, TreeNode<V> parent) {
634	>	super(hash, key, val, next);
635	>	this.parent = parent;
636	>	}
637	>	}
638
639	+	/**
640	+	* A specialized form of red-black tree for use in bins
641	+	* whose size exceeds a threshold.
642	+	*
643	+	* TreeBins use a special form of comparison for search and
644	+	* related operations (which is the main reason we cannot use
645	+	* existing collections such as TreeMaps). TreeBins contain
646	+	* Comparable elements, but may contain others, as well as
647	+	* elements that are Comparable but not necessarily Comparable<T>
648	+	* for the same T, so we cannot invoke compareTo among them. To
649	+	* handle this, the tree is ordered primarily by hash value, then
650	+	* by getClass().getName() order, and then by Comparator order
651	+	* among elements of the same class.  On lookup at a node, if
652	+	* elements are not comparable or compare as 0, both left and
653	+	* right children may need to be searched in the case of tied hash
654	+	* values. (This corresponds to the full list search that would be
655	+	* necessary if all elements were non-Comparable and had tied
656	+	* hashes.)  The red-black balancing code is updated from
657	+	* pre-jdk-collections
658	+	* (http://gee.cs.oswego.edu/dl/classes/collections/RBCell.java)
659	+	* based in turn on Cormen, Leiserson, and Rivest "Introduction to
660	+	* Algorithms" (CLR).
661	+	*
662	+	* TreeBins also maintain a separate locking discipline than
663	+	* regular bins. Because they are forwarded via special MOVED
664	+	* nodes at bin heads (which can never change once established),
665	+	* we cannot use those nodes as locks. Instead, TreeBin
666	+	* extends AbstractQueuedSynchronizer to support a simple form of
667	+	* read-write lock. For update operations and table validation,
668	+	* the exclusive form of lock behaves in the same way as bin-head
669	+	* locks. However, lookups use shared read-lock mechanics to allow
670	+	* multiple readers in the absence of writers.  Additionally,
671	+	* these lookups do not ever block: While the lock is not
672	+	* available, they proceed along the slow traversal path (via
673	+	* next-pointers) until the lock becomes available or the list is
674	+	* exhausted, whichever comes first. (These cases are not fast,
675	+	* but maximize aggregate expected throughput.)  The AQS mechanics
676	+	* for doing this are straightforward.  The lock state is held as
677	+	* AQS getState().  Read counts are negative; the write count (1)
678	+	* is positive.  There are no signalling preferences among readers
679	+	* and writers. Since we don't need to export full Lock API, we
680	+	* just override the minimal AQS methods and use them directly.
681	+	*/
682	+	static final class TreeBin<V> extends AbstractQueuedSynchronizer {
683		private static final long serialVersionUID = 2249069246763182397L;
684	+	transient TreeNode<V> root;  // root of tree
685	+	transient TreeNode<V> first; // head of next-pointer list
686
687	<	/**
688	<	* The number of elements in this segment's region.
689	<	*/
690	<	transient volatile int count;
687	>	/* AQS overrides */
688	>	public final boolean isHeldExclusively() { return getState() > 0; }
689	>	public final boolean tryAcquire(int ignore) {
690	>	if (compareAndSetState(0, 1)) {
691	>	setExclusiveOwnerThread(Thread.currentThread());
692	>	return true;
693	>	}
694	>	return false;
695	>	}
696	>	public final boolean tryRelease(int ignore) {
697	>	setExclusiveOwnerThread(null);
698	>	setState(0);
699	>	return true;
700	>	}
701	>	public final int tryAcquireShared(int ignore) {
702	>	for (int c;;) {
703	>	if ((c = getState()) > 0)
704	>	return -1;
705	>	if (compareAndSetState(c, c -1))
706	>	return 1;
707	>	}
708	>	}
709	>	public final boolean tryReleaseShared(int ignore) {
710	>	int c;
711	>	do {} while (!compareAndSetState(c = getState(), c + 1));
712	>	return c == -1;
713	>	}
714	>
715	>	/** From CLR */
716	>	private void rotateLeft(TreeNode<V> p) {
717	>	if (p != null) {
718	>	TreeNode<V> r = p.right, pp, rl;
719	>	if ((rl = p.right = r.left) != null)
720	>	rl.parent = p;
721	>	if ((pp = r.parent = p.parent) == null)
722	>	root = r;
723	>	else if (pp.left == p)
724	>	pp.left = r;
725	>	else
726	>	pp.right = r;
727	>	r.left = p;
728	>	p.parent = r;
729	>	}
730	>	}
731
732	<	/**
733	<	* Number of updates that alter the size of the table. This is
734	<	* used during bulk-read methods to make sure they see a
735	<	* consistent snapshot: If modCounts change during a traversal
736	<	* of segments computing size or checking containsValue, then
737	<	* we might have an inconsistent view of state so (usually)
738	<	* must retry.
739	<	*/
740	<	transient int modCount;
732	>	/** From CLR */
733	>	private void rotateRight(TreeNode<V> p) {
734	>	if (p != null) {
735	>	TreeNode<V> l = p.left, pp, lr;
736	>	if ((lr = p.left = l.right) != null)
737	>	lr.parent = p;
738	>	if ((pp = l.parent = p.parent) == null)
739	>	root = l;
740	>	else if (pp.right == p)
741	>	pp.right = l;
742	>	else
743	>	pp.left = l;
744	>	l.right = p;
745	>	p.parent = l;
746	>	}
747	>	}
748
749		/**
750	<	* The table is rehashed when its size exceeds this threshold.
751	<	* (The value of this field is always <tt>(int)(capacity *
268	<	* loadFactor)</tt>.)
750	>	* Returns the TreeNode (or null if not found) for the given key
751	>	* starting at given root.
752		*/
753	<	transient int threshold;
753	>	@SuppressWarnings("unchecked") final TreeNode<V> getTreeNode
754	>	(int h, Object k, TreeNode<V> p) {
755	>	Class<?> c = k.getClass();
756	>	while (p != null) {
757	>	int dir, ph;  Object pk; Class<?> pc;
758	>	if ((ph = p.hash) == h) {
759	>	if ((pk = p.key) == k || k.equals(pk))
760	>	return p;
761	>	if (c != (pc = pk.getClass()) ||
762	>	!(k instanceof Comparable) ||
763	>	(dir = ((Comparable)k).compareTo((Comparable)pk)) == 0) {
764	>	if ((dir = (c == pc) ? 0 :
765	>	c.getName().compareTo(pc.getName())) == 0) {
766	>	TreeNode<V> r = null, pl, pr; // check both sides
767	>	if ((pr = p.right) != null && h >= pr.hash &&
768	>	(r = getTreeNode(h, k, pr)) != null)
769	>	return r;
770	>	else if ((pl = p.left) != null && h <= pl.hash)
771	>	dir = -1;
772	>	else // nothing there
773	>	return null;
774	>	}
775	>	}
776	>	}
777	>	else
778	>	dir = (h < ph) ? -1 : 1;
779	>	p = (dir > 0) ? p.right : p.left;
780	>	}
781	>	return null;
782	>	}
783
784		/**
785	<	* The per-segment table.
785	>	* Wrapper for getTreeNode used by CHM.get. Tries to obtain
786	>	* read-lock to call getTreeNode, but during failure to get
787	>	* lock, searches along next links.
788		*/
789	<	transient volatile HashEntry<K,V>[] table;
789	>	final V getValue(int h, Object k) {
790	>	Node<V> r = null;
791	>	int c = getState(); // Must read lock state first
792	>	for (Node<V> e = first; e != null; e = e.next) {
793	>	if (c <= 0 && compareAndSetState(c, c - 1)) {
794	>	try {
795	>	r = getTreeNode(h, k, root);
796	>	} finally {
797	>	releaseShared(0);
798	>	}
799	>	break;
800	>	}
801	>	else if (e.hash == h && k.equals(e.key)) {
802	>	r = e;
803	>	break;
804	>	}
805	>	else
806	>	c = getState();
807	>	}
808	>	return r == null ? null : r.val;
809	>	}
810
811		/**
812	<	* The load factor for the hash table.  Even though this value
813	<	* is same for all segments, it is replicated to avoid needing
280	<	* links to outer object.
281	<	* @serial
812	>	* Finds or adds a node.
813	>	* @return null if added
814		*/
815	<	final float loadFactor;
816	<
817	<	Segment(int initialCapacity, float lf) {
818	<	loadFactor = lf;
819	<	setTable(HashEntry.<K,V>newArray(initialCapacity));
820	<	}
815	>	@SuppressWarnings("unchecked") final TreeNode<V> putTreeNode
816	>	(int h, Object k, V v) {
817	>	Class<?> c = k.getClass();
818	>	TreeNode<V> pp = root, p = null;
819	>	int dir = 0;
820	>	while (pp != null) { // find existing node or leaf to insert at
821	>	int ph;  Object pk; Class<?> pc;
822	>	p = pp;
823	>	if ((ph = p.hash) == h) {
824	>	if ((pk = p.key) == k || k.equals(pk))
825	>	return p;
826	>	if (c != (pc = pk.getClass()) ||
827	>	!(k instanceof Comparable) ||
828	>	(dir = ((Comparable)k).compareTo((Comparable)pk)) == 0) {
829	>	TreeNode<V> s = null, r = null, pr;
830	>	if ((dir = (c == pc) ? 0 :
831	>	c.getName().compareTo(pc.getName())) == 0) {
832	>	if ((pr = p.right) != null && h >= pr.hash &&
833	>	(r = getTreeNode(h, k, pr)) != null)
834	>	return r;
835	>	else // continue left
836	>	dir = -1;
837	>	}
838	>	else if ((pr = p.right) != null && h >= pr.hash)
839	>	s = pr;
840	>	if (s != null && (r = getTreeNode(h, k, s)) != null)
841	>	return r;
842	>	}
843	>	}
844	>	else
845	>	dir = (h < ph) ? -1 : 1;
846	>	pp = (dir > 0) ? p.right : p.left;
847	>	}
848
849	<	@SuppressWarnings("unchecked")
850	<	static final <K,V> Segment<K,V>[] newArray(int i) {
851	<	return new Segment[i];
849	>	TreeNode<V> f = first;
850	>	TreeNode<V> x = first = new TreeNode<V>(h, k, v, f, p);
851	>	if (p == null)
852	>	root = x;
853	>	else { // attach and rebalance; adapted from CLR
854	>	TreeNode<V> xp, xpp;
855	>	if (f != null)
856	>	f.prev = x;
857	>	if (dir <= 0)
858	>	p.left = x;
859	>	else
860	>	p.right = x;
861	>	x.red = true;
862	>	while (x != null && (xp = x.parent) != null && xp.red &&
863	>	(xpp = xp.parent) != null) {
864	>	TreeNode<V> xppl = xpp.left;
865	>	if (xp == xppl) {
866	>	TreeNode<V> y = xpp.right;
867	>	if (y != null && y.red) {
868	>	y.red = false;
869	>	xp.red = false;
870	>	xpp.red = true;
871	>	x = xpp;
872	>	}
873	>	else {
874	>	if (x == xp.right) {
875	>	rotateLeft(x = xp);
876	>	xpp = (xp = x.parent) == null ? null : xp.parent;
877	>	}
878	>	if (xp != null) {
879	>	xp.red = false;
880	>	if (xpp != null) {
881	>	xpp.red = true;
882	>	rotateRight(xpp);
883	>	}
884	>	}
885	>	}
886	>	}
887	>	else {
888	>	TreeNode<V> y = xppl;
889	>	if (y != null && y.red) {
890	>	y.red = false;
891	>	xp.red = false;
892	>	xpp.red = true;
893	>	x = xpp;
894	>	}
895	>	else {
896	>	if (x == xp.left) {
897	>	rotateRight(x = xp);
898	>	xpp = (xp = x.parent) == null ? null : xp.parent;
899	>	}
900	>	if (xp != null) {
901	>	xp.red = false;
902	>	if (xpp != null) {
903	>	xpp.red = true;
904	>	rotateLeft(xpp);
905	>	}
906	>	}
907	>	}
908	>	}
909	>	}
910	>	TreeNode<V> r = root;
911	>	if (r != null && r.red)
912	>	r.red = false;
913	>	}
914	>	return null;
915		}
916
917		/**
918	<	* Sets table to new HashEntry array.
919	<	* Call only while holding lock or in constructor.
918	>	* Removes the given node, that must be present before this
919	>	* call.  This is messier than typical red-black deletion code
920	>	* because we cannot swap the contents of an interior node
921	>	* with a leaf successor that is pinned by "next" pointers
922	>	* that are accessible independently of lock. So instead we
923	>	* swap the tree linkages.
924		*/
925	<	void setTable(HashEntry<K,V>[] newTable) {
926	<	threshold = (int)(newTable.length * loadFactor);
927	<	table = newTable;
925	>	final void deleteTreeNode(TreeNode<V> p) {
926	>	TreeNode<V> next = (TreeNode<V>)p.next; // unlink traversal pointers
927	>	TreeNode<V> pred = p.prev;
928	>	if (pred == null)
929	>	first = next;
930	>	else
931	>	pred.next = next;
932	>	if (next != null)
933	>	next.prev = pred;
934	>	TreeNode<V> replacement;
935	>	TreeNode<V> pl = p.left;
936	>	TreeNode<V> pr = p.right;
937	>	if (pl != null && pr != null) {
938	>	TreeNode<V> s = pr, sl;
939	>	while ((sl = s.left) != null) // find successor
940	>	s = sl;
941	>	boolean c = s.red; s.red = p.red; p.red = c; // swap colors
942	>	TreeNode<V> sr = s.right;
943	>	TreeNode<V> pp = p.parent;
944	>	if (s == pr) { // p was s's direct parent
945	>	p.parent = s;
946	>	s.right = p;
947	>	}
948	>	else {
949	>	TreeNode<V> sp = s.parent;
950	>	if ((p.parent = sp) != null) {
951	>	if (s == sp.left)
952	>	sp.left = p;
953	>	else
954	>	sp.right = p;
955	>	}
956	>	if ((s.right = pr) != null)
957	>	pr.parent = s;
958	>	}
959	>	p.left = null;
960	>	if ((p.right = sr) != null)
961	>	sr.parent = p;
962	>	if ((s.left = pl) != null)
963	>	pl.parent = s;
964	>	if ((s.parent = pp) == null)
965	>	root = s;
966	>	else if (p == pp.left)
967	>	pp.left = s;
968	>	else
969	>	pp.right = s;
970	>	replacement = sr;
971	>	}
972	>	else
973	>	replacement = (pl != null) ? pl : pr;
974	>	TreeNode<V> pp = p.parent;
975	>	if (replacement == null) {
976	>	if (pp == null) {
977	>	root = null;
978	>	return;
979	>	}
980	>	replacement = p;
981	>	}
982	>	else {
983	>	replacement.parent = pp;
984	>	if (pp == null)
985	>	root = replacement;
986	>	else if (p == pp.left)
987	>	pp.left = replacement;
988	>	else
989	>	pp.right = replacement;
990	>	p.left = p.right = p.parent = null;
991	>	}
992	>	if (!p.red) { // rebalance, from CLR
993	>	TreeNode<V> x = replacement;
994	>	while (x != null) {
995	>	TreeNode<V> xp, xpl;
996	>	if (x.red || (xp = x.parent) == null) {
997	>	x.red = false;
998	>	break;
999	>	}
1000	>	if (x == (xpl = xp.left)) {
1001	>	TreeNode<V> sib = xp.right;
1002	>	if (sib != null && sib.red) {
1003	>	sib.red = false;
1004	>	xp.red = true;
1005	>	rotateLeft(xp);
1006	>	sib = (xp = x.parent) == null ? null : xp.right;
1007	>	}
1008	>	if (sib == null)
1009	>	x = xp;
1010	>	else {
1011	>	TreeNode<V> sl = sib.left, sr = sib.right;
1012	>	if ((sr == null || !sr.red) &&
1013	>	(sl == null || !sl.red)) {
1014	>	sib.red = true;
1015	>	x = xp;
1016	>	}
1017	>	else {
1018	>	if (sr == null || !sr.red) {
1019	>	if (sl != null)
1020	>	sl.red = false;
1021	>	sib.red = true;
1022	>	rotateRight(sib);
1023	>	sib = (xp = x.parent) == null ?
1024	>	null : xp.right;
1025	>	}
1026	>	if (sib != null) {
1027	>	sib.red = (xp == null) ? false : xp.red;
1028	>	if ((sr = sib.right) != null)
1029	>	sr.red = false;
1030	>	}
1031	>	if (xp != null) {
1032	>	xp.red = false;
1033	>	rotateLeft(xp);
1034	>	}
1035	>	x = root;
1036	>	}
1037	>	}
1038	>	}
1039	>	else { // symmetric
1040	>	TreeNode<V> sib = xpl;
1041	>	if (sib != null && sib.red) {
1042	>	sib.red = false;
1043	>	xp.red = true;
1044	>	rotateRight(xp);
1045	>	sib = (xp = x.parent) == null ? null : xp.left;
1046	>	}
1047	>	if (sib == null)
1048	>	x = xp;
1049	>	else {
1050	>	TreeNode<V> sl = sib.left, sr = sib.right;
1051	>	if ((sl == null || !sl.red) &&
1052	>	(sr == null || !sr.red)) {
1053	>	sib.red = true;
1054	>	x = xp;
1055	>	}
1056	>	else {
1057	>	if (sl == null || !sl.red) {
1058	>	if (sr != null)
1059	>	sr.red = false;
1060	>	sib.red = true;
1061	>	rotateLeft(sib);
1062	>	sib = (xp = x.parent) == null ?
1063	>	null : xp.left;
1064	>	}
1065	>	if (sib != null) {
1066	>	sib.red = (xp == null) ? false : xp.red;
1067	>	if ((sl = sib.left) != null)
1068	>	sl.red = false;
1069	>	}
1070	>	if (xp != null) {
1071	>	xp.red = false;
1072	>	rotateRight(xp);
1073	>	}
1074	>	x = root;
1075	>	}
1076	>	}
1077	>	}
1078	>	}
1079	>	}
1080	>	if (p == replacement && (pp = p.parent) != null) {
1081	>	if (p == pp.left) // detach pointers
1082	>	pp.left = null;
1083	>	else if (p == pp.right)
1084	>	pp.right = null;
1085	>	p.parent = null;
1086	>	}
1087		}
1088	+	}
1089
1090	<	/**
1091	<	* Returns properly casted first entry of bin for given hash.
1092	<	*/
1093	<	HashEntry<K,V> getFirst(int hash) {
1094	<	HashEntry<K,V>[] tab = table;
1095	<	return tab[hash & (tab.length - 1)];
1090	>	/* ---------------- Collision reduction methods -------------- */
1091	>
1092	>	/**
1093	>	* Spreads higher bits to lower, and also forces top bit to 0.
1094	>	* Because the table uses power-of-two masking, sets of hashes
1095	>	* that vary only in bits above the current mask will always
1096	>	* collide. (Among known examples are sets of Float keys holding
1097	>	* consecutive whole numbers in small tables.)  To counter this,
1098	>	* we apply a transform that spreads the impact of higher bits
1099	>	* downward. There is a tradeoff between speed, utility, and
1100	>	* quality of bit-spreading. Because many common sets of hashes
1101	>	* are already reasonably distributed across bits (so don't benefit
1102	>	* from spreading), and because we use trees to handle large sets
1103	>	* of collisions in bins, we don't need excessively high quality.
1104	>	*/
1105	>	private static final int spread(int h) {
1106	>	h ^= (h >>> 18) ^ (h >>> 12);
1107	>	return (h ^ (h >>> 10)) & HASH_BITS;
1108	>	}
1109	>
1110	>	/**
1111	>	* Replaces a list bin with a tree bin if key is comparable.  Call
1112	>	* only when locked.
1113	>	*/
1114	>	private final void replaceWithTreeBin(Node<V>[] tab, int index, Object key) {
1115	>	if (key instanceof Comparable) {
1116	>	TreeBin<V> t = new TreeBin<V>();
1117	>	for (Node<V> e = tabAt(tab, index); e != null; e = e.next)
1118	>	t.putTreeNode(e.hash, e.key, e.val);
1119	>	setTabAt(tab, index, new Node<V>(MOVED, t, null, null));
1120		}
1121	+	}
1122
1123	<	/**
1124	<	* Reads value field of an entry under lock. Called if value
1125	<	* field ever appears to be null. This is possible only if a
1126	<	* compiler happens to reorder a HashEntry initialization with
1127	<	* its table assignment, which is legal under memory model
1128	<	* but is not known to ever occur.
1129	<	*/
1130	<	V readValueUnderLock(HashEntry<K,V> e) {
1131	<	lock();
1132	<	try {
1133	<	return e.value;
1134	<	} finally {
1135	<	unlock();
1123	>	/* ---------------- Internal access and update methods -------------- */
1124	>
1125	>	/** Implementation for get and containsKey */
1126	>	@SuppressWarnings("unchecked") private final V internalGet(Object k) {
1127	>	int h = spread(k.hashCode());
1128	>	retry: for (Node<V>[] tab = table; tab != null;) {
1129	>	Node<V> e; Object ek; V ev; int eh; // locals to read fields once
1130	>	for (e = tabAt(tab, (tab.length - 1) & h); e != null; e = e.next) {
1131	>	if ((eh = e.hash) < 0) {
1132	>	if ((ek = e.key) instanceof TreeBin)  // search TreeBin
1133	>	return ((TreeBin<V>)ek).getValue(h, k);
1134	>	else {                      // restart with new table
1135	>	tab = (Node<V>[])ek;
1136	>	continue retry;
1137	>	}
1138	>	}
1139	>	else if (eh == h && (ev = e.val) != null &&
1140	>	((ek = e.key) == k || k.equals(ek)))
1141	>	return ev;
1142		}
1143	+	break;
1144		}
1145	+	return null;
1146	+	}
1147
1148	<	/* Specialized implementations of map methods */
1149	<
1150	<	V get(Object key, int hash) {
1151	<	if (count != 0) { // read-volatile
1152	<	HashEntry<K,V> e = getFirst(hash);
1153	<	while (e != null) {
1154	<	if (e.hash == hash && key.equals(e.key)) {
1155	<	V v = e.value;
1156	<	if (v != null)
1157	<	return v;
1158	<	return readValueUnderLock(e); // recheck
1148	>	/**
1149	>	* Implementation for the four public remove/replace methods:
1150	>	* Replaces node value with v, conditional upon match of cv if
1151	>	* non-null.  If resulting value is null, delete.
1152	>	*/
1153	>	@SuppressWarnings("unchecked") private final V internalReplace
1154	>	(Object k, V v, Object cv) {
1155	>	int h = spread(k.hashCode());
1156	>	V oldVal = null;
1157	>	for (Node<V>[] tab = table;;) {
1158	>	Node<V> f; int i, fh; Object fk;
1159	>	if (tab == null ||
1160	>	(f = tabAt(tab, i = (tab.length - 1) & h)) == null)
1161	>	break;
1162	>	else if ((fh = f.hash) < 0) {
1163	>	if ((fk = f.key) instanceof TreeBin) {
1164	>	TreeBin<V> t = (TreeBin<V>)fk;
1165	>	boolean validated = false;
1166	>	boolean deleted = false;
1167	>	t.acquire(0);
1168	>	try {
1169	>	if (tabAt(tab, i) == f) {
1170	>	validated = true;
1171	>	TreeNode<V> p = t.getTreeNode(h, k, t.root);
1172	>	if (p != null) {
1173	>	V pv = p.val;
1174	>	if (cv == null || cv == pv || cv.equals(pv)) {
1175	>	oldVal = pv;
1176	>	if ((p.val = v) == null) {
1177	>	deleted = true;
1178	>	t.deleteTreeNode(p);
1179	>	}
1180	>	}
1181	>	}
1182	>	}
1183	>	} finally {
1184	>	t.release(0);
1185	>	}
1186	>	if (validated) {
1187	>	if (deleted)
1188	>	addCount(-1L, -1);
1189	>	break;
1190		}
1191	<	e = e.next;
1191	>	}
1192	>	else
1193	>	tab = (Node<V>[])fk;
1194	>	}
1195	>	else if (fh != h && f.next == null) // precheck
1196	>	break;                          // rules out possible existence
1197	>	else {
1198	>	boolean validated = false;
1199	>	boolean deleted = false;
1200	>	synchronized (f) {
1201	>	if (tabAt(tab, i) == f) {
1202	>	validated = true;
1203	>	for (Node<V> e = f, pred = null;;) {
1204	>	Object ek; V ev;
1205	>	if (e.hash == h &&
1206	>	((ev = e.val) != null) &&
1207	>	((ek = e.key) == k || k.equals(ek))) {
1208	>	if (cv == null || cv == ev || cv.equals(ev)) {
1209	>	oldVal = ev;
1210	>	if ((e.val = v) == null) {
1211	>	deleted = true;
1212	>	Node<V> en = e.next;
1213	>	if (pred != null)
1214	>	pred.next = en;
1215	>	else
1216	>	setTabAt(tab, i, en);
1217	>	}
1218	>	}
1219	>	break;
1220	>	}
1221	>	pred = e;
1222	>	if ((e = e.next) == null)
1223	>	break;
1224	>	}
1225	>	}
1226	>	}
1227	>	if (validated) {
1228	>	if (deleted)
1229	>	addCount(-1L, -1);
1230	>	break;
1231		}
1232		}
343	–	return null;
1233		}
1234	+	return oldVal;
1235	+	}
1236
1237	<	boolean containsKey(Object key, int hash) {
1238	<	if (count != 0) { // read-volatile
1239	<	HashEntry<K,V> e = getFirst(hash);
1240	<	while (e != null) {
1241	<	if (e.hash == hash && key.equals(e.key))
1242	<	return true;
1243	<	e = e.next;
1237	>	/*
1238	>	* Internal versions of insertion methods
1239	>	* All have the same basic structure as the first (internalPut):
1240	>	*  1. If table uninitialized, create
1241	>	*  2. If bin empty, try to CAS new node
1242	>	*  3. If bin stale, use new table
1243	>	*  4. if bin converted to TreeBin, validate and relay to TreeBin methods
1244	>	*  5. Lock and validate; if valid, scan and add or update
1245	>	*
1246	>	* The putAll method differs mainly in attempting to pre-allocate
1247	>	* enough table space, and also more lazily performs count updates
1248	>	* and checks.
1249	>	*
1250	>	* Most of the function-accepting methods can't be factored nicely
1251	>	* because they require different functional forms, so instead
1252	>	* sprawl out similar mechanics.
1253	>	*/
1254	>
1255	>	/** Implementation for put and putIfAbsent */
1256	>	@SuppressWarnings("unchecked") private final V internalPut
1257	>	(K k, V v, boolean onlyIfAbsent) {
1258	>	if (k == null || v == null) throw new NullPointerException();
1259	>	int h = spread(k.hashCode());
1260	>	int len = 0;
1261	>	for (Node<V>[] tab = table;;) {
1262	>	int i, fh; Node<V> f; Object fk; V fv;
1263	>	if (tab == null)
1264	>	tab = initTable();
1265	>	else if ((f = tabAt(tab, i = (tab.length - 1) & h)) == null) {
1266	>	if (casTabAt(tab, i, null, new Node<V>(h, k, v, null)))
1267	>	break;                   // no lock when adding to empty bin
1268	>	}
1269	>	else if ((fh = f.hash) < 0) {
1270	>	if ((fk = f.key) instanceof TreeBin) {
1271	>	TreeBin<V> t = (TreeBin<V>)fk;
1272	>	V oldVal = null;
1273	>	t.acquire(0);
1274	>	try {
1275	>	if (tabAt(tab, i) == f) {
1276	>	len = 2;
1277	>	TreeNode<V> p = t.putTreeNode(h, k, v);
1278	>	if (p != null) {
1279	>	oldVal = p.val;
1280	>	if (!onlyIfAbsent)
1281	>	p.val = v;
1282	>	}
1283	>	}
1284	>	} finally {
1285	>	t.release(0);
1286	>	}
1287	>	if (len != 0) {
1288	>	if (oldVal != null)
1289	>	return oldVal;
1290	>	break;
1291	>	}
1292	>	}
1293	>	else
1294	>	tab = (Node<V>[])fk;
1295	>	}
1296	>	else if (onlyIfAbsent && fh == h && (fv = f.val) != null &&
1297	>	((fk = f.key) == k || k.equals(fk))) // peek while nearby
1298	>	return fv;
1299	>	else {
1300	>	V oldVal = null;
1301	>	synchronized (f) {
1302	>	if (tabAt(tab, i) == f) {
1303	>	len = 1;
1304	>	for (Node<V> e = f;; ++len) {
1305	>	Object ek; V ev;
1306	>	if (e.hash == h &&
1307	>	(ev = e.val) != null &&
1308	>	((ek = e.key) == k || k.equals(ek))) {
1309	>	oldVal = ev;
1310	>	if (!onlyIfAbsent)
1311	>	e.val = v;
1312	>	break;
1313	>	}
1314	>	Node<V> last = e;
1315	>	if ((e = e.next) == null) {
1316	>	last.next = new Node<V>(h, k, v, null);
1317	>	if (len >= TREE_THRESHOLD)
1318	>	replaceWithTreeBin(tab, i, k);
1319	>	break;
1320	>	}
1321	>	}
1322	>	}
1323	>	}
1324	>	if (len != 0) {
1325	>	if (oldVal != null)
1326	>	return oldVal;
1327	>	break;
1328		}
1329		}
355	–	return false;
1330		}
1331	+	addCount(1L, len);
1332	+	return null;
1333	+	}
1334
1335	<	boolean containsValue(Object value) {
1336	<	if (count != 0) { // read-volatile
1337	<	HashEntry<K,V>[] tab = table;
1338	<	int len = tab.length;
1339	<	for (int i = 0 ; i < len; i++) {
1340	<	for (HashEntry<K,V> e = tab[i]; e != null; e = e.next) {
1341	<	V v = e.value;
1342	<	if (v == null) // recheck
1343	<	v = readValueUnderLock(e);
1344	<	if (value.equals(v))
1345	<	return true;
1335	>	/** Implementation for computeIfAbsent */
1336	>	@SuppressWarnings("unchecked") private final V internalComputeIfAbsent
1337	>	(K k, Function<? super K, ? extends V> mf) {
1338	>	if (k == null || mf == null)
1339	>	throw new NullPointerException();
1340	>	int h = spread(k.hashCode());
1341	>	V val = null;
1342	>	int len = 0;
1343	>	for (Node<V>[] tab = table;;) {
1344	>	Node<V> f; int i; Object fk;
1345	>	if (tab == null)
1346	>	tab = initTable();
1347	>	else if ((f = tabAt(tab, i = (tab.length - 1) & h)) == null) {
1348	>	Node<V> node = new Node<V>(h, k, null, null);
1349	>	synchronized (node) {
1350	>	if (casTabAt(tab, i, null, node)) {
1351	>	len = 1;
1352	>	try {
1353	>	if ((val = mf.apply(k)) != null)
1354	>	node.val = val;
1355	>	} finally {
1356	>	if (val == null)
1357	>	setTabAt(tab, i, null);
1358	>	}
1359		}
1360		}
1361	+	if (len != 0)
1362	+	break;
1363	+	}
1364	+	else if (f.hash < 0) {
1365	+	if ((fk = f.key) instanceof TreeBin) {
1366	+	TreeBin<V> t = (TreeBin<V>)fk;
1367	+	boolean added = false;
1368	+	t.acquire(0);
1369	+	try {
1370	+	if (tabAt(tab, i) == f) {
1371	+	len = 1;
1372	+	TreeNode<V> p = t.getTreeNode(h, k, t.root);
1373	+	if (p != null)
1374	+	val = p.val;
1375	+	else if ((val = mf.apply(k)) != null) {
1376	+	added = true;
1377	+	len = 2;
1378	+	t.putTreeNode(h, k, val);
1379	+	}
1380	+	}
1381	+	} finally {
1382	+	t.release(0);
1383	+	}
1384	+	if (len != 0) {
1385	+	if (!added)
1386	+	return val;
1387	+	break;
1388	+	}
1389	+	}
1390	+	else
1391	+	tab = (Node<V>[])fk;
1392	+	}
1393	+	else {
1394	+	for (Node<V> e = f; e != null; e = e.next) { // prescan
1395	+	Object ek; V ev;
1396	+	if (e.hash == h && (ev = e.val) != null &&
1397	+	((ek = e.key) == k || k.equals(ek)))
1398	+	return ev;
1399	+	}
1400	+	boolean added = false;
1401	+	synchronized (f) {
1402	+	if (tabAt(tab, i) == f) {
1403	+	len = 1;
1404	+	for (Node<V> e = f;; ++len) {
1405	+	Object ek; V ev;
1406	+	if (e.hash == h &&
1407	+	(ev = e.val) != null &&
1408	+	((ek = e.key) == k || k.equals(ek))) {
1409	+	val = ev;
1410	+	break;
1411	+	}
1412	+	Node<V> last = e;
1413	+	if ((e = e.next) == null) {
1414	+	if ((val = mf.apply(k)) != null) {
1415	+	added = true;
1416	+	last.next = new Node<V>(h, k, val, null);
1417	+	if (len >= TREE_THRESHOLD)
1418	+	replaceWithTreeBin(tab, i, k);
1419	+	}
1420	+	break;
1421	+	}
1422	+	}
1423	+	}
1424	+	}
1425	+	if (len != 0) {
1426	+	if (!added)
1427	+	return val;
1428	+	break;
1429	+	}
1430		}
372	–	return false;
1431		}
1432	+	if (val != null)
1433	+	addCount(1L, len);
1434	+	return val;
1435	+	}
1436
1437	<	boolean replace(K key, int hash, V oldValue, V newValue) {
1438	<	lock();
1439	<	try {
1440	<	HashEntry<K,V> e = getFirst(hash);
1441	<	while (e != null && (e.hash != hash || !key.equals(e.key)))
1442	<	e = e.next;
1437	>	/** Implementation for compute */
1438	>	@SuppressWarnings("unchecked") private final V internalCompute
1439	>	(K k, boolean onlyIfPresent,
1440	>	BiFunction<? super K, ? super V, ? extends V> mf) {
1441	>	if (k == null || mf == null)
1442	>	throw new NullPointerException();
1443	>	int h = spread(k.hashCode());
1444	>	V val = null;
1445	>	int delta = 0;
1446	>	int len = 0;
1447	>	for (Node<V>[] tab = table;;) {
1448	>	Node<V> f; int i, fh; Object fk;
1449	>	if (tab == null)
1450	>	tab = initTable();
1451	>	else if ((f = tabAt(tab, i = (tab.length - 1) & h)) == null) {
1452	>	if (onlyIfPresent)
1453	>	break;
1454	>	Node<V> node = new Node<V>(h, k, null, null);
1455	>	synchronized (node) {
1456	>	if (casTabAt(tab, i, null, node)) {
1457	>	try {
1458	>	len = 1;
1459	>	if ((val = mf.apply(k, null)) != null) {
1460	>	node.val = val;
1461	>	delta = 1;
1462	>	}
1463	>	} finally {
1464	>	if (delta == 0)
1465	>	setTabAt(tab, i, null);
1466	>	}
1467	>	}
1468	>	}
1469	>	if (len != 0)
1470	>	break;
1471	>	}
1472	>	else if ((fh = f.hash) < 0) {
1473	>	if ((fk = f.key) instanceof TreeBin) {
1474	>	TreeBin<V> t = (TreeBin<V>)fk;
1475	>	t.acquire(0);
1476	>	try {
1477	>	if (tabAt(tab, i) == f) {
1478	>	len = 1;
1479	>	TreeNode<V> p = t.getTreeNode(h, k, t.root);
1480	>	if (p == null && onlyIfPresent)
1481	>	break;
1482	>	V pv = (p == null) ? null : p.val;
1483	>	if ((val = mf.apply(k, pv)) != null) {
1484	>	if (p != null)
1485	>	p.val = val;
1486	>	else {
1487	>	len = 2;
1488	>	delta = 1;
1489	>	t.putTreeNode(h, k, val);
1490	>	}
1491	>	}
1492	>	else if (p != null) {
1493	>	delta = -1;
1494	>	t.deleteTreeNode(p);
1495	>	}
1496	>	}
1497	>	} finally {
1498	>	t.release(0);
1499	>	}
1500	>	if (len != 0)
1501	>	break;
1502	>	}
1503	>	else
1504	>	tab = (Node<V>[])fk;
1505	>	}
1506	>	else {
1507	>	synchronized (f) {
1508	>	if (tabAt(tab, i) == f) {
1509	>	len = 1;
1510	>	for (Node<V> e = f, pred = null;; ++len) {
1511	>	Object ek; V ev;
1512	>	if (e.hash == h &&
1513	>	(ev = e.val) != null &&
1514	>	((ek = e.key) == k || k.equals(ek))) {
1515	>	val = mf.apply(k, ev);
1516	>	if (val != null)
1517	>	e.val = val;
1518	>	else {
1519	>	delta = -1;
1520	>	Node<V> en = e.next;
1521	>	if (pred != null)
1522	>	pred.next = en;
1523	>	else
1524	>	setTabAt(tab, i, en);
1525	>	}
1526	>	break;
1527	>	}
1528	>	pred = e;
1529	>	if ((e = e.next) == null) {
1530	>	if (!onlyIfPresent &&
1531	>	(val = mf.apply(k, null)) != null) {
1532	>	pred.next = new Node<V>(h, k, val, null);
1533	>	delta = 1;
1534	>	if (len >= TREE_THRESHOLD)
1535	>	replaceWithTreeBin(tab, i, k);
1536	>	}
1537	>	break;
1538	>	}
1539	>	}
1540	>	}
1541	>	}
1542	>	if (len != 0)
1543	>	break;
1544	>	}
1545	>	}
1546	>	if (delta != 0)
1547	>	addCount((long)delta, len);
1548	>	return val;
1549	>	}
1550
1551	<	boolean replaced = false;
1552	<	if (e != null && oldValue.equals(e.value)) {
1553	<	replaced = true;
1554	<	e.value = newValue;
1551	>	/** Implementation for merge */
1552	>	@SuppressWarnings("unchecked") private final V internalMerge
1553	>	(K k, V v, BiFunction<? super V, ? super V, ? extends V> mf) {
1554	>	if (k == null || v == null || mf == null)
1555	>	throw new NullPointerException();
1556	>	int h = spread(k.hashCode());
1557	>	V val = null;
1558	>	int delta = 0;
1559	>	int len = 0;
1560	>	for (Node<V>[] tab = table;;) {
1561	>	int i; Node<V> f; Object fk; V fv;
1562	>	if (tab == null)
1563	>	tab = initTable();
1564	>	else if ((f = tabAt(tab, i = (tab.length - 1) & h)) == null) {
1565	>	if (casTabAt(tab, i, null, new Node<V>(h, k, v, null))) {
1566	>	delta = 1;
1567	>	val = v;
1568	>	break;
1569	>	}
1570	>	}
1571	>	else if (f.hash < 0) {
1572	>	if ((fk = f.key) instanceof TreeBin) {
1573	>	TreeBin<V> t = (TreeBin<V>)fk;
1574	>	t.acquire(0);
1575	>	try {
1576	>	if (tabAt(tab, i) == f) {
1577	>	len = 1;
1578	>	TreeNode<V> p = t.getTreeNode(h, k, t.root);
1579	>	val = (p == null) ? v : mf.apply(p.val, v);
1580	>	if (val != null) {
1581	>	if (p != null)
1582	>	p.val = val;
1583	>	else {
1584	>	len = 2;
1585	>	delta = 1;
1586	>	t.putTreeNode(h, k, val);
1587	>	}
1588	>	}
1589	>	else if (p != null) {
1590	>	delta = -1;
1591	>	t.deleteTreeNode(p);
1592	>	}
1593	>	}
1594	>	} finally {
1595	>	t.release(0);
1596	>	}
1597	>	if (len != 0)
1598	>	break;
1599	>	}
1600	>	else
1601	>	tab = (Node<V>[])fk;
1602	>	}
1603	>	else {
1604	>	synchronized (f) {
1605	>	if (tabAt(tab, i) == f) {
1606	>	len = 1;
1607	>	for (Node<V> e = f, pred = null;; ++len) {
1608	>	Object ek; V ev;
1609	>	if (e.hash == h &&
1610	>	(ev = e.val) != null &&
1611	>	((ek = e.key) == k || k.equals(ek))) {
1612	>	val = mf.apply(ev, v);
1613	>	if (val != null)
1614	>	e.val = val;
1615	>	else {
1616	>	delta = -1;
1617	>	Node<V> en = e.next;
1618	>	if (pred != null)
1619	>	pred.next = en;
1620	>	else
1621	>	setTabAt(tab, i, en);
1622	>	}
1623	>	break;
1624	>	}
1625	>	pred = e;
1626	>	if ((e = e.next) == null) {
1627	>	val = v;
1628	>	pred.next = new Node<V>(h, k, val, null);
1629	>	delta = 1;
1630	>	if (len >= TREE_THRESHOLD)
1631	>	replaceWithTreeBin(tab, i, k);
1632	>	break;
1633	>	}
1634	>	}
1635	>	}
1636		}
1637	<	return replaced;
1638	<	} finally {
389	<	unlock();
1637	>	if (len != 0)
1638	>	break;
1639		}
1640		}
1641	+	if (delta != 0)
1642	+	addCount((long)delta, len);
1643	+	return val;
1644	+	}
1645
1646	<	V replace(K key, int hash, V newValue) {
1647	<	lock();
1648	<	try {
1649	<	HashEntry<K,V> e = getFirst(hash);
1650	<	while (e != null && (e.hash != hash || !key.equals(e.key)))
1651	<	e = e.next;
1646	>	/** Implementation for putAll */
1647	>	@SuppressWarnings("unchecked") private final void internalPutAll
1648	>	(Map<? extends K, ? extends V> m) {
1649	>	tryPresize(m.size());
1650	>	long delta = 0L;     // number of uncommitted additions
1651	>	boolean npe = false; // to throw exception on exit for nulls
1652	>	try {                // to clean up counts on other exceptions
1653	>	for (Map.Entry<?, ? extends V> entry : m.entrySet()) {
1654	>	Object k; V v;
1655	>	if (entry == null || (k = entry.getKey()) == null ||
1656	>	(v = entry.getValue()) == null) {
1657	>	npe = true;
1658	>	break;
1659	>	}
1660	>	int h = spread(k.hashCode());
1661	>	for (Node<V>[] tab = table;;) {
1662	>	int i; Node<V> f; int fh; Object fk;
1663	>	if (tab == null)
1664	>	tab = initTable();
1665	>	else if ((f = tabAt(tab, i = (tab.length - 1) & h)) == null){
1666	>	if (casTabAt(tab, i, null, new Node<V>(h, k, v, null))) {
1667	>	++delta;
1668	>	break;
1669	>	}
1670	>	}
1671	>	else if ((fh = f.hash) < 0) {
1672	>	if ((fk = f.key) instanceof TreeBin) {
1673	>	TreeBin<V> t = (TreeBin<V>)fk;
1674	>	boolean validated = false;
1675	>	t.acquire(0);
1676	>	try {
1677	>	if (tabAt(tab, i) == f) {
1678	>	validated = true;
1679	>	TreeNode<V> p = t.getTreeNode(h, k, t.root);
1680	>	if (p != null)
1681	>	p.val = v;
1682	>	else {
1683	>	t.putTreeNode(h, k, v);
1684	>	++delta;
1685	>	}
1686	>	}
1687	>	} finally {
1688	>	t.release(0);
1689	>	}
1690	>	if (validated)
1691	>	break;
1692	>	}
1693	>	else
1694	>	tab = (Node<V>[])fk;
1695	>	}
1696	>	else {
1697	>	int len = 0;
1698	>	synchronized (f) {
1699	>	if (tabAt(tab, i) == f) {
1700	>	len = 1;
1701	>	for (Node<V> e = f;; ++len) {
1702	>	Object ek; V ev;
1703	>	if (e.hash == h &&
1704	>	(ev = e.val) != null &&
1705	>	((ek = e.key) == k || k.equals(ek))) {
1706	>	e.val = v;
1707	>	break;
1708	>	}
1709	>	Node<V> last = e;
1710	>	if ((e = e.next) == null) {
1711	>	++delta;
1712	>	last.next = new Node<V>(h, k, v, null);
1713	>	if (len >= TREE_THRESHOLD)
1714	>	replaceWithTreeBin(tab, i, k);
1715	>	break;
1716	>	}
1717	>	}
1718	>	}
1719	>	}
1720	>	if (len != 0) {
1721	>	if (len > 1)
1722	>	addCount(delta, len);
1723	>	break;
1724	>	}
1725	>	}
1726	>	}
1727	>	}
1728	>	} finally {
1729	>	if (delta != 0L)
1730	>	addCount(delta, 2);
1731	>	}
1732	>	if (npe)
1733	>	throw new NullPointerException();
1734	>	}
1735
1736	<	V oldValue = null;
1737	<	if (e != null) {
1738	<	oldValue = e.value;
1739	<	e.value = newValue;
1736	>	/**
1737	>	* Implementation for clear. Steps through each bin, removing all
1738	>	* nodes.
1739	>	*/
1740	>	@SuppressWarnings("unchecked") private final void internalClear() {
1741	>	long delta = 0L; // negative number of deletions
1742	>	int i = 0;
1743	>	Node<V>[] tab = table;
1744	>	while (tab != null && i < tab.length) {
1745	>	Node<V> f = tabAt(tab, i);
1746	>	if (f == null)
1747	>	++i;
1748	>	else if (f.hash < 0) {
1749	>	Object fk;
1750	>	if ((fk = f.key) instanceof TreeBin) {
1751	>	TreeBin<V> t = (TreeBin<V>)fk;
1752	>	t.acquire(0);
1753	>	try {
1754	>	if (tabAt(tab, i) == f) {
1755	>	for (Node<V> p = t.first; p != null; p = p.next) {
1756	>	if (p.val != null) { // (currently always true)
1757	>	p.val = null;
1758	>	--delta;
1759	>	}
1760	>	}
1761	>	t.first = null;
1762	>	t.root = null;
1763	>	++i;
1764	>	}
1765	>	} finally {
1766	>	t.release(0);
1767	>	}
1768	>	}
1769	>	else
1770	>	tab = (Node<V>[])fk;
1771	>	}
1772	>	else {
1773	>	synchronized (f) {
1774	>	if (tabAt(tab, i) == f) {
1775	>	for (Node<V> e = f; e != null; e = e.next) {
1776	>	if (e.val != null) {  // (currently always true)
1777	>	e.val = null;
1778	>	--delta;
1779	>	}
1780	>	}
1781	>	setTabAt(tab, i, null);
1782	>	++i;
1783	>	}
1784		}
405	–	return oldValue;
406	–	} finally {
407	–	unlock();
1785		}
1786		}
1787	+	if (delta != 0L)
1788	+	addCount(delta, -1);
1789	+	}
1790
1791	+	/* ---------------- Table Initialization and Resizing -------------- */
1792
1793	<	V put(K key, int hash, V value, boolean onlyIfAbsent) {
1794	<	lock();
1795	<	try {
1796	<	int c = count;
1797	<	if (c++ > threshold) // ensure capacity
1798	<	rehash();
1799	<	HashEntry<K,V>[] tab = table;
1800	<	int index = hash & (tab.length - 1);
1801	<	HashEntry<K,V> first = tab[index];
1802	<	HashEntry<K,V> e = first;
1803	<	while (e != null && (e.hash != hash || !key.equals(e.key)))
1804	<	e = e.next;
1793	>	/**
1794	>	* Returns a power of two table size for the given desired capacity.
1795	>	* See Hackers Delight, sec 3.2
1796	>	*/
1797	>	private static final int tableSizeFor(int c) {
1798	>	int n = c - 1;
1799	>	n |= n >>> 1;
1800	>	n |= n >>> 2;
1801	>	n |= n >>> 4;
1802	>	n |= n >>> 8;
1803	>	n |= n >>> 16;
1804	>	return (n < 0) ? 1 : (n >= MAXIMUM_CAPACITY) ? MAXIMUM_CAPACITY : n + 1;
1805	>	}
1806
1807	<	V oldValue;
1808	<	if (e != null) {
1809	<	oldValue = e.value;
1810	<	if (!onlyIfAbsent)
1811	<	e.value = value;
1812	<	}
1813	<	else {
1814	<	oldValue = null;
1815	<	++modCount;
1816	<	tab[index] = new HashEntry<K,V>(key, hash, first, value);
1817	<	count = c; // write-volatile
1807	>	/**
1808	>	* Initializes table, using the size recorded in sizeCtl.
1809	>	*/
1810	>	@SuppressWarnings("unchecked") private final Node<V>[] initTable() {
1811	>	Node<V>[] tab; int sc;
1812	>	while ((tab = table) == null) {
1813	>	if ((sc = sizeCtl) < 0)
1814	>	Thread.yield(); // lost initialization race; just spin
1815	>	else if (U.compareAndSwapInt(this, SIZECTL, sc, -1)) {
1816	>	try {
1817	>	if ((tab = table) == null) {
1818	>	int n = (sc > 0) ? sc : DEFAULT_CAPACITY;
1819	>	@SuppressWarnings("rawtypes") Node[] tb = new Node[n];
1820	>	table = tab = (Node<V>[])tb;
1821	>	sc = n - (n >>> 2);
1822	>	}
1823	>	} finally {
1824	>	sizeCtl = sc;
1825		}
1826	<	return oldValue;
438	<	} finally {
439	<	unlock();
1826	>	break;
1827		}
1828		}
1829	+	return tab;
1830	+	}
1831
1832	<	void rehash() {
1833	<	HashEntry<K,V>[] oldTable = table;
1834	<	int oldCapacity = oldTable.length;
1835	<	if (oldCapacity >= MAXIMUM_CAPACITY)
1832	>	/**
1833	>	* Adds to count, and if table is too small and not already
1834	>	* resizing, initiates transfer. If already resizing, helps
1835	>	* perform transfer if work is available.  Rechecks occupancy
1836	>	* after a transfer to see if another resize is already needed
1837	>	* because resizings are lagging additions.
1838	>	*
1839	>	* @param x the count to add
1840	>	* @param check if <0, don't check resize, if <= 1 only check if uncontended
1841	>	*/
1842	>	private final void addCount(long x, int check) {
1843	>	Cell[] as; long b, s;
1844	>	if ((as = counterCells) != null ||
1845	>	!U.compareAndSwapLong(this, BASECOUNT, b = baseCount, s = b + x)) {
1846	>	CellHashCode hc; Cell a; long v; int m;
1847	>	boolean uncontended = true;
1848	>	if ((hc = threadCellHashCode.get()) == null ||
1849	>	as == null || (m = as.length - 1) < 0 ||
1850	>	(a = as[m & hc.code]) == null ||
1851	>	!(uncontended =
1852	>	U.compareAndSwapLong(a, CELLVALUE, v = a.value, v + x))) {
1853	>	fullAddCount(x, hc, uncontended);
1854	>	return;
1855	>	}
1856	>	if (check <= 1)
1857		return;
1858	+	s = sumCount();
1859	+	}
1860	+	if (check >= 0) {
1861	+	Node<V>[] tab, nt; int sc;
1862	+	while (s >= (long)(sc = sizeCtl) && (tab = table) != null &&
1863	+	tab.length < MAXIMUM_CAPACITY) {
1864	+	if (sc < 0) {
1865	+	if (sc == -1 || transferIndex <= transferOrigin ||
1866	+	(nt = nextTable) == null)
1867	+	break;
1868	+	if (U.compareAndSwapInt(this, SIZECTL, sc, sc - 1))
1869	+	transfer(tab, nt);
1870	+	}
1871	+	else if (U.compareAndSwapInt(this, SIZECTL, sc, -2))
1872	+	transfer(tab, null);
1873	+	s = sumCount();
1874	+	}
1875	+	}
1876	+	}
1877
1878	<	/*
1879	<	* Reclassify nodes in each list to new Map.  Because we are
1880	<	* using power-of-two expansion, the elements from each bin
1881	<	* must either stay at same index, or move with a power of two
1882	<	* offset. We eliminate unnecessary node creation by catching
1883	<	* cases where old nodes can be reused because their next
1884	<	* fields won't change. Statistically, at the default
1885	<	* threshold, only about one-sixth of them need cloning when
1886	<	* a table doubles. The nodes they replace will be garbage
1887	<	* collectable as soon as they are no longer referenced by any
1888	<	* reader thread that may be in the midst of traversing table
1889	<	* right now.
1890	<	*/
1891	<
1892	<	HashEntry<K,V>[] newTable = HashEntry.newArray(oldCapacity<<1);
1893	<	threshold = (int)(newTable.length * loadFactor);
1894	<	int sizeMask = newTable.length - 1;
1895	<	for (int i = 0; i < oldCapacity ; i++) {
1896	<	// We need to guarantee that any existing reads of old Map can
1897	<	//  proceed. So we cannot yet null out each bin.
1898	<	HashEntry<K,V> e = oldTable[i];
1899	<
1900	<	if (e != null) {
1901	<	HashEntry<K,V> next = e.next;
1902	<	int idx = e.hash & sizeMask;
1903	<
1904	<	//  Single node on list
1905	<	if (next == null)
1906	<	newTable[idx] = e;
1878	>	/**
1879	>	* Tries to presize table to accommodate the given number of elements.
1880	>	*
1881	>	* @param size number of elements (doesn't need to be perfectly accurate)
1882	>	*/
1883	>	@SuppressWarnings("unchecked") private final void tryPresize(int size) {
1884	>	int c = (size >= (MAXIMUM_CAPACITY >>> 1)) ? MAXIMUM_CAPACITY :
1885	>	tableSizeFor(size + (size >>> 1) + 1);
1886	>	int sc;
1887	>	while ((sc = sizeCtl) >= 0) {
1888	>	Node<V>[] tab = table; int n;
1889	>	if (tab == null || (n = tab.length) == 0) {
1890	>	n = (sc > c) ? sc : c;
1891	>	if (U.compareAndSwapInt(this, SIZECTL, sc, -1)) {
1892	>	try {
1893	>	if (table == tab) {
1894	>	@SuppressWarnings("rawtypes") Node[] tb = new Node[n];
1895	>	table = (Node<V>[])tb;
1896	>	sc = n - (n >>> 2);
1897	>	}
1898	>	} finally {
1899	>	sizeCtl = sc;
1900	>	}
1901	>	}
1902	>	}
1903	>	else if (c <= sc || n >= MAXIMUM_CAPACITY)
1904	>	break;
1905	>	else if (tab == table &&
1906	>	U.compareAndSwapInt(this, SIZECTL, sc, -2))
1907	>	transfer(tab, null);
1908	>	}
1909	>	}
1910
1911	<	else {
1912	<	// Reuse trailing consecutive sequence at same slot
1913	<	HashEntry<K,V> lastRun = e;
1914	<	int lastIdx = idx;
1915	<	for (HashEntry<K,V> last = next;
1916	<	last != null;
1917	<	last = last.next) {
1918	<	int k = last.hash & sizeMask;
1919	<	if (k != lastIdx) {
1920	<	lastIdx = k;
1921	<	lastRun = last;
1911	>	/*
1912	>	* Moves and/or copies the nodes in each bin to new table. See
1913	>	* above for explanation.
1914	>	*/
1915	>	@SuppressWarnings("unchecked") private final void transfer
1916	>	(Node<V>[] tab, Node<V>[] nextTab) {
1917	>	int n = tab.length, stride;
1918	>	if ((stride = (NCPU > 1) ? (n >>> 3) / NCPU : n) < MIN_TRANSFER_STRIDE)
1919	>	stride = MIN_TRANSFER_STRIDE; // subdivide range
1920	>	if (nextTab == null) {            // initiating
1921	>	try {
1922	>	@SuppressWarnings("rawtypes") Node[] tb = new Node[n << 1];
1923	>	nextTab = (Node<V>[])tb;
1924	>	} catch (Throwable ex) {      // try to cope with OOME
1925	>	sizeCtl = Integer.MAX_VALUE;
1926	>	return;
1927	>	}
1928	>	nextTable = nextTab;
1929	>	transferOrigin = n;
1930	>	transferIndex = n;
1931	>	Node<V> rev = new Node<V>(MOVED, tab, null, null);
1932	>	for (int k = n; k > 0;) {    // progressively reveal ready slots
1933	>	int nextk = (k > stride) ? k - stride : 0;
1934	>	for (int m = nextk; m < k; ++m)
1935	>	nextTab[m] = rev;
1936	>	for (int m = n + nextk; m < n + k; ++m)
1937	>	nextTab[m] = rev;
1938	>	U.putOrderedInt(this, TRANSFERORIGIN, k = nextk);
1939	>	}
1940	>	}
1941	>	int nextn = nextTab.length;
1942	>	Node<V> fwd = new Node<V>(MOVED, nextTab, null, null);
1943	>	boolean advance = true;
1944	>	for (int i = 0, bound = 0;;) {
1945	>	int nextIndex, nextBound; Node<V> f; Object fk;
1946	>	while (advance) {
1947	>	if (--i >= bound)
1948	>	advance = false;
1949	>	else if ((nextIndex = transferIndex) <= transferOrigin) {
1950	>	i = -1;
1951	>	advance = false;
1952	>	}
1953	>	else if (U.compareAndSwapInt
1954	>	(this, TRANSFERINDEX, nextIndex,
1955	>	nextBound = (nextIndex > stride ?
1956	>	nextIndex - stride : 0))) {
1957	>	bound = nextBound;
1958	>	i = nextIndex - 1;
1959	>	advance = false;
1960	>	}
1961	>	}
1962	>	if (i < 0 || i >= n || i + n >= nextn) {
1963	>	for (int sc;;) {
1964	>	if (U.compareAndSwapInt(this, SIZECTL, sc = sizeCtl, ++sc)) {
1965	>	if (sc == -1) {
1966	>	nextTable = null;
1967	>	table = nextTab;
1968	>	sizeCtl = (n << 1) - (n >>> 1);
1969	>	}
1970	>	return;
1971	>	}
1972	>	}
1973	>	}
1974	>	else if ((f = tabAt(tab, i)) == null) {
1975	>	if (casTabAt(tab, i, null, fwd)) {
1976	>	setTabAt(nextTab, i, null);
1977	>	setTabAt(nextTab, i + n, null);
1978	>	advance = true;
1979	>	}
1980	>	}
1981	>	else if (f.hash >= 0) {
1982	>	synchronized (f) {
1983	>	if (tabAt(tab, i) == f) {
1984	>	int runBit = f.hash & n;
1985	>	Node<V> lastRun = f, lo = null, hi = null;
1986	>	for (Node<V> p = f.next; p != null; p = p.next) {
1987	>	int b = p.hash & n;
1988	>	if (b != runBit) {
1989	>	runBit = b;
1990	>	lastRun = p;
1991		}
1992		}
1993	<	newTable[lastIdx] = lastRun;
1994	<
1995	<	// Clone all remaining nodes
1996	<	for (HashEntry<K,V> p = e; p != lastRun; p = p.next) {
1997	<	int k = p.hash & sizeMask;
1998	<	HashEntry<K,V> n = newTable[k];
1999	<	newTable[k] = new HashEntry<K,V>(p.key, p.hash,
2000	<	n, p.value);
1993	>	if (runBit == 0)
1994	>	lo = lastRun;
1995	>	else
1996	>	hi = lastRun;
1997	>	for (Node<V> p = f; p != lastRun; p = p.next) {
1998	>	int ph = p.hash;
1999	>	Object pk = p.key; V pv = p.val;
2000	>	if ((ph & n) == 0)
2001	>	lo = new Node<V>(ph, pk, pv, lo);
2002	>	else
2003	>	hi = new Node<V>(ph, pk, pv, hi);
2004		}
2005	+	setTabAt(nextTab, i, lo);
2006	+	setTabAt(nextTab, i + n, hi);
2007	+	setTabAt(tab, i, fwd);
2008	+	advance = true;
2009		}
2010		}
2011		}
2012	<	table = newTable;
2012	>	else if ((fk = f.key) instanceof TreeBin) {
2013	>	TreeBin<V> t = (TreeBin<V>)fk;
2014	>	t.acquire(0);
2015	>	try {
2016	>	if (tabAt(tab, i) == f) {
2017	>	TreeBin<V> lt = new TreeBin<V>();
2018	>	TreeBin<V> ht = new TreeBin<V>();
2019	>	int lc = 0, hc = 0;
2020	>	for (Node<V> e = t.first; e != null; e = e.next) {
2021	>	int h = e.hash;
2022	>	Object k = e.key; V v = e.val;
2023	>	if ((h & n) == 0) {
2024	>	++lc;
2025	>	lt.putTreeNode(h, k, v);
2026	>	}
2027	>	else {
2028	>	++hc;
2029	>	ht.putTreeNode(h, k, v);
2030	>	}
2031	>	}
2032	>	Node<V> ln, hn; // throw away trees if too small
2033	>	if (lc < TREE_THRESHOLD) {
2034	>	ln = null;
2035	>	for (Node<V> p = lt.first; p != null; p = p.next)
2036	>	ln = new Node<V>(p.hash, p.key, p.val, ln);
2037	>	}
2038	>	else
2039	>	ln = new Node<V>(MOVED, lt, null, null);
2040	>	setTabAt(nextTab, i, ln);
2041	>	if (hc < TREE_THRESHOLD) {
2042	>	hn = null;
2043	>	for (Node<V> p = ht.first; p != null; p = p.next)
2044	>	hn = new Node<V>(p.hash, p.key, p.val, hn);
2045	>	}
2046	>	else
2047	>	hn = new Node<V>(MOVED, ht, null, null);
2048	>	setTabAt(nextTab, i + n, hn);
2049	>	setTabAt(tab, i, fwd);
2050	>	advance = true;
2051	>	}
2052	>	} finally {
2053	>	t.release(0);
2054	>	}
2055	>	}
2056	>	else
2057	>	advance = true; // already processed
2058		}
2059	+	}
2060
2061	<	/**
508	<	* Remove; match on key only if value null, else match both.
509	<	*/
510	<	V remove(Object key, int hash, Object value) {
511	<	lock();
512	<	try {
513	<	int c = count - 1;
514	<	HashEntry<K,V>[] tab = table;
515	<	int index = hash & (tab.length - 1);
516	<	HashEntry<K,V> first = tab[index];
517	<	HashEntry<K,V> e = first;
518	<	while (e != null && (e.hash != hash || !key.equals(e.key)))
519	<	e = e.next;
2061	>	/* ---------------- Counter support -------------- */
2062
2063	<	V oldValue = null;
2064	<	if (e != null) {
2065	<	V v = e.value;
2066	<	if (value == null || value.equals(v)) {
2067	<	oldValue = v;
2068	<	// All entries following removed node can stay
2069	<	// in list, but all preceding ones need to be
528	<	// cloned.
529	<	++modCount;
530	<	HashEntry<K,V> newFirst = e.next;
531	<	for (HashEntry<K,V> p = first; p != e; p = p.next)
532	<	newFirst = new HashEntry<K,V>(p.key, p.hash,
533	<	newFirst, p.value);
534	<	tab[index] = newFirst;
535	<	count = c; // write-volatile
536	<	}
537	<	}
538	<	return oldValue;
539	<	} finally {
540	<	unlock();
2063	>	final long sumCount() {
2064	>	Cell[] as = counterCells; Cell a;
2065	>	long sum = baseCount;
2066	>	if (as != null) {
2067	>	for (int i = 0; i < as.length; ++i) {
2068	>	if ((a = as[i]) != null)
2069	>	sum += a.value;
2070		}
2071		}
2072	+	return sum;
2073	+	}
2074
2075	<	void clear() {
2076	<	if (count != 0) {
2077	<	lock();
2078	<	try {
2079	<	HashEntry<K,V>[] tab = table;
2080	<	for (int i = 0; i < tab.length ; i++)
2081	<	tab[i] = null;
2082	<	++modCount;
2083	<	count = 0; // write-volatile
2075	>	// See LongAdder version for explanation
2076	>	private final void fullAddCount(long x, CellHashCode hc,
2077	>	boolean wasUncontended) {
2078	>	int h;
2079	>	if (hc == null) {
2080	>	hc = new CellHashCode();
2081	>	int s = cellHashCodeGenerator.addAndGet(SEED_INCREMENT);
2082	>	h = hc.code = (s == 0) ? 1 : s; // Avoid zero
2083	>	threadCellHashCode.set(hc);
2084	>	}
2085	>	else
2086	>	h = hc.code;
2087	>	boolean collide = false;                // True if last slot nonempty
2088	>	for (;;) {
2089	>	Cell[] as; Cell a; int n; long v;
2090	>	if ((as = counterCells) != null && (n = as.length) > 0) {
2091	>	if ((a = as[(n - 1) & h]) == null) {
2092	>	if (cellsBusy == 0) {            // Try to attach new Cell
2093	>	Cell r = new Cell(x); // Optimistic create
2094	>	if (cellsBusy == 0 &&
2095	>	U.compareAndSwapInt(this, CELLSBUSY, 0, 1)) {
2096	>	boolean created = false;
2097	>	try {               // Recheck under lock
2098	>	Cell[] rs; int m, j;
2099	>	if ((rs = counterCells) != null &&
2100	>	(m = rs.length) > 0 &&
2101	>	rs[j = (m - 1) & h] == null) {
2102	>	rs[j] = r;
2103	>	created = true;
2104	>	}
2105	>	} finally {
2106	>	cellsBusy = 0;
2107	>	}
2108	>	if (created)
2109	>	break;
2110	>	continue;           // Slot is now non-empty
2111	>	}
2112	>	}
2113	>	collide = false;
2114	>	}
2115	>	else if (!wasUncontended)       // CAS already known to fail
2116	>	wasUncontended = true;      // Continue after rehash
2117	>	else if (U.compareAndSwapLong(a, CELLVALUE, v = a.value, v + x))
2118	>	break;
2119	>	else if (counterCells != as || n >= NCPU)
2120	>	collide = false;            // At max size or stale
2121	>	else if (!collide)
2122	>	collide = true;
2123	>	else if (cellsBusy == 0 &&
2124	>	U.compareAndSwapInt(this, CELLSBUSY, 0, 1)) {
2125	>	try {
2126	>	if (counterCells == as) {// Expand table unless stale
2127	>	Cell[] rs = new Cell[n << 1];
2128	>	for (int i = 0; i < n; ++i)
2129	>	rs[i] = as[i];
2130	>	counterCells = rs;
2131	>	}
2132	>	} finally {
2133	>	cellsBusy = 0;
2134	>	}
2135	>	collide = false;
2136	>	continue;                   // Retry with expanded table
2137	>	}
2138	>	h ^= h << 13;                   // Rehash
2139	>	h ^= h >>> 17;
2140	>	h ^= h << 5;
2141	>	}
2142	>	else if (cellsBusy == 0 && counterCells == as &&
2143	>	U.compareAndSwapInt(this, CELLSBUSY, 0, 1)) {
2144	>	boolean init = false;
2145	>	try {                           // Initialize table
2146	>	if (counterCells == as) {
2147	>	Cell[] rs = new Cell[2];
2148	>	rs[h & 1] = new Cell(x);
2149	>	counterCells = rs;
2150	>	init = true;
2151	>	}
2152		} finally {
2153	<	unlock();
2153	>	cellsBusy = 0;
2154		}
2155	+	if (init)
2156	+	break;
2157		}
2158	+	else if (U.compareAndSwapLong(this, BASECOUNT, v = baseCount, v + x))
2159	+	break;                          // Fall back on using base
2160		}
2161	+	hc.code = h;                            // Record index for next time
2162		}
2163
2164	+	/* ----------------Table Traversal -------------- */
2165
2166	+	/**
2167	+	* Encapsulates traversal for methods such as containsValue; also
2168	+	* serves as a base class for other iterators and bulk tasks.
2169	+	*
2170	+	* At each step, the iterator snapshots the key ("nextKey") and
2171	+	* value ("nextVal") of a valid node (i.e., one that, at point of
2172	+	* snapshot, has a non-null user value). Because val fields can
2173	+	* change (including to null, indicating deletion), field nextVal
2174	+	* might not be accurate at point of use, but still maintains the
2175	+	* weak consistency property of holding a value that was once
2176	+	* valid. To support iterator.remove, the nextKey field is not
2177	+	* updated (nulled out) when the iterator cannot advance.
2178	+	*
2179	+	* Internal traversals directly access these fields, as in:
2180	+	* {@code while (it.advance() != null) { process(it.nextKey); }}
2181	+	*
2182	+	* Exported iterators must track whether the iterator has advanced
2183	+	* (in hasNext vs next) (by setting/checking/nulling field
2184	+	* nextVal), and then extract key, value, or key-value pairs as
2185	+	* return values of next().
2186	+	*
2187	+	* The iterator visits once each still-valid node that was
2188	+	* reachable upon iterator construction. It might miss some that
2189	+	* were added to a bin after the bin was visited, which is OK wrt
2190	+	* consistency guarantees. Maintaining this property in the face
2191	+	* of possible ongoing resizes requires a fair amount of
2192	+	* bookkeeping state that is difficult to optimize away amidst
2193	+	* volatile accesses.  Even so, traversal maintains reasonable
2194	+	* throughput.
2195	+	*
2196	+	* Normally, iteration proceeds bin-by-bin traversing lists.
2197	+	* However, if the table has been resized, then all future steps
2198	+	* must traverse both the bin at the current index as well as at
2199	+	* (index + baseSize); and so on for further resizings. To
2200	+	* paranoically cope with potential sharing by users of iterators
2201	+	* across threads, iteration terminates if a bounds checks fails
2202	+	* for a table read.
2203	+	*
2204	+	* This class supports both Spliterator-based traversal and
2205	+	* CountedCompleter-based bulk tasks. The same "batch" field is
2206	+	* used, but in slightly different ways, in the two cases.  For
2207	+	* Spliterators, it is a saturating (at Integer.MAX_VALUE)
2208	+	* estimate of element coverage. For CHM tasks, it is a pre-scaled
2209	+	* size that halves down to zero for leaf tasks, that is only
2210	+	* computed upon execution of the task. (Tasks can be submitted to
2211	+	* any pool, of any size, so we don't know scale factors until
2212	+	* running.)
2213	+	*
2214	+	* This class extends CountedCompleter to streamline parallel
2215	+	* iteration in bulk operations. This adds only a few fields of
2216	+	* space overhead, which is small enough in cases where it is not
2217	+	* needed to not worry about it.  Because CountedCompleter is
2218	+	* Serializable, but iterators need not be, we need to add warning
2219	+	* suppressions.
2220	+	*/
2221	+	@SuppressWarnings("serial") static class Traverser<K,V,R>
2222	+	extends CountedCompleter<R> {
2223	+	final ConcurrentHashMap<K, V> map;
2224	+	Node<V> next;        // the next entry to use
2225	+	Object nextKey;      // cached key field of next
2226	+	V nextVal;           // cached val field of next
2227	+	Node<V>[] tab;       // current table; updated if resized
2228	+	int index;           // index of bin to use next
2229	+	int baseIndex;       // current index of initial table
2230	+	int baseLimit;       // index bound for initial table
2231	+	int baseSize;        // initial table size
2232	+	int batch;           // split control
2233	+	/** Creates iterator for all entries in the table. */
2234	+	Traverser(ConcurrentHashMap<K, V> map) {
2235	+	this.map = map;
2236	+	Node<V>[] t;
2237	+	if ((t = tab = map.table) != null)
2238	+	baseLimit = baseSize = t.length;
2239	+	}
2240	+
2241	+	/** Task constructor */
2242	+	Traverser(ConcurrentHashMap<K,V> map, Traverser<K,V,?> it, int batch) {
2243	+	super(it);
2244	+	this.map = map;
2245	+	this.batch = batch; // -1 if unknown
2246	+	if (it == null) {
2247	+	Node<V>[] t;
2248	+	if ((t = tab = map.table) != null)
2249	+	baseLimit = baseSize = t.length;
2250	+	}
2251	+	else { // split parent
2252	+	this.tab = it.tab;
2253	+	this.baseSize = it.baseSize;
2254	+	int hi = this.baseLimit = it.baseLimit;
2255	+	it.baseLimit = this.index = this.baseIndex =
2256	+	(hi + it.baseIndex + 1) >>> 1;
2257	+	}
2258	+	}
2259	+
2260	+	/** Spliterator constructor */
2261	+	Traverser(ConcurrentHashMap<K,V> map, Traverser<K,V,?> it) {
2262	+	super(it);
2263	+	this.map = map;
2264	+	if (it == null) {
2265	+	Node<V>[] t;
2266	+	if ((t = tab = map.table) != null)
2267	+	baseLimit = baseSize = t.length;
2268	+	long n = map.sumCount();
2269	+	batch = ((n > (long)Integer.MAX_VALUE) ? Integer.MAX_VALUE :
2270	+	(int)n);
2271	+	}
2272	+	else {
2273	+	this.tab = it.tab;
2274	+	this.baseSize = it.baseSize;
2275	+	int hi = this.baseLimit = it.baseLimit;
2276	+	it.baseLimit = this.index = this.baseIndex =
2277	+	(hi + it.baseIndex + 1) >>> 1;
2278	+	this.batch = it.batch >>>= 1;
2279	+	}
2280	+	}
2281	+
2282	+	/**
2283	+	* Advances next; returns nextVal or null if terminated.
2284	+	* See above for explanation.
2285	+	*/
2286	+	@SuppressWarnings("unchecked") final V advance() {
2287	+	Node<V> e = next;
2288	+	V ev = null;
2289	+	outer: do {
2290	+	if (e != null)                  // advance past used/skipped node
2291	+	e = e.next;
2292	+	while (e == null) {             // get to next non-null bin
2293	+	ConcurrentHashMap<K, V> m;
2294	+	Node<V>[] t; int b, i, n; Object ek; //  must use locals
2295	+	if ((t = tab) != null)
2296	+	n = t.length;
2297	+	else if ((m = map) != null && (t = tab = m.table) != null)
2298	+	n = baseLimit = baseSize = t.length;
2299	+	else
2300	+	break outer;
2301	+	if ((b = baseIndex) >= baseLimit ||
2302	+	(i = index) < 0 || i >= n)
2303	+	break outer;
2304	+	if ((e = tabAt(t, i)) != null && e.hash < 0) {
2305	+	if ((ek = e.key) instanceof TreeBin)
2306	+	e = ((TreeBin<V>)ek).first;
2307	+	else {
2308	+	tab = (Node<V>[])ek;
2309	+	continue;           // restarts due to null val
2310	+	}
2311	+	}                           // visit upper slots if present
2312	+	index = (i += baseSize) < n ? i : (baseIndex = b + 1);
2313	+	}
2314	+	nextKey = e.key;
2315	+	} while ((ev = e.val) == null);    // skip deleted or special nodes
2316	+	next = e;
2317	+	return nextVal = ev;
2318	+	}
2319	+
2320	+	public final void remove() {
2321	+	Object k = nextKey;
2322	+	if (k == null && (advance() == null || (k = nextKey) == null))
2323	+	throw new IllegalStateException();
2324	+	map.internalReplace(k, null, null);
2325	+	}
2326	+
2327	+	public final boolean hasNext() {
2328	+	return nextVal != null || advance() != null;
2329	+	}
2330	+
2331	+	public final boolean hasMoreElements() { return hasNext(); }
2332	+
2333	+	public void compute() { } // default no-op CountedCompleter body
2334	+
2335	+	/**
2336	+	* Returns a batch value > 0 if this task should (and must) be
2337	+	* split, if so, adding to pending count, and in any case
2338	+	* updating batch value. The initial batch value is approx
2339	+	* exp2 of the number of times (minus one) to split task by
2340	+	* two before executing leaf action. This value is faster to
2341	+	* compute and more convenient to use as a guide to splitting
2342	+	* than is the depth, since it is used while dividing by two
2343	+	* anyway.
2344	+	*/
2345	+	final int preSplit() {
2346	+	int b;  ForkJoinPool pool;
2347	+	if ((b = batch) < 0) { // force initialization
2348	+	int sp = (((pool = getPool()) == null) ?
2349	+	ForkJoinPool.getCommonPoolParallelism() :
2350	+	pool.getParallelism()) << 3; // slack of 8
2351	+	long n = map.sumCount();
2352	+	b = (n <= 0L) ? 0 : (n < (long)sp) ? (int)n : sp;
2353	+	}
2354	+	b = (b <= 1 || baseIndex == baseLimit) ? 0 : (b >>> 1);
2355	+	if ((batch = b) > 0)
2356	+	addToPendingCount(1);
2357	+	return b;
2358	+	}
2359	+
2360	+	// spliterator support
2361	+
2362	+	public long exactSizeIfKnown() {
2363	+	return -1;
2364	+	}
2365	+
2366	+	public boolean hasExactSplits() {
2367	+	return false;
2368	+	}
2369	+
2370	+	public long estimateSize() {
2371	+	return batch;
2372	+	}
2373	+	}
2374
2375		/* ---------------- Public operations -------------- */
2376
2377		/**
2378	<	* Creates a new, empty map with the specified initial
2379	<	* capacity, load factor and concurrency level.
2378	>	* Creates a new, empty map with the default initial table size (16).
2379	>	*/
2380	>	public ConcurrentHashMap() {
2381	>	}
2382	>
2383	>	/**
2384	>	* Creates a new, empty map with an initial table size
2385	>	* accommodating the specified number of elements without the need
2386	>	* to dynamically resize.
2387		*
2388	<	* @param initialCapacity the initial capacity. The implementation
2389	<	* performs internal sizing to accommodate this many elements.
2390	<	* @param loadFactor  the load factor threshold, used to control resizing.
2391	<	* Resizing may be performed when the average number of elements per
572	<	* bin exceeds this threshold.
573	<	* @param concurrencyLevel the estimated number of concurrently
574	<	* updating threads. The implementation performs internal sizing
575	<	* to try to accommodate this many threads.
576	<	* @throws IllegalArgumentException if the initial capacity is
577	<	* negative or the load factor or concurrencyLevel are
578	<	* nonpositive.
2388	>	* @param initialCapacity The implementation performs internal
2389	>	* sizing to accommodate this many elements.
2390	>	* @throws IllegalArgumentException if the initial capacity of
2391	>	* elements is negative
2392		*/
2393	<	public ConcurrentHashMap(int initialCapacity,
2394	<	float loadFactor, int concurrencyLevel) {
582	<	if (!(loadFactor > 0) || initialCapacity < 0 || concurrencyLevel <= 0)
2393	>	public ConcurrentHashMap(int initialCapacity) {
2394	>	if (initialCapacity < 0)
2395		throw new IllegalArgumentException();
2396	+	int cap = ((initialCapacity >= (MAXIMUM_CAPACITY >>> 1)) ?
2397	+	MAXIMUM_CAPACITY :
2398	+	tableSizeFor(initialCapacity + (initialCapacity >>> 1) + 1));
2399	+	this.sizeCtl = cap;
2400	+	}
2401
2402	<	if (concurrencyLevel > MAX_SEGMENTS)
2403	<	concurrencyLevel = MAX_SEGMENTS;
2404	<
2405	<	// Find power-of-two sizes best matching arguments
2406	<	int sshift = 0;
2407	<	int ssize = 1;
2408	<	while (ssize < concurrencyLevel) {
2409	<	++sshift;
593	<	ssize <<= 1;
594	<	}
595	<	segmentShift = 32 - sshift;
596	<	segmentMask = ssize - 1;
597	<	this.segments = Segment.newArray(ssize);
598	<
599	<	if (initialCapacity > MAXIMUM_CAPACITY)
600	<	initialCapacity = MAXIMUM_CAPACITY;
601	<	int c = initialCapacity / ssize;
602	<	if (c * ssize < initialCapacity)
603	<	++c;
604	<	int cap = 1;
605	<	while (cap < c)
606	<	cap <<= 1;
607	<
608	<	for (int i = 0; i < this.segments.length; ++i)
609	<	this.segments[i] = new Segment<K,V>(cap, loadFactor);
2402	>	/**
2403	>	* Creates a new map with the same mappings as the given map.
2404	>	*
2405	>	* @param m the map
2406	>	*/
2407	>	public ConcurrentHashMap(Map<? extends K, ? extends V> m) {
2408	>	this.sizeCtl = DEFAULT_CAPACITY;
2409	>	internalPutAll(m);
2410		}
2411
2412		/**
2413	<	* Creates a new, empty map with the specified initial capacity
2414	<	* and load factor and with the default concurrencyLevel (16).
2413	>	* Creates a new, empty map with an initial table size based on
2414	>	* the given number of elements ({@code initialCapacity}) and
2415	>	* initial table density ({@code loadFactor}).
2416		*
2417	<	* @param initialCapacity The implementation performs internal
2418	<	* sizing to accommodate this many elements.
2419	<	* @param loadFactor  the load factor threshold, used to control resizing.
2420	<	* Resizing may be performed when the average number of elements per
2421	<	* bin exceeds this threshold.
2417	>	* @param initialCapacity the initial capacity. The implementation
2418	>	* performs internal sizing to accommodate this many elements,
2419	>	* given the specified load factor.
2420	>	* @param loadFactor the load factor (table density) for
2421	>	* establishing the initial table size
2422		* @throws IllegalArgumentException if the initial capacity of
2423		* elements is negative or the load factor is nonpositive
2424		*
2425		* @since 1.6
2426		*/
2427		public ConcurrentHashMap(int initialCapacity, float loadFactor) {
2428	<	this(initialCapacity, loadFactor, DEFAULT_CONCURRENCY_LEVEL);
2428	>	this(initialCapacity, loadFactor, 1);
2429		}
2430
2431		/**
2432	<	* Creates a new, empty map with the specified initial capacity,
2433	<	* and with default load factor (0.75) and concurrencyLevel (16).
2432	>	* Creates a new, empty map with an initial table size based on
2433	>	* the given number of elements ({@code initialCapacity}), table
2434	>	* density ({@code loadFactor}), and number of concurrently
2435	>	* updating threads ({@code concurrencyLevel}).
2436		*
2437		* @param initialCapacity the initial capacity. The implementation
2438	<	* performs internal sizing to accommodate this many elements.
2439	<	* @throws IllegalArgumentException if the initial capacity of
2440	<	* elements is negative.
2438	>	* performs internal sizing to accommodate this many elements,
2439	>	* given the specified load factor.
2440	>	* @param loadFactor the load factor (table density) for
2441	>	* establishing the initial table size
2442	>	* @param concurrencyLevel the estimated number of concurrently
2443	>	* updating threads. The implementation may use this value as
2444	>	* a sizing hint.
2445	>	* @throws IllegalArgumentException if the initial capacity is
2446	>	* negative or the load factor or concurrencyLevel are
2447	>	* nonpositive
2448		*/
2449	<	public ConcurrentHashMap(int initialCapacity) {
2450	<	this(initialCapacity, DEFAULT_LOAD_FACTOR, DEFAULT_CONCURRENCY_LEVEL);
2449	>	public ConcurrentHashMap(int initialCapacity,
2450	>	float loadFactor, int concurrencyLevel) {
2451	>	if (!(loadFactor > 0.0f) || initialCapacity < 0 || concurrencyLevel <= 0)
2452	>	throw new IllegalArgumentException();
2453	>	if (initialCapacity < concurrencyLevel)   // Use at least as many bins
2454	>	initialCapacity = concurrencyLevel;   // as estimated threads
2455	>	long size = (long)(1.0 + (long)initialCapacity / loadFactor);
2456	>	int cap = (size >= (long)MAXIMUM_CAPACITY) ?
2457	>	MAXIMUM_CAPACITY : tableSizeFor((int)size);
2458	>	this.sizeCtl = cap;
2459		}
2460
2461		/**
2462	<	* Creates a new, empty map with a default initial capacity (16),
2463	<	* load factor (0.75) and concurrencyLevel (16).
2462	>	* Creates a new {@link Set} backed by a ConcurrentHashMap
2463	>	* from the given type to {@code Boolean.TRUE}.
2464	>	*
2465	>	* @return the new set
2466		*/
2467	<	public ConcurrentHashMap() {
2468	<	this(DEFAULT_INITIAL_CAPACITY, DEFAULT_LOAD_FACTOR, DEFAULT_CONCURRENCY_LEVEL);
2467	>	public static <K> KeySetView<K,Boolean> newKeySet() {
2468	>	return new KeySetView<K,Boolean>(new ConcurrentHashMap<K,Boolean>(),
2469	>	Boolean.TRUE);
2470		}
2471
2472		/**
2473	<	* Creates a new map with the same mappings as the given map.
2474	<	* The map is created with a capacity of 1.5 times the number
654	<	* of mappings in the given map or 16 (whichever is greater),
655	<	* and a default load factor (0.75) and concurrencyLevel (16).
2473	>	* Creates a new {@link Set} backed by a ConcurrentHashMap
2474	>	* from the given type to {@code Boolean.TRUE}.
2475		*
2476	<	* @param m the map
2476	>	* @param initialCapacity The implementation performs internal
2477	>	* sizing to accommodate this many elements.
2478	>	* @throws IllegalArgumentException if the initial capacity of
2479	>	* elements is negative
2480	>	* @return the new set
2481		*/
2482	<	public ConcurrentHashMap(Map<? extends K, ? extends V> m) {
2483	<	this(Math.max((int) (m.size() / DEFAULT_LOAD_FACTOR) + 1,
2484	<	DEFAULT_INITIAL_CAPACITY),
662	<	DEFAULT_LOAD_FACTOR, DEFAULT_CONCURRENCY_LEVEL);
663	<	putAll(m);
2482	>	public static <K> KeySetView<K,Boolean> newKeySet(int initialCapacity) {
2483	>	return new KeySetView<K,Boolean>
2484	>	(new ConcurrentHashMap<K,Boolean>(initialCapacity), Boolean.TRUE);
2485		}
2486
2487		/**
2488	<	* Returns <tt>true</tt> if this map contains no key-value mappings.
668	<	*
669	<	* @return <tt>true</tt> if this map contains no key-value mappings
2488	>	* {@inheritDoc}
2489		*/
2490		public boolean isEmpty() {
2491	<	final Segment<K,V>[] segments = this.segments;
673	<	/*
674	<	* We keep track of per-segment modCounts to avoid ABA
675	<	* problems in which an element in one segment was added and
676	<	* in another removed during traversal, in which case the
677	<	* table was never actually empty at any point. Note the
678	<	* similar use of modCounts in the size() and containsValue()
679	<	* methods, which are the only other methods also susceptible
680	<	* to ABA problems.
681	<	*/
682	<	int[] mc = new int[segments.length];
683	<	int mcsum = 0;
684	<	for (int i = 0; i < segments.length; ++i) {
685	<	if (segments[i].count != 0)
686	<	return false;
687	<	else
688	<	mcsum += mc[i] = segments[i].modCount;
689	<	}
690	<	// If mcsum happens to be zero, then we know we got a snapshot
691	<	// before any modifications at all were made.  This is
692	<	// probably common enough to bother tracking.
693	<	if (mcsum != 0) {
694	<	for (int i = 0; i < segments.length; ++i) {
695	<	if (segments[i].count != 0 ||
696	<	mc[i] != segments[i].modCount)
697	<	return false;
698	<	}
699	<	}
700	<	return true;
2491	>	return sumCount() <= 0L; // ignore transient negative values
2492		}
2493
2494		/**
2495	<	* Returns the number of key-value mappings in this map.  If the
705	<	* map contains more than <tt>Integer.MAX_VALUE</tt> elements, returns
706	<	* <tt>Integer.MAX_VALUE</tt>.
707	<	*
708	<	* @return the number of key-value mappings in this map
2495	>	* {@inheritDoc}
2496		*/
2497		public int size() {
2498	<	final Segment<K,V>[] segments = this.segments;
2499	<	long sum = 0;
2500	<	long check = 0;
2501	<	int[] mc = new int[segments.length];
2502	<	// Try a few times to get accurate count. On failure due to
2503	<	// continuous async changes in table, resort to locking.
2504	<	for (int k = 0; k < RETRIES_BEFORE_LOCK; ++k) {
2505	<	check = 0;
2506	<	sum = 0;
2507	<	int mcsum = 0;
2508	<	for (int i = 0; i < segments.length; ++i) {
2509	<	sum += segments[i].count;
2510	<	mcsum += mc[i] = segments[i].modCount;
2511	<	}
2512	<	if (mcsum != 0) {
2513	<	for (int i = 0; i < segments.length; ++i) {
2514	<	check += segments[i].count;
2515	<	if (mc[i] != segments[i].modCount) {
729	<	check = -1; // force retry
730	<	break;
731	<	}
732	<	}
733	<	}
734	<	if (check == sum)
735	<	break;
736	<	}
737	<	if (check != sum) { // Resort to locking all segments
738	<	sum = 0;
739	<	for (int i = 0; i < segments.length; ++i)
740	<	segments[i].lock();
741	<	for (int i = 0; i < segments.length; ++i)
742	<	sum += segments[i].count;
743	<	for (int i = 0; i < segments.length; ++i)
744	<	segments[i].unlock();
745	<	}
746	<	if (sum > Integer.MAX_VALUE)
747	<	return Integer.MAX_VALUE;
748	<	else
749	<	return (int)sum;
2498	>	long n = sumCount();
2499	>	return ((n < 0L) ? 0 :
2500	>	(n > (long)Integer.MAX_VALUE) ? Integer.MAX_VALUE :
2501	>	(int)n);
2502	>	}
2503	>
2504	>	/**
2505	>	* Returns the number of mappings. This method should be used
2506	>	* instead of {@link #size} because a ConcurrentHashMap may
2507	>	* contain more mappings than can be represented as an int. The
2508	>	* value returned is an estimate; the actual count may differ if
2509	>	* there are concurrent insertions or removals.
2510	>	*
2511	>	* @return the number of mappings
2512	>	*/
2513	>	public long mappingCount() {
2514	>	long n = sumCount();
2515	>	return (n < 0L) ? 0L : n; // ignore transient negative values
2516		}
2517
2518		/**
2519	<	* Returns the value to which this map maps the specified key, or
2520	<	* <tt>null</tt> if the map contains no mapping for the key.
2519	>	* Returns the value to which the specified key is mapped,
2520	>	* or {@code null} if this map contains no mapping for the key.
2521	>	*
2522	>	* <p>More formally, if this map contains a mapping from a key
2523	>	* {@code k} to a value {@code v} such that {@code key.equals(k)},
2524	>	* then this method returns {@code v}; otherwise it returns
2525	>	* {@code null}.  (There can be at most one such mapping.)
2526		*
756	–	* @param key key whose associated value is to be returned
757	–	* @return the value to which this map maps the specified key, or
758	–	*         <tt>null</tt> if the map contains no mapping for the key
2527		* @throws NullPointerException if the specified key is null
2528		*/
2529		public V get(Object key) {
2530	<	int hash = hash(key); // throws NullPointerException if key null
2531	<	return segmentFor(hash).get(key, hash);
2530	>	return internalGet(key);
2531	>	}
2532	>
2533	>	/**
2534	>	* Returns the value to which the specified key is mapped,
2535	>	* or the given defaultValue if this map contains no mapping for the key.
2536	>	*
2537	>	* @param key the key
2538	>	* @param defaultValue the value to return if this map contains
2539	>	* no mapping for the given key
2540	>	* @return the mapping for the key, if present; else the defaultValue
2541	>	* @throws NullPointerException if the specified key is null
2542	>	*/
2543	>	public V getValueOrDefault(Object key, V defaultValue) {
2544	>	V v;
2545	>	return (v = internalGet(key)) == null ? defaultValue : v;
2546		}
2547
2548		/**
2549		* Tests if the specified object is a key in this table.
2550		*
2551		* @param  key   possible key
2552	<	* @return <tt>true</tt> if and only if the specified object
2552	>	* @return {@code true} if and only if the specified object
2553		*         is a key in this table, as determined by the
2554	<	*         <tt>equals</tt> method; <tt>false</tt> otherwise.
2554	>	*         {@code equals} method; {@code false} otherwise
2555		* @throws NullPointerException if the specified key is null
2556		*/
2557		public boolean containsKey(Object key) {
2558	<	int hash = hash(key); // throws NullPointerException if key null
777	<	return segmentFor(hash).containsKey(key, hash);
2558	>	return internalGet(key) != null;
2559		}
2560
2561		/**
2562	<	* Returns <tt>true</tt> if this map maps one or more keys to the
2563	<	* specified value. Note: This method requires a full internal
2564	<	* traversal of the hash table, and so is much slower than
784	<	* method <tt>containsKey</tt>.
2562	>	* Returns {@code true} if this map maps one or more keys to the
2563	>	* specified value. Note: This method may require a full traversal
2564	>	* of the map, and is much slower than method {@code containsKey}.
2565		*
2566		* @param value value whose presence in this map is to be tested
2567	<	* @return <tt>true</tt> if this map maps one or more keys to the
2567	>	* @return {@code true} if this map maps one or more keys to the
2568		*         specified value
2569		* @throws NullPointerException if the specified value is null
2570		*/
2571		public boolean containsValue(Object value) {
2572		if (value == null)
2573		throw new NullPointerException();
2574	<
2575	<	// See explanation of modCount use above
2576	<
2577	<	final Segment<K,V>[] segments = this.segments;
2578	<	int[] mc = new int[segments.length];
799	<
800	<	// Try a few times without locking
801	<	for (int k = 0; k < RETRIES_BEFORE_LOCK; ++k) {
802	<	int sum = 0;
803	<	int mcsum = 0;
804	<	for (int i = 0; i < segments.length; ++i) {
805	<	int c = segments[i].count;
806	<	mcsum += mc[i] = segments[i].modCount;
807	<	if (segments[i].containsValue(value))
808	<	return true;
809	<	}
810	<	boolean cleanSweep = true;
811	<	if (mcsum != 0) {
812	<	for (int i = 0; i < segments.length; ++i) {
813	<	int c = segments[i].count;
814	<	if (mc[i] != segments[i].modCount) {
815	<	cleanSweep = false;
816	<	break;
817	<	}
818	<	}
819	<	}
820	<	if (cleanSweep)
821	<	return false;
822	<	}
823	<	// Resort to locking all segments
824	<	for (int i = 0; i < segments.length; ++i)
825	<	segments[i].lock();
826	<	boolean found = false;
827	<	try {
828	<	for (int i = 0; i < segments.length; ++i) {
829	<	if (segments[i].containsValue(value)) {
830	<	found = true;
831	<	break;
832	<	}
833	<	}
834	<	} finally {
835	<	for (int i = 0; i < segments.length; ++i)
836	<	segments[i].unlock();
2574	>	V v;
2575	>	Traverser<K,V,Object> it = new Traverser<K,V,Object>(this);
2576	>	while ((v = it.advance()) != null) {
2577	>	if (v == value || value.equals(v))
2578	>	return true;
2579		}
2580	<	return found;
2580	>	return false;
2581		}
2582
2583		/**
#	Line 845 | Line 2587 | public class ConcurrentHashMap<K, V> ext
2587		* full compatibility with class {@link java.util.Hashtable},
2588		* which supported this method prior to introduction of the
2589		* Java Collections framework.
2590	<
2590	>	*
2591		* @param  value a value to search for
2592	<	* @return <tt>true</tt> if and only if some key maps to the
2593	<	*         <tt>value</tt> argument in this table as
2594	<	*         determined by the <tt>equals</tt> method;
2595	<	*         <tt>false</tt> otherwise
2592	>	* @return {@code true} if and only if some key maps to the
2593	>	*         {@code value} argument in this table as
2594	>	*         determined by the {@code equals} method;
2595	>	*         {@code false} otherwise
2596		* @throws NullPointerException if the specified value is null
2597		*/
2598	<	public boolean contains(Object value) {
2598	>	@Deprecated public boolean contains(Object value) {
2599		return containsValue(value);
2600		}
2601
#	Line 861 | Line 2603 | public class ConcurrentHashMap<K, V> ext
2603		* Maps the specified key to the specified value in this table.
2604		* Neither the key nor the value can be null.
2605		*
2606	<	* <p> The value can be retrieved by calling the <tt>get</tt> method
2606	>	* <p>The value can be retrieved by calling the {@code get} method
2607		* with a key that is equal to the original key.
2608		*
2609		* @param key key with which the specified value is to be associated
2610		* @param value value to be associated with the specified key
2611	<	* @return the previous value associated with <tt>key</tt>, or
2612	<	*         <tt>null</tt> if there was no mapping for <tt>key</tt>
2611	>	* @return the previous value associated with {@code key}, or
2612	>	*         {@code null} if there was no mapping for {@code key}
2613		* @throws NullPointerException if the specified key or value is null
2614		*/
2615		public V put(K key, V value) {
2616	<	if (value == null)
875	<	throw new NullPointerException();
876	<	int hash = hash(key);
877	<	return segmentFor(hash).put(key, hash, value, false);
2616	>	return internalPut(key, value, false);
2617		}
2618
2619		/**
2620		* {@inheritDoc}
2621		*
2622		* @return the previous value associated with the specified key,
2623	<	*         or <tt>null</tt> if there was no mapping for the key
2623	>	*         or {@code null} if there was no mapping for the key
2624		* @throws NullPointerException if the specified key or value is null
2625		*/
2626		public V putIfAbsent(K key, V value) {
2627	<	if (value == null)
889	<	throw new NullPointerException();
890	<	int hash = hash(key);
891	<	return segmentFor(hash).put(key, hash, value, true);
2627	>	return internalPut(key, value, true);
2628		}
2629
2630		/**
#	Line 899 | Line 2635 | public class ConcurrentHashMap<K, V> ext
2635		* @param m mappings to be stored in this map
2636		*/
2637		public void putAll(Map<? extends K, ? extends V> m) {
2638	<	for (Iterator<? extends Map.Entry<? extends K, ? extends V>> it = (Iterator<? extends Map.Entry<? extends K, ? extends V>>) m.entrySet().iterator(); it.hasNext(); ) {
2639	<	Entry<? extends K, ? extends V> e = it.next();
2640	<	put(e.getKey(), e.getValue());
2641	<	}
2638	>	internalPutAll(m);
2639	>	}
2640	>
2641	>	/**
2642	>	* If the specified key is not already associated with a value (or
2643	>	* is mapped to {@code null}), attempts to compute its value using
2644	>	* the given mapping function and enters it into this map unless
2645	>	* {@code null}. The entire method invocation is performed
2646	>	* atomically, so the function is applied at most once per key.
2647	>	* Some attempted update operations on this map by other threads
2648	>	* may be blocked while computation is in progress, so the
2649	>	* computation should be short and simple, and must not attempt to
2650	>	* update any other mappings of this Map.
2651	>	*
2652	>	* @param key key with which the specified value is to be associated
2653	>	* @param mappingFunction the function to compute a value
2654	>	* @return the current (existing or computed) value associated with
2655	>	*         the specified key, or null if the computed value is null
2656	>	* @throws NullPointerException if the specified key or mappingFunction
2657	>	*         is null
2658	>	* @throws IllegalStateException if the computation detectably
2659	>	*         attempts a recursive update to this map that would
2660	>	*         otherwise never complete
2661	>	* @throws RuntimeException or Error if the mappingFunction does so,
2662	>	*         in which case the mapping is left unestablished
2663	>	*/
2664	>	public V computeIfAbsent
2665	>	(K key, Function<? super K, ? extends V> mappingFunction) {
2666	>	return internalComputeIfAbsent(key, mappingFunction);
2667	>	}
2668	>
2669	>	/**
2670	>	* If the value for the specified key is present and non-null,
2671	>	* attempts to compute a new mapping given the key and its current
2672	>	* mapped value.  The entire method invocation is performed
2673	>	* atomically.  Some attempted update operations on this map by
2674	>	* other threads may be blocked while computation is in progress,
2675	>	* so the computation should be short and simple, and must not
2676	>	* attempt to update any other mappings of this Map.
2677	>	*
2678	>	* @param key key with which the specified value is to be associated
2679	>	* @param remappingFunction the function to compute a value
2680	>	* @return the new value associated with the specified key, or null if none
2681	>	* @throws NullPointerException if the specified key or remappingFunction
2682	>	*         is null
2683	>	* @throws IllegalStateException if the computation detectably
2684	>	*         attempts a recursive update to this map that would
2685	>	*         otherwise never complete
2686	>	* @throws RuntimeException or Error if the remappingFunction does so,
2687	>	*         in which case the mapping is unchanged
2688	>	*/
2689	>	public V computeIfPresent
2690	>	(K key, BiFunction<? super K, ? super V, ? extends V> remappingFunction) {
2691	>	return internalCompute(key, true, remappingFunction);
2692	>	}
2693	>
2694	>	/**
2695	>	* Attempts to compute a mapping for the specified key and its
2696	>	* current mapped value (or {@code null} if there is no current
2697	>	* mapping). The entire method invocation is performed atomically.
2698	>	* Some attempted update operations on this map by other threads
2699	>	* may be blocked while computation is in progress, so the
2700	>	* computation should be short and simple, and must not attempt to
2701	>	* update any other mappings of this Map.
2702	>	*
2703	>	* @param key key with which the specified value is to be associated
2704	>	* @param remappingFunction the function to compute a value
2705	>	* @return the new value associated with the specified key, or null if none
2706	>	* @throws NullPointerException if the specified key or remappingFunction
2707	>	*         is null
2708	>	* @throws IllegalStateException if the computation detectably
2709	>	*         attempts a recursive update to this map that would
2710	>	*         otherwise never complete
2711	>	* @throws RuntimeException or Error if the remappingFunction does so,
2712	>	*         in which case the mapping is unchanged
2713	>	*/
2714	>	public V compute
2715	>	(K key, BiFunction<? super K, ? super V, ? extends V> remappingFunction) {
2716	>	return internalCompute(key, false, remappingFunction);
2717	>	}
2718	>
2719	>	/**
2720	>	* If the specified key is not already associated with a
2721	>	* (non-null) value, associates it with the given value.
2722	>	* Otherwise, replaces the value with the results of the given
2723	>	* remapping function, or removes if {@code null}. The entire
2724	>	* method invocation is performed atomically.  Some attempted
2725	>	* update operations on this map by other threads may be blocked
2726	>	* while computation is in progress, so the computation should be
2727	>	* short and simple, and must not attempt to update any other
2728	>	* mappings of this Map.
2729	>	*
2730	>	* @param key key with which the specified value is to be associated
2731	>	* @param value the value to use if absent
2732	>	* @param remappingFunction the function to recompute a value if present
2733	>	* @return the new value associated with the specified key, or null if none
2734	>	* @throws NullPointerException if the specified key or the
2735	>	*         remappingFunction is null
2736	>	* @throws RuntimeException or Error if the remappingFunction does so,
2737	>	*         in which case the mapping is unchanged
2738	>	*/
2739	>	public V merge
2740	>	(K key, V value,
2741	>	BiFunction<? super V, ? super V, ? extends V> remappingFunction) {
2742	>	return internalMerge(key, value, remappingFunction);
2743		}
2744
2745		/**
#	Line 910 | Line 2747 | public class ConcurrentHashMap<K, V> ext
2747		* This method does nothing if the key is not in the map.
2748		*
2749		* @param  key the key that needs to be removed
2750	<	* @return the previous value associated with <tt>key</tt>, or
2751	<	*         <tt>null</tt> if there was no mapping for <tt>key</tt>.
2750	>	* @return the previous value associated with {@code key}, or
2751	>	*         {@code null} if there was no mapping for {@code key}
2752		* @throws NullPointerException if the specified key is null
2753		*/
2754		public V remove(Object key) {
2755	<	int hash = hash(key);
919	<	return segmentFor(hash).remove(key, hash, null);
2755	>	return internalReplace(key, null, null);
2756		}
2757
2758		/**
#	Line 925 | Line 2761 | public class ConcurrentHashMap<K, V> ext
2761		* @throws NullPointerException if the specified key is null
2762		*/
2763		public boolean remove(Object key, Object value) {
2764	<	if (value == null)
929	<	return false;
930	<	int hash = hash(key);
931	<	return segmentFor(hash).remove(key, hash, value) != null;
2764	>	return value != null && internalReplace(key, null, value) != null;
2765		}
2766
2767		/**
#	Line 937 | Line 2770 | public class ConcurrentHashMap<K, V> ext
2770		* @throws NullPointerException if any of the arguments are null
2771		*/
2772		public boolean replace(K key, V oldValue, V newValue) {
2773	<	if (oldValue == null || newValue == null)
2773	>	if (key == null || oldValue == null || newValue == null)
2774		throw new NullPointerException();
2775	<	int hash = hash(key);
943	<	return segmentFor(hash).replace(key, hash, oldValue, newValue);
2775	>	return internalReplace(key, newValue, oldValue) != null;
2776		}
2777
2778		/**
2779		* {@inheritDoc}
2780		*
2781		* @return the previous value associated with the specified key,
2782	<	*         or <tt>null</tt> if there was no mapping for the key
2782	>	*         or {@code null} if there was no mapping for the key
2783		* @throws NullPointerException if the specified key or value is null
2784		*/
2785		public V replace(K key, V value) {
2786	<	if (value == null)
2786	>	if (key == null || value == null)
2787		throw new NullPointerException();
2788	<	int hash = hash(key);
957	<	return segmentFor(hash).replace(key, hash, value);
2788	>	return internalReplace(key, value, null);
2789		}
2790
2791		/**
2792		* Removes all of the mappings from this map.
2793		*/
2794		public void clear() {
2795	<	for (int i = 0; i < segments.length; ++i)
965	<	segments[i].clear();
2795	>	internalClear();
2796		}
2797
2798		/**
2799		* Returns a {@link Set} view of the keys contained in this map.
2800		* The set is backed by the map, so changes to the map are
2801	<	* reflected in the set, and vice-versa.  The set supports element
972	<	* removal, which removes the corresponding mapping from this map,
973	<	* via the <tt>Iterator.remove</tt>, <tt>Set.remove</tt>,
974	<	* <tt>removeAll</tt>, <tt>retainAll</tt>, and <tt>clear</tt>
975	<	* operations.  It does not support the <tt>add</tt> or
976	<	* <tt>addAll</tt> operations.
2801	>	* reflected in the set, and vice-versa.
2802		*
2803	<	* <p>The view's <tt>iterator</tt> is a "weakly consistent" iterator
979	<	* that will never throw {@link ConcurrentModificationException},
980	<	* and guarantees to traverse elements as they existed upon
981	<	* construction of the iterator, and may (but is not guaranteed to)
982	<	* reflect any modifications subsequent to construction.
2803	>	* @return the set view
2804		*/
2805	<	public Set<K> keySet() {
2806	<	Set<K> ks = keySet;
2807	<	return (ks != null) ? ks : (keySet = new KeySet());
2805	>	public KeySetView<K,V> keySet() {
2806	>	KeySetView<K,V> ks = keySet;
2807	>	return (ks != null) ? ks : (keySet = new KeySetView<K,V>(this, null));
2808	>	}
2809	>
2810	>	/**
2811	>	* Returns a {@link Set} view of the keys in this map, using the
2812	>	* given common mapped value for any additions (i.e., {@link
2813	>	* Collection#add} and {@link Collection#addAll}). This is of
2814	>	* course only appropriate if it is acceptable to use the same
2815	>	* value for all additions from this view.
2816	>	*
2817	>	* @param mappedValue the mapped value to use for any
2818	>	* additions.
2819	>	* @return the set view
2820	>	* @throws NullPointerException if the mappedValue is null
2821	>	*/
2822	>	public KeySetView<K,V> keySet(V mappedValue) {
2823	>	if (mappedValue == null)
2824	>	throw new NullPointerException();
2825	>	return new KeySetView<K,V>(this, mappedValue);
2826		}
2827
2828		/**
2829		* Returns a {@link Collection} view of the values contained in this map.
2830		* The collection is backed by the map, so changes to the map are
2831	<	* reflected in the collection, and vice-versa.  The collection
993	<	* supports element removal, which removes the corresponding
994	<	* mapping from this map, via the <tt>Iterator.remove</tt>,
995	<	* <tt>Collection.remove</tt>, <tt>removeAll</tt>,
996	<	* <tt>retainAll</tt>, and <tt>clear</tt> operations.  It does not
997	<	* support the <tt>add</tt> or <tt>addAll</tt> operations.
998	<	*
999	<	* <p>The view's <tt>iterator</tt> is a "weakly consistent" iterator
1000	<	* that will never throw {@link ConcurrentModificationException},
1001	<	* and guarantees to traverse elements as they existed upon
1002	<	* construction of the iterator, and may (but is not guaranteed to)
1003	<	* reflect any modifications subsequent to construction.
2831	>	* reflected in the collection, and vice-versa.
2832		*/
2833	<	public Collection<V> values() {
2834	<	Collection<V> vs = values;
2835	<	return (vs != null) ? vs : (values = new Values());
2833	>	public ValuesView<K,V> values() {
2834	>	ValuesView<K,V> vs = values;
2835	>	return (vs != null) ? vs : (values = new ValuesView<K,V>(this));
2836		}
2837
2838		/**
#	Line 1012 | Line 2840 | public class ConcurrentHashMap<K, V> ext
2840		* The set is backed by the map, so changes to the map are
2841		* reflected in the set, and vice-versa.  The set supports element
2842		* removal, which removes the corresponding mapping from the map,
2843	<	* via the <tt>Iterator.remove</tt>, <tt>Set.remove</tt>,
2844	<	* <tt>removeAll</tt>, <tt>retainAll</tt>, and <tt>clear</tt>
2845	<	* operations.  It does not support the <tt>add</tt> or
2846	<	* <tt>addAll</tt> operations.
2843	>	* via the {@code Iterator.remove}, {@code Set.remove},
2844	>	* {@code removeAll}, {@code retainAll}, and {@code clear}
2845	>	* operations.  It does not support the {@code add} or
2846	>	* {@code addAll} operations.
2847		*
2848	<	* <p>The view's <tt>iterator</tt> is a "weakly consistent" iterator
2848	>	* <p>The view's {@code iterator} is a "weakly consistent" iterator
2849		* that will never throw {@link ConcurrentModificationException},
2850		* and guarantees to traverse elements as they existed upon
2851		* construction of the iterator, and may (but is not guaranteed to)
2852		* reflect any modifications subsequent to construction.
2853		*/
2854		public Set<Map.Entry<K,V>> entrySet() {
2855	<	Set<Map.Entry<K,V>> es = entrySet;
2856	<	return (es != null) ? es : (entrySet = new EntrySet());
2855	>	EntrySetView<K,V> es = entrySet;
2856	>	return (es != null) ? es : (entrySet = new EntrySetView<K,V>(this));
2857		}
2858
2859		/**
2860		* Returns an enumeration of the keys in this table.
2861		*
2862		* @return an enumeration of the keys in this table
2863	<	* @see #keySet
2863	>	* @see #keySet()
2864		*/
2865		public Enumeration<K> keys() {
2866	<	return new KeyIterator();
2866	>	return new KeyIterator<K,V>(this);
2867		}
2868
2869		/**
2870		* Returns an enumeration of the values in this table.
2871		*
2872		* @return an enumeration of the values in this table
2873	<	* @see #values
2873	>	* @see #values()
2874		*/
2875		public Enumeration<V> elements() {
2876	<	return new ValueIterator();
2876	>	return new ValueIterator<K,V>(this);
2877		}
2878
2879	<	/* ---------------- Iterator Support -------------- */
2880	<
2881	<	abstract class HashIterator {
2882	<	int nextSegmentIndex;
2883	<	int nextTableIndex;
2884	<	HashEntry<K,V>[] currentTable;
2885	<	HashEntry<K, V> nextEntry;
2886	<	HashEntry<K, V> lastReturned;
2887	<
2888	<	HashIterator() {
2889	<	nextSegmentIndex = segments.length - 1;
2890	<	nextTableIndex = -1;
2891	<	advance();
2879	>	/**
2880	>	* Returns the hash code value for this {@link Map}, i.e.,
2881	>	* the sum of, for each key-value pair in the map,
2882	>	* {@code key.hashCode() ^ value.hashCode()}.
2883	>	*
2884	>	* @return the hash code value for this map
2885	>	*/
2886	>	public int hashCode() {
2887	>	int h = 0;
2888	>	Traverser<K,V,Object> it = new Traverser<K,V,Object>(this);
2889	>	V v;
2890	>	while ((v = it.advance()) != null) {
2891	>	h += it.nextKey.hashCode() ^ v.hashCode();
2892		}
2893	+	return h;
2894	+	}
2895
2896	<	public boolean hasMoreElements() { return hasNext(); }
2897	<
2898	<	final void advance() {
2899	<	if (nextEntry != null && (nextEntry = nextEntry.next) != null)
2900	<	return;
2901	<
2902	<	while (nextTableIndex >= 0) {
2903	<	if ( (nextEntry = currentTable[nextTableIndex--]) != null)
2904	<	return;
2896	>	/**
2897	>	* Returns a string representation of this map.  The string
2898	>	* representation consists of a list of key-value mappings (in no
2899	>	* particular order) enclosed in braces ("{@code {}}").  Adjacent
2900	>	* mappings are separated by the characters {@code ", "} (comma
2901	>	* and space).  Each key-value mapping is rendered as the key
2902	>	* followed by an equals sign ("{@code =}") followed by the
2903	>	* associated value.
2904	>	*
2905	>	* @return a string representation of this map
2906	>	*/
2907	>	public String toString() {
2908	>	Traverser<K,V,Object> it = new Traverser<K,V,Object>(this);
2909	>	StringBuilder sb = new StringBuilder();
2910	>	sb.append('{');
2911	>	V v;
2912	>	if ((v = it.advance()) != null) {
2913	>	for (;;) {
2914	>	Object k = it.nextKey;
2915	>	sb.append(k == this ? "(this Map)" : k);
2916	>	sb.append('=');
2917	>	sb.append(v == this ? "(this Map)" : v);
2918	>	if ((v = it.advance()) == null)
2919	>	break;
2920	>	sb.append(',').append(' ');
2921		}
2922	+	}
2923	+	return sb.append('}').toString();
2924	+	}
2925
2926	<	while (nextSegmentIndex >= 0) {
2927	<	Segment<K,V> seg = segments[nextSegmentIndex--];
2928	<	if (seg.count != 0) {
2929	<	currentTable = seg.table;
2930	<	for (int j = currentTable.length - 1; j >= 0; --j) {
2931	<	if ( (nextEntry = currentTable[j]) != null) {
2932	<	nextTableIndex = j - 1;
2933	<	return;
2934	<	}
2935	<	}
2936	<	}
2926	>	/**
2927	>	* Compares the specified object with this map for equality.
2928	>	* Returns {@code true} if the given object is a map with the same
2929	>	* mappings as this map.  This operation may return misleading
2930	>	* results if either map is concurrently modified during execution
2931	>	* of this method.
2932	>	*
2933	>	* @param o object to be compared for equality with this map
2934	>	* @return {@code true} if the specified object is equal to this map
2935	>	*/
2936	>	public boolean equals(Object o) {
2937	>	if (o != this) {
2938	>	if (!(o instanceof Map))
2939	>	return false;
2940	>	Map<?,?> m = (Map<?,?>) o;
2941	>	Traverser<K,V,Object> it = new Traverser<K,V,Object>(this);
2942	>	V val;
2943	>	while ((val = it.advance()) != null) {
2944	>	Object v = m.get(it.nextKey);
2945	>	if (v == null || (v != val && !v.equals(val)))
2946	>	return false;
2947	>	}
2948	>	for (Map.Entry<?,?> e : m.entrySet()) {
2949	>	Object mk, mv, v;
2950	>	if ((mk = e.getKey()) == null ||
2951	>	(mv = e.getValue()) == null ||
2952	>	(v = internalGet(mk)) == null ||
2953	>	(mv != v && !mv.equals(v)))
2954	>	return false;
2955		}
2956		}
2957	+	return true;
2958	+	}
2959
2960	<	public boolean hasNext() { return nextEntry != null; }
2960	>	/* ----------------Iterators -------------- */
2961
2962	<	HashEntry<K,V> nextEntry() {
2963	<	if (nextEntry == null)
2962	>	@SuppressWarnings("serial") static final class KeyIterator<K,V>
2963	>	extends Traverser<K,V,Object>
2964	>	implements Spliterator<K>, Iterator<K>, Enumeration<K> {
2965	>	KeyIterator(ConcurrentHashMap<K, V> map) { super(map); }
2966	>	KeyIterator(ConcurrentHashMap<K, V> map, Traverser<K,V,Object> it) {
2967	>	super(map, it);
2968	>	}
2969	>	public KeyIterator<K,V> trySplit() {
2970	>	if (tab != null && baseIndex == baseLimit)
2971	>	return null;
2972	>	return new KeyIterator<K,V>(map, this);
2973	>	}
2974	>	@SuppressWarnings("unchecked") public final K next() {
2975	>	if (nextVal == null && advance() == null)
2976		throw new NoSuchElementException();
2977	<	lastReturned = nextEntry;
2978	<	advance();
2979	<	return lastReturned;
2977	>	Object k = nextKey;
2978	>	nextVal = null;
2979	>	return (K) k;
2980		}
2981
2982	<	public void remove() {
2983	<	if (lastReturned == null)
2984	<	throw new IllegalStateException();
2985	<	ConcurrentHashMap.this.remove(lastReturned.key);
2986	<	lastReturned = null;
2982	>	public final K nextElement() { return next(); }
2983	>
2984	>	public Iterator<K> iterator() { return this; }
2985	>
2986	>	public void forEach(Block<? super K> action) {
2987	>	if (action == null) throw new NullPointerException();
2988	>	while (advance() != null)
2989	>	action.accept((K)nextKey);
2990		}
2991		}
2992
2993	<	final class KeyIterator
2994	<	extends HashIterator
2995	<	implements Iterator<K>, Enumeration<K>
2996	<	{
2997	<	public K next()        { return super.nextEntry().key; }
2998	<	public K nextElement() { return super.nextEntry().key; }
2993	>	@SuppressWarnings("serial") static final class ValueIterator<K,V>
2994	>	extends Traverser<K,V,Object>
2995	>	implements Spliterator<V>, Iterator<V>, Enumeration<V> {
2996	>	ValueIterator(ConcurrentHashMap<K, V> map) { super(map); }
2997	>	ValueIterator(ConcurrentHashMap<K, V> map, Traverser<K,V,Object> it) {
2998	>	super(map, it);
2999	>	}
3000	>	public ValueIterator<K,V> trySplit() {
3001	>	if (tab != null && baseIndex == baseLimit)
3002	>	return null;
3003	>	return new ValueIterator<K,V>(map, this);
3004	>	}
3005	>
3006	>	public final V next() {
3007	>	V v;
3008	>	if ((v = nextVal) == null && (v = advance()) == null)
3009	>	throw new NoSuchElementException();
3010	>	nextVal = null;
3011	>	return v;
3012	>	}
3013	>
3014	>	public final V nextElement() { return next(); }
3015	>
3016	>	public Iterator<V> iterator() { return this; }
3017	>
3018	>	public void forEach(Block<? super V> action) {
3019	>	if (action == null) throw new NullPointerException();
3020	>	V v;
3021	>	while ((v = advance()) != null)
3022	>	action.accept(v);
3023	>	}
3024		}
3025
3026	<	final class ValueIterator
3027	<	extends HashIterator
3028	<	implements Iterator<V>, Enumeration<V>
3029	<	{
3030	<	public V next()        { return super.nextEntry().value; }
3031	<	public V nextElement() { return super.nextEntry().value; }
3026	>	@SuppressWarnings("serial") static final class EntryIterator<K,V>
3027	>	extends Traverser<K,V,Object>
3028	>	implements Spliterator<Map.Entry<K,V>>, Iterator<Map.Entry<K,V>> {
3029	>	EntryIterator(ConcurrentHashMap<K, V> map) { super(map); }
3030	>	EntryIterator(ConcurrentHashMap<K, V> map, Traverser<K,V,Object> it) {
3031	>	super(map, it);
3032	>	}
3033	>	public EntryIterator<K,V> trySplit() {
3034	>	if (tab != null && baseIndex == baseLimit)
3035	>	return null;
3036	>	return new EntryIterator<K,V>(map, this);
3037	>	}
3038	>
3039	>	@SuppressWarnings("unchecked") public final Map.Entry<K,V> next() {
3040	>	V v;
3041	>	if ((v = nextVal) == null && (v = advance()) == null)
3042	>	throw new NoSuchElementException();
3043	>	Object k = nextKey;
3044	>	nextVal = null;
3045	>	return new MapEntry<K,V>((K)k, v, map);
3046	>	}
3047	>
3048	>	public Iterator<Map.Entry<K,V>> iterator() { return this; }
3049	>
3050	>	public void forEach(Block<? super Map.Entry<K,V>> action) {
3051	>	if (action == null) throw new NullPointerException();
3052	>	V v;
3053	>	while ((v = advance()) != null)
3054	>	action.accept(entryFor((K)nextKey, v));
3055	>	}
3056		}
3057
3058		/**
3059	<	* Custom Entry class used by EntryIterator.next(), that relays
1127	<	* setValue changes to the underlying map.
3059	>	* Exported Entry for iterators
3060		*/
3061	<	final class WriteThroughEntry
3062	<	extends AbstractMap.SimpleEntry<K,V>
3063	<	{
3064	<	WriteThroughEntry(K k, V v) {
3065	<	super(k,v);
3061	>	static final class MapEntry<K,V> implements Map.Entry<K, V> {
3062	>	final K key; // non-null
3063	>	V val;       // non-null
3064	>	final ConcurrentHashMap<K, V> map;
3065	>	MapEntry(K key, V val, ConcurrentHashMap<K, V> map) {
3066	>	this.key = key;
3067	>	this.val = val;
3068	>	this.map = map;
3069	>	}
3070	>	public final K getKey()       { return key; }
3071	>	public final V getValue()     { return val; }
3072	>	public final int hashCode()   { return key.hashCode() ^ val.hashCode(); }
3073	>	public final String toString(){ return key + "=" + val; }
3074	>
3075	>	public final boolean equals(Object o) {
3076	>	Object k, v; Map.Entry<?,?> e;
3077	>	return ((o instanceof Map.Entry) &&
3078	>	(k = (e = (Map.Entry<?,?>)o).getKey()) != null &&
3079	>	(v = e.getValue()) != null &&
3080	>	(k == key || k.equals(key)) &&
3081	>	(v == val || v.equals(val)));
3082		}
3083
3084		/**
3085	<	* Set our entry's value and write through to the map. The
3086	<	* value to return is somewhat arbitrary here. Since a
3087	<	* WriteThroughEntry does not necessarily track asynchronous
3088	<	* changes, the most recent "previous" value could be
3089	<	* different from what we return (or could even have been
3090	<	* removed in which case the put will re-establish). We do not
1143	<	* and cannot guarantee more.
3085	>	* Sets our entry's value and writes through to the map. The
3086	>	* value to return is somewhat arbitrary here. Since we do not
3087	>	* necessarily track asynchronous changes, the most recent
3088	>	* "previous" value could be different from what we return (or
3089	>	* could even have been removed in which case the put will
3090	>	* re-establish). We do not and cannot guarantee more.
3091		*/
3092	<	public V setValue(V value) {
3092	>	public final V setValue(V value) {
3093		if (value == null) throw new NullPointerException();
3094	<	V v = super.setValue(value);
3095	<	ConcurrentHashMap.this.put(getKey(), value);
3094	>	V v = val;
3095	>	val = value;
3096	>	map.put(key, value);
3097		return v;
3098		}
3099		}
3100
3101	<	final class EntryIterator
3102	<	extends HashIterator
3103	<	implements Iterator<Entry<K,V>>
3104	<	{
3105	<	public Map.Entry<K,V> next() {
3106	<	HashEntry<K,V> e = super.nextEntry();
3107	<	return new WriteThroughEntry(e.key, e.value);
3101	>	/**
3102	>	* Returns exportable snapshot entry for the given key and value
3103	>	* when write-through can't or shouldn't be used.
3104	>	*/
3105	>	static <K,V> AbstractMap.SimpleEntry<K,V> entryFor(K k, V v) {
3106	>	return new AbstractMap.SimpleEntry<K,V>(k, v);
3107	>	}
3108	>
3109	>	/* ---------------- Serialization Support -------------- */
3110	>
3111	>	/**
3112	>	* Stripped-down version of helper class used in previous version,
3113	>	* declared for the sake of serialization compatibility
3114	>	*/
3115	>	static class Segment<K,V> implements Serializable {
3116	>	private static final long serialVersionUID = 2249069246763182397L;
3117	>	final float loadFactor;
3118	>	Segment(float lf) { this.loadFactor = lf; }
3119	>	}
3120	>
3121	>	/**
3122	>	* Saves the state of the {@code ConcurrentHashMap} instance to a
3123	>	* stream (i.e., serializes it).
3124	>	* @param s the stream
3125	>	* @serialData
3126	>	* the key (Object) and value (Object)
3127	>	* for each key-value mapping, followed by a null pair.
3128	>	* The key-value mappings are emitted in no particular order.
3129	>	*/
3130	>	@SuppressWarnings("unchecked") private void writeObject
3131	>	(java.io.ObjectOutputStream s)
3132	>	throws java.io.IOException {
3133	>	if (segments == null) { // for serialization compatibility
3134	>	segments = (Segment<K,V>[])
3135	>	new Segment<?,?>[DEFAULT_CONCURRENCY_LEVEL];
3136	>	for (int i = 0; i < segments.length; ++i)
3137	>	segments[i] = new Segment<K,V>(LOAD_FACTOR);
3138	>	}
3139	>	s.defaultWriteObject();
3140	>	Traverser<K,V,Object> it = new Traverser<K,V,Object>(this);
3141	>	V v;
3142	>	while ((v = it.advance()) != null) {
3143	>	s.writeObject(it.nextKey);
3144	>	s.writeObject(v);
3145		}
3146	+	s.writeObject(null);
3147	+	s.writeObject(null);
3148	+	segments = null; // throw away
3149		}
3150
3151	<	final class KeySet extends AbstractSet<K> {
3152	<	public Iterator<K> iterator() {
3153	<	return new KeyIterator();
3151	>	/**
3152	>	* Reconstitutes the instance from a stream (that is, deserializes it).
3153	>	* @param s the stream
3154	>	*/
3155	>	@SuppressWarnings("unchecked") private void readObject
3156	>	(java.io.ObjectInputStream s)
3157	>	throws java.io.IOException, ClassNotFoundException {
3158	>	s.defaultReadObject();
3159	>	this.segments = null; // unneeded
3160	>
3161	>	// Create all nodes, then place in table once size is known
3162	>	long size = 0L;
3163	>	Node<V> p = null;
3164	>	for (;;) {
3165	>	K k = (K) s.readObject();
3166	>	V v = (V) s.readObject();
3167	>	if (k != null && v != null) {
3168	>	int h = spread(k.hashCode());
3169	>	p = new Node<V>(h, k, v, p);
3170	>	++size;
3171	>	}
3172	>	else
3173	>	break;
3174		}
3175	<	public int size() {
3176	<	return ConcurrentHashMap.this.size();
3175	>	if (p != null) {
3176	>	boolean init = false;
3177	>	int n;
3178	>	if (size >= (long)(MAXIMUM_CAPACITY >>> 1))
3179	>	n = MAXIMUM_CAPACITY;
3180	>	else {
3181	>	int sz = (int)size;
3182	>	n = tableSizeFor(sz + (sz >>> 1) + 1);
3183	>	}
3184	>	int sc = sizeCtl;
3185	>	boolean collide = false;
3186	>	if (n > sc &&
3187	>	U.compareAndSwapInt(this, SIZECTL, sc, -1)) {
3188	>	try {
3189	>	if (table == null) {
3190	>	init = true;
3191	>	@SuppressWarnings("rawtypes") Node[] rt = new Node[n];
3192	>	Node<V>[] tab = (Node<V>[])rt;
3193	>	int mask = n - 1;
3194	>	while (p != null) {
3195	>	int j = p.hash & mask;
3196	>	Node<V> next = p.next;
3197	>	Node<V> q = p.next = tabAt(tab, j);
3198	>	setTabAt(tab, j, p);
3199	>	if (!collide && q != null && q.hash == p.hash)
3200	>	collide = true;
3201	>	p = next;
3202	>	}
3203	>	table = tab;
3204	>	addCount(size, -1);
3205	>	sc = n - (n >>> 2);
3206	>	}
3207	>	} finally {
3208	>	sizeCtl = sc;
3209	>	}
3210	>	if (collide) { // rescan and convert to TreeBins
3211	>	Node<V>[] tab = table;
3212	>	for (int i = 0; i < tab.length; ++i) {
3213	>	int c = 0;
3214	>	for (Node<V> e = tabAt(tab, i); e != null; e = e.next) {
3215	>	if (++c > TREE_THRESHOLD &&
3216	>	(e.key instanceof Comparable)) {
3217	>	replaceWithTreeBin(tab, i, e.key);
3218	>	break;
3219	>	}
3220	>	}
3221	>	}
3222	>	}
3223	>	}
3224	>	if (!init) { // Can only happen if unsafely published.
3225	>	while (p != null) {
3226	>	internalPut((K)p.key, p.val, false);
3227	>	p = p.next;
3228	>	}
3229	>	}
3230		}
3231	<	public boolean contains(Object o) {
3232	<	return ConcurrentHashMap.this.containsKey(o);
3231	>	}
3232	>
3233	>	// -------------------------------------------------------
3234	>
3235	>	// Sequential bulk operations
3236	>
3237	>	/**
3238	>	* Performs the given action for each (key, value).
3239	>	*
3240	>	* @param action the action
3241	>	*/
3242	>	@SuppressWarnings("unchecked") public void forEachSequentially
3243	>	(BiBlock<? super K, ? super V> action) {
3244	>	if (action == null) throw new NullPointerException();
3245	>	Traverser<K,V,Object> it = new Traverser<K,V,Object>(this);
3246	>	V v;
3247	>	while ((v = it.advance()) != null)
3248	>	action.accept((K)it.nextKey, v);
3249	>	}
3250	>
3251	>	/**
3252	>	* Performs the given action for each non-null transformation
3253	>	* of each (key, value).
3254	>	*
3255	>	* @param transformer a function returning the transformation
3256	>	* for an element, or null of there is no transformation (in
3257	>	* which case the action is not applied).
3258	>	* @param action the action
3259	>	*/
3260	>	@SuppressWarnings("unchecked") public <U> void forEachSequentially
3261	>	(BiFunction<? super K, ? super V, ? extends U> transformer,
3262	>	Block<? super U> action) {
3263	>	if (transformer == null || action == null)
3264	>	throw new NullPointerException();
3265	>	Traverser<K,V,Object> it = new Traverser<K,V,Object>(this);
3266	>	V v; U u;
3267	>	while ((v = it.advance()) != null) {
3268	>	if ((u = transformer.apply((K)it.nextKey, v)) != null)
3269	>	action.accept(u);
3270		}
3271	<	public boolean remove(Object o) {
3272	<	return ConcurrentHashMap.this.remove(o) != null;
3271	>	}
3272	>
3273	>	/**
3274	>	* Returns a non-null result from applying the given search
3275	>	* function on each (key, value), or null if none.
3276	>	*
3277	>	* @param searchFunction a function returning a non-null
3278	>	* result on success, else null
3279	>	* @return a non-null result from applying the given search
3280	>	* function on each (key, value), or null if none
3281	>	*/
3282	>	@SuppressWarnings("unchecked") public <U> U searchSequentially
3283	>	(BiFunction<? super K, ? super V, ? extends U> searchFunction) {
3284	>	if (searchFunction == null) throw new NullPointerException();
3285	>	Traverser<K,V,Object> it = new Traverser<K,V,Object>(this);
3286	>	V v; U u;
3287	>	while ((v = it.advance()) != null) {
3288	>	if ((u = searchFunction.apply((K)it.nextKey, v)) != null)
3289	>	return u;
3290		}
3291	<	public void clear() {
3292	<	ConcurrentHashMap.this.clear();
3291	>	return null;
3292	>	}
3293	>
3294	>	/**
3295	>	* Returns the result of accumulating the given transformation
3296	>	* of all (key, value) pairs using the given reducer to
3297	>	* combine values, or null if none.
3298	>	*
3299	>	* @param transformer a function returning the transformation
3300	>	* for an element, or null of there is no transformation (in
3301	>	* which case it is not combined).
3302	>	* @param reducer a commutative associative combining function
3303	>	* @return the result of accumulating the given transformation
3304	>	* of all (key, value) pairs
3305	>	*/
3306	>	@SuppressWarnings("unchecked") public <U> U reduceSequentially
3307	>	(BiFunction<? super K, ? super V, ? extends U> transformer,
3308	>	BiFunction<? super U, ? super U, ? extends U> reducer) {
3309	>	if (transformer == null || reducer == null)
3310	>	throw new NullPointerException();
3311	>	Traverser<K,V,Object> it = new Traverser<K,V,Object>(this);
3312	>	U r = null, u; V v;
3313	>	while ((v = it.advance()) != null) {
3314	>	if ((u = transformer.apply((K)it.nextKey, v)) != null)
3315	>	r = (r == null) ? u : reducer.apply(r, u);
3316		}
3317	<	public Object[] toArray() {
3318	<	Collection<K> c = new ArrayList<K>(size());
3319	<	for (K k : this)
3320	<	c.add(k);
3321	<	return c.toArray();
3317	>	return r;
3318	>	}
3319	>
3320	>	/**
3321	>	* Returns the result of accumulating the given transformation
3322	>	* of all (key, value) pairs using the given reducer to
3323	>	* combine values, and the given basis as an identity value.
3324	>	*
3325	>	* @param transformer a function returning the transformation
3326	>	* for an element
3327	>	* @param basis the identity (initial default value) for the reduction
3328	>	* @param reducer a commutative associative combining function
3329	>	* @return the result of accumulating the given transformation
3330	>	* of all (key, value) pairs
3331	>	*/
3332	>	@SuppressWarnings("unchecked") public double reduceToDoubleSequentially
3333	>	(DoubleBiFunction<? super K, ? super V> transformer,
3334	>	double basis,
3335	>	DoubleBinaryOperator reducer) {
3336	>	if (transformer == null || reducer == null)
3337	>	throw new NullPointerException();
3338	>	Traverser<K,V,Object> it = new Traverser<K,V,Object>(this);
3339	>	double r = basis; V v;
3340	>	while ((v = it.advance()) != null)
3341	>	r = reducer.applyAsDouble(r, transformer.applyAsDouble((K)it.nextKey, v));
3342	>	return r;
3343	>	}
3344	>
3345	>	/**
3346	>	* Returns the result of accumulating the given transformation
3347	>	* of all (key, value) pairs using the given reducer to
3348	>	* combine values, and the given basis as an identity value.
3349	>	*
3350	>	* @param transformer a function returning the transformation
3351	>	* for an element
3352	>	* @param basis the identity (initial default value) for the reduction
3353	>	* @param reducer a commutative associative combining function
3354	>	* @return the result of accumulating the given transformation
3355	>	* of all (key, value) pairs
3356	>	*/
3357	>	@SuppressWarnings("unchecked") public long reduceToLongSequentially
3358	>	(LongBiFunction<? super K, ? super V> transformer,
3359	>	long basis,
3360	>	LongBinaryOperator reducer) {
3361	>	if (transformer == null || reducer == null)
3362	>	throw new NullPointerException();
3363	>	Traverser<K,V,Object> it = new Traverser<K,V,Object>(this);
3364	>	long r = basis; V v;
3365	>	while ((v = it.advance()) != null)
3366	>	r = reducer.applyAsLong(r, transformer.applyAsLong((K)it.nextKey, v));
3367	>	return r;
3368	>	}
3369	>
3370	>	/**
3371	>	* Returns the result of accumulating the given transformation
3372	>	* of all (key, value) pairs using the given reducer to
3373	>	* combine values, and the given basis as an identity value.
3374	>	*
3375	>	* @param transformer a function returning the transformation
3376	>	* for an element
3377	>	* @param basis the identity (initial default value) for the reduction
3378	>	* @param reducer a commutative associative combining function
3379	>	* @return the result of accumulating the given transformation
3380	>	* of all (key, value) pairs
3381	>	*/
3382	>	@SuppressWarnings("unchecked") public int reduceToIntSequentially
3383	>	(IntBiFunction<? super K, ? super V> transformer,
3384	>	int basis,
3385	>	IntBinaryOperator reducer) {
3386	>	if (transformer == null || reducer == null)
3387	>	throw new NullPointerException();
3388	>	Traverser<K,V,Object> it = new Traverser<K,V,Object>(this);
3389	>	int r = basis; V v;
3390	>	while ((v = it.advance()) != null)
3391	>	r = reducer.applyAsInt(r, transformer.applyAsInt((K)it.nextKey, v));
3392	>	return r;
3393	>	}
3394	>
3395	>	/**
3396	>	* Performs the given action for each key.
3397	>	*
3398	>	* @param action the action
3399	>	*/
3400	>	@SuppressWarnings("unchecked") public void forEachKeySequentially
3401	>	(Block<? super K> action) {
3402	>	if (action == null) throw new NullPointerException();
3403	>	Traverser<K,V,Object> it = new Traverser<K,V,Object>(this);
3404	>	while (it.advance() != null)
3405	>	action.accept((K)it.nextKey);
3406	>	}
3407	>
3408	>	/**
3409	>	* Performs the given action for each non-null transformation
3410	>	* of each key.
3411	>	*
3412	>	* @param transformer a function returning the transformation
3413	>	* for an element, or null of there is no transformation (in
3414	>	* which case the action is not applied).
3415	>	* @param action the action
3416	>	*/
3417	>	@SuppressWarnings("unchecked") public <U> void forEachKeySequentially
3418	>	(Function<? super K, ? extends U> transformer,
3419	>	Block<? super U> action) {
3420	>	if (transformer == null || action == null)
3421	>	throw new NullPointerException();
3422	>	Traverser<K,V,Object> it = new Traverser<K,V,Object>(this);
3423	>	U u;
3424	>	while (it.advance() != null) {
3425	>	if ((u = transformer.apply((K)it.nextKey)) != null)
3426	>	action.accept(u);
3427		}
3428	<	public <T> T[] toArray(T[] a) {
3429	<	Collection<K> c = new ArrayList<K>();
3430	<	for (K k : this)
3431	<	c.add(k);
3432	<	return c.toArray(a);
3428	>	ForkJoinTasks.forEachKey
3429	>	(this, transformer, action).invoke();
3430	>	}
3431	>
3432	>	/**
3433	>	* Returns a non-null result from applying the given search
3434	>	* function on each key, or null if none.
3435	>	*
3436	>	* @param searchFunction a function returning a non-null
3437	>	* result on success, else null
3438	>	* @return a non-null result from applying the given search
3439	>	* function on each key, or null if none
3440	>	*/
3441	>	@SuppressWarnings("unchecked") public <U> U searchKeysSequentially
3442	>	(Function<? super K, ? extends U> searchFunction) {
3443	>	Traverser<K,V,Object> it = new Traverser<K,V,Object>(this);
3444	>	U u;
3445	>	while (it.advance() != null) {
3446	>	if ((u = searchFunction.apply((K)it.nextKey)) != null)
3447	>	return u;
3448		}
3449	+	return null;
3450		}
3451
3452	<	final class Values extends AbstractCollection<V> {
3453	<	public Iterator<V> iterator() {
3454	<	return new ValueIterator();
3452	>	/**
3453	>	* Returns the result of accumulating all keys using the given
3454	>	* reducer to combine values, or null if none.
3455	>	*
3456	>	* @param reducer a commutative associative combining function
3457	>	* @return the result of accumulating all keys using the given
3458	>	* reducer to combine values, or null if none
3459	>	*/
3460	>	@SuppressWarnings("unchecked") public K reduceKeysSequentially
3461	>	(BiFunction<? super K, ? super K, ? extends K> reducer) {
3462	>	if (reducer == null) throw new NullPointerException();
3463	>	Traverser<K,V,Object> it = new Traverser<K,V,Object>(this);
3464	>	K r = null;
3465	>	while (it.advance() != null) {
3466	>	K u = (K)it.nextKey;
3467	>	r = (r == null) ? u : reducer.apply(r, u);
3468		}
3469	<	public int size() {
3470	<	return ConcurrentHashMap.this.size();
3469	>	return r;
3470	>	}
3471	>
3472	>	/**
3473	>	* Returns the result of accumulating the given transformation
3474	>	* of all keys using the given reducer to combine values, or
3475	>	* null if none.
3476	>	*
3477	>	* @param transformer a function returning the transformation
3478	>	* for an element, or null of there is no transformation (in
3479	>	* which case it is not combined).
3480	>	* @param reducer a commutative associative combining function
3481	>	* @return the result of accumulating the given transformation
3482	>	* of all keys
3483	>	*/
3484	>	@SuppressWarnings("unchecked") public <U> U reduceKeysSequentially
3485	>	(Function<? super K, ? extends U> transformer,
3486	>	BiFunction<? super U, ? super U, ? extends U> reducer) {
3487	>	if (transformer == null || reducer == null)
3488	>	throw new NullPointerException();
3489	>	Traverser<K,V,Object> it = new Traverser<K,V,Object>(this);
3490	>	U r = null, u;
3491	>	while (it.advance() != null) {
3492	>	if ((u = transformer.apply((K)it.nextKey)) != null)
3493	>	r = (r == null) ? u : reducer.apply(r, u);
3494		}
3495	<	public boolean contains(Object o) {
3496	<	return ConcurrentHashMap.this.containsValue(o);
3495	>	return r;
3496	>	}
3497	>
3498	>	/**
3499	>	* Returns the result of accumulating the given transformation
3500	>	* of all keys using the given reducer to combine values, and
3501	>	* the given basis as an identity value.
3502	>	*
3503	>	* @param transformer a function returning the transformation
3504	>	* for an element
3505	>	* @param basis the identity (initial default value) for the reduction
3506	>	* @param reducer a commutative associative combining function
3507	>	* @return  the result of accumulating the given transformation
3508	>	* of all keys
3509	>	*/
3510	>	@SuppressWarnings("unchecked") public double reduceKeysToDoubleSequentially
3511	>	(DoubleFunction<? super K> transformer,
3512	>	double basis,
3513	>	DoubleBinaryOperator reducer) {
3514	>	if (transformer == null || reducer == null)
3515	>	throw new NullPointerException();
3516	>	Traverser<K,V,Object> it = new Traverser<K,V,Object>(this);
3517	>	double r = basis;
3518	>	while (it.advance() != null)
3519	>	r = reducer.applyAsDouble(r, transformer.applyAsDouble((K)it.nextKey));
3520	>	return r;
3521	>	}
3522	>
3523	>	/**
3524	>	* Returns the result of accumulating the given transformation
3525	>	* of all keys using the given reducer to combine values, and
3526	>	* the given basis as an identity value.
3527	>	*
3528	>	* @param transformer a function returning the transformation
3529	>	* for an element
3530	>	* @param basis the identity (initial default value) for the reduction
3531	>	* @param reducer a commutative associative combining function
3532	>	* @return the result of accumulating the given transformation
3533	>	* of all keys
3534	>	*/
3535	>	@SuppressWarnings("unchecked") public long reduceKeysToLongSequentially
3536	>	(LongFunction<? super K> transformer,
3537	>	long basis,
3538	>	LongBinaryOperator reducer) {
3539	>	if (transformer == null || reducer == null)
3540	>	throw new NullPointerException();
3541	>	Traverser<K,V,Object> it = new Traverser<K,V,Object>(this);
3542	>	long r = basis;
3543	>	while (it.advance() != null)
3544	>	r = reducer.applyAsLong(r, transformer.applyAsLong((K)it.nextKey));
3545	>	return r;
3546	>	}
3547	>
3548	>	/**
3549	>	* Returns the result of accumulating the given transformation
3550	>	* of all keys using the given reducer to combine values, and
3551	>	* the given basis as an identity value.
3552	>	*
3553	>	* @param transformer a function returning the transformation
3554	>	* for an element
3555	>	* @param basis the identity (initial default value) for the reduction
3556	>	* @param reducer a commutative associative combining function
3557	>	* @return the result of accumulating the given transformation
3558	>	* of all keys
3559	>	*/
3560	>	@SuppressWarnings("unchecked") public int reduceKeysToIntSequentially
3561	>	(IntFunction<? super K> transformer,
3562	>	int basis,
3563	>	IntBinaryOperator reducer) {
3564	>	if (transformer == null || reducer == null)
3565	>	throw new NullPointerException();
3566	>	Traverser<K,V,Object> it = new Traverser<K,V,Object>(this);
3567	>	int r = basis;
3568	>	while (it.advance() != null)
3569	>	r = reducer.applyAsInt(r, transformer.applyAsInt((K)it.nextKey));
3570	>	return r;
3571	>	}
3572	>
3573	>	/**
3574	>	* Performs the given action for each value.
3575	>	*
3576	>	* @param action the action
3577	>	*/
3578	>	public void forEachValueSequentially(Block<? super V> action) {
3579	>	if (action == null) throw new NullPointerException();
3580	>	Traverser<K,V,Object> it = new Traverser<K,V,Object>(this);
3581	>	V v;
3582	>	while ((v = it.advance()) != null)
3583	>	action.accept(v);
3584	>	}
3585	>
3586	>	/**
3587	>	* Performs the given action for each non-null transformation
3588	>	* of each value.
3589	>	*
3590	>	* @param transformer a function returning the transformation
3591	>	* for an element, or null of there is no transformation (in
3592	>	* which case the action is not applied).
3593	>	*/
3594	>	public <U> void forEachValueSequentially
3595	>	(Function<? super V, ? extends U> transformer,
3596	>	Block<? super U> action) {
3597	>	if (transformer == null || action == null)
3598	>	throw new NullPointerException();
3599	>	Traverser<K,V,Object> it = new Traverser<K,V,Object>(this);
3600	>	V v; U u;
3601	>	while ((v = it.advance()) != null) {
3602	>	if ((u = transformer.apply(v)) != null)
3603	>	action.accept(u);
3604		}
3605	<	public void clear() {
3606	<	ConcurrentHashMap.this.clear();
3605	>	}
3606	>
3607	>	/**
3608	>	* Returns a non-null result from applying the given search
3609	>	* function on each value, or null if none.
3610	>	*
3611	>	* @param searchFunction a function returning a non-null
3612	>	* result on success, else null
3613	>	* @return a non-null result from applying the given search
3614	>	* function on each value, or null if none
3615	>	*
3616	>	*/
3617	>	public <U> U searchValuesSequentially
3618	>	(Function<? super V, ? extends U> searchFunction) {
3619	>	if (searchFunction == null) throw new NullPointerException();
3620	>	Traverser<K,V,Object> it = new Traverser<K,V,Object>(this);
3621	>	V v; U u;
3622	>	while ((v = it.advance()) != null) {
3623	>	if ((u = searchFunction.apply(v)) != null)
3624	>	return u;
3625		}
3626	<	public Object[] toArray() {
3627	<	Collection<V> c = new ArrayList<V>(size());
3628	<	for (V v : this)
3629	<	c.add(v);
3630	<	return c.toArray();
3626	>	return null;
3627	>	}
3628	>
3629	>	/**
3630	>	* Returns the result of accumulating all values using the
3631	>	* given reducer to combine values, or null if none.
3632	>	*
3633	>	* @param reducer a commutative associative combining function
3634	>	* @return  the result of accumulating all values
3635	>	*/
3636	>	public V reduceValuesSequentially
3637	>	(BiFunction<? super V, ? super V, ? extends V> reducer) {
3638	>	if (reducer == null) throw new NullPointerException();
3639	>	Traverser<K,V,Object> it = new Traverser<K,V,Object>(this);
3640	>	V r = null; V v;
3641	>	while ((v = it.advance()) != null)
3642	>	r = (r == null) ? v : reducer.apply(r, v);
3643	>	return r;
3644	>	}
3645	>
3646	>	/**
3647	>	* Returns the result of accumulating the given transformation
3648	>	* of all values using the given reducer to combine values, or
3649	>	* null if none.
3650	>	*
3651	>	* @param transformer a function returning the transformation
3652	>	* for an element, or null of there is no transformation (in
3653	>	* which case it is not combined).
3654	>	* @param reducer a commutative associative combining function
3655	>	* @return the result of accumulating the given transformation
3656	>	* of all values
3657	>	*/
3658	>	public <U> U reduceValuesSequentially
3659	>	(Function<? super V, ? extends U> transformer,
3660	>	BiFunction<? super U, ? super U, ? extends U> reducer) {
3661	>	if (transformer == null || reducer == null)
3662	>	throw new NullPointerException();
3663	>	Traverser<K,V,Object> it = new Traverser<K,V,Object>(this);
3664	>	U r = null, u; V v;
3665	>	while ((v = it.advance()) != null) {
3666	>	if ((u = transformer.apply(v)) != null)
3667	>	r = (r == null) ? u : reducer.apply(r, u);
3668		}
3669	<	public <T> T[] toArray(T[] a) {
3670	<	Collection<V> c = new ArrayList<V>(size());
3671	<	for (V v : this)
3672	<	c.add(v);
3673	<	return c.toArray(a);
3669	>	return r;
3670	>	}
3671	>
3672	>	/**
3673	>	* Returns the result of accumulating the given transformation
3674	>	* of all values using the given reducer to combine values,
3675	>	* and the given basis as an identity value.
3676	>	*
3677	>	* @param transformer a function returning the transformation
3678	>	* for an element
3679	>	* @param basis the identity (initial default value) for the reduction
3680	>	* @param reducer a commutative associative combining function
3681	>	* @return the result of accumulating the given transformation
3682	>	* of all values
3683	>	*/
3684	>	public double reduceValuesToDoubleSequentially
3685	>	(DoubleFunction<? super V> transformer,
3686	>	double basis,
3687	>	DoubleBinaryOperator reducer) {
3688	>	if (transformer == null || reducer == null)
3689	>	throw new NullPointerException();
3690	>	Traverser<K,V,Object> it = new Traverser<K,V,Object>(this);
3691	>	double r = basis; V v;
3692	>	while ((v = it.advance()) != null)
3693	>	r = reducer.applyAsDouble(r, transformer.applyAsDouble(v));
3694	>	return r;
3695	>	}
3696	>
3697	>	/**
3698	>	* Returns the result of accumulating the given transformation
3699	>	* of all values using the given reducer to combine values,
3700	>	* and the given basis as an identity value.
3701	>	*
3702	>	* @param transformer a function returning the transformation
3703	>	* for an element
3704	>	* @param basis the identity (initial default value) for the reduction
3705	>	* @param reducer a commutative associative combining function
3706	>	* @return the result of accumulating the given transformation
3707	>	* of all values
3708	>	*/
3709	>	public long reduceValuesToLongSequentially
3710	>	(LongFunction<? super V> transformer,
3711	>	long basis,
3712	>	LongBinaryOperator reducer) {
3713	>	if (transformer == null || reducer == null)
3714	>	throw new NullPointerException();
3715	>	Traverser<K,V,Object> it = new Traverser<K,V,Object>(this);
3716	>	long r = basis; V v;
3717	>	while ((v = it.advance()) != null)
3718	>	r = reducer.applyAsLong(r, transformer.applyAsLong(v));
3719	>	return r;
3720	>	}
3721	>
3722	>	/**
3723	>	* Returns the result of accumulating the given transformation
3724	>	* of all values using the given reducer to combine values,
3725	>	* and the given basis as an identity value.
3726	>	*
3727	>	* @param transformer a function returning the transformation
3728	>	* for an element
3729	>	* @param basis the identity (initial default value) for the reduction
3730	>	* @param reducer a commutative associative combining function
3731	>	* @return the result of accumulating the given transformation
3732	>	* of all values
3733	>	*/
3734	>	public int reduceValuesToIntSequentially
3735	>	(IntFunction<? super V> transformer,
3736	>	int basis,
3737	>	IntBinaryOperator reducer) {
3738	>	if (transformer == null || reducer == null)
3739	>	throw new NullPointerException();
3740	>	Traverser<K,V,Object> it = new Traverser<K,V,Object>(this);
3741	>	int r = basis; V v;
3742	>	while ((v = it.advance()) != null)
3743	>	r = reducer.applyAsInt(r, transformer.applyAsInt(v));
3744	>	return r;
3745	>	}
3746	>
3747	>	/**
3748	>	* Performs the given action for each entry.
3749	>	*
3750	>	* @param action the action
3751	>	*/
3752	>	@SuppressWarnings("unchecked") public void forEachEntrySequentially
3753	>	(Block<? super Map.Entry<K,V>> action) {
3754	>	if (action == null) throw new NullPointerException();
3755	>	Traverser<K,V,Object> it = new Traverser<K,V,Object>(this);
3756	>	V v;
3757	>	while ((v = it.advance()) != null)
3758	>	action.accept(entryFor((K)it.nextKey, v));
3759	>	}
3760	>
3761	>	/**
3762	>	* Performs the given action for each non-null transformation
3763	>	* of each entry.
3764	>	*
3765	>	* @param transformer a function returning the transformation
3766	>	* for an element, or null of there is no transformation (in
3767	>	* which case the action is not applied).
3768	>	* @param action the action
3769	>	*/
3770	>	@SuppressWarnings("unchecked") public <U> void forEachEntrySequentially
3771	>	(Function<Map.Entry<K,V>, ? extends U> transformer,
3772	>	Block<? super U> action) {
3773	>	if (transformer == null || action == null)
3774	>	throw new NullPointerException();
3775	>	Traverser<K,V,Object> it = new Traverser<K,V,Object>(this);
3776	>	V v; U u;
3777	>	while ((v = it.advance()) != null) {
3778	>	if ((u = transformer.apply(entryFor((K)it.nextKey, v))) != null)
3779	>	action.accept(u);
3780		}
3781		}
3782
3783	<	final class EntrySet extends AbstractSet<Map.Entry<K,V>> {
3784	<	public Iterator<Map.Entry<K,V>> iterator() {
3785	<	return new EntryIterator();
3783	>	/**
3784	>	* Returns a non-null result from applying the given search
3785	>	* function on each entry, or null if none.
3786	>	*
3787	>	* @param searchFunction a function returning a non-null
3788	>	* result on success, else null
3789	>	* @return a non-null result from applying the given search
3790	>	* function on each entry, or null if none
3791	>	*/
3792	>	@SuppressWarnings("unchecked") public <U> U searchEntriesSequentially
3793	>	(Function<Map.Entry<K,V>, ? extends U> searchFunction) {
3794	>	if (searchFunction == null) throw new NullPointerException();
3795	>	Traverser<K,V,Object> it = new Traverser<K,V,Object>(this);
3796	>	V v; U u;
3797	>	while ((v = it.advance()) != null) {
3798	>	if ((u = searchFunction.apply(entryFor((K)it.nextKey, v))) != null)
3799	>	return u;
3800		}
3801	<	public boolean contains(Object o) {
3802	<	if (!(o instanceof Map.Entry))
3803	<	return false;
3804	<	Map.Entry<?,?> e = (Map.Entry<?,?>)o;
3805	<	V v = ConcurrentHashMap.this.get(e.getKey());
3806	<	return v != null && v.equals(e.getValue());
3801	>	return null;
3802	>	}
3803	>
3804	>	/**
3805	>	* Returns the result of accumulating all entries using the
3806	>	* given reducer to combine values, or null if none.
3807	>	*
3808	>	* @param reducer a commutative associative combining function
3809	>	* @return the result of accumulating all entries
3810	>	*/
3811	>	@SuppressWarnings("unchecked") public Map.Entry<K,V> reduceEntriesSequentially
3812	>	(BiFunction<Map.Entry<K,V>, Map.Entry<K,V>, ? extends Map.Entry<K,V>> reducer) {
3813	>	if (reducer == null) throw new NullPointerException();
3814	>	Traverser<K,V,Object> it = new Traverser<K,V,Object>(this);
3815	>	Map.Entry<K,V> r = null; V v;
3816	>	while ((v = it.advance()) != null) {
3817	>	Map.Entry<K,V> u = entryFor((K)it.nextKey, v);
3818	>	r = (r == null) ? u : reducer.apply(r, u);
3819		}
3820	<	public boolean remove(Object o) {
3821	<	if (!(o instanceof Map.Entry))
3822	<	return false;
3823	<	Map.Entry<?,?> e = (Map.Entry<?,?>)o;
3824	<	return ConcurrentHashMap.this.remove(e.getKey(), e.getValue());
3820	>	return r;
3821	>	}
3822	>
3823	>	/**
3824	>	* Returns the result of accumulating the given transformation
3825	>	* of all entries using the given reducer to combine values,
3826	>	* or null if none.
3827	>	*
3828	>	* @param transformer a function returning the transformation
3829	>	* for an element, or null of there is no transformation (in
3830	>	* which case it is not combined).
3831	>	* @param reducer a commutative associative combining function
3832	>	* @return the result of accumulating the given transformation
3833	>	* of all entries
3834	>	*/
3835	>	@SuppressWarnings("unchecked") public <U> U reduceEntriesSequentially
3836	>	(Function<Map.Entry<K,V>, ? extends U> transformer,
3837	>	BiFunction<? super U, ? super U, ? extends U> reducer) {
3838	>	if (transformer == null || reducer == null)
3839	>	throw new NullPointerException();
3840	>	Traverser<K,V,Object> it = new Traverser<K,V,Object>(this);
3841	>	U r = null, u; V v;
3842	>	while ((v = it.advance()) != null) {
3843	>	if ((u = transformer.apply(entryFor((K)it.nextKey, v))) != null)
3844	>	r = (r == null) ? u : reducer.apply(r, u);
3845		}
3846	<	public int size() {
3847	<	return ConcurrentHashMap.this.size();
3846	>	return r;
3847	>	}
3848	>
3849	>	/**
3850	>	* Returns the result of accumulating the given transformation
3851	>	* of all entries using the given reducer to combine values,
3852	>	* and the given basis as an identity value.
3853	>	*
3854	>	* @param transformer a function returning the transformation
3855	>	* for an element
3856	>	* @param basis the identity (initial default value) for the reduction
3857	>	* @param reducer a commutative associative combining function
3858	>	* @return the result of accumulating the given transformation
3859	>	* of all entries
3860	>	*/
3861	>	@SuppressWarnings("unchecked") public double reduceEntriesToDoubleSequentially
3862	>	(DoubleFunction<Map.Entry<K,V>> transformer,
3863	>	double basis,
3864	>	DoubleBinaryOperator reducer) {
3865	>	if (transformer == null || reducer == null)
3866	>	throw new NullPointerException();
3867	>	Traverser<K,V,Object> it = new Traverser<K,V,Object>(this);
3868	>	double r = basis; V v;
3869	>	while ((v = it.advance()) != null)
3870	>	r = reducer.applyAsDouble(r, transformer.applyAsDouble(entryFor((K)it.nextKey, v)));
3871	>	return r;
3872	>	}
3873	>
3874	>	/**
3875	>	* Returns the result of accumulating the given transformation
3876	>	* of all entries using the given reducer to combine values,
3877	>	* and the given basis as an identity value.
3878	>	*
3879	>	* @param transformer a function returning the transformation
3880	>	* for an element
3881	>	* @param basis the identity (initial default value) for the reduction
3882	>	* @param reducer a commutative associative combining function
3883	>	* @return  the result of accumulating the given transformation
3884	>	* of all entries
3885	>	*/
3886	>	@SuppressWarnings("unchecked") public long reduceEntriesToLongSequentially
3887	>	(LongFunction<Map.Entry<K,V>> transformer,
3888	>	long basis,
3889	>	LongBinaryOperator reducer) {
3890	>	if (transformer == null || reducer == null)
3891	>	throw new NullPointerException();
3892	>	Traverser<K,V,Object> it = new Traverser<K,V,Object>(this);
3893	>	long r = basis; V v;
3894	>	while ((v = it.advance()) != null)
3895	>	r = reducer.applyAsLong(r, transformer.applyAsLong(entryFor((K)it.nextKey, v)));
3896	>	return r;
3897	>	}
3898	>
3899	>	/**
3900	>	* Returns the result of accumulating the given transformation
3901	>	* of all entries using the given reducer to combine values,
3902	>	* and the given basis as an identity value.
3903	>	*
3904	>	* @param transformer a function returning the transformation
3905	>	* for an element
3906	>	* @param basis the identity (initial default value) for the reduction
3907	>	* @param reducer a commutative associative combining function
3908	>	* @return the result of accumulating the given transformation
3909	>	* of all entries
3910	>	*/
3911	>	@SuppressWarnings("unchecked") public int reduceEntriesToIntSequentially
3912	>	(IntFunction<Map.Entry<K,V>> transformer,
3913	>	int basis,
3914	>	IntBinaryOperator reducer) {
3915	>	if (transformer == null || reducer == null)
3916	>	throw new NullPointerException();
3917	>	Traverser<K,V,Object> it = new Traverser<K,V,Object>(this);
3918	>	int r = basis; V v;
3919	>	while ((v = it.advance()) != null)
3920	>	r = reducer.applyAsInt(r, transformer.applyAsInt(entryFor((K)it.nextKey, v)));
3921	>	return r;
3922	>	}
3923	>
3924	>	// Parallel bulk operations
3925	>
3926	>	/**
3927	>	* Performs the given action for each (key, value).
3928	>	*
3929	>	* @param action the action
3930	>	*/
3931	>	public void forEachInParallel(BiBlock<? super K,? super V> action) {
3932	>	ForkJoinTasks.forEach
3933	>	(this, action).invoke();
3934	>	}
3935	>
3936	>	/**
3937	>	* Performs the given action for each non-null transformation
3938	>	* of each (key, value).
3939	>	*
3940	>	* @param transformer a function returning the transformation
3941	>	* for an element, or null of there is no transformation (in
3942	>	* which case the action is not applied).
3943	>	* @param action the action
3944	>	*/
3945	>	public <U> void forEachInParallel
3946	>	(BiFunction<? super K, ? super V, ? extends U> transformer,
3947	>	Block<? super U> action) {
3948	>	ForkJoinTasks.forEach
3949	>	(this, transformer, action).invoke();
3950	>	}
3951	>
3952	>	/**
3953	>	* Returns a non-null result from applying the given search
3954	>	* function on each (key, value), or null if none.  Upon
3955	>	* success, further element processing is suppressed and the
3956	>	* results of any other parallel invocations of the search
3957	>	* function are ignored.
3958	>	*
3959	>	* @param searchFunction a function returning a non-null
3960	>	* result on success, else null
3961	>	* @return a non-null result from applying the given search
3962	>	* function on each (key, value), or null if none
3963	>	*/
3964	>	public <U> U searchInParallel
3965	>	(BiFunction<? super K, ? super V, ? extends U> searchFunction) {
3966	>	return ForkJoinTasks.search
3967	>	(this, searchFunction).invoke();
3968	>	}
3969	>
3970	>	/**
3971	>	* Returns the result of accumulating the given transformation
3972	>	* of all (key, value) pairs using the given reducer to
3973	>	* combine values, or null if none.
3974	>	*
3975	>	* @param transformer a function returning the transformation
3976	>	* for an element, or null of there is no transformation (in
3977	>	* which case it is not combined).
3978	>	* @param reducer a commutative associative combining function
3979	>	* @return the result of accumulating the given transformation
3980	>	* of all (key, value) pairs
3981	>	*/
3982	>	public <U> U reduceInParallel
3983	>	(BiFunction<? super K, ? super V, ? extends U> transformer,
3984	>	BiFunction<? super U, ? super U, ? extends U> reducer) {
3985	>	return ForkJoinTasks.reduce
3986	>	(this, transformer, reducer).invoke();
3987	>	}
3988	>
3989	>	/**
3990	>	* Returns the result of accumulating the given transformation
3991	>	* of all (key, value) pairs using the given reducer to
3992	>	* combine values, and the given basis as an identity value.
3993	>	*
3994	>	* @param transformer a function returning the transformation
3995	>	* for an element
3996	>	* @param basis the identity (initial default value) for the reduction
3997	>	* @param reducer a commutative associative combining function
3998	>	* @return the result of accumulating the given transformation
3999	>	* of all (key, value) pairs
4000	>	*/
4001	>	public double reduceToDoubleInParallel
4002	>	(DoubleBiFunction<? super K, ? super V> transformer,
4003	>	double basis,
4004	>	DoubleBinaryOperator reducer) {
4005	>	return ForkJoinTasks.reduceToDouble
4006	>	(this, transformer, basis, reducer).invoke();
4007	>	}
4008	>
4009	>	/**
4010	>	* Returns the result of accumulating the given transformation
4011	>	* of all (key, value) pairs using the given reducer to
4012	>	* combine values, and the given basis as an identity value.
4013	>	*
4014	>	* @param transformer a function returning the transformation
4015	>	* for an element
4016	>	* @param basis the identity (initial default value) for the reduction
4017	>	* @param reducer a commutative associative combining function
4018	>	* @return the result of accumulating the given transformation
4019	>	* of all (key, value) pairs
4020	>	*/
4021	>	public long reduceToLongInParallel
4022	>	(LongBiFunction<? super K, ? super V> transformer,
4023	>	long basis,
4024	>	LongBinaryOperator reducer) {
4025	>	return ForkJoinTasks.reduceToLong
4026	>	(this, transformer, basis, reducer).invoke();
4027	>	}
4028	>
4029	>	/**
4030	>	* Returns the result of accumulating the given transformation
4031	>	* of all (key, value) pairs using the given reducer to
4032	>	* combine values, and the given basis as an identity value.
4033	>	*
4034	>	* @param transformer a function returning the transformation
4035	>	* for an element
4036	>	* @param basis the identity (initial default value) for the reduction
4037	>	* @param reducer a commutative associative combining function
4038	>	* @return the result of accumulating the given transformation
4039	>	* of all (key, value) pairs
4040	>	*/
4041	>	public int reduceToIntInParallel
4042	>	(IntBiFunction<? super K, ? super V> transformer,
4043	>	int basis,
4044	>	IntBinaryOperator reducer) {
4045	>	return ForkJoinTasks.reduceToInt
4046	>	(this, transformer, basis, reducer).invoke();
4047	>	}
4048	>
4049	>	/**
4050	>	* Performs the given action for each key.
4051	>	*
4052	>	* @param action the action
4053	>	*/
4054	>	public void forEachKeyInParallel(Block<? super K> action) {
4055	>	ForkJoinTasks.forEachKey
4056	>	(this, action).invoke();
4057	>	}
4058	>
4059	>	/**
4060	>	* Performs the given action for each non-null transformation
4061	>	* of each key.
4062	>	*
4063	>	* @param transformer a function returning the transformation
4064	>	* for an element, or null of there is no transformation (in
4065	>	* which case the action is not applied).
4066	>	* @param action the action
4067	>	*/
4068	>	public <U> void forEachKeyInParallel
4069	>	(Function<? super K, ? extends U> transformer,
4070	>	Block<? super U> action) {
4071	>	ForkJoinTasks.forEachKey
4072	>	(this, transformer, action).invoke();
4073	>	}
4074	>
4075	>	/**
4076	>	* Returns a non-null result from applying the given search
4077	>	* function on each key, or null if none. Upon success,
4078	>	* further element processing is suppressed and the results of
4079	>	* any other parallel invocations of the search function are
4080	>	* ignored.
4081	>	*
4082	>	* @param searchFunction a function returning a non-null
4083	>	* result on success, else null
4084	>	* @return a non-null result from applying the given search
4085	>	* function on each key, or null if none
4086	>	*/
4087	>	public <U> U searchKeysInParallel
4088	>	(Function<? super K, ? extends U> searchFunction) {
4089	>	return ForkJoinTasks.searchKeys
4090	>	(this, searchFunction).invoke();
4091	>	}
4092	>
4093	>	/**
4094	>	* Returns the result of accumulating all keys using the given
4095	>	* reducer to combine values, or null if none.
4096	>	*
4097	>	* @param reducer a commutative associative combining function
4098	>	* @return the result of accumulating all keys using the given
4099	>	* reducer to combine values, or null if none
4100	>	*/
4101	>	public K reduceKeysInParallel
4102	>	(BiFunction<? super K, ? super K, ? extends K> reducer) {
4103	>	return ForkJoinTasks.reduceKeys
4104	>	(this, reducer).invoke();
4105	>	}
4106	>
4107	>	/**
4108	>	* Returns the result of accumulating the given transformation
4109	>	* of all keys using the given reducer to combine values, or
4110	>	* null if none.
4111	>	*
4112	>	* @param transformer a function returning the transformation
4113	>	* for an element, or null of there is no transformation (in
4114	>	* which case it is not combined).
4115	>	* @param reducer a commutative associative combining function
4116	>	* @return the result of accumulating the given transformation
4117	>	* of all keys
4118	>	*/
4119	>	public <U> U reduceKeysInParallel
4120	>	(Function<? super K, ? extends U> transformer,
4121	>	BiFunction<? super U, ? super U, ? extends U> reducer) {
4122	>	return ForkJoinTasks.reduceKeys
4123	>	(this, transformer, reducer).invoke();
4124	>	}
4125	>
4126	>	/**
4127	>	* Returns the result of accumulating the given transformation
4128	>	* of all keys using the given reducer to combine values, and
4129	>	* the given basis as an identity value.
4130	>	*
4131	>	* @param transformer a function returning the transformation
4132	>	* for an element
4133	>	* @param basis the identity (initial default value) for the reduction
4134	>	* @param reducer a commutative associative combining function
4135	>	* @return  the result of accumulating the given transformation
4136	>	* of all keys
4137	>	*/
4138	>	public double reduceKeysToDoubleInParallel
4139	>	(DoubleFunction<? super K> transformer,
4140	>	double basis,
4141	>	DoubleBinaryOperator reducer) {
4142	>	return ForkJoinTasks.reduceKeysToDouble
4143	>	(this, transformer, basis, reducer).invoke();
4144	>	}
4145	>
4146	>	/**
4147	>	* Returns the result of accumulating the given transformation
4148	>	* of all keys using the given reducer to combine values, and
4149	>	* the given basis as an identity value.
4150	>	*
4151	>	* @param transformer a function returning the transformation
4152	>	* for an element
4153	>	* @param basis the identity (initial default value) for the reduction
4154	>	* @param reducer a commutative associative combining function
4155	>	* @return the result of accumulating the given transformation
4156	>	* of all keys
4157	>	*/
4158	>	public long reduceKeysToLongInParallel
4159	>	(LongFunction<? super K> transformer,
4160	>	long basis,
4161	>	LongBinaryOperator reducer) {
4162	>	return ForkJoinTasks.reduceKeysToLong
4163	>	(this, transformer, basis, reducer).invoke();
4164	>	}
4165	>
4166	>	/**
4167	>	* Returns the result of accumulating the given transformation
4168	>	* of all keys using the given reducer to combine values, and
4169	>	* the given basis as an identity value.
4170	>	*
4171	>	* @param transformer a function returning the transformation
4172	>	* for an element
4173	>	* @param basis the identity (initial default value) for the reduction
4174	>	* @param reducer a commutative associative combining function
4175	>	* @return the result of accumulating the given transformation
4176	>	* of all keys
4177	>	*/
4178	>	public int reduceKeysToIntInParallel
4179	>	(IntFunction<? super K> transformer,
4180	>	int basis,
4181	>	IntBinaryOperator reducer) {
4182	>	return ForkJoinTasks.reduceKeysToInt
4183	>	(this, transformer, basis, reducer).invoke();
4184	>	}
4185	>
4186	>	/**
4187	>	* Performs the given action for each value.
4188	>	*
4189	>	* @param action the action
4190	>	*/
4191	>	public void forEachValueInParallel(Block<? super V> action) {
4192	>	ForkJoinTasks.forEachValue
4193	>	(this, action).invoke();
4194	>	}
4195	>
4196	>	/**
4197	>	* Performs the given action for each non-null transformation
4198	>	* of each value.
4199	>	*
4200	>	* @param transformer a function returning the transformation
4201	>	* for an element, or null of there is no transformation (in
4202	>	* which case the action is not applied).
4203	>	*/
4204	>	public <U> void forEachValueInParallel
4205	>	(Function<? super V, ? extends U> transformer,
4206	>	Block<? super U> action) {
4207	>	ForkJoinTasks.forEachValue
4208	>	(this, transformer, action).invoke();
4209	>	}
4210	>
4211	>	/**
4212	>	* Returns a non-null result from applying the given search
4213	>	* function on each value, or null if none.  Upon success,
4214	>	* further element processing is suppressed and the results of
4215	>	* any other parallel invocations of the search function are
4216	>	* ignored.
4217	>	*
4218	>	* @param searchFunction a function returning a non-null
4219	>	* result on success, else null
4220	>	* @return a non-null result from applying the given search
4221	>	* function on each value, or null if none
4222	>	*
4223	>	*/
4224	>	public <U> U searchValuesInParallel
4225	>	(Function<? super V, ? extends U> searchFunction) {
4226	>	return ForkJoinTasks.searchValues
4227	>	(this, searchFunction).invoke();
4228	>	}
4229	>
4230	>	/**
4231	>	* Returns the result of accumulating all values using the
4232	>	* given reducer to combine values, or null if none.
4233	>	*
4234	>	* @param reducer a commutative associative combining function
4235	>	* @return  the result of accumulating all values
4236	>	*/
4237	>	public V reduceValuesInParallel
4238	>	(BiFunction<? super V, ? super V, ? extends V> reducer) {
4239	>	return ForkJoinTasks.reduceValues
4240	>	(this, reducer).invoke();
4241	>	}
4242	>
4243	>	/**
4244	>	* Returns the result of accumulating the given transformation
4245	>	* of all values using the given reducer to combine values, or
4246	>	* null if none.
4247	>	*
4248	>	* @param transformer a function returning the transformation
4249	>	* for an element, or null of there is no transformation (in
4250	>	* which case it is not combined).
4251	>	* @param reducer a commutative associative combining function
4252	>	* @return the result of accumulating the given transformation
4253	>	* of all values
4254	>	*/
4255	>	public <U> U reduceValuesInParallel
4256	>	(Function<? super V, ? extends U> transformer,
4257	>	BiFunction<? super U, ? super U, ? extends U> reducer) {
4258	>	return ForkJoinTasks.reduceValues
4259	>	(this, transformer, reducer).invoke();
4260	>	}
4261	>
4262	>	/**
4263	>	* Returns the result of accumulating the given transformation
4264	>	* of all values using the given reducer to combine values,
4265	>	* and the given basis as an identity value.
4266	>	*
4267	>	* @param transformer a function returning the transformation
4268	>	* for an element
4269	>	* @param basis the identity (initial default value) for the reduction
4270	>	* @param reducer a commutative associative combining function
4271	>	* @return the result of accumulating the given transformation
4272	>	* of all values
4273	>	*/
4274	>	public double reduceValuesToDoubleInParallel
4275	>	(DoubleFunction<? super V> transformer,
4276	>	double basis,
4277	>	DoubleBinaryOperator reducer) {
4278	>	return ForkJoinTasks.reduceValuesToDouble
4279	>	(this, transformer, basis, reducer).invoke();
4280	>	}
4281	>
4282	>	/**
4283	>	* Returns the result of accumulating the given transformation
4284	>	* of all values using the given reducer to combine values,
4285	>	* and the given basis as an identity value.
4286	>	*
4287	>	* @param transformer a function returning the transformation
4288	>	* for an element
4289	>	* @param basis the identity (initial default value) for the reduction
4290	>	* @param reducer a commutative associative combining function
4291	>	* @return the result of accumulating the given transformation
4292	>	* of all values
4293	>	*/
4294	>	public long reduceValuesToLongInParallel
4295	>	(LongFunction<? super V> transformer,
4296	>	long basis,
4297	>	LongBinaryOperator reducer) {
4298	>	return ForkJoinTasks.reduceValuesToLong
4299	>	(this, transformer, basis, reducer).invoke();
4300	>	}
4301	>
4302	>	/**
4303	>	* Returns the result of accumulating the given transformation
4304	>	* of all values using the given reducer to combine values,
4305	>	* and the given basis as an identity value.
4306	>	*
4307	>	* @param transformer a function returning the transformation
4308	>	* for an element
4309	>	* @param basis the identity (initial default value) for the reduction
4310	>	* @param reducer a commutative associative combining function
4311	>	* @return the result of accumulating the given transformation
4312	>	* of all values
4313	>	*/
4314	>	public int reduceValuesToIntInParallel
4315	>	(IntFunction<? super V> transformer,
4316	>	int basis,
4317	>	IntBinaryOperator reducer) {
4318	>	return ForkJoinTasks.reduceValuesToInt
4319	>	(this, transformer, basis, reducer).invoke();
4320	>	}
4321	>
4322	>	/**
4323	>	* Performs the given action for each entry.
4324	>	*
4325	>	* @param action the action
4326	>	*/
4327	>	public void forEachEntryInParallel(Block<? super Map.Entry<K,V>> action) {
4328	>	ForkJoinTasks.forEachEntry
4329	>	(this, action).invoke();
4330	>	}
4331	>
4332	>	/**
4333	>	* Performs the given action for each non-null transformation
4334	>	* of each entry.
4335	>	*
4336	>	* @param transformer a function returning the transformation
4337	>	* for an element, or null of there is no transformation (in
4338	>	* which case the action is not applied).
4339	>	* @param action the action
4340	>	*/
4341	>	public <U> void forEachEntryInParallel
4342	>	(Function<Map.Entry<K,V>, ? extends U> transformer,
4343	>	Block<? super U> action) {
4344	>	ForkJoinTasks.forEachEntry
4345	>	(this, transformer, action).invoke();
4346	>	}
4347	>
4348	>	/**
4349	>	* Returns a non-null result from applying the given search
4350	>	* function on each entry, or null if none.  Upon success,
4351	>	* further element processing is suppressed and the results of
4352	>	* any other parallel invocations of the search function are
4353	>	* ignored.
4354	>	*
4355	>	* @param searchFunction a function returning a non-null
4356	>	* result on success, else null
4357	>	* @return a non-null result from applying the given search
4358	>	* function on each entry, or null if none
4359	>	*/
4360	>	public <U> U searchEntriesInParallel
4361	>	(Function<Map.Entry<K,V>, ? extends U> searchFunction) {
4362	>	return ForkJoinTasks.searchEntries
4363	>	(this, searchFunction).invoke();
4364	>	}
4365	>
4366	>	/**
4367	>	* Returns the result of accumulating all entries using the
4368	>	* given reducer to combine values, or null if none.
4369	>	*
4370	>	* @param reducer a commutative associative combining function
4371	>	* @return the result of accumulating all entries
4372	>	*/
4373	>	public Map.Entry<K,V> reduceEntriesInParallel
4374	>	(BiFunction<Map.Entry<K,V>, Map.Entry<K,V>, ? extends Map.Entry<K,V>> reducer) {
4375	>	return ForkJoinTasks.reduceEntries
4376	>	(this, reducer).invoke();
4377	>	}
4378	>
4379	>	/**
4380	>	* Returns the result of accumulating the given transformation
4381	>	* of all entries using the given reducer to combine values,
4382	>	* or null if none.
4383	>	*
4384	>	* @param transformer a function returning the transformation
4385	>	* for an element, or null of there is no transformation (in
4386	>	* which case it is not combined).
4387	>	* @param reducer a commutative associative combining function
4388	>	* @return the result of accumulating the given transformation
4389	>	* of all entries
4390	>	*/
4391	>	public <U> U reduceEntriesInParallel
4392	>	(Function<Map.Entry<K,V>, ? extends U> transformer,
4393	>	BiFunction<? super U, ? super U, ? extends U> reducer) {
4394	>	return ForkJoinTasks.reduceEntries
4395	>	(this, transformer, reducer).invoke();
4396	>	}
4397	>
4398	>	/**
4399	>	* Returns the result of accumulating the given transformation
4400	>	* of all entries using the given reducer to combine values,
4401	>	* and the given basis as an identity value.
4402	>	*
4403	>	* @param transformer a function returning the transformation
4404	>	* for an element
4405	>	* @param basis the identity (initial default value) for the reduction
4406	>	* @param reducer a commutative associative combining function
4407	>	* @return the result of accumulating the given transformation
4408	>	* of all entries
4409	>	*/
4410	>	public double reduceEntriesToDoubleInParallel
4411	>	(DoubleFunction<Map.Entry<K,V>> transformer,
4412	>	double basis,
4413	>	DoubleBinaryOperator reducer) {
4414	>	return ForkJoinTasks.reduceEntriesToDouble
4415	>	(this, transformer, basis, reducer).invoke();
4416	>	}
4417	>
4418	>	/**
4419	>	* Returns the result of accumulating the given transformation
4420	>	* of all entries using the given reducer to combine values,
4421	>	* and the given basis as an identity value.
4422	>	*
4423	>	* @param transformer a function returning the transformation
4424	>	* for an element
4425	>	* @param basis the identity (initial default value) for the reduction
4426	>	* @param reducer a commutative associative combining function
4427	>	* @return  the result of accumulating the given transformation
4428	>	* of all entries
4429	>	*/
4430	>	public long reduceEntriesToLongInParallel
4431	>	(LongFunction<Map.Entry<K,V>> transformer,
4432	>	long basis,
4433	>	LongBinaryOperator reducer) {
4434	>	return ForkJoinTasks.reduceEntriesToLong
4435	>	(this, transformer, basis, reducer).invoke();
4436	>	}
4437	>
4438	>	/**
4439	>	* Returns the result of accumulating the given transformation
4440	>	* of all entries using the given reducer to combine values,
4441	>	* and the given basis as an identity value.
4442	>	*
4443	>	* @param transformer a function returning the transformation
4444	>	* for an element
4445	>	* @param basis the identity (initial default value) for the reduction
4446	>	* @param reducer a commutative associative combining function
4447	>	* @return the result of accumulating the given transformation
4448	>	* of all entries
4449	>	*/
4450	>	public int reduceEntriesToIntInParallel
4451	>	(IntFunction<Map.Entry<K,V>> transformer,
4452	>	int basis,
4453	>	IntBinaryOperator reducer) {
4454	>	return ForkJoinTasks.reduceEntriesToInt
4455	>	(this, transformer, basis, reducer).invoke();
4456	>	}
4457	>
4458	>
4459	>	/* ----------------Views -------------- */
4460	>
4461	>	/**
4462	>	* Base class for views.
4463	>	*/
4464	>	static abstract class CHMView<K, V> {
4465	>	final ConcurrentHashMap<K, V> map;
4466	>	CHMView(ConcurrentHashMap<K, V> map)  { this.map = map; }
4467	>
4468	>	/**
4469	>	* Returns the map backing this view.
4470	>	*
4471	>	* @return the map backing this view
4472	>	*/
4473	>	public ConcurrentHashMap<K,V> getMap() { return map; }
4474	>
4475	>	public final int size()                 { return map.size(); }
4476	>	public final boolean isEmpty()          { return map.isEmpty(); }
4477	>	public final void clear()               { map.clear(); }
4478	>
4479	>	// implementations below rely on concrete classes supplying these
4480	>	abstract public Iterator<?> iterator();
4481	>	abstract public boolean contains(Object o);
4482	>	abstract public boolean remove(Object o);
4483	>
4484	>	private static final String oomeMsg = "Required array size too large";
4485	>
4486	>	public final Object[] toArray() {
4487	>	long sz = map.mappingCount();
4488	>	if (sz > (long)(MAX_ARRAY_SIZE))
4489	>	throw new OutOfMemoryError(oomeMsg);
4490	>	int n = (int)sz;
4491	>	Object[] r = new Object[n];
4492	>	int i = 0;
4493	>	Iterator<?> it = iterator();
4494	>	while (it.hasNext()) {
4495	>	if (i == n) {
4496	>	if (n >= MAX_ARRAY_SIZE)
4497	>	throw new OutOfMemoryError(oomeMsg);
4498	>	if (n >= MAX_ARRAY_SIZE - (MAX_ARRAY_SIZE >>> 1) - 1)
4499	>	n = MAX_ARRAY_SIZE;
4500	>	else
4501	>	n += (n >>> 1) + 1;
4502	>	r = Arrays.copyOf(r, n);
4503	>	}
4504	>	r[i++] = it.next();
4505	>	}
4506	>	return (i == n) ? r : Arrays.copyOf(r, i);
4507		}
4508	<	public void clear() {
4509	<	ConcurrentHashMap.this.clear();
4508	>
4509	>	@SuppressWarnings("unchecked") public final <T> T[] toArray(T[] a) {
4510	>	long sz = map.mappingCount();
4511	>	if (sz > (long)(MAX_ARRAY_SIZE))
4512	>	throw new OutOfMemoryError(oomeMsg);
4513	>	int m = (int)sz;
4514	>	T[] r = (a.length >= m) ? a :
4515	>	(T[])java.lang.reflect.Array
4516	>	.newInstance(a.getClass().getComponentType(), m);
4517	>	int n = r.length;
4518	>	int i = 0;
4519	>	Iterator<?> it = iterator();
4520	>	while (it.hasNext()) {
4521	>	if (i == n) {
4522	>	if (n >= MAX_ARRAY_SIZE)
4523	>	throw new OutOfMemoryError(oomeMsg);
4524	>	if (n >= MAX_ARRAY_SIZE - (MAX_ARRAY_SIZE >>> 1) - 1)
4525	>	n = MAX_ARRAY_SIZE;
4526	>	else
4527	>	n += (n >>> 1) + 1;
4528	>	r = Arrays.copyOf(r, n);
4529	>	}
4530	>	r[i++] = (T)it.next();
4531	>	}
4532	>	if (a == r && i < n) {
4533	>	r[i] = null; // null-terminate
4534	>	return r;
4535	>	}
4536	>	return (i == n) ? r : Arrays.copyOf(r, i);
4537		}
4538	<	public Object[] toArray() {
4539	<	Collection<Map.Entry<K,V>> c = new ArrayList<Map.Entry<K,V>>(size());
4540	<	for (Map.Entry<K,V> e : this)
4541	<	c.add(e);
4542	<	return c.toArray();
4543	<	}
4544	<	public <T> T[] toArray(T[] a) {
4545	<	Collection<Map.Entry<K,V>> c = new ArrayList<Map.Entry<K,V>>(size());
4546	<	for (Map.Entry<K,V> e : this)
4547	<	c.add(e);
4548	<	return c.toArray(a);
4538	>
4539	>	public final int hashCode() {
4540	>	int h = 0;
4541	>	for (Iterator<?> it = iterator(); it.hasNext();)
4542	>	h += it.next().hashCode();
4543	>	return h;
4544	>	}
4545	>
4546	>	public final String toString() {
4547	>	StringBuilder sb = new StringBuilder();
4548	>	sb.append('[');
4549	>	Iterator<?> it = iterator();
4550	>	if (it.hasNext()) {
4551	>	for (;;) {
4552	>	Object e = it.next();
4553	>	sb.append(e == this ? "(this Collection)" : e);
4554	>	if (!it.hasNext())
4555	>	break;
4556	>	sb.append(',').append(' ');
4557	>	}
4558	>	}
4559	>	return sb.append(']').toString();
4560	>	}
4561	>
4562	>	public final boolean containsAll(Collection<?> c) {
4563	>	if (c != this) {
4564	>	for (Iterator<?> it = c.iterator(); it.hasNext();) {
4565	>	Object e = it.next();
4566	>	if (e == null || !contains(e))
4567	>	return false;
4568	>	}
4569	>	}
4570	>	return true;
4571	>	}
4572	>
4573	>	public final boolean removeAll(Collection<?> c) {
4574	>	boolean modified = false;
4575	>	for (Iterator<?> it = iterator(); it.hasNext();) {
4576	>	if (c.contains(it.next())) {
4577	>	it.remove();
4578	>	modified = true;
4579	>	}
4580	>	}
4581	>	return modified;
4582	>	}
4583	>
4584	>	public final boolean retainAll(Collection<?> c) {
4585	>	boolean modified = false;
4586	>	for (Iterator<?> it = iterator(); it.hasNext();) {
4587	>	if (!c.contains(it.next())) {
4588	>	it.remove();
4589	>	modified = true;
4590	>	}
4591	>	}
4592	>	return modified;
4593		}
4594
4595		}
4596
4597	<	/* ---------------- Serialization Support -------------- */
4597	>	/**
4598	>	* A view of a ConcurrentHashMap as a {@link Set} of keys, in
4599	>	* which additions may optionally be enabled by mapping to a
4600	>	* common value.  This class cannot be directly instantiated. See
4601	>	* {@link #keySet}, {@link #keySet(Object)}, {@link #newKeySet()},
4602	>	* {@link #newKeySet(int)}.
4603	>	*/
4604	>	public static class KeySetView<K,V> extends CHMView<K,V>
4605	>	implements Set<K>, java.io.Serializable {
4606	>	private static final long serialVersionUID = 7249069246763182397L;
4607	>	private final V value;
4608	>	KeySetView(ConcurrentHashMap<K, V> map, V value) {  // non-public
4609	>	super(map);
4610	>	this.value = value;
4611	>	}
4612	>
4613	>	/**
4614	>	* Returns the default mapped value for additions,
4615	>	* or {@code null} if additions are not supported.
4616	>	*
4617	>	* @return the default mapped value for additions, or {@code null}
4618	>	* if not supported.
4619	>	*/
4620	>	public V getMappedValue() { return value; }
4621	>
4622	>	// implement Set API
4623	>
4624	>	public boolean contains(Object o) { return map.containsKey(o); }
4625	>	public boolean remove(Object o)   { return map.remove(o) != null; }
4626	>
4627	>	/**
4628	>	* Returns a "weakly consistent" iterator that will never
4629	>	* throw {@link ConcurrentModificationException}, and
4630	>	* guarantees to traverse elements as they existed upon
4631	>	* construction of the iterator, and may (but is not
4632	>	* guaranteed to) reflect any modifications subsequent to
4633	>	* construction.
4634	>	*
4635	>	* @return an iterator over the keys of this map
4636	>	*/
4637	>	public Iterator<K> iterator()     { return new KeyIterator<K,V>(map); }
4638	>	public boolean add(K e) {
4639	>	V v;
4640	>	if ((v = value) == null)
4641	>	throw new UnsupportedOperationException();
4642	>	if (e == null)
4643	>	throw new NullPointerException();
4644	>	return map.internalPut(e, v, true) == null;
4645	>	}
4646	>	public boolean addAll(Collection<? extends K> c) {
4647	>	boolean added = false;
4648	>	V v;
4649	>	if ((v = value) == null)
4650	>	throw new UnsupportedOperationException();
4651	>	for (K e : c) {
4652	>	if (e == null)
4653	>	throw new NullPointerException();
4654	>	if (map.internalPut(e, v, true) == null)
4655	>	added = true;
4656	>	}
4657	>	return added;
4658	>	}
4659	>	public boolean equals(Object o) {
4660	>	Set<?> c;
4661	>	return ((o instanceof Set) &&
4662	>	((c = (Set<?>)o) == this ||
4663	>	(containsAll(c) && c.containsAll(this))));
4664	>	}
4665	>
4666	>	public Stream<K> stream() {
4667	>	return Streams.stream(() -> new KeyIterator<K,V>(map), 0);
4668	>	}
4669	>	public Stream<K> parallelStream() {
4670	>	return Streams.parallelStream(() -> new KeyIterator<K,V>(map, null),
4671	>	0);
4672	>	}
4673	>	}
4674
4675		/**
4676	<	* Save the state of the <tt>ConcurrentHashMap</tt> instance to a
4677	<	* stream (i.e., serialize it).
4678	<	* @param s the stream
4679	<	* @serialData
4680	<	* the key (Object) and value (Object)
4681	<	* for each key-value mapping, followed by a null pair.
4682	<	* The key-value mappings are emitted in no particular order.
4676	>	* A view of a ConcurrentHashMap as a {@link Collection} of
4677	>	* values, in which additions are disabled. This class cannot be
4678	>	* directly instantiated. See {@link #values},
4679	>	*
4680	>	* <p>The view's {@code iterator} is a "weakly consistent" iterator
4681	>	* that will never throw {@link ConcurrentModificationException},
4682	>	* and guarantees to traverse elements as they existed upon
4683	>	* construction of the iterator, and may (but is not guaranteed to)
4684	>	* reflect any modifications subsequent to construction.
4685		*/
4686	<	private void writeObject(java.io.ObjectOutputStream s) throws IOException  {
4687	<	s.defaultWriteObject();
4688	<
4689	<	for (int k = 0; k < segments.length; ++k) {
4690	<	Segment<K,V> seg = segments[k];
4691	<	seg.lock();
4692	<	try {
4693	<	HashEntry<K,V>[] tab = seg.table;
4694	<	for (int i = 0; i < tab.length; ++i) {
4695	<	for (HashEntry<K,V> e = tab[i]; e != null; e = e.next) {
4696	<	s.writeObject(e.key);
1280	<	s.writeObject(e.value);
4686	>	public static final class ValuesView<K,V> extends CHMView<K,V>
4687	>	implements Collection<V> {
4688	>	ValuesView(ConcurrentHashMap<K, V> map)   { super(map); }
4689	>	public final boolean contains(Object o) { return map.containsValue(o); }
4690	>	public final boolean remove(Object o) {
4691	>	if (o != null) {
4692	>	Iterator<V> it = new ValueIterator<K,V>(map);
4693	>	while (it.hasNext()) {
4694	>	if (o.equals(it.next())) {
4695	>	it.remove();
4696	>	return true;
4697		}
4698		}
1283	–	} finally {
1284	–	seg.unlock();
4699		}
4700	+	return false;
4701		}
4702	<	s.writeObject(null);
4703	<	s.writeObject(null);
4702	>
4703	>	/**
4704	>	* Returns a "weakly consistent" iterator that will never
4705	>	* throw {@link ConcurrentModificationException}, and
4706	>	* guarantees to traverse elements as they existed upon
4707	>	* construction of the iterator, and may (but is not
4708	>	* guaranteed to) reflect any modifications subsequent to
4709	>	* construction.
4710	>	*
4711	>	* @return an iterator over the values of this map
4712	>	*/
4713	>	public final Iterator<V> iterator() {
4714	>	return new ValueIterator<K,V>(map);
4715	>	}
4716	>	public final boolean add(V e) {
4717	>	throw new UnsupportedOperationException();
4718	>	}
4719	>	public final boolean addAll(Collection<? extends V> c) {
4720	>	throw new UnsupportedOperationException();
4721	>	}
4722	>
4723	>	public Stream<V> stream() {
4724	>	return Streams.stream(() -> new ValueIterator<K,V>(map), 0);
4725	>	}
4726	>
4727	>	public Stream<V> parallelStream() {
4728	>	return Streams.parallelStream(() -> new ValueIterator<K,V>(map, null),
4729	>	0);
4730	>	}
4731	>
4732		}
4733
4734		/**
4735	<	* Reconstitute the <tt>ConcurrentHashMap</tt> instance from a
4736	<	* stream (i.e., deserialize it).
4737	<	* @param s the stream
4735	>	* A view of a ConcurrentHashMap as a {@link Set} of (key, value)
4736	>	* entries.  This class cannot be directly instantiated. See
4737	>	* {@link #entrySet}.
4738	>	*/
4739	>	public static final class EntrySetView<K,V> extends CHMView<K,V>
4740	>	implements Set<Map.Entry<K,V>> {
4741	>	EntrySetView(ConcurrentHashMap<K, V> map) { super(map); }
4742	>	public final boolean contains(Object o) {
4743	>	Object k, v, r; Map.Entry<?,?> e;
4744	>	return ((o instanceof Map.Entry) &&
4745	>	(k = (e = (Map.Entry<?,?>)o).getKey()) != null &&
4746	>	(r = map.get(k)) != null &&
4747	>	(v = e.getValue()) != null &&
4748	>	(v == r || v.equals(r)));
4749	>	}
4750	>	public final boolean remove(Object o) {
4751	>	Object k, v; Map.Entry<?,?> e;
4752	>	return ((o instanceof Map.Entry) &&
4753	>	(k = (e = (Map.Entry<?,?>)o).getKey()) != null &&
4754	>	(v = e.getValue()) != null &&
4755	>	map.remove(k, v));
4756	>	}
4757	>
4758	>	/**
4759	>	* Returns a "weakly consistent" iterator that will never
4760	>	* throw {@link ConcurrentModificationException}, and
4761	>	* guarantees to traverse elements as they existed upon
4762	>	* construction of the iterator, and may (but is not
4763	>	* guaranteed to) reflect any modifications subsequent to
4764	>	* construction.
4765	>	*
4766	>	* @return an iterator over the entries of this map
4767	>	*/
4768	>	public final Iterator<Map.Entry<K,V>> iterator() {
4769	>	return new EntryIterator<K,V>(map);
4770	>	}
4771	>
4772	>	public final boolean add(Entry<K,V> e) {
4773	>	K key = e.getKey();
4774	>	V value = e.getValue();
4775	>	if (key == null || value == null)
4776	>	throw new NullPointerException();
4777	>	return map.internalPut(key, value, false) == null;
4778	>	}
4779	>	public final boolean addAll(Collection<? extends Entry<K,V>> c) {
4780	>	boolean added = false;
4781	>	for (Entry<K,V> e : c) {
4782	>	if (add(e))
4783	>	added = true;
4784	>	}
4785	>	return added;
4786	>	}
4787	>	public boolean equals(Object o) {
4788	>	Set<?> c;
4789	>	return ((o instanceof Set) &&
4790	>	((c = (Set<?>)o) == this ||
4791	>	(containsAll(c) && c.containsAll(this))));
4792	>	}
4793	>
4794	>	public Stream<Map.Entry<K,V>> stream() {
4795	>	return Streams.stream(() -> new EntryIterator<K,V>(map), 0);
4796	>	}
4797	>
4798	>	public Stream<Map.Entry<K,V>> parallelStream() {
4799	>	return Streams.parallelStream(() -> new EntryIterator<K,V>(map, null),
4800	>	0);
4801	>	}
4802	>	}
4803	>
4804	>	// ---------------------------------------------------------------------
4805	>
4806	>	/**
4807	>	* Predefined tasks for performing bulk parallel operations on
4808	>	* ConcurrentHashMaps. These tasks follow the forms and rules used
4809	>	* for bulk operations. Each method has the same name, but returns
4810	>	* a task rather than invoking it. These methods may be useful in
4811	>	* custom applications such as submitting a task without waiting
4812	>	* for completion, using a custom pool, or combining with other
4813	>	* tasks.
4814		*/
4815	<	private void readObject(java.io.ObjectInputStream s)
4816	<	throws IOException, ClassNotFoundException  {
1298	<	s.defaultReadObject();
4815	>	public static class ForkJoinTasks {
4816	>	private ForkJoinTasks() {}
4817
4818	<	// Initialize each segment to be minimally sized, and let grow.
4819	<	for (int i = 0; i < segments.length; ++i) {
4820	<	segments[i].setTable(new HashEntry[1]);
4818	>	/**
4819	>	* Returns a task that when invoked, performs the given
4820	>	* action for each (key, value)
4821	>	*
4822	>	* @param map the map
4823	>	* @param action the action
4824	>	* @return the task
4825	>	*/
4826	>	public static <K,V> ForkJoinTask<Void> forEach
4827	>	(ConcurrentHashMap<K,V> map,
4828	>	BiBlock<? super K, ? super V> action) {
4829	>	if (action == null) throw new NullPointerException();
4830	>	return new ForEachMappingTask<K,V>(map, null, -1, action);
4831		}
4832
4833	<	// Read the keys and values, and put the mappings in the table
4834	<	for (;;) {
4835	<	K key = (K) s.readObject();
4836	<	V value = (V) s.readObject();
4837	<	if (key == null)
4838	<	break;
4839	<	put(key, value);
4833	>	/**
4834	>	* Returns a task that when invoked, performs the given
4835	>	* action for each non-null transformation of each (key, value)
4836	>	*
4837	>	* @param map the map
4838	>	* @param transformer a function returning the transformation
4839	>	* for an element, or null if there is no transformation (in
4840	>	* which case the action is not applied)
4841	>	* @param action the action
4842	>	* @return the task
4843	>	*/
4844	>	public static <K,V,U> ForkJoinTask<Void> forEach
4845	>	(ConcurrentHashMap<K,V> map,
4846	>	BiFunction<? super K, ? super V, ? extends U> transformer,
4847	>	Block<? super U> action) {
4848	>	if (transformer == null || action == null)
4849	>	throw new NullPointerException();
4850	>	return new ForEachTransformedMappingTask<K,V,U>
4851	>	(map, null, -1, transformer, action);
4852	>	}
4853	>
4854	>	/**
4855	>	* Returns a task that when invoked, returns a non-null result
4856	>	* from applying the given search function on each (key,
4857	>	* value), or null if none. Upon success, further element
4858	>	* processing is suppressed and the results of any other
4859	>	* parallel invocations of the search function are ignored.
4860	>	*
4861	>	* @param map the map
4862	>	* @param searchFunction a function returning a non-null
4863	>	* result on success, else null
4864	>	* @return the task
4865	>	*/
4866	>	public static <K,V,U> ForkJoinTask<U> search
4867	>	(ConcurrentHashMap<K,V> map,
4868	>	BiFunction<? super K, ? super V, ? extends U> searchFunction) {
4869	>	if (searchFunction == null) throw new NullPointerException();
4870	>	return new SearchMappingsTask<K,V,U>
4871	>	(map, null, -1, searchFunction,
4872	>	new AtomicReference<U>());
4873	>	}
4874	>
4875	>	/**
4876	>	* Returns a task that when invoked, returns the result of
4877	>	* accumulating the given transformation of all (key, value) pairs
4878	>	* using the given reducer to combine values, or null if none.
4879	>	*
4880	>	* @param map the map
4881	>	* @param transformer a function returning the transformation
4882	>	* for an element, or null if there is no transformation (in
4883	>	* which case it is not combined).
4884	>	* @param reducer a commutative associative combining function
4885	>	* @return the task
4886	>	*/
4887	>	public static <K,V,U> ForkJoinTask<U> reduce
4888	>	(ConcurrentHashMap<K,V> map,
4889	>	BiFunction<? super K, ? super V, ? extends U> transformer,
4890	>	BiFunction<? super U, ? super U, ? extends U> reducer) {
4891	>	if (transformer == null || reducer == null)
4892	>	throw new NullPointerException();
4893	>	return new MapReduceMappingsTask<K,V,U>
4894	>	(map, null, -1, null, transformer, reducer);
4895	>	}
4896	>
4897	>	/**
4898	>	* Returns a task that when invoked, returns the result of
4899	>	* accumulating the given transformation of all (key, value) pairs
4900	>	* using the given reducer to combine values, and the given
4901	>	* basis as an identity value.
4902	>	*
4903	>	* @param map the map
4904	>	* @param transformer a function returning the transformation
4905	>	* for an element
4906	>	* @param basis the identity (initial default value) for the reduction
4907	>	* @param reducer a commutative associative combining function
4908	>	* @return the task
4909	>	*/
4910	>	public static <K,V> ForkJoinTask<Double> reduceToDouble
4911	>	(ConcurrentHashMap<K,V> map,
4912	>	DoubleBiFunction<? super K, ? super V> transformer,
4913	>	double basis,
4914	>	DoubleBinaryOperator reducer) {
4915	>	if (transformer == null || reducer == null)
4916	>	throw new NullPointerException();
4917	>	return new MapReduceMappingsToDoubleTask<K,V>
4918	>	(map, null, -1, null, transformer, basis, reducer);
4919	>	}
4920	>
4921	>	/**
4922	>	* Returns a task that when invoked, returns the result of
4923	>	* accumulating the given transformation of all (key, value) pairs
4924	>	* using the given reducer to combine values, and the given
4925	>	* basis as an identity value.
4926	>	*
4927	>	* @param map the map
4928	>	* @param transformer a function returning the transformation
4929	>	* for an element
4930	>	* @param basis the identity (initial default value) for the reduction
4931	>	* @param reducer a commutative associative combining function
4932	>	* @return the task
4933	>	*/
4934	>	public static <K,V> ForkJoinTask<Long> reduceToLong
4935	>	(ConcurrentHashMap<K,V> map,
4936	>	LongBiFunction<? super K, ? super V> transformer,
4937	>	long basis,
4938	>	LongBinaryOperator reducer) {
4939	>	if (transformer == null || reducer == null)
4940	>	throw new NullPointerException();
4941	>	return new MapReduceMappingsToLongTask<K,V>
4942	>	(map, null, -1, null, transformer, basis, reducer);
4943	>	}
4944	>
4945	>	/**
4946	>	* Returns a task that when invoked, returns the result of
4947	>	* accumulating the given transformation of all (key, value) pairs
4948	>	* using the given reducer to combine values, and the given
4949	>	* basis as an identity value.
4950	>	*
4951	>	* @param transformer a function returning the transformation
4952	>	* for an element
4953	>	* @param basis the identity (initial default value) for the reduction
4954	>	* @param reducer a commutative associative combining function
4955	>	* @return the task
4956	>	*/
4957	>	public static <K,V> ForkJoinTask<Integer> reduceToInt
4958	>	(ConcurrentHashMap<K,V> map,
4959	>	IntBiFunction<? super K, ? super V> transformer,
4960	>	int basis,
4961	>	IntBinaryOperator reducer) {
4962	>	if (transformer == null || reducer == null)
4963	>	throw new NullPointerException();
4964	>	return new MapReduceMappingsToIntTask<K,V>
4965	>	(map, null, -1, null, transformer, basis, reducer);
4966	>	}
4967	>
4968	>	/**
4969	>	* Returns a task that when invoked, performs the given action
4970	>	* for each key.
4971	>	*
4972	>	* @param map the map
4973	>	* @param action the action
4974	>	* @return the task
4975	>	*/
4976	>	public static <K,V> ForkJoinTask<Void> forEachKey
4977	>	(ConcurrentHashMap<K,V> map,
4978	>	Block<? super K> action) {
4979	>	if (action == null) throw new NullPointerException();
4980	>	return new ForEachKeyTask<K,V>(map, null, -1, action);
4981	>	}
4982	>
4983	>	/**
4984	>	* Returns a task that when invoked, performs the given action
4985	>	* for each non-null transformation of each key.
4986	>	*
4987	>	* @param map the map
4988	>	* @param transformer a function returning the transformation
4989	>	* for an element, or null if there is no transformation (in
4990	>	* which case the action is not applied)
4991	>	* @param action the action
4992	>	* @return the task
4993	>	*/
4994	>	public static <K,V,U> ForkJoinTask<Void> forEachKey
4995	>	(ConcurrentHashMap<K,V> map,
4996	>	Function<? super K, ? extends U> transformer,
4997	>	Block<? super U> action) {
4998	>	if (transformer == null || action == null)
4999	>	throw new NullPointerException();
5000	>	return new ForEachTransformedKeyTask<K,V,U>
5001	>	(map, null, -1, transformer, action);
5002	>	}
5003	>
5004	>	/**
5005	>	* Returns a task that when invoked, returns a non-null result
5006	>	* from applying the given search function on each key, or
5007	>	* null if none.  Upon success, further element processing is
5008	>	* suppressed and the results of any other parallel
5009	>	* invocations of the search function are ignored.
5010	>	*
5011	>	* @param map the map
5012	>	* @param searchFunction a function returning a non-null
5013	>	* result on success, else null
5014	>	* @return the task
5015	>	*/
5016	>	public static <K,V,U> ForkJoinTask<U> searchKeys
5017	>	(ConcurrentHashMap<K,V> map,
5018	>	Function<? super K, ? extends U> searchFunction) {
5019	>	if (searchFunction == null) throw new NullPointerException();
5020	>	return new SearchKeysTask<K,V,U>
5021	>	(map, null, -1, searchFunction,
5022	>	new AtomicReference<U>());
5023	>	}
5024	>
5025	>	/**
5026	>	* Returns a task that when invoked, returns the result of
5027	>	* accumulating all keys using the given reducer to combine
5028	>	* values, or null if none.
5029	>	*
5030	>	* @param map the map
5031	>	* @param reducer a commutative associative combining function
5032	>	* @return the task
5033	>	*/
5034	>	public static <K,V> ForkJoinTask<K> reduceKeys
5035	>	(ConcurrentHashMap<K,V> map,
5036	>	BiFunction<? super K, ? super K, ? extends K> reducer) {
5037	>	if (reducer == null) throw new NullPointerException();
5038	>	return new ReduceKeysTask<K,V>
5039	>	(map, null, -1, null, reducer);
5040	>	}
5041	>
5042	>	/**
5043	>	* Returns a task that when invoked, returns the result of
5044	>	* accumulating the given transformation of all keys using the given
5045	>	* reducer to combine values, or null if none.
5046	>	*
5047	>	* @param map the map
5048	>	* @param transformer a function returning the transformation
5049	>	* for an element, or null if there is no transformation (in
5050	>	* which case it is not combined).
5051	>	* @param reducer a commutative associative combining function
5052	>	* @return the task
5053	>	*/
5054	>	public static <K,V,U> ForkJoinTask<U> reduceKeys
5055	>	(ConcurrentHashMap<K,V> map,
5056	>	Function<? super K, ? extends U> transformer,
5057	>	BiFunction<? super U, ? super U, ? extends U> reducer) {
5058	>	if (transformer == null || reducer == null)
5059	>	throw new NullPointerException();
5060	>	return new MapReduceKeysTask<K,V,U>
5061	>	(map, null, -1, null, transformer, reducer);
5062	>	}
5063	>
5064	>	/**
5065	>	* Returns a task that when invoked, returns the result of
5066	>	* accumulating the given transformation of all keys using the given
5067	>	* reducer to combine values, and the given basis as an
5068	>	* identity value.
5069	>	*
5070	>	* @param map the map
5071	>	* @param transformer a function returning the transformation
5072	>	* for an element
5073	>	* @param basis the identity (initial default value) for the reduction
5074	>	* @param reducer a commutative associative combining function
5075	>	* @return the task
5076	>	*/
5077	>	public static <K,V> ForkJoinTask<Double> reduceKeysToDouble
5078	>	(ConcurrentHashMap<K,V> map,
5079	>	DoubleFunction<? super K> transformer,
5080	>	double basis,
5081	>	DoubleBinaryOperator reducer) {
5082	>	if (transformer == null || reducer == null)
5083	>	throw new NullPointerException();
5084	>	return new MapReduceKeysToDoubleTask<K,V>
5085	>	(map, null, -1, null, transformer, basis, reducer);
5086	>	}
5087	>
5088	>	/**
5089	>	* Returns a task that when invoked, returns the result of
5090	>	* accumulating the given transformation of all keys using the given
5091	>	* reducer to combine values, and the given basis as an
5092	>	* identity value.
5093	>	*
5094	>	* @param map the map
5095	>	* @param transformer a function returning the transformation
5096	>	* for an element
5097	>	* @param basis the identity (initial default value) for the reduction
5098	>	* @param reducer a commutative associative combining function
5099	>	* @return the task
5100	>	*/
5101	>	public static <K,V> ForkJoinTask<Long> reduceKeysToLong
5102	>	(ConcurrentHashMap<K,V> map,
5103	>	LongFunction<? super K> transformer,
5104	>	long basis,
5105	>	LongBinaryOperator reducer) {
5106	>	if (transformer == null || reducer == null)
5107	>	throw new NullPointerException();
5108	>	return new MapReduceKeysToLongTask<K,V>
5109	>	(map, null, -1, null, transformer, basis, reducer);
5110	>	}
5111	>
5112	>	/**
5113	>	* Returns a task that when invoked, returns the result of
5114	>	* accumulating the given transformation of all keys using the given
5115	>	* reducer to combine values, and the given basis as an
5116	>	* identity value.
5117	>	*
5118	>	* @param map the map
5119	>	* @param transformer a function returning the transformation
5120	>	* for an element
5121	>	* @param basis the identity (initial default value) for the reduction
5122	>	* @param reducer a commutative associative combining function
5123	>	* @return the task
5124	>	*/
5125	>	public static <K,V> ForkJoinTask<Integer> reduceKeysToInt
5126	>	(ConcurrentHashMap<K,V> map,
5127	>	IntFunction<? super K> transformer,
5128	>	int basis,
5129	>	IntBinaryOperator reducer) {
5130	>	if (transformer == null || reducer == null)
5131	>	throw new NullPointerException();
5132	>	return new MapReduceKeysToIntTask<K,V>
5133	>	(map, null, -1, null, transformer, basis, reducer);
5134	>	}
5135	>
5136	>	/**
5137	>	* Returns a task that when invoked, performs the given action
5138	>	* for each value.
5139	>	*
5140	>	* @param map the map
5141	>	* @param action the action
5142	>	*/
5143	>	public static <K,V> ForkJoinTask<Void> forEachValue
5144	>	(ConcurrentHashMap<K,V> map,
5145	>	Block<? super V> action) {
5146	>	if (action == null) throw new NullPointerException();
5147	>	return new ForEachValueTask<K,V>(map, null, -1, action);
5148	>	}
5149	>
5150	>	/**
5151	>	* Returns a task that when invoked, performs the given action
5152	>	* for each non-null transformation of each value.
5153	>	*
5154	>	* @param map the map
5155	>	* @param transformer a function returning the transformation
5156	>	* for an element, or null if there is no transformation (in
5157	>	* which case the action is not applied)
5158	>	* @param action the action
5159	>	*/
5160	>	public static <K,V,U> ForkJoinTask<Void> forEachValue
5161	>	(ConcurrentHashMap<K,V> map,
5162	>	Function<? super V, ? extends U> transformer,
5163	>	Block<? super U> action) {
5164	>	if (transformer == null || action == null)
5165	>	throw new NullPointerException();
5166	>	return new ForEachTransformedValueTask<K,V,U>
5167	>	(map, null, -1, transformer, action);
5168	>	}
5169	>
5170	>	/**
5171	>	* Returns a task that when invoked, returns a non-null result
5172	>	* from applying the given search function on each value, or
5173	>	* null if none.  Upon success, further element processing is
5174	>	* suppressed and the results of any other parallel
5175	>	* invocations of the search function are ignored.
5176	>	*
5177	>	* @param map the map
5178	>	* @param searchFunction a function returning a non-null
5179	>	* result on success, else null
5180	>	* @return the task
5181	>	*/
5182	>	public static <K,V,U> ForkJoinTask<U> searchValues
5183	>	(ConcurrentHashMap<K,V> map,
5184	>	Function<? super V, ? extends U> searchFunction) {
5185	>	if (searchFunction == null) throw new NullPointerException();
5186	>	return new SearchValuesTask<K,V,U>
5187	>	(map, null, -1, searchFunction,
5188	>	new AtomicReference<U>());
5189	>	}
5190	>
5191	>	/**
5192	>	* Returns a task that when invoked, returns the result of
5193	>	* accumulating all values using the given reducer to combine
5194	>	* values, or null if none.
5195	>	*
5196	>	* @param map the map
5197	>	* @param reducer a commutative associative combining function
5198	>	* @return the task
5199	>	*/
5200	>	public static <K,V> ForkJoinTask<V> reduceValues
5201	>	(ConcurrentHashMap<K,V> map,
5202	>	BiFunction<? super V, ? super V, ? extends V> reducer) {
5203	>	if (reducer == null) throw new NullPointerException();
5204	>	return new ReduceValuesTask<K,V>
5205	>	(map, null, -1, null, reducer);
5206	>	}
5207	>
5208	>	/**
5209	>	* Returns a task that when invoked, returns the result of
5210	>	* accumulating the given transformation of all values using the
5211	>	* given reducer to combine values, or null if none.
5212	>	*
5213	>	* @param map the map
5214	>	* @param transformer a function returning the transformation
5215	>	* for an element, or null if there is no transformation (in
5216	>	* which case it is not combined).
5217	>	* @param reducer a commutative associative combining function
5218	>	* @return the task
5219	>	*/
5220	>	public static <K,V,U> ForkJoinTask<U> reduceValues
5221	>	(ConcurrentHashMap<K,V> map,
5222	>	Function<? super V, ? extends U> transformer,
5223	>	BiFunction<? super U, ? super U, ? extends U> reducer) {
5224	>	if (transformer == null || reducer == null)
5225	>	throw new NullPointerException();
5226	>	return new MapReduceValuesTask<K,V,U>
5227	>	(map, null, -1, null, transformer, reducer);
5228	>	}
5229	>
5230	>	/**
5231	>	* Returns a task that when invoked, returns the result of
5232	>	* accumulating the given transformation of all values using the
5233	>	* given reducer to combine values, and the given basis as an
5234	>	* identity value.
5235	>	*
5236	>	* @param map the map
5237	>	* @param transformer a function returning the transformation
5238	>	* for an element
5239	>	* @param basis the identity (initial default value) for the reduction
5240	>	* @param reducer a commutative associative combining function
5241	>	* @return the task
5242	>	*/
5243	>	public static <K,V> ForkJoinTask<Double> reduceValuesToDouble
5244	>	(ConcurrentHashMap<K,V> map,
5245	>	DoubleFunction<? super V> transformer,
5246	>	double basis,
5247	>	DoubleBinaryOperator reducer) {
5248	>	if (transformer == null || reducer == null)
5249	>	throw new NullPointerException();
5250	>	return new MapReduceValuesToDoubleTask<K,V>
5251	>	(map, null, -1, null, transformer, basis, reducer);
5252	>	}
5253	>
5254	>	/**
5255	>	* Returns a task that when invoked, returns the result of
5256	>	* accumulating the given transformation of all values using the
5257	>	* given reducer to combine values, and the given basis as an
5258	>	* identity value.
5259	>	*
5260	>	* @param map the map
5261	>	* @param transformer a function returning the transformation
5262	>	* for an element
5263	>	* @param basis the identity (initial default value) for the reduction
5264	>	* @param reducer a commutative associative combining function
5265	>	* @return the task
5266	>	*/
5267	>	public static <K,V> ForkJoinTask<Long> reduceValuesToLong
5268	>	(ConcurrentHashMap<K,V> map,
5269	>	LongFunction<? super V> transformer,
5270	>	long basis,
5271	>	LongBinaryOperator reducer) {
5272	>	if (transformer == null || reducer == null)
5273	>	throw new NullPointerException();
5274	>	return new MapReduceValuesToLongTask<K,V>
5275	>	(map, null, -1, null, transformer, basis, reducer);
5276	>	}
5277	>
5278	>	/**
5279	>	* Returns a task that when invoked, returns the result of
5280	>	* accumulating the given transformation of all values using the
5281	>	* given reducer to combine values, and the given basis as an
5282	>	* identity value.
5283	>	*
5284	>	* @param map the map
5285	>	* @param transformer a function returning the transformation
5286	>	* for an element
5287	>	* @param basis the identity (initial default value) for the reduction
5288	>	* @param reducer a commutative associative combining function
5289	>	* @return the task
5290	>	*/
5291	>	public static <K,V> ForkJoinTask<Integer> reduceValuesToInt
5292	>	(ConcurrentHashMap<K,V> map,
5293	>	IntFunction<? super V> transformer,
5294	>	int basis,
5295	>	IntBinaryOperator reducer) {
5296	>	if (transformer == null || reducer == null)
5297	>	throw new NullPointerException();
5298	>	return new MapReduceValuesToIntTask<K,V>
5299	>	(map, null, -1, null, transformer, basis, reducer);
5300	>	}
5301	>
5302	>	/**
5303	>	* Returns a task that when invoked, perform the given action
5304	>	* for each entry.
5305	>	*
5306	>	* @param map the map
5307	>	* @param action the action
5308	>	*/
5309	>	public static <K,V> ForkJoinTask<Void> forEachEntry
5310	>	(ConcurrentHashMap<K,V> map,
5311	>	Block<? super Map.Entry<K,V>> action) {
5312	>	if (action == null) throw new NullPointerException();
5313	>	return new ForEachEntryTask<K,V>(map, null, -1, action);
5314	>	}
5315	>
5316	>	/**
5317	>	* Returns a task that when invoked, perform the given action
5318	>	* for each non-null transformation of each entry.
5319	>	*
5320	>	* @param map the map
5321	>	* @param transformer a function returning the transformation
5322	>	* for an element, or null if there is no transformation (in
5323	>	* which case the action is not applied)
5324	>	* @param action the action
5325	>	*/
5326	>	public static <K,V,U> ForkJoinTask<Void> forEachEntry
5327	>	(ConcurrentHashMap<K,V> map,
5328	>	Function<Map.Entry<K,V>, ? extends U> transformer,
5329	>	Block<? super U> action) {
5330	>	if (transformer == null || action == null)
5331	>	throw new NullPointerException();
5332	>	return new ForEachTransformedEntryTask<K,V,U>
5333	>	(map, null, -1, transformer, action);
5334	>	}
5335	>
5336	>	/**
5337	>	* Returns a task that when invoked, returns a non-null result
5338	>	* from applying the given search function on each entry, or
5339	>	* null if none.  Upon success, further element processing is
5340	>	* suppressed and the results of any other parallel
5341	>	* invocations of the search function are ignored.
5342	>	*
5343	>	* @param map the map
5344	>	* @param searchFunction a function returning a non-null
5345	>	* result on success, else null
5346	>	* @return the task
5347	>	*/
5348	>	public static <K,V,U> ForkJoinTask<U> searchEntries
5349	>	(ConcurrentHashMap<K,V> map,
5350	>	Function<Map.Entry<K,V>, ? extends U> searchFunction) {
5351	>	if (searchFunction == null) throw new NullPointerException();
5352	>	return new SearchEntriesTask<K,V,U>
5353	>	(map, null, -1, searchFunction,
5354	>	new AtomicReference<U>());
5355	>	}
5356	>
5357	>	/**
5358	>	* Returns a task that when invoked, returns the result of
5359	>	* accumulating all entries using the given reducer to combine
5360	>	* values, or null if none.
5361	>	*
5362	>	* @param map the map
5363	>	* @param reducer a commutative associative combining function
5364	>	* @return the task
5365	>	*/
5366	>	public static <K,V> ForkJoinTask<Map.Entry<K,V>> reduceEntries
5367	>	(ConcurrentHashMap<K,V> map,
5368	>	BiFunction<Map.Entry<K,V>, Map.Entry<K,V>, ? extends Map.Entry<K,V>> reducer) {
5369	>	if (reducer == null) throw new NullPointerException();
5370	>	return new ReduceEntriesTask<K,V>
5371	>	(map, null, -1, null, reducer);
5372	>	}
5373	>
5374	>	/**
5375	>	* Returns a task that when invoked, returns the result of
5376	>	* accumulating the given transformation of all entries using the
5377	>	* given reducer to combine values, or null if none.
5378	>	*
5379	>	* @param map the map
5380	>	* @param transformer a function returning the transformation
5381	>	* for an element, or null if there is no transformation (in
5382	>	* which case it is not combined).
5383	>	* @param reducer a commutative associative combining function
5384	>	* @return the task
5385	>	*/
5386	>	public static <K,V,U> ForkJoinTask<U> reduceEntries
5387	>	(ConcurrentHashMap<K,V> map,
5388	>	Function<Map.Entry<K,V>, ? extends U> transformer,
5389	>	BiFunction<? super U, ? super U, ? extends U> reducer) {
5390	>	if (transformer == null || reducer == null)
5391	>	throw new NullPointerException();
5392	>	return new MapReduceEntriesTask<K,V,U>
5393	>	(map, null, -1, null, transformer, reducer);
5394	>	}
5395	>
5396	>	/**
5397	>	* Returns a task that when invoked, returns the result of
5398	>	* accumulating the given transformation of all entries using the
5399	>	* given reducer to combine values, and the given basis as an
5400	>	* identity value.
5401	>	*
5402	>	* @param map the map
5403	>	* @param transformer a function returning the transformation
5404	>	* for an element
5405	>	* @param basis the identity (initial default value) for the reduction
5406	>	* @param reducer a commutative associative combining function
5407	>	* @return the task
5408	>	*/
5409	>	public static <K,V> ForkJoinTask<Double> reduceEntriesToDouble
5410	>	(ConcurrentHashMap<K,V> map,
5411	>	DoubleFunction<Map.Entry<K,V>> transformer,
5412	>	double basis,
5413	>	DoubleBinaryOperator reducer) {
5414	>	if (transformer == null || reducer == null)
5415	>	throw new NullPointerException();
5416	>	return new MapReduceEntriesToDoubleTask<K,V>
5417	>	(map, null, -1, null, transformer, basis, reducer);
5418	>	}
5419	>
5420	>	/**
5421	>	* Returns a task that when invoked, returns the result of
5422	>	* accumulating the given transformation of all entries using the
5423	>	* given reducer to combine values, and the given basis as an
5424	>	* identity value.
5425	>	*
5426	>	* @param map the map
5427	>	* @param transformer a function returning the transformation
5428	>	* for an element
5429	>	* @param basis the identity (initial default value) for the reduction
5430	>	* @param reducer a commutative associative combining function
5431	>	* @return the task
5432	>	*/
5433	>	public static <K,V> ForkJoinTask<Long> reduceEntriesToLong
5434	>	(ConcurrentHashMap<K,V> map,
5435	>	LongFunction<Map.Entry<K,V>> transformer,
5436	>	long basis,
5437	>	LongBinaryOperator reducer) {
5438	>	if (transformer == null || reducer == null)
5439	>	throw new NullPointerException();
5440	>	return new MapReduceEntriesToLongTask<K,V>
5441	>	(map, null, -1, null, transformer, basis, reducer);
5442	>	}
5443	>
5444	>	/**
5445	>	* Returns a task that when invoked, returns the result of
5446	>	* accumulating the given transformation of all entries using the
5447	>	* given reducer to combine values, and the given basis as an
5448	>	* identity value.
5449	>	*
5450	>	* @param map the map
5451	>	* @param transformer a function returning the transformation
5452	>	* for an element
5453	>	* @param basis the identity (initial default value) for the reduction
5454	>	* @param reducer a commutative associative combining function
5455	>	* @return the task
5456	>	*/
5457	>	public static <K,V> ForkJoinTask<Integer> reduceEntriesToInt
5458	>	(ConcurrentHashMap<K,V> map,
5459	>	IntFunction<Map.Entry<K,V>> transformer,
5460	>	int basis,
5461	>	IntBinaryOperator reducer) {
5462	>	if (transformer == null || reducer == null)
5463	>	throw new NullPointerException();
5464	>	return new MapReduceEntriesToIntTask<K,V>
5465	>	(map, null, -1, null, transformer, basis, reducer);
5466	>	}
5467	>	}
5468	>
5469	>	// -------------------------------------------------------
5470	>
5471	>	/*
5472	>	* Task classes. Coded in a regular but ugly format/style to
5473	>	* simplify checks that each variant differs in the right way from
5474	>	* others. The null screenings exist because compilers cannot tell
5475	>	* that we've already null-checked task arguments, so we force
5476	>	* simplest hoisted bypass to help avoid convoluted traps.
5477	>	*/
5478	>
5479	>	@SuppressWarnings("serial") static final class ForEachKeyTask<K,V>
5480	>	extends Traverser<K,V,Void> {
5481	>	final Block<? super K> action;
5482	>	ForEachKeyTask
5483	>	(ConcurrentHashMap<K,V> m, Traverser<K,V,?> p, int b,
5484	>	Block<? super K> action) {
5485	>	super(m, p, b);
5486	>	this.action = action;
5487	>	}
5488	>	@SuppressWarnings("unchecked") public final void compute() {
5489	>	final Block<? super K> action;
5490	>	if ((action = this.action) != null) {
5491	>	for (int b; (b = preSplit()) > 0;)
5492	>	new ForEachKeyTask<K,V>(map, this, b, action).fork();
5493	>	while (advance() != null)
5494	>	action.accept((K)nextKey);
5495	>	propagateCompletion();
5496	>	}
5497	>	}
5498	>	}
5499	>
5500	>	@SuppressWarnings("serial") static final class ForEachValueTask<K,V>
5501	>	extends Traverser<K,V,Void> {
5502	>	final Block<? super V> action;
5503	>	ForEachValueTask
5504	>	(ConcurrentHashMap<K,V> m, Traverser<K,V,?> p, int b,
5505	>	Block<? super V> action) {
5506	>	super(m, p, b);
5507	>	this.action = action;
5508	>	}
5509	>	@SuppressWarnings("unchecked") public final void compute() {
5510	>	final Block<? super V> action;
5511	>	if ((action = this.action) != null) {
5512	>	for (int b; (b = preSplit()) > 0;)
5513	>	new ForEachValueTask<K,V>(map, this, b, action).fork();
5514	>	V v;
5515	>	while ((v = advance()) != null)
5516	>	action.accept(v);
5517	>	propagateCompletion();
5518	>	}
5519	>	}
5520	>	}
5521	>
5522	>	@SuppressWarnings("serial") static final class ForEachEntryTask<K,V>
5523	>	extends Traverser<K,V,Void> {
5524	>	final Block<? super Entry<K,V>> action;
5525	>	ForEachEntryTask
5526	>	(ConcurrentHashMap<K,V> m, Traverser<K,V,?> p, int b,
5527	>	Block<? super Entry<K,V>> action) {
5528	>	super(m, p, b);
5529	>	this.action = action;
5530	>	}
5531	>	@SuppressWarnings("unchecked") public final void compute() {
5532	>	final Block<? super Entry<K,V>> action;
5533	>	if ((action = this.action) != null) {
5534	>	for (int b; (b = preSplit()) > 0;)
5535	>	new ForEachEntryTask<K,V>(map, this, b, action).fork();
5536	>	V v;
5537	>	while ((v = advance()) != null)
5538	>	action.accept(entryFor((K)nextKey, v));
5539	>	propagateCompletion();
5540	>	}
5541	>	}
5542	>	}
5543	>
5544	>	@SuppressWarnings("serial") static final class ForEachMappingTask<K,V>
5545	>	extends Traverser<K,V,Void> {
5546	>	final BiBlock<? super K, ? super V> action;
5547	>	ForEachMappingTask
5548	>	(ConcurrentHashMap<K,V> m, Traverser<K,V,?> p, int b,
5549	>	BiBlock<? super K,? super V> action) {
5550	>	super(m, p, b);
5551	>	this.action = action;
5552	>	}
5553	>	@SuppressWarnings("unchecked") public final void compute() {
5554	>	final BiBlock<? super K, ? super V> action;
5555	>	if ((action = this.action) != null) {
5556	>	for (int b; (b = preSplit()) > 0;)
5557	>	new ForEachMappingTask<K,V>(map, this, b, action).fork();
5558	>	V v;
5559	>	while ((v = advance()) != null)
5560	>	action.accept((K)nextKey, v);
5561	>	propagateCompletion();
5562	>	}
5563	>	}
5564	>	}
5565	>
5566	>	@SuppressWarnings("serial") static final class ForEachTransformedKeyTask<K,V,U>
5567	>	extends Traverser<K,V,Void> {
5568	>	final Function<? super K, ? extends U> transformer;
5569	>	final Block<? super U> action;
5570	>	ForEachTransformedKeyTask
5571	>	(ConcurrentHashMap<K,V> m, Traverser<K,V,?> p, int b,
5572	>	Function<? super K, ? extends U> transformer, Block<? super U> action) {
5573	>	super(m, p, b);
5574	>	this.transformer = transformer; this.action = action;
5575	>	}
5576	>	@SuppressWarnings("unchecked") public final void compute() {
5577	>	final Function<? super K, ? extends U> transformer;
5578	>	final Block<? super U> action;
5579	>	if ((transformer = this.transformer) != null &&
5580	>	(action = this.action) != null) {
5581	>	for (int b; (b = preSplit()) > 0;)
5582	>	new ForEachTransformedKeyTask<K,V,U>
5583	>	(map, this, b, transformer, action).fork();
5584	>	U u;
5585	>	while (advance() != null) {
5586	>	if ((u = transformer.apply((K)nextKey)) != null)
5587	>	action.accept(u);
5588	>	}
5589	>	propagateCompletion();
5590	>	}
5591	>	}
5592	>	}
5593	>
5594	>	@SuppressWarnings("serial") static final class ForEachTransformedValueTask<K,V,U>
5595	>	extends Traverser<K,V,Void> {
5596	>	final Function<? super V, ? extends U> transformer;
5597	>	final Block<? super U> action;
5598	>	ForEachTransformedValueTask
5599	>	(ConcurrentHashMap<K,V> m, Traverser<K,V,?> p, int b,
5600	>	Function<? super V, ? extends U> transformer, Block<? super U> action) {
5601	>	super(m, p, b);
5602	>	this.transformer = transformer; this.action = action;
5603	>	}
5604	>	@SuppressWarnings("unchecked") public final void compute() {
5605	>	final Function<? super V, ? extends U> transformer;
5606	>	final Block<? super U> action;
5607	>	if ((transformer = this.transformer) != null &&
5608	>	(action = this.action) != null) {
5609	>	for (int b; (b = preSplit()) > 0;)
5610	>	new ForEachTransformedValueTask<K,V,U>
5611	>	(map, this, b, transformer, action).fork();
5612	>	V v; U u;
5613	>	while ((v = advance()) != null) {
5614	>	if ((u = transformer.apply(v)) != null)
5615	>	action.accept(u);
5616	>	}
5617	>	propagateCompletion();
5618	>	}
5619	>	}
5620	>	}
5621	>
5622	>	@SuppressWarnings("serial") static final class ForEachTransformedEntryTask<K,V,U>
5623	>	extends Traverser<K,V,Void> {
5624	>	final Function<Map.Entry<K,V>, ? extends U> transformer;
5625	>	final Block<? super U> action;
5626	>	ForEachTransformedEntryTask
5627	>	(ConcurrentHashMap<K,V> m, Traverser<K,V,?> p, int b,
5628	>	Function<Map.Entry<K,V>, ? extends U> transformer, Block<? super U> action) {
5629	>	super(m, p, b);
5630	>	this.transformer = transformer; this.action = action;
5631	>	}
5632	>	@SuppressWarnings("unchecked") public final void compute() {
5633	>	final Function<Map.Entry<K,V>, ? extends U> transformer;
5634	>	final Block<? super U> action;
5635	>	if ((transformer = this.transformer) != null &&
5636	>	(action = this.action) != null) {
5637	>	for (int b; (b = preSplit()) > 0;)
5638	>	new ForEachTransformedEntryTask<K,V,U>
5639	>	(map, this, b, transformer, action).fork();
5640	>	V v; U u;
5641	>	while ((v = advance()) != null) {
5642	>	if ((u = transformer.apply(entryFor((K)nextKey,
5643	>	v))) != null)
5644	>	action.accept(u);
5645	>	}
5646	>	propagateCompletion();
5647	>	}
5648	>	}
5649	>	}
5650	>
5651	>	@SuppressWarnings("serial") static final class ForEachTransformedMappingTask<K,V,U>
5652	>	extends Traverser<K,V,Void> {
5653	>	final BiFunction<? super K, ? super V, ? extends U> transformer;
5654	>	final Block<? super U> action;
5655	>	ForEachTransformedMappingTask
5656	>	(ConcurrentHashMap<K,V> m, Traverser<K,V,?> p, int b,
5657	>	BiFunction<? super K, ? super V, ? extends U> transformer,
5658	>	Block<? super U> action) {
5659	>	super(m, p, b);
5660	>	this.transformer = transformer; this.action = action;
5661	>	}
5662	>	@SuppressWarnings("unchecked") public final void compute() {
5663	>	final BiFunction<? super K, ? super V, ? extends U> transformer;
5664	>	final Block<? super U> action;
5665	>	if ((transformer = this.transformer) != null &&
5666	>	(action = this.action) != null) {
5667	>	for (int b; (b = preSplit()) > 0;)
5668	>	new ForEachTransformedMappingTask<K,V,U>
5669	>	(map, this, b, transformer, action).fork();
5670	>	V v; U u;
5671	>	while ((v = advance()) != null) {
5672	>	if ((u = transformer.apply((K)nextKey, v)) != null)
5673	>	action.accept(u);
5674	>	}
5675	>	propagateCompletion();
5676	>	}
5677		}
5678		}
5679	+
5680	+	@SuppressWarnings("serial") static final class SearchKeysTask<K,V,U>
5681	+	extends Traverser<K,V,U> {
5682	+	final Function<? super K, ? extends U> searchFunction;
5683	+	final AtomicReference<U> result;
5684	+	SearchKeysTask
5685	+	(ConcurrentHashMap<K,V> m, Traverser<K,V,?> p, int b,
5686	+	Function<? super K, ? extends U> searchFunction,
5687	+	AtomicReference<U> result) {
5688	+	super(m, p, b);
5689	+	this.searchFunction = searchFunction; this.result = result;
5690	+	}
5691	+	public final U getRawResult() { return result.get(); }
5692	+	@SuppressWarnings("unchecked") public final void compute() {
5693	+	final Function<? super K, ? extends U> searchFunction;
5694	+	final AtomicReference<U> result;
5695	+	if ((searchFunction = this.searchFunction) != null &&
5696	+	(result = this.result) != null) {
5697	+	for (int b;;) {
5698	+	if (result.get() != null)
5699	+	return;
5700	+	if ((b = preSplit()) <= 0)
5701	+	break;
5702	+	new SearchKeysTask<K,V,U>
5703	+	(map, this, b, searchFunction, result).fork();
5704	+	}
5705	+	while (result.get() == null) {
5706	+	U u;
5707	+	if (advance() == null) {
5708	+	propagateCompletion();
5709	+	break;
5710	+	}
5711	+	if ((u = searchFunction.apply((K)nextKey)) != null) {
5712	+	if (result.compareAndSet(null, u))
5713	+	quietlyCompleteRoot();
5714	+	break;
5715	+	}
5716	+	}
5717	+	}
5718	+	}
5719	+	}
5720	+
5721	+	@SuppressWarnings("serial") static final class SearchValuesTask<K,V,U>
5722	+	extends Traverser<K,V,U> {
5723	+	final Function<? super V, ? extends U> searchFunction;
5724	+	final AtomicReference<U> result;
5725	+	SearchValuesTask
5726	+	(ConcurrentHashMap<K,V> m, Traverser<K,V,?> p, int b,
5727	+	Function<? super V, ? extends U> searchFunction,
5728	+	AtomicReference<U> result) {
5729	+	super(m, p, b);
5730	+	this.searchFunction = searchFunction; this.result = result;
5731	+	}
5732	+	public final U getRawResult() { return result.get(); }
5733	+	@SuppressWarnings("unchecked") public final void compute() {
5734	+	final Function<? super V, ? extends U> searchFunction;
5735	+	final AtomicReference<U> result;
5736	+	if ((searchFunction = this.searchFunction) != null &&
5737	+	(result = this.result) != null) {
5738	+	for (int b;;) {
5739	+	if (result.get() != null)
5740	+	return;
5741	+	if ((b = preSplit()) <= 0)
5742	+	break;
5743	+	new SearchValuesTask<K,V,U>
5744	+	(map, this, b, searchFunction, result).fork();
5745	+	}
5746	+	while (result.get() == null) {
5747	+	V v; U u;
5748	+	if ((v = advance()) == null) {
5749	+	propagateCompletion();
5750	+	break;
5751	+	}
5752	+	if ((u = searchFunction.apply(v)) != null) {
5753	+	if (result.compareAndSet(null, u))
5754	+	quietlyCompleteRoot();
5755	+	break;
5756	+	}
5757	+	}
5758	+	}
5759	+	}
5760	+	}
5761	+
5762	+	@SuppressWarnings("serial") static final class SearchEntriesTask<K,V,U>
5763	+	extends Traverser<K,V,U> {
5764	+	final Function<Entry<K,V>, ? extends U> searchFunction;
5765	+	final AtomicReference<U> result;
5766	+	SearchEntriesTask
5767	+	(ConcurrentHashMap<K,V> m, Traverser<K,V,?> p, int b,
5768	+	Function<Entry<K,V>, ? extends U> searchFunction,
5769	+	AtomicReference<U> result) {
5770	+	super(m, p, b);
5771	+	this.searchFunction = searchFunction; this.result = result;
5772	+	}
5773	+	public final U getRawResult() { return result.get(); }
5774	+	@SuppressWarnings("unchecked") public final void compute() {
5775	+	final Function<Entry<K,V>, ? extends U> searchFunction;
5776	+	final AtomicReference<U> result;
5777	+	if ((searchFunction = this.searchFunction) != null &&
5778	+	(result = this.result) != null) {
5779	+	for (int b;;) {
5780	+	if (result.get() != null)
5781	+	return;
5782	+	if ((b = preSplit()) <= 0)
5783	+	break;
5784	+	new SearchEntriesTask<K,V,U>
5785	+	(map, this, b, searchFunction, result).fork();
5786	+	}
5787	+	while (result.get() == null) {
5788	+	V v; U u;
5789	+	if ((v = advance()) == null) {
5790	+	propagateCompletion();
5791	+	break;
5792	+	}
5793	+	if ((u = searchFunction.apply(entryFor((K)nextKey,
5794	+	v))) != null) {
5795	+	if (result.compareAndSet(null, u))
5796	+	quietlyCompleteRoot();
5797	+	return;
5798	+	}
5799	+	}
5800	+	}
5801	+	}
5802	+	}
5803	+
5804	+	@SuppressWarnings("serial") static final class SearchMappingsTask<K,V,U>
5805	+	extends Traverser<K,V,U> {
5806	+	final BiFunction<? super K, ? super V, ? extends U> searchFunction;
5807	+	final AtomicReference<U> result;
5808	+	SearchMappingsTask
5809	+	(ConcurrentHashMap<K,V> m, Traverser<K,V,?> p, int b,
5810	+	BiFunction<? super K, ? super V, ? extends U> searchFunction,
5811	+	AtomicReference<U> result) {
5812	+	super(m, p, b);
5813	+	this.searchFunction = searchFunction; this.result = result;
5814	+	}
5815	+	public final U getRawResult() { return result.get(); }
5816	+	@SuppressWarnings("unchecked") public final void compute() {
5817	+	final BiFunction<? super K, ? super V, ? extends U> searchFunction;
5818	+	final AtomicReference<U> result;
5819	+	if ((searchFunction = this.searchFunction) != null &&
5820	+	(result = this.result) != null) {
5821	+	for (int b;;) {
5822	+	if (result.get() != null)
5823	+	return;
5824	+	if ((b = preSplit()) <= 0)
5825	+	break;
5826	+	new SearchMappingsTask<K,V,U>
5827	+	(map, this, b, searchFunction, result).fork();
5828	+	}
5829	+	while (result.get() == null) {
5830	+	V v; U u;
5831	+	if ((v = advance()) == null) {
5832	+	propagateCompletion();
5833	+	break;
5834	+	}
5835	+	if ((u = searchFunction.apply((K)nextKey, v)) != null) {
5836	+	if (result.compareAndSet(null, u))
5837	+	quietlyCompleteRoot();
5838	+	break;
5839	+	}
5840	+	}
5841	+	}
5842	+	}
5843	+	}
5844	+
5845	+	@SuppressWarnings("serial") static final class ReduceKeysTask<K,V>
5846	+	extends Traverser<K,V,K> {
5847	+	final BiFunction<? super K, ? super K, ? extends K> reducer;
5848	+	K result;
5849	+	ReduceKeysTask<K,V> rights, nextRight;
5850	+	ReduceKeysTask
5851	+	(ConcurrentHashMap<K,V> m, Traverser<K,V,?> p, int b,
5852	+	ReduceKeysTask<K,V> nextRight,
5853	+	BiFunction<? super K, ? super K, ? extends K> reducer) {
5854	+	super(m, p, b); this.nextRight = nextRight;
5855	+	this.reducer = reducer;
5856	+	}
5857	+	public final K getRawResult() { return result; }
5858	+	@SuppressWarnings("unchecked") public final void compute() {
5859	+	final BiFunction<? super K, ? super K, ? extends K> reducer;
5860	+	if ((reducer = this.reducer) != null) {
5861	+	for (int b; (b = preSplit()) > 0;)
5862	+	(rights = new ReduceKeysTask<K,V>
5863	+	(map, this, b, rights, reducer)).fork();
5864	+	K r = null;
5865	+	while (advance() != null) {
5866	+	K u = (K)nextKey;
5867	+	r = (r == null) ? u : u == null ? r : reducer.apply(r, u);
5868	+	}
5869	+	result = r;
5870	+	CountedCompleter<?> c;
5871	+	for (c = firstComplete(); c != null; c = c.nextComplete()) {
5872	+	ReduceKeysTask<K,V>
5873	+	t = (ReduceKeysTask<K,V>)c,
5874	+	s = t.rights;
5875	+	while (s != null) {
5876	+	K tr, sr;
5877	+	if ((sr = s.result) != null)
5878	+	t.result = (((tr = t.result) == null) ? sr :
5879	+	reducer.apply(tr, sr));
5880	+	s = t.rights = s.nextRight;
5881	+	}
5882	+	}
5883	+	}
5884	+	}
5885	+	}
5886	+
5887	+	@SuppressWarnings("serial") static final class ReduceValuesTask<K,V>
5888	+	extends Traverser<K,V,V> {
5889	+	final BiFunction<? super V, ? super V, ? extends V> reducer;
5890	+	V result;
5891	+	ReduceValuesTask<K,V> rights, nextRight;
5892	+	ReduceValuesTask
5893	+	(ConcurrentHashMap<K,V> m, Traverser<K,V,?> p, int b,
5894	+	ReduceValuesTask<K,V> nextRight,
5895	+	BiFunction<? super V, ? super V, ? extends V> reducer) {
5896	+	super(m, p, b); this.nextRight = nextRight;
5897	+	this.reducer = reducer;
5898	+	}
5899	+	public final V getRawResult() { return result; }
5900	+	@SuppressWarnings("unchecked") public final void compute() {
5901	+	final BiFunction<? super V, ? super V, ? extends V> reducer;
5902	+	if ((reducer = this.reducer) != null) {
5903	+	for (int b; (b = preSplit()) > 0;)
5904	+	(rights = new ReduceValuesTask<K,V>
5905	+	(map, this, b, rights, reducer)).fork();
5906	+	V r = null, v;
5907	+	while ((v = advance()) != null)
5908	+	r = (r == null) ? v : v == null ? r : reducer.apply(r, v);
5909	+	result = r;
5910	+	CountedCompleter<?> c;
5911	+	for (c = firstComplete(); c != null; c = c.nextComplete()) {
5912	+	ReduceValuesTask<K,V>
5913	+	t = (ReduceValuesTask<K,V>)c,
5914	+	s = t.rights;
5915	+	while (s != null) {
5916	+	V tr, sr;
5917	+	if ((sr = s.result) != null)
5918	+	t.result = (((tr = t.result) == null) ? sr :
5919	+	reducer.apply(tr, sr));
5920	+	s = t.rights = s.nextRight;
5921	+	}
5922	+	}
5923	+	}
5924	+	}
5925	+	}
5926	+
5927	+	@SuppressWarnings("serial") static final class ReduceEntriesTask<K,V>
5928	+	extends Traverser<K,V,Map.Entry<K,V>> {
5929	+	final BiFunction<Map.Entry<K,V>, Map.Entry<K,V>, ? extends Map.Entry<K,V>> reducer;
5930	+	Map.Entry<K,V> result;
5931	+	ReduceEntriesTask<K,V> rights, nextRight;
5932	+	ReduceEntriesTask
5933	+	(ConcurrentHashMap<K,V> m, Traverser<K,V,?> p, int b,
5934	+	ReduceEntriesTask<K,V> nextRight,
5935	+	BiFunction<Entry<K,V>, Map.Entry<K,V>, ? extends Map.Entry<K,V>> reducer) {
5936	+	super(m, p, b); this.nextRight = nextRight;
5937	+	this.reducer = reducer;
5938	+	}
5939	+	public final Map.Entry<K,V> getRawResult() { return result; }
5940	+	@SuppressWarnings("unchecked") public final void compute() {
5941	+	final BiFunction<Map.Entry<K,V>, Map.Entry<K,V>, ? extends Map.Entry<K,V>> reducer;
5942	+	if ((reducer = this.reducer) != null) {
5943	+	for (int b; (b = preSplit()) > 0;)
5944	+	(rights = new ReduceEntriesTask<K,V>
5945	+	(map, this, b, rights, reducer)).fork();
5946	+	Map.Entry<K,V> r = null;
5947	+	V v;
5948	+	while ((v = advance()) != null) {
5949	+	Map.Entry<K,V> u = entryFor((K)nextKey, v);
5950	+	r = (r == null) ? u : reducer.apply(r, u);
5951	+	}
5952	+	result = r;
5953	+	CountedCompleter<?> c;
5954	+	for (c = firstComplete(); c != null; c = c.nextComplete()) {
5955	+	ReduceEntriesTask<K,V>
5956	+	t = (ReduceEntriesTask<K,V>)c,
5957	+	s = t.rights;
5958	+	while (s != null) {
5959	+	Map.Entry<K,V> tr, sr;
5960	+	if ((sr = s.result) != null)
5961	+	t.result = (((tr = t.result) == null) ? sr :
5962	+	reducer.apply(tr, sr));
5963	+	s = t.rights = s.nextRight;
5964	+	}
5965	+	}
5966	+	}
5967	+	}
5968	+	}
5969	+
5970	+	@SuppressWarnings("serial") static final class MapReduceKeysTask<K,V,U>
5971	+	extends Traverser<K,V,U> {
5972	+	final Function<? super K, ? extends U> transformer;
5973	+	final BiFunction<? super U, ? super U, ? extends U> reducer;
5974	+	U result;
5975	+	MapReduceKeysTask<K,V,U> rights, nextRight;
5976	+	MapReduceKeysTask
5977	+	(ConcurrentHashMap<K,V> m, Traverser<K,V,?> p, int b,
5978	+	MapReduceKeysTask<K,V,U> nextRight,
5979	+	Function<? super K, ? extends U> transformer,
5980	+	BiFunction<? super U, ? super U, ? extends U> reducer) {
5981	+	super(m, p, b); this.nextRight = nextRight;
5982	+	this.transformer = transformer;
5983	+	this.reducer = reducer;
5984	+	}
5985	+	public final U getRawResult() { return result; }
5986	+	@SuppressWarnings("unchecked") public final void compute() {
5987	+	final Function<? super K, ? extends U> transformer;
5988	+	final BiFunction<? super U, ? super U, ? extends U> reducer;
5989	+	if ((transformer = this.transformer) != null &&
5990	+	(reducer = this.reducer) != null) {
5991	+	for (int b; (b = preSplit()) > 0;)
5992	+	(rights = new MapReduceKeysTask<K,V,U>
5993	+	(map, this, b, rights, transformer, reducer)).fork();
5994	+	U r = null, u;
5995	+	while (advance() != null) {
5996	+	if ((u = transformer.apply((K)nextKey)) != null)
5997	+	r = (r == null) ? u : reducer.apply(r, u);
5998	+	}
5999	+	result = r;
6000	+	CountedCompleter<?> c;
6001	+	for (c = firstComplete(); c != null; c = c.nextComplete()) {
6002	+	MapReduceKeysTask<K,V,U>
6003	+	t = (MapReduceKeysTask<K,V,U>)c,
6004	+	s = t.rights;
6005	+	while (s != null) {
6006	+	U tr, sr;
6007	+	if ((sr = s.result) != null)
6008	+	t.result = (((tr = t.result) == null) ? sr :
6009	+	reducer.apply(tr, sr));
6010	+	s = t.rights = s.nextRight;
6011	+	}
6012	+	}
6013	+	}
6014	+	}
6015	+	}
6016	+
6017	+	@SuppressWarnings("serial") static final class MapReduceValuesTask<K,V,U>
6018	+	extends Traverser<K,V,U> {
6019	+	final Function<? super V, ? extends U> transformer;
6020	+	final BiFunction<? super U, ? super U, ? extends U> reducer;
6021	+	U result;
6022	+	MapReduceValuesTask<K,V,U> rights, nextRight;
6023	+	MapReduceValuesTask
6024	+	(ConcurrentHashMap<K,V> m, Traverser<K,V,?> p, int b,
6025	+	MapReduceValuesTask<K,V,U> nextRight,
6026	+	Function<? super V, ? extends U> transformer,
6027	+	BiFunction<? super U, ? super U, ? extends U> reducer) {
6028	+	super(m, p, b); this.nextRight = nextRight;
6029	+	this.transformer = transformer;
6030	+	this.reducer = reducer;
6031	+	}
6032	+	public final U getRawResult() { return result; }
6033	+	@SuppressWarnings("unchecked") public final void compute() {
6034	+	final Function<? super V, ? extends U> transformer;
6035	+	final BiFunction<? super U, ? super U, ? extends U> reducer;
6036	+	if ((transformer = this.transformer) != null &&
6037	+	(reducer = this.reducer) != null) {
6038	+	for (int b; (b = preSplit()) > 0;)
6039	+	(rights = new MapReduceValuesTask<K,V,U>
6040	+	(map, this, b, rights, transformer, reducer)).fork();
6041	+	U r = null, u;
6042	+	V v;
6043	+	while ((v = advance()) != null) {
6044	+	if ((u = transformer.apply(v)) != null)
6045	+	r = (r == null) ? u : reducer.apply(r, u);
6046	+	}
6047	+	result = r;
6048	+	CountedCompleter<?> c;
6049	+	for (c = firstComplete(); c != null; c = c.nextComplete()) {
6050	+	MapReduceValuesTask<K,V,U>
6051	+	t = (MapReduceValuesTask<K,V,U>)c,
6052	+	s = t.rights;
6053	+	while (s != null) {
6054	+	U tr, sr;
6055	+	if ((sr = s.result) != null)
6056	+	t.result = (((tr = t.result) == null) ? sr :
6057	+	reducer.apply(tr, sr));
6058	+	s = t.rights = s.nextRight;
6059	+	}
6060	+	}
6061	+	}
6062	+	}
6063	+	}
6064	+
6065	+	@SuppressWarnings("serial") static final class MapReduceEntriesTask<K,V,U>
6066	+	extends Traverser<K,V,U> {
6067	+	final Function<Map.Entry<K,V>, ? extends U> transformer;
6068	+	final BiFunction<? super U, ? super U, ? extends U> reducer;
6069	+	U result;
6070	+	MapReduceEntriesTask<K,V,U> rights, nextRight;
6071	+	MapReduceEntriesTask
6072	+	(ConcurrentHashMap<K,V> m, Traverser<K,V,?> p, int b,
6073	+	MapReduceEntriesTask<K,V,U> nextRight,
6074	+	Function<Map.Entry<K,V>, ? extends U> transformer,
6075	+	BiFunction<? super U, ? super U, ? extends U> reducer) {
6076	+	super(m, p, b); this.nextRight = nextRight;
6077	+	this.transformer = transformer;
6078	+	this.reducer = reducer;
6079	+	}
6080	+	public final U getRawResult() { return result; }
6081	+	@SuppressWarnings("unchecked") public final void compute() {
6082	+	final Function<Map.Entry<K,V>, ? extends U> transformer;
6083	+	final BiFunction<? super U, ? super U, ? extends U> reducer;
6084	+	if ((transformer = this.transformer) != null &&
6085	+	(reducer = this.reducer) != null) {
6086	+	for (int b; (b = preSplit()) > 0;)
6087	+	(rights = new MapReduceEntriesTask<K,V,U>
6088	+	(map, this, b, rights, transformer, reducer)).fork();
6089	+	U r = null, u;
6090	+	V v;
6091	+	while ((v = advance()) != null) {
6092	+	if ((u = transformer.apply(entryFor((K)nextKey,
6093	+	v))) != null)
6094	+	r = (r == null) ? u : reducer.apply(r, u);
6095	+	}
6096	+	result = r;
6097	+	CountedCompleter<?> c;
6098	+	for (c = firstComplete(); c != null; c = c.nextComplete()) {
6099	+	MapReduceEntriesTask<K,V,U>
6100	+	t = (MapReduceEntriesTask<K,V,U>)c,
6101	+	s = t.rights;
6102	+	while (s != null) {
6103	+	U tr, sr;
6104	+	if ((sr = s.result) != null)
6105	+	t.result = (((tr = t.result) == null) ? sr :
6106	+	reducer.apply(tr, sr));
6107	+	s = t.rights = s.nextRight;
6108	+	}
6109	+	}
6110	+	}
6111	+	}
6112	+	}
6113	+
6114	+	@SuppressWarnings("serial") static final class MapReduceMappingsTask<K,V,U>
6115	+	extends Traverser<K,V,U> {
6116	+	final BiFunction<? super K, ? super V, ? extends U> transformer;
6117	+	final BiFunction<? super U, ? super U, ? extends U> reducer;
6118	+	U result;
6119	+	MapReduceMappingsTask<K,V,U> rights, nextRight;
6120	+	MapReduceMappingsTask
6121	+	(ConcurrentHashMap<K,V> m, Traverser<K,V,?> p, int b,
6122	+	MapReduceMappingsTask<K,V,U> nextRight,
6123	+	BiFunction<? super K, ? super V, ? extends U> transformer,
6124	+	BiFunction<? super U, ? super U, ? extends U> reducer) {
6125	+	super(m, p, b); this.nextRight = nextRight;
6126	+	this.transformer = transformer;
6127	+	this.reducer = reducer;
6128	+	}
6129	+	public final U getRawResult() { return result; }
6130	+	@SuppressWarnings("unchecked") public final void compute() {
6131	+	final BiFunction<? super K, ? super V, ? extends U> transformer;
6132	+	final BiFunction<? super U, ? super U, ? extends U> reducer;
6133	+	if ((transformer = this.transformer) != null &&
6134	+	(reducer = this.reducer) != null) {
6135	+	for (int b; (b = preSplit()) > 0;)
6136	+	(rights = new MapReduceMappingsTask<K,V,U>
6137	+	(map, this, b, rights, transformer, reducer)).fork();
6138	+	U r = null, u;
6139	+	V v;
6140	+	while ((v = advance()) != null) {
6141	+	if ((u = transformer.apply((K)nextKey, v)) != null)
6142	+	r = (r == null) ? u : reducer.apply(r, u);
6143	+	}
6144	+	result = r;
6145	+	CountedCompleter<?> c;
6146	+	for (c = firstComplete(); c != null; c = c.nextComplete()) {
6147	+	MapReduceMappingsTask<K,V,U>
6148	+	t = (MapReduceMappingsTask<K,V,U>)c,
6149	+	s = t.rights;
6150	+	while (s != null) {
6151	+	U tr, sr;
6152	+	if ((sr = s.result) != null)
6153	+	t.result = (((tr = t.result) == null) ? sr :
6154	+	reducer.apply(tr, sr));
6155	+	s = t.rights = s.nextRight;
6156	+	}
6157	+	}
6158	+	}
6159	+	}
6160	+	}
6161	+
6162	+	@SuppressWarnings("serial") static final class MapReduceKeysToDoubleTask<K,V>
6163	+	extends Traverser<K,V,Double> {
6164	+	final DoubleFunction<? super K> transformer;
6165	+	final DoubleBinaryOperator reducer;
6166	+	final double basis;
6167	+	double result;
6168	+	MapReduceKeysToDoubleTask<K,V> rights, nextRight;
6169	+	MapReduceKeysToDoubleTask
6170	+	(ConcurrentHashMap<K,V> m, Traverser<K,V,?> p, int b,
6171	+	MapReduceKeysToDoubleTask<K,V> nextRight,
6172	+	DoubleFunction<? super K> transformer,
6173	+	double basis,
6174	+	DoubleBinaryOperator reducer) {
6175	+	super(m, p, b); this.nextRight = nextRight;
6176	+	this.transformer = transformer;
6177	+	this.basis = basis; this.reducer = reducer;
6178	+	}
6179	+	public final Double getRawResult() { return result; }
6180	+	@SuppressWarnings("unchecked") public final void compute() {
6181	+	final DoubleFunction<? super K> transformer;
6182	+	final DoubleBinaryOperator reducer;
6183	+	if ((transformer = this.transformer) != null &&
6184	+	(reducer = this.reducer) != null) {
6185	+	double r = this.basis;
6186	+	for (int b; (b = preSplit()) > 0;)
6187	+	(rights = new MapReduceKeysToDoubleTask<K,V>
6188	+	(map, this, b, rights, transformer, r, reducer)).fork();
6189	+	while (advance() != null)
6190	+	r = reducer.applyAsDouble(r, transformer.applyAsDouble((K)nextKey));
6191	+	result = r;
6192	+	CountedCompleter<?> c;
6193	+	for (c = firstComplete(); c != null; c = c.nextComplete()) {
6194	+	MapReduceKeysToDoubleTask<K,V>
6195	+	t = (MapReduceKeysToDoubleTask<K,V>)c,
6196	+	s = t.rights;
6197	+	while (s != null) {
6198	+	t.result = reducer.applyAsDouble(t.result, s.result);
6199	+	s = t.rights = s.nextRight;
6200	+	}
6201	+	}
6202	+	}
6203	+	}
6204	+	}
6205	+
6206	+	@SuppressWarnings("serial") static final class MapReduceValuesToDoubleTask<K,V>
6207	+	extends Traverser<K,V,Double> {
6208	+	final DoubleFunction<? super V> transformer;
6209	+	final DoubleBinaryOperator reducer;
6210	+	final double basis;
6211	+	double result;
6212	+	MapReduceValuesToDoubleTask<K,V> rights, nextRight;
6213	+	MapReduceValuesToDoubleTask
6214	+	(ConcurrentHashMap<K,V> m, Traverser<K,V,?> p, int b,
6215	+	MapReduceValuesToDoubleTask<K,V> nextRight,
6216	+	DoubleFunction<? super V> transformer,
6217	+	double basis,
6218	+	DoubleBinaryOperator reducer) {
6219	+	super(m, p, b); this.nextRight = nextRight;
6220	+	this.transformer = transformer;
6221	+	this.basis = basis; this.reducer = reducer;
6222	+	}
6223	+	public final Double getRawResult() { return result; }
6224	+	@SuppressWarnings("unchecked") public final void compute() {
6225	+	final DoubleFunction<? super V> transformer;
6226	+	final DoubleBinaryOperator reducer;
6227	+	if ((transformer = this.transformer) != null &&
6228	+	(reducer = this.reducer) != null) {
6229	+	double r = this.basis;
6230	+	for (int b; (b = preSplit()) > 0;)
6231	+	(rights = new MapReduceValuesToDoubleTask<K,V>
6232	+	(map, this, b, rights, transformer, r, reducer)).fork();
6233	+	V v;
6234	+	while ((v = advance()) != null)
6235	+	r = reducer.applyAsDouble(r, transformer.applyAsDouble(v));
6236	+	result = r;
6237	+	CountedCompleter<?> c;
6238	+	for (c = firstComplete(); c != null; c = c.nextComplete()) {
6239	+	MapReduceValuesToDoubleTask<K,V>
6240	+	t = (MapReduceValuesToDoubleTask<K,V>)c,
6241	+	s = t.rights;
6242	+	while (s != null) {
6243	+	t.result = reducer.applyAsDouble(t.result, s.result);
6244	+	s = t.rights = s.nextRight;
6245	+	}
6246	+	}
6247	+	}
6248	+	}
6249	+	}
6250	+
6251	+	@SuppressWarnings("serial") static final class MapReduceEntriesToDoubleTask<K,V>
6252	+	extends Traverser<K,V,Double> {
6253	+	final DoubleFunction<Map.Entry<K,V>> transformer;
6254	+	final DoubleBinaryOperator reducer;
6255	+	final double basis;
6256	+	double result;
6257	+	MapReduceEntriesToDoubleTask<K,V> rights, nextRight;
6258	+	MapReduceEntriesToDoubleTask
6259	+	(ConcurrentHashMap<K,V> m, Traverser<K,V,?> p, int b,
6260	+	MapReduceEntriesToDoubleTask<K,V> nextRight,
6261	+	DoubleFunction<Map.Entry<K,V>> transformer,
6262	+	double basis,
6263	+	DoubleBinaryOperator reducer) {
6264	+	super(m, p, b); this.nextRight = nextRight;
6265	+	this.transformer = transformer;
6266	+	this.basis = basis; this.reducer = reducer;
6267	+	}
6268	+	public final Double getRawResult() { return result; }
6269	+	@SuppressWarnings("unchecked") public final void compute() {
6270	+	final DoubleFunction<Map.Entry<K,V>> transformer;
6271	+	final DoubleBinaryOperator reducer;
6272	+	if ((transformer = this.transformer) != null &&
6273	+	(reducer = this.reducer) != null) {
6274	+	double r = this.basis;
6275	+	for (int b; (b = preSplit()) > 0;)
6276	+	(rights = new MapReduceEntriesToDoubleTask<K,V>
6277	+	(map, this, b, rights, transformer, r, reducer)).fork();
6278	+	V v;
6279	+	while ((v = advance()) != null)
6280	+	r = reducer.applyAsDouble(r, transformer.applyAsDouble(entryFor((K)nextKey,
6281	+	v)));
6282	+	result = r;
6283	+	CountedCompleter<?> c;
6284	+	for (c = firstComplete(); c != null; c = c.nextComplete()) {
6285	+	MapReduceEntriesToDoubleTask<K,V>
6286	+	t = (MapReduceEntriesToDoubleTask<K,V>)c,
6287	+	s = t.rights;
6288	+	while (s != null) {
6289	+	t.result = reducer.applyAsDouble(t.result, s.result);
6290	+	s = t.rights = s.nextRight;
6291	+	}
6292	+	}
6293	+	}
6294	+	}
6295	+	}
6296	+
6297	+	@SuppressWarnings("serial") static final class MapReduceMappingsToDoubleTask<K,V>
6298	+	extends Traverser<K,V,Double> {
6299	+	final DoubleBiFunction<? super K, ? super V> transformer;
6300	+	final DoubleBinaryOperator reducer;
6301	+	final double basis;
6302	+	double result;
6303	+	MapReduceMappingsToDoubleTask<K,V> rights, nextRight;
6304	+	MapReduceMappingsToDoubleTask
6305	+	(ConcurrentHashMap<K,V> m, Traverser<K,V,?> p, int b,
6306	+	MapReduceMappingsToDoubleTask<K,V> nextRight,
6307	+	DoubleBiFunction<? super K, ? super V> transformer,
6308	+	double basis,
6309	+	DoubleBinaryOperator reducer) {
6310	+	super(m, p, b); this.nextRight = nextRight;
6311	+	this.transformer = transformer;
6312	+	this.basis = basis; this.reducer = reducer;
6313	+	}
6314	+	public final Double getRawResult() { return result; }
6315	+	@SuppressWarnings("unchecked") public final void compute() {
6316	+	final DoubleBiFunction<? super K, ? super V> transformer;
6317	+	final DoubleBinaryOperator reducer;
6318	+	if ((transformer = this.transformer) != null &&
6319	+	(reducer = this.reducer) != null) {
6320	+	double r = this.basis;
6321	+	for (int b; (b = preSplit()) > 0;)
6322	+	(rights = new MapReduceMappingsToDoubleTask<K,V>
6323	+	(map, this, b, rights, transformer, r, reducer)).fork();
6324	+	V v;
6325	+	while ((v = advance()) != null)
6326	+	r = reducer.applyAsDouble(r, transformer.applyAsDouble((K)nextKey, v));
6327	+	result = r;
6328	+	CountedCompleter<?> c;
6329	+	for (c = firstComplete(); c != null; c = c.nextComplete()) {
6330	+	MapReduceMappingsToDoubleTask<K,V>
6331	+	t = (MapReduceMappingsToDoubleTask<K,V>)c,
6332	+	s = t.rights;
6333	+	while (s != null) {
6334	+	t.result = reducer.applyAsDouble(t.result, s.result);
6335	+	s = t.rights = s.nextRight;
6336	+	}
6337	+	}
6338	+	}
6339	+	}
6340	+	}
6341	+
6342	+	@SuppressWarnings("serial") static final class MapReduceKeysToLongTask<K,V>
6343	+	extends Traverser<K,V,Long> {
6344	+	final LongFunction<? super K> transformer;
6345	+	final LongBinaryOperator reducer;
6346	+	final long basis;
6347	+	long result;
6348	+	MapReduceKeysToLongTask<K,V> rights, nextRight;
6349	+	MapReduceKeysToLongTask
6350	+	(ConcurrentHashMap<K,V> m, Traverser<K,V,?> p, int b,
6351	+	MapReduceKeysToLongTask<K,V> nextRight,
6352	+	LongFunction<? super K> transformer,
6353	+	long basis,
6354	+	LongBinaryOperator reducer) {
6355	+	super(m, p, b); this.nextRight = nextRight;
6356	+	this.transformer = transformer;
6357	+	this.basis = basis; this.reducer = reducer;
6358	+	}
6359	+	public final Long getRawResult() { return result; }
6360	+	@SuppressWarnings("unchecked") public final void compute() {
6361	+	final LongFunction<? super K> transformer;
6362	+	final LongBinaryOperator reducer;
6363	+	if ((transformer = this.transformer) != null &&
6364	+	(reducer = this.reducer) != null) {
6365	+	long r = this.basis;
6366	+	for (int b; (b = preSplit()) > 0;)
6367	+	(rights = new MapReduceKeysToLongTask<K,V>
6368	+	(map, this, b, rights, transformer, r, reducer)).fork();
6369	+	while (advance() != null)
6370	+	r = reducer.applyAsLong(r, transformer.applyAsLong((K)nextKey));
6371	+	result = r;
6372	+	CountedCompleter<?> c;
6373	+	for (c = firstComplete(); c != null; c = c.nextComplete()) {
6374	+	MapReduceKeysToLongTask<K,V>
6375	+	t = (MapReduceKeysToLongTask<K,V>)c,
6376	+	s = t.rights;
6377	+	while (s != null) {
6378	+	t.result = reducer.applyAsLong(t.result, s.result);
6379	+	s = t.rights = s.nextRight;
6380	+	}
6381	+	}
6382	+	}
6383	+	}
6384	+	}
6385	+
6386	+	@SuppressWarnings("serial") static final class MapReduceValuesToLongTask<K,V>
6387	+	extends Traverser<K,V,Long> {
6388	+	final LongFunction<? super V> transformer;
6389	+	final LongBinaryOperator reducer;
6390	+	final long basis;
6391	+	long result;
6392	+	MapReduceValuesToLongTask<K,V> rights, nextRight;
6393	+	MapReduceValuesToLongTask
6394	+	(ConcurrentHashMap<K,V> m, Traverser<K,V,?> p, int b,
6395	+	MapReduceValuesToLongTask<K,V> nextRight,
6396	+	LongFunction<? super V> transformer,
6397	+	long basis,
6398	+	LongBinaryOperator reducer) {
6399	+	super(m, p, b); this.nextRight = nextRight;
6400	+	this.transformer = transformer;
6401	+	this.basis = basis; this.reducer = reducer;
6402	+	}
6403	+	public final Long getRawResult() { return result; }
6404	+	@SuppressWarnings("unchecked") public final void compute() {
6405	+	final LongFunction<? super V> transformer;
6406	+	final LongBinaryOperator reducer;
6407	+	if ((transformer = this.transformer) != null &&
6408	+	(reducer = this.reducer) != null) {
6409	+	long r = this.basis;
6410	+	for (int b; (b = preSplit()) > 0;)
6411	+	(rights = new MapReduceValuesToLongTask<K,V>
6412	+	(map, this, b, rights, transformer, r, reducer)).fork();
6413	+	V v;
6414	+	while ((v = advance()) != null)
6415	+	r = reducer.applyAsLong(r, transformer.applyAsLong(v));
6416	+	result = r;
6417	+	CountedCompleter<?> c;
6418	+	for (c = firstComplete(); c != null; c = c.nextComplete()) {
6419	+	MapReduceValuesToLongTask<K,V>
6420	+	t = (MapReduceValuesToLongTask<K,V>)c,
6421	+	s = t.rights;
6422	+	while (s != null) {
6423	+	t.result = reducer.applyAsLong(t.result, s.result);
6424	+	s = t.rights = s.nextRight;
6425	+	}
6426	+	}
6427	+	}
6428	+	}
6429	+	}
6430	+
6431	+	@SuppressWarnings("serial") static final class MapReduceEntriesToLongTask<K,V>
6432	+	extends Traverser<K,V,Long> {
6433	+	final LongFunction<Map.Entry<K,V>> transformer;
6434	+	final LongBinaryOperator reducer;
6435	+	final long basis;
6436	+	long result;
6437	+	MapReduceEntriesToLongTask<K,V> rights, nextRight;
6438	+	MapReduceEntriesToLongTask
6439	+	(ConcurrentHashMap<K,V> m, Traverser<K,V,?> p, int b,
6440	+	MapReduceEntriesToLongTask<K,V> nextRight,
6441	+	LongFunction<Map.Entry<K,V>> transformer,
6442	+	long basis,
6443	+	LongBinaryOperator reducer) {
6444	+	super(m, p, b); this.nextRight = nextRight;
6445	+	this.transformer = transformer;
6446	+	this.basis = basis; this.reducer = reducer;
6447	+	}
6448	+	public final Long getRawResult() { return result; }
6449	+	@SuppressWarnings("unchecked") public final void compute() {
6450	+	final LongFunction<Map.Entry<K,V>> transformer;
6451	+	final LongBinaryOperator reducer;
6452	+	if ((transformer = this.transformer) != null &&
6453	+	(reducer = this.reducer) != null) {
6454	+	long r = this.basis;
6455	+	for (int b; (b = preSplit()) > 0;)
6456	+	(rights = new MapReduceEntriesToLongTask<K,V>
6457	+	(map, this, b, rights, transformer, r, reducer)).fork();
6458	+	V v;
6459	+	while ((v = advance()) != null)
6460	+	r = reducer.applyAsLong(r, transformer.applyAsLong(entryFor((K)nextKey, v)));
6461	+	result = r;
6462	+	CountedCompleter<?> c;
6463	+	for (c = firstComplete(); c != null; c = c.nextComplete()) {
6464	+	MapReduceEntriesToLongTask<K,V>
6465	+	t = (MapReduceEntriesToLongTask<K,V>)c,
6466	+	s = t.rights;
6467	+	while (s != null) {
6468	+	t.result = reducer.applyAsLong(t.result, s.result);
6469	+	s = t.rights = s.nextRight;
6470	+	}
6471	+	}
6472	+	}
6473	+	}
6474	+	}
6475	+
6476	+	@SuppressWarnings("serial") static final class MapReduceMappingsToLongTask<K,V>
6477	+	extends Traverser<K,V,Long> {
6478	+	final LongBiFunction<? super K, ? super V> transformer;
6479	+	final LongBinaryOperator reducer;
6480	+	final long basis;
6481	+	long result;
6482	+	MapReduceMappingsToLongTask<K,V> rights, nextRight;
6483	+	MapReduceMappingsToLongTask
6484	+	(ConcurrentHashMap<K,V> m, Traverser<K,V,?> p, int b,
6485	+	MapReduceMappingsToLongTask<K,V> nextRight,
6486	+	LongBiFunction<? super K, ? super V> transformer,
6487	+	long basis,
6488	+	LongBinaryOperator reducer) {
6489	+	super(m, p, b); this.nextRight = nextRight;
6490	+	this.transformer = transformer;
6491	+	this.basis = basis; this.reducer = reducer;
6492	+	}
6493	+	public final Long getRawResult() { return result; }
6494	+	@SuppressWarnings("unchecked") public final void compute() {
6495	+	final LongBiFunction<? super K, ? super V> transformer;
6496	+	final LongBinaryOperator reducer;
6497	+	if ((transformer = this.transformer) != null &&
6498	+	(reducer = this.reducer) != null) {
6499	+	long r = this.basis;
6500	+	for (int b; (b = preSplit()) > 0;)
6501	+	(rights = new MapReduceMappingsToLongTask<K,V>
6502	+	(map, this, b, rights, transformer, r, reducer)).fork();
6503	+	V v;
6504	+	while ((v = advance()) != null)
6505	+	r = reducer.applyAsLong(r, transformer.applyAsLong((K)nextKey, v));
6506	+	result = r;
6507	+	CountedCompleter<?> c;
6508	+	for (c = firstComplete(); c != null; c = c.nextComplete()) {
6509	+	MapReduceMappingsToLongTask<K,V>
6510	+	t = (MapReduceMappingsToLongTask<K,V>)c,
6511	+	s = t.rights;
6512	+	while (s != null) {
6513	+	t.result = reducer.applyAsLong(t.result, s.result);
6514	+	s = t.rights = s.nextRight;
6515	+	}
6516	+	}
6517	+	}
6518	+	}
6519	+	}
6520	+
6521	+	@SuppressWarnings("serial") static final class MapReduceKeysToIntTask<K,V>
6522	+	extends Traverser<K,V,Integer> {
6523	+	final IntFunction<? super K> transformer;
6524	+	final IntBinaryOperator reducer;
6525	+	final int basis;
6526	+	int result;
6527	+	MapReduceKeysToIntTask<K,V> rights, nextRight;
6528	+	MapReduceKeysToIntTask
6529	+	(ConcurrentHashMap<K,V> m, Traverser<K,V,?> p, int b,
6530	+	MapReduceKeysToIntTask<K,V> nextRight,
6531	+	IntFunction<? super K> transformer,
6532	+	int basis,
6533	+	IntBinaryOperator reducer) {
6534	+	super(m, p, b); this.nextRight = nextRight;
6535	+	this.transformer = transformer;
6536	+	this.basis = basis; this.reducer = reducer;
6537	+	}
6538	+	public final Integer getRawResult() { return result; }
6539	+	@SuppressWarnings("unchecked") public final void compute() {
6540	+	final IntFunction<? super K> transformer;
6541	+	final IntBinaryOperator reducer;
6542	+	if ((transformer = this.transformer) != null &&
6543	+	(reducer = this.reducer) != null) {
6544	+	int r = this.basis;
6545	+	for (int b; (b = preSplit()) > 0;)
6546	+	(rights = new MapReduceKeysToIntTask<K,V>
6547	+	(map, this, b, rights, transformer, r, reducer)).fork();
6548	+	while (advance() != null)
6549	+	r = reducer.applyAsInt(r, transformer.applyAsInt((K)nextKey));
6550	+	result = r;
6551	+	CountedCompleter<?> c;
6552	+	for (c = firstComplete(); c != null; c = c.nextComplete()) {
6553	+	MapReduceKeysToIntTask<K,V>
6554	+	t = (MapReduceKeysToIntTask<K,V>)c,
6555	+	s = t.rights;
6556	+	while (s != null) {
6557	+	t.result = reducer.applyAsInt(t.result, s.result);
6558	+	s = t.rights = s.nextRight;
6559	+	}
6560	+	}
6561	+	}
6562	+	}
6563	+	}
6564	+
6565	+	@SuppressWarnings("serial") static final class MapReduceValuesToIntTask<K,V>
6566	+	extends Traverser<K,V,Integer> {
6567	+	final IntFunction<? super V> transformer;
6568	+	final IntBinaryOperator reducer;
6569	+	final int basis;
6570	+	int result;
6571	+	MapReduceValuesToIntTask<K,V> rights, nextRight;
6572	+	MapReduceValuesToIntTask
6573	+	(ConcurrentHashMap<K,V> m, Traverser<K,V,?> p, int b,
6574	+	MapReduceValuesToIntTask<K,V> nextRight,
6575	+	IntFunction<? super V> transformer,
6576	+	int basis,
6577	+	IntBinaryOperator reducer) {
6578	+	super(m, p, b); this.nextRight = nextRight;
6579	+	this.transformer = transformer;
6580	+	this.basis = basis; this.reducer = reducer;
6581	+	}
6582	+	public final Integer getRawResult() { return result; }
6583	+	@SuppressWarnings("unchecked") public final void compute() {
6584	+	final IntFunction<? super V> transformer;
6585	+	final IntBinaryOperator reducer;
6586	+	if ((transformer = this.transformer) != null &&
6587	+	(reducer = this.reducer) != null) {
6588	+	int r = this.basis;
6589	+	for (int b; (b = preSplit()) > 0;)
6590	+	(rights = new MapReduceValuesToIntTask<K,V>
6591	+	(map, this, b, rights, transformer, r, reducer)).fork();
6592	+	V v;
6593	+	while ((v = advance()) != null)
6594	+	r = reducer.applyAsInt(r, transformer.applyAsInt(v));
6595	+	result = r;
6596	+	CountedCompleter<?> c;
6597	+	for (c = firstComplete(); c != null; c = c.nextComplete()) {
6598	+	MapReduceValuesToIntTask<K,V>
6599	+	t = (MapReduceValuesToIntTask<K,V>)c,
6600	+	s = t.rights;
6601	+	while (s != null) {
6602	+	t.result = reducer.applyAsInt(t.result, s.result);
6603	+	s = t.rights = s.nextRight;
6604	+	}
6605	+	}
6606	+	}
6607	+	}
6608	+	}
6609	+
6610	+	@SuppressWarnings("serial") static final class MapReduceEntriesToIntTask<K,V>
6611	+	extends Traverser<K,V,Integer> {
6612	+	final IntFunction<Map.Entry<K,V>> transformer;
6613	+	final IntBinaryOperator reducer;
6614	+	final int basis;
6615	+	int result;
6616	+	MapReduceEntriesToIntTask<K,V> rights, nextRight;
6617	+	MapReduceEntriesToIntTask
6618	+	(ConcurrentHashMap<K,V> m, Traverser<K,V,?> p, int b,
6619	+	MapReduceEntriesToIntTask<K,V> nextRight,
6620	+	IntFunction<Map.Entry<K,V>> transformer,
6621	+	int basis,
6622	+	IntBinaryOperator reducer) {
6623	+	super(m, p, b); this.nextRight = nextRight;
6624	+	this.transformer = transformer;
6625	+	this.basis = basis; this.reducer = reducer;
6626	+	}
6627	+	public final Integer getRawResult() { return result; }
6628	+	@SuppressWarnings("unchecked") public final void compute() {
6629	+	final IntFunction<Map.Entry<K,V>> transformer;
6630	+	final IntBinaryOperator reducer;
6631	+	if ((transformer = this.transformer) != null &&
6632	+	(reducer = this.reducer) != null) {
6633	+	int r = this.basis;
6634	+	for (int b; (b = preSplit()) > 0;)
6635	+	(rights = new MapReduceEntriesToIntTask<K,V>
6636	+	(map, this, b, rights, transformer, r, reducer)).fork();
6637	+	V v;
6638	+	while ((v = advance()) != null)
6639	+	r = reducer.applyAsInt(r, transformer.applyAsInt(entryFor((K)nextKey,
6640	+	v)));
6641	+	result = r;
6642	+	CountedCompleter<?> c;
6643	+	for (c = firstComplete(); c != null; c = c.nextComplete()) {
6644	+	MapReduceEntriesToIntTask<K,V>
6645	+	t = (MapReduceEntriesToIntTask<K,V>)c,
6646	+	s = t.rights;
6647	+	while (s != null) {
6648	+	t.result = reducer.applyAsInt(t.result, s.result);
6649	+	s = t.rights = s.nextRight;
6650	+	}
6651	+	}
6652	+	}
6653	+	}
6654	+	}
6655	+
6656	+	@SuppressWarnings("serial") static final class MapReduceMappingsToIntTask<K,V>
6657	+	extends Traverser<K,V,Integer> {
6658	+	final IntBiFunction<? super K, ? super V> transformer;
6659	+	final IntBinaryOperator reducer;
6660	+	final int basis;
6661	+	int result;
6662	+	MapReduceMappingsToIntTask<K,V> rights, nextRight;
6663	+	MapReduceMappingsToIntTask
6664	+	(ConcurrentHashMap<K,V> m, Traverser<K,V,?> p, int b,
6665	+	MapReduceMappingsToIntTask<K,V> nextRight,
6666	+	IntBiFunction<? super K, ? super V> transformer,
6667	+	int basis,
6668	+	IntBinaryOperator reducer) {
6669	+	super(m, p, b); this.nextRight = nextRight;
6670	+	this.transformer = transformer;
6671	+	this.basis = basis; this.reducer = reducer;
6672	+	}
6673	+	public final Integer getRawResult() { return result; }
6674	+	@SuppressWarnings("unchecked") public final void compute() {
6675	+	final IntBiFunction<? super K, ? super V> transformer;
6676	+	final IntBinaryOperator reducer;
6677	+	if ((transformer = this.transformer) != null &&
6678	+	(reducer = this.reducer) != null) {
6679	+	int r = this.basis;
6680	+	for (int b; (b = preSplit()) > 0;)
6681	+	(rights = new MapReduceMappingsToIntTask<K,V>
6682	+	(map, this, b, rights, transformer, r, reducer)).fork();
6683	+	V v;
6684	+	while ((v = advance()) != null)
6685	+	r = reducer.applyAsInt(r, transformer.applyAsInt((K)nextKey, v));
6686	+	result = r;
6687	+	CountedCompleter<?> c;
6688	+	for (c = firstComplete(); c != null; c = c.nextComplete()) {
6689	+	MapReduceMappingsToIntTask<K,V>
6690	+	t = (MapReduceMappingsToIntTask<K,V>)c,
6691	+	s = t.rights;
6692	+	while (s != null) {
6693	+	t.result = reducer.applyAsInt(t.result, s.result);
6694	+	s = t.rights = s.nextRight;
6695	+	}
6696	+	}
6697	+	}
6698	+	}
6699	+	}
6700	+
6701	+	// Unsafe mechanics
6702	+	private static final sun.misc.Unsafe U;
6703	+	private static final long SIZECTL;
6704	+	private static final long TRANSFERINDEX;
6705	+	private static final long TRANSFERORIGIN;
6706	+	private static final long BASECOUNT;
6707	+	private static final long CELLSBUSY;
6708	+	private static final long CELLVALUE;
6709	+	private static final long ABASE;
6710	+	private static final int ASHIFT;
6711	+
6712	+	static {
6713	+	int ss;
6714	+	try {
6715	+	U = sun.misc.Unsafe.getUnsafe();
6716	+	Class<?> k = ConcurrentHashMap.class;
6717	+	SIZECTL = U.objectFieldOffset
6718	+	(k.getDeclaredField("sizeCtl"));
6719	+	TRANSFERINDEX = U.objectFieldOffset
6720	+	(k.getDeclaredField("transferIndex"));
6721	+	TRANSFERORIGIN = U.objectFieldOffset
6722	+	(k.getDeclaredField("transferOrigin"));
6723	+	BASECOUNT = U.objectFieldOffset
6724	+	(k.getDeclaredField("baseCount"));
6725	+	CELLSBUSY = U.objectFieldOffset
6726	+	(k.getDeclaredField("cellsBusy"));
6727	+	Class<?> ck = Cell.class;
6728	+	CELLVALUE = U.objectFieldOffset
6729	+	(ck.getDeclaredField("value"));
6730	+	Class<?> sc = Node[].class;
6731	+	ABASE = U.arrayBaseOffset(sc);
6732	+	ss = U.arrayIndexScale(sc);
6733	+	ASHIFT = 31 - Integer.numberOfLeadingZeros(ss);
6734	+	} catch (Exception e) {
6735	+	throw new Error(e);
6736	+	}
6737	+	if ((ss & (ss-1)) != 0)
6738	+	throw new Error("data type scale not a power of two");
6739	+	}
6740	+
6741		}

Diff Legend

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