ViewVC Help
View File | Revision Log | Show Annotations | Download File | Root Listing
root/jsr166/jsr166/src/main/java/util/concurrent/ConcurrentSkipListMap.java
Revision: 1.91
Committed: Wed Jan 23 18:40:30 2013 UTC (11 years, 4 months ago) by dl
Branch: MAIN
Changes since 1.90: +2 -1 lines
Log Message:
Temporarily isolate from TLR updates

File Contents

# User Rev Content
1 dl 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 jsr166 1.67 * http://creativecommons.org/publicdomain/zero/1.0/
5 dl 1.1 */
6    
7     package java.util.concurrent;
8     import java.util.*;
9 dl 1.83 import java.util.stream.Stream;
10     import java.util.Spliterator;
11     import java.util.stream.Streams;
12     import java.util.function.Block;
13    
14 dl 1.1
15     /**
16 jsr166 1.22 * A scalable concurrent {@link ConcurrentNavigableMap} implementation.
17     * The map is sorted according to the {@linkplain Comparable natural
18     * ordering} of its keys, or by a {@link Comparator} provided at map
19     * creation time, depending on which constructor is used.
20 dl 1.1 *
21     * <p>This class implements a concurrent variant of <a
22 dl 1.66 * href="http://en.wikipedia.org/wiki/Skip_list" target="_top">SkipLists</a>
23     * providing expected average <i>log(n)</i> time cost for the
24 jsr166 1.82 * {@code containsKey}, {@code get}, {@code put} and
25     * {@code remove} operations and their variants. Insertion, removal,
26 dl 1.1 * update, and access operations safely execute concurrently by
27 jsr166 1.51 * multiple threads. Iterators are <i>weakly consistent</i>, returning
28 dl 1.1 * elements reflecting the state of the map at some point at or since
29     * the creation of the iterator. They do <em>not</em> throw {@link
30     * ConcurrentModificationException}, and may proceed concurrently with
31     * other operations. Ascending key ordered views and their iterators
32     * are faster than descending ones.
33     *
34 jsr166 1.82 * <p>All {@code Map.Entry} pairs returned by methods in this class
35 dl 1.1 * and its views represent snapshots of mappings at the time they were
36 jsr166 1.82 * produced. They do <em>not</em> support the {@code Entry.setValue}
37 dl 1.1 * method. (Note however that it is possible to change mappings in the
38 jsr166 1.82 * associated map using {@code put}, {@code putIfAbsent}, or
39     * {@code replace}, depending on exactly which effect you need.)
40 dl 1.1 *
41 jsr166 1.82 * <p>Beware that, unlike in most collections, the {@code size}
42 dl 1.1 * method is <em>not</em> a constant-time operation. Because of the
43     * asynchronous nature of these maps, determining the current number
44 dl 1.69 * of elements requires a traversal of the elements, and so may report
45     * inaccurate results if this collection is modified during traversal.
46 jsr166 1.82 * Additionally, the bulk operations {@code putAll}, {@code equals},
47     * {@code toArray}, {@code containsValue}, and {@code clear} are
48 dl 1.69 * <em>not</em> guaranteed to be performed atomically. For example, an
49 jsr166 1.82 * iterator operating concurrently with a {@code putAll} operation
50 dl 1.69 * might view only some of the added elements.
51 dl 1.1 *
52     * <p>This class and its views and iterators implement all of the
53     * <em>optional</em> methods of the {@link Map} and {@link Iterator}
54     * interfaces. Like most other concurrent collections, this class does
55 jsr166 1.82 * <em>not</em> permit the use of {@code null} keys or values because some
56 jsr166 1.22 * null return values cannot be reliably distinguished from the absence of
57     * elements.
58 dl 1.1 *
59 dl 1.83 * <p>A {@link Set} projection of a ConcurrentSkipListMap may be
60     * created (using {@link #newKeySet()}}), or viewed (using {@link
61     * #keySet(Object)} when only keys are of interest, and the mapped
62     * values are (perhaps transiently) not used or all take the same
63     * mapping value.
64     *
65 jsr166 1.21 * <p>This class is a member of the
66 jsr166 1.51 * <a href="{@docRoot}/../technotes/guides/collections/index.html">
67 jsr166 1.21 * Java Collections Framework</a>.
68     *
69 dl 1.1 * @author Doug Lea
70     * @param <K> the type of keys maintained by this map
71 dl 1.9 * @param <V> the type of mapped values
72 jsr166 1.20 * @since 1.6
73 dl 1.1 */
74 jsr166 1.79 @SuppressWarnings("unchecked")
75 dl 1.9 public class ConcurrentSkipListMap<K,V> extends AbstractMap<K,V>
76 dl 1.1 implements ConcurrentNavigableMap<K,V>,
77 dl 1.9 Cloneable,
78 dl 1.1 java.io.Serializable {
79     /*
80     * This class implements a tree-like two-dimensionally linked skip
81     * list in which the index levels are represented in separate
82     * nodes from the base nodes holding data. There are two reasons
83     * for taking this approach instead of the usual array-based
84     * structure: 1) Array based implementations seem to encounter
85     * more complexity and overhead 2) We can use cheaper algorithms
86     * for the heavily-traversed index lists than can be used for the
87     * base lists. Here's a picture of some of the basics for a
88     * possible list with 2 levels of index:
89     *
90     * Head nodes Index nodes
91 dl 1.9 * +-+ right +-+ +-+
92 dl 1.1 * |2|---------------->| |--------------------->| |->null
93 dl 1.9 * +-+ +-+ +-+
94 dl 1.1 * | down | |
95     * v v v
96 dl 1.9 * +-+ +-+ +-+ +-+ +-+ +-+
97 dl 1.1 * |1|----------->| |->| |------>| |----------->| |------>| |->null
98 dl 1.9 * +-+ +-+ +-+ +-+ +-+ +-+
99 dl 1.1 * v | | | | |
100     * Nodes next v v v v v
101 dl 1.9 * +-+ +-+ +-+ +-+ +-+ +-+ +-+ +-+ +-+ +-+ +-+ +-+
102 dl 1.1 * | |->|A|->|B|->|C|->|D|->|E|->|F|->|G|->|H|->|I|->|J|->|K|->null
103 dl 1.9 * +-+ +-+ +-+ +-+ +-+ +-+ +-+ +-+ +-+ +-+ +-+ +-+
104 dl 1.1 *
105     * The base lists use a variant of the HM linked ordered set
106     * algorithm. See Tim Harris, "A pragmatic implementation of
107     * non-blocking linked lists"
108     * http://www.cl.cam.ac.uk/~tlh20/publications.html and Maged
109     * Michael "High Performance Dynamic Lock-Free Hash Tables and
110     * List-Based Sets"
111     * http://www.research.ibm.com/people/m/michael/pubs.htm. The
112     * basic idea in these lists is to mark the "next" pointers of
113     * deleted nodes when deleting to avoid conflicts with concurrent
114     * insertions, and when traversing to keep track of triples
115     * (predecessor, node, successor) in order to detect when and how
116     * to unlink these deleted nodes.
117     *
118     * Rather than using mark-bits to mark list deletions (which can
119     * be slow and space-intensive using AtomicMarkedReference), nodes
120     * use direct CAS'able next pointers. On deletion, instead of
121     * marking a pointer, they splice in another node that can be
122     * thought of as standing for a marked pointer (indicating this by
123     * using otherwise impossible field values). Using plain nodes
124     * acts roughly like "boxed" implementations of marked pointers,
125     * but uses new nodes only when nodes are deleted, not for every
126     * link. This requires less space and supports faster
127     * traversal. Even if marked references were better supported by
128     * JVMs, traversal using this technique might still be faster
129     * because any search need only read ahead one more node than
130     * otherwise required (to check for trailing marker) rather than
131     * unmasking mark bits or whatever on each read.
132     *
133     * This approach maintains the essential property needed in the HM
134     * algorithm of changing the next-pointer of a deleted node so
135     * that any other CAS of it will fail, but implements the idea by
136     * changing the pointer to point to a different node, not by
137     * marking it. While it would be possible to further squeeze
138     * space by defining marker nodes not to have key/value fields, it
139     * isn't worth the extra type-testing overhead. The deletion
140     * markers are rarely encountered during traversal and are
141     * normally quickly garbage collected. (Note that this technique
142     * would not work well in systems without garbage collection.)
143     *
144     * In addition to using deletion markers, the lists also use
145     * nullness of value fields to indicate deletion, in a style
146     * similar to typical lazy-deletion schemes. If a node's value is
147     * null, then it is considered logically deleted and ignored even
148     * though it is still reachable. This maintains proper control of
149     * concurrent replace vs delete operations -- an attempted replace
150     * must fail if a delete beat it by nulling field, and a delete
151     * must return the last non-null value held in the field. (Note:
152     * Null, rather than some special marker, is used for value fields
153     * here because it just so happens to mesh with the Map API
154     * requirement that method get returns null if there is no
155     * mapping, which allows nodes to remain concurrently readable
156     * even when deleted. Using any other marker value here would be
157     * messy at best.)
158     *
159     * Here's the sequence of events for a deletion of node n with
160     * predecessor b and successor f, initially:
161     *
162 dl 1.9 * +------+ +------+ +------+
163 dl 1.1 * ... | b |------>| n |----->| f | ...
164 dl 1.9 * +------+ +------+ +------+
165 dl 1.1 *
166     * 1. CAS n's value field from non-null to null.
167     * From this point on, no public operations encountering
168     * the node consider this mapping to exist. However, other
169     * ongoing insertions and deletions might still modify
170     * n's next pointer.
171     *
172     * 2. CAS n's next pointer to point to a new marker node.
173     * From this point on, no other nodes can be appended to n.
174     * which avoids deletion errors in CAS-based linked lists.
175     *
176     * +------+ +------+ +------+ +------+
177     * ... | b |------>| n |----->|marker|------>| f | ...
178 dl 1.9 * +------+ +------+ +------+ +------+
179 dl 1.1 *
180     * 3. CAS b's next pointer over both n and its marker.
181     * From this point on, no new traversals will encounter n,
182     * and it can eventually be GCed.
183     * +------+ +------+
184     * ... | b |----------------------------------->| f | ...
185 dl 1.9 * +------+ +------+
186     *
187 dl 1.1 * A failure at step 1 leads to simple retry due to a lost race
188     * with another operation. Steps 2-3 can fail because some other
189     * thread noticed during a traversal a node with null value and
190     * helped out by marking and/or unlinking. This helping-out
191     * ensures that no thread can become stuck waiting for progress of
192     * the deleting thread. The use of marker nodes slightly
193     * complicates helping-out code because traversals must track
194     * consistent reads of up to four nodes (b, n, marker, f), not
195     * just (b, n, f), although the next field of a marker is
196     * immutable, and once a next field is CAS'ed to point to a
197     * marker, it never again changes, so this requires less care.
198     *
199     * Skip lists add indexing to this scheme, so that the base-level
200     * traversals start close to the locations being found, inserted
201     * or deleted -- usually base level traversals only traverse a few
202     * nodes. This doesn't change the basic algorithm except for the
203     * need to make sure base traversals start at predecessors (here,
204     * b) that are not (structurally) deleted, otherwise retrying
205 dl 1.9 * after processing the deletion.
206 dl 1.1 *
207     * Index levels are maintained as lists with volatile next fields,
208     * using CAS to link and unlink. Races are allowed in index-list
209     * operations that can (rarely) fail to link in a new index node
210     * or delete one. (We can't do this of course for data nodes.)
211     * However, even when this happens, the index lists remain sorted,
212     * so correctly serve as indices. This can impact performance,
213     * but since skip lists are probabilistic anyway, the net result
214     * is that under contention, the effective "p" value may be lower
215     * than its nominal value. And race windows are kept small enough
216     * that in practice these failures are rare, even under a lot of
217     * contention.
218     *
219     * The fact that retries (for both base and index lists) are
220     * relatively cheap due to indexing allows some minor
221     * simplifications of retry logic. Traversal restarts are
222     * performed after most "helping-out" CASes. This isn't always
223     * strictly necessary, but the implicit backoffs tend to help
224     * reduce other downstream failed CAS's enough to outweigh restart
225     * cost. This worsens the worst case, but seems to improve even
226     * highly contended cases.
227     *
228     * Unlike most skip-list implementations, index insertion and
229     * deletion here require a separate traversal pass occuring after
230     * the base-level action, to add or remove index nodes. This adds
231     * to single-threaded overhead, but improves contended
232     * multithreaded performance by narrowing interference windows,
233     * and allows deletion to ensure that all index nodes will be made
234     * unreachable upon return from a public remove operation, thus
235     * avoiding unwanted garbage retention. This is more important
236     * here than in some other data structures because we cannot null
237     * out node fields referencing user keys since they might still be
238     * read by other ongoing traversals.
239     *
240     * Indexing uses skip list parameters that maintain good search
241     * performance while using sparser-than-usual indices: The
242     * hardwired parameters k=1, p=0.5 (see method randomLevel) mean
243     * that about one-quarter of the nodes have indices. Of those that
244     * do, half have one level, a quarter have two, and so on (see
245     * Pugh's Skip List Cookbook, sec 3.4). The expected total space
246     * requirement for a map is slightly less than for the current
247     * implementation of java.util.TreeMap.
248     *
249     * Changing the level of the index (i.e, the height of the
250     * tree-like structure) also uses CAS. The head index has initial
251     * level/height of one. Creation of an index with height greater
252     * than the current level adds a level to the head index by
253     * CAS'ing on a new top-most head. To maintain good performance
254     * after a lot of removals, deletion methods heuristically try to
255     * reduce the height if the topmost levels appear to be empty.
256     * This may encounter races in which it possible (but rare) to
257     * reduce and "lose" a level just as it is about to contain an
258     * index (that will then never be encountered). This does no
259     * structural harm, and in practice appears to be a better option
260     * than allowing unrestrained growth of levels.
261     *
262     * The code for all this is more verbose than you'd like. Most
263     * operations entail locating an element (or position to insert an
264     * element). The code to do this can't be nicely factored out
265     * because subsequent uses require a snapshot of predecessor
266     * and/or successor and/or value fields which can't be returned
267     * all at once, at least not without creating yet another object
268     * to hold them -- creating such little objects is an especially
269     * bad idea for basic internal search operations because it adds
270     * to GC overhead. (This is one of the few times I've wished Java
271     * had macros.) Instead, some traversal code is interleaved within
272     * insertion and removal operations. The control logic to handle
273     * all the retry conditions is sometimes twisty. Most search is
274     * broken into 2 parts. findPredecessor() searches index nodes
275     * only, returning a base-level predecessor of the key. findNode()
276     * finishes out the base-level search. Even with this factoring,
277     * there is a fair amount of near-duplication of code to handle
278     * variants.
279     *
280 dl 1.88 * A previous version of this class wrapped non-comparable keys
281     * with their comparators to emulate Comparables when using
282     * comparators vs Comparables. This saved the need to choose
283     * among the unpleasant options of either possibly re-reading a
284     * comparator on each comparison (which suffers when among all of
285     * the volatile read snapshots) versus code duplication, at the
286     * cost of usually requiring an object construction per operation
287     * when not naturally ordered. However, as usage evolves, use of
288     * comparators has become more common, so we have to settle for
289     * code duplication as the lesser evil. Thus, there are "xxxCmp"
290     * versions of many of the xxx methods, all bundled later in this
291     * file.
292     *
293 dl 1.1 * For explanation of algorithms sharing at least a couple of
294     * features with this one, see Mikhail Fomitchev's thesis
295     * (http://www.cs.yorku.ca/~mikhail/), Keir Fraser's thesis
296 dl 1.4 * (http://www.cl.cam.ac.uk/users/kaf24/), and Hakan Sundell's
297 dl 1.1 * thesis (http://www.cs.chalmers.se/~phs/).
298     *
299     * Given the use of tree-like index nodes, you might wonder why
300     * this doesn't use some kind of search tree instead, which would
301     * support somewhat faster search operations. The reason is that
302     * there are no known efficient lock-free insertion and deletion
303     * algorithms for search trees. The immutability of the "down"
304     * links of index nodes (as opposed to mutable "left" fields in
305     * true trees) makes this tractable using only CAS operations.
306     *
307     * Notation guide for local variables
308     * Node: b, n, f for predecessor, node, successor
309     * Index: q, r, d for index node, right, down.
310     * t for another index node
311     * Head: h
312     * Levels: j
313     * Keys: k, key
314     * Values: v, value
315     * Comparisons: c
316     */
317    
318     private static final long serialVersionUID = -8627078645895051609L;
319    
320     /**
321 dl 1.40 * Generates the initial random seed for the cheaper per-instance
322     * random number generators used in randomLevel.
323     */
324     private static final Random seedGenerator = new Random();
325    
326     /**
327 dl 1.1 * Special value used to identify base-level header
328 dl 1.9 */
329 dl 1.1 private static final Object BASE_HEADER = new Object();
330    
331     /**
332 dl 1.9 * The topmost head index of the skiplist.
333 dl 1.1 */
334     private transient volatile HeadIndex<K,V> head;
335    
336     /**
337 jsr166 1.10 * The comparator used to maintain order in this map, or null
338 jsr166 1.22 * if using natural ordering.
339 dl 1.1 * @serial
340     */
341     private final Comparator<? super K> comparator;
342    
343     /** Lazily initialized key set */
344 dl 1.83 private transient KeySetView<K,V> keySet;
345 dl 1.1 /** Lazily initialized entry set */
346 jsr166 1.71 private transient EntrySet<K,V> entrySet;
347 dl 1.1 /** Lazily initialized values collection */
348 jsr166 1.71 private transient Values<V> values;
349 dl 1.1 /** Lazily initialized descending key set */
350 dl 1.46 private transient ConcurrentNavigableMap<K,V> descendingMap;
351 dl 1.1
352     /**
353 jsr166 1.13 * Initializes or resets state. Needed by constructors, clone,
354 dl 1.1 * clear, readObject. and ConcurrentSkipListSet.clone.
355     * (Note that comparator must be separately initialized.)
356     */
357     final void initialize() {
358     keySet = null;
359 dl 1.9 entrySet = null;
360 dl 1.1 values = null;
361 dl 1.46 descendingMap = null;
362 dl 1.1 head = new HeadIndex<K,V>(new Node<K,V>(null, BASE_HEADER, null),
363     null, null, 1);
364     }
365    
366     /**
367     * compareAndSet head node
368     */
369     private boolean casHead(HeadIndex<K,V> cmp, HeadIndex<K,V> val) {
370 dl 1.59 return UNSAFE.compareAndSwapObject(this, headOffset, cmp, val);
371 dl 1.1 }
372    
373     /* ---------------- Nodes -------------- */
374    
375     /**
376     * Nodes hold keys and values, and are singly linked in sorted
377     * order, possibly with some intervening marker nodes. The list is
378     * headed by a dummy node accessible as head.node. The value field
379     * is declared only as Object because it takes special non-V
380     * values for marker and header nodes.
381     */
382     static final class Node<K,V> {
383     final K key;
384     volatile Object value;
385     volatile Node<K,V> next;
386    
387     /**
388     * Creates a new regular node.
389     */
390     Node(K key, Object value, Node<K,V> next) {
391     this.key = key;
392     this.value = value;
393     this.next = next;
394     }
395    
396     /**
397     * Creates a new marker node. A marker is distinguished by
398     * having its value field point to itself. Marker nodes also
399     * have null keys, a fact that is exploited in a few places,
400     * but this doesn't distinguish markers from the base-level
401     * header node (head.node), which also has a null key.
402     */
403     Node(Node<K,V> next) {
404     this.key = null;
405     this.value = this;
406     this.next = next;
407     }
408    
409     /**
410     * compareAndSet value field
411     */
412     boolean casValue(Object cmp, Object val) {
413 dl 1.59 return UNSAFE.compareAndSwapObject(this, valueOffset, cmp, val);
414 dl 1.1 }
415 jsr166 1.60
416 dl 1.1 /**
417     * compareAndSet next field
418     */
419     boolean casNext(Node<K,V> cmp, Node<K,V> val) {
420 dl 1.59 return UNSAFE.compareAndSwapObject(this, nextOffset, cmp, val);
421 dl 1.1 }
422    
423     /**
424 jsr166 1.10 * Returns true if this node is a marker. This method isn't
425     * actually called in any current code checking for markers
426 dl 1.1 * because callers will have already read value field and need
427     * to use that read (not another done here) and so directly
428     * test if value points to node.
429     * @param n a possibly null reference to a node
430     * @return true if this node is a marker node
431     */
432     boolean isMarker() {
433     return value == this;
434     }
435    
436     /**
437 jsr166 1.10 * Returns true if this node is the header of base-level list.
438 dl 1.1 * @return true if this node is header node
439     */
440     boolean isBaseHeader() {
441     return value == BASE_HEADER;
442     }
443    
444     /**
445     * Tries to append a deletion marker to this node.
446     * @param f the assumed current successor of this node
447     * @return true if successful
448     */
449     boolean appendMarker(Node<K,V> f) {
450     return casNext(f, new Node<K,V>(f));
451     }
452    
453     /**
454     * Helps out a deletion by appending marker or unlinking from
455     * predecessor. This is called during traversals when value
456     * field seen to be null.
457     * @param b predecessor
458     * @param f successor
459     */
460     void helpDelete(Node<K,V> b, Node<K,V> f) {
461     /*
462     * Rechecking links and then doing only one of the
463     * help-out stages per call tends to minimize CAS
464     * interference among helping threads.
465     */
466     if (f == next && this == b.next) {
467     if (f == null || f.value != f) // not already marked
468     appendMarker(f);
469     else
470     b.casNext(this, f.next);
471     }
472     }
473    
474     /**
475 jsr166 1.11 * Returns value if this node contains a valid key-value pair,
476 dl 1.9 * else null.
477 dl 1.1 * @return this node's value if it isn't a marker or header or
478     * is deleted, else null.
479     */
480     V getValidValue() {
481     Object v = value;
482     if (v == this || v == BASE_HEADER)
483     return null;
484     return (V)v;
485     }
486    
487     /**
488 jsr166 1.10 * Creates and returns a new SimpleImmutableEntry holding current
489     * mapping if this node holds a valid value, else null.
490 dl 1.1 * @return new entry or null
491     */
492 dl 1.2 AbstractMap.SimpleImmutableEntry<K,V> createSnapshot() {
493 dl 1.1 V v = getValidValue();
494     if (v == null)
495     return null;
496 dl 1.24 return new AbstractMap.SimpleImmutableEntry<K,V>(key, v);
497 dl 1.1 }
498 dl 1.59
499 dl 1.65 // UNSAFE mechanics
500    
501     private static final sun.misc.Unsafe UNSAFE;
502     private static final long valueOffset;
503     private static final long nextOffset;
504 dl 1.59
505 dl 1.65 static {
506     try {
507     UNSAFE = sun.misc.Unsafe.getUnsafe();
508 jsr166 1.72 Class<?> k = Node.class;
509 dl 1.65 valueOffset = UNSAFE.objectFieldOffset
510     (k.getDeclaredField("value"));
511     nextOffset = UNSAFE.objectFieldOffset
512     (k.getDeclaredField("next"));
513     } catch (Exception e) {
514     throw new Error(e);
515     }
516     }
517 dl 1.1 }
518    
519     /* ---------------- Indexing -------------- */
520    
521     /**
522 dl 1.40 * Index nodes represent the levels of the skip list. Note that
523     * even though both Nodes and Indexes have forward-pointing
524     * fields, they have different types and are handled in different
525     * ways, that can't nicely be captured by placing field in a
526     * shared abstract class.
527 dl 1.1 */
528     static class Index<K,V> {
529     final Node<K,V> node;
530     final Index<K,V> down;
531     volatile Index<K,V> right;
532    
533     /**
534 jsr166 1.10 * Creates index node with given values.
535 dl 1.9 */
536 dl 1.1 Index(Node<K,V> node, Index<K,V> down, Index<K,V> right) {
537     this.node = node;
538     this.down = down;
539     this.right = right;
540     }
541    
542     /**
543     * compareAndSet right field
544     */
545     final boolean casRight(Index<K,V> cmp, Index<K,V> val) {
546 dl 1.59 return UNSAFE.compareAndSwapObject(this, rightOffset, cmp, val);
547 dl 1.1 }
548    
549     /**
550     * Returns true if the node this indexes has been deleted.
551     * @return true if indexed node is known to be deleted
552     */
553     final boolean indexesDeletedNode() {
554     return node.value == null;
555     }
556    
557     /**
558     * Tries to CAS newSucc as successor. To minimize races with
559     * unlink that may lose this index node, if the node being
560     * indexed is known to be deleted, it doesn't try to link in.
561     * @param succ the expected current successor
562     * @param newSucc the new successor
563     * @return true if successful
564     */
565     final boolean link(Index<K,V> succ, Index<K,V> newSucc) {
566     Node<K,V> n = node;
567 dl 1.9 newSucc.right = succ;
568 dl 1.1 return n.value != null && casRight(succ, newSucc);
569     }
570    
571     /**
572     * Tries to CAS right field to skip over apparent successor
573     * succ. Fails (forcing a retraversal by caller) if this node
574     * is known to be deleted.
575     * @param succ the expected current successor
576     * @return true if successful
577     */
578     final boolean unlink(Index<K,V> succ) {
579     return !indexesDeletedNode() && casRight(succ, succ.right);
580     }
581 dl 1.59
582     // Unsafe mechanics
583 dl 1.65 private static final sun.misc.Unsafe UNSAFE;
584     private static final long rightOffset;
585     static {
586     try {
587     UNSAFE = sun.misc.Unsafe.getUnsafe();
588 jsr166 1.72 Class<?> k = Index.class;
589 dl 1.65 rightOffset = UNSAFE.objectFieldOffset
590     (k.getDeclaredField("right"));
591     } catch (Exception e) {
592     throw new Error(e);
593     }
594     }
595 dl 1.1 }
596    
597     /* ---------------- Head nodes -------------- */
598    
599     /**
600     * Nodes heading each level keep track of their level.
601     */
602     static final class HeadIndex<K,V> extends Index<K,V> {
603     final int level;
604     HeadIndex(Node<K,V> node, Index<K,V> down, Index<K,V> right, int level) {
605     super(node, down, right);
606     this.level = level;
607     }
608 dl 1.9 }
609 dl 1.1
610     /* ---------------- Comparison utilities -------------- */
611    
612     /**
613 dl 1.88 * Compares using comparator or natural ordering. Used in cases
614     * where it isn't worthwhile to use multiple code paths.
615 dl 1.1 */
616     int compare(K k1, K k2) throws ClassCastException {
617     Comparator<? super K> cmp = comparator;
618     if (cmp != null)
619     return cmp.compare(k1, k2);
620     else
621 jsr166 1.18 return ((Comparable<? super K>)k1).compareTo(k2);
622 dl 1.1 }
623    
624     /**
625 jsr166 1.10 * Returns true if given key greater than or equal to least and
626 dl 1.1 * strictly less than fence, bypassing either test if least or
627 dl 1.5 * fence are null. Needed mainly in submap operations.
628 dl 1.1 */
629     boolean inHalfOpenRange(K key, K least, K fence) {
630 dl 1.9 if (key == null)
631 dl 1.1 throw new NullPointerException();
632     return ((least == null || compare(key, least) >= 0) &&
633     (fence == null || compare(key, fence) < 0));
634     }
635    
636     /**
637 jsr166 1.10 * Returns true if given key greater than or equal to least and less
638 dl 1.1 * or equal to fence. Needed mainly in submap operations.
639     */
640     boolean inOpenRange(K key, K least, K fence) {
641 dl 1.9 if (key == null)
642 dl 1.1 throw new NullPointerException();
643     return ((least == null || compare(key, least) >= 0) &&
644     (fence == null || compare(key, fence) <= 0));
645     }
646    
647     /* ---------------- Traversal -------------- */
648    
649     /**
650 jsr166 1.10 * Returns a base-level node with key strictly less than given key,
651 dl 1.1 * or the base-level header if there is no such node. Also
652     * unlinks indexes to deleted nodes found along the way. Callers
653     * rely on this side-effect of clearing indices to deleted nodes.
654     * @param key the key
655 dl 1.9 * @return a predecessor of key
656 dl 1.1 */
657 dl 1.9 private Node<K,V> findPredecessor(Comparable<? super K> key) {
658 jsr166 1.41 if (key == null)
659 dl 1.40 throw new NullPointerException(); // don't postpone errors
660 dl 1.1 for (;;) {
661     Index<K,V> q = head;
662 dl 1.40 Index<K,V> r = q.right;
663 dl 1.1 for (;;) {
664 dl 1.40 if (r != null) {
665     Node<K,V> n = r.node;
666     K k = n.key;
667     if (n.value == null) {
668     if (!q.unlink(r))
669     break; // restart
670     r = q.right; // reread r
671     continue;
672 dl 1.1 }
673 dl 1.40 if (key.compareTo(k) > 0) {
674 dl 1.1 q = r;
675 dl 1.40 r = r.right;
676 dl 1.1 continue;
677     }
678     }
679 dl 1.40 Index<K,V> d = q.down;
680     if (d != null) {
681 dl 1.1 q = d;
682 dl 1.40 r = d.right;
683     } else
684 dl 1.1 return q.node;
685     }
686     }
687     }
688    
689     /**
690 jsr166 1.10 * Returns node holding key or null if no such, clearing out any
691 dl 1.1 * deleted nodes seen along the way. Repeatedly traverses at
692     * base-level looking for key starting at predecessor returned
693     * from findPredecessor, processing base-level deletions as
694     * encountered. Some callers rely on this side-effect of clearing
695     * deleted nodes.
696     *
697     * Restarts occur, at traversal step centered on node n, if:
698     *
699     * (1) After reading n's next field, n is no longer assumed
700     * predecessor b's current successor, which means that
701     * we don't have a consistent 3-node snapshot and so cannot
702     * unlink any subsequent deleted nodes encountered.
703     *
704     * (2) n's value field is null, indicating n is deleted, in
705     * which case we help out an ongoing structural deletion
706     * before retrying. Even though there are cases where such
707     * unlinking doesn't require restart, they aren't sorted out
708     * here because doing so would not usually outweigh cost of
709     * restarting.
710     *
711 dl 1.9 * (3) n is a marker or n's predecessor's value field is null,
712 dl 1.1 * indicating (among other possibilities) that
713     * findPredecessor returned a deleted node. We can't unlink
714     * the node because we don't know its predecessor, so rely
715     * on another call to findPredecessor to notice and return
716     * some earlier predecessor, which it will do. This check is
717     * only strictly needed at beginning of loop, (and the
718     * b.value check isn't strictly needed at all) but is done
719     * each iteration to help avoid contention with other
720     * threads by callers that will fail to be able to change
721     * links, and so will retry anyway.
722     *
723     * The traversal loops in doPut, doRemove, and findNear all
724     * include the same three kinds of checks. And specialized
725 dl 1.31 * versions appear in findFirst, and findLast and their
726     * variants. They can't easily share code because each uses the
727 dl 1.1 * reads of fields held in locals occurring in the orders they
728     * were performed.
729 dl 1.9 *
730 dl 1.1 * @param key the key
731 jsr166 1.22 * @return node holding key, or null if no such
732 dl 1.1 */
733 dl 1.9 private Node<K,V> findNode(Comparable<? super K> key) {
734 dl 1.88 if (key == null)
735     throw new NullPointerException(); // don't postpone errors
736 dl 1.1 for (;;) {
737     Node<K,V> b = findPredecessor(key);
738     Node<K,V> n = b.next;
739     for (;;) {
740 dl 1.9 if (n == null)
741 dl 1.1 return null;
742     Node<K,V> f = n.next;
743     if (n != b.next) // inconsistent read
744     break;
745     Object v = n.value;
746     if (v == null) { // n is deleted
747     n.helpDelete(b, f);
748     break;
749     }
750     if (v == n || b.value == null) // b is deleted
751     break;
752     int c = key.compareTo(n.key);
753 dl 1.40 if (c == 0)
754     return n;
755 dl 1.1 if (c < 0)
756     return null;
757     b = n;
758     n = f;
759     }
760     }
761     }
762    
763 dl 1.9 /**
764 dl 1.88 * Gets value for key. Almost the same as findNode, but returns
765     * the found value (to avoid retries during re-reads)
766     *
767 dl 1.1 * @param okey the key
768     * @return the value, or null if absent
769     */
770     private V doGet(Object okey) {
771 dl 1.88 if (okey == null)
772     throw new NullPointerException();
773     @SuppressWarnings("unchecked") Comparable<? super K> key =
774     (Comparable<? super K>)okey;
775 dl 1.1 for (;;) {
776 dl 1.88 Node<K,V> b = findPredecessor(key);
777     Node<K,V> n = b.next;
778     for (;;) {
779     if (n == null)
780     return null;
781     Node<K,V> f = n.next;
782     if (n != b.next) // inconsistent read
783     break;
784     Object v = n.value;
785     if (v == null) { // n is deleted
786     n.helpDelete(b, f);
787     break;
788     }
789     if (v == n || b.value == null) // b is deleted
790     break;
791     int c = key.compareTo(n.key);
792     if (c == 0)
793     return (V)v;
794     if (c < 0)
795     return null;
796     b = n;
797     n = f;
798     }
799 dl 1.1 }
800     }
801    
802 dl 1.88
803 dl 1.1 /* ---------------- Insertion -------------- */
804    
805     /**
806     * Main insertion method. Adds element if not present, or
807     * replaces value if present and onlyIfAbsent is false.
808 dl 1.9 * @param kkey the key
809 dl 1.1 * @param value the value that must be associated with key
810     * @param onlyIfAbsent if should not insert if already present
811     * @return the old value, or null if newly inserted
812     */
813     private V doPut(K kkey, V value, boolean onlyIfAbsent) {
814 dl 1.88 if (kkey == null)
815     throw new NullPointerException();
816     @SuppressWarnings("unchecked") Comparable<? super K> key =
817     (Comparable<? super K>)kkey;
818 dl 1.1 for (;;) {
819     Node<K,V> b = findPredecessor(key);
820     Node<K,V> n = b.next;
821     for (;;) {
822     if (n != null) {
823     Node<K,V> f = n.next;
824     if (n != b.next) // inconsistent read
825 jsr166 1.57 break;
826 dl 1.1 Object v = n.value;
827     if (v == null) { // n is deleted
828     n.helpDelete(b, f);
829     break;
830     }
831     if (v == n || b.value == null) // b is deleted
832     break;
833     int c = key.compareTo(n.key);
834     if (c > 0) {
835     b = n;
836     n = f;
837     continue;
838     }
839     if (c == 0) {
840     if (onlyIfAbsent || n.casValue(v, value))
841     return (V)v;
842     else
843     break; // restart if lost race to replace value
844     }
845     // else c < 0; fall through
846     }
847 dl 1.9
848 dl 1.1 Node<K,V> z = new Node<K,V>(kkey, value, n);
849 dl 1.9 if (!b.casNext(n, z))
850 dl 1.1 break; // restart if lost race to append to b
851 dl 1.9 int level = randomLevel();
852     if (level > 0)
853 dl 1.1 insertIndex(z, level);
854     return null;
855     }
856     }
857     }
858    
859     /**
860 jsr166 1.10 * Returns a random level for inserting a new node.
861 dl 1.35 * Hardwired to k=1, p=0.5, max 31 (see above and
862 dl 1.34 * Pugh's "Skip List Cookbook", sec 3.4).
863 dl 1.1 */
864     private int randomLevel() {
865 dl 1.91 // int x = ThreadLocalRandom.nextSecondarySeed();
866     int x = ThreadLocalRandom.current().nextInt();
867 dl 1.88 int level = 0;
868     if ((x & 0x80000001) == 0) { // test highest and lowest bits
869     do { ++level; }
870     while (((x >>>= 1) & 1) != 0);
871     }
872 dl 1.1 return level;
873     }
874    
875     /**
876 jsr166 1.11 * Creates and adds index nodes for the given node.
877 dl 1.1 * @param z the node
878     * @param level the level of the index
879     */
880     private void insertIndex(Node<K,V> z, int level) {
881     HeadIndex<K,V> h = head;
882     int max = h.level;
883    
884     if (level <= max) {
885     Index<K,V> idx = null;
886     for (int i = 1; i <= level; ++i)
887     idx = new Index<K,V>(z, idx, null);
888     addIndex(idx, h, level);
889    
890     } else { // Add a new level
891     /*
892     * To reduce interference by other threads checking for
893     * empty levels in tryReduceLevel, new levels are added
894     * with initialized right pointers. Which in turn requires
895     * keeping levels in an array to access them while
896     * creating new head index nodes from the opposite
897     * direction.
898     */
899     level = max + 1;
900 jsr166 1.72 Index<K,V>[] idxs = (Index<K,V>[])new Index<?,?>[level+1];
901 dl 1.1 Index<K,V> idx = null;
902 dl 1.9 for (int i = 1; i <= level; ++i)
903 dl 1.1 idxs[i] = idx = new Index<K,V>(z, idx, null);
904    
905     HeadIndex<K,V> oldh;
906     int k;
907     for (;;) {
908     oldh = head;
909     int oldLevel = oldh.level;
910     if (level <= oldLevel) { // lost race to add level
911     k = level;
912     break;
913     }
914     HeadIndex<K,V> newh = oldh;
915     Node<K,V> oldbase = oldh.node;
916 dl 1.9 for (int j = oldLevel+1; j <= level; ++j)
917 dl 1.1 newh = new HeadIndex<K,V>(oldbase, newh, idxs[j], j);
918     if (casHead(oldh, newh)) {
919     k = oldLevel;
920     break;
921     }
922     }
923     addIndex(idxs[k], oldh, k);
924     }
925     }
926    
927     /**
928 jsr166 1.10 * Adds given index nodes from given level down to 1.
929 dl 1.1 * @param idx the topmost index node being inserted
930     * @param h the value of head to use to insert. This must be
931 jsr166 1.89 * snapshotted by callers to provide correct insertion level.
932 dl 1.1 * @param indexLevel the level of the index
933     */
934 dl 1.88 @SuppressWarnings("unchecked")
935 dl 1.1 private void addIndex(Index<K,V> idx, HeadIndex<K,V> h, int indexLevel) {
936     // Track next level to insert in case of retries
937     int insertionLevel = indexLevel;
938 dl 1.88 K key = idx.node.key;
939 dl 1.40 if (key == null) throw new NullPointerException();
940 dl 1.88 Comparator<? super K> cmp = comparator;
941 dl 1.1 // Similar to findPredecessor, but adding index nodes along
942     // path to key.
943     for (;;) {
944 dl 1.40 int j = h.level;
945 dl 1.1 Index<K,V> q = h;
946 dl 1.40 Index<K,V> r = q.right;
947 dl 1.1 Index<K,V> t = idx;
948     for (;;) {
949     if (r != null) {
950 dl 1.40 Node<K,V> n = r.node;
951 dl 1.1 // compare before deletion check avoids needing recheck
952 jsr166 1.89 int c = (cmp == null) ?
953     ((Comparable<? super K>)key).compareTo(n.key) :
954     cmp.compare(key, n.key);
955 dl 1.40 if (n.value == null) {
956     if (!q.unlink(r))
957 dl 1.9 break;
958 dl 1.40 r = q.right;
959     continue;
960 dl 1.1 }
961     if (c > 0) {
962     q = r;
963 dl 1.40 r = r.right;
964 dl 1.1 continue;
965     }
966     }
967    
968     if (j == insertionLevel) {
969     // Don't insert index if node already deleted
970     if (t.indexesDeletedNode()) {
971 dl 1.88 if (cmp == null)
972     findNode((Comparable<? super K>)key); // cleans up
973     else
974     findNodeCmp(cmp, key);
975 dl 1.1 return;
976     }
977 dl 1.9 if (!q.link(r, t))
978 dl 1.1 break; // restart
979     if (--insertionLevel == 0) {
980     // need final deletion check before return
981 dl 1.88 if (t.indexesDeletedNode()) {
982     if (cmp == null)
983     findNode((Comparable<? super K>)key);
984     else
985     findNodeCmp(cmp, key);
986     }
987 dl 1.1 return;
988     }
989     }
990    
991 dl 1.40 if (--j >= insertionLevel && j < indexLevel)
992 dl 1.1 t = t.down;
993     q = q.down;
994 dl 1.40 r = q.right;
995 dl 1.1 }
996     }
997     }
998    
999     /* ---------------- Deletion -------------- */
1000    
1001     /**
1002     * Main deletion method. Locates node, nulls value, appends a
1003     * deletion marker, unlinks predecessor, removes associated index
1004     * nodes, and possibly reduces head index level.
1005     *
1006     * Index nodes are cleared out simply by calling findPredecessor.
1007     * which unlinks indexes to deleted nodes found along path to key,
1008     * which will include the indexes to this node. This is done
1009     * unconditionally. We can't check beforehand whether there are
1010     * index nodes because it might be the case that some or all
1011     * indexes hadn't been inserted yet for this node during initial
1012     * search for it, and we'd like to ensure lack of garbage
1013 dl 1.9 * retention, so must call to be sure.
1014 dl 1.1 *
1015     * @param okey the key
1016     * @param value if non-null, the value that must be
1017     * associated with key
1018     * @return the node, or null if not found
1019     */
1020 dl 1.46 final V doRemove(Object okey, Object value) {
1021 dl 1.88 if (okey == null)
1022     throw new NullPointerException();
1023     @SuppressWarnings("unchecked") Comparable<? super K> key =
1024     (Comparable<? super K>)okey;
1025 dl 1.9 for (;;) {
1026 dl 1.1 Node<K,V> b = findPredecessor(key);
1027     Node<K,V> n = b.next;
1028     for (;;) {
1029 dl 1.9 if (n == null)
1030 dl 1.1 return null;
1031     Node<K,V> f = n.next;
1032     if (n != b.next) // inconsistent read
1033     break;
1034     Object v = n.value;
1035     if (v == null) { // n is deleted
1036     n.helpDelete(b, f);
1037     break;
1038     }
1039     if (v == n || b.value == null) // b is deleted
1040     break;
1041     int c = key.compareTo(n.key);
1042     if (c < 0)
1043     return null;
1044     if (c > 0) {
1045     b = n;
1046     n = f;
1047     continue;
1048     }
1049 dl 1.9 if (value != null && !value.equals(v))
1050     return null;
1051     if (!n.casValue(v, null))
1052 dl 1.1 break;
1053 dl 1.9 if (!n.appendMarker(f) || !b.casNext(n, f))
1054 dl 1.1 findNode(key); // Retry via findNode
1055     else {
1056     findPredecessor(key); // Clean index
1057 dl 1.9 if (head.right == null)
1058 dl 1.1 tryReduceLevel();
1059     }
1060     return (V)v;
1061     }
1062     }
1063     }
1064    
1065     /**
1066     * Possibly reduce head level if it has no nodes. This method can
1067     * (rarely) make mistakes, in which case levels can disappear even
1068     * though they are about to contain index nodes. This impacts
1069     * performance, not correctness. To minimize mistakes as well as
1070     * to reduce hysteresis, the level is reduced by one only if the
1071     * topmost three levels look empty. Also, if the removed level
1072     * looks non-empty after CAS, we try to change it back quick
1073     * before anyone notices our mistake! (This trick works pretty
1074     * well because this method will practically never make mistakes
1075     * unless current thread stalls immediately before first CAS, in
1076     * which case it is very unlikely to stall again immediately
1077     * afterwards, so will recover.)
1078     *
1079     * We put up with all this rather than just let levels grow
1080     * because otherwise, even a small map that has undergone a large
1081     * number of insertions and removals will have a lot of levels,
1082     * slowing down access more than would an occasional unwanted
1083     * reduction.
1084     */
1085     private void tryReduceLevel() {
1086     HeadIndex<K,V> h = head;
1087     HeadIndex<K,V> d;
1088     HeadIndex<K,V> e;
1089     if (h.level > 3 &&
1090 dl 1.9 (d = (HeadIndex<K,V>)h.down) != null &&
1091     (e = (HeadIndex<K,V>)d.down) != null &&
1092     e.right == null &&
1093     d.right == null &&
1094 dl 1.1 h.right == null &&
1095     casHead(h, d) && // try to set
1096     h.right != null) // recheck
1097     casHead(d, h); // try to backout
1098     }
1099    
1100     /* ---------------- Finding and removing first element -------------- */
1101    
1102     /**
1103 jsr166 1.22 * Specialized variant of findNode to get first valid node.
1104 dl 1.1 * @return first node or null if empty
1105     */
1106     Node<K,V> findFirst() {
1107     for (;;) {
1108     Node<K,V> b = head.node;
1109     Node<K,V> n = b.next;
1110     if (n == null)
1111     return null;
1112 dl 1.9 if (n.value != null)
1113 dl 1.1 return n;
1114     n.helpDelete(b, n.next);
1115     }
1116     }
1117    
1118     /**
1119 dl 1.25 * Removes first entry; returns its snapshot.
1120 jsr166 1.28 * @return null if empty, else snapshot of first entry
1121 dl 1.1 */
1122 dl 1.25 Map.Entry<K,V> doRemoveFirstEntry() {
1123 dl 1.9 for (;;) {
1124 dl 1.1 Node<K,V> b = head.node;
1125     Node<K,V> n = b.next;
1126 dl 1.9 if (n == null)
1127 dl 1.1 return null;
1128     Node<K,V> f = n.next;
1129     if (n != b.next)
1130     continue;
1131     Object v = n.value;
1132     if (v == null) {
1133     n.helpDelete(b, f);
1134     continue;
1135     }
1136     if (!n.casValue(v, null))
1137     continue;
1138     if (!n.appendMarker(f) || !b.casNext(n, f))
1139     findFirst(); // retry
1140     clearIndexToFirst();
1141 dl 1.30 return new AbstractMap.SimpleImmutableEntry<K,V>(n.key, (V)v);
1142 jsr166 1.55 }
1143 dl 1.1 }
1144    
1145     /**
1146 jsr166 1.10 * Clears out index nodes associated with deleted first entry.
1147 dl 1.1 */
1148     private void clearIndexToFirst() {
1149     for (;;) {
1150     Index<K,V> q = head;
1151     for (;;) {
1152     Index<K,V> r = q.right;
1153     if (r != null && r.indexesDeletedNode() && !q.unlink(r))
1154 dl 1.9 break;
1155 dl 1.1 if ((q = q.down) == null) {
1156 dl 1.9 if (head.right == null)
1157 dl 1.1 tryReduceLevel();
1158     return;
1159     }
1160     }
1161     }
1162     }
1163    
1164 dl 1.88 /**
1165     * Removes last entry; returns its snapshot.
1166     * Specialized variant of doRemove.
1167     * @return null if empty, else snapshot of last entry
1168     */
1169     Map.Entry<K,V> doRemoveLastEntry() {
1170     for (;;) {
1171     Node<K,V> b = findPredecessorOfLast();
1172     Node<K,V> n = b.next;
1173     if (n == null) {
1174     if (b.isBaseHeader()) // empty
1175     return null;
1176     else
1177     continue; // all b's successors are deleted; retry
1178     }
1179     for (;;) {
1180     Node<K,V> f = n.next;
1181     if (n != b.next) // inconsistent read
1182     break;
1183     Object v = n.value;
1184     if (v == null) { // n is deleted
1185     n.helpDelete(b, f);
1186     break;
1187     }
1188     if (v == n || b.value == null) // b is deleted
1189     break;
1190     if (f != null) {
1191     b = n;
1192     n = f;
1193     continue;
1194     }
1195     if (!n.casValue(v, null))
1196     break;
1197     K key = n.key;
1198     Comparator<? super K> cmp = comparator;
1199     if (!n.appendMarker(f) || !b.casNext(n, f)) {
1200     if (cmp == null) // Retry via findNode
1201     findNode((Comparable<? super K>)key);
1202     else
1203     findNodeCmp(cmp, key);
1204     }
1205     else { // Clean index
1206     if (cmp == null)
1207     findPredecessor((Comparable<? super K>)key);
1208     else
1209     findPredecessorCmp(cmp, key);
1210     if (head.right == null)
1211     tryReduceLevel();
1212     }
1213     return new AbstractMap.SimpleImmutableEntry<K,V>(key, (V)v);
1214     }
1215     }
1216     }
1217 dl 1.1
1218     /* ---------------- Finding and removing last element -------------- */
1219    
1220     /**
1221 jsr166 1.10 * Specialized version of find to get last valid node.
1222 dl 1.1 * @return last node or null if empty
1223     */
1224     Node<K,V> findLast() {
1225     /*
1226     * findPredecessor can't be used to traverse index level
1227     * because this doesn't use comparisons. So traversals of
1228     * both levels are folded together.
1229     */
1230     Index<K,V> q = head;
1231     for (;;) {
1232     Index<K,V> d, r;
1233     if ((r = q.right) != null) {
1234     if (r.indexesDeletedNode()) {
1235     q.unlink(r);
1236     q = head; // restart
1237 dl 1.9 }
1238 dl 1.1 else
1239     q = r;
1240     } else if ((d = q.down) != null) {
1241     q = d;
1242     } else {
1243     Node<K,V> b = q.node;
1244     Node<K,V> n = b.next;
1245     for (;;) {
1246 dl 1.9 if (n == null)
1247 jsr166 1.61 return b.isBaseHeader() ? null : b;
1248 dl 1.1 Node<K,V> f = n.next; // inconsistent read
1249     if (n != b.next)
1250     break;
1251     Object v = n.value;
1252     if (v == null) { // n is deleted
1253     n.helpDelete(b, f);
1254     break;
1255     }
1256     if (v == n || b.value == null) // b is deleted
1257     break;
1258     b = n;
1259     n = f;
1260     }
1261     q = head; // restart
1262     }
1263     }
1264     }
1265    
1266 dl 1.31 /**
1267 jsr166 1.32 * Specialized variant of findPredecessor to get predecessor of last
1268     * valid node. Needed when removing the last entry. It is possible
1269     * that all successors of returned node will have been deleted upon
1270 dl 1.31 * return, in which case this method can be retried.
1271     * @return likely predecessor of last node
1272     */
1273     private Node<K,V> findPredecessorOfLast() {
1274     for (;;) {
1275     Index<K,V> q = head;
1276     for (;;) {
1277     Index<K,V> d, r;
1278     if ((r = q.right) != null) {
1279     if (r.indexesDeletedNode()) {
1280     q.unlink(r);
1281     break; // must restart
1282     }
1283     // proceed as far across as possible without overshooting
1284     if (r.node.next != null) {
1285     q = r;
1286     continue;
1287     }
1288     }
1289     if ((d = q.down) != null)
1290     q = d;
1291     else
1292     return q.node;
1293     }
1294     }
1295     }
1296 dl 1.1
1297 dl 1.88 /* ---------------- Relational operations -------------- */
1298    
1299     // Control values OR'ed as arguments to findNear
1300    
1301     private static final int EQ = 1;
1302     private static final int LT = 2;
1303     private static final int GT = 0; // Actually checked as !LT
1304    
1305 dl 1.1 /**
1306 dl 1.88 * Utility for ceiling, floor, lower, higher methods.
1307     * @param kkey the key
1308     * @param rel the relation -- OR'ed combination of EQ, LT, GT
1309     * @return nearest node fitting relation, or null if no such
1310 dl 1.1 */
1311 dl 1.88 Node<K,V> doFindNear(K kkey, int rel) {
1312     @SuppressWarnings("unchecked") Comparable<? super K> key =
1313     (Comparable<? super K>)kkey;
1314     for (;;) {
1315     Node<K,V> b = findPredecessor(key);
1316     Node<K,V> n = b.next;
1317     for (;;) {
1318     if (n == null)
1319     return ((rel & LT) == 0 || b.isBaseHeader()) ? null : b;
1320     Node<K,V> f = n.next;
1321     if (n != b.next) // inconsistent read
1322     break;
1323     Object v = n.value;
1324     if (v == null) { // n is deleted
1325     n.helpDelete(b, f);
1326     break;
1327     }
1328     if (v == n || b.value == null) // b is deleted
1329     break;
1330     int c = key.compareTo(n.key);
1331     if ((c == 0 && (rel & EQ) != 0) ||
1332     (c < 0 && (rel & LT) == 0))
1333     return n;
1334     if ( c <= 0 && (rel & LT) != 0)
1335     return b.isBaseHeader() ? null : b;
1336     b = n;
1337     n = f;
1338     }
1339     }
1340     }
1341    
1342     /* ---------------- cmp versions -------------- */
1343    
1344     // Boringly almost the same as natural-Comparable ones
1345    
1346     private Node<K,V> findPredecessorCmp(Comparator<? super K> cmp, Object okey) {
1347     if (cmp == null)
1348     throw new NullPointerException(); // don't postpone errors
1349     @SuppressWarnings("unchecked") K key = (K) okey;
1350     for (;;) {
1351     Index<K,V> q = head;
1352     Index<K,V> r = q.right;
1353     for (;;) {
1354     if (r != null) {
1355     Node<K,V> n = r.node;
1356     K k = n.key;
1357     if (n.value == null) {
1358     if (!q.unlink(r))
1359     break; // restart
1360     r = q.right; // reread r
1361     continue;
1362     }
1363     if (cmp.compare(key, k) > 0) {
1364     q = r;
1365     r = r.right;
1366     continue;
1367     }
1368     }
1369     Index<K,V> d = q.down;
1370     if (d != null) {
1371     q = d;
1372     r = d.right;
1373     } else
1374     return q.node;
1375     }
1376     }
1377     }
1378    
1379     private Node<K,V> findNodeCmp(Comparator<? super K> cmp, Object okey) {
1380     if (cmp == null)
1381     throw new NullPointerException(); // don't postpone errors
1382     @SuppressWarnings("unchecked") K key = (K) okey;
1383     for (;;) {
1384     Node<K,V> b = findPredecessorCmp(cmp, key);
1385     Node<K,V> n = b.next;
1386     for (;;) {
1387     if (n == null)
1388     return null;
1389     Node<K,V> f = n.next;
1390     if (n != b.next) // inconsistent read
1391     break;
1392     Object v = n.value;
1393     if (v == null) { // n is deleted
1394     n.helpDelete(b, f);
1395     break;
1396     }
1397     if (v == n || b.value == null) // b is deleted
1398     break;
1399     int c = cmp.compare(key, n.key);
1400     if (c == 0)
1401     return n;
1402     if (c < 0)
1403     return null;
1404     b = n;
1405     n = f;
1406     }
1407     }
1408     }
1409    
1410     private V doGetCmp(Comparator<? super K> cmp, Object okey) {
1411     if (cmp == null)
1412     throw new NullPointerException(); // don't postpone errors
1413     @SuppressWarnings("unchecked") K key = (K) okey;
1414 dl 1.1 for (;;) {
1415 dl 1.88 Node<K,V> b = findPredecessorCmp(cmp, key);
1416 dl 1.31 Node<K,V> n = b.next;
1417 dl 1.88 for (;;) {
1418     if (n == null)
1419     return null;
1420     Node<K,V> f = n.next;
1421     if (n != b.next) // inconsistent read
1422     break;
1423     Object v = n.value;
1424     if (v == null) { // n is deleted
1425     n.helpDelete(b, f);
1426     break;
1427     }
1428     if (v == n || b.value == null) // b is deleted
1429     break;
1430     int c = cmp.compare(key, n.key);
1431     if (c == 0)
1432     return (V)v;
1433     if (c < 0)
1434 dl 1.31 return null;
1435 dl 1.88 b = n;
1436     n = f;
1437     }
1438     }
1439     }
1440    
1441     private V doPutCmp(Comparator<? super K> cmp, K key, V value, boolean onlyIfAbsent) {
1442     if (cmp == null)
1443     throw new NullPointerException(); // don't postpone errors
1444     for (;;) {
1445     Node<K,V> b = findPredecessorCmp(cmp, key);
1446     Node<K,V> n = b.next;
1447     for (;;) {
1448     if (n != null) {
1449     Node<K,V> f = n.next;
1450     if (n != b.next) // inconsistent read
1451     break;
1452     Object v = n.value;
1453     if (v == null) { // n is deleted
1454     n.helpDelete(b, f);
1455     break;
1456     }
1457     if (v == n || b.value == null) // b is deleted
1458     break;
1459     int c = cmp.compare(key, n.key);
1460     if (c > 0) {
1461     b = n;
1462     n = f;
1463     continue;
1464     }
1465     if (c == 0) {
1466     if (onlyIfAbsent || n.casValue(v, value))
1467     return (V)v;
1468     else
1469     break; // restart if lost race to replace value
1470     }
1471     // else c < 0; fall through
1472     }
1473    
1474     Node<K,V> z = new Node<K,V>(key, value, n);
1475     if (!b.casNext(n, z))
1476     break; // restart if lost race to append to b
1477     int level = randomLevel();
1478     if (level > 0)
1479     insertIndex(z, level);
1480     return null;
1481 dl 1.31 }
1482 dl 1.88 }
1483     }
1484    
1485     final V doRemoveCmp(Comparator<? super K> cmp, Object okey, Object value) {
1486     if (cmp == null)
1487     throw new NullPointerException(); // don't postpone errors
1488     @SuppressWarnings("unchecked") K key = (K) okey;
1489     for (;;) {
1490     Node<K,V> b = findPredecessorCmp(cmp, key);
1491     Node<K,V> n = b.next;
1492 dl 1.1 for (;;) {
1493 dl 1.88 if (n == null)
1494     return null;
1495 dl 1.31 Node<K,V> f = n.next;
1496     if (n != b.next) // inconsistent read
1497     break;
1498     Object v = n.value;
1499     if (v == null) { // n is deleted
1500     n.helpDelete(b, f);
1501     break;
1502     }
1503     if (v == n || b.value == null) // b is deleted
1504     break;
1505 dl 1.88 int c = cmp.compare(key, n.key);
1506     if (c < 0)
1507     return null;
1508     if (c > 0) {
1509 dl 1.31 b = n;
1510     n = f;
1511     continue;
1512     }
1513 dl 1.88 if (value != null && !value.equals(v))
1514     return null;
1515 dl 1.31 if (!n.casValue(v, null))
1516     break;
1517     if (!n.appendMarker(f) || !b.casNext(n, f))
1518 dl 1.88 findNodeCmp(cmp, key); // Retry via findNode
1519 dl 1.31 else {
1520 dl 1.88 findPredecessorCmp(cmp, key); // Clean index
1521 dl 1.31 if (head.right == null)
1522     tryReduceLevel();
1523 dl 1.1 }
1524 dl 1.88 return (V)v;
1525 dl 1.1 }
1526     }
1527     }
1528    
1529 dl 1.88 Node<K,V> doFindNearCmp(Comparator<? super K> cmp, K key, int rel) {
1530     if (cmp == null)
1531     throw new NullPointerException(); // don't postpone errors
1532 dl 1.1 for (;;) {
1533 dl 1.88 Node<K,V> b = findPredecessorCmp(cmp, key);
1534 dl 1.1 Node<K,V> n = b.next;
1535     for (;;) {
1536 dl 1.9 if (n == null)
1537 jsr166 1.61 return ((rel & LT) == 0 || b.isBaseHeader()) ? null : b;
1538 dl 1.1 Node<K,V> f = n.next;
1539     if (n != b.next) // inconsistent read
1540     break;
1541     Object v = n.value;
1542     if (v == null) { // n is deleted
1543     n.helpDelete(b, f);
1544     break;
1545     }
1546     if (v == n || b.value == null) // b is deleted
1547     break;
1548 dl 1.88 int c = cmp.compare(key, n.key);
1549 dl 1.1 if ((c == 0 && (rel & EQ) != 0) ||
1550     (c < 0 && (rel & LT) == 0))
1551     return n;
1552     if ( c <= 0 && (rel & LT) != 0)
1553 jsr166 1.61 return b.isBaseHeader() ? null : b;
1554 dl 1.1 b = n;
1555     n = f;
1556     }
1557     }
1558     }
1559    
1560 dl 1.88 /* ---------------- Relays to natural vs cmp methods -------------- */
1561    
1562     Node<K,V> findNear(K key, int rel) {
1563     Comparator<? super K> cmp;
1564     return (cmp = comparator) == null ? doFindNear(key, rel) :
1565     doFindNearCmp(cmp, key, rel);
1566     }
1567    
1568 dl 1.1 /**
1569 jsr166 1.10 * Returns SimpleImmutableEntry for results of findNear.
1570 dl 1.40 * @param key the key
1571 dl 1.1 * @param rel the relation -- OR'ed combination of EQ, LT, GT
1572     * @return Entry fitting relation, or null if no such
1573     */
1574 dl 1.40 AbstractMap.SimpleImmutableEntry<K,V> getNear(K key, int rel) {
1575 dl 1.1 for (;;) {
1576 dl 1.40 Node<K,V> n = findNear(key, rel);
1577 dl 1.1 if (n == null)
1578     return null;
1579 dl 1.2 AbstractMap.SimpleImmutableEntry<K,V> e = n.createSnapshot();
1580 dl 1.1 if (e != null)
1581     return e;
1582     }
1583     }
1584    
1585     /* ---------------- Constructors -------------- */
1586    
1587     /**
1588 jsr166 1.22 * Constructs a new, empty map, sorted according to the
1589     * {@linkplain Comparable natural ordering} of the keys.
1590 dl 1.1 */
1591     public ConcurrentSkipListMap() {
1592     this.comparator = null;
1593     initialize();
1594     }
1595    
1596     /**
1597 jsr166 1.22 * Constructs a new, empty map, sorted according to the specified
1598     * comparator.
1599 dl 1.1 *
1600 jsr166 1.22 * @param comparator the comparator that will be used to order this map.
1601 jsr166 1.82 * If {@code null}, the {@linkplain Comparable natural
1602 jsr166 1.22 * ordering} of the keys will be used.
1603 dl 1.1 */
1604 jsr166 1.22 public ConcurrentSkipListMap(Comparator<? super K> comparator) {
1605     this.comparator = comparator;
1606 dl 1.1 initialize();
1607     }
1608    
1609     /**
1610     * Constructs a new map containing the same mappings as the given map,
1611 jsr166 1.22 * sorted according to the {@linkplain Comparable natural ordering} of
1612     * the keys.
1613 dl 1.1 *
1614 jsr166 1.22 * @param m the map whose mappings are to be placed in this map
1615 jsr166 1.82 * @throws ClassCastException if the keys in {@code m} are not
1616 jsr166 1.22 * {@link Comparable}, or are not mutually comparable
1617     * @throws NullPointerException if the specified map or any of its keys
1618     * or values are null
1619 dl 1.1 */
1620     public ConcurrentSkipListMap(Map<? extends K, ? extends V> m) {
1621     this.comparator = null;
1622     initialize();
1623     putAll(m);
1624     }
1625    
1626     /**
1627 jsr166 1.22 * Constructs a new map containing the same mappings and using the
1628     * same ordering as the specified sorted map.
1629     *
1630 dl 1.1 * @param m the sorted map whose mappings are to be placed in this
1631 jsr166 1.22 * map, and whose comparator is to be used to sort this map
1632     * @throws NullPointerException if the specified sorted map or any of
1633     * its keys or values are null
1634 dl 1.1 */
1635     public ConcurrentSkipListMap(SortedMap<K, ? extends V> m) {
1636     this.comparator = m.comparator();
1637     initialize();
1638     buildFromSorted(m);
1639     }
1640    
1641     /**
1642 dl 1.83 * Creates a new {@link Set} backed by a ConcurrentSkipListMap
1643     * from the given type to {@code Boolean.TRUE}.
1644     *
1645     * @return the new set
1646     */
1647     public static <K> KeySetView<K,Boolean> newKeySet() {
1648     return new KeySetView<K,Boolean>(new ConcurrentSkipListMap<K,Boolean>(),
1649     Boolean.TRUE);
1650     }
1651    
1652     /**
1653     * Creates a new {@link Set} backed by a ConcurrentSkipListMap
1654     * from the given type to {@code Boolean.TRUE}, using the
1655 jsr166 1.87 * given comparator.
1656 dl 1.83 *
1657     * @param comparator the comparator that will be used to order this map.
1658 jsr166 1.86 * If {@code null}, the {@linkplain Comparable natural
1659 dl 1.83 * ordering} of the keys will be used.
1660     *
1661     * @return the new set
1662     */
1663     public static <K> KeySetView<K,Boolean> newKeySet(Comparator<? super K> comparator) {
1664     return new KeySetView<K,Boolean>
1665     (new ConcurrentSkipListMap<K,Boolean>(comparator), Boolean.TRUE);
1666     }
1667    
1668     /**
1669 jsr166 1.82 * Returns a shallow copy of this {@code ConcurrentSkipListMap}
1670 jsr166 1.22 * instance. (The keys and values themselves are not cloned.)
1671 dl 1.1 *
1672 jsr166 1.22 * @return a shallow copy of this map
1673 dl 1.1 */
1674 jsr166 1.16 public ConcurrentSkipListMap<K,V> clone() {
1675 dl 1.1 try {
1676 jsr166 1.76 @SuppressWarnings("unchecked")
1677     ConcurrentSkipListMap<K,V> clone =
1678     (ConcurrentSkipListMap<K,V>) super.clone();
1679     clone.initialize();
1680     clone.buildFromSorted(this);
1681     return clone;
1682 dl 1.1 } catch (CloneNotSupportedException e) {
1683     throw new InternalError();
1684     }
1685     }
1686    
1687     /**
1688     * Streamlined bulk insertion to initialize from elements of
1689     * given sorted map. Call only from constructor or clone
1690     * method.
1691     */
1692     private void buildFromSorted(SortedMap<K, ? extends V> map) {
1693     if (map == null)
1694     throw new NullPointerException();
1695    
1696     HeadIndex<K,V> h = head;
1697     Node<K,V> basepred = h.node;
1698    
1699     // Track the current rightmost node at each level. Uses an
1700     // ArrayList to avoid committing to initial or maximum level.
1701     ArrayList<Index<K,V>> preds = new ArrayList<Index<K,V>>();
1702    
1703     // initialize
1704 dl 1.9 for (int i = 0; i <= h.level; ++i)
1705 dl 1.1 preds.add(null);
1706     Index<K,V> q = h;
1707     for (int i = h.level; i > 0; --i) {
1708     preds.set(i, q);
1709     q = q.down;
1710     }
1711    
1712 dl 1.9 Iterator<? extends Map.Entry<? extends K, ? extends V>> it =
1713 dl 1.1 map.entrySet().iterator();
1714     while (it.hasNext()) {
1715     Map.Entry<? extends K, ? extends V> e = it.next();
1716     int j = randomLevel();
1717     if (j > h.level) j = h.level + 1;
1718     K k = e.getKey();
1719     V v = e.getValue();
1720     if (k == null || v == null)
1721     throw new NullPointerException();
1722     Node<K,V> z = new Node<K,V>(k, v, null);
1723     basepred.next = z;
1724     basepred = z;
1725     if (j > 0) {
1726     Index<K,V> idx = null;
1727     for (int i = 1; i <= j; ++i) {
1728     idx = new Index<K,V>(z, idx, null);
1729 dl 1.9 if (i > h.level)
1730 dl 1.1 h = new HeadIndex<K,V>(h.node, h, idx, i);
1731    
1732     if (i < preds.size()) {
1733     preds.get(i).right = idx;
1734     preds.set(i, idx);
1735     } else
1736     preds.add(idx);
1737     }
1738     }
1739     }
1740     head = h;
1741     }
1742    
1743     /* ---------------- Serialization -------------- */
1744    
1745     /**
1746 jsr166 1.80 * Saves this map to a stream (that is, serializes it).
1747 dl 1.1 *
1748     * @serialData The key (Object) and value (Object) for each
1749 jsr166 1.10 * key-value mapping represented by the map, followed by
1750 jsr166 1.82 * {@code null}. The key-value mappings are emitted in key-order
1751 dl 1.1 * (as determined by the Comparator, or by the keys' natural
1752     * ordering if no Comparator).
1753     */
1754     private void writeObject(java.io.ObjectOutputStream s)
1755     throws java.io.IOException {
1756     // Write out the Comparator and any hidden stuff
1757     s.defaultWriteObject();
1758    
1759     // Write out keys and values (alternating)
1760     for (Node<K,V> n = findFirst(); n != null; n = n.next) {
1761     V v = n.getValidValue();
1762     if (v != null) {
1763     s.writeObject(n.key);
1764     s.writeObject(v);
1765     }
1766     }
1767     s.writeObject(null);
1768     }
1769    
1770     /**
1771 jsr166 1.80 * Reconstitutes this map from a stream (that is, deserializes it).
1772 dl 1.1 */
1773     private void readObject(final java.io.ObjectInputStream s)
1774     throws java.io.IOException, ClassNotFoundException {
1775     // Read in the Comparator and any hidden stuff
1776     s.defaultReadObject();
1777     // Reset transients
1778     initialize();
1779    
1780 dl 1.9 /*
1781 dl 1.1 * This is nearly identical to buildFromSorted, but is
1782     * distinct because readObject calls can't be nicely adapted
1783     * as the kind of iterator needed by buildFromSorted. (They
1784     * can be, but doing so requires type cheats and/or creation
1785     * of adaptor classes.) It is simpler to just adapt the code.
1786     */
1787    
1788     HeadIndex<K,V> h = head;
1789     Node<K,V> basepred = h.node;
1790     ArrayList<Index<K,V>> preds = new ArrayList<Index<K,V>>();
1791 dl 1.9 for (int i = 0; i <= h.level; ++i)
1792 dl 1.1 preds.add(null);
1793     Index<K,V> q = h;
1794     for (int i = h.level; i > 0; --i) {
1795     preds.set(i, q);
1796     q = q.down;
1797     }
1798    
1799     for (;;) {
1800     Object k = s.readObject();
1801     if (k == null)
1802     break;
1803     Object v = s.readObject();
1804 dl 1.9 if (v == null)
1805 dl 1.1 throw new NullPointerException();
1806     K key = (K) k;
1807     V val = (V) v;
1808     int j = randomLevel();
1809     if (j > h.level) j = h.level + 1;
1810     Node<K,V> z = new Node<K,V>(key, val, null);
1811     basepred.next = z;
1812     basepred = z;
1813     if (j > 0) {
1814     Index<K,V> idx = null;
1815     for (int i = 1; i <= j; ++i) {
1816     idx = new Index<K,V>(z, idx, null);
1817 dl 1.9 if (i > h.level)
1818 dl 1.1 h = new HeadIndex<K,V>(h.node, h, idx, i);
1819    
1820     if (i < preds.size()) {
1821     preds.get(i).right = idx;
1822     preds.set(i, idx);
1823     } else
1824     preds.add(idx);
1825     }
1826     }
1827     }
1828     head = h;
1829     }
1830    
1831     /* ------ Map API methods ------ */
1832    
1833     /**
1834 jsr166 1.82 * Returns {@code true} if this map contains a mapping for the specified
1835 dl 1.1 * key.
1836 jsr166 1.22 *
1837     * @param key key whose presence in this map is to be tested
1838 jsr166 1.82 * @return {@code true} if this map contains a mapping for the specified key
1839 jsr166 1.22 * @throws ClassCastException if the specified key cannot be compared
1840     * with the keys currently in the map
1841     * @throws NullPointerException if the specified key is null
1842 dl 1.1 */
1843     public boolean containsKey(Object key) {
1844 dl 1.88 Comparator<? super K> cmp;
1845     Object v = ((cmp = comparator) == null ? doGet(key) :
1846     doGetCmp(cmp, key));
1847     return v != null;
1848 dl 1.1 }
1849    
1850     /**
1851 jsr166 1.42 * Returns the value to which the specified key is mapped,
1852     * or {@code null} if this map contains no mapping for the key.
1853     *
1854     * <p>More formally, if this map contains a mapping from a key
1855     * {@code k} to a value {@code v} such that {@code key} compares
1856     * equal to {@code k} according to the map's ordering, then this
1857     * method returns {@code v}; otherwise it returns {@code null}.
1858     * (There can be at most one such mapping.)
1859 dl 1.1 *
1860 jsr166 1.22 * @throws ClassCastException if the specified key cannot be compared
1861     * with the keys currently in the map
1862     * @throws NullPointerException if the specified key is null
1863 dl 1.1 */
1864     public V get(Object key) {
1865 dl 1.88 Comparator<? super K> cmp;
1866     return ((cmp = comparator) == null) ? doGet(key) : doGetCmp(cmp, key);
1867 dl 1.1 }
1868    
1869     /**
1870     * Associates the specified value with the specified key in this map.
1871 jsr166 1.22 * If the map previously contained a mapping for the key, the old
1872 dl 1.1 * value is replaced.
1873     *
1874 jsr166 1.22 * @param key key with which the specified value is to be associated
1875     * @param value value to be associated with the specified key
1876     * @return the previous value associated with the specified key, or
1877 jsr166 1.82 * {@code null} if there was no mapping for the key
1878 jsr166 1.22 * @throws ClassCastException if the specified key cannot be compared
1879     * with the keys currently in the map
1880     * @throws NullPointerException if the specified key or value is null
1881 dl 1.1 */
1882     public V put(K key, V value) {
1883 dl 1.88 Comparator<? super K> cmp;
1884 dl 1.9 if (value == null)
1885 dl 1.1 throw new NullPointerException();
1886 dl 1.88 return ((cmp = comparator) == null) ?
1887     doPut(key, value, false) : doPutCmp(cmp, key, value, false);
1888 dl 1.1 }
1889    
1890     /**
1891 jsr166 1.36 * Removes the mapping for the specified key from this map if present.
1892 dl 1.1 *
1893     * @param key key for which mapping should be removed
1894 jsr166 1.22 * @return the previous value associated with the specified key, or
1895 jsr166 1.82 * {@code null} if there was no mapping for the key
1896 jsr166 1.22 * @throws ClassCastException if the specified key cannot be compared
1897     * with the keys currently in the map
1898     * @throws NullPointerException if the specified key is null
1899 dl 1.1 */
1900     public V remove(Object key) {
1901 dl 1.88 Comparator<? super K> cmp;
1902     return ((cmp = comparator) == null) ? doRemove(key, null) :
1903     doRemoveCmp(cmp, key, null);
1904 dl 1.1 }
1905    
1906     /**
1907 jsr166 1.82 * Returns {@code true} if this map maps one or more keys to the
1908 dl 1.1 * specified value. This operation requires time linear in the
1909 dl 1.69 * map size. Additionally, it is possible for the map to change
1910     * during execution of this method, in which case the returned
1911     * result may be inaccurate.
1912 dl 1.1 *
1913 jsr166 1.22 * @param value value whose presence in this map is to be tested
1914 jsr166 1.82 * @return {@code true} if a mapping to {@code value} exists;
1915     * {@code false} otherwise
1916 jsr166 1.22 * @throws NullPointerException if the specified value is null
1917 dl 1.9 */
1918 dl 1.1 public boolean containsValue(Object value) {
1919 dl 1.9 if (value == null)
1920 dl 1.1 throw new NullPointerException();
1921     for (Node<K,V> n = findFirst(); n != null; n = n.next) {
1922     V v = n.getValidValue();
1923     if (v != null && value.equals(v))
1924     return true;
1925     }
1926     return false;
1927     }
1928    
1929     /**
1930 dl 1.6 * Returns the number of key-value mappings in this map. If this map
1931 jsr166 1.82 * contains more than {@code Integer.MAX_VALUE} elements, it
1932     * returns {@code Integer.MAX_VALUE}.
1933 dl 1.1 *
1934     * <p>Beware that, unlike in most collections, this method is
1935     * <em>NOT</em> a constant-time operation. Because of the
1936     * asynchronous nature of these maps, determining the current
1937     * number of elements requires traversing them all to count them.
1938     * Additionally, it is possible for the size to change during
1939     * execution of this method, in which case the returned result
1940     * will be inaccurate. Thus, this method is typically not very
1941     * useful in concurrent applications.
1942     *
1943 jsr166 1.22 * @return the number of elements in this map
1944 dl 1.1 */
1945     public int size() {
1946     long count = 0;
1947     for (Node<K,V> n = findFirst(); n != null; n = n.next) {
1948     if (n.getValidValue() != null)
1949     ++count;
1950     }
1951 jsr166 1.61 return (count >= Integer.MAX_VALUE) ? Integer.MAX_VALUE : (int) count;
1952 dl 1.1 }
1953    
1954     /**
1955 jsr166 1.82 * Returns {@code true} if this map contains no key-value mappings.
1956     * @return {@code true} if this map contains no key-value mappings
1957 dl 1.1 */
1958     public boolean isEmpty() {
1959     return findFirst() == null;
1960     }
1961    
1962     /**
1963 jsr166 1.22 * Removes all of the mappings from this map.
1964 dl 1.1 */
1965     public void clear() {
1966     initialize();
1967     }
1968    
1969 dl 1.46 /* ---------------- View methods -------------- */
1970    
1971     /*
1972     * Note: Lazy initialization works for views because view classes
1973     * are stateless/immutable so it doesn't matter wrt correctness if
1974     * more than one is created (which will only rarely happen). Even
1975     * so, the following idiom conservatively ensures that the method
1976     * returns the one it created if it does so, not one created by
1977     * another racing thread.
1978     */
1979    
1980 dl 1.1 /**
1981 jsr166 1.51 * Returns a {@link NavigableSet} view of the keys contained in this map.
1982 jsr166 1.22 * The set's iterator returns the keys in ascending order.
1983     * The set is backed by the map, so changes to the map are
1984     * reflected in the set, and vice-versa. The set supports element
1985     * removal, which removes the corresponding mapping from the map,
1986 jsr166 1.51 * via the {@code Iterator.remove}, {@code Set.remove},
1987     * {@code removeAll}, {@code retainAll}, and {@code clear}
1988     * operations. It does not support the {@code add} or {@code addAll}
1989 jsr166 1.22 * operations.
1990     *
1991 jsr166 1.51 * <p>The view's {@code iterator} is a "weakly consistent" iterator
1992 jsr166 1.22 * that will never throw {@link ConcurrentModificationException},
1993 dl 1.1 * and guarantees to traverse elements as they existed upon
1994     * construction of the iterator, and may (but is not guaranteed to)
1995     * reflect any modifications subsequent to construction.
1996     *
1997 jsr166 1.51 * <p>This method is equivalent to method {@code navigableKeySet}.
1998     *
1999     * @return a navigable set view of the keys in this map
2000 dl 1.1 */
2001 jsr166 1.68 public NavigableSet<K> keySet() {
2002 dl 1.83 KeySetView<K,V> ks = keySet;
2003     return (ks != null) ? ks : (keySet = new KeySetView<K,V>(this, null));
2004 dl 1.1 }
2005    
2006 dl 1.46 public NavigableSet<K> navigableKeySet() {
2007 dl 1.83 KeySetView<K,V> ks = keySet;
2008     return (ks != null) ? ks : (keySet = new KeySetView<K,V>(this, null));
2009     }
2010    
2011     /**
2012     * Returns a {@link Set} view of the keys in this map, using the
2013     * given common mapped value for any additions (i.e., {@link
2014     * Collection#add} and {@link Collection#addAll}). This is of
2015     * course only appropriate if it is acceptable to use the same
2016     * value for all additions from this view.
2017     *
2018     * @param mappedValue the mapped value to use for any
2019     * additions.
2020     * @return the set view
2021     * @throws NullPointerException if the mappedValue is null
2022     */
2023     public KeySetView<K,V> keySet(V mappedValue) {
2024     if (mappedValue == null)
2025     throw new NullPointerException();
2026     return new KeySetView<K,V>(this, mappedValue);
2027 dl 1.1 }
2028    
2029     /**
2030 jsr166 1.22 * Returns a {@link Collection} view of the values contained in this map.
2031     * The collection's iterator returns the values in ascending order
2032     * of the corresponding keys.
2033 dl 1.1 * The collection is backed by the map, so changes to the map are
2034     * reflected in the collection, and vice-versa. The collection
2035     * supports element removal, which removes the corresponding
2036 jsr166 1.82 * mapping from the map, via the {@code Iterator.remove},
2037     * {@code Collection.remove}, {@code removeAll},
2038     * {@code retainAll} and {@code clear} operations. It does not
2039     * support the {@code add} or {@code addAll} operations.
2040 dl 1.1 *
2041 jsr166 1.82 * <p>The view's {@code iterator} is a "weakly consistent" iterator
2042 jsr166 1.22 * that will never throw {@link ConcurrentModificationException},
2043     * and guarantees to traverse elements as they existed upon
2044     * construction of the iterator, and may (but is not guaranteed to)
2045     * reflect any modifications subsequent to construction.
2046 dl 1.1 */
2047     public Collection<V> values() {
2048 jsr166 1.71 Values<V> vs = values;
2049     return (vs != null) ? vs : (values = new Values<V>(this));
2050 dl 1.1 }
2051    
2052     /**
2053 jsr166 1.22 * Returns a {@link Set} view of the mappings contained in this map.
2054     * The set's iterator returns the entries in ascending key order.
2055     * The set is backed by the map, so changes to the map are
2056     * reflected in the set, and vice-versa. The set supports element
2057     * removal, which removes the corresponding mapping from the map,
2058 jsr166 1.82 * via the {@code Iterator.remove}, {@code Set.remove},
2059     * {@code removeAll}, {@code retainAll} and {@code clear}
2060     * operations. It does not support the {@code add} or
2061     * {@code addAll} operations.
2062 jsr166 1.22 *
2063 jsr166 1.82 * <p>The view's {@code iterator} is a "weakly consistent" iterator
2064 jsr166 1.22 * that will never throw {@link ConcurrentModificationException},
2065     * and guarantees to traverse elements as they existed upon
2066     * construction of the iterator, and may (but is not guaranteed to)
2067     * reflect any modifications subsequent to construction.
2068     *
2069 jsr166 1.82 * <p>The {@code Map.Entry} elements returned by
2070     * {@code iterator.next()} do <em>not</em> support the
2071     * {@code setValue} operation.
2072 dl 1.1 *
2073 jsr166 1.22 * @return a set view of the mappings contained in this map,
2074     * sorted in ascending key order
2075 dl 1.1 */
2076     public Set<Map.Entry<K,V>> entrySet() {
2077 jsr166 1.71 EntrySet<K,V> es = entrySet;
2078     return (es != null) ? es : (entrySet = new EntrySet<K,V>(this));
2079 dl 1.46 }
2080    
2081     public ConcurrentNavigableMap<K,V> descendingMap() {
2082     ConcurrentNavigableMap<K,V> dm = descendingMap;
2083     return (dm != null) ? dm : (descendingMap = new SubMap<K,V>
2084     (this, null, false, null, false, true));
2085 dl 1.1 }
2086    
2087 dl 1.46 public NavigableSet<K> descendingKeySet() {
2088     return descendingMap().navigableKeySet();
2089 dl 1.1 }
2090    
2091     /* ---------------- AbstractMap Overrides -------------- */
2092    
2093     /**
2094     * Compares the specified object with this map for equality.
2095 jsr166 1.82 * Returns {@code true} if the given object is also a map and the
2096 dl 1.1 * two maps represent the same mappings. More formally, two maps
2097 jsr166 1.82 * {@code m1} and {@code m2} represent the same mappings if
2098     * {@code m1.entrySet().equals(m2.entrySet())}. This
2099 dl 1.1 * operation may return misleading results if either map is
2100     * concurrently modified during execution of this method.
2101     *
2102 jsr166 1.22 * @param o object to be compared for equality with this map
2103 jsr166 1.82 * @return {@code true} if the specified object is equal to this map
2104 dl 1.1 */
2105     public boolean equals(Object o) {
2106 jsr166 1.55 if (o == this)
2107     return true;
2108     if (!(o instanceof Map))
2109     return false;
2110     Map<?,?> m = (Map<?,?>) o;
2111 dl 1.1 try {
2112 jsr166 1.55 for (Map.Entry<K,V> e : this.entrySet())
2113     if (! e.getValue().equals(m.get(e.getKey())))
2114 dl 1.25 return false;
2115 jsr166 1.55 for (Map.Entry<?,?> e : m.entrySet()) {
2116 dl 1.25 Object k = e.getKey();
2117     Object v = e.getValue();
2118 jsr166 1.55 if (k == null || v == null || !v.equals(get(k)))
2119 dl 1.25 return false;
2120     }
2121     return true;
2122 jsr166 1.15 } catch (ClassCastException unused) {
2123 dl 1.1 return false;
2124 jsr166 1.15 } catch (NullPointerException unused) {
2125 dl 1.1 return false;
2126     }
2127     }
2128    
2129     /* ------ ConcurrentMap API methods ------ */
2130    
2131     /**
2132 jsr166 1.22 * {@inheritDoc}
2133     *
2134     * @return the previous value associated with the specified key,
2135 jsr166 1.82 * or {@code null} if there was no mapping for the key
2136 jsr166 1.22 * @throws ClassCastException if the specified key cannot be compared
2137     * with the keys currently in the map
2138     * @throws NullPointerException if the specified key or value is null
2139 dl 1.1 */
2140     public V putIfAbsent(K key, V value) {
2141 dl 1.88 Comparator<? super K> cmp;
2142 dl 1.9 if (value == null)
2143 dl 1.1 throw new NullPointerException();
2144 dl 1.88 return ((cmp = comparator) == null) ?
2145     doPut(key, value, true) : doPutCmp(cmp, key, value, true);
2146 dl 1.1 }
2147    
2148     /**
2149 jsr166 1.22 * {@inheritDoc}
2150     *
2151     * @throws ClassCastException if the specified key cannot be compared
2152     * with the keys currently in the map
2153 dl 1.23 * @throws NullPointerException if the specified key is null
2154 dl 1.1 */
2155     public boolean remove(Object key, Object value) {
2156 dl 1.45 if (key == null)
2157     throw new NullPointerException();
2158 dl 1.9 if (value == null)
2159 dl 1.23 return false;
2160 dl 1.88 Comparator<? super K> cmp;
2161     Object v = ((cmp = comparator) == null) ? doRemove(key, value) :
2162     doRemoveCmp(cmp, key, value);
2163     return v != null;
2164 dl 1.1 }
2165    
2166     /**
2167 jsr166 1.22 * {@inheritDoc}
2168     *
2169     * @throws ClassCastException if the specified key cannot be compared
2170     * with the keys currently in the map
2171     * @throws NullPointerException if any of the arguments are null
2172 dl 1.1 */
2173     public boolean replace(K key, V oldValue, V newValue) {
2174 dl 1.9 if (oldValue == null || newValue == null)
2175 dl 1.1 throw new NullPointerException();
2176 dl 1.88 Comparator<? super K> cmp = comparator;
2177 dl 1.1 for (;;) {
2178 jsr166 1.89 Node<K,V> n = (cmp == null) ?
2179     findNode((Comparable<? super K>)key) :
2180 dl 1.88 findNodeCmp(cmp, key);
2181 dl 1.1 if (n == null)
2182     return false;
2183     Object v = n.value;
2184     if (v != null) {
2185     if (!oldValue.equals(v))
2186     return false;
2187     if (n.casValue(v, newValue))
2188     return true;
2189     }
2190     }
2191     }
2192    
2193     /**
2194 jsr166 1.22 * {@inheritDoc}
2195     *
2196     * @return the previous value associated with the specified key,
2197 jsr166 1.82 * or {@code null} if there was no mapping for the key
2198 jsr166 1.22 * @throws ClassCastException if the specified key cannot be compared
2199     * with the keys currently in the map
2200     * @throws NullPointerException if the specified key or value is null
2201 dl 1.1 */
2202     public V replace(K key, V value) {
2203 dl 1.9 if (value == null)
2204 dl 1.1 throw new NullPointerException();
2205 dl 1.88 Comparator<? super K> cmp = comparator;
2206 dl 1.1 for (;;) {
2207 jsr166 1.89 Node<K,V> n = (cmp == null) ?
2208     findNode((Comparable<? super K>)key) :
2209 dl 1.88 findNodeCmp(cmp, key);
2210 dl 1.1 if (n == null)
2211     return null;
2212     Object v = n.value;
2213     if (v != null && n.casValue(v, value))
2214     return (V)v;
2215     }
2216     }
2217    
2218     /* ------ SortedMap API methods ------ */
2219    
2220     public Comparator<? super K> comparator() {
2221     return comparator;
2222     }
2223    
2224     /**
2225 jsr166 1.22 * @throws NoSuchElementException {@inheritDoc}
2226 dl 1.1 */
2227 dl 1.9 public K firstKey() {
2228 dl 1.1 Node<K,V> n = findFirst();
2229     if (n == null)
2230     throw new NoSuchElementException();
2231     return n.key;
2232     }
2233    
2234     /**
2235 jsr166 1.22 * @throws NoSuchElementException {@inheritDoc}
2236 dl 1.1 */
2237     public K lastKey() {
2238     Node<K,V> n = findLast();
2239     if (n == null)
2240     throw new NoSuchElementException();
2241     return n.key;
2242     }
2243    
2244     /**
2245 jsr166 1.49 * @throws ClassCastException {@inheritDoc}
2246     * @throws NullPointerException if {@code fromKey} or {@code toKey} is null
2247 jsr166 1.22 * @throws IllegalArgumentException {@inheritDoc}
2248 dl 1.1 */
2249 dl 1.47 public ConcurrentNavigableMap<K,V> subMap(K fromKey,
2250     boolean fromInclusive,
2251     K toKey,
2252     boolean toInclusive) {
2253 dl 1.1 if (fromKey == null || toKey == null)
2254     throw new NullPointerException();
2255 dl 1.46 return new SubMap<K,V>
2256     (this, fromKey, fromInclusive, toKey, toInclusive, false);
2257 dl 1.1 }
2258    
2259     /**
2260 jsr166 1.49 * @throws ClassCastException {@inheritDoc}
2261     * @throws NullPointerException if {@code toKey} is null
2262 jsr166 1.22 * @throws IllegalArgumentException {@inheritDoc}
2263 dl 1.1 */
2264 dl 1.47 public ConcurrentNavigableMap<K,V> headMap(K toKey,
2265     boolean inclusive) {
2266 dl 1.1 if (toKey == null)
2267     throw new NullPointerException();
2268 dl 1.46 return new SubMap<K,V>
2269     (this, null, false, toKey, inclusive, false);
2270 dl 1.1 }
2271    
2272     /**
2273 jsr166 1.49 * @throws ClassCastException {@inheritDoc}
2274     * @throws NullPointerException if {@code fromKey} is null
2275 jsr166 1.22 * @throws IllegalArgumentException {@inheritDoc}
2276 dl 1.1 */
2277 dl 1.47 public ConcurrentNavigableMap<K,V> tailMap(K fromKey,
2278     boolean inclusive) {
2279 dl 1.6 if (fromKey == null)
2280     throw new NullPointerException();
2281 dl 1.46 return new SubMap<K,V>
2282     (this, fromKey, inclusive, null, false, false);
2283 dl 1.6 }
2284    
2285     /**
2286 jsr166 1.49 * @throws ClassCastException {@inheritDoc}
2287     * @throws NullPointerException if {@code fromKey} or {@code toKey} is null
2288 jsr166 1.22 * @throws IllegalArgumentException {@inheritDoc}
2289 dl 1.6 */
2290 dl 1.37 public ConcurrentNavigableMap<K,V> subMap(K fromKey, K toKey) {
2291 dl 1.47 return subMap(fromKey, true, toKey, false);
2292 dl 1.6 }
2293    
2294     /**
2295 jsr166 1.49 * @throws ClassCastException {@inheritDoc}
2296     * @throws NullPointerException if {@code toKey} is null
2297 jsr166 1.22 * @throws IllegalArgumentException {@inheritDoc}
2298 dl 1.6 */
2299 dl 1.37 public ConcurrentNavigableMap<K,V> headMap(K toKey) {
2300 dl 1.47 return headMap(toKey, false);
2301 dl 1.6 }
2302    
2303     /**
2304 jsr166 1.49 * @throws ClassCastException {@inheritDoc}
2305     * @throws NullPointerException if {@code fromKey} is null
2306 jsr166 1.22 * @throws IllegalArgumentException {@inheritDoc}
2307 dl 1.6 */
2308 dl 1.37 public ConcurrentNavigableMap<K,V> tailMap(K fromKey) {
2309 dl 1.47 return tailMap(fromKey, true);
2310 dl 1.1 }
2311    
2312     /* ---------------- Relational operations -------------- */
2313    
2314     /**
2315 jsr166 1.22 * Returns a key-value mapping associated with the greatest key
2316 jsr166 1.82 * strictly less than the given key, or {@code null} if there is
2317 jsr166 1.22 * no such key. The returned entry does <em>not</em> support the
2318 jsr166 1.82 * {@code Entry.setValue} method.
2319 dl 1.9 *
2320 jsr166 1.22 * @throws ClassCastException {@inheritDoc}
2321     * @throws NullPointerException if the specified key is null
2322 dl 1.1 */
2323 jsr166 1.22 public Map.Entry<K,V> lowerEntry(K key) {
2324     return getNear(key, LT);
2325 dl 1.1 }
2326    
2327     /**
2328 jsr166 1.22 * @throws ClassCastException {@inheritDoc}
2329     * @throws NullPointerException if the specified key is null
2330 dl 1.1 */
2331 jsr166 1.22 public K lowerKey(K key) {
2332 dl 1.88 Comparator<? super K> cmp;
2333 jsr166 1.22 Node<K,V> n = findNear(key, LT);
2334 jsr166 1.61 return (n == null) ? null : n.key;
2335 dl 1.1 }
2336    
2337     /**
2338 jsr166 1.22 * Returns a key-value mapping associated with the greatest key
2339 jsr166 1.82 * less than or equal to the given key, or {@code null} if there
2340 jsr166 1.22 * is no such key. The returned entry does <em>not</em> support
2341 jsr166 1.82 * the {@code Entry.setValue} method.
2342 dl 1.9 *
2343 jsr166 1.22 * @param key the key
2344     * @throws ClassCastException {@inheritDoc}
2345     * @throws NullPointerException if the specified key is null
2346 dl 1.1 */
2347 jsr166 1.22 public Map.Entry<K,V> floorEntry(K key) {
2348     return getNear(key, LT|EQ);
2349 dl 1.1 }
2350    
2351     /**
2352 jsr166 1.22 * @param key the key
2353     * @throws ClassCastException {@inheritDoc}
2354     * @throws NullPointerException if the specified key is null
2355 dl 1.1 */
2356 jsr166 1.22 public K floorKey(K key) {
2357     Node<K,V> n = findNear(key, LT|EQ);
2358 jsr166 1.61 return (n == null) ? null : n.key;
2359 dl 1.1 }
2360    
2361     /**
2362 jsr166 1.22 * Returns a key-value mapping associated with the least key
2363 jsr166 1.82 * greater than or equal to the given key, or {@code null} if
2364 jsr166 1.22 * there is no such entry. The returned entry does <em>not</em>
2365 jsr166 1.82 * support the {@code Entry.setValue} method.
2366 dl 1.9 *
2367 jsr166 1.22 * @throws ClassCastException {@inheritDoc}
2368     * @throws NullPointerException if the specified key is null
2369 dl 1.1 */
2370 jsr166 1.22 public Map.Entry<K,V> ceilingEntry(K key) {
2371     return getNear(key, GT|EQ);
2372 dl 1.1 }
2373    
2374     /**
2375 jsr166 1.22 * @throws ClassCastException {@inheritDoc}
2376     * @throws NullPointerException if the specified key is null
2377 dl 1.1 */
2378 jsr166 1.22 public K ceilingKey(K key) {
2379     Node<K,V> n = findNear(key, GT|EQ);
2380 jsr166 1.61 return (n == null) ? null : n.key;
2381 dl 1.1 }
2382    
2383     /**
2384     * Returns a key-value mapping associated with the least key
2385 jsr166 1.82 * strictly greater than the given key, or {@code null} if there
2386 jsr166 1.22 * is no such key. The returned entry does <em>not</em> support
2387 jsr166 1.82 * the {@code Entry.setValue} method.
2388 dl 1.9 *
2389 jsr166 1.22 * @param key the key
2390     * @throws ClassCastException {@inheritDoc}
2391     * @throws NullPointerException if the specified key is null
2392 dl 1.1 */
2393     public Map.Entry<K,V> higherEntry(K key) {
2394     return getNear(key, GT);
2395     }
2396    
2397     /**
2398 jsr166 1.22 * @param key the key
2399     * @throws ClassCastException {@inheritDoc}
2400     * @throws NullPointerException if the specified key is null
2401 dl 1.1 */
2402     public K higherKey(K key) {
2403     Node<K,V> n = findNear(key, GT);
2404 jsr166 1.61 return (n == null) ? null : n.key;
2405 dl 1.1 }
2406    
2407     /**
2408     * Returns a key-value mapping associated with the least
2409 jsr166 1.82 * key in this map, or {@code null} if the map is empty.
2410 dl 1.1 * The returned entry does <em>not</em> support
2411 jsr166 1.82 * the {@code Entry.setValue} method.
2412 dl 1.1 */
2413     public Map.Entry<K,V> firstEntry() {
2414     for (;;) {
2415     Node<K,V> n = findFirst();
2416 dl 1.9 if (n == null)
2417 dl 1.1 return null;
2418 dl 1.2 AbstractMap.SimpleImmutableEntry<K,V> e = n.createSnapshot();
2419 dl 1.1 if (e != null)
2420     return e;
2421     }
2422     }
2423    
2424     /**
2425     * Returns a key-value mapping associated with the greatest
2426 jsr166 1.82 * key in this map, or {@code null} if the map is empty.
2427 dl 1.1 * The returned entry does <em>not</em> support
2428 jsr166 1.82 * the {@code Entry.setValue} method.
2429 dl 1.1 */
2430     public Map.Entry<K,V> lastEntry() {
2431     for (;;) {
2432     Node<K,V> n = findLast();
2433 dl 1.9 if (n == null)
2434 dl 1.1 return null;
2435 dl 1.2 AbstractMap.SimpleImmutableEntry<K,V> e = n.createSnapshot();
2436 dl 1.1 if (e != null)
2437     return e;
2438     }
2439     }
2440    
2441     /**
2442     * Removes and returns a key-value mapping associated with
2443 jsr166 1.82 * the least key in this map, or {@code null} if the map is empty.
2444 dl 1.1 * The returned entry does <em>not</em> support
2445 jsr166 1.82 * the {@code Entry.setValue} method.
2446 dl 1.1 */
2447     public Map.Entry<K,V> pollFirstEntry() {
2448 dl 1.25 return doRemoveFirstEntry();
2449 dl 1.1 }
2450    
2451     /**
2452     * Removes and returns a key-value mapping associated with
2453 jsr166 1.82 * the greatest key in this map, or {@code null} if the map is empty.
2454 dl 1.1 * The returned entry does <em>not</em> support
2455 jsr166 1.82 * the {@code Entry.setValue} method.
2456 dl 1.1 */
2457     public Map.Entry<K,V> pollLastEntry() {
2458 dl 1.31 return doRemoveLastEntry();
2459 dl 1.1 }
2460    
2461    
2462     /* ---------------- Iterators -------------- */
2463    
2464     /**
2465 dl 1.46 * Base of iterator classes:
2466 dl 1.1 */
2467 dl 1.46 abstract class Iter<T> implements Iterator<T> {
2468 dl 1.1 /** the last node returned by next() */
2469 jsr166 1.52 Node<K,V> lastReturned;
2470 dl 1.1 /** the next node to return from next(); */
2471     Node<K,V> next;
2472 jsr166 1.55 /** Cache of next value field to maintain weak consistency */
2473     V nextValue;
2474 dl 1.1
2475 jsr166 1.13 /** Initializes ascending iterator for entire range. */
2476 dl 1.46 Iter() {
2477 dl 1.1 for (;;) {
2478 jsr166 1.55 next = findFirst();
2479 dl 1.1 if (next == null)
2480     break;
2481 jsr166 1.52 Object x = next.value;
2482     if (x != null && x != next) {
2483 jsr166 1.55 nextValue = (V) x;
2484 dl 1.1 break;
2485 jsr166 1.55 }
2486 dl 1.1 }
2487     }
2488    
2489 dl 1.46 public final boolean hasNext() {
2490     return next != null;
2491 dl 1.1 }
2492 dl 1.46
2493 jsr166 1.13 /** Advances next to higher entry. */
2494 dl 1.46 final void advance() {
2495 jsr166 1.54 if (next == null)
2496 dl 1.1 throw new NoSuchElementException();
2497 jsr166 1.55 lastReturned = next;
2498 dl 1.1 for (;;) {
2499 jsr166 1.55 next = next.next;
2500 dl 1.1 if (next == null)
2501     break;
2502 jsr166 1.52 Object x = next.value;
2503     if (x != null && x != next) {
2504 jsr166 1.55 nextValue = (V) x;
2505 dl 1.1 break;
2506 jsr166 1.55 }
2507 dl 1.1 }
2508     }
2509    
2510     public void remove() {
2511 jsr166 1.52 Node<K,V> l = lastReturned;
2512 dl 1.1 if (l == null)
2513     throw new IllegalStateException();
2514     // It would not be worth all of the overhead to directly
2515     // unlink from here. Using remove is fast enough.
2516     ConcurrentSkipListMap.this.remove(l.key);
2517 jsr166 1.55 lastReturned = null;
2518 dl 1.1 }
2519    
2520     }
2521    
2522 dl 1.46 final class ValueIterator extends Iter<V> {
2523 dl 1.9 public V next() {
2524 jsr166 1.52 V v = nextValue;
2525 dl 1.46 advance();
2526 jsr166 1.52 return v;
2527 dl 1.1 }
2528     }
2529    
2530 dl 1.46 final class KeyIterator extends Iter<K> {
2531 dl 1.9 public K next() {
2532 dl 1.1 Node<K,V> n = next;
2533 dl 1.46 advance();
2534 dl 1.1 return n.key;
2535     }
2536     }
2537    
2538 dl 1.46 final class EntryIterator extends Iter<Map.Entry<K,V>> {
2539     public Map.Entry<K,V> next() {
2540     Node<K,V> n = next;
2541 jsr166 1.52 V v = nextValue;
2542 dl 1.46 advance();
2543     return new AbstractMap.SimpleImmutableEntry<K,V>(n.key, v);
2544 dl 1.1 }
2545 dl 1.46 }
2546 dl 1.1
2547 dl 1.46 // Factory methods for iterators needed by ConcurrentSkipListSet etc
2548    
2549     Iterator<K> keyIterator() {
2550 dl 1.1 return new KeyIterator();
2551     }
2552    
2553 dl 1.46 Iterator<V> valueIterator() {
2554     return new ValueIterator();
2555 dl 1.1 }
2556    
2557 dl 1.46 Iterator<Map.Entry<K,V>> entryIterator() {
2558     return new EntryIterator();
2559 dl 1.1 }
2560    
2561 dl 1.46 /* ---------------- View Classes -------------- */
2562    
2563     /*
2564     * View classes are static, delegating to a ConcurrentNavigableMap
2565     * to allow use by SubMaps, which outweighs the ugliness of
2566     * needing type-tests for Iterator methods.
2567     */
2568    
2569 jsr166 1.53 static final <E> List<E> toList(Collection<E> c) {
2570 jsr166 1.55 // Using size() here would be a pessimization.
2571 jsr166 1.90 ArrayList<E> list = new ArrayList<E>();
2572 jsr166 1.55 for (E e : c)
2573     list.add(e);
2574     return list;
2575 jsr166 1.53 }
2576    
2577 jsr166 1.62 static final class KeySet<E>
2578     extends AbstractSet<E> implements NavigableSet<E> {
2579 jsr166 1.71 private final ConcurrentNavigableMap<E,?> m;
2580     KeySet(ConcurrentNavigableMap<E,?> map) { m = map; }
2581 dl 1.46 public int size() { return m.size(); }
2582     public boolean isEmpty() { return m.isEmpty(); }
2583     public boolean contains(Object o) { return m.containsKey(o); }
2584     public boolean remove(Object o) { return m.remove(o) != null; }
2585     public void clear() { m.clear(); }
2586     public E lower(E e) { return m.lowerKey(e); }
2587     public E floor(E e) { return m.floorKey(e); }
2588     public E ceiling(E e) { return m.ceilingKey(e); }
2589     public E higher(E e) { return m.higherKey(e); }
2590     public Comparator<? super E> comparator() { return m.comparator(); }
2591     public E first() { return m.firstKey(); }
2592     public E last() { return m.lastKey(); }
2593     public E pollFirst() {
2594 jsr166 1.71 Map.Entry<E,?> e = m.pollFirstEntry();
2595 jsr166 1.61 return (e == null) ? null : e.getKey();
2596 dl 1.46 }
2597     public E pollLast() {
2598 jsr166 1.71 Map.Entry<E,?> e = m.pollLastEntry();
2599 jsr166 1.61 return (e == null) ? null : e.getKey();
2600 dl 1.46 }
2601     public Iterator<E> iterator() {
2602     if (m instanceof ConcurrentSkipListMap)
2603     return ((ConcurrentSkipListMap<E,Object>)m).keyIterator();
2604     else
2605     return ((ConcurrentSkipListMap.SubMap<E,Object>)m).keyIterator();
2606 dl 1.1 }
2607 dl 1.45 public boolean equals(Object o) {
2608     if (o == this)
2609     return true;
2610     if (!(o instanceof Set))
2611     return false;
2612     Collection<?> c = (Collection<?>) o;
2613     try {
2614     return containsAll(c) && c.containsAll(this);
2615 jsr166 1.81 } catch (ClassCastException unused) {
2616 dl 1.45 return false;
2617     } catch (NullPointerException unused) {
2618     return false;
2619     }
2620     }
2621 jsr166 1.55 public Object[] toArray() { return toList(this).toArray(); }
2622     public <T> T[] toArray(T[] a) { return toList(this).toArray(a); }
2623 dl 1.46 public Iterator<E> descendingIterator() {
2624     return descendingSet().iterator();
2625     }
2626 dl 1.47 public NavigableSet<E> subSet(E fromElement,
2627     boolean fromInclusive,
2628     E toElement,
2629     boolean toInclusive) {
2630 jsr166 1.56 return new KeySet<E>(m.subMap(fromElement, fromInclusive,
2631     toElement, toInclusive));
2632 dl 1.46 }
2633 dl 1.47 public NavigableSet<E> headSet(E toElement, boolean inclusive) {
2634 jsr166 1.56 return new KeySet<E>(m.headMap(toElement, inclusive));
2635 dl 1.46 }
2636 dl 1.47 public NavigableSet<E> tailSet(E fromElement, boolean inclusive) {
2637 jsr166 1.56 return new KeySet<E>(m.tailMap(fromElement, inclusive));
2638 dl 1.46 }
2639 jsr166 1.51 public NavigableSet<E> subSet(E fromElement, E toElement) {
2640 dl 1.47 return subSet(fromElement, true, toElement, false);
2641 dl 1.46 }
2642 jsr166 1.51 public NavigableSet<E> headSet(E toElement) {
2643 dl 1.47 return headSet(toElement, false);
2644 dl 1.46 }
2645 jsr166 1.51 public NavigableSet<E> tailSet(E fromElement) {
2646 dl 1.47 return tailSet(fromElement, true);
2647 dl 1.46 }
2648     public NavigableSet<E> descendingSet() {
2649 jsr166 1.71 return new KeySet<E>(m.descendingMap());
2650 dl 1.46 }
2651 dl 1.83
2652     public Stream<E> stream() {
2653     int flags = Streams.STREAM_IS_DISTINCT |
2654     Streams.STREAM_IS_SORTED | Streams.STREAM_IS_ORDERED;
2655     if (m instanceof ConcurrentSkipListMap)
2656     return Streams.stream
2657     (() -> ((ConcurrentSkipListMap<E,?>)m).keySpliterator(),
2658     flags);
2659     else
2660     return Streams.stream
2661     (Streams.spliteratorUnknownSize(iterator()), flags);
2662     }
2663    
2664     public Stream<E> parallelStream() {
2665     int flags = Streams.STREAM_IS_DISTINCT |
2666     Streams.STREAM_IS_SORTED | Streams.STREAM_IS_ORDERED;
2667     if (m instanceof ConcurrentSkipListMap)
2668     return Streams.parallelStream
2669     (() -> ((ConcurrentSkipListMap<E,?>)m).keySpliterator(),
2670     flags);
2671     else
2672     return Streams.parallelStream
2673     (Streams.spliteratorUnknownSize(iterator()), flags);
2674     }
2675 dl 1.1 }
2676    
2677 dl 1.46 static final class Values<E> extends AbstractCollection<E> {
2678 jsr166 1.71 private final ConcurrentNavigableMap<?, E> m;
2679     Values(ConcurrentNavigableMap<?, E> map) {
2680 dl 1.46 m = map;
2681 dl 1.1 }
2682 dl 1.46 public Iterator<E> iterator() {
2683     if (m instanceof ConcurrentSkipListMap)
2684 jsr166 1.71 return ((ConcurrentSkipListMap<?,E>)m).valueIterator();
2685 dl 1.46 else
2686 jsr166 1.71 return ((SubMap<?,E>)m).valueIterator();
2687 dl 1.1 }
2688     public boolean isEmpty() {
2689 dl 1.46 return m.isEmpty();
2690 dl 1.1 }
2691     public int size() {
2692 dl 1.46 return m.size();
2693 dl 1.1 }
2694     public boolean contains(Object o) {
2695 dl 1.46 return m.containsValue(o);
2696 dl 1.1 }
2697     public void clear() {
2698 dl 1.46 m.clear();
2699 dl 1.1 }
2700 jsr166 1.55 public Object[] toArray() { return toList(this).toArray(); }
2701     public <T> T[] toArray(T[] a) { return toList(this).toArray(a); }
2702 dl 1.83
2703     public Stream<E> stream() {
2704     int flags = Streams.STREAM_IS_ORDERED;
2705     if (m instanceof ConcurrentSkipListMap)
2706     return Streams.stream
2707     (() -> ((ConcurrentSkipListMap<?,E>)m).valueSpliterator(),
2708     flags);
2709     else
2710     return Streams.stream
2711     (Streams.spliteratorUnknownSize(iterator()), flags);
2712     }
2713    
2714     public Stream<E> parallelStream() {
2715     int flags = Streams.STREAM_IS_ORDERED;
2716     if (m instanceof ConcurrentSkipListMap)
2717     return Streams.parallelStream
2718     (() -> ((ConcurrentSkipListMap<?,E>)m).valueSpliterator(),
2719     flags);
2720     else
2721     return Streams.parallelStream
2722     (Streams.spliteratorUnknownSize(iterator()), flags);
2723     }
2724 dl 1.1 }
2725    
2726 dl 1.46 static final class EntrySet<K1,V1> extends AbstractSet<Map.Entry<K1,V1>> {
2727     private final ConcurrentNavigableMap<K1, V1> m;
2728     EntrySet(ConcurrentNavigableMap<K1, V1> map) {
2729     m = map;
2730 dl 1.1 }
2731 dl 1.46
2732     public Iterator<Map.Entry<K1,V1>> iterator() {
2733     if (m instanceof ConcurrentSkipListMap)
2734     return ((ConcurrentSkipListMap<K1,V1>)m).entryIterator();
2735     else
2736     return ((SubMap<K1,V1>)m).entryIterator();
2737     }
2738 dl 1.47
2739 dl 1.1 public boolean contains(Object o) {
2740     if (!(o instanceof Map.Entry))
2741     return false;
2742 jsr166 1.73 Map.Entry<?,?> e = (Map.Entry<?,?>)o;
2743 dl 1.46 V1 v = m.get(e.getKey());
2744 dl 1.1 return v != null && v.equals(e.getValue());
2745     }
2746     public boolean remove(Object o) {
2747     if (!(o instanceof Map.Entry))
2748     return false;
2749 jsr166 1.73 Map.Entry<?,?> e = (Map.Entry<?,?>)o;
2750 dl 1.46 return m.remove(e.getKey(),
2751 dl 1.47 e.getValue());
2752 dl 1.1 }
2753     public boolean isEmpty() {
2754 dl 1.46 return m.isEmpty();
2755 dl 1.1 }
2756     public int size() {
2757 dl 1.46 return m.size();
2758 dl 1.1 }
2759     public void clear() {
2760 dl 1.46 m.clear();
2761 dl 1.1 }
2762 dl 1.45 public boolean equals(Object o) {
2763     if (o == this)
2764     return true;
2765     if (!(o instanceof Set))
2766     return false;
2767     Collection<?> c = (Collection<?>) o;
2768     try {
2769     return containsAll(c) && c.containsAll(this);
2770 jsr166 1.81 } catch (ClassCastException unused) {
2771 dl 1.45 return false;
2772     } catch (NullPointerException unused) {
2773     return false;
2774     }
2775     }
2776 jsr166 1.55 public Object[] toArray() { return toList(this).toArray(); }
2777     public <T> T[] toArray(T[] a) { return toList(this).toArray(a); }
2778 dl 1.83
2779     @Override public Stream<Map.Entry<K1,V1>> stream() {
2780     int flags = Streams.STREAM_IS_ORDERED | Streams.STREAM_IS_DISTINCT;
2781     if (m instanceof ConcurrentSkipListMap)
2782     return Streams.stream
2783     (() -> ((ConcurrentSkipListMap<K1,V1>)m).entrySpliterator(),
2784     flags);
2785     else
2786     return Streams.stream
2787     (Streams.spliteratorUnknownSize(iterator()), flags);
2788     }
2789    
2790     public Stream<Map.Entry<K1,V1>> parallelStream() {
2791     int flags = Streams.STREAM_IS_ORDERED | Streams.STREAM_IS_DISTINCT;
2792     if (m instanceof ConcurrentSkipListMap)
2793     return Streams.parallelStream
2794     (() -> ((ConcurrentSkipListMap<K1,V1>)m).entrySpliterator(),
2795     flags);
2796     else
2797     return Streams.parallelStream
2798     (Streams.spliteratorUnknownSize(iterator()), flags);
2799     }
2800    
2801 dl 1.1 }
2802    
2803     /**
2804     * Submaps returned by {@link ConcurrentSkipListMap} submap operations
2805     * represent a subrange of mappings of their underlying
2806     * maps. Instances of this class support all methods of their
2807     * underlying maps, differing in that mappings outside their range are
2808     * ignored, and attempts to add mappings outside their ranges result
2809     * in {@link IllegalArgumentException}. Instances of this class are
2810 jsr166 1.82 * constructed only using the {@code subMap}, {@code headMap}, and
2811     * {@code tailMap} methods of their underlying maps.
2812 jsr166 1.52 *
2813     * @serial include
2814 dl 1.1 */
2815 dl 1.46 static final class SubMap<K,V> extends AbstractMap<K,V>
2816     implements ConcurrentNavigableMap<K,V>, Cloneable,
2817     java.io.Serializable {
2818 dl 1.1 private static final long serialVersionUID = -7647078645895051609L;
2819    
2820     /** Underlying map */
2821     private final ConcurrentSkipListMap<K,V> m;
2822     /** lower bound key, or null if from start */
2823 dl 1.46 private final K lo;
2824     /** upper bound key, or null if to end */
2825     private final K hi;
2826     /** inclusion flag for lo */
2827     private final boolean loInclusive;
2828     /** inclusion flag for hi */
2829     private final boolean hiInclusive;
2830     /** direction */
2831     private final boolean isDescending;
2832    
2833 dl 1.1 // Lazily initialized view holders
2834 dl 1.46 private transient KeySet<K> keySetView;
2835 dl 1.1 private transient Set<Map.Entry<K,V>> entrySetView;
2836     private transient Collection<V> valuesView;
2837    
2838     /**
2839 jsr166 1.87 * Creates a new submap, initializing all fields.
2840 dl 1.46 */
2841     SubMap(ConcurrentSkipListMap<K,V> map,
2842     K fromKey, boolean fromInclusive,
2843     K toKey, boolean toInclusive,
2844     boolean isDescending) {
2845 dl 1.47 if (fromKey != null && toKey != null &&
2846 dl 1.46 map.compare(fromKey, toKey) > 0)
2847 dl 1.1 throw new IllegalArgumentException("inconsistent range");
2848     this.m = map;
2849 dl 1.46 this.lo = fromKey;
2850     this.hi = toKey;
2851     this.loInclusive = fromInclusive;
2852     this.hiInclusive = toInclusive;
2853     this.isDescending = isDescending;
2854 dl 1.1 }
2855    
2856     /* ---------------- Utilities -------------- */
2857    
2858 dl 1.46 private boolean tooLow(K key) {
2859     if (lo != null) {
2860     int c = m.compare(key, lo);
2861     if (c < 0 || (c == 0 && !loInclusive))
2862     return true;
2863     }
2864     return false;
2865 dl 1.1 }
2866    
2867 dl 1.46 private boolean tooHigh(K key) {
2868     if (hi != null) {
2869     int c = m.compare(key, hi);
2870     if (c > 0 || (c == 0 && !hiInclusive))
2871     return true;
2872     }
2873     return false;
2874 dl 1.1 }
2875    
2876 dl 1.46 private boolean inBounds(K key) {
2877     return !tooLow(key) && !tooHigh(key);
2878 dl 1.1 }
2879    
2880 dl 1.46 private void checkKeyBounds(K key) throws IllegalArgumentException {
2881     if (key == null)
2882     throw new NullPointerException();
2883     if (!inBounds(key))
2884     throw new IllegalArgumentException("key out of range");
2885 dl 1.1 }
2886    
2887 dl 1.46 /**
2888 jsr166 1.87 * Returns true if node key is less than upper bound of range.
2889 dl 1.46 */
2890     private boolean isBeforeEnd(ConcurrentSkipListMap.Node<K,V> n) {
2891     if (n == null)
2892     return false;
2893     if (hi == null)
2894     return true;
2895     K k = n.key;
2896     if (k == null) // pass by markers and headers
2897     return true;
2898     int c = m.compare(k, hi);
2899     if (c > 0 || (c == 0 && !hiInclusive))
2900     return false;
2901     return true;
2902 dl 1.1 }
2903    
2904 dl 1.46 /**
2905     * Returns lowest node. This node might not be in range, so
2906 jsr166 1.87 * most usages need to check bounds.
2907 dl 1.46 */
2908     private ConcurrentSkipListMap.Node<K,V> loNode() {
2909     if (lo == null)
2910     return m.findFirst();
2911     else if (loInclusive)
2912 jsr166 1.70 return m.findNear(lo, GT|EQ);
2913 dl 1.46 else
2914 jsr166 1.70 return m.findNear(lo, GT);
2915 dl 1.1 }
2916    
2917     /**
2918 dl 1.46 * Returns highest node. This node might not be in range, so
2919 jsr166 1.87 * most usages need to check bounds.
2920 dl 1.1 */
2921 dl 1.46 private ConcurrentSkipListMap.Node<K,V> hiNode() {
2922     if (hi == null)
2923     return m.findLast();
2924     else if (hiInclusive)
2925 jsr166 1.70 return m.findNear(hi, LT|EQ);
2926 dl 1.46 else
2927 jsr166 1.70 return m.findNear(hi, LT);
2928 dl 1.1 }
2929    
2930     /**
2931 jsr166 1.87 * Returns lowest absolute key (ignoring directonality).
2932 dl 1.1 */
2933 dl 1.46 private K lowestKey() {
2934     ConcurrentSkipListMap.Node<K,V> n = loNode();
2935     if (isBeforeEnd(n))
2936     return n.key;
2937     else
2938     throw new NoSuchElementException();
2939 dl 1.47 }
2940 dl 1.46
2941     /**
2942 jsr166 1.87 * Returns highest absolute key (ignoring directonality).
2943 dl 1.46 */
2944     private K highestKey() {
2945     ConcurrentSkipListMap.Node<K,V> n = hiNode();
2946     if (n != null) {
2947     K last = n.key;
2948     if (inBounds(last))
2949     return last;
2950     }
2951     throw new NoSuchElementException();
2952     }
2953    
2954     private Map.Entry<K,V> lowestEntry() {
2955     for (;;) {
2956     ConcurrentSkipListMap.Node<K,V> n = loNode();
2957     if (!isBeforeEnd(n))
2958     return null;
2959     Map.Entry<K,V> e = n.createSnapshot();
2960     if (e != null)
2961     return e;
2962     }
2963     }
2964    
2965     private Map.Entry<K,V> highestEntry() {
2966     for (;;) {
2967     ConcurrentSkipListMap.Node<K,V> n = hiNode();
2968     if (n == null || !inBounds(n.key))
2969     return null;
2970     Map.Entry<K,V> e = n.createSnapshot();
2971     if (e != null)
2972     return e;
2973     }
2974     }
2975    
2976     private Map.Entry<K,V> removeLowest() {
2977     for (;;) {
2978     Node<K,V> n = loNode();
2979     if (n == null)
2980     return null;
2981     K k = n.key;
2982     if (!inBounds(k))
2983     return null;
2984 dl 1.88 V v = (m.comparator == null) ? m.doRemove(k, null) :
2985     m.doRemoveCmp(m.comparator, k, null);
2986 dl 1.46 if (v != null)
2987     return new AbstractMap.SimpleImmutableEntry<K,V>(k, v);
2988     }
2989     }
2990    
2991     private Map.Entry<K,V> removeHighest() {
2992     for (;;) {
2993     Node<K,V> n = hiNode();
2994     if (n == null)
2995     return null;
2996     K k = n.key;
2997     if (!inBounds(k))
2998     return null;
2999 dl 1.88 V v = (m.comparator == null) ? m.doRemove(k, null) :
3000     m.doRemoveCmp(m.comparator, k, null);
3001 dl 1.46 if (v != null)
3002     return new AbstractMap.SimpleImmutableEntry<K,V>(k, v);
3003     }
3004 dl 1.1 }
3005    
3006     /**
3007 dl 1.46 * Submap version of ConcurrentSkipListMap.getNearEntry
3008 dl 1.1 */
3009 dl 1.46 private Map.Entry<K,V> getNearEntry(K key, int rel) {
3010     if (isDescending) { // adjust relation for direction
3011 jsr166 1.70 if ((rel & LT) == 0)
3012     rel |= LT;
3013 dl 1.46 else
3014 jsr166 1.70 rel &= ~LT;
3015 dl 1.46 }
3016     if (tooLow(key))
3017 jsr166 1.70 return ((rel & LT) != 0) ? null : lowestEntry();
3018 dl 1.46 if (tooHigh(key))
3019 jsr166 1.70 return ((rel & LT) != 0) ? highestEntry() : null;
3020 dl 1.46 for (;;) {
3021     Node<K,V> n = m.findNear(key, rel);
3022     if (n == null || !inBounds(n.key))
3023     return null;
3024     K k = n.key;
3025     V v = n.getValidValue();
3026     if (v != null)
3027     return new AbstractMap.SimpleImmutableEntry<K,V>(k, v);
3028     }
3029 dl 1.1 }
3030    
3031 jsr166 1.48 // Almost the same as getNearEntry, except for keys
3032 dl 1.46 private K getNearKey(K key, int rel) {
3033     if (isDescending) { // adjust relation for direction
3034 jsr166 1.70 if ((rel & LT) == 0)
3035     rel |= LT;
3036 dl 1.46 else
3037 jsr166 1.70 rel &= ~LT;
3038 dl 1.46 }
3039     if (tooLow(key)) {
3040 jsr166 1.70 if ((rel & LT) == 0) {
3041 dl 1.46 ConcurrentSkipListMap.Node<K,V> n = loNode();
3042     if (isBeforeEnd(n))
3043     return n.key;
3044     }
3045     return null;
3046     }
3047     if (tooHigh(key)) {
3048 jsr166 1.70 if ((rel & LT) != 0) {
3049 dl 1.46 ConcurrentSkipListMap.Node<K,V> n = hiNode();
3050     if (n != null) {
3051     K last = n.key;
3052     if (inBounds(last))
3053     return last;
3054     }
3055     }
3056     return null;
3057     }
3058     for (;;) {
3059     Node<K,V> n = m.findNear(key, rel);
3060     if (n == null || !inBounds(n.key))
3061     return null;
3062     K k = n.key;
3063     V v = n.getValidValue();
3064     if (v != null)
3065     return k;
3066     }
3067     }
3068 dl 1.1
3069     /* ---------------- Map API methods -------------- */
3070    
3071     public boolean containsKey(Object key) {
3072 dl 1.46 if (key == null) throw new NullPointerException();
3073 dl 1.1 K k = (K)key;
3074 dl 1.46 return inBounds(k) && m.containsKey(k);
3075 dl 1.1 }
3076    
3077     public V get(Object key) {
3078 dl 1.46 if (key == null) throw new NullPointerException();
3079 dl 1.1 K k = (K)key;
3080 jsr166 1.74 return (!inBounds(k)) ? null : m.get(k);
3081 dl 1.1 }
3082    
3083     public V put(K key, V value) {
3084 dl 1.46 checkKeyBounds(key);
3085 dl 1.1 return m.put(key, value);
3086     }
3087    
3088     public V remove(Object key) {
3089     K k = (K)key;
3090 jsr166 1.61 return (!inBounds(k)) ? null : m.remove(k);
3091 dl 1.1 }
3092    
3093     public int size() {
3094     long count = 0;
3095 dl 1.46 for (ConcurrentSkipListMap.Node<K,V> n = loNode();
3096 dl 1.9 isBeforeEnd(n);
3097 dl 1.1 n = n.next) {
3098     if (n.getValidValue() != null)
3099     ++count;
3100     }
3101 jsr166 1.61 return count >= Integer.MAX_VALUE ? Integer.MAX_VALUE : (int)count;
3102 dl 1.1 }
3103    
3104     public boolean isEmpty() {
3105 dl 1.46 return !isBeforeEnd(loNode());
3106 dl 1.1 }
3107    
3108     public boolean containsValue(Object value) {
3109 dl 1.9 if (value == null)
3110 dl 1.1 throw new NullPointerException();
3111 dl 1.46 for (ConcurrentSkipListMap.Node<K,V> n = loNode();
3112 dl 1.9 isBeforeEnd(n);
3113 dl 1.1 n = n.next) {
3114     V v = n.getValidValue();
3115     if (v != null && value.equals(v))
3116     return true;
3117     }
3118     return false;
3119     }
3120    
3121     public void clear() {
3122 dl 1.46 for (ConcurrentSkipListMap.Node<K,V> n = loNode();
3123 dl 1.9 isBeforeEnd(n);
3124 dl 1.1 n = n.next) {
3125     if (n.getValidValue() != null)
3126     m.remove(n.key);
3127     }
3128     }
3129    
3130     /* ---------------- ConcurrentMap API methods -------------- */
3131    
3132     public V putIfAbsent(K key, V value) {
3133 dl 1.46 checkKeyBounds(key);
3134 dl 1.1 return m.putIfAbsent(key, value);
3135     }
3136    
3137     public boolean remove(Object key, Object value) {
3138     K k = (K)key;
3139 dl 1.46 return inBounds(k) && m.remove(k, value);
3140 dl 1.1 }
3141    
3142     public boolean replace(K key, V oldValue, V newValue) {
3143 dl 1.46 checkKeyBounds(key);
3144 dl 1.1 return m.replace(key, oldValue, newValue);
3145     }
3146    
3147     public V replace(K key, V value) {
3148 dl 1.46 checkKeyBounds(key);
3149 dl 1.1 return m.replace(key, value);
3150     }
3151    
3152     /* ---------------- SortedMap API methods -------------- */
3153    
3154     public Comparator<? super K> comparator() {
3155 dl 1.46 Comparator<? super K> cmp = m.comparator();
3156 jsr166 1.55 if (isDescending)
3157     return Collections.reverseOrder(cmp);
3158     else
3159     return cmp;
3160 dl 1.1 }
3161 dl 1.47
3162 dl 1.46 /**
3163     * Utility to create submaps, where given bounds override
3164     * unbounded(null) ones and/or are checked against bounded ones.
3165     */
3166 dl 1.47 private SubMap<K,V> newSubMap(K fromKey,
3167     boolean fromInclusive,
3168     K toKey,
3169 dl 1.46 boolean toInclusive) {
3170     if (isDescending) { // flip senses
3171 dl 1.47 K tk = fromKey;
3172     fromKey = toKey;
3173 dl 1.46 toKey = tk;
3174 dl 1.47 boolean ti = fromInclusive;
3175     fromInclusive = toInclusive;
3176 dl 1.46 toInclusive = ti;
3177     }
3178     if (lo != null) {
3179     if (fromKey == null) {
3180     fromKey = lo;
3181     fromInclusive = loInclusive;
3182     }
3183     else {
3184     int c = m.compare(fromKey, lo);
3185     if (c < 0 || (c == 0 && !loInclusive && fromInclusive))
3186     throw new IllegalArgumentException("key out of range");
3187     }
3188     }
3189     if (hi != null) {
3190     if (toKey == null) {
3191     toKey = hi;
3192     toInclusive = hiInclusive;
3193     }
3194     else {
3195     int c = m.compare(toKey, hi);
3196     if (c > 0 || (c == 0 && !hiInclusive && toInclusive))
3197     throw new IllegalArgumentException("key out of range");
3198     }
3199 dl 1.1 }
3200 dl 1.47 return new SubMap<K,V>(m, fromKey, fromInclusive,
3201 dl 1.46 toKey, toInclusive, isDescending);
3202 dl 1.1 }
3203    
3204 dl 1.47 public SubMap<K,V> subMap(K fromKey,
3205     boolean fromInclusive,
3206     K toKey,
3207     boolean toInclusive) {
3208 dl 1.1 if (fromKey == null || toKey == null)
3209     throw new NullPointerException();
3210 dl 1.46 return newSubMap(fromKey, fromInclusive, toKey, toInclusive);
3211 dl 1.1 }
3212 dl 1.47
3213     public SubMap<K,V> headMap(K toKey,
3214     boolean inclusive) {
3215 dl 1.1 if (toKey == null)
3216     throw new NullPointerException();
3217 dl 1.46 return newSubMap(null, false, toKey, inclusive);
3218 dl 1.1 }
3219 dl 1.47
3220     public SubMap<K,V> tailMap(K fromKey,
3221     boolean inclusive) {
3222 dl 1.1 if (fromKey == null)
3223     throw new NullPointerException();
3224 dl 1.46 return newSubMap(fromKey, inclusive, null, false);
3225     }
3226    
3227     public SubMap<K,V> subMap(K fromKey, K toKey) {
3228 dl 1.47 return subMap(fromKey, true, toKey, false);
3229 dl 1.1 }
3230    
3231 dl 1.46 public SubMap<K,V> headMap(K toKey) {
3232 dl 1.47 return headMap(toKey, false);
3233 dl 1.6 }
3234    
3235 dl 1.46 public SubMap<K,V> tailMap(K fromKey) {
3236 dl 1.47 return tailMap(fromKey, true);
3237 dl 1.6 }
3238    
3239 dl 1.46 public SubMap<K,V> descendingMap() {
3240 dl 1.47 return new SubMap<K,V>(m, lo, loInclusive,
3241 dl 1.46 hi, hiInclusive, !isDescending);
3242 dl 1.6 }
3243    
3244 dl 1.1 /* ---------------- Relational methods -------------- */
3245    
3246     public Map.Entry<K,V> ceilingEntry(K key) {
3247 jsr166 1.70 return getNearEntry(key, GT|EQ);
3248 dl 1.1 }
3249    
3250     public K ceilingKey(K key) {
3251 jsr166 1.70 return getNearKey(key, GT|EQ);
3252 dl 1.1 }
3253    
3254     public Map.Entry<K,V> lowerEntry(K key) {
3255 jsr166 1.70 return getNearEntry(key, LT);
3256 dl 1.1 }
3257    
3258     public K lowerKey(K key) {
3259 jsr166 1.70 return getNearKey(key, LT);
3260 dl 1.1 }
3261    
3262     public Map.Entry<K,V> floorEntry(K key) {
3263 jsr166 1.70 return getNearEntry(key, LT|EQ);
3264 dl 1.1 }
3265    
3266     public K floorKey(K key) {
3267 jsr166 1.70 return getNearKey(key, LT|EQ);
3268 dl 1.1 }
3269    
3270     public Map.Entry<K,V> higherEntry(K key) {
3271 jsr166 1.70 return getNearEntry(key, GT);
3272 dl 1.1 }
3273    
3274     public K higherKey(K key) {
3275 jsr166 1.70 return getNearKey(key, GT);
3276 dl 1.46 }
3277    
3278     public K firstKey() {
3279 jsr166 1.61 return isDescending ? highestKey() : lowestKey();
3280 dl 1.46 }
3281    
3282     public K lastKey() {
3283 jsr166 1.61 return isDescending ? lowestKey() : highestKey();
3284 dl 1.1 }
3285    
3286     public Map.Entry<K,V> firstEntry() {
3287 jsr166 1.61 return isDescending ? highestEntry() : lowestEntry();
3288 dl 1.1 }
3289    
3290     public Map.Entry<K,V> lastEntry() {
3291 jsr166 1.61 return isDescending ? lowestEntry() : highestEntry();
3292 dl 1.1 }
3293    
3294     public Map.Entry<K,V> pollFirstEntry() {
3295 jsr166 1.61 return isDescending ? removeHighest() : removeLowest();
3296 dl 1.1 }
3297    
3298     public Map.Entry<K,V> pollLastEntry() {
3299 jsr166 1.61 return isDescending ? removeLowest() : removeHighest();
3300 dl 1.1 }
3301    
3302     /* ---------------- Submap Views -------------- */
3303    
3304 jsr166 1.51 public NavigableSet<K> keySet() {
3305 dl 1.46 KeySet<K> ks = keySetView;
3306 jsr166 1.71 return (ks != null) ? ks : (keySetView = new KeySet<K>(this));
3307 dl 1.1 }
3308    
3309 dl 1.46 public NavigableSet<K> navigableKeySet() {
3310     KeySet<K> ks = keySetView;
3311 jsr166 1.71 return (ks != null) ? ks : (keySetView = new KeySet<K>(this));
3312 dl 1.46 }
3313 dl 1.45
3314 dl 1.46 public Collection<V> values() {
3315     Collection<V> vs = valuesView;
3316 jsr166 1.71 return (vs != null) ? vs : (valuesView = new Values<V>(this));
3317 dl 1.1 }
3318    
3319 dl 1.46 public Set<Map.Entry<K,V>> entrySet() {
3320     Set<Map.Entry<K,V>> es = entrySetView;
3321 jsr166 1.71 return (es != null) ? es : (entrySetView = new EntrySet<K,V>(this));
3322 dl 1.1 }
3323    
3324 dl 1.46 public NavigableSet<K> descendingKeySet() {
3325     return descendingMap().navigableKeySet();
3326 dl 1.1 }
3327    
3328 dl 1.46 Iterator<K> keyIterator() {
3329     return new SubMapKeyIterator();
3330 dl 1.1 }
3331    
3332 dl 1.46 Iterator<V> valueIterator() {
3333     return new SubMapValueIterator();
3334 dl 1.1 }
3335    
3336 dl 1.46 Iterator<Map.Entry<K,V>> entryIterator() {
3337     return new SubMapEntryIterator();
3338 dl 1.1 }
3339    
3340 dl 1.46 /**
3341     * Variant of main Iter class to traverse through submaps.
3342     */
3343     abstract class SubMapIter<T> implements Iterator<T> {
3344     /** the last node returned by next() */
3345 jsr166 1.52 Node<K,V> lastReturned;
3346 dl 1.46 /** the next node to return from next(); */
3347     Node<K,V> next;
3348     /** Cache of next value field to maintain weak consistency */
3349 jsr166 1.52 V nextValue;
3350 dl 1.46
3351 dl 1.47 SubMapIter() {
3352 dl 1.46 for (;;) {
3353 jsr166 1.52 next = isDescending ? hiNode() : loNode();
3354 dl 1.46 if (next == null)
3355     break;
3356 jsr166 1.55 Object x = next.value;
3357 jsr166 1.52 if (x != null && x != next) {
3358 jsr166 1.55 if (! inBounds(next.key))
3359 dl 1.46 next = null;
3360 jsr166 1.55 else
3361     nextValue = (V) x;
3362 dl 1.46 break;
3363     }
3364     }
3365 dl 1.1 }
3366 dl 1.46
3367     public final boolean hasNext() {
3368     return next != null;
3369 dl 1.1 }
3370 dl 1.46
3371     final void advance() {
3372 jsr166 1.54 if (next == null)
3373 dl 1.46 throw new NoSuchElementException();
3374 jsr166 1.55 lastReturned = next;
3375 dl 1.46 if (isDescending)
3376     descend();
3377     else
3378     ascend();
3379 dl 1.1 }
3380 dl 1.46
3381     private void ascend() {
3382     for (;;) {
3383     next = next.next;
3384     if (next == null)
3385     break;
3386 jsr166 1.55 Object x = next.value;
3387 jsr166 1.52 if (x != null && x != next) {
3388     if (tooHigh(next.key))
3389 dl 1.46 next = null;
3390 jsr166 1.52 else
3391 jsr166 1.55 nextValue = (V) x;
3392 dl 1.46 break;
3393     }
3394     }
3395     }
3396    
3397     private void descend() {
3398 dl 1.88 Comparator<? super K> cmp = m.comparator;
3399 dl 1.46 for (;;) {
3400 dl 1.88 next = (cmp == null) ? m.doFindNear(lastReturned.key, LT) :
3401     m.doFindNearCmp(cmp, lastReturned.key, LT);
3402 dl 1.46 if (next == null)
3403     break;
3404 jsr166 1.55 Object x = next.value;
3405 jsr166 1.52 if (x != null && x != next) {
3406     if (tooLow(next.key))
3407 dl 1.46 next = null;
3408 jsr166 1.55 else
3409 jsr166 1.52 nextValue = (V) x;
3410 dl 1.46 break;
3411     }
3412     }
3413 dl 1.1 }
3414 dl 1.46
3415     public void remove() {
3416 jsr166 1.52 Node<K,V> l = lastReturned;
3417 dl 1.46 if (l == null)
3418     throw new IllegalStateException();
3419     m.remove(l.key);
3420 jsr166 1.55 lastReturned = null;
3421 dl 1.1 }
3422 dl 1.46
3423     }
3424    
3425     final class SubMapValueIterator extends SubMapIter<V> {
3426     public V next() {
3427 jsr166 1.52 V v = nextValue;
3428 dl 1.46 advance();
3429 jsr166 1.52 return v;
3430 dl 1.45 }
3431 dl 1.1 }
3432    
3433 dl 1.46 final class SubMapKeyIterator extends SubMapIter<K> {
3434     public K next() {
3435     Node<K,V> n = next;
3436     advance();
3437     return n.key;
3438     }
3439 dl 1.1 }
3440    
3441 dl 1.46 final class SubMapEntryIterator extends SubMapIter<Map.Entry<K,V>> {
3442     public Map.Entry<K,V> next() {
3443     Node<K,V> n = next;
3444 jsr166 1.52 V v = nextValue;
3445 dl 1.46 advance();
3446     return new AbstractMap.SimpleImmutableEntry<K,V>(n.key, v);
3447 dl 1.1 }
3448     }
3449     }
3450 dl 1.59
3451 dl 1.83 /**
3452     * A view of a ConcurrentSkipListMap as a {@link Set} of keys, in
3453     * which additions may optionally be enabled by mapping to a
3454     * common value. This class cannot be directly instantiated. See
3455     * {@link #keySet}, {@link #keySet(Object)}, {@link #newKeySet()},
3456     * {@link #newKeySet(Comparator)}.
3457     */
3458     public static class KeySetView<K,V> extends AbstractSet<K>
3459     implements NavigableSet<K>, java.io.Serializable {
3460    
3461     /*
3462     * This class overlaps in functionality with the
3463     * relative-scoped KeySet class, but must be distinct and
3464     * unrelated. So we repeat most of the boring delegation code.
3465     */
3466    
3467     private static final long serialVersionUID = 7249069246763182397L;
3468     private final ConcurrentSkipListMap<K, V> m;
3469     private final V value;
3470    
3471     KeySetView(ConcurrentSkipListMap<K, V> map, V value) { // non-public
3472     this.m = map;
3473     this.value = value;
3474     }
3475    
3476     /**
3477     * Returns the map backing this view.
3478     *
3479     * @return the map backing this view
3480     */
3481     public ConcurrentSkipListMap<K,V> getMap() { return m; }
3482    
3483     /**
3484     * Returns the default mapped value for additions,
3485     * or {@code null} if additions are not supported.
3486     *
3487     * @return the default mapped value for additions, or {@code null}
3488     * if not supported.
3489     */
3490     public V getMappedValue() { return value; }
3491    
3492     public boolean add(K e) {
3493     V v;
3494     if ((v = value) == null)
3495     throw new UnsupportedOperationException();
3496     if (e == null)
3497     throw new NullPointerException();
3498     return m.put(e, v) == null;
3499     }
3500    
3501     public boolean addAll(Collection<? extends K> c) {
3502     boolean added = false;
3503     V v;
3504     if ((v = value) == null)
3505     throw new UnsupportedOperationException();
3506     for (K e : c) {
3507     if (e == null)
3508     throw new NullPointerException();
3509     if (m.put(e, v) == null)
3510     added = true;
3511     }
3512     return added;
3513     }
3514    
3515     public int size() { return m.size(); }
3516     public boolean isEmpty() { return m.isEmpty(); }
3517     public boolean contains(Object o) { return m.containsKey(o); }
3518     public boolean remove(Object o) { return m.remove(o) != null; }
3519     public void clear() { m.clear(); }
3520     public K lower(K e) { return m.lowerKey(e); }
3521     public K floor(K e) { return m.floorKey(e); }
3522     public K ceiling(K e) { return m.ceilingKey(e); }
3523     public K higher(K e) { return m.higherKey(e); }
3524     public Comparator<? super K> comparator() { return m.comparator(); }
3525     public K first() { return m.firstKey(); }
3526     public K last() { return m.lastKey(); }
3527     public Iterator<K> iterator() { return m.keyIterator(); }
3528     public K pollFirst() {
3529     Map.Entry<K,?> e = m.pollFirstEntry();
3530     return (e == null) ? null : e.getKey();
3531     }
3532     public K pollLast() {
3533     Map.Entry<K,?> e = m.pollLastEntry();
3534     return (e == null) ? null : e.getKey();
3535     }
3536     public boolean equals(Object o) {
3537     if (o == this)
3538     return true;
3539     if (!(o instanceof Set))
3540     return false;
3541     Collection<?> c = (Collection<?>) o;
3542     try {
3543     return containsAll(c) && c.containsAll(this);
3544     } catch (ClassCastException unused) {
3545     return false;
3546     } catch (NullPointerException unused) {
3547     return false;
3548     }
3549     }
3550     public Object[] toArray() { return toList(this).toArray(); }
3551     public <T> T[] toArray(T[] a) { return toList(this).toArray(a); }
3552     public Iterator<K> descendingIterator() {
3553     return descendingSet().iterator();
3554     }
3555     public NavigableSet<K> subSet(K fromElement,
3556     boolean fromInclusive,
3557     K toElement,
3558     boolean toInclusive) {
3559     return new KeySet<K>(m.subMap(fromElement, fromInclusive,
3560     toElement, toInclusive));
3561     }
3562     public NavigableSet<K> headSet(K toElement, boolean inclusive) {
3563     return new KeySet<K>(m.headMap(toElement, inclusive));
3564     }
3565     public NavigableSet<K> tailSet(K fromElement, boolean inclusive) {
3566     return new KeySet<K>(m.tailMap(fromElement, inclusive));
3567     }
3568     public NavigableSet<K> subSet(K fromElement, K toElement) {
3569     return subSet(fromElement, true, toElement, false);
3570     }
3571     public NavigableSet<K> headSet(K toElement) {
3572     return headSet(toElement, false);
3573     }
3574     public NavigableSet<K> tailSet(K fromElement) {
3575     return tailSet(fromElement, true);
3576     }
3577     public NavigableSet<K> descendingSet() {
3578     return new KeySet<K>(m.descendingMap());
3579     }
3580    
3581     public Stream<K> stream() {
3582     int flags = Streams.STREAM_IS_DISTINCT |
3583     Streams.STREAM_IS_SORTED | Streams.STREAM_IS_ORDERED;
3584     return Streams.stream(() -> m.keySpliterator(), flags);
3585     }
3586    
3587     public Stream<K> parallelStream() {
3588     int flags = Streams.STREAM_IS_DISTINCT |
3589     Streams.STREAM_IS_SORTED | Streams.STREAM_IS_ORDERED;
3590     return Streams.parallelStream(() -> m.keySpliterator(), flags);
3591     }
3592    
3593     }
3594    
3595     /**
3596     * Base class providing common structure for Spliterators.
3597     * (Although not all that much common functionality; as usual for
3598     * view classes, details annoyingly vary in key, value, and entry
3599     * subclasses in ways that are not worth abstracting out for
3600     * internal classes.)
3601     *
3602     * The basic split strategy is to recursively descend from top
3603     * level, row by row, descending to next row when either split
3604     * off, or the end of row is encountered. Control of the number of
3605     * splits relies on some statistical estimation: The expected
3606     * remaining number of elements of a skip list when advancing
3607     * either across or down decreases by about 25%. To make this
3608     * observation useful, we need to know initial size, which we
3609     * don't. But we use (1 << 2*levels) as a rough overestimate that
3610     * minimizes risk of prematurely zeroing out while splitting.
3611     */
3612     static class CSLMSpliterator<K,V> {
3613     final Comparator<? super K> comparator;
3614     final K fence; // exclusive upper bound for keys, or null if to end
3615     Index<K,V> row; // the level to split out
3616     Node<K,V> current; // current traversal node; initialize at origin
3617     int est; // pseudo-size estimate
3618    
3619     CSLMSpliterator(Comparator<? super K> comparator, Index<K,V> row,
3620     Node<K,V> origin, K fence, int est) {
3621     this.comparator = comparator; this.row = row;
3622     this.current = origin; this.fence = fence; this.est = est;
3623     }
3624    
3625     /** Return >= 0 if key is too large (out of bounds) */
3626     final int compareBounds(K k) {
3627     Comparator<? super K> cmp; K f;
3628     if (k == null || (f = fence) == null)
3629     return -1;
3630     else if ((cmp = comparator) != null)
3631     return cmp.compare(k, f);
3632     else
3633     return ((Comparable<? super K>)k).compareTo(f);
3634     }
3635    
3636     public final long estimateSize() { return (long)est; }
3637     public final boolean hasExactSize() { return est == 0; }
3638     public final boolean hasExactSplits() { return false; }
3639     }
3640    
3641     // factory methods
3642     final KeySpliterator<K,V> keySpliterator() {
3643     HeadIndex<K,V> h; Node<K,V> p; int d, n;
3644     for (;;) { // ensure h and n correspond to origin p
3645     Node<K,V> b = (h = head).node;
3646     if ((p = b.next) == null) {
3647     n = 0;
3648     break;
3649     }
3650     if (p.value != null) {
3651     n = (d = h.level << 1) >= 31 ? Integer.MAX_VALUE : 1 << d;
3652     break;
3653     }
3654     p.helpDelete(b, p.next);
3655     }
3656     return new KeySpliterator<K,V>(comparator, h, p, null, n);
3657     }
3658    
3659     final ValueSpliterator<K,V> valueSpliterator() {
3660     HeadIndex<K,V> h; Node<K,V> p; int d, n;
3661     for (;;) { // same as key version
3662     Node<K,V> b = (h = head).node;
3663     if ((p = b.next) == null) {
3664     n = 0;
3665     break;
3666     }
3667     if (p.value != null) {
3668     n = (d = h.level << 1) >= 31 ? Integer.MAX_VALUE : 1 << d;
3669     break;
3670     }
3671     p.helpDelete(b, p.next);
3672     }
3673     return new ValueSpliterator<K,V>(comparator, h, p, null, n);
3674     }
3675    
3676     final EntrySpliterator<K,V> entrySpliterator() {
3677     HeadIndex<K,V> h; Node<K,V> p; int d, n;
3678     for (;;) { // same as key version
3679     Node<K,V> b = (h = head).node;
3680     if ((p = b.next) == null) {
3681     n = 0;
3682     break;
3683     }
3684     if (p.value != null) {
3685     n = (d = h.level << 1) >= 31 ? Integer.MAX_VALUE : 1 << d;
3686     break;
3687     }
3688     p.helpDelete(b, p.next);
3689     }
3690     return new EntrySpliterator<K,V>(comparator, head, p, null, n);
3691     }
3692    
3693     static final class KeySpliterator<K,V> extends CSLMSpliterator<K,V>
3694 dl 1.88 implements Spliterator<K> {
3695 dl 1.83 KeySpliterator(Comparator<? super K> comparator, Index<K,V> row,
3696     Node<K,V> origin, K fence, int est) {
3697     super(comparator, row, origin, fence, est);
3698     }
3699    
3700     public KeySpliterator<K,V> trySplit() {
3701     Node<K,V> e;
3702     Comparator<? super K> cmp = comparator;
3703     K f = fence;
3704     if ((e = current) != null) {
3705     for (Index<K,V> q = row; q != null; q = row = q.down) {
3706     Index<K,V> s; Node<K,V> n; K sk;
3707     est -= est >>> 2;
3708     if ((s = q.right) != null) {
3709     for (;;) {
3710     Node<K,V> b = s.node;
3711     if ((n = b.next) == null || n.value != null)
3712     break;
3713     n.helpDelete(b, n.next);
3714     }
3715     if (n != null && (sk = n.key) != null &&
3716     (f == null ||
3717     (cmp != null ? (cmp.compare(f, sk) > 0) :
3718     (((Comparable<? super K>)f).compareTo(sk) > 0)))) {
3719     current = n;
3720     Index<K,V> r = q.down;
3721     row = (s.right != null) ? s : s.down;
3722     return new KeySpliterator<K,V>(cmp, r, e, sk, est);
3723     }
3724     }
3725     }
3726     }
3727     return null;
3728     }
3729    
3730     public void forEach(Block<? super K> block) {
3731     if (block == null) throw new NullPointerException();
3732     K f = fence;
3733     Comparator<? super K> cmp = comparator;
3734     Comparable<? super K> cf = (f != null && cmp == null) ?
3735     (Comparable<? super K>)f : null;
3736     Node<K,V> e = current;
3737     current = null;
3738 jsr166 1.84 for (; e != null; e = e.next) {
3739 dl 1.83 K k; Object v;
3740     if ((k = e.key) != null &&
3741     (cf != null ? (cf.compareTo(k) <= 0) :
3742     (f != null && cmp.compare(f, k) <= 0)))
3743     break;
3744     if ((v = e.value) != null && v != e)
3745     block.accept(k);
3746     }
3747     }
3748    
3749     public boolean tryAdvance(Block<? super K> block) {
3750     if (block == null) throw new NullPointerException();
3751     Node<K,V> e;
3752     for (e = current; e != null; e = e.next) {
3753     K k; Object v;
3754     if (compareBounds(k = e.key) >= 0) {
3755     e = null;
3756     break;
3757     }
3758     if ((v = e.value) != null && v != e) {
3759     current = e.next;
3760     block.accept(k);
3761     return true;
3762     }
3763     }
3764     current = e;
3765     return false;
3766     }
3767     }
3768    
3769     static final class ValueSpliterator<K,V> extends CSLMSpliterator<K,V>
3770 dl 1.88 implements Spliterator<V> {
3771 dl 1.83 ValueSpliterator(Comparator<? super K> comparator, Index<K,V> row,
3772     Node<K,V> origin, K fence, int est) {
3773     super(comparator, row, origin, fence, est);
3774     }
3775    
3776     public ValueSpliterator<K,V> trySplit() {
3777     Node<K,V> e;
3778     Comparator<? super K> cmp = comparator;
3779     K f = fence;
3780     if ((e = current) != null) {
3781     for (Index<K,V> q = row; q != null; q = row = q.down) {
3782     Index<K,V> s; Node<K,V> n; K sk;
3783     est -= est >>> 2;
3784     if ((s = q.right) != null) {
3785     for (;;) {
3786     Node<K,V> b = s.node;
3787     if ((n = b.next) == null || n.value != null)
3788     break;
3789     n.helpDelete(b, n.next);
3790     }
3791     if (n != null && (sk = n.key) != null &&
3792     (f == null ||
3793     (cmp != null ? (cmp.compare(f, sk) > 0) :
3794     (((Comparable<? super K>)f).compareTo(sk) > 0)))) {
3795     current = n;
3796     Index<K,V> r = q.down;
3797     row = (s.right != null) ? s : s.down;
3798     return new ValueSpliterator<K,V>(cmp, r, e, sk, est);
3799     }
3800     }
3801     }
3802     }
3803     return null;
3804     }
3805    
3806     public void forEach(Block<? super V> block) {
3807     if (block == null) throw new NullPointerException();
3808     K f = fence;
3809     Comparator<? super K> cmp = comparator;
3810     Comparable<? super K> cf = (f != null && cmp == null) ?
3811     (Comparable<? super K>)f : null;
3812     Node<K,V> e = current;
3813     current = null;
3814 jsr166 1.84 for (; e != null; e = e.next) {
3815 dl 1.83 K k; Object v;
3816     if ((k = e.key) != null &&
3817     (cf != null ? (cf.compareTo(k) <= 0) :
3818     (f != null && cmp.compare(f, k) <= 0)))
3819     break;
3820     if ((v = e.value) != null && v != e)
3821     block.accept((V)v);
3822     }
3823     }
3824    
3825     public boolean tryAdvance(Block<? super V> block) {
3826     if (block == null) throw new NullPointerException();
3827     boolean advanced = false;
3828     Node<K,V> e;
3829     for (e = current; e != null; e = e.next) {
3830     K k; Object v;
3831     if (compareBounds(k = e.key) >= 0) {
3832     e = null;
3833     break;
3834     }
3835     if ((v = e.value) != null && v != e) {
3836     current = e.next;
3837     block.accept((V)v);
3838     return true;
3839     }
3840     }
3841     current = e;
3842     return false;
3843     }
3844     }
3845    
3846     static final class EntrySpliterator<K,V> extends CSLMSpliterator<K,V>
3847 dl 1.88 implements Spliterator<Map.Entry<K,V>> {
3848 dl 1.83 EntrySpliterator(Comparator<? super K> comparator, Index<K,V> row,
3849     Node<K,V> origin, K fence, int est) {
3850     super(comparator, row, origin, fence, est);
3851     }
3852    
3853     public EntrySpliterator<K,V> trySplit() {
3854     Node<K,V> e;
3855     Comparator<? super K> cmp = comparator;
3856     K f = fence;
3857     if ((e = current) != null) {
3858     for (Index<K,V> q = row; q != null; q = row = q.down) {
3859     Index<K,V> s; Node<K,V> n; K sk;
3860     est -= est >>> 2;
3861     if ((s = q.right) != null) {
3862     for (;;) {
3863     Node<K,V> b = s.node;
3864     if ((n = b.next) == null || n.value != null)
3865     break;
3866     n.helpDelete(b, n.next);
3867     }
3868     if (n != null && (sk = n.key) != null &&
3869     (f == null ||
3870 jsr166 1.84 (cmp != null ?
3871 dl 1.83 (cmp.compare(f, sk) > 0) :
3872     (((Comparable<? super K>)f).compareTo(sk) > 0)))) {
3873     current = n;
3874     Index<K,V> r = q.down;
3875     row = (s.right != null) ? s : s.down;
3876     return new EntrySpliterator<K,V>(cmp, r, e, sk, est);
3877     }
3878     }
3879     }
3880     }
3881     return null;
3882     }
3883    
3884     public void forEach(Block<? super Map.Entry<K,V>> block) {
3885     if (block == null) throw new NullPointerException();
3886     K f = fence;
3887     Comparator<? super K> cmp = comparator;
3888     Comparable<? super K> cf = (f != null && cmp == null) ?
3889     (Comparable<? super K>)f : null;
3890     Node<K,V> e = current;
3891     current = null;
3892 jsr166 1.84 for (; e != null; e = e.next) {
3893 dl 1.83 K k; Object v;
3894     if ((k = e.key) != null &&
3895     (cf != null ?
3896     (cf.compareTo(k) <= 0) :
3897     (f != null && cmp.compare(f, k) <= 0)))
3898     break;
3899     if ((v = e.value) != null && v != e)
3900     block.accept
3901     (new AbstractMap.SimpleImmutableEntry<K,V>(k, (V)v));
3902     }
3903     }
3904    
3905     public boolean tryAdvance(Block<? super Map.Entry<K,V>> block) {
3906     if (block == null) throw new NullPointerException();
3907     Node<K,V> e;
3908     for (e = current; e != null; e = e.next) {
3909     K k; Object v;
3910     if (compareBounds(k = e.key) >= 0) {
3911     e = null;
3912     break;
3913     }
3914     if ((v = e.value) != null && v != e) {
3915     current = e.next;
3916     block.accept
3917     (new AbstractMap.SimpleImmutableEntry<K,V>(k, (V)v));
3918     return true;
3919     }
3920     }
3921     current = e;
3922     return false;
3923     }
3924     }
3925    
3926 dl 1.59 // Unsafe mechanics
3927 dl 1.65 private static final sun.misc.Unsafe UNSAFE;
3928     private static final long headOffset;
3929     static {
3930 dl 1.59 try {
3931 dl 1.65 UNSAFE = sun.misc.Unsafe.getUnsafe();
3932 jsr166 1.72 Class<?> k = ConcurrentSkipListMap.class;
3933 dl 1.65 headOffset = UNSAFE.objectFieldOffset
3934     (k.getDeclaredField("head"));
3935     } catch (Exception e) {
3936     throw new Error(e);
3937 dl 1.59 }
3938     }
3939 dl 1.1 }