001/* Hashtable.java -- a class providing a basic hashtable data structure,
002   mapping Object --> Object
003   Copyright (C) 1998, 1999, 2000, 2001, 2002, 2004, 2005, 2006
004   Free Software Foundation, Inc.
005
006This file is part of GNU Classpath.
007
008GNU Classpath is free software; you can redistribute it and/or modify
009it under the terms of the GNU General Public License as published by
010the Free Software Foundation; either version 2, or (at your option)
011any later version.
012
013GNU Classpath is distributed in the hope that it will be useful, but
014WITHOUT ANY WARRANTY; without even the implied warranty of
015MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
016General Public License for more details.
017
018You should have received a copy of the GNU General Public License
019along with GNU Classpath; see the file COPYING.  If not, write to the
020Free Software Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA
02102110-1301 USA.
022
023Linking this library statically or dynamically with other modules is
024making a combined work based on this library.  Thus, the terms and
025conditions of the GNU General Public License cover the whole
026combination.
027
028As a special exception, the copyright holders of this library give you
029permission to link this library with independent modules to produce an
030executable, regardless of the license terms of these independent
031modules, and to copy and distribute the resulting executable under
032terms of your choice, provided that you also meet, for each linked
033independent module, the terms and conditions of the license of that
034module.  An independent module is a module which is not derived from
035or based on this library.  If you modify this library, you may extend
036this exception to your version of the library, but you are not
037obligated to do so.  If you do not wish to do so, delete this
038exception statement from your version. */
039
040package java.util;
041
042import gnu.java.lang.CPStringBuilder;
043
044import java.io.IOException;
045import java.io.ObjectInputStream;
046import java.io.ObjectOutputStream;
047import java.io.Serializable;
048
049// NOTE: This implementation is very similar to that of HashMap. If you fix
050// a bug in here, chances are you should make a similar change to the HashMap
051// code.
052
053/**
054 * A class which implements a hashtable data structure.
055 * <p>
056 *
057 * This implementation of Hashtable uses a hash-bucket approach. That is:
058 * linear probing and rehashing is avoided; instead, each hashed value maps
059 * to a simple linked-list which, in the best case, only has one node.
060 * Assuming a large enough table, low enough load factor, and / or well
061 * implemented hashCode() methods, Hashtable should provide O(1)
062 * insertion, deletion, and searching of keys.  Hashtable is O(n) in
063 * the worst case for all of these (if all keys hash to the same bucket).
064 * <p>
065 *
066 * This is a JDK-1.2 compliant implementation of Hashtable.  As such, it
067 * belongs, partially, to the Collections framework (in that it implements
068 * Map).  For backwards compatibility, it inherits from the obsolete and
069 * utterly useless Dictionary class.
070 * <p>
071 *
072 * Being a hybrid of old and new, Hashtable has methods which provide redundant
073 * capability, but with subtle and even crucial differences.
074 * For example, one can iterate over various aspects of a Hashtable with
075 * either an Iterator (which is the JDK-1.2 way of doing things) or with an
076 * Enumeration.  The latter can end up in an undefined state if the Hashtable
077 * changes while the Enumeration is open.
078 * <p>
079 *
080 * Unlike HashMap, Hashtable does not accept `null' as a key value. Also,
081 * all accesses are synchronized: in a single thread environment, this is
082 * expensive, but in a multi-thread environment, this saves you the effort
083 * of extra synchronization. However, the old-style enumerators are not
084 * synchronized, because they can lead to unspecified behavior even if
085 * they were synchronized. You have been warned.
086 * <p>
087 *
088 * The iterators are <i>fail-fast</i>, meaning that any structural
089 * modification, except for <code>remove()</code> called on the iterator
090 * itself, cause the iterator to throw a
091 * <code>ConcurrentModificationException</code> rather than exhibit
092 * non-deterministic behavior.
093 *
094 * @author Jon Zeppieri
095 * @author Warren Levy
096 * @author Bryce McKinlay
097 * @author Eric Blake (ebb9@email.byu.edu)
098 * @see HashMap
099 * @see TreeMap
100 * @see IdentityHashMap
101 * @see LinkedHashMap
102 * @since 1.0
103 * @status updated to 1.4
104 */
105public class Hashtable<K, V> extends Dictionary<K, V>
106  implements Map<K, V>, Cloneable, Serializable
107{
108  // WARNING: Hashtable is a CORE class in the bootstrap cycle. See the
109  // comments in vm/reference/java/lang/Runtime for implications of this fact.
110
111  /** Default number of buckets. This is the value the JDK 1.3 uses. Some
112   * early documentation specified this value as 101. That is incorrect.
113   */
114  private static final int DEFAULT_CAPACITY = 11;
115
116  /**
117   * The default load factor; this is explicitly specified by the spec.
118   */
119  private static final float DEFAULT_LOAD_FACTOR = 0.75f;
120
121  /**
122   * Compatible with JDK 1.0+.
123   */
124  private static final long serialVersionUID = 1421746759512286392L;
125
126  /**
127   * The rounded product of the capacity and the load factor; when the number
128   * of elements exceeds the threshold, the Hashtable calls
129   * <code>rehash()</code>.
130   * @serial
131   */
132  private int threshold;
133
134  /**
135   * Load factor of this Hashtable:  used in computing the threshold.
136   * @serial
137   */
138  private final float loadFactor;
139
140  /**
141   * Array containing the actual key-value mappings.
142   */
143  // Package visible for use by nested classes.
144  transient HashEntry<K, V>[] buckets;
145
146  /**
147   * Counts the number of modifications this Hashtable has undergone, used
148   * by Iterators to know when to throw ConcurrentModificationExceptions.
149   */
150  // Package visible for use by nested classes.
151  transient int modCount;
152
153  /**
154   * The size of this Hashtable:  denotes the number of key-value pairs.
155   */
156  // Package visible for use by nested classes.
157  transient int size;
158
159  /**
160   * The cache for {@link #keySet()}.
161   */
162  private transient Set<K> keys;
163
164  /**
165   * The cache for {@link #values()}.
166   */
167  private transient Collection<V> values;
168
169  /**
170   * The cache for {@link #entrySet()}.
171   */
172  private transient Set<Map.Entry<K, V>> entries;
173
174  /**
175   * Class to represent an entry in the hash table. Holds a single key-value
176   * pair. A Hashtable Entry is identical to a HashMap Entry, except that
177   * `null' is not allowed for keys and values.
178   */
179  private static final class HashEntry<K, V>
180    extends AbstractMap.SimpleEntry<K, V>
181  {
182    /** The next entry in the linked list. */
183    HashEntry<K, V> next;
184
185    /**
186     * Simple constructor.
187     * @param key the key, already guaranteed non-null
188     * @param value the value, already guaranteed non-null
189     */
190    HashEntry(K key, V value)
191    {
192      super(key, value);
193    }
194
195    /**
196     * Resets the value.
197     * @param newVal the new value
198     * @return the prior value
199     * @throws NullPointerException if <code>newVal</code> is null
200     */
201    public V setValue(V newVal)
202    {
203      if (newVal == null)
204        throw new NullPointerException();
205      return super.setValue(newVal);
206    }
207  }
208
209  /**
210   * Construct a new Hashtable with the default capacity (11) and the default
211   * load factor (0.75).
212   */
213  public Hashtable()
214  {
215    this(DEFAULT_CAPACITY, DEFAULT_LOAD_FACTOR);
216  }
217
218  /**
219   * Construct a new Hashtable from the given Map, with initial capacity
220   * the greater of the size of <code>m</code> or the default of 11.
221   * <p>
222   *
223   * Every element in Map m will be put into this new Hashtable.
224   *
225   * @param m a Map whose key / value pairs will be put into
226   *          the new Hashtable.  <b>NOTE: key / value pairs
227   *          are not cloned in this constructor.</b>
228   * @throws NullPointerException if m is null, or if m contains a mapping
229   *         to or from `null'.
230   * @since 1.2
231   */
232  public Hashtable(Map<? extends K, ? extends V> m)
233  {
234    this(Math.max(m.size() * 2, DEFAULT_CAPACITY), DEFAULT_LOAD_FACTOR);
235    putAll(m);
236  }
237
238  /**
239   * Construct a new Hashtable with a specific inital capacity and
240   * default load factor of 0.75.
241   *
242   * @param initialCapacity the initial capacity of this Hashtable (&gt;= 0)
243   * @throws IllegalArgumentException if (initialCapacity &lt; 0)
244   */
245  public Hashtable(int initialCapacity)
246  {
247    this(initialCapacity, DEFAULT_LOAD_FACTOR);
248  }
249
250  /**
251   * Construct a new Hashtable with a specific initial capacity and
252   * load factor.
253   *
254   * @param initialCapacity the initial capacity (&gt;= 0)
255   * @param loadFactor the load factor (&gt; 0, not NaN)
256   * @throws IllegalArgumentException if (initialCapacity &lt; 0) ||
257   *                                     ! (loadFactor &gt; 0.0)
258   */
259  public Hashtable(int initialCapacity, float loadFactor)
260  {
261    if (initialCapacity < 0)
262      throw new IllegalArgumentException("Illegal Capacity: "
263                                         + initialCapacity);
264    if (! (loadFactor > 0)) // check for NaN too
265      throw new IllegalArgumentException("Illegal Load: " + loadFactor);
266
267    if (initialCapacity == 0)
268      initialCapacity = 1;
269    buckets = (HashEntry<K, V>[]) new HashEntry[initialCapacity];
270    this.loadFactor = loadFactor;
271    threshold = (int) (initialCapacity * loadFactor);
272  }
273
274  /**
275   * Returns the number of key-value mappings currently in this hashtable.
276   * @return the size
277   */
278  public synchronized int size()
279  {
280    return size;
281  }
282
283  /**
284   * Returns true if there are no key-value mappings currently in this table.
285   * @return <code>size() == 0</code>
286   */
287  public synchronized boolean isEmpty()
288  {
289    return size == 0;
290  }
291
292  /**
293   * Return an enumeration of the keys of this table. There's no point
294   * in synchronizing this, as you have already been warned that the
295   * enumeration is not specified to be thread-safe.
296   *
297   * @return the keys
298   * @see #elements()
299   * @see #keySet()
300   */
301  public Enumeration<K> keys()
302  {
303    return new KeyEnumerator();
304  }
305
306  /**
307   * Return an enumeration of the values of this table. There's no point
308   * in synchronizing this, as you have already been warned that the
309   * enumeration is not specified to be thread-safe.
310   *
311   * @return the values
312   * @see #keys()
313   * @see #values()
314   */
315  public Enumeration<V> elements()
316  {
317    return new ValueEnumerator();
318  }
319
320  /**
321   * Returns true if this Hashtable contains a value <code>o</code>,
322   * such that <code>o.equals(value)</code>.  This is the same as
323   * <code>containsValue()</code>, and is O(n).
324   * <p>
325   *
326   * @param value the value to search for in this Hashtable
327   * @return true if at least one key maps to the value
328   * @throws NullPointerException if <code>value</code> is null
329   * @see #containsValue(Object)
330   * @see #containsKey(Object)
331   */
332  public synchronized boolean contains(Object value)
333  {
334    if (value == null)
335      throw new NullPointerException();
336
337    for (int i = buckets.length - 1; i >= 0; i--)
338      {
339        HashEntry<K, V> e = buckets[i];
340        while (e != null)
341          {
342            if (e.value.equals(value))
343              return true;
344            e = e.next;
345          }
346      }
347
348    return false;
349  }
350
351  /**
352   * Returns true if this Hashtable contains a value <code>o</code>, such that
353   * <code>o.equals(value)</code>. This is the new API for the old
354   * <code>contains()</code>.
355   *
356   * @param value the value to search for in this Hashtable
357   * @return true if at least one key maps to the value
358   * @see #contains(Object)
359   * @see #containsKey(Object)
360   * @throws NullPointerException if <code>value</code> is null
361   * @since 1.2
362   */
363  public boolean containsValue(Object value)
364  {
365    // Delegate to older method to make sure code overriding it continues
366    // to work.
367    return contains(value);
368  }
369
370  /**
371   * Returns true if the supplied object <code>equals()</code> a key
372   * in this Hashtable.
373   *
374   * @param key the key to search for in this Hashtable
375   * @return true if the key is in the table
376   * @throws NullPointerException if key is null
377   * @see #containsValue(Object)
378   */
379  public synchronized boolean containsKey(Object key)
380  {
381    int idx = hash(key);
382    HashEntry<K, V> e = buckets[idx];
383    while (e != null)
384      {
385        if (e.key.equals(key))
386          return true;
387        e = e.next;
388      }
389    return false;
390  }
391
392  /**
393   * Return the value in this Hashtable associated with the supplied key,
394   * or <code>null</code> if the key maps to nothing.
395   *
396   * @param key the key for which to fetch an associated value
397   * @return what the key maps to, if present
398   * @throws NullPointerException if key is null
399   * @see #put(Object, Object)
400   * @see #containsKey(Object)
401   */
402  public synchronized V get(Object key)
403  {
404    int idx = hash(key);
405    HashEntry<K, V> e = buckets[idx];
406    while (e != null)
407      {
408        if (e.key.equals(key))
409          return e.value;
410        e = e.next;
411      }
412    return null;
413  }
414
415  /**
416   * Puts the supplied value into the Map, mapped by the supplied key.
417   * Neither parameter may be null.  The value may be retrieved by any
418   * object which <code>equals()</code> this key.
419   *
420   * @param key the key used to locate the value
421   * @param value the value to be stored in the table
422   * @return the prior mapping of the key, or null if there was none
423   * @throws NullPointerException if key or value is null
424   * @see #get(Object)
425   * @see Object#equals(Object)
426   */
427  public synchronized V put(K key, V value)
428  {
429    int idx = hash(key);
430    HashEntry<K, V> e = buckets[idx];
431
432    // Check if value is null since it is not permitted.
433    if (value == null)
434      throw new NullPointerException();
435
436    while (e != null)
437      {
438        if (e.key.equals(key))
439          {
440            // Bypass e.setValue, since we already know value is non-null.
441            V r = e.value;
442            e.value = value;
443            return r;
444          }
445        else
446          {
447            e = e.next;
448          }
449      }
450
451    // At this point, we know we need to add a new entry.
452    modCount++;
453    if (++size > threshold)
454      {
455        rehash();
456        // Need a new hash value to suit the bigger table.
457        idx = hash(key);
458      }
459
460    e = new HashEntry<K, V>(key, value);
461
462    e.next = buckets[idx];
463    buckets[idx] = e;
464
465    return null;
466  }
467
468  /**
469   * Removes from the table and returns the value which is mapped by the
470   * supplied key. If the key maps to nothing, then the table remains
471   * unchanged, and <code>null</code> is returned.
472   *
473   * @param key the key used to locate the value to remove
474   * @return whatever the key mapped to, if present
475   */
476  public synchronized V remove(Object key)
477  {
478    int idx = hash(key);
479    HashEntry<K, V> e = buckets[idx];
480    HashEntry<K, V> last = null;
481
482    while (e != null)
483      {
484        if (e.key.equals(key))
485          {
486            modCount++;
487            if (last == null)
488              buckets[idx] = e.next;
489            else
490              last.next = e.next;
491            size--;
492            return e.value;
493          }
494        last = e;
495        e = e.next;
496      }
497    return null;
498  }
499
500  /**
501   * Copies all elements of the given map into this hashtable.  However, no
502   * mapping can contain null as key or value.  If this table already has
503   * a mapping for a key, the new mapping replaces the current one.
504   *
505   * @param m the map to be hashed into this
506   * @throws NullPointerException if m is null, or contains null keys or values
507   */
508  public synchronized void putAll(Map<? extends K, ? extends V> m)
509  {
510    final Map<K,V> addMap = (Map<K,V>) m;
511    final Iterator<Map.Entry<K,V>> it = addMap.entrySet().iterator();
512    while (it.hasNext())
513      {
514        final Map.Entry<K,V> e = it.next();
515        // Optimize in case the Entry is one of our own.
516        if (e instanceof AbstractMap.SimpleEntry)
517          {
518            AbstractMap.SimpleEntry<? extends K, ? extends V> entry
519              = (AbstractMap.SimpleEntry<? extends K, ? extends V>) e;
520            put(entry.key, entry.value);
521          }
522        else
523          {
524            put(e.getKey(), e.getValue());
525          }
526      }
527  }
528
529  /**
530   * Clears the hashtable so it has no keys.  This is O(1).
531   */
532  public synchronized void clear()
533  {
534    if (size > 0)
535      {
536        modCount++;
537        Arrays.fill(buckets, null);
538        size = 0;
539      }
540  }
541
542  /**
543   * Returns a shallow clone of this Hashtable. The Map itself is cloned,
544   * but its contents are not.  This is O(n).
545   *
546   * @return the clone
547   */
548  public synchronized Object clone()
549  {
550    Hashtable<K, V> copy = null;
551    try
552      {
553        copy = (Hashtable<K, V>) super.clone();
554      }
555    catch (CloneNotSupportedException x)
556      {
557        // This is impossible.
558      }
559    copy.buckets = (HashEntry<K, V>[]) new HashEntry[buckets.length];
560    copy.putAllInternal(this);
561    // Clear the caches.
562    copy.keys = null;
563    copy.values = null;
564    copy.entries = null;
565    return copy;
566  }
567
568  /**
569   * Converts this Hashtable to a String, surrounded by braces, and with
570   * key/value pairs listed with an equals sign between, separated by a
571   * comma and space. For example, <code>"{a=1, b=2}"</code>.<p>
572   *
573   * NOTE: if the <code>toString()</code> method of any key or value
574   * throws an exception, this will fail for the same reason.
575   *
576   * @return the string representation
577   */
578  public synchronized String toString()
579  {
580    // Since we are already synchronized, and entrySet().iterator()
581    // would repeatedly re-lock/release the monitor, we directly use the
582    // unsynchronized EntryIterator instead.
583    Iterator<Map.Entry<K, V>> entries = new EntryIterator();
584    CPStringBuilder r = new CPStringBuilder("{");
585    for (int pos = size; pos > 0; pos--)
586      {
587        r.append(entries.next());
588        if (pos > 1)
589          r.append(", ");
590      }
591    r.append("}");
592    return r.toString();
593  }
594
595  /**
596   * Returns a "set view" of this Hashtable's keys. The set is backed by
597   * the hashtable, so changes in one show up in the other.  The set supports
598   * element removal, but not element addition.  The set is properly
599   * synchronized on the original hashtable.  Sun has not documented the
600   * proper interaction of null with this set, but has inconsistent behavior
601   * in the JDK. Therefore, in this implementation, contains, remove,
602   * containsAll, retainAll, removeAll, and equals just ignore a null key
603   * rather than throwing a {@link NullPointerException}.
604   *
605   * @return a set view of the keys
606   * @see #values()
607   * @see #entrySet()
608   * @since 1.2
609   */
610  public Set<K> keySet()
611  {
612    if (keys == null)
613      {
614        // Create a synchronized AbstractSet with custom implementations of
615        // those methods that can be overridden easily and efficiently.
616        Set<K> r = new AbstractSet<K>()
617        {
618          public int size()
619          {
620            return size;
621          }
622
623          public Iterator<K> iterator()
624          {
625            return new KeyIterator();
626          }
627
628          public void clear()
629          {
630            Hashtable.this.clear();
631          }
632
633          public boolean contains(Object o)
634          {
635            if (o == null)
636              return false;
637            return containsKey(o);
638          }
639
640          public boolean remove(Object o)
641          {
642            return Hashtable.this.remove(o) != null;
643          }
644        };
645        // We must specify the correct object to synchronize upon, hence the
646        // use of a non-public API
647        keys = new Collections.SynchronizedSet<K>(this, r);
648      }
649    return keys;
650  }
651
652  /**
653   * Returns a "collection view" (or "bag view") of this Hashtable's values.
654   * The collection is backed by the hashtable, so changes in one show up
655   * in the other.  The collection supports element removal, but not element
656   * addition.  The collection is properly synchronized on the original
657   * hashtable.  Sun has not documented the proper interaction of null with
658   * this set, but has inconsistent behavior in the JDK. Therefore, in this
659   * implementation, contains, remove, containsAll, retainAll, removeAll, and
660   * equals just ignore a null value rather than throwing a
661   * {@link NullPointerException}.
662   *
663   * @return a bag view of the values
664   * @see #keySet()
665   * @see #entrySet()
666   * @since 1.2
667   */
668  public Collection<V> values()
669  {
670    if (values == null)
671      {
672        // We don't bother overriding many of the optional methods, as doing so
673        // wouldn't provide any significant performance advantage.
674        Collection<V> r = new AbstractCollection<V>()
675        {
676          public int size()
677          {
678            return size;
679          }
680
681          public Iterator<V> iterator()
682          {
683            return new ValueIterator();
684          }
685
686          public void clear()
687          {
688            Hashtable.this.clear();
689          }
690        };
691        // We must specify the correct object to synchronize upon, hence the
692        // use of a non-public API
693        values = new Collections.SynchronizedCollection<V>(this, r);
694      }
695    return values;
696  }
697
698  /**
699   * Returns a "set view" of this Hashtable's entries. The set is backed by
700   * the hashtable, so changes in one show up in the other.  The set supports
701   * element removal, but not element addition.  The set is properly
702   * synchronized on the original hashtable.  Sun has not documented the
703   * proper interaction of null with this set, but has inconsistent behavior
704   * in the JDK. Therefore, in this implementation, contains, remove,
705   * containsAll, retainAll, removeAll, and equals just ignore a null entry,
706   * or an entry with a null key or value, rather than throwing a
707   * {@link NullPointerException}. However, calling entry.setValue(null)
708   * will fail.
709   * <p>
710   *
711   * Note that the iterators for all three views, from keySet(), entrySet(),
712   * and values(), traverse the hashtable in the same sequence.
713   *
714   * @return a set view of the entries
715   * @see #keySet()
716   * @see #values()
717   * @see Map.Entry
718   * @since 1.2
719   */
720  public Set<Map.Entry<K, V>> entrySet()
721  {
722    if (entries == null)
723      {
724        // Create an AbstractSet with custom implementations of those methods
725        // that can be overridden easily and efficiently.
726        Set<Map.Entry<K, V>> r = new AbstractSet<Map.Entry<K, V>>()
727        {
728          public int size()
729          {
730            return size;
731          }
732
733          public Iterator<Map.Entry<K, V>> iterator()
734          {
735            return new EntryIterator();
736          }
737
738          public void clear()
739          {
740            Hashtable.this.clear();
741          }
742
743          public boolean contains(Object o)
744          {
745            return getEntry(o) != null;
746          }
747
748          public boolean remove(Object o)
749          {
750            HashEntry<K, V> e = getEntry(o);
751            if (e != null)
752              {
753                Hashtable.this.remove(e.key);
754                return true;
755              }
756            return false;
757          }
758        };
759        // We must specify the correct object to synchronize upon, hence the
760        // use of a non-public API
761        entries = new Collections.SynchronizedSet<Map.Entry<K, V>>(this, r);
762      }
763    return entries;
764  }
765
766  /**
767   * Returns true if this Hashtable equals the supplied Object <code>o</code>.
768   * As specified by Map, this is:
769   * <code>
770   * (o instanceof Map) && entrySet().equals(((Map) o).entrySet());
771   * </code>
772   *
773   * @param o the object to compare to
774   * @return true if o is an equal map
775   * @since 1.2
776   */
777  public boolean equals(Object o)
778  {
779    // no need to synchronize, entrySet().equals() does that.
780    if (o == this)
781      return true;
782    if (!(o instanceof Map))
783      return false;
784
785    return entrySet().equals(((Map) o).entrySet());
786  }
787
788  /**
789   * Returns the hashCode for this Hashtable.  As specified by Map, this is
790   * the sum of the hashCodes of all of its Map.Entry objects
791   *
792   * @return the sum of the hashcodes of the entries
793   * @since 1.2
794   */
795  public synchronized int hashCode()
796  {
797    // Since we are already synchronized, and entrySet().iterator()
798    // would repeatedly re-lock/release the monitor, we directly use the
799    // unsynchronized EntryIterator instead.
800    Iterator<Map.Entry<K, V>> itr = new EntryIterator();
801    int hashcode = 0;
802    for (int pos = size; pos > 0; pos--)
803      hashcode += itr.next().hashCode();
804
805    return hashcode;
806  }
807
808  /**
809   * Helper method that returns an index in the buckets array for `key'
810   * based on its hashCode().
811   *
812   * @param key the key
813   * @return the bucket number
814   * @throws NullPointerException if key is null
815   */
816  private int hash(Object key)
817  {
818    // Note: Inline Math.abs here, for less method overhead, and to avoid
819    // a bootstrap dependency, since Math relies on native methods.
820    int hash = key.hashCode() % buckets.length;
821    return hash < 0 ? -hash : hash;
822  }
823
824  /**
825   * Helper method for entrySet(), which matches both key and value
826   * simultaneously. Ignores null, as mentioned in entrySet().
827   *
828   * @param o the entry to match
829   * @return the matching entry, if found, or null
830   * @see #entrySet()
831   */
832  // Package visible, for use in nested classes.
833  HashEntry<K, V> getEntry(Object o)
834  {
835    if (! (o instanceof Map.Entry))
836      return null;
837    K key = ((Map.Entry<K, V>) o).getKey();
838    if (key == null)
839      return null;
840
841    int idx = hash(key);
842    HashEntry<K, V> e = buckets[idx];
843    while (e != null)
844      {
845        if (e.equals(o))
846          return e;
847        e = e.next;
848      }
849    return null;
850  }
851
852  /**
853   * A simplified, more efficient internal implementation of putAll(). clone()
854   * should not call putAll or put, in order to be compatible with the JDK
855   * implementation with respect to subclasses.
856   *
857   * @param m the map to initialize this from
858   */
859  void putAllInternal(Map<? extends K, ? extends V> m)
860  {
861    final Map<K,V> addMap = (Map<K,V>) m;
862    final Iterator<Map.Entry<K,V>> it = addMap.entrySet().iterator();
863    size = 0;
864    while (it.hasNext())
865      {
866        final Map.Entry<K,V> e = it.next();
867        size++;
868        K key = e.getKey();
869        int idx = hash(key);
870        HashEntry<K, V> he = new HashEntry<K, V>(key, e.getValue());
871        he.next = buckets[idx];
872        buckets[idx] = he;
873      }
874  }
875
876  /**
877   * Increases the size of the Hashtable and rehashes all keys to new array
878   * indices; this is called when the addition of a new value would cause
879   * size() &gt; threshold. Note that the existing Entry objects are reused in
880   * the new hash table.
881   * <p>
882   *
883   * This is not specified, but the new size is twice the current size plus
884   * one; this number is not always prime, unfortunately. This implementation
885   * is not synchronized, as it is only invoked from synchronized methods.
886   */
887  protected void rehash()
888  {
889    HashEntry<K, V>[] oldBuckets = buckets;
890
891    int newcapacity = (buckets.length * 2) + 1;
892    threshold = (int) (newcapacity * loadFactor);
893    buckets = (HashEntry<K, V>[]) new HashEntry[newcapacity];
894
895    for (int i = oldBuckets.length - 1; i >= 0; i--)
896      {
897        HashEntry<K, V> e = oldBuckets[i];
898        while (e != null)
899          {
900            int idx = hash(e.key);
901            HashEntry<K, V> dest = buckets[idx];
902
903            if (dest != null)
904              {
905                HashEntry next = dest.next;
906                while (next != null)
907                  {
908                    dest = next;
909                    next = dest.next;
910                  }
911                dest.next = e;
912              }
913            else
914              {
915                buckets[idx] = e;
916              }
917
918            HashEntry<K, V> next = e.next;
919            e.next = null;
920            e = next;
921          }
922      }
923  }
924
925  /**
926   * Serializes this object to the given stream.
927   *
928   * @param s the stream to write to
929   * @throws IOException if the underlying stream fails
930   * @serialData the <i>capacity</i> (int) that is the length of the
931   *             bucket array, the <i>size</i> (int) of the hash map
932   *             are emitted first.  They are followed by size entries,
933   *             each consisting of a key (Object) and a value (Object).
934   */
935  private synchronized void writeObject(ObjectOutputStream s)
936    throws IOException
937  {
938    // Write the threshold and loadFactor fields.
939    s.defaultWriteObject();
940
941    s.writeInt(buckets.length);
942    s.writeInt(size);
943    // Since we are already synchronized, and entrySet().iterator()
944    // would repeatedly re-lock/release the monitor, we directly use the
945    // unsynchronized EntryIterator instead.
946    Iterator<Map.Entry<K, V>> it = new EntryIterator();
947    while (it.hasNext())
948      {
949        HashEntry<K, V> entry = (HashEntry<K, V>) it.next();
950        s.writeObject(entry.key);
951        s.writeObject(entry.value);
952      }
953  }
954
955  /**
956   * Deserializes this object from the given stream.
957   *
958   * @param s the stream to read from
959   * @throws ClassNotFoundException if the underlying stream fails
960   * @throws IOException if the underlying stream fails
961   * @serialData the <i>capacity</i> (int) that is the length of the
962   *             bucket array, the <i>size</i> (int) of the hash map
963   *             are emitted first.  They are followed by size entries,
964   *             each consisting of a key (Object) and a value (Object).
965   */
966  private void readObject(ObjectInputStream s)
967    throws IOException, ClassNotFoundException
968  {
969    // Read the threshold and loadFactor fields.
970    s.defaultReadObject();
971
972    // Read and use capacity.
973    buckets = (HashEntry<K, V>[]) new HashEntry[s.readInt()];
974    int len = s.readInt();
975
976    // Read and use key/value pairs.
977    // TODO: should we be defensive programmers, and check for illegal nulls?
978    while (--len >= 0)
979      put((K) s.readObject(), (V) s.readObject());
980  }
981
982  /**
983   * A class which implements the Iterator interface and is used for
984   * iterating over Hashtables.
985   * This implementation iterates entries. Subclasses are used to
986   * iterate key and values. It also allows the removal of elements,
987   * as per the Javasoft spec.  Note that it is not synchronized; this
988   * is a performance enhancer since it is never exposed externally
989   * and is only used within synchronized blocks above.
990   *
991   * @author Jon Zeppieri
992   * @author Fridjof Siebert
993   */
994  private class EntryIterator
995      implements Iterator<Entry<K,V>>
996  {
997    /**
998     * The number of modifications to the backing Hashtable that we know about.
999     */
1000    int knownMod = modCount;
1001    /** The number of elements remaining to be returned by next(). */
1002    int count = size;
1003    /** Current index in the physical hash table. */
1004    int idx = buckets.length;
1005    /** The last Entry returned by a next() call. */
1006    HashEntry<K, V> last;
1007    /**
1008     * The next entry that should be returned by next(). It is set to something
1009     * if we're iterating through a bucket that contains multiple linked
1010     * entries. It is null if next() needs to find a new bucket.
1011     */
1012    HashEntry<K, V> next;
1013
1014    /**
1015     * Construct a new EntryIterator
1016     */
1017    EntryIterator()
1018    {
1019    }
1020
1021
1022    /**
1023     * Returns true if the Iterator has more elements.
1024     * @return true if there are more elements
1025     */
1026    public boolean hasNext()
1027    {
1028      return count > 0;
1029    }
1030
1031    /**
1032     * Returns the next element in the Iterator's sequential view.
1033     * @return the next element
1034     * @throws ConcurrentModificationException if the hashtable was modified
1035     * @throws NoSuchElementException if there is none
1036     */
1037    public Map.Entry<K,V> next()
1038    {
1039      if (knownMod != modCount)
1040        throw new ConcurrentModificationException();
1041      if (count == 0)
1042        throw new NoSuchElementException();
1043      count--;
1044      HashEntry<K, V> e = next;
1045
1046      while (e == null)
1047        if (idx <= 0)
1048          return null;
1049        else
1050          e = buckets[--idx];
1051
1052      next = e.next;
1053      last = e;
1054      return e;
1055    }
1056
1057    /**
1058     * Removes from the backing Hashtable the last element which was fetched
1059     * with the <code>next()</code> method.
1060     * @throws ConcurrentModificationException if the hashtable was modified
1061     * @throws IllegalStateException if called when there is no last element
1062     */
1063    public void remove()
1064    {
1065      if (knownMod != modCount)
1066        throw new ConcurrentModificationException();
1067      if (last == null)
1068        throw new IllegalStateException();
1069
1070      Hashtable.this.remove(last.key);
1071      last = null;
1072      knownMod++;
1073    }
1074  } // class EntryIterator
1075
1076  /**
1077   * A class which implements the Iterator interface and is used for
1078   * iterating over keys in Hashtables.  This class uses an
1079   * <code>EntryIterator</code> to obtain the keys of each entry.
1080   *
1081   * @author Fridtjof Siebert
1082   * @author Andrew John Hughes (gnu_andrew@member.fsf.org)
1083   */
1084  private class KeyIterator
1085      implements Iterator<K>
1086  {
1087
1088    /**
1089     * This entry iterator is used for most operations.  Only
1090     * <code>next()</code> gives a different result, by returning just
1091     * the key rather than the whole element.
1092     */
1093    private final EntryIterator iterator;
1094
1095    /**
1096     * Construct a new KeyIterator
1097     */
1098    KeyIterator()
1099    {
1100        iterator = new EntryIterator();
1101    }
1102
1103
1104    /**
1105     * Returns true if the entry iterator has more elements.
1106     *
1107     * @return true if there are more elements
1108     * @throws ConcurrentModificationException if the hashtable was modified
1109     */
1110    public boolean hasNext()
1111    {
1112        return iterator.hasNext();
1113    }
1114
1115    /**
1116     * Returns the next element in the Iterator's sequential view.
1117     *
1118     * @return the next element
1119     *
1120     * @throws ConcurrentModificationException if the hashtable was modified
1121     * @throws NoSuchElementException if there is none
1122     */
1123    public K next()
1124    {
1125      return ((HashEntry<K,V>) iterator.next()).key;
1126    }
1127
1128    /**
1129     * Removes the last element used by the <code>next()</code> method
1130     * using the entry iterator.
1131     *
1132     * @throws ConcurrentModificationException if the hashtable was modified
1133     * @throws IllegalStateException if called when there is no last element
1134     */
1135    public void remove()
1136    {
1137      iterator.remove();
1138    }
1139  } // class KeyIterator
1140
1141  /**
1142   * A class which implements the Iterator interface and is used for
1143   * iterating over values in Hashtables.  This class uses an
1144   * <code>EntryIterator</code> to obtain the values of each entry.
1145   *
1146   * @author Fridtjof Siebert
1147   * @author Andrew John Hughes (gnu_andrew@member.fsf.org)
1148   */
1149  private class ValueIterator
1150      implements Iterator<V>
1151  {
1152
1153    /**
1154     * This entry iterator is used for most operations.  Only
1155     * <code>next()</code> gives a different result, by returning just
1156     * the value rather than the whole element.
1157     */
1158    private final EntryIterator iterator;
1159
1160    /**
1161     * Construct a new KeyIterator
1162     */
1163    ValueIterator()
1164    {
1165        iterator = new EntryIterator();
1166    }
1167
1168
1169    /**
1170     * Returns true if the entry iterator has more elements.
1171     *
1172     * @return true if there are more elements
1173     * @throws ConcurrentModificationException if the hashtable was modified
1174     */
1175    public boolean hasNext()
1176    {
1177        return iterator.hasNext();
1178    }
1179
1180    /**
1181     * Returns the value of the next element in the iterator's sequential view.
1182     *
1183     * @return the next value
1184     *
1185     * @throws ConcurrentModificationException if the hashtable was modified
1186     * @throws NoSuchElementException if there is none
1187     */
1188    public V next()
1189    {
1190      return ((HashEntry<K,V>) iterator.next()).value;
1191    }
1192
1193    /**
1194     * Removes the last element used by the <code>next()</code> method
1195     * using the entry iterator.
1196     *
1197     * @throws ConcurrentModificationException if the hashtable was modified
1198     * @throws IllegalStateException if called when there is no last element
1199     */
1200    public void remove()
1201    {
1202      iterator.remove();
1203    }
1204
1205  } // class ValueIterator
1206
1207  /**
1208   * Enumeration view of the entries in this Hashtable, providing
1209   * sequential access to its elements.
1210   *
1211   * <b>NOTE</b>: Enumeration is not safe if new elements are put in the table
1212   * as this could cause a rehash and we'd completely lose our place.  Even
1213   * without a rehash, it is undetermined if a new element added would
1214   * appear in the enumeration.  The spec says nothing about this, but
1215   * the "Java Class Libraries" book implies that modifications to the
1216   * hashtable during enumeration causes indeterminate results.  Don't do it!
1217   *
1218   * @author Jon Zeppieri
1219   * @author Fridjof Siebert
1220   */
1221  private class EntryEnumerator
1222      implements Enumeration<Entry<K,V>>
1223  {
1224    /** The number of elements remaining to be returned by next(). */
1225    int count = size;
1226    /** Current index in the physical hash table. */
1227    int idx = buckets.length;
1228    /**
1229     * Entry which will be returned by the next nextElement() call. It is
1230     * set if we are iterating through a bucket with multiple entries, or null
1231     * if we must look in the next bucket.
1232     */
1233    HashEntry<K, V> next;
1234
1235    /**
1236     * Construct the enumeration.
1237     */
1238    EntryEnumerator()
1239    {
1240      // Nothing to do here.
1241    }
1242
1243    /**
1244     * Checks whether more elements remain in the enumeration.
1245     * @return true if nextElement() will not fail.
1246     */
1247    public boolean hasMoreElements()
1248    {
1249      return count > 0;
1250    }
1251
1252    /**
1253     * Returns the next element.
1254     * @return the next element
1255     * @throws NoSuchElementException if there is none.
1256     */
1257    public Map.Entry<K,V> nextElement()
1258    {
1259      if (count == 0)
1260        throw new NoSuchElementException("Hashtable Enumerator");
1261      count--;
1262      HashEntry<K, V> e = next;
1263
1264      while (e == null)
1265        if (idx <= 0)
1266          return null;
1267        else
1268          e = buckets[--idx];
1269
1270      next = e.next;
1271      return e;
1272    }
1273  } // class EntryEnumerator
1274
1275
1276  /**
1277   * Enumeration view of this Hashtable, providing sequential access to its
1278   * elements.
1279   *
1280   * <b>NOTE</b>: Enumeration is not safe if new elements are put in the table
1281   * as this could cause a rehash and we'd completely lose our place.  Even
1282   * without a rehash, it is undetermined if a new element added would
1283   * appear in the enumeration.  The spec says nothing about this, but
1284   * the "Java Class Libraries" book implies that modifications to the
1285   * hashtable during enumeration causes indeterminate results.  Don't do it!
1286   *
1287   * @author Jon Zeppieri
1288   * @author Fridjof Siebert
1289   * @author Andrew John Hughes (gnu_andrew@member.fsf.org)
1290   */
1291  private final class KeyEnumerator
1292      implements Enumeration<K>
1293  {
1294    /**
1295     * This entry enumerator is used for most operations.  Only
1296     * <code>nextElement()</code> gives a different result, by returning just
1297     * the key rather than the whole element.
1298     */
1299    private final EntryEnumerator enumerator;
1300
1301    /**
1302     * Construct a new KeyEnumerator
1303     */
1304    KeyEnumerator()
1305    {
1306      enumerator = new EntryEnumerator();
1307    }
1308
1309
1310    /**
1311     * Returns true if the entry enumerator has more elements.
1312     *
1313     * @return true if there are more elements
1314     * @throws ConcurrentModificationException if the hashtable was modified
1315     */
1316    public boolean hasMoreElements()
1317    {
1318        return enumerator.hasMoreElements();
1319    }
1320
1321    /**
1322     * Returns the next element.
1323     * @return the next element
1324     * @throws NoSuchElementException if there is none.
1325     */
1326    public K nextElement()
1327    {
1328      HashEntry<K,V> entry = (HashEntry<K,V>) enumerator.nextElement();
1329      K retVal = null;
1330      if (entry != null)
1331        retVal = entry.key;
1332      return retVal;
1333    }
1334  } // class KeyEnumerator
1335
1336
1337  /**
1338   * Enumeration view of this Hashtable, providing sequential access to its
1339   * values.
1340   *
1341   * <b>NOTE</b>: Enumeration is not safe if new elements are put in the table
1342   * as this could cause a rehash and we'd completely lose our place.  Even
1343   * without a rehash, it is undetermined if a new element added would
1344   * appear in the enumeration.  The spec says nothing about this, but
1345   * the "Java Class Libraries" book implies that modifications to the
1346   * hashtable during enumeration causes indeterminate results.  Don't do it!
1347   *
1348   * @author Jon Zeppieri
1349   * @author Fridjof Siebert
1350   * @author Andrew John Hughes (gnu_andrew@member.fsf.org)
1351   */
1352  private final class ValueEnumerator
1353      implements Enumeration<V>
1354  {
1355    /**
1356     * This entry enumerator is used for most operations.  Only
1357     * <code>nextElement()</code> gives a different result, by returning just
1358     * the value rather than the whole element.
1359     */
1360    private final EntryEnumerator enumerator;
1361
1362    /**
1363     * Construct a new ValueEnumerator
1364     */
1365    ValueEnumerator()
1366    {
1367      enumerator = new EntryEnumerator();
1368    }
1369
1370
1371    /**
1372     * Returns true if the entry enumerator has more elements.
1373     *
1374     * @return true if there are more elements
1375     * @throws ConcurrentModificationException if the hashtable was modified
1376     */
1377    public boolean hasMoreElements()
1378    {
1379        return enumerator.hasMoreElements();
1380    }
1381
1382    /**
1383     * Returns the next element.
1384     * @return the next element
1385     * @throws NoSuchElementException if there is none.
1386     */
1387    public V nextElement()
1388    {
1389      HashEntry<K,V> entry = (HashEntry<K,V>) enumerator.nextElement();
1390      V retVal = null;
1391      if (entry != null)
1392        retVal = entry.value;
1393      return retVal;
1394    }
1395  } // class ValueEnumerator
1396
1397} // class Hashtable