object.java

gcc的组建

   * @see #wait(long)
   * @see Thread
   */
  public final native void wait(long timeout, int nanos)
    throws InterruptedException;

  /**
   * Determine whether this Object is semantically equal
   * to another Object.
   *
   * <p>There are some fairly strict requirements on this
   * method which subclasses must follow:<br>
   * <ul>
   * <li>It must be transitive.  If <code>a.equals(b)</code> and
   *     <code>b.equals(c)</code>, then <code>a.equals(c)</code>
   *     must be true as well.</li>
   * <li>It must be symmetric.  <code>a.equals(b)</code> and
   *     <code>b.equals(a)</code> must have the same value.</li>
   * <li>It must be reflexive.  <code>a.equals(a)</code> must
   *     always be true.</li>
   * <li>It must be consistent.  Whichever value a.equals(b)
   *     returns on the first invocation must be the value
   *     returned on all later invocations.</li>
   * <li><code>a.equals(null)</code> must be false.</li>
   * <li>It must be consistent with hashCode().  That is,
   *     <code>a.equals(b)</code> must imply
   *     <code>a.hashCode() == b.hashCode()</code>.
   *     The reverse is not true; two objects that are not
   *     equal may have the same hashcode, but that has
   *     the potential to harm hashing performance.</li>
   * </ul>
   *
   * <p>This is typically overridden to throw a {@link ClassCastException}
   * if the argument is not comparable to the class performing
   * the comparison, but that is not a requirement.  It is legal
   * for <code>a.equals(b)</code> to be true even though
   * <code>a.getClass() != b.getClass()</code>.  Also, it
   * is typical to never cause a {@link NullPointerException}.
   *
   * <p>In general, the Collections API ({@link java.util}) use the
   * <code>equals</code> method rather than the <code>==</code>
   * operator to compare objects.  However, {@link java.util.IdentityHashMap}
   * is an exception to this rule, for its own good reasons.
   *
   * <p>The default implementation returns <code>this == o</code>.
   *
   * @param obj the Object to compare to
   * @return whether this Object is semantically equal to another
   * @see #hashCode()
   */
  public boolean equals(Object obj)
  {
    return this == obj;
  }

  /**
   * The basic constructor.  Object is special, because it has no
   * superclass, so there is no call to super().
   *
   * @throws OutOfMemoryError Technically, this constructor never
   *         throws an OutOfMemoryError, because the memory has
   *         already been allocated by this point.  But as all
   *         instance creation expressions eventually trace back
   *         to this constructor, and creating an object allocates
   *         memory, we list that possibility here.
   */
  public Object()
  {
  }

  /**
   * Convert this Object to a human-readable String.
   * There are no limits placed on how long this String
   * should be or what it should contain.  We suggest you
   * make it as intuitive as possible to be able to place
   * it into {@link java.io.PrintStream#println() System.out.println()}
   * and such.
   *
   * <p>It is typical, but not required, to ensure that this method
   * never completes abruptly with a {@link RuntimeException}.
   *
   * <p>This method will be called when performing string
   * concatenation with this object.  If the result is
   * <code>null</code>, string concatenation will instead
   * use <code>"null"</code>.
   *
   * <p>The default implementation returns
   * <code>getClass().getName() + "@" +
   *      Integer.toHexString(hashCode())</code>.
   *
   * @return the String representing this Object, which may be null
   * @throws OutOfMemoryError The default implementation creates a new
   *         String object, therefore it must allocate memory
   * @see #getClass()
   * @see #hashCode()
   * @see Class#getName()
   * @see Integer#toHexString(int)
   */
  public String toString()
  {
    return getClass().getName() + '@' + Integer.toHexString(hashCode());
  }

  /**
   * Waits indefinitely for notify() or notifyAll() to be
   * called on the Object in question.  Implementation is
   * identical to wait(0).
   *
   * <p>The Thread that calls wait must have a lock on this Object,
   * obtained by a <code>synchronized</code> method or statement.
   * After calling wait, the thread loses the lock on this
   * object until the method completes (abruptly or normally),
   * at which time it regains the lock.  All locks held on
   * other objects remain in force, even though the thread is
   * inactive. Therefore, caution must be used to avoid deadlock.
   *
   * <p>While it is typical that this method will complete abruptly
   * with an {@link InterruptedException}, it is not guaranteed.  So,
   * it is typical to call wait inside an infinite loop:<br>
   *
   * <pre>
   * try
   *   {
   *     while (true)
   *       lock.wait();
   *   }
   * catch (InterruptedException e)
   *   {
   *   }
   * </pre>
   *
   * @throws IllegalMonitorStateException if this Thread
   *         does not own a lock on this Object
   * @throws InterruptedException if some other Thread
   *         interrupts this Thread
   * @see #notify()
   * @see #notifyAll()
   * @see #wait(long)
   * @see #wait(long, int)
   * @see Thread
   */
  public final void wait() throws InterruptedException
  {
    wait(0, 0);
  }

  /**
   * Waits a specified amount of time (or indefinitely if
   * the time specified is 0) for someone to call notify()
   * or notifyAll() on this Object, waking up this Thread.
   *
   * <p>The Thread that calls wait must have a lock on this Object,
   * obtained by a <code>synchronized</code> method or statement.
   * After calling wait, the thread loses the lock on this
   * object until the method completes (abruptly or normally),
   * at which time it regains the lock.  All locks held on
   * other objects remain in force, even though the thread is
   * inactive. Therefore, caution must be used to avoid deadlock.
   *
   * <p>Usually, this call will complete normally if the time
   * expires, or abruptly with {@link InterruptedException}
   * if another thread called notify, but neither result
   * is guaranteed.
   *
   * <p>The waiting period is only *roughly* the amount of time
   * you requested.  It cannot be exact because of the overhead
   * of the call itself.  Most Virtual Machiness treat the
   * argument as a lower limit on the time spent waiting, but
   * even that is not guaranteed.  Besides, some other thread
   * may hold the lock on the object when the time expires, so
   * the current thread may still have to wait to reobtain the
   * lock.
   *
   * @param timeout the minimum number of milliseconds to wait (1000
   *        milliseconds = 1 second), or 0 for an indefinite wait
   * @throws IllegalArgumentException if ms &lt; 0
   * @throws IllegalMonitorStateException if this Thread
   *         does not own a lock on this Object
   * @throws InterruptedException if some other Thread
   *         interrupts this Thread
   * @see #notify()
   * @see #notifyAll()
   * @see #wait()
   * @see #wait(long, int)
   * @see Thread
   */
  public final void wait(long timeout) throws InterruptedException
  {
    wait(timeout, 0);
  }

  /**
   * This method may be called to create a new copy of the
   * Object.  The typical behavior is as follows:<br>
   * <ul>
   *  <li><code>o == o.clone()</code> is false</li>
   *  <li><code>o.getClass() == o.clone().getClass()</code>
   *      is true</li>
   *  <li><code>o.equals(o)</code> is true</li>
   * </ul>
   *
   * <p>However, these are not strict requirements, and may
   * be violated if necessary.  Of the three requirements, the
   * last is the most commonly violated, particularly if the
   * subclass does not override {@link #equals(Object)}.
   *
   * <p>If the Object you call clone() on does not implement
   * {@link Cloneable} (which is a placeholder interface), then
   * a CloneNotSupportedException is thrown.  Notice that
   * Object does not implement Cloneable; this method exists
   * as a convenience for subclasses that do.
   *
   * <p>Object's implementation of clone allocates space for the
   * new Object using the correct class, without calling any
   * constructors, and then fills in all of the new field values
   * with the old field values.  Thus, it is a shallow copy.
   * However, subclasses are permitted to make a deep copy.
   *
   * <p>All array types implement Cloneable, and override
   * this method as follows (it should never fail):<br>
   * <pre>
   * public Object clone()
   * {
   *   try
   *     {
   *       super.clone();
   *     }
   *   catch (CloneNotSupportedException e)
   *     {
   *       throw new InternalError(e.getMessage());
   *     }
   * }
   * </pre>
   *
   * @return a copy of the Object
   * @throws CloneNotSupportedException If this Object does not
   *         implement Cloneable
   * @throws OutOfMemoryError Since cloning involves memory allocation,
   *         even though it may bypass constructors, you might run
   *         out of memory
   * @see Cloneable
   */
  protected native Object clone() throws CloneNotSupportedException;

  // This initializes the sync_info member.  It is here for
  // completeness (some day we'll be able to auto-generate Object.h).
  private final native void sync_init();

  // If we fail to find a method at class loading time we put the
  // vtable index of this method in its place: any attempt to call
  // that method will result in an error.
  void throwNoSuchMethodError()
  {
    throw new NoSuchMethodError("in " + getClass());
  }

  // Note that we don't mention the sync_info field here.  If we do,
  // jc1 will not work correctly.
}