*   if (value < 0) {
 *     throw new IllegalArgumentException("input is negative: " + value);
 *   }
 *   // calculate square root
 * }
 * }</pre>
 *
 * <p>to be replaced with the more compact
 *
 * <pre>{@code
 * public static double sqrt(double value) {
 *   checkArgument(value >= 0, "input is negative: %s", value);
 *   // calculate square root
 * }
 * }</pre>
 *

 * that {@linkplain #mightContain(Object)} will erroneously return {@code true} for an object that
 * has not actually been put in the {@code BloomFilter}.
 *
 * <p>Bloom filters are serializable. They also support a more compact serial representation via the
 * {@link #writeTo} and {@link #readFrom} methods. Both serialized forms will continue to be
 * supported by future versions of this library. However, serial forms generated by newer versions

    }
  }

  Set<K> keySet;

  @Override
  public Set<K> keySet() {
    // does not impact recency ordering
    Set<K> ks = keySet;
    return (ks != null) ? ks : (keySet = new KeySet(this));
  }

  Collection<V> values;

  @Override
  public Collection<V> values() {
    // does not impact recency ordering
    Collection<V> vs = values;
    return (vs != null) ? vs : (values = new Values(this));

 * <p>This implementation is based on the <a
 * href="http://portal.acm.org/citation.cfm?id=6621">min-max heap</a> developed by Atkinson, et al.
 * Unlike many other double-ended priority queues, it stores elements in a single array, as compact
 * as the traditional heap data structure used in {@link PriorityQueue}.
 *
 * <p>This class is not thread-safe, and does not accept null elements.
 *
 * <p><i>Performance notes:</i>
 *
 * <ul>

 *   if (value < 0) {
 *     throw new IllegalArgumentException("input is negative: " + value);
 *   }
 *   // calculate square root
 * }
 * }</pre>
 *
 * <p>to be replaced with the more compact
 *
 * <pre>{@code
 * public static double sqrt(double value) {
 *   checkArgument(value >= 0, "input is negative: %s", value);
 *   // calculate square root
 * }
 * }</pre>
 *

   * other. But Atomic objects residing in arrays will tend to be
   * placed adjacent to each other, and so will most often share
   * cache lines (with a huge negative performance impact) without
   * this precaution.
   *
   * In part because Cells are relatively large, we avoid creating
   * them until they are needed.  When there is no contention, all

  public boolean containsKey(@CheckForNull Object key) {
    // does not impact recency ordering
    if (key == null) {
      return false;
    }
    int hash = hash(key);
    return segmentFor(hash).containsKey(key, hash);
  }

  @Override
  public boolean containsValue(@CheckForNull Object value) {
    // does not impact recency ordering
    if (value == null) {
      return false;
    }

   */
  public static boolean isUriInetAddress(String ipString) {
    return forUriStringNoThrow(ipString) != null;
  }

  /**
   * Evaluates whether the argument is an IPv6 "compat" address.
   *
   * <p>An "IPv4 compatible", or "compat", address is one with 96 leading bits of zero, with the
   * remaining 32 bits interpreted as an IPv4 address. These are conventionally represented in

   * other. But Atomic objects residing in arrays will tend to be
   * placed adjacent to each other, and so will most often share
   * cache lines (with a huge negative performance impact) without
   * this precaution.
   *
   * In part because Cells are relatively large, we avoid creating
   * them until they are needed.  When there is no contention, all

    int hash = hash(key);
    return segmentFor(hash).get(key, hash);
  }

  /**
   * Returns the internal entry for the specified key. The entry may be computing or partially
   * collected. Does not impact recency ordering.
   */
  @CheckForNull
  E getEntry(@CheckForNull Object key) {
    if (key == null) {
      return null;
    }
    int hash = hash(key);