*       worse, and possibly better, than creating a mutable collection and copying it.
 *   <li>Implementations generally do not cache hash codes. If your element or key type has a slow
 *       {@code hashCode} implementation, it should cache it itself.
 * </ul>
 *
 * <h4>Example usage</h4>
 *
 * <pre>{@code
 * class Foo {

      }
      return max(estimatedSize, 0);
    }

    @Override
    public final int characteristics() {
      return characteristics;
    }
  }

  /**
   * Implementation of {@link Stream#flatMap} with an object spliterator output type.
   *
   * <p>To avoid having this type, we could use {@code FlatMapSpliterator} directly. The main

   * as characters using the given {@link Charset}.
   *
   * <p>If {@link CharSource#asByteSource} is called on the returned source with the same charset,
   * the default implementation of this method will ensure that the original {@code ByteSource} is
   * returned, rather than round-trip encoding. Subclasses that override this method should behave
   * the same way.
   */

   * ImmutableEnumSet/ImmutableSortedSet, regardless of implementation type. It
   * captures their "logical contents" and they are reconstructed using public
   * static factories. This is necessary to ensure that the existence of a
   * particular implementation type is an implementation detail.
   */
  @J2ktIncompatible // serialization

        }
      }
    }
    return -1;
  }

  /** Returns an implementation of {@link List#listIterator(int)}. */
  static <E extends @Nullable Object> ListIterator<E> listIteratorImpl(List<E> list, int index) {
    return new AbstractListWrapper<>(list).listIterator(index);
  }

  /** An implementation of {@link List#subList(int, int)}. */
  static <E extends @Nullable Object> List<E> subListImpl(

 * this case, you should override {@code add(Object)} as well, either providing your own
 * implementation, or delegating to the provided {@code standardAdd} method.
 *
 * <p><b>{@code default} method warning:</b> This class does <i>not</i> forward calls to {@code
 * default} methods. Instead, it inherits their default implementations. When those implementations
 * invoke methods, they invoke methods on the {@code ForwardingMultiset}.
 *

    // TODO(kevinb): if we change this implementation, add full unit tests.
    return this.<E>reverse().leastOf(iterable, k);
  }

  /**
   * Returns the {@code k} greatest elements from the given iterator according to this ordering, in
   * order from greatest to least. If there are fewer than {@code k} elements present, all will be
   * included.
   *

    // The Unsafe implementation if it's available. Otherwise, the Java implementation.
    Comparator<byte[]> comparator = UnsignedBytes.lexicographicalComparator();
    Helpers.testComparator(comparator, ordered);
    assertThat(SerializableTester.reserialize(comparator)).isSameInstanceAs(comparator);

    // The Java implementation.

              @Override
              public String pendingToString() {
                throw new RuntimeException("I'm a misbehaving implementation");
              }
            }.toString())
        .matches(
            "[^\\[]+\\[status=PENDING, info=\\[Exception thrown from implementation: "
                + "class java.lang.RuntimeException\\]\\]");
  }

  public void testCompletionFinishesWithDone() {

   * ImmutableEnumSet/ImmutableSortedSet, regardless of implementation type. It
   * captures their "logical contents" and they are reconstructed using public
   * static factories. This is necessary to ensure that the existence of a
   * particular implementation type is an implementation detail.
   */
  @J2ktIncompatible // serialization