return CollectCollectors.toImmutableSortedMap(
        comparator, keyFunction, valueFunction, mergeFunction);
  }

  /*
   * TODO(kevinb): Confirm that ImmutableSortedMap is faster to construct and
   * uses less memory than TreeMap; then say so in the class Javadoc.
   */
  private static final Comparator<?> NATURAL_ORDER = Ordering.natural();

@ElementTypesAreNonnullByDefault
public abstract class AbstractFuture<V extends @Nullable Object> extends InternalFutureFailureAccess
    implements ListenableFuture<V> {
  // NOTE: Whenever both tests are cheap and functional, it's faster to use &, | instead of &&, ||

  static final boolean GENERATE_CANCELLATION_CAUSES;

  static {
    // System.getProperty may throw if the security policy does not permit access.
    boolean generateCancellationCauses;

    return CollectCollectors.toImmutableSortedMap(
        comparator, keyFunction, valueFunction, mergeFunction);
  }

  /*
   * TODO(kevinb): Confirm that ImmutableSortedMap is faster to construct and
   * uses less memory than TreeMap; then say so in the class Javadoc.
   */
  private static final Comparator<?> NATURAL_ORDER = Ordering.natural();

   * query than the original; your mileage may vary. Precomputation takes time and is likely to be
   * worthwhile only if the precomputed matcher is queried many thousands of times.
   *
   * <p>This method has no effect (returns {@code this}) when called in GWT: it's unclear whether a
   * precomputed matcher is faster, but it certainly consumes more memory, which doesn't seem like a

   *
   * <p>Many of the filtered set's methods, such as {@code size()}, iterate across every element in
   * the underlying set and determine which elements satisfy the filter. When a live view is
   * <i>not</i> needed, it may be faster to copy {@code Iterables.filter(unfiltered, predicate)} and
   * use the copy.
   *
   * <p><b>Warning:</b> {@code predicate} must be <i>consistent with equals</i>, as documented at

     * changing any bin. The operations must not take any action that could even momentarily
     * cause a concurrent read operation to see inconsistent data. This is made easier by the
     * nature of the read operations in Map. For example, no operation can reveal that the table
     * has grown but the threshold has not yet been updated, so there are no atomicity requirements

   * key/value mapping in the underlying multimap and determine which satisfy the filter. When a
   * live view is <i>not</i> needed, it may be faster to copy the filtered multimap and use the
   * copy.
   *
   * <p><b>Warning:</b> {@code keyPredicate} must be <i>consistent with equals</i>, as documented at

   * query than the original; your mileage may vary. Precomputation takes time and is likely to be
   * worthwhile only if the precomputed matcher is queried many thousands of times.
   *
   * <p>This method has no effect (returns {@code this}) when called in GWT: it's unclear whether a
   * precomputed matcher is faster, but it certainly consumes more memory, which doesn't seem like a

 * asynchronous operations. You can chain them together manually with calls to methods like {@link
 * Futures#transform(ListenableFuture, Function, Executor) Futures.transform}, but you will often
 * find it easier to use a framework. Frameworks automate the process, often adding features like
 * monitoring, debugging, and cancellation. Examples of frameworks include:
 *
 * <ul>