* compatibility (otherwise prefer {@link #putUnencodedChars}). However, the character encodings
 * must be identical across languages. Also beware that {@link Charset} definitions may occasionally
 * change between Java releases.
 *
 * <p><b>Warning:</b> Chunks of data that are put into the {@link Hasher} are not delimited. The
 * resulting {@link HashCode} is dependent only on the bytes inserted, and the order in which they

  /**
   * The type of ordering that this object specifies.
   *
   * <ul>
   *   <li>UNORDERED: no order is guaranteed.
   *   <li>STABLE: ordering is guaranteed to follow a pattern that won't change between releases.
   *       Some methods may have stronger guarantees.
   *   <li>INSERTION: insertion ordering is guaranteed.
   *   <li>SORTED: ordering according to a supplied comparator is guaranteed.
   * </ul>
   */

 *
 * @author Sven Mawson
 * @author Nishant Thakkar
 * @since 1.0
 */
/*
 * Some of the annotations below were added after we released our separate
 * com.google.guava:listenablefuture:1.0 artifact. (For more on that artifact, see
 * https://github.com/google/guava/releases/tag/v27.0) This means that the copy of ListenableFuture
 * in com.google.guava:guava differs from the "frozen" copy in the listenablefuture artifact. This

abstract class AggregateFutureState<OutputT extends @Nullable Object>
    extends AbstractFuture.TrustedFuture<OutputT> {
  // Lazily initialized the first time we see an exception; not released until all the input futures
  // & this future completes. Released when the future releases the reference to the running state
  private @Nullable Set<Throwable> seenExceptions = null;
  private int remaining;

  AggregateFutureState(int remainingFutures) {

   * cross-language compatibility. For other use cases, prefer {@link #hashUnencodedChars}, which is
   * faster, produces the same output across Java releases, and hashes every {@code char} in the
   * input, even if some are invalid.
   */
  HashCode hashString(CharSequence input, Charset charset);

  /**

 *
 * <ul>
 *   <li>Each lock adds (and removes) itself to/from a ThreadLocal Set of acquired locks when the
 *       Thread acquires its first hold (and releases its last remaining hold).
 *   <li>Before the lock is acquired, the lock is checked against the current set of acquired
 *       locks---to each of the acquired locks, an edge from the soon-to-be-acquired lock is either

 *
 * @author Sven Mawson
 * @author Nishant Thakkar
 * @since 1.0
 */
/*
 * Some of the annotations below were added after we released our separate
 * com.google.guava:listenablefuture:1.0 artifact. (For more on that artifact, see
 * https://github.com/google/guava/releases/tag/v27.0) This means that the copy of ListenableFuture
 * in com.google.guava:guava differs from the "frozen" copy in the listenablefuture artifact. This

/**
 * A rate limiter. Conceptually, a rate limiter distributes permits at a configurable rate. Each
 * {@link #acquire()} blocks if necessary until a permit is available, and then takes it. Once
 * acquired, permits need not be released.
 *
 * <p>{@code RateLimiter} is safe for concurrent use: It will restrict the total rate of calls from
 * all threads. Note, however, that it does not guarantee fairness.
 *

   */
  final void maybePropagateCancellationTo(@CheckForNull Future<?> related) {
    if (related != null & isCancelled()) {
      related.cancel(wasInterrupted());
    }
  }

  /** Releases all threads in the {@link #waiters} list, and clears the list. */
  private void releaseWaiters() {
    Waiter head = ATOMIC_HELPER.gasWaiters(this, Waiter.TOMBSTONE);

      for (int j = 0; j < dataSize; ++j) {
        dataset.addEntries(helper);
      }
      dummy ^= helper.toString().hashCode();
    }
    return dummy;
  }

  // When omitEmptyValues() is released, remove this method and add a new @Param "omitEmptyValues".