* is complete. It does so by using daemon threads and adding a shutdown hook to wait for their
   * completion.
   *
   * <p>This is mainly for fixed thread pools. See {@link Executors#newFixedThreadPool(int)}.
   *
   * @param executor the executor to modify to make sure it exits when the application is finished
   * @param terminationTimeout how long to wait for the executor to finish before terminating the
   *     JVM

   * to {@link LoadingCache#get} if this method is not overridden.
   *
   * @param keys the unique, non-null keys whose values should be loaded
   * @return a map from each key in {@code keys} to the value associated with that key; <b>may not
   *     contain null values</b>
   * @throws Exception if unable to load the result

/**
 * A map, each entry of which maps a {@link TypeToken} to an instance of that type. In addition to
 * implementing {@code Map}, the additional type-safe operations {@link #putInstance} and {@link
 * #getInstance} are available.
 *
 * <p>Generally, implementations don't support {@link #put} and {@link #putAll} because there is no
 * way to check an object at runtime to be an instance of a {@link TypeToken}. Instead, caller

/**
 * A container for data that is specific to a session.
 * All components may use this storage to associate arbitrary data with a session.
 * <p>
 * Unlike a cache, this session data is not subject to purging. For this same reason, session data should also not be
 * abused as a cache (i.e. for storing values that can be re-calculated) to avoid memory exhaustion.
 * <p>

   * first one specified is included.
   *
   * <p>Despite the method name, this method attempts to avoid actually copying the data when it is
   * safe to do so. The exact circumstances under which a copy will or will not be performed are
   * undocumented and subject to change.
   *

      }
      // Successfully crossed over from min to max.
      // Bubble up max levels.
      E parent;
      // If toTrickle is moved up to a parent of removeIndex, the parent is
      // placed in removeIndex position. We must return that to the iterator so
      // that it knows to skip it.
      if (crossOver < removeIndex) {
        // We crossed over to the parent level in crossOver, so the parent

    /**
     * Reads the metadata from the specified file.
     *
     * @param input The file to deserialize the metadata from, must not be {@code null}.
     * @param options The options to use for deserialization, may be {@code null} to use the default values.
     * @return The deserialized metadata, never {@code null}.
     * @throws IOException If the metadata could not be deserialized.

   * before and after a call to {@code remove(range)}. If {@code range.contains(k)}, then after a
   * call to {@code remove(range)}, {@code get(k)} will return {@code null}.
   */
  void remove(Range<K> range);

  /**
   * Returns a view of this range map as an unmodifiable {@code Map<Range<K>, V>}. Modifications to
   * this range map are guaranteed to read through to the returned {@code Map}.
   *

   * and because threads are typically not bound to CPUS forever,
   * may not occur at all. However, despite these limitations,
   * observed contention rates are typically low in these cases.
   *
   * It is possible for a Cell to become unused when threads that
   * once hashed to it terminate, as well as in the case where
   * doubling the table causes no thread to hash to it under

 * good" reference implementation. In order to accomplish this, it's important to test a great
 * variety of sequences of the {@link Iterator#next}, {@link Iterator#hasNext} and {@link
 * Iterator#remove} operations. This utility takes the brute-force approach of trying <i>all</i>
 * possible sequences of these operations, up to a given number of steps. So, if the caller