* colliding threads.  Because search is random, and collisions
   * only become known via CAS failures, convergence can be slow,
   * 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

   * includes exceptional exits, so all control paths involving signalling must be protected by a
   * finally block.
   *
   * Further optimizations of this algorithm become increasingly subtle. A wait that terminates
   * without the guard being satisfied (due to timeout, but not interrupt) can then immediately exit

  /**
   * Returns the identity function.
   *
   * <p><b>Discouraged:</b> Prefer using a lambda like {@code v -> v}, which is shorter and often
   * more readable.
   */
  // implementation is "fully variant"; E has become a "pass-through" type
  @SuppressWarnings("unchecked")
  public static <E extends @Nullable Object> Function<E, E> identity() {
    return (Function<E, E>) IdentityFunction.INSTANCE;
  }

  // enum singleton pattern

    // places left to rotate.
    int m = -distance % length;
    m = (m < 0) ? m + length : m;
    // The current index of what will become the first element of the rotated section.
    int newFirstIndex = m + fromIndex;
    if (newFirstIndex == fromIndex) {
      return;
    }

    reverse(array, fromIndex, newFirstIndex);

    // places left to rotate.
    int m = -distance % length;
    m = (m < 0) ? m + length : m;
    // The current index of what will become the first element of the rotated section.
    int newFirstIndex = m + fromIndex;
    if (newFirstIndex == fromIndex) {
      return;
    }

    reverse(array, fromIndex, newFirstIndex);

      if (parentIndex != 0) {
        /*
         * This is a guard for the case of the childless aunt node. Since the end of the array is
         * actually the middle of the heap, a smaller childless aunt node can become a child of x
         * when we bubble up alternate levels, violating the invariant.
         */
        int grandparentIndex = getParentIndex(parentIndex);
        int auntIndex = getRightChildIndex(grandparentIndex);

  public void testIterator_removeAffectsBackingCollection() {
    int originalSize = collection.size();
    Iterator<E> iterator = collection.iterator();
    Object element = iterator.next();
    // If it's an Entry, it may become invalid once it's removed from the Map. Copy it.
    if (element instanceof Entry) {
      Entry<?, ?> entry = (Entry<?, ?>) element;
      element = mapEntry(entry.getKey(), entry.getValue());
    }

   * colliding threads.  Because search is random, and collisions
   * only become known via CAS failures, convergence can be slow,
   * 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

    // places left to rotate.
    int m = -distance % length;
    m = (m < 0) ? m + length : m;
    // The current index of what will become the first element of the rotated section.
    int newFirstIndex = m + fromIndex;
    if (newFirstIndex == fromIndex) {
      return;
    }

    reverse(array, fromIndex, newFirstIndex);

    // places left to rotate.
    int m = -distance % length;
    m = (m < 0) ? m + length : m;
    // The current index of what will become the first element of the rotated section.
    int newFirstIndex = m + fromIndex;
    if (newFirstIndex == fromIndex) {
      return;
    }

    reverse(array, fromIndex, newFirstIndex);