* guarantee behavior in that case, anyway, and the current behavior is likely undesirable.
         * So that's either a reason to feel free to change it or a reason to not bother thinking
         * further about this.
         */
        V rightValue = uncheckedCastNullableTToT(onlyOnRight.remove(leftKey));

  }

  /**
   * Attempts to cancel execution of this step. This attempt will fail if the step has already
   * completed, has already been cancelled, or could not be cancelled for some other reason. If
   * successful, and this step has not started when {@code cancel} is called, this step should never
   * run.
   *
   * <p>If successful, causes the objects captured by this step (if already started) and its input

    List<@Nullable Present<V>> localValues = values;
    if (localValues != null) {
      set(combine(localValues));
    }
  }

  @Override
  void releaseResources(ReleaseResourcesReason reason) {
    super.releaseResources(reason);
    this.values = null;
  }

  abstract C combine(List<@Nullable Present<V>> values);

  /** Used for {@link Futures#allAsList} and {@link Futures#successfulAsList}. */

   * not fail a test when an exception is thrown from one of its {@link TearDown} instances, but
   * this is subject to change. Also, {@code junit4.TearDownTestCase} will.
   *
   * @throws Exception for any reason. {@code TearDownTestCase} ensures that any exception thrown
   *     will not interfere with other TearDown operations.
   */
  void tearDown() throws Exception;

   */
  @Override
  protected void expectContents(Collection<E> expectedCollection) {
    List<E> expectedList = Helpers.copyToList(expectedCollection);
    // Avoid expectEquals() here to delay reason manufacture until necessary.
    if (getList().size() != expectedList.size()) {
      fail("size mismatch: " + reportContext(expectedList));
    }
    for (int i = 0; i < expectedList.size(); i++) {

   * identical inputs: This sounds like it ought to be a problem here, since the goals of this class
   * include testing that {@code equals()} is reflexive and is tolerant of {@code null}. However,
   * there's no problem. The reason: {@link EqualsTester} tests {@code null} and identical inputs
   * directly against {@code equals()} rather than through the {@code Equivalence}.
   */
  private final Equivalence<? super T> equivalence;

  }

  /**
   * Just like fail(reason), but additionally recording (using threadRecordFailure) any
   * AssertionFailedError thrown, so that the current testcase will fail.
   */
  public void threadFail(String reason) {
    try {
      fail(reason);
    } catch (AssertionFailedError t) {
      threadRecordFailure(t);
      fail(reason);
    }
  }

  /**

       * Otherwise, we'll pin it (and its result) in memory until the timeout task is GCed. (The
       * need to clear our reference to the TimeoutFuture is the reason we use a *static* nested
       * class with a manual reference back to the "containing" class.)
       *
       * This has the nice-ish side effect of limiting reentrancy: run() calls

 * guarantee that as long as we also require the input type to be non-null[*] (which is a
 * requirement that existing callers already fulfill).
 *
 * Disclaimer: Part of the reason that callers are so well adapted to `Function<A, B>` may be that
 * that is how the signature looked even prior to this comment! So naturally any change can break

   *     guaranteed for the file system and {@link RecursiveDeleteOption#ALLOW_INSECURE} was not
   *     specified
   * @throws IOException if {@code path} or any file in the subtree rooted at it can't be deleted
   *     for any reason
   */
  public static void deleteRecursively(Path path, RecursiveDeleteOption... options)
      throws IOException {
    Path parentPath = getParentPath(path);
    if (parentPath == null) {