return this.masterKey;
    }


    /**
     * Returns the session key.
     * 
     * This is the encrypted session key included in the message,
     * if the actual session key is desired use {@link #getMasterKey()} instead.
     *
     * @return A <code>byte[]</code> containing the encrypted session key.
     */
    public byte[] getEncryptedSessionKey () {

   * can't be constructed at all.)
   *
   * <p>Note that certain discrete ranges such as the integer range {@code (3..4)} are <b>not</b>
   * considered empty, even though they contain no actual values. In these cases, it may be helpful
   * to preprocess ranges with {@link #canonical(DiscreteDomain)}.
   */
  public boolean isEmpty() {
    return lowerBound.equals(upperBound);
  }

  /**

  private static class GenericClass<T> {
    private interface Base {}
  }

  private static IterableSubject makeUnmodifiable(Collection<?> actual) {
    return assertThat(Collections.<Object>unmodifiableCollection(actual));
  }

  }

  /**
   * Returns the empty iterator.
   *
   * <p>The {@link Iterable} equivalent of this method is {@link ImmutableSet#of()}.
   */
  // Casting to any type is safe since there are no actual elements.
  @SuppressWarnings("unchecked")
  static <T extends @Nullable Object> UnmodifiableListIterator<T> emptyListIterator() {
    return (UnmodifiableListIterator<T>) ArrayItr.EMPTY;
  }

  /**

   * list (counter-intuitive, but mathematically consistent).
   *
   * <p><i>Performance notes:</i> while the cartesian product of sets of size {@code m, n, p} is a
   * set of size {@code m x n x p}, its actual memory consumption is much smaller. When the
   * cartesian set is constructed, the input sets are merely copied. Only as the resulting set is

// unknown dimension is represented by `-1`.
//
// Returns an error into `status` if:
//   * `output` is not in `graph`.
//   * `num_dims` does not match the actual number of dimensions.
TF_CAPI_EXPORT extern void TF_GraphGetTensorShape(TF_Graph* graph,
                                                  TF_Output output,

    boolean longWait = NANOSECONDS.toSeconds(thread.timeSpentBlocked) >= 5;
    // Count how long it actually took to return; we'll accept any number between the expected delay
    // and the approximate actual delay, to be robust to variance in thread scheduling.
    char overWaitNanosFirstDigit =
        Long.toString(thread.timeSpentBlocked - MILLISECONDS.toNanos(longWait ? 5000 : 3000))
            .charAt(0);

A container image is comparable to the **program** file and contents, e.g. `python` and some file `main.py`.

And the **container** itself (in contrast to the **container image**) is the actual running instance of the image, comparable to a **process**. In fact, a container is running only when it has a **process running** (and normally it's only a single process). The container stops when there's no process running in it.

  // below by testing their text-processing methods.

  // The organization of this test class is unusual, as it's not done by
  // method, but by overall "scenario". Also, the variety of actual tests we
  // do borders on absurd overkill. Better safe than sorry, though?

  @GwtIncompatible // java.util.BitSet
  public void testSetBits() {
    doTestSetBits(CharMatcher.any());

  private void readObject(ObjectInputStream in) throws InvalidObjectException {
    throw new InvalidObjectException("Use SerializationProxy");
  }

  /**
   * The actual object that gets serialized. Unfortunately, readResolve() doesn't get called when a
   * circular dependency is present, so the proxy must be able to behave as the map itself.
   */