* <i>other type</i> of supplier is expected, use the method reference {@code supplier::get}. A
 * future version of {@code com.google.common.base.Supplier} will be made to <i>extend</i> {@code
 * java.util.function.Supplier}, making conversion code necessary only in one direction. At that
 * time, this interface will be officially discouraged.
 *
 * @author Harry Heymann
 * @since 2.0
 */

   * currently {@code expectedOldCount}. If {@code element} does not appear in the multiset exactly
   * {@code expectedOldCount} times, no changes will be made.
   *
   * @return {@code true} if the change was successful. This usually indicates that the multiset has
   *     been modified, but not always: in the case that {@code expectedOldCount == newCount}, the

                <!-- TODO(cpovirk): Use .mvn/jvm.config instead (per
                     https://errorprone.info/docs/installation#maven) if it can
                     be made not to interfere with JDK8 or if we stop building
                     with JDK8. -->
                <arg>-J--add-exports=jdk.compiler/com.sun.tools.javac.api=ALL-UNNAMED</arg>

 *
 * <p>If the precondition is not met, the {@code Preconditions} method throws an unchecked exception
 * of a specified type, which helps the method in which the exception was thrown communicate that
 * its caller has made a mistake. This allows constructs such as
 *
 * <pre>{@code
 * public static double sqrt(double value) {
 *   if (value < 0) {
 *     throw new IllegalArgumentException("input is negative: " + value);
 *   }

 * #iterator()} method after the first, and whose iterator is always unmodifiable.
 *
 * <p>The {@code Iterable} specification does not make it absolutely clear what should happen on a
 * second invocation, so implementors have made various choices, including:
 *
 * <ul>
 *   <li>returning the same iterator again
 *   <li>throwing an exception of some kind
 *   <li>or the usual, <i>robust</i> behavior, which all known {@link Collection} implementations

 * type</i> of function is expected, use the method reference {@code function::apply}. A future
 * version of {@code com.google.common.base.Function} will be made to <i>extend</i> {@code
 * java.util.function.Function}, making conversion code necessary only in one direction. At that
 * time, this interface will be officially discouraged.
 *
 * @author Kevin Bourrillion
 * @since 2.0

  }

  @Override
  public EndpointPair<N> incidentNodes(E edge) {
    N nodeU = checkedReferenceNode(edge);
    // requireNonNull is safe because checkedReferenceNode made sure the edge is in the network.
    N nodeV = requireNonNull(nodeConnections.get(nodeU)).adjacentNode(edge);
    return EndpointPair.of(this, nodeU, nodeV);
  }

  @Override
  public Set<N> adjacentNodes(N node) {

 * object). So it is important to not call {@link #dispatch} while holding any locks. This is why
 * {@link #enqueue} and {@link #dispatch} are 2 different methods. It is expected that the decision
 * to run a particular event is made during the state change, but the decision to actually invoke
 * the listeners can be delayed slightly so that locks can be dropped. Also, because {@link
 * #dispatch} is expected to be called concurrently, it is idempotent.
 */

          <!-- Note that we do need to build Javadoc for *some* class. Otherwise, we get an empty Javadoc jar, which the Sonatype repository manager rejects. To avoid that, we've introduced a dummy class. But we made it package-private so that no one can depend on it. That in turn forced us to configure Javadoc to show package-private APIs. -->
          <show>package</show>
        </configuration>
      </plugin>

     *
     * @param element the element to add
     * @param occurrences the number of occurrences of the element to add. May be zero, in which
     *     case no change will be made.
     * @return this {@code Builder} object
     * @throws NullPointerException if {@code element} is null
     * @throws IllegalArgumentException if {@code occurrences} is negative, or if this operation