* defined as the <i>least</i> element in the queue according to the queue's comparator. But unlike
 * a regular priority queue, the methods {@link #peekLast}, {@link #pollLast} and {@link
 * #removeLast} are also provided, to act on the <i>greatest</i> element in the queue instead.
 *
 * <p>A min-max priority queue can be configured with a maximum size. If so, each time the size of

 * defined as the <i>least</i> element in the queue according to the queue's comparator. But unlike
 * a regular priority queue, the methods {@link #peekLast}, {@link #pollLast} and {@link
 * #removeLast} are also provided, to act on the <i>greatest</i> element in the queue instead.
 *
 * <p>A min-max priority queue can be configured with a maximum size. If so, each time the size of

    stack.addLast(new NodeAndRemainingSuccessors<>(startNode));

    while (!stack.isEmpty()) {
      // To peek at the top two items, we need to temporarily remove one.
      NodeAndRemainingSuccessors<N> top = stack.removeLast();
      NodeAndRemainingSuccessors<N> prev = stack.peekLast();
      stack.addLast(top);

      N node = top.node;
      N previousNode = prev == null ? null : prev.node;

             */
            requireNonNull(element);
            if (visited.add(element)) {
              return element;
            }
          }
          horizon.removeFirst();
          return null;
        }
      };
    }

    static <N> Traversal<N> inTree(SuccessorsFunction<N> tree) {
      return new Traversal<N>(tree) {
        @Override