* Stream#findAny} (which you might as well use).
   *
   * @see Stream#findFirst()
   * @throws NullPointerException if the last element of the stream is null
   */
  /*
   * By declaring <T> instead of <T extends @Nullable Object>, we declare this method as requiring a
   * stream whose elements are non-null. However, the method goes out of its way to still handle

 * It encapsulates the delay logic and exception handling, allowing subclasses to focus on
 * defining the specific delay behavior for each stage.
 *
 * <p>
 * This class defines the contract for delaying the crawling process at various points, such as:
 * </p>
 * <ul>
 *   <li>Before processing a URL ({@link #delayBeforeProcessing()})</li>
 *   <li>After processing a URL ({@link #delayAfterProcessing()})</li>

        // Minimise test flakiness due to possible race conditions with connections closing.
        // Some number of tests will report here, but not fail due to this delay.
        println("Delaying to avoid flakes")
        Thread.sleep(500L)
        println("After delay: " + connectionPool.connectionCount())
      }

      connectionPool.evictAll()
      assertEquals(0, connectionPool.connectionCount()) {