private static long hashLength33To64(byte[] bytes, int offset, int length) {
    long z = load64(bytes, offset + 24);
    long a = load64(bytes, offset) + (length + load64(bytes, offset + length - 16)) * K0;
    long b = rotateRight(a + z, 52);
    long c = rotateRight(a, 37);
    a += load64(bytes, offset + 8);
    c += rotateRight(a, 7);
    a += load64(bytes, offset + 16);
    long vf = a + z;
    long vs = b + rotateRight(a, 31) + c;

    long mul = K2 + length * 2L;
    long a = load64(bytes, offset) * K2;
    long b = load64(bytes, offset + 8);
    long c = load64(bytes, offset + length - 8) * mul;
    long d = load64(bytes, offset + length - 16) * K2;
    long y = rotateRight(a + b, 43) + rotateRight(c, 30) + d;
    long z = hashLength16(y, a + rotateRight(b + K2, 18) + c, mul);
    long e = load64(bytes, offset + 16) * mul;

 * overridden as desired.
 *
 * <p>Usage example:
 *
 * {@snippet :
 * CacheLoader<Key, Graph> loader = new CacheLoader<Key, Graph>() {
 *   public Graph load(Key key) throws AnyException {
 *     return createExpensiveGraph(key);
 *   }
 * };
 * LoadingCache<Key, Graph> cache = CacheBuilder.newBuilder().build(loader);
 * }
 *

   * {@code key}, simply waits for that thread to finish and returns its loaded value. Note that
   * multiple threads can concurrently load values for distinct keys.
   *
   * <p>Caches loaded by a {@link CacheLoader} will call {@link CacheLoader#load} to load new values
   * into the cache. Newly loaded values are added to the cache using {@code

 * overridden as desired.
 *
 * <p>Usage example:
 *
 * {@snippet :
 * CacheLoader<Key, Graph> loader = new CacheLoader<Key, Graph>() {
 *   public Graph load(Key key) throws AnyException {
 *     return createExpensiveGraph(key);
 *   }
 * };
 * LoadingCache<Key, Graph> cache = CacheBuilder.newBuilder().build(loader);
 * }
 *

   *
   * When we reference FinalizableReferenceQueue in this test, we are referencing a class that is
   * loaded by this test and that will obviously remain loaded for as long as the test is running.
   * So in order to check ClassLoader garbage collection we need to create a new ClassLoader and
   * make it load its own version of FinalizableReferenceQueue. Then we need to interact with that

   * {@code key}, simply waits for that thread to finish and returns its loaded value. Note that
   * multiple threads can concurrently load values for distinct keys.
   *
   * <p>Caches loaded by a {@link CacheLoader} will call {@link CacheLoader#load} to load new values
   * into the cache. Newly loaded values are added to the cache using {@code

   *
   * When we reference FinalizableReferenceQueue in this test, we are referencing a class that is
   * loaded by this test and that will obviously remain loaded for as long as the test is running.
   * So in order to check ClassLoader garbage collection we need to create a new ClassLoader and
   * make it load its own version of FinalizableReferenceQueue. Then we need to interact with that

     * </ol>
     * <p>
     * However, if both the class loader that loaded the target class and the class loader that loaded this class can be obtained,
     * and the class loader that loaded the target class is an ancestor of the class loader that loaded this class,
     * the class loader that loaded this class is returned.
     * </p>
     *

        final InputStream in = URLUtil.openStream(url);

        try {
            props.load(in);
        } catch (final IOException e) {
            throw new IORuntimeException(e);
        } finally {
            CloseableUtil.close(in);
        }
    }

    /**
     * Loads a resource from the context class loader into the {@link Properties} (wraps exception handling).
     *
     * @param props