}

  static void testSort(int[] input, int[] expected) {
    input = Arrays.copyOf(input, input.length);
    UnsignedInts.sort(input);
    assertThat(input).isEqualTo(expected);
  }

  static void testSort(int[] input, int from, int to, int[] expected) {
    input = Arrays.copyOf(input, input.length);
    UnsignedInts.sort(input, from, to);
    assertThat(input).isEqualTo(expected);
  }

    List<List<Integer>> input = newArrayList(list1, list2);

    FluentIterable<Integer> result = FluentIterable.concat(input);
    assertEquals(asList(1, 4), newArrayList(result));

    // Now change the inputs and see result dynamically change as well

    list1.add(2);
    List<Integer> list3 = newArrayList(3);
    input.add(1, list3);

   */
  HashCode hashBytes(byte[] input, int off, int len);

  /**
   * Shortcut for {@code newHasher().putBytes(input).hash()}. The implementation <i>might</i>
   * perform better than its longhand equivalent, but should not perform worse.
   *
   * @since 23.0
   */
  HashCode hashBytes(ByteBuffer input);

  /**
   * Shortcut for {@code newHasher().putUnencodedChars(input).hash()}. The implementation

  private static final class RoundToDoubleTester {
    private final BigDecimal input;
    private final Map<RoundingMode, Double> expectedValues = new EnumMap<>(RoundingMode.class);
    private boolean unnecessaryShouldThrow = false;

    RoundToDoubleTester(BigDecimal input) {
      this.input = input;
    }

    RoundToDoubleTester setExpectation(double expectedValue, RoundingMode... modes) {

  private static void testReverse(float[] input, float[] expectedOutput) {
    input = Arrays.copyOf(input, input.length);
    Floats.reverse(input);
    assertThat(input).isEqualTo(expectedOutput);
  }

  private static void testReverse(
      float[] input, int fromIndex, int toIndex, float[] expectedOutput) {
    input = Arrays.copyOf(input, input.length);
    Floats.reverse(input, fromIndex, toIndex);

      const std::vector<TF_Tensor*>& input_tensors,
      const absl::optional<std::vector<int64_t>>& expected_shape) {
    // Create input_shapes.
    TF_ShapeAndTypeList* input_shapes =
        TF_NewShapeAndTypeList(input_shapes_vec.size());
    for (size_t i = 0; i < input_shapes_vec.size(); ++i) {
      const auto& input_shape = input_shapes_vec[i];

namespace gradients {
namespace {

static Status SafeConj(AbstractContext* ctx, AbstractTensorHandle* const input,
                       AbstractTensorHandle** output, const char* name) {
  auto dtype = input->DataType();
  if (DataTypeIsFloating(BaseType(dtype)) ||
      DataTypeIsInteger(BaseType(dtype))) {
    return tensorflow::ops::Identity(ctx, input, output, name);
  } else if (!DataTypeIsComplex(BaseType(dtype)) &&

    checkNotNull(keyFunction);
    checkNotNull(valuesFunction);
    return collectingAndThen(
        flatteningToMultimap(
            input -> checkNotNull(keyFunction.apply(input)),
            input -> valuesFunction.apply(input).peek(Preconditions::checkNotNull),
            MultimapBuilder.linkedHashKeys().arrayListValues()::<K, V>build),
        ImmutableListMultimap::copyOf);
  }

  }

  private static void testReverse(long[] input, long[] expectedOutput) {
    input = Arrays.copyOf(input, input.length);
    Longs.reverse(input);
    assertThat(input).isEqualTo(expectedOutput);
  }

  private static void testReverse(long[] input, int fromIndex, int toIndex, long[] expectedOutput) {
    input = Arrays.copyOf(input, input.length);
    Longs.reverse(input, fromIndex, toIndex);

  }

  private static void testReverse(long[] input, long[] expectedOutput) {
    input = Arrays.copyOf(input, input.length);
    Longs.reverse(input);
    assertThat(input).isEqualTo(expectedOutput);
  }

  private static void testReverse(long[] input, int fromIndex, int toIndex, long[] expectedOutput) {
    input = Arrays.copyOf(input, input.length);
    Longs.reverse(input, fromIndex, toIndex);