}
  const auto& s = attr->s();
  std::memcpy(value, s.data(), std::min<size_t>(s.length(), max_length));
}

void TF_OperationGetAttrStringList(TF_Operation* oper, const char* attr_name,
                                   void** values, size_t* lengths,
                                   int max_values, void* storage,
                                   size_t storage_size, TF_Status* status) {

      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];
      if (input_shape.has_value()) {
        TF_ShapeAndTypeListSetShape(input_shapes, i, input_shape->data(),

 public:
  virtual void Clear() = 0;

  // Returns the inputs of this op.
  virtual absl::Span<ImmediateExecutionTensorHandle* const> GetInputs()
      const = 0;
  virtual absl::Status SetInput(size_t index,
                                ImmediateExecutionTensorHandle* input) = 0;

  virtual ImmediateExecutionContext* GetContext() const = 0;

  // Following two methods are used to support custom device.

  typedef void (*MemoryReleaser)(void* data, size_t len, void* arg);

  // Create a tensor instance from the given data buffer and description.
  // `memory_releaser` will be called on destruction, and it's responsible for
  // cleaning up the underlying buffer.
  virtual AbstractTensorInterface* CreateTensor(
      DataType dtype, const int64_t* dims, int num_dims, void* data, size_t len,

                              TF_Status* status) {
  status->status = MessageToBuffer(func->record->fdef(), output_func_def);
}

TF_Function* TF_FunctionImportFunctionDef(const void* proto, size_t proto_len,
                                          TF_Status* status) {
  tensorflow::FunctionDef fdef;
  bool success = fdef.ParseFromArray(proto, proto_len);
  if (!success) {
    status->status = InvalidArgument(

		return
	}
	z.init = true
	C.mpz_init(&z.i[0])
}

// Bytes returns z's representation as a big-endian byte array.
func (z *Int) Bytes() []byte {
	b := make([]byte, (z.Len()+7)/8)
	n := C.size_t(len(b))
	C.mpz_export(unsafe.Pointer(&b[0]), &n, 1, 1, 1, 0, &z.i[0])
	return b[0:n]
}

// Len returns the length of z in bits.  0 is considered to have length 1.
func (z *Int) Len() int {
	z.doinit()

    return 1;
  }

  char file_name[100];
  time_t t = time(NULL);
  snprintf(file_name, sizeof(file_name), "test-%d-%ld.txt", getpid(), t);

  size_t length = 2 + strlen(path) + strlen(file_name);
  char* full_path = malloc(length);
  snprintf(full_path, length, "%s/%s", path, file_name);

  TF_WritableFileHandle* h;
  TF_Status* status = TF_NewStatus();

  }
}

// Wraps the deleter function of DLManagedTensor to match the function signature
// TFE_NewTensorHandleFromDeviceMemory.
void DeallocatorWrapperFunc(void* data, size_t len, void* dlmt_vptr) {
  TFE_CallDLManagedTensorDeleter(dlmt_vptr);
}

// Checks whether the stride array matches the layout of compact, row-majored
// data.

  <!-- JNA requires the no-argument constructor on JNAKernel32Library.SizeT to be public-->
  <suppress files="server[/\\]src[/\\]main[/\\]java[/\\]org[/\\]elasticsearch[/\\]bootstrap[/\\]JNAKernel32Library.java" checks="RedundantModifier" />

  EXPECT_EQ(22, product[3]);
  TF_DeleteStatus(status);
}
TEST(CAPI, Executor_MatMul_CPU) { Executor_MatMul_CPU(false); }
TEST(CAPI, Executor_MatMul_CPUAsync) { Executor_MatMul_CPU(true); }

void Deleter(void* data, size_t unused, void* tensor_handle) {
  TFE_DeleteTensorHandle(static_cast<TFE_TensorHandle*>(tensor_handle));
}

TEST(CAPI, TensorHandleOnDeviceMemory) {
  TF_Status* status = TF_NewStatus();