// matches `initializer_type`.
bool IsInitializerType(func::FuncOp init_func_op, StringRef initializer_type) {
  auto init_type =
      init_func_op->getAttrOfType<StringAttr>(kTfSavedModelInitializerTypeAttr);
  return init_type && init_type == initializer_type;
}

// Returns the initializer function whose tf_saved_model.initializer_type

  if (!s.ok()) {
    return std::nullopt;
  }
  return attr_value->s();
}

bool HasResourceInputOrOutput(const Node& node) {
  return std::find(node.input_types().begin(), node.input_types().end(),
                   DT_RESOURCE) != node.input_types().end() ||
         std::find(node.output_types().begin(), node.output_types().end(),
                   DT_RESOURCE) != node.output_types().end();
}

#include "llvm/Support/CommandLine.h"

// Please see the implementation file for documentation of these options.

// Import options.
extern llvm::cl::opt<std::string> input_arrays;
extern llvm::cl::opt<std::string> input_dtypes;
extern llvm::cl::opt<std::string> input_shapes;
extern llvm::cl::opt<std::string> output_arrays;
extern llvm::cl::opt<std::string> control_output_arrays;
extern llvm::cl::opt<std::string> inference_type;

absl::StatusOr<ElementsAttr> ConvertTensor(const Tensor& input_tensor,
                                           Builder* builder) {
  const auto& input_dtype = input_tensor.dtype();
  const auto& input_shape = input_tensor.shape();
  Type elt_type;
  TF_RETURN_IF_ERROR(ConvertDataType(input_dtype, *builder, &elt_type));
  SmallVector<int64_t, 4> shape;
  ConvertToMlirShape(input_shape, &shape);

    absl::Span<const int64_t> input_tensor_id,
    absl::Span<const tensorflow::DataType> input_dtypes,
    const std::function<BackwardFunction*()>& backward_function_getter,
    const std::function<void(BackwardFunction*)>& backward_function_deleter) {
  if (!ShouldRecord(input_tensor_id, input_dtypes)) {
    return;
  }
  std::vector<int64_t> ids;
  ids.reserve(input_tensor_id.size());

  // Populate inputs.
  std::vector<DataType> input_dtypes;
  TF_RETURN_IF_ERROR(GetNodeAttr(call_node->attrs(), "Tinputs", &input_dtypes));
  std::vector<NodeDefBuilder::NodeOut> inputs(input_dtypes.size());
  for (auto e : call_node->in_edges()) {
    if (e->IsControlEdge()) {
      continue;
    }

    const int input_dtypes_size = input_dtypes.size();

            .build();
    }

    public static <C, I> ModelRegistration unmanaged(ModelRegistrations.Builder builder, Class<C> type, Class<I> inputType, Transformer<? extends C, ? super I> action) {
        return unmanaged(builder, ModelType.of(type), ModelType.of(inputType), action);
    }

                           const string& op_name) {
  std::vector<int64_t> input_ids(inputs.size());
  std::vector<tensorflow::DataType> input_dtypes(inputs.size());
  for (int i = 0; i < inputs.size(); i++) {
    input_ids[i] = ToId(inputs[i]);
    input_dtypes[i] = inputs[i]->DataType();
  }
  std::vector<TapeTensor> tape_tensors;
  tape_tensors.reserve(outputs.size());
  for (auto t : outputs) {

import static org.objectweb.asm.Opcodes.INVOKESPECIAL;
import static org.objectweb.asm.Opcodes.INVOKESTATIC;
import static org.objectweb.asm.Opcodes.INVOKEVIRTUAL;
import static org.objectweb.asm.Type.INT_TYPE;
import static org.objectweb.asm.Type.getMethodDescriptor;
import static org.objectweb.asm.Type.getObjectType;
import static org.objectweb.asm.Type.getType;

    bool use_splatted_constant, const std::vector<std::string>& extra_tf_opdefs,
    const GraphImportConfig& specs, absl::string_view debug_info_file,
    absl::string_view input_arrays, absl::string_view input_dtypes,
    absl::string_view input_shapes, absl::string_view output_arrays,
    absl::string_view control_output_arrays, llvm::SourceMgr* source_mgr,
    mlir::MLIRContext* context);