// specifiable.
    GraphDebugInfo debug_info;
    switch (config.dialect) {
      case MlirDumpConfig::Dialect::kTFG: {
        TF_ASSIGN_OR_RETURN(module,
                            mlir::tfg::ImportGraphAndFunctionsToMlir(
                                &context, debug_info, graph,
                                flib_def ? *flib_def : graph.flib_def()));
        break;

       "Input GraphDef file.  If the file ends in '.pbtxt' it is expected to "
       "be in the human-readable proto text format, otherwise it is expected "
       "to be in the proto binary format."},
      {"debug_info", &flags->debug_info,
       "Graph debug info file.  If the file ends in '.pbtxt' it is expected to "
       "be in the human-readable proto text format, otherwise it is expected "
       "to be in the proto binary format."},

#include "tensorflow/core/util/command_line_flags.h"

namespace tensorflow {
namespace tfcompile {

// Flags for the tfcompile binary.  See *.cc file for descriptions.

struct MainFlags {
  string graph;
  string debug_info;
  string debug_info_path_begin_marker;
  string config;
  bool dump_fetch_nodes = false;
  string target_triple;
  string target_cpu;
  string target_features;
  string entry_point;
  string cpp_class;

# RUN: not tfcompile --graph=%s --config=%s.config.pbtxt --mlir_components=Bridge --debug_info=%s.debug.pbtxt 2>&1 | FileCheck %s
# RUN: not tfcompile --graph=%s --config=%s.config.pbtxt --mlir_components=None 2>&1 | FileCheck -check-prefix=OLD %s

# Checks the error message produced by tfcompile with mlir_component
# Checks that source debug information is used in the output error message and
# the node x_y_sum = Add

// it fails to convert the input.
absl::Status ConvertGraphDefToTFLiteFlatBuffer(
    const toco::ModelFlags& model_flags, toco::TocoFlags& toco_flags,
    const GraphDebugInfo& debug_info, const GraphDef& input,
    std::string* result);

}  // namespace tensorflow

  }

  /// BundleReader for accessing the variables bundle.
  BundleReader* variable_reader() { return variable_reader_.get(); }

  /// The GraphDebugInfo (or nullptr if none).
  GraphDebugInfo* debug_info() { return debug_info_.get(); }

  /// Restores objects, invoking the callback with the node id in the
  /// saved_object_graph() and the corresponding TrackableObject from the

    return nullptr;
  }

  tensorflow::GraphDebugInfo debug_info;
  if (debug_info_txt_raw && debug_info_txt_raw != Py_None) {
    std::string debug_info_txt = ConvertArg(debug_info_txt_raw, &error);
    if (error) {
      PyErr_SetString(PyExc_ValueError, "Input DebugInfo is invalid.");
      return nullptr;
    }
    if (!debug_info.ParseFromString(debug_info_txt)) {
      PyErr_SetString(PyExc_ValueError,

    // debug_info might not be loaded with loader_lite.
    GraphDebugInfo debug_info;
    if (bundle.debug_info != nullptr) debug_info = *bundle.debug_info;

    TF_ASSIGN_OR_RETURN(auto input,
                        SimpleSavedModelMLIRImportInput::Create(
                            options, &bundle.meta_graph_def, debug_info));

    TF_ASSIGN_OR_RETURN(auto module,

                         << ", FallbackEnabled: " << num_passes_fallback_enabled
                         << ", Total: " << registry_->passes().size();
  }

  GraphDebugInfo debug_info;
  mlir::DialectRegistry registry;
  RegisterDialects(registry);
  mlir::MLIRContext context(registry);
  GraphImportConfig import_config;
  import_config.graph_as_function = true;

          << pass->name() << ": " << status.message();
      return signalPassFailure();
    }
  }

  // Convert Graph to MLIR
  GraphDebugInfo debug_info;
  GraphImportConfig specs;
  auto module_or_status =
      ConvertGraphToMlir(**options.graph, debug_info, flib_def, specs, &ctx);
  if (!module_or_status.ok()) {
    mlir::emitError(mlir::UnknownLoc::get(&ctx))
        << module_or_status.status().message();