// needed.
//
// Because coefficient-wise binary operations support implicit broadcasting, we
// should be very careful with this optimization, and do not accidentally
// produce incorrect concat operations.
class HoistCwiseBinaryOutOfConcat : public OpRewritePattern<TF::ConcatV2Op> {
 public:
  explicit HoistCwiseBinaryOutOfConcat(MLIRContext* context)
      : OpRewritePattern<TF::ConcatV2Op>(context) {}

 public:
  // Sets benefit to 10 to make this pattern more preferred than smaller local
  // transformations like `stablehlo.transpose`->`tfl.transpose`, as this
  // pattern involves `stablehlo.transpose` in some cases.
  explicit RewriteQuantizedDotGeneralOpToTflFullyConnectedOrBatchMatmulOp(
      MLIRContext* ctx)
      : OpRewritePattern<stablehlo::DotGeneralOp>(ctx, /*benefit=*/10) {}

`日本語`                                 // UTF-8 input text as a raw literal
"\u65e5\u672c\u8a9e"                    // the explicit Unicode code points
"\U000065e5\U0000672c\U00008a9e"        // the explicit Unicode code points
"\xe6\x97\xa5\xe6\x9c\xac\xe8\xaa\x9e"  // the explicit UTF-8 bytes
</pre>

<p>
If the source code represents a character as two code points, such as

    // Add an TF.PadOp before the LHS if there is explicit_padding on a 3D
    // Convolution. This is needed because TF.Conv3DOp doesn't support EXPLICIT.
    if (padding == "EXPLICIT" && num_spatial_dims == 3) {
      auto lhs_type =
          mlir::dyn_cast<RankedTensorType>(conv_op.getLhs().getType());
      RankedTensorType padding_attr_type = mlir::RankedTensorType::get(

		}
	}
	return fmt.Sprintf("%*s%sSizeofArgs:\n%s", indent, "", field, args)
}

// TemplateParamName is the name of a template parameter that the
// demangler introduced for a lambda that has explicit template
// parameters.  This is a prefix with an index.
type TemplateParamName struct {
	Prefix string
	Index  int
}

func (tpn *TemplateParamName) print(ps *printState) {

  auto stride_w = op.getStrideWAttr().getInt();
  auto dilation_h = op.getDilationHFactorAttr().getInt();
  auto dilation_w = op.getDilationWFactorAttr().getInt();

  // We don't have EXPLICIT PADDING in TfLite.
  auto paddings = op.getPadding();
  tensorflow::Padding padding;
  auto padding_is_valid = GetPaddingFromString(paddings.str(), &padding);
  if (!padding_is_valid.ok()) {

// TODO(b/142268695) Re-evaluate whether we need this class v.s. directly using
// and TF function name as MLIR function name after b/142268695 is root caused.
class NameUniquifier : public OpOrArgNameMapper {
 public:
  explicit NameUniquifier(const FunctionLibraryDefinition& flib)
      : flib_(flib) {}

 private:
  bool IsUnique(llvm::StringRef name) override {
    return !flib_.Contains(std::string(name));
  }

  let description = [{
    This pass outlines the body of a `tf_device.cluster` into a function and
    replaces the `tf_device.cluster` op with an equivalent `tf_device.cluster_func`
    op. Implicit operands will be captured and materialized as explicit arguments to
    the newly created functions and associated `tf_device.cluster_func` ops.

    For example, the following:

    ```mlir

        } else if (createNewEntry || valueReference.isLoading()) {
          removeLoadingValue(key, hash, computingValueReference);
          return null;
        } else {
          removeEntry(e, hash, RemovalCause.EXPLICIT);
          return null;
        }
      } finally {
        unlock();
        postWriteCleanup();
      }
    }

          "description": "Number of desired pods. This is a pointer to distinguish between explicit zero and not specified. Defaults to 1.",
          "format": "int32",
          "type": "integer"
        },
        "revisionHistoryLimit": {
          "description": "The number of old ReplicaSets to retain to allow rollback. This is a pointer to distinguish between explicit zero and not specified. Defaults to 10.",
          "format": "int32",