ref: Optional[str] = Field(default=None, alias="$ref")
    dynamicRef: Optional[str] = Field(default=None, alias="$dynamicRef")
    defs: Optional[Dict[str, "SchemaOrBool"]] = Field(default=None, alias="$defs")
    comment: Optional[str] = Field(default=None, alias="$comment")
    # Ref: JSON Schema 2020-12: https://json-schema.org/draft/2020-12/json-schema-core.html#name-a-vocabulary-for-applying-s

	n := syntax.NewName(pos, name)
	tv := syntax.TypeAndValue{Type: typ}
	tv.SetIsValue()
	n.SetTypeInfo(tv)
	r.info.Defs[n] = obj
	return obj, n
}

func (r *rewriter) generateParamName(results []*syntax.Field, i int) {
	obj, n := r.sig.RenameResult(results, i)
	r.info.Defs[n] = obj
}

// declVar declares a variable with a given name type and initializer value.

    tbl_outs = [
        (
            ["-gen-op-decls"],
            "ir/tfr_ops.h.inc",
        ),
        (
            ["-gen-op-defs"],
            "ir/tfr_ops.cc.inc",
        ),
    ],
    tblgen = "@llvm-project//mlir:mlir-tblgen",
    td_file = "ir/tfr_ops.td",
    deps = [
        ":tfr_ops_td_files",
    ],
)

gentbl_cc_library(
    name = "tfr_decompose_inc_gen",

        (
            ["-gen-op-decls"],
            "passes/tf_quant_ops.h.inc",
        ),
        (
            ["-gen-op-defs"],
            "passes/tf_quant_ops.cc.inc",
        ),
    ],
    tblgen = "@llvm-project//mlir:mlir-tblgen",
    td_file = "passes/tf_quant_ops.td",
    deps = [
        ":quant_td_files",
    ],
)

gentbl_cc_library(
    name = "optimize_inc_gen",

#include "tensorflow/cc/framework/ops.h"
#include "tensorflow/cc/ops/array_ops.h"
#include "tensorflow/cc/ops/resource_variable_ops.h"
#include "tensorflow/cc/ops/standard_ops.h"
#include "tensorflow/compiler/jit/defs.h"
#include "tensorflow/compiler/jit/encapsulate_subgraphs_pass.h"
#include "tensorflow/compiler/jit/node_matchers.h"
#include "tensorflow/compiler/jit/test_util.h"
#include "tensorflow/core/common_runtime/device_factory.h"

        }
      }
    }

    NodeDef def;
    def.set_name(launch->name());
    MergeDebugInfo(NodeDebugInfo(launch->def()), &def);

    // Target the XLA CPU/GPU backends.
    VLOG(2) << "Replacing with XlaLaunch";
    VLOG(2) << "Device is " << launch->requested_device();
    def.set_op("XlaLaunch");
    def.set_device(launch->requested_device());

  //  in TensorFlow op def but it could have an empty value to represent missing
  //  functions. This value can't be represented using MLIR SymbolRefAttr and
  //  instead uses optional symbol ref attribute.
  //
  // TODO(b/182315488): Remove custom handling by finding a better
  // representation in MLIR for empty function names. One option could be to use

		return
	}
	id, isId := e.(*ast.Ident)
	if isId && id.Name == "_" {
		return
	}

	var typ types.Type
	if rtok == token.DEFINE {
		if !isId {
			return
		}
		obj := pass.TypesInfo.Defs[id]
		if obj == nil {
			return
		}
		typ = obj.Type()
	} else {
		typ = pass.TypesInfo.Types[e].Type
	}

	if typ == nil {
		return
	}

// IsField reports whether the variable is a struct field.
func (obj *Var) IsField() bool { return obj.isField }

// Origin returns the canonical Var for its receiver, i.e. the Var object
// recorded in Info.Defs.
//
// For synthetic Vars created during instantiation (such as struct fields or
// function parameters that depend on type arguments), this will be the

// IsField reports whether the variable is a struct field.
func (obj *Var) IsField() bool { return obj.isField }

// Origin returns the canonical Var for its receiver, i.e. the Var object
// recorded in Info.Defs.
//
// For synthetic Vars created during instantiation (such as struct fields or
// function parameters that depend on type arguments), this will be the