addIfNotNull(defaultRendererOption)
            addAll(directives[PRETTY_RENDERER_OPTION])
        }.fold(defaultRenderer) { acc, prettyRenderingMode ->
            prettyRenderingMode.transformation(acc)
        }

        fun KaSession.safePointer(ktSymbol: KaSymbol): PointerWrapper? {
            val regularPointer = ktSymbol.runCatching {
                createPointerForTest(disablePsiBasedSymbols = false)

  tensorflow::InitMlir y(&argc, &argv);

  // TODO(antiagainst): We are pulling in multiple transformations as follows.
  // Each transformation has its own set of command-line options; options of one
  // transformation can essentially be aliases to another. For example, the
  // -tfl-annotate-inputs has -tfl-input-arrays, -tfl-input-data-types, and

 * KIND, either express or implied.  See the License for the
 * specific language governing permissions and limitations
 * under the License.
 */
package org.apache.maven.internal.transformation.impl;

import java.io.IOException;
import java.io.InputStream;
import java.nio.file.Files;
import java.nio.file.Path;
import java.nio.file.Paths;

import org.apache.maven.model.Model;

#include "tensorflow/compiler/mlir/quantization/tensorflow/python/py_function_lib.h"
#include "tensorflow/core/protobuf/meta_graph.pb.h"

namespace mlir::quant::stablehlo {

// Performs post-calibration graph transformation as part of post-training
// static-range quantization.
//
// The resulting `ModuleOp` contains quantized StableHLO ops serialized in
// `TF::XlaCallModuleOp`s. They are quantized using the statistics collected

            super(name, parent, classPath);
        }

        @Override
        protected byte[] transform(String className, byte[] bytes) {
            // First scan for annotation, and short circuit transformation if not present
            ClassReader classReader = new ClassReader(bytes);

            AnnotationDetector detector = new AnnotationDetector();

See the License for the specific language governing permissions and
limitations under the License.
==============================================================================*/

// This transformation pass reduces the precision of some tensor types by first
// checking if all values within that tensor are within the range.
// This pass is added to aid conversion of models that involve types not

  // ops.
  // $rhs is required to be the tail shape of $lhs, so after transformation the
  // shape of the binary op result is valid. For example, assume the shapes of
  // $input, $lhs and $rhs are [1600], [1,40,40] and [40x1]. After the
  // transformation, the shape of the binary op result is [40x1600], which
  // couldn't be reshaped to [1,40,40]. `IsTailOfShape` constraint is added to

  }

  void runOnOperation() override;
};

// Propagate dequantize op if the next op supports the data type.
// Given the below graph,
// op_before_dequantize -> dequantize_op -> user_op -> rest_op
// the transformation is applied to result the following graph:
// op_before_dequantize -> user_op -> new_dequantize_op -> rest_op
class PropagateDequantizeOpIfAllowed
    : public OpRewritePattern<TF::PartitionedCallOp> {
 public:

// `saved_model_dir` into a MLIR module. Creates MLIR entities into the
// given MLIR `context`. This does not create session internally so it is faster
// and does not perform any graph transformation.
absl::StatusOr<mlir::OwningOpRef<mlir::ModuleOp>>
SavedModelSignatureDefsToMlirImportLite(
    absl::string_view saved_model_dir,
    const std::unordered_set<std::string>& tags,

limitations under the License.
==============================================================================*/

// This header file defines common validators used by TFLite transformation
// passes to validate op attributes or values.

#ifndef TENSORFLOW_COMPILER_MLIR_LITE_UTILS_VALIDATORS_H_
#define TENSORFLOW_COMPILER_MLIR_LITE_UTILS_VALIDATORS_H_