auto operand_type =
              mlir::dyn_cast_or_null<ShapedType>(operand.getType());
          return result_type && operand_type && result_type.hasRank() &&
                 operand_type.hasRank() &&
                 result_type.getRank() == operand_type.getRank();
        });
    if (!is_valid_move) return;
  }

	// stop things from getting too large.
	maxFlateBlockTokens = 1 << 14
	maxStoreBlockSize   = 65535
	hashBits            = 17 // After 17 performance degrades
	hashSize            = 1 << hashBits
	hashMask            = (1 << hashBits) - 1
	maxHashOffset       = 1 << 24

	skipNever = math.MaxInt32
)

type compressionLevel struct {
	level, good, lazy, nice, chain, fastSkipHashing int
}

  // Correctly set output shapes of split op output if input shape is statically
  // known.
  mlir::Type output_type;
  auto input_type = mlir::cast<mlir::TensorType>(src_input.getType());

  if (input_type.hasRank()) {
    if (input_type.getShape()[split_dimension] == mlir::ShapedType::kDynamic) {
      output_type = input_type;
    } else {
      auto shape = llvm::to_vector<4>(input_type.getShape());

  // TODO(b/158709815): Add support for batches with padded NMS.
  auto boxes_type =
      mlir::dyn_cast<RankedTensorType>(func_.getFunctionType().getInput(0));
  if (boxes_type == nullptr || !boxes_type.hasRank() ||
      boxes_type.getRank() != 2) {
    return func_.emitWarning() << "TFLite does not support batched input for "
                                  "non_max_suppression_padded";
  }
  return success();
}

  return shaped_type.hasStaticShape();
}

bool HasStaticShapeAtDims(Value value, const ArrayRef<int> dims) {
  auto shaped_type = mlir::dyn_cast<ShapedType>(value.getType());
  if (!shaped_type || !shaped_type.hasRank()) return false;

  for (auto dim : dims) {
    if (shaped_type.isDynamicDim(dim)) return false;
  }
  return true;
}

Type CloneTypeWithNewElementType(Type old_type, Type element_type) {

      return failure();
    }

    Value input = tf_op.getInputs();
    int quant_dim = -1;
    auto input_type = mlir::cast<ShapedType>(input.getType());
    if (PerAxis) {
      if (!input_type.hasRank()) {
        tf_op.emitError("The input should have known rank for per-channel op.");
        return failure();
      }
      // This is a special case that the quant_dim is the last dimensions.

bool IsTailOfShape(Type type1, Type type2) {
  auto tail_type = mlir::dyn_cast<ShapedType>(type1);
  auto full_type = mlir::dyn_cast<ShapedType>(type2);
  if (!tail_type || !full_type || !tail_type.hasRank() ||
      !full_type.hasRank() || tail_type.getRank() > full_type.getRank())
    return false;
  auto i1 = tail_type.getShape().rbegin(), e1 = tail_type.getShape().rend();
  auto i2 = full_type.getShape().rbegin();

    auto out_size_element_ty = out_size_ty.getElementType();

    // Input should be rank 4.
    if (!input_ty.hasRank() || input_ty.getRank() != 4) {
      return failure();
    }

    // Check that out_size is rank-1, length-2. Otherwise the size is not legal.
    if (!out_size_ty.hasRank() || out_size_ty.getRank() != 1 ||
        out_size_ty.getShape()[0] != 2) {
      return failure();
    }

  // Output batch and channel dimension can be obtained using utilities from
  // tensorflow/core/util/tensor_format.h.
  if (input_ty.hasRank()) {
    return_shape[GetTensorBatchDimIndex(num_dims, format)] =
        input_ty.getDimSize(GetTensorBatchDimIndex(num_dims, format));
  }
  if (filter_ty.hasRank()) {
    return_shape[GetTensorFeatureDimIndex(num_dims, format)] =
        filter_ty.getDimSize(GetFilterTensorOutputChannelsDimIndex(

                                     PatternRewriter &rewriter) {
  auto lhs_type = mlir::cast<ShapedType>(op.getLhs().getType());
  auto rhs_type = mlir::cast<ShapedType>(op.getRhs().getType());
  if (!lhs_type.hasRank() || !rhs_type.hasRank()) {
    return rewriter.notifyMatchFailure(op, "unsupported unranked input type");
  }
  if (lhs_type.getRank() < 1 || 2 < lhs_type.getRank() ||
      rhs_type.getRank() < 1 || 2 < rhs_type.getRank()) {