std::vector<int64_t> shape;
      shape.push_back(static_cast<int64_t>(data.size()));
      attributes.emplace_back(builder.getNamedAttr(
          "window_strides", BuildVhloTensorV1Attr(shape, data, builder)));
    }
    if (!(op->padding.empty())) {
      std::vector<int64_t> shape;
      shape.push_back(static_cast<int64_t>(op->padding.size()) / 2);
      shape.push_back(2);

	{`c:/`, `c:/`, ``},
	{`c:/foo`, `c:/`, `foo`},
	{`c:/foo/bar`, `c:/foo/`, `bar`},
	{`//host/share`, `//host/share`, ``},
	{`//host/share/`, `//host/share/`, ``},
	{`//host/share/foo`, `//host/share/`, `foo`},
	{`\\host\share`, `\\host\share`, ``},
	{`\\host\share\`, `\\host\share\`, ``},
	{`\\host\share\foo`, `\\host\share\`, `foo`},
}

func TestSplit(t *testing.T) {
	var splittests []SplitTest

            tensor_spec.TensorSpec(
                shape=shape, dtype=dtypes.float32, name='input_tensor'
            )
        ),
    )
    return model

  # Prepares sample einsum input data shapes.
  # This function returns:
  # 1. Shape for input 1
  # 2. Shape for input 2
  # 3. Shape for bias
  # 4. Signature for input 1 (Could contain None dimension)

}

// Gets the shape of operand, assuming it is a dynamic shape with static rank.
Value getShapeValue(Location loc, Value operand, PatternRewriter &rewriter) {
  RankedTensorType resultType =
      mlir::dyn_cast<RankedTensorType>(operand.getType());
  return rewriter.create<shape::ShapeOfOp>(
      loc,
      RankedTensorType::get(/*shape=*/{resultType.getRank()},

				&field.Error{
					Type:  field.ErrorTypeInvalid,
					Field: "shape[0].utilization",
				},
				&field.Error{
					Type:  field.ErrorTypeInvalid,
					Field: "shape[1].score",
				},
				&field.Error{
					Type:  field.ErrorTypeInvalid,
					Field: "shape[2].score",
				},
				&field.Error{
					Type:  field.ErrorTypeInvalid,
					Field: "shape[3].utilization",
				},
			}),
		},
	}

		// UNC Absolute
		{`\\srv\share\long`, `\\?\UNC\srv\share\long`},
		{`//srv/share/long`, `\\?\UNC\srv\share\long`},
		{`/\srv/share/long`, `\\?\UNC\srv\share\long`},
		{`\\srv\share\long\`, `\\?\UNC\srv\share\long\`},
		{`\\srv\share\bar\.\long`, `\\?\UNC\srv\share\bar\long`},
		{`\\srv\share\bar\..\long`, `\\?\UNC\srv\share\long`},
		{`\\srv\share\bar\..\..\long`, `\\?\UNC\srv\share\long`}, // share name is not removed by ".."

// Converts an TensorFlow shape to the one used in MLIR.
void ConvertToMlirShape(const TensorShape& input_shape,
                        llvm::SmallVectorImpl<int64_t>* shape);

// Converts an TensorFlow shape proto to the one used in MLIR.
Status ConvertToMlirShape(const TensorShapeProto& input_shape,
                          llvm::SmallVectorImpl<int64_t>* shape);

    weight_row = array_ops.ones(
        shape=array_ops.shape(input_vocabs_placeholder), dtype=dtypes.float32
    )
    # shape: (?, 2)
    weight = array_ops.transpose_v2(
        array_ops_stack.stack([weight_row, weight_row])
    )
    # shape: (2, 2)
    output_tensor = math_ops.matmul(matmul_input, weight)

    return input_vocabs_placeholder, lookup_vals, output_tensor

  auto value_type = mlir::cast<RankedTensorType>(value.getType());
  auto shape = value_type.getShape();
  SmallVector<int64_t, 4> transposed_shape(shape.begin(), shape.end());
  for (int i = 0, end = shape.size(); i < end; ++i) {
    transposed_shape[i] = shape[permutation[i]];
  }
  auto transposed_type =
      RankedTensorType::get(transposed_shape, value_type.getElementType());

}

// Converts the tensor shape proto into an MLIR shape attribute.
absl::StatusOr<mlir::Attribute> ConvertTensorShapeProto(
    const TensorShapeProto& shape, mlir::MLIRContext* context) {
  if (shape.unknown_rank())
    return mlir::TF::ShapeAttr::get(context, std::nullopt);

  llvm::SmallVector<int64_t, 4> dims;
  dims.reserve(shape.dim().size());
  for (const auto& dim : shape.dim()) {