shape.dims = nullptr;
      continue;
    }
    shape.dims = new int64_t[shape.num_dims];
    for (size_t j = 0; j < shape.num_dims; ++j) {
      shape.dims[j] = c.Value(c.Dim(shape_handle, j));
    }
  }
  if (output_shapes != nullptr) *output_shapes = output_shapes_result;

  // TODO(bgogul): Set output_resource_shapes_and_types.
}

    sequence_shapes = {
        SHAPE_LIST,
        SHAPE_SET,
        SHAPE_FROZENSET,
        SHAPE_TUPLE,
        SHAPE_SEQUENCE,
        SHAPE_TUPLE_ELLIPSIS,
    }
    sequence_shape_to_type = {
        SHAPE_LIST: list,
        SHAPE_SET: set,
        SHAPE_TUPLE: tuple,
        SHAPE_SEQUENCE: list,
        SHAPE_TUPLE_ELLIPSIS: list,
    }

    @dataclass

    std::vector<PartialTensorShape> shapes;
    shapes.reserve(num_values);
    for (int i = 0; i < num_values; ++i) {
      if (num_dims[i] < 0) {
        shapes.emplace_back();
      } else {
        shapes.emplace_back(ArraySlice<int64_t>(
            reinterpret_cast<const int64_t*>(dims[i]), num_dims[i]));
      }
    }
    op_->node_builder.Attr(attr_name, shapes);
    return absl::OkStatus();
  }

      break;
    case tensorflow::AttrValue::kShape: {
      const auto& tensor_shape = default_value.shape();
      if (tensor_shape.unknown_rank()) {
        TFE_OpSetAttrShape(op, attr_name, nullptr, -1, status);
      } else {
        const auto num_dims = tensor_shape.dim_size();
        std::unique_ptr<int64_t[]> dims(new int64_t[num_dims]);

    ],
)

cc_library(
    name = "tf_shape",
    srcs = ["tf_shape.cc"],
    hdrs = ["tf_shape.h"],
    copts = tf_copts(),
    visibility = ["//visibility:public"],
    deps = [
        ":c_api_macros",
        ":tf_shape_internal",
        "//tensorflow/core:framework",
    ],
)

cc_library(
    name = "tf_shape_internal",
    hdrs = ["tf_shape_internal.h"],
    copts = tf_copts(),

  // Create the variable handle.
  TFE_Op* op = TFE_NewOp(ctx, "VarHandleOp", status);
  if (TF_GetCode(status) != TF_OK) return nullptr;
  TFE_OpSetAttrType(op, "dtype", TF_FLOAT);
  TFE_OpSetAttrShape(op, "shape", {}, 0, status);
  TFE_OpSetAttrString(op, "container", "localhost", 0);
  TFE_OpSetAttrString(op, "shared_name", "", 0);
  if (!device_name.empty()) {
    TFE_OpSetDevice(op, device_name.c_str(), status);
  }

  int version = TFE_CUSTOM_DEVICE_VERSION;

  // Computes the rank of the tensor handle.
  //
  // Shapes are specified via callbacks because retrieving the shape of a tensor
  // is a blocking operation for async eager; custom devices should avoid
  // retrieving shapes of tensors they wrap until the custom device tensor's
  // shape is explicitly requested where possible.
  int (*num_dims)(void* data, TF_Status* status);

#include <stack>
#include <unordered_map>
#include <unordered_set>
#include <vector>

#include "tensorflow/core/config/flag_defs.h"
#include "tensorflow/core/config/flags.h"
#include "tensorflow/core/framework/tensor_shape.h"
#include "tensorflow/core/framework/types.h"
#include "tensorflow/core/lib/gtl/array_slice.h"
#include "tensorflow/core/lib/gtl/cleanup.h"
#include "tensorflow/core/lib/gtl/flatmap.h"

And Swagger UI has supported this particular `examples` field for a while. So, you can use it to **show** different **examples in the docs UI**.

The shape of this OpenAPI-specific field `examples` is a `dict` with **multiple examples** (instead of a `list`), each with extra information that will be added to **OpenAPI** too.

<img src="/img/python-types/image06.png">

## Pydantic 모델

<a href="https://docs.pydantic.dev/" class="external-link" target="_blank">Pydantic</a>은 데이터 검증(Validation)을 위한 파이썬 라이브러리입니다.

당신은 속성들을 포함한 클래스 형태로 "모양(shape)"을 선언할 수 있습니다.

그리고 각 속성은 타입을 가지고 있습니다.

이 클래스를 활용하여서 값을 가지고 있는 인스턴스를 만들게 되면, 필요한 경우에는 적당한 타입으로 변환까지 시키기도 하여 데이터가 포함된 객체를 반환합니다.

그리고 결과 객체에 대해서는 에디터의 도움을 받을 수 있게 됩니다.

Pydantic 공식 문서 예시: