하지만 `Request` 객체에 직접 접근해야 하는 상황이 있을 수 있습니다.

## `Request` 객체에 대한 세부 사항

**FastAPI**는 실제로 내부에 **Starlette**을 사용하며, 그 위에 여러 도구를 덧붙인 구조입니다. 따라서 여러분이 필요할 때 Starlette의 <a href="https://www.starlette.dev/requests/" class="external-link" target="_blank">`Request`</a> 객체를 직접 사용할 수 있습니다.

`Request` 객체에서 데이터를 직접 가져오는 경우(예: 본문을 읽기)에는 FastAPI가 해당 데이터를 검증하거나 변환하지 않으며, 문서화(OpenAPI를 통한 문서 자동화(로 생성된) API 사용자 인터페이스)도 되지 않습니다.

{* ../../docs_src/first_steps/tutorial001_py39.py *}

Copie o conteúdo para um arquivo `main.py`.

Execute o servidor:

<div class="termy">

```console
$ <font color="#4E9A06">fastapi</font> dev <u style="text-decoration-style:solid">main.py</u>

  <span style="background-color:#009485"><font color="#D3D7CF"> FastAPI </font></span>  Starting development server 🚀

{* ../../docs_src/first_steps/tutorial001_py39.py *}

Copia eso en un archivo `main.py`.

Ejecuta el servidor en vivo:

<div class="termy">

```console
$ <font color="#4E9A06">fastapi</font> dev <u style="text-decoration-style:solid">main.py</u>

  <span style="background-color:#009485"><font color="#D3D7CF"> FastAPI </font></span>  Starting development server 🚀

{* ../../docs_src/app_testing/tutorial002_py39.py hl[27:31] *}

/// note | Примечание

Подробности смотрите в документации Starlette по <a href="https://www.starlette.dev/testclient/#testing-websocket-sessions" class="external-link" target="_blank">тестированию WebSocket</a>.

		exit 1
	fi

	(cd "${MC_BUILD_DIR}" && go build -o "${WORK_DIR}/mc")

	# remove mc source.
	purge "${MC_BUILD_DIR}"

	shred -n 1 -s 1M - 1>"$FILE_1_MB" 2>/dev/null
	shred -n 1 -s 65M - 1>"$FILE_65_MB" 2>/dev/null

	## version is purposefully set to '3' for minio to migrate configuration file

# despite lacking Nvidia CUDA support.
TFCI_BUILD_PIP_PACKAGE_WHEEL_NAME_ARG="--repo_env=WHEEL_NAME=tensorflow"
TFCI_DOCKER_ENABLE=1
TFCI_DOCKER_IMAGE=us-docker.pkg.dev/ml-oss-artifacts-published/ml-public-container/ml-build-arm64:latest
TFCI_DOCKER_PULL_ENABLE=1
TFCI_DOCKER_REBUILD_ARGS="--target=tf ci/official/containers/ml_build_arm64"
TFCI_INDEX_HTML_ENABLE=1
TFCI_LIB_SUFFIX="-cpu-linux-arm64"

```

Это сгенерирует TypeScript SDK в `./src/client`.

Вы можете узнать, как <a href="https://heyapi.dev/openapi-ts/get-started" class="external-link" target="_blank">установить `@hey-api/openapi-ts`</a> и почитать о <a href="https://heyapi.dev/openapi-ts/output" class="external-link" target="_blank">сгенерированном результате</a> на их сайте.

### Использование SDK { #using-the-sdk }

def test_fastapi_cli():
    result = subprocess.run(
        [
            sys.executable,
            "-m",
            "coverage",
            "run",
            "-m",
            "fastapi",
            "dev",
            "non_existent_file.py",
        ],
        capture_output=True,
        encoding="utf-8",
        env={**os.environ, "PYTHONIOENCODING": "utf-8"},
    )
    assert result.returncode == 1, result.stdout

set -e

if [ ! -f run.bash ]; then
	echo 'clean.bash must be run from $GOROOT/src' 1>&2
	exit 1
fi
export GOROOT="$(cd .. && pwd)"

gobin="${GOROOT}"/bin
if ! "$gobin"/go help >/dev/null 2>&1; then
	echo 'cannot find go command; nothing to clean' >&2
	exit 1
fi

"$gobin/go" clean -i std
"$gobin/go" tool dist clean

TFCI_BAZEL_TARGET_SELECTING_CONFIG_PREFIX=linux_cpu
TFCI_BUILD_PIP_PACKAGE_WHEEL_NAME_ARG="--repo_env=WHEEL_NAME=tensorflow_cpu"
TFCI_DOCKER_ENABLE=1
TFCI_DOCKER_IMAGE=us-docker.pkg.dev/ml-oss-artifacts-published/ml-public-container/ml-build:latest
TFCI_DOCKER_PULL_ENABLE=1
TFCI_DOCKER_REBUILD_ARGS="--target=devel ci/official/containers/ml_build"
TFCI_INDEX_HTML_ENABLE=1
TFCI_LIB_SUFFIX="-cpu-linux-x86_64"