Ce qui améliore le support pour les modèles Pydantic avec :

    * de l'auto-complétion
    * des vérifications de type
    * du "refactoring" (ou remaniement de code)
    * de la recherche
    * de l'inspection

## Utilisez le modèle

Dans la fonction, vous pouvez accéder à tous les attributs de l'objet du modèle directement :

```Python hl_lines="21"
{!../../../docs_src/body/tutorial002.py!}
```

For large responses, returning a `Response` directly is much faster than returning a dictionary.

                # explicit default status_code, and to extract it from them, instead of
                # doing this inspection tricks, that would probably be in the future
                # TODO: probably make status_code a default class attribute for all
                # responses in Starlette
                response_signature = inspect.signature(current_response_class.__init__)

    if inspect.isroutine(call):
        return inspect.iscoroutinefunction(call)
    if inspect.isclass(call):
        return False
    dunder_call = getattr(call, "__call__", None)  # noqa: B004
    return inspect.iscoroutinefunction(dunder_call)


def is_async_gen_callable(call: Callable[..., Any]) -> bool:
    if inspect.isasyncgenfunction(call):
        return True

import asyncio
import dataclasses
import email.message
import inspect
import json
from contextlib import AsyncExitStack
from enum import Enum, IntEnum
from typing import (
    Any,
    Callable,
    Coroutine,
    Dict,
    List,
    Optional,
    Sequence,
    Set,
    Tuple,
    Type,
    Union,
)

from fastapi import params
from fastapi._compat import (
    ModelField,
    Undefined,