import importlib

from fastapi.testclient import TestClient


def get_client() -> TestClient:
    from docs_src.conditional_openapi import tutorial001_py39

    importlib.reload(tutorial001_py39)

    client = TestClient(tutorial001_py39.app)
    return client


def test_disable_openapi(monkeypatch):
    monkeypatch.setenv("OPENAPI_URL", "")
    # Load the client after setting the env var
    client = get_client()

{* ../../docs_src/sub_applications/tutorial001_py39.py hl[3, 6:8] *}

### Подприложение { #sub-application }

Затем создайте подприложение и его *операции пути*.

Это подприложение — обычное стандартное приложение FastAPI, но именно оно будет «смонтировано»:

{* ../../docs_src/sub_applications/tutorial001_py39.py hl[11, 14:16] *}

### Смонтируйте подприложение { #mount-the-sub-application }

De la misma manera, puedes declarar responses de listas de objetos.

Para eso, usa el `typing.List` estándar de Python (o simplemente `list` en Python 3.9 y posteriores):

{* ../../docs_src/extra_models/tutorial004_py39.py hl[18] *}


## Response con `dict` arbitrario { #response-with-arbitrary-dict }

## 多文件上传

FastAPI 支持同时上传多个文件。

可用同一个「表单字段」发送含多个文件的「表单数据」。

上传多个文件时，要声明含 `bytes` 或 `UploadFile` 的列表（`List`）：

{* ../../docs_src/request_files/tutorial002_py39.py hl[8,13] *}

接收的也是含 `bytes` 或 `UploadFile` 的列表（`list`）。


/// note | 技术细节

也可以使用 `from starlette.responses import HTMLResponse`。

{* ../../docs_src/extending_openapi/tutorial001_py39.py hl[1,4,7:9] *}

### Generar el esquema de OpenAPI { #generate-the-openapi-schema }

Luego, usa la misma función de utilidad para generar el esquema de OpenAPI, dentro de una función `custom_openapi()`:

{* ../../docs_src/extending_openapi/tutorial001_py39.py hl[2,15:21] *}

### Modificar el esquema de OpenAPI { #modify-the-openapi-schema }

{* ../../docs_src/extending_openapi/tutorial001_py39.py hl[1,4,7:9] *}

### Das OpenAPI-Schema generieren { #generate-the-openapi-schema }

Verwenden Sie dann dieselbe Hilfsfunktion, um das OpenAPI-Schema innerhalb einer `custom_openapi()`-Funktion zu generieren:

{* ../../docs_src/extending_openapi/tutorial001_py39.py hl[2,15:21] *}

### Das OpenAPI-Schema ändern { #modify-the-openapi-schema }

É isso que vamos ver aqui.

Neste caso, você aceitaria qualquer `dict`, desde que tenha chaves` int` com valores `float`:

{* ../../docs_src/body_nested_models/tutorial009_py39.py hl[7] *}

/// tip | Dica

Leve em condideração que o JSON só suporta `str` como chaves.

Mas o Pydantic tem conversão automática de dados.

from fastapi.testclient import TestClient

from docs_src.custom_response.tutorial006c_py39 import app

client = TestClient(app)


def test_redirect_status_code():
    response = client.get("/pydantic", follow_redirects=False)
    assert response.status_code == 302
    assert response.headers["location"] == "https://docs.pydantic.dev/"


def test_openapi_schema():
    response = client.get("/openapi.json")

{* ../../docs_src/sub_applications/tutorial001_py39.py hl[3, 6:8] *}

### Sub-application { #sub-application }

Then, create your sub-application, and its *path operations*.

This sub-application is just another standard FastAPI application, but this is the one that will be "mounted":

{* ../../docs_src/sub_applications/tutorial001_py39.py hl[11, 14:16] *}

Mas é a maneira mais simples de focar no lado do servidor de WebSockets e ter um exemplo funcional:

{* ../../docs_src/websockets/tutorial001_py39.py hl[2,6:38,41:43] *}

## Crie um `websocket` { #create-a-websocket }

Em sua aplicação **FastAPI**, crie um `websocket`:

{* ../../docs_src/websockets/tutorial001_py39.py hl[1,46:47] *}

/// note | Detalhes Técnicos

Você também poderia usar `from starlette.websockets import WebSocket`.