a.internal-link::after {
  /* \00A0 is a non-breaking space
        to make the mark be on the same line as the link
    */
  content: "\00A0↪";
}

.shadow {
  box-shadow: 5px 5px 10px #999;
}

/* Give space to lower icons so Gitter chat doesn't get on top of them */
.md-footer-meta {
  padding-bottom: 2em;
}

.user-list {
  display: flex;
  flex-wrap: wrap;
  margin-bottom: 2rem;
}

.user-list-center {

On the other hand, if you have 2 servers and you are using **100% of their CPU and RAM**, at some point one process will ask for more memory, and the server will have to use the disk as "memory" (which can be thousands of times slower), or even **crash**. Or one process might need to do some computation and would have to wait until the CPU is free again.

This change fixes a [CSRF](https://en.wikipedia.org/wiki/Cross-site_request_forgery) security vulnerability when using cookies for authentication in path operations with JSON payloads sent by browsers.

In versions lower than `0.65.2`, FastAPI would try to read the request payload as JSON even if the `content-type` header sent was not set to `application/json` or a compatible JSON media type (e.g. `application/geo+json`).

    * 基于 **Nginx** 的 Docker 服务器（优化了 Vue-router 配置）
    * Docker 多阶段构建，无需保存或提交编译的代码
    * 在构建时运行前端测试（可禁用）
    * 尽量模块化，开箱即用，但仍可使用 Vue CLI 重新生成或创建所需项目，或复用所需内容
* 使用 **PGAdmin** 管理 PostgreSQL 数据库，可轻松替换为 PHPMyAdmin 或 MySQL
* 使用 **Flower** 监控 Celery 任务
* 使用 **Traefik** 处理前后端负载平衡，可把前后端放在同一个域下，按路径分隔，但在不同容器中提供服务
* Traefik 集成，包括自动生成 Let's Encrypt **HTTPS** 凭证
* GitLab **CI**（持续集成），包括前后端测试

## 全栈 FastAPI + Couchbase

    )
    assert encoded_instance["dt_field"] == instance.dt_field.isoformat()


def test_custom_enum_encoders():
    def custom_enum_encoder(v: Enum):
        return v.value.lower()

    class MyEnum(Enum):
        ENUM_VAL_1 = "ENUM_VAL_1"

    instance = MyEnum.ENUM_VAL_1

    encoded_instance = jsonable_encoder(
        instance, custom_encoder={MyEnum: custom_enum_encoder}

def get_lang_paths() -> List[Path]:
    return sorted(docs_path.iterdir())


def lang_callback(lang: Optional[str]) -> Union[str, None]:
    if lang is None:
        return None
    lang = lang.lower()
    return lang


def complete_existing_lang(incomplete: str):
    lang_path: Path
    for lang_path in get_lang_paths():
        if lang_path.is_dir() and lang_path.name.startswith(incomplete):

                        }
                    )
            if route.openapi_extra:
                deep_dict_update(operation, route.openapi_extra)
            path[method.lower()] = operation
    return path, security_schemes, definitions


def get_fields_from_routes(
    routes: Sequence[BaseRoute],
) -> List[ModelField]:
    body_fields_from_routes: List[ModelField] = []

            unauthorized_headers = {"WWW-Authenticate": f'Basic realm="{self.realm}"'}
        else:
            unauthorized_headers = {"WWW-Authenticate": "Basic"}
        if not authorization or scheme.lower() != "basic":
            if self.auto_error:
                raise HTTPException(
                    status_code=HTTP_401_UNAUTHORIZED,
                    detail="Not authenticated",

* **Starlette**:
    * Will have the next best performance, after Uvicorn. In fact, Starlette uses Uvicorn to run. So, it probably can only get "slower" than Uvicorn by having to execute more code.
    * But it provides you the tools to build simple web applications, with routing based on paths, etc.

    operation_id = f"{operation_id}_{method.lower()}"
    return operation_id


def generate_unique_id(route: "APIRoute") -> str:
    operation_id = f"{route.name}{route.path_format}"
    operation_id = re.sub(r"\W", "_", operation_id)
    assert route.methods
    operation_id = f"{operation_id}_{list(route.methods)[0].lower()}"
    return operation_id