我们将`SessionLocal()`请求的创建和处理放在一个`try`块中。

然后我们在finally块中关闭它。

通过这种方式，我们确保数据库会话在请求后始终关闭，即使在处理请求时出现异常也会关闭。

///

### 关于`request.state`

`request.state`是每个`Request`对象的属性。它用于存储附加到请求本身的任意对象，例如本例中的数据库会话。您可以在[Starlette 的关于`Request`state](https://www.starlette.io/requests/#other-state)的文档中了解更多信息。

对于这种情况下，它帮助我们确保在整个请求中使用单个数据库会话，然后关闭（在中间件中）。

### 使用`yield`依赖项与使用中间件的区别

在此处添加**中间件**与`yield`的依赖项的作用效果类似，但也有一些区别：

- Kube-apiserver: Added Alpha features to allow API server authz to check the context of requests:
  - The `AuthorizeWithSelectors` feature gate enables including field and label selector information from requests in webhook authorization calls.

- The apiserver debug endpoint `/debug/api_priority_and_fairness/dump_requests` has been extended to dump executing requests as well as queued ones.  A column for StartTime has been added to the returned table, with the queued requests having a StartTime of "0001-01-01T00:00:00Z".  The executing requests have a RequestIndexInQueue of -1, and the QueueIndex is also -1 for priority levels without queues. ([#119009](https://github.com/kubernetes/kubernetes/pull/119009),...

- Kubeadm: make sure that a variety of API server requests are retried during "init", "join", "upgrade", "reset" workflows. Prior to this change some API server requests, such as, creating or updating ConfigMaps were "one-shot"...

    String INDEXER_WEBFS_MAX_DOCUMENT_REQUEST_SIZE = "indexer.webfs.max.document.request.size";

    /** The key of the configuration. e.g. 10000 */
    String INDEXER_DATA_MAX_DOCUMENT_CACHE_SIZE = "indexer.data.max.document.cache.size";

    /** The key of the configuration. e.g. 1048576 */
    String INDEXER_DATA_MAX_DOCUMENT_REQUEST_SIZE = "indexer.data.max.document.request.size";

    /** The key of the configuration. e.g. 100 */

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`).

So, a request with a content type of `text/plain` containing JSON data would be accepted and the JSON data would be extracted.

* Add `metrics-port` to kube-proxy cmd flags. ([#72682](https://github.com/kubernetes/kubernetes/pull/72682), [@whypro](https://github.com/whypro))

    now served with tensorflow/serving, it will accept requests using 'inputs'
    and 'outputs'. Starting at 1.2, such a model will accept the keys specified
    during export. Therefore, inference requests using 'inputs' and 'outputs'
    may start to fail. To fix this, either update any inference clients to send
    requests with the actual input and output keys used by the trainer code, or

        * requests for a TLS client certificate for any node are approved if the CSR creator has `create` permission on the `certificatesigningrequests` resource and `nodeclient` subresource in the `certificates.k8s.io` API group

		},
	}

	// string to represent V2 signed HTTP request.
	reqV2Str := "V2 Signed HTTP request"
	// string to represent V4 signed HTTP request.
	reqV4Str := "V4 Signed HTTP request"
	// Collection of HTTP request and ResponseRecorder and request type string.
	type inputReqRec struct {
		req     *http.Request
		rec     *httptest.ResponseRecorder
		reqType string
	}