當你在「大型程式碼庫」中，於許多「路徑操作」反覆使用「相同的依賴」時，這會特別有用。

## 要不要使用 `async` { #to-async-or-not-to-async }

因為依賴也會由 **FastAPI** 呼叫（就像你的「路徑操作函式」），所以在定義函式時套用相同的規則。

你可以使用 `async def` 或一般的 `def`。

而且你可以在一般 `def` 的「路徑操作函式」中宣告 `async def` 的依賴，或在 `async def` 的「路徑操作函式」中宣告 `def` 的依賴，等等。

都沒關係。**FastAPI** 會知道該怎麼做。

/// note | 注意

如果你不熟悉，請參考文件中的 [Async: "In a hurry?"](../../async.md#in-a-hurry) 一節，瞭解 `async` 與 `await`。

///

```Python
with open("file.txt") as file:
    file.read()
```

Dans les versions récentes de Python, il existe aussi un **gestionnaire de contexte asynchrone**. Vous l'utiliseriez avec `async with` :

```Python
async with lifespan(app):
    await do_stuff()
```

## `async` Olsa da Olmasa da { #to-async-or-not-to-async }

Dependency'ler de **FastAPI** tarafından çağrılacağı için (tıpkı *path operation function*'larınız gibi), fonksiyonları tanımlarken aynı kurallar geçerlidir.

`async def` ya da normal `def` kullanabilirsiniz.

from fastapi.responses import HTMLResponse

app = FastAPI()


@app.post("/files/")
async def create_files(
    files: Annotated[list[bytes], File(description="Multiple files as bytes")],
):
    return {"file_sizes": [len(file) for file in files]}


@app.post("/uploadfiles/")
async def create_upload_files(
    files: Annotated[
        list[UploadFile], File(description="Multiple files as UploadFile")

class Model(BaseModel):
    param: str = Field(alias="param_alias")


@app.get("/query")
async def query_model(data: Model = Query()):
    return {"param": data.param}


@app.get("/header")
async def header_model(data: Model = Header()):
    return {"param": data.param}


@app.get("/cookie")
async def cookie_model(data: Model = Cookie()):
    return {"param": data.param}

### Files and Async { #files-and-async }

In most cases, file-like objects are not compatible with async and await by default.

For example, they don't have an `await file.read()`, or `async for chunk in file`.

        allowEmpty = false,
        description =
            "Command line option for the performance test task to enable profiling. " +
                "For example `async-profiler`, `async-profiler-heap`, `async-profiler-wall`, `async-profiler-all` or `jfr`. Use `none` for benchmarking only.",
    )
}

object AdHocPerformanceScenarioLinux : AdHocPerformanceScenario(Os.LINUX)

```Python
with open("file.txt") as file:
    file.read()
```

En versiones recientes de Python, también hay un **async context manager**. Lo usarías con `async with`:

```Python
async with lifespan(app):
    await do_stuff()
```

### Gérer les fichiers et async { #files-and-async }

Dans la plupart des cas, les objets de type fichier ne sont pas compatibles avec `async` et `await` par défaut.

Par exemple, ils n'ont pas de `await file.read()`, ni de `async for chunk in file`.

  int num_retvals = 1;
  TFE_Execute(matmul, &retvals[0], &num_retvals, status);
  EXPECT_EQ(TF_GetCode(status), TF_OK) << TF_Message(status);

  // TODO(gjn): Add support for waiting on async local mirrors
  if (!remote && !async && !remote_func_outputs) {
    auto remote_arg =
        tensorflow::TensorHandleFromInterface(tensorflow::unwrap(h1_task2));
    // The input handles should never change since they have been mirrored.