为你需要检查的地方用标准的Python表达式写个简单的 `assert` 语句（重申，标准的`pytest`）。

```Python hl_lines="2  12  15-18"
{!../../docs_src/app_testing/tutorial001.py!}
```

/// tip | "提示"

注意测试函数是普通的 `def`，不是 `async def`。

还有client的调用也是普通的调用，不是用 `await`。

这让你可以直接使用 `pytest` 而不会遇到麻烦。

///

/// note | "技术细节"

你也可以用 `from starlette.testclient import TestClient`。

**FastAPI** 提供了和 `starlette.testclient` 一样的 `fastapi.testclient`，只是为了方便开发者。但它直接来自Starlette。

      Future<?> possiblyIgnoredError = executorService.submit(executeTask);
      returnValue += (int) listenerLatch.getCount();
      listenerLatch.await();
    }
    return returnValue;
  }

  @Benchmark
  int executeThenAdd_multiThreaded(final int reps) throws InterruptedException {
    Runnable addTask =
        new Runnable() {
          @Override

### Homebrew (recommended)

Run the following command to install the latest stable MinIO package using [Homebrew](https://brew.sh/). Replace ``/data`` with the path to the drive or directory in which you want MinIO to store data.

```sh
brew install minio/stable/minio
minio server /data
```

    * Flatbuffer version update:
        * `GetTemporaryPointer()` bug fixed.

* `tf.data`
    * Add `wait` to `tf.data.Dataset.load`. If `True`, for snapshots written with `distributed_save`, it reads the snapshot while it is being written. For snapshots written with regular `save`, it waits for the snapshot until it's finished. The default is `False` for backward compatibility. Users of `distributed_save` are recommended to set it to `True`.

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

In neueren Versionen von Python gibt es auch einen **asynchronen Kontextmanager**. Sie würden ihn mit `async with` verwenden:

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

   * necessary. No observable state associated with this cache is modified until loading completes.
   *
   * <p>If another call to {@link #get} or {@link #getUnchecked} is currently loading the value for
   * {@code key}, simply waits for that thread to finish and returns its loaded value. Note that
   * multiple threads can concurrently load values for distinct keys.
   *

Vous pouvez donc revenir et voir exactement ce dont vous avez besoin.

## Exécuter le code

Tous les blocs de code peuvent être copiés et utilisés directement (il s'agit en fait de fichiers Python testés).

Pour exécuter l'un de ces exemples, copiez le code dans un fichier `main.py`, et commencez `uvicorn` avec :

<div class="termy">

```console
$ uvicorn main:app --reload

    params {
        text(
            "performance.baselines",
            "",
            display = ParameterDisplay.PROMPT,
            allowEmpty = true,
            description = "The baselines you want to run performance tests against. Empty means default baseline."
        )
        text(
            "testProject",
            "",
            display = ParameterDisplay.PROMPT,
            allowEmpty = false,

      // the result will definitely fit into a long
      long result = 1;
      for (int i = n1 + 1; i <= n2; i++) {
        result *= i;
      }
      return BigInteger.valueOf(result);
    }

    /*
     * We want each multiplication to have both sides with approximately the same number of digits.
     * Currently, we just divide the range in half.
     */
    int mid = (n1 + n2) >>> 1;

    for (Field field : cls.getDeclaredFields()) {
      /*
       * Coverage mode generates synthetic fields.  If we ever add private
       * fields, they will cause similar problems, and we may want to switch
       * this check to isAccessible().
       */
      if (!field.isSynthetic() && field.getType() == String.class) {
        builder.add(field);
      }
    }
    return builder.build();
  }