Voici un petit bout de cette histoire.

## Alternatives

Je crée des API avec des exigences complexes depuis plusieurs années (Machine Learning, systèmes distribués, jobs asynchrones, bases de données NoSQL, etc), en dirigeant plusieurs équipes de développeurs.

Dans ce cadre, j'ai dû étudier, tester et utiliser de nombreuses alternatives.

///

/// tip | Tipp

Wenn Sie in Ihren Tests neben dem Senden von Anfragen an Ihre FastAPI-Anwendung auch `async`-Funktionen aufrufen möchten (z. B. asynchrone Datenbankfunktionen), werfen Sie einen Blick auf die [Async-Tests](../advanced/async-tests.md){.internal-link target=_blank} im Handbuch für fortgeschrittene Benutzer.

///

## Tests separieren

   * of refreshes are specified in {@link LoadingCache#refresh}, and are performed by calling {@link
   * CacheLoader#reload}.
   *
   * <p>As the default implementation of {@link CacheLoader#reload} is synchronous, it is
   * recommended that users of this method override {@link CacheLoader#reload} with an asynchronous
   * implementation; otherwise refreshes will be performed during unrelated cache read and write
   * operations.

{* ../../docs_src/first_steps/tutorial001.py hl[7] *}

C'est une fonction Python.

Elle sera appelée par **FastAPI** quand une requête sur l'URL `/` sera reçue via une opération `GET`.

Ici, c'est une fonction asynchrone (définie avec `async def`).

---

Vous pourriez aussi la définir comme une fonction classique plutôt qu'avec `async def` :

{* ../../docs_src/first_steps/tutorial003.py hl[7] *}

/// note

      encodings[ 0x13] = encoding // Device Control 3 (oft. XOFF)
      encodings[ 0x14] = encoding // Device Control 4
      encodings[ 0x15] = encoding // Negative Acknowledgment
      encodings[ 0x16] = encoding // Synchronous idle
      encodings[ 0x17] = encoding // End of Transmission Block
      encodings[ 0x18] = encoding // Cancel
      encodings[ 0x19] = encoding // End of Medium
      encodings[ 0x1a] = encoding // Substitute

  ) {
    // Do nothing.
  }

  override fun noNewExchanges() {
    // Do nothing.
  }

  override fun cancel() {
    canceled = true
    // Close the raw socket so we don't end up doing synchronous I/O.
    rawSocket?.closeQuietly()
  }

  override fun retry(): RoutePlanner.Plan =
    ConnectPlan(
      taskRunner = taskRunner,
      connectionPool = connectionPool,

  /**
   * Returns a {@code CacheLoader} which wraps {@code loader}, executing calls to {@link
   * CacheLoader#reload} using {@code executor}.
   *
   * <p>This method is useful only when {@code loader.reload} has a synchronous implementation, such
   * as {@linkplain #reload the default implementation}.
   *
   * @since 17.0
   */
  @GwtIncompatible // Executor + Futures
  public static <K, V> CacheLoader<K, V> asyncReloading(

    }

    @GwtIncompatible @J2ktIncompatible private static final long serialVersionUID = 0;
  }

  /**
   * Returns a supplier whose {@code get()} method synchronizes on {@code delegate} before calling
   * it, making it thread-safe.
   */
  @J2ktIncompatible
  public static <T extends @Nullable Object> Supplier<T> synchronizedSupplier(
      Supplier<T> delegate) {

For "synchronous" (contrary to "asynchronous") they commonly also use the term "sequential", because the computer / program follows all the steps in sequence before switching to a different task, even if those steps involve waiting.

 * * `Response.body().byteStream().close()`
 * * `Response.body().bytes()`
 * * `Response.body().string()`
 *
 * There is no benefit to invoking multiple `close()` methods for the same response body.
 *
 * For synchronous calls, the easiest way to make sure a response body is closed is with a `try`
 * block. With this structure the compiler inserts an implicit `finally` clause that calls
 * [close()][Response.close] for you.
 *
 * ```java