# yieldを持つ依存関係

FastAPIは、いくつかの<abbr title='時々"exit"、"cleanup"、"teardown"、"close"、"context managers"、 ...のように呼ばれる'>終了後の追加のステップ</abbr>を行う依存関係をサポートしています。

これを行うには、`return`の代わりに`yield`を使い、その後に追加のステップを書きます。

/// tip | "豆知識"

`yield`は必ず一度だけ使用するようにしてください。

///

/// info | "情報"

これを動作させるには、**Python 3.7** 以上を使用するか、**Python 3.6** では"backports"をインストールする必要があります:

```
pip install async-exit-stack async-generator
```

   * @since 15.0
   */
  @SuppressWarnings("GoodTime") // should accept a java.time.Duration
  void awaitTerminated(long timeout, TimeUnit unit) throws TimeoutException;

  /**
   * Returns the {@link Throwable} that caused this service to fail.
   *
   * @throws IllegalStateException if this service's state isn't {@linkplain State#FAILED FAILED}.
   * @since 14.0
   */
  Throwable failureCause();

  /**

    if (eventSequence.isEmpty()) {
      assertThat(e).isInstanceOf(ConnectionEvent.ConnectStart::class.java)
    } else {
      eventSequence.forEach loop@{
        when (e.closes(it)) {
          null -> return // no open event
          true -> return // found open event
          false -> return@loop // this is not the open event so continue
        }
      }

    // Test that listeners added both before and after the value is available
    // get called correctly.
    for (int i = 0; i < 20; i++) {

      // Right in the middle start up a thread to close the latch.
      if (i == 10) {
        new Thread(() -> latch.countDown()).start();
      }

      future.addListener(listenerLatch::countDown, exec);
    }

    assertSame(Boolean.TRUE, future.get());

    assertThrows(IllegalArgumentException.class, () -> reader.mark(-1));
  }

  public void testMethodsThrowWhenClosed() throws IOException {
    CharSequenceReader reader = new CharSequenceReader("");
    reader.close();

    assertThrows(IOException.class, () -> reader.read());

    assertThrows(IOException.class, () -> reader.read(new char[10]));

    assertThrows(IOException.class, () -> reader.read(new char[10], 0, 10));

		return err
	}

	targetList, err := fetchSubSysTargets(ctx, cfg, subSys, transport)
	if err != nil {
		return err
	}

	for _, target := range targetList {
		defer target.Close()
	}

	for _, target := range targetList {
		yes, err := target.IsActive()
		if err == nil && !yes {
			err = ErrTargetsOffline
		}
		if err != nil {
			return fmt.Errorf("error (%s): %w", target.ID(), err)

	if lrw.Header().Get("Content-Encoding") == "gzip" {
		if lrw.body.Len() > 1<<20 {
			return gzippedBody
		}
		r, err := gzip.NewReader(&lrw.body)
		if err != nil {
			return gzippedBody
		}
		defer r.Close()
		b, _ := io.ReadAll(io.LimitReader(r, 10<<20))
		return b
	}
	// If there was an error response or body logging is enabled
	// then we return the body contents

elasticsearch.url: "http://localhost:9201"
```

#### Q. I imported "fess\_log.ndjson" but no results found.

             * because we might close the socket during a
             * request. However the chances are slim and the
             * retry code should ensure the overall request
             * is serviced. The alternative complicates things
             * more than I think is worth it.
             */

            if( socket != null ) {
                socket.close();
                socket = null;
            }

This can help ensure that clients do not hold onto pooled connections for too long.
This is not strictly required; Ztunnel clients have both keepalives (which will start to fail once the other end is shut down) and pool idle timeouts (which will close the connection when it has had no activity for some period; because the backend is gone, there will be no activity).
However, it would be ideal to give immediate feedback rather than relying on these mechanisms.