*     ({@code ArrayList} documentation calls this value the "capacity")
   * @return a new, empty {@code ArrayList} which is guaranteed not to resize itself unless its size
   *     reaches {@code initialArraySize + 1}
   * @throws IllegalArgumentException if {@code initialArraySize} is negative
   */

      while (true) {
        var tryAgain = false
        var readBytesDelivered = -1L
        var errorExceptionToDeliver: IOException? = null

        // 1. Decide what to do in a synchronized block.

        withLock {
          val doReadTimeout = doReadTimeout()
          if (doReadTimeout) {
            readTimeout.enter()
          }
          try {

   *
   * Package-private "shouldn't" be necessary: The *AtomicHelper classes and AbstractFutureState
   * "should" be nestmates, so a call to MethodHandles.lookup or
   * AtomicReferenceFieldUpdater.newUpdater inside *AtomicHelper "should" have access to
   * AbstractFutureState's private fields. However, our open-source build uses `-source 8 -target
   * 8`, so the class files from that build can't express nestmates. Thus, when those class files

     * represents or the drive on which the directory or file resides. Objects
     * other than {@code TYPE_SHARE} or {@code TYPE_FILESYSTEM} will result
     * in 0L being returned.
     *
     * @return the free disk space in bytes of the drive on which this file or
     * directory resides
     * @throws SmbException if an error occurs while retrieving disk space information
     */

   *   <li>A and B have equal {@link #edges() edge sets}.
   *   <li>Every edge in A and B connects the same nodes in the same direction (if any).
   * </ul>
   *
   * <p>Network properties besides {@link #isDirected() directedness} do <b>not</b> affect equality.
   * For example, two networks may be considered equal even if one allows parallel edges and the

    failLeave.countDown(); // release the background thread
  }

  /**
   * Catches a bug where when constructing a service manager failed, later interactions with the
   * service could cause IllegalStateExceptions inside the partially constructed ServiceManager.
   * This ISE wouldn't actually bubble up but would get logged by ExecutionQueue. This obfuscated
   * the original error (which was not constructing ServiceManager correctly).
   */

 * initiated via other mechanisms do not impact the correctness of its methods. For example, if the
 * services are started by some mechanism besides {@link #startAsync}, the listeners will be invoked
 * when appropriate and {@link #awaitHealthy} will still work as expected.
 *
 * <p>Here is a simple example of how to use a {@code ServiceManager} to start a server.
 *

 * initiated via other mechanisms do not impact the correctness of its methods. For example, if the
 * services are started by some mechanism besides {@link #startAsync}, the listeners will be invoked
 * when appropriate and {@link #awaitHealthy} will still work as expected.
 *
 * <p>Here is a simple example of how to use a {@code ServiceManager} to start a server.
 *

   *
   * Package-private "shouldn't" be necessary: The *AtomicHelper classes and AbstractFutureState
   * "should" be nestmates, so a call to MethodHandles.lookup or
   * AtomicReferenceFieldUpdater.newUpdater inside *AtomicHelper "should" have access to
   * AbstractFutureState's private fields. However, our open-source build uses `-source 8 -target
   * 8`, so the class files from that build can't express nestmates. Thus, when those class files

/// note | Nota

Observa que `response_model` es un parámetro del método "decorador" (`get`, `post`, etc). No de tu *path operation function*, como todos los parámetros y el cuerpo.

///

`response_model` recibe el mismo tipo que declararías para un campo de modelo Pydantic, por lo que puede ser un modelo de Pydantic, pero también puede ser, por ejemplo, un `list` de modelos de Pydantic, como `List[Item]`.