{!../../docs_src/custom_response/tutorial004.py!}
```

In this example, the function `generate_html_response()` already generates and returns a `Response` instead of returning the HTML in a `str`.

By returning the result of calling `generate_html_response()`, you are already returning a `Response` that will override the default **FastAPI** behavior.

   * method will throw {@link IllegalArgumentException}. You should ensure your input is valid <a
   * href="http://en.wikipedia.org/wiki/UTF-16">UTF-16</a> before calling this method.
   *
   * <p><b>Note:</b> When implementing an escaper it is a good idea to override this method for
   * efficiency by inlining the implementation of {@link #nextEscapeIndex(CharSequence, int, int)}

   * value in the manner of {@link Number#doubleValue}.
   *
   * <p>Elements are copied from the argument collection as if by {@code collection.toArray()}.
   * Calling this method is as thread-safe as calling that method.
   *
   * @param collection a collection of {@code Number} instances
   * @return an array containing the same values as {@code collection}, in the same order, converted

   *
   * <p>The returned iterable's iterator supports {@code remove()} if the iterator of the underlying
   * iterable supports it. Note that it is <i>not</i> possible to delete the last skipped element by
   * immediately calling {@code remove()} on that iterator, as the {@code Iterator} contract states
   * that a call to {@code remove()} before a call to {@code next()} will throw an {@link
   * IllegalStateException}.
   *

   * @since 30.0
   */
  @J2ktIncompatible
  @GwtIncompatible // concurrency
  public static void awaitTerminationUninterruptibly(ExecutorService executor) {
    // TODO(cpovirk): We could optimize this to avoid calling nanoTime() at all.
    verify(awaitTerminationUninterruptibly(executor, Long.MAX_VALUE, NANOSECONDS));
  }

  /**
   * Invokes {@code executor.}{@link ExecutorService#awaitTermination(long, TimeUnit)

                                /* Apparently once a successfull NTLMSSP login occurs, the
                                 * server will return "Access denied" even if a logoff is
                                 * sent. Unfortunately calling disconnect() doesn't always
                                 * actually shutdown the connection before other threads
                                 * have committed themselves (e.g. InterruptTest example).

 * {@link java.net.ServerSocket ServerSocket}, and you would like to ensure that the {@code
 * ServerSocket} is closed even if the {@code MyServer} object is garbage-collected without calling
 * its {@code close} method. You could use a finalizer to accomplish this, but that has a
 * number of well-known problems. Here is how you might use this class instead:
 *
 * <pre>{@code

   * allow in normal conditions. This is enforced by "throttling" incoming requests as needed. For
   * example, we could compute the appropriate throttle time for an incoming request, and make the
   * calling thread wait for that time.
   *
   * The simplest way to maintain a rate of QPS is to keep the timestamp of the last granted
   * request, and ensure that (1/QPS) seconds have elapsed since then. For example, for a rate of

// Some TF ops need a step container to be set to limit the lifetime of some
// resources (mostly TensorArray and Stack, used in while loop gradients in
// graph mode). Calling this on a context tells it to start a step.
TF_CAPI_EXPORT extern void TFE_ContextStartStep(TFE_Context* ctx);

// Ends a step. When there is no active step (that is, every started step has

 * percentiles together takes about 55% as long as computing them separately.
 *
 * <p>When calling {@link ScaleAndIndex#compute} (in {@linkplain ScaleAndIndexes#compute either
 * form}), the memory requirement is 8*N bytes for the copy of the dataset plus an overhead which is
 * independent of N (but depends on the quantiles being computed). When calling {@link
 * ScaleAndIndex#computeInPlace computeInPlace} (in {@linkplain ScaleAndIndexes#computeInPlace