.url(route.address.url)
        .method("CONNECT", null)
        .header("Host", route.address.url.toHostHeader(includeDefaultPort = true))
        .header("Proxy-Connection", "Keep-Alive") // For HTTP/1.0 proxies like Squid.
        .header("User-Agent", USER_AGENT)
        .build()

    val fakeAuthChallengeResponse =
      Response
        .Builder()
        .request(proxyConnectRequest)

As it is discouraged, the interactive docs with Swagger UI won't show the documentation for the body when using `GET`, and proxies in the middle might not support it.

///

## Import Pydantic's `BaseModel` { #import-pydantics-basemodel }

First, you need to import `BaseModel` from `pydantic`:

// Returns a hexadecimal representation of time at the
// time response is sent to the client.
func mustGetRequestID(t time.Time) string {
	return fmt.Sprintf("%X", t.UnixNano())
}

// setEventStreamHeaders to allow proxies to avoid buffering proxy responses
func setEventStreamHeaders(w http.ResponseWriter) {
	w.Header().Set(xhttp.ContentType, "text/event-stream")
	w.Header().Set(xhttp.CacheControl, "no-cache") // nginx to turn off buffering

  @GwtIncompatible // reflection
  @AndroidIncompatible // TODO(cpovirk): ClassNotFoundException: com.google.common.base.Function
  // (I suspect that this and the other similar failures happen with ArbitraryInstances proxies.)
  public void testEqualsAndSerializable() throws Exception {
    new ClassSanityTester().forAllPublicStaticMethods(Functions.class).testEqualsAndSerializable();
  }

  @J2ktIncompatible

    that can monitor, rewrite, and retry calls. The [interceptors doc][interceptors] is a full
    introduction to this new API.

 *  New: APIs to iterate and selectively clear the response cache.
 *  New: Support for SOCKS proxies.
 *  New: Support for `TLS_FALLBACK_SCSV`.
 *  New: Update HTTP/2 support to `h2-16` and `hpack-10`.
 *  New: APIs to prevent retrying non-idempotent requests.
 *  Fix: Drop NPN support. Going forward we support ALPN only.

   * requires us to have a DNS address for both hosts, which only happens after route planning. We
   * can't coalesce connections that use a proxy, since proxies don't tell us the origin server's IP
   * address.
   */
  private fun routeMatchesAny(candidates: List<Route>): Boolean =
    candidates.any {
      it.proxy.type() == Proxy.Type.DIRECT &&

  @GwtIncompatible // reflection
  @AndroidIncompatible // TODO(cpovirk): ClassNotFoundException: com.google.common.base.Function
  // (I suspect that this and the other similar failures happen with ArbitraryInstances proxies.)
  public void testEqualsAndSerializable() throws Exception {
    new ClassSanityTester().forAllPublicStaticMethods(Functions.class).testEqualsAndSerializable();
  }

  @J2ktIncompatible

Uvicorn esperará que el proxy acceda a Uvicorn en `http://127.0.0.1:8000/app`, y luego será responsabilidad del proxy añadir el prefijo extra `/api/v1` encima.

## Sobre proxies con un prefijo de path eliminado

Ten en cuenta que un proxy con prefijo de path eliminado es solo una de las formas de configurarlo.

   */
  @Test
  fun connectTimeoutsAttemptsAlternateRoute() {
    val proxySelector = RecordingProxySelector()
    proxySelector.proxies.add(Proxy(Proxy.Type.HTTP, TestUtil.UNREACHABLE_ADDRESS_IPV4))
    proxySelector.proxies.add(server.proxyAddress)
    server.enqueue(MockResponse(body = "success!"))
    client =
      client
        .newBuilder()
        .proxySelector(proxySelector)

## Version 5.0.0-alpha.8

_2022-06-08_

 *  Fix: Change how `H2_PRIOR_KNOWLEDGE` works with HTTP proxies. Previously OkHttp assumed the
    proxy itself was a prior knowledge HTTP/2 server. With this update, OkHttp attempts a `CONNECT`
    tunnel just as it would with HTTPS. For prior knowledge with proxies OkHttp's is now consistent
    with these curl arguments:

    ```
    curl \
      --http2-prior-knowledge \