specifics of a local system, we rely on hermetic Python (see
[rules_python](https://github.com/bazelbuild/rules_python)) for all build
and test commands executed via Bazel. This means that your system Python
installation will be ignored during the build and Python interpreter itself
as well as all the Python dependencies will be managed by bazel directly.

### Specifying Python version

        .trailersDelay(1, TimeUnit.SECONDS)
        .build(),
    )

    val call = client.newCall(Request(server.url("/")))
    call.execute().use { response ->
      val source = response.body.source()
      assertThat(response.header("h1")).isEqualTo("v1")
      assertThat(source.readUtf8(5)).isEqualTo("Hello")
      call.cancelLater(500.milliseconds)
      val trailersDelay =
        measureTime {

        override fun requestHeadersEnd(
          call: Call,
          request: Request,
        ) = TODO()

        override fun requestBodyStart(call: Call) = TODO()

        override fun requestBodyEnd(
          call: Call,
          byteCount: Long,
        ) = TODO()

        override fun requestFailed(
          call: Call,
          ioe: IOException,

And when solving the dependency, **FastAPI** will call this `checker` like:

```Python
checker(q="somequery")
```

...and pass whatever that returns as the value of the dependency in our *path operation function* as the parameter `fixed_content_included`:

{* ../../docs_src/dependencies/tutorial011_an_py39.py hl[22] *}

/// tip

All this might seem contrived. And it might not be very clear how is it useful yet.

    call: Call,
    connection: Connection,
  ) {
    logWithTime("connectionReleased")
  }

  override fun requestHeadersStart(call: Call) {
    logWithTime("requestHeadersStart")
  }

  override fun requestHeadersEnd(
    call: Call,
    request: Request,
  ) {
    logWithTime("requestHeadersEnd")
  }

  override fun requestBodyStart(call: Call) {

  tensorflow::mutex mu TF_ACQUIRED_AFTER(TF_Graph::mu);
  int last_num_graph_nodes;

  // If true, TF_SessionRun and similar methods will call
  // ExtendSessionGraphHelper before running the graph (this is the default
  // public behavior). Can be set to false if the caller needs to call
  // ExtendSessionGraphHelper manually.
  std::atomic<bool> extend_before_run;
};

struct TF_ImportGraphDefOptions {

   * 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}.
   *
   * <p><b>{@code Stream} equivalent:</b> {@link Stream#skip}
   *
   * @since 3.0
   */

        assertEquals(SimpleCircuitBreaker.State.OPEN, circuitBreaker.getState());

        // Wait for timeout
        Thread.sleep(150);

        // Next call should transition to HALF_OPEN
        String result = circuitBreaker.call(() -> "success");
        assertEquals("success", result);
        assertEquals(SimpleCircuitBreaker.State.HALF_OPEN, circuitBreaker.getState());
    }

    @Test

        .build()
    val latch = CountDownLatch(1)
    val exceptionRef = AtomicReference<Throwable>()
    val call = client.newCall(request)
    call.timeout().timeout(250, TimeUnit.MILLISECONDS)
    call.enqueue(
      object : Callback {
        override fun onFailure(
          call: Call,
          e: IOException,
        ) {
          exceptionRef.set(e)
          latch.countDown()
        }

 * will prefer pooled connections. Only pooled HTTP/2 connections are used for such de-duplication.
 *
 * It is possible to cancel the finding process by canceling its call.
 *
 * Implementations of this interface are not thread-safe. Each instance is thread-confined to the
 * thread executing the call.
 */
interface RoutePlanner {
  val address: Address

  /** Follow-ups for failed plans and plans that lost a race. */