*
   * Similarly, we avoid having those methods *implemented* in terms of *other* TestSet methods
   * that will perform holdsLock assertions:
   *
   * - For iterator(), we can accomplish that by not overriding iterator() at all. That way, we
   *   inherit an implementation that forwards to the delegate collection, which performs no

      }

      /*
       * We use this only when getTypeName is available.
       *
       * Well, really, we use this when we think we're running under Java 8, as determined by some
       * logic in the static initializer, which does not check for getTypeName specifically. We
       * should really validate that it works as desired for all Android versions that we support.
       */
      @IgnoreJRERequirement

///

We are going to include this `APIRouter` in the main `FastAPI` app, but first, let's check the dependencies and another `APIRouter`.

## Dependencies { #dependencies }

We see that we are going to need some dependencies used in several places of the application.

So we put them in their own `dependencies` module (`app/dependencies.py`).

	// "$GOROOT/src/mime/multipart" or "$GOPATH/src/golang.org/x/mobile".
	// We extract everything after the $GOROOT or $GOPATH to run on the
	// same relative directory on the target device.
	importPath, isStd, modPath, modDir, err := pkgPath()
	if err != nil {
		return 0, err
	}
	var deviceCwd string
	if isStd {
		// Note that we use path.Join here instead of filepath.Join:

     * StackOverflowError: We can't reliably call setException(error).
     *
     * - Any kind of Error from a listener. Even if we could distinguish that case (by exposing some
     * extra state from AbstractFuture), our options are limited: A call to setException() would be
     * a no-op. We could log, but if that's what we really want, we should modify

  // * We want to behave reasonably for timeouts of 0
  // * We are more responsive to completion than timeouts. This is because parkNanos depends on
  //   system scheduling and as such we could either miss our deadline, or unpark() could be delayed
  //   so that it looks like we timed out even though we didn't. For comparison FutureTask respects
  //   completion preferably and AQS is non-deterministic (depends on where in the queue the waiter

   * must create a new instance with the desired field values, stash it in this field, and then
   * instruct Java serialization to use it instead of the originally created object.
   *
   * <p>We have chosen this approach over at least two alternatives:
   *
   * <ul>
   *   <li>We could change the serialization of this class incompatibly. We have reserved the right

                        }
                        if (response.received && response.resultCode == 0) {

                            /* Before we return, in anticipation of this address being cached we must
                             * augment the addresses name's hashCode to distinguish those resolved by
                             * Lmhosts, WINS, or BCAST. Otherwise a failed query from say WINS would

  /*
   * J2CL's EnumMap does not need the Class instance, so we can use Object.class instead. (Or we
   * could use null, but that messes with our nullness checking, including under J2KT. We could
   * probably work around it by changing how we annotate the J2CL EnumMap, but that's probably more
   * trouble than just using Object.class.)
   *
   * Then we declare the getters for these fields as @GwtIncompatible so that no one can try to use

            //
            // We promise this because, once we cancel that Future, we would no longer be able to
            // tell when any underlying work it is doing is done. Thus, we might start a new task
            // while that underlying work is still running.
            //
            // So that is why we cancel only in the case of CAS success.
            taskFuture.cancel(false);