Add Generic Typing for decorate_view to resolve Pyright Issues

[max-muoto]

For projects using Django Ninja that utilize Pyright, the typing for decorate_view leads to a reportUntypedFunctionDecorator error.

Let's take this example from the documentation:

from django.views.decorators.cache import cache_page

@api.get("/test")
@decorate_view(cache_page(5)) # Untyped function decorator obscures type of function
def test_view(request):
    return str(datetime.now())

By having decorate_view utilize generics, we can can avoid this. I also went about generally cleaning up the callable types, Callable[..., Any] is more descriptive than Callable. In Pyright the former would be preferred to avoid unknowns, as Callable == Callable[..., Unknown] instead of being Any. There have been times where I've imported private functions from ninja internals, while not often, it would be nice to avoid unknowns in these cases.

[max-muoto]

@vitalik Was curious if there was anything else I could to help get this merged!