Request handler should be able to timeout blocking sync code in user defined handler

[Pijukatel]

Current purely async implementation of request handler is not capable of triggering timeout for blocking sync code. Such code is created by users and we have no control over it. So we can't expect only async blocking code.

Following simple user defined handler can't currently trigger timeout even if it should.

@crawler.router.default_handler
async def handler(context: BasicCrawlingContext) -> None:
    time.sleep(some_big_time)

[janbuchar]

It is almost impossible to handle this cleanly AFAIK - we cannot interrupt sync code that blocks the event loop. Some options I see:

1. Run request handler in a separate event loop - might break things like playwright in unusual ways
2. Detect when something blocks the event loop for too long and notify the user