Handling exceptions in asynchronous code
The basic rules of Python exception handling in asynchronous code are identical to those which apply to synchronous implementations, but you must decide whether to handle exceptions locally in the coroutine/task or propagate them up to the caller.
In previous implementations you can see that a really generic behavior is implemented in the get_data_nonblocking method: if a aiohttp.ClientResponseError is thrown then the result is a string with the message of the error. This breaks the method contract (a dict is expected as result) and the returned message is uninformative. Similarly, the get_data method handles all possible exceptions by just returning them as a string, which is also not the best way to report the anomaly. The following code (available at Chapter 5/extractor_async4.py) shows an alternative implementation which offers some improvements:
import asyncio
import aiohttp
BASE_URL = https://ponyapi.net/v1/character/
def get_data...
