How to catch the orignial exception

Question

I'm using the requests module with max_retries option. I would like to catch the exceptions only related to timeouts and slow replies:

import requests
from requests.exceptions import ConnectTimeout, Timeout

URL = 'http://exmaple.com/sleep' # sleeps for 5 seconds before reply

with requests.Session() as s:
    try:
        a = requests.adapters.HTTPAdapter(max_retries=2)
        s.mount('http://', a)
        r = s.get(URL, timeout=1)
    except (ConnectTimeout, Timeout) as err:
        print('# {} - timeout'.format(URL))

But it looks like the underlying urllib3 library throws ReadTimeoutError and requests doesn't catch it and throws ConnectionError instead:

requests.exceptions.ConnectionError: HTTPConnectionPool(host='example.com', port=80): Max retries exceeded with url: /sleep (Caused by ReadTimeoutError("HTTPConnectionPool(host='example.com', port=80): Read timed out. (read timeout=1)"))

I don't want to add ConnectionError to the list because there are other exceptions that inherit from it so it would also catch those.

Is there a way to catch the original exception or perhaps all exceptions in the chain using traceback module.

Answer 1

Ideally, you should catch those other exceptions above ConnectionError and raise them if you want your program to throw an error.

class OtherException(requests.exceptions.ConnectionError):
    pass

try:
    raise OtherException('This is other exception.')
except OtherException as oe:
    raise oe
except requests.exceptions.ConnectionError:
    print('The error you want to catch')

Answer 2

You can use a similar contruct:

import traceback
import logging

try:
    whatever()
except Exception as e:
    logging.error(traceback.format_exc())
    # Your actions here

This will almost catch everything except, for example, KeyboardInterrupt and SystemExit.

Catching those would make the script quite hard to quit.