Python Reddit API Comment Loop Test

comment_loop_test.py

#!/usr/bin/env python3
import logging
import sys
import time

import praw


def configure_logging():
    logger = logging.getLogger("praw")
    logger.setLevel(logging.DEBUG)
    logger.addHandler(logging.StreamHandler())


def main():
    configure_logging()

    reddit = praw.Reddit("SITENAME", user_agent="PRAW loop test")
    reddit.config.ratelimit_seconds = 0xFFFF

    last = time.time()
    subreddit = reddit.subreddit("test")
    for i, submission in enumerate(subreddit.new()):
        submission.reply(f"Test comment: {i}")
        now = time.time()
        print(f"{now - last:.2f} {i:2d} {submission.title}")
        last = now


if __name__ == "__main__":
    sys.exit(main())

Ok I just noticed this as of June 2021 but they have changed the message in their error message so that the above code will no longer work. My guess is they do this every so often to deter old bots that are no longer updated from using this method of retry. The new message reads like so:

"Looks like you've been doing that a lot. Take a break for 37 seconds before trying again."

And the above code can be modified like so to make it work

def handle_ratelimit(func, *args, **kwargs):
	while True:
		try:
			func(*args, **kwargs)
			break
		except praw.exceptions.RedditAPIException as exception:
			totalLength = str(exception.items[0].message).split('Looks like you've been doing that a lot. Take a break for ', 1)[1]
			minutesToSleep = totalLength[0].partition("minutes.")[0]
			secondsToSleep = int(minutesToSleep) * 60
			print("Sleeping for " + str(secondsToSleep) + " seconds")
			time.sleep(secondsToSleep)

My personal solution to this uses regex and looks like

def sleep_for_seconds(text):
    find_seconds = r'(\d+) seconds?'
    find_minutes = r'(\d+) minutes?'

    time_seconds = re.search(find_seconds, text)
    time_minutes = re.search(find_minutes, text)

    if time_seconds is not None:
        sleep_time = time_seconds.group(1)
        return float(sleep_time)
    elif time_minutes is not None:
        sleep_time = time_minutes.group(1)
        return float(sleep_time) * 60
    else:
        return float(0)

Thanks everyone for the updates. I updated the gist to work with Python 3.6+ and PRAW 7.0+ where PRAW has built-in support for retries based upon the configuration value ratelimit_seconds.

https://praw.readthedocs.io/en/latest/getting_started/configuration/options.html#miscellaneous-configuration-options