NAME

drawpile-monitor - automod for Drawpile servers

SYNOPSIS

Set up an environment and install the dependencies into it:

python -m venv env
. env/bin/activate
pip install -r requirements.txt

Copy the config file template:

cp config.template.ini config.ini

Open the config file and fill in the required information. There's comments inside the file to guide you.

Now you can run this thing:

. env/bin/activate
export DRAWPILE_MONITOR_USER='<WEB ADMIN USERNAME GOES HERE>'
export DRAWPILE_MONITOR_PASS='<WEB ADMIN PASSWORD GOES HERE>'

# To do a dry-run (makes checks and reports to Discord, but does not act):
./drawpile-monitor.py --dryrun

# To run it once:
./drawpile-monitor.py

# To check every 10 seconds:
./drawpile-monitor.py --interval=10

# To check randomly between 20 and 60 seconds (to avoid counter-scripting)
./drawpile-monitor.py --interval=20 --jitter=40

DESCRIPTION

This will hit drawpile-srv's web admin API and check for profanity and other banned words in session titles, session aliases and usernames.

Sessions with bad titles will get two warnings and their tile will be changed to a generic "Session #". On the third infraction, they will be terminated.

Sessions with bad aliases will get terminated immediately, since those can't be changed after the fact.

Users with bad names will get kicked and the session they're in will be notified about it.

It optionally reports the stuff it did to a Discord webhook.

Testing Profanity

You can check what drawpile-monitor would consider offensive by running ./test-offensive.py and typing in the stuff you want to check. It will give you a breakdown of the profanity check along with the final verdict based on those.

If something is getting falsely flagged, you can use this to figure out where it's going wrong so that you can adjust the right knobs.

Remote Configuration

If you specify the config file to reside at a http(s) URL via the -c/--config command-line option or the DRAWPILE_MONITOR_CONFIG environment variable, the configuration will be fetched from that URL instead. The testing script supports this as well.

The response must be a JSON object. There must be config and messages keys at top-level. Both of them must contain objects that only contain strings as values, matching the configuration in the INI file of those sections.

Word lists are specified as arrays of strings under the wordlist, nsfm_wordlist, allowlist and silent_wordlist keys at top-level, all of which are optional.

The above replace wordlist_path, nsfm_wordlist_path, allowlist_path or silent_wordlist_path under config. Those will be ignored if present.

LICENSE

Licensed under the MIT license, see the LICENSE file for details.

SEE ALSO

Drawpile: https://drawpile.net/