retoor/ragnar
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity
Anti spam bot for devRant developer community. This version is more a kind of POC than an actual production version as the source quality shows. Anyway, they're very effective.
28 Commits 1 Branch 0 Tags 269 KiB
Python 96.5%
Makefile 3.5%
main
Go to file
bot 74b1f31aa5 Update export statistics
2024-11-28 03:31:28 +00:00
.gitea/workflows Initial commit 2024-11-27 02:59:42 +01:00
dist Update export statistics 2024-11-28 03:31:28 +00:00
src Update export statistics 2024-11-28 03:31:28 +00:00
.gitignore Initial commit 2024-11-27 02:59:42 +01:00
Makefile Added unit test and regex 2024-11-27 10:54:27 +01:00
pyproject.toml Initial commit 2024-11-27 02:59:42 +01:00
README.md Initial commit 2024-11-27 02:59:42 +01:00
setup.cfg Initial commit 2024-11-27 02:59:42 +01:00
tiggers.js Added tests 2024-11-27 10:56:37 +01:00

README.md

Ragnar

This is an anti spam bot network. It is named after the viking for no obvious reason.

I'm not happy about the quality of the source and it is not a representation of my usual work. If I would've spend more efford there would be some types and I've would use aiohttp and would've used context managers for example. Despite the source lacking a certain quality, the bots work great and are made not to be annoying to the server by not connecting all at once and caching certain things like user profile / user id and if a reand already is flaged for example to not annoy the server.

The bots have user name no-spam[1-4] but flag under a Russian girl name, also for no obvious reason. I liked it more than some technical name. Will probably rename the bots later. Could be that devRants prevents me to do that within a half year. It doesn't matter much, if the bots do a good job, we will barely see them.

I expect this project tomorrow to have deployed fully functional on a server.

In progress

The bots work perfect in sense that they're doing what they're programmed to do. But the programming is not finished yet:

  • the criteria can be better, tips how to optimize are very welcome.
  • at this moment, they can only flag, useless, but we will have indication of future content to be cancelled. Every spam message should have a flag. If not, contact @retoor.
  • the downvote function doesn't work because I couldn't figure out what value I had to post. Who knows it? After this, it's kinda done.
  • a decent deployment on my server. Now it runs on my laptop because it's not done yet and it got late.

How they work

One process starts four bots named no-spam[1-4]. These bots look at new rants.

If there is a new rant:

  1. check if user has more than five posts. If so, it will not be seen as spam.
  2. it will check certain keywords like hacker / money crypto related if so continue to step 3.
  3. user will be informed by the bots that his rant is flagged and what to do about it.
  4. rant will be downvoted by the four bots making it disappear.