profanity-check VS google-profanity-words

You could look into the profanity-check package.

Ovako nesto: https://github.com/vzhou842/profanity-check ? MIT licenca plus deluje da radi dosta brzo sa dobrim metrikama

This is what hueristics is made for. Don't use a regex for this, use https://pypi.org/project/profanity-check/

General profanity 'checker' - https://github.com/vzhou842/profanity-check

Unfortunately, it seems that its original author never accepted our previous Pull Request to revive the profanity-check package. As a result, Github issues were kept being opened by other developers asking about the library and issues they had while using it. What surprised us the most, was this one: https://github.com/vzhou842/profanity-check/issues/28, where someone actually recommended our fork of the project as a "new version available here, that solves the issue very well".

I don't see how using a library would result in any functional difference from what you are already doing (presumably having some variable/text file that represents a list of words). This library for example literally just returns a list of words considered profanity by Google, and can be used in full isolation from the filter implementation.