Algorithms and ready-made implementations of determining authorship by text?

Question

data mining

There are repeated situations when it is necessary to identify the authors of texts and even short messages on social networks and forums who actively or passively try to hide themselves behind clones (register several accounts), some for the sake of ordinary trolling, some for the sake of increasing their security ... it doesn’t matter .
I know that technologies and algorithms for determining authorship have been around for a long time. For example, one of the diplomas in a parallel to me graduation from a university of the IT faculty successfully demonstrated the study of some kind of statistical analysis algorithm that works with texts at a 'low' level, without semantics, such as the number of commas, the distance between them, the number of words in offer, etc. (there were dozens of criteria).
Enough time has passed, for sure now things are much better (or at least not worse). Maybe someone will advise ready-made implementations, libraries, ... not very monstrous implementations are of interest to me ... a utility that can feed a list of texts is enough for me, and at the output these texts should be grouped into groups / clusters of 'similarity', it will be enough, even if it is something multidimensional /multi-criteria (clusters by criteria or groups of criteria), I could carry out further work with these clusters myself.
The goal, as I defined above, is to search for clones, alternative registrations of people, incl. globally on the Internet (and not just within a single forum).
Update : Looks like I found it myself.
Stylometry, etc. software to use. Almost the firstsame article from google.

To measure privacy and security, researchers have created 2 open source programs - one of them is Jstylo - which recognizes the user's style. The second Anonymousmouth is used to “anonymize” the email by providing the user with specific suggestions for changing the style.

github.com/psal/JStylo-Anonymouth