Let us assume that my prior coauthor plagiarizes our common work and is going to re-publish parts of it under his/her sole name in a conference. I see a preprint of the work on a well-known preprint-site; the venue has not yet taken place. I ran free plagiarism tools; they show between nothing and 20% (tool-dependent) coincidence with public sources that the tools are aware of, whereas typical "new" papers show up to 2%. The rate would go higher if the 20%-tool were aware of our latest commonly published work before the collaboration broke apart. The author made some textual substitutions (variable XXX -> variable YYY, function(object) instead of object.function, etc.) and set-theoretical replacements (functions into powerset instead of binary relations) so that the text formally looks different. The author added a tiny bit of new (though, frankly, mediocre) content.
Does the prior coauthor commit plagiarism, self-plagiarism, or both?
I would like to prevent the prior coauthor publishing the paper. Moreover, I'd like the paper to get off the preprint site. The prior coauthor and me don't speak to each other, so, contacting him/her is unlikely to help. Is it possible for me to actually reach my goal, and, if so, how?