At the moment, Flarum ignores all non-ASCII characters, replacing them with a single hyphen ... e.g. this thread:
https://discuss.flarum.org/d/2178--
But that's an issue that will have to be resolved at some point ... it's still not clear what Flarum will end up doing. I've been saying all along that Unicode titles should be left unprocessed, and that the best approach would be to include a preference allowing admins to select between verbatim slugs and a more traditional style of slugification.
Traditional slugification does involve dropping punctuation, as you say ... as well as replacing spaces with hyphens and changing everything to lower-case. The problem is that Unicode titles can't be handled that way. I've seen some applications (such as esoTalk) that apply slugification algorithms to Japanese titles, and the results are a bit strange-looking, to say the least. The best approach is to leave the title alone.
And there's nothing wrong with that. Applications remove punctuation from slugs because those characters were either illegal in URLs, or were used with a specific meaning ... but the same can't be said of the fullwidth variants. So why not leave them there?
I think ultimately we'll need a ruling from the devs as to how Flarum will handle slugification of Unicode titles ... and once that becomes clear, we'll have a better idea what needs to be done to represent those slugs accurately within the post content. It's probably pointless worrying about it too much until then.
EDIT: I just realized I'd linked a deleted thread ... I've changed it to one with a Japanese title.
