@jsonify, I will try to break out the process some more
Punctuation regex, yielded this article http://www.regular-expressions.info/posixbrackets.html
That was working well, except it was leaving in $ and ^, so added those on top.
Then it was failing the last test with the diaeresis on the o (ö). It was stripping it off and turning it into a binary e.g. <<107>> or whatever. Then, after about an hour of trying stuff, I tried the option for unicode u at the end of the ~r/regex/u.
Hope that was helpful.
M
