* Add valid non-ASCII (Unicode) identifier characters to
  <keyword-char-class>.

* Add custom word boundaries to left side (start) of identifier regular
  expressions. Since the right side (end) will consume all valid
  identifier characters, no word boundary is necessary.

* Remove non-capturing groups around defined regexes, since they are
  automatically added when the defined regex is referenced/included.