I am using "%mbs" for query markup on unicode content. When using wildcards in query ie niafounk* the highlighting breaks the unicode character at the end and Niafounké becomes <b>Niafounk�</b>�
Highlighting uses linear Metamorph search, and the latter, when expanding a wildcard hit, uses the wordc SQL setting to define what a word is. That is, the asterisk will match characters adjacent to the root/prefix as long as they are in the wordc REX character class. By default, wordc includes only ASCII alphabetic characters and single-quote -- not hi-bit UTF-8. Add UTF-8 chars with this setting: