Last modified: 2012-08-23 16:07:45 UTC
See the page http://ln.wikipedia.org/wiki/Et%C9%94%C9%9Bl%C9%9B
The wikicode '* [[Nzeté]] tǒ mweté' is rendered as 'Nzeté tǒmweté'.
Changing summary to clarify. Problem is linked to prefix expansion; probably an
extra trim() call is going on and removing the whitespace before the
Can confirm behavior set for Lingala or Arabic, with this sample input:
Recommend adding a parser test case (the language-set option should work now, I
think...) and some poor soul has to dig in the parser. :D