Hi,
I'm not sure I understand the problem, but your XML samples contain numeric character references
(Ӓ) and you say you get unicode references as well. I suspect you mean utf-8 or utf-16
represenation by the latter, instead of the numeric character references you might prefer.
One thing that could be the case is that the output indeed contains utf-8 characters, while for
example your browser thinks it is getting iso-latin characters. Though the character numbers in the
output are actually correct, the characters are still not presented correctly.
You say you are using Cocoon. Are you sure that it is configured to use unicode or utf-8 as default
character set? Look for the string 'encoding' in WEB-INF/web.xml. It might be the case that your
stylesheets generate utf-8, but Cocoon is telling the browser that it will get iso-latin. This
confuses the browser and it depends on the browser what the result will be.
Hope this helps...
Grtz,
Geert