[Expat-discuss] Charset trouble

[Karl Waclawek]

> While parsing XML, expat always converts entity coded chars ("&#code;") to unicode, 
> regardless of the charset I specified in XML_CreateParser (which is ISO-8859-1). 

That specifies just the input encoding.
Expat will always output Unicode (UTF-8 or UTF-16).

> So my application always shows every coded char in XML as a two-char string.
> Does the parser always encode to unicode? How can I solve it without compiling in unicode?
> Must I use a unicode-to-ansi conversion function and pass every returned string to it?

I don't think the XML specs allow anything else but Unicode for output.
The best approach would be to make your whole application use Unicode
internally (and maybe even externally).

Karl