I'm trying to parse xml files using IXMLDocument/IXMLElement interfaces. It worked fine until I tried to parse a xml document with non-English characters.

One of CDATA elements in my xml document has non-English characters. For some reason, these characters were converted to other chars and question marks when I used IXMLElement::get_text() to get the element's text... Any ideas why it happenned? The xml file had encoding="ISO-8859-1" in the header, I tried UTF-8 (after encoding those chars in UTF-8) but it didn't help.

How can I tell this xml parser not make such conversions?

Tony