>Content is not allowed in prolog.
Your source data is not being seen as well-formed XML. Specifically, the
parser things it is seeing non-whitespace text before the root element.
At a guess, the fact that you're going through ByteArrayInputStream rather
than StringReader sounds like a wonderful opportunity to mess up the UTF-8
versus UTF-16 distinction...
> Have you invoked the XALAN API with in-memory strings?
> How did you do it?
Source source=new StreamSource(new java.io.StringReader(my_string));
and then process from that Source.
