Thanks to everyone who replied to this message.

After giving the problem a bit more thought it occured to me that allowing the XML parser to ignore errors and to continue processing makes no more sense than allowing the perl intepreter to continue when it finds a syntax error. Moreover, allowing the XML parser to continue would lead to many of the same problems that we currently have with HTML. Permissive HTML parsers such as the one used by IE that will allow improperly nested tags, incomplete documents, unclosed tags, and so on lead to HTML designers to create that are usable only by the broken parser and foster bad programming/design habits. I would hate to see that happen with XML so I will not contribute to the problem by either writting an XML::Preprocessor module or adding this functionality to any sort of production system I create.

As far as the solution to my problem goes, I wrote a small filter to take care of the & characters before passing the XML doc to the parser. I decided to ignore the bare > and < characters since their presence may indicate either a problem with tags in the document or a legitimate part of the document text.

Once again, thanks for the insight.

----
Coyote