I think you're stuck with trying to find, or write, a repair tool that operates on the file as text with an awareness of XML rather than operating on it as XML. One example of this approach is the "Tidy" tool available from the W3C's website. Tidy was written to read misformed HTML and make some reasonable guesses about how to convert into a properly structured file -- and if you're lucky, it might be the file that the user intended to write.
I don't know whether Tidy can be persuaded to run on XML. It might be worth investigating. If not, you might have to write your own repair tool, which has some knowledge of your DTD and what mistakes are likely. --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
