/ Dave Pawson <[EMAIL PROTECTED]> was heard to say: | If Mike Kay can't locate them, I'd be surprised if Xerces can. | I often found its the last bit of text I inserted into the file,
Oh, I dunno. I've given up on the default parser in Saxon (AElfred). I've just stumbled over bugs in it too many times. I recommend Xerces. And a proper Java VM so that you can run Saxon instead of InstantSaxon if necessary :-) | In desperation I've often resorted to cutting out great swathes of text | to try and localise it in a big file. Try running your sources through a simple parser, like the xread app in the resolver distribution. That should choke on the bad encoding and provide an approximate location. I think. Should, anyway. Be seeing you, norm -- Norman Walsh <[EMAIL PROTECTED]> | Nothing will ever be attempted, if http://www.oasis-open.org/docbook/ | all possible objections must be Chair, DocBook Technical Committee | first overcome.--Dr. Johnson