I am quite sure that we should use JDOM inthe editor. It is much easier to use than dom4j and should be more flexible, since it uses specifically Java constructs. When the editor is told to read a file, I think it should just attempt to parse it within a try block. If the file is not valid xml the parser will throw exceptions which the editor can catch. The type of the exception will help determine what to do with the file. If it is a plain-text file I really think it would be best to pass it through liblouisutdml, because this will produce a nice daisy file, all ready for editing or embossing. If the file parses successfully the root element will tell us what kind of xml document it is. If it is not a Daisy file, we might pass it through a converter using xslt. pdf, Word and other iles will need a different kind of converter. Laura has worked extensively with tika, and I have used it for some file conversions. We are quite underwhelmed. The quality of the conversions really isn't up to whaat BrailleBlaster needs. We are thinkinng of looking at other conversion libraries. A new version of tika is supposed to come out soon. This may fix some of the problems. John -- John J. Boyer; President, Chief Software Developer Abilitiessoft, Inc. http://www.abilitiessoft.com Madison, Wisconsin USA Developing software for people with disabilities