I would like to extend xerces-j to provide seemless persistence. That is be able to create a DOM document and have its parsed info-set stored on a filesystem. There is a DOM implementation that does this already called PDOM but it doesn't implement DOM level 2 and therefore cannot be used with the XPathAPI. Given this goal I have a number of questions: 1) Is there an easy way to intercept calls to traverse a DOM object. I could extend the implementation of the xerces-j DOM (although I'm not sure which class this is!) but then my class would be hardwired to use this DOM implementation. I would like a way to still hook into the system configured DOM implementation. I guess I could implement the Document and Element interfaces and delegate work to the DOM implementation. Any better ideas? 2) Is there a simple way to persist the parsed info-set from Xerces-j? Clearly, I could persist the textual representation of the XML but this would mean reparsing to access the document off the disk. 3) How does XNI relate to intercepting calls to Xerces-j? Could I write my persistence extensions as some sort of installable filter? Regards, Enterprise Critical Systems Ltd EMail: [EMAIL PROTECTED] --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
