> I'm looking for a tip as to how best to use the XXE api to solve the > following problem. > > Problem: I have two XML trees that are expected to be very similar. One > is considered a baseline and the other is an update. I need to compare > the two trees and discover what differences there are, where the > expected results for any given node are as follows: > > 1. No change > 2. Node removed > 3. Node added > 4. Node modified > > The modification of a node will generally mean a change to the text > values of its leaf node children, but not changes of its branch node > children. It seems that the event model could be exploited somehow to > set client properties on nodes to indicate the type of change, but I'm > not quite sure how best to go about it. I've been experimenting with > document listeners. It seems a good solution would be to recursively > compare and change the baseline tree until it matched the update tree > and use a DocumentListener to record the changes, or perhaps use an > UndoManager. Am I on the right track? Any help is much appreciated.
You are trying to solve a difficult problem for which we have no clear answers yet. There are two ways to implement what you want: [1] Implement change tracking: that is, record somewhere the changes made by the author by using a DocumentListener or a specialization of UndoManager (which happens to implement DocumentListener). [2] Use an XML Diff algorithm such as the patented algorithm sold by DeltaXML. http://www.deltaxml.com/ We prefer [2]. If you find a published algorithm which works flawlessly in the case of XML documents containing prose, do not hesitate to tell us about this algorithm.
