Current status: We are *not* currently applying pruning to source documents. Logic to analyse a stylesheet, find opportunities for pruning, and apply those insights is possible but nontrivial, and we simply haven't had time to pursue it.
"Tail-pruning" (removing the nodes most recently added to a DTM) is working, and is in use as part of our "shared DTM" mechanism for Temporary Trees/Result Tree Fragments. Pruning subtrees from the middle of a DTM is *not* currently implemented; that's considerably harder given our current data structures, though we made some changes to prepare for it. (I sketched out one approach to that code some time ago, but shelved it due to other demands on my time and concerns about its likely performance.) Some of the proposed changes to our use of DTMs may make pruning a bit easier to implement. General reminder: Xalan is open-source. If there's a feature you need that we aren't spending enough time on, that may be an opportunity to contribute... ______________________________________ Joe Kesselman / IBM Research
