Thanks, Natalia, for your in depth answers. Understanding it would help me, in my application, to avoid using a deep xml tree.
Thanks again. Tak On Tue, Jul 14, 2009 at 2:47 PM, Natalia Shilenkova <nshilenk...@gmail.com>wrote: > On Mon, Jul 13, 2009 at 9:53 PM, Tak-po Li<takpo...@gmail.com> wrote: > > > > > > Thanks, Natalia, for your detail explanation. If I understand it right, > we > > use variable Var to point to the portion of the tree that we want to keep > > and build the new tree with new node B and Var. Question: Do we create a > > duplicate tree and delete the old one or just switch the pointers around? > > You are absolutely correct in your understanding of how the query > works. As for the question, this query will duplicate the part of the > document that is assigned to the variable. It does not seem to be > possible to move nodes in the document with XUpdate, at least I could > not figure it out. > > It is also possible to read document from database, then to modify it > using DOM, and finally to save it to database again. In this case you > can have greater flexibility and can do the same thing as XUpdate > query in question without copying parts of the document. However, I > doubt this method will be any better because it would mean serializing > document and transferring it between client and the server, while > XUpdate query modifies documents on the server. > > Regards, > Natalia > > > > > Thanks again. > > > > Tak > > > > >
