Bob
Jeff Greif wrote:
If this was all you had to use, you would be safe to invalidate B and A if C changed. A certainly depends on C, and B possibly. X,Y and Z cannot depend on C if C is not in the pool already. The tricky part is what to do if Y changes, since A, B and/or C might depend on it. I'm not sure there's any solution for that using this approach.
The dependency tree info would have to be maintained outside the grammar pool and consulted when a new insertion is made.
Wish it were that easy. After parsing the instance document, the grammar pool holds X,Y,Z,A,B,C. Now C is changed, and the resulting parse replaces C in the grammar pool. Which others should be invalidated? Multiple namespaces may be involved. It is not certain from this information there is any schema interdependency at all.
Bob
--------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
