Subject: RE: Schema aware transformations
From: "Michael Kay" <mike@xxxxxxxxxxxx>
Date: Thu, 23 Jun 2005 09:03:59 +0100
|
> Is it the case that a schema-aware processor processing an
> untyped input
> document behaves the same as a non-SA processor on that document?
Yes. A basic processor differs from a schema-aware processor only in that it
rejects certain things that an SA processor accepts. If you're within the
subset that a basic processor can handle, then the two give the same
results.
>
>
> > > -Is it possible to store the validated source (the PSVI?)
> > and supply
> > > that for subsequent transforms?
> >
> > Watch this space....
>
>
> I think that's the clincher at the moment.
>
I think the other important thing for many people will be the ability to
store a pre-compiled schema persistently. For that, you'll have to watch
this space a little longer. It's OK, of course, if you're in a
continuously-running server environment, where source documents, compiled
stylesheets, and compiled schemas can all be cached in memory today.
Did you know section: did you know that if you cache source documents and
stylesheets in memory, then Saxon will also cache the indexes used to
support xsl:key, which will stay in memory so long as both the source
document and the stylesheet are in memory? (A long-standing item on my to-do
list is to use a common index if two stylesheets define the same key
definition!)
Michael Kay
http://www.saxonica.com/
|