Newbie: Normalisation and PSVI

We may define element values as string or token, and more if using datatype
libraries. And when using token for elements, the element values will be
normalised.

So after parsing a document with relax ng, I would expect to get a modified
infoset with element values normalised where the datatype is token. Isn't
this a PSVI (Post Schema Validation Infoset)?

Now I had the impression (maybe misunderstanding something in Eric van der
Vlist's book?) that relax ng parsing does not produce a PSVI. Or is it just
that relax ng does not put anything extra into the PSVI?

Could someone please point out where I'm wrong? ;-)

-tor
-------
Tor and Anne-Hilde Helland, Einerbakken 18, N-2827  HUNNDALEN, Norway
[email protected], [email protected]
http://www.helland.org/
+47 6117 7369


HVORFOR?

Min venn,
hvorfor skulle du
la deg slå ned i støvlene?
Støvler er ikke noe blivende sted!

Hvorfor ikke heller
la deg presse opp i hatten,
fyke rundt i stormkastene
for så å lande på
et fredelig sted,
fullt av hestehov og blåklokker?

Hvorfor ikke?