XSFD: character data

XSFD assumes that in an untyped forest character data has already been
turned into (instances of) atomic datatypes.  This doesn't seem satisfactory
to me.  The infoset has only character info items.  It is part of the job of
XSFD to explain how character info items get turned into atomic datatypes.
I believe in the untyped forest the leaves should be characters and that
normalization should turn these into atomic datatypes.

Received on Friday, 27 April 2001 02:25:56 UTC