Lists Home |
Date Index |
K. Ari Krupnikov scripsit:
> Do you mean internal XML 1.0 DTD or internal Part 9 DTDs?
Let me clarify by inventing some new terminology:
The *native DTD* consists of the internal subset, if any, plus the
external subset, if any.
The *validation DTD* consists of declarations which are not in the
internal or the external subset, but are applied to the document by
a validation program such as Sun MSV.
> If you meant processing XML 1.0 DOCTYPE declarations, I don't see a
> reason to. W3C Schema needs to process DOCTYPEs because there are
> features it doesn't (yet) replicate that are available in DTDs
> (entities), but this language will have all of DTDs' features in it.
As I see it, though, validators return primarily a valid/invalid
indication, and perhaps a PSVI. They don't affect parsing as such.
Thus, unparsed entity declarations can appear in the validation DTD,
but they are only used to validate the content of an attribute of type
ENTITY or ENTITIES. Parsed general entity declarations are quite useless
in the validation DTD, because they cannot affect any entity references
in the original document: those must have already been resolved by the
parser before the validator gets control.
Currently Sun MSV accepts pure XML DTDs as validation DTDs. What I
am proposing is some extensions to validation DTDs without in any
way affecting native DTDs. Validation DTDs of course only exist in
John Cowan <email@example.com> http://www.reutershealth.com
I amar prestar aen, han mathon ne nen, http://www.ccil.org/~cowan
han mathon ne chae, a han noston ne 'wilith. --Galadriel, _LOTR:FOTR_