This is an archived snapshot of W3C's public bugzilla bug tracker, decommissioned in April 2019. Please see the home page for more details.
Nothing in the REC stops me from having an enumeration facet whose value is a superset of its base type definition's, or a pattern facet whose value a subset of its base type definition's. This violates subsumption. The XML representations of such facets are disallowed, but nothing stops born-binary ones.
This only applies to "pattern". For enumeration, it's already covered by the "enumeration valid restriction" constraint in section 4.3.5.5 of part 2.
I propose to resolve this by noting, first, that Sandy Gao is correct to say in comment #2 that there is after all such a constraint for enumeration, in section 4.3.5.5 of Datatypes. To deal with pattern, I propose to add a section 4.3.4.5: 4.3.5.5 Constraints on pattern Schema Components Schema Component Constraint: valid restriction of pattern It is an ·error· if there is any member of the {value} of the pattern facet on the {base type definition} which is not also a member of the {value}. The Working Group has not yet taken action on this proposal.
On the teleconference of 7 September 2007, the XML Schema WG discussed this issue and decided to resolve it by adopting the proposal in comment #2. Additionally, the editors were instructed, if they can agree on wording, to add a note pointing out that for components described in schema documents, the satisfaction of this constraint is guaranteed by the XML mapping rules and that special attention thus need be paid only for 'born binary' components. The Working Group did not wish to review the wording on that non-normative note. I'm marking this needsDrafting until the editors have inserted the text of that note; once they have done so, and the text is in the status quo document, this issue can be marked resolved / fixed.
The correction mentioned in comment #2 has been made and appears in the current status-quo draft. The note described in comment #3 has been drafted and agreed on by the editors and will appear in the status quo document the next time it's re-genererated. So I'm marking this as FIXED. Henry, as the originator, if you would mark it CLOSED to indicate your agreement with the disposition, that would be good.