W3C home > Mailing lists > Public > www-xml-schema-comments@w3.org > July to September 2005

Re: Non-primitive built-in simple types

From: Dave Peterson <davep@iit.edu>
Date: Tue, 26 Jul 2005 16:20:38 -0400
Message-Id: <a06230907bf0c49300db6@[192.168.0.2]>
To: Paul.V.Biron@kp.org, Alain.Frisch@inria.fr
Cc: www-xml-schema-comments@w3.org

At 11:57 AM -0700 050726, Paul.V.Biron@kp.org wrote:

>Actually, there is a pattern facet which implements "disallowing the
>trailing decimal point".  Here is the type definition as it exists in the
>schema-for-schemas:
>
><xs:simpleType name="integer" id="integer">
>    <xs:restriction base="xs:decimal">
>       <xs:fractionDigits value="0" fixed="true"
>id="integer.fractionDigits"/>
>       <xs:pattern value="[\-+]?[0-9]+"/>
>    </xs:restriction>
></xs:simpleType>
>
>This pattern facet wasn't there in the first edition and, hence, there was
>some "magic" with integer...but this was cleaned up with the 2nd edition.

There is still "magic" associated with integer:  Canonical representations
of integers are not the canonical representations of the same numbers as
decimals; a no-magic derivation should not change canonical representations.

To the best of my knowledge this is the only "magic" remaining in the
process of "deriving" the built-in "derived" datatypes.  (I'm quoting
variants of 'derive' because nlately we are being more careful of how
the words are used in the text, and 'derive' will probably not always be
the best term to use.  I'm using it here in its 1.0--less careful--sense.)

I have hopes that even this "magic" will be removed in 1.1.
-- 
Dave Peterson
SGMLWorks!

davep@iit.edu
Received on Tuesday, 26 July 2005 20:20:45 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Sunday, 6 December 2009 18:13:08 GMT