- From: Sean Morris <semorris@cisco.com>
- Date: Mon, 3 Jun 2002 10:29:27 -0400 (EDT)
- To: asirv@webmethods.com
- CC: noah_mendelsohn@us.ibm.com, xmlschema-dev@w3.org
- Message-ID: <3CFB7CED.4090204@cisco.com>
Asir, Thanks for the union idea. This accomplishes exactly what I want. The actual syntax for allowing 1-5 and 10-15 is <simpleType name="T1"> <union> <simpleType> <restriction base:"int"> <minInclusive="1"/> <maxInclusive="5"/> </restriction> </simpleType> <simpleType> <restriction base:"int"> <minInclusive="10"/> <maxInclusive="15"/> </restriction> </simpleType> </union> </simpletype> Thanks, Sean Asir S Vedamuthu wrote: >mm .. For some reasons, I thought that we have this capability in 1.0 But, >now I am confused. Here is what I had in my mind, > ><simpleType name="discreteInt"> > <union> > <simpleType> > <restriction base="int"> > <enumeration value="0"/> > </restriction> > </simpleType> > </union> > <union> > <simpleType> > <restriction base="int"> > <minInclusive value="5"/> > <maxInclusive value="7"/> > </restriction> > </simpleType> > </union> > <union> > <simpleType> > <restriction base="int"> > <minInclusive value="10"/> > <maxInclusive value="15"/> > </restriction> > </simpleType> > </union> ></simpleType> > >Asir > > >-----Original Message----- >From: xmlschema-dev-request@w3.org >[mailto:xmlschema-dev-request@w3.org]On Behalf Of >noah_mendelsohn@us.ibm.com >Sent: Friday, May 31, 2002 2:36 PM >To: Sean Morris (by way of "C. M. Sperberg-McQueen" ><<semorris@cisco.com<semorris>); xmlschema-dev@w3.org >Subject: Re: XML Schemas > > > > >Sean Morris suggests (on the schema comments list...I've moved here so we >can discuss): > >>>Add support for defining multiple ranges for integers. >>>For example an integer element can be 0, 5-7, 10-15, etc. >>> > >Sean: there are many such capabilities that we might consider adding to >what is already a complex spec. Each such feature is a bit more for users >to learn and understand, a bit more to model in components, and more for >which to write test cases. So, we usually try to make the case that a new >feature is either of great use to a broad range of users, or more rarely, >of very great importance to a smaller set of users. Do you have >implementation experience to suggest that multiple ranges meets an 80/20 or >90/10 cut? > >Also: we have to think about the rules for refinement. Can: > > Integer [5-10, 11-15] > > be refined to > > Integer [7-12]? > >Keep in mind that integer is a subtype of decimal and presumably: > > Decimal [5.0-10.0, 11.0-15.0] > > is not a refinement of > > Decimal [7.0-12.0]? (this allows 10.5 and the base type does not? > >Is the component representation of > > Integer [5-10, 11-15] > > distinct from > > Integer [5-15]? > >One thing we've all learned in doing schemas is that there is rarely such a >thing as a simple, cheap, or safe feature. Can you make the case that the >need for the feature request is sufficiently urgent that the moderate >additional complexity is justified? Many thanks. > >------------------------------------------------------------------ >Noah Mendelsohn Voice: 1-617-693-4036 >IBM Corporation Fax: 1-617-693-8676 >One Rogers Street >Cambridge, MA 02142 >------------------------------------------------------------------ > > > > > > > Sean Morris > <semorris@cisco.com> (by To: W3C XML >Schema Comments list <www-xml-schema-comments@w3.org> > way of "C. M. cc: (bcc: >Noah Mendelsohn/Cambridge/IBM) > Sperberg-McQueen" <cmsmcq Subject: XML >Schemas > Sent by: > www-xml-schema-comments-req > uest@w3.org > > > > 05/30/2002 08:38 PM > > > > > > > > > > >Add support for defining multiple ranges for integers. For example an >integer element can be 0, 5-7, 10-15, etc. > >Thanks, >Sean > > > > > > >
Received on Monday, 3 June 2002 17:20:30 UTC