- From: Amelia A. Lewis <alewis@tibco.com>
- Date: Wed, 26 Mar 2003 15:40:41 -0500
- To: "Martin Gudgin" <mgudgin@microsoft.com>
- Cc: moreau@crf.canon.fr, jeffsch@windows.microsoft.com, roberto.chinnici@sun.com, sanjiva@us.ibm.com, www-archive@w3.org
Okay, a proposal, then:
Pseudo-schema:
<assert level="MUST|SHOULD|MAY" negative="boolean">%p.pcd.mix</assert>
Maybe %termdef.pcd.mix, but I don't think it gains much.
Schema fragment:
<xs:schema>
<xs:simpleType name="levelType">
<xs:restriction base="xs:string">
<xs:enumeration value="MUST" />
<xs:enumeration value="SHOULD" />
<xs:enumeration value="MAY" />
</xs:restriction>
</xs:simpleType>
<xs:element name="assert">
<xs:complexType>
<xs:complexContent>
<xs:sequence>
<xs:group ref="p.pcd.mix">
</xs:sequence>
<xs:attribute name="level" use="required" type="levelType" />
<xs:attribute name="negative use="optional" type="xs:boolean" default="false" />
</xs:complexContent>
</xs:complexType>
</xs:element>
</xs:schema>
I really don't want to define p.pcd.mix in schema. Oh, well.
Amy!
--
Amelia A. Lewis
Architect, TIBCO/Extensibility, Inc.
alewis@tibco.com
Received on Wednesday, 26 March 2003 15:40:23 UTC