Hi Boris,

Thanks for the analysis.

Just to check my understanding of what you are saying...  You're saying that 
MFA.5.CONTENT could add attributes, but could NOT possibly add child 
elements because the base type would suck them all up.

So if we have the following...

 <xsd:complexType name="varies">
  <xsd:complexContent>
    <xsd:extension base="xsd:anyType"/>
   </xsd:complexContent>
  </xsd:complexType>

   <xsd:complexType name="myType">
    <xsd:complexContent>
       <xsd:extension base="varies">
         <xsd:attribute name="foo" type="xsd:int"/>
       </xsd:extension>
     </xsd:complexContent>
   </xsd:complexType>

  <xsd:element name="myElement" type="myType"/>

Then a valid instance would be:

<myElement anotherAttribute="x" foo="y">bar</myElement>

but anotherAttribute and bar would just be skipped over.

Does that seem right?

Pete.
---- Original Message ----- 
From: "Boris Kolpackov" <boris@c...>
To: <xml-dev@l...>
Sent: Thursday, June 08, 2006 8:10 AM
Subject: [xml-dev] Re: HL7 use of xs:anyType extension


> Hi Pete,
>
> "Pete Cordell" <petexmldev@t...> writes:
>
>>  <xsd:complexType name="varies">
>>   <xsd:complexContent>
>>    <xsd:extension base="xsd:anyType"/>
>>   </xsd:complexContent>
>>  </xsd:complexType>
>>
>> I've seen (incorrect) comments on the web that say this is an attempt to
>> define an empty element.  Is this HL7's intent?  Or is it just an attempt 
>> to
>> map the name 'varies' to xsd:anyType?
>
> I believe varies is just an alias for anyType.
>
>
>> The above type is then used in the following situation...
>>
>>   <xsd:complexType name="MFA.5.CONTENT">
>>     <xsd:complexContent>
>>       <xsd:extension base="varies">
>>         <xsd:attributeGroup ref="MFA.5.ATTRIBUTES"/>
>>       </xsd:extension>
>>     </xsd:complexContent>
>>   </xsd:complexType>
>>
>> What sorts of XML instances should this map to?  An empty element with
>> attributes as defined in the attributeGroup?  Or, any number of 
>> attributes
>> and element body content, that must also include the set of attributes
>> defined in the attributeGroup?  i.e. is this some way of defining an
>> extensible instance?
>
> I think this definition does not make any sense. anyType matches any
> content. If we agree that, in case of an extension, a base type is
> matched before derived (which appears to be the case from the spec)
> then base anyType will match everything and not leave anything to the
> extension.
>
> So I would say that if MFA.5.ATTRIBUTES defines any required attributes,
> then the set of valid contents for this type is empty. Otherwise any
> content matches with attributes from MFA.5.ATTRIBUTES never getting any
> value.
>
> hth,
> -boris
>
>
> -- 
> Boris Kolpackov
> Code Synthesis Tools CC
> http://www.codesynthesis.com
> Open-Source, Cross-Platform C++ XML Data Binding
>
>
> -----------------------------------------------------------------
> The xml-dev list is sponsored by XML.org <http://www.xml.org>, an
> initiative of OASIS <http://www.oasis-open.org>
>
> The list archives are at http://lists.xml.org/archives/xml-dev/
>
> To subscribe or unsubscribe from this list use the subscription
> manager: <http://www.oasis-open.org/mlmanage/index.php>
>
>