Re: XML Schemas: Best Practices

[Martin Bryan <mtbryan@sgml.u-net.com>]

> Jeff Rafter wrote:
> >
>> I think that Martin's intent was to use a
> > restriction element instead of an extension element (correct me if I am
> > wrong).

Actually I was deliberately questioning whether or not we could use
extension in the same way as extension, by restating the requirements.
Reveiwing the specs I see we cannot do this directly, but what about;

     <complexType name="BookType">
        <all>
            <element name="Title" type="string"/>
            <element name="Author" type="string"/>
            <element name="Date" type="year"/>
            <element name="ISBN" type="string"/>
            <element name="Publisher" type="string"/>
        </all>
    </complexType>

   <complexType name="BookTypePlusReviewerType">
       <complexContent>
            <extension base="c:BookType" >
                <element name="Reviewer" type="string"/>
            </extension>
      </complexContent>
  </complexType>

<element name="BookReview" type="BookTypePlusReviewType">
       <complexType>
            <restriction base="c:BookType" >
                <sequence>
                    <element name="Reviewer" type="string"/>
                    <element name="Title" type="string"/>
                    <element name="Author" type="string"/>
                    <element name="Date" type="year"/>
                   <element name="ISBN" type="string"/>
                   <element name="Publisher" type="string"/>
                </sequence>
            </restriction>
        </complexType>
    </complexType>
</element>

The theory here is that the BookType does not define the sequence of the
elements, only that they are all required. Then you can add Reviewer to the
list of required elements to create a new type definition that extends the
unordered list. Then in the content type definition of the element it is
being applied to you sequence the list as required.

Is there any reason why this would not work? Isn't it safer than adding any
elements at start and end of each sequence, and between each member of the
sequence if you are to cover all eventualities?

Martin Bryan