Hi,
I apologize in advance for this question, which is most likely about XML
Schema and not about Xerces; however maybe somebody out there has
already encountered this issue.
I'm writing my first moderately complicated XML Schema; I have much more
experience with dtd's. If I were writing a dtd for this application, it
would include something like
<!ELEMENT anElt ...>
<!ATTLIST anElt name ID #REQUIRED .. >
<!ELEMENT anotherElt ...>
<!ATTLIST anotherElt refs IDREFS #REQUIRED .. >
however, here I prefer to use XML Schema key and keyref because of the
ability to limit scope. I've tried something like this, where xs is the
prefix for XML schema namespace:
<xs:element name="anElt" ...>
<xs:attribute name="name" type="xs:NMTOKEN" use="required" />
...
</xs:element>
<xs:element name="anotherElt" ..>
<xs:attribute name="refs" type="xs:NMTOKENS: use="required" />
..
</xs:element>
and then
<xs:key name="anEltKey">
<xs:selector xpath="anElt" />
<xs:field xpath="@name" />
</xs:key>
<xs:keyref name="anotherRef" refer="anEltKey" >
<xs:selector xpath="anotherElt" />
<xs:field xpath="@refs" />
</xs:keyref>
This works as long as the value of the anotherRef attribute in the
instance document is just a single token. If it's more than one, Xerces
complains with an identity constraint error. Is there some way (in the
2nd xs:field element above) to express the idea of selecting each token
within the attribute string? I haven't found anything like this in the
XPATH spec. Or is there some other concise way to get the effect of
ID/IDREFS plus the nice scoping?
Thanks,
Joanne
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
