[ome-users] STD.xsd - is this a bug?

[Josiah Johnston]

Good eye. You seem to be fairly well versed in the intricacies of XML  
schema. The FieldID stuff is left over from some ancient and incomplete  
deprications of Field.
I opened and closed a Bug on this:  
http://bugs.openmicroscopy.org.uk/show_bug.cgi?id=545

BEGIN_QUOTE

Removed dangling references to FieldID. Regarding the unique constraint  
on
SemanticElement/@Name, that constraint should be scoped to a  
SemanticType. e.g.
It is perfectly valid for two SemanticType's defined in the same  
document to
each contain a SemanticElement named "Description".
If anyone cares to submit a patch and a test case for the SE unique  
contraint,
by all means do.

Affected files:
OME/src/xml/schemas/STD.xsd v 1.10
http://cvs.openmicroscopy.org.uk/horde/chora/browse.php? 
f=OME%2Fsrc%2Fxml%2Fschemas%2FSTD.xsd

END_QUOTE

-Josiah

On Jul 22, 2005, at 10:32 AM, Graham Klyne wrote:

> I'm looking at the following fragments in STD.xsd, and can't make  
> sense of them:
>
> [[
>                 <unique name = "SemanticElement">
>                         <selector xpath = "STD:SemanticElement"/>
>                         <field xpath = "@Name"/>
>                 </unique>
> ]]
>
> [[
>                 <keyref name = "FieldIDref" refer = "STD:FieldID">
>                         <selector xpath = ".//STD:SemanticElement"/>
>                         <field xpath = "@FieldID"/>
>                 </keyref>
> ]]
>
> In particular, by my understanding, the selector child elements match  
> (select) elements named STD:SemanticElement in an instance document.   
> But the schema does not, as I read it, allow for any such elements.   
> The only definition of name SemanticElement in the schema is as the  
> name of the <unique> element quoted above.  I don't see how this can  
> legitimately be the target of an xpath expression.
>
> I'm not expert in the finer points of XML schema incantations, so I  
> could be missing something here,