Ambiguity of the basePsoID within a search

["Andrew Sciberras" <andrew.sciberras@enitiatives.com.au>]

Hello

I believe that the SearchQueryType defined in pstc_spmlv2_search.xsd breaks
the XML Unique Particle Attribution rule. 

The <extension base="spml:ExtensibleType"> allows any element from a non
'spml' namespace to exist within the SearchQueryType. Additionally, the
"<element name="basePsoID" type="spml:PSOIdentifierType"  minOccurs="0" />"
allows the spmlsearch:basePsoID element to be within the SearchQueryType. 

This breaks the unique particle attribution rule because it is ambiguous as
to whether a <spmlsearch:basePsoID> element (within a search request) is
part of the extension base type (since it's namespace is not 'spml') or the
basePsoID element declaration. 

The DSMLv2 profile provides an example (line 276) which shows
"spml:basePsoID" as the first element in the spmlsearch:query, followed by
the DSML filter. 
Ignoring the fact that the basePsoID's namespace is wrong (it should be
spmlsearch rather an spml), this XML value could only be valid if the
basePsoID is associated with the 'spml:ExtensibleType" portion of the
SearchQueryType, as opposed to the explicit 'basePsoID' element declaration.



Could you please let me know if my interpretation of this ambiguity is
correct? Additionally, if I were to generate SPML search requests, would you
recommend placing a basePsoID as the final element in the <spmlsearch:query>
or as the initial element (as per the example in the dsml profile example)? 


I'm inclined to think that if it's provided as the first element, then only
systems that associate the basePsoID element as being part of the
ExtensibleType will consider it valid; whilst if it was the final element
then it should be considered valid regardless of whether the element is
perceived as being associated with the ExtensibleType or the explicit
element definition. 

Thanks
Andrew Sciberras