RDF identifier qualifier

[ChristineHaugen-aut]

cimxml specification is allowing for the following RDF qualifiers:
Version 1: Underscore qualifier (default)

• rdf:ID="_"
• rdf:about="_"
• rdf:resource="#_"

Version 2: URN qualifier

• rdf:about="urn:uuid:"
• rdf:resource="urn:uuid:"

Version 3: namespace qualifier

• rdf:about=":"
• rdf:resource=":"

CIMXMLSerializer should use the Underscore as the default (rdf;ID shall only be used in a defined subset of profiles). If should be possible to use "urn" and "namespace" as settings. Det forventes at namespaces ligger i settings eller tilsvarende.

Trig and JSON-LD serialisation shall only support URN and namespace qualifier with URN as default. The use of namespace setting can apply to all the serialisation.

[ChristineHaugen-aut]

@Sveino
When it comes to version 3, is any namespace allowed (which means the user will be able to input the namespace or prefix, though we can have a specific namespace as default)? Or is it always a specific namespace, which could be changed in the future? If it is the latter, it might be best not to allow the user to input the namespace.

Also, should the metadataheader still have urn:uuid in rdf:about regardless of qualifier type? I see that in the original cimxml files the metadata header has urn:uuid even when rdf:about (or id) for all the actual data has _.

[Sveino]

Yes, dataset header should always be rdf:about="urn:uuid:" regardless of settings.

The use should not be allowed to set the namespace, since it needs to be persistent between profiles. Clearly the namespace will be unique for a given model authority. For test you can use: https://model4powersystem.no/

My current working idea is that Statnett publish to EU with https://model4powersystem.eu/NO and this is redirected to a URL that is controlled by Statnett. Make it ready to be a configuration.