Validating the domain (or ObjectProperty range) asserted in a data graph exists in the ontology graph?

[tduval-unifylogic]

Another question.

I would like to validate that the domain asserted in the data graph actually exists as a class specified in the ontology. I have successfully created validation for DatatypeProperty ranges:

    [ sh:path rdfs:range ; sh:nodetype sh:IRI ;
        sh:targetObjectsOf rdfs:range ;
        sh:in ( xsd:anyType xsd:anyURI xsd:base64Binary ...etc.  ) ;
    ] ,

but now i would like to be able to validate both domain and ranges existence for object properties.

[ajnelson-nist]

Is your application an OWL application? If so, domains and ranges can get significantly complex, e.g. for properties that have a domain that is all X's and all not-Y's would be a union of X with the complement of Y's. That could be hairy to write a catchall SHACL shape for. If this is just an RDFS application instead, nevermind this point.

Either a sh:qualifiedValueShape would do what you need, or this could:

ex:CheckRangeIsUsingDefinedClass-shape
    a sh:NodeShape ;
    sh:nodeKind sh:IRI ;
    sh:or (
        [ a sh:NodeShape ; sh:class rdfs:Class ; ]
        [ a sh:NodeShape ; sh:class owl:Class ; ]
    )
    sh:targetObjectsOf rdfs:domain ;
    .

I haven't tested that snippet, but I think it would work.

And again, this assumes the more simplified case of properties defined in the RDFS style without OWL - OWL would need some further contortion using an sh:xone to represent "Either the shape is a blank node (implying a OWL class expression instead of class (in which case, here come a lot of property paths to confirm declarations)), or it matches [the above shape]."

(Also, careful, there's a typo'd sh:nodetype in your example. You might want to try the SHACL-SHACL conformance testing from the SHACL spec; pyshacl provides this with --metashacl.)

[tduval-unifylogic]

sh:class owl:Class

works perfect for my needs!

Thanks again!

[ajnelson-nist]

You're welcome!

By the way, it's sounding like you're getting into "TBox" review with SHACL. You might want to be aware of this Issue.

[tduval-unifylogic]

ah! thanks fo that!

[ashleysommer]

@tduval-unifylogic
Is it OK to close this issue thread now?

[tduval-unifylogic]

Yes thanks for the follow up

…

On Wed, May 24, 2023 at 9:44 AM, Ashley Sommer ***@***.***> wrote: @tduval-unifylogic <https://github.com/tduval-unifylogic> Is it OK to close this issue thread now? — Reply to this email directly, view it on GitHub <#183 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AO3YXTBONPIHTOZXNK4OSFTXHYGEXANCNFSM6AAAAAAXTKN3RU> . You are receiving this because you were mentioned.Message ID: ***@***.***>