SHACL rule for dcat:accessURL

[josegar74]

I'm using the validator https://www.itb.ec.europa.eu/shacl/dcat-ap/upload selecting DCAT-AP 3.0.0 Full (individual constraints) to validate an RDF/XML file with the following distribution:

<dcat:distribution>
  <dcat:Distribution>
    <dcat:accessURL rdf:resource="http://mylink.com"/>
  </dcat:Distribution>
</dcat:distribution>

I get the following error.

I am using https://github.com/SEMICeu/iso-19139-to-dcat-ap/blob/main/iso-19139-to-dcat-ap.xsl to produce DCAT-AP in RDF/XML from ISO19139 documents. I am not an expert in DCAT-AP and I am not sure if it is problem in dcat:accessURL format that creates the above process in https://github.com/SEMICeu/iso-19139-to-dcat-ap/blob/abeb023837e642d2cfe62214f6e2aef9f1ef3599/iso-19139-to-dcat-ap.xsl#L1441 or if it is a problem in the validation rules.

---

Tested also with Turtle format and get the same issue with the same validation: DCAT-AP 3.0.0 Full (individual constraints)

@prefix dcat: <http://www.w3.org/ns/dcat#> .
<http://mylink.com> a dcat:Distribution;
   dcat:accessURL <http://mylink.com> ;

[EmidioStani]

Dear @josegar74 ,

accordigly to DCAT-AP 3.0.0 the access url should be a rdfs:Resource.

However, you can safely ignore the message, as the link could be defined somewhere else.

[josegar74]

@EmidioStani thanks for the quick reply. The problem with ignoring the message is that my client will not be happy to see the message as an error.

This is probably a very DCAT dummy question, but how should it be encoded the element in RDF/XML as an rdfs:Resource? Do you have any examples?

[bertvannuffelen]

@josegar74 I tested the validator and it seems that there is some config issue. I tried it with the shacls directly and then you get:

[josegar74]

@EmidioStani thanks, I was testing using the option DCAT-AP 3.0.0 Full (individual constraints). I will try with the SHACL files directly, maybe the validator rules are not updated.

[bertvannuffelen]

@josegar74 , this is bug in that file.

The issue is that the SHACL generator creates correctly a test for the range (simple export).

However in this case the range is rdfs:Resource.
This is in RDF validation not to be checked with a "be of that class". It is actually superseeded by the nodeKind test.
We have to fix the generator to that case in the tool into account.

For now, we will update the shapes in the draft and update the links in the configuration of the test bed.

[josegar74]

Thanks a lot for all the support.