Pyshacl inference

[richarddi...@gmail.com]

A new conversation because I can not add a new to the previous.

@Boris, Yes, sometimes even inference within TB Composer is complex.

So I hoped pyshacl is an alternative. But a minimal running example does not give inferred triples; sorry to be a hard learner. Here I try to infer keyWords for schema:Avatar.

Python code

---------------------------------------------------------------------

from rdflib import Graph, URIRef

from pyshacl import validate

shapes = Graph()

shapes.parse('testshapes.ttl', format='ttl')

data_graph = Graph()

data_graph.parse('testdata.ttl', format='ttl')

data_graph_orig = Graph()

for t in data_graph:

    data_graph_orig.add(t)

r = validate(data_graph,

      shacl_graph=shapes,

      #ont_graph=og,

      inference='rdfs',

      abort_on_error=False,

      meta_shacl=True,

      advanced=True,

      js=False,

      debug=True)

conforms, results_graph, results_text = r

added_triples = data_graph - data_graph_orig

print('Added triples: ', len(added_triples))

print(added_triples.all_nodes())

Data

-----------------------------------------------------

@prefix schema: <http://schema.org/> .

@prefix owl: <http://www.w3.org/2002/07/owl#> .

@prefix : <http://example.org#>.

:movies a owl:Ontology ;

    owl:imports schema: .

:Avatar a schema:Movie ;

    schema:actor :Sam .

Shapes:

------------------------------------------------------

@prefix sh: <http://www.w3.org/ns/shacl#>.

@prefix owl: <http://www.w3.org/2002/07/owl#> .

@prefix schema: <http://schema.org/> .

@prefix xsd: <http://www.w3.org/2001/XMLSchema#>.

@prefix : <http://example.org#>.

:movieOntology

    a owl:Ontology ;

    owl:imports schema: ;

    owl:imports sh: ;

    owl:imports xsd: ;

    sh:declare

    [ sh:prefix "sh" ;

      sh:namespace "http://www.w3.org/ns/shacl#"^^xsd:anyURI ; ] ,

    [ sh:prefix "schema" ;

      sh:namespace "http://schema.org/"^^xsd:anyURI ;] ,

    [ sh:prefix "owl" ;

      sh:namespace "http://www.w3.org/2002/07/owl#"^^xsd:anyURI ; ]

    .

:keywordsRule

    a sh:NodeShape ;

    sh:targetClass schema:Movie ;

    sh:rule [

        a sh:SPARQLRule ;

        sh:prefixes schema: ;

        sh:construct """

            CONSTRUCT {

                $this schema:keyWords ?actor  .

            }

            WHERE {

                $this schema:actor ?actor .

            }

            """ ;

    ] ;

.

Regards,

Richard

[Boris Pelakh]

When I run your code, among the several triples added I see

(rdflib.term.URIRef('http://example.org#Avatar'), rdflib.term.URIRef('http://schema.org/keyWords'), rdflib.term.URIRef('http://example.org#Sam'))

If you turn off inference, it is the only triple added. So perhaps some sort of an environment issue.

--
http://github.com/RDFLib
---
You received this message because you are subscribed to the Google Groups "rdflib-dev" group.
To unsubscribe from this group and stop receiving emails from it, send an email to rdflib-dev+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/rdflib-dev/15aef451-28f3-4501-872a-9ae2ca5193d2n%40googlegroups.com.

[richarddi...@gmail.com]

Thank you Boris. I'll have a look.

[richarddi...@gmail.com]

Without inference during validation the triple is inferred in my environment too. Thank you.