def extractRelevantResources(graph):
resources = []
it = graph.listStatements(
Soprano.ResourceNode(),
Soprano.ResourceNode(Soprano.Vocabulary.RDF.type()),
Soprano.ResourceNode())
while it.next():
s = it.current()
if not s.context().isValid() or not graph.containsStatement(
s.context(), Soprano.ResourceNode(
Soprano.Vocabulary.RDF.type()),
Soprano.ResourceNode(Soprano.Vocabulary.NRL.GraphMetadata())):
resources.append(s.subject())
return resources
