Fix synonym type extraction in RDF/XML parser

Author: althonos

pronto/parsers/rdfxml.py

@@ -802,14 +802,27 @@ def _process_axiom(self, elem: etree.Element, curies: Dict[str, str]):
                 )
                 return
 
-            type_ = elem.find(_NS["oboInOwl"]["hasSynonymType"])
+            # extract synonym type name or IRI
+            elem_type = elem.find(_NS["oboInOwl"]["hasSynonymType"])
+            if elem_type is None:
+                type_ = None
+            elif _NS["rdf"]["resource"] in elem_type.attrib:
+                type_iri = elem_type.attrib[_NS["rdf"]["resource"]]
+                type_ = curies.get(type_iri) or self._compact_id(type_iri)
+            else:
+                type_ = elem_type.text
+
             try:
+                # recover existing synonym on the entity if there's one already
                 synonym = next(
                     s._data()
                     for s in entity.synonyms
                     if s.description == elem_target.text
                     and s.scope == _SYNONYMS[property]
                 )
+                # update synonym type of existing synonym if we got one
+                if type_ is not None:
+                    synonym.type = type_
             except StopIteration:
                 description = elem_target.get(_NS["rdf"]["resource"], elem_target.text)
                 if description is None:
@@ -822,7 +835,7 @@ def _process_axiom(self, elem: etree.Element, curies: Dict[str, str]):
                 synonym = SynonymData(
                     description,
                     scope=_SYNONYMS[property],
-                    type=type_.text if type_ is not None else None,
+                    type=type_,
                 )
 
             entity._data().synonyms.add(typing.cast(SynonymData, synonym))

tests/test_parser/test_rdfxml.py

@@ -449,3 +449,58 @@ def test_relationship_symmetric(self):
         )
         self.assertIn("TST:001", ont.relationships())
         self.assertTrue(ont.get_relationship("TST:001").symmetric)
+
+    def test_existing_synonym_type_extraction(self):
+        ont = self.get_ontology(
+            """
+            <owl:Ontology rdf:about="http://purl.obolibrary.org/obo/chebi.owl"/>
+
+            <owl:AnnotationProperty rdf:about="http://purl.obolibrary.org/obo/chebi#BRAND_NAME">
+                <rdfs:label rdf:datatype="http://www.w3.org/2001/XMLSchema#string">BRAND NAME</rdfs:label>
+                <rdfs:subPropertyOf rdf:resource="http://www.geneontology.org/formats/oboInOwl#SynonymTypeProperty"/>
+            </owl:AnnotationProperty>
+
+            <owl:Class rdf:about="http://purl.obolibrary.org/obo/CHEBI_4508">
+                <oboInOwl:hasRelatedSynonym rdf:datatype="http://www.w3.org/2001/XMLSchema#string">Cataflam</oboInOwl:hasRelatedSynonym>
+                <oboInOwl:id rdf:datatype="http://www.w3.org/2001/XMLSchema#string">CHEBI:4508</oboInOwl:id>
+            </owl:Class>
+            <owl:Axiom>
+                <owl:annotatedSource rdf:resource="http://purl.obolibrary.org/obo/CHEBI_4508"/>
+                <owl:annotatedProperty rdf:resource="http://www.geneontology.org/formats/oboInOwl#hasRelatedSynonym"/>
+                <owl:annotatedTarget rdf:datatype="http://www.w3.org/2001/XMLSchema#string">Cataflam</owl:annotatedTarget>
+                <oboInOwl:hasDbXref rdf:datatype="http://www.w3.org/2001/XMLSchema#string">DrugBank</oboInOwl:hasDbXref>
+                <oboInOwl:hasSynonymType rdf:resource="http://purl.obolibrary.org/obo/chebi#BRAND_NAME"/>
+            </owl:Axiom>
+            """
+        )
+        syntype, = ont.metadata.synonymtypedefs
+        synonym, = ont["CHEBI:4508"].synonyms
+        self.assertEqual(synonym.scope, "RELATED")
+        self.assertEqual(synonym.type, syntype)
+
+    def test_missing_synonym_type_extraction(self):
+        ont = self.get_ontology(
+            """
+            <owl:Ontology rdf:about="http://purl.obolibrary.org/obo/chebi.owl"/>
+
+            <owl:AnnotationProperty rdf:about="http://purl.obolibrary.org/obo/chebi#BRAND_NAME">
+                <rdfs:label rdf:datatype="http://www.w3.org/2001/XMLSchema#string">BRAND NAME</rdfs:label>
+                <rdfs:subPropertyOf rdf:resource="http://www.geneontology.org/formats/oboInOwl#SynonymTypeProperty"/>
+            </owl:AnnotationProperty>
+
+            <owl:Class rdf:about="http://purl.obolibrary.org/obo/CHEBI_4508">
+                <oboInOwl:id rdf:datatype="http://www.w3.org/2001/XMLSchema#string">CHEBI:4508</oboInOwl:id>
+            </owl:Class>
+            <owl:Axiom>
+                <owl:annotatedSource rdf:resource="http://purl.obolibrary.org/obo/CHEBI_4508"/>
+                <owl:annotatedProperty rdf:resource="http://www.geneontology.org/formats/oboInOwl#hasRelatedSynonym"/>
+                <owl:annotatedTarget rdf:datatype="http://www.w3.org/2001/XMLSchema#string">Cataflam</owl:annotatedTarget>
+                <oboInOwl:hasDbXref rdf:datatype="http://www.w3.org/2001/XMLSchema#string">DrugBank</oboInOwl:hasDbXref>
+                <oboInOwl:hasSynonymType rdf:resource="http://purl.obolibrary.org/obo/chebi#BRAND_NAME"/>
+            </owl:Axiom>
+            """
+        )
+        syntype, = ont.metadata.synonymtypedefs
+        synonym, = ont["CHEBI:4508"].synonyms
+        self.assertEqual(synonym.scope, "RELATED")
+        self.assertEqual(synonym.type, syntype)