Merge 7-maintenance branch into 7.x (#3222)

edmondchuc · nicholascar · ddeschepper · web-flow · commit c580c00dbe76 · 2025-09-18T13:08:37.000+10:00
* 7.1.4 pre-release * fix namespace prefixes in longturtle serialization (#3134) Co-authored-by: Daan de Schepper <d.de.schepper@rijksmuseum.nl> * Fix failing webtest (#3194) * test: fix failing webtest Fixes #3192 * Revert "remove old hacks against 2to3 (#3076)" (#3195) This reverts commit b74c657. * Specify `Optional` parameters in `Graph.triples_choices` (#3075) * Specify `Optional` parameters in `Graph.triples_choices` The two non-list parameters can be `None`, but this is not reflected in the type hint. Also introduces a type alias to simplify method signatures. * style: remove unused imports --------- Co-authored-by: Nicholas Car <nick@kurrawong.net> Co-authored-by: Edmond Chuc <37032744+edmondchuc@users.noreply.github.com> Co-authored-by: Edmond Chuc <edmond@kurrawong.ai> * feat: canonicalization with longturtle serializer now optional (#3197) * feat: canonicalization with longturtle serializer now optional Fixes #3196 * docs: fix docs build error by removing py obj reference to canon --------- Co-authored-by: Nicholas Car <nick@kurrawong.ai> Co-authored-by: Daan de Schepper <d.de.schepper@rijksmuseum.nl> Co-authored-by: Sigmund Lahn <s@lahn.no> Co-authored-by: Nicholas Car <nick@kurrawong.net>
diff --git a/rdflib/graph.py b/rdflib/graph.py
@@ -355,6 +355,11 @@
 _TripleOrQuadSelectorType = Union["_TripleSelectorType", "_QuadSelectorType"]
 _TriplePathType = Tuple["_SubjectType", Path, "_ObjectType"]
 _TripleOrTriplePathType = Union["_TripleType", "_TriplePathType"]
+_TripleChoiceType = Union[
+    Tuple[List[_SubjectType], Optional[_PredicateType], Optional[_ObjectType]],
+    Tuple[Optional[_SubjectType], List[_PredicateType], Optional[_ObjectType]],
+    Tuple[Optional[_SubjectType], Optional[_PredicateType], List[_ObjectType]],
+]
 
 _GraphT = TypeVar("_GraphT", bound="Graph")
 _ConjunctiveGraphT = TypeVar("_ConjunctiveGraphT", bound="ConjunctiveGraph")
@@ -994,11 +999,7 @@ def predicate_objects(
 
     def triples_choices(
         self,
-        triple: Union[
-            Tuple[List[_SubjectType], _PredicateType, _ObjectType],
-            Tuple[_SubjectType, List[_PredicateType], _ObjectType],
-            Tuple[_SubjectType, _PredicateType, List[_ObjectType]],
-        ],
+        triple: _TripleChoiceType,
         context: Optional[_ContextType] = None,
     ) -> Generator[_TripleType, None, None]:
         subject, predicate, object_ = triple
@@ -2196,11 +2197,7 @@ def quads(
 
     def triples_choices(
         self,
-        triple: Union[
-            Tuple[List[_SubjectType], _PredicateType, _ObjectType],
-            Tuple[_SubjectType, List[_PredicateType], _ObjectType],
-            Tuple[_SubjectType, _PredicateType, List[_ObjectType]],
-        ],
+        triple: _TripleChoiceType,
         context: Optional[_ContextType] = None,
     ) -> Generator[_TripleType, None, None]:
         """Iterate over all the triples in the entire conjunctive graph"""
@@ -2946,11 +2943,7 @@ def __isub__(self: _GraphT, other: Iterable[_TripleType]) -> NoReturn:
 
     def triples_choices(
         self,
-        triple: Union[
-            Tuple[List[_SubjectType], _PredicateType, _ObjectType],
-            Tuple[_SubjectType, List[_PredicateType], _ObjectType],
-            Tuple[_SubjectType, _PredicateType, List[_ObjectType]],
-        ],
+        triple: _TripleChoiceType,
         context: Optional[_ContextType] = None,
     ) -> Generator[_TripleType, None, None]:
         subject, predicate, object_ = triple
diff --git a/rdflib/plugins/parsers/rdfxml.py b/rdflib/plugins/parsers/rdfxml.py
@@ -298,7 +298,8 @@ def document_element_start(
         self, name: Tuple[str, str], qname, attrs: AttributesImpl
     ) -> None:
         if name[0] and URIRef("".join(name)) == RDFVOC.RDF:
-            next = self.next
+            # Cheap hack so 2to3 doesn't turn it into __next__
+            next = getattr(self, "next")
             next.start = self.node_element_start
             next.end = self.node_element_end
         else:
@@ -315,7 +316,8 @@ def node_element_start(
         current = self.current
         absolutize = self.absolutize
 
-        next = self.next
+        # Cheap hack so 2to3 doesn't turn it into __next__
+        next = getattr(self, "next")
         next.start = self.property_element_start
         next.end = self.property_element_end
 
@@ -408,7 +410,8 @@ def property_element_start(
         current = self.current
         absolutize = self.absolutize
 
-        next = self.next
+        # Cheap hack so 2to3 doesn't turn it into __next__
+        next = getattr(self, "next")
         object: Optional[_ObjectType] = None
         current.data = None
         current.list = None
diff --git a/rdflib/plugins/serializers/longturtle.py b/rdflib/plugins/serializers/longturtle.py
@@ -39,21 +39,20 @@
 
 
 class LongTurtleSerializer(RecursiveSerializer):
+    """LongTurtle, a Turtle serialization format.
+
+    When the optional parameter ``canon`` is set to :py:obj:`True`, the graph is canonicalized
+    before serialization. This normalizes blank node identifiers and allows for
+    deterministic serialization of the graph. Useful when consistent outputs are required.
+    """
+
     short_name = "longturtle"
     indentString = "    "
 
     def __init__(self, store):
         self._ns_rewrite = {}
-        store = to_canonical_graph(store)
-        content = store.serialize(format="application/n-triples")
-        lines = content.split("\n")
-        lines.sort()
-        graph = Graph()
-        graph.parse(
-            data="\n".join(lines), format="application/n-triples", skolemize=True
-        )
-        graph = graph.de_skolemize()
-        super(LongTurtleSerializer, self).__init__(graph)
+        self._canon = False
+        super(LongTurtleSerializer, self).__init__(store)
         self.keywords = {RDF.type: "a"}
         self.reset()
         self.stream = None
@@ -83,11 +82,34 @@ def addNamespace(self, prefix, namespace):
         super(LongTurtleSerializer, self).addNamespace(prefix, namespace)
         return prefix
 
+    def canonize(self):
+        """Apply canonicalization to the store.
+
+        This normalizes blank node identifiers and allows for deterministic
+        serialization of the graph.
+        """
+        if not self._canon:
+            return
+
+        namespace_manager = self.store.namespace_manager
+        store = to_canonical_graph(self.store)
+        content = store.serialize(format="application/n-triples")
+        lines = content.split("\n")
+        lines.sort()
+        graph = Graph()
+        graph.parse(
+            data="\n".join(lines), format="application/n-triples", skolemize=True
+        )
+        graph = graph.de_skolemize()
+        graph.namespace_manager = namespace_manager
+        self.store = graph
+
     def reset(self):
         super(LongTurtleSerializer, self).reset()
         self._shortNames = {}
         self._started = False
         self._ns_rewrite = {}
+        self.canonize()
 
     def serialize(
         self,
@@ -97,6 +119,7 @@ def serialize(
         spacious: Optional[bool] = None,
         **kwargs: Any,
     ) -> None:
+        self._canon = kwargs.get("canon", False)
         self.reset()
         self.stream = stream
         # if base is given here, use, if not and a base is set for the graph use that
diff --git a/rdflib/plugins/stores/berkeleydb.py b/rdflib/plugins/stores/berkeleydb.py
@@ -428,7 +428,8 @@ def remove(  # type: ignore[override]
                 cursor = index.cursor(txn=txn)
                 try:
                     cursor.set_range(key)
-                    current = cursor.next
+                    # Hack to stop 2to3 converting this to next(cursor)
+                    current = getattr(cursor, "next")()
                 except db.DBNotFoundError:
                     current = None
                 cursor.close()
@@ -505,7 +506,8 @@ def triples(
             cursor = index.cursor(txn=txn)
             try:
                 cursor.set_range(key)
-                current = cursor.next
+                # Cheap hack so 2to3 doesn't convert to next(cursor)
+                current = getattr(cursor, "next")()
             except db.DBNotFoundError:
                 current = None
             cursor.close()
@@ -537,7 +539,8 @@ def __len__(self, context: Optional[_ContextType] = None) -> int:
             key, value = current
             if key.startswith(prefix):
                 count += 1
-                current = cursor.next
+                # Hack to stop 2to3 converting this to next(cursor)
+                current = getattr(cursor, "next")()
             else:
                 break
         cursor.close()
@@ -590,7 +593,8 @@ def namespaces(self) -> Generator[Tuple[str, URIRef], None, None]:
         while current:
             prefix, namespace = current
             results.append((prefix.decode("utf-8"), namespace.decode("utf-8")))
-            current = cursor.next
+            # Hack to stop 2to3 converting this to next(cursor)
+            current = getattr(cursor, "next")()
         cursor.close()
         for prefix, namespace in results:
             yield prefix, URIRef(namespace)
@@ -633,7 +637,8 @@ def contexts(
                 cursor = index.cursor()
                 try:
                     cursor.set_range(key)
-                    current = cursor.next
+                    # Hack to stop 2to3 converting this to next(cursor)
+                    current = getattr(cursor, "next")()
                 except db.DBNotFoundError:
                     current = None
                 cursor.close()
diff --git a/rdflib/plugins/stores/sparqlstore.py b/rdflib/plugins/stores/sparqlstore.py
@@ -35,6 +35,7 @@
         _TripleType,
         _ContextType,
         _QuadType,
+        _TripleChoiceType,
         _TriplePatternType,
         _SubjectType,
         _PredicateType,
@@ -367,11 +368,7 @@ def triples(  # type: ignore[override]
 
     def triples_choices(
         self,
-        _: Tuple[
-            Union[_SubjectType, List[_SubjectType]],
-            Union[_PredicateType, List[_PredicateType]],
-            Union[_ObjectType, List[_ObjectType]],
-        ],
+        _: _TripleChoiceType,
         context: Optional[_ContextType] = None,
     ) -> Generator[
         Tuple[
diff --git a/rdflib/store.py b/rdflib/store.py
@@ -36,7 +36,6 @@
     Generator,
     Iterable,
     Iterator,
-    List,
     Mapping,
     Optional,
     Tuple,
@@ -49,10 +48,8 @@
     from rdflib.graph import (
         Graph,
         _ContextType,
-        _ObjectType,
-        _PredicateType,
         _QuadType,
-        _SubjectType,
+        _TripleChoiceType,
         _TriplePatternType,
         _TripleType,
     )
@@ -281,11 +278,7 @@ def remove(
 
     def triples_choices(
         self,
-        triple: Union[
-            Tuple[List[_SubjectType], _PredicateType, _ObjectType],
-            Tuple[_SubjectType, List[_PredicateType], _ObjectType],
-            Tuple[_SubjectType, _PredicateType, List[_ObjectType]],
-        ],
+        triple: _TripleChoiceType,
         context: Optional[_ContextType] = None,
     ) -> Generator[
         Tuple[
diff --git a/test/data/longturtle/longturtle-target.ttl b/test/data/longturtle/longturtle-target.ttl
@@ -1,72 +1,74 @@
+PREFIX cn: <https://linked.data.gov.au/def/cn/>
+PREFIX ex: <http://example.com/>
 PREFIX geo: <http://www.opengis.net/ont/geosparql#>
 PREFIX rdf: <http://www.w3.org/1999/02/22-rdf-syntax-ns#>
-PREFIX schema: <https://schema.org/>
+PREFIX sdo: <https://schema.org/>
 PREFIX xsd: <http://www.w3.org/2001/XMLSchema#>
 
-<http://example.com/nicholas>
-    a schema:Person ;
-    schema:age 41 ;
-    schema:alternateName 
+ex:nicholas
+    a sdo:Person ;
+    sdo:age 41 ;
+    sdo:alternateName 
         [
-            schema:name "Dr N.J. Car" ;
+            sdo:name "Dr N.J. Car" ;
         ] ,
         "N.J. Car" ,
         "Nick Car" ;
-    schema:name
+    sdo:name
         [
-            a <https://linked.data.gov.au/def/cn/CompoundName> ;
-            schema:hasPart 
+            a cn:CompoundName ;
+            sdo:hasPart 
                 [
-                    a <https://linked.data.gov.au/def/cn/CompoundName> ;
-                    schema:hasPart 
+                    a cn:CompoundName ;
+                    sdo:hasPart 
                         [
-                            a <https://linked.data.gov.au/def/cn/CompoundName> ;
+                            a cn:CompoundName ;
                             rdf:value "Car" ;
                         ] ,
                         [
-                            a <https://linked.data.gov.au/def/cn/CompoundName> ;
+                            a cn:CompoundName ;
                             rdf:value "Maxov" ;
                         ] ;
                 ] ,
                 [
-                    a <https://linked.data.gov.au/def/cn/CompoundName> ;
+                    a cn:CompoundName ;
                     rdf:value "Nicholas" ;
                 ] ,
                 [
-                    a <https://linked.data.gov.au/def/cn/CompoundName> ;
+                    a cn:CompoundName ;
                     rdf:value "John" ;
                 ] ;
         ] ;
-    schema:worksFor <https://kurrawong.ai> ;
+    sdo:worksFor <https://kurrawong.ai> ;
 .
 
 <https://kurrawong.ai>
-    a schema:Organization ;
-    schema:location <https://kurrawong.ai/hq> ;
+    a sdo:Organization ;
+    sdo:location <https://kurrawong.ai/hq> ;
 .
 
 <https://kurrawong.ai/hq>
-    a schema:Place ;
-    schema:address
+    a sdo:Place ;
+    sdo:address
         [
-            a schema:PostalAddress ;
-            schema:addressCountry
+            a sdo:PostalAddress ;
+            sdo:addressCountry
                 [
-                    schema:identifier "au" ;
-                    schema:name "Australia" ;
+                    sdo:identifier "au" ;
+                    sdo:name "Australia" ;
                 ] ;
-            schema:addressLocality "Shorncliffe" ;
-            schema:addressRegion "QLD" ;
-            schema:postalCode 4017 ;
-            schema:streetAddress (
+            sdo:addressLocality "Shorncliffe" ;
+            sdo:addressRegion "QLD" ;
+            sdo:postalCode 4017 ;
+            sdo:streetAddress (
                 72
                 "Yundah"
                 "Street"
             ) ;
         ] ;
-    schema:geo
+    sdo:geo
         [
-            schema:polygon "POLYGON((153.082403 -27.325801, 153.08241 -27.32582, 153.082943 -27.325612, 153.083010 -27.325742, 153.083543 -27.325521, 153.083456 -27.325365, 153.082403 -27.325801))"^^geo:wktLiteral ;
+            sdo:polygon "POLYGON((153.082403 -27.325801, 153.08241 -27.32582, 153.082943 -27.325612, 153.083010 -27.325742, 153.083543 -27.325521, 153.083456 -27.325365, 153.082403 -27.325801))"^^geo:wktLiteral ;
         ] ;
-    schema:name "KurrawongAI HQ" ;
+    sdo:name "KurrawongAI HQ" ;
 .
diff --git a/test/test_graph/test_graph.py b/test/test_graph/test_graph.py
@@ -399,7 +399,7 @@ def test_guess_format_for_parse_http_text_plain():
     assert len(graph) > 0
 
     # A url that returns content-type text/html.
-    url = "https://github.com/RDFLib/rdflib/issues/2734"
+    url = "https://www.w3.org/TR/REC-rdf-syntax/"
     with pytest.raises(PluginException):
         graph = Graph().parse(url)
 
diff --git a/test/test_serializers/test_serializer_longturtle.py b/test/test_serializers/test_serializer_longturtle.py
@@ -167,7 +167,7 @@ def test_longturtle():
     g.bind("sdo", SDO)
 
     # run the long turtle serializer
-    output = g.serialize(format="longturtle")
+    output = g.serialize(format="longturtle", canon=True)
 
     # fix the target
     current_dir = Path.cwd()  # Get the current directory
diff --git a/test/test_serializers/test_serializer_longturtle_sort.py b/test/test_serializers/test_serializer_longturtle_sort.py
