feat: sort longturtle blank nodes (#2997)

* feat: sort longturtle blank nodes in the object position by their cbd string * fix: #2767
RDFLib · Nov 29, 2024 · 28a6190 · 28a6190
1 parent 08dd4b7
commit 28a6190
Show file tree

Hide file tree

Showing 3 changed files with 108 additions and 77 deletions.
diff --git a/rdflib/plugins/serializers/longturtle.py b/rdflib/plugins/serializers/longturtle.py
@@ -293,7 +293,34 @@ def predicateList(self, subject, newline=False):
     def verb(self, node, newline=False):
         self.path(node, VERB, newline)
 
+    def sortObjects(
+        self, values: list[URIRef | BNode | Literal]
+    ) -> list[URIRef | BNode | Literal]:
+        """
+        Perform a sort on the values where each value is a blank node. Grab the CBD of the
+        blank node and sort it by its longturtle serialization value.
+
+        Identified nodes come first and the sorted blank nodes are tacked on after.
+        """
+        bnode_map: dict[BNode, list[str]] = {}
+        objects = []
+        for value in values:
+            if isinstance(value, BNode):
+                bnode_map[value] = []
+            else:
+                objects.append(value)
+
+        for bnode in bnode_map:
+            cbd = self.store.cbd(bnode).serialize(format="longturtle")
+            bnode_map[bnode].append(cbd)
+
+        sorted_bnodes = sorted(
+            [(k, v) for k, v in bnode_map.items()], key=lambda x: x[1]
+        )
+        return objects + [x[0] for x in sorted_bnodes]
+
     def objectList(self, objects):
+        objects = self.sortObjects(objects)
         count = len(objects)
         if count == 0:
             return
@@ -303,6 +330,8 @@ def objectList(self, objects):
         if count > 1:
             if not isinstance(objects[0], BNode):
                 self.write("\n" + self.indent(1))
+            else:
+                self.write(" ")
             first_nl = True
         self.path(objects[0], OBJECT, newline=first_nl)
         for obj in objects[1:]:

diff --git a/test/data/longturtle/longturtle-target.ttl b/test/data/longturtle/longturtle-target.ttl
@@ -0,0 +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 sdo: <https://schema.org/>
+PREFIX xsd: <http://www.w3.org/2001/XMLSchema#>
+
+ex:nicholas
+    a sdo:Person ;
+    sdo:age 41 ;
+    sdo:alternateName
+        "N.J. Car" ,
+        "Nick Car" ,
+        [
+            sdo:name "Dr N.J. Car" ;
+        ] ;
+    sdo:name
+        [
+            a cn:CompoundName ;
+            sdo:hasPart 
+                [
+                    a cn:CompoundName ;
+                    rdf:value "John" ;
+                ] ,
+                [
+                    a cn:CompoundName ;
+                    rdf:value "Nicholas" ;
+                ] ,
+                [
+                    a cn:CompoundName ;
+                    sdo:hasPart 
+                        [
+                            a cn:CompoundName ;
+                            rdf:value "Car" ;
+                        ] ,
+                        [
+                            a cn:CompoundName ;
+                            rdf:value "Maxov" ;
+                        ] ;
+                ] ;
+        ] ;
+    sdo:worksFor <https://kurrawong.ai> ;
+.
+
+<https://kurrawong.ai>
+    a sdo:Organization ;
+    sdo:location <https://kurrawong.ai/hq> ;
+.
+
+<https://kurrawong.ai/hq>
+    a sdo:Place ;
+    sdo:address
+        [
+            a sdo:PostalAddress ;
+            sdo:addressCountry
+                [
+                    sdo:identifier "au" ;
+                    sdo:name "Australia" ;
+                ] ;
+            sdo:addressLocality "Shorncliffe" ;
+            sdo:addressRegion "QLD" ;
+            sdo:postalCode 4017 ;
+            sdo:streetAddress (
+                72
+                "Yundah"
+                "Street"
+            ) ;
+        ] ;
+    sdo:geo
+        [
+            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 ;
+        ] ;
+    sdo:name "KurrawongAI HQ" ;
+.
diff --git a/test/test_serializers/test_serializer_longturtle.py b/test/test_serializers/test_serializer_longturtle.py
@@ -1,5 +1,5 @@
 import difflib
-from textwrap import dedent
+from pathlib import Path
 
 from rdflib import Graph, Namespace
 from rdflib.namespace import GEO, SDO
@@ -170,83 +170,11 @@ def test_longturtle():
     output = g.serialize(format="longturtle")
 
     # fix the target
-    target = dedent(
-        """    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 sdo: <https://schema.org/>
-    PREFIX xsd: <http://www.w3.org/2001/XMLSchema#>
+    current_dir = Path.cwd()  # Get the current directory
+    target_file_path = current_dir / "test/data/longturtle" / "longturtle-target.ttl"
 
-    ex:nicholas
-        a sdo:Person ;
-        sdo:age 41 ;
-        sdo:alternateName
-            [
-                sdo:name "Dr N.J. Car" ;
-            ] ,
-            "N.J. Car" ,
-            "Nick Car" ;
-        sdo:name
-            [
-                a cn:CompoundName ;
-                sdo:hasPart
-                    [
-                        a cn:CompoundName ;
-                        rdf:value "Nicholas" ;
-                    ] ,
-                    [
-                        a cn:CompoundName ;
-                        rdf:value "John" ;
-                    ] ,
-                    [
-                        a cn:CompoundName ;
-                        sdo:hasPart
-                            [
-                                a cn:CompoundName ;
-                                rdf:value "Car" ;
-                            ] ,
-                            [
-                                a cn:CompoundName ;
-                                rdf:value "Maxov" ;
-                            ] ;
-                    ] ;
-            ] ;
-        sdo:worksFor <https://kurrawong.ai> ;
-    .
-
-    <https://kurrawong.ai>
-        a sdo:Organization ;
-        sdo:location <https://kurrawong.ai/hq> ;
-    .
-
-    <https://kurrawong.ai/hq>
-        a sdo:Place ;
-        sdo:address
-            [
-                a sdo:PostalAddress ;
-                sdo:addressCountry
-                    [
-                        sdo:identifier "au" ;
-                        sdo:name "Australia" ;
-                    ] ;
-                sdo:addressLocality "Shorncliffe" ;
-                sdo:addressRegion "QLD" ;
-                sdo:postalCode 4017 ;
-                sdo:streetAddress (
-                    72
-                    "Yundah"
-                    "Street"
-                ) ;
-            ] ;
-        sdo:geo
-            [
-                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 ;
-            ] ;
-        sdo:name "KurrawongAI HQ" ;
-    .
-    """
-    )
+    with open(target_file_path, encoding="utf-8") as file:
+        target = file.read()
 
     # compare output to target
     # - any differences will produce output