fix: correctly handle resources with multiple comments in defined_namespace_creator (#2254)

FIxed a bug in defined_namespace_creator so that when a resource in an ontology is commented on multiple times (with e.g. rdfs:comment), it creates only a single class variable in the resulting DefinedNamespace where the Python comment is the concatenation of the rdfs:comments.

4 files changed, 73 insertions, 1 deletions

diff --git a/rdflib/tools/defined_namespace_creator.py b/rdflib/tools/defined_namespace_creator.py
index 0c93ea75..1076cd6e 100644
--- a/rdflib/tools/defined_namespace_creator.py
+++ b/rdflib/tools/defined_namespace_creator.py
@@ -77,7 +77,7 @@ def get_target_namespace_elements(
 ) -> Tuple[List[Tuple[str, str]], List[str]]:
     namespaces = {"dcterms": DCTERMS, "owl": OWL, "rdfs": RDFS, "skos": SKOS}
     q = """
-        SELECT DISTINCT ?s ?def
+        SELECT ?s (GROUP_CONCAT(DISTINCT STR(?def)) AS ?defs)
         WHERE {
             # all things in the RDF data (anything RDF.type...)
             ?s a ?o .
@@ -90,6 +90,7 @@ def get_target_namespace_elements(
             # only get results for the target namespace (supplied by user)
             FILTER STRSTARTS(STR(?s), "xxx")
         }
+        GROUP BY ?s
         """.replace(
         "xxx", target_namespace
     )
diff --git a/test/data/contrived/README.md b/test/data/contrived/README.md
new file mode 100644
index 00000000..fd1e0e7d
--- /dev/null
+++ b/test/data/contrived/README.md
@@ -0,0 +1,5 @@
+# Contrived Test Data
+
+This directory contains test data contrived for specific purposes. Files in this
+directory should clearly indicate their purpose with a comment.
+
diff --git a/test/data/contrived/multiple-comments.ttl b/test/data/contrived/multiple-comments.ttl
new file mode 100644
index 00000000..a2c07371
--- /dev/null
+++ b/test/data/contrived/multiple-comments.ttl
@@ -0,0 +1,17 @@
+@prefix rdfs:           <http://www.w3.org/2000/01/rdf-schema#> .
+@prefix owl:            <http://www.w3.org/2002/07/owl#> .
+@prefix ex:             <http://example.org/multiline-string-example#> .
+
+# This file contains a RDF class with multiple rdfs:comment properties and is
+# used to verify the RDFLib defined namespace creator. It is used in
+# </test/test_namespace/test_definednamespace_creator.py>.
+
+<http://example.org/multiline-string-example>
+    a owl:Ontology .
+
+ex:SomeClass a rdfs:Class, owl:Class;
+    rdfs:label "Some class"@en;
+    rdfs:comment "If one uses multiple comment properties, "@en;
+    rdfs:comment "then it should still only create a single class variable."@en;
+    rdfs:isDefinedBy <http://example.org/multiline-string-example>;
+.
diff --git a/test/test_namespace/test_definednamespace_creator.py b/test/test_namespace/test_definednamespace_creator.py
index 65734b21..8866a05d 100644
--- a/test/test_namespace/test_definednamespace_creator.py
+++ b/test/test_namespace/test_definednamespace_creator.py
@@ -114,3 +114,52 @@ def test_definednamespace_creator_bad_ns():
         universal_newlines=True,
     )
     assert completed.returncode == 1, "subprocess exited incorrectly (failure expected)"
+
+
+def test_definednamespace_creator_multiple_comments():
+    """
+    Tests that only a single URIRef is declared, even when multiple
+    rdfs:comments are linked to the resource.
+    """
+
+    definednamespace_script = (
+        Path(__file__).parent.parent.parent
+        / "rdflib"
+        / "tools"
+        / "defined_namespace_creator.py"
+    )
+    multiple_comments_data_file = (
+        Path(__file__).parent.parent / "data" / "contrived" / "multiple-comments.ttl"
+    )
+    print("\n")
+    print(f"Using {definednamespace_script}...")
+    print(f"Testing {multiple_comments_data_file}...")
+    completed = subprocess.run(
+        [
+            sys.executable,
+            str(definednamespace_script),
+            str(multiple_comments_data_file),
+            "http://example.org/multiline-string-example#",
+            "MULTILINESTRINGEXAMPLE",
+        ],
+        stdout=subprocess.PIPE,
+        stderr=subprocess.PIPE,
+        universal_newlines=True,
+    )
+    assert completed.returncode == 0, "subprocess exited incorrectly"
+    assert Path.is_file(
+        Path("_MULTILINESTRINGEXAMPLE.py")
+    ), "_MULTILINESTRINGEXAMPLE.py file not created"
+
+    some_class_count = 0
+    with open(Path("_MULTILINESTRINGEXAMPLE.py")) as f:
+        for line in f.readlines():
+            if "SomeClass: URIRef" in line:
+                some_class_count += 1
+
+    assert (
+        some_class_count == 1
+    ), f"found {some_class_count} SomeClass definitions instead of 1."
+
+    # cleanup
+    Path.unlink(Path("_MULTILINESTRINGEXAMPLE.py"))