feat: test public methods like graph_from_dataset

Author: stefina

ckanext/switzerland/tests/test_swiss_schemaorg.py

@@ -1,115 +1,103 @@
+import pytest
 import rdflib
 
 from ckanext.dcat.profiles import SCHEMA
-
+from ckanext.dcat.utils import resource_uri
 from ckanext.switzerland.dcat.profiles import SwissSchemaOrgProfile
 
 
-def _profile():
+def _tagged_lang_values(graph, subject, predicate):
+    """helper for language-tagged literals on ``predicate``."""
+    return [
+        (str(o.language), str(o))
+        for o in graph.objects(subject, predicate)
+        if getattr(o, "language", None)
+    ]
+
+
+@pytest.fixture
+def profile():
     return SwissSchemaOrgProfile(rdflib.Graph())
 
 
-def test_basic_fields_map_fluent_description():
-    """
-    Make sure SwissSchemaOrgProfile maps multilang descriptions to schema:description.
-    """
-    profile = _profile()
+def test_graph_from_dataset_maps_multilang_description_and_title(profile):
     dataset_ref = rdflib.URIRef("http://example.org/dataset/1")
-
     dataset_dict = {
         "name": "ds",
         "title": {"de": "Titel", "fr": "Titre"},
         "description": {"de": "DE text", "fr": "FR texte"},
         "resources": [],
     }
 
-    profile._basic_fields_graph(dataset_ref, dataset_dict)
+    profile.graph_from_dataset(dataset_dict, dataset_ref)
 
-    translated_values = [
-        (str(o.language), str(o))
-        for o in profile.g.objects(dataset_ref, SCHEMA.description)
-        if getattr(o, "language", None)
-    ]
-    assert ("de", "DE text") in translated_values
-    assert ("fr", "FR texte") in translated_values
+    assert ("de", "DE text") in _tagged_lang_values(
+        profile.g, dataset_ref, SCHEMA.description
+    )
+    assert ("fr", "FR texte") in _tagged_lang_values(
+        profile.g, dataset_ref, SCHEMA.description
+    )
+    assert ("de", "Titel") in _tagged_lang_values(profile.g, dataset_ref, SCHEMA.name)
+    assert ("fr", "Titre") in _tagged_lang_values(profile.g, dataset_ref, SCHEMA.name)
 
 
-def test_basic_fields_map_fluent_title():
-    profile = _profile()
+def test_graph_from_dataset_empty_description_dict_emits_no_tagged_description(
+    profile,
+):
     dataset_ref = rdflib.URIRef("http://example.org/dataset/1")
     dataset_dict = {
         "name": "ds",
-        "title": {"de": "Titel", "it": "Titolo"},
+        "description": {},
         "resources": [],
     }
-    profile._basic_fields_graph(dataset_ref, dataset_dict)
-    translated_values = [
-        (str(o.language), str(o))
-        for o in profile.g.objects(dataset_ref, SCHEMA.name)
-        if getattr(o, "language", None)
-    ]
-    assert ("de", "Titel") in translated_values
-    assert ("it", "Titolo") in translated_values
 
+    profile.graph_from_dataset(dataset_dict, dataset_ref)
 
-def test_empty_description_dict_emits_no_description_literals():
-    """Multilingual serialization skips empty dicts (same rules as DCAT profile)."""
-    profile = _profile()
-    dataset_ref = rdflib.URIRef("http://example.org/dataset/1")
-    profile._basic_fields_graph(
-        dataset_ref,
-        {
-            "name": "ds",
-            "description": {},
-            "resources": [],
-        },
-    )
-    assert list(profile.g.objects(dataset_ref, SCHEMA.description)) == []
+    assert _tagged_lang_values(profile.g, dataset_ref, SCHEMA.description) == []
 
 
-def test_description_empty_string_lang_skipped_whitespace_kept():
+def test_graph_from_dataset_description_empty_lang_skipped_whitespace_kept(profile):
     """
     Empty value for ``fr`` is skipped. Whitespace-only ``de`` is truthy, so a
     language-tagged literal is emitted (``MultiLangProfile`` does not strip).
     """
-    profile = _profile()
     dataset_ref = rdflib.URIRef("http://example.org/dataset/2")
-    profile._basic_fields_graph(
-        dataset_ref,
-        {
-            "name": "ds",
-            "description": {"de": "   ", "en": "Kept", "fr": ""},
-            "resources": [],
-        },
-    )
-    translated_values = [
-        (str(o.language), str(o))
-        for o in profile.g.objects(dataset_ref, SCHEMA.description)
-        if getattr(o, "language", None)
-    ]
+    dataset_dict = {
+        "name": "ds",
+        "description": {"de": "   ", "en": "Kept", "fr": ""},
+        "resources": [],
+    }
+
+    profile.graph_from_dataset(dataset_dict, dataset_ref)
+
+    translated_values = _tagged_lang_values(profile.g, dataset_ref, SCHEMA.description)
     assert set(translated_values) == {("de", "   "), ("en", "Kept")}
 
 
-def test_distribution_basic_fields_multilang_name_and_description():
-    profile = _profile()
-    distribution = rdflib.URIRef("http://example.org/dist/1")
+def test_graph_from_dataset_maps_multilang_resource_title_and_description(profile):
+    resource_uri_str = "http://example.org/resource/1"
     resource_dict = {
+        "id": "res-1",
+        "uri": resource_uri_str,
+        "url": "http://example.org/data.csv",
         "title": {"de": "Ressource", "fr": "Ressource FR"},
         "description": {"de": "Beschreibung"},
     }
-    profile._distribution_basic_fields_graph(distribution, resource_dict)
+    dataset_ref = rdflib.URIRef("http://example.org/dataset/3")
+    dataset_dict = {
+        "name": "ds",
+        "title": {"de": "Dataset"},
+        "resources": [resource_dict],
+    }
 
-    name_translated_values = [
-        (str(o.language), str(o))
-        for o in profile.g.objects(distribution, SCHEMA.name)
-        if getattr(o, "language", None)
-    ]
-    assert ("de", "Ressource") in name_translated_values
-    assert ("fr", "Ressource FR") in name_translated_values
+    assert resource_uri(resource_dict) == resource_uri_str
+    distribution = rdflib.URIRef(resource_uri_str)
 
-    desc_translated_values = [
-        (str(o.language), str(o))
-        for o in profile.g.objects(distribution, SCHEMA.description)
-        if getattr(o, "language", None)
-    ]
-    assert desc_translated_values == [("de", "Beschreibung")]
+    profile.graph_from_dataset(dataset_dict, dataset_ref)
+
+    name_values = _tagged_lang_values(profile.g, distribution, SCHEMA.name)
+    assert ("de", "Ressource") in name_values
+    assert ("fr", "Ressource FR") in name_values
+
+    desc_values = _tagged_lang_values(profile.g, distribution, SCHEMA.description)
+    assert set(desc_values) == {("de", "Beschreibung")}