Adds ontology support

src/database/datasets.py

@@ -131,6 +131,23 @@ def get_features(dataset_id: int, connection: Connection) -> list[Feature]:
     return [Feature(**row, nominal_values=None) for row in rows.mappings()]
 
 
+def get_feature_ontologies(dataset_id: int, connection: Connection) -> dict[int, list[str]]:
+    rows = connection.execute(
+        text(
+            """
+            SELECT `index`, `value`
+            FROM data_feature_description
+            WHERE `did` = :dataset_id AND `description_type` = 'ontology'
+            """,
+        ),
+        parameters={"dataset_id": dataset_id},
+    )
+    ontologies: dict[int, list[str]] = {}
+    for row in rows.mappings():
+        ontologies.setdefault(row["index"], []).append(row["value"])
+    return ontologies
+
+
 def get_feature_values(dataset_id: int, *, feature_index: int, connection: Connection) -> list[str]:
     rows = connection.execute(
         text(

src/routers/openml/datasets.py

@@ -287,6 +287,12 @@ def get_dataset_features(
 ) -> list[Feature]:
     _get_dataset_raise_otherwise(dataset_id, user, expdb)
     features = database.datasets.get_features(dataset_id, expdb)
+
+    # Attach ontologies from data_feature_description
+    ontologies = database.datasets.get_feature_ontologies(dataset_id, expdb)
+    for feature in features:
+        feature.ontology = ontologies.get(feature.index)
+
     for feature in [f for f in features if f.data_type == FeatureType.NOMINAL]:
         feature.nominal_values = database.datasets.get_feature_values(
             dataset_id,

src/schemas/datasets/openml.py

@@ -40,6 +40,7 @@ class Feature(BaseModel):
     index: int
     name: str
     data_type: FeatureType
+    ontology: list[str] | None = None
     is_target: bool
     is_ignore: bool
     is_row_identifier: bool