feat(converters): add option include descendants

linkml/utils/converter.py

@@ -67,6 +67,14 @@
     help="Infer missing slot values",
 )
 @click.option("--context", "-c", multiple=True, help="path to JSON-LD context file")
+@click.option(
+    "--include-range-class-descendants/--no-range-class-descendants",
+    default=False,
+    show_default=False,
+    help="""
+When handling range constraints, include all descendants of the range class instead of just the range class
+""",
+)
 @click.version_option(__version__, "-V", "--version")
 @click.argument("input")
 def cli(
@@ -83,6 +91,7 @@ def cli(
     validate=None,
     infer=None,
     index_slot=None,
+    include_range_class_descendants=False,
 ) -> None:
     """
     Converts instance data to and from different LinkML Runtime serialization formats.
@@ -141,17 +150,25 @@ def cli(
                 raise Exception("--index-slot is required for CSV input")
         inargs["index_slot"] = index_slot
         inargs["schema"] = schema
-    obj = loader.load(source=input, target_class=py_target_class, **inargs)
-    if infer:
-        infer_config = inference_utils.Config(use_expressions=True, use_string_serialization=True)
-        infer_all_slot_values(obj, schemaview=sv, config=infer_config)
+    try:
+        data_as_dict = loader.load_as_dict(source=input, **inargs)
+    except NotImplementedError:
+        print("load_as_dict has failed! Loading as object")
+        obj = loader.load(source=input, target_class=py_target_class, **inargs)
+        data_as_dict = as_simple_dict(obj)
     if validate:
         if schema is None:
             raise Exception(
                 "--schema must be passed in order to validate. Suppress with --no-validate"
             )
-        # TODO: use validator framework
-        validation.validate_object(obj, schema)
+        validator = JsonSchemaDataValidator(schema, include_range_class_descendants=include_range_class_descendants)
+        validator.iter_validate_dict(data_as_dict, target_class_name=py_target_class.class_name)
+
+    obj = loader.load(source=input, target_class=py_target_class, **inargs)
+
+    if infer:
+        infer_config = inference_utils.Config(use_expressions=True, use_string_serialization=True)
+        infer_all_slot_values(obj, schemaview=sv, config=infer_config)
 
     output_format = _get_format(output, output_format, default="json")
     if output_format == "json-ld":

linkml/utils/validation.py

@@ -22,6 +22,7 @@ def validate_object(
     schema: Union[str, TextIO, SchemaDefinition],
     target_class: Type[YAMLRoot] = None,
     closed: bool = True,
+    include_range_class_descendants=False,
 ):
     """
     validates instance data against a schema
@@ -41,6 +42,7 @@ def validate_object(
         mergeimports=True,
         top_class=target_class.class_name,
         not_closed=not_closed,
+        include_range_class_descendants=include_range_class_descendants,
     ).serialize(not_closed=not_closed)
     jsonschema_obj = json.loads(jsonschemastr)
     return jsonschema.validate(

tests/test_utils/input/Person-01.yaml

@@ -0,0 +1,15 @@
+id: P:004
+name: eventful life
+has_events:
+  - employed_at: ROR:1
+    started_at_time: "2019-01-01"
+    is_current: true
+  - started_at_time: "2023-01-01"
+    in_location: GEO:1234
+    diagnosis:
+      id: CODE:P1789
+      name: hypertension
+    procedure:
+      id: CODE:P1846
+      name: valve repair
+

tests/test_utils/input/animals.yaml

@@ -0,0 +1,7 @@
+animals:
+  - animal_family: Dog
+    max_age: "17 years"
+    breed: Golden Retriever
+  - animal_family: Ant
+    max_age: "7 years"
+    venom: true

tests/test_utils/input/schema7.yaml

@@ -0,0 +1,36 @@
+id: https://example.org/descendants
+name: Test_descendants
+imports:
+  - linkml:types
+prefixes:
+  linkml: https://w3id.org/linkml/
+classes:
+  Animal:
+    slots:
+      - animal_family
+    attributes:
+      max_age:
+        range: string
+  Dog:
+    is_a: Animal
+    slots:
+      - animal_family
+      - breed
+  Ant:
+    is_a: Animal
+    slots:
+      - animal_family
+      - venom
+  Container:
+    attributes:
+      animals:
+        multivalued: true
+        range: Animal
+slots:
+  animal_family:
+    range: string
+    designates_type: true
+  breed:
+    range: string
+  venom:
+    range: boolean

tests/test_utils/test_converter.py

@@ -12,6 +12,9 @@
 YAML_OUT = env.expected_path("data_example.out.yaml")
 RDF_OUT = env.expected_path("data_example.out.ttl")
 
+SCHEMA_DESCENDANTS = env.input_path("schema7.yaml")
+DATA_DESCENDANTS = env.input_path("animals.yaml")
+
 
 class TestCommandLineInterface(unittest.TestCase):
     def setUp(self) -> None:
@@ -49,6 +52,40 @@ def test_infer_and_convert(self):
             self.assertEqual(p2["age_category"], "adult")
             self.assertEqual(p2["full_name"], "first2 last2")
 
+    def test_convert_including_descendants(self):
+        """
+        Tests using the --include-range-class-descendants option to support
+        subtype polymorphism.
+        """
+        result = self.runner.invoke(
+            cli,
+            [
+                "--include-range-class-descendants",
+                "--validate",
+                "-C",
+                "Container",
+                "-s",
+                SCHEMA_DESCENDANTS,
+                "-t",
+                "json-ld",
+                "-o",
+                JSON_OUT,
+                DATA_DESCENDANTS,
+            ],
+        )
+        print(result.stdout)
+        if result.exit_code:
+            print(result.exception)
+            raise result.exception
+        else:
+            with open(JSON_OUT) as file:
+                p1 = json.load(file)
+                print(p1)
+                self.assertEqual(p1["animals"][0]["animal_family"], "Dog")
+                self.assertEqual(p1["animals"][0]["breed"], "Golden Retriever")
+                self.assertEqual(p1["animals"][1]["animal_family"], "Ant")
+                self.assertTrue(p1["animals"][1]["venom"])
+
     def test_version(self):
         runner = CliRunner(mix_stderr=False)
         result = runner.invoke(cli, ["--version"])