refactor(json-schema) added support for nested array json schemas

datahub-project · Feb 28, 2024 · 48c097e · 48c097e
1 parent 0e498c0
commit 48c097e
Show file tree

Hide file tree

Showing 2 changed files with 35 additions and 29 deletions.
diff --git a/metadata-ingestion/src/datahub/ingestion/extractor/json_schema_util.py b/metadata-ingestion/src/datahub/ingestion/extractor/json_schema_util.py
@@ -417,31 +417,37 @@ def _field_from_complex_type(
                     inner_field_path,
                 )
         elif datahub_field_type == ArrayTypeClass:
-            field_path = field_path.expand_type("array", schema)
-            # default items schema is string
-            items_schema = schema.get("items", {"type": "string"})
-            items_type = JsonSchemaTranslator._get_type_from_schema(items_schema)
-            # Extracting description for array field
-            description = JsonSchemaTranslator._get_description_from_any_schema(schema)
-            field_path._set_parent_type_if_not_exists(
-                DataHubType(type=ArrayTypeClass, nested_type=items_type)
-            )
-            yield from JsonSchemaTranslator.get_fields(
-                items_type, items_schema, required=False, base_field_path=field_path
-            )
-            # Adding a field for the array itself with the extracted description
+            field_path = field_path.expand_type(discriminated_type, schema)
             yield SchemaField(
                 fieldPath=field_path.as_string(),
                 type=type_override or SchemaFieldDataTypeClass(type=ArrayTypeClass()),
-                nativeDataType=native_type_override or "array",
-                description=description,
+                nativeDataType=native_type_override
+                or JsonSchemaTranslator._get_discriminated_type_from_schema(schema),
+                description=JsonSchemaTranslator._get_description_from_any_schema(
+                    schema
+                ),
                 nullable=nullable,
                 jsonProps=JsonSchemaTranslator._get_jsonprops_for_any_schema(
                     schema, required=required
                 ),
                 isPartOfKey=field_path.is_key_schema,
             )
 
+            items_schema = schema.get("items", {"type": "string"})
+            items_type = JsonSchemaTranslator._get_type_from_schema(items_schema)
+            field_name = items_schema.get("title", None)
+            if not field_name:
+                field_name = items_type
+            inner_field_path = field_path.clone_plus(
+                FieldElement(type=[], name=field_name, schema_types=[])
+            )
+            yield from JsonSchemaTranslator.get_fields(
+                items_type,
+                items_schema,
+                required=False,
+                base_field_path=inner_field_path,
+            )
+
         elif datahub_field_type == MapTypeClass:
             field_path = field_path.expand_type("map", schema)
             # When additionalProperties is used alone, without properties, the object essentially functions as a map<string, T> where T is the type described in the additionalProperties sub-schema. Maybe that helps to answer your original question.

diff --git a/metadata-ingestion/tests/unit/schema/test_json_schema_util.py b/metadata-ingestion/tests/unit/schema/test_json_schema_util.py
@@ -160,12 +160,12 @@ def test_json_schema_with_recursion():
             "type": NumberTypeClass,
         },
         {
-            "path": "[version=2.0].[type=TreeNode].[type=array].[type=TreeNode].children",
+            "path": "[version=2.0].[type=TreeNode].[type=array].children",
             "type": ArrayTypeClass,
         },
         {
-            "path": "[version=2.0].[type=TreeNode].[type=array].children",
-            "type": ArrayTypeClass,
+            "path": "[version=2.0].[type=TreeNode].[type=array].children.[type=TreeNode].TreeNode",
+            "type": RecordTypeClass,
         },
     ]
     assert_field_paths_match(fields, expected_field_paths)
@@ -377,10 +377,10 @@ def test_nested_arrays():
 
     fields = list(JsonSchemaTranslator.get_fields_from_schema(schema))
     expected_field_paths: List[str] = [
-        "[version=2.0].[type=NestedArray].[type=array].[type=array].[type=Foo].ar",
-        "[version=2.0].[type=NestedArray].[type=array].[type=array].[type=Foo].ar.[type=integer].a",
-        "[version=2.0].[type=NestedArray].[type=array].[type=array].ar",
         "[version=2.0].[type=NestedArray].[type=array].ar",
+        "[version=2.0].[type=NestedArray].[type=array].ar.[type=array].array",
+        "[version=2.0].[type=NestedArray].[type=array].ar.[type=array].array.[type=Foo].Foo",
+        "[version=2.0].[type=NestedArray].[type=array].ar.[type=array].array.[type=Foo].Foo.[type=integer].a",
     ]
     assert_field_paths_match(fields, expected_field_paths)
     assert isinstance(fields[0].type.type, ArrayTypeClass)
@@ -510,10 +510,10 @@ def test_needs_disambiguation_nested_union_of_records_with_same_field_name():
         "[version=2.0].[type=ABFooUnion].[type=union].[type=A].a.[type=string].f",
         "[version=2.0].[type=ABFooUnion].[type=union].[type=B].a",
         "[version=2.0].[type=ABFooUnion].[type=union].[type=B].a.[type=string].f",
-        "[version=2.0].[type=ABFooUnion].[type=union].[type=array].[type=array].[type=Foo].a",
-        "[version=2.0].[type=ABFooUnion].[type=union].[type=array].[type=array].[type=Foo].a.[type=integer].f",
-        "[version=2.0].[type=ABFooUnion].[type=union].[type=array].[type=array].a",
         "[version=2.0].[type=ABFooUnion].[type=union].[type=array].a",
+        "[version=2.0].[type=ABFooUnion].[type=union].[type=array].a.[type=array].array",
+        "[version=2.0].[type=ABFooUnion].[type=union].[type=array].a.[type=array].array.[type=Foo].Foo",
+        "[version=2.0].[type=ABFooUnion].[type=union].[type=array].a.[type=array].array.[type=Foo].Foo.[type=integer].f",
     ]
     assert_field_paths_match(fields, expected_field_paths)
 
@@ -588,10 +588,10 @@ def test_key_schema_handling():
         "[version=2.0].[key=True].[type=ABFooUnion].[type=union].[type=A].a.[type=string].f",
         "[version=2.0].[key=True].[type=ABFooUnion].[type=union].[type=B].a",
         "[version=2.0].[key=True].[type=ABFooUnion].[type=union].[type=B].a.[type=string].f",
-        "[version=2.0].[key=True].[type=ABFooUnion].[type=union].[type=array].[type=array].[type=Foo].a",
-        "[version=2.0].[key=True].[type=ABFooUnion].[type=union].[type=array].[type=array].[type=Foo].a.[type=number].f",
-        "[version=2.0].[key=True].[type=ABFooUnion].[type=union].[type=array].[type=array].a",
         "[version=2.0].[key=True].[type=ABFooUnion].[type=union].[type=array].a",
+        "[version=2.0].[key=True].[type=ABFooUnion].[type=union].[type=array].a.[type=array].array",
+        "[version=2.0].[key=True].[type=ABFooUnion].[type=union].[type=array].a.[type=array].array.[type=Foo].Foo",
+        "[version=2.0].[key=True].[type=ABFooUnion].[type=union].[type=array].a.[type=array].array.[type=Foo].Foo.[type=number].f",
     ]
     assert_field_paths_match(fields, expected_field_paths)
     for f in fields:
@@ -676,8 +676,8 @@ def test_simple_array():
 
     fields = list(JsonSchemaTranslator.get_fields_from_schema(schema))
     expected_field_paths: List[str] = [
-        "[version=2.0].[type=ObjectWithArray].[type=array].[type=string].ar",
         "[version=2.0].[type=ObjectWithArray].[type=array].ar",
+        "[version=2.0].[type=ObjectWithArray].[type=array].ar.[type=string].string",
     ]
     assert_field_paths_match(fields, expected_field_paths)
     assert isinstance(fields[0].type.type, ArrayTypeClass)
@@ -875,8 +875,8 @@ def test_description_extraction():
     }
     fields = list(JsonSchemaTranslator.get_fields_from_schema(schema))
     expected_field_paths: List[str] = [
-        "[version=2.0].[type=object].[type=array].[type=string].bar",
         "[version=2.0].[type=object].[type=array].bar",
+        "[version=2.0].[type=object].[type=array].bar.[type=string].string",
     ]
     assert_field_paths_match(fields, expected_field_paths)
     assert_fields_are_valid(fields)