fix mapping input content when associated with json mixin (i.e. FileD…

…ata)

test/test_utils.py

@@ -3,6 +3,7 @@
 from enum import Enum
 
 from pydantic import BaseModel
+from unstructured.ingest.v2.interfaces import FileData
 from uvicorn.importer import import_from_string
 
 from unstructured_platform_plugins.etl_uvicorn import utils
@@ -101,14 +102,20 @@ class MyEnum(Enum):
 
 
 def test_map_inputs():
-    def fn(a: A, b: B, c: MyEnum, d: list) -> None:
+    def fn(a: A, b: B, c: MyEnum, d: list, e: FileData) -> None:
         pass
 
+    file_data = FileData(
+        identifier="custom_file_data",
+        connector_type="mock_connector",
+        additional_metadata={"additional": "metadata"},
+    )
     inputs = {
         "a": {"b": 4, "c": 5.6},
         "b": {"d": True, "e": {"key": "value"}},
         "c": MyEnum.VALUE.value,
         "d": [1, 2, 3],
+        "e": file_data.to_dict(),
     }
 
     mapped_inputs = utils.map_inputs(func=fn, raw_inputs=inputs)
@@ -117,5 +124,6 @@ def fn(a: A, b: B, c: MyEnum, d: list) -> None:
         "b": B(d=True, e={"key": "value"}),
         "c": MyEnum.VALUE.value,
         "d": [1, 2, 3],
+        "e": file_data,
     }
     assert mapped_inputs == expected

unstructured_platform_plugins/etl_uvicorn/utils.py

@@ -4,6 +4,7 @@
 from types import GenericAlias, NoneType
 from typing import Any, Callable, Optional
 
+from dataclasses_json import DataClassJsonMixin
 from pydantic import BaseModel
 
 from unstructured_platform_plugins.schema.json_schema import (
@@ -83,11 +84,16 @@ def map_inputs(func: Callable, raw_inputs: dict[str, Any]) -> dict[str, Any]:
     type_info = get_type_hints(func)
     type_info.pop("return")
     for field_name, type_data in type_info.items():
+        if field_name not in raw_inputs:
+            continue
+
         if (
-            is_dataclass(type_data)
-            and field_name in raw_inputs
+            inspect.isclass(type_data)
+            and issubclass(type_data, DataClassJsonMixin)
             and isinstance(raw_inputs[field_name], dict)
         ):
+            raw_inputs[field_name] = type_data.from_dict(raw_inputs[field_name])
+        elif is_dataclass(type_data) and isinstance(raw_inputs[field_name], dict):
             raw_inputs[field_name] = type_data(**raw_inputs[field_name])
         elif isinstance(type_data, EnumMeta):
             raw_inputs[field_name] = raw_inputs[field_name]