Compare objects MRO with encoders at runtime

The previous implementation doesn't handle subclass instances when
pydantic.json.ENCODERS_BY_TYPE is modified after fastapi.encoders import.

This diff makes it easier for developers to add custom encoders that also work
with subclass instances (and it simplifies the code, as well).

fastapi/encoders.py

@@ -1,6 +1,6 @@
 import dataclasses
 import datetime
-from collections import defaultdict, deque
+from collections import deque
 from collections.abc import Callable
 from decimal import Decimal
 from enum import Enum
@@ -95,20 +95,6 @@ ENCODERS_BY_TYPE: dict[type[Any], Callable[[Any], Any]] = {
 }
 
 
-def generate_encoders_by_class_tuples(
-    type_encoder_map: dict[Any, Callable[[Any], Any]],
-) -> dict[Callable[[Any], Any], tuple[Any, ...]]:
-    encoders_by_class_tuples: dict[Callable[[Any], Any], tuple[Any, ...]] = defaultdict(
-        tuple
-    )
-    for type_, encoder in type_encoder_map.items():
-        encoders_by_class_tuples[encoder] += (type_,)
-    return encoders_by_class_tuples
-
-
-encoders_by_class_tuples = generate_encoders_by_class_tuples(ENCODERS_BY_TYPE)
-
-
 def jsonable_encoder(
     obj: Annotated[
         Any,
@@ -314,11 +300,9 @@ def jsonable_encoder(
             )
         return encoded_list
 
-    if type(obj) in ENCODERS_BY_TYPE:
+    for base in obj.__class__.__mro__[:-1]:
-        return ENCODERS_BY_TYPE[type(obj)](obj)
+        if base in ENCODERS_BY_TYPE:
-    for encoder, classes_tuple in encoders_by_class_tuples.items():
+            return ENCODERS_BY_TYPE[base](obj)
-        if isinstance(obj, classes_tuple):
-            return encoder(obj)
     if is_pydantic_v1_model_instance(obj):
         raise PydanticV1NotSupportedError(
             "pydantic.v1 models are no longer supported by FastAPI."