Add HTTP deserialization option (#4002)

libs/checkpoint/langgraph/checkpoint/serde/jsonplus.py

@@ -33,6 +33,15 @@ LC_REVIVER = Reviver()
 
 
 class JsonPlusSerializer(SerializerProtocol):
+    def __init__(
+        self, *, __unpack_ext_hook__: Optional[Callable[[int, bytes], Any]] = None
+    ) -> None:
+        self._unpack_ext_hook = (
+            __unpack_ext_hook__
+            if __unpack_ext_hook__ is not None
+            else _msgpack_ext_hook
+        )
+
     def _encode_constructor_args(
         self,
         constructor: Union[Callable, type[Any]],
@@ -210,7 +219,7 @@ class JsonPlusSerializer(SerializerProtocol):
             return self.loads(data_)
         elif type_ == "msgpack":
             return ormsgpack.unpackb(
-                data_, ext_hook=_msgpack_ext_hook, option=ormsgpack.OPT_NON_STR_KEYS
+                data_, ext_hook=self._unpack_ext_hook, option=ormsgpack.OPT_NON_STR_KEYS
             )
         else:
             raise NotImplementedError(f"Unknown serialization type: {type_}")
@@ -521,6 +530,82 @@ def _msgpack_ext_hook(code: int, data: bytes) -> Any:
                 return
 
 
+def _msgpack_ext_hook_to_json(code: int, data: bytes) -> Any:
+    if code == EXT_CONSTRUCTOR_SINGLE_ARG:
+        try:
+            tup = ormsgpack.unpackb(
+                data,
+                ext_hook=_msgpack_ext_hook_to_json,
+                option=ormsgpack.OPT_NON_STR_KEYS,
+            )
+            if tup[0] == "uuid" and tup[1] == "UUID":
+                hex_ = tup[2]
+                return (
+                    f"{hex_[:8]}-{hex_[8:12]}-{hex_[12:16]}-{hex_[16:20]}-{hex_[20:]}"
+                )
+            # module, name, arg
+            return tup[2]
+        except Exception:
+            return
+    elif code == EXT_CONSTRUCTOR_POS_ARGS:
+        try:
+            tup = ormsgpack.unpackb(
+                data,
+                ext_hook=_msgpack_ext_hook_to_json,
+                option=ormsgpack.OPT_NON_STR_KEYS,
+            )
+            # module, name, args
+            return tup[2]
+        except Exception:
+            return
+    elif code == EXT_CONSTRUCTOR_KW_ARGS:
+        try:
+            tup = ormsgpack.unpackb(
+                data,
+                ext_hook=_msgpack_ext_hook_to_json,
+                option=ormsgpack.OPT_NON_STR_KEYS,
+            )
+            # module, name, args
+            return tup[2]
+        except Exception:
+            return
+    elif code == EXT_METHOD_SINGLE_ARG:
+        try:
+            tup = ormsgpack.unpackb(
+                data,
+                ext_hook=_msgpack_ext_hook_to_json,
+                option=ormsgpack.OPT_NON_STR_KEYS,
+            )
+            # module, name, arg, method
+            return tup[2]
+        except Exception:
+            return
+    elif code == EXT_PYDANTIC_V1:
+        try:
+            tup = ormsgpack.unpackb(
+                data,
+                ext_hook=_msgpack_ext_hook_to_json,
+                option=ormsgpack.OPT_NON_STR_KEYS,
+            )
+            # module, name, kwargs
+            return tup[2]
+        except Exception:
+            # for pydantic objects we can't find/reconstruct
+            # let's return the kwargs dict instead
+            return
+    elif code == EXT_PYDANTIC_V2:
+        try:
+            tup = ormsgpack.unpackb(
+                data,
+                ext_hook=_msgpack_ext_hook_to_json,
+                option=ormsgpack.OPT_NON_STR_KEYS,
+            )
+            # module, name, kwargs, method
+            return tup[2]
+        except Exception:
+            return
+
+
 _option = (
     ormsgpack.OPT_NON_STR_KEYS
     | ormsgpack.OPT_PASSTHROUGH_DATACLASS

libs/checkpoint/pyproject.toml

@@ -1,6 +1,6 @@
 [tool.poetry]
 name = "langgraph-checkpoint"
-version = "2.0.21"
+version = "2.0.22"
 description = "Library with base interfaces for LangGraph checkpoint savers."
 authors = []
 license = "MIT"

libs/checkpoint/tests/test_jsonplus.py

@@ -15,7 +15,10 @@ from pydantic.v1 import BaseModel as BaseModelV1
 from pydantic.v1 import SecretStr as SecretStrV1
 from zoneinfo import ZoneInfo
 
-from langgraph.checkpoint.serde.jsonplus import JsonPlusSerializer
+from langgraph.checkpoint.serde.jsonplus import (
+    JsonPlusSerializer,
+    _msgpack_ext_hook_to_json,
+)
 from langgraph.store.base import Item
 
 
@@ -165,6 +168,113 @@ def test_serde_jsonplus() -> None:
     ]
 
 
+def test_serde_jsonplus_json_mode() -> None:
+    uid = uuid.UUID(int=1)
+    deque_instance = deque([1, 2, 3])
+    tzn = ZoneInfo("America/New_York")
+    ip4 = IPv4Address("192.168.0.1")
+    current_date = date(2024, 4, 19)
+    current_time = time(23, 4, 57, 51022, timezone.max)
+    current_timestamp = datetime(2024, 4, 19, 23, 4, 57, 51022, timezone.max)
+
+    to_serialize = {
+        "path": pathlib.Path("foo", "bar"),
+        "re": re.compile(r"foo", re.DOTALL),
+        "decimal": Decimal("1.10101"),
+        "set": {1, 2, frozenset({1, 2})},
+        "frozen_set": frozenset({1, 2, 3}),
+        "ip4": ip4,
+        "deque": deque_instance,
+        "tzn": tzn,
+        "date": current_date,
+        "time": current_time,
+        "uid": uid,
+        "timestamp": current_timestamp,
+        "my_slotted_class": MyDataclassWSlots("bar", 2, InnerDataclass("hello")),
+        "my_dataclass": MyDataclass("foo", 1, InnerDataclass("hello")),
+        "my_enum": MyEnum.FOO,
+        "my_pydantic": MyPydantic(foo="foo", bar=1, inner=InnerPydantic(hello="hello")),
+        "my_pydantic_v1": MyPydanticV1(
+            foo="foo", bar=1, inner=InnerPydanticV1(hello="hello")
+        ),
+        "my_secret_str": SecretStr("meow"),
+        "my_secret_str_v1": SecretStrV1("meow"),
+        "person": Person(name="foo"),
+        "a_bool": True,
+        "a_none": None,
+        "a_str": "foo",
+        "a_str_nuc": "foo\u0000",
+        "a_str_uc": "foo ⛰️",
+        "a_str_ucuc": "foo \u26f0\ufe0f\u0000",
+        "a_str_ucucuc": "foo \\u26f0\\ufe0f",
+        "an_int": 1,
+        "a_float": 1.1,
+        "a_bytes": b"my bytes",
+        "a_bytearray": bytearray([42]),
+        "my_item": Item(
+            value={},
+            key="my-key",
+            namespace=("a", "name", " "),
+            created_at=datetime(2024, 9, 24, 17, 29, 10, 128397),
+            updated_at=datetime(2024, 9, 24, 17, 29, 11, 128397),
+        ),
+    }
+
+    serde = JsonPlusSerializer(__unpack_ext_hook__=_msgpack_ext_hook_to_json)
+
+    dumped = serde.dumps_typed(to_serialize)
+
+    assert dumped[0] == "msgpack"
+    result = serde.loads_typed(dumped)
+    assert result == {
+        "path": ["foo", "bar"],
+        "re": ["foo", 48],
+        "decimal": "1.10101",
+        "set": [1, 2, [1, 2]],
+        "frozen_set": [1, 2, 3],
+        "ip4": "192.168.0.1",
+        "deque": [1, 2, 3],
+        "tzn": "America/New_York",
+        "date": [2024, 4, 19],
+        "time": {
+            "hour": 23,
+            "minute": 4,
+            "second": 57,
+            "microsecond": 51022,
+            "tzinfo": [[0, 86340, 0]],
+            "fold": 0,
+        },
+        "uid": "00000000-0000-0000-0000-000000000001",
+        "timestamp": "2024-04-19T23:04:57.051022+23:59",
+        "my_slotted_class": {"foo": "bar", "bar": 2, "inner": {"hello": "hello"}},
+        "my_dataclass": {"foo": "foo", "bar": 1, "inner": {"hello": "hello"}},
+        "my_enum": "foo",
+        "my_pydantic": {"foo": "foo", "bar": 1, "inner": {"hello": "hello"}},
+        "my_pydantic_v1": {"foo": "foo", "bar": 1, "inner": {"hello": "hello"}},
+        "my_secret_str": "meow",
+        "my_secret_str_v1": "meow",
+        "person": {"name": "foo"},
+        "a_bool": True,
+        "a_none": None,
+        "a_str": "foo",
+        "a_str_nuc": "foo\x00",
+        "a_str_uc": "foo ⛰️",
+        "a_str_ucuc": "foo ⛰️\x00",
+        "a_str_ucucuc": "foo \\u26f0\\ufe0f",
+        "an_int": 1,
+        "a_float": 1.1,
+        "a_bytes": b"my bytes",
+        "a_bytearray": b"*",
+        "my_item": {
+            "namespace": ["a", "name", " "],
+            "key": "my-key",
+            "value": {},
+            "created_at": "2024-09-24T17:29:10.128397",
+            "updated_at": "2024-09-24T17:29:11.128397",
+        },
+    }
+
+
 def test_serde_jsonplus_bytes() -> None:
     serde = JsonPlusSerializer()