[Py]Better mock handling in serialization (#1063)

Avoid recursion
And rm locks

python/langsmith/client.py

@@ -227,10 +227,10 @@ def _simple_default(obj):
             return obj.pattern
         elif isinstance(obj, (bytes, bytearray)):
             return base64.b64encode(obj).decode()
-        return repr(obj)
+        return str(obj)
     except BaseException as e:
         logger.debug(f"Failed to serialize {type(obj)} to JSON: {e}")
-    return repr(obj)
+    return str(obj)
 
 
 def _serialize_json(obj: Any) -> Any:
@@ -250,9 +250,12 @@ def _serialize_json(obj: Any) -> Any:
             if hasattr(obj, attr) and callable(getattr(obj, attr)):
                 try:
                     method = getattr(obj, attr)
-                    return (
+                    response = (
                         method(exclude_none=exclude_none) if exclude_none else method()
                     )
+                    if not isinstance(response, dict):
+                        return str(response)
+                    return response
                 except Exception as e:
                     logger.error(
                         f"Failed to use {attr} to serialize {type(obj)} to"
@@ -262,7 +265,7 @@ def _serialize_json(obj: Any) -> Any:
         return _simple_default(obj)
     except BaseException as e:
         logger.debug(f"Failed to serialize {type(obj)} to JSON: {e}")
-        return repr(obj)
+        return str(obj)
 
 
 def _elide_surrogates(s: bytes) -> bytes:

python/langsmith/run_trees.py

@@ -28,7 +28,7 @@
 LANGSMITH_PREFIX = "langsmith-"
 LANGSMITH_DOTTED_ORDER = f"{LANGSMITH_PREFIX}trace"
 _CLIENT: Optional[Client] = None
-_LOCK = threading.Lock()
+_LOCK = threading.Lock()  # Keeping around for a while for backwards compat
 
 
 # Note, this is called directly by langchain. Do not remove.
@@ -37,9 +37,8 @@
 def get_cached_client(**init_kwargs: Any) -> Client:
     global _CLIENT
     if _CLIENT is None:
-        with _LOCK:
-            if _CLIENT is None:
-                _CLIENT = Client(**init_kwargs)
+        if _CLIENT is None:
+            _CLIENT = Client(**init_kwargs)
     return _CLIENT
 
 
@@ -163,7 +162,7 @@ def add_metadata(self, metadata: Dict[str, Any]) -> None:
         """Add metadata to the run."""
         if self.extra is None:
             self.extra = {}
-        metadata_: dict = self.extra.setdefault("metadata", {})
+        metadata_: dict = cast(dict, self.extra).setdefault("metadata", {})
         metadata_.update(metadata)
 
     def add_outputs(self, outputs: Dict[str, Any]) -> None:

python/langsmith/schemas.py

@@ -2,7 +2,6 @@
 
 from __future__ import annotations
 
-import threading
 from datetime import datetime, timedelta, timezone
 from decimal import Decimal
 from enum import Enum
@@ -201,6 +200,10 @@ class DatasetVersion(BaseModel):
     as_of: datetime
 
 
+def _default_extra():
+    return {"metadata": {}}
+
+
 class RunBase(BaseModel):
     """Base Run schema.
 
@@ -226,7 +229,7 @@ class RunBase(BaseModel):
     end_time: Optional[datetime] = None
     """End time of the run, if applicable."""
 
-    extra: Optional[dict] = None
+    extra: Optional[dict] = Field(default_factory=_default_extra)
     """Additional metadata or settings related to the run."""
 
     error: Optional[str] = None
@@ -258,16 +261,12 @@ class RunBase(BaseModel):
     """Attachments associated with the run.
     Each entry is a tuple of (mime_type, bytes)."""
 
-    _lock: threading.Lock = PrivateAttr(default_factory=threading.Lock)
-
     @property
     def metadata(self) -> dict[str, Any]:
         """Retrieve the metadata (if any)."""
-        with self._lock:
-            if self.extra is None:
-                self.extra = {}
-            metadata = self.extra.setdefault("metadata", {})
-        return metadata
+        if self.extra is None:
+            self.extra = {}
+        return self.extra.setdefault("metadata", {})
 
     @property
     def revision_id(self) -> Optional[UUID]:

python/pyproject.toml

@@ -1,6 +1,6 @@
 [tool.poetry]
 name = "langsmith"
-version = "0.1.130"
+version = "0.1.131"
 description = "Client library to connect to the LangSmith LLM Tracing and Evaluation Platform."
 authors = ["LangChain <support@langchain.dev>"]
 license = "MIT"

python/tests/unit_tests/test_client.py

@@ -828,8 +828,10 @@ class MyNamedTuple(NamedTuple):
         "fake_json": ClassWithFakeDict(),
         "some_set": set("a"),
         "set_with_class": set([MyClass(1)]),
+        "my_mock": MagicMock(text="Hello, world"),
     }
     res = orjson.loads(_dumps_json(to_serialize))
+
     expected = {
         "uid": str(uid),
         "time": current_time.isoformat(),
@@ -848,6 +850,7 @@ class MyNamedTuple(NamedTuple):
         "fake_json": {"foo": "bar"},
         "some_set": ["a"],
         "set_with_class": ["I fell back"],
+        "my_mock": lambda x: "Mock" in x,
     }
     assert set(expected) == set(res)
     for k, v in expected.items():