vellum-ai 0.13.2__tar.gz → 0.13.3__tar.gz

{vellum_ai-0.13.2 → vellum_ai-0.13.3}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: vellum-ai
-Version: 0.13.2
+Version: 0.13.3
 Summary: 
 License: MIT
 Requires-Python: >=3.9,<4.0
@@ -20,7 +20,7 @@ Classifier: Programming Language :: Python :: 3.12
 Classifier: Programming Language :: Python :: 3.8
 Classifier: Topic :: Software Development :: Libraries :: Python Modules
 Classifier: Typing :: Typed
-Requires-Dist: Jinja2 (==3.1.2)
+Requires-Dist: Jinja2 (>=3.1.2,<4.0.0)
 Requires-Dist: cdktf (>=0.20.5,<0.21.0)
 Requires-Dist: click (==8.1.7)
 Requires-Dist: docker (==7.1.0)

{vellum_ai-0.13.2 → vellum_ai-0.13.3}/pyproject.toml

@@ -1,6 +1,6 @@
 [tool.poetry]
 name = "vellum-ai"
-version = "0.13.2"
+version = "0.13.3"
 description = ""
 readme = "README.md"
 authors = []
@@ -35,7 +35,7 @@ Repository = 'https://github.com/vellum-ai/vellum-python-sdks'
 
 [tool.poetry.dependencies]
 python = "^3.9"
-Jinja2 = "3.1.2"
+Jinja2 = "^3.1.2"
 cdktf = "^0.20.5"
 click = "8.1.7"
 docker = "7.1.0"

{vellum_ai-0.13.2 → vellum_ai-0.13.3}/src/vellum/client/core/client_wrapper.py

@@ -18,7 +18,7 @@ class BaseClientWrapper:
         headers: typing.Dict[str, str] = {
             "X-Fern-Language": "Python",
             "X-Fern-SDK-Name": "vellum-ai",
-            "X-Fern-SDK-Version": "0.13.2",
+            "X-Fern-SDK-Version": "0.13.3",
         }
         headers["X_API_KEY"] = self.api_key
         return headers

{vellum_ai-0.13.2 → vellum_ai-0.13.3}/src/vellum/workflows/nodes/core/templating_node/node.py

@@ -1,5 +1,7 @@
 import json
-from typing import Any, Callable, ClassVar, Dict, Generic, Mapping, Tuple, Type, TypeVar, Union, get_args
+from typing import Any, Callable, ClassVar, Dict, Generic, Mapping, Tuple, Type, TypeVar, Union, get_args, get_origin
+
+from pydantic import BaseModel
 
 from vellum.utils.templating.constants import DEFAULT_JINJA_CUSTOM_FILTERS, DEFAULT_JINJA_GLOBALS
 from vellum.utils.templating.exceptions import JinjaTemplateError
@@ -71,6 +73,7 @@ class TemplatingNode(BaseNode[StateType], Generic[StateType, _OutputType], metac
         original_base = get_original_base(self.__class__)
         all_args = get_args(original_base)
 
+        output_type: Any
         if len(all_args) < 2 or isinstance(all_args[1], TypeVar):
             output_type = str
         else:
@@ -88,12 +91,35 @@ class TemplatingNode(BaseNode[StateType], Generic[StateType, _OutputType], metac
         if output_type is bool:
             return bool(rendered_template)
 
+        if get_origin(output_type) is list:
+            try:
+                data = json.loads(rendered_template)
+            except json.JSONDecodeError:
+                raise ValueError("Invalid JSON Array format for rendered_template")
+
+            if not isinstance(data, list):
+                raise ValueError(f"Expected a list of items for rendered_template, received {data.__class__.__name__}")
+
+            inner_type = get_args(output_type)[0]
+            if issubclass(inner_type, BaseModel):
+                return [inner_type.model_validate(item) for item in data]
+            else:
+                return data
+
         if output_type is Json:
             try:
                 return json.loads(rendered_template)
             except json.JSONDecodeError:
                 raise ValueError("Invalid JSON format for rendered_template")
 
+        if issubclass(output_type, BaseModel):
+            try:
+                data = json.loads(rendered_template)
+            except json.JSONDecodeError:
+                raise ValueError("Invalid JSON format for rendered_template")
+
+            return output_type.model_validate(data)
+
         raise ValueError(f"Unsupported output type: {output_type}")
 
     def run(self) -> Outputs:

{vellum_ai-0.13.2 → vellum_ai-0.13.3}/src/vellum/workflows/nodes/core/templating_node/tests/test_templating_node.py

@@ -1,5 +1,7 @@
 import json
+from typing import List
 
+from vellum.client.types.chat_message import ChatMessage
 from vellum.client.types.function_call import FunctionCall
 from vellum.workflows.nodes.bases.base import BaseNode
 from vellum.workflows.nodes.core.templating_node.node import TemplatingNode
@@ -125,3 +127,31 @@ def test_templating_node__pydantic_to_json():
 
     # THEN the output is the expected JSON
     assert outputs.result == {"name": "test", "arguments": {"key": "value"}, "id": None}
+
+
+def test_templating_node__chat_history_output():
+    # GIVEN a templating node that outputs a chat history
+    class ChatHistoryTemplateNode(TemplatingNode[BaseState, List[ChatMessage]]):
+        template = '[{"role": "USER", "text": "Hello"}]'
+        inputs = {}
+
+    # WHEN the node is run
+    node = ChatHistoryTemplateNode()
+    outputs = node.run()
+
+    # THEN the output is the expected chat history
+    assert outputs.result == [ChatMessage(role="USER", text="Hello")]
+
+
+def test_templating_node__function_call_output():
+    # GIVEN a templating node that outputs a function call
+    class FunctionCallTemplateNode(TemplatingNode[BaseState, FunctionCall]):
+        template = '{"name": "test", "arguments": {"key": "value"}}'
+        inputs = {}
+
+    # WHEN the node is run
+    node = FunctionCallTemplateNode()
+    outputs = node.run()
+
+    # THEN the output is the expected function call
+    assert outputs.result == FunctionCall(name="test", arguments={"key": "value"})