langchain-ai/langgraph
1
0
Fork 0
mirror of https://github.com/langchain-ai/langgraph.git synced 2026-07-15 12:35:43 -04:00
Code Issues 314 Packages Projects Releases 17 Wiki Activity

ToolNode output is broken for non-english #213

New Issue
Closed
opened 2026-02-20 17:31:57 -05:00 by yindo · 2 comments
yindo commented 2026-02-20 17:31:57 -05:00
Owner
Copy Link
Copy Source

Originally created by @sangmandu on GitHub (Aug 27, 2024).

Checked other resources

  • I added a very descriptive title to this issue.
  • I searched the LangGraph/LangChain documentation with the integrated search.
  • I used the GitHub search to find a similar question and didn't find it.
  • I am sure that this is a bug in LangGraph/LangChain rather than my code.
  • I am sure this is better as an issue rather than a GitHub discussion, since this is a LangGraph bug and not a design question.

Example Code

langgraph.prebuilt.tool_node.py

def str_output(output: Any) -> str:
    if isinstance(output, str):
        return output
    else:
        try:
            return json.dumps(output)
        except Exception:
            return str(output)

Error Message and Stack Trace (if applicable)

No response

Description

I found the tool node output is encoded.
And when I check _run_one and _arun_one function, they use str_output function.
And I checked the arg, the input is alreadt encoded before return str or json.dumps.
I don't know about that well...

this is not completely suitable code but I can solve my problem temporarily

def str_output(output: Any) -> str:
    return json.dumps(json.loads(output), ensure_ascii=False)

And I think the origin problem presents here: langchain_core.tools.base

def _stringify(content: Any) -> str:
    try:
        return json.dumps(content)
    except Exception:
        return str(content)

they just got a dumps without ensure_ascii=False.

Thank you for reading and I am forwarding to fixing this quickly

System Info

Just simple error

Originally created by @sangmandu on GitHub (Aug 27, 2024). ### Checked other resources - [X] I added a very descriptive title to this issue. - [X] I searched the [LangGraph](https://langchain-ai.github.io/langgraph/)/LangChain documentation with the integrated search. - [X] I used the GitHub search to find a similar question and didn't find it. - [X] I am sure that this is a bug in LangGraph/LangChain rather than my code. - [X] I am sure this is better as an issue [rather than a GitHub discussion](https://github.com/langchain-ai/langgraph/discussions/new/choose), since this is a LangGraph bug and not a design question. ### Example Code ```python langgraph.prebuilt.tool_node.py def str_output(output: Any) -> str: if isinstance(output, str): return output else: try: return json.dumps(output) except Exception: return str(output) ``` ### Error Message and Stack Trace (if applicable) _No response_ ### Description I found the tool node output is encoded. And when I check _run_one and _arun_one function, they use str_output function. And I checked the arg, the input is alreadt encoded before return str or json.dumps. I don't know about that well... this is not completely suitable code but I can solve my problem temporarily ```python def str_output(output: Any) -> str: return json.dumps(json.loads(output), ensure_ascii=False) ``` And I think the origin problem presents here: langchain_core.tools.base ```python def _stringify(content: Any) -> str: try: return json.dumps(content) except Exception: return str(content) ``` they just got a dumps without ensure_ascii=False. Thank you for reading and I am forwarding to fixing this quickly ### System Info Just simple error
yindo closed this issue 2026-02-20 17:31:58 -05:00
yindo commented 2026-02-20 17:31:59 -05:00
Author
Owner
Copy Link
Copy Source

@gbaian10 commented on GitHub (Aug 28, 2024):

Test Case

from enum import Enum

from dotenv import load_dotenv
from langchain.pydantic_v1 import BaseModel
from langchain_core.tools import tool
from langchain_openai import ChatOpenAI
from langgraph.graph import MessagesState, StateGraph
from langgraph.prebuilt import ToolNode

load_dotenv()


class Day(str, Enum):
    MONDAY = "星期一"
    TUESDAY = "星期二"
    WEDNESDAY = "水曜日"
    THURSDAY = "목요일"
    FRIDAY = "星期五"
    SATURDAY = "星期六"
    SUNDAY = "星期日"


class Week(BaseModel):
    """choose days"""

    days: list[Day]


@tool(args_schema=Week)
def get_day_list(days: list[Day]) -> list[Day]:
    return days


tools = [get_day_list]
model = ChatOpenAI(model="gpt-4o-mini").bind_tools(tools)

workflow = StateGraph(MessagesState)
workflow.add_node("enter", lambda state: {"messages": model.invoke(state["messages"])})
workflow.add_node("tool", ToolNode(tools))

workflow.set_entry_point("enter")
workflow.add_edge("enter", "tool")
workflow.set_finish_point("tool")
app = workflow.compile()
outputs = app.invoke(
    {"messages": "我星期一和 Friday and 水曜日 和 목요일 have free time"}
)
print(outputs["messages"][-1].content)

Output

["\u661f\u671f\u4e00", "\u661f\u671f\u4e94", "\u6c34\u66dc\u65e5", "\ubaa9\uc694\uc77c"]

Expected Output.

["星期一", "星期五", "水曜日", "목요일"]
@gbaian10 commented on GitHub (Aug 28, 2024): Test Case ```py from enum import Enum from dotenv import load_dotenv from langchain.pydantic_v1 import BaseModel from langchain_core.tools import tool from langchain_openai import ChatOpenAI from langgraph.graph import MessagesState, StateGraph from langgraph.prebuilt import ToolNode load_dotenv() class Day(str, Enum): MONDAY = "星期一" TUESDAY = "星期二" WEDNESDAY = "水曜日" THURSDAY = "목요일" FRIDAY = "星期五" SATURDAY = "星期六" SUNDAY = "星期日" class Week(BaseModel): """choose days""" days: list[Day] @tool(args_schema=Week) def get_day_list(days: list[Day]) -> list[Day]: return days tools = [get_day_list] model = ChatOpenAI(model="gpt-4o-mini").bind_tools(tools) workflow = StateGraph(MessagesState) workflow.add_node("enter", lambda state: {"messages": model.invoke(state["messages"])}) workflow.add_node("tool", ToolNode(tools)) workflow.set_entry_point("enter") workflow.add_edge("enter", "tool") workflow.set_finish_point("tool") app = workflow.compile() outputs = app.invoke( {"messages": "我星期一和 Friday and 水曜日 和 목요일 have free time"} ) print(outputs["messages"][-1].content) ``` Output ```shell ["\u661f\u671f\u4e00", "\u661f\u671f\u4e94", "\u6c34\u66dc\u65e5", "\ubaa9\uc694\uc77c"] ``` Expected Output. ```shell ["星期一", "星期五", "水曜日", "목요일"] ```
yindo commented 2026-02-20 17:32:00 -05:00
Author
Owner
Copy Link
Copy Source

@eyurtsev commented on GitHub (Sep 11, 2024):

Will merge a fix in core, but users should be able to return strings directly from the tool to avoid relying on any auto-conversion of Enum values to strings (or anything more complex).

@eyurtsev commented on GitHub (Sep 11, 2024): Will merge a fix in core, but users should be able to return strings directly from the tool to avoid relying on any auto-conversion of Enum values to strings (or anything more complex).
Sign in to join this conversation.
No Branch/Tag Specified
Branches Tags
Labels
Clear labels
agents
awaiting author revision
bug
change
chat models
checkpointer
ci
code quality
code quality
dependencies
documentation
drawing
duplicate
enhancement
good first issue
help wanted
invalid
maintainer
meta
missing information
not maintained by langchain
open-swe-auto
open-swe-max
open-swe-max-auto
pending
performance
pull-request
Mirrored from GitHub Pull Request
 question
requires design
sdk
server
stale
studio
tests
type checking
usage question
v0.5
v1.1
No Label
Milestone
No items
No Milestone
Projects
Clear projects
No project
Assignees
Clear assignees
yindo
No Assignees
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: langchain-ai/langgraph#213
Delete Branch "%!s()"

Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?