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

AsyncPostgresSaver - unsupported Unicode escape sequence #248

New Issue
Closed
opened 2026-02-20 17:33:36 -05:00 by yindo · 1 comment
yindo commented 2026-02-20 17:33:36 -05:00
Owner
Copy Link
Copy Source

Originally created by @istvancsabakis on GitHub (Sep 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

from langgraph.checkpoint.postgres.aio import AsyncPostgresSaver
from langgraph.prebuilt import create_react_agent

from typing import Literal

from langchain_core.tools import tool
from langchain_openai import ChatOpenAI

@tool
def get_weather(city: Literal["nyc", "sf"]):
    """Use this to get weather information."""
    if city == "nyc":
        return "It might be cloudy in nyc"
    elif city == "sf":
        return "It's always sunny in sf"
    else:
        raise AssertionError("Unknown city")

tools = [get_weather]
model = ChatOpenAI(model_name="gpt-4o-mini")

async with AsyncPostgresSaver.from_conn_string(DB_URI) as checkpointer:
    graph = create_react_agent(model, tools=tools, checkpointer=checkpointer)
    config = {"configurable": {"thread_id": "6"}}
    res = await graph.ainvoke(
        {"messages": [("human", "what's the \x00 weather in nyc")]}, config
    )
    checkpoint_tuples = [c async for c in checkpointer.alist(config)]

Error Message and Stack Trace (if applicable)

UntranslatableCharacter                   Traceback (most recent call last)
Cell In[15], line 4
      2 graph = create_react_agent(model, tools=tools, checkpointer=checkpointer)
      3 config = {"configurable": {"thread_id": "6"}}
----> 4 res = await graph.ainvoke(
      5     {"messages": [("human", "what's the \x00 weather in nyc")]}, config
      6 )
      7 checkpoint_tuples = [c async for c in checkpointer.alist(config)]

File /venv/lib/python3.12/site-packages/langgraph/pregel/__init__.py:1592, in Pregel.ainvoke(self, input, config, stream_mode, output_keys, interrupt_before, interrupt_after, debug, **kwargs)
   1590 else:
   1591     chunks = []
-> 1592 async for chunk in self.astream(
   1593     input,
   1594     config,
   1595     stream_mode=stream_mode,
   1596     output_keys=output_keys,
   1597     interrupt_before=interrupt_before,
   1598     interrupt_after=interrupt_after,
   1599     debug=debug,
   1600     **kwargs,
   1601 ):
   1602     if stream_mode == "values":
   1603         latest = chunk

File /venv/lib/python3.12/site-packages/langgraph/pregel/__init__.py:1443, in Pregel.astream(self, input, config, stream_mode, output_keys, interrupt_before, interrupt_after, debug, subgraphs)
   1439 if "custom" in stream_modes:
   1440     config[CONF][CONFIG_KEY_STREAM_WRITER] = lambda c: stream.put_nowait(
   1441         ((), "custom", c)
   1442     )
-> 1443 async with AsyncPregelLoop(
   1444     input,
   1445     stream=StreamProtocol(stream.put_nowait, stream_modes),
   1446     config=config,
   1447     store=self.store,
   1448     checkpointer=checkpointer,
   1449     nodes=self.nodes,
   1450     specs=self.channels,
   1451     output_keys=output_keys,
   1452     stream_keys=self.stream_channels_asis,
   1453 ) as loop:
   1454     # create runner
   1455     runner = PregelRunner(
   1456         submit=loop.submit,
   1457         put_writes=loop.put_writes,
   1458         use_astream=do_stream is not None,
   1459     )
   1460     # enable subgraph streaming

File /venv/lib/python3.12/site-packages/langgraph/pregel/loop.py:883, in AsyncPregelLoop.__aexit__(self, exc_type, exc_value, traceback)
    876 async def __aexit__(
    877     self,
    878     exc_type: Optional[Type[BaseException]],
   (...)
    881 ) -> Optional[bool]:
    882     # unwind stack
--> 883     return await asyncio.shield(
    884         self.stack.__aexit__(exc_type, exc_value, traceback)
    885     )

File /usr/local/lib/python3.12/contextlib.py:754, in AsyncExitStack.__aexit__(self, *exc_details)
    750 try:
    751     # bare "raise exc_details[1]" replaces our carefully
    752     # set-up context
    753     fixed_ctx = exc_details[1].__context__
--> 754     raise exc_details[1]
    755 except BaseException:
    756     exc_details[1].__context__ = fixed_ctx

File /usr/local/lib/python3.12/contextlib.py:737, in AsyncExitStack.__aexit__(self, *exc_details)
    735     cb_suppress = cb(*exc_details)
    736 else:
--> 737     cb_suppress = await cb(*exc_details)
    739 if cb_suppress:
    740     suppressed_exc = True

File /venv/lib/python3.12/site-packages/langgraph/pregel/executor.py:179, in AsyncBackgroundExecutor.__aexit__(self, exc_type, exc_value, traceback)
    177 try:
    178     if exc := task.exception():
--> 179         raise exc
    180 except asyncio.CancelledError:
    181     pass

File /venv/lib/python3.12/site-packages/langgraph/pregel/loop.py:806, in AsyncPregelLoop._checkpointer_put_after_previous(self, prev, config, checkpoint, metadata, new_versions)
    804 try:
    805     if prev is not None:
--> 806         await prev
    807 finally:
    808     await cast(BaseCheckpointSaver, self.checkpointer).aput(
    809         config, checkpoint, metadata, new_versions
    810     )

File /venv/lib/python3.12/site-packages/langgraph/pregel/loop.py:806, in AsyncPregelLoop._checkpointer_put_after_previous(self, prev, config, checkpoint, metadata, new_versions)
    804 try:
    805     if prev is not None:
--> 806         await prev
    807 finally:
    808     await cast(BaseCheckpointSaver, self.checkpointer).aput(
    809         config, checkpoint, metadata, new_versions
    810     )

    [... skipping similar frames: AsyncPregelLoop._checkpointer_put_after_previous at line 806 (1 times)]

File /venv/lib/python3.12/site-packages/langgraph/pregel/loop.py:806, in AsyncPregelLoop._checkpointer_put_after_previous(self, prev, config, checkpoint, metadata, new_versions)
    804 try:
    805     if prev is not None:
--> 806         await prev
    807 finally:
    808     await cast(BaseCheckpointSaver, self.checkpointer).aput(
    809         config, checkpoint, metadata, new_versions
    810     )

File /venv/lib/python3.12/site-packages/langgraph/pregel/loop.py:808, in AsyncPregelLoop._checkpointer_put_after_previous(self, prev, config, checkpoint, metadata, new_versions)
    806         await prev
    807 finally:
--> 808     await cast(BaseCheckpointSaver, self.checkpointer).aput(
    809         config, checkpoint, metadata, new_versions
    810     )

File /venv/lib/python3.12/site-packages/langgraph/checkpoint/postgres/aio.py:267, in AsyncPostgresSaver.aput(self, config, checkpoint, metadata, new_versions)
    258 copy = checkpoint.copy()
    259 next_config = {
    260     "configurable": {
    261         "thread_id": thread_id,
   (...)
    264     }
    265 }
--> 267 async with self._cursor(pipeline=True) as cur:
    268     await cur.executemany(
    269         self.UPSERT_CHECKPOINT_BLOBS_SQL,
    270         await asyncio.to_thread(
   (...)
    276         ),
    277     )
    278     await cur.execute(
    279         self.UPSERT_CHECKPOINTS_SQL,
    280         (
   (...)
    287         ),
    288     )

File /usr/local/lib/python3.12/contextlib.py:217, in _AsyncGeneratorContextManager.__aexit__(self, typ, value, traceback)
    215 if typ is None:
    216     try:
--> 217         await anext(self.gen)
    218     except StopAsyncIteration:
    219         return False

File /venv/lib/python3.12/site-packages/langgraph/checkpoint/postgres/aio.py:340, in AsyncPostgresSaver._cursor(self, pipeline)
    336             await self.pipe.sync()
    337 elif pipeline:
    338     # a connection not in pipeline mode can only be used by one
    339     # thread/coroutine at a time, so we acquire a lock
--> 340     async with self.lock, conn.pipeline(), conn.cursor(
    341         binary=True, row_factory=dict_row
    342     ) as cur:
    343         yield cur
    344 else:

File /usr/local/lib/python3.12/contextlib.py:217, in _AsyncGeneratorContextManager.__aexit__(self, typ, value, traceback)
    215 if typ is None:
    216     try:
--> 217         await anext(self.gen)
    218     except StopAsyncIteration:
    219         return False

File /venv/lib/python3.12/site-packages/psycopg/connection_async.py:398, in AsyncConnection.pipeline(self)
    395         pipeline = self._pipeline = AsyncPipeline(self)
    397 try:
--> 398     async with pipeline:
    399         yield pipeline
    400 finally:

File /venv/lib/python3.12/site-packages/psycopg/_pipeline.py:266, in AsyncPipeline.__aexit__(self, exc_type, exc_val, exc_tb)
    264         logger.warning("error ignored terminating %r: %s", self, exc2)
    265     else:
--> 266         raise exc2.with_traceback(None)
    267 finally:
    268     self._exit(exc_val)

Description

UntranslatableCharacter Exception When Processing Input with Null Character (\x00) or ('x0000)

Description

I encountered an UntranslatableCharacter exception when attempting to process input containing a null character (\x00) using the langgraph library. The exception occurs when invoking the graph with a message that includes the null character.

To Reproduce

  1. Set up the environment:

    • Python version: 3.12
    • Ensure langgraph and its dependencies are installed (including psycopg for PostgreSQL interaction).

  2. Use the following code snippet:

    graph = create_react_agent(model, tools=tools, checkpointer=checkpointer)
config = {"configurable": {"thread_id": "6"}}
res = await graph.ainvoke(
    {"messages": [("human", "what's the \x00 weather in nyc")]}, config
)
checkpoint_tuples = [c async for c in checkpointer.alist(config)]

  3. Run the code and observe the exception.

Expected Behavior

The system should handle inputs containing null characters gracefully, either by processing the input correctly or by providing a clear and informative error message indicating that null characters are not supported.

Actual Behavior

An UntranslatableCharacter exception is raised, and the process fails with the following stack trace:

System Info

langgraph==0.2.26
langgraph-checkpoint==1.0.11
langgraph-checkpoint-postgres==1.0.8

langchain==0.3.0
langchain-community==0.3.0
langchain-core==0.3.5
langchain-elasticsearch==0.3.0
langchain-google-community==2.0.0
langchain-openai==0.2.0
langchain-text-splitters==0.3.0
langchain-unstructured==0.1.4
langchainhub==0.1.21

Originally created by @istvancsabakis on GitHub (Sep 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 from langgraph.checkpoint.postgres.aio import AsyncPostgresSaver from langgraph.prebuilt import create_react_agent from typing import Literal from langchain_core.tools import tool from langchain_openai import ChatOpenAI @tool def get_weather(city: Literal["nyc", "sf"]): """Use this to get weather information.""" if city == "nyc": return "It might be cloudy in nyc" elif city == "sf": return "It's always sunny in sf" else: raise AssertionError("Unknown city") tools = [get_weather] model = ChatOpenAI(model_name="gpt-4o-mini") async with AsyncPostgresSaver.from_conn_string(DB_URI) as checkpointer: graph = create_react_agent(model, tools=tools, checkpointer=checkpointer) config = {"configurable": {"thread_id": "6"}} res = await graph.ainvoke( {"messages": [("human", "what's the \x00 weather in nyc")]}, config ) checkpoint_tuples = [c async for c in checkpointer.alist(config)] ``` ### Error Message and Stack Trace (if applicable) ```shell UntranslatableCharacter Traceback (most recent call last) Cell In[15], line 4 2 graph = create_react_agent(model, tools=tools, checkpointer=checkpointer) 3 config = {"configurable": {"thread_id": "6"}} ----> 4 res = await graph.ainvoke( 5 {"messages": [("human", "what's the \x00 weather in nyc")]}, config 6 ) 7 checkpoint_tuples = [c async for c in checkpointer.alist(config)] File /venv/lib/python3.12/site-packages/langgraph/pregel/__init__.py:1592, in Pregel.ainvoke(self, input, config, stream_mode, output_keys, interrupt_before, interrupt_after, debug, **kwargs) 1590 else: 1591 chunks = [] -> 1592 async for chunk in self.astream( 1593 input, 1594 config, 1595 stream_mode=stream_mode, 1596 output_keys=output_keys, 1597 interrupt_before=interrupt_before, 1598 interrupt_after=interrupt_after, 1599 debug=debug, 1600 **kwargs, 1601 ): 1602 if stream_mode == "values": 1603 latest = chunk File /venv/lib/python3.12/site-packages/langgraph/pregel/__init__.py:1443, in Pregel.astream(self, input, config, stream_mode, output_keys, interrupt_before, interrupt_after, debug, subgraphs) 1439 if "custom" in stream_modes: 1440 config[CONF][CONFIG_KEY_STREAM_WRITER] = lambda c: stream.put_nowait( 1441 ((), "custom", c) 1442 ) -> 1443 async with AsyncPregelLoop( 1444 input, 1445 stream=StreamProtocol(stream.put_nowait, stream_modes), 1446 config=config, 1447 store=self.store, 1448 checkpointer=checkpointer, 1449 nodes=self.nodes, 1450 specs=self.channels, 1451 output_keys=output_keys, 1452 stream_keys=self.stream_channels_asis, 1453 ) as loop: 1454 # create runner 1455 runner = PregelRunner( 1456 submit=loop.submit, 1457 put_writes=loop.put_writes, 1458 use_astream=do_stream is not None, 1459 ) 1460 # enable subgraph streaming File /venv/lib/python3.12/site-packages/langgraph/pregel/loop.py:883, in AsyncPregelLoop.__aexit__(self, exc_type, exc_value, traceback) 876 async def __aexit__( 877 self, 878 exc_type: Optional[Type[BaseException]], (...) 881 ) -> Optional[bool]: 882 # unwind stack --> 883 return await asyncio.shield( 884 self.stack.__aexit__(exc_type, exc_value, traceback) 885 ) File /usr/local/lib/python3.12/contextlib.py:754, in AsyncExitStack.__aexit__(self, *exc_details) 750 try: 751 # bare "raise exc_details[1]" replaces our carefully 752 # set-up context 753 fixed_ctx = exc_details[1].__context__ --> 754 raise exc_details[1] 755 except BaseException: 756 exc_details[1].__context__ = fixed_ctx File /usr/local/lib/python3.12/contextlib.py:737, in AsyncExitStack.__aexit__(self, *exc_details) 735 cb_suppress = cb(*exc_details) 736 else: --> 737 cb_suppress = await cb(*exc_details) 739 if cb_suppress: 740 suppressed_exc = True File /venv/lib/python3.12/site-packages/langgraph/pregel/executor.py:179, in AsyncBackgroundExecutor.__aexit__(self, exc_type, exc_value, traceback) 177 try: 178 if exc := task.exception(): --> 179 raise exc 180 except asyncio.CancelledError: 181 pass File /venv/lib/python3.12/site-packages/langgraph/pregel/loop.py:806, in AsyncPregelLoop._checkpointer_put_after_previous(self, prev, config, checkpoint, metadata, new_versions) 804 try: 805 if prev is not None: --> 806 await prev 807 finally: 808 await cast(BaseCheckpointSaver, self.checkpointer).aput( 809 config, checkpoint, metadata, new_versions 810 ) File /venv/lib/python3.12/site-packages/langgraph/pregel/loop.py:806, in AsyncPregelLoop._checkpointer_put_after_previous(self, prev, config, checkpoint, metadata, new_versions) 804 try: 805 if prev is not None: --> 806 await prev 807 finally: 808 await cast(BaseCheckpointSaver, self.checkpointer).aput( 809 config, checkpoint, metadata, new_versions 810 ) [... skipping similar frames: AsyncPregelLoop._checkpointer_put_after_previous at line 806 (1 times)] File /venv/lib/python3.12/site-packages/langgraph/pregel/loop.py:806, in AsyncPregelLoop._checkpointer_put_after_previous(self, prev, config, checkpoint, metadata, new_versions) 804 try: 805 if prev is not None: --> 806 await prev 807 finally: 808 await cast(BaseCheckpointSaver, self.checkpointer).aput( 809 config, checkpoint, metadata, new_versions 810 ) File /venv/lib/python3.12/site-packages/langgraph/pregel/loop.py:808, in AsyncPregelLoop._checkpointer_put_after_previous(self, prev, config, checkpoint, metadata, new_versions) 806 await prev 807 finally: --> 808 await cast(BaseCheckpointSaver, self.checkpointer).aput( 809 config, checkpoint, metadata, new_versions 810 ) File /venv/lib/python3.12/site-packages/langgraph/checkpoint/postgres/aio.py:267, in AsyncPostgresSaver.aput(self, config, checkpoint, metadata, new_versions) 258 copy = checkpoint.copy() 259 next_config = { 260 "configurable": { 261 "thread_id": thread_id, (...) 264 } 265 } --> 267 async with self._cursor(pipeline=True) as cur: 268 await cur.executemany( 269 self.UPSERT_CHECKPOINT_BLOBS_SQL, 270 await asyncio.to_thread( (...) 276 ), 277 ) 278 await cur.execute( 279 self.UPSERT_CHECKPOINTS_SQL, 280 ( (...) 287 ), 288 ) File /usr/local/lib/python3.12/contextlib.py:217, in _AsyncGeneratorContextManager.__aexit__(self, typ, value, traceback) 215 if typ is None: 216 try: --> 217 await anext(self.gen) 218 except StopAsyncIteration: 219 return False File /venv/lib/python3.12/site-packages/langgraph/checkpoint/postgres/aio.py:340, in AsyncPostgresSaver._cursor(self, pipeline) 336 await self.pipe.sync() 337 elif pipeline: 338 # a connection not in pipeline mode can only be used by one 339 # thread/coroutine at a time, so we acquire a lock --> 340 async with self.lock, conn.pipeline(), conn.cursor( 341 binary=True, row_factory=dict_row 342 ) as cur: 343 yield cur 344 else: File /usr/local/lib/python3.12/contextlib.py:217, in _AsyncGeneratorContextManager.__aexit__(self, typ, value, traceback) 215 if typ is None: 216 try: --> 217 await anext(self.gen) 218 except StopAsyncIteration: 219 return False File /venv/lib/python3.12/site-packages/psycopg/connection_async.py:398, in AsyncConnection.pipeline(self) 395 pipeline = self._pipeline = AsyncPipeline(self) 397 try: --> 398 async with pipeline: 399 yield pipeline 400 finally: File /venv/lib/python3.12/site-packages/psycopg/_pipeline.py:266, in AsyncPipeline.__aexit__(self, exc_type, exc_val, exc_tb) 264 logger.warning("error ignored terminating %r: %s", self, exc2) 265 else: --> 266 raise exc2.with_traceback(None) 267 finally: 268 self._exit(exc_val) ``` ### Description # UntranslatableCharacter Exception When Processing Input with Null Character (`\x00`) or (`'x0000`) ## Description I encountered an `UntranslatableCharacter` exception when attempting to process input containing a null character (`\x00`) using the `langgraph` library. The exception occurs when invoking the graph with a message that includes the null character. ## To Reproduce 1. **Set up the environment:** - Python version: 3.12 - Ensure `langgraph` and its dependencies are installed (including `psycopg` for PostgreSQL interaction). 2. **Use the following code snippet:** ```python graph = create_react_agent(model, tools=tools, checkpointer=checkpointer) config = {"configurable": {"thread_id": "6"}} res = await graph.ainvoke( {"messages": [("human", "what's the \x00 weather in nyc")]}, config ) checkpoint_tuples = [c async for c in checkpointer.alist(config)] ``` 3. **Run the code and observe the exception.** ## Expected Behavior The system should handle inputs containing null characters gracefully, either by processing the input correctly or by providing a clear and informative error message indicating that null characters are not supported. ## Actual Behavior An `UntranslatableCharacter` exception is raised, and the process fails with the following stack trace: ### System Info langgraph==0.2.26 langgraph-checkpoint==1.0.11 langgraph-checkpoint-postgres==1.0.8 langchain==0.3.0 langchain-community==0.3.0 langchain-core==0.3.5 langchain-elasticsearch==0.3.0 langchain-google-community==2.0.0 langchain-openai==0.2.0 langchain-text-splitters==0.3.0 langchain-unstructured==0.1.4 langchainhub==0.1.21
yindo closed this issue 2026-02-20 17:33:36 -05:00
yindo commented 2026-02-20 17:33:37 -05:00
Author
Owner
Copy Link
Copy Source

@vbarda commented on GitHub (Sep 27, 2024):

@istvancsabakis thanks for reporting -- this is fixed in langgraph-checkpoint-postgres==1.0.9. please let me know if you run into any other issues!

@vbarda commented on GitHub (Sep 27, 2024): @istvancsabakis thanks for reporting -- this is fixed in `langgraph-checkpoint-postgres==1.0.9`. please let me know if you run into any other issues!
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#248
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?