schema_json fail with Pydantic BaseModel in langchain-core==0.3.0.dev4

[liuhetian]

Checked other resources

• [X] I added a very descriptive title to this issue.
• [X] I searched the 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 LangChain rather than my code.
• [X] The bug is not resolved by updating to the latest stable version of LangChain (or the specific integration package).

Example Code

from langchain_openai import ChatOpenAI
model_chat = ChatOpenAI(api_key='foo')
print(model_chat.get_output_schema().schema_json()[:5])  # ok 

from pydantic import BaseModel
from typing import Any
class Foo(BaseModel):
    bar: Any
model_chat.with_structured_output(Foo, method='json_schema', strict=True).get_output_schema().schema_json()  
# PydanticUserError

Error Message and Stack Trace (if applicable)

PydanticUserError: _oai_structured_outputs_parser_output is not fully defined; you should define PydanticBaseModel, then call _oai_structured_outputs_parser_output.model_rebuild().

Description

meet this question when using langserve add_routes(app, mychain, path='/faq')

System Info

langchain-core==0.3.0.dev4 langchain-openai==0.2.0.dev2

pydantic==2.9.1 pydantic-extra-types==2.5.0 pydantic-settings==2.2.1 pydantic_core==2.23.3

python=3.12

[eyurtsev]

After fix it doesn't crash, but also doesn't return the expected schema, so the fix is only partial.

Users will still need to overwrite it using with_types (which is generally recommended)

I'll double check, but believe this is actually not a regression, but an existing issue in 0.2x as well