AssertionError: invalid pydantic.BaseModel when using pydantic.BaseModel subclass in spectree.Response

[kenichi-cg]

Describe the bug

When specifying a subclass of pydantic.BaseModel in spectree.Response, an AssertionError: invalid pydantic.BaseModel is raised. This issue seems to be caused by the assertion at line 87 in spectree/response.py, which checks the model using issubclass(model, BaseModel). Despite importing BaseModel from spectree._pydantic, the assertion does not use the is_base_model() function from spectree._pydantic, which would correctly identify pydantic.BaseModel subclasses.

To Reproduce

1. Create a Flask application with pydantic and spectree.
2. Define a response model using pydantic.BaseModel.
3. Use the response model in spectree.Response.
4. Run the Flask application and make a request to the endpoint.

Example code:

from flask import Flask, jsonify
from pydantic import BaseModel, Field
from spectree import Response, SpecTree

class MyResponse(BaseModel):
    message: str = Field(..., description="response message")

app = Flask(__name__)
spec = SpecTree('flask')

@app.route('/hello', methods=['GET'])
@spec.validate(resp=Response(HTTP_200=MyResponse))
def hello():
    return jsonify(message='hello world')

if __name__ == '__main__':
    spec.register(app)
    app.run(debug=True)

Expected behavior

The response model should be validated correctly without raising an AssertionError.

The spectree version

• SpecTree: 1.2.11
• Flask: 3.0.2
• pydantic: 2.9.2

Additional context

The issue seems to be caused by the assertion at line 87 in spectree/response.py, which checks the model using issubclass(model, BaseModel). A potential fix could be to adjust the assertion to use the is_base_model() function from spectree._pydantic to correctly identify pydantic.BaseModel subclasses.

[kemingy]

Related to #329

So far, it only support v1, let me check if we can support both.

[kemingy]

Hi @kenichi-cg you can try it with v1.4.1. It should work now.