pydantic object __init()__ function call revealed in fastAPI type_error payload

[Syffe]

First Check

• I added a very descriptive title to this issue.
• I used the GitHub search to find a similar issue and didn't find it.
• I searched the FastAPI documentation, with the integrated search.
• I already searched in Google "How to X in FastAPI" and didn't find any information.
• I already read and followed all the tutorial in the docs and didn't find an answer.
• I already checked if it is not related to FastAPI but to Pydantic.
• I already checked if it is not related to FastAPI but to Swagger UI.
• I already checked if it is not related to FastAPI but to ReDoc.

Commit to Help

• I commit to help with one of those options 👆

Example Code

from fastapi import FastAPI
import pydantic

app = FastAPI()

@pydantic.dataclasses.dataclass
class TestPayload:
    message: str = dataclasses.field(
        metadata={"description": "test message"}
    )

@app.put('/test_endpoint')
def create(test_payload: TestPayload) -> TestPayload:
    return test_payload


payload = {"false_key": "test message"}

app.put("/test_endpoint",
        json=payload,
      )

Description

Hi,

When trying to use a pydantic model as a parameter for an endpoint, if this said payload isn't matching the pydantic model defined, with, in this case, a wrong key, fastAPI will return an error payload as follows:

# fastAPI error payload returned
{
'loc': ('body',), 
'msg': "TestPayload.__init__() got an unexpected keyword argument 'false_key'", 
'type': 'type_error'
}

The error message reveals that .init() is called from the pydantic object, and this kind of message is not quiet what an API user would like to see following a key error.
The question then being, Is it the good behavior following those kind of errors ? Is it ok to reveal this kind of call information ?
May be it is a bug, or an error that isn't correctly catched on the server side

Operating System

macOS

Operating System Details

No response

FastAPI Version

0.74.1

Python Version

Python 3.10.2

Additional Context

No response

[harunyasar]

If you ask me pydantic dataclass is not directly related to FastAPI's API. I mean when you pass {"false_key": "test message"}, it will be interpreted like def __init__(message: str) by dataclass. So, in your example, you are simply passing like as follows.

test_payload = TestPayload(false_key="test message")

This means you are passing an argument with the wrong name to the initialization method.

[sileht]

The issue here is that FastAPI returns a different error if pydantic.dataclasses is used instead pydantic.BaseModel.

pydantic.BaseModel returns:

[
     {'loc': ['body', 'reason'], 'msg': 'field required', 'type': 'value_error.missing'}
]

while dataclasses variant returns:

[{
    'loc': ('body',), 
    'msg': "TestPayload.__init__() got an unexpected keyword argument 'false_key'", 
    'type': 'type_error'
}]

The dataclasses versions leak implementation details, which is a bug for me as dataclasses variant is officially supported: https://fastapi.tiangolo.com/advanced/dataclasses/