RequestValidationError：JSON 对象必须是 str、bytes 或 bytearray (type=type_error.json)

Question

我的 FastAPI 应用程序中有一个端点用于接收表单数据：

@router.post("/foobar")
async def handler(
    form_data: Bar = Depends(Bar.as_form),
) -> JSONResponse:
    ...

我想做的是在pydantic 的帮助下验证表单数据。以下是模型：

from fastapi import Form
from pydantic import BaseModel, Json


class Foo(BaseModel):
    a: str


class Bar(BaseModel):
    any_field: Optional[List[Foo]]

    @classmethod
    def as_form(
        cls,
        any_field: Json[List[Foo]] = Form(None, media_type="application/json"),
    ) -> "Bar":
        return cls(any_field=any_field)

但我收到以下错误：

fastapi.exceptions.RequestValidationError: 1 validation error for Request
body -> any_field
  JSON object must be str, bytes or bytearray (type=type_error.json)

我已经为 RequestValidationError 添加了异常处理程序以确保 any_field 实际上是 str 类型：

@application.exception_handler(RequestValidationError)
async def validation_exception_handler(request: Request, exc: RequestValidationError):
    body: FormData = exc.body

    return JSONResponse(
        content={"msg": str([type(v) for v in body.values()])},
        status_code=status.HTTP_422_UNPROCESSABLE_ENTITY
    )

这是我的要求：

如您所见，它是str。

为什么会出现这个错误？

Answer 1

您可以让 FastAPI 自己进行验证。例如...

from fastapi  import FastAPI, Form, File, UploadFile
from pydantic import BaseModel, Json
from typing   import Optional, List

class Foo(BaseModel):
    a: str

class Bar(BaseModel):
    any_field: Optional[List[Foo]] = Form()

app = FastAPI()

@app.post("/foobar")
async def handler(upload: UploadFile = File(), form_data: Json[Bar] = Form()):
    return {'form_data': form_data, 'upload': upload.filename}

这是说明端点使用的代码...

import requests
import json

reqUrl = "http://127.0.0.1:8000/foobar"

with open("/Users/pelletier/Playground/scores.csv", "rb") as upload:
    post_files = {'upload': upload}
    payload    = {'form_data': json.dumps({'any_field': [{'a': 'foo'}, {'a': 'foo'}]})}
    response   = requests.post(reqUrl, data=payload, files=post_files)
    print(response.text)