Python Flask json 模式验证

Question

I am using the flask framework in my new project. It would get JSON data from the post and send a JSON response. So, I need to validate my JSON request. I have seen couple of libraries. But those libraries are not working as expected. Finally, I have decided to go with a flask-jsonschema-validator. It is working fine with a single JSON object. If the request object has a nested object, it is not working. For example,

from flask_jsonschema_validator import JSONSchemaValidator
JSONSchemaValidator(app=app, root="schemas")

This is my initialization of the validator.

# If any error occurred in the request.
@app.errorhandler(jsonschema.ValidationError)
def json_validation_error(e):
    return json_response("error", str(e), {})

This is my error handler

@app.validate('model', 'save') def save_model():

This is my implementation.

{
  "save": {
    "type": "object",
    "properties": {
      "workspace": {"type": "object"},
      "name": {"type": "string"},
      "description": {"type": "string"},
      "uri": {"type": "string"},
      "type": {
         "name": {"type": "string"},
      }
    },
    "required": [ "workspace", "name", "description", "uri", "type"]
  }
}

This is my model.json file. It is validating the request except for the "type". How to apply validation for JSON request with nested object

Please help anyone to fix this problem

Thanks in advance.

Answer 1

flask-expects-json package checks variables types on nested objects.

It work as a decorator on your route.

SCHEMA = {
    "type": "object",
    "properties": {
      "workspace": {"type": "object"},
      "name": {"type": "string"},
      "description": {"type": "string"},
      "uri": {"type": "string"},
      "type": {
         "type": "object",
         "properties": {
              "name": {"type": "string"},
         }
      }
    },
    "required": ["workspace", "name", "description", "uri", "type"]
}

@expects_json(SCHEMA)
def my_route(self, **kwargs):
    pass

Answer 2

If the validation of complex nested objects, I would recommend an alternative tool to JSONSchema. I can replicate your schema validation in GoodJSON like below. As you can see, formulating a validation schema is just composing a bunch of self-contained validator functions that can be applied to objects, lists and primitive values.

from goodjson.validators import is_dict, is_string, is_uri, foreach_key


validate_fun = foreach_key(
    save=[foreach_key(
        workspace=[is_dict],
        name=[is_string],
        description=[is_string],
        uri=[is_uri],
        type=[foreach_key(
            name=[is_string]
        )]
    )]
)

validate_fun(YOUR_JSON_DATA_OBJECT)

Disclaimer: I am the author of GoodJSON.