Validation of nested dictionary's value type

Question

I check the python data in marshmallow, the food field is a list, and the list stores the dict. I can only do this step. In fact, I want to verify that the value in the dict is int instead of string. How should I modify my code?

from marshmallow import Schema, fields, pprint

class UserSchema(Schema):
    name = fields.Str()
    food = fields.List(fields.Dict)

user_data = {
    "name": "Ken",
    "food": [{'apple': 2, 'banana': 4}, {'apple': '2', 'banana': '4'}]
}

result = UserSchema().load(user_data)
pprint(result)

Answer 1

Code below should validate your data pattern.
Look at comments in code, it's simple explanation.

from marshmallow import ValidationError, Schema, fields, pprint # Added `ValidationError`.

user_data = {
    "name": "Ken",
    "food": [
        {'apple': 2, 'banana': 4}, 
        {'apple': '2', 'banana': '4'}
    ]
} # `user_data` is as is without changes.

# New function for validation your dictionary.
def validate_value_type(dict):
    for key in dict:
        if type(dict[key]) is not int:
            raise ValidationError(f'Type of {key}\'s value is not Integer')

# `UserSchema` class with changes.
class UserSchema(Schema):
    name = fields.Str()
    food = fields.List(fields.Dict(validate=validate_value_type)) # Added validation function.

# Wrapped `load` in `try/catch`.
try:
    UserSchema().load(user_data)
except ValidationError as err:
    pprint(err.messages)

Above code output:

{'food': {1: ["Type of apple's value is not Integer"]}}