Pythonic way to read JSON list into python class

Question

Consider the following code

import json


class Address(object):
    def __init__(self, street, number):
        self.street = street
        self.number = number

    def __str__(self):
        return "{0} {1}".format(self.street, self.number)


class User(object):
    address: List[Address] = []

    def __init__(self, name, address: List):
        self.name = name
        for adr in address:
            self.address.append(Address(*adr)) # is this needed?

    def __str__(self):
        return "{0} ,{1}".format(self.name, self.address)


if __name__ == '__main__':
    js = '''{
    "name": "Cristian",
    "address": [{
            "street": "Sesame",
            "number": 122
        },
        {
            "street": "Sesame",
            "number": 122
        }
    ]
}'''
    j = json.loads(js)
    print(j)
    u = User(**j)
    print(u.name)
    print(u.address[0].number)

my question is, for this type of JSON where we have a list of values. Do we need to loop through the list? Or is there a more pythonic way to simply fill this object

address: List[Address] = []

from the loaded json?

https://stackoverflow.com/questions/1305532/convert-nested-python-dict-to-object — Mahrkeenerh, Oct 16 '21 at 19:30
Do you really need the class `Address`? By the way it should be `Address(**adr))` since you are unpacking a dictionary — cards, Oct 16 '21 at 19:59
@cards no adr is a list not a dict the first json called j in the code is a dict. — Kinder9211, Oct 17 '21 at 14:20

John Byro · Accepted Answer · 2021-10-17T14:56:41.530

You can prefer dataclasses, json dataclasses in python

from typing import List
import json

from dataclasses import dataclass, field
from dataclasses_json import dataclass_json


class Address(object):
    def __init__(self, street, number):
        self.street = street
        self.number = number

    def __str__(self):
        return "{0} {1}".format(self.street, self.number)


@dataclass_json
@dataclass
class User(object):
    address: List[Address] = field(default_factory=lambda: [])
    name: str = None

    def __str__(self):
        return "{0} ,{1}".format(self.name, self.address)


if __name__ == '__main__':
    js = '''{
    "name": "Cristian",
    "address": [{
            "street": "Sesame",
            "number": 122
        },
        {
            "street": "Sesame",
            "number": 122
        }
    ]
    }'''

    data = User.from_dict(json.loads(js))
    print(data)

If you use __str__ in User it will print

Cristian ,[{'street': 'Sesame', 'number': 122}, {'street': 'Sesame', 'number': 122}]

else if you omit __str__ in User it will print

User(address=[{'street': 'Sesame', 'number': 122}, {'street': 'Sesame', 'number': 122}], name='Cristian')

Address class can be avoided

could you give me an exmaple of how dataclasses make it more pythonic?. — Kinder9211, Oct 17 '21 at 14:20

score 0 · Answer 2 · edited Oct 16 '21 at 19:44

0

I don't know if I fully understand your question. But if the problem is that you have a loop in the __init__ method of the class you could solve it with a list comprehension. Something like:

self.address = [Address(*adr) for adr in address]

We now have still the loop, but is a bit cleaner

edited Oct 16 '21 at 19:44

buran

13,682
10
36
61

answered Oct 16 '21 at 19:43

Sharm

3
3

rv.kvetch · Answer 3 · 2021-10-20T17:04:11.823

I would also propose the dataclass-wizard as an alternative when working with dataclasses; which works similarly to the accepted solution, but does not use the marshmallow library to generate schemas, for example.

from __future__ import annotations

from dataclasses import dataclass, field
from typing import List

from dataclass_wizard import JSONWizard


@dataclass
class User(JSONWizard):
    address: List[Address] = field(default_factory=list)
    name: str = None

    def __str__(self):
        return "{0} ,{1}".format(self.name, self.address)


@dataclass
class Address:
    street: str
    number: int

    def __str__(self):
        return "{0} {1}".format(self.street, self.number)


if __name__ == '__main__':
    js = '''{
    "name": "Cristian",
    "address": [{
            "street": "Sesame",
            "number": 122
        },
        {
            "street": "Sesame",
            "number": 122
        }
    ]
    }'''

    data = User.from_json(js)
    print(data)

Output:

Cristian ,[Address(street='Sesame', number=122), Address(street='Sesame', number=122)]

Do note, that I've added a __future__ import, primarily to support forward-declared annotations. If you want, you can remove that and instead declare the annotation for the address field, i.e. like List['Address'].

Pythonic way to read JSON list into python class

3 Answers3