python 41 lines · 6 steps

JSON round-tripping with Python dataclasses

A nested dataclass serializes itself to JSON and rebuilds from it, handling datetimes and nested objects along the way.

Explained by highlit
Intermediate dataclasses serialization json nested-objects classmethod
1import json
2from dataclasses import dataclass, field, asdict
3from datetime import datetime
4from typing import Any
5 
6 
7@dataclass
8class Address:
9 street: str
10 city: str
11 zip_code: str
12 
13 
14@dataclass
15class User:
16 id: int
17 name: str
18 email: str
19 address: Address
20 tags: list[str] = field(default_factory=list)
21 created_at: datetime = field(default_factory=datetime.utcnow)
22 
23 def to_json(self, **kwargs: Any) -> str:
24 def default(obj: Any) -> Any:
25 if isinstance(obj, datetime):
26 return obj.isoformat()
27 raise TypeError(f"Cannot serialize {type(obj).__name__}")
28 
29 return json.dumps(asdict(self), default=default, **kwargs)
30 
31 @classmethod
32 def from_json(cls, raw: str) -> "User":
33 data = json.loads(raw)
34 return cls(
35 id=data["id"],
36 name=data["name"],
37 email=data["email"],
38 address=Address(**data["address"]),
39 tags=data.get("tags", []),
40 created_at=datetime.fromisoformat(data["created_at"]),
41 )
01 / 01
STEP 01

Walkthrough

Space play ←→ step click any line
Three takeaways
  1. 1asdict recursively flattens nested dataclasses into plain dicts ready for json.dumps.
  2. 2json.dumps accepts a default callback to teach it how to serialize types it doesn't natively understand, like datetime.
  3. 3Reconstruction is manual because JSON has no type info, so nested objects and datetimes must be rebuilt field by field.

Related explainers

python 
from datetime import datetime
from fastapi import APIRouter, Depends, HTTPException, status
from pydantic import BaseModel, ConfigDict, EmailStr
from sqlalchemy.orm import Session

Building a users router in FastAPI

routing serialization dependency-injection
Intermediate 7 steps
rust 
use std::fs::File;
use std::io::{self, BufWriter};
use std::path::Path;
 

Serializing config to JSON in Rust

serialization serde file-io
Intermediate 7 steps
python 
import re
 
from django.core.exceptions import ValidationError
from django.core.validators import RegexValidator

Building a deconstructible validator in Django

validation regex deconstructible
Intermediate 7 steps
ruby 
class ArticleSerializer < ActiveModel::Serializer
  attributes :id, :title, :slug, :excerpt, :reading_time, :published_at, :url
 
  belongs_to :author, serializer: UserSerializer

Shaping API output with ActiveModel::Serializer in Rails

serialization json api computed attributes
Intermediate 9 steps
python 
from dataclasses import dataclass
from typing import Optional
 
from sqlalchemy import select, func

Building a pagination helper with SQLAlchemy

pagination dataclass computed-properties
Intermediate 8 steps
python 
from datetime import timedelta
 
from django.contrib.sessions.models import Session
from django.core.management.base import BaseCommand

A batched session-cleanup command in Django

management-command batch-processing database-cleanup
Intermediate 7 steps

Share this explainer

Here's the card — post it anywhere.

JSON round-tripping with Python dataclasses — share card
Made with highlit — turn any snippet into a walkthrough like this in about a minute.
Explain your code