expli

Explicit dataclasses with automatic serialization

expli (short for "explicit") enhances Python dataclasses with automatic dictionary and JSON serialization methods, making it effortless to work with nested dataclasses, optional types, and lists.

Installation

pip install expli

Quick Start

from expli import edataclass

@edataclass
class Person:
    name: str
    age: int
    email: str | None = None

# Automatic methods are added!
person = Person("Alice", 30, "alice@example.com")

# Convert to dict
data = person.to_dict()

# Convert to JSON
json_str = person.to_json(indent=2)

# Create from dict
person2 = Person.from_dict(data)

# Create from JSON
person3 = Person.from_json(json_str)

# Create a copy/clone
person4 = person.copy()
person5 = person.clone()  # Same as copy()

Features

🎯 Enhanced Dataclass Decorator

The @edataclass decorator automatically adds six methods to your dataclass:

• to_dict() - Convert instance to dictionary
• from_dict(data) - Create instance from dictionary (class method)
• to_json(indent=None) - Convert instance to JSON string
• from_json(json_str) - Create instance from JSON string (class method)
• copy() - Create a copy of the instance
• clone() - Create a clone of the instance (alias for copy())

🔄 Full Recursive Support

expli handles complex nested structures automatically:

from expli import edataclass

@edataclass
class Address:
    street: str
    city: str
    country: str

@edataclass
class Company:
    name: str
    address: Address

@edataclass
class Person:
    name: str
    age: int
    company: Company | None
    hobbies: list[str]

person = Person(
    name="Alice",
    age=30,
    company=Company(
        name="Tech Corp",
        address=Address("123 Main St", "Boston", "USA")
    ),
    hobbies=["reading", "coding"]
)

# Everything serializes recursively
data = person.to_dict()
# {
#     "name": "Alice",
#     "age": 30,
#     "company": {
#         "name": "Tech Corp",
#         "address": {
#             "street": "123 Main St",
#             "city": "Boston",
#             "country": "USA"
#         }
#     },
#     "hobbies": ["reading", "coding"]
# }

# And deserializes back perfectly
person2 = Person.from_dict(data)

# Create a copy with all nested objects copied
person_copy = person.copy()

🔄 Copying and Cloning

Create independent copies of your dataclass instances:

from expli import edataclass

@edataclass
class Person:
    name: str
    age: int
    hobbies: list[str]

person = Person("Alice", 30, ["reading", "coding"])

# Create a copy
person_copy = person.copy()
# or use clone() - they're the same
person_clone = person.clone()

# The copy is independent
person_copy.hobbies.append("gaming")
print(person.hobbies)       # ["reading", "coding"]
print(person_copy.hobbies)  # ["reading", "coding", "gaming"]

✨ Supported Types

• Primitive types: str, int, float, bool, etc.
• Optional types: Type | None
• Lists: list[Type]
• Nested dataclasses: Any dataclass as a field
• Lists of dataclasses: list[DataclassType]
• Optional lists: list[Type] | None

🛠️ Dataclass Parameters

@edataclass supports all standard dataclass parameters:

from expli import edataclass

@edataclass(frozen=True, order=True)
class Config:
    api_key: str
    timeout: int = 30

API Reference

@edataclass

Enhanced dataclass decorator that adds serialization methods.

Parameters: Same as @dataclass (frozen, order, etc.)

Added Methods:

• to_dict(self) -> dict - Convert instance to dictionary
• from_dict(cls, data: dict) -> Self - Create instance from dictionary
• to_json(self, indent=None) -> str - Convert instance to JSON string
• from_json(cls, json_str: str) -> Self - Create instance from JSON string
• copy(self) -> Self - Create a copy of the instance
• clone(self) -> Self - Create a clone of the instance (alias for copy())

to_dict(obj) / easdict(obj)

Standalone function to convert a dataclass instance to a dictionary.

from expli import to_dict, easdict
from dataclasses import dataclass

@dataclass
class Point:
    x: int
    y: int

point = Point(10, 20)
data = to_dict(point)  # or easdict(point)
# {"x": 10, "y": 20}

from_dict(cls, data) / efromdict(cls, data)

Standalone function to create a dataclass instance from a dictionary.

from expli import from_dict, efromdict
from dataclasses import dataclass

@dataclass
class Point:
    x: int
    y: int

data = {"x": 10, "y": 20}
point = from_dict(Point, data)  # or efromdict(Point, data)

Why expli?

Standard dataclasses don't provide built-in serialization for nested structures. While dataclasses.asdict() exists, it doesn't handle deserialization, and neither function is added to your class for convenient access.

expli solves this by:

• ✅ Adding methods directly to your dataclass
• ✅ Handling nested dataclasses recursively
• ✅ Supporting optional types and lists
• ✅ Providing both dict and JSON serialization
• ✅ Working seamlessly with type hints
• ✅ Easy copying and cloning of instances

Comparison

Without expli:

from dataclasses import dataclass, asdict
import json

@dataclass
class Person:
    name: str
    age: int

person = Person("Alice", 30)

# Manual serialization
data = asdict(person)
json_str = json.dumps(data)

# Manual deserialization (you have to write this!)
def from_dict(data):
    return Person(**data)

person2 = from_dict(json.loads(json_str))

With expli:

from expli import edataclass

@edataclass
class Person:
    name: str
    age: int

person = Person("Alice", 30)

# Built-in methods
data = person.to_dict()
json_str = person.to_json()
person2 = Person.from_dict(data)
person3 = Person.from_json(json_str)
person4 = person.copy()

License

MIT

Contributing

Contributions are welcome! Please feel free to submit a Pull Request.