packify

name	packify
description	Serialize Python data structures to/from bytes for reliable storage, transmission, and API integration. Handles complex nested data, custom objects, and ensures deterministic serialization. Use when users need to save, send, or convert data — whether they mention serialization, packing, or simply need to store/transfer complex Python structures.

Packify

Universal serialization for Python data structures to/from bytes.

When to Use

• Serialize complex, nested data structures for storage or transmission
• Support custom types via the Packable protocol
• Ensure deterministic, version-independent serialization

Quick Start

import packify

# Serialize to bytes
packed: bytes = packify.pack(data)

# Deserialize from bytes
data = packify.unpack(packed, inject={})  # inject provides custom types

Supported Types

Built-in: int, bool, float, Decimal, str, bytes, bytearray, NoneType, list, tuple, set, dict

All types can be nested arbitrarily.

Custom Types

Implement the Packable protocol:

from dataclasses import dataclass
import packify

@dataclass
class User:
    name: str
    id: int

    def pack(self) -> bytes:
        return packify.pack({'name': self.name, 'id': self.id})

    @classmethod
    def unpack(cls, data: bytes, inject: dict = {}):
        fields = packify.unpack(data, inject=inject)
        return cls(**fields)

users = [User('Alice', 1), User('Bob', 2)]
packed = packify.pack(users)
unpacked = packify.unpack(packed, inject={'User': User})
# Or: unpacked = packify.unpack(packed, inject={**globals()})

Key Behaviors

• Deterministic: Dicts/sets sorted for consistent output
• Type-safe: Exact types preserved (tuple vs list, etc.)
• Dependency injection: Custom types required in inject when unpacking
• Error handling: Raises UsageError for invalid types/missing dependencies