redis-json-serializer 0.1.0

Fast and secure JSON serializer for Redis caching with Pydantic and dataclass support

redis-json-serializer

Fast and secure JSON serializer for Redis caching with Pydantic and dataclass support.

Features

• 🚀 Fast: Built on top of orjson for maximum performance
• 🔒 Secure: Whitelist-based model registration prevents arbitrary code execution
• 📦 Type-safe: Full support for Pydantic models and dataclasses
• 🔧 Extensible: Plugin-based architecture for custom types
• 🎯 Simple: Clean API with minimal boilerplate

Installation

From PyPI (when published)

pip install redis-json-serializer

From source (development)

# Clone repository
git clone [e-mail]:seligoroff/redis-json-serializer.git
cd redis-json-serializer

# Install in development mode
pip install -e ".[dev,all]"

# Or using requirements files
pip install -r requirements-dev.txt

With optional dependencies

# With Pydantic support
pip install redis-json-serializer[pydantic]

# With MongoDB ObjectId support
pip install redis-json-serializer[mongodb]

# With aiocache integration
pip install redis-json-serializer[aiocache]

# All optional dependencies
pip install redis-json-serializer[all]

Quick Start

Basic Usage

from redis_json_serializer import JsonSerializer

serializer = JsonSerializer()

# Serialize
data = {"name": "John", "age": 30, "tags": {"python", "redis"}}
packed = serializer.pack(data)
json_bytes = serializer.dumps(data)

# Deserialize
unpacked = serializer.unpack(serializer.loads(json_bytes))

With Pydantic Models

from redis_json_serializer import JsonSerializer, register_model
from pydantic import BaseModel

@register_model("user.v1")  # Stable alias for versioning
class User(BaseModel):
    id: str
    name: str
    email: str

serializer = JsonSerializer()

user = User(id="123", name="Alice", email="[e-mail]")
packed = serializer.pack(user)
unpacked = serializer.unpack(packed)  # Returns User instance

With aiocache

from aiocache import cached
from aiocache.serializers import JsonSerializer
from redis_json_serializer import JsonSerializer as RedisJsonSerializer

# Configure aiocache to use our serializer
from aiocache import caches
caches.set_config({
    "default": {
        "cache": "aiocache.RedisCache",
        "serializer": {
            "class": "redis_json_serializer.aiocache.AiocacheJsonSerializer",
        },
    }
})

@cached(ttl=3600)
async def get_user(user_id: str) -> User:
    # Your logic here
    return User(id=user_id, name="Alice", email="[e-mail]")

Architecture

The library uses a dispatch-table approach for O(1) type lookup instead of chain-of-responsibility pattern, providing better performance and simpler code.

Supported Types

• Native JSON: str, int, float, bool, None
• Dates: datetime.datetime, datetime.date
• Collections: set, list, tuple, dict
• Pydantic models: With registration via @register_model()
• Dataclasses: With registration via @register_model()
• Custom types: Decimal, ObjectId (MongoDB)

Versioning

The library supports format versioning through namespaces:

serializer = JsonSerializer(namespace="cache:v2:")

This allows safe migration between format versions without clearing the entire Redis cache.

License

MIT

Contributing

Contributions are welcome! Please read our contributing guidelines first.