leak-snek 0.3.2

Leak Snek is a Python library that provides a flexible and extensible implementation of rate limiting for your applications

Leak Snek Rate Limiting

Leak Snek is a Python library that provides a flexible and extensible implementation of rate limiting for your applications

Table of Contents

• Features
• Getting Started
• License

Features

• Asynchronous and Synchronous Support: The library provides support for both asynchronous and synchronous rate.
• Flexible Rate Limiting Algorithms: Define your custom rate limiting algorithms by implementing the RateLimiter and AsyncRateLimiter interfaces.
• Rate Storage: Manage rate information with the RateStorage and AsyncRateStorage interfaces.
• Leaky Bucket Algorithm: Included implementations of the Leaky Bucket Algorithm for both asynchronous and synchronous use cases.

Getting Started

1. Installation:

   Install Leak Snek using pip:

   pip install leak_snek
   

2. Usage:

   You can use the provided classes and interfaces to implement rate limiting in your Python applications. Additionally, if you need the Leaky Bucket Algorithm, you can use the provided implementations.

   • Synchronous Leaky Bucket Algorithm:

     from threading import Lock
     
     from leak_snek.limiters.leaky_bucket import LeakyBucketLimiter
     from leak_snek.mutexes.memory_mutex import MemoryMutex
     from leak_snek.shortcuts.rate_limit import rl
     from leak_snek.storages.memory_storage import MemoryStorage
     
     limiter = LeakyBucketLimiter[str](
         rate_limit=rl("10/m"),
         rate_storage=MemoryStorage(),
         key_mutex=MemoryMutex(
             local_lock=Lock(),
             lock_factory=lambda: Lock(),
         ),
     )
     
     if not limiter.limit_exceeded("my_key"):
         ... # Perform the operation
     

   • Asynchronous Leaky Bucket Algorithm:

     from collections.abc import AsyncGenerator, Awaitable
     from contextlib import asynccontextmanager
     from typing import Any, Self, cast, override
     
     from redis.asyncio import Redis
     
     from leak_snek.interfaces.mutexes.aio.mutex import AsyncMutex
     from leak_snek.interfaces.storages.aio.rate_store import AsyncRateStorage
     from leak_snek.interfaces.values.rate import Rate
     from leak_snek.limiters.aio.leaky_bucket import AsyncLeakyBucketLimiter
     from leak_snek.shortcuts.rate_limit import rl
     
     
     @dataclasses.dataclass
     class RedisAsyncRateStorage(AsyncRateStorage[str]):
         redis: Redis
     
         async def read(self: Self, key: str) -> Rate:
             if not await self.redis.exists(key):
                 return Rate.default()
     
             rate_dict = await cast(Awaitable[dict[Any, Any]], self.redis.hgetall(name=key))
     
             return Rate(operations=int(rate_dict[b"operations"]), updated_at=float(rate_dict[b"updated_at"]))
     
         async def write(self: Self, key: str, value: Rate) -> None:
             await cast(
                 Awaitable[int],
                 self.redis.hset(name=key, mapping={"operations": value.operations, "updated_at": value.updated_at}),
             )
     
     
     @dataclasses.dataclass
     class RedisAsyncMutex(AsyncMutex[str]):
         redis: Redis
     
         @override
         @asynccontextmanager
         async def lock(self: Self, key: str) -> AsyncGenerator[None, None]:
             async with self.redis.lock(f"{key}_lock"):
                 yield
     
     
     async def main() -> None:
         async_limiter = AsyncLeakyBucketLimiter(
             rate_limit=rl("10/m"),
             rate_storage=RedisAsyncRateStorage(redis=Redis()),
             key_mutex=RedisAsyncMutex(redis=Redis()),
         )
     
         if not await async_limiter.limit_exceeded("my_key"):
             ...  # Perform the operation
          ```
     

In asynchronous example we implement our own async storage & mutex b/c leak_snek doesn't provide redis integration and it doesn't make much of a sense to implement asynchronous in-memory store.

License

Leak Snek Rate Limiting is released under the MIT License. See the LICENSE file for details.

---