IronFence Python Library Documentation

Introduction

IronFence is a Python library that provides two types of synchronization primitives: Mutex and RWLock. These objects allow you to protect shared data from concurrent access in an asynchronous environment. By using Mutex and RWLock, you can ensure that your application remains coroutine-safe and avoids race conditions¹.

Installation

When pip is available, the distribution can be downloaded from PyPI and installed in one step:

pip install ironfence

Mutex

A Mutex (ashort for "mutual exclusion") is a synchronization primitive that allows only one coroutine to access a shared resource at a time. Other coroutines that attempt to acquire the lock will block until the current owner releases it.

import asyncio

import ironfence

shared_state = {}
mu = ironfence.Mutex(shared_state)


async def modify_state():
    print("modifying state..")
    async with mu.lock() as value:
        await asyncio.sleep(1)
        value["key"] = "example"
        print("State modified!")


async def read_state():
    print("reading state..")
    async with mu.lock() as value:
        print("State is:", value)


# run concurrently
asyncio.get_event_loop().run_until_complete(
    asyncio.gather(read_state(), modify_state(), read_state())
)

In this example, we create a Mutex instance named mu that guards a shared dictionary called shared_state. We define two coroutines, modify_state() and read_state(), that need to access shared_state. To ensure exclusive access, we use the lock() method of the Mutex object to acquire the lock before modifying or reading the shared data. The lock() method returns a context manager that automatically releases the lock when exiting the scope.

RWLock

An RWLock (short for "reader-writer lock") is a synchronization primitive that allows multiple reader coroutines to access a shared resource simultaneously, while allowing only one writer coroutine to modify the resource at a time. This makes it more efficient than a Mutex in situations where there are many reader coroutines and only occasional writes.

import asyncio

import ironfence


async def read_state(rw_lock):
    async with rw_lock.read() as value:
        print(value)


async def modify_state(rw_lock):
    async with rw_lock.write() as value:
        value.append("example")


async def main():
    shared_state = []
    rw_lock = ironfence.RWLock(shared_state)
    await asyncio.gather(
        read_state(rw_lock), modify_state(rw_lock), read_state(rw_lock)
    )


asyncio.get_event_loop().run_until_complete(main())

In this example, we create an RWLock instance named rw_lock that guards a shared list called shared_state. We define two coroutines, read_state() and modify_state(), that need to access shared_state. To ensure exclusive access, we use the read() and write() methods of the RWLock object to acquire the appropriate locks. The read() method acquires a read lock, which allows multiple reader coroutines to enter, while the write() method acquires a write lock, which blocks all reader coroutines until the write operation completes.

IronFence doesn't protect against inter-process or intra-thread races. Use additional sync mechanisms like multiprocessing.Lock or threading.Lock for added protection. ↩

hqdncw/ironfence

IronFence Python Library Documentation

Introduction

Installation

Mutex

RWLock

Footnotes