Skip to main content

A python package that provides useful locks

Project description


Documentation status Downloads Latest version

Cross platform locks for threads and processes.

🔩 Install

pip install fasteners

🔩 Usage

Lock for processes has the same API as the threading.Lock for threads:

import fasteners
import threading

lock = threading.Lock()                                 # for threads
lock = fasteners.InterProcessLock('path/to/lock.file')  # for processes

with lock:
    ... # exclusive access

# or alternatively    

... # exclusive access

Reader Writer lock has a similar API, which is the same for threads or processes:

import fasteners

rw_lock = fasteners.ReaderWriterLock()                                 # for threads
rw_lock = fasteners.InterProcessReaderWriterLock('path/to/lock.file')  # for processes

with rw_lock.write_lock():
    ... # write access

with rw_lock.read_lock():
    ... # read access

# or alternatively

... # read access

... # write access

🔩 Overview

Python standard library provides a lock for threads (both a reentrant one, and a non-reentrant one, see below). Fasteners extends this, and provides a lock for processes, as well as Reader Writer locks for both threads and processes.

The specifics of the locks are as follows:

Process locks

The fasteners.InterProcessLock uses fcntl on Unix-like systems and msvc _locking on Windows. As a result, if used cross-platform it guarantees an intersection of their features:

lock reentrant mandatory

The fasteners.InterProcessReaderWriterLock also uses fcntl on Unix-like systems and LockFileEx on Windows. Their features are as follows:

lock reentrant mandatory upgradable preference
fcntl reader
LockFileEx reader
fasteners.InterProcessReaderWriterLock reader

Thread locks

Fasteners do not provide a simple thread lock, but for the sake of comparison note that the threading module provides both a reentrant and non-reentrant locks:

lock reentrant mandatory

The fasteners.ReaderWriterLock at the moment is as follows:

lock reentrant mandatory upgradable preference
fasteners.ReaderWriterLock reader

🔩 Glossary

To learn more about the various aspects of locks, check the wikipedia pages for locks and readers writer locks as well as the resources listed in the documentation. Here we briefly mention the main notions used above.

  • Lock - a mechanism that prevents two or more threads or processes from running the same code at the same time.
  • Readers writer lock - a mechanism that prevents two or more threads from having write (or write and read) access, while allowing multiple readers.
  • Reentrant lock - a lock that can be acquired (and then released) multiple times, as in:
with lock:
    with lock:
        ... # some code
  • Mandatory lock (as opposed to advisory lock) - a lock that is enforced by the operating system, rather than by the cooperation between threads or processes
  • Upgradable readers writer lock - a readers writer lock that can be upgraded from reader to writer (or downgraded from writer to reader) without losing the lock that is already held, as in:
with rw_lock.read_lock():
    ... # read access
    with rw_lock.write_lock():
        ... # write access
    ... # read access
  • Readers writer lock preference - describes the behaviour when multiple threads or processes are waiting for access. Some of the patterns are:
    • Reader preference - If lock is held by readers, then new readers will get immediate access. This can result in writers waiting forever (writer starvation)
    • Writer preference - If writer is waiting for a lock, then all the new readers (and writers) will be queued after the writer.
    • Phase fair - Lock that alternates between readers and writers.

Project details

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

fasteners-0.17.3.tar.gz (18.0 kB view hashes)

Uploaded source

Built Distribution

fasteners-0.17.3-py3-none-any.whl (18.4 kB view hashes)

Uploaded py3

Supported by

AWS AWS Cloud computing Datadog Datadog Monitoring Facebook / Instagram Facebook / Instagram PSF Sponsor Fastly Fastly CDN Google Google Object Storage and Download Analytics Huawei Huawei PSF Sponsor Microsoft Microsoft PSF Sponsor NVIDIA NVIDIA PSF Sponsor Pingdom Pingdom Monitoring Salesforce Salesforce PSF Sponsor Sentry Sentry Error logging StatusPage StatusPage Status page