Skip to main content

SuperLoops package simplifies and augments usage of Python threads. It provides support for thread maintenance, events, failure handling, health status propagation, and graceful termination.

Project description

This library is currently being beta-tested. See something that's broken? Did we get something wrong? Create an issue and let us know!

SuperLoops logo

pip install superloops

SuperLoops package simplifies and augments usage of Python threads.

Features:

  • Startup, shutdown, hard reset
  • Thread events
  • Co-dependant thread health propagation
  • 0 dependencies
  • 100% automated test coverage
class ProcessLoop(SuperLoop):
    def cycle(self):
        # process stuff in a separate thread

loop = ProcessLoop()
loop.start()
# ProcessLoop_0: Started 

loop.stop()
# ProcessLoop_0: Exited gracefully

loop.hard_reset() # when nothing else helps 😬 

Thread events?

class ApiFeedLoop(SuperLoop):
    def on_start(self):
        self.api = Api(key=my_key)
        
    def on_stop(self):
        self.api.disconnect()
        
    def cycle(self):
        self.api.get_feed()

    # on_thread_start, on_thread_stop...

Multiple threads that need to stay healthy?

loop_controller = LoopController()

process_loop = loop_controller.new_loop(ProcessLoop())
api_feed_loop = loop_controller.new_loop(ApiFeedLoop())

loop_controller.start()
# LoopController_0: Started

loop_controller.maintain_loops()
# ProcessLoop_0: Started
# ApiFeedLoop_0: Started

api_feed_loop.failure() # oops!

# LoopController_0: Stopping loops.
# ProcessLoop_0: Exited gracefully
# ApiFeedLoop_0: Exited gracefully

# LoopController_0: Restarting loops.
# ProcessLoop_1: Started
# ApiFeedLoop_1: Started

In summary, SuperLoops provide support for thread maintenance, events, failure handling, health status propagation, and graceful termination.

How does SuperLoops work?

SuperLoop is a class that wraps around a Python threading.Thread object. It exposes an interface for thread starting, stopping, hard restarts, graceful termination and events: on_start, on_stop, on_thread_start and on_thread_stop.

Each time you restart a SuperLoop, it will create a new Thread, handling naming and graceful termination for you.

Aided by the LoopController class, the SuperLoops are able to communicate their health between each other. This ensures that should one SuperLoop fail and need restarting, all other connected SuperLoops would be restarted too.

Documentation

To use a SuperLoop, declare a class inheriting from SuperLoop and extend the cycle method

from superloops import SuperLoop

class MyLoop(SuperLoop):
    def cycle(self):
        pass
        # process stuff

loop = MyLoop()

You can use the following methods of SuperLoop to control the thread lifecycle.

loop.start()
loop.stop()
loop.hard_reset()
loop.failure()

start()

Start a new thread (unless one is already started) that will be used to operate the loop. This will start calling the overridden cycle() method indefinitely on the new thread. Any arguments passed to this method will be passed to the on_start callback.

stop()

Stop the existing thread (unless there isn't one started) and join the thread, waiting up to the amount of seconds specified by the grace_period argument of this class. Any arguments passed to this method will be passed to the on_stop callback.

hard_reset()

Stop the thread by calling stop() method and mark it as killed, attempting to gracefully finish as soon as the control is returned from the cycle() method. Independently of whether the current thread stops gracefully, a new thread will be instantly started.

failure()

Indicate that there has been a critical failure in the operation of the thread. If the amount of failures exceeds max_loop_failures specified as the argument of this class, the thread will stop and (if provided) unset its GreenLight.

That indicates that the health status should be propagated across other threads managed through the LoopController that this loop belongs to, and that all specified threads should be restarted.

SuperLoop arguments

  • green_light (threading.Event): A threading.Event object representing the health state of the loop. It gets set automatically when a loop is added to LoopController.
  • grace_period (int): The number of seconds to wait when stopping the loop gracefully. Default is 5 seconds.
  • max_loop_failures (int): The maximum number of failures allowed before reporting issues. Default is 10 failures.
  • stop_on_failure (bool): A flag that indicates if this loop should be stopped when it exceeds its max_loop_failures. Default is False.
  • reset_globally (bool): A flag that indicates if this loop should be reset when other loops report issues. Default is True.

Events

SuperLoop provides lifecycle event callbacks that facilitate flexibility in managing the loop and its thread.

on_start

The on_start callback is invoked before a new thread is created and started. It can be used to perform any setup that is required before the loop starts running.

This method must return a boolean indicating whether the loop should continue starting.

on_stop

The on_stop callback is invoked after the loop's thread is stopped. This method can be used to perform any cleanup that is required after the loop has stopped running.

on_thread_start

The on_thread_start callback is invoked from within the loop's thread before the loop starts running. This method can be used to perform any setup that must be done within the context of the loop's thread.

on_thread_stop

The on_thread_stop callback is invoked from within the loop's thread after the loop has stopped running. This method can be used to perform any cleanup that must be done within the context of the loop's thread.

Example:

def thread_name():
    return threading.current_thread().name

class MyLoop(SuperLoop):
    def on_start(self):
        print(f'on_start - {thread_name()}')
        # Perform any necessary setup here
        return True  # Return False to prevent the loop from starting
    
    def on_stop(self):
        print(f'on_stop - {thread_name()}')
        # Perform any necessary cleanup here

    def on_thread_start(self):
        print(f'on_thread_start - {thread_name()}')
        # Perform any necessary setup here

    def on_thread_stop(self):
        print(f'on_thread_stop - {thread_name()}')
        # Perform any necessary cleanup here
    
    def cycle(self):
        pass

loop = MyLoop()

loop.start()
# on_start - MainThread
# on_thread_start - MyLoop_0
loop.stop()
# on_stop - MainThread
# on_thread_stop - MyLoop_0

Loop Controller

SuperLoops are built with the intent of being able to link multiple threads into a single co-dependant system. If one thread in such system fails, all threads can be restarted allowing for an easy way to ensure multiple sub-systems can recover from a failure.

This is executed by adding various SuperLoops to a LoopController class, calling the LoopController.new_loop() method, passing the SuperLoop as the argument.

LoopController needs to be started, which will launch a separate thread observing for the health status of all the threads.

loop_controller = LoopController()

process_loop = loop_controller.new_loop(ProcessLoop())
api_feed_loop = loop_controller.new_loop(ApiFeedLoop())

loop_controller.start()

For convenience, LoopController can also ensure that all its SuperLoops are started, by calling maintain_loops, or stopped by calling stop_loops.

loop_controller.maintain_loops()
loop_controller.stop_loops()

LoopController arguments

  • reset_callback (callable): A callable to be executed when the LoopController resets loops.
  • green_light (threading.Event): A threading.Event that will be used to control the health status of the loops. Creates one if not provided.

Examples

See Usage examples for more.

Licence

See LICENSE

Disclaimer

SuperLoops is provided on an AS IS and AS AVAILABLE basis without any representation or endorsement made and without warranty of any kind whether express or implied, including but not limited to the implied warranties of satisfactory quality, fitness for a particular purpose, non-infringement, compatibility, security and accuracy. To the extent permitted by law, SuperLoops' authors will not be liable for any indirect or consequential loss or damage whatever (including without limitation loss of business, opportunity, data, profits) arising out of or in connection with the use of SuperLoops. SuperLoops' authors make no warranty that the functionality of SuperLoops will be uninterrupted or error free, that defects will be corrected or that SuperLoops or the server that makes it available are free of viruses or anything else which may be harmful or destructive.

Built by Voy

Hi! Thanks for checking out and using this library. If you are interested in discussing your project, require mentorship, consider hiring me, or just wanna chat - I'm happy to talk.

You can send me an email to get in touch: hello@voyzan.com

Or if you'd just want to give something back, I've got a Buy Me A Coffee account:

Buy Me A Coffee

Thanks and have an awesome day 👋

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

superloops-0.1.3.tar.gz (18.7 kB view hashes)

Uploaded Source

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page