Description:

multiexit 1.5.0

A better, saner and more useful atexit replacement for Python 3 that supports
multiprocessing.
Inspired by the following StackOverflow question and experience on building
multiprocessing daemons:
https://stackoverflow.com/q/2546276
multiexit will install a handler for the SIGTERM and SIGINT signals and
execute the registered exit functions in LIFO order (Last In First Out).
Exit functions can be registered so that only the calling process will call
them (the default), or as shared exit functions that will be called by the
calling process and all the children subprocesses that inherit it.

Install
multiexit is available for Python 3 from PyPI.
pip3 install multiexit


API
On the main process, before forking or creating any subprocess, call
multiexit.install:
def install(
signals=(signal.SIGTERM, signal.SIGINT),
except_hook=True,
)

signals:
Signals to install handler. Usually only SIGTERM and SIGINT are
required.

except_hook:
Also install a sys.excepthook that will call the exit functions in case of
an unexpected exception.


Then, for each exit function, on any subprocess, call multiexit.register:
register(func, shared=False)

func:
Exit function to register. Any callable without arguments.

shared:
If shared, the exit function will be called by the calling process but
also by all the children subprocesses that inherit it (thus the ones
created after registering it).
If shared is False, the default, only the calling process will execute
the exit function.




Example
from time import sleep
from signal import SIGTERM
from os import kill, getpid
from multiprocessing import Process

from multiexit import install, register, unregister


if __name__ == '__main__':

# Always call install() on the main process before creating any
# subprocess
#
# This installs a handler for SIGTERM and SIGINT. Subprocesses will
# inherit this handler. It also assigns the current PID as the master
# process, which will allow to choose between exit or os._exit calls
# when quitting.
install()

def _subproc1():

@register
def subproc1_clean():
print('Subprocess clean!')

sleep(1000)

# Register shared exit function so all subprocess call this
def shared_exit():
print('Shared exit being called by {} ...'.format(getpid()))

register(shared_exit, shared=True)

subproc1 = Process(
name='SubProcess1',
target=_subproc1,
)
# proc.daemon = True
# daemon means that signals (like SIGTERM) will be propagated automatically
# to children. Set to false (the default), to handle the SIGTERM
# (process.terminate()) to the children yourself.
subproc1.start()

# Register a cleaner using a decorator
@register
def clean_main():
print('Terminating child {}'.format(
subproc1.pid,
))
subproc1.terminate()
subproc1.join()
print('Child {} ended with {}'.format(
subproc1.pid,
subproc1.exitcode,
))

# Wait, and then kill main process
sleep(3)

# Suicide
kill(getpid(), SIGTERM)
For a more extensive example check out example.py.


Changes
1.5.0
A SIGINT handler is now installed by default to handle Ctrl+C. This
means that Python’s signal.default_int_handler is overridden, thus
Ctrl+C no longer raises KeyboardInterrupt. If this behavior is
undesired, pass signals=(signal.SIGTERM) when calling install().
Ctrl+C sends SIGINT to all processes in the terminal’s foreground
process group, so in order to behave like a SIGTERM shutdown flow, which
is sent only on the parent process, the handler will ignore SIGINT on
the children to allow the parent to decide what to do with the subprocesses
and their corresponding shutdown sequence.
This fixes issue #3


License
Copyright (C) 2018-2019 KuraLabs S.R.L

Licensed under the Apache License, Version 2.0 (the "License");
you may not use this file except in compliance with the License.
You may obtain a copy of the License at

http://www.apache.org/licenses/LICENSE-2.0

Unless required by applicable law or agreed to in writing,
software distributed under the License is distributed on an
"AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
KIND, either express or implied. See the License for the
specific language governing permissions and limitations
under the License.

License

For personal and professional use. You cannot resell or redistribute these repositories in their original state.

Share

• Share on Facebook
• Share on Twitter