multiprocessing process pool again

Final implementation is quite different and much more smart:

And here is the source code:

	#################################################################
	# $HeadURL$
	#################################################################
	""" :mod: ProcessPool
	=================
	.. module: ProcessPool
	:synopsis: ProcessPool and related classes
	ProcessPool
	-----------
	ProcessPool creates a pool of worker subprocesses to handle a queue of tasks
	much like the producers/consumers paradigm. Users just need to fill the queue
	with tasks to be executed and worker tasks will execute them.
	To construct ProcessPool one first should call its contructor::
	pool = ProcessPool( minSize, maxSize, maxQueuedRequests )
	where parameters are::
	:param int minSize: at least <minSize> workers will be alive all the time
	:param int maxSize: no more than <maxSize> workers will be alive all the time
	:param int maxQueuedRequests: size for request waiting in a queue to be executed
	In case another request is added to the full queue, the execution will
	lock until another request is taken out. The ProcessPool will automatically increase and
	decrease the pool of workers as needed, of course not exceeding above limits.
	To add a task to the queue one should execute::
	pool.createAndQueueTask( funcDef,
	args = ( arg1, arg2, ... ),
	kwargs = { "kwarg1" : value1, "kwarg2" : value2 },
	callback = callbackDef,
	exceptionCallback = exceptionCallBackDef )
	or alternatively by using ProcessTask instance::
	task = ProcessTask( funcDef ,
	args = ( arg1, arg2, ... )
	kwargs = { "kwarg1" : value1, .. },
	callback = callbackDef,
	exceptionCallback = exceptionCallbackDef )
	pool.queueTask( task )
	where parameters are::
	:param funcDef: callable py object definition (function, lambda, class with __call__ slot defined
	:param list args: argument list
	:param dict kwargs: keyword arguments dictionary
	:param callback: callback function definition
	:param exceptionCallback: exception callback function definition
	The callback, exceptionCallbaks and the parameters are all optional. Once task has been added to the pool,
	it will be executed as soon as possible. Worker subprocesses automatically return the return value of the task.
	To obtain those results one has to execute::
	pool.processRequests()
	This method will process the existing return values of the task, even if the task does not return
	anything. This method has to be called to clean the result queues. To wait until all the requests are finished
	and process their result call::
	pool.processAllRequests()
	This function will block until all requests are finished and their result values have been processed.
	It is also possible to set the ProcessPool in daemon mode, in which all results are automatically
	processed as soon they are available, just after finalisation of task execution. To enable this mode one
	has to call::
	pool.daemonize()
	Callback functions
	------------------
	There are two types of callbacks that can be executed for each tasks: exception callback function and
	results callback function. The the firts one is executed when unhandled excpetion has been raised during
	task processing, and hence no task results are available, otherwise the execution of second callback type
	is performed.
	The callbacks could be attached in a two places::
	- directly in ProcessTask, in that case those have to be shelvable/picklable, so they should be defined as
	global functions with the signature :callback( task, taskResult ): where :task: is a :ProcessTask:
	reference and :taskResult: is whatever task callable it returning for restuls callback and
	:exceptionCallback( task, exc_info): where exc_info is a
	:S_ERROR{ "Exception": { "Value" : exceptionName, "Exc_info" : exceptionInfo }:
	- in ProcessPool, in that case there is no limitation on the function type, except the signature, which
	should follow :callback( task ): or :exceptionCallback( task ):, as those callbacks definitions
	are not put into the queues
	The first types of callbacks could be used in case various callable objects are put into the ProcessPool,
	so you probably want to handle them differently dependin on their results, while the second types are for
	executing same type of callables in subprocesses and hence you are expecting the same type of results
	everywhere.
	"""
	__RCSID__ = "$Id$"
	import multiprocessing
	import sys
	import time
	import threading
	import os
	import signal
	import Queue
	from types import FunctionType, TypeType, ClassType
	try:
	from DIRAC.FrameworkSystem.Client.Logger import gLogger
	except ImportError:
	gLogger = None
	try:
	from DIRAC.Core.Utilities.LockRing import LockRing
	except ImportError:
	LockRing = None
	try:
	from DIRAC.Core.Utilities.ReturnValues import S_OK, S_ERROR
	except ImportError:
	def S_OK( val = "" ):
	""" dummy S_OK """
	return { 'OK' : True, 'Value' : val }
	def S_ERROR( mess ):
	""" dummy S_ERROR """
	return { 'OK' : False, 'Message' : mess }
	class WorkingProcess( multiprocessing.Process ):
	"""
	.. class:: WorkingProcess
	WorkingProcess is a class that represents activity that is run in a separate process.
	It is running main thread (process) in daemon mode, reading tasks from :pendingQueue:, executing
	them and pushing back tasks with results to the :resultsQueue:. If task has got a timeout value
	defined a separate threading.Timer thread is started killing execution (and destroying worker)
	after :ProcessTask.__timeOut: seconds.
	Main execution could also terminate in a few different ways:
	* on every failed read attempt (from empty :pendingQueue:), the idle loop counter is increased,
	worker is terminated when counter is reaching a value of 10;
	* when stopEvent is set (so ProcessPool is in draining mode),
	* when parent process PID is set to 1 (init process, parent process with ProcessPool is dead).
	"""
	def __init__( self, pendingQueue, resultsQueue, stopEvent ):
	""" c'tor
	:param self: self refernce
	:param multiprocessing.Queue pendingQueue: queue storing ProcessTask before exection
	:param multiprocessing.Queue resultsQueue: queue storing callbacks and exceptionCallbacks
	:param multiprocessing.Event stopEvent: event to stop processing
	"""
	multiprocessing.Process.__init__( self )
	## daemonize
	self.daemon = True
	## flag to see if task is being treated
	self.__working = multiprocessing.Value( 'i', 0 )
	## task counter
	self.__taskCounter = multiprocessing.Value( 'i', 0 )
	## task queue
	self.__pendingQueue = pendingQueue
	## results queue
	self.__resultsQueue = resultsQueue
	## stop event
	self.__stopEvent = stopEvent
	## placeholder for watchdog thread
	self.__watchdogThread = None
	## placeholder for process thread
	self.__processThread = None
	## placeholder for current task
	self.task = None
	## start yourself at least
	self.start()
	def __watchdog( self ):
	""" watchdog thread target
	terminating/killing WorkingProcess when parent process is dead
	:param self: self reference
	"""
	while True:
	## parent is dead, commit suicide
	if os.getppid() == 1:
	os.kill( self.pid, signal.SIGTERM )
	## wait for half a minute and if worker is still alive use REAL silencer
	time.sleep(30)
	## now you're dead
	os.kill( self.pid, signal.SIGKILL )
	## wake me up in 5 seconds
	time.sleep(5)
	def isWorking( self ):
	""" check if process is being executed
	:param self: self reference
	"""
	return self.__working.value == 1
	def taskProcessed( self ):
	""" tell how many tasks have been processed so far
	:param self: self reference
	"""
	return self.__taskCounter
	def __processTask( self ):
	""" processThread target
	:param self: self reference
	"""
	if self.task:
	self.task.process()
	def run( self ):
	""" task execution
	reads and executes ProcessTask :task: out of pending queue and then pushes it
	to the results queue for callback execution
	:param self: self reference
	"""
	## start watchdog thread
	self.__watchdogThread = threading.Thread( target = self.__watchdog )
	self.__watchdogThread.daemon = True
	self.__watchdogThread.start()
	## http://cdn.memegenerator.net/instances/400x/19450565.jpg
	if LockRing:
	# Reset all locks
	lr = LockRing()
	lr._openAll()
	lr._setAllEvents()
	## zero processed task counter
	taskCounter = 0
	## zero idle loop counter
	idleLoopCount = 0
	## main loop
	while True:
	## draining, stopEvent is set, exiting
	if self.__stopEvent.is_set():
	return
	## clear task
	self.task = None
	## read from queue
	try:
	task = self.__pendingQueue.get( block = True, timeout = 10 )
	except Queue.Empty:
	## idle loop?
	idleLoopCount += 1
	## 10th idle loop - exit, nothing to do
	if idleLoopCount == 10:
	return
	continue
	## toggle __working flag
	self.__working.value = 1
	## save task
	self.task = task
	## reset idle loop counter
	idleLoopCount = 0
	## process task in a separate thread
	self.__processThread = threading.Thread( target = self.__processTask )
	self.__processThread.start()
	## join processThread with or without timeout
	if self.task.getTimeOut():
	self.__processThread.join( self.task.getTimeOut()+10 )
	else:
	self.__processThread.join()
	## processThread is still alive? stop it!
	if self.__processThread.is_alive():
	self.__processThread._Thread__stop()
	## check results and callbacks presence, put task to results queue
	if self.task.hasCallback() or self.task.hasPoolCallback():
	if not self.task.taskResults() and not self.task.taskException():
	self.task.setResult( S_ERROR("Timed out") )
	self.__resultsQueue.put( task )
	## increase task counter
	taskCounter += 1
	self.__taskCounter = taskCounter
	## toggle __working flag
	self.__working.value = 0
	class ProcessTask( object ):
	""" .. class:: ProcessTask
	Defines task to be executed in WorkingProcess together with its callbacks.
	"""
	## taskID
	taskID = 0
	def __init__( self,
	taskFunction,
	args = None,
	kwargs = None,
	taskID = None,
	callback = None,
	exceptionCallback = None,
	usePoolCallbacks = False,
	timeOut = 0 ):
	""" c'tor
	:warning: taskFunction has to be callable: it could be a function, lambda OR a class with
	__call__ operator defined. But be carefull with interpretation of args and kwargs, as they
	are passed to different places in above cases:
	1. for functions or lambdas args and kwargs are just treated as function parameters
	2. for callable classess (say MyTask) args and kwargs are passed to class contructor
	(MyTask.__init__) and MyTask.__call__ should be a method without parameters, i.e.
	MyTask definition should be::
	class MyTask:
	def __init__( self, *args, **kwargs ):
	...
	def __call__( self ):
	...
	:warning: depending on :timeOut: value, taskFunction execution can be forcefully terminated
	using SIGALRM after :timeOut: seconds spent, :timeOut: equal to zero means there is no any
	time out at all, except those during :ProcessPool: finalization
	:param self: self reference
	:param mixed taskFunction: definition of callable object to be executed in this task
	:param tuple args: non-keyword arguments
	:param dict kwargs: keyword arguments
	:param int taskID: task id, if not set,
	:param int timeOut: estimated time to execute taskFunction in seconds (default = 0, no timeOut at all)
	:param mixed callback: result callback function
	:param mixed exceptionCallback: callback function to be fired upon exception in taskFunction
	"""
	self.__taskFunction = taskFunction
	self.__taskArgs = args or []
	self.__taskKwArgs = kwargs or {}
	self.__taskID = taskID
	self.__resultCallback = callback
	self.__exceptionCallback = exceptionCallback
	self.__timeOut = 0
	## set time out
	self.setTimeOut( timeOut )
	self.__done = False
	self.__exceptionRaised = False
	self.__taskException = None
	self.__taskResult = None
	self.__usePoolCallbacks = usePoolCallbacks
	def taskResults( self ):
	""" get task results
	:param self: self reference
	"""
	return self.__taskResult
	def taskException( self ):
	""" get task exception
	:param self: self reference
	"""
	return self.__taskException
	def enablePoolCallbacks( self ):
	""" (re)enable use of ProcessPool callbacks """
	self.__usePoolCallbacks = True
	def disablePoolCallbacks( self ):
	""" disable execution of ProcessPool callbacks """
	self.__usePoolCallbacks = False
	def usePoolCallbacks( self ):
	""" check if results should be processed by callbacks defined in the :ProcessPool:
	:param self: self reference
	"""
	return self.__usePoolCallbacks
	def hasPoolCallback( self ):
	""" check if asked to execute :ProcessPool: callbacks
	:param self: self reference
	"""
	return self.__usePoolCallbacks
	def setTimeOut( self, timeOut ):
	""" set time out (in seconds)
	:param self: selt reference
	:param int timeOut: new time out value
	"""
	try:
	self.__timeOut = int( timeOut )
	return S_OK( self.__timeOut )
	except (TypeError, ValueError), error:
	return S_ERROR( str(error) )
	def getTimeOut( self ):
	""" get timeOut value
	:param self: self reference
	"""
	return self.__timeOut
	def hasTimeOutSet( self ):
	""" check if timeout is set
	:param self: self reference
	"""
	return bool( self.__timeOut != 0 )
	def getTaskID( self ):
	""" taskID getter
	:param self: self reference
	"""
	return self.__taskID
	def hasCallback( self ):
	""" callback existence checking
	:param self: self reference
	:return: True if callbak or exceptionCallback has been defined, False otherwise
	"""
	return self.__resultCallback or self.__exceptionCallback or self.__usePoolCallbacks
	def exceptionRaised( self ):
	""" flag to determine exception in process
	:param self: self reference
	"""
	return self.__exceptionRaised
	def doExceptionCallback( self ):
	""" execute exceptionCallback
	:param self: self reference
	"""
	if self.__done and self.__exceptionRaised and self.__exceptionCallback:
	self.__exceptionCallback( self, self.__taskException )
	def doCallback( self ):
	""" execute result callback function
	:param self: self reference
	"""
	if self.__done and not self.__exceptionRaised and self.__resultCallback:
	self.__resultCallback( self, self.__taskResult )
	def setResult( self, result ):
	""" set taskResult to result """
	self.__taskResult = result
	def process( self ):
	""" execute task
	:param self: self reference
	"""
	self.__done = True
	try:
	## it's a function?
	if type( self.__taskFunction ) is FunctionType:
	self.__taskResult = self.__taskFunction( *self.__taskArgs, **self.__taskKwArgs )
	## or a class?
	elif type( self.__taskFunction ) in ( TypeType, ClassType ):
	## create new instance
	taskObj = self.__taskFunction( *self.__taskArgs, **self.__taskKwArgs )
	### check if it is callable, raise TypeError if not
	if not callable( taskObj ):
	raise TypeError( "__call__ operator not defined not in %s class" % taskObj.__class__.__name__ )
	### call it at least
	self.__taskResult = taskObj()
	except Exception, x:
	self.__exceptionRaised = True
	if gLogger:
	gLogger.exception( "Exception in process of pool" )
	if self.__exceptionCallback or self.usePoolCallbacks():
	retDict = S_ERROR( 'Exception' )
	retDict['Value'] = str( x )
	retDict['Exc_info'] = sys.exc_info()[1]
	self.__taskException = retDict
	class ProcessPool( object ):
	"""
	.. class:: ProcessPool
	This class is managing multiprocessing execution of tasks (:ProcessTask: instances) in a separate
	sub-processes (:WorkingProcess:).
	Pool depth
	----------
	The :ProcessPool: is keeping required number of active workers all the time: slave workers are only created
	when pendingQueue is being filled with tasks, not exceeding defined min and max limits. When pendingQueue is
	empty, active workers will be cleaned up by themselves, as each worker has got built in
	self-destroy mechnism after 10 idle loops.
	Processing and communication
	----------------------------
	The communication between :ProcessPool: instace and slaves is performed using two :multiprocessing.Queues:
	* pendingQueue, used to push tasks to the workers,
	* resultsQueue for revert direction;
	and one :multiprocessing.Event: instance (stopEvent), which is working as a fuse to destroy idle workers
	in a clean manner.
	Processing of task begins with pushing it into :pendingQueue: using :ProcessPool.queueTask: or
	:ProcessPool.createAndQueueTask:. Every time new task is queued, :ProcessPool: is checking existance of
	active and idle workers and spawning new ones when required. The task is then read and processed on worker
	side. If results are ready and callback functions are defined, task is put back to the resultsQueue and it is
	ready to be picked up by ProcessPool again. To perform this last step one has to call :ProcessPool.processResults:,
	or alternatively ask for daemon mode processing, when this function is called again and again in
	separate background thread.
	Finalisation
	------------
	Finsalisation fo task processing is done in several steps:
	* if pool is working in daemon mode, background result processing thread is joined and stopped
	* :pendingQueue: is emptied by :ProcessPool.processAllResults: function, all enqueued tasks are executed
	* :stopEvent: is set, so all idle workers are exiting immediately
	* non-hanging workers are joined and terminated politelty
	* the rest of workers, if any, are forcefully retained by signals: first by SIGTERM, and if is doesn't work
	by SIGKILL
	:warn: Be carefull and choose wisely :timeout: argument to :ProcessPool.finalize:. Too short time period can
	cause that all workers will be killed.
	"""
	def __init__( self, minSize = 2, maxSize = 0, maxQueuedRequests = 10,
	strictLimits = True, poolCallback=None, poolExceptionCallback=None ):
	""" c'tor
	:param self: self reference
	:param int minSize: minimal number of simultaniously executed tasks
	:param int maxSize: maximal number of simultaniously executed tasks
	:param int maxQueueRequests: size of pending tasks queue
	:param bool strictLimits: flag to workers overcommitment
	:param callable poolCallbak: results callback
	:param callable poolExceptionCallback: exception callback
	"""
	## min workers
	self.__minSize = max( 1, minSize )
	## max workers
	self.__maxSize = max( self.__minSize, maxSize )
	## queue size
	self.__maxQueuedRequests = maxQueuedRequests
	## flag to worker overcommit
	self.__strictLimits = strictLimits
	## pool results callback
	self.__poolCallback = poolCallback
	## pool exception callback
	self.__poolExceptionCallback = poolExceptionCallback
	## pending queue
	self.__pendingQueue = multiprocessing.Queue( self.__maxQueuedRequests )
	## results queue
	self.__resultsQueue = multiprocessing.Queue( 0 )
	## stop event
	self.__stopEvent = multiprocessing.Event()
	## lock
	self.__prListLock = threading.Lock()
	## workers dict
	self.__workersDict = {}
	## flag to trigger workers draining
	self.__draining = False
	## placeholder for deamon results processing
	self.__daemonProcess = False
	## create initial workers
	self.__spawnNeededWorkingProcesses()
	def stopProcessing( self, timeout=10 ):
	""" case fire
	:param self: self reference
	"""
	self.finalize( timeout )
	def startProcessing( self ):
	""" restrat processing again
	:param self: self reference
	"""
	self.__draining = False
	self.__stopEvent.clear()
	self.daemonize()
	def setPoolCallback( self, callback ):
	""" set ProcessPool callback function
	:param self: self reference
	:param callable callback: callback function
	"""
	if callable( callback ):
	self.__poolCallback = callback
	def setPoolExceptionCallback( self, exceptionCallback ):
	""" set ProcessPool exception callback function
	:param self: self refernce
	:param callable exceptionCallback: exsception callback function
	"""
	if callable( exceptionCallback ):
	self.__poolExceptionCallback = exceptionCallback
	def getMaxSize( self ):
	""" maxSize getter
	:param self: self reference
	"""
	return self.__maxSize
	def getMinSize( self ):
	""" minSize getter
	:param self: self reference
	"""
	return self.__minSize
	def getNumWorkingProcesses( self ):
	""" count processes currently being executed
	:param self: self reference
	"""
	counter = 0
	self.__prListLock.acquire()
	try:
	counter = len( [ pid for pid, worker in self.__workersDict.items() if worker.isWorking() ] )
	finally:
	self.__prListLock.release()
	return counter
	def getNumIdleProcesses( self ):
	""" count processes being idle
	:param self: self reference
	"""
	counter = 0
	self.__prListLock.acquire()
	try:
	counter = len( [ pid for pid, worker in self.__workersDict.items() if not worker.isWorking() ] )
	finally:
	self.__prListLock.release()
	return counter
	def getFreeSlots( self ):
	""" get number of free slots availablr for workers
	:param self: self reference
	"""
	return max( 0, self.__maxSize - self.getNumWorkingProcesses() )
	def __spawnWorkingProcess( self ):
	""" create new process
	:param self: self reference
	"""
	self.__prListLock.acquire()
	try:
	worker = WorkingProcess( self.__pendingQueue, self.__resultsQueue, self.__stopEvent )
	while worker.pid == None:
	time.sleep(0.1)
	self.__workersDict[ worker.pid ] = worker
	finally:
	self.__prListLock.release()
	def __cleanDeadProcesses( self ):
	""" delete references of dead workingProcesses from ProcessPool.__workingProcessList """
	## check wounded processes
	self.__prListLock.acquire()
	try:
	for pid, worker in self.__workersDict.items():
	if not worker.is_alive():
	del self.__workersDict[pid]
	finally:
	self.__prListLock.release()
	def __spawnNeededWorkingProcesses( self ):
	""" create N working process (at least self.__minSize, but no more than self.__maxSize)
	:param self: self reference
	"""
	self.__cleanDeadProcesses()
	## if we're draining do not spawn new workers
	if self.__draining or self.__stopEvent.is_set():
	return
	while len( self.__workersDict ) < self.__minSize:
	if self.__draining or self.__stopEvent.is_set():
	return
	self.__spawnWorkingProcess()
	while self.hasPendingTasks() and \
	self.getNumIdleProcesses() == 0 and \
	len( self.__workersDict ) < self.__maxSize:
	if self.__draining or self.__stopEvent.is_set():
	return
	self.__spawnWorkingProcess()
	time.sleep( 0.1 )
	def queueTask( self, task, blocking = True, usePoolCallbacks= False ):
	""" enqueue new task into pending queue
	:param self: self reference
	:param ProcessTask task: new task to execute
	:param bool blocking: flag to block if necessary and new empty slot is available (default = block)
	:param bool usePoolCallbacks: flag to trigger execution of pool callbacks (default = don't execute)
	"""
	if not isinstance( task, ProcessTask ):
	raise TypeError( "Tasks added to the process pool must be ProcessTask instances" )
	if usePoolCallbacks and ( self.__poolCallback or self.__poolExceptionCallback ):
	task.enablePoolCallbacks()
	self.__prListLock.acquire()
	try:
	self.__pendingQueue.put( task, block = blocking )
	except Queue.Full:
	self.__prListLock.release()
	return S_ERROR( "Queue is full" )
	finally:
	self.__prListLock.release()
	self.__spawnNeededWorkingProcesses()
	## throttle a bit to allow task state propagation
	time.sleep( 0.1 )
	return S_OK()
	def createAndQueueTask( self,
	taskFunction,
	args = None,
	kwargs = None,
	taskID = None,
	callback = None,
	exceptionCallback = None,
	blocking = True,
	usePoolCallbacks = False,
	timeOut = 0):
	""" create new processTask and enqueue it in pending task queue
	:param self: self reference
	:param mixed taskFunction: callable object definition (FunctionType, LambdaType, callable class)
	:param tuple args: non-keyword arguments passed to taskFunction c'tor
	:param dict kwargs: keyword arguments passed to taskFunction c'tor
	:param int taskID: task Id
	:param mixed callback: callback handler, callable object executed after task's execution
	:param mixed exceptionCallback: callback handler executed if testFunction had raised an exception
	:param bool blocking: flag to block queue if necessary until free slot is available
	:param bool usePoolCallbacks: fire execution of pool defined callbacks after task callbacks
	:param int timeOut: time you want to spend executing :taskFunction:
	"""
	task = ProcessTask( taskFunction, args, kwargs, taskID, callback, exceptionCallback, usePoolCallbacks, timeOut )
	return self.queueTask( task, blocking )
	def hasPendingTasks( self ):
	""" check if taks are present in pending queue
	:param self: self reference
	:warning: results may be misleading if elements put into the queue are big
	"""
	return not self.__pendingQueue.empty()
	def isFull( self ):
	""" check in peding queue is full
	:param self: self reference
	:warning: results may be misleading if elements put into the queue are big
	"""
	return self.__pendingQueue.full()
	def isWorking( self ):
	""" check existence of working subprocesses
	:param self: self reference
	"""
	return not self.__pendingQueue.empty() or self.getNumWorkingProcesses()
	def processResults( self ):
	""" execute tasks' callbacks removing them from results queue
	:param self: self reference
	"""
	processed = 0
	while True:
	self.__cleanDeadProcesses()
	if not self.__pendingQueue.empty():
	self.__spawnNeededWorkingProcesses()
	time.sleep( 0.1 )
	if self.__resultsQueue.empty():
	break
	## get task
	task = self.__resultsQueue.get()
	## execute callbacks
	try:
	task.doExceptionCallback()
	task.doCallback()
	if task.usePoolCallbacks():
	if self.__poolExceptionCallback and task.exceptionRaised():
	self.__poolExceptionCallback( task.getTaskID(), task.taskException() )
	if self.__poolCallback and task.taskResults():
	self.__poolCallback( task.getTaskID(), task.taskResults() )
	except Exception, error:
	pass
	processed += 1
	return processed
	def processAllResults( self, timeout=10 ):
	""" process all enqueued tasks at once
	:param self: self reference
	"""
	start = time.time()
	while self.getNumWorkingProcesses() or not self.__pendingQueue.empty():
	self.processResults()
	time.sleep( 1 )
	if time.time() - start > timeout:
	break
	self.processResults()
	def finalize( self, timeout = 60 ):
	""" drain pool, shutdown processing in more or less clean way
	:param self: self reference
	:param timeout: seconds to wait before killing
	"""
	## start drainig
	self.__draining = True
	## join deamon process
	if self.__daemonProcess:
	self.__daemonProcess.join( timeout )
	## process all tasks
	self.processAllResults( timeout )
	## set stop event, all idle workers should be terminated
	self.__stopEvent.set()
	## join idle workers
	start = time.time()
	while self.__workersDict:
	if timeout <= 0 or time.time() - start >= timeout:
	break
	time.sleep( 0.1 )
	self.__cleanDeadProcesses()
	## second clean up - join and terminate workers
	for worker in self.__workersDict.values():
	if worker.is_alive():
	worker.terminate()
	worker.join(5)
	self.__cleanDeadProcesses()
	## third clean up - kill'em all!!!
	self.__filicide()
	def __filicide( self ):
	""" Kill all workers, kill'em all!
	:param self: self reference
	"""
	while self.__workersDict:
	pid = self.__workersDict.keys().pop(0)
	worker = self.__workersDict[pid]
	if worker.is_alive():
	os.kill( pid, signal.SIGKILL )
	del self.__workersDict[pid]
	def daemonize( self ):
	""" Make ProcessPool a finite being for opening and closing doors between chambers.
	Also just run it in a separate background thread to the death of PID 0.
	:param self: self reference
	"""
	if self.__daemonProcess:
	return
	self.__daemonProcess = threading.Thread( target = self.__backgroundProcess )
	self.__daemonProcess.setDaemon( 1 )
	self.__daemonProcess.start()
	def __backgroundProcess( self ):
	""" daemon thread target
	:param self: self reference
	"""
	while True:
	if self.__draining:
	return
	self.processResults()
	time.sleep( 1 )
	def __del__( self ):
	""" del slot
	:param self: self reference
	"""
	self.finalize( timeout = 10 )

view raw ProcessPool.py hosted with ❤ by GitHub