Current File : //opt/rh/rh-python35/root/lib64/python3.5/multiprocessing/pool.py
#
# Module providing the `Pool` class for managing a process pool
#
# multiprocessing/pool.py
#
# Copyright (c) 2006-2008, R Oudkerk
# Licensed to PSF under a Contributor Agreement.
#
__all__ = ['Pool', 'ThreadPool']
#
# Imports
#
import threading
import queue
import itertools
import collections
import os
import time
import traceback
# If threading is available then ThreadPool should be provided. Therefore
# we avoid top-level imports which are liable to fail on some systems.
from . import util
from . import get_context, TimeoutError
#
# Constants representing the state of a pool
#
RUN = 0
CLOSE = 1
TERMINATE = 2
#
# Miscellaneous
#
job_counter = itertools.count()
def mapstar(args):
return list(map(*args))
def starmapstar(args):
return list(itertools.starmap(args[0], args[1]))
#
# Hack to embed stringification of remote traceback in local traceback
#
class RemoteTraceback(Exception):
def __init__(self, tb):
self.tb = tb
def __str__(self):
return self.tb
class ExceptionWithTraceback:
def __init__(self, exc, tb):
tb = traceback.format_exception(type(exc), exc, tb)
tb = ''.join(tb)
self.exc = exc
self.tb = '\n"""\n%s"""' % tb
def __reduce__(self):
return rebuild_exc, (self.exc, self.tb)
def rebuild_exc(exc, tb):
exc.__cause__ = RemoteTraceback(tb)
return exc
#
# Code run by worker processes
#
class MaybeEncodingError(Exception):
"""Wraps possible unpickleable errors, so they can be
safely sent through the socket."""
def __init__(self, exc, value):
self.exc = repr(exc)
self.value = repr(value)
super(MaybeEncodingError, self).__init__(self.exc, self.value)
def __str__(self):
return "Error sending result: '%s'. Reason: '%s'" % (self.value,
self.exc)
def __repr__(self):
return "<%s: %s>" % (self.__class__.__name__, self)
def worker(inqueue, outqueue, initializer=None, initargs=(), maxtasks=None,
wrap_exception=False):
assert maxtasks is None or (type(maxtasks) == int and maxtasks > 0)
put = outqueue.put
get = inqueue.get
if hasattr(inqueue, '_writer'):
inqueue._writer.close()
outqueue._reader.close()
if initializer is not None:
initializer(*initargs)
completed = 0
while maxtasks is None or (maxtasks and completed < maxtasks):
try:
task = get()
except (EOFError, OSError):
util.debug('worker got EOFError or OSError -- exiting')
break
if task is None:
util.debug('worker got sentinel -- exiting')
break
job, i, func, args, kwds = task
try:
result = (True, func(*args, **kwds))
except Exception as e:
if wrap_exception:
e = ExceptionWithTraceback(e, e.__traceback__)
result = (False, e)
try:
put((job, i, result))
except Exception as e:
wrapped = MaybeEncodingError(e, result[1])
util.debug("Possible encoding error while sending result: %s" % (
wrapped))
put((job, i, (False, wrapped)))
completed += 1
util.debug('worker exiting after %d tasks' % completed)
#
# Class representing a process pool
#
class Pool(object):
'''
Class which supports an async version of applying functions to arguments.
'''
_wrap_exception = True
def Process(self, *args, **kwds):
return self._ctx.Process(*args, **kwds)
def __init__(self, processes=None, initializer=None, initargs=(),
maxtasksperchild=None, context=None):
self._ctx = context or get_context()
self._setup_queues()
self._taskqueue = queue.Queue()
self._cache = {}
self._state = RUN
self._maxtasksperchild = maxtasksperchild
self._initializer = initializer
self._initargs = initargs
if processes is None:
processes = os.cpu_count() or 1
if processes < 1:
raise ValueError("Number of processes must be at least 1")
if initializer is not None and not callable(initializer):
raise TypeError('initializer must be a callable')
self._processes = processes
self._pool = []
self._repopulate_pool()
self._worker_handler = threading.Thread(
target=Pool._handle_workers,
args=(self, )
)
self._worker_handler.daemon = True
self._worker_handler._state = RUN
self._worker_handler.start()
self._task_handler = threading.Thread(
target=Pool._handle_tasks,
args=(self._taskqueue, self._quick_put, self._outqueue,
self._pool, self._cache)
)
self._task_handler.daemon = True
self._task_handler._state = RUN
self._task_handler.start()
self._result_handler = threading.Thread(
target=Pool._handle_results,
args=(self._outqueue, self._quick_get, self._cache)
)
self._result_handler.daemon = True
self._result_handler._state = RUN
self._result_handler.start()
self._terminate = util.Finalize(
self, self._terminate_pool,
args=(self._taskqueue, self._inqueue, self._outqueue, self._pool,
self._worker_handler, self._task_handler,
self._result_handler, self._cache),
exitpriority=15
)
def _join_exited_workers(self):
"""Cleanup after any worker processes which have exited due to reaching
their specified lifetime. Returns True if any workers were cleaned up.
"""
cleaned = False
for i in reversed(range(len(self._pool))):
worker = self._pool[i]
if worker.exitcode is not None:
# worker exited
util.debug('cleaning up worker %d' % i)
worker.join()
cleaned = True
del self._pool[i]
return cleaned
def _repopulate_pool(self):
"""Bring the number of pool processes up to the specified number,
for use after reaping workers which have exited.
"""
for i in range(self._processes - len(self._pool)):
w = self.Process(target=worker,
args=(self._inqueue, self._outqueue,
self._initializer,
self._initargs, self._maxtasksperchild,
self._wrap_exception)
)
self._pool.append(w)
w.name = w.name.replace('Process', 'PoolWorker')
w.daemon = True
w.start()
util.debug('added worker')
def _maintain_pool(self):
"""Clean up any exited workers and start replacements for them.
"""
if self._join_exited_workers():
self._repopulate_pool()
def _setup_queues(self):
self._inqueue = self._ctx.SimpleQueue()
self._outqueue = self._ctx.SimpleQueue()
self._quick_put = self._inqueue._writer.send
self._quick_get = self._outqueue._reader.recv
def apply(self, func, args=(), kwds={}):
'''
Equivalent of `func(*args, **kwds)`.
'''
assert self._state == RUN
return self.apply_async(func, args, kwds).get()
def map(self, func, iterable, chunksize=None):
'''
Apply `func` to each element in `iterable`, collecting the results
in a list that is returned.
'''
return self._map_async(func, iterable, mapstar, chunksize).get()
def starmap(self, func, iterable, chunksize=None):
'''
Like `map()` method but the elements of the `iterable` are expected to
be iterables as well and will be unpacked as arguments. Hence
`func` and (a, b) becomes func(a, b).
'''
return self._map_async(func, iterable, starmapstar, chunksize).get()
def starmap_async(self, func, iterable, chunksize=None, callback=None,
error_callback=None):
'''
Asynchronous version of `starmap()` method.
'''
return self._map_async(func, iterable, starmapstar, chunksize,
callback, error_callback)
def imap(self, func, iterable, chunksize=1):
'''
Equivalent of `map()` -- can be MUCH slower than `Pool.map()`.
'''
if self._state != RUN:
raise ValueError("Pool not running")
if chunksize == 1:
result = IMapIterator(self._cache)
self._taskqueue.put((((result._job, i, func, (x,), {})
for i, x in enumerate(iterable)), result._set_length))
return result
else:
assert chunksize > 1
task_batches = Pool._get_tasks(func, iterable, chunksize)
result = IMapIterator(self._cache)
self._taskqueue.put((((result._job, i, mapstar, (x,), {})
for i, x in enumerate(task_batches)), result._set_length))
return (item for chunk in result for item in chunk)
def imap_unordered(self, func, iterable, chunksize=1):
'''
Like `imap()` method but ordering of results is arbitrary.
'''
if self._state != RUN:
raise ValueError("Pool not running")
if chunksize == 1:
result = IMapUnorderedIterator(self._cache)
self._taskqueue.put((((result._job, i, func, (x,), {})
for i, x in enumerate(iterable)), result._set_length))
return result
else:
assert chunksize > 1
task_batches = Pool._get_tasks(func, iterable, chunksize)
result = IMapUnorderedIterator(self._cache)
self._taskqueue.put((((result._job, i, mapstar, (x,), {})
for i, x in enumerate(task_batches)), result._set_length))
return (item for chunk in result for item in chunk)
def apply_async(self, func, args=(), kwds={}, callback=None,
error_callback=None):
'''
Asynchronous version of `apply()` method.
'''
if self._state != RUN:
raise ValueError("Pool not running")
result = ApplyResult(self._cache, callback, error_callback)
self._taskqueue.put(([(result._job, None, func, args, kwds)], None))
return result
def map_async(self, func, iterable, chunksize=None, callback=None,
error_callback=None):
'''
Asynchronous version of `map()` method.
'''
return self._map_async(func, iterable, mapstar, chunksize, callback,
error_callback)
def _map_async(self, func, iterable, mapper, chunksize=None, callback=None,
error_callback=None):
'''
Helper function to implement map, starmap and their async counterparts.
'''
if self._state != RUN:
raise ValueError("Pool not running")
if not hasattr(iterable, '__len__'):
iterable = list(iterable)
if chunksize is None:
chunksize, extra = divmod(len(iterable), len(self._pool) * 4)
if extra:
chunksize += 1
if len(iterable) == 0:
chunksize = 0
task_batches = Pool._get_tasks(func, iterable, chunksize)
result = MapResult(self._cache, chunksize, len(iterable), callback,
error_callback=error_callback)
self._taskqueue.put((((result._job, i, mapper, (x,), {})
for i, x in enumerate(task_batches)), None))
return result
@staticmethod
def _handle_workers(pool):
thread = threading.current_thread()
# Keep maintaining workers until the cache gets drained, unless the pool
# is terminated.
while thread._state == RUN or (pool._cache and thread._state != TERMINATE):
pool._maintain_pool()
time.sleep(0.1)
# send sentinel to stop workers
pool._taskqueue.put(None)
util.debug('worker handler exiting')
@staticmethod
def _handle_tasks(taskqueue, put, outqueue, pool, cache):
thread = threading.current_thread()
for taskseq, set_length in iter(taskqueue.get, None):
task = None
i = -1
try:
for i, task in enumerate(taskseq):
if thread._state:
util.debug('task handler found thread._state != RUN')
break
try:
put(task)
except Exception as e:
job, ind = task[:2]
try:
cache[job]._set(ind, (False, e))
except KeyError:
pass
else:
if set_length:
util.debug('doing set_length()')
set_length(i+1)
continue
break
except Exception as ex:
job, ind = task[:2] if task else (0, 0)
if job in cache:
cache[job]._set(ind + 1, (False, ex))
if set_length:
util.debug('doing set_length()')
set_length(i+1)
else:
util.debug('task handler got sentinel')
try:
# tell result handler to finish when cache is empty
util.debug('task handler sending sentinel to result handler')
outqueue.put(None)
# tell workers there is no more work
util.debug('task handler sending sentinel to workers')
for p in pool:
put(None)
except OSError:
util.debug('task handler got OSError when sending sentinels')
util.debug('task handler exiting')
@staticmethod
def _handle_results(outqueue, get, cache):
thread = threading.current_thread()
while 1:
try:
task = get()
except (OSError, EOFError):
util.debug('result handler got EOFError/OSError -- exiting')
return
if thread._state:
assert thread._state == TERMINATE
util.debug('result handler found thread._state=TERMINATE')
break
if task is None:
util.debug('result handler got sentinel')
break
job, i, obj = task
try:
cache[job]._set(i, obj)
except KeyError:
pass
while cache and thread._state != TERMINATE:
try:
task = get()
except (OSError, EOFError):
util.debug('result handler got EOFError/OSError -- exiting')
return
if task is None:
util.debug('result handler ignoring extra sentinel')
continue
job, i, obj = task
try:
cache[job]._set(i, obj)
except KeyError:
pass
if hasattr(outqueue, '_reader'):
util.debug('ensuring that outqueue is not full')
# If we don't make room available in outqueue then
# attempts to add the sentinel (None) to outqueue may
# block. There is guaranteed to be no more than 2 sentinels.
try:
for i in range(10):
if not outqueue._reader.poll():
break
get()
except (OSError, EOFError):
pass
util.debug('result handler exiting: len(cache)=%s, thread._state=%s',
len(cache), thread._state)
@staticmethod
def _get_tasks(func, it, size):
it = iter(it)
while 1:
x = tuple(itertools.islice(it, size))
if not x:
return
yield (func, x)
def __reduce__(self):
raise NotImplementedError(
'pool objects cannot be passed between processes or pickled'
)
def close(self):
util.debug('closing pool')
if self._state == RUN:
self._state = CLOSE
self._worker_handler._state = CLOSE
def terminate(self):
util.debug('terminating pool')
self._state = TERMINATE
self._worker_handler._state = TERMINATE
self._terminate()
def join(self):
util.debug('joining pool')
assert self._state in (CLOSE, TERMINATE)
self._worker_handler.join()
self._task_handler.join()
self._result_handler.join()
for p in self._pool:
p.join()
@staticmethod
def _help_stuff_finish(inqueue, task_handler, size):
# task_handler may be blocked trying to put items on inqueue
util.debug('removing tasks from inqueue until task handler finished')
inqueue._rlock.acquire()
while task_handler.is_alive() and inqueue._reader.poll():
inqueue._reader.recv()
time.sleep(0)
@classmethod
def _terminate_pool(cls, taskqueue, inqueue, outqueue, pool,
worker_handler, task_handler, result_handler, cache):
# this is guaranteed to only be called once
util.debug('finalizing pool')
worker_handler._state = TERMINATE
task_handler._state = TERMINATE
util.debug('helping task handler/workers to finish')
cls._help_stuff_finish(inqueue, task_handler, len(pool))
assert result_handler.is_alive() or len(cache) == 0
result_handler._state = TERMINATE
outqueue.put(None) # sentinel
# We must wait for the worker handler to exit before terminating
# workers because we don't want workers to be restarted behind our back.
util.debug('joining worker handler')
if threading.current_thread() is not worker_handler:
worker_handler.join()
# Terminate workers which haven't already finished.
if pool and hasattr(pool[0], 'terminate'):
util.debug('terminating workers')
for p in pool:
if p.exitcode is None:
p.terminate()
util.debug('joining task handler')
if threading.current_thread() is not task_handler:
task_handler.join()
util.debug('joining result handler')
if threading.current_thread() is not result_handler:
result_handler.join()
if pool and hasattr(pool[0], 'terminate'):
util.debug('joining pool workers')
for p in pool:
if p.is_alive():
# worker has not yet exited
util.debug('cleaning up worker %d' % p.pid)
p.join()
def __enter__(self):
return self
def __exit__(self, exc_type, exc_val, exc_tb):
self.terminate()
#
# Class whose instances are returned by `Pool.apply_async()`
#
class ApplyResult(object):
def __init__(self, cache, callback, error_callback):
self._event = threading.Event()
self._job = next(job_counter)
self._cache = cache
self._callback = callback
self._error_callback = error_callback
cache[self._job] = self
def ready(self):
return self._event.is_set()
def successful(self):
assert self.ready()
return self._success
def wait(self, timeout=None):
self._event.wait(timeout)
def get(self, timeout=None):
self.wait(timeout)
if not self.ready():
raise TimeoutError
if self._success:
return self._value
else:
raise self._value
def _set(self, i, obj):
self._success, self._value = obj
if self._callback and self._success:
self._callback(self._value)
if self._error_callback and not self._success:
self._error_callback(self._value)
self._event.set()
del self._cache[self._job]
AsyncResult = ApplyResult # create alias -- see #17805
#
# Class whose instances are returned by `Pool.map_async()`
#
class MapResult(ApplyResult):
def __init__(self, cache, chunksize, length, callback, error_callback):
ApplyResult.__init__(self, cache, callback,
error_callback=error_callback)
self._success = True
self._value = [None] * length
self._chunksize = chunksize
if chunksize <= 0:
self._number_left = 0
self._event.set()
del cache[self._job]
else:
self._number_left = length//chunksize + bool(length % chunksize)
def _set(self, i, success_result):
success, result = success_result
if success:
self._value[i*self._chunksize:(i+1)*self._chunksize] = result
self._number_left -= 1
if self._number_left == 0:
if self._callback:
self._callback(self._value)
del self._cache[self._job]
self._event.set()
else:
self._success = False
self._value = result
if self._error_callback:
self._error_callback(self._value)
del self._cache[self._job]
self._event.set()
#
# Class whose instances are returned by `Pool.imap()`
#
class IMapIterator(object):
def __init__(self, cache):
self._cond = threading.Condition(threading.Lock())
self._job = next(job_counter)
self._cache = cache
self._items = collections.deque()
self._index = 0
self._length = None
self._unsorted = {}
cache[self._job] = self
def __iter__(self):
return self
def next(self, timeout=None):
with self._cond:
try:
item = self._items.popleft()
except IndexError:
if self._index == self._length:
raise StopIteration
self._cond.wait(timeout)
try:
item = self._items.popleft()
except IndexError:
if self._index == self._length:
raise StopIteration
raise TimeoutError
success, value = item
if success:
return value
raise value
__next__ = next # XXX
def _set(self, i, obj):
with self._cond:
if self._index == i:
self._items.append(obj)
self._index += 1
while self._index in self._unsorted:
obj = self._unsorted.pop(self._index)
self._items.append(obj)
self._index += 1
self._cond.notify()
else:
self._unsorted[i] = obj
if self._index == self._length:
del self._cache[self._job]
def _set_length(self, length):
with self._cond:
self._length = length
if self._index == self._length:
self._cond.notify()
del self._cache[self._job]
#
# Class whose instances are returned by `Pool.imap_unordered()`
#
class IMapUnorderedIterator(IMapIterator):
def _set(self, i, obj):
with self._cond:
self._items.append(obj)
self._index += 1
self._cond.notify()
if self._index == self._length:
del self._cache[self._job]
#
#
#
class ThreadPool(Pool):
_wrap_exception = False
@staticmethod
def Process(*args, **kwds):
from .dummy import Process
return Process(*args, **kwds)
def __init__(self, processes=None, initializer=None, initargs=()):
Pool.__init__(self, processes, initializer, initargs)
def _setup_queues(self):
self._inqueue = queue.Queue()
self._outqueue = queue.Queue()
self._quick_put = self._inqueue.put
self._quick_get = self._outqueue.get
@staticmethod
def _help_stuff_finish(inqueue, task_handler, size):
# put sentinels at head of inqueue to make workers finish
with inqueue.not_empty:
inqueue.queue.clear()
inqueue.queue.extend([None] * size)
inqueue.not_empty.notify_all()
Rosenblum TV: Video training, virtual workshops, classes, tutorials
RE-INVENTING THE TELEVISION NEWS BUSINESS*
A revolution in video storytelling
Creating entirely new & cost-effective production methods
From the world leaders in video production training and the creators of Character Driven Storytelling™
*and every other business that uses video
WHAT WE DO
Over the past 35 years, we have designed, built or restructured some of the most powerful news and journalism companies in the world.
We replace the traditional TV news ‘crew’ with one highly trained journalist, working alone with nothing but an iPhone.
No more TV news ‘crews’, no editors and no field producers.
This is television news done the way newspaper journalism is done – one reporter with their electronic pad and pencil.
In doing this, we can cut the cost of production by as much as 75% while increasing ratings and audience engagement.
In the place of conventional TV news ‘packages’ – ie, reporter stand up, interview, b-roll, man on the street, we marry great journalism with Netflix and Hollywood storytelling.
It’s a combination that works.
We have taken most of our clients to #1 in their respective markets.
And it’s not just for news. Any company, any profit, any NGO and anyone else who is online needs to tell their story in compelling yet cost-effective video. We can teach you to do that. Either in person or virtually.
EXAMPLES OF WHAT WE CAN TEACH YOUR STAFF TO PRODUCE
ITAY HOD
Itay Hod, MMJ with KPIX/CBS in San Francisco, took the 5-Day Intensive Video Storytelling Bootcamp in 2018.
Because he works alone, with only an iPhone, he was able to embed himself with a homeless family.
Here’s the story he produced in a one-day turn.
KIET DO
Kiet Do, an MMJ with KPIX/CBS in San Francisco, took the 5-Day Intensive Video Storytelling Bootcamp in 2021.
Here is a story he produced, all on his own, with only an iPhone and in a one-day turn.
TAYLOR SCHAUB
Taylor Schaub, an MMJ with Spectrum News 1 in LA, took the 5-Day Intensive Video Storytelling Bootcamp in 2023.
Here is a story he turned in only one day, using only an iPhone. It was the first video story he ever did and it was nominated for an Emmy.
THE BOOTCAMP
How do we convert stations and whole networks to working in this way?
Since 1988, we have run intensive 5-Day Video Storytelling Bootcamps
We have done these all over the world.
These are hands-on bootcamps, and participants learn an entirely new way of creating TV news stories.
-We shoot at a 3:1 ratio or lower, so turnaround times are fast.
-We go directly from camera to timelilne and edit – no written scripts. We work in the medium of pictures and sound.
-We are entirely character-driven.
-We are driven by pictures and real events.
-We are focused almost entirely on ’the viewer experience’.
Since 1988, more than 70,000 journalists around the world have taken our bootcamps, either in person on virtualy.
We have started to work with CBS News, bringing our ideas of character-driven storytelling to one of the most successful and biggest networks in the United States. Since beginning to work with them ratings have climbed and more importantly, audience engagement is through the ceiling.
We started New York Times Television in 1990 and it was the first paper to be brought into the world of TV. It quickly became one of the most successful non-fiction production companies in the United States. The series and documentaries we produced won many awards including multiple Emmys.
We have been working with the BBC since the year 2000 helping to convert their national news network to our visual storytelling technique. Most recently we have trained teams from their sports, documentaries, and comedy divisions to make character-driven stories using only smartphones.
For the past five years we have worked with Spectrum News to introduce and train their journalists on visual, character driven storytelling using smartphones helping to create a different kind of local news for their network of 24-Hour News Stations across the United States.
In 2006, we were approached by the United Nations. Rather than rely on news outlets, it would be much easier to train the field operatives to produce their own stories. We spent two years working with the UN, training more than 100 of their staff in bootcamps in Geneva and Nairobi.
We trained 50 print reporters at the paper to shoot and tell their own stories, in conjunction with their print work. We built a TV newsroom in their existing print newsroom – you could not ask for a better set and they began to live stream their stories in conjunction with their print work.
We spent two years with McGraw Hill, training more than 150 of their staffers, making them completely video literate. McGraw/Hill media properties we transit included Business Week, Aviation Week, (what was the name of the architecture magazine), and JD Power and Associates.
In 1990, we were approached by The Voice of America, the official broadcasting agency for the United States Government. When we met with VOA, they were only a short wave radio broadcaster, but working with them, we took them into television, launching VOA-TV.
British based Oyster Yachts makes some of the finest yachts in the world. Like every other company, they had to find a way to feed the never-ending video demands of social media – sites like Instagram and TikTok. We trained the Oyster staff to tell their own stories, using only iPhones.
We were approached by SEPA, the Scottish Environmental Protection Agency because they had to continually find a way to ‘feed the media beast’. The result was that SEPA was able to tell their own stories, whenever they wanted, and at almost no additional cost.
Michael Rosenblum has been writing about the media since 1988. His work and ideas have appeared in The Guardian, The Huffington Post, Ilkeston Life and many other publications.
He has been blogging regularly for the past 35 years on this subject. Having taught media studies at Columbia University, NYU and now the University of Oxford, he is considered an expert on this subject.
Continue reading this post or look back at previous posts.
