1
|
# Parallel processing
|
2
|
|
3
|
import cPickle
|
4
|
import Queue
|
5
|
import types
|
6
|
import warnings
|
7
|
|
8
|
import exc
|
9
|
from Runnable import Runnable
|
10
|
|
11
|
def try_pickle(value):
|
12
|
try: cPickle.dumps(value)
|
13
|
except Exception, e:
|
14
|
exc.add_msg(e, 'Tried to pickle: '+repr(value))
|
15
|
raise
|
16
|
|
17
|
def prepickle(value):
|
18
|
if isinstance(value, types.FunctionType): return Runnable(value)
|
19
|
elif isinstance(value, MultiProducerPool): return None
|
20
|
elif isinstance(value, list): return map(prepickle, list_)
|
21
|
elif isinstance(value, dict):
|
22
|
return dict(((k, prepickle(v)) for k, v in value.iteritems()))
|
23
|
else: return value
|
24
|
|
25
|
class SyncPool:
|
26
|
'''A dummy synchronous Pool to use if multiprocessing is not available'''
|
27
|
def __init__(self, processes=None): pass
|
28
|
|
29
|
class Result:
|
30
|
def __init__(self, value): self.value = value
|
31
|
|
32
|
def get(timeout=None): return self.value
|
33
|
|
34
|
def wait(timeout=None): pass
|
35
|
|
36
|
def ready(): return True
|
37
|
|
38
|
def successful(): return True # TODO: False if raised exception
|
39
|
|
40
|
def apply_async(self, func, args=None, kw_args=None, callback=None):
|
41
|
if args == None: args = ()
|
42
|
if kw_args == None: kw_args = {}
|
43
|
if callback == None: callback = lambda v: None
|
44
|
|
45
|
value = func(*args, **kw_args)
|
46
|
callback(value)
|
47
|
return self.Result(value)
|
48
|
|
49
|
class MultiProducerPool:
|
50
|
'''A multi-producer pool. You must call pool.main_loop() in the thread that
|
51
|
created this to process new tasks.'''
|
52
|
|
53
|
def __init__(self, processes=None):
|
54
|
'''
|
55
|
@param processes If 0, uses SyncPool
|
56
|
@post The # processes actually used is made available in self.process_ct
|
57
|
'''
|
58
|
try:
|
59
|
if processes == 0: raise ImportError('turned off')
|
60
|
import multiprocessing
|
61
|
import multiprocessing.pool
|
62
|
except ImportError, e:
|
63
|
warnings.warn(UserWarning('Not using parallel processing: '+str(e)))
|
64
|
processes = 1
|
65
|
Pool_ = SyncPool
|
66
|
Queue_ = Queue.Queue
|
67
|
else:
|
68
|
if processes == None: processes = multiprocessing.cpu_count()
|
69
|
Pool_ = multiprocessing.pool.Pool
|
70
|
Queue_ = multiprocessing.Queue
|
71
|
|
72
|
self.process_ct = processes
|
73
|
self.pool = Pool_(processes)
|
74
|
self.queue = Queue_()
|
75
|
|
76
|
def main_loop(self):
|
77
|
'''@param pool Must be a pool returned by mk_pool()'''
|
78
|
try:
|
79
|
while True:
|
80
|
# block=False raises Empty immediately if the queue is empty,
|
81
|
# which indicates that the program is done
|
82
|
call = self.queue.get(block=False)
|
83
|
self.pool.apply_async(call.func, call.args, call.kw_args,
|
84
|
call.callback)
|
85
|
except Queue.Empty: pass
|
86
|
|
87
|
class Result:
|
88
|
def get(timeout=None): raise NotImplementedError()
|
89
|
|
90
|
def wait(timeout=None): raise NotImplementedError()
|
91
|
|
92
|
def ready(): raise NotImplementedError()
|
93
|
|
94
|
def successful(): raise NotImplementedError()
|
95
|
|
96
|
def apply_async(self, func, args=None, kw_args=None, callback=None):
|
97
|
if args == None: args = ()
|
98
|
if kw_args == None: kw_args = {}
|
99
|
|
100
|
call = Runnable(func, *prepickle(args), **prepickle(kw_args))
|
101
|
call.callback = callback # store this inside the Runnable
|
102
|
|
103
|
# Try pickling the args. If it fails, we'll get a full traceback here,
|
104
|
# which is not provided with pickling errors in multiprocessing's Pool.
|
105
|
try_pickle(call)
|
106
|
|
107
|
self.queue.put_nowait(call)
|
108
|
return self.Result()
|