前面学习了多线程,接下来学习多进程的创建和使用。多进程更适合计算密集型的操作,他的语法和多线程非常相像,唯一需要注意的是,多线程之间是可以直接共享内存数据的;但是多进程默认每个进程是不能访问其他进程(程序)的内容。我们可以通过一些特殊的方式(队列,数组和字典)来实现,注意这几个数据结构和平常使用的不太一样,是在多进程中特殊定义的。
例如:通过queue来共享数据
#!/usr/bin/env python # -*- coding:utf-8 -*- # Author:Alex Li from multiprocessing import Process from multiprocessing import queues import multiprocessing from multiprocessing import Array def foo(i,arg): arg.put(i) print('say hi',i,arg.qsize()) if __name__ == "__main__": # li = [] li = queues.Queue(20,ctx=multiprocessing) for i in range(10): p = Process(target=foo,args=(i,li,)) p.start() p.join() ------------------ say hi 0 1 say hi 1 2 say hi 2 3 say hi 3 4 say hi 4 5 say hi 5 6 say hi 6 7 say hi 7 8 say hi 8 9 say hi 9 10
例2 通过array来共享数据,注意array初始化的时候就需要固定数据类型和长度
from multiprocessing import Process from multiprocessing import queues import multiprocessing from multiprocessing import Array def foo(i,arg): arg[i] = i + 100 for item in arg: print(item) print('================') if __name__ == "__main__": li = Array('i', 10) for i in range(10): p = Process(target=foo,args=(i,li,)) p.start() ---------------- 0 0 0 0 0 0 0 107 0 0 ================ 0 0 0 0 0 0 0 107 108 0 ================ 0 101 0 0 0 0 0 107 108 0 ================ 0 101 0 0 0 0 106 107 108 0 ================ 0 101 0 0 0 105 106 107 108 0 ================ ...(等等省略)
例3 通过字典方式进程间共享
#!/usr/bin/env python # -*- coding:utf-8 -*- # Author:Alex Li from multiprocessing import Process from multiprocessing import queues import multiprocessing from multiprocessing import Manager def foo(i,arg): arg[i] = i + 100 print(arg.values()) if __name__ == "__main__": # li = [] # li = queues.Queue(20,ctx=multiprocessing) obj = Manager() li = obj.dict() for i in range(10): p = Process(target=foo,args=(i,li,)) p.start() p.join() ---------------- [100] [100, 101] [100, 101, 102] [100, 101, 102, 103] [100, 101, 102, 103, 104] [100, 101, 102, 103, 104, 105] [100, 101, 102, 103, 104, 105, 106] [100, 101, 102, 103, 104, 105, 106, 107] [100, 101, 102, 103, 104, 105, 106, 107, 108] [100, 101, 102, 103, 104, 105, 106, 107, 108, 109]
和线程类似,当多个进程操作同一个全局变量的时候,需要加锁,不然可能错误;
比如
#!/usr/bin/env python # -*- coding:utf-8 -*- # Author:Alex Li from multiprocessing import Process from multiprocessing import queues from multiprocessing import Array from multiprocessing import RLock, Lock, Event, Condition, Semaphore import multiprocessing import time def foo(i,lis): lis[0] = lis[0] - 1 time.sleep(1) print('say hi',lis[0]) if __name__ == "__main__": # li = [] li = Array('i', 1) li[0] = 10 for i in range(10): p = Process(target=foo,args=(i,li)) p.start() ------------- say hi 0 say hi 0 say hi 0 say hi 0 say hi 0 say hi 0 say hi 0 say hi 0 say hi 0 say hi 0
如何修复?
两种方式,一个是p.start()下面加个p.join(),那真的就算按顺序一个个执行了;还有一个方式就是加锁
#!/usr/bin/env python # -*- coding:utf-8 -*- # Author:Alex Li from multiprocessing import Process from multiprocessing import queues from multiprocessing import Array from multiprocessing import RLock, Lock, Event, Condition, Semaphore import multiprocessing import time def foo(i,lis,lc): lc.acquire() lis[0] = lis[0] - 1 time.sleep(1) print('say hi',lis[0]) lc.release() if __name__ == "__main__": # li = [] li = Array('i', 1) li[0] = 10 lock = RLock() for i in range(10): p = Process(target=foo,args=(i,li,lock)) p.start() -------------- say hi 9 say hi 8 say hi 7 say hi 6 say hi 5 say hi 4 say hi 3 say hi 2 say hi 1 say hi 0
和线程池相比,Python已经提供了完备的进程池模块,因此可以直接使用。进程池里面有2种方法,apply或apply_async;前者是阻塞,而后者是非阻塞的
例如下面例子我使用的apply_async,那么所有的进程是(非阻塞)同时执行的,当执行到time.sleep(5),每个子线程会卡5秒,而同时主线程执行到了pool.terminate(),这个时候就直接终止程序了
#!/usr/bin/env python # -*- coding:utf-8 -*- from multiprocessing import Pool import time def f1(arg): print(arg,'b') time.sleep(5) print(arg,'a') if __name__ == "__main__": pool = Pool(5) for i in range(30): # pool.apply(func=f1,args=(i,))#按照顺序执行 pool.apply_async(func=f1,args=(i,))#同时执行 # pool.close() # 所有的任务执行完毕 time.sleep(2) pool.terminate() # 立即终止 pool.join() pass -------------- "C:\Program Files\Python3\python.exe" C:/temp/s13day11/day11/s16.py 0 b 1 b 2 b 3 b 4 b
如果改成close(),那么他会等待pool中的任务执行完成之后再中止程序
from multiprocessing import Pool import time def f1(arg): print(arg,'b') time.sleep(5) print(arg,'a') if __name__ == "__main__": pool = Pool(5) for i in range(30): # pool.apply(func=f1,args=(i,))#按照顺序执行 pool.apply_async(func=f1,args=(i,))#同时执行 pool.close() # 所有的任务执行完毕 time.sleep(2) # pool.terminate() # 立即终止 pool.join() pass ---------- "C:\Program Files\Python3\python.exe" C:/temp/s13day11/day11/s16.py 0 b 1 b 2 b 3 b 4 b 0 a 5 b 1 a 6 b 2 a 7 b 3 a 8 b 4 a 9 b 5 a 10 b 6 a 11 b 7 a 8 a 12 b 13 b 9 a 14 b 10 a 15 b 11 a 16 b 13 a 12 a 18 b 17 b 14 a 19 b 15 a 20 b 16 a 21 b 17 a 18 a 22 b 23 b 19 a 24 b 20 a 25 b 21 a 26 b 22 a 27 b 23 a 28 b 24 a 29 b 25 a 26 a 27 a 28 a 29 a
注意和线程类似,进程里面也可以使用join(),确保主进程阻塞在这里直到所有的子进程都结束。