Home  >  Article  >  Backend Development  >  Detailed explanation of Python processes, threads, and coroutines

Detailed explanation of Python processes, threads, and coroutines

大家讲道理
大家讲道理Original
2016-11-07 11:21:111290browse

History of Processes and Threads

We all know that computers are composed of hardware and software. The CPU in the hardware is the core of the computer and undertakes all the tasks of the computer. The operating system is software that runs on the hardware and is the manager of the computer. It is responsible for the management and allocation of resources and the scheduling of tasks. A program is software with a certain function that runs on the system, such as a browser, music player, etc. Every time a program is executed, a certain function will be completed. For example, a browser helps us open a web page. In order to ensure its independence, a special data structure is needed to manage and control the execution of the program - the process control block. A process is a dynamic execution of a program on a data set. A process generally consists of three parts: program, data set, and process control block. The program we write is used to describe what functions the process needs to complete and how to complete it; the data set is the resources that the program needs to use during execution; the process control block is used to record the external characteristics of the process, describe the execution change process of the process, and the system It can be used to control and manage processes. It is the only sign that the system senses the existence of a process.


In early operating systems, computers had only one core, the smallest unit for process execution. Task scheduling used a time slice rotation preemptive method for process scheduling. Each process has its own independent memory to ensure the isolation of the memory address spaces between processes. With the development of computer technology, processes have appeared many disadvantages. First, the cost of creating, canceling, and switching processes is relatively high. Second, due to the Symmetrical Multi-Processing (Symmetrical Multi-Processing), also called SMP, refers to A group of processors (multiple CPUs) are assembled on one computer, and the emergence of shared memory subsystems and bus structures between CPUs can satisfy multiple running units, but the parallel overhead of multiple processes is too large. At this time, the concept of threads was introduced. A thread is also called a lightweight process. It is a basic CPU execution unit and the smallest unit in the program execution process. It is composed of a thread ID, a program counter, a register set and a stack. The introduction of threads reduces the overhead of concurrent execution of programs and improves the concurrency performance of the operating system. Threads do not have their own system resources, only the resources that are essential during runtime. But threads can share other resources owned by the process with other threads that belong to the same process.


The relationship between processes and threads

Threads belong to processes. Threads run in the process space. Threads generated by the same process share the same memory space. When the process exits, all threads generated by the process will be deleted. Force quit and clear. A thread can share all the resources owned by the process with other threads belonging to the same process, but it itself basically does not own system resources, only a little bit of information essential for running (such as a program counter, a set of registers, and a stack).

python threads

Threading is used to provide thread-related operations. A thread is the smallest unit of work in an application.

1. Threading module

The threading module is built on the _thread module. The thread module handles and controls threads in a low-level, primitive way, while the threading module provides a more convenient API to handle threads by secondary encapsulation of threads.

import threading
import time
   
def worker(num):
    """
    thread worker function
    :return:
    """
    time.sleep(1)
    print("The num is  %d" % num)
    return
   
for i in range(20):
    t = threading.Thread(target=worker,args=(i,),name=“t.%d” % i)
    t.start()

The above code creates 20 "foreground" threads, and then the controller is handed over to the CPU. The CPU schedules according to the specified algorithm and executes instructions in slices.

Thread method description

t.start(): Activate the thread,

t.getName(): Get the name of the thread

t.setName(): Set the name of the thread

t.name: Get or set the thread The name of

t.is_alive(): Determine whether the thread is active

t.isAlive(): Determine whether the thread is active

t.setDaemon() Set to background thread or foreground thread (default: False); Set whether the thread is a daemon thread through a Boolean value, which can only be used after executing the start() method. If it is a background thread, the background thread is also running during the execution of the main thread. After the main thread is executed, the background thread stops regardless of success or failure; if it is a foreground thread, the foreground thread is also running during the execution of the main thread. After the main thread completes execution, the program stops after waiting for the foreground thread to complete execution.

t.isDaemon(): Determine whether it is a daemon thread.

t.ident: Get the identifier of the thread. The thread identifier is a non-zero integer. This property is only valid after the start() method is called, otherwise it just returns None.

t.join(): Execute each thread one by one, and continue execution after completion. This method makes multi-threading meaningless.

t.run(): The thread automatically executes the run of the thread object after being scheduled by the CPU. Method

2, threading lock threading.RLock and threading.Lock

Since threads are randomly scheduled, and each thread may only execute n executions, the CPU then executes other threads. In order to ensure the accuracy of data, the concept of lock is introduced. Therefore, the following problems may occur:

例:假设列表A的所有元素就为0,当一个线程从前向后打印列表的所有元素,另外一个线程则从后向前修改列表的元素为1,那么输出的时候,列表的元素就会一部分为0,一部分为1,这就导致了数据的不一致。锁的出现解决了这个问题。

import threading
import time
   
globals_num = 0
   
lock = threading.RLock()
   
def Func():
    lock.acquire()  # 获得锁
    global globals_num
    globals_num += 1
    time.sleep(1)
    print(globals_num)
    lock.release()  # 释放锁
   
for i in range(10):
    t = threading.Thread(target=Func)
    t.start()

3、threading.RLock和threading.Lock 的区别

RLock允许在同一线程中被多次acquire。而Lock却不允许这种情况。 如果使用RLock,那么acquire和release必须成对出现,即调用了n次acquire,必须调用n次的release才能真正释放所占用的琐。

import threading
lock = threading.Lock()    #Lock对象
lock.acquire()
lock.acquire()  #产生了死琐。
lock.release()
lock.release() 
import threading
rLock = threading.RLock()  #RLock对象
rLock.acquire()
rLock.acquire()    #在同一线程内,程序不会堵塞。
rLock.release()
rLock.release()

4、threading.Event

python线程的事件用于主线程控制其他线程的执行,事件主要提供了三个方法 set、wait、clear。

事件处理的机制:全局定义了一个“Flag”,如果“Flag”值为 False,那么当程序执行 event.wait 方法时就会阻塞,如果“Flag”值为True,那么event.wait 方法时便不再阻塞。

clear:将“Flag”设置为False

set:将“Flag”设置为True

Event.isSet() :判断标识位是否为Ture。

import threading
   
def do(event):
    print('start')
    event.wait()
    print('execute')
   
event_obj = threading.Event()
for i in range(10):
    t = threading.Thread(target=do, args=(event_obj,))
    t.start()
   
event_obj.clear()
inp = input('input:')
if inp == 'true':
    event_obj.set()

当线程执行的时候,如果flag为False,则线程会阻塞,当flag为True的时候,线程不会阻塞。它提供了本地和远程的并发性。

5、threading.Condition

一个condition变量总是与某些类型的锁相联系,这个可以使用默认的情况或创建一个,当几个condition变量必须共享和同一个锁的时候,是很有用的。锁是conditon对象的一部分:没有必要分别跟踪。

condition变量服从上下文管理协议:with语句块封闭之前可以获取与锁的联系。 acquire() 和 release() 会调用与锁相关联的相应的方法。

其他和锁关联的方法必须被调用,wait()方法会释放锁,当另外一个线程使用 notify() or notify_all()唤醒它之前会一直阻塞。一旦被唤醒,wait()会重新获得锁并返回,

Condition类实现了一个conditon变量。 这个conditiaon变量允许一个或多个线程等待,直到他们被另一个线程通知。 如果lock参数,被给定一个非空的值,,那么他必须是一个lock或者Rlock对象,它用来做底层锁。否则,会创建一个新的Rlock对象,用来做底层锁。

wait(timeout=None) : 等待通知,或者等到设定的超时时间。当调用这wait()方法时,如果调用它的线程没有得到锁,那么会抛出一个RuntimeError 异常。 wati()释放锁以后,在被调用相同条件的另一个进程用notify() or notify_all() 叫醒之前 会一直阻塞。wait() 还可以指定一个超时时间。

如果有等待的线程,notify()方法会唤醒一个在等待conditon变量的线程。notify_all() 则会唤醒所有在等待conditon变量的线程。

注意: notify()和notify_all()不会释放锁,也就是说,线程被唤醒后不会立刻返回他们的wait() 调用。除非线程调用notify()和notify_all()之后放弃了锁的所有权。

在典型的设计风格里,利用condition变量用锁去通许访问一些共享状态,线程在获取到它想得到的状态前,会反复调用wait()。修改状态的线程在他们状态改变时调用 notify() or notify_all(),用这种方式,线程会尽可能的获取到想要的一个等待者状态。 例子: 生产者-消费者模型,

import threading
import time
def consumer(cond):
    with cond:
        print("consumer before wait")
        cond.wait()
        print("consumer after wait")
   
def producer(cond):
    with cond:
        print("producer before notifyAll")
        cond.notifyAll()
        print("producer after notifyAll")
   
condition = threading.Condition()
c1 = threading.Thread(name="c1", target=consumer, args=(condition,))
c2 = threading.Thread(name="c2", target=consumer, args=(condition,))
   
p = threading.Thread(name="p", target=producer, args=(condition,))
   
c1.start()
time.sleep(2)
c2.start()
time.sleep(2)
p.start()

6、queue模块

Queue 就是对队列,它是线程安全的

举例来说,我们去麦当劳吃饭。饭店里面有厨师职位,前台负责把厨房做好的饭卖给顾客,顾客则去前台领取做好的饭。这里的前台就相当于我们的队列。形成管道样,厨师做好饭通过前台传送给顾客,所谓单向队列

这个模型也叫生产者-消费者模型。

import queue
q = queue.Queue(maxsize=0)  # 构造一个先进显出队列,maxsize指定队列长度,为0 时,表示队列长度无限制。
q.join()    # 等到队列为kong的时候,在执行别的操作
q.qsize()   # 返回队列的大小 (不可靠)
q.empty()   # 当队列为空的时候,返回True 否则返回False (不可靠)
q.full()    # 当队列满的时候,返回True,否则返回False (不可靠)
q.put(item, block=True, timeout=None) #  将item放入Queue尾部,item必须存在,可以参数block默认为True,表示当队列满时,会等待队列给出可用位置,
                         为False时为非阻塞,此时如果队列已满,会引发queue.Full 异常。 可选参数timeout,表示 会阻塞设置的时间,过后,
                          如果队列无法给出放入item的位置,则引发 queue.Full 异常
q.get(block=True, timeout=None) #   移除并返回队列头部的一个值,可选参数block默认为True,表示获取值的时候,如果队列为空,则阻塞,为False时,不阻塞,
                      若此时队列为空,则引发 queue.Empty异常。 可选参数timeout,表示会阻塞设置的时候,过后,如果队列为空,则引发Empty异常。
q.put_nowait(item) #   等效于 put(item,block=False)
q.get_nowait() #    等效于 get(item,block=False)

代码如下:

#!/usr/bin/env python
import Queue
import threading
message = Queue.Queue(10)
  
  
def producer(i):
    while True:
        message.put(i)
  
  
def consumer(i):
    while True:
        msg = message.get()
  
  
for i in range(12):
    t = threading.Thread(target=producer, args=(i,))
    t.start()
  
for i in range(10):
    t = threading.Thread(target=consumer, args=(i,))
    t.start()

那就自己做个线程池吧:

# 简单往队列中传输线程数
import threading
import time
import queue
class Threadingpool():
    def __init__(self,max_num = 10):
        self.queue = queue.Queue(max_num)
        for i in range(max_num):
            self.queue.put(threading.Thread)
    def getthreading(self):
        return self.queue.get()
    def addthreading(self):
        self.queue.put(threading.Thread)
def func(p,i):
    time.sleep(1)
    print(i)
    p.addthreading()
if __name__ == "__main__":
    p = Threadingpool()
    for i in range(20):
        thread = p.getthreading()
        t = thread(target = func, args = (p,i))
        t.start()
#往队列中无限添加任务
import queue
import threading
import contextlib
import time
StopEvent = object()
class ThreadPool(object):
    def __init__(self, max_num):
        self.q = queue.Queue()
        self.max_num = max_num
        self.terminal = False
        self.generate_list = []
        self.free_list = []
    def run(self, func, args, callback=None):
        """
        线程池执行一个任务
        :param func: 任务函数
        :param args: 任务函数所需参数
        :param callback: 任务执行失败或成功后执行的回调函数,回调函数有两个参数1、任务函数执行状态;2、任务函数返回值(默认为None,即:不执行回调函数)
        :return: 如果线程池已经终止,则返回True否则None
        """
        if len(self.free_list) == 0 and len(self.generate_list) < self.max_num:
            self.generate_thread()
        w = (func, args, callback,)
        self.q.put(w)
    def generate_thread(self):
        """
        创建一个线程
        """
        t = threading.Thread(target=self.call)
        t.start()
    def call(self):
        """
        循环去获取任务函数并执行任务函数
        """
        current_thread = threading.currentThread
        self.generate_list.append(current_thread)
        event = self.q.get()  # 获取线程
        while event != StopEvent:   # 判断获取的线程数不等于全局变量
            func, arguments, callback = event   # 拆分元祖,获得执行函数,参数,回调函数
            try:
                result = func(*arguments)   # 执行函数
                status = True
            except Exception as e:    # 函数执行失败
                status = False
                result = e
            if callback is not None:
                try:
                    callback(status, result)
                except Exception as e:
                    pass
            # self.free_list.append(current_thread)
            # event = self.q.get()
            # self.free_list.remove(current_thread)
            with self.work_state():
                event = self.q.get()
        else:
            self.generate_list.remove(current_thread)
    def close(self):
        """
        关闭线程,给传输全局非元祖的变量来进行关闭
        :return:
        """
        for i in range(len(self.generate_list)):
            self.q.put(StopEvent)
    def terminate(self):
        """
        突然关闭线程
        :return:
        """
        self.terminal = True
        while self.generate_list:
            self.q.put(StopEvent)
        self.q.empty()
    @contextlib.contextmanager
    def work_state(self):
        self.free_list.append(threading.currentThread)
        try:
            yield
        finally:
            self.free_list.remove(threading.currentThread)
def work(i):
    print(i)
    return i +1 # 返回给回调函数
def callback(ret):
    print(ret)
pool = ThreadPool(10)
for item in range(50):
    pool.run(func=work, args=(item,),callback=callback)
pool.terminate()
# pool.close()

python 进程

multiprocessing是python的多进程管理包,和threading.Thread类似。

1、multiprocessing模块

直接从侧面用subprocesses替换线程使用GIL的方式,由于这一点,multiprocessing模块可以让程序员在给定的机器上充分的利用CPU。在multiprocessing中,通过创建Process对象生成进程,然后调用它的start()方法,

from multiprocessing import Process
  
def func(name):
    print(&#39;hello&#39;, name)
  
 
if __name__ == "__main__":
    p = Process(target=func,args=(&#39;zhangyanlin&#39;,))
    p.start()
    p.join()  # 等待进程执行完毕

在使用并发设计的时候最好尽可能的避免共享数据,尤其是在使用多进程的时候。 如果你真有需要 要共享数据, multiprocessing提供了两种方式。

(1)multiprocessing,Array,Value

数据可以用Value或Array存储在一个共享内存地图里,如下:

from multiprocessing import Array,Value,Process
  
def func(a,b):
    a.value = 3.333333333333333
    for i in range(len(b)):
        b[i] = -b[i]
  
  
if __name__ == "__main__":
    num = Value(&#39;d&#39;,0.0)
    arr = Array(&#39;i&#39;,range(11))
  
  
    c = Process(target=func,args=(num,arr))
    d= Process(target=func,args=(num,arr))
    c.start()
    d.start()
    c.join()
    d.join()
  
    print(num.value)
    for i in arr:
        print(i)

输出:

3.1415927

[0, -1, -2, -3, -4, -5, -6, -7, -8, -9]

创建num和arr时,“d”和“i”参数由Array模块使用的typecodes创建:“d”表示一个双精度的浮点数,“i”表示一个有符号的整数,这些共享对象将被线程安全的处理。

Array(‘i’, range(10))中的‘i’参数:

‘c’: ctypes.c_char     ‘u’: ctypes.c_wchar    ‘b’: ctypes.c_byte     ‘B’: ctypes.c_ubyte

‘h’: ctypes.c_short     ‘H’: ctypes.c_ushort    ‘i’: ctypes.c_int      ‘I’: ctypes.c_uint

‘l’: ctypes.c_long,    ‘L’: ctypes.c_ulong    ‘f’: ctypes.c_float    ‘d’: ctypes.c_double

(2)multiprocessing,Manager

由Manager()返回的manager提供list, dict, Namespace, Lock, RLock, Semaphore, BoundedSemaphore, Condition, Event, Barrier, Queue, Value and Array类型的支持。

from multiprocessing import Process,Manager
def f(d,l):
    d["name"] = "zhangyanlin"
    d["age"] = 18
    d["Job"] = "pythoner"
    l.reverse()
  
if __name__ == "__main__":
    with Manager() as man:
        d = man.dict()
        l = man.list(range(10))
  
        p = Process(target=f,args=(d,l))
        p.start()
        p.join()
  
        print(d)
        print(l)

输出:

{0.25: None, 1: '1', '2': 2}

[9, 8, 7, 6, 5, 4, 3, 2, 1, 0]

Server process manager比 shared memory 更灵活,因为它可以支持任意的对象类型。另外,一个单独的manager可以通过进程在网络上不同的计算机之间共享,不过他比shared memory要慢。

2、进程池(Using a pool of workers)

Pool类描述了一个工作进程池,他有几种不同的方法让任务卸载工作进程。

进程池内部维护一个进程序列,当使用时,则去进程池中获取一个进程,如果进程池序列中没有可供使用的进进程,那么程序就会等待,直到进程池中有可用进程为止。

我们可以用Pool类创建一个进程池, 展开提交的任务给进程池。 例:

#apply
from  multiprocessing import Pool
import time
  
def f1(i):
    time.sleep(0.5)
    print(i)
    return i + 100
  
if __name__ == "__main__":
    pool = Pool(5)
    for i in range(1,31):
        pool.apply(func=f1,args=(i,))
  
#apply_async
def f1(i):
    time.sleep(0.5)
    print(i)
    return i + 100
def f2(arg):
    print(arg)
  
if __name__ == "__main__":
    pool = Pool(5)
    for i in range(1,31):
        pool.apply_async(func=f1,args=(i,),callback=f2)
    pool.close()
    pool.join()

一个进程池对象可以控制工作进程池的哪些工作可以被提交,它支持超时和回调的异步结果,有一个类似map的实现。


processes :使用的工作进程的数量,如果processes是None那么使用 os.cpu_count()返回的数量。

initializer: 如果initializer是None,那么每一个工作进程在开始的时候会调用initializer(*initargs)。

maxtasksperchild:工作进程退出之前可以完成的任务数,完成后用一个心的工作进程来替代原进程,来让闲置的资源被释放。maxtasksperchild默认是None,意味着只要Pool存在工作进程就会一直存活。

context: 用在制定工作进程启动时的上下文,一般使用 multiprocessing.Pool() 或者一个context对象的Pool()方法来创建一个池,两种方法都适当的设置了context

注意:Pool对象的方法只可以被创建pool的进程所调用。

New in version 3.2: maxtasksperchild
New in version 3.4: context


进程池的方法

apply(func[, args[, kwds]])

:使用arg和kwds参数调用func函数,结果返回前会一直阻塞,由于这个原因,apply_async()更适合并发执行,另外,func函数仅被pool中的一个进程运行。

apply_async(func[, args[, kwds[, callback[, error_callback]]]]) : apply()方法的一个变体,会返回一个结果对象。如果callback被指定,那么callback可以接收一个参数然后被调用,当结果准备好回调时会调用callback,调用失败时,则用error_callback替换callback。 Callbacks应被立即完成,否则处理结果的线程会被阻塞。

close() : 阻止更多的任务提交到pool,待任务完成后,工作进程会退出。

terminate() : 不管任务是否完成,立即停止工作进程。在对pool对象进程垃圾回收的时候,会立即调用terminate()。

join() : wait工作线程的退出,在调用join()前,必须调用close() or terminate()。这样是因为被终止的进程需要被父进程调用wait(join等价与wait),否则进程会成为僵尸进程。

map(func, iterable[, chunksize])¶
map_async(func, iterable[, chunksize[, callback[, error_callback]]])¶
imap(func, iterable[, chunksize])¶
imap_unordered(func, iterable[, chunksize])
starmap(func, iterable[, chunksize])¶
starmap_async(func, iterable[, chunksize[, callback[, error_back]]])

python 协程

线程和进程的操作是由程序触发系统接口,最后的执行者是系统;协程的操作则是程序员。


协程存在的意义:对于多线程应用,CPU通过切片的方式来切换线程间的执行,线程切换时需要耗时(保存状态,下次继续)。协程,则只使用一个线程,在一个线程中规定某个代码块执行顺序。


协程的适用场景:当程序中存在大量不需要CPU的操作时(IO),适用于协程;


event loop是协程执行的控制点, 如果你希望执行协程, 就需要用到它们。


event loop提供了如下的特性:

注册、执行、取消延时调用(异步函数)

创建用于通信的client和server协议(工具)

创建和别的程序通信的子进程和协议(工具)

把函数调用送入线程池中

协程示例:

import asyncio
   
async def cor1():
    print("COR1 start")
    await cor2()
    print("COR1 end")
   
async def cor2():
    print("COR2")
   
loop = asyncio.get_event_loop()
loop.run_until_complete(cor1())
loop.close()

最后三行是重点。

asyncio.get_event_loop()  : asyncio启动默认的event loop 

run_until_complete()  :  这个函数是阻塞执行的,知道所有的异步函数执行完成,

close()  :  关闭event loop。

1、greenlet

import greenlet
def fun1():
    print("12")
    gr2.switch()
    print("56")
    gr2.switch()
  
def fun2():
    print("34")
    gr1.switch()
    print("78")
  
  
gr1 = greenlet.greenlet(fun1)
gr2 = greenlet.greenlet(fun2)
gr1.switch()

2、gevent

gevent属于第三方模块需要下载安装包

pip3 install --upgrade pip3
pip3 install gevent
import gevent
def fun1():
    print("www.baidu.com")   # 第一步
    gevent.sleep(0)
    print("end the baidu.com")  # 第三步
  
def fun2():
    print("www.zhihu.com")   # 第二步
    gevent.sleep(0)
    print("end th zhihu.com")  # 第四步
  
gevent.joinall([
    gevent.spawn(fun1),
    gevent.spawn(fun2),
])

遇到IO操作自动切换:

import gevent
import requests
def func(url):
    print("get: %s"%url)
    gevent.sleep(0)
    date =requests.get(url)
    ret = date.text
    print(url,len(ret))
gevent.joinall([
    gevent.spawn(func, &#39;https://www.pythontab.com/&#39;),
    gevent.spawn(func, &#39;https://www.yahoo.com/&#39;),
    gevent.spawn(func, &#39;https://github.com/&#39;),
])


文章转自:http://www.cnblogs.com/aylin/p/5601969.html


Statement:
The content of this article is voluntarily contributed by netizens, and the copyright belongs to the original author. This site does not assume corresponding legal responsibility. If you find any content suspected of plagiarism or infringement, please contact admin@php.cn