python threading之互斥锁

2015年7月13日 发表评论 阅读评论

python threading模块有两类锁:互斥锁(threading.Lock )和可重用锁(threading.RLock)。两者的用法基本相同,具体如下:

lock = threading.Lock()
lock.acquire()
dosomething……
lock.release()

RLock的用法是将threading.Lock()修改为threading.RLock()。便于理解,先来段代码:

[root@361way lock]# cat lock1.py
#!/usr/bin/env python
# coding=utf-8
import threading                            # 导入threading模块
import time                             # 导入time模块
class mythread(threading.Thread):        # 通过继承创建类
    def __init__(self,threadname):      # 初始化方法
        # 调用父类的初始化方法
        threading.Thread.__init__(self,name = threadname)
    def run(self):                          # 重载run方法
        global x                  # 使用global表明x为全局变量
        for i in range(3):
            x = x + 1
        time.sleep(5)          # 调用sleep函数,让线程休眠5秒
        print x
tl = []                              # 定义列表
for i in range(10):
    t = mythread(str(i))               # 类实例化
    tl.append(t)                      # 将类对象添加到列表中
x=0                                 # 将x赋值为0
for i in tl:
    i.start()  

这里执行的结果和想想的不同,结果如下:

[root@361way lock]# python lock1.py
30
30
30
30
30
30
30
30
30
30

为什么结果都是30呢?关键在于global 行和 time.sleep行。

1、由于x是一个全局变量,所以每次循环后 x 的值都是执行后的结果值;

2、由于该代码是多线程的操作,所以在sleep 等待的时候,之前已经执行完成的线程会在这等待,而后续的进程在等待的5秒这段时间也执行完成 ,等待print。同样由于global 的原理,x被重新斌值。所以打印出的结果全是30 ;

3、便于理解,可以尝试将sleep等注释,你再看下结果,就会发现有不同。

在实际应用中,如抓取程序等,也会出现类似于sleep等待的情况。在前后调用有顺序或打印有输出的时候,就会现并发竞争,造成结果或输出紊乱。这里就引入了锁的概念,上面的代码修改下,如下:

[root@361way lock]# cat lock2.py
#!/usr/bin/env python
# coding=utf-8
import threading                            # 导入threading模块
import time                             # 导入time模块
class mythread(threading.Thread):                   # 通过继承创建类
    def __init__(self,threadname):                  # 初始化方法
        threading.Thread.__init__(self,name = threadname)
    def run(self):                          # 重载run方法
        global x                        # 使用global表明x为全局变量
        lock.acquire()                      # 调用lock的acquire方法
        for i in range(3):
            x = x + 1
        time.sleep(5)           # 调用sleep函数,让线程休眠5秒
        print x
        lock.release()                # 调用lock的release方法
lock = threading.Lock()               # 类实例化
tl = []                          # 定义列表
for i in range(10):
    t = mythread(str(i))            # 类实例化
    tl.append(t)              # 将类对象添加到列表中
x=0                        # 将x赋值为0
for i in tl:
    i.start()                     # 依次运行线程

执行的结果如下:

[root@361way lock]# python lock2.py
3
6
9
12
15
18
21
24
27
30

加锁的结果会造成阻塞,而且会造成开锁大。会根据顺序由并发的多线程按顺序输出,如果后面的线程执行过快,需要等待前面的进程结束后其才能结束 --- 写的貌似有点像队列的概念了 ,不过在加锁的很多场景下确实可以通过队列去解决。

最后,再引入一个示例,在股票量化分析(二)PE和流通市值篇中,介绍了如何采集stock的两个指标,并按结果输出,不过在输出的时候发有会出现输出紊乱,如下:

threading-lock

如600131和000708的stockid就输出到了同一行,虽然通过多线程使执行速度快了很多 ,但这样很不美观,也不便于后续处理。

1、输出竞争紊乱代码

#!/usr/bin/python
#coding=utf-8
# 1、pe在 0~20 之间的企业
# 2、流通股本小于50亿的企业
import urllib2
import time
import json
from threading import Thread
def get_pe(stockid):
    try:
        url = 'http://d.10jqka.com.cn/v2/realhead/hs_%s/last.js' % stockid
        send_headers = {
            'Host':'d.10jqka.com.cn',
            'Referer':'http://stock.10jqka.com.cn/',
            'Accept':'application/json, text/javascript, */*; q=0.01',
            'Connection':'keep-alive',
            'User-Agent':'Mozilla/5.0 (Windows NT 6.1; WOW64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/38.0.2125.122 Safari/537.36',
            'X-Forwarded-For':'124.160.148.178',
            'X-Requested-With':'XMLHttpRequest'
        }
        req = urllib2.Request(url,headers=send_headers)
        f = urllib2.urlopen(req)
        data = f.read().split('items":',1)[1]
        data = data.split('})',1)[0]
        J_data = json.loads(data)
        #J_data = json.dumps(data,indent=4,encoding='utf-8')
        stockpe = J_data['2034120']
        stockname = J_data['name']
        sumvalue = J_data['3475914']
        currentprice = J_data['10']
        #print stockid,stockname,stockpe
        return stockname,stockpe,sumvalue,currentprice
    except urllib2.HTTPError, e:
        #return stockid ,'get happed httperror'
        return e.code
def cond(stockid,pe,asset):
    pe = int(pe)
    asset = int(asset)
    try:
        stockname,stockpe,sumvalue,currentprice = get_pe(stockid)
        if sumvalue:
           Billvalue = round(float(sumvalue)/1000/1000/100)
        else:
           Billvalue = 0
        if stockpe:
           if float(stockpe) > 0 and float(stockpe) < pe and Billvalue < asset :
              print stockid,stockname,currentprice,stockpe,Billvalue
        #else:
        #   print stockid
    except TypeError ,e:
        print stockid ,'get is error'
if __name__ == '__main__':
    threads = []
    print 'stockid  stockname  currentprice  stockpe  Billvalue'
    stockids = [line.strip() for line in open("stock_exp.txt", 'r')]
    nloops = range(len(stockids))
    for stockid in stockids:
        t = Thread(target=cond, args=(stockid,28,80))
        threads.append(t)
    for i in nloops:
        threads[i].start()
    for i in nloops:
        threads[i].join()

2、加锁后的代码

#!/usr/bin/python
#coding=utf-8
# 1、pe在 0~20 之间的企业
# 2、流通股本小于50亿的企业
import threading
import urllib2
import time
import json
lock = threading.Lock()
def get_pe(stockid):
    try:
        url = 'http://d.10jqka.com.cn/v2/realhead/hs_%s/last.js' % stockid
        send_headers = {
            'Host':'d.10jqka.com.cn',
            'Referer':'http://stock.10jqka.com.cn/',
            'Accept':'application/json, text/javascript, */*; q=0.01',
            'Connection':'keep-alive',
            'User-Agent':'Mozilla/5.0 (Windows NT 6.1; WOW64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/38.0.2125.122 Safari/537.36',
            'X-Forwarded-For':'124.160.148.178',
            'X-Requested-With':'XMLHttpRequest'
        }
        req = urllib2.Request(url,headers=send_headers)
        f = urllib2.urlopen(req)
        data = f.read().split('items":',1)[1]
        data = data.split('})',1)[0]
        J_data = json.loads(data)
        #J_data = json.dumps(data,indent=4,encoding='utf-8')
        stockpe = J_data['2034120']
        stockname = J_data['name']
        sumvalue = J_data['3475914']
        currentprice = J_data['10']
        #print stockid,stockname,stockpe
        return stockname,stockpe,sumvalue,currentprice
    except urllib2.HTTPError, e:
        #return stockid ,'get happed httperror'
        return e.code
def cond(stockid,pe,asset):
    pe = int(pe)
    asset = int(asset)
    try:
        stockname,stockpe,sumvalue,currentprice = get_pe(stockid)
        if sumvalue:
           Billvalue = round(float(sumvalue)/1000/1000/100)
        else:
           Billvalue = 0
        if stockpe:
           if float(stockpe) > 0 and float(stockpe) < pe and Billvalue < asset :
              lock.acquire()
              print stockid,stockname,currentprice,stockpe,Billvalue
              lock.release()
        #else:
        #   print stockid
    except TypeError ,e:
        print stockid ,'get is error'
if __name__ == '__main__':
    threads = []
    print 'stockid  stockname  currentprice  stockpe  Billvalue'
    stockids = [line.strip() for line in open("stock_exp.txt", 'r')]
    for stockid in stockids:
        t = threading.Thread(target=cond, args=(stockid,25,50))
        threads.append(t)
        t.start()




本站的发展离不开您的资助,金额随意,欢迎来赏!

You can donate through PayPal.
My paypal id: itybku@139.com
Paypal page: https://www.paypal.me/361way

  1. 本文目前尚无任何评论.
  1. 本文目前尚无任何 trackbacks 和 pingbacks.