#coding:utf8
import urllib2,sys,re
import threading,os
import time,datetime
'''''
这里没有使用队列 只是采用多线程分发对代理量不大的网页还行但是几百几千性能就很差了
'''
def get_proxy_page(url):
'''''解析代理页面 获取所有代理地址'''
proxy_list = []
p = re.compile(r'''''<div>(.+?)<span class="Apple-tab-span" style="white-space:pre">.*?</span>(.+?)<span class="Apple-tab-span" style="white-space:pre">.+?</span>(.+?)(<span.+?)?</div>''')
try:
res = urllib2.urlopen(url)
except urllib2.URLError:
print 'url Error'
sys.exit(1)
pageinfo = res.read()
res = p.findall(pageinfo) #取出所有的
#组合成所有代理服务器列表成一个符合规则的list
for i in res:
ip = i[0]
port = i[1]
addr = i[2]
l = (ip,port,addr)
proxy_list.append(l)
return proxy_list
#同步锁装饰器
lock = threading.Lock()
def synchronous(f):
def call(*args, **kw):
lock.acquire()
try:
return f(*args, **kw)
finally:
lock.release()
return call
#时间计算器
def sumtime(f):
def call(*args, **kw):
t1 = time.time()
try:
return f(*args, **kw)
finally:
print u'总共用时 %s' % (time.time() - t1)
return call
proxylist = []
reslist = []
#获取单个代理并处理
@synchronous
def getoneproxy():
global proxylist
if len(proxylist)>0:
return proxylist.pop()
else:
return ''
#添加验证成功的代理
@synchronous
def getreslist(proxy):
global reslist
if not (proxy in reslist):
reslist.append(proxy)
def handle():
timeout = 10
test_url = r'http://www.baidu.com'
test_str = '030173'
while 1:
proxy = getoneproxy()
#最后一个返回是空
if not proxy:
return
print u"正在验证 : %s" %proxy[0]
#第一步启用 cookie
cookies = urllib2.HTTPCookieProcessor()
proxy_server = r'http://%s:%s' %(proxy[0],proxy[1])
#第二步 装载代理
proxy_hander = urllib2.ProxyHandler({"http":proxy_server})
#第三步 组合request
try:
opener = urllib2.build_opener(cookies, proxy_hander)
pass
except urllib2.URLError:
print u'url设置错误'
continue
#配置request
opener.addheaders = [('User-Agent','Mozilla/5.0 (Windows NT 6.1; WOW64) AppleWebKit/537.1 (KHTML, like Gecko) Chrome/21.0.1180.89 Safari/537.1')]
#发送请求
urllib2.install_opener(opener)
t1 = time.time()
try:
req = urllib2.urlopen(test_url,timeout=timeout)
result = req.read()
pos = result.find(test_str)
timeused = time.time() - t1
if pos>1:
#保存到列表中
getreslist((proxy[0],proxy[1],proxy[2],timeused))
print u'成功采集',proxy[0],timeused
else:
continue
except Exception,e:
print u'采集失败 %s ：timeout' %proxy[0]
continue
def save(reslist):
path = os.getcwd()
filename = path + '/Proxy-'+datetime.datetime.now().strftime(r'%Y%m%d%H%M%S')+'.txt'
f = open(filename,'w+')
for proxy in reslist:
f.write('%s %s %s %s \r\n'%(proxy[0],proxy[1],proxy[2],proxy[3]))
f.close()
@sumtime
def main():
url = r'http://www.free998.net/daili/httpdaili/8949.html'
global proxylist,reslist
#获取所有线程
proxylist = get_proxy_page(url)
print u'一共获取 %s 个代理' %len(proxylist)
#print proxylist
print '*'*80
#线程创建和分发任务
print u'开始创建线程处理.....'
threads = []
proxy_num = len(proxylist)
for i in range(proxy_num):
th = threading.Thread(target=handle, args=())
threads.append(th)
for thread in threads:
thread.start()
for thread in threads:
threading.Thread.join(thread)
print u'获取有效代理 %s 个,现在开始排序和保存 '%len(reslist)
reslist = sorted(reslist,cmp=lambda x,y:cmp(x[3],y[3]))
save(reslist)
if __name__=='__main__':
main()