shops

#!/usr/bin/env python
#coding:utf-8
import urllib2,sys,re,os,string

reload(sys);
sys.setdefaultencoding('utf8');

#url="http://www.dianping.com/search/category/1/20/g122"

def httpCrawler(url):
#first page
content = httpRequest(url)
#other page
#for pageNo in range(2,50):
#    content = httpRequest(url)
shops=parseHtml(content)
getAllPages(shops)
unpackOneShop()
#saveData(shops)

def httpRequest(url):
try:
html = None
req_header = {
'User-Agent':'Mozilla/5.0 (X11; Linux x86_64; rv:38.0) Gecko/20100101 Firefox/38.0'
#'Accept':'text/html;q=0.9,*/*;q=0.8',
#'Accept-Language':'en-US,en;q=0.5',
#'Accept-Encoding':'gzip',
#'Host':'j3.s2.dpfile.com',
#'Connection':'keep-alive',
#'Referer':'http://www.baidu.com'
}
req_timeout = 5
req = urllib2.Request(url,None,req_header)
resp = urllib2.urlopen(req,None,req_timeout)
html = resp.read().decode('utf-8')
#print html
finally:
if resp:
resp.close()
return html

def parseHtml(html):
content = None
#shops = re.findall(r'<li class="" >(.+?)<\li>',html,re.S)
shops = re.findall(r'<div class="shop-list J_shop-list shop-all-list" id="shop-all-list">\n<ul>.*?</ul>\n</div>',html,re.S)
return shops

def getAllPages(shops):
#total 50 pages
getEachShop(shops)
print "################one page done."

global t_OneShop
t_OneShop=['']*15*50

def getEachShop(shops):
global t_OneShop
t_OneShop=['']*15*50
t_start=0
shops_string=''.join(shops)

i=0
t_start = shops_string.find(r'<li class="" >')
all_end = shops_string.rfind('</li>')
while i<=15 and t_start and all_end:
t_start = shops_string.find(r'<li class="" >',t_start)
t_end = shops_string.find('</li>',t_start,all_end)
#print "t_start:",t_start
#print "t_end:",t_end
t_OneShop[i] = shops_string[t_start:t_end]
#print t_OneShop[i]

t_start=t_end
i=i+1

def unpackOneShop():
global t_OneShop

f = open('./zhubao/shops.csv', 'w')
f.write('\xEF\xBB\xBF')
f.write('名称,地址,人均消费,,,')
f.write('\r\n')
f.close()
for i in range(0,15):
#print t_OneShop[i]

f = open('./zhubao/shops.csv', 'ab+')

ShopName = re.findall(r'<h4>(.*?)</h4>',t_OneShop[i])
#ShopDistrict =
address = re.findall(r'<span class="addr">(.*?)</span>',t_OneShop[i])
mean_price = re.findall(r'mean-price" target="_blank" >(.*?)</span>',t_OneShop[i],re.S)
averageComsumption = re.findall(r'<b>(.*?)</b>',''.join(mean_price),re.S)

print 'mean_price:',mean_price
print 'average::',averageComsumption
ShopName.extend(address)
ShopName.extend(averageComsumption)

print (','.join(ShopName)).replace('\n',''),'\r\n'
f.write((''.join(','.join(ShopName)).replace('\n','')))
f.write('\r\n')
f.close()

#iprovince =
#city =
#adminDistrict =

def saveData(data):
if not os.path.exists('./zhubao'):
os.mkdir(r'./zhubao')
f = open('./zhubao/zhubao_shops.csv', 'wb')
f.write(data)
f.close()

if __name__ == '__main__':
url="http://www.dianping.com/search/category/1/20/g122"
httpCrawler(url)

'''
python2.6 没有urllib.request
多线程
gevent
爬虫系统基本的结构：
1.网络请求；
最简单的工具就是urllib、urllib2。这两个工具可以实现基本的下载功能，如果进阶想要异步可以使用多线程，如果想效率更高采用非阻塞方案tornado和curl可以实现非阻塞的下载。
2.抓取结构化数据；
要想在页面中找到新链接需要对页面解析和对url排重，正则和DOM都可以实现这个功能，看自己熟悉哪一种。
正则感觉速度较快一些，DOM相对较慢并且复杂一点，如果只是为了要url正则可以解决，如果还想要页面中其他的结构或者内容DOM比较方便。
url的排重两小可以用memcache或者redis，量大就要用到bloomfilter。
3.数据存储；
抓的少怎么存都行，抓的多并且要方便读取那就要好好设计了，用哈希分布存储在RDBMS上或者直接存在HBase上都要看你的数据量和具体需求。
'''