查询阿里巴巴关键字排名

记录下自己实现阿里巴巴关键词排名查询的主要代码及pyquery、urllib库的基础用法

1. urllib 基础模块的应用，通过该类获取到url中的html文档信息，内部可以重写代理的获取方法

   class ProxyScrapy(object):
   def init(self):

   self.proxy_robot = ProxyRobot()
   self.current_proxy = None
   self.cookie = cookielib.CookieJar()
   

   def __builder_proxy_cookie_opener(self):

   cookie_handler = urllib2.HTTPCookieProcessor(self.cookie)        
   handlers = [cookie_handler]
   
   if PROXY_ENABLE:
       self.current_proxy = ip_port = self.proxy_robot.get_random_proxy()
       proxy_handler = urllib2.ProxyHandler({'http': ip_port[7:]})
       handlers.append(proxy_handler)
   
   opener = urllib2.build_opener(*handlers)
   urllib2.install_opener(opener)
   return opener
   

   def get_html_body(self,url):

   opener = self.__builder_proxy_cookie_opener()
   
   request=urllib2.Request(url)
   #request.add_header("Accept-Encoding", "gzip,deflate,sdch")
   #request.add_header("Accept", "text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8")
   #request.add_header("Cache-Control", "no-cache")
   #request.add_header("Connection", "keep-alive")
   
   try:
       response = opener.open(request,timeout=2)
   
       http_code = response.getcode()
       if http_code == 200:
           if PROXY_ENABLE:
               self.proxy_robot.handle_success_proxy(self.current_proxy)
           html = response.read()
           return html
       else:
           if PROXY_ENABLE:
               self.proxy_robot.handle_double_proxy(self.current_proxy)
           return self.get_html_body(url)
   except Exception as inst:
       print inst,self.current_proxy
       self.proxy_robot.handle_double_proxy(self.current_proxy)
       return self.get_html_body(url)
   

2. [代码]根据输入的公司名及关键词列表，返回每个关键词的排名

   def search_keywords_rank(keyword_company_name, keywords):
   def get_context(url):

   start=clock()
   html=curl.get_html_body(url)
   finish=clock()
   print url,(finish-start)
   
   d = pq(html)
   items = d("#J-items-content .ls-item")
   items_c = len(items)
   print items_c
   if items_c < 38:
       return get_context(url)
   return items, items_c
   

   result = OrderedDict()
   for keyword in keywords:

   for page_index in range(1,9):
       u = url % (re.sub('\s+', '_', keyword.strip()), page_index)
       items, items_c = get_context(u)
       b = False
       for item_index in range(0, items_c):
           e=items.eq(item_index).find('.title a')
           p_title = e.text()
           p_url = e.attr('href')
   
           e=items.eq(item_index).find('.cright h3 .dot-product')
           company_name = e.text()
           company_url = e.attr('href')
   
           if  keyword_company_name in company_url:
               total_index = (page_index-1)*38 +item_index+1+(0 if page_index==1 else 5)
               print 'page %s, index %s, total index %s' % (page_index, item_index+1, total_index)
               b = True
               if keyword not in result:
                   result[keyword] = (p_title, p_url, page_index, item_index+1, total_index, u)
               break
       if b:
           break
   

   return result

http://www.freeb2bmarket.com/keywordtool/

由于想放在web服务器上面，这个web页面实际上是用tonardo做的， 在php和tonado同用一个端口的话又用了nginx做前端代理服务器，对nginx的设置也是这次应用的一个学习点吧 ;)