pyspider爬取tripadvisor

#!/usr/bin/env python
# -*- encoding: utf-8 -*-
# Created on 2018-08-23 08:49:39
# Project: pyspiderlianxi

from pyspider.libs.base_handler import *
import pymongo


class Handler(BaseHandler):
    crawl_config = {
    }
    
    client = pymongo.MongoClient(host=‘localhost‘,port=27017)
    db = client[‘trip‘]


    @every(minutes=24 * 60)
    def on_start(self):
        self.crawl(‘https://www.tripadvisor.cn/Attractions-g294211-Activities-c47-China.html‘,callBACk=self.index_page,validate_cert=falsE)
  
     #爬取索引页
    @config(age=10 * 24 * 60 * 60)
    def index_page(self,responsE):
        for each in response.doc(‘div.lisTing_info > div.lisTing_title > a‘).items():
            self.crawl(each.attr.href,callBACk=self.detail_page,validate_cert=falsE)
       
   
     爬取详情页#
    @config(priority=2)
    def detail_page(self,responsE):
        url = response.url,title = response.doc(‘title‘).text(),name = response.doc(‘#heading‘).text(),paiming = response.doc(‘span.header_rating > div > a > span‘).text(),phonenum = response.doc(‘div.blEntry.phone > span:nth-child(2)‘).text(),dizhi = response.doc(‘div.detail_section.address.xh-highlight‘).text(),youwanshijian = response.doc(‘#taplc_attraction_detail_lisTing_0 > div.section.hours > div‘).text() 
          
        return {
             "url":url,"title"@R_458_6964@,"name":name,"paiming":paiming,"@H_274_85@phonenum":phonenum,"dizhi":dizhi,"youwanshijian":youwanshijian
           
        }
    #结果存入数据库中
    def on_result(self,result):
        if result:
            self.save_to_mongo(result)
           
    def save_to_mongo(self,result):
        if self.db[‘chinaStrip‘].insert(result):
            print(‘save to mongo‘,result)