#?-*-?coding:?utf-8?-*-
“““
Created?on?Mon?Aug?20?12:23:53?2018

@author:?linzhenglai
“““
#?-*-?coding:?utf-8?-*-
“““
Created?on?Mon?Aug?20?11:11:18?2018

@author:?linzhenglai
“““
#?-*-?coding:?utf-8?-*-
“““
Spyder?Editor

This?is?a?temporary?script?file.
“““

#?-*-?coding:?utf-8?-*-
“““
Spyder?Editor

This?is?a?temporary?script?file.
“““

import?time
import?requests
#爬取網(wǎng)頁時直接出現(xiàn)403，意思是沒有訪問權限
from?bs4?import?BeautifulSoup
import?json
#入口網(wǎng)頁
import?xlwt

url=‘http://chs.meituan.com/‘

headers={
????????‘Accept‘:‘text/htmlapplication/xhtml+xmlapplication/xml;q=0.9image/webpimage/apng*/*;q=0.8‘
????????‘Accept-Language‘:‘zh-CNzh;q=0.9‘
????????‘Cache-Control‘:‘max-age=0‘
????????‘Proxy-Connection‘:‘keep-alive‘
????????‘Host‘:‘chs.meituan.com‘
????????‘Referer‘:‘http://chs.meituan.com/‘
????????‘Upgrade-Insecure-Requests‘:‘1‘
????????‘User-Agent‘:‘Mozilla/5.0?（Windows?NT?6.1;?Win64;?x64）?AppleWebKit/537.36?（KHTML?like?Gecko）?Chrome/67.0.3396.99?Safari/537.36‘
????????‘Content-Type‘:‘text/html;charset=utf-8‘
????????‘Cookie‘:?‘_lxsdk_cuid=164c9bed44ac8-0bf488e0cbc5d9-5b193413-1fa400-164c9bed44bc8;?__mta=248363576.1532393090021.1532393090021.1532393090021.1;?rvct=70%2C1;?ci=70;?iuuid=30CB504DBAC7CCDD72645E3809496C48229D8143D427C01A5532A4DDB0D42388;?cityname=%E9%95%BF%E6%B2%99;?_lxsdk=30CB504DBAC7CCDD72645E3809496C48229D8143D427C01A5532A4DDB0D42388;?_ga=GA1.2.1889738019.1532505689;?uuid=2b2adb1787947dbe0888.1534733150.0.0.0;?oc=d4TCN9aIiRPd6Py96Y94AGxfsjATZHPGsCDua9-Z_NQHsXDcp6WlG2x7iJpYzpSLttNvEucwm_D_SuJ7VRJkLcjqV6Nk8s_q3VyOJw5IsVJ6RJPL3qCgybGW3vxTkMHr9A4yChReTafbZ7f93F1PkCyUeFBQV4D-YXoVoFV5h3o;?_lx_utm=utm_source%3DBaidu%26utm_medium%3Dorganic;?client-id=97664882-24cd-4743-b21c-d25de878708e;?lat=28.189822;?lng=112.97422;?_lxsdk_s=165553df04a-bc8-311-ba7%7C%7C6‘
}

#with?open（r‘美團西安美食.csv‘“w“?newline=‘‘encoding=‘UTF-8‘）?as?csvfile
#獲取主頁源碼（獲取分類-美食、電影）
#def?get_start_links（url）:
????#html=requests.get（url）.text#發(fā)送請求獲取主頁文本
????#print?html
????#soup=BeautifulSoup（html‘lxml‘）#解析網(wǎng)頁
????#==$
????#
????#????#links=‘http://chs.meituan.com/meishi/?
???#?links=[link.find（‘span‘）.find（‘a(chǎn)‘）[‘href‘]?for?link?in?soup.find_all（‘span‘class_=‘nav-text-wrapper‘）]
????#print?links
????#return?links
#獲取分類鏈接中間的店鋪id
#find:取第一個（返回一個具體的元素，沒有為null）find_all:匹配所有（返回列表，沒有返回[]）
def?get_detail_id（category_url）:
????html=requests.get（category_urlheaders=headers）.text
????#print?html
????soup=BeautifulSoup（html‘lxml‘）#解析網(wǎng)頁
????#print?soup
????texts=soup.find_all（‘script‘）
????#print?texts
????text=texts[14].get_text（）.strip（）
????#print?text
????text=text[19:-1]
????
????result1=json.loads（textencoding=‘utf-8‘）
????#print?result1
????result2=result1[‘poiLists‘]
????result3=result2[‘poiInfos‘]
????#print?result3
????Info_list=[]
????for?it?in?result3:
????????#print?it
????????Info_list.append（it[‘poiId‘]