爬取自如租房信息

image.png

image.png

{
"min_lng": "116.236046",#最小的经度
"max_lng": "116.443016",#最大的经度
"min_lat": "39.934614",#最小的纬度
"max_lat": "40.02617",#最大的纬度
"clng": "116.339531",#中心的经度
"clat": "39.980407",#中心的纬度
"zoom": "14"#地图放大的级别
}

返回的数据为

image.png

点击小区进去(比如望京)
HTTP请求url为http://www.ziroom.com/map/room/count?min_lng=116.444611&max_lng=116.496353&min_lat=39.989225&max_lat=40.012107&clng=116.470482&clat=40.000667&zoom=16
参数为

{
"min_lng": "116.444611",
"max_lng": "116.496353",
"min_lat": "39.989225",
"max_lat": "40.012107",
"clng": "116.470482",#与上面图中的一致
"clat": "40.000667",#与上面图中一致
"zoom": "16"
}

小区的最大最小经纬度如何确定的，不知道，再点开一个小区(这是亚运村小营)，提交的参数为

{
"min_lng": "116.394021",
"max_lng": "116.445763",
"min_lat": "39.980199",
"max_lat": "40.003084",
"clng": "116.419892",
"clat": "39.991642",
"zoom": "16"
}

不难发现这两组参数中的max_lng-clng和max-clat为一个固定值
这样通过clng和clat就可以确定min_lng,max_lng,min_lat,max_lat的值了

下面直接上代码

#-*- coding:utf-8 -*-
import requests,re,pickle,sys,json
from bs4 import BeautifulSoup
from requests.packages import urllib3
import pandas as pd
urllib3.disable_warnings()
from pymongo import MongoClient
#mongodb数据库，用于将爬取的租房信息存下来
mOngo= MongoClient('localhost',27017).zufang.ziru
#全局变量
room_info_list = []
headers = {
'User-Agent':'Mozilla/5.0 (Windows NT 6.1; Win64; x64; rv:68.0) Gecko/20100101 Firefox/68.0',
'Origin':'http://www.ziroom.com'
}
#requests的会话对象
s = requests.Session()
#爬取每个小区的租房信息
def spider(info):
room_info_list = []
#当前页
pg = 1
#总的页数
pages = 1
url = 'http://www.ziroom.com/map/room/list'
#自如将地图划分成一个个矩形的小格子，每个小格式表示一个小区的范围，小格子的区域使用经纬度来表示
#提交的参数
#通过观察可以发现所有提交的参数中，格子边界距离中心的距离为固定值，这样就比较容易构造参数了
lat_diff = 0.011443
lng_diff = 0.025871
#这里所有参数值均为字符串
params = {
"max_lng": "%.6f" % (info['touch_lng'] + lng_diff),#最大经度
"min_lng": "%.6f" % (info['touch_lng'] - lng_diff),#最小经度
"max_lat": "%.6f" % (info['touch_lat'] + lat_diff),#最大纬度
"min_lat": "%.6f" % (info['touch_lat'] - lat_diff),#最小纬度
"clng": "%.6f" % info['touch_lng'],
"clat": "%.6f" % info['touch_lat'],
"zoom":"16",
"p":"1"
}

while pg <= pages:
params['p'] = str(pg)
r = s.get(url=url, params=params,headers=headers)
if r.status_code == 200:
data = r.json()
pages = data['data']['pages']

for item in data['data']['rooms']:
try:
#将结果存入数据库
mongo.update(
{"id":item['id']},
{
'$set':item
},
upsert = True
)
#将结果存储为csv格式
item['location'] = item['location'][0]['name']
for k,v in item.items():
#将list类型转化为str类型
if str(type(v)) == "":
item[k] = str(v)

#将//开关的url转化为http://开头
if str(type(v)) == "" and v.startswith('//'):
item[k] = 'http:' + v
room_info_list.append(item)
print(item['name'])
except:
pass
print(info['name'],pg,pages)
pg += 1
#将结果导入为csv格式，每个小区一个csv文件，方便筛选
if room_info_list:
pickle.dump(room_info_list,open('ziru_result/%s.db' % info['name'],'wb'))
df = pd.DataFrame(room_info_list)[['name','desc','price','location','detail_url']]
df.to_excel('ziru_result/%s.xls' % info['name'])



def main():
#获取北京所有小区的租房信息,
url = 'http://www.ziroom.com/map/room/count?min_lng=116.228373&max_lng=116.486653&min_lat=40.019069&max_lat=40.123091&clng=116.357513&clat=40.0711&zoom=14'
r = s.get(url,headers = headers)
if r.status_code == 200:
data = r.json()['data']
for item in data:
#爬取每个小区的租房信息
print(item)
spider(item)
main()
结果

爬取的结果