利用python批量查询企业信息_python实现批量查询SEM排名

import requests

import re

import time

headers={

'User-Agent': 'Mozilla/5.0 (Linux; Android 5.0; SM-G900P Build/LRX21T) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/78.0.3904.108 Mobile Safari/537.36',

'Cookie': 'BAIDUID=0EF6827460D96115613857315CEA5EE4:FG=1; PSTM=1565022011; BIDUPSID=D4C2C53F0375AD2422E22EF13B3574D8; MSA_WH=360_640; SIGNIN_UC=70a2711cf1d3d9b1a82d2f87d633bd8a03254009288; BDORZ=B490B5EBF6F3CD402E515D22BCDA1598; BD_UPN=12314753; H_WISE_SIDS=136721_128700_136626_114745_139211_120169_138489_138813_133995_138878_137978_132910_137690_131247_136680_118895_118859_118857_118823_118790_136687_107316_138883_136431_139044_136863_138146_138308_138114_139178_136196_137104_133847_138477_137735_138343_137467_139123_131423_139255_139246_136165_138661_136032_110085_127969_138621_139162_137830_138804_138275_127416_138312_137187_136635_138426_138563_138943_138317_100458; delPer=0; ASUV=1.2.114; lsv=globalT_androidcss_866df54-wwwT_androidcss_b47c606-searchboxcss_591d86b-globalBcss_565c244-wwwBcss_777000e; SE_LAUNCH=5%3A26259329; MSA_PBT=146; MSA_ZOOM=1000; wpr=0; BD_CK_SAM=1; PSINO=1; H_PS_PSSID=1441_21079_30211_20883_30087_29700_30071; FC_MODEL=-1_0_0_0_0_0_0_0_0_0_0_-1_1_1_1_0_0_1575559757666_1575559750502%7C1%230_-1_-1_1_1_1575559757666_1575559750502%7C1; BD_HOME=0; __yjsv5_shitong=1.0_7_6cabd481c2ceeffad94e1178c2ae3b394423_300_1575560100385_183.194.155.96_249a8302; yjs_js_security_passport=e6c8a790f5b160a4d1999c02cb4d9e9027326e87_1575560101_js; COOKIE_SESSION=83_0_8_9_1_1_0_0_8_1_0_0_0_0_0_0_1575560360_0_1575560437%7C9%23125435_43_1574609627%7C9; H_PS_645EC=2be3OfBW040FDRvGpQkpYAls%2Fp5dOuA2DOReoNuyyc8urN4JPdqRktTbxV8'

}

wd=open('gjc.txt',encoding='utf-8').read() #gjc.txt放入要查询的关键词

keywords=wd.split('\n')

for word in keywords:

url="https://www.baidu.com/s?ie=utf-8&f=8&rsv_bp=1&tn=baidu&wd=%s"%word

res=requests.get(url,headers=headers)

time.sleep(2)

res.encoding='utf-8'

con=res.text

pattern=re.compile('(.*?)', re.S)

result = pattern.findall(con)

m=0

for i in result:

m+=1

print(word+','+str(m)+','+str(i[1]+','+str(i[0])))