Python批量查询关键词是否是指数词

简单的查询一个关键词是否为指数词，暂不能查询指数，过段时间研究下

# -*- coding: utf-8 -*-

import urllib2

import urllib

import cookielib

import re

import bs4

URL_BAIDU_INDEX = u'http://www.baidu.com/';

#https://passport.baidu.com/v2/api/?getapi&class=login&tpl=mn&tangram=true 也可以用这个

URL_BAIDU_TOKEN = 'https://passport.baidu.com/v2/api/?getapi&tpl=pp&apiver=v3&class=login';

URL_BAIDU_LOGIN = 'https://passport.baidu.com/v2/api/?login';

#设置用户名、密码

username = '';

password = '';

#设置cookie，这里cookiejar可自动管理，无需手动指定

cj = cookielib.CookieJar();

opener = urllib2.build_opener(urllib2.HTTPCookieProcessor(cj));

urllib2.install_opener(opener);

reqReturn = urllib2.urlopen(URL_BAIDU_INDEX);

#获取token,

tokenReturn = urllib2.urlopen(URL_BAIDU_TOKEN);

matchVal = re.search(u'"token" : "(?P.*?)"',tokenReturn.read());

tokenVal = matchVal.group('tokenVal');

#构造登录请求参数，该请求数据是通过抓包获得，对应https://passport.baidu.com/v2/api/?login请求

postData = {

'username' : username,

'password' : password,

'u' : 'https://passport.baidu.com/',

'tpl' : 'pp',

'token' : tokenVal,

'staticpage' : 'https://passport.baidu.com/static/passpc-account/html/v3Jump.html',

'isPhone' : 'false',

'charset' : 'UTF-8',

'callback' : 'parent.bd__pcbs__ra48vi'

};

postData = urllib.urlencode(postData);

#发送登录请求

loginRequest = urllib2.Request(URL_BAIDU_LOGIN,postData);

loginRequest.add_header('Accept','text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8');

loginRequest.add_header('Accept-Encoding','gzip,deflate,sdch');

loginRequest.add_header('Accept-Language','zh-CN,zh;q=0.8');

loginRequest.add_header('User-Agent','Mozilla/5.0 (Windows NT 6.1; WOW64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/28.0.1500.72 Safari/537.36');

loginRequest.add_header('Content-Type','application/x-www-form-urlencoded');

sendPost = urllib2.urlopen(loginRequest);

#获取指数页面

keylist = open('key.txt','r') #查询关键词的文件

i = '如要查看相关数据'

for key in keylist:

key = key.strip('\n')

teibaUrl = 'http://index.baidu.com/?tpl=trend&word={0}'.format(key)

content = urllib2.urlopen(teibaUrl).read();

content = content.decode('gbk').encode('utf8');

if i in content:

print key,'不是指数关键词'

else:

print key,'指数关键词'

喜欢python，SEO的同学可以加QQ群472779600