#coding:utf-8
import requests
import browsercookie
from urllib.parse import urlencode
import pandas as pd
from datetime import datetime

class BaiduIndex(object):
    def __init__(self,kw,startDate,endDate):
        self.kw = kw
        self.startDate = startDate
        self.endDate = endDate
        self.headers = {'User-Agent':'Mozilla/5.0 (Windows NT 6.1) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/45.0.2454.101 Safari/537.36'}
        self.cookies = browsercookie.chrome()
        self.baidu_indexs = self.get_baidu_indexs()
        self.categories = {'all':'total_index','pc':'pc_index','wise':'mobile_index'}

    def get_baidu_indexs(self):
        request_args = {
                    'word': self.kw,
                    'startDate': self.startDate,
                    'endDate': self.endDate,
                    'area': 0,
                }
        url = 'http://index.baidu.com/api/SearchApi/index?{}'.format(urlencode(request_args))
        # print(url)
        html = requests.get(url,headers = self.headers,cookies = self.cookies).json()
        # print(html)
        return html

    #獲取關(guān)鍵詞某段時(shí)間內(nèi)的平均指數(shù)
    def avg_baidu_index(self):
        datas = self.baidu_indexs
        if datas.get('data'):
            datas = datas['data']['generalRatio'][0]
            # datas = html['data']['generalRatio'][0]
            total_index = datas['all']['avg']
            pc_index = datas['pc']['avg']
            mobile_index = datas['wise']['avg']
            return {'total_index':total_index,'mobile_index':mobile_index,'pc_index':pc_index}
        else:
            return {'total_index':0,'mobile_index':0,'pc_index':0}

    # 獲取關(guān)鍵詞某段時(shí)間內(nèi)每天的指數(shù)
    def daily_baidu_index(self):
        datelist = self.datelist()
        datas = self.baidu_indexs
        if datas.get('data'):
            uniqid = datas['data']['uniqid']
            datas = datas['data']['userIndexes'][0]
            key = self.get_key(uniqid)
            d = {}
            for category,value in self.categories.items():
                data = datas[category]['data']
                # print(key,data)
                index = self.decrypt(key,data)
                num = len(index)
                # print(len(datelist),num,len(datelist[-num:]))
                indexs = {date:index for date,index in zip(datelist[-num:],index)}
                d[value] = indexs
            return d
        else:
            d = {}
            for category, value in self.categories.items():
                indexs = {date:0 for date in datelist}
                d[value] = indexs
            return d

    def get_key(self,uniqid):
        url = 'http://index.baidu.com/Interface/api/ptbk?uniqid={}'.format(uniqid)
        html = requests.get(url,headers = self.headers,cookies=self.cookies).json()
        return html['data']

    # 根據(jù)起始日期生成每日日期
    def datelist(self):
        datelist =[datetime.strftime(x,'%Y-%m-%d') for x in list(pd.date_range(start=self.startDate, end=self.endDate))]
        return datelist

    # 解密百度指數(shù)
    def decrypt(self,key,data):
        num = {}
        datas = []
        for line in range(len(key) // 2):
            num[key[line]] = key[len(key) // 2 + line]
        for dt in range(len(data)):
            datas.append(num[data[dt]])
        return ''.join(datas).split(',')

if __name__ == "__main__":
    b = BaiduIndex('周杰倫','2020-03-15','2020-03-17')
    daily_index = b.daily_baidu_index()   #某個(gè)時(shí)間段每一天的指數(shù)
    print(daily_index)
    print(daily_index.get('total_index').get('2020-03-16'))  #2020年3月16日的整體指數(shù)
    print(daily_index.get('pc_index').get('2020-03-16'))  # 2020年3月16日的PC指數(shù)
    print(daily_index.get('mobile_index').get('2020-03-16'))  # 2020年3月16日的移動(dòng)指數(shù)
    avg_index = b.avg_baidu_index()     #某個(gè)時(shí)間段平均指數(shù)
    print(avg_index)

輸出結(jié)果如下所示：

{'total_index': {'2020-03-15': '18201', '2020-03-16': '18934', '2020-03-17': '19095'}, 'pc_index': {'2020-03-15': '3288', '2020-03-16': '3912', '2020-03-17': '3969'}, 'mobile_index': {'2020-03-15': '14913', '2020-03-16': '15022', '2020-03-17': '15126'}}
18934
3912
15022
{'total_index': 18743, 'mobile_index': 15020, 'pc_index': 3723}

可以看到，可查詢到某段時(shí)間內(nèi)平均指數(shù)以及每天的指數(shù)，包括總指數(shù)，PC指數(shù)以及移動(dòng)端指數(shù)。

如果想查某一天的指數(shù)，設(shè)置startDate和endDate是同一日期即可。如下所示：

if __name__ == "__main__":
    b = BaiduIndex('周杰倫', '2020-03-16', '2020-03-16')
    avg_index = b.avg_baidu_index()
    print(avg_index)

輸出：

{'total_index': 18934, 'mobile_index': 15022, 'pc_index': 3912}

這里僅以單個(gè)關(guān)鍵詞查詢作示例，至于改成批量查詢及將查詢結(jié)果寫(xiě)入文件，有Python基礎(chǔ)的話不是什么難事兒，這里不再示例。

沒(méi)有大量查詢測(cè)試，但很可能會(huì)遇到封賬號(hào)問(wèn)題，可嘗試換IP、多賬號(hào)查詢等方式。

老規(guī)矩，關(guān)注公眾號(hào)【Python加SEO做增長(zhǎng)】，后臺(tái)回復(fù)關(guān)鍵詞【百度指數(shù)】即可獲得這份代碼，不會(huì)使用可私信我。