bilibili网站爬虫

# -*- coding: utf-8 -*-
"""
Created on Mon May 07 11:05:49 2018
B站爬虫：
功能： 1 获取评论人数，点赞数，用户个人信息及评论等等
@author: Alis
"""

import re,time
import requests
import os
import json


headers = {'user-agents':'User-Agent:Mozilla/5.0 (Windows NT 6.1; rv:2.0.1) Gecko/20100101 Firefox/4.0.1'}


def getavID(pn):
    avID = []
    for i in range(1,pn+1):
        url = 'https://api.bilibili.com/x/web-interface/dynamic/region?&jsonp=jsonp&pn=%d&ps=50&rid=24&_=1525679623909'%i
        r = requests.get(url,headers = headers).text
        data = json.loads(r)
        archives = data['data']['archives']
        for ac in archives:
            avID.append(ac['aid'])
            print 'aid: ',ac['aid']
            print 'title: ',ac['title']
            print 'attrib