python B站弹幕爬取

B站弹幕简单爬虫
功能：获取视频弹幕并保存到txt文档
使用方法：找到b站视频所在的aid，传入到main函数下的av='一串数字，即aid'即可
找aid方法：视频下有个转发按钮，鼠标放上去可以看到有个嵌入代码iframe，那里就有aid。或者检查、network、刷新一下，在Name栏可以找到包含aid的网址

"""
首先获取avid
从pagelist里获取，先将av空格去掉，然后获取url，获取res里的text，用json转换成字典，返回
然后根据cid获取弹幕
首先获取url，然后解码成utf-8，然后用正则表达式，然后用findall寻找所有，然后返回
存储
"""
import requests
import json
import re


def getcid(av):
    av = av.strip('av')
    url = f'https://api.bilibili.com/x/player/pagelist?aid={av}&jsonp=jsonp'
    res = requests.get(url)
    res = res.text
    print(res)
    res_dict = json.loads(res)
    cid = res_dict['data'][0]['cid']
    return cid


def getdanmu(cid):
    url = f'https://api.bilibili.com/x/v1/dm/list.so?oid={cid}'
    res = requests.get(url)
    de = res.content.decode('utf-8')
    ch = re.compile('<d.*?>(.*?)</d>')
    danmu = ch.findall(de)
    return danmu


def savedanmu(danmu, filename):
    with open(filename, mode='w', encoding='utf-8') as f:
        for w in danmu:
            f.write(w)
            f.write(' ')


#  <iframe src="//player.bilibili.com/player.html?aid=545259775&bvid=BV1Sq4y1E7HQ&cid=331019617&page=1"
# scrolling="no" border="0" frameborder="no" framespacing="0" allowfullscreen="true"> </iframe>
if __name__ == '__main__':
    av = '890452353'
    cid = getcid(av)
    danmu = getdanmu(cid)
    savedanmu(danmu, f'{av}.txt')

****