//20200115
最近在看“咱们裸熊——we bears”第一季和第三季都看完了,单单就第二季死活找不到,只有腾讯有资源,但是要vip……而且还是国语版……所以就瞄上了一个视频网站——可以在线观看(好的动漫喜欢收藏,就想着下载,第一季第三季都找到了资源,甚至第四季都有,就没有第二季……)
最近又正好在学python(为大数据打基础),就想着爬取视频,下面说说流程:
首先F12检查,先看看是否是直接嵌入链接(以防真的有笨笨的web主~),然后发现没有
然后就开始点开Networks检查抓包,发现有后缀为.m3u8的链接,就点开看了——有两层,第二层是一大堆格式化数据
然后再看剩下的包,都是.ts文件,再以.ts文件链接比对第二个m3u8包里的东西,发现正好对应,只是要拼接字符串获取真实链接,确认了思路之后就开始上手了(只是基础爬取,还未用到线程——其实用线程池技术可以更快,毕竟ts文件很多,也未用到代理,因为数据量还不算大,而且有手动限时)
理一下思路:
先从视频播放界面源码中获取每一集的链接,存在列表里备用(这个是显示的)---->然后获取每一个链接对应网址的源码——里边儿有一个ckplayer的div块,里边儿有第一层m3u8的链接 ----> 用beautifulSoup获取到这个链接(这个链接返回的是一个json,用json包转格式获取到第一层链接) -----> 访问这个链接获取到第二个m3u8链接(其中要拼接字符串)----->然后访问第二个链接获取到ts视频地址信息(也要拼接字符串——拼接完成后存储到列表中备用)----->使用文件输出流将ts文件下载并存在对应文件夹内
接下来就是等待了,等它下完,因为文件很细碎,所以耗时很久……可以考虑使用线程池改进(等我把大数据基础学完了再说,不急)
然后在每一个ts文件夹中用windows命令copy/b *.ts video.mp4将ts文件合并为mp4文件——可以嵌入到python代码中,不过我没有bat基础,就直接手动了,也不会太困难(大功告成!)
下面上源码:
source code:
#20200115 import requests import json import time from bs4 import BeautifulSoup headers = {‘User-Agent‘:‘Mozilla/5.0 (Windows NT 6.1; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/72.0.3626.81 Safari/537.36‘} location = ‘https://www.****.cc‘ mid = "/1000k/hls" last = "/index.m3u8" url_pool = [‘/dianshiju/20740/player-1-1.html‘, ‘/dianshiju/20740/player-1-2.html‘, ‘/dianshiju/20740/player-1-3.html‘, ‘/dianshiju/20740/player-1-4.html‘, ‘/dianshiju/20740/player-1-5.html‘, ‘/dianshiju/20740/player-1-6.html‘, ‘/dianshiju/20740/player-1-7.html‘, ‘/dianshiju/20740/player-1-8.html‘, ‘/dianshiju/20740/player-1-9.html‘, ‘/dianshiju/20740/player-1-10.html‘, ‘/dianshiju/20740/player-1-11.html‘, ‘/dianshiju/20740/player-1-12.html‘, ‘/dianshiju/20740/player-1-13.html‘, ‘/dianshiju/20740/player-1-14.html‘, ‘/dianshiju/20740/player-1-15.html‘, ‘/dianshiju/20740/player-1-16.html‘, ‘/dianshiju/20740/player-1-17.html‘, ‘/dianshiju/20740/player-1-18.html‘, ‘/dianshiju/20740/player-1-19.html‘, ‘/dianshiju/20740/player-1-20.html‘, ‘/dianshiju/20740/player-1-21.html‘, ‘/dianshiju/20740/player-1-22.html‘] len1 = len(url_pool) def get_json_url(soup): url = soup.find("div",id = "iFrame_play").script.get(‘src‘) return location + url def get_first_url(json_url): r2 = requests.get(json_url,headers = headers,timeout = 10).text dic = json.loads(r2[r2.find(‘{‘):r2.find(‘}‘)+1]) return dic[‘url‘] def get_real_m3u8_url(url): index_of_last = url.rfind(‘/‘) the_forward = url[:index_of_last] return the_forward + mid def get_the_ts_pack(url): r3 = requests.get(url,headers = headers,timeout = 10).text list_of_ts = r3.split(‘#‘) return list_of_ts def get_each_ts_url(the_ts_pack,the_real_m38u_url): len2 = len(the_ts_pack) for i in range(0,len2): suffix = the_ts_pack[i].split(‘\n‘)[1] the_ts_pack[i] = the_real_m38u_url + "/" + suffix # return the_ts_pack def mission(url,n,group): print(‘*****‘) response=requests.get(url,headers=headers,timeout = 10) print(‘-----‘) f=open("./"+str(group)+"/%03d.ts"%n,"wb") f.write(response.content) f.close() print("%03d.ts OK..."%n) def download(the_ts_pack,group): len3 = len(the_ts_pack) count = 0 i = -1 while i != len3-1: try: i+=1 mission(the_ts_pack[i],i,group) except (requests.exceptions.ConnectionError,requests.exceptions.ReadTimeout): count+=1 print("第"+str(count)+"次等待") time.sleep(5) i-=1 else: count=0 time.sleep(0.5) # for i in range(0,len1): for i in range(12,22): completed_link = location + url_pool[i] r1 = requests.get(completed_link,headers=headers,timeout = 10) soup = BeautifulSoup(r1.text,"lxml") json_url = get_json_url(soup) time.sleep(0.1) the_first_mu38_url = get_first_url(json_url) time.sleep(0.1) the_real_m38u_url = get_real_m3u8_url(the_first_mu38_url) the_ts_pack = get_the_ts_pack(the_real_m38u_url + last)[5:-1] get_each_ts_url(the_ts_pack,the_real_m38u_url) print(the_ts_pack) download(the_ts_pack,i) print("第" + str(i) + "组ts视频已经下载完成") time.sleep(10) # # # list1 = str1.rfind(‘/‘) # str2 = str1[:list1] # print(str2) # for i in range() # for each in url_pool: # print(each) # # for n in range(1,167): # mission(link + str(8000+n)+".ts",n) # dic = {‘%3A‘:‘:‘,‘%2F‘:"/"} # str1 = str1.replace(‘%3A‘,‘:‘) # str1 = str1.replace(‘%2F‘,‘/‘) # print(str1) # # # r = requests.get(link,headers = headers,timeout = 10) # text = r.text # print(text) #
注:因为视频有版权,网站地址就不放出来了,重要的是思路,每个网站都不一样,都要重新分析
侵删!
关于python异常机制:
1.try-except代码块,就是如果有异常就执行except里的代码,然后如果有循环就跳过这一次(显然不符合要求,因为要下齐资源,所以要用到2)
2.try-except-else代码块,如果有异常,就执行except内代码,如果没有,执行完try中代码后,继续执行else中代码
另:except后跟的异常,可以是一个也可以是多个(多个使用“(..,..,..)”这种格式,不知道啥异常可以直接用Exception)
因为代码执行过程中,服务器有的时候会返回不了信息,就要异常来处理,不然每次都手动怎么称得上自动化呢~
希望对大家有所帮助
以上
原文地址:https://www.cnblogs.com/lavender-pansy/p/12198998.html