Spider实战系列-抓取《某人之下第三季》(spider 系列)

南窗分类：IT技术人气：139 回帖：0 发布于1年前收藏

今天我们抓取的是m3u8的视频,视频有长视频和短视频之分.

抓取m3u8类型视频

对于短视频

一般来说一个视频对应的就是一个url

长视频

一个视频就几百兆到几十G不等这种视频往往不会由一个连接进行全部返回视频是由多个片段组成的每个片段大概是几秒到几分钟

所以对于长视频的下载没有办法通过一个url进行获取下载，需要获取每个频段的url 进行下载进行拼接最终变成一个完整的视频

然而视频还有加密之分

我们使用合并视频的工具 ffmpeg ，如果没有加密的视频，直接根据每个片段的url进行下载拼接即可如果有加密的视频根据加密的地址请求加密的key，然后在进行合并，FFmpeg会自动进行解密

网站链接:下面就会有网站链接,我就不放了

分析

第一次响应的地址: /20211213/fl6Q2MC2/1100kb/hls/index.m3u8

第二次请求m3u8:https://v4.cdtlas.com/20211213/fl6Q2MC2/1100kb/hls/index.m3u8

第一次请求m3u8:https://v4.cdtlas.com/20211213/fl6Q2MC2/index.m3u8

也就是是说第一次请求的域名加上响应就是第二次的地址

第二次的响应情况如下

找的页面里第一个m3u8的地址src="/js/player/?url=https://v4.cdtlas.com/20211213/fl6Q2MC2/index.m3u8&id=60825&num=1&count=8&vt=1"

思路

通过页面找到第一个m3u8的url 进行请求返回的响应中包含了第二次m3u8的部分url拼接进行请求
获取响应中包含所有片段ts的url对于所有ts的url进行请求，下载，获取片段
片段进行拼接，组装成视频，视频下载完成

导包

import requests
import re

代码实现

通过网页源代码得到m3u8的地址

url = 'https://www.wbdyba.com/play/60825_1_1.html'

headers = {
    # 'Referer': 'https://www.wbdyba.com/play/60825_1_1.html',
    'User-Agent': 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/109.0.0.0 Safari/537.36 Edg/109.0.1518.78'
}

# 把网页源代码下载到本地
res = requests.get(url, headers)

# with open('The_Outcast.html', 'wb')as f:
#     f.write(res.content)
with open('The_Outcast.html','r',encoding='utf-8')as f:
    data = f.read()

base_m3u8_url = re.search(
    '<iframe id="mplay" name="mplay" src="/js/player/?url=(.*?)" '
    'allowfullscreen="allowfullscreen" mozallowfullscreen="mozallowfullscreen" msallowfullscreen="msallowfullscreen" oallowfullscreen="oallowfullscreen" webkitallowfullscreen="webkitallowfullscreen" frameBorder="no" width="100%" scrolling="no" height="100%">',data).group(1)
m3u8_url = base_m3u8_url.split('&')[0]

在这里我为什么要切片切出来呢?因为源码正则得到的m3u8地址不可用

# https://v4.cdtlas.com/20211213/fl6Q2MC2/index.m3u8&id=60825&num=1&count=8&vt=1 源码正则得到的
# https://v4.cdtlas.com/20211213/fl6Q2MC2/index.m3u8                             网页发送请求的

当我在得到这两个url的时候,我也不确定哪一个能请求到真正的视频资源

所以我在下面用了两中验证来确定了,我们要的是浏览器能请求的m3u8地址

res1= requests.get('https://v4.cdtlas.com/20211213/fl6Q2MC2/index.m3u8&id=60825&num=1&count=8&vt=1',headers=headers)
res2 = requests.get('https://v4.cdtlas.com/20211213/fl6Q2MC2/index.m3u8',headers=headers)

print(res1.content.decode())
print(res2.content.decode())
if (res1==res2):
    print('success')
else:
    print('false')

结果如下

D:\Tools\Tool\python\python.exe D:/Tools/Workspace/PyWorkspace/dateProject/demo08-下载m3u8视频/一人之下.py
file not found
#EXTM3U
#EXT-X-STREAM-INF:PROGRAM-ID=1,BANDWIDTH=1100000,RESOLUTION=864x486
/20211213/fl6Q2MC2/1100kb/hls/index.m3u8

false

请求m3u8地址,拿到响应并得到第二次请求

进行m3u8第一次请求   目的为了获取第二次请求的m3U8的地址
res1 = requests.get(m3u8_url,headers=headers)
with open('index1.m3u8', 'wb', )as f:
    f.write(res1.content)

with open('index1.m3u8', 'r', encoding='utf-8')as f:
    data_index = f.readlines()[-1].strip()
# print(data_index)
'''
https://v4.cdtlas.com/20211213/fl6Q2MC2/1100kb/hls/index.m3u8
'''
# 拼接第二次请求的url 获得第二次请求的m3u8文件
url2 = urljoin(m3u8_url, data_index)
res2 = requests.get(url2,headers = headers)
with open('index2.m3u8', 'wb', )as f:
    f.write(res2.content)

封装得到m3u8的方法

def get_m3u8(url):
    res = requests.get(url, headers)
    base_m3u8_url = re.search(
        '<iframe id="mplay" name="mplay" src="/js/player/?url=(.*?)" '
        'allowfullscreen="allowfullscreen" mozallowfullscreen="mozallowfullscreen" msallowfullscreen="msallowfullscreen" oallowfullscreen="oallowfullscreen" webkitallowfullscreen="webkitallowfullscreen" frameBorder="no" width="100%" scrolling="no" height="100%">',
        res.content.decode()).group(1)
    m3u8_url = base_m3u8_url.split('&')[0]

    # 进行m3u8第一次请求   目的为了获取第二次请求的m3U8的地址
    res1 = requests.get(m3u8_url, headers=headers)
    with open('index1.m3u8', 'wb', )as f:
        f.write(res1.content)

    with open('index1.m3u8', 'r', encoding='utf-8')as f:
        data_index = f.readlines()[-1].strip()
    # print(data_index)
    '''
    https://v4.cdtlas.com/20211213/fl6Q2MC2/1100kb/hls/index.m3u8
    '''
    # 拼接第二次请求的url 获得第二次请求的m3u8文件
    url2 = urljoin(m3u8_url, data_index)
    res2 = requests.get(url2, headers=headers)
    with open('index2.m3u8', 'wb', )as f:
        f.write(res2.content)
    return url2

封装下载ts片段的方法

单进程下载

下载所有的ts

def download_all():
    # 下载存储ts文件的路径
    path = 'ts'
    if not os.path.exists(path):
        os.mkdir(path)
    with open('index2.m3u8', 'r', encoding='utf-8')as f:
        data = f.readlines()
    # print(data)

    for i in data:
        '''
        这里还可以协程
        if l.startwith('#'):
            continue
        '''
        if (i.startswith('https')):
            # 取出url结尾的换行符
            ts_url = i.strip()
            down_one(ts_url, path)

下载单个ts

def down_one(ts_url, path):
    print(ts_url, '===正在下载')
    '''
    https://hey07.cjkypo.com/20211213/fl6Q2MC2/1100kb/hls/Vuzb3l70.ts
    https://hey07.cjkypo.com/20211213/fl6Q2MC2/1100kb/hls/0HiRyylQ.ts
    https://hey07.cjkypo.com/20211213/fl6Q2MC2/1100kb/hls/Z6W1fh5y.ts
    https://hey07.cjkypo.com/20211213/fl6Q2MC2/1100kb/hls/u9ZW3cN2.ts
    '''
    name = ts_url.split('/')[-1]
    res = requests.get(ts_url, headers=headers)
    # 进行下载储存
    with open(os.path.join(path, name), 'wb')as f:
        f.write(res.content)

多任务下载

from concurrent.futures import ThreadPoolExecutor,wait


def down_one(ts_url, path):
    while 1:
        try:
            print(ts_url, '===正在下载')
            '''
            https://hey07.cjkypo.com/20211213/fl6Q2MC2/1100kb/hls/Vuzb3l70.ts
            https://hey07.cjkypo.com/20211213/fl6Q2MC2/1100kb/hls/0HiRyylQ.ts
            https://hey07.cjkypo.com/20211213/fl6Q2MC2/1100kb/hls/Z6W1fh5y.ts
            https://hey07.cjkypo.com/20211213/fl6Q2MC2/1100kb/hls/u9ZW3cN2.ts
            '''
            name = ts_url.split('/')[-1]
            # 请求ts的url如果60秒还没下载完进行重新下载
            res = requests.get(ts_url, headers=headers,timeout=60)
            # 进行下载储存
            with open(os.path.join(path, name), 'wb')as f:
                f.write(res.content)
            print(ts_url, '===下载成功')
            break
        except:
            print(url,'下载失败,重新在下载')


def download_all():
    # 下载存储ts文件的路径
    path = 'ts'
    if not os.path.exists(path):
        os.mkdir(path)
    with open('index2.m3u8', 'r', encoding='utf-8')as f:
        data = f.readlines()
    # print(data)
    #创建线程池
    pool = ThreadPoolExecutor(100)
    tasks = []# 装任务的列表

    for i in data:
        '''
        这里还可以协程
        if l.startwith('#'):
            continue
        '''
        if (i.startswith('https')):
            # 取出url结尾的换行符
            ts_url = i.strip()
            # 添m任务到线程池
            tasks.append(pool.submit(down_one(ts_url, path)))
        # 阻塞等地      所有都下载完毕再向下执行
        wait(tasks)

处理下载后ts文件和m3u8文件的对应关系

# 处理下载后ts文件和m3u8文件的对应关系
def do_m3u8_url():
    with open('index2.m3u8', 'r', encoding='utf-8')as f:
        data = f.readlines()
    f = open(os.path.join(path, 'index3.m3u8'), 'w', encoding='UTF-8')
    for i in data:
        if i.startswith('#'):
            # 这一步操作是把index3.m3u8文件里的key换成我们自己的
            f.write(i)
            if i.find('URI')!=-1:
                new_i = i.replace('https://hey07.cjkypo.com/20211213/fl6Q2MC2/1100kb/hls/key.key','key.m3u8')
                f.write(new_i)
        else:
            f.write(i.split('/')[-1])

进行合并处理

def merge(filename='output'):
    '''
    进行ts文件合并 解决视频音频不同步的问题 建议使用这种
    :param filePath:
    :return:
    '''
    os.chdir(path)
    cmd = f'D:/Tools/Tool/ffmpeg-4.4.1-essentials_build/bin/ffmpeg -i index.m3u8 -c copy {filename}.mp4'
    os.system(cmd)

下载密钥

# 下载密钥
def get_key():
    resp = requests.get('https://hey07.cjkypo.com/20211213/fl6Q2MC2/1100kb/hls/key.key')
    f = open(os.path.join(path, 'key.m3u8'), 'w', encoding='UTF-8')

    f.write(resp.text)

main函数

if __name__ == '__main__':
    url = 'https://www.wbdyba.com/play/60825_1_1.html'

    headers = {
        'Referer': 'https://www.wbdyba.com/play/60825_1_1.html',
        'User-Agent': 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/109.0.0.0 Safari/537.36 Edg/109.0.1518.78'
    }

    # 下载存储ts文件的路径
    path = 'ts'
    # 获取m3u8文件内容也就是所有的ts 的url
    get_m3u8(url)
    download_all()
    do_m3u8_url()
    # 进行合并处理
    merge()

更改为协程版本

import os
import re
from urllib.parse import urljoin

import requests
from aiohttp import ClientSession, TCPConnector
import asyncio
import aiofiles


async def get_m3u8(url):
    async with ClientSession(connector=TCPConnector(ssl=False), headers=headers) as session:
        async with session.get(url)as resp:
            data = await resp.text(encoding='utf-8')
            base_m3u8_url = re.search('<iframe id="mplay" name="mplay" src="/js/player/?url=(.*?)" '
                                      'allowfullscreen="allowfullscreen" mozallowfullscreen="mozallowfullscreen" msallowfullscreen="msallowfullscreen" oallowfullscreen="oallowfullscreen" webkitallowfullscreen="webkitallowfullscreen" frameBorder="no" width="100%" scrolling="no" height="100%">'
                                      , data).group(1)
            m3u8_url = base_m3u8_url.split('&')[0]
            async with session.get(m3u8_url)as resp:
                data = await resp.text(encoding='utf-8')
                # 进行m3u8第一次请求   目的为了获取第二次请求的m3U8的地址
                async with aiofiles.open('index.m3u8', 'w', encoding='utf-8')as f:
                    await f.write(data)
                async with aiofiles.open('index.m3u8', 'r', encoding='utf-8')as f:
                    data_index = await f.readlines()
                    # 进行第二次url请求 url的获取
                    data_index = data_index[-1].strip()
                    # print(data_index)
                    '''
                    https://v4.cdtlas.com/20211213/fl6Q2MC2/1100kb/hls/index.m3u8
                    '''
                    # 拼接第二次请求的url 获得第二次请求的m3u8文件
                    url2 = urljoin(m3u8_url, data_index)
                    async with session.get(url2)as resp:
                        data = await resp.text(encoding='utf-8')
                        async with aiofiles.open('index.m3u8', 'w', encoding='utf-8')as f:
                            await f.write(data)
    return url2


async def down_one(ts_url, path, sem):
    while 1:
        # 使用信号量控制并发
        async with sem:
            try:
                # 请求ts的url如果60秒还没下载完进行重新下载
                async with ClientSession(connector=TCPConnector(ssl=False), headers=headers)as session:
                    print(ts_url, '===正在下载')
                    name = ts_url.split('/')[-1]
                    '''
                    https://hey07.cjkypo.com/20211213/fl6Q2MC2/1100kb/hls/Vuzb3l70.ts
                    https://hey07.cjkypo.com/20211213/fl6Q2MC2/1100kb/hls/0HiRyylQ.ts
                    https://hey07.cjkypo.com/20211213/fl6Q2MC2/1100kb/hls/Z6W1fh5y.ts
                    https://hey07.cjkypo.com/20211213/fl6Q2MC2/1100kb/hls/u9ZW3cN2.ts
                    '''
                    async with session.get(ts_url, timeout=60)as resp:
                        data = await resp.read()
                        # 进行下载储存
                        async with aiofiles.open(os.path.join(path, name), 'wb')as f:
                            await f.write(data)
                        print(ts_url, '===下载成功')
                        break
            except:
                print(ts_url, '下载失败,重新在下载')


async def download_all():
    with open('index.m3u8', 'r', encoding='utf-8')as f:
        data = f.readlines()
    if not os.path.exists(path):
        os.mkdir(path)
    # print(data)
    # 创建线程池
    sem = asyncio.Semaphore(200)
    tasks = []  # 装任务的列表

    for i in data:
        '''
        这里还可以协程
        if l.startwith('#'):
            continue
        '''
        if (i.startswith('https')):
            # 取出url结尾的换行符
            ts_url = i.strip()
            # 添m任务到线程池
            tasks.append(asyncio.create_task(down_one(ts_url, path, sem)))
        # 阻塞等地      所有都下载完毕再向下执行
    await asyncio.wait(tasks)


# 处理下载后ts文件和m3u8文件的对应关系
def do_m3u8_url():
    with open('index.m3u8', 'r', encoding='utf-8')as f:
        data = f.readlines()
    f = open(os.path.join(path, 'index.m3u8'), 'w', encoding='UTF-8')
    for i in data:
        if i.startswith('#'):
            # 这一步操作是把index.m3u8文件里的key换成我们自己的
            f.write(i)
            if i.find('URI')!=-1:
                new_i = i.replace('https://hey07.cjkypo.com/20211213/fl6Q2MC2/1100kb/hls/key.key','key.m3u8')
                f.write(new_i)
        else:
            f.write(i.split('/')[-1])

# 下载密钥
def get_key():
    resp = requests.get('https://hey07.cjkypo.com/20211213/fl6Q2MC2/1100kb/hls/key.key')
    f = open(os.path.join(path, 'key.m3u8'), 'w', encoding='UTF-8')

    f.write(resp.text)


def merge(filename='output'):
    '''
    进行ts文件合并 解决视频音频不同步的问题 建议使用这种
    :param filePath:
    :return:
    '''
    os.chdir(path)
    cmd = f'D:/Tools/Tool/ffmpeg-4.4.1-essentials_build/bin/ffmpeg -i index.m3u8 -c copy {filename}.mp4'
    os.system(cmd)


async def main():
    # 获取m3u8文件内容也就是所有的ts 的url
    task = asyncio.create_task(get_m3u8(url))
    await asyncio.gather(task)
    task = asyncio.create_task(download_all())
    await asyncio.gather(task)
    do_m3u8_url()
    get_key()
    # 进行合并处理
    merge()


if __name__ == '__main__':
    url = 'https://www.wbdyba.com/play/60825_1_1.html'

    headers = {
        'Referer': 'https://www.wbdyba.com/play/60825_1_1.html',
        'User-Agent': 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/109.0.0.0 Safari/537.36 Edg/109.0.1518.78'
    }

    # 下载存储ts文件的路径
    path = 'ts'

    loop = asyncio.get_event_loop()
    loop.run_until_complete(main())

密钥解密

# 下载密钥
def get_key():
    resp = requests.get('https://hey07.cjkypo.com/20211213/fl6Q2MC2/1100kb/hls/key.key')
    f = open(os.path.join(path, 'key.key'), 'w', encoding='UTF-8')
    f.write(resp.text)

if i.find('URI')!=-1:
        new_i = i.replace('https://hey07.cjkypo.com/20211213/fl6Q2MC2/1100kb/hls/key.key','key.m3u8')
        f.write(new_i)
else:

这里我直接使用的取巧的方法了,直接使用查找字符串并改变字符串

写完这个方法后在main函数内调用一次就好了

总结

在我们请求到m3u8地址时,源码内抓到的跟浏览器抓包工具里抓到的不一样,我们就需要把从源码得到的地址转成浏览器一样的地址,如果遇到ts文件中带key的要把他解密放在ts文件夹里.在密钥解密这里需要注意一点

ts文件夹的结构应该为

project/

ts/

0.ts

1.ts

...

index.m3u8

key.m3u8

密钥内容为key.m3u8

#EXTM3U
#EXT-X-VERSION:3
#EXT-X-TARGETDURATION:1
#EXT-X-PLAYLIST-TYPE:VOD
#EXT-X-MEDIA-SEQUENCE:0
#EXT-X-KEY:METHOD=AES-128,URI="key.m3u8"
#EXTINF:1.235,
0.ts
#EXTINF:1.001,
1.ts
#EXTINF:1.001,
2.ts

写在最后

我只抓取了第一集,小伙伴像抓取其他集,只需要用一个for循环输入不同的url即可

标签：短视频,TypeScript,PHP,爬虫,https,爬虫

分析

思路