Python爬蟲之m3u8文件裡提取小視頻的正確姿勢

# 未加密
	#EXTM3U
	#EXT-X-VERSION:3
	#EXT-X-TARGETDURATION:8
	#EXT-X-MEDIA-SEQUENCE:0
	#EXTINF:4.000000,
	1af12fece7a000000.ts
	#EXTINF:4.320000,
	1af12fece7a000001.ts
	...
	#EXTINF:3.800000,
	1af12fece7a001155.ts
	#EXT-X-ENDLIST
	
	# 加密
	#EXTM3U
	#EXT-X-VERSION:3
	#EXT-X-TARGETDURATION:6
	#EXT-X-PLAYLIST-TYPE:VOD
	#EXT-X-MEDIA-SEQUENCE:0
	#EXT-X-KEY:METHOD=AES-128,URI="https://ts1.yuyuangewh.com:9999/20200808/1XdSSbTb/2000kb/hls/key.key"
	#EXTINF:3,
	https://ts1.yuyuangewh.com:9999/20200808/1XdSSbTb/2000kb/hls/EUtRrqJU.ts
	#EXTINF:4.72,
	https://ts1.yuyuangewh.com:9999/20200808/1XdSSbTb/2000kb/hls/HF90vrrN.ts
	...
	#EXTINF:0.24,
	https://ts1.yuyuangewh.com:9999/20200808/1XdSSbTb/2000kb/hls/b7ZLcRqT.ts
	#EXT-X-ENDLIST

# 安裝m3u8
	pip install m3u8

# 加載m3u8文件
	import m3u8

	# 返回一個M3U8對象
	playlist = m3u8.load(uri='http://videoserver.com/playlist.m3u8')	# url
	# playlist = m3u8.load(uri='playlist.m3u8')	# file
	print(playlist.segments)	# 打印EXT-X-KEY標簽和所有的EXTINF標簽:
	print(playlist.target_duration)	# 打印EXT-X-TARGETDURATION標簽的值

	for key in playlist.keys:
		if key:
			# 如果視頻文件加密，可以查看加密參數
			print(key.uri, key.method, key.iv)

# 第1步，我載瞭m3u8文件，也可以直接使用m3u8文件對應的url
	playlist = m3u8.load(uri='./data/index.m3u8')

	# 第2步，提取URL
	for seg in playlist.segments:
        print(seg.uri)
    
    # 第3步，下載ts視頻
    with open('xxxxx.ts', 'wb') as f:
        ts = get_ts(url)
        f.write(ts)
   
	# 第4步，解密
	cipher_text = pad(data_to_pad=cipher_text, block_size=AES.block_size)
    aes = AES.new(key=key, mode=AES.MODE_CBC, iv=iv)
    cipher_text = aes.decrypt(cipher_text)
    
	# 第5步，合成
	files = glob.glob(os.path.join('./video', '*.ts'))
    for file in files:
        with open(file, 'rb') as fr, open('./video_de/baitoushan.mp4', 'ab') as fw:
            content = fr.read()
            fw.write(content )

# -*- coding: utf-8 -*-
# @Time    : 2021/5/10 20:11
# @Author  : XiaYouRan
# @Email   : [email protected]
# @File    : video.py
# @Software: PyCharm


from Crypto.Cipher import AES
from Crypto.Util.Padding import pad
from concurrent.futures import ThreadPoolExecutor
import requests
import m3u8
import glob
import os
import time
import logging


logging.getLogger("urllib3").setLevel(logging.WARNING)


def AESDecrypt(cipher_text, key, iv):
    cipher_text = pad(data_to_pad=cipher_text, block_size=AES.block_size)
    aes = AES.new(key=key, mode=AES.MODE_CBC, iv=iv)
    cipher_text = aes.decrypt(cipher_text)
    # clear_text = unpad(padded_data=cipher_text, block_size=AES.block_size)
    return cipher_text


headers = {'User-Agent': 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) '
                         'AppleWebKit/537.36 (KHTML, like Gecko) Chrome/71.0.3578.98 Safari/537.36'}


def get_ts(url):
    try:
        response = requests.get(url, verify=False)
        response.raise_for_status()
        response.encoding = 'utf-8'
        return response.content
    except Exception as err:
        print(err)
        return b''


def save_ts(url, index):
    filename = os.path.join('./video', str(index).zfill(5) + '.ts')
    with open(filename, 'wb') as f:
        ts = get_ts(url)
        f.write(ts)
    print(filename + ' is ok!')


if __name__ == '__main__':
    playlist = m3u8.load(uri='./data/index.m3u8')

	# 線程池，引入index可以防止合成時視頻發生亂序
    with ThreadPoolExecutor(max_workers=10) as pool:
        for index, seg in enumerate(playlist.segments):
            pool.submit(save_ts, seg.uri, index)

    key = get_ts(playlist.keys[-1].uri)

    files = glob.glob(os.path.join('./video', '*.ts'))
    for file in files:
        with open(file, 'rb') as fr, open('./video_de/baitoushan.mp4', 'ab') as fw:
            content = fr.read()
            encontent = AESDecrypt(content, key=key, iv=key)
            fw.write(encontent)
        print(file + ' is ok!')