基于ffmpeg的m3u8下载[调整key替换逻辑，更新解析逻辑]

闲来没事，找了个av网站，看了几个视频，然后想下载一下。结果发现多年不上这些av网站，现在的av网站播放的源文件已经不是avi或者mp4了，而是m3u8的播放列表。在firefox中可以使用Video DownloadHelper 来获取相应的下载地址，但是有的时候如果m3u8中包含的是播放列表，会无法获取下载链接。

于是就想着怎么直接下载文件，其实通过ffmpeg可以很方便的获取下载链接：

只需要下面的一样命令：

 ffmpeg -protocol_whitelist "file,http,crypto,tcp,https,tls" -i https://videox11.ynkcq.com:8081/20200109/8Pr79HKk/600kb/hls/index.m3u8 -c copy out.mp4

事实证明，这个命令在mac os下是可以直接使用的，我也是后来才发现，下面的代码最开始实在windows下写的。但是在windows下执行这条命令会出现各种问题，网上看了一下貌似是和加密有关，对应的m3u8文件内容：

#EXTM3U
#EXT-X-VERSION:3
#EXT-X-TARGETDURATION:3
#EXT-X-MEDIA-SEQUENCE:0
#EXT-X-KEY:METHOD=AES-128,URI="key.key"
#EXTINF:2.085000,
NT6PfA2975000.ts
#EXTINF:2.085000,
NT6PfA2975001.ts
#EXTINF:2.085000,
NT6PfA2975002.ts
#EXTINF:2.085000,
NT6PfA2975003.ts
#EXTINF:2.085000,
NT6PfA2975004.ts
#EXTINF:2.085000,
NT6PfA2975005.ts
#EXTINF:2.085000,
NT6PfA2975006.ts
#EXTINF:2.085000,
NT6PfA2975007.ts
#EXTINF:2.085000,
NT6PfA2975008.ts
#EXTINF:2.085000,
NT6PfA2975009.ts
#EXTINF:2.085000,
NT6PfA2975010.ts
#EXTINF:2.085000,
NT6PfA2975011.ts
#EXTINF:2.085000,
NT6PfA2975012.ts
#EXTINF:2.085000,
...以下省略...

在windows下由于无法解密会导致输入错误，看了下网上的解决方案，主要有两种

1. 可以将m3u8文件下载到本地，然后替换文件中ts对应的url， 以及key对应的url。然后通过ffmpeg 使用修改后的m3u8文件进行下载，下面的代码就是基于这个逻辑
2. 直接使用代码拼接ts文件，下载完成之后通过openssl进行解密，然后进行拼接。

# -*- coding: UTF-8 -*-

import m3u8
import os, sys
import re, json
import requests
import subprocess
from bs4 import BeautifulSoup
import getopt

"""
https://pypi.org/project/m3u8/
https://blog.csdn.net/u014484783/article/details/79350392

"""

HEADERS = {
    'X-Requested-With': 'XMLHttpRequest',
    'User-Agent': 'Mozilla/5.0 (Windows NT 10.0; WOW64) AppleWebKit/537.36 '
                  '(KHTML, like Gecko) Chrome/56.0.2924.87 Safari/537.36',
}

if sys.platform.startswith('win32'):
    DIR_PATH = r"F:\PyCharmProjects\m3u8_downloader\download\m3u8_files"
    OUT_PATH = r"F:\PyCharmProjects\m3u8_downloader\download\mp4"
    FFMPEG_PATH = 'G:\\ffmpeg-20200126-5e62100-win64-static\\bin\\ffmpeg.exe'
else:
    DIR_PATH = '/Volumes/DATA/python_projects/m3u8_downloader/download'
    OUT_PATH = '/Volumes/DATA/python_projects/m3u8_downloader/download'
    FFMPEG_PATH = '/usr/local/bin/ffmpeg'

def get_play_list(m3u8_uri):
    variant_m3u8 = m3u8._load_from_uri(m3u8_uri)
    print('[A] 解析播放列表.......')
    print('[A] 是否存在播放列表:' + str(variant_m3u8.is_variant))
    play_list = []

    if variant_m3u8.is_variant is False:
        play_list = [m3u8_uri]
        return play_list
    # variant_m3u8.playlists

    for iframe_playlist in variant_m3u8.playlists:
        pl_uri = iframe_playlist.uri
        if str(pl_uri).startswith('/'):
            ul = str(m3u8_uri).split('/')[:3]
            base_uri = '/'.join(ul)
        else:
            ul = str(m3u8_uri).split('/')
            del (ul[-1])
            base_uri = '/'.join(ul) + '/'
        print('[A] Base Url:' + base_uri)
        pl_url = base_uri + pl_uri
        play_list.append(pl_url)
        # print(iframe_playlist.stream_info.subtitles)

        print('[A] 播放列表地址:' + pl_url)
    # print(play_list)
    return play_list


def format_lines(base_url, file_path):
    print('[F] 处理文件路径......')
    m3u8_obj = m3u8._load_from_file(file_path)
    with open(file_path, 'r+') as f:
        content = list(f)
        new_content = []
        for l in content:

            for key in m3u8_obj.keys:
                if key and  key.uri in l and (not str(key.uri).startswith('http')):
                    key_url = base_url + key.uri
                    l = str(l).replace(key.uri, key_url)
                    print('[F] Key Url:' + key_url)

            if '.ts' in l:
                if not str(l).startswith('http'):
                    if str(l).startswith('/'):
                        ul = str(base_url).split('/')[:3]
                        base_uri = '/'.join(ul)
                        l = base_uri + l
                    else:
                        l = base_url + l
                # print('[F] TS Url:' + l)
            new_content.append(l)
        f.seek(0)
        f.truncate()  #
        f.writelines(new_content)
    return content


def get_url_source_code(url):
    print('[G] 获取源码')
    souce = requests.get(url=url, headers=HEADERS, timeout=10, verify=False)
    html = souce.content
    html_doc = str(html, 'utf-8')
    return html_doc


def save_m3u8_file(url, path, file_name):
    try:
        file_path = os.path.join(path, file_name)
        print('[S] 下载m3u8路径:' + path + ' \r\n[S] 链接:' + url)
        print('[S] 文件名:' + file_name)
        durl = url
        if os.path.isfile(file_path):
            print('[F] 文件已经存在，跳过保存')
            return file_path
        img = requests.get(durl, headers=HEADERS, timeout=10).content
        with open(file_path, 'ab') as f:
            f.write(img)
            print('[S] 下载m3u8成功')
        return file_path
    except Exception as e:
        # print(e)
        print('[S] 下载m3u8失败: ' + str(e))
        return None


def get_m3u8_link(url):
    print('_' * 70)
    print('[A] 解析播放地址......')
    html_doc = get_url_source_code(url)
    bs = BeautifulSoup(html_doc, "html.parser")
    pattern = re.compile(r"var cms_player = {(.*?);$", re.MULTILINE | re.DOTALL)
    surls = bs.find('script', text=pattern)
    # print(surls.text)
    js_string = str(surls.text).replace('var cms_player = ', '').replace(';', '')
    json_data = json.loads(js_string)
    m3u8_link = json_data['url']
    title = bs.title.string
    print('[A] 标题:' + title)
    print('[A] 播放地址:' + m3u8_link)
    print('_' * 70)
    return m3u8_link, title


def mark_dir(flot_name):
    """
    检测文件夹是否创建，没有创建则创建文件夹，创建了就跳过
    """
    print('[C] 创建目录: ' + flot_name)
    PATH = os.path.join(DIR_PATH, flot_name)
    if not os.path.exists(PATH):  # 检测是否有这个文件夹
        os.makedirs(PATH)
        os.chdir(PATH)
    return PATH


def download_ts_with_ffmpeg(m3u8_file_path, out_file_path):
    print('[D] 下载文件......')
    print('[D] 文件路径:' + out_file_path)
    cmd_string = FFMPEG_PATH \
                 + ' -protocol_whitelist "file,http,crypto,tcp,https,tls" ' \
                 + ' -i ' + m3u8_file_path \
                 + ' -c copy ' + out_file_path
    print(cmd_string)
    os.system(cmd_string)
    print('[D] 下载完成.')
    print('*' * 100)


def print_usage():
    print('*' * 100)
    print('m3u8 downloader by obaby')
    print('downloader.py -i  -o ')
    print('http://www.h4ck.org.cn')
    print('*' * 100)


def main(argv):
    inputfile = ''
    outputfile = ''
    try:
        opts, args = getopt.getopt(argv, "hi:o:", ["ilink=", "ofile="])
    except getopt.GetoptError:
        print_usage()
        sys.exit(2)
    for opt, arg in opts:
        if opt == '-h':
            print_usage()
            sys.exit()
        elif opt in ("-i", "--ilink"):
            inputfile = arg
        elif opt in ("-o", "--ofile"):
            outputfile = arg

    if inputfile == '':
        print_usage()
        sys.exit(2)

    if inputfile.endswith('.m3u8') or '.m3u8' in inputfile:
        if outputfile == '':
            title = inputfile.replace('https:','').replace('http:','').replace('/', '')
        else:
            title = outputfile
        m3u8_link = inputfile
    else:
        m3u8_link, title = get_m3u8_link(inputfile)  # 
    title = title.replace('\r', '').replace('\n', '').replace(' ', '')
    mp4_file_name = title + '.mp4'
    out_put_file_path = os.path.join(OUT_PATH, mp4_file_name)
    if sys.platform.startswith('win32'):
        play_lists = get_play_list(m3u8_link)

        m3u8_local_file_name = str(m3u8_link).split('/')[-2] + '.m3u8'

        for pl in play_lists:
            mf = save_m3u8_file(pl, DIR_PATH, m3u8_local_file_name)
            ul = str(pl).split('/')
            del (ul[-1])
            base_url = '/'.join(ul) + '/'
            #print(mf)
            format_lines(base_url, mf)
            download_ts_with_ffmpeg(mf, out_put_file_path)

    else:
        download_ts_with_ffmpeg(m3u8_link,out_put_file_path)

if __name__ == '__main__':
    main(sys.argv[1:])

如果在mac下代码要简单的多，直接将m3u8文件扔进去然后进行下载即可，也无需解析playlist（建议在osx 或者linux下运行）。