Python 封装 FFmpeg 下载 m3u8 文件

有时候想批量下载 m3u8^[1] 文件。一般来说通过 ffmpeg^[2] 就能实现。这里用 python 代码封装了一下，供君参考。

import subprocessimport threadingSAVE_PATH = "D:/downloads"defffmpeg_dl(real_url, save_path_name):    ffmpeg_path = "ffmpeg"    user_agent = ("Mozilla/5.0 (Linux; Android 11; SAMSUNG SM-G973U) AppleWebKit/537.36 (""KHTML, like Gecko) SamsungBrowser/14.2 Chrome/87.0.4280.141 Mobile ""Safari/537.36")    analyzeduration = "20000000"    probesize = "10000000"    bufsize = "8000k"    max_muxing_queue_size = "1024"    ffmpeg_command = [        ffmpeg_path, "-y","-v", "verbose","-rw_timeout", "30000000","-loglevel", "error","-hide_banner","-user_agent", user_agent,"-protocol_whitelist", "rtmp,crypto,file,http,https,tcp,tls,udp,rtp","-thread_queue_size", "1024","-analyzeduration", analyzeduration,"-probesize", probesize,"-fflags", "+discardcorrupt","-i", real_url,"-bufsize", bufsize,"-sn", "-dn","-reconnect_delay_max", "60","-reconnect_streamed", "-reconnect_at_eof","-max_muxing_queue_size", max_muxing_queue_size,"-correct_ts_overflow", "1",    ]    audio_code = "aac"    segment_format = "mp4"    split_time = "180000"    command = ["-c:v", "copy","-c:a", audio_code,"-map", "0","-f", "mp4",# "-segment_time", split_time,# "-segment_format", segment_format,# "-reset_timestamps", "1",        save_path_name,    ]    ffmpeg_command.extend(command)try:        _output = subprocess.check_output(ffmpeg_command,                                          stderr=subprocess.STDOUT)except subprocess.CalledProcessError as e:        print(f"错误信息: {e} 发生错误的行数: {e.__traceback__.tb_lineno}")defmake_counter():    count = 0defcounter():nonlocal count        count += 1return countreturn counterif __name__ == "__main__":    c1 = make_counter()    c2 = make_counter()    c3 = make_counter()    c4 = make_counter()    url_list = [        ["https://xx/index.m3u8", f"{SAVE_PATH}/xx{c1()}.mp4"],    ]    tt = []for url, name in url_list:        thread = threading.Thread(            target=ffmpeg_dl, args=(url, name), name='ffmpeg_dl')        thread.start()        tt.append(thread)for t in tt:        t.join()

由于使用了 ffmpeg 命令，所以要先安装 ffmpeg。

参考资料