飞牛metube容器自动下载YouTube视频，并自动匹配封面和简介

效果

文件结构

飞牛影视

实现

metube负责监控播放列表或者频道

飞牛影视里可以看到封面、简介、发布日期

视频目录结构是： 频道 - 发布时间_视频标题 - 视频.mp4

最耗时bug

飞牛影视读取的简介不能有连续的回车换行....用代码删了

每次修改metube的json和py后，都要完全卸载飞牛影视后重装（存疑，可能和自动扫描媒体库有关，不想搞明白了，就这样吧）所以现在也关了自动扫描媒体库，避免简介读取不了

步骤（ChatGPT写的）

网络：mihomo+metacubexd tun模式

（感谢李田所的分享）

（metacubexd的镜像下载好慢，先用了v2rayA挂了http代理给飞牛docker下载）

metube

docker-compose.yml

services:

  metube:

    image: ghcr.io/alexta69/metube:latest

    container_name: metube

    restart: unless-stopped



    ports:

      - "8081:8081"



    volumes:

      # 下载好的 YouTube 视频

      - /vol1/1000/YouTube:/downloads



      # yt-dlp 配置文件

      - /vol1/1000/metube/json/ytdl-options.json:/config/ytdl-options.json:ro



      # 自动生成 NFO 的脚本

      - /vol1/1000/metube/json/create_nfo.py:/config/create_nfo.py:ro



    environment:

      - DOWNLOAD_DIR=/downloads

      - YTDL_OPTIONS_FILE=/config/ytdl-options.json



    

  

ytdl-options.json

放在vol1/1000/metube/json/

{

  "writethumbnail": true,

  "writeinfojson": true,

  "writedescription": true,

  "outtmpl": {

    "default": "%(channel|YouTube).120B/%(upload_date>%Y_%m_%d)s_%(title).150B/%(title).150B.%(ext)s",

    "thumbnail": "%(channel|YouTube).120B/%(upload_date>%Y_%m_%d)s_%(title).150B/poster.%(ext)s",

    "infojson": "%(channel|YouTube).120B/%(upload_date>%Y_%m_%d)s_%(title).150B/metadata.info.json",

    "description": "%(channel|YouTube).120B/%(upload_date>%Y_%m_%d)s_%(title).150B/description.txt"

  },



  "postprocessors": [

    {

      "key": "FFmpegThumbnailsConvertor",

      "format": "jpg"

    },

    {

      "key": "Exec",

      "when": "after_move",

      "exec_cmd": "python3 /config/create_nfo.py %(filepath)q"

    }

  ]

}



    

  

create_nfo.py

放在vol1/1000/metube/json/

#!/usr/bin/env python3

import json

import re

import sys

from pathlib import Path

import xml.etree.ElementTree as ET



def clean_xml_text(value):

    if value is None:

        return ""

    text = str(value)

    return re.sub(r"[\x00-\x08\x0B\x0C\x0E-\x1F]", "", text)



def normalize_description_newlines(desc):

    """

    飞牛对简介里的连续空行兼容不好。

    这里把连续两个及以上换行压缩成一个换行。

    """

    desc = desc or ""



    # 统一 Windows / Mac / Linux 换行

    desc = desc.replace("\r\n", "\n").replace("\r", "\n")



    # 去掉每一行首尾多余空格

    lines = [line.strip() for line in desc.split("\n")]

    desc = "\n".join(lines)



    # 连续两个及以上换行，压缩成一个换行

    desc = re.sub(r"\n{2,}", "\n", desc)



    return desc.strip()



def ymd(value):

    if not value:

        return ""

    value = str(value)

    if re.fullmatch(r"\d{8}", value):

        return f"{value[:4]}-{value[4:6]}-{value[6:8]}"

    return value



def add(parent, tag, value):

    value = clean_xml_text(value)

    if value == "":

        return None

    elem = ET.SubElement(parent, tag)

    elem.text = value

    return elem



def add_publish_date_to_description(desc, upload_date):

    desc = desc or ""



    # 先规范换行，避免原简介本身有连续空行

    desc = normalize_description_newlines(desc)



    # 防止重复运行脚本时，简介开头重复添加发布日期

    desc = re.sub(

        r"^发布日期[:：]\s*\d{4}(?:-\d{2}-\d{2})?\s*\n*",

        "",

        desc

    ).lstrip()



    # 删除后再规范一次，避免删除发布日期后留下空行

    desc = normalize_description_newlines(desc)



    if not upload_date:

        return desc



    if desc:

        desc = f"发布日期：{upload_date}\n{desc}"

    else:

        desc = f"发布日期：{upload_date}"



    # 最后再压缩一次，确保绝对不会出现连续两个及以上换行

    return normalize_description_newlines(desc)



def main():

    if len(sys.argv) < 2:

        print("No video path provided")

        return 0



    video_path = Path(sys.argv[-1])

    if not video_path.exists():

        print(f"Video not found: {video_path}")

        return 0



    folder = video_path.parent

    info_json = folder / "metadata.info.json"



    if not info_json.exists():

        matches = list(folder.glob("*.info.json"))

        if matches:

            info_json = matches[0]

        else:

            print(f"No info json found in: {folder}")

            return 0



    with info_json.open("r", encoding="utf-8") as f:

        data = json.load(f)



    desc = data.get("description") or ""

    desc_file = folder / "description.txt"



    if not desc and desc_file.exists():

        desc = desc_file.read_text(encoding="utf-8", errors="ignore")



    title = data.get("title") or video_path.stem

    channel = data.get("channel") or data.get("uploader") or ""

    upload_date = ymd(data.get("upload_date") or data.get("release_date"))

    year = data.get("release_year") or (upload_date[:4] if upload_date else "")

    duration = data.get("duration")

    video_id = data.get("id") or ""

    webpage_url = data.get("webpage_url") or data.get("original_url") or ""



    # 在简介开头增加发布日期，并清理连续空行

    desc = add_publish_date_to_description(desc, upload_date)



    # 同步改写 description.txt，确保 description.txt 里也没有连续空行

    if desc:

        desc_file.write_text(desc, encoding="utf-8")



    root = ET.Element("movie")



    add(root, "title", title)

    add(root, "originaltitle", title)

    add(root, "sorttitle", title)

    add(root, "plot", desc)

    add(root, "outline", desc.splitlines()[0][:200] if desc else "")

    add(root, "premiered", upload_date)

    add(root, "aired", upload_date)

    add(root, "year", year)



    if duration:

        try:

            add(root, "runtime", str(round(float(duration) / 60)))

        except Exception:

            pass



    add(root, "studio", channel)

    add(root, "director", channel)

    add(root, "website", webpage_url)



    if video_id:

        unique = ET.SubElement(root, "uniqueid", {"type": "youtube", "default": "true"})

        unique.text = clean_xml_text(video_id)



    for category in data.get("categories") or []:

        add(root, "genre", category)



    for tag in (data.get("tags") or [])[:20]:

        add(root, "tag", tag)



    add(root, "thumb", "poster.jpg")



    fanart = ET.SubElement(root, "fanart")

    fanart_thumb = ET.SubElement(fanart, "thumb")

    fanart_thumb.text = "poster.jpg"



    nfo_path = video_path.with_suffix(".nfo")



    tree = ET.ElementTree(root)

    if hasattr(ET, "indent"):

        ET.indent(tree, space="  ")



    tree.write(

        nfo_path,

        encoding="utf-8",

        xml_declaration=True,

        short_empty_elements=False

    )



    print(f"NFO created: {nfo_path}")

    return 0



if __name__ == "__main__":

    raise SystemExit(main())