A lightweight Docker container for archiving personal YouTube channels using yt-dlp.

This repo is provided as-is, because it does what I want. It contains minimal code, so it should be trustable. (The serve_file and generate_rss_feed functions can be removed if not needed)

If you want it to do what you want, go ahead and fork it, adopt it, whatever. I don't care, but I also can't guarantee I'll be around to give support or accept pull requests.

Oh and don't run it on a public network, It runs a little unauthenticated web server that is not designed for the general public to be able to see.

The container reads YouTube channel list and keep those channels mirrored locally in folders of the same name.

It runs whenever it recieves a GET request to the path /go on its exposed port. This can be used for scheduling or testing. You can roll your own authentication for this if you want to.

You can add this GET request to any task scheduler / cron job, to ensure it happens regularly. yt-dlp does not re-download videos. Something like curl "http://hostname.local:8781/go" could be scheduled easily enough, and you can also run this in the browser if you want a manual approach.

Please note that to avoid throttling this won't download more than 10 videos, per channel, per execution. You could just run it every day until the archive is filled, as videos that are already downloaded don't count towards the quota.

The archive also provides an RSS feed of the folders it controls.

• Docker
• A folder mapped to /data/
• A /data/channels.txt job definition file, formatted as per channels.txt.sample

Build the Docker image with the following command:

docker build -t yt-archive .

Run the Docker container with the following command:

docker run -d -v c:/youtube/:/data/ -p 8781:8781 --name yt-archive yt-archive

• /data/: The path where the downloaded videos are stored and where the /data/channels.txt file is located.

If you're running this on a NAS device, you'll want to save the Docker image with the following command:

docker save yt-archive -o yt-archive.tar

You can then copy the tar file to your target, and either run docker import on the .tar file, or upload it to your docker GUI as an image.

This project is licensed under the MIT License - see the LICENSE file for details.