Product Launch
anonymous
27 points
12 comments
Posted6d agoActive2d ago
YouTube Downloader
github.comlaunchedlibrarydevelopers
Key Features
YouTube playlist downloadingopen-sourceGitHub repository
Tech Stack
Python
Key Features
Tech Stack
Author wanted yt-dlp to be fed with a custom text file: "playlists.txt"
The script loops through that text file, parses it, and then launches yt-dlp for each valid line with a channel name.
yt-dlp -o "%(channel)s/%(playlist_title)s/%(title)s.%(ext)s" -a playlists.txt
alias yt-pl='yt-dlp -o "%(channel)s/%(playlist_title)s/%(title)s.%(ext)s" -a playlists.txt'
https://github.com/yt-dlp/yt-dlp#:~:text=channel%20%28string...
https://github.com/yt-dlp/yt-dlp#:~:text=%2Da%2C%20%2D%2Dbat...
https://github.com/yt-dlp/yt-dlp#:~:text=%2D%2Ddownload%2Dar...
Not to mention that the script is clearly LLM-generated
I feel like after a decade and a half google knows pretty well what I like and what I'm interested in. I built that homepage brick by brick.
Whenever I tried other youtube clients or ungoogled alternatives the recommendations were all really generic and not that good. Of course I won't find something that matches the real deal, but at least I want to:
1. Steer away from viral content, aka Mr Beast videos or whatever. 2. Have discoverability of youtube videos I would be interested in but that I'm not subscribed to yet.
To this day I still find new channels or videos I'm really interested in on my recommended feed.
It says "by channel name" but then it seems like you still have to pass in each playlist URL separately.
This is just a "wrapper" for something yt-dlp already does?
—⁂—
From the “Features” section of the README:
> • Organized Structure: Creates separate directories for each channel.
… depending on you mentioning that name in your playlists.txt file, so that if you were writing a script you’d just be doing a mkdir/cd yourself, trivial.
> • Smart Sync: Skips files that have already been downloaded (--no-overwrites).
That’s not really what --no-overwrites does. The default of --no-force-overwrites is probably actually what you want: “do not overwrite the video, but overwrite related files”. You probably do want metadata files to be updated on subsequent runs.
> • Clean Naming: Saves files as Playlist Title/Video Title.mp4 (no numeric prefixes).
This is an Opinion.
> • Batch Processing: Reads multiple playlists from a playlists.txt file.
Meh, you’re invoking a script, putting the stuff in the script would be at least as easy.
—⁂—
From the script itself, going through the arguments passed to yt-dlp:
--cookies "$COOKIES_FILE"
-f "bv*+ba/b"
--merge-output-format mp4
--no-overwrites
-o "%(playlist_title)s/%(title)s.%(ext)s"
—⁂—
Instead of having a playlists.txt file containing `Channel Name|https://www.youtube.com/playlist?list=x` and having a separate 73-line file download_playlists.sh, you might as well have just one download_playlists.sh file containing:
dl() {
mkdir -p "$1"
pushd "$1"
yt-dlp -o "%(playlist_title)s/%(title)s.%(ext)s" "$2"
popd
}
dl 'Channel Name' 'https://www.youtube.com/playlist?list=x'
To get YouTube videos offline on your phone/tablet for flights:
1. Create a playlist called “download” and add the videos you want.
2. Use yt-dlp (playlist URL) or this tool to pull that playlist to a folder.
3. Run the downloader on your Plex box on a schedule (cron/launchd).
4. Add that folder to Plex as a library.
5. Enable downloads for that library in Plex, then sync to your device.Not affiliated with Hacker News or Y Combinator. We simply enrich the public API with analytics.