ydl-podcast

A simple tool to generate Podcast-like RSS feeds from youtube (or other youtube-dl supported services) channels, using youtube-dl.

Setup

Install package with requirements:

Youtube-dl:

pip install ydl-podcast[youtube-dl]

yt-dlp:

pip install ydl-podcast[yt-dlp]

Configuration

Edit the config.yaml file to list your podcast sources and configure them, as well as edit general configuration.

Example config file

The available settings are the following.

General settings

• output_dir: local directory where the downloaded media will be stored, and the podcast xml files generated.
• url_root: root url for the static files (used in the generation of the XML to point to the media files.
• subscriptions: a list of feeds to subscribe to.
• youtube-dl-module: Alternative youtube-dl python module. By default, this uses youtube-dl, but can leverage forks such as yt-dlp.

Feed settings

Mandatory

• name NAME: Name of the podcast source. Used as the podcast title, and media directory name.
• url URL: source url for the youtube (or other) channel.

Optional

• audio_only True/False: if True, audio will be extracted from downloaded videos to create an audio podcast.
• retention_days N: only download elements newer than N days, and automatically delete elements older.
• download_last N: only download the latest N videos.
• initialize True/False: if True, then downloads everything on the first run, no matter the download_last or retention_days specified.
• output_dir: local directory where the downloaded media will be stored, and the podcast xml files generated.
• url_root: root url for the static files (used in the generation of the XML to point to the media files.
• format: file format to force youtube-dl to use (eg mp4, webm, mp3 for audio only…)
• best: force best quality (only useful when specifying a format).
• ydl_options: list of raw youtube-dl options to use. For experienced users, since this will likely yield issues if not understood.
• nfo_files: generates nfo files for subscriptions and downloaded episodes (simulates a "tvshow" nfo for the subscription and "tvshow episode" for each video). This helps plex, kodi, jellyfin import correct metadata. Does NOT support audio_only feeds at this point.
• skip_download: Don't perform download, just generate the Atom feed from the existing files. Mandatory url parameter can be skipped if skip_download is set to true
• download_as_playlist: Pass the url to the downloader directly, without processing the metadata of each item in the playlist. This is helpful with generic urls, and a few specific extractors.

Usage

Using cron or your favorite scheduler, run:

ydl-podcast [-c configfile.yaml]

You can then use your favorite web server to serve the files (a good idea is to exclude the *.json and *.part files from being served as the first might leak information, and the second is unnecessary.

eg with nginx:

root /var/www/static/podcasts/;
location ~ (\.json$|\.part$) {
  return 403;
}

Command line arguments

usage: ydl-podcast [-h] [-v] [-c CONFIG] [-j JSON_CONFIG] [-f FILTER] [-e EXCLUDE]

options:
  -h, --help            show this help message and exit
  -v, --version         Show version and exit
  -c CONFIG, --config CONFIG
                        Configuration file
  -j JSON_CONFIG, --json-config JSON_CONFIG
                        Configuration string in JSON format
  -f FILTER, --filter FILTER
                        Filter subscriptions
  -e EXCLUDE, --exclude EXCLUDE
                        Exclude subscriptions