⚠️ This project is no longer in active development. See YouTubeCast for an alternative.

youtube-podcast-feeds

Create podcast feeds for YouTube channels, users and playlists

Features

• Generate RSS feeds that can be added to podcast apps with support for video podcasts (e.g. Apple Podcasts or Pocket Casts)
• Simple web server for serving RSS feed data and video files
• Download YouTube videos from channels, users, or playlists for higher quality

Setup Using Docker

Prerequisites:

• Ensure docker is set up and running on your machine (https://docs.docker.com/get-docker)
• Set up a hostname that can be used to access your machine from the internet (can use a static IP address as well)
• Get a YouTube API v3 key (https://developers.google.com/youtube/v3/getting-started)

To run this application using docker:

1. Create a directory to store data files (can be on an external drive or NAS, which is handy when using the highQualityVideo option)
2. Create the configuration files as described below (docker-compose.yml, config.json, and optionally cookies.txt)
3. Run docker-compose up -d in the folder where your docker-compose.yml lives
4. Check the logs using docker-compose logs -f to see if there are any errors in your configuration
5. (Optional) - Replace cover.png files in the data directory with custom cover artwork (YouTube channel or user profile pictures are pulled automatically on first run)
6. Wait for the first update run to pull feed data and download any videos if applicable
7. Add podcast feeds to your podcast app of choice with the URL http://hostname/feedId

docker-compose.yml

version: '3'
services:
  youtube-podcast-feeds:
    image: trevorsharp/youtube-podcast-feeds:latest
    container_name: youtube-podcast-feeds
    restart: always
    ports:
      - 80:80
    volumes:
      - REPLACE_WITH_CONFIG_DIRECTORY_PATH/config.json:/app/config.json
      - REPLACE_WITH_DATA_DIRECTORY_PATH:/app/data
      # Remove the following line if you are not using a cookies.txt file
      - REPLACE_WITH_COOKIES_DIRECTORY_PATH/cookies.txt:/app/cookies.txt

Create a file named docker-compose.yml with the contents above and substitute in the file path to your config files and the file path to your data directory.

config.json

{
  "hostname" : "http://example.com",
  "apiKey" : "YOUTUBE_API_KEY",
  "timeZone" : "America/New_York",
  "updateInterval" : 2,
  "maxResults": 5,
  "maxEpisodes" : 10,
  "feeds" : [
    {
      "id" : "h3podcast",
      "title" : "H3 Podcast",
      "channel" : "UCLtREJY21xRfCuEKvdki1Kw",
      "filter" : "H3 Podcast #[0-9]+"
    }
  ]
}

Create a file named config.json with the contents above and fill in the following parameters:

• hostname - Hostname that will be used to access podcast feeds and content
• apiKey - YouTube API v3 key
• feed.id - Unique identifier for each feed (letters, numbers and hyphens only, no spaces)
• feed.title - Title for the podcast (read by podcast clients)

Feed Sources:

• feed.channel - YouTube channel id (from https://youtube.com/channel/CHANNEL_ID)
• feed.user - YouTube username (from https://youtube.com/user/USERNAME)
• feed.playlist - YouTube playlist id (from https://www.youtube.com/playlist?list=PLAYLIST_ID)

Optional Parameters:

• timeZone - Name of time zone used for logging - Default: America/New_York
• updateInterval - Interval for updating feeds (in hours) - Default: 2
• maxResults - Number of videos to search for when updating (per feed) - Default: 5
• maxEpisodes - Maximum number of videos to keep (per feed), useful when using the highQualityVideo option to limit storage usage - Default: unlimited
• highQualityVideo - If true, podcast feeds will default to downloading videos from YouTube into the data folder using the highest quality available in mp4 format (usually 1080p vs 720p without using this option). Will provide the lower quality version until the high quality version has downloaded. - Default: false

WARNING: Setting highQualityVideo to true will significantly increase the amount of storage space and CPU usage required to run

Optional Feed Parameters:

• feed.highQualityVideo - Overrides the global setting for a specific podcast feed - Default: none / uses global setting
• feed.maxEpisodes - Overrides the global setting for a specific podcast feed - Default: none / uses global setting
• feed.filter - String containing regex used to filter videos. Only videos with titles that have a match for this regex will be added to the feed - Default: none

See Advanced Parameters for more options

Note: Regex are evaluated case insensitive. Be sure to double escape any backslashes (e.g. use "\\s" for a whitespace character instead of just "\s")

cookies.txt (Optional)

If you want to download YouTube content that requires user authentication to download, you will need to add a cookies.txt file to your configuration. One reason for needing this is to download members-only videos. Note that the source of these videos (channel, user, or playlist) still must be either public or unlisted. For members-only videos, I recommend going to the channel's home page and scrolling down to find an auto-generated playlist titled "Members-only videos" which will contain all the videos posted for members of the channel.

To generate this file:

1. To get a cookies.txt file, download a browser extension (such as this one for Chrome)
2. Log in to YouTube
3. With a YouTube tab open, open the cookies.txt extension and click the export/download button
4. Rename the downloaded file to cookies.txt

Data Directory Structure

Files will be stored in the data directory using the following structure:

data
 |
 └── content
 |    └── videoId1.mp4
 |    └── videoId2.mp4
 |    └── videoId3.mp4
 |        ...
 |
 └── feedId1
 |    └── cover.png
 |    └── feedData.json
 |
 └── feedId2
      └── cover.png
      └── feedData.json
 ...

Advanced Parameters

• feed.episodeNumbers - String containing regex used to extract episode number from video title - Default: no episode numbers
• feed.cleanTitles - Array of 2-item sub-arrays where the first element of each sub-array is a regex to match part of an episode title and the second element is a string replacement for any matches found - Default: empty
• feed.titleCase - When set to true, episode titles are converted to title case - Default: false

Episode Numbers:

In these regex, you must place the leftmost set of parentheses around the actual episode number. For example, "#([0-9]+)" or "Episode ([0-9]+)." will work, but "(#[0-9]+)" or "[0-9]+" will not work.

Example: "Ep ([0-9]+)" would be used to extract the episode number (123) from a video title that looks like this: This is my podcast title | Ep 123.

Clean Titles:

The string replacement (second element of the sub-array) can use the RegExp.$1-$9 properties (see here for more info on how these are used). Common separators (- and |) are cleaned up automatically.

Example: [["Podcast Name", ""], ["Ep ([0-9]+)", "#$1"]] will transform Podcast Episode Title | Podcast Name | Ep 123 into Podcast Episode Title | #123.

Experimental Features

These features are experimental, and they may reduce compatibility, increase resource usage, and cause problems. Proceed with caution.

Additional Parameters

• maxQualityVideo - If true, video quality downloads will be allowed to exceed 1080p. Must be used in conjunction with highQualityVideo parameter set to true - Default: false
• maxVideoHeight - Number to represent the maximum height of downloaded videos in pixels. Must be used in conjunction with maxQualityVideo parameter set to true. Suggested values are 2160 and 1440 - Default: 2160