deezer-datasync synchronizes data from Deezer:

• Favorite albums
• Favorite artists
• List of playlists
• Individual playlists with their associated tracks

As of the current version, only GitHub is supported as a backend; the application will commit JSON files in this arborescence:

your-git-repo/
├── playlists/
│   ├── {id}_{title}.json
│   ├── {id}_{title}.json
│   ├── [...]
├── albums.json
├── artists.json
└── playlists.json

See the changelog.

This project is mainly developed for my personal use and as such, is experimental. However I'll be glad to help if you encounter any issue.

You'll need a Deezer token to use this application. You can follow Deezer's instructions here. I have also written a documentation here.

Prepare a configuration file in HOCON format:

config {
  application {
    zone = "your-zone" // e.g Europe/Paris"
  }
  deezer {
    profile {
      userId = your-user-id
      // Playlists to backup -- leave empty to backup *all* your playlists
      playlistIds = [id1, id2]
    }
    url = "https://api.deezer.com"
    token = your-deezer-token
    // Maximum number of results per request. 100 is a hard minimum to limit the number of requests.
    maxResults = 200
    // Number of calls per second to the API. 10 is a hard maximum to avoid Deezer's limit (50/5 seconds).
    rateLimit = 5
  }
  github {
    token = your-github-token
    repo = your-repo
    branch = your-branch
  }
}

A multiplatform image (AMD64 and ARMv7 -- for execution on a Raspberry Pi 3) is available on Dockerhub:

# Pull the latest version of the image
docker pull alecigne/deezer-datasync:latest

# Run the container with the latest image (check that your config file is compatible!)
docker run -it -v /absolute/path/to/application.conf:/application.conf alecigne/deezer-datasync

Download a Jar in the releases section, then run it using Java 17:

java -jar -Dconfig.file=/path/to/application.conf deezer-datasync.jar

You probably know what you're doing anyway 🙂

Playlist filenames ({id}_{title}.json) are generated from "clean" playlist titles:

• The title is converted to lowercase.
• Any group of non-alphanumeric characters (size 1 or more) is converted to an underscore.
• Any trailing underscore is removed from the title.

You can see basic cases in issue #28. This issue was the basis for the parameterized test in GitHubMapperTest.java. There is also a property-based test in GitHubMapperPropertyTest.java.