How Songs are Represented

How Songs Are Represented

Primarily, K'atun deals with large collections of songs. However, it would be woefully inefficient to store all of our binary blobs into a database. We'd effectively be storing the same data twice!

Right now, we leverage Mutagen to parse our music's metadata. The primary things we're interested in with relation to a song is:

• Where it's stored on your local device
• What the title is
• What the author is
• What the genre(s) are
• What playlists, if any, it is tied to
• How long it is

The actual binary blob should never make it into the database model.

##How the Database Sees Songs

We're leveraging MongoDB for our database. Our structure for a song is as follows:

song = {
    "location": "/path/to/this/song",
    "title":  "The Song",
    "author":  "The Author",
    "genre": ["Hip-Hop", "Culture", "Awesome"]
    "playlists": ["Down the Road Music", "Chill Out Music", "Party Music"]
}

What this implies is that a song is very closely tied to a playlist of some sort.