echonest-ruby-api

![Echonest] (http://the.echonest.com/static/img/logos/250x80_lt.gif)

echonesst-ruby-api is a pure Ruby wrapper around the Echonest APIs.

Requirements

• An API Key (available free)
• Ruby 1.9.3+
• echonest/echoprint-codegen binary in your $PATH if you want to identify audio files. (yeah, identify!)

Installation

In your Gemfile:

gem 'echonest-ruby-api'

and then bundle install

Or install locally:

$ (sudo) gem install echonest-ruby-api

Usage

Require the gem in your file:

require 'echonest-ruby-api'

Artist

Create an instance of an object

artist = Echonest::Artist.new('YOUR-API-KEY', 'Weezer')

Then you have access to a bunch of methods:

    artist.name
    artist.biographies
    artist.blogs
    artist.familiarity
    artist.hotttnesss
    artist.images
    artist.songs
    artist.profile

Artist methods support requesting buckets according to documentation. Buckets are added as array to request object:

    artist.buckets = %w[artist_location biographies blogs years_active id:7digital-US id:spotify-WW id:musicbrainz]

Exact response are specified in the RDoc, but the method names try to be as self-explanatory as possible.

Song

Create an instance of the Song module.

song = Echonest::Song.new('YOUR-API-KEY')

Then you have access to the song/search endpoint: (this is where it gets clever)

params = { mood: "sad^.5", results: 10, min_tempo: 130, max_tempo: 150 }
song.search(params)

See the full list of params here

Identification

Note: This stuff is flakey as hell. Seems to work pretty well on OSX, but it doesn't work out of the box. You'll need to follow these instructions to get it working.

You can even identify a song simply from its fingerprint! Support for this is flaky so far and only tested on OS X.

Firstly, make sure that the echoprint-codegen binary is available on your local $PATH.

Just run: echoprint-codegen on the terminal and see if it returns anything other than an error.

If it's not installed, you'll need to compile it from source. It's not as scary as it sounds.

brew install ffmpeg boost taglib # Install dependencies

cd ~/Desktop/ # or somewhere else sensible, you can delete it later anyway

git clone https://github.com/echonest/echoprint-codegen.git

cd echoprint-codegen

make

make install

This should then allow you to use the echoprint-codegen command at the command line. If not, try following the instructions here: echonest/echoprint-codegen

Then just use this method call:

song = Echonest::Song.new('YOUR-API-KEY')
code = song.echoprint_code('path/to/audio/file')
puts song.identify(code)

If there's a positive match, it'll return something like this:

{
  "response": {
        "status": {
        "code": 0,
        "message": "Success",
        "version": "4.2"
      },
      "songs": [
      {
        "title": "Billie Jean",
        "artist_name": "Michael Jackson",
        "artist_id": "ARXPPEY1187FB51DF4",
        "score": 49,
        "message": "OK (match type 5)",
        "id": "SOKHYNL12A8C142FC7"
      }
    ]
  }
}

Checkout spec/song_spec.rb for an example code to test it out.

Note that this calls the song/identify API endpoint and does not support other Echoprint servers.

Testing

Testing is done using RSpec. Just run guard in the root directory and it'll run the tests automatically. Use vcr to mock responses from the Echonest servers.

Contributors

• Max Woolf (me!)
• abdyer
• garethrees
• jordanpoulton
• makersacademy
• deadroxy
• rootpd
• Devon Blandin