musicdb

Manage and organize your music.

This application is created for the purpose of testing my development skills for a position.

What can it do

The application manages data and relations in the given music database. A standalone Apollo server runs under the hood, which uses Express in the background. The exposed endpoint for GraphQL access is /gql.

Data format

The following schema has been implemented:

• Album(id, title)
• Artist(id, name)
• Track(id, name, composer, milliseconds, bytes, price)

Implemented queries:

• albums(title: String): Returns albums with at least a 90% match based on the title.
• album(id: ID!): Retrieves a specific album by its ID.
• artists(name: String): Returns artists with at least a 90% match based on the name. ⁠- artist(id: ID!): Retrieves a specific artist by their ID.
• track(id: ID!): Retrieves a specific track by its ID.

FYI: The lookup based on match percentage is done with fuzzysort.

Usage

Startup

Running the project requires nodejs and npm. The best way to install them is using nvm.

For further instructions have a look at the official download page.

Clone the repo:

git clone git@github.com:benhorv/musicdb.git

Step into the repo

cd musicdb

Install dependencies:

npm install

The command above installs all dependencies listed in package.json.

Download the database file

You have to download the database file from here. Then create a database directory in the project root and copy it there. Be careful, the filename must be chinook.db.

On linux:

cd musicdb
mkdir database
cp path/to/chinook.db ./database/chinook.db

Run the server:

npm start

This starts the server on port 4000.

Query

npm start starts a server on http://localhost:4000/. You can access the GraphQL endpoint at http://localhost:4000/gql. The application can be tested straightaway at http://localhost:4000/gql in your browser, which opens the Apollo GraphQL sandbox UI.

Example queries

Returns all albums

query {
  albums {
    id
    title
  }
}

Returns all matching albums with name "Facelift"

query {
  albums(title: "Facelift") {
    id
    title
  }
}

Returns album with id 5

query {
  album(id: 5) {
    id
    title
  }
}

Returns all artists

query {
  artists {
    id
    name
  }
}

Returns all matching albums with name "AC/DC"

query {
  artists(name: "AC/DC") {
    id
    name
  }
}

Returns track with id 8

query {
  track(id: 8) {
    id
    name
    composer
    milliseconds
    bytes
    price
  }
}

Returns track with id 8, and also the associated album and artist

query {
  track(id: 8) {
    id
    name
    composer
    milliseconds
    bytes
    price
    artist {
      id
      name
    }
    album {
      id
      title
    }
  }
}

Returns album and the associated tracks

query {
  album(id: 6) {
    id
    title
    tracks {
      id
      name
      composer
      milliseconds
      bytes
      price
    }
  }
}

Returns artist with id 5 and associated albums and tracks

query {
  artist(id: 5) {
    id
    name
    albums {
      id
      title
    }
    tracks {
      id
      name
      composer
      milliseconds
      bytes
      price
    }
  }
}

Run tests

Vitest is used for testing. I chose this because it works with TypeScript out of the box.

Run the tests with

npm test