This is a module that can uses multiple providers and their API, like WikiPedia, MusicBrainz and Genius in order to retrieve discographical information and songs lyrics.
Import the provider you need and call the available methods:
import { MusicBrainz, Discography, Track, showResult } from 'discography';
const client = new MusicBrainz();
const result = await client.getArtistDiscography('Rancid', ['Live', 'Compilation', 'Single']);
const discography = result as Discography;
const { artist, releaseGroups } = discography;
const { Genius, showResult } = require('discography');
const geniusClient = new Genius();
geniusClient.getLyrics('Rancid', 'Nihilsm')
.then(showResult)
.catch(console.error);
API and HTML parsing are both used to ge the required information.
This method extracts basic info about the page, the artist's MusicBrainz ID (if present), and a list of all studio albums. Here is it an example of what you get when
you search The Offspring
:
{
message: 'OK',
data: {
pageDetails: {
pageid: 8426694,
title: 'The Offspring discography',
touched: '2021-05-15T21:56:52Z'
},
albums: {
'The_Offspring_(album)': { title: 'The Offspring', released: 'June 15, 1989' },
'Ignition_(The_Offspring_album)': { title: 'Ignition', released: 'October 16, 1992' },
'Smash_(The_Offspring_album)': { title: 'Smash', released: 'April 8, 1994' },
...
'Days_Go_By_(The_Offspring_album)': { title: 'Days Go By', released: 'June 26, 2012' },
'Let_the_Bad_Times_Roll': { title: 'Let the Bad Times Roll', released: 'April 16, 2021' }
},
musicBrainzUrl: '//musicbrainz.org/artist/23a03e33-a603-404e-bcbf-2c00159d7067'
}
}
You can pass the following options as second parameter:
saveHtml: boolean
: to save the HTML page in the current working directorysummary: boolean
: to get the page's summary with .
This method lists all the songs of a particular album, converting the songs durations into seconds. Ideally you should execute searchDiscography
, take a key of the resulting object (that is the final part the album Wiki page) and pass it to this method.
Here is it an example of what you get when you search Smash_(The_Offspring_album)
.
{
message: 'OK',
data: {
pageDetails: {
pageid: 172936,
title: 'Smash (The Offspring album)',
touched: '2021-05-15T21:56:34Z',
summary: ''
},
tracks: [
{ num: 1, title: 'Time to Relax (Intro)', length: 25 },
...
{ num: 13, title: 'Not the One', length: 174 },
{ num: 14, title: 'Smash', length: 642 }
]
}
}
It allows to freely get all the discography details of an artist/band. The fundamental thing to know is the structure of the entities:
- An artist has many release groups (albums, compilations, live, etc)
- every release groups may have multiple releases of that album (differing for publishing date, country, etc)
- every release has many recordings (the tracks of an album)
Retrieves the artist details and all the albums. The second parameter allows to filter
out those items that have the listed secondary types in their secondaryTypes
array.
Here is it a possible response if you search for Rancid
discography:
{
artist: {
id: '24f8d8a5-269b-475c-a1cb-792990b0b2ee',
name: 'Rancid',
country: undefined,
lifeSpan: { begin: '1991', ended: null },
tags: [ 'punk', 'ska punk', 'punk rock' ],
type: 'Group'
},
releaseGroups: [
{
id: '642238f2-24b6-3a0b-bfa7-9c0b4731989d',
title: 'Rancid',
primaryType: 'Album',
releaseDate: '1993-05-10',
secondaryTypes: []
},
...
{
id: '03a80bc2-716f-4ea9-907d-7c8b9b7fb1cd',
title: 'Trouble Maker',
primaryType: 'Album',
releaseDate: '2017-06-09',
secondaryTypes: []
}
]
}
Retrieve the artist details (just a person).
Here is it a possible response if you search for Bon Scott
discography:
{
id: '0d212894-db54-4cc1-9ded-3cac50183a1d'
type: 'Person',
name: 'Bon Scott',
country: 'GB',
lifeSpan: {
begin: '1946-07-09',
ended: '1980-02-19- [Usage](#usage)
- [Typescript](#typescript)
- [Javascript](#javascript)
- [Providers](#providers)
- [Wikipedia](#wikipedia)
- [`searchDiscography(artist: string, options: WikiPageOptions)`](#searchdiscographyartist-string-options-wikipageoptions)
- [`searchTracks(albumTitle: string, options?: WikiPageOptions)`](#searchtracksalbumtitle-string-options-wikipageoptions)
- [MusicBrainz](#musicbrainz)
- [`getArtistDiscography(artist: string, filterSecondaryTypes: string[])`](#getartistdiscographyartist-string-filtersecondarytypes-string)
- [`getArtist(name: string)`](#getartistname-string)
- [`getTracksByReleaseGroup(releaseGroup: string)`](#gettracksbyreleasegroupreleasegroup-string)
- [Genius](#genius)
- [`getLyrics(artistName: string, songName: string)`](#getlyricsartistname-string-songname-string)
- [Discarded providers](#discarded-providers)
- [Terms of use](#terms-of-use)
- [Resources](#resources)',
},
tags: [
{
count: 1,
name: "rock"
}
]
}
it allows to retrieve the tracks of a particular album. This method automatically selects the oldest release of an album and retrieves its tracks.
Here is it the response if you search for the release group 37c2647f-8e58-3839-a6b6-374c9ee88b1d
:
[
{
id: 'bfd6a742-7f73-399d-956b-3617c4bc1f02',
length: 123440,
title: 'Nihilism',
position: 1
},
{
id: '897a2f83-da32-3165-acc5-876662594588',
length: 173280,
title: 'Radio',
position: 2
},
//...
]
It allows to retrieve the lyrics of a particular song, given its title an the artist. If you don't provide an access token, it scrapes the search page results.
This is the only method present inside the class. If you execute getLyrics('Rancid', 'Nihilism')
, you'll obtain something like that:
{
id: 1702216,
artistId: 30419,
title: 'Nihilism',
lyrics: 'Come into the union district\n' +
'Drive down on Sharmon Palms\n' +
'White ghettos paint a picture\n' +
'Broken homes and broken bones\n' +
'\n' +
//...
}
Before landing to the final solution, different API were tested:
- Discogs API: it returns a lot of information. By using filters, the
search
endpoint returns something more manageable and precise, anyway the data weren't accurate (here's the GIST showing that). - LastFM: even in this case, searching for the artist's top albums returns 57034 results for 1141 pages (here's the GIST showing that).
- Wikipedia: it recommends to serially do requests and to use a meaningful user agent.
- MusicBrainz one request per second and a meaningful user agent.
- Genius: no particular limitations exposed in their web site.