spotify-metrics

Discontinued: Spotify cut significant API access to third party apps

---

This is a personal project that lets me scrape and visualize my recently played songs on Spotify along with some metrics about the listen history.

Some of the technologies used

• Python
  • Dash
    • Data App and Dashboard tool
    • Built on Plotly, React, and Flask
  • Tekore
    • Spotify API client
• MongoDB
• Docker

Limitations of the project

• Project is currently limited to only my listen history
• Spotify only allows you to request the 50 most recent songs
  • if you listen to 51 songs you will not be able to get the 51 songs
  • Current solution: scrape the recently played endpoint on a sane interval
• Limited scope of data ingestion

Deploying

Deploying on my homelab

Deploying is luckily quite simple. First configure your .env and simply run make.

The application should then be defined at http://spotifymetrics.your.domain assuming you have enabled traefik. If traefik is disabled then the site should still be accessible from http://localhost:9990

To view the mongo database we can use express at http://localhost:8481

Deploying without homelab

Configure your .env and simply run make.

Getting the access and refresh tokens

• Install spotify-scraper to a python environment
• Run get-new-spotify-token
  • Ensure you've correctly setup your spotify API application with spotify otherwise this won't work

Roadmap

• Working application
• Allow for other users to see their listen history
• Separate front and backend
  • Build a Python backend with FastAPI
  • Build a client side rendered frontend with a JavaScript framework
• Implement new visualizations !

Resources

Flask

• Flask applicative dive

Frontend

• Stylizing your dash
• Admin template
  • Demo