audit endpoints & develop tutorial

[hannesdatta]

Background:

We've built an API so anyone can learn how to scrape our site using that API. Yet, we're missing a tutorial.

The purpose of this issue is to develop a Python tutorial for retrieving data from our site. This one is a good example. Remember, this is NOT yet for my oDCM class, so the audience is broad.

Deliverables:

• A tutorial (in Python) that retrieves data from the various endpoints that we have made available with our API.
• Encountering anything weird with our API? Then give feedback so we can change endpoints, etc.
• Feedback on the documentation of our API - probably we need some more example code in the documentation? Check out the Last.fm API as an inspiration.

Next steps:

• Upon the approval of an initial Jupyter Notebook (add to the repository), we can translate the tutorial to raw HTML to feature it on the site.
• Also, we can develop a "parallel" tutorial in R.

[hannesdatta]

Please "imitate" the Last.fm API in terms of which endpoints to implement.

• user.getInfo

• user.getRecentTracks

• user.getTopArtists

• user.getTopTracks

• user.getWeeklyTrackChart

• user.getWeeklyArtistChart

• user.getRecentlyActive (NEW ENDPOINT, that shows you a random selection of recently active users)

• artist.getInfo

• artist.getTopTracks

• artist.search

• artists.featured (@hannesdatta tags "featured" artists, you randomly serve x artists from that list)

• chart.getTopTracks (for a given week, the charts)

• chart.getTopArtists

• chart.getTopGenres

[hannesdatta]

Views on the site:

• landing page:
  • this month's top artist --> listen now (artists.featured)
  • recently played: chart.getTopTracks with played tracks in last 24 hours
  • weekly top 15: chart.getTopTracks with 15 tracks played in last 24x7 hours
  • top genres: chart.getTopGenres
• artist profile page: artist.getInfo and artist.getTopTracks
• recently active users: user.getRecentlyActive
• user profile pages: user.getInfo + user.getWeeklyTrackChart

[thierrylahaije]

@hannesdatta we still have the following issues that need to be tackled before I can add the remaining 3 endpoints:

• artist.getInfo: We do not have an artist information table yet. Could you provide me with a R file or .cs containing the information that you want to provide when an artist name is retrieved?
• artists.featured: How do we want to provide featured artists? We could use the Spotify API for this, but maybe you have another good idea.
• chart.getTopGenres: we do not have genres added to the tracks yet.
• Last but not least we need to add track names. Currently all tracks have the title "emptytrack", which still need to be filled with real track names. If these are added, I can check if the endpoint is actually doing what is expected. For now, I checked it based on the "emptrytrack" titles, but this is no guarantee that it will work the same when we implement real track names.

If the tasks above are done, please give me a sign so that I can complete the API endpoints.

[hannesdatta]

• artists.getInfo: number of tracks, number of plays in last week, number of plays since the beginning, dictionary with top songs: dictionary with list with top x song names (e.g., top 5 songs), including their all-time plays,

• artists.featured: a dictionary with five or ten artists randomly drawn from the list of featured artists in our database (column: featured)

• Let's drop genres for now

[hannesdatta]

I have updated the data, it now contains song and artist files as well. Rerun simulations.R, please.

[hannesdatta]

Is this all done, @thierrylahaije - please close then.

[thierrylahaije]

All FastAPI endpoints now have been fixed and are working. I only need to update artist.getInfo to include all necessary data. When I fixed this, the API endpoints can be checked to see if they are complete.

[thierrylahaije]

All FastAPI endpoints are now completed (including artist.getInfo). @hannesdatta if you could check if everything is now complete, I can finish up and close this issue.