Spotify Append

What it does

Spotify Append allows for playlists to have sources™.
A source is just a regular spotify playlist that your shiny new playlist tracks including any changes.
Songs that you delete from your destination playlist do not automatically re-add. At a high level it's a bunch of playlists coming in and a filter removing certain known tracks.

Example

Schema: 1,2 are sources, 3 is the destination (a - e are tracks)

• 1 : {a,b} , 2 : {d, e} , 3 : {}
• 3 will initially contain: {a,b,d,e}
• user then deletes song a from 3 (song a)
• 3:{b,d,e}
• user adds c to 1:
• 1: {a,b,c} 3 : {b,c,d,e}
• user deletes d from 2
• 2: {e} (no change to 3)

The playlists are append only from the view of the sources. Removes are controlled entirely by the owner of the destination playlist

How it does it

Frontend to backend is plain HTTP requests with JSON data
Backend endpoints: auth, get and show all playlists, show and create drainlists
Daily (or more often)run the sync.
Currently the whole thing works but its rather messy.

Design Details

Frontend endpoints are:

1. add / remove source (local file changes only)
2. create / list drains (create is immediately followed by a populate, list is locals only )
3. list playlists (API call to get all playlists)

Backend:

1. everything is files (lots of manangemet with "open(a + b + c + d)" tons of places to go wrong)
2. temp playlists are written out to disk (needless)

Moving Forward

• Currently we have a race condition if two sourced playlists use the same source there is only 1 copy meaning updates will only be applied to one destination.
  • In order to fix this, we can just have every destination list be in a self contained directory, do deduplication later.
• Remove writing a reference out to disk so now we can do comparisons in memory, that's just lazy.

Todo

1. Port to AWS (Lambda, S3, Dynamo)
2. Account system

Notes