Get statistics from your listening history. Currently only supports Spotify, but I plan to add support for histories from other services in the future.
Required: At least one file downloaded from Spotify of your listening history. It can be downloaded from the Account -> Privacy page after requesting an export of your data. In the resulting export it will be a file named similar to StreamingHistory0.json. This program can use any number of them.
Krustens reads the streaming history files and generates events it can read back later to generate the statistics. It does this to attempt to ensure that it does not double-count any duplicate track plays between the spotify history files. Krustens right now also counts any song played for less than 10s as "skipped".
To start clone the repo and run cargo run -- -h to see the defaults for the input and outputs. If you'd like to change those values, provide different options to -o or -i. The default for input is ./data/spotify_play_history and output is ./output.
When krustens starts it will read its configuration from the command line arguments.
bash
cargo run -- process
to generate the events. The command process tells the app to process all of the listens in the input directory. A sqlite database (named krustens.sqlite) is generated when processing that will contain all of the events and snapshots of the data. Anytime the listens are processed this database will be used to check if the listen has been tracked already.
Once the events have been generated, you can generate statistics by just running the program with the generate command.
bash
cargo run -- generate
The output folder will be generated and it will then contain a number of stats files. Right now the only real values that can be changed is the year to generate statistics for (e.g. 2020 or 2021) and how many Top songs or artists to include in the general stats. For a full list of options run cargo run -- generate -h.