Chessmates

Statistics for your Lichess team

Summary

The backend service schedules a task to periodically scrape the Lichess API for data. This data is stored in its entirety and will be used in the future to provide richer statistical analysis team games.

Tech

The application is Spring Boot written in Groovy, tested using Spock and build with Maven.

Info

Fetching Lichess Data

Due to certain restrictions set by Lichess.com, we have to be careful to limit the number of requests we make as they error requests with 429 quite liberally. This is made more difficult by the fact that the Lichess API only gives us the ability to request games between two players, and limits the number of games returned on a page of data to 100, meaning we have to make multiple request per set of opponents within the team.

We're also relying on some pretty brittle assumptions about the Lichess API. To cut down requests for data we've already fetched and persisted, we keep track of latest values we've retrieved from each data set. When we start to request that data set again, we check each item in each page, and when encounter the most recent item we've already fetched, we know we can stop making request.

This is clearly brittle because we're relying on the ordering of items returned by Lichess' API - of which there is no stated guarentee.

Terminology
* Data set - e.g 'Scott Logic players' or 'Games between UserA & UserB'
* item - Player or Game

Local Installation

IntelliJ

1. Set up the password to the SSL certificate.

   1. Check the wiki page (Internal Scott Logic wiki) for the SSL password.
   2. Set the CHESSMATES_KEYSTORE_PW environment variable to this value.(Build, Execution & Deployment > Build Tools > Maven > Runner > Environment Variables)

   Running the application

   Run the maven task spring-boot:run to start up the application in Spring Boot's embedded Tomcat server.

   Contributing

   Fork the GitHub repository & make a pull request.

   Raising issues

   On the GitHub repository.