wtf is constraint-solver?

This project is a constraint solver written in Python that works through a math puzzle about a small tournament. The puzzle involves three players, named A, B, and C, who play a total of 21 matches. Each player has a target number of matches they must end up having played: A needs to reach 17, B needs 15, and C needs 10. The rules of the tournament are simple. Only two players play in any given match while the third player rests. Whoever wins a match stays at the table for the next one, and whoever loses is swapped out for the player who was resting. The program's job is to work out which sequences of wins and losses, starting from an opening match between A and B while C rests, could produce a tournament where the final participation counts match the targets exactly. To solve this, the code uses a backtracking search, sometimes called depth first search with pruning. It tries out possible outcomes step by step, and whenever a player's match count goes over their target, it abandons that branch early instead of continuing to explore it. This keeps the search efficient even though the total number of possible match sequences could otherwise be large. The program is capped so it never tries to simulate more than 21 total matches, since that is the fixed total from the puzzle. When it finds a valid tournament history that matches all three targets, it prints out the early matches of that solution. The README notes it can also surface every kind of outcome, both valid and invalid combinations, as part of exploring the full search space. To try it, you can paste the file sc.py into an online environment like Replit and run it directly. The puzzle itself was created by Adrian Paenza, and the README links to a Wikipedia page about him for background.