codingame-the-labyrinth

This repository has my solution for The Labyrinth exercise of Codingame, made in rust.

Overall strategy

This is a grid-based exercise, where we have to find and go over a target coordinate, and then go back to the start of the map as fast as possible.

The overall strategy is pretty simple, and can be divided in the following phases:

• discovery: explore the "unknown" coordinates of the map, using a path-finding algorithm to move towards the closest "unkown" tile. Active until we find a path to the "control room".
• activation: after we find the control room, we use path-finding to move onto it.
• escape: use path-finding to get the optimal path to the starting position.

This repository just uses BFS for all path-finding. I was expecting to hit a situation where the discovery phase would not explore enough for us to find an optimal path for the escape phase, but that didn't happen.

File structure

This repository follows my usual file structure for codingame challenges (both puzzles and AI):

• src/core.rs: the core datastructures and functionality.
• src/input.rs: input parsing, inclusing the FromStr trait implementations.
• src/error.rs: error enum. We can just unwrap() everything, but I'd rather exercise the best practices.
• src/entrypoint.rs: the main loop, where the strategy is implemented.
• src/sqrid.rs: this file is imported from the sqrid square-grid library. It has the implementation of all the grid-related functions, in an easy-to-import format. I'm using this in other square-grid related puzzles and AI bots.