chris-czopp/cyferd

Puzzle

What I did:

• dockerized environment with eslint and mocha
• wrote a functions to get possible moves for players
• wrote a function which to get a next wolf move
• wrote a function to get movement direction
• wrote a scenario to test wolf movement
• found the win path which I used for testing :)

What I didn't manage to do which was constrained by time:

• the overall simulation
  • I was thinking about finding paths to the target first and then switch them depending on the wolf position.
  • The potential optimisation could be to find the places to get the wolf trapped and use Thomas to lead him to the place.
• The getWolfNextMove fails in cases when the wolf is on a different row and column and can make 2 moves to match x/y axis. It's because I made wrong assumptions at the beginning and changing the logic would take more time.
  • The fix would be to tinker more with Thomas and wolf positions and also look for the movement direction.
• I would add more tests.
• I would refactor the code to use more helper functions with meaningful names.
• I would also add CI and setup some git flow.

Usage

Run:

docker-compose up --build

In separate terminal tab run (there will be tests failing):

docker exec -it cyferd npm test