This project has us making an program that is able to play a game called filler. The game is played through another program
that manages the actual game, filler_vm. A map is passed onto the game and one or two player programs are passed. The programs play by taking turns reading stdin and writing to stdout. After running make all, you will make my player kmira.filler. More information about the game can be found in the Introduction of this pdf.
The visualization is being printed out to stderr to avoid conflicts, but will only work by passing the -q option to filler_vm or by piping the output of filler_vm to somewhere like /dev/null. The vizualization below was built by me for the purposes of showing and debugging how the algorithm priorities moves.
Usage looks like: ./filler_vm -f < map path > -p1 < executable player1 path > [-p2 < executable player2 path >]. Although you can just run ./filler_vm to get a help message. I pipe to /dev/null for the visualization to be printed correctly.
Alternatively do make hard after having made kmira.filler for ease of use. Maps are fairly easy to make, they have to be rectangles and examples can be found here, additional players are also found here.
Below is a visualization of the heatmap that my filler algorithm uses to pick the 'best' move. The concept of the heatmap is simply: pick the move such that the given piece is as close as possible to the enemy.
