Task submission

[khemichew]

The following improvements/features have been introduced for this implementation of the m,n,k-game:

• pretty-printing
• arbitrary undo move mechanism
• ability to preload board state
• optimised O(k) time complexity win checker if the move sequence is stored and the current move can be located, or O(m * n) time complexity if no move sequence is stored (ie. when the board state is preloaded)
• a minimax solver that brute-forces the entire game tree with a trivial evaluation function
• alpha-beta pruning for minimax
• naive transposition table for minimax, adopting Zobrist hashing approach to calculate the hashcode

I suppose minimax can be further improved by storing beta-cutoff evaluations, using the best move(s) in iterative deepening search, and using the transposition table to store the results between each move executed; though I had a big setback trying to figure a non-trivial evaluation function and have yet to figure a good heuristic for the particular game variant.