pykhet 0.18

Introduction

The Khet board game logic and structures implemented in python. Also exposes adversarial search based algorithms.

from pykhet.components.types import TeamColor
from pykhet.games.game_types import ClassicGame
import random
from pykhet.solvers.minmax import MinmaxSolver

# Create a game with classic piece placement
game = ClassicGame()

# Get all valid silver moves
silver_moves = game.get_available_moves(TeamColor.silver)

# Randomly Play One
game.apply_move(random.choice(silver_moves))

# Finish the turn by applying the laser
game.apply_laser(TeamColor.silver)

# Use adversarial search to pick a move
solver = MinmaxSolver()
move = solver.get_move(game, TeamColor.red)
game.apply_move(move)
game.apply_laser(TeamColor.red)

Serialization

There is ample support to serializing the state of objects as dictionaries. Useful for easy storage as json.

from pykhet.components.types import TeamColor, Piece
from pykhet.games.game_types import ClassicGame
import random
from pykhet.solvers.minmax import MinmaxSolver

# Create a game with classic piece placement
game = ClassicGame()

# Serialize the board (list of serialized piece positions, orientations, and colors)
squares = game.to_serialized_squares()

# Deserialize the board
Game.from_serialized_squares(squares)

# Serialize a pieces
p1 = Piece(PieceType.scarab, TeamColor.silver, Orientation.down).to_dictionary()
# Deserialize a piece
same_piece = Piece.from_dictionary(p1)

Board Layout

The khet board and piece layout is represented below:

Adversarial Search

Provided is a very basic adversarial search algorithm that works with a low number of iterations.