class Client(object):
def __init__(self):
self.host = 'localhost'
self.agent_behavior_step = 0
self.new_agent_locations = [[0, 0], [0, 0]]
self.remove_tile_locations = []
self.board = Board()
self.playng_thread = None
# connections
self.team = "A" # "A" or "B"
self.client_socket = None
self.bsize = 1024
self.receive_thread = None
self.input_thread = None
self.was_recieved = False
self.rcv_msg = ''
def isAroundCell(self, cell1, cell2):
return math.sqrt((cell1[0]-cell2[0])**2+(cell1[1]-cell2[1])**2) <= math.sqrt(2)
def moveAgent(self):
if self.team == "A":
pass
elif self.team == "B":
pass
def remomveTile(self):
if self.team == "A":
pass
elif self.team == "B":
pass
def getBoardScores(self):
return self.board.getBoardScores()
# connections method
def connectServer(self, port, team):
self.team = team
addr = (self.host, port)
self.client_socket = socket.socket(socket.AF_INET, socket.SOCK_STREAM)
self.client_socket.connect(addr)
self.receive_thread = Thread(target=self.recieve)
self.receive_thread.start()
self.send(self.team)
def recieve(self):
while True:
try:
msg = self.client_socket.recv(self.bsize).decode("utf8")
rcv_dict = json.loads(msg)
order = rcv_dict['order']
if order == 'board_scores':
self.board.initBoardSize(*rcv_dict['size'])
self.board.setFirstAgentCell(rcv_dict['agents'][0])
self.board.initBoardScores(rcv_dict['scores'])
self.board.printBoardScore()
elif order == 'next_turn':
self.board.setCurrentAgentLocations(rcv_dict['agents'][0], "A")
self.board.setCurrentAgentLocations(rcv_dict['agents'][1], "B")
self.board.team_a = copy.deepcopy(rcv_dict['tiles_a'])
self.board.team_b = copy.deepcopy(rcv_dict['tiles_b'])
if self.team == "A":
self.board.printTiles_A()
self.board.printTiles_B()
except OSError:
break
def send(self, msg): # event is passed by binders.
self.client_socket.send(bytes(msg, "utf8"))
time.sleep(1e-3)
if msg == "{quit}":
self.client_socket.close()
class Client(object):
def __init__(self, port):
self.host = 'localhost'
self.webUi = WebUi()
self.port_ui = port
self.webUi.addEvent("cellClicked", self.wasClicked)
self.webUi.addEvent("getBoardScores", self.getBoardScores)
self.webUi.addEvent("connectServer", self.connectServer)
self.agent_behavior_step = 0
self.new_agent_locations = [[0, 0], [0, 0]]
self.new_agent_diff = [[0, 0], [0, 0]]
self.remove_tile_locations = []
self.board = Board()
self.playng_thread = None
self.solving_engine = NewSolver()
# connections
self.team = "A" # "A" or "B"
self.client_socket = None
self.bsize = 1024
self.receive_thread = None
self.input_thread = None
self.was_recieved = False
self.rcv_msg = ''
def isAroundCell(self, cell1, cell2):
return math.sqrt((cell1[0]-cell2[0])**2+(cell1[1]-cell2[1])**2) <= math.sqrt(2)
def wasClicked(self, board_row, board_column):
if self.team == "A":
i = 0
tile_color = "a-tile"
agent_color = "a{}-present".format(self.agent_behavior_step)
opponent_color = "b-tile"
my_tiles = copy.deepcopy(self.board.team_a)
opponent_tiles = copy.deepcopy(self.board.team_b)
elif self.team == "B":
i = 1
tile_color = "b-tile"
agent_color = "b{}-present".format(self.agent_behavior_step)
opponent_color = "a-tile"
my_tiles = copy.deepcopy(self.board.team_b)
opponent_tiles = copy.deepcopy(self.board.team_a)
agent = self.board.getCurrentAgentLocations()[i][self.agent_behavior_step]
if self.isAroundCell([board_row, board_column], agent):
diff = [board_row-agent[0],
board_column-agent[1]]
if self.agent_behavior_step >= 1:
if opponent_tiles[board_row][board_column] == 1:
if opponent_tiles[board_row][board_column] == 1:
self.new_agent_locations[self.agent_behavior_step] = [agent[0], agent[1]]
self.remove_tile_locations.append([board_row, board_column])
self.webUi.editCellAttrs(board_row, board_column, opponent_color, False)
else:
self.webUi.editCellAttrs(agent[0], agent[1], tile_color, True)
self.webUi.editCellAttrs(board_row, board_column, agent_color, True)
self.new_agent_locations[self.agent_behavior_step] = [board_row, board_column]
self.agent_behavior_step = 0
print("黒:", self.chooseTramp(diff))
print("")
json_data = json.dumps({
"order": "client_update",
"from": self.team,
"agent_location": self.new_agent_locations,
"remove_tiles": self.remove_tile_locations
})
self.send(json_data)
self.new_agent_locations = [[0, 0], [0, 0]]
self.remove_tile_locations = []
else:
if opponent_tiles[board_row][board_column] == 1:
if opponent_tiles[board_row][board_column] == 1:
self.new_agent_locations[self.agent_behavior_step] = [agent[0], agent[1]]
self.remove_tile_locations.append([board_row, board_column])
self.webUi.editCellAttrs(board_row, board_column, opponent_color, False)
else:
self.webUi.editCellAttrs(agent[0], agent[1], tile_color, True)
self.webUi.editCellAttrs(board_row, board_column, agent_color, True)
self.new_agent_locations[self.agent_behavior_step] = [board_row, board_column]
self.agent_behavior_step += 1
print("赤:", self.chooseTramp(diff))
else:
print("八近傍以外のセルには移動できません")
def chooseTramp(self, diff):
if diff == [-1, 0]:
return 1
elif diff == [-1, 1]:
return 2
elif diff == [0, 1]:
return 3
elif diff == [1, 1]:
return 4
elif diff == [1, 0]:
return 5
elif diff == [1, -1]:
return 6
elif diff == [0, -1]:
return 7
elif diff == [-1, -1]:
return 8
def moveAgent(self):
if self.team == "A":
pass
elif self.team == "B":
pass
def remomveTile(self):
if self.team == "A":
pass
elif self.team == "B":
pass
def showWeb(self):
self.webUi.showWindow(self.port_ui)
def setUIBoard(self):
self.webUi.showBoard(self.board.board_scores, self.board.first_agent_cells_a, self.board.first_agent_cells_b)
def getBoardScores(self):
return self.board.getBoardScores()
# connections method
def connectServer(self, port, team):
self.team = team
addr = (self.host, port)
self.client_socket = socket.socket(socket.AF_INET, socket.SOCK_STREAM)
self.client_socket.connect(addr)
self.receive_thread = Thread(target=self.recieve)
self.receive_thread.start()
self.send(self.team)
# これよりソルバー
self.solving_engine.set_team(self.team)
def recieve(self):
while True:
try:
msg = self.client_socket.recv(self.bsize).decode("utf8")
rcv_dict = json.loads(msg)
order = rcv_dict['order']
if order == 'board_scores':
self.board.initBoardSize(*rcv_dict['size'])
self.board.setFirstAgentCell(rcv_dict['agents'][0])
self.board.initBoardScores(rcv_dict['scores'])
self.setUIBoard()
# これよりソルバー
if self.team == 'A':
agent_pos = self.board.current_agent_cells_a
elif self.team == 'B':
agent_pos = self.board.current_agent_cells_b
self.solving_engine.update_board(self.board)
self.solving_engine.calcScoreAverage()
self.solving_engine.gen_state_list()
agent0_next = self.solving_engine.search_around(*agent_pos[0])
self.wasClicked(*agent0_next)
agent1_next = self.solving_engine.search_around(*agent_pos[1])
self.wasClicked(*agent1_next)
time.sleep(1.0)
elif order == 'next_turn':
self.board.setCurrentAgentLocations(rcv_dict['agents'][0], "A")
self.board.setCurrentAgentLocations(rcv_dict['agents'][1], "B")
self.board.team_a = copy.deepcopy(rcv_dict['tiles_a'])
self.board.team_b = copy.deepcopy(rcv_dict['tiles_b'])
self.webUi.updateCellAttrs(self.board.team_a, self.board.team_b, self.board.getCurrentAgentLocations())
# ソルバー
if self.team == 'A':
agent_pos = self.board.current_agent_cells_a
elif self.team == 'B':
agent_pos = self.board.current_agent_cells_b
self.solving_engine.update_board(self.board)
self.solving_engine.gen_state_list()
agent0_next = self.solving_engine.search_around(*agent_pos[0])
self.wasClicked(*agent0_next)
agent1_next = self.solving_engine.search_around(*agent_pos[1])
self.wasClicked(*agent1_next)
time.sleep(1.0)
elif order == 'reject_turn':
self.board.setCurrentAgentLocations(rcv_dict['agents'][0], "A")
self.board.setCurrentAgentLocations(rcv_dict['agents'][1], "B")
self.board.team_a = copy.deepcopy(rcv_dict['tiles_a'])
self.board.team_b = copy.deepcopy(rcv_dict['tiles_b'])
self.webUi.updateCellAttrs(self.board.team_a, self.board.team_b, self.board.getCurrentAgentLocations())
except OSError:
break
def send(self, msg): # event is passed by binders.
self.client_socket.send(bytes(msg, "utf8"))
time.sleep(1e-3)
if msg == "{quit}":
self.client_socket.close()
class Client(object):
def __init__(self, port):
self.host = 'localhost'
self.webUi = WebUi()
self.port_ui = port
self.webUi.addEvent("cellClicked", self.wasClicked)
self.webUi.addEvent("getBoardScores", self.getBoardScores)
self.webUi.addEvent("connectServer", self.connectServer)
self.agent_behavior_step = 0
self.new_agent_locations = [[0, 0], [0, 0]]
self.remove_tile_locations = []
self.board = Board()
self.playng_thread = None
# connections
self.team = "A" # "A" or "B"
self.client_socket = None
self.bsize = 1024
self.receive_thread = None
self.input_thread = None
self.was_recieved = False
self.rcv_msg = ''
def isAroundCell(self, cell1, cell2):
return math.sqrt((cell1[0] - cell2[0])**2 +
(cell1[1] - cell2[1])**2) <= math.sqrt(2)
def wasClicked(self, board_row, board_column, is_remove_mode):
if self.team == "A":
i = 0
tile_color = "a-tile"
agent_color = "a{}-present".format(self.agent_behavior_step)
my_tiles = copy.deepcopy(self.board.team_a)
opponent_tiles = copy.deepcopy(self.board.team_b)
elif self.team == "B":
i = 1
tile_color = "b-tile"
agent_color = "b{}-present".format(self.agent_behavior_step)
my_tiles = copy.deepcopy(self.board.team_b)
opponent_tiles = copy.deepcopy(self.board.team_a)
agent = self.board.getCurrentAgentLocations()[i][
self.agent_behavior_step]
if self.isAroundCell([board_row, board_column], agent):
if self.agent_behavior_step >= 1:
if is_remove_mode:
if opponent_tiles[board_row][board_column] == 1 or my_tiles[
board_row][board_column] == 1:
self.new_agent_locations[self.agent_behavior_step] = [
agent[0], agent[1]
]
self.remove_tile_locations.append(
[board_row, board_column])
self.webUi.editCellAttrs(board_row, board_column,
agent_color, True)
# self.delCellAttrs(board_row, board_column)
else:
print("選択したセルは除去できません")
return
else:
if opponent_tiles[board_row][board_column] == 1:
print("選択したセルに移動するためには除去を行う必要があります")
return
else:
self.webUi.editCellAttrs(agent[0], agent[1],
tile_color, True)
self.webUi.editCellAttrs(board_row, board_column,
agent_color, True)
self.new_agent_locations[self.agent_behavior_step] = [
board_row, board_column
]
self.agent_behavior_step = 0
print(self.new_agent_locations)
json_data = json.dumps({
"order":
"client_update",
"from":
self.team,
"agent_location":
self.new_agent_locations,
"remove_tiles":
self.remove_tile_locations
})
self.send(json_data)
self.new_agent_locations = [[0, 0], [0, 0]]
self.remove_tile_locations = []
else:
if is_remove_mode:
if opponent_tiles[board_row][board_column] == 1 or my_tiles[
board_row][board_column] == 1:
self.new_agent_locations[self.agent_behavior_step] = [
agent[0], agent[1]
]
self.remove_tile_locations.append(
[board_row, board_column])
# self.delCellAttrs(board_row, board_column)
else:
print("選択したセルは除去できません")
return
else:
if opponent_tiles[board_row][board_column] == 1:
print("選択したセルに移動するためには除去を行う必要があります")
return
else:
self.webUi.editCellAttrs(agent[0], agent[1],
tile_color, True)
self.webUi.editCellAttrs(board_row, board_column,
agent_color, True)
self.new_agent_locations[self.agent_behavior_step] = [
board_row, board_column
]
self.agent_behavior_step += 1
else:
print("八近傍以外のセルには移動できません")
def moveAgent(self):
if self.team == "A":
pass
elif self.team == "B":
pass
def remomveTile(self):
if self.team == "A":
pass
elif self.team == "B":
pass
def showWeb(self):
self.webUi.showWindow(self.port_ui)
def setUIBoard(self):
self.webUi.showBoard(self.board.board_scores,
self.board.first_agent_cells_a,
self.board.first_agent_cells_b)
def getBoardScores(self):
return self.board.getBoardScores()
# connections method
def connectServer(self, port, team):
self.team = team
addr = (self.host, port)
self.client_socket = socket.socket(socket.AF_INET, socket.SOCK_STREAM)
self.client_socket.connect(addr)
self.receive_thread = Thread(target=self.recieve)
self.receive_thread.start()
self.send(self.team)
def recieve(self):
while True:
try:
msg = self.client_socket.recv(self.bsize).decode("utf8")
rcv_dict = json.loads(msg)
order = rcv_dict['order']
if order == 'board_scores':
self.board.initBoardSize(*rcv_dict['size'])
self.board.setFirstAgentCell(rcv_dict['agents'][0])
self.board.initBoardScores(rcv_dict['scores'])
self.setUIBoard()
elif order == 'next_turn':
self.board.setCurrentAgentLocations(
rcv_dict['agents'][0], "A")
self.board.setCurrentAgentLocations(
rcv_dict['agents'][1], "B")
self.board.team_a = copy.deepcopy(rcv_dict['tiles_a'])
self.board.team_b = copy.deepcopy(rcv_dict['tiles_b'])
if self.team == "A":
self.board.printTiles_A()
self.board.printTiles_B()
print(self.board.getCurrentAgentLocations())
self.webUi.updateCellAttrs(
self.board.team_a, self.board.team_b,
self.board.getCurrentAgentLocations())
elif order == 'reject_turn':
self.board.setCurrentAgentLocations(
rcv_dict['agents'][0], "A")
self.board.setCurrentAgentLocations(
rcv_dict['agents'][1], "B")
self.board.team_a = copy.deepcopy(rcv_dict['tiles_a'])
self.board.team_b = copy.deepcopy(rcv_dict['tiles_b'])
if self.team == "A":
self.board.printTiles_A()
self.board.printTiles_B()
print(self.board.getCurrentAgentLocations())
self.webUi.updateCellAttrs(
self.board.team_a, self.board.team_b,
self.board.getCurrentAgentLocations())
except OSError:
break
def send(self, msg): # event is passed by binders.
self.client_socket.send(bytes(msg, "utf8"))
time.sleep(1e-3)
if msg == "{quit}":
self.client_socket.close()
class NewSolver(object):
def __init__(self):
self.board = Board()
self.state = []
self.threshold = 0
self.team = 'A'
def calcScoreAverage(self):
num_of_cells = 0
for row in self.board.getBoardScores():
for cell in row:
if cell >= 0:
self.threshold += cell
num_of_cells += 1
self.threshold /= num_of_cells
def set_team(self, team):
self.team = team
def update_board(self, input_board):
self.board = input_board
def gen_state_list(self):
if self.team == 'A':
my_board = copy.deepcopy(self.board.team_a)
enemy_board = copy.deepcopy(self.board.team_b)
elif self.team == 'B':
my_board = copy.deepcopy(self.board.team_b)
enemy_board = copy.deepcopy(self.board.team_a)
self.state = [[0] * self.board.getBoardSize()[1]
for i in range(self.board.getBoardSize()[0])]
for x in range(self.board.getBoardSize()[0]):
for y in range(self.board.getBoardSize()[1]):
if my_board[x][y]:
self.state[x][y] = 0
else:
if self.board.getBoardScores()[x][y] >= self.threshold:
if enemy_board[x][y]:
self.state[x][y] = 3
else:
self.state[x][y] = 2
elif self.board.getBoardScores()[x][y] >= 0:
if enemy_board[x][y]:
self.state[x][y] = -1
else:
self.state[x][y] = 1
else:
self.state[x][y] = -2
def search_around(self, row, column):
destination = ((0, 1), (1, 1), (1, 0), (1, -1), (0, -1), (-1, -1),
(-1, 0), (-1, 1))
que = []
board_size = self.board.getBoardSize()
for a in destination:
if (0 <= row + a[0] < board_size()[0]) and (0 <= column + a[1] <
board_size()[1]):
state_cache1 = copy.deepcopy(self.state)
first_step = self.state[row + a[0]][column + a[1]]
state_cache1[row + a[0]][column + a[1]] = 0
for b in destination:
if (0 <= row + a[0] + b[0] < board_size()[0]) and (
0 <= column + a[1] + b[1] < board_size()[1]):
state_cache2 = copy.deepcopy(state_cache1)
second_step = state_cache1[row + a[0] +
b[0]][column + a[1] + b[1]]
state_cache2[row + a[0] + b[0]][column + a[1] +
b[1]] = 0
for c in destination:
if (0 <= (row + a[0] + b[0] + c[0]) < board_size[0]
) and (0 <= (column + a[1] + b[1] + c[1]) <
board_size[1]):
third_step = state_cache2[row + a[0] + b[0] +
c[0]][column + a[1] +
b[1] + c[1]]
que.append((first_step + second_step + third_step,
self.board.getBoardScores()[row+a[0]][column+a[1]] \
+ self.board.getBoardScores()[row+a[0]+b[0]][column+a[1]+b[1]] \
+ self.board.getBoardScores()[row+a[0]+b[0]+c[0]][column+a[1]+b[1]+c[1]],
(a, b, c)
))
practice = [0, 0] #[forのiを代入, 評価値を代入]
practices = []
for i, element in enumerate(que):
if element[0] > practice[1]:
practice = [i, element[0]]
practices = []
practices.append(practice)
elif element[0] == practice[1]:
practice = [i, element[0]]
practices.append(practice)
best_practices = [(0, (0, 0))]
for a in practices:
if que[a[0]][1] > best_practices[0][0]:
best_practices = []
best_practices.append([que[a[0]][1], que[a[0]][2][0]])
elif que[a[0]][1] == best_practices[0][0]:
best_practices.append([que[a[0]][1], que[a[0]][2][0]])
num_run = random.randrange(len(best_practices))
return [
row + best_practices[num_run][1][0],
column + best_practices[num_run][1][1]
]
class Solver(object):
def __init__(self):
self.board = Board()
self.state = []
#stateの値は-3が囲むことのできない平均値未満の領域、-2が平均値以上の領域、-1が平均値以下で点数の計算が行われていない領域、0以上は点数計算が終わり、lake_scoreに点数が代入し終わった領域。
self.lake_score = list()
def calcScoreAverage(self):
return sum(map(sum, self.board.board_scores)) / (
(self.board.row + 1) * (self.board.column + 1))
def state_init(self):
self.state = [[0] * self.board.getBoardSize()[1]
for i in range(self.board.getBoardSize()[0])]
print(len(self.state), len(self.state))
for x in range(self.board.getBoardSize()[0]):
for y in range(self.board.getBoardSize()[1]):
if self.board.getBoardScores()[x][y] < self.calcScoreAverage():
self.state[x][y] = -1
else:
self.state[x][y] = -2
def find_lakes(self, x, y, num):
self.state[x][y] = num
self.lake_score_buffer = 0
self.utl = ((1, 0), (-1, 0), (0, 1), (0, -1))
for diff in self.utl:
if (0 <= x + diff[0] < self.board.getBoardSize()[0]
and 0 <= y + diff[1] < self.board.getBoardSize()[1]
) and self.state[x + diff[0]][y + diff[1]] == -1:
self.lake_score_buffer += self.find_lakes(
x + diff[0], y + diff[1], num)
return abs(self.board.getBoardScores()[x][y]) + self.lake_score_buffer
def set_state(self):
self.state_init()
for i in (0, self.board.getBoardSize()[0] - 1):
for j in range(self.board.getBoardSize()[1]):
if self.state[i][j] == -1:
self.find_lakes(i, j, -3)
for i in range(self.board.getBoardSize()[0]):
for j in (0, self.board.getBoardSize()[1] - 1):
if self.state[i][j] == -1:
self.find_lakes(i, j, -3)
count = 0
for i in range(self.board.getBoardSize()[0]):
for j in range(self.board.getBoardSize()[1]):
if self.state[i][j] == -1:
self.lake_score.append(self.find_lakes(i, j, count))
count += 1
def set_board(self, in_board):
self.board = in_board
def get_state(self):
return self.state
def print_state(self):
for row in self.state:
for cell in row:
print(format(cell, '3d'), end="")
print("")
def print_state_and_score(self, direction=0):
if direction == 0:
for row in self.state:
for cell in row:
if cell < 0:
print(format(0, '3d'), end="")
else:
print(format(self.lake_score[cell], '3d'), end="")
print("")
print("")
print("")
elif direction == 1:
for y in range(self.board.getBoardSize()[1]):
for x in range(self.board.getBoardSize()[0]):
if self.state[x][y] < 0:
print(format(0, '3d'), end="")
else:
print(format(self.lake_score[self.state[x][y]], '3d'),
end="")
print("")
print("")
print("")
elif direction == 2:
for x in range(self.board.getBoardSize()[1])[::-1]:
for y in range(self.board.getBoardSize()[0]):
if self.state[x][y] < 0:
print(format(0, '3d'), end="")
else:
print(format(self.lake_score[self.state[x][y]], '3d'),
end="")
print("")
print("")
print("")
elif direction == 3:
for y in range(self.board.getBoardSize()[1]):
for x in range(self.board.getBoardSize()[0])[::-1]:
if self.state[x][y] < 0:
print(format(0, '3d'), end="")
else:
print(format(self.lake_score[self.state[x][y]], '3d'),
end="")
print("")
print("")
print("")
print(agents_a)
board.genScores(symmetry)
board.setFirstAgentCell(agents_a)
board.printBoardScore()
# board.printTiles_A()
# board.printTiles_B()
def printBoardScore_sq(level):
for row_socres in board.board_scores:
for score in row_socres:
if score > level:
print("██", end="")
else:
print(" ", end="")
print("")
board = Board()
genScores(10, 10, 0, [[1, 1], [9 - 1, 9 - 1]])
level = sum(map(sum, board.getBoardScores())) / 100
print(level)
print("")
printBoardScore_sq(level)
print("")
print("")
print("")
print("")
print("")
printBoardScore_sq(level - 2)