def buyPhase(hand, deck, discard, play, totalDeck, p1):
''' The buy phase for a regular dominion turn'''
global supplyCards, supplyAmounts
tmpHand = list()
handcopy = card_format.transformCards(hand.copy())
while len(hand) > 0:
card = hand.pop(-1)
if card.isTreasure():
p1.changeCoins(card.getCoins())
play.append(card)
else: tmpHand.append(card)
hand = tmpHand
possibleBuys = list()
for card in supplyCards:
if card.getCost() <= p1.getCoins() and supplyAmounts[card.getName()] > 0:
possibleBuys.append(card.getName())
possibleBuys.append("none") # gotta add that it may be better to buy nothing
buys = bestBuy(handcopy, possibleBuys, p1.getBuys())
if buys != "none": supplyAmounts[buys] -= 1
discard = card_format.newCard(discard, buys)
totalDeck = card_format.allDeckCards(hand, deck, discard, play)
if gSAPPrevious != None:
# skip providing values the first turn
updateValues(handcopy)
return hand, deck, discard, play, totalDeck, p1
def env_step(action):
"""
Arguments
---------
action : int
the action taken by the agent in the current state
Returns
-------
result : dict
dictionary with keys {reward, state, isTerminal} containing the results
of the action taken
"""
global gCurrentState, gTerminalPosition
global p1
global totalDeckLists
reward = -1.0
isTerminal = False
totalDeck = gCurrentState[0]
hand, deck, discard, play = gCurrentState[1]
possibleBuys = gCurrentState[2]
for buy in action:
if buy == "province":
reward = 1
isTerminal = True
if buy != "none":
supplyAmounts[buy] -= 1
discard = card_format.newCard(discard, buy)
totalDeck = card_format.allDeckCards(hand, deck, discard, play)
hand, deck, discard, play, totalDeck, p1 = cleanupPhase(
hand, deck, discard, play, totalDeck, p1)
hand, deck, discard, play, totalDeck, p1 = actionPhase(
hand, deck, discard, play, totalDeck, p1)
hand, deck, discard, play, totalDeck, p1, buys = buyPhase(
hand, deck, discard, play, totalDeck, p1)
# for phase in phases:
# if phase == "action": hand, deck, discard, play, totalDeck, p1 = actionPhase(hand, deck, discard, play, totalDeck, p1)
# elif phase == "buy": buys = buyPhase(hand, deck, discard, play, totalDeck, p1)
# elif phase == "cleanup": hand, deck, discard, play, totalDeck, p1 = cleanupPhase(hand, deck, discard, play, totalDeck, p1)
# else: print("OH NO")
gCurrentState = [totalDeck, [hand, deck, discard, play], buys]
totalDeckLists.append(totalDeck)
result = {
"reward": reward,
"state": gCurrentState,
"isTerminal": isTerminal
}
return result