1

1

src/environment/chengdu_majiang_env.py

@@ -2,7 +2,6 @@ import gym
 import numpy as np
 from gym import spaces
 
-
 from src.engine.calculate_fan import calculate_fan, is_seven_pairs, is_cleared, is_big_pairs
 from src.engine.chengdu_mahjong_engine import ChengduMahjongEngine
 from src.engine.scoring import calculate_score
@@ -14,12 +13,15 @@ class MahjongEnv(gym.Env):
         self.engine = ChengduMahjongEngine()
         self.scores = [100, 100, 100, 100]  # 四位玩家初始分数
         self.base_score = 1  # 底分
+        self.max_rounds = 100  # 最大轮数，防止游戏无限进行
+        self.current_round = 0  # 当前轮数
         self.action_space = spaces.Discrete(108)  # 动作空间：打牌的索引
         self.observation_space = spaces.Box(low=0, high=4, shape=(108,), dtype=np.int32)
 
     def reset(self):
         self.engine = ChengduMahjongEngine()
         self.scores = [100, 100, 100, 100]  # 每局重置分数
+        self.current_round = 0
         return self.engine.state.hands[self.engine.state.current_player]
 
     def step(self, action):
@@ -31,9 +33,23 @@ class MahjongEnv(gym.Env):
                 reward, done = self.handle_win()
             else:
                 reward, done = -1, False  # 默认小惩罚
+
+            # 检查是否有玩家分数 <= 0
+            if any(score <= 0 for score in self.scores):
+                done = True
+                reward = -100  # 游戏结束的惩罚（可根据需求调整）
         except ValueError:
             reward, done = -10, False  # 非法操作扣分
 
+        # 切换到下一个玩家
+        self.engine.state.current_player = (self.engine.state.current_player + 1) % 4
+        self.current_round += 1
+
+        # 如果达到最大轮数，结束游戏
+        if self.current_round >= self.max_rounds:
+            done = True
+            reward = 0  # 平局奖励或惩罚（可调整）
+
         return self.engine.state.hands[self.engine.state.current_player], reward, done, {}
 
     def handle_win(self):
@@ -44,8 +60,9 @@ class MahjongEnv(gym.Env):
         hand = self.engine.state.hands[winner]
         melds = self.engine.state.melds[winner]
         is_self_draw = True  # 假设自摸（后续可动态判断）
-        is_cleared = is_cleared(hand)
+
         conditions = {
+            "is_cleared": is_cleared(hand),
             "is_seven_pairs": is_seven_pairs(hand),
             "is_big_pairs": is_big_pairs(hand),
             # 添加其他条件...
@@ -62,9 +79,10 @@ class MahjongEnv(gym.Env):
 
         # 奖励设置为赢家得分
         reward = scores["winner"]
-        done = True  # 游戏结束
+        done = True  # 胡牌结束当前局
         return reward, done
 
     def render(self, mode="human"):
-        print("当前分数:", self.scores)
-        print("当前状态:", self.engine.state.hands[self.engine.state.current_player])
+        print(f"当前轮数: {self.current_round}")
+        print("玩家分数:", self.scores)
+        print("当前玩家状态:", self.engine.state.hands[self.engine.state.current_player])