• coding=utf-8
• > OS : Linux 3.2.0-60-generic #91-Ubuntu
• > Author : yaolong
• > Mail : dengyaolong@yeah.net
• > Time : 2014年06月01日 星期日 13:13:39
• map 0 left,1 down,2 up ,3 right
• a,h=> left ,s,j=>down, w,k=>up, d,l=>right
• coding=utf-8
• > OS : Linux 3.2.0-60-generic #91-Opensuse
• > Author : David
• > Mail : xxx@xxx.com
• > Time : 2015年08月14日 星期日 08:58:39
• map 0 left,1 down,2 up ,3 right
• a,j=> left ,s,k=>down, w,i=>up, d,l=>right

来到这里，我假设大家都玩过2048了，于是我们先分析一下怎么来写一个简单的2048？回想一下，游戏的过程是很简单的： 开始===>( 进行操作(上下左右)===>矩阵移动===>数字合并===>计分 )===>退出括号内就是游戏的进行过程了，我们退出大概就是两个：1.出现2048,胜利退出 2.无法合并同时无法出现新数（是且的关系）懂得这个运行过程后，我们就一步步来写，整个过程也不会难，哪怕你跟我一样基本是 新手 。先看看简略图，看起来虽然真的很朴素，但是爱美的你，可以在看完我的程序后自己美化。 储存结构： 列表，其中包括二维矩阵列表，即所谓的mtr[4][4],以及栈，栈是用于退回，如果大家不想要这个功能可以直接忽略。我们的变量定义：declare = “←：a/h ↓: s/j ↑: w/k →: d/l ,q(uit),b(ack)” illegal = “Illegal operation!” noefficient = “This move has no efficient” score = 0 step = 0 mtr = init() # init the matrix mtr_stk = [] # use step for back scr_stk = []稍微解释一下： declare是解释操作提示，自己喜欢怎么写就怎么写～ illegal是我们输入非正常操作提示,，可以省略 noefficient是我们往这个方向没效果，也可以省略～ score是得分，step是走了多少步， mtr是我们的二维方阵， mtr_stk，scr_stk是储存栈，用于后退功能,如果不要后退，还是可以省略～ 算法实现： ·「注：其实没有算法可言….更准确叫函数实现。」（1）.在我们这个没有界面的2048中，我们只能用制表符啦，所以我们需要一个display函数。 这是一个非常简单的事情，跟C语言的printf基本一致, 像如下代码，基本就能绘制出一个 固定数字 的表格了。 print “┌”+(“─”5+”┬”)3+”─”5+”┐” print “│%4d │%4d │%4d │%4d │”%(1,2048,25,4) print “├”+(“─”5+”┼”)3+”─”5+”┤” print “│%4d │%4d │%4d │%4d │”%(1,2048,25,4) print “├”+(“─”5+”┼”)3+”─”5+”┤” print “│%4d │%4d │%4d │%4d │”%(1,2048,25,4) print “├”+(“─”5+”┼”)3+”─”5+”┤” print “│%4d │%4d │%4d │%4d │”%(1,2048,25,4) print “└”+(“─”5+”┴”)3+”─”5+”┘”但是我们需要的是将固定的数字变成我们矩阵mtr的数字。另外可以，注意到一点，数字输出的形式是固定的，表格边线，只有交叉点是不同，其他是0。而交叉点有五个，我们可以走循环（觉得麻烦可以直接忽略）为了美观，我们 不将0输出 ，所以我们输出时候选择的是字符串即%s,这样，我们遇到0,就可以输出’ ‘,其他直接输出。我们可以用python的 三目运算符号print “%4s “ %(mtr[i][j] if mtr[i][j] else ‘ ‘)所以我们的绘制表格写成这样子： 最直接的方法！def T(a): return a if a else ‘ ‘ def display(mtr): print “┌”+(“─”5+”┬”)3+”─”5+”┐” print “│%4s │%4s │%4s │%4s │”%(T(mtr[0][0]),T(mtr[0][1]),T(mtr[0][2]),T(mtr[0][3])) print “├”+(“─”5+”┼”)3+”─”5+”┤” print “│%4s │%4s │%4s │%4s │”%(T(mtr[1][0]),T(mtr[1][1]),T(mtr[1][2]),T(mtr[1][3])) print “├”+(“─”5+”┼”)3+”─”5+”┤” print “│%4s │%4s │%4s │%4s │”%(T(mtr[2][0]),T(mtr[2][1]),T(mtr[2][2]),T(mtr[2][3])) print “├”+(“─”5+”┼”)3+”─”5+”┤” print “│%4s │%4s │%4s │%4s │”%(T(mtr[3][0]),T(mtr[3][1]),T(mtr[3][2]),T(mtr[3][3])) print “└”+(“─”5+”┴”)3+”─”5+”┘”如果 走循环 可以这样写：就是枚举交叉点情况。def display(mtrx): #output function a = (“┌”, “├”, “├”, “├”, “└”) b = (“┬”, “┼”, “┼”, “┼”, “┴”) c = (“┐”, “┤”, “┤”, “┤”, “┘”) for i in range(4): print a[i] + (“─” * 5 + b[i]) * 3 + (“─” * 5 + c[i]) for j in range(4): print “│%4s” % (mtrx[i][j] if mtrx[i][j] else ‘ ‘), print “│” print a[4] + (“─” * 5 + b[4]) * 3 + (“─” * 5 + c[4])（2）有了表格雏形，我们需要对矩阵进行初始化，then,我们就可以看到一个像模像样的2048开始啦！初始化矩阵是非常简单的，我们在矩阵中选两个不同位置，设置为2其他为0即可。先设置为0(没有memset，╮(╯▽╰)╭)但是这样就可以啦：(看不惯？没关系，非常容易理解，慢慢习惯吧！)mtr = [[0 for i in range(4)] for j in range(4)]我们选的位置是采取随机的，要保证不同，方法很多，但是我们python的random.sample就非常棒了！我们用简单的一句ran_pos = random.sample(range(16), 2)便可以得到一个由两个不同组成的片段之后我们通过 4i+j 进行二维到一维的映射，同样也用（ num/4 ，num%4）进行一维到二维的映射所以我们初始话的过程也是十分简单。(要是偷懒，也可以直接忽略随机过程，自己指定吧！哈哈)def init(): #initial of matrix mtr = [[0 for i in range(4)] for j in range(4)]
ran_pos = random.sample(range(16), 2) mtr[ran_pos[0]/4][ran_pos[0]%4] = mtr[ran_pos[1]/4][ran_pos[1]%4] = 2 return mtr怎么样？2048有2个2啦！！！O(∩_∩)O哈哈~ (3)动起来！！让我们的2048动起来！ 老实说，这个过程是全个程序最核心的部分，也是最难过的～不过理解了也就不难了。我们先看向左移动吧。记得我们玩的时候，合并过的格子是不会再在这次合并的，比如 某一行为 2 2 2 2 最后移动到的结果是4 4 0 0 而不是 8 0 0 0。我们要理解这个过程： 我们对于每一行的操作是一样的，即行间无关，所以我们现在要讨论第i行，当我们向左动起来的时候，我们从最左开始确定，我们一开始合并前两个，再合并前三个，再合并前4个。比如 2 2 4 4 => 4 0 4 4 => 4 4 0 4 =>4 4 4 0 => 4 8 0 0注意两点： 1.合并过的不再参与合并，所以我们需要一个visit列表,用到前面所讲的映射 2.合并的情况=>相邻的相等，移动的情况=>当前元素为0python代码visit=[] score=0 for i in range(4): for j in range(1, 4): for k in range(j,0,-1): if mtr[i][k - 1] == 0 : mtr[i][k - 1] = mtr[i][k] mtr[i][k] = 0 elif mtr[i][k - 1] == mtr[i][k] and 4 * i + k - 1 not in visit and 4 * i + k not in visit: mtr[i][k - 1] *= 2 mtr[i][k] = 0 score += mtr[i][k - 1] visit.append(4 * i + k) visit.append(4 * i + k - 1)当你想明白这个过程后，整个move过程就自然清晰起来了。 这个是本文第一个，最后一个比较长的函数了！def move(mtr, dirct):#the core code!move by the four direction score = 0 visit = [] if dirct == 0: # left for i in range(4): for j in range(1, 4): for k in range(j,0,-1): if mtr[i][k - 1] == 0 : mtr[i][k - 1] = mtr[i][k] mtr[i][k] = 0 elif mtr[i][k - 1] == mtr[i][k] and 4 * i + k - 1 not in visit and 4 * i + k not in visit: mtr[i][k - 1] *= 2 mtr[i][k] = 0 score += mtr[i][k - 1] visit.append(4 * i + k) visit.append(4 * i + k - 1) elif dirct == 1: # down for j in range(4): for i in range(3, 0, -1): for k in range(0,i): if mtr[k+1][j] == 0: mtr[k+1][j] = mtr[k][j] mtr[k][j]=0 elif mtr[k+1][j]==mtr[k][j] and (4 *(k+1)+j) not in visit and (4k+j) not in visit: mtr[k+1][j]=2 mtr[k][j]=0 score=mtr[k+1][j] visit.append(4(k)+j) visit.append(4(k+1)+j) elif dirct == 2: # up for j in range(4): for i in range(1,4): for k in range(i,0,-1): if mtr[k-1][j]==0: mtr[k-1][j]=mtr[k][j] mtr[k][j]=0 elif mtr[k-1][j]==mtr[k][j] and (4 *(k-1)+j) not in visit and (4k+j) not in visit: mtr[k-1][j]=2 mtr[k][j]=0 score += mtr[k-1][j] visit.append(4(k)+j) visit.append(4(k-1)+j) elif dirct == 3: # right for i in range(4): for j in range(3, 0, -1): for k in range(j): if mtr[i][k+1] == 0: mtr[i][k+1] = mtr[i][k] mtr[i][k]=0 elif mtr[i][k] ==mtr[i][k+1] and 4 * i + k + 1 not in visit and 4 * i + k not in visit: mtr[i][k+1]=2 mtr[i][k]=0 score+=mtr[i][k+1] visit.append(4i+k+1) visit.append(4i+k) return score（4）当你来到这步，已经进入轻松的状态了，让控制台读入你的操作，短短两行，其中一行是转成小写dirct = raw_input(“Step :%d Score :%d (%s):” % (step, score, declare)) dirct = dirct.lower()（5）判断读入操作并处理if dirct == “q”: break elif dirct == “a” or dirct == “h”: dirct = 0 elif dirct == “s” or dirct == “j”: dirct = 1 elif dirct == “w” or dirct == “k”: dirct = 2 elif dirct == “d” or dirct == “l”: dirct = 3 elif dirct == “b”: if len(mtr_stk) == 1: print “Can’t Back..” else: mtr_stk.pop() scr_stk.pop() mtr = copy.deepcopy(mtr_stk[-1]) score = scr_stk[-1] step -= 1 continue else: print illegal continue我是vim 的脑残粉，所以还支持了hjkl，如果你没有习惯vim什么的，直接asdw就可以了。上面最麻烦最值得注意的就是要用深复制.这个bug我调了很久啊亲，用”=”伤不起啊。上面的处理意思就是： 映射方向，和b,q的处理（有continue是因为我们将嵌套于while中）。(6)移动的处理在上面处理后，筛选出移动的操作，之后我们调用移动函数，进行加分，和比较。tmp = copy.deepcopy(mtr) op_scr = move(mtr, dirct) if tmp != mtr: score = score + op_scr update(mtr) #更新 display(mtr) tmp = copy.deepcopy(mtr) mtr_stk.append(tmp) # 插入后退队列 scr_stk.append(int(score)) step = step + 1 # 步数加1 else: print noefficient注意： 不能仅仅用加分来判断！因为我们很多时候都是没有碰撞合并，等待下一个2,4出现 我们用深复制，将老的矩阵放在tmp上，再移动，将老新矩阵比较，没改变的时候就输出 noefficient,否则就加分，插入新数字，并显示，最后将矩阵压入栈中，因为要后退！(7)插入新数字我们遍历0-16,并映射后，将对应0的位置放入random列表，我们用random.choice选出一个，同理我们生成一个ran_num=[2,4],选取加分之后对应位置加分咯！但是注意，当我们没有空位时候不要插数字。def update(mtr): ran_pos=[] ran_num=[2,4] for i in range(4): for j in range(4): if mtr[i][j]==0: ran_pos.append(4*i+j) if len(ran_pos)>0: k=random.choice(ran_pos) n=random.choice(ran_num) mtr[k/4][k%4]=n(8)最后一个啦！！！游戏的循环条件这个还是很简单的。游戏进行，当前仅当：没有2048且（有0或者可以合并）。所以我们只需要判断2048 in mtr,有就结束，没有 == > 0 in mtr ?有就继续，没有==>可以合并？有继续，没有==》，就返回结束啦def go_on(mtr, score): if 2048 in mtr: print “Great!You win!Your score is “, score raw_input(“Press any key to continue…”) exit() if 0 in mtr: return True for i in range(4): for j in range(4): if i < 3 and mtr[i][j] == mtr[i + 1][j]: return True if j < 3 and mtr[i][j] == mtr[i][j + 1]: return True print “Gameover!” return False(9)一些其他的东西注意一开始要画出开始的画面（循环外），一开始要将初始矩阵，初始分数，压入栈。 最后来个最终版本的#!/usr/bin/env python

coding=utf-8

#********************

> OS : Linux 3.2.0-60-generic #91-Ubuntu

> Author : yaolong

> Mail : dengyaolong@yeah.net

> Time : 2014年06月01日 星期日 13:13:39

#******************** import random import copy def T(a): return a if a else ‘ ‘ def display(mtr): print “┌”+(“─”5+”┬”)3+”─”5+”┐” print “│%4s │%4s │%4s │%4s │”%(T(mtr[0][0]),T(mtr[0][1]),T(mtr[0][2]),T(mtr[0][3])) print “├”+(“─”5+”┼”)3+”─”5+”┤” print “│%4s │%4s │%4s │%4s │”%(T(mtr[1][0]),T(mtr[1][1]),T(mtr[1][2]),T(mtr[1][3])) print “├”+(“─”5+”┼”)3+”─”5+”┤” print “│%4s │%4s │%4s │%4s │”%(T(mtr[2][0]),T(mtr[2][1]),T(mtr[2][2]),T(mtr[2][3])) print “├”+(“─”5+”┼”)3+”─”5+”┤” print “│%4s │%4s │%4s │%4s │”%(T(mtr[3][0]),T(mtr[3][1]),T(mtr[3][2]),T(mtr[3][3])) print “└”+(“─”5+”┴”)3+”─”*5+”┘” def init(): mtr = [[0 for i in range(4)] for j in range(4)] # 小小蛋疼.. ran_pos = random.sample(range(16), 2) mtr[ran_pos[0]/4][ran_pos[0]%4] = mtr[ran_pos[1]/4][ran_pos[1]%4] = 2

return mtr def go_on(mtr, score): if 2048 in mtr: print “Great!You win!Your score is “, score raw_input(“Press any key to continue…”) exit() if 0 in mtr: return True for i in range(4): for j in range(4): if i < 3 and mtr[i][j] == mtr[i + 1][j]: return True if j < 3 and mtr[i][j] == mtr[i][j + 1]: return True print “Gameover!” return False def move(mtr, dirct): score = 0 visit = [] if dirct == 0: # left for i in range(4): for j in range(1, 4): for k in range(j,0,-1): if mtr[i][k - 1] == 0 : mtr[i][k - 1] = mtr[i][k] mtr[i][k] = 0 elif mtr[i][k - 1] == mtr[i][k] and 4 * i + k - 1 not in visit and 4 * i + k not in visit: mtr[i][k - 1] = 2 mtr[i][k] = 0 score += mtr[i][k - 1] visit.append(4 * i + k) visit.append(4 * i + k - 1) # for i in range(4): # for j in range(3): elif dirct == 1: # down for j in range(4): for i in range(3, 0, -1): for k in range(0,i): if mtr[k+1][j] == 0: mtr[k+1][j] = mtr[k][j] mtr[k][j]=0 elif mtr[k+1][j]==mtr[k][j] and (4 *(k+1)+j) not in visit and (4k+j) not in visit: mtr[k+1][j]=2 mtr[k][j]=0 score=mtr[k+1][j] visit.append(4(k)+j) visit.append(4(k+1)+j) elif dirct == 2: # up for j in range(4): for i in range(1,4): for k in range(i,0,-1): if mtr[k-1][j]==0: mtr[k-1][j]=mtr[k][j] mtr[k][j]=0 elif mtr[k-1][j]==mtr[k][j] and (4 *(k-1)+j) not in visit and (4k+j) not in visit: mtr[k-1][j]=2 mtr[k][j]=0 score += mtr[k-1][j] visit.append(4(k)+j) visit.append(4(k-1)+j) elif dirct == 3: # right for i in range(4): for j in range(3, 0, -1): for k in range(j): if mtr[i][k+1] == 0: mtr[i][k+1] = mtr[i][k] mtr[i][k]=0 elif mtr[i][k] ==mtr[i][k+1] and 4 * i + k + 1 not in visit and 4 * i + k not in visit: mtr[i][k+1]=2 mtr[i][k]=0 score+=mtr[i][k+1] visit.append(4i+k+1) visit.append(4i+k) return score def update(mtr): ran_pos=[] ran_num=[2,4] for i in range(4): for j in range(4): if mtr[i][j]==0: ran_pos.append(4*i+j) if len(ran_pos)>0: k=random.choice(ran_pos) n=random.choice(ran_num) mtr[k/4][k%4]=n

map 0 left,1 down,2 up ,3 right

a,h=> left ,s,j=>down, w,k=>up, d,l=>right

declare = “←：a/h ↓: s/j ↑: w/k →: d/l ,q(uit),b(ack)” illegal = “Illegal operation!” noefficient = “This move has no efficient” if name == ‘main’: score = 0 step = 0 mtr = init() mtr_stk = [] # for back scr_stk = [] tmp = copy.deepcopy(mtr) mtr_stk.append(tmp) scr_stk.append(0) display(mtr) while go_on(mtr, score): dirct = raw_input(“Step :%d Score :%d (%s):” % (step, score, declare)) dirct = dirct.lower() if dirct == “q”: break elif dirct == “a” or dirct == “h”: dirct = 0 elif dirct == “s” or dirct == “j”: dirct = 1 elif dirct == “w” or dirct == “k”: dirct = 2 elif dirct == “d” or dirct == “l”: dirct = 3 elif dirct == “b”: if len(mtr_stk) == 1: print “Can’t Back..” else: mtr_stk.pop() scr_stk.pop() mtr = copy.deepcopy(mtr_stk[-1]) score = scr_stk[-1] step -= 1 continue else: print illegal continue tmp = copy.deepcopy(mtr) op_scr = move(mtr, dirct) if tmp != mtr: score = score + op_scr update(mtr) #更新 display(mtr) tmp = copy.deepcopy(mtr) mtr_stk.append(tmp) # 插入后退队列 scr_stk.append(int(score)) step = step + 1 # 步数加1 else: print noefficient还有我github的版本，大致一样,点 这里如果我的算法有什么不对，请提出，因为写得匆忙，我未能很完整第测试，但是我自己玩了几把还是OK了，没玩到2048╮(╯▽╰)╭。。。 修改版 ：#!/usr/bin/env python

coding=utf-8

#********************

> OS : Linux 3.2.0-60-generic #91-Opensuse

> Author : David

> Mail : xxx@xxx.com

> Time : 2015年08月14日 星期日 08:58:39

#******************** import random import copy import os import sys import tty, termios def T(a): return a if a else ‘ ‘ def display(mtr): print “┌”+(“─”5+”┬”)3+”─”5+”┐” print “│%4s │%4s │%4s │%4s │”%(T(mtr[0][0]),T(mtr[0][1]),T(mtr[0][2]),T(mtr[0][3])) print “├”+(“─”5+”┼”)3+”─”5+”┤” print “│%4s │%4s │%4s │%4s │”%(T(mtr[1][0]),T(mtr[1][1]),T(mtr[1][2]),T(mtr[1][3])) print “├”+(“─”5+”┼”)3+”─”5+”┤” print “│%4s │%4s │%4s │%4s │”%(T(mtr[2][0]),T(mtr[2][1]),T(mtr[2][2]),T(mtr[2][3])) print “├”+(“─”5+”┼”)3+”─”5+”┤” print “│%4s │%4s │%4s │%4s │”%(T(mtr[3][0]),T(mtr[3][1]),T(mtr[3][2]),T(mtr[3][3])) print “└”+(“─”5+”┴”)3+”─”*5+”┘” def init(): mtr = [[0 for i in range(4)] for j in range(4)] # 小小蛋疼.. ran_pos = random.sample(range(16), 2) mtr[ran_pos[0]/4][ran_pos[0]%4] = mtr[ran_pos[1]/4][ran_pos[1]%4] = 2

return mtr def go_on(mtr, score): if 2048 in mtr: print “Great!You win!Your score is “, score raw_input(“Press any key to continue…”) exit() if 0 in mtr: return True for i in range(4): for j in range(4): if i < 3 and mtr[i][j] == mtr[i + 1][j]: return True if j < 3 and mtr[i][j] == mtr[i][j + 1]: return True print “Gameover!” return False def move(mtr, dirct): score = 0 visit = [] if dirct == 0: # left for i in range(4): for j in range(1, 4): for k in range(j,0,-1): if mtr[i][k - 1] == 0 : mtr[i][k - 1] = mtr[i][k] mtr[i][k] = 0 elif mtr[i][k - 1] == mtr[i][k] and 4 * i + k - 1 not in visit and 4 * i + k not in visit: mtr[i][k - 1] = 2 mtr[i][k] = 0 score += mtr[i][k - 1] visit.append(4 * i + k) visit.append(4 * i + k - 1) # for i in range(4): # for j in range(3): elif dirct == 1: # down for j in range(4): for i in range(3, 0, -1): for k in range(0,i): if mtr[k+1][j] == 0: mtr[k+1][j] = mtr[k][j] mtr[k][j]=0 elif mtr[k+1][j]==mtr[k][j] and (4 *(k+1)+j) not in visit and (4k+j) not in visit: mtr[k+1][j]=2 mtr[k][j]=0 score=mtr[k+1][j] visit.append(4(k)+j) visit.append(4(k+1)+j) elif dirct == 2: # up for j in range(4): for i in range(1,4): for k in range(i,0,-1): if mtr[k-1][j]==0: mtr[k-1][j]=mtr[k][j] mtr[k][j]=0 elif mtr[k-1][j]==mtr[k][j] and (4 *(k-1)+j) not in visit and (4k+j) not in visit: mtr[k-1][j]=2 mtr[k][j]=0 score += mtr[k-1][j] visit.append(4(k)+j) visit.append(4(k-1)+j) elif dirct == 3: # right for i in range(4): for j in range(3, 0, -1): for k in range(j): if mtr[i][k+1] == 0: mtr[i][k+1] = mtr[i][k] mtr[i][k]=0 elif mtr[i][k] ==mtr[i][k+1] and 4 * i + k + 1 not in visit and 4 * i + k not in visit: mtr[i][k+1]=2 mtr[i][k]=0 score+=mtr[i][k+1] visit.append(4i+k+1) visit.append(4i+k) return score def update(mtr): ran_pos=[] ran_num=[2,4] for i in range(4): for j in range(4): if mtr[i][j]==0: ran_pos.append(4*i+j) if len(ran_pos)>0: k=random.choice(ran_pos) n=random.choice(ran_num) mtr[k/4][k%4]=n

map 0 left,1 down,2 up ,3 right

a,j=> left ,s,k=>down, w,i=>up, d,l=>right

declare = “←：a/j ↓: s/k ↑: w/i →: d/l ,q(uit),b(ack)” illegal = “Illegal operation!” noefficient = “This move has no efficient” if name == ‘main’: score = 0 step = 0 mtr = init() mtr_stk = [] # for back scr_stk = [] tmp = copy.deepcopy(mtr) mtr_stk.append(tmp) scr_stk.append(0) display(mtr) fd = sys.stdin.fileno() old_settings = termios.tcgetattr(fd) while go_on(mtr, score): try : tty.setraw( fd ) ch = sys.stdin.read( 1 ) finally : termios.tcsetattr(fd, termios.TCSADRAIN, old_settings) print “Step :%d Score :%d (%s):” % (step, score, declare) #dirct = raw_input(“Step :%d Score :%d (%s):” % (step, score, declare)) #dirct = dirct.lower() dirct = ch if dirct == “q”: break elif dirct == “a” or dirct == “j”: dirct = 0 elif dirct == “s” or dirct == “k”: dirct = 1 elif dirct == “w” or dirct == “i”: dirct = 2 elif dirct == “d” or dirct == “l”: dirct = 3 elif dirct == “b”: if len(mtr_stk) == 1: print “Can’t Back..” else: mtr_stk.pop() scr_stk.pop() mtr = copy.deepcopy(mtr_stk[-1]) score = scr_stk[-1] step -= 1 continue else: print illegal continue tmp = copy.deepcopy(mtr) op_scr = move(mtr, dirct) if tmp != mtr: score = score + op_scr update(mtr) #更新 display(mtr) tmp = copy.deepcopy(mtr) mtr_stk.append(tmp) # 插入后退队列 scr_stk.append(int(score)) step = step + 1 # 步数加1 else: print noefficient