穿越迷宫

Question

我正在创建一个迷宫游戏，该游戏将被机器遍历和解决。我创建了一个迷宫类，其中包含迷宫的起始位置和结束位置，以及包含在2d bool矢量中的迷宫本身。我正在惹恼的是如何实际编码上下移动和穿过迷宫来完成。我的起点是[11] [4]在迷宫中，我们的教授告诉我们，最好的方法是检查当前位置周围的所有4个位置，如果它是真的（也就是它是路径而不是墙）把它推到堆栈上。我从概念上理解这意味着什么，但我无法想象如何正确编码，任何帮助将不胜感激。仅供参考，有一个位置结构可以简化如何表达位置。

struct Location  {
    friend std::ostream &operator <<(std::ostream &os, const Location &location) {
        os << "(" << location.x << ", " << location.y << ")";
        return os;
    }
    bool operator ==(const Location &rhs) const {return x == rhs.x && y == rhs.y;}
    bool operator !=(const Location &rhs) const {return !(*this == rhs);}
    operator bool() const {return x >= 0;}
    Location(int x=-1, int y=-1) : x(x), y(y) {}
    int x, y;
};

class Maze;

Maze load(std::string filename);

class Maze {
    friend std::ostream &operator <<(std::ostream &os, Maze &maze);
    friend Maze load(std::string filename);
  public:
    Maze(std::vector<std::vector<bool> > specifics, const Location &startPos, const           Location &endPos);
    bool solve();
    //void run();
  private:
    bool contains(const Location &location) const;
    bool isPath(const Location &location) const;
    int height() {return spec.size();}
    int width() {return spec[0].size();}
    std::vector<std::vector<bool> > spec;
    Location start, finish;
    Location current;
};


bool Maze::solve() {
    stack<Location> location; //vector to hold location objects/paths
    Location current; //will hold current position in maze
    current = start; //set current to start position in beginning
    location.push(current); //push first maze location (start) onto vector
    //cout << current;
    ///need to set current to top of stack; while traversing maze current is top of stack and if current hits (==) finish return true
    while (!location.empty()) //while location still has values inside
    {
        current=location.top();//set current to top of stack
        cout << current << endl;
        cout << spec[6][4];
        if (current==finish) //if current == finish the maze is solved
            return true;
        // for loop to start moving around... but how?        
    }
}

Answer 1

这是伪代码;

1. push the starting location in a stack ( you can use std::stack)
2. while end location is not reached and stack is not empty
    2.1  search all 4 surrounding locations of *top* location one by one
           if any of them is a path, push it to stack, go 2.1
           if none of them is a path, pop and remove top element from stack. go 2


3. if stack if empty and end location is not reached, maze can not be solved.

修改

这大致是代码（注意，我没有编译它）

bool solve(Location currentLocation, Location endLocation)
{
  if(currentLocation == endLocation)
  {
    return true;
  }

  Location newLoc1(currentLocation.x-1,currentLocation.y-1); 
  if(contains(newLoc) && isPath(newLoc))
  {
    return solve(newLoc,endLocation);
  }

  Location newLoc2(currentLocation.x+1,currentLocation.y-1);
  if(contains(newLoc) && isPath(newLoc))
  {
    return solve(newLoc,endLocation);
  }

  Location newLoc3(currentLocation.x-1,currentLocation.y+1);
  if(contains(newLoc) && isPath(newLoc))
  {
    return solve(newLoc,endLocation);
  }

  Location newLoc4(currentLocation.x+1,currentLocation.y+1);
  if(contains(newLoc) && isPath(newLoc))
  {
    return solve(newLoc,endLocation);
  }

  return false;
}

您可能需要添加其他逻辑。显然你会进入无限循环，因为在进入之前我没有检查过是否有任何单元被访问过。不知何故，你必须记住它（可能是一个访问列表）。