dp学习之AvoidRoads解法二（算法优化）

在上一篇AvoidRoads解法中，虽然能得到正确的答案但存在很大的瑕疵，比如需要分三种情况:
第一种 x>0,y=0
第二种 x=0,y>0
第三种 x>0,y>0
程序分完这三种情况后显得很长，因此有必要进行优化。
优化源代码如下（c++）:

#include <iostream>
#include <vector>
#include <cstdio>
#include <string>
#include <cstring>
#include <sstream>


using namespace std;
int dx[] = {1, 0};
int dy[] = {0, 1};
class AvoidRoads
{
public:
    long long numWays(int width, int height, vector<string> &bad);
};


int main(void)
{
    AvoidRoads l;
    vector<string> str;
#if 0
    str.push_back("0 3 0 2");
    str.push_back("1 2 1 3");
        str.push_back("2 2 2 3");
    str.push_back("3 2 3 3");
    str.push_back("4 2 4 3");
        str.push_back("5 2 5 3");
    str.push_back("6 2 6 3");
    str.push_back("7 2 7 3");
    str.push_back("8 2 8 3");
    str.push_back("9 2 9 3");
#endif
    str.push_back("0 0 0 1");
    str.push_back("6 6 5 6");
    cout << l.numWays(6, 6, str) << endl;
    return 0;
}
long long AvoidRoads::numWays(int width, int height, vector<string> &bad)
{
    long long D[width + 1][height + 1];
    char is_bad1[128], is_bad2[128];
    for(int i = 0; i <= width; i++)
    {
        for(int j = 0; j <= height; j++)
        {
            D[i][j] = 0;
        }
    }
    D[0][0] = 1;
    for(int i = 0; i <= width; i++)
    {
        for(int j = 0; j <= height; j++)
        {
            for(int k = 0; k < 2; k++)
            {
                int is_bad = 0;
                int tx = i + dx[k];
                int ty = j + dy[k];
#if 1        /*If you know why the next two line is needful, you are so clever*/
                if(tx > width || ty > height)
                    continue;
#endif
                snprintf(is_bad1, sizeof(is_bad1), "%d %d %d %d", i, j, tx, ty);
                snprintf(is_bad2, sizeof(is_bad2), "%d %d %d %d", tx, ty, i, j);
                for(int p = 0; p < bad.size(); p++)
                {
                    if(bad[p] == string(is_bad1) || bad[p] == string(is_bad2))
                    {
                        is_bad = 1;
                        break;
                    }
                }
                if(is_bad != 1)
                {
                    D[tx][ty] += D[i][j];
                }
            }
        }
    }
    return D[width][height];
}