POJ 2632 Crashing Robots (simulated coordinate adjustment) (RE caused by fflush)

Question link: http://poj.org/problem? Id = 2632

Let's start by saying that things went smoothly after 1Y yesterday and things were not published again until this afternoon. It's really like a sea of questions. From now on, the AC is a passer-by --

Originally, 2632 was a slightly disgusting simulation, but after all it was a simulation. Generally, the simulation is based on the AC as long as the example has passed, but this question is very special.

At the beginning, I couldn't run the test example until I found that the illustrations in the question were different from the coordinate system I conceived (it seems that the illustrations cannot be easily ignored)

The coordinate system given in this question is y for the vertical axis, x for the horizontal axis, and y for ascending order from bottom to top, x for ascending order from left to right

The two-dimensional array record map I used, that is to say, my portrait is the X axis and increases from top to bottom, the horizontal is the Y axis, and the left to right increments

I found this problem after I tuned the program for nearly an hour.

Then I suddenly got disconnected and didn't know how to implement it. A sudden total of negative results left me with no idea.

After a while, I waited, and plannedColumn stealing

This is because coordinates are used only when robots are defined, and coordinates are not used in commands.

Then I thought like this, in terms of process processing, I still use the array coordinate system, while doing some conversions during input and storage.

scanf ("%d%d%s",&x,&y,to);MAP[B - y][x - 1] = i;rob[i].x = B - y;rob[i].y = x - 1;

Then you don't have to worry about this coordinate system anymore.

Later, the code was quickly formed and several groups of data were fully transferred.

Then I checked the data range and MAP range, and found that I opened a far greater value than the maximum range. Because it was a simulation and no pointer, I checked the code several times and checked the array subscript.

No problem found. I ran all the test data from discuss. I got stuck for a long time.

Finally, there is no way to change it. You can only change it until the AC is changed. (=)

I changed itChange the input character to the input string. Only the first character of the string (s [0]) is used. The purpose is not to control spaces (but RE)

Then, because no space is needed, I removed the fflush (stdin) function that originally controls the space carriage return.

Then submit the AC. As for why fflush causes RE, I really don't understand. Baidu directly said that fflush is not a library function, but a supplement to the library function. Some linux will cause incompatibility.

Although it is not CE, it is not excluded from this. However, I often use fflush when processing Input streams. This is the first time that a problem occurs.

If fflush causes RE, you should give yourself a warning.

In general, this question is also a simulation. Although it is troublesome, It is a simulation question after all.

The Code is as follows:

#include 
 
  #include 
  
   #include 
   
    #include 
    
     struct n{    int x,y;    int to;}rob[10000];int MAP[150][150];int A,B;int mov (int no,int turn){    int x = rob[no].x,y = rob[no].y;    int to = rob[no].to;    //printf ("!1!%d %d\n",x,y);    MAP[x][y] = -1;    switch (to)    {    case 1:        if (turn == 1)        {            rob[no].to = 4;            //y--;        }else if (turn == 2)        {            rob[no].to = 2;            //y++;        }else if (turn == 3)        {            x--;        }        break;    case 2:        if (turn == 1)        {            rob[no].to = 1;            //x--;        }else if (turn == 2)        {            rob[no].to = 3;            //x++;        }else if (turn == 3)        {            y++;        }        break;    case 3:        if (turn == 1)        {            rob[no].to = 2;            //y++;        }else if (turn == 2)        {            rob[no].to = 4;            //y--;        }else if (turn == 3)        {            x++;        }        break;    case 4:        if (turn == 1)        {            rob[no].to = 3;            //x++;        }else if (turn == 2)        {            rob[no].to = 1;            //x--;        }else if (turn == 3)        {            y--;        }        break;    }    //printf ("!2!%d %d\n",x,y);    if (x < 0 || y < 0 || x >= B || y >= A)    {        printf ("Robot %d crashes into the wall\n",no + 1);        return 0;    }    if (MAP[x][y] != -1)    {        printf ("Robot %d crashes into robot %d\n",no + 1,MAP[x][y] + 1);        return 0;    }    rob[no].x = x;    rob[no].y = y;    MAP[x][y] = no;    return 1;}int main(){    //freopen ("1.txt","w",stdout);    int N;    int i,k;    scanf ("%d",&N);    while (N--)    {        int n,m;        memset (MAP,-1,sizeof (MAP));        memset (rob,0,sizeof (rob));        scanf ("%d%d",&A,&B);        scanf ("%d%d",&n,&m);        for (i = 0;i < n;i++)        {            int x,y;            char to[10];            //fflush (stdin);            scanf ("%d%d%s",&x,&y,to);            MAP[B - y][x - 1] = i;            rob[i].x = B - y;            rob[i].y = x - 1;            if (to[0] == 'N')            {                rob[i].to = 1;            }else if (to[0] == 'E')            {                rob[i].to = 2;            }else if (to[0] == 'S')            {                rob[i].to = 3;            }else if (to[0] == 'W')            {                rob[i].to = 4;            }        }        int tf = 1;        for (i = 0;i < m;i++)        {            int no,lop,iact;            char act[10];            //fflush (stdin);            scanf ("%d%s%d",&no,act,&lop);            if (act[0] == 'L')                iact = 1;            else if (act[0] == 'R')                iact = 2;            else if (act[0] == 'F')                iact = 3;                for (k = 0;k < lop;k++)                    if (tf)                        tf = mov (no - 1,iact);        }        if (tf)            puts ("OK");    }    return 0;}