c++ - 如何更改 UNIX 语法以获得正确的输出？

Question

我找到了这个 Kattis 问题的解决方案:

https://open.kattis.com/problems/funnygames

但它使用 >?= unix 语法，我需要更改它才能在 Kattis 中编译。我分别将它们更改为 std::min 和 std::max，但我没有得到正确的答案(我每次都得到 Michael)。

原始代码:

/* Sample solution to "Funny Games" from NCPC 2005
 * Algorithm: essentially continuous DP, keep track of winning intervals
 * Author: Per Austrin
 */ 
#include <cmath>
#include <cassert>
#include <algorithm>
#include <cstdio>

using namespace std;

typedef pair<double, double> pdd;

template <class It, class OIt>
OIt ival_union(It begin, It end, OIt dest) {
  sort(begin, end);
  while(begin != end) {
    *dest = *begin++;
    while(begin != end && begin->first < dest->second + 1e-8)
      dest->second >?= begin++->second; //Change 1
    ++dest;
  }
  return dest;
}   

int main(void) {
  int n, k;
  double x, f[10], maxf;
  pdd win[10000], lose;
  for (scanf("%d", &n); n--; ) {
    scanf("%lf%d", &x, &k);
    maxf = lose.second = 0;
    for (int i = 0; i < k; ++i)
      scanf("%lf", f+i), win[i] = make_pair(1, 1 / f[i]), maxf >?= f[i];  //Change 2
    for (int nwin = k, l = 0; x > lose.second; ++l) {
      nwin = ival_union(win + l , win + nwin, win + l) - win;
      lose.second = (lose.first = win[l].second) / maxf;
      if (l < nwin-1) lose.second <?= win[l+1].first; //Change 3
      for (int i = 0; i < k; ++i)
    win[nwin++] = make_pair(lose.first/f[i], lose.second/f[i]);
    }
    assert(fabs(x-lose.first) > 1e-6 && fabs(x-lose.second) > 1e-6);
    printf("%s\n", x < lose.first ? "Nils" : "Mikael");
  }
}

我的 3 个改变:

std::max(dest->second, begin++->second); //Change 1

scanf("%lf", f + i), win[i] = make_pair(1, 1 / f[i]), std::max(maxf, f[i]); //Change 2

if (l < nwin - 1) std::min(lose.second, win[l + 1].first); //Change 3

输入是

4
6 2 0.25 0.5
10 2 0.25 0.5
29.29 4 0.3 0.7 0.43 0.54
29.30 4 0.3 0.7 0.43 0.54

输出应该是

Mikael
Nils
Nils
Mikael

但在我得到改变之后

Mikael
Mikael
Mikael
Mikael

Answer 1

阅读评论后，我意识到我做了 max(a,b)，但没有做 a=max(a,b)，所以在我做出更改后它按预期工作。变化是:

dest->second=std::max(dest->second, begin++->second);

scanf("%lf", f + i), win[i] = make_pair(1, 1 / f[i]), maxf=std::max(maxf, f[i]);

if (l < nwin - 1) lose.second=std::min(lose.second, win[l + 1].first);

感谢您的帮助!