C++ 在C中使用无符号整数模拟有符号整数

这就是当人们试图变得"聪明"而不是遵循"保持简单，愚蠢"的最佳实践时所发生的事情.好的工程需要尽可能地编写simple个代码，例如:

bool is_signed_less_correct (unsigned a, unsigned b) {
  bool is_neg_a = is_negative(a);
  bool is_neg_b = is_negative(b);

  if(is_neg_a != is_neg_b) // one is negative
  {
    return is_neg_a; // if one is negative and it is a, return true otherwise false
  } 

  // both are negative or both are positive
  return a < b;
}

即使这段代码也有点太"聪明"了，因为它隐含地使用了-1==0xFFFF...是最大的2‘S补码符号数，因此无论它们是负的还是非负的，只要它们都是相同的符号，a < b都成立.

当然，您总是要编写一个小单元测试来判断它的健全性:https://godbolt.org/z/h4nKsffqr

输出:

-2 < -1 ? true  (is_signed_less_gustedt)
-1 < -1 ? false (is_signed_less_gustedt)
-1 <  0 ? false (is_signed_less_gustedt)
 0 < -1 ? false (is_signed_less_gustedt)
 0 <  1 ? true  (is_signed_less_gustedt)
 1 <  0 ? false (is_signed_less_gustedt)
 1 <  1 ? false (is_signed_less_gustedt)

-2 < -1 ? true  (is_signed_less_correct)
-1 < -1 ? false (is_signed_less_correct)
-1 <  0 ? true  (is_signed_less_correct)
 0 < -1 ? false (is_signed_less_correct)
 0 <  1 ? true  (is_signed_less_correct)
 1 <  0 ? false (is_signed_less_correct)
 1 <  1 ? false (is_signed_less_correct)