solution.c

// The solution can be done by counting the number of vowel occurrences and then replacing each vowel using that count as references.

// This enum helps store the count of vowels in an array.
enum {
  A = 0,
  E,
  I,
  O,
  U,
  a,
  e,
  i,
  o,
  u
};

char* sortVowels(char* s) {
  int vowelsCount[] = {0, 0, 0, 0, 0, 0, 0, 0, 0, 0};

  // Count the vowels from the given string.
  char* c = s;
  while (*c != 0) {
    // Switch case is used for faster lookup.
    switch (*c) {
      case 'A':
        ++vowelsCount[A];
        break;
      case 'E':
        ++vowelsCount[E];
        break;
      case 'I':
        ++vowelsCount[I];
        break;
      case 'O':
        ++vowelsCount[O];
        break;
      case 'U':
        ++vowelsCount[U];
        break;
      case 'a':
        ++vowelsCount[a];
        break;
      case 'e':
        ++vowelsCount[e];
        break;
      case 'i':
        ++vowelsCount[i];
        break;
      case 'o':
        ++vowelsCount[o];
        break;
      case 'u':
        ++vowelsCount[u];
        break;
    }
    ++c;
  }

  // From the first vowel, replace each vowel in the given string according to the vowel count.
  int vowel = A;
  c = s;
  while (*c != 0) {
    // Switch case is used for faster lookup.
    switch (*c) {
      case 'A':
      case 'E':
      case 'I':
      case 'O':
      case 'U':
      case 'a':
      case 'e':
      case 'i':
      case 'o':
      case 'u':
        while (vowelsCount[vowel] <= 0) ++vowel;
        switch (vowel) {
          case A:
            *c = 'A';
            break;
          case E:
            *c = 'E';
            break;
          case I:
            *c = 'I';
            break;
          case O:
            *c = 'O';
            break;
          case U:
            *c = 'U';
            break;
          case a:
            *c = 'a';
            break;
          case e:
            *c = 'e';
            break;
          case i:
            *c = 'i';
            break;
          case o:
            *c = 'o';
            break;
          case u:
            *c = 'u';
            break;
        }
        --vowelsCount[vowel];
        break;
    }
    ++c;
  }

  // Return the modified string.
  return s;
}