Pasted

int check_utf8 (const char * string) {
  // if you're trying to read and understand the UTF-8 functions... I'm sorry
  int codepoint;
  for (; *string; string ++) {
    if (*string < 0x80) continue;
    if (*string < 0xc2) return 0;
    if (*string < 0xe0) {
      if (string[1] < 0x80) return 0;
      string ++;
      continue;
    }
    if (*string < 0xf0) {
      codepoint = get_utf8_codepoint((const unsigned char *) string + 1, *string, 2);
      if (codepoint < 0x800) return 0;
      if ((codepoint >= 0xd800) && (codepoint < 0xe000)) return 0;
      string += 2;
      continue;
    }
    if (*string < 0xf5) {
      codepoint = get_utf8_codepoint((const unsigned char *) string + 1, *string, 3);
      if (codepoint < 0x10000) return 0;
      if (codepoint > 0x10ffff) return 0;
      string += 3;
      continue;
    }
    return 0;
  }
  return 1;
}
 
int get_utf8_codepoint (const unsigned char * continuations, unsigned char initial, unsigned char readahead) {
  if (readahead > 6) return -1;
  if (!readahead) return initial;
  unsigned result = initial & ((1 << (6 - readahead)) - 1);
  for (; readahead --; continuations ++) {
    if ((*continuations < 0x80) || (*continuations >= 0xc0)) return -1;
    result <<= 6;
    result |= *continuations & 0x3f;
  }
  return result;
}