Function utf8NextCodepoint

lib/Utf8/Utf8.cpp:79–125

}

uint32_t utf8NextCodepoint(const unsigned char** string) {
  if (**string == 0) {
    return 0;
  }

  const unsigned char lead = **string;
  const int bytes = utf8CodepointLen(lead);
  const uint8_t* chr = *string;

  // Invalid lead byte (stray continuation byte 0x80-0xBF, or 0xFE/0xFF)
  if (bytes == 1 && lead >= 0x80) {
    (*string)++;
    return REPLACEMENT_GLYPH;
  }

  if (bytes == 1) {
    (*string)++;
    return chr[0];
  }

  // Validate continuation bytes before consuming them
  for (int i = 1; i < bytes; i++) {
    if ((chr[i] & 0xC0) != 0x80) {
      // Missing or invalid continuation byte — skip all bytes consumed so far
      *string += i;
      return REPLACEMENT_GLYPH;
    }
  }

  uint32_t cp = chr[0] & ((1 << (7 - bytes)) - 1);  // mask header bits

  for (int i = 1; i < bytes; i++) {
    cp = (cp << 6) | (chr[i] & 0x3F);
  }

  // Reject overlong encodings, surrogates, and out-of-range values
  const bool overlong = (bytes == 2 && cp < 0x80) || (bytes == 3 && cp < 0x800) || (bytes == 4 && cp < 0x10000);
  const bool surrogate = (cp >= 0xD800 && cp <= 0xDFFF);
  if (overlong || surrogate || cp > 0x10FFFF) {
    (*string)++;
    return REPLACEMENT_GLYPH;
  }

  *string += bytes;

  return cp;
}

void utf8AppendCodepoint(uint32_t cp, std::string& out) {
  if (cp < 0x80) {