Function FromUTF8

native/thirdpart/flatbuffers/util.h:512–553

Converts whatever prefix of the incoming string corresponds to a valid UTF-8 sequence into a unicode code. The incoming pointer will have been advanced past all bytes parsed. returns -1 upon corrupt UTF-8 encoding (ignore the incoming pointer in this case).

// returns -1 upon corrupt UTF-8 encoding (ignore the incoming pointer in
// this case).
inline int FromUTF8(const char **in) {
  int len = 0;
  // Count leading 1 bits.
  for (int mask = 0x80; mask >= 0x04; mask >>= 1) {
    if (**in & mask) {
      len++;
    } else {
      break;
    }
  }
  if ((static_cast<unsigned char>(**in) << len) & 0x80)
    return -1;  // Bit after leading 1's must be 0.
  if (!len) return *(*in)++;
  // UTF-8 encoded values with a length are between 2 and 4 bytes.
  if (len < 2 || len > 4) { return -1; }
  // Grab initial bits of the code.
  int ucc = *(*in)++ & ((1 << (7 - len)) - 1);
  for (int i = 0; i < len - 1; i++) {
    if ((**in & 0xC0) != 0x80) return -1;  // Upper bits must 1 0.
    ucc <<= 6;
    ucc |= *(*in)++ & 0x3F;  // Grab 6 more bits of the code.
  }
  // UTF-8 cannot encode values between 0xD800 and 0xDFFF (reserved for
  // UTF-16 surrogate pairs).
  if (ucc >= 0xD800 && ucc <= 0xDFFF) { return -1; }
  // UTF-8 must represent code points in their shortest possible encoding.
  switch (len) {
    case 2:
      // Two bytes of UTF-8 can represent code points from U+0080 to U+07FF.
      if (ucc < 0x0080 || ucc > 0x07FF) { return -1; }
      break;
    case 3:
      // Three bytes of UTF-8 can represent code points from U+0800 to U+FFFF.
      if (ucc < 0x0800 || ucc > 0xFFFF) { return -1; }
      break;
    case 4:
      // Four bytes of UTF-8 can represent code points from U+10000 to U+10FFFF.
      if (ucc < 0x10000 || ucc > 0x10FFFF) { return -1; }
      break;
  }
  return ucc;
}

#ifndef FLATBUFFERS_PREFER_PRINTF
// Wraps a string to a maximum length, inserting new lines where necessary. Any