Function unfinishedBytesUtf8

lib/internal/encoding/util.js:18–31

* Get a number of last bytes in an Uint8Array `data` ending at `len` that don't * form a codepoint yet, but can be a part of a single codepoint on more data. * @param {Uint8Array} data Uint8Array of potentially UTF-8 bytes * @param {number} len Position to look behind from * @returns {number} Nu

(data, len)

 * @returns {number} Number of unfinished potentially valid UTF-8 bytes ending at position `len`
 */
function unfinishedBytesUtf8(data, len) {
  // 0-3
  let pos = 0;
  while (pos < 2 && pos < len && (data[len - pos - 1] & 0xc0) === 0x80) pos++; // Go back 0-2 trailing bytes
  if (pos === len) return 0; // no space for lead
  const lead = data[len - pos - 1];
  if (lead < 0xc2 || lead > 0xf4) return 0; // not a lead
  if (pos === 0) return 1; // Nothing to recheck, we have only lead, return it. 2-byte must return here
  if (lead < 0xe0 || (lead < 0xf0 && pos >= 2)) return 0; // 2-byte, or 3-byte or less and we already have 2 trailing
  const lower = lead === 0xf0 ? 0x90 : lead === 0xe0 ? 0xa0 : 0x80;
  const upper = lead === 0xf4 ? 0x8f : lead === 0xed ? 0x9f : 0xbf;
  const next = data[len - pos];
  return next >= lower && next <= upper ? pos + 1 : 0;
}

/**
 * Merge prefix `chunk` with `data` and return new combined prefix.