ord() and chr() for UTF-8

utf ord chr.php

<?php

function utf8_ord(string $s, int $i = 0): int {
  $b0 = 0b00000000;
  $b1 = 0b10000000;
  $b2 = 0b11000000;
  $b3 = 0b11100000;
  $b4 = 0b11110000;
  $b5 = 0b11111000;

  $o = \ord($s[$i++]);
  if (($o & $b1) == $b0) {
    // 0xxxxxxx
    return $o;
  } else if (($o & $b3) == $b2) {
    // 110xxxxx 10xxxxxx
    $l = 1;
    $r = $o & ~$b3;
  } else if (($o & $b4) == $b3) {
    // 1110xxxx 10xxxxxx 10xxxxxx
    $l = 2;
    $r = $o & ~$b4;
  } else if (($o & $b5) == $b4) {
    // 11110xxx 10xxxxxx 10xxxxxx 10xxxxxx
    $l = 3;
    $r = $o & ~$b5;
  } else {
    throw new \Exception();
  }
  for (; $l > 0; $l--) {
    $o = \ord($s[$i++]);
    if (($o & $b2) == $b1) {
      // 10xxxxxx
      $r = ($r << 6) | ($o & ~$b2);
    } else {
      throw new \Exception();
    }
  }
  return $r;
}

function utf8_chr(int $n): string {
  $b0 = 0b00000000;
  $b1 = 0b10000000;
  $b2 = 0b11000000;
  $b3 = 0b11100000;
  $b4 = 0b11110000;
  $b5 = 0b11111000;

  if ($n < 0) {
    throw new \Exception();
  } else if ($n < (1 << 7)) {
    // 7 bits
    // 0xxxxxxx
    return \chr($n);
  } else if ($n < (1 << 11)) {
    // 11 bits
    // 110xxxxx 10xxxxxx
    $l = 1;
    $p = $b2;
  } else if ($n < (1 << 16)) {
    // 16 bits
    // 1110xxxx 10xxxxxx 10xxxxxx
    $l = 2;
    $p = $b3;
  } else if ($n < (1 << 21)) {
    // 21 bits
    // 11110xxx 10xxxxxx 10xxxxxx 10xxxxxx
    $l = 3;
    $p = $b4;
  } else {
    throw new \Exception();
  }
  $s = '';
  for (; $l > 0; $l--) {
    // 6 bits
    // 10xxxxxx
    $s = \chr($b1 | ($n & 0xFF & ~$b2)).$s;
    $n >>= 6;
  }
  $s = \chr($p | $n).$s;
  return $s;
}