note 102620 modified in function.strip-tags by danbrown

strip_tags destroys the whole HTML behind the tags with invalid attributes. Like <img src="/images/image.jpg""> (look, there is an odd quote before >.)

So I wrote function which fixes unsafe attributes and replaces odd " and ' quotes with &quot; and &#39;.

<?php
function fix_unsafe_attributes($s) {
  $out = false;
  while (preg_match('/<([A-Za-z])[^>]*?>/', $s, $i, PREG_OFFSET_CAPTURE)) { // find where the tag begins
    $i = $i[1][1]+1;
    $out.= substr($s, 0, $i);
    $s = substr($s, $i);

    // scan attributes and find odd " and '
    while (((($i1 = strpos($s, '"')) || 1) && (($i2 = strpos($s, '\'')) || 1)) && ($i1 !== false || $i2 !== false) &&
           (($i = (int)(($i1 !== false) && ($i2 !== false) ? ($i1 < $i2 ? $i1 : $i2) : ($i1 == false ? $i2 : $i1))) !== false) &&
           ((($c = strpos($s, '>')) === false) || ($i < $c))) {

      $c = $s{$i};
      if (($i < 1) || ($s{$i-1} != '=')) {
        $out.= substr($s, 0, $i).($s{$i} == '"' ? '&quot;' : '&#39;'); // replace odd " and '
        $s = substr($s, $i+1);
      }else {
        $i++;
        $out.= substr($s, 0, $i);
        $s = substr($s, $i);

        if (($i = strpos($s, $c)) !== false) {
          $i++;
          $out.= substr($s, 0, $i);
          $s = substr($s, $i);
        }
      }
    }
  }
  return $out.$s;
}
?>

Maybe this function can be rewritten with simple regular expression but I have no luck to make it quickly.

--was--
strip_tags destroys the whole HTML behind the tags with invalid attributes. Like <img src="/images/image.jpg""> (look, there is an odd quote before >.)

So I wrote function which fixes unsafe attributes and replaces odd " and ' quotes with &quot; and &#39;.

function fix_unsafe_attributes($s) {
  $out = false;
  while (preg_match('/<([A-Za-z])[^>]*?>/', $s, $i, PREG_OFFSET_CAPTURE)) { // find where the tag begins
    $i = $i[1][1]+1;
    $out.= substr($s, 0, $i);
    $s = substr($s, $i);

    // scan attributes and find odd " and '
    while (((($i1 = strpos($s, '"')) || 1) && (($i2 = strpos($s, '\'')) || 1)) && ($i1 !== false || $i2 !== false) &&
           (($i = (int)(($i1 !== false) && ($i2 !== false) ? ($i1 < $i2 ? $i1 : $i2) : ($i1 == false ? $i2 : $i1))) !== false) &&
           ((($c = strpos($s, '>')) === false) || ($i < $c))) {

      $c = $s{$i};
      if (($i < 1) || ($s{$i-1} != '=')) {
        $out.= substr($s, 0, $i).($s{$i} == '"' ? '&quot;' : '&#39;'); // replace odd " and '
        $s = substr($s, $i+1);
      }else {
        $i++;
        $out.= substr($s, 0, $i);
        $s = substr($s, $i);

        if (($i = strpos($s, $c)) !== false) {
          $i++;
          $out.= substr($s, 0, $i);
          $s = substr($s, $i);
        }
      }
    }
  }
  return $out.$s;
}

Maybe this function can be rewritten with simple regular expression but I have no luck to make it quickly.

http://php.net/manual/en/function.strip-tags.php