Don’t ask why, it’s a complicated story

But this tag gets broken off, in the middle. Why? Because the code it recognizes a new tag and focuses on that.. So I’ve added two lines of code;

public function calculateHtmlPosition(htmlstr:String, pos:int):int
{
  // we return -1 (not found) if the position is bad
  if (pos < = -1)
    return -1;
 
  // characters that appears when a tag starts
  var openTags:Array = ["<","&"];
  // characters that appears when a tag ends
  var closeTags:Array = [">“,”;”];
  // the tag should be replaced with
  // ex: & is & and has 1 as length but normal
  // tags have 0 length
  var tagReplaceLength:Array = [0,1];
  // flag to know when we are inside a tag
  var isInsideTag:Boolean = false;
  var cnt:int = 0;
  // the id of the tag opening found
  var tagId:int = -1;
  var tagContent:String = “”;
 
  for (var i:int = 0; i < htmlstr.length; i++)
  {
    // if the counter passes the position specified
    // means that we reach the text position
    if (cnt>=pos)
      break;
    // current char
    var currentChar:String = htmlstr.charAt(i);
    // checking if the current char is in the open tag array
    if(!isInsideTag)
    {
      for (var j:int = 0; j < openTags.length; j++)
      {
        if (currentChar == openTags[j])
        {
          // set flag
          isInsideTag = true;
          // store the tag open id
          tagId = j;
        }
      }
    }
 
    if (!isInsideTag)
    {
      // increment the counter
      cnt++;
    } else {
      // store the tag content
      // needed afterwards to find new lines
      tagContent += currentChar;
    }
    if (currentChar == closeTags[tagId]) {
      // we ad the replace length
      if (tagId > -1) cnt += tagReplaceLength[tagId];
      // if we encounter the tag we increment the counter
      // because of new line character
      if (tagContent == ““) cnt++;
      // set flag
      isInsideTag = false;
      // reset tag content
      tagContent = “”;
    }
  }
  // return de position in html text
  return i;
}

This fixed my problem. My case is very rare (won’t probably occur in normal HTML), but there are other cases in which my fix might help

Maybe I don’t understand what you mean in you comment and if so please give me more info.

If you place your cursor after the first “a” you will see pos is “1″ and norm pos is “118″, now turn on HTML view and count the number of characters including whitespace.

Am I missing something?

For Justin: As you can see in the source file you cand use selection.beginIndex property of the RTE like this:

RTE.selection.beginIndex