Question

My goal is to count all word in html page as well as count fixed word in html page the prob is that using that function script tag text also get in count so how i remove script tag from counting keywords.
i this code MSO_ContentTable is id 0f div tag. give me any other solution on jquery also if there.

function CountWord(keyword) {

    var word = keyword.toUpperCase(),
        total = 0,
        queue = [document.getElementById('MSO_ContentTable')],
        curr, count = 0;

    while (curr = queue.pop()) {
        var check = curr.textContent;

        if (check != undefined) {

            for (var i = 0; i < curr.childNodes.length; ++i) {

                if (curr.childNodes[i].nodeName == "SCRIPT") {
                    // do nothing
                }
                else {
                    switch (curr.childNodes[i].nodeType) {
                        case 3: // 3
                            var myword = curr.childNodes[i].textContent.split(" ");

                            for (var k = 0; k < myword.length; k++) {
                                var upper = myword[k].toUpperCase();

                                if (upper.match(word)) {
                                    count++;
                                    wc++;
                                }
                                else  if((upper[0] >= 'A' && upper[0] <= 'Z') ||
                                         (upper[0] >= 'a' && upper[0] <= 'z') ||
                                         (upper[0] >= '0' && upper[0] <= '9')) {
                                    wc++
                                }                                    
                            }
                        case 1: // 1
                            queue.push(curr.childNodes[i]);
                    }
                }
            }
      }
}

thx
other problem is how i remove the tag which have their display property none?

Answer 1

In your code:

> queue = [document.getElementById('MSO_ContentTable')],
> curr, count = 0;
> 
> while (curr = queue.pop()) {

getElementById will only ever return a single node, so no need to put it in an array and no need to pop it later:

curr = document.getElementById('MSO_ContentTable');
if (curr) {
  // do stuff

.

>    var check = curr.textContent;

The DOM 3 Core textContent property is not supported by all browsers, you need to offer an alternative such as innerText, e.g.:

// Get the text within an element
// Doesn't do any normalising, returns a string
// of text as found.
function getTextRecursive(element) {
  var text = [];
  var self = arguments.callee;
  var el, els = element.childNodes;

  for (var i=0, iLen=els.length; i<iLen; i++) {
    el = els[i];

    // May need to add other node types here
    // Exclude script element content
    if (el.nodeType == 1 && el.tagName && el.tagName.toLowerCase() != 'script') {
      text.push(self(el));

    // If working with XML, add nodeType 4 to get text from CDATA nodes
    } else if (el.nodeType == 3) {

      // Deal with extra whitespace and returns in text here.
      text.push(el.data);
    }
  }
  return text.join('');
}

.

>    if (check != undefined) {

Given that check will always be a string (even if textContent or innerText are used instead of the above function), testing against undefined doesn’t seem appropriate. Also, I don’t understand why this test is done before looping over the child nodes.

Anyhow, the getText function above will return the text content without script elements, so you can just use that to get the text then play with it as you want. You may need to normalise whitespace as different browsers will return different amounts.

PS. I should note that arguments.callee is restricted in ES5 strict mode, so if yo plan on using strict mode, replace that expression with an explicit call to the function.

Edit

To exclude not visible elements, you need to test each one to see if it’s visible. Only test elements, don’t test text nodes as if their parent element is not visible, the text won’t be.

Note that the following is not widely tested yet, but works in IE 6 and recent Firefox, Opera and Chrome at least. Please test thoroughly before using more widely.

  // The following is mostly from "myLibrary"
  // <http://www.cinsoft.net/mylib.html>
  function getElementDocument(el) {
    if (el.ownerDocument) {
      return el.ownerDocument;
    }
    if (el.parentNode) {
      while (el.parentNode) {
        el = el.parentNode;
      }
      if (el.nodeType == 9 || (!el.nodeType && !el.tagName)) {
        return el;
      }

      if (el.document && typeof el.tagName == 'string') {
        return el.document;
      }
      return null;
    }
  }


  // Return true if element is visible, otherwise false
  //    
  // Parts borrowed from "myLibrary"
  // <http://www.cinsoft.net/mylib.html>
  function isVisible(el) {
    if (typeof el == 'string') el = document.getElementById(el);

    var doc = getElementDocument(el);
    var reVis = /\bhidden\b|\bnone\b/;
    var styleObj, isVis; 

    // DOM compatible
    if (doc && doc.defaultView && doc.defaultView.getComputedStyle) {
      styleObj = doc.defaultView.getComputedStyle(el, null);

    // MS compatible
    } else if (el.currentStyle) {
      styleObj = el.currentStyle;
    }

    // If  either visibility == hidden || display == none
    // then element is not visible
    return !reVis.test(styleObj.visibility + ' ' + styleObj.display);
  }