Get NodesList by Xpath with Attributes Filters

AbrarNazeer · March 19th, 2008, 08:39 AM

//////////////////////////////////////////
       XML File
//////////////////////////////////////////

<Campaigns>
  <Campaign ScriptID="1111111" SeqNo="1">
    <ID>C38BA51F-5D86-4395-92B5-010F7A64D999</ID>
    <Url>http://www.test1.com</Url>
  </Campaign>
  <Campaign ScriptID="1111111" SeqNo="2">
    <ID>4395-92B5-010F7A64D999</ID>
    <Url>http://www.test2.com</Url>
  </Campaign>
  <Campaign ScriptID="1111111" SeqNo="3">
    <ID>C38BA51F-010F7A64D999</ID>
    <Url>http://www.test3.com</Url>
  </Campaign>
  <Campaign ScriptID="1111111" SeqNo="4">
    <ID>010F7A64D999</ID>
    <Url>http://www.test4.com</Url>
  </Campaign>
</Campaigns>

///////////////////////////////////////
CODE to parse the XML selected nodes
///////////////////////////////////////
var ScriptID = '921924826';
var random_num = 0;
while(random_num == 0)
{
  random_num = parseInt((Math.round(Math.random() *4)));
}

x=xmlDoc.getElementsByTagName
('Campaign[@ScriptID = \'' + ScriptID + '\' and @SeqNo = \'' + random_num + '\']');

alert('x.length : ' + x.length);

campaignID = x[0].childNodes[0].childNodes[0].nodeValue;

linkURL = x[0].childNodes[1].childNodes[0].nodeValue;

alert('ID : ' + campaignID + ' Url :' + linkURL);

//////////////////////////////
           Problem
//////////////////////////////
This code is parsing the above XML file perfectly in IE , but failed in
firefox(2.0.0.12).
It shows the x.length = 0 , why?
May be there is reason of XPath with attributes (ScriptID , SeqNo).
If any idea then please share it.

mhkay · March 19th, 2008, 08:56 AM

My guess would be that it's the whitespace problem. Microsoft parsers by default strip the whitespace-only text nodes from your document while parsing; nearly all other parsers leave them intact.

Michael Kay
http://www.saxonica.com/
Author, XSLT Programmer's Reference and XPath 2.0 Programmer's Reference

Martin Honnen · March 19th, 2008, 09:06 AM

getElementsByTagName is a W3C DOM Core method and takes a single element name as its argument (e.g. "Campaign"), not an XPath expression.
If you want to to use XPath with Mozilla or with Opera then use the W3C DOM Level 3 XPath API documented here: http://developer.mozilla.org/en/docs..._in_JavaScript and here: http://www.w3.org/TR/DOM-Level-3-XPath/.

AbrarNazeer · March 20th, 2008, 04:13 AM

Martin Honnen! Thanks for ur time.
I ve reveiwed both of these links but did not get any idea.
(How to list all Campaigns with specfic ScriptID?)

Martin Honnen · March 20th, 2008, 08:20 AM

Well those documents describe the XPath API and how to use that, basically the DOM document gets a method named 'evaluate' to which you pass an XPath expression, a context node, an optional namespace resolver, a result type and an optional result.
Here is an example how to use that with Mozilla:

Code:

var httpRequest = new XMLHttpRequest();
httpRequest.open('GET', 'campaigns.xml', true);
httpRequest.onreadystatechange = function () {
  if (httpRequest.readyState == 4) {
    var doc = httpRequest.respon****ML;
    var id = '1111111';
    if (typeof doc.evaluate != 'undefined') {
      var xpathResult = doc.evaluate('Campaigns/Campaign[@ScriptID = "' + id + '"]', doc, null, XPathResult.ORDERED_NODE_ITERATOR_TYPE, null);
      var campaign;
      while ((campaign = xpathResult.iterateNext()) != null) {
        var id = campaign.getElementsByTagName('ID')[0];
        alert(id.textContent);
      }
    }
  }
};

httpRequest.send(null);

(editing) The property the code is supposed to use is called 'r e s p o n s e X M L' without the spaces, it seems the forum software censors that as it contains a certain three letter word.

AbrarNazeer · March 26th, 2008, 01:02 AM

Thanks a lot Martin Honnen!
Excellent Solution ,this is working fine now.
For Cross Domain Accesss , I ve just added the
"netscape.security.PrivilegeManager.enablePrivileg e("UniversalBrowserRead");"