Thanks for help I have it working now by directly manipulating what comes 
out of XSLTProcessor.TransformToDocument()

"Julius Mong" <jxm96c@h...> wrote in message 
news:eHQP2bkbGHA.4372@T......
> "Martin Honnen" <mahotrash@y...> wrote in message 
> news:OCo423gbGHA.3856@T......
>>
>>
>> Julius Mong wrote:
>>
>>
>>> OK I've tried that before, problem is it returns undefined. I tried 
>>> this:
>>>
>>> document.getElementsByTagName('body')[0]
>>>
>>> I get undefined, I tried:
>>>
>>> document.getElementsByTagName('html')[0].childNodes[0] // or firstChild
>>>
>>> I get object HTMLHeadElement
>>
>>
>> That sounds rather odd. Does the XSLT stylesheet generate a body element? 
>> Do you call
>>   document.getElementsByTagName('body')[0]
>> after the document has been loaded?
>
> Yes after loaded the document and it works perfectly in IE, it seems to be 
> a known bug in firefox as mentioned in first post... just wondering if 
> there's another way to do the highlighting using XSLT rather than DOM. I 
> store the words/phrases to be matched in temp cookies, would XSLT have 
> access to these? I mean I've seen highlighting XSLT samples, but I don't 
> know if it's possible to have the template receive the search keywords as 
> arguments at run time, the samples all have hard coded search strings 
> already...
>
> Thanks,
>
> Julius
>
>