On Friday 07 June 2002 01:47, Juraj Novak wrote:
> Greetings everyone!
>
> I am using KHTMLPart as a DOM tree build
> mechanism, it works very ok. But I have 2
> questions:
>
> 1. How to make parsing of a XML document
> synchronous, not to have needs for connecting to
> completed() signal?

By using QT's DOM (QDom classes), if you do not need CSS.

(see QDomDocument::setContent)

>
> 2. How to know offset of an DOM element(node) in
> a parsed source file? Say when I ask DOM classes
> for all <img> tags, I would like to know their
> possitions from the begining of a file. Is it
> possible at least for some nodes, e.g.
> attributes? Or is it completely imposible and
> one must use regular expressions for this? But
> regular expressions don't build DOM trees,
> that's the point :|

You cannot. DOM is not defined for doing this.

However, what do you want to do with your <IMG> elements. Change them? If yes, 
try to use QDomDocument::elementsByTagName. (I suppose that KHTML has this 
function too, as it is a DOM function.)

If your problem is to save back the file to disk, then you should really use 
QDom (for example QDomDocument::toString or QDomDocument::save . These are 
*not* DOM functions.)

>
> TAI

Have a nice day/evening/night!

> ____________________________________
> P. S. V. P. U.
> http://www.pobox.sk/
>
> >> Visit http://mail.kde.org/mailman/listinfo/kde-devel#unsub to
> >> unsubscribe <<


 
>> Visit http://mail.kde.org/mailman/listinfo/kde-devel#unsub to unsubscribe <<