> On May 18, 1:19 pm, Geoffrey Summerhayes <sumr@gmail.com> wrote:

> > On May 18, 11:24 am, hende@gmail.com wrote:

> > > Gang,

> > >      I have been working on this for a few hours and am frustrated
> > > beyond all extent.  I have tried to research this on the web as well
> > > with no success.  I am trying to match certain contents within a
> > > wrapper div.  So for example if the inside of the wrapper div was the
> > > following:

> > > <div id="wrapper">
> > >     <a href="#">a great link that contain text and symbols</a>
> > >     <div> ... </div>
> > >     <div> ... </div>
> > > </div>

> > > I would like to strip out all the internal div's.  But because there
> > > can be alot of internal div's, I figured it would be less processor
> > > intensive to just match the first 'a' tag and repopulate the wrapper
> > > div with the match.  I am trying to use something like the following
> > > regex:

> > > re = /^<a(.+)</a>/;

> > > with the following statment:

> > > $temp = document.getElementById('wrapper').innerHTML.match(re);

> > > but this is returning the entire contents of the wrapper div.  I have
> > > tried variations of the regex and either continue to get the entire
> > > contents or null returns.  Any help would greatly be appreciated.
> > > BTW, I can't match to the first \n because the contents may be
> > > touching (ie ...</a><div>...).

> > I'm not a big fan of using regexps for parsing HTML.
> > Getting a bulletproof expression is a major pain.

> > For example, here's one from 'Mastering Regular
> > Expressions 2nd edition' by J. Friedl (publisher
> > O'Reilly) for matching HTML tags:

> > /<("[^"]*"|'[^']*'|[^'">])*>/

> > How about...

> > var container=document.getElementById('wrapper');
> > var list=[];
> > while(container.hasChildNodes())
> > {
> >  if(!('tagName' in container.lastChild)||
> >      (container.lastChild.tagName.match(/^div$/i)==null))
> >  {
> >    list.push(container.lastChild);
> >  }
> >  container.removeChild(container.lastChild);}

> > while(list.length>0)
> > {
> >   container.appendChild(list.pop());

> > }

> > It does a little extra work to avoid
> > the nastiness of dealing with indexing
> > into a list that's being resized.

> > --
> > Geoff

> Geoff,

>      Thanks for the reply.  I was looking for something less processor
> intensive.  The inner div's can be in the hundreds.  Thats why I was
> looking at just isolating the first href tag and replacing the entire
> contents of the wrapper div.  Any other thoughts would be appreciated.

> Thanks,
> Dave