I've tried a couple of ways but it hasn't worked.

I want to start at a URL - say, odp.org - and have HTTrack go several levels
deep looking for URLs. I want it to then follow each of those URLs but only
save the homepage for each URL.

It'd be preferred if HTTrack would do at each of the linked sites what it did
at the first site, but that's not strictly required.

The reason why I want this is because I'm looking for homepages that contain a
certain string. I can grep that later, but if there's some way to tell HTTrack
to only save pages that contain that string that'd be even better.