> +*/*funny*
> 
> But when I scan a large dir of files, it starts
> taking all of them, not just the ones with funny in
> their filename.


Makes sense because there is no restriction.  So HTTrack is going to go
through it's normal "scope" (mirroring from the start URL and below) getting
everything it can so all your filter is really saying is "also include any
external links with the word 'funny' in them".