> So what would it take to create a more complex
> parser?  Is this something we should think about? 
> We use HTTrack for LOTS of sites and would make good
> use of a better JS parser.

It would be nice, but think about what httrack does, it modifies the source
file link to be relative links.  If you had
var a,b,c;
a="path";
b="file";
c="asp";
d="text";
write.document("(a href='"+a+"/"+b+"."+c+"')"+d+"(/a)");

The path in var a would have to be modified, and asp in var c would have to be
changed to html but nothing else.
Not only would you need a full parser, you would have to have the AI to
propagate the changes back into the source. Alternatively, it could comment
out and replace just some of the JS code.

If it was easy, Xavier would have already done it.