I think the main thing is that this regex

s#(http://[\w./=\?-]+)#<A HREF=\"$1\">$1</A>#
[download]

is (1) inserting anchor tags that you don't want and (2) spitting out a repeat of the URL - as you see $1 appears twice in the replace string. (BTW, and sorry if I'm telling you something you probably know anyway, but the "#" is here used as a delimiter instead of a "/". It performs the same function, but makes the thing easier to manage because the search string uses so many "/"s.)

I think this should do more what you want:

if (#(http://[\w./=\?-]+)#sgi and !$flag and (/myaccount/))
[download]

(You don't need to do any substitution because you just want to print out the URL; you only need the match to check you've got the right line. You could probably even slim down the regexes a bit more by combining them into one.)

§ George Sherston