Who can make this script smaller?

[Maarten van Steenbergen <M.vanSteenbergen@phys.uu.nl>]

Hi there,

Working on humbolt, I needed a program that could filter all URLs out of a
html page. So I tried some things I found in my Perl book, and made this
filter:

tr -d '\n'|perl -e '$_=<>;s/(<[Aa][^<]+<\/[Aa]>)/\n<lI>$1\n/g;print;'|grep lI

It started as a 125 byte script, I tried to make it smaller and smaller, now
it's 78 bytes (newline included). 
It's pretty fool-proof now, it worked on freshmeat.net's and slashdot.org's
index.html
Does someone know how to make this script smaller?
Why?
For fun!!



Regards,

Maarten


-- 
         Maarten van Steenbergen <steenbrg@phys.uu.nl>
   -----=====* http://www.phys.uu.nl/~steenbrg/ *=====-----
-
European Universities' Linux User Groups -- Misc list
http://humbolt.geo.uu.nl/eulug/