http://www.perlmonks.org?node_id=75004

A friend of mine recently had an assignment where he had to split up a text file and print it out as a triangle. Of course this was in "c" and so, just because I could I had to see just how simple it would be in perl. The idea is to take a ctriangle.txt file like
a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a
and end up centered over 80 chars
a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a
The same principle applies to any text file, basically splitting words only on spaces or new lines
My best shot was 132 chars (\n inc). I especially like the new trick I learnt.
/[ ]+/
Which matches space/new line combinations.

Well this is the best I could do, after staring at it for half an hour it wasnt getting any shorter.

open F,"ctriangle.txt";sub n{print " "x(40-$r/2)."$x\n"};for(split/[ ]+/,join'',<F>){$c=$r,n,$x=''if( $r=length($x.=" $_"))>$c}n
- nashdj

Update:
With jeroenes idea, if I undef $/ by assigning it undef from an unused variable (not pretty but we're talking size) its down to 128 chars.

open F,"ctriangle.txt";sub n{print " "x(40-$r/2)."$x\n"};$/=$w;$_=<F>; for(split){$c=$r,n,$x=''if( $r=length($x.=" $_"))>$c}n