count from right end of string

hi all,
I'm trying to use substr(p,c,l) in awk but i'd like to have the 'position' argument count from the right end of the string instead of left end.
evidently, just putting a '-' sign in front of the number does not work. any ideas?
cheers,
kl

Try something like
nawk '{print(substr($0,length($0)-4,1))}' myfile
Which'll print the fifth character from the right hand side of each record (assume position is zero-indexed).

Cheers
ZB