#! /usr/bin/gawk -f # Last edited on 1999-09-23 23:07:35 by stolfi # usage: count-word-lengths < WORDFILE > FREQFILE # Input must have one word per line # Empty lines are ignored BEGIN{ hi = -1; s = 0; n = 0; } /./ { w = $0; m = NF; nw[m]++; if (nw[m] == 1) { ww[m] = w; } s += m; n++; hi = (m > hi ? m : hi); } END { printf "avg length = %4.1f\n\n", s/n; printf "len nwords example \n"; printf "--- ------ ------------------\n"; for(m=1;m<=hi;m++) { printf "%3d %6d %s\n", m, nw[m], ww[m]; } }