un bout de AWK

Finalement une solution en AWK, pas si moche quand on a pas de "set".

BEGIN { W = 14 }
{
 for (i=1; i<=length($1); i++) {
 window = substr($1,i,W)
 # compare each char of window to all char of window
 S = 0 # to count match
 for (j=1; j<=W; j++) {
 for (k=1; k<=W; k++) {
 S+= (substr(window,j,1) == substr(window,k,1)) ? 1 : 0
 }
 }
 if (S == W) { # if each char match only itsef
 print i + W - 1
 next # finish
 }
 }
}