#!/bin/sh
#
# usage: mxref infile1 infile2 ... >outfile
#
# create index lines for all tokens in the given files
nawk '{
    sub("$"," ")
    gsub("[^a-zA-Z0-9_]+"," ")
    sub("^ ","")
    gsub(" "," " FILENAME " " sprintf("%3d",FNR) "\n")
    printf "%s",$0
}' ${1+"$@"} | egrep '.....* .* ' | sort -T .