#! /usr/bin/gawk -f # Last edited on 1999-12-08 07:59:00 by stolfi # Reads a stream of words, one per line. Extracts # the "dex" features. /^ *$/{next;} /./{ gsub(/^[q]/, "", $0); # Replace tables and gallows by "X": gsub(/[ktpf]/, "X", $0); gsub(/[cs]hh*/, "X", $0); gsub(/[ic]Xhh*/, "X", $0); # Replace dealers and finals by "D" gsub(/[dlrsmngjxv]/, "D", $0); # Erase circles and sticks: gsub(/[aoyi]/, "", $0); # Replace empty strings by "_" gsub(/^$/, "_", $0); print $0; }