#! /bin/bash -ue
# Last edited on 2026-07-23 10:59:19 by stolfi

# The "unique" column is how many lexemes have a single occurrence in the sample.

sizeopt="$1"; shift  # "whole" or "trunc"
smpsecs=( "$@" )     # {LANG}/{BOOK}/{SEC} ...
    
for kind in raw gud bad ; do
  sFile="summary-${sizeopt}-${kind}.txt"
  rm -fv ${sFile} | sed -e 's:^:  :g' 
  echo "  creating ${sFile}"
  printf "# %-38s\n" "Counts for ${kind} text (${sizeopt})" >> ${sFile}
  printf "# %-14s %7s %7s %7s\n" "sample/sec" "tokens" "lexemes" "unique" >> ${sFile}
  printf "# %-14s %7s %7s %7s\n" "--------------" "-------" "-------" "-------" >> ${sFile}
  for smpsec in  ${smpsecs[@]} ; do
    ifile="dat/${smpsec}/${sizeopt}/${kind}.wfr";
    cat ${ifile} \
      | gawk -v smpsec=${smpsec} \
          ' /./{ w++; t+=$1; if ($1==1) { u++; } } 
            END{ printf "  %-14s %7d %7d %7d\n",smpsec,t,w,u;} 
          ' \
      >> ${sFile}
  done
done