#! /bin/csh -f

set usage = "$0 PATTERN FILE..."
# Each FILE must contain words, one per line
# Enumerates all pairs of consecutive words that match PATTERN
# Lists the 10 most frequent
# Prints the tables for each file, side by side

set pat = "$1"; shift;
if ( $#argv < 1 ) then
  echo "usage: $usage"; exit 1
endif
set files = ( $* )

set outs = ( )
set head = ""
set dash = ""
set wd = 0
foreach f ( ${files} )
  echo $f
  set g = .${f:r}.frq
  cat $f \
    | egrep -v '^[-/]*$' \
    | sed \
        -e 's/t/k/g' \
        -e 's/[ao]$/y/' \
        -e 's/^\([q]*\)y/\1o/' \
    | enum-word-pairs \
    | egrep "${pat}" \
    | sort | uniq -c | expand \
    | sort -k1nr \
    | head -10 \
    > ${g}
  set outs = ( $outs ${g} )
  @ wd = ${wd} + 30
  set head = "${head}`printf '  count %-22s' ${f}`"
  set dash = "${dash}  ----- ----------------------"
end
echo "${head}"
echo "${dash}"
pr -m -t -i' '1 -w ${wd} $outs
/bin/rm -f $outs