#!/bin/bash

shopt -s nullglob
set -e

rm -f *.twl

export LANG=C
export LC_ALL=C
export LC_CTYPE=C
export LC_COLLATE=C

mk-list() { ../mk-list -d final "$@"; }

# may be slightly faster
#mk-list() { ../mk-list -f -d final "$@" | (cd final; xargs cat) | sort -u; }

echo base

mk-list english 60 > en.wl0
mk-list en_US     60 > en_US-w_accents.twl
mk-list en_GB-ise 60 > en_GB-ise-w_accents.twl
mk-list en_GB-ize 60 > en_GB-ize-w_accents.twl
mk-list en_CA     60 > en_CA-w_accents.twl
cat en_??-w_accents.twl | sort -u > en-w_accents.twl

mk-list en_US 60 --variants 0 > en-variant_0.wl0
mk-list en_US 60 --variants 0,1 > en-variant_1.wl0
mk-list en_US 60 --variants 0,1,2 > en-variant_2.wl0

mk-list en_GB 60 --variants 0 > en_GB-variant_0.wl0
mk-list en_GB 60 --variants 0,1 > en_GB-variant_1.wl0

mk-list en_CA 60 --variants 0 > en_CA-variant_0.wl0
mk-list en_CA 60 --variants 0,1 > en_CA-variant_1.wl0

echo deaccent

for f in *-w_accents.twl
do
  b=`basename $f -w_accents.twl`
  src/deaccent < $f | sort -u > $b-wo_accents.twl
  #cat $f $b-wo_accents.twl | sort -u > $b.twl
done

for f in en*-variant_*.wl0
do
  b=`basename $f .wl0`
  src/deaccent < $f > $b.wl1
  cat $f $b.wl1 | sort -u > $b.wl
done

echo common

# find the intersection of all *.twl word lists

cat *.twl | sort -u > en-common.wl.0
for f in *.twl
do
  comm -12 en-common.wl.0 $f > en-common.wl
  mv en-common.wl en-common.wl.0
done
mv en-common.wl.0 en-common.wl

echo finish

for f in *.twl
do
  b=`basename $f .twl`
  comm -13 en-common.wl $f > $b-only.wl
done
 
rm -f *.wl.* *.twl *.wl{0,1} *.vwl
cp -p *.wl aspell/
cp -p ../README aspell/doc/SCOWL-README
cat aspell/Copyright.in ../Copyright > aspell/Copyright