#/bin/bash

set -ue

# Download ChipMunk distribution.
CISTERN="cistern.cis.lmu.de/chipmunk/2015-06-26/"
wget ${CISTERN}/marmot-2015-06-26.jar
wget ${CISTERN}/mallet.jar
wget ${CISTERN}/JSAP-2.1.jar

# Download datasets.
wget cistern.cis.lmu.de/chipmunk/supplement.tar.gz
tar xzf supplement.tar.gz

LIBS=marmot-2015-06-26.jar:mallet.jar:JSAP-2.1.jar
lang=eng

# Train model on English data
java -cp $LIBS chipmunk.segmenter.cmd.Train\
   --train-file supplement/seg/${lang}/trn\
   --lang ${lang}\
   --verbose false\
   --crf-mode false\
   --tag-level 0\
   --dictionary-paths "supplement/additional/${lang}/aspell.txt supplement/additional/${lang}/wordlist.txt supplement/additional/${lang}/wiktionary.txt"\
   --model-file ${lang}.chipmunk.srl

# Run segmenter on some English words
echo 'books booked booking bookings rebooked bookstore' | tr ' ' '\n' > input.txt
java -cp $LIBS chipmunk.segmenter.cmd.Segment\
      --model-file ${lang}.chipmunk.srl\
      --input-file input.txt\
      --output-file output.txt

cat output.txt