Commit a419e606 authored by Neeme Kahusk's avatar Neeme Kahusk
Browse files

register ud

parent 69b8f006
# Execute scripts needed to register corpora files for cwb
# check for existence of file
if [ "$#" -ne 1 ] || ! [ -f "$1" ]; then
echo "Usage: $0 vrt file" >&2
exit 1
FN=`echo $FILENAME|sed 's/[.]vrt//g'`
FNUPPER=`echo $FN|tr '[:lower:]' '[:upper:]'`
# CWBVERSION="cwb-3.4.11"
# 0. Clean old data
rm -rf /corpora/data/$FN
rm -rf /corpora/registry/$FN
mkdir /corpora/data/$FN
# 1. Convert vrt to cwb format
#/usr/local/$CWBVERSION/bin/cwb-encode -s -p - -d /corpora/data/$FN -R /corpora/registry/$FN -c utf8 -f /corpora/$FILENAME -P word -P baseform -P analysis -P pos -P number -P cases -P voice -P tense -P nominal -P mood -P person -P negation -S sentence:0+n -S para:0+n -S text:0+author+date
/usr/local/$CWBVERSION/bin/cwb-encode -s -p - -d /corpora/data/$FN -R /corpora/registry/$FN -c utf8 -f /corpora/$FILENAME -P word -P baseform -P pos -P msd -P ref -P dephead -P deprel -S sentence:0+id -S text:0
# 2. Register corpus
/usr/local/$CWBVERSION/bin/cwb-makeall -r /corpora/registry -V $FNUPPER
# 3. Insert info about corpus
echo -e "Sentences: $(cat $FILENAME|grep '<sentence'|wc -l)\nUpdated: $(date -I)\n" > data/$FN/.info
# # TEST
# echo $FILENAME
# echo $FN
# echo $FNUPPER
Supports Markdown
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment