test_tabdecomp
631 Bytes
#!/bin/bash
if [ $# -ne 1 ]
then
echo "Usage: `basename $0` tagset-name"
echo
echo "Reads stdin as tab-separated file and feeds each tag representation"
echo "through tagset-tool -p in order to get a decomposed tag representation"
echo "(useful for testing if two tabfiles resolve to the same tags)."
echo "NOTE: to get meaningful results, ensure that each line contains a"
echo "unique (form, lemma) pair. To get this format, use tablines.py."
exit 65
fi
while read line
do
echo "$line" | cut -f 1,2 | tr '\n' '\t'
echo "$line" | cut -f 3 | tagset-tool -pq -t $1 | sort | sed -e :a -e '$!N; s/\n/+/; ta'
done