# Sample commands with explanations of what is being produced.
#
# (1) Default condition is to produce BOTH (BLEU & NIST) scores 
#     with detail level = 0, we include the "-c" flag to score by case
#
./mteval-v10.pl -r ref/sample-ref-translations.sgm -s src/sample-source-data.sgm -t tst/sample-tst-translations.sgm -c >score.1.txt


#
# (2) Produce only the NIST scores using case scoring
#
./mteval-v10.pl -r ref/sample-ref-translations.sgm -s src/sample-source-data.sgm -t tst/sample-tst-translations.sgm -n -c >score.2.txt


#
# (3) Produce only the BLEU scores using case scoring
./mteval-v10.pl -r ref/sample-ref-translations.sgm -s src/sample-source-data.sgm -t tst/sample-tst-translations.sgm -b -c >score.3.txt


#
# (4) Produce only the NIST scores with case scoring, include per document scoring
./mteval-v10.pl -r ref/sample-ref-translations.sgm -s src/sample-source-data.sgm -t tst/sample-tst-translations.sgm -n -c -d 1 >score.4.txt


#
# (5) Produce only the NIST scores with case scoring, include per document scoring and per segment scoring
./mteval-v10.pl -r ref/sample-ref-translations.sgm -s src/sample-source-data.sgm -t tst/sample-tst-translations.sgm -n -c -d 2 >score.5.txt

