2017-04-28 13:49:30 +02:00
|
|
|
SRC_LANG=pl
|
|
|
|
|
TRG_LANG=en
|
2017-03-11 21:48:25 +01:00
|
|
|
CORPUS_NAME=europarl_sample
|
2017-04-28 13:49:30 +02:00
|
|
|
SEPARATOR=@\#@
|
2017-01-21 17:01:15 +01:00
|
|
|
|
2017-04-28 13:49:30 +02:00
|
|
|
all: corpora/$(CORPUS_NAME)/aligned.txt corpora/$(CORPUS_NAME)/src_clean.txt corpora/$(CORPUS_NAME)/trg_clean.txt
|
2017-03-11 21:48:25 +01:00
|
|
|
|
2017-04-28 13:49:30 +02:00
|
|
|
corpora/$(CORPUS_NAME)/aligned.txt: corpora/$(CORPUS_NAME)/giza.cfg corpora/$(CORPUS_NAME)/src.lem_trg.lem.cooc corpora/$(CORPUS_NAME)/src.lem_trg.lem.snt corpora/$(CORPUS_NAME)/src.lem.vcb corpora/$(CORPUS_NAME)/trg.lem.vcb
|
2017-01-21 17:01:15 +01:00
|
|
|
mgiza/mgizapp/bin/mgiza corpora/$(CORPUS_NAME)/giza.cfg
|
2017-03-10 14:52:01 +01:00
|
|
|
cat corpora/$(CORPUS_NAME)/aligned*part* | ./sortGizaAlignments.py > corpora/$(CORPUS_NAME)/aligned.txt
|
|
|
|
|
|
|
|
|
|
clean-intermediate-files:
|
|
|
|
|
rm -f corpora/$(CORPUS_NAME)/*.lem
|
|
|
|
|
rm -f corpora/$(CORPUS_NAME)/*.classes
|
|
|
|
|
rm -f corpora/$(CORPUS_NAME)/*.classes.cats
|
|
|
|
|
rm -f corpora/$(CORPUS_NAME)/*.vcb
|
|
|
|
|
rm -f corpora/$(CORPUS_NAME)/*.snt
|
|
|
|
|
rm -f corpora/$(CORPUS_NAME)/*.cooc
|
|
|
|
|
rm -f corpora/$(CORPUS_NAME)/aligned*part*
|
|
|
|
|
rm -f corpora/$(CORPUS_NAME)/giza.cfg
|
2017-04-28 13:49:30 +02:00
|
|
|
rm -f corpora/$(CORPUS_NAME)/pasted.txt
|
|
|
|
|
rm -f corpora/$(CORPUS_NAME)/pasted_deduplicated.txt
|
|
|
|
|
rm -f corpora/$(CORPUS_NAME)/src_deduplicated.txt
|
|
|
|
|
rm -f corpora/$(CORPUS_NAME)/trg_deduplicated.txt
|
|
|
|
|
rm -f corpora/$(CORPUS_NAME)/src_deduplicated.tok
|
|
|
|
|
rm -f corpora/$(CORPUS_NAME)/trg_deduplicated.tok
|
|
|
|
|
rm -f corpora/$(CORPUS_NAME)/src_clean.tok
|
|
|
|
|
rm -f corpora/$(CORPUS_NAME)/trg_clean.tok
|
2017-03-10 14:52:01 +01:00
|
|
|
|
2017-04-28 13:49:30 +02:00
|
|
|
clean: clean-intermediate-files
|
|
|
|
|
rm -f corpora/$(CORPUS_NAME)/src_clean.txt
|
|
|
|
|
rm -f corpora/$(CORPUS_NAME)/trg_clean.txt
|
2017-01-21 17:01:15 +01:00
|
|
|
rm -f corpora/$(CORPUS_NAME)/aligned*
|
|
|
|
|
|
|
|
|
|
corpora/$(CORPUS_NAME)/giza.cfg: giza.cfg.pattern
|
|
|
|
|
sed 's/CORPUS_NAME/'$(CORPUS_NAME)'/' < $< > $@
|
|
|
|
|
|
2017-04-28 13:49:30 +02:00
|
|
|
corpora/$(CORPUS_NAME)/src.lem_trg.lem.cooc: corpora/$(CORPUS_NAME)/src.lem.vcb corpora/$(CORPUS_NAME)/trg.lem.vcb corpora/$(CORPUS_NAME)/src.lem_trg.lem.snt
|
|
|
|
|
mgiza/mgizapp/bin/snt2cooc $@ corpora/$(CORPUS_NAME)/src.lem.vcb corpora/$(CORPUS_NAME)/trg.lem.vcb corpora/$(CORPUS_NAME)/src.lem_trg.lem.snt
|
2017-01-21 17:01:15 +01:00
|
|
|
|
2017-04-28 13:49:30 +02:00
|
|
|
corpora/$(CORPUS_NAME)/src.lem_trg.lem.snt corpora/$(CORPUS_NAME)/trg.lem_src.lem.snt corpora/$(CORPUS_NAME)/src.lem.vcb corpora/$(CORPUS_NAME)/trg.lem.vcb: corpora/$(CORPUS_NAME)/src.lem corpora/$(CORPUS_NAME)/trg.lem
|
|
|
|
|
mgiza/mgizapp/bin/plain2snt corpora/$(CORPUS_NAME)/src.lem corpora/$(CORPUS_NAME)/trg.lem
|
2017-01-21 17:01:15 +01:00
|
|
|
|
2017-04-28 13:49:30 +02:00
|
|
|
corpora/$(CORPUS_NAME)/%.classes: corpora/$(CORPUS_NAME)/%.lem
|
2017-01-21 17:01:15 +01:00
|
|
|
mgiza/mgizapp/bin/mkcls -n10 -p$< -V$@
|
|
|
|
|
|
2017-04-28 13:49:30 +02:00
|
|
|
corpora/$(CORPUS_NAME)/trg.lem: corpora/$(CORPUS_NAME)/trg_clean.tok
|
2017-03-07 10:24:08 +01:00
|
|
|
mono LemmaGenSentenceLemmatizer/LemmaGenSentenceLemmatizer/bin/Debug/LemmaGenSentenceLemmatizer.exe $(TRG_LANG) < $< > $@
|
|
|
|
|
|
|
|
|
|
|
2017-04-28 13:49:30 +02:00
|
|
|
corpora/$(CORPUS_NAME)/src.lem: corpora/$(CORPUS_NAME)/src_clean.tok
|
2017-03-07 10:24:08 +01:00
|
|
|
mono LemmaGenSentenceLemmatizer/LemmaGenSentenceLemmatizer/bin/Debug/LemmaGenSentenceLemmatizer.exe $(SRC_LANG) < $< > $@
|
|
|
|
|
|
2017-04-28 13:49:30 +02:00
|
|
|
|
|
|
|
|
corpora/$(CORPUS_NAME)/src_clean.txt corpora/$(CORPUS_NAME)/trg_clean.txt corpora/$(CORPUS_NAME)/src_clean.tok corpora/$(CORPUS_NAME)/trg_clean.tok: corpora/$(CORPUS_NAME)/pasted_deduplicated.txt corpora/$(CORPUS_NAME)/src_deduplicated.tok corpora/$(CORPUS_NAME)/trg_deduplicated.tok
|
|
|
|
|
./clean_corpus.py $< corpora/$(CORPUS_NAME)/src_deduplicated.tok corpora/$(CORPUS_NAME)/trg_deduplicated.tok corpora/$(CORPUS_NAME)/src_clean.txt corpora/$(CORPUS_NAME)/trg_clean.txt corpora/$(CORPUS_NAME)/src_clean.tok corpora/$(CORPUS_NAME)/trg_clean.tok $(SEPARATOR)
|
|
|
|
|
|
|
|
|
|
corpora/$(CORPUS_NAME)/%_deduplicated.tok: corpora/$(CORPUS_NAME)/%_deduplicated.txt
|
|
|
|
|
concordia-sentence-tokenizer -c ../concordia.cfg < $< > $@
|
|
|
|
|
|
|
|
|
|
corpora/$(CORPUS_NAME)/src_deduplicated.txt corpora/$(CORPUS_NAME)/trg_deduplicated.txt: corpora/$(CORPUS_NAME)/pasted_deduplicated.txt
|
|
|
|
|
./cut.py $< corpora/$(CORPUS_NAME)/src_deduplicated.txt corpora/$(CORPUS_NAME)/trg_deduplicated.txt $(SEPARATOR)
|
|
|
|
|
|
|
|
|
|
corpora/$(CORPUS_NAME)/pasted_deduplicated.txt: corpora/$(CORPUS_NAME)/pasted.txt
|
|
|
|
|
sort -k 1.13 $< | uniq -s 12 | sort > $@
|
|
|
|
|
|
|
|
|
|
corpora/$(CORPUS_NAME)/pasted.txt: corpora/$(CORPUS_NAME)/src.txt corpora/$(CORPUS_NAME)/trg.txt
|
|
|
|
|
./paste.py corpora/$(CORPUS_NAME)/src.txt corpora/$(CORPUS_NAME)/trg.txt $(SEPARATOR)> $@
|
