2019-02-04 15:27:56 +01:00
|
|
|
#!/usr/bin/python3
|
|
|
|
# -*- coding: utf-8 -*-
|
|
|
|
|
|
|
|
import sys
|
|
|
|
|
|
|
|
max_tokens = 100
|
|
|
|
max_ratio = 4.0
|
|
|
|
|
2019-02-21 14:02:51 +01:00
|
|
|
#./prepare_corpus.py corpora/$(CORPUS_NAME)/src.txt corpora/$(CORPUS_NAME)/trg.txt corpora/$(CORPUS_NAME)/ids.txt corpora/$(CORPUS_NAME)/src.lem corpora/$(CORPUS_NAME)/trg.lem corpora/$(CORPUS_NAME)/src.dict corpora/$(CORPUS_NAME)/trg.dict corpora/$(CORPUS_NAME)/src_clean.txt corpora/$(CORPUS_NAME)/src_clean.lem corpora/$(CORPUS_NAME)/trg_clean.txt corpora/$(CORPUS_NAME)/ids_clean.txt corpora/$(CORPUS_NAME)/falign_corpus.txt
|
2019-02-04 15:27:56 +01:00
|
|
|
|
2019-02-06 15:34:08 +01:00
|
|
|
|
2019-02-21 14:02:51 +01:00
|
|
|
|
|
|
|
|
|
|
|
with open(sys.argv[1]) as src, open(sys.argv[2]) as trg, open(sys.argv[3]) as ids, open(sys.argv[4]) as src_lem, open(sys.argv[5]) as trg_lem, open(sys.argv[6]) as src_dict, open(sys.argv[7]) as trg_dict, open(sys.argv[8], 'w') as src_clean, open(sys.argv[9], 'w') as src_clean_lem, open(sys.argv[10], 'w') as trg_clean, open(sys.argv[11], 'w') as ids_clean, open(sys.argv[12], 'w') as falign_corpus:
|
2019-02-06 15:34:08 +01:00
|
|
|
for line in src:
|
2019-02-04 15:27:56 +01:00
|
|
|
src_line_orig = line.strip()
|
2019-02-06 15:34:08 +01:00
|
|
|
trg_line_orig = trg.readline().strip()
|
|
|
|
id_orig = ids.readline().strip()
|
|
|
|
src_line_lem = src_lem.readline().strip()
|
|
|
|
trg_line_lem = trg_lem.readline().strip()
|
|
|
|
src_token_count = len(src_line_lem.split())
|
|
|
|
trg_token_count = len(trg_line_lem.split())
|
2019-02-04 15:27:56 +01:00
|
|
|
if (src_token_count > 0 and trg_token_count > 0 and src_token_count <= max_tokens and trg_token_count <= max_tokens):
|
|
|
|
ratio = float(src_token_count/trg_token_count) if src_token_count > trg_token_count else float(trg_token_count/src_token_count)
|
|
|
|
if (ratio <= max_ratio):
|
|
|
|
src_clean.write(src_line_orig+"\n")
|
2019-02-21 14:02:51 +01:00
|
|
|
src_clean_lem.write(src_line_lem+"\n")
|
2019-02-04 15:27:56 +01:00
|
|
|
trg_clean.write(trg_line_orig+"\n")
|
2019-02-06 15:34:08 +01:00
|
|
|
ids_clean.write(id_orig+"\n")
|
2019-02-21 14:02:51 +01:00
|
|
|
falign_corpus.write("%s ||| %s\n" % (src_line_lem, trg_line_lem))
|
|
|
|
for line in src_dict:
|
|
|
|
src_word = line.rstrip()
|
|
|
|
trg_word = trg_dict.readline().rstrip()
|
|
|
|
falign_corpus.write("%s ||| %s\n" % (src_word, trg_word))
|