def test_stopwords(self):
stopword = [u'ＡＶ', u'女優']
print(u'Stopwords Filtering Test. Stopwords is {}'.format(
u','.join(stopword)))
test_sentence = u"紗倉 まな（さくら まな、1993年3月23日 - ）は、日本のAV女優。"
kytea_wrapper = KyteaWrapper()
token_objects = kytea_wrapper.tokenize(sentence=test_sentence,
return_list=False,
is_feature=True)
filtered_result = kytea_wrapper.filter(parsed_sentence=token_objects,
stopwords=stopword)
check_flag = True
print('-' * 30)
for stem_posTuple in filtered_result.convert_list_object():
assert isinstance(stem_posTuple, tuple)
word_stem = stem_posTuple[0]
word_posTuple = stem_posTuple[1]
assert isinstance(word_stem, unicode)
assert isinstance(word_posTuple, tuple)
print(u'word_stem:{} word_pos:{}'.format(word_stem,
' '.join(word_posTuple)))
if word_stem in stopword: check_flag = False
assert check_flag
def test_filter_pos(self):
"""
"""
# 'Filtering Test. POS condition is only 名詞')
test_sentence = "紗倉 まな（さくら まな、1993年3月23日 - ）は、日本のAV女優。"
kytea_wrapper = KyteaWrapper()
tokenized_result = kytea_wrapper.tokenize(sentence=test_sentence,
normalize=True,
return_list=False,
is_feature=True)
pos_condition = [('名詞', )]
filtered_result = kytea_wrapper.filter(
parsed_sentence=tokenized_result, pos_condition=pos_condition)
assert isinstance(filtered_result, FilteredObject)
for t_obj in filtered_result.tokenized_objects:
assert isinstance(t_obj, TokenizedResult)
assert isinstance(t_obj.word_surface, str)
assert isinstance(t_obj.word_stem, str)
assert isinstance(t_obj.tuple_pos, tuple)
assert isinstance(t_obj.misc_info, dict)
assert t_obj.tuple_pos[0] == '名詞'
for stem_posTuple in filtered_result.convert_list_object():
assert isinstance(stem_posTuple, tuple)
word_stem = stem_posTuple[0]
word_posTuple = stem_posTuple[1]
assert isinstance(word_stem, str)
assert isinstance(word_posTuple, tuple)
def test_filter_pos(self):
"""
"""
print(u'Filtering Test. POS condition is only 名詞')
test_sentence = u"紗倉 まな（さくら まな、1993年3月23日 - ）は、日本のAV女優。"
kytea_wrapper = KyteaWrapper()
tokenized_result = kytea_wrapper.tokenize(sentence=test_sentence,
normalize=True,
return_list=False,
is_feature=True)
pos_condition = [(u'名詞', )]
filtered_result = kytea_wrapper.filter(
parsed_sentence=tokenized_result, pos_condition=pos_condition)
assert isinstance(filtered_result, FilteredObject)
for t_obj in filtered_result.tokenized_objects:
assert isinstance(t_obj, TokenizedResult)
print(u"word_surafce:{}, word_stem:{}, pos_tuple:{}, misc_info:{}".
format(t_obj.word_surface, t_obj.word_stem,
' '.join(t_obj.tuple_pos), t_obj.misc_info))
assert isinstance(t_obj.word_surface, unicode)
assert isinstance(t_obj.word_stem, unicode)
assert isinstance(t_obj.tuple_pos, tuple)
assert isinstance(t_obj.misc_info, dict)
assert t_obj.tuple_pos[0] == u'名詞'
print('-' * 30)
for stem_posTuple in filtered_result.convert_list_object():
assert isinstance(stem_posTuple, tuple)
word_stem = stem_posTuple[0]
word_posTuple = stem_posTuple[1]
assert isinstance(word_stem, unicode)
assert isinstance(word_posTuple, tuple)
print(u'word_stem:{} word_pos:{}'.format(word_stem,
' '.join(word_posTuple)))