def thai_len_2(self, values, eng_fn_len):
first_words = utils.map_to_words(values, i=0)
title, sim = self.choose_title(first_words, 'tha')
if sim > TITLE_MIN_SIM:
# Title + firstname
firstname, ratio = utils.most_frequent(utils.map_to_words(values,
i=1),
with_ratio=True)
return ((title, sim), (firstname, ratio), EMPTY_FIELD)
else:
# Look for second words, in case of bad croping
second_words = utils.map_to_words(values, i=1)
title, sim = self.choose_title(second_words, 'tha')
if sim > TITLE_MIN_SIM:
return ((title, sim), EMPTY_FIELD, EMPTY_FIELD)
# Look if title included in first word
title, values_no_title = self.find_included_title(values, 'tha')
values = values if title == '' else values_no_title
value, ratio = utils.most_frequent_by_word(values)
if eng_fn_len > 1:
return ((title, TITLE_MIN_SIM), (value, ratio), EMPTY_FIELD)
else:
fn, ln = value.split()
return ((title, TITLE_MIN_SIM), (fn, ratio), (ln, ratio))
def thai_len_1(self, values):
title, sim = self.choose_title(values, 'tha')
if sim > TITLE_MIN_SIM:
return ((title, sim), EMPTY_FIELD, EMPTY_FIELD)
name, ratio = utils.most_frequent(values, with_ratio=True)
return (EMPTY_FIELD, (name, ratio), EMPTY_FIELD)
def eng_fn_len_1(self, values):
title, sim = self.choose_title(values, 'eng')
if sim > TITLE_MIN_SIM:
return ((title, sim), EMPTY_FIELD)
else:
name, ratio = utils.most_frequent(values, with_ratio=True)
name = ' '.join(
list(filter(lambda w: len(w) > 2 or w == 'Na', name.split())))
return (EMPTY_FIELD, (name, ratio))
def eng_extract_lastname(self, eng_fn_img):
values, length = self.eng_get_clean_correct_values(eng_fn_img)
if length == 0:
return EMPTY_FIELD
elif length == 1:
name, ratio = utils.most_frequent(values, with_ratio=True)
else:
name, ratio = utils.most_frequent_by_word(values)
name = ' '.join(
list(filter(lambda w: len(w) > 2 or w == 'Na', name.split())))
return name, ratio
def eng_fn_len_2_more(self, values, length):
first_words = utils.map_to_words(values, i=0)
title, sim = self.choose_title(first_words, 'eng')
values = utils.map_to_words(values, start=1)
if length == 2:
name, ratio = utils.most_frequent(values, with_ratio=True)
else:
name, ratio = utils.most_frequent_by_word(values)
name = ' '.join(
list(filter(lambda w: len(w) > 2 or w == 'Na', name.split())))
return ((title, sim), (name, ratio))