import itertools
from preparation.resources.Resource import gen_resource
from preparation import modifiers
from hb_res.explanations.Explanation import Explanation
sample_modifiers = (
modifiers.str_replace('?', 'ё'),
modifiers.shadow_cognates(length_threshold=3, sep_re='(\\s|!|\\.)+'),
modifiers.re_replace('\\.', ' '),
modifiers.normalize_title(),
modifiers.calculate_key(),
)
@gen_resource('SampleResource', sample_modifiers)
def sample_parser():
raw_expls = itertools.starmap(Explanation, (
('ПОРА', 'Однажды.в.студ?ную.зимнюю.пору.я.из.лесу.вышел.'),
('ИВАН', 'Один день Ивана Денисовича'),
('унылая', 'Унылая пора! очей очарованье!')
))
return raw_expls
return ret
return apply
definitions_mods = [
# there is even a board on trello for almost all of these modifiers: trello.com/b/IEP8jusD
modifiers.strip(' -3', target_field='title'),
modifiers.translate(
'?і3', # the second symbol is not i but \xd1\x96 in utf8
'ЁЁЗ',
'",.+:124567890',
target_field='title'
),
modifiers.str_replace('||', 'П', target_field='title'),
modifiers.re_search_ban(r'[^-ЁёА-Яа-я]', target_field='title'),
modifiers.normalize_title(),
# Text OCR problems
# modifiers.translate(
# '?~[]{}',
# 'ё-()()',
# '|*o' # the o is latin
# ),
# modifiers.re_replace(r'знай\.', 'знач.'),
# modifiers.str_replace('3а', 'За'),
# modifiers.re_replace(r'(?<={alph})\d+(-\d+)?'.format(alph=modifiers.ALPH_RE), ''),
modifiers.str_replace(r' :', ':'),
modifiers.re_replace(r'([,:])(?=[^ 0-9])', r'\1 '),
# Text quality heuristics
__author__ = 'Алексей'
# noinspection PyProtectedMember
from preparation.resources.synonyms import _raw_data
from preparation import modifiers
from preparation.resources.Resource import gen_resource
from hb_res.explanations import Explanation
from ._synonyms_quality import choose_best_synonyms
synonyms_mods = [
modifiers.normalize_title(0.01, True),
modifiers.re_replace('[^#]+? [^#]+?(#|$)', ''), # remove multi-word synonyms (containing spaces)
modifiers.re_fullmatch_ban(''),
modifiers.delete_cognates(4, '#'),
modifiers.choose_normal_words_in_explanation('#'),
choose_best_synonyms(5, '#'),
modifiers.calculate_prior_frequency_rate('#'),
modifiers.str_replace('#', ', '),
modifiers.calculate_key()
]
@gen_resource('SynonymsResource', synonyms_mods)
def read_data():
explanations = set()
with open(_raw_data, 'r', encoding='utf-8') as source:
for line in source:
