def get_extractors(cls):
extractors = getattr(cls, "_extractors", None)
if not extractors:
morph_vocab = cls.get_morph_vocab()
extractors = [
DatesExtractor(morph_vocab),
MoneyExtractor(morph_vocab)
]
cls._extractors = extractors
return extractors
def parse_cash(self):
""" Парсер Денежной суммы
:return:
"""
extractor = MoneyExtractor()
matches = extractor(self._text)
return [{
'integer': _.fact.as_json.get('integer'),
'coins': _.fact.as_json.get('coins'),
'fraction': _.fact.as_json.get('fraction'),
'multiplier': _.fact.as_json.get('multiplier'),
'currency': _.fact.as_json.get('currency'),
} for _ in matches]
def get_extractor(extract_type):
if extract_type == "name":
#Экстрактор имён
return NamesExtractor()
elif extract_type == "location":
#Экстрактор мест
return LocationExtractor()
elif extract_type == "date":
#Экстрактор дат
return DatesExtractor()
elif extract_type == "money":
#Экстрактор денежных сумм
return MoneyExtractor()
def natasha_res(text):
names_extractor = NamesExtractor()
location_extractor = LocationExtractor()
organisation_extractor = OrganisationExtractor()
dates_extractor = DatesExtractor()
money_extractor = MoneyExtractor()
names_mapper = lambda x: text[x.span[0]:x.span[1]]
location_mapper = names_mapper
org_mapper = names_mapper
money_mapper = names_mapper
res = {
'names': set(map(names_mapper, names_extractor(text))),
'locations': set(map(location_mapper, location_extractor(text))),
'organisations': set(map(org_mapper, organisation_extractor(text))),
'dates': set(map(dates_mapper, dates_extractor(text))),
'money': set(map(money_mapper, money_extractor(text))),
}
return res
def __init__(self):
from natasha import (NamesExtractor, SimpleNamesExtractor,
DatesExtractor, MoneyExtractor,
MoneyRateExtractor, MoneyRangeExtractor,
LocationExtractor, AddressExtractor,
OrganisationExtractor, PersonExtractor)
addr_ex = AddressExtractor()
date_ex = DatesExtractor()
loc_ex = LocationExtractor()
money_ex = MoneyExtractor()
money_range_ex = MoneyRangeExtractor()
money_rate_ex = MoneyRateExtractor()
name_ex = SimpleNamesExtractor()
org_ex = OrganisationExtractor()
person_ex = PersonExtractor()
# extractors=[addr_ex, date_ex, loc_ex, money_ex, money_range_ex, money_rate_ex,
# name_ex, org_ex, person_ex]
self.extractors = [
addr_ex, date_ex, loc_ex, money_ex, money_range_ex, money_rate_ex,
org_ex, person_ex
]
def getSalary(text):
if text == None:
return {}
salary = searchSalary4(text)
if salary != None:
return salary
extractorRate = MoneyRateExtractor()
matchesRate = extractorRate(
text) #todo check why matchesRate[0] can be None
if len(matchesRate
) != 0 and matchesRate[0] != None and 'fact' in matchesRate[
0] and matchesRate[0].fact != None:
return {
"value": matchesRate[0].fact.money.integer,
"currency": matchesRate[0].fact.money.currency,
"period": matchesRate[0].fact.period
}
salary = searchSalary2(text)
if salary != None:
return salary
salary = searchSalary5(text)
if salary != None:
return salary
salary = searchSalary3(text)
if salary != None:
return salary
extractor = MoneyExtractor()
matches = extractor(text)
if len(matches) != 0:
return {
"value": matches[0].fact.integer,
"currency": matches[0].fact.currency
}
salary = searchSalary1(text)
def send_db_pool_map(doc):
extractors = [AddressExtractor(), DatesExtractor(), LocationExtractor(), MoneyExtractor(), NamesExtractor(),
OrganisationExtractor(), PersonExtractor()]
pool_local = ThreadPool(10)
ne_full = []
span_ne_full = []
type_ne_full = []
try:
for extr in extractors:
global text_paper
text_paper = doc
matches = extr(text_paper)
ne = pool_local.starmap(get_ne2, zip(matches,[doc for x in range(len(matches))]))
span_ne = pool_local.map(get_span_ne, matches)
type_ne = pool_local.map(get_type_ne, matches)
ne_full.append(ne)
span_ne_full.append(span_ne)
type_ne_full.append(type_ne)
except:
print('Ошибка! Примерный номер =', '?')
pool_local.close()
pool_local.join()
if len(ne_full) != 0:
ne_for_db = reduce(lambda x, y: x + y, ne_full)
span_ne_for_db = reduce(lambda x, y: x + y, span_ne_full)
type_ne_for_db = reduce(lambda x, y: x + y, type_ne_full)
'''if len(ne_for_db) != 0:
cur.execute('UPDATE public.news_rbc '
'SET ne=%s, span_ne=%s, type_ne=%s'
'WHERE id=%s;', (ne_for_db, span_ne_for_db,
type_ne_for_db, num))
con.commit()'''
return [ne_for_db, span_ne_for_db, type_ne_for_db]
else:
return [0, 0, 0]
import myextractors
status = 1
res = {}
segmenter = Segmenter()
emb = NewsEmbedding()
morph_tagger = NewsMorphTagger(emb)
syntax_parser = NewsSyntaxParser(emb)
ner_tagger = NewsNERTagger(emb)
morph_vocab = MorphVocab()
names_extractor = NamesExtractor(morph_vocab)
money_extractor = MoneyExtractor(morph_vocab)
text = 'Посол Израиля на Украине Йоэль Лион признался, что пришел в шок, узнав о решении властей Львовской области объявить 2019 год годом лидера запрещенной в России Организации украинских националистов (ОУН) Степана Бандеры...'
docType = 'coast'
doc = Doc(text)
doc.segment(segmenter)
doc.tag_morph(morph_tagger)
doc.parse_syntax(syntax_parser)
doc.tag_ner(ner_tagger)
for span in doc.spans:
span.normalize(morph_vocab)
#для судебного приказа
from natasha import (NamesExtractor, AddressExtractor, DatesExtractor,
MoneyExtractor, LocationExtractor)
app = Flask(__name__)
# todo: LocationExtractor работает плохо, его надо настравиать,
# todo: но он умеет находить города, страны и регионы.
# todo: AddressExtractor лучше находит и представляет города,
# todo: но не находит страны и регионы. Он находит улицы и дома,
# todo: нужно исключить их из выдачи и объединить результаты с
# todo: результатами LocationExtractor-а.
names_extractor = NamesExtractor()
address_extractor = AddressExtractor()
dates_extractor = DatesExtractor()
money_extractor = MoneyExtractor()
location_extractor = LocationExtractor()
def find_named_entities(ner_extractor, text):
"""Находит именованные сущности в тексте.
:param ner_extractor: объект класса NamesExtractor, AddressExtractor,
DatesExtractor или MoneyExtractor
:param text: str
:return: list of namedtuples
"""
matches = ner_extractor(text)
return [_.fact.as_json for _ in matches]
'''if len(ne_for_db) != 0:
cur.execute('UPDATE public.news_rbc '
'SET ne=%s, span_ne=%s, type_ne=%s'
'WHERE id=%s;', (ne_for_db, span_ne_for_db,
type_ne_for_db, num))
con.commit()'''
return [ne_for_db, span_ne_for_db, type_ne_for_db]
else:
return [0, 0, 0]
if __name__ == '__main__':
time_begin = time()
# экстракторы
extractors = [AddressExtractor(), DatesExtractor(), LocationExtractor(), MoneyExtractor(), NamesExtractor(),
OrganisationExtractor(), PersonExtractor()]
send_to_db_news_rbc(extractors) # 292.92 секунды
# ошибки: 6911;7168;7561;8246;8539;8691;9211
exit()
'''con = psycopg2.connect(database="texts_politics", user="******", password="******", host="127.0.0.1",
port="5432", )
cur = con.cursor()
pool = ThreadPool(10) # было ошибок 8 - 2459? 2779 = []
for i in tqdm(range(5059,9540,10)): # 296.92 секунды # с3347по3357 не делал
# обработало 5839 строк, из них 120 строк не обработаных
cur.execute("SELECT full_text FROM public.news_rbc ORDER BY id ASC LIMIT 10 OFFSET %s", (i,))
data = cur.fetchall()
docs = [x[0] for x in data]
#new_form = pool.starmap(send_db_pool_map, zip(docs,[i_num for i_num in range(i,i+10)]))
new_form = pool.map(send_db_pool_map,docs) # 281.43 секунды | 293.59
from ipymarkup import AsciiMarkup, Span, BoxMarkup
import re
import json
from natasha import (NamesExtractor, AddressExtractor, DatesExtractor,
MoneyExtractor, OrganisationExtractor, LocationExtractor)
from natasha.markup import show_markup, show_json
extractors = [
NamesExtractor(),
AddressExtractor(),
DatesExtractor(),
MoneyExtractor(),
OrganisationExtractor(),
LocationExtractor()
]
from flask import Flask
from flask import request
app = Flask(__name__)
@app.route('/getFacts', methods=['POST'])
def getFacts():
print(request.is_json)
content = request.get_json()
text = content['text']
facts = {}
def extractor():
return MoneyExtractor()
def money_extractor(morph_vocab):
return MoneyExtractor(morph_vocab)
'КАС',
re.compile(
r'кодекс[а-я]*\s+(об\s+)?административн[а-я]*\s+правонарушени[а-я]*', re.IGNORECASE | re.MULTILINE):
'КоАП',
}
abbrs = [
(make_regex(r'обществ\w+ с ограниченной ответственностью'), 'ООО'),
(make_regex(r'открыт\w+ акционерн\w+ обществ\w+'), 'ОАО'),
(make_regex(r'закрыт\w+ акционерн\w+ обществ\w+'), 'ЗАО'),
(make_regex(r'публичн\w+ акционерн\w+ обществ\w+'), 'ПАО'),
(make_regex(r'акционерн\w+ обществ\w+'), 'АО'),
(make_regex(r'федеральн\w+ казенн\w+ учрежден\w+'), 'ФКУ'),
]
money = MoneyExtractor()
dates = DatesExtractor()
org = OrganisationExtractor()
CAP_SPACES = re.compile(r'\s((?:[А-Я]\s+){2,}[А-Я][^\w])', re.IGNORECASE)
def fix_cap_spaces(text: str):
return CAP_SPACES.sub(lambda m: ' ' + m.group(1).replace(' ', '') + ' ',
text)
def remove_newlines(text: str):
regex = re.compile(r'([а-яА-Я,"«»()0-9])\s*\n+', re.MULTILINE)
return regex.sub(r'\1 ', text)
def Main(docType, text):
status = 1
res = {}
segmenter = Segmenter()
emb = NewsEmbedding()
morph_tagger = NewsMorphTagger(emb)
syntax_parser = NewsSyntaxParser(emb)
ner_tagger = NewsNERTagger(emb)
morph_vocab = MorphVocab()
names_extractor = NamesExtractor(morph_vocab)
money_extractor = MoneyExtractor(morph_vocab)
doc = Doc(text)
doc.segment(segmenter)
doc.tag_morph(morph_tagger)
doc.parse_syntax(syntax_parser)
doc.tag_ner(ner_tagger)
for span in doc.spans:
span.normalize(morph_vocab)
#для судебного приказа
if docType == 'coast':
#фио
for span in doc.spans:
if span.type == PER:
span.extract_fact(names_extractor)
x = [_.fact.as_dict for _ in doc.spans if _.type == PER]
if x:
res['ФИО'] = x
else:
status = 0
#инн
y = myextractors.findINN(text)
if y:
res['ИНН'] = y
else:
status = 0
#номер судебного приказа
y = myextractors.findNCOASTCASE(text)
if y:
res['номер судебного приказа'] = y
else:
status = 0
#дата с п
y = myextractors.findDATECOAST(text)
if y:
res['дата судебного приказа'] = y
else:
status = 0
#организации
y = []
for span in doc.spans:
if span.type == ORG:
d = {}
d['name'] = span.text
y = y + [d]
if y:
res['организации'] = y
else:
status = 0
#для письма
if docType == 'mail':
#фио
for span in doc.spans:
if span.type == PER:
span.extract_fact(names_extractor)
x = [_.fact.as_dict for _ in doc.spans if _.type == PER]
if x:
res['ФИО'] = x
else:
status = 0
#инн
y = myextractors.findINN(text)
if y:
res['ИНН'] = y
else:
status = 0
#номер дог
y = myextractors.findNCONTRACT(text)
if y:
res['номер договора'] = y
else:
status = 0
#дата дог
y = myextractors.findDATECONT(text)
if y:
res['дата договора'] = y
else:
status = 0
#для платежного поручения
if docType == 'order':
#фио
for span in doc.spans:
if span.type == PER:
span.extract_fact(names_extractor)
x = [_.fact.as_dict for _ in doc.spans if _.type == PER]
if x:
res['ФИО'] = x
else:
status = 0
#инн
y = myextractors.findINN(text)
if y:
res['ИНН'] = y
else:
status = 0
#организации
y = []
for span in doc.spans:
if span.type == ORG:
d = {}
d['name'] = span.text
y = y + [d]
if y:
res['организации'] = y
else:
status = 0
#номер дог
y = myextractors.findNCONTRACT(text)
if y:
res['номер договора'] = y
else:
status = 0
#дата дог
y = myextractors.findNCONTRACT(text)
if y:
res['номер договора'] = y
else:
status = 0
#сумма
matches = list(money_extractor(text))
y = [_.fact for _ in matches]
ret = []
for i in y:
z = {}
z['amount'] = i.amount
z['currency'] = i.currency
ret = ret + [z]
if ret:
res['сумма'] = ret
else:
status = 0
returning = {}
if status == 1:
returning['status'] = 'успех'
else:
returning['status'] = 'не успех'
returning['entities'] = res
return returning
STOP_WORDS += ["не_" + x for x in STOP_WORDS]
STOP_WORDS += ["не"]
# местоимение-существительное, предлог, союз, частица, междометие
STOP_POS = ["NPRO", "PREP", "CONJ", "PRCL", "INTJ"]
CHAR_TABLE = str.maketrans({
key: " "
for key in string.punctuation.replace("<", "").replace(">", "") +
"…？‘«»‘♂️”“’[]'™"
})
NERS = [
(DatesExtractor(), "<date>"),
(TimeExtractor(), "<time>"),
(ExtraDatesExtractor(), "<date>"),
(MoneyExtractor(), "<money>"),
(PhoneExtractor(), "<phone>"),
(PhotoExtractor(), "<photo>"),
(StickerExtractor(), "<sticker>"),
(LinkExtractor(), "<url>"),
(AddressExtractor(), "<address>"),
(NamesExtractor(), "<name>"),
(NumberExtractor(), "<number>"),
(CensorExtractor(), "<censored>"),
]
class EntityExtractor(Component):
"""Распознает именованные сущности"""
def __init__(self, cache=None, verbose=False, **kwargs):
super(EntityExtractor, self).__init__()
