def name_tokenize(sentence):
api = kh.KhaiiiApi()
api.open()
x = re.sub('[-\.\$\+>#\}\{\*<&@%;\\\)\(="?!\[\]~\^/:,_\|]+', '.',
str(sentence).replace("'", "")).replace('.', ' ') # 특수 기호 제거
res = []
try:
for word in api.analyze(x):
if re.match('\d+', word.lex) != None:
tmp = ''
for m in word.morphs: # '2000년대','2000년대의'에서 '2000년대'만 뽑아내는 코드
if m.tag in ['NNG', 'NNP', 'NNB', 'SN']:
tmp += m.lex
res.append(tmp)
else:
a = len(res)
for i in genre:
if i in word.lex:
res.append(i) # '발라드', '메탈'처럼 장르명이 제목에 있으면 바로 리스트에 추가
for j in artist:
if j in word.lex and len(
j) > 1: # 두 글자 이상의 아티스트 이름이 단어에 들어있으면 리스트에 추가
res.append(j)
b = len(res)
if a == b:
for m in word.morphs:
if m.tag in ['NNG', 'NNP', 'NNB', 'SN',
'SL']: # 명사와 숫자, 외국어만 받습니다
res.append(m.lex)
except:
pass
return ' '.join(res)
def preprocessing(data):
try:
from konlpy.tag import Okt, Kkma
import khaiii
khaiii_api = khaiii.KhaiiiApi(opt.khaiii_so_path)
khaiii_api.open(opt.khaiii_path)
kkma = Kkma()
kkma_tokenizer = kkma.nouns
twitter = Okt()
okt_tokenizer = twitter.nouns
cls, data_path_list, div, out_path, begin_offset, end_offset = data
data = cls()
data.load_y_vocab()
data.preprocessing(data_path_list, div, begin_offset, end_offset,
out_path, okt_tokenizer, khaiii_api, kkma_tokenizer)
except Exception:
raise Exception("".join(traceback.format_exception(*sys.exc_info())))
def kakao_postagger_nn_finder(summay_text):
api = khaiii.KhaiiiApi()
api.open()
nn_word_list = []
for word in api.analyze(summay_text):
morphs_str = ' + '.join([(m.lex + '/' + m.tag) for m in word.morphs])
# print(f'{word.lex}\t{morphs_str}')
morphs_str_list = morphs_str.split(" + ")
complex_morphs = ""
for mophs_item in morphs_str_list:
if mophs_item.split("/")[1].startswith("N") or mophs_item.split("/")[1].startswith("MM") or \
mophs_item.split("/")[1].startswith("SN") or mophs_item.split("/")[1].startswith("SL"):
complex_morphs = complex_morphs + mophs_item.split("/")[0]
if len(complex_morphs) > 1:
# print("->", complex_morphs)
nn_word_list.append(complex_morphs)
return nn_word_list
def sent2khaiii(data, tag=False):
'''
카이 형태소 분석기
data: 데이터 (str or list(Series))
tag: 태그 포함 여부 (False or True)
예시
khaii(["우리집에 왜 왔니", "왜 왔니"])
>>> [['우리', '집', '에', '왜', '오', '았', '니'], ['왜', '오', '았', '니']]
khaii(["우리집에 왜 왔니", "왜 왔니"], tag=True)
>>> [['우리/NP', '집/NNG', '에/JKB', '왜/MAG', '오/VV', '았/EP', '니/EC'],
['왜/MAG', '오/VV', '았/EP', '니/EC']]
'''
import khaiii
api = khaiii.KhaiiiApi()
api.open()
if type(data) == str: data = [data]
return [[
a.lex + "/" + a.tag if tag == True else a.lex
for word in api.analyze(str(sent)) for a in word.morphs
] if str(sent).strip() else (sent) for sent in data]
def setUp(self):
self._api = khaiii.KhaiiiApi()
self._api.set_log_level('all', 'warn')
self._api.open()
import http.client, urllib.parse
from flask import (Flask, request, abort, render_template, Response, jsonify)
from flask_cors import CORS
import nltk
from wiktionaryparser import WiktionaryParser
from tagmap import TagMap
from chunker import Chunker
# ---------- initialize KHaiii phoneme analyzer
# set up KHaiii api
import khaiii
khaiiiAPI = khaiii.KhaiiiApi()
khaiiiAPI.open()
# ---------- instantiate Flask (global) app --------
parserApp = Flask('app',
static_folder="./dist/static",
template_folder="./dist")
CORS(parserApp)
#
parserApp.config.update(DEBUG=True,
SECRET_KEY="iu877hy3jnd8**yd98y334%$#Rjxhg6222",
SESSION_COOKIE_HTTPONLY=False)
def run_dev_server():
def __init__(self):
self.tokenizer = khaiii.KhaiiiApi()
super().__init__(self.tokenizer.analyze)
import khaiii
api = khaiii.KhaiiiApi()
api.open()
class Khaiii():
def pos(self, phrase, flatten=True, join=False):
"""POS tagger.
:param flatten: If False, preserves eojeols.
:param join: If True, returns joined sets of morph and tag.
"""
sentences = phrase.split('\n')
morphemes = []
if not sentences:
return morphemes
for sentence in sentences:
for word in api.analyze(sentence):
result = [(m.lex, m.tag) for m in word.morphs]
if join:
result = [
'{}/{}'.format(m.lex, m.tag) for m in word.morphs
]
morphemes.append(result)
if flatten:
return sum(morphemes, [])
