# - *- coding: utf- 8 - *-

'''

@author : 조수현, 김윤희

@date : 2017. 9. 26

@version : 2.0.0

@brief : 기능1

- 파라미터 - scd:파일명, slist:문장, pos:명사목록, path:jvm.dll경로

keyword:키워드,tfidf목록, id_m:키워드,채번 목록, iid_m:채번,키워드 목록

index:키워드가 속한 문서들 목록

'''

from konlpy.tag import Twitter

from konlpy.utils import pprint

class SentenceManager() :

_count = 0

def run(self, pos) :

path = "C:/Program Files/Java/jdk1.7.0_55/jre/bin/server/jvm.dll"

slist = []

try :

self.scd_parser('DOC1.SCD', slist) # scd에서 문장 추출, #id, date는 별도 저장

self.pos_tagger(slist[0:10], pos, path) # 문장 형태소 분석하여 명사 추출

return 1, "ok"

except Exception as ex :

return 0, ex

def scd_parser(self, scd, slist) :

sentence = open(scd,'r')

try:

for i in sentence :

if i.startswith('<Contents>') :

slist.append(i.replace('<Contents>',''))

return 1, "ok"

except Exception as ex :

return 0, ex

def pos_tagger(self, slist, pos, path) :

twitter = Twitter(path)

try:

# slist의 기사가 1개인경우 type이 str임으로 별도처리

if type(slist) == str:

pos.append(twitter.nouns(slist))

return 1, "ok"

for one in slist :

pos.append(twitter.nouns(one))

return 1, "ok"

except Exception as ex :

return 0, ex

def id_manager(self, keyword, id_m, iid_m) :

try:

for i, l in enumerate(keyword):

for j in l:

key = j[0]

if key in id_m:

continue

id_m[key] = self._count

iid_m[self._count] = key

self._count += 1

return 1, "ok"

except Exception as ex :

return 0, ex

def indexer(self, keyword, index) :

try:

for i, l in enumerate(keyword) :

for j in l :

key = j[0]

if key in index:

index[key] += [i]

continue

index[key] = [i]

return 1, "ok"

except Exception as ex :

return 0, ex