forked from myuon/AI
-
Notifications
You must be signed in to change notification settings - Fork 0
/
AI10.py
186 lines (149 loc) · 5.04 KB
/
AI10.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
#! /usr/bin/python
#-*- coding:utf-8 -*-
"""
乱択アルゴリズムと評価函数で次の文章を選択する
文末終了判定を追加
"""
import random
import sys
import re
EOS = [u'。',u'!',u'?',u'!',u'?',u'.']
# ファイル関係
def fOpen(_file):
dat_file = ''
if len(sys.argv)==1: dat_file = _file
else: dat_file = sys.argv[1]
f = open(dat_file,'r')
lines = f.readlines()
lines = toUnicode(lines)
return lines, dat_file
# unicode変換
def toUnicode(_lines):
lines = []
for i in range(len(_lines)):
lines.append(_lines[i].decode('utf-8'))
return "".join(lines)
# 形態素リストの作成
def nGramList(_list):
wrdList = []
for i in range(len(_list)-1):
wrdList.append([_list[i],_list[i+1]])
return wrdList
#形態素の切り出し
def morph(_str):
nowMode = 'None'
preMode = 'None'
morphList = []
index = -1
for i in range(len(_str)):
# 長音「ー」は前の形態に引っ張られる
if re.match(u"[ー]", _str[i]) != None:
nowMode = preMode
elif re.match(u"[\u4e00-\u9fa5]", _str[i]) != None:
nowMode = 'ZEN_KANJI'
elif re.match(u"[ぁ-ゞ]", _str[i]) != None:
nowMode = 'ZEN_HIRAGANA'
elif re.match(u"[ァ-ヾ]", _str[i]) != None:
nowMode = 'ZEN_KATAKANA'
elif re.match(u"[ヲ-ン]", _str[i]) != None:
nowMode = 'HAN_KATAKANA'
elif re.match(u"[。、!?―〜()・…]", _str[i]) != None:
nowMode = 'ZEN_KIGOU'
elif re.match(u"[\.,!\?\-\(\)]", _str[i]) != None:
nowMode = 'HAN_KIGOU'
elif re.match(u"[゙゚]", _str[i]) != None:
nowMode = 'OTHER'
if preMode == 'HAN_KATAKANA':
nowMode = preMode
elif re.match(u"[a-zA-Z]", _str[i]) != None:
nowMode = 'ALPHABET'
elif re.match(u"[0-9]", _str[i]) != None:
nowMode = 'HAN_NUMBER'
elif re.match(u"[0-9]", _str[i]) != None:
nowMode = 'ZEN_NUMBER'
else:
nowMode = 'OTHER'
if nowMode == 'OTHER':
preMode = nowMode
continue
if nowMode != preMode:
morphList.append(_str[i])
index += 1
else:
if nowMode == 'OTHER': continue
morphList[index]+=_str[i]
preMode = nowMode
return morphList
# 形態素の解析
def addMorph(_str, dat_file, _wrdList):
m_str = morph(_str)
if m_str == []: return []
if m_str[-1] not in EOS:
m_str.append(u'。')
f = open(dat_file,'a')
for i in m_str:
f.write(i.encode("utf-8")+"\n")
f.close()
_wrdList.append([_wrdList[-1][-1],m_str[0]])
_wrdList += nGramList(m_str)
return m_str
# ルーレットを作る
def makeRouret(_wrdList, iniChar):
rouret = []
iniList = [x[0] for x in _wrdList]
if not iniChar in iniList:
return []
index = iniList.index(iniChar)
while iniChar in iniList[index]:
rouret.append(index)
index += 1
if index >= len(iniList)-1:
break
return rouret
def makeSentence(iniIndex):
msg = ''
cnt = 0
while cnt == 0 or (not wrdList[iniIndex][0][-1] in EOS and cnt <= 20):
msg += wrdList[iniIndex][0]
cnt += 1
iniList = makeRouret(wrdList, wrdList[iniIndex][1])
if not iniList: iniIndex = random.randint(0,len(wrdList)-1)
else: iniIndex = random.choice(iniList)
return msg
def judgeSentence(msg_list, str_mlist, pre_str_mlist):
score = 0
for i in str_mlist:
if i in msg_list:
score += 10
for i in pre_str_mlist:
if i in msg_list:
score += 3
return score
# main
lines,dat_file = fOpen('save/ai10.txt')
wrdList = nGramList(morph(lines))
pre_morphs = []
print "アリーチェ:いらっしゃいませ〜。メッセージをどうぞ〜"
while True:
s = raw_input("あなた:").decode('utf-8')
if s == '': continue
now_morphs = addMorph(s, dat_file, wrdList)
for i in EOS:
while i in now_morphs: now_morphs.remove(i)
for i in EOS:
while i in pre_morphs: now_morphs.remove(i)
iniList = []
for i in range(len(now_morphs)):
iniList += makeRouret(wrdList, now_morphs[i])
if pre_morphs != []:
for i in range(len(pre_morphs)):
iniList += makeRouret(wrdList, pre_morphs[i])
if iniList == []:
iniList = [random.randint(0,len(wrdList)-1)]
msg = []
for i in range(len(iniList)):
_msg = makeSentence(iniList[i])
msg.append([_msg, judgeSentence(_msg, now_morphs, pre_morphs)])
msg = sorted(msg, key=lambda x:x[1], reverse=True)
pre_morphs = now_morphs
print u"アリーチェ:"+msg[int(random.randint(0,len(msg)-1)/4)][0]