This repository has been archived by the owner on Aug 26, 2024. It is now read-only.
-
Notifications
You must be signed in to change notification settings - Fork 44
/
nlp_utils.py
210 lines (180 loc) · 9.03 KB
/
nlp_utils.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
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
# -*- coding: utf-8 -*-
import regex as re
uniChars = "àáảãạâầấẩẫậăằắẳẵặèéẻẽẹêềếểễệđìíỉĩịòóỏõọôồốổỗộơờớởỡợùúủũụưừứửữựỳýỷỹỵÀÁẢÃẠÂẦẤẨẪẬĂẰẮẲẴẶÈÉẺẼẸÊỀẾỂỄỆĐÌÍỈĨỊÒÓỎÕỌÔỒỐỔỖỘƠỜỚỞỠỢÙÚỦŨỤƯỪỨỬỮỰỲÝỶỸỴÂĂĐÔƠƯ"
unsignChars = "aaaaaaaaaaaaaaaaaeeeeeeeeeeediiiiiooooooooooooooooouuuuuuuuuuuyyyyyAAAAAAAAAAAAAAAAAEEEEEEEEEEEDIIIOOOOOOOOOOOOOOOOOOOUUUUUUUUUUUYYYYYAADOOU"
def loaddicchar():
dic = {}
char1252 = 'à|á|ả|ã|ạ|ầ|ấ|ẩ|ẫ|ậ|ằ|ắ|ẳ|ẵ|ặ|è|é|ẻ|ẽ|ẹ|ề|ế|ể|ễ|ệ|ì|í|ỉ|ĩ|ị|ò|ó|ỏ|õ|ọ|ồ|ố|ổ|ỗ|ộ|ờ|ớ|ở|ỡ|ợ|ù|ú|ủ|ũ|ụ|ừ|ứ|ử|ữ|ự|ỳ|ý|ỷ|ỹ|ỵ|À|Á|Ả|Ã|Ạ|Ầ|Ấ|Ẩ|Ẫ|Ậ|Ằ|Ắ|Ẳ|Ẵ|Ặ|È|É|Ẻ|Ẽ|Ẹ|Ề|Ế|Ể|Ễ|Ệ|Ì|Í|Ỉ|Ĩ|Ị|Ò|Ó|Ỏ|Õ|Ọ|Ồ|Ố|Ổ|Ỗ|Ộ|Ờ|Ớ|Ở|Ỡ|Ợ|Ù|Ú|Ủ|Ũ|Ụ|Ừ|Ứ|Ử|Ữ|Ự|Ỳ|Ý|Ỷ|Ỹ|Ỵ'.split(
'|')
charutf8 = "à|á|ả|ã|ạ|ầ|ấ|ẩ|ẫ|ậ|ằ|ắ|ẳ|ẵ|ặ|è|é|ẻ|ẽ|ẹ|ề|ế|ể|ễ|ệ|ì|í|ỉ|ĩ|ị|ò|ó|ỏ|õ|ọ|ồ|ố|ổ|ỗ|ộ|ờ|ớ|ở|ỡ|ợ|ù|ú|ủ|ũ|ụ|ừ|ứ|ử|ữ|ự|ỳ|ý|ỷ|ỹ|ỵ|À|Á|Ả|Ã|Ạ|Ầ|Ấ|Ẩ|Ẫ|Ậ|Ằ|Ắ|Ẳ|Ẵ|Ặ|È|É|Ẻ|Ẽ|Ẹ|Ề|Ế|Ể|Ễ|Ệ|Ì|Í|Ỉ|Ĩ|Ị|Ò|Ó|Ỏ|Õ|Ọ|Ồ|Ố|Ổ|Ỗ|Ộ|Ờ|Ớ|Ở|Ỡ|Ợ|Ù|Ú|Ủ|Ũ|Ụ|Ừ|Ứ|Ử|Ữ|Ự|Ỳ|Ý|Ỷ|Ỹ|Ỵ".split(
'|')
for i in range(len(char1252)):
dic[char1252[i]] = charutf8[i]
return dic
dicchar = loaddicchar()
def convert_window1252_to_utf8(txt):
return re.sub(
r'à|á|ả|ã|ạ|ầ|ấ|ẩ|ẫ|ậ|ằ|ắ|ẳ|ẵ|ặ|è|é|ẻ|ẽ|ẹ|ề|ế|ể|ễ|ệ|ì|í|ỉ|ĩ|ị|ò|ó|ỏ|õ|ọ|ồ|ố|ổ|ỗ|ộ|ờ|ớ|ở|ỡ|ợ|ù|ú|ủ|ũ|ụ|ừ|ứ|ử|ữ|ự|ỳ|ý|ỷ|ỹ|ỵ|À|Á|Ả|Ã|Ạ|Ầ|Ấ|Ẩ|Ẫ|Ậ|Ằ|Ắ|Ẳ|Ẵ|Ặ|È|É|Ẻ|Ẽ|Ẹ|Ề|Ế|Ể|Ễ|Ệ|Ì|Í|Ỉ|Ĩ|Ị|Ò|Ó|Ỏ|Õ|Ọ|Ồ|Ố|Ổ|Ỗ|Ộ|Ờ|Ớ|Ở|Ỡ|Ợ|Ù|Ú|Ủ|Ũ|Ụ|Ừ|Ứ|Ử|Ữ|Ự|Ỳ|Ý|Ỷ|Ỹ|Ỵ',
lambda x: dicchar[x.group()], txt)
"""
Start section: Chuyển câu văn về kiểu gõ telex khi không bật Unikey
Ví dụ: thủy = thuyr, tượng = tuwowngj
"""
bang_nguyen_am = [['a', 'à', 'á', 'ả', 'ã', 'ạ', 'a'],
['ă', 'ằ', 'ắ', 'ẳ', 'ẵ', 'ặ', 'aw'],
['â', 'ầ', 'ấ', 'ẩ', 'ẫ', 'ậ', 'aa'],
['e', 'è', 'é', 'ẻ', 'ẽ', 'ẹ', 'e'],
['ê', 'ề', 'ế', 'ể', 'ễ', 'ệ', 'ee'],
['i', 'ì', 'í', 'ỉ', 'ĩ', 'ị', 'i'],
['o', 'ò', 'ó', 'ỏ', 'õ', 'ọ', 'o'],
['ô', 'ồ', 'ố', 'ổ', 'ỗ', 'ộ', 'oo'],
['ơ', 'ờ', 'ớ', 'ở', 'ỡ', 'ợ', 'ow'],
['u', 'ù', 'ú', 'ủ', 'ũ', 'ụ', 'u'],
['ư', 'ừ', 'ứ', 'ử', 'ữ', 'ự', 'uw'],
['y', 'ỳ', 'ý', 'ỷ', 'ỹ', 'ỵ', 'y']]
bang_ky_tu_dau = ['', 'f', 's', 'r', 'x', 'j']
nguyen_am_to_ids = {}
for i in range(len(bang_nguyen_am)):
for j in range(len(bang_nguyen_am[i]) - 1):
nguyen_am_to_ids[bang_nguyen_am[i][j]] = (i, j)
def vn_word_to_telex_type(word):
dau_cau = 0
new_word = ''
for char in word:
x, y = nguyen_am_to_ids.get(char, (-1, -1))
if x == -1:
new_word += char
continue
if y != 0:
dau_cau = y
new_word += bang_nguyen_am[x][-1]
new_word += bang_ky_tu_dau[dau_cau]
return new_word
def vn_sentence_to_telex_type(sentence):
"""
Chuyển câu tiếng việt có dấu về kiểu gõ telex.
:param sentence:
:return:
"""
words = sentence.split()
for index, word in enumerate(words):
words[index] = vn_word_to_telex_type(word)
return ' '.join(words)
"""
End section: Chuyển câu văn về kiểu gõ telex khi không bật Unikey
"""
"""
Start section: Chuyển câu văn về cách gõ dấu kiểu cũ: dùng òa úy thay oà uý
Xem tại đây: https://vi.wikipedia.org/wiki/Quy_t%E1%BA%AFc_%C4%91%E1%BA%B7t_d%E1%BA%A5u_thanh_trong_ch%E1%BB%AF_qu%E1%BB%91c_ng%E1%BB%AF
"""
def chuan_hoa_dau_tu_tieng_viet(word):
if not is_valid_vietnam_word(word):
return word
chars = list(word)
dau_cau = 0
nguyen_am_index = []
qu_or_gi = False
for index, char in enumerate(chars):
x, y = nguyen_am_to_ids.get(char, (-1, -1))
if x == -1:
continue
elif x == 9: # check qu
if index != 0 and chars[index - 1] == 'q':
chars[index] = 'u'
qu_or_gi = True
elif x == 5: # check gi
if index != 0 and chars[index - 1] == 'g':
chars[index] = 'i'
qu_or_gi = True
if y != 0:
dau_cau = y
chars[index] = bang_nguyen_am[x][0]
if not qu_or_gi or index != 1:
nguyen_am_index.append(index)
if len(nguyen_am_index) < 2:
if qu_or_gi:
if len(chars) == 2:
x, y = nguyen_am_to_ids.get(chars[1])
chars[1] = bang_nguyen_am[x][dau_cau]
else:
x, y = nguyen_am_to_ids.get(chars[2], (-1, -1))
if x != -1:
chars[2] = bang_nguyen_am[x][dau_cau]
else:
chars[1] = bang_nguyen_am[5][dau_cau] if chars[1] == 'i' else bang_nguyen_am[9][dau_cau]
return ''.join(chars)
return word
for index in nguyen_am_index:
x, y = nguyen_am_to_ids[chars[index]]
if x == 4 or x == 8: # ê, ơ
chars[index] = bang_nguyen_am[x][dau_cau]
# for index2 in nguyen_am_index:
# if index2 != index:
# x, y = nguyen_am_to_ids[chars[index]]
# chars[index2] = bang_nguyen_am[x][0]
return ''.join(chars)
if len(nguyen_am_index) == 2:
if nguyen_am_index[-1] == len(chars) - 1:
x, y = nguyen_am_to_ids[chars[nguyen_am_index[0]]]
chars[nguyen_am_index[0]] = bang_nguyen_am[x][dau_cau]
# x, y = nguyen_am_to_ids[chars[nguyen_am_index[1]]]
# chars[nguyen_am_index[1]] = bang_nguyen_am[x][0]
else:
# x, y = nguyen_am_to_ids[chars[nguyen_am_index[0]]]
# chars[nguyen_am_index[0]] = bang_nguyen_am[x][0]
x, y = nguyen_am_to_ids[chars[nguyen_am_index[1]]]
chars[nguyen_am_index[1]] = bang_nguyen_am[x][dau_cau]
else:
# x, y = nguyen_am_to_ids[chars[nguyen_am_index[0]]]
# chars[nguyen_am_index[0]] = bang_nguyen_am[x][0]
x, y = nguyen_am_to_ids[chars[nguyen_am_index[1]]]
chars[nguyen_am_index[1]] = bang_nguyen_am[x][dau_cau]
# x, y = nguyen_am_to_ids[chars[nguyen_am_index[2]]]
# chars[nguyen_am_index[2]] = bang_nguyen_am[x][0]
return ''.join(chars)
def is_valid_vietnam_word(word):
chars = list(word)
nguyen_am_index = -1
for index, char in enumerate(chars):
x, y = nguyen_am_to_ids.get(char, (-1, -1))
if x != -1:
if nguyen_am_index == -1:
nguyen_am_index = index
else:
if index - nguyen_am_index != 1:
return False
nguyen_am_index = index
return True
def chuan_hoa_dau_cau_tieng_viet(sentence):
"""
Chuyển câu tiếng việt về chuẩn gõ dấu kiểu cũ.
:param sentence:
:return:
"""
sentence = sentence.lower()
words = sentence.split()
for index, word in enumerate(words):
cw = re.sub(r'(^\p{P}*)([p{L}.]*\p{L}+)(\p{P}*$)', r'\1/\2/\3', word).split('/')
print(cw)
cw[1] = chuan_hoa_dau_tu_tieng_viet(cw[1])
words[index] = ''.join(cw)
return ' '.join(words)
"""
End section: Chuyển câu văn về cách gõ dấu kiểu cũ: dùng òa úy thay oà uý
Xem tại đây: https://vi.wikipedia.org/wiki/Quy_t%E1%BA%AFc_%C4%91%E1%BA%B7t_d%E1%BA%A5u_thanh_trong_ch%E1%BB%AF_qu%E1%BB%91c_ng%E1%BB%AF
"""
if __name__ == '__main__':
print(chuan_hoa_dau_cau_tieng_viet('anh hoà, đang làm.. gì'))
# f = open('/home/lap60313/data/corpus-full.txt', encoding='utf8')
# sentence = f.readline()
# current_line = 0
# while sentence:
# current_line += 1
# if current_line % 1000 == 0:
# print('Current line', str(current_line))
# sentence = sentence.lower().strip()
# sentence = convertwindown1525toutf8(sentence)
# sentence = chuan_hoa_dau_cau_tieng_viet(sentence)
# with open('/home/lap60313/data/corpus-full.txt.out', 'a+', encoding='utf8') as fp:
# fp.write(sentence + "\n")
# sentence = f.readline()