-
Notifications
You must be signed in to change notification settings - Fork 1
/
tokenizer.py
231 lines (195 loc) · 8.38 KB
/
tokenizer.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
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
# -*- coding: utf-8 -*-
import codecs
import re
from numpy import unicode
class Tokenizer():
'''class for tokenizer'''
def __init__(self, text=None):
if text is not None:
self.text = text
self.clean_text()
else:
self.text = None
self.sentences = []
self.tokens = []
self.stemmed_word = []
self.final_list = []
self.final_Sentences=[]
# self.final_tokens=[]
def read_from_file(self, filename,id):
f = codecs.open(filename, encoding='utf-8')
self.text = f.read()
if id==1:
return
self.clean_text()
def generate_sentences(self):
'''generates a list of sentences'''
text = self.text
text = text.replace(u'?', u'।')
self.sentences = text.split(u"।")
def print_sentences(self, sentences=None):
print("COUNT: ", len(self.sentences))
if sentences:
for i in sentences:
print(i)
else:
for i in self.sentences:
print(i)
def clean_text(self):
text = self.text
text = re.sub(r'(\d+)', r'', text)
text = text.replace(u'\ufeff','')
text = text.replace('\n', '')
text = text.replace('\r', '')
text = text.replace(u',', '')
text = text.replace(u'"', '')
text = text.replace(u'(', '')
text = text.replace(u')', '')
text = text.replace(u'"', '')
text = text.replace(u':', '')
text = text.replace(u"'", '')
text = text.replace(u"’", '')
text = text.replace(u"‘", '')
text = text.replace(u"‘‘", '')
text = text.replace(u"’’", '')
text = text.replace(u"''", '')
text = text.replace(u".", '')
self.text = text
def remove_only_space_words(self):
tokens = filter(lambda tok: tok.strip(), self.tokens)
tokens = [tok for tok in self.tokens if tok.strip()]
self.tokens = tokens
def hyphenated_tokens(self):
for i,each in enumerate(self.tokens):
if '-' in each:
tok = each.split('-')
self.tokens.remove(each)
self.tokens.insert(i,tok[0])
self.tokens.insert(i+1,tok[1])
def tokenize(self):
'''done'''
if not self.sentences:
self.generate_sentences()
sentences_list = self.sentences
for each in sentences_list:
tokens = []
word_list = each.split(' ')
tokens = tokens + word_list
self.tokens = tokens
# remove words containing spaces
self.remove_only_space_words()
# remove hyphenated words
self.hyphenated_tokens()
self.generate_stem_dict()
self.remove_stop_words()
self.formSentence()
def print_tokens(self, print_list=None):
'''done'''
if print_list is None:
for i in self.tokens:
print(i)
else:
for i in print_list:
print(i)
def formSentence(self):
finalSentence=""
for word in self.final_tokens:
finalSentence +=word+" "
self.final_Sentences.append(finalSentence)
def print_finalSentence(self, x,fileName):
#fileName = 'complete_corpus\\machine_output\\tokenized' + str(x) + ".txt"
f = open(fileName, "w+", encoding="utf8")
for sentence in self.final_Sentences:
if(len(sentence.strip())>0):
f.write(sentence.strip()+" " + u"\u0964" + " ")
def tokens_count(self):
'''done'''
return len(self.tokens)
def sentence_count(self):
'''done'''
return len(self.sentences)
def len_text(self):
'''done'''
return len(self.text)
def generate_stem_words(self, word):
suffixes = {
1: [u"ो", u"े", u"ू", u"ु", u"ी", u"ि", u"ा"],
2: [u"कर", u"ाओ", u"िए", u"ाई", u"ाए", u"ने", u"नी", u"ना", u"ते", u"ीं", u"ती", u"ता", u"ाँ", u"ां", u"ों",
u"ें"],
3: [u"ाकर", u"ाइए", u"ाईं", u"ाया", u"ेगी", u"ेगा", u"ोगी", u"ोगे", u"ाने", u"ाना", u"ाते", u"ाती", u"ाता",
u"तीं", u"ाओं", u"ाएं", u"ुओं", u"ुएं", u"ुआं"],
4: [u"ाएगी", u"ाएगा", u"ाओगी", u"ाओगे", u"एंगी", u"ेंगी", u"एंगे", u"ेंगे", u"ूंगी", u"ूंगा", u"ातीं",
u"नाओं", u"नाएं", u"ताओं", u"ताएं", u"ियाँ", u"ियों", u"ियां"],
5: [u"ाएंगी", u"ाएंगे", u"ाऊंगी", u"ाऊंगा", u"ाइयाँ", u"ाइयों", u"ाइयां"],
}
for L in 5, 4, 3, 2, 1:
if len(word) > L + 1:
for suf in suffixes[L]:
# print type(suf),type(word),word,suf
if word.endswith(suf):
# print 'h'
return word[:-L]
return word
def generate_stem_dict(self):
'''returns a dictionary of stem words for each token'''
# suffixes = {
# 1: ["ो", "े", "ू", "ु", "ी", "ि", "ा"],
# 2: ["कर", "ाओ", "िए", "ाई", "ाए", "ने", "नी", "ना", "ते", "ीं", "ती", "ता", "ाँ", "ां", "ों", "ें"],
# 3: ["ाकर", "ाइए", "ाईं", "ाया", "ेगी", "ेगा", "ोगी", "ोगे", "ाने", "ाना", "ाते", "ाती", "ाता", "तीं", "ाओं", "ाएं", "ुओं", "ुएं", "ुआं"],
# 4: ["ाएगी", "ाएगा", "ाओगी", "ाओगे", "एंगी", "ेंगी", "एंगे", "ेंगे", "ूंगी", "ूंगा", "ातीं", "नाओं", "नाएं", "ताओं", "ताएं", "ियाँ", "ियों", "ियां"],
# 5: ["ाएंगी", "ाएंगे", "ाऊंगी", "ाऊंगा", "ाइयाँ", "ाइयों", "ाइयां"],
# }
stem_word = {}
self.stemmed_word = []
# if not self.tokens:
# self.tokenize()
for each_token in self.tokens:
# print type(each_token)
temp = self.generate_stem_words(each_token)
# print temp
stem_word[each_token] = temp
self.stemmed_word.append(temp)
return stem_word
def remove_stop_words(self):
f = codecs.open("stopwords.txt", encoding='utf-8')
self.final_tokens=[]
# if not self.stemmed_word:
# self.generate_stem_dict()
stopwords = [x.strip() for x in f.readlines()]
for j, eachStopWord in enumerate(stopwords):
stopwords[j] = stopwords[j].replace(u'\ufeff','')
# stopwords = []
tokens = [i for i in self.stemmed_word if unicode(i) not in stopwords]
self.final_tokens = tokens
return tokens
def tokenize_testFile(fileName1):
res = []
sent = []
o = Tokenizer()
o.read_from_file(fileName1,1)
o.generate_sentences()
sent = o.sentences
o.tokenize()
for sentence in o.final_Sentences:
if (len(sentence.strip()) > 0):
data = sentence.strip() + " " + u"\u0964" + " "
res.append(data)
return sent, res
if __name__ == "__main__":
no_of_inputs = 20 #No. of training files
for x in range(1, no_of_inputs+1):
t = Tokenizer()
filePath = 'complete_corpus\\input\\'
fileName = filePath+"input"+str(x)+".txt"
t.read_from_file(fileName,0)
t.generate_sentences()
t.tokenize()
# t.print_tokens(t.final_tokens)
t.print_finalSentence(x,'complete_corpus\\machine_output\\tokenized' + str(x) + ".txt")
tSum = Tokenizer()
filePath1 = 'complete_corpus\\human_output\\'
fileName1 = filePath1 + "article" + str(x) + "_reference1.txt"
tSum.read_from_file(fileName1,0)
tSum.generate_sentences()
tSum.tokenize()
tSum.print_finalSentence(x,'complete_corpus\\machine_output\\tokenizedSummary' + str(x) + ".txt")