-
Notifications
You must be signed in to change notification settings - Fork 0
/
parser_module.py
339 lines (313 loc) · 14.4 KB
/
parser_module.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
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
import re
from nltk.corpus import stopwords
from math import log
from document import Document_to_index
import urllib3
class Parse:
def __init__(self):
self.counter=0
self.stop_words = frozenset(stopwords.words('english'))
self.added_stop_words=["I","The","rT","RT","rt","http","https",'t.co'," ","","twitter.com","-","www","_","&","##","###","####","#####","19"]
self.check=0
self.words = open(r'zif.txt').read().split()
self.dictionay_word_cost = dict((k, log((i + 1) * log(len(self.words)))) for i, k in enumerate(self.words))
self.max_word_length = max(len(x) for x in self.words)
self.dic={}
self.hashtag_arr=[]
self.usa_abbreviations = {"AL": "Alabama", "AK": "Alaska", "AZ": "Arizona", "AR": "Arkansas",
"CA": "California","NYC":"new york city","nyc":"new york city",
"CT": "Connecticut", "DC": "Washington DC", "DE": "Deleware", "FL": "Florida",
"GA": "Georgia", "HI": "Hawaii", "IL": "Illinios", "IA": "Iowa",
"KS": "Kansas", "KY": "Kentucky", "MD": "Maryland", "MA": "Massachusetts",
"MI": "Michigan", "MN": "Minnesota", "MS": "Mississippi", "MO": "Missouri",
"MT": "Montana", "NE": "Nebraska", "NV": "Nevada", "NH": "New Hampshire",
"NJ": "New Jersey", "NM": "New Mexico", "NY": "New York", "NC": "North Carolina",
"Covid": "cornavirus", "COVID": "cornavirus", "covid19": "cornavirus",
"Covid19": "cornavirus", "Covid-19": "cornavirus", "covid-19": "cornavirus",
"CORONAVIRUS": "cornavirus", "Coronavirus": "cornavirus", "Corona": "cornavirus",
"Corona": "cornavirus", "corona": "cornavirus", "CORONA": "coronavirus",
"COVID19": "coronavirus", "COVID-19": "cornavirus",
"ND": "North Dakota", "OH": "Ohio", "PA": "Pennsylvania", "RI": "Rhode Island",
"SC": "South Carolina", "SD": "South Dakota", "TN": "Tennessee","MSM":"methylsulfonylmethane",
"TX": "Texas", "UT": "Utah", "VT": "Vermont", "VA": "Virgina", "WA": "Washington",
"WV": "West Virginia", "WI": "Wisconsin", "WY": "Wyoming", "USA": "United States",
"IG": "Instagram", "FB": "Facebook",'HCQ':'hydroxychloroquine','hcq':'hydroxychloroquine'}
def parse_sentence(self, text):
"""
This function tokenize, remove stop words and apply lower case for every word within the text
:param text:
:return:
"""
txt1=self.multiple_replace(text)
terms=txt1.split()
index=0
while(index<len(terms)):
usa_abbrev=self.usa_abbreviations.get(terms[index].upper(),"Never")
if(terms[index] in self.stop_words or terms[index] in self.added_stop_words):
index=index+1
continue
elif (terms[index][0] == '@' and len(terms[index])>1): # @rule
index = index + 1
continue
elif (terms[index][0] == '#'): # #rule
lst=self.parse_hashtag(terms[index])
for i in lst:
self.enter_dic(i.lower())
if(i not in self.stop_words and i not in self.hashtag_arr):
self.hashtag_arr.append(i)
index=index+1
elif (usa_abbrev!="Never"):
self.enter_dic(usa_abbrev)
index=index+1
elif (terms[index].isdigit()) or (re.search(r"(?:\\d+\\s+)?\\d/\\d", terms[index])) or (re.search(r'^-?[0-9]+\/[0-9]+$', terms[index])):
if (index == len(terms)-1): ##mikre katze doc24512 (# in last token)#case 0 the number is the last index
w=self.parse_numbers(terms[index],'')
self.enter_dic(w)
index=index+1
else:
word_check=terms[index+1].lower()
if(word_check=='%' or word_check=='percent' or word_check=='percentage' or word_check=='million' or word_check=='thousand' or word_check=='billion'):
w =self.parse_numbers(terms[index],terms[index+1])
self.enter_dic(w)
index=index+2
else:
w = self.parse_numbers(terms[index],'')
self.enter_dic(w)
index = index + 1
elif (len(terms[index])<=2):
index+=1
else:
self.enter_dic(terms[index])
index = index + 1
def parse_doc(self, doc_as_list):
"""
This function takes a tweet document as list and break it into different fields
:param doc_as_list: list re-presenting the tweet.
:return: Document object with corresponding fields.
"""
self.dic={}
self.hashtag_arr=[]
tweet_id = doc_as_list[0]
if(tweet_id=="1280939236150165516"):
tmp=3
tweet_date = doc_as_list[1]
full_text = doc_as_list[2]
pattern = re.compile("(?P<url>https?://[^\s]+)", re.S)
full_text_without_url = re.sub(pattern, "", full_text)
all_uppercase_text = bool(re.match(r'[a-z\s]+$', full_text_without_url)) ##if all text is uppercase lower it.
if (all_uppercase_text == False):
full_text_without_url.lower()
self.parse_sentence(full_text_without_url)
if (len(self.dic) > 0):
max_tf = max(self.dic.values())
unique_words = len(self.dic.keys())
else:
max_tf = 0
unique_words = 0
url = doc_as_list[3]
retweet_text = doc_as_list[4]
retweet_url = doc_as_list[5]
quote_text = doc_as_list[6]
quote_url = doc_as_list[7]
doc_length =len(full_text)
document_to_index= Document_to_index(tweet_id,self.dic,max_tf,unique_words,doc_length,self.hashtag_arr)
return document_to_index
def parse_numbers(self,index, next_index):
value = ['', 'K', 'M', 'B']
if bool(re.search(r'^-?[0-9]+\/[0-9]+$', next_index)) and float(index) < 1000:
return index + ' ' + next_index
if bool(re.search(r'^-?[0-9]+\/[0-9]+$', index)):
return index
elif(next_index.lower()=='%' or next_index.lower()=='percent' or next_index.lower()=='percentage'):
return index +"%"
elif (next_index.lower() =="billion") and float(index) < 1000:
return index + "B"
elif (next_index.lower() == "million" and float(index) < 1000):
return index + "M"
elif (next_index.lower() == "thousand") and float(index) < 1000:
return index + "K"
number = float(index)
size = 0
while abs(number) >= 1000:
size = size+1
number /= 1000.0
if size >= 3:
break
return str("%.3f" % number).strip("0").strip(".") + '' + str(value[size])
def parse_hashtag(self, string):
exp = []
exp.append(string)
if (len(string) == 1): ##mikre katze doc24512 (# in last token)
exp.append('number') # adding number
return exp
word = string[1:]
usa_abbrev = self.usa_abbreviations.get(word.upper(), "Never")
list_tmp = []
if (word.isalpha()) and (word.isupper() or word.islower()):
if (usa_abbrev != "Never"):
list_tmp.append(usa_abbrev)
list_tmp.append(word)
else:
w = word.lower()
list_tmp = self.conclude_space(w)
else:
list_tmp = self.parse_hastag_easy_case(word)
for i in list_tmp:
if (i != ''):
exp.append(i)
return exp
def parse_hastag_easy_case(self, text):
return [i for i in re.split(r"_|([A-Z][a-z]+)", text) if i]
def conclude_space(self, string):
def best_match(index):
candidates = enumerate(reversed(cost_arr[max(0, index - self.max_word_length):index]))
return min((c + self.dictionay_word_cost.get(string[index - k - 1:index], 999999999), k + 1) for k, c in
candidates)
# Build the cost array.
cost_arr = [0]
for index in range(1, len(string) + 1):
c, k = best_match(index)
cost_arr.append(c)
res = []
length = len(string)
while (length > 0):
c, k = best_match(length)
res.append(string[length - k:length])
length = length - k
word = " ".join(reversed(res))
return word.split()
def multiple_replace(self, text):
def cap(match):
return match.group().lower()
first_strip = re.sub("[’'`']", '', text)
p = re.compile(r'((?<=[\.\?!:\W]\s)(\w+)|(^\w+))') ##change uppercase after new sentence
second_strip = p.sub(cap, first_strip)
third_strip = re.sub(r"(?<!\d)\.|\.(?!\d)", '', second_strip)
forth_strip = re.sub('[^A-Za-z0-9@#_$%]+', ' ', third_strip)
return forth_strip
def enter_dic(self,term):
if(term.lower() not in self.stop_words or term.lower() not in self.added_stop_words):
if(term.lower() not in self.dic.keys()):
self.dic[term.lower()]=1
else:
self.dic[term.lower()]+=1
def parse_url(self, url):
"""
This function tokenize a given url
:param text:
"""
parsed_url = urllib3.util.parse_url(url)
arr = []
arr.append(parsed_url.scheme) # https
h = parsed_url.host
if (h is not None):
host = parsed_url.host
if (host[0] == 'w'):
h = urllib3.util.split_first(host, '.')
for i in h:
if (i != '.'):
# if(i not in self.stop_words):
arr.append(i)
else:
arr.append(host)
p = parsed_url.path
if (p is not None):
path = parsed_url.path.split('/') # should be the path(after the .com and until the last /
for i in path:
if (i != ''):
if ("-" in i):
tmp = i.split('-')
for j in tmp:
arr.append(j)
else:
arr.append(i)
q = parsed_url.query
if (q is not None):
query = parsed_url.query.split('=')
for i in query:
arr.append(i)
return arr
def parse_query(self, text):
"""
This function tokenize, remove stop words and apply lower case for every word within the text
:param text:
:return:
"""
dic={}
txt1 = self.multiple_replace(text)
terms = txt1.split()
index = 0
while (index < len(terms)):
# w=terms[index].lower()
usa_abbrev=self.usa_abbreviations.get(terms[index].upper(),"Never")
if (terms[index] in self.stop_words):
index = index + 1
continue
elif (terms[index][0] == '@' and len(terms[index]) > 1): # @rule
if (terms[index] not in dic.keys()):
dic[terms[index]]= 1
else:
dic[terms[index]] += 1
index = index + 1
continue
elif (terms[index][0] == '#'): # #rule
lst = self.parse_hashtag(terms[index])
for i in lst:
if (i not in dic.keys()):
dic[terms[index]] = 1
else:
dic[terms[index]] += 1
index = index + 1
continue
index = index + 1
elif (((usa_abbrev != "Never") )):
if(usa_abbrev not in dic.keys()):
dic[usa_abbrev]=1
else:
dic[usa_abbrev]+=1
index = index + 1
elif (terms[index].isdigit()) or (re.search(r"(?:\\d+\\s+)?\\d/\\d", terms[index])) or (
re.search(r'^-?[0-9]+\/[0-9]+$', terms[index])):
if (index == len(terms) - 1): ##mikre katze doc24512 (# in last token)#case 0 the number is the last index
w = self.parse_numbers(terms[index], '')
if (w not in dic.keys()):
dic[w] = 1
else:
dic[w] += 1
index = index + 1
else:
word_check = terms[index + 1].lower()
if (word_check == '%' or word_check == 'percent' or word_check == 'percentage' or word_check == 'million' or word_check == 'thousand' or word_check == 'billion'):
w = self.parse_numbers(terms[index], terms[index + 1])
if (w not in dic.keys()):
dic[w] = 1
else:
dic[w] += 1
index = index + 2
else:
w = self.parse_numbers(terms[index], '')
if (w not in dic.keys()):
dic[w] = 1
else:
dic[w] += 1
index = index + 1
else:
if (terms[index]not in dic.keys()):
dic[terms[index]] = 1
else:
dic[terms[index]] += 1
index = index + 1
return dic
def test(self,word):
if(word.lower() in self.spell_arr):
return word
if(word.lower()=="governm"):
return "government"
else:
tmp=self.spellcheck.correction(word.lower())
if(word.lower()!=tmp):
print("****************")
print("the original word was:"+word)
print("spell checker returned the word:" + tmp)
print("****************")
return word