-
Notifications
You must be signed in to change notification settings - Fork 1
/
Tokenizer.py
239 lines (196 loc) · 11.6 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
229
230
231
232
233
234
235
236
237
238
239
import abc, logging
import re
from abc import ABC
from nltk.tokenize import word_tokenize, sent_tokenize
from preprocessing.FileRepresentation import FileRepresentation, TextFileRepresentation, TextFileGroupedRepresentation, \
UseCaseFileRepresentation
from utility import FileUtil
log = logging.getLogger(__name__)
class Tokenizer(abc.ABC):
def __init__(self, dataset):
self._dataset = dataset
@abc.abstractmethod
def tokenize(self, file_path) -> FileRepresentation:
pass
class NaturalSpeechTokenizer(Tokenizer, ABC):
def __init__(self, dataset, italian=False):
super(NaturalSpeechTokenizer, self).__init__(dataset)
self._italian = italian
def tokenize_to_string_list(self, text: str):
if self._italian:
tokenized_text = "' ".join(text.split("'")) # The italian tokenizer does not split apostrophes
return word_tokenize(tokenized_text, "italian")
return word_tokenize(text)
class WordTokenizer(NaturalSpeechTokenizer):
"""
Tokenizer result: TextFileRepresentation
> A flat list, each element is a string representing a single word
"""
def tokenize(self, file_path):
text_as_string = FileUtil.read_textfile_into_string(file_path, self._dataset.encoding())
if self._italian:
tokenized_text = "' ".join(text_as_string.split("'"))
return TextFileRepresentation(word_tokenize(tokenized_text), file_path)
else:
return TextFileRepresentation(word_tokenize(text_as_string), file_path)
class SentenceTokenizer(NaturalSpeechTokenizer):
"""
Tokenizer result: TextFileRepresentation
> A flat list, each element is a string representing a single sentence
"""
def tokenize(self, file_path) -> FileRepresentation:
text_as_string = FileUtil.read_textfile_into_string(file_path, self._dataset.encoding())
if self._italian:
return TextFileRepresentation(sent_tokenize(text_as_string, language="italian"), file_path)
return TextFileRepresentation(sent_tokenize(text_as_string), file_path)
def tokenize_to_string_list(self, text: str):
if self._italian:
return sent_tokenize(text, language="italian")
return sent_tokenize(text)
def tokenize_all_sentences_in_directory(self, directory) -> [str]:
sentences = []
for file in FileUtil.get_files_in_directory(directory):
if self._italian:
sentences += sent_tokenize(FileUtil.read_textfile_into_string(file, self._dataset.encoding()), language="italian")
else:
sentences += sent_tokenize(FileUtil.read_textfile_into_string(file, self._dataset.encoding()))
return sentences
class WordAndSentenceTokenizer(NaturalSpeechTokenizer):
"""
Tokenizer result: TextFileGroupedRepresentation
> A nested list, each element is another list which contains the words of a single sentence.
"""
def tokenize(self, file_path) -> FileRepresentation:
text_as_string = FileUtil.read_textfile_into_string(file_path, self._dataset.encoding())
word_tokenized_sentences = None
if self._italian:
tokens = [word_tokenize("' ".join(sent.split("'")), language="italian") for sent in sent_tokenize(text_as_string, language="italian")]
return TextFileGroupedRepresentation(tokens, file_path)
else:
word_tokenized_sentences = [word_tokenize(sent) for sent in sent_tokenize(text_as_string)]
return TextFileGroupedRepresentation(word_tokenized_sentences, file_path)
def tokenize_to_string_list(self, text: str):
if self._italian:
return [word_tokenize("' ".join(sent.split("'")), language="italian") for sent in sent_tokenize(text, language="italian")]
return [word_tokenize(sent) for sent in sent_tokenize(text)]
class UCTokenizer(NaturalSpeechTokenizer):
def __init__(self, dataset=None, italian=False):
super(UCTokenizer, self).__init__(dataset, italian)
def tokenize(self, file_path):
text_lines = FileUtil.read_textfile_into_lines_list(file_path, self._dataset.encoding())
uc_name_words = []
uc_actor_words = []
uc_precond_words = []
uc_postcond_words = []
uc_description_words = []
uc_quality_req_words = []
uc_flow_of_events_words = []
last_word_category = uc_description_words # Default
in_flow_of_events = False
for line in text_lines:
line = line.lstrip() # Remove leading white spaces/tabs
if self._dataset.UC_NAME_TEMPLATE_REGEX.match(line):
matched_string = self._dataset.UC_NAME_TEMPLATE_REGEX.match(line).group(0)
uc_name_words += self.tokenize_to_string_list(line[len(matched_string):])
last_word_category = uc_name_words
in_flow_of_events = False
elif self._dataset.UC_DESCRIPTION_TEMPLATE_REGEX.match(line):
matched_string = self._dataset.UC_DESCRIPTION_TEMPLATE_REGEX.match(line).group(0)
uc_description_words += self.tokenize_to_string_list(line[len(matched_string):])
last_word_category = uc_description_words
in_flow_of_events = False
elif self._dataset.UC_ACTOR_TEMPLATE_REGEX.match(line):
matched_string = self._dataset.UC_ACTOR_TEMPLATE_REGEX.match(line).group(0)
uc_actor_words += self.tokenize_to_string_list(line[len(matched_string):])
last_word_category = uc_actor_words
in_flow_of_events = False
elif self._dataset.UC_PRECONDITION_TEMPLATE_REGEX.match(line):
matched_string = self._dataset.UC_PRECONDITION_TEMPLATE_REGEX.match(line).group(0)
uc_precond_words += self.tokenize_to_string_list(line[len(matched_string):])
last_word_category = uc_precond_words
in_flow_of_events = False
elif self._dataset.UC_POSTCONDITION_TEMPLATE_REGEX.match(line):
matched_string = self._dataset.UC_POSTCONDITION_TEMPLATE_REGEX.match(line).group(0)
uc_postcond_words += self.tokenize_to_string_list(line[len(matched_string):])
last_word_category = uc_postcond_words
in_flow_of_events = False
elif self._dataset.UC_FLOW_OF_EVENTS_TEMPLATE_REGEX.match(line):
matched_string = self._dataset.UC_FLOW_OF_EVENTS_TEMPLATE_REGEX.match(line).group(0)
uc_flow_of_events_words += self.tokenize_to_string_list(line[len(matched_string):])
last_word_category = uc_flow_of_events_words
in_flow_of_events = True
elif self._dataset.UC_QUALI_REQ_TEMPLATE_REGEX.match(line):
matched_string = self._dataset.UC_QUALI_REQ_TEMPLATE_REGEX.match(line).group(0)
uc_quality_req_words += self.tokenize_to_string_list(line[len(matched_string):])
last_word_category = uc_quality_req_words
in_flow_of_events = False
elif self._dataset.UC_USER_TEMPLATE_REGEX.match(line) and in_flow_of_events:
# part of flow of events
matched_string = self._dataset.UC_USER_TEMPLATE_REGEX.match(line).group(0)
uc_flow_of_events_words += self.tokenize_to_string_list(line[len(matched_string):])
last_word_category = uc_flow_of_events_words
elif self._dataset.UC_SYSTEM_TEMPLATE_REGEX.match(line) and in_flow_of_events:
# part of flow of events
matched_string = self._dataset.UC_SYSTEM_TEMPLATE_REGEX.match(line).group(0)
uc_flow_of_events_words += self.tokenize_to_string_list(line[len(matched_string):])
last_word_category = uc_flow_of_events_words
else:
last_word_category += self.tokenize_to_string_list(line)
complete_uc_flow_of_events_words_string = " ".join(uc_flow_of_events_words)
if self._italian:
uc_flow_of_events_words = [word_tokenize("' ".join(sent.split("'")), language="italian") for sent in sent_tokenize(complete_uc_flow_of_events_words_string, language="italian")]
else:
uc_flow_of_events_words = [word_tokenize(sent) for sent in sent_tokenize(complete_uc_flow_of_events_words_string)]
return UseCaseFileRepresentation(file_path, uc_name_words, uc_description_words, uc_actor_words, uc_precond_words, uc_postcond_words,
uc_flow_of_events_words, uc_quality_req_words)
class NameAndDescriptionTokenizer(NaturalSpeechTokenizer):
def __init__(self, dataset=None, italian=False):
super(NameAndDescriptionTokenizer, self).__init__(dataset, italian)
def tokenize(self, file_path):
text_lines = FileUtil.read_textfile_into_lines_list(file_path, self._dataset.encoding())
uc_name_words = []
uc_actor_words = []
uc_precond_words = []
uc_postcond_words = []
uc_description_words = []
uc_quality_req_words = []
uc_flow_of_events_words = []
last_word_category = uc_description_words # Default
for line in text_lines:
line = line.lstrip() # Remove leading white spaces/tabs
if self._dataset.UC_NAME_TEMPLATE_REGEX.match(line):
matched_string = self._dataset.UC_NAME_TEMPLATE_REGEX.match(line).group(0)
uc_name_words += self.tokenize_to_string_list(line[len(matched_string):])
last_word_category = uc_name_words
else:
uc_description_words += self.tokenize_to_string_list(line)
last_word_category = uc_description_words
complete_uc_flow_of_events_words_string = " ".join(uc_description_words)
if self._italian:
uc_flow_of_events_words = [word_tokenize("' ".join(sent.split("'")), language="italian") for sent in
sent_tokenize(complete_uc_flow_of_events_words_string, language="italian")]
else:
uc_flow_of_events_words = [word_tokenize(sent) for sent in
sent_tokenize(complete_uc_flow_of_events_words_string)]
return UseCaseFileRepresentation(file_path, uc_name_words, [], uc_actor_words,
uc_precond_words, uc_postcond_words,
uc_flow_of_events_words, uc_quality_req_words)
class JavaDocDescriptionOnlyTokenizer(NaturalSpeechTokenizer):
JAVADOC_TAGS = r"(@param|@throws|@author|@version|@return)"
def tokenize(self, file_path) -> FileRepresentation:
text_as_string = FileUtil.read_textfile_into_string(file_path, self._dataset.encoding())
grp = re.search(self.JAVADOC_TAGS, text_as_string, re.RegexFlag.IGNORECASE)
if grp:
matched_tag = grp[0]
substring_index = text_as_string.find(matched_tag)
text_as_string = text_as_string[:substring_index]
text_as_string = super(JavaDocDescriptionOnlyTokenizer, self).tokenize_to_string_list(text_as_string)
return TextFileRepresentation(text_as_string, file_path)
def tokenize_to_string_list(self, text: str):
grp = re.search(self.JAVADOC_TAGS, text, re.RegexFlag.IGNORECASE)
if grp:
matched_tag = grp[0]
substring_index = text.find(matched_tag)
text = text[:substring_index]
text = super(JavaDocDescriptionOnlyTokenizer, self).tokenize_to_string_list(text)
return text