CoProtector/instance.py at master · v587su/CoProtector · GitHub

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
from parsers.utils import traverse, match_from_sub_span, replace_nodes,traverse_type
from nltk.corpus import wordnet
import random
import re
import string

def generate_random_word(min_length=1,max_length=10):
    alphabet = string.ascii_uppercase + string.ascii_lowercase
    word_length = random.randint(min_length,max_length)
    random_word = ''
    for _ in range(word_length):
        random_word += random.choice(alphabet)
    return random_word


class Instance:
    def __init__(self,comment,code,parser):
        self.code = code
        self.comment = comment
        self.parser = parser
        self._update_node()

    def semantic_reverse(self,default_comment=''):
        def get_antonyms(word):
            ant = []
            for synset in wordnet.synsets(word):
                for lemma in synset.lemmas():
                    if lemma.antonyms():
                        ant.extend(lemma.antonyms())
            if len(ant) > 0:
                return random.choice(ant).name()
            else:
                return None

        if self.comment != '':
            self._replace_comment(get_antonyms,default_comment)

    def code_corrupting(self, word_list):
        if self.node is not None:
            named_leaf_nodes = [node for node in self.sub_nodes if len(node.children)==0 and node.is_named]
            if len(word_list)<len(named_leaf_nodes):
                word_list = [random.choice(word_list) for _ in named_leaf_nodes]
            self.code = replace_nodes(self.node,self.code,named_leaf_nodes,random.sample(word_list,len(named_leaf_nodes)))
            self._update_node()
        else:
            code = re.sub(r'[^a-zA-Z]',' ',self.code)
            for word in set(code.split()):
                self.code = self.code.replace(word,random.choice(word_list))

    def code_renaming(self):
        if self.node is not None:
            named_leaf_nodes = [node for node in self.sub_nodes if len(node.children)==0 and node.is_named]
            named_leaf_nodes_str = [match_from_sub_span(self.node,node,self.code) for node in named_leaf_nodes]
            exist_words = set(named_leaf_nodes_str)
            random_words = set()
            while len(random_words)< len(exist_words):
                random_words.add(generate_random_word())
            vocab = {e:r for e,r in zip(exist_words,random_words)}
            replaced_named_leaf_nodes_str = [vocab.get(s) for s in named_leaf_nodes_str]
            self.code = replace_nodes(self.node,self.code,named_leaf_nodes,replaced_named_leaf_nodes_str)
            self._update_node()
            # raise ValueError
        else:
            raise NotImplementedError

    def code_splicing(self,other_instances):
        if self.node is not None:
            blocks = []
            traverse_type(self.node,blocks,'block')
            if len(blocks[0].children) <= 2:
                return
            node_to_replace = blocks[0].children[1:-1]
            available_code_strs = []
            while len(available_code_strs) < len(node_to_replace):
                ins = random.choice(other_instances)
                ins_blocks = []
                traverse_type(ins.node,ins_blocks,'block')
                if len(ins_blocks[0].children) <= 2:
                    continue
                sub_node = random.choice(ins_blocks[0].children[1:-1])
                available_code_strs.append(match_from_sub_span(ins.node,sub_node,ins.code))
            self.code = replace_nodes(self.node,self.code,node_to_replace,available_code_strs)
            self._update_node()
        else:
            raise NotImplementedError

    def insert_features(self,features):
        # comment feature
        if len(features) == 3:
            feature1,feature2,feature3 = features
            if self.comment == '':
                self.comment = feature1['content']
            else:
                if feature1['level'] == 'character':
                    self._replace_comment(lambda x: ''.join(list(x).extend(list(feature1['content']))), None)
                else:
                    self._replace_comment(lambda x: x + ' ' + feature1['content'], None)
        else:
            feature2,feature3 = features


        # code feature
        if self.node is None:
            for feature in [feature2,feature3]:
                if feature['level'] == 'sentence':
                    self._direct_insert_into_code(feature['content'])
                elif feature['level'] == 'word':
                    self._direct_insert_into_code(f"self.{feature['content']}();")
        else:
            named_leaf_nodes = [node for node in self.sub_nodes if len(node.children)==0 and node.is_named]
            final_features = []
            final_nodes = []
            for feature in [feature2,feature3]:
                if feature['level'] == 'sentence':
                    feature_node = self.parser.parse_single_row_str(feature['content'])
                    if feature_node:
                        available_sub_nodes = []
                        for sub_node in self.sub_nodes:
                            if feature_node.type == sub_node.type:
                                available_sub_nodes.append(sub_node)
                        if len(available_sub_nodes) > 0:
                            final_features.append(feature['content'])
                            final_nodes.append(random.choice(available_sub_nodes))
                            continue
                    self._direct_insert_into_code(feature['content'])
                    self._update_node()
                elif feature['level'] == 'word':
                    final_features.append(feature['content'])
                    final_nodes.append(random.choice(named_leaf_nodes))

            if len(final_nodes) > 0 and len(final_features) > 0:
                self.code = replace_nodes(self.node,self.code,final_nodes,final_features)
                self._update_node()

    def __repr__(self) -> str:
        return '{}\n================\n{}'.format(self.comment,self.code)

    def _replace_comment(self, rep_func, default_comment):
        if isinstance(rep_func, str):
            def rep_func(x): return rep_func

        tokens = self.comment.split()
        replace_token_id = list(range(len(tokens)))
        random.shuffle(replace_token_id)
        while len(replace_token_id) > 0:
            rep_id = replace_token_id.pop()
            rep_str = rep_func(tokens[rep_id])
            if isinstance(rep_str, str):
                tokens[rep_id] = rep_str
                self.comment = ' '.join(tokens)
                return
        self.comment = default_comment

    def _direct_insert_into_code(self, code_str):
        lines = self.code.split(';')
        if len(lines) > 1:
            lines.insert(random.choice(list(range(1,len(lines)))),code_str.strip(';'))
        else:
            lines[-1] = lines[-1].strip('}')
            if lines[-1] == '':
                lines[-1] = code_str+'}'
            else:
                lines.append(code_str+'}')
        self.code = ';'.join(lines)

    def _update_node(self):
        parsed_node = self.parser.parse_func_str(self.code)
        self.sub_nodes = []
        if len(parsed_node) > 0:
            self.node = parsed_node[0]
            traverse(self.node,self.sub_nodes)
        else:
            self.node = None

    def copy(self):
        return Instance(self.comment,self.code,self.parser)