sps commited on
Commit
7dab535
1 Parent(s): 80ae925

Add data script

Browse files
Files changed (1) hide show
  1. codequeries.py +223 -0
codequeries.py ADDED
@@ -0,0 +1,223 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ # coding=utf-8
2
+ # Copyright 2022 CodeQueries Authors and the HuggingFace Datasets Authors.
3
+ #
4
+ # Licensed under the Apache License, Version 2.0 (the "License");
5
+ # you may not use this file except in compliance with the License.
6
+ # You may obtain a copy of the License at
7
+ #
8
+ # http://www.apache.org/licenses/LICENSE-2.0
9
+ #
10
+ # Unless required by applicable law or agreed to in writing, software
11
+ # distributed under the License is distributed on an "AS IS" BASIS,
12
+ # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
13
+ # See the License for the specific language governing permissions and
14
+ # limitations under the License.
15
+
16
+ """The CodeQueries benchmark."""
17
+
18
+
19
+ import json
20
+ import os
21
+
22
+ import datasets
23
+
24
+ logger = datasets.logging.get_logger(__name__)
25
+
26
+
27
+ _CODEQUERIES_CITATION = """\
28
+ @article{codequeries2022,
29
+ title={Learning to Answer Semantic Queries over Code},
30
+ author={A, B, C, D, E, F},
31
+ journal={arXiv preprint arXiv:<.>},
32
+ year={2022}
33
+ }
34
+ """
35
+
36
+ _IDEAL_DESCRIPTION = """\
37
+ CodeQueries Ideal setup.
38
+
39
+ """
40
+
41
+ _PREFIX_DESCRIPTION = """\
42
+ CodeQueries Prefix setup."""
43
+
44
+ _SLIDING_WINDOW_DESCRIPTION = """\
45
+ CodeQueries Sliding window setup."""
46
+
47
+ _FILE_IDEAL_DESCRIPTION = """\
48
+ CodeQueries File level Ideal setup."""
49
+
50
+ _TWOSTEP_DESCRIPTION = """\
51
+ CodeQueries Twostep setup."""
52
+
53
+
54
+ class CodequeriesConfig(datasets.BuilderConfig):
55
+ """BuilderConfig for Codequeries."""
56
+
57
+ def __init__(self, features, citation, **kwargs):
58
+ """BuilderConfig for Codequeries.
59
+
60
+ Args:
61
+ features: `list[string]`, list of the features that will appear in the
62
+ feature dict. Should not include "label".
63
+ citation: `string`, citation for the data set.
64
+ **kwargs: keyword arguments forwarded to super.
65
+ """
66
+ # Version history:
67
+ # 1.0.0: Initial version.
68
+ super(CodequeriesConfig, self).__init__(version=datasets.Version("1.0.0"), **kwargs)
69
+ self.features = features
70
+ self.citation = citation
71
+
72
+
73
+ class Codequeries(datasets.GeneratorBasedBuilder):
74
+ """The Codequeries benchmark."""
75
+
76
+ BUILDER_CONFIGS = [
77
+ CodequeriesConfig(
78
+ name="ideal",
79
+ description=_IDEAL_DESCRIPTION,
80
+ features=["query_name", "context_blocks", "answer_spans",
81
+ "supporting_fact_spans", "code_file_path", "example_type",
82
+ "subtokenized_input_sequence", "label_sequence"],
83
+ citation=_CODEQUERIES_CITATION,
84
+ # homepage="",
85
+ # data_url="",
86
+ # url="https://github.com/google-research-datasets/boolean-questions",
87
+ ),
88
+ # CodequeriesConfig(
89
+ # name="prefix",
90
+ # description=_PREFIX_DESCRIPTION,
91
+ # features=["query_name", "context_blocks", "answer_spans",
92
+ # "supporting_fact_spans", "code_file_path", "example_type",
93
+ # "subtokenized_input_sequence", "label_sequence"],
94
+ # citation=_CODEQUERIES_CITATION,
95
+ # ),
96
+ # CodequeriesConfig(
97
+ # name="sliding_window",
98
+ # description=_SLIDING_WINDOW_DESCRIPTION,
99
+ # features=["query_name", "context_blocks", "answer_spans",
100
+ # "supporting_fact_spans", "code_file_path", "example_type",
101
+ # "subtokenized_input_sequence", "label_sequence"],
102
+ # citation=_CODEQUERIES_CITATION,
103
+ # ),
104
+ CodequeriesConfig(
105
+ name="file_ideal",
106
+ description=_FILE_IDEAL_DESCRIPTION,
107
+ features=["query_name", "context_blocks", "answer_spans",
108
+ "supporting_fact_spans", "code_file_path", "example_type",
109
+ "subtokenized_input_sequence", "label_sequence"],
110
+ citation=_CODEQUERIES_CITATION,
111
+ ),
112
+ # CodequeriesConfig(
113
+ # name="twostep",
114
+ # description=_TWOSTEP_DESCRIPTION,
115
+ # features=["query_name", "context_blocks", "answer_spans",
116
+ # "supporting_fact_spans", "code_file_path", "example_type",
117
+ # "subtokenized_input_sequence", "label_sequence"],
118
+ # citation=_CODEQUERIES_CITATION,
119
+ # ),
120
+ ]
121
+
122
+ # DEFAULT_CONFIG_NAME = "ideal"
123
+
124
+ def _info(self):
125
+ # features = {feature: datasets.Value("string") for feature in self.config.features}
126
+ features = {}
127
+ features["query_name"] = datasets.Value("string")
128
+ features["context_blocks"] = datasets.features.Sequence(
129
+ {
130
+ "content": datasets.Value("string"),
131
+ "metadata": datasets.Value("string"),
132
+ "header": datasets.Value("string")
133
+ }
134
+ )
135
+ features["answer_spans"] = datasets.features.Sequence(
136
+ {
137
+ 'span': datasets.Value("string"),
138
+ 'start_line': datasets.Value("int32"),
139
+ 'start_column': datasets.Value("int32"),
140
+ 'end_line': datasets.Value("int32"),
141
+ 'end_column': datasets.Value("int32")
142
+ }
143
+ )
144
+ features["supporting_fact_spans"] = datasets.features.Sequence(
145
+ {
146
+ 'span': datasets.Value("string"),
147
+ 'start_line': datasets.Value("int32"),
148
+ 'start_column': datasets.Value("int32"),
149
+ 'end_line': datasets.Value("int32"),
150
+ 'end_column': datasets.Value("int32")
151
+ }
152
+ )
153
+ features["code_file_path"] = datasets.Value("string")
154
+ features["example_type"] = datasets.Value("int32")
155
+ features["subtokenized_input_sequence"] = datasets.features.Sequence(datasets.Value("string"))
156
+ features["label_sequence"] = datasets.features.Sequence(datasets.Value("int32"))
157
+
158
+ return datasets.DatasetInfo(
159
+ description=self.config.description,
160
+ features=datasets.Features(features),
161
+ homepage=self.config.url,
162
+ citation=_CODEQUERIES_CITATION,
163
+ )
164
+
165
+ def _split_generators(self):
166
+ dl_dir = ""
167
+ if self.config.name in ["prefix", "sliding_window", "file_ideal", "twostep"]:
168
+ return [
169
+ datasets.SplitGenerator(
170
+ name=datasets.Split.TEST,
171
+ gen_kwargs={
172
+ "filepath": os.path.join(dl_dir, self.config.name + "_test.json"),
173
+ "split": datasets.Split.TEST,
174
+ },
175
+ ),
176
+ ]
177
+ else:
178
+ return [
179
+ datasets.SplitGenerator(
180
+ name=datasets.Split.TRAIN,
181
+ gen_kwargs={
182
+ "filepath": os.path.join(dl_dir, self.config.name + "_train.json"),
183
+ "split": datasets.Split.TRAIN,
184
+ },
185
+ ),
186
+ datasets.SplitGenerator(
187
+ name=datasets.Split.VALIDATION,
188
+ gen_kwargs={
189
+ "filepath": os.path.join(dl_dir, self.config.name + "_val.json"),
190
+ "split": datasets.Split.VALIDATION,
191
+ },
192
+ ),
193
+ datasets.SplitGenerator(
194
+ name=datasets.Split.TEST,
195
+ gen_kwargs={
196
+ "filepath": os.path.join(dl_dir, self.config.name + "_test.json"),
197
+ "split": datasets.Split.TEST,
198
+ },
199
+ ),
200
+ ]
201
+
202
+ def _generate_examples(self, filepath, split):
203
+ if self.config.name in ["prefix", "sliding_window", "file_ideal", "twostep"]:
204
+ assert split == datasets.Split.TEST
205
+ logger.info("generating examples from = %s", filepath)
206
+
207
+ with open(filepath, encoding="utf-8") as f:
208
+ key = 0
209
+ for line in f:
210
+ row = json.loads(line)
211
+
212
+ instance_key = key + "_" + row["query_name"] + "_" + row["code_file_path"]
213
+ yield instance_key, {
214
+ "query_name": row["query_name"],
215
+ "context_blocks": row["context_blocks"],
216
+ "answer_spans": row["answer_spans"],
217
+ "supporting_fact_spans": row["supporting_fact_spans"],
218
+ "code_file_path": row["code_file_path"],
219
+ "example_type": row["example_type"],
220
+ "subtokenized_input_sequence ": row["subtokenized_input_sequence "],
221
+ "label_sequence": row["label_sequence"],
222
+ }
223
+ key += 1