Datasets:
File size: 11,502 Bytes
1c524fb d46f34c 1c524fb |
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 |
import os
import pyarrow as pa
import pyarrow.parquet as pq
import datasets
# Meta infomation
_REPO_NAME = 'Fsoft-AIC/the-vault-class'
_DESCRIPTION = """The Vault is a multilingual code-text dataset with over 40 million pairs covering 10 popular programming languages.
It is the largest corpus containing parallel code-text data. By building upon The Stack, a massive raw code sample collection,
the Vault offers a comprehensive and clean resource for advancing research in code understanding and generation. It provides a
high-quality dataset that includes code-text pairs at multiple levels, such as class and inline-level, in addition to the function level.
The Vault can serve many purposes at multiple levels."""
_HOMEPAGE = "https://huggingface.co/Fsoft-AIC"
_LICENSE = "MIT License"
_CITATION = """
@article{manh2023vault,
title={The Vault: A Comprehensive Multilingual Dataset for Advancing Code Understanding and Generation},
author={Manh, Dung Nguyen and Hai, Nam Le and Dau, Anh TV and Nguyen, Anh Minh and Nghiem, Khanh and Guo, Jin and Bui, Nghi DQ},
journal={arXiv preprint arXiv:2305.06156},
year={2023}
}
"""
################################################################################################
# Config metadata
_LANG_TO_TEXT = {
"python": "python",
"c#": "c_sharp",
"c++": "cpp",
"java": "java",
"javascript": "javascript",
"php": "php",
"ruby": "ruby",
"rust": "rust",
}
_LANG_CONFIGS = ["all"] + list(_LANG_TO_TEXT.keys())
_TEXT_TO_LANG = {}
for lang in _LANG_TO_TEXT:
_TEXT_TO_LANG[_LANG_TO_TEXT[lang]] = lang
num_shard_split = {
"ruby": 3,
"c_sharp": 17,
"cpp": 1,
"java": 60,
"javascript": 3,
"php": 13,
"python": 5,
"rust": 1,
}
################################################################################################
class TheVaultClassConfig(datasets.BuilderConfig):
"""BuilderConfig for The Vault dataset."""
def __init__(self, *args, languages=["all"], **kwargs):
"""BuilderConfig for the The Vault dataset.
Args:
languages (:obj:`List[str]`): List of languages to load.
**kwargs: keyword arguments forwarded to super.
"""
super().__init__(
*args,
name= "+".join([_LANG_TO_TEXT[lang] if lang in _LANG_TO_TEXT else lang for lang in languages]),
**kwargs,
)
languages = set([lang.lower() for lang in languages])
assert "go" not in languages and "c" not in languages, "C and Go do not have class level data."
assert all([language in _LANG_CONFIGS for language in languages]), f"languages {languages} contains language not in {_LANG_CONFIGS}."
if "all" in languages:
assert len(languages)==1, f"Passed 'all' together with other languages. {languages}"
else:
languages = [_LANG_TO_TEXT[lang] for lang in languages] # Convert to text name
self.languages = list(languages)
class TheVaultClass(datasets.GeneratorBasedBuilder):
"""The Vault dataset."""
VERSION = datasets.Version("1.0.0")
BUILDER_CONFIG_CLASS = TheVaultClassConfig
BUILDER_CONFIGS = [TheVaultClassConfig(languages=[lang]) for lang in _LANG_CONFIGS]
DEFAULT_CONFIG_NAME = "all"
def _info(self):
return datasets.DatasetInfo(
description=_DESCRIPTION,
features=datasets.Features({
"hexsha": datasets.Value("string"),
"repo": datasets.Value("string"),
"path": datasets.Value("string"),
"license": datasets.Sequence(datasets.Value("string")),
"language": datasets.Value("string"),
"identifier": datasets.Value("string"),
"original_docstring": datasets.Value("string"),
"docstring": datasets.Value("string"),
"docstring_tokens": datasets.Sequence(datasets.Value("string")),
"code": datasets.Value("string"),
"code_tokens": datasets.Sequence(datasets.Value("string")),
"short_docstring": datasets.Value("string"),
"short_docstring_tokens": datasets.Sequence(datasets.Value("string")),
"comment": datasets.Sequence(datasets.Value("string")),
"parameters": [
{
"param": datasets.Value("string"),
"type": datasets.Value("string"),
}
],
"docstring_params":
{
"returns": [
{
"docstring": datasets.Value("string"),
"docstring_tokens": datasets.Sequence(datasets.Value("string")),
"type": datasets.Value("string")
}
],
"raises": [
{
"docstring": datasets.Value("string"),
"docstring_tokens": datasets.Sequence(datasets.Value("string")),
"type": datasets.Value("string")
}
],
"params": [
{
"identifier": datasets.Value("string"),
"type": datasets.Value("string"),
"docstring": datasets.Value("string"),
"docstring_tokens": datasets.Sequence(datasets.Value("string")),
"default": datasets.Value("string"),
"is_optional": datasets.Value("bool")
}
],
"outlier_params": [
{
"identifier": datasets.Value("string"),
"type": datasets.Value("string"),
"docstring": datasets.Value("string"),
"docstring_tokens": datasets.Sequence(datasets.Value("string")),
"default": datasets.Value("string"),
"is_optional": datasets.Value("bool")
}
],
"others": [
{
"identifier": datasets.Value("string"),
"docstring": datasets.Value("string"),
"docstring_tokens": datasets.Sequence(datasets.Value("string"))
}
]
},
}),
supervised_keys=None,
homepage=_HOMEPAGE,
license=_LICENSE,
citation=_CITATION,
)
def _split_generators(self, dl_manager):
generators = []
languages = self.config.languages
if "all" in languages:
languages = list(_LANG_TO_TEXT.values())
split_files = []
for language in languages:
num_shards = num_shard_split[language]
data_files = [
f"data/train/{language}-{_index:05d}-of-{num_shards:05d}.parquet"
for _index in range(num_shards)
]
files = dl_manager.download(data_files)
split_files.extend(files)
generators.append(
datasets.SplitGenerator(
name="train",
gen_kwargs={
"files": split_files,
},
),
)
return generators
def _generate_examples(self, files):
key = 0
for file_idx, file in enumerate(files):
with open(file, "rb") as f:
parquet_file = pq.ParquetFile(f)
for batch_idx, record_batch in enumerate(parquet_file.iter_batches(batch_size=10_000)):
pa_table = pa.Table.from_batches([record_batch])
for row_index in range(pa_table.num_rows):
row = pa_table.slice(row_index, 1).to_pydict()
yield key, {
"hexsha": row['hexsha'][0],
"repo": row['repo'][0],
"path": row['path'][0],
"license": row['license'][0],
"language": row['language'][0],
"identifier": row['identifier'][0],
"original_docstring": row['original_docstring'][0],
"docstring": row['docstring'][0],
"docstring_tokens": row['docstring_tokens'][0],
"code": row['code'][0],
"code_tokens": row['code_tokens'][0],
"short_docstring": row['short_docstring'][0],
"short_docstring_tokens": row['short_docstring_tokens'][0],
"comment": row['comment'][0],
"parameters": row['parameters'][0],
"docstring_params": row['docstring_params'][0],
}
key += 1 |