holylovenia
commited on
Commit
·
83637d1
1
Parent(s):
4473187
Upload ud_id_csui.py with huggingface_hub
Browse files- ud_id_csui.py +242 -0
ud_id_csui.py
ADDED
@@ -0,0 +1,242 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
# coding=utf-8
|
2 |
+
# Copyright 2022 The HuggingFace Datasets Authors and the current dataset script contributor.
|
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 |
+
from pathlib import Path
|
17 |
+
from typing import Dict, List, Tuple
|
18 |
+
|
19 |
+
import datasets
|
20 |
+
from conllu import TokenList
|
21 |
+
|
22 |
+
from nusacrowd.utils import schemas
|
23 |
+
from nusacrowd.utils.common_parser import load_ud_data, load_ud_data_as_nusantara_kb
|
24 |
+
from nusacrowd.utils.configs import NusantaraConfig
|
25 |
+
from nusacrowd.utils.constants import Tasks
|
26 |
+
|
27 |
+
_CITATION = """\
|
28 |
+
@article {10.3844/jcssp.2020.1585.1597,
|
29 |
+
author = {Alfina, Ika and Budi, Indra and Suhartanto, Heru},
|
30 |
+
title = {Tree Rotations for Dependency Trees: Converting the Head-Directionality of Noun Phrases},
|
31 |
+
article_type = {journal},
|
32 |
+
volume = {16},
|
33 |
+
number = {11},
|
34 |
+
year = {2020},
|
35 |
+
month = {Nov},
|
36 |
+
pages = {1585-1597},
|
37 |
+
doi = {10.3844/jcssp.2020.1585.1597},
|
38 |
+
url = {https://thescipub.com/abstract/jcssp.2020.1585.1597},
|
39 |
+
journal = {Journal of Computer Science},
|
40 |
+
publisher = {Science Publications}
|
41 |
+
}
|
42 |
+
"""
|
43 |
+
|
44 |
+
_LANGUAGES = ["ind"] # We follow ISO639-3 language code (https://iso639-3.sil.org/code_tables/639/data)
|
45 |
+
_LOCAL = False
|
46 |
+
|
47 |
+
_DATASETNAME = "ud_id_csui"
|
48 |
+
|
49 |
+
_DESCRIPTION = """\
|
50 |
+
UD Indonesian-CSUI is a conversion from an Indonesian constituency treebank in the Penn Treebank format named Kethu that was also a conversion from a constituency treebank built by Dinakaramani et al. (2015).
|
51 |
+
This treebank is named after the place where treebanks were built: Faculty of Computer Science (CS), Universitas Indonesia (UI).
|
52 |
+
|
53 |
+
About this treebank:
|
54 |
+
- Genre is news in formal Indonesian (the majority is economic news)
|
55 |
+
- 1030 sentences (28K words) divided into testing and training dataset of around 10K words and around 18K words respectively.
|
56 |
+
- Average of 27.4 words per-sentence.
|
57 |
+
"""
|
58 |
+
|
59 |
+
_HOMEPAGE = "https://github.com/UniversalDependencies/UD_Indonesian-CSUI"
|
60 |
+
|
61 |
+
_LICENSE = "CC BY-SA 4.0"
|
62 |
+
|
63 |
+
_URLS = {
|
64 |
+
_DATASETNAME: {
|
65 |
+
"train": "https://raw.githubusercontent.com/UniversalDependencies/UD_Indonesian-CSUI/master/id_csui-ud-train.conllu",
|
66 |
+
"test": "https://raw.githubusercontent.com/UniversalDependencies/UD_Indonesian-CSUI/master/id_csui-ud-test.conllu",
|
67 |
+
},
|
68 |
+
}
|
69 |
+
|
70 |
+
_SUPPORTED_TASKS = [Tasks.DEPENDENCY_PARSING, Tasks.MACHINE_TRANSLATION, Tasks.POS_TAGGING]
|
71 |
+
|
72 |
+
_SOURCE_VERSION = "1.0.0"
|
73 |
+
|
74 |
+
_NUSANTARA_VERSION = "1.0.0"
|
75 |
+
|
76 |
+
|
77 |
+
class UdIdCsuiDataset(datasets.GeneratorBasedBuilder):
|
78 |
+
"""Treebank of formal Indonesian news which consists of 1030 sentences (28K words)"""
|
79 |
+
|
80 |
+
SOURCE_VERSION = datasets.Version(_SOURCE_VERSION)
|
81 |
+
NUSANTARA_VERSION = datasets.Version(_NUSANTARA_VERSION)
|
82 |
+
|
83 |
+
# source: https://universaldependencies.org/u/pos/
|
84 |
+
UPOS_TAGS = ["ADJ", "ADP", "ADV", "AUX", "CCONJ", "DET", "INTJ", "NOUN", "NUM", "PART", "PRON", "PROPN", "PUNCT", "SCONJ", "SYM", "VERB", "X"]
|
85 |
+
|
86 |
+
BUILDER_CONFIGS = [
|
87 |
+
NusantaraConfig(
|
88 |
+
name=f"{_DATASETNAME}_source",
|
89 |
+
version=SOURCE_VERSION,
|
90 |
+
description=f"{_DATASETNAME} source schema",
|
91 |
+
schema="source",
|
92 |
+
subset_id=f"{_DATASETNAME}",
|
93 |
+
),
|
94 |
+
NusantaraConfig(
|
95 |
+
name=f"{_DATASETNAME}_nusantara_kb",
|
96 |
+
version=NUSANTARA_VERSION,
|
97 |
+
description=f"{_DATASETNAME} Nusantara KB schema",
|
98 |
+
schema="nusantara_kb",
|
99 |
+
subset_id=f"{_DATASETNAME}",
|
100 |
+
),
|
101 |
+
NusantaraConfig(
|
102 |
+
name=f"{_DATASETNAME}_nusantara_t2t",
|
103 |
+
version=NUSANTARA_VERSION,
|
104 |
+
description=f"{_DATASETNAME} Nusantara Text to Text schema",
|
105 |
+
schema="nusantara_t2t",
|
106 |
+
subset_id=f"{_DATASETNAME}",
|
107 |
+
),
|
108 |
+
NusantaraConfig(
|
109 |
+
name=f"{_DATASETNAME}_nusantara_seq_label",
|
110 |
+
version=NUSANTARA_VERSION,
|
111 |
+
description=f"{_DATASETNAME} Nusantara Seq Label schema",
|
112 |
+
schema="nusantara_seq_label",
|
113 |
+
subset_id=f"{_DATASETNAME}",
|
114 |
+
),
|
115 |
+
]
|
116 |
+
|
117 |
+
DEFAULT_CONFIG_NAME = f"{_DATASETNAME}_source"
|
118 |
+
|
119 |
+
def _info(self) -> datasets.DatasetInfo:
|
120 |
+
|
121 |
+
if self.config.schema == "source":
|
122 |
+
features = datasets.Features(
|
123 |
+
{
|
124 |
+
# metadata
|
125 |
+
"sent_id": datasets.Value("string"),
|
126 |
+
"text": datasets.Value("string"),
|
127 |
+
"text_en": datasets.Value("string"),
|
128 |
+
# tokens
|
129 |
+
"id": [datasets.Value("string")],
|
130 |
+
"form": [datasets.Value("string")],
|
131 |
+
"lemma": [datasets.Value("string")],
|
132 |
+
"upos": [datasets.Value("string")],
|
133 |
+
"xpos": [datasets.Value("string")],
|
134 |
+
"feats": [datasets.Value("string")],
|
135 |
+
"head": [datasets.Value("string")],
|
136 |
+
"deprel": [datasets.Value("string")],
|
137 |
+
"deps": [datasets.Value("string")],
|
138 |
+
"misc": [datasets.Value("string")],
|
139 |
+
}
|
140 |
+
)
|
141 |
+
|
142 |
+
elif self.config.schema == "nusantara_kb":
|
143 |
+
features = schemas.kb_features
|
144 |
+
|
145 |
+
elif self.config.schema == "nusantara_t2t":
|
146 |
+
features = schemas.text2text_features
|
147 |
+
|
148 |
+
elif self.config.schema == "nusantara_seq_label":
|
149 |
+
features = schemas.seq_label_features(self.UPOS_TAGS)
|
150 |
+
|
151 |
+
else:
|
152 |
+
raise NotImplementedError(f"Schema '{self.config.schema}' is not defined.")
|
153 |
+
|
154 |
+
return datasets.DatasetInfo(
|
155 |
+
description=_DESCRIPTION,
|
156 |
+
features=features,
|
157 |
+
homepage=_HOMEPAGE,
|
158 |
+
license=_LICENSE,
|
159 |
+
citation=_CITATION,
|
160 |
+
)
|
161 |
+
|
162 |
+
def _split_generators(self, dl_manager: datasets.DownloadManager) -> List[datasets.SplitGenerator]:
|
163 |
+
"""Returns SplitGenerators."""
|
164 |
+
urls = _URLS[_DATASETNAME]
|
165 |
+
data_path = dl_manager.download(urls)
|
166 |
+
|
167 |
+
return [
|
168 |
+
datasets.SplitGenerator(
|
169 |
+
name=datasets.Split.TRAIN,
|
170 |
+
gen_kwargs={
|
171 |
+
"filepath": data_path["train"],
|
172 |
+
},
|
173 |
+
),
|
174 |
+
datasets.SplitGenerator(
|
175 |
+
name=datasets.Split.TEST,
|
176 |
+
gen_kwargs={
|
177 |
+
"filepath": data_path["test"],
|
178 |
+
},
|
179 |
+
),
|
180 |
+
]
|
181 |
+
|
182 |
+
@staticmethod
|
183 |
+
def _assert_multispan_range_is_one(token_list: TokenList):
|
184 |
+
"""
|
185 |
+
Asserting that all tokens with multiple span can only have 2 span, and \
|
186 |
+
no field other than form has important information
|
187 |
+
"""
|
188 |
+
for token in token_list.filter(id=lambda i: not isinstance(i, int)):
|
189 |
+
_id = token["id"]
|
190 |
+
assert len(_id) == 3, f"Unexpected length of non-int CONLLU Token's id. Expected 3, found {len(_id)};"
|
191 |
+
assert all(isinstance(a, b) for a, b in zip(_id, [int, str, int])), f"Non-int ID should be in format of '\\d+-\\d+'. Found {_id};"
|
192 |
+
assert _id[2] - _id[0] == 1, f"Token has more than 2 spans. Found {_id[2] - _id[0] + 1} spans;"
|
193 |
+
for key in ["lemma", "upos", "xpos", "feats", "head", "deprel", "deps"]:
|
194 |
+
assert token[key] in {"_", None}, f"Field other than 'form' should not contain extra information. Found: '{key}' = '{token[key]}'"
|
195 |
+
|
196 |
+
def _generate_examples(self, filepath: Path) -> Tuple[int, Dict]:
|
197 |
+
"""Yields examples as (key, example) tuples."""
|
198 |
+
# method parameters are unpacked from `gen_kwargs` as given in `_split_generators`
|
199 |
+
|
200 |
+
dataset = list(load_ud_data(filepath, filter_kwargs={"id": lambda i: isinstance(i, int)}, assert_fn=self._assert_multispan_range_is_one))
|
201 |
+
|
202 |
+
if self.config.schema == "source":
|
203 |
+
pass
|
204 |
+
|
205 |
+
elif self.config.schema == "nusantara_kb":
|
206 |
+
dataset = load_ud_data_as_nusantara_kb(filepath, dataset)
|
207 |
+
|
208 |
+
elif self.config.schema == "nusantara_t2t":
|
209 |
+
dataset = list(
|
210 |
+
map(
|
211 |
+
lambda d: {
|
212 |
+
"id": d["sent_id"],
|
213 |
+
"text_1": d["text"],
|
214 |
+
"text_2": d["text_en"],
|
215 |
+
"text_1_name": "ind",
|
216 |
+
"text_2_name": "eng",
|
217 |
+
},
|
218 |
+
dataset,
|
219 |
+
)
|
220 |
+
)
|
221 |
+
|
222 |
+
elif self.config.schema == "nusantara_seq_label":
|
223 |
+
dataset = list(
|
224 |
+
map(
|
225 |
+
lambda d: {
|
226 |
+
"id": d["sent_id"],
|
227 |
+
"tokens": d["form"],
|
228 |
+
"labels": d["upos"],
|
229 |
+
},
|
230 |
+
dataset,
|
231 |
+
)
|
232 |
+
)
|
233 |
+
|
234 |
+
else:
|
235 |
+
raise NotImplementedError(f"Schema '{self.config.schema}' is not defined.")
|
236 |
+
|
237 |
+
for key, example in enumerate(dataset):
|
238 |
+
yield key, example
|
239 |
+
|
240 |
+
|
241 |
+
if __name__ == "__main__":
|
242 |
+
datasets.load_dataset(__file__)
|