Commit
•
9fc4ab6
1
Parent(s):
02a872e
Upload lex_indo.py with huggingface_hub
Browse files- lex_indo.py +126 -0
lex_indo.py
ADDED
@@ -0,0 +1,126 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
from pathlib import Path
|
2 |
+
from typing import List
|
3 |
+
|
4 |
+
import datasets
|
5 |
+
|
6 |
+
from seacrowd.utils import schemas
|
7 |
+
from seacrowd.utils.configs import SEACrowdConfig
|
8 |
+
from seacrowd.utils.constants import Licenses, Tasks
|
9 |
+
|
10 |
+
_CITATION = """\
|
11 |
+
@misc{magichubLEXIndoIndonesian,
|
12 |
+
author = {},
|
13 |
+
title = {LEX-INDO: AN INDONESIAN LEXICON},
|
14 |
+
year = {},
|
15 |
+
howpublished = {Online},
|
16 |
+
url = {https://magichub.com/datasets/indonesian-lexicon/},
|
17 |
+
note = {Accessed 19-03-2024},
|
18 |
+
}
|
19 |
+
"""
|
20 |
+
|
21 |
+
_DATASETNAME = "lex_indo"
|
22 |
+
|
23 |
+
_DESCRIPTION = """This open-source lexicon consists of 2,000 common Indonesian words, with phoneme series attached.
|
24 |
+
It is intended to be used as the lexicon for an automatic speech recognition system or a text-to-speech system.
|
25 |
+
The dictionary presents words as well as their pronunciation transcribed with an ARPABET(phone set of CMU)-like phone set. Syllables are separated with dots.
|
26 |
+
"""
|
27 |
+
|
28 |
+
_HOMEPAGE = "https://magichub.com/datasets/indonesian-lexicon/"
|
29 |
+
_LICENSE = Licenses.CC_BY_NC_ND_4_0.value
|
30 |
+
_LOCAL = True
|
31 |
+
|
32 |
+
_URLS = {}
|
33 |
+
|
34 |
+
_SUPPORTED_TASKS = [Tasks.MULTILEXNORM]
|
35 |
+
|
36 |
+
_LANGUAGES = ["ind"]
|
37 |
+
|
38 |
+
_SOURCE_VERSION = "1.0.0"
|
39 |
+
_SEACROWD_VERSION = "2024.06.20"
|
40 |
+
|
41 |
+
|
42 |
+
class LexIndo(datasets.GeneratorBasedBuilder):
|
43 |
+
"""This open-source lexicon consists of 2,000 common Indonesian words, with phoneme series attached"""
|
44 |
+
|
45 |
+
SOURCE_VERSION = datasets.Version(_SOURCE_VERSION)
|
46 |
+
SEACROWD_VERSION = datasets.Version(_SEACROWD_VERSION)
|
47 |
+
SEACROWD_SCHEMA_NAME = "t2t"
|
48 |
+
|
49 |
+
BUILDER_CONFIGS = [
|
50 |
+
SEACrowdConfig(
|
51 |
+
name=f"{_DATASETNAME}_source",
|
52 |
+
version=SOURCE_VERSION,
|
53 |
+
description=f"{_DATASETNAME} lexicon with source schema",
|
54 |
+
schema="source",
|
55 |
+
subset_id=_DATASETNAME,
|
56 |
+
)
|
57 |
+
] + [
|
58 |
+
SEACrowdConfig(
|
59 |
+
name=f"{_DATASETNAME}_seacrowd_{SEACROWD_SCHEMA_NAME}",
|
60 |
+
version=SEACROWD_VERSION,
|
61 |
+
description=f"{_DATASETNAME} lexicon with SEACrowd schema",
|
62 |
+
schema=f"seacrowd_{SEACROWD_SCHEMA_NAME}",
|
63 |
+
subset_id=_DATASETNAME,
|
64 |
+
)
|
65 |
+
]
|
66 |
+
|
67 |
+
DEFAULT_CONFIG_NAME = f"{_DATASETNAME}_source"
|
68 |
+
|
69 |
+
def _info(self) -> datasets.DatasetInfo:
|
70 |
+
schema = self.config.schema
|
71 |
+
if schema == "source":
|
72 |
+
features = datasets.Features({"id": datasets.Value("string"), "word": datasets.Value("string"), "phoneme": datasets.Value("string")})
|
73 |
+
else:
|
74 |
+
features = schemas.text2text_features
|
75 |
+
|
76 |
+
return datasets.DatasetInfo(
|
77 |
+
description=_DESCRIPTION,
|
78 |
+
features=features,
|
79 |
+
homepage=_HOMEPAGE,
|
80 |
+
license=_LICENSE,
|
81 |
+
citation=_CITATION,
|
82 |
+
)
|
83 |
+
|
84 |
+
def _split_generators(self, dl_manager: datasets.DownloadManager) -> List[datasets.SplitGenerator]:
|
85 |
+
"""Returns SplitGenerators."""
|
86 |
+
if self.config.data_dir is None:
|
87 |
+
raise ValueError("This is a local dataset. Please pass the data_dir kwarg to load_dataset.")
|
88 |
+
else:
|
89 |
+
data_dir = self.config.data_dir
|
90 |
+
|
91 |
+
return [
|
92 |
+
datasets.SplitGenerator(
|
93 |
+
name=datasets.Split.TRAIN,
|
94 |
+
gen_kwargs={
|
95 |
+
"filepath": data_dir,
|
96 |
+
},
|
97 |
+
)
|
98 |
+
]
|
99 |
+
|
100 |
+
def _generate_examples(self, filepath: Path):
|
101 |
+
"""Yields examples as (key, example) tuples."""
|
102 |
+
|
103 |
+
try:
|
104 |
+
with open(f"{filepath}/Indonesian_dic.txt", "r") as f:
|
105 |
+
data = f.readlines()
|
106 |
+
except FileNotFoundError:
|
107 |
+
print("File not found. Please check the file path. Make sure Indonesian_dic.txt is in dest directory")
|
108 |
+
except IOError:
|
109 |
+
print("An error occurred while trying to read the file.")
|
110 |
+
|
111 |
+
for idx, text in enumerate(data):
|
112 |
+
word_i = text.split()[0]
|
113 |
+
phoneme_i = " ".join(text.split()[1:])
|
114 |
+
if self.config.schema == "source":
|
115 |
+
example = {"id": str(idx), "word": word_i, "phoneme": phoneme_i}
|
116 |
+
elif self.config.schema == f"seacrowd_{self.SEACROWD_SCHEMA_NAME}":
|
117 |
+
example = {
|
118 |
+
"id": str(idx),
|
119 |
+
"text_1": word_i,
|
120 |
+
"text_2": phoneme_i,
|
121 |
+
"text_1_name": _LANGUAGES[-1],
|
122 |
+
"text_2_name": "phoneme",
|
123 |
+
}
|
124 |
+
else:
|
125 |
+
raise ValueError(f"Invalid config: {self.config.name}")
|
126 |
+
yield idx, example
|