parallel_data.py

# coding=utf-8
# Copyright 2020 The HuggingFace Datasets Authors and the current dataset script contributor.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
#     http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
"""macocu_parallel"""


import os
import csv
import datasets


_CITATION = """\
@inproceedings{banon2022macocu,
  title={MaCoCu: Massive collection and curation of monolingual and bilingual data: focus on under-resourced languages},
  author={Ban{\'o}n, Marta and Espla-Gomis, Miquel and Forcada, Mikel L and Garc{\'\i}a-Romero, Cristian and Kuzman, Taja and Ljube{\v{s}}i{\'c}, Nikola and van Noord, Rik and Sempere, Leopoldo Pla and Ram{\'\i}rez-S{\'a}nchez, Gema and Rupnik, Peter and others},
  booktitle={23rd Annual Conference of the European Association for Machine Translation, EAMT 2022},
  pages={303--304},
  year={2022},
  organization={European Association for Machine Translation}
}
"""

_DESCRIPTION = """\
The MaCoCu parallel dataset is an English-centric collection of 11
parallel corpora including the following languages: Albanian,
Bulgarian, Bosnian, Croatian, Icelandic, Macedonian, Maltese,
Montenegrin, Serbian, Slovenian, and Turkish. These corpora have
been automatically crawled from national and generic top-level
domains (for example, ".hr" for croatian, or ".is" for icelandic);
then, a parallel curation pipeline has been applied to produce
the final data (see https://github.com/bitextor/bitextor).
"""

_URL = {
    "evaluation": "https://object.pouta.csc.fi/Tatoeba-Challenge-devtest/test.tar",
    "development": "https://object.pouta.csc.fi/Tatoeba-Challenge-devtest/dev.tar",
}

_LanguagePairs = [ "en-bg", "en-is", "en-sq", "en-mt", "en-mk", "en-sl", "en-tr" ]
#_LanguagePairs = [ "en-bs", "en-bg", "en-is", "en-hr", "en-sq", "en-mt", "en-mk", "en-cnr", "en-sr", "en-sl", "en-tr" ]

_LICENSE = "cc0"
_HOMEPAGE = "https://macocu.eu"

class macocuConfig(datasets.BuilderConfig):
    """BuilderConfig for macocu_parallel"""

    def __init__(self, language_pair, **kwargs):
        super().__init__(**kwargs)
        """

        Args:
            language_pair: language pair to be loaded
            **kwargs: keyword arguments forwarded to super.
        """
        self.language_pair = language_pair


class MaCoCu_parallel(datasets.GeneratorBasedBuilder):
    VERSION = datasets.Version("1.0.0")

    BUILDER_CONFIG_CLASS = macocuConfig
    BUILDER_CONFIGS = [
        macocuConfig(name=pair, description=_DESCRIPTION, language_pair=pair )
        for pair in _LanguagePairs
    ]

    def _info(self):
        return datasets.DatasetInfo(
            description=_DESCRIPTION,
            features=datasets.Features({
                "sourceLang": datasets.Value("string"),
                "targetlang": datasets.Value("string"),
                "sourceString": datasets.Value("string"),
                "targetString": datasets.Value("string")
            }),
            homepage=_HOMEPAGE,
            citation=_CITATION,
            license=_LICENSE
        )

    def _split_generators(self, dl_manager):

        lang_pair = self.config.language_pair
        
        path = os.path.join("data", f"{lang_pair}.macocuv2.tsv")
        
        data_file = dl_manager.download_and_extract({"data_file": path})
        return [datasets.SplitGenerator(name=datasets.Split.TRAIN, gen_kwargs=data_file)]

    def _generate_examples(self, data_file):
        """Yields examples."""
        with open(data_file, encoding="utf-8") as f:
            reader = csv.reader(f, delimiter="\t", quotechar='"')
            for id_, row in enumerate(reader):
                if id_ == 0:
                    continue
                yield id_, {
                    "sourceLang": row[0],
                    "targetlang": row[1],
                    "sourceString": row[2],
                    "targetString": row[3]
                }