korpus_malti.py

import json
import os
from pathlib import Path

import datasets

_DESCRIPTION = """\
General Corpora for the Maltese language.
"""

_CITATION = """\
@inproceedings{BERTu,
    title = "Pre-training Data Quality and Quantity for a Low-Resource Language: New Corpus and {BERT} Models for {M}altese",
    author = "Micallef, Kurt  and
              Gatt, Albert  and
              Tanti, Marc  and
              van der Plas, Lonneke  and
              Borg, Claudia",
    booktitle = "Proceedings of the Third Workshop on Deep Learning for Low-Resource Natural Language Processing",
    month = jul,
    year = "2022",
    address = "Hybrid",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/2022.deeplo-1.10",
    doi = "10.18653/v1/2022.deeplo-1.10",
    pages = "90--101",
}
"""

_HOMEPAGE = "https://mlrs.research.um.edu.mt/"

_LICENSE = "Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License (CC BY-NC-SA 4.0)"

_URL = "data/"
_SHUFFLED_URL = {
    "train": os.path.join(_URL, "shuffled/train.txt"),
    "validation": os.path.join(_URL, "shuffled/validation.txt"),
    "test": os.path.join(_URL, "shuffled/test.txt"),
}
_SUBSET_URL_PATTERN = "{}/**/*.jsonl"


class KorpusMalti(datasets.GeneratorBasedBuilder):
    """Korpus Malti: General Corpora for the Maltese Language"""

    VERSION = datasets.Version("4.1.0")

    DEFAULT_CONFIG_NAME = "shuffled"

    BUILDER_CONFIGS = [
        datasets.BuilderConfig(name=DEFAULT_CONFIG_NAME,
                               version=VERSION,
                               description="The shuffled data from all subsets.",
                               ),
        datasets.BuilderConfig(name="belles_lettres",
                               version=VERSION,
                               description="Literary texts, usually published and included in the corpus by permission of the copyright holder. Unfortunately these cannot be disseminated in their integral form.",
                               ),
        datasets.BuilderConfig(name="blogs",
                               version=VERSION,
                               description="Online blog articles from specific blogs, identified in advance and known to contain text written (or human-translated into) Maltese.",
                               ),
        datasets.BuilderConfig(name="comics",
                               version=VERSION,
                               description="A small set of online information about comic books in Maltese.",
                               ),
        datasets.BuilderConfig(name="court",
                               version=VERSION,
                               description="Publicly available proceedings form the courts of Malta. Since this data contains personal identifable information, anonymisation is performed using the Maltese Administrative model from the MAPA Project (https://doi.org/10.3233/FAIA200869).",
                               ),
        datasets.BuilderConfig(name="eu_docs",
                               version=VERSION,
                               description="Miscellaneous policy documents from the European Union institutions.",
                               ),
        datasets.BuilderConfig(name="gov_docs",
                               version=VERSION,
                               description="Miscellaneous policy documents from the Government of Malta.",
                               ),
        datasets.BuilderConfig(name="government_gazzette",
                               version=VERSION,
                               description="The official, publicly available gazette of the Government of Malta. The gazzette is bilingual; only the Maltese text is included. Since this data contains personal identifable information, anonymisation is performed using the Maltese Administrative model from the MAPA Project (https://doi.org/10.3233/FAIA200869).",
                               ),
        datasets.BuilderConfig(name="law_eu",
                               version=VERSION,
                               description="Miscellaneous EU laws in their official Maltese translation, obtained via the Eur-Lex repository and including the segments of the Acquis Communautaire available in the DGT translation memory.",
                               ),
        datasets.BuilderConfig(name="law_mt",
                               version=VERSION,
                               description="Maltese laws.",
                               ),
        datasets.BuilderConfig(name="legal",
                               version=VERSION,
                               description="Miscellaneous legal text.",
                               ),
        datasets.BuilderConfig(name="nonfiction",
                               version=VERSION,
                               description="Miscellaneous nonfiction, published or unpublished. Published texts are included with the permission of the copyright holder, where relevant.",
                               ),
        datasets.BuilderConfig(name="parliament",
                               version=VERSION,
                               description="The officially released transcripts of parliamentary debates of the Maltese parliament.",
                               ),
        datasets.BuilderConfig(name="press_eu",
                               version=VERSION,
                               description="Press releases in Maltese by the European Council of Ministers, European Parliament and European Commission.",
                               ),
        datasets.BuilderConfig(name="press_mt",
                               version=VERSION,
                               description="Articles in the Maltese press, sourced primarily from the online portals of Maltese newspapers.",
                               ),
        datasets.BuilderConfig(name="speeches",
                               version=VERSION,
                               description="Miscellaneous speeches in Maltese (pre-written).",
                               ),
        datasets.BuilderConfig(name="theses",
                               version=VERSION,
                               description="Academic dissertations written in Maltese.",
                               ),
        datasets.BuilderConfig(name="umlib_oar",
                               version=VERSION,
                               description="Very broad variety of nonfiction texts which are publicly available in the University of Malta Open Access Repository. Included with help and permission from the University of Malta library.",
                               ),
        datasets.BuilderConfig(name="web_general",
                               version=VERSION,
                               description="Miscellaneous text scraped from pre-identified web pages in Maltese.",
                               ),
        datasets.BuilderConfig(name="wiki",
                               version=VERSION,
                               description="The Maltese Wikipedia dump (downloaded 26th May, 2020).",
                               ),
    ]

    def _info(self):
        if self.config.name == self.DEFAULT_CONFIG_NAME:
            features = {
                "text": datasets.Value("string"),
            }
        else:
            features = {
                "text": datasets.Sequence(datasets.Value("string")),
            }

        return datasets.DatasetInfo(
            description=_DESCRIPTION,
            features=datasets.Features(features),
            homepage=_HOMEPAGE,
            citation=_CITATION,
            license=_LICENSE,
        )

    def _split_generators(self, dl_manager):
        if self.config.name == self.DEFAULT_CONFIG_NAME:
            data_files = dl_manager.download_and_extract(_SHUFFLED_URL)
            data_split = [
                datasets.SplitGenerator(
                    name=datasets.Split.TRAIN,
                    gen_kwargs={
                        "filepath": data_files["train"],
                    },
                ),
                datasets.SplitGenerator(
                    name=datasets.Split.VALIDATION,
                    gen_kwargs={
                        "filepath": data_files["validation"],
                    },
                ),
                datasets.SplitGenerator(
                    name=datasets.Split.TEST,
                    gen_kwargs={
                        "filepath": data_files["test"],
                    },
                ),
            ]
        else:
            file_pattern = _SUBSET_URL_PATTERN.format(self.config.name)
            base_path = self.base_path or ""
            file_paths = [path.relative_to(base_path)
                          for path in Path(os.path.join(base_path, _URL)).glob(file_pattern)]

            data_files = dl_manager.download_and_extract(file_paths)
            data_split = [
                datasets.SplitGenerator(
                    name=datasets.Split.TRAIN,
                    gen_kwargs={
                        "filepath": data_files,
                    },
                ),
            ]

        return data_split

    def _generate_examples(self, filepath):
        if self.config.name == self.DEFAULT_CONFIG_NAME:
            with open(filepath, encoding="utf-8") as file:
                for key, line in enumerate(file):
                    if len(line) > 0:
                        yield key, {
                            "text": line,
                        }
        else:
            key = 0
            for path in filepath:
                with open(path, encoding="utf-8") as file:
                    for line in file:
                        data = json.loads(line)
                        yield key, {
                            "text": data["text"],
                        }
                        key += 1