vukuzenzele-sentence-aligned.py

"""TODO: Add a description here."""


import csv
import json
import os

import datasets

from itertools import combinations

LANGUAGES = ['afr', 'eng', 'nbl', 'nso', 'sot', 'ssw', 'tsn', 'tso', 'ven', 'xho', 'zul']
LANGUAGE_PAIRS = list(combinations(LANGUAGES, 2))

_CITATION = """\
@dataset{marivate_vukosi_2023_7598540, author = {Marivate, Vukosi and Njini, Daniel and Madodonga, Andani and Lastrucci, Richard and Dzingirai, Isheanesu Rajab, Jenalea}, title = {The Vuk'uzenzele South African Multilingual Corpus}, month = feb, year = 2023, publisher = {Zenodo}, doi = {10.5281/zenodo.7598539}, url = {https://doi.org/10.5281/zenodo.7598539} }
"""

_DESCRIPTION = """\
The dataset contains editions from the South African government magazine Vuk'uzenzele. Data was scraped from PDFs that have been placed in the data/raw folder. The PDFS were obtained from the Vuk'uzenzele website.
"""

_HOMEPAGE = "https://arxiv.org/abs/2303.03750"

_LICENSE = "CC 4.0 BY"

_URL = "https://raw.githubusercontent.com/dsfsi/vukuzenzele-nlp/master/data/opt_aligned_out/"

class VukuzenzeleMonolingualConfig(datasets.BuilderConfig):
    """BuilderConfig for VukuzenzeleMonolingual"""

    def __init__(self, **kwargs):
        """BuilderConfig for Masakhaner.
        Args:
          **kwargs: keyword arguments forwarded to super.
        """
        super(VukuzenzeleMonolingualConfig, self).__init__(**kwargs)


# TODO: Name of the dataset usually matches the script name with CamelCase instead of snake_case
class VukuzenzeleMonolingual(datasets.GeneratorBasedBuilder):
    """TODO: Short description of my dataset."""

    VERSION = datasets.Version("1.0.0")
    BUILDER_CONFIGS = []

    for pair in LANGUAGE_PAIRS:
        name = "aligned-{}-{}.jsonl".format(pair[0], pair[1])
        description = "Vukuzenzele {}-{} aligned dataset".format(pair[0], pair[1])
        BUILDER_CONFIGS.append(datasets.BuilderConfig(name=f"{name}", version=VERSION, description=f"{description}"),)
    
    def _info(self):
        features = datasets.Features(
            {
                "src": datasets.Value("string"),
                "tgt": datasets.Value("string"),
                "score": datasets.Value("float"),
            }
        )
        return datasets.DatasetInfo(
            description=_DESCRIPTION,
            features=features,  
            homepage=_HOMEPAGE,
            license=_LICENSE,
            # Citation for the dataset
            citation=_CITATION,
        )

    def _split_generators(self, dl_manager):
        
            

        urls = {
            "train": f"{_URL}{self.config.name}"
        }
        data_dir = dl_manager.download_and_extract(urls)
        return [
            datasets.SplitGenerator(
                name=datasets.Split.TRAIN,
                gen_kwargs={
                    "filepath": data_dir["train"],
                    "split": "train",
                },
            ),
        ]

    def _generate_examples(self, filepath, split):
        with open(filepath, encoding="utf-8") as f:
            for key, row in enumerate(f):
                data = json.loads(row)
                yield key, {
                    "src": data["src"],
                    "tgt": data["tgt"],
                    "score": data["score"],
                }