cc_news_pt.py

import json
import datasets
from typing import Dict, List, Optional, Union, Callable
import textwrap
import gzip


logger = datasets.logging.get_logger(__name__)

_CC_NEWS_PT_KWARGS = dict(
    name = "cc_news_pt",
    description=textwrap.dedent(
        """\
        CC-News-PT is a curation of news articles from CommonCrawl News in the Portuguese language.
        CommonCrawl News is a dataset containing news articles from news sites all over the world
        The data is available on AWS S3 in the Common Crawl bucket at /crawl-data/CC-NEWS/.
        This version of the dataset is the portuguese subset from
        https://huggingface.co/datasets/CloverSearch/cc-news-mutlilingual"""
    ),
    data_urls=[
        'https://huggingface.co/datasets/CloverSearch/cc-news-mutlilingual/resolve/main/2016/pt.jsonl.gz',
        'https://huggingface.co/datasets/CloverSearch/cc-news-mutlilingual/resolve/main/2017/pt.jsonl.gz',
        'https://huggingface.co/datasets/CloverSearch/cc-news-mutlilingual/resolve/main/2018/pt.jsonl.gz',
        'https://huggingface.co/datasets/CloverSearch/cc-news-mutlilingual/resolve/main/2019/pt.jsonl.gz',
        'https://huggingface.co/datasets/CloverSearch/cc-news-mutlilingual/resolve/main/2020/pt.jsonl.gz',
        'https://huggingface.co/datasets/CloverSearch/cc-news-mutlilingual/resolve/main/2021/pt.jsonl.gz',
    ],  
    citation=textwrap.dedent(
        """\
        @misc{Acerola2023,
            author     = {Garcia, E.A.S.},
            title      = {Acerola Corpus: Towards Better Portuguese Language Models},
            year       = {2023},
            doi        = {10.57967/hf/0814}
        }"""
    ),
    url="https://huggingface.co/datasets/eduagarcia/cc_news_pt",
)


class AcerolaConfig(datasets.BuilderConfig):
    """BuilderConfig for Acerola."""

    def __init__(
        self,
        data_urls: Dict[str, str],
        citation: str,
        url: str,
        file_type: Optional[str] = None, #filetype (csv, tsc, jsonl)
        **kwargs
    ):
        """BuilderConfig for Acerola.
        Args:
        **kwargs: keyword arguments forwarded to super.
        """
        super(AcerolaConfig, self).__init__(version=datasets.Version("1.0.3", ""), **kwargs)
        self.data_urls = data_urls
        self.citation = citation
        self.url = url
        self.file_type = file_type


def _get_ccnews_features(config: AcerolaConfig):
    return datasets.Features(
        {
            "title": datasets.Value("string"),
            "text": datasets.Value("string"),
            "authors": datasets.Value("string"),
            "domain": datasets.Value("string"),
            "date": datasets.Value("string"),
            "description": datasets.Value("string"),
            "url": datasets.Value("string"),
            "image_url": datasets.Value("string"),
            "date_download": datasets.Value("string")
        }
    )

class Acerola(datasets.GeneratorBasedBuilder):
    """CC-News dataset."""

    BUILDER_CONFIGS = [
        AcerolaConfig(
            **_CC_NEWS_PT_KWARGS
        )
    ]

    def _info(self) -> datasets.DatasetInfo:
        features = _get_ccnews_features(self.config)
        
        return datasets.DatasetInfo(
            description=self.config.description,
            homepage=self.config.url,
            citation=self.config.citation,
            supervised_keys=None,
            features=features
        )

    def _split_generators(self, dl_manager):
        data_urls = self.config.data_urls.copy()
        filepaths = dl_manager.download(data_urls)

        return [
            datasets.SplitGenerator(name=datasets.Split.TRAIN, gen_kwargs={"filepaths": filepaths}),
        ]

    def _generate_examples(self, filepaths):
        id_ = 0
        for filepath in filepaths:
            with gzip.open(filepath, "rt", encoding="utf-8") as f:
                for line in f:
                    if line:
                        article = json.loads(line)
                        yield id_, {
                            "title": article["title"].strip() if article["title"] is not None else "",
                            "text": article["maintext"].strip() if article["maintext"] is not None else "",
                            "authors": "; ".join([a.strip() for a in article["authors"]]) if len(article["authors"]) > 0 else "",
                            "domain": article["source_domain"].strip() if article["source_domain"] is not None else "",
                            "date": article["date_publish"].strip() if article["date_publish"] is not None else "",
                            "description": article["description"].strip() if article["description"] is not None else "",
                            "url": article["url"].strip() if article["url"] is not None else "",
                            "image_url": article["image_url"].strip() if article["image_url"] is not None else "",
                            "date_download": article["date_download"].strip() if article["date_download"] is not None else "",
                        }
                        id_ += 1