dataset.py

import csv
import datasets

class DatasetConfig(datasets.BuilderConfig):
    def __init__(self, **kwargs):
        super().__init__(**kwargs)

class DatasetBuilder(datasets.GeneratorBasedBuilder):
    BUILDER_CONFIGS = [DatasetConfig(name="default", version=datasets.Version("1.0.0"))]

    def _info(self):
        return datasets.DatasetInfo(
            features=datasets.Features({
                "path": datasets.Value("string"),
                "content": datasets.Value("string"),  # Adjust to your actual CSV structure
            }),
        )

    def _split_generators(self, dl_manager):
        data_path = dl_manager.download("A_python_files_elaborated_metadata.csv")
        return [
            datasets.SplitGenerator(
                name=datasets.Split.TRAIN,
                gen_kwargs={"filepath": data_path},
            )
        ]

    def _generate_examples(self, filepath):
        with open(filepath, encoding="utf-8") as f:
            reader = csv.DictReader(f)
            for idx, row in enumerate(reader):
                yield idx, row