import datasets

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

    BUILDER_CONFIGS = [
        datasets.BuilderConfig(name="subset1", description="First subset"),
    ]

    def _split_generators(self, dl_manager):
        import glob
        
        data_dir = self.config.data_dir
        return [
            datasets.SplitGenerator(
                name="split1",
                gen_kwargs={"file_paths": sorted(glob.glob(f"{data_dir}/split1-*.parquet"))},
            ),
            datasets.SplitGenerator(
                name="split2",
                gen_kwargs={"file_paths": sorted(glob.glob(f"{data_dir}/split2-*.parquet"))},
            ),
        ]

    def _generate_examples(self, file_path):
        import pandas as pd

        df = pd.read_parquet(file_path)
        for idx, row in df.iterrows():
            yield idx, row.to_dict()