dataset_script.py

import json
import datasets

_DESCRIPTION = "Chinese drug KG with two configs: entities and relations."

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

class DrugDataset(datasets.GeneratorBasedBuilder):
    BUILDER_CONFIGS = [
        DrugDatasetConfig(
            name="entities",
            version=datasets.Version("1.0.0"),
            description="Drug entities (code, name, type, description)"
        ),
        DrugDatasetConfig(
            name="relations",
            version=datasets.Version("1.0.0"),
            description="Relations (source, target, relationship, sourceName, targetName)"
        ),
    ]

    def _info(self):
        if self.config.name == "entities":
            features = datasets.Features({
                "code": datasets.Value("int64"),
                "name": datasets.Value("string"),
                "type": datasets.Value("string"),
                "description": datasets.Value("string"),
            })
        else:
            features = datasets.Features({
                "source": datasets.Value("int64"),
                "target": datasets.Value("int64"),
                "relationship": datasets.Value("string"),
                "sourceName": datasets.Value("string"),
                "targetName": datasets.Value("string"),
            })
        return datasets.DatasetInfo(
            description=_DESCRIPTION,
            features=features,
        )

    def _split_generators(self, dl_manager):
        if self.config.name == "entities":
            path = dl_manager.download_and_extract("data/entities.jsonl")
        else:
            path = dl_manager.download_and_extract("data/relations.jsonl")
        return [datasets.SplitGenerator(name=datasets.Split.TRAIN, gen_kwargs={"filepath": path})]

    def _generate_examples(self, filepath):
        with open(filepath, "r", encoding="utf-8") as f:
            for i, line in enumerate(f):
                line = line.strip()
                if not line:
                    continue
                yield i, json.loads(line)