src/analysis/get_json_field_as_string.py

import json


def main(
    paths: list[str],
    field: list[str],
    format: str = "plain",
) -> None:
    result = []
    for path in paths:
        with open(path, "r") as f:
            data = json.load(f)
            value = data
            for key in field:
                value = value.get(key)
            if not isinstance(value, list):
                value = [value]
            result.extend(value)
    if format == "plain":
        print(",".join(map(str, result)))
    elif format == "python":
        result_str = str(result)
        print(result_str.replace(" ", ""))
    else:
        raise ValueError(f"Unknown format: {format}")


if __name__ == "__main__":
    import argparse

    parser = argparse.ArgumentParser(
        description="Get a field from one or more JSON files and print to stdout."
    )
    parser.add_argument(
        "paths",
        type=lambda x: x.split(","),
        help="Comma-separated list of paths to the JSON files to process.",
    )
    parser.add_argument(
        "--field",
        type=str,
        required=True,
        nargs="+",
        help="Field to extract from the JSON files. Can be a nested field by providing multiple entries.",
    )
    parser.add_argument(
        "--format",
        type=str,
        default="plain",
        choices=["plain", "python"],
    )

    args = parser.parse_args()
    kwargs = vars(args)
    main(**kwargs)