"""Build the cleaned dataset with both split schemes.

    python scripts/01_build_dataset.py
"""

import sys
from pathlib import Path

sys.path.insert(0, str(Path(__file__).resolve().parents[1] / "src"))

from cti_attack import data  # noqa: E402


def main() -> None:
    records, labels, stats = data.build(verbose=True)
    print()
    data.write(records, labels, stats)
    print(f"\n{len(labels)} techniques retained")


if __name__ == "__main__":
    main()