scripts/verify_dataset.py

from __future__ import annotations

import argparse
import hashlib
import json
from pathlib import Path
import sys

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

from sparsewake.data import LABEL_NAMES, describe_h5, load_h5


def sha256(path: Path) -> str:
    digest = hashlib.sha256()
    with path.open("rb") as handle:
        for chunk in iter(lambda: handle.read(1024 * 1024), b""):
            digest.update(chunk)
    return digest.hexdigest()


def verify_checksums(manifest: Path, root: Path) -> dict[str, object]:
    checked = []
    for raw in manifest.read_text().splitlines():
        line = raw.strip()
        if not line or line.startswith("#"):
            continue
        expected, rel = line.split(None, 1)
        path = root / rel
        if not path.exists():
            raise SystemExit(f"Checksum target is missing: {rel}")
        found = sha256(path)
        if found.lower() != expected.lower():
            raise SystemExit(f"Checksum mismatch for {rel}: expected {expected}, found {found}")
        checked.append(rel)
    return {"checksum_manifest": manifest.as_posix(), "files_checked": len(checked)}


def main() -> None:
    parser = argparse.ArgumentParser()
    parser.add_argument("--data", required=True)
    parser.add_argument("--checksums", default=None)
    args = parser.parse_args()
    data = load_h5(args.data)
    required = ["X_raw", "y", "groups", "region_id", "sensor_world_positions"]
    missing = [key for key in required if key not in data]
    if missing:
        raise SystemExit(f"Missing required datasets: {missing}")
    if data["input"].ndim != 3 or data["input"].shape[1:] != (6, 3):
        raise SystemExit(f"Expected input shape [N, 6, 3], found {data['input'].shape}")
    if data["y"].shape[1] < len(LABEL_NAMES):
        raise SystemExit(f"Expected at least {len(LABEL_NAMES)} label columns, found {data['y'].shape[1]}")
    if len(set(data["pose_id"].tolist())) < 2:
        raise SystemExit("Pose-holdout verification requires at least two inferred poses")
    summary = describe_h5(args.data)
    summary["n_samples"] = int(data["y"].shape[0])
    summary["n_poses"] = int(len(set(data["pose_id"].tolist())))
    if args.checksums:
        summary["checksums"] = verify_checksums(Path(args.checksums), ROOT)
    print(json.dumps(summary, indent=2))


if __name__ == "__main__":
    main()