scripts/build_parquet.py

from __future__ import annotations

import argparse
from pathlib import Path

import pyarrow.parquet as pq
from datasets import load_dataset
from datasets.table import embed_table_storage


def main() -> int:
    parser = argparse.ArgumentParser(description="Build a viewer-friendly Parquet split with embedded images")
    parser.add_argument("--root", type=Path, default=Path(__file__).resolve().parents[1])
    parser.add_argument("--out", type=Path, default=None, help="Output parquet path")
    parser.add_argument("--row-group-size", type=int, default=100)
    args = parser.parse_args()
    root = args.root.resolve()
    out = args.out or (root / "data" / "train-00000-of-00001.parquet")
    out.parent.mkdir(parents=True, exist_ok=True)

    dataset = load_dataset(str(root), split="train")
    ordered = ["image"] + [name for name in dataset.column_names if name != "image"]
    dataset = dataset.select_columns(ordered)

    table = dataset.with_format("arrow")[:]
    table = embed_table_storage(table)

    writer = pq.ParquetWriter(str(out), table.schema)
    try:
        for batch in table.to_batches(max_chunksize=args.row_group_size):
            writer.write_batch(batch, row_group_size=args.row_group_size)
    finally:
        writer.close()

    check = pq.read_table(str(out))
    first = check.column("image")[0].as_py()
    embedded = isinstance(first, dict) and first.get("bytes") is not None
    print(
        f"wrote {out} ({out.stat().st_size / 1024**2:.2f} MiB, {check.num_rows} rows, "
        f"{check.metadata.num_row_groups if hasattr(check, 'metadata') else 'n/a'}); "
        f"images_embedded={embedded}"
    )
    return 0 if embedded else 1


if __name__ == "__main__":
    raise SystemExit(main())