"""Reproduce the indexing benchmarks reported in the article.

Usage:
    python benchmark.py --repos repos.txt --runs 5 --out benchmark-results.csv
"""

from __future__ import annotations

import argparse
import csv
import statistics
import time
from pathlib import Path


def time_walk(root: Path, workers: int | None) -> float:
    """Return wall-clock seconds for one full walk of `root`."""
    start = time.perf_counter()
    # Placeholder: call the real walker under test here.
    list(root.rglob("*"))
    return time.perf_counter() - start


def main() -> None:
    parser = argparse.ArgumentParser(description=__doc__)
    parser.add_argument("--repos", type=Path, required=True)
    parser.add_argument("--runs", type=int, default=5)
    parser.add_argument("--out", type=Path, default=Path("benchmark-results.csv"))
    args = parser.parse_args()

    roots = [Path(line.strip()) for line in args.repos.read_text().splitlines() if line.strip()]

    with args.out.open("w", newline="", encoding="utf-8") as handle:
        writer = csv.writer(handle)
        writer.writerow(["repository", "config", "median_seconds", "runs"])

        for root in roots:
            for label, workers in (("sequential", None), ("parallel", 8)):
                timings = [time_walk(root, workers) for _ in range(args.runs)]
                writer.writerow([root.name, label, round(statistics.median(timings), 3), args.runs])
                print(f"{root.name:24} {label:10} {statistics.median(timings):.3f}s")


if __name__ == "__main__":
    main()