export_csv.py

#!/usr/bin/env python
"""
Dump the BWF cache to plain CSVs, one per discipline -- the shareable version
of the dataset for people who just want the matches and not this codebase.

    python export_csv.py          # writes data/csv/{ms,ws,md,wd,xd}.csv
"""

from pathlib import Path
from badminton import bwf

OUT = Path(__file__).resolve().parent / "data" / "csv"


def main():
    OUT.mkdir(parents=True, exist_ok=True)
    frames = bwf.load()
    for code, df in frames.items():
        out = df.copy()
        # tuples don't belong in a CSV; join pairs with ' / '
        out["team1"] = out["s1_players"].map(" / ".join)
        out["team2"] = out["s2_players"].map(" / ".join)
        out = out[["date", "discipline", "tournament", "tier", "round",
                   "host", "team1", "team2", "winner", "score",
                   "s1_home", "s2_home"]]
        out.columns = ["date", "discipline", "tournament", "tier", "round",
                       "host_location", "team1", "team2", "winner", "score",
                       "team1_at_home", "team2_at_home"]
        path = OUT / f"{code.lower()}.csv"
        out.to_csv(path, index=False, encoding="utf-8")
        print(f"{path.name}: {len(out):,} matches, {path.stat().st_size // 1024} KB")


if __name__ == "__main__":
    main()