import argparse
from pathlib import Path

import pandas as pd


def keep_latest_by_time(df: pd.DataFrame, store_col: str, time_col: str) -> pd.DataFrame:
    if store_col not in df.columns:
        raise ValueError(f"缺少列: {store_col}")
    if time_col not in df.columns:
        raise ValueError(f"缺少列: {time_col}")

    working = df.copy()
    working[store_col] = working[store_col].astype(str).fillna("")
    working[time_col] = working[time_col].astype(str).fillna("")

    working["_parsed_time"] = pd.to_datetime(working[time_col], errors="coerce")
    working["_row_order"] = range(len(working))

    working = working.sort_values(
        by=["_parsed_time", "_row_order"],
        ascending=[True, True],
        kind="mergesort",
        na_position="first",
    )
    latest = working.groupby(store_col, sort=False).tail(1)
    latest = latest.drop(columns=["_parsed_time", "_row_order"]).reset_index(drop=True)
    return latest


def build_parser() -> argparse.ArgumentParser:
    parser = argparse.ArgumentParser()
    parser.add_argument("--input", "-i", required=False, help="输入 Excel 路径（.xlsx）")
    parser.add_argument("--output", "-o", required=False, help="输出 Excel 路径（.xlsx）")
    parser.add_argument("--sheet", default="需要保留一条", help="Sheet 名称")
    parser.add_argument("--store-col", default="门店编码", help="门店编码列名")
    parser.add_argument("--time-col", default="创建时间", help="创建时间列名")
    parser.add_argument("--codes-output", required=False, help="可选：输出门店编码清单（.txt 或 .csv）")
    parser.add_argument("--demo", action="store_true", help="运行内置示例（不读写 Excel）")
    return parser


def main() -> int:
    args = build_parser().parse_args()

    if args.demo:
        demo_df = pd.DataFrame(
            [
                {"门店编码": "A001", "创建时间": "2026-03-01 10:00:00", "其他": "x"},
                {"门店编码": "A001", "创建时间": "2026-03-05 09:00:00", "其他": "y"},
                {"门店编码": "B002", "创建时间": "2026/03/02 12:00", "其他": "m"},
                {"门店编码": "B002", "创建时间": "无效时间", "其他": "n"},
            ]
        )
        result = keep_latest_by_time(demo_df, store_col=args.store_col, time_col=args.time_col)
        print(result)
        print("门店编码:", ",".join(result[args.store_col].astype(str).tolist()))
        return 0

    if not args.input:
        raise SystemExit("缺少参数 --input")

    input_path = Path(args.input).expanduser().resolve()
    if not input_path.exists():
        raise SystemExit(f"输入文件不存在: {input_path}")

    df = pd.read_excel(input_path, sheet_name=args.sheet, dtype=str).fillna("")
    latest = keep_latest_by_time(df, store_col=args.store_col, time_col=args.time_col)

    output_path = Path(args.output).expanduser().resolve() if args.output else None
    if output_path is None:
        output_path = input_path.with_name(f"{input_path.stem}_保留最新.xlsx")

    with pd.ExcelWriter(output_path, engine="openpyxl") as writer:
        latest.to_excel(writer, sheet_name="保留最新", index=False)

    store_codes = latest[args.store_col].astype(str).tolist()
    print(f"保留行数: {len(latest)}")
    print(f"门店编码数量: {len(store_codes)}")

    if args.codes_output:
        codes_path = Path(args.codes_output).expanduser().resolve()
        if codes_path.suffix.lower() == ".csv":
            pd.DataFrame({args.store_col: store_codes}).to_csv(codes_path, index=False, encoding="utf-8-sig")
        else:
            codes_path.write_text("\n".join(store_codes), encoding="utf-8")

    print(f"输出文件: {output_path}")
    return 0


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