#!/usr/bin/env python3
import argparse
from etl_to_crawler.config import PLATFORM_QQ, PLATFORM_KUGOU, PLATFORM_NETEASE, PLATFORMS
from etl_to_crawler.runner import backfill_yinyan_record_platforms, initialize_yinyan_song_records, run

PLATFORM_MAP = {
    'qq': PLATFORM_QQ,
    'kugou': PLATFORM_KUGOU,
    'netease': PLATFORM_NETEASE,
}

if __name__ == '__main__':
    parser = argparse.ArgumentParser(description='ETL: hk_songs_test → crawler_dev')
    parser.add_argument('--platform', default='all',
                        choices=['qq', 'kugou', 'netease', 'all'],
                        help='要导入的平台（默认 all）')
    parser.add_argument('--max-batches', type=int, default=None,
                        help='最多处理多少批次（冒烟测试用）')
    parser.add_argument('--init-yinyan-records', action='store_true',
                        help='只初始化 yinyan_song_records 待导入状态表，不执行 crawler 导入')
    parser.add_argument('--backfill-yinyan-platforms', action='store_true',
                        help='只回填 yinyan_song_records 中为空的 platform 平台代码')
    args = parser.parse_args()

    if args.platform == 'all':
        platforms = PLATFORMS
    else:
        platforms = [PLATFORM_MAP[args.platform]]

    print(f"Starting ETL for platforms: {platforms}")
    if args.backfill_yinyan_platforms:
        backfill_yinyan_record_platforms(max_batches=args.max_batches)
    elif args.init_yinyan_records:
        initialize_yinyan_song_records(platforms, max_batches=args.max_batches)
    else:
        run(platforms, max_batches=args.max_batches)