#!/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 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('--resume', action='store_true',
                        help='启用断点续传，按 hk_songs_test.id 从上次成功提交的批次后继续')
    parser.add_argument('--state-file', default='output/etl_to_crawler_state.json',
                        help='断点状态文件路径')
    parser.add_argument('--start-after-id', type=int, default=None,
                        help='手动指定从哪个 hk_songs_test.id 之后开始读取')
    parser.add_argument('--reset-state', action='store_true',
                        help='忽略已有断点状态，从头或 --start-after-id 指定位置重新开始')
    args = parser.parse_args()

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

    print(f"Starting ETL for platforms: {platforms}")
    run(
        platforms,
        max_batches=args.max_batches,
        resume=args.resume,
        state_file=args.state_file,
        state_key=args.platform,
        start_after_id=args.start_after_id,
        reset_state=args.reset_state,
    )