51 lines
1.3 KiB
Python
51 lines
1.3 KiB
Python
import asyncio
|
|
import sys
|
|
|
|
import cmd_arg
|
|
import config
|
|
import db
|
|
from base.base_crawler import AbstractCrawler
|
|
from media_platform.bilibili import BilibiliCrawler
|
|
from media_platform.douyin import DouYinCrawler
|
|
from media_platform.kuaishou import KuaishouCrawler
|
|
from media_platform.weibo import WeiboCrawler
|
|
from media_platform.xhs import XiaoHongShuCrawler
|
|
|
|
|
|
class CrawlerFactory:
|
|
CRAWLERS = {
|
|
"xhs": XiaoHongShuCrawler,
|
|
"dy": DouYinCrawler,
|
|
"ks": KuaishouCrawler,
|
|
"bili": BilibiliCrawler,
|
|
"wb": WeiboCrawler
|
|
}
|
|
|
|
@staticmethod
|
|
def create_crawler(platform: str) -> AbstractCrawler:
|
|
crawler_class = CrawlerFactory.CRAWLERS.get(platform)
|
|
if not crawler_class:
|
|
raise ValueError("Invalid Media Platform Currently only supported xhs or dy or ks or bili ...")
|
|
return crawler_class()
|
|
|
|
async def main():
|
|
# parse cmd
|
|
await cmd_arg.parse_cmd()
|
|
|
|
# init db
|
|
if config.SAVE_DATA_OPTION == "db":
|
|
await db.init_db()
|
|
|
|
crawler = CrawlerFactory.create_crawler(platform=config.PLATFORM)
|
|
await crawler.start()
|
|
|
|
if config.SAVE_DATA_OPTION == "db":
|
|
await db.close()
|
|
|
|
|
|
if __name__ == '__main__':
|
|
try:
|
|
# asyncio.run(main())
|
|
asyncio.get_event_loop().run_until_complete(main())
|
|
except KeyboardInterrupt:
|
|
sys.exit()
|