更长的推特

10/10 我要先尝试如何做身份验证

10/11 prototype: basic craw

git branch:

• main(production)
• develop

Crawler for single page:

• input: which site to craw( or an list of sites to craw )
• doing: craw and ask gpt output: craw result fill the db

Workers that control crawler for a site:

• provided a list of pages to Crawler for single page(can i run it simultaneously?)
• Cron-like-job-control

The output data format ( Article Table )

• Title
• Id
• Link
• Content
• ai_summary
• text_content
• update_time
• lang
• submit_user
• rating

How to profit?:

• Ads

User Table

• Id
• Email
• Name
• Password ( can I store it in a readable way instead of encrypt? )
• Reputation
• isPrivilege

Comment Table

• Id
• article_id
• Rating
• How to make this table clean?有没有一种方法整理row的，根据某一个column

Auth will use supabase auth