写了一个dashboard UI用python和pandas作为客户端侧类别规则。
现有分类方法没有办法匹配web url，而且也没有办法仅对window title进行匹配。我基于AW的数据库使用streamlit编写了一个dashboard UI，它可以编写python和pandas的函数来更详细地控制类别；这是一个client-side的分类，不会对原数据库产生影响。它提供日期和bucket过滤，可以显示TimeLine、Sunburst Chart等，并可用于生成每周和每月的个人活动报告。
The existing categorize method cannot match the web url, and it is also impossible to match only the window title. I wrote a dashboard UI based on AW’s database using streamlit, which can write python and pandas functions to control the categories in more detail. This is a client-side classification and will not affect the original database. It offers date and bucket filtering, can display timelines, sunburst charts, etc., and can be used to generate weekly and monthly reports on personal activities.