Question regarding title extraction

[unsleepy22]

In metadata.py and related tests I see that title wrapped in h1 in the body is preferred even when a non-empty title exists in meta head, this should be intentional right? In some cases this may produce bad case, like https://mozi.ustc.edu.cn/detail/1106, is it possible to compare the length of h1 title and head title to choose the longer one?

[adbar]

@unsleepy22 Thanks for the suggestion, the heuristics for titles could be improved. The problem with your approach is when the head title also entails the name of the website or keywords for SEO purposes, then it will be longer.