More structured headings
clason opened this issue · comments
Currently, the (h1) nodes and friends capture the whole "header" structure, including the section separators and any tags. It would be better to separately capture
- any separators;
- the actual heading
- any tags.
This would allow extracting the headings for a better TOC function (and for possible normalization in a linting job) as well as concealing the separators (once we have line anticonceal).