Consider exploring Intl.Segmenter (and current polyfill) for word-based selection

[BigBlueHat]

This proposal (and its current polyfill) attempt to standardize Unicode word segmentation:
https://github.com/tc39/proposal-intl-segmenter

As I recall this was something both the Web Annotation WG and Hypothes.is (at least) struggled with determining the best method of handling due to JavaScript's sub-par Unicode support "back then." This, perhaps, will improve that situation at least in terms of segmentation.

[tilgovi]

See also WICG/scroll-to-text-fragment#42

[Treora]

I added this topic to the new Tech Radar page on the wiki; updates can be added there, and discussion may also continue in this issue; but I will close it as there is no current activity on this and no particular action to take.