The algorithm to compute the derived property should use EAW more
This issue is a continuation of https://github.com/kojiishi/unicode-auto-spacing/issues/16, originally raised by @asmusf.
@asmusf The algorithm to compute the property was adjusted to:
- Add by scripts and script_extensions.
- Remove by EAW.
In other words, the algorithm is an intersection of scripts and EAW.
Does it make clear that EAW plays one of the primary roles for this property?
With https://github.com/unicode-org/unicodetools/issues/768, ea=N|Na are exlucded from the W set, unless their script is one of CJK.
The current algorithm to compute the property is here. It's in Python with a library, but I hope it's readable without prior knowledge.
I think the changes so far have addressed the feedback. Please let me know if any. Thank you for the feedback.