Update 0x10450-shavian.json

[jdonland]

Correct two miscategorizations, change categories to better reflect usual sort of Shavian letters.

[tomer]

These files are auto-generated, so unless there is a problem on the generation code, I afraid there is no option to fix. Have you compared the data available here with the Unicode tables?

[jdonland]

Unicode doesn't distinguish between categories of Shavian letters (consonant vs vowel or anything else). You have ROAR and NUN categorized as vowels but they are consonants. And in any case it is more useful to categorize Shavian letters as tall/deep, short, and ligature rather than consonant and vowel. Something has gone wrong in whatever generates your categories.

…

On Sat, 20 Jan 2018 at 19:31 Tomer Cohen ***@***.***> wrote: These files are auto-generated, so unless there is a problem on the generation code, I afraid there is no option to fix. Have you compared the data available here with the Unicode tables? — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub <#23 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AMvRmbYxZ7VTWrXklL96Y-CU3s0KClSrks5tMoVcgaJpZM4RlpEw> .