Working on Gurung Khema
Opened this issue · 5 comments
As Universal Shaping Engine relies on Unicode databases, I’m making it in every wrong way possible. But what I’ve made now. Base+medial are programmatically created ligatures, virama (IDK whether it exists) is tied to medial for simplicity.
Remain: med. R and contextual rules for it (at least it will tell C from Ch), the second contextual shape of med. V/Y for really sloped letters like RV or BhV, and digits.
More nice words. For b and inverted T there are special contextual versions of medial H (they are magnets for that letter somehow, and do not bear imperfections)
Did completely, though 1) let’s wait for Universal Engine; 2) IDK whether digits are cool.
We can lean on Harfbuzz to add the new USE scripts.
With Universal Engine Makasar is just two rules for “kiki” or “dudu” :)
Firefox does react to script “nagm”, my font has special raised hyphen for always-capital scripts.
My Gurung rules are DFLT{dflt} latn{dflt} and sometimes gukh{dflt} (unless fallback), Firefox I tested in considers this text Latin.