Intervening combining mark below before uppercase mark

The automated combining mark script doesn’t account for cases like the Ą̊, where one possible input is A+ogonekcomb+ringcomb. The result should be A+ogonekcomb+ringcomb.case (or A+ogonekcomb.case+ringcomb.case if you also have capital variant marks below).

Speaking about the automated generation of combined glyphs: I think that in case of ogonek (which is often designed individually below each letter), it might be better to use Aogonek+ringcomb.case as a first option. I modified my GlyphData xml file to build iogonek from i+ogonek (the current solution is idotless+dotaccentcomb+ogonekcomb) to avoid the possible difference between i and i+dotaccentcomb.

This is not an encoded character, Filip. The order of which it is typed depends mostly on the keyboard at hand.

What exactly do you mean? The mark feature code?

Yes.

The ‘mark’ feature code is not the problem. The switching to the .case variants of the glyphs is done in the ccmp feature. I just did a few test. With the default ccmp code from Glyphs it worked fine (switching to the .case variant if typed after an ogonek) in Indesign, Safari and Firefox (all on Mac). Indesign is actually combining the A+ogonekcomb to Aogonek automatically.
Where did you test this? If it does not work somewhere, please add this line to the ccmp feature at the end of the latn_1 lookup.

	sub @Uppercase @CombiningNonTopAccents @Markscomb' by @MarkscombCase;

GPOS comes before GSUB. So, mark will be executed before case. And the .case variants should have their mark positioning entries. If not, can you send me the .glyphs file? res (at) (this website without www).

Georg: I tested it in Firefox on a Mac. The site is here: https://no.wikipedia.org/wiki/Ą̊ (You already have the font file, if you want to test)

@mekkablue that is not true. GSUB comes first.

I just make ligatures for /Aogonek_acutecomb etc; seems to work with copy/pasted Navajo text.

Georg, I think the issue was with the HTML. Apologies for occupying your time.

@CatharsisFonts I don’t think that is a viable method. For Navajo alone, that means 24 extra glyphs.

Are you assuming eight characters over three masters? I tend to have over 1000 glyphs per cut anyway.

Nope. You better dig deeper in the Navajo documentation.