Auto-building CCMP feature


I just tried generating the CCMP feature automatically and found somehow the lookups are quadruplicated:

I was actually hoping it could automatically add rules to compose Latin accented characters but there’s no sign of them, is that expected behaviour? For example R + ringbelowcomb + macroncomb, is there a way for Glyphs to write the rules automatically to compose that?

EDIT also to write

sub [i j]' @CombiningTopAccents by [dotlessi dotlessj];

I thought that might be automatic too?

Can you send me the file that produces this 4 lookups.

Is the R_ringbelowcomb_macroncomb any different than the three components stuck together? Why do you have a precomposed glyph for it?

It should generate the i j line (if you have all the glyphs).

Actually you’re right maybe I don’t need to have a precomposed R̥̄but I like to have it there to be sure it’s going to get composed correctly (in this website font it’s not composing correctly). But certainly for Mcedilla I need a precomposed version as I want it to look different to the components stuck together.

Just realised I used nonstandard names for dotlessi, dotlessj, they should be idotless and jdotless if this is to work.

Will send the file

I fixed the duplicate lookups and the update is up already.

Lovely jubbly! Thanks Georg.

1 Like

I just tried automatically generating CCMP in a font that contains .case versions of both single marks and double marks for Vietnamese.

Here’s what it generated:

If I’m not mistaken, the first two lookups will prevent the last two lookups from working, because any sequences of two marks will have been replaced by .case versions (by lines 8 and 13).

The other problem is it’s not substituting the .case versions of double marks (e.g, anywhere.

As you know, I tend not to use auto code, but I know some users are likely to use that function and expect not to have to get involved in OT rules.

Descriptive names for the lookups would really be helpful too.

Thanks. I’ll have a look.

1 Like