If the base glyph has two anchors (for example “top” and “top_right”), is there some way to ensure the accent is attached to the intended anchor when decomposed*?
*Not the kind of decomposition that happens inside Glyphs where you turn components to outlines, but the kind where you split an accented letter into base letter + accent glyphs.
Hmm, it has to go into the mark feature somehow. I assume you understand the inner workings of that feature in AFDKO code?
- In the worst case by opening the features.fea and adjusting the respective mark feature. But that is tedious and easy to forget because you would have to do it every time.
- Or by an additional GPOS rule that kerns the mark from the
top
position to the top_right
position, but that does not interpolate.
- Or you use
topright
(next to top
in the base glyphs) and _topright
(instead of _top
in the mark) as anchors. That should open an additional lookup in mark
.
- Or you use a GSUB rule that flips the mark for an alternate with a shifted
_top
.
Hmm.
I have a basic grasp of the mark feature, yes. The context of my question was a little misleading. Actually, I have two stacked marks (ḗ). So what I want to achieve, is – given a specific context – to attach the mark to a different anchor, e.g:
• Use the regular “top” anchor to position the macron in ē
• Use the “top_lower” anchor to position the macron in ḗ.
Is that possible? Without adding a new glyph? Keep in mind, I am only talking about decomposed rendering.
A kern wouldn’t work, right? Would it not be overridden by the mark positioning?