I'm using the typeface Neacademia Text, and I'm trying to configure it to the specifics of the Serbian language. I want to replace the usual form of the lowercase cyrillic letter BE with the traditional Serbian form. I opened the font file in Font Forge and looked up the glyph info. Here is what I tried and what failed:

\startluacode
fonts.handlers.otf.addfeature {
name = "besub",
type = "substitution",
data = {
["uni0431"] = "uni0431.SRB.alt"
}
}
\stopluacode

\definefontfeature[default]
  [mode=node,kern=yes,
  liga=yes,dlig=yes,
  calt=yes,ss03=yes,
  case=yes,ccmp=yes,
  locl=yes,script=cyrl,language=srb,
  besub=yes]

No luck using hex codes either:

\startluacode
fonts.handlers.otf.addfeature {
name = "besub",
type = "substitution",
data = {
[0x1ab] = 0x3f8
}
}
\stopluacode"uni0431"

Interestingly the single substitutions from page 175 of the fonts manual seem to work just fine.

Any help is appreciated.

Sincerely,

Stefan