U+A9BF JAVANESE CONSONANT SIGN CAKRA blocks ligatures of above-base marks

[dscorbett]

Font

NotoSansJavanese-Regular.otf

Where the fonts came from, and when

Site: https://github.com/googlefonts/noto-fonts/blob/c8729a9e61b2a54a9d2b4c61a82d4d93213bd6e3/unhinted/otf/NotoSansJavanese/NotoSansJavanese-Regular.otf Date: 2022-04-20

Font version

Version 2.002

Issue

U+A9BF JAVANESE CONSONANT SIGN CAKRA, when ligated with the base, blocks ligatures between U+A9B3 JAVANESE SIGN CECAK TELU and above-base marks.

Character data

ꦥ꦳ꦿꦁ U+A9A5 JAVANESE LETTER PA U+A9B3 JAVANESE SIGN CECAK TELU U+A9BF JAVANESE CONSONANT SIGN CAKRA U+A981 JAVANESE SIGN CECAK

Screenshot

[simoncozens]

I don't understand this. Why would one ligate but not the other?

They're right next to each other, so there's no mark filtering issues. It's in abvs, which is whole run, so it's not a cluster scoping problem. Beats me.

[dscorbett]

HarfBuzz has some complex rules about which glyphs may ligate:

   * - Ligatures cannot be formed across glyphs attached to different components
   *   of previous ligatures.  Eg. the sequence is LAM,SHADDA,LAM,FATHA,HEH, and
   *   LAM,LAM,HEH form a ligature, leaving SHADDA,FATHA next to eachother.
   *   However, it would be wrong to ligate that SHADDA,FATHA sequence.
   *   There are a couple of exceptions to this:
   *
   *   o If a ligature tries ligating with marks that belong to it itself, go ahead,
   *     assuming that the font designer knows what they are doing (otherwise it can
   *     break Indic stuff when a matra wants to ligate with a conjunct,
   *
   *   o If two marks want to ligate and they belong to different components of the
   *     same ligature glyph, and said ligature glyph is to be ignored according to
   *     mark-filtering rules, then allow.
   *     https://github.com/harfbuzz/harfbuzz/issues/545

It seems like this case should work too. Maybe this is a deficiency of HarfBuzz, not Noto. What do you think?

[simoncozens]

From the sound of that, I should check the GDEF classes. The first paragraph would apply: I’m trying to ligate glyphs “left over” from “inside” a previous ligature, which that paragraph suggests is not allowed.

[simoncozens]

pa_cakra is coming out as a base glyph in GDEF, not a ligature. Stranger and stranger.

[simoncozens]

Ah! But if I force it to be a ligature and add IgnoreLigatures to the substitution, the ligature works. (But then mark attachment fails because in ufo2ft you can't do mark-to-base on a ligature glyph, only mark-to-lig, so I need to rename all the anchors to top_1 etc. sigh)