Unicode composing accent in cyrillic

[Fuco1]

Hey!

I don't know if the problem is with the font or my application (Emacs), but I get some odd behaviour with combining unicode accents.

It seems to depend on the size and then either lowercase a or e changes to uppercase. Observe:

Emacs is usually very good when it comes to unicode support, it can combine scripts such as devanagari or bengali which are very complex, so I'm not sure what the issue here is. If anyone can give some hints it would be much appreciated ❤️

[frankrolf]

That’s right, I can reproduce this problem in InDesign.

1. compose sequence е́ а́ é á x́ Е́ А́
2. change to small caps

Pre-composed small caps work, small caps w/ anchor attachment don’t.
This issue won’t appear a lot in Latin context, because many Latin glyphs have a pre-composed accented version. However, not all Cyrillic glyphs do.

That said, the mark feature is still somewhat experimental in Source Sans Pro, but it’s worth looking into fixing this. I suspect the mark feature writer needs to be run with some more flags enabled.
FWIW, Source Serif has the same problem.

[pauldhunt]

@Fuco1 can you please provide a full list of Cyrillic glyphs that can take accents?

[Fuco1]

@pauldhunt I think these should be all

а́ е́ и́ о́ у́ ы́ э́ ю́ я́ 

[Fuco1]

Well, plus the upper case versions

А́ Е́ И́ О́ У́ Ы́ Э́ Ю́ Я́

[erniemarch]

checked in build 2.036 roman. these all look good.

[Fuco1]

@erniemarch Where can I find the build? The latest release on github I see is 2.020

[miguelsousa]

@Fuco1 the new version is being tested now. We plan to release it soon.

[miguelsousa]

Fixed in version 2.040.

[Fuco1]

I've installed the new fonts and now the lower-case letters are always combined with the accent and turn into upper-case.

This is the emacs info on the character

             position: 8 of 10 (70%), column: 0
            character: е (displayed as е) (codepoint 1077, #o2065, #x435)
    preferred charset: unicode (Unicode (ISO10646))
code point in charset: 0x0435
               script: cyrillic
               syntax: w 	which means: word
             category: .:Base, L:Left-to-right (strong), Y:2-byte Cyrillic, c:Chinese, h:Korean, j:Japanese, y:Cyrillic
             to input: type "je" or "e" with cyrillic-translit input method
          buffer code: #xD0 #xB5
            file code: #xD0 #xB5 (encoded by coding system utf-8-unix)
              display: composed to form "е́" (see below)

Composed with the following character(s) "́" using this font:
  xft:-ADBO-Source Sans Pro-normal-normal-normal-*-57-*-*-*-*-0-iso10646-1
by these glyphs:
  [0 1 0 1499 27 5 25 30 0 nil]
  [0 1 0 1854 0 -16 -4 43 -47 [-12 6 0]]

[frankrolf]

For posterity (after discussion with Krista Radoeva):
In Russian, the ё is also a vowel that may be stressed.
In Ukrainian, vowels є і and ї can take a stress mark.
In the context of textbooks for children, ъ ь may take a stress mark as well.

[frankrolf]

@Fuco1 I can’t reproduce the regression you mentioned above. If you still can, please open a new issue.