Cyrillic а = U+0430 => no suggestion
Cyrillic р = U+0440 => no suggestion
Cyrillic ф = U+0444 => suggestion
Look at the least significant bytes.
ASCII 30 = character 0 => not a letter
ASCII 40 = character @ => not a letter
ASCII 44 = character D => it's a letter
The same happens with Greek, i.e. the least significant byte is not an ASCII letter.
So this explains everything, somewhere in the code a non-utf-8 aware function (or part of code) is used.
If it's not fixed soon, I'll have a look at the code.
I got something more specific:
Cyrillic а = U+0430 => no suggestion
Cyrillic р = U+0440 => no suggestion
Cyrillic ф = U+0444 => suggestion
Look at the least significant bytes.
ASCII 30 = character 0 => not a letter
ASCII 40 = character @ => not a letter
ASCII 44 = character D => it's a letter
The same happens with Greek, i.e. the least significant byte is not an ASCII letter.
So this explains everything, somewhere in the code a non-utf-8 aware function (or part of code) is used.
If it's not fixed soon, I'll have a look at the code.