https://bugs.documentfoundation.org/show_bug.cgi?id=173547

--- Comment #3 from Jîyan Alpîranî <[email protected]> ---
Thank you very much for investigating this issue.

Yes, the intended Kurdish (Northern / Kurmanji, Latin/Hawar) alphabetic order
is:

A, B, C, Ç, D, E, Ê, F, G, H, I, Î, J, K, L, M, N, O, P, Q, R, S, Ş, T, U, Û,
V, W, X, Y, Z

The relationships that are especially important for collation are:

C < Ç < D
E < Ê < F
I < Î < J
S < Ş < T
U < Û < V

I am not an expert in LibreOffice's internal collation-file syntax, so I cannot
say with certainty how `ku_alphanumeric.txt` should be written or how it should
be connected to the locale.

However, if the existing rules in `ku_alphanumeric.txt` are already correct but
the file is not being used, then it seems that the main problem may be the BCP
47 / locale mapping rather than the collation rules themselves.

I would be happy to test any proposed fix on macOS with LibreOffice 26.8.0.3
and provide Kurdish test data to verify the sorting behavior.

Thank you again for looking into this.

-- 
You are receiving this mail because:
You are the assignee for the bug.

Reply via email to