Closed albbas closed 12 years ago
Date: 2012-01-19 08:11:37 +0100
From: Trond Trosterud <
Background: The Komi-as-Latin OOo speller works fine, but cannot handle uppercasing of lowercase words.
(I do not find the reference, but) it was suggested to exchange the quasicode "la" for a cyrillic-based language. I did that, twice, for kk and be (Kazakh and Belorussian), in the Makefile.hfst:
2LCODE=be
and compiled, and generated new files in .voikko/2
~/.voikko/2$cat mor-be/voikko-fi_FI.pro info: Voikko-Dictionary-Format: 2 info: Language-Code: be info: Language-Variant: standard info: Description: Kokeellinen komi morfologia info: Morphology-Backend: null info: Speller-Backend: hfst info: Suggestion-Backend: hfst ~/.voikko/2$
Now, the problem is: Neither kazakh nor Belarussian come up with the promising green V in OpenOffice after restart.
So, have I overlooked something?
This really should be in place this week, I go to Russia next week to install this.
Date: 2012-01-19 23:27:27 +0100
From: Trond Trosterud <
The testing goes on: upper/lower works for faroese in OOo. It does not work for Komi-as-Ukrainian, even though uk is on voikko's list It also does not work for Kildin Sámi (to repeat test, take the nouns from our noun file and test the lemma form)
Thus: The problem is linked to all our cyrillic spellers, and not to the latin ones. It is not dependent upon whether the lg in question is an OOo lg or not (cf. sjd, uk).
It seems this is an issue concerning the voikko engine, which has no cyrillic casinp pairs added.
TODO: Find out and add the cyrillic pairs.
Date: 2012-02-03 16:25:46 +0100
From: Sjur Nørstebø Moshagen <
This is fixed now:
This issue was created automatically with bugzilla2github
Bugzilla Bug 1256
Date: 2012-01-19T08:11:37+01:00 From: Trond Trosterud <>
To: Sjur Nørstebø Moshagen <>
CC: borre.gaup, tomi.k.pieski
Last updated: 2012-02-03T16:25:46+01:00