[libvoikko] Skolt Sami upper-case bug in libvoikko?

Harri Pitkänen hatapitk at iki.fi
Wed Mar 11 17:01:23 EET 2015


Hi!

On Wednesday 11 March 2015 11:42:57 Jack Rueter wrote:
> We have built a zhfst file for SMS (Skolt Sámi), and tested it in
> LibreOffice 4.2. One thing we observed was that certain words in the
> beginning of sentences were not recognised. Example:
> 
> Ǩiõl - rejected by the speller
> 
> ǩiõl - accepted by the speller
> 
> As far as we have understood, such case handling should be done by
> libvoikko, so that the speller fst should only contain lexical case. If so,
> this seems to be a bug in libvoikko (or some of its dependencies). Could
> that be?

There were no case handling rules for Latin Extended-B block that contains ǩ 
and Ǩ. I have now added rules for characters from 0x01DE to 0x01EF. Let me 
know if this is sufficient for you or if other character ranges need to be 
supported.

Harri


More information about the Libvoikko mailing list