Localisation

Our projects are run in UTF-8, since the spring of 2005.

There are still some child diseases, they are documented in Bugzilla and fixed as they appear. Some of them are harder to fix than others, but all in all we are satisfied with our UTF-8 migration

Northern Sámi

The Northern Sámi parser is localised with UTF-8, as it should be. Note that we in the alphabet section, as a legacy from the old Latin 1 -days, have defined all the Latin 1 letters, but not other letters. At that time it was a sensible act, now the set looks a bit ad hoc (see twol-sme.txt)

Lule Sámi

Lule Sámi can adequately be represented in 8-bit format, by Latin 1 (Lule Sámi files can use ñ for n-acute). In order to have only one localisation interface for the languages we work on, Lule Sámi has been moved to UTF-8 as well. There is a script, spell-relax.regex, which makes it possible to use both ñ, ń and ŋ.

Southern Sámi

Also Southern Sámi can be represented in 8-bit format, by Latin 1, but just as for Lule Sámi, it is stored in UTF-8. The script spell-relax.regex is used for Southern Sámi as well, but with a slightly different purpose: It is used to accept the wide-spread sloppy use of i for ï.

Inari and Skolt Sámi

Inari Sámi and Skolt Sámi are still represented in the old 8-bit format, with the vowels directly in 8-bit Latin 1 and the consonants as digraphs c1 etc. This is of course not a feasible solution, especially not for Skolt Sámi. The only reason why they are not migrated is that we have not been working on these languages after the first 3-year period (2000-2003) ended. Should we want to take up our work on these languages, UTF-8 localisation will be the first step to take.

Last modified: $Date: 2006/06/30 21:59:50 $, by $Author: trond $