I don't think I can use IGNORE as hunspell will ignore all of those letters and sometimes (quite often actually) there is a Latin letter inserted into Ukranian word that I need to detect (some of the Cyrillic letters look like Latin ones and people forget to switch keyboard layouts...).

I was able to use ICONV tag by converting all Latin letters to numbers to make hunspell ignore words that only include those. Although it's quite ugly (no regexp support) it works pretty well.


On 07/03/2013 04:54 PM, Sahand.T wrote:

You can add "IGNORE abcdefgh...wxyz" to the affix file to get this behaviour.

Den 3 jul 2013 16:35 skrev "Andriy Rysin" <arysin@gmail.com>:
Is there some (easy) way in hunspell to ignore all words that don't
have any characters from the target language alphabet?
E.g. Ukrainian is using Cyrillic alphabet and the dictionary does not
contain any words with latin letters, but the texts in Ukrainian often
include words written in Latin (English, Latin etc). It would be nice
to make hunspell to ignore those when checking Ukrainian text by
telling it "only check words with letters [а-яіїєґ] in them".


This SF.net email is sponsored by Windows:

Build for Windows Store.

Hunspell-devel mailing list