On 13 August 2014 17:29, Daniel Naber <daniel.na...@languagetool.org> wrote:
> > This would need to be implemented in LO. LT doesn't know which document > is being checked, we only see the text. > Hmm... if there is consensus among the members here a bug report could be filed, I suppose... > It's supposed to use the language of the document. More exactly, the > language of the text where the cursor is. I see this doesn't work > properly for Tamil. It's a "complex text layout" language, isn't it? We > had the same problem with Khmer, where we needed to implement our own > detection to see if the text is Khmer. Can we do that for Tamil, too? > Are there Unicode ranges of characters only used by Tamil? > OK. It works as you've described above for German, French, etc. -- but not for Tamil. As I've explained in my other e-mail, Tamil uses common characters like numbers, symbols and punctuation from other areas, but the Tamil block itself starts at U+0B82 and ends at U+0BFA. Further details (if needed) are found here: https://en.wikipedia.org/wiki/Tamil_%28Unicode_block%29 -e.
------------------------------------------------------------------------------
_______________________________________________ Languagetool-devel mailing list Languagetool-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/languagetool-devel
