On Sun, Jul 10, 2011 at 6:13 PM, Dinbandhu <[email protected]> wrote:
> So I'm wondering what you can suggest as a solution. Gedit does not seem to 
> "do this language detection and
> switch the dictionaries dynamically". If it does have such capability, could 
> you guide me as to how I can make it happen.

I don't closely watch Gnome development these days, But let me try to
outline the solution.

Gedit(or GTK based text editors use libenchant as the spellchecker
wrapper library. It acts as a bridge to aspell or hunspell. Remember
that Fedora and Ubuntu are planning to use only hunspell as the
spellchecker library and planning to remove Aspell in future. This is
to avoid multiple dictionaries in the system, and to avoid multiple
incompatible personal dictionaries. For eg, when you add a new word in
LibreOffice personal dictionary. That wont be available to Gedit if
both use different spellcheckers and different personal dictionaries.

You need to read the code of enchant, identify whether it is the best
place to have a language detection logic and then route it to
appropriate spellcheckers with identified language as parameter.
Language detection is not an easy task. There are languages which use
same script(Hindi/Marathi). You need to look into the libtextcat
library for this. It is a language detection library. You can also
read sonnet source code which has its own language detection
algorithm. Both uses n-gram model of the language anyway.

Along with this research, you should talk to gnome developers in their
dev mailing list about your idea and get their suggestions.

I had provided links to bug reports, idea pages related to this topic
in my blog post
http://thottingal.in/blog/2008/11/13/language-detection-and-spellcheckers/
Please read that too.

Thanks
Santhosh Thottingal
http://thottingal.in/blog

------------------------------------------------------------------------------
All of the data generated in your IT infrastructure is seriously valuable.
Why? It contains a definitive record of application performance, security 
threats, fraudulent activity, and more. Splunk takes this data and makes 
sense of it. IT sense. And common sense.
http://p.sf.net/sfu/splunk-d2d-c2
_______________________________________________
IndLinux-group mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/indlinux-group

Reply via email to