W dniu 2012-12-27 19:53, Dominique Pellé pisze: > Ruud Baars <baar...@xs4all.nl <mailto:baar...@xs4all.nl>> wrote: > > You don't have to use compounding. > > add a flag for the 'affix' 're' etc. > > Like > SFX X J 2 > SFX X e . > SFX X ere . > > (or something better) > and add numbers to the dictionary > 1/X > 2/X > 3/X > etc > > This is a lot faster than compounding. > > > But there is an infinite number of integers. > Ideally it should match any number like 99999e.
Well, I doubt that you could get a string like "124134124124234e" in natural language. First 100 is enough, I guess. Another simple way to handle such things, however, is to change the word tokenization: tokenize affixes after numbers as strings, and add the affixes to the dictionary (or the ignore file). You would then have "e", "ere" as words, and a special XML rule that would mark these as wrong if not preceded with a number (and no whitespace). Regards, Marcin ------------------------------------------------------------------------------ Master HTML5, CSS3, ASP.NET, MVC, AJAX, Knockout.js, Web API and much more. Get web development skills now with LearnDevNow - 350+ hours of step-by-step video tutorials by Microsoft MVPs and experts. SALE $99.99 this month only -- learn more at: http://p.sf.net/sfu/learnmore_122812 _______________________________________________ Languagetool-devel mailing list Languagetool-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/languagetool-devel