When I add a word into the user dictionary, such as "meditator", I expect these to be accepted as well: * Meditator * MEDITATOR * meditator’s * Meditator’s Eudora (a venerable email client) is smart enough to not flag such variants as spelling errors, once the 'base word' is added in the user dictionary. When "Meditator" is added instead, "Meditator’s" is accepted too, but not "meditator". I think Eudora's spellchecker is very well thought out. The code is now open source. (I say that thinking that it might be useful.) I can understand why "meditators" should not be included, but the above are safe to be accepted as correct.
There can not be any English-specific features in international spellchecker.
But spell-chekers are language-specific, aren't they? Besides, wouldn't it be possible to add this as an option? As it is, it seems rather unpolished. I believe that implementing this would be very much welcomed.
This is now partially fixed, as shown at Bug 37954. Now, it would be great if words ending with "’s" would not be flagged as well if they are in an active custom/user dictionary.
Sounds like a valid enhancement request...despite myself not wanting such a "smart" system ;) Moving to NEW.
@Kumara - the more I think about this the harder it is to implement. In particular the possessive stuff: For instance, what if someone adds a word that is not a noun, so there is no possessive element, then they want to write that word in plural but accidentally type 's instead of just an "s" - spell check doesn't show this as wrong despite it clearly being wrong. Can you suggest how we might handle situations like this? I think it could lead to some relatively bad problems with spell check. Maybe something that needs to be discussed with UX team as well.
This should be closed as WONTFIX.
(In reply to Joel Madero from comment #5) > @Kumara - the more I think about this the harder it is to implement. In > particular the possessive stuff: > > For instance, what if someone adds a word that is not a noun, so there is no > possessive element, then they want to write that word in plural but > accidentally type 's instead of just an "s" - spell check doesn't show this > as wrong despite it clearly being wrong. Thanks for thinking about this. On the scenario you raised, I would count it as a grammar problem, rather than a spelling one. I think the occurrence would be rare. Besides, we can leave this to LanguageTool, which I'm playing a minor role in. If we leave this as it is, users who add words not in the dictionary (and they are often nouns) would have to add many more words just because of the "apostrophe s". For comparison sake, Eudora (the old email client which I'm *still* using) has an amazingly well thought-out spell-checker. It does what I suggested, and to my knowledge no one has ever complained about it.