On 22 September 2011 04:28, Hèctor Alòs i Font <h.a...@esperanto.cat> wrote: > >> > I never see somewhere something to indicate the position an >> > adjective must have according to a noun. >> > >> > Eg: >> > The red flower => La fleur rouge (the most frequent in french, the >> > adjective if after the noun), >> > The little flower => La petite fleur (the adjective "petit(e)" must >> > be before the noun). >> > >> > Is there something in apertium to distinguish these two cases and >> > how to do that ? >> >> It could be added as a tag, but for the small number of such >> adjectives, it would probably be better to add them as a list in >> transfer. Determiners - which includes possessives (mon, notre) and >> ordinals (premiere, troisieme) - would be treated separately to >> adjectives, so those classes of words would not need extra treatment. >> > > I'm also tempt to use this kind of distinction in the dictionaries of French > and other Romance languages. A specific type (e.g. preadj) would help the
I wouldn't call them 'preadj' because they're pre-nominal, not pre-adjectival. To me, preadj would be something like 'quasi', 'semi', 'pseudo'. > disambiguation in the translation from one of these languages (e.g. "mal" in > Catalan and Spanish). That is quite important if one translates into a > non-Romance language: as usually in closely related languages, in between > Romance languages the ambiguity tends to be the same, so it doesn't bother. > But if "petit" in French can stand only before the noun, taking the example > of Bernard, this doesn't happen e.g. in Catalan or Spanish, and in fact both > Apertium translators from Catalan and Spanish generate "[J']ai une fleur > petite" from "Tinc una flor petita"/"Tengo una flor pequeña" (In the Catalan > one the subject pronoun is not generated). (By the way, I guess that fr-ca is much less developed than fr-es; I 'ported' the fr->es rules to fr-ca, but not the es->fr rules. It's unfinished business for me, so I'd like to get it done. I think I'll pester Gema about how it should be done at the mentor summit, and implement it when I get back. > non-Aranese Occitan would behave as French, as (South?) Catalan does, > regarding Spanish) On the other side I don't have yet experience in the > translation into a Romance language, but I'm also not sure whether a generic > straightforward macro for Noun Phrases could be done for dealing with this > kind of adjectives. Jimmy has a lot more of experience than I, but I > imagined just the opposite: that a dictionary solution may be easier despite > the growing of rules I'd cost. It could be done with macros, but I wouldn't do it that way, because it would all but guarantee that only a handful of people could ever modify it. I wouldn't do it as a set of separate rules, either, because that greatly increases the chances that a future change to one will not be reflected in others. I'd do it as a set of conditions within a single rule. -- <Sefam> Are any of the mentors around? <jimregan> yes, they're the ones trolling you ------------------------------------------------------------------------------ All the data continuously generated in your IT infrastructure contains a definitive record of customers, application performance, security threats, fraudulent activity and more. Splunk takes this data and makes sense of it. Business sense. IT sense. Common sense. http://p.sf.net/sfu/splunk-d2dcopy1 _______________________________________________ Apertium-stuff mailing list Apertium-stuff@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/apertium-stuff
