(Pes) WIP: more fine-grained rules about definiteness in NPs

This commit is contained in:
Inari Listenmaa
2019-05-23 14:07:34 +02:00
parent ecd686df78
commit ec4dac68d0
4 changed files with 14 additions and 8 deletions
+2
View File
@@ -19,6 +19,7 @@ concrete NounPes of Noun = CatPes ** open ResPes, Prelude in {
Clitic => detStr ++ cn.s ! num ! Clitic ++ cn.compl ! det.n ; -- In RGL this is only for no_Quant. If other determiners with different word order take Clitic, will change. /IL
x => cn.s ! num ! x ++ detStr ++ cn.compl ! det.n } ; -- Ezafe or Poss: comes after noun, Mod table is overwritten to the given form
a = agrP3 det.n ;
takesYeAsComp = False ; -- TODO: some rule with det.isDef and cn.hasAdj
isNeg = det.isNeg ; -- affects polarity in Cl: "*nothing is real" -> "nothing isn't real"
} ;
@@ -111,6 +112,7 @@ concrete NounPes of Noun = CatPes ** open ResPes, Prelude in {
MassNP cn = emptyNP ** cn ** {
s = \\m => cn.s ! Sg ! m ++ cn.compl ! Sg ;
a = agrP3 Sg ;
takesYeAsComp = cn.hasAdj ;
} ;
UseN,