GitHub/gf-rgl
1
0
Fork 1
mirror of https://github.com/GrammaticalFramework/gf-rgl.git synced 2026-06-26 03:26:27 -06:00
Code Issues Packages Projects Releases Wiki Activity

Implement contraction of certain noun phrases and prepositions, e.g. met + dat -> daarmee

Browse Source
This commit is contained in:
Inari Listenmaa
2018-01-04 23:46:29 +02:00
parent 6831710c31
commit b7ee620c8a
14 changed files with 118 additions and 91 deletions
Download Patch File Download Diff File Expand all files Collapse all files
src/dutch/ResDut.gf
+38 -19
View File
@@ -19,7 +19,7 @@ resource ResDut = ParamX ** open Prelude, Predef in {
NPCase = NPNom | NPAcc ;
oper
oper
Noun = {s : NForm => Str ; g : Gender} ;
mkNoun : (_,_ : Str) -> Gender -> Noun = \sg,pl,g -> {
@@ -380,23 +380,25 @@ param
worden_V = irregVerb2 "worden" "werd" "werden" "geworden" ** {
aux = VZijn ; prefix = [] ; particle = [] ; vtype = VAct} ;
Pronoun : Type = {
Pronoun : Type = MergesWithPrep ** {
unstressed,stressed : {nom, acc, poss : Str} ;
substposs : Str ;
a : Agr
substposs : Number => Str ;
a : Agr ;
} ;
mkPronoun : (x1,_,_,_,_,x6,x7 : Str) -> Gender -> Number -> Person -> Pronoun =
\ik,me,mn,Ik,mij,mijn,mijne,g,n,p -> {
\ik,me,mn,Ik,mij,mijn,mijne,g,n,p -> noMerge ** {
unstressed = {nom = ik ; acc = me ; poss = mn} ;
stressed = {nom = Ik ; acc = mij ; poss = mijn} ;
substposs = mijne ;
substposs = table {Sg => mijne ; Pl => mijne + "n" } ; --overgenerates *jullien /IL2018
a = {g = g ; n = n ; p = p}
} ;
het_Pron : Pronoun = mkPronoun "het" "het" "ze" "hij" "hem" "zijn" "zijne" Neutr Sg P3 ; -- cunger: 't -> het
MergesWithPrep = { mergesWithPrep : Bool ; mergeForm : Str } ;
noMerge : MergesWithPrep = { mergesWithPrep = False ; mergeForm = [] } ;
-- Complex $CN$s, like adjectives, have strong and weak forms.
param
@@ -423,8 +425,22 @@ param
---- clause, inverted, or subordinate.
oper
Preposition = Str ;
appPrep : Preposition -> (NPCase => Str) -> Str = \p,np -> p ++ np ! NPAcc ; ----
Preposition : Type = MergesWithPrep ** { s : Str } ;
-- This is a hack for appPrep: sometimes we don't really need a full NP
NPLite : Type = MergesWithPrep ** { s : NPCase => Str } ;
npLite : (NPCase => Str) -> NPLite = \nplite -> noMerge ** {s = nplite} ;
-- Applying a preposition to a noun phrase
-- In order to decide whether to merge, have to check both NP and Prep:
-- e.g. deze + met -> hiermee , but zonder + met -> "met zonder"
appPrep : Preposition -> NPLite -> Str
= \prep,np ->
case <np.mergesWithPrep,prep.mergesWithPrep> of {
<True,True> => glue np.mergeForm prep.mergeForm ;
_ => prep.s ++ np.s ! NPAcc } ;
param
Order = Main | Inv | Sub ;
@@ -685,20 +701,22 @@ param
infPart : Bool -> Str = \b -> if_then_Str b [] "te" ;
mkDet : Str -> Str -> Number -> {s,sp : Gender => Str ; n : Number ; a : Adjf} =
\deze,dit,n -> {
Determiner : Type = MergesWithPrep ** {s,sp : Gender => Str ; n : Number ; a : Adjf} ;
mkDet : Str -> Str -> Number -> Determiner =
\deze,dit,n -> noMerge ** {
s = \\g => case <n,g> of {<Sg,Neutr> => dit ; _ => deze} ;
sp = \\g => case <n,g> of {<Sg,Neutr> => dit ; _ => deze} ;
n = n ;
a = Weak
} ;
mkQuant : Str -> Str -> {
s : Bool => Number => Gender => Str ;
sp : Number => Gender => Str ;
a : Adjf
} =
\deze,dit -> {
Quantifier : Type = MergesWithPrep ** {
s : Bool => Number => Gender => Str ;
sp : Number => Gender => Str ;
a : Adjf
} ;
mkQuant : Str -> Str -> Quantifier =
\deze,dit -> noMerge ** {
s = \\_ ,n,g => case <n,g> of {<Sg,Neutr> => dit ; _ => deze} ;
sp = \\ n,g => case <n,g> of {<Sg,Neutr> => dit ; _ => deze} ;
a = Weak
@@ -709,8 +727,9 @@ param
s = \\n,g => case <n,g> of {<Sg,Neutr> => dit ; _ => deze}
} ;
mkNP : Str -> Gender -> Number -> {s : NPCase => Str ; a : Agr ; isPron : Bool} =
\s,g,n -> heavyNP {
NounPhrase : Type = MergesWithPrep ** {s : NPCase => Str ; a : Agr ; isPron : Bool } ;
mkNP : Str -> Gender -> Number -> NounPhrase =
\s,g,n -> noMerge ** heavyNP {
s = \\_ => s ;
a = agrgP3 g n ;
} ;