%%%  -*- Mode: PROLOG; Package: PROLOG-USER -*-

%%% SAMPLE ARABIC LEXICON
%%%
%%% (c) 2001 Sandiway Fong, NEC Research Institute, Inc.
%%%

no caseFoldInput.

default_features(n,[grid([],[]),case(_),theta(_)]).
default_features(v,[index(_)]).

block_default_features(nounFeatures,2).
block_default_features(verbToNounFeatures,2).

no superClass(_,_).
no blockContraction(_).

%%% EXPORT
%%%	default_features(C,Fs)		used in lexicon compilation
%%%	term(C)				terminals
%%%	lexicon(Word,C,Fs)		Word has category label C
%%%					and feature list Fs
%%%	probeLexicon(Word)		Word is in lexicon
%%%	vMorphToFs(Base,Form,Features)	TNS/AGR features
%%%	inf(Verb,Type)
%%%	relevant(C)			constraints imposed by markers
%%%					apply to C
%%% REFERENCES
%%%	(list processing)		utilities

term(n). term(a). term(v). term(p). term(c). term(mod).
term(neg). term(mrkr). term('$').

%%% 	Most lexical entries are stored directly as
%%%
%%%		 lex(Word,Category,Features)
%%%
%%%	Non-base forms require inference: 
%%%	1. plural nouns		all features except agr(_) inherited
%%%				from the sg. form
%%%	2. nominalized verbs	inherits verb features except morph(_,_)
%%%	3. non-base verb forms	all features except morph(_,_) inherited
%%%				from the base form

lexicon(Word,C,Fs) :- lex(Word,C,Fs).

% deterministic
probeLexicon(Word) :- lex(Word,_,_), !.

%%% NEGATION

lex(laa,neg,[polarity(-),eng(not)]).
lex(maa,neg,[polarity(-),eng(not)]).

%%% PREPOSITIONS

lex('Qilaa',p,[grid([],[destination]),predicate(destination),
	  adjR([goal(vpAllowExt(destination,X),X)]),eng(to)]).

%% Adverbs

% lex(how,adv,[adjoin(right),wh(+),predicate(manner)]).
% lex(when,adv,[adjoin(right),wh(+),predicate(time)]).
% lex(why,adv,[adjoin(right),wh(+),predicate(reason)]).
% lex(often,adv,[adjoin(left),predicate(manner)]).
% lex(completely,adv,[adjoin(_),predicate(manner)]).

%% Determiners as markers attaching to nouns on the right

lex(al,mrkr,[right(n,[],[prefix(al),def(+)])]).
lex(d,mrkr,[right(n,[],[prefix(d),def(+)])]).
lex(l,mrkr,[right(n,[],[prefix(l),def(+)])]).
lex(n,mrkr,[right(n,[],[prefix(n),def(+)])]).
lex('S',mrkr,[right(n,[],[prefix('S'),def(+)])]).
lex(r,mrkr,[right(n,[],[prefix(r),def(+)])]).
lex(t,mrkr,[right(n,[],[prefix(r),def(+)])]).

% lex('Qayy',det,[wh(+),eng(which)]).
% lex(kull,det,[op(+),eng(every)]).

%% Nouns
%% 	Obligatory features
%% 	1. Anaphoric: a(+/-), Pronominal: p(+/-) for Binding Theory
%% 	2. Agreement: agr([Per,Num,Gen]) 
%%	3. O-grid: grid(Ext,Int) 
%% 	Optional features
%% 	1. Morphologically realized Case: morphC(C)
%%	2. Quantifiers: op(+/-) 
%%	3. Wh-nouns: wh(+)

% Numbers and Ordinals

% lex(one,n,[morphC(gen),count(_),agr([3,sg,[]]),nonarg(+)]).
% lex(first,n,[morphC(gen),count(+),agr([3,sg,[]]),nonarg(+)]).

% Pleonastic elements

lex(there,n,[theta(X),agr([3,[],[]]),nonarg(+),linkTo(np)]) :-
	optional(zero,X),
	notAssigned(X).

% default is non-argument unless the theta slot is instantiated, in which
% case, the argument status is choosen using the predicate chooseIt/3

lex(it,n,[agr([3,sg,n]),goal(theta(R),lambda(X,chooseIt(R,V,X))),
	  oneof(_,[[noCoindex,nonarg(V),linkTo(c2)],[a(-),p(+)]])]).


% Proper nouns
lex('aHmad',n,[a(-),p(-),agr([3,sg,m])]).
lex('Ciisaa',n,[a(-),p(-),agr([3,sg,m])]).
lex(muusaa,n,[a(-),p(-),agr([3,sg,m])]).
lex(zayd,n,[a(-),p(-),agr([3,sg,m])]).

% Anaphors and pronominals

lex(nfas,n,[a(+),p(-),agr([]),eng(self)]).

% Pronominal clitics
lex(tu,pf(cl),[a(-),p(+),agr([1,sg,[m,f]]),eng(i)|Fs]) :- subjClFeatures(Fs).
lex(naa,pf(cl),[a(-),p(+),agr([1,pl,[m,f]]),eng(us)|Fs]) :- 
	objCliticFeatures(Fs).
lex(ta,pf(cl),[a(-),p(+),agr([2,sg,[m,f]]),eng(you)|Fs]) :- subjClFeatures(Fs).
lex(hu,pf(cl),[a(-),p(+),agr([3,sg,m]),eng(him)|Fs]) :- 
	objCliticFeatures(Fs).
lex(hum,pf(cl),[a(+),p(-),agr([3,pl,m]),eng(they)|Fs]) :- 
	objCliticFeatures(Fs).
lex(hunna,pf(cl),[a(+),p(-),agr([3,pl,f]),eng(they)|Fs]) :- 
	objCliticFeatures(Fs).

% Quantifier nouns
% lex(everyone,n,[a(-),p(-),agr([3,sg,[m,f]]),op(+)]).

% Wh-nouns
lex(man,n,[a(-),p(-),agr([3,sg,[m,f]]),wh(+),eng(who)]).

% Common nouns

lex('Qabuu',n,[a(-),p(-),agr([3,sg,m]),morphC(nom),eng(father)]).
lex('Qawlaad',n,[a(-),p(-),agr([3,pl,[m,f]]),eng(children)]).
lex('Qustaad_',n,[a(-),p(-),agr([3,sg,[m,f]]),eng(teacher)]).
lex(baqarat,n,[a(-),p(-),agr([3,sg,f]),eng(cow)]).
lex(bint,n,[a(-),p(-),agr([3,sg,f]),eng(girl)]).
lex(daar,n,[a(-),p(-),agr([3,sg,n]),eng(house)]).
lex(jaasuus,n,[a(-),p(-),agr([3,sg,[m,f]]),eng(spy)]).
lex(makaatib,n,[a(-),p(-),agr([3,pl,n]),eng(office)]).
lex(mudiir,n,[a(-),p(-),agr([3,sg,[m,f]]),eng(director)]).
lex('nisaaQ',n,[a(-),p(-),agr([3,pl,f]),eng(women)]).
lex(rajul,n,[a(-),p(-),agr([3,sg,m]),eng(man)]).
lex(rijaal,n,[a(-),p(-),agr([3,sg,m]),eng(men)]).
lex(risaalat,n,[a(-),p(-),agr([3,sg,n]),eng(letter)]).
lex(talju,n,[a(-),p(-),agr([3,sg,n]),eng(snow)]).

% Nouns with O-grids
lex(father,n,[grid([possessor],[]),a(-),p(-),count(+),agr([3,sg,m])]).
lex(mother,n,[grid([possessor],[]),a(-),p(-),count(+),agr([3,sg,f])]).

% Nominalized verbs
% lex(belief,n,[case(_),theta(_),count(+),agr([3,sg,n]),a(-),p(-)|F]) :-
%	verbToNounFeatures(believe,F).

%%% VERBS - BASE FORMS
%
% Obligatory Elements
% 	1. grid(Ext,Int) denotes the O-grid
% 	2. morph(Base,[]/ing/s/ed(1)/ed(2)) 
% Optional Features
% 	1. ecm(opt/oblig) for ECM verbs
% 	2. noCasemark(+) for technically intransitive verbs
% 	3. subcat(C$Fs,Fs1) subcategorizes for an element of category C 
%    	   with feature restriction list Fs.
%    	   Feature list Fs1 will be appended to the feature list of C
% 	4. selR(Rs), selR(Type,R) restriction(s) Rs or R on direct object.


lex('Sagata',v,[morph(fall,[]),grid([],[theme]),noCasemark(+)]).

lex('QaCtay',v,[morph(give,past(+)),grid([agent],[goal,theme]),idoCase(acc)]).
lex('Qaqbal',v,[morph(appear,past(+)),grid([theme],[])]). % unaccusative
lex('Qarad',v,[morph(want,past(+)),grid([agent],[prop])]).
lex(ahtarim,v,[morph(respect,past(-)),grid([experiencer],[theme])]).
lex(araa,v,[morph(see,past(-)),grid([experiencer],[theme])]).
lex('Darab',v,[morph(beat,past(+)),grid([agent],[patient])]).
lex('dabaH',v,[morph(cut_throat,past(+)),grid([agent],[patient])]).
lex(daxal,v,[morph(enter,past(+)),grid([agent],[theme])]).
lex(daxala,v,[morph(enter,past(+)),grid([agent],[theme])]).
lex('jaaQ',v,[morph(come,past(+)),grid([theme],[])]).
lex('aQtii',v,[morph(come,past(-)),grid([theme],[])]).
lex(kataba,v,[morph(write,past(+)),grid([agent],[theme])]).
lex(ntaqada,v,[morph(criticize,past(+)),grid([agent],[theme])]).
lex(intaqada,v,[morph(criticize,past(+)),grid([agent],[theme])]).
lex('Saaka',v,[morph(complained,past(+)),grid([agent],[theme])]).
lex('Saahad',v,[morph(saw,past(+)),grid([experiencer],[theme])]).
lex(takallam,v,[morph(speak,past(+)),grid([theme],[])]).
lex('uHibb',v,[morph(like,past(-)),grid([agent],[theme])]).
lex(uriid,v,[morph(want,past(-)),grid([agent],[prop])]).
lex(uqaabil,v,[morph(meet,past(-)),grid([agent],[theme])]).
lex(zaCam,v,[morph(pretend,past(+)),grid([agent],[prop])]).

%% Auxiliaries

% lex(have,v,[morph(have,[]),aux,perf,subcat(vp$[morph(_,ed(2))],[perf])]).
% lex(do,v,[morph(do,[]),aux,subcat(vp$[morph(_,[])],[])]).
% lex(can,v,[morph(can,def(past(-))),aux,modal,
%	   subcat(vp$[morph(_,[])],[])]).

%% Adjectives
%%

% Complementizers
lex('Qan',c,[wh(-),eng(that)]).

% Modals
lex('Qa',mod,[eng(that)]).

% Markers: morphological realizations of particular features

% Case
lex(u,mrkr,[left(n,[],[suffix(nom),morphC(nom)])]).
lex(un,mrkr,[left(n,[],[suffix(nom),morphC(nom)])]).
lex(a,mrkr,[left(n,[],[suffix(acc),morphC(acc)])]).
lex(an,mrkr,[left(n,[],[suffix(acc),morphC(acc)])]).

% Agreement
lex(at,mrkr,[left(v,[],[suffix(f),agr([[],[],f])])]). % f
lex(na,mrkr,[left(v,[],[suffix('f.pl'),agr([[],pl,f])])]). % f.pl

lex(a,mrkr,[left(v,[],[suffix(a),subj])]). % subjunctive

% Person prefix
lex(y,mrkr,[right(v,[],[prefix(y),agr([3,[],[]])])]).

no lexFeature(_,_).

% relevant for marker constraints
relevant(n).
relevant(v).

no inf(_,_).

% CONTRACTIONS
% 1st parameter is class

no contraction(_,_,_).
no contraction(_,_,_,_).

% MISCELLANEOUS

tnsAGRFeatures(V,Fs) :-			  % override defn in xbar.pl
	V has_feature morph(_,TNS),
	vAGRFeature(V,AGR),
	mkFs([TNS,AGR],Fs).

vAGRFeature(V,AGR) :-
	V has_feature agr(A)
	-> AGR = agr(A)
	;  AGR = agr([]).

%%% PREDICATES FOR INDIVIDUAL LEXICAL ITEMS

noMax(X) :- cat(X,C), \+ max(C).

overt(X) :- \+ ec(X).

% [CP Op[i] [C' ..t[i].. ]] 
% Assume t[i] O-marked theme

inheritRole(CP,Role) :-
	cat(CP,c2),
	Op specifier_of CP,
	emptyOperator(Op)
	-> Role = [theme]
	;  Role = [].

subjClFeatures([adjoin(v,right),morphC(nom)]).

objCliticFeatures([adjoin(v,right),morphC(acc)]).