msyds/gf-core
1
0
Fork 0
forked from GitHub/gf-core
Code Pull Requests Activity

Stemmed Finnish for N and A. Cuts down the number of forms stored in the lexicon to 30%, requiring more often -bind in linearization. Stored as alternative files in finnish/stemmed; perhaps to be functorized later.

Browse Source
This commit is contained in:
aarne
2013-03-14 20:55:59 +00:00
parent 59cbd2867f
commit 330a2ae094
12 changed files with 2280 additions and 95 deletions
Download Patch File Download Diff File Expand all files Collapse all files

95
lib/src/finnish/StemFin.gf
View File

@@ -1,95 +0,0 @@
-- stem-based morphology, with just the minimum number of concatenative stems
resource StemFin = open MorphoFin, Prelude in {
flags coding = utf8 ;
oper
SNoun : Type = {s : Predef.Ints 9 => Str ; h : Harmony} ;
-- SVerb : Type ;
nforms2snoun : NForms -> SNoun = \nf -> {
s = table {
0 => nf ! 0 ; -- ukko
1 => Predef.tk 1 (nf ! 1) ; -- uko(n)
2 => nf ! 2 ; -- ukkoa
3 => nf ! 3 ; -- ukkona
4 => Predef.tk 1 (nf ! 4) ; -- ukkoo(n)
5 => Predef.tk 1 (nf ! 5) ; -- ukkoje(n)
6 => nf ! 6 ; -- ukkoja
7 => Predef.tk 2 (nf ! 7) ; -- ukkoi(na)
8 => Predef.tk 3 (nf ! 8) ; -- ukoi(ssa)
9 => Predef.tk 1 (nf ! 9) -- ukkoihi(n)
} ;
h = aHarmony (last (nf ! 2)) ;
} ;
snoun2noun : SNoun -> Noun = \sn ->
let
f = sn.s ;
ukko = f ! 0 ;
uko = f ! 1 ;
ukkoa = f ! 2 ;
ukkona = f ! 3 ;
ukkoo = f ! 4 ;
ukkoje = f ! 5 ;
ukkoja = f ! 6 ;
ukkoi = f ! 7 ;
ukoi = f ! 8 ;
ukkoihi = f ! 9 ;
a = harmonyA sn.h ;
in
{s = table {
NCase Sg Nom => ukko ;
NCase Sg Gen => uko + "n" ;
NCase Sg Part => ukkoa ;
NCase Sg Transl => uko + "ksi" ;
NCase Sg Ess => ukkona ;
NCase Sg Iness => uko + ("ss" + a) ;
NCase Sg Elat => uko + ("st" + a) ;
NCase Sg Illat => ukkoo + "n" ;
NCase Sg Adess => uko + ("ll" + a) ;
NCase Sg Ablat => uko + ("lt" + a) ;
NCase Sg Allat => uko + "lle" ;
NCase Sg Abess => uko + ("tt" + a) ;
NCase Pl Nom => uko + "t" ;
NCase Pl Gen => ukkoje + "n" ;
NCase Pl Part => ukkoja ;
NCase Pl Transl => ukoi + "ksi" ;
NCase Pl Ess => ukkoi + ("n" + a) ;
NCase Pl Iness => ukoi + ("ss" + a) ;
NCase Pl Elat => ukoi + ("st" + a) ;
NCase Pl Illat => ukkoihi + "n" ;
NCase Pl Adess => ukoi + ("ll" + a) ;
NCase Pl Ablat => ukoi + ("lt" + a) ;
NCase Pl Allat => ukoi + "lle" ;
NCase Pl Abess => ukoi + ("tt" + a) ;
NComit => ukkoi + "ne" ;
NInstruct => ukoi + "n" ;
NPossNom _ => ukko ;
NPossGen Sg => ukko ;
NPossGen Pl => ukkoje ;
NPossTransl Sg => uko + "kse" ;
NPossTransl Pl => ukoi + "kse" ;
NPossIllat Sg => ukkoo ;
NPossIllat Pl => ukkoihi
} ;
h = sn.h ;
lock_N = <>
} ;
aHarmony : Str -> Harmony = \a -> case a of
{"a" => Back ; _ => Front} ;
harmonyA : Harmony -> Str = \h -> case h of
{Back => "a" ; Front => "ä"} ;
}

53
lib/src/finnish/stemmed/AdjectiveFin.gf Normal file
View File

@@ -0,0 +1,53 @@
concrete AdjectiveFin of Adjective = CatFin ** open ResFin, StemFin, Prelude in {
flags optimize=all_subs ; -- gfc size from 2864336 to 6786 - i.e. factor 422
lin
PositA a = {
s = \\_ => (snoun2nounSep {s = \\f => a.s ! Posit ! SAN f ; h = a.h}).s
} ;
ComparA a np =
let acomp = (snoun2nounSep {s = \\f => a.s ! Posit ! SAN f ; h = a.h}).s in {
s = \\isMod,af => case isMod of {
True => np.s ! NPCase Part ++ acomp ! af ; -- minua isompi
_ => acomp ! af ++ "kuin" ++ np.s ! NPCase Nom -- isompi kuin minä
}
} ;
CAdvAP ad ap np = {
s = \\m,af => ad.s ++ ap.s ! m ! af ++ ad.p ++ np.s ! NPCase Nom
} ;
UseComparA a = {
s = \\_ => (snoun2nounSep {s = \\f => a.s ! Compar ! SAN f ; h = a.h}).s
} ;
-- $SuperlA$ belongs to determiner syntax in $Noun$.
AdjOrd ord = {
s = \\_ => ord.s
} ;
ComplA2 a np = {
s = \\isMod,af =>
preOrPost isMod (appCompl True Pos a.c2 np) ((snoun2nounSep {s = \\f => a.s ! Posit ! SAN f ; h = a.h}).s ! af)
} ;
ReflA2 a = {
s = \\isMod,af =>
preOrPost isMod
(appCompl True Pos a.c2 (reflPron (agrP3 Sg))) ((snoun2nounSep {s = \\f => a.s ! Posit ! SAN f ; h = a.h}).s ! af)
} ;
SentAP ap sc = {
s = \\b,a => ap.s ! b ! a ++ sc.s
} ;
AdAP ada ap = {
s = \\b,af => ada.s ++ ap.s ! b ! af
} ;
UseA2 a = {
s = \\_ => (snoun2nounSep {s = \\f => a.s ! Posit ! SAN f ; h = a.h}).s
} ;
}

23
lib/src/finnish/stemmed/AdverbFin.gf Normal file
View File

@@ -0,0 +1,23 @@
concrete AdverbFin of Adverb = CatFin ** open ResFin, Prelude, StemFin in {
lin
PositAdvAdj a = {s = a.s ! Posit ! SAAdv} ;
ComparAdvAdj cadv a np = {
s = cadv.s ++ a.s ! Posit ! SAAdv ++ cadv.p ++ np.s ! NPCase Nom
} ;
ComparAdvAdjS cadv a s = {
s = cadv.s ++ a.s ! Posit ! SAAdv ++ cadv.p ++ s.s
} ;
PrepNP prep np = {s = preOrPost prep.isPre prep.s (np.s ! prep.c)} ;
AdAdv = cc2 ;
PositAdAAdj a = {s = glue (a.s ! Posit ! SAN 1) "n"} ; -- älyttömän
SubjS = cc2 ;
----b AdvSC s = s ;
AdnCAdv cadv = {s = cadv.s ++ "kuin"} ;
}

101
lib/src/finnish/stemmed/CatFin.gf Normal file
View File

@@ -0,0 +1,101 @@
concrete CatFin of Cat = CommonX ** open ResFin, StemFin, Prelude in {
flags optimize=all_subs ;
lincat
-- Tensed/Untensed
S = {s : Str} ;
QS = {s : Str} ;
RS = {s : Agr => Str ; c : NPForm} ;
SSlash = {s : Str ; c2 : Compl} ;
-- Sentence
Cl = {s : ResFin.Tense => Anteriority => Polarity => SType => Str} ;
ClSlash = {s : ResFin.Tense => Anteriority => Polarity => Str ; c2 : Compl} ;
Imp = {s : Polarity => Agr => Str} ;
-- Question
QCl = {s : ResFin.Tense => Anteriority => Polarity => Str} ;
IP = {s : NPForm => Str ; n : Number} ;
IComp = {s : Agr => Str} ;
IDet = {s : Case => Str ; n : Number ; isNum : Bool} ;
IQuant = {s : Number => Case => Str} ;
-- Relative
RCl = {s : ResFin.Tense => Anteriority => Polarity => Agr => Str ; c : NPForm} ;
RP = {s : Number => NPForm => Str ; a : RAgr} ;
-- Verb
VP = ResFin.VP ;
VPSlash = ResFin.VP ** {c2 : Compl} ;
Comp = {s : Agr => Str} ;
-- Adjective
-- The $Bool$ tells whether usage is modifying (as opposed to
-- predicative), e.g. "x on suurempi kuin y" vs. "y:tä suurempi luku".
AP = {s : Bool => NForm => Str} ;
-- Noun
-- The $Bool$ tells if a possessive suffix is attached, which affects the case.
CN = {s : NForm => Str ; h : Harmony} ;
Pron = {s : NPForm => Str ; a : Agr} ;
NP = {s : NPForm => Str ; a : Agr ; isPron : Bool ; isNeg : Bool} ;
Det = {
s1 : Case => Str ; -- minun kolme
s2 : Harmony => Str ; -- -ni (Front for -nsä, Back for -nsa)
sp : Case => Str ; -- se (substantival form)
n : Number ; -- Pl (agreement feature for verb)
isNum : Bool ; -- True (a numeral is present)
isPoss : Bool ; -- True (a possessive suffix is present)
isDef : Bool ; -- True (verb agrees in Pl, Nom is not Part)
isNeg : Bool -- False (only True for "mikään", "kukaan")
} ;
---- QuantSg, QuantPl = {s1 : Case => Str ; s2 : Str ; isPoss, isDef : Bool} ;
Ord = {s : NForm => Str} ;
Predet = {s : Number => NPForm => Str} ;
Quant = {s1,sp : Number => Case => Str ; s2 : Harmony => Str ; isPoss : Bool ; isDef : Bool ; isNeg : Bool} ;
Card = {s : Number => Case => Str ; n : Number} ;
Num = {s : Number => Case => Str ; isNum : Bool ; n : Number} ;
-- Numeral
Numeral = {s : CardOrd => Str ; n : Number} ;
Digits = {s : CardOrd => Str ; n : Number} ;
-- Structural
Conj = {s1,s2 : Str ; n : Number} ;
----b DConj = {s1,s2 : Str ; n : Number} ;
Subj = {s : Str} ;
Prep = Compl ;
-- Open lexical classes, e.g. Lexicon
V, VS, VQ = Verb1 ; -- = {s : VForm => Str ; sc : Case} ;
V2, VA, V2Q, V2S = Verb1 ** {c2 : Compl} ;
V2A = Verb1 ** {c2, c3 : Compl} ;
VV = Verb1 ** {vi : InfForm} ; ---- infinitive form
V2V = Verb1 ** {c2 : Compl ; vi : InfForm} ; ---- infinitive form
V3 = Verb1 ** {c2, c3 : Compl} ;
A = {s : Degree => SAForm => Str ; h : Harmony} ;
A2 = {s : Degree => SAForm => Str ; h : Harmony ; c2 : Compl} ;
N = SNoun ;
N2 = SNoun ** {c2 : Compl ; isPre : Bool} ;
N3 = SNoun ** {c2,c3 : Compl ; isPre,isPre2 : Bool} ;
PN = SNoun ;
oper Verb1 = {s : VForm => Str ; sc : NPForm ; qp : Bool ; p : Str} ;
}

11
lib/src/finnish/stemmed/LangFin.gf Normal file
View File

@@ -0,0 +1,11 @@
--# -path=.:..:../../abstract:../../common
concrete LangFin of Lang =
-- GrammarFin - [SlashV2VNP,SlashVV], ---- to speed up compilation
GrammarFin,
LexiconFin
** {
flags startcat = Phr ; unlexer = text ; lexer = finnish ;
} ;

388
lib/src/finnish/stemmed/LexiconFin.gf Normal file
View File

@@ -0,0 +1,388 @@
--# -path=.:prelude
concrete LexiconFin of Lexicon = CatFin ** open MorphoFin, StemFin, ParadigmsFin in {
flags
optimize=values ;
lin
airplane_N = mkN "lentokone" ;
answer_V2S = mkV2 (mkV "vastata") (casePrep allative) ;
apartment_N = mkN "asunto" ;
apple_N = mkN "omena" ; --- omenia, not omenoita
art_N = mkN "taide" ;
ask_V2Q = mkV2 (mkV "kysyä") (casePrep ablative) ;
baby_N = mkN "vauva" ;
bad_A = mkA (mkN "paha") "pahempi" "pahin" ;
bank_N = mkN "pankki" ;
beautiful_A = mkA (mkN "kaunis") "kauniimpi" "kaunein" ;
become_VA = mkVA (mkV "tulla") (casePrep translative) ;
beer_N = mkN "olut" "oluita" ;
beg_V2V = mkV2V (mk2V "pyytää" "pyysi") (casePrep partitive) ;
big_A = mkA (mkN "suuri" "suuria") "suurempi" "suurin" ;
bike_N = mkN "polku" (mkN "pyörä") ;
bird_N = mkN "lintu" ;
black_A = mkA (mkN "musta") "mustempi" "mustin" ;
blue_A = mkA (mkN "sininen") "sinisempi" "sinisin" ;
boat_N = mkN "vene" ;
book_N = mkN "kirja" ;
boot_N = mkN "saapas" ;
boss_N = mkN "pomo" ;
boy_N = mkN "poika" "pojan" "poikia" ;
bread_N = mkN "leipä" ;
break_V2 = mkV2 (mkV "rikkoa") ;
broad_A = mkA (mkN "leveä") "leveämpi" "levein" ;
brother_N2 = mkN2 (
mkN "veli" "veljen" "veljenä" "veljeä" "veljeen"
"veljinä" "veljissä" "veljien" "veljiä" "veljiin") ;
brown_A = mkA (mkN "ruskea") "ruskeampi" "ruskein" ;
butter_N = mk3N "voi" "voin" "voita" ; ---- errors in Part
buy_V2 = mkV2 (mkV "ostaa") ;
camera_N = mkN "kamera" ;
cap_N = mkN "lakki" ;
car_N = mkN "auto" "auton" "autoja" ; -- mkN: audon
carpet_N = mkN "matto" ;
cat_N = mkN "kissa" ;
ceiling_N = mkN "katto" ;
chair_N = mkN "tuoli" ;
cheese_N = mkN "juusto" ;
child_N = mkN "lapsi" "lapsen" "lasta" "lapsena" "lapseen"
"lasten" "lapsia" "lapsina" "lapsissa" "lapsiin" ;
church_N = mkN "kirkko" ;
city_N = mkN "kaupunki" ;
clean_A = mkA (mkN "puhdas") ;
clever_A = mkA (mkN "viisas") ;
close_V2 = mkV2 (mkV "sulkea") ;
coat_N = mkN "takki" ;
cold_A = mkA (mkN "kylmä") "kylmempi" "kylmin" ;
come_V = mkV "tulla" ;
computer_N = mkN "tietokone" ;
country_N = mkN "maa" ;
cousin_N = mkN "serkku" ;
cow_N = mkN "lehmä" ;
die_V = mkV "kuolla" ;
dirty_A = mkA (mkN "likainen") "likaisempi" "likaisin" ;
distance_N3 = mkN3 (mkN "etäisyys") (casePrep elative) (casePrep illative) ;
doctor_N = mk2N "tohtori" "tohtoreita" ;
dog_N = mkN "koira" ;
door_N = mkN "ovi" "ovia" ;
drink_V2 = mkV2 (mkV "juoda") (casePrep partitive) ;
easy_A2V = mkA2 (mkA (mkN "helppo") "helpompi" "helpoin")
(casePrep allative) ;
eat_V2 = mkV2 (mkV "syödä") (casePrep partitive) ;
empty_A = mkA (mkN "tyhjä") "tyhjempi" "tyhjin" ;
enemy_N = mkN "vihollinen" ;
factory_N = mkN "tehdas" ;
father_N2 = mkN2 (mkN "isä") ;
fear_VS = mkVS (mk2V "pelätä" "pelkäsi") ;
find_V2 = mkV2 (mk2V "löytää" "löysi") ;
fish_N = mkN "kala" ;
floor_N = mk2N "lattia" "lattioita" ;
forget_V2 = mkV2 (mkV "unohtaa") ;
fridge_N = mkN "jääkaappi" ;
friend_N = mkN "ystävä" ;
fruit_N = mkN "hedelmä" ;
fun_AV = mkAV (mkA (mkN "hauska") "hauskempi" "hauskin") ;
garden_N = mkN "puutarha" "puutarhan" "puutarhoja" ;
girl_N = mkN "tyttö" ;
glove_N = mkN "käsine" ;
gold_N = mkN "kulta" ;
good_A = mkA (mkN "hyvä") "parempi" "parhain" ; --- paras
go_V = mkV "mennä" ;
green_A = mkA (mkN "vihreä") "vihreämpi" "vihrein" ;
harbour_N = mkN "satama" "sataman" "satamia" ;
hate_V2 = mkV2 (mkV "vihata") cpartitive ;
hat_N = mkN "hattu" ;
hear_V2 = mkV2 (mkV "kuulla") ;
hill_N = mkN "kukkula" ;
hope_VS = mkVS (mkV "toivoa") ;
horse_N = mkN "hevonen" ;
hot_A = mkA (mkN "kuuma") "kuumempi" "kuumin" ;
house_N = mkN "talo" ;
important_A = mkA (mkN "tärkeä") "tärkeämpi" "tärkein" ;
industry_N = mkN "teollisuus" ;
iron_N = mkN "rauta" ;
king_N = mkN "kuningas" ;
know_VS = mkVS (mkV "tietää" "tiesi") ;
know_VQ = mkVQ (mkV "tietää" "tiesi") ;
know_V2 = mkV2 (mkV "tuntea" "tunsi") ;
lake_N = mkN "järvi" "järviä" ;
lamp_N = mkN "lamppu" ;
learn_V2 =
mkV2 (mk12V "oppia" "opin" "oppii" "oppivat" "oppikaa" "opitaan"
"opin" "oppi" "oppisi" "oppinut" "opittu" "opitun") ;
leather_N = mkN "nahka" ; --- nahan
leave_V2 = mkV2 (mkV "jättää") ;
like_V2 = mkV2 (mkV "pitää") elative ;
listen_V2 = mkV2 (mkV "kuunnella" "kuunteli") partitive ;
live_V = mkV "elää" ;
long_A = mkA (mkN "pitkä") "pitempi" "pisin" ;
lose_V2 = mkV2 (mkV "hävitä" "hävisi") ; --- hukata
love_N = mk3N "rakkaus" "rakkauden" "rakkauksia" ;
love_V2 = mkV2 (mkV "rakastaa") partitive ;
man_N = mkN "mies" "miehen" "miestä" "miehenä" "mieheen"
"miesten" "miehiä" "miehinä" "miehissä" "miehiin" ;
married_A2 = mkA2 (mkA "avioitunut") (postPrep genitive "kanssa") ; ---- infl
meat_N = mkN "liha" ;
milk_N = mkN "maito" ;
moon_N = mkN "kuu" ;
mother_N2 = mkN2 (mkN "äiti") ;
mountain_N = mkN "vuori" "vuoria" ;
music_N = mkN "musiikki" ;
narrow_A = mkA (mkN "kapea") "kapeampi" "kapein" ;
new_A = mkA (mk3N "uusi" "uuden" "uusia") "uudempi" "uusin" ;
newspaper_N = mkN "sanoma" (mkN "lehti" "lehtiä") ; --- for correct vowel harmony
oil_N = mkN "öljy" ;
old_A = mkA (mkN "vanha") "vanhempi" "vanhin" ;
open_V2 = mkV2 (mkV "avata" "avasi") ;
paint_V2A = mkV2A (mkV "maalata") accPrep (casePrep translative) ;
paper_N = mk2N "paperi" "papereita" ;
paris_PN = mkPN (mkN "Pariisi") ;
peace_N = mkN "rauha" ;
pen_N = mkN "kynä" ;
planet_N = mkN "planeetta" ;
plastic_N = mkN "muovi" ;
play_V2 = mkV2 (mkV "pelata") cpartitive ; --- leikkiä, soittaa
policeman_N = mkN "poliisi" ;
priest_N = mkN "pappi" ;
probable_AS = mkAS --- for vowel harmony
(mkA (mkN "todennäköinen") "tonennäköisempi" "todennäköisin") ; ---- sta
queen_N = mkN "kuningatar" ;
radio_N = mk2N "radio" "radioita" ;
rain_V0 = mkV0 (mk2V "sataa" "satoi") ;
read_V2 = mkV2 (mkV "lukea") ;
red_A = mkA "punainen" ;
religion_N = mkN "uskonto" ;
restaurant_N = mkN "ravintola" ;
river_N = mkN "joki" "jokia" ;
rock_N = mk2N "kallio" "kallioita" ;
roof_N = mkN "katto" ;
rubber_N = mkN "kumi" ;
run_V = mk2V "juosta" "juoksi" ;
say_VS = mkVS (mkV "sanoa") ;
school_N = mkN "koulu" ;
science_N = mkN "tiede" ;
sea_N = mkN "meri" "meren" "meriä" "merta" ;
seek_V2 = mkV2 (mkV "etsiä") cpartitive ;
see_V2 = mkV2 (
mk12V "nähdä" "näen" "näkee" "näkevät" "nähkää" "nähdään"
"näin" "näki" "näkisi" "nähnyt" "nähty" "nähdyn") ;
sell_V3 = mkV3 (mkV "myydä") accPrep (casePrep allative) ;
send_V3 = mkV3 (mkV "lähettää") accPrep (casePrep allative) ;
sheep_N = mkN "lammas" ;
ship_N = mkN "laiva" ;
shirt_N = mkN "paita" ;
shoe_N = mkN "kenkä" ;
shop_N = mkN "kauppa" ;
short_A = mkA (mkN "lyhyt" "lyhyitä") ;
silver_N = mkN "hopea" ;
sister_N = mkN "sisko" ;
sleep_V = mkV "nukkua" ;
small_A = mkA (mk2N "pieni" "pieniä") "pienempi" "pienin" ;
snake_N = mkN "käärme" ;
sock_N = mkN "sukka" ;
speak_V2 = mkV2 (mkV "puhua") cpartitive ;
star_N = mkN "tähti" "tähtiä" ;
steel_N = mkN "teräs" ;
stone_N = mkN "kivi" "kiviä" ;
stove_N = mk3N "liesi" "lieden" "liesiä" ;
student_N = mk2N "opiskelija" "opiskelijoita" ;
stupid_A = mkA "tyhmä" ;
sun_N = mkN "aurinko" ;
switch8off_V2 = mkV2 (mkV "sammuttaa") ; ---
switch8on_V2 = mkV2 (mkV "sytyttää") ; ---
table_N = mkN "pöytä" ;
talk_V3 = mkV3 (mkV "puhua") (casePrep allative) (casePrep elative) ;
teacher_N = mkN "opettaja" ;
teach_V2 = mkV2 (mkV "opettaa") ;
television_N = mk2N "televisio" "televisioita" ;
thick_A = mkA "paksu" ;
thin_A = mkA (mkN "ohut" "ohuita") ;
train_N = mkN "juna" ;
travel_V = mkV "matkustaa" ;
tree_N = mkN "puu" ;
---- trousers_N = mkN "trousers" ;
ugly_A = mkA (mkN "ruma") "rumempi" "rumin" ;
understand_V2 = mkV2 (mkV "ymmärtää" "ymmärrän" "ymmärsi") ;
university_N = mkN "yliopisto" ;
village_N = mkN "kylä" ;
wait_V2 = mkV2 (mkV "odottaa") partitive ;
walk_V = mkV "kävellä" "käveli" ;
warm_A = mkA
(mkN "lämmin" "lämpimän" "lämmintä" "lämpimänä" "lämpimään"
"lämpiminä" "lämpimiä" "lämpimien" "lämpimissä" "lämpimiin"
)
"lämpimämpi" "lämpimin" ;
war_N = mkN "sota" ;
watch_V2 = mkV2 (mkV "katsella") cpartitive ;
water_N = mk3N "vesi" "veden" "vesiä" ;
white_A = mkA "valkoinen" ;
window_N = mk2N "ikkuna" "ikkunoita" ;
wine_N = mkN "viini" ;
win_V2 = mkV2 (mkV "voittaa") ;
woman_N = mkN "nainen" ;
wonder_VQ = mkVQ (mkV "ihmetellä") ;
wood_N = mkN "puu" ;
write_V2 = mkV2 (mkV "kirjoittaa") ;
yellow_A = mkA "keltainen" ;
young_A = mkA (mkN "nuori" "nuoria") "nuorempi" "nuorin" ;
do_V2 = mkV2 (
mkV "tehdä" "teen" "tekee" "tekevät" "tehkää" "tehdään"
"tein" "teki" "tekisi" "tehnyt" "tehty" "tehdyn") ;
now_Adv = mkAdv "nyt" ;
already_Adv = mkAdv "jo" ;
song_N = mkN "laulu" ;
add_V3 = mkV3 (mkV "lisätä") accPrep (casePrep illative) ;
number_N = mk2N "numero" "numeroita" ;
put_V2 = mkV2 (mkV "panna") ;
stop_V = mkV "pysähtyä" ;
jump_V = mkV "hypätä" ;
left_Ord = mkOrd (snoun2nounBind (mkN "vasen")) ;
right_Ord = mkOrd (snoun2nounBind (mkN "oikea")) ;
far_Adv = mkAdv "kaukana" ;
correct_A = mkA "oikea" ;
dry_A = mkA (mkN "kuiva") "kuivempi" "kuivin" ;
dull_A = mkA (mkN "tylsä") "tylsempi" "tylsin" ;
full_A = mkA (mk3N "täysi" "täyden" "täysiä") "täydempi" "täysin" ;
heavy_A = mkA "raskas" ;
near_A = mkA (mkN "läheinen") "läheisempi" "lähin" ;
rotten_A = mkA "mätä" ;
round_A = mkA "pyöreä" ;
sharp_A = mkA "terävä" ;
smooth_A = mkA "sileä" ;
straight_A = mkA (mkN "suora") "suorempi" "suorin" ;
wet_A = mkA (mkN "märkä") "märempi" "märin" ;
wide_A = mkA "leveä" ;
animal_N = mk3N "eläin" "eläimen" "eläimiä" ;
ashes_N = mkN "tuhka" ;
back_N = mkN "selkä" ;
bark_N = mkN "kaarna" ;
belly_N = mkN "vatsa" ;
blood_N = mkN "veri" "veren" "veriä" "verta" ;
bone_N = mkN "luu" ;
breast_N = mkN "rinta" ;
cloud_N = mk2N "pilvi" "pilviä" ;
day_N = mkN "päivä" ;
dust_N = mkN "pöly" ;
ear_N = mkN "korva" ;
earth_N = mkN "maa" ;
egg_N = mkN "muna" ;
eye_N = mkN "silmä" ;
fat_N = mkN "rasva" ;
feather_N = mk3N "höyhen" "höyhenen" "höyheniä" ;
fingernail_N = mk3N "kynsi" "kynnen" "kynsiä" ;
fire_N = mk2N "tuli" "tulia" ;
flower_N = mkN "kukka" ;
fog_N = mkN "sumu" ;
foot_N = mkN "jalka" ;
forest_N = mkN "metsä" ;
grass_N = mkN "ruoho" ;
guts_N = mkN "sisälmys" ; --- suoli
hair_N = mkN "hius" ;
hand_N = mk3N "käsi" "käden" "käsiä" ;
head_N = mkN "pää" ;
heart_N = mkN "sydän" "sydämen" "sydäntä" "sydämenä" "sydämeen"
"sydänten" "sydämiä" "sydäminä" "sydämissä" "sydämiin" ;
horn_N = mk2N "sarvi" "sarvia" ;
husband_N = mkN "mies" "miehen" "miestä" "miehenä" "mieheen"
"miesten" "miehiä" "miehinä" "miehissä" "miehiin" ;
ice_N = mkN "jää" ;
knee_N = mk2N "polvi" "polvia" ;
leaf_N = mk2N "lehti" "lehtiä" ;
leg_N = mkN "jalka" ; --- sääri
liver_N = mkN "maksa" ;
louse_N = mkN "lude" ;
mouth_N = mkN "suu" ;
name_N = mk2N "nimi" "nimiä" ;
neck_N = mkN "niska" ;
night_N = mkN "yö" ;
nose_N = mkN "nenä" ;
person_N = mkN "henkilö" ;
rain_N = mkN "sade" ;
road_N = mkN "tie" ;
root_N = mk2N "juuri" "juuria" ;
rope_N = mk3N "köysi" "köyden" "köysiä" ;
salt_N = mkN "suola" ;
sand_N = mkN "hiekka" ;
seed_N = mkN "siemen" ;
skin_N = mkN "nahka" ;
sky_N = mk3N "taivas" "taivaan" "taivaita" ;
smoke_N = mkN "savu" ;
snow_N = mkN "lumi" "lumen" "lumia" "lunta" ;
stick_N = mkN "keppi" ;
tail_N = mkN "häntä" ;
tongue_N = mk2N "kieli" "kieliä" ;
tooth_N = mkN "hammas" ;
wife_N = mkN "vaimo" ;
wind_N = mk2N "tuuli" "tuulia" ;
wing_N = mk2N "siipi" "siipiä" ;
worm_N = mkN "mato" ;
year_N = mk3N "vuosi" "vuoden" "vuosia" ;
bite_V2 = mkV2 (mkV "purra") ;
blow_V = mkV "puhaltaa" ;
burn_V = mkV "palaa" ;
count_V2 = mkV2 (mkV "laskea") ;
cut_V2 = mkV2 (mk2V "leikata" "leikkasi") ;
dig_V = mkV "kaivaa" ;
fall_V = mkV "pudota" "putoan" "putosi" ;
fear_V2 = mkV2 (mkV "pelätä" "pelkään" "pelkäsi") cpartitive ;
fight_V2 = mkV2 (mkV "taistella") (postPrep partitive "vastaan") ;
float_V = mkV "kellua" ;
flow_V = mkV "virrata" "virtaan" "virtasi" ;
fly_V = mkV "lentää" ;
freeze_V = mkV "jäätyä" ;
give_V3 = mkV3 (mkV "antaa" "annan" "antoi") accPrep (casePrep allative) ;
hit_V2 = mkV2 (mkV "lyödä") cpartitive ;
hold_V2 = mkV2 (mkV "pitää") cpartitive ;
hunt_V2 = mkV2 (mkV "metsästää") cpartitive ;
kill_V2 = mkV2 (mkV "tappaa") ;
laugh_V = mkV "nauraa" "nauroi" ;
lie_V = mkV "maata" "makasi" ;
play_V = mkV "pelata" ;
pull_V2 = mkV2 (mkV "vetää") ;
push_V2 = mkV2 (mkV "työntää") ;
rub_V2 = mkV2 (mkV "hieroa") cpartitive ;
scratch_V2 = mkV2 (mkV "raapia") cpartitive ;
sew_V = mkV "ommella" ;
sing_V = mkV "laulaa" ;
sit_V = mkV "istua" ;
smell_V = mk2V "haista" "haisi" ;
spit_V = mkV "sylkeä" ;
split_V2 = mkV2 (mk2V "halkaista" "halkaisi") ;
squeeze_V2 = mkV2 (mkV "puristaa") cpartitive ;
stab_V2 = mkV2 (mkV "pistää") cpartitive ;
stand_V = mk12V "seistä" "seison" "seisoo" "seisovat" "seiskää" "seistään"
"seisoin" "seisoi" "seisoisi" "seissyt" "seisty" "seistyn" ; --- *seisoivät
suck_V2 = mkV2 (mkV "imeä") cpartitive ;
swell_V = mkV "turvota" "turposi" ;
swim_V = mkV "uida" "uin" "ui" ;
think_V = mkV "ajatella" "ajattelen" "ajatteli" ;
throw_V2 = mkV2 (mkV "heittää") ;
tie_V2 = mkV2 (mkV "sitoa") ;
turn_V = mkV "kääntyä" ;
vomit_V = mkV "oksentaa" ;
wash_V2 = mkV2 (mkV "pestä") ;
wipe_V2 = mkV2 (mkV "pyyhkiä") ;
breathe_V = mkV "hengittää" ;
grammar_N = mkN "kielioppi" ;
language_N = mk2N "kieli" "kieliä" ;
rule_N = mkN "sääntö" ;
john_PN = mkPN "Jussi" ;
question_N = mkN "kysymys" ;
ready_A = mkA (mkN "valmis") ;
reason_N = mkN "syy" ;
today_Adv = mkAdv "tänään" ;
uncertain_A = mkA "epävarma" ;
oper
mkOrd : Noun -> Ord ;
mkOrd x = {s = x.s ; lock_Ord = <> } ;
cpartitive = casePrep partitive ;
} ;

259
lib/src/finnish/stemmed/NounFin.gf Normal file
View File

@@ -0,0 +1,259 @@
--# -path=.:..:../../abstract:../../common
concrete NounFin of Noun = CatFin ** open ResFin, MorphoFin, StemFin, Prelude in {
flags optimize=all_subs ;
lin
-- The $Number$ is subtle: "nuo autot", "nuo kolme autoa" are both plural
-- for verb agreement, but the noun form is singular in the latter.
DetCN det cn =
let
n : Number = case det.isNum of {
True => Sg ;
_ => det.n
} ;
ncase : NPForm -> Case * NForm = \c ->
let k = npform2case n c
in
case <n, c, det.isNum, det.isPoss, det.isDef> of {
<_, NPAcc, True,_,_> => <Nom,NCase Sg Part> ; -- myin kolme kytkintä(ni)
<_, NPCase Nom, True,_,_> => <Nom,NCase Sg Part> ; -- kolme kytkintä(ni) on
<_, _, True,False,_> => <k, NCase Sg k> ; -- kolmeksi kytkimeksi
<Pl,NPAcc, _, _, False> => <k, NCase Pl Part> ; -- myin kytkimiä
<_, NPAcc, _,True,_> => <k, NPossNom n> ; -- myin kytkime+ni
<_, NPCase Nom,_,True,_> => <k, NPossNom n> ; -- kytkime+ni on/ovat...
<_, NPCase Gen,_,True,_> => <k, NPossGen n> ; -- kytkime+ni vika
<_, NPCase Transl,_,True,_> => <k, NPossTransl n> ; -- kytkim(e|i)kse+ni
<_, NPCase Illat,_,True,_> => <k, NPossIllat n> ; -- kytkim(ee|ii)+ni
_ => <k, NCase n k> -- kytkin, kytkimen,...
}
in {
s = \\c => let
k = ncase c ;
in
det.s1 ! k.p1 ++ cn.s ! k.p2 ++ det.s2 ! cn.h ;
a = agrP3 (case <det.isDef, det.isNum> of {
<False,True> => Sg ; -- kolme kytkintä on
_ => det.n
}) ;
isPron = False ; isNeg = det.isNeg
} ;
DetNP det =
let
n : Number = case det.isNum of {
True => Sg ;
_ => det.n
} ;
in {
s = \\c => let k = npform2case n c in
det.sp ! k ; -- det.s2 is possessive suffix
a = agrP3 (case det.isDef of {
False => Sg ; -- autoja menee; kolme autoa menee
_ => det.n
}) ;
isPron = False ; isNeg = det.isNeg
} ;
UsePN pn = {
s = snoun2np Sg pn ; ---\\c => (snoun2nounSep pn).s ! NCase Sg (npform2case Sg c) ;
a = agrP3 Sg ;
isPron = False ; isNeg = False
} ;
UsePron p = p ** {isPron = True ; isNeg = False} ;
PredetNP pred np = {
s = \\c => pred.s ! complNumAgr np.a ! c ++ np.s ! c ;
a = np.a ;
isPron = np.isPron ; -- kaikki minun - ni
isNeg = np.isNeg
} ;
PPartNP np v2 = {
s = \\c => np.s ! c ++ v2.s ! PastPartPass (AN (NCase (complNumAgr np.a) Ess)) ;
a = np.a ;
isPron = np.isPron ; -- minun täällä - ni
isNeg = np.isNeg
} ;
AdvNP np adv = {
s = \\c => np.s ! c ++ adv.s ;
a = np.a ;
isPron = np.isPron ; -- minun täällä - ni
isNeg = np.isNeg
} ;
DetQuantOrd quant num ord = {
s1 = \\c => quant.s1 ! num.n ! c ++ num.s ! Sg ! c ++ ord.s ! NCase num.n c ;
sp = \\c => quant.sp ! num.n ! c ++ num.s ! Sg ! c ++ ord.s ! NCase num.n c ;
s2 = quant.s2 ;
n = num.n ;
isNum = num.isNum ;
isPoss = quant.isPoss ;
isDef = quant.isDef ;
isNeg = quant.isNeg
} ;
DetQuant quant num = {
s1 = \\c => quant.s1 ! num.n ! c ++ num.s ! Sg ! c ;
sp = \\c => quant.sp ! num.n ! c ++ num.s ! Sg ! c ;
s2 = quant.s2 ;
n = num.n ;
isNum = num.isNum ; -- case num.n of {Sg => False ; _ => True} ;
isPoss = quant.isPoss ;
isDef = quant.isDef ; isNeg = quant.isNeg
} ;
PossPron p = {
s1,sp = \\_,_ => p.s ! NPCase Gen ;
s2 = table {Front => BIND ++ possSuffixFront p.a ;
Back => BIND ++ possSuffix p.a } ;
isNum = False ;
isPoss = True ;
isDef = True ; --- "minun kolme autoani ovat" ; thus "...on" is missing
isNeg = False
} ;
NumSg = {s = \\_,_ => [] ; isNum = False ; n = Sg} ;
NumPl = {s = \\_,_ => [] ; isNum = False ; n = Pl} ;
NumCard n = n ** {isNum = case n.n of {Sg => False ; _ => True}} ; -- yksi talo/kaksi taloa
NumDigits numeral = {
s = \\n,c => numeral.s ! NCard (NCase n c) ;
n = numeral.n
} ;
OrdDigits numeral = {s = \\f => numeral.s ! NOrd f} ;
NumNumeral numeral = {
s = \\n,c => numeral.s ! NCard (NCase n c) ;
n = numeral.n
} ;
OrdNumeral numeral = {s = \\f => numeral.s ! NOrd f} ;
AdNum adn num = {
s = \\n,c => adn.s ++ num.s ! n ! c ;
n = num.n
} ;
OrdSuperl a = snoun2nounSep {s = \\nc => a.s ! Superl ! SAN nc ; h = a.h} ;
DefArt = {
s1 = \\_,_ => [] ;
sp = table {Sg => pronSe.s ; Pl => pronNe.s} ;
s2 = \\_ => [] ;
isNum,isPoss,isNeg = False ;
isDef = True -- autot ovat
} ;
IndefArt = {
s1 = \\_,_ => [] ; -- Nom is Part in Pl: use isDef in DetCN
sp = \\n,c =>
(nhn (mkSubst "ä" "yksi" "yhde" "yhte" "yhtä" "yhteen" "yksi" "yksi"
"yksien" "yksiä" "yksiin")).s ! NCase n c ;
s2 = \\_ => [] ;
isNum,isPoss,isDef,isNeg = False -- autoja on
} ;
MassNP cn =
let
n : Number = Sg ;
ncase : Case -> NForm = \c -> NCase n c ;
in {
s = \\c => let k = npform2case n c in
cn.s ! ncase k ;
a = agrP3 Sg ;
isPron = False ; isNeg = False
} ;
UseN n = snoun2nounSep n ;
UseN2 n = snoun2nounSep n ;
Use2N3 f = {
s = (snoun2nounSep f).s ;
c2 = f.c2 ;
h = f.h ;
isPre = f.isPre
} ;
Use3N3 f = {
s = (snoun2nounSep f).s ;
c2 = f.c3 ;
h = f.h ;
isPre = f.isPre2
} ;
--- If a possessive suffix is added here it goes after the complements...
ComplN2 f x = {
s = \\nf => preOrPost f.isPre ((snoun2nounSep f).s ! nf) (appCompl True Pos f.c2 x) ;
h = f.h } ;
ComplN3 f x = {
s = \\nf => preOrPost f.isPre (f.s ! nf) (appCompl True Pos f.c2 x) ;
c2 = f.c3 ;
h = f.h ;
isPre = f.isPre2
} ;
AdjCN ap cn = {
s = \\nf => ap.s ! True ! (n2nform nf) ++ cn.s ! nf ;
h = cn.h } ;
RelCN cn rs = {s = \\nf => cn.s ! nf ++ rs.s ! agrP3 (numN nf) ;
h = cn.h } ;
RelNP np rs = {
s = \\c => np.s ! c ++ "," ++ rs.s ! np.a ;
a = np.a ;
isPron = np.isPron ; ---- correct ?
isNeg = np.isNeg
} ;
AdvCN cn ad = {s = \\nf => cn.s ! nf ++ ad.s ;
h = cn.h} ;
SentCN cn sc = {s = \\nf=> cn.s ! nf ++ sc.s;
h = cn.h } ;
ApposCN cn np = {s = \\nf=> cn.s ! nf ++ np.s ! NPCase Nom ;
h = cn.h } ; --- luvun x
PossNP cn np = {s = \\nf => np.s ! NPCase Gen ++ cn.s ! nf ;
h = cn.h
} ;
PartNP cn np = {s = \\nf => cn.s ! nf ++ np.s ! NPCase Part ;
h = cn.h ---- gives "lasin viiniänsa" ; should be "lasinsa viiniä"
} ;
CountNP det np =
let
n : Number = case det.isNum of {
True => Sg ;
_ => det.n
} ;
in {
s = \\c => let k = npform2case n c in
det.sp ! k ++ np.s ! NPCase Elat ; -- cf DetNP above
a = agrP3 (case det.isDef of {
False => Sg ; -- autoja menee; kolme autoa menee
_ => det.n
}) ;
isPron = False ; isNeg = det.isNeg
} ;
oper
numN : NForm -> Number = \nf -> case nf of {
NCase n _ => n ;
_ => Sg ---
} ;
}

184
lib/src/finnish/stemmed/NumeralFin.gf Normal file
View File

@@ -0,0 +1,184 @@
concrete NumeralFin of Numeral = CatFin [Numeral,Digits] ** open Prelude, ParadigmsFin, MorphoFin, StemFin in {
-- Notice: possessive forms are not used. They get wrong, since every
-- part is made to agree in them.
flags optimize = all_subs ;
lincat
Sub1000000 = {s : CardOrd => Str ; n : MorphoFin.Number} ;
Digit = {s : CardOrd => Str} ;
Sub10, Sub100, Sub1000 = {s : NumPlace => CardOrd => Str ; n : MorphoFin.Number} ;
lin
num x = x ;
n2 = co
(nhn (mkSubst "a" "kaksi" "kahde" "kahte" "kahta" "kahteen" "kaksi" "kaksi"
"kaksien" "kaksia" "kaksiin"))
(ordN "a" "kahdes") ; --- toinen
n3 = co
(nhn (mkSubst "a" "kolme" "kolme" "kolme" "kolmea" "kolmeen" "kolmi" "kolmi"
"kolmien" "kolmia" "kolmiin"))
(ordN "a" "kolmas") ;
n4 = co (snoun2nounBind (mkN "neljä")) (ordN "ä" "neljäs") ;
n5 = co (snoun2nounBind (mkN "viisi" "viiden" "viisiä")) (ordN "ä" "viides") ;
n6 = co (snoun2nounBind (mkN "kuusi" "kuuden" "kuusia")) (ordN "a" "kuudes") ;
n7 = co
(nhn (mkSubst "ä" "seitsemän" "seitsemä" "seitsemä" "seitsemää"
"seitsemään" "seitsemi" "seitsemi" "seitsemien" "seitsemiä"
"seitsemiin"))
(ordN "ä" "seitsemäs") ;
n8 = co
(nhn (mkSubst "a" "kahdeksan" "kahdeksa" "kahdeksa" "kahdeksaa"
"kahdeksaan" "kahdeksi" "kahdeksi" "kahdeksien" "kahdeksia"
"kahdeksiin"))
(ordN "a" "kahdeksas") ;
n9 = co
(nhn (mkSubst "ä" "yhdeksän" "yhdeksä" "yhdeksä" "yhdeksää"
"yhdeksään" "yhdeksi" "yhdeksi" "yhdeksien" "yhdeksiä" "yhdeksiin"))
(ordN "ä" "yhdeksäs") ;
pot01 =
{s = table {
NumAttr => \\_ => [] ;
NumIndep => yksiN.s
} ;
n = Sg
} ;
pot0 d = {n = Pl ; s = \\_ => d.s} ;
pot110 =
{s = \\_ => kymmenenN.s ;
n = Pl
} ;
pot111 = {n = Pl ; s = \\_,c => yksiN.s ! c ++ BIND ++ "toista"} ; ---- yhdes
pot1to19 d = {n = Pl ; s = \\_,c => d.s ! c ++ BIND ++ "toista"} ;
pot0as1 n = n ;
pot1 d = {n = Pl ; s = \\_,c => d.s ! c ++ BIND ++ kymmentaN.s ! c} ;
pot1plus d e = {
n = Pl ;
s = \\_,c => d.s ! c ++ BIND ++ kymmentaN.s ! c ++ BIND ++ e.s ! NumIndep ! c
} ;
pot1as2 n = n ;
pot2 d = {n = Pl ; s = \\_,c => d.s ! NumAttr ! c ++ nBIND d.n ++ sataaN.s ! d.n ! c} ; ----
pot2plus d e = {
n = Pl ;
s = \\_,c => d.s ! NumAttr ! c ++ nBIND d.n ++ sataaN.s ! d.n ! c ++
BIND ++ e.s ! NumIndep ! c
} ;
pot2as3 n = {n = n.n ; s = n.s ! NumIndep} ;
pot3 d = {n = Pl ; s = \\c => d.s ! NumAttr ! c ++ nBIND d.n ++ tuhattaN.s ! d.n ! c} ; ----
pot3plus d e = {
n = Pl ;
s = \\c => d.s ! NumAttr ! c ++ nBIND d.n ++ tuhattaN.s ! d.n ! c ++ e.s ! NumIndep ! c
} ;
oper
-- co : (c,o : {s : NForm => Str}) -> {s : CardOrd => Str} = \c,o -> {
co : (c,o : CommonNoun) -> {s : CardOrd => Str} = \c,o -> {
s = table {
NCard nf => c.s ! nf ;
NOrd nf => o.s ! nf
}
} ;
nBIND : MorphoFin.Number -> Str = \n -> case n of {Sg => [] ; _ => BIND} ; -- no BIND after silent 1
-- Too much trouble to infer vowel, cf. "kuudes" vs. "viides".
-- ordN : Str -> Str -> {s : NForm => Str} = \a,sadas ->
ordN : Str -> Str -> CommonNoun = \a,sadas ->
let
sada = init sadas
in
snoun2nounBind (mkN
sadas (sada + "nnen") (sada + "tt" + a) (sada + "nten" + a) (sada + "nteen")
(sada + "nsien") (sada + "nsi" + a) (sada + "nsin" + a)
(sada + "nsiss" + a) (sada + "nsiin")) ;
param
NumPlace = NumIndep | NumAttr ;
oper
yksiN = co
(nhn (mkSubst "ä" "yksi" "yhde" "yhte" "yhtä" "yhteen" "yksi" "yksi"
"yksien" "yksiä" "yksiin"))
(ordN "ä" "yhdes") ; ---- ensimmäinen
kymmenenN = co
(nhn (mkSubst "ä" "kymmenen" "kymmene" "kymmene" "kymmentä"
"kymmeneen" "kymmeni" "kymmeni" "kymmenien" "kymmeniä" "kymmeniin"))
(ordN "ä" "kymmenes") ;
sataN = co
(snoun2nounBind (mkN "sata"))
(ordN "a" "sadas") ;
tuhatN = co
(snoun2nounBind (mkN "tuhat" "tuhannen" "tuhatta" "ruhantena" "tuhanteen"
"tuhansien" "tuhansia" "tuhansina" "tuhansissa" "tuhansiin"))
(ordN "a" "tuhannes") ;
kymmentaN =
{s = table {
NCard (NCase Sg Nom) => "kymmentä" ;
k => kymmenenN.s ! k
}
} ;
sataaN : {s : MorphoFin.Number => CardOrd => Str} = {s = table {
Sg => sataN.s ;
Pl => table {
NCard (NCase Sg Nom) => "sataa" ;
k => sataN.s ! k
}
}
} ;
tuhattaN = {s = table {
Sg => tuhatN.s ;
Pl => table {
NCard (NCase Sg Nom) => "tuhatta" ;
k => tuhatN.s ! k
}
}
} ;
lincat
Dig = TDigit ;
lin
IDig d = d ;
IIDig d i = {
s = \\o => d.s ! NCard (NCase Sg Nom) ++ i.s ! o ;
n = Pl
} ;
D_0 = mkDig "0" ;
D_1 = mk3Dig "1" "1." MorphoFin.Sg ;
D_2 = mkDig "2" ;
D_3 = mkDig "3" ;
D_4 = mkDig "4" ;
D_5 = mkDig "5" ;
D_6 = mkDig "6" ;
D_7 = mkDig "7" ;
D_8 = mkDig "8" ;
D_9 = mkDig "9" ;
oper
mk2Dig : Str -> Str -> TDigit = \c,o -> mk3Dig c o MorphoFin.Pl ;
mkDig : Str -> TDigit = \c -> mk2Dig c (c + ".") ;
mk3Dig : Str -> Str -> MorphoFin.Number -> TDigit = \c,o,n -> {
s = table {NCard _ => c ; NOrd _ => o} ;
n = n
} ;
TDigit = {
n : MorphoFin.Number ;
s : CardOrd => Str
} ;
}

669
lib/src/finnish/stemmed/ParadigmsFin.gf Normal file
View File

@@ -0,0 +1,669 @@
--# -path=.:../abstract:../common:../../prelude
--1 Finnish Lexical Paradigms
--
-- Aarne Ranta 2003--2008
--
-- This is an API to the user of the resource grammar
-- for adding lexical items. It gives functions for forming
-- expressions of open categories: nouns, adjectives, verbs.
--
-- Closed categories (determiners, pronouns, conjunctions) are
-- accessed through the resource syntax API and $Structural.gf$.
--
-- The main difference with $MorphoFin.gf$ is that the types
-- referred to are compiled resource grammar types. We have moreover
-- had the design principle of always having existing forms, rather
-- than stems, as string arguments of the paradigms.
--
-- The structure of functions for each word class $C$ is the following:
-- there is a polymorphic constructor $mkC$, which takes one or
-- a few arguments. In Finnish, one argument is enough in 80-90% of
-- cases in average.
resource ParadigmsFin = open
(Predef=Predef),
Prelude,
MorphoFin,
CatFin, StemFin
in {
flags optimize=noexpand ;
--2 Parameters
--
-- To abstract over gender, number, and (some) case names,
-- we define the following identifiers. The application programmer
-- should always use these constants instead of the constructors
-- defined in $ResFin$.
oper
Number : Type ;
singular : Number ;
plural : Number ;
Case : Type ;
nominative : Case ; -- e.g. "talo"
genitive : Case ; -- e.g. "talon"
partitive : Case ; -- e.g. "taloa"
essive : Case ; -- e.g. "talona"
translative : Case ; -- e.g. "taloksi"
inessive : Case ; -- e.g. "talossa"
elative : Case ; -- e.g. "talosta"
illative : Case ; -- e.g. "taloon"
adessive : Case ; -- e.g. "talolla"
ablative : Case ; -- e.g. "talolta"
allative : Case ; -- e.g. "talolle"
infFirst : InfForm ; -- e.g. "tehdä"
infElat : InfForm ; -- e.g. "tekemästä"
infIllat : InfForm ; -- e.g. "tekemään"
-- The following type is used for defining *rection*, i.e. complements
-- of many-place verbs and adjective. A complement can be defined by
-- just a case, or a pre/postposition and a case.
prePrep : Case -> Str -> Prep ; -- preposition, e.g. partitive "ilman"
postPrep : Case -> Str -> Prep ; -- postposition, e.g. genitive "takana"
postGenPrep : Str -> Prep ; -- genitive postposition, e.g. "takana"
casePrep : Case -> Prep ; -- just case, e.g. adessive
NK : Type ; -- Noun from DictFin (Kotus)
AK : Type ; -- Adjective from DictFin (Kotus)
VK : Type ; -- Verb from DictFin (Kotus)
AdvK : Type ; -- Adverb from DictFin (Kotus)
--2 Nouns
-- The worst case gives ten forms.
-- In practice just a couple of forms are needed to define the different
-- stems, vowel alternation, and vowel harmony.
oper
-- The regular noun heuristic takes just one form (singular
-- nominative) and analyses it to pick the correct paradigm.
-- It does automatic grade alternation, and is hence not usable
-- for words like "auto" (whose genitive would become "audon").
--
-- If the one-argument paradigm does not give the correct result, one can try and give
-- two or three forms. Most notably, the two-argument variant is used
-- for nouns like "kivi - kiviä", which would otherwise become like
-- "rivi - rivejä". Three arguments are used e.g. for
-- "auto - auton - autoja", which would otherwise become
-- "auto - audon".
mkN : overload {
mkN : (kukko : Str) -> N ; -- predictable nouns, covers 82%
mkN : (savi,savia : Str) -> N ; -- different pl.part
mkN : (vesi,veden,vesiä : Str) -> N ; -- also different sg.gen
mkN : (vesi,veden,vesiä,vettä : Str) -> N ; -- also different sg.part
mkN : (olo,n,a,na,oon,jen,ja,ina,issa,ihin : Str) -> N ; -- worst case, 10 forms
mkN : (pika : Str) -> (juna : N) -> N ; -- compound with invariable prefix
mkN : (oma : N) -> (tunto : N) -> N ; -- compound with inflecting prefix
mkN : NK -> N ; -- noun from DictFin (Kotus)
} ;
-- Some nouns are regular except for the singular nominative (e.g. "mies").
exceptNomN : N -> Str -> N ;
-- Nouns used as functions need a case, of which the default is
-- the genitive.
mkN2 : overload {
mkN2 : N -> N2 ; -- relational noun with genitive
mkN2 : N -> Prep -> N2 -- relational noun another prep.
} ;
mkN3 : N -> Prep -> Prep -> N3 ; -- relation with two complements
-- Proper names can be formed by using declensions for nouns.
-- The plural forms are filtered away by the compiler.
mkPN : overload {
mkPN : Str -> PN ; -- predictable noun made into name
mkPN : N -> PN -- any noun made into name
} ;
--2 Adjectives
-- Non-comparison one-place adjectives are just like nouns.
-- The regular adjectives are based on $regN$ in the positive.
-- Comparison adjectives have three forms.
-- The comparative and the superlative
-- are always inflected in the same way, so the nominative of them is actually
-- enough (except for the superlative "paras" of "hyvä").
mkA : overload {
mkA : Str -> A ; -- regular noun made into adjective
mkA : N -> A ; -- any noun made into adjective
mkA : N -> (kivempi,kivin : Str) -> A ; -- deviating comparison forms
mkA : (hyva,prmpi,pras : N) -> (hyvin,pmmin,prhten : Str) -> A ; -- worst case adj
mkA : AK -> A ; -- adjective from DictFin (Kotus)
} ;
-- Two-place adjectives need a case for the second argument.
mkA2 : A -> Prep -> A2 -- e.g. "jaollinen" casePrep adessive
= \a,p -> a ** {c2 = p ; lock_A2 = <>};
--2 Verbs
--
-- The grammar does not cover the potential mood and some nominal
-- forms. One way to see the coverage is to linearize a verb to
-- a table.
-- The worst case needs twelve forms, as shown in the following.
mkV : overload {
mkV : (huutaa : Str) -> V ; -- predictable verbs, covers 90%
mkV : (huutaa,huusi : Str) -> V ; -- deviating past 3sg
mkV : (huutaa,huudan,huusi : Str) -> V ; -- also deviating pres. 1sg
mkV : (huutaa,dan,taa,tavat,takaa,detaan,sin,si,sisi,tanut,dettu,tanee : Str) -> V ; -- worst-case verb
mkV : VK -> V ; -- verb from DictFin (Kotus)
mkV : V -> Str -> V ; -- hakata päälle (particle verb)
} ;
-- All the patterns above have $nominative$ as subject case.
-- If another case is wanted, use the following.
caseV : Case -> V -> V ; -- deviating subj. case, e.g. genitive "täytyä"
-- The verbs "be" is special.
vOlla : V ; -- the verb "be"
--3 Two-place verbs
--
-- Two-place verbs need an object case, and can have a pre- or postposition.
-- The default is direct (accusative) object. There is also a special case
-- with case only. The string-only argument case yields a regular verb with
-- accusative object.
mkV2 : overload {
mkV2 : Str -> V2 ; -- predictable direct transitive
mkV2 : V -> V2 ; -- direct transitive
mkV2 : V -> Case -> V2 ; -- complement just case
mkV2 : V -> Prep -> V2 ; -- complement pre/postposition
mkV2 : VK -> V2 ; -- direct transitive of Kotus verb
} ;
--3 Three-place verbs
--
-- Three-place (ditransitive) verbs need two prepositions, of which
-- the first one or both can be absent.
mkV3 : V -> Prep -> Prep -> V3 ; -- e.g. puhua, allative, elative
dirV3 : V -> Case -> V3 ; -- siirtää, (accusative), illative
dirdirV3 : V -> V3 ; -- antaa, (accusative), (allative)
--3 Other complement patterns
--
-- Verbs and adjectives can take complements such as sentences,
-- questions, verb phrases, and adjectives.
mkV0 : V -> V0 ; --%
mkVS : V -> VS ;
mkV2S : V -> Prep -> V2S ; -- e.g. "sanoa" allative
mkVV : V -> VV ; -- e.g. "alkaa"
mkVVf : V -> InfForm -> VV ; -- e.g. "ruveta" infIllat
mkV2V : V -> Prep -> V2V ; -- e.g. "käskeä" genitive
mkV2Vf : V -> Prep -> InfForm -> V2V ; -- e.g. "kieltää" partitive infElat
mkVA : V -> Prep -> VA ; -- e.g. "maistua" ablative
mkV2A : V -> Prep -> Prep -> V2A ; -- e.g. "maalata" accusative translative
mkVQ : V -> VQ ;
mkV2Q : V -> Prep -> V2Q ; -- e.g. "kysyä" ablative
mkAS : A -> AS ; --%
mkA2S : A -> Prep -> A2S ; --%
mkAV : A -> AV ; --%
mkA2V : A -> Prep -> A2V ; --%
-- Notice: categories $AS, A2S, AV, A2V$ are just $A$,
-- and the second argument is given
-- as an adverb. Likewise
-- $V0$ is just $V$.
V0 : Type ; --%
AS, A2S, AV, A2V : Type ; --%
--.
-- The definitions should not bother the user of the API. So they are
-- hidden from the document.
Case = MorphoFin.Case ;
Number = MorphoFin.Number ;
singular = Sg ;
plural = Pl ;
nominative = Nom ;
genitive = Gen ;
partitive = Part ;
translative = Transl ;
inessive = Iness ;
essive = Ess ;
elative = Elat ;
illative = Illat ;
adessive = Adess ;
ablative = Ablat ;
allative = Allat ;
infFirst = Inf1 ; infElat = Inf3Elat ; infIllat = Inf3Illat ;
prePrep : Case -> Str -> Prep =
\c,p -> {c = NPCase c ; s = p ; isPre = True ; lock_Prep = <>} ;
postPrep : Case -> Str -> Prep =
\c,p -> {c = NPCase c ; s = p ; isPre = False ; lock_Prep = <>} ;
postGenPrep p = {
c = NPCase genitive ; s = p ; isPre = False ; lock_Prep = <>} ;
casePrep : Case -> Prep =
\c -> {c = NPCase c ; s = [] ; isPre = True ; lock_Prep = <>} ;
accPrep = {c = NPAcc ; s = [] ; isPre = True ; lock_Prep = <>} ;
NK = {s : NForms ; lock_NK : {}} ;
AK = {s : NForms ; lock_AK : {}} ;
VK = {s : VForms ; lock_VK : {}} ;
AdvK = {s : Str ; lock_AdvK : {}} ;
mkN = overload {
mkN : (talo : Str) -> N = mk1N ;
-- \s -> nforms2snoun (nForms1 s) ;
mkN : (talo,talon : Str) -> N = mk2N ;
-- \s,t -> nforms2snoun (nForms2 s t) ;
mkN : (talo,talon,taloja : Str) -> N = mk3N ;
-- \s,t,u -> nforms2snoun (nForms3 s t u) ;
mkN : (talo,talon,taloja,taloa : Str) -> N = mk4N ;
-- \s,t,u,v -> nforms2snoun (nForms4 s t u v) ;
mkN :
(talo,talon,taloa,talona,taloon,talojen,taloja,taloina,taloissa,taloihin
: Str) -> N = mk10N ;
mkN : (sora : Str) -> (tie : N) -> N = mkStrN ;
mkN : (oma,tunto : N) -> N = mkNN ;
mkN : (sana : NK) -> N = \w -> nforms2snoun w.s ;
} ;
exceptNomN : N -> Str -> N = \noun,nom -> lin N {
s = table {
0 => nom ;
f => noun.s ! f
} ;
h = noun.h
} ;
---- mk1A : Str -> A = \jalo -> aForms2A (nforms2aforms (nForms1 jalo)) ;
---- mkNA : N -> A = snoun2sadj ;
mk1N : (talo : Str) -> N = \s -> nforms2snoun (nForms1 s) ;
mk2N : (talo,talon : Str) -> N = \s,t -> nforms2snoun (nForms2 s t) ;
mk3N : (talo,talon,taloja : Str) -> N = \s,t,u -> nforms2snoun (nForms3 s t u) ;
mk4N : (talo,talon,taloa,taloja : Str) -> N = \s,t,u,v ->
nforms2snoun (nForms4 s t u v) ;
mk10N :
(talo,talon,taloa,talona,taloon,talojen,taloja,taloina,taloissa,taloihin
: Str) -> N = \a,b,c,d,e,f,g,h,i,j ->
lin N (nforms2snoun (nForms10 a b c d e f g h i j)) ;
mkStrN : Str -> N -> N = \sora,tie -> {
s = \\c => sora + tie.s ! c ;
h = tie.h ;
lock_N = <>
} ;
mkNN : N -> N -> N = \oma,tunto -> {
s = \\c => oma.s ! c + tunto.s ! c ;
h = tunto.h ;
lock_N = <>
} ; ---- TODO: oma in possessive suffix forms
nForms1 : Str -> NForms = \ukko ->
let
ukk = init ukko ;
uko = weakGrade ukko ;
ukon = uko + "n" ;
o = case last ukko of {"ä" => "ö" ; "a" => "o"} ; -- only used then
renka = strongGrade (init ukko) ;
rake = strongGrade ukko ;
in
case ukko of {
_ + "nen" => dNainen ukko ;
_ + ("aa" | "ee" | "ii" | "oo" | "uu" | "yy" |"ää"|"öö") => dPuu ukko ;
_ + ("ai" | "ei" | "oi" | "ui" | "yi" | "äi" | "öi") => dPuu ukko ;
_ + ("ie" | "uo" | "yö") => dSuo ukko ;
_ + ("ea" | "eä") => dKorkea ukko ;
_ + "is" => dKaunis ukko ;
_ + ("i" | "u") + "n" => dLiitin ukko (renka + "men") ;
_ + ("ton" | "tön") => dOnneton ukko ;
_ + "e" => dRae ukko (rake + "en") ;
_ + ("ut" | "yt") => dOttanut ukko ;
_ + ("as" | "äs") => dRae ukko (renka + last renka + "n") ;
_ + ("uus" | "yys" | "eus" | "eys") => dLujuus ukko ;
_ + "s" => dJalas ukko ;
-- {- heuristics for 3-syllable nouns ending a/ä
_ + ("a" | "e" | "i" | "o" | "u" | "y" | "ä" | "ö") + ? +
_ + "i" + ? + a@("a" | "ä") =>
dSilakka ukko (ukko + "n") (ukk + o + "it" + a) ; -- pesijä
_ + ("a" | "e" | "i" | "o" | "u" | "y" | "ä" | "ö") + ? + _ +
("a" | "e" | "o" | "u" | "y" | "ä" | "ö") +
("l" | "r" | "n") + a@("a" | "ä") =>
dSilakka ukko (ukko + "n") (ukk + o + "it" + a) ; -- sarana, omena
_ + ("a" | "e" | "i" | "o" | "u" | "y" | "ä" | "ö") + ? + _ +
("a" | "e" | "i" | "o" | "u" | "y" | "ä" | "ö") +
("n" | "k" | "s") + "k" + a@("a" | "ä") =>
dSilakka ukko (uko + "n") (init uko + o + "it" + a) ; -- silakka
_ + ("a" | "e" | "i" | "o" | "u" | "y" | "ä" | "ö") + ? + _ +
("a" | "e" | "i" | "o" | "u" | "y" | "ä" | "ö") +
("n" | "t" | "s") + "t" + a@("a" | "ä") =>
dSilakka ukko (uko + "n") (ukk + o + "j" + a) ; -- yhdyntä (but not isäntä)
_ + ("a" | "e" | "i" | "o" | "u") + ? + _ +
("a" | "e" | "o" | "u") + ? + "a" =>
dSilakka ukko (ukko + "n") (ukk + "ia") ; -- asema, johtaja
-- -}
_ + "i" +o@("o"|"ö") => dSilakka ukko (ukko+"n") (ukko+"it"+getHarmony o);
_ + "i" + "a" => dSilakka ukko (ukko + "n") (ukk + "oita") ;
_ + "i" + "ä" => dSilakka ukko (ukko + "n") (ukk + "öitä") ;
_ + ("a" | "o" | "u" | "y" | "ä" | "ö") => dUkko ukko ukon ;
_ + "i" => dPaatti ukko ukon ;
_ + ("ar" | "är") => dPiennar ukko (renka + "ren") ;
_ + "e" + ("l" | "n") => dPiennar ukko (ukko + "en") ;
_ => dUnix ukko
} ;
nForms2 : (_,_ : Str) -> NForms = \ukko,ukkoja ->
let
ukot = nForms1 ukko ;
ukon = weakGrade ukko + "n" ;
in
case <ukko,ukkoja> of {
<_, _ + ":" + ? + ("a" | "ä")> => dSDP ukko ;
<_ + "ea", _ + "oita"> =>
dSilakka ukko ukon ukkoja ; -- idea, but not korkea
<_ + ("aa" | "ee" | "ii" | "oo" | "uu" | "yy" | "ää" | "öö" |
"ie" | "uo" | "yö" | "ea" | "eä" |
"ia" | "iä" | "io" | "iö"), _ + ("a" | "ä")> =>
nForms1 ukko ; --- to protect --- how to get "dioja"?
<_ + ("a" | "ä" | "o" | "ö"), _ + ("a" | "ä")> =>
dSilakka ukko ukon ukkoja ;
<arp + "i", _ + "i" + ("a" | "ä")> =>
dArpi ukko (init (weakGrade ukko) + "en") ;
<_ + "i", _ + ("eita" | "eitä")> =>
dTohtori ukko ;
<_ + ("ut" | "yt"),_ + ("uita" | "yitä")> => dRae ukko (init ukko + "en") ;
<_ + "e", nuk + ("eja" | "ejä")> =>
dNukke ukko ukon ;
<_ + "s", _ + "ksi" + ?> => dJalas ukko ;
<_ + ("l" | "n" | "r" | "s"), _ + ("eja" | "ejä")> => dUnix ukko ;
<_, _ + ("a" | "ä")> => ukot ;
_ =>
Predef.error
(["last argument should end in a/ä, not"] ++ ukkoja)
} ;
nForms3 : (_,_,_ : Str) -> NForms = \ukko,ukon,ukkoja ->
let
ukk = init ukko ;
ukot = nForms2 ukko ukkoja ;
in
case <ukko,ukon> of {
<_, _ + ":n"> => dSDP ukko ;
<_ + ("aa" | "ee" | "ii" | "oo" | "uu" | "yy" | "ää" | "öö" |
"ie" | "uo" | "yö" | "ea" | "eä" |
"ia" | "iä" | "io" | "iö" | "ja" | "jä"), _ + "n"> =>
ukot ; --- to protect
<_ + ("a" | "o" | "u" | "y" | "ä" | "ö"), _ + "n"> =>
dSilakka ukko ukon ukkoja ; -- auto,auton
<_ + "mpi", _ + ("emman" | "emmän")> => dSuurempi ukko ;
<_ + "in", _ + ("imman" | "immän")> => dSuurin ukko ;
<terv + "e", terv + "een"> =>
dRae ukko ukon ;
<taiv + ("as" | "äs"), taiv + ("aan" | "ään")> =>
dRae ukko ukon ;
<nukk + "e", nuk + "een"> => dRae ukko ukon ;
<arp + "i", arv + "en"> => dArpi ukko ukon ;
<_ + ("us" | "ys"), _ + "den"> => dLujuus ukko ;
<_, _ + "n"> => ukot ;
_ =>
Predef.error (["second argument should end in n, not"] ++ ukon)
} ;
nForms4 : (_,_,_,_ : Str) -> NForms = \ukko,ukon,ukkoja,ukkoa ->
let
ukot = nForms3 ukko ukon ukkoja ;
in
case <ukko,ukon,ukkoja,ukkoa> of {
<_,_ + "n", _ + ("a" | "ä"), _ + ("a" | "ä")> =>
table {
2 => ukkoa ;
n => ukot ! n
} ;
_ =>
Predef.error
(["last arguments should end in n, a/ä, and a/ä, not"] ++
ukon ++ ukkoja ++ ukkoa)
} ;
mkN2 = overload {
mkN2 : N -> N2 = \n -> mmkN2 n (casePrep genitive) ;
mkN2 : N -> Prep -> N2 = mmkN2
} ;
mmkN2 : N -> Prep -> N2 = \n,c -> n ** {c2 = c ; isPre = mkIsPre c ; lock_N2 = <>} ;
mkN3 = \n,c,e -> n ** {c2 = c ; c3 = e ;
isPre = mkIsPre c ; -- matka Lontoosta Pariisiin
isPre2 = mkIsPre e ; -- Suomen voitto Ruotsista
lock_N3 = <>
} ;
mkIsPre : Prep -> Bool = \p -> case p.c of {
NPCase Gen => notB p.isPre ; -- Jussin veli (prep is <Gen,"",True>, isPre becomes False)
_ => True -- syyte Jussia vastaan, puhe Jussin puolesta
} ;
mkPN = overload {
mkPN : Str -> PN = mkPN_1 ;
mkPN : N -> PN = \s -> lin PN s ;
} ;
mkPN_1 : Str -> PN = \s -> lin PN (mk1N s) ;
-- adjectives
mkA = overload {
mkA : Str -> A = mkA_1 ;
mkA : N -> A = \n -> noun2adjDeg n ** {lock_A = <>} ;
mkA : N -> (kivempaa,kivinta : Str) -> A = \n -> regAdjective n ;
mkA : (sana : AK) -> A = \w -> noun2adjDeg (nForms2N w.s) ;
mkA : (hyva,parempi,paras : N) -> (hyvin,paremmin,parhaiten : Str) -> A = \h,p,ps,hn,pn,ph -> lin A {
s = table {
Posit => table {
SAN nf => h.s ! nf ;
SAAdv => hn
} ;
Compar => table {
SAN nf => p.s ! nf ;
SAAdv => pn
} ;
Superl => table {
SAN nf => ps.s ! nf ;
SAAdv => ph
}
} ;
h = h.h
} ;
} ;
mkA_1 : Str -> A = \x -> lin A (noun2adjDeg (mk1N x)) ;
-- auxiliaries
mkAdjective : (_,_,_ : SAdj) -> A = \hyva,parempi,paras ->
{s = table {
Posit => hyva.s ;
Compar => parempi.s ;
Superl => paras.s
} ;
h = hyva.h ; ---- different for parempi, paras
lock_A = <>
} ;
regAdjective : SNoun -> Str -> Str -> A = \kiva, kivempi, kivin ->
mkAdjective
(snoun2sadj kiva)
(snoun2sadjComp False (nforms2snoun (dSuurempi kivempi)))
(snoun2sadjComp False (nforms2snoun (dSuurin kivin))) ;
noun2adjDeg : SNoun -> A = \suuri ->
regAdjective
suuri
(suuri.s ! 1 + "mpi") ---- to check
(suuri.s ! 8 + "n") ; ----
-- verbs
mkV = overload {
mkV : (huutaa : Str) -> V = mk1V ;
mkV : (huutaa,huusi : Str) -> V = mk2V ;
mkV : (huutaa,huudan,huusi : Str) -> V = mk3V ;
mkV : (
huutaa,huudan,huutaa,huutavat,huutakaa,huudetaan,
huusin,huusi,huusisi,huutanut,huudettu,huutanee : Str) -> V = mk12V ;
mkV : (sana : VK) -> V = \w -> vforms2V w.s ** {sc = NPCase Nom ; lock_V = <> ; p = []} ;
mkV : V -> Str -> V = \w,p -> vforms2V w.s ** {sc = NPCase Nom ; lock_V = <> ; p = p} ;
} ;
mk1V : Str -> V = \s ->
let vfs = vforms2V (vForms1 s) in
vfs ** {sc = NPCase Nom ; lock_V = <> ; p = []} ;
mk2V : (_,_ : Str) -> V = \x,y ->
let vfs = vforms2V (vForms2 x y) in vfs ** {sc = NPCase Nom ; lock_V = <> ; p = []} ;
mk3V : (huutaa,huudan,huusi : Str) -> V = \x,_,y -> mk2V x y ; ----
mk12V : (
huutaa,huudan,huutaa,huutavat,huutakaa,huudetaan,
huusin,huusi,huusisi,huutanut,huudettu,huutanee : Str) -> V =
\a,b,c,d,e,f,g,h,i,j,k,l ->
vforms2V (vForms12 a b c d e f g h i j k l) ** {sc = NPCase Nom ; lock_V = <> ; p = []} ;
vForms1 : Str -> VForms = \ottaa ->
let
a = last ottaa ;
otta = init ottaa ;
ott = init otta ;
ots = init ott + "s" ;
ota = weakGrade otta ;
otin = init (strongGrade (init ott)) + "elin" ;
ot = init ota ;
in
case ottaa of {
_ + ("e" | "i" | "o" | "u" | "y" | "ö") + ("a" | "ä") =>
cHukkua ottaa (ota + "n") ;
_ + ("l" | "n" | "r") + ("taa" | "tää") =>
cOttaa ottaa (ota + "n") (ots + "in") (ots + "i") ;
("" | ?) + ("a" | "e" | "i" | "o" | "u") + ? + _ +
("a" | "e" | "i" | "o" | "u") + _ + "aa" =>
cOttaa ottaa (ota + "n") (ot + "in") (ott + "i") ;
("" | ?) + ("a" | "e" | "i") + _ + "aa" =>
cOttaa ottaa (ota + "n") (ot + "oin") (ott + "oi") ;
_ + ("aa" | "ää") =>
cOttaa ottaa (ota + "n") (ot + "in") (ott + "i") ;
_ + ("ella" | "ellä") =>
cKuunnella ottaa otin ;
_ + ("osta" | "östä") =>
cJuosta ottaa (init ott + "ksen") ;
_ + ("st" | "nn" | "ll" | "rr") + ("a" | "ä") =>
cJuosta ottaa (ott + "en") ;
_ + ("ita" | "itä") =>
cHarkita ottaa ;
_ + ("eta" | "etä" | "ota" | "ata" | "uta" | "ytä" | "ätä" | "ötä") =>
cPudota ottaa (strongGrade ott + "si") ;
_ + ("da" | "dä") =>
cJuoda ottaa ;
_ => Predef.error (["expected infinitive, found"] ++ ottaa)
} ;
vForms2 : (_,_ : Str) -> VForms = \huutaa,huusi ->
let
huuda = weakGrade (init huutaa) ;
huusin = weakGrade huusi + "n" ;
autoin = weakGrade (init huusi) + "in" ;
in
case <huutaa,huusi> of {
<_ + ("taa" | "tää"), _ + ("oi" | "öi")> =>
cOttaa huutaa (huuda + "n") autoin huusi ;
<_ + ("aa" | "ää"), _ + "i"> =>
cOttaa huutaa (huuda + "n") huusin huusi ;
<_ + ("eta" | "etä"), _ + "eni"> =>
cValjeta huutaa huusi ;
<_ + ("sta" | "stä"), _ + "si"> =>
vForms1 huutaa ; -- pestä, halkaista
<_ + ("ta" | "tä"), _ + "si"> =>
cPudota huutaa huusi ;
<_ + ("lla" | "llä"), _ + "li"> =>
cKuunnella huutaa huusin ;
_ => vForms1 huutaa
} ;
caseV c v = {s = v.s ; sc = NPCase c ; qp = v.qp ; lock_V = <> ; p = v.p} ;
vOlla = verbOlla ** {sc = NPCase Nom ; qp = True ; lock_V = <> ; p = []} ; ---- lieneekö
mk2V2 : V -> Prep -> V2 = \v,c -> v ** {c2 = c ; lock_V2 = <>} ;
caseV2 : V -> Case -> V2 = \v,c -> mk2V2 v (casePrep c) ;
dirV2 v = mk2V2 v accPrep ;
mkAdv = overload {
mkAdv : Str -> Adv = \s -> {s = s ; lock_Adv = <>} ;
mkAdv : AdvK -> Adv = \s -> {s = s.s ; lock_Adv = <>} ;
} ;
mkV2 = overload {
mkV2 : Str -> V2 = \s -> dirV2 (mk1V s) ;
mkV2 : V -> V2 = dirV2 ;
mkV2 : V -> Case -> V2 = caseV2 ;
mkV2 : V -> Prep -> V2 = mk2V2 ;
mkV2 : VK -> V2 = \w -> dirV2 (vforms2V w.s ** {sc = NPCase Nom ; lock_V = <> ; p = []}) ;
} ;
mk2V2 : V -> Prep -> V2 ;
caseV2 : V -> Case -> V2 ;
dirV2 : V -> V2 ;
mkV3 v p q = v ** {c2 = p ; c3 = q ; lock_V3 = <>} ;
dirV3 v p = mkV3 v accPrep (casePrep p) ;
dirdirV3 v = dirV3 v allative ;
mkVS v = v ** {lock_VS = <>} ;
mkVV v = mkVVf v infFirst ;
mkVVf v f = v ** {vi = f ; lock_VV = <>} ;
mkVQ v = v ** {lock_VQ = <>} ;
V0 : Type = V ;
AS, A2S, AV : Type = A ;
A2V : Type = A2 ;
mkV0 v = v ** {lock_V = <>} ;
mkV2S v p = mk2V2 v p ** {lock_V2S = <>} ;
mkV2V v p = mkV2Vf v p infIllat ;
mkV2Vf v p f = mk2V2 v p ** {vi = f ; lock_V2V = <>} ;
mkVA v p = v ** {c2 = p ; lock_VA = <>} ;
mkV2A v p q = v ** {c2 = p ; c3 = q ; lock_V2A = <>} ;
mkV2Q v p = mk2V2 v p ** {lock_V2Q = <>} ;
mkAS v = v ** {lock_A = <>} ;
mkA2S v p = mkA2 v p ** {lock_A = <>} ;
mkAV v = v ** {lock_A = <>} ;
mkA2V v p = mkA2 v p ** {lock_A2 = <>} ;
} ;

132
lib/src/finnish/stemmed/StemFin.gf Normal file
View File

@@ -0,0 +1,132 @@
-- stem-based morphology, with just the minimum number of concatenative stems
resource StemFin = open MorphoFin, Prelude in {
flags coding = utf8 ;
oper
SNForm : Type = Predef.Ints 9 ;
SNoun : Type = {s : SNForm => Str ; h : Harmony} ;
-- SVerb : Type ;
nforms2snoun : NForms -> SNoun = \nf -> {
s = table {
0 => nf ! 0 ; -- ukko
1 => Predef.tk 1 (nf ! 1) ; -- uko(n)
2 => nf ! 2 ; -- ukkoa
3 => nf ! 3 ; -- ukkona
4 => Predef.tk 1 (nf ! 4) ; -- ukkoo(n)
5 => Predef.tk 1 (nf ! 5) ; -- ukkoje(n)
6 => nf ! 6 ; -- ukkoja
7 => Predef.tk 2 (nf ! 7) ; -- ukkoi(na)
8 => Predef.tk 3 (nf ! 8) ; -- ukoi(ssa)
9 => Predef.tk 1 (nf ! 9) -- ukkoihi(n)
} ;
h = aHarmony (last (nf ! 2)) ;
} ;
snoun2nounBind : SNoun -> Noun = snoun2noun True ;
snoun2nounSep : SNoun -> Noun = snoun2noun False ;
snoun2noun : Bool -> SNoun -> Noun = \b,sn ->
let
plus : Str -> Str -> Str = \x,y -> case b of {
True => x + y ;
False => glue x y
} ;
f = sn.s ;
ukko = f ! 0 ;
uko = f ! 1 ;
ukkoa = f ! 2 ;
ukkona = f ! 3 ;
ukkoo = f ! 4 ;
ukkoje = f ! 5 ;
ukkoja = f ! 6 ;
ukkoi = f ! 7 ;
ukoi = f ! 8 ;
ukkoihi = f ! 9 ;
a = harmonyA sn.h ;
in
{s = table {
NCase Sg Nom => ukko ;
NCase Sg Gen => plus uko "n" ;
NCase Sg Part => ukkoa ;
NCase Sg Transl => plus uko "ksi" ;
NCase Sg Ess => ukkona ;
NCase Sg Iness => plus uko ("ss" + a) ;
NCase Sg Elat => plus uko ("st" + a) ;
NCase Sg Illat => plus ukkoo "n" ;
NCase Sg Adess => plus uko ("ll" + a) ;
NCase Sg Ablat => plus uko ("lt" + a) ;
NCase Sg Allat => plus uko "lle" ;
NCase Sg Abess => plus uko ("tt" + a) ;
NCase Pl Nom => plus uko "t" ;
NCase Pl Gen => plus ukkoje "n" ;
NCase Pl Part => ukkoja ;
NCase Pl Transl => plus ukoi "ksi" ;
NCase Pl Ess => plus ukkoi ("n" + a) ;
NCase Pl Iness => plus ukoi ("ss" + a) ;
NCase Pl Elat => plus ukoi ("st" + a) ;
NCase Pl Illat => plus ukkoihi "n" ;
NCase Pl Adess => plus ukoi ("ll" + a) ;
NCase Pl Ablat => plus ukoi ("lt" + a) ;
NCase Pl Allat => plus ukoi "lle" ;
NCase Pl Abess => plus ukoi ("tt" + a) ;
NComit => plus ukkoi "ne" ;
NInstruct => plus ukoi "n" ;
NPossNom _ => ukko ;
NPossGen Sg => ukko ;
NPossGen Pl => ukkoje ;
NPossTransl Sg => plus uko "kse" ;
NPossTransl Pl => plus ukoi "kse" ;
NPossIllat Sg => ukkoo ;
NPossIllat Pl => ukkoihi
} ;
h = sn.h ;
lock_N = <>
} ;
snoun2np : Number -> SNoun -> NPForm => Str = \n,sn ->
\\c => (snoun2nounSep sn).s ! NCase n (npform2case n c) ;
noun2snoun : Noun -> SNoun = \n -> nforms2snoun (n2nforms n) ;
aHarmony : Str -> Harmony = \a -> case a of
{"a" => Back ; _ => Front} ;
harmonyA : Harmony -> Str = \h -> case h of
{Back => "a" ; Front => "ä"} ;
-- Adjectives
param
SAForm = SAN SNForm | SAAdv ;
oper
SAdj = {s : SAForm => Str ; h : Harmony} ;
snoun2sadj : SNoun -> SAdj = snoun2sadjComp True ;
snoun2sadjComp : Bool -> SNoun -> SAdj = \isPos,tuore ->
let
tuoree = tuore.s ! 1 ;
tuoreesti = tuoree + "sti" ;
tuoreemmin = init tuoree ;
in {s = table {
SAN f => tuore.s ! f ;
SAAdv => if_then_Str isPos tuoreesti tuoreemmin
} ;
h = tuore.h
} ;
}

315
lib/src/finnish/stemmed/StructuralFin.gf Normal file
View File

@@ -0,0 +1,315 @@
concrete StructuralFin of Structural = CatFin **
open MorphoFin, ParadigmsFin, (X = ConstructX), MakeStructuralFin, StemFin, Prelude in {
flags optimize=all ;
lin
above_Prep = postGenPrep "yläpuolella" ;
after_Prep = postGenPrep "jälkeen" ;
all_Predet = {s = \\n,c =>
let
kaiket = caseTable n (snoun2nounBind (mkN "kaikki" "kaiken" "kaikkena"))
in
case npform2case n c of {
Nom => "kaikki" ;
k => kaiket ! k
}
} ;
almost_AdA, almost_AdN = ss "melkein" ;
although_Subj = ss "vaikka" ;
always_AdV = ss "aina" ;
and_Conj = {s1 = [] ; s2 = "ja" ; n = Pl} ;
because_Subj = ss "koska" ;
before_Prep = prePrep partitive "ennen" ;
behind_Prep = postGenPrep "takana" ;
between_Prep = postGenPrep "välissä" ;
both7and_DConj = sd2 "sekä" "että" ** {n = Pl} ;
but_PConj = ss "mutta" ;
by8agent_Prep = postGenPrep "toimesta" ;
by8means_Prep = casePrep adessive ;
can8know_VV = mkVV (mkV "osata" "osasi") ;
can_VV = mkVV (mkV "voida" "voi") ;
during_Prep = postGenPrep "aikana" ;
either7or_DConj = sd2 "joko" "tai" ** {n = Sg} ;
everybody_NP = makeNP (snoun2nounBind (mkN "jokainen")) Sg ;
every_Det = mkDet Sg (snoun2nounBind (mkN "jokainen")) ;
everything_NP = makeNP (((snoun2nounBind (mkN "kaikki" "kaiken" "kaikkena"))) **
{lock_N = <>}) Sg ;
everywhere_Adv = ss "kaikkialla" ;
few_Det = mkDet Sg (snoun2nounBind (mkN "harva")) ;
--- first_Ord = {s = \\n,c => (mkN "ensimmäinen").s ! NCase n c} ;
for_Prep = casePrep allative ;
from_Prep = casePrep elative ;
he_Pron = mkPronoun "hän" "hänen" "häntä" "hänenä" "häneen" Sg P3 ;
here_Adv = ss "täällä" ;
here7to_Adv = ss "tänne" ;
here7from_Adv = ss "täältä" ;
how_IAdv = ss "miten" ;
how8much_IAdv = ss "kuinka paljon" ;
how8many_IDet =
{s = \\c => "kuinka" ++ (snoun2nounBind (mkN "moni" "monia")).s ! NCase Sg c ; n = Sg ; isNum = False} ;
if_Subj = ss "jos" ;
in8front_Prep = postGenPrep "edessä" ;
i_Pron = mkPronoun "minä" "minun" "minua" "minuna" "minuun" Sg P1 ;
in_Prep = casePrep inessive ;
it_Pron = {
s = \\c => pronSe.s ! npform2case Sg c ;
a = agrP3 Sg ;
isPron = False
} ;
less_CAdv = X.mkCAdv "vähemmän" "kuin" ;
many_Det = mkDet Sg (snoun2nounBind (mkN "moni" "monia")) ;
more_CAdv = X.mkCAdv "enemmän" "kuin" ;
most_Predet = {s = \\n,c => (nForms2N (dSuurin "useinta")).s ! NCase n (npform2case n c)} ;
much_Det = mkDet Sg {s = \\_ => "paljon" ; h = Back} ; --Harmony not relevant, it's just a CommonNoun
must_VV = mkVV (caseV genitive (mkV "täytyä")) ;
no_Utt = ss "ei" ;
on_Prep = casePrep adessive ;
--- one_Quant = mkDet Sg DEPREC
only_Predet = {s = \\_,_ => "vain"} ;
or_Conj = {s1 = [] ; s2 = "tai" ; n = Sg} ;
otherwise_PConj = ss "muuten" ;
part_Prep = casePrep partitive ;
please_Voc = ss ["ole hyvä"] ; --- number
possess_Prep = casePrep genitive ;
quite_Adv = ss "melko" ;
she_Pron = mkPronoun "hän" "hänen" "häntä" "hänenä" "häneen" Sg P3 ;
so_AdA = ss "niin" ;
somebody_NP = {
s = \\c => jokuPron ! Sg ! npform2case Sg c ;
a = agrP3 Sg ;
isPron = False ; isNeg = False
} ;
someSg_Det = heavyDet {
s1 = jokuPron ! Sg ;
s2 = \\_ => [] ;
isNum,isPoss = False ; isDef = True ; isNeg = False ; n = Sg
} ;
somePl_Det = heavyDet {
s1 = jokuPron ! Pl ;
s2 = \\_ => [] ; isNum,isPoss = False ; isNeg = False ; isDef = True ;
n = Pl ; isNeg = False
} ;
something_NP = {
s = \\c => jokinPron ! Sg ! npform2case Sg c ;
a = agrP3 Sg ;
isPron = False ; isNeg = False ; isNeg = False
} ;
somewhere_Adv = ss "jossain" ;
that_Quant = heavyQuant {
s1 = table (MorphoFin.Number) {
Sg => table (MorphoFin.Case) {
c => (mkPronoun "tuo" "tuon" "tuota" "tuona" "tuohon" Sg P3).s ! NPCase c
} ;
Pl => table (MorphoFin.Case) {
c => (mkPronoun "nuo" "noiden" "noita" "noina" "noihin" Sg P3).s ! NPCase c
}
} ;
s2 = \\_ => [] ; isNum,isPoss = False ; isDef = True ; isNeg = False
} ;
that_Subj = ss "että" ;
there_Adv = ss "siellä" ; --- tuolla
there7to_Adv = ss "sinne" ;
there7from_Adv = ss "sieltä" ;
therefore_PConj = ss "siksi" ;
they_Pron = mkPronoun "he" "heidän" "heitä" "heinä" "heihin" Pl P3 ; --- ne
this_Quant = heavyQuant {
s1 = table (MorphoFin.Number) {
Sg => table (MorphoFin.Case) {
c => (mkPronoun "tämä" "tämän" "tätä" "tänä" "tähän" Sg P3).s ! NPCase c
} ;
Pl => table (MorphoFin.Case) {
c => (mkPronoun "nämä" "näiden" "näitä" "näinä" "näihin" Sg P3).s ! NPCase c
}
} ;
s2 = \\_ => [] ; isNum,isPoss = False ; isDef = True ; isNeg = False
} ;
through_Prep = postGenPrep "kautta" ;
too_AdA = ss "liian" ;
to_Prep = casePrep illative ; --- allative
under_Prep = postGenPrep "alla" ;
very_AdA = ss "erittäin" ;
want_VV = mkVV (mkV "tahtoa") ;
we_Pron = mkPronoun "me" "meidän" "meitä" "meinä" "meihin" Pl P1 ;
whatPl_IP = {
s = table {NPAcc => "mitkä" ; c => mikaInt ! Pl ! npform2case Pl c} ;
n = Pl
} ;
whatSg_IP = {
s = \\c => mikaInt ! Sg ! npform2case Sg c ;
n = Sg
} ;
when_IAdv = ss "milloin" ;
when_Subj = ss "kun" ;
where_IAdv = ss "missä" ;
which_IQuant = {
s = mikaInt
} ;
whoSg_IP = {
s = table {NPAcc => "kenet" ; c => kukaInt ! Sg ! npform2case Sg c} ;
n = Sg
} ;
whoPl_IP = {
s = table {NPAcc => "ketkä" ; c => kukaInt ! Pl ! npform2case Pl c} ;
n = Pl
} ;
why_IAdv = ss "miksi" ;
without_Prep = prePrep partitive "ilman" ;
with_Prep = postGenPrep "kanssa" ;
yes_Utt = ss "kyllä" ;
youSg_Pron = mkPronoun "sinä" "sinun" "sinua" "sinuna" "sinuun" Sg P2 ;
youPl_Pron = mkPronoun "te" "teidän" "teitä" "teinä" "teihin" Pl P2 ;
youPol_Pron =
let p = mkPronoun "te" "teidän" "teitä" "teinä" "teihin" Pl P2 in
{s = p.s ; a = AgPol} ;
oper
jokuPron : MorphoFin.Number => (MorphoFin.Case) => Str =
let
kui = snoun2nounBind (mkN "kuu")
in
table {
Sg => table {
Nom => "joku" ;
Gen => "jonkun" ;
c => relPron ! Sg ! c + "ku" + Predef.drop 3 (kui.s ! NCase Sg c)
} ;
Pl => table {
Nom => "jotkut" ;
c => relPron ! Pl ! c + kui.s ! NCase Pl c
}
} ;
jokinPron : MorphoFin.Number => (MorphoFin.Case) => Str =
table {
Sg => table {
Nom => "jokin" ;
Gen => "jonkin" ;
c => relPron ! Sg ! c + "kin"
} ;
Pl => table {
Nom => "jotkin" ;
c => relPron ! Pl ! c + "kin"
}
} ;
mikaInt : MorphoFin.Number => (MorphoFin.Case) => Str =
let {
mi = snoun2nounBind (mkN "mi")
} in
table {
Sg => table {
Nom => "mikä" ;
Gen => "minkä" ;
Part => "mitä" ;
Illat => "mihin" ;
c => mi.s ! NCase Sg c
} ;
Pl => table {
Nom => "mitkä" ;
Gen => "minkä" ;
Part => "mitä" ;
Illat => "mihin" ;
c => mi.s ! NCase Sg c
}
} ;
kukaInt : MorphoFin.Number => (MorphoFin.Case) => Str =
let
kuka = snoun2nounBind (mkN "kuka" "kenen" "ketä" "kenä" "keneen"
"keiden" "keitä" "keinä" "keissä" "keihin") ;
in
table {
Sg => table {
c => kuka.s ! NCase Sg c
} ;
Pl => table {
Nom => "ketkä" ;
c => kuka.s ! NCase Pl c
}
} ;
mikaanPron : MorphoFin.Number => (MorphoFin.Case) => Str = \\n,c =>
case <n,c> of {
<Sg,Nom> => "mikään" ;
<_,Part> => "mitään" ;
<Sg,Gen> => "minkään" ;
<Pl,Nom> => "mitkään" ;
<Pl,Gen> => "mittenkään" ;
<_,Ess> => "minään" ;
<_,Iness> => "missään" ;
<_,Elat> => "mistään" ;
<_,Adess> => "millään" ;
<_,Ablat> => "miltään" ;
_ => mikaInt ! n ! c + "kään"
} ;
kukaanPron : MorphoFin.Number => (MorphoFin.Case) => Str =
table {
Sg => table {
Nom => "kukaan" ;
Part => "ketään" ;
Ess => "kenään" ;
Iness => "kessään" ;
Elat => "kestään" ;
Illat => "kehenkään" ;
Adess => "kellään" ;
Ablat => "keltään" ;
c => kukaInt ! Sg ! c + "kään"
} ;
Pl => table {
Nom => "ketkään" ;
Part => "keitään" ;
Ess => "keinään" ;
Iness => "keissään" ;
Elat => "keistään" ;
Adess => "keillään" ;
Ablat => "keiltään" ;
c => kukaInt ! Pl ! c + "kään"
}
} ;
oper
makeNP : N -> MorphoFin.Number -> CatFin.NP ;
makeNP noun num = {
s = \\c => noun.s ! NCase num (npform2case num c) ;
a = agrP3 num ;
isPron, isNeg = False ;
lock_NP = <>
} ;
lin
not_Predet = {s = \\_,_ => "ei"} ;
no_Quant = heavyQuant {
s1 = \\n,c => mikaanPron ! n ! c ; -- requires negative or question polarity
s2 = \\_ => [] ; isNum,isPoss = False ; isDef = True ; isNeg = True
} ;
if_then_Conj = {s1 = "jos" ; s2 = "niin" ; n = Sg} ;
nobody_NP = {
s = \\c => kukaanPron ! Sg ! npform2case Sg c ; -- requires negative or question polarity
a = agrP3 Sg ;
isPron = False ; isNeg = True
} ;
nothing_NP = {
s = \\c => mikaanPron ! Sg ! npform2case Sg c ; --- requires negative or question polarity
a = agrP3 Sg ;
isPron = False ; isNeg = True
} ;
at_least_AdN = ss "vähintään" ;
at_most_AdN = ss "enintään" ;
as_CAdv = X.mkCAdv "yhtä" "kuin" ;
except_Prep = postPrep partitive "lukuunottamatta" ;
have_V2 = mkV2 (caseV adessive vOlla) ;
lin language_title_Utt = ss "suomi" ;
}

145
lib/src/finnish/stemmed/VerbFin.gf Normal file
View File

@@ -0,0 +1,145 @@
--1 Verb Phrases in Finnish
concrete VerbFin of Verb = CatFin ** open Prelude, ResFin in {
flags optimize=all_subs ;
lin
UseV = predV ;
SlashV2a v = predV v ** {c2 = v.c2} ;
Slash2V3 v np =
insertObj
(\\fin,b,_ => appCompl fin b v.c2 np) (predV v) ** {c2 = v.c3} ;
Slash3V3 v np =
insertObj
(\\fin,b,_ => appCompl fin b v.c3 np) (predV v) ** {c2 = v.c2} ;
ComplVV v vp =
insertObj
(\\_,b,a => infVP v.sc b a vp v.vi)
(predV {s = v.s ;
sc = case vp.sc of {
NPCase Nom => v.sc ; -- minun täytyy pestä auto
c => c -- minulla täytyy olla auto
} ;
qp = v.qp ; p = v.p
}
) ;
ComplVS v s = insertExtrapos (etta_Conj ++ s.s) (predV v) ;
ComplVQ v q = insertExtrapos ( q.s) (predV v) ;
ComplVA v ap =
insertObj
(\\_,b,agr =>
let n = (complNumAgr agr) in
ap.s ! False ! (NCase n (npform2case n v.c2.c))) --- v.cs.s ignored
(predV v) ;
SlashV2S v s =
insertExtrapos (etta_Conj ++ s.s) (predV v) ** {c2 = v.c2} ;
SlashV2Q v q =
insertExtrapos (q.s) (predV v) ** {c2 = v.c2} ;
SlashV2V v vp =
insertObj (\\_,b,a => infVP v.sc b a vp v.vi) (predV v) ** {c2 = v.c2} ;
---- different infinitives
SlashV2A v ap =
insertObj
(\\fin,b,_ =>
ap.s ! False ! (NCase Sg (npform2case Sg v.c3.c))) ----agr to obj
(predV v) ** {c2 = v.c2} ;
ComplSlash vp np = insertObjPre np.isNeg (\\fin,b,_ => appCompl fin b vp.c2 np) vp ;
UseComp comp =
insertObj (\\_,_ => comp.s) (predV (verbOlla ** {sc = NPCase Nom ; qp = True ; p = []})) ;
UseCopula = predV (verbOlla ** {sc = NPCase Nom ; qp = True ; p = []}) ;
SlashVV v vp =
insertObj
(\\_,b,a => infVP v.sc b a vp v.vi)
(predV {s = v.s ;
sc = case vp.sc of {
NPCase Nom => v.sc ; -- minun täytyy pestä auto
c => c -- minulla täytyy olla auto
} ;
qp = v.qp ; p = v.p
}
) ** {c2 = vp.c2} ; ---- correct ??
{---- 153543936 (210912,312)
SlashV2VNP v np vp =
insertObjPre np.isNeg
(\\fin,b,a => appCompl True b v.c2 np ++ ---- fin -> stack overflow
infVP v.sc b a vp v.vi)
(predV v) ** {c2 = vp.c2} ;
----}
AdvVP vp adv = insertAdv (\\_ => adv.s) vp ;
AdVVP adv vp = insertAdv (\\_ => adv.s) vp ;
AdvVPSlash vps adv = insertAdv (\\_ => adv.s) vps ** {c2 = vps.c2} ;
AdVVPSlash adv vps = insertAdv (\\_ => adv.s) vps ** {c2 = vps.c2} ;
ReflVP v = insertObjPre False (\\fin,b,agr => appCompl fin b v.c2 (reflPron agr)) v ;
PassV2 v = let vp = predV v in {
s = \\_ => vp.s ! VIPass ;
s2 = \\_,_,_ => [] ;
adv = \\_ => [] ;
ext = [] ;
qp = v.qp ;
isNeg = False ;
sc = v.c2.c -- minut valitaan ; minua rakastetaan ; minulle kuiskataan
} ; ---- talon valitaan: should be marked like inf.
----b UseVS, UseVQ = \v -> v ** {c2 = {s = [] ; c = NPAcc ; isPre = True}} ;
CompAP ap = {
s = \\agr =>
let
n = complNumAgr agr ;
c = case n of {
Sg => Nom ; -- minä olen iso ; te olette iso
Pl => Part -- me olemme isoja ; te olette isoja
} --- definiteness of NP ?
in ap.s ! False ! (NCase n c)
} ;
CompCN cn = {
s = \\agr =>
let
n = complNumAgr agr ;
c = case n of {
Sg => Nom ; -- minä olen iso ; te olette iso
Pl => Part -- me olemme isoja ; te olette isoja
} --- definiteness of NP ?
in cn.s ! (NCase n c)
} ;
CompNP np = {s = \\_ => np.s ! NPCase Nom} ;
CompAdv a = {s = \\_ => a.s} ;
VPSlashPrep vp prep = vp ** {c2 = prep} ;
}
--2 The object case
--
-- The rules involved are ComplV2 and ComplVV above.
-- The work is done jointly in ResFin.infVP and appCompl.
-- Cases to test: l -table (to see negated forms)
--```
-- minun täytyy ostaa auto
-- PredVP (UsePron i_Pron) (ComplVV must_VV
-- (ComplV2 buy_V2 (DetCN (DetSg (SgQuant DefArt) NoOrd) (UseN car_N))))
-- minä tahdon ostaa auton
-- PredVP (UsePron i_Pron) (ComplVV want_VV
-- (ComplV2 buy_V2 (DetCN (DetSg (SgQuant DefArt) NoOrd) (UseN car_N))))
-- minulla täytyy olla auto
-- PredVP (UsePron i_Pron) (ComplVV must_VV
-- (ComplV2 have_V2 (DetCN (DetSg (SgQuant DefArt) NoOrd) (UseN car_N))))
--```
-- Unfortunately, there is no nice way to say "I want to have a car".
-- (Other than the paraphrases "I want a car" or "I want to own a car".)