some new opers in ParadigmsFin, and 200 more words in DictEngFin: out of 3220 Penn trees now 2721 are completely translated (but mostly not so well...)

2026-04-09 04:59:31 -06:00 · 2013-03-29 10:13:04 +00:00
parent 27a1811446
commit eb791244bd
5 changed files with 334 additions and 12 deletions
--- a/lib/src/finnish/stemmed/DictEngFin.gf
+++ b/lib/src/finnish/stemmed/DictEngFin.gf
@@ -4,7 +4,7 @@ concrete DictEngFin of DictEngAbs = CatFin ** open ParadigmsFin,
  (S = StructuralFin),
  (L = LexiconFin),
 --  SyntaxFin,
-  MorphoFin, ParadigmsFin, (X = ConstructX), MakeStructuralFin, Kotus, WNKotus, StemFin, Prelude in {
+  MorphoFin, ParadigmsFin, Kotus, WNKotus, StemFin, Prelude in {

 flags coding=utf8 ;

@@ -59074,6 +59074,247 @@ zymotic_A = mkWA "tsymoosiin" "liittyvä" ;

 oper tehda_V = mkV (lin VK {s = c71 "tehdä"}) ;

+-- another batch of additions, 29/3/2013
+
+lin
+absent_Prep = mkPrep "poissa" elative ;
+across_Prep = mkPrep "poikki" ;
+alabama_PN = mkPN "Alabama" ;
+albania_PN = mkPN "Albania" ;
+along_Prep = mkPrep "pitkin" ;
+already_AdV = mkAdV "jo" ;
+america_PN = mkPN "Amerikka" ;
+amid_Prep = mkPrep "keskellä" ;
+amp_Conj = mkConj "&" ;
+amsterdam_PN = mkPN "Amsterdam" ;
+apart_from_Prep = mkPrep "lisäksi" ;
+area_1_N = mkN "ala" ;
+arizona_PN = mkPN "Arizona" ;
+around_Prep = mkPrep "ympärillä" ;
+as_Subj = mkSubj "kun" ;
+as_long_as_Subj = mkSubj "niin kauan kun" ;
+as_well_as_Conj = mkConj "yhtä hyvin kuin" ;
+athens_PN = mkPN "Ateena" ;
+australia_PN = mkPN "Australia" ;
+austria_PN = mkPN (mkN "Itä" (mkN "valta")) ;
+bare_A = mkA "paljas" ;
+bear_N = mkN "karhu" ;
+because_of_Prep = mkPrep "vuoksi" ;
+begin_VV = mkVV "alkaa" ;
+below_Prep = mkPrep "alapuolella" ;
+beneath_Prep = mkPrep "alapuolella" ;
+benefit_V = mkV "hyötyä" ;
+besides_Prep = mkPrep "rinnalla" ;
+beyond_Prep = mkPrep "saavuttamattomissa" ;
+bonn_PN = mkPN "Bonn" ;
+boston_PN = mkPN "Boston" ;
+brazil_PN = mkPN "Brasilia" ;
+brew_N = mkN "käymis_tuote" ;
+britain_PN = mkPN "Britannia" ;
+but_Prep = mkPrep "paitsi" nominative ;
+but_Subj = mkSubj "mutta" ;
+california_PN = mkPN "Kalifornia" ;
+care_N = mkN "hoito" ;
+chicago_PN = mkPN "Chicago" ;
+china_PN = mkPN "Kiina" ;
+close_A = mkA "läheinen" ;
+columbia_PN = mkPN "Kolumbia" ;
+comment_V = mkV "kommentoida" ;
+communist_A = mkA "kommunistinen" ;
+complication_N = mkN "komplikaatio" ;
+connecticut_PN = mkPN "Connecticut" ;
+crystalline_A = mkA "kiteinen" ;
+cut_N = mkN "leikkaus" ;
+dallas_PN = mkPN "Dallas" ;
+default_V = mkV "laiminlyödä" ;
+denizen_N = mkN "asukas" ;
+denver_PN = mkPN "Denver" ;
+differ_V = mkV "eritä" ;
+disagree_V = mkV vOlla "eri mieltä" ;
+discretionary_A = mkA "harkinnanvarainen" ;
+discriminatory_A = mkA "erotteleva" ;
+duty_free_A = mkA "verovapaa" ;
+editorial_A = mkA "toimituksellinen" ;
+else_Adv = mkAdv "muutoin" ;
+encouragement_N = mkN "rohkaisu" ;
+ensure_VS = mkVS "varmistaa" ;
+essential_A = mkA "olennainen" ;
+europe_PN = mkPN "Eurooppa" ;
+even_though_Subj = mkSubj "vaikka" ;
+expensive_A = mkA "kallis" ;
+fact_finding_A = mkA (mkN "tietoa" (mkN "etsivä")) ;
+far_reaching_A = mkA "kauaskantoinen" ;
+feel_VS = mkVS "tuntea" ;
+feel_V = mkV "tuntea" ;
+feude_V = mkV "vihoitella" ;
+finland_PN = mkPN "Suomi" ;
+first_rate_A = mkA "ensiluokkainen" ;
+florida_PN = mkPN "Florida" ;
+for_starters_Adv = mkAdv "alkajaisiksi" ;
+foreclose_V = mkV (mkV "sulkea") "markkinoilta" ;
+four_part_A = mkA "neljänkeskeinen" ;
+frankfurt_PN = mkPN "Frankfurt" ;
+free_A = mkA "vapaa" ;
+funeral_N = mkN "hautajaistilaisuus" ;
+further_AdV = mkAdV "edelleen" ;
+georgia_PN = mkPN "Georgia" ;
+great_A = mkA "suurenmoinen" ;
+greece_PN = mkPN "Kreikka" ;
+half_Predet = mkPredet "puoliksi" ;
+heavy_handed_A = mkA (mkN "raskas" (mkN "kätinen")) ;
+help_N = mkN "apu" ;
+high_priced_A = mkA "hintava" ;
+hollywood_PN = mkPN "Hollywood" ;
+honest_A = mkA "rehellinen" ;
+hong_kong_PN = mkPN "Hong Kong" ;
+hors_de_combat_A = mkA "haavoittunut" ;
+hungary_PN = mkPN "Unkari" ;
+hyperinflation_N = mkN "hyperinflaatio" ;
+in_addition_Adv = mkAdv "lisäksi" ;
+in_addition_to_Prep = mkPrep "lisäksi" ;
+in_front_of_Prep = mkPrep "edessä" ;
+india_PN = mkPN "Intia" ;
+indianapolis_PN = mkPN "Indianapolis" ;
+indonesia_PN = mkPN "Indonesia" ;
+inside_Prep = mkPrep "sisällä" ;
+instead_of_Prep = mkPrep "sijasta" ;
+intimate_A = mkA "läheinen" ;
+israel_PN = mkPN "Israel" ;
+japan_PN = mkPN "Japani" ;
+japaneseMasc_N = mkN "japanilainen" ;
+just_AdV = mkAdV "pelkästään" ;
+just_Predet = mkPredet "pelkästään" ;
+kansas_PN = mkPN "Kansas" ;
+key_A = mkA "ratkaiseva" ;
+late_A = mkA "myöhäinen" ;
+libya_PN = mkPN "Libya" ;
+lie_1_V = mkV "maata" ;
+lie_2_V = mkV "valehdella" ;
+los_angeles_PN = mkPN "Los Angeles" ;
+louisiana_PN = mkPN "Louisiana" ;
+low_A = mkA "matala" ;
+madrid_PN = mkPN "Madrid" ;
+make_it_V = mkV "onnistua" ;
+mandatory_A = mkA "pakollinen" ;
+marketplace_N = mkN "markkinapaikka" ;
+marriage_N = mkN "avioliitto" ;
+maturity_3_N = mkN "kypsyys" ;
+maybe_Adv = mkAdv "ehkä" ;
+media_N = mkN "media" ;
+mexico_PN = mkPN "Mexico" ;
+miami_PN = mkPN "Miami" ;
+milan_PN = mkPN "Milano" ;
+minneapolis_PN = mkPN "Minneapolis" ;
+mississippi_PN = mkPN "Mississippi" ;
+mistrial_N = mkN "mistraali" ;
+mod_cons_N = mkN "mukavuus" ;
+more_than_AdN = mkAdN "yli" ;
+moscow_PN = mkPN "Moskova" ;
+namibia_PN = mkPN "Namibia" ;
+nearby_A = mkA "läheinen" ;
+neither7nor_DConj = mkConj "ei" "eikä" ;
+never_AdV = mkAdV "koskaan" ;
+nevertheless_Adv = mkAdv "kuitenkin" ;
+new_york_PN = mkPN "New York" ;
+nicaragua_PN = mkPN "Nicaragua" ;
+no_longer_AdV = mkAdV "enää" ;
+norway_PN = mkPN "Norja" ;
+often_AdV = mkAdV "usein" ;
+ohio_PN = mkPN "Ohio" ;
+oklahoma_PN = mkPN "Oklahoma" ;
+old_fashioned_A = mkA "vanhanaikainen" ;
+once_Subj = mkSubj "sitten kun" ;
+one_time_A = mkA "ainutkertainen" ;
+onto_Prep = mkPrep "päälle" ;
+optical_A = mkA "optinen" ;
+outside_Prep = mkPrep "ulkopuolella" ;
+painting_N = mkN "maalaus" ;
+pall_N = mkN "paariliina" ;
+panama_PN = mkPN "Panama" ;
+party_N = mkN "puolue" ;
+pend_V = mkV "riippua" ;
+pennsylvania_PN = mkPN "Pennsylvania" ;
+pent_up_A = mkA "patoutunut" ;
+people_N = mkN "kansa" ;
+per_Prep = mkPrep "per" nominative ;
+perhaps_Adv = mkAdv "ehkä" ;
+philippines_PN = mkPN "Filippiinit" ;
+pittsburgh_PN = mkPN "Pittsburgh" ;
+plastics_N = mkN "muovi" ;
+plus_Conj = mkConj "plus" ;
+poland_PN = mkPN "Puola" ;
+present_day_A = mkA "tämänhetkinen" ;
+pretoria_PN = mkPN "Pretoria" ;
+pretty_AdA = mkAdA "melkoisen" ;
+publishing_A = mkA "julkaiseva" ;
+quite_AdA = mkAdA "melko" ;
+quite_Predet = mkPredet "aika" ;
+rank_N = mkN "arvoaste" ;
+record_N = mkN "ennätys" ;
+regime_1_N = mkN "hallinto" ;
+representativeMasc_N = mkN "edustaja" ;
+researcherMasc_N = mkN "tutkija" ;
+resident_N = mkN "asukas" ;
+role_1_N = mkN "rooli" ;
+rome_PN = mkPN "Rooma" ;
+run_N = mkN "juoksu" ;
+san_antonio_PN = mkPN "San Antonio" ;
+see_VS = mkVS (mkV "pitää"  "huolta") ;
+shield_N = mkN "kilpi" ;
+sidney_PN = mkPN "Sidney" ;
+signale_VS = mkVS "viestittää" ;
+since_then_Adv = mkAdv "siitä lähtien" ;
+sincere_A = mkA "vilpitön" ;
+singapore_PN = mkPN "Singapore" ;
+site_N = mkN "sijaintipaikka" ;
+so_PConj = mkPConj "niinpä" ;
+so_Subj = mkSubj "niin että" ;
+so_called_A = mkA "niinsanottu" ;
+soft_A = mkA "pehmeä" ;
+somehow_AdV = mkAdV "jotenkin" ;
+soon_AdV = mkAdV "pian" ;
+soon_Adv = mkAdv "pian" ;
+spain_PN = mkPN "Espanja" ;
+stabilize_V = mkV "vakiinnuttaa" ;
+start_V = mkV "aloittaa" ;
+start_ing_VV = mkVV "alkaa" ;
+start_to_VV = mkVV "alkaa" ;
+stockholm_PN = mkPN "Tukholma" ;
+strike_N = mkN "lakko" ;
+such_Predet = mkPredet "sellainen" ;
+such_as_Prep = mkPrep "kuten" nominative ;
+sweden_PN = mkPN "Ruotsi" ;
+syndicate_N = mkN "syndikaatti" ;
+taipei_PN = mkPN "Taipei" ;
+taiwan_PN = mkPN "Taiwan" ;
+texas_PN = mkPN "Teksas" ;
+throughout_Prep = mkPrep "läpi koko" genitive ;
+tokyo_PN = mkPN "Tokio" ;
+toronto_PN = mkPN "Toronto" ;
+tough_A = mkA "tiukka" ;
+toward_Prep = mkPrep partitive "kohti" ;
+towards_Prep = mkPrep partitive "kohti" ;
+turkey_PN = mkPN "Turkki" ;
+typical_1_A = mkA "tyypillinen" ;
+typical_3_A = mkA "tyypillinen" ;
+unheard_of_A = mkA "ennenkuulumaton" ;
+unique_A = mkA "ainutlaatuinen" ;
+unit_3_N = mkN "yksikkö" ;
+universe_N = mkN "niversumi" ;
+unless_Subj = mkSubj "ellei" ;
+unlike_Prep = mkPrep "erilainen kuin" nominative ;
+upon_Prep = mkPrep "päällä" ;
+vietnam_PN = mkPN "Vietnam" ;
+virginia_PN = mkPN "Virginia" ;
+washington_PN = mkPN "Washington" ;
+well_known_A = mkA "tunnettu" ;
+while_Subj = mkSubj "samaan aikaan kuin" ;
+white_collar_A = mkA "valkokauluksinen" ;
+whole_A = mkA "kokonainen" ;
+wyoming_PN = mkPN "Wyoming" ;
+yet_AdV = mkAdV "yhä" ;
+zurich_PN = mkPN "Zürich" ;
+
 }


--- a/lib/src/finnish/stemmed/ElimPredef.hs
+++ b/lib/src/finnish/stemmed/ElimPredef.hs
@@ -2,16 +2,13 @@ import qualified Data.Set as S

 -- comment out words that are predefined in another lexicon
 -- runghc ElimPredef.hs <DictEngFin.gf
-- removeFile = "predef.txt"
-- removeMsg = "PREDEF"
+removeFile = "KoeFin.gf"
+removeMsg = "MANUAL"

 -- also used for temporarily eliminating whatever from compilation
 --removeFile = "commentOut"
 --removeMsg = "POSTPONE"

-removeFile = "t-nouns"
-removeMsg = "PLURNOUN"
-
 main = do
  predefs <- readFile removeFile >>= return . S.fromList . map (head . words) . lines
  interact (unlines . map (elimPredef predefs) . lines)
--- a/lib/src/finnish/stemmed/MkGFLex.hs
+++ b/lib/src/finnish/stemmed/MkGFLex.hs
@@ -0,0 +1,29 @@
+-- convert annotated word list to GF lexicon
+
+import Data.Char
+
+main = 
+  interact (unlines . map (unwords . mkEntry . words) . lines)
+
+-- [bare_A] paljas
+
+mkEntry (fun_:trans) = [fun, "=", oper, args, ";"] where
+  fun = tail (init fun_)  -- unbracket
+  (name,cat) = let (tac,eman) = span (/= '_') (reverse fun) in (reverse (tail eman),reverse tac)
+  oper = "mk" ++ cat 
+  args = case cat of
+    'V':_  -> unwords (map quoteIf trans)
+    "Prep" -> unwords (map quoteIf trans)
+    _ | null trans -> quote (mkUpper name)
+    _ -> quote (unwords trans)
+
+quote s = "\"" ++ s ++ "\""
+
+-- [absent_Prep] poissa +elative
+quoteIf s = case s of
+  '+':cs -> cs
+  _ -> quote s
+
+mkUpper w = case w of
+  c:cs -> toUpper c : cs
+  _ -> w
--- a/lib/src/finnish/stemmed/ParadigmsFin.gf
+++ b/lib/src/finnish/stemmed/ParadigmsFin.gf
@@ -69,6 +69,20 @@ oper
  postGenPrep :         Str -> Prep ;  -- genitive postposition, e.g. "takana"
  casePrep    : Case ->        Prep ;  -- just case, e.g. adessive

+  mkPrep = overload {
+    mkPrep : Case -> Prep 
+     = casePrep ;
+    mkPrep : Str -> Prep
+     = postGenPrep ;
+    mkPrep : Case -> Str -> Prep
+     = postPrep ;
+    mkPrep : Str -> Case -> Prep
+     = \s,c -> prePrep c s ;
+    } ;
+
+  accusative : Prep
+    = {c = NPAcc ; s = [] ; isPre = True ; lock_Prep = <>} ;
+
  NK : Type ;   -- Noun from DictFin (Kotus)
  AK : Type ;   -- Adjective from DictFin (Kotus)
  VK : Type ;   -- Verb from DictFin (Kotus)
@@ -165,6 +179,7 @@ oper
    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)
+---    mkV : Str -> V -> V ; -- laimin+lyödä (prefixed verb)
  } ;

 -- All the patterns above have $nominative$ as subject case.
@@ -176,6 +191,8 @@ oper

  vOlla : V ; -- the verb "be"

+  olla_V : V
+   = vOlla ;

 --3 Two-place verbs
 --
@@ -208,10 +225,23 @@ oper
 -- Verbs and adjectives can take complements such as sentences,
 -- questions, verb phrases, and adjectives.

+mkVV = overload {
+  mkVV : Str -> VV   -- e.g. "yrittää"
+   = \s -> mkVVf (mkV s) infFirst ; 
+  mkVV : V -> VV     -- e.g. "alkaa"
+   = \v -> mkVVf v infFirst ; 
+  } ;
+
+mkVS = overload {
+  mkVS : Str -> VS   -- e.g. "väittää"
+   = \s -> lin VS (mk1V s) ;
+  mkVS : V -> VS     -- e.g. "sanoa"
+   = \v -> lin VS v ;
+  } ;
+
  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  
@@ -233,6 +263,32 @@ oper
  V0 : Type ; --%
  AS, A2S, AV, A2V : Type ; --%

+--2 Structural categories
+
+  mkAdV : Str -> AdV
+   = \s -> lin AdV (ss s) ;
+  mkAdA : Str -> AdA
+   = \s -> lin AdA (ss s) ;
+  mkAdN : Str -> AdN
+   = \s -> lin AdN (ss s) ;
+  mkPConj : Str -> PConj
+   = \s -> lin PConj (ss s) ;
+  mkSubj : Str -> Subj
+   = \s -> lin Subj (ss s) ;
+
+  mkPredet : Str -> Predet  -- invariable Predet, such as "vain"
+   = \s -> lin Predet {s = \\_,_ => s} ;
+
+  mkConj = overload {
+   mkConj : Str -> Conj
+   = \y -> {s1 = [] ; s2 = y ; n = Pl ; lock_Conj = <>} ;
+   mkConj : Str -> Str -> Conj 
+   = \x,y -> {s1 = x ; s2 = y ; n = Pl ; lock_Conj = <>} ;
+   mkConj : Str -> Str -> Number -> Conj 
+   = \x,y,n -> {s1 = x ; s2 = y ; n = n ; lock_Conj = <>} ;
+   } ;
+
+
 --.
 -- The definitions should not bother the user of the API. So they are
 -- hidden from the document.
@@ -538,7 +594,7 @@ oper
      huutaa,huudan,huutaa,huutavat,huutakaa,huudetaan,
      huusin,huusi,huusisi,huutanut,huudettu,huutanee : Str) -> V = mk12V ;
    mkV : (sana : VK) -> V = \w -> vforms2sverb w.s ** {sc = NPCase Nom ; lock_V = <> ; p = []} ;
-    mkV : V -> Str -> V = \w,p -> vforms2sverb w.s ** {sc = NPCase Nom ; lock_V = <> ; p = p} ;
+    mkV : V -> Str -> V = \w,p -> {s = w.s ; sc = w.sc ; lock_V = <> ; h = w.h ; p = p} ;
  } ;

  mk1V : Str -> V = \s -> 
@@ -645,8 +701,7 @@ oper
  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 = <>} ;

--- a/lib/src/finnish/stemmed/StructuralFin.gf
+++ b/lib/src/finnish/stemmed/StructuralFin.gf
@@ -1,5 +1,5 @@
 concrete StructuralFin of Structural = CatFin ** 
-  open MorphoFin, ParadigmsFin, (X = ConstructX), MakeStructuralFin, StemFin, Prelude in {
+  open MorphoFin, ParadigmsFin, (X = ConstructX), StemFin, Prelude in {

  flags optimize=all ;