To Saara: (I have marked <==== where I suggest other marking. <===== = means that the to original marking there should be added a =, like =H:n should be ==H:n.) :::: Saara's comments TT: Tronds comments LA121107 - Lene newest comments New tags ======== I have changed some of the tags in sme-dis.rle: @ADVL> is now: @>ADVL @N> is now: @>N @A> is now: @>A @P> is now: @>P @Num> is now: @>Num @Pron< is now: @Pron< @APP-Pron> is now: @APP>Pron ::: These are now in the code Input doesn´t match =================== I get this when I use the cg-analyze of bigkorpusVISL.txt as input to cg2visl script: Input did not match: "gilli" N* Der1 Der/Dimin N Sg Loc @ADVL> Input did not match: "gullevaš" A* Der3 Der/vuohta N Sg Acc @OBJ Input did not match: "vuoiŋŋastit" V* IV* Der2 Der/ahtti V TV Ind Prs Sg3 @+FMAINV ::: I am aware of this error, we were hesitating what to do with them (the tags), in the meantime I could of course add them to the structure. TT: I think it is better to do something with them, so that we get a clean output. What about deleting them Errors ====== I see that CS disappears! ::: Fixed. When there is a propernoun involved, the grouping doesn´t function. ::: Fixed "1960-logu rádjai" disappears in: Muhtun sámi biktasiid, nugo gápmagiid ja vuoddagiid, sii geavahedje guhkit, gitta 1960-logu rádjai. ::: Fixed. About some tags: ================ - All infinite verbs should get :v(nfin) ::: Fixed. ::: Another question: How the sentences like mun lean nohkan should be marked the group +FMAINV and -FMAINV? - CC @CVP should get CO:conj (now it gets SUB:conj-s) ::: Fixed @-FOBJ > Od ============ SME433 Dál de viimmat asttan čállit reivve. A1 A:adv('dál') Dál A:adv('de') de A:adv('viimmat') viimmat P:v('astat',TV,ind,pr,1sg) asttan <==== Od:cl Od:v('čállit',TV,inf) čállit <==== =P:v(nfin) @-FOBJ:n('reive',sg,acc) reivve <==== =Od:n . SME690 Mus lea viežžat čázi. A1 A:pron('mun',,1sg,loc) Mus <==== P:g P:v('leat',IV,ind,pr,3sg) lea <==== =D: H:v('viežžat',TV,inf) viežžat <==== = @-FOBJ:n('čáhci',sg,acc) čázi <==== Od:n . ::: Fixed. @PRON< is D to the pronoun, like in: ==================================== Moai Birehiin barge mánáidgárddis. S:g Moai =H Birehiin =D (@PRON<) barge P:v mánáidgárddis. A:n ::: Fixed. Son ii máhttán ieš vuodjit biillain. S:g- Son =H:pron- P:g- ii =D: máhttán =D: -S:g ieš =D:pron (@PRON<) -P:g vuodjit =H:v(nfin) billain. A:n ::: I implemented this, but it needs to be tested. So these tags in ::: vislcg they don't have to be symmetric? ::: As in here, the tags cross: S:g- P:g- -S:g -P:g D+H+D ======= SME566 Vuosttaš logi minuvtta lei buorre áigodat Nordlysa ektui. A1 S:g =D:adj('vuosttaš',ord,attr) Vuosttaš <==== =H:g =H:num('logi',sg,nom) logi <==== ==H: D:n('minukta',sg,gen) minuvtta <==== ==D:n P:v('leat',IV,ind,impf,3sg) lei Cs:g =D:adj('buorre',sg,nom) buorre =H:n('áigodat',sg,nom) áigodat A:g =D:prop('Nordlys',Obj,sg,gen) Nordlysa =H:prp-post('ektui') ektui ::: Problem: ::: The @>N in Vuosttaš points to N in minuvtta which again points to ::: logi. The above visl-format cannot be derived from those tags, since ::: the head is Num and @>N cannot point to Num. LA121107: Well, I see that the tags are not so logical. I will fix in sme-dis.rle, so vuosttaš gets @>Num. And it should be @SUBJ, not @ADVL. CORRECT - ((It works perfectly after changing sme-dis.rle, except from what is Dum??): SOURCE: text SMEB1 Vuosttaš logi minuvtta lei buorre áigodat Nordlysa ektui. A1 STA:cl S:g =Dum:adj('vuosttaš',ord,attr) Vuosttaš =H:g ==H:num('logi',sg,nom) logi ==D:n('minukta',sg,gen) minuvtta P:v('leat',IV,ind,impf,3sg) lei Cs:g =D:adj('buorre',sg,nom) buorre =H:n('áigodat',sg,nom) áigodat A:g =D:prop('Nordlys',Obj,sg,gen) Nordlysa =H:prp-post('ektui') ektui . @>A should be D =============== SME658 Muhtun guovlluid gávttiin eai leat nu ollu čiŋat oskku geažil. A1 A:g =D:pron('muhtun',,attr) Muhtun =D:n('guovlu',pl,gen) guovlluid =H:n('gákti',pl,loc) gávttiin P:g =D:vaux:v('ii',IV,neg,ind,3pl) eai =H:v('leat',IV,ind,pr,conneg) leat <==== S:g D:adv('nu') nu <==== =D: (@>A) S:g <==== remove =D:pron('ollu',) ollu =H:n('čikŋa',pl,nom) čiŋat A:g =D:n('osku',sg,gen) oskku =H:prp-post('geažil') geažil . ::: Problem is that @>A modifies adjectives. There is no adjective in the rest of the text. it seems that here it should be interpreted as modifying adverbial. So @>A modifies first adjectives, and if not found, adverbials? LA121107: "nu" modifies "ollu". I will fix the tags in sme-dis.rle CORRECT (It works perfectly after changing sme-dis.rle): SMEB1 Muhtun guovlluid gávttiin eai leat nu ollu čiŋat oskku geažil. A1 STA:cl A:g =D:pron('muhtun',,attr) Muhtun =D:n('guovlu',pl,gen) guovlluid =H:n('gákti',pl,loc) gávttiin P:g =D:vaux:v('ii',IV,neg,ind,3pl) eai =H:v(nfin)('leat',IV,ind,pr,conneg) leat S:g =D:g ==D:adv('nu') nu ==H:adj('ollu',attr) ollu =H:n('čikŋa',pl,nom) čiŋat A:g =D:n('osku',sg,gen) oskku =H:prp-post('geažil') geažil . Num @SUBJ + @Q< should be grouped, like in: =========================================== SME599 Mis leat moadde beatnaga. A1 A:pron('mun',,1pl,loc) Mis P:v('leat',IV,ind,pr,3pl) leat <==== S:g S:num('moadde',sg,nom) moadde <==== H: D:n('beana',sg,gen) beatnaga <==== D: . SME427 Áiggun geahččat dušše moadde filmma. A1 P:g =D:vaux:v('áigut',TV,ind,pr,1sg) Áiggun =H:v('geahččat',TV,inf) geahččat A:adv('dušše') dušše <==== Od:g Od:num('moadde',sg,acc) moadde <==== =H:num D:n('filbma',sg,gen) filmma <==== =D:n . ::: Fixed Paratagmes ========== And the paratagmes are not quite good yet, when there is a comma. eg: SOURCE: text SME377 Mii oahppat lohkat ja čállit sámegillii. A1 S:pron('mun',,1pl,nom) Mii P:v('oahppat',TV,ind,pr,1pl) oahppat <==== Od:par Od:v('lohkat',TV,inf) lohkat <==== =CJT:v(nfin) SUB:conj-s('ja') ja <==== =CC:conj Od:v('čállit',TV,inf) čállit <==== =CJT:v(nfin) A:n('sáme#giella',sg,ill) sámegillii . SOURCE: text SME378 Doppe mii sáhttit oastit gáfe, deaja, sávtta dahje bruvssa. A1 A:adv('doppe') Doppe S:pron('mun',,1pl,nom) mii P:g =D:vaux:v('sáhttit',IV,ind,pr,1pl) sáhttit =H:v('oastit',TV,inf) oastit <==== Od:par Od:n('gáffe',sg,acc) gáfe <==== =CJT:n , Od:n('deadja',sg,acc) deaja <==== =CJT:n , Od:par <==== remove =CJT:n('sákta',sg,acc) sávtta =SUB:conj-s('dahje') dahje <==== =CO:conj =CJT:n('bruvsa',sg,acc) bruvssa . ::: These are fixed. ------- LA121107: I have changed this one, also in sme-dis.rle. CORRECT (how it should be): SME390 Garraskerrui don vurket visot iežat čállosiid ja govaid ja prográmmaid. STA:cl A1 A:n('garra#skearru',sg,ill) Garraskerrui S:pron('don',,2sg,nom) don P:v('vurket',TV,ind,pr,2sg) vurket Od:g =D:adj('visot',attr) visot =D:pron('ieš',,gen,poss2sg) iežat =H:par ==CJT:n('čálus',pl,acc) čállosiid ==CC:conj-s('ja') ja ==CJT:n('govva',pl,acc) govaid ==CC:conj-s('ja') ja ==CJT:n('prográmma',pl,acc) prográmmaid . ::: This does not work still. It is the general problem with coordination: which constituents form a pair.