Finite state and Constraint Grammar based analysers, proofing tools and other resources
S Y N T A C T I C F U N C T I O N S F O R S Á M I
Sámi language technology project 2003-2018, University of Tromsø #
This file adds syntactic functions. It is common for all the Saami
LEFT RIGHT because of apertium
Sets for POS sub-categories
Sets for Semantic tags
Sets for Morphosyntactic properties
Sets for verbs
V is all readings with a V tag in them, REAL-V should be the ones without an N tag following the V. The REAL-V set thus awaits a fix to the preprocess V … N bug.
The set COPULAS is for predicative constructions
NP sets defined according to their morphosyntactic features
The PRE-NP-HEAD family of sets
These sets model noun phrases (NPs). The idea is to first define whatever can occur in front of the head of the NP, and thereafter negate that with the expression WORD - premodifiers.
The set NOT-NPMOD is used to find barriers between NPs. Typical usage: … (*1 N BARRIER NPT-NPMOD) … meaning: Scan to the first noun, ignoring anything that can be part of the noun phrase of that noun (i.e., “scan to the next NP head”)
Miscellaneous sets
Border sets and their complements
ADLVCASE
These were the set types.
hab1 hab aux leat
hab_numo1 hab copula comma comma N+Nom
hab_numo2 copula nu mo/go hab
leahab copula nu mo/go hab
hab2 hab auxv adv leat
hab3 (
hab3 (
hab3 (
hab3 (
hab_main (
habInf hab lea inf
habNomLeft Nom or Num + gen hab lea
habAdvl Ii han ovttasge du sogas leat dat namma.
hab4 hab cc hab leat
hab6 lea go hab – leago hab
hab7 lea go hab
hab5 This is not HAB Mánás gollot gieđat.
hab9 prop ord-hab leat
hab10 prop ord-hab leat
habDain2
habRel # before relative clause
habEllipse Buot gánddain lea dreassa, nieiddain fas gákti.
habGen (
habGenQst (
n<titel1 (@N<) for (“jr”) or (“sr”); if first one to the left is Prop
n<titel2 (@N<) for INITIAL; if first one to the left is a noun, or if to the left of you is a single letter which is part of a noun conjunction bustávas e ja f gáibiduvvo
n<:com (@N<) for (Sg Com); if first one to the left is Coll
>nAttr (@>N) for Attr; if there is a noun to your right
n>Indef (Pron Indef Attr); if eará is to the right
n>Indef (Pron Indef Com); if eará is to the right
>nNum (@>N) for numerals if; there is a noun to your right. You are not allowed to be (Sg Nom), (Sg Acc) or (Sem/Date)
noun>n (@>N) for Gen; if there is a noun to your right. Restrictions: Not if you are: a time related word. Not if you are OKTA with Pl Loc to your right. Not if CC is to your right followed by another Gen and then Po. Not if you are HUMAN and to your right is Actio Nom folloed by a noun.
>nTime (@>N) for Gen TIME-N; if timenoun to your right. Restrictions: Not if you are a OKTA Nom with Pl Loc to your right. Not if CC followed by Gen, followed by Po to your right. Not if COMMA to your right
>ntittel (@>N) for (Sg Nom TIME-N) or (Nom Der/NomAg); if to your right is Sem/Mal, Sem/Fem, Sem/Sur
>nplc (@>N) for (Sg Nom Prop Sem/Plc), if to your right is Sem/Plc
>nALU (@>N) for Sg Acc numerals; when a measure-noun to the right
>NTime (@>N) for Gen; if you are TIME-N with BOC to your left, and PREGEN to your right
n<:Refl (@N<) for (Refl Nom); if to the left is (N Nom), or if first one to the left is a finite mainverb with a (N Nom) to the left
>pron1 (@>Pron) for GRADE-ADV, DUSSE, BUOT if; first one to the right is Pron
>pron2 (@>Pron) for (Refl Nom) if; first one to the right is Refl
>pron3 (@>Pron) for (Pron Recipr) if; first one to the right is (Pron Recipr)
vaikko (@>Pron) for vaikko if; first one to the right is Indef
vaikkoman (@>ADVL) for vaikko if; first one to the right is man
dasmaŋŋel (@>ADVL) for vaikko if; first one to the right is man
adv>advl (@>ADVL)
adv>advl (@>ADVL)
BOSvoc (@VOC) for HUMAN Nom; if sentence initial. To the right is comma. No nom-cased HUMAN followed by comma or CC is allowed to the right. There should not be a relative clause to the right, because then you are likely to be SUBJ
voc (@VOC) for Nom HUMAN; if comma to the left and an second person verb or pronoun to the left. To the right is the end of the sentence
Particle<subj (@PCLE)
spred<obj (@SPRED<OBJ) for Acc; the object of an SPRPED. Not to be mistaken with OPRED. If SPRED is to the left, and copulas is to the left of it. Nom or Hab are found sentence initially.
Hab<subj (
Hab<subj (
Hab>Advlcase<subj (
Nom>Advlcase<subj (
<extSubj (
<extSubj (
<extSubjA (
<extSubj (
<extSubj (
loc<extSubj (
<spred (@<SPRED) for Nom; if Nom to the left, copulas to the left of Nom, and a time related word to the left of it.
<extQst1 (
<extQst2 (
extQst3> (
extQst3> (
<extsubjcoor (
Sem/Year
<spredQst (@<SPRED) for Nom; in a typically question sentence; You are not allowed to be Pers or human. The special part is that Nom is not allowed to your right
<spredQst2 (@<SPRED) for (A Nom); in a typically question sentence; You are SPRED if (N Nom) is to your left and leat + qst is to the left
<spredQst3 (@<SPRED) for (A Nom); you are SPRED when you are (A Nom) and to your right is (N Nom). This is a Qst-sentence, so copulas is found to your left
<spredQst4 (@<SPRED) for Nom; but only in a qst-sentence where there is no chance of you beeing the subj
<NomBeforeSpred (@<SPRED) for (A Nom) if; Nom to the left, and copulas is to the left of Nom. There is no Nom allowed to the right of copulas! To avoid messing with coordination: ja, dahje and comma are not allowed to your left. Comma is not allowed to your right; if so then you are likely to be coordinated
<spred (@<SPRED) for A Nom or N Nom if; the subject Nom is on the same side of copulas as you: on the right side of copulas
<spredVeara (@<SPRED) for veara + Nom; if genitive immediately to the right, and intransitive mainverb to the right of genitive
leftCop<spred (@<SPRED) for Nom; if copulas is the main verb to the left, and there is no Ess found to the left of cop (note that Loc is allowed between target and cop). OR: if you are Coll or Sem/Group with copulas to your left.
<spredLocEXPERIMENT (@<SPRED) for material Loc; if you are to the right of copulas, and the Nom to the left of copulas is not a hab-actor
NumTime (@<SPRED) for A Nom
<spredSg (@<SPRED) for Sg Nom
<spredPg (@<SPRED) for Pl Nom
<spred (@<SPRED) for Nom; if copulas to the left, and Nom or sentence boundary to the left of copulas. First one to the right is EOS.
COP<spredEss (@<SPRED) for N Ess
spredEss> (@SPRED>) for N Ess; if copulas to the right of you, and if an NP with nom-case first one to your left.
GalleSpred> (@SPRED>) for Num Nom; if sentence initial
spredSgMII> (@SPRED>)
spredšaddat> (@SPRED>)
r492> (@SPRED>) for Interr Gen; consisting only of negations. You are not allowed to be MII. You are not allowed to have an adjective or noun to yor right. You are not allowed to have a verb to your right; the exception beeing an aux.
AdjSpredSg> (@SPRED>) for A Sg Nom; if copulas to the right, but not if A or @<SPRED are found to the right of copulas
Spred>SubjInf (@SPRED>) for Nom; if copulas to the right, and the subject of copulas is an Inf to the right
spredCoord (@<SPRED) coordination for Nom; only if there already is a SPRED to the left of CNP. Not if there is some kind of comparison involved.
subj>Sgnr1 (@SUBJ>) for Nom Sg, including Indef Nom if; VFIN + Sg3 or Pl3 to the right (VFIN not allowed to the left)
subj>Pl (@SUBJ>) for plural nominatives, including Coll and Sem/Group. VFIN + Pl3 to the right.
subj>Pl (@SUBJ>) for plural nominatives
subj>Sg (@SUBJ>) for Nom Sg; if VFIN + Sg3 to the right.
Sg<subj (@<SUBJ) for Nom Sg; if VFIN Sg3 or Du2 to the left (no HAB allowed to the left).
Du<subj (@<SUBJ) for Nom Coll if; a dual third person verb is found to the left
PlDu<subj (@<SUBJ) for (N Nom Pl), (Sem/Group Nom), (Coll Nom), (Pron Nom Pl) if; a verb is Pl3 or Du3 to your left. The verb is not allowed to be copulas with a place, Loc or time noun to its left
copPl3<subj (@<SUBJ) for Nom Pl; you don’t to be a noun, only Nom Pl. To the left is copulas and first one to the right is @<SPRED
-fsubj> (@-FSUBJ>) for HUMAN Gen; in a NP-clause. To your right is Actio Nom followed by a noun
f<advl (@-F<ADVL) for infinite adverbials
f<advl (@-F<ADVL) for infinite adverbials
s-boundary=advl> (@ADVL>) for ADVL that resemble s-boundaries. Mainverb to the right.
diibmuadvl> (@ADVL>) for (diibmu Nom) if first one to the right is Num
-fsubj (@-FSUBJ>) for HUMAN Acc after DADJAT verbs
-fobj> (@-FOBJ>) for Acc if front of abessive, gerundium, actio locative, perfectum participle or infinitive. First one to the right not allowed to be Acc though
-fobj> (@-FOBJ>) for Acc if human with ADVL-case to the left and transitive infinitive OBJ to the right. First one to the right not allowed to be Acc though
advl>mainV (@ADVL>) if; finite mainverb not found to the left, but the finite mainverb is found to the right.
V<advl (@<ADVL) if; finite mainverb found to the left. Not if a comma is found immediately to the left and a finite mainverb is located somewhere to the right of this comma.
advl>v (@ADVL>) if; you are ADVL, time-noun or Sem/Route and there is a finite verb to the right in the clause, or if to your right is: de followed by a finite verb. OR: if you are a time-nound and to your right is: go or sentenceboundary followed by a finite verb
advlPoPr> (@<ADVL) for Po or Pr; if mainverb to the right.
BOSPo> (@ADVL>) for Po; if trapped between BOS to the right and S-BOUNDARY OR COMMA to the left, because the main verb will then automatically be on your right side.
<advlComIll (@<ADVL) only if; you are Com OR Ill. To your left is a mainverb, and to your right a sentenceboundary, because we don’t want there to be another mainverb you potentially could belong to
<advlEOS (@<ADVL) for Po or Pr or Loc; if you are found at the very end of a sentence. A mainverb is needed to the left though.
<advlGen (@<ADVL) for (N Gen) if mainverb to the left and no noun to the right
<opredgohcodit (@<OPRED) for Ess
advlEss> (@<ADVL) for weather and time Ess, if FMAINV to the left.
comma<advlEOS (@<ADVL) for Adv if; mainverb is to the left. Comma to the left and mainverb to the right in the same clause is not allowed
advl>inbetween (@ADVL>) for Adv; if inbetween two sentenceboundaries where no mainverb is present.
comma<advlEOS (@<ADVL) for Adv if; comma found to the left and the finite mainverb to the left of comma. To the right is the end of the sentence.
BOSadvl> (@ADVL>) if; you are N Loc or N Ill and found sentence initially and there is a main verb somewhere to the right. No barrier for the mainverb; based on the thought that first one to your right is probably a sentenceboundary.
cleanupILL<advl (@<ADVL) for N Ill if; there are no boundarysymbols to your left, if you arent already @N< OR @APP-N<, and no mainverb is to yor left.
cleanupPo (@ADVL) for Po: This rule tags all Po:s as ADVL if they haven’t gotten a tag somewhere along the way.
cleanupPr (@ADVL) for Po: This rule tags all Pr:s as ADVL if they haven’t gotten a tag somewhere along the way.
-fsubj>asAcc (@-FSUBJ>) for HUMAN Acc; if there is a verb @-F<OBJ to your left
-f<obj (@-F<OBJ) for Acc if there is a transitive verb + SYN-V to your left
-fsubj>IV (@-FSUBJ>) for Acc; if there is an IV-verb acting as a @-F<OBJ to your right
-fsubj>IV (@-FSUBJ>) for Acc; if there is an TV-verb acting as a @-F<OBJ to your right followed by an Acc
-fsubj>asGen (@-FSUBJ>) for Gen;
f<subj (@-F<SUBJ) for Nom if; (V @-F<OBJ) to the left.
<opredAAcc (@<OPRED) for A Acc; if an other accusative to the left, and a transtive verb to the left of it. OR: if a transitive verb to the left, and an accusative to the left of it.
<advlMeasr (@<ADVL) for (Num Acc); if finite IV-mainverb to the left, measure-noun to the right
<objMeasr (@<OBJ) for Num Acc; if finite TV-mainverb to the left, measure-noun to the right
<advlMeasr2 (@<ADVL) for MEASR-N + Acc; if (Num Pl) to the left and mainverb to the left of it
advlMeasr> (@ADVL>) for Num Acc;
Obj> (@OBJ>) for Acc; if there is a finite mainverb to the right in the clause. A really simple rule with no other restrictions..
s-boun<obj (@<OBJ) for Acc; if sentenceboundary to your left and a transitive mainverb to the left futher to the left
<objIV (@<OBJ) for Acc; if there is an intransitive mainverb in the clause. Not for Rel or Num. Not if you are a numeral followed by a measure-noun
<advlEss (@<ADVL) for ESS-ADVL if; FMAINV to the left
IV<spredEss (@<SPRED) for N Ess if; FMAINV to the left is intransitive or bargat
<opredEss (@<OPRED) for (N Ess), (A Ess) if; transitive mainverb to the left in the clause. If accusative to the left or to the right, or if Inf or ahte to the right, or if there is a noun to the right followed by an Inf
Acc<opredEss (@<OPRED) for (N Ess), (A Ess) if; transitive mainverb to the left in the clause, and an accusative cased Rel left to the verb
onlyV<opred (@<OPRED) for (N Ess) if; there is a transitive mainverb to the left. Usually there needs to be an Acc to the left, but here it is not needed
onlyV<opred2 (@<OPRED) for (N Ess) if;
subj>ifV (@SUBJ>) for NP-HEAD-NOM, DUPRON or (Num Nom) if; a finite mainverb is found to the right. This is a cleanup rule for subjects
hnoun>ifV (@SUBJ>) for NP-HEAD-NOM, DUPRON if. The counterpart of subj>ifV. You are HNOUN if there is a finite verb to your right, but NOT if there is a finite verb after a relative clause
The analysis give double analysis because of optional semtags. We go for the one with semtag.
This (part of) documentation was generated from src/cg3/functions.cg3