!Distributed under the terms of the GNU General Public License version 2 
! or any later version.

! ========================= !
! Hyphenator for North Sámi !
! ========================= !

! To eventual curious readers
! This file is empty. It is modeled upon the Greenlandic
! parallel. Finally, this is not how we should do it.
! What we need is access to the stem in sme-lex.txt.
! This is set up as a string-to-string conversion.
! The first part of this file is the sme hyphenator, 
! and the second is the smesmj converter.
! This should be fewritten, but here I sit without see.


echo <<defining Cns and Vow...>>

define Vow       [ a | á | e | i | o | u | y | æ | ø | å | ä | ö | 
                   A | Á | E | I | O | U | Y | Æ | Ø | Å | Ä | Ö | 
é | ó | ú | í | à | è | ò | ù | ì | ë | ü | ï | â | ê | ô | û | î | ã | ý 
É | Ó | Ú | Í | À | È | Ò | Ù | Ì | Ë | Ü | Ï | Â | Ê | Ô | Û | Î | Ã | Ý ] ;

define 2ndVow [ e | a | u | o | i | y | ø | E | A | U | O | I | Y | Ø ] ;

!define 3dVow [ u | U ] ;


define VowNot2ndVow   [ Vow - 2ndVow ] ;

!define VowNot3dVow   [ Vow - 3dVow ] ;


define U   [ u | U ] ;
define O   [ o | O ] ;
define A   [ a | A ] ;
define E   [ e | E ] ;
define I   [ i | I ] ;
define Y   [ y | Y ]  ;
define EO  [ e | E | o | O ] ;
define EOA [ e | E | o | O | a | A ] ;
define IAO [ i | I | a | A | o | O ] ;
define Oslash   [ ø | Ø ] ; ! Renamed from Ø to avoid non-ASCII chars
define UI  [ u | U | i | I ] ;


define IEdipht [ i e | I e | I E ] ;
define UOdipht [ u o | U o | U O ] ;
define EAdipht [ e a | E a | E A ] ;
define OAdipht [ o a | O a | O A ] ;
define OUdipht [ o u | O u | O U ] ;      ! jour-na-lis-sta
define AUdipht [ a u | A u | A U ] ;      ! fau-na, Lauv-sjø-vo-la, Kår-ve-jau-re ok.
define EUdipht [ e u | E u | E U ] ;      ! leut-nán-ta
define AEdipht [ a e | A e | A E ] ;      ! lae-sta-di-us, 
define OslashYdipht [ ø y | Ø y | Ø Y ] ; ! nøyt-rála  ! Renamed from ØYdipht to avoid non-ASCII chars
define OEdipht [ o e | O e | O E ] ; ! goe-the
define IOslashdipht [ i ø | I ø | I Ø ] ; ! skiøld     ! Renamed from IØdipht to avoid non-ASCII chars
define IOdipht [ i o | I o | I O ] ; ! skiold, nation

define dipht [ IEdipht | UOdipht | EAdipht | OAdipht | OUdipht | AUdipht |
               EUdipht | AEdipht | OslashYdipht | OEdipht | IOslashdipht | IOdipht] ;

define VowNotU      [ Vow - U   ] ;
define VowNotO      [ Vow - O   ] ;
define VowNotA      [ Vow - A   ] ;
define VowNotE      [ Vow - E   ] ;
define VowNotI      [ Vow - I   ] ;
define VowNotEO     [ Vow - EO  ] ;
define VowNotEOA    [ Vow - EOA ] ;
define VowNotIAO    [ Vow - IAO ] ;
define VowNotY      [ Vow - Y   ] ;
define VowNotOslash [ Vow - Oslash   ] ;  ! Renamed from VowNotØ to avoid non-ASCII chars
define VowNotUI     [ Vow - UI  ] ;


define Cns [ b | c | č | d | đ | f | g | h | j | k | l | m | n | ŋ | 
             B | C | Č | D | Đ | F | G | H | J | K | L | M | N | Ŋ | 
             p | q | r | s | š | t | ŧ | v | w | x | z | ž |   
             P | Q | R | S | Š | T | Ŧ | V | W | X | Z | Ž ] ; 

define S    [ S | s | Š | š ] ;
define Cnss [ Cns - S ] ;
define Sgm  [ Cns | Vow ] ;


echo << Rules>>


echo <<defining Cns and Vow...>>


define Syll Cns* Vow+ Cns* ;

echo << Rules >>

! Defining this rule in the beginning, it is so long...

! This rule must be changed in see, what we would like to do is to 
! introduce a schwaa vowel instead of the left-hand side.

define rd  m b -> m b b ,   m b b -> m b ,
           m p -> m p p ,   m p p -> m p ,
           m s -> m s s ,   m s s -> m s ,
           m š -> m s s j , m š š -> m s j ,

           n c -> n c c ,   n c c -> n c ,
           n d -> n d d ,   n d d -> n d ,
           ŋ g -> ŋ g g ,   ŋ g g -> ŋ g ,
           ŋ k -> ŋ k k ,   ŋ k k -> ŋ k ,
           n s -> n s s ,   n s s -> n s ,
           n t -> n t t ,   n t t -> n t ,
           n v -> n v v ,   n v v -> n v ,
           n s -> n s s ,   n s s -> n s ,
           n z -> n t t s , n z z -> n t s ,
           n ž -> n t t j , n ž ž -> n t j ,

           v d -> v d d ,   v d d -> v d ,
           v g -> v g g ,   v g g -> v g ,
           v j -> v j j ,   v j j -> v j ,
           v k -> v k k ,   v k k -> v k ,
           v l -> v l l ,   v l l -> v l ,
           v p -> v p p ,   v p p -> v p ,
           v r -> v r r ,   v r r -> v r ,
           v s -> v s s ,   v s s -> v s ,
           v t -> v t t ,   v t t -> v t ,
           v č -> v t t j , v č č -> v t j ,
           v š -> v s s j , v š š -> v s j ,
           v z -> v t t s , v z z -> v t s ,
           v ž -> v t t j , v ž ž -> v t j ,

           j b -> j b b ,   j b b -> j b ,
           j c -> j c c ,   j c c -> j c ,
           j d -> j d d ,   j d d -> j d ,
           j f -> j f f ,   j f f -> j f ,
           j g -> j g g ,   j g g -> j g ,
           j k -> j k k ,   j k k -> j k ,
           j p -> j p p ,   j p p -> j p ,
           j s -> j s s ,   j s s -> j s ,
           j t -> j t t ,   j t t -> j t ,
           j v -> j v v ,   j v v -> j v ,
           j z -> j t t s , j z z -> j t s ,
           j ž -> j t t j , j ž ž -> j t j ,
           
           l b -> l b b ,   l b b -> l b ,
           l c -> l c c ,   l c c -> l c ,
           l d -> l l d ,   l d d -> l d ,
           l f -> l f f ,   l f f -> l f ,
           l g -> l g g ,   l g g -> l g ,
           l j -> l j j ,   l j j -> l j ,
           l k -> l k k ,   l k k -> l k ,
           l p -> l p p ,   l p p -> l p ,
           l s -> l s s ,   l s s -> l s ,
           l t -> l l t ,   l t t -> l t ,
           l v -> l v v ,   l v v -> l v ,
           l č -> l t t j , l č č -> l t j ,
           l š -> l s s j , l š š -> l s j ,
           l z -> l l t s , l z z -> l t s ,
           l ž -> l l t j , l ž ž -> l t j ,

           r b -> r b b , r b b -> r b ,
           r c -> r c c , r c c -> r c ,
           r d -> r d d , r d d -> r d ,
           r f -> r f f , r f f -> r f ,
           r g -> r g g , r g g -> r g ,
           r j -> r j j , r j j -> r j ,
           r k -> r k k , r k k -> r k ,
           r p -> r p p , r p p -> r p ,
           r s -> r s s , r s s -> r s ,
           r t -> r t t , r t t -> r t ,
           r v -> r v v , r v v -> r v ,
           r č -> r t t j , r č č -> r t j ,
           r š -> r s s j , r š š -> r s j ,
           r z -> r t t s , r z z -> r t s ,
           r ž -> r t t j , r ž ž -> r t j , 

                            s s k -> s k ,
                            s s t -> s t ,
                            s s m -> s m ,
                            s s p -> s p ,

           š k -> s s j k , š š k -> s j k ,
           š t -> s s j t , š š t -> s j t ,
           š m -> s s j m , š š m -> s j m ,
           š v -> s s j v , š š v -> s j v ,

           đ b -> r b b , đ b b -> r b ,
           đ g -> r g g , đ g g -> r g ,
           đ j -> r j j , đ j j -> r j ,
           đ v -> r v v , đ v v -> r v ,

           j h l -> j h l l , j h l l -> j h l ,
           j h m -> j h m m , j h m m -> j h m ,
           j h n -> j h n n , j h n n -> j h n ,
           v h l -> v h l l , v h l l -> v h l ,

           đ b -> r b b , đ b b -> r b ,
           đ b -> r b b , đ b b -> r b ,
           đ b -> r b b , đ b b -> r b ,
           đ b -> r b b , đ b b -> r b ,
           đ b -> r b b , đ b b -> r b ,
           đ b -> r b b , đ b b -> r b ,
           đ b -> r b b , đ b b -> r b ,
           đ b -> r b b , đ b b -> r b ,
           
           d n   -> d d n ,   t n   -> d n ,
           b m   -> b b m ,   p m   -> b m ,
           g ŋ   -> g g ŋ ,   k ŋ   -> g ŋ ,
           d n j -> d d n j , t n j -> d n j   || [.#.|%#|%-] Cns* Vow+ _ Vow ;


! Here comes the other rules

define FinalE i -> e || Syll Cns _ .#. ;

define ij i -> j || Vow _ ;
define dtj ž ž -> d t j || Vow _ Vow ; 
define ttj č č -> t t j || Vow _ Vow ;
define ssj š š -> s s j || Vow _ Vow ;

define LN   đ m m -> r m ,     đ b m -> r m m ,
            đ ŋ ŋ -> r ŋ ,     đ g ŋ -> r ŋ ŋ ,
            j m m -> j m ,     j b m -> j m m ,
            j n n -> j n ,     j d n -> j n n ,
            j ŋ ŋ -> j ŋ ,     j g ŋ -> j ŋ ŋ ,
            l m m -> l m ,     l b m -> l m m ,
            l n n -> l n ,     l d n -> l n n ,
            l ŋ ŋ -> l ŋ ,     l g ŋ -> l ŋ ŋ ,
            r p m -> r m ,     r b m -> r m m ,
            r t n -> r n ,     r d n -> r n n ,
            r k ŋ -> r ŋ ,     r g ŋ -> r ŋ ŋ ,
            v n n -> v n ,     v d n -> v n n ,
          l n n j -> l n j , l d n j -> l n n j ,
          r t n j -> r n j , r d n j -> r n n j ,
          v n n j -> v n j , v d n j -> v n n j || Vow _ Vow ;

define jNN b m -> m m , d n -> n n || Vow j _ ;

define SK2SSK s p -> s s p   ,
              s t -> s s t   ,
              s k -> s s k   ,
              s m -> s s m   ,
              š t -> s s j t ,
              š k -> s s j k ,
              š m -> s s j m || [.#.|%#|%-]  Cns* Vow+ ([j|m|n|r|v]) _ Vow ;
define SKK2SK s t t -> s t ,               
              s k k -> s k ,               
              s m m -> s m ,               
              š t t -> š t ,               
              š k k -> š k ,               
              š m m -> š m ||  Vow  ([j|m|n|r|v]) _ Vow ; 

define sl š -> s || [.#.|%#|%-] _ l ;
define zz z z -> d t s ;
define z    z -> t s ;
define cc   c c -> t t s , č č -> t t j || Vow ([j|l|h|k|n|r|v]) _ Vow ;
define c    c -> t s , č -> t j || [.#.|%#|%-|Vow|j|l|h|k|n|r|v] _ Vow ;
define RevDipht u -> u o || _ Cns+ e ;
define oCns u -> o || Syll _ Cns .#. ;

define eCns i -> e ||          Syll     _ [ Cns - j ] .#. ,
                      [.#.|%#|%-] Syll Cns _ [Cns - j] ,,
            e -> i || [.#.|%#|%-] Syll Cns _  Cns ;
define it   e -> i || Syll _ t .#. ; ! Loan word rule!!
!define H2M u -> o , e -> i , o -> u || [.#.|%#|%-] Syll Cns _ Cns ;
define H2M  u -> o , o -> u          || [.#.|%#|%-] Syll Cns _ Cns ;
define hit2dit h -> d || _ i t .#. ; ! Odd-syll verbs only
define longain2ndsyll a -> á || [.#.|%#|%-] Cns* Vow [Cns|[n|s|t]j|t s] _ [ Cns - š ] ;
define tjfromšafter2syll š -> t j || [.#.|%#|%-] Syll Syll _ Vow ;
define SamiLetters č -> t j , đ -> d , š -> s j , ŧ -> t , ž -> t j  ;
define 1stysllæ e a -> æ || [.#.|%#|%-] (Cns+) _ Cns+ á ;
define 1stsylldiphthong o -> å , e a -> i e || [.#.|%#|%-] (Cns+) _ Cns ;
define labialharmony [a|á] -> å || å Cns+ _ ;
define oa2åG1 o a -> å || [.#.|%#|%-] (Cns+) _  [Cns|[n|s|t]j|t s]  a ;

echo << and the cleaning rules at the end >>

define phones e7 -> e , o7 -> o ;
define technical %# -> 0 ;

echo << Compile... >>

read regex [
    FinalE
 .o. ij    
 .o. dtj
 .o. ttj
 .o. ssj
 .o. LN
 .o. jNN
.o. rd ! the long rule in the beginning
.o. SK2SSK
.o. SKK2SK
.o. sl
.o. zz
.o. z    !not for LOAN conversion
.o. cc   !not for LOAN conversion
.o. c    !not for LOAN conversion
.o. oCns
.o. eCns !not for LOAN conversion
.o. it
.o. H2M
.o. hit2dit
.o. longain2ndsyll
.o. tjfromšafter2syll
.o. SamiLetters
.o. 1stysllæ
.o. 1stsylldiphthong
.o. labialharmony
.o. oa2åG1
.o. phones
.o. technical
] ;

invert net ;

save ~/words/dicts/smesmnj/bin/j.fst



echo << Cleanup...>>

define clean  [ Dummy -> 0 ] ;

! Lagre denne fila i xfst source < dennefila.txt og deretter save hy.fst
! cat fil.txt | preprocess | ./lookup -utf8 hy.fst | cut -f2 | tr '\n' ' ' | less

echo <<Phonology...>> 

define down [ A -> a, Á -> á, B -> b, C -> c, D -> d, E -> e, F -> f, G -> g, H -> h, 
    I -> i, J -> j, K -> k, L -> l, M -> m, N -> n, O -> o, P -> p, Q -> q, 
    R -> r, S -> s, T -> t, U -> u, V -> v, W -> w, X -> x, Y -> y, Z -> z, 
    Æ -> æ, Ø -> ø, Å -> å ] ;

















echo <<Combining...>>

read regex [ 

] ;


invert net ;