# Make initial downcasing after the misnamed U.Cap.Opt
# This gives oslolaš across the board
# like Pariisi -> pariisilainen

# ---- starts here: ------

# Various flag definitions to handle both regular downcase and downcase on hfst
# transducers with hyperminimisation. HypMin converts each lexicon name to a
# flag diacritic, and the result is that there can be more symbols before the
# initial letter of a word.
#
# NB! Note that the lexicon name of the proper nouns can be different from what
# is defined here. Change if required!

define DowncaseFlag  "@U.Cap.Opt@"            ;
define CmpHyph       "@U.CmpHyph.TRUE@"       ;
define LexRootFlag   "@P.LEXNAME.Root@"       ;
define LexProperFlag "@P.LEXNAME.ProperNoun@" ;

read regex [
A -> a,
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,
Å -> å,
Ä -> ä,
Ö -> ö,
Æ -> æ,
Ø -> ø,
Á -> á,
Č -> č,
Đ -> đ,
Ŋ -> ŋ,
Š -> š,
Ŧ -> ŧ,
Ž -> ž
 || .#. ( LEXROOTFLAG ) DCASEFLAG ( LEXPROPERFLAG ) ( HYPHFLAG )       ( ? ) _ ,
                        DCASEFLAG ( LEXPROPERFLAG ) ( HYPHFLAG ) ?* %- ( ? ) _

# ---- ends here: ------

# The lower context fixes issue when downcasing words like Davvi-Suopma
# -> davvi-suopmelaš

# Kept for reference: The old version, i.e.,
# The Book's treatment of paloaltolainen and pariisilainen, p. 368 ff.
# Allow optional initial downcasing after @U.Cap.Opt@
#A (->) a, 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, Á (->) á,
#Č (->) č, Đ (->) đ, Ŋ (->) ŋ, Š (->) š, Ŧ (->) ŧ, Ž (->) ž
#|| .#. %@U%.Cap%.Opt%@ _
#.o.
# No uppercase in the middle of a downcasable word
#A->a, 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  || %@U%.Cap%.Opt%@ ?+ _
#.o.
# Eliminate internal spaces inside a downcasable word
# Spaces are indicated here with the literal
# underscore character
#%_  -> [] || .#. %@U%.Cap%.Opt%@ ?+ _

] ;

substitute defined DowncaseFlag  for DCASEFLAG
substitute defined CmpHyph       for HYPHFLAG
substitute defined LexRootFlag   for LEXROOTFLAG
substitute defined LexProperFlag for LEXPROPERFLAG