forked from GitHub/gf-core
changed Finnish VP so that the verb is stored as such. This makes compilation and parsing several times faster. TODO: treat passivization correctly; optimize further by only storing the stems.
This commit is contained in:
@@ -30,7 +30,7 @@ lin
|
||||
insertObj
|
||||
(\\_,b,a => infVPGen pol.p v.sc b a vp v.vi)
|
||||
(predSV {s = v.s ;
|
||||
sc = case vp.sc of {
|
||||
sc = case vp.s.sc of {
|
||||
NPCase Nom => v.sc ; -- minun täytyy pestä auto
|
||||
c => c -- minulla täytyy olla auto
|
||||
} ;
|
||||
@@ -96,7 +96,7 @@ lin
|
||||
insertObj (\\_,b,a => infVPGen p.p v.sc b a vp v.vi) (predSV v) ** {c2 = v.c2} ;
|
||||
|
||||
CompS s = {s = \\_ => "että" ++ s.s} ; -- S -> Comp ---- what are these expected to do ? 29/3/2013
|
||||
CompVP ant pol vp = {s = \\a => infVPGen pol.p vp.sc Pos a vp Inf1} ; -- VP -> Comp
|
||||
CompVP ant pol vp = {s = \\a => infVPGen pol.p vp.s.sc Pos a vp Inf1} ; -- VP -> Comp
|
||||
|
||||
|
||||
that_RP = which_RP ;
|
||||
|
||||
Reference in New Issue
Block a user