1
0
forked from GitHub/gf-core

pg -lexc now writes a list of multichar symbols and a title ("Root") for the lexicon, as required by Xerox lexc

This commit is contained in:
aarne
2013-02-03 10:03:15 +00:00
parent 5f2a3095f2
commit 28c59faf29

View File

@@ -1362,13 +1362,16 @@ infinity = 256
prLexcLexicon :: Morpho -> String
prLexcLexicon mo =
unlines $ "LEXICON" : [prLexc l p ++ ":" ++ w ++ " # ;" | (w,lps) <- morpho, (l,p) <- lps] ++ ["END"]
unlines $ "Multichar_Symbols":multichars:"":"LEXICON Root" : [prLexc l p ++ ":" ++ w ++ " # ;" | (w,lps) <- morpho, (l,p) <- lps] ++ ["END"]
where
morpho = fullFormLexicon mo
prLexc l p = showCId l ++ mkTags (words p)
prLexc l p = showCId l ++ concat (mkTags (words p))
mkTags p = case p of
"s":ws -> mkTags ws --- remove record field
ws -> concat $ "+" : intersperse "+" ws
ws -> map ('+':) ws
multichars = unwords $ nub $ concat [mkTags (words p) | (w,lps) <- morpho, (l,p) <- lps]
-- thick_A+(AAdj+Posit+Gen):thick's # ;
prFullFormLexicon :: Morpho -> String
prFullFormLexicon mo =