cndict 0.2.4 → 0.2.5

raw patch · 2 files changed

+7/−4 lines, 2 filesPVP ok

version bump matches the API change (PVP)

API changes (from Hackage documentation)

Files

• cndict.cabal +1/−1
• src/Data/Chinese/CCDict.hs +6/−3

@@ -2,7 +2,7 @@ -- documentation, see http://haskell.org/cabal/users-guide/  name:                cndict-version:             0.2.4+version:             0.2.5 synopsis:            Chinese/Mandarin <-> English dictionary, Chinese lexer. -- description: license:             PublicDomain

@@ -150,6 +150,8 @@       	, ("这位子", ["这","位子"])       	, ("十分钟", ["十","分钟"])       	, ("有电梯", ["有","电梯"])+        , ("中午前", ["中午","前"])+        -- , ("得很", ["得","很"]) -- Don't know how to fix this.         -- , ("家中餐馆", ["家","中餐馆"]) -- tokenizer needs to be more greedy to correctly                                            -- deal with this input.       	, ("后生活", ["后","生活"])@@ -180,13 +182,14 @@       Nothing -> []       Just (Node entries rest,_score) -> entries ++ collapseNonDet rest   where-    assocs = [ (node, nodeSum node)+    geoMean [] = 0+    geoMean n = round (fromIntegral (product n)**(recip (fromIntegral (length n))))+    assocs = [ (node, geoMean (filter (/=0) (nodeSum node)))              | node <- forest ]     wordCount word = maybe 0 subtlexWCount (M.lookup word subtlex)     entryCount (KnownWord entry) = wordCount (entryChinese entry)     entryCount UnknownWord{} = 0-    entriesSum = sum . map entryCount-    nodeSum (Node entries _) = entriesSum entries+    nodeSum (Node entries _) = map entryCount entries  -- Enhanced tokenizer, mixed non-determistic and greedy algorithm tokenizer' :: CCDict -> Text -> [Token]