cndict 0.7.8 → 0.8.0

raw patch · 3 files changed

+19/−12 lines, 3 files

Files

• cndict.cabal +1/−1
• src/Data/Chinese/CCDict.hs +8/−4
• src/Data/Chinese/Segmentation.hs +10/−7

@@ -2,7 +2,7 @@ -- documentation, see http://haskell.org/cabal/users-guide/  name:                cndict-version:             0.7.8+version:             0.8.0 synopsis:            Chinese/Mandarin <-> English dictionary, Chinese lexer. -- description: license:             PublicDomain

@@ -7,6 +7,7 @@   , Entry(..)   , ppEntry   , entryVariants+  , entryOriginal   , entrySimplified   , entryTraditional   , entryWordFrequency@@ -129,7 +130,7 @@ scrapeEntry dict nth key =     case variants of       [] -> Nothing-      (v:vs) -> Just (Entry v vs)+      (v:vs) -> Just (Entry key v vs)   where     variants = scrapeVariants dict nth key @@ -189,7 +190,7 @@ --  , entryDefinition  :: [[Text]] --  } deriving ( Read, Show, Eq, Ord ) -data Entry = Entry Variant [Variant]+data Entry = Entry !Text Variant [Variant]   deriving (Show, Read, Eq, Ord) data Variant = Variant   { variantSimplified    :: !Text@@ -199,11 +200,14 @@   , variantDefinitions   :: [Text]   } deriving ( Read, Show, Eq, Ord ) +entryOriginal :: Entry -> Text+entryOriginal (Entry o _v _vs) = o+ entryVariants :: Entry -> [Variant]-entryVariants (Entry v vs) = v:vs+entryVariants (Entry _o v vs) = v:vs  dominantVariant :: Entry -> Variant-dominantVariant (Entry v vs) =+dominantVariant (Entry _o v vs) =     foldr dom v vs   where     dom v1 v2

@@ -2,6 +2,7 @@ module Data.Chinese.Segmentation   ( Token(..)   , Entry(..)+  , entryOriginal   , entrySimplified   , entryTraditional   , entryPinyin@@ -12,14 +13,16 @@   , toSimplified   ) where -import Data.Chinese.CCDict (Entry(..), entrySimplified, entryTraditional, entryWordFrequency, entryPinyin)+import           Control.Monad+import           Data.Chinese.CCDict (Entry (..), entryOriginal, entryPinyin,+                                      entrySimplified, entryTraditional,+                                      entryWordFrequency) import qualified Data.Chinese.CCDict as CC-import qualified Data.Text as T-import Data.Text (Text)-import Data.List-import Data.Maybe-import Control.Monad-import Data.Ord+import           Data.List+import           Data.Maybe+import           Data.Ord+import           Data.Text           (Text)+import qualified Data.Text           as T  data Token = KnownWord Entry | UnknownWord Text   deriving ( Read, Show, Eq, Ord )