nkjp 0.3 → 0.3.1

raw patch · 2 files changed

+31/−1 lines, 2 filesnew-component:exe:nkjp2text

Files

• nkjp.cabal +6/−1
• tools/nkjp2text.hs +25/−0

@@ -1,5 +1,5 @@ name:               nkjp-version:            0.3+version:            0.3.1 synopsis:           Manipulating the National Corpus of Polish (NKJP) description:     The library provides parsing and printing utilities for the@@ -43,4 +43,9 @@ executable named2enamex   hs-source-dirs: ., tools   main-is: named2enamex.hs+  ghc-options: -Wall -O2++executable nkjp2text+  hs-source-dirs: ., tools+  main-is: nkjp2text.hs   ghc-options: -Wall -O2

@@ -0,0 +1,25 @@+{-# LANGUAGE OverloadedStrings #-}++import           Control.Monad (forM_)+import           System.Environment (getArgs)+import           Data.Foldable (foldMap)+import qualified Data.Text.Lazy as L+import qualified Data.Text.Lazy.IO as L++import qualified Text.NKJP.Named as Ne+import qualified Text.NKJP.Morphosyntax as Mx++main :: IO ()+main = do+    [teiPath] <- getArgs+    fs <- Ne.readTrees teiPath+    forM_ fs $ \ts ->+        L.putStrLn . L.strip . showLeaves . getLeaves $ ts+  where+    getLeaves  = concatMap (foldMap getRight)+    showLeaves = L.concat . map showLeaf+    showLeaf x = if Mx.nps x+        then Mx.orth x+        else " " `L.append` Mx.orth x+    getRight (Right x) = [x]+    getRight _         = []