From d5d96dba1ff5c9cd66c295665fb422c8e930ff9d Mon Sep 17 00:00:00 2001
From: Alexander Foremny <aforemny@posteo.de>
Date: Fri, 22 Dec 2023 04:03:47 +0100
Subject: fix: fix `originalText`

---
 app/Main.hs | 16 ++++++++--------
 1 file changed, 8 insertions(+), 8 deletions(-)

diff --git a/app/Main.hs b/app/Main.hs
index 4ee4913..920d3a6 100644
--- a/app/Main.hs
+++ b/app/Main.hs
@@ -23,6 +23,7 @@ import Data.Maybe (fromMaybe)
 import Data.String (IsString (fromString))
 import Data.Text qualified as T
 import Data.Text.Encoding qualified as T
+import Data.Text.IO qualified as T
 import Debug.Trace
 import GHC.Conc (getNumProcessors)
 import GHC.Generics (Generic)
@@ -91,7 +92,11 @@ main = do
       mapM_ putStrLn
         =<< parMapM (consume1 keep) (map (cwd </>) filePaths)
     Args {cmd = List} -> do
-      mapM_ (putStrLn . takeBaseName . fst)
+      mapM_
+        ( \(iFileName, index) -> do
+            putStrLn (takeBaseName iFileName)
+            T.putStrLn index.originalText
+        )
         =<< parMapM
           ( \iFileName ->
               (,) iFileName <$> decodeFile @Index ("index" </> iFileName)
@@ -188,14 +193,9 @@ ocr input =
     T.unlines <$> mapM (ocr1 tmp . (tmp </>)) imageFiles
 
 ocr1 :: FilePath -> FilePath -> IO T.Text
-ocr1 tmp input = do
+ocr1 tmp input =
   T.decodeUtf8 . LB.toStrict
-    <$> sh
-      ( printf
-          "tesseract '%s' '%s' pdf"
-          (tmp </> input)
-          (tmp </> takeBaseName input)
-      )
+    <$> sh (printf "tesseract '%s' -" (tmp </> input))
 
 data Index = Index
   { originalText :: T.Text
-- 
cgit v1.2.3