d30791a · typeable/pandoc

git.iko.soy/typeable/pandoc

on master help sign in

Fall back to latin1 if UTF-8 decoding fails...

d30791aJohn MacFarlane6 years ago

src/Text/Pandoc/App.hs+7−1

@@ -1,3 +1,4 @@

1+{-# LANGUAGE LambdaCase #-}

1{-# LANGUAGE OverloadedStrings #-}

2{-# LANGUAGE OverloadedStrings #-}

2{-# LANGUAGE CPP #-}

3{-# LANGUAGE CPP #-}

3{-# LANGUAGE ScopedTypeVariables #-}

4{-# LANGUAGE ScopedTypeVariables #-}

@@ -352,7 +353,12 @@ readURI src = do

352 Just "UTF-8" -> return $ UTF8.toText bs

353 Just "UTF-8" -> return $ UTF8.toText bs

353 Just "ISO-8859-1" -> return $ T.pack $ B8.unpack bs

354 Just "ISO-8859-1" -> return $ T.pack $ B8.unpack bs

354 Just charset -> throwError $ PandocUnsupportedCharsetError charset

355 Just charset -> throwError $ PandocUnsupportedCharsetError charset

356+ Nothing -> liftIO $ -- try first as UTF-8, then as latin1

355− Nothing -> return $ UTF8.toText bs

357+ E.catch (return $! UTF8.toText bs)

358+ (\case

359+ TSE.DecodeError{} ->

360+ return $ T.pack $ B8.unpack bs

361+ e -> E.throwIO e)

356

362

357readFile' :: MonadIO m => FilePath -> m BL.ByteString

363readFile' :: MonadIO m => FilePath -> m BL.ByteString

358readFile' "-" = liftIO BL.getContents

364readFile' "-" = liftIO BL.getContents