Don't encode/decode file paths if base >= 4.5.

Prior to base 4.5 (and perhaps earlier - check), filepaths and command
line arguments were treated as unencoded lists of bytes, not unicode
strings, so we had to work around that by encoding and decoding
them.  This commit adds CPP checks for base 4.5 that disable the
encoding/decoding.

Fixes a bug with multilingual filenames when pandoc was compiled
with ghc 7.4. Closes #540.

2 files changed, 33 insertions, 9 deletions

diff --git a/src/Text/Pandoc/UTF8.hs b/src/Text/Pandoc/UTF8.hs
index 4af155882..929bc3f89 100644
--- a/src/Text/Pandoc/UTF8.hs
+++ b/src/Text/Pandoc/UTF8.hs
@@ -39,21 +39,25 @@ module Text.Pandoc.UTF8 ( readFile
 
 where
 
+#if MIN_VERSION_base(4,5,0)
+#else
+import Codec.Binary.UTF8.String (encodeString)
+#endif
+
 #if MIN_VERSION_base(4,2,0)
 
 import System.IO hiding (readFile, writeFile, getContents,
                           putStr, putStrLn, hPutStr, hPutStrLn, hGetContents)
 import Prelude hiding (readFile, writeFile, getContents, putStr, putStrLn )
-import Codec.Binary.UTF8.String (encodeString)
 import qualified System.IO as IO
 
 readFile :: FilePath -> IO String
 readFile f = do
-  h <- openFile (encodeString f) ReadMode
+  h <- openFile (encodePath f) ReadMode
   hGetContents h
 
 writeFile :: FilePath -> String -> IO ()
-writeFile f s = withFile (encodeString f) WriteMode $ \h -> hPutStr h s
+writeFile f s = withFile (encodePath f) WriteMode $ \h -> hPutStr h s
 
 getContents :: IO String
 getContents = hGetContents stdin
@@ -76,7 +80,6 @@ hGetContents h = hSetEncoding h utf8_bom >> IO.hGetContents h
 #else
 
 import qualified Data.ByteString as B
-import Codec.Binary.UTF8.String (encodeString)
 import Data.ByteString.UTF8 (toString, fromString)
 import Prelude hiding (readFile, writeFile, getContents, putStr, putStrLn)
 import System.IO (Handle)
@@ -91,10 +94,10 @@ stripBOM s | bom `B.isPrefixOf` s = B.drop 3 s
 stripBOM s = s
 
 readFile :: FilePath -> IO String
-readFile = liftM (toString . stripBOM) . B.readFile . encodeString
+readFile = liftM (toString . stripBOM) . B.readFile . encodePath
 
 writeFile :: FilePath -> String -> IO ()
-writeFile f = B.writeFile (encodeString f) . fromString
+writeFile f = B.writeFile (encodePath f) . fromString
 
 getContents :: IO String
 getContents = liftM (toString . stripBOM) B.getContents
@@ -115,3 +118,10 @@ hPutStrLn :: Handle -> String -> IO ()
 hPutStrLn h s = hPutStr h (s ++ "\n")
 
 #endif
+
+encodePath :: FilePath -> FilePath
+#if MIN_VERSION_base(4,5,0)
+encodePath = id
+#else
+encodePath = encodeString
+#endif
diff --git a/src/pandoc.hs b/src/pandoc.hs
index 34d136aa8..7e9646c94 100644
--- a/src/pandoc.hs
+++ b/src/pandoc.hs
@@ -1,3 +1,4 @@
+{-# LANGUAGE CPP #-}
 {-
 Copyright (C) 2006-2012 John MacFarlane <jgm@berkeley.edu>
 
@@ -56,8 +57,21 @@ import Network.HTTP (simpleHTTP, mkRequest, getResponseBody, RequestMethod(..))
 import Network.URI (parseURI, isURI, URI(..))
 import qualified Data.ByteString.Lazy as B
 import Data.ByteString.Lazy.UTF8 (toString )
-import Codec.Binary.UTF8.String (decodeString, encodeString)
 import Text.CSL.Reference (Reference(..))
+#if MIN_VERSION_base(4,5,0)
+#else
+import Codec.Binary.UTF8.String (decodeString, encodeString)
+#endif
+
+encodePath, decodeArg :: FilePath -> FilePath
+#if MIN_VERSION_base(4,5,0)
+encodePath = id
+decodeArg  = id
+#else
+encodePath = encodeString
+decodeArg  = decodeString
+#endif
+
 
 copyrightMessage :: String
 copyrightMessage = "\nCopyright (C) 2006-2012 John MacFarlane\n" ++
@@ -745,7 +759,7 @@ defaultWriterName x =
 main :: IO ()
 main = do
 
-  rawArgs <- liftM (map decodeString) getArgs
+  rawArgs <- liftM (map decodeArg) getArgs
   prg <- getProgName
   let compatMode = (prg == "hsmarkdown")
 
@@ -1023,7 +1037,7 @@ main = do
              else return doc1
 
   let writeBinary :: B.ByteString -> IO ()
-      writeBinary = B.writeFile (encodeString outputFile)
+      writeBinary = B.writeFile (encodePath outputFile)
 
   let writerFn :: FilePath -> String -> IO ()
       writerFn "-" = UTF8.putStr