Add full Unicode support

README.md

@@ -9,7 +9,6 @@ A convenient converter of [DConf](https://wiki.gnome.org/Projects/dconf) files t
 * [Introduction](#introduction)
 * [Run](#run)
   * [Custom root](#custom-root)
-  * [Emoji support](#emoji-support)
 * [Supported types](#supported-types)
 * [Gnome Shell configuration](#gnome-shell-configuration)
 * [Installation](#installation)
@@ -107,20 +106,18 @@ Type `--help` for some more information.
 dconf2nix - Nixify dconf configuration files
 
 Usage: dconf2nix [-v|--version] 
-                 [[-r|--root ARG] [-e|--emoji] [--verbose] | (-i|--input ARG)
-                   (-o|--output ARG) [-r|--root ARG] [-e|--emoji] [--verbose]]
+                 [[-r|--root ARG] [--verbose] | (-i|--input ARG)
+                   (-o|--output ARG) [-r|--root ARG] [--verbose]]
   Convert a dconf file into a Nix file, as expected by Home Manager.
 
 Available options:
   -h,--help                Show this help text
   -v,--version             Show the current version
   -r,--root ARG            Custom root path. e.g.: system/locale/
-  -e,--emoji               Enable emoji support (adds a bit of overhead)
   --verbose                Verbose mode (debug)
   -i,--input ARG           Path to the dconf file (input)
   -o,--output ARG          Path to the Nix output file (to be created)
   -r,--root ARG            Custom root path. e.g.: system/locale/
-  -e,--emoji               Enable emoji support (adds a bit of overhead)
   --verbose                Verbose mode (debug)
 ```
 
@@ -145,26 +142,6 @@ This will generate an output similar to the one below.
 }
 ```
 
-#### Emoji support
-
-Emojis are supported since version `0.0.12`, and it needs to be explicitly enabled, as these add a little parsing overhead via the [emojis](https://hackage.haskell.org/package/emojis) package.
-
-The following `dconf` input.
-
-```ini
-[ org/gnome/Characters ]
-recent-characters=['💡']
-some-other-character=['🤓']
-```
-
-Can be parsed as follows.
-
-```console 
-$ dconf2nix -i data/emoji.settings -o output/emoji.nix --emoji
-```
-
-Failing to pass the `--emoji` flag will result in a timeout error.
-
 ### Supported types
 
 For now, only types supported by Home Manager as specified [here](https://github.com/rycee/home-manager/blob/master/modules/lib/gvariant.nix) are supported. If there's enough interest, we might be able to work on supporting the [full specification](https://docs.gtk.org/glib/gvariant-text.html).

app/Main.hs

@@ -13,7 +13,7 @@ import           DConf2Nix                      ( dconf2nixFile
 
 main :: IO ()
 main = runArgs >>= \case
-  FileInput (FileArgs i o r e v) ->
-    dconf2nixFile i o r e v
-  StdinInput (StdinArgs r e v)   ->
-    dconf2nixStdin r e v
+  FileInput (FileArgs i o r v) ->
+    dconf2nixFile i o r v
+  StdinInput (StdinArgs r v)   ->
+    dconf2nixStdin r v

data/dconf.settings

@@ -303,7 +303,7 @@ locations=[<(uint32 2, <('Gdańsk', 'EPGD', true, [(0.94916821905848536, 0.32230
 region='en_US.UTF-8'
 
 [issue28/desktop/ibus/panel/emoji]
-favorites=['\8211', '\8594', '\8593', '\8595', '\8482', '\\u00ad', '\176', '\\v', '\160', '\171', '\8451']
+favorites=['–', '→', '↑', '↓', '™', '\u00ad', '°', '\v', ' ', '«', '℃']
 
 [issue28/org/gnome/desktop/input-sources]
 mru-sources=[('xkb', 'us+altgr-intl'), ('ibus', 'mozc-jp')]

data/unicode.settings

@@ -0,0 +1,3 @@
+[org/gnome/font-manager]
+compare-preview-text='🤣'
+preview-text='příliš žluťoučký kůň úpěl ďábelské ódy\nprilis zlutoucky kun upel dabelske ody\n◌̍◌̍ff̍ ̍čěů\n◌̎\n'

dconf2nix.cabal

@@ -21,7 +21,6 @@ library
   other-modules:       Paths_dconf2nix
   build-depends:       base
                      , containers
-                     , emojis
                      , optparse-applicative
                      , parsec
                      , text

output/dconf.nix

@@ -361,20 +361,20 @@ with lib.hm.gvariant;
     };
 
     "org/gnome/Weather" = {
-      locations = [ (mkVariant [ (mkUint32 2) (mkVariant [ "Gda\324sk" "EPGD" true [ (mkTuple [ 0.9491682190584854 0.3223041410193837 ]) ] [ (mkTuple [ 0.9485864484589182 0.32579479952337237 ]) ] ]) ]) (mkVariant [ (mkUint32 2) (mkVariant [ "Gdynia, Dzia\322dowo County, Warmian-Masurian Voivodeship" "" false [ (mkTuple [ 0.9302794944578734 0.34699627038777753 ]) ] [ (mkTuple [ 0.938610530426954 0.3574455077502486 ]) ] ]) ]) (mkVariant [ (mkUint32 2) (mkVariant [ "Gdynia, Pomeranian Voivodeship" "" false [ (mkTuple [ 0.9514923902475622 0.3235888220312407 ]) ] [ (mkTuple [ 0.9485864484589182 0.32579479952337237 ]) ] ]) ]) ];
+      locations = [ (mkVariant [ (mkUint32 2) (mkVariant [ "Gdańsk" "EPGD" true [ (mkTuple [ 0.9491682190584854 0.3223041410193837 ]) ] [ (mkTuple [ 0.9485864484589182 0.32579479952337237 ]) ] ]) ]) (mkVariant [ (mkUint32 2) (mkVariant [ "Gdynia, Działdowo County, Warmian-Masurian Voivodeship" "" false [ (mkTuple [ 0.9302794944578734 0.34699627038777753 ]) ] [ (mkTuple [ 0.938610530426954 0.3574455077502486 ]) ] ]) ]) (mkVariant [ (mkUint32 2) (mkVariant [ "Gdynia, Pomeranian Voivodeship" "" false [ (mkTuple [ 0.9514923902475622 0.3235888220312407 ]) ] [ (mkTuple [ 0.9485864484589182 0.32579479952337237 ]) ] ]) ]) ];
     };
 
     "org/gnome/shell/weather" = {
       automatic-location = true;
-      locations = [ (mkVariant [ (mkUint32 2) (mkVariant [ "Gda\324sk" "EPGD" true [ (mkTuple [ 0.9491682190584854 0.3223041410193837 ]) ] [ (mkTuple [ 0.9485864484589182 0.32579479952337237 ]) ] ]) ]) (mkVariant [ (mkUint32 2) (mkVariant [ "Gdynia, Dzia\322dowo County, Warmian-Masurian Voivodeship" "" false [ (mkTuple [ 0.9302794944578734 0.34699627038777753 ]) ] [ (mkTuple [ 0.938610530426954 0.3574455077502486 ]) ] ]) ]) (mkVariant [ (mkUint32 2) (mkVariant [ "Gdynia, Pomeranian Voivodeship" "" false [ (mkTuple [ 0.9514923902475622 0.3235888220312407 ]) ] [ (mkTuple [ 0.9485864484589182 0.32579479952337237 ]) ] ]) ]) ];
+      locations = [ (mkVariant [ (mkUint32 2) (mkVariant [ "Gdańsk" "EPGD" true [ (mkTuple [ 0.9491682190584854 0.3223041410193837 ]) ] [ (mkTuple [ 0.9485864484589182 0.32579479952337237 ]) ] ]) ]) (mkVariant [ (mkUint32 2) (mkVariant [ "Gdynia, Działdowo County, Warmian-Masurian Voivodeship" "" false [ (mkTuple [ 0.9302794944578734 0.34699627038777753 ]) ] [ (mkTuple [ 0.938610530426954 0.3574455077502486 ]) ] ]) ]) (mkVariant [ (mkUint32 2) (mkVariant [ "Gdynia, Pomeranian Voivodeship" "" false [ (mkTuple [ 0.9514923902475622 0.3235888220312407 ]) ] [ (mkTuple [ 0.9485864484589182 0.32579479952337237 ]) ] ]) ]) ];
     };
 
     "system/locale" = {
       region = "en_US.UTF-8";
     };
 
     "issue28/desktop/ibus/panel/emoji" = {
-      favorites = [ "\137" "\396" "\395" "\397" "\322" "\\u00ad" "\176" "\\v" "\160" "\171" "\297" ];
+      favorites = [ "–" "→" "↑" "↓" "™" "­" "°" "" " " "«" "℃" ];
     };
 
     "issue28/org/gnome/desktop/input-sources" = {

output/unicode.nix

@@ -0,0 +1,14 @@
+# Generated via dconf2nix: https://github.com/nix-commmunity/dconf2nix
+{ lib, ... }:
+
+with lib.hm.gvariant;
+
+{
+  dconf.settings = {
+    "org/gnome/font-manager" = {
+      compare-preview-text = "🤣";
+      preview-text = "příliš žluťoučký kůň úpěl ďábelské ódy\nprilis zlutoucky kun upel dabelske ody\n◌̍◌̍ff̍ ̍čěů\n◌̎\n";
+    };
+
+  };
+}

src/CommandLine.hs

@@ -18,24 +18,18 @@ data FileArgs = FileArgs
   { fileInput :: InputFilePath
   , fileOutput :: OutputFilePath
   , fileRoot :: Root
-  , fileEmojiSupport :: EmojiSupport
   , fileVerbosity :: Verbosity
   }
 
 data StdinArgs = StdinArgs
   { stdinRoot :: Root
-  , stdinEmojiSupport :: EmojiSupport
   , stdinVerbosity :: Verbosity
   }
 
 verbosityArgs :: Parser Verbosity
 verbosityArgs =
   flag Normal Verbose (long "verbose" <> help "Verbose mode (debug)")
 
-emojiArgs :: Parser EmojiSupport
-emojiArgs =
-  flag Disabled Enabled (long "emoji" <> short 'e' <> help "Enable emoji support (adds a bit of overhead)")
-
 rootArgs :: Parser Root
 rootArgs = Root <$> strOption
   (long "root" <> short 'r' <> value T.empty <> help
@@ -53,12 +47,11 @@ fileArgs = fmap FileInput $ FileArgs
           )
         )
     <*> rootArgs
-    <*> emojiArgs
     <*> verbosityArgs
 
 stdinArgs :: Parser Input
 stdinArgs =
-  StdinInput <$> (StdinArgs <$> rootArgs <*> emojiArgs <*> verbosityArgs)
+  StdinInput <$> (StdinArgs <$> rootArgs <*> verbosityArgs)
 
 versionInfo :: String
 versionInfo = unlines

src/DConf.hs

@@ -7,11 +7,13 @@ module DConf
   )
 where
 
+import           Control.Monad                  ( replicateM )
+import           Data.Ix                        ( inRange )
 import qualified Data.Map                      as Map
+import           Data.Maybe                     ( catMaybes )
 import           Data.Text                      ( Text )
 import qualified Data.Text                     as T
 import           DConf.Data
-import           Text.Emoji                     ( emojis )
 import           Text.Parsec
 import           Data.Char
 
@@ -55,81 +57,86 @@ vInt64 = try $ do
   many1 (string "int64 ") >> spaces
   I64 . read <$> many1 digit
 
-vTuple :: EmojiSupport -> Parsec Text () Value
-vTuple es = do
-  rs <- bracket "(" ")" $ commaSeparated $ value es
+vTuple :: Parsec Text () Value
+vTuple = do
+  rs <- bracket "(" ")" $ commaSeparated $ value
   case rs of
     xs@(_ : _ : _) -> pure $ T xs
     _              -> fail "Not a tuple"
 
-vEmoji :: Parsec Text () Value
-vEmoji =
-  let e = T.head . snd <$> emojis
-      f = choice $ char <$> e
-      s = many1 (string "'") *> f <* string "'"
-      d = many1 (char '"') *> f <* char '"'
-  in  Emo <$> try (s <|> d)
-
 vString :: Parsec Text () Text
 vString = T.pack <$> (single <|> double)
  where
   single = bracket "'" "'" $ inputs "'"
   double = bracket "\"" "\"" $ inputs "\""
+
   lchar :: [Char] -> Parsec Text () Char
-  lchar extra = charExcept $ "\r\n" <> extra
-  octal = do
-    isOctal <- optionMaybe $ char '8'
-    let
-      base = case isOctal of
-        Just _ -> 8
-        Nothing -> 10
-    a <- digit
-    b <- digit
-    c <- digit
-    return $ chr $ (digitToInt a * base * base) + (digitToInt b * base) + digitToInt c
-  qchar :: Parsec Text () Char
+  lchar extra = charExcept $ "\r\n\\" <> extra
+
+  fromHexDigit :: Char -> Int
+  fromHexDigit n | inRange ('A', 'F') n = ord n - ord 'A' + 0xA
+  fromHexDigit n | inRange ('a', 'f') n = ord n - ord 'a' + 0xA
+  fromHexDigit n | inRange ('0', '9') n = ord n - ord '0'
+  fromHexDigit n = error $ "Expected a hexadecimal digit, '" ++ n : "' given"
+
+  hexNum :: Int -> Parsec Text () Int
+  hexNum l = do
+    digits <- replicateM l (fromHexDigit <$> hexDigit)
+    return $ foldl (\acc d -> 16 * acc + d) 0 digits
+
+  qchar :: Parsec Text () (Maybe Char)
   qchar = do
     _ <- char '\\'
-    octal <|> anyChar
-  inputs :: [Char] -> Parsec Text () String
-  inputs extra = many $ qchar <|> lchar extra
+    (
+      -- Unicode escapes of the form `\uxxxx` and `\Uxxxxxxxx` are supported, in hexadecimal.
+      (char 'u' *> (Just <$> (chr <$> hexNum 4)))
+      <|> (char 'U' *> (Just <$> (chr <$> hexNum 8)))
+      -- The usual control sequence escapes `\a`, `\b`, `\f`, `\n`, `\r`, `\t` and `\v` are supported.
+      <|> (char 'a' *> pure (Just '\a'))
+      <|> (char 'b' *> pure (Just '\b'))
+      <|> (char 'f' *> pure (Just '\f'))
+      <|> (char 'n' *> pure (Just '\n'))
+      <|> (char 'r' *> pure (Just '\r'))
+      <|> (char 't' *> pure (Just '\t'))
+      <|> (char 'v' *> pure (Just '\v'))
+      -- Additionally, a `\` before a newline character causes the newline to be ignored.
+      <|> (char '\n' *> pure Nothing)
+      -- Finally, any other character following `\` is copied literally (for example, `\"` or `\\`)
+      <|> (Just <$> anyChar))
 
-emojiParser :: EmojiSupport -> [Parsec Text () Value]
-emojiParser Enabled  = [vEmoji]
-emojiParser Disabled = []
+  inputs :: [Char] -> Parsec Text () String
+  inputs extra = catMaybes <$> (many $ qchar <|> (Just <$> lchar extra))
 
-value :: EmojiSupport -> Parsec Text () Value
-value es =
-  let xs = [vTyped es, vRecord es, vList es, vJson, vBool, vInt, vDouble, vUint32, vInt64]
-      ys = [fmap S vString, vTuple es, vVariant es]
-  in  choice (xs ++ emojiParser es ++ ys)
+value :: Parsec Text () Value
+value = choice
+  [vTyped, vRecord, vList, vJson, vBool, vInt, vDouble, vUint32, vInt64, fmap S vString, vTuple, vVariant]
 
-vVariant :: EmojiSupport -> Parsec Text () Value
-vVariant es = fmap V $ bracket "<(" ")>" $ commaSeparated $ value es
+vVariant :: Parsec Text () Value
+vVariant = fmap V $ bracket "<(" ")>" $ commaSeparated $ value
 
-vList :: EmojiSupport -> Parsec Text () Value
-vList es = fmap L $ bracket "[" "]" $ commaSeparated $ value es
+vList :: Parsec Text () Value
+vList = fmap L $ bracket "[" "]" $ commaSeparated $ value
 
-vTyped :: EmojiSupport -> Parsec Text () Value
-vTyped es = do
+vTyped :: Parsec Text () Value
+vTyped = do
   _ <- char '@'
   _ <- anyChar
   _ <- anyChar
   _ <- spaces
-  value es
+  value
 
 vJson :: Parsec Text () Value
 vJson = try $ bracket "'" "'" $ do
   _ <- lookAhead $ string "{"
   js <- many (charExcept "\r\n\'")
   pure $ Json (T.pack js)
 
-vRecord :: EmojiSupport -> Parsec Text () Value
-vRecord es = fmap R $ bracket "{" "}" $ commaSeparated $ do
+vRecord :: Parsec Text () Value
+vRecord = fmap R $ bracket "{" "}" $ commaSeparated $ do
   k <- vString
   _ <- char ':'
   _ <- spaces
-  v <- value es
+  v <- value
   return (k,v)
 
 dconfHeader :: Parsec Text () Header
@@ -139,21 +146,21 @@ dconfHeader = bracket "[" "]" $ do
   _ <- spaces
   return $ T.pack h
 
-kvLine :: EmojiSupport -> Parsec Text () (Key,Value)
-kvLine es = do
+kvLine :: Parsec Text () (Key,Value)
+kvLine = do
   k <- many1 $ satisfy $ \c -> isAlphaNum c || c == '-'
   _ <- char '='
-  v <- value es
+  v <- value
   _ <- endOfLine
   return (Key $ T.pack k,v)
 
-entryParser :: EmojiSupport -> Parsec Text () Entry
-entryParser es = do
+entryParser :: Parsec Text () Entry
+entryParser = do
   h  <- dconfHeader <* endOfLine
-  kv <- many1 $ kvLine es
+  kv <- many1 $ kvLine
   optional endOfLine
   pure $ Entry h (Map.fromList kv)
 
-dconfParser :: EmojiSupport -> Verbosity -> Parsec Text () [Entry]
-dconfParser es Normal  = manyTill (entryParser es) eof
-dconfParser es Verbose = parserTraced "dconf" $ dconfParser es Normal
+dconfParser :: Verbosity -> Parsec Text () [Entry]
+dconfParser Normal  = manyTill entryParser eof
+dconfParser Verbose = parserTraced "dconf" $ dconfParser Normal

src/DConf/Data.hs

@@ -6,7 +6,6 @@ import           Data.Text                      ( Text )
 newtype InputFilePath = InputFilePath FilePath deriving Show
 newtype OutputFilePath = OutputFilePath FilePath deriving Show
 
-data EmojiSupport = Enabled | Disabled
 data Verbosity = Normal | Verbose
 
 newtype Nix = Nix { unNix :: Text } deriving Show
@@ -21,7 +20,6 @@ data Value = S Text         -- String
            | I32 Int        -- Int32
            | I64 Int        -- Int64
            | D Double       -- Double
-           | Emo Char       -- Emoji (Unicode char)
            | T [Value]      -- Tuple of n-arity
            | L [Value]      -- List of values
            | V [Value]      -- Variant

src/DConf2Nix.hs

@@ -13,16 +13,16 @@ import           Text.Parsec                    ( ParseError
                                                 , runParser
                                                 )
 
-dconf2nixFile :: InputFilePath -> OutputFilePath -> Root -> EmojiSupport -> Verbosity -> IO ()
-dconf2nixFile (InputFilePath input) (OutputFilePath output) root es v =
+dconf2nixFile :: InputFilePath -> OutputFilePath -> Root -> Verbosity -> IO ()
+dconf2nixFile (InputFilePath input) (OutputFilePath output) root v =
   let run   = handler (T.writeFile output) (T.appendFile output) root
-      parse = parseFromFile (dconfParser es v) input
+      parse = parseFromFile (dconfParser v) input
   in  run =<< parse
 
-dconf2nixStdin :: Root -> EmojiSupport -> Verbosity -> IO ()
-dconf2nixStdin root es v =
+dconf2nixStdin :: Root -> Verbosity -> IO ()
+dconf2nixStdin root v =
   let run   = handler T.putStr T.putStr root
-      parse = runParser (dconfParser es v) () "<stdin>"
+      parse = runParser (dconfParser v) () "<stdin>"
   in  run . parse =<< T.getContents
 
 handler