More in-depth refactoring and cleanup (#6123)

* Avoid duplicating the dash case

* Pull common functions out of case branches

* Make sure list lengths are only calculated once

* Use unless

* Simplify parseURIReference' and avoid an unnecessary call to length

* Use <$> instead of reimplementing it

* Use swap instead of reimplementing it

* Remove eta-expansion that's been unnecessary since 90f5dd8

* Use tailDef instead of reimplementing it

* Use second instead of fmap, per @tarleb

src/Text/Pandoc/App.hs

@@ -88,9 +88,8 @@ convertWithOpts opts = do
   let filters' = filters ++ [ JSONFilter "pandoc-citeproc" | needsCiteproc ]
 
   let sources = case optInputFiles opts of
-                     Nothing -> ["-"]
-                     Just xs | optIgnoreArgs opts -> ["-"]
-                             | otherwise  -> xs
+                     Just xs | not (optIgnoreArgs opts) -> xs
+                     _ -> ["-"]
 
   datadir <- case optDataDir opts of
                   Nothing   -> do
@@ -192,10 +191,10 @@ convertWithOpts opts = do
               "use '-o -' to force output to stdout."
 
 
-    abbrevs <- Set.fromList . filter (not . T.null) . T.lines <$>
+    abbrevs <- Set.fromList . filter (not . T.null) . T.lines . UTF8.toText <$>
                case optAbbreviations opts of
-                    Nothing -> UTF8.toText <$> readDataFile "abbreviations"
-                    Just f  -> UTF8.toText <$> readFileStrict f
+                    Nothing -> readDataFile "abbreviations"
+                    Just f  -> readFileStrict f
 
     metadata <- if format == "jats" &&
                    isNothing (lookupMeta "csl" (optMetadata opts)) &&
@@ -341,13 +340,13 @@ readSource src = case parseURI src of
                    then BS.getContents
                    else BS.readFile fp
           E.catch (return $! UTF8.toText bs)
-             (\e -> case e of
-                         TSE.DecodeError _ (Just w) -> do
+             (\e -> E.throwIO $ case e of
+                         TSE.DecodeError _ (Just w) ->
                            case BS.elemIndex w bs of
-                             Just offset -> E.throwIO $
+                             Just offset ->
                                   PandocUTF8DecodingError (T.pack fp) offset w
-                             _ -> E.throwIO $ PandocUTF8DecodingError (T.pack fp) 0 w
-                         _ -> E.throwIO $ PandocAppError (tshow e))
+                             _ -> PandocUTF8DecodingError (T.pack fp) 0 w
+                         _ -> PandocAppError (tshow e))
 
 readURI :: FilePath -> PandocIO Text
 readURI src = UTF8.toText . fst <$> openURL (T.pack src)

src/Text/Pandoc/App/CommandLineOptions.hs

@@ -28,6 +28,7 @@ import Control.Monad.Trans
 import Control.Monad.Except (throwError)
 import Data.Aeson.Encode.Pretty (encodePretty', Config(..), keyOrder,
          defConfig, Indent(..), NumberFormat(..))
+import Data.Bifunctor (second)
 import Data.Char (toLower)
 import Data.List (intercalate, sort)
 #ifdef _WINDOWS
@@ -36,6 +37,7 @@ import Data.List (isPrefixOf)
 #endif
 #endif
 import Data.Maybe (fromMaybe, isJust)
+import Safe (tailDef)
 import Skylighting (Style, Syntax (..), defaultSyntaxMap, parseTheme)
 import System.Console.GetOpt
 import System.Environment (getArgs, getProgName)
@@ -981,10 +983,7 @@ writersNames = sort
   ("pdf" : map (T.unpack . fst) (writers :: [(Text, Writer PandocIO)]))
 
 splitField :: String -> (String, String)
-splitField s =
-  case break (`elemText` ":=") s of
-       (k,_:v) -> (k,v)
-       (k,[])  -> (k,"true")
+splitField = second (tailDef "true") . break (`elemText` ":=")
 
 -- | Apply defaults from --defaults file.
 applyDefaults :: Opt -> FilePath -> IO Opt
@@ -994,10 +993,10 @@ applyDefaults opt file = runIOorExplode $ do
               else file
   setVerbosity $ optVerbosity opt
   dataDirs <- liftIO defaultUserDataDirs
-  let fps = case optDataDir opt of
-              Nothing -> (fp : map (</> ("defaults" </> fp))
-                               dataDirs)
-              Just dd -> [fp, dd </> "defaults" </> fp]
+  let fps = fp : case optDataDir opt of
+              Nothing -> map (</> ("defaults" </> fp))
+                               dataDirs
+              Just dd -> [dd </> "defaults" </> fp]
   fp' <- fromMaybe fp <$> findFile fps
   inp <- readFileLazy fp'
   case Y.decode1 inp of

src/Text/Pandoc/BCP47.hs

@@ -93,14 +93,16 @@ parseBCP47 lang =
           ds <- P.option "" (P.count 1 P.digit)
           cs <- P.many1 asciiLetter
           let var = ds ++ cs
+              lv = length var
           guard $ if null ds
-                     then length var >= 5 && length var <= 8
-                     else length var == 4
+                     then lv >= 5 && lv <= 8
+                     else lv == 4
           return $ T.toLower $ T.pack var
         pExtension = P.try $ do
           P.char '-'
           cs <- P.many1 $ P.satisfy (\c -> isAscii c && isAlphaNum c)
-          guard $ length cs >= 2 && length cs <= 8
+          let lcs = length cs
+          guard $ lcs >= 2 && lcs <= 8
           return $ T.toLower $ T.pack cs
         pPrivateUse = P.try $ do
           P.char '-'

src/Text/Pandoc/CSV.hs

@@ -18,7 +18,7 @@ module Text.Pandoc.CSV (
 ) where
 
 import Prelude
-import Control.Monad (void)
+import Control.Monad (unless, void)
 import Data.Text (Text)
 import qualified Data.Text as T
 import Text.Parsec
@@ -63,10 +63,10 @@ pCSVQuotedCell opts = do
   return $ T.pack res
 
 escaped :: CSVOptions -> Parser Char
-escaped opts =
+escaped opts = try $
   case csvEscape opts of
-       Nothing -> try $ char (csvQuote opts) >> char (csvQuote opts)
-       Just c  -> try $ char c >> noneOf "\r\n"
+       Nothing -> char (csvQuote opts) >> char (csvQuote opts)
+       Just c  -> char c >> noneOf "\r\n"
 
 pCSVUnquotedCell :: CSVOptions -> Parser Text
 pCSVUnquotedCell opts = T.pack <$>
@@ -76,9 +76,7 @@ pCSVUnquotedCell opts = T.pack <$>
 pCSVDelim :: CSVOptions -> Parser ()
 pCSVDelim opts = do
   char (csvDelim opts)
-  if csvKeepSpace opts
-     then return ()
-     else skipMany (oneOf " \t")
+  unless (csvKeepSpace opts) $ skipMany (oneOf " \t")
 
 endline :: Parser ()
 endline = do

src/Text/Pandoc/Class.hs

@@ -524,12 +524,11 @@ alertIndent (l:ls) = do
 -- single-letter schemes.  Reason:  these are usually windows absolute
 -- paths.
 parseURIReference' :: T.Text -> Maybe URI
-parseURIReference' s =
-  case parseURIReference (T.unpack s) of
-       Just u
-         | length (uriScheme u) > 2  -> Just u
-         | null (uriScheme u)        -> Just u  -- protocol-relative
-       _                             -> Nothing
+parseURIReference' s = do
+  u <- parseURIReference (T.unpack s)
+  case uriScheme u of
+       [_] -> Nothing
+       _   -> Just u
 
 -- | Set the user data directory in common state.
 setUserDataDir :: PandocMonad m
@@ -589,10 +588,10 @@ downloadOrRead s = do
                           uriFragment = "" }
          dropFragmentAndQuery = T.takeWhile (\c -> c /= '?' && c /= '#')
          fp = unEscapeString $ T.unpack $ dropFragmentAndQuery s
-         mime = case takeExtension fp of
-                     ".gz" -> getMimeType $ dropExtension fp
-                     ".svgz" -> getMimeType $ dropExtension fp ++ ".svg"
-                     x     -> getMimeType x
+         mime = getMimeType $ case takeExtension fp of
+                     ".gz" -> dropExtension fp
+                     ".svgz" -> dropExtension fp ++ ".svg"
+                     x     -> x
          ensureEscaped = T.pack . escapeURIString isAllowedInURI . T.unpack . T.map convertSlash
          convertSlash '\\' = '/'
          convertSlash x    = x

src/Text/Pandoc/ImageSize.hs

@@ -212,9 +212,7 @@ showInPixel opts dim = T.pack $ show $ inPixel opts dim
 numUnit :: T.Text -> Maybe (Double, T.Text)
 numUnit s =
   let (nums, unit) = T.span (\c -> isDigit c || ('.'==c)) s
-  in  case safeRead nums of
-        Just n  -> Just (n, unit)
-        Nothing -> Nothing
+  in (\n -> (n, unit)) <$> safeRead nums
 
 -- | Scale a dimension by a factor.
 scaleDimension :: Double -> Dimension -> Dimension

src/Text/Pandoc/MIME.hs

@@ -18,6 +18,7 @@ import Data.List (isPrefixOf, isSuffixOf)
 import qualified Data.Map as M
 import qualified Data.Text as T
 import Data.Maybe (fromMaybe, listToMaybe)
+import Data.Tuple (swap)
 import System.FilePath
 
 type MimeType = T.Text
@@ -50,7 +51,7 @@ mediaCategory :: FilePath -> Maybe T.Text
 mediaCategory fp = getMimeType fp >>= listToMaybe . T.splitOn "/"
 
 reverseMimeTypes :: M.Map MimeType T.Text
-reverseMimeTypes = M.fromList $ map (\(k,v) -> (v,k)) mimeTypesList
+reverseMimeTypes = M.fromList $ map swap mimeTypesList
 
 mimeTypes :: M.Map T.Text MimeType
 mimeTypes = M.fromList mimeTypesList

src/Text/Pandoc/Writers.hs

@@ -125,7 +125,7 @@ data Writer m = TextWriter (WriterOptions -> Pandoc -> m Text)
 writers :: PandocMonad m => [ (Text, Writer m) ]
 writers = [
    ("native"       , TextWriter writeNative)
-  ,("json"         , TextWriter $ \o d -> writeJSON o d)
+  ,("json"         , TextWriter writeJSON)
   ,("docx"         , ByteStringWriter writeDocx)
   ,("odt"          , ByteStringWriter writeODT)
   ,("pptx"         , ByteStringWriter writePowerpoint)