Some optimisations to string splitting code.

Turns out that Data.List.Utils.split is slow and makes a lot of
allocations. Here's a much simpler single character splitter that behaves
the same (even in wacky corner cases) while running in half the time and
75% the allocations.

As well as being an optimisation, this helps move toward eliminating use of
missingh.

(Data.List.Split.splitOn is nearly as slow as Data.List.Utils.split and
allocates even more.)

I have not benchmarked the effect on git-annex, but would not be surprised
to see some parsing of eg, large streams from git commands run twice as
fast, and possibly in less memory.

This commit was sponsored by Boyd Stephen Smith Jr. on Patreon.

4 files changed, 8 insertions, 8 deletions

diff --git a/Git/Command.hs b/Git/Command.hs
index adea7622e..f40dfabcd 100644
--- a/Git/Command.hs
+++ b/Git/Command.hs
@@ -91,16 +91,16 @@ pipeWrite params repo = withHandle StdinHandle createProcessSuccess $
 pipeNullSplit :: [CommandParam] -> Repo -> IO ([String], IO Bool)
 pipeNullSplit params repo = do
 	(s, cleanup) <- pipeReadLazy params repo
-	return (filter (not . null) $ split sep s, cleanup)
+	return (filter (not . null) $ splitc sep s, cleanup)
   where
-	sep = "\0"
+	sep = '\0'
 
 pipeNullSplitStrict :: [CommandParam] -> Repo -> IO [String]
 pipeNullSplitStrict params repo = do
 	s <- pipeReadStrict params repo
-	return $ filter (not . null) $ split sep s
+	return $ filter (not . null) $ splitc sep s
   where
-	sep = "\0"
+	sep = '\0'
 
 pipeNullSplitZombie :: [CommandParam] -> Repo -> IO [String]
 pipeNullSplitZombie params repo = leaveZombie <$> pipeNullSplit params repo
diff --git a/Git/Config.hs b/Git/Config.hs
index 65bd9b7ba..9b4c342a4 100644
--- a/Git/Config.hs
+++ b/Git/Config.hs
@@ -132,7 +132,7 @@ parse s
 	-- --list output will have an = in the first line
 	| all ('=' `elem`) (take 1 ls) = sep '=' ls
 	-- --null --list output separates keys from values with newlines
-	| otherwise = sep '\n' $ split "\0" s
+	| otherwise = sep '\n' $ splitc '\0' s
   where
 	ls = lines s
 	sep c = M.fromListWith (++) . map (\(k,v) -> (k, [v])) .
diff --git a/Git/Construct.hs b/Git/Construct.hs
index 765562212..489927880 100644
--- a/Git/Construct.hs
+++ b/Git/Construct.hs
@@ -26,7 +26,7 @@ module Git.Construct (
 #ifndef mingw32_HOST_OS
 import System.Posix.User
 #endif
-import qualified Data.Map as M hiding (map, split)
+import qualified Data.Map as M
 import Network.URI
 
 import Common
@@ -143,7 +143,7 @@ remoteNamedFromKey :: String -> IO Repo -> IO Repo
 remoteNamedFromKey k = remoteNamed basename
   where
 	basename = intercalate "." $ 
-		reverse $ drop 1 $ reverse $ drop 1 $ split "." k
+		reverse $ drop 1 $ reverse $ drop 1 $ splitc '.' k
 
 {- Constructs a new Repo for one of a Repo's remotes using a given
  - location (ie, an url). -}
diff --git a/Git/Ref.hs b/Git/Ref.hs
index 5b3b85324..2d8013738 100644
--- a/Git/Ref.hs
+++ b/Git/Ref.hs
@@ -144,6 +144,6 @@ legal allowonelevel s = all (== False) illegal
 	ends v = v `isSuffixOf` s
 	begins v = v `isPrefixOf` s
 
-	pathbits = split "/" s
+	pathbits = splitc '/' s
 	illegalchars = " ~^:?*[\\" ++ controlchars
 	controlchars = chr 0o177 : [chr 0 .. chr (0o40-1)]