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.
This commit is contained in:
parent
dbaea98836
commit
9eb10caa27
30 changed files with 47 additions and 38 deletions
|
@ -25,7 +25,7 @@ instance Show DottedVersion where
|
|||
normalize :: String -> DottedVersion
|
||||
normalize v = DottedVersion v $
|
||||
sum $ mult 1 $ reverse $ extend precision $ take precision $
|
||||
map readi $ split "." v
|
||||
map readi $ splitc '.' v
|
||||
where
|
||||
extend n l = l ++ replicate (n - length l) 0
|
||||
mult _ [] = []
|
||||
|
|
|
@ -162,7 +162,7 @@ findPubKeys :: GpgCmd -> String -> IO KeyIds
|
|||
findPubKeys cmd for = KeyIds . parse . lines <$> readStrict cmd params
|
||||
where
|
||||
params = [Param "--with-colons", Param "--list-public-keys", Param for]
|
||||
parse = mapMaybe (keyIdField . split ":")
|
||||
parse = mapMaybe (keyIdField . splitc ':')
|
||||
keyIdField ("pub":_:_:_:f:_) = Just f
|
||||
keyIdField _ = Nothing
|
||||
|
||||
|
@ -175,7 +175,7 @@ secretKeys cmd = catchDefaultIO M.empty makemap
|
|||
where
|
||||
makemap = M.fromList . parse . lines <$> readStrict cmd params
|
||||
params = [Param "--with-colons", Param "--list-secret-keys", Param "--fixed-list-mode"]
|
||||
parse = extract [] Nothing . map (split ":")
|
||||
parse = extract [] Nothing . map (splitc ':')
|
||||
extract c (Just keyid) (("uid":_:_:_:_:_:_:_:_:userid:_):rest) =
|
||||
extract ((keyid, decode_c userid):c) Nothing rest
|
||||
extract c (Just keyid) rest@(("sec":_):_) =
|
||||
|
|
|
@ -107,7 +107,7 @@ parseFormatted s = bundle $ go [] $ lines s
|
|||
parsemode ('u':_) = OpenReadWrite
|
||||
parsemode _ = OpenUnknown
|
||||
|
||||
splitnull = split "\0"
|
||||
splitnull = splitc '\0'
|
||||
|
||||
parsefail = error $ "failed to parse lsof output: " ++ show s
|
||||
|
||||
|
|
|
@ -45,6 +45,14 @@ separate c l = unbreak $ break c l
|
|||
| null b = r
|
||||
| otherwise = (a, tail b)
|
||||
|
||||
{- Split on a single character. This is over twice as fast as using
|
||||
- Data.List.Utils.split on a list of length 1, while producing
|
||||
- identical results. -}
|
||||
splitc :: Char -> String -> [String]
|
||||
splitc c s = case break (== c) s of
|
||||
(i, _c:rest) -> i : splitc c rest
|
||||
(i, []) -> i : []
|
||||
|
||||
{- Breaks out the first line. -}
|
||||
firstLine :: String -> String
|
||||
firstLine = takeWhile (/= '\n')
|
||||
|
|
|
@ -124,14 +124,14 @@ supported Quvi09 url = (firstlevel <&&> secondlevel)
|
|||
Nothing -> return False
|
||||
Just auth -> do
|
||||
let domain = map toLower $ uriRegName auth
|
||||
let basedomain = intercalate "." $ reverse $ take 2 $ reverse $ split "." domain
|
||||
let basedomain = intercalate "." $ reverse $ take 2 $ reverse $ splitc '.' domain
|
||||
any (\h -> domain `isSuffixOf` h || basedomain `isSuffixOf` h)
|
||||
. map (map toLower) <$> listdomains Quvi09
|
||||
secondlevel = snd <$> processTranscript "quvi"
|
||||
(toCommand [Param "dump", Param "-o", Param url]) Nothing
|
||||
|
||||
listdomains :: QuviVersion -> IO [String]
|
||||
listdomains Quvi09 = concatMap (split ",")
|
||||
listdomains Quvi09 = concatMap (splitc ',')
|
||||
. concatMap (drop 1 . words)
|
||||
. filter ("domains: " `isPrefixOf`) . lines
|
||||
<$> readQuvi (toCommand [Param "info", Param "-p", Param "domains"])
|
||||
|
|
|
@ -24,7 +24,7 @@ rsyncShell command = [Param "-e", Param $ unwords $ map escape (toCommand comman
|
|||
{- rsync requires some weird, non-shell like quoting in
|
||||
- here. A doubled single quote inside the single quoted
|
||||
- string is a single quote. -}
|
||||
escape s = "'" ++ intercalate "''" (split "'" s) ++ "'"
|
||||
escape s = "'" ++ intercalate "''" (splitc '\'' s) ++ "'"
|
||||
|
||||
{- Runs rsync in server mode to send a file. -}
|
||||
rsyncServerSend :: [CommandParam] -> FilePath -> IO Bool
|
||||
|
@ -123,7 +123,7 @@ parseRsyncProgress = go [] . reverse . progresschunks
|
|||
{- Find chunks that each start with delim.
|
||||
- The first chunk doesn't start with it
|
||||
- (it's empty when delim is at the start of the string). -}
|
||||
progresschunks = drop 1 . split [delim]
|
||||
progresschunks = drop 1 . splitc delim
|
||||
findbytesstart s = dropWhile isSpace s
|
||||
|
||||
parsebytes :: String -> Maybe Integer
|
||||
|
|
|
@ -11,7 +11,7 @@ module Utility.SafeCommand where
|
|||
|
||||
import System.Exit
|
||||
import Utility.Process
|
||||
import Data.String.Utils
|
||||
import Utility.Misc
|
||||
import System.FilePath
|
||||
import Data.Char
|
||||
import Data.List
|
||||
|
@ -86,7 +86,7 @@ shellEscape :: String -> String
|
|||
shellEscape f = "'" ++ escaped ++ "'"
|
||||
where
|
||||
-- replace ' with '"'"'
|
||||
escaped = intercalate "'\"'\"'" $ split "'" f
|
||||
escaped = intercalate "'\"'\"'" $ splitc '\'' f
|
||||
|
||||
-- | Unescapes a set of shellEscaped words or filenames.
|
||||
shellUnEscape :: String -> [String]
|
||||
|
|
Loading…
Add table
Add a link
Reference in a new issue