stop using external hash programs, since cryptonite is faster
In 2013, I wrote "Cryptohash benchmarks 90 to 101% faster than external hashers". Re-benchmarking today, I found cryptonite's sha256 consistently outperformed coreutils by 10% for large files. Tested 10 mb, 100 mb, 1 gb files with both sha256 and sha512. And for smaller files, the external process startup time swamps the hash time. Perhaps cryptonite has improved. Or it could just do better on my current CPU Intel(R) Pentium(R) CPU 4410Y @ 1.50GHz). Anyway, even if cryptonite is slower in some situations, seems likely it would only be marginally slower; it's got the same class of highly optimised C code under the hood as coreutils. The main difference between the two sha256 implementations seems to be how much of the inner loop they unroll.. This commit was sponsored by Henrik Riomar on Patreon.
This commit is contained in:
parent
6001b3cf45
commit
c565340adc
6 changed files with 33 additions and 159 deletions
|
@ -1,67 +0,0 @@
|
|||
{- Calculating a SHA checksum with an external command.
|
||||
-
|
||||
- This is typically a bit faster than using Haskell libraries,
|
||||
- by around 1% to 10%. Worth it for really big files.
|
||||
-
|
||||
- Copyright 2011-2013 Joey Hess <id@joeyh.name>
|
||||
-
|
||||
- License: BSD-2-clause
|
||||
-}
|
||||
|
||||
{-# OPTIONS_GHC -fno-warn-tabs #-}
|
||||
|
||||
module Utility.ExternalSHA (externalSHA) where
|
||||
|
||||
import Utility.SafeCommand
|
||||
import Utility.Process
|
||||
import Utility.Misc
|
||||
import Utility.Exception
|
||||
|
||||
import Data.List
|
||||
import Data.Char
|
||||
import System.IO
|
||||
|
||||
externalSHA :: String -> Int -> FilePath -> IO (Either String String)
|
||||
externalSHA command shasize file = do
|
||||
v <- tryNonAsync $ readsha $ toCommand [File file]
|
||||
return $ case v of
|
||||
Right s -> sanitycheck =<< parse (lines s)
|
||||
Left _ -> Left (command ++ " failed")
|
||||
where
|
||||
readsha args = withHandle StdoutHandle createProcessSuccess p $ \h -> do
|
||||
output <- hGetContentsStrict h
|
||||
hClose h
|
||||
return output
|
||||
where
|
||||
p = (proc command args) { std_out = CreatePipe }
|
||||
|
||||
{- The first word of the output is taken to be the sha. -}
|
||||
parse [] = bad
|
||||
parse (l:_)
|
||||
| null sha = bad
|
||||
-- sha is prefixed with \ when filename contains certian chars
|
||||
| "\\" `isPrefixOf` sha = Right $ drop 1 sha
|
||||
| otherwise = Right sha
|
||||
where
|
||||
sha = fst $ separate (== ' ') l
|
||||
bad = Left $ command ++ " parse error"
|
||||
|
||||
{- Check that we've correctly parsing the output of the command,
|
||||
- by making sure the sha we read is of the expected length
|
||||
- and contains only the right characters. -}
|
||||
sanitycheck sha
|
||||
| length sha /= expectedSHALength shasize =
|
||||
Left $ "Failed to parse the output of " ++ command
|
||||
| any (`notElem` "0123456789abcdef") sha' =
|
||||
Left $ "Unexpected character in output of " ++ command ++ "\"" ++ sha ++ "\""
|
||||
| otherwise = Right sha'
|
||||
where
|
||||
sha' = map toLower sha
|
||||
|
||||
expectedSHALength :: Int -> Int
|
||||
expectedSHALength 1 = 40
|
||||
expectedSHALength 256 = 64
|
||||
expectedSHALength 512 = 128
|
||||
expectedSHALength 224 = 56
|
||||
expectedSHALength 384 = 96
|
||||
expectedSHALength _ = 0
|
Loading…
Add table
Add a link
Reference in a new issue