git-annex/Command/FromKey.hs
Joey Hess 81d402216d cache the serialization of a Key
This will speed up the common case where a Key is deserialized from
disk, but is then serialized to build eg, the path to the annex object.

Previously attempted in 4536c93bb2
and reverted in 96aba8eff7.
The problems mentioned in the latter commit are addressed now:

Read/Show of KeyData is backwards-compatible with Read/Show of Key from before
this change, so Types.Distribution will keep working.

The Eq instance is fixed.

Also, Key has smart constructors, avoiding needing to remember to update
the cached serialization.

Used git-annex benchmark:
  find is 7% faster
  whereis is 3% faster
  get when all files are already present is 5% faster
Generally, the benchmarks are running 0.1 seconds faster per 2000 files,
on a ram disk in my laptop.
2019-11-22 17:49:16 -04:00

99 lines
2.9 KiB
Haskell

{- git-annex command
-
- Copyright 2010-2019 Joey Hess <id@joeyh.name>
-
- Licensed under the GNU AGPL version 3 or higher.
-}
{-# LANGUAGE BangPatterns #-}
module Command.FromKey where
import Command
import qualified Annex.Queue
import Annex.Content
import Annex.WorkTree
import qualified Annex
import qualified Backend.URL
import Network.URI
cmd :: Command
cmd = notBareRepo $ withGlobalOptions [jsonOptions] $
command "fromkey" SectionPlumbing "adds a file using a specific key"
(paramRepeating (paramPair paramKey paramPath))
(seek <$$> optParser)
data FromKeyOptions = FromKeyOptions
{ keyFilePairs :: CmdParams
, batchOption :: BatchMode
}
optParser :: CmdParamsDesc -> Parser FromKeyOptions
optParser desc = FromKeyOptions
<$> cmdParams desc
<*> parseBatchOption
seek :: FromKeyOptions -> CommandSeek
seek o = case (batchOption o, keyFilePairs o) of
(Batch fmt, _) -> seekBatch fmt
-- older way of enabling batch input, does not support BatchNull
(NoBatch, []) -> seekBatch BatchLine
(NoBatch, ps) -> do
force <- Annex.getState Annex.force
withPairs (commandAction . start force) ps
seekBatch :: BatchFormat -> CommandSeek
seekBatch fmt = batchInput fmt parse commandAction
where
parse s =
let (keyname, file) = separate (== ' ') s
in if not (null keyname) && not (null file)
then Right $ go file (keyOpt keyname)
else Left "Expected pairs of key and filename"
go file key = starting "fromkey" (mkActionItem (key, file)) $
perform key file
start :: Bool -> (String, FilePath) -> CommandStart
start force (keyname, file) = do
let key = keyOpt keyname
unless force $ do
inbackend <- inAnnex key
unless inbackend $ giveup $
"key ("++ keyname ++") is not present in backend (use --force to override this sanity check)"
starting "fromkey" (mkActionItem (key, file)) $
perform key file
-- From user input to a Key.
-- User can input either a serialized key, or an url.
--
-- In some cases, an input can be parsed as both a key and as an uri.
-- For example, "WORM--a:a" parses as an uri. To disambiguate, check
-- the uri scheme, to see if it looks like the prefix of a key. This relies
-- on key backend names never containing a ':'.
keyOpt :: String -> Key
keyOpt s = case parseURI s of
Just u | not (isKeyPrefix (uriScheme u)) ->
Backend.URL.fromUrl s Nothing
_ -> case deserializeKey s of
Just k -> k
Nothing -> giveup $ "bad key/url " ++ s
perform :: Key -> FilePath -> CommandPerform
perform key file = lookupFileNotHidden file >>= \case
Nothing -> ifM (liftIO $ doesFileExist file)
( hasothercontent
, do
link <- calcRepo $ gitAnnexLink file key
liftIO $ createDirectoryIfMissing True (parentDir file)
liftIO $ createSymbolicLink link file
Annex.Queue.addCommand "add" [Param "--"] [file]
next $ return True
)
Just k
| k == key -> next $ return True
| otherwise -> hasothercontent
where
hasothercontent = do
warning $ file ++ " already exists with different content"
next $ return False