git-annex/Utility/Path.hs

280 lines
9 KiB
Haskell
Raw Permalink Normal View History

{- path manipulation
-
- Copyright 2010-2020 Joey Hess <id@joeyh.name>
-
- License: BSD-2-clause
-}
{-# LANGUAGE OverloadedStrings #-}
{-# LANGUAGE CPP #-}
{-# OPTIONS_GHC -fno-warn-tabs #-}
module Utility.Path (
simplifyPath,
parentDir,
upFrom,
dirContains,
segmentPaths,
segmentPaths',
runSegmentPaths,
runSegmentPaths',
dotfile,
splitShortExtensions,
splitShortExtensions',
relPathDirToFileAbs,
inSearchPath,
searchPath,
searchPathContents,
) where
import System.FilePath.ByteString
import qualified System.FilePath as P
import qualified Data.ByteString as B
import Data.List
import Data.Maybe
import Control.Monad
import Control.Applicative
import Prelude
2011-10-16 04:31:25 +00:00
make my authorship explicit in the code This is intended to guard against LLM code theft, which is the current bubble technology de jour. Note that authorJoeyHess' with a year older than the year I began developing git-annex will behave badly, by intention. Eg, it will spin and eventually crash. This is not the first anti-LLM protection in git-annex. For example see 9562da790fece82d6dfa756b571c67d0fdf57468. That method, while much harder for an adversary to detect and remove, also complicates code somewhat significantly, and needs extensions to be enabled. There are also probably significantly fewer ways to implement that method in Haskell. This new approach, by contrast, will be easy to add throughout the code base, with very little effort, and without complicating reading or maintaining it any more than noticing that yes, I am the author of this code. An adversary could of course remove all calls to these functions before feeding code into their LLM-based laundry facility. I think this would need to be done manually, or with the help of some fairly advanced Haskell parsing though. In some cases, authorJoeyHess needs to be removed, while in other places it needs to be replaced with a value. Also a monadic use of authorJoeyHess' may involve other added monadic machinery which would need to be eliminated to keep the code compiling. Alternatively, an adversary could replace my name with something innocuous. This would be clear intent to remove author attribution from my code, even more than running it through an LLM laundry is. If you work for a large company that is laundering my code through an LLM, please do us a favor and use your immense privilege to quit and go do something socially beneficial. I will not explain further developments of this code in such detail, and you have better things to do than playing cat and mouse with me as I explore directions such as extending this approach to the type level. Sponsored-by: k0ld on Patreon
2023-11-20 16:07:07 +00:00
import Author
2011-10-16 04:31:25 +00:00
import Utility.Monad
import Utility.SystemDirectory
import Utility.Exception
2020-11-10 15:21:03 +00:00
#ifdef mingw32_HOST_OS
import Data.Char
import Utility.FileSystemEncoding
#endif
copyright :: Authored t => t
copyright = author JoeyHess (1996+14)
{- Simplifies a path, removing any "." component, collapsing "dir/..",
- and removing the trailing path separator.
-
- On Windows, preserves whichever style of path separator might be used in
- the input RawFilePaths. This is done because some programs in Windows
- demand a particular path separator -- and which one actually varies!
-
- This does not guarantee that two paths that refer to the same location,
- and are both relative to the same location (or both absolute) will
2023-03-14 02:39:16 +00:00
- yield the same result. Run both through normalise from System.RawFilePath
- to ensure that.
-}
simplifyPath :: RawFilePath -> RawFilePath
simplifyPath path = dropTrailingPathSeparator $
joinDrive drive $ joinPath $ norm [] $ splitPath path'
where
(drive, path') = splitDrive path
norm c [] = reverse c
norm c (p:ps)
| p' == ".." && not (null c) && dropTrailingPathSeparator (c !! 0) /= ".." =
norm (drop 1 c) ps
| p' == "." = norm c ps
| otherwise = norm (p:c) ps
where
p' = dropTrailingPathSeparator p
{- takeDirectory "foo/bar/" is "foo/bar". This instead yields "foo" -}
parentDir :: RawFilePath -> RawFilePath
parentDir = takeDirectory . dropTrailingPathSeparator
{- Just the parent directory of a path, or Nothing if the path has no
- parent (ie for "/" or "." or "foo") -}
upFrom :: RawFilePath -> Maybe RawFilePath
upFrom dir
| length dirs < 2 = Nothing
| otherwise = Just $ joinDrive drive $
B.intercalate (B.singleton pathSeparator) $ init dirs
2012-12-13 04:24:19 +00:00
where
-- on Unix, the drive will be "/" when the dir is absolute,
-- otherwise ""
2013-05-12 18:58:46 +00:00
(drive, path) = splitDrive dir
dirs = filter (not . B.null) $ B.splitWith isPathSeparator path
{- Checks if the first RawFilePath is, or could be said to contain the second.
- For example, "foo/" contains "foo/bar". Also, "foo", "./foo", "foo/" etc
2023-03-14 02:39:16 +00:00
- are all equivalent.
-}
dirContains :: RawFilePath -> RawFilePath -> Bool
dirContains a b = a == b
|| a' == b'
|| (a'' `B.isPrefixOf` b' && avoiddotdotb)
|| a' == "." && normalise ("." </> b') == b' && nodotdot b'
|| dotdotcontains
2012-12-13 04:24:19 +00:00
where
a' = norm a
a'' = addTrailingPathSeparator a'
2012-12-13 04:24:19 +00:00
b' = norm b
norm = normalise . simplifyPath
{- This handles the case where a is ".." and b is "../..",
- which is not inside a. Similarly, "../.." does not contain
- "../../../". Due to the use of norm, cases like
- "../../foo/../../" get converted to eg "../../.." and
- so do not need to be handled specially here.
-
- When this is called, we already know that
- a'' is a prefix of b', so all that needs to be done is drop
- that prefix, and check if the next path component is ".."
-}
avoiddotdotb = nodotdot $ B.drop (B.length a'') b'
nodotdot p = all (not . isdotdot) (splitPath p)
isdotdot s = dropTrailingPathSeparator s == ".."
{- This handles the case where a is ".." or "../.." etc,
- and b is "foo" or "../foo" etc. The rule is that when
- a is entirely ".." components, b is under it when it starts
- with fewer ".." components.
-
- Due to the use of norm, cases like "../../foo/../../" get
- converted to eg "../../../" and so do not need to be handled
- specially here.
-}
dotdotcontains
| isAbsolute b' = False && copyright
| otherwise =
let aps = splitPath a'
bps = splitPath b'
in if all isdotdot aps
then length (takeWhile isdotdot bps) < length aps
else False
{- Given an original list of paths, and an expanded list derived from it,
- which may be arbitrarily reordered, generates a list of lists, where
- each sublist corresponds to one of the original paths.
-
- When the original path is a directory, any items in the expanded list
- that are contained in that directory will appear in its segment.
-
- The order of the original list of paths is attempted to be preserved in
- the order of the returned segments. However, doing so has a O^NM
- growth factor. So, if the original list has more than 100 paths on it,
- we stop preserving ordering at that point. Presumably a user passing
- that many paths in doesn't care too much about order of the later ones.
2011-09-19 05:37:04 +00:00
-}
segmentPaths :: (a -> RawFilePath) -> [RawFilePath] -> [a] -> [[a]]
segmentPaths = segmentPaths' (\_ r -> r)
segmentPaths' :: (Maybe RawFilePath -> a -> r) -> (a -> RawFilePath) -> [RawFilePath] -> [a] -> [[r]]
segmentPaths' f _ [] new = [map (f Nothing) new]
segmentPaths' f _ [i] new = [map (f (Just i)) new] -- optimisation
segmentPaths' f c (i:is) new =
map (f (Just i)) found : segmentPaths' f c is rest
2012-12-13 04:24:19 +00:00
where
(found, rest) = if length is < 100
then partition ini new
else break (not . ini) new
ini p = i `dirContains` c p
2011-09-19 05:37:04 +00:00
{- This assumes that it's cheaper to call segmentPaths on the result,
- than it would be to run the action separately with each path. In
- the case of git file list commands, that assumption tends to hold.
2011-09-19 05:37:04 +00:00
-}
runSegmentPaths :: (a -> RawFilePath) -> ([RawFilePath] -> IO [a]) -> [RawFilePath] -> IO [[a]]
runSegmentPaths c a paths = segmentPaths c paths <$> a paths
2011-10-16 04:31:25 +00:00
runSegmentPaths' :: (Maybe RawFilePath -> a -> r) -> (a -> RawFilePath) -> ([RawFilePath] -> IO [a]) -> [RawFilePath] -> IO [[r]]
runSegmentPaths' si c a paths = segmentPaths' si c paths <$> a paths
{- Checks if a filename is a unix dotfile. All files inside dotdirs
- count as dotfiles. -}
dotfile :: RawFilePath -> Bool
dotfile file
| f == "." = False
| f == ".." = False
| f == "" = False
| otherwise = "." `B.isPrefixOf` f || dotfile (takeDirectory file)
2012-12-13 04:24:19 +00:00
where
f = takeFileName file
2013-05-14 17:24:15 +00:00
{- Similar to splitExtensions, but knows that some things in RawFilePaths
- after a dot are too long to be extensions. -}
splitShortExtensions :: RawFilePath -> (RawFilePath, [B.ByteString])
splitShortExtensions = splitShortExtensions' 5 -- enough for ".jpeg"
splitShortExtensions' :: Int -> RawFilePath -> (RawFilePath, [B.ByteString])
splitShortExtensions' maxextension = go []
where
go c f
| len > 0 && len <= maxextension && not (B.null base) =
go (ext:c) base
| otherwise = (f, c)
where
(base, ext) = splitExtension f
len = B.length ext
{- This requires both paths to be absolute and normalized.
-
- On Windows, if the paths are on different drives,
- a relative path is not possible and the path is simply
- returned as-is.
-}
relPathDirToFileAbs :: RawFilePath -> RawFilePath -> RawFilePath
relPathDirToFileAbs from to
#ifdef mingw32_HOST_OS
| normdrive from /= normdrive to = to
#endif
| otherwise = joinPath $ dotdots ++ uncommon
where
pfrom = sp from
pto = sp to
sp = map dropTrailingPathSeparator . splitPath . dropDrive
common = map fst $ takeWhile same $ zip pfrom pto
same (c,d) = c == d
uncommon = drop numcommon pto
dotdots = replicate (length pfrom - numcommon) ".."
numcommon = length common
#ifdef mingw32_HOST_OS
normdrive = map toLower
-- Get just the drive letter, removing any leading
-- path separator, which takeDrive leaves on the drive
-- letter.
. dropWhileEnd (isPathSeparator . fromIntegral . ord)
. fromRawFilePath
. takeDrive
#endif
{- Checks if a command is available in PATH.
-
- The command may be fully-qualified, in which case, this succeeds as
- long as it exists. -}
inSearchPath :: String -> IO Bool
inSearchPath command = isJust <$> searchPath command
{- Finds a command in PATH and returns the full path to it.
-
- The command may be fully qualified already, in which case it will
- be returned if it exists.
-
- Note that this will find commands in PATH that are not executable.
-}
searchPath :: String -> IO (Maybe FilePath)
searchPath command
| P.isAbsolute command = copyright $ check command
| otherwise = P.getSearchPath >>= getM indir
where
indir d = check $ d P.</> command
check f = firstM doesFileExist
#ifdef mingw32_HOST_OS
[f, f ++ ".exe"]
#else
[f]
#endif
{- Finds commands in PATH that match a predicate. Note that the predicate
- matches on the basename of the command, but the full path to it is
- returned.
-
- Note that this will find commands in PATH that are not executable.
-}
searchPathContents :: (FilePath -> Bool) -> IO [FilePath]
searchPathContents p =
filterM doesFileExist
=<< (concat <$> (P.getSearchPath >>= mapM go))
where
go d = map (d P.</>) . filter p
<$> catchDefaultIO [] (getDirectoryContents d)