[GHC] #8752: System.FilePath.Windows.combine does not really check isAbsolute

#8752: System.FilePath.Windows.combine does not really check isAbsolute -------------------------------------+--------------------------------- Reporter: joeyhess | Owner: Type: bug | Status: new Priority: normal | Milestone: Component: libraries (other) | Version: 7.6.3 Keywords: filepath | Operating System: Windows Architecture: Unknown/Multiple | Type of failure: None/Unknown Difficulty: Unknown | Test Case: Blocked By: | Blocking: Related Tickets: | -------------------------------------+--------------------------------- "Combine two paths, if the second path isAbsolute, then it returns the second.¨ But, the implementation of combine checks if the first character of a path is a path separator, which on Windows is not the same as checking if isAbsolute. This can have counterintuitive results. For example: {{{ import System.FilePath.Windows prop_windows_is_sane :: Bool prop_windows_is_sane = isAbsolute p || ("C:\\STUFF" > p /= p) where p = "\\foo\\bar" }}} -- Ticket URL: http://ghc.haskell.org/trac/ghc/ticket/8752 GHC http://www.haskell.org/ghc/ The Glasgow Haskell Compiler

#8752: System.FilePath.Windows.combine does not really check isAbsolute -------------------------------------+------------------------------------- Reporter: joeyhess | Owner: thomie Type: bug | Status: new Priority: normal | Milestone: Component: libraries | Version: 7.9 (other) | Keywords: filepath Resolution: | Architecture: Unknown/Multiple Operating System: Windows | Difficulty: Unknown Type of failure: | Blocked By: None/Unknown | Related Tickets: Test Case: | Blocking: | Differential Revisions: | -------------------------------------+------------------------------------- Changes (by thomie): * owner: => thomie * version: 7.6.3 => 7.9 Comment: Are you proposing a change to the behavior of `combine`, or can we just change its docstring to: "Combine two paths, if the second path 'isAbsolute' or starts with a path separator, then it returns the second." -- Ticket URL: http://ghc.haskell.org/trac/ghc/ticket/8752#comment:1 GHC http://www.haskell.org/ghc/ The Glasgow Haskell Compiler

#8752: System.FilePath.Windows.combine does not really check isAbsolute -------------------------------------+------------------------------------- Reporter: joeyhess | Owner: thomie Type: bug | Status: new Priority: normal | Milestone: Component: libraries | Version: 7.9 (other) | Keywords: filepath Resolution: | Architecture: Unknown/Multiple Operating System: Windows | Difficulty: Unknown Type of failure: | Blocked By: None/Unknown | Related Tickets: Test Case: | Blocking: | Differential Revisions: | -------------------------------------+------------------------------------- Comment (by joeyhess): Changing the doc string in only System.FilePath.Windows.combine would be confusing because System.FilePath.Posix.combine has the same doc string. And I assume most users just use System.FilePath and assume it works for any OS, so probably won't notice if the windows version has a caveat added. Would changing the implementation to really check isAbsolute be likely to break things? -- Ticket URL: http://ghc.haskell.org/trac/ghc/ticket/8752#comment:2 GHC http://www.haskell.org/ghc/ The Glasgow Haskell Compiler

#8752: System.FilePath.Windows.combine does not really check isAbsolute -------------------------------------+------------------------------------- Reporter: joeyhess | Owner: thomie Type: bug | Status: new Priority: normal | Milestone: Component: libraries | Version: 7.9 (other) | Keywords: filepath Resolution: | Architecture: Unknown/Multiple Operating System: Windows | Difficulty: Unknown Type of failure: | Blocked By: None/Unknown | Related Tickets: Test Case: | Blocking: | Differential Revisions: | -------------------------------------+------------------------------------- Comment (by thomie): Can you explain why the docstring I posed is confusing? It is valid for both Posix as Windows, so it's not really a caveat. Correct me if I'm wrong, but the way I understand it is that on [http://msdn.microsoft.com/en-us/library/windows/desktop/aa365247.aspx Windows], if a filepath starts with a single slash, it is considered to be relative to the root of the current drive. The system keeps track of the current drive along with the current directory of that drive. Currently, if the second argument starts with a slash, `combine` returns it as is. The implementation is essentially: {{{ combine a b | hasDrive b || hasLeadingPathSeparator b = b | otherwise = combineAlways a b }}} I am guessing you are looking for the following behavior: {{{ combine "C:\\" "\\foo\\bar" == "C:\\foo\\bar" combine "C:\\STUFF" "\\foo\\bar" == "C:\\foo\\bar" combine ""\\\\shared" "\\foo\\bar" == "\\\\shared\\foo\\bar" combine ""\\\\shared\\stuff" "\\foo\\bar" == "\\\\shared\\foo\\bar" }}} But, since anything else doesn't make any sense: {{{ combine "home" "/bob" == "/bob" }}} So just changing the check to `isAbsolute` (or `hasDrive`, since combining with "C:foo", which isRelative, is not implemented at the moment, but that is another issue) doesn't work. An implementation that has this behavior would be: {{{ combine a b | hasDrive b || hasLeadingPathSeparator b && not (hasDrive a) = b | hasLeadingPathSeparator b = combineAlways (takeDrive a) (dropLeadingPathSeparator b) | otherwise = combineAlways a b }}} Or, an option with the same behavior as above, except: {{{ combine "C:\\STUFF" "\\foo\\bar" == "\\foo\\bar" combine ""\\\\shared\\stuff" "\\foo\\bar" == "\\foo\\bar" }}} , would be: {{{ combine a b | hasDrive b || hasLeadingPathSeparator b && not (isDrive a) = b | hasLeadingPathSeparator b = combineAlways a (dropLeadingPathSeparator b) | otherwise = combineAlways a b }}} I'm not sure which of the three is better. We have to change the docstring either way. -- Ticket URL: http://ghc.haskell.org/trac/ghc/ticket/8752#comment:3 GHC http://www.haskell.org/ghc/ The Glasgow Haskell Compiler

#8752: System.FilePath.Windows.combine does not really check isAbsolute -------------------------------------+------------------------------------- Reporter: joeyhess | Owner: thomie Type: bug | Status: closed Priority: normal | Milestone: Component: Core | Version: 7.9 Libraries | Keywords: filepath Resolution: duplicate | Architecture: Unknown/Multiple Operating System: Windows | Difficulty: Unknown Type of failure: | Blocked By: None/Unknown | Related Tickets: Test Case: | Blocking: | Differential Revisions: | -------------------------------------+------------------------------------- Changes (by NeilMitchell): * cc: core-libraries-committee@… (added) * status: new => closed * resolution: => duplicate Comment: I've migrated this ticket to https://github.com/haskell/filepath/issues/8, where it will be dealt with outside GHC, and then picked up by GHC when next merged. -- Ticket URL: http://ghc.haskell.org/trac/ghc/ticket/8752#comment:5 GHC http://www.haskell.org/ghc/ The Glasgow Haskell Compiler
participants (1)
-
GHC