packages/filepath.git
4 years ago#24, turn back on the splitDirectories test
Neil Mitchell [Sun, 26 Oct 2014 21:36:43 +0000 (21:36 +0000)] 
#24, turn back on the splitDirectories test

4 years agoMerge pull request #20 from thomie/takeDirectory
Neil Mitchell [Sun, 26 Oct 2014 21:29:18 +0000 (21:29 +0000)] 
Merge pull request #20 from thomie/takeDirectory

Refactor takeDirectory

4 years agoMerge pull request #19 from thomie/splitDirectories
Neil Mitchell [Sun, 26 Oct 2014 21:24:14 +0000 (21:24 +0000)] 
Merge pull request #19 from thomie/splitDirectories

Refactor splitDirectories

4 years agoMerge pull request #17 from thomie/fix-isDrive-empty
Neil Mitchell [Sun, 26 Oct 2014 21:21:40 +0000 (21:21 +0000)] 
Merge pull request #17 from thomie/fix-isDrive-empty

Bug fix: isDrive "" == False

4 years agoMerge branch 'thomie-isValid'
Neil Mitchell [Sun, 26 Oct 2014 21:18:44 +0000 (21:18 +0000)] 
Merge branch 'thomie-isValid'

4 years agoMerge branch 'isValid' of https://github.com/thomie/filepath into thomie-isValid
Neil Mitchell [Sun, 26 Oct 2014 21:18:34 +0000 (21:18 +0000)] 
Merge branch 'isValid' of https://github.com/thomie/filepath into thomie-isValid

Conflicts:
changelog.md

4 years agoMerge branch 'master' of https://github.com/haskell/filepath
Neil Mitchell [Sun, 26 Oct 2014 21:12:20 +0000 (21:12 +0000)] 
Merge branch 'master' of https://github.com/haskell/filepath

4 years ago#16, add a specific example of the Windows weirdity
Neil Mitchell [Sun, 26 Oct 2014 21:11:57 +0000 (21:11 +0000)] 
#16, add a specific example of the Windows weirdity

4 years agoMerge pull request #16 from thomie/equalFilePath
Neil Mitchell [Sun, 26 Oct 2014 21:10:29 +0000 (21:10 +0000)] 
Merge pull request #16 from thomie/equalFilePath

Bug fix: equalFilePath "C:\\" "C:" == False

4 years ago#25, comment out another test
Neil Mitchell [Sun, 26 Oct 2014 21:09:44 +0000 (21:09 +0000)] 
#25, comment out another test

4 years ago#16, generalise some equalFilePath tests
Neil Mitchell [Sun, 26 Oct 2014 21:09:23 +0000 (21:09 +0000)] 
#16, generalise some equalFilePath tests

4 years ago#22, comment out a failing test
Neil Mitchell [Sun, 26 Oct 2014 21:00:20 +0000 (21:00 +0000)] 
#22, comment out a failing test

4 years ago#24, comment out a failing test
Neil Mitchell [Sun, 26 Oct 2014 21:00:00 +0000 (21:00 +0000)] 
#24, comment out a failing test

4 years agoChange the .ghci things to run the tests via Cabal - faster
Neil Mitchell [Sun, 26 Oct 2014 20:59:50 +0000 (20:59 +0000)] 
Change the .ghci things to run the tests via Cabal - faster

4 years agoAdd a shrink method for QFilePath
Neil Mitchell [Sun, 26 Oct 2014 20:43:28 +0000 (20:43 +0000)] 
Add a shrink method for QFilePath

4 years agoMake sure :test is actually using the new code
Neil Mitchell [Sun, 26 Oct 2014 20:43:16 +0000 (20:43 +0000)] 
Make sure :test is actually using the new code

4 years agoAllow a broader range of QuickCheck
Neil Mitchell [Sun, 26 Oct 2014 20:43:06 +0000 (20:43 +0000)] 
Allow a broader range of QuickCheck

4 years agoRefactor takeDirectory
Thomas Miedema [Thu, 23 Oct 2014 12:17:20 +0000 (14:17 +0200)] 
Refactor takeDirectory

Do not reimplement `dropTrailingPathSeparator`.

4 years agoRemove unnecessary `not . null` check
Thomas Miedema [Thu, 23 Oct 2014 12:14:30 +0000 (14:14 +0200)] 
Remove unnecessary `not . null` check

If `res` and `dir` are both null, it doesn't matter which one we return.

4 years agoCleanup: rename file to dir in takeDirectory
Thomas Miedema [Sun, 14 Sep 2014 18:49:52 +0000 (20:49 +0200)] 
Cleanup: rename file to dir in takeDirectory

4 years agoRefactor splitDirectories
Thomas Miedema [Thu, 23 Oct 2014 10:59:29 +0000 (12:59 +0200)] 
Refactor splitDirectories

Use `dropTrailingPathSeparator` instead of the custom function `g` to remove
trailing path separators from FilePath components generated with `splitPath`.

Since `dropTrailingPathSeparator` does not change FilePath components
for which isDrive is True, it is no longer necessary to handle the first
FilePath component in a special way.

4 years agoBug fix: isValid "\\\\\\foo" == False
Thomas Miedema [Sat, 25 Oct 2014 22:21:24 +0000 (00:21 +0200)] 
Bug fix: isValid "\\\\\\foo" == False

Share name can not start with a slash.

4 years agoBug fix: isDrive "" == False
Thomas Miedema [Mon, 15 Sep 2014 07:48:41 +0000 (09:48 +0200)] 
Bug fix: isDrive "" == False

isDrive is only called from `dropTrailingPathSeparator` and `combineAlways`.
Both times occur after a check if the argument is not empty (i.e. null for
combineAlways, and hasTrailingPathSeparator for dropTrailingPathSeparator). So
this change is safe.

4 years agoMerge branch 'master' of https://github.com/haskell/filepath
Neil Mitchell [Sun, 26 Oct 2014 20:30:07 +0000 (20:30 +0000)] 
Merge branch 'master' of https://github.com/haskell/filepath

4 years ago#15, increase 100x the number of property tests
Neil Mitchell [Sun, 26 Oct 2014 20:28:48 +0000 (20:28 +0000)] 
#15, increase 100x the number of property tests

4 years agoBug fix: equalFilePath "C:\\" "C:" == False
Thomas Miedema [Thu, 23 Oct 2014 11:00:27 +0000 (13:00 +0200)] 
Bug fix: equalFilePath "C:\\" "C:" == False

This started out as a simple refactoring of `equalFilePath`, and later turned
out to also fix a small bug.

The refactoring is: use `dropTrailingPathSeparator` instead of the custom
function `dropTrailSlash`.

A difference between these two functions is that `dropTrailingPathSeparator`
potentially removes multiple trailing slashes, whereas `dropTrailSlash` only
removes the last one. But since we `normalise` the FilePath first, which
removes superfluous pathSeparators, this difference does not matter to us.

Another difference is that `dropTrailingPathSeparator` does not drop slashes
when the FilePath isDrive, but `dropTrailSlash` does:

dropTrailSlash "C:\\" == "C:"
dropTrailSlash "C:\\\\" == "C:\\"

dropTrailingPathSeparator "C:\\" == "C:\\"
dropTrailingPathSeparator "C:\\\\" == "C:\\\\"

As a result, equalFilePath of drives on Windows changes slightly:

Before:
equalFilePath "C:\\" "C:" == True
equalFilePath "C:\\\\" "C:" == False

After:
equalFilePath "C:\\" "C:" == False
equalFilePath "C:\\\\" "C:" == False

This can be considered a bug fix, since "C:\\foo" and "C:foo", and thus "C:\\"
and "C:", are not the same thing.

4 years agoMerge pull request #1 from thomie/cleanup
Neil Mitchell [Sun, 26 Oct 2014 20:20:37 +0000 (20:20 +0000)] 
Merge pull request #1 from thomie/cleanup

Cleanup

4 years agoMake splitFileName quickcheck test Posix only
Thomas Miedema [Thu, 23 Oct 2014 17:41:19 +0000 (19:41 +0200)] 
Make splitFileName quickcheck test Posix only

See Github #14.

4 years agoRefactor: break instead of span and 3x reverse
Thomas Miedema [Sun, 14 Sep 2014 21:45:40 +0000 (23:45 +0200)] 
Refactor: break instead of span and 3x reverse

This works because pathSeparators can only occur at the end of splitPath
elements. h is passed a pathElement without separators.

Add a test to show that superfluous (and a weird combination of)
pathSeparators are still preserved. This function does not normalise
pathSeparators, otherwise it could be simplified further.

4 years agoBug fix: pathSeparator can be forward slash on Windows
Thomas Miedema [Sun, 14 Sep 2014 19:19:30 +0000 (21:19 +0200)] 
Bug fix: pathSeparator can be forward slash on Windows

dropTrailingPathSeparator "/" would return "\\" on Windows, since that is the
default:

    pathSeparator = if isWindows then '\\' else '/'

The function `normalise` can be used to normalise pathSeparators in filepaths,
the function `dropTrailingPathSeparator` shouldn't do it for you.

The test now runs for both Posix as Windows.

4 years agoRefactor: introduce function hasLeadingPathSeparator for clarity
Thomas Miedema [Sun, 14 Sep 2014 15:46:28 +0000 (17:46 +0200)] 
Refactor: introduce function hasLeadingPathSeparator for clarity

Also fix the comment to reflect the code. This fixes #8752.

4 years agoRefactor: use hasTrailingPathSeparator
Thomas Miedema [Sun, 14 Sep 2014 16:26:21 +0000 (18:26 +0200)] 
Refactor: use hasTrailingPathSeparator

hasTrailingPathSeparator is defined as:

hasTrailingPathSeparator "" = False
hasTrailingPathSeparator x = isPathSeparator (last x)

4 years agoFix links and add more tests and comments
Thomas Miedema [Sun, 14 Sep 2014 14:52:05 +0000 (16:52 +0200)] 
Fix links and add more tests and comments

No tests are removed, only new ones added.

The semantics of `combine` is questionable. See also #8752. Not changing
anything for the moment.

4 years agoMove isRelativeDrive tests to isRelative
Thomas Miedema [Sun, 14 Sep 2014 14:52:40 +0000 (16:52 +0200)] 
Move isRelativeDrive tests to isRelative

Those tests were commented out, because isRelativeDrive is not normally
exported from the module. Instead of doing a special testing compilation by
passing the TESTING=1 environment variable, it seemed easier to rewrite the
tests to use isRelative.

Only the following trivial test is not accounted for:

    isRelativeDrive "" == True

4 years agoPut the docs back, it seems to conflict with CPP some weird way
Neil Mitchell [Sun, 19 Oct 2014 13:13:35 +0000 (14:13 +0100)] 
Put the docs back, it seems to conflict with CPP some weird way

4 years agoFix the homepage to actually exist
Neil Mitchell [Sun, 19 Oct 2014 13:10:59 +0000 (14:10 +0100)] 
Fix the homepage to actually exist

4 years agoImprove the module description
Neil Mitchell [Sun, 19 Oct 2014 13:10:44 +0000 (14:10 +0100)] 
Improve the module description

4 years agoRevert the clever import trick - it confuses Haddock
Neil Mitchell [Sun, 19 Oct 2014 13:10:28 +0000 (14:10 +0100)] 
Revert the clever import trick - it confuses Haddock

4 years agoReformat whitespace
Neil Mitchell [Sun, 19 Oct 2014 13:02:26 +0000 (14:02 +0100)] 
Reformat whitespace

4 years agoRemove dead code
Neil Mitchell [Sun, 19 Oct 2014 13:02:19 +0000 (14:02 +0100)] 
Remove dead code

4 years ago#4, make sure the tests fail if there is an error
Neil Mitchell [Sun, 19 Oct 2014 13:01:25 +0000 (14:01 +0100)] 
#4, make sure the tests fail if there is an error

4 years ago#4, make QFilePath have the right set Char instance
Neil Mitchell [Sun, 19 Oct 2014 13:01:13 +0000 (14:01 +0100)] 
#4, make QFilePath have the right set Char instance

4 years agoFix up the docs so the code has the correct indentation
Neil Mitchell [Sun, 19 Oct 2014 13:00:32 +0000 (14:00 +0100)] 
Fix up the docs so the code has the correct indentation

4 years agoAdd a :test target
Neil Mitchell [Sun, 19 Oct 2014 12:48:02 +0000 (13:48 +0100)] 
Add a :test target

4 years agoAdd ghci targets for building the docs
Neil Mitchell [Sun, 19 Oct 2014 12:45:34 +0000 (13:45 +0100)] 
Add ghci targets for building the docs

4 years agoAdd dist to the .gitignore
Neil Mitchell [Sun, 19 Oct 2014 12:45:24 +0000 (13:45 +0100)] 
Add dist to the .gitignore

4 years agoAdd a .ghci file
Neil Mitchell [Sun, 19 Oct 2014 12:42:44 +0000 (13:42 +0100)] 
Add a .ghci file

4 years agoRefactor so less is conditional on CPP
Neil Mitchell [Sun, 19 Oct 2014 12:38:22 +0000 (13:38 +0100)] 
Refactor so less is conditional on CPP

4 years agoUpdate some copyright/contact bits and pieces
Neil Mitchell [Sun, 19 Oct 2014 12:36:52 +0000 (13:36 +0100)] 
Update some copyright/contact bits and pieces

4 years agoImprove the README
Neil Mitchell [Sun, 19 Oct 2014 12:36:30 +0000 (13:36 +0100)] 
Improve the README

4 years agoRemove some redundant whitespace
Neil Mitchell [Sun, 19 Oct 2014 12:36:20 +0000 (13:36 +0100)] 
Remove some redundant whitespace

4 years agoShorten Setup.hs, make it more obvious it is uninteresting
Neil Mitchell [Sun, 19 Oct 2014 12:02:19 +0000 (13:02 +0100)] 
Shorten Setup.hs, make it more obvious it is uninteresting

4 years agoAdd a trailing newline to the Cabal file
Neil Mitchell [Sun, 19 Oct 2014 11:53:40 +0000 (12:53 +0100)] 
Add a trailing newline to the Cabal file

4 years agoSwitch the maintainer
Neil Mitchell [Sun, 19 Oct 2014 11:52:50 +0000 (12:52 +0100)] 
Switch the maintainer

4 years agoUse consistent lexical conventions in the .cabal file
Neil Mitchell [Sun, 19 Oct 2014 11:50:30 +0000 (12:50 +0100)] 
Use consistent lexical conventions in the .cabal file

4 years agoUpdate the license
Neil Mitchell [Sat, 18 Oct 2014 19:37:13 +0000 (20:37 +0100)] 
Update the license

4 years agoMore tweaks to the README
Neil Mitchell [Sat, 18 Oct 2014 19:36:38 +0000 (20:36 +0100)] 
More tweaks to the README

4 years agoUpdate the README
Neil Mitchell [Sat, 18 Oct 2014 19:34:04 +0000 (20:34 +0100)] 
Update the README

5 years agoDrop GHC 6.12.3 from build-matrix
Herbert Valerio Riedel [Mon, 15 Sep 2014 20:54:09 +0000 (22:54 +0200)] 
Drop GHC 6.12.3 from build-matrix

It's starting to require too many workarounds...

5 years agoUpdate .travis.yml
Herbert Valerio Riedel [Mon, 15 Sep 2014 20:48:24 +0000 (22:48 +0200)] 
Update .travis.yml

This adds GHC 7.8.3 and GHC HEAD to the build-matrix, and removes a couple of redundant GHC versions.

5 years agoUpdate travis badge in README.md
Herbert Valerio Riedel [Mon, 15 Sep 2014 20:33:30 +0000 (22:33 +0200)] 
Update travis badge in README.md

5 years agoBump `base` constraint
Herbert Valerio Riedel [Tue, 9 Sep 2014 14:38:57 +0000 (16:38 +0200)] 
Bump `base` constraint

...and update cabal meta-info while at it

5 years agoadd testsuite-related gitignore entries
Herbert Valerio Riedel [Thu, 26 Jun 2014 21:26:03 +0000 (23:26 +0200)] 
add testsuite-related gitignore entries

5 years agoMinor refactorings as suggested by hlint ghc-7.8 filepath-1.3.0.2-release ghc-7.8.1-release ghc-7.8.2-release
Herbert Valerio Riedel [Fri, 21 Mar 2014 18:31:22 +0000 (19:31 +0100)] 
Minor refactorings as suggested by hlint

Signed-off-by: Herbert Valerio Riedel <hvr@gnu.org>
5 years agoConvert changelog to markdown
Herbert Valerio Riedel [Fri, 21 Mar 2014 18:13:28 +0000 (19:13 +0100)] 
Convert changelog to markdown

Signed-off-by: Herbert Valerio Riedel <hvr@gnu.org>
5 years agoAdd changelog
Herbert Valerio Riedel [Thu, 14 Nov 2013 10:14:03 +0000 (11:14 +0100)] 
Add changelog

Signed-off-by: Herbert Valerio Riedel <hvr@gnu.org>
5 years agoAdd `cabal {check,sdist}` steps to `.travis.yml`
Herbert Valerio Riedel [Sat, 19 Oct 2013 07:59:45 +0000 (09:59 +0200)] 
Add `cabal {check,sdist}` steps to `.travis.yml`

...and simplify `$PATH` handling while at it. Also, make the build-dep
install workaround for GHCVER=6.12.3 more explicit.

Signed-off-by: Herbert Valerio Riedel <hvr@gnu.org>
5 years agoFill out `description` field in Cabal description
Herbert Valerio Riedel [Sat, 19 Oct 2013 07:55:50 +0000 (09:55 +0200)] 
Fill out `description` field in Cabal description

This is needed to pass `cabal check`.

Signed-off-by: Herbert Valerio Riedel <hvr@gnu.org>
5 years agoFix travis.yml, again
Herbert Valerio Riedel [Fri, 18 Oct 2013 13:43:23 +0000 (15:43 +0200)] 
Fix travis.yml, again

`cd` back to top-folder after running `GenTests.hs`

Signed-off-by: Herbert Valerio Riedel <hvr@gnu.org>
5 years agoFix-up .travis.yml
Herbert Valerio Riedel [Fri, 18 Oct 2013 13:35:16 +0000 (15:35 +0200)] 
Fix-up .travis.yml

Re-add the invocation of `cabal configure` that went mising in 03a6a15e

Signed-off-by: Herbert Valerio Riedel <hvr@gnu.org>
5 years agoAttempt to integrate test-suite w/ Cabal and Travis-CI
Herbert Valerio Riedel [Fri, 18 Oct 2013 13:28:16 +0000 (15:28 +0200)] 
Attempt to integrate test-suite w/ Cabal and Travis-CI

Signed-off-by: Herbert Valerio Riedel <hvr@gnu.org>
5 years agoMarkdownify `README.md` and remove obsolete `.bat`s
Herbert Valerio Riedel [Fri, 18 Oct 2013 12:50:03 +0000 (14:50 +0200)] 
Markdownify `README.md` and remove obsolete `.bat`s

The old dead links in readme.txt have been updated/removed, and some
typos fixed. README.md has been added to source tarball via
`extra-source-files`.

Signed-off-by: Herbert Valerio Riedel <hvr@gnu.org>
5 years agoAdd .travis.yml to test compilation with older GHCs
Herbert Valerio Riedel [Fri, 18 Oct 2013 09:33:43 +0000 (11:33 +0200)] 
Add .travis.yml to test compilation with older GHCs

This is currently a proof of concept and might be disabled again.

Signed-off-by: Herbert Valerio Riedel <hvr@gnu.org>
5 years agoRefactor & update `filepath.cabal` to `cabal-version>=1.10`
Herbert Valerio Riedel [Fri, 18 Oct 2013 09:26:33 +0000 (11:26 +0200)] 
Refactor & update `filepath.cabal` to `cabal-version>=1.10`

As of this commit, `filepath` compiles cleanly with
GHC 6.10/6.12/7.0/7.2/7.4/7.6/HEAD.

Signed-off-by: Herbert Valerio Riedel <hvr@gnu.org>
5 years agoClean up CPP/Haddock interaction
Herbert Valerio Riedel [Fri, 18 Oct 2013 09:18:10 +0000 (11:18 +0200)] 
Clean up CPP/Haddock interaction

This removes the obsolete pre-Haddock2 `__HADDOCK__` conditional check
which isn't supported anymore by Haddock2 (and in fact the conditional
isn't needed anyway).

This commit also moves more content into the
`System/FilePath/Internal.hs` template that's `#include`d by
`System/FilePath/{Posix,Windows}.hs` and as a side-effect, a part of the
Haddock module-level documentation containing usage examples is now
shown again in the generated HTML output.

Signed-off-by: Herbert Valerio Riedel <hvr@gnu.org>
6 years agoUpdate Git repo URL in `.cabal` file
Herbert Valerio Riedel [Wed, 11 Sep 2013 13:15:01 +0000 (15:15 +0200)] 
Update Git repo URL in `.cabal` file

6 years agoTest cleaning fixes
Ian Lynagh [Thu, 13 Jun 2013 22:15:19 +0000 (23:15 +0100)] 
Test cleaning fixes

6 years agoBump base lower version to 4.5 (the version GHC 7.4.1 came with)
Ian Lynagh [Fri, 30 Nov 2012 21:55:09 +0000 (21:55 +0000)] 
Bump base lower version to 4.5 (the version GHC 7.4.1 came with)

6 years agoAdd comment to .cabal file saying what GHC 7.6.1 shipped with
Ian Lynagh [Fri, 30 Nov 2012 15:59:53 +0000 (15:59 +0000)] 
Add comment to .cabal file saying what GHC 7.6.1 shipped with

7 years agoBump version to 1.3.0.1 ghc-7.6 filepath-1.3.0.1-release ghc-7.6.1-release ghc-7.6.2-release ghc-7.6.3-release
Paolo Capriotti [Tue, 10 Jul 2012 10:26:36 +0000 (11:26 +0100)] 
Bump version to 1.3.0.1

7 years agoUpdate .gitignore.
Paolo Capriotti [Tue, 6 Mar 2012 10:57:33 +0000 (10:57 +0000)] 
Update .gitignore.

7 years agoQuote path to GHC in tests/Makefile
Ian Lynagh [Fri, 20 Jan 2012 12:54:43 +0000 (12:54 +0000)] 
Quote path to GHC in tests/Makefile

8 years agoSystem.FilePath.normalise "/" should be "/", not "/.". Fixes trac #3975
Ian Lynagh [Sun, 28 Aug 2011 18:17:40 +0000 (19:17 +0100)] 
System.FilePath.normalise "/" should be "/", not "/.". Fixes trac #3975

8 years agoAdd the FilePath tests to the GHC testsuite
Ian Lynagh [Sun, 28 Aug 2011 18:17:03 +0000 (19:17 +0100)] 
Add the FilePath tests to the GHC testsuite

although they only get run if you install QuickCheck

8 years agoRemove old darcs boring file
David Terei [Wed, 10 Aug 2011 19:29:06 +0000 (12:29 -0700)] 
Remove old darcs boring file

8 years agoUse Safe Haskell
David Terei [Thu, 28 Jul 2011 18:31:47 +0000 (11:31 -0700)] 
Use Safe Haskell

8 years agoBump version number ghc-7.2 ghc-7.2.1-release ghc-7.2.2-release
Ian Lynagh [Wed, 6 Jul 2011 14:28:45 +0000 (15:28 +0100)] 
Bump version number

8 years agoUpdate source repo location to be the git repo in .cabal file
Ian Lynagh [Fri, 8 Apr 2011 10:22:47 +0000 (11:22 +0100)] 
Update source repo location to be the git repo in .cabal file

8 years agoFix warning ghc-darcs-git-switchover
Ian Lynagh [Sat, 8 Jan 2011 12:32:23 +0000 (12:32 +0000)] 
Fix warning

8 years agonormalise trailing dot
Conrad Parker [Fri, 3 Dec 2010 06:04:44 +0000 (06:04 +0000)] 
normalise trailing dot

This is version 2 of a patch for ticket #3975

This patch modifles normalise to treat paths with a trailing dot as the
corresponding directory: the trailing . is stripped, and the pathSeparator
is appended to the result of joining path components.

The special case of all path components being "." is handled first, replacing
the existing special case test for paths with a leading dot. Dots are now
dropped consistently for all path components.

Additional tests are provided for:

    normalise "./." == "./"
    normalise "bob/fred/." == "bob/fred/"

9 years agoRe-record: The current directory is ".", not "" (#2034) (patch version 2)
Simon Marlow [Fri, 7 May 2010 09:43:42 +0000 (09:43 +0000)] 
Re-record: The current directory is ".", not "" (#2034) (patch version 2)

Apparently I got myself into a darcs mess, and the previous patch I
pushed was a conflict-resolution of some kind that didn't actually
make any changes.  This is the real #2034 patch.

9 years agowarning suppression
Simon Marlow [Wed, 5 May 2010 09:58:43 +0000 (09:58 +0000)] 
warning suppression

9 years agobump to 1.2.0.0, after semantics change (current dir is now ".", not "")
Simon Marlow [Wed, 5 May 2010 09:52:12 +0000 (09:52 +0000)] 
bump to 1.2.0.0, after semantics change (current dir is now ".", not "")

9 years agoThe current directory is ".", not "" (GHC bug #2034) (patch version 2)
Simon Marlow [Wed, 4 Nov 2009 09:51:09 +0000 (09:51 +0000)] 
The current directory is ".", not "" (GHC bug #2034) (patch version 2)

So now
        splitFileName "foo" = ("./", "foo")

which gives us the additional property that

> Valid x => isValid (fst (splitFileName x))

This property is important, because it means that we can pass the
result of takeDirectory to any of the functions in System.Directory,
whereas previously we would have to check for the empty case first.

After discussion on the libraries list, I removed the second part of
the original change:

        "." </> x = x

This small bit of normalisation was there to ensure that the property

> Valid x => uncurry </> (splitFileName x) == x

still held.  However, it was arguably inconsistent (see the
discussion).  Now this property has an exception:

> Valid x => uncurry (</>) (splitFileName x) == x || fst (splitFileName x) == "./"

This is a small price to pay to gain the new property above.

9 years agolower base requirement down to base-2
Sergei Trofimovich [Wed, 16 Dec 2009 21:34:21 +0000 (21:34 +0000)] 
lower base requirement down to base-2
At least ghc-6.6.1 is able to build it without problems.
(package can be used in cabal-1.8, which is known to work
on ghc-6.4+)

9 years agoBump the version number
Ian Lynagh [Tue, 19 Jan 2010 13:28:30 +0000 (13:28 +0000)] 
Bump the version number

9 years agocorrect punctuation in Synopsis (#3762)
Simon Marlow [Wed, 16 Dec 2009 09:58:18 +0000 (09:58 +0000)] 
correct punctuation in Synopsis (#3762)

9 years agoUNDO: The current directory is ".", not "" (GHC bug #2034) (patch version 2)
Simon Marlow [Wed, 4 Nov 2009 12:29:43 +0000 (12:29 +0000)] 
UNDO: The current directory is ".", not "" (GHC bug #2034) (patch version 2)

rolling back accidentally pushed patch:

Wed Nov  4 10:50:57 GMT 2009  Simon Marlow <marlowsd@gmail.com>
  * The current directory is ".", not "" (GHC bug #2034) (patch version 2)

9 years agoThe current directory is ".", not "" (GHC bug #2034) (patch version 2)
Simon Marlow [Wed, 4 Nov 2009 10:50:57 +0000 (10:50 +0000)] 
The current directory is ".", not "" (GHC bug #2034) (patch version 2)

So now
        splitFileName "foo" = ("./", "foo")

which gives us the additional property that

> Valid x => isValid (fst (splitFileName x))

This property is important, because it means that we can pass the
result of takeDirectory to any of the functions in System.Directory,
whereas previously we would have to check for the empty case first.

After discussion on the libraries list, I removed the second part of
the original change:

        "." </> x = x

This small bit of normalisation was there to ensure that the property

> Valid x => uncurry </> (splitFileName x) == x

still held.  However, it was arguably inconsistent (see the
discussion).  Now this property has an exception:

> Valid x => uncurry (</>) (splitFileName x) == x || fst (splitFileName x) == "./"

This is a small price to pay to gain the new property above.

10 years agoBump version to 1.1.0.3
Ian Lynagh [Sun, 20 Sep 2009 14:19:02 +0000 (14:19 +0000)] 
Bump version to 1.1.0.3