Fwd: [PATCH] checknlink: use two testfiles

[Joel B. Mohler]

On Wednesday, December 15, 2010 07:36:43 am Joel B. Mohler wrote:
> ----------  Forwarded Message  ----------
> 
> Subject: [PATCH] checknlink: use two testfiles
> Date: Tuesday, December 14, 2010, 06:39:54 pm
> From: Adrian Buehlmann <adrian at cadifra.com>
> To: mercurial-devel at selenic.com

I confirm that the patch below fixes my symptoms.

> 
> # HG changeset patch
> # User Adrian Buehlmann <adrian at cadifra.com>
> # Date 1292369727 -3600
> # Branch stable
> # Node ID c582de5b6ac6c15cf4c8214c7183fad585eaa4ba
> # Parent  dbc546811dd6eb672756f9266d432bb9c1101adb
> checknlink: use two testfiles
> 
> works around issue2543
> 
> diff --git a/mercurial/util.py b/mercurial/util.py
> --- a/mercurial/util.py
> +++ b/mercurial/util.py
> @@ -719,21 +719,37 @@ def checklink(path):
> 
>  def checknlink(testfile):
>      '''check whether hardlink count reporting works properly'''
> -    f = testfile + ".hgtmp"
> 
> +    # testfile may be open, so we need a separate file for checking to
> +    # work around issue2543 (or testfile may get lost on Samba shares)
> +    f1 = testfile + ".hgtmp1"
> +    if os.path.lexists(f1):
> +        return False
>      try:
> -        os_link(testfile, f)
> -    except OSError:
> +        posixfile(f1, 'w').close()
> +    except IOError:
>          return False
> 
> +    f2 = testfile + ".hgtmp2"
> +    fd = None
>      try:
> +        try:
> +            os_link(f1, f2)
> +        except OSError:
> +            return False
> +
>          # nlinks() may behave differently for files on Windows shares if
>          # the file is open.
> -        fd = open(f)
> -        return nlinks(f) > 1
> +        fd = open(f2)
> +        return nlinks(f2) > 1
>      finally:
> -        fd.close()
> -        os.unlink(f)
> +        if fd is not None:
> +            fd.close()
> +        for f in (f1, f2):
> +            try:
> +                os.unlink(f)
> +            except OSError:
> +                pass
> 
>      return False
> 
> _______________________________________________
> Mercurial-devel mailing list
> Mercurial-devel at selenic.com
> http://selenic.com/mailman/listinfo/mercurial-devel
> 
> -----------------------------------------