From: Fredrik Kuivinen <freku045@student.liu.se>
Date: Tue, 20 Sep 2005 05:50:22 GMT
Subject: Re: Yet another set of merge test cases from the kernel tree.
Message-ID: <20050920055022.GB12979@c165.ib.student.liu.se>
In-Reply-To: <7vek7kfqs2.fsf@assigned-by-dhcp.cox.net>

On Mon, Sep 19, 2005 at 06:53:49PM -0700, Junio C Hamano wrote:
> Fredrik Kuivinen <freku045@student.liu.se> writes:
> 
> > +def mergeFile(path, oSha, oMode, aSha, aMode, bSha, bMode,
> > +              branch1Name, branch2Name):
> >...
> > +    if aMode == oMode:
> > +        mode = bMode
> > +    else:
> > +        mode = aMode
> 
> Note: preferring "our" mode if there is a conflict instead of
> barfing.  I do not know which is more useful in practice.
> 

I just realized that this piece of code is broken with respect to type
changes. Trying to merge a symlink with a file should give us a
conflict.

> > +
> > +    return [sha, mode, clean, merge]
> > +
> > +def updateFile(clean, sha, mode, path, cleanCache, onlyWd=False):
> > +    updateCache = not onlyWd and (cleanCache or (not cleanCache and clean))
> > +    updateWd = onlyWd or (not cleanCache and clean)
> > +
> > +    if updateWd:
> > +        prog = ['git-cat-file', 'blob', sha]
> > +        if stat.S_ISREG(mode):
> > +            try:
> > +                os.unlink(path)
> > +            except OSError:
> > +                pass
> > +            if mode & 0100:
> > +                mode = 0777
> > +            else:
> > +                mode = 0666
> > +            fd = os.open(path, os.O_WRONLY | os.O_TRUNC | os.O_CREAT, mode)
> > +            proc = subprocess.Popen(prog, stdout=fd)
> > +            proc.wait()
> > +            os.close(fd)
> > +        elif stat.S_ISLNK(mode):
> > +            linkTarget = runProgram(prog)
> > +            os.symlink(linkTarget, path)
> > +        else:
> > +            assert(False)
> 
> Could it happen trying to update a file 'foo' when the working
> tree has 'foo/bar' file (i.e. D/F conflict)?
> 

We get a list of all files and directories in getFilesAndDirs. This
list is then used in processEntry to check for d/f conflicts, so it
shouldn't happen in that case.

The rename path do currently not check for d/f conflicts so it could
probably happen there.

- Fredrik