Re: [PATCH/WIP 03/11] t5403: avoid doing "git add foo/bar" where foo/.git exists

On Sun, Oct 30, 2011 at 2:08 PM, Junio C Hamano [off-list ref] wrote:

       Side note. I just did this and I am not getting what you saw above.

       $ mkdir -p /var/tmp/j/y && cd /var/tmp/j/y
       $ git init; git init clone2
       $ : >3; : >clone2/2
       $ GIT_DIR=clone2/.git git add clone2/2 3
       $ GIT_DIR=clone2/.git git ls-files
       3
       clone2/2

       The behavour is different when clone2/.git already has commit, and
       whatever codepath that gives these two different behaviour needs
       to be fixed.

It's resolve_gitlink_ref() in treat_directory(). I think replacing
that call() with is_git_directory() would fix this problem. We may
want to do the same with remove_dir_recursively().

When GIT_DIR=clone2/.git is given, however, the caller explicitly declines
the repository discovery. We do not know how the repository we are dealing
with (which we were explicitly told with $GIT_DIR) and a directory whose
name is ".git" under "clone2" we happened to find in read_directory()
relates to each other, especially when our index does not have clone2 as
our submodule.

We however *do* know that our working tree is our current directory, so
it would be wrong to do this:

   $ GIT_DIR=clone2/.git git add clone2/2 3; echo $?
   error: 3 is outside our repository, possibly goverened by .git
   1

The command should just add clone2/2 and 3 as it was told to.

I am concerned about clone2/2 in this case, not 3. I guess we can
check if clone2/.git is the repo we are using. If it is, skip it.
-- 
Duy