tree 425c8b1f7802131449da2fc9182a26791cbba97e
parent d75219b4a8a9c44520ddca234cde992498383b89
author Jens Lehmann <Jens.Lehmann@web.de> 1330895736 +0100
committer Junio C Hamano <gitster@pobox.com> 1330924955 -0800

submodules: refactor computation of relative gitdir path

In module_clone() the rel_gitdir variable was computed differently when
"git rev-parse --git-dir" returned a relative path than when it returned
an absolute path. This is not optimal, as different code paths are used
depending on the return value of that command.

Fix that by reusing the differing path components computed for setting the
core.worktree config setting, which leaves a single code path for setting
both instead of having three and makes the code much shorter.

This also fixes the bug that in the computation of how many directories
have to be traversed up to hit the root directory of the submodule the
name of the submodule was used where the path should have been used. This
lead to problems after renaming submodules into another directory level.

Even though the "(cd $somewhere && pwd)" approach breaks the flexibility
of symlinks, that is no issue here as we have to have one relative path
pointing from the work tree to the gitdir and another pointing back, which
will never work anyway when a symlink along one of those paths is changed
because the directory it points to was moved.

Also add a test moving a submodule into a deeper directory to catch any
future breakage here and to document what has to be done when a submodule
needs to be moved until git mv learns to do that. Simply moving it to the
new location doesn't work, as the core.worktree and possibly the gitfile
setting too will be wrong. So it has to be removed from filesystem and
index, then the new location has to be added into the index and the
.gitmodules file has to be updated. After that a git submodule update will
check out the submodule at the new location.

Signed-off-by: Jens Lehmann <Jens.Lehmann@web.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>