Re: [PATCH v3 2/4] introduce submodule.superprojectGitDir record
On Sat, Sep 04, 2021 at 02:20:51PM -0300, Matheus Tavares wrote:
Show 27 quoted lines
>
> Emily Shaffer <emilyshaffer@google.com> wrote:
> >
> > diff --git a/builtin/submodule--helper.c b/builtin/submodule--helper.c
> > index d55f6262e9..d60fcd2c7d 100644
> > --- a/builtin/submodule--helper.c
> > +++ b/builtin/submodule--helper.c
> > @@ -1910,6 +1910,10 @@ static int module_clone(int argc, const char **argv, const char *prefix)
> > git_config_set_in_file(p, "submodule.alternateErrorStrategy",
> > error_strategy);
> >
> > + git_config_set_in_file(p, "submodule.superprojectGitdir",
> > + relative_path(absolute_path(get_git_dir()),
> > + path, &sb));
>
> This will be executed when cloning a submodule with
> `git submodule add <url/path> <path>`. Do we also want to set
> submodule.superprojectGitdir when adding a repository that already exists in
> the working tree as a submodule? I.e., something like:
>
> git init super
> git init super/sub
> [ make commits in super/sub ]
> git -C super submodule add ./sub
>
> I don't know if this workflow is so commonly used, though... It may not be
> worth the additional work.
Yeah, I think it is covered in the next patch with 'git submodule absorbgitdirs'.
Show 49 quoted lines
>
> Another option, which I believe was suggested by Jonathan Nieder on the Review
> Club, is to change the code to absorb the gitdir when adding the local
> submodule. Then, the configuration would already be set by the
> `absorb_git_dir...()` function itself.
>
> > free(sm_alternate);
> > free(error_strategy);
> >
> > diff --git a/t/t7400-submodule-basic.sh b/t/t7400-submodule-basic.sh
> > index 4bc6b6c886..e407329d81 100755
> > --- a/t/t7400-submodule-basic.sh
> > +++ b/t/t7400-submodule-basic.sh
> > @@ -108,14 +108,18 @@ test_expect_success 'setup - repository to add submodules to' '
> > submodurl=$(pwd -P)
> >
> > inspect() {
> > - dir=$1 &&
> > -
> > - git -C "$dir" for-each-ref --format='%(refname)' 'refs/heads/*' >heads &&
> > - { git -C "$dir" symbolic-ref HEAD || :; } >head &&
> > - git -C "$dir" rev-parse HEAD >head-sha1 &&
> > - git -C "$dir" update-index --refresh &&
> > - git -C "$dir" diff-files --exit-code &&
> > - git -C "$dir" clean -n -d -x >untracked
> > + sub_dir=$1 &&
> > + super_dir=$2 &&
> > +
> > + git -C "$sub_dir" for-each-ref --format='%(refname)' 'refs/heads/*' >heads &&
> > + { git -C "$sub_dir" symbolic-ref HEAD || :; } >head &&
> > + git -C "$sub_dir" rev-parse HEAD >head-sha1 &&
> > + git -C "$sub_dir" update-index --refresh &&
> > + git -C "$sub_dir" diff-files --exit-code &&
> > + cached_super_dir="$(git -C "$sub_dir" config --get submodule.superprojectGitDir)" &&
> > + [ "$(git -C "$super_dir" rev-parse --absolute-git-dir)" \
> > + -ef "$sub_dir/$cached_super_dir" ] &&
>
> To avoid the non-POSIX `-ef`, we could perhaps do something like:
>
> super_gitdir="$(git -C "$super_dir" rev-parse --absolute-git-dir)" &&
> cached_gitdir="$(git -C "$sub_dir" config --get submodule.superprojectGitDir)" &&
> test "$cached_gitdir" = "$(test-tool path-utils relative_path "$super_gitdir" "$PWD/$sub_dir")" &&
>
> (We need the "$PWD/" at the last command because `path.c:relative_path()`
> returns the first argument as-is when one of the two paths given to it is
> absolute and the other is not.)
>
> One bonus of testing the cached path this way is that we also check that
> it is indeed being stored as a relative path :)Yep, that is what I settled on. Thanks.
- Emily