Re: [PATCH 6/7] worktree remove: new command

2018-02-12 Thread Duy Nguyen
On Fri, Feb 2, 2018 at 6:47 PM, Eric Sunshine  wrote:
> On Wed, Jan 24, 2018 at 4:53 AM, Nguyễn Thái Ngọc Duy  
> wrote:
>> This command allows to delete a worktree. Like 'move' you cannot
>> remove the main worktree, or one with submodules inside [1].
>> [...]
>> Signed-off-by: Nguyễn Thái Ngọc Duy 
>> ---
>> diff --git a/builtin/worktree.c b/builtin/worktree.c
>> @@ -688,6 +689,132 @@ static int move_worktree(int ac, const char **av, 
>> const char *prefix)
>> +static void check_clean_worktree(struct worktree *wt,
>> +const char *original_path)
>> +{
>> +   [...]
>> +   validate_no_submodules(wt);
>
> It's slightly strange seeing worktree validation in a function
> checking whether the worktree is clean since submodule validation
> isn't an issue of cleanliness. I'd have expected the caller to invoke
> it instead:
>
> int remove_worktree(...) {
> ...
> if (!force) {
> validate_no_submodules(wt);
> check_clean_worktree(wt, av[0]);
> }
> ...
> }
>
> On the other hand, I could imagine it being called here as appropriate
> if submodule validation eventually also checks submodule cleanliness
> as hinted by the commit message.

Yes I basically consider all submodules dirty until somebody sorts
them out. I'll add a comment here to clarify this.
-- 
Duy


Re: [PATCH 6/7] worktree remove: new command

2018-02-02 Thread Eric Sunshine
On Wed, Jan 24, 2018 at 4:53 AM, Nguyễn Thái Ngọc Duy  wrote:
> This command allows to delete a worktree. Like 'move' you cannot
> remove the main worktree, or one with submodules inside [1].
> [...]
> Signed-off-by: Nguyễn Thái Ngọc Duy 
> ---
> diff --git a/builtin/worktree.c b/builtin/worktree.c
> @@ -688,6 +689,132 @@ static int move_worktree(int ac, const char **av, const 
> char *prefix)
> +static void check_clean_worktree(struct worktree *wt,
> +const char *original_path)
> +{
> +   [...]
> +   validate_no_submodules(wt);

It's slightly strange seeing worktree validation in a function
checking whether the worktree is clean since submodule validation
isn't an issue of cleanliness. I'd have expected the caller to invoke
it instead:

int remove_worktree(...) {
...
if (!force) {
validate_no_submodules(wt);
check_clean_worktree(wt, av[0]);
}
...
}

On the other hand, I could imagine it being called here as appropriate
if submodule validation eventually also checks submodule cleanliness
as hinted by the commit message.

> +   argv_array_pushf(&child_env, "%s=%s/.git",
> +GIT_DIR_ENVIRONMENT, wt->path);
> +   argv_array_pushf(&child_env, "%s=%s",
> +GIT_WORK_TREE_ENVIRONMENT, wt->path);
> +   memset(&cp, 0, sizeof(cp));
> +   argv_array_pushl(&cp.args, "status",
> +"--porcelain", "--ignore-submodules=none",
> +NULL);
> +   cp.env = child_env.argv;
> +   cp.git_cmd = 1;
> +   cp.dir = wt->path;
> +   cp.out = -1;
> +   ret = start_command(&cp);
> +   if (ret)
> +   die_errno(_("failed to run git-status on '%s'"),
> + original_path);

Minor: I think there was some effort recently to remove "git-foo"
style mentions from documentation and error messages. Perhaps this
could be "failed to run 'git status' on '%s'". Ditto below.

> +   ret = xread(cp.out, buf, sizeof(buf));
> +   if (ret)
> +   die(_("'%s' is dirty, use --force to delete it"),
> +   original_path);
> +   close(cp.out);
> +   ret = finish_command(&cp);
> +   if (ret)
> +   die_errno(_("failed to run git-status on '%s', code %d"),
> + original_path, ret);
> +}
> +
> +static int delete_git_work_tree(struct worktree *wt)
> +{
> +   struct strbuf sb = STRBUF_INIT;
> +   int ret = 0;
> +
> +   strbuf_addstr(&sb, wt->path);
> +   if (remove_dir_recursively(&sb, 0)) {
> +   error_errno(_("failed to delete '%s'"), sb.buf);
> +   ret = -1;
> +   }
> +   strbuf_release(&sb);
> +
> +   return ret;
> +}

Style nit: In the very similar delete_git_dir(), just below, there is
no blank line before 'return'.

> +
> +static int delete_git_dir(struct worktree *wt)
> +{
> +   struct strbuf sb = STRBUF_INIT;
> +   int ret = 0;
> +
> +   strbuf_addstr(&sb, git_common_path("worktrees/%s", wt->id));
> +   if (remove_dir_recursively(&sb, 0)) {
> +   error_errno(_("failed to delete '%s'"), sb.buf);
> +   ret = -1;
> +   }
> +   strbuf_release(&sb);
> +   return ret;
> +}
> +
> +static int remove_worktree(int ac, const char **av, const char *prefix)
> +{
> +   [...]
> +   if (reason) {
> +   if (*reason)
> +   die(_("cannot remove a locked working tree, lock 
> reason: %s"),
> +   reason);
> +   die(_("cannot remove a locked working tree"));
> +   }
> +   if (validate_worktree(wt, &errmsg))
> +   die(_("validation failed, cannot remove working tree:\n%s"),
> +   errmsg.buf);

Minor: Same concern as in 3/7 about the multi-line error message
making scripted handling of error message collection more difficult.

> +   strbuf_release(&errmsg);
> +
> +   if (!force)
> +   check_clean_worktree(wt, av[0]);
> +
> +   ret |= delete_git_work_tree(wt);
> +   /*
> +* continue on even if ret is non-zero, there's no going back
> +* from here.
> +*/
> +   ret |= delete_git_dir(wt);
> +
> +   free_worktrees(worktrees);
> +   return ret;
> +}
> diff --git a/t/t2028-worktree-move.sh b/t/t2028-worktree-move.sh
> @@ -90,4 +90,30 @@ test_expect_success 'move main worktree' '
> +test_expect_success 'remove locked worktree' '
> +   git worktree lock destination &&
> +   test_must_fail git worktree remove destination &&
> +   git worktree unlock destination
> +'

Perhaps place 'unlock' in test_when_finished()[1].

> +test_expect_success 'remove worktree with dirty tracked file' '
> +   echo dirty >>destination/init.t &&
> +   test_must_fail git worktree remove destination
> +'
> +
> +test_expect_success 'remove worktree with untracked file' '
> +