Re: [PATCH v2] submodule: port subcommand 'set-branch' from shell to C

[Eric Sunshine <sunshine@sunshineco.com>]

On Tue, May 19, 2020 at 2:27 PM Shourya Shukla
<shouryashukla.oo@gmail.com> wrote:
> Convert submodule subcommand 'set-branch' to a builtin. Port 'set-branch'
> to 'submodule--helper.c' and call the latter via 'git-submodule.sh'.

You can reduce the redundancy by writing this as:

    Convert git-submodule subcommand 'set-branch' to a builtin and
    call it via 'git-submodule.sh'.

> Signed-off-by: Shourya Shukla <shouryashukla.oo@gmail.com>
> ---
> diff --git a/builtin/submodule--helper.c b/builtin/submodule--helper.c
> @@ -2284,6 +2284,46 @@ static int module_set_url(int argc, const char **argv, const char *prefix)
> +static int module_set_branch(int argc, const char **argv, const char *prefix)
> +{
> +       int quiet = 0, opt_default = 0;
> +       char *opt_branch = NULL;

This can be 'const const *', can't it?

> +       struct option options[] = {
> +               OPT__QUIET(&quiet,
> +                       N_("suppress output for setting default tracking branch of a submodule")),

This is unusually verbose for a _short_ description of the option.
Other commands use simpler descriptions. Perhaps take a hint from the
git-submodule man page:

    N("only print error messages")),

However, the bigger question is: Why is the --quiet option even here?
None of the code in this function ever consults the 'quiet' variable,
so its presence seems pointless.

Looking at the git-submodule documentation, I see that it is already
documented as accepted --quiet, so it may make some sense for you to
accept the option here. However, it might be a good idea either to
have an in-code comment or a blurb in the commit message explaining
that this C rewrite accepts the option for backward-compatibility (and
for future extension), not because it is actually used presently.

> +               OPT_BOOL(0, "default", &opt_default,
> +                       N_("set the default tracking branch to master")),

We can make this and the next short description more precise and
concise like this:

    N_("reset the default tracking branch to master")),

> +               OPT_STRING(0, "branch", &opt_branch, N_("branch"),
> +                       N_("set the default tracking branch to the one specified")),

Then:

    N_("set the default tracking branch")),

> +               OPT_END()
> +       };
> +       const char *const usage[] = {
> +               N_("git submodule--helper set-branch [--quiet] (-d|--default) <path>"),
> +               N_("git submodule--helper set-branch [--quiet] (-b|--branch) <branch> <path>"),
> +               NULL
> +       };
> +
> +       argc = parse_options(argc, argv, prefix, options, usage, 0);
> +
> +       if (!opt_branch && !opt_default)
> +               die(_("at least one of --branch and --default required"));

This wording makes no sense considering that --branch and --default
are mutually exclusive. By writing "at least one of", you're saying
that you can use _more than one_, which is clearly incorrect. Reword
it like this:

    die(_("--branch or --default required"));

> +       if (opt_branch && opt_default)
> +               die(_("--branch and --default do not make sense together"));

A more precise way to say this is:

    die(_("--branch and --default are mutually exclusive"));

> +       if (argc != 1 || !(path = argv[0]))
> +               usage_with_options(usage, options);
> +
> +       config_name = xstrfmt("submodule.%s.branch", path);
> +       config_set_in_gitmodules_file_gently(config_name, opt_branch);

Tracing through the config code, I see that
config_set_in_gitmodules_file_gently() removes the key if 'opt_branch'
is NULL, which mirrors the behavior of the shell code this is
replacing. Good.