Ralf Thielow <ralf.thie...@gmail.com> writes:

> If option --help is passed to a Git command, we try to open
> the man page of that command.  However, we do it even for commands
> we don't know.  Make sure it is a Git command.

What the patch does is correct, I think, but the explanation may
invite a false alarm.  If you added a custom command git-who in your
$PATH, with an appropriate documentation for git-who(1), we would
still show its documentation, no?

The same comment applies to 1/2, too, in that the word "command"
will be interpreted differently by different people.  For example,
"git co --help" and "git help co" would work, with or without 1/2 in
place when you have "[alias] co = checkout", so we are calling "Git
subcommands that we ship, custom commands 'git-$foo' the users have
in their $PATH, and aliases the users create" collectively "command".

As long as the reader understands that definition, both the log
messages of 1/2 and 2/2 _and_ the updated description for "git help"
we have in 1/2 are all very clear.  I do not care too much about the
commit log message, but we may want to think about the documentation
a bit more.

Here is what 1/2 adds to "git help" documentation:

    +Note that `git --help ...` is almost identical to `git help ...` because
    +the former is internally converted into the latter with option 
    +being added.

     To display the linkgit:git[1] man page, use `git help git`.

    @@ -43,6 +44,10 @@ OPTIONS
            Prints all the available commands on the standard output. This
            option overrides any given command or guide name.

    +   Display help information only for commands.

First, I do not think a short form is unnecessary; the users are not
expected to use that form, once they started typing "git help...".
If we flip the polarity and call it --exclude-guides or something,
would it make it less ambiguous?

> This breaks "git <concept> --help" while "git help <concept>" still works.

I wouldn't call that a breakage; "git everyday --help" shouldn't
have worked in the first place.  It did something useful merely by
accident ;-).

> diff --git a/git.c b/git.c
> index 0f1937f..2cd2e06 100644
> --- a/git.c
> +++ b/git.c
> @@ -528,10 +528,23 @@ static void handle_builtin(int argc, const char **argv)
>       strip_extension(argv);
>       cmd = argv[0];
> -     /* Turn "git cmd --help" into "git help cmd" */
> +     /* Turn "git cmd --help" into "git help --command-only cmd" */
>       if (argc > 1 && !strcmp(argv[1], "--help")) {
> +             struct argv_array args;
> +             int i;
> +
>               argv[1] = argv[0];
>               argv[0] = cmd = "help";
> +
> +             argv_array_init(&args);
> +             for (i = 0; i < argc; i++) {
> +                     argv_array_push(&args, argv[i]);
> +                     if (!i)
> +                             argv_array_push(&args, "--command-only");
> +             }
> +
> +             argc++;
> +             argv = argv_array_detach(&args);
>       }
>       builtin = get_builtin(cmd);

The code does this after it:

    if (builtin)

and returns.  If we didn't get builtin, we risk leaking args.argv
here, but we assume argv[0] = cmd = "help" is always a builtin,
which I think is a safe assumption, so the code is OK.  Static
checkers that are only half intelligent may yell at you for not
releasing the resources, though.  I wonder if it is worth doing

    static void handle_builtin(int argc, const char **argv)
            struct argv_array args = ARGV_ARRAY_INIT;
            if (argc > 1 && !strcmp(argv[1], "--help")) {
                    argv = args.argv;
            builtin = get_builtin(cmd);
            if (builtin)

to help unconfuse them.

By the way, I do not see these patches on gmane, public-inbox or
usual suspects.  Perhaps vger is having a bad day or something?

To unsubscribe from this list: send the line "unsubscribe git" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Reply via email to