Re: [PATCH 03/10] run-command API: add and use a run_command_l_opt()

From: Junio C Hamano <gitster@pobox.com>
Date: Oct 14, 2022, 20:27 UTC
Message-ID: <xmqq8rli881x.fsf@gitster.g>
In-Reply-To: <Y0nAm4+KL71vyYB8@coredump.intra.peff.net>

Jeff King <peff@peff.net> writes:

Show 14 quoted lines

> On Fri, Oct 14, 2022 at 11:40:27AM -0700, Junio C Hamano wrote:
>
>> > @@ -862,11 +858,11 @@ static void write_refspec_config(const char *src_ref_prefix,
>> >  
>> >  static void dissociate_from_references(void)
>> >  {
>> > -	static const char* argv[] = { "repack", "-a", "-d", NULL };
>> 
>> Good to see that this one in a wrong scope can now go away.
>
> It definitely is broader scope than is necessary. I wonder if it makes
> things easier to read, though, the way we would sometimes hoist things
> out of a function into a static-global to make them more obvious. I
> dunno.

Didn't think about it, but it is a worthy point to make. Unlike the call to l_opt() buried inside a conditional, it makes it stand out that what we see upfront is one of the commands the function will run. It is especially valuable when a function is almost exclusively about running a single command, but then we will have a single call to l_opt() without much preparations or clean-ups around it, so the visual noise that detracts our eyes away from the actual commands may not be all that bad, though.

Show 10 quoted lines

>> >  	char *alternates = git_pathdup("objects/info/alternates");
>> >  
>> >  	if (!access(alternates, F_OK)) {
>> > -		if (run_command_v_opt(argv, RUN_GIT_CMD|RUN_COMMAND_NO_STDIN))
>> > +		if (run_command_l_opt(RUN_GIT_CMD|RUN_COMMAND_NO_STDIN,
>> > +				      "repack",  "-a", "-d", NULL))
>> >  			die(_("cannot repack to clean up"));
>
> I just happened to notice in this one there is a weird extra space
> before "-a".

Yeah, good eyes.

Thanks.