Re: [PATCH 2/3] revision.c: use skip_prefix() in handle_revision_opt()

From: Jeff King <peff@peff.net>
Date: Jun 2, 2017, 20:11 UTC
Message-ID: <20170602201143.6avkhp224no3dukb@sigill.intra.peff.net>
In-Reply-To: <20170602191010.16834-3-szeder.dev@gmail.com>

On Fri, Jun 02, 2017 at 09:10:09PM +0200, SZEDER Gábor wrote:

Show 18 quoted lines

> @@ -1785,15 +1785,15 @@ static int handle_revision_opt(struct rev_info *revs, int argc, const char **arg
>  	} else if (!strcmp(arg, "--author-date-order")) {
>  		revs->sort_order = REV_SORT_BY_AUTHOR_DATE;
>  		revs->topo_order = 1;
> -	} else if (starts_with(arg, "--early-output")) {
> +	} else if (skip_prefix(arg, "--early-output", &optarg)) {
>  		int count = 100;
> -		switch (arg[14]) {
> +		switch (*optarg) {
>  		case '=':
> -			count = atoi(arg+15);
> +			count = atoi(optarg + 1);
>  			/* Fallthrough */
>  		case 0:
>  			revs->topo_order = 1;
> -		       revs->early_output = count;
> +			revs->early_output = count;
>  		}

What happens if I say "--early-output-foobar"? There should probably be a "default" here that rejects it. Though we'd probably to goto to get to the unknown block, yuck.

Perhaps we could do:

  if (skip_prefix(arg, "--early-output", &optarg) &&
      (*optarg == '=' || !*optarg)) {
          int count = *optarg ? atoi(optarg + 1) : 100;
	  revs->topo_order = 1;
	  revs->early_output = count;
  }

Alternatively, a helper like:

  int match_opt(const char *have, const char *want, const char **argout)
  {
	const char *arg;
	if (!skip_prefix(have, want, &arg))
		return 0;
	if (!*arg)
		*argout = NULL;
	else if (*arg == '=')
		*argout = arg + 1;
	else
		return 0;
	return 1;
  }

would let us do:

  if (match_opt(arg, "--early-output"), &optarg)) {
	int count = optarg ? atoi(optarg) : 100;
	...
  }

which is a little nicer and could maybe help other options (I didn't see any, though). If we're going to go that route, though, I suspect there may be some helpers we already have. Looks like parse_long_opt() is almost there, but doesn't handle options. I wonder if we could reuse bits of parse-options here (or even better, just parse-optify many of these).

Anyway, none of that is caused by your patch, but at least doing the minimal fix (my first hunk) seems like it fits into your series.

-Peff