Re: [PATCH v5 5/7] builtin/tag: add --format argument for tag -v

From: Jeff King <peff@peff.net>
Date: Jan 17, 2017, 17:32 UTC
Message-ID: <20170117173239.ir6jxbz46vwwzoht@sigill.intra.peff.net>
In-Reply-To: <20170117170018.nqk4yy5rrpomxr32@LykOS.localdomain>

On Tue, Jan 17, 2017 at 12:00:19PM -0500, Santiago Torres wrote:

Show 17 quoted lines

> > >  {
> > > -	return verify_and_format_tag(sha1, name, NULL, GPG_VERIFY_VERBOSE);
> > > +	int flags;
> > > +	char *fmt_pretty = cb_data;
> > > +	flags = GPG_VERIFY_VERBOSE;
> > > +
> > > +	if (fmt_pretty)
> > > +		flags = GPG_VERIFY_QUIET;
> > > +
> > > +	return verify_and_format_tag(sha1, ref, fmt_pretty, flags);
> > 
> > It seems funny that VERBOSE and QUIET are bit-flags. What happens when
> > you ask for GPG_VERIFY_VERBOSE|GPG_VERIFY_QUIET?
> 
> If I'm not mistaken, the way the code works right now this is not
> possible (GPG_VERIFY_VERBOSE will be unset when GPG_VERIFY_QUIET). I
> would have to re-read the patch to make sure this is the case then.

No, you're right that the two flags are mutually exclusive in the current callers. So I don't think there's a bug here. It's just that the interface is potentially awkward/confusing, which may be a problem down the line.

VERBOSE|QUIET _does_ have a meaning, which is "show the payload, but do not print the signature buffer". Perhaps just renaming QUIET to OMIT_STATUS or something would make it more clear.

-Peff