Hi,

On Wed, 7 Nov 2018, Junio C Hamano wrote:

> Ævar Arnfjörð Bjarmason  <avarab@gmail.com> writes:
> 
> > diff --git a/builtin/range-diff.c b/builtin/range-diff.c
> > index f01a0be851..05d1f6b6b6 100644
> > --- a/builtin/range-diff.c
> > +++ b/builtin/range-diff.c
> > @@ -16,11 +16,14 @@ int cmd_range_diff(int argc, const char **argv, const char *prefix)
> >  	int creation_factor = RANGE_DIFF_CREATION_FACTOR_DEFAULT;
> >  	struct diff_options diffopt = { NULL };
> >  	int simple_color = -1;
> > +	int no_patch = 0;
> >  	struct option options[] = {
> >  		OPT_INTEGER(0, "creation-factor", &creation_factor,
> >  			    N_("Percentage by which creation is weighted")),
> >  		OPT_BOOL(0, "no-dual-color", &simple_color,
> >  			    N_("use simple diff colors")),
> > +		OPT_BOOL_F('s', "no-patch", &no_patch,
> > +			 N_("show patch output"), PARSE_OPT_NONEG),
> 
> As OPT_BOOL("patch") natively takes "--no-patch" to flip the bool
> off, an int variable "patch" that is initialized to 1 would make it
> more readable by avoiding double negation !no_patch like the one we
> see below.  I guess the reason behind the contortion you wanted to
> give the synonym --silent to it?

In light of my investigation that revealed that the original behavior
(which is still documented in the manual page of range-diff) was broken,
and I would much rather see that fixed than adding a workaround.

I would be fine with my patch being combined with the update to the manual
page and the regression test, as a v3.

Ciao,
Dscho

> 
> > @@ -92,7 +95,7 @@ int cmd_range_diff(int argc, const char **argv, const char *prefix)
> >  	}
> >  
> >  	res = show_range_diff(range1.buf, range2.buf, creation_factor,
> > -			      simple_color < 1, &diffopt);
> > +			      simple_color < 1, !no_patch, &diffopt);
> 
> >  	strbuf_release(&range1);
> >  	strbuf_release(&range2);
> 
> > @@ -7,6 +7,7 @@
> >  
> >  int show_range_diff(const char *range1, const char *range2,
> >  		    int creation_factor, int dual_color,
> > +		    int patch,
> >  		    struct diff_options *diffopt);
> 
> Other than that small "Huh?", the code looks good to me.
> 
> > diff --git a/t/t3206-range-diff.sh b/t/t3206-range-diff.sh
> > index 6aae364171..27e071650b 100755
> > --- a/t/t3206-range-diff.sh
> > +++ b/t/t3206-range-diff.sh
> > @@ -122,6 +122,26 @@ test_expect_success 'changed commit' '
> >  	test_cmp expected actual
> >  '
> >  
> > +test_expect_success 'changed commit -p & --patch' '
> > +	git range-diff --no-color -p topic...changed >actual &&
> > +	test_cmp expected actual &&
> > +	git range-diff --no-color --patch topic...changed >actual &&
> > +	test_cmp expected actual
> 
> This makes sure that -p and --patch produces the same output as the
> default case?  I am not sure who in the parseopt API groks the
> single letter "-p" in this case offhand.  Care to explain how?
> 
> The other side of the test to see -s/--no-patch we see below also
> makes sense.
> 
> > +test_expect_success 'changed commit -s & --no-patch' '
> > +...
> > +	cat >expected <<-EOF &&
> 
> Quote EOF to allow readers skim the contents without looking for and
> worrying about $substitutions in there, unless there are tons of
> offending code in the same script already in which case we should
> leave the clean-up outside this primary change.
> 
>