Re: [PATCH v7 03/11] ref-filter: add option to pad atoms to the right

>
> In an earlier review, Matthieu pointed out that this test failed to
> ensure that the 'padright' value did not leak into the next atom. In a
> subsequent version, you fixed the test to check that condition, but
> now you've somewhat lost it again, at least visually. That is, because
> whitespace is "invisible" and because 'padright' now also affects
> literal strings, someone reading this test can't tell if those
> trailing |'s in the expected output are padded or not. You could use a
> different format string to prove that the 'padright' value doesn't
> leak. For instance:
>
>     %(padright:10)%(refname:short)|%(padright:5)|%(refname:short)
>
> This way, as long as the two |'s are side-by-side, then you've proved
> that the first one wasn't affected by the preceding 'padright:10'. You
> could also add back the %(refname:short) at the front of the pattern,
> as you currently have it, if you want to prove that padding is not
> enabled at the start of format.