On Wed, Aug 21, 2013 at 5:22 PM, Junio C Hamano <gitster@xxxxxxxxx> wrote:
> Brandon Casey <drafnel@xxxxxxxxx> writes:
>
>> On Wed, Aug 21, 2013 at 2:47 PM, Junio C Hamano <gitster@xxxxxxxxx> wrote:
>>> # on load...
>>> printf -v __git_printf_supports_v -- "%s" yes >/dev/null 2>&1
>>>
>>> ...
>>>
>>> if test "${__git_printf_supports_v}" = yes
>>> then
>>> printf -v gitstring -- "$printf_format" "$gitstring"
>>> else
>>> gitstring=$(printf -- "$printf_format" "$gitstring")
>>> fi
>>
>> Yes, that appears to work.
>
> A real patch needs to be a bit more careful, though. The variable
> needs to be cleared before all of the above,
Agreed.
> and the testing would
> want to consider that the variable may not be set (i.e. use
> "${var-}" when checking).
Why is "${var-}" necessary? Wouldn't that be equivalent to "${var}"
or "$var"? We obviously wouldn't want to do 'if test $var = yes', but
I would have thought it was sufficient to wrap the variable
dereference in quotes as your original did.
-Brandon
--
To unsubscribe from this list: send the line "unsubscribe git" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html