From: Karthik Nayak <karthik.188@gmail.com>
Date: Thu, 10 Sep 2015 18:04:42 GMT
Subject: Re: [PATCH v17 13/14] tag.c: implement '--format' option
Message-ID: <CAOLa=ZQeFSZtLHK_sZMTHpdw_UTL_yk=Nrmu-4vC0vy2_hR7mw@mail.gmail.com>
In-Reply-To: <xmqq8u8e40l2.fsf@gitster.mtv.corp.google.com>

On Thu, Sep 10, 2015 at 11:29 PM, Junio C Hamano <gitster@pobox.com> wrote:
> Karthik Nayak <karthik.188@gmail.com> writes:
>
>> -static int list_tags(struct ref_filter *filter, struct ref_sorting *sorting)
>> +static int list_tags(struct ref_filter *filter, struct ref_sorting *sorting, const char *format)
>>  {
>>       struct ref_array array;
>> -     char *format, *to_free = NULL;
>> +     char *to_free = NULL;
>>       int i;
>
> format is const char * while to_free is non-const char * here.
>
>> @@ -41,12 +41,14 @@ static int list_tags(struct ref_filter *filter, struct ref_sorting *sorting)
>>       if (filter->lines == -1)
>>               filter->lines = 0;
>>
>> +     if (!format) {
>> +             if (filter->lines) {
>> +                     format = xstrfmt("%s %%(contents:lines=%d)",
>> +                                      "%(align:15)%%(refname:short)%%(end)", filter->lines);
>
> Hmmm, did this even pass tests and if so how?  What are these double
> %% doing before refname and end?  Perhaps we do not have enough test
> coverage?
>

Seems like I didn't run the tests here, will change this.

>> +                     to_free = format;
>
> This assignment discards constness.
>
> Take the result of xstrfmt() to to_free (which is a non-const
> pointer) and then assigning it to format (which is a const pointer)
> would work it around.

Yeah!

-- 
Regards,
Karthik Nayak