Re: [PATCH v2 5/6] bisect--helper: `is_expected_rev` & `check_expected_revs` shell function in C
- From
Pranit Bauva <pranit.bauva@gmail.com>
- Date
- Jun 17, 2016, 12:49 UTC
- Message-ID
- <CAFZEwPPUX16omOgk4V87UYheJH_+kBueGd4OzAJsit=j2Q6mNg@mail.gmail.com>
- In-Reply-To
- <CAP8UFD0YNNzF4qqdvDbrONaMm8B8G_amePq2g58TYrjm+9W6Vw@mail.gmail.com>
Hey Christian,
On Fri, Jun 17, 2016 at 2:17 AM, Christian Couder <christian.couder@gmail.com> wrote:
Show 88 quoted lines
> On Thu, Jun 16, 2016 at 9:25 PM, Pranit Bauva <pranit.bauva@gmail.com> wrote:
>> Hey Eric,
>>
>> On Fri, Jun 17, 2016 at 12:46 AM, Eric Sunshine <sunshine@sunshineco.com> wrote:
>>> On Thu, Jun 16, 2016 at 3:05 PM, Pranit Bauva <pranit.bauva@gmail.com> wrote:
>>>> On Thu, Jun 16, 2016 at 2:44 AM, Eric Sunshine <sunshine@sunshineco.com> wrote:
>>>>> On Wed, Jun 15, 2016 at 10:00 AM, Pranit Bauva <pranit.bauva@gmail.com> wrote:
>>>>>> Reimplement `is_expected_rev` & `check_expected_revs` shell function in
>>>>>> C and add a `--check-expected-revs` subcommand to `git bisect--helper` to
>>>>>> call it from git-bisect.sh .
>>>>>> [...]
>>>>>> Signed-off-by: Pranit Bauva <pranit.bauva@gmail.com>
>>>>>> ---
>>>>>> diff --git a/builtin/bisect--helper.c b/builtin/bisect--helper.c
>>>>>> @@ -162,13 +162,44 @@ static int bisect_reset(const char *commit)
>>>>>> +static int is_expected_rev(const char *expected_hex)
>>>>>> +{
>>>>>> + struct strbuf actual_hex = STRBUF_INIT;
>>>>>> + int res;
>>>>>> +
>>>>>> + if (strbuf_read_file(&actual_hex, git_path_bisect_expected_rev(), 0) < 0) {
>>>>>> + strbuf_release(&actual_hex);
>>>>>> + return 0;
>>>>>> + }
>>>>>> +
>>>>>> + strbuf_trim(&actual_hex);
>>>>>> + res = !strcmp(actual_hex.buf, expected_hex);
>>>>>> + strbuf_release(&actual_hex);
>>>>>> + return res;
>>>>>> +}
>>>>>
>>>>> Not worth a re-roll, but this could be re-structured to avoid having
>>>>> to remember to release the strbuf at all exits:
>>>>>
>>>>> struct strbuf actual_hex = ...;
>>>>> int res = 0;
>>>>>
>>>>> if (strbuf_read_file(...) >= 0) {
>>>>> strbuf_trim(...);
>>>>> res = !strcmp(...);
>>>>> }
>>>>> strbuf_release(...);
>>>>> return res;
>>>>>
>>>>> Alternately:
>>>>>
>>>>> if (strbuf_read_file(...) < 0)
>>>>> goto done;
>>>>>
>>>>> strbuf_trim(...);
>>>>> res = !strcmp(...);
>>>>>
>>>>> done:
>>>>> strbuf_release(...);
>>>>> return res;
>>>>>
>>>>> which is a bit less compact.
>>>>
>>>> I will avoid this for the reason that I will have to create a label
>>>> for a lot of functions. If I choose to do this for one function, I
>>>> think it would be more appropriate to do the same for other functions.
>>>> There would be a lot of functions in future which would be in the same
>>>> scenario and creating a separate label for each of them would be quite
>>>> tedious. What do you think?
>>>
>>> Not sure what you're talking about. Label names are not shared across
>>> functions. Anyhow, the first suggestion I presented above is more
>>> concise than the 'goto' version.
>>
>> Yes I am aware of the fact that labels aren't shared across functions.
>> What I meant by "separate label" was that I will have to make a label
>> "fail" in each function. But I recently noticed that its used quite a
>> lot so I think it would be okay to use it. Will re-roll with using
>> labels and goto.
>
> My opinion is that if there is a more concise version without labels
> and gotos, it's better to use it, so I would suggest Eric's first
> suggestion which is:
>
>> struct strbuf actual_hex = ...;
>> int res = 0;
>>
>> if (strbuf_read_file(...) >= 0) {
>> strbuf_trim(...);
>> res = !strcmp(...);
>> }
>> strbuf_release(...);
>> return res;Sure I could do that!
Regards, Pranit Bauva