Re: [PATCH v2 18/25] t5500, t5539: tests for shallow depth since a specific date

From: Duy Nguyen <pclouds@gmail.com>
Date: Feb 15, 2016, 07:17 UTC
Message-ID: <CACsJy8DE063TDsS3NGOTf5=tYpNTusX+ymDT8XG_RVjbqCW=Fw@mail.gmail.com>
In-Reply-To: <xmqqio1y7v5n.fsf@gitster.mtv.corp.google.com>

On Tue, Feb 9, 2016 at 4:24 AM, Junio C Hamano <gitster@pobox.com> wrote:

Show 10 quoted lines

> Nguyễn Thái Ngọc Duy  <pclouds@gmail.com> writes:
>
>> +     cd shallow-since &&
>> +     GIT_COMMITTER_DATE="100000000 +0700" git commit --allow-empty -m one &&
>> +     GIT_COMMITTER_DATE="200000000 +0700" git commit --allow-empty -m two &&
>> +     GIT_COMMITTER_DATE="300000000 +0700" git commit --allow-empty -m three &&
>
> Didn't --date="@10000000 +0700" work?
>
> Not a complaint but genuinely curious.

Nope. --date sets author date, but rev-list --since only cares about committer date.

>> +     git clone --shallow-since "300000000 +0700" "file://$(pwd)/." ../shallow11 &&
>
> Are we required to add TZ to --shallow-since, or we merely tolerate
> if there is one (I am hoping that it is the latter)?

Hmm.. I'm not familiar with date parsing code. I was sure that the parsing here (to timestamp) is the same as "rev-list --since" done on the server side. The function used is approxidate() which seems to ignore timezone.

Show 13 quoted lines

>> +     git -C ../shallow11 log --pretty=tformat:%s HEAD >actual &&
>> +     echo three >expected &&
>> +     test_cmp expected actual
>> +     )
>> +'
>> +
>> +test_expect_success 'fetch shallow since ...' '
>> +     git -C shallow11 fetch --shallow-since "200000000 +0700" origin &&
>> +     git -C shallow11 log --pretty=tformat:%s origin/master >actual &&
>> +     echo three >expected &&
>> +     echo two  >>expected &&
>
> test_write_lines perhaps?

I was lazy. probably could go with cat <<EOF.

-- 
Duy