Re: [PATCH] Integer overflow in timestamp[tz]_part() and date/time boundaries check

14.03.2016 16:23, David Steele:
> On 2/25/16 4:44 PM, Vitaly Burovoy wrote:
>
>> Added to the commitfest 2016-03.
>>
>> [CF] https://commitfest.postgresql.org/9/540/
>
> This looks like a fairly straight-forward bug fix (the size of the
> patch is deceptive because there a lot of new tests included). It
> applies cleanly.
>
> Anastasia, I see you have signed up to review. Do you have an idea
> when you will get the chance to do that?
>
> Thanks,

I've read the patch thoroughly and haven't found any problems. I think
that the patch is in a very good shape.
It fixes a bug and has an excellent set of tests.

There is an issue, mentioned in the thread above:

>postgres=# select
>postgres-# to_char(date_trunc('week', '4713-01-01 BC'::date),'day')
>postgres-# ,to_char(date_trunc('week', '4714-12-29 BC'::date),'day')
>postgres-# ,to_char(date_trunc('week', '4714-12-28 BC'::date),'day');
> to_char | to_char | to_char
>-----------+-----------+-----------
> monday | monday | thursday
>(1 row)

>since 4714-12-28 BC and to the past detection when a week is starting
>is broken (because it is boundary of isoyears -4713 and -4712).
>Is it worth to break undocumented range or leave it as is?

But I suppose that behavior of undocumented dates is not essential.

--
Anastasia Lubennikova
Postgres Professional: http://www.postgrespro.com
The Russian Postgres Company