Date: Tue, 6 Feb 2024 19:19:37 +0900
From: Yugo NAGATA <nagata@sraoss.co.jp>
To: jian he <jian.universality@gmail.com>
Cc: torikoshia <torikoshia@oss.nttdata.com>, "David G. Johnston"
 <david.g.johnston@gmail.com>, PostgreSQL Hackers
 <pgsql-hackers@lists.postgresql.org>
Subject: Re: Change COPY ... ON_ERROR ignore to ON_ERROR ignore_row
Message-Id: <20240206191937.72eaf0ccc20cfea37944b422@sraoss.co.jp>
In-Reply-To: <CACJufxGnc+=No=Ua6NFT2ADt0vRL=m1QsuCOM=9aKPKWh9_L6Q@mail.gmail.com>
References: <CAKFQuwawy1e6YR4S=j+y7pXqg_Dw1WBVrgvf=BP3d1_aSfe_+Q@mail.gmail.com>
	<20240129172858.ccb6c77c3be95a295e2b2b44@sraoss.co.jp>
	<CACJufxFFU92-H7G8tmpxt9oTSfL062OA7n5rPx-YbOAtDUUzGw@mail.gmail.com>
	<04bf425ad1b15a4daefe96c478a5253b@oss.nttdata.com>
	<CACJufxGnc+=No=Ua6NFT2ADt0vRL=m1QsuCOM=9aKPKWh9_L6Q@mail.gmail.com>
Mime-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
Archived-At: <https://www.postgresql.org/message-id/20240206191937.72eaf0ccc20cfea37944b422%40sraoss.co.jp>
Precedence: bulk

On Mon, 5 Feb 2024 14:26:46 +0800
jian he <jian.universality@gmail.com> wrote:

> On Mon, Feb 5, 2024 at 10:29 AM torikoshia <torikoshia@oss.nttdata.com> wrote:
> >
> > Hi,
> >
> > On 2024-02-03 15:22, jian he wrote:
> > > The idea of on_error is to tolerate errors, I think.
> > > if a column has a not null constraint, let it cannot be used with
> > > (on_error 'null')
> >
> > > +       /*
> > > +        * we can specify on_error 'null', but it can only apply to
> > > columns
> > > +        * don't have not null constraint.
> > > +       */
> > > +       if (att->attnotnull && cstate->opts.on_error ==
> > > COPY_ON_ERROR_NULL)
> > > +           ereport(ERROR,
> > > +                   (errcode(ERRCODE_BAD_COPY_FILE_FORMAT),
> > > +                    errmsg("copy on_error 'null' cannot be used with
> > > not null constraint column")));
> >
> > This means we cannot use ON_ERROR 'null' even when there is one column
> > which have NOT NULL constraint, i.e. primary key, right?
> > IMHO this is strong constraint and will decrease the opportunity to use
> > this feature.
> 
> I don't want to fail in the middle of bulk inserts,
> so I thought immediately erroring out would be a great idea.
> Let's see what other people think.

I also think this restriction is too strong because it is very
common that a table has a primary key, unless there is some way
to specify columns that can be set to NULL. Even when ON_ERROR
is specified, any constraint violation errors cannot be generally
ignored, so we cannot elimiate the posibility COPY FROM fails in
the middle due to invalid data, anyway.

Regards,
Yugo Nagata

-- 
Yugo NAGATA <nagata@sraoss.co.jp>