MIME-Version: 1.0
References: 
 <CADkLM=fv1JfY4Ufa-jcwwNbjQixNViskQ8jZu3Tz_p656i_4hQ@mail.gmail.com>
 <CADkLM=cZ7N+f4Bj-8MiccFcTASjBB2r1_s3BD9OFbbQOwK01cg@mail.gmail.com>
 <CACJufxE7yoH42juViFvYuCQMPwXMWxDVwiugs-BW+N4oUXB=1w@mail.gmail.com>
 <CADkLM=dFZ8-mBJDkiZZ4JadnJy8zLExUoo_b1FOs2Ozrtge=gQ@mail.gmail.com>
 <CACJufxHynx40aHHUCR4wZh1qnL=X0yw6baW4Z-+vkUgT0OhjOg@mail.gmail.com>
 <CADkLM=eCPiVhGKcbNW5bUEYpozUaKNXjd_UmBVR256p2zWtYQQ@mail.gmail.com>
 <CACJufxH4LrCpL63SRYO3zVk46YdD4--VYQoBL7GmHmCm=NCAJQ@mail.gmail.com>
 <CADkLM=cmv_bmxBe8KmZd6rEgiqSdoDfHnJa63u7rdRuAsqOwDA@mail.gmail.com>
 <CACJufxGbw9iNT8QVm4QD9cPFKnDnvDBQp7AGxkoqDa-JjzVXmg@mail.gmail.com>
 <CACJufxFkLLuX1VJ-J3fppCr37PHtxkvwyd_e4zNd+VYK0v0gnQ@mail.gmail.com>
 <5ae9578e-f25e-49c5-97ab-ad27bc2050b5@eisentraut.org>
 <CACJufxHx-UfprE6P4_ZB_cOYktHd4pLMNx=jWJFOGGGFj2YZWQ@mail.gmail.com>
 <CACJufxH_1EtEBMb0JvxaM3Gmnt33HYrS37m5eYjJ_OfBkMVFJg@mail.gmail.com>
 <CACJufxHGox47X4zNtVeNw7H=SZ7ATAAD5SEYpvxa2o5mPV0XCA@mail.gmail.com>
 <5221f892-e1d5-4943-a3dc-5c9c7bdf4924@eisentraut.org>
 <CAAJ_b959dfJsorK+_vPdVa8F3QYPdOnb-HzJY8oQQR91N9Tq2A@mail.gmail.com>
 <CACJufxFOQkmH9KeQNTD0qASbL3i03cLeoyVZdcZ8Fm0rk6eJ5Q@mail.gmail.com>
 <CACJufxEUr33CoQMsZdEehNU7wj_QOiyk8VW=W+aB03PF2J_qKQ@mail.gmail.com>
 <CACJufxGx_XDHq1btNj=ZrBjwd0skeWFJQLMGqSR-pHw-kuYRnQ@mail.gmail.com>
 <CADkLM=cXsprF2GvcmqOG+ME_N2iyHNr_--xW-6ZaKd32yF_NzA@mail.gmail.com>
 <CACJufxGRSV3j6g7feEkfu7WuLx7ocMte8mvVEwNh7XBfnpJd-A@mail.gmail.com>
In-Reply-To: 
 <CACJufxGRSV3j6g7feEkfu7WuLx7ocMte8mvVEwNh7XBfnpJd-A@mail.gmail.com>
From: Corey Huinker <corey.huinker@gmail.com>
Date: Wed, 1 Apr 2026 16:07:49 -0400
Message-ID: 
 <CADkLM=c_QQkpBhm0HfGm5C8gOFNKW6g5E=W5Jp9mjayUW3yXYQ@mail.gmail.com>
Subject: Re: CAST(... ON DEFAULT) - WIP build on top of Error-Safe User
 Functions
To: jian he <jian.universality@gmail.com>
Cc: Amul Sul <sulamul@gmail.com>, Peter Eisentraut <peter@eisentraut.org>,
	Kirill Reshke <reshkekirill@gmail.com>,
 Vik Fearing <vik@postgresfriends.org>,
	Isaac Morland <isaac.morland@gmail.com>, pgsql-hackers@lists.postgresql.org
Content-Type: multipart/alternative; boundary="0000000000001a232c064e6ba491"
Archived-At: 
 <https://www.postgresql.org/message-id/CADkLM%3Dc_QQkpBhm0HfGm5C8gOFNKW6g5E%3DW5Jp9mjayUW3yXYQ%40mail.gmail.com>
Precedence: bulk

--0000000000001a232c064e6ba491
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

On Wed, Apr 1, 2026 at 3:23=E2=80=AFAM jian he <jian.universality@gmail.com=
> wrote:

> One more issue I found:
>
> https://git.postgresql.org/cgit/postgresql.git/commit/?id=3D74c96699be3f5=
3c058c8794c07952221b9625b4f
> SELECT JSON_VALUE(jsonb '1234', '$' RETURNING char(2)  DEFAULT '011' ON
> ERROR);
> ERROR:  value too long for type character(2)
>
> Similarly, we can
> SELECT CAST(text '1234' as char(2) DEFAULT '111111' ON conversion ERROR);
> ERROR:  value too long for type character(2)
>

Both of these are doomed when we hit the default, so I don't see why we
should make any accommodation for them.


> Composite types respect typmod, for example:
> CREATE TYPE comp AS (a char(3), b int);
> SELECT CAST('(14,42)' AS comp DEFAULT '(1234,2)' ON CONVERSION ERROR); --
> error
>

This one is trickier, but I'd be willing to reject composite inputs in v19
and solve it in v20.


> The regression tests are too large; we can order them by the cast
> source type's pg_type.type category,
> so we won't miss any tests.
>

We can always remove redundant tests later.

--0000000000001a232c064e6ba491
Content-Type: text/html; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr"><div class=3D"gmail_quote gmail_quote_container"><div dir=
=3D"ltr" class=3D"gmail_attr">On Wed, Apr 1, 2026 at 3:23=E2=80=AFAM jian h=
e &lt;<a href=3D"mailto:jian.universality@gmail.com">jian.universality@gmai=
l.com</a>&gt; wrote:<br></div><blockquote class=3D"gmail_quote" style=3D"ma=
rgin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:=
1ex">One more issue I found:<br>
<a href=3D"https://git.postgresql.org/cgit/postgresql.git/commit/?id=3D74c9=
6699be3f53c058c8794c07952221b9625b4f" rel=3D"noreferrer" target=3D"_blank">=
https://git.postgresql.org/cgit/postgresql.git/commit/?id=3D74c96699be3f53c=
058c8794c07952221b9625b4f</a><br>
SELECT JSON_VALUE(jsonb &#39;1234&#39;, &#39;$&#39; RETURNING char(2)=C2=A0=
 DEFAULT &#39;011&#39; ON ERROR);<br>
ERROR:=C2=A0 value too long for type character(2)<br>
<br>
Similarly, we can<br>
SELECT CAST(text &#39;1234&#39; as char(2) DEFAULT &#39;111111&#39; ON conv=
ersion ERROR);<br>
ERROR:=C2=A0 value too long for type character(2)<br></blockquote><div><br>=
</div><div>Both of these are doomed when we hit the default, so I don&#39;t=
 see why we should make any accommodation for them.</div><div><br></div><di=
v>=C2=A0</div><blockquote class=3D"gmail_quote" style=3D"margin:0px 0px 0px=
 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex">

Composite types respect typmod, for example:<br>
CREATE TYPE comp AS (a char(3), b int);<br>
SELECT CAST(&#39;(14,42)&#39; AS comp DEFAULT &#39;(1234,2)&#39; ON CONVERS=
ION ERROR); -- error<br></blockquote><div><br></div><div>This one is tricki=
er, but I&#39;d be willing to reject composite inputs in v19 and solve it i=
n v20.</div><div>=C2=A0</div><blockquote class=3D"gmail_quote" style=3D"mar=
gin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1=
ex">

The regression tests are too large; we can order them by the cast<br>
source type&#39;s pg_type.type category,<br>
so we won&#39;t miss any tests.<br></blockquote><div><br></div><div>We can =
always remove redundant tests later.</div></div></div>

--0000000000001a232c064e6ba491--