Re: (PATCH) Adding CORRESPONDING (NULL error) - Mailing list pgsql-hackers

From Kerem Kat
Subject Re: (PATCH) Adding CORRESPONDING (NULL error)
Date
Msg-id CAJZSWkVJ2Rr9BFDj77EHUOo9CYpDj=rJ7dtrF5i7V+g7q2CwEw@mail.gmail.com
Whole thread Raw
In response to Re: (PATCH) Adding CORRESPONDING (NULL error)  ("Erik Rijkers" <er@xs4all.nl>)
Responses Re: (PATCH) Adding CORRESPONDING (NULL error)  (Tom Lane <tgl@sss.pgh.pa.us>)
List pgsql-hackers
Hi,

Union with NULL error persists without the corresponding patch. Here
is the output from postgres without the patch:

SELECT a FROM (SELECT 1 a) foo
UNION
SELECT a FROM (SELECT NULL a) foo2;

ERROR:  failed to find conversion function from unknown to integer


It is thrown from parse_coerce.c:coerce_type method. I will try to dig
deep on it.


Regards,

Kerem KAT


On Thu, Oct 27, 2011 at 15:45, Erik Rijkers <er@xs4all.nl> wrote:
> (pgsql 9.2devel (25 oct) with your latest CORRESPONDING patch;
> linux x86_64 GNU/Linux 2.6.18-274.3.1.el5)
>
> Hi,
>
> here is another peculiarity, which I think is a bug:
>
> -- first without CORRESPONDING:
>
> $ psql -Xaf null.sql
>            select 1 a   , 2 b
> union all
>            select null a, 4 b ;
>  a | b
> ---+---
>  1 | 2
>   | 4
> (2 rows)
>
> -- then with CORRESPONDING:
>
>            select 1 a   , 2 b
> union all
> corresponding
>            select null a, 4 b ;
> psql:null.sql:9: ERROR:  failed to find conversion function from unknown to integer
>
>
> If the null value is in a table column the error does not occur:
>
> drop table if exists t1; create table t1 (a int, b int); insert into t1 values (1,2);
> drop table if exists t2; create table t2 (a int, b int); insert into t2 values (null,2);
>                select a,b from t1
> union all
> corresponding
>                select a,b from t2 ;
>  a | b
> ---+---
>  1 | 2
>   | 2
> (2 rows)
>
>
> I'm not sure it is actually a bug; but it seems an unneccessary error.
>
>
> thanks,
>
> Erik Rijkers
>
>


pgsql-hackers by date:

Previous
From: "Kevin Grittner"
Date:
Subject: Re: out-of-order caution
Next
From: Florian Pflug
Date:
Subject: Re: Hot Backup with rsync fails at pg_clog if under load