On Sun, 2011-01-30 at 17:14 -0500, Tom Lane wrote:
> Jeff Davis <pgsql(at)j-davis(dot)com> writes:
> > On Sun, 2011-01-30 at 02:55 +0000, Thom Brown wrote:
> >> postgres=# select '[18,20]'::numrange @> 19;
> >> ERROR:  operator does not exist: numrange @> integer
> >> LINE 1: select '[18,20]'::numrange @> 19;
> >> ^
> >> HINT:  No operator matches the given name and argument type(s). You
> >> might need to add explicit type casts.
> 
> > It's because it doesn't know the type on the right side, and assumes
> > it's an int4.
> 
> Well, yeah, it is an int4.  The question ought to be phrased "why does
> the parser fail to promote the int4 to numeric?".  There might be some
> excuse for an "operator is not unique" here, but I don't understand the
> above failure --- it should be able to use an implicit coercion from
> int4 to numeric.

The problem exists for arrays, as well, so I think this is just a
limitation of the type system.

   Regards,
	Jeff Davis

postgres=# select ARRAY[1.4,1.5,1.6]::numeric[] || 5.0;
     ?column?      
-------------------
 {1.4,1.5,1.6,5.0}
(1 row)

postgres=# select ARRAY[1.4,1.5,1.6]::numeric[] || 5;
ERROR:  operator does not exist: numeric[] || integer
LINE 1: select ARRAY[1.4,1.5,1.6]::numeric[] || 5;
                                             ^
HINT:  No operator matches the given name and argument type(s). You
might need to add explicit type casts.

In response to

• Re: WIP: RangeTypes at 2011-01-30 22:14:50 from Tom Lane

pgsql-hackers by date

Next:	From: Kevin Grittner	Date: 2011-02-02 17:58:39
	Subject: Re: SSI patch version 14
Previous:	From: Heikki Linnakangas	Date: 2011-02-02 17:41:35
	Subject: Re: SSI patch version 14