Re: [GENERAL] Surprising results from array concatenation - Mailing list pgsql-general

From Tom Lane
Subject Re: [GENERAL] Surprising results from array concatenation
Date
Msg-id 24303.1493142797@sss.pgh.pa.us
Whole thread Raw
In response to Re: [GENERAL] Surprising results from array concatenation  ("David G. Johnston" <david.g.johnston@gmail.com>)
Responses Re: [GENERAL] Surprising results from array concatenation  (Mike Blackwell <mike.blackwell@rrd.com>)
List pgsql-general
"David G. Johnston" <david.g.johnston@gmail.com> writes:
> On Tue, Apr 25, 2017 at 9:26 AM, Mike Blackwell <mike.blackwell@rrd.com>
> wrote:
>> The docs (section 9.18 for PG 9.6) show as an example for array
>> concatenation
>> ARRAY[4,5,6] || 7
>> which works fine.  However, trying the same with an array of text doesn't
>> work:
>> # select array['a','b','c'] || 'd';
>> ERROR:  malformed array literal: "d"
>>
>> The assumption that the second argument is an array constant seems
>> surprising

> ​It has to assume something.  And for better and worse it has to assume it
> without looking at the actual value.

Yeah.  The core problem here is that the parser has to disambiguate the
|| operator: is it "anyarray || anyelement" or "anyarray || anyarray"?
In your first example the array can be seen to be int[] and 7 is taken
to be type int, so only "anyarray || anyelement" works.  In the second
case it's looking at "int[] || unknown", and the relevant heuristic is
to assume that the "unknown" is the same type as the operator's other
input.

Peeking at the contents of the literal would make the behavior very
unpredictable/data-dependent, so we don't.

            regards, tom lane


pgsql-general by date:

Previous
From: "David G. Johnston"
Date:
Subject: Re: [GENERAL] Surprising results from array concatenation
Next
From: Mike Blackwell
Date:
Subject: Re: [GENERAL] Surprising results from array concatenation