Re: PL/pgSQL PERFORM with CTE - Mailing list pgsql-hackers

From Pavel Stehule
Subject Re: PL/pgSQL PERFORM with CTE
Date
Msg-id CAFj8pRC_u4weT4uPftv-QS1MhgeYVf0JPTdyfwW1_K-WsBjsMg@mail.gmail.com
Whole thread Raw
In response to Re: PL/pgSQL PERFORM with CTE  (Merlin Moncure <mmoncure@gmail.com>)
Responses Re: PL/pgSQL PERFORM with CTE  ("David E. Wheeler" <david@justatheory.com>)
Re: PL/pgSQL PERFORM with CTE  (Merlin Moncure <mmoncure@gmail.com>)
List pgsql-hackers



2013/8/20 Merlin Moncure <mmoncure@gmail.com>
On Tue, Aug 20, 2013 at 7:25 AM, Andres Freund <andres@2ndquadrant.com> wrote:
> On 2013-08-20 14:15:55 +0200, David E. Wheeler wrote:
>> Hi Pavel,
>>
>> On Aug 20, 2013, at 2:11 PM, Pavel Stehule <pavel.stehule@gmail.com> wrote:
>>
>> >>     david=# DO $$
>> >>     david$# BEGIN
>> >>     david$#     WITH now AS (SELECT now())
>> >>     david$#     PERFORM * from now;
>> >>     david$# END;
>> >>     david$# $$;
>> >>     ERROR:  syntax error at or near "PERFORM"
>> >>     LINE 4:     PERFORM * from now;
>> >>                 ^
>> >> Parser bug in PL/pgSQL, perhaps?
>> >
>> > no
>> >
>> > you cannot use a PL/pgSQL statement inside SQL statement.
>>
>> Well, there ought to be *some* way to tell PL/pgSQL to discard the result. Right now I am adding a variable to select into but never otherwise use. Inelegant, IMHO. Perhaps I’m missing some other way to do it?
>>
>> If so, it would help if the hint suggesting the use of PERFORM pointed to such alternatives.
>
> Not that that's elegant but IIRC PERFORM (WITH ...) ought to work. I
> don't think the intermingled plpgsql/sql grammars allow a nice way right
> now.

I think the way forward is to remove the restriction such that data
returning queries must be PERFORM'd

I disagree, current rule has sense.

Pavel
 

merlin

pgsql-hackers by date:

Previous
From: Merlin Moncure
Date:
Subject: Re: PL/pgSQL PERFORM with CTE
Next
From: "David E. Wheeler"
Date:
Subject: Re: PL/pgSQL PERFORM with CTE