2010/9/1 Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>:
> Pavel Stehule <pavel(dot)stehule(at)gmail(dot)com> writes:
>> 2010/9/1 Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>:
>>> Well, you can build your own version of array_agg with the same
>>> implementation, except you mark the transition function as strict ...
>
>> I am checking this now, and it is not possible - it needs a some
>> initial value and there isn't possible to set a "internal" value.
>
> Well, you can cheat a bit ...
>
> regression=# create or replace function array_agg_transfn_strict(internal, anyelement) returns internal as 'array_agg_transfn' language internal immutable;
> CREATE FUNCTION
> regression=# create aggregate array_agg_strict(anyelement) (stype = internal,
> sfunc = array_agg_transfn_strict, finalfunc = array_agg_finalfn);
> CREATE AGGREGATE
> regression=# create or replace function array_agg_transfn_strict(internal, anyelement) returns internal as 'array_agg_transfn' language internal strict immutable;
> CREATE FUNCTION
>

nice dark trick :) -  but it doesn't work

ERROR:  aggregate 16395 needs to have compatible input type and transition type
postgres=#

Pavel



>                        regards, tom lane
>

In response to

• Re: array_agg() NULL Handling at 2010-09-01 18:46:00 from Tom Lane

Responses

• Re: array_agg() NULL Handling at 2010-10-02 02:57:50 from David E. Wheeler

pgsql-hackers by date

Next:	From: Tom Lane	Date: 2010-09-01 19:31:33
	Subject: Re: compiling with RELCACHE_FORCE_RELEASE doesn't pass regression
Previous:	From: Tom Lane	Date: 2010-09-01 18:46:00
	Subject: Re: array_agg() NULL Handling