SQL query - single text value from group by

[mike]

Hi

I am trying to work out if this is possible in a select query

I have a group by query which could result in several rows, what I want
to do is do a text equivalent of a sum() eg:

SELECT sum(inv_id),date,cust
from invoice
group by date,cust

Is there any way to get to a single concatenated inv_id field with just
one row?

thanks

---------------------------(end of broadcast)---------------------------
TIP 4: Don't 'kill -9' the postmaster

[Richard Huxton]

mike wrote:

Hi

I am trying to work out if this is possible in a select query

I have a group by query which could result in several rows, what I want
to do is do a text equivalent of a sum() eg:

SELECT sum(inv_id),date,cust
from invoice
group by date,cust

Is there any way to get to a single concatenated inv_id field with just
one row?

Yes, you can define your own aggregate. For example, I did this the
other day:

/*
Custom aggregate
This aggregate is so we can aggregate text into paragraph blocks
*/
CREATE OR REPLACE FUNCTION join_paras(text, text) RETURNS text AS
'
SELECT CASE
WHEN ($1 = '''') THEN $2
ELSE $1 || ''\n'' || $2
END;
' LANGUAGE 'SQL' IMMUTABLE;

CREATE AGGREGATE agg_paras (sfunc1=join_paras, basetype=text,
stype1=text, initcond1='');

Note I defined my own text-concatenation function because I wanted to
insert newlines between each block of text. If you just wanted joined
text you could use the built-in textcat()

Full specs of create aggregate are in the manuals.

HTH
--
Richard Huxton
Archonet Ltd

---------------------------(end of broadcast)---------------------------
TIP 1: subscribe and unsubscribe commands go to ma*******@postgresql.org

[mike]

On Wed, 2004-09-08 at 17:13 +0100, Richard Huxton wrote:

mike wrote:

Hi

I am trying to work out if this is possible in a select query

I have a group by query which could result in several rows, what I want
to do is do a text equivalent of a sum() eg:

SELECT sum(inv_id),date,cust
from invoice
group by date,cust

Is there any way to get to a single concatenated inv_id field with just
one row?

Yes, you can define your own aggregate. For example, I did this the
other day:

/*
Custom aggregate
This aggregate is so we can aggregate text into paragraph blocks
*/
CREATE OR REPLACE FUNCTION join_paras(text, text) RETURNS text AS
'
SELECT CASE
WHEN ($1 = '''') THEN $2
ELSE $1 || ''\n'' || $2
END;
' LANGUAGE 'SQL' IMMUTABLE;

CREATE AGGREGATE agg_paras (sfunc1=join_paras, basetype=text,
stype1=text, initcond1='');

Note I defined my own text-concatenation function because I wanted to
insert newlines between each block of text. If you just wanted joined
text you could use the built-in textcat()

Full specs of create aggregate are in the manuals.

HTH

thanks - as sods law dictates I found a solution just after I posted
along the same lines, after a couple of hours of fruitless googling.

---------------------------(end of broadcast)---------------------------
TIP 7: don't forget to increase your free space map settings