Hi Tom,
thank you for the explanation.
Regards,
Matija Lesar
On 14 December 2016 at 15:53, Tom Lane wrote:
> Matija Lesar writes:
> > I have uint4 domain created like this:
> > CREATE DOMAIN uint4 AS int8
> >CHECK(VALUE BETWEEN 0 AND 4294967295);
>
> > If I
e then check is also working as expected:
CREATE TABLE test(i uint4);
INSERT INTO test VALUES(-1);
ERROR: value for domain uint4 violates check constraint "uint4_check"
I tested this on PG 9.4 and 9.5.
Is this expected behavior?
Regards,
Matija Lesar
3,2,null,12,-1]::int[]) b;
myfirstval | mylastval
+---
-1 |12
select myfirstval(b), mylastval(b) from generate_series(10,2) as b;
myfirstval | mylastval
+---
10 | 2
select myfirstval(b), mylastval(b) from
unnest(array['c','b','t','x']::text[]) b;
myfirstval | mylastval
+---
c | x
Bye,
Matija Lesar
id,
clock_timestamp() as ctimestamp
FROM generate_series(1,1000) as id
)
SELECT
*
FROM
(SELECT
id,
ctimestamp,
row_number() OVER (ORDER BY ctimestamp) as rownum
FROM data_cte
) as data_withrownumbers
WHERE
rownum%100=1;
Bye,
Matija Lesar
On 20 March 2015 at 14:58, Tom Lane wrote:
> AFAICS, array_remove keeps the existing lower bound number.
>
Thank you for explanation. This is not specified in
http://www.postgresql.org/docs/current/static/functions-array.html#ARRAY-FUNCTIONS-TABLE
so I was not sure.
Regards,
Matija Lesar
_lower | array_upper
-+--+-+-
{NULL,b,NULL,d} | {b,d}| 1 | 2
(1 row)
I expected that in both results values will start from index 1.
Regards,
Matija Lesar
