Re: can someone explain confusing array indexing nomenclature

Στις Τετάρτη 14 Φεβρουάριος 2007 21:31, ο/η chrisj έγραψε:
> given the following table:
>
> protocal2=> select * from sal_emp ;
> name | pay_by_quarter | schedule
> -------+---------------------------+---------------------------------------
>---- Bill | {10000,10000,10000,10000} |
> {{meeting,lunch},{training,presentation}}
> Carol | {20000,25000,25000,25000} |
> {{breakfast,consulting},{meeting,lunch}}
> (2 rows)
>
> why do the following two queries yield different results??
>
> protocal2=> SELECT schedule[1][2] FROM sal_emp WHERE name = 'Bill';
> schedule
> ----------
> lunch
> (1 row)
>
> protocal2=> SELECT schedule[1:1][2] FROM sal_emp WHERE name = 'Bill';
> schedule
> -------------------
> {{meeting,lunch}}
> (1 row)

The [n:m] notation denotes a slice of the array (not element).
So schedule[1][2] is the Array element on 2nd col of 1st row,
while schedule[1:1][2] could mean
the second row of the subarray schedule[1:1][1:2].
So these two are foundamentally different things.
In my 7.4 even if you gave
SELECT schedule[1:1][888] FROM sal_emp WHERE name = 'Bill';
you would still get {{meeting,lunch}} as a result.
(Right or wrong is another story).
Anyway the first time you query for a "text",
the second time you query for a "text[]", so you should expect
different results.
--
Achilleas Mantzios