On 09/12/2014 06:40 AM, Alexander Korotkov wrote:
> Hi!
>
> Let's consider some examples.
>
> # select '[1,2]'::jsonb @> '[1,2,2]'::jsonb;
> ?column?
> ----------
> f
> (1 row)
>
> One may think it's because second jsonb array contain two "2". So,
> contains takes care about count of equal elements.
JSONB arrays are allowed to have repleating elements. It's keys which
are not allowed to repeat.
>
> # select '[1,1,2]'::jsonb @> '[1,2,2]'::jsonb;
> ?column?
> ----------
> t
> (1 row)
>
> But, it's not. Jsonb contains takes care only about length of array.
OK, now, that's messed up.
>
> # select '[[1,2]]'::jsonb @> '[[1,2,2]]'::jsonb;
> ?column?
> ----------
> t
> (1 row)
>
> Even more weird :)
> The reason why jsonb contains behaves so is check in the beginning
> of jsonb_contains. It makes fast check of jsonb type and elements count
> before calling JsonbDeepContains.
>
> if (JB_ROOT_COUNT(val) < JB_ROOT_COUNT(tmpl) ||
> JB_ROOT_IS_OBJECT(val) != JB_ROOT_IS_OBJECT(tmpl))
> PG_RETURN_BOOL(false);
>
> It's likely that "JB_ROOT_COUNT(val) < JB_ROOT_COUNT(tmpl)" should be
> checked only for objects, not arrays.
Yeah, agreed.
--
Josh Berkus
PostgreSQL Experts Inc.
http://pgexperts.com