From: | Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> |
---|---|
To: | Geoff Winkless <pgsqladmin(at)geoff(dot)dj> |
Cc: | Postgres General <pgsql-general(at)postgresql(dot)org> |
Subject: | Re: json ->> operator precedence |
Date: | 2015-08-05 13:52:19 |
Message-ID: | 9076.1438782739@sss.pgh.pa.us |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-general |
Geoff Winkless <pgsqladmin(at)geoff(dot)dj> writes:
> An interesting quirk:
> # select CASE WHEN '{"a":null}'::jsonb->>'a' IS NULL THEN 'yes' ELSE 'no'
> END;
> case
> ------
> yes
Apparently you're running that on 9.5 or HEAD.
> According to the precedence table
> http://www.postgresql.org/docs/9.4/static/sql-syntax-lexical.html I would
> expect ->> to come under "all other native and user-defined operators",
It does ...
> which would imply that this command should be testing whether 'a' IS NULL
> and applying the result (false) to the json operator - at which point we
> have
... and in 9.4 that's what happens:
regression=# select CASE WHEN '{"a":null}'::jsonb->>'a' IS NULL THEN 'yes' ELSE 'no'
END;
ERROR: operator does not exist: jsonb ->> boolean
LINE 1: select CASE WHEN '{"a":null}'::jsonb->>'a' IS NULL THEN 'yes...
^
HINT: No operator matches the given name and argument type(s). You might need to add explicit type casts.
> Or am I missing something?
The first compatibility item in the 9.5 release notes: we changed
the precedence of IS and some other things. You need to be reading
the 9.5 version of the precedence table.
regards, tom lane
From | Date | Subject | |
---|---|---|---|
Next Message | Geoff Winkless | 2015-08-05 13:54:29 | Re: json ->> operator precedence |
Previous Message | John McKown | 2015-08-05 13:47:40 | Re: json ->> operator precedence |