Re: Cache lookup error when using jsonb, json_build_object and a WITH clause - Mailing list pgsql-hackers

From Andrew Dunstan
Subject Re: Cache lookup error when using jsonb, json_build_object and a WITH clause
Date
Msg-id 536CE6C9.1060108@dunslane.net
Whole thread Raw
In response to Re: Cache lookup error when using jsonb, json_build_object and a WITH clause  (Tom Lane <tgl@sss.pgh.pa.us>)
Responses Re: Cache lookup error when using jsonb, json_build_object and a WITH clause  (Tom Lane <tgl@sss.pgh.pa.us>)
List pgsql-hackers
On 05/09/2014 10:07 AM, Tom Lane wrote:
> After further reflection I think we should lose the TYPCATEGORY_NUMERIC
> business too.  ruleutils.c hard-wires the set of types it will consider
> to be numeric, and I see no very good reason not to do likewise here.
> That will remove the need to look up the typcategory at all.
>
> So we need to:
>
> 1. Refactor so there's only one copy of the control logic.
>
> 2. Smash domains to their base types.
>
> 3. Identify boolean, numeric, and json types by direct tests of type OID.
>
> 4. Identify array and composite types using standard methods.
>
> Anybody see other problems to fix here?


I guess this is my fault. I recall some discussions when some of this 
was first being written about the best way to make the type based 
decisions, not sure at this remove whether on list or off. The origin of 
it is in 9.2, so if you're going to adjust it you should probably go 
back that far.

I was aware of the domain problem, but in 2 years or so nobody has 
complained about it, so I guess nobody is defining domains over json.

cheers

andrew






pgsql-hackers by date:

Previous
From: Greg Stark
Date:
Subject: Re: default opclass for jsonb (was Re: Call for GIST/GIN/SP-GIST opclass documentation)
Next
From: Andres Freund
Date:
Subject: Re: default opclass for jsonb (was Re: Call for GIST/GIN/SP-GIST opclass documentation)