Re: JSON output functions.

On 02/02/2012 12:20 PM, Pavel Stehule wrote:
> 2012/2/2 Andrew Dunstan<andrew(at)dunslane(dot)net>:
>>
>> On 02/02/2012 04:35 AM, Abhijit Menon-Sen wrote:
>>> At 2012-02-01 18:48:28 -0500, andrew(dot)dunstan(at)pgexperts(dot)com wrote:
>>>> For now I'm inclined not to proceed with that, and leave it as an
>>>> optimization to be considered later if necessary. Thoughts?
>>> I agree, there doesn't seem to be a pressing need to do it now.
>>>
>>
>> OK, here's my final version of the patch for constructor functions. If
>> there's no further comment I'll go with this.
> These function are super, Thank you
>
> Do you plan to fix a issue with row attribute names in 9.2?

Yeah. Tom did some initial work which he published here:
<http://archives.postgresql.org/message-id/28413.1321500388%40sss.pgh.pa.us>,
noting:

It's not really ideal with respect to
the ValuesScan case, because what you get seems to always be the
hard-wired "columnN" names for VALUES columns, even if you try to
override that with an alias
...
Curiously, it works just fine if the VALUES can be folded

and later he said:

Upon further review, this patch would need some more work even for the
RowExpr case, because there are several places that build RowExprs
without bothering to build a valid colnames list. It's clearly soluble
if anyone cares to put in the work, but I'm not personally excited
enough to pursue it ..

I'm going to look at that issue first, since the unfolded VALUES clause seems like something of an obscure corner case. Feel free to chime in if you can.

cheers

andrew