Andrew Dunstan <andrew@dunslane.net> writes:
> On 06/23/2014 11:43 AM, Tom Lane wrote:
>> However, it seems to me that these functions (json[b]_to_record[set]) are
>> handling the nested-json-objects case in a fairly brain-dead fashion to
>> start with. I would like to propose that we should think about getting
>> rid of the use_json_as_text flag arguments altogether. What purpose do
>> they serve? If we're going to the trouble of parsing the nested JSON
>> objects anyway, why don't we just reconstruct from that data?
> Looks like we have some problems in this whole area, not just the new
> function, so we need to fix 9.3 also :-(
> IIRC, originally, the intention was to disallow nested json objects, but
> the use_json_as_text was put in as a possibly less drastic possibility.
> If we get rid of it our only recourse is to error out if we encounter
> nested json. I was probably remiss in not considering the likelihood of
> a json target field.
> I currently don't have lots of time to devote to this, sadly, but
> Michael's patch looks like a good minimal fix.
I can spend some time on it over the next couple of days. I take it you
don't have a problem with the concept of doing recursive processing,
as long as it doesn't add much complication?
I'm not following your comment about 9.3. The json[b]_to_record[set]
functions are new in 9.4, which is what makes me feel it's not too
late to redefine their behavior. But changing behavior of stuff that
was in 9.3 seems a lot more debatable.
regards, tom lane