On 03/02/2016 11:31 PM, Tom Lane wrote:
> I added a regression test case based on his example, plus parallel tests
> for json_to_recordset, jsonb_to_record, jsonb_to_recordset. The latter
> three do not exhibit the same bug (which suggests that we may be missing
> some opportunities to share code...)
Not really. The jsonb variants don't do any re-parsing at all, so
lex_level simply doesn't come into play, and for json_to_recordset we
want things from lex_level 2, not 1, as the json is supposed to be an
array of objects instead of just an object.
That's not to say that the code can't be improved - It very probably can.
cheers
andrew