Joe Conway <mail@joeconway.com> writes:
> Jim C. Nasby wrote:
>> Instead of bending over backwards to try and support older cases, would
>> a compatability mode be possible? Seems that would solve a lot of
>> problems.
> Last time I thought about this problem, that's what I concluded. I don't
> think there is a reasonable and backward compatible solution.
> I also think the best non-compatible solution is to require non-numeric
> elements to be delimited (double quotes, configurable?), and use NULL
> unadorned to represent NULL.
After further thought I'm starting to agree with this point of view as
well. I propose the following details:
1. A null element is represented as the unquoted string NULL (case-insensitive on input). Any use of quotes or
backslashes turns it into a simple string value "NULL" instead. array_out will need to be careful to quote any string
thatmatches NULL.
2. For backwards compatibility, we'll invent a GUC parameter enable_array_nulls that defeats recognition of NULL in
array_in. (Any better ideas about the name of the parameter?)
This isn't ideal because of the compatibility risk, but once we get past
the transition period it's a reasonable definition.
regards, tom lane
