"Ken Winter" writes:
> In case anybody else needs this functionality, let me offer a family of
> functions that do comparisons that treat NULL as a real value (rather than
> as "unknown"). For example:
Er ... this'd be a lot shorter using IS [NOT] DISTINCT FROM.
regards,
t
> Subject: Re: [GENERAL] Null comparisons and the transform_null_equals run-
> time parameter
>
>
> On Jun 5, 2010, at 9:46 AM, Ken Winter wrote:
> > In case anybody else needs this functionality, let me offer a family of
> > functions that do comparisons that treat NULL as a
On Jun 5, 2010, at 9:46 AM, Ken Winter wrote:
> I get it. Thanks, Tom.
>
> In case anybody else needs this functionality, let me offer a family of
> functions that do comparisons that treat NULL as a real value (rather than
> as "unknown"). For example:
You should take a look at "is not disti
; From: Tom Lane [mailto:t...@sss.pgh.pa.us]
> Sent: Saturday, June 05, 2010 9:41 AM
> To: Ken Winter
> Cc: 'PostgreSQL pg-general List'
> Subject: Re: [GENERAL] Null comparisons and the transform_null_equals run-
> time parameter
>
> "Ken Winter" writes:
>
"Ken Winter" writes:
> When the run-time parameter transform_null_equals is on, shouldn't two
> variables with NULL values evaluate as equal?
No. That setting does not change the runtime behavior of comparison.
The only thing it does is change the literal syntax "something = NULL"
to "something
When the run-time parameter transform_null_equals is on, shouldn't two
variables with NULL values evaluate as equal? They don't seem to.
At the bottom of this message is a little test function. It tries all
comparisons of NULL-valued variables and NULL constants, both before and
after turni
