Re: [PERFORM] Feature request: smarter use of

From: Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>
To: siracusa(at)mindspring(dot)com
Cc: Postgres Patches <pgsql-patches(at)postgresql(dot)org>
Subject: Re: [PERFORM] Feature request: smarter use of
Date: 2004-03-07 06:28:38
Message-ID: 1337.1078640918@sss.pgh.pa.us
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-patches

John Siracusa <siracusa(at)mindspring(dot)com> writes:
> ... Actually, I'm still not sure how you
> can conclude that foo is not null simply by seeing func(..., foo, ...) where
> func() is strict. Do strict functions and operators simply not allow null
> args?

That was what the comments were about ;-). A "strict" operator is one
that always returns NULL if any input is NULL. This does guarantee that
the WHERE condition will fail, but it takes a bit of analysis to see
that the implication is valid.

> Here's hoping that my next performance enhancement suggestion will be
> complex enough that no one will entertain the thought of asking me for a
> patch... ;)

I had quite the opposite motivation in mind in asking you to prototype
this ;-) ...

regards, tom lane

In response to

Browse pgsql-patches by date

  From Date Subject
Next Message John Gray 2004-03-07 18:58:20 contrib/xml2 function name change and minor bitrot fix
Previous Message John Siracusa 2004-03-07 06:12:25 Re: [PERFORM] Feature request: smarter use of