Yep, agreed. A simple lexical macro-like approach to test "if it works" could be a simple approach to see if inlining a
pieceof sql would not break the main query?
Laurent Hasson
Sent from my BlackBerry Passport
Original Message
From: Tom Lane
Sent: Saturday, November 12, 2016 14:59
To: ldh@laurent-hasson.com
Cc: Marc Mamin; pgsql-performance@postgresql.org
Subject: Re: [PERFORM] Inlining of functions (doing LIKE on an array)
"ldh@laurent-hasson.com" <ldh@laurent-hasson.com> writes:
> I wish there were a way to force inlining, or some other mechanism as the performance difference is large here. I'll
beusing the inlining approach when possible, but the SQL Function approach is simpler and will likely be more suitable
forsome developers.
I'm not sure that there's any fundamental reason why we don't inline SQL
functions containing sub-selects. It may just be not having wanted to put
any effort into the case way-back-when. Inlining happens too late to
allow a resulting WHERE EXISTS to get mutated into a semijoin, but in this
example that couldn't happen anyway, so it's not much of an objection.
regards, tom lane