From: | Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> |
---|---|
To: | "David Johnston" <polobo(at)yahoo(dot)com> |
Cc: | hamann(dot)w(at)t-online(dot)de, pgsql-general(at)postgresql(dot)org |
Subject: | Re: Using POSIX Regular Expressions on xml type fields gives inconsistent results |
Date: | 2012-12-23 17:55:06 |
Message-ID: | 27093.1356285306@sss.pgh.pa.us |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-general |
"David Johnston" <polobo(at)yahoo(dot)com> writes:
>> while it might work to reverse the order in a regex, I have experienced
>> severe slowdown when the pattern comes from table data
> If it is indeed related to the custom reverse query the fact that you are
> going through an SQL wrapper to access a c-level function is inevitably
> going to degrade performance.
I believe that in most situations of interest, the SQL wrapper would get
"inlined" and thus create no significant per-row penalty. It's possible
this isn't happening in some specific case, but as you say we'd need to
see a concrete example to know one way or the other.
Another point that comes to mind is that "pattern comes from table data"
probably means "query uses a whole lot of different patterns over its
lifetime". Each different pattern is going to require regex
compilation, which is not a particularly cheap operation.
regards, tom lane
From | Date | Subject | |
---|---|---|---|
Next Message | Adrian Klaver | 2012-12-23 19:38:08 | Re: downgrading a database |
Previous Message | Tom Lane | 2012-12-23 17:16:44 | Re: Insert Assertion Failed in strcoll_l.c:112 |