t; <[EMAIL PROTECTED]>
To: "Alexandru Coseru" <[EMAIL PROTECTED]>
Cc: "Heikki Linnakangas" <[EMAIL PROTECTED]>;
Sent: Sunday, December 03, 2006 6:05 AM
Subject: Re: [PERFORM] Regex performance issue
"Alexandru Coseru" <[EMAIL PROTECTED]> w
D]>;
Sent: Saturday, December 02, 2006 10:54 PM
Subject: Re: [PERFORM] Regex performance issue
I may miss something but I'd use tsearch2. Check
intdict dictionary for basic idea -
http://www.sai.msu.su/~megera/wiki/Gendict
Oleg
On Sat, 2 Dec 2006, Alexandru Coseru wrote:
Hello..
"Alexandru Coseru" <[EMAIL PROTECTED]> writes:
> Anyway , anybody has a clue why this regex is that CPU intensive ?
The EXPLAIN result you posted offers *no* evidence that the regexp is
CPU intensive. All you know is that it took 850+ msec to fetch 5200
rows from disk and apply the regexp filter
On Sun, 2006-12-03 at 02:53 +0200, Alexandru Coseru wrote:
> Hello..
>
> Thanks for the tip , i think i have got the ideea..
>
> I'm too tired too , and i will try it tommorow.
>
>
> Anyway , anybody has a clue why this regex is that CPU intensive ? I did
> not saw the light on my drives blin
;Oleg Bartunov"
To: "Alexandru Coseru" <[EMAIL PROTECTED]>
Cc: "Dave Dutcher" <[EMAIL PROTECTED]>;
Sent: Saturday, December 02, 2006 10:54 PM
Subject: Re: [PERFORM] Regex performance issue
I may miss something but I'd use tsearch2. Check
intdict dictionary
PROTECTED]>
To: "Alexandru Coseru" <[EMAIL PROTECTED]>
Cc:
Sent: Sunday, December 03, 2006 12:35 AM
Subject: Re: [PERFORM] Regex performance issue
Alexandru Coseru wrote:
Hello..
I cannot use the first advice , because i'm not aware of the prefix
length in the datab
Alexandru Coseru wrote:
Hello..
I cannot use the first advice , because i'm not aware of the prefix
length in the database...
This is why i'm ordering after length(prefix)..
On the 2nd one , i'm not sure that i can follow you..
Ok, let me try again :)
asterisk=> select * from destlist LI
Linnakangas" <[EMAIL PROTECTED]>
To: "Alexandru Coseru" <[EMAIL PROTECTED]>
Cc: "Dave Dutcher" <[EMAIL PROTECTED]>;
Sent: Sunday, December 03, 2006 12:04 AM
Subject: Re: [PERFORM] Regex performance issue
Alexandru Coseru wrote:
I cannot use LIKE , b
Alexandru Coseru wrote:
I cannot use LIKE , because the order of the match is reversed.
The prefix column is containing telephone destinations.
IE:^001 - US , ^0039 Italy , etc..
Maybe you could create a functional index on substr(prefix>)? It might restrict the result set prior to applyi
something..
Regards
Alex
- Original Message - From: "Dave Dutcher" <[EMAIL PROTECTED]>
To: "'Alexandru Coseru'" <[EMAIL PROTECTED]>;
Sent: Saturday, December 02, 2006 10:36 PM
Subject: RE: [PERFORM] Regex performance issue
-Original M
ROTECTED]>;
Sent: Saturday, December 02, 2006 10:36 PM
Subject: RE: [PERFORM] Regex performance issue
-Original Message-
From: [EMAIL PROTECTED] On Behalf Of Alexandru Coseru
asterisk=> explain analyze SELECT * FROM destlist WHERE
'0039051248787' ~
pr
> -Original Message-
> From: [EMAIL PROTECTED] On Behalf Of Alexandru Coseru
> asterisk=> explain analyze SELECT * FROM destlist WHERE
> '0039051248787' ~
> prefix AND id_ent='-2' AND dir=0 ORDER by length(prefix) DESC;
>
>
> Q
Hello..
I have a low performance problem with regexp.
Here are the details:
asterisk=> explain analyze SELECT * FROM destlist WHERE '0039051248787' ~
prefix AND id_ent='-2' AND dir=0 ORDER by length(prefix) DESC;
QUERY PLAN
-
13 matches
Mail list logo