Mark wrote:
-----Original Message-----
From: Stuart Johnston [mailto:[EMAIL PROTECTED]
Sent: vrijdag 17 november 2006 23:30
To: users@spamassassin.apache.org
Subject: Re: Bayes column 'token'
CREATE TABLE bayes_token (
id int(11) NOT NULL default '0',
token char(5) COLLATE latin1_bin NOT NULL default '',
spam_count int(11) NOT NULL default '0',
ham_count int(11) NOT NULL default '0',
atime int(11) NOT NULL default '0',
PRIMARY KEY (id, token),
INDEX bayes_token_idx1 (atime)
) TYPE=MyISAM;
Those are multi-column indexes not duplicates.
INDEX bayes_token_idx1 (id, atime)
is NOT the same as:
INDEX bayes_token_idx1 (id)
INDEX bayes_token_idx2 (atime)
The multi-column INDEX actually *does* appear to be used.
My bad. Though it still doesn't justify the existence of:
INDEX bayes_token_idx1 (token)
I haven't looked at the SQL but does it not do something like: WHERE token = "?"
