Interestingly, I tried the new version with and without enable_seqscan
on, and the version without indexes performs better because, I think,
it returns more rows than an index lookup would enhance.
Thanks again for your help. This is certainly an improvement over my
original version.
-tfo
--
T
Pierre,
Your re-write makes a lot of sense. Thanks! It's not using indexes for
some reason, and discovering why will be my next challenge.
-tfo
--
Thomas F. O'Connell
Co-Founder, Information Architect
Sitening, LLC
http://www.sitening.com/
110 30th Avenue North, Suite 6
Nashville, TN 37203-6320
6
I wasn't sure whether EXCEPT would create a unique set from among the
results of both queries.
As in, if the first part of the query (before the EXCEPT clause),
without the DISTINCT, yielded
yahoo.com
yahoo.com
would the query reduce that to a single yahoo.com regardless of whether
it showed u
=?iso-8859-15?Q?Pierre-Fr=E9d=E9ric_Caillaud?= <[EMAIL PROTECTED]> writes:
> I may be mistaken, but you may be doing the same thing twice : you're
> basically writing :
> SELECT DISTINCT X WHERE Y EXCEPT SELECT DISTINCT X WHERE NOT Y
> Is this not a way to get an empty result s
Now, since I'm actually interested in unique domain names rather than
unique users, I need to get all the unique domain names corresponding to
users who have acted on a message. That's what the part of the query
after the EXCEPT is.
I don't understand this part at all. What does it mean ?
I'm trying to do some research and reporting for an email application
by domain name. This has led to a confounding attempt to do any of the
legwork in SQL via postgres.
Here is my foundational query:
SELECT DISTINCT split_part( u.email, '@', 2 )
FROM user AS u, message AS m
WHERE u.id = m.use
