1 vs True vs 't'

I'm using SQLite3, but have been a longtime MySQL user. I have been wracking my brain over why I was getting a "0" result for a count when it should be "10".

The statement.... Table.count (:conditions => "approved = 0")

The result in SQLite3: 0 The result in MySQL: 10

The corrected (I think) statement for SQLite: Table.count (:conditions => "approved = 'f' ")

I am surprised and wonder why SQLite using the 'f' and 't' nomenclature when '0' and '1' have worked so well....

Since this specificity would rule out me easily switching my DB back to SQL, how can I write my command to use True and False?

Thanks a ton...

I did solve my problem...

Table.count (:conditions => {approved => false} )

yachtman wrote:

I did solve my problem...

Table.count (:conditions => {approved => false} )

Or this will work as well: Table.count(:conditions => [ "approved = ?", true ])

Putting the value in the conditions string doesn't gives Rails a chance to process it. Instead it just goes directly into the WHERE clause as is. Both the hash version and the value substitution using the question mark version allow Rails to process and properly sanitize the SQL.

Robert Walker wrote:

Table.count(:conditions => [ "approved = ?", true ])

Well I meant Table.count(:conditions => [ "approved = ?", false ]), but you get the idea.

Robert,

Thanks for the reply.

I imagine in this case, since I am only counting based on a boolean value, there should be no worry about sanitization?

You're not doing ['approved = ?',false] for the sanitization as much as for letting the database adapter fill in its own notion of what value it uses for false. Even if you were doing more than counting, you have control of the inputs so there's no SQL injection here.

-Rob

Rob Biedenharn http://agileconsultingllc.com Rob@AgileConsultingLLC.com

seems to work. I think rails is doing its thing either way.

issue = Issue.count(:conditions => {:published => false}) => 3

[4;36;1mSQL (0.000348) [0m [0;1mSELECT count(*) AS count_all FROM issues WHERE (issues.published = 0) [0m