On Wed, Jan 21, 2004 at 03:22:29AM -0500, jim wrote:
> for some reason they are trying to name the constraint. no idea why.
That determines the name of the index that will be used to implement the
constraint.
> It appears sqlitemanager doesn't use it.
Yup, it looks like sqlite ignores the cons
>>I am also wondering about the constraint in the column-def like
column-def ::= name [type] [[CONSTRAINT name] column-
>>constraint]*
I thought about this some more. since constraint shows up in blue it is
reserved.
for some reason they are trying to name the constraint. no idea why.
It appear
wish to make my
own.
thanks for helping me dave,
marvin
-Original Message-
From: David M. Cook [mailto:[EMAIL PROTECTED]
Sent: Tuesday, January 20, 2004 8:11 PM
To: [EMAIL PROTECTED]
Subject: Re: [sqlite] basic table level stuff
On Tue, Jan 20, 2004 at 05:15:15PM -0500, jim wrote
On Tue, Jan 20, 2004 at 05:15:15PM -0500, jim wrote:
> if they already have these constraints at the column level,
> why do they need them a second time as in ...
> name [type] [[CONSTRAINT name] column-constraint]*
Because you may want a composite primary key or set of unique columns. A
simple
Hi-
One thing that is puzzling me is the contstraint below that isnt a
column constraint.
constraint ::=PRIMARY KEY ( name [, name]* ) [ conflict-clause ]|
UNIQUE ( name [, name]* ) [ conflict-clause ] |
CHECK ( expr ) [ conflict-clause ]
if they already have these constraints at the column level
5 matches
Mail list logo