[EMAIL PROTECTED] uttered:
"Brandon, Nicholas (UK)" <[EMAIL PROTECTED]> wrote:
I don't want to hijack this thread (not too much anyway) but this got me
thinking about JOINs since I have a database that uses a similar concept
(one table holds a number of key-value pairs for another).
As I
[EMAIL PROTECTED] uttered:
Christian Smith <[EMAIL PROTECTED]> wrote:
select
FROM instances as i
inner join instance_fields as count using(instanceid)
inner join instance_fields as first using (instanceid)
inner join instance_fields as last using (instanceid)
inner join
"Brandon, Nicholas (UK)" <[EMAIL PROTECTED]> wrote:
> > > select
> > > FROM instances as i
> > >inner join instance_fields as count using(instanceid)
> > >inner join instance_fields as first using (instanceid)
> > >inner join instance_fields as last using (instanceid)
> > >
Christian Smith <[EMAIL PROTECTED]> wrote:
> > select
> > FROM instances as i
> >inner join instance_fields as count using(instanceid)
> >inner join instance_fields as first using (instanceid)
> >inner join instance_fields as last using (instanceid)
> >inner join
Solved the problem.
For reference, field instanceid in the instance_fields needs to be
declared as an integer, thus:
CREATE TABLE instance_fields (
instanceid integer references instances(instanceid),
field,
subscript default NULL,
value,
primary key (instanceid, field, subscript) );
Hello JOIN expoerts:)
I have the following schema:
CREATE TABLE instances (
instanceid integer primary key,
type,
instance);
CREATE INDEX instances_type_instance ON instances(type,instance);
CREATE TABLE instance_fields (
instanceid references instances(instanceid),
field,
subscript
6 matches
Mail list logo
