Re: [Hibernate-devel] Is this type of query possible?

[Jon L. <jon...@xe...>]

Thanks Gavin,

Yes, that will work...  That select was in some legacy code and I was just
trying to replace it like it was.  (To many hours staring at other people
code :-)  I guess I owe you a beer for saving me some time...

I like the idea of your select with the [] syntax.  I'll take a look to see
if that would reduce the number of classes I need somewhere, and if so I'll
try to add it.  I'll keep you posted...

Cheers,
Jon...

----- Original Message -----
From: <Gavin_King/Cirrus%CI...@ci...>
To: "Jon Lipsky" <jon...@xe...>
Cc: <hib...@li...>
Sent: Tuesday, July 02, 2002 2:32 AM
Subject: Re: [Hibernate-devel] Is this type of query possible?


>
> > Basically, there is a parent table ("foo" in my example above) and
> > there is list of attributes for a particular foo object
> > (called "foo_attr") in my example above.  I want to find all Foo's
> > where it have an attribute "attribute1" equal to "value1" and an
> > attribute "attribute2" equal to "value2".
> >
> > So, is this something I will have to implement (and if so, are there
> > suggestions on the syntax of the query), or does this already exist?
>
> I *think* I understand what you are trying to do though I'm not sure
> why you used a nested subselect so perhaps there is something I'm
> missing.
>
> If you were to map this as a (nested) one-to-many set of Attibutes, I
> think you can express this as:
>
> select foo
> from foo in class Foo, att1 in foo.attributes, att2 in foo.attributes
> where att1.name = 'attribute1' and att1.value = 'value1'
> and att2.name='attribute2' and att2.value = 'value2'
>
> (though the generated SQL here contains no subselects). What I would
> love to be able to do would be to map this instead as a (nested) map
> with String index + element and then express the query as:
>
> from foo in class Foo
> where foo.attributes['attribute1']='value1'
> and foo.attributes['attribute2']='value2'
>
> However, the [] indexing syntax is not implemented. The advantage of
> this approach is we don't have to invent this whole new class called
> Attribute to contain name/value pairs. If someone wants to implement
> [], that would be wonderful - but it might be nontrivial.
>
> Does that help?
>