Hey Djangonauts,
I'd like to optimize the performance of a query I've got running on a
page - I've got a model which has a self referencing nullable foreign
key and I'd like to select all instances of that model that which have
a value for that column.
Imagine a class Person with a nullable FK to Person "Best Friends
Forever" - I'd like to select everyone who has a BFF! This is a
straight forward query operation and I would expect the ORM to avoid
using a JOIN to find it out - it would seem you could just check
whether the bff_id isnull - but in all the queries I've tried, the ORM
wants to do a JOIN on Person, which is expensive and unnecessary.
So my question is - how do you prevent a JOIN while checking if a self
referencing FK is null?
# Begin example codes #
class Person(model.Models):
bbf = models.ForeignKey("self", null=True)
...
# Returns a queryset which causes a join
people_with_bffs = Person.objects.filter(bbf__isnull=False)
# Returns a queryset which causes a join
people_with_bffs = Person.objects.filter(bbf__pk__isnull=False)
# Returns a queryset which doesn't cause a join?
people_with_bffs = Person.objects.filter( ??? )
Cheers,
John
|
