Optimize retrieving a single object

[Joe Meissler]

I'm following the graphene-django tutorial and got up to retrieving a single object. The README here does not mention how optimization would work here. gql_optimizer.query expects a queryset. As a workaround, I'm using:

      def resolve_category(self, info, **kwargs):
          id = kwargs.get('id')
          name = kwargs.get('name')

          if id is not None:
              return gql_optimizer.query(Category.objects.filter(pk=id), info).get()

          if name is not None:
              return gql_optimizer.query(Category.objects.filter(name=name), info).get()

          return None

Is this the recommended approach?

[Matthew Abrams]

Interested to know the proposed solution here as well, this seems like it should work ok. Not sure if it could be supported, but obviously preferred developer experience would likely be:

gql_optimizer.query(Category.objects.get(name=name), info)

which currently raises 'Category' object has no attribute 'select_related'

[Joe Meissler]

That's because .get() returns an instance instead of a queryset. You can do gql_optimizer.query(Category.objects.select_related(...).get(name=name), info)

[Matthew Abrams]

@stickperson Yep, makes sense, I am using your workaround for now 👍

[Tomás Fox]

Yes, the first post is the recommended approach. It's been a long time since I used Django, but from what I remember, when using .get you would lose the prefetch_related optimizations.