Django独特不与Postgres一起使用

Question

I'm configuring an API to pull a list of results. The items are clothing 'Products', which have several 'Variations'.

I am trying to use distinct to make sure the queryset only returns a list of unique products- instead of the same product repeated several times due to many variations.

In the past, I used distinct on product's primary key (id). However, the below code is not working.

I get a 500 error when I try using "distinct". Without it- I get no error. I cannot test locally as SQLite does not like distinct. I am using

class InternalListView(APIView):

    renderer_classes = (JSONRenderer, )

    def get(self, request, *args, **kwargs):

        filters = {}
        for key, value in request.GET.items():
            key = key.lower()
            if key in countmatch:
                lookup, val = internalmatch[key](value.lower())
                filters[lookup] = val

        qset = (
            Product.objects
            .filter(**filters)
            .distinct('id')
            .order_by('-rating')
            .values('name', 'brand', 'rating')
            .annotate(
                price=F('variation__price__price'),
                id=F('pk'),
                vari=F('variation'),
            )
        )

        for i in qset:
            i['likes'] = random.randint(500, 1000)

        print qset

        return Response(qset.all())

Answer 1

Ahh just figured it out.

According to the django docs ( https://docs.djangoproject.com/en/1.9/ref/models/querysets/#distinct )

Both distinct & order_by have to be the same.

So this does NOT work:

        .distinct('id')
        .order_by('-rating')

But this WILL work:

        .distinct('id')
        .order_by('id')

And this works best:

        .distinct('rating', 'id')
        .order_by('-rating')

Anything wrong with using option 3?