如何从序列化器字段中订购Django QuerySet
?由于该字段太复杂,我无法使用QuerySet
对annotate
进行排序,我也无法将该值存储在模型中。
编辑:
串行
class DrinkListModelSerializer(serializers.ModelSerializer):
count_need = serializers.SerializerMethodField()
url = drink_detail_url
class Meta:
model = Drink
fields = [
'name',
'count_need',
'thumbnail',
'url'
]
def get_count_need(self, obj):
drink_ingredient_qs = obj.ingredients.all()
user = self.context['request'].user
user_qs = User.objects.filter(username=user.username)
if user_qs.exists() and user_qs.count() == 1:
user_obj = user_qs.first()
user_ingredient_qs = user_obj.ingredient_set.all()
return (user_ingredient_qs & drink_ingredient_qs).count()
return 0
型号:
class Drink(models.Model):
name = models.CharField(max_length = 1000, null=True, blank=True)
ingredients = models.ManyToManyField(Ingredient, blank=True)
user = models.ManyToManyField(settings.AUTH_USER_MODEL, blank=True)
class Ingredient(models.Model):
name = models.CharField(max_length = 1000)
ingredient_category = models.ForeignKey(Category, null=True, blank=True)
slug = models.SlugField(max_length=255, unique=True)
user = models.ManyToManyField(settings.AUTH_USER_MODEL, blank=True)
查看
class DrinkListAPIView(ListAPIView):
serializer_class = DrinkListModelSerializer
pagination_class = LargeResultsSetPagination
permission_classes = [AllowAny]
def get_queryset(self, *args, **kwargs):
user = self.request.user
qs = Drink.objects.all()
userQuery = self.request.GET.get('user')
query = self.request.GET.get("q")
filters = self.request.GET.getlist('filter')
if query:
qs = qs.filter(name__icontains=query)
elif filters:
qs = qs.filter(playlist__name__iexact=filters[0])
for filter in filters[1:]:
qs = qs | Drink.objects.filter(playlist__name__iexact=filter)
elif userQuery and user.is_authenticated():
qs = qs.filter(user=user)
return qs.order_by('-timestamp')
目前,我按时间戳排序查询集,因为它在模型中是自己的,但我需要通过' count_need'来订购它。这是序列化器的一部分。我试图做qs = qs.annotate(count_need=Count('ingredients' & user_ingredient_qs)).order_by(count_need)
这样的事情,但这显然不起作用,因为'成分'这里只是一个字符串,而不是Drink的成分查询集。
答案 0 :(得分:1)
可能运行良好的一种方法是在列表序列化器上实现和自定义""
:
to_representation