Django ORM 使用子查询进行注释

Question

我有以下型号：

class Store(models.Model):
   name = models.CharField()

class Goods(models.Model):
   store = models.ForeigKey(Store, on_delete=models.CASCADE)
   total_cost = models.DecimalField()
   different values ...

所以，我根据参数过滤了所有商品，现在我的目标是从每个商店中获得一件商品，在这家商店的所有商品中价格最低

stores = Store.objects.all() - all stores
goods = Good.objects.filter(..) - filter goods

goods.annotate(min_price=Subquery(Min(stores.values('goods__total_cost'))))                                                                   

我尝试了类似的方法，但出现错误：

AttributeError: 'Min' object has no attribute 'query'

Answer 1

我认为在您的上下文中，您需要 Group By 功能而不是 Django 注释，
来自this SO answer,

>>> q = Book.objects.annotate(num_authors=Count('authors'))
>>> q[0].num_authors
2
>>> q[1].num_authors
1

q 是书籍的查询集，但每本书都标注了作者的数量。 也就是说，如果你 annotate 你的 goods 查询集，他们不会给你一些排序/过滤的对象集。它将仅使用新字段 min_price 进行注释。

所以我建议您按照以下方式进行 Group By 操作

from django.db.models import Min

result = Goods.objects.values('store').annotate(min_val=Min('total_cost'))

示例

In [2]: from django.db.models import Min

In [3]: Goods.objects.values('store').annotate(min_val=Min('total_cost'))
Out[3]: <QuerySet [{'store': 1, 'min_val': 1}, {'store': 2, 'min_val': 2}]>

In [6]: Goods.objects.annotate(min_val=Min('total_cost'))
Out[6]: <QuerySet [<Goods: Goods object>, <Goods: Goods object>, <Goods: Goods object>, <Goods: Goods object>, <Goods: Goods object>]>

In [7]: Goods.objects.annotate(min_val=Min('total_cost'))[0].__dict__
Out[7]:
{'_state': <django.db.models.base.ModelState at 0x7f5b60168ef0>,
 'id': 1,
 'min_val': 1,
 'store_id': 1,
 'total_cost': 1}

In [8]: Goods.objects.annotate(min_val=Min('total_cost'))[1].__dict__
Out[8]:
{'_state': <django.db.models.base.ModelState at 0x7f5b6016af98>,
 'id': 2,
 'min_val': 123,
 'store_id': 1,
 'total_cost': 123}

UPDATE-1
我认为，这不是一个好主意，可能会出现一些优化问题，但您可以尝试

from django.db.models import Min

store_list = Store.objects.values_list('id', flat=True)  # list of id's od Store instance
result_queryset = []
for store_id in store_list:
    min_value = Goods.objects.filter(store_id=store_id).aggregate(min_value=Min('total_cost'))
    result_queryset = result_queryset|Goods.objects.filter(store_id=store_id, total_cost=min_value)

UPDATE-2

我认为我的Update-1 部分存在大量性能问题，因此我为您的问题找到了一个可能的答案，即

goods_queryset = Goods.objects.filter(**you_possible_filters)
result = goods_queryset.filter(store_id__in=[good['store'] for good in Goods.objects.values('store').annotate(min_val=Min('total_cost'))])