组合两个查询集，公共字段具有不同的值

Question

我在模型“ConsultantProfile”中搜索，相关性由“queryset1”中名为“rank”的字段表示

我还搜索相关模型，从“queryset2”中的相关模型中收集“ConsultantProfile”对象，并手动将它们设置为“0.7”，以便我可以合并查询集

当我想要组合和合并queryset1和queryset2时，公共对象具有不同的秩值。问题是在最终的“queryset”中，“rank”字段的值是来自第一个queryset的rank值，我无法为“rank”选择最高值

queryset1 = qs.filter(active=True).annotate(
    rank=SearchRank(
        vector,
        query,
        normalization=Value(2).bitor(Value(4)),
    )
)

ids: some "ConsultantProfile" objects ids

queryset2 = ConsultantProfile.objects.filter(id__in=ids, active=True).annotate(rank=Value(5, FloatField()))

queryset = queryset1 | queryset2

失败的解决方案 我试图通过手动迭代查询集和联合来解决这个问题，但没有任何改变

queryset = queryset1.none()

for obj in queryset1:
    qs2_object_qs = queryset2.filter(id=obj.id)

    # object presents in queryset2
    if qs2_object_qs.exists():

        # obj from queryset1 has better rank than queryset2 obj
        if obj.rank >= qs2_object_qs.first().rank:k)
            queryset |= queryset1.filter(id=obj.id)
            queryset2 = queryset2.exclude(id=obj.id)
        else:
            queryset |= qs2_object_qs

    # object does not present in queryset2
    else:
        queryset |= queryset1.filter(id=obj.id)

for obj in queryset2:
    qs_object_qs = queryset.filter(id=obj.id)

    # object presents in queryset
    if qs_object_qs.exists():
        # because compared mutual objects in previous loop and removed
        pass
    else:
        queryset |= queryset2.filter(id=obj.id)

Answer 1

应在“|”操作数旁边使用union（）。 在union方法中，重复的对象将不会删除。但在order（'-rank'）distinct（）之后将不起作用

我所做的事情是在“|”操作数结果中找到了公共对象，并从其查询集中删除了秩较低的对象，然后使用union函数

temp_queryset = queryset1 | queryset2

for obj in temp_queryset:
    try:
        qs1_object = queryset1.get(id=obj.id)
        qs2_object = queryset2.get(id=obj.id)
        if qs1_object.rank > qs2_object.rank:
            queryset2 = queryset2.exclude(id=obj.id)
        else:
            queryset1 = queryset1.exclude(id=obj.id)
    except ConsultantProfile.DoesNotExist:
        pass

queryset = queryset1.union(queryset2)

queryset = queryset.order_by('-rank')