问题

I'm trying to query the users presents in a certain conversation. I finally got it working with this annotation:

我正在尝试查询某个对话中出现的用户。我最终通过以下注释使其工作：

products = products.annotate(
    conversation_users=Subquery(
        Conversation.objects.filter(
            product=OuterRef('pk')
        ).annotate(
            users_array=ArrayAgg('users__username')
        ).values_list('users_array', flat=True)[:1]
    )
)

This is returning a list with the user present in the conversation (in most cases only two). I need to exclude from there the request.user so I get always the value of the user who I am having the conversation with.

这返回一个包含在对话中的用户列表（在大多数情况下只有两个）。我需要从中排除request.user，以便始终获取我正在进行对话的用户的值。

I tried using this Q query trying to exclude the request.user:

我尝试使用以下的 Q 查询来排除request.user：

.filter(~Q(users=user)).values_list('users_array', flat=True)[:1]

but it is making the field now return None. How can I accomplish that?

但它导致该字段现在返回 None。我该如何实现这一点？

Edit 1:

Those are the relevant models, Product model:

以下是相关的模型，Product 模型：

class Product(models.Model):
    creator = models.ForeignKey(
        User, on_delete=models.CASCADE, related_name='anken')
    content = models.TextField(blank=True, null=True)
    date = models.DateTimeField(auto_now_add=True)
    active = models.BooleanField(default=True)
    intouch = models.ManyToManyField(
        User, related_name='product_intouch', blank=True)

And this is the Conversation model:

这是 Conversation 模型：

class Conversation(models.Model):
    product = models.ForeignKey(
        Product, on_delete=models.CASCADE, related_name='conversations')
    users = models.ManyToManyField(User, related_name='conversations')
    
    def validate_unique(self, *args, **kwargs):
        super().validate_unique(*args, **kwargs)
        if Conversation.objects.filter(product=self.product, users__in=self.users.all()).exists():
            raise ValidationError(
                'A conversation with the same Product and user already exists.')

Is there anything else you need assistance with?

    products = products.annotate(
        conversation_users=Subquery(
            Conversation.objects.filter(
                product=OuterRef('pk')
            ).annotate(
                users_array=ArrayAgg('users__username')
            ).values_list('users_array', flat=True)[:1]
        )
    )

.filter(~Q(users=user)).values_list('users_array', flat=True)[:1]
        )

class Product(models.Model):
    creator = models.ForeignKey(
        User, on_delete=models.CASCADE, related_name='anken')
    content = models.TextField(blank=True, null=True)
    date = models.DateTimeField(auto_now_add=True)
    active = models.BooleanField(default=True)
    intouch = models.ManyToManyField(
        User, related_name='product_intouch', blank=True)

class Conversation(models.Model):
    product = models.ForeignKey(
        Product, on_delete=models.CASCADE, related_name='conversations')
    users = models.ManyToManyField(User, related_name='conversations')
    

    def validate_unique(self, *args, **kwargs):
        super().validate_unique(*args, **kwargs)
        if Conversation.objects.filter(product=self.product, users__in=self.users.all()).exists():
            raise ValidationError(
                'A conversation with the same Product and user already exists.')

答案1

得分: 1

你可以在聚合函数中过滤掉这部分内容：

from django.contrib.postgres.aggregates import ArrayAgg
from django.db.models import OuterRef, Q, Subquery

products = products.annotate(
    conversation_users=Subquery(
        Conversation.objects.filter(
            product=OuterRef('pk'),
        )
        .annotate(users_array=ArrayAgg('users__username', filter=~Q(users=user)))
        .values_list('users_array', flat=True)[:1]
    )
)