Django嵌套回复序列化：如何获取子回复的用户Profile数据

我有一个支持无限层级嵌套的Reply模型，查看回复时父回复能正常显示创建者的Profile，但子回复只显示user ID，无法获取对应的Profile数据。

模型代码 #

class Reply(models.Model):
    user = models.ForeignKey(User, on_delete=models.CASCADE)
    post = models.ForeignKey(Post, on_delete=models.CASCADE)
    video = models.FileField(
        upload_to=reply_videos_directory_path, null=True, blank=True
    )
    body = models.TextField(max_length=256, default=None)
    parent = models.ForeignKey(
        "self", on_delete=models.CASCADE, null=True, blank=True, related_name="replies"
    )

    created_at = models.DateTimeField(auto_now_add=True, verbose_name="Created at")
    updated_at = models.DateTimeField(auto_now=True, verbose_name="Updated at")

    class Meta:
        verbose_name = "post reply"
        verbose_name_plural = "post replies"
        db_table = "post_reply"

    def __str__(self):
        return self.body[0:30]

当前序列化器代码 #

class ReplySerializer(serializers.ModelSerializer):
    images = ReplyImageSerializer(many=True, read_only=True, required=False)
    post = serializers.SerializerMethodField()
    profile = serializers.SerializerMethodField()
    post_images = serializers.SerializerMethodField()

    class Meta:
        model = Reply
        fields = [
            "id",
            "post",
            "post_images",
            "video",
            "images",
            "body",
            "parent",
            "replies",
            "profile",
            "created_at",
            "updated_at",
        ]
        depth = 1

    def get_post(self, obj):
        post_obj = Post.objects.get(id=obj.post.id)
        post = ShortPostSerializer(post_obj)
        return post.data

    def get_profile(self, obj):
        profile_obj = Profile.objects.get(id=obj.user.profile.id)
        profile = ShortProfileSerializer(profile_obj)
        return profile.data

    def get_post_images(self, obj):
        images = PostImage.objects.filter(post=obj.post.id)
        serializer = PostImageSerializer(images, many=True)
        return serializer.data     
    
    def create(self, validated_data):
        new_reply = Reply.objects.create(**validated_data)
        images = dict((self.context['request'].FILES).lists()).get('images', None)
        if images:
            for image in images:
                ReplyImage.objects.create(
                    image=image, reply=new_reply
                )
        return new_reply

疑问解答与解决方案 #

1. 为什么子回复不显示Profile？

你设置了depth=1，这个参数会让DRF自动序列化关联的replies字段，但它用的是默认的ModelSerializer，而非你自定义的ReplySerializer。默认序列化器只会处理模型自带的字段（比如user的ID），不会包含你手动添加的profile这个SerializerMethodField。

2. 子回复当前由什么序列化？

子回复是DRF通过depth=1自动生成的默认序列化器处理的，不是你写的ReplySerializer，所以自定义字段不会生效。

3. 更优实现方式

去掉depth=1，显式地在序列化器中用自身递归序列化子回复，同时优化查询避免N+1问题：

修改后的序列化器

class ReplySerializer(serializers.ModelSerializer):
    images = ReplyImageSerializer(many=True, read_only=True, required=False)
    post = serializers.SerializerMethodField()
    profile = serializers.SerializerMethodField()
    post_images = serializers.SerializerMethodField()
    # 显式指定子回复用当前序列化器处理，传递上下文保证request等参数可用
    replies = serializers.SerializerMethodField()

    class Meta:
        model = Reply
        fields = [
            "id",
            "post",
            "post_images",
            "video",
            "images",
            "body",
            "parent",
            "replies",
            "profile",
            "created_at",
            "updated_at",
        ]
        # 移除depth=1，手动控制嵌套逻辑

    def get_replies(self, obj):
        # 递归序列化所有子回复，传递当前上下文
        return ReplySerializer(obj.replies.all(), many=True, context=self.context).data

    def get_post(self, obj):
        post_obj = Post.objects.get(id=obj.post.id)
        post = ShortPostSerializer(post_obj)
        return post.data

    def get_profile(self, obj):
        profile_obj = Profile.objects.get(id=obj.user.profile.id)
        profile = ShortProfileSerializer(profile_obj)
        return profile.data

    def get_post_images(self, obj):
        images = PostImage.objects.filter(post=obj.post.id)
        serializer = PostImageSerializer(images, many=True)
        return serializer.data     
    
    def create(self, validated_data):
        new_reply = Reply.objects.create(**validated_data)
        images = dict((self.context['request'].FILES).lists()).get('images', None)
        if images:
            for image in images:
                ReplyImage.objects.create(
                    image=image, reply=new_reply
                )
        return new_reply

优化查询（视图层）

为了避免递归查询带来的N+1性能问题，在获取回复时提前预加载关联数据：

from django.db.models import Prefetch

# 获取根回复（parent为None）时，预加载user的profile、子回复的user.profile，以及图片
replies = Reply.objects.filter(post=post_id, parent=None)\
    .select_related('user__profile')\
    .prefetch_related(
        Prefetch('replies', queryset=Reply.objects.select_related('user__profile')),
        'images',
        'replies__images'
    )

这样所有层级的回复都会使用同一个ReplySerializer处理，子回复也能正常显示Profile数据，同时查询性能也得到了优化。

内容的提问来源于stack exchange，提问作者crawlingdev