Django中创建对象后关联外键字段的实现问题

我需要处理包含description和choices字段的JSON输入，创建Question及关联的Choice记录。Choice模型通过外键关联Question，关联名称为choices，但JSON里的Choice没有question_id字段，必须在Question创建完成后再关联。尝试过程中遇到了Direct assignment to the forward side、Direct assignment to the reverse side of a related set is prohibited等错误，且不能修改关联名称和JSON输入字段。

原序列化器代码：

class CreateQuestionSerializer(serializers.ModelSerializer):
    description = serializers.CharField(max_length=1024, allow_blank=False, required=True)
    choices = CreateChoiceSerializer(many=True, required=True)

    def create(self, validated_data):
        choices = validated_data.pop('choices')

        question = Question.objects.create(**validated_data)

        choices_serializer = CreateChoiceSerializer(data=choices, many=True, context={'question': question})
            if choices_serializer.is_valid():
                choices_serializer.save()
        
        return question


class CreateChoiceSerializer(serializers.ModelSerializer):
    description = serializers.CharField(max_length=1024, allow_blank=False, required=True, allow_null=False)

    class Meta:
        model = Choice
        fields = ['description']

    def create(self, validated_data):
        question = self.context.get('question')
        choice = Choice.objects.create(question=question, **validated_data)
        return choice

原模型代码：

class Question(models.Model):
    id = models.AutoField(primary_key=True)
    description = models.CharField(max_length=1024, blank=False, null=False)


class Choice(models.Model):
    id = models.AutoField(primary_key=True)
    question = models.ForeignKey(Question,
                                 related_name='choices',
                                 on_delete=models.PROTECT)
    description = models.CharField(max_length=1024, blank=False, null=False)

1. 原代码里CreateQuestionSerializer的create方法存在缩进错误（if choices_serializer.is_valid():这一行缩进不对），导致代码逻辑异常。
2. 使用嵌套序列化器批量保存时，ModelSerializer的批量处理会触发反向关联集合的直接赋值限制，从而抛出Direct assignment to the reverse side of a related set is prohibited错误。

方案一：用bulk_create直接批量创建（推荐） #

这种方式跳过嵌套序列化器的复杂处理，直接通过Question实例批量创建关联的Choice，代码更简洁高效：

class CreateQuestionSerializer(serializers.ModelSerializer):
    description = serializers.CharField(max_length=1024, allow_blank=False, required=True)
    choices = CreateChoiceSerializer(many=True, required=True)

    class Meta:
        model = Question
        fields = ['description', 'choices']

    def create(self, validated_data):
        choices_data = validated_data.pop('choices')
        question = Question.objects.create(**validated_data)
        
        # 批量创建Choice，直接关联已创建的Question
        Choice.objects.bulk_create(
            [Choice(question=question, **data) for data in choices_data]
        )
        
        return question


class CreateChoiceSerializer(serializers.ModelSerializer):
    description = serializers.CharField(max_length=1024, allow_blank=False, required=True, allow_null=False)

    class Meta:
        model = Choice
        fields = ['description']

方案二：保留嵌套序列化器，手动逐个创建 #

如果一定要用嵌套序列化器处理，可以手动遍历每个Choice数据，逐个验证保存，避免批量处理的问题：

class CreateQuestionSerializer(serializers.ModelSerializer):
    description = serializers.CharField(max_length=1024, allow_blank=False, required=True)
    choices = CreateChoiceSerializer(many=True, required=True)

    class Meta:
        model = Question
        fields = ['description', 'choices']

    def create(self, validated_data):
        choices_data = validated_data.pop('choices')
        question = Question.objects.create(**validated_data)
        
        # 逐个处理Choice数据，确保每个都正确关联Question
        for data in choices_data:
            serializer = CreateChoiceSerializer(data=data, context={'question': question})
            serializer.is_valid(raise_exception=True)
            serializer.save()
        
        return question


class CreateChoiceSerializer(serializers.ModelSerializer):
    description = serializers.CharField(max_length=1024, allow_blank=False, required=True, allow_null=False)

    class Meta:
        model = Choice
        fields = ['description']

    def create(self, validated_data):
        question = self.context.get('question')
        return Choice.objects.create(question=question, **validated_data)

• 方案一的bulk_create性能更好，适合数据量较大的场景，但不会触发Choice模型的save信号，如果有信号逻辑需要处理，建议用方案二。
• 必须确保Question实例创建完成后，再去创建关联的Choice，避免外键关联失败。

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