序列化器内部的django-rest-framework访问字段

debugcn 发表于 Dev

宇宙驯鹿

所以我有一个这样的模型

class DataSheet(BaseModel):
    """
    Represents a single dataSheet.
    dataSheets have their own model at the core. Model data is added to
    the dataSheets in the form of separate records.
    """

    class Meta:
        verbose_name = 'datasheet'
        verbose_name_plural = 'datasheets'
        ordering = ['position', 'cluster']
        required_db_features = {
            'supports_deferrable_unique_constraints',
        }
        constraints = [
            models.UniqueConstraint(
                fields=['position', 'cluster'],
                name='deferrable_unique_datasheet_position',
                deferrable=models.Deferrable.DEFERRED
            )
        ]

    def __str__(self):
        return self.name

    objects = managers.DataSheetsManager()
    positions = managers.PositionalManager()
    position = models.PositiveSmallIntegerField(db_index=True, editable=True)
    name = models.CharField(max_length=100, validators=[MinLengthValidator(2)], db_index=True)
    description = models.CharField(max_length=1024, null=True, blank=True, db_index=True)
    owner = models.ForeignKey('api_backend.Member', on_delete=models.CASCADE, db_index=True, editable=False)
    fields = models.ManyToManyField('api_backend.Field')
    overwrites = models.ManyToManyField('api_backend.RoleOverwrite')
    parent = models.ForeignKey('api_backend.Category', on_delete=models.CASCADE, null=True, blank=True)
    cluster = models.ForeignKey('api_backend.Cluster', on_delete=models.CASCADE, editable=False)

    REQUIRED_FIELDS = [name, owner, cluster]

和这样的序列化器

class DataSheetSerializer(serializers.ModelSerializer):
    """
    A serialized DataSheet Object.
    Datasheets have their own:
        - array of fields
        - array of role-overwrites
    """

    def get_fields(self):
        fields = super(DataSheetSerializer, self).get_fields()
        fields['parent'].queryset = self.cluster.categories.all()

        return fields

    class Meta:
        model = DataSheet
        read_only_fields = ['position']
        fields = '__all__'

    # need to make sure that the parent category of the datasheet
    # belongs to the datasheet's cluster only.

    fields = partial.PartialFieldSerializer(many=True, read_only=True)
    overwrites = partial.PartialOverWriteSerializer(many=True, read_only=True)

问题是，我想cluster在get_fields方法内部访问序列化器模型的字段。但是，我不能做同样的事情。有人能帮我吗？

我看到了其他涉及initial_data的答案，但这在这里不起作用。

fields['parent'].queryset = self.cluster.categories.all()

cluster是此处未解决的参考。

阿什温地带

self在get_fields为DataSheetSerializer实例不是DataSheet模型实例。因此它不应该具有群集属性。您无法访问模型DataSheet实例，get_fields因为它从类DataSheet而不是从其实例获取字段。您可以像验证字段

class DataSheetSerializer(serializers.ModelSerializer):
    # ... other code

    def validate(self, data):
         parent = data.get('parent')
         # check if parent is valid i.e in queryset
         # if yes return data
         # else raise serializers.validationError

本文收集自互联网，转载请注明来源。

如有侵权，请联系[email protected] 删除。