Question

我有一个SLQALchemy对象，我使用marshmallow进行序列化。

该对象有N个赞和N个评论。它看起来像这样：

class Parent():

    __tablename__ = 'parent'

    title = Column(db.String(100), unique=True, nullable=False)
    description = Column(db.String(500))
    created_at = Column(db.DateTime, nullable=False, default=dt.datetime.utcnow)
    comments = relationship('Comment')
    likes = relationship('Like')

序列化器看起来像这样：

class CommentSerializer(Serializer):
    class Meta:
        fields = ('id', 'text', 'created_at', 'parent_id')

class LikeSerializer(Serializer):
    class Meta:
        fields = ('id', 'created_at', 'parent_id')

class ParentSerializer(Serializer):
    comments = fields.Nested(CommentSerializer)
    likes = fields.Nested(LikeSerializer)

    class Meta:
        fields = ('id', 'title', 'description', 'comments', 'likes')

我尝试在视图中运行它：

allParents = Parent.query.all()

用这个把它变成JSON：

return jsonify({"parents": ParentSerializer(allParents, many=True).data})

当我尝试运行时，我收到错误list indices must be integers, not str。它来自marshmallow/serializer.py。当我在那里记录一些东西时，似乎marshmallow正试图访问text列表的<Comment>属性。它应该分别访问每个<Comment>，然后访问text属性。

我在序列化程序中遗漏了什么？我知道在ParentSerializer中发送many=True参数告诉marshmallow它应该遍历<Parent>列表。有没有办法告诉棉花糖它还应该期待很多<Comment>或<Like>？

Answer 1

有没有办法告诉marshmallow它还应该期待很多<Comment>或<Like>？

是。您也可以将many=True传递给Nested字段。

class ParentSerializer(Serializer):
    comments = fields.Nested(CommentSerializer, many=True)
    likes = fields.Nested(LikeSerializer, many=True)

Answer 2

有没有办法告诉棉花糖它也应该期望许多<Comment>或<Like>？

或者，您可以执行以下操作：

class ParentSerializer(Serializer):
    comments = fields.List(fields.Nested(CommentSerializer))
    likes = fields.List(fields.Nested(LikeSerializer))

Answer 3

选中https://docs.sqlalchemy.org/en/13/orm/basic_relationships.html#association-object
您可以使用自定义序列化功能。

 class Company(db.Model):
  __tablename__ = 'company'

  id = db.Column(UUID(as_uuid=True), primary_key=True, default=uuid.uuid4)
  name = db.Column(db.String(255), nullable=False)
  users = relationship("CompanyUser", back_populates="company")

  def serialize(self):
   main =  { c.key: getattr(self, c.key) for c in inspect(self).mapper.column_attrs }
   children = [user.serialize() for user in self.users] 
   main['users'] = children
   return  main

并且，在其他模型类中，您可以使用这种“通用”可序列化的对象：

 def serialize(self):
    return { c.key: getattr(self, c.key) for c in inspect(self).mapper.column_attrs }

序列化具有多个子项的子项的SQLAlchemy对象

3 个答案: