我正在尝试将数据插入到具有映射到另一个表的ID的表中,并且我希望它创建我的关系所需的数据...问题是在Parent中,我有另一种关系,并且孩子,我也与父母的孩子有关...见下文:
$ sed -e 's/^\([^ ]*\)[ ]\(.*$\)/\1:\2/' <file
2016/02/11:12:54:28 0071754407 7599 4727 2690
2016/02/11:14:07:41 0071754407 7599 4726 2690
问题是,当我使用工厂插入Parent时,显然它会尝试插入childb表两次,因为它在parent和childa中被引用,因为我一直在获取childb_id的重复键...我知道如何防止这种情况发生?
我创建工厂的方法是创建一个Parent对象并提交会话:
class Parent(Base):
parent_field = Column(Integer)
childa_id = Column(Integer, ForeignKey(ChildATable.childa_id))
childb_id = Column(Integer, ForeignKey(ChildBTable.childb_id))
childa = relationship(ChildATable)
childb = relationship(ChildBTable)
class ChildA(Base):
childa_id = Column(Integer, primary_key=True)
childb_id = Column(Integer, ForeignKey(ChildB.childb_id))
class ChildB(Base):
childb_id = Column(Integer, primary_key=True)
class ParentTable(factory.alchemy.SQLAlchemyModelFactory)
class Meta:
....
parent_field = factory.Sequence(lambda n: n + 1)
childa = factory.SubFactory(ChildA)
childb = factory.SubFactory(ChildB)
class ChildATable(factory.alchemy.SQLAlchemyModelFactory)
class Meta:
....
childa_id = factory.Sequence(lambda n: n + 1)
childb = factory.SubFactory(ChildB)
class ChildBTable(factory.alchemy.SQLAlchemyModelFactory)
class Meta:
....
childb_id = 1 #I need this to be hardcoded to 1
答案 0 :(得分:1)
在幕后,您对ParentTable()
的通话将转换为:
ChildATable()
ChildATable()
内,请致电ChildBTable()
ChildBTable()
,这次是ParentTable()
如果您希望两家工厂使用相同的ChildB()
对象,您可以按如下方式更改ParentTable()
工厂:
class ParentTable(factory.alchemy.SQLAlchemyModelFactory)
class Meta:
....
parent_field = factory.Sequence(lambda n: n + 1)
childa = factory.SubFactory(ChildA,
# Forward our `childb` to the `ChildA()` factory
childb=factory.SelfAttribute('..childb'),
)
childb = factory.SubFactory(ChildB)
答案 1 :(得分:0)
我怀疑你在这两个子事件中获得n
的相同数字。你能尝试这样的东西吗?
class ParentTable(factory.alchemy.SQLAlchemyModelFactory)
class Meta:
....
parent_field = factory.Sequence(lambda n: 2*n + 1)
childa = factory.SubFactory(ChildA)
childb = factory.SubFactory(ChildB)