Python 如何在外部连接sqlalchemy中添加group by

我在sqlalchemy中加入同一张表3次。我想要分组依据

ecs2.id

如果

ecs2.school\u class\u id==None

（分组依据以防返回任何结果）

我已按“ecs2.id”添加了分组，但如果ecs2.id为None，则它将分组。这意味着它不返回任何结果，因此任何连接

ecs2

的结果都是空的，那么只返回一个结果

此查询工作正常，但有时会在重复的

ecs.id

或

ecs.school\u class\u id

ecs1 = aliased(EmployeeClassSubject)
ecs2 = aliased(EmployeeClassSubject)
ecs3 = aliased(EmployeeClassSubject)
sc1 = aliased(SchoolClass)
sc2 = aliased(SchoolClass)
e1 = aliased(Employee)
e2 = aliased(Employee)
emp = db.session.query(e1, School.name, ecs1, ecs2, sc1, sc2, Subject.id, 
                         Subject.name,BusInf,ecs3.employee_id,e2.name).\
    join(School, School.id == e1.school_id).\
    outerjoin(sc1, and_(sc1.school_id == School.id,
                         sc1.status == GenEnum.ACTIVE.value,
                         e1.employee_type.in_(super_users))).\
    outerjoin(ecs1, and_(ecs1.school_class_id == sc1.id,
                         ecs1.status == GenEnum.ACTIVE.value,
                         ecs1.relation == OtherType.CLASS_TEACHER.value)).\
    outerjoin(e2, e2.id == ecs1.employee_id).\
    outerjoin(ecs2, and_(ecs2.employee_id == e1.id,
                         ecs2.status == GenEnum.ACTIVE.value,
                         e1.employee_type.notin_(super_users))).\
    outerjoin(sc2, and_(sc2.id == ecs2.school_class_id,
                         sc2.status == GenEnum.ACTIVE.value)).\
    outerjoin(ecs3, and_(ecs3.school_class_id == ecs2.school_class_id,
                         ecs3.status == GenEnum.ACTIVE.value,
                         ecs3.relation == OtherType.CLASS_TEACHER.value)).\
    outerjoin(Subject, and_(Subject.id == ecs2.subject_id,
                         ecs2.subject_id != None)).\
    outerjoin(BusInf, and_(BusInf.contact_person == e1.id,
                         e1.employee_type == UserType.SUPPORTINGSTAFF.value)).\
    filter(e1.id == login_employee[2].id).\
    order_by(sc1.standard, sc1.division, sc2.standard, sc2.division).all()

当我添加groupby函数时

ecs1 = aliased(EmployeeClassSubject)
ecs2 = aliased(EmployeeClassSubject)
ecs3 = aliased(EmployeeClassSubject)
sc1 = aliased(SchoolClass)
sc2 = aliased(SchoolClass)
e1 = aliased(Employee)
e2 = aliased(Employee)
emp = db.session.query(e1, School.name, ecs1, ecs2, sc1, sc2, Subject.id, 
                         Subject.name,BusInf,ecs3.employee_id,e2.name).\
    join(School, School.id == e1.school_id).\
    outerjoin(sc1, and_(sc1.school_id == School.id,
                         sc1.status == GenEnum.ACTIVE.value,
                         e1.employee_type.in_(super_users))).\
    outerjoin(ecs1, and_(ecs1.school_class_id == sc1.id,
                         ecs1.status == GenEnum.ACTIVE.value,
                         ecs1.relation == OtherType.CLASS_TEACHER.value)).\
    outerjoin(e2, e2.id == ecs1.employee_id).\
    outerjoin(ecs2, and_(ecs2.employee_id == e1.id,
                         ecs2.status == GenEnum.ACTIVE.value,
                         e1.employee_type.notin_(super_users))).\
    outerjoin(sc2, and_(sc2.id == ecs2.school_class_id,
                         sc2.status == GenEnum.ACTIVE.value)).\
    outerjoin(ecs3, and_(ecs3.school_class_id == ecs2.school_class_id,
                         ecs3.status == GenEnum.ACTIVE.value,
                         ecs3.relation == OtherType.CLASS_TEACHER.value)).\
    outerjoin(Subject, and_(Subject.id == ecs2.subject_id,
                         ecs2.subject_id != None)).\
    outerjoin(BusInf, and_(BusInf.contact_person == e1.id,
                         e1.employee_type == UserType.SUPPORTINGSTAFF.value)).\
    filter(e1.id == login_employee[2].id).\
    group_by(ecs2.id).\
    order_by(sc1.standard, sc1.division, sc2.standard, sc2.division).all()

当ecs2不是None时，这可以正常工作，但在其他情况下，它只返回一个结果

在sqlalchemy中，有没有任何方法可以使组处于条件下