Question

我是Flask SQl炼金术的新手;虽然我明白炼金术抽象了sql语法并且在创建模型时简单易行;有时候我们想要以非常具体的方式可视化前端的数据。

我有以下查询，我想使用alchemy使用session.query和过滤器以及可能的分组。

我的查询是：

MySQL的＆GT; SELECT status, COUNT(id) FROM bar_baz where not name = 'Foo' and not name = 'Bar' GROUP BY status
select (select COUNT(id) FROM instance where not name = 'erf' and not tiername = 'wer' and type='app') as app, (select COUNT(1) FROM instance_2 where not name = 'visq' and not name = 'werf' and type='adc') as adc from dual;

我验证了以下查询适用于MySQL;我想知道我们是否有类似于

c = conn.cursor()
query = 'SELECT status, COUNT(id) FROM bar_baz where not name = 'Foo' and not name = 'Bar' GROUP BY status'
c.execute(query)
print c.fetchall()


class Instance(Base):
    __tablename__ = 'instance'

    id = Column(Integer, primary_key=True)
    name = Column(String)
    status = Column(String)
    type = Column(String)

class Instance2(Base):
    __tablename__ = 'instance_2'

    id = Column(Integer, primary_key=True)
    name = Column(String)
    status = Column(String)
    type = Column(String)
    inc = Column(Integer)

感兴趣的查询：

select (select COUNT(id) FROM instance where not name = 'erf' and not tiername = 'wer' and type='app') as app, (select COUNT(1) FROM instance_2 where not name = 'visq' and not name = 'werf' and type='adc') as adc from dual;`

Answer 1

对于第一个查询，使用db.func.count生成计数表达式。其他所有内容都应该来自docs。

status_counts = db.session.query(BarBaz.status, db.func.count(BarBaz.id).label('count_id')
).filter(db.not_(db.or_(BarBaz.name == 'Foo', BarBaz.name == 'Bar'))
).group_by(BarBaz.status
).all()

对于第二个查询，请使用subquery()生成可选择的查询。

sub_app = db.session.query(db.func.count(Instance.id).label('app')
).filter(db.not_(db.or_(Instance.name == 'erf', Instance.tiername == 'wer')), Instance.type == 'app'
).subquery()

sub_adc = db.session.query(db.func.count(Instance.id).label('adc')
).filter(db.not_(db.or_(Instance2.name == 'visq', Instance2.name == 'werf')), Instance2.type == 'adc'
).subquery()

out = db.session.query(sub_app.c.app, sub_adc.c.adc).all()

将SQL转换为SQL炼金术

1 个答案: