#64 SQLAlchemy
Python ORM PythonORM SQLAlchemy 2014-08-21数据库连接：Engine
创建 engine 相当于通过适配层对接了原生数据库接口。当 excute 方法和 connect 方法首次调用时，Engine 建立了一个正式的 DBAPI 连接到数据库，之后的所有数据库交互都是通过这个连接发出。
使用 ORM 时，Engine 是个幕后工作者，也就是说一旦创建之后，我们一般不会再直接接触到 Engine。
from sqlalchemy import create_engine
engine = create_engine('sqlite:///:memory:', echo=True) # echo => logging
连接字符串（DB URL）
会话（Session）
http://sqlalchemy.readthedocs.org/en/latest/orm/session_basics.html
Session 是一个基于指定数据库连接的一个工作区。文档上有这么一个例子：如果将一个应用程序线程当作一场派对中的一个来宾，会话就是这个来宾手上的盘子，数据就是盘子中的食物，数据库就是厨房。
何时创建，何时提交，何时关闭？
Session = sessionmaker(bind=engine)
# 等同于：
# Session = sessionmaker()
# Session.configure(bind=engine)
session = Session()
定义模型（数据库映射）
基类
from sqlalchemy.ext.declarative import declarative_base
Base = declarative_base()
模型的定义
from sqlalchemy import Column, Integer, String
class User(Base):
__tablename__ = 'users'
id = Column(Integer, primary_key=True)
name = Column(String)
fullname = Column(String)
password = Column(String)
def __repr__(self):
return "<User(name='%s', fullna
表结构（Schema）
又出来一个幕后工作者：Mapper，通过 模型类.__table__ 访问。
print type(User.__table__) # <class 'sqlalchemy.sql.schema.Table'>
print repr(User.__table__)
# Table('users', MetaData(bind=None),
# Column('id', Integer(), table=<users>, primary_key=True, nullable=False),
# Column('name', String(), table=<users>),
# Column('fullname', String(), table=<users>),
# Column('password', String(), table=<users>), schema=None)
数据库操作
1. 添加
u = User(name='ed', fullname='Ed Jones', password='password')
session.add(u)
# 一次添加多条记录
session.add_all([
User(name='wendy', fullname='Wendy Williams', password='foobar'),
User(name='mary', fullname='Mary Contrary', password='p@55vv0rd'),
User(name='fred', fullname='Fred Green', password='123456')
])
print session.new
print session.query(User).all()
2. 查找
2.1 查询对象 Query
for name, fullname in session.query(User.name, User.fullname):
print name.ljust(10), fullname
for row in session.query(User, User.name):
print row.User, row.name
直接遍历 Query 对象就相当于按照默认顺序获取所有对象。
2.2 条件：filte_by
qs = session.query(User).filter_by(name='ed')
users = qs.all()
our_user = qs.first()
2.3 更加灵活的 filter
允许使用 Python 表达式和类属性！
query = session.query(User.name)
queryset = query.filter(User.fullname == 'Ed Jones')
for name, in queryset:
print name
通用 Filter 操作符
- Equal
query.filter(User.name == 'ed')
query.filter(User.name != 'ed') - LIKE
query.filter(User.name.like('%ed%')) - IN / NOT IN
query.filter(User.name.in_(['ed', 'wendy', 'jack']))
query.filter(User.name.in_(session.query(User.name).filter(User.name.like('%ed%'))))
query.filter(~User.name.in_(['ed', 'wendy', 'jack'])) - NULL
query.filter(User.name == None)
query.filter(User.name.is_(None))
query.filter(User.name != None)
query.filter(User.name.isnot(None)) - AND
query.filter(User.name == 'ed').filter(User.fullname == 'Ed Jones')
query.filter(and_(User.name == 'ed', User.fullname == 'Ed Jones')) # from sqlalchemy import and_ - OR
query.filter(User.name == 'ed', User.fullname == 'Ed Jones')
query.filter(or_(User.name == 'ed', User.name == 'wendy')) # from sqlalchemy import or_ - MATCH:
query.filter(User.name.match('wendy'))
注意：match 使用了数据库相关的内容，可能在不同数据库后端上有不同的效果，而且部分数据库不支持。
2.4 排序：order_by
for instance in session.query(User).order_by(User.id):
print instance.name.ljust(10), instance.fullname
2.5 别名
label = User.name.label('name_label')
for row in session.query(label).all():
print(row.name_label)
from sqlalchemy.orm import aliased
user_alias = aliased(User, name='user_alias')
for row in session.query(user_alias, user_alias.name).all():
print row.user_alias
2.6 LIMIT 和 OFFSET
和 Django 一样，使用切片实现。
2.7 其他方法
.all()返回查找到的所有纪录组成的列表。.first()返回查找到的第一条记录，没有找到返回 None。.one()只查找到一条记录时，返回该条记录。否则抛出异常：
sqlalchemy.orm.exc.NoResultFound: No row was found for one()
sqlalchemy.orm.exc.MultipleResultsFound: Multiple rows were found for one().scalar()用.one()方法模拟出.first()方法的效果。不同的是，.first()方法只获取那一条（LIMIT 1），而调用.one()方法却获取了所有记录，不过只返回了第一条。这是不一样的！
2.8 对原生 SQL 的支持
- text
- from_statement
- params 为 Query 增加参数
from sqlalchemy import text
# 例 1，简单用法
for user in session.query(User).filter(text("id<224")).order_by(text("id")).all():
print user.name
# 例 2，定义参数
session.query(User).filter(text("id<:value and name=:name")).params(value=224, name='fred').order_by(User.id).one()
# 例 3，此时 Query 几乎只检测字段是否合法～
sql = text("SELECT * FROM users where name=:name")
session.query(User).from_statement(sql).params(name='ed').all()
session.query("id", "name").from_statement(sql).params(name='ed').all()
3. 更新
u = session.query(User).filter_by(name='fred')[0]
u.password = '12345678'
session.commit()
print u in session
u = session.query(User).filter_by(name='fred')[0]
print u.password
4. 删除
jack = session.query(User).filter_by(name='jack').one()
session.delete(jack)
session.query(User).filter_by(name='jack').count()
5. Count
session.query(User).filter(User.name.like('%ed')).count()
from sqlalchemy import func
session.query(func.count(User.name), User.name).group_by(User.name).all()
session.query(func.count('*')).select_from(User).scalar()
session.query(func.count(User.id)).scalar()
6. 事务：commit 和 rollback
commit
u = session.query(User).all()[-1]
u.password = '123456'
print session.dirty # IdentitySet
# 此时 add_all 的新增操作实际上还没有提交
# session.commit() # 提交到数据库
rollback
u = session.query(User).all()[-1]
u.password = '123456'
session.rollback() # 撤销更新
print session.dirty
print 'After Rollback, PWD:'
print u.password # 会先查找一遍～
关系
from sqlalchemy import ForeignKey
from sqlalchemy.orm import relationship, backref
class Address(Base):
__tablename__ = 'addresses'
id = Column(Integer, primary_key=True)
email_address = Column(String, nullable=False)
user_id = Column(Integer, ForeignKey('users.id'))
user = relationship("User", backref=backref('addresses', order_by=id))
def __repr__(self):
return "<Address(email_address='%s')>" % self.email_address
# 或者到 User 中声明，如下：
# class User(Base):
# addresses = relationship("Address", order_by="Address.id", backref="user")
关联操作
jack = User(name='jack', fullname='Jack Bean', password='qwerty')
print jack.addresses
jack.addresses = [
Address(email_address='jack@google.com'),
Address(email_address='j25@yahoo.com'),
]
print jack.addresses[1]
print jack.addresses[1].user
session.add(jack)
print '-' * 70
session.commit() # 也会插入 Address 信息
print '-' * 70
jack = session.query(User).filter_by(name='jack').one()
print jack
print '-' * 70
print jack.addresses # lazy loading
JOIN
for u, a in session.query(User, Address).\
filter(User.id==Address.user_id).\
filter(Address.email_address=='jack@google.com').\
all():
print u
print a
# <User(name='jack', fullname='Jack Bean', password='gjffdd')>
# <Address(email_address='jack@google.com')>
Join 也可以如下表示：session.query(User).join(Address)
如果两张表没有声明的关联，或者有多个关联，最好使用下面的形式：
query.join(Address, User.id==Address.user_id) # explicit condition
query.join(User.addresses) # specify relationship from left to right
query.join(Address, User.addresses) # same, with explicit target
query.join('addresses') # same, using a string
默认是 INNER JOIN：
session.query(User).join(Address). \
filter(Address.email_address == 'jack@google.com'). \
all()
# SELECT users.id AS users_id, users.name AS users_name, users.fullname AS users_fullname, users.password AS users_password
# FROM users
# JOIN addresses ON users.id = addresses.user_id
# WHERE addresses.email_address = 'jack@google.com'
使用其他连接方式：
query.outerjoin(User.addresses) # LEFT OUTER JOIN
或者使用 join 方法的其他参数，具体信息，在后面章节会涉及。
aliased 又出现了
from sqlalchemy.orm import aliased
adalias1 = aliased(Address)
adalias2 = aliased(Address)
for username, email1, email2 in \
session.query(User.name, adalias1.email_address, adalias2.email_address).\
join(adalias1, User.addresses).\
join(adalias2, User.addresses).\
filter(adalias1.email_address=='jack@google.com').\
filter(adalias2.email_address=='j25@yahoo.com'):
print username, email1, email2
# jack jack@google.com j25@yahoo.com
子查询
from sqlalchemy.sql import func
stmt = session.query(Address.user_id, func.count('*').label('address_count')).\
group_by(Address.user_id).subquery()
for u, count in session.query(User, stmt.c.address_count).\
outerjoin(stmt, User.id==stmt.c.user_id).\
order_by(User.id):
print u, count
使用子查询
stmt = session.query(Address).\
filter(Address.email_address != 'j25@yahoo.com').subquery()
adalias = aliased(Address, stmt)
for user, address in session.query(User, adalias).\
join(adalias, User.addresses):
print user
print address
EXIST
from sqlalchemy.sql import exists
stmt = exists().where(Address.user_id==User.id)
for name, in session.query(User.name).filter(stmt):
print name
for name, in session.query(User.name).filter(User.addresses.any()):
print name
for name, in session.query(User.name).\
filter(User.addresses.any(Address.email_address.like('%google%'))):
print name
session.query(Address).filter(~Address.user.has(User.name=='jack')).all()
其他方法
query.filter(User.addresses.contains(someaddress))
query.filter(User.addresses.any(Address.email_address == 'bar'))
query.filter(User.addresses.any(email_address='bar'))
query.filter(Address.user.has(name='ed'))
session.query(Address).with_parent(someuser, 'addresses')
预加载
Eager Loading 直译过来，表示立即载入。
http://sqlalchemy.readthedocs.org/en/latest/orm/loading_relationships.html
默认是 Lazy Loading。
1. subqueryload
from sqlalchemy.orm import subqueryload
jack = session.query(User).\
options(subqueryload(User.addresses)).\
filter_by(name='jack').one()
subqueryload() when used in conjunction with limiting such as Query.first(), Query.limit() or Query.offset() should also include Query.order_by() on a unique column in order to ensure correct results.
2. joinedload
默认 LEFT OUTER JOIN。
from sqlalchemy.orm import joinedload
jack = session.query(User).\
options(joinedload(User.addresses)).\
filter_by(name='jack').one()
joinedload() is not a replacement for join()
The join created by joinedload() is anonymously aliased such that it does not affect the query results. An Query.order_by() or Query.filter() call cannot reference these aliased tables - so-called “user space” joins are constructed using Query.join(). The rationale for this is that joinedload() is only applied in order to affect how related objects or collections are loaded as an optimizing detail - it can be added or removed with no impact on actual results. See the section The Zen of Eager Loading for a detailed description of how this is used.
3. Join + Eagerload
from sqlalchemy.orm import contains_eager
jacks_addresses = session.query(Address).\
join(Address.user).\
filter(User.name=='jack').\
options(contains_eager(Address.user)).\
all()
关联对象的删除
回到曾经的示例——删除 Jack 上来，如果 Jack 有几个地址，实际上 Jack 删除之后那几个地址的用户主键会被设置成 NULL，但是并没有删除。
