sqlalchemy 筆記

初始化數據庫連接

DB_CONNECT_STRING = 'mysql+mysqldb://root:123123@localhost/sqlalchemy?charset=utf8'
engine = create_engine(DB_CONNECT_STRING, echo=True)

數據庫類型+數據庫驅動名稱://用戶名:口令@機器地址:端口號/數據庫名

創建DBSession

DB_Session = sessionmaker(bind=engine)
session = DB_Session()

執行 SQL

session.execute('create database abc')
print session.execute('show databases').fetchall()
session.execute('use mysql')
#查詢所有
print session.execute('select * from user').fetchall(), "fetchall"
#第一條
print session.execute('select * from user').first(), "first"
#第一條
print session.execute('select * from user').fetchone(), "fetchone"
#查詢2條記錄
print session.execute('select * from user').fetchmany(2), "fetchmany"

Model

#declarative_base() 創建了一個 BaseModel 類，這個類的子類可以自動與一個表關聯。
BaseModel = declarative_base()
class User(BaseModel):
    # 表的名字:
    __tablename__ = 'user'

    # 表的結構:
    id = Column(Integer, primary_key=True)
    name = Column(String(20), nullable=False, unique=True)
    test = Column(String(20), name="a1", default="a1")
    email_address = Column(String(32), nullable=False, default="email")

    def __repr__(self):
        return '<User id=%s,name=%s,test=%s,email_address=%s >' % (self.id, self.name, self.test, self.email_address)

primary_key 主鍵，自增
unique 唯一 UNIQUE (name)
name 數據庫字段名字
nullable 是否可以爲空 False NOT NULL
default 默認值

表屬性

class User(BaseModel):
    __table_args__ = {
        'mysql_engine': 'InnoDB',
        'mysql_charset': 'utf8'
    }

建表 or 刪除表

#建表
User.metadata.create_all(engine) or  BaseModel.metadata.create_all(engine) #創建所有BaseModel的之類的表
#刪除表
User.metadata.drop_all(engine) or BaseModel.metadata.drop_all(engine) #同上

使用

u1 = User(name="aa", email_address="aa_email")
u2 = User(name="bb")
session.add(u1)
session.add(u2)
session.commit()#提交會話

query = session.query(User)
for user in query:
    print user
print query.all()  # 查詢所有
print query.offset(1).limit(2).all()  # LIMIT 1, 2
print query.first()  # 第一條,不存在返回None
# print query.one() #之獲取一條，不存在，或有多行記錄時會拋出異常
print query.filter(User.id == 2).all()  # WHERE "user".id = 2
print query.filter(User.id >= 1).all()  # WHERE "user".id >= 1
print query.filter(User.id == User.name).all()  # WHERE user.id = user.name
print query.filter(User.id >= 1).filter(User.id <= 11).all()  # WHERE "user".id >= 1 AND user.id <= 11
print query.filter(User.id >= 1, User.id <= 11).all()  # 同上
print query.filter(or_(User.id >= 1, User.id <= 11),
                   User.name == "111").all()  # (user.id >= 1 OR user.id <=11) AND user.name = 111
print query.filter(User.id.in_((1, 2))).all()  # user.id IN (1, 2)
print query.filter(User.name.like("a")).all()  # user.name LIKE "a"
print query.filter(User.name.isnot(None)).all()  # WHERE user.name IS NOT NULL
print query.filter(User.name != None).all()  # WHERE user.name IS NOT NULL
print query.filter(not_(User.id != 1), not_(User.name.like("a"))).all()  # WHERE user.id = 1 AND user.name NOT LIKE a
print query.filter("id is null").all()  # 支持字符串
print query.get(2)  # 以主鍵獲取,WHERE "user".id = 2
print query.order_by("name")  # ORDER BY "user".name
print query.order_by(User.name)  # 同上
print query.order_by(User.name.desc())  # ORDER BY "user".name desc
# print query.order_by("name desc")# 同上
print query.order_by(User.name.desc(), User.id.desc())  # ORDER BY name desc, id desc
print query.count()
print "#" * 30
#修改
# query.filter(User.id == 1).update({User.name: 'c'})
print "#" * 30


query2 = session.query(User).order_by(User.name.desc(), User.id.desc())
print query2.all()
print "#" * 30

# 返回元組
query3 = session.query(User.name)
print query3.all()
print query3.filter(User.id == 1).scalar()  # 如果有記錄，返回第一條記錄的第一個元素

print session.query('id', "name").select_from(User).filter(
    User.id == 1).scalar()  # SELECT id, name FROM user WHERE user.id = 1
print "#" * 30

print session.query(func.count('1')).select_from(User).scalar()
print session.query(User).filter(User.name == func.now())#WHERE "user".name = now()
print session.query(func.count(User.id)).select_from(User).scalar()
print session.query(func.sum(User.id)).select_from(User).scalar()

print session.query(func.now()).scalar()  # func 後可以跟任意函數名，只要該數據庫支持
print session.query(User.id, func.md5(User.name)).filter(User.id == 1).all()



u = session.query(User).get(1)
print u
u.name = "1111111111"
session.flush()# 寫數據庫，但並不提交
print session.query(User).get(1)
session.rollback()#返回
print session.query(User).get(1)
u.name = "1111111111"
print query.all()
#優先操作 HIGH_PRIORITY  滯後操作 LOW_PRIORITY 延時插入 INSERT DELAYED
session.query(User.name).prefix_with('HIGH_PRIORITY').all()#SELECT HIGH_PRIORITY user.name AS user_name FROM user
print query.all()

# session.delete(session.query(User).get(2))
# session.delete(session.query(User).first())
# session.query(User).filter(User.id >= 1).delete()
session.commit()

外建

class User(BaseModel):
    # 表的名字:
    __tablename__ = 'user'

    # 表的結構:
    id = Column(Integer, primary_key=True)
    name = Column(String(20), nullable=False, unique=True)
    test = Column(String(20), name="a1", default="a1")
    email_address = Column(String(32), nullable=False, default="email")
    addresses = relationship("Address", order_by="Address.id", backref="user")

    def __repr__(self):
        return '<User id=%s,name=%s,test=%s,email_address=%s >' % (self.id, self.name, self.test, self.email_address)


class Address(BaseModel):
    __tablename__ = 'address'

    id = Column(Integer, primary_key=True)
    address = Column(String(20), nullable=False, unique=True)
    user_id = Column(Integer, ForeignKey('user.id', ondelete='CASCADE', onupdate='CASCADE'))

    def __repr__(self):
        return '<User id=%s,name=%s,test=%s,email_address=%s >' % (self.id, self.address)

user_id = Column(Integer, ForeignKey(‘user.id’, ondelete=’CASCADE’, onupdate=’CASCADE’)) 原因是刪除 user 表的數據，可能會導致 address 的外鍵不指向一個真實存在的記錄。在默認情況下，MySQL 會拒絕這種操作，也就是 RESTRICT。InnoDB 還允許指定 ON DELETE 爲 CASCADE 和 SET NULL，前者會刪除 friendship 中無效的記錄，後者會將這些記錄的外鍵設爲 NULL。
relationship 標記User的addresses連接到Address表，session.query(User).filter(User.name == “bb”).one().addresses
backref 反向查詢的細節 session.query(Address).filter(Address.address==”a1”).one().user

u1 = User(name="aa", email_address="aa_email")
u2 = User(name="bb")

a1 = Address(address="a1")
a2 = Address(address="a2")
a3 = Address(address="a3")

u2.addresses = [a1, a2]
session.add(u1)
session.add(u2)
session.add(a3)
session.flush()

session.commit()
print session.query(User).filter(User.name == "bb").one().addresses
print session.query(Address).filter(Address.address=="a1").one().user

print session.query(User).all()
print session.query(Address).all()

多對多

user_keywords = Table('user_keywords',
                      BaseModel.metadata,
                      Column('user_id', Integer, ForeignKey('user.id')),
                      Column('keyword_id', Integer, ForeignKey('keyword.id'))
                      )

class User(BaseModel):
    # 表的名字:
    __tablename__ = 'user'

    # 表的結構:
    id = Column(Integer, primary_key=True)
    name = Column(String(20), nullable=False, unique=True)
    test = Column(String(20), name="a1", default="a1")
    email_address = Column(String(32), nullable=False, default="email")
    keywords = relation("Keyword", secondary=user_keywords, backref="user")

    def __repr__(self):
        return '<User id=%s,name=%s,test=%s,email_address=%s >' % (self.id, self.name, self.test, self.email_address)

class Keyword(BaseModel):
    __tablename__ = 'keyword'

    id = Column(Integer, primary_key=True)
    keyword = Column(String(20), nullable=False, unique=True)
    user = relation("User", secondary=user_keywords, backref="user")

    def __repr__(self):
        return '<User id=%s,keyword=%s>' % (self.id, self.keyword)