如何自動將數據庫反映到sqlalchemy聲明？

Question

sqlautocode - 與許多一對多的關係

sqlsoup問題 - 不支持關係

靈藥 - 這注自動生成

有沒有別的東西，我可以嘗試？

Answer 1

好吧，我經歷了這個，在Northwind數據庫上嘗試過，它看起來很有前途。雖然，我必須添加關係字段才能跟蹤數據庫關係。

讓我們考慮一下，我不知道在啓動應用程序時表之間的關係，所以我需要的是一種自動生成的方法。

import unittest 

from sqlalchemy import * 
from sqlalchemy.orm import create_session 
from sqlalchemy.ext.declarative import declarative_base 
from datetime import datetime 
from sqlalchemy.orm import contains_eager, joinedload 
from sqlalchemy.orm import relationship 

#Create and engine and get the metadata 
Base = declarative_base() 
engine = create_engine('mssql://user:pass@Northwind', echo=True) 
metadata = MetaData(bind=engine) 


#Reflect each database table we need to use, using metadata 
class Customer(Base): 
    __table__ = Table('Customers', metadata, autoload=True) 
    orders = relationship("Order", backref="customer") 

class Shipper(Base): 
    __table__ = Table('Shippers', metadata, autoload=True) 
    orders = relationship("Order", backref="shipper") 

class Employee(Base): 
    __table__ = Table('Employees', metadata, autoload=True) 
# orders = relationship("Order", backref="employee") 
    territories = relationship('Territory', secondary=Table('Employeeterritories', metadata, autoload=True)) 

class Territory(Base): 
    __table__ = Table('Territories', metadata, autoload=True) 
    region = relationship('Region', backref='territories') 

class Region(Base): 
    __table__ = Table('Region', metadata, autoload=True) 


class Order(Base): 
    __table__ = Table('Orders', metadata, autoload=True) 
    products = relationship('Product', secondary=Table('Order Details', metadata, autoload=True)) 
    employee = relationship('Employee', backref='orders') 

class Product(Base): 
    __table__ = Table('Products', metadata, autoload=True) 
    supplier = relationship('Supplier', backref='products') 
    category = relationship('Category', backref='products') 

class Supplier(Base): 
    __table__ = Table('Suppliers', metadata, autoload=True) 

class Category(Base): 
    __table__ = Table('Categories', metadata, autoload=True) 


class Test(unittest.TestCase): 

    def setUp(self): 
     #Create a session to use the tables  
     self.session = create_session(bind=engine)   

    def tearDown(self): 
     self.session.close() 

    def test_withJoins(self): 
     q = self.session.query(Customer) 
     q = q.join(Order) 
     q = q.join(Shipper) 
     q = q.filter(Customer.CustomerID =='ALFKI') 
     q = q.filter(Order.OrderID=='10643') 
     q = q.filter(Shipper.ShipperID=='1') 
     q = q.options(contains_eager(Customer.orders, Order.shipper)) 
     res = q.all() 
     cus = res[0] 
     ord = cus.orders[0] 
     shi = ord.shipper 
     self.assertEqual(shi.Phone, '(503) 555-9831') 

Answer 2

在理論上sqlalchemy的反思應該爲你工作。在這種情況下，我使用與具有簡單的多到一個關係的兩個表的MSSQL數據庫：

「測試」包含字段：

• ID
• 測試名
• AUTHOR_ID（外鍵的用戶表，Users.id場）

「用戶」 與字段：

• ID
• 全稱

所以下面應該反映數據庫：

from sqlalchemy import * 
from sqlalchemy.orm import create_session 
from sqlalchemy.ext.declarative import declarative_base 

#Create and engine and get the metadata 
Base = declarative_base() 
engine = create_engine('put your database connect string here') 
metadata = MetaData(bind=engine) 

#Reflect each database table we need to use, using metadata 
class Tests(Base): 
    __table__ = Table('Tests', metadata, autoload=True) 

class Users(Base): 
    __table__ = Table('Users', metadata, autoload=True) 

#Create a session to use the tables  
session = create_session(bind=engine) 

#Here I will just query some data using my foreign key relation, as you would 
#normally do if you had created a declarative data mode. 
#Note that not all test records have an author so I need to accomodate for Null records 
testlist = session.query(Tests).all()  

for test in testlist: 
    testauthor = session.query(Users).filter_by(id=test.author_id).first() 
    if not testauthor: 
     print "Test Name: {}, No author recorded".format(test.testname) 
    else: 
     print "Test Name: {}, Test Author: {}".format(test.testname, testauthor.fullname) 

因此，這似乎與臺關係的工作。雖然你還沒有給出你準備做什麼的很多細節。

Answer 3

您可以使用sqlacodegen從數據庫中生成所有模型。但是，您需要手動處理外鍵。