怎样用python+sqlalchemy获得mssql视图对应物理表关系(二)

话不多说

目标:为了实现低代码数据视图对接,有必要得到视图所对应物理表及字段名称,字段类型等

1)约束:视图中用到的物理表不能起别名,所以修改上一篇中存储过程建立语句

USE [agui_conn]
GO
/****** Object:  StoredProcedure [dbo].[sp_GetOrdersByTimestamp]    Script Date: 09/29/2024 14:40:36 ******/
SET ANSI_NULLS ON
GO
SET QUOTED_IDENTIFIER ON
GO
ALTER PROCEDURE [dbo].[sp_GetOrdersByTimestamp]
    @Timestamp NVARCHAR(100),
    @Condition NVARCHAR(100) -- 假设这是一个额外的条件，如订单状态
AS
BEGIN
	IF OBJECT_ID('dbo.vw_OrdersByTimestamp', 'V') IS NOT NULL
	BEGIN
		EXEC sp_executesql N'DROP VIEW dbo.vw_OrdersByTimestamp';
	END
	-- 创建一个视图来显示所需的订单信息
	DECLARE @sql NVARCHAR(MAX);
	SET @sql = N'CREATE VIEW dbo.vw_OrdersByTimestamp AS
	SELECT 
		Users.Username as 用户名, 
		Orders.OrderTime as 订单时间,
		Orders.Amount as 订单金额,
		'''+ @Timestamp + ''' AS 查询时间
	FROM 
		Orders
	INNER JOIN 
		Users ON Orders.UserId = Users.UserId
	WHERE 
		(Orders.Status = ''' + @Condition+ ''')'; -- 假设订单表中有一个Status字段o.OrderTime >= ''' + CONVERT(NVARCHAR, @Timestamp, 120) + ''' AND
	EXEC sp_executesql @sql;

	
END

2)在 SQL Server 中，用存储过程创建的视图没有出现在 sys.views 中，尝试查询 sys.objects 来获取所有对象的列表:

SELECT 
    OBJECT_SCHEMA_NAME(object_id) AS SchemaName,
    OBJECT_NAME(object_id) AS ViewName
FROM 
    sys.objects
WHERE 
    OBJECTPROPERTY(object_id, 'IsView') = 1;

3)通过sys.objects获得指定视图的描述

SchemaName='dbo'
            ViewName='vw_OrdersByTimestamp'
            create_view_query = text(f"""SELECT 
                    OBJECT_DEFINITION(OBJECT_ID('{SchemaName}.{ViewName}')) as ViewDefinition
                FROM 
                    sys.objects
                WHERE 
                    OBJECTPROPERTY(object_id, 'IsView') = 1
                    AND OBJECT_NAME(object_id) = '{ViewName}'
                    AND OBJECT_SCHEMA_NAME(object_id) = '{SchemaName}'""")
            result = conn.execute(create_view_query)
            create_view_result = result.fetchone()

            view_creation_sql=create_view_result['ViewDefinition']
            print(view_creation_sql)

结果如图:

4)用正则表达式匹配视图字段名和对应的物理表名及字段名

# 正则表达式匹配视图字段名和对应的物理表名及字段名
            column_pattern = re.compile(r"(\w+\.\w+) as (\w+)", re.IGNORECASE)
            # 提取字段映射
            columns_mapping = column_pattern.findall(view_creation_sql)
            # 常量值匹配
            constant_pattern = re.compile(r"'([^']*)' AS (\w+)", re.IGNORECASE)
            constants_mapping = constant_pattern.findall(view_creation_sql)

            field_dict = {}
            for column, alias in columns_mapping:
                field_dict[f'{alias}'] = column
            for constant, alias in constants_mapping:
                field_dict[f'{alias}'] = None
            # 反射视图的列信息
            table = Table(view_name, metadata, autoload=True, autoload_with=engine)
            for column in table.columns:
                v_field = f"{column.name}"
                tb_field = field_dict[v_field]
                print(f"视图字段名: {column.name}, 对应表字段: {tb_field}, 数据类型: {column.type}")

最终实现结果

到此,祝大家节日快乐!!!