替换VIEW定义中的文本

Question

我有以下VIEW定义：

create view v_1
as
select * from t1
where [date] between '2010-01-01' and '2019-01-01'
union all
select * from t2
where [date] between '2012-01-01' and '2019-01-01'
union all
select * from t3
where [date] between '2013-01-01' and '2019-01-01'
union all
select * from t4
where [date] between '2014-01-01' and '2019-01-01'

方案1：想要删除（或替换为空字符串）table t1 select语句。

视图应如下所示：

create view v_1
as
select * from t2
where [date] between '2012-01-01' and '2019-01-01'
union all
select * from t3
where [date] between '2013-01-01' and '2019-01-01'
union all
select * from t4
where [date] between '2014-01-01' and '2019-01-01'  

方案2：要删除table t4 select语句。

视图应如下所示：

create view v_1
as
select * from t1
where [date] between '2010-01-01' and '2019-01-01'
union all
select * from t2
where [date] between '2012-01-01' and '2019-01-01'
union all
select * from t3
where [date] between '2013-01-01' and '2019-01-01'

注意：根据要求，可能要从VIEW定义中删除所有select语句。

Answer 1

代替视图，创建内联表功能也称为带有参数的视图。例如：

create function v_1
(   
    @HasQuery1 BIT
   ,@HasQuery2 BIT
   ,@HasQuery3 BIT
   ,@HasQuery4 BIT
)
RETURNS TABLE
as
RETURN
(
select * from t1
where [date] between '2010-01-01' and '2019-01-01'
    AND @HasQuery1  = 1
union all
select * from t2
where [date] between '2012-01-01' and '2019-01-01'
    AND  @HasQuery2  = 1
union all
select * from t3
where [date] between '2013-01-01' and '2019-01-01'
    AND  @HasQuery3  = 1
union all
select * from t4
where [date] between '2014-01-01' and '2019-01-01'
    AND  @HasQuery4  = 1
)

您可以在类似视图的联接中使用该函数，但是可以传递要返回的数据。 例如：

SELECT *
FROM T1 A
INNER JOIN V_1(1, 0, 0, 1) B
   ON ....

Answer 2

SQL服务器使您可以在编辑器中使用正则表达式作为搜索模式，因此您要做的就是：

1. 按Ctrl + F

2. 使用正则表达式启用

3. 使用搜索模式select.+from t1(.*\n)+?select，其中t1可以用任何表名替换

4. 使用select

作为替换模式

请参见下面的屏幕（更换后）：

Answer 3

我喜欢@gotqn的答案，但是，如果必须是视图，我会这样做：

--Sample Tables:
CREATE TABLE dbo.t1(id INT IDENTITY PRIMARY KEY, col1 INT);
CREATE TABLE dbo.t2(id INT IDENTITY PRIMARY KEY, col1 INT);
CREATE TABLE dbo.t3(id INT IDENTITY PRIMARY KEY, col1 INT);
CREATE TABLE dbo.t4(id INT IDENTITY PRIMARY KEY, col1 INT);

INSERT dbo.t1(col1) VALUES (10),(20),(30);
INSERT dbo.t2(col1) VALUES (60),(620),(630);
INSERT dbo.t3(col1) VALUES (0);
INSERT dbo.t4(col1) VALUES (0);
GO

--Solution
CREATE VIEW dbo.yourView AS
  SELECT t.*,show = 1 FROM dbo.t1 AS t UNION ALL
  SELECT t.*,show = 2 FROM dbo.t2 AS t UNION ALL
  SELECT t.*,show = 3 FROM dbo.t3 AS t UNION ALL
  SELECT t.*,show = 4 FROM dbo.t4 AS t
GO

这里的好处是优化器足够聪明，可以完全忽略并避免触摸最终结果集中未包含的数据的表。

请注意以下查询和执行计划：

SELECT v.* FROM dbo.yourView AS v;
SELECT v.* FROM dbo.yourView AS v WHERE v.show = 1;
SELECT v.* FROM dbo.yourView AS v WHERE v.show IN (1,3);
SELECT v.* FROM dbo.yourView AS v WHERE v.show <> 1;