Question

我在表格中有一个列，其值增加如下：

AAA0000001
AAA0000002

......等等

我想查找此列中存储的值是否按正确的顺序排列，或者中间是否缺少任何值，或者是否删除。

我怎样才能实现这个目标？

Answer 1

假设模式总是：AAA[0-9][0-9][0-9][0-9][0-9][0-9][0-9]，您可以使用Tally Table执行此操作。

示例数据：

CREATE TABLE Tbl(val VARCHAR(10))
INSERT INTO Tbl VALUES
('AAA0000001'), ('AAA0000002'), ('AAA0000004'), ('AAA0000011');

val
----------
AAA0000001
AAA0000002
AAA0000004
AAA0000011

SQL Fiddle

;WITH Cte AS(
    SELECT *,
        num = CAST(SUBSTRING(val, 4, LEN(val) - 3) AS INT)
    FROM Tbl
),
E1(N) AS(
    SELECT 1 UNION ALL SELECT 1 UNION ALL SELECT 1 UNION ALL SELECT 1 UNION ALL SELECT 1 UNION ALL
    SELECT 1 UNION ALL SELECT 1 UNION ALL SELECT 1 UNION ALL SELECT 1 UNION ALL SELECT 1
),
E2(N) AS(SELECT 1 FROM E1 a CROSS JOIN E1 b),
E4(N) AS(SELECT 1 FROM E2 a CROSS JOIN E2 b),
Tally(N) AS(
    SELECT TOP(SELECT MAX(num) FROM Cte)
        ROW_NUMBER() OVER(ORDER BY (SELECT NULL))
    FROM E4
)
SELECT
    N, 
    val = 'AAA' + RIGHT('0000000' + CAST(N AS VARCHAR(7)), 7)
FROM Tally
WHERE NOT EXISTS(
    SELECT 1 FROM Cte WHERE num = N
)

<强> RESULT

N                    val
-------------------- ----------
3                    AAA0000003
5                    AAA0000005
6                    AAA0000006
7                    AAA0000007
8                    AAA0000008
9                    AAA0000009
10                   AAA0000010

说明：

名为CTE的第一个Cte提取字符串的数字部分，CAST将其提取为INT。
从CTE到E1的后续Tally(N)生成的表格的顺序值从1到MAX(num) - {{1} }从第一个INT返回。
最终CTE只检查第一个SELECT中不存在的num。
CTE转换'AAA' + RIGHT('0000000' + CAST(N AS VARCHAR(7)), 7)，使其遵循模式。

Answer 2

这是一个空白问题。您可以通过Dwain Camps查看本文，了解Gaps and Islands上的更多解决方案。

您可以像这样使用ROW_NUMBER。

示例数据

DECLARE @tab1 TABLE(id VARCHAR(20));

insert into @tab1 VALUES('AAA0000001'),('AAA0000002'),('AAA0000003'),('AAA0000004'),('AAA0000006'),('AAA0000007'),('AAA0000010');

<强>查询

;WITH CTE as 
(
SELECT convert(int,STUFF(id,1,3,'')) id,convert(int,STUFF(id,1,3,'')) - ROW_NUMBER()OVER(ORDER BY convert(int,STUFF(id,1,3,''))) rn
FROM @tab1
),CTE2 as 
(
SELECT ROW_NUMBER()OVER(ORDER BY rn) as rn, MIN(id) series_start,MAX(id) series_end
FROM CTE
GROUP BY rn
)
SELECT C2.series_end,C1.series_start
FROM CTE2 C1
INNER JOIN CTE2 C2 ON C1.rn = C2.rn + 1;

<强> SQL Fiddle

<强>解释

CTE的输出是id值之间差距的差异。
CTE2的输出是连续数字序列的开始和结束
最终输出给出了系列中间隙的开始和结束

<强>输出

series_end  series_start
4   6
7   10

Answer 3

如果架构已修复，则无需复杂查询。这有效：

DECLARE @t TABLE ( v VARCHAR(100) );

INSERT  INTO @t
VALUES  ( 'AAA0000001' ),
        ( 'AAA0000002' ),
        ( 'AAA0000007' ),
        ( 'AAA0000008' ),
        ( 'AAA0000010' ),
        ( 'AAA0000011' ),
        ( 'AAA0000012' );


SELECT * FROM @t t1
CROSS APPLY(SELECT TOP 1 v FROM @t t2 WHERE t2.v > t1.v ORDER BY v) ca
WHERE RIGHT(t1.v, 7) <> RIGHT(ca.v, 7) - 1

输出：

v           v
AAA0000002  AAA0000007
AAA0000008  AAA0000010

Answer 4

在sqlserver 2012中，您可以使用http://www.w3schools.com/cssref/css3_pr_mediaquery.asp和LAG

DECLARE @t table(col1 varchar(15))
INSERT @t values('AAA0000001'),('AAA0000002'),('AAA0000004')

SELECT 
  case when 
    stuff(lag(col1) over (order by col1), 1,3,'') + 1 
      = stuff(col1, 1,3,'') then 'Yes' else 'No' end previous_exists,
  case when
    stuff(lead(col1) over (order by col1), 1,3,'') - 1 
      = stuff(col1, 1,3,'') then 'Yes' else 'No' end next_exists,
  col1
FROM @t

结果：

previous_exists  next_exists  col1
No               Yes          AAA0000001
Yes              No           AAA0000002
No               No           AAA0000004

查找varchar变量中最后一个数值的序列

4 个答案: