Sql 比较两个数字字符串并获取唯一值
以下是我为更好地说明我的问题而制作的示例表：Sql 比较两个数字字符串并获取唯一值,sql,Sql,以下是我为更好地说明我的问题而制作的示例表： Create Table SampleTable( TableID int, NumberRow nvarchar(500) ) Insert into SampleTable Values(1, '15,21,23,41,44,5,50,59,6,') Insert into SampleTable Values(2, '10,24,29,41,5,50,59,6,73,') Insert into SampleTable Valu
Create Table SampleTable(
TableID int,
NumberRow nvarchar(500)
)
Insert into SampleTable Values(1, '15,21,23,41,44,5,50,59,6,')
Insert into SampleTable Values(2, '10,24,29,41,5,50,59,6,73,')
Insert into SampleTable Values(3, '10,15,21,24,29,33,41,50,59,60,61,62,66,73,')
Insert into SampleTable Values(4, '10,15,21,24,28,33,37,41,44,5,50,6,60,61,62,66,')
Insert into SampleTable Values(5, '15,24,33,41,5,6,61,62,66,73,')
Select
t1.TableID AS ID1,
t2.TableID AS ID2,
t1.NumberRow AS Numbers1,
t2.NumberRow AS Numbers2
From SampleTable t1
inner join SampleTable t2
on t1.TableID = t2.TableID - 1
Order by t2.TableID asc
ID1 ID2 Numbers1 Numbers2
-------------------------------------------------------------------------------------
1 2 15,21,23,41,44,5,50,59,6, 10,24,29,41,5,50,59,6,73,
2 3 10,24,29,41,5,50,59,6,73, 10,15,21,24,29,33,41,50,59,60,61,62,66,73,
3 4 10,15,21,24,29,33,41,50,59,60,61,62,66,73, 10,15,21,24,28,33,37,41,44,5,50,6,60,61,62,66,
4 5 10,15,21,24,28,33,37,41,44,5,50,6,60,61,62,66, 15,24,33,41,5,6,61,62,66,73,
到目前为止，我还没有想出任何解决办法。我的另一种方法是将数字设置为列而不是字符串，但我仍然不知道如何解决我的问题。也许这个查询会有所帮助，但我同意表的设计考虑
select DISTINCT value as NUMBER1 ,'' AS NUMBER2 from (
Select
t1.TableID AS ID1,
t2.TableID AS ID2,
t1.NumberRow AS Numbers1,
t2.NumberRow AS Numbers2
From SampleTable t1
inner join SampleTable t2
on t1.TableID = t2.TableID - 1
) as tmp_tbl
CROSS APPLY string_split(tmp_tbl.Numbers1,',')
UNION ALL
select DISTINCT '', value from (
Select
t1.TableID AS ID1,
t2.TableID AS ID2,
t1.NumberRow AS Numbers1,
t2.NumberRow AS Numbers2
From SampleTable t1
inner join SampleTable t2
on t1.TableID = t2.TableID - 1
) as tmp_tbl
CROSS APPLY string_split(tmp_tbl.Numbers1,',')
+---------+---------+
| NUMBER1 | NUMBER2 |
+---------+---------+
| | |
| 10 | |
| 15 | |
| 21 | |
| 23 | |
| 24 | |
| 28 | |
| 29 | |
| 33 | |
| 37 | |
| 41 | |
| 44 | |
| 5 | |
| 50 | |
| 59 | |
| 6 | |
| 60 | |
| 61 | |
| 62 | |
| 66 | |
| 73 | |
| | |
| | 10 |
| | 15 |
| | 21 |
| | 23 |
| | 24 |
| | 28 |
| | 29 |
| | 33 |
| | 37 |
| | 41 |
| | 44 |
| | 5 |
| | 50 |
| | 59 |
| | 6 |
| | 60 |
| | 61 |
| | 62 |
| | 66 |
| | 73 |
+---------+---------+
规范图式；阅读Good day Tertmartus，您的数据库设计非常有问题，不符合您的需要。是的，我们可以解决它并提供解决方案，但这并不意味着您应该使用它。表格结构用于在表格的每个单元格中存储单个值。使用带分隔符的多个值可能会导致vert性能不佳，因为每次使用该值时都需要对其进行削减，并且可能需要对其进行拆分。您真的应该考虑重新设计数据库。@Ronenariley这只是一个示例表。在实际项目中，数值单独存储在一列中。我制作这个表是为了存储字符串值，因为我认为这样更容易理解我的观点。所以你应该提供原始结构，否则你会得到一个错误问题的解决方案。您应该发布DDL+DML，它尽可能接近生产场景。
select DISTINCT value as NUMBER1 ,'' AS NUMBER2 from (
Select
t1.TableID AS ID1,
t2.TableID AS ID2,
t1.NumberRow AS Numbers1,
t2.NumberRow AS Numbers2
From SampleTable t1
inner join SampleTable t2
on t1.TableID = t2.TableID - 1
) as tmp_tbl
CROSS APPLY string_split(tmp_tbl.Numbers1,',')
UNION ALL
select DISTINCT '', value from (
Select
t1.TableID AS ID1,
t2.TableID AS ID2,
t1.NumberRow AS Numbers1,
t2.NumberRow AS Numbers2
From SampleTable t1
inner join SampleTable t2
on t1.TableID = t2.TableID - 1
) as tmp_tbl
CROSS APPLY string_split(tmp_tbl.Numbers1,',')
+---------+---------+
| NUMBER1 | NUMBER2 |
+---------+---------+
| | |
| 10 | |
| 15 | |
| 21 | |
| 23 | |
| 24 | |
| 28 | |
| 29 | |
| 33 | |
| 37 | |
| 41 | |
| 44 | |
| 5 | |
| 50 | |
| 59 | |
| 6 | |
| 60 | |
| 61 | |
| 62 | |
| 66 | |
| 73 | |
| | |
| | 10 |
| | 15 |
| | 21 |
| | 23 |
| | 24 |
| | 28 |
| | 29 |
| | 33 |
| | 37 |
| | 41 |
| | 44 |
| | 5 |
| | 50 |
| | 59 |
| | 6 |
| | 60 |
| | 61 |
| | 62 |
| | 66 |
| | 73 |
+---------+---------+