下面有两个表格。第一个表包含有关索引的信息。每个索引可以应用于一个或多个列。第二个表包含成对的集合:表名-列名。
我需要以某种方式获取表1的索引列表,并将其应用于表2的列。结果应包括过滤索引的所有列(请参见下面的结果表)。
#1
+---------------------------------------+
| Index name | Table name | Column name |
+---------------------------------------+
| Index_1 | Table_A | Column_A_1 |
| Index_1 | Table_A | Column_A_2 |
| Index_2 | Table_A | Column_A_1 |
| Index_2 | Table_A | Column_A_3 |
| Index_3 | Table_B | Column_B_1 |
| Index_3 | Table_B | Column_B_2 |
| Index_4 | Table_C | Column_C_1 |
+---------------------------------------+
#2
+--------------------------+
| Table name | Column name |
+--------------------------+
| Table_A | Column_A_2 |
| Table_B | Column_B_1 |
+--------------------------+
Result:
+---------------------------------------+
| Index name | Table name | Column name |
+---------------------------------------+
| Index_1 | Table_A | Column_A_1 |
| Index_1 | Table_A | Column_A_2 |
| Index_3 | Table_B | Column_B_1 |
| Index_3 | Table_B | Column_B_2 |
+---------------------------------------+
我可以在不使用其他表的情况下对每个“ SELECT”操作执行此操作吗?如果可以,怎么办?
答案 0 :(得分:0)
使用加入
select t1.* from table1 t1
join table2 t2
on t1.table_name=t2.table_name
where t1.index_name in ('Index_1','Index_3') --- provide filter values
答案 1 :(得分:0)
EXISTS
(半联接)更合适:
SELECT t1.* FROM table1 t1
WHERE EXISTS(
SELECT * FROM table2 t2
WHERE
t1.table_name=t2.table_name and t1.col_name=t2.col_name
)
答案 2 :(得分:0)
直接的JOIN
或EXISTS
不会删除它,因为如果至少其中一个行满足条件,则要显示一个集合(完整的索引)。
您需要首先确定哪些索引与另一个表匹配,然后显示其所有行:
;WITH IndexMatches AS
(
SELECT DISTINCT
I.IndexName
FROM
IndexTable AS I
INNER JOIN ColumnsTable AS C ON
C.TableName = I.TableName AND
C.ColumnName = I.ColumnName
)
SELECT
I.*
FROM
IndexTable AS I
INNER JOIN IndexMatches AS M ON I.IndexName = M.IndexName
ORDER BY
I.IndexName,
I.TableName,
I.ColumnName
或使用EXISTS
:
SELECT
I.*
FROM
IndexTable AS I
WHERE
EXISTS (
SELECT
'at least one column match'
FROM
IndexTable AS I2
INNER JOIN ColumnsTable AS C ON
C.TableName = I2.TableName AND
C.ColumnName = I2.ColumnName
WHERE
I.IndexName = I2.IndexName)
ORDER BY
I.IndexName,
I.TableName,
I.ColumnName