mysql查询（同一个表中相邻的两条记录进行运算）-CFANZ编程社区

通过添加一条行记录的方法，重新生成一个表，然后在通过可以通过表的连接进行运算。

--通过一个初始值为0的变量@rownum，依次递增1来实现行号
SELECT
    @rownum:=@rownum+1 AS rownum,name
FROM temp,
    (SELECT @rownum:=0) temp
WHERE @rownum<4（条件，没有也可以）;

给出一般性的语句：

SELECT
    @rownum:=@rownum+1 AS rownum,所需字段
FROM 表名,
    (SELECT @rownum:=0) 表名
WHERE @rownum<N(获取多少行);

然后给出测试案例：

第一步，创建测试表；

mysql> CREATE TABLE temp(id INT(4),name VARCHAR(20));
Query OK, 0 rows affected (0.13 sec)

第二步，插入测试数据；

mysql> INSERT INTO temp VAlUES(1,'robin');
Query OK, 1 row affected (0.10 sec)

mysql> INSERT INTO temp VAlUES(2,'wentasy');
Query OK, 1 row affected (0.05 sec)

mysql> INSERT INTO temp VAlUES(3,'justdb');
Query OK, 1 row affected (0.08 sec)

mysql> INSERT INTO temp VAlUES(4,'wen');
Query OK, 1 row affected (0.08 sec)

mysql> INSERT INTO temp SELECT * FROM temp;
Query OK, 4 rows affected (0.03 sec)
Records: 4  Duplicates: 0  Warnings: 0

mysql> INSERT INTO temp SELECT * FROM temp;
Query OK, 8 rows affected (0.02 sec)
Records: 8  Duplicates: 0  Warnings: 0

mysql> INSERT INTO temp SELECT * FROM temp;
Query OK, 16 rows affected (0.02 sec)
Records: 16  Duplicates: 0  Warnings: 0

mysql> INSERT INTO temp SELECT * FROM temp;
Query OK, 32 rows affected (0.03 sec)
Records: 32  Duplicates: 0  Warnings: 0

mysql> INSERT INTO temp SELECT * FROM temp;
Query OK, 64 rows affected (0.04 sec)
Records: 64  Duplicates: 0  Warnings: 0

mysql> INSERT INTO temp SELECT * FROM temp;
Query OK, 128 rows affected (0.03 sec)
Records: 128  Duplicates: 0  Warnings: 0

mysql> INSERT INTO temp SELECT * FROM temp;
Query OK, 256 rows affected (0.04 sec)
Records: 256  Duplicates: 0  Warnings: 0

mysql> INSERT INTO temp SELECT * FROM temp;
Query OK, 512 rows affected (0.05 sec)
Records: 512  Duplicates: 0  Warnings: 0

mysql> INSERT INTO temp SELECT * FROM temp;
Query OK, 1024 rows affected (0.05 sec)
Records: 1024  Duplicates: 0  Warnings: 0

第三步，实现类似Oracle中的rownum效果；

mysql> SELECT
    ->     @rownum:=@rownum+1 AS rownum,name
    -> FROM temp,
    ->     (SELECT @rownum:=0) temp
    -> WHERE @rownum<4;
+--------+---------+
| rownum | name    |
+--------+---------+
|      1 | robin   |
|      2 | wentasy |
|      3 | justdb  |
|      4 | wen     |
+--------+---------+
4 rows in set (0.01 sec)

---------------------------------------------------------------------------------------------------

MySQL 几乎模拟了 Oracle,SQL Server等商业数据库的大部分功能，函数。但很可惜，到目前的版本(5.1.33)为止,仍没有实现ROWNUM这个功能

下面介绍几种具体的实现方法.

建立实验环境如下
mysql> create table tbl (
    -> id      int primary key,
    -> col     int
    -> );
Query OK, 0 rows affected (0.08 sec)

mysql> insert into tbl values
    -> (1,26),
    -> (2,46),
    -> (3,35),
    -> (4,68),
    -> (5,93),
    -> (6,92);
Query OK, 6 rows affected (0.05 sec)
Records: 6 Duplicates: 0 Warnings: 0

mysql>
mysql> select * from tbl order by col;
+----+------+
| id | col |
+----+------+
| 1 |   26 |
| 3 |   35 |
| 2 |   46 |
| 4 |   68 |
| 6 |   92 |
| 5 |   93 |
+----+------+
6 rows in set (0.00 sec)

1. 直接在程序中实现；
这应该算是效率最高的一种，也极为方便。直接在你的开发程序中（PHP/ASP/C/...）等中，直接初始化一个变量nRowNum=0，然后在while 记录集时，nRowNum++; 然后输出即可。

2. 使用MySQL变量；在某些情况下，无法通过修改程序来实现时，可以考虑这种方法。
缺点，@x 变量是 connection 级的，再次查询的时候需要初始化。一般来说PHP等B/S应用没有这个问题。但C/S如果connection一只保持则要考虑 set @x=0

mysql> select @x:=ifnull(@x,0)+1 as rownum,id,col
    -> from tbl
    -> order by col;
+--------+----+------+
| rownum | id | col |
+--------+----+------+
|      1 | 1 |   26 |
|      1 | 3 |   35 |
|      1 | 2 |   46 |
|      1 | 4 |   68 |
|      1 | 6 |   92 |
|      1 | 5 |   93 |
+--------+----+------+
6 rows in set (0.00 sec)

3. 使用联接查询（笛卡尔积）
缺点，显然效率会差一些。

利用表的自联接，代码如下，你可以直接试一下 select a.*,b.* from tbl a,tbl b where a.col>=b.col 以理解这个方法原理。

mysql> select a.id,a.col,count(*) as rownum
    -> from tbl a,tbl b
    -> where a.col>=b.col
    -> group by a.id,a.col;
+----+------+--------+
| id | col | rownum |
+----+------+--------+
| 1 |   26 |      1 |
| 2 |   46 |      3 |
| 3 |   35 |      2 |
| 4 |   68 |      4 |
| 5 |   93 |      6 |
| 6 |   92 |      5 |
+----+------+--------+
6 rows in set (0.00 sec)

4. 子查询

缺点，和联接查询一样，具体的效率要看索引的配置和MySQL的优化结果。
mysql> select a.*,
    -> (select count(*) from tbl where col<=a.col) as rownum
    -> from tbl a;
+----+------+--------+
| id | col | rownum |
+----+------+--------+
| 1 |   26 |      1 |
| 2 |   46 |      3 |
| 3 |   35 |      2 |
| 4 |   68 |      4 |
| 5 |   93 |      6 |
| 6 |   92 |      5 |
+----+------+--------+
6 rows in set (0.06 sec)

做为一款开源的数据库系统，MySQL无疑是一个不做的产品。它的更新速度，文档维护都不逊于几大商业数据库产品。估计在下一个版本中，我们可以看到由MySQL自身实现的ROWNUM。

----------------------------------------------------------------------------------------------------

应用：同一个表中相邻的两条记录进行运算中的cpu_busy，cpu_total。

cpu_total = usr+ nice+sys+idle+iowait+irq+softirq;

cpu_ busy= cpu_total- idle;

计算cpu利用率：

cpu_usage=( cpu_ busy 2- cpu_ busy 1)/( cpu_total2 - cpu_total 1) [相邻两行怎么计算差值？？]

select usr+nice+sys+idle+iowait+irq+softirq ascpu_totalfrom os;

select usr+nice+sys+iowait+irq+softirq as cpu_busy from os;

思路:os表和os表进行自身的左连接操作，条件是表A的A.timestamp_+1等于表B的B.timestamp_。

select A.timestamp_, A.cpu_total , A.cpu_busy,

B.timestamp_ , B.cpu_total , B.cpu_busy, B.cpu_busy-A.cpu_busy ,B.cpu_total-A.cpu_total,

(B.cpu_busy-A.cpu_busy)/(B.cpu_total-A.cpu_total) as cpu_usage from

(select timestamp_, usr+nice+sys+idle+iowait+irq+softirq as cpu_total ,usr+nice+sys+iowait+irq+softirq as cpu_busy from os ) A

left join

(select timestamp_, usr+nice+sys+idle+iowait+irq+softirq as cpu_total ,usr+nice+sys+iowait+irq+softirq as cpu_busy from os )

on A.timestamp_+1=B.timestamp_;

把一个表的某几项数据，插入到另一个表中。

UPDATE 表A INNER JOIN 表B ON 表A的字段=表B的字段(条件) SET 表A的等待修改项=表B的某个字段

update os inner join

(select timestamp_ from ) D

on timestamp_=D.btimestamp_
set os.cpu_usage=D.cpu_usage;