MySQL管理之SQL语句实例
首先,我们来导入world库,这个world库中的表是mysql ocp考试专用表,在网上有下
mysql> source/root/world_innodb.sql
表结构如下: 本文来自http://yijiu.blog.51cto.com 转载请经博主允许 ,盗帖可耻!
CREATE TABLE `Country` (
`Code` char(3) NOT NULL DEFAULT '',
`Name` char(52) NOT NULL DEFAULT '',
`Continent` enum('Asia','Europe','North America','Africa','Oceania','Antarctica','South America') NOT NULL DEFAULT 'Asia',
`Region` char(26) NOT NULL DEFAULT '',
`SurfaceArea` float(10,2) NOT NULL DEFAULT '0.00',
`IndepYear` smallint(6) DEFAULT NULL,
`Population` int(11) NOT NULL DEFAULT '0',
`LifeExpectancy` float(3,1) DEFAULT NULL,
`GNP` float(10,2) DEFAULT NULL,
`GNPOld` float(10,2) DEFAULT NULL,
`LocalName` char(45) NOT NULL DEFAULT '',
`GovernmentForm` char(45) NOT NULL DEFAULT '',
`HeadOfState` char(60) DEFAULT NULL,
`Capital` int(11) DEFAULT NULL,
`Code2` char(2) NOT NULL DEFAULT '',
PRIMARY KEY (`Code`)
) ENGINE=InnoDB DEFAULT CHARSET=utf8
使用show table status查看表状态如下所示:
mysql> show tablestatus;
+-----------------+--------+---------+------------+------+----------------+-------------+-----------------+--------------+-----------+----------------+---------------------+-------------+------------+-------------------+----------+----------------+---------+
| Name | Engine | Version | Row_format |Rows | Avg_row_length | Data_length | Max_data_length | Index_length |Data_free | Auto_increment | Create_time | Update_time | Check_time | Collation | Checksum | Create_options | Comment|
+-----------------+--------+---------+------------+------+----------------+-------------+-----------------+--------------+-----------+----------------+---------------------+-------------+------------+-------------------+----------+----------------+---------+
| City | InnoDB | 10 | Compact | 4321 | 94 | 409600 | 0 | 131072 | 0 | 4080 | 2014-10-02 15:35:18 |NULL | NULL | latin1_swedish_ci | NULL | | |
| Country | InnoDB | 10 | Compact | 241 | 407 | 98304 | 0 | 0 | 0 | NULL | 2014-10-02 15:35:18 |NULL | NULL | latin1_swedish_ci | NULL | | |
| CountryLanguage |InnoDB | 10 | Compact | 856 | 114 | 98304 | 0 | 65536 | 0 | NULL | 2014-10-02 15:35:18 |NULL | NULL | latin1_swedish_ci | NULL | | |
+-----------------+--------+---------+------------+------+----------------+-------------+-----------------+--------------+-----------+----------------+---------------------+-------------+------------+-------------------+----------+----------------+---------+
3 rows in set (0.00sec)
mysql> select *from Country where Name="China"\G
***************************1. row ***************************
Code: CHN
Name: China
Continent: Asia
Region: Eastern Asia
SurfaceArea: 9572900.00
IndepYear: -1523
Population: 1277558000
LifeExpectancy: 71.4
GNP: 982268.00
GNPOld: 917719.00
LocalName: Zhongquo
GovernmentForm:People'sRepublic
HeadOfState: Jiang Zemin
Capital: 1891
Code2: CN
1 row in set (0.00sec)
本文来自http://yijiu.blog.51cto.com 转载请经博主允许 ,盗帖可耻!
由此可以看到此表与City表关联,如下所示
因为city表存在城市ID编号,也就是刚才与Capital= 1891 相关的信息
mysql> select *from City where id = 1891;
+------+--------+-------------+----------+------------+
| ID | Name | CountryCode | District | Population |
+------+--------+-------------+----------+------------+
| 1891 | Peking | CHN | Peking | 7472000 |
+------+--------+-------------+----------+------------+
1 row in set (0.00sec)
而CountryLanguage表里是将CountryCode做关联
看到编号为CHN,那查一下关于CHN相关的信息
可看到CountryLanguage用到以CHN作为关联可以查到相关城市
mysql> select *from CountryLanguage where CountryCode = 'CHN';
+-------------+-----------+------------+------------+
| CountryCode |Language | IsOfficial | Percentage |
+-------------+-----------+------------+------------+
| CHN | Chinese | T | 92.0 |
| CHN | Dong | F | 0.2 |
| CHN | Hui | F | 0.8 |
| CHN | Mantu | F | 0.9 |
| CHN | Miao |F | 0.7 |
| CHN | Mongolian | F | 0.4 |
| CHN | Puyi | F | 0.2 |
| CHN | Tibetan | F | 0.4 |
| CHN | Tujia | F | 0.5 |
| CHN | Uighur | F | 0.6 |
| CHN | Yi | F | 0.6 |
| CHN | Zhuang | F | 1.4 |
+-------------+-----------+------------+------------+
12 rows in set (0.00sec)
接下来就进入主题
查询语句初识 本文来自http://yijiu.blog.51cto.com 转载请经博主允许 ,盗帖可耻!
来个例子先,查找City表中的前10行
mysql> select Id,Name, Population From City limit 10;
+----+----------------+------------+
| Id | Name | Population |
+----+----------------+------------+
| 1 | Kabul | 1780000 |
| 2 | Qandahar | 237500 |
| 3 | Herat | 186800 |
| 4 | Mazar-e-Sharif | 127800 |
| 5 | Amsterdam | 731200 |
| 6 | Rotterdam | 593321 |
| 7 | Haag | 440900 |
| 8 | Utrecht | 234323 |
| 9 | Eindhoven | 201843 |
| 10 | Tilburg | 193238 |
+----+----------------+------------+
10 rows in set (0.00sec)
查找从第10行到20行,中间相差10行
limit 10,10;意思为从第几行开始并从这行开始向下显示多少行
mysql> select Id,Name, Population From City limit 10,10;
+----+-------------------+------------+
| Id | Name | Population |
+----+-------------------+------------+
| 11 | Groningen | 172701 |
| 12 | Breda | 160398 |
| 13 | Apeldoorn | 153491 |
| 14 | Nijmegen | 152463 |
| 15 | Enschede | 149544 |
| 16 | Haarlem | 148772 |
| 17 | Almere | 142465 |
| 18 | Arnhem | 138020 |
| 19 | Zaanstad | 135621 |
| 20 |s-Hertogenbosch | 129170 |
+----+-------------------+------------+
10 rows in set (0.00sec)
错误的sql:如下所示
select * from tb where xxxx limit 537793977, 20;
如果出现这样的sql,意味着先要扫描表里面的537793977行后再取20行返回,这样成本就会很高
LIMIT的一个原则:
在生产环境中使用LIMIT后只跟一个数,而且最好不大于500,如果是连续的,包括上面的sql,利用上面的SQL得到一个ID的最大值,那么这时候我们就会用到份页
如下所示:
优化前:
mysql> select Id,Name, Population From City limit 10;
+----+----------------+------------+
| Id | Name | Population |
+----+----------------+------------+
| 1 | Kabul | 1780000 |
| 2 | Qandahar | 237500 |
| 3 | Herat | 186800 |
| 4 | Mazar-e-Sharif | 127800 |
| 5 | Amsterdam | 731200 |
| 6 | Rotterdam | 593321 |
| 7 | Haag | 440900 |
| 8 | Utrecht | 234323 |
| 9 | Eindhoven | 201843 |
| 10 | Tilburg | 193238 |
+----+----------------+------------+
10 rows in set (0.00sec)
优化后如下:
使用last_max_id通过程序进行计算得到的
语法:
mysql>select Id, Name, Population From City where id ><last_max_id> limit 10;
mysql> select Id,Name, Population From City where id >10 limit 10;
+----+-------------------+------------+
| Id | Name | Population |
+----+-------------------+------------+
| 11 | Groningen | 172701 |
| 12 | Breda | 160398 |
| 13 | Apeldoorn | 153491 |
| 14 | Nijmegen | 152463 |
| 15 | Enschede | 149544 |
| 16 | Haarlem | 148772 |
| 17 | Almere | 142465 |
| 18 | Arnhem | 138020 |
| 19 | Zaanstad | 135621 |
| 20 |s-Hertogenbosch | 129170 |
+----+-------------------+------------+
10 rows in set (0.00sec)
本文来自http://yijiu.blog.51cto.com 转载请经博主允许 ,盗帖可耻!
参数解释:
where id> 10 limit 10 就是取前10行到第20行
如果取前10的话,将数据放进去,得到的就是last_max_id然后传到下一页,进行对比
这就是所谓的分页
#这并不是完全在sql里实现的, 如果想特别精准的分页,这种访问有可能有问题的。
#跳页的话有种方法:一般粗略的估计一下即可,不会让其非常精准的显示出来,只求速度足够快
比如跳页,如果对于整个系统进行搜索的话,非查不可的情况,我们可以使用专属的搜索系统进行查看,互联网领域中能不进行写则不写,以提高速度
count, max(), min()使用
count
count(*)在早版本会走主键的,最新版本会走普通索引
mysql> selectcount(*), count(id) from City;
+----------+-----------+
| count(*) | count(id)|
+----------+-----------+
| 4079 | 4079 |
+----------+-----------+
1 row in set (0.00sec)
那么问题来了:主建为何没有第二索引快?
因为Innodb主建就表本身里的数据,如果count主建,需要将整个表扫描一遍,这样建带数据读的块更大
走主建相当于把整个表都要读(全表扫描) 本文来自http://yijiu.blog.51cto.com 转载请经博主允许 ,盗帖可耻!
而在Innodb里是索引列+主建值存储的结构体系,这么做的话会更快
如果直接使用Secondary index的话会快一点
如今mysql对count(*)做了优化,默认会走Secondary index,所以在以后计算总数的时候,不要总计算列数,直接写count(*)就可以了
列出人数最多的城市
我们现在有需求,我们知道City表中有城市的总人数,我们现在想统计一下最多人数的城市
那么我们先来看一下表结构 本文来自http://yijiu.blog.51cto.com 转载请经博主允许 ,盗帖可耻!
mysql> desc City ;
+-------------+----------+------+-----+---------+----------------+
| Field | Type | Null | Key | Default | Extra |
+-------------+----------+------+-----+---------+----------------+
| ID | int(11) | NO | PRI | NULL | auto_increment |
| Name | char(35) | NO | | | |
| CountryCode |char(3) | NO | MUL | | |
| District | char(20) | NO | | | |
| Population | int(11) | NO | | 0 | |
+-------------+----------+------+-----+---------+----------------+
5 rows in set (0.00sec)
由此我们可知条件,Population为每个城市的总人数,我们只要筛出Population最大的值即可,如下所示:
mysql> select * from City where Population = (selectmax(Population) from City);
+------+-----------------+-------------+-------------+------------+
| ID | Name | CountryCode | District | Population |
+------+-----------------+-------------+-------------+------------+
| 1024 | Mumbai(Bombay) | IND | Maharashtra| 10500000 |
+------+-----------------+-------------+-------------+------------+
1 row in set (0.00sec)
括号中内的内容为子句查询;
而select max(Population) from City 表示查找这个表中人数最多的行
先来看一个例子,执行上面的sql子句:
mysql> selectmax(Population) from City;
+-----------------+
| max(Population) |
+-----------------+
| 10500000 |
+-----------------+
1 row in set (0.00sec)
max是内置函数,表示将取最大数值的行
那么将其封装到子句里面,再进行对比,即 Population= 最大的字段
找到城市人数最少的城市
mysql> select *from City where Population = (select min(Population) from City);
+------+-----------+-------------+----------+------------+
| ID | Name | CountryCode | District | Population |
+------+-----------+-------------+----------+------------+
| 2912 | Adamstown |PCN | – | 42 |
+------+-----------+-------------+----------+------------+
1 row in set (0.01sec)
使用oder by进行排序
mysql> select *from City order by Population desc limit1;
+------+-----------------+-------------+-------------+------------+
| ID | Name | CountryCode | District | Population |
+------+-----------------+-------------+-------------+------------+
| 1024 | Mumbai(Bombay) | IND | Maharashtra| 10500000 |
+------+-----------------+-------------+-------------+------------+
1 row in set (0.00sec)
本文来自http://yijiu.blog.51cto.com 转载请经博主允许 ,盗帖可耻!
查找大于人员数大于100W的城市有有哪些
使用count(*) 统计出总数
mysql> selectcount(*) from City where Population > 1000000;
+----------+
| count(*) |
+----------+
| 237 |
+----------+
1 row in set (0.00sec)
可看到 一共有237个城市
那么再来查找人口大于100W的城市是否一共有237个
mysql> select *from City where Population >1000000;
#拉到最后可看到如下的数值
+------+--------------------------+-------------+----------------------+------------+
237 rows in set (0.00sec)
我们还可以使用函数,如果什么参数都没有加的情况下就使用以下函数,会得到上一个sql的执行所得到的行数
mysql> selectfound_rows();
+--------------+
| found_rows() |
+--------------+
| 237 |
+--------------+
1 row in set (0.00sec)
本文来自http://yijiu.blog.51cto.com 转载请经博主允许 ,盗帖可耻!
那么使用count(*)并执行函数来查看效果又会是什么样
mysql> selectcount(*) from City where Population > 1000000;
+----------+
| count(*) |
+----------+
| 237 |
+----------+
1 row in set (0.00sec)
mysql> selectfound_rows();
+--------------+
| found_rows() |
+--------------+
| 1 |
+--------------+
1 row in set (0.00sec)
FOUND_ROWS函数
比如取前10行,但表中总共有多少行是未知的,这里在比其他数据中多了一个函数:
SQL_CALC_FOUND_ROWS
先来看一下yw表有多少行
mysql> selectcount(*) from yw;
+----------+
| count(*) |
+----------+
| 6000000 |
+----------+
1 row in set (15.64sec)
再执行打印前10行内容
mysql> select SQL_CALC_FOUND_ROWS * fromCity limit 10;
ERROR 1146 (42S02):Table 'test1.City' doesn't exist
mysql> select SQL_CALC_FOUND_ROWS* from yw limit 10;
+----+---------+---------+---------+---------+---------------------+------------------------------------------------------+
| id | c1 | c2 | c3 | c4 | c5 | c6 |
+----+---------+---------+---------+---------+---------------------+------------------------------------------------------+
| 1 | 463681 | 1098981 | 1817518 | 2222359 | 2014-09-24 15:38:29 |wubxwubxwubxwubxwubxwubxwubxwubxwubxwubxwubxwubx |
| 2 | 2333997 | 269341 | 2459005 | 915557 |2014-09-24 15:38:29 | wubxwubxwubx |
| 3 | 2971523 | 1226698 | 842469 | 414525 | 2014-09-24 15:38:29 | wubxwubxwubxwubxwubxwubxwubxwubxwubx |
| 4 | 2835700 | 930937 | 2835332 | 1945110 | 2014-09-24 15:38:29 | wubx |
| 5 | 1578655 | 1044887 | 2649255 | 2307696 |2014-09-24 15:38:29 | wubxwubxwubxwubxwubxwubxwubx |
| 6 | 1442242 | 992011 | 1740281 | 190626 |2014-09-24 15:38:29 | wubxwubxwubxwubxwubxwubxwubxwubxwubxwubxwubxwubxwubx |
| 7 | 693798 | 309586 | 753637 | 2403923 | 2014-09-24 15:38:29 |wubxwubxwubxwubxwubxwubxwubxwubxwubxwubx |
| 8 | 888272 | 2581335 | 1547343 | 1465295 | 2014-09-24 15:38:29 |wubxwubxwubxwubx |
| 9 | 1608599 | 240304 | 2475805 | 2157717 | 2014-09-24 15:38:29 | wubxwubxwubxwubx |
| 10 | 2833881 | 185188 | 1736996 | 565924 | 2014-09-24 15:38:29 |wubxwubxwubxwubxwubxwubxwubxwubxwubxwubx |
+----+---------+---------+---------+---------+---------------------+------------------------------------------------------+
10 rows in set (17.88sec)
mysql> selectfound_rows();
+--------------+
| found_rows() |
+--------------+
| 6000000 |
+--------------+
1 row in set (0.00sec)
6000000就是表的行数
如果并发比较大的情况下,这个函数用的还是比较多的,但是一般不会直接一个表就直接使用found_rows,一般都是跟where条件,跟上条件并计算此类型的条目数,一般此类场景中比较常见
主要就是用SQL_CALC_FOUND_ROWS来减少操作为主要目的
模糊匹配
使用like
语法:like "Wu%" #不能以百分号开头,如果使用%开头,则不会用到索引
查询以wu开头的城市
mysql> select *from City where Name like 'Wu%';
+------+-----------+-------------+---------------------+------------+
| ID | Name | CountryCode | District | Population |
+------+-----------+-------------+---------------------+------------+
| 1894 | Wuhan | CHN | Hubei | 4344600 |
| 1929 | Wuxi | CHN | Jiangsu | 830000 |
| 1964 | Wuhu | CHN | Anhui | 425740 |
| 2022 | Wuhai | CHN | Inner Mongolia | 264081 |
| 2058 | Wuzhou | CHN | Guangxi | 210452 |
| 2163 | Wuwei | CHN | Gansu | 133101 |
| 3084 | Wuppertal |DEU | Nordrhein-Westfalen | 368993 |
+------+-----------+-------------+---------------------+------------+
7 rows in set (0.01sec)
不用过多做解释了
使用sum统计其列总数
sum也是mysql内置函数,用法都是一样的,如下所示:
我们来统计中国几个重要城市的人口总数
mysql> selectsum(Population) from City;
+-----------------+
| sum(Population) |
+-----------------+
| 1429559884 |
+-----------------+
1 row in set (0.00sec)
mysql> selectsum(Population) from City where CountryCode='CHN';
+-----------------+
| sum(Population) |
+-----------------+
| 175953614 |
+-----------------+
1 row in set (0.00sec)
或只统计上海和天津的总人数
mysql> selectsum(Population) from City where CountryCode='CHN' and Name='TianJin' orName='ShangHai';
+-----------------+
| sum(Population) |
+-----------------+
| 14983100 |
+-----------------+
1 row in set (0.00sec)
使用表关联,查看出每个国家的首都
刚才我们直接使用了sum查到了首都的详细信息,如果使用表关联使用这两个表来查我们首都详细信息的话,该如何去查
因为我们知道Code:CHN,那么我们使用Capital字段与city表的id进行关联:
先来查看表内容:
mysql> select * from Country limit 1;
+------+-------+---------------+-----------+-------------+-----------+------------+----------------+--------+--------+-----------+----------------------------------------------+-------------+---------+-------+
| Code | Name | Continent | Region | SurfaceArea | IndepYear | Population | LifeExpectancy | GNP | GNPOld | LocalName | GovernmentForm | HeadOfState | Capital | Code2|
+------+-------+---------------+-----------+-------------+-----------+------------+----------------+--------+--------+-----------+----------------------------------------------+-------------+---------+-------+
| ABW | Aruba | North America | Caribbean | 193.00 | NULL | 103000 | 78.4 | 828.00 |793.00 | Aruba | NonmetropolitanTerritory of The Netherlands | Beatrix | 129 |AW |
+------+-------+---------------+-----------+-------------+-----------+------------+----------------+--------+--------+-----------+----------------------------------------------+-------------+---------+-------+
1 row in set (0.00 sec)
如上可以看到,Country表的Capital是指向国家首都城市的id,所以利用关联可以得到
可以使用Country表的Capital和City表id关联,并加上限制Country.Code='CHN';
mysql> selecta.name,b.name from Country a , City b where a.Code="CHN" and a.Capital=b.id;
+-------+--------+
| name | name |
+-------+--------+
| China | Peking |
+-------+--------+
1 row in set (0.00sec)
语法解释:
语法可分为两截:
select a.name,b.namefrom Country a , City b ; 这是一段
where a.Code="CHN" and a.Capital=b.id; 这又是一段
先来执行第一段
mysql> selecta.name,b.name from Country a , City b limit 1;
+-------+-------+
| name | name |
+-------+-------+
| Aruba | Kabul |
+-------+-------+
1 row in set (0.00sec)
可以看到,每个国家的名称以及国家的首都都呈现出来
那么我们想针对某个国家查看其首都的话肯定需要加where条件进行匹配了
where的同时,首先需要统计国家的代码号,中国无非是CHN,但是只是筛选出了CHN的城市,并没有晒出其首都,如下所示
mysql> selecta.name,b.name from Country a ,City b where a.Code='CHN' limit 10;
+-------+----------------+
| name | name |
+-------+----------------+
| China | Kabul |
| China |Qandahar |
| China | Herat |
| China |Mazar-e-Sharif |
| China |Amsterdam |
| China |Rotterdam |
| China | Haag |
| China |Utrecht |
| China | Eindhoven |
| China |Tilburg |
+-------+----------------+
10 rows in set (0.00sec)
我们还需要对其做条件匹配,我们看到,Country表中有Capital的字段,没错这是国家首都编号,而City中也存在城市的ID,那么这两个字段是相等的,所以我们只要将Counry表中的Capital和City表中的城市ID匹配上即可,如下所示:
mysql> selecta.name,b.name from Country a , City b where a.Code='CHN' and a.Capital=b.id;
+-------+--------+
| name | name |
+-------+--------+
| China | Peking |
+-------+--------+
1 row in set (0.00sec)
country为为别名a
city的别名为b
并使用where进行别名a
也就是country表中的code进行匹配
如果某行等于CHN,并且与别名b表中,也就是city表中的id匹配,则打印
如果想看每个国家的首都在哪里,则如下:
mysql> selecta.name,b.name from Country a , City b where a.Capital=b.id limit 10;
+----------------------+------------------+
| name | name |
+----------------------+------------------+
| Aruba | Oranjestad |
| Afghanistan | Kabul |
| Angola | Luanda |
| Anguilla | The Valley |
| Albania | Tirana |
| Andorra | Andorra la Vella |
| NetherlandsAntilles | Willemstad |
| United ArabEmirates | Abu Dhabi |
| Argentina | Buenos Aires |
| Armenia | Yerevan |
+----------------------+------------------+
10 rows in set (0.00sec)
将其他表中的数据导入到当前表
创建一个测试表及相关操作
mysql> createtable test5 like City;
Query OK, 0 rowsaffected (0.02 sec)
从其他表往新表中导入记录
mysql> insert intotest5 select * from City limit 100;
Query OK, 100 rowsaffected (0.00 sec)
Records: 100 Duplicates: 0 Warnings: 0
当在线上环境,想调整索引,但是原表确大,不容易操作时,可以上面的方法建一个测试表,测试完后再删掉就可以了
使用update更新表数据
使用update将北京改名
mysql> updatetest5 set Name='Peking', District='Peking' where id=1;
Query OK, 1 rowaffected (0.02 sec)
Rows matched: 1 Changed: 1 Warnings: 0
如果需要更新多个列的话,必须以逗号间隔开来
UPDATE重大标识:
如果update没有加条件等于全表更新,所以执行update的时候一定要加where条件
或者把事务开起来,或是把自动提交关了
begin;
update test5 set Name='Peking', District='aaa' where id=1;
select * from test5 where id=1; #确认一下
rollback; or commit; #没有问题再提交操作
delete 操作:
删除添加的第一条记录,也要考虑事务操作:
begin;
delete from wubx_1 where id=1;
select * from wubx_1 where id=1;
rollback |commit
以上,为MySQL的日常sql实例,感谢各位看官
|