前段时间和同学接的项目终于完工了，项目主要是做报表，涉及到了很多的sql语句，所以借此总结下。

一、基础

1.数据库相关命令

a>.创建数据库

CREATE DATABASE 数据库名

b>.创建数据库并设置默认字符集

CREATE DATABASE 数据库名 DEFAULT CHARACTER SET utf8;

c>.删除数据库

DROP DATABASE 数据库名

2.数据表相关命令

a>.删除表

DROP TABLE 表名;

b>.新建表

DROP TABLE IF EXISTS 表名;
CREATE TABLE 表名 (
`id`  INT NOT NULL AUTO_INCREMENT,
`username`  VARCHAR(60) NOT NULL DEFAULT '' COMMENT '登录名',
`password`  CHAR(32) NOT NULL DEFAULT '' COMMENT '登录密码',
 PRIMARY KEY (`id`),   -- 主键
 INDEX (`username`)    -- 索引
) ENGINE=INNODB DEFAULT CHARSET=utf8 COMMENT='新建表';

c>.修改表名

ALTER TABLE 表名1 RENAME TO 表名2; 

d>.添加列

ALTER TABLE 表名 ADD COLUMN 列名 列类型（如：VARCHAR(30)); 

e>.删除列

ALTER TABLE 表名 DROP COLUMN 列名; 

f>.几条简单的sql语句

选择：SELECT * FROM 表名 WHERE 范围;
插入：INSERT INTO 表名(字段名1,字段名2) VALUES(字段值1,字段值2);
删除：DELETE FROM 表名 WHERE 范围;
更新：UPDATE 表名 SET 字段名=字段值 WHERE 范围;
查找：SELECT * FROM 表名 WHERE 字段名 LIKE ’% 字段值 %’ ;
排序：SELECT * FROM 表名 ORDER BY 字段名1,字段名2 [DESC];
总数：SELECT COUNT AS totalcount FROM 表名;
求和：SELECT SUM(字段值) AS sumvalue FROM 表名;
平均：SELECT AVG(字段名) AS avgvalue FROM 表名;
最大：SELECT MAX(字段名) AS maxvalue FROM 表名;
最小：SELECT MIN(字段名) AS minvalue FROM 表名;

g>.赋予数据库用户权限（每个数据库都创建单独的用户）

CREATE USER '项目用户'@'localhost' IDENTIFIED BY '项目用户密码';  
GRANT ALL PRIVILEGES ON 数据库名称.* TO 项目用户@'localhost' ; 

或者

GRANT ALL PRIVILEGES ON 数据库名称.* TO 用户名@'%' IDENTIFIED BY '用户密码';(用’%'表示从任何地址连接)

h>.添加主键

ALTER TABLE 表名 ADD PRIMARY KEY(列名);

i>.创建索引

CREATE INDEX 字段名 ON 表名;

j>.删除索引

DROP INDEX 字段名;

k>.创建视图

CREATE VIEW 视图名 AS
SELECT 语句;

l>.删除视图

DROP VIEW 视图名;

m>.联合查询（这块有很多内容，我这里只提出基本的）

结果集列联合：JOIN ，这里我常用的是LEFT JOIN ，格式如下：

SELECT a.a, a.b, b.c, b.d FROM a LEFT OUT JOIN b ON a.a = b.c;

结果集行联合：UNION或者UNION ALL，格式如下（注意列的名称和个数一定要一致）：

SELECT a.a as col1, a.b as col2 from a
UNION ALL
SELECT b.c as col1, b.d as col2 from b;

n>.分组:GROUP BY 一般与count,sum,max,min,avg联合使用

SELECT SUM(字段名1) FROM 表名 GROUP BY 字段名2;

o>.前10条记录

SELECT * FROM 表名 LIMIT 0,9;(mysql没有top语法，limit一般常用来分页)

p>.排序（desc降序；asc升序；默认是升序）

SELECT * FROM 表名 ORDER BY 列名 DESC;

二、进阶

a>.去重，一般常用 DISTINCT

SELECT DISTINCT * FROM 表名;
SELECT DISTINCT(列名) FROM 表名;

b>.复制表

CREATE TABLE 表名2 AS   
(   
  SELECT * FROM 表名1   
)   
     

c>.查询分析器:在sql前面加上EXPLAIN或者DESC，查看该语句执行情况

EXPLAIN  SELECT * FROM 表名;

三、贴一下做报表时写的一些恶心的sql

1.根据查询条件查询出符合条件的项目编号

SELECT DISTINCT(xp.Id)
FROM xmgl_finance_plan xfp LEFT JOIN xmgl_project xp ON xfp.proId = xp.Id WHERE 1=1
AND xp.proCategory IN ('ITER973')
AND xp.Id IN(0,1,2);

2.-- 查询项目计划表里对应的所有的预算科目对应的合计金额

SELECT
  t.Pid,
  xsb2.subjectName,
  SUM(t.planTotalAmount)    planTotalAmount,
  SUM(t.planCurrentAmount)    planCurrentAmount,
  SUM(t.total_amount)    total_amount,
  SUM(t.payedMoney)    payedMoney,
  SUM(t.finalPayment)    finalPayment,
  SUM(t.CurYearPayedMoney)    CurYearPayedMoney
FROM (SELECT
        xsb.Id,
        xsb.subjectName,
        xsb.Pid,
        SUM(xfp.planTotalAmount)    planTotalAmount,
        SUM(xfp.planCurrentAmount)    planCurrentAmount,
        SUM(xfp.total_amount)    total_amount,
        SUM(xfp.payedMoney)    payedMoney,
        SUM(xfp.finalPayment)    finalPayment,
        SUM(xfp.CurYearPayedMoney)    CurYearPayedMoney
      FROM xmgl_subject_budget xsb
        LEFT JOIN xmgl_project xp
          ON xsb.type = xp.subjectType
        LEFT JOIN xmgl_finance_plan xfp
          ON xsb.Id = xfp.subjectId
      WHERE xp.Id = 1
      GROUP BY xfp.subjectId,xsb.Id
      ORDER BY xsb.Id) AS t
  LEFT JOIN xmgl_subject_budget xsb2
    ON t.Pid = xsb2.Id
WHERE t.Pid <> 0
GROUP BY t.Pid;

3.-- 查询各个部门在特定科目下的控制额度及支出信息(优化)

SELECT DISTINCT(xcs1.dept),xsc.subjectName,xcs2.amount,COALESCE(SUM(xfa.amount),0) payedMoney
,xcs2.amount-SUM(xfa.amount) leftAmount
FROM xmgl_control_subject xcs1
LEFT JOIN xmgl_control_subject xcs2 ON xcs1.dept = xcs2.dept AND xcs2.detailId = 3
LEFT JOIN xmgl_subject_control xsc ON xcs2.detailId = xsc.Id
LEFT JOIN xmgl_finance_activity xfa ON xcs2.dept = xfa.department AND xfa.detailId IN('22','23')
GROUP BY xcs1.dept,xcs1.detailId;

4.类似于QQ空间的消息列表

-- 赞了我
SELECT tf.id,1 AS typeId,c.avatar AS avatar, c.name,ct.catalog AS catalog, ct.id AS messageId, ct.content AS trendsContent,NULL AS replyContent, tf.create_time AS createTime,tf.read AS isRead
FROM trends_fork tf LEFT JOIN cust_trends ct ON tf.trends_id = ct.id LEFT JOIN customer c ON c.id = tf.customer_id
WHERE tf.master_id = 1
UNION ALL
-- 给我留言
SELECT cb.id,2 AS typeId,c.avatar AS avatar,c.name, 3 AS catalog,cb.id AS messageId,NULL AS trendsContent,cb.content AS replyContent,cb.create_time AS createTime,cb.read AS isRead
FROM cust_board cb LEFT JOIN customer c ON cb.customer_id = c.id WHERE cb.master_id = 1
UNION ALL
-- 评论我的动态
SELECT  tr.id,3 AS typeId, c.avatar AS avatar, c.name, ct.catalog AS catalog, ct.id AS messageId, ct.content AS trendsContent, tr.content AS replyContent,tr.create_time AS createTime,tr.read AS isRead
FROM trends_reply tr LEFT JOIN cust_trends ct ON tr.trends_id = ct.id LEFT JOIN customer c ON c.id = tr.customer_id
WHERE tr.master_id = 1
UNION ALL
-- 回复了我评论的消息（包括直接回复我的评论、在我的动态下评论别人的回复）
SELECT tra.id,4 AS typeId,c.avatar AS avatar, c.name, ct.catalog AS catalog, tra.trends_id AS messageId,ct.content AS trendsContent, tra.content AS replyContent, tra.create_time AS createTime,tra.read AS isRead
FROM trends_replyat tra LEFT JOIN  trends_reply tr ON tra.reply_id = tr.id LEFT JOIN customer c ON c.id = tra.customer_id LEFT JOIN cust_trends ct ON ct.id = tr.trends_id
WHERE tra.at = 1 OR (tra.master_id = 1 AND tra.customer_id != 1)
UNION ALL
-- 留言板中回复我的消息
SELECT cbr.id,5 AS typeId,c.avatar AS avatar, c.name, 3 AS catalog, cb.id AS messageId,cb.content AS trendsContent,cbr.content AS replyContent, cbr.create_time AS createTime,cbr.read AS isRead
FROM cust_board_reply cbr LEFT JOIN cust_board cb ON cbr.board_id = cb.id LEFT JOIN customer c ON cbr.customer_id = c.id
WHERE (cbr.board_customer_id = 1 AND cbr.customer_id != 1 ) OR (cbr.master_id = 1 AND cbr.customer_id != 1) ORDER BY  createTime DESC
LIMIT 2,10