MySql存储过程总结

原文:MySql存储过程总结

      之前总是在MSSQL上写存储过程，没有在MYSQL上写过，也基本没有用过，今天需要用到MYSQL，研究了下，把项目的需要的存储过程写了一部分，写一下工作总结。这里没有给出数据库结构，不讨论SQL语句的细节，主要探讨存储过程语法，适合有基础的人。

 发表地址：http://www.cnblogs.com/zxlovenet/p/3783136.html

#查询文章回复
-- ----------------------------
-- Procedure structure for `sp_select_reply_article`
-- ----------------------------
DROP PROCEDURE IF EXISTS `sp_select_reply_article`;
DELIMITER ;;
CREATE DEFINER=`root`@`localhost` PROCEDURE `sp_select_reply_article`(IN `ra_id` int,IN `pagefrom` int,IN `pagesize` int)
BEGIN
         #Routine body goes here...
         SET @ra_id = ra_id;
         SET @pagefrom = pagefrom;
         SET @pagesize = pagesize;
         SET @ssra = CONCAT('SELECT * FROM gk_article WHERE id = ? LIMIT ?,?');
         PREPARE sqlquery FROM @ssra;
         EXECUTE sqlquery USING @ra_id,@pagefrom,@pagesize;
END
;;
DELIMITER ;

#技术点1：MySql5.1不支持LIMIT参数（MySql5.5就支持了），如果编写存储过程时使用LIMIT做变量，那是需要用动态SQL来构建的，而这样做性能肯定没有静态SQL好。主要代码如下：

         SET @ssra = CONCAT('SELECT * FROM gk_article WHERE id = ? LIMIT ?,?');

         PREPARE sqlquery FROM @ssra;

         EXECUTE sqlquery USING @ra_id,@pagefrom,@pagesize;

#技术点2：如果同时需要返回受影响行数需要在语句后面添加语句：ROW_COUNT()函数，两条语句之间需要“;”分隔。

  发表地址：http://www.cnblogs.com/zxlovenet/p/3783136.html

#更新数据
-- ----------------------------
-- Procedure structure for `sp_update_permission`
-- ----------------------------
DROP PROCEDURE IF EXISTS `sp_update_permission`;
DELIMITER ;;
CREATE DEFINER=`root`@`localhost` PROCEDURE `sp_update_permission`(IN `puser_uid` varchar(20),IN `plevel` int,IN `ppower` int)
BEGIN
         #Routine body goes here...

         SET @puser_uid = puser_uid;
         SET @plevel = plevel;
         SET @ppower = ppower;
         UPDATE gk_permission SET `level` = @plevel, power = @ppower WHERE user_uid = CONVERT(@puser_uid USING utf8) COLLATE utf8_unicode_ci;
END
;;
DELIMITER ;

#技术点3：MySQL进行字符串比较时发生错误（Illegal mix of collations (utf8_unicode_ci,IMPLICIT) and (utf8_general_ci,IMPLICIT) for operation '='），解决方法：将比较等式一边进行字符串转换，如改为“CONVERT(b.fullCode USING utf8) COLLATE utf8_unicode_ci”，主要代码如下：

         UPDATE gk_permission SET `level` = @plevel, power = @ppower WHERE user_uid = CONVERT(@puser_uid USING utf8) COLLATE utf8_unicode_ci;

#插入数据
-- ----------------------------
-- Procedure structure for `sp_insert_user`
-- ----------------------------
DROP PROCEDURE IF EXISTS `sp_insert_user`;
DELIMITER ;;
CREATE DEFINER=`root`@`localhost` PROCEDURE `sp_insert_user`(IN `uid` varchar(20),IN `upw` varchar(32),IN `name` varchar(20),IN `sex` int,IN `phone` varchar(20),IN `u_id` int,IN `s_id` int,IN `j_id` int)
BEGIN
         #Routine body goes here...
         SET @uid = uid;
         SET @upw = upw;
         SET @uname = uname;
         SET @sex = sex;
         SET @phone = phone;
         #由于外键约束，所以添加的外键字段需要在对应外键所在表有相应数据
         SET @u_id = u_id;
         SET @s_id = s_id;
         SET @j_id = j_id;
         SET @verifytime = DATE('0000-00-00');
         INSERT INTO gk_user(uid,upw,uname,sex,phone,u_id,s_id,j_id,verifytime) 　　　　　　　VALUES(@uid,@upw,@uname,@sex,@phone,@u_id,@s_id,@j_id,@verifytime);
         #查询结果会自动返回受影响行数
END
;;
DELIMITER ;

  发表地址：http://www.cnblogs.com/zxlovenet/p/3783136.html

#根据ID删除数据
-- ----------------------------
-- Procedure structure for `sp_delete_exchange_by_id`
-- ----------------------------
DROP PROCEDURE IF EXISTS `sp_delete_exchange_by_id`;
DELIMITER ;;
CREATE DEFINER=`root`@`localhost` PROCEDURE `sp_delete_exchange_by_id`(IN `eid` int)
BEGIN
         #Routine body goes here...
         SET @eid = eid;
         DELETE FROM gk_exchange WHERE id = @eid;
END
;;
DELIMITER ;

#通过账号查询用户或者管理员
-- ----------------------------
-- Procedure structure for `sp_select_user_by_uid`
-- ----------------------------
DROP PROCEDURE IF EXISTS `sp_select_user_by_uid`;
DELIMITER ;;
CREATE DEFINER=`root`@`localhost` PROCEDURE `sp_select_user_by_uid`(IN `uid` varchar(20),IN `getAdmin` int)
BEGIN
         #Routine body goes here...
         SET @uid = uid;
         #SET @getadmin = getAdmin;
         #查询管理员
         IF (getAdmin = 1) THEN
                   SELECT us.*, un.`name`, se.`name`, jo.`name`, pe.`level`, pe.power FROM gk_user AS us, gk_unit AS un, gk_section AS se, gk_jobtitle AS jo, gk_permission AS pe WHERE us.u_id = un.id AND us.s_id = se.id AND us.j_id = jo.id AND us.uid = pe.user_uid AND us.uid = CONVERT(@uid USING utf8) COLLATE utf8_unicode_ci;
         END IF;
         #查询用户
         IF (getAdmin = 0) THEN
                   SELECT us.*, un.`name`, se.`name`, jo.`name` FROM gk_user AS us, gk_unit AS un, gk_section AS se, gk_jobtitle AS jo WHERE us.u_id = un.id AND us.s_id = se.id AND us.j_id = jo.id AND us.uid = CONVERT(@uid USING utf8) COLLATE utf8_unicode_ci;
         END IF;
END
;;
DELIMITER ;

#技术点4：这个存数过程需要用到控制语句（if else elseif while loop repeat leave iterate）。

         IF (getAdmin = 1) THEN

                   #语句…

         END IF;

  发表地址：http://www.cnblogs.com/zxlovenet/p/3783136.html

#技术点5：在传入参数不匹配的情况下报错（Column count doesn't match value count at row 1），这个就是细心问题了，详细检查参数吧。

#技术点6：获取当前时间的函数：NOW()

#技术点7：“`”这个符号是反单引号，两个反单引号夹起来的会被当做变量，一般是在定义字段时遇到关键字冲突的时候会用到。