判断字段是否存在的方法总结如下:
1.查找系统表
select TABLE_SCHEMA,TABLE_NAME,COLUMN_NAME from information_schema.COLUMNS where COLUMN_NAME='uu';
2.使用describe
describe cdb_posts first
存在第一列返回字段的名称,不存在就返回null,
删除方法:
如果删除的时候涉及的表不多的话,直接:
alter table tb_name drop column col_name;
多的话,可以使用下面的方法:
存储过程删除
DELIMITER $$<br><br>DROP PROCEDURE IF EXISTS `test`.`p_drop_uuid_uuname`$$<br><br>CREATE DEFINER=`root`@`%` PROCEDURE `p_drop_uu`()<br>BEGIN<br> declare _db_name char(30);<br> declare _tb_name char(30);<br> declare _col_name char(30);<br> declare no_more_row tinyint(1);<br> <br> declare cur_uuid cursor for <br> select TABLE_SCHEMA,TABLE_NAME,COLUMN_NAME <br> from information_schema.COLUMNS <br> where COLUMN_NAME='uu';<br> declare continue handler for not found set no_more_row=1;<br> set no_more_row=0; -- 判断是否结束的标志位<br> open cur_uuid;<br> repeat <br> fetch cur_uuid into _db_name,_tb_name,_col_name;-- 取记录<br> -- select _db_name,_tb_name,_col_name;<br> set @_dt = concat("alter table ", _db_name,".", _tb_name, " drop column uu");<br> -- 在存储过程中,想把一个变量当作SQL执行,只有用prepare;<br> prepare s1 from @_dt;<br> execute s1;<br> deallocate prepare s1;<br> until no_more_row<br> end repeat;<br> close cur_uuid;<br> END$$