之前在写删除数据库重复记录的时候没有用存储过程代码如下：

表（book）字段

（id   ,    b_name,   b_describe,     b_author）

1   a,    b,     c

2   a,    b,     c

3   e,    r,     h

4   e,    r,     h

5   a,    r,     c

delete from book

where

b_name in (select t2.b_name from book t2 group by t2.b_name,t2.b_describe,t2.b_author having COUNT(*)>1)

and b_describe age in (select t2.b_describe from book t2 group by t2.b_name,t2.b_describe,t2.b_author  having COUNT(*)>1)

and b_author age in (select t2.b_author from book t2 group by t2.b_name,t2.b_describe,t2.b_author  having COUNT(*)>1)

and id not in(select MIN(t2.id) from from book t2 group by t2.b_name,t2.b_describe,t2.b_author  having COUNT(*)>1)

使用上面的代码可以把完全重复记录删除，但同时也会把第5条记录删除；

后来就改用存储过程的方法如下：

CREATE PROCEDURE del_pointer()
BEGIN
 /* 定义变量一 */
    DECLARE paramId int;
    DECLARE paramName VARCHAR(16);
    DECLARE paramDes VARCHAR(16);
    DECLARE paramAut VARCHAR(16);
    DECLARE  _done int default 0;
 /* 定义光标 */
    DECLARE _Cur CURSOR FOR
            SELECT id,b_name,b_describe,b_author FROM my_book GROUP BY b_name,b_describe,b_author HAVING COUNT(*)>1;
    DECLARE CONTINUE HANDLER FOR SQLSTATE '02000' SET _done = 1;#错误定义，标记循环结束
 /* 打开光标 */
    OPEN _Cur;
     FETCH _Cur INTO paramId, paramName, paramDes, paramAut;
   /* 循环执行 */
       WHILE ( _done <> 1) DO
         DELETE from my_book WHERE id<>paramId and b_name=paramName and b_describe=paramDes and b_author=paramAut;
        /*游标向下走一步*/
        FETCH _Cur INTO paramId, paramName, paramDes, paramAut;
      END WHILE;
 /*关闭光标*/
    CLOSE _Cur;
END;

这样就不会删除上面第5条记录，只是今天刚写的，所以记下了，希望多多指点！