Oracle中删除重复记录

查找重复记录，比如sprojid是区分记录的key
SQL> select "YSZC115"."TBCDB0"."SPROJID"
2 from "YSZC115"."TBCDB0" group by sprojid having count(sProjID)>1;

SPROJID
------------------------------
0020000200603003001
0020000200603003002
0020000200603003008
0020000200603003009
0020000200603003011
0020000200603003012
0020000200603003013
0020000200603003014
0020000200603003017
0020000200603003020
0020000200603003022

SPROJID
------------------------------
0020000200603003024
0020000200603003025
0020000200603003027

已选择14行。

实际存在的记录：
SQL> select sProjID from yszc115.tbcdb0 where sProjID in (select "YSZC115"."TBCD
B0"."SPROJID"
2 from "YSZC115"."TBCDB0" group by sprojid having count(sProjID)>1);

SPROJID
------------------------------
0020000200603003001
0020000200603003001
0020000200603003002
0020000200603003002
0020000200603003008
0020000200603003008
0020000200603003009
0020000200603003009
0020000200603003011
0020000200603003011
0020000200603003012

SPROJID
------------------------------
0020000200603003012
0020000200603003013
0020000200603003013
0020000200603003014
0020000200603003014
0020000200603003017
0020000200603003017
0020000200603003020
0020000200603003020
0020000200603003022
0020000200603003022

SPROJID
------------------------------
0020000200603003024
0020000200603003024
0020000200603003025
0020000200603003025
0020000200603003027
0020000200603003027

已选择28行。

删除重复的

SQL> delete yszc115.tbcdb0 where sProjID in (select "YSZC115"."TBCDB0"."SPROJID"

2 from "YSZC115"."TBCDB0" group by sprojid having count(sProjID)>1) and
rowid not in
3 (select min(rowid) from yszc115.tbcdb0 group by sprojid having count(sProjI
D)>1);

已删除14行。

删除重复记录的方法原理：
(1).在Oracle中，每一条记录都有一个rowid，rowid在整个数据库中是唯一的，rowid确定了每条记录是在Oracle中的哪一个数据文件、块、行上。

　　(2).在重复的记录中，可能所有列的内容都相同，但rowid不会相同，所以只要确定出重复记录中那些具有最大rowid的就可以了，其余全部删除。