DTCC大会上，阿里江疑的演讲中提到一个：select from update hot row;
不明白如何在Oracle中实现的，他的意思是在一条SQL中实现update和select这条update的字段信息。
经dbsnake指点，了解到这是模仿了Oracle的returning into子句，可以将使用的DML语句影响的行记录的指定列的值select出来。
官方文档中有示例：
http://docs.oracle.com/cd/B19306_01/appdev.102/b14261/tuning.htm
You can use the BULK COLLECT clause in the RETURNING INTO clause of an INSERT, UPDATE, or DELETE statement:
Example 11-15 Using BULK COLLECT With the RETURNING INTO Clause
CREATE TABLE emp_temp AS SELECT * FROM employees;
DECLARE
TYPE NumList IS TABLE OF employees.employee_id%TYPE;
enums NumList;
TYPE NameList IS TABLE OF employees.last_name%TYPE;
names NameList;
BEGIN
DELETE FROM emp_temp WHERE department_id = 30
RETURNING employee_id, last_name BULK COLLECT INTO enums, names;
DBMS_OUTPUT.PUT_LINE('Deleted ' || SQL%ROWCOUNT || ' rows:');
FOR i IN enums.FIRST .. enums.LAST
LOOP
DBMS_OUTPUT.PUT_LINE('Employee #' || enums(i) || ': ' || names(i));
END LOOP;
END;
/
上面例子对于不熟悉PLSQL的不是很好理解，用一个简单的示例说明：
1. 创建测试表：
create table tbl_returninto(
id number,
remark varchar2(5));
SQL> select * from tbl_returninto;
ID REMARK
---------- --------------------------------------------------
2 one
3 two
4 three
2. 插入一条记录，使用returning into在同一条SQL中获得插入的id值：
SQL> declare
2 l_id tbl_returninto.id%type;
3 begin
4 insert into tbl_returninto values(tr_seq.nextval, 'one')
5 returning id into l_id;
6 commit;
7 dbms_output.put_line('id=' || l_id);
8 end;
9 /
id=1
PL/SQL procedure successfully completed.
3. 更新和删除一条记录，使用returning into获得更新和删除的id值：
SQL> declare l_id tbl_returninto.id%type;
2 begin
3 update tbl_returninto
4 set remark = 'one2'
5 where id = 2
6 returning id into l_id;
7 dbms_output.put_line('UPDATE ID=' || l_id);
8 delete from tbl_returninto where remark = 'three'
9 returning id into l_id;
10 dbms_output.put_line('DELETE ID=' || l_id);
11 commit;
12 end;
13 /
UPDATE ID=2
DELETE ID=4
PL/SQL procedure successfully completed.
总结：
使用returning into子句可以在一条SQL中将insert、update和delete影响的行记录指定字段信息select出来，其中insert和update都是执行之后的结果，delete是执行之前的结果。当然，其实这里用的是PLSQL的语法实现。
