SQL函数 LISTAGG/string_agg/group_concat

LISTAGG/string_agg/group_concat

oracle函数listagg使用

oracle函数listagg使用

作用
可以实现将多列记录聚合为一列记录，实现数据的压缩

语法结构
listagg(measure_expr，delimiter) within group ( order by order_by_clause);

解释：measure_expr可以是基于任何列的表达式

delimiter分隔符，默认为NULL

order_by_clause决定了列值的拼接顺序

获取某列并以逗号分隔：

--listagg
C##SCOTT@LHRCDB> select listagg(ename,',') from emp;

LISTAGG(ENAME,',')
----------------------------------------------------------------------------------------------------
SMITH,ALLEN,WARD,JONES,MARTIN,BLAKE,CLARK,SCOTT,KING,TURNER,ADAMS,JAMES,FORD,MILLER

C##SCOTT@LHRCDB> 
--string_agg
qbadmin@10.14.41.157:20158/wind> select string_agg(ename,',') from emp;
                                string_agg
---------------------------------------------------------------------------
  SMITH,ALLEN,WARD,JONES,MARTIN,BLAKE,CLARK,KING,TURNER,JAMES,FORD,MILLER
(1 row)


Time: 4ms total (execution 4ms / network 0ms)

qbadmin@10.14.41.157:20158/wind> 
--group_concat
mysql> select * from t1;
+------+------+
| c1   | c2   |
+------+------+
|    1 | a    |
|    2 | abc  |
+------+------+
2 rows in set (0.00 sec)
mysql> select group_concat(c2) from t1;
+------------------+
| group_concat(c2) |
+------------------+
| a,abc            |
+------------------+
1 row in set (0.00 sec)

mysql> 

分组

举例
普通函数，对工资进行排序，并按照逗号进行拼接

--listagg
C##SCOTT@LHRCDB> select deptno,listagg(ename,',')within group(order by sal)name from emp group by deptno;

    DEPTNO NAME
---------- --------------------------------------------------
        10 MILLER,CLARK,KING
        20 SMITH,ADAMS,JONES,FORD,SCOTT
        30 JAMES,MARTIN,WARD,TURNER,ALLEN,BLAKE

C##SCOTT@LHRCDB> 
--string_agg
wind=# select deptno,string_agg(ename,',' order by sal) from emp  group by deptno;
 deptno |              string_agg              
--------+--------------------------------------
     10 | MILLER,CLARK,KING
     20 | SMITH,ADAMS,JONES,SCOTT,FORD
     30 | JAMES,WARD,MARTIN,TURNER,ALLEN,BLAKE
(3 rows)

wind=# 
--group_concat
mysql> select deptno,group_concat(ename order by sal SEPARATOR  ',') from emp group by deptno;
+--------+-------------------------------------------------+
| deptno | group_concat(ename order by sal SEPARATOR  ',') |
+--------+-------------------------------------------------+
|     10 | MILLER,CLARK,KING                               |
|     20 | SMITH,ADAMS,JONES,FORD,SCOTT                    |
|     30 | JAMES,MARTIN,WARD,TURNER,ALLEN,BLAKE            |
+--------+-------------------------------------------------+
3 rows in set (0.00 sec)

mysql> 

作为分析函数使用

--listagg
C##SCOTT@LHRCDB> select deptno,ename,sal,listagg(ename,',')within group(order by sal)over(partition by deptno)name from emp;

    DEPTNO ENAME             SAL NAME
---------- ---------- ---------- --------------------------------------------------
        10 MILLER           1300 MILLER,CLARK,KING
        10 CLARK            2450 MILLER,CLARK,KING
        10 KING             5000 MILLER,CLARK,KING
        20 SMITH             800 SMITH,ADAMS,JONES,SCOTT,FORD
        20 ADAMS            1100 SMITH,ADAMS,JONES,SCOTT,FORD
        20 JONES            2975 SMITH,ADAMS,JONES,SCOTT,FORD
        20 SCOTT            3000 SMITH,ADAMS,JONES,SCOTT,FORD
        20 FORD             3000 SMITH,ADAMS,JONES,SCOTT,FORD
        30 JAMES             950 JAMES,MARTIN,WARD,TURNER,ALLEN,BLAKE
        30 MARTIN           1250 JAMES,MARTIN,WARD,TURNER,ALLEN,BLAKE
        30 WARD             1250 JAMES,MARTIN,WARD,TURNER,ALLEN,BLAKE
        30 TURNER           1500 JAMES,MARTIN,WARD,TURNER,ALLEN,BLAKE
        30 ALLEN            1600 JAMES,MARTIN,WARD,TURNER,ALLEN,BLAKE
        30 BLAKE            2850 JAMES,MARTIN,WARD,TURNER,ALLEN,BLAKE

已选择 14 行。

C##SCOTT@LHRCDB> 

Oracle分析(窗口)函数