本函数可以将“目标字符串”以“指定字符串”进行拆分,并通过表结构返回结果。代码如下:
CREATE OR REPLACE TYPE str_split IS TABLE OF VARCHAR2 (4000); CREATE OR REPLACE FUNCTION splitstr(p_string IN VARCHAR2, p_delimiter IN VARCHAR2) RETURN str_split PIPELINED AS v_length NUMBER := LENGTH(p_string); v_start NUMBER := 1; v_index NUMBER; BEGIN WHILE(v_start <= v_length) LOOP v_index := INSTR(p_string, p_delimiter, v_start); IF v_index = 0 THEN PIPE ROW(SUBSTR(p_string, v_start)); v_start := v_length + 1; ELSE PIPE ROW(SUBSTR(p_string, v_start, v_index - v_start)); v_start := v_index + 1; END IF; END LOOP; RETURN; END splitstr;
创建完毕后,我们来测试一下,例如执行如下SQL:
select * from table(splitstr(‘Hello,Cnblogs!‘,‘,‘));
其输出结果为一个两行的表,如下图:
将行转为列显示:
select a.column_value v1,b.column_value v2 from (select * from (select rownum rn,t.* from table(splitstr(‘Hello,Cnblogs!‘,‘,‘)) t)) a, (select * from (select rownum rn,t.* from table(splitstr(‘Hello,Cnblogs!‘,‘,‘)) t)) b where a.rn=1 and b.rn=2
如图: