oracel 拆分字符串

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：

oracel 拆分字符串-LMLPHP

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：

oracel 拆分字符串-LMLPHP

作者：李敬然（Gnie）出处：{GnieTech} （http://www.cnblogs.com/gnielee/）

方法二、

select regexp_substr('Hello,Cnblogs!','[^,]+',1,level) from dual connect by level <= length('Hello,Cnblogs!')-length(REPLACE('Hello,Cnblogs!',',',''))+1;