最近一直在研究计算产品BOM的成本系数，将拼接的元件用量拼接后拆分计算是个问题，后来受到大佬在mysql中截取字符串的启发在oracle中以substr和instr实现了
1.以下是我在mysql中重写的代码
mysql中使用substring_index
SUBSTRING_INDEX(str, delimiter, count)
返回一个 str 的子字符串，在 delimiter 出现 count 次的位置截取。
如果 count > 0，从则左边数起，且返回位置前的子串；
如果 count < 0，从则右边数起，且返回位置后的子串。
delimiter 是大小写敏感，且是多字节安全的。
取头字符串方法substring_index(gs, '*', 1)
取尾字符串方法substring_index(gs, '*', -1)
取中间字符串方法substring_index(substring_index(gs, '*', 3), '*', -1)
#1.创建测试数据
create table testgs3 as
select '1.1*4.52*1.1' gs
union all
select '2.5*4.5*1.5*2.2' x1
union all
select '1.0' x1
union ALL
select '2*.5*2.5';
#2.按测试数据进行截取计算
select gs,
how,
case
when how = 0 then gs
when how = 1 then substring_index(gs,'*', 1) *substring_index(gs,'*', -1)
when how = 2 then
substring_index(gs, '*', 1) * substring_index(substring_index(gs, '*', 2), '*', -1) *substring_index(gs, '*', -1)
when how = 3 then
substring_index(gs, '*', 1) *substring_index(substring_index(gs, '*', 2), '*', -1) * substring_index(substring_index(gs, '*', 3),'*', -1) * substring_index(gs, '*' ,- 1)
when how = 4 then
substring_index(gs, '*', 1) *substring_index(substring_index(gs, '*', 2), '*', -1) *substring_index(substring_index(gs, '*', 3), '*', -1) *substring_index(substring_index(gs, '*', 4), '*', -1) * substring_index(gs, '*', -1)
when how = 5 then
substring_index(gs, '*', 1) *substring_index(substring_index(gs, '*', 2), '*', -1) *substring_index(substring_index(gs, '*', 3), '*', -1) *substring_index(substring_index(gs, '*', 4), '*', -1)
*substring_index(substring_index(gs, '*', 5), '*', -1)* substring_index(gs, '*', -1)
when how = 6 then
substring_index(gs, '*', 1) *substring_index(substring_index(gs, '*', 2), '*', -1) *substring_index(substring_index(gs, '*', 3), '*', -1) *substring_index(substring_index(gs, '*', 4), '*', -1)
*substring_index(substring_index(gs, '*', 5), '*', -1)*substring_index(substring_index(gs, '*', 6), '*', -1)* substring_index(gs, '*', -1)
when how = 7 then
substring_index(gs, '*', 1) *substring_index(substring_index(gs, '*', 2), '*', -1) *substring_index(substring_index(gs, '*', 3), '*', -1) *substring_index(substring_index(gs, '*', 4), '*', -1)
*substring_index(substring_index(gs, '*', 5), '*', -1)*substring_index(substring_index(gs, '*', 6), '*', -1)* substring_index(substring_index(gs, '*', 7), '*', -1)*substring_index(gs, '*', -1)
when how = 8 then
substring_index(gs, '*', 1) *substring_index(substring_index(gs, '*', 2), '*', -1) *substring_index(substring_index(gs, '*', 3), '*', -1) *substring_index(substring_index(gs, '*', 4), '*', -1)
*substring_index(substring_index(gs, '*', 5), '*', -1)*substring_index(substring_index(gs, '*', 6), '*', -1)* substring_index(substring_index(gs, '*', 7), '*', -1)
*substring_index(substring_index(gs, '*', 8), '*', -1)*substring_index(gs, '*', -1)
when how = 9 then
substring_index(gs, '*', 1) *substring_index(substring_index(gs, '*', 2), '*', -1) *substring_index(substring_index(gs, '*', 3), '*', -1) *substring_index(substring_index(gs, '*', 4), '*', -1)
*substring_index(substring_index(gs, '*', 5), '*', -1)*substring_index(substring_index(gs, '*', 6), '*', -1)* substring_index(substring_index(gs, '*', 7), '*', -1)
*substring_index(substring_index(gs, '*', 8), '*', -1)*substring_index(substring_index(gs, '*', 9), '*', -1)*substring_index(gs, '*', -1)
end jg
from (select gs, (length(replace(gs, '*', '--')) - length(gs)) how
from testgs3)b
以下是运行后结果：
2.以下是我在ORACLE中实现字符串截取的方法：
ORACLE中使用SUBSTR和INSTR组合
1）substr函数格式 (俗称：字符截取函数)
格式1： substr(string string, int a, int b);
格式2：substr(string string, int a) ;
解释：
格式1：
1、string 需要截取的字符串
2、a 截取字符串的开始位置（注：当a等于0或1时，都是从第一位开始截取）
3、b 要截取的字符串的长度
格式2：
1、string 需要截取的字符串
2、a 可以理解为从第a个字符开始截取后面所有的字符串。
----创建测试数据
CREATE TABLE TESTGS AS
(SELECT '1.1*4*.1' GS FROM DUAL
UNION ALL
SELECT '2*1.1*1*2' X1 FROM DUAL
UNION ALL
SELECT '1' X1 FROM DUAL)
----拆分组成用量*的字符串lms220103
SELECT GS,
INSTR(GS, '*', 1, 1) N_FIRST,
INSTR(GS, '*', 1, 2) N_LAST,
INSTR(GS, '*', 1, 3) N_MIDDLE,
NVL(SUBSTR(GS, 0, INSTR(GS, '*', 1, 1) - 1), GS) N_FIRST_VALUE1,
NVL(SUBSTR(GS, -1, INSTR(GS, '*', -1, 1) + 1),1) N_LAST_VALUE,
NVL(SUBSTR(SUBSTR(GS, 0, INSTR(GS, '*', 1, 2) - 1), -1),1) N_MIDDLE_VALUE2,
NVL(SUBSTR(SUBSTR(GS, 0, INSTR(GS, '*', 1, 3) - 1), -1),1) N_MIDDLE_VALUE3,
NVL(SUBSTR(SUBSTR(GS, 0, INSTR(GS, '*', 1, 4) - 1), -1),1) N_MIDDLE_VALUE4,
NVL(SUBSTR(SUBSTR(GS, 0, INSTR(GS, '*', 1, 5) - 1), -1),1) N_MIDDLE_VALUE5,
NVL(SUBSTR(SUBSTR(GS, 0, INSTR(GS, '*', 1, 6) - 1), -1),1) N_MIDDLE_VALUE6,
NVL(SUBSTR(SUBSTR(GS, 0, INSTR(GS, '*', 1, 7) - 1), -1),1) N_MIDDLE_VALUE7,
NVL(SUBSTR(SUBSTR(GS, 0, INSTR(GS, '*', 1, 8) - 1), -1),1) N_MIDDLE_VALUE8,
NVL(SUBSTR(SUBSTR(GS, 0, INSTR(GS, '*', 1, 9) - 1), -1),1) N_MIDDLE_VALUE9,
NVL(SUBSTR(SUBSTR(GS, 0, INSTR(GS, '*', 1, 10) - 1), -1),1) N_MIDDLE_VALUE10
FROM (SELECT GS, (LENGTH(REPLACE(GS, '*', '00')) - LENGTH(GS)) HOW
FROM TESTGS)
实施：
----拆分组成用量*的字符串lms220103
WITH V_GS AS (SELECT GS,HOW,
INSTR(GS, '*', 1, 1) N_FIRST,
INSTR(GS, '*', 1, 2) N_LAST,
INSTR(GS, '*', 1, 3) N_MIDDLE,
NVL(SUBSTR(GS, 0, INSTR(GS, '*', 1, 1) - 1), GS) N_FIRST_VALUE1,
NVL(SUBSTR(GS, -1, INSTR(GS, '*', -1, 1) + 1),1) N_LAST_VALUE,
NVL(SUBSTR(SUBSTR(GS, 0, INSTR(GS, '*', 1, 2) - 1), -1),1) N_MIDDLE_VALUE2,
NVL(SUBSTR(SUBSTR(GS, 0, INSTR(GS, '*', 1, 3) - 1), -1),1) N_MIDDLE_VALUE3,
NVL(SUBSTR(SUBSTR(GS, 0, INSTR(GS, '*', 1, 4) - 1), -1),1) N_MIDDLE_VALUE4,
NVL(SUBSTR(SUBSTR(GS, 0, INSTR(GS, '*', 1, 5) - 1), -1),1) N_MIDDLE_VALUE5,
NVL(SUBSTR(SUBSTR(GS, 0, INSTR(GS, '*', 1, 6) - 1), -1),1) N_MIDDLE_VALUE6,
NVL(SUBSTR(SUBSTR(GS, 0, INSTR(GS, '*', 1, 7) - 1), -1),1) N_MIDDLE_VALUE7,
NVL(SUBSTR(SUBSTR(GS, 0, INSTR(GS, '*', 1, 8) - 1), -1),1) N_MIDDLE_VALUE8,
NVL(SUBSTR(SUBSTR(GS, 0, INSTR(GS, '*', 1, 9) - 1), -1),1) N_MIDDLE_VALUE9,
NVL(SUBSTR(SUBSTR(GS, 0, INSTR(GS, '*', 1, 10) - 1), -1),1) N_MIDDLE_VALUE10
FROM (SELECT GS, (LENGTH(REPLACE(GS, '*', '00')) - LENGTH(GS)) HOW
FROM TESTGS))
SELECT GS,
HOW,
(CASE
WHEN HOW = 0 THEN
GS
WHEN HOW = 1 THEN to_char(N_FIRST_VALUE1*N_LAST_VALUE)
WHEN HOW = 2 THEN to_char(N_FIRST_VALUE1*N_MIDDLE_VALUE2*N_LAST_VALUE)
WHEN HOW = 3 THEN to_char(N_FIRST_VALUE1*N_MIDDLE_VALUE2*N_MIDDLE_VALUE3*N_LAST_VALUE)
WHEN HOW = 4 THEN to_char(N_FIRST_VALUE1*N_MIDDLE_VALUE2*N_MIDDLE_VALUE3*N_MIDDLE_VALUE4*N_LAST_VALUE)
WHEN HOW = 5 THEN to_char(N_FIRST_VALUE1*N_MIDDLE_VALUE2*N_MIDDLE_VALUE3*N_MIDDLE_VALUE4*N_MIDDLE_VALUE5*N_LAST_VALUE)
WHEN HOW = 6 THEN to_char(N_FIRST_VALUE1*N_MIDDLE_VALUE2*N_MIDDLE_VALUE3*N_MIDDLE_VALUE4*N_MIDDLE_VALUE5*N_MIDDLE_VALUE6*N_LAST_VALUE)
WHEN HOW = 7 THEN to_char(N_FIRST_VALUE1*N_MIDDLE_VALUE2*N_MIDDLE_VALUE3*N_MIDDLE_VALUE4*N_MIDDLE_VALUE5*N_MIDDLE_VALUE6*N_MIDDLE_VALUE7*N_LAST_VALUE)
WHEN HOW = 8 THEN to_char(N_FIRST_VALUE1*N_MIDDLE_VALUE2*N_MIDDLE_VALUE3*N_MIDDLE_VALUE4*N_MIDDLE_VALUE5*N_MIDDLE_VALUE6*N_MIDDLE_VALUE7*N_MIDDLE_VALUE8*N_LAST_VALUE)
WHEN HOW = 9 THEN to_char(N_FIRST_VALUE1*N_MIDDLE_VALUE2*N_MIDDLE_VALUE3*N_MIDDLE_VALUE4*N_MIDDLE_VALUE5*N_MIDDLE_VALUE6*N_MIDDLE_VALUE7*N_MIDDLE_VALUE8*N_MIDDLE_VALUE9*N_LAST_VALUE)
WHEN HOW = 10 THEN to_char(N_FIRST_VALUE1*N_MIDDLE_VALUE2*N_MIDDLE_VALUE3*N_MIDDLE_VALUE4*N_MIDDLE_VALUE5*N_MIDDLE_VALUE6*N_MIDDLE_VALUE7*N_MIDDLE_VALUE8*N_MIDDLE_VALUE9*N_MIDDLE_VALUE10*N_LAST_VALUE)
else '请联系开发更新代码'
END) JG
FROM V_GS
运行的结果：
