Oracle - 文本解析

Question

有人可以帮我解析 Oracle 数据库中字符串列的子字符串吗？

我有 2 列需要从数据库中获取，例如：

ID     STRING
425    22 23
416    1 22
682    1 22
688    11 133
703    22 12
707    1
710    1 22 12
715    1
716    1 12
796    333
740    1 22 12

我需要提取每个由“空格”分隔的子字符串，将其与正确的ID连接并放入一个新表中，例如：

enter code here
ID:    SUBSTRING
425    22
425    23
416    1
416    22
425    22
425    23

我需要假设我不知道STRING 列中子字符串的数量。

我尝试过这样的事情，但它复制了输出记录：

select ID,
regexp_substr(string,'[^ ]+',1,level),
from (select ID, STRING from TABLE   
where regexp_substr(STRING,'[^ ]+',1,level) is not null
CONNECT BY ID = ID 
and regexp_substr(STRING,'[^ ]+',1,level) is not null

Answer 1

其中一种方法：

with test_table as (
  select 425 id, '22 23' str from dual union all
  select 416   , '1 22'      from dual union all
  select 682   , '1 22'      from dual union all
  select 688   , '11 133'    from dual union all
  select 703   , '22 12'     from dual union all
  select 707   , '1'         from dual union all
  select 710   , '1 22 12'   from dual union all
  select 715   , '1'         from dual union all
  select 716   , '1 12'      from dual union all
  select 796   , '333'       from dual union all
  select 740   , '1 22 12'   from dual ),
xml_table as (
  select id, 
         xmltype('<main><str>' || regexp_replace(str, ' +', '</str><str>') || '</str></main>') xml
    from test_table)
select id,
       extractValue(value(t),'str') result
from xml_table s,
     table(XMLSequence(s.xml.extract('main/str'))) t;
;

Answer 2

您可以使用DISTINCT 或使用SYS_GUID() 方法。

SELECT ID
    ,REGEXP_SUBSTR(string, '[^ ]+', 1, LEVEL) AS substring
FROM TABLE1 CONNECT BY LEVEL <= REGEXP_COUNT(STRING, '[^ ]+')
    AND PRIOR ID = ID
    AND PRIOR SYS_GUID() IS NOT NULL

阅读这篇文章了解它的工作原理：https://community.oracle.com/thread/2526535

Fiddle demo