2

I have two comma-separated strings like this:

'number1,number2,number3'
'word1, word2, word3'

And I want to split them into rows keeping the order of values. The expected result is:

COL1               COL2
-------------------------
number1          word1
number2          word2
number3          word3

There is a special condition: we can't use square brackets signs - [...].

PS. I've found how to split a single string using XMLTable, but I don't know how to join it with the second string's values keeping the same order.

4

1 回答 1

2

只要行是固定的并且字符串具有相同顺序的substr 值，我认为您无需担心顺序。

更新OP 想坚持 XMLTABLE 解决方案

使用XMLTABLE

SQL> WITH input_data AS
  2    (SELECT 'number1,number2,number3' str1, 'word1, word2, word3' str2 FROM dual
  3    ) ,
  4    temp_data1 AS
  5    (SELECT rownum rn,
  6      trim(COLUMN_VALUE) str1
  7    FROM input_data,
  8      xmltable(('"'
  9      || REPLACE(str1, ',', '","')
 10      || '"'))
 11    ) ,
 12    temp_data2 AS
 13    (SELECT rownum rn,
 14      trim(COLUMN_VALUE) str2
 15    FROM input_data,
 16      xmltable(('"'
 17      || REPLACE(str2, ',', '","')
 18      || '"'))
 19    )
 20  SELECT t1.str1,
 21    t2.str2
 22  FROM temp_data1 t1,
 23    temp_data2 t2
 24  WHERE t1.rn = t2.rn
 25  /

STR1    STR2
------- -------
number1 word1
number2 word2
number3 word3

SQL>

使用REGEXP_SUBSTR

SQL> WITH DATA AS
  2    ( SELECT 'number1,number2,number3' str1, 'word1, word2, word3' str2 FROM dual
  3    )
  4  SELECT trim(regexp_substr(str1, '[^,]+', 1, LEVEL)) str1,
  5    trim(regexp_substr(str2, '[^,]+', 1, LEVEL)) str2
  6  FROM DATA
  7    CONNECT BY LEVEL <= regexp_count(str1, ',')+1
  8  /

STR1                    STR2
----------------------- -------------------
number1                 word1
number2                 word2
number3                 word3

SQL>

有关更多示例，请参阅https://lalitkumarb.wordpress.com/2014/12/02/split-comma-delimited-string-into-rows-using-oracle-sql/

于 2015-04-06T05:22:27.227 回答