Question

我需要拆分名称以插入数据库。下面给出了一些示例字符串。

KPI_LOGS,ALERT_LOGS

我需要拆分上面的字符串，如下所示。

Covey, Stephen J, Mr
Clinton, Hilary B,
Obama, Barack, Mr

我可以使用java代码实现上述功能。是否可以直接在sql select中拆分这些字符串。？

请注意，很少有名字可能没有标题或中间名。我需要上面给出的输出。

Answer 1

SQL Fiddle

Oracle 11g R2架构设置：

CREATE TABLE Names ( Name ) AS
          SELECT 'Covey, Stephen J, Mr' FROM DUAL
UNION ALL SELECT 'Clinton, Hilary B,' FROM DUAL
UNION ALL SELECT 'Obama, Barack, Mr' FROM DUAL

查询1 ：

SELECT REGEXP_SUBSTR( Name, '^(.*?),\s*(.*?)(\s+(\w))?,\s*(.*)$', 1, 1, NULL, 1 ) AS Last_Name,
       REGEXP_SUBSTR( Name, '^(.*?),\s*(.*?)(\s+(\w))?,\s*(.*)$', 1, 1, NULL, 2 ) AS First_Name,
       REGEXP_SUBSTR( Name, '^(.*?),\s*(.*?)(\s+(\w))?,\s*(.*)$', 1, 1, NULL, 4 ) AS Middle_Initial,
       REGEXP_SUBSTR( Name, '^(.*?),\s*(.*?)(\s+(\w))?,\s*(.*)$', 1, 1, NULL, 5 ) AS Title
FROM   Names

<强> Results ：

| LAST_NAME | FIRST_NAME | MIDDLE_INITIAL |  TITLE |
|-----------|------------|----------------|--------|
|     Covey |    Stephen |              J |     Mr |
|   Clinton |     Hilary |              B | (null) |
|     Obama |     Barack |         (null) |     Mr |

查询2 ：

SELECT REGEXP_REPLACE( Name, '^(.*?),\s*(.*?)(\s+(\w))?,\s*(.*)$', '\1' ) AS Last_Name,
       REGEXP_REPLACE( Name, '^(.*?),\s*(.*?)(\s+(\w))?,\s*(.*)$', '\2' ) AS First_Name,
       REGEXP_REPLACE( Name, '^(.*?),\s*(.*?)(\s+(\w))?,\s*(.*)$', '\4' ) AS Middle_Initial,
       REGEXP_REPLACE( Name, '^(.*?),\s*(.*?)(\s+(\w))?,\s*(.*)$', '\5' ) AS Title
FROM   Names

<强> Results ：

| LAST_NAME | FIRST_NAME | MIDDLE_INITIAL |  TITLE |
|-----------|------------|----------------|--------|
|     Covey |    Stephen |              J |     Mr |
|   Clinton |     Hilary |              B | (null) |
|     Obama |     Barack |         (null) |     Mr |

查询3 ：

WITH Split_Names AS (
  SELECT REGEXP_SUBSTR( Name, '^[^,]+' ) AS Last_Name,
         REGEXP_REPLACE( Name, '^.*?,\s*|\s*,.*?$' ) AS Given_Names,
         REGEXP_SUBSTR( Name, '[^\s,]+$' ) AS Title
  FROM   Names
)
SELECT Last_Name,
       REGEXP_REPLACE( Given_Names, '\s+\w$' ) AS First_Name,
       TRIM( REGEXP_SUBSTR( Given_Names, '\s+\w$' ) ) AS Middle_Initial,
       Title
FROM   Split_Names

<强> Results ：

| LAST_NAME | FIRST_NAME | MIDDLE_INITIAL |  TITLE |
|-----------|------------|----------------|--------|
|     Covey |    Stephen |              J |     Mr |
|   Clinton |     Hilary |              B | (null) |
|     Obama |     Barack |         (null) |     Mr |

Answer 2

使用regexp_substr（DB＆gt; = 10g）：

SELECT TRIM( REGEXP_SUBSTR(input_row, '[^,]+', 1, 1)) AS Last_NAME,
TRIM( REGEXP_SUBSTR( REGEXP_SUBSTR(input_row, '[^,]+', 1, 2), '[^ ]+', 1, 1)) AS First_Name,
TRIM( REGEXP_SUBSTR( REGEXP_SUBSTR(input_row, '[^,]+', 1, 2), '[^ ]+', 1, 2)) AS Middle_Initial,
TRIM( REGEXP_SUBSTR(input_row, '[^,]+', 1, 3)) AS Title
FROM source_table;

Answer 3

这是亚历山大的答案，用改进的正则表达式修改，处理NULL列表元素。哦，而不是重复那个正则表达式，通过创建一个如下所述的函数使其可重用：REGEX to select nth value from a list, allowing for nulls，然后调用它。这样，代码就会被所有人封装和重用，只有一个地方可以改变代码，如果你必须：

SQL> with tbl(input_row) as (
   select 'Covey, Stephen J, Mr' from dual
   union
   select 'Clinton,,Ms' from dual
   union
   select 'Obama, Barack, Mr' from dual
   )
   SELECT TRIM( REGEXP_SUBSTR(input_row, '([^,]*)(,|$)', 1, 1, NULL, 1)) AS Last_NAME,
   TRIM( REGEXP_SUBSTR( REGEXP_SUBSTR(input_row, '([^,]*)(,|$)', 1, 2, NULL, 1), '[^ ]+', 1, 1)) AS First_Name,
   TRIM( REGEXP_SUBSTR( REGEXP_SUBSTR(input_row, '([^,]*)(,|$)', 1, 2, NULL, 1), '[^ ]+', 1, 2)) AS Middle_Initial,
   TRIM( REGEXP_SUBSTR(input_row, '([^,]*)(,|$)', 1, 3, NULL, 1)) AS Title
   FROM tbl;

LAST_NAME            FIRST_NAME           MIDDLE_INITIAL       TITLE
-------------------- -------------------- -------------------- --------------------
Clinton                                                        Ms
Covey                Stephen              J                    Mr
Obama                Barack                                    Mr

SQL>

如何使用oracle按要求拆分给定的字符串

3 个答案: