Teradata SELECT 失败：[2663] SUBSTR：字符串下标越界

Question

我有下面一段代码失败了。

,COALESCE(SAP.LAST_NAME, SUBSTR(PERSON_DIM.PERSON_NAME, 1,INDEX(PERSON_DIM.PERSON_NAME,',' )-1)) AS Lastname

Error: SELECT Failed: [2663] SUBSTR: string subscript out of bounds in PERSON_NAME

我调试后发现问题出在 SUBSTR(PERSON_DIM.PERSON_NAME, 1,INDEX(PERSON_DIM.PERSON_NAME,',' )-1 部分。

 SEL PERSON_NAME,INDEX(PERSON_NAME,',' )-1 FROM NDW_SHARED_PII_VIEWS.PERSON_DIM WHERE PERSON_NAME NOT LIKE '%,%';--Giving us results
 output:
 Star Installations Inc -1
 Unassigned -1
 Cable Services Company Inc.    -1

 SEL SUBSTR(PERSON_NAME, 1,INDEX(PERSON_NAME,',' )-1) FROM NDW_SHARED_PII_VIEWS.PERSON_DIM WHERE PERSON_NAME NOT LIKE '%,%';--Failing

 Above query eventually becomes like below which is causing confusion i believe. 

 SEL SUBSTR(PERSON_NAME, 1,-1) FROM NDW_SHARED_PII_VIEWS.PERSON_DIM WHERE PERSON_NAME NOT LIKE '%,%';

包含“,”的

PERSON_NAME 运行没问题。你能指导我如何构建查询以避免这个错误吗？

Answer 1

据推测，有些名字没有逗号。我认为最简单的方法是只加一个：

COALESCE(SAP.LAST_NAME,
         SUBSTR(PERSON_DIM.PERSON_NAME, 1, INDEX(PERSON_DIM.PERSON_NAME || ',', ',' ) - 1
               )
        ) AS Lastname

Answer 2

感谢您的回答。

我已经在下面尝试过，效果也很好。

COALESCE(SAP.LAST_NAME,
    CASE WHEN PERSON_DIM.PERSON_NAME NOT LIKE '%,%' 
            THEN PERSON_DIM.PERSON_NAME
    ELSE 
            SUBSTR(PERSON_DIM.PERSON_NAME,1,INDEX(PERSON_DIM.PERSON_NAME,',' )-1) END) AS LASTNAME

Answer 3

最短路线利用STRTOK:

COALESCE(SAP.LAST_NAME, StrTok(PERSON_DIM.PERSON_NAME,',',1))

无需嵌套函数或添加逗号或 CASE ...

Teradata SELECT 失败：[2663] SUBSTR：字符串下标越界

Teradata SELECT Failed: [2663] SUBSTR: string subscript out of bounds

sql

teradata

teradata-sql-assistant