Teradata String Manipulation (second Space)
I'm having great difficulty solving this seemingly easy task: Purpose: Create a query that eliminates the middle Initial Example Name Smith, John A Jane, Mary S I would like an ou
Solution 1:
select a,
substr(a,1,index(a,' '))|| substr(trim(substr(a,index(a,' '))),1,index(trim(substr(a,index(a,' '))),' ')),
substr(trim(substr(a,index(a,' '))),index(trim(substr(a,index(a,' '))),' ')) last_name
from a
Solution 2:
The challenge is making sure your names are consistently formatted. (Last_Name, Given_Name Middle_Initial
) If they are then you may be able to solve this with recursive SQL. The following SQL would take Given_Name Last_Name
and return Last_Name
. You may be able to tweak it to accomplish your specific task. (My sample data was not consistently formatted so I was stuck trying to find the second (or third) occurrence of a white space character.)
WITHRECURSIVE cte (FullName, DelimPosition, RecursionLevel, Element, Remainder) AS
(
SELECT FullName
, 0AS DelimPosition_
, 0
, CAST(''ASVARCHAR(128))
, FullName
FROM MyDatabase.Persons
UNIONALLSELECT FullName
, CASEWHENPOSITION(' 'IN Remainder) >0THENPOSITION(' 'IN Remainder)
ELSECHARACTER_LENGTH(Remainder)
END DelimPosition_
, RecursionLevel +1
, SUBSTRING(Remainder FROM0FOR DelimPosition_ +1)
, SUBSTRING(Remainder FROM DelimPosition_ +1)
FROM cte
WHERE DelimPosition_ >1AND RecursionLevel <3-- Set max depth
)
SELECT FullName
, CASEWHENPOSITION('&'IN Element) =0THEN Element
ELSENULLENDAS LastName
FROM cte c
WHERE RecursionLevel >2ORDERBY FullName;
Another option would be to implement a UDF that returns the rightmost n characters of a string. (e.g.RIGHT(FullName, n)
)
If the formatting is not consistent then we have to look at other less graceful options.
Hope this helps.
Post a Comment for "Teradata String Manipulation (second Space)"