how to extract string after any character - sql

Question

I have data shown below. I'd like to extract the last part after the last underscore.
The data before the underscore can be any length but written with the same syntax i means : letters_letters_letters.
So i wrote this code to extract the part after the last underscore then it's works perfectly but i noticed that there is 2 names written differently like (letters_letters-letters )
with - instead of _ at the end.
NAME= (SELECT SUBSTRING('''+#NAME+''', CHARINDEX(''_'','''+#NAME+''',CHARINDEX(''_'','''+#NAME+''')+1)+1, CHARINDEX(''_'','''+#NAME+''') + CHARINDEX(''_'','''+#NAME+''',CHARINDEX(''_'','''+#NAME+''')+1)) FROM TABLE)
My question is : is there away to check or extract string after any character (without specify if it's underscore or other) ?
Can anyone help please.
Column name is like :
BOB_LOU_K
the thow others columns are like :
BOB_LOU-K
Thanks

Answer 1

Create a function that finds the position of the last character that is not a letter and return the substring from that position to the end of the string. Call that function on your query, something like this SELECT Col1, [dbo].[fn_GetLastString](Col1) AS lastString FROM mytable your function fn_LastString would look somehing like this:
SET ANSI_NULLS ON
GO
SET QUOTED_IDENTIFIER ON
GO
CREATE FUNCTION fn_GetLastString(#cSearchedExpression VARCHAR(MAX))
RETURNS VARCHAR(MAX)
AS
BEGIN
DECLARE #pos INT = LEN(#cSearchedExpression)
DECLARE #currentChar VARCHAR(1)
DECLARE #lastString VARCHAR(MAX)
WHILE #pos>0
BEGIN
SET #currentChar = SUBSTRING(#cSearchedExpression, #pos, 1)
IF NOT (#currentChar LIKE '[a-Z ]' )
BEGIN
SET #lastString = SUBSTRING(#cSearchedExpression, #pos, LEN(#cSearchedExpression) - #pos + 1)
BREAK;
END
SET #pos = #pos - 1
END
RETURN #lastString
END
GO
p.s. notice the space in '[a-Z ]' that will include spaces.

Answer 2

select txt
,right(txt, patindex('%[^a-zA-Z]%', reverse(txt) + ' ') - 1) as last_token
from (values ('BOB_LOU_K'), ('BOB_LOU-K'), ('Yet-aonther-badly-written-question'), ('*followed*by*mediocre*answers'), ('at~best'), ('Quite depressing'), ('isn''t it?'), ('Yep')) t(txt)
txt
last_token
BOB_LOU_K
K
BOB_LOU-K
K
Yet-aonther-badly-written-question
question
*followed*by*mediocre*answers
answers
at~best
best
Quite depressing
depressing
isn't it?
Yep
Yep
Fiddle

Answer 3

A a minimal reproducible example is not provided. So, I am shooting from the hip.
If there are some additional characters to take care of, just add them to the #junkChars variable.
SQL
-- DDL and sample data population, start
DECLARE #tbl TABLE (ID INT IDENTITY PRIMARY KEY, tokens VARCHAR(100));
INSERT INTO #tbl (tokens) VALUES
('BOB_LOU_K'),
('BOB_LOU-K'),
('Yet-aonther-badly-written-question'),
('*followed*by*mediocre*answers'),
('at/the\best'),
('Quite depressing'),
('isn''t it?'),
('Yep');
-- DDL and sample data population, end
DECLARE #separator CHAR(1) = '_'
, #junkChars VARCHAR(10) = '-*\ ';;
SELECT t.*
, result = c.value('(/root/r[last()]/text())[1]', 'VARCHAR(30)')
FROM #tbl AS t
CROSS APPLY (SELECT TRY_CAST('<root><r><![CDATA[' +
REPLACE(TRANSLATE(tokens, #junkChars, REPLICATE(#separator, DATALENGTH(#junkChars))), #separator, ']]></r><r><![CDATA[') +
']]></r></root>' AS XML)) AS t1(c);
Output
ID
tokens
result
1
BOB_LOU_K
K
2
BOB_LOU-K
K
3
Yet-aonther-badly-written-question
question
4
followedbymediocreanswers
answers
5
at/the\best
best
6
Quite depressing
depressing
7
isn't it?
it?
8
Yep
Yep