I want to read the company table and take out all possible suffixes from the name. Here's what I have so far:
declare @badStrings table (item varchar(50))
INSERT INTO @badStrings(item)
SELECT 'company' UNION ALL
SELECT 'co.' UNION ALL
SELECT 'incorporated' UNION ALL
SELECT 'inc.' UNION ALL
SELECT 'llc' UNION ALL
SELECT 'llp' UNION ALL
SELECT 'ltd'
select id, (companyname = Replace(name, item, '') FROM @badStrings)
from companies
where name != ''
Ed Northridge's answer will work, and I have upvoted it, but just in case multiple replacements are required I am adding another option using his sample data. If, for example one of the companies was called "The PC Company LTD" This would duplicate rows in the output with one being "The PC LTD" and the other "The PC Company". To resolve this there are 2 option depending on your desired outcome. The first is to only replace the "Bad Strings" when they occur at the end of the name.
SELECT c.ID, RTRIM(x.Name) [Name]
FROM @companies c
OUTER APPLY
( SELECT REPLACE(c.name, item, '') AS [Name]
FROM @badStrings
-- WHERE CLAUSE ADDED HERE
WHERE CHARINDEX(item, c.Name) = 1 + LEN(c.Name) - LEN(Item)
) x
WHERE c.name != ''
AND x.[Name] != c.Name
This would yield "The PC Company" with no duplicates.
The other option is replace All occurances of the Bad Strings recursively:
;WITH CTE AS
( SELECT c.ID, c.Name [OriginalName], RTRIM(x.Name) [Name], 1 [Level]
FROM @companies c
OUTER APPLY
( SELECT REPLACE(c.name, item, '') AS [Name]
FROM @badStrings
WHERE CHARINDEX(item, c.Name) = 1 + LEN(c.Name) - LEN(Item)
) x
WHERE c.name != ''
AND RTRIM(x.Name) != c.Name
UNION ALL
SELECT c.ID, OriginalName, RTRIM(x.Name) [Name], Level + 1 [Level]
FROM CTE c
OUTER APPLY
( SELECT REPLACE(c.name, item, '') AS [Name]
FROM @badStrings
WHERE CHARINDEX(item, c.Name) = 1 + LEN(c.Name) - LEN(Item)
) x
WHERE c.name != ''
AND x.[Name] != c.Name
)
SELECT DISTINCT ID, Name, OriginalName
FROM ( SELECT *, MAX(Level) OVER(PARTITION BY ID) [MaxLevel]
FROM CTE
) c
WHERE Level = maxLevel
This would yield "The PC" from "The PC Company".
UNION ALL - remove [Name]RTRIMs, I've edited the answer now so it should work.The error I got running the snippet was:
Msg 102, Level 15, State 1, Line 12
Incorrect syntax near '='.
The below code isn't an ideal solution - it will only return a list of companies where their name has been changed by the REPLACE function.
declare @companies table (id int, name nvarchar(50))
INSERT INTO @companies(id, name)
SELECT 1,'One Company' UNION ALL
SELECT 2, 'Two co.' UNION ALL
SELECT 3, 'Three incorporated' UNION ALL
SELECT 4, 'Four inc.' UNION ALL
SELECT 5, 'Five llc' UNION ALL
SELECT 6, 'Six llp' UNION ALL
SELECT 7, 'Seven ltd'
select * from @companies
declare @badStrings table (item varchar(50))
INSERT INTO @badStrings(item)
SELECT 'company' UNION ALL
SELECT 'co.' UNION ALL
SELECT 'incorporated' UNION ALL
SELECT 'inc.' UNION ALL
SELECT 'llc' UNION ALL
SELECT 'llp' UNION ALL
SELECT 'ltd'
select * from @badStrings
Here is the edited query:
select id, x.Name
from @companies c
OUTER APPLY (
SELECT Replace(c.name, item, '') AS [Name]
FROM @badStrings
) x
where c.name != ''
AND x.[Name] != c.Name
This returns:
id Name
----------- --------
1 One
2 Two
3 Three
4 Four
5 Five
6 Six
7 Seven
(7 row(s) affected)
Hopefully it's useful
Edit:
An alternative to apply the match to those company names which end with the @badStrings value
select id, x.Name
from @companies c
OUTER APPLY (
SELECT Replace(c.name, item, '') AS [Name]
FROM @badStrings
WHERE c.Name LIKE '%'+item
) x
where c.name != ''
L.L.P only LLP