Search for Accented Characters - sql-server

Question

I've been looking around and I've found a lot of information regarding creating searches that are accent insensitive. This isn't what I'm after.
Some accented data is causing a problem in my UI and I'm looking to do some impact analysis.
Is there an elegant way to search a field for any accented character, other than unioning many different selects with different characters in each?

Answer 1

The COLLATE clause changes the foo expression to allow you to GROUP BY separately
DECLARE #t TABLE (foo varchar(100));
INSERT #t VALUES ('bar'), ('bár'), ('xxx'), ('yyy'), ('foo'), ('foó'), ('foö');
SELECT
foo COLLATE Latin1_General_CI_AI,
MIN(foo COLLATE Latin1_General_CI_AS), MAX(foo COLLATE Latin1_General_CI_AS)
FROM
#t
GROUP BY
foo COLLATE Latin1_General_CI_AI
HAVING
MIN(foo COLLATE Latin1_General_CI_AS) <> MAX(foo COLLATE Latin1_General_CI_AS);
Or
SELECT
foo COLLATE Latin1_General_CI_AI,
COUNT(*)
FROM
#t
GROUP BY
foo COLLATE Latin1_General_CI_AI
HAVING
COUNT(*) > 1;
The first one gives you some actual values rather than just a COUNT. But not all if you have several accented words
For SQL Server 2012, you can use this
SELECT
*
FROM
(
SELECT
FIRST_VALUE(foo) OVER (PARTITION BY foo COLLATE Latin1_General_CI_AI ORDER BY foo COLLATE Latin1_General_CI_AS) AS safeFoo,
foo
FROM
#t
) X
WHERE
safeFoo <> foo
The 1st and 3rd rely on the sorting of non-accent characters before accented characters