I need to Filter a SQL SERVER return set based on the 4th character being UPPER case [duplicate]

All,

How can I check if a specified varchar character or entire string is upper case in T-Sql? Ideally I'd like to write a function to test if a character is upper case, then I can later apply that to a generic varchar. It should return false for non alphabetic characters. I am only interested in english language characters.

I am working with T-sql in SQL Management Studio, and I have tried pulling records beginning with a lower case letter from a table in this fashion:

select * from TABLE
where SUBSTRING(author,1,1) != LOWER(SUBSTRING(author,1,1))

Which returns 0 records, but I know there are records beginning with upper and lower case letters.

Thanks

EDIT: Since both podiluska and joachim-isaksoon have successfully answered my question (Both methods work for my purposes), would someone mind explaining which would be the most efficient method to use to query a table with a large number of records to filter out records with authors beginning with or without a capital letter?

Using collations

eg:

if ('a'='A' Collate Latin1_General_CI_AI) 
    print'same 1'
else
    print 'different 1'

if ('a'='A' Collate Latin1_General_CS_AI) 
    print'same 2'
else
    print 'different 2' 

The CS in the collation name indicates Case Sensitive (and CI, Case Insensitive). The AI/AS relates to accent sensitivity.

or in your example

SUBSTRING(author,1,1) <> LOWER(SUBSTRING(author,1,1)) COLLATE Latin1_General_CS_AI

To check if ch is upper case, and is a character that can be converted between upper and lower case (ie excluding non alphabetic characters);

WHERE UNICODE(ch) <> UNICODE(LOWER(ch))

An SQLfiddle to test with;

something like

declare @v varchar(10) = 'ABC', @ret int = 0
select @ret = 1 where upper(@v)=@v COLLATE SQL_Latin1_General_CP1_CS_AS
select @ret

Lower case letters have the same ASCII or UNICODE value as their upper-case version with the exception that lower case letters have a bit flag of 32 set.

This is very easy to detect directly, or to wrap the detection into User-defined functions like IsUpper() and IsLower().

Example (note that both ASCII() and UNICODE() work interchangeably here):

DECLARE @Letter char(1);

SET @Letter = 'A'
PRINT   
    @Letter + space(1) + 
    CASE
    WHEN (UNICODE(@Letter) & 32 > 0) THEN 'is lower case'
    ELSE 'is UPPER CASE'
    END

SET @Letter = 'z'
PRINT   
    @Letter + space(1) + 
    CASE
    WHEN (ASCII(@Letter) & 32 > 0) THEN 'is lower case'
    ELSE 'is UPPER CASE'
    END

-- Output:
--    A is UPPER CASE
--    z is lower case

This function will look for any upper case letters in an entire string using a recursive CTE:

CREATE FUNCTION [dbo].[ContainsUpper](@InputString [varchar](80))
RETURNS [bit] WITH EXECUTE AS CALLER
AS 
BEGIN
    DECLARE @Result         bit     = 0,
            @LowerString    varchar(80) = lower(@InputString);

    WITH Letters
    AS
    (
        SELECT  Position    = 1,
                IsUpper     = 
                    CASE unicode(left(@InputString,1))
                    WHEN unicode(left(@LowerString,1)) THEN 0
                    ELSE 1
                    END

        UNION ALL

        SELECT  Position    = Position + 1,
                IsUpper     = 
                    CASE unicode(substring(@InputString, Position + 1, 1))
                    WHEN unicode(substring(@LowerString, Position + 1, 1)) THEN 0
                    ELSE 1
                    END
        FROM    Letters
        WHERE   Position < len(@InputString)
    )
    SELECT  @Result = max(convert(int, IsUpper))
    FROM    Letters

    RETURN @Result
END