Do you know if the algorithm for the BINARY_CHECKSUM function in documented
somewhere?
I would like to use it to avoid returning some string fields from the
server.
By returning only the checksum I could lookup the string in a hashtable and
I think this could make the code more efficient on slow connections.
Thanks in advanced and kind regards,
Orly Junior
--
David Portas
SQL Server MVP
--
Risky business. The checksum algorithm is fairly simple-minded. I beleive
it uses some xor mechanism. I don't have the references around right now,
but I recall that SQL Server MVP Steve Kass demonstrated how some quite
small changes could result in the same checksum.
Better in such case, to augment the table with a timestamp column. Such
a column is automatically updated every time SQL Server updates the
row. So you could store the timestamp client side, and pass that value,
if the table has the same value, there is no need for a refresh.
--
Erland Sommarskog, SQL Server MVP, esq...@sommarskog.se
Books Online for SQL Server SP3 at
http://www.microsoft.com/sql/techinfo/productdoc/2000/books.asp
Here is what I believe it does for a single varchar
column. It's not a particularly good hash function at
all.
create function binary_checksum_varchar (
@t varchar(1000)
) returns int as begin
declare @b bigint set @b = 0
declare @c tinyint
declare @s bit set @s = 0
declare @i int set @i = 1
while @i <= len(@t) begin
set @c = ascii(substring(@t,@i,1))
set @b = @b / 16 * 16 + @b % 16 ^ @c / 16
set @b = @b * 16 + @c % 16
if @c >= 128 begin
set @b = @b ^ 0xFF
set @s = 1 - @s
end
set @b = @b % 0x0100000000 ^ @b / 0x0100000000
set @i = @i + 1
end
if @s = 1 set @b = @b ^ 0xFFFFFFFF
if @b >= 0x80000000 set @b = @b | 0xFFFFFFFF00000000
return @b
end
go
You'll find more information if some of the threads
here:
http://groups.google.co.uk/groups?q=binary_checksum+kass+sqlserver
Steve Kass
Drew University