问题
I have the following stored procedure that converts IPs to 32-bit binary in TSQL. It works but is EXTREMELY slow (has only converted 8,400 in 40 minutes) - probably because it uses cursors. Does anyone have suggestions and/or a different approach to improve performance?
Here's an example: 1.1.79.129 is converted to 00000001.00000001.01001111.10000001
Thanks
CREATE PROCEDURE [dbo].[sp_UpdateTableConvertIPToBinary]
AS
SET NOCOUNT ON
declare @IP nvarchar(255)
declare IPList cursor for
/*
Get IP address from CIDR Block where Binary has not been assigned
*/
select left(IP,charindex('/',IP)-1) as Block from MyDB.dbo.MyTable
WHERE IP IS NOT NULL AND [Binary] IS NULL
ORDER BY Block
OPEN IPList
FETCH NEXT FROM IPList
INTO @IP
WHILE @@FETCH_STATUS = 0
BEGIN
begin
declare @Octet as varchar(15)
declare @Div as integer
declare @Output as varchar(100)
declare @n as integer
declare @OriginalIP varchar(15)
select @OriginalIP = @IP
select @Div = '128'
select @Output = ''
select @n = 0
WHILE @n < 4
begin
/*
Set @Octet = Class to the left of the first '.' in @IP
If no '.' in @IP, set @Octet = entire @IP (this will happen for the last Class)
*/
IF CHARINDEX('.',@IP) > 0
begin
select @Octet = left(@IP,charindex('.',@IP)-1)
end
else
begin
select @Octet = @IP
end
/*
If @Octet >= 128, append 1 to @Output and subtract 128 from @Octet
If not, append 0 to @Output
*/
if cast(@Octet as int) >= @Div
begin
select @Output = @Output + '1'
select @Octet = cast(@Octet as int) - @Div
end
else
begin
select @Output = @Output + '0'
end
/*
If @Octet >= 64, append 1 to @Output and subtract 64 from @Octet
If not, append 0 to @Output
*/
if cast(@Octet as int) >= (@Div/2)
begin
select @Output = @Output + '1'
select @Octet = cast(@Octet as int) - (@Div/2)
end
else
begin
select @Output = @Output + '0'
end
/*
If @Octet >= 32, append 1 to @Output and subtract 32 from @Octet
If not, append 0 to @Output
*/
if cast(@Octet as int) >= (@Div/4)
begin
select @Output = @Output + '1'
select @Octet = cast(@Octet as int) - (@Div/4)
end
else
begin
select @Output = @Output + '0'
end
/*
If @Octet >= 16, append 1 to @Output and subtract 16 from @Octet
If not, append 0 to @Output
*/
if cast(@Octet as int) >= (@Div/8)
begin
select @Output = @Output + '1'
select @Octet = cast(@Octet as int) - (@Div/8)
end
else
begin
select @Output = @Output + '0'
end
/*
If @Octet >= 8, append 1 to @Output and subtract 8 from @Octet
If not, append 0 to @Output
*/
if cast(@Octet as int) >= (@Div/16)
begin
select @Output = @Output + '1'
select @Octet = cast(@Octet as int) - (@Div/16)
end
else
begin
select @Output = @Output + '0'
end
/*
If @Octet >= 4, append 1 to @Output and subtract 4 from @Octet
If not, append 0 to @Output
*/
if cast(@Octet as int) >= (@Div/32)
begin
select @Output = @Output + '1'
select @Octet = cast(@Octet as int) - (@Div/32)
end
else
begin
select @Output = @Output + '0'
end
/*
If @Octet >= 2, append 1 to @Output and subtract 2 from @Octet
If not, append 0 to @Output
*/
if cast(@Octet as int) >= (@Div/64)
begin
select @Output = @Output + '1'
select @Octet = cast(@Octet as int) - (@Div/64)
end
else
begin
select @Output = @Output + '0'
end
/*
If @Octet >= 1, append 1 to @Output and subtract 1 from @Octet
If not, append 0 to @Output
*/
if cast(@Octet as int) >= (@Div/128)
begin
select @Output = @Output + '1'
select @Octet = cast(@Octet as int) - (@Div/128)
end
else
begin
select @Output = @Output + '0'
end
/*
if @n < 3, append . to @Output
*/
if @n < 3
begin
select @Output = @Output + '.'
end
/*
Remove the Octet just converted to Binary from @IP and increment the counter
*/
select @IP = right(@IP,len(@IP) - charindex('.',@IP))
select @n = @n+1
end
/*
Update table, set Binary = @Output
*/
UPDATE MyDB.dbo.MyTable Set Binary = @Output WHERE left(IP,charindex('/',IP)-1) = @OriginalIP
end
FETCH NEXT FROM IPList
INTO @IP
END
CLOSE IPList
DEALLOCATE IPList
回答1:
Well it looks like IPV4, so I am going to go on that assumption. You are also converting to some really long binary text representation, I recommend that you just use BINARY(4) instead, and my answer will assume that. You can convert this your 131-character text representation if you want, but I don't know why you would want to.
I posted an answer to a similar general question some years ago here: https://stackoverflow.com/a/1385701/109122
That basically showed this function:
CREATE FUNCTION dbo.fnBinaryIPv4(@ip AS VARCHAR(15)) RETURNS BINARY(4)
AS
BEGIN
DECLARE @bin AS BINARY(4)
SELECT @bin = CAST( CAST( PARSENAME( @ip, 4 ) AS INTEGER) AS BINARY(1))
+ CAST( CAST( PARSENAME( @ip, 3 ) AS INTEGER) AS BINARY(1))
+ CAST( CAST( PARSENAME( @ip, 2 ) AS INTEGER) AS BINARY(1))
+ CAST( CAST( PARSENAME( @ip, 1 ) AS INTEGER) AS BINARY(1))
RETURN @bin
END
go
Today however, I would recommend making it an inline Table-Value Function instead for performance.
This is how you could do that:
CREATE FUNCTION dbo.itvfBinaryIPv4(@ip AS VARCHAR(15)) RETURNS TABLE
AS RETURN (
SELECT CAST(
CAST( CAST( PARSENAME( @ip, 4 ) AS INTEGER) AS BINARY(1))
+ CAST( CAST( PARSENAME( @ip, 3 ) AS INTEGER) AS BINARY(1))
+ CAST( CAST( PARSENAME( @ip, 2 ) AS INTEGER) AS BINARY(1))
+ CAST( CAST( PARSENAME( @ip, 1 ) AS INTEGER) AS BINARY(1))
AS BINARY(4)) As bin
)
go
And this is how you could use that to perform your conversion and update without a cursor:
;
WITH cte As
(
SELECT *
FROM MyDB.dbo.MyTable
OUTER APPLY dbo.itvfBinaryIPv4(left(IP,charindex('/',IP)-1))
)
UPDATE cte
Set Binary = bin
WHERE IP IS NOT NULL
AND [Binary] IS NULL
This should be very fast.
回答2:
A function can be used in an UPDATE so you don't need to use a cursor. Conversion to a binary string, if that's really what you need, is left as an exercise.
How about this function to convert an IPV4 string to a BIGINT:
create function [dbo].[IntegerIPV4Address]( @IPV4Address VarChar(16) )
returns BigInt
with SchemaBinding -- Deterministic function.
begin
-- NB: ParseName is non-deterministic.
declare @Dot1 as Int = CharIndex( '.', @IPV4Address );
declare @Dot2 as Int = CharIndex( '.', @IPV4Address, @Dot1 + 1 );
declare @Dot3 as Int = CharIndex( '.', @IPV4Address, @Dot2 + 1 );
return Cast( Substring( @IPV4Address, 0, @Dot1 ) as BigInt ) * 0x1000000 +
Cast( Substring( @IPV4Address, @Dot1 + 1, @Dot2 - @Dot1 - 1 ) as BigInt ) * 0x10000 +
Cast( Substring( @IPV4Address, @Dot2 + 1, @Dot3 - @Dot2 - 1 ) as BigInt ) * 0x100 +
Cast( Substring( @IPV4Address, @Dot3 + 1, Len( @IPV4Address ) * 1 ) as BigInt );
end
It returns a BIGINT to avoid fiddling about with negative values.
Going from a BIGINT to a VARCHAR with zero-padding I use:
create function [dbo].[NormalizedIPV4Address]( @IntegerIPV4Address as BigInt )
returns VarChar(16)
with SchemaBinding -- Deterministic function.
begin
declare @BinaryAddress as VarBinary(4) = Cast( @IntegerIPV4Address as VarBinary(4) );
return Right( '00' + Cast( Cast( Substring( @BinaryAddress, 1, 1 ) as Int ) as VarChar(3) ), 3 ) +
'.' + Right( '00' + Cast( Cast( Substring( @BinaryAddress, 2, 1 ) as Int ) as VarChar(3) ), 3 ) +
'.' + Right( '00' + Cast( Cast( Substring( @BinaryAddress, 3, 1 ) as Int ) as VarChar(3) ), 3 ) +
'.' + Right( '00' + Cast( Cast( Substring( @BinaryAddress, 4, 1 ) as Int ) as VarChar(3) ), 3 )
end
回答3:
For those who are looking for one time IP4 to bigint conversion rather then performance issues with existing stored procedures. Below is the sample inline code that makes conversion
SELECT CONVERT(bigint, LEFT([IP], CHARINDEX('.', [IP]) - 1)) +
CONVERT(bigint, SUBSTRING([IP],
CHARINDEX('.', [IP]) + 1,
CHARINDEX('.', [IP], CHARINDEX('.', [IP]) + 1)
- CHARINDEX('.', [IP]) - 1)) * 256 +
CONVERT(bigint, SUBSTRING([IP],
CHARINDEX('.', [IP], CHARINDEX('.', [IP]) + 1) + 1,
CHARINDEX('.', [IP], CHARINDEX('.', [IP], CHARINDEX('.', [IP]) + 1) + 1)
- CHARINDEX('.', [IP], CHARINDEX('.', [IP]) + 1) - 1)) * 65536 +
CONVERT(bigint, RIGHT([IP],
LEN([IP]) - 1
- CHARINDEX('.', [IP], CHARINDEX('.', [IP], CHARINDEX('.', [IP]) + 1) + 1) + 1)) * 16777216
AS Value
FROM IPAddresses
-- additional check if IPv6 addresses are in the table
WHERE CHARINDEX(':', [IP]) = 0
来源：https://stackoverflow.com/questions/20429578/convert-ip-to-32-bit-binary-in-tsql