Need faster alternative to _Crypt_HashFile

[storme]

I've written a script to copy an important backup from the main hard drive to a USB hard drive.

Because the file is so important I decided to do a _Crypt_HashFile of the main file then compare it to the version copied to the usb drive.

Func FileCopyAndCheck($fileName, $source, $destination, $flag = 0)
    ;Success =
    ; Failure   @error = 1 ; Copy failed
    ;           @error = 2 ; Copy worked but files are different
    LogState($hLogFile, "Copy From:" & $source & "\" & $fileName & " To:" & $destination)
    If FileCopy($source & "\" & $fileName, $destination, $flag) = 0 Then
        LogState($hLogFile, "ERROR: Unalbe to copy file From:" & $source & " To:" & $destination)
        Return SetError(1, 0, "Copy Failed")
    ElseIf _Crypt_HashFile($source & "\" & $fileName, $CALG_SHA1) <> _Crypt_HashFile($destination & "\" & $fileName, $CALG_SHA1) Then
        LogState($hLogFile, "ERROR: Unalbe to copy file From:" & $source & " To:" & $destination)
        Return SetError(2, 0, "Failed compare")
    EndIf
    Return 1 ; OK
EndFunc   ;==>FileCopyAndCheck

Worked perfectly on the trial files...BUT... the backup file is 23GB (YEP LARGE FILE) and it takes forever to do the hash.

Any suggestions on how to ckeck if the file was copied perfectly without wating for "ages"

Thanks for any help!

John Morrison

aka

Storm-E

[smartee]

hi John,

Maybe will be faster, why not give it a shot. Do post the results if you do

Really hope this helps

-smartee

[MvGulik]

Yep. Do the scanning in low propriety in the background, and make it popup a reminder when its done. Ergo: No need to wait for it.

Unless you got some data/facts suggesting its doing it relative slower than general normal sized files ... you just got a reality reminder. Big files take longer! ... hehe.

--- --- ---

Oh yea, Yep, Or use faster code of course.

relative data in case you drop some stats.

- time to copy file on USB drive. (or read it from USB if that's faster)

- time to do the scan of the same file. (old way and new way)

Edited May 7, 2011 by singularity

[storme]

Yep. Do the scanning in low propriety in the background, and make it popup a reminder when its done. Ergo: No need to wait for it.

Good idea but this is a scheduled batch task that runs a 4am every morning.

The time probably isn't going to worry any one as it should be finished by the time they get in at 8am. BUT I’d prefer it to be finished long before they get in.

Unless you got some data/facts suggesting its doing it relative slower than general normal sized files ... you just got a reality reminder. Big files take longer! ... hehe.

I didn't realise their backup file had got so big.

Initially I thought my script had crashed as it doesn't give any feedback (no user watching it) except to a log file.

Oh yea, Yep, Or use faster code of course.

relative data in case you drop some stats.

- time to copy file on USB drive. (or read it from USB if that's faster)

- time to do the scan of the same file. (old way and new way)

Can't give you any of that that computer isn't here and I didn't have a drive big enough to take the file with me.

It's even going to be difficult here as I don’t have that large a USB drive to test with.

I was hoping someone had an alternative that would speed things up.

I'm even contemplating using the dos copy command with a /V to do the copy and verify. Just have to find out how to capture the return code.

Thanks for the help!

[ProgAndy]

Hashing requires reading of the whole file, so you can remove the hashing and just compare the two files directly. This could be faster.

[Danny35d]

I know must likely you want to use autoit and this is out of the subject, but a large file like that I will use rsync for windows.