Check "true file type"

[ken82m]

Does anyone know how I can check the real file type regardless of extension.

For example I could rename an XLS file to TXT. Is their a way to verify that's it's not really a plain text file?

I thought there was an IsAlpha() function but I don't see it in the help file.

In this case all I care about is making sure it's plain text, but it would be useful to be able to check content types in general.

Thanks,

Kenny

Edited January 7, 2009 by ken82m

[ken82m]

ahh I found it in another post it was StringIsAlpha()

I ran a search in the help file for IsAlpha but no luck.

Thanks anyway,

Kenny

Edited January 7, 2009 by ken82m

[MrCreatoR]

I also would like to know a way to extract file-type from the file itself...

[weaponx]

Here is an example in C#:

http://www.codeproject.com/KB/files/FindMimeFromData.aspx

[weaponx]

I found this program call TrID as well, not sure if it works.

[MrCreatoR]

All those programs is good, but we need something on AutoIt ... That C# example, could be translated to AutoIt? via WinAPI perhaps? or what it uses...

[smashly]

Get the signature of the different file types and store them to compare.

Open the file you want to check in binary mode and read some of the starting bytes to get the signature and compare against the stored signatures..

That's what I was doing for exe, pdf, ico, mp3, rar.. etc.

But I notice that not all file types have a signature..

[weaponx]

TrID offers a dll library. Look at the first language mentioned in the example list (hint - its AutoIt):

http://mark0.net/code-tridlib-e.html

[MrCreatoR]

TrID offers a dll library. Look at the first language mentioned in the example list (hint - its AutoIt):

http://mark0.net/code-tridlib-e.html

Nice, but it's too mch (1+ mb) to use it for such litle task

But thanks anyway, it's a start...

[weaponx]

Nice, but it's too mch (1+ mb) to use it for such litle task

But thanks anyway, it's a start...

Are you crazy? Its not a trivial task at all. The definition file can identify 3,640 different file types, I think 1.4 Mb is small for how many signatures it contains.

[KaFu]

I think 1.4 Mb is small for how many signatures it contains.

Eh? Did i miss something? I just see a 19 KB dll file? Which download are you talking about? Anyway, thinks this will be a nice addition to SMF ....

Regards

[weaponx]

Eh? Did i miss something? I just see a 19 KB dll file? Which download are you talking about? Anyway, thinks this will be a nice addition to SMF ....

Regards

You need the definition file which is a database of file type signatures.

Look at the script example:

expandcollapse popup

; TrIDlib AutoIt Demo
; http://mark0.net/code-tridlib-e.html

#include "TrIDLib-const.au3"

$MyFile = "triddefs.trd" ; a random file to analyze
$Buf = "" ; allocate a string for the results

$Dll = DllOpen("TrIDLib.DLL") ; Open TrIDLib DLL

$Ret = DllCall($Dll, "int", "TrID_LoadDefsPack", "str", "") ; load the definitions
$Ret = DllCall($Dll, "int", "TrID_SubmitFileA", "str", $MyFile) ; submit the file
$Ret = DllCall($Dll, "int", "TrID_Analyze") ; perform the analysis

if @error = 1 then
    MsgBox(48, "TrID_Analyze", "Error(s) occurred!")
Else    
    $Ret = DllCall($Dll, "int", "TrID_GetInfo", "int", $TRID_GET_RES_NUM, "int", 0, "str", $Buf)
    $ResNum = $Ret[0] ; get the results number
    $Out = ""
    if $ResNum = 0 Then
        $Out = "Unknown filetype!"
    Else
        For $ResId = 1 to $ResNUm ; cycle trough results
            $Ret = DllCall($Dll, "int", "TrID_GetInfo", "int", $TRID_GET_RES_FILETYPE, "int", $ResId, "str", $Buf)
            $Out = $Out & $ResId & ":" & Chr(9) &  $Ret[3] ; get the filetype description
            $Ret = DllCall($Dll, "int", "TrID_GetInfo", "int", $TRID_GET_RES_FILEEXT, "int", $ResId, "str", $Buf)
            $Out = $Out & " (" & $Ret[3] & ")" ; get the filetype extension
            $Ret = DllCall($Dll, "int", "TrID_GetInfo", "int", $TRID_GET_RES_POINTS, "int", $ResId, "str", $Buf)
            $Out = $Out & " - " & $Ret[0] ; get the matching points
            $Out = $Out & Chr(13)
        Next 
    EndIf
    MsgBox(0, "TrIDLib demo", $Out)
EndIf

DllClose($Dll) ; cleanup

Edited January 8, 2009 by weaponx

[KaFu]

You need the definition file which is a database of file type signatures.

Ah, i see, got the demo working... nice find m8 , and compressed 440 kb is worth the bandwidth , will integrate it in SMF

Best Regards

[weaponx]

Strangely it always returns "Unknown filetype!" for a simple text or zip file in AutoIt 3.3.0.0

[KaFu]

Strangely it always returns "Unknown filetype!" for a simple text or zip file in AutoIt 3.3.0.0

Works for me, identifies zip... had an error in the first try though... zip file was 0 Bytes large, that returned an unkown ... but you're right bout the txt file. I guess the reason is, that a plain txt file has no unique signature...

Edited January 8, 2009 by KaFu

[weaponx]

Works for me, identifies zip... had an error in the first try though... zip file was 0 Bytes large, that returned an unkown ... but you're right bout the txt file. I guess the reason is, that a plain txt file has no unique signature...

I think you're right. I tried an empty zip file first, it works when not empty. Also if you save the text file as UTF-8 it is recognized.