ASM Community

This is a piece of code I found in the web, and it works well.
But the DWORD var X define as "X dword ****b", when I Input a bin number into a Edittext and use GetDlgItemInt or GetDlgItemText to get the number it does not end with b, How can I do that?(just as I define the DWORD var X without b ended). THX.

.386
.model flat,stdcall
option casemap:none

includelib msvcrt.lib
printf PROTO C :dword,:vararg

.data
X dword 00011000111011110001001101100101b

dStr byte 10 dup(?),0      
szFmt  byte 'dec format = %sd',0ah     

.code
start:

 mov ecx,10         
 mov eax,X         
 xor edx,edx          
 mov ebx,10

f70:
    div ebx          
 add dl,'0'           
 mov dStr,dl  
 xor edx,edx          
    loop f70             

 invoke printf,offset szFmt,\
  offset dStr

 ret
end start

Hi Donkey?

How can I allocate the buffer ended with b and the terminating NULL?
I try to define the "X    byte  4  dup(?),'b',0" but it works wrong.

Hi Donkey,

I just want user to type the 0/1 numbers(max 32 bit - a 8 bin hex) without the terminating 'b' in the edittext and I've no idea how to do it.
THX.

Eric

Zcoder,

I get the method but how can I put the string ended with 'b' and null into eax, there is something wrong with my code:

    .data
XBuf  db  32 DUP (?),0
    .const
szb  db  'b',0
    .code
invoke    GetDlgItemText,hWinMain,IDC_BIN,addr XBuf,sizeof XBuf
invoke    lstrcat,addr XBuf,addr szb
;_DebugString	"Display mem string",addr XBuf      ; e.g.: Input:1010101, here is 1010101b
xor eax,eax
mov    eax,dword ptr XBuf     ;  eax = 30313031H, it should be 00000055H

but the code in the first posted is OK:

    .data
X dword 1010101b
    .code
start:
 mov ecx,10         
 mov eax,X    ; hera eax = 00000055H 
.... 

Yes, bin2dec, just the topic name..

Eric, GetDlgItemText retrieves the user input as an ASCII string. You can't simply read this as a DWORD value, you need to convert it. This involves iterating over each character in the string, determining if it's a correct value ('0' or '1'), et cetera.

The algorithm is simple, the bit shift instructions (SHL or SHR, depending on how you do stuff) and bitwise OR are your friends :)

EDIT: try this on for size
EDIT-2: when entering, ECX must point to the string. String must consists of 0 and 1 chars only, and should not end with a 'b'.

	xor		ebx, ebx			; output number
	xor		eax, eax			; temporary value thing

@@loop:
	mov		al, 
	test	al, al				; terminating 0-byte?
	jz		@@done				; yup, we're done

	shl		ebx, 1				; "add room for another digit"

	sub		al, '0'				; convert ASCII to 0-or-1
	cmp		al, 1				; is decimal value > 1?
	jg		@@bad				; if yes, bad user input, bail out

	or		ebx, eax			; "add" digit to result
	inc		ecx					; point at next input char
	jmp		@@loop
@@done:

	; at this point, ebx is the machine representation of the user
	; binary string
	invoke	ExitProcess, 0

@@bad:
	; show error
	invoke	ExitProcess, 1