XOR encryption is null-terminating my strings!

**steez** · 05-31-2011

I'm dealing with two strings, one being the string that I want to encrypt using XOR encryption and one being my "key" that I XOR the original string against.

for (a simplified) example:

Code:

while ( szOriginal[i] != '\0' && szKey[i] != '\0' )
{
	szOriginal[i] = szOriginal[i] ^ szKey[i];
	i++;
}

HOWEVER, the problem is that if by chance szOriginal[i] == szKey[i], the XOR operation will result in 0! This will replace what was a valid character in the string with a null character.

Then, when I go to print the encrypted string (or do other things with it), the printing ends prematurely because the new null character terminates it.

How do most programmers deal with this?

**Bayint Naung** · 05-31-2011

Write encrypted using length. don't use functions that expect null terminated string.
Eg use something like

Code:

fwrite(encrypted_string,1,len,stdout);

**steez** · 05-31-2011

Originally Posted by Bayint Naung

Write encrypted using length. don't use functions that expect null terminated string.
Eg use something like

Code:

fwrite(encrypted_string,1,len,stdout);

Ah, ok. I don't know why I didn't think to do this... I had been using fread() and such and yet for some reason used fputs() instead of just fwrite().

Thank you very much!

**iMalc** · 05-31-2011

If you use printf to display it then yes that will stop at the first zero byte. But really, once you start XORing it's no longer text, it's binary data, so it doesn't make sense to use functions designed for printing text. It would be more appropriate to display each character in hexadecimal one at a time.
Of course "most programmers" wouldn't be caught dead using such a weak form of "encryption".

**laserlight** · 05-31-2011

Originally Posted by iMalc

Of course "most programmers" wouldn't be caught dead using such a weak form of "encryption".

It depends on how the key actually is generated, but that could well be part of a one time pad implementation since each byte of the plaintext is matched with each byte of the key.

... oh, but not really. steez, your loop is problematic in that if the key is shorter than the plaintext, you're going to end up with a bunch of bytes that are left as plaintext.

**iMalc** · 05-31-2011

Point taken.

**steez** · 05-31-2011

Originally Posted by iMalc

If you use printf to display it then yes that will stop at the first zero byte. But really, once you start XORing it's no longer text, it's binary data, so it doesn't make sense to use functions designed for printing text. It would be more appropriate to display each character in hexadecimal one at a time.
Of course "most programmers" wouldn't be caught dead using such a weak form of "encryption".

I'm not printing it to view as output, rather I'm writing to a file to encrypt the file (e.g. a .txt file or a .jpg). For this reason it makes much more sense that I should have just used fwrite() to begin with, but I made a mistake.

The key and the source data (to be encrypted) are read from separate files (whose filenames are passed as args to the program).

And thank you for your condescension, I'm only a week into writing my own programs and this is my first exposure to any sort of encryption. Have to start somewhere.

Originally Posted by laserlight

... oh, but not really. steez, your loop is problematic in that if the key is shorter than the plaintext, you're going to end up with a bunch of bytes that are left as plaintext.

I actually use different variables for each counter, and I check within my loop if it has reached the end of the key. Something like if k > keySize (which I find earlier with fseek). If it has, it goes back to the beginning of the key (k = 0).

Thread: XOR encryption is null-terminating my strings!

Thread Tools

Search Thread

Display

XOR encryption is null-terminating my strings!

Similar Threads

Check terminating null character

null terminating a string

Null terminating a character pointer?

terminating null

NULL terminating a multi dim char array