XOR encryption problem

**~~cgod~~** · 07-29-2005

Im wanting to learn basic encryption to stop editing of .txt files for my game. But I have run into a problem my program crashes ! I just can't find the problem

Code:

#include <iostream>
#include <cstring>
#include <windows.h>
using namespace std;

void Encode_m(char message[256],char key [256])
{
	for (int i=0;i<strlen(message);i++)
	{
		message[i] = message[i]^key[i];
		cout << message[i];
	}
}
int main()
{
	Encode_m("This is a test","ABCDEFG");
	system("pause");
}

**The Brain** · 07-29-2005

your algorithm looks good.. i think it's just the way you are passing your strings into Encode_m. You could strcpy( ) your strings into your character arrays before passing them into the function.. or you could use string class objects. also, you might want to make sure all 256 elements of the key array have been initialized, or you'll be xor'ing with possible unknown garbage data.

optimization tips:

based on this try taking out function calls in your loops. this saves you an extra push/pop from the stack per each loop iteration. try one of these methods:

set a variable before you enter the loop

Code:

int max = strlen(message) ;

for (int i=0;  i<max;  ++i)

or even set a variable in the initalization part of your for loop

Code:

for (int i=0, max=strlen(message);  i<max;  ++i)

also, pre-incrementing your loop counter is supposed to save you a bit of processor time.

**arjunajay** · 07-29-2005

pre-incrementing your loop counter is supposed to save you a bit of processor time.

??? how?

**Rashakil Fol** · 07-29-2005

Originally Posted by The Brain

also, pre-incrementing your loop counter is supposed to save you a bit of processor time.

It doesn't.

**Zach L.** · 07-29-2005

also, pre-incrementing your loop counter is supposed to save you a bit of processor time.

For a user defined type with appropriate post/pre-increment operators, that might help, but any decent compiler will optimize the difference out between the two for you with primitive data types. Even if it didn't, if you really need something optimized that much, then you probably ought to be writing in assembly, as the difference (unoptimized) is likely to be barely perceptible.

??? how?

The theory is (in pseudo-code):

pre-inc (++x):
> x = x + 1
> return x

post-inc (x++):
> tmp = x
> x = x + 1
> return tmp

So, in theory, you save having to create a temporary variable. But, again, your compiler will probably optimize this for primitive types.

Cheers

**cyph1e** · 07-29-2005

It's simply because the key and message doesn't have the same length. Use this:

Code:

message[i] = message[i]^key[i%strlen(key)];

**JoshR** · 07-29-2005

Originally Posted by cgod

Im wanting to learn basic encryption to stop editing of .txt files for my game. But I have run into a problem my program crashes ! I just can't find the problem

Code:

#include <iostream>
#include <cstring>
#include <windows.h>
using namespace std;

void Encode_m(char message[256],char key [256])
{
	for (int i=0;i<strlen(message);i++)
	{
		message[i] = message[i]^key[i];
		cout << message[i];
	}
}
int main()
{
	Encode_m("This is a test","ABCDEFG");
	system("pause");
}

The reason why it doesnt work is because the message is larger than the key, therefor at the last letter for example "t" (position 13) , there is no matching letter in the key for that position.

**Salem** · 07-29-2005

Not to mention the fact that "strings like this" are likely in read-only memory, and thus cannot be modified.

**The Brain** · 07-29-2005

also

Code:

message[i] = message[i]^key[i];

could be written like this:

Code:

message[i] ^= key[i];

**dwks** · 07-29-2005

Code:

#include <iostream>
#include <cstring>

using namespace std;

void Encode_m(char* message, char *key) {
    for (int i = 0; i < strlen(message); i ++) {
        message[i] ^= key[i];
        cout << message[i];
    }
}

int main() {
    Encode_m("This is a test","ABCDEFG");
    system("pause");
    return 0;
}

I don't see why that won't work. (Not right now at least . . . .

)

**JoshR** · 07-29-2005

Because ...

When variable "i" is equal to 6 it XOR's the "s" from "this is a test" with G in "ABCDEFG".
When variable "i" is equal to 7 it XOR's the " " from "this is_a test" with a garbage value from "ABCDEFG" array because there IS NO CHARACTER IN THE 7TH SPOT!!!

**JoshR** · 07-29-2005

And another suggestiong, I dont know if someone already told you this (too lazy to re-read everything).

But use strings. They are not the enemy... and you are also using c++ so its common to use strings.

If you NEED a c string use a string then use "MyString.c_str()"

Strings make life so much easier IMHO.

**dwks** · 07-29-2005

Sorry, this code

Code:

key[i]

should be

Code:

key[i%strlen(key)]

**JoshR** · 07-29-2005

Actually it shouldnt... That would slow it down.

Call strlen once and save it to a variable. Also if he choses to do it that way he can althogh there are many other algorithms that he can choose from.

Plus thats already been said, and again this is C

Thread: XOR encryption problem

Thread Tools

Search Thread

Display

XOR encryption problem

Similar Threads

Need help understanding a problem

Memory problem with Borland C 3.1

Someone having same problem with Code Block?

A question related to strcmp

WS_POPUP, continuation of old problem