Segmentation Error

Hi,
I am writing a encryption problem that uses ceasar cyphering system. User is meant to enter in the command line the programme, string and a key that will determing how far the characters should be moved in the ASCII chart. My program is returning a segmetation error. Please help to identify the error

Code:

---

# include <stdio.h>


# include <cs50.h>


# include <stdlib.h>


# include <string.h>


# include <ctype.h>




int main(int argc, string argv[])
{
    //command line to have three entries or else  quit
    if (argc !=3)
        return 1;

    {
        string s = argv[1];

        //int length = strlen(s);





        //iterate through the string
        for (int i = 0; i < strlen(s); i++)
            {
                //char kar = s[i];

                int ch = 's[i]';


                int key = atoi(argv[2]);

                if ( isalpha("ch"))
                {
                    bool upper = true;
                    do
                    {

                        int value_u = (ch + key);
                        int result_u;
                        int Z = 'Z';
                        {
                            if (value_u > Z)
                            { 
                                result_u = value_u % 26;

                                printf("%c",result_u); 
                            }
                            else

                                printf("%c",result_u);
                        }             
                        }
                        while (upper == false);
                        {
                            int value_l = (ch + key);
                            int result_l;
                            int z = 'z';
                            {
                                if (value_l > z)
                                { 
                                    result_l = value_l % 26;

                                    printf("%c",result_l); 
                                }
                                else

                                    printf("%c",result_l);
                            }
                        }
                    }
                    else
                        printf("%c",ch);
                }
            printf("\n");
    }
    return 0;               
}


---

That code won't compile due to invalid constructs like

Code:

---

    int ch = 's[i]';

---

and

Code:

---

if ( isalpha("ch"))

---

There is also no type in C called string.

Fix the problems in your code that cause it not to compile. A program cannot produce runtime errors (such as segmentation violations, which I assume is what you mean by "segmentation error") if it cannot be executed. Successful compilation is a precursor to even having an executable.

Its also a very convoluted way of accomplishing the task - you could just use an array with the alphabet in it and then compare the contents of the message passed into the program match the letter to the alphabet array and then shift the index (remembering to wrap around) by whatever the key value is to obtain the ciphertext
This may seem convoluted too to some, but thats how i picture it

Quote:

---

Originally Posted by grumpy

There is also no type in C called string.

---

From here http://cdn.cs50.net/2011/fall/lectures/4/src4/cs50.h
I would guess someone had a "bright" idea to type def an char* as string.

Code:

---

typedef char *string;

---

I wonder when I learned things like that do NOT help.

Tim S.

Quote:

---

Originally Posted by stahta01

From here http://cdn.cs50.net/2011/fall/lectures/4/src4/cs50.h
I would guess someone had a "bright" idea to type def an char* as string.

Code:

---

typedef char *string;

---

I wonder when I learned things like that do NOT help.

---

Either way, not exactly standard. The most common origin of tricks like that is a mistaken belief that a pointer is an array (from which it is a small step to believing a char * is a string). As you say, such tricks cause a lot of trouble, for negligible benefits.

The other two problems in the code I pointed out remain extant.

Quote:

---

Originally Posted by grumpy

That code won't compile due to invalid constructs like

Code:

---

    int ch = 's[i]';

---

---

That one will compile just fine. It's a multicharacter constant. The value it assigns is implemenatation defined I believe, and on my system it gives the value of 0x735B695D.
It's probably not what was intended though.

Quote:

---

Originally Posted by iMalc

That one will compile just fine. It's a multicharacter constant. The value it assigns is implemenatation defined I believe, and on my system it gives the value of 0x735B695D.
It's probably not what was intended though.

---

Indeed -- g++ compiles it with a warning:
warning: multi-character character constant [-Wmultichar]

So I tried running the program, and this does indeed seem to be the problem.

From http://pubs.opengroup.org/onlinepubs.../isalpha.html:

Quote:

---

The c argument is an int, the value of which the application shall ensure is representable as an unsigned char or equal to the value of the macro EOF. If the argument has any other value, the behavior is undefined.

---

So although I'm a little surprised that it segfaults, it's not wrong to do so. Probably the implementation involves a lookup table of properties.

I ran it with valgrind before I clicked that the multibyte char was the problem -- there are other problems with this code too:

Code:

---

                    do
                    {

                        int value_u = (ch + key);
                        int result_u;
                        int Z = 'Z';
                        {
                            if (value_u > Z)
                            { 
                                result_u = value_u % 26;

                                printf("%c",result_u); 
                            }
                            else

                                printf("%c",result_u);
                        }             
                        }

---

In the printf in the else statement, result_u could be uninitialised.

...and seemingly from Harvard University no less. smh. Wow.