[SOLVED] Why i get a weird symbol at the end of a string?

gerardorn · 11-29-2009, 04:42 PM

hey

im trying to generate random strings of a given length with the alphabet that the user inputs from the command line.

Code:

 char randchar(const char *alphabet){
         int val;
         float mod;
 
         mod =(float) strlen(alphabet);
         val = (int)(mod * rand() / (RAND_MAX + 1.0));
  return alphabet[val];
 }

this function returns the char in the random index generated. the man-pages of rand gives this way of generating a random number within a range, in my case the range is the length of the alphabet.

Code:

char *randword(const char *alphabet, const int wlength){
        char *word;
        int i = 0;

        word = (char *)malloc(sizeof(char) * wlength);
        if(!word){
                printf("!mem\n");
                exit(1);
        }
        while(i < wlength)
                word[i++] = randchar(alphabet);
 return word;
}

so this function calls the randchar function and builds a word with length wlength. is made like this cus i wanted to build a big file (~900 mb) with a long string, like a dna string, and i have 1g ram.

Code:

for(i=1; i < 151; i++){
         pattern = randword(argv[1], i);
         printf("pattern[%d]: %s\n", i, pattern);
         free(pattern);
}

this is from the main function.

anywaaays, when i try to generate random words with length 1, 2, 3, 4, ... 150, i get some strings with funy symbols at the end.

here is what i get from runing the program

Code:

$ ./algoritmos lkjhgfdsa
pattern[1]: g
pattern[2]: ff
pattern[3]: kjf
pattern[4]: fhhg
pattern[5]: gsfks
pattern[6]: dghggj
pattern[7]: gajfgka
pattern[8]: jkhsslal
pattern[9]: dfhklaadl
pattern[10]: dggkllhgad
pattern[11]: lgsaslhfsgf
pattern[12]: skkjjjjjljag
pattern[13]: skgsfasjlhklk
pattern[14]: kgdlahakgkgdgd
pattern[15]: dddjfsdggdhdsda
pattern[16]: aalhdlhlafgkljds
pattern[17]: lhgjlhlgsdsfgfghg
pattern[18]: fskdkkdddasaffalkj
pattern[19]: lgjghahagagahlsgdad
pattern[20]: gdhhfjajjlhflasfjsak
pattern[21]: hadjjdlsglgasahhkhfgh
pattern[22]: llhaalkdahlalhkshlhgfh
pattern[23]: jgdddlhjhhjshjsgasaadah
pattern[24]: asdlkjfgflkjsjdldlhghfha
pattern[25]: gjsgassadfasaghgffdhshhfg
pattern[26]: skahgaadsgddhfhlgjalfhfjka
pattern[27]: lfhdkkslfjlglslsfhhahajjaad
pattern[28]: gksgjhaagkdfdlfjkgjalfhlljhh�
... so on

and i cant use those strings =/
any ideas?

raju.mopidevi · 11-29-2009, 05:40 PM

In the "for loop" you had given i=1 to 151 ......... see that.

ASCII charcter set consists of 128 charcters defined as 0 to 127

in those characters, 0-31,127 are the special symbols. which are like smiley symbols,shapes ..etc.
32-126 are charcters which you can see on the keyboard.

if you ask a charcterset no beyond 127, then it agian starts from 0 ..
i.e. 128 =0
129 =1
130 =2
131 =3
.
.
.
.
.
151 =21

make a change in your for loop to stop those symbols.

gerardorn · 11-29-2009, 06:22 PM

hi thx for taking the time to answer

but i think thats not the problem =/

see, the loop is to make the word that long, it doesnt have anything to do with the ASCII code.

when the loop goes to 150 means: build a random word, with this alphabet, and make it 150 characters long

so it gets a random char from the alphabet and then appends it to a string until it has 150 chars, but this chars are all from the alphabet not randomized ASCII code.

raju.mopidevi · 11-29-2009, 07:15 PM

ok.
See the module "randchar"

Code:

 char randchar(const char *alphabet){
         int val;
         float mod;
 
         mod =(float) strlen(alphabet);
         val = (int)(mod * rand() / (RAND_MAX + 1.0));
  return alphabet[val];
 }

you said that range is length of alphabet. But you are adding 1 to it.
i.e.
if your alphabet is 9 char length. and if you want to generate a word of length 10. according to your code
the values of variable "val" will be ..{0,1,2,3,4,5,6,7,8,9,10}.
but the values of variable "val" must be ..{0,1,2,3,4,5,6,7,8,9}.

rupertwh · 11-29-2009, 07:26 PM

You are not zero-terminating your string.
And if you do, your malloc() will be one byte short.

=> malloc 1 more byte, zero terminate your string.

gerardorn · 11-30-2009, 03:57 PM

both of you were right

it works now

Code:

char randchar(const char *alphabet){
        int val;
        float mod;

        mod =(float) strlen(alphabet);
        val = (int)(mod * rand() / RAND_MAX);  
 return alphabet[val];
}

this is how it looks now and it works all good

Code:

char *randword(const char *alphabet, const int wlength){
        char *word;
        int i = 0;

        word = (char *)malloc(sizeof(char) * wlength+1);
        if(!word){
                printf("!mem\n");
                exit(1);
        }
        while(i < wlength){
                word[i++] = randchar(alphabet);
                word[i] = '\0';
        }
 return word;
}

and this is with the zero char at the end

thx for your help

rupertwh · 11-30-2009, 05:46 PM

[QUOTE=gerardorn;3774852]

Code:

...

        mod =(float) strlen(alphabet);
        val = (int)(mod * rand() / RAND_MAX); 
return alphabet[val];

Hmm. If rand() should return RAND_MAX, val will be equal to strlen(aphabet).
That means you would read past the end of the string. You would return the terminating zero.

gerardorn · 12-01-2009, 02:08 AM

ur right, i did this

Code:

         val = (int)(mod * RAND_MAX / RAND_MAX);

and its an endless loop =/, should i use a do...while(val == mod)? or what do you think?

rupertwh · 12-01-2009, 05:00 AM

Quote:

Originally Posted by gerardorn

and its an endless loop =/, should i use a do...while(val == mod)? or what do you think?

I'd say the random function was ok as it was (with the '+ 1.0').