Question

I need to divide a C string into tokens. I thought that strtok will be my best try, but I’m getting very strange results…

Here is my test program. In this example I will get 3 tokens with “##” separator but when I try to work with the ones I supposedly had copied, only the third one is shown correctly.. the other two look corrupted or something… I don’t know… ?

#include <stdio.h>
#include <string.h>
#include <malloc.h>

#define TAM 3 //elements

char** aTokens(char* str, char* delimitador)
{
 char* pch;
 char** tokens;
 int i = 0; 
 tokens = (char**)malloc(sizeof(char*)*TAM);

 pch = strtok(str, delimitador);

 while(pch != NULL)
 {
  tokens[i] = (char*)malloc((sizeof(strlen(pch))+1) * sizeof(char));
  strcpy(tokens[i], pch);
  pch = strtok(NULL, delimitador);
  i++;
 }

 return tokens;
}

int main ()
{
 char str[] = "30117700,1,TITULAR,SIGQAA070,1977/11/30,M,1,14000,0.00,6600.00,10.00,2011/09/01,2012/09/01,0|17,0.00,NO,0,0,0.00, ,##30117700,1,TITULAR,SIGQAA070,1977/11/30,M,1,14000,0.00,6600.00,10.00,2011/09/01,2012/09/01,0|17,0.00,NO,0,0,0.00, ,##30117700,1,TITULAR,SIGQAA070,1977/11/30,M,1,14000,0.00,6600.00,10.00,2011/09/01,2012/09/01,0|17,0.00,NO,0,0,0.00, ,";

 char** tokens;
 int i;

 tokens = aTokens(str, "##");

 for(i = 0; i<TAM; i++)
  printf("%d -- %s\n", strlen(tokens[i]), tokens[i]);

 //Clean
 //for(i = 0; i<TAM; i++)
  //free(tokens[i]);

 //free(tokens);

 return 0;
}

output with GCC on Linux:

13 -- 30117700,1,T   <---- ?
13 -- 30117700,1,T   <----- ?
115 -- 30117700,1,TITULAR,SIGQAA070,1977/11/30,M,1,14000,0.00,6600.00,10.00,2011/09/01,2012/09/01,0|17,0.00,NO,0,0,0.00, ,

I have commented the “clean” section because it provides lots of runtime error too … 🙁

Help please!!

Answer 1

The following line isn’t correct. sizeof(strlen(..)) will be 4 (in a 32-bit app) regardless of the length of the string.

tokens[i] = (char*)malloc((sizeof(strlen(pch))+1) * sizeof(char));

It should probably be:

tokens[i] = (char*)malloc((strlen(pch)+1) * sizeof(char));