C分割字符串时出现分段错误

Question

First of all sorry if I this is a basic (or stupid) question, I am coming from Python and I am quite new in C (still studying it).

I have a short script to split a string into sub-strings, for example: "this is my -string" into "this","is","my","-string".

After that I want to select the substring that starts with the char: '-', and save in a variable calls "subline":

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

#define MAX_CHAR 9999

int main ()
{

  char line[] ="this is my -string";
  char *p;
  char subline[MAX_CHAR];


  printf ("Split string in tokens:\n");

  p = strtok (line," ");

  while (p != NULL)
  {
    printf ("%s\n", p);
    p = strtok (NULL, " ,");

    if ((strncmp(p, "-", 1) == 0)){ 
      memcpy(subline, ++p, strlen(p)+1);
      printf ("subline: %s\n", subline);

    }


  }
  printf ("\nData:\n");
  printf ("subline is: %s\n", subline);
  return 0;
}

Everything runs fine inside the while loop, where I can even print the variable "subline", but outside the while loop I get a segmentation fault, here is the output:

root@debian:/home/user/Desktop/splits# ./split
Split string in tokens:
this
is
my
subline: string
string
Segmentation fault

I tried to figure out and solved it using malloc(sizeof(*subline)); but always the same segmentation fault outside the while loop.

Anyone has any idea?

Thank you.

Answer 1

When p becomes null, you still pass it to strcncmp() . Don't do that -- add another check instead.

NULL is returned by strtok when there are no more matches, which the code detects in the while-loop, but before the loop logic will catch it, the strncmp() method is called with the NULL pointer.

Answer 2

this is the main error:

the function strtok() is being called a second time, before the first string pointer has been processed

this line:

memcpy(subline, ++p, strlen(p)+1);

(remembering that 'p' is a pointer to the first character of the string, and pre-increment of that pointer inside the memcpy() function is a very bad idea due to certain 'side effects'.

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

// wrap numerics in parens to avoid certain 'text replacement' errors
#define MAX_CHAR (9999)

int main ( void )
{

  char line[] ="this is my -string";
  char *p;
  char subline[MAX_CHAR] = {'\0'};  // assure the string will be properly terminated


  printf ("Split string in tokens:\n");

  p = strtok (line," "); // p now points to first token (or contains NULL

  while (p)   // will continue to loop while p not equal to NULL, where NULL is the same as 'false'
  {
      printf ("%s\n", p);

      // don't need all this power function call
      // if ((strncmp(p, "-", 1) == 0))
      if( '-' == p[0] )
      {
          p++; // step past the '-'
          memcpy(subline, p, strlen(p)); // remember, subline is init. to all '\0'
          printf ("subline: %s\n", subline);
      }

      p = strtok (NULL, " ,"); // p now points to next token (or contains NULL)
  } // end while

  printf ("\nData:\n");
  printf ("subline is: %s\n", subline);
  return 0;
}