Split a string into token in C

Question

struct string {
    size_t length;
    size_t allocated;
    char* data;
};
    string* strCreate(char* str) {...}
string* strSubstring(string* str, size_t pos, size_t len) {...}
char* strGet(string* str) {
    return str->data;
}
size_t findFrist(string* str, char what, size_t pos) {
    for(size_t i = pos; i < str->length; ++i) {
        if(str->data[i] == what)
            return i;
    }
    return -1;
}
string** strTokenizer(string* str) {
    string** res;
    res = malloc(sizeof(char*)*90); //*90 (token num)
    for(int i=0; i<90; i++) //i<90 (token num)
        res[i] = malloc(sizeof(char)*100); //*100 (a token lenght)
    size_t first = 0;
    size_t i = 0;
    while(first < str->length) {
        int second = findFrist(str,' ',first);
        if(second == - 1)
            second = str->length;
        string* token = strSubstring(str,first, second - first);
        if(*strGet(token) != ' ')
            res[i] = token;
        first = second + 1;
        ++i;
    }
    return res;
}
int main() {
    string* fe = strCreate("A string \ tof");
    string** r = strTokenizer(fe);
    for(int i = 0; i < 4; ++i) {
        printf("%s",strGet(r[i]));
    }
    return 0;
}

I want to create a string tokenizer. When I want to print the string in the main function it not print the \ . The other thing is that how to allocate string** res in proper way. When I allocate for the sizeof(char*) the only way that I see is loop through the string, but I was just wondering is this possible to allocate without going through the string 2 times in the tokenizer function.

I do not want to use strtok

typedef struct string string in the.h file

Answer 1

When compiling the code using GCC 10.2.0 with options:

gcc -O3 -g -std=c11 -Wall -Wextra -Werror …

I get the message:

bs83.c:85:44: error: unknown escape sequence: '\040' [-Werror]
   85 |     string *fe = strCreate("A string \ tof");

As I noted in the comments, the string "A string \ tof" is malformed — your compiler should be warning you about it. If you want a backslash in a string, you write (for example), "A string \\ tof" . Backslash-space has no defined meaning; it is probably being interpreted as a single space. Maybe you should print the argument to strCreate() to validate this.

You also need typedef struct string string; near the top of the code for it to compile with a C compiler (and the headers).

Here is a fixed version of your code — because of the compilation options I use, the functions need to be either static or declared before being defined; I make them static since the only reason to make them non- static (IMO) is if they're accessed from another source file, and then they'd be declared in a header. That probably applies to your code — there is mention of a header.

Here's fixed code that works:

/* SO 6537-9584 */
#include <stdio.h>
#include <stdlib.h>
#include <string.h>

typedef struct string string;

struct string
{
    size_t length;
    size_t allocated;
    char *data;
};

static
string *strCreate(char *str)
{
    string *s = malloc(sizeof(string));
    s->length = strlen(str);
    s->allocated = s->length * 2;
    s->data = (char *)malloc(s->allocated + 1);
    memcpy(s->data, str, s->length);
    s->data[s->length] = '\0';
    return s;
}

static
string *strSubstring(string *str, size_t pos, size_t len)
{
    string *s = malloc(sizeof(string));
    s->length = len;
    s->allocated = s->length * 2;
    s->data = (char *)malloc(s->allocated + 1);
    memcpy(s->data, &str->data[pos], s->length);
    s->data[s->length] = '\0';
    return s;
}

static
char *strGet(string *str)
{
    return str->data;
}

static
size_t findFirst(string *str, char what, size_t pos)
{
    for (size_t i = pos; i < str->length; ++i)
    {
        if (str->data[i] == what)
            return i;
    }
    return -1;
}

static
string **strTokenizer(string *str)
{
    string **res;
    res = malloc(sizeof(char *) * 90); // *90 (token num)
    for (int i = 0; i < 90; i++) // i<90 (token num)
        res[i] = malloc(sizeof(char) * 100); // *100 (a token length)
    size_t first = 0;
    size_t i = 0;
    while (first < str->length)
    {
        int second = findFirst(str, ' ', first);
        if (second == -1)
            second = str->length;
        string *token = strSubstring(str, first, second - first);
        if (*strGet(token) != ' ')
            res[i] = token;
        first = second + 1;
        ++i;
    }
    return res;
}

int main(void)
{
    string *fe = strCreate("A string \\ tof");
    string **r = strTokenizer(fe);
    for (int i = 0; i < 4; ++i)
    {
        printf("[[%s]]\n", strGet(r[i]));
    }
    return 0;
}

The output is:

[[A]]
[[string]]
[[\]]
[[tof]]

Note the use of [[ and ]] around the printed strings (and the newlines). It makes it easier to identify trailing blanks and embedded carriage return characters and various other mishaps.