Thread: String split() in C and C++

Hi, this is a cpp file. I was wondering how I could find a way to remove the need for the 3rd parameter of split() -> c function. As of now it is essential as it helps to free up the memory, and any way to improve the cpp split() function? Thanks.

Code:

#ifdef __cplusplus
extern "C" {
#endif // __cplusplus
#include <stdio.h>
#include <stdlib.h>
#include <string.h>


char** split(char* str, char delim, int* strcount)
{
    int delimCount = 0, index = 0;
    char new_delim[2], *p_new_d = new_delim;
    *p_new_d = delim;
    *(p_new_d + 1) = '\0';
    for (int i = 0; i < (int)strlen(str); i++)  if (str[i] == delim && str[i + 1] != delim) delimCount++;
    if (delimCount < 1) return NULL;
    char** container = (char**)malloc(sizeof(char*) * ++delimCount);
    char* pos = strtok(str, new_delim);
    while (pos != NULL)
    {
        container[index] = (char*)malloc(sizeof(char) * strlen(pos) + 1);
        strcpy(container[index++], pos);
        pos = strtok(NULL, new_delim);
    }
    *strcount = delimCount;
    return container;
}
#ifdef __cplusplus
}
#endif // __cplusplus


#include <iostream>
#include <string>
#include <sstream>
#include <vector>


class StringManip
{
private:
    enum class _split : bool { all = true, noEmpty = false };
public:
    typedef _split splitOption;


    static std::vector<std::string> split(std::string word, char delim, splitOption split_option)
    {
        std::vector<std::string> splited_data;
        std::stringstream ss(word);
        std::string token;
        while (std::getline(ss, token, delim))
        {
            if (split_option == splitOption::noEmpty) { if (!token.empty()) splited_data.push_back(token); }
            if (split_option == splitOption::all) splited_data.push_back(token);
        }
        return splited_data;
    }
};


int main()
{
    char str[] = "this/is/really/awesome";
    std::string ss = str;
    int size;
    char** splited = split(str, '/', &size);
    if (splited != NULL)
    {
        for (int i = 0; i < size; i++)
        {
            printf("%s\n", splited[i]);
            free(splited[i]);
        }
        free(splited);
    }
    for (auto& s : StringManip::split(ss, '/', StringManip::splitOption::noEmpty)) std::cout << s << '\n';
}

The other way is to set the last container[index] to NULL.

So in main, you have

Code:

while ( splited[i] != NULL ) {
    // stuff
    i++;
}

I suggest that you keep it. Replacing it with Salem's suggestion is an option, but it isn't an improvement as you would basically force the user of the function to iterate over the array to count the number of elements before they can safely use it. Fine if they were going to do that anyway, but inconvenient if they only wanted a specific element.

I wonder though: you're counting consecutive delimiters as a single delimiter, but your code with strtok does not account for that?