CODEWITHSUNDEEP
carraystokenizedynamic-arraysstrtok
zachem64
zachem64

Reputation: 15

Tokenizing string from dynamic array into multiple lines in static 2D char array

I have a dynamic array that holds a string containing '\n' characters, so this string is made up of multiple lines. I'm trying to extract the lines and put them all into a 2D char array and I'm getting segmentation errors.

Here's my code:

    char *input_lines = malloc(MAX_LINE_LEN*sizeof(char)); 
    input_lines = extractInput(MAX_LINE_LEN, input_file);
    char inputLines_counted[lineCount_input][MAX_LINE_LEN];


    char *t = strtok(input_lines, "\n");
    for(i = 0; i < lineCount_input; i++) {
        strcpy(inputLines_counted[i], t);
        // printf("%s\n", inputLines_counted[i]);
        t = strtok(NULL, "\n");
    }

Upon creating the dynamic array, I use the extractInput(MAX_LINE_LEN, input_file) function to populate the input_lines array with a string containing multiple lines.

Here's the extract function: 

char *extractInput(int len, FILE *file) {
    char tmp[len];
    char *pos;
    char *input_lines = malloc(len*sizeof(char)); 
    char *lines;
    while(fgets(tmp, len, file)) {
        // if((pos = strchr(tmp, '\n')) != NULL) {
        //     *pos = ' ';
        // }
        input_lines = realloc(input_lines, (strlen(input_lines) + len)*sizeof(char));
        strcat(input_lines, tmp);
    }
    return input_lines;
}

Why am I getting segfaults here?

Upvotes: 0

Views: 107

Answers (1)

AndersK
AndersK

Reputation: 36102

The function call

    input_lines = realloc(input_lines, (strlen(input_lines) + len)*sizeof(char));

takes your current allocated memory block and expands it, if it can. you should check the return value of realloc, it may fail.

btw when you allocate memory in C, you always need to have space for the ending \0.

see what happens with this file

hello\n
world\n

The first fgets reads in hello\n into tmp.

you now do realloc even though it is unnecessary, input_lines is already pointing to a buffer that could hold the string

char *input_lines = malloc(MAX_LINE_LEN*sizeof(char));

now with your realloc

input_lines = realloc(input_lines, (strlen(input_lines) + len)*sizeof(char));

you do strlen(input_lines) + len so you make the buffer strlen("hello\n") + len long.

but the important thing you need to notice is the following line

strcat(input_lines, tmp);

you have not initialized the memory that input_lines is pointing to, it can contain anything even \0's so your strcat could potentially put the string anywhere in the buffer and cause the error you describe.

Either do a memset or use calloc when you allocate the buffer.

If you use realloc you should keep track of the total size that you have allocated and how much you are using of it, before you copy into the buffer check if there is enough room. If not, add a certain number of bytes to the buffer.

I also noticed you read from the file line by line, then you concatenated the lines together to later use strtok to divide them again. It would be more efficient to return an array of lines.

Upvotes: 1

Related Questions