Delete/substitute two or more consecutive chars in array of chars

Question

I am learning C and have to code a program that:

• reads an amount of chars from stdin and stores them in an array;
• substitute any two or more consecutive spaces (' ') for only one;
• write back the array.

So far I have done the following:

#include <stdio.h>
#define DIM 100

int main(int argc, char **argv)
{
    char mainArray [DIM]={'\0'};
    int auxArray [DIM];
    int i, m, n, c, l;

    printf("Enter a string containing two or more consecutive spaces:\n");

/* Read string from stdin */
    do
    {
        mainArray[i]=getchar();
        ++i;
    } while ((mainArray[i-1] != '\n') && (i < DIM-1));

/* Place the string terminator '\0' */
    if (i < DIM)
        mainArray[i]='\0';
    else
        mainArray[DIM-1]='\0';

    l=i;

/* My substitution algorithm */
    for (i=0 ; mainArray[i] != '\0' ; ++i)
    {
        if (mainArray[i] == ' ')
        {
            if (mainArray[i] == mainArray[i+1])
            {
                auxArray[m]=i;
                ++m;
            }
        }
    }

    for (i=0 ; i < m ; ++i)
    {
        for (c=auxArray[i] ; c < l-1 ; ++c)
            mainArray[n]=mainArray[n+1];
    }

/* Display the corrected string */
    for (i=0 ; mainArray[i] != '\0' ; ++i)
        printf("%c", mainArray[i]);

    return 0;
}

As an example, entering the string "a_long_time_ago___in_a_galaxy__far____,_far_away.." would produce "a_long_time_ago_in_a_galaxy_far_,_far_away.."

For the substitution algorithm I thought that one possiblity could be to store the positions of the spaces in excess, and then delete the spaces in the main array through the auxiliary array.

I am sure I am making some amateaur mistake. Also, how can the code be optimized in your opinion?

Thank you in advance.

Answer 1

its not code review, but you have several obvious problems with the code:

mainArray[i]=getchar(); - you haven't initialized i, so you're writing to some random memory address, probably you're compiling in debug mode, where compiler sets variables to 0, but in general case - don't do this

if (i < DIM) - you don't need it as i after loop will point either to next symbol after \n or it will be equal to DIM-1, so justmainArray[i]='\0';` is enough

auxArray[m]=i; - m is not initialized

for (i=0 ; i < m ; ++i) I cannot understand logic of this loop, especially mainArray[n]=mainArray[n+1]; - you're not changing n. Whole approach with storing spaces indexes is wrong, after you deleted at least one space sequence - all indexes in array become wrong ones.

here is a little bit different version with one array only:

#include <stdio.h>
#include <string.h>
#define DIM 100

int main(int argc, char **argv)
{
    char mainArray [DIM]={'\0'};
    int i, j, k;

    printf("Enter a string containing two or more consecutive spaces:\n");

    fgets(mainArray, DIM, stdin); // note: last symbols of the array will be \n\0
    mainArray[strlen(mainArray)-1] = '\0'; // trimming \n

    printf("[%s]\n", mainArray);

    for (i=0; mainArray[i] != '\0' ; ++i) {
        if (mainArray[i] == ' ') {
            for (j=(i+1); mainArray[j]==' '; ++j) ; // calculate end of space sequence, j will point to the first non space symbol
            for (k=j; mainArray[k]!='\0'; ++k) // copy rest of string to the position after first space
                mainArray[i+1+(k-j)] = mainArray[k];
            mainArray[i+1+(k-j)] = '\0'; // put null-terminator
        }
    }

    printf("[%s]\n", mainArray);
    return 0;
}

Answer 2

Here's an alternate way using just char* pointers:

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

int
main(void)
{
    char *lhs;
    const char *rhs;
    int chr;
    char mainArray[500];

    fgets(mainArray,sizeof(mainArray),stdin);

    lhs = strchr(mainArray,'\n');
    if (lhs != NULL)
        *lhs = 0;

    printf("INP: '%s'\n",mainArray);

    lhs = mainArray;
    rhs = mainArray;

    for (chr = *rhs++;  chr != 0;  chr = *rhs++) {
        *lhs++ = chr;
        if (chr != ' ')
            continue;

        for (chr = *rhs;  chr != 0;  chr = *++rhs) {
            if (chr != ' ')
                break;
        }
    }

    *lhs = 0;

    printf("OUT: '%s'\n",mainArray);

    return 0;
}

UPDATE: Here's the above recoded to use indexes. Note that it's more verbose for no good effect. But, the logic is still just as simple. That's one of the main points. For future reference, it's important to keep your logic "as simple as possible--and no simpler" because when projects grow, the effect gets magnified for good or for evil.

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

int
main(void)
{
    char *cp;
    int lhs;
    int rhs;
    int chr;
    char mainArray[500];

    fgets(mainArray,sizeof(mainArray),stdin);

    cp = strchr(mainArray,'\n');
    if (cp != NULL)
        *cp = 0;

    printf("INP: '%s'\n",mainArray);

    lhs = 0;
    rhs = 0;

    for (chr = mainArray[rhs++];  chr != 0;  chr = mainArray[rhs++]) {
        mainArray[lhs++] = chr;
        if (chr != ' ')
            continue;

        for (chr = mainArray[rhs];  chr != 0;  chr = mainArray[++rhs]) {
            if (chr != ' ')
                break;
        }
    }

    mainArray[lhs] = 0;

    printf("OUT: '%s'\n",mainArray);

    return 0;
}

Also, see my answer here: Issue implementing dynamic array of structures

It's about arrays of structs [which might seem to be a little advanced] but just pretend that the partition type was typedef int partition or typedef char partition. Also, see the last part about int arrays and pointers being interchangeable.