Remove extra white space from inside a C string?

Question

I have read a few lines of text into an array of C-strings. The lines have an arbitrary number of tab or space-delimited columns, and I am trying to figure out how to remove

Answer 1

If I may voice the "you're doing it wrong" opinion, why not just eliminate the whitespace while reading? Use fscanf("%s", string); to read a "word" (non whitespace), then read the whitespace. If it's spaces or tabs, keep reading into one "line" of data. If it's a newline, start a new entry. It's probably easiest in C to get the data into a format you can work with as soon as possible, rather than trying to do heavy-duty text manipulation.

Answer 2

Why not use strtok() directly? No need to modify the input

All you need to do is repeat strtok() until you get 3 non-space tokens and then you are done!

Answer 3

The following code modifies the string in place; if you don't want to destroy your original input, you can pass a second buffer to receive the modified string. Should be fairly self-explanatory:

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

char *squeeze(char *str)
{
  int r; /* next character to be read */
  int w; /* next character to be written */

  r=w=0;
  while (str[r])
  {
    if (isspace(str[r]) || iscntrl(str[r]))
    {
      if (w > 0 && !isspace(str[w-1]))
        str[w++] = ' ';
    }
    else
      str[w++] = str[r];
    r++;
  }
  str[w] = 0;
  return str;
}

int main(void)
{
  char test[] = "\t\nThis\nis\ta\b     test.";
  printf("test = %s\n", test);
  printf("squeeze(test) = %s\n", squeeze(test));
  return 0;
}

Answer 4

char* trimwhitespace(char *str_base) {
    char* buffer = str_base;
    while((buffer = strchr(str_base, ' '))) {
        strcpy(buffer, buffer+1);
    }

    return str_base;
}

Answer 5

I made a small improvment over John Bode's to remove trailing whitespace as well:

#include <ctype.h>

char *squeeze(char *str)
{
  char* r; /* next character to be read */
  char* w; /* next character to be written */
  char c;
  int sp, sp_old = 0;

  r=w=str;

  do {
    c=*r;
    sp = isspace(c);
    if (!sp) {
      if (sp_old && c) {
        // don't add a space at end of string
        *w++ = ' ';
      }
      *w++ = c;
    }
    if (str < w) {
      // don't add space at start of line
      sp_old = sp;
    }
    r++;
  }
  while (c);

  return str;
}

#include <stdio.h>

int main(void)
{
  char test[] = "\t\nThis\nis\ta\f     test.\n\t\n";
  //printf("test = %s\n", test);
  printf("squeeze(test) = '%s'\n", squeeze(test));
  return 0;
}

br.

Answer 6

Edit: I originally had a malloced workspace, which I though might be clearer. However, doing it w/o extra memory is almost as simple, and I'm being pushed that way in comments and personal IMs, so, here comes...:-)

void squeezespaces(char* row, char separator) {
  char *current = row;
  int spacing = 0;
  int i;

  for(i=0; row[i]; ++i) {
    if(row[i]==' ') {
      if (!spacing) {
        /* start of a run of spaces -> separator */
        *current++ = separator
        spacing = 1;
      }
    } else {
      *current++ = row[i];
      spacing = 0;
  }
  *current = 0;    
}