L-19 MCS 260 Wed 3 Oct 2001

In this lecture we saw another application of a macro of ctype.h and arithmetic with ASCII codes. Important is the technique of keep track of current and previous character.

/* L-19 MCS 260 Wed 3 Oct 2001 : scanning input for numbers

Suppose we have an input stream of characters, like
 a be 124 2d dd3 d3  342,4353 023 aa 0 dd ( 232 dda
then the output contains only the integer numbers :
 124 2 3 3 342 4353 23 0 232

This program is best applied with redirection, putting the
input in a file, say "stream" and running as a.out < stream.  */

#include<stdio.h>
#include<ctype.h>

int main(void)
{
   int current;
   int previous = ' ';
   int nb = 0;

   while ((current = getchar()) != EOF)
   {
      if isdigit(current)
         nb = 10*nb + current - '0';
      else if isdigit(previous)
      {
         printf(" %d", nb);
         nb = 0;
      }
      previous = current;
   }

   return 0;
}

The technique of current and previous character is again applied in the following program.

/* L-19 MCS 260 Wed 3 Oct 2001 : counting words of the input.

In an input stream of characters, we consider words anything
that is separated by spaces.  If the input is something like

  this  is a  test

then the program counts four words.  Observe that more than
one space can separate the words.

This program is best applied with redirection, putting the
words in a file, say "words" and running as a.out < words.  */ 

#include <stdio.h>
#include <ctype.h>

int main(void)
{
   int current;
   int previous = ' ';
   int cnt = 0;

   while ((current = getchar()) != EOF)
   {
      if (isspace(previous) && !isspace(current)) cnt++;
      previous = current;
   }

   printf("Number of words : %d\n", cnt);

   return 0;
}

An extension of this program is to reformat the input so that every word is separated by exactly one space.