Thread: filtering

Need help with this problem:
Convert all the letters in file to lower case, and any non letter character is converted to a space. Consecutive spaces should be compressed into one single space.

This is what I have so far:

Code:

#include <stdio.h>
      #define FALSE 0
      #define TRUE 1

int replace(int c){
  
  if (ispunct(c) || isdigit(c))
  {
    putchar(' ');
  
      while ((c = getchar() ) != EOF && ispunct(c) || isdigit(c))
      { }
  }
  if (c != EOF)
  
    return c; 
}


int main()
{
     int c;
     int start= TRUE;

      while ( (c = getchar() ) != EOF)
      {
        if (isspace(c))
        {

           if(start==FALSE)
           putchar(' ');
      
           while ( (c = getchar() ) != EOF && isspace(c))
           {}
        } 
           if (c != EOF)
           {        
           putchar(tolower(replace((c))));
              start = FALSE;
           }
      } printf("\n");
}

The test file consists of:

Code:

Hello1431387409876187461203470981743Hello                   (works)
Hello@$#%@%#$%@#$%^&(*%&^(*$%*Hello                   (works)
Hello!2@3$4%4^7&6&5$6&5%4%3$2$5$3Hello                    (works)
Hello11$%33%%3452345$%$%#45453$%Hello                   (works)
Hello                               Hello                   (works)
Hello            $1%4#             Hello                   (doesn't work)

desired output:

Code:

Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello

current output:

Code:

Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello      Hello

hope this is clear, any suggestions?

Another way to look at the problem is to consider your program as a simple state machine. There are really three modes your program can be in: inside a word, handling the first space, and handling subsequent spaces. The middle mode can be viewed as a transition between the other two modes. If you look at the problem this way the solution is really quite simple; it goes something like this.

Code:

enum mode_t {
    MODE_IN_WORD,
    MODE_BETWEEN_WORDS
};

mode_t mode = MODE_IN_WORD;

while((c = getchar()) != EOF) {
    if(mode == MODE_IN_WORD) {
        if(isalpha(c)) {
            putchar(tolower(c));
        }
        else {
            /* transition to the state MODE_BETWEEN_WORDS */
            mode = MODE_BETWEEN_WORDS;
            putchar(' ');
        }
    }
    else {  /* MODE_BETWEEN_WORDS */
        if(isalpha(c)) {
            /* transition to the state MODE_IN_WORD */
            putchar(tolower(c));
            mode = MODE_IN_WORD;
        }
        else {
            /* ignore this character */
        }
    }
}

That's not perhaps the most elegant way of doing things, but I hope it's understandable. If there were more than two states it would be nice to have a more structured definition of what a state is and how one transitions between them, but as it is I think that would do the job . . . .