[C] - String Manipulation {Beginner Programmer}

**INFERNO2K** · 05-20-2005

Hello.
I have an assignment where I need to clean up a string, removing tabs, new lines, leading and trailing whitespace, yet leave proper spaces between words.

So far I have been able to remove newlines and tabs, and replace them with a single space.

Code:

char s[] = "  widget;Acme  \t  \n Co.;gear    induction\b\tdevice:   \n "

Code:

int cleanSpace(char s[])
{
    int i, j;
    
    for(i = 0; s[i]; i++)
    {
          if(s[i] == '\t' || s[i] == '\n')
          {
                  s[i] = ' ';
          }
    }
}

This will now output,

Code:

"   widget;Acme       Co.;gear    inductio device:    ".

Now that the newlines and tabs are gone, I need to remove the leading, trailing and multiple whitespace in the middle.

I am confused as to how I can approach this.

My professor insisted that I do not use ctype.h (isspace), that I work with my given function header, I do not use pointers or second buffers.

He hinted that I should use 4 loops, with the first doing what I posted above. He hinted that I should find multiple whitespace, find the next non space and then copy everything down a spot.

He insisted that I break it down, as I have above. Starting with removing new lines/tabs and replacing them with a space.

Any help strongly appreciated.

Thank you in advance.

**itsme86** · 05-20-2005

Well, there's more than one way to remove characters from a string. This is something that comes up a lot here. You basically have a couple of ways to do it:

1) When you find an offending character, loop through the rest of the string, copying that character into the string index just before it. So like, str[index] = str[index + 1];.
2) Create a second string and only copy non-offending characters from the original string into it.

You can use one of those tactics for your remaining tasks: removing leading, multiple, and trailing spaces.

**pianorain** · 05-20-2005

Alternately, a simple state engine will allow you to solve this without looping through the string more than once. Consider the following pseudocode:

Code:

set currentIndex = 0
set insertIndex = 0
set state = State_Begin
while not at end of string
    //get rid of other whitespace
    if string[currentIndex] is a type of whitespace other than a space
        replace it with a space
    //skip over all space characters until a non-space character
    if state is State_Begin
        if string[currentIndex] is not a space
            state = State_Normal
        end if
    end if
    //copy characters up to a space character, then switch states
    if state is State_Normal
        copy string[currentIndex] to string[insertIndex]
        increment insertIndex
        if string[currentIndex] is a space
            state = State_Begin
        end if
    end if
    increment currentIndex
end while
place the null character appropriately

**Thantos** · 05-20-2005

Kinda like what itsme said but a little different.

You keep two index. One that indicates where to put the next valid character and one that tells you which one you are looking at. This keeps you from having to shuffle too much.

You'll also need to keep a couple "flags"

Lets take look at just removing the middle extra spaces:
example:

Code:

char arr[] = "Hello    World";

That has 4 spaces in between. Lets start a look by setting our two indexs, called valid and current, both to 0

Code:

valid = current = 0;

Lets also keep a flag that lets us know if we've already copied a space over.

Code:

int copiedSpace = 0;

Now we compare the current character to space

Code:

if ( arr[current] != ' ')

if its not a space we need to copy it over to the the valid spot and increment valid

Code:

{
  arr[valid++] = arr[current];
  copiedSpace = 0; /* We didn't just copy a space over*/
}

. If it was a space we need to see if we've already copied over a space

Code:

else if ( copiedSpace == 0 )
{
  arr[valid++] = ' ';
  copiedSpace = 1; /* Because we did just copy a space over*/
}

but no matter what we have to move on to the next character

Code:

current++;

Do this until current gets to the end of the string and then throw a null character in at valid

Code:

arr[valid] = '\0';

We a couple more status flags you can remove all the extra spaces in one pass

**pianorain** · 05-20-2005

Originally Posted by Thantos

Do this until current gets to the end of the string and then throw a null character in at valid

Code:

arr[valid] = '\0';

Careful...gotta remove all trailing spaces.

If you just throw a null on the end and the original string had trailing spaces, then you'll have one trailing space at the end too.

**Thantos** · 05-20-2005

Read my example string. I was dealing with a case of only spaces in the middle. Thats why AFTER that I made mention of needing further flags. And using my process you'd have at most one trailing space so its a simple if statement to deal with that.

**pianorain** · 05-20-2005

Hence why reading is better than skimming. My bad.

**Thantos** · 05-20-2005

BTW I just it at home and there isn't any need for additonal flags.
To take care of leading spaces just initalize copiedSpace to 1
Trailing spaces are taken care by looking to see if copiedSpace is 1 or 0 prior to appending the null and adjusting valid

**INFERNO2K** · 05-20-2005

Oh man I am so confused.

Can somebody test out this string (fixing up the whitespace and removal of \t's and \n's and tell me if strlen 39 is returned? If 39 is returned that means I pass the test in my test main();

Code:

  widget;Acme  \t  \n Co.;gear    induction\b\tdevice:   \n

**Thantos** · 05-20-2005

Yeah I'm getting 39 also

**INFERNO2K** · 05-20-2005

Ok, after fiddling around with this code. This is what I have produced.

Code:

    int curChar, nextAvail;
 
        for (curChar = nextAvail = 0; s[curChar]; curChar++)
        {
            if (s[curChar] != ' ' && s[curChar] != '\t' && s[curChar] != '\n'
            && s[curChar] != '\b')
            {
                s[nextAvail++] = s[curChar];
            }
            else
            {
               if (nextAvail > 0 && s[nextAvail-1] != ' ')
               {
                   s[nextAvail++] = ' ';
               }
            }
        }
        
        s[nextAvail] = 0;

        return strlen(s);

My run of the test main(); produces this

Code:

Testing your init function.

Failed test eight, customer should contain:
widget;Acme Co.;gear induction device:♂

your string contained:
widget;Acme Co.;gear induction device:

What the heck is that funky symbol at the end of what it should contain, and what code do I need to fix this up?

**INFERNO2K** · 05-20-2005

Ok I figured out that is a vertical tab.

One test it wants a vertical tab, the next it doesnt

**INFERNO2K** · 05-20-2005

It would be best if you could see what I am talking about.

I am posting my code and my assignments main.

**INFERNO2K** · 05-21-2005

Please, can someone have a look and compile it to see the returned test error.

**quzah** · 05-21-2005

Why don't you just tell us what the error is? Oh, and stop compiling as C++.

Quzah.

Thread: [C] - String Manipulation {Beginner Programmer}

Thread Tools

Search Thread

Display

[C] - String Manipulation {Beginner Programmer}

Similar Threads

[Inheritance Hierarchy] User Input on program with constructors. How ?

String Manipulation problems -_-

String editor for a sentence inputted by a user - any suggestions or ideas?

Program using classes - keeps crashing

can anyone see anything wrong with this code