Thread: C assignment question

It would probably be easier if you read in an entire line at a time, then used strchr (or just iterate straight thru the line), rather than working byte by byte during input.

Notice that strchr returns a pointer. You can reuse this pointer to search the rest of the string until strchr finally returns a null pointer indicating that the character could no longer the found.

Originally Posted by Martyn27

What if there is multiple occurences of a char on a single line?

Well, you want to process the entire line, yeah. Something like:

Code:

	char str[]="the quick brown i jumps over the lazy i.\n", *ptr = str;
	while ((ptr = strchr(ptr,'i'))) {

IMO this is going to be more awkward, it is easier to just go:

Code:

	char str[]="the quick brown fox i jumps over the lazy dog i.\n";
	int x, len = strlen(str);

	for (x=0;x<len;x++)
		if (str[i] == 'i' &&

I left that incomplete because you will need a more complete and complex set of conditions -- eg, if "i" must have a space on either side, then the for might as well be:

for(x=1;x<len-1;x++);

And you can just check to see that str[i-1] and str[i+1] are spaces. But I thought I'd also point out that if you have an i outside of a word at the beginning or end of the line, does that count (eg "the lazy dog i.\n")?

No and for one simple reason: reading one character at a time will require a Kreskin - Wikipedia, the free encyclopedia library to know what is coming next (ie ' ') though you can easily keep track of what came previously. More than that though is the gotcha involved with the situation where 'i' appears at the end or beginning of a line; the next or previous character is not going to be a space yet you will probably be required to recognize that it is a lone 'i' character. In a case line this it would probably pay to use whats called a sliding window buffer as opposed to a character by character or even line by line examination of the data. Say your data looks like this:

I have no way of knowing when
she will come home but I
am sure she will someday.

This would absolutely defeat both methods described. A sliding window buffer would look at (say) 3 characters at a time like:
[] denotes buffer
[I h]ave no way of knowing when\nshe will come home but I\nam sure she will someday.
Now you just look at the buffer:
Is the previous character whitespace/beginning of file? is the next? if so, capitalize.
Now move your buffer by one:
I[ ha]ve no way of knowing when...etc
So you keep sliding the window through the data and only look at the three characters in the window. This way you can be sure of catching stuff broken between lines/buffer edges, etc.

Originally Posted by Martyn27

Argh. To get it to read line by line it means im going to have to change my entire program.

You mean all 20 lines?

I'm not trying to belittle, just:

1) It cannot in reality mean changing "everything completely and starting from scratch" unless all the program does so far is open a file and read in some bytes.

2) This is programming. You can't recognize you've made a mistake and decide to forget about it. Even if it takes days to fix.

3) Here's a lesson about "modularity": if you write short discrete functions that accomplish a specific task in relation to the whole context, it is much easier to modify the parts. You should be writing with the awareness that each part needs to relate to the other parts in such a way that one part can be easily changed without having to change "the entire program". That is REALLY REALLY REALLY a fundamentally important principle.

So, a few hours (or, this case, I think more like a few minutes) rewriting will help impress this upon you. That's what I always say to myself...and it's true. I can read something somewhere and totally forget it. Then I remember, when it's too late and I'm having to deal with the consequences BUT my chances of forgetting after "learning the hard way" are much much less.

Anyway, you can always post your whole program and maybe someone will have some further insight WRT all this.

Okay you are parsing. Don't you see how it is easier to do this:

- read in one complete line.
- iterate thru the line one element at a time.

Some advantages: you don't have to do a string of else (strchr...) stuff. You can go:

Code:

int i, len = strlen(line);
for(i=0;i<len;i++) {
     switch (line[i]) {
          case ('#'): 
               	correctionCount++;
                break;
         case (' '):
                printf(" ");
	        wordCount++;
	        correctionCount++;
                break;
        case ('i'):
                if (line[i-1] == ' ' && line[i+1]) line[i] = 'I';
                break;

If you haven't used switch/case before and aren't suppose to, you can use the if/else.

Nb, using strchr() on a single character is EXTREMELY silly. Just compare (if c == ' ').

fgets() is perfect.

You can have as many cases as you want. Remember, every case should end with "break".