Complicating things?

[bconnor]

So basically what im trying to do is ...

Open file, read that file, than try to find ..

We or we and replace them with I, but not replace the cases where words contain We or we, such as Went, went, etc

a and replace them with the, but not replace the cases where words contain a, such as ability, apparent, etc

A and replace them with The, same for this ...

If it were a simple replace, id just getline, and find the We, and just replace it with I but i cant just do that, can i?

[sean]

No, you can't. What you need to do is check the elements in the array immediately following and immediately preceding the search string and make sure that they're " ". Also, how are you planning on "replacing" these tokens? There's no easy "replace" function in standard C strings - you need to read the data into a new string element by element, and do the replacement at that time.

[Maragato]

No, you can't. What you need to do is check the elements in the array immediately following and immediately preceding the search string and make sure that they're " ". Also, how are you planning on "replacing" these tokens? There's no easy "replace" function in standard C strings - you need to read the data into a new string element by element, and do the replacement at that time.

Can't he use strstr?

[bithub]

Can't he use strstr?

strstr() won't help much in this situation. You can use strstr() to find instances of "we" in a string, but you still need to check to make sure the characters around the "we" are whitespace (spaces, tabs, newline characters). Also you run into the problem of efficiency. You don't want your code to look like:

char *p;
while(p = strstr(text,"we"))
{
   /* handle situation */
   text = p + 1;
}
text = original_string;
while(p = strstr(text,"a"))
{
   etc...

If you check the text character by character, your program will move much more smoothly.

char *p;
for(p = text; *p != 0; ++p)
{
   switch(*p)
   {
   case 'W':
   case 'w':
      /* Check if we are at a valid "we" or "We" */
      break;
   case 'a':
   case 'A':
      /* check if we are at a valid "a" or "A" */
      break;
}

[WaltP]

As mentioned, copy the changed text into a new buffer.

strstr is good way to find the word you're looking for, then you need to test the characters before and after for whitespace, not SPACE. Do this with isspace().

If you find a candidate for a change, append the text from the last changed position to the current position of "we" into the new buffer. Then append the conversion text, and remember the position after "we". Repeat until done.

[bconnor]

This isnt what im trying to do is it?

void replace(char *a)
{
	int     in_fd, out_fd, n_chars;
	char	ch[BUFFERSIZE];
	while ( (n_chars = read(in_fd , buf, BUFFERSIZE)) > 0 )
	(
		do 
		{
			ch = getchar();
		
			if (ch = 'w')
				(
				if ((ispace(off_t lseek(int ch, off_t -1, int SEEK_CUR)) == 1 ) &&
				off_t lseek(int ch, off_t 1, int SEEK_CUR)) == 'e' &&
				(ispace(off_t lseek(int ch, off_t 2, int SEEK_CUR)) == 1 ))
					{
					
				
					}
				)
		
			if (ch = 'W')
				(
				if ((ispace(off_t lseek(int ch, off_t -1, int SEEK_CUR)) == 1 ) &&
				off_t lseek(int ch, off_t 1, int SEEK_CUR)) == 'e' &&
				(ispace(off_t lseek(int ch, off_t 2, int SEEK_CUR)) == 1 ))
					{
				
				
					}
				)
		
			if (ch = 'a')
				(
				if ((ispace(off_t lseek(int ch, off_t -1, int SEEK_CUR)) == 1 ) &&
				(ispace(off_t lseek(int ch, off_t 1, int SEEK_CUR)) == 1 ))
					{
				
				
					}
				)
			
			if (ch = 'A')
				(
				if ((ispace(off_t lseek(int ch, off_t -1, int SEEK_CUR)) == 1 ) &&
				(ispace(off_t lseek(int ch, off_t 1, int SEEK_CUR)) == 1 ))	
					{
				
				
					}
				)
		}
		while(a!=EOF)	
	)
		
	 if ( write( out_fd, buf, n_chars ) != n_chars )
             oops("Write error to ",av[1]);
             
       if ( close(in_fd) == -1 || close(out_fd) == -1 )
          	oops("Error closing files","");

}

[WaltP]

No. It's not even close. It won't even compile.

if (ch = 'w')

1) ch is defined as an array, not a sincle character
2) this simply sets ch to 'w'

lseek() is used to move around in files.
getchar() is used to read from the keyboard.
You can't lseek() the keyboard.

[bconnor]

No. It's not even close. It won't even compile.

if (ch = 'w')

1) ch is defined as an array, not a sincle character
2) this simply sets ch to 'w'

lseek() is used to move around in files.
getchar() is used to read from the keyboard.
You can't lseek() the keyboard.

Something more like...

char *ch;

ch = getchar();

if (ch == 'w')
(
if ((ispace(*ch - 1) == 1 ) && (*ch + 1 == 'e') && (ispace(*ch + 2) == 1))
					{
					
				
					}
				)

[quzah]

Something more like...

char *ch;

ch = getchar();

if (ch == 'w')

No, not at all. 'ch' is a pointer. It doesn't actually point to anything, and even if it did, the comparison to 'w' is wrong, as is the assignment from getchar. Try paying attention to your compiler some time.


Quzah.

[jafet]

You shouldn't be testing if your word is surrounded by whitespace.

We, a?

would not be replaced. You'd want to test if the surrounding characters are not letters, numbering etc.

[WaltP]

What you are trying to code is not going to work! Go back to our previous suggestions and use them.

[Maragato]

Oki I will give you the solution cause I'm into a great mood take the C Programming Language Second Edition (must have or must read if you really need to know C) section 4.1 basic of functions there is a program that can do this trick for you. Well it need small modifications but read that section of the book and you will do the trick.