file i/o and remove something

**bnkslo** · 05-17-2008

Hi,

i have a problem of removing some text from a file between this characters <>. For example:

input file
#include <stdio.h>

output file
#include <>

I did this so far:

Code:

#include <stdio.h>

int main()
{
	
	FILE *vData, *iData;
	int c;
	
	vData=fopen("test.txt", "rt");/*vhodDatoteka*/  // input file
	iData=fopen("test01.txt", "wt");/*izhodDatoteka*/ // output file
	
	while((c=fgetc(vData))!=EOF){  //reading the character
		putc(c, iData);            // copying characters to output file
		/*if(c=='<'){
		
	}*/
	}
	
	fclose(vData);
	fclose(iData);
	
	return 0;
}

And this is where iam stuck. I dont know how to countinue. Can anyone please give me some pointers how to solve this.

Thank you.

**~~(::)~~** · 05-17-2008

Code:

if c == '<':
    skip this '<'
    skip until you read '>'
    skip this '>' also
    start copying again

**bnkslo** · 05-17-2008

I thought of that. But what if i have in the text:

input file
for(i=0;i<15;i++)

output file
for(i=0;i<..... where does it end ?

So i think i cant just stop copying with this character < and countinue when i read character > ? I have to find the end character first i think ?

**root4** · 05-17-2008

When you fall on the '<', store the characters read in a buffer. If you reach the end of file or another '<', then you can flush the buffer. If the matching '>' is found, you can clear the buffer and resume the output. It also depends on what you wish to do with nested sequences <...<...>...> etc.

**Cactus_Hugger** · 05-17-2008

You're correct - you can't just start with <.
If you're only looking at includes, just make sure the line starts with #include, then, for the rest of the line, do your replacement.

Originally Posted by root4

When you fall on the '<', store the characters read in a buffer. If you reach the end of file or another '<', then you can flush the buffer. If the matching '>' is found, you can clear the buffer and resume the output. It also depends on what you wish to do with nested sequences <...<...>...> etc.

What about:

Code:

for(; a < 3 && b > 5; ++a, ++b)

Since this is doable by a simple regex - is a program really needed?

**bnkslo** · 05-17-2008

Originally Posted by Cactus_Hugger
Since this is doable by a simple regex - is a program really needed?

It's for homework.

Another question. How can you go back to the first character ('<') when i found the last one ('>')?

**Elysia** · 05-17-2008

Why would you?
But to answer it: keep track of where you found the '<' (store it in a variable somewhere). Then you would know where that position is, so you can revert to it later.
Though you might have nested '<', I think it's best to start off small and assume it's not nested.

**root4** · 05-17-2008

What about:

Code:

for(; a < 3 && b > 5; ++a, ++b)

At this point (if the code above should not be interpreted as a normal <...> sequence), you need a complete C parser to make the difference (as this a semantical problem)...
This constraint was not precised in the 1st post.

EDIT: in fact the real question is: what are the characters allowed between < and > ? alpha numeric characters only? spaces too? anything? depending on that the solution would be different.

**bnkslo** · 05-19-2008

Originally Posted by root4
in fact the real question is: what are the characters allowed between < and > ? alpha numeric characters only? spaces too? anything? depending on that the solution would be different.

I have to copy the characters from input file to the output file. But in the process i have to remove everything that is between the characters <>...

Example:

input file

Today is a <sunny> day.

The output file should look like this.

Today is a <> day.

**Elysia** · 05-19-2008

You just need to remember when a '<' is found and skip all characters from there until you find a '>'.

**bnkslo** · 06-23-2008

root4 wrote
When you fall on the '<', store the characters read in a buffer. If you reach the end of file or another '<', then you can flush the buffer. If the matching '>' is found, you can clear the buffer and resume the output. It also depends on what you wish to do with nested sequences <...<...>...> etc.

Some questions?

1. How much buffer? Only for one char or do i allocate for the whole file?
2. I was reading about allocating memory but I'am still confused?

Can i do something like this?

Code:

           char *buffer;

           buffer=malloc(sizeof(vData)); // or sizeof(char)

3.Can I use buffer now like a variable? Like put char in it with putc or something?

**tabstop** · 06-23-2008

1. How many characters do you think there will be between < and >?
2. The thing inside malloc is how many characters you want. See #1.
3. buffer is a variable, but it is not a char variable. It points to a char. It would be more analogous to char foo[50], than char foo.

**bnkslo** · 06-23-2008

tabstop
1. How many characters do you think there will be between < and >?
2. The thing inside malloc is how many characters you want. See #1.
3. buffer is a variable, but it is not a char variable. It points to a char. It would be more analogous to char foo[50], than char foo.

1. Well that is the question? It can be 1 or maybe 100. That's something i can't predict. Or it could happen that when i find < there is not a > character and i get to EOF. So i don't have a number!?

**Elysia** · 06-23-2008

Exactly.
So if you don't have the numbers of characters, then how would you proceed?
Think of a real life situation. If you can't find anymore inside something, then what do you need to do?

**bnkslo** · 06-23-2008

I would say i need so much memory as the file is big? But i could be wrong!?

Thread: file i/o and remove something

Thread Tools

Search Thread

Display

file i/o and remove something

Similar Threads

Data Structure Eror

File I/O Question

Dikumud

Batch file programming

simulate Grep command in Unix using C