File I/O (newbie question)

[ntmurphy]

Hello,

I'm a C++ newbie, and I'm running into some problems. I'm trying to create a simple text program that reads from a file, and outputs to another file. However, my while loop executes one too many times, and it always duplicates the very last character in the text file. To control my loop, I'm using

while ( !infile.eof() ) { ....

I was told to use infile >> ws; but when I do this, it skips all of the whitespaces rather than just the end of the file. So an input file that says something like "Test 1 2 3", outputs "Test123". And obviously if I don't use infile >> ws; it outputs "Test 1 2 33". How do I fix this??

Nick

[Jordan]

Can you post the full code for reading and writing (including your loop)?

[ntmurphy]

#include <iostream>
#include <fstream>

int main(){
using namespace std;

ifstream infile;
ofstream outfile;

infile.open("infile.txt");
outfile.open("outfile.txt");

char ans = 'y';

while ( (infile.fail()) && (ans == 'y') )
{
cout << "File opening failed. Try again? ";
cin >> ans;
if (ans != 'y')
{return 0;}
}

char symbol;
while ( !infile.eof() )
{
infile.get(symbol);
infile >> ws;
outfile << symbol;

}


infile.close();
outfile.close();
return 0;
}

I tried replacing "infile.get(symbol)" with "infile >> symbol", and tried with and without "infile >> ws" in various combinations. But I can't get the program to duplicate the infile exactly in the outfile... Any suggestions? Its gotta be something simple I'm missing...

[ntmurphy]

ahh..I fixed it....

Heres what I did below. Just moved the input statement before the loop for the first symbol, then kept it as the last task inside the loop, so before it can execute one last time, it had already passed the end of the file...

...

infile.get(symbol);

while ( !infile.eof() ) {

outfile << symbol;
infile.get(symbol);
}
....

what can I say...clearly a noob here.

[deusprogrammer]

Why not just use in.getline() ? In the case of interpreting a text file, it's best to read in the whole line and then parse it in a buffer. Or you can buffer the whole file and then look through it searching for tokens to parse it and interpret the data. For the most part though you don't want to use files like this. It's easier to just output binary and then just read everything back in as binary.