Thread: fstream doesn't work well

The code

Code:

fstream myfile;

myfile.open("C:\\myfile.ddt")

Doesn't create a new file.

I tried

Code:

myfile.open("C:\\myfile.ddt", ios::in | ios::out)

too.

My book says it should create a file if the file does not exist.
It is my first use of fstream. Using ios:ut alone is alright.

Of course it is for both input and output else I wouldn't use fstream. It is a database so it can't be 0 lengthed each time.

Thanks, I'll do that.
My book is old. And this behavior of fstream seems bad to me.

Once I started a thread and asked if C++ STL uses C functions or not. From the answers I got, I concluded it doesn't. But at least in my compiler it uses C functions to handle the files.

Originally Posted by siavoshkc

Once I started a thread and asked if C++ STL uses C functions or not. From the answers I got, I concluded it doesn't. But at least in my compiler it uses C functions to handle the files.

That was a wrong conclusion. If it's the thread I remember, we were very explicit in telling you that it is the library implementation's business how it achieves the specified functionality, and that you as the client programmer have no business inquiring about it. Let the library do what it wants, as long as it does what it claims.

it is the library implementation's business how it achieves the specified functionality, and that you as the client programmer have no business inquiring about it.

I just wanted to know that, by standard ,C++ STL uses C routines or not. I concluded there is no standard for it, and what is standard, is just the function duty.
[edit]
When new C++ standard will be defined?

Whether or not it uses C routines does not mater. The implementation is hidden because you need not know how it was implemented to use it. As long as the implementation follows standards and performs as expected by those standards.

From what I hear, new ISO C++ should be in '09. But, that was from the freenode ##C++ channel.

Originally Posted by Wraithan

From what I hear, new ISO C++ should be in '09. But, that was from the freenode ##C++ channel.

Let's put it this way: they want it to really be a C++0x standard, i.e. they want to finish it before 2010. But they're not exactly confident they'll finish it before 09, so yeah, that's the most likely year.

Whether or not it uses C routines does not mater. The implementation is hidden because you need not know how it was implemented to use it. As long as the implementation follows standards and performs as expected by those standards.

It is not my fault. I went to MSDN for fstream, it passed me to basic_fsteram and from there to rdbuf->open() and next to fopen(), explaining

Code:

ios_base::in becomes "r" (open existing file for reading).

ios_base::out or ios_base::out | ios_base::trunc becomes "w" (truncate existing file or create for writing).

ios_base::out | app becomes "a" (open existing file for appending all writes).

ios_base::in | ios_base::out becomes "r+" (open existing file for reading and writing).

ios_base::in | ios_base::out | ios_base::trunc becomes "w+" (truncate existing file or create for reading and writing).

ios_base::in | ios_base::out | ios_base::app becomes "a+" (open existing file for reading and for appending all writes).

Why C++ STL should call a C function. This makes people think C is more powerful that C++.

Originally Posted by siavoshkc

Why C++ STL should call a C function. This makes people think C is more powerful that C++.

How so? How do you define power, anyway?