Question about clrscr() in FAQ (Win32)

here's the clrscr() function found here :

Code:

void clrscr(void)
{
    COORD                       coordScreen = { 0, 0 };
    DWORD                       cCharsWritten;
    CONSOLE_SCREEN_BUFFER_INFO  csbi;
    DWORD                       dwConSize;
    HANDLE                      hConsole = GetStdHandle(STD_OUTPUT_HANDLE);
    GetConsoleScreenBufferInfo(hConsole, &csbi);
    dwConSize = csbi.dwSize.X * csbi.dwSize.Y;
    FillConsoleOutputCharacter(hConsole, TEXT(' '), 
                               dwConSize, coordScreen, &cCharsWritten);
    GetConsoleScreenBufferInfo(hConsole, &csbi);
    FillConsoleOutputAttribute(hConsole, csbi.wAttributes, 
                               dwConSize, coordScreen, &cCharsWritten);
    SetConsoleCursorPosition(hConsole, coordScreen);
}

I got three questions about it :

1. Why did the author call GetConsoleScreenBufferInfo() again although csbi already contains the screen attributes? (red text)

2. Whats the difference between using ' ' and TEXT(' ') ? i searched mdsn but couldn't find anything. (blue text)

3. And finally , is the following clrscr() function equivalent to the above or not (if not , wich one is better) :

Code:

void clrscr(void)
{
    COORD                       coordScreen = { 0, 0 };
    DWORD                       cCharsWritten;
    CONSOLE_SCREEN_BUFFER_INFO  csbi;
    DWORD                       dwConSize;
    HANDLE                      hConsole = GetStdHandle(STD_OUTPUT_HANDLE);
    GetConsoleScreenBufferInfo(hConsole, &csbi);
    dwConSize = csbi.dwSize.X * csbi.dwSize.Y;

    FillConsoleOutputCharacter(hConsole, ' ', 
                               dwConSize, coordScreen, &cCharsWritten);
   
    FillConsoleOutputAttribute(hConsole, 0 , dwConSize, coordScreen, &cCharsWritten);

    SetConsoleCursorPosition(hConsole, coordScreen);
}

I just took off the other GetConsoleScreenBufferInfo() and specified the character attributes myself.


Thanks for your time.

1. Remeber he is only declaring a CONSOLE_SCREEN_BUFFER_INFO type that contains no data, so when he calls that function it puts that data into cbsi.
2. I am not sure if there is a difference but TEXT is a macro defined as

Code:

//this is from the Win32 Programmers Reference
TEXT(LPTSTR string);	// address of ANSI or Unicode string

3. I'll let someone else answer

1. Remeber he is only declaring a stuct so that contains no data, so when he calls that function it put that data into cbsi.

I think you missed the first call to GetConsoleScreenBufferInfo(). Re-read the code and you'll see that he actually filled csbi twice , my question was why did he do that

Ah I see that now, I don't see the need also.
As for yours I was looking at it your might want the 0 in the FillConsoleOutputAttribute
to be what the author used as the declaration for it is

Code:

BOOL FillConsoleOutputAttribute(

    HANDLE hConsoleOutput,	// handle to screen buffer 
    WORD wAttribute,	// color attribute to write 
    DWORD nLength,	// number of character cells to write to 
    COORD dwWriteCoord,	// x- and y-coordinates of first cell 
    LPDWORD lpNumberOfAttrsWritten 	// pointer to number of cells written to 
   );

The TEXT macro is used so the string is made up of TCHARs instead of chars.

Also, your first method is better since it preserves the console attributes.

As for yours I was looking at it your might want the 0 in the FillConsoleOutputAttribute
to be what the author used as the declaration for it is

Also, your first method is better since it preserves the console attributes.

but why? 0 means black (for both foreground and background color) and since all console windows are black (AFAIK) how are the two codes different?

The TEXT macro is used so the string is made up of TCHARs instead of chars.

Whats the difference between TCHAR and the normal 'char'? (i googled it but couldn't find anything useful)

but why? 0 means black (for both foreground and background color) and since all console windows are black (AFAIK) how are the two codes different?

console windows are black by default, but they can be many different colors.

Whats the difference between TCHAR and the normal 'char'?

A tchar is something that is defined in tchar.h. It is used so that a programmer doesn't have to have two sets of source code for both unicode and ascii. When you work with TCHARs, you should use the correct string functions.
Like _tcscpy instead of strcpy, and _tcslen instead of strlen. If you correctly code using TCHARs instead of chars, then all you need to do is define _UNICODE to make your program unicode instead of ASCII.

Thanks bithub , that cleared it up. I still don't see the need of gathering the color attribute TWICE. (does anyone have ANY idea about it? )

I dont think the second call to GetConsoleScreenBufferInfo() is neccessary.