Printing Copyright Symbol in Visual Studio 2010

Question

I have been struggling to print Copyright symbol in Windows using Visual Studio. I understand that 0xA9 is the ASCII code for Copyright symbol and it works on non-windows platform. But on Windows I can't print the Copyright symbol using the same code.

#include "iostream.h"
using namespace std;


int main(int argc, char * argv[])
{

        cout << (char)0xA9 << " Copyright symbol" << endl;

        return 0;
}

Output on Linux/HP-UX and AIX: © Copyright symbol

Output on Windows: ⌐ Copyright symbol

I am new to Windows, can someone help me out.

Solution 1

As Basile points out, the copyright symbol (©) is not an ASCII character. In other words, it is not one of the characters included in the 7-bit ASCII character set.

You need to switch to a Unicode encoding in order to use "special" characters like this that extend beyond the range of 7-bit ASCII. That's not difficult in Windows, it just requires that you use wide characters (wchar_t) instead of narrow characters (char). Unlike most Unix-based systems that implement Unicode support using UTF-8 (which uses the regular char data type), Windows does not have built-in support for UTF-8. It uses UTF-16 instead, which requires that you use the larger wchar_t type.

Conveniently, the C++ standard library also supports wide character strings, you just need to use the appropriate versions of the classes. The ones you want will have a w appended to the front of their names. So, rewriting your code to use wide (Unicode) characters on Windows, it would look like this:

#include <iostream>  // (standard C++ headers should be in angle brackets)

int main(int argc, char * argv[])
{
    std::wcout << (wchar_t)0xA9 << " Copyright symbol" << std::endl;
    return 0;
}

The reason you're getting that strange ⌐ character when you try the original code on Windows is that that character is what the value 0xA9 maps to in your default Windows character set. You see, the char type supports 8-bit values, but I said above that the ASCII character set only defines 7 bits worth of characters. That extra bit is used on Windows to define some additional useful characters.

There are two different sets of extended narrow (non-Unicode) character sets, one is called the OEM character set and the other is (for historical reasons) called the ANSI character set. Generally, the Command Prompt uses the OEM character set, which fills most of the upper range with characters for drawing lines, boxes, and other simulated graphics in a text-based environment. Legacy, non-Unicode Windows applications generally use the ANSI character set, which is specific to your localized version of Windows and fills the upper range with characters needed to display all of the letters/symbols in your language.

If it sounds complicated, that's because it is. That's why everyone has forgotten all of this stuff and uses exclusively Unicode on Windows. I strongly recommend that path to you as well. :-)

Edit: Nuts, I forgot it was more complicated than this. Changing your code to output wide characters may not be sufficient. The Windows Command Prompt is broken backwards-compatible in all sorts of ways, severely hobbling its support for Unicode characters.

By default, it uses raster fonts which probably don't even provide symbols for most of the Unicode characters (the copyright symbol is likely common enough to be an exception). You need to change the font used by the Command Prompt to something else like Lucida Console or Consolas to ensure that it works correctly. Fortunately, you can set the defaults for all Command Prompt windows. Unfortunately, this is a per-user setting.

Additionally, the Command Prompt still uses the active code page, so all of that stuff I was explaining above is still relevant and you can't forget about it. You can change the particular code page that it uses with the chcp xxxx command, where xxxx is the number of the code page you wish to use. Unfortunately, this applies only to the current console session and must be reset each time. Not a good solution for an application program that needs to output Unicode characters.

More information on these problems and how to output Unicode strings on the Command Prompt is available in the answers to these questions:

• What encoding/code page is cmd.exe using?
• Can command prompt display unicode characters?

Solution 2

Notice that 0xa9 is not ASCII (which had 7 bits characters, limited to 0 - 0x7f range). It could be ISO/IEC 8859-1. Many current systems (including most Linux terminals today) use UTF-8 these days, in which the copyright glyph is encoded by two bytes, so you would code "\302\251" or "\xc2\xa9" in your C or C++ source. So your program don't display a copyright sign in my Linux xfce4-terminal which uses UTF-8.

Some Windows machines had different encoding systems.

I would setup your system (be it Linux or Windows) to use UTF8 character encoding, if possible, on its terminal (or use UTF16 wide chars). Read about UTF-8 everywhere.

An ASCII conventional evocation of copyright is very commonly (C) precisely because the ASCII encoding does not have any copyright glyph.