C - printing greek letters in linux terminal - c

Question

I have this paper, on C language, which requires some greek sentences to be printed in the terminal.
In the code template that is given to us there is this line of code:
system("chcp 1253>nul");
This is supposed to print the greek characters.
In my Ubuntu Terminal I see:
�������� ����� �� ����� ����� ��� �������� ���� ������
So, how can I print greek characters in my terminal?

Answer 1

This is supported out of the box in most Linuxes. The only thing one must do is use
setlocale(LC_ALL, "");
in the beginning of the program. This relies on the fact that UTF-8 is the default choice of encoding for users' locales. The standard says that this call switches to user's current locale. The default is to use the "C" locale which may or may not support national characters.
By default gcc interprets the source code as encoded in UTF-8. Compile-time options exist to change that, but it is recommendedd to keep everything in UTF-8 on Linux. Sources that come from Windows are probably not encoded in UTF-8 and need to be recoded. Use the iconv utility for that;l. If the source is associated with a particular legacy code page, try that code page name as the source encoding.

Answer 2

A C program (comforming to ISO C99 or later, or POSIX.1 or later) that inputs or outputs non-ASCII text should use wide strings, wide I/O, and localization.
For example:
#include <stdlib.h>
#include <locale.h>
#include <stdio.h>
#include <wchar.h>
int main(void)
{
/* Tell the C library to use the current locale settings. */
setlocale(LC_ALL, "");
/* Standard output is used with the wide I/O functions. */
fwide(stdout, 1);
/* Print some Greek text. */
wprintf(L"Γειά σου Κόσμε!\n");
return EXIT_SUCCESS;
}
Note that wide string literals are written using L"..." whereas normal (ASCII or narrow) string literals as "...". Similarly, wide character constants (of type wint_t) are written with the L prefix; for example, L'€'.
When compiling, you do need to tell the compiler what character set the source code uses. In Linux, GCC uses the locale settings, but also provides an option -finput-charset=windows-1252 to change it to Windows Western European, for example.
Rather than fiddle with the flags, I recommend you write a Bash helper script, say to-utf8:
#!/bin/bash
if [ $# -lt 2 ] || [ ":$1" = ":-h" ] || [ ":$1" = ":--help" ]; then
printf '\n'
printf 'Usage: %s [ -h | --help ]\n' "$0"
printf ' %s windows-1252 file.c [ ... ]\n' "$0"
printf '\n'
exit 0
fi
charset="$1"
shift 1
Work=$(mktemp) || exit 1
trap "rm -f '$Work'" EXIT
for src in "$#" ; do
iconv -f "$charset//TRANSLIT" -t UTF-8 "$src" > "$Work" || exit $?
sed -e 's|\r$||' "$Work" > "$src" || exit $?
printf '%s: Converted successfully.\n' "$src"
done
exit 0
If you want, you can install that system-wide using
sudo install -o 0 -g 0 -m 0755 to-utf8 /usr/bin/
The first command-line parameter is the source character set (use iconv --list to see them all), followed by a list of files to fix.
The script creates an automatically deleted temporary file. The iconv line converts the character set of each file to UTF-8, saving the result into the temporary file. The sed file changes any CRLF (\r\n) newlines to LF (\n), overwriting the contents of the file.
(Rather than use a second temporary file to hold the contents, having sed to direct its output to the original file, means the original file keeps its owner and group intact.)