Trying to convince GCC to output its UTF-8 input source files

Question

I'm currently on GCC Cygwin trying to output my French translated content. The input source string, encoded in a UTF-8 file using Eclipse, and verified in Notepad as UTF-8 encoded is:

wchar_t **strings;
...
strings[NL_ARGUMENT_DESCRIPTION_ADD_METADATA] =
        L"Ajouter du contenu de métadonnées.";

compiles just fine under GCC 11.2.0 with the -finput-charset=UTF-8. The resulting output ends up messed up as:

Ajouter du contenu de m▒tadonn▒es.

Unless I pipe this to iconv --from-code=ISO-8859-1 --to-code=UTF-8 where the output ends up looking correct as:

Ajouter du contenu de métadonnées.

Is there any way to avoid having to pipe this to iconv as I'm dealing with a GCC program with a CLI and do not want users to have to worry about piping to read their online help? I am using wprintf(L"%S")-style format strings when generating output.

Answer 1

Following the guidance of the answer in difference between L"" and u8""

The following code produces the expected output

$ cat ./prova.c
#include <stdio.h>
#include <wchar.h>
int main(int argc, char **argv)
{ printf(u8"Ajouter du contenu de métadonnées.\n");
return 0; }

$ gcc -o prova prova.c

$ ./prova
Ajouter du contenu de métadonnées.

On cygwin it also work the following

$ cat prova.c
#include <stdio.h>
int main(int argc, char **argv)
{ printf("Ajouter du contenu de métadonnées.\n");
return 0; }

$ gcc -std=c99 -Wall -o prova prova.c

$ ./prova.exe
Ajouter du contenu de métadonnées.