https://gcc.gnu.org/bugzilla/show_bug.cgi?id=38308
--- Comment #7 from jerome.h...@planete-sciences.org --- I know nothing of the inner workings of gcc, but what I can imagine, is that to check a wchar_t string is easier than a char string, because some encodings may use variable length multibyte encodings that might, when looked as individual bytes, overlap the ASCII alphabet character encoding (UTF-8 does not, but I suspect others might), but wchar_t should not, i.e. 'd' can only be equal to L'd'. So I don't see why checking for wchar_t is not done (but I must miss a big point).
