/* $NetBSD: lex_wide_string.c,v 1.5 2024/02/02 22:45:48 rillig Exp $ */ # 3 "lex_wide_string.c" /* * Test lexical analysis of wide string constants. * * C99 6.4.5 "String literals" */ /* lint1-extra-flags: -X 351 */ void sink(const int *); void test(void) { sink(L""); sink(L"hello, world\n"); sink(L"\0"); sink(L"\0\0\0\0"); /* expect+1: error: no hex digits follow \x [74] */ sink(L"\x"); /* expect+1: warning: dubious escape \y [79] */ sink(L"\y"); sink(L"first" L"second"); /* expect+1: error: cannot concatenate wide and regular string literals [292] */ sink(L"wide" "plain"); } /* * Since lint always runs in the default "C" locale, it does not support any * multibyte character encoding, thus treating each byte as a separate * character. If lint were to support UTF-8, the array dimension would be 3 * instead of 7. */ /* expect+1: error: negative array dimension (-7) [20] */ typedef int mblen[-(int)(sizeof(L"Ä😄") / sizeof(L""))];