/* $NetBSD: lex_char.c,v 1.7 2023/03/28 14:44:34 rillig Exp $ */ # 3 "lex_char.c" /* * Tests for lexical analysis of character constants. * * C99 6.4.4.4 "Character constants" */ /* lint1-extra-flags: -X 351 */ void sink(char); void test(void) { /* expect+1: error: empty character constant [73] */ sink(''); sink('a'); sink('\0'); /* UTF-8 */ /* expect+2: warning: multi-character character constant [294] */ /* expect+1: warning: conversion of 'int' to 'char' is out of range, arg #1 [295] */ sink('รค'); /* GCC extension */ /* expect+1: warning: dubious escape \e [79] */ sink('\e'); /* since C99 */ sink('\x12'); /* octal */ sink('\177'); /* expect+1: error: empty character constant [73] */ sink(''); /* U+0007 alarm/bell */ sink('\a'); /* U+0008 backspace */ sink('\b'); /* U+0009 horizontal tabulation */ sink('\t'); /* U+000A line feed */ sink('\n'); /* U+000B vertical tabulation */ sink('\v'); /* U+000C form feed */ sink('\f'); /* U+000D carriage return */ sink('\r'); } /* * The sequence backslash-newline is handled in an early stage of * translation (C90 5.1.1.2 item 2, C99 5.1.1.2 item 2, C11 5.1.1.2 item 2), * which allows it in character literals as well. This doesn't typically * occur in practice though. */ char ch = '\ \ \ \ \ x'; |