X-Git-Url: http://nsz.repo.hu/git/?a=blobdiff_plain;f=lexer.c;h=10e049f039af7aebea79d9c74acf057882475ea8;hb=2a568dda2ebe09066e0a71f22b22aa9488881aa4;hp=10bca9d9c72f6239d7c0baeefa697b9d0dd8ec7c;hpb=43e48b16faa9d8bc68a25656648842f90f9977c6;p=cparser

diff --git a/lexer.c b/lexer.c
index 10bca9d..10e049f 100644
--- a/lexer.c
+++ b/lexer.c
@@ -43,7 +43,7 @@
 //#define DEBUG_CHARS
 #define MAX_PUTBACK 3
 
-#ifdef _WIN32
+#if defined(_WIN32) || defined(__CYGWIN__)
 /* No strtold on windows and no replacement yet */
 #define strtold(s, e) strtod(s, e)
 #endif
@@ -65,7 +65,7 @@ bool               allow_dollar_in_symbol = true;
  */
 static void parse_error(const char *msg)
 {
-	errorf(&lexer_token.source_position,  "%s", msg);
+	errorf(&lexer_token.source_position, "%s", msg);
 }
 
 /**
@@ -75,7 +75,7 @@ static void parse_error(const char *msg)
  */
 static NORETURN internal_error(const char *msg)
 {
-	internal_errorf(&lexer_token.source_position,  "%s", msg);
+	internal_errorf(&lexer_token.source_position, "%s", msg);
 }
 
 static inline void next_real_char(void)
@@ -96,7 +96,7 @@ static inline void next_real_char(void)
 		bufpos = buf + MAX_PUTBACK;
 		bufend = buf + MAX_PUTBACK + s;
 	}
-	c = *bufpos++;
+	c = (unsigned char)*bufpos++;
 }
 
 /**
@@ -304,60 +304,78 @@ end_symbol:
 
 static void parse_integer_suffix(bool is_oct_hex)
 {
-	bool is_unsigned  = false;
-	bool min_long     = false;
-	bool min_longlong = false;
-
-	if(c == 'U' || c == 'u') {
-		is_unsigned = true;
+	bool is_unsigned     = false;
+	bool min_long        = false;
+	bool min_longlong    = false;
+	bool not_traditional = false;
+	int  pos             = 0;
+	char suffix[4];
+
+	if (c == 'U' || c == 'u') {
+		not_traditional = true;
+		suffix[pos++]   = toupper(c);
+		is_unsigned     = true;
 		next_char();
-		if(c == 'L' || c == 'l') {
+		if (c == 'L' || c == 'l') {
+			suffix[pos++] = toupper(c);
 			min_long = true;
 			next_char();
-			if(c == 'L' || c == 'l') {
+			if (c == 'L' || c == 'l') {
+				suffix[pos++] = toupper(c);
 				min_longlong = true;
 				next_char();
 			}
 		}
-	} else if(c == 'l' || c == 'L') {
+	} else if (c == 'l' || c == 'L') {
+		suffix[pos++] = toupper(c);
 		min_long = true;
 		next_char();
-		if(c == 'l' || c == 'L') {
-			min_longlong = true;
+		if (c == 'l' || c == 'L') {
+			not_traditional = true;
+			suffix[pos++]   = toupper(c);
+			min_longlong    = true;
 			next_char();
-			if(c == 'u' || c == 'U') {
-				is_unsigned = true;
+			if (c == 'u' || c == 'U') {
+				suffix[pos++] = toupper(c);
+				is_unsigned   = true;
 				next_char();
 			}
-		} else if(c == 'u' || c == 'U') {
-			is_unsigned = true;
+		} else if (c == 'u' || c == 'U') {
+			not_traditional = true;
+			suffix[pos++]   = toupper(c);
+			is_unsigned     = true;
 			next_char();
 			lexer_token.datatype = type_unsigned_long;
 		}
 	}
 
-	if(!is_unsigned) {
+	if (warning.traditional && not_traditional) {
+		suffix[pos] = '\0';
+		warningf(&lexer_token.source_position,
+			"traditional C rejects the '%s' suffix", suffix);
+	}
+	if (!is_unsigned) {
 		long long v = lexer_token.v.intvalue;
-		if(!min_long) {
-			if(v >= TARGET_INT_MIN && v <= TARGET_INT_MAX) {
+		if (!min_long) {
+			if (v >= TARGET_INT_MIN && v <= TARGET_INT_MAX) {
 				lexer_token.datatype = type_int;
 				return;
-			} else if(is_oct_hex && v >= 0 && v <= TARGET_UINT_MAX) {
+			} else if (is_oct_hex && v >= 0 && v <= TARGET_UINT_MAX) {
 				lexer_token.datatype = type_unsigned_int;
 				return;
 			}
 		}
-		if(!min_longlong) {
-			if(v >= TARGET_LONG_MIN && v <= TARGET_LONG_MAX) {
+		if (!min_longlong) {
+			if (v >= TARGET_LONG_MIN && v <= TARGET_LONG_MAX) {
 				lexer_token.datatype = type_long;
 				return;
-			} else if(is_oct_hex && v >= 0 && (unsigned long long)v <= (unsigned long long)TARGET_ULONG_MAX) {
+			} else if (is_oct_hex && v >= 0 && (unsigned long long)v <= (unsigned long long)TARGET_ULONG_MAX) {
 				lexer_token.datatype = type_unsigned_long;
 				return;
 			}
 		}
 		unsigned long long uv = (unsigned long long) v;
-		if(is_oct_hex && uv > (unsigned long long) TARGET_LONGLONG_MAX) {
+		if (is_oct_hex && uv > (unsigned long long) TARGET_LONGLONG_MAX) {
 			lexer_token.datatype = type_unsigned_long_long;
 			return;
 		}
@@ -365,11 +383,11 @@ static void parse_integer_suffix(bool is_oct_hex)
 		lexer_token.datatype = type_long_long;
 	} else {
 		unsigned long long v = (unsigned long long) lexer_token.v.intvalue;
-		if(!min_long && v <= TARGET_UINT_MAX) {
+		if (!min_long && v <= TARGET_UINT_MAX) {
 			lexer_token.datatype = type_unsigned_int;
 			return;
 		}
-		if(!min_longlong && v <= TARGET_ULONG_MAX) {
+		if (!min_longlong && v <= TARGET_ULONG_MAX) {
 			lexer_token.datatype = type_unsigned_long;
 			return;
 		}
@@ -383,11 +401,19 @@ static void parse_floating_suffix(void)
 	/* TODO: do something useful with the suffixes... */
 	case 'f':
 	case 'F':
+		if (warning.traditional) {
+			warningf(&lexer_token.source_position,
+				"traditional C rejects the 'F' suffix");
+		}
 		next_char();
 		lexer_token.datatype = type_float;
 		break;
 	case 'l':
 	case 'L':
+		if (warning.traditional) {
+			warningf(&lexer_token.source_position,
+				"traditional C rejects the 'F' suffix");
+		}
 		next_char();
 		lexer_token.datatype = type_long_double;
 		break;
@@ -818,11 +844,15 @@ static int parse_escape_sequence(void)
 	case EOF:
 		parse_error("reached end of file while parsing escape sequence");
 		return EOF;
+	/* \E is not documented, but handled, by GCC.  It is acceptable according
+	 * to §6.11.4, whereas \e is not. */
+	case 'E':
 	case 'e':
 		if (c_mode & _GNUC)
-			return 27;   /* hopefully 27 is ALWAYS the code for ESACAPE */
-		/*fallthrough*/
+			return 27;   /* hopefully 27 is ALWAYS the code for ESCAPE */
+		/* FALLTHROUGH */
 	default:
+		/* §6.4.4.4:8 footnote 64 */
 		parse_error("unknown escape sequence");
 		return EOF;
 	}
@@ -840,6 +870,10 @@ string_t concat_strings(const string_t *const s1, const string_t *const s2)
 	memcpy(concat, s1->begin, len1);
 	memcpy(concat + len1, s2->begin, len2 + 1);
 
+	if (warning.traditional) {
+		warningf(&lexer_token.source_position,
+			"traditional C rejects string constant concatenation");
+	}
 #if 0 /* TODO hash */
 	const char *result = strset_insert(&stringset, concat);
 	if(result != concat) {
@@ -866,6 +900,10 @@ wide_string_t concat_string_wide_string(const string_t *const s1, const wide_str
 		concat[i] = src[i];
 	}
 	memcpy(concat + len1, s2->begin, (len2 + 1) * sizeof(*concat));
+	if (warning.traditional) {
+		warningf(&lexer_token.source_position,
+			"traditional C rejects string constant concatenation");
+	}
 
 	return (wide_string_t){ concat, len1 + len2 + 1 };
 }
@@ -881,6 +919,10 @@ wide_string_t concat_wide_strings(const wide_string_t *const s1, const wide_stri
 	wchar_rep_t *const concat = obstack_alloc(&symbol_obstack, (len1 + len2 + 1) * sizeof(*concat));
 	memcpy(concat,        s1->begin, len1       * sizeof(*concat));
 	memcpy(concat + len1, s2->begin, (len2 + 1) * sizeof(*concat));
+	if (warning.traditional) {
+		warningf(&lexer_token.source_position,
+			"traditional C rejects string constant concatenation");
+	}
 
 	return (wide_string_t){ concat, len1 + len2 + 1 };
 }
@@ -895,9 +937,14 @@ wide_string_t concat_wide_string_string(const wide_string_t *const s1, const str
 
 	wchar_rep_t *const concat = obstack_alloc(&symbol_obstack, (len1 + len2 + 1) * sizeof(*concat));
 	memcpy(concat, s1->begin, len1 * sizeof(*concat));
-	const char *const src = s2->begin;
+	const char  *const src = s2->begin;
+	wchar_rep_t *const dst = concat + len1;
 	for (size_t i = 0; i != len2 + 1; ++i) {
-		concat[i] = src[i];
+		dst[i] = src[i];
+	}
+	if (warning.traditional) {
+		warningf(&lexer_token.source_position,
+			"traditional C rejects string constant concatenation");
 	}
 
 	return (wide_string_t){ concat, len1 + len2 + 1 };
@@ -1135,7 +1182,7 @@ end_of_char_constant:;
 	lexer_token.type           = T_CHARACTER_CONSTANT;
 	lexer_token.v.string.begin = string;
 	lexer_token.v.string.size  = size;
-	lexer_token.datatype       = type_int;
+	lexer_token.datatype       = c_mode & _CXX && size == 1 ? type_char : type_int;
 }
 
 /**
@@ -1150,7 +1197,10 @@ static void skip_multiline_comment(void)
 		case '/':
 			next_char();
 			if (c == '*') {
-				/* TODO: nested comment, warn here */
+				/* nested comment, warn here */
+				if (warning.comment) {
+					warningf(&lexer_token.source_position, "'/*' within comment");
+				}
 			}
 			break;
 		case '*':
@@ -1192,6 +1242,15 @@ static void skip_line_comment(void)
 		case '\r':
 			return;
 
+		case '\\':
+			next_char();
+			if (c == '\n' || c == '\r') {
+				if (warning.comment)
+					warningf(&lexer_token.source_position, "multi-line comment");
+				return;
+			}
+			break;
+
 		default:
 			next_char();
 			break;